INDEX TO RECORD OF TECHNICAL INFORMATION/DOCUMENTS
SUPPORTING THE PROPOSED RULE TO AMEND 40 CFR PART 136
TO INCORPORATE METHOD 1613
Page
Volume No.
I. Method 1613 and Method Performance Studies
A. Method 1613: Tetra- through Octa- I-A 001
Chlorinated Dioxins and Furans by Isotope
Dilution HRGC/HRMS, USEPA, Office of Water
Regulations and Standards (OWRS),
Industrial Technology Division (ITD),
Revision A, April 1990.
B. Performance Evaluation of Method 1613, I-B 045
USEPA, OWRS, ITD, March 1990.
C. Study Plan for the Evaluation of Method I-D 202
1613, USEPA, OWRS, ITD, May 1990.
D. Summary Report, USEPA ITD, Method Detection I-C 231
Limit Study for Method 1613 Determination
of 2,3,7,8-TCDD and 2,3,7,8-TCDF, USEPA,
OWRS, ITD, May 1990.
II. Sources Referenced in Method 1613
A. Method 8290: Analytical Procedures and II-A 240
Quality Assurance for Multimedia Analysis
of Polychlorinated Dibenzo-p-dioxins and
Dibenzofurans by High-Resolution Gas
Chromatography/High-Resolytion Mass
Spectrometry, prepared for Environmental
Monitoring Systems Laboratory-Las Vegas,
USEPA, by Yves Tondeur, June 1987.
B. Measurement of 2,3,7,8-Tetrachlorinated II-B 372
Dibenzo-p-dioxin (TCDD) and 2,3,7,8-
Tetrachlorinated Dibenzofuran (TCDF) in
Pulp, Sludges, Process Samples and
Wastewaters from Pulp and Paper Mills, by
Wright State University, Dayton, Ohio, June
1988.
C. NCASI Procedures for the Preparation and II-C 414
Isomer Specific Analysis of Pulp and Paper
Industry Samples for 2,3,7,8-TCDD and
2,3,7,8- TCDF, by National Council of the
Paper Industry for Air and Stream
Improvement, New York, Technical Bulletin
No. 551, May 1989.
-------
INDEX TO RECORD OF TECHNICAL INFORMATION/DOCUMENTS
SUPPORTING THE PROPOSED RULE TO AMEND 40 CFR PART 136
TO INCORPORATE METHOD 1613 (CONT.)
Page
Volume No.
D. Analytical Procedures and Quality Assurance II-D 476
Plan for the Determination of PCDD/PCDF in
Fish, by Environmental Research Laboratory,
USEPA, April 1988.
E. Determination of Tetra-, Hexa-, Hepta-, and II-E 520
Octachlorodibenzo-p-dioxin Isomers in
Particulate Samples at Parts per Trillion
Levels, by L.L. Lamparski and T.J.
Nestrick, Analytical Chemistry. Volume 52:
2045-2054, 1980.
F. Novel Extraction Device for the II-F 530
Determination of Chlorinated Dibenzo-p-
dioxins (PCDDs) and Dibenzofurans (PCDFs)
in Matrices Containing Water,by L.L.
Lamparski and T.J. Nestrick, Chemosphere,
Volume 19: 27-31, 1989.
G. Control of Interferences in the Analysis of II-G 535
Human Adipose Tissue for 2,3,7,8-Tetra-
chlorodibenzo-p-dioxin, by D.G. Patterson,
et. al., Environmental Toxicological
Chemistry. Volume 5: 355-360, 1986.
H. Protocol for the Analysis of 2,3,7,8- II-H 541
Tetrachlorodibenzo-p-dioxin by High-
Resolution Gas Chromatography/High-
Resolution Mass Spectrometry, by John S.
Stanley and Thomas M. Sack, Environmental
Monitoring Systems Laboratory-Las Vegas,
USEPA, EPA 600/4-86-004, January 1986.
I. Method 613 -- 2,3,7,8-Tetrachlorodibenzo-p- II-I 697
dioxin, Section 4.1, 49 Federal Register
43234, October 26, 1984.
J. Interpretation of Percent Recovery Data, by II-J 703
L. P. Provost and R.S. Elder, American
Laboratory. Volume 15: 56-83, 1983.
ii 11/90
-------
Method 1613 Revision A October 1990
Tetra- through Octa- Chlorinated Dioxins and Furans
by Isotope Dilution HRGC/HRMS
1 SCOPE AND APPLICATION
1.1 This method is designed to meet the survey
requirements of the USEPA ITD. The method
is used to determine the tetra- through
octa- chlorinated dibenzo-p-dioxins and
dibenzofurans associated with the Clean
Water Act (as amended 1987); the Resource
Conservation and Recovery Act (as amended
1986); and the Comprehensive Environmental
Response, Compensation and Liability Act
(as amended 1986); and other dioxin and
furan compounds amenable to high resolu-
tion capillary column gas chromatography
(HRGC)/high resolution mass spectrometry
(HRMS). Specificity is provided for de-
termination of the 17 2,3,7,8-substituted
polychlorinated dibenzo-p-dioxins (PCDD)
and polychlorinated dibenzofurans (PCOF).
1.2 The method is based on EPA, industry, com-
mercial laboratory, and academic methods
(References 1-6).
1.3 The compounds listed in Table 1 may be
determined in waters, soils, sludges, and
other matrices by this method.
1.4 The detection limits of the method are
usually dependent on the level of inter-
ferences rather than instrumental limita-
tions. The levels in Table 2 typify the
minimum quantities that can be determined
in environmental samples using the method.
1.5 The GCHS portions of the method are for
use only by analysts experienced with
HRGC/HRMS or under the close supervision
of such qualified persons. Each labora-
tory that uses this method must demon-
strate the ability to generate acceptable
results using the procedure in Section
8.2.
2 SUMMARY OF METHOD
2.1 Stable isotopicaUy labeled analogs of 15
of the PCDDs and PCOFs are added to each
sample prior to extraction. Samples con-
taining coarse solids are prepared for
extraction by grinding or homogenization.
Water samples are filtered and then
extracted with methylene chloride using
separatory funnel procedures; the particu-
lates from the water samples, soils, and
other finely divided solids are extracted
using a combined Soxhlet extraction/Dean-
Stark azeotropic distillation (Reference
7). Prior to cleanup and analysis, the
extracts of the filtered water and the
particulates are combined.
2.2 After extraction, 37C14-labeled 2,3,7,8-
TCDD is added to each extract to measure
the efficiency of the cleanup process.
Samples cleanup may include back extrac-
tion with acid and/or base, and gel perme-
ation, alumina, silica gel, and activated
carbon chromatography. High performance
liquid chromatography (HPLC) can be used
for further isolation of the 2,3,7,8-
isomers or other specific isomers or
congeners.
2.3 After cleanup, the extract is concentrated
to near dryness. Immediately prior to
injection, two internal standards are
added to each extract, and 'a 1 uL aliquot
of the extract is injected into the gas
chromatograph. The analytes are separated
by the GC and detected by a high resolu-
tion (>10,000) mass spectrometer. Two
exact masses (m/z's) are monitored for
each analyte. The isotopicaUy labeled
compounds serve to correct for the
variability of the analytical technique.
2.4 Dioxins and furans are identified by
comparing GC retention times and the ion
abundance ratios of the m/z's with the
corresponding retention time ranges of
authentic standards and the theoretical
ion abundance ratios of the exact m/z's;
Isomers and congeners are identified when
the retention times and m/z abundance
ratios agree within pre-defined limits..
By using a GC column or columns capable of
resolving the 2,3,7,8-substituted isomers
from all other tetra- isomers, the
2,3,7,8-substituted isomers are identified
when the retention time and m/z abundance
ratios agree within pre-defined limits o*f
the retention times and exact m/z ratios
of authentic standards.
2.5 Quantitative analysis is performed by GCMS
using selected ion current profile (SICP)
areas, in one of two ways.
2.5.1 For the 15 2,3,7,8-substituted isomers for
which labeled analogs are available (see
001
-------
Table 1), the GCMS system is calibrated
and the compound concentration is deter-
mined using an isotope dilution technique.
Although a labeled analog of the octa-
chlorinated dibenzofuran (OCDF) is avail-
able, using high resolution mass spectrom-
etry it produces an m/z that may interfere
with the identification and quantitat ion
of the unlabeled octachlorinated dibenzo-
p-dioxin (OCDD). Therefore, this labeled
analog has not been included in the cali-
bration standards, and the un labeled OCDF
is quantitared against the labeled OCDO.
Because the labeled analog of 1,2,3,7,8,9-
HxCOD is used as an internal standard
(i.e., not added before extraction of the
sample), it cannot be used to quantitate
the un labeled compound by strict isotope
dilution procedures. Therefore, the unla-
beled 1,2,3,7,8,9-HxCDD is quantitated
using the average of the responses of the
labeled analogs of the other two 2,3,7,8-
substituted HxCDO's, 1,2,3,4,7,8-HxCDO and
1,2,3,6,7,8-HxCDD. As a result, the con-
centration of the unlabeled 1,2,3,7,8,9-
HxCDD is corrected for the average
recovery of the other two HxCDD's.
2.5.2 For non-2,3,7,8-substituted isomers and
the total concentrations of all isomers
within a level of chlorination (i.e.,
total TCDD), concentrations are determined
using response factors from the calibra-
tion of labeled analogs at the same level
of chlorination.
2.6 The quality of the analysis is assured
through reproducible calibration and test-
ing of the extraction, cleanup, and GCMS
systems.
3 CONTAMINATION AND INTERFERENCES
3.1 Solvents, reagents, glassware, and other
sample processing hardware may yield arti-
facts and/or elevated baselines causing
misinterpretation of chromatograms (Ref-
erences 8-9). Specific selection of
reagents and purification of solvents by
distillation in all-glass systems may be
required. Where possible, reagents are
cleaned by extraction or solvent rinse.
3.2 Proper cleaning of glassware is extremely
important because glassware may not only
contaminate the samples, but may also
remove the analytes of interest by
adsorption on the glass surface.
3.2.1 Glassware should be rinsed with solvent
and washed with a detergent solution as
soon after use as is practical. Sonica-
tion of glassware containing a detergent
solution for approximately 30 seconds may
aid in cleaning. Glassware with removable
parts, particularly separatory funnels
with teflon stopcocks, must be
disassembled prior to detergent washing.
3.2.2 After detergent washing, glassware should
be immediately rinsed first with methanol,
then with hot tap water. The tap water
rinse is followed by another methanol
rinse, then acetone, and then methylene
chloride.
3.2.3 Do not bake reusable glassware in an oven
as a routine part of cleaning. Baking may
be warranted after particularly dirty
samples are encountered, but should be
minimized, as repeated baking of glassware
may cause active sites on the glass
surface that will irreversibly adsorb
PCDDs/PCDFs.
3.2.4 Immediately prior to use, Soxhlet extrac-
tion glassware should be pre-extracted
with toluene for approximately 3 hours.
See Section 11.1.2.3. Separatory funnels
should be shaken with - methylene
chloride/toluene .(80/20 mixture) for 2
minutes, drained, and then shaken with
pure methylene chloride for 2 minutes.
3.3 All materials used in the analysis shall
be demonstrated to be free from interfer-
ences by running reference matrix blanks
initially and with each sample set
(samples started through the extraction
process on a given 12-hour shift, to a
maximum of 20 samples). The reference
matrix blank must simulate, as closely as
possible, the sample matrix under test.
Reagent water (Section 6.6.1) is used to
simulate water samples; playground sand
(Section 6.6.2) or white quartz sand
(Section 6.3.2) can be used to simulate
soils; filter paper (Section 6.6.3) is
used to simulate papers and similar
materials; other materials (Section 6.6.4)
can be used to simulate other matrices.
3.4 Interferences coextracted from samples
will vary considerably from source to
source, depending on the diversity of the
site being sampled. Interfering compounds
may be present at concentrations several
orders of magnitude higher than the PCDDs
and PCDFs. The most frequently encoun-
tered interferences are chlorinated-
biphenyls, methoxy biphenyls, hydroxy-
-------
diphenyl ethers, benzylphenyl ethers,
polynuclear aromatics, and pesticides.
Because very low levels of PCDDs and PCOFs
are measured by this method, the elimina-
tion of interferences \s essential. The
cleanup steps given in Section 12 can be
used to reduce or eliminate these inter-
ferences and thereby permit reliable
determination of the PCDOs and PCOFs at
the levels shown in Table 2.
3.5 Each piece of reusable glassware should be
numbered in such a fashion that the
laboratory can associate all reusable
glassware with the processing of a parti-
cular sample. This will assist the
laboratory in: 1) tracking down possible
sources of contamination for individual
samples, 21 identifying glassware assoc-
iated with highly contaminated samples
that may require extra cleaning, and 3)
determining when glassware should be
discarded.
4 SAFETY
4.1 The toxicity or careinogenicity of each
compound or reagent used in this method
has not been precisely determined;
however, each chemical compound should be
treated as a potential health hazard.
Exposure to these compounds should be
reduced to the lowest possible level.
4.1.1 The 2,3,7,8-TCDO isomer has been found to
. . be acnegenic, carcinogenic, and terato-
genic in laboratory animal studies. It is
soluble in water to approximately 200 ppt
and in organic solvents to 0.14 percent.
On the basis of the available
toxicological and physical properties of
2,3,7,8-TCDD, all of the PCDDs and PCDFs
should be handled only by highly trained
personnel thoroughly familiar with
handling and cautionary procedures, and
who understand the associated risks.
4.1.2 It is recommended that the laboratory pur-
chase dilute standard solutions of the
analytes in this method. However, if
primary solutions are prepared, they shall
be prepared in a hood, and a NIOSH/MESA
approved toxic gas respirator shall be
worn when high concentrations are handled.
4.2 The laboratory is responsible for main-
taining a current awareness file of OSHA
regulations regarding the safe handling of
the chemicals specified in this method. A
reference file of data handling sheets
should also be made available to all
personnel involved in these analyses.
Additional information on laboratory
safety can be found in References 10-13.
The references and bibliography at the end
of Reference 13 are particularly compre-
hensive in dealing with the general
subject of laboratory safety.
4.3 The PCDOs and PCDFs and samples suspected
to contain these compounds are handled
using essentially the same techniques
employed in handling radioactive or
infectious materials. Well-ventilated,
controlled access laboratories are
required. Assistance in evaluating the
health hazards of particular laboratory
conditions may be obtained from certain
consulting laboratories and from State
Departments of Health or Labor, many of
which have an industrial health service.
The PCDDs and PCOFs are extremely toxic to
laboratory animals. Each laboratory must
develop a strict safety program for
handling the PCDDs and PCDFs. The follow-
ing practices are recommended (References
2 and 14).
4.3.1 Facility -- When finely divided samples
(dusts, soils, dry chemicals) are handled,
all operations (including removal of
samples from sample containers, weighing,
transferring, and mixing), should be
performed in a glove box demonstrated to
be leak tight or in a fume hood demon-
strated to have adequate air flow. Gross
losses to the laboratory ventilation
system must not be allowed. Handling of
the dilute solutions normally used in
analytical and animal work presents no
inhalation hazards except in the case of
an accident.
4.3.2 Protective equipment -- Throwaway plastic
gloves, apron or lab coat, safety glasses
or mask, and a glove box or fume hood
adequate for radioactive work should be
utilized. During analytical operations
which may give rise to aerosols or dusts,
personnel should wear respirators equipped
with activated carbon filters. Eye
protection equipment (preferably full face
shields) must be worn while working with
exposed samples or pure analytical
standards. Latex gloves are commonly used
to reduce exposure of the hands. When
handling samples suspected or known to
contain high concentrations of the PCDDs
or PCDFs, an additional set of gloves can
also be worn beneath the latex gloves.
003
-------
4.3.3 Training -- Workers must be trained in the
proper method of removing contaminated
gloves and clothing without contacting the
exterior surfaces.
4.3.4 Personal hygiene -• Thorough washing of
hands and forearms after each manipulation
and before breaks (coffee, lunch, and
shift).
4.3.5 Confinement -- Isolated work area, posted
with signs, segregated glassware and
tools, plastic absorbent paper on bench
tops.
4.3.6 Effluent vapors -- The effluents of sample
splitters for the gas chromatograph and
roughing pumps on the GC/MS should pass
through either a column of activated char-
coal or be bubbled through a trap contain-
ing oil or high-boiling alcohols.
4.3.7 Waste Handling and Disposal
4.3.7.1 Handling •- Good technique includes
minimizing contaminated waste. Plastic
bag liners should be used in waste cans.
Janitors and other personnel must be
trained in the safe handling of waste.
4.3.7.2 Disposal
4.3.7.2.1 The PCDDs and PCDFs decompose above 800
°C. Low-1 eve I waste such as absorbent
paper, tissues, animal remains, and
plastic gloves may be burned in an appro-
priate incinerator. Gross quantities
(milligrams) should be packaged securely
and disposed through commercial or govern-
mental channels which are capable of
handling extremely toxic wastes.
4.3.7.2.2 Liquid or soluble waste should be dis-
solved in methanol or ethanol and
irradiated with ultraviolet light with a
wavelength greater than 290 nm for several
days. (Use F 40 8L lamps or equivalent.)
Analyze liquid wastes and dispose of the
solutions when the PCDDs and PCDFs can no
longer be detected.
4.3.8 Decontamination
4.3.8.1 Personal Decontamination -- Use any mild
soap with plenty of scrubbing action.
4.3.8.2 Glassware, tools, and surfaces
Chlorothene NU Solvent (Trademark of the
Dow Chemical Company) is the least toxic
solvent shown to be effective. Satis-
factory cleaning may be accomplished by
4.3.9
4.3.10
4.3.11
5
5.1
5.1.1.2
5.1.1.3
5.1.1.4
rinsing with Chlorothene, then washing
with any detergent and water. If glass-
ware is first rinsed with solvent, then
the dish water may be disposed of in the
sewer. Given the cost of disposal, it is
prudent to minimize solvent wastes.
Laundry -- Clothing known to be contami-
nated should be collected in plastic bags.
Persons who convey the bags and launder
the clothing should be advised of the
hazard and trained in proper handling.
The clothing may be put into a washer
without contact if the launderer knows of
the potential problem. The washer should
be run through a cycle before being used
again for other clothing.
Wipe tests -- A useful method of determin-
ing cleanliness of work surfaces and tools
is to wipe the surface with a piece of
filter paper. Extraction and analysis by
"GC can achieve a limit of detection of 0.1
ug per wipe. Less than 0.1 ug per wipe
indicates acceptable cleanliness; anything
higher warrants further cleaning. More
than 10 ug on a wipe constitutes an acute
hazard and requires prompt cleaning before
further use of the equipment or work
space, and indicates that unacceptable
work practices have been employed.
Accidents -- Remove contaminated clothing
immediately, taking precautions not to
contaminate skin or other articles. Wash
exposed skin vigorously and repeatedly
until medical attention is obtained.
for discrete or
APPARATUS AND MATERIALS
Sampli ng equ i pment
composite sampling.
Sample bottles and caps
Liquid samples (waters, sludges and simi-
lar materials containing five percent
solids or less) •• Sample bottle, amber
glass, 1.1 L minimum, with screw cap.
Solid samples {soils, sediments, sludges,
paper pulps, filter cake, compost, and
similar materials that contain more than
five percent solids) -- Sample bottle,
wide mouth, amber glass, 500 mL minimum.
If amber bottles are not available,
samples shall be protected from light.
Bottle caps -- Threaded to fit sample
bottles. Caps shall be lined with Teflon.
004
-------
5.1.1.5 Cleaning
5.1.1.5.1 Bottles are detergent water washed, then
solvent rinsed before use.
5.1.1.5.2 Liners are detergent water washed, then
rinsed with reagent water (Section 6.6.1)
and then solvent, and baked- at
approximately 200 °C for one hour minimum
prior to use.
5.1.2 Compositing equipment -- Automatic or
manual compositing system incorporating
glass containers cleaned per bottle
cleaning procedure above. Glass or Teflon
tubing only shall be used. If the sampler
uses a peristaltic pump, a minimum length
of compressible silicone rubber tubing may
be used in the pump only. Before use, the
tubing shall be thoroughly rinsed with
methanol, followed by repeated rinsings
with reagent water to minimize sample
contamination. An integrating flow meter
is used to collect proportional composite
samples.
5.2 Equipment for glassware cleaning
5.2.1 Laboratory sink with.overhead fume hood
5.3 Equipment for sample preparation
5.3.1 Laboratory fume hood of sufficient size to
contain the sample preparation equipment
listed below
5.3.2 Glove box (optional)
5.3.3 Tissue homogenizer -- VirTis Model 45
Macro homogenizer (American Scientific
Products H-3515, or equivalent) with
stainless steel Macro-shaft and Turbo-
shear blade.
5.3.4 Meat grinder -- Hobart, or equivalent,
with 3-5 mm holes in inner plate.
5.3.5 Equipment for determining percent moisture
5.3.5.1 Oven, capable of maintaining a temperature
of 110 ±5 °C.
5.3.5.2 Oessicator
5.3.6 Balances
5.3.6.1 Analytical -- Capable of weighing 0.1 mg.
5.3.6.2 Top loading -- Capable of weighing 10 mg.
5.4 Extraction apparatus
5.4.1 Water samples
5.4.1.1 pH meter, with combination
electrode.
glass
5.4.1.2 pH paper, wide range (Hydrion Papers, or
equivalent).
5.4.1.3 Graduated cylinder, 1 L capacity
5.4.1.4 1 L filtration flasks with side arm, for
use in vacuum filtration of water samples.
5.4.1.5 Separatory funnels -- 250, 500; and 2000
ml, with Teflon stop cocks.
5.4.2 Soxhlet/Dean-Stark
(Figure 1)
(SOS)
extractor
FIGURE 1 Soxhlet/Dean-Stark Extractor
005
-------
5.4.2.1 Soxhlet -- 50 mm i.d., 200 mL capacity
with 500 mL flask (Cal-Glass LG-6900, or
equivalent, except substitute 500 mL round
bottom flask for 300 mL flat bottom
flask).
5.4.2.2 Thimble -- 43 x 123 to fit Soxhlet (Cal-
Glass LG-6901-122, or equivalent).
5.4.2.3 Moisture trap -- Dean Stark or Barret with
Teflon stopcock, to fit Soxhlet.
5.4.2.4 Heating mantle -- Hemispherical, to fit
500 ml round bottom flask (Cal-Glass LG-
8801-112, or equivalent).
5.4.2.5 Variable transformer -- Powerstat (or
equivalent), 110 volt. 10 amp.
5.4.3 Beakers, 400-500 mL
5.4.4 Spatulas -- Stainless steel
5.5 Filtration apparatus
5.5.1 Pyrex glass wool -- Solvent extracted by
SOS for three hours minimum. (NOTE:
Baking glass wool may cause active sites
that will irreversibly adsorb
PCDOs/PCOFs.)
5.5.2 Glass funnel -- 125-250 mL
5.5.3 Glass fiber filter paper (Whatman GF/D, or
equivalent)
5.5.4 Drying column -- 15 .to 20 mm i.d. Pyrex
chromatographic column equipped with
coarse glass frit or glass wool plug.
5.5.5 Buchner funnel, 15 cm.
5.5.6 Glass fiber filter paper for above.
5.5.7 Pressure filtration apparatus -- Millipore
YT30 142 HW, or equivalent.
5.6 Centrifuge apparatus
5.6.1 Centrifuge -- Capable of rotating 500 mL
centrifuge bottles or 15 mL centrifuge
tubes at 5,000 rpm minimum
5.6.2 Centrifuge bottles -- 500 mL, with screw
caps, to f\t centrifuge
5.6.3 Centrifuge tubes -- 12-15 mL, with screw
caps, to fit centrifuge
5.7 Cleanup apparatus
5.7.1 Automated gel permeation chromatograph
(Analytical Biochemical Labs, Inc,
Columbia, HO, Model GPC Autoprep 1002, or
equivalent).
5.7.1.1 Column -- 600-700 mm x 25 mm i.d., packed
with 70 g of SX-3 Bio-beads (Bio-Rad
Laboratories, Richmond, CA, or
equivalent).
5.7.1.2 Syringe, 10 mL, with Luer fitting.
5.7.1.3 Syringe filter holder, stainless steel,
and glass fiber or Teflon filters (Gelman
4310, or equivalent).
5.7.1.4 UV detectors -- 254-nm, preparative or
semi-prep flow cell: (Isco, Inc., Type 6;
Schmadzu, 5 mm path length; Beckman-Altex
152U, 8 uL micro-prep flow cell, 2 mm
path; Pharmacia UV-1, 3 mm flow cell; LDC
Milton-Roy UV-3, monitor #1203; or
equivalent).
5.7.2 Reverse phase high performance liquid
chromatograph
5.7.2.1 Column oven and detector -- Perkin-Elmer
Model LC-65T (or equivalent) operated at
0.02 AUFS at 235 nm.
5.7.2.2 Injector -- Rheodyne 7120 (or equivalent)
with 50 uL sample loop.
5.7.2.3 Column -- Two 6.2 x 250 "mm Zorbax-OOS
columns in series (DuPont Instruments
Division, Wilmington, DE, or equivalent),
operated at 50 "C with 2.0 mL/min methanol
isocratic effluent.
5.7.2.4 Pump -- Altex 110A (or equivalent).
5.7.3 Pipets
5.7.3.1 Disposable, Pasteur, 150 mm x 5 mm i.d.
(Fisher Scientific 13-678-6A, or
equivalent).
5.7.3.2 Disposable, serological, 10 mL (6 mm
i.d.).
5.7.4 Chromatographic columns
5.7.4.1 150 mm ,x 8 mm i.d., (Kontes K-420155, or
equivalent) with coarse glass frit or
glass wool plug and 250 mL reservoir.
OOG
-------
5.7.5
5.7.4.2 200 mm x 15 mm i.d., with coarse glass
frit or glass wool plug and 250 ml
reserve i r.
Oven -- For storage of adsorbents, capable
of maintaining a temperature of 130 ±5 °C.
5.8 Concentration apparatus
5.8.1 Rotary evaporator -- Buchi/Brinkman-
American Scientific No. E5045-10 or
equivalent, equipped with a variable
temperature water bath.
5.8.1.1 A vacuum source is required for use of the
rotary evaporator. It must be equipped
with a shutoff valve at the evaporator,
and preferably, have a vacuum gauge.
5.8.1.2 A recirculating water pump and chiller are
recommended, as use of tap water for
cooling the evaporator wastes large
volumes of water and can lead to
inconsistent performance as water
temperatures and pressures vary.
5.8.1.3 Round bottom flasks -- 100 raL and 500 mL
or larger, with ground glass fitting
compatible with the rotary evaporator.
5.8.2 Kuderna-Danish (K-D)
5.8.2.1 Concentrator tube--10mL, graduated (Kontes
K-570050-1025, or equivalent) with
calibration verified. Ground glass
stopper (size 19/22 joint) is used to
prevent evaporation of extracts.
5.8.2.2 Evaporation flask--500 mL (Kontes K-
570001-0500, or equivalent), attached to
concentrator tube with springs (Kontes K-
662750-0012).
5.8.2.3 Snyder column--three ball macro (Kontes K-
503000-0232, or equivalent).
5.8.2.4 Boiling chips
5.8.2.4.1 Glass or silicon carbide--approx 10/40
mesh, extracted with rnethylene chloride
and baked at 450 °C for one h minimum.
5.8.2.4.2 Teflon (optional)--extracted
methylene chloride.
with
5.8.2.5 Water bath—heated, with concentric ring
cover, capable of maintaining a
temperature within +/- 2 °C, installed in
a fume hood.
5.8.3 Nitrogen blowdown apparatus -- Equipped
with water bath controlled at 35-40 °C (N-
Evap, Organomation Associates, Inc., South
Berlin, MA, or equivalent), installed in a
fume hood.
5.8.4 Sample vials -- Amber glass, 2-5 mL with
Teflon-lined screw cap.
5.9 Gas chromatograph -- Shall have split less
or on-column injection port for capillary
column, temperature program with
isothermal hold, and shall meet all of the
performance specifications in Section 7.
5.9.1 GC Column for PCDDs and PCDFs and for
isomer specificity for 2,3,7,8-TCDO -- 60
±5 m x 0.32 ±0.02 mm i.d.; 0.25 urn 5%
phenyl, 94% methyl, 1% vinyl silicone
bonded phase fused silica capillary column
(J & U DB-5, or equivalent).
5.9.2 GC Column for isomer specificity for
2,3,7,8-TCDF -- 30 ±5 m x 0.32 ±0.02 mm
i.d.; 0.25 urn bonded phase fused silica
capillary column (J & U 08-225, or
equivalent).
5.10 Mass spectrometer -- 28-40 eV electron
impact ionization, shall be capable of
repetitively selectively monitoring 12
exact m/z's minimum at high resolution
(>10,000) during a period of approximately
1 second, and shall meet all of the
performance specifications in Section 7.
5.11 GCHS interface -- The mass spectrometer
(MS) shall be interfaced to the GC such
that the end of the capillary column
terminates within 1 cm of the ion source
but does not intercept the electron or ion
beams.
5.12 Data system -- Capable of collecting,
recording and storing MS data.
6 REAGENTS AND STANDARDS
6.1 pH adjustment and back extraction
6.1.1 Potassium hydroxide -- Dissolve 20 g
reagent grade KOH in 100 mL reagent water.
6.1.2 Sulfuric acid -- Reagent grade (specific
gravity 1.84).
6.1.3 Sodium chloride -- Reagent grade, prepare
a five percent (w/v) solution in reagent
water.
007
-------
6.2 Solution drying and evaporation
6.2.1 Solution drying -- Sodium sulfate, reagent
grade, granular anhydrous (Baker 3375, or
equivalent), rinsed with methylene
chloride (20 mL/g). baked at 400 °C for
one hour minimum, cooled in a dessicator,
and stored in a pre-cleaned glass bottle
with screw cap that prevents moisture, from
entering. If after heating the sodium
sulfate develops a noticeable grayish cast
(due to the presence of carbon in the
crystal matrix), that batch of reagent is
not suitable for use and should be
discarded. Extraction with methylene
chloride (as opposed to simple rinsing)
and baking at a lower temperature may
produce sodium sulfate that is suitable
for use.
6.2.2 Prepurified nitrogen
6.3 Extraction
6.3.1 Solvents -- Acetone, toluene, cyclohexane,
hexane, nonane, methanol, methylene
chloride, and nonane: distilled-in-glass,
pesticide quality, lot certified to be
free of interferences.
6.3.2 White quartz sand, 60/70 mesh -- For
Soxhlet/Dean-Stark extraction, (Aldrich
Chemical Co, Milwaukee UI Cat No.
27,437-9, or equivalent). Bake at 450 "C
for four hours minimum.
6.4 GPC calibration solution -- Solution
containing 300 mg/mL corn oil, 15 mg/mL
bis(2-ethylhexyl) phthalate, 1.4 mg/mL
pentachlorophenol, 0.1 rag/mL perylene, and
0.5 mg/mL sulfur
6.5 Adsorbents for sample cleanup
6.5.1 Silica gel
6.5.1.1 Activated silica gel -- Bio-Si I A, 100-200
mesh (Bio-Rad 131-1340, or equivalent),
rinsed with methylene chloride, baked at
180 °C for one hour minimum, cooled in a
dessicator, and stored in a pre-cleaned
glass bottle with screw cap that prevents
moisture from entering.
i
6.5.1.2 Acid silica gel (30 percent w/w)
Thoroughly mix 44.0 g of concentrated
sulfuric acid with 100.0 g of activated
silica gel in a clean container. Break up
aggregates with a stirring rod until a
uniform mixture is obtained. Store in a
screw-capped bottle with Teflon-lined cap.
6.5.1.3 Basic silica gel -- Thoroughly nix 30 g of
IN sodium hydroxide with 100 g of
activated silica gel in a clean container.
Break up aggregates with a stirring rod
until a uniform mixture is obtained.
Store in a screw-capped bottle with
Teflon-lined cap.
6.5.2 Alumina -- Either one of two types of
alumina, acid or basic, may be used in the
cleanup of sample extracts, provided that
the laboratory can meet the performance
specifications for the recovery of labeled
compounds described in Section 8.3. The
same type of alumina must be used for all
samples, including those used to
demonstrate initial precision and accuracy
(Section 8.2) and ongoing precision and
accuracy (Section 14.5).
6.5.2.1 Acid alumina -- Bio-Rad Laboratories 132-
1340 Acid Alumina AG 4 (or equivalent).
Activate by heating to 130 °C for 12 hours
minimum.
6.5.2.2 Basic alumina -- Bio-Rad Laboratories 132-
1240 Basic Alumina AG 10 (or equivalent).
Activate by heating to 600 °C for 24 hours
minimum. Alternatively, activate by
heating alumina in a tube furnace at 650-
700 °C under an air flow of approximately
400 cc/min. Do not heat over 700 °C, as
this can lead to reduced capacity for
retaining the analytes. Store at 130 °C
in a covered flask. Use within five days
of baking.
6.5.3 AX-21/Celite
6.5.3.1 Activated carbon -- AX-21 (Anderson
Development Company, Adrian, MI, or
equivalent). Prewash with methanol and
dry in vacuo at 110 "C.
6.5.3.2
6.5.3.3
Celite 545
equivalent).
(Supelco 2-0199, or
Thoroughly mix 5.35 g AX-21 and 62.0 g
Celite 545 to produce a 7.9% w/w mixture.
Activate the mixture at 130 "C for six
hours minimum. Store in a dessicator.
6.6 Reference matrices
6.6.1 Reagent water -- Water in which the PCDDs
and PCDFs and interfering compounds are
not detected by this method.
8
003
-------
6.6.2 High solids reference matrix -- Playground
sand or similar material in which the
PCDDs and PCDFs and interfering compounds
are not detected by this method. May be
prepared by extraction with methylene
chloride and/or baking at '450 °C for four
hours minimum.
6.6.3 Filter paper -- Gelman type A (or equiva-
lent) glass fiber filter paper in which
the PCOOs and PCDFs and interfering con-
pounds are not detected by this method.
Cut the paper to simulate the surface area
of the paper sample being tested.
6.6.4 Other matrices -- This method may be
verified on any matrix by performing the
tests given in Section 8.2. Ideally, the
matrix should be free of the PCODs and
PCDFs, but in no case shall the background
level of the PCDDs and PCDFs in the
reference matrix exceed three times the
minimum levels given in Table 2. If low
background levels of the PCDDs and PCDFs
are present in the reference matrix, the
spike level of the analytes used in
Section 8.2 should be increased to provide
a spike-to-background ratio in the range
of 1/1 to 5/1 (Reference 15).
6.7 Standard solutions -- Purchased as solu-
tions or mixtures with certification to
their purity, concentration, and authen-
ticity, or prepared from materials of
known purity and composition. If compound
purity is 98 percent or greater, the
weight may be used without correction to
compute the concentration of the standard.
When not being used, standards are stored
in the dark at room temperature in screw-
capped vials with Teflon-lined caps. A
mark is placed on the vial at the level of
the solution so that solvent evaporation
loss can be detected. If solvent loss has
occurred, the solution should be replaced.
6.8 Stock solutions
6.8.1 Preparation -- Prepare in nonane per the
steps below or purchase as dilute solu-
tions (Cambridge Isotope Laboratories,
Cambridge, MA, or equivalent). Observe
the safety precautions in Section 4, and
the recommendation in Section 4.1.2.
6.8.2 Dissolve an appropriate amount of assayed
reference material in solvent. For
example, weigh 1-2 mg of 2,3,7,8-TCDD to
three significant figures in a 10 mL
ground glass stoppered volumetric flask
and fill to the mark with nonane. After
the TCDD is completely dissolved, transfer
the solution to a clean 15 ml vial with
Teflon-lined cap.
6.8.3 Stock standard solutions should be checked
for signs of degradation prior to the
preparation of calibration or performance
test standards. Reference standards that
can be used to determine the accuracy of
calibration standards are available from
Cambridge Isotope Laboratories.
6.9 "Secondary standard -- Using stock solu-
tions (Section 6.8), prepare secondary
standard solutions containing the com-
pounds and concentrations shown in Table 4
in nonane.
6.10 Labeled compound stock standard -- From
stock standard solutions prepared as
above, or from purchased mixtures, prepare
this standard to contain the labeled com-
pounds at the concentrations shown in
Table 4 in nonane. This solution is
diluted with acetone prior to use (Section
10.3.2).
6.11 Cleanup standard - Prepare Cl^-2,3,7,8-
TCDD at the concentration shown in Table 4
in nonane.
6.12 Internal standard -- Prepare at the con-
centration shown in Table 4 in nonane.
6.13 Calibration standards (CS1 through CSS) --
Combine the solutions in Sections 6.9,
6.10, 6.11, and 6.12 to produce the five
calibration solutions shown in Table 4 in
nonane. These solutions permit the rela-
tive response (labeled to unlabeled) and
response factor to be measured as a func-
tion of concentration. The CS3 standard
is used for calibration verification
(VER).
6.14 Precision and recovery standard (PAR) --
Used for determination of initial (Section
8.2) and ongoing (Section 14.5) precision
and accuracy. This solution contains the
analytes and labeled compounds at the con-
centrations listed in Table 4 in nonane.
This solution is diluted with acetone
prior to use (Section 10.3.4).
6.15 GC retention time window defining solu-
tions -- Used to define the beginning and
ending retention times for the dioxin and
furan isomers.
00:)
-------
6.15.1 DB-5 column window defining standards --
Cambridge Isotope Laboratories ED-1732-A
(dioxins) and ED-1731-A (furans), or
equivalent, containing the compounds
listed in Table 5.
6.16 Isomer specificity test standards -- Used
to demonstrate isomer specificity for the
2.3,7,8-tetra- isomers of dioxin and
furan.
6.16.1 Standards for the DB-5 column •- Cambridge
Isotope Laboratories ED-908, ED-908-C, or
ED-935, or equivalent, containing the
compounds listed in Table 5.
6.16.2 Standards for the 08-225 column
Cambridge Isotope Laboratories EF-937 or
EF-938, or equivalent, containing the
compounds listed in Table 5.
6.17 Stability of solutions -- Standard
solutions used for quantitative purposes
(Sections 6.9-6.14) shall be analyzed
within 48 hours of preparation and on a
monthly basis thereafter for signs of
degradation. Standards will remain
acceptable if the peak area at the
quant i tat ion ai/z remains within ±15
percent of the area obtained in the
initial analysis of the standard. Any
standards failing to meet this criterion
should be assayed against reference
standards, as in Section 6.8.3., before
further use.
7 CALIBRATION
7.1 Assemble the GCMS and establish the
operating conditions necessary to meet the
relative retention time specifications in
Table 2.
7.1.1 The following GC operating conditions may
be used for guidance and adjusted as
needed to meet the relative retention time
specifications in Table 2:
Injector temp: 270 °C
Interface temp: 290 °C
Initial temp and time: 200 °C, 2 min
Temp Program: 200-220 °C at 5 °C/min
220 °C for 16 min
220-235 °C at 5 "C/min
235 °C for 7 min
235-330 °C at 5 "C/min
NOTE: All portions of the column which
connect the GC to the ion source shall
remain at the interface temperature
specified above during analysis, to
preclude condensation of less volatile
compounds.
7.1.2 Mass spectrometer (MS) resolution
Obtain a selected ion current profile
(SICP) of each analyte in Table 4 at the
two exact masses specified in Table 3 and
at >10,000 resolving power by injecting an
authentic standard of the PCDDs and PCOFs
either singly or as part of a mixture in
which there is no interference between
closely eluted components, using the
procedure in Section 13.
7.1.2.1 The analysis time for PCDDs and PCDFs may
exceed the long-term mass stability of the
mass spectrometer. Because the instrument
is operated in the high-resolution mode,
mass drifts of a few ppm (e.g., 5 ppm- in
mass) can have serious adverse effects on
instrument performance. Therefore, a
mass-drift correction is mandatory. A
lock-mass ion from the reference compound
(PFK) is used for tuning the mass
spectrometer. The lock-mass ion is
dependent on the masses of the ions
monitored within each descriptor, as shown
in Table 3. The level of the reference
compound (PFK) metered into the ion
chamber during HRGC/HRMS analyses should
be adjusted so that the amplitude of the
most intense selected lock-mass ion signal
(regardless of the descriptor number) does
not exceed 10 percent of the full-scale
deflection for a given set of detector
parameters. Under those conditions,
sensitivity changes that might occur
during the analysis can be more
effectively monitored. NOTE: Excessive
PFK (or any other reference substance) may
cause noise problems and contamination of
the ion source resulting in an increase in
time lost in cleaning the source.
7.1.2.2 By using a PFK molecular leak, tune the
instrument to meet the minimum required
resolving power of 10,000 (10 percent
valley) at m/z 304.9824 (PFK) or any other
reference signal close to m/z 303.9016
(from TCDF). By using the peak matching
unit and the PFK reference peak, verify
that the exact mass of m/z 380.9760 (PFK)
is within 5 ppm of the required value.
7.2 Ion abundance ratios, minimum levels,
signal-to-noise ratios, and absolute
retention times -- Inject the CS1
10
010
-------
calibration solution (Table 4) per the
procedure in Section 13 and the conditions
in Table 2.
7.2.1 Measure the SICP areas for each analyte
and compute the ion •abundance ratios
specified in Table 3A. Compare the
computed ratio to the theoretical ratio
given in Table 3A.
7.2.1.1 The groups of m/z's to be monitored are
shown in Table 3. Each group or
descriptor shall be monitored in succes-
sion as a function of GC retention time to
ensure that all PCDDs and PCDFs are
detected. The theoretical abundance
ratios for the m/z's are given in Table
3A, along with the control limits of each
ratio.
7.2.1.2 The mass spectrometer shall be operated in
a mass drift correction mode, using per-
fluorokerosene (PFIO to provide lock
masses. The lock mass for each group of
m/z's is shown in Table 3. Each lock mass
shall be monitored and shall not vary by
more than ±10 percent throughout its
respective retention time window. Varia-
tions of the lock mass by more than 10
percent indicate the presence of coeluting
interferences that may significantly
reduce the sensitivity of the mass
spectrometer. Re-injection of another
aliquot of the sample extract will not
resolve the problem. Additional cleanup
of the extract may be required to remove
the interferences.
7.2.2 All PCDDs and PCDFs shall be within their
respective ratios; otherwise, the mass
spectrometer shall be adjusted and this
test repeated until the m/z ratios fall
within the limits specified. If the
adjustment alters the resolution of the
mass spectrometer, resolution shall be
verified (Section 7.1) prior to repeat of
the test.
7.2.3 Verify that the HRGC/HRMS instrument meets
the minimum levels in Table 2. The peaks
representing both unlabeled and labeled
analytes in the calibration standards must
have a signal-to-noise ratio (S/N) greater
than or 'equal to 10; otherwise, the mass
spectrometer shall be adjusted and this
test repeated until the minimum levels in
Table 2 are met.
7.2.4 The absolute retention time of
"12
1,2,3,4-TCDD (Section 6.12) shall exceed
25.0 minutes on the DB-5 column, and the
retention time of C12-1,2,3,4-TCDO shall
exceed 15.0 minutes on the DB-225 column;
otherwise, the GC temperature program
shall be adjusted and this test repeated
until the above-stated minimum retention
time criteria are met.
7.3 Retention time windows -- Analyze the
window defining mixtures (Section 6.15)
using the procedure in Section 13 (Figures
2A-2D). Table 5 gives the elution order
(first/last) of the compound pairs.
7.4 Isomer specificity
7.4.1 Analyze the isomer specificity test
standards (Section 6.16) using the
procedure in Section 13.
7.4.2 Compute the percent valley between the GC
peaks that elute most closely to the
2,3,7,8- TCDD and TCDF isomers, on their
respective columns, per Figure 3.
7.4.3 Verify that the height of the valley
between the most closely eluted isomers
and the 2,3,7,8- isomers is less than 25
percent (computed as 100 x/y in Figure 3).
If the valley exceeds 25 percent, adjust
the analytical conditions and repeat the
test or replace the GC column and recali-
brate (Section 7.2 through 7.4).
7.5 Calibration with isotope dilution
Isotope dilution is used for the 15
2,3,7,8-substituted PCDOs and PCDFs with
labeled compounds added to the samples
prior to extraction, and for 1,2,3,7,8,9-
HxCDD and OCDF (see Section 16.1). The
reference compound for each unlabeled
compound is shown in Table 6.
7.5.1 A calibration curve encompassing the
concentration range is prepared for each
compound to be determined. The relative
response (RR) (unlabeled to labeled) vs.
concentration in standard solutions is
plotted or computed using a linear regres-
sion. Relative response is determined
according to the procedures described
below. A minimum of five data points are
employed for calibration.
7.5.2 The relative response of each unlabeled
PCDD/PCDF and its labeled analog is
determined using the area responses of
Oil
11
-------
6-MAY-88 Sir: Voltage 705 Sys: D85US
Sample 1 Injection 1 Group 2 Mass 303.9016
100
80
60
40
20
1,3,6,8-TCDF
Norm: 3044
1,2,8.9-TCDF
25:20 26:40 28:00 29:20 30:40 32:00 33:20 34:40 36:00 37:20 38:40
6-MAY-88 Sir: Voltage 705 Sys: DB5US
Sample 1 Injection 1 Group 2 Mass 319.8965
100
80-
60
40-
20
1,3,6,8-TCDD
Norm: 481
1,2,8.9-TCDD
1,2.3.7/1,2.3,8-TCDD
2,3,7,8-TCDD
25:20 26:40 28:00 29:20 30:40- 32:00^x33:20 34:40 36:00 37:20 38:40
FIGURE 2A First and Last Eluted Tetra- Oioxin and Furan Isomers
12
012
-------
6-MAY-88 Sir: Voltage 705 Sys: DB5US
Sample 1 Injection 1 Group 2 Mass 339.8597
100
80
60
40
20-
Norm: 652
1,3,4.6,8-PeCDF
1,2,3,8,9-PeCDF
29:20 30:40 32:00 33:20 34:40 36:00 37:20 38:40
6-MAY-88 Sir: Voltage 705 Sys: DB5US
Sample 1 Injection 1 Group 2 Mass 355.8546
100-1
80
60-
40-
20-
1,2,4,7,9-PeCDD
Norm:
1,2,3,8,9-PeCDD
29:20 30:40 32:00 33:20 34:40 36:00 37:20 38:40
FIGURE 2B First and Last Eluted Penta- Dioxin and Furan Isomers
013
13
-------
6-MAY-88 Sir: Voltage 705 Sys: DB5US
Sample 1 Injection 1 Group 3 Mass 373.8208
100
80
60
40-
20-
Norm: 560
1.2.3.4,6.8-HxCDF
1.2,3,4.8,9-HxCDF
39:30 40:00 40:30 41:00 41:30 42:00 42:30 43:00 43:30 44:00 44:30
6-MAY-88 Sir: Voltage 705 Sys: DB5US
Sample 1 Injection 1 Group 3 Mass 389.8156
100
00
60
40-
20-
1.2,4,6,7,9/1,2,4,6,8,9-HxCDO
Norm: 384
1,2,3,4,6,7-HxCDO
39:30 40:00 40:30 41:00 41:30 42:00 42:30 43:00 43:30 44:00 44:30
FIGURE 2C First and Last Eluted Hexa- Dioxin and Furan Isomers
14
014
-------
6-MAY-88 Sir: Voltage 705 Sys: D85US
Sample 1 Injection 1 Group 4 Mass 407.7818
1,2,3,4,6,7,8-HpCDF
100 I .. ' •- Norm: 336
80
60^
40
20
1.2,3,4,7,8,9-HpCDF
45:20 46:40 48:00 49:20 50:40 52:00 53:20 54:40 56:00 57:20
6-MAY-88 Sir: Voltage 705 Sys: DB5US
Sample 1 Injection 1 Group 4 Mass 423.7766
1,2,3,4,6,7,9-HpCDO
,X~
80-
60-
40
20
Norm: 282
1,2,3,4.6,7.8-HpCDD
45:20 46:40 48:00 49:20 50:40 52:00 53:20 54:40 56:00 57:20
6-MAY-88 Sir: Voltage 705 Sys: DB5US
Sample 1 Injection 1 Group 4 Mass 441.7428
100, , Norm: 13
OCDF
I
80-j
60-
40
20-I
\
45:20 46:40 48:00 49:20 50:40 52:00 53:20 54:40 56:00 57:20
6-MAY-88 Sir: Voltage 705 ' Sys: DB5US
Sample 1 Injection 1 Group 4 Mass 457.7377
100-, , OCDD Norm:
80-I
60
40
20-I
45:20 46:40 48:00 49:20 50:40 52:00 53:20 54:40 56.00 57:20
FIGURE 2D First and Last Eluted Hepta- Dioxin and Furan Isomers
015
15
-------
3A DB225 Column
21-APR-88 Sir: Voltage 705 Sys: DB225
Sample 1 Injection 1 Group 1 Mass 305.8987
Text: COLUMN PERFORMANCE
2,3,7,8-TCDF
Norm:
3466
100
1,2,3,9-TCDF
20
16:10 16:20 16:30 16:40 16:50 17:00 17:10 17:20 17:30 17:40 17:50 18:00
3B DBS Column
100
FIGURE 3 Valley between 2,3,7,8- Tetra Dioxin and Furan Isomers and Other Closely Eluted Isomers
16
016
-------
7.5.3
both the primary and secondary m/z's
specified in Table 3, for each calibration
standard, as follows:
RR =
(A
1
l
(A,
where,
A and A are the areas of the primary
and secondary m/z's for the unlabeled
compound.
A. and A. are the areas of the primary
and secondary m/z's for the labeled
compound.
C. is the concentration of the labeled
compound in the calibration standard.
C is the concentration of the unlabeled
compound in the calibration standard.
To calibrate the analytical system by
isotope dilution, inject a 1.0 uL aliquot
of calibration standards CS1 through CSS
(Section 6.13 and Table 4) using the
procedure in Section 13 and the conditions
in Table 2. Compute the relative response
at each concentration.
7.5.4 Linearity -- If the relative response for
any compound is constant (less than 20
percent coefficient of variation) over the
5-point calibration range, an averaged
relative response may be used for that
compound; otherwise, the complete calibra-
tion curve for that compound shall be used
over the 5-point calibration range.
7.6 Calibration by internal standard -- The
internal standard method is applied to
determination of non-2,3,7,8-substituted
compounds having no labeled analog in this
method, and to measurement of labeled
compounds for intralaboratory statistics
(Sections 8.4 and 14.5.4).
7.6.1 Response factors -- Calibrat.ion requires
the determination of response factors (RF)
defined by the following equation:
1
RF = ' s
where,
A and A are the areas of the primary
and secondary m/z's for the compound
7.6.2
7.6.3
7.7
7.8
7.8.1
7.8.2
to be calibrated.
only one m/z for
See Table 3.)
(NOTE: There is
Cl4-2,3,7,8-TCDD.
A. and A. are the areas of the primary
and secondary m/z'
internal standard.
for the GCMS
C. is the concentration of the GCMS
internal standard (Section 6.12 and
Table 4).
C is the concentration of the compound in
the calibration standard.
To calibrate the analytical system by
internal standard, inject a 1.0 uL aliquot
of calibration standards CS1 through CSS
(Section 6.13 and Table 4) using the
procedure in Section 13 and the conditions
in Table 2. Compute the response factor
(RF) at each concentration.
Linearity -- If the response factor (RF)
for any compound is constant (less than 35
percent coefficient of variation) over the
5-point ^calibration range, an averaged
response factor may be used for that
compound; otherwise, the complete calibra-
tion curve for that compound shall be used
over the 5-point range.
Combined calibration -- By using calibra-
tion solutions (Section 6.13 and Table 4)
containing the unlabeled and labeled
compounds, and the internal standards, a
single set of analyses can be used to
produce calibration curves for the isotope
dilution and internal standard methods.
These curves are verified each shift
(Section 14.3) by analyzing the calibra-
tion verification standard (VER, Table 4).
Recalibration is required if calibration
verification criteria (Section 14.3.4)
cannot be met.
Data storage -- MS data shall
collected, recorded, and stored.
be
Data acquisition -- The signal at each .
exact m/z shall be collected repetitively
throughout the monitoring period and
stored on a mass storage device.
Response factors and multipoint
calibrations -- The data system shall be
used to record and maintain lists of
response factors (response ratios for
017
17
-------
isotope dilution) and multipoint
calibration curves. Computations of
relative standard deviation (coefficient
of variation) shall be used to test
calibration linearity. Statistics on
initial performance (Section 8.2) and
ongoing performance (Section 14.5) shall
be computed and maintained.
8 QUALITY ASSURANCE/QUALITY COMTROL
8.1 Each laboratory that uses this method is
required to operate a formal quality
assurance program (Reference 16). The
minimum requirements of this program
consist of an initial demonstration of
laboratory capability, analysis of samples
spiked with labeled compounds to evaluate
and document data quality, and analysis of
standards and blanks as tests of continued
performance. Laboratory performance is
compared to established performance
criteria to determine if the results of
analyses meet the performance characteris-
tics of the method. If the method is to
be applied routinely to samples containing
high solids with very little moisture
(e.g., soils, filter cake, compost) or to
an alternate matrix, the high solids
reference matrix (Section 6.6.2) or the
alternate matrix (Section 6.6.4) is sub-
stituted for the reagent water matrix
(Section 6.6.1) in all performance tests.
8.1.1 The analyst shall make an initial demon-
stration of the ability to generate
acceptable accuracy and precision with
this method. This ability is established
as described in Section 8.2.
8.1.2 The analyst is permitted to modify this
method to improve separations or lower the
costs of measurements, provided that all
performance specifications are met. Each
time a modification is made to the method,
the analyst is required to repeat the pro-
cedures in Sections 7.2 through 7.4 and
Section 8.2 to demonstrate method
performance.
8.1.3 Analyses of blanks are required to demon-
strate freedom from contamination (Section
3.2). The procedures and criteria for
analysis of a blank are described in
Section 8.5.
8.1.4 The laboratory shall spike all samples
with labeled compounds to monitor method
performance. This test is described in
Section 8.3. When results of these spikes
indicate atypical method performance for
samples, the samples are diluted to bring
method performance within acceptable
limits. Procedures for dilutions are
given in Section 16.4.
8.1.5 The laboratory shall, on an ongoing basis,
demonstrate through calibration verifica-
tion and the analysis of the precision and
recovery standard that the analytical sys-
tem is in control. These procedures are
described in Sections 14.1 through 14.5.
8.1.6 The laboratory shall maintain records to
define the quality of data that is gener-
ated. Development of accuracy statements
is described in Section 8.4.
8.2 Initial precision and accuracy -- To
establish the ability to generate accept-
able precision and accuracy, the analyst
shall perform the following operations.
8.2.1 For low solids (aqueous samples), extract,
concentrate, and analyze four 1-liter
aliquots of reagent water spiked with the
diluted precision and recovery standard
(PAR) (Sections 6.14 and 10.3.4) according
to the procedures in Sections 10 through
13. For an alternate sample matrix, four
aliquots of the alternate matrix are used.
All sample processing steps, including
preparation (Section 10), extraction
(Section 11), and cleanup (Section 12)
that are to be used for processing samples
shall be included in this test.
8.2.2 Using results of the set of four analyses,
compute the average concentration (X) of
the extracts in ng/mL and the standard
deviation of the concentration (s) in
ng/mL for each compound, by isotope
dilution for PCODs and PCDFs with a
labeled analog, and by internal standard
for labeled compounds. Calculate the
recovery of the labeled compounds.
8.2.3 For each unlabeled and labeled compound,
compare s and X with the. corresponding
limits for initial precision and accuracy
in Table 7. If s and X for all compounds
meet the acceptance criteria, system
performance is acceptable and analysis of
blanks and samples may begin. If,
however, any individual s exceeds the
precision limit or any individual X falls
outside the range for accuracy, system
18
013
o
-------
performance is unacceptable for that
compound. Correct the problem and repeat
the test (Section 8.2). The concentration
limits in Table 7 for labeled compounds
are based on the requirement that the
recovery of each labeled 'compound be in
the range of 25-150%.
8.3 The laboratory shall spike alt samples and
OC aliquots with the diluted labeled
compound spiking solution (Sections 6.10
and 10.3.2) to assess method performance
on the sample matrix.
8.3.1
8.3.2
8.3.3
8.4.1
8.4.2
Analyze each sample according to
procedures in Sections 10 through 13.
the
Compute the percent recovery (R) of the
labeled compounds in the labeled compound
spiking standard and the cleanup standard
using the internal standard method
(Section 7.6).
The recovery of each labeled compound must
be within 25-150%. If the recovery of any
compound falls outside of these limits,
method performance i's unacceptable for
that compound in that sample. To overcome
such difficulties, water samples are
diluted and smaller amounts of soils,
sludges, sediments and other matrices are
reanalyzed per Section 17.
8.4 Method accuracy for samples shall be
assessed and records shall be maintained.
After the analysis of five samples of a
given matrix type (water, soil, sludge,
pulp, etc) for which the labeled compound
spiking standards pass the tests in
Section 8.3, compute the average percent
recovery (R) and the standard deviation of
the percent recovery (SR) for the labeled
compounds only. Express the accuracy
assessment as a percent recovery interval
from R
2SR to R
2S for each matrix.
K
For example, if R = 90% and SR = 10% for
five analyses of pulp, the accuracy
interval is expressed as 70-110%.
Update the accuracy assessment for each
compound in each matrix on a regular basis
(e.g., after each 5-10 new accuracy
measurements).
8.5 Blanks -- Reference matrix blanks -are
analyzed to demonstrate freedom from
contamination (Section 3.2).
S.5.1 Extract and concentrate a 1-liter reagent
water blank (Section 6.6.1), high solids
reference matrix blank (Section 6.6.2),
paper matrix blank (Section 6.6.3) or
alternate reference matrix blank (Section
6.6.4) with each sample set (samples
started through the extraction process on
the same 12-hour shift, to a maximum of HO
samples). Analyze the blank immediately
after analysis of the precision and
recovery standard (Section 14.5) to
demonstrate freedom from contamination.
8.5.2 If any of the PCDDs or PCDFs (Table 1) or
any potentially interfering compound is
found in blank at greater than the minimum
level (Table 2), assuming a response
factor of 1 relative to the C.^-1,2,3,4-
TCOD internal standard for compounds not
listed in Table 1, analysis of samples is
halted until the source of contamination
is eliminated and a blank shows no
evidence of contamination at this level.
NOTE: All samples associated with a
contaminated method blank must be re-
extracted and reanalyzed before the
results may be reported for regulatory
compliance purposes.
8.6 The specifications contained in this
method can be met if the apparatus used is
calibrated properly and then maintained in
a calibrated state. The standards used
for calibration (Section 7), calibration
verification (Section 14.3), and for
initial (Section 8.2) and ongoing (Section
14.5) precision and recovery should be
identical, so that the most precise
results will be obtained. A GCMS instru-
ment will provide the most reproducible
results if dedicated to the settings and
conditions required for the analyses of
PCDDs and PCDFs by this method.
8.7 Depending on specific program require-
ments, field replicates may be collected
to determine the precision of the sampling
technique, and spiked samples may be
required to determine the accuracy of the
analysis when the internal standard method
is used.
9 SAMPLE COLLECTION,
HANDLING
PRESERVATION, AND
9.1 Collect samples in amber glass containers
following conventional sampling practices
(Reference 17). Aqueous samples which
flow freely are collected in refrigerated
19
-------
bottles using automatic sampling equip-
ment. Solid samples are collected as grab
samples using wide mouth jars.
9.2 Maintain samples at 0-4 °C in the dark
from the time of collection until
extraction. If residual chlorine is
present in aqueous samples, add 80 mg
sodium thiosulfate per liter of water.
EPA Methods 330.4 and 330.5 may be used to
measure residual chlorine (Reference 18).
9.3 Perform sample analysis within 40 days of
extraction.
10 SAMPLE PREPARATION
The sample preparation process involves
modifying the physical form of the sample
so that the PCOOs and PCDFs can be
extracted efficiently. In general, the
samples must be in a liquid form or in the
form of finely divided solids in order for
efficient extraction to take place. Table
8 lists the phase(s) and quantity
extracted for various sample matrices.
Samples containing a solid phase and
samples containing particle sizes larger
than 1 mm require preparation prior to
extraction. Because PCODs/PCDFs are
strongly associated with particulates, the
preparation of aqueous samples is depen-
dent on the solids content of the sample.
Aqueous samples containing less than one
percent solids are extracted in a
separator/ funnel. A smaller sample
aliquot is used for aqueous samples
containing one percent solids or more.
For samples expected or known to contain
high levels of the PCDDs and/or PCDFs, the
smallest sample size representative of the
entire sample should be used, and the
sample extract should be diluted, if
necessary, per Section 16.4.
10.1 Determine percent solids
10.1.1 Weigh 5-10 g of sample (to three signifi-
cant figures) into a tared beaker. NOTE:
This aliquot is used only for determining
the solids content of the sample, not for
analysis of PCDDs/PCDFs.
10.1.2 Dry overnight (12 hours minimum) at 110 ±5
°C, and cool in a dessicator.
10.1.3 Calculate percent solids as follows:
x 100
% sol ids =
weight of sample after drying
weight of sample before drying
10.2 Determine particle size
10.2.1 Spread the dried sample from Section
10.1.2 on a piece of filter paper or
aluminum foil in a fume hood or glove box.
10.2.2 Estimate the size of the particles in the
sample. If the size of the largest
particles is greater than 1 mm, the
particle size must be reduced to 1 mm or
less prior to extraction.
10.3 Preparation of aqueous samples containing
one percent solids or less -- The
extraction procedure for aqueous samples
containing less than or equal to one
percent solids involves filtering the
sample, extracting the particulate phase
and the filtrate separately, and combining
the extracts for analysis. The aqueous-
portion is extracted by shaking with
methylene chloride in a separatory funnel.
The particulate material is extracted
using the SOS procedure.
10.3.1 Mark the original level of the sample on
the sample bottle for reference. Weigh
the sample in the bottle on a top loading
balance to ±1 g.
10.3.2 Dilute a sufficient volume of the labeled
compound stock solution by a factor of 50
with acetone to prepare the labeled
compound spiking solution. 1.0 mL of the
diluted solution is required for each
sample, but no more solution should be
prepared than can be used in one day.
Spike 1.0 mL of the diluted solution into
the sample bottle. Cap the bottle and mix
the sample by careful shaking. Allow the
sample to equilibrate for 1-2 hours, with
occasional shaking.
10.3.3 For each sample or sample set (to a
maximum of 20 samples) to be extracted
during the same 12-hour shift, place two
1.0 liter aliquots of reagent water in
clean 2 liter separatory flasks.
10.3.4 Spike 1.0 mL of the diluted labeled
compound spiking standard (Section 6.10)
into one reagent water aliquot. This
aliquot will serve as the blank. Dilute
10 uL of the precision and recovery
20
020
-------
standard (Section 6.14) to 2.0 mt_ with
acetone. Spike 1.0 mL of the diluted
precision and recovery standard into the
remaining reagent water aliquot. This
aliquot will serve as the PAR (Section
14.5).
10.3.5 Assemble a Buchner funnel on top of a
clean 1 L filtration flask. Apply a
vacuum to the flask, and pour the entire
contents of the sample bottle through a
glass fiber filter (Section 5.5.4) in the
Buchner funnel, swirling the sample
remaining in the bottle to suspend any
particulates.
10.3.6 Rinse the sample bottle twice with 5 ml of
reagent water to transfer any remaining
particulates onto the filter.
10.3.7 Rinse the any particulates off the sides
of the Buchner funnel with small quanti-
ties of reagent water.
10.3.8 Weigh the empty sample bottle on a top-
loading balance to ±1 g. Determine the
weight of the sample by difference. Do
not discard the bottle at this point.
10.3.9 Extract the filtrates using the procedures
in Section 11.1.1.
10.3.10 Extract the particulates using the proce-
dures in Section 11.1.2.
10.4 Preparation of samples containing greater
than one percent solids
10.4.1 Weigh a well-mixed aliquot of each sample
(of the same matrix type) sufficient to
provide 10 g of dry solids (based on the
solids determination in 10.1.3) into a
clean beaker or glass jar.
10.4.2 Spike 1.0 ml of the diluted labeled
compound spiking solution (Section 10.3.2)
into the sample aliquot(s).
10.4.3 For each sample or sample set (to a
maximum of 20 samples) to be extracted
during the same 12-hour shift, weigh two
10 g aliquots of the appropriate reference
matrix (Section 6.6) into clean beakers or
glass jars.
10.4.4 Spikre 1.0 mL of the diluted labeled
compound spiking solution into one
reference matrix aliquot. This aliquot
will serve as the blank. Spike 1.0 mL of
the diluted precision and recovery
standard (Section 10.3.4) into the
remaining reference matrix aliquot. This
aliquot will serve as the PAR (Section
14.5).
10.4.5 Stir or tumble and equilibrate the
aliquots for 1-2 hours.
10.4.6 Extract the aliquots using the procedures
in Section 11.
10.5 Multiphase samples
10.5.1 Pressure filter the sample, blank, and PAR
aliquots through Whatman GF/0 glass fiber
filter paper. If necessary, centrifuge
these aliquots for 30 minutes at greater
than 5000 rpm prior to filtration.
10.5.2 Discard any aqueous phase (if present).
Remove any non-aqueous liquid (if present)
and reserve for recombination with the
extract of the solid phase (Section
11.1.2.5). Prepare the filter papers of
the sample and QC aliquots for particle
size reduction and blending (Section
10.6).
10.6 Sample grinding, homogenization, or blend-
ing -- Samples with particle sizes greater
than '1 mm (as determined by Section
10.2.2) are subjected to grinding, homo-
genization, or blending. The method of
reducing particle size to less than 1 mm
is matrix dependent. In general, hard
particles can be reduced by grinding with
a mortar and pestle. Softer particles can
be reduced by grinding in a Wiley mill or
meat grinder, by homogenization, or by
blending.
10.6.1 Each size reducing preparation procedure
on each matrix shall be verified by run-
ning the tests in Section 8.2 before the
procedure is employed routinely.
10.6.2 The grinding, homogenization, or blending
procedures shall be carried out in a glove
box or fume hood to prevent particles from
contaminating the work environment.
10.6.3 Grinding -- Tissue samples, certain papers
and pulps, slurries, and amorphous solids
can be ground in a Wiley mill or heavy
duty meat grinder. In some cases, reduc-
ing the temperature of the sample to
freezing or to dry ice or liquid nitrogen
temperatures can aid in the grinding
process. Grind the sample aliquots from
Section 10.4.5 or 10.5.2 in a clean
021
21
-------
grinder. Do not allow the sample tempera-
ture to exceed 50 °C. Grind the blank and
reference matrix aliquots using a clean
grinder.
10.6.4 Homogenization or blending -- Particles
that are not ground effectively, or
particles greater than 1 mm in size after
grinding, can often be reduced in size by
high speed homogenization or blending.
Homogenize and/or blend the sample, blank,
and PAR aliquots from Section 10.4.5,
10.5.2, or 10.6.3.
10.6.5 Extract the aliquots using the procedures
in Section 11.
11 EXTRACTION AND CONCENTRATION
11.1 Extraction of filtrates -- extract the
aqueous samples, blanks, and PAR aliquots
according to the following procedures.
11.1.1 Pour the filtered aqueous sample from the
filtration flask into a 2-1 separatory
funnel. Rinse the flask twice with 5 ml
of reagent water and add these rinses to
the separatory funnel. Add 60 mL methy-
lene chloride to the 'Sample bottle
(Section 10.3.8)°, seal,and shake 60
seconds to rinse the inner surface.
11.1.2 Transfer the solvent to the separatory
funnel and extract the sample by shaking
the funnel for 2 minutes with periodic
venting. Allow the organic layer to
separate from the water phase for a
minimum of 10 minutes. If the emulsion
interface between layers is more than one-
third the volume of the solvent layer,
employ mechanical techniques to complete
the phase separation (e.g., a glass stir-
ring rod). Drain the methylene chloride
extract into a sol vent-rinsed glass funnel
approximately one-ha If full of clean
sodium sulfate. Set up the glass funnel
so that it will drain directly into a
solvent-rinsed 500-mL K-0 concentrator
fitted with a 10 ml concentrator tube.
NOTE: Experience with aqueous samples
high in dissolved organic materials (e.g.,
paper mill effluents) has shown that acid-
ification of the sample prior to
extraction may reduce the formation of
emulsions. Paper industry methods suggest
that the addition of up to 400 mL of
ethanol to a 1 L effluent sample may also
reduce emulsion formation. However,
studies by the Agency to date suggest that
the effect may be a result of the dilution
of the sample, and that the addition of
reagent water may serve the same function.
Mechanical techniques may still be neces-
sary to complete the phase separation. If
either of these techniques is utilized,
the laboratory must perform the startup
tests described in Section 8.2 using the
same techniques.
11.1.3 Extract the water sample two more times
using 60 mL of fresh methylene chloride
each time. Drain each extract through the
funnel containing the sodium sulfate into
the K-0 concentrator. After the third
extraction, rinse the separatory funnel
with at least 20 mL of fresh methylene
chloride, and drain this rinse through the
sodium sulfate into the concentrator.
Repeat this rinse at least twice.
11.1.4 The extract of the filtrate must be
concentrated before it is combined with
the extract of the particulates for
further cleanup. Add one or two clean
boiling chips to the receiver and attach a
three-ball macro Snyder column. Pre-wet
the column by adding approximately 1 mL of
hexane through the top. Place the K-D
apparatus in a hot water bath so that the
entire lower rounded surface of the flask
is bathed with steam.
11.1.5 Adjust the vertical position of the
apparatus and the water temperature as
required to complete the concentration in
15-20 minutes. At the proper rate of
distillation, the balls of the column will
actively chatter but the chambers will not
flood.
11.1.6 When the liquid has reached an apparent
volume of 1 mL, remove the K-D apparatus
from the -bath and allow the solvent to
drain and cool for at least 10 minutes.
Remove the Snyder column and rinse the
flask and its lower joint into the concen-
trator tube with 1-2 ml of hexane. A 5 mu
syringe is recommended for this operation.
11.1.7 The concentrated extracts of the filtrate
and the particulates are combined using
the procedures in Section 11.2.13.
11.2 Soxhlet/Dean-Stark extraction of solids --
Extract the solid samples, particulates,
blanks, and PAR aliquots using the follow-
ing procedure.
11.2.1 Charge a clean extraction thimble with 5.0
g of 100/200 mesh silica (Section 6.5.1.1)
22
022
-------
and 100 g of quartz sand (Section 6.3.2).
NOTE: Do not disturb the silica layer
throughout the extraction process.
11.2.2 Place the thimble in a clean extractor.
Place 30-40 ml of toluene' in the receiver
and 200-250 mL of toluene in the flask.
11.2.3 Pre-extract the glassware by heating the
flask until the toluene is boiling. When
properly adjusted, 1-2 drops of toluene
per second will fall from the condenser
tip into the receiver. Extract the
apparatus for three hours minimum.
11.2.4 After pre-extraction, cool and disassemble
the apparatus. Rinse the thimble with
toluene and allow to air dry.
11.2.5 Load the wet sample from Sections 10.4.6,
10.5.2, 10.6.3, or 10.6.4, and any non-
aqueous liquid from Section 10.5.2 into
the thimble and manually mix into the sand
layer with a clean metal spatula carefully
breaking up any large lumps of sample. If
the material to be extracted is the
particulate matter from the filtration of
an aqueous sample, add the filter paper to
the thimble also.
11.2.6 Reassemble the pre-extracted SOS apparatus
and add a fresh charge of toluene to the
receiver and reflux flask.
11.2.7 Apply power to the heating mantle to begin
refluxing. Adjust the reflux rate to
match the rate of percolation through the
sand and silica beds until water removal
lessens the restriction to toluene flow.
Check the apparatus for foaming frequently
during the first 2 hours of extraction.
If foaming occurs, reduce the reflux rate
until foaming subsides.
11.2.8 Drain the water from the receiver at 1-2
hours and 8-9 hours, or sooner if the
receiver fills with water. Reflux the
sample for a total of 16-24 hours. Cool
and disassemble the apparatus. Record the
total volume of water collected.
11.2.9 Remove the distilling flask. Drain the
water from the Oean Stark receiver and add
any toluene in the receiver to the extract
in the flask.
11.2.10 For solid samples, the extract must be
concentrated to approximately 10 mL prior
to back extraction. For the particulates
filtered from an aqueous sample, the
extract must be concentrated prior to
combining with the extract of the
filtrate. Therefore, add one or two clean
boiling chips to the round bottom flask
and attach a three-ball macro Snyder
column. Pre-wet the column by adding
approximately 1 mL of toluene through the
top. Place the round bottom flask in a
heating mantle and apply heat as required
to complete the concentration in 15-20
minutes. At the proper rate of distilla-
tion, the balls of the column will
actively chatter but the chambers will not
flood.
11.2.11 When the liquid has reached an apparent
volume of 10 mL, remove the round bottom
flask from the heating mantle and allow
the solvent to drain and cool for at least
10 minutes. Remove the Snyder column.
11.2.12 If the extract is from a solid sample, not
the particulates from an aqueous sample,
transfer the concentrated extract to a 250
mL separatory funnel. Rinse the flask
with toluene and add the rinse to the
separatory funnel. Proceed with back
extraction per Section 11.3.
11.2.13 If the extract is from the partjculates
from an aqueous sample, it must be com-
bined with the concentrated extract of the
filtrate (Section 11.1.7) prior to back
extraction. Assemble the glass funnel
filled approximately one-half full with
sodium sulfate from Section 11.1.2 such
that the funnel will drain into the K-D
concentrator from Section 11.1.7 contain-
ing the concentrated methylene chloride
extract of the filtrate. Pour the concen-
trated toluene extract of the particulates
through the sodium sulfate into the K-0
concentrator. Rinse the round-bottom
flask with three 15-20 mL volumes of
hexane, and pour each rinse through the
sodium sulfate into the K-0 concentrator.
Add one or two fresh boiling chips to the
receiver and attach the three-ball macro
Snyder column to the K-0 concentrator.
Pre-wet the column by adding approximately
1 mL of hexane to the top of the column.
Concentrate the combined extract to
approximately 10 mL (the volume of the
toluene). Remove the K-0 apparatus from
the bath and allow the solvent to drain
and cool for at least 10 minutes. Remove
the Snyder column. Transfer the contents
of the K-0 concentrator to a pre-rinsed
250 mL separatory funnel. Rinse the flask
023
23
-------
and lower joint with three 5 ml volumes of
hexane, and add each rinse to the separa-
tory funnel. Proceed with back extraction
per Section 11.3.
11.3 Back extraction with base and acid
11.3.1 Spike 1.0 mL of the cleanup standard
(Section 6.11) into the separatory funnels
containing the sample and QC extracts
(Section 11.2.12 or 11.2.13).
11.3.2 Partition the extract against 50 mL of
potassium hydroxide solution (Section
6.1.1). Shake for 2 minutes with periodic
venting into a hood. Remove and discard
the aqueous layer. Repeat the base wash-
ing until no color is visible in the
aqueous layer, to a maximum of four wash-
ings. Minimize contact time between the
extract and the base to prevent degrada-
tion of the PCDDs and PCDFs. Stronger
potassium hydroxide solutions may be
employed for back extraction, provided
that the laboratory meets the specifica-
tions for labeled compound recovery and
demonstrates acceptable performance using
the procedures in Section 8.2.
11.3.3 Partition the extract against 50 mL of
sodium chloride solution (Section 6.1.3)
in the same way as with base. Discard the
aqueous layer.
11.3.4 Partition the extract against 50 mL of
sulfuric acid (Section 6.1.2) in the same
way as with base. Repeat the acid washing
until no color is visible in the aqueous
layer, to a maximum of four washings.
11.3.5 Repeat the partitioning against sodium
chloride solution and discard the aqueous
layer.
11.3.6 Pour each extract through a drying column
containing 7 to 10 cm of anhydrous sodium
sulfate. Rinse the separatory funnel with
30-50 mL of toluene and pour through the
drying column. Collect each extract in a
500 mL round bottom flask. Concentrate
and clean up the samples and QC aliquots
per Sections 11.4 and 12.
11.4 Macro-concentration -- Concentrate the
extracts in separate 100-mL round bottom
flasks on a rotary evaporator.
11.4.1 Assemble the rotary evaporator according
to manufacturer's instructions, and warm
the water bath to 45 °C. On a daily
basis, preclean the rotary evaporator by
concentrating 100 mL of clean extraction
solvent through the system. Archive both
the concentrated solvent and the solvent
in the catch flask for contamination check
if necessary. Between samples, three 2-3
mL aliquots of toluene should be rinsed
down the feed tube into a waste beaker.
11.4.2 Attach the round bottom flask containing
the sample extract to the rotary evapora-
tor. Slowly apply vacuum to the system,
and begin rotating the sample flask.
11.4.3 Lower the flask into the water bath and
adjust the speed of rotation and the
temperature as required to complete the
concentration in 15-20 minutes. At the
proper rate of concentration, the flow of
solvent into the receiving flask will be
steady, but no bumping or visible boiling
of the extract will occur. NOTE: If the
rate of concentration is too fast, analyte
loss may occur.
11.4.4 When the liquid in the concentration flask
has reached an apparent volume of 2 mL,
remove the flask from the water bath and
stop the rotation. Slowly and carefully,
admit air into the system. Be sure not to
open the valve so quickly that the sample
is blown out of the flask. Rinse the feed
tube with approximately 2 mL of hexane.
11.4.5 Transfer the extract to a vial using three
2-3 mL rinses of hexane. Proceed with
micro-concentration and solvent exchange.
11.5 Micro-concentration and solvent exchange
11.5.1 Toluene extracts to be subjected to GPC or
HPLC cleanup are exchanged into methylene
chloride. Extracts that are to be cleaned
up using silica gel, alumina, and/or AX-
21/Celite are exchanged into hexane.
11.5.2 Transfer the vial containing the sample
extract to a nitrogen evaporation device.
Adjust the flow of nitrogen so that the
surface of the solvent is just visibly
disturbed. NOTE: A large vortex, in the
solvent may cause analyte loss.
11.5.3 Lower the vial into a 45 3C water bath and
continue concentrating.
11.5.4 When the volume of the liquid is aoproxi-
mately 100 uL, add 2-3 mL of the desired
solvent (methylene chloride or hexane) and
24
02-1
-------
11.5.5
11.5.6
11.5.7
11.5.8
12
12.1
12.1.2
12.1.3
continue concentration to approximately
100 uL. Repeat the addition of solvent
and concentrate once more.
If the extract is to be cleaned up by GPC
or HPLC, adjust the volume of the extract
to 5.0 ml with methylene chloride.
Proceed with GPC cleanup (Section 12.2).
If the extract is to be cleaned up by
column chromatography (alumina, silica
gel, AX-21/Celite), bring the final volume
to 1.0 ml with hexane. Proceed with
column cleanups (Sections 12.3-12.5).
For extracts to be concentrated for
injection into the GCHS -- add 10 uL of
nonane to the vial. Evaporate the solvent
to the level of the nonane. Evaporate the
hexane in the vial to the level of the
nonane.
Seal the vial and label with the sample
number. Store in the dark at room temper-
ature until ready for GCMS analysis.
EXTRACT CLEANUP
Cleanup may not be necessary for
relatively clean samples (e.g., treated
effluents, groundwater, drinking water).
If particular circumstances require the
use of a cleanup procedure, the analyst
may use any or all of the procedures below
or any other appropriate procedure.
Before using a cleanup procedure, the
analyst must demonstrate that the require-
ments of Section 8.2 can be met using the
cleanup procedure.
Gel permeation chromatography (Section
12.2) removes many high molecular weight
interferences that cause GC column
performance to degrade. It may be used
for all soil and sediment extracts and may
be used for water extracts that are
expected to contain high molecular weight
organic compounds (e.g., polymeric
materials, humic acids).
Acid, neutral, and basic silica gel, and
alumina (Sections 12.3 and 12.4) are used
to remove nonpolar and polar
interferences.
AX-21/Celite (Section 12.5) is
remove nonpolar interferences.
used to
12.1.4 HPLC (Section 12.6) is. used to provide
specificity . for the 2,3,7,8-substituted
and other PCDO and PCOF isomers.
12.2 Gel permeation chromatography (GPC)
12.2.1 Column packing
12.2.1.1 Place 70-75 g of SX-3 Bio-beads in a 400-
500 ml beaker.
12.2.1.2 Cover the beads with methylene chloride
and allow to swell overnight (12 hours
minimum).
12.2.1.3 Transfer the swelled beads to the column
and pump solvent through the column, from
bottom to top, at 4.5-5.5 mL/min prior to
connecting the column to the detector.
12.2.1.4 After purging the column with solvent for
1-2 hours, adjust the column head pressure
to 7-10 psig and purge for 4-5 hours to
remove air. Maintain a head pressure of
7-10 psig. Connect the column to the
detector.
12.2.2 Column calibration
12.2.2.1 Load 5 mL of the calibration solution
(Section 6.4) into the sample loop.
12.2.2.2 Inject the calibration solution and record
the signal from the detector. The elution
pattern will be corn oil, bis(2-ethyl
hexyl) phthalate, pentachlorophenol,
perylene, and sulfur.
12.2.2.3 Set the "dump time" to allow >85 percent
removal of the corn oil and '85 percent
collection of the phthalate.
12.2.2.4 Set the "collect time" to the peak minimum
between perylene and sulfur.
12.2.2.5 Verify the calibration with the calibra-
tion solution after every 20 extracts.
Calibration is verified if the recovery of
the pentachlorophenol is greater than 85
percent. If calibration is not verified,
the system shall be recalibrated using the
calibration solution, and the previous 20
samples shall be re-extracted and cleaned
up using the calibrated GPC system.
12.2.3 Extract cleanup -- GPC requires that the
column not be overloaded. The column
specified in this method is designed to
handle a maximum of 0.5 g of high
molecular weight material in a 5 mL
0
or,-
25
-------
extract. If the extract is known or
expected to contain more than 0.5 g, the
extract is split into a liquors for GPC and
the aliquots are combined after elution
from the column. The residue content of
the extract may be obtained
gravimetricaUy by evaporating the solvent
from a 50 uL aliquot.
12.2.3.1 Filter the extract or load through the
filter holder to remove particulates.
Load the 5.0 mL extract onto the column.
12.2.3.2 Elute the extract using the calibration
data determined in Section 12.2.2.
Collect the eluate in a clean 400-500 ml
beaker.
12.2.3.3 Rinse the sample loading tube thoroughly
with methylene chloride between extracts
to prepare for the next sample.
12.2.3.4 If a particularly dirty extract is
encountered, a 5.0 ml methylene chloride
blank shall be run through the system to
check for carry-over.
12.2.3.5 Concentrate the eluate per Section 11.2.1,
11.2.2, and 11.3.1 or 11.3.2 for further
cleanup or for injection into the GCHS.
12.3 Silica gel cleanup
12.3.1 Place a glass wool plug in a 15 mm i.d.
chromatography column. Pack the column in
the following order (bottom to top): 1 g
silica gel (Section 6.5.1.1), four g basic
silica gel (Section 6.5.1.3), 1 g silica
gel, 8 g acid silica gel (Section
6.5.1.2), 2 g silica gel. Tap the column
to settle the adsorbents.
12.3.2 Pre-rinse the column with 50-100 mL of
hexane. Close the stopcock when the
hexane is within 1 mm of the sodium
sulfate. Discard the eluate. Check the
column for channeling. If channeling is
present, discard the column and prepare
another.
12.3.3 Apply the concentrated extract to the
column. Open the stopcock until the
extract is within 1 mm of the sodium
sulfate.
12.3.4 Rinse the receiver twice with 1 mL
portions of hexane and apply separately to
the column. Elute the PCDDs/PCDFs with
100 mL hexane and collect the eluate.
12.3.5 Concentrate the eluate per Section 11.4 or
11.5 for further cleanup or for injection
into the HPLC or GCMS.
12.3.6 For extracts of samples known to contain
large quantities of other organic
compounds (such as paper mill effluents)
it may be advisable to increase the
capacity of the sflica gel column. This
may be accomplished by increasing the
strengths of the acid and basic silica
gels. The acid silica gel (Section
6.5.1.2) may be increased in strength to
as much as 44% w/w (7.9 g su If uric acid
added to 10 g silica gel). The basic
silica gel (Section 6.5.1.3) may be
increased in strength to as much as 33%
w/w (50 mL IN NaOH added to 100 g silica
gel). NOTE: The use of stronger acid
silica gel (44% w/w) may lead to charring
of organic compounds in some extracts.
The charred material may retain some of
the analytes and lead to lower recoveries
of PCOOs/PCDFs. Increasing the strengths
of the acid and basic silica gel may also
require different volumes of hexane than
those specified above, to elute the
analytes off the column. Therefore, the
performance of the method after such
modifications must be verified by the
procedures in Section 8.2.
12.4 Alumina cleanup
12.4.1 Place a glass wool plug in a 15 mm i.d.
chromatography column.
12.4.2 If using acid alumina, pack the column by
adding 6 g acid alumina (Section 6.5.2.1).
If using basic alumina, substitute 6 g
basic alumina (Section 6.5.2.2). Tap the
column to settle the adsorbents.
12.4.3 Pre-rinse the column with 50-100 mL of
hexane. Close the stopcock when the
hexane is within 1 mm of the alumina.
12.4.4 Discard the eluate. Check the column for
channeling. If channeling is present,
discard the column and prepare another.
12.4.5 Apply the concentrated extract to the
column. Open the stopcock until the
extract is within 1 mm of the alumina.
12.4.6 Rinse the receiver twice with 1 mL
portions of hexane and apply separately to
the column. Elute the interfering
compounds with 100 mL hexane and discard
the eluate.
26
02
-------
12.A.7 The choice of eluting solvents will.depend
on the choice of alumina (acid or basic)
made in Section 12.4.2.
12.4.7.1 If using acid alumina, elute the PCDDs and
PCDFs from the column with 20 ml methylene
chloride:hexane (20:80 v/v). Collect the
eluate.
12.4.7.2 If using basic alumina, elute the PCDOs
and PCDFs from the column with 20 mL
methylene chloriderhexane (50:50 v/v).
Collect the eluate.
12.4.8 Concentrate the eluate per Section 11.4 or
11.5 for further cleanup or for injection
into the HPLC or GCHS.
12.5 AX-21/Celite
12.5.1 Cut both ends from a 10 ml disposable
serological pipet to produce a 10 cm
column. Fire polish both ends and flare
both ends if desired. Insert a glass wool
plug at one end, then pack the column with
1 g of the activated AX-21/Celite to form
a 2 cm long adsorbent bed. Insert a glass
wool plug on top of the bed to hold the
adsorbent in place.
12.5.2 Pre-rinse the column with five mL of
toluene followed by 2 ml methylene
chloride:methanol:toluene (15:4:1 v/v), 1
ml methylene chloride:cyclohexane (1:1
v/v), and five mL hexane. If the flow
rate of eluate exceeds 0.5 mL per min,
discard the column.
12.5.3 When the solvent is within 1 mm of the
column packing, apply the sample extract
to the column. Rinse the sample container
twice with 1 mL portions of hexane and
apply separately to the column. Apply 2
mL of hexane to complete the transfer.
12.5.4 Elute the interfering compounds with 2 mL
of hexane, 2 mL of methylene
chloride:cyclohexane (1:1 v/v), and 2 mL
of methylene chloride:methanol:toluene
(15:4:1 v/v). Discard the eluate.
12.5.5 Invert the column and elute the PCDDs and
PCDFs with 20 mL of toluene. If carbon
particles are present in the eluate,
filter through glass fiber filter paper.—
12.5.6 Concentrate the eluate per Section 11.4 or
11.5 for further cleanup or for injection
into the HPLC or GCMS.
12.6 HPLC (Reference 6)
12.6.1 Column calibration
12.6.1.1 Prepare a calibration standard containing
the 2,3,7,8- isomers and/or other isomers
of interest at a concentration of.approxi-
mately 500 pg/uL in methylene chloride.
12.6.1.2 Inject 30 uL of the calibration solution
into the HPLC and record the signal from
the detector. Collect the eluant for re-
use. The elution order will be the tetra-
through octa-isomers.
12.6.1.3 Establish the collect time for the tetra-
i seniors and for the other isomers of
interest. Following calibration, flush
the injection system with copious
quantities of methylene chloride, includ-
ing a minimum of five 50-uL injections
while the detector is monitored, to ensure
that residual PCDDs and PCDFs are removed
from the system.
12.6.1.4 Verify the calibration with the calibra-
tion solution after every 20 extracts.
Calibration is verified if the recovery of
the PCDDs and PCDFs from the calibration
standard (Section 12.6.1.1) is 75-125
percent compared to the calibration
(Section 12.6.1.2). If calibration is not
verified, the system shall be recalibrated
using the calibration solution, and the
previous 20 samples shall be re-extracted
and cleaned up using the calibrated
system.
12.6.2 Extract cleanup -- HPLC requires that the
column not be overloaded. The column
specified in this method is designed to
handle a maximum of 30 uL of extract. If
the extract cannot be concentrated to less
than 30 uL, it is split into fractions and
the fractions are combined after elution
from the column.
12.6.2.1 Rinse the sides of the vial twice with 30
uL of methylene chloride and reduce to 30
uL with the blowdown apparatus.
12.6.2.2 Inject the 30 uL extract into the HPLC.
12.6.2.3 Elute the extract using the calibration
data determined in 12.6.1. Collect the
fraction(s) in a clean 20 mL concentrator
tube containing 5 mL of hexaneracetone
(1:1 v/v).
027
n
27
-------
12.6.2.4 If an extract containing greater than 100
ng/mL of total PCOD or PCDF is encoun-
tered, a 30 uL methylene chloride blank
shall be run through the system to check
for carry-over.
12.6.2.5 Concentrate the eluate per Section 11.5
for injection into the GCHS.
13 HRGC/HRMS ANALYSIS
13.1 Establish the operating conditions given
in Section 7.1.
13.2 Add 10 uL of the internal standard solu-
tion (Section 6.12) to the sample extract
immediately prior to injection to minimize
the possibility of loss by evaporation,
adsorption, or reaction. If an extract is
to be reanalyzed and evaporation has
occurred, do not add more instrument
internal standard solution. Rather, bring
the extract back to its previous volume
(e.g., 19 uL) with pure nonane only.
13.3 Inject 1.0 uL of the concentrated extract
containing the internal standard solution,
using on-column or split less injection.
Start the GC column initial isothermal
hold upon injection. Start MS data
collection after the solvent peak elutes.
Stop data col lection.after the octachloro-
dioxin and furan have eluted. Return the
column to the initial temperature for
analysis of the next extract or standard.
14 SYSTEM AND LABORATORY PERFORMANCE
14.1 At the beginning of each 12-hour shift
during which analyses are performed, GCMS
system performance and calibration are
verified for all unlabeled and labeled
compounds. For these tests, analysis of
the CS3 calibration verification (VER)
standard (Section 6.13 and Table 4) and
the isomer specificity test standards
(Sections 6.16 and Table 5) shall be used
to verify all performance criteria.
Adjustment and/or recalibration (per
Section 7) shall be performed until all
performance criteria are met. Only after
all performance criteria are met may
samples, blanks, and precision and
recovery standards be analyzed.
14.2 MS resolution -- A static resolving power
of at least 10,000 (10 percent valley
definition) must be demonstrated at appro-
priate masses before any analysis is
performed. Static resolving power checks
must be performed at the beginning and at
the end of each 12-hour shift according to
procedures in Section 7.1.2. Corrective
actions must be implemented whenever the
resolving power does not meet the
requirement.
14.3 Calibration verification
14.3.1 Inject the VER ' standard using the
procedure in Section 13.
14.3.2 The m/z abundance ratios for all PCDDs and
PCDFs shall be within the limits in Table
3A; otherwise, the mass spectrometer shall
be adjusted until the m/z abundance ratios
fall within the limits specified, and the
verification test (Section 14.3.1)
repeated. If the adjustment alters the
resolution of the mass spectrometer, reso-
lution shall be verified (Section 7.1.2)
prior to repeat of the verification test.
14.3.3 The peaks representing each unlabeled and
labeled compound in the VER standard must
be present with a S/N of at least 10;
otherwise, the mass spectrometer shall be
adjusted and the verification test
(Section 14.3.1) repeated.
14.3.4 Compute the concentration of each
unlabeled compound by isotope dilution
(Section 7.5) for those compounds that
have labeled analogs (Table 1). Compute
the concentration of the labeled compounds
by the internal standard method. These
concentrations are computed based on the
averaged relative response and averaged
response factor from the calibration data
in Section 7.
14.3.5 For each compound, compare the concentra-
tion with the calibration verification
limit in Table 7. If all compounds meet
the acceptance criteria, calibration has
been verified. If, however, any compound
fails, the measurement system is not
performing properly for that compound. In
this event, prepare a fresh calibration
standard or correct the problem causing
the failure and repeat the resolution
(Section 14.2) and verification (Section
14.3.1) tests, or recalibrate (Section 7).
14.4 Retention times and GC resolution
14.4.1 Retention times
28
-------
14.4.1.1 Absolute --The absolute retention times
of the 3C12-1,2,3,4-TCDD and C^-
1.2,3,7.8.9-HxCDD GCMS internal standards
shall be within ±15 seconds of the
retention times obtained during calibra-
tion (Section 7.2.4).
14.4.1.2 Relative -- The relative retention times
of unlabeled and labeled PCDDs and PCDFs
shall be within the limits given in Table
2.
14.4.2 GC resolution
14.4.2.1 Inject the isomer specificity standards
(Section 6.16) on their respective
columns.
14.4.2.2 The valley height between 2,3,7,8-TCDO and
the other tetra- dioxin isomers at m/z
319.8965, and between 2.3,7,8-TCDF and the
other tetra- furan isomers at m/z 303.9016
shall not exceed 25 percent on their
respective columns (Figure 3).
14.4.3 If the absolute retention time of any
compound is not within the limits
specified or the 2,3,7,8- isomers are not
resolved, the GC is not performing
properly. In this event, adjust the GC
and repeat the verification test (Section
14.3.1) or recalibrate (Section 7).
14.5 Ongoing precision and accuracy
14.5.1 Analyze the extract of the diluted
precision and recovery standard (PAR)
(Section 10.3.4 or 10.4.4) prior to
analysis of samples from the same set.
14.5.2 Compute the concentration of each PCDD and
PCDF by isotope dilution for those
compounds that have labeled analogs
(Section 7.5). Compute the concentration
of each labeled compound by the internal
standard method.
14.5.3 For each unlabeled and labeled compound,
compare the concentration with the limits
for ongoing accuracy in Table 7. If all
compounds meet the acceptance criteria,
system performance is acceptable and
analysis of blanks and samples may
proceed. If, however, any individual
concentration falls outside of the range
given, the extraction/concentration
processes are not being performed properly
for that compound. In this event, correct
the problem, re-extract the sample set
(Section 10) and repeat the ongoing
precision and recovery test (Section
14.5). The concentration limits in Table
7 for labeled compounds are based on the
requirement that the recovery of each
labeled compound be in the range of 25-
150%.
14.5.4 Add results which pass the specifications
in Section 14.5.3 to initial and previous
ongoing data for each compound in each
matrix. Update QC charts to form a
graphic representation of continued
laboratory performance. Develop a state-
ment of laboratory accuracy for each PCDD
and PCDF in each matrix type by calculat-
ing the average percent recovery (R) and
the standard deviation of percent recovery
(SR). Express the accuracy as a recovery
interval from R - 2S to R + 2SR. For
example, if R = 95% and SR = 5%, the
accuracy is 85-105%.
15 QUALITATIVE DETERMINATION
For a gas chromatographic peak to be
identified as a PCDD or PCDF (either a
unlabeled or a labeled compound), it must
meet all of the criteria in Sections 15.1-
15.4.
15.1 The signals for the two exact m/z's being
monitored (Table 3) must be present, and
must maximize within + 2 seconds of one
another.
15.2 The signal-to-noise ratio (S/N) of each of
the two exact m/z's must be greater than
or equal to 2.5 for a sample extract, and
greater than or equal to 10 for a calibra-
tion standard (see Sections 7.2.3 and
14.3.3).
15.3 The ratio of the integrated ion currents
of both the exact m/z's monitored must be
within the limits in Table 3A.
15.4 The relative retention time of the peaks
representing a unlabeled 2,3,7,8-
substituted PCDD or PCDF must be within
the limits given in Table 2. The
retention time of peaks representing non-
2,3,7,8-substituted PCDDs or PCDFs must be
within the retention time windows
established in Section 7.3.
15.5 Confirmatory analysis -- Isomer
specificity for all of the 2,3,7,8-substi-
tuted analytes cannot be attained by
analysis on the DB-5 (or equivalent) GC
column alone. The lack of specificity is
29
-------
of greatest concern for the un labeled
2,3,7,8-TCDF. Therefore, any sample in
which 2.3,7,8-TCDF is identified by
analysis on a OB-5 (or equivalent) GC
column must have a confirmatory analysis
performed on a DB-225, SP-2330, or equiva-
lent GC column. The operating conditions
in Section 7.1.1 may be adjusted for
analyses on the second GC column, but the
GCHS must meet the mass resolution and
calibration specifications in Section 7.
15.6 If any gas chromatographic peak meets the
identification criteria in Sections 15.1,
15.2, and 15.4, but does not meet the ion
abundance ratio criterion (Section 15.3),
and is not a labeled analog, that sample
must be analyzed on a second GC column, as
in Section 15.5 above. Interferences co-
eluting in either of the two m/z's may
cause the ion abundance ratio to fall out-
side of the limits in Table 3A. If the
ion abundance ratio of the peak fails to
meet the criteria on the second GC column,
then the peak does not represent a PCDD or
PCDF. If the peak does meet all of the
criteria in Sections 15.1-15.4 on the
second GC column, then calculate the
concentration of that peak from the
analysis on the second GC column, accord-
ing to the procedures in Section 16.
15.7 If any gas chromatograpoic peak that
represents a labeled analog does not meet
all of the identification criteria in
Sections 15.1-15.4 on the second GC
column, then the results may not be
reported for regulatory compliance
purposes and a new aliquot of the sample
must be extracted and analyzed.
16 QUANTITATIVE DETERMINATION
16.1 Isotope dilution -- By adding a known
amount of a labeled compound to every
sample prior to extraction, correction for
recovery of the un labeled compound can be
made because the unlabeled compound and
its labeled analog exhibit similar effects
upon extraction, concentration, and gas
chromatography. Relative response (RR)
values are used in conjunction with cali-
bration data described in Section 7.5 to
determine concentrations directly, so long
as labeled compound spiking levels are
constant, using the following equation:
where, C£x is the concentration at the
unlabeled compound in the extract and the
other terms are as defined in Section
7.5.2.
16.1.1 Because of a potential interference, the
labeled analog of OCDF is not added to the
sample. Therefore, this unlabeled analyte
is quant itated against the labeled OCDD.
As a result, the concentration of
unlabeled OCDF is corrected for the
recovery of the labeled OCDD. In
instances where OCDD and OCDF behave
differently during sample extraction,
concentration, and cleanup procedures,
this may decrease the accuracy of the OCDF
results. However, given the low toxicity
of this compound relative to the other
dioxins and furans, the potential decrease
in accuracy is not considered significant.
16.1.2 Because the labeled analog of 1,2.3,7,8,9-
HxCDD is used as an internal standard
(i.e., not added before extraction of the
sample), it cannot be used to quantitate
the unlabeled compound by strict isotope
dilution procedures. Therefore, the
unlabeled 1,2,3,7,8,9-HxCDD is quantitated
using the average of the responses of the
labeled analogs of the other two 2,3,7,8-
substituted HxCDD's, 1,2,3,4,7,8-HxCDD and
1,2,3,6,7,8-HxCDD. As a result, the
concentration of the unlabeled
1.2,3,7,8,9-HxCDD is corrected for the
average recovery of the other two HxCDD's.
16.1.3 Any peaks representing non-2,3,7,8-substi-
tuted dioxins or furans are quantitated
using an average of the response factors
from all of the labeled 2,3,7,8- isomers
in the same level of chlorination.
16.2 Internal standard -- Compute the concen-
trations of the C-labeled analogs and
the C-labeled cleanup standard in the
extract using the response factors deter-
mined from calibration data (Section 7.6)
and the following equation:
(ng/mL) =
Cis
(Ajs
Ais) RF
C (ng/mi.) =
ex
(A + A ) C.
n n is
. 2
(A' -•• A ) RR
where, C is the concentration of the
compound in the extract and the other
terms are as defined in Section 7.6.1.
(NOTE: There is only one m/z for the
37Cl-labeled standard.)
30
u '
-------
16.3 The concentration of the unlabeled
compound in the solid phase of the sample
is computed using the concentration of the
compound in the extract and the weight of
the solids (Section 10), as follows:
(C x V )
Concentration = ex ex
in solid (ng/Kg) w
s
where,
V is the extract volume in ml.
W is the sample weight in Kg.
16.4 The concentration of the unlabeled
compound in the aqueous phase of the
sample is computed using the concentration
of the compound in the extract and the
volume of water extracted (Section 10.3),
as follows:
(C
Vex>
Vs
Concentration
in aqueous phase
(P9/L)
where,
V is the extract volume in ml.
V is the sample volume in liters.
16.5 If the SICP areas at the quantitation
m/z's for any compound exceed the calibra-
tion range of the system, a smaller sample
aliquot is extracted.
16.5.1 For aqueous samples containing one percent
solids or less, dilute 100 ml, 10 ml,
etc., of sample to 1 liter with reagent
water and extract per Section 11.
16.5.2 For samples containing greater than one
percent solids, extract an amount of
sample equal to 1/10, '1/100, etc., of the
amount determined in Section 10.1.3.
Extract per Section 10.4.
16.5.3 If a smaller sample size will not be
representative of the entire sample,
dilute the sample extract by a factor of
10, adjust the concentration of the
instrument internal standard to 100 pg/uL
in the extract, and analyze an aliquot of
this diluted extract by the internal
standard method.
16.6 Results are reported to three significant
figures for the unlabeled and labeled
isomers found in all standards, blanks,
and samples. For aqueous samples, the
units are pg/L; for samples containing
greater than one percent solids (soils,
sediments, filter cake, compost), the
units are ng/Kg based on the dry weight of
the sample.
16.6.1 Results for samples which have been
diluted are reported at the least dilute
level at which the areas at the quantita-
tion m/z's are within the calibration
range (Section 16.5).
16.6.2 For unlabeled compounds having a labeled
analog, results are reported at the least
dilute level at which the area at the
quantitation m/z is within the calibration
range (Section 16.5) and the labeled
compound recovery is within the normal
range for the method (Section 17.4).
16.6.3 Additionally, the total concentrations of
all isomers in an individual level of
chlorination (i.e.,. total TCDD, total
PeCDD, etc.) are reported to three signi-
ficant figures in units of pg/L, for both
dioxins and furans. The total or ng/Kg
concentration in each level of chlorina-
tion is the sum of the concentrations of
all isomers identified in that level,
including any non-2,3,7,8-substituted
isomers.
17 ANALYSIS OF COMPLEX SAMPLES
17.1 Some samples may contain high levels (>10
ng/L; >1000 ng/Kg) of the compounds of
interest, interfering compounds, and/or
polymeric materials. Some extracts will
not concentrate to 10 uL (Section 11);
others may overload the GC column and/or
mass spectrometer.
17.2 Analyze a smaller aliquot .of the sample
(Section 16.4) when the extract will not
concentrate to 20 uL after all cleanup
procedures have been exhausted.
17.3 Recovery of labeled compound spiking
standards -- In most samples, recoveries
of the labeled compound spiking standards
will be similar to those from reagent
water or from the alternate matrix
(Section 6.6). If recovery is outside of
the 25-150% range, a diluted sample
(Section 16.4) shall be analyzed. If the
recoveries of the labeled compound spiking
standards in the diluted sample are
outside of the limits (per the criteria
above), then the verification standard
(Section 14.3) shall be analyzed and
calibration verified (Section 14.3.4). If
the calibration cannot be verified, a new
031
31
-------
calibration must be performed and the
original sample extract reanalyzed. If
the calibration is verified and the
diluted sample does not meet the limits
for labeled compound recovery, then the
method does not apply to the sample being
analyzed and the result may not be
reported for regulatory compliance
purposes.
18 METHOD PERFORMANCE
The performance specifications in this
method are based on the analyses of more
than 400 samples, representing matrices
from at least five industrial categories.
These specifications will be updated
periodically as more data are received,
and each time the procedures in the method
are revised.
REFERENCES
1 Tondeur, Yves, "Method 8290: Analytical
Procedures and Quality Assurance for
Multimedia Analysis of Polychlorinated
Oibenzo-p-dioxins and Dibenzofurans by
High-Resolution Gas Chromatography/High-
Resolution Mass Spectrometry", USEPA EMSL,
Las Vegas, Nevada, June 1987.
2 "Measurement of 2,3,7,8-Tetrachlorinated
Dibenzo-p-dioxin (TCOO) and 2,3,7,8-Tetra-
chlorinated Dibenzofuran (TCDF) in Pulp,
Sludges, Process Samples and Waste-waters
from Pulp and Paper Mills", Wright State
University, Dayton, OH 45435, June 1988.
3 "HCASI Procedures for the Preparation and
Isomer Specific Analysis of Pulp and Paper
Industry Samples for 2,3,7,8-TCDD and
2,3,7,8- TCDF". National Council of the
Paper Industry for Air and Stream Improve-
ment, 260 Madison Avenue, New York, NY
10016, Technical Bulletin No. 551, Pre-
release Copy, July 1988.
4 "Analytical Procedures and Quality
Assurance Plan for the Determination of
PCDD/PCDF in Fish", USEPA. Environmental
Research Laboratory, 6201 Congdon
Boulevard, Duluth, MN 55804, April 1988.
5 Tondeur, Yves, "Proposed GC/MS Methodology
for the Analysis of PCDDs and PCDFs in
Special Analytical Services Samples",
Triangle Laboratories, Inc., 801-10
Capitola Or, Research Triangle Park, NC
27713, January 1988; updated by personal
communication September 1988.
6 Lamparski, L.L., and Nestrick, T.J.,
"Determination of Tetra-, Hexa-, Hepta-,
and Octachlorodibenzo-p-dioxin Isomers in
Particulate Samples at Parts per Trillion
Levels", Analytical Chemistry. 52: 2045-
2054, 1980.
7 Lamparski, L.L., and Nestrick, T.J.,
"Novel Extraction Device for the
Determination of Chlorinated Dibenzo-p-
dioxins (PCDDs) and Dibenzofurans (PCDFs)
in Matrices Containing Water",
Chemosphere. 19:27-31, 1989.
8 Patterson, D.G., et. al. "Control of
Interferences in the Analysis of Human
Adipose Tissue for 2,3,7,8-Tetra-
chlorodibenzo-p-dioxin". Environmental
Toxicological Chemistry. 5: 355-360, 1986.
9 Stanley, John S., and Sack, Thomas M.,
"Protocol for the Analysis of 2,3,7,8-
Tetrachlorodibenzo-p-dioxin by High-
Resolution Gas Chromatography/High-
Resolution Mass Spectrometry", USEPA EMSL,
Las Vegas, Nevada 89114, EPA 600/4-86-004,
January 1986.
10 "Working with Carcinogens", DHEW, PHS,
CDC, NIOSH. Publication 77-206, August
1977.
11 "OSHA Safety and Health Standards, General
Industry" OSHA 2206, 29 CFR 1910, January
1976.
12 "Safety in Academic Chemistry
Laboratories", ACS Committee on Chemical
Safety, 1979.
13 "Standard Methods for the Examination of
Water and Wastewater", 16th edition and
later revisions, American Public Health
Association, 1015 15th St, N.W.,
Washington, DC 20005, 46: Section 108
(Safety), 1985.
14 "Method 613 -- 2,3,7,8-Tetrachlorodibenzo-
p-dioxin". 40 CFR 136 (49 FR 43234),
October 26, 1984, Section 4.1.
15 Provost, L.P., and Elder, R.S.,
"Interpretation of Percent Recovery Data",
American Laboratory. 15: 56-83, 1983.
16 "Handbook of Analytical Quality Control in
Water and Wastewater Laboratories", USEPA
EMSL, Cincinnati, OH 45268, EPA-600/4-79-
019, March 1979.
32
03
-------
17 "Standard Practice for Sampling Water",
ASTH Annual Book of Standards, ASTH, 1916
Race Street, Philadelphia, PA 19103-1187,
1980.
18 "Methods 330.4 and 330.5 for Total
Residual Chlorine", USEPA, EMSL, Cincin-
nati, OH 45268, EPA 600/4-70-020, March
1979.
19 Barnes, Donald G., Kutz, Frederick U., and
Bottimore, David P., "Update of Toxicity
Equivalency Factors (TEFs) for Estimating
Risks Associated with Exposures to
Mixtures of Chlorinated Dibenzo-p-Dioxins
and Dibenzofurans (CDDs/CDFs)", Risk
Assessment Forum, USEPA, Washington, DC
20460, February 1989.
033
33
-------
Table 1
POLYCHLORINATED DIBENZODIOXINS AND FURANS DETERMINED BY ISOTOPE DILUTION AND INTERNAL STANDARD
HIGH RESOLUTION GAS CHROMATOGRAPHY (HRGO/HIGH RESOLUTION MASS SPECTROMETRY (HRHS)
PCDDs/PCDFs (1)
Isomer/Congener CAS Registry
2,3,7,8-TCDD 1746-01-6
Total -TCDD 41903-57-5
2,3,7.8-TCDF 51207-31-9
Total -TCDF 55722-27-5
1,2,3,7,8-PeCDD 40321-76-4
Total -PeCDD 36088-22-9
1,2,3,7,8-PeCDF 57117-41-6
2,3,4,7,8-PeCDF 57117-31-4
Total-PeCDF 30402-15-4
1,2,3.4,7.8-HxCDD 39227-28-6
1,2,3,6,7,8-HxCDD 57653-85-7
1,2.3,7.8,9-HxCOD 19408-74-3
Total-HxCDD 34465-4608
1,2.3.4.7,8-HxCDF 70648-26-9
1,2,3,6,7.8-HxCDF 57117-44-9
1,2,3,7,8,9-HxCDF 72918-21-9
2,3,4,6,7,8-HxCDF 60851-34-5
Total-HxCDF 55684-94-1
1,2,3,4,6,7-,8-HpCDD 35822-46-9
Total-HpCDD 37871-00-4
1,2.3,4,6,7,8-HpCDF ' 67562-39-4
1,2,3,4,7,8,9-HpCOF 55673-89-7
Total-HpCDF 38998-75-3
OCDD 3268-87-9
OCOF 39001-02-0
(1) Polychlorinated dioxins and furans
TCDD = Tetrachlorodibenzo-p-dioxin
PeCDD = Pentachlorodibenzo-p-dioxin
HxCDD = Hexachlorodibenzo-p-dioxin
HpCDO = Heptachlorodibenzo-p-dioxin
OCDD = Octachlorodibenzo-p-dioxin
Labeled Analog
13C12-2,3,7,8-TCDD
37Cl4-2,3,7,8-TCDD
13C12-2,3,7,8-TCDF
13C12-1,2.3,7,8-PeCDD
13C12-1,2,3.7,8-PeCDF
13C12-2.3,4,7,8-PeCOF
13C,,-1,2,3,4,7,8-HxCDD
Cl2-1,2,3,6,7.8-HxCDD
13C12-1.2,3,7,8,9-HxCDD<2>
13C12-1,2,3,4,7,8-HxCDF
13C12-1,2,3.6,7.8-HxCDF
13C,--1. 2,3, 7,8,9- HxCDF
C,,-2,3,4,6,7,8-HxCDF
13C12-1,2,3,4,6,7,8-HpCDD
13C12-1, 2,3,4. 6,7,8-HpCOF
13C12-1,2,3,4,7.8,9-HpCDF
13C12-OCDD
none
TCDF = Tetrachlorodibenzofuran
PeCDF = Pentachlorodibenzofuran
HxCDF = Hexachlorodibenzofuran
HpCDF = Heptachlorodibenzofuran
OCDF = Octachlorodibenzofuran
CAS Registry
76523-40-5
85508-50-5
89059-46-1
109719-79-1
109719-77-9
116843-02-8
109719-80-4
109719-81-5
109719-82-6
114423-98-2
116843-03-9
116843-04-0
116843-05-1
109719-83-7
109719-84-8
109719-94-0
114423-97-1
(2) Labeled analog is used as an internal standard and therefore is not used for quant i tat ion of the native
compound.
34
031
-------
Table 2
RETENTION TIMES AND MINIMUM LEVELS FOR PCDDs AND PCDFs
Minimum Level (1)
Compound
Compounds using C..--1,2,3,4-TCDD
Native Compounds
2,3,7,8-TCDF
2,3,7,8-TCDD
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,7,8-PeCDD
Labeled Compounds
;3C12-2,3.7,8-TCDF
1JC12-1.2.3,4-TCDD
^3C12-2,3,7,8-TCDD
37Cl4-2,3,7,8-TCOD
13C12-1,2,3,7,8-PeCDF
13C12-2,3,4,7,8-PeCDF
13C12-1,2,3,7,8-PeCDD
Compounds using C12-1,2,3,7,8,9-
Native Compounds
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDF
1,2,3,4,6,7,8-HpCDD
1,2,3,4,7,8,9-HpCDF
OCDD
OCDF
Labeled Compounds
13C12-1,2,3,4,7,8-HxCDF
13C,_-1,2,3.6,7,8-HxCDF
C.2-1,2,3,7,8,9-HxCDF
13C12-2,3,4,6,7,8-HxCDF
13C12-1,2,3,4,7,8-HxCDD
13C12- 1,2,3,6,7,8-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,4,6,7,a-HpCDF
13C.-- 1,2,3,4,6,7,8-HpCDD
C,--1,2,3,4,7,8,9-HpCDF
C12-OCDD
Retention
Time
Reference
as internal standard
J3C12-2,3.7.8-TCDF
X2-2,3,7,8-TCOD
C12-1,2,3,7,8-PeCDF
13C12-2,3,4,7,8-PeCDF
C12-1,2,3,7,8-PeCDD
13C12-1,2,3,4-TCOD
13C.,-1,2.3.4-TCDD
13 12
I3C12-1.2,3.4-TCOO
^•I^S^TCDD
"C12-1,2,3,4-TCOO
13C12-1,2,3,4-TCDD
HxCDD as internal standard
13C.,-1,2,3,4,7,8-HxCDF
C12-1.2,3,6,7.8-HxCDF
13C12-1,2,3,7,8,9'HxCDF
13C,_-2,3,4,6,7,8-HxCDF
C12-1,2.3,4,7.8-HxCDD
13C12-1,2,3,6,7,8-HxCDD
13C12-1,2,3,6.7,8-HxCDD
13C12-1,2,3,4,6.7,8-HpCDF
13C.,- 1,2,3,4,6,7,8-HpCDD
C12-1,2.3,4,7,8,9-HpCDF
13C..--OCOD
13 12
C12-OCOD
13C.2-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
Relative Water Solid
Retention pg/L ng/kg
Time ppq ppt
0.993 -
0.993 -
0.918 -
0.999 -
0.987 -
0.931 -
1.000 •
0.993 -
1.002 -
1.091 -
1.123 -
1.134 -
0.986 -
0.973 -
0.937 -
0.999 -
0.999 -
0.992 -
0.986 -
0.930 -
0.986 -
0.896 -
0.996 -
0.995 -
0.947 -
0.940 -
0.993 -
0.971 -
0.974 -
0.975 -
1.000 -
0.953 -
1.023 -
1.024 -
1.050 -
1.009 10 1
1.009 10 1
1.076 50 5
1.001 50 5
1.016 50 5
0.994
1.000
1.036
1.013
1.371
1.408
1.428
1.015 50 5
1.025 50 5
1.068 50 5
1.001 50 5
1.001 50 5
1.009 50 5
1.016 50 5
1.022 50 5
1.016 50 5
1.079 50 5
1.005 100 10
1.013 100 10
0.992
1.006
1.017
1.000
1.002
1.006
1.000
1.172
1.125
1.148
1.275
Extract
pg/uL
ppb
0.5
0.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
5.0
5.0
(1) Level at which the analytical system will give acceptable SICP and calibration.
0
O r-
35
-------
DESCRIPTORS, MASSES, M/Z TYPES,
Table 3
AND ELEMENTAL COMPOSITIONS OF THE CDDs AND CDFs (1)
Descriptor
Number
1
2
3
Accurate
m/z (2)
292.9825
303.9016
305.8987
315.9419
317.9389
319.8965
321 .8936
327.8847
330.9792
331 .9368
333.9339
375.8364
339.8597
341 .8567
351.9000
353.8970
354.9792
355.8546
357.8516
367.8949
369.8919
409.7974
373.8208
375.8178
383.8639
385.8610
389.8157
391.8127
392.9760
401.8559
403.8529
430.9729
445.7555
m/z
Type
Lock
M
M+2
M
M+2
M
M+2
M
QC
M
M+2
M+2
M+2
M+4
M+2
M+4
Lock
M+2
M+4
M+2
M+4
M+2
M+2
M+4
M
M+2
M+2
M+4
Lock
M+2
M+4
QC
M+4
Elemental Composition
C7F11
C12 H4 35C14 0
C12 H4 35cl337(:l °
13C12 H4 35C14 0
13C12 H, 35C13 37Cl 0
C12 H4 35cl4 °2
C.- H. 35Cl, 37Cl 0,
12 4 3 2
C12 H4 37cl4 °2
C7F13
13C12 H, 35C14 O,
13 35 37
C12 H4 C13 Ct °2
C12 H4 35C15 37Cl 0
C,2 H3 35C14 37Cl 0
C12 H, 35C13 37C12 0
13 35 . 37
C12 H3 C14 Cl °
13 35 37
C12 H3 C13 C12 °
C0 F.,
9 13
C H 35Cl 37Cl 0
C12 H3 35C13 37C12 °2
13 35 37
C12 H3 C14 Cl °2
13 35 37
C12 H3 CL3 C12 °2
C12 H3 35cl6 37cl °
C12 H2 35cl5 3/Ct °
C12 H2 35cl4 37cl2 °
13C H 35
C12 H2 C16 °
13C12 H, 35C15 37Cl 0
C,, H2 35C15 37Cl 02
C12 H2 35cl4 37cl2 °2
C0 F -
9 15
13 35 37
C12 H2 C15 Cl °2
13 35 37
C12 H2 C14 C12 °2
C9 F13
Compound
(3)
PFK
TCDF
TCDF
TCDF(4)
TCDF(4)
TCDD
TCDD
TCDD(S)
PFK
TCDD(4)
TCDO(4)
HxCDPE
PeCDF
PeCDF
PeCDF(4)
PeCDF(4)
PFK
PeCDD
PeCDD
PeCDD (4)
PeCDD(4)
HpCDPE
HxCDF
HxCDF
HxCDF(4)
HxCDF(4)
HxCDD
HxCDD
PFK
HxCDD(4)
HxCDD(4)
PFK
OCDPE
Primary
m/z?
Yes
Yes
Yes
-
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
36
036
-------
Table 3 (continued)
DESCRIPTORS, MASSES, M/Z TYPES, AND ELEMENTAL COMPOSITIONS OF THE CDDs AND CDFs (1)
Descriptor Accurate
Number m/z (2)
(1)
(2)
(3)
4 407.7818
409.7789
417.8253
419.8220
423.7766
425.7737
430.9729
435.8169
437.8140
479.7165
5 441.7428
442.9728
443.7399
457.7377
459.7348
469.7779
471.7750
513.6775
From Reference 5
Nuclidic masses used:
H = 1.007825
0 = 15.994915
m/z
Type
M+2
M+4
M
M+2
M+2
M+4
Lock
M+2
M+4
M+4
M+2
Lock
M+4
M+2
M+4
M+2
M+4
M+4
C = 12.00000
35Cl = 34.968853
Elemental Composition
C-2 H 35C16 37Cl 0
C-2 H 35C15 37C12 0
13C12 H 35C17 0
13C12 H 35C16 37Cl 0
C-2 H 35Cl, 37Cl 02
C.2 H 35CU 37CU °2
C9F17
13C12 H 35C16 37Cl 02
17 75 77
C12 H C15 C12 °2
C12 H 35C17 37C12 0
75 77
c12 "ci7 a'ct o
C10 F17
_ 35_. 37_. _.
C12 C16 C12 °
C 35Cl 37Cl 0
C12 C17 Cl °2
C., 35Cl, 37Cl- 0,
12 6 22
13C12 35C17 37Cl 0,
13_ 35.. 37.. _
C12 C16 C12 °2
C 35Cl 37Cl 0
C12 C18 C12 °
13C = 13.003355
37Cl = 36.965903
Compound
(3)
HpCDF
HppDF
HpCDF(4)
HpCDF(4)
HpCDD
HpCDD
PFK
HpCDD (4)
HpCDD(4)
NCDPE
OCDF
PFK
OCDF
OCDD
OCDD
OCDD(4)
OCDD(4)
DCDPE
F = 18.9984
Primary
m/z?
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Compound abbreviations:
Chlorinated dibenzo-p-
dioxins
TCDD = Tetrachlorodibenzo-p-dioxin
PeCDD = Pentachlorodibenzo-p-dioxin
HxCDD = Hexachlorodibenzo-p-dioxin
HpCDD = Heptachlorodibenzo-p-dioxin
OCDD = Octachlorodibenzo-p-dioxin
Chlorinated diphenyl ethers
HxCDPE =
HpCDPE =
OCDPE
NCDPE =
DCDPE
Hexachlorodiphenyl ether
Heptachlorodiphenyl ether
Octachlorodiphenyl ether
Nonachlorodiphenyl ether
Decachlorodiphenyl ether
Chlorinated dibenzofurans
TCDF = Tetrachlorodibenzofuran
PeCDF = Pentachlorodibenzofuran
HxCDF = Hexachlorodibenzofuran
HpCDF = Heptachlorodibenzofuran
Lock mass and OC compound
PFK = Perfluorokerosene
(4) Labeled compound
(5) There is only one m/z for Cl,-2,3,7,8-TCDD (cleanup standard).
037
37
-------
Table 3A
THEORETICAL ION ABUNDANCE RATIOS AND CONTROL LIMITS
NO. Of
Chlorine
Atoms
4 <2>
5
6
6 (3)
7
7 (4)
a
m/z's
Forming
Ratio
M/M+2
M+2/M+4
M+2/M+4
M/M+2
M+2/M+4
M/M+2
M+2/M+4
Theoretical
Ratio
0.77
1.55
1.24
0.51
1.05
0.44
0.89
Control
Lower
0.65
1.32
1.05
0.43
0.88
0.37
0.76
Limitsd)
Upper
0.89
1.78
1.43
0.59
1.20
0.51
1.02
(1) Represent + 15% windows around the theoretical ion
abundance ratios.
(2) Does not apply to Cl.-2,3,7,8-TCDD (cleanup
standard).
(3) Used for 13C-HxCDF only.
(4) Used for 13C-HpCDF only.
38
038
-------
Table 4
CONCENTRATIONS OF SOLUTIONS CONTAINING LABELED AND UNLABELED PCDDS AND PCDFS -•
STOCK AND SPIKING SOLUTIONS
Compound
Labeled
Compound
Stock
Solution (1)
(ng/mL)
Labeled
Compound
Spiking
Solution (2)
(ng/mL)
PAR
Stock
Solution (3)
(ng/mL)
Cleanup
Standard
Spiking
Solution (4)
(ng/mL)
Internal
Standard
Spiking
Solution (5)
(ng/mL)
Native CDDs and CDFs
2,3,7,8-TCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDD
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDD
1,2,3,6,7,3-HxCDD
1,2,3,7,8.9-HxCDD
1,2,3.4,7,8-HxCDF
1.2,3,6,7,8-HxCDF
1,2,3.7,8,9-HxCDF
2,3,4,6,7.8-HxCDF
1,2.3,4,6,7,8-HpCDD
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDD
OCDF
Labeled CDDs and CDFs
13C12-2.3,7.8-TCDD
13
13
13,
13C12-2,3,7,8-TCDF
C12-1,2.3,7,8-PeCDD
C12-1.2,3,7,8-PeCDF
'c12-2.3,4.7,8-PeCDF
Cl2-1,2.3,4.7,8-HxCDD
C12-1,2,3,6,7.8-HxCDD
C12-1,2,3,4,7,8-HxCDF
C12-1,2,3,6,7,8-HxCDF
13C12-1,2,3,7,8,9-HxCDF
13C12-2,3,4,6,7,8-HxCDF
13C12-1,2,3,4,6,7,8-HpCDD
13C12-1,2,3,4,6f7,8-HpCOF
13ci:)-i,2,3,4,7,8f9-HpcDF
13,
13
13
"12
13,
C-2-OCDD
Cleanup Standard
37Cl4-2,3,7,8-TCDD
Internal Standards
13C12-1.2,3,4-TCDD
13
C12-1,2,3,7,8,9-HxCDD
100
100
100
100
100
100
100
100
100
100
100
100
100
100
200
40
40
200
200
200
200
200
200
200
200
200
200
200
200
200
400
400
0.8
200
200
(1) Section 6.10 - prepared in nonane and diluted to prepare spiking solution.
(2) Section 10.3.2 - prepared from stock solution daily.
(3) Precision and Recovery (PAR) standard. Section 6.14 • prepared in nonane and diluted to prepare spiking
solution in Section 10.3.4.
(4) Section 6.11 - prepared in nonane.
(5) Section 6.12 - prepared in nonane.
039
39
-------
Table 4 (continued)
CONCENTRATIONS OF SOLUTIONS CONTAINING LABELED AND UNLABELED PCDDS AND PCDFS
CALIBRATION AND VERIFICATION SOLUTIONS
Compound
Native CDDs and CDFs
2,3,7.8-TCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDD
1,2,3,7,3-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,4,6,7,8-HpCDD
1,2,3,4.6,7.8-HpCDF
1,2,3,4,7.8,9-HpCDF
OCDD
OCDF
Labeled CDDs and CDFs
13C12-2,3,7,8-TCDD
13C12-2.3,7.8-TCOF
'3C12-1,2,3,7,8-PeCDD
13Cl2-1,2,3,7,8-PeCOF
13C12-2.3,4,7.8-PeCOF
C12-1,2,3,4,7,8-HxCOD
13C12-1,2,3,6,7,8-HxCDD
13C12-1,2,3,4,7,8-HxCDF
13C,,-1,2,3,6,7,8-HxCDF
^C^-I^.S./.S^-HxCDF
13C12-2,3,4,6,7,8-HxCDF
13C,,-1,2,3,4,6,7,8-HpCOD
IOC.,-1,2,3,4,6,7,8-HpCDF
-17 \f.
C,-- 1,2.3, 4. 7,8,9- HpCDF
12 13
Cleanup Standard
37Cl4-2,3,7,8-TCDD
Internal Standards
13C12-1,2,3,4-TCDD
13C12-1,2,3,7,8,9-HxCDD
CS1
(ng/mL)
0.5
0.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
5.0
5.0
100
100
100
100
100
100
100
100
100
100
100
100
100
100
200
0.5
100
100
CS2
(ng/mL)
2
2
10
10
10
10
10
10
10
10
10
10
10
10
10
20
20
100
100
100
100
100
100
100
100
100
100
100
100
100
100
200
2
100
100
VER(6)
CS3
(ng/mL)
10
10
50
50
50
50
50
50
50
50
50
50
50
50
50
100
100
100.
100
100
100
100
100
100
100
100
100
100
100
100
100
200
10
100
100
CS4
(ng/mL)
40
40
200
200
200
200
200
200
200
200
200
200
200
200
200
400
400
100
100
100
100
100
100
100
100
100
100
100
100
100
100
200
40
100
100
CSS
(ng/mL)
200
200
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
100
100
100
100
100
100
100
100
100
100
100
100
100
100
200
200
100
100
(6) Section 14.3 - calibration verification (VER) solution.
40
040
-------
Table 5
GC RETENTION TIME WINDOW DEFINING STANDARD MIXTURES AND
ISOMER SPECIFICITY TEST STANDARD MIXTURES
DB-5 Column GC Retention Time Window Defining Standard
(Section 6.15)
Congener First Eluted Last Eluted
TCDF
TCDD
PeCDF
PeCDD
HxCDF
HxCDD
HpCDF
HpCDO
1,3,6,8-
1,3,6,8-
1,3,4,6.8-
1,2,4,7,9-
1,2,3,4,6,8-
1,2,4,6,7,9-
1,2,3,4,6,7,8-
1,2,3,4,6,7,9-
1.2,8,9-
1,2,8,9-
1,2,3,8,9-
1,2,3,8,9-
1,2,3,4,8,9-
1,2,3,4,6,7-
1,2,3,4,7,8,9-
1,2,3,4,6,7,8-
DB-5 TCDD Isomer Specificity Test Standard
(Section 6.16.1)
1,2,3,4-TCDD 1,2,3,7-TCDD
1,2,7,8-TCDD 1,2,3,8-TCDD
1,4,7,8-TCOD 2,3,7,8-TCDD
DB-225 Column TCDF Isomer Specificity Test Standard
(Section 6.16.2)
2,3,4,7-TCDF
2,3,7,8-TCDF
1,2,3,9-TCDF
04i
41
-------
Table 6
REFERENCE COMPOUNDS FOR QUANT ITAT IOH OF NATIVE AND LABELED PCODS AND PCDFS
Native PCDDs and PCDFs
Reference Compound
1
1
1
1
1
1
1
1
1
2
p
,2
.2
1,2
1.2
2,3
.2.3
.2.3
? 3
? 3
? 3
.2.3
,3.4
1 4
,3.4
.3.4
2,3,
2,3,
,3,7
,3,7
,4,
,4,
,6,
7
4
f>
,7,
,6,
6
.6,
.7.
7
7
7
8
7
7
3
7
7
7
8
7,3-TCDD
7,3-TCDF
,8-PeCDD
,8-PeCDF
,S-PeCDF
,8-HxCDD
,8-HxCDD
9-HxCDD
8-HxCDF
8-HxCDF
,9-HxCDF
,8-HxCDF
8-HpCDD
,8-HpCDF
,9-HpCDF
OCDD
OCDF
13c
13c
1-*
1
13c
13c
13c
1-1
13c
13c
c
13C
-C12
13C12
C
C12
JC
12
12
12
12
12
12
-1
-1
-1
13C .2
12 '
13
13C12-2,
12-1.2.3
12-1,2,3
12-2
-1 ?
-1 ?
-1 ?
-1 ?
-1 ?
-2.3
2 3
.2.3
.2.3
.3,4
T 4
1 6
3 4
T 0
3 7
.4.6
4 6
,4.6
,4,7
3, 7,8-TCDD
3,7,8-TCDF
,7,8-PeCDD
,7,8-PeCDF
,7,
7
7
7
7
a
,7,
7
.7,
.3,
n
8-PeCDF
8-HxCDD
8-HxCDD
(1)
8-HxCDF
8-HxCOF
9-HxCDF
8-HxCDF
8-HpCDD
8-HpCDF
9-HpCOF
"C1?-OCDD
c
12-OCDD
(1) 1,2,3,7,8,9-HxCOD is quantified using the average
1,2,3,6,7.8-HxCDD.
Labeled PCDDs and PCDFs
13C,,-2. 3, 7,8-TCDD
13
IJC12-2,3,7,8-TCDF
13C12-1,2,3,7,8-PeCDD
13C12-1.2,3,7,8-PeCDF
13C12-2,3,4,7,8-PeCDF
13C12-1,2,3,4,7,8-HxCDD
13C12-1,2,3,6,7,8-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1.2,3,4,7,8-HxCDF
13C,,-1,2,3,6.7,8-HxCDF
13C12"2,3,4,6.7,8-HxCDF
13C12-1,2,3,4,6.7,8-HpCDD
13C12-1,2,3,4,6,7,8-HpCDF
12 13^ .QCDD
37Cl4-2, 3, 7,8-TCDD
ponses for the 13C,--1,2,3,4,
Reference Compound
13C.,-1,2,3,4-TCDD
13
C12-1,2,3,4-TCDD
13C12-1,2,3,4-TCDD
13C12-1,2,3,4-TCDD
13C12-1,2,3,4-TCDD
13C.,-1,2,3,7,8,9-HxCOD
.-12
C.2-1,2,3,7,8,9-HxCDD
13Cl2-1,2,3,7,8,9-HxCDD
13C.,-1,2,3,7,8,9-HXCDD
12
C12-1,2,3,7,8,9-HxCOD
13C12-1.2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,7,8,9-HxCDD
13C.,-1,2,3,7,8,9-HxCDD
13 12 ' ' ' ' ' D
°C12-1,2,3,7,8,9-HxCDD
13C12-1,2,3,4-TCDD
,7,8-HxCDD and 13C.--
42
042
-------
Table 7
ACCEPTANCE CRITERIA FOR PERFORMANCE TESTS
Compound
2,3,7,8-TCDO
2,3,7.8-TCDF
1.2,3,7,8-PeCDD
1,2,3,7,8-PeCDF
2.3,4,7,8-PeCDF
1,2.3,4,7,8-HxCDD
1,2.3,6,7,8-HxCDD
1,2,3,7.8,9-HxCDD
1.2.3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
1,a,3,7,8,9-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3.4.6,7,8-HpCDD
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCOF
OCDD
OCOF
13C12-2,3.7,8-TCDD
13C12-2.3,7,8-TCDF
1X
l:>C12-1,2.3,7,8-PeCDD
13C12-1.2.3,7,8-PeCDF
13C12-2.3,4.7.8-PeCDF
•]•»
C,,-1, 2,3,4.7, 8-HxCDD
l:>C.,-i.2.3,6,7,8-HxCDD
.,12
C,,-1,2,3,4,7,8-HxCDF
•!•» '^
'JC..,-1,2,3,6.7,8-HxCDF
.,12
°C,, -1,2.3. 7,8, 9-HxCDF
.,12
C.--2,3,4,6,7,8-HxCDF
13 12
'^.,-1,2,3,4,6,7,8-8^00
C.,-1,2,3,4,6.7,8-HpCDF
12 ...... K-
C., -1,2,3,4, 7.8, 9-HpCDF
13
c.2-ocoo
37Cl4-2,3.7.8-TCDD
Test
Cone. (1) s
(ng/tnL) (ng/mL)
10
10
50
50
50
50
50
50
50
50
50
50
50
50
50
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
200
10
1.5
2.0
4.2
4.6
4.2
5.5
5.5
9.5
6.3
4.0
4.0
5.0
6.4
3.6
4.2
13.0
45.0
-
-
-
-
-
-
.
-
-
-
.
-
-
-
-
-
IPR (2)
X
( ng/mL )
3.9 -
3.2 -
47.5 -
44.2 -
45.3 -
30.9 -
33.2 -
22.7 -
25.2 -
39.1 -
37.9 -
27.4 -
27.4 -
39.5 -
36.6 -
69.4 -
46.1 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
50.0 -
2.5 -
20.6
26.8
50.5
54.0
50.3
70.2
65.9
90.9
92.0
54.4
62.9
85.5
76.5
62.1
64.9
154.6
139.8
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
300.0
15.0
OPR(2)
(ng/mL)
5.9 -
6.6 -
35.6 -
36.7 -
37.8 -
35.1 -
33.3 -
31.8 -
36.9 -
34.8 -
37.1 -
35.7 -
37.5 -
37.4 -
36.9 -
75.6 -
69.5 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
25.0 -
50.0 -
2.5 -
14.2
12.7
58.1
57.3
56.9
60.4
64.4
61.2
58.8
58.8
55.7
60.0
56.8
60.5
60.6
118.7
127.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
150.0
300.0
15.0
VER
(ng/mL)
8.6 -
8.8 -
44.2 -
46.7 -
47.2 -
37.6 -
39.7 -
42.6 -
41.5 -
40.5 -
45.7 -
44.1 -
41.6 -
43.1. -
43.6 -
87.5 -
83.9 -
90.0 •
87.7 -
80.6 -
81.8 -
83.0 -
76.1 -
84.0 -
85.2 -
85.0 -
89.5 -
85.7 -
82.2 -
88.5 -
89.0 -
164.2 -
6.1 -
11.6
11.3
56.6
53.5
53.0
66.5
63.0
58.7
60.2
61.7
54.5
56.7
60.2
58.0
57.3
114.4
119.2
111.2
114.0
124.0
122.3
120.5
131.3
119.1
117.4
117.7
111.7.
116.7
121.6
113.1
112.4
243.6
11.6
(1) All specifications are given as concentrations in the final extract or standard solution.
(2) s = standard deviation of the concentration; X = average concentration. Concentration limits for labeled
compounds in IPR and OPR aliquots are based on requirements for labeled compound recovery of 25-150% (Sections
8.2.3 and 14.5.3).
043
43
-------
Table S
SAMPLE PHASE AND QUANTITY EXTRACTED FOR VARIOUS MATRICES
Sample Matrix (1)
SINGLE PHASE
Aqueous
Solid
Organic
MULTIPHASE
Liquid/Solid
Aqueous/solid
Example
Drinking water
Groundwater
Treated wastewater
Dry soi I
Compost
Ash
Waste solvent
Waste oil
Organic polymer
Wet soil
Percent Quant i ty
Solids Phase Extracted
<1 (2) 1000 mL
'1
>20 Solid 10 9
<1 Organic 10 g
1-30 Solid 10 g
Organic/solid
Liquid/Liquid
Aqueous/organ ic
Aqueous/organic/
solid
Untreated effluent
Digested municipal sludge
Filter cake
Paper pulp
Tissue
Industrial sludge
Oily waste
tn-process effluent
Untreated effluent
Drum waste
Untreated effluent
Drum waste
1-100
>1
Both
Organic
Organic
& solid
10 g
10 g
10 g
(1) The exact matrix may be vague for some samples. In general, when the CDDs and CDFs are in contact with a
multiphase system in which one of the phases is water, they will be preferentially dispersed in or adsorbed
on the alternate phase, because of their low solubility in water.
(2) Aqueous samples are filtered after spiking with labeled analogs. The filtrate and the material trapped on
the filter are extracted separately, and then the extracts are combined for cleanup and analysis.
44
04
-------
PERFORMANCE EVALUATION OF METHOD 1613
. High Resolution Gas Chromatography/High Resolution Mass Spectrometry
Determination of Tetra- through Octa-Chlorinated Dibenzo-p-Dioxins and Dibenzofurans
by Isotope Dilution
March 1990
USEPA OFFICE OF WATER
OFFICE OF WATER REGULATIONS AND STANDARDS
INDUSTRIAL TECHNOLOGY DIVISION (WH-552)
SAMPLE CONTROL CENTER
Washington, DC 20460
'i' J
-------
TABLE OF CONTENTS
Section
Introduction
Determination of Method Requirements
Method Selection and Initial Development
Single Laboratory Testing
Confirmatory Testing
Interim Method Testing
Field Validation Study
Conclusions
Outliers
Estimation of Variance Components
Derivation of Quality Control Limits for Accuracy
Derivation of Quality Control Limits for Precision
Calibration Linearity
Relative Retention Time
Method 1613,12 September 1988 Draft
Method 1613, July 1989
Page Number
1
1
2
4
11
12
14
19
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E
Appendix F
Appendix G
Appendix H
04G
-------
UST OF TABLES
Table Description
1 Comparison of Candidate Methods
2 Soxhlet/Dean Stark versus Soxhlet t Test Results
3 Method Detection Limit
4 1613 Field Validation Study - Samples by Industry and Lab
5 Initial Precision and Recovery - All Labs
6 Ongoing Precision and Recovery - All Labs
7 Precision and Recovery of Labeled Compounds in EPA Samples
8 Calibration Linearity QC Limits
9 Calibration Verification QC Limits
10 Relative Retention Time QC Limits
11A Initial Precision and Recovery QC Limits for Accuracy
11B Initial Precision and Recovery QC Limits for Precision
12 Ongoing Quality Assurance Limits
13 Precision and Recovery of Labeled Compounds - EPA Samples, All Labs
14 Precision and Recovery of Labeled Compounds - EPA Samples, Labs 2
and 3
15 QC Limits for Precision and Recovery of Labeled Compounds - EPA
Samples
047
-------
LIST OF FIGURES
Figure Description
1 Initial Precision and Recovery of Labeled and Native
Compounds - Frequency Distribution Plot of All Labs Combined
2 Ongoing Precision and Recovery of Labeled and Native
Compounds - Frequency Distribution Plot of All Labs Combined
3 Precision and Recovery of Labeled Compounds
Frequency Distribution Plot by Lab
4 Precision and Recovery of Labeled Compounds
Frequency Distribution Plot of All Labs Combined
111
048
-------
1. INTRODUCTION
In 1976 the U.S. Supreme Court issued a consent decree requiring the Environmental Protection
Agency (EPA) to measure and limit 65 compounds and classes of compounds in effluents discharged to
receiving waters in the United States. The list of 65 was subsequently refined by EPA to a list of 129 specific
analytes termed the "Priority Pollutants" and codified as the Section 307(a) list of "toxic pollutants" in the
1977 Clean Water Act (CWA) amendments. Priority Pollutant Number 129 is 2,3,7,8 -tetrachlorodibenzo-
p-dioxin (23,7,8-TCDD), one of the most toxic substances known to man.
Currently, EPA Method 613 is the only method for the analysis of 2,3,7,8-TCDD that has been
promulgated under CWA Section 304(h). This method, developed in the late 1970's, utilizes gas
chromatography and low resolution mass spectrometry, with minimal extract cleanup steps. The detection
limit for 2,3,7,8-TCDD in water is 2000 ppq for Method 613. As a result, it does not achieve the detection
limit for 23,7,8-TCDD currently required by the Agency for analysis of treated effluents (10 - 25 ppq).
Method 1613 was developed by the Agency in response to the need for analyses of treated effluents at
low levels of 23,7,8-TCDD (10 - 25 ppq). It was designed for regulatory development purposes and
compliance monitoring under the National Pollutant Discharge Elimination System (NPDES, CWA Section
402).
Method 1613 is a survey method designed to quantitate 23,7,8-TCDD as well as the other sixteen
23,7,8 substituted dioxins and furans. It employs gas chromatography coupled with a state-of-the-art high
resolution mass spectrometric method for the analysis. Method 1613 incorporates the Agency's 500/600
series QA/QC program, including rigorous start-up tests and on-going demonstrations of laboratory
performance. It employs isotope dilution as a means of quantifying the analytes of interest. Isotope dilution
is a technique in which stable isotopically labeled analogs of the target compounds are added to each sample
and used to quantify the native analytes present, in order to reduce the variability of the analysis and correct
for recovery bias.
This report presents results of various single-laboratory development studies, sample and QA/QC
analyses performed to date using Method 1613. The objective is to demonstrate that the Method achieves
acceptable precision and bias across a range of industrial effluents and contract laboratories, and is therefore
a valid method for promulgation as a Section 304(h) alternate test procedure for the analysis of 2,3,7,8-
TCDD. A secondary objective is the promulgation of Method 1613 under Section 304(h) for the analysis of
the 23,7,8 substituted dioxin and furan isomers other than 2378-TCDD.
2. DETERMINATION OF METHOD REQUIREMENTS
The CWA (as amended 1987) requires the Agency to consider water quality-based effluent
limitations in conjunction with traditional Best Available Technology (BAT) efforts. The Agency's water
quality criteria for 23,7,8-TCDD of 0.013 ppq (at a 10"6 estimated human cancer risk) is below the limit of
detection of current analytical techniques.
Section 304(m) of the CWA requires the Agency to periodically review existing effluent guidelines,
and to develop guidleines for new industries and new compounds. In response to this requirement, the
Office of Water (OW) will address the issue of regulation of 23,7,8-TCDD in the treated effluents of several
industries. In order to collect data to support guideline development and support current regional NPDES
efforts, OW must measure 23,7,8-TCDD in treated effluents at the lowest levels practical using available
analytical technology. Further, methods utilized by OW must contain a rigorous QA/QC program
consistent with the 600/1600 series methods that have been promulgated under 304(h).
040
-------
Therefore, the method implemented by the Office of Water for the measurement of 2,3,7,8-TCDD
must meet or exceed the following data quality objectives (DQOs):
o The method must be appropriate for nationwide use by EPA and contract laboratories as well as
by the regulated community. This requires a method that is sufficiently rugged that minor
variations in use do not significantly affect the usability (precision and bias) of the data obtained.
o The method must utilize high resolution gas chromatography combined with high resolution mass
spectrometry (HRGC/HRMS). The HRMS instrument must have a resolving power in excess of
10,000 atomic mass units.
o Sample preparation procedures must minimize sample handling and therefore reduce the
potential for contamination.
o Extraction and cleanup procedures must be sufficient to allow removal of interferences so that
target detection limits can be reached.
o Isotope dilution quantitation must be utilized to allow precise estimates of analytical variability,
as this technique has proven useful in the development of BAT effluent limits for other
compounds.
o A Limit of Quantitation (LOQ) for 2^,7,8-TCDD of 1 ppt for solid samples and 10 ppq for
effluents is required.
o In order to collect data on other potentially toxic dioxins and furans, the method must measure
the 16 other 2,3,7,8 substituted dioxins and furans.
Due to the recent discovery of 2,3,7,8-TCDD in pulp and paper export vectors, EPA Regional
personnel are currently working on NPDES permits for 2,3,7,8-TCDD, and require a standard method for
nationwide use. Industry representatives argue that the Agency can not require utilization of a method that
is not promulgated under 304(h), and therefore, the only method that can be utilized is Method 613.
However, given that the MDL published in Method 613 is 2000 ppq, Method 613 would be inappropriate for
monitoring a pollutant for which the water quality criteria is five orders of magnitude lower (0.013 ppq).
3. METHOD SELECTION AND INITIAL DEVELOPMENT
After reviewing OWs method requirements, several meetings were conducted with EPA Regional
personnel and various dioxin experts from commercial and industrial laboratories. Concurrent with these
meetings, ITD Sample Control Center staff reviewed seven candidate methods and reviewed data for 2,3,7,8-
TCDD analysis from contract laboratories using various existing methods. The information shown in Table
1 is a result of these efforts.
050
-------
TABLE 1 - COMPARISON OF CANDIDATE METHODS
METHOD POSITIVE ATTRIBUTES NEGATIVE ATTRIBUTES
613 304(h) Promulgation Low Resolution MS
500/600 Series QA/QC MDL2000ppq
Limited to 2,3,7,8-TCDD
ERL-Duluth Extensive use Used only by ORD lab
MDL adequate Primarily Fish Tissue
Wright State Extensive use on Used primarily by WSU
University pulp and paper matrix Labor intensive, High cost
Hybrid GC column
Limited to 2,3,7,8-TCDD/F
NCASI Appropriate matrices Used only by NCASI
High Cost
Limited to 2,3,7,8-TCDD/F
Dow Chemical Extraction & Cleanup Low Resolution MS
Used only by Dow
Limited to 2,3,7,8-TCDD/F
8290 Extensive use Published version is "draft",
Extensively modified,
Not standardized across labs
Triangle Highly refined 8290 "Proprietary"
Laboratories
"8290x"
After review of the candidate methods, OW decided to develop a method specifically designed to
meet the data quality objectives listed in Section 2 above. The initial version of Method 1613 was prepared
by combining the "best" features of the candidate methods and adding other features required to meet the
stated data quality objectives. Those features included the following:
o The EMSL-Cincinnati format and the 600 series QA/QC Program were taken from Method 613
o Sample preparation features were taken from the ERL-Duluth, Wright State University, NCASI,
Dow, and 8290 methods
o Chromatography and mass spectrometry specifications were taken from 8290
o Isotope dilution quantitation was specified, based on ITD's lengthy experience with Method 1625
The initial draft of Method 1613 is dated 12 September 1988, and employs isotope dilution
techniques and a 600 series QA/QC program (Attached as Appendix G).
051
-------
4. SINGLE LABORATORY TESTING
Two concurrent single laboratory tests were initiated by ITD during the early development of Method
1613. The first test involved the use of a new extraction technique for solid matrices. The second test
involved the use of a new extraction technique for aqueous matrices. Both techniques were taken from work
performed at Dow, and were modified for use in Method 1613. A third single laboratory test was initiated by
ITD to determine the Method Detection Limit (MDL) of the July 1989 version of the method.
4.1 Soxhlet-Dean Stark (SDS) Extraction Evaluation
A soxhlet extraction procedure is specified in many analytical methods for the extraction of
polychlorinated dibenzo-/?-dioxins and polychlorinated dibenzofurans (PCDDs/PCDFs) from solid and
semi-solid matrices such as soil and sludge. Basically, the procedure involves the repeated refluxing of an
organic solvent such that the solvent percolates through the sample matrix and extracts the compounds of
interest by dissolution.
Typically, when using soxhlet extraction, steps must be taken to remove the water from the sample
matrix prior to extraction because the organic solvents used for extraction are not water miscible. This is
particularly true when extracting samples with high moisture contents such as sludge. Techniques such as
filtration and centrifugation have been employed to remove the water in other analytical procedures. The
addition of sodium sulfate to adsorb the water from the sample has been extensively employed in the
analysis of organic compounds from environmental matrices. Each of these techniques involves additional
handling of the sample, and therefore increases the potential for introduction of contaminants, loss of
analytes, or loss of the entire sample. Since each sample handling step has the potential to increase the
variability of the data produced by the overall analytical method, it is critical to minimize sample handling
steps when dealing with 2,3,7,8-TCDD at the levels addressed in the Method 1613 (10-25 ppq).
Another technique for removing water from a sludge sample and other types of wet samples involves
the use of a Dean Stark water separator in conjunction with a soxhlet extractor. This piece of glassware fits
between the soxhlet extractor and the condenser, and offsets the condenser to one side. This offset provides
space for a dog leg that drops straight down from the bottom of the condenser. Ground glass joints connect
the separator to the soxhlet extractor and the condenser, and the dog leg ends in either a graduated receiving
tube or a stopcock. The resulting combined apparatus is referred to as the Soxhlet-Dean Stark apparatus,
abbreviated as SDS.
Water is removed from the sample during extraction by the process of azeotropic distillation.
Originally designed for processes involving xylene, the separator works equally well with toluene, the
extraction solvent specified in Method 1613, because toluene and water form an azeotrope that boils at a
temperature of 85 °C. When the azeotropic vapor condenses, the liquid drops into the dog leg, where the
toluene floats on top of the more dense water, and eventually flows back into the distilling flask. The use of
a separator with a stopcock allows the laboratory to draw off the collected water without interrupting the
extraction process.
The SDS extraction technique offers the potential for enhancing analytical precision and decreasing
bias by reducing the number of sample handling steps and more effectively extracting the sample. It has a
noteworthy advantage over simple soxhlet extraction in that the percent moisture in the sample may be
determined directly from the sample being extracted, rather than from another aliquot of the sample which
may not be truly representative of the aliquot that is extracted.
The SDS combination offers a significant advantage over the use of sodium sulfate to adsorb the
water from the sample matrix. Sodium sulfate drys the sample by hydrating itself with the water in the
0-!
-------
sample. During the process of extraction, some of this water of hydration may be lost back to the sample or
to the solvent and, as a result, the dehydrated sodium sulfate may seal off pores in the surface of the solid
matrix. This process effectively traps the analytes of interest within the matrix, thus preventing their
extraction.
Another potential problem with the use of sodium sulfate in the analysis of very low levels of
PCDDs/PCDFs is the loss of analytes by absorption on contaminants in the reagent itself. The reagent is
typically purified of any organic contaminants by heating it in a muffle furnace at high temperatures. Any
organic material present is charred, often giving the reagent a light gray cast. While the heat treatment
effectively prevents this organic material from being extracted from the sodium sulfate, PCDDs/PCDFs are
strongly adsorbed by activated carbon, and the charred material represents a source of activated carbon.
Because of the potential loss of analytes through the use of sodium sulfate, ITD has chosen to avoid the use
of this reagent during the extraction of samples.
Although Dow had published data on the use of the SDS in other solid matrices, the first phase in the
validation of Method 1613 consisted of an intralaboratory study to ascertain the comparability of the SDS
procedure with the more commonly used soxhlet procedure when applied to municipal sewage sludge.
4.1.1 Experimental Design
To demonstrate the comparability of the SDS procedure proposed for EPA Method 1613 with the
soxhlet procedure currently employed in other analytical methods for PCDDs/PCDFs, a five gallon sample
of sewage sludge was sent to Triangle Laboratories for analyses. ITD SAS 131, Episode 1519 consisted of
nine analyses of the sludge sample, as follows:
- 3 analyses of the unspiked sludge, extracted by soxhlet
- 3 analyses of sludge spiked with PCDDs/PCDFs, extracted by soxhlet
- 3 analyses of sludge spiked with PCDDs/PCDFs, extracted by SDS
A preliminary examination of the unspiked sludge data indicated that very few of the PCDDs/PCDFs
were detected in these samples. Hence, it would have been impossible to collect statistically meaningful
results from an experiment in which the unspiked sludge was extracted by both procedures. Therefore, the
unspiked data were not considered further.
The sludge was spiked by the laboratory with all 17 2,3,7,8-substituted PCDD/PCDF isomers prior to
extraction. Replicate aliquots were extracted by soxhlet alone and by the SDS procedure. Because this work
was done during the earliest stages of development of Method 1613, Triangle Laboratories utilized Method
"8290x" for the analyses. (For the purpose of testing the SDS procedure, this instrumental aspects of "8290x"
were deemed to be sufficiently similar to Method 1613). The resulting data from Triangle laboratories were
evaluated to determine if the SDS procedure provided comparable or better results than the soxhlet
procedure alone.
Data for all the samples were reviewed to determine if they met the requirements of Method "8290x"
for identification of analytes. All data were consistent with those requirements.
4.1.2 Statistical Results
Because the extraction procedures were compared in a controlled experiment rather than during the
routine analyses of field samples for other purposes, a more rigorous statistical analysis of the results was
possible. The results of the triplicate analyses of the spiked sludge by both the soxhlet and SDS procedures
were compared using standard statistical techniques. The mean concentrations of the three measurements
of each isomer were determined for each extraction procedure. For each isomer, the mean concentration for
053
-------
the SDS procedure was compared to the mean concentration for the Soxhlet procedure using a two-tailed r-
test. The r-test was repeated for each of the 17 isomers.
The results of those tests are given in Table 2. The table lists the mean concentration of each isomer
for the samples extracted by both methods and the variance (s2) of each isomer by each method. The t
statistic was calculated for each pair of means, and is given in Table 2. The value of t for each pair of means
was used to determine the probability, P(r,4), that the mean concentrations of each isomer determined by
the different extraction procedures were not the same. The null hypothesis was that the means were not
different The null hypothesis would be rejected, i.e. the means were different, if the probability value that
was calculated was less than 0.05.
All 17 isomers were detected in all six samples, with one exception. 1,2,3,4,7,8,9-HpCDF was not
detected in one of the soxhlet-extracted replicates. In the case of this isomer, the r-test was performed two
ways: once using the estimated detection limit (EDL) calculated by the laboratory in place of the
undetected concentration; and once using a value of 0.0 for the undetected concentration. The results of
both r-tests appear near the bottom of Table 2.
Only two analytes yielded values for P(r,4) that were less than 0.05. Those analytes were 1,2,3,4,6,7,8-
HpCDD, with P equal to 0.038, and 23,7,8-TCDF, with P equal to 0.032. OCDD had a P value of 0.060, but
none of the results for the other analytes were close to the 0.05 level. The r-test results indicate that, for
1,2,3,4,6,7,8-HpCDD and 2,3,7,8-TCDF, there is a statistically significant difference in the mean
concentrations of these analytes extracted by the two procedures. In both cases, the SDS procedure yielded
a higher mean concentration than the soxhlet alone. Therefore, it would appear that for at least two
analytes, the SDS extraction procedure gave higher results than extraction by soxhlet alone, and for the
remaining 15 analytes, the SDS results comparable to those from the soxhlet alone.
4.13 Limitations
The r-test used for these data assumes that the variances of the two means being compared are equal,
and can be estimated from the sample data. All the r-tests use a value of 4 for the number of degrees of
freedom for the test, i.e. 3+3-2. A review of the data in Table 2 suggests that the variances may not be equal
and that, in general, the variance of the SDS data is less than that of the Soxhlet data (12 of 17 compounds
had lower variances by SDS). The notable exceptions to this trend are the variances for the two isomers
which were statistically different between the two extraction methods. The variances for 1,2,3,4,6,7,8-
HpCDD and 2,3,7,8-TCDF were higher for the SDS data.
Therefore, r-tests were performed for several of the isomers assuming that the variances are not
equal. This version of the r-test calculates the same r statistic as is reported on Table 2, but uses the sample
variances and the numbers of observations for each isomer (n) to determine a different number of degrees of
freedom (n') than is used when the variances are assumed to be equal. This new value for the degrees of
freedom was calculated for four of the isomers, and resulted in n'<0. Values of n' less than zero have no
validity, indicating that there are not enough data to perform this variation of the r-test.
The method used by Triangle Laboratories also includes isotopically labeled "surrogates". Unlike the
labeled analogs used in an isotope dilution method such as Method 1613, the surrogates are not used for
quantifying the target analytes, but may be used to judge the effectiveness of the sample extraction
procedures. Triangle reported data for the recovery of these surrogates and, in general, the recoveries were
good for both extraction methods. Two-tailed r-tests were performed on some of the surrogate recovery data
as well. No statistically significant differences between the mean recoveries of the surrogates could be
found, presumably also due to the small size of the data set.
In an attempt to overcome the small number of observations in the data set, the data for most of the
native isomers were grouped together and tested. Eleven of the 17 spiked isomers had mean concentrations
near 200 ppt. Data for these isomers were retained in the second data set tested.
054
-------
TABLE 2 - SOXHLET/DEAN STARK VERSUS SOXHLET t TEST RESULTS
SDS
Soxhlet
Isomer
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,23,4,6,7,8-HpCDD
OCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
23,4,7,8-PeCDF
1,23,4,7,8-HxCDF
1,23,6,7,8-HxCDF
23,4,6,7,8-HxCDF
1,23,7,8,9-HxCDF
1,23,4,6,7,8-HpCDF
1,23,4,7,8,9-HpCDF
mean cone.
34.449
200.503
232.960
212.913
220.462
496.677
3717.909
58.707
204.043
170.667
248.912
223.459
217.313
181.578
262.347
174.869
s2
6.121
15.124
55.071
43.294
8.738
5342.758
9.17 105
16.595
4.826
22.333
5.510
60.549
90.934
91.152
335.027
452.480
mean cone.
37.410
209.472
236.235
218.159
218.760
322.147
1711.635
48.494
206.290
183.667
252.408
222.163
228.919
178.551
242.249
160.569
s2
3.486
90.973
47.987
356.928
59.923
1039.967
2.37 105
3.228
73.55
37.333
188.022
451.416
358.418
222.782
1320.180
1283.639
t
-1.351
-1.231
-0.456
-0.371
0.290
3.089
2.64
3.244
-0.359
-2.380
-0.355
0.081
-0.774
0.242
0.698
0.485
Pfr.41
0.125
0.288
0.673
0.730
0.780
0.038 *
0.060
0.032 *
0.756
0.078
0.756
0.940
0.482
0.822
0.524
0.653
(using EDL of 189.0 for missing soxhlet value)
1,2,3,4,7,8,9-HpCDF 174.869 452.480 97.330 7751.0
(using 0.0 for missing soxhlet value)
OCDF 478.340 814.661 511.207 49760
* indicates analytes whose mean concentrations are significantly different at the 0.05 level.
All concentration data are given in parts-per-trillion, as reported by the laboratory
1.211
0.207
0.292
0.850
055
-------
Two of the other six isomers had mean concentrations that were much lower or much higher than 200
ppt. OCDD was found at relatively high concentrations (900-1900 ppt) in the unspiked sample aliquots, and
1,2,3,4,6,7,8-HpCDD was consistently found in the unspiked aliquots at levels around 100 ppt. Of the
remaining four isomers, 2,3,7,8-TCDD, 2,3,7,8-TCDF, and OCDF were not consistently found in the
unspiked aliquots, and 1,2,3,4,7,8,9-HpCDF was the isomer which was not detected in one of the spiked
aliquots. Therefore, data for these six isomers were not included in the second data set.
The data for the three replicate extractions by the two extraction procedures for the eleven remaining
isomers yielded 33 data points for each extraction method. Despite this number of data points, the mean
concentrations of all eleven isomers combined were not significantly different for the two procedures at the
0.05 probability level. The variation in the concentrations of each isomer due to the instrumental analysis
itself may be sufficiently large to obscure any difference between the extraction procedures.
4.1.4 Conclusions
Based on this study, the data for 1,23,4,6,7,8-HpCDD and 2,3,7,8-TCDF do indicate a significant
difference between the extraction procedures, with the SDS extraction yielding higher mean concentrations
(recoveries) of these two isomers. However, one cannot differentiate between the means of the
concentrations of 15 of the spiked analytes determined using the two extraction procedures. The lack of
differentiation between those mean concentrations may reflect the small size of the data sets tested.
Although the strength of this conclusion is limited by the size of the data set, the data indicate that
the SDS procedure is as good an extraction method as the soxhlet procedure alone, and better for at least
two isomers. Given this, and the advantages that fewer sample handling steps are involved in the SDS
procedure and the percent solids content of the sample may be determined directly from the aliquot
extracted for analysis, the SDS procedure was incorporated into Method 1613.
4.2 Erlenmeyer Flask Extraction Evaluation
The September 1988 draft of Method 1613 contained procedures for extracting aqueous samples
through the use of a continuous liquid-liquid extractor. This procedure has been used in numerous other
Agency methods for extraction of organic compounds from aqueous samples.
The continuous liquid-liquid extraction procedure was criticized by various reviewers of the draft
method for several reasons. The commercial laboratories argued that it was expensive and time-consuming
to perform. LTD was less concerned about the time and cost elements if the procedure would result in a
significant improvement in data quality, as measured through the precision and accuracy of labeled
compound recoveries.
Representatives from Dow Chemical argued that they had seen only minimal improvements in
precision through the use of the continuous liquid-liquid extraction procedure, and that it posed a large risk
for contaminating the samples. The continuous liquid-liquid extractors are difficult to adequately clean
between samples, and thus at the very low levels of interest for PCDDs/PCDFs, could pose a significant risk
of cross-contamination between samples. Dow suggested that ITD adopt a procedure that Dow had used for
many years for extraction of aqueous samples. Their procedure involves extracting the sample in the
original bottle by adding a lighter than water solvent and a magnetic stirring bar to the bottle, and drawing a
vortex of solvent into the sample for several hours. The advantages appeared to be simplicity and lack of
sample transfer steps, thus lowering the risk of contamination. As practiced by Dow on their own samples,
the extraction is sufficiently efficient that the water and the used sample bottle are essentially "clean", i.e.
free of measurable PCDDs/PCDFs, and thus the used bottle containing the extracted water is simply capped
and discarded.
0
5G
-------
The use of a lighter than water solvent, typically toluene or benzene, had additional advantages.
After extracting the aqueous phase of the sample (the filtrate), the actual volume of solvent containing the
material extracted from the filtrate would be added to the SDS apparatus and used to extract the particulates
collected on the filter. Thus, the two fractions are extracted with one portion of solvent, and there is no
need to combine two separate extracts. Using the same solvent also eliminates the need for a solvent
exchange step that would be required when a heavier than water solvent and a lighter than water solvent are
combined.
As a result of these discussions, the use of an erlenmeyer flask and a magnetic stirring bar to extract
water samples was incorporated in the second draft of Method 1613. Given the time constraints of the ITD
method development activities at the time, no separate evaluation study was designed. Rather, two
laboratories performing analyses for ITD during the 304(m) Pulp and Paper guidelines review utilized this
procedure for aqueous samples. However, several problems became apparent immediately. They included:
o Need to filter the samples before extraction
o Difficulties in having samplers leave sufficient room in the bottle for the solvent
o Difficulty with the quantitative transfer of the extract out of the bottle
Dow has much more control over the entire sampling and analytical process, and thus were not concerned
about filtering samples beforehand. They know that their aqueous samples are very low in solids, and can
control how much water is added to each bottle during sampling. The quantitative transfer of the extract is
of less concern to Dow because they do not use a true isotope dilution method for quantifying the analytes.
In order to accommodate the need for filtration of the samples before extraction, ITD devised a
scheme to filter the sample through a glass fiber filter directly into an erlenmeyer flask of approximately 2
liter volume. About 100 mL of toluene was added to this flask along with a magnetic stirring bar. The flask
was then placed on a magnetic stirrer, and extracted for about an hour. While this worked on many samples,
there were a significant number of samples where emulsions prevented the effective extraction of the
samples. In these instances, multiple extractions were performed and the extracts added together.
ITD anticipated difficulty in having a variety of sampling crews leave 100-150 mL of empty headspace
in a nominal 1 liter sample bottle. Both ITD and Dow ruled out the possibility of withdrawing and
discarding a representative aliquot of the sample from a completely filled bottle in order to make room for
the extraction solvent. The use of erlenmeyer flask also solved the potential problem of the samplers
overfilling the sample bottles, thereby not leaving space for the addition of the solvent.
Despite the numerous changes made to the extraction procedure, neither commercial laboratory was
able to reliably achieve recoveries of the isotopically labelled compound added to each sample before
extraction that met the QC criteria (25-150% recovery) specified in the draft method . One laboratory had
very poor results, and was not interested in its further use. The second laboratory made numerous efforts to
make the procedure work. They attempted to use multiple extractions, combining three 100 mL solvent
extracts before cleanup and analysis. They encountered difficulties with several darkly-colored samples,
where they could not see the vortex through the sample itself, which led to poor recoveries if the sample was
"understirred11, and led to severe emulsions if the sample was "overstirred".
The quantitative transfer of the extract also caused some problems. The Dow procedure involves
adding reagent water to the bottle (or flask, in this case), until the solvent is driven up into the narrow neck.
From the neck, it is simply drawn off in a clean disposable pipette. Additional reagent water is added to
ensure all the solvent is recovered. Unfortunately, in practice, complete transfer is not possible, and the loss
of the labeled compounds contained in whatever volume of solvent is not collected caused the recoveries of
the labeled compounds to fall below the method specifications. Because Dow does not practice isotope
dilution quantitation, their method does not contain such rigid specifications for labeled compound
recovery.
057
-------
In August and September 1989, ITD met with both laboratories, and discussed the problem with
several other laboratories familiar with PCDD/PCDF analyses at the Dioxin '89 meeting in Toronto. As a
result, the erlenmeyer vortex extraction procedure was eliminated, in favor of the more traditional
separately funnel extraction procedure. However, the separately funnel extraction requires a heavier than
water solvent. While easily accommodated in the extraction of water samples, the use of methylene chloride
for extraction precluded the possibility of using the same actual volume of solvent in the SDS extraction of
the filtered paniculate fraction. Therefore, Method 1613 was rewritten, resulting in a solvent exchange of
the methylene chloride to toluene, and combination of the two extracts prior to further cleanup.
43 Method Detection Limit (MDL) Study
In preparation for promulgation of Method 1613 under Section 304(h), a single laboratory method
detection limit study was initiated by ITD in December 1989. The basic design of the study was dictated by
the procedure for determining MDLs specified in Appendix B of 40 CFR Part 136, as published in the
October 26,1984 Federal Register. Briefly, this procedure specifies the following:
o Spiking at least 7 aliquots of reagent water with the analytes of interest
o Spiking levels in the range of one to five times the laboratory's estimate of the detection limit of
each analyte
o Analysis of all replicates and calculation of a mean and standard deviation of the concentration of
each analyte
o Calculating the MDL as the standard deviation times the Student's r value for (n-1) degrees of
freedom, where n is the number of replicates
The specifications in 40 CFR 136 list four ways in which to determine an estimate of the detection
limit of the method. One way is to estimate the concentration in a sample that corresponds to an
instrumental signal-to-noise ratio of 2.5 to 5.0. Based on calculations from the single laboratory performing
the study, the lowest concentration of 2,3,7,8-TCDD in a sample which produces a signal that is 2.5 times the
background instrumental noise is 3.3 ppq. However, the concentration is determined by isotope dilution,
and the calculation assumes that the recovery of the labeled compound (the isotope) is 100%. Because the
draft of Method 1613 allows data acceptance when labeled compound recovery is as low as 25%, the
concentration of 3.3 ppq was adjusted upward by the inverse of the worst-case recovery, i.e. 3.3/0.25, which
equals 13.2 ppq. Following the specifications in 40 CFR 136, the spiking level was set at 25 ppq for TCDD,
within the range of 1 to 5 times the 13.2 ppq estimated detection limit. Based on similarities in instrumental
response to the other analytes, this spiking level was used for the tetra, penta, and hexa chlorinated dioxins
and furans, and 50 ppq was used for the hepta and octa compounds, which typically yield a lower
instrumental response.
The mean concentrations of each analyte from the replicate analyses are given in Table 3, along with
the calculated MDL for each. One of the QC criteria for the qualitative identification of PCDDs/PCDFs by
selected ion monitoring gas chromatography and mass spectrometry is the ratio of the abundance of the two
ions monitored for each analyte. Interferences (non-target analytes) that elute in the retention time window
of the dioxins and furans will cause the ion abundance ratio to vary from the theoretical abundances of the
two ions. When the ion abundance for a chromatographic peak falls outside established acceptance
windows (+ 15% of theoretical), the analyte cannot be positively identified. The risk of an interference
causing the ion abundance ratio to fail the QC criteria increases as the concentration of the analyte
decreases. As can be seen by the numbers of observations for each analyte listed in Table 3, most of the
analytes failed to meet the identification criteria at least once among the seven replicates. The results were
worst for several of the hepta chlorinated dioxins and furans, to which the instrument is less sensitive than to
the analytes with fewer chlorine atoms attached. Although less than ideal, the MDL values in Table 3 were
calculated using only those values that met the qualitative identification criteria.
10
058
-------
Except for OCDD, all the MDL values fall within a general range of 2-30 ppq. The large variability of
the OCDD measurements is believed to be due to the ubiquitous presence of this compound in the
environment. The OCDD results were not corrected for the levels of OCDD found in the blanks associated
with the samples, although the procedure in 40 CFR 136 may permit such corrections. Given the low
toxicity of OCDD and a correspondingly lower degree of concern with the presence of OCDD in
environmental samples, the high MDL for OCDD was not deemed to be a problem at this time.
Except for the OCDD, the MDL values in Table 3 are all below the "minimum levels" listed in
Method 1613. The minimum levels correspond to the concentration in a sample equivalent to the
concentration of each analyte in the lowest of the calibration standards, assuming 100% recovery of the
labeled compounds added to the sample and used for quantification by isotope dilution. Thus, the MDL
data presented here indicate that the method is capable of measuring PCDDs and PCDFs at levels at least as
low as the minimum levels specified in the method description.
Ultimately, the MDLs need to be determined in a variety of additional matrices, by additional
laboratories, and at additional spiking levels.
5. CONFIRMATORY TESTING
5.1 Laboratory Testing
Beginning with the September 12, 1988 draft of the method, and continuing with the revisions
through July 1989, a total of three commercial laboratories have utilized Method 1613 for contract analyses
for ITD. Those laboratories are:
o Triangle Laboratories, Research Triangle Park, NC
o Twin City Testing, St. Paul, MN
o California Analytical Laboratories, W. Sacramento, CA
Each of these laboratories submitted comments and suggestions to ITD based on their use of the method.
The results of the analysis of field samples by these laboratories are discussed in the section of this report on
Field Validation.
5.2 Peer Review
In addition to the comments received from the laboratories listed above, the draft method was
circulated to a variety of other Agency and industry reviewers. Comments were received from the following:
o Terry Nestrick and Les Lamparski, Dow Chemical, Midland, MI
o Dr. Tang and Dr. Alwan, Region V CRL, Chicago, IL
o Dr. Jill Henes, ChemWest Laboratories, W. Sacramento, CA
o Gary Robertson, Lockheed Environmental Services Co., Las Vegas, NV
o Dale Rushneck, Chemex International, Fort Collins, CO
o Joan Fisk, Analytical Methods Implementation Section, OERR, USEPA, Washington, DC
o Dr. Larry LaFleur, National Council of the Paper Industry for Air and Stream Improvement
(NCASI), Eugene, OR.
o John Stanley, Midwest Research Institute, Kansas City, MO
o Heather Cavalier, Cambridge Isotope Laboratories, Cambridge, MA
11
053
-------
In addition to a large number of editorial comments regarding typographical errors, erroneous
paragraph citations, etc., numerous technical comments were received. The comments from Dow Chemical
on the extraction procedures are discussed at length in the section on the single laboratory testing
performed.
Three of the commercial laboratories and a commercial provider of standards commented on the
concentrations of the calibration standards specified in the original draft of the method, as well as the
concentrations of the precision and recovery standard (PAR). As a result, the concentrations of some of the
standards were adjusted.
The solvent used for the standards drew several comments, resulting in a change from iso-octane to
nonane for the calibration standards, and the addition of a procedure for diluting the nonane-based spiking
solutions with acetone, to make them more easily incorporated into water samples.
Dow Chemical also recommended against baking the reusable glassware at high temperatures, as such
baking promotes the formation of active sites on the glass surface which may cause the loss of analytes
during sampling processing. As a result, kiln baking of glassware was removed from the method, and the use
of disposable glassware was maximized.
Whereas most other dioxin methods specify the use of benzene for extracting solid matrices, OERR
recommended a change to toluene for such extractions. Given the health concerns involved in the use and
disposal of benzene, this change was well received by the commercial laboratories. Given that toluene forms
an azeotrope with water that boils at a higher temperature than the benzene-water azeotrope, the use of
toluene in the SDS extraction process is even more effective than benzene.
In February of 1990, Yves Tondeur of Triangle Laboratories provided data that indicated that the
multi-layer alumina column specified in the July 1989 revision of 1613 may lead to selective losses of the
tetra-chlorinated isomers. Inconsistent recoveries of these isomers had been noted at other laboratories.
ITD speculated that the multi-layer column may not be sufficiently rugged for widespread use.
NCASI had numerous comments on the specifics of sample extraction and cleanup. Many of these
were based on their experience with pulp and paper matrices. In particular, they were concerned that
Method 1613 did not give sufficient guidance on dealing with the emulsions that may form during the
extraction of water samples, and that the multi-layered alumina column used for extract cleanup may not be
adequate for many sample extracts.
As the result of a meeting between ITD and NCASI, held in February 1990, Method 1613 will be
modified to incorporate additional guidance on treating emulsions. Also, after review of Triangle's data on
the multi-layer alumina column and consideration of NCASI's comments on the column, a single-layer
alumina column will be used instead of the current multi-layer cleanup column.
6. Interim Method Description
A copy of the July 1989 revision of Method 1613 is attached as Appendix H. This version of the
method contains the changes made in response to all the comments cited above except those made in
February 1990 from Triangle and NCASI.
The method is written in a format consistent with that of EMSL-Cincinnati and other ITD isotope
dilution methods, including specific section for:
o Scope and Application
o Summary of the method
o Contamination and Interferences
12
OGO
-------
o Safety
o Apparatus and materials
o Reagents and standards
o Calibration
o Quality Assurance and Quality Control
o Sample collection, preservation, and handling
o Sample preparation
o Extraction and concentration
o Extract cleanup
o Analysis
o System and laboratory performance
o Qualitative determination
o Analysis of complex samples
o Method performance
o References
The Quality Assurance and Quality Control specifications of Method 1613 include a variety of
operations consistent with the 600 and 1600 series methods. The laboratory must perform a series of start
up tests to demonstrate their ability to perform the method adequately in a reference matrix The start up
tests include spiking at least four aliquots of a reference matrix, typically reagent water or a clean solid
matrix, with all 17 native analytes, and carrying these spiked samples through the entire analytical
procedure. The recovery of the native and labeled analytes from the reference matrix are compared to the
specifications in the Method to determine if the laboratory is performing adequately.
All samples and blanks are spiked with 15 isotopically labeled PCDDs/PCDFs prior to extraction.
Each extract is spiked with an additional labeled compound prior to cleanup. The IS compounds spiked
prior to extraction are used for the quantification of the native analytes by isotope dilution, and result in
recovery-corrected data for the native analytes. The recoveries of these labeled compounds and the one
compound added prior to cleanup are compared to the Method specifications for recovery. In this fashion,
the performance of the method is evaluated for each sample analysis. The accuracy of the method, as
measured by the recovery of the labeled compounds, is re-evaluated by the laboratory after 5-10 sample
analyses.
The method requires the analysis of blanks that are extracted with each set of samples, up to a
maximum of 20 samples per blank. Data from the analysis of the blank is provided along with the sample
data, and the levels of PCDDs/PCDFs in the sample are not adjusted for the levels in blanks. Limits are
placed on the amount of any analyte detected in the blank. Given the ubiquitous nature and low toxicity of
OCDD, background levels of this analyte are not a significant cause for concern.
Guidance is given in the method on the use of field replicates and spiked samples in evaluating the
performance of the method and the laboratory.
The performance of the laboratory and the entire analytical system is evaluated on a daily basis. Each
day on which samples are extracted, the laboratory must prepare a reference matrix aliquot spiked with the
Precision and Recovery Standard. This standard contains the 17 native analytes as well as the 15 labeled
analogs, and is carried through the entire analytical procedure, in a fashion similar to the start up test
aliquots. The recovery of the native and labeled compounds in this spiked sample are judged against the
13 081
-------
requirements of the method, and may require re-extraction of all the associated samples if the spiked sample
does not meet the specifications.
The resolution of the mass spectrometer is evaluated each 12 hours, to determine that it meets the
method specification for static resolving power of at least 10,000. The resolution of the gas chromatograph
is evaluated through the analysis of an isomer specificity standard which contains the PCDDs/PCDFs which
are most difficult to separate on the two GC columns specified in the method. The analysis of the GC
window defining mixture serves to verify that the laboratory is searching for the analytes in the appropriate
retention time windows.
The calibration of the GCMS system is evaluated each 12 hours by the analysis of a calibration
verification standard. If the results of this analysis fail to meet the specifications of the method, the
laboratory must analyze a new set of five initial calibration standards that pass the criteria for initial
calibration.
7. FIELD VALIDATION STUDY
7.1 Scope
Method 1613 was developed to support 304(m) studies in the Pulp and Paper and Petroleum
Refining industries. However, prior to use in these industries 1613 was used in the National Sewage Sludge
Survey and Pesticides industry studies. Method 1613 was also used for analysis of samples from the
Superfund Discharges to POTWs Study conducted under 304(m). Three laboratories participated in the
field validation study. Table 4 provides a breakdown of the number of samples by Industrial Category and
Laboratory. When discussing performance data from a specific laboratory the laboratory will be referred to
as Lab 1, Lab 2,... etc, rather than by name.
TABLE 4
1613 Field Validation Study Samples by Industry and Lab
Industry Lab 1 Lab 2 Lab 3 Total
NSSS 110 123 0 233
Pesticides 090 9
Petroleum .0 7 12 19
Pulp & Paper 0 32 0 32
SF/POTW 0 17 0 17
Total 108 188 12 310
7.1.1 National Sewage Sludge Survey
In December of 1988, OW required a method for the analysis of dioxins and furans in municipal
sewage sludge in support of the National Sewage Sludge Survey (NSSS). This survey was being conducted to
provide data for the 40 CFR Part 503 regulation (CWA Section 405(d)). The decision was made to use
Method 1613 (September 12 1988 revision) for the analysis of samples from the NSSS. Special Analytical
Services (SAS) subcontract numbers 140 and 141 were awarded to Twin Cities Testing (TCT) and
ENSECO-California Analytical (CAL) respectively. As part of the SAS solicitation, each solicited
14
082
-------
laboratory submitted comments on the clarity and completeness of the method write-up. These comments
were taken into account during subsequent method revisions. TCT supplied extensive written comments
many of which were incorporated into future revisions. The comments from these laboratories are among
those cited earlier in this report
Approximately eighty samples which had been stored frozen since sampling were split into two
batches and sent to TCT and CAL. An additional 160 samples were taken and shipped to TCT and CAL.
7.1.2 Pesticide Formulators and Packagers Study
Nine samples from a study of four pesticide formulator and packager facilities were sent to TCT for
analysis, in support of CWA Section 402 effluent guideline development.
7.1.3 304(m) Industry Studies
Sixty-eight samples from three CWA Section 304(m) industries were analyzed using Method 1613
during FY89 studies.
As the result of reports from Canada of the discovery of PCDDs and PCDFs in the effluents from
petroleum refinery operations, a study of US refineries was undertaken by OW in August of 1989.
Approximately 35 samples of effluents and sludges were shipped to TCT and Triangle Laboratories for
analysis.
Thirty-two samples were taken in support of the ITD Pulp and Paper Study. These samples were
analyzed by TCT over the course of six months, starting in the spring of 1989.
Seventeen analyses for PCDDs and PCDFs were performed by TCT in support of studies of
discharges from Superfund sites to POTWs.
7.2 Method and Laboratory Performance
Prior to sample analysis using Method 1613, a laboratory must complete the start-up tests (Initial
Precision and Recovery,Method Section 8.2). The laboratory must also analyze an ongoing precision and
accuracy aliquot (Method Section 14.5) with each sample set. The results of these analyses must display
acceptable precision and recovery, as defined in the method.
Table 5 presents the results of the start-up tests (IPR) performed by each laboratory. Table 6 details
the results of the Ongoing Precision and Accuracy (OPR) analyses by each laboratory. The data presented
in these two tables are not differentiated by laboratory. Both tables divide the compounds into two groups.
The first group contains the 15 native 2,3,7,8-substituted PCDDs and PCDFs that are quantified by isotope
dilution. The second group contains the 16 labeled compounds and the two native compounds that are
quantified by internal standard.
As can be seen in Table 5, the initial precision of and recovery data indicate that all the laboratories
are able to achieve acceptable recoveries of the native and labeled compounds spiked into the reference
matrices. For 2,3,7,8-TCDD, the mean recovery across all labs was 88.5%. The mean recoveries of all other
native compounds were even higher. The fact that the mean recovery of OCDD was above 100% is due to
the great difficulty in eliminating background levels of this compound. The data reported in these tables
have not been adjusted for the levels of OCDD found in the blanks. The standard deviatons of the native
compound recoveries are also relatively low. No native compound had a standard deviation across all labs of
greater than 20%.
15 063
-------
The recoveries of the labeled compounds in Table 5 are generally lower than those of the native
compounds. The recoveries of 2,3,7,8-TCDD-C13 and 2^,7,8-TCDF-C13 are on average around 50%. The
Cl3 labeled TCDD standard that is added after extraction but prior to cleanup has a similar average
recovery. This result confirms the belief of the laboratories that the multi-layered alumina column used for
extract cleanup may have caused the preferential loss of the tetrachlorinated compounds. As discussed
earlier, this cleanup column is being replaced in the latest revision of the method.
The results in Table 6 for the ongoing precision and recovery data across all laboratories indicate
similiar trends. The native compound recoveries were all above 90%, and the labeled compound recoveries
were all above 50%, though lowest for the tetrachlorinated compounds.
In general, the EPR and OPR recoveries from the participating laboratories appear to be
homogeneous.
By comparing the recoveries of the native compounds with their labeled counterparts, one can see the
value of isotope dilution for quantification. Despite losses of some labeled compounds as high as 50%
during the extract cleanup, the precision of the results for the native compounds, measured as the standard
deviation of the recoveries, is considerably better than for the two native compounds quantified by internal
standard techniques.
7.2.5 Field Validation Study Results
The data for the precision and recovery of the labeled compounds spiked into the field samples are
presented by Industry and Matrix in Table 7. As these were actual field samples from a wide variety of
sources, no "true" values existed for the concentrations of the native PCDDs and PCDFs. Therefore, Table 7
contains no data for native compound recoveries.
As can be seen in Table 7, there were significant differences in the labeled analog recoveries in the
samples from the various industries. However, the recoveries generally fell within the method specifications
despite these differences. The recoveries are generally higher in the solid matrices.
The results from the analyses of these field samples were used to develop quality control
specifications for the revised Method, as discussed below.
7.3 Development of Quality Control Specifications
A major objective of this method evaluation was to produce performance specifications from the data
collected to date and apply these specifications to the next method revision in. order to better gauge
laboratory performance.
Quality control specifications were calculated by constructing statistical prediction intervals for
future observations of a quantity of interest using estimates determined in this study. Recovery limits are
calculated using the results of a variance component analysis of the percent recoveries on three subsets of
data; IPR samples, OPR samples and EPA (field) samples. The inter- and intralaboratory variance
components of the logarithms of the recoveries were estimated (Sg2 and SA2), along with the log mean
response (m) by the Type 1 method, using PROC VARCOMP (SAS). Details of the variance components
analysis are given in Appendix B.
The percent recoveries of these compounds have been assumed to follow a log normal distribution
throughout the analysis described in this section. The log normal distribution has been frequently and
effectively applied to model pollutant concentrations, and agrees with the physical interpretation of non-
negative concentration values. Limits derived from this assumption are always non-negative. Descriptive
16 064
-------
and summary statistics calculated for these data support the assumption of log normality for most
compounds. Although some compounds appear to be normally distributed, log normal distributions were
assumed in order to attain non-negative limits for all specifications.
Compound-specific performance specifications have been determined at a 95% confidence level.
Using these specifications, each tested compound would have a 5% chance of falling outside QC limits.
The specifications resulting from the three-laboratory evaluation are detailed below. Statistical
methods are discussed in Appendices A through F.
7.3.1 Data Screening
The data utilized in this study were screened for outliers both on a laboratory and individual point
basis. The labeled analog recoveries in field samples analyzed at Lab 1 were on the order of 40-50% lower,
in similar samples, than those for Labs 2 and 3. This tended to increase the spread of the data and create the
appearance of a separate population of recovery data. Figure 3 illustrates this effect. The inclusion of data
from Lab 1 resulted in percent recovery specifications that were extremely wide and not representative of
current laboratory performance. The lower recoveries in these samples from Lab 1 were due to problems
associated with the earliest version of Method 1613, particularly regarding the extraction procedures.
Therefore, data from Lab 1 were not utilized for the development of recovery specifications for
labeled analogs in field samples. The remaining recovery values were then screened individually for outliers.
A robust quantile method based on the median and the interquartile range was applied, as described in
Appendix A.
7.3.2 Instrument Performance Specifications
7.3.2.1 Calibration Linearity
In order to calculate the concentration of each compound in a sample, the measured instrument
response is compared to the response obtained from a series of calibration samples analyzed at known
concentrations (Method 1613, Section 7). This allows an evaluation of instrument response across a range
of concentrations of the analytes. In this study, the coefficient of variation (CV) of the instrument response
achieved by the participating laboratories was examined. A linearity specification was calculated for each
compound and the existing data were tested against this specification. Table 8 details this analysis and
Appendix E describes the statistical approach. In general the specifications obtained compare favorably
with the existing CV specification of 20% for compounds quantitated by isotope dilution (Method 1613,
Section 7.5.5) and 35% for compounds quantitated by internal standard (Method 1613, Section 7.6.1.2).
7.3.3 Calibration Verification
Method 1613 Section 14.3 addresses calibration verification. Concentration limits for calibration
verification (VER) are required to insure that acceptable instrument response (measured as concentration)
is maintained over time. Specifications for upper and lower concentration limits were developed using
laboratory-submitted VER analyses. These specifications are detailed in Table 9. Submitted data were then
tested against the developed specifications to determine the percentage of compounds that would fall
outside the limits. Table 7 of Method 1613 will be updated to include these specifications. The statistical
approach is detailed in Appendix C.
17
085
-------
7.3.4 Relative Retention Time
The relative retention time (RRT) of the native compound versus the labeled analog is a critical
specification for isotope dilution quantitation. Current relative retention time (RRT) specifications
contained in Table 2 of Method 1613 were developed prior to applying isotope dilution to the analysis of the
2,3,7,8-substituted dioxin and furan isomers, and these specifications were extrapolated from the RRT of
chlorinated compounds in Method 1625. The extrapolated RRT specifications for native compounds of
approximately 0.999 -1.001 have proven to be too narrow. Method 1613 looks at a much narrower range of
analytes, and uses a slower temperature program for GC. As a result, the analytes are spread out across the
chromatogram further, and the differences between the retention times of the native and labeled compounds
increase. Therefore, the RRT limits will be adjusted according to the specfications in Table 10 of this
report, and will vary by compound. Appendix F details the statistical analyses used in developing these
specifications.
7.3.5 Initial Precision and Accuracy
Method 1613 requires the analysis of four initial precision and accuracy aliquots (EPR, Method
Section 8.2) prior to the analysis of any samples. The arithmetic average (mean) and standard deviation (sd)
of the four results generated for each labeled and native compound are computed per the Method. The
mean and sd must meet the limits specified in the Method in order to insure acceptable precision and bias
for subsequent analyses. As part of this study, limits were constructed for accuracy and precision. Table
11A details the EPR accuracy limits, expressed as mean percent recovery, and Appendix C provides the
derivation of these limits. Table 11B provides the the IPR precision limits expressed as sd of the recoveries,
and Appendix D provides the derivation of these limits. The existing limits in the Method will be updated to
relfect the limits generated in this study.
7.3.6 Ongoing Precision and Accuracy
In each batch of samples analyzed by the laboratory, one ongoing precision and accuracy aliquot is
analyzed (OPR, Method 1613 Section 14.5). Mean recovery is evaluated against specified limits to ensure
analytical control. Mean percent recovery specifications developed during this study are presented in Table
12. The derivation of these limits is given in Appendix C The existing limits in the Method will be updated
to relfect the limits generated in this study.
7.3.7 Labeled Compound Recovery in EPA Samples
The data for the recoveries of labeled compounds from the various field validation samples are
presented in Table 13. This table contains the data from all three laboratories, including those data from
Lab 1 discussed above which exhibited significantly lower recoveries. Table 14 contains the recovery data for
only Lab 2 and Lab 3. As can be seen by comparing these two tables, the means, medians, and standard
deviations of the recoveries increased significantly when the data from Lab 1 were screened out of the
population. The differences in the extraction and analyses procedures that these data represent are
indicative of the improvements made to Method 1613 during the field vlaidation studies.
Table 15 contains the limits generated for recovery of labled compounds from field samples. These
limits were generated from the field samples results for Lab 2 and Lab 3 only, as discussed above. Generally
speaking, the ranges of recoveries for each compound derived from these data are within the current
specifications (25-150%) of Method 1613, with some predicted recoveries above the current limits.
These limits for labled compound recovery in field samples will be incorporated into the Method, and
re-evaluated as more and more data are collected by the Agency using the latest version of the procedures.
18
06G
-------
8. Conclusions " '
Method 1613 was developed by OW for the analysis of PCDDs and PCDFs in effluents in other
matrices such as solids and sludges. The data presented here demonstrate that the method is sufficiently
rugged to be used by a variety of laboratories. The precision and accuracy of the method meet the needs of
the Agency for a survey method for dioxins and furans. The method contains an extensive QA/QC program
that allows the data user to evaluate the quality of individual sample results. This QA/QC program was built
into the method from the start, not simply added on to an existing method. The benefits of this approach
were evident during the development of the final method, where data on labeled compound recoveries were
instrumental in evaluating various sample extraction and clean up procedures.
For 2,3,7,8-TCDD and 2,3,7,8,-TCDF, Method 1613 is approximately 200 times more sensitive than
Method 613, the existing method promulgated under Section 304(h). Validation data for these two
compounds from the analyses of over 300 field samples indicate acceptable levels of precision and bias for
these compounds. Therefore, Method 1613 is proposed as a nationwide Alternative Test Procedure under
Section 304(h) for the analysis of 2,3,7,8-TCDD and 23,7,8-TCDF.
Given that no method for the analysis of the penta- through octachlorinated dioxins and furans has
been proposed or promulgated under Section 304(h), Method 1613 is proposed as the nationwide primary
test procedure for these analytes.
Finally, the results of the evaluation presented here and the public comments received during the
course of this evaluation have been used to revise the July 1989 version of Method 1613 for nationwide use
in support of effluent guideline development and NPDES monitoring.
References
1. "Ambient Water Quality Criteria for 2,3,7,8 - Tetrachloro-dibenzo-p-dioxinn, USEPA, OWRS,
Washington DC 20460, EPA 440/5-84-007, February 1984.
2. Tondeur, Yves, "Method 8290: Analytical Procedures and Quality Assurance for Multimedia Analysis
of Polychlorinated Dibenzo-p-dioxins and Dibenzorurans by High-Resolution Gas
Chromatography/High-Resolution Mass Spectrometry," USEPA, EMSL-Las Vegas, Nevada, June
1987.
3. "Measurement of 23,7,8-Tetrachlorinated Dibenzo-p-dioxin (TCDD) and 23,7,8-Tetrachlorinated
Dibenzofuran (TCDF) in Pulp, Sludges, Process Samples and Wastewaters from Pulp and Paper
Mills", Wright State University, Dayton OH 45435, June 1988.
4. "NCASI Procedures for the Preparation and Isomer Specific Analysis of Pulp and Paper Industry
Samples for 2,3,7,8-TCDD and 23,7,8- TCDF", National Council of the Paper Industry for Air and
Stream Improvement, 260 Madison Av, New York NY 10016, Technical Bulletin No. 551, Pre-
release Copy, July 1988.
5. "Analytical Procedures and Quality Assurance Plan for the Determination of PCDD/PCDF in Fish",
U.S. Environmental Protection Agency, Environmental Research Laboratory, 6201 Congdon Blvd.,
Duluth MN 55804, April 1988.
6. Yves Tondeur, "Proposed GC/MS Methodology for the Analysis of PCDDs and PCDFs in Special
Analytical Services Samples", Triangle Laboratories, Inc., 801-10 Capitola Dr, Research Triangle
Park NC 27713, January 1988; updated by personal communication September 1988.
7. Lamparski, L.L., and Nestrick, TJ., "Determination of Tetra-, Hexa-, Hepta-, and
Octachlorodibenzo-p-dioxin Isomers in Paniculate Samples at Parts per Trillion Levels", "Anal.
Chem." 52,2045-2054 (1980).
19
08 V
-------
8. Lamparski, L.L., and Nestrick, TJ., "Novel Extraction Device for the Determination of Chlorinated
Dibenzo-p-dioxins (PCDDs) and Dibenzofurans (PCDFs) in Matrices Containing Water", Personal
Communication, July 1988.
9. Patterson, D.G., et. al. "Control of Interferences in the Analysis of Human Adipose Tissue for 2,3,7,8-
Tetrachlorodibenzo-p-dioxin", "Environ. Toxicol. Chem.," 5,355-360 (1986).
10. Stanley, John S., and Sack, Thomas M., "Protocol for the Analysis of 2,3,7,8-Tetrachlorodibenzo-p-
dioxin by High-Resolution Gas Chromatography/High-Resolution Mass Spectrometry", USEPA,
Environmental Monitoring Systems Laboratory, Las Vegas NV 89114, EPA 600/4-86-004, January
1986.
11. "Method 613 - 23,7,8-Tetrachlorodibenzo-p-dioxin", 40 CFR136 (49 FR 43234), October 26,1984.
12. "Guidelines and Format for EMSL-Cincinnati Methods", USEPA, EMSL, Cincinnati, OH 45268,
EPA-600/8-83-020, August 1983.
13. "Guidelines for Selection and Validation of USEPA's Measurement Methods", USEPA, ORD,
OADEMQA, Draft, August 1987.
14. Eynon, B.P., Maxwell, C and Valder, A., "Interlaboratory Validation of U.S. Environmental
Protection Agency Method 1625A", EPA Contract 68-01-6192, July 1984.
15. Hoaglin, D.C, Mosteller, F., and Tukey, J.S., "Understanding Robust and Exploratory Data Analysis",
John Wiley and Sons, New York, pp 37-39,1983.
16. Milliken, G.A., Johnson, D.E., "Analysis of Messy Data", Lifetime Learning Publications, California,
pp 238-241,1984.
17. Wilson, A.L., "Approach for Achiecing Comparable Results from a Number of Laboratories", The
Analyst, Vol. 104 No. 1237, p 273,1979.
18. "USEPA Development Document for Existing Source Pretreatment Standards for the Electroplating
Point Source Category", EPA Document 440/1-79/003, August 1979.
20
088
-------
U S E P A
INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE 3
METHOD DETECTION
LIMIT
COMPOUND
2378-TCDD
12378-PECDD
123478-HXCOO
123678-HXCDD
1234678-HPCDD
OCDD
2378-TCOF
12378-PECDF
23478-PECDF
123478-HXCDF
123678-HXCDF
123789-HXCDF
234678-HXCDF
1234678-HPCDF
1234789-HPCDF
COMPOUND
^ 123789-HXCDD
X-C OCDF
NUMBER
DBS
6
6
7
6
5
7
6
6
6
6
7
6
6
7
3
NUMBER
OBS
5
6
QUANTITATION=ISOTOPE DILUTION
MEAN
26.3
27.6
31.3
26.1
52.3
195.6
27.2
27.7
30.6
38.8
28.0
25.4
25.9
67.9
70.2
STD
DEVIATION
0.8
2.0
1.7
2.6
4.9
97.4
1.4
1.7
1.7
2.4
0.5
1.3
1.9
6.0
T-VALUE
3.365
3.365
3.143
3.365
3.747
3.143
3.365
3.365
3.365
3.365
3.143
3.365
3.365
3.143
6.965
QUANTITATION=INTERNAL STANDARD
MEAN STD T-VALUE
22.4
75.8
STD
DEVIATION
2.1
8.2
3.747
3.365
METHOD
DET LIMIT
2.7
6.7
5.3
8.7
18.2
306.2
4.7
5.8
5.7
8.1
1.7
4.3
6.4
19.0
20.3
METHOD
DET LIMIT
8.0
27.6
-------
USEPA INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
COMPOUND
2378-TCDD
12378-PECDD
123478-HXCDD
123678-HXCDO
1234678-HPCDD
OCDD
2378-TCDF
12378-PECOF
23478-PECDF
123470-HXCDF
123678-HXCOF
123789-HXCDF
234678-HXCDF
1234678-HPCDF
1234789-HPCDF
COMPOUND
o
^- 2378-TCDD-C13
__• 2378-TCDD-CL37
v— 12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
123789-HXCDD
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
OCDF
NO OF
OBSERVATINS
21
19
21
21
22
18
22
20
21
22
21
18
22
19
20
NO OF
OBSERVATINS
22
22
22
21
20
22
22
22
20
21
21
21
20
22
17
22
22
22
TABLE 5
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
SUMMMARY STATISTICS
ALL LABS COMBINED
QUANTITATION=ISOTOPE DILUTION
MEAN
88.5
96.6
92.2
93.9
92.2
102.
91.
96.
95.5
97.7
92.3
MEDIAN
A
.3
.5
97.7
W.2
96.8
85.0
97.0
93
90
93
102
94.2
98.0
95.0
96.2
92.0
98.0
98.0
100.0
98.5
qUANTITATION=INTEHNAL STANDARD
MEAN MEDIAN
56.4
57.0
76.6
72.5
72.6
91.7
82.5
80.6
50.3
70.1
72.9
63.4
67.8
72.7
69.4
67.6
73.6
97.8
67.0
56.5
71.1
66.2
68.0
90.0
83.8
83.5
58.5
72.0
72.0
63.1
63.3
69.9
67.1
62.3
75.5
103.6
STD
DEVIATION
14.4
5.0
9.2
10.7
12.9
10.3
19.6
5.6
5.4
13.6
5.9
4.5
11.9
4.8
7.5
STD
DEVIATION
22.0
25.8
30.3
15.4
17.5
17.6
13.7
25.3
19.7
26.3
25.2
13.4
15.3
16.1
15.8
14.0
16.9
24.2
MINIMUM
66.0
87.0
76.0
78.0
72.0
82.0
55.0
85.0
86.0
73.0
82.0
89.0
79.0
89.0
61.0
MINIMUM
19.0
21.0
26.0
54.0
37.4
64.0
60.0
34.9
16.0
36.0
38.0
46.0
47.0
49.0
49.0
48.0
46.6
56.0
MAXIMUM
118.0
105.0
105.0
114.0
120.0
124.0
125.0
104.0
105.0
123.0
103.0
108.0
128.0
108.0
108.0
MAXIMUM
86.2
. 95.1
140.0
101.0
97.7
133.0
113.0
120.0
77.3
129.0
133.0
98.0
99.0
106.0
109.0
95.8
101.0
150.0
-------
U S E P A
INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE 6
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
SUMMMARY STATISTICS
ALL LABS COMBINED
COMPOUND
2378-TCDD
12378-PECDD
123478-HXCDD
123678-HXCDD
1234678-HPCDD
OCDD
2378-TCDF
12378-PECDF
23478-PECDF
123478-HXCDF
123678-HXCDF
123789-HXCDF
234678-HXCOF
1234678-HPCDF
1234789-HPCDF
------- muuti
NO OF
OBSERVATINS
48
47
45
48
42
47
47
44
46
47
47
45
47
48
47
A i M i Aun-
MEAN
90.6
91.5
92.1
92-. 8
91.7
95.1
92.5
92.1
92.7
93.7
91.0
90.9
91.0
95.7
95.1
Aauiurc uj
MEDIAN
89.5
93.0
90.5
90.5
94.0
96.0
92.5
92.0
92.5
95.0
90.0
93.0
93.0
96.0
96.0
STD
DEVIATION
13.8
9.2
11.2
13.1
7.2
8.7
12.0
7.6
8.6
9.4
9.3
7.8
9.4
9.8
10.3
MINIMUM
58.0
76.0
67.0
70.0
80.0
80.0
66.0
77.0
74.0
77.0
74.0
75.0
70.0
75.0
80.0
MAXIMUM
125.0
115.0
115.0
125.0
105.0
116.0
125.0
108.0
111.0
112.0
110.0
105.0
110.0
118.0
120.0
COMPOUND
o
is. 2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
123789-HXCDD
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
OCDF
NO OF MEAN MEDIAN STD
OBSERVATINS
50
49
50
46
48
48
49
49
50
49
47
50
49
49
35
50
48
44
55.2
57.4
69.9
73.1
75.2
88.6
69.4
55.8
55.6
65.1
64.4
68.4
70.6
69.5
69.8
63.4
66.2
93.2
58.0
60.0
68.0
71.0
71.0
88.0
66.0
54.0
57.5
62.0
59.0
70.5
70.0
70.0
70.0
61.5
68.0
92.0
DEVIATION
22.8
24.2
28.0
13.4
15.1
12.5
16.8
18.4
21.0
23.5
23.9
15.1
14.8
14.5
14.0
13.5
17.4
9.8
MINIMUM
11.0
9.6
10.0
49.0
45.0
69.0
36.0
11.0
11.0
35.0
5.7
37.0
37.0
37.0
42.0
38.0
30.9
77.5
MAXIMUM
96.0
113.0
135.0
101.0
109.0
115.0
108.0
97.0
91.0
115.0
115.0
103.0
101.0
100.0
96.0
93.0
97.0
119.5
-------
USEPA INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE 7
PRECISION AND RECOVERY OF LABELED
SUMMARY STATISTICS
BY INO CATC MATRIX)
COMPOUNDS
IND_CAT=PULP + PAPER MATRIX=AQUEOUS QUANTITATION=INTERNAL STANDARD
COMPOUND
2378-TCDD-C13
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
NO OF
OBS
MEAN
MEDIAN
1 31.0
1 36.0
1 41.0
1 32.0
1 24.0
1 20.0
1 26.0
31.0
34.0
34.0
31.0
33.0
20.0
1 22.0
31.0
36.0
41.0
32.0
24.0
20.0
26.0
31.0
34.0
34.0
31.0
33.0
20.0
22.0
STD
DEVIATION
MINIMUM MAXIMUM
31.0
36.0
41.0
32.0
24.0
20.0
26.0
31.0
34.0
34.0
31.0
33.0
20.0
22.0
31.0
36.0
41.0
32.0
24.0
20.0
26.0
31.0
34.0
34.0
31.0
33.0
20.0
22.0
IND_CAT=PULP + PAPER MATRIX=SOLID QUANTITATION=INTERNAL STANDARD
o
COMPOUND
2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDO-C13
123678-HXCDD-C13
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCOF-C13
NO OF
OBS
MEAN
MEDIAN
STD
DEVIATION
MINIMUM
MAXIMUM
31
31
31
31
31
31
31
31
31
31
31
31
31
31
31
31
56.2
57.8
85.5
80.5
72.5
58.1
48.5
51.8
73.6
85.3
72.9
68.7
70.7
66.8
55.2
60.6
59.0
60.0
84.0
79.0
74.0
57.0
46.0
52.0
71.0
87.0
70.0
68.0
70.0
67.0
55.0
61.0
14.7
15.1
13.2
10.3
11.6
8.7
7.6
14.4
10.2
12.4
9.7
10.6
9.8
8.7
8.8
8.7
27.0
29.0
63.0
66.0
51.0
41.0
33.0
28.0
53.0
63.0
58.0
52.0
55.0
45.0
40.0
39.0
79.0
83.0
113.0
113,0
112.0
90.0
72.0
79.0
92.0
110.0
104.0
102.0
103.0
87.0
85.0
89.0
-------
U S E P A
INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE 7
PRECISION AND RECOVERY OF LABELED
SUMMARY STATISTICS
BY IND CATC MATRIX)
COMPOUNDS
IND_CAT=PESTICIDES MATRIX=AQUEOUS qUANTITATION=INTERNAL STANDARD
COMPOUND NO OF MEAN MEDIAN STD MINIMUM MAXIMUM
2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
1234678-HPCDD-C13
OCDO-C13
2378-TCDF-C13
J2378-PECDF-CJ3
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
NO OF
OBS
STD
DEVIATION
8
8
8
8
8
8
8
8
8
8
8
8
7
7
8
8
41.0
58 .'3
60.4
54.8
54.9
51.5
48.0
38.8
49.9
54.0
48.0
53.3
61.0
58.4
48.4
53.1
43.0
59.0
64.5
57.5
54.0
52.0
48.0
41.0
53.5
58.i
52.0
56.0
59.0
55 0
50.5
54.0
18.9
17.1
24.1
20.1
21.1
23.1
20.6
18.5
20.1
20.9
18.0
20.4
8.3
7.9
22.1
22.8
11.0
32.0
8.0
9.0
15.0
17.0
13.0
10.0
8.0
7.0
8.0
9.0
55.0
53.0
8.0
13.0
71.0
82.0
69.0
75.0
87.0
90.0
74.0
71.0
75.0
79.0
69.0
62.0
79.0
75.0
87.0
90.0
IND_CAT=PESTICIDES MATRIX=SOLID C|UANTITATION=INTERNAL STANDARD
COMPOUND
2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
NO OF
OBS
MEAN
MEDJAN
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
76.0
78.0
64.0
85.0
59.0
71.0
68.0
69.0
60.0
54.0
63.0
68.0
66.0
67.0
58.0
67.0
76.0
78.0
64.0
85.0
59.0
71.0
68.0
69.0
60.0
54.0
63.0
68.0
66.0
67.0
56.0
67.0
STD
DEVIATION
MINIMUM
76.0
78.0
64.0
85.0
59.0
71.0
68.0
69.0
60.0
54.0
63.0
68.0
66.0
67.0
58.0
67.0
MAXIMUM
76.0
78.0
64.0
85.0
59.0
71.0
68.0
69.0
60.0
54.0
63.0
68.0
66.0
67.0
58.0
67.0
o
-------
USEPA INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE 7
PRECISION AND RECOVERY OF LABELED
SUMMARY STATISTICS
BY IND CAT(MATRIX)
COMPOUNDS
IND_CAT=PETROLEUM REF. MATRIX=AQUEOUS QUANTITATION=INTERNAL STANDARD
COMPOUND NO OF MEAN MEDIAN STD MINIMUM MAXIMUM
2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
1234678-HPCDO-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
NO OF
OBS
STD
DEVIATION
17
17
17
17
17
17
17
17
17
17
17
17
17
15
17
17
62.1
56.8
62.7
76.4
70.5
59.6
46.6
69.0
64.1
64.0
63.3
64.3
60.7
69.0
52.6
52.8
62.0
59.1
63.1
79.9
76.9
64.1
51.6
62.2
59.2
60.0
68.0
68.3
58.4
69.4
58.3
53.0
18.8
19.5
11.9
27.6
14.2
16.6
16.4
22.2
12.4
14.3
13.5
14.8
15.8
10.4
14.4
15.0
23.6
21.0
34.3
41.7
42.8
17.5
10.0
34.7
39.5
37.3
38.7
39.7
34.9
52.0
11.4
11.9
103.0
88.9
82.5
151.0
92.4
84.7
68.5
112.0
95.1
101.0
81.6
86.7
90.2
87.5
67.0
74.0
o
IND_CAT=PETROLEUM REF. MATRIX=OTHER QUANTITATION=INTERNAL STANDARD
COMPOUND NO OF MEAN MEDIAN STD MINIMUM MAXIMUM
2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123678-HXCDD-C13
123678-HXCDD-C13
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECOF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
1237tt9-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
NO OF
OBS
STD
DEVIATION
16
16
16
16
16
16
16
16
16
16
15
16
16
14
16
16
51.2
45.3
53.3
68.5
66.7
58.1
38.3
60.0
56.6
56.0
70.9
66.6
60.2
64.2
55.5
55.2
48.6
42.2
43.4
68.2
64.3
55.6
33.9
56.4
57.0
55.6
66.6
63.3
50.3
63.5
54.3
51.4
24.6
22.8
26.7
26.6
26.7
19.0
15.2
19.4
21.6
23.9
17.9
25.9
21.1
15.5
22.5
22.1
7.6
5.8
7.1
4.4
7.3
27.2
14.6
31.3
18.5
3.5
46.4
3.9
22.7
42.4
7.0
24.1
95.1
83.8
98.1
103.0
103.0
88.0
66.0
104.0
102.0
106.0
96.0
106.0
91.0
65.4
84.0
86.0
-------
U S E P A
INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE 7
PRECISION ANO RECOVERY OF LABELED
SUMMARY STATISTICS
BY IND CAT(MATRIX)
COMPOUNDS
IND_CAT=PETROLEUM REF. MATRIX=SOLID QUANTITATION=INTERNAL STANDARD
COMPOUND
2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23476-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
NO OF
OBS
2
2
1
2
1
1
2
2
2
2
2
2
2
2
2
2
MEAN
51.3
46.8
68.4
76.8
65.0
41.0
45.3
55.6
64.4
59.5
65.7
57.6
56.1
60.1
46.8
38.9
MEDIAN
51.3
46.8
68.4
76.8
65.0
41.0
45.3
55.6
64.4
59.5
65.7
57.6
56.1
60.1
46.8
38.9
STD
DEVIATION
11.0
7.4
,
14.4
.
.
30.7
12.2
10.7
2.1
13.2
0.6
11.1
9.8
14.4
10.0
MINIMUM
43.5
41.6
68.4
66.7
65.0
41.0
23.6
47.0
56.9
58.0
56.4
57.2
48.3
53.2
36.6
31.8
MAXIMI
59.0
52.0
68.4
87.0
65.0
41.0
67.0
64.3
72.0
61.0
75.0
58.0
64.0
67.0
57.0
46.0
o
-------
USEPA INDUSTRIAL TECHNOLOGY DIVISION
METHOD 1613 PERFORMANCE EVALUATION
TABLE 7
PRECISION AND RECOVERY OF LABELED COMPOUNDS
SUMMARY STATISTICS
BY IND_CAT(MATRIX)
IND_CAT=SUPERFUND/POTM MATRIX=AQUEOUS QUANTITATIOH=INTERNAL STANDARD
COMPOUND NO OF MEAN MEDIAN STD MINIMUM MAXIMUM
DBS DEVIATION
2378-TCDD-C13 14 23.7 18.5 17.3 3.0 54.0
2378-TCDD-CL37 14 36.8 36.0 16.5 7.0 64.0
12378-PECDD-C13 14 43.6 37.5 21.5 12.0 75.0
123478-HXCDD-C13 14 52.8 59.5 23.4 20.0 82.0
123678-HXCDD-C13 14 51.6 61.5 23.0 20.0 78.0
1234678-HPCDD-C13 14 52.1 68.0 27.6 14.0 80.0
OCDO-C13 14 46.9 59.0 26.0 11.0 73.0
2378-TCDF-C13 14 24.7 19 0 16.8 4.0 51.0
12378-PECDF-C13 14 43.8 41.5 22.8 14.0 79.0
23476-PECDF-C13 14 52.6 51.5 23.4 19.0 88.0
123478-HXCDF-C13 14 45.8 52.0 20.9 15.0 70.0
123678-HXCDF-C13 14 48.7 58.5 23.0 18.0 77.0
123789-HXCDF-C13 14 50.8 59.5 21.9 20.0 74.0
234678-HXCDF-C13 9 69.8 70.0 11.4 56.0 93.0
1234678-HPCDF-C13 14 46.4 51.5 25.6 12.0 77.0
1234789-HPCDF-C13 14 55.4 69.0 30.8 14.0 93.0
IND_CAT=SUPERFUND/POTW MATRIX=SOLID QUANTITATION=INTERNAL STANDARD
COMPOUND NO OF MEAN MEDIAN STD MINIMUM MAXIMUM
£-5 OBS DEVIATION
"^ 2378-TCDD-C13 3 77.3 79.0 3.8 73.0 80.0
C" 2378-TCDD-CL37 3 79.3 79.0 1.5 78.0 81.0
12378-PECDD-C13 3 93.3 90.0 6.7 £9.0 101.0
123478-HXCDD-C13 3 69.0 67.0 15.1 55.0 85.0
123678-HXCDD-C13 3 70.3 66.0 7.5 66.0 79.0
1234676-HPCDD-C13 3 74.0 69.0 9.5 68.0 85.0
OCDD-C13 3 62.7 77.0 10.7 76.0 95.0
2378-TCDF-C13 3 75.3 76.0 2.1 73.0 77.0
12376-PECDF-C13 3 85.0 79.0 11.3 78.0 98.0
23478-PECDF-C13 3 86.0 84.0 3.5 84.0 90.0
123478-HXCDF-C13 3 61.3 58.0 5.8 58.0 68.0
123678-HXCDF-C13 3 64.7 61.0 9.1 58.0 75.0
123789-HXCDF-C13 3 60.3 55.0 9.2 55.0 71.0
234678-HXCDF-C13 3 64.7 62.0 6.4 60.0 72.0
1234678-HPCDF-C13 3 60.7 57.0 8.1 55.0 70.0
1234789-HPCDF-C13 3 76.0 70.0 H.3 69.0 89.0
-------
U S E P A
INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE 7
PRECISION AND RECOVERY OF LABELED
SUMMARY STATISTICS
BY IND CATCMATRIX)
COMPOUNDS
IND_CAT=NATL SEWAGE SLUD MATRIX=SOLID QUANTITATIOH=INTERNAL STANDARD
COMPOUND
2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-CI3
1234789-HPCDF-C13
NO OF
OBS
200
199
201
201
201
201
197
201
201
196
194
201
200
122
199
201
MEAN
64.0
64.9
58.7
70.3
70.5
71.9
51.9
57.2
57.7
46.4
65.2
66.7
67.3
79.3
55.8
64.1
MEDIAN
76.5
78.0
64.0
83.0
84.0
77.0
55.0
67.0
62.0
47.0
75.0
78.0
75.0
78.5
62.0
74.0
STD
DEVIATION
32.6
35.5
31.5
36.7
34.5
33.4
29.0
25.6
25.6
26.9
26.3
32.8
25.9
11.2
24.1
30.6
MINIMUM
3.0
2.6
3.0
2.8
5.7
9.6
2.2
3.5
6.8
2.0
8.0
3.2
10.0
49.0
4.7
6.2
MAXIMU1
119.0
139.0
134.0
137.0
122.0
132.0
118.0
111.0
120.0
121.0
116.0
117.0
118.0
106.0
111.0
119.0
o
-vl
c;
-------
USEPA INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE 8
CALIBRATION LINEARITY LIMITS ~ INTERNAL STANDARD AND ISOTOPE DILUTION
__ _____ ____ ___ ____________ MUATIIi.1
COMPOUND
2378-TCDD
12378-PECDD
123478-HXCDD
123678-HXCDD
1234678-HPCDD
OCDD
2378-TCDF
12378-PECDF
23478-PECDF
123478-HXCDF
123678-HXCDF
123789-HXCDF
234678-HXCDF
1234678-HPCDF
1234789-HPCDF
AI_UN-_3UIU
DEGREES
OF
FREEDOM
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
ire UILUI J.UN •
CV a 95X
LIMIT
11.43
12.15
15.07
13.59
13.44
10.05
12.46
9.11
8.40
11.63
12.90
12.05
11.04
24.19
25.24
7. OUT OF
PRED_LMT
0.00
0.00
0.00
0.00
0.00
50.00
0.00
0.00
0.00
0.00
0.00
50.00
0.00
0.00
0.00
COMPOUND DEGREES CV 3 95X
-J 2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
123789-HXCDD
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
OCDF
OF
FREEDOM
14
14
14
14
14
14
14
14
14
14
14
14
14
13
14
14
14
14
LIMIT
6.21
10.29
35.89
13.88
22.31
12.21
12.76
18.47
15.85
18.76
17.53
14.74
14.52
14.47
13.94
26.98
13.97
16.51
7. OUT OF
PRED_LMT
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
50.00
0.00
0.00
0.00
0.00
0.00
-------
USEPA INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
TABLE 9
ONGOING CALIBRATION VERIFICATION LIMITS
ALL LABS
DIVISION
COMPOUND
2378-TCDD
12378-PECDD
123478-HXCDD
123678-HXCDD
1234678-HPCDD
OCDD
2378-TCDF
12378-PECDF
23478-PECDF
123478-HXCDF
123678-HXCDF
123789-HXCDF
234678-HXCDF
1234678-HPCDF
1234789-HPCDF
COMPOUND
O
CO 2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
123789-HXCDD
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCOF-C13
123789-HXCUF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
OCDF
NO OF
OBS
11
11
11
11
11
11
11
11
11
11
11
11
11
11
i A i xun-Aa
SPIKE
LEVEL
10
50
50
50
50
100
10
50
50
50
50
50
50
50
uiurc UXI.UIA
LOWER
95X
PRED_LMT
8.6
44.2
37.6
39.7
41.6
87.5
8.8
46.7
47.2
41.5
40.5
45.9
44.1
43.1
UPPER
95X
PRED_LMT
11.6
56.6
66.5
63.0
60.2
114.4
11.3
53.5
53.0
60.2
61.7
54.5
56.7
58.0
11 50 43.6 57.3
NO OF SPIKE LOWER UPPER
OBS LEVEL 95X 95X
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
100
10
100
100
100
50
100
200
100
100
100
100
100
100
100
100
100
100
PRED.LMT
90.0
8.6
80.6
76.1
84.0
42.6
82.2
164.2
87.7
81.8
83.0
85.2
85.0
89.5
85.7
88.5
89.0
83.9
PRED_LMT
111.2
11.6
124.0
131.3
119.1
58.7
121.6
243.6
114.0
122.3
120.5
117.4
117.7
111.7
116.7
113.1
112.4
119.2
X OUT OF
PRED.LMT
0.00
0.00
0.00
0.00
0.00
0.00
0.00
9.09
0.00
0.00
0.00
0.00
9.09
0.00
9.09
7. OUT OF
PRED.LMT
9.09
9.09
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
9.09
9.09
0.00
9.09
18.18
0.00
-------
USEPA INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE 10
RELATIVE RETENTION TIMES — INTERNAL STANDARD AND ISOTOPE DILUTION
ALL LABS
COMPOUND
2378-TCDD
12378-PECOD
123478-HXCDD
123678-HXCDD
1234678-HPCDD
OCDD
2378-TCDF
12378-PECDF
23478-PECDF
123478-HXCDF
123678-HXCDF
123789-HXCDF
234678-HXCDF
1234678-HPCDF
1234789-HPCDF
COMPOUND
C£- 2378-TCDD-C13
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
123789-HXCDD-C13
123789-HXCOD
1234678-HPCDD-C13
OCDO-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
OCDF
NO OF
DBS
95
96
96
96
96
96
95
96
96
96
96
96
69
96
96
NO OF
DBS
94
94
94
94
94
94
94
94
94
94
94
94
94
94
69
94
94
96
<*UAn 1 J. 1 A 1
MEAN
1.001
1.001
1.000
1.001
1.001
1.000
1.001
0.997
1.000
1.000
0.999
1.003
1.000
0.976
•lun-jauiurc i
STD
DEVIATION
0.003
0.005
0.000
0.003
0.006
0.002
0.003
0.030
0.000
0.005
0.010
0.025
0.000
0.017
0.988 0.035
Ol lAMTTTATTnM-TMTFDMAI
MwAn i x i M i £ur(— xn i cnnAi.
MEAN STD
DEVIATION
1.015
1.281
0.988
0.990
1.000
1.001
1.074
1.162
0.963
1.231
1.265
0.970
0.973
1.005
0.985
1.063
1.086
1.004
0.008
0.056
0.005
0.006
0.000
0.006
0.019
0.043
0.012
0.053
0.054
0.009
0.013
0.004
0.006
0.041
0.023
0.004
LOWER
95X
PRED_LMT
0.993
0.987
0.999
0.992
0.986
0.996
0.993
0.918
1.000
0.986
0.973
0.937
1.000
0.930
0.896
LOWER
95X
PRED.LMT
0.993
1.134
0.974
0.975
1.000
0.986
1.023
1.050
0.931
1.091
1.123
0.947
0.940
0.993
0.971
0.953
1.024
0.995
UPPER
95X
PRED_LMT
1.009
1.016
1.001
1.009
1.016
1.005
1.009
1.076
1.001
1.015
1.025
1.068
1.001
1.022
1.079
UPPER
95X
PRED_LMT
1.036
1.428
1.002
1.006
1.000
1.016
1.125
1.275
0.994
1.371
1.408
0.992
1.006
1.017
1.000
1.172
1.148
1.013
'/. OUT OF
PRED.LMT
4.21
1.04
3.13
4.17
3.13
3.13
4.21
1.04
3.13
1.04
1.04
1.04
1.45
0.00
0.00
X OUT OF
PRED_LMT
0.00
0.00
0.00
1.06
0.00
2.13
0.00
0.00
0.00
0.00
0.00
1.06
1.06
1.06
0.00
0.00
0.00
0.00
!• ~.:-7
A H
-------
USEPA INDUSTRIAL TECHNOLOGY DIVISION
METHOD 1613 PERFORMANCE EVALUATION
TABLE HA
INITIAL PRECISION AND RECOVERY — LABELED AMD NATIVE COMPOUNDS
START-UP LIMITS FOR ACCURACY
ALL LABS
COMPOUND 95 X LOWER
LIMIT
2378-TCDD 38.67
12378-PECDD 94.87
123478-HXCDD 61.80
123678-HXCDD 66.43
1234678-HPCDO 54.91
OCDO 69.37
2378-TCDF 31.56
12378-PECDF 88.54
e3478-PECDF 90.69
123478-HXCDF 51.32
123678-HXCDF 78.20
123789-HXCDF 75.85
234678-HXCDF 54.80
1E34678-HPCDF 79.12
1234789-HPCDF 73.32
ft U1LUI1UN —
95 X UPPER
LIMIT
205.96
100.72
140.35
131.82
153.00
154.63
267.82
108.17
100.46
183.94
108.86
125.85
171.53
124.23
129.83
X OUT OF
PRED_LMT
0.00
60.00
0.00
0.00
0.00
0.00
0.00
20.00
20.00
0.00
0.00
0.00
0.00
0.00
0.00
O COMPOUND 95 X LOWER 95 X UPPER
C 5 LIMIT
1 2378-TCDD-C13 26.00
2378-TCDD-CL37 31.06
12378-PECDD-C13 20.98
123478-HXCDD-C13 50.75
123678-HXCDD-C13 48.82
123789-HXCDD 45.39
1234678-HPCDD-C13 75.16
OCDD-C13 46.15
2378-TCDF-C13 16.57
12378-PECDF-C13 15.15
23478-PECDF-C13 17.47
123478-HXCDF-C13 51.78
123678-HXCDF-C13 44.86
123789-HXCDF-C13 60.57
234678-HXCDF-C13 52.58
1234678-HPCDF-C13 55.78
1234789-HPCDF-C13 41.24
OCDF 82.88
LIMIT
121.20
104.58
261.19
108.33
127.96
181.84
90.73
139.76
211.08
331.51
317.78
81.95
116.22
87.34
98.78
81.71
129.06
116.03
X OUT OF
PRED_LMT
0.00
20.00
0.00
0.00
0.00
0.00
40.00
0.00
0.00
0.00
0.00
20.00
0.00
20.00
0.00
20.00
0.00
60.00
-------
USEPA INDUSTRIAL TECHNOLOGY DIVISION
METHOD 1613 PERFORMANCE EVALUATION
TABLE 11B
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
START-UP LIMITS FOR PRECISION
ALL UBS
COMPOUND
2378-TCDO
12378-PECDD
123478-HXCDD
123678-HXCOD
1234678-HPCDD
OCOD
2378-TCDF
12378-PECDF
23478-PECDF
123478-HXCDF
123678-HXCDF
123789-HXCDF
234678-HXCDF
1234678-HPCOF
1234789-HPCDF
n-j.au lure uiuu
95 7. UPPER
LIMIT
13.13
8.43
11.02
11.72
12.73
12.63
19.87
9.29
8.49
12.50
7.26
7.21
10.18
7.06
8.32
X OUT OF
PRED_LMT
0.00
20.00
20.00
20.00
0.00
20.00
0.00
20.00
20.00
0.00
20.00
20.00
20.00
20.00
20.00
COMPOUND 95 X UPPER X OUT OF
o
GO 2378-TCDD-C13
L^. 2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
123789-HXCDD
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
OCDF
LIMIT
45.92
55.61
38.27
24.47
42.54
18.77
23.58
45.14
53.43
31.24
32.48
23.57
29.79
26.03
29.30
21.86
22.80
45.15
PRED_LMT
0.00
0.00
0.00
20.00
20.00
20.00
0.00
0.00
0.00
20.00
20.00
0.00
0.00
0.00
0.00
0.00
20.00
0.00
,
-------
U S E P A INDUSTRIAL TECHNOLOGY DIVISION
METHOD 1613 PERFORMANCE EVALUATION
TABLE 12
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
ONGOING QUALITY ASSURANCE LIMITS
ALL LABS
——---..----—--- HUMPH i i«i iun-iouiu
COMPOUND 95 X LOWER
LIMIT
2378-TCDD 59.49
12378-PECDD 71.34
123478-HXCDO 70.15
123678-HXCDD 66.61
1234678-HPCDD 74.98
OCOD 75.58
2378-TCDF 66.09
12378-PECDF 73.43
23478-PECDF 75.61
123478-HXCDF 73.86
123678-HXCDF 69.65
123789-HXCDF 74.20
234678-HXCDF 71.40
1234678-HPCDF 74.89
95 •/. UPPER
LIMIT
142.30
116.22
120.83
128.70
113.54
118.74
127.44
114.64
113.81
117.64
117.65
111.37
120.08
120.98
1234789-HPCOF 73.81 121.13
COMPOUND 95 7. LOWER 95 7. UPPER
O LIMIT LIMIT
CO
•r; 2378-TCDD-C13 29.36
-------
U S E P A
INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE 13
PRECISION AND RECOVERY OF LABELED
SUMMARY STATISTICS
ALL LABS COMBINED
COMPOUNDS
QUANTITATION=INTERNAL STANDARD
COMPOUND
2378-TCDO-C13
2378-TCOO-CL37
12378-PECDD-C13
123478-HXCDO-C13
123678-HXCDD-C13
1234678-HPCDD-C13
OCDD-C13
2378-TCOF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
NO OF
OBSERVATINS
293
291
293
294
293
293
290
294
294
289
286
294
292
204
292
294
MEAN
59.8
61.1
61.2
70.4
69.0
67.2
50.4
55.5
59.1
53.1
64.6
65.4
65.7
74.0
54.8
61.7
MEDIAN
64.0
64.0
66.0
78.0
75.0
68.0
51.0
61.0
62.0
57.0
70.0
70.
6?.
74.
58.
65.0
STD
DEVIATION
30.3
31.8
29.6
32.9
30.7
30.2
26.0
24.8
23.9
27.3
25.4
29.1
23.5
12.7
22.2
27.7
MINIMUM
3.0
2.6
3.0
.2.8
5.7
9.6
2.2
3.5
6.8
2.0
8.0
3.2
10.0
42.4
4.7
6.2
MAXIMUM
119.0
139.0
134.0
151.0
122.0
132.0
118.0
112.0
120.0
121.0
116.0
117.0
118.0
106.0
111.0
119.0
o
CO
Co
^
tE.T
-------
o
CO
USEPA INDUSTRIAL TECHNOLOGY DIVISION
METHOD 1613 PERFORMANCE EVALUATION
TABLE 1*
PRECISION AND RECOVERY OF LABELED COMPOUNDS
SUMMARY STATISTICS
LAB 2 AND LAB 3 COMBINED
QUANTITATION=INTERNAL STANDARD
COMPOUND NO OF MEAN MEDIAN STD MINIMUM MAXIMUM
OBSERVATINS DEVIATION
2378-TCDD-C13 216 71.6 77.5 25.1 3.0 119.0
2378-TCDD-CL37 214 74.2 78.0 25.7 5.8 139.0
12378-PECDO-C13 215 72.6 75.0 23.2 7.1 134.0
123478-HXCDD-C13 216 85.3 87.9 22.9 4.4 151.0
123678-HXCDD-C13 215 82.4 85.0 22.1 7.3 122.0
1234678-HPCDD-C13 215 78.0 77.0 26.0 14.0 132.0
OCDD-C13 215 59.5 59.0 21.2 10.0 118.0
2378-TCOF-C13 216 64.6 68.0 20.3 4.0 112.0
12378-PECDF-C13 216 67.4 69.0 18.9 8.0 120.0
23478-PECDF-C13 215 62.8 63.0 22.0 3.5 121.0
123478-HXCDF-C13 215 75.1 77.0 17.9 8.0 116.0
123678-HXCDF-C13 216 77.9 80.0 21.1 3.9 117.0
123789-HXCDF-C13 215 74.8 76.0 17.8 20.0 118.0
234678-HXCDF-C13 204 74.0 74.2 12.7 42.4 106.0
1234678-HPCDF-C13 214 62.9 64.3 17.6 7.0 111.0
1234789-HPCDF-C13 216 72.6 75.0 21.3 11.9 119.0
-------
o
oo
USEPA INDUSTRIAL TECHNOLOGY DIVISION
METHOD 1613 PERFORMANCE EVALUATION
TABLE 15
PRECISION AND RECOVERY OF LABELED COMPOUNDS
QUALITY ASSURANCE LIMITS
QUANTITATION=INTERNAL STANDARD
COMPOUND 95 X LOWER 95 X UPPER X OUT OF
LIMIT LIMIT PRED_LMT
2378-TCDD-C13 33.09 168.23 9.72
2378-TCDD-CL37 27.97 209.43 4.19
12378-PECDD-C13 36.92 152.56 8.33
123478-HXCDD-C13 52.11 145.05 8.80
123678-HXCDD-C13 44.91 156.80 6.02
1234678-HPCDD-C13 29.84 198.48 4.63
OCDD-C13 25.94 138.15 6.48
2378-TCDF-C13 41.31 110.93 14.81
12378-PECOF-C13 41.59 116.59 10.65
23478-PECOF-C13 29.14 131.21 7.41
123478-HXCDF-C13 48.76 122.80 7.87
123678-HXCDF-C13 41.92 149.47 5.09
123789-HXCDF-C13 37.68 152.32 4.63
234678-HXCDF-C13 49.41 110.13 7.87
1234678-HPCDF-C13 33.93 121.66 5.09
1234789-HPCDF-C13 28.41 192.74 4.63
-------
Appendix A
Outliers
To examine the data for outlying values, Hoaglin et al. (1983) provides an
approach based on the interquartile range. When observations came from a
normal distribution, points outside the following range are rejected
M + - tt> • IQR
- 2 N (.75)
where ,
M = median
IQR = interquartile range
N(l - a) = (1 - a)th percentile of normal distribution
The data were screened for outlier values using the above method. This
method of screening was applied to all three types of data sets (IPR, OPR and
EPA- type Samples) on the logarithms of the amounts. A total of 29
observations were identified as outliers in IPR- type samples (722 obs) and
deleted. From OPR-type samples, 89 observations were deleted (1624 obs). A
total of 245 observations were identified as outliers in EPA- type samples
(3455 obs) and deleted.
086
-------
Appendix B
Estimation of Variance Components
For the purpose of calculating quality control limits from the data in this
study, the variance components model assumes that the logarithm of the
percent recovery X^- measured by laboratory i and replicate j can be written
as:
log (Xjj) = fj. + EL + Ay
where,
i =1,2,3 labs,
j = l,2,...n£ replicate measurements at lab i,
p = average response,
. 2
E^ «= random interlaboratory effect with mean 0 and variance a , and
III
2
^11 = rand°m intralaboratory effect with mean 0 and variance a^.
The variance components analysis was performed by the Type 1 method using
PROC VARCOMP(SAS) to estimate the inter- and intralaboratory variance
components of the logarithms of the percent recovery. The MIVQUE method also
showed comparable results. Tables B-l through B-3 give the results of the
variance components analysis for each sample type (IPR, OPR, EPA),
re'spectively. For each compound, the total numbers of observations, the
logarithmic mean M, the square roots of the variance components Sr.
(interlaboratory) and S. (intralaboratory), and the percentage of the total
variance due to interlaboratory variation is computed as:
100 * f ——E-—— 1
087
-------
U S E P A
INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE B-l
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
RESULTS OF VARIANCE COMPONENT ANALYSIS
ALL LABS
COMPOUND
2378-TCDD
12378-PECDD
123478-HXCDD
123678-HXCDD
1234678-HPCDD
OCDD
2378-TCDF
12378-PECDF
23478-PECDF
123478-HXCDF
123678-HXCDF
123789-HXCDF
234678-HXCDF
1234678-HPCDF
1234789-HPCDF
COMPOUND
^ 2378-TCDD-C13
CO 2378-TCDD-CL37
CO 12378-PECDD-C13
123478-HXCOD-C13
123678-HXCDD-C13
123789-HXCDD
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCOF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
OCDF
NO OF
OBSERVATIONS
22
19
22
21
22
19
22
21
21
22
21
21
22
21
i j. IMI iun-1
NO OF
LABS
3
3
3
3
3
3
3
3
3
3
3
3
3
3
auiurc i
LOG
MEAN
4.49
4.58
4.53
4.54
4.51
4.64
4.51
4.58
4.56
4.57
4.52
4.58
4.57
4.60
20 3 4.58
NO OF NO OF LOG
OBSERVATIONS LABS MEAN
22
22
22
22
22
22
22
22
22
22
22
22
22
22
ie
22
22
22
3
3
3
3
3
3
3
3
3
3
3
3
3
3
2
3
3
3
3.93
3.93
4.26
4.29
4.32
4.50
4.40
4.33
3.97
4.23
4.28
4.15
4.25
4.26
4.25
4.19
4.27
4.55
WITHIN
(S_A)
0.09
0.05
0.07
0.08
0.09
0.08
0.13
0.06
0.06
0.08
0.05
0.05
0.06
0.04
0.05
WITHIN
(S_A)
0.45
0.51
0.30
0.20
0.31
0.13
0.17
0.33
0.48
0.26
0.25
0.22
0.24
0.22
0.24
0.20
0.19
0.27
BETWEEN(S_E)
0.19
0.00
0.11
0.10
0.13
0.10
0.24
0.03
0.02
0.14
0.05
0.07
0.13
0.06
0.07
BETHEEN(S_E)
0.24
0.16
0.36
0.13
0.16
0.18
0.00
0.18
0.39
0.40
0.38
0.07
0.16
0.05
0.09
0.06
0.18
0.02
7. VAR DUE -
TO LAB
80.34
0.00
67.40
60.53
70.53
64.08
76.00
22.93
7.29
75.36
50.08
67.24
79.08
63.48
66.03
'/. VAR DUE
TO LAB
22.75
8.58
58.60
27.80
20.65
66.39
0.00
23.66
40.25
70.08
68.67
9.59
29.12
5.08
12.25
7.99
46.53
0.47
-------
USEPA INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE B-2
ONGOING PRECISION AND RECOVERY ~ LABELED AND NATIVE COMPOUNDS
RESULTS OF VARIANCE COMPONENT ANALYSIS
ALL LABS
COMPOUND
2378-TCOD
12378-PECDD
123478-HXCDD
123678-HXCDD
1234678-HPCDD
OCDD
2378-TCDF
12378-PECDF
23478-PECDF
123478-HXCDF
123678-HXCDF
123789-HXCDF
234678-HXCDF
1234678-HPCDF
1234789-HPCDF
COMPOUND
o
C» 2378-TCDD-C13
1C 2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
123789-HXCDD
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDr-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
OCDF
— — — — mj
NO OF
OBSERVATIONS
49
47
44
49
44
47
47
44
45
47
47
44
47
48
IMTililAllU
NO OF
LABS
3
3
3
3
2
3
3
3
3
3
3
3
3 •
3
n-isuiurt ui
LOG
MEAN
4.52192
4.51145
4.52248
4.52818
4.52470
4.55107
4.51935
4.51909
4.53005
4.53493
4.50562
4.50980
4.52824
4.55582
WITHIN
(S_AJ
0.15
0.10
0.11
0.15
0.09
0.09
0.12
0.07
0.09
0.10
0.10
0.08
0.10
0.10
47 3 4.54922 0.11
NO OF NO OF LOG WITHIN
OBSERVATIONS
42
45
49
47
49
49
50
47
44
49
48
47
47
48
35
50
47
46
LABS
3
3
3
3
3
3
3
3
3
3
3
3
3
3
2
3
3
3
MEAN
4.09185
4.04677
4.19112
4.28357
4.31097
4.48022
4.21991
4.01236
4.07703
4.11370
4.16151
4.23546
4.25858
4.23199
4.22599
4.12521
4.19957
4.54296
(S_A)
0.28
0.37
0.29
0.17
0.20
0.14
0.26
0.28
0.28
0.32
0.31
0.19
0.19
0.20
0.21
0.23
0.26
0.11
BETWEEN! (S_E)
0.10
0.04
0.03
0.00
0.00
0.04
0.06
0.05
0.00
0.01
0.05
0.03
0.04
0.03
0.00
BETMEEN((S_E)
0.11
0.14
0.38
0.12
0.11
0.02
0.00
0.12
0.00
0.22
0.28
0.01
0.00
0.03
0.00
0.00
0.05
0.06
'/. VAR.DUE
TO LAB
30.38
15.49
6.31
0.00
0.00
18.56
21.38
29.84
0.00
0.73
23.16
11.83
14.63
6.10
0.00
7. VAR DUE
TO LAB
12.91
11.97
62.77
34.40
23.34
1.19
0.00
14.04
0.00
32.58
45.28
0.15
0.00
2.66
0.00
0.00
3.20
20.84
-------
U S E P A
INDUSTRIAL TECHNOLOGY
METHOD 1613 PERFORMANCE EVALUATION
DIVISION
TABLE B-3
PRECISION AND RECOVERY OF LABELED COMPOUNDS
RESULTS OF VARIANCE COMPONENT ANALYSIS
ALL LABS
COMPOUND
2378-TCDD-C13
2378-TCDD-CL37
12378-PECDD-C13
123478-HXCDD-C13
123678-HXCDD-C13
1234678-HPCDD-C13
OCDD-C13
2378-TCDF-C13
12378-PECDF-C13
23478-PECDF-C13
123478-HXCDF-C13
123678-HXCDF-C13
123789-HXCDF-C13
234678-HXCDF-C13
1234678-HPCDF-C13
1234789-HPCDF-C13
NO OF
OBSERVATIONS
195
195
194
202
202
206
202
194
197
206
199
204
204
203
205
202
HiAun-j.nii
NO OF
LABS
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
iKrwu SIM
L06
MEAN
4.31
4.34
4.32
4.47
4.43
4.34
4.09
4.21
4.24
4.12
4.35
4.37
4.33
4.30
4.16
4.30
WITHIN
(S_A)
0.27
0.27
0.24
0.21
0.21
0.30
0.30
0.23
0.21
0.34
0.17
0.21
0.18
0.17
0.23
0.22
BETWEEN
(S_E)
0.20
0.27
0.16
0.08
0.15
0.23
0.18
0.00
0.09
0.02
0.09
0.15
0.18
0.05
0.14
0.26
X VAR DUE
TO LAB
35.33
49.86
31.03
11.47
32.89
37.34
26.13
0.00
15.50
0.47
22.51
34.51
49.98
8.55
26.31
57.84
o
-------
Appendix C
Derivation of Quality Control Limits For Accuracy
If we observe a test series X X»j independently chosen from a normal
distribution in the unknown mean M and unknown variance a , the mean and
variance can be estimated by:
N
X = 1 ^S x.
N
S2 - \ (xL-x)2
A (1-p) 100 percent confidence interval for a single independent future
observation X from the same distribution can be constructed by noting that
x - x has mean 0 and variance (1 + 1/N) a , hence:
x € (x ± tN.x (l-p/2) J(1+1/N) S)
with probability exactly (1-p), where £„-! is the inverse cumulative t
distribution with N-l degrees of freedom. All of the quality control limits
formulae for accuracy used in this report are extensions of this concept.
Log Normal Data
If X.; has a log normal distribution with logarithmic mean n and_logarithmic
variance a , the limits can be derived by letting Y- = a log (X-) and
computing Y and S . The prediction interval for the future value of
Y- = log (X-) can be exponentiated to obtain
X £ (Exp [Y - tN.1(l-p/2) Jl-H/N S ]
(Exp [Y + tN^d-p/2) Jl+l/N Syl)
with probability 1-p.
Average of Log Normal Values
Because the start-up test is to be based on the arithmetic average of four
observations, we consider the case of predicting future x when the data are
drawn from a ^og normal distribution, with parameters fj. and a . For small
values of n, Xn can be assumed to have log normal distribution (see
References 1 and 2).
091
-------
Let Y = log (Xn)
We derive a prediction interval for Y and then exponentiate it to produce a
prediction interval for Xn.
If X-^, X-2> Xj, X^ have log normal distribution,
— o
Xn has mean m = exp (M + 1/2 a )
2 2
and variance = m
n
where , n = 4 , and
K2 = [exp(a
Using converging Taylor's series on f(x) = log(x) ,
K2 = [exp(a2 - 1]
1 k2
Y =• log (X ) will have a mean' = M + — r
n 2 2 n
2
and a variance = —
n
2 2
Since /* and a are estimated by Y,, and S as before, we have:
v f ^7 i o2 i ky 1
I " 2 y 2 n J
2 2
where, k^ = exp (S.~)-l) has asymptotic mean zero and variance approximately:
kf + £1 + I fi I]
n N 4 [ nj
(N-l)
oo o 22
Using the fact that k - a for small a to combine S^ and k^ terms^
Therefore, an approximate lOO(l-p) percent confidence interval for XR can be
computed as:
1 2 lkVl
Exp | Yn + - S^ y - tN , (1
. 2 y 2 n I
Exp v- ^ c2
Y" + 2 y" 2 n + tN-l(1'P/2) ' S
- k'y s2y l 2 s4y
where, S = —- + —- + - (I - l/n)* -
n N 2 (N-l)
Variance Components
Let Xij = M + Ei + A.J
032
-------
where,
i - 1,
j = 1,
Ei =
2, . .
2, . .
N (0, c
N (0, <
. i;
. J;
J)'
and
The estimates M, Sg, and Sf of the p, Sg, and a^ respectively are
obtainable through a variance component analysis. Since the difference X-M
has mean 0 and asymptotic variance °v + api + CTE/^ + CTA/*^ ' ancl
approximate prediction interval is given by:
M - t
- p/2)S, M
- P/2)S
where,
S =
c2 o2
S2 S2
I IJ
It will often be necessary to know the number of degrees of freedom (d) that
can be assigned to S. The degrees of freedom are given by Wilson (1979).
When the between laboratories mean square (BLMS) is much greater than the
within laboratory mean square (WLMS), a' close approximation given by:
d
1(1-1) [BLMS + (J-l)WLMSp
I(BLMS)2 + (I-1)(J-1)(WLMS)2
when, BLM < WLMS use d ~ I(J-l).
Applications
The limits for the arithmetic average (X ) of the four tests are obtained by
combining the variance components estimates to give:
exp
M
1
+ —
2
2
^
n
t(d,l-p/2)
(S
K
K
n
N
(N-L)
o 2
where M, S., and Sg are as above, Kf =• exp (Sf)-l, n=4, N is the number of
observations in the study, and L is the number of laboratories.
033
-------
The ongoing precison and recovery limits are obtained from the analysis of
the OPR type samples as:
exp
M ±
The limits for labeled compound recoveries are obtained from the analysis of
the EPA type samples as:
exp
M ±
SE
+ *l +
s£
L
+ »2
N
The ongoing calibration verification limits are obtained from the analysis of
VER-type samples as:
exp
ln(MU) ± t(d,(l-p)/2)SA
where MU is the corresponding spike level.
034
-------
,=
a -i»/.v
Appendix D
Derivation of Quality Control Limits For Precision
We need to determine the distribution of:
n
S L-
where n=4 and X^ are distributed log normally, the distribution of S depends
on a , but n can be removed from consideration by a transformation such that:
S' =
2
Finally, for a range of a values, the upper quantiles of S' were determined
by a simulation which was performed with SAS , using 10,000 replicates, and
the quantiles were estimated with PROG UNIVARIATE.
The precision limit on the standard deviation is given by:
exp(M)
where Q is the upper quantile at 1-p of SA and K is the approximate
correction factor for the estimation of S. , approximated as- (see Reference
2):
Fn-l,d(l-P)
K
and F and C are the inverse cumulative distributions of the F and Chi-squared
distributions, respectively, and d is the df of S^.
-------
Appendix E
Calibration Linearity
In order to calculate the concentration of each compound in a sample , a
calibration curve is applied to the peak area of the compound and of the
reference compound obtained from the Gas Chromatograph . This calibration
curve is constructed by the analysis of a series of calibration samples at
known concentrations .
For isotop dilution, the ratio, of the peak areas and concentrations to those
of a labeled compound are used for calibration
A/Alabel = f( c/clabel)
and the unknown concentration in a sample is constructed from the area ratio
and the known level of the standard spiked into the sample by
) clabel j
i
In estimating the calibration curve, a range of calibration samples are
used in order to evaluate the response of the instrument over its
performance range. In this study, five points were obtained for each
laboratory, at levels specified in the method (only two labs provided data
for this).
The simplest form of the response curve is proportional response curve
f(x) =ax
The random variation in the calibration response can be assumed to have a
proportional error structure, i.e., for repeated measurements the area ratio
is distributed around f (C/clabel)i as
f(C/Clabel)(l+0/
where £ has mean zero and variance a ^ i ! independent of C/C^abel\ In the case
of linear response curve
I
A/Alabel =a(C/Clabel)
and rewritting in terms of the response factor RF give
RF =>(A/Alabel)/(c/Clabel) - a(l + O-
090
-------
Hence the coefficient of variation of the response factor would be
a2a2/a = a
a constant for all concentration ratios.
If a linear propotional calibration curve is to be fitted to a calibration
set containing values AI/ A2, ..., A5, Alabell, Alabei2, •••• Alabel5,
cl' C2,.-., C5,Clabell, Clabei2' •••» Clabel5 j
, the best estimate of the calibration coefficient can be derived form the
formula for a weighted regression^ (see reference 3) as
(Ai/Alabeli)/(Ci/Ciabeli) - Vn V KF± I
Where RF is the average response factor
Linearity Specification calculation
Linearity specification is calculated based on the CAL type samples. RF was
calculated for each compound. A mild pre- screening of the RF was done.
Then, appropriate limits for testing the goddness of fit of a linear
calibration were obtained. For each compound and laboratory, the variance
among midpoint calibrations' were obtained. And this variance was
standardized by the square of the mean response factor for all calibration
points. This stadardized variance was then averaged across all laboratories,
weighted by the degrees of freedom for each laboratory, to come up with an
overall standardized variance of response factors o2|, with d total degrees
of freedom. Then for calibration, assuming the response is truely linear
with proportional error structure, the test for linearity used was to
compare the coefficient of variation of RFlf RF2,
with
C V Limit = 100 ( a2 FM_lfd(.95) ) V2
where N is the number of RFs . F is the inverse of the cumilative F
distribution, d is the degrees of freedom for o2[ These values are
calculated in Table 8
These limits were applied to actual calibration data to find out how many
points are there outside this specification.
Appendix F
• •
Relative Retention Time
,o °97
-------
Appendix F
Relative Retention Time
Before analysis, a mild pre-screening was done to relative retention time
data. Relative retention was calculated from the absolute retention time of
compound and its reference compound. Nominal scale analyses were used for
these calculations, the 95 percent confidence limits for the mean of the
quantity is computed as
x" *
where
T =
SN = [ £(Xi-"x)2/N-l]1/2
I
and fcN-l is the inverse cumilative distribution functon of the t
distribution with N-l degrees of freedom. The prediction limits are given by
X ±
. The limits are given in Table 10 . Also given are the percentage of times
the observations fell outside the calculated limits.
11
038
-------
References for Appendices
Eynon B.P, Maxwell C., and Valder A. Interlaboratorv Validation of U.S.
Environmental Protection Agency Method 1625A. USEPA Contract 68-01-6192, June
1984.
Hoaglin, D.C., Hosteller, -P.,-and Turkey, J.S. Understanding Robust and
Exploratory Data Analysis. John Wiley and Sons, New York, pp 37-39, 1983.
Milliken G.A., Johnson D.E. Analysis of Messy Data. Lifetime Learning
Publications, CA, pp 238-241, 1984.
Wilson, A.L. Approach for Achieving Comparable Analytical Results From a
Number of Laboratories. The Analyst, Vol. 104 No. 1237, p 273, 1979.
Hunt, D.T., Wilson, A.L. The Chemical Analysis of Water. The Royal Society of
Chemistry, London, pp 252-254, 1986.
U.S. EPA Development Document for Existing Source Pretreatment Standards for
the Electroplating Point Source Category, USEPA, Document 440/1-79/003,
August 1979.
090
-------
FIGURE 1
INITIAL PRECISION AND RECOVERY OF LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT OF ALL LABS COMBINED
100
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
COOE=2378-TCDF-C13
150.
130.
S
c
H
E
M
A
T
I
C 110.
F
0 90.0
R
O
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
. FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=2378-TCDF
150.
130.
110.
90.0
O
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY - LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLCT ~ ALL LABS COMBINED
CODE=2378-TCDD-C13
150.
130.
S
c
H
E
M
A
T
I
C 110.
F
0 90.0
R
O
Go
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=237B-TCDD-CL37
150.
130.
110.
F
0 90.0
R
V
A
H ^-*
I
A
B """
L 70.0
E
o
*- + -*
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED '
CODE=2378-TCDD
150.
130.
S
c
H
E
M
A
T
I
C 110.
F
0 90.0
R
O
v
A
R
I
A ^
B
L 70.0
E
•f
* *
I I
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=23478-PECDF-C13
150.
130.
110.
F
0 90.0
R
O
70.0
50.0
30.0
*-+-#
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED .
CODE=2347fl-PECDF
150.
130.
110.
90.0
I
I
•f *
I I
*-+-*
+ +
I
o
^ J
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=234678-HXCDF-C13
150.
130.
110.
90.0
O
CC
70.0
+ *
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=33
-------
INITIAL PRECISION AND RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123789-HXCDF-C13
150.
130.
S
c
H
E
M
A
T
I
C 110.
F
0 90.0
R
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123789-HXCDF
150.
130.
110.
90.0
* *
I + I
+ +
I
0
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — UBELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=123789-HXCDD
150.
130.
110.
+ 1
90.0
•f +
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=12378-PECDF-C13
150.
130.
S
c
H
E
M
A
T
I
C 110.
90.0
70.0
50.0
•f +
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=12378-PECDF
150.
130.
110.
I
I
*
0 90.0
R
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=1237fl-PECDD-C13
150.
130.
110.
F
0 90.0
R
V
A \-
< 5
B
L 70.0
E
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=12378-PECDD
150.
130.
110.
90.0
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=123676-HXCDF-C13
150.
130.
S
C
H
E
M
A
T
I
C 110.
90.0
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY ~ LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=123678-HXCDF
150.
130.
110.
90.0
*
0
I
I
+-•—+
*-*-*
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=123678-HXCDD-C13
150.
130.
S
c
H
E
M
A
T
I
C 110.
F
0 90.0
R
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=12367fl-HXCDD
150.
130.
110.
F
0 90.0
R
ro
o
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=1234789-HPCDF-C13
150.
130.
110.
F
0 90.0
R
FO
70.0
*- + -*
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY ~ LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PIOT ~ ALL LABS COMBINED
CODE=12347*9-HPCDF
150.
130.
110.
F
0 90.0
R
*
I
I
I
* --- *
I + I
+ --- +
I
*
*
*
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123478-HXCDF-C13
150.
130.
S
C
H
E
M ,
A
T
I
C 110.
F
0 90.0
R
fO
CO
70.0
*-+-*
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=123478-HXCDF
150.
130.
S
c
H
E
M
A
T
I
C 110.
«-•»-*
F
0 90.0
R
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123478-HXCDD-C13
150.
130.
110.
0 90.0
R
V
A
R
I
A
B
L 70.0
E
ro
C/!
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY ~ LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=1E3478-HXCDD
150.
130.
110.
90.0
I I
I I
*-+-*
I I
rc
cr
70.0
50.0
30.0
-------
INITIAL PRECISION ANO RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123467fl-HPCDF-C13
150.
130.
110.
F
0 90.0
R
70.0
I +
I I
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=123467fl-HPCDF
150.
130.
110.
0 90.0
R
ro
CO
70.0
50.0
30.0
*-+-*
+ 1
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=1234678-HPCDD-C13
150.
130.
110.
F
0 90.0
R
ro
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=1E34678-HPCDD
150.
130.
S
C
H
E
M
A
T
I
C 110.
K *
•f
90.0
70.0
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=OCDF
150.
F
0
R
V
A
R
I
A
B
L
E
P
R
E
C
130.
S
C
H
E
M
A
T
I
C 110.
P
L
0
T
S
90.0
70.0
50.0
30.0
+—t
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=OCDO-C13
150.
130.
S
C
H
E
M
A
T
I
c no.
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
-------
INITIAL PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=OCDD
150.
130.
110.
S
C
H
E
tl
A
T
I
C
P
L
0
T
S
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
P
R
E
C
50.0
30.0
I I
I I
I + I
* *
I I
+ +
I
I
I
I
0
-------
FIGURE 2
ONGOING PRECISION AND RECOVERY OF LABELED AND NATIVE
COMPOUNDS
FREQUENCY DISTRIBUTION PLOT OF ALL LABS COMBINED
134
-------
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT « ALL LABS COMBINED
COOE=2378-TCDF-C13
150.
130.
S
C
H
E
tl
A
T
I
C 110.
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
LO
CO
«-+-*
-------
ONGOING PRECISION AND RECOVERY — LABELED ArlPWKlIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=Z37S-TCDF
150.
130.
110.
F
0 90.0
H
70.0
50.0
30.0
CO
-------
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=2378-TCDD-C13
150.
130.
S
C
H
E
M
A
T
I
C 110.
90.0
70.0
50.0
30.0
oc
-------
ONGOING PRECISION AND RECOVERY -- LABELED ATOTlATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=2378-TCDD-CL37
150.
130.
110.
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
CO
CO
-------
ONGOING PRECISION AMD RECOVERY — LABELED AMD NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
COOE=2378-TCDD
150.
130.
110.
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
tt-t-*
CO
-------
ONGOING PRECISION AND RECOVERY — LABELED AHJWPtlVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=23<*78-PECDF-C13
150.
130.
S
C
H
E
n
A
T
I
c no.
F
0 90.0
R
70.0
50.0
30.0
«•."*(
-------
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=23<*78-PECDF
150.
130.
110.
90.0
0
I
I
I
I
+ +
I I
I I
*-+-*
I I
+ +
I
I
I
0
70.0
50.0
30.0
-------
ONGOING PRECISION AND RECOVERY -- LABELED AJATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=a34678-HXCDF-C13
150.
130.
110.
90.0
70.0
50.0
30.0
-------
ONGOING PRECISION AND RECOVERY.-- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LA3S COMBINED
CODE=23467,l-HXCDF
150.
130.
S
C
H
E
M
A
T
I
C 110.
0 90.0
R
70.0
50.0
30.0
I I
I I
* *
I > I
I I
I I
-------
ONGOING PRECISION AND RECOVERY ~ LABELED^BT NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123789--HXCDF-C13
150.
130.
S
C
H
E
M
A
T
I
c no.
p
L
o
T
S
F
0
R
V
A
R
I
A
B
L
E
90.0
70.0
*-+-*
P
R
E
C
50.0
30.0
-------
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123789-HXCDF
150.
130.
110.
f
0 90.0
R
70.0
50.0
30.0
I
I
•f 1
I I
* *
-------
ONGOING PRECISION AKD RECOVERY — LABELED ATWTATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=123789-HXCDD
150.
130.
110.
F
0 90.0
R
70.0
50.0
30.0
-------
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=12378-PECDF-C13
150.
130.
110.
90.0
70.0
50.0
30.0
-------
ONGOING PRECISION AND RECOVERY -- LABELED AriVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=12378-PECDF
150.
130.
S
C
H
E
M
A
T
I
c no.
F
0 90.0
R
70.0
50.0
30.0
0
I
I
I
I
4 «•
I I
»- + -»
I I
+ +
I
I
I
0
0
CO
-------
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=12378-PECDD-C13
150.
130.
S
C
H
E
M
A
T
I
C 110.
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
-------
ONGOING PRECISION AND RECOVERY — LABELED A^^IATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=12378-PECDD
150.
130.
110.
F
0 90.0
R
70.0
50.0
30.0
•f +
I I
*-+-*
I I
I I
I I
-------
ONGOING PRECISION AND RECOVERY ~ LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
COD E=12 36 78-HXCD F-C13
150.
130.
S
C
H
E
M
A
T
I
C 110.
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
*-+-*
-------
ONGOING PRECISION AND RECOVERY — LABELED
FREQUENCY DISTRIBUTION PLOT — ALL
COOE=133678-HXCDF
NATIVE COMPOUNDS
LABS COMBINED
150.
130.
S
C
H
E
M
A
T
I
C
P
L
0
T
S
F
0
R
V
A
R
I
A
B
L
E
110.
10
90.0
I I
*- + -*
I I
70.0
P
R
E
C
50.0
30.0
-------
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123678-HXCDD-C13
150.
130.
S
C
H
E
M
A
T
I
C 110.
0 90.0
R
70.0
50.0
30.0
CO
to
4 4
*—*
I I
4 4
-------
ONGOING PRECISION AND RECOVERY -- LABELED A^WIftTIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123678-HXCDO
150.
130.
110.
90.0
I I
I I
I + I
* *
I I
I I
•f +
70.0
50.0
30.0
-------
ONGOING PRECISION AMD RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=1234789-HPCDF-C13
150.
130.
110.
LO
LQ
90.0
70.0
*—*
50.0
30.0
-------
ONGOING PRECISION AND RECOVERY — LABELEB^TO NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
COOE=1234789-HPCDF
150.
130.
no.
s
c
H
E
M
A
T
I
C
P
L
0
T
S
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
P
R
E
C
50.0
30.0
I I
I I
I I
t --- +
I
I
CO
LO
•H
-------
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUND?
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=123478-HXCDF-C13
150.
130.
110.
LO
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
-------
p
R
E
C
ONGOING PRECISION AND RECOVERY -- LABELE^fib NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123478-HXCDF
150.
130.
110.
S
C
H
E
M
A
T
I
C
P
L
0
T
S
F
0 90.0
R
V
A
R
I
A
B
L
E
70.0
50.0
30.0
I
I
*-+-*
I
I
I
* I
CO
LO
-------
ONGOING PRECISION AMD RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLCT ~ ALL LABS COMBINED
= ia3478-IIXCDD-C13
150.
130.
S
c
H
E
n
A
T
I
C 110.
f
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
I I
I I
I I
* *
+ -f
-------
ONGOING PRECISION AND RECOVERY -- LABELED TWT NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=123<>78-HXCDD
150.
130.
110.
90.0
I I
I I
I I
I I
« *
•f +
70.0
50.0
30.0
-------
ONGOING PRECISION AND RECOVERY ~ LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=1234678-HPCDF-C13
150.
130.
S
C
H
E
tt
A
T
I
C 110.
90.0
V
A
R
I
A
B
L 70.0
E
50.0
30.0
CD
*—*
-------
ONGOING PRECISION AND RECOVERY « LABELED TfBTNATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=123<«678-HPCDF
150.
130.
110.
F
0 90.0
R
70.0
50.0
30.0
+ *
I I
X—f-*
I I
I I
I I
4. +
CD
-------
ONGOING PRECISION AND RECOVERY — LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=1234678-HPCDD-C13
150.
130.
S
C
H
E
M
A
T
I
C 110.
F
0 90.0
R
70.0
50.0
30.0
CJD
-------
ONGOING PRECISION AND RECOVERY — LABELED^PB NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123<*678-HPCDD
150.
130.
S
C
H
E
M
A
T
I
C 110.
P
L
0
T
S
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
P
R
E
C
50.0
30.0
I I
-------
ONGOING PRECISION AND RECOVERY — LABELED AMD NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=OCDF
150.
130.
S
C
H
E
M
A
T
I
C
P
L
0
T
S
F
0
R
V
A
R
I
A
B
L
E
110.
*-+-*
90.0
10
CD
70.0
P
R
E
C
50.0
30.0
-------
p
R
E
C
ONGOING PRECISION AND RECOVERY — LABELED^WT NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=OCDD-C13
150.
130.
S
C
H
E
M
A
T
I
C 110.
P
L
0
T
S
F
0 90.0
R
V
A
R
I
A
B
L
E
70.0
50.0
30.0
CD
CD
*—*
-------
ONGOING PRECISION ANO RECOVERY -- LABELED AND NATIVE COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=OCDD
150.
130.
110.
S
C
H
E
M
A
T
I
C
P
L
0
T
S
F
0 90.0
R
V
A
R
I
A
B
L 70.0
e
p
R
E
C
50.0
30.0
I I
I I
CD
-------
-, • ";".--. '. 3
FIGURES
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT BY LAB
168
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTIOf4 PLOT — BY LAB
CODE=Z378-TCDF-C13
150.
130.
S
C
H
E
M
A
T
I
c no.
90.0
70.0
50.0
30.0
LAB
LAB
i i
i i
i i
-*-+-*-
LABI
—o —
LAB2
LAB3
-------
PRECISION AND RECOVERY OF LABE LEO^WftPOUNDS
FREQUENCY DISTRIBUTION PLOT — BY LAB
CODE=2378-TCDD-C13
150.
130.
110.
F
0 90.0
R
70.0
50.0
I I
I I
[Ail-
—o—
LAB2
I
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTE IN PLOT -- BY LAB
CODE=2378-Tt OD-CL37
150.
130.
S
C
H
E
M
A
T
I
C 110.
90.0
70.0
50.0
•f 4-
I I
I I
"'LAB!"
o
—o—
LAB2
t
*
-LAB3-
-------
PRECISION AND RECOVERY OF LABELED^PGHPOUNDS
FREQUENCY DISTRIBUTION PLOT — BY LAB
CODE=23<»78-PECDr-C13
150.
130.
110.
CM
r-
90.0
70.0
*
*
0
0
•f •»•
*-+-*
*-+_*
50.0
LAB2
te3
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — BY LAB
COD E=Z346 78-HXCD F-C13
150.
130.
S
C
H
E
M
A
T
I
C 110.
90.0
70.0
50.0
I I
I I
I I
*-*-*
I I
I I
*-+-*
LAB2
LAB3
-------
PRECISION AND RECOVERY OF LABELED^OMPCUNDS
FREQUENCY DISTRIBUTION PLOT —' BY LAB
CODE=1237S9-HXCDF-C13
150.
130.
S
C
H
E
II
A
T
I
C 110.
F
0 90.0
R
70.0
50.0
Oil-
*
--*
UB2
-— I
"UBS"
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- BY LAB
COOE=12378-PECDF-C13
150.
130.
110.
90.0
70.0
50.0
*
ft
LO
r-
ft~f-»
LABI
LAB2
+ 1
I
I
I
I
~~LAB3~
-------
PRECISION AND RECOVERY OF LABELED
FREQUENCY DISTRIBUTION PLOT — BY LAB
CODE=13378-PECDD-C13
150.
130.
110.
F
0 90.0
R
70.0
50.0
CD
*-+-*
I I
I I
I I
•'LABI'
0
0
0
LABE
•f
1
"U
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — BY LAB
COD E=12 36 78-HXCD F-C13
150.
130.
110.
f>
f
0 90.0
R
70.0
50.0
+ 1
Ail'
LAB
—0—
LAB2
"LA!
B3
-------
PRECISION AND RECOVERY Or LABELED^CflPOUNDS
FREQUENCY DISTRIBUTION PLOT -- BY LAB
CODE=1E3678-HXCDD-C13
150.
130.
110.
t 1
F
0 90.0
R
70.0
50.0
* *
I I
I I
I I
I I
0
LAB2
tiii"
LABI
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — BY LAB
CODE=1Z34789-HPCDF-C13
150.
130.
S
c
H
E
M
A
T
I
C 110.
F
0 90.0
R
70.0
50.0
•f +
1
1
1
1
+ 1
_l_l
^
^ H
1
i-
*-+-*
CD
LAB2
AB3
-------
PRECISION AND RECOVERY OF LABELED^fflPOUNDS
FREQUENCY DISTRIBUTION PLOT — BY LAB
COOE=123478-HXCOF-C13
150.
130.
no.
90.0
70.0
50.0
*-+-*
LAB2
LA83
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — BY LAB
CODE=123478-HXCDD-C13
150.
130.
110.
90.0
70.0
50.0
1 1
I 1
1 1
1 I
1 _*-|
CO
LAB
LAB2
LAB3
-------
PRECISION AND RECOVERY OF LABELED
FREQUENCY DISTRIBUTION PLOT — BY LAB
CODE=1234676-HPCDF-C13
150.
130.
S
C
H
E
ft
A
T
I
c no.
90.0
70.0
50.0
cvj
oo
*—*
*-+-*
+---*
* tt-
LAB1
—o —
LAB2
TAB!
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — BY LAB
CODE=1234678-HPCDD-C13
150.
130.
110.
CO
CO
90.0
70.0
50.0
[Ail
UB2
I
I
I
LABS"
-------
PRECISION AND RECOVERY CH: LABELED
FREQUENCY DISTRIBUTION PLOT — BY
CODE=OCOD-C13
IPOUNDS
LAB
150.
130.
110.
f
0 90.0
R
70.0
50.0
00
4 +
•f—
I
LAB2
Jii"
LAB
-------
^
FIGURE 4
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT OF ALL LABS COMBINED
185
-------
PRECISION AND RECOVERY CP LABELED^BtlPOUNDS
FREQUENCY DISTRIBUTION PLCT — ALL LABS COMBINED
CODE=2378-1CDF-C13
150.
130.
110.
CO
90.0
70.0
50.0
30.0
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODE=2378-TCOD-C13
150.
130.
110.
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
CO
-------
PRECISION AND RECOVERY OF LABE LED^PBMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=2378-TCOD-CL37
ISO.
130.
110.
CO
co
90.0
70.0
50.0
30.0
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=23<*7fl-PECDF-C13
150.
130.
S
C
H
E
M
A
T
I
C 110.
90.0
70.0
50.0
30.0
CD
CO
-------
PRECISION AND RECOVERY OF LABELED "
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=234678-HXCDF-C13
150.
130.
110.
F
0 90.0
R
70.0
50.0
30.0
O
CD
*-+-*
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
COOE=123789-HXCDF-C13
150.
130.
110.
90.0
70.0
50.0
30.0
-------
PRECISION AND RECOVERY OF LABELED^ElPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=1237e-PECDF-C13
150.
130.
110.
0 90.0
R
70.0
50.0
30.0
(73
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=1Z378-PECDD-C13
150.
130.
110.
CO
a:
90.0
70.0
50.0
30.0
-------
PRECISION AND RECOVERY OF LABELE0^5(1POUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123678-HXCDF-C13
150.
130.
S
c
H
E
M
A
T
I
C 110.
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
CD
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123678-HXCDO-C13
150.
130.
110.
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
LO
CO
*—*
-------
PRECISION AND RECOVERY OF LABELED^RMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=1334789-HPCDF-C13
ISO.
130.
S
C
H
E
M
A
T
I
C 110.
F
0 90.0
R
V
A
R
I
A
B
L 70.0
E
50.0
30.0
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT ~ ALL LABS COMBINED
CODEM23478-HXCDF-C13
150.
130.
S
C
H
E
(1
A
T
I
C 110.
90.0
70.0
50.0
30.0
-------
PRECISION AND RECOVERY OF LABELEB^COMPOUHDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=123478-HXCDD-C13
150.
130.
S
C
H
E
M
A
T
I
C
P
L
0
T
S
F
0
R
V
A
R
I
A
B
L
E
110.
CO
90.0
70.0
50.0
30.0
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=1E34678-HPCDF-C13
150.
130.
S
c
H
E
M
A
T
I
C 110.
F
0 90.0
R
70.0
50.0
30.0
03
+—*
-------
PRECISION AND RECOVERY OF LABELgS^OMPOUNDS
FREQUENCY DISTRIBUTION PLOT — ALL LABS COMBINED
CODE=1234678-HPCDD-C13
150.
130.
S
C
H
E
M
A
T
I
C
P
L
0
T
S
F
0
R
V
A
R
I
A
B
L
E
110.
o
CM
90.0
70.0
*-+-*
P
R
E
C
50.0
30.0
-------
PRECISION AND RECOVERY OF LABELED COMPOUNDS
FREQUENCY DISTRIBUTION PLOT -- ALL LABS COMBINED
CODE=OCDD-C13
150.
130.
S
C
H
E
n
A
T
I
C 110.
90.0
70.0
50.0
30.0
O
CM
*-+-*
-------
USEPA OFFICE OF WATER
STUDY PLAN FOR THE EVALUATION OF METHOD 1613
May 1990
USEPA Office of Water
Office of Water Regulations and Standards
Industrial Technology Divsion (WH-552)
Washington, DC 20460
-------
TABLE OF CONTENTS
Study Plan Section 1
Data Reporting Forms Section 2
Description of Quality Control Solutions Section 3
Description of PE Spike Solutions Section 4
203
-------
SECTION 1
STUDY PLAN FOR THE EVALUATION OF EPA METHOD 1613
204
-------
Introduction
The USEPA Office of Water intends to promulgate Method 1613 on a nationwide basis under the
Clean Water Act (as amended, 1987) for the determination of polychlorinated dibenzo-p-dioxins and furans
(dioxins and furans). The objective of this study is to evaluate the use of high resolution gas
chromatography mass spectrometry (HRGC/HRMS) for the determination of dioxins and furans in
industrial effluents. The data from the study will be used to assess the performance of the method, to
determine the intra-laboratory and inter-laboratory components of variability for the method, and to
generate improved method specifications. Laboratories successfully completing this study will be qualified
to perform Method 1613 for the Agency and other clients.
Because of the problems associated with shipping large volumes of wastewater effluent to all
participants, the Agency has prepared a spiking solution to be added to reagent water. This spiking solution
resulted from the batch extraction of wastewaters collected from several pulp and paper facilities and
combined prior to extraction.
Sample Preparation
Each laboratory will receive two (2) flame-sealed ampules containing the spiking solutions, along
with specific instructions for preparing two (2) synthetic samples (the "Performance Samples") from the
solutions.
Extraction and Analysis
Each prepared sample will be spiked with the 15 labeled analogs of the 2,3,7,8-substituted PCDDs
and PCDFs listed in Table 1 of Method 1613. After spiking, the samples will be extracted according to the
procedures, in Method 1613, filtering water samples before extraction. The particulates collected on the
filter are to be extracted separately, using the Soxhlet-Dean Stark apparatus described in the method. The
filtered water is extracted in a separatory funnel, and combined with the paniculate extract after solvent
exchange.
The remainder of the analytical procedures are detailed in the method itself, and must be followed as
written.
Quality Assurance/Quality Control
As with all the 600 and 1600 series EPA methods, there are a series of prescribed QA/QC analyses
that are essential to the process. This study seeks to evaluate these aspects of the method as well.
Therefore, participating laboratories must be prepared to provide data for the following non-sample
analyses:
1. Start-up tests in Section 8.2 (Initial Precision and Accuracy) and Ongoing Precision and
Accuracy (Section 14.5)
The Initial Precision and Accuracy test consists of the analysis of four aliquots of reagent water
spiked with the native and labeled analytes listed in Table 1 (PAR solution, Sections 6.14, and
8.2 of the method). Note that these analyses are required in addition to the two samples
prepared from the pulp and paper spiking solution. They are used to demonstrate the
laboratory's ability to generate data of acceptable initial precision and recovery (IPR) in a
reference matrix Laboratories who have generated IPR data previously through the use of
Method 1613. may submit those data.
£ U O
-------
In addition, if the analyses of standards and samples for this study continue beyond a 12-hour
shift beginning with the injection of the first calibration standard, the Ongoing Precision and
Accuracy analysis is required. This test demonstrates that ongoing precision and recovery
(OPR) meets the QC limits specified in the Method (see Section 14.5). If the analyses of the
standards and samples above can be completed in a single 12-hour time period, no OPR
analysis is required.
2. Method blank in Section 10.4.4
This reagent water aliquot spiked with the labeled analogs in Table 1 serves as a check on
contamination. One blank must be extracted along with the Performance Samples, and
submitted with that data.
3. Initial Calibration and Calibration Verification Data in Sections 7.5 and 14.1
A five point initial calibration must be performed using the compounds and concentrations
listed in Table 4 of the method. Unless all the samples and blanks, including the start-up test
aliquots, are analyzed in the 12 hour period immediately after the initial calibration, a
calibration verification standard must be analyzed at the begining of the 12 hour shift, as
described in Section 14.1.
Deliverables
The deliverables for this study consist of the following materials:
1. A narrative discussion of the data outlining any problems encountered during the analyses,
and detailing all steps taken to overcome them. Additional comments on the method write-
up and specifications are welcome.
2. Sample concentration data calculated according to the specifications in the method,
recoveries of labeled compounds in all samples and standards, and calibration data, including
response factors for all native and labeled compounds. Draft forms have been provided with
this outline. These forms may be photocopied and completed by hand, or the laboratory may
substitute forms in another format, so long as all data elements contained on the draft forms
appear in the other format as well.
3. Raw data, in the form of selected ion current profiles, data system reports, or handwritten
tabulations of instrument data, containing peak areas, for all analyses, including calibration
standards, start-up test aliquots, samples, and blanks. These data, in conjunction with the
forms above, must be sufficient to allow an independant reviewer to reconstruct all
calculations and quantitations performed by the laboratory.
4. Copies of log books, bench notes, etc., detailing the processing of samples and sample
extracts.
The data described above should be sent to:
USEPA Sample Control Center
P. O. Box 1407
Alexandria, VA 22313
USA
or, if sent by overnight courier, use street address of:
0 *i
\j o
-------
USEPA Sample Control Center
300 North Lee Street
Alexandria, VA 22314
USA
Statistical Analysis
Standard statistical methods for the analysis of collaborative tests will be used to analyze the data
submitted by the laboratories. Samples will be coded and their assignment to laboratories randomized so
that the identity of each sample and its contents will be unknown to the laboratory performing the analysis.
Statistical procedures (see Youden, WJ., Statistical Techniques for Collaborative Tests. Association of
Official Analytical Chemists, 1973) will be used to estimate intra-laboratory and inter-laboratory
components of variability and assess precision and accuracy for different concentration levels.
Questions
Questions regarding participation in the study or the method in particular should be addressed to Or.
Harry McCarty, Sample Control Center, at the above address, by telephone at 703-557-5040, or by telefax at
703-683-0378.
-------
SECTION 2
DATA REPORTING FORMS FOR THE EVALUATION OF EPA METHOD 1613
:08
-------
USEPA Industrial Technology Division
Sample Control Center
P. O. Box 1407 - Alexandria, VA 22313 ,,
703/557-5040 FTS 8-557-5040 ;1
COVER PAGE - PCDD/PCDF ANALYSES DATA PACKAGE
Lab Name: Contract No.: SAS No.:
Episode No.: EPA Method No.: Method Issue/Rev. Date;
Industrial Category: Program:
EPA Sample No. Lab Sample ID
Comments: Narrative Report is attached. (Yes)
I certify that this data package is in compliance with the terms and
conditions of the contract, both technically and for completeness, for
other than the conditions detailed in the Narrative Report. Release of
the data contained in this hardcopy data package (and in the data submitted on
magnetic media, if data is submitted on magnetic media) , has been
authorized by the Laboratory Manager or the Manager's designee, as verified
by the following signature.
Signature: Name:
Date: Title:
3/90
2 Of)
-------
USEPA — ITD • ' J
FORM 1A EPA SAMPLE NO.
PCDD/PCDF ANALYSIS DATA SHEET
Use for Sample and Blank Results
Lab Name: Episode No.:
Contract No.: SAS No.: Lab Sample ID:
Matrix (aqueous/solid/leachate): Sample Wt/Vol: • g or mL:
Sample Receipt Date: Initial Calibration Date:
Ext. Date: Shift: Instrument ID:
Analysis Date: Time: GC Column ID:
Extract Volume (uL): Sample Data Filename:
Injection Volume (uL): Blank Data Filename:
Dilution Factor: Cal. Ver. Data Filename:
Concentration Units (pg/L or ng/Kg dry weight): % Solids:
CONCENTRATION DETECTION ION ABUND. RRT
ANALYTE FOUND LIMIT RATIO (1) (1)
2378-TCDD
12378-PeCDD
123478-HxCDD ._
123678-HxCDD
123789-HXCDD
1234678-HpCDD
OCDD
2378-TCDF
12378-PeCDF
23478-PeCDF
123478-HXCDF
123678-HxCDF
123789-HxCDF
234678-HxCDF "
1234678-HpCDF
1234789-HpCDF
OCDF
Total TCDD
Total PeCDD
Total HxCDD
Total HpCDD
Total TCDF
Total PeCDF
Total HxCDF
Total HpCDF
(1) Contract-required limits for RRTs and ion abundance ratios are specified
in Tables 2 and 3A, respectively, Method 1613.
4/90
-------
USEPA — ITD
FORM IB EPA SAMPLE NO.
'PCDD/PCDF CONFIRMATION ANALYSIS DATA SHEET
Lab Name: Episode No.:
Contract No.: SAS No.: Lab Sample ID:
Matrix (aqueous/solid/leachate): Sample Wt/Vol: g or mL:
Sample Receipt Date: Initial Calibration Date:
Ext. Date: Shift: Instrument ID:
Analysis Date: Time: GC Column ID:
Extract Volume (uL): Sample Data Filename:
Injection Volume (uL): Blank Data Filename:
Dilution Factor: Cal. Ver. Data Filename:
Concentration Units (pg/L or ng/Kg dry weight): % Solids:
CONCENTRATION DETECTION ION ABUND. RRT
ANALYTE FOUND LIMIT RATIO (1) (1)
2378-TCDD
12378-PeCDD
123478-HxCDD
123678-HxCDD
123789-HXCDD
1234678-HpCDD
OCDD
2378-TCDF
12378-PeCDF
23478-PeCDF
123478-HXCDF
123678-HxCDF
123789-HxCDF
234678-HxCDF
1234678-HpCDF
1234789-HpCDF
OCDF
(1) Contract-required limits for RRTs and ion abundance ratios are specified
in Tables 2 and 3A, respectively, Method 1613.
4/90
21.1
-------
. . -.•„. - ••-.-. I'- • - • ..-• •#•••••• • t
DSEPA - ITD
FORM 2
PCDD/PCDF LABELED COMPOUND AND EPA SAMPLE1 NO.
CLEANUP STANDARD RECOVERIES
Lab Name: Episode No.:
Contract No.: SAS No.: Lab Sample ID:
Matrix (aqueous/solid/leachate) : Sample Wt/Vol: g or mL:
Sample Receipt Date: Initial Calibration Date:
Ext. Date: Shift: Instrument ID:
Analysis Date: Time: GC Column ID:
Extract Volume (uL): Sample Data Filename:
Injection Volume (uL): Blank Data Filename:
Dilution Factor: Cal. Ver. Data Filename:
Concentration Units (pg/L or ng/Kg dry weight): % Solids:
ION
SPIKE CONC. R(%) ABUND. RRT
CONC. FOUND (1) RATIO (2) (2)
LABELED COMPOUNDS
13C-2378-TCDD
13C-12378-PeCDD .
13C-123478-HXCDD •_
13C-123678-HXCDD
13C-1234678-HpCDD
13C-OCDD
13C-2378-TCDF
13C-12378-PeCDF
13C-23478-PeCDF
13C-123478-HXCDF
13C-123678-HXCDF
13C-123789-HXCDF
13C-234678-HXCDF
13C-1234678-HpCDF
13C-1234789-HpCDF
CLEANUP STANDARD
37C14-2378-TCDD
(1) Contract-required limits for percent recovery (R) are 25-150% (Section
8.3.3, Method 1613).
(2) Contract-required limits for RRTs and ion abundance ratios are specified
in Tables 2 and 3A, respectively, Method 1613.
4/90
1 '>
-L »„
-------
USEPA - ITD ; •";-
FORM 3A
PCDD/PCDF INITIAL CALIBRATION RELATIVE RESPONSES
Lab Name
Episode No.:
Contract No.:
SAS No.:
Initial Calibration Date:
Instrument ID:
CS1 Data Filename:
CS2 Data Filename:
CS3 Data Filename:
NATIVE ANALYTES
2378-TCDD
12378-PeCDD
123478-HXCDD
123678-HXCDD
123789-HxCDD (2)
1234678-HpCDD
OCDD
2378-TCDF
12378-PeCDF
23478-PeCDF
123478-HxCDF
123678-HxCDF
123789-HXCDF
234678-HxCDF
1234678-HpCDF
1234789-HpCDF
OCDF (3)
GC Column ID:
CS4 Data Filename:
CSS Data Filename:
RELATIVE RESPONSE (RR)
MEAN
RR
CS1
CS2
CS3
CS4
CSS
Cv
(%RSD)
(1)
(1) For contract Cv specifications, see Section 7.5.4, Method 1613.
(2) Response Ratios are calculated relative to the labeled analogs of the
other two HxCDDs (Section 16.1.2, Method 1613).
(3) Response Ratios are calculated relative to the labeled analog of OCDD
(Section 16.1.1, Method 1613).
4/90
213
-------
Lab Name:
DSEPA - ITD
FORM 3B
PCDD/PCDF INITIAL CALIBRATION RESPONSE FACTORS
Episode No.:
Contract No.:
SAS No.:
Initial Calibration Date:
Instrument ID:
CS1 Data Filename:
CS2 Data Filename:
CS3 Data Filename:
GC Column ID:
CS4 Data Filename:
CSS Data Filename:
RESPONSE FACTOR (RF)
MEAN
RF
LABELED COMPOUNDS
13C-2378-TCDD
13C-12378-PeCDD
13C-123478-HXCDD
13C-123678-HXCDD
13C-1234678-HpCDD
13C-OCDD
13C-2378-TCDF
13C-12378-PeCDF
13C-23478-PeCDF
13C-123478-HXCDF
13C-123678-HXCDF
13C-123789-HXCDF
13C-234678-HXCDF
13C-1234678-HpCDF
13C-1234789-HpCDF
CLEANUP STANDARD
37C14-2378-TCDD
CS1
CS2
CS3
CS4
CSS
CV
(%RSD)
(1)
(1) For assignment of labeled compounds to internal standards, see Table 2. F
contract Cv specifications, see Section 7.6.3, Method 1613.
4/90
214
-------
Lab Name:
•'••"" .:,: ... USBPA - ITD
FORM 3C
PCOD/PCDF INITIAL CALIBRATION ION ABUNDANCE RATIOS
Episode No.:
Contract No . :
SAS No.:
Initial Calibration Date:
Instrument ID:
CS1 Data Filename:
CS2 Data Filename:
CS3 Data Filename:
NATIVE ANALYTES
2378 TCDD
12378 PeCDD
123478 HxCDD
123678 HXCDD
123789 HxCDD
1234678 HpCDD
OCDD
2378 TCDF
12378 PeCDF
23478 PeCDF
123478 HxCDF
123678 HxCDF
123789 HXCDF
234678 HxCDF
1234678 HpCDF
1234789 HpCDF
GC Column ID:
CS4 Data Filename:
CSS Data Filename:
M/Z'S ION ABUNDANCE RATIO
FORMING
RATIO (1) CS1 CS2 CS3 CS4
M/M+2
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M/M+2
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
QC
LIMITS
CSS (2)
0.65-0.89
1.32-1.78
1.05-1.^
1.05-l.M
1.05-1.40
0.88-1.20
0.76-1.02
0.65-0.89
1.32-1.78
1.32-1.78
1.05-1.43
1.05-1.43
1.05-1.43
1.05-1.43
0.88-1.20
0.88-1.20
OCDF
M+2/M+4
0.76-1.02
(1) See Table 3, Method 1613, for m/z specifications.
(2) Ion Abundance Ratio Control Limits from Table 3A, Method 1613.
91 r;i
>i. J. U
4/90
-------
USEPA - ITD
Lab Name:
FORM 3D
PCDD/PCDF INITIAL CALIBRATION ION ABUNDANCE RATIOS
Episode No.:
Contract No.:
SAS No.:
Initial Calibration Date:
Instrument ID:
GC Column ID:
CS1 Data Filename:
CS2 Data Filename:
CS3 Data Filename:
LABELED COMPOUNDS
13C-2378-TCDD
13C-12378-PeCDD
13C-123478-HXCDD
13C-123678-HXCDD
13C-1234678-HpCDD
13C-OCDD
13C-2378-TCDF
13C-12378-PeCDF
13C-23478-PeCDF
13C-123478-HXCDF
13C-123678-HXCDF
13C-123789-HXCDF
13C-234678-HXCDF
13C-1234678-HpCDF
13C-1234789-HpCDF
CS4 Data Filename:
CSS Data Filename:
M/Z'S
FORMING
RATIO (1) CS1
M/M+2
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M/M+2
M+2/M+4
M/M+2
M/M+2
M/M+2
M/M+2
M/M+2
M/M+2
ION ABUNDANCE RATIO QC
LIMITS
CS2 CS3 CS4 CSS (2)
0.65-0.89
1.32-1.78
1.05-1.43
1.05-1.43
0.88-1.20
0.76-1.02
0.65-0.89
1.32-1.78
1.32-1.78
0.43-0.59
0.43-0.59
0.43-0.59
0.43-0.59
0.37-0.51
0.37-0.51
(1) See Table 3, Method 1613, for m/z specifications.
(2) Ion Abundance Ratio Control Limits from Table 3A, Method 1613
216
4/90
-------
Lab Name:
USEPA - ITD '"•
FORM 4A
PCDD/PCDF CALIBRATION VERIFICATION
Episode No.
Contract No . :
SAS No.:
Initial Calibration Date:
Instrument ID:
VER Data Filename:
NATIVE ANALYTES
2378 TCDD
12378 PeCDD
123478 HxCDD
123678 HXCDD
123789 HXCDD
1234678 HpCDD
OCDD
2378 TCDF
12378 PeCDF
23478 PeCDF
123478 HXCDF
123678 HXCDF
123789 HXCDF
234678 HXCDF
1234678 HpCDF
1234789 HpCDF
OCDF
M/Z'S
FORMING
RATIO (1)
M/M+2
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M/M+2
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
GC Column ID:
Analysis Date:
ION QC
ABUND. LIMITS
RATIO (2)
0.65-0.89
1.32-1.78
1.05-1.43
1.05-1.43
1.05-1.43
0.88-1.20
0.76-1.02
0.65-0.89
1.32-1.78
1.32-1.78
1.05-1.43
1.05-1.43
1.05-1.43
1.05-1.43
0.88-1.20
0.88-1.20
0.76-1.02
Time:
CONC.
CONC. RANGE (3)
FOUND (ng/mL)
8.6-11.6
44.2-56.6
37.6-66.5
39.7-63.0
42.6-58.7
41.6-60.2
87.5-114.4
8.8-11.3
46.7-53.5
47.2-53.0
41.5-60.2
40.5-61.7
45.7-54.5
44.1-56.7
43.1-58.0
43.6-57.3
83.9-119.2
(1) See Table 3, Method 1613, for m/z specifications.
(2) Ion Abundance Ratio Control Limits as specified in Table 3A, Method 1613.
(3) Contract-required concentration range as specified in Table 7, Method
1613, under VER.
217
4/90
-------
Lab Name:
USEPA -MTD
FORM 4B
PCDD/PCDF CALIBRATION VERIFICATION
Episode No.:
Contract No.:
SAS No.:
Initial Calibration Date:
Instrument ID:
GC Column ID:
VER Data Filename:
Analysis Date:
Time:
LABELED COMPOUNDS
13C-2378-TCDD
13C-12378-PeCDD
13C-123478-HXCDD
13C-123678-HXCDD
13C-1234678-HpCDD
13C-OCDD
13C-2378-TCDF
13C-12378-PeCDF
130-23478-PeCDF
13C-123478-HXCDF
13C-123678-HXCDF
13C-123789-HXCDF
13C-234678-HXCDF
13C-1234678-HpCDF
13C-1234789-HpCDF
M/Z'S
FORMING
RATIO (1)
M/M+2
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M+2/M+4
M/M+2
M+2/M+4
M+2/M+4
M/M+2
M/M+2
M/M+2
M/M+2
M/M+2
M/M+2
ION QC
ABUND. LIMITS CONC.
RATIO (2) FOUND
0.65-0.89
1.32-1.78
1.05-1.43
1.05-1.43
0.88-1.20
0.76-1.02
0.65-0.89
1.32-1.78
1.32-1.78
0.43-0.59
0.43-0.59
0.43-0.59
0.43-0.59
0.37-0.51
0.37-0.51
CONC.
RANGE (3)
(ng/mL)
90.0-111.2
80.6-124.0
76.1-131.3
84.0-119.1
82.2-121.6
164.2-243.6
87.7-114.0
81.8-122.3
83.0-120.5
85.2-117.4
85.0-117.7
89.5-111.7
85.7-116.7
88.5-113.1
89.0-112.4
CLEANUP STANDARD
37C14-2378-TCDD (4)
24.4-46.4
(1) See Table 3, Method 1613, for m/z specifications.
(2) Ion Abundance Ratio Control Limits from-Table 3A, Method 1613.
(3) Contract-required concentration range, as specified in Table 7, Method
1613, under VER.
(4) No ion abundance ratio; report concentration found.
218
4/90
-------
Lab Name:
JJSBPA. --ITD
FORM 5
PCDD/FCDF ,ST WINDOW AND ISOMER SPECIFICITY STANDARDS
Episode No.:
Contract No.:
Instrument ID:
SAS No.
RT Window Data Filename:
DB-5 IS Data Filename:
DB-225 IS Data Filename:
Init. Calibration Date:
Analysis Date:
Analysis Date:
Analysis Date:
DB-5 RT WINDOW DEFINING STANDARDS RESULTS
ISOMERS
1368-TCDD (F)
1289-TCDD (L)
12479-PeCDD (F)
12389-PeCDD (L)
124679-HxCDD (F)
123467-HxCDD (L)
1234679-HpCDD (F)
1234678-HpCDD (L)
ABSOLUTE
RT
ISOMERS
1368-TCDF (F)
1289-TCDF (L)
13468-PeCDF (F)
12389-PeCDF (L)
123468-HXCDF (F)
123489-HXCDF (L)
1234678-HpCDF (F)
1234789-HpCDF (L)
Time:
Time:
Time:
ABSOLUTE
RT
(F) = First eluting isomer (DB-5); (L) = Last eluting isomer (DB-5).
ISOMER SPECIFICITY (IS) TEST STANDARDS RESULTS
ISOMERS
1234-TCDD
1278-TCDD
1278-TCDD
1478-TCDD
1478-TCDD
1237-TCDD
1237-TCDD
1238-TCDD
% VALLEY HEIGHT
BETWEEN
COMPARED PEAKS (1)
ISOMERS
1238-TCDD
2378-TCDD
2347-TCDF
2378-TCDF
2378-TCDF
1239-TCDF
% VALLEY HEIGHT
BETWEEN
COMPARED PEAKS (1)
(1) To meet contract requirements, % Valley Height Between Compared
Peaks shall not exceed 25% (Section 14.4.2.2, Method 1613).
4/90
21f)
-------
:-- USEPA - ITO
FORM 6A
PCDD/PCDF RELATIVE RETENTION TIMES
Lab Name:
Episode No.:
Contract No.:
Instrument ID:
Analysis Date:
SAS No.:
Init. Cal. Date:
GC Column ID:
Time:
CS3 or VER Data Filename:
Compounds Using 13C-1234-TCDD as Internal Standard
NATIVE ANALYTES
2378-TCDF
2378-TCDD
12378-PeCDF
23478-PeCDF
12378-PeCDD
LABELED COMPOUNDS
13C-2378-TCDF
13C-2378-TCDD
37C-2378-TCDD
13C-12378-PeCDF
13C-23478-PeCDF
13C-12378-PeCDD
RETENTION TIME
REFERENCE
13C-2378-TCDF
13C-2378-TCDD
13C-12378-PeCDF
13C-23478-PeCDF
13C-12378-PeCDD
13C-1234-TCDD
13C-1234-TCDD
13C-1234-TCDD
13C-1234-TCDD
13C-1234-TCDD
13C-1234-TCDD
RRT
RRT
QC LIMITS (1)
0.993-1.009
0.993-1.009
0.918-1.076
0.999-1.001
0.987-1.016
0.931-0.994
0.993-1.036
1.002-1.013
1.091-1.371
1.123-1.408
1.134-1.428
(1) Contract-required limits for Relative Retention Times (RRT) as specified
in Table 2, Method 1613.
4/90
-------
USEPA - TD
FORM 6B
PCDD/PCDF RELATIVE RETENTION TIMES
Lab Name:
Episode No.:
Contract No.:
Instrument ID:
Analysis Date:
SAS No.:
Init. Cal. Date:
GC Column ID:
Time:
.CS3 or VER Data Filename:
Compounds Using 13C-123789-HXCDD as Internal Standard
RETENTION TIME
NATIVE ANALYTES REFERENCE RRT
123478-
123678-
123789-
234678-
123478-
123678-
123789-
1234678
1234678
1234789
OCDD
OCDF
HxCDF
HxCDF
HxCDF
HxCDF
HxCDD
HxCDD
HxCDD
-HpCDF
-HpCDD
-HpCDF
LABELED COMPOUNDS
13C
13C
13C
13C
ISC-
ISC
13C-
13C-
13C-
13C-
-123478-
-123678-
-234678-
-123789-
-123478-
-123678-
-1234678
-1234678
-1234789
-OCDD
HxCDF
HxCDF
HxCDF
HxCDF
HxCDD
HxCDD
-HpCDF
-HpCDD
-HpCDF
13C-123478-HXCDF
13C-123678-HXCDF
13C-123789-HXCDF
13C-124678-HXCDF
13C-123478-HXCDD
13C-123678-HXCDD
13C-123678-HXCDD
13C-1234678-HpCDF
13C-1234678-HpCDD
13C-1234789-HpCDF
13C-OCDD
13C-OCDD
13C-123789-HXCDD
13C-123789-HXCDD
13C-123789-HXCDD
13C-123789-HXCDD
13C-123789-HXCDD
13C-123789-HXCDD
13C-123789-HXCDD
13C-123789-HXCDD
13C-123789-HXCDD
13C-123789-HXCDD
RRT
QC LIMITS (1)
0.986
0.973
0.937
999
999
992
0.986
0.930
0.986
0.896
0.996
-1.015
0.
0.
0.
,025
,068
,001
.001
,009
,016
,022
-1.016
-1.
,079"
-1.005
0.995-1.013
0.947-
0.940-
0.971-
0.993-
0.974-
0.975-
0.953-
1.023-
1.024-
1.050-
•0.992
•1.006
1.000
1.017
1.002
1.006
1.172
1.125
1.148
1.275
(1) Contract-required limits for Relative Retention Times (RRT) as specified
in Table 2, Method 1613.
4/90
-------
Lab Name:
USEPA - ITD
FORM 7A ,
PCDD/PCDF INITIAL PRECISION AND RECOVERY (IPR)
Episode No.:
Contract No.:
SAS No.:
Matrix (agueous/solid/leachate):
Ext. Date: Shift:
Repl Data Filename:
Rep2 Data Filename:
Rep3 Data Filename:
Rep4 Data Filename:
Analysis Date:
Analysis Date:
Analysis Date:
Analysis Date:
Time:
Time:
Time:
Time:
Cone. Units (pg/L or ng/Kg dry weight):
NATIVE ANALYTES
2378-TCDD
12378-PeCDD
123478-HxCDD
123678-HxCDD
123789-HxCDD
1234678-HpCDD
OCDD
2378-TCDF
12378-PeCDF
23478-PeCDF
123478-HXCDF
123678-HxCDF
123789-HxCDF
234678-HxCDF
1234678-HpCDF
1234789-HpCDF
OCDF
SPIKE
CONG.
REP 1
CONC.
FOUND
REP 2
CONC.
FOUND
REP 3
CONC.
FOUND
REP 4
CONC.
FOUND
X
(1)
s
(1)
(1) X = average concentration; s = standard deviation of the concentration.
Contract-required limits for X and s are specified in Table 7, Method 1613
4/90
'}
-------
USEPA'- ITD
FORM 7B
PCDD/PCDF INITIAL PRECISION , AND RECOVERY (IPR)
Lab Name: Episode No.:
Contract No.: SAS No.:
Matrix (agueous/solid/leachate):
Ext. Date: Shift:
Repl Data Filename: Analysis Date: Time:
Rep2 Data Filename: Analysis Date: Time:
Rep3 Data Filename: Analysis Date: Time:
Rep4 Data Filename: Analysis Date: Time:
Cone. Units (pg/L or ng/Kg dry weight):
REP 1 REP 2 REP 3 REP 4
SPIKE CONC. CONC. CONC. CONG. X
CONG. FOUND FOUND FOUND FOUND (1)
LABELED COMPOUNDS
13C-2378-TCDD
13C-12378-PeCDD
13C-123478-HXCDD
13C-123678-HXCDD
13C-1234678-HpCDD
13C-OCDD
13C-2378-TCDF
13C-12378-PeCDF
13C-23478-PeCDF
13C-123478-HXCDF
13C-123678-HXCDF
13C-123789-HXCDF
13C-234678-HXCDF
13C-1234678-HpCDF
13C-1234789-HpCDF
CLEANUP STANDARD
37C14-2378-TCDD
(1) X = average concentration. Contract-required concentration limits (X) fo
IPR are specified in Table 7, Method 1613.
4/90
223
-------
USEPA - ITD
Lab Name:
Contract No.:
, FORM 8 A
PCDD/PCDF ONGOING PRECISION AND RECOVERY (OPR)
Episode No.:
SAS No.:
Matrix (aqueous/solid/leachate):
Ext. Date: Shift:
PAR Data Filename:
Analysis Date: Time:
Cone. Units (pg/L or ng/Kg dry weight):
SPIKE
CONC.
CONC.
FOUND
NATIVE ANALYTES
2378-TCDD
12378-PeCDD
123478-HxCDD
123678-HXCDD
123789-HxCDD
1234678-HpCDD
OCDD
2378-TCDF
12378-PeCDF
23478-PeCDF
123478-HxCDF
123678-HxCDF
123789-HxCDF
234678-HxCDF
1234678-HpCDF
1234789-HpCDF
OCDF
OPR CONC.
LIMITS (1)
5.9
35.6
35.1
33.3
31.8
37.5
75.6
6.6
36.7
37.8
36.9
34.8
37.1
35.7
37.4
36.9
14.2
58.1
60.4
64.4
61.2
56.8
118.7
12.7
57.3
56.9
58.8
58.8
55.7
60.0
60.5
60.6
69.5 - 127.0
(1) Contract-required concentration limits for OPR as specified in Table 7,
Method 1613.
4/90
-------
USEPA - ITD
Lab Name:
FORM 8B
PCDD/PCDF ONGOING PRECISION AND RECOVER*
Episode No.:
Contract No.:
SAS No.:
Matrix (aqueous/solid/leachate):
Ext. Date: Shift:
PAR Data Filename:
Analysis Date: Time:
Cone. Units (pg/L or ng/Kg dry weight):
SPIKE
CONG.
CONC.
FOUND
LABELED COMPOUNDS
13C-2378-TCDD
13C-12378-PeCDD
13C-123478-HXCDD
13C-123678-HXCDD
13C-1234678-HpCDD
13C-OCDD
13C-2378-TCDF
13C-12378-PeCDF
13C-23478-P6CDF
13C-123478-HXCDF
13C-123678-HXCDF
13C-123789-HXCDF
13C-234678-HXCDF
13C-1234678-HpCDF
13C-1234789-HpCDF
CLEANUP STANDARD
37C14-2378-TCDD
OPR CONC.
LIMITS (1)
25.0
25.0
25.0
25.0
25.0
50.0
25.0
25.0
25.0
25.0
25.0
25.0
25.0
25.0
25.0
150.0
150.0
150.0
150.0
150.0
300.0
150.0
150.0
150.0-
150.0
150.0
150.0
150.0
150.0
150.0
10.0 - 60.0
(1) Contract-required concentration limits for OPR as specified in Table 7,
Method 1613. Labeled compound concentration limits are based on
required percent recovery of 25-150% (Section 14.5, Method 1613).
4/90
9 9
-------
SECTION 3
QUALITY CONTROL SOLUTIONS PROVIDED TO ANALYTICAL LABORATORIES
FOR THE EVALUATION OF EPA METHOD 1613
226
-------
Table 1
Calibration Solutions - concentrations in ng/mL
Native CDDs and CDFs CS1 CS2 CS3 CS4 CSS
2,3,7,8-TCDD 0.5 2 10 40 200
23,7,8-TCDF 0.5 2 10 . 40 200
1,2,3,7,8-PeCDD 2.5 10 50 200 1000
1,23,7,8-PeCDF 2.5 10 50 200 1000
23,4,7,8-PeCDF 2.5 10 50 200 1000
1,23,4,7,8-HxCDD 2.5 10 50 200 1000
1,23,6,7,8-HxCDD 2.5 10 50 200 1000
1,23,7,8,9-HxCDD 2.5 10 50 200 1000
1,2,3,4,7,8-HxCDF ' 2.5 10 50 200 1000
1,2,3,6,7,8-HxCDF 2.5 10 50 200 1000
1,23,7,8,9-HxCDF 2.5 10 50 200 1000
23,4,6,7,8-HxCDF 2.5 10 50 200 1000
1,23,4,6,7,8-HpCDD 2.5 10 50 200 1000
1,2,3,4,6,7,8-HpCDF 2.5 10 50 200 1000
U,3A7,8,9-HpCDF 2.5 10 50 200 1000
OCDD 5.0 20 100 400 2000
OCDF 5.0 20 100 400 2000
Labeled Compounds
13C12-23,7,8-TCDD 100 100 100 100 100
13C12-23,7,8-TCDF 100 100 100 100 100
13C12-l,2v3,7,8-PeCDD 100 100 100 100 100
13C12-l,23,7,8-PeCDF 100 100 100 100 100
13C12-23,4,7,8-PeCDF 100 100 100 100 100
13C12-l,2,3,4,7,8-HxCDD 100 100 100 100 100
13C12-l,23,6,7,8-HxCDD 100 100 100 100 100
13C12-l,2r3,4,7,8-HxCDF 100 100 100 100 100
13Cu-l,23,6,7,8-HxCDF 100 100 100 100 100
13C12-l,23,7,8,9-HxCDF 100 100 100 100 100
13C12-2,3,4,6,7,8-HxCDF 100 100 100 100 100
13C12-l,23,4,6,7,8-HpCDD 100 100 100 100 100
13C12-l,2,3,4,6,7,8-HpCDF 100 100 100 100 100
13C12-l,23,4,7,8,9-HpCDF 100 100 100 100 100
13C12-OCDD 200 200 200 200 200
37C14-23,7,8-TCDD 0.5 2 10 40 200
13C12-1,23,4-TCDD 100 100 100 100 100
13C12-l,23,7,8,9HxCDD 100 100 100 100 100
-------
Table 2
Labeled Compound Spiking Solution
Concentration
Labeled Compound (ng/mH
13C12-2,3,7,8-TCDD 100
13C12-2,3,7,8-TCDF 100
"C12-l,2,3,7,8-PeCDD 100
"C12-l,2,3,7,8-PeCDF 100
13c12-2,3,4,7,8-PeCDF 100
^C12-l,2,3,4,7,8-HxCDD 100
13C12-l,2,3,6,7,8-HxCDD 100
1?Cl2-l'2'3'4»7.8-HxCDF 100
^C12-l,2,3,6,7,8-HxCDF 100
"C12-l,2,3,7,8,9-HxCDF 100
^C12-2,3,4,6,7,8-HxCDF 100
"C12-1,2,3,4,6,7,8-HPCDD 100
"C12-l,2,3,4,6,7,8-HpCDF 100
"C^-lAW&o-HpCDF 100
13C12-OCDD 200
-------
Table3
Precision and Recovery Standard Solution
Compound
2,3,7,8-TCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDD
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,4,6,7,8-HpCDD
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDD
OCDF
Concentration Tng/ml)
40
40
200
200
200
200
200
200
200
200
200
200
200
200
200
400
400
9 9
-------
SECTION 4
PE SPIKE SOLUTIONS
FOR THE EVALUATION OF EPA METHOD 1613
NOTE: PE spike solutions will be included in the report after
laboratories have completed study analyses and reported
data.
230
-------
SUMMARY REPORT
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
INDUSTRIAL TECHNOLOGY DIVISION
METHOD DETECTION LIMIT STUDY
FOR METHOD 1613 DETERMINATION OF
2,3,7,8-TCDD AND 2,3,7,8-TCDF
May 1990
USEPA Office of Water
Office of Water Regulations and Standards
Industrial Technology Division (WH-552)
Washington, DC 20460
231
-------
TABLE OF CONTENTS
Method Detection Limit Study 1
Table 1 - Analytical Results 3
Table 2 - Standard Deviation and Method Detection Limits 4
Attachment 1- Appendix B to Part 136 5
Definition and Procedure for the Determination of the
Method Detection Limit - Revision 1.11
-------
METHOD DETECTION LIMIT (MDL) STUDY FOR
METHOD 1613 DETERMINATION OF 2,3,7,8-TCDD AND 2,3,7,8-TCDF
As part of the development of Method 1613 for the analysis of PCDDs and PCDFs, a method
detection limit (MDL) study was undertaken in December 1989. The basic design of this study was in
accordance with the procedure for determining MDLs specified in Appendix B of 40 CFR Part 136, as
published in the October 26, 1984 Federal Register (Attachment 1). The major requirements of this
procedure are:
o At least seven (7) aliquots of reagent water must be spiked with the analytes of interest
o Spike levels should be in the range of one to five times the laboratory's estimate of the detection
limit of each analyte
o Analyze all replicates and calculate a mean and standard deviation of the concentration of each
analyte
o Calculate the MDL as the standard deviation times the Students t value for (n-1) degrees of
freedom, where n is the number of replicates.
According to 40CFR 136, the MDL is defined as the minimum concentration of a substance that can be
measured and reported with 99% confidence that the analyte concentration is greater than zero and is
determined from analysis of a sample in a given matrix containing the analyte.
For the purposes of this study, the laboratory chose to analyze eight replicate samples instead of the
minimum of seven.
Estimated Detection Limits
As noted above, the Federal Register procedure requires that each replicate is spiked with a solution
containing target analytes at a concentration between one and five times the laboratory's estimated detection
limit. The specifications in 40 CFR 136 also list four ways in which to determine an estimate of the detection
limit of the method. The first of these options is to estimate the concentration value that corresponds to an
instrument signal-to-noise ratio of 2.5 to 5.0. Using this criteria, the laboratory calculated an estimated
detection limit of 25 ppq for 2,3,7,8-TCDD and 2,3,7,8-TCDF as follows.
o The lowest calibration solution (CC1) in Method 1613 has a 2,3,7,8-TCDD concentration of 0.5
ng/mL and presents a peak with a signal to noise ratio of at least 10.
o Each sample analyzed by Method 1613 has a final volume of 20 uL. By multiplying the final
volume of the sample times the concentration of TCDD in the CC1, it was determined that the
sample with a concentration equivalent to that of the lowest calibration standard would have a
final TCDD concentration of 10 pg/L or 10 ppq. This concentration is equal to the "Minimum
Level" described in Method 1613, and is derived in the same fashion.
o The laboratory equated the Minimum Level in Method 1613 with the American Chemical
Society's (ACS) concept of the Limit of Quantitation (LOQ). The ACS further defines a Limit
of Detection (LOD) as approximately one third of the LOQ. Thus, the laboratory set their
estimated detection limit as one third the Minimum Level, or 3.3 ppq. This figure represents the
estimated detection limit at 100% recovery, and should yield an instrumental signal of at least
2.5 times the background noise.
o Since Method 1613 specifies analyte quantitation by isotope dilution and allows data acceptance
when labeled compound recovery is as low as 25%, the estimated detection limit was adjusted by
the laboratory to account for the worst-case recovery. Thus 3.3 ppq was divided by 0.25, and a
worst case estimated detection limit was calculated as 13.2 ppq.
1
O '< f)
-------
Spike Levels
The Federal Register specifies that each of the replicates be spiked with each analyte to yield a
concentration between one and five times the estimated detection limit. The laboratory chose to use spike
solutions containing approximately twice the estimated detection limit of 13.2 ppq. Thus, the 2,3,7,8-TCDD
and 2,3,7,8-TCDF isomers were both spiked at 25 ppq.
Analytical Results and Calculation of Variance, Standard Deviation and Method Detection Limit
Table 1 provides the analytical results for 2,3,7,8-TCDD and 2,3,7,8-TCDF in each of the eight
replicates. The Federal Register requires that the Method Detection Limit for each analyte be calculated as
the standard deviation of the replicate analyses multiplied by the Students t value for (n-1) degrees of
freedom, where n equals the number of replicates. Table 2 provides the mean concentrations of both
isomers from the replicate analyses along with standard deviations and the calculated MDLs. The calculated
MDL for 2,3,7,8-TCDD is 5.6 ppq and the calculated MDL for 2,3,7,8-TCDF is 1.7 ppq.
The MDL values of 5.6 ppq and 1.7 ppq for TCDD and TCDF in Table 2 are below the "minimum
levels" of 10 ppq listed in Method 1613 for these isomers. As indicated above, the minimum levels
correspond to the concentration in a sample equivalent to the concentration of each analyte in the lowest of
the calibration standards, assuming 100% recovery of the labeled compounds added to the sample and used
for quantification by isotope dilution. Thus, the MDL data presented here indicate that the method is
capable of measuring 2,3,7,8-TCDD and 2,3,7,8-TCDFs at levels at least as low as the minimum levels
specified in the method description.
234
2
-------
Table 1
Analytical Results
ISOMER
2378-TCDD
12378-PeCDD
123478-HxCDD
123678-HxCDD
123789-HxCDD
1234678-HpCDD
OCDD
2378-TCDF
12378PeCDF
23478-PeCDF
123478-HxCDF
123678-HxCDF
234678-HxCDF
123789-HxCDF
1234678-HpCDF
1234789-HpCDF
OCDF
SMX1R
33.1
24.2
33.8
59.6
30.2
802
8250
33.3
31.4
31.2
48.9
37.9
43.3
41.8
174
88.6
833
SMX2R
32.6
26.2
33.7
28.3
29.8
51.2
153
34.3
30.7
33.4
44.6
33.9
35.1
39.0
76.3
84.1
70.4
SMX3R
30.9
22.7
36.0
27.9
21.2
53.0
169
33.3
28.9
31.4
44.6
35.3
34.5
34.9
75.8
83.9
80.3
SMX4R
30.3
22.4
33.2
29.0
28.3
109
967
32.4
28.8
31.5
40.5
33.8
34.2
39.5
82.0
73.2
123
SMX5R
32.0
24.1
32.4
29.4
28.5
51.6
161
33.3
30.3
31.4
42.8
34.0
34.6
38.1
75.8
82.2
87.1
SMX6R
27.1
23.1
33.7
28.6
19.6
82.0
581
32.8
30.1
32.0
42.8
30.9
34.6
35.3
76.6
78.5
136
SMX7R
31.6
22.5
28.7
26.2
23.1
53.1
201
33.1
29.9
31.0
41.5
31.0
32.9
36.7
72.7
82.1
58.5
SMX8R
31.7
23.1
34.1
33.8
30.1
208
2110
33.6
30.6
29.3
43.3
34.1
36.1
37.6
101
84.7
249
235
-------
Table 2
Standard Deviation and Method Detection Limits
Isomer
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
1,23,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDD
OCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
23,4,6,7,8-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDD
No. of
Observations
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
Mean
31.2
23.5
33.2
32.8
26.3
176.2
1574
33.3
30.1
31.4
43.6
33.9
37.9
35.7
91.8
82.2
204.7
Standard
Deviation
1.9
1.3
2.1
11.0
4.3
258.4
2780
0.6
0.9
1.1
2.5
2.2
2.3
3.2
34.4
4.6
261.0
Student's
r-Value
2.998
2.998
2.998
2.998
2.998
2.998
2.998
2.998
2.998
2.998
2.998
2.998
2.998
2.998
2.998
2.998
2.998
Method
Detection Limit
5.6
3.8
6.2
33.0
13.0
774.8
8333
1.7
2.7
3.4
7.6
6.7
6.8
9.6
103.1
13.8
782.5
236
-------
ATTACHMENT 1
APPENDIX B TO PART 136
DEFINITION AND PROCEDURE FOR THE DETERMINATION OF THE METHOD DETECTION LIMIT
REVISION 1.11
(EXCERPTED FROM 40 CFR 136, OCTOBER 26,1984)
-------
198 ~ Federal Register /'VoL 49. No. 209 / Friday.-October 26. 1984 7 Roles and Regulations
Appendix B to Part 138—Definition utd
Procedure for the Determination of the
Method Detection Unit—Revision Ul
Definition
The method detection limit (MDL) U
defined a* the minimum concentration of a
substance that can be measured and reported
with 99% confidence that the analyte
concentration is greater than zero and U
determined from analysis of a sample in a
given matrix containing the analyte.
Scope and Application
This procedure is designed for applicability
to a wide variety of sample types ranging
from reagent (blank) water containing
analyte to wastewater containing analyte.
The MDL for an analytical procedure may
vary as a function of sample type. The
procedure requires a complete, specific, and
well defined analytical method. It is essential
that all sample processing steps of the
analytical method be included in the
determination of the method detection limit.
The MDL obtained by this procedure is
used to judge the significance of a single
measurement of a future sample.
The MDL procedure was designed for
applicability to a broad variety of physical
and chemical methods. To accomplish this.
the procedure was made device- or
instrument-independent.
Procedure
1. Make an estimate of the detection limit
using one of the following:
(a) The concentration value that
corresponds to an instrument signal/noise in
the range of 2.5 to 5.
(b) The concentration equivalent of three
times the standard deviation of replicate
instrumental measurements of the analyte in
reagent water.
(c) That region of the standard curve where
there is a significant change in sensitivity.
i.e.. a break in the slope of the standard
curve.
(d) Instrumental limitations.
It is recognized that the experience of the
analyst is important to this process.
However, the analyst must include the above
considerations in the initial estimate of the
detection limit.
2. Prepare reagent (blank) water that is as
free of analyte as possible. Reagent or
interference free water is defined as a water
sample in which analyte and interferent
concentrations are not detected at the
method detection limit of each analyte of
interest Interferences are defined as
systematic errors in the measured analytical
signal of an established procedure caused by
the presence of interfering species
(interferentL The interferent concentration is
presupposed to be normally distributed in
representative samples of a given matrix.
3. (a) If the MDL is to be determined in
reagent (blank) water, prepare a laboratory
standard (analyte in reagent water} at a
concentration which is at least equal to or in
the same concentration range as the
estimated method detection limit
(Recommend between 1 and 5 times the
estimated method detection limit) Proceed to
Step 4.
(b) If the MDL is to be determined in
another sample matrix analyze the •ample. If
the measured tare) of the analyte is in the
recommended range of one to fiVe time* the
estimated detection Hnrit proceed to Step 4.
If the measured level of analyte isles* Utan
the estimated detection limit add a known
amount of analyte to bring the lerd of
analyte between one and five times the
estimated detection limit
If the measured level of analyte is greater
than five times the estimated detection limit
there are two options.
(1) Obtain another sample with a lower
level of analyte in the same matrix if
possible.
(2) The sample may be used as is for
determining the method detection limit if the
analyte level does not exceed 10 times the
MDL of the analyte in reagent water. The
variance of the analytical method changes as
the analyte concentration increases from the
MDL hence the MDL determined under these
circumstances may not truly reflect method
variance at lower analyte concentrations.
4. (a) Take a minimum of seven aliquots of
the sample to be used to calculate the method
detection limit and process each through the
entire analytical method. Make all
computations according to the defined
method with final results in the method
reporting units. If a blank measurement is
required to calculate the measured level of
analyte. obtain a separate blank
measurement for each sample aliquot
analyzed. The average blank measurement is
subtracted from the respective sample
measurements.
(b) It may be economically and technically
desirable to evaluate the estimated method
detection limit before proceeding with 4a.
This will: (1) Prevent repeating this entire
procedure when the costs of analyses are
high and (2) insure that the procedure is being
conducted at the correct concentration. It is
quite possible that an inflated MDL will be
calculated from data obtained at many times
the real MDL even though the level of analyte
is less than five times the calculated method
detection limit. To insure that the estimate of
the method detection limit is a good estimate.
it is necessary to determine that a lower
concentration of analyte will not result in a
significantly lower method detection limit.
Take two aliquots of the sample to be used to
calculate the method detection limit and
process each through the entire method.
including blank measurements as described
above in 4a. Evaluate these data:
(1) If these measurements indicate the
sample is in desirable range for
determination of the MDL. take five
additional aliquots and proceed. Use all
seven measurements for calculation of the
MDL
(2) If these measurements indicate the
sample is not in correct range, reestimate the
MDL obtain new sample as in 3 and repeat
either 4a or 4b.
5. Calculate the variance (S*) and standard
deviation (S) of the replicate measurements.
as follows:
where:
X,: i=l to n. = are the analytical results in
the final method reporting units obtained
from the n sample aliquots and I refers
to the sum of the X values from i = l to n.
6. (a) Compute the MDL as follows:
MDL =
Ual.l-* • O.«t) (S)
where:
MDL
the method detection limit
U.-i.].. . .«») = the students' I value
appropriate for a 99% confidence level
and a standard deviation estimate with
n-1 degrees of freedom. See Table.
S = standard deviation of the replicate
analyses.
(b) The 95% confidence interval estimates
for the MDL derived in 6a are computed
according to the following equations derived
from percentiles of the chi square over
degrees of freedom distribution (,'/df)-
LCL = 0.64 MDL
UCL = 2.20 MDL
where: LCL and UCL are the lower and
upper 95% confidence limits respectively
based on seven aliquots.
7. Optional iterative procedure to verify the
reasonableness of the estimate of the MDL
and subsequent MDL determinations.
(a) If this is the initial attempt to compute
MDL based on the estimate of MDL
formulated in Step 1. take the MDL as
calculated in Step 6, spike in the matrix at the
calculated MDL and proceed through the
procedure starting with Step 4.
(b] If this is the second or later iteration of
the MDL calculation, use S' from the current
MDL calculation and S1 from the previous
MDL calculation to compute the F-ratio. The
F-ran'o is calculated by substituting the larger
S'into the numerator SA- and the others into
the denominator SV The computed F-ratio is
then compared with the F-ratio found in the
table which is 3.05 as follows: if S'A/
S*B<3.05. then compute the pooled standard
deviation by the following equation:
-I-T-1)
if S'A/S1.>3.05. respike at the most recent
calculated MDL and process the samples
through the procedure starting with Step 4. If
the most recent calculated MDL does not
permit qualitative identification when
samples are spiked at that level report the
MDL as a concentration between the current
and previous MDL which permits qualitative
identification.
(c) Use the SMM as calculated in 7b to
compute the final MDL according to the
following equation:
238
-------
Federal Register / Vol 49.'No. 209 / Friday. October 26, 1984 / Rules and Regulations
199
MDL
where 2381 i< equal to lot. .-. = .99).
(d) The 95* confidence limits for MDL
derived in 7c are computed according to the
following equations derived from precentiles
of the chi squared over degrees of freedom
distribution.
LO.=072MDL
UCL-1.65 MDL
where LCL and UCL are the lower and upper
95% confidence limits respectively baaed on
14 aliquots.
TABLES Of STUDENTS' t VALUES AT THE 99
PERCENT CONFIDENCE LEVEL
NUR*MT of rapiotts
7
B , . ..
o
1ft
11 ,, ,
Iff <
9*
?tt ......
}T
fft
Oft ,, , ,
-
"•ST
frwXxn
(0.1)
6
7
8
9
10
10
20
a
30
60
00
1... -.
3.143
2.998
2.896
2.821
1764
2.6Q2
2.528
2.485
2.457
2-390
2.326
Reporting
The analytical method used must be
specifically identified by number or title and
the MDL for each analyte expressed in the
appropriate method reporting units. If the
analytical method permits options which
affect the method detection limit, these
conditions must be specified with the MDL
value. The sample matrix used to determine
.the MDL must also be identified with MDL
fcralue. Report the mean analyte level with the
IMDL and indicate if the MDL procedure was
iterated. If a laboratory standard or a sample
that contained a known amount analyte was
used for this determination, also report the
mean recovery.
If the level of analyte in the sample was
below the determined MDL or does not
exceed 10 times the MDL of the analyte in
reagent water, do not report a value for the
MDL.
Appendix C to Part 136—Inductively
Coupled Plasma—Atomic Emission
Spectrometric Method for Trace Element
Analysis of Water and Wastes Method
200.7
1. Scope and Application
1.1 This method may be used for the
determination of dissolved, suspended, or
total elements in drinking water, surface
water, and domestic and industrial
wastewaters.
1.2 Dissolved elements are determined in
filtered and acidified samples. Appropriate
steps must be taken in all analyse* to ensure
that potential interferences are taken into
account This is especially true when
dissolved solids exceed 1500 mg/L, (See
j jOI>Total elements are determined after
appropriate digestion procedure* are
oerformed. Since digestion technique*
Se dissolved solid, content of the .
sample*, appropriate step* most be taken to
correct for potential interference effect*. (See
sections.)
1.4 Table 1 KsU elements for which this
method applies along with recommended
wavelengths and typical estimated
instrumental detection limits using
conventional pneumatic nebulization. Actual
working detection limits are sample
dependent and as the sample matrix varies.
these concentrations may also vary. In time,
other elements may be added as more
information becomes available and as
required-
1.5 Because of the differences between
various makes and models of satisfactory
instruments, no detailed instrumental
operating instructions can be provided.
Instead, the analyst is referred to the
instruction provided by the manufacturer of
the particular instrument.
2. Summary of Method
2.1 The method describes a technique for
the simultaneous or sequential multielement
determination of trace elements in solution.
The basis of the method is die measurement
of atomic emission by an optical
spectroscopic technique. Samples are
nebulized and the aerosol that is produced is
transported to the plasma torch where
excitation occurs. Characteristic atomic-line
emission spectra are produced by a radio-
frequency inductively coupled plasma (1CP).
The spectra are dispersed by a grating
spectrometer and the intensities of the lines
are monitored by pbotomoltiplier tubes. The
photocurrents from the photomultiplier tubes
are processed and controlled by a computer
system. A background correction technique is
required to compensate for variable
background contribution to the determination
of trace elements. Background must be
measured adjacent to analyte lines on
samples during analysis. The position
selected for the background intensity
measurement, on either or both sides of the
analytical line, will be determined by the
complexity of the spectrum adjacent to the
analyte line. The position used must be free
of spectral interference and reflect the same
change in background intensity as occurs at
the analyte wavelength measured.
Background correction is not required in
cases of line broadening where a background
correction measurement would actually
degrade the analytical result The possibility
of additional interferences named in 5.1 (and
tests for their presence as described in 5-2)
should also be recognized and appropriate
corrections made.
3. Definitions
3.1 Dissolved—Those elements which
will pass through a 0.45 fim membrane filter.
3.2 Suspended—Those elements which
are retained by a 0.45 /im membrane filter.
3.3 Total—The concentration determined
on an unfiltered sample following vigorous
digestion (Section 93). or the sum of the
dissolved plus suspended concentrations.
(Section 11 phis 9-2L
3.4 Total recoverable—The coccentntion
determined on an unfiltered •ampb following
treatment with hot. dilute mineral acid
(Section 9.4).
15 Instrumental detection limit—The
concentration equivalent to a signal, due to
the analyte. which is equal to three times the
standard deviation of a series often replicate
measurements of a reagent blank signal at
the same wavelength.
M Sensitivity—The slope of the
analytical curve. i.e- functional relationship
between emission intensity and
concentration.
3J Instrument check standard—A
multielement standard of known
concentration* prepared by the analyst to
monitor and verify instrument performance
on a daily basis. (See 7.6.1)
3JJ Interference check sample—A
solution containing both interfering and
analyte elements of known concentration
that can be used to verify background and
interelement correction factors. (See 7.6L2.)
3.9 Quality control sample—A solution
obtained from an outside source having
known, concentration values to be used to
verify the calibration standards. (See 7.6J)
3.10 Calibration standards—A series of
known standard solutions used by the
analyst for calibration of the instrument (i.e_
preparation of the analytical curve). (See 7.4)
3.11 Linear dynamic range—The
concentration range over which the
analytical carve remains linear.
3.12 Reagent blank—A volume of
deionized. distilled water containing the
same acid matrix as the calibration standards
carried through the entire analytical scheme.
(See 7 52)
3.13 Calibration blank—A volume of
deionized. distilled water acidified with
HNCs and Ha (See 7.5.1)
3.14 Method of standard addition—The
standard addition technique involves the use
of the unknown and the unknown plus a
known amount of standard. (See 10.8.1.)
* Safety
4.1 The toxicity of carcinogeniciry of each
reagent used in this method has not been
precisely defined; however, each chemical
compound should be treated as a potential
health hazard. From this viewpoint, exposure
to these chemicals must be reduced to the
lowest possible level by whatever means
available. The laboratory is repsonsible for
maintaining a current awareness file of
OSHA regulations regarding the safe
handling of the chemicals specified in this
method. A reference file of material data
handling sheets should also be made
available to all personnel involved in the
chemical analysis. Additional references to
laboratory safety are available and have
been identified <'** "••-* "•» for the
information of the analyst
^Interferences
5.1 Several types of interference effects
may contribute to inaccuracies in the
determination of trace elements. They can be
summarized a* follows:
5.1.1 Spectral interferences can be
categorized a* (1) overlap of a spectral line
from another element; (2) unresolved overlap
of molecular band spectra; (3) background
cuutriuutioA If out continuous or
recombination phenomena; and (4)
background contribution from stray light from
23H
-------
! DRAFT
Master 2-2
24 May 1987
METHOD 8290
ANALYTICAL PROCEDURES AND QUALITY ASSURANCE
FOR MULTIMEDIA ANALYSIS
OF
POLYCHLORINATED DIBKNZO-p~DIOXINS
AND
POLYCHLORINATED DIBENZOFURANS
BY
HIGH-RESOLUTION GAS CHROMATOGRAPHY/HIGH-RESOLUTION MASS
SPECTROMETRY
(Exhibits D and E)
by
Yves Tondeur
June 1987
Project Officer
Werner F. Beckert
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada 89193-3478
240
-------
Draft
': 24 May 1987
DP^ p. r*^T> ;
rtAFl ;
NOTICE
This document is a preliminary draft. It has not been formally released
by the University of Nevada Environmental Research Center or the U.S. Environ-
mental Protection Agency, and it should not at this stage be construed to
represent University or Agency policy. It is circulated for comments on its
technical merit and policy implications.
241
-------
Draft
24 May 1987
DRAFT
FOREWORD
In January 1986, the Environmental Protection Agency published an analy-
tical protocol, Protocol for the Analysis of 2,3,7,8-Tetrachlorodibenzo-p-
Dioxin (TCDD) by High-Resolution Gas Chroraatography/High-Resolution Mass
Spectroraetry (HRGC/HRMS) (EPA 600/4-86-004), aimed at the determination of part-
per-trillion and sub-part-per-trillion levels of 2,3,7,8-TCDD and of total TCDD
in soil, sediment and aqueous samples. The January 1986 document was intended
to be a stepping stone for the realization of a more comprehensive method that
would include all the polychlorinated dibenzodioxin (PCDD) and polychlorinated
dibenzofuran (PCDF) congeners present in a broader spectrum of environmentally
significant matrices.
The present report constitutes a draft addressing the analytical proce-
dures (Exhibit D) and quality assurance (Exhibit E, quality assessment and
control) requirements sections of the future analytical protocol for the
analysis of PCDDs and PCDFs by HRGC/HRMS; i.e., Method 8290. At times, refer-
ence to other exhibits (e.g., Exhibit C) are made, even though these sections
have not been prepared. The format used for this report is similar to the
format used for other EPA TCDD protocols. Figures and tables are, however,
grouped at the end of Exhibit D. A final version of Method 8290 is expected
following peer review of this draft report and the completion of the single-
laboratory evaluation. Elements included in this Method 8290 have been taken
from a variety of sources, such as the EPA Region VII low-resolution mass
spectrometry (LRMS) TCDD protocol, the aforementioned high-resolution mass
ii
942
'- - ^C 4-M4
-------
spectrometry TCDD protocol',
DRAFF
Draft
24 May 1987
Method 8280
-------
Draft
24 May 1987
TABLE OF CONTENTS
Foreword ii
Abbreviations and Symbols v
Analytical Methods (Exhibit D)
1. Scope and Application D-l
2. Summary of the Method D-2
3. Definitions D-5
4. Interferences D-10
5. Safety D-ll
6. Apparatus and Equipment D-16
7. Reagents and Standard Solutions D-23
8. System Performance Criteria D-27
9. Calibration D-32
10. Quality Assessment/Quality Control Procedures D-40
11. Sample Preservation D-41
12. Extraction and Cleanup Procedures D-44
13. Analytical Procedures D-59
14. Calculations D-63
APPENDIX A: PROCEDURE FOR THE COLLECTION, HANDLING, ANALYSIS, AND
REPORTING REQUIREMENTS OF WIPE TESTS PERFORMED WITHIN THE
LABORATORY D-71
APPENDIX B: STANDARDS TRACEABILITY PROCEDURE D-75
APPENDIX C: SIGNAL-TO-NOISE DETERMINATION METHOD D-81
Figures D-84
Tables D-93
Quality Assurance Requirements (Exhibit E)
1. Summary of QA/QC Analyses E-l
2. Quality Assessment/Quality Control E-2
3. Laboratory Evaluation Procedures E-12
iv
244
-------
DRAFT
Draft
24 May 1987
\i
LIST OF ABBREVIATIONS AND SYMBOLS
A — Integrated Ion abundance
ADC — Analogue-to-digltal conversion
AX-21 — Type of carbon adsorbent
C — Concentration
CDC — Center for Disease Control
CDWG — Chlorinated Dioxins Workgroup
0 C — Degree centigrade
13C — Carbon-13 labeled
cm — Centimeter
DB-5 — Type of fused-silica capillary column
DS — Data system
EDL — Estimated detection limit
EMPC — Estimated maximum possible concentration
EMSL-LV -- Environmental Monitoring System Laboratory, Las Vegas
EPA — Environmental Protection Agency
g — Gram
GC — Gas chromatography or gas chromatograph
GC/MS — Gas chromatography/mass spectrometry
HEPA — High-efficiency partlculate absorbant
HpCDD — Heptachlorodibenzodioxin
HpCDF — Heptachlorodlbenzofuran
HRGC/HRMS — High-resolution gas chromatography/high-resolution
mass spectrometry
HxCDD — Hexachlorodibenzodioxin
HxCDF — Hexachlorodibenzofuran
IFB — Invitation for Bid
IS — Internal Standard
KD — Kuderna-Danish
L — Liter
MB — Method blank
MCL — Method calibration limit
mL — Milliliter
mm — Millimeter
M/AM — Mass spectrometer resolving power
MS — Matrix spike
MSD — matrix spike duplicate
OCDD — Octachlorodibenzodioxin
OCDF — Octachlorodlbenzofuran
OSHA — Occupational Safety and Health Administration
PCB — Polychlorinated blphenyl
PCDD — Polychlorlnated dibenzodioxln
PCDPE — Polychlorinated diphenyl ether
PCDF — Polychlorlnated dlbenzofuran
PE — Performance evaluation
O A C
•c --i O
-------
Draft
24 May 1987
PEM
PeCDD
PeCDF
PFK
PS
ppm
ppt
Q
QA
QA/QC
rpm
RPD
RRF
RRF
RRT
RS
S
SAS
SES
SICP
SIM
SMO
S/N
SOP
SP-2330
Still-
bottom
TCDD
TEF
V
v/v
W
WTE
uL
Performance evaluation material
Pentachlorodlbenzodloxln
Pentachlorodibenzofuran
Pe rfluoroke rosene
Plcogram
Part per million
Part per trillion
Amount of substance
Quality Assurance or Quality Assessment
Quality Assessment/Quality Control
Revolutions per minute
Relative percent difference
Relative response factor
Mean relative response factor
Relative retention time
Recovery standard
EPA reference standard solution
Special Analytical Service
Site evaluation sheet
Selected ion current profile
Selected ion monitoring
Sample Management Office
Signal-to-noise ratio
Standard Operating Procedure
Type of fused-sllica capillary column
Name of a matrix that is used as a noun
Tetrachlorodibenzodloxin
Toxicity Equivalency Factor
Volume
Volume/volume
Weight or laboratory working standard
Wipe test experiment
Microllter
vi
248
-------
DRAFT
Draft
24 May 1987
ANALYTICAL METHODS
(EXHIBIT D)
247
-------
Draft
24 May 1987
EXHIBIT D
1. Scope and Application
1.1 This method provides procedures for the detection and quantitative measure-
ment of 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD), polychlorinated
dibenzo-p-dioxins (tetra- through octachlorinated homologues; PCDDs), and
polychlorinated dibenzofurans (tetra- through octachlorinated homologues;
PCDFs) in a variety of environmental matrices and at part-per-trillion
(ppt) concentrations. The analytical method calls for the use of high-
resolution gas chromatography and high-resolution mass spectrometry (HRGC/
HRMS) on purified sample extracts. Table 1 lists the various sample types
covered by this analytical protocol, the 2,3,7,8-TCDD-based method calibra-
tion limits (MCLs) and other germane information. Analysis of a one-tenth
aliquot of the sample permits measurement of concentrations up to 10 times
the upper MCL (Table 1). Samples containing concentrations of specific
congeneric analytes (PCDDs and PCDFs) considered within the scope of this
method that are greater than the upper MCL must be analyzed by a protocol
designed for such concentration levels. An optional method for reporting
the analytical results using a 2,3,7,8-TCDD toxicity equivalency factor
(TEF) is described.
D-l
248
-------
1.2 The sensitivity of this method is dependent upon the level of interferences
within a given matrix. Actual limits of detection and quantification will
be provided based on the single- or multi-laboratory evaluation of this
protocol, and on examining the data gathered by the Sample Management
Office (SMO) from Special Analytical Services (SAS) performed over the
past few years.
1.3 This method is designed for use by analysts who are experienced with
residue analysis and skilled in high-resolution gas chromatography/high-
resolution mass spectrometry (HRGC/HRMS).
1.4 Because of the extreme toxicity of many of these compounds, the analyst
must take the necessary precautions to prevent exposure to materials known
or believed to contain PCDDs or PCDFs. It is the responsibility of the
laboratory personnel to ensure that safe handling procedures are employed.
2. Summary of the Method
2.1 This procedure uses matrix-specific extraction, analyte-specific cleanup,
and high-resolution capillary column gas chroraatography/high-resolution
mass spectrometry (HRGC/HRMS) techniques.
2.2 If interferences are encountered, the method provides selected cleanup
procedures to aid the analyst in their elimination. A simplified
analysis flow chart is shown in Figure 1.
D-2
2 4 fl
-------
2.3 A specified amount (see Table 1) of soil, sediment, fly ash, water,
sludge (Including paper pulp), still-bottom, fuel oil, chemical reactor
residue, fish tissue, or human adipose tissue Is spiked with a solution
containing specified amounts of each of the nine Isotopically (^C^)
labeled PCDDs/PCDFs listed In Column 1 of Table 2. The sample Is then
o
extracted according to a matrix-specific extraction procedure. The extrac-
tion procedures are: a) toluene (or benzene) Soxhlet extraction for soil,
sediment and fly ash samples; b) methylene chloride liquid-liquid extrac-
tion for water samples; c) toluene (or benzene) Dean-Stark extraction for
fuel oils and aqueous sludges; d) toluene (or benzene) extraction for
still-bottoms; e) hexane/methylene chloride Soxhlet extraction for fish
tissue and paper pulp; and f) methylene chloride extraction for human
adipose tissue. The decision for the selection of an extraction procedure
for chemical reactor residue samples is based on the appearance (consistency,
viscosity) of the samples. Generally, they can be handled according to
the procedure used for still-bottom (or chemical sludge) samples.
2.4 The extracts are submitted to an.acid-base washing treatment and dried.
Following a solvent exchange step, the residue Is cleaned up by column
chromatography on neutral alumina and carbon on Celite 545®. The extract
from adipose tissue is treated with silica gel impregnated with sulfuric
acid before chromatography on acidic silica gel, neutral alumina, and
carbon on Celite 545®. Fish tissue and paper pulp are subjected to an
acid wash treatment only prior to chromatography or neutral alumina and
• carbon/Celite. The preparation of the final extract for HRGC/HRMS
analysis is accomplished by adding, to the concentrated carbon column
D-3
25
-------
eluate, 10 to 50 uL uL (depending on the matrix type) of a tridecane
solution containing 50 pg/uL of each of the two recovery standards
13C12-1,2,3,4-TCDD and 13C12-1,2,3,7,8,9-HxCDD (Table 2). The former is
used to determine the percent recoveries of tetra- and pentachlorinated
PCDD/PCDF congeners while the latter is used for the determination of
L.
hexa-, hepta- and octa-chlorinated PCDD/PCDF congeners percent recoveries.
2.5 One to two uL of the concentrated extract are injected into an HRGC/HRMS
system capable of performing selected ion monitoring at resolving powers
of at least 10,000 (10 percent valley definition).
2.6 The identification of OCDD and nine of the fifteen 2,3,7,8-substituted
congeners (Table 3), for which a *3C-labeled standard is available in the
sample fortification and recovery standard solutions (Table 2), is based
on their elution at their exact retention time (-1 to -4-3 seconds from the
respective internal or recovery standard signal) and the simultaneous
detection of the two most abundant ions in the molecular ion region. The
remaining six 2,3,7,8-substituted congeners (i.e., 2,3,4,7,8-PeCDF;
1,2,3,4,7,8-HxCDD; 1,2,3,6,7,8-HxCDF; 1,2,3,7,8,9-HxCDF; 2,3,4,6,7,8-HxCDF,
and 1,2,3,4,7,8,9-HpCDF), for which no carbon-labeled internal standards
are available in the sample fortification solution, and all other identified
PCDD/PCDF congeners are identified by their relative retention times
falling within their respective PCDD/PCDF retention time windows, as estab-
lished by using a GC column performance evaluation solution, and the
simultaneous detection of the two most abundant ions in the molecular
ion region. The identification of OCDF is based on its retention time
D-4
251
-------
1 ^
relative to Cj2~OCDD and the simultaneous detection of the two most
abundant ions in the molecular Ion region. Confirmation is based on a
comparison of the ratio of the Integrated Ion abundance of the molecular
Ion species to their theoretical abundance ratio.
2.7 Quantification of the individual congeners, total PCDDs and total PCDFs is
achieved in conjunction with the establishment of a multipoint (seven
points) calibration curve for each homologue, during which each cali-
bration solution is analyzed once.
3. Definitions
3.1 Polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans
(PCDFs): Compounds (Figure 2) that contain from one to eight chlorine
atoms. The fifteen 2,3,7,8-substltuted PCDDs (totaling 75) and PCDFs
(totaling 135) are shown in Table 3. The number of isomers at different
chlorination levels is shown in Table 4.
3.2 Homologous series: Defined as a group of chlorinated dibenzodioxins or
dibenzofurans having a specific number of chlorine atoms.
3.3 Isomer: Defined by the arrangement of chlorine atoms within an
homologous series. For example, 2,3,7,8-TCDD is a TCDD isomer.
3.4 Congener: Any isomer of any homologous series.
D-5
O !T
-------
3.5 Internal Standard: An Internal standard is a c^-labeled analogue of a
congener chosen from the compounds listed in Table 3 and of OCDD. Internal
standards are added to all samples including method blanks and quality con-
trol samples before extraction, and they are used to measure the concentra-
tion of the analytes. Nine internal standards are used in this method.
There is one for each of the dioxin and furan homologues (except for OCDF)
with the degree of chlorination ranging from four to eight.
3.6 Recovery Standard: Recovery standards (two) are used to determine the
percent recoveries for PCDDs and PCDFs. The 13Cj2-l,2,3,4-TCDD is used to
measure the percent recoveries of the tetra- and pentachlorinated dioxins
and furans while ^^Cj2~l»2,3,7,8,9-HxCDD permits the recovery determination
of the hexa-, hepta- and octachlorinated homologues. They are added to
the final sample extract before HRGC/HRMS analysis. Furthermore, ^^C-.^'
1,2,3,7,8,9-HxCDD is used for the identification of the unlabeled analogue
present in sample extracts (this exhibit, Section 2.6).
3.7 High-Resolution Concentration Calibration Solutions (Table 5): Solutions
(tridecane) containing known amounts of 17 selected PCDDs and PCDFs, nine
internal standards ( C^'labeled PCDDs/PCDFs), and two carbon-labeled
recovery standards (this exhibit, Section 3.6); the set of seven solutions
is used to determine the instrument response of the unlabeled analytes
relative to the internal standards and of the internal standards relative
to the recovery standards.
D-6
253
-------
3.8 Sample Fortificatior
DRAFT
solution (isooctane) containing
the nine internal standards, which is used to spike all samples before
extraction and cleanup.
3.9 Recovery Standard Solution (Table 2): A tridecane solution containing the
two recovery standards, which is added to the final sample extract before
HRGC/HRMS analysis.
3.10 Field Blank: A portion of a sample representative of the matrix under
consideration, which is free of any PCDDs/PCDFs.
3.11 Laboratory Method Blank: A blank prepared in the laboratory and carried
through all analytical procedure steps except the addition of a sample
aliquot to the extraction vessel.
3.12 Rinsate: A portion of solvent used to rinse sampling equipment. The
rinsate is analyzed to demonstrate that samples were not contaminated
during sampling.
3.13 GC Column Performance Check Mixture: A tridecane solution containing a
mixture of selected PCDD/PCDF standards including the first and last
eluters for each homologous series, which is used to demonstrate continued
acceptable performance of the capillary column (i.e., < 25 percent valley
separation of 2,3,7,8-TCDD from all the other 21 TCDD isomers) and to
define the homologous PCDD/PCDF retention time windows.
D-7
-------
3.14 Performance Evaluation Materials: Representative sample portions
containing known amounts of certain unlabeled PCDD/PCDF congeners (in
particular the ones having a 2,3,7,8-substitution pattern). Representa-
tive interferences may be present. PEMs are obtained from the EPA EMSL-LV
and submitted to potential contract laboratories, who must analyze these
and obtain acceptable results before being awarded a contract for sample
analyses (see 1FB Pre-Award Bid Confirmations). PEMSs are also included
as unspecified ("blind") quality control (QC) samples in any sample batch
submitted to a laboratory for analysis.
3.15 Relative Response Factor: Response of the mass spectrometer to a known
amount of an analyte relative to a known amount of an internal standard.
3.16 Estimated Level of Method Blank Contamination: The response from a signal
occurring in the homologous PCDD/PCDF retention time windows, at any of
the masses monitored, is used to calculate the level of contamination
in the method blank, as described in Section 14 (this exhibit). The
results from such calculations must be reported along with the data
obtained on the samples belonging to the batch associated with the method
blank.
Reporting a method blank contamination level for any of the 2,3,7,8-
substituted congeners except OCDD and OCDF that exceeds 10 percent
of the desired detection limit would invalidate the results and require
automatic sample reruns (Exhibit C) for all positive samples found in
that batch of samples. A positive sample is defined as a sample found to
D-8
-------
contain at least one 2,3,7,8-substituted PCDD/PCDF congener (except OCDD
and OCDF). A valid method blank run is an analysis during which all
internal standard signals are characterized by S/N of at least 10:1.
3.17 Sample Rerun: Extraction of another portion of the sample followed by
extract cleanup and extract analysis.
3.18 Extract Reanalysis: Analysis by HRGC/HRMS of another aliquot of the
final extract.
3.19 Mass Resolution Check: Standard method used to demonstrate a static
resolving power of 10,000 minimum (10 percent valley definition).
3.20 Method Calibration Limits (MCLs): For a given sample size, a final
extract volume, and the lowest and highest concentration calibration
solutions, the lower and upper MCLs delineate the region of quantification
for which the HRGC/HRMS system was calibrated with standard solutions.
3.21 HRGC/HRMS Method Blank (MB): This additional QC check analysis corresponds
to a 2-uL injection of the method blank extract into the GC column and a
complete (tetra- through octachlorinated congeners) HRGC/HRMS analysis.
Such a QC check is required following a calibration run and before the
daily analysis of the first sample extract. Acceptable HRGC/HRMS method
blanks (see this exhibit, Section 3.16, for guidelines) must be obtained
before sample extracts can be analyzed.
D-9
25G
-------
3.22 Matrix Spike (MS): A sample which is spiked with a known amount of the
matrix spike fortification solution (this exhibit, Section 3.24) prior
to the extraction step. The recoveries of the matrix spike compounds are
determined; they are used to estimate the effect of the sample matrix
upon the analytical methodology.
3.23 Matrix Spike Duplicate (MSB): A second portion of the same sample as
used in the matrix spike analysis and which is treated like the matrix
spike sample.
3.24 Matrix Spike Fortification Solution: Solution used to prepare the MS and
MSD samples. It contains all unlabeled analytes listed in Table 5 at con-
centrations corresponding to the HRCC 3. The solution also contains all
internal standards used in the sample fortification solution at concen-
trations as shown in Table 2.
4. Interferences
4.1 Solvents, reagents, glassware and other sample processing hardware may
yield discrete artifacts or elevated baselines that may cause misinter-
pretation of the chromatographic data (see references 1 and 2 at the
end of this Section). All of these materials must be demonstrated to
be free from Interferents under the conditions of analysis by running
laboratory method blanks. Analysts should avoid using PVC gloves.
D-10
-------
4.2 The use of high-purity reagents and solvents helps minimize interference
problems. Purification of solvents by distillation in all-glass systems
may be necessary.
4.3 Interferents co-extracted from the sample will vary considerably from
matrix to matrix. PCDDs and PCDFs are often associated with other
interfering chlorinated substances such as polychlorinated biphenyls
(PCBs), polychlorinated diphenyl ethers (PCDPEs), polychlorinated
naphthalenes, and polychlorinated xanthenes that may be found at con-
centrations several orders of magnitude higher than the analytes of
interest. Retention times of target analytes must be verified using
reference standards. These values must correspond to the retention time
windows established in Section 8.1.3 (this exhibit). While certain clean-
up techniques are provided as part of this method, unique samples may
require additional cleanup steps to achieve lower detection limits.
4.4 A high-resolution capillary column (60 m DB-5) is used to resolve as many
PCDD and PCDF isomers as possible; however, no single column is known to
resolve all isomers. The use of several capillary columns will, in fact,
be necessary during the determination of the toxicity equivalency factors
(TEFs) (this exhibit, Section 14.7).
References:
1. "Control of Interferences in the Analysis of Human Adipose Tissue
for 2,3,7,8-Tetrachlorodlbenzo-p-dioxin". D. G. Patterson et al.,
Environ. Toxicol. Chem. 5, 355-360 (1986).
D-ll
258
-------
2. "Protocol for the Analysis of 2,3,7,8-TCDD by HRGC/HRMS".
J. S. Stanley and T. M. Sack, EPA 600/4-86-004.
5. Safety
5.1 The following safety practices are exerpted directly from EPA Method 613,
Section 4 (July 1982 version) and amended for use in conjunction with
this method.
Other PCDDs and PCDFs containing chlorine atoms in positions 2,3,7,8 are
known to have toxicities comparable to that of 2,3,7,8-TCDD. The
analyst should note that finely divided dry soils contaminated with PCDDs
and PCDFs are particularly hazardous because of the potential for inhala-
tion and ingestion. It is recommended that such samples be processed in
a confined environment, such as a hood or a glove box. Laboratory
personnel handling these types of samples should also wear masks fitted
with charcoal filter absorbent media to prevent inhalation of dust.
5.2 The toxicity or carcinogenicity of each reagent used in this method is
not precisely defined; however, each chemical compound should be treated
as a potential health hazard. From this viewpoint, exposure to these
chemicals must be kept to a minimum by whatever means available. The
laboratory is responsible for maintaining a current awareness file of
OSHA regulations regarding the safe handling of the chemicals specified
in this method. A reference file of material safety data sheets should
also be made available to all personnel involved in the chemical analysis.
D-12
-------
Additional references to laboratory safety are given in references 1-3
(see end of Section 5, this exhibit). Benzene and 2,3,7,8-TCDD have been
identified as suspected human or mammalian carcinogens.
5.3 Each laboratory must develop a strict safety program for the handling of
2,3,7,8-TCDD. The laboratory practices listed below are recommended.
5.3.1 Contamination of the laboratory will be minimized by conducting most of
the manipulations in a hood.
5.3.2 The effluents of sample splitters for the gas chromatograph and roughing
pumps on the HRGC/HRMS system should pass through either a column of ac-
tivated charcoal or be bubbled through a trap containing oil or high-
boiling alcohols.
5.3.3 Liquid waste should be dissolved in methanol or ethanol and irradiated
with ultraviolet light at a wavelength less than 290 nm for several days
(use F 40 BL lamps or equivalent). Using this analytical method, analyze
the liquid wastes and dispose of the solutions when 2,3,7,8-TCDD can no
longer be detected.
5.4 Some of the following precautions were issued by Dow Chemical U.S.A.
(revised 11/78) for safe handling of 2,3,7,8-TCDD in the laboratory and
amended for use in conjunction with this method.
5.4.1 The following statements on safe handling are as complete as possible on
D-13
-------
the basis of available toxicological information. The precautions for
safe handling and use are necessarily general in nature since detailed,
specific recommendations can be made only for the particular exposure
and circumstances of each individual use. Assistance in evaluating the
health hazards of particular plant conditions may be obtained from
certain consulting laboratories and from State Departments of Health or
of Labor, many of which have an industrial health service. The 2,3,7,8-
TCDD isomer is extremely toxic to certain kinds of laboratory animals.
However, it has been handled for years without injury in analytical and
biological laboratories. Techniques used in handling radioactive and
infectious materials are applicable to 2,3,7,8-TCDD.
5.4.1.1 Protective Equipment: Throw-away plastic gloves, apron or lab coat,
safety glasses and laboratory hood adequate for radioactive work.
5.4.1.2 Training:. Workers must be trained in the proper method of removing
contaminated gloves and clothing without contacting the exterior
surfaces.
5.4.1.3 Personal Hygiene: Thorough washing of hands and forearms after each
manipulation and before breaks (coffee, lunch, and shift).
5.4.1.4 Confinement: Isolated work area, posted with signs, segregated glass-
ware and tools, plastic-backed absorbent paper on benchtops.
5.4.1.5 Waste: Good technique includes minimizing contaminated waste.
D-14
OO
61
-------
Plastic bag liners should be used in waste cans.
5.4.1.6 Disposal of Hazardous Wastes: Refer to the November 7, 1986 issue of
the Federal Register on Land Ban Rulings for details concerning the
handling of dioxin-containing wastes.
5.4.1.7 Decontamination: Personnel - any mild soap with plenty of scrubbing
action. Glassware, tools and surfaces - Chlorothene NU Solvent (Trade-
mark of the Dow Chemical Company) is the least toxic solvent shown to
be effective. Satisfactory cleaning may be accomplished by rinsing
with Chlorothene, then washing with any detergent and water. Dish
water may be disposed to the sewer after percolation through a char-
coal bed filter. It is prudent to minimize solvent wastes because
they require special disposal through commercial sources that are
expensive.
5.4.1.8 Laundry: Clothing known to be contaminated should be disposed with
the precautions described under "Disposal of Hazardous Wastes".
Laboratory coats or other clothing worn in 2,3,7,8-TCDD work area may
be laundered. Clothing should be collected in plastic bags. Persons
who convey the bags and launder the clothing should be advised of the
hazard and trained in proper handling. The clothing may be put into a
washer without contact if the launderer knows the problem. The washer
should be run through one full cycle before being used again for other
clothing.
D-15
-------
5.A.1.9 Wipe Tests: A useful method of determining cleanliness of work
surfaces and tools is to wipe the surface with a piece of filter
paper, extract the filter paper and analyze the extract.
NOTE: Appendix A describes a procedure for the collection, handling,
analysis, and reporting requirements of wipe tests performed within
the laboratory. The results and decision making processes are based
on the presence of 2,3,7,8-substituted PCDD/PCDFs.
5.4.1.10 Inhalation: Any procedure that may produce airborne contamination
must be carried out with good ventilation. Gross losses to a venti-
lation system must not be allowed. Handling of the dilute solutions
normally used in analytical and animal work presents no significant
inhalation hazards except in case of an accident.
5.4.1.11 Accidents: Remove contaminated clothing immediately, taking precau-
tions not to contaminate skin or other articles. Wash exposed skin
vigorously and repeatedly until medical attention is obtained.
References:
1. "Carcinogens - Working with Carcinogens", Department of Health,
Education, and Welfare, Public Health Service, Center for Disease
Control. National Institute for Occupational Safety and Health.
Publication No. 77-206, August 1977.
2. "OSHA Safety and Health Standards, General Industry", (29 CFR 1910),
D-16
9«
283
-------
Occupational Safety and Health Administration, OSHA 2206 (revised
January 1976).
3. "Safety in Academic Chemistry Laboratories", American Chemical Society
Publication, Committee on Chemical Safety (3rd Edition, 1979.)
6. Apparatus and Equipment
6.1 High-Resolution Gas Chromatograph/High-Resolution Mass Spectrometer/Data
System (HRGC/HRMS/DS).
6.1.1 The GC must be equipped for temperature programming, and all required
accessories must be available, such as syringes, gases, and capillary
columns. The GC injection port must be designed for capillary
columns. The use of splitless injection techniques is recommended.
On-column 1-ul injections can be used on the 60-m DB-5 column. The use
of a moving needle injection port is also acceptable. When using the
method described in this protocol, a 2-uL injection volume is used
consistently (i.e., the injection volumes for all extracts, blanks,
calibration solutions and the performance check samples are 2 uL).
One-uL injections are allowed; however, laboratories are encouraged to
remain consistent throughout the analyses by using the same injection
volume at all times.
6.1.2 Gas Chromatograph/Mass Spectrometer (GC/MS) Interface—The GC/MS interface
components should withstand 350° C. The interface must be designed so
D-17
284
-------
that the separation of 2,3,7,8-TCDD from the other TCDD isomers achieved
in the gas chromatographic column is not appreciably degraded. Cold
spots or active surfaces (adsorption sites) in the GC/MS interface can
cause peak tailing and peak broadening. It is recommended that the GC
column be fitted directly into the mass spectrometer ion source without
being exposed to the ionizing electron beam. Graphite ferrules should
be avoided in the injection port because they may adsorb the PCDDs and
PCDFs. Vespel1" or equivalent ferrules are recommended.
6.1.3 Mass Spectrometer—The static resolving power of the instrument must be
maintained at a minimum of 10,000 (10 percent valley). The mass spec-
trometer must be operated in a selected ion monitoring (SIM) mode with
a total cycle time (including the voltage reset time) of one second or
less (this exhibit, Section 9.1.4.1). At a minimum, the ions listed in
Table 6 for each of the five SIM descriptors must be monitored. Note
that with the exception of the last descriptor (OCDD/OCDF), all the
descriptors contain 10 ions. The selection (Table 6) of the molecular
ions M and M+2 for 13c-HxCDF and 13c-HpCDF rather than M+2 and M+A (for
consistency) is to eliminate, even under high-resolution mass spectrometric
conditions, interferences occuring in these two ion channels for samples
containing high levels of native HxCDDs and HpCDDs. It is important to
maintain the same set of ions for both calibration and sample extract
analyses. The selection of the lock-mass ion is left to the performing
laboratory. The recommended mass spectrometer tuning conditions (this
exhibit, Section 8.2.3) are based on the groups of monitored ions shown
in Table 6.
D-18
I f-\ [—
OO
-------
6.1.4 Data System—A dedicated data system Is employed to control the rapid
multiple ion monitoring process and to acquire the data. Quantification
data (peak areas or peak heights) and SIM traces (displays of intensities
of each ion signal being monitored including the lock-mass ion as a
function of time) must be acquired during the analyses and stored.
Quantifications may be reported based upon computer-generated peak areas
or upon measured peak heights (chart recording). The data system must
be capable of acquiring data at a minimum of 10 ions in a single scan.
It is also recommended to have a data system capable of switching to
different sets of ions (descriptors) at specified times during an HRGC/
HUMS acquisition. The data system should be able to provide hard copies
of individual ion chromatograms for selected gas chromatographic time
intervals. It should also be able to acquire mass-spectral peak profiles
(this exhibit, Section 8.2.4) and provide hard copies of peak profiles
to demonstrate the required resolving power. The data system should
also permit the measurement of noise on the base line.
NOTE: The detector ADC zero setting must allow peak-to-peak measurement
of the noise on the base line of every monitored channel and allow for
good estimation of the instrument resolving power. In Figure 3, the
effect of different zero settings on the measured resolving power is shown,
6.2 GC Column
In order to have an isomer-specific determination for 2,3,7,8-TCDD and to
allow the detection of OCDD/OCDF within a reasonable time interval in one
D-19
266
-------
HRGC/HRMS analysis, the 60-m DB-5 fused-silica capillary column is recom-
mended. Minimum acceptance criteria must be demonstrated and documented
(this exhibit, Section 8.1). At the beginning of each 12-hour period
(after mass resolution is demonstrated) during which sample extracts or
concentration calibration solutions will be analyzed, column operating
conditions must be attained for the required separation on the column to
be used for samples. Operating conditions known to produce acceptable
results with the recommended column are shown in Table 7.
6.3 Miscellaneous Equipment and Materials
The following list of items does not necessarily constitute an exhaustive
compendium of the equipment needed for this analytical method.
6.3.1 Nitrogen evaporation apparatus with variable flow rate.
6.3.2 Balances capable of accurately weighing to 0.01 g and 0.0001 g.
6.3.3 Centrifuge.
6.3.4 Water bath, equipped with concentric ring covers and capable of being
temperature-controlled within + 2° C.
6.3.5 Stainless steel or glass container large enough to hold contents of
one-pint sample containers.
D-20
B*~1
/
-------
6.3.6 Glove box.
6.3.7 Drying oven.
6.3.8 Stainless steel spoons and spatulas.
6.3.9 Laboratory hoods.
6.3.10 Pipets, disposable, Pasteur, 150 mm long x 5 mm ID.
6.3.11 Pipets, disposable, serological, 10 mL, for the preparation of the
carbon column specified in Section 7.1.2.
6.3.12 Reacti-vial, 2 mL, silanized amber glass.
6.3.13 Stainless steel meatgrinder with a 3- to 5-mm hole size inner plate.
6.3.14 Separatory funnels, 125 mL.
6.3.15 Kuderna-Danish concentrator, 500 mL, fitted with 10-mL concentrator
tube and three-ball Snyder column.
6.3.16 Teflon1" boiling chips (or equivalent), washed with hexane before use.
6.3.17 Chromatographic column, glass, 300 mm x 10.5 mm, fitted with Teflon
stopcock.
D-21
268
-------
6.3.18 Adaptors for concentrator tubes.
6.3.19 Glass fiber filters.
6.3.20 Dean-Stark trap, 5 or 10 mL, with T-joints, condenser and 125-mL flask.
6.3.21 Continuous liquid-liquid extractor.
6.3.22 All-glass Soxhlet apparatus, 500-mL flask.
6.3.23 Glass funnels, sized to hold 170 mL of liquid.
6.3.24 Desiccator.
6.3.25 Solvent reservoir (125 mL), Kontes; 12.35 cm diameter (special order
item), compatible with gravity carbon column.
6.3.26 Rotary evaporator with a temperature-controlled water bath.
6.3.27 High-speed tissue homogenizer, equipped with an EN-8 probe or
equivalent.
6.3.28 Glass wool, extracted with methylene chloride, dried and stored in a
clean glass jar.
NOTE: Reuse of glassware should be minimized to avoid the risk of
D-22
26!)
-------
_DRAFT •
contamination. All glas-sware""ttra-t-i« ..reused must bV-aerupuiot:
- • .. _ -"-*•-.»_,/ (
cleaned as soon as possible after use, applying the following procedure:
Rinse glassware with the last solvent used in it, then with high-purity
acetone and hexane. Wash with hot detergent water. Rinse with copious
amounts of tap water and several portions of distilled water. Drain, dry
and heat in a muffle furnace at 400° C for 15 to 30 minutes. Volumetric
glassware must not be heated in a muffle furnace. Some thermally stable
materials (such as PCBs) may not be removed by heating in a muffle
furnace. In these cases, rinsing with high-purity acetone and
hexane may be substituted for muffle-furnace heating. After the
glassware is dry and cool, rinse it with hexane and store it inverted
or capped with solvent-rinsed aluminum foil in a clean environment.
7. Reagents and Standard Solutions
7.1 Column Chromatography Reagents
7.1.1 Alumina, neutral, Super 1, Woelm®, 80/200 mesh. Store in a sealed
container at room temperature in a desiccator over self-indicating
silica gel.
7.1.2 Carbopak C (80 to 100 mesh, Supelco 1-1025) and Celite 545® (Supelco).
Preparation of the Carbopak C/Celite 545® column: Thoroughly mix
3.6 g Carbopak C (80 to 100 mesh) and 16.4 g Celite 545® in a 40-mL
vial. Activate the mixture at 130° C for 6 hours, then store it in a
desiccator. Cut off both ends of a 10-mL disposable serological pipet
D-23
270
-------
to give a 4-inch long column. Fine-polish both ends and flare, if
desired. Insert a glass-wool plug at one end, then pack the column with
0.64 g of the activated Carbopak C/Celite 545® mixture to form a 2-cm
long absorbant bed. Cap the packing with another glass-wool plug.
7.2 Reagents
7.2.1 Sulfuric acid, concentrated, ACS grade, specific gravity 1.84.
7.2.2 Potassium hydroxide, ACS grade, 20 percent (w/v) in distilled water.
7.2.3 Sodium chloride, analytical reagent, 5 percent (w/v) in distilled
water.
7.2.4 Potassium carbonate, anyhdrous, analytical reagent.
7.3 Desiccating Agent
7.3.1 Sodium sulfate, granular, anhydrous; use as such.
7.4 Solvents
7.4.1 High-purity, distilled-in-glass or highest available purity: methylene
chloride, hexane, benzene, methanol, tridecane, isooctane, toluene,
cyclohexane, and acetone.
D-24
-------
7.5 Calibration Solutions
/.5.1 High-Resolution Concentration Calibration Solutions (Table 5) — Seven
triaecane solutions containing unlabeied (totaling 17) and carbon-labeied
(.totaling Li} fUJUs and t-cuts at known concentrations used to canorate
tne instrument. The concentration ranges are nomoiogue dependent, witn
the lowest values associated with the tetra- and pentachlorinated
dioxins and furans (2.5 pg/uL; and the highest for the octachlorinated
congeners (1000 pg/uL;.
/.5.2 These high-resolution concentration calibration solutions may be obtained
from the Quality Assurance Division, US EPA, Las Vegas, Nevada. However,
additional secondary standards must be obtained from commercial sources,
and solutions must be prepared in the contractor laboratory. Trace-
ability (Appendix B) of standards must be verified against EPA-supplied
standard solutions. Such procedures will be documented by laboratory
standard operating procedures (SOP) as required in IFB Preaward Bid
Confirmations, part 2.f.(4). It Is the responsibility of the laboratory
to ascertain that the calibration solutions received (or prepared) are
indeed at the appropriate concentrations before they are used to analyze
samples. A recommended traceablllty procedure for PCDD/PCDF standards
is described in Appendix B.
7.5.3 Store the concentration calibration solutions in 1-mL minivlals at
room temperature in the dark.
D-25
-------
7.6 GC Column Performance Check Solution
This solution contains the firstand last-eluting isomers for each homolo-
gous series from tetra- through hepta-chlorinated congeners. The solution
also contains a series of other TCDD isomers for the purpose of documenting
the chromatographic resolution. The *^Cj2~2,3,7,8-TCDD is also present.
The laboratory is required to use tridecane as the solvent and adjust the
volume so that the final concentration does not exceed 100 pg/uL per
congener. Table 8 summarized the qualitative composition (minimum
requirement) of this performance evaluation solution.
NOTE: The use of a PCDD/PCDF-containing fly-ash extract is allowed but
the qualitative equivalency of the fly-ash extract to the EPA solution
should be demonstrated for each fly-ash extract.
7.7 Sample Fortification Solution
This isooctane solution contains the nine internal standards at the nominal
concentrations that are listed in Table 2. The solution contains at least
one carbon-labeled standard for each homologous series, and it is used to
measure the concentrations of the native substances. (Note that ^c- -QCDF
is not present in the solution.)
7.8 Recovery Standard Solution
This tridecane solution contains two recovery standards (*-*C -1,2,3,4-
TCDD and 13C12-1,2,3,7,8,9-HxCDD) at a nominal concentration of 50 pg/uL
D-26
273
-------
per compound. Ten to titty UL ot this solution will be spiked into each
sample extract before the tinal concentration step and tiKU(J/HKMS analysis.
b. System Performance Criteria
System performance criteria are presented beiow. me laboratory may use the
recommended GC column described in Section 6.2 (.this exhibit;. It must be
documented that ail applicable system performance criteria specified in
Section 8.1 (this exhibit; were met before analysis ot any sample Is per-
formed, lable / provides recommended lie conditions tnat can be used to
sacisiy cne required criteria, figure 4 provides a typical i^-nour analysis
sequence wnereoy tne response factors and mass spectrometer resolving
power checks must be performed at the beginning and the end ot each iz-hour
period ot operation. A. tJC column performance check is only required at the
beginning of each 12-hour period during which samples are analyzed. An
HRGC/HRMS method blank run (this exhibit, Section 3.21) is required between
a calibration run and the first sample run. The same method blank extract
may thus be analyzed more than once if the number of samples within a batch
requires more than 12 hours of analyses.
8.1 GC Column Performance
8.1.1 Inject 2 uL (this exhibit, Section 6.1.1) of the column performance
check solution (this exhibit, Section 7.6) and acquire selected ion
monitoring (SIM) data as described in Section 6.1.3 (this exhibit) within
a total cycle time of £ 1 second (this exhibit, Section 9.1.4.1).
D-27
f~l I
i 4
-------
8.1.2 The chromatographic separation between 2,3,7,8-TCDD and the peaks repre-
senting any other TCDD isomers must be resolved with a valley of < 25
percent (Figure 5), where
Valley Percent = (x/y) (100)
x = measured as In Figure 5 from the 2,3,7,8-closest TCDD elutlng
Isomer, and
y - the peak height of 2,3,7,8-TCDD.
It Is the responsibility of the laboratory to verify the conditions
suitable for the appropriate resolution of 2,3,7,8-TCDD from all other
TCDD isomers. The GC column performance check solution also contains the
known first and last PCDD/PCDF eluters under the conditions specified In
this protocol. Their retention times are used to determine the eight
homologue retention time windows that are used for qualitative (this
exhibit, Section 13.4.1) and quantitative purposes. All peaks (that
Includes 13C12-2,3,7,8-TCDD) must be labeled and identified on the
chromatograms. Furthermore, all first eluters of a homologous series
must be labeled with the letter F, and all last eluters of a homologous
series must be labeled with the letter L (Figure 5 shows an example of
peak labeling for TCDD isomers). Any individual selected ion current
profile (SICP) (for the tetras, this would be the SICP for m/z 322 and
m/z 304) or the reconstructed homologue Ion current (for the tetras,
this would correspond to m/z 320 + m/z 322 + m/z 304 + m/z 306)
D-28
275
-------
constitutes an acceptable form of data presentation. An SICP for
the labeled compounds (e.g. , m/z 334 for labeled TCDD) is also required.
8.1.3 The retention times for the switching of SIM ions characteristic of one
homologous series to the next higher homologous series must be indicated
in the SICP. Accurate switching at the appropriate times is absolutely
necessary for accurate monitoring of these compounds. Allowable toler-
ance on the daily verification with the GC performance check solution
should be better than 10 seconds for the absolute retention times of all
the components of the mixture. Particular caution should be excercised
for the switching time between the last tetrachlorinated congener (i.e.,
1,2,8,9-TCDD) and the first pentachlorinated congener (i.e., 1,3,4,6,8-
PeCDF), as these two compounds elute within 15 seconds of each other on
the 60-m DB-5 column. A laboratory with a GC/MS system that is not
capable of detecting both congeners (1,2,8,9-TCDD and 1,3,4,6,8-PeCDF)
within one analysis must indicate in the case narrative of its report
which congener (only one is permitted) was missed.
8.2 Mass Spectrometer Performance
8.2.1 The mass spectrometer must be operated in the electron ionlzatlon mode.
A static resolving power of at least 10,000 (10 percent valley defini-
tion) must be demonstrated at appropriate masses before any analysis is
performed (this exhibit, Section 13). Static resolving power checks
must be performed at the beginning and at the end of each 12-hour period
of operation. However, it is recommended that a visual check (i.e.,
D-29
276
-------
documentation is not required) of the static resolution be made by using
the peak matching unit before and after each analysis. Corrective
actions must be implemented whenever the resolving power does not meet
the requirement.
8.2.2 Chromatography time for PCDDs and PCDFs exceeds the long-term mass
stability of the mass spectrometer. Because the instrument is operated
in the high-resolution mode, mass drifts of a few ppm (e.g., 5 ppm in
mass) can have serious adverse effects on the instrument performances.
Therefore, a mass-drift correction is mandatory. To that effect, it is
recommended to select a lock-mass ion from the reference compound (PFK
is recommended) used for tuning the mass spectrometer. The selection of
the lock-mass ion is dependent on the masses of the ions monitored
within each descriptor. Table 6 offers some suggestions for the lock-
mass ions. However, an acceptable lock-mass ion at any mass between the
lightest and heaviest ion in each descriptor can be used to monitor and
correct mass drifts. The level of the reference compound (PFK) metered
into the ion chamber during HRGC/HRMS analyses should be adjusted so
that the amplitude of the most intense selected lock-mass ion signal
(regardless of the descriptor number) does not exceed 10 percent of the
full-scale deflection for a given set of detector parameters. Under
those conditions, sensitivity changes that might occur during the
analysis can be more effectively monitored.
NOTE: Excessive PFK (or any other reference substance) may cause noise
problems and contamination of the ion source resulting in an increase in
D-30
2r-i -•)
I (
-------
downtime for source cleaning.
8.2.3 By using a PFK molecular leak, tune the Instrument to meet the minimum
required resolving power of 10,000 (10 percent valley) at m/z 304.9824
(PFK) or any other reference signal close to m/z 303.9016 (from TCDF).
By using the peak matching unit and the aforementioned PFK reference
peak, verify that the exact mass of m/z 380.9760 (PFK) is within 5 ppm
of the required value. Note that the selection of the low- and high-mass
ions must be such that they provide the largest voltage jump performed
in any of the five mass descriptors (Table 6).
8.2.4 Documentation of the instrument resolving power must then be accomplished
by recording the peak profile of the high-mass reference signal (m/z
380.9760) obtained during the above peak matching experiment by using
the low-mass PFK ion at m/z 304.9824 as a reference. The minimum
resolving power of 10,000 must be demonstrated on the high-mass ion
while it is transmitted at a lower accelerating voltage than the low-mass
reference ion, which Is transmitted at full sensitivity. The format of
the peak profile representation (Figure 6) must allow manual determina-
tion of the resolution, i.e., the horizontal axis must be a calibrated
mass scale (amu or ppm per division). The result of the peak width
measurement (performed at 5 percent of the maximum, which corresponds to
the 10-percent valley definition) must appear on the hard copy and
cannot exceed 100 ppm at m/z 380.9760 (or 0.038 amu at that particular
mass).
D-31
278
-------
9. Calibration
DRAFT
9.1 Initial Calibration
Initial calibration is required before any samples are analyzed for PCDDs
and PCDFs. Initial calibration is also required if any routine calibration
(this exhibit, Section 9.3) does not meet the required criteria listed in
Section 9.4 (this exhibit).
9.1.1 All seven high-resolution concentration calibration solutions listed in
Table 5 must be used for the initial calibration.
9.1.2 Tune the Instrument with PFK as described in Section 8.2.3 (this exhibit).
9.1.3 Inject 2 uL of the GC column performance check solution (this exhibit,
Section 7.6) and acquire SIM mass spectral data as described earlier In
Section 8.1 (this exhibit). The total cycle time must be £ 1 second.
The laboratory must not perform any further analysis until it is demon-
strated and documented that the criterion listed in Section 8.1.2 (this
exhibit) was met.
9.1.4 By using the same GC (this exhibit, Section 6.2) and mass spectrometer
(this exhibit, Section 6.1.3) conditions that produced acceptable results
with the column performance check solution, analyze a 2-uL portion of
each of the seven concentration calibration solutions once with the
following mass spectrometer operating parameters.
D-32
97Q
^ i <..»
-------
9.I.A.I The total cycle time for data acquisition must be < 1 second. The
total cycle time includes the sum of all the dwell times and voltage
reset times.
9.1.4.2 Acquire SIM data for all the ions listed in the five descriptors of
Table 6.
9.1.4.3 The ratio of integrated ion current for the ions appearing in Table 9
(homologous series quantification ions) must be within the indicated
control limits (set for each homologous series).
9.1.4.4 The ratio of integrated ion current for the ions belonging to the
carbon-labeled internal and recovery standards must be within the
control limits stipulated in Table 9.
NOTE: Sections 9.1.4.3 and 9.1.4.4 (this exhibit) require that 17 ion
ratios from Section 9.1.4.3 and 11 ion ratios from Section 9.1.4.4 be
within the specified control limits simultaneously in one run. It is
the laboratory's responsibility to take corrective action if the ion
abundance ratios are outside the limits.
9.1.4.5 For each SICP and for each GC signal corresponding to the elution of a
target analyte and of its labeled standards, the signal-to-noise ratio
(S/N) must be better than or equal to 2.5. Appendix C describes the
procedure to be followed for the measurement of the S/N from con-
spicuously weak signals. This measurement is required for any GC
D-33
280
-------
peak that has an apparent S/N of less than 5:1. The result of the
calculation must appear on the SICP above the GC peak in question.
9.1.4.6 Referring to Table 10, calculate the 17 relative response factors
(RRF) for unlabeled target analytes [RRF(n); n « 1 to 17] relative to
their appropriate internal standards (Table 5) and the nine RRFs for
the labeled 13C12 internal standards [RRF(m); m = 18 to 26)] relative
to the two recovery standards according to the following formulae:
RRF(n) = —
Ais
RRF(m) .
Qis ' Ars
where
Ax = sum of the integrated ion abundances of the quantification
ions (Tables 6 and 9) for unlabeled PCDDs/PCDFs,
Ais = sum °f tne integrated ion abundances of the quantification
ions (Tables 6 and 9) for the labeled internal standards,
Ars = sum of the integrated ion abundances of the quantification
ions (Tables 6 and 9) for the labeled recovery standards,
D-34
-------
Qj_s = quantity of the internal standard injected (pg),
Qrs = quantity of the recovery standard injected (pg), and
Qx = quantity of the unlabeled PCDD/PCDF analyte injected (pg).
The RRF(n) and RRF(m) are dimensionless quantities; the
units used to express QIS, Qrs anc* QX must be the same.
9.1.4.7 Calculate the RRF(n)s and their respective percent relative standard
deviations (%RSD) for the seven calibration solutions:
7
RRF(n) = 1/7 I RRF^n) ,
where n represents a particular PCDD/PCDF (2,3,7,8-substituted) con-
gener (n = 1 to 17; Table 10), and j is the injection number (or
calibration solution number; j = 1 to 7).
9.1.4.8 The relative response factors to be used for the determination of the
concentration of total isomers in a homologous series (Table 10) are
calculated as follows:
9.1.4.8.1 For congeners that belong to a homologous series containing only
one isomer (e.g., OCDD and OCDF) or only one 2,3,7,8-substituted
isomer (Table 4; TCDD, PeCDD, HpCDD, and TCDF), the mean RRF used
D-35
9H9
*~ O <^.
-------
will be the same as the mean RRF determined in Section 9.1.4.7 (this
exhibit).
13
C12~°CDF as an
NOTE: The calibration solutions do not contain
internal standard. This is because a minimum resolving power of
12,000 is required to resolve the [M+6]+ ion of 13C12-OCDF from the
[M+2]+ ion of OCDD (and [M+4]+ from 13C12-OCDF with [M]+ of OCDD).
Therefore, the RRF for OCDF is calculated relative to
9.1.4.8.2 For congeners that belong to a homologous series containing more
than one 2,3,7,8-substituted isomer (Table 4), the mean RRF used
for those homologous series will be the mean of the RRFs calculated
for all individual 2,3,7,8-substituted congeners using the equation
below:
RRF(k)
1 t
I
t n=l
RRFT
where
k = 27 to 30 (Table 10), with 27 = PeCDF; 28 = HxCDF;
29 = HxCDD; and 30 = HpCDF,
t = total number of 2,3,7,8-substituted isomers present in
the calibration solutions (Table 5) for each homologous
series (e.g., two for PeCDF, four for HxCDF, three for
HxCDD, two for HpCDF).
D-36
283
-------
NOTE: Presumably, the HRGC/HRMS response factors of different isomers
within a homologous series are different. However, this analytical
protocol will make the assumption that the HRGC/HRMS responses of all
isomers in a homologous series that do not have the 2,3,7,8-substitution
pattern are the same as the responses of one or more of the 2,3,7,8-
substituted isomer(s) in that homologous series.
9.1.4.9 Relative response factors [RRF(m)] to be used for the determination
of the percent recoveries for the nine internal standards are calcu-
lated as follows:
RRF(m)
Aism *
* Ars
1 7
RRF(m) = - £ RRFj(m),
7 J-l
where:
m = 18 to 26 (congener type) and j = 1 to 7 (injection number),
Aism = sum of the integrated ion abundances of the quantification ions
(Tables 6 and 9) for a given internal standard (m = 18 to 26),
Ars = sum of the integrated ion abundances of the quantification ions
(Tables 6 and 9) for the appropriate recovery standard (see Table 5,
footnotes),
D-37
284
-------
J)RAFT j
Q,.., and Qtf.m = quantities of, respectfvely",~Tnii5~TECovery standard (rs)
rs i. s
and a particular internal standard (is = m) injected
(pg),
RRF(m) = relative response factor of a particular internal
standard (m) relative to an appropriate recovery
standard, as determined from one injection, and
RRF(m) = calculated mean relative response factor of a particular
internal standard (m) relative to an appropriate recovery
standard, as determined from the seven initial calibra-
tion injections (j).
9.2 Criteria for Acceptable Calibration
The criteria listed below for acceptable calibration must be met before
the analysis is performed.
9.2.1 The percent relative standard deviations for the mean response factors
[RRF(n) and RRF(m)] from each of the 26 determinations (17 for the
unlabeled standards and 9 for the labeled reference compounds) must be
less than 20 percent.
9.2.2 The S/N for the GC signals present in every SICP (including the
ones for the labeled standards) must be > 2.5.
9.2.3 The isotoplc ratios (Table 9) must be within the specified control limits.
D-38
23;
O
-------
NOTE: If the criterion for acceptable calibration listed in Section
9.2.1 (this exhibit) is met, the analyte-specific RRF can then be con-
sidered independent of the analyte quantity for the calibration concen-
tration range. The mean RRFs will be used for all calculations until
the routine calibration criteria (this exhibit, Section 9.4) are no
longer met. At such time, new mean RRFs will be calculated from a new
set of injections of the calibration solutions.
9.3 Routine Calibration (Continuing Calibration Check)
Routine calibrations must be performed at the beginning of a 12-hour
period after successful mass resolution and GC resolution performance
checks. A routine calibration is also required at the end of a 12-hour
shift.
9.3.1 Inject 2 uL of the concentration calibration solution HRCC-3 containing
10 pg/uL of tetra- and pentachlorinated congeners, 25 pg/uL of hexa-
and heptachlorinated congeners, 50 pg/uL of octachlorinated congeners,
and the respective internal and recovery standards (Table 5). By using
the same HRGC/HRMS conditions as used in Sections 6.1.3 and 6.2 (this
exhibit), determine and document an acceptable calibration as provided in
Section 9.4 (this exhibit).
9.4 Criteria for Acceptable Routine Calibration
The following criteria must be met before further analysis is performed.
D-39
23G
-------
If these criteria are not met, corrective action must be taken.
9.4.1 The measured RRFs [RRF(n) for the unlabeled standards] obtained during
the routine calibration runs must be within 20 percent of the mean
values established during the initial calibration (this exhibit, Section
9.1.4.7).
9.4.2 The measured RRFs (RRF(m) for the labeled standards] obtained during
the routine calibration runs must be within 20 percent of the mean
values established during the initial calibration (this exhibit, Section
9.1.4.9).
9.4.3 The ion-abundance ratios (Table 9) must be within the allowed control
limits.
9.4.4 If either one of the above criteria (this exhibit, Sections 9.4.1 and
9.4.2) is not satisfied, the entire initial calibration process (this
exhibit, Section 9.1) must be repeated. If the ion-abundance ratio
criterion (this exhibit, Section 9.4.3) is not satisfied, refer to the
note in Section 9.1.4.4 (this exhibit) for resolution.
NOTE: An initial calibration must be carried out whenever the HRCC-3,
the sample fortification or the recovery standard solution is replaced
by a new solution from a different lot.
10. Quality Assessment/Quality Control Procedures
D-40
287
-------
See Exhibit E for QA/QC requirements.
11. Sample Preservation
11.1 The sample collection, shipping, handling, and chain-of-custody procedures
are not described in this document. Sample collection personnel will, to
the extent possible, homogenize samples in the field before filling the
sample containers. This should minimize or eliminate the necessity for
sample homogenization in the laboratory. The analyst should make a judg-
ment, based on the appearance of the sample, regarding the necessity for
additional mixing. If the sample is clearly inhomogeneous, the entire
contents should be transferred to a glass or stainless steel pan for
mixing with a stainless steel spoon or spatula before removal of a
sample portion for analysis.
11.2 Grab and composite samples must be collected in glass containers.
Conventional sampling practices must be followed. The bottle must not be
prewashed with sample before collection. Sampling equipment must be free
of potential sources of contamination.
11.3 Grinding or Blending of Fish Samples.
If not otherwise specified by the EPA, the whole fish (frozen) should be
blended or ground to provide a homogeneous sample. The use of a stain-
less steel meatgrinder with a 3- to 5-mm hole size inner plate is recom-
mended. In some circumstances, analysis of fillet or specific organs of
fish may be requested by the EPA. If so requested by the EPA, the above
whole fish requirement is superseded.
D-41
288
-------
11.4 With the exception of the fish and adipose tissues, which must be stored
at -20° C, all samples must be stored at 4° C, extracted within 30 days
and completely analyzed within 45 days of collection.
11.5 Phase Separation - This is a guideline for phase separation on very wet
(>25 percent water) soil and sediment samples. Place a 50-g portion in a
suitable centrifuge bottle and centrifuge for 30 minutes at 2,000 rpm.
Remove the bottle and mark the interface level on the bottle. Estimate
the relative volume of each phase. With a disposable pipet, transfer the
liquid layer into a clean bottle. Mix the solid with a stainless steel
spatula and remove a portion to be weighed and analyzed (percent moisture
determination, extraction). Return the remaining solid portion to the
original sample bottle (empty) or to a clean sample bottle that is properly
labeled, and store it as appropriate. Analyze the solid phase by using
only the soil and sediment method. Take note of and report the estimated
volume of liquid before disposing of the liquid as a liquid waste.
CAUTION: Finely divided soils and sediments contaminated with PCDDs/PCDFs
are hazardous because of the potential for inhalation or ingestion of
particles containing PCDDs/PCDFs (including 2,3,7,8-TCDD). Such samples
should be handled in a confined environment (i.e., a closed hood or a
glove box).
11.6 Soil, Sediment or Paper Sludge (Pulp) Percent Moisture Determination.
The percent moisture of soil or sediment samples showing detectable
D-42
239
-------
levels (see note below) of at least one 2,3,7,8-substituted PCDD/PCDF
congener is determined according to the following recommended procedure.
Weigh a 9.5- to 10.5-g portion of the soil or sediment sample (+ 0.5 g)
to three significant figures. Dry it to constant weight at 100° C in an
adequately ventilated oven. Allow the sample to cool in a desiccator.
Weigh the dried solid to three significant figures. Calculate and report
the percent moisture on Form (to be determined). Do not use this solid
portion of the sample for extraction, but instead dispose of it as
hazardous waste. The pulp sample (10 g) should be dried overnight in a
fume hood.
NOTE: Until detection limits are determined (Section 1.2, this exhibit),
the lower MCLs (Table 1) may be used to estimate.the minimum detectable
levels.
Weight of wet soil - Weight of dry soil
Percent moisture = . x 100
Weight of wet soil
11.7 Fish Tissue Lipid Content Determination
The percent lipid of fish samples showing detectable levels (see Section
11.6 note; this exhibit) of at least one 2,3,7,8-substituted PCDD/PCDF
congener is determined as follows:
Use a separate portion (2 g) of the ground frozen fish sample. Blend it
D-43
290
-------
with 6 g anhydrous sodium sulfate, pour the mixture in a 1-cm i.d.
glass column and extract the lipids by passing two 25-mL portions of
methylene chloride through the column and collecting the extract in a
tared 100-mL round-bottom flask. Concentrate the extract on a rotary
evaporator until constant weight is attained. The percent lipid is
calculated using the following expression:
Weight of residue from extraction (in g)
Percent lipid = x 100
Weight of fish tissue portion (in g)
Dispose of the lipid residue as a hazardous waste if the results of the
analysis indicate the presence of PCDOs or PCDFs.
11.8 Adipose Tissue Lipid Content Determination
Details for the determination of the adipose tissue lipid content are
provided in Section 12.11.3 (this exhibit).
12. Extraction and Cleanup Procedures
12.1 Internal standard addition. Use a portion of 1 g to 1000 g (typical sam-
ple size requirements for each type of matrix are given in Section 12.2
of this exhibit and in Table 1) of the sample to be analyzed. Transfer
the sample portion to a tared flask and determine its weight. Except for
adipose tissue, add an appropriate quantity of the sample fortification
mixture (this exhibit, Section 3.8) to the sample. All samples should be
D-44
291
-------
spiked with 100 uL of the sample fortification mixture to give internal
standard concentrations as indicated in Table 1. As an example, for
l^Cj2~2»3,7,8-TCDD, a 10-g soil sample requires the addition of 1000 pg of
13C12-2,3,7,8-TCDD to give the requisite 100 ppt fortification level. For
the fortification of soil, sediment, fly ash, water and fish tissue
samples, mix the 100 uL sample fortification solution with 1.5 mL ace-
tone. Do not dilute the isooctane solution for the other matrices. The
fortification of adipose tissue is carried out at the time of homogeniza-
tion (this exhibit, Section 12.11.2.3).
12.2 Extraction
The extraction and purification procedures for biological tissue samples
are described in Sections 12.10 (fish tissue) and 12.11 (adipose tissue)
of this exhibit.
12.2.1 Sludge/Fuel Oil. Extract aqueous sludge samples by refluxing a sample
(e.g., 2 g) with 50 mL toluene (or benzene) in a 125-mL flask fitted
with a Dean-Stark water separator. Continue refluxing the sample
until all the water is removed. Cool the sample, filter the toluene
(or benzene) extract through a glass-fiber filter, or equivalent, into
a 100-mL round-bottom flask. Rinse the filter with 10 mL toluene (or
benzene), and combine the extract and rinsate. Concentrate the combined
solutions to near dryness on a rotary evaporator at 50° C (toluene) or
a Kudema-Danish (KD) apparatus (benzene). Use of an inert gas to
concentrate the extract is also permitted. Proceed with Section 12.2.4
below.
D-45
292
-------
NOTE: If the labeled sludge sample dissolves In toluene, treat it
according to the instructions in Section 12.2.2 below. If the labeled
sludge sample originates from pulp (paper mills), treat it according
to the instructions starting in Section 12.10.1 but without the addition
of sodium sulfate.
12.2.2 Still-Bottom. Extract still-bottom samples by mixing a sample portion
(e.g., 1.0 g) with 10 mL toluene (or benzene) in a small beaker and
filtering the solution through a glass-fiber filter (or equivalent)
into a 50-mL round-bottom flask. Rinse the beaker and filter with 10
mL toluene (or benzene). Concentrate the combined toluene (or benzene)
solutions to near dryness on a rotary evaporator at 50° C. A KD appa-
ratus can be used if benzene is the extraction solvent. Proceed with
Section 12.2.4 below.
12.2.3 Fly Ash. Extract fly ash samples by placing a sample portion (e.g., 10
g) and an equivalent amount of anhydrous sodium sulfate in a Soxhlet
extraction apparatus charged with 100 mL toluene (or benzene), and
extract for 16 hours using a three cycle/hour schedule. Cool and
filter the toluene (or benzene) extract through a glass-fiber filter
into a 500-mL round-bottom flask. Rinse the filter with 5 mL toluene
(or benzene). Concentrate the combined toluene (or benzene) solutions
to near dryness on a rotary evaporator (toluene) at 50° C or a KD
apparatus (benzene). Proceed with Section 12.2.4 below.
12.2.4 Transfer the residue to a 125-mL separatory funnel using 15 mL hexane.
D-46
293
-------
Rinse the flask with two 5-mL portions of hexane and add the rinses to
the funnel. Shake two minutes with 50 mL of 5 percent sodium chloride
solution, discard the aqueous layer and proceed with Section 12.3
(this exhibit).
12.2.5 Soil. Add 10 g anhydrous sodium sulfate to the soil sample portion
(e.g., 10 g) and mix thoroughly with a stainless steel spatula. After
breaking up any lumps, place the soil/sodium sulfate mixture in the
Soxhlet apparatus on top of a glass-wool plug (the use of an extraction
thimble is optional). Add 200 to 250 mL benzene (or toluene) to the
Soxhlet apparatus and reflux for 24 hours. The solvent must cycle
completely through the system at least three times per hour.
12.2.5.1 Transfer the extract from Section 12.2.5 to a KD apparatus mounted
with a three-ball Snyder column (or to a 500-mL round-bottom flask
for evaporating the toluene on a rotary evaporator).
12.2.5.2 Add a Teflon™ or an equivalent boiling chip. Concentrate in a 70° C
water bath to an apparent volume of 10 mL. Remove the apparatus from
the water bath and allow it to cool for 5 minutes.
12.2.5.3 Add 50 mL hexane and a new boiling chip to the KD flask. Concen-
trate in a water bath to an apparent volume of 10 mL. Remove the
apparatus from the water bath and allow to cool for 5 minutes.
12.2.5.4 Remove and invert the Snyder column, and rinse it down into the KD
D-47
294
-------
apparatus with two 1-mL portions of hexane. Decant the contents of
the KD apparatus and concentrator tube Into a 125-mL separatory
funnel. Rinse the RD apparatus with two additional 5-mL portions of
hexane, and add the rinsates to the funnel. Proceed with Section
12.3 (this exhibit).
12.2.6 Aqueous Samples. Mark the water meniscus on the side of the 1-L sample
bottle for later determination of the exact sample volume. Pour the
entire sample (approximately 1-L) into a 2-L separatory funnel. Proceed
with Section 12.2.6.1 (this exhibit).
NOTE: A continuous liquid-liquid extractor may be used in place of a
separatory funnel when experience with a sample from a given source
indicates that a serious emulsion problem will result or an emulsion is
encountered when using a separatory funnel. Add 60 mL methylene chloride
to the sample bottle, seal, and shake for 30 seconds to rinse the inner
surface. Transfer the solvent to the extractor. Repeat the sample bot-
tle rinse with an additional 50- to 100-mL portion of methylene chloride
and add the rinsate to the extractor. Add 200 to 500 mL methylene
chloride to the distilling flask, add sufficient reagent water to ensure
proper operation, and extract for 24 hours. Allow to cool, then detach
the distilling flask. Dry and concentrate the extract as described in
Sections 12.2.6.1 and 12.2.6.2 (this exhibit). Proceed with Section
12.2.6.3 (this exhibit).
12.2.6.1 Add 60 mL methylene chloride to the sample bottle, seal, and shake for
D-48
295
-------
30 seconds to rinse the inner surface. Transfer the solvent to the
separatory funnel and extract the sample by shaking the funnel for
two minutes with periodic venting. Allow the organic layer to sepa-
rate from the water phase for a minimum of 10 minutes. If the emul-
sion interface between layers is more than one-third the volume of
the solvent layer, the analyst must employ mechanical techniques to
complete the phase separation. Collect the methylene chloride into a
KD apparatus (mounted with a 10-mL concentrator tube) by passing the
sample extracts through a filter funnel packed with a glass-wool plug
and 5 g anhydrous sodium sulfate. Repeat the extraction twice with
fresh 60-ml portions of methylene chloride. After the third extrac-
tion, rinse the sodium sulfate with an additional 30 mL methylene
chloride to ensure quantitative transfer. Combine all extracts and
the rinsate in the KD apparatus.
12.2.6.2 Attach a Snyder column and concentrate the extract on a water bath
until the apparent volume of the liquid is 5 mL. Remove the KD
apparatus and allow it to drain and cool for at least 10 minutes.
Remove the Snyder column, add 50 mL hexane, re-attach the Snyder
column and concentrate to approximately 5 mL. Add a new boiling chip
to the KD apparatus before proceeding with the second concentration
step. Rinse the flask and the lower joint with two 5-mL portions
of hexane and combine the rinsates with the extract to give a final
volume of about 15 mL.
12.2.6.3 Determine the original sample volume by filling the sample bottle to
D-49
296
-------
the mark with water and transferring the water to a 1000-mL graduated
cylinder. Record the sample volume to the nearest 5 mL. Proceed
with Section 12.3 (this exhibit).
12.3 Partition the extract (15 mL hexane) against 40 mL of 20 percent (w/v)
aqueous potassium hydroxide (KOH). Shake for two minutes. Remove and
discard the aqueous layer (bottom). Repeat the base washing until no
color is visible in the bottom layer (perform a maximum of four base
washings). Strong base (KOH) is known to degrade certain PCDDs/PCDFs,
so contact time must be minimized.
12.4 Partition the extract (15 mL hexane) against 40 mL of 5 percent (w/v)
aqueous sodium chloride. Shake for two minutes. Remove and discard the
aqueous layer (bottom).
12.5 Partition the extract against 40 mL concentrated sulfuric acid. Shake
for two minutes. Remove and discard the sulfuric acid layer (bottom).
Repeat the acid washing until no color is visible in the acid layer
(perform a maximum of four acid washings).
12.6 Partition the extract against 40 mL of five percent (w/v) sodium chloride.
Shake for two minutes. Remove and discard the aqueous layer (bottom).
Dry the extract by pouring it through a funnel containing anhydrous
sodium sulfate and collect it in a 50-mL round-bottom flask. Rinse the
sodium sulfate with two 15-mL portions of hexane, add the rinsates to the
50-mL flask, and concentrate the hexane solution to near dryness on a
D-50
9 O - ?
&a ^
-------
rotary evapotator (35° C water bath), making sure all traces of toluene
(when applicable) are removed. (Use of blow-down with an inert gas to
concentrate the extract is also permitted.)
12.7 Pack a gravity column (glass, 300 mm x 10.5 mm), fitted with a Teflon™
stopcock, in the following manner: Insert a glass-wool plug into the
bottom of the column. Add a 4-g layer of sodium sulfate. Add a 4-g
layer of Woelm® Super 1 neutral alumina. Tap the top of the column
gently. Woelm® Super 1 neutral alumina need not be activated or cleaned
before use, but it should be stored in a sealed desiccator. Add a 4-g
layer of anhydrous sodium sulfate to cover the alumina. Elute with 10 mL
hexane and close the stopcock just before exposure of the sodium sulfate
layer to air. Discard the eluate. Check the column for channeling. If
channeling is present, discard the column. Do not tap a wetted column.
12.8 Dissolve the residue from Section 12.6 (this exhibit) in 2 mL hexane and
apply the hexane solution to the top of the column. Elute with enough
hexane (3-4 mL) to complete the transfer of the sample cleanly to the
surface of the alumina. Discard the eluate.
12.8.1 Elute with 10 mL of 8 percent (v/v) methyleiie chloride in hexane.
12.8.2 Elute the PCDDs and PCDFs from the column with 15 mL of 60 percent
(v/v) methylene chloride in hexane and collect this fraction in a
conical shaped (15 mL) concentrator tube.
D-51
298
-------
12.9 Carbon Column Cleanup
Prepare a Carbopak C/Celite 545® column as described in Section 7.1.2
(this exhibit).
12.9.1 With a carefully regulated stream of nitrogen, concentrate the
60-percent fraction (this exhibit, Section 12.8.2) to about 2 mL.
Rinse the Carbopak C/Celite 545® with 5 raL toluene followed by 2 mL of
75:20:5 methylene chloride/methanol/benzene, 1 mL of 1:1 cyclohexane/
methylene chloride, and 5 mL hexane. The flow rate should be less than
0.5 mL/min. Discard the rinsates. While the column is still wet with
hexane, add the sample concentrate to the top of the column. Rinse the
concentrator tube which contained the sample concentrate twice with
1 mL hexane and add the rinsates to the top of the column. Elute the
column sequentiallly with two 2-mL portions of hexane, 2 mL cyclohexane/
methylene chloride (50:50, v/v), and 2 mL methylene chloride/methanol/
benzene (75:20:5, v/v). Combine these eluates; this combined fraction
may be used as a check on column efficiency. Now turn the column
upside down and elute the PCDD/PCDF fraction with 20 mL toluene.
Verify that no carbon fines are present in the eluate.
12.9.2 Concentrate the toluene fraction to about 1 mL on a rotary evaporator
by using a water bath at 50° C. Carefully transfer the concentrate into
a 1-tnL minivial and, again at elevated temperature (50° C), reduce the
volume to about 100 uL using a stream of nitrogen and a sand bath.
Rinse the rotary evaporator flask three times with 300 uL of a solution
D-52
290
-------
of 1 percent toluene In methylene chloride. Add 10 uL for soil, sedi-
ment, and water, or 50 uL for sludge, still-bottom and fly ash of the
tridecane recovery standard solution. Store the sample at room tempera-
ture in the dark.
12.10 Extraction and Purification Procedures for Fish and Paper Pulp Samples
12.10.1 Add 30 g anhydrous sodium sulfate to a 10-g portion of a homogeneous
fish sample (this exhibit, Section 11.3) and mix thoroughly with a
stainless steel spatula. After breaking up any lumps, place the
fish/sodium sulfate mixture in the Soxhlet apparatus on top of a glass-
wool plug. Add 200 mL hexane/methylene chloride (1:1) to the Soxhlet
apparatus and reflux for 12 hours. The solvent must cycle completely
through the system at least three times per hour. Follow the same
procedure for the dried (this exhibit, Section 11.6) paper pulp samples.
12.10.2 Transfer the fish or paper pulp extract from Section 12.10.1 to a KD
apparatus equipped with a Snyder column.
12.10.3 Add a Teflon™ or an equivalent boiling chip. Concentrate the extract
in a water bath to an apparent volume of 10 mL. Remove the apparatus
from the water bath and allow to cool for 5 minutes.
12.10.4 Add 50 mL isooctane and a new boiling chip to the KD flask. Concentrate
in a water bath to an apparent volume of 5 mL. Remove the apparatus
from the water bath and allow to cool for 5 minutes.
D-53
300
-------
NOTE: The methylene chloride must have been completely removed before
proceeding with the next step.
12.10.5 Remove and Invert the Snyder Column and rinse it into the KD apparatus
with two 1-mL portions of hexane. Decant the contents of the KD
apparatus and concentrator tube into a 125-mL separatory funnel.
Rinse the KD apparatus with two additional 5-mL portions of hexane and
add the rinsates to the funnel. Proceed with the cleanup according to
the Instructions starting in Section 12.5 (this exhibit).
12.11 Extraction and Purification Procedures for Human Adipose Tissue
12.11.1 Human adipose tissue samples must be stored at -20° C from the time of
collection until the time of analysis. The use of chlorinated mate-
rials during the collection of the sample must be avoided. Samples
are handled with stainless steel forceps, spatulas, or scissors. All
sample bottles (glass) are cleaned as specified in the note appearing
in Section 6.3 (this exhibit). Teflon"1-!ined caps should be used.
12.11.2 Adipose Tissue Extraction Procedure
12.11.2.1 Weigh to the nearest 0.01 g a 10-g portion of a frozen adipose
tissue sample into a culture tube (2.2 x 15 cm).
'NOTE: The sample size may be smaller, depending on availability.
In such a situation, the analyst is required to adjust the volume of
D-54
301
-------
the internal standard solution added to the sample to meet the for-
tification level stipulated in Table 1.
12.11.2.2 Allow the adipose tissue specimen to reach room temperature (up to 2
hours).
12.11.2.3 Add 10 mL methylene chloride and 100 uL of the sample fortification
solution. Homogenize the mixture for approximately 1 minute with a
tissue homogenizer.
12.11.2.4 Allow the mixture to separate, and remove the methylene chloride
extract from the residual solid material with a disposable pipet.
Percolate the methylene chloride through a filter funnel containing
a clean glass-wool plug and 10 g anhydrous sodium sulfate.' Collect
the dried extract in a graduated 100-mL volumetric flask.
12.11.2.5 Add a second 10-mL portion of methylene chloride to the sample and
homogenize for 1 minute. Decant the solvent, dry it, and transfer
it to the 100-mL volumetric flask (this exhibit, Section 12.11.2.4).
12.11.2.6 Rinse the culture tube with at least two additional portions of
methylene chloride (10 mL each), and transfer the entire contents
to the filter funnel containing the anhydrous sodium sulfate. Rinse
the filter funnel and the anhydrous sodium sulfate contents with
additional methylene chloride (20 to 40 mL) into the 100-mL flask.
Discard the sodium sulfate.
D-55
302
-------
_
12.11.2.7 Adjust the volume to- Ebe"."rDO.=inL*vmark-witij 'inethylene chloride.
12.11.3 Adipose Tissue Lipid Content Determination
12.11.3.1 Preweigh a clean 1-dram glass vial to the nearest 0.0001 g on an
analytical balance tared to zero.
12.11.3.2 Accurately transfer 1.0 mL of the final extract (100 mL) from Section
12.11.2.6 (this exhibit) to the 1-dram vial. Reduce the volume of
the extract on a water bath (50-60° C) by a gentle stream of
purified nitrogen until an oily residue remains. Nitrogen blow-down
is continued until a constant weight is achieved.
12.11.3.3 Accurately weigh the 1-dram vial with the residue to the nearest
0.0001 g and calculate the weight of the lipid present in the vial
based on the difference of the weights.
12.11.3.4 Calculate the percent lipid content of the original sample to the
nearest 0.1 percent as shown below:
x vext
Lipid Content, LC (%) = x 100
wat x val
where
V/ir = weight of the lipid residue to the nearest 0.0001 g
calculated from Section 12.11.3.3 (this exhibit),
D-56
303 ^
-------
Vext = total volume (100 mL) of the extract in mL from
Section 12.11.2.6 (this exhibit)
Wat = weight of the original adipose tissue sample to the
nearest 0.01 g from Section 12.11.2.1 (this exhibit),
and
= volume of the aliquot of the final extract in mL
used for the quantitative measure of the lipid residue
(1.0 mL).
12.11.3.5 Record the lipid residue measured in Section 12.11.3.3 (this exhibit)
and the percent lipid content from Section 12.11.3.4 (this exhibit).
12. 11. A Adipose Tissue Extract Concentration
12.11.4.1 Quantitatively transfer the remaining extract volume (99.0 mL) to a
500-mL round-bottom flask. Rinse the volumetric flask with 20 to 30
mL of additional methylene chloride to ensure quantitative transfer.
12.11.4.2 Concentrate the extract on a rotary evaporator and a water
bath at 40°C until an oily residue remains.
12.11.5 Adipose Tissue Extract Cleanup Procedures
12.11.5.1 Add 200 mL hexane to the lipid residue in the 500-mL Erlenmeyer
D-57
304
-------
flask and swirl the flask to dissolve the residue.
12.11.5.2 Slowly add, with stirring, 100 g of 40-percent w/w sulfuric-acid-
impregnated silica gel. Stir with a magnetic stirrer for two hours
at room temperature.
12.11.5.3 Allow the solid phase to settle and decant the liquid through a
powder funnel containing 20 g anhydrous sodium sulfate into another
500-mL Erlenmeyer flask.
12.11.5.4 Rinse the solid phase with two 50-mL portions of hexane. Stir each
rinse for 15 minutes, decant, and dry as described under Section
12.11.5.3. Combine the hexane extracts from Section 12.11.5.3
(this exhibit) with the rinses.
12.11.5.5 Rinse the sodium sulfate in the powder funnel with an additional
25 mL hexane and combine this rinse with the hexane extracts from
Section 12.11.5.4 (this exhibit).
12.11.5.6 Prepare an acidic silica column as follows: Pack a 2-cm x 10-cm
chromatographic column with a glass-wool plug, add approximately
20 mL hexane, add 4 g silica gel and allow to settle, then add 16 g
of 40-percent w/w sulfuric-acid-impregnated-silica gel and allow to
settle. Elute the excess hexane from the column until the solvent
level reaches the top of the chromatographic packing. Verify that
the column does not have any air bubbles and channels.
D-58
SOS
-------
12.11.5.7 Quantitatively transfer the hexane extract from the Erlenmeyer flask
(this exhibit, Sections 12.11.5.3 through 12.11.5.5) to the silica
gel column reservoir. Allow the hexane extract to percolate through
the column and collect the eluate in a 500-mL KD apparatus.
12.11.5.8 Complete the elution by percolating 50 mL hexane through the column
into the KD apparatus. Concentrate the eluate on a steam bath to
approximately 5 mL. Use nitrogen blow-down to bring the final
volume to about 100 uL.
NOTE: If the silica gel impregnated with 40-percent sulfuric acid
is highly discolored throughout the length of the adsorbent bed,
the cleaning procedure must be repeated beginning with Section
12.11:5.1 (this exhibit).
12.11.5.9 The extract is ready for the alumina and carbon cleanups described
in Sections 12.7 through 12.9.2 (this exhibit).
13. Analytical Procedures.
13.1 Remove the sample extract or blank from storage. With a stream of dry,
purified nitrogen, reduce the extract volume to 10 uL or 50 uL (the
volume of the tridecane recovery standard solution) as stipulated above
(this exhibit, Section 12.9.2).
13.2 Inject a 2-uL aliquot of the extract into the GC, operated under the
D-59
-------
conditions previously used (this exhibit, Section 6.2) to produce accept-
able results with the performance check solution.
13.3 Acquire SIM data according to Section 6.1.3 (this exhibit). Use the same
acquisition and mass spectrometer operating conditions previously used to
determine the relative response factors (this exhibit, Sections 9.1.4.6
through 9.1.4.9). Ions characteristic for polychlorinated diphenyl
ethers are included in the descriptors listed in Table 6. Their presence
is to monitor their interference during the characterization of PCDFs.
NOTE: The acquisition period must at least encompass the PCDD/PCDF
overall retention time window previously determined (Section 8.1, this
exhibit). Selected ion current profiles (SICP) for the lock-mass ions
(one per mass descriptor) must also be recorded and included in the data
package as deliverables. These SICPs must be true representations of the
evolution of the lock-mass ions amplitudes during the HRGC/HRMS run.
(See this exhibit, Section 8.2.2 for the proper level of reference compound
to be metered into the ion chamber.) It is recommended to examine the
lock-mass ion SICP for obvious basic sensitivity and stability changes
of the instrument during the GC/MS run that could affect the measurements
[Y. Tondeur et al., Anal. Chem. 56, 1344 (1984)]. Report any discrepancies
in the case narrative.
13.4 Identification Criteria
For a gas chromatographic peak to be identified as a PCDD or PCDF, it
D-60
o r>
OU
-------
must meet all of the following criteria:
13.4.1 Relative Retention Times.
13.4.1.1 For 2,3,7,8-substituted congeners, which have an isotopically labeled
internal or recovery standard present in the sample extract (this
represents a total of 10 congeners including OCDD; Tables 2 and 3),
the relative retention time (RRT; at maximum peak height) of the
sample components (i.e., the two ions used for quantification purposes
listed in Table 6) must be within -1 and +3 seconds of the retention
time of the peak for the isotopically labeled internal or recovery
standard at m/z corresponding to the first characteristic ion (of the
set of two; Table 6) to obtain a positive identification of these
nine 2,3,7,8-substituted PCDDs/PCDFs and OCDD.
13.A.1.2 For 2,3,7,8-substituted compounds, that do not have an isotopically
labeled internal standard present in the sample extract (this repre-
sents a total of six congeners; Table 3), the relative retention time
must fall within the established homologous retention time windows by
analyzing the column performance check solution (this exhibit, Section
8.1.3). Identification of OCDF is based on its retention time rela-
tive to ^Cj2~°CDD as determined from the daily routine calibration
results.
13.4.1.3 For non-2,3,7,8-substituted compounds (tetra through octa; totaling
119 congeners), the retention time must be within the corresponding
D-61
308
-------
homologous retention time windows established by analyzing the column
performance check solution (this exhibit, Section 8.1.3).
13.4.1.4 The ion current responses for both ions used for quantitative pur-
poses (e.g., for TCDDs: m/z 319.8465 and 321.8936) must reach maximum
simultaneously (+ 2 seconds).
13.4.1.5 The ion current responses for both ions used for the labeled stan-
dards (e.g., for 13C12-TCDD: m/z 331.9368 and m/z 333.9339) must
reach maximum simultaneously (+ 2 seconds).
NOTE: The analyst is required to verify the presence of 1,2,8,9-TCDD
and 1,3,4,6,8-PeCDF (this exhibit, Section 8.1.3) in the SICPs of the
daily performance checks. Should either one compound be missing, the
analyst is required to report that observation with the results
associated with the sample batch as it may indicate a potential
problem with the ability to detect all the PCDDs/PCDFs.
\
13.4.2 Ion Abundance Ratios
13.4.2.1 The integrated ion current for the two ions used for quantification
purposes must have a ratio between the lower and upper limits
established for the homologous series to which the peak is assigned.
See Sections 9.1.4.3 and 9.1.4.4 (this exhibit) and Table 9 for
details.
D-62
3'
-------
13.4.3 Signal-to-Noise Ratio
13.4.3.1 All ion current intensities must be £ 2.5 times noise level for posi-
tive identification of a PCDD/PCDF compound or a group of coeluting
isomers. Appendix C describes the procedure to be followed for the
determination of the S/N.
13.4.4 Polychlorinated Diphenyl Ether Interferences
13.4.4.1 In addition to the above criteria, the identification of a GC peak as
a PCDF can only be made if no signal having a S/N >^ 2.5 is detected,
at the same retention time (+ 2 seconds), in the corresponding PCDPE
channel.
14. Calculations
14.1 For gas chromatographic peaks that have met the criteria outlined in
Sections 13.4.1.1 through 13.4.3.1 (this exhibit), calculate the concen-
tration of the PCDD or PCDF compounds using the formula:
Ax x Qls
Als x W x RRF(n)
where
concentration of unlabeled PCDD/PCDF congeners (or group of
coeluting isomers within an homologous series) in pg/g,
D-63
n -« ,}
Ol(J
-------
Ax = sum of the integrated ion abundances of the quantification
ions (Table 6) for unlabeled PCDDs/PCDFs,
Ais = sum °f *-he integrated ion abundances of the quantification ions
(Table 6) for the labeled internal standards,
Qis = quantity, in pg, of the internal standard added to the sample
before extraction,
W «• weight, in g, of the sample (solid or liquid), and
RRF(n) = calculated mean relative response factor for the analyte
[RRF(n) with n = 1 to 17; Section 9.1.4.7, this exhibit].
If the analyte is identified as one of the 2,3,7,8-substituted PCDDs
or PCDFs , RRF(n) is the value calculated using the equation in Section
9.1.4.7 (this exhibit). However, if it is a non-2,3,7,8-substituted
congener, the RRF(k) value is the one calculated using the equation in
Section 9.1.4.8.2 (this exhibit). [RRF(k) with k = 27 to 30.]
14.2 Calculate the percent recovery of the nine internal standards measured in
the sample extract, using the formula:
* Qrs
Internal standard percent recoverr = — x 100
Qis x Ars x
where
D-64
311
-------
sum of the integrated Ion abundances of the quantification
ions (Table 6) for the labeled internal standard,
Ars = sum of the integrated ion abundances of the quantification
ions (Table 6) for the labeled recovery standard; the selection
of the recovery standard depends on the type of congeners (see
Table 5, footnotes),
Qis ° quantity, in pg, of the internal standard added to the sample
before extraction,
Qrs » quantity, in pg, of the recovery standard added to the
cleaned-up sample residue before HRGC/HRMS analysis, and
RRF(m) = calculated mean relative response factor for the labeled
internal standard relative to the appropriate (see Table 5,
footnotes) recovery standard. This represents the mean
obtained in Section 9.1.4.9 (this exhibit) [RRF(m) with
m = 18 to 26].
NOTE: For human adipose tissue, adjust the percent
recoveries by adding 1 percent to the calculated value.
14.3 If the concentration in the 10-uL or 50-uL final extract of any of the
fifteen 2,3,7,8-substituted PCDD/PCDF compounds (Table 3) exceeds the
upper method calibration limits (MCL) listed in Table 1 (e.g., 200 pg/uL
D-65
312
-------
for TCDD in soil), the linear range of response versus concentration may
have been exceeded, and, after contacting EPA/SHO, a reanalysis of the
sample (using one tenth aliquot) should be undertaken. The volumes of
the internal and recovery standard solutions should remain the same as
described for the sample preparation (this exhibit, Sections 12.1 to
12.9.3). For the other congeners (including OCDD), however, report the
measured concentration and Indicate that the value exceeds the MCL.
14.4 The total concentration for each homologous series of PCDD and PCDF is
calculated by summing up the concentrations of all positively identified
isomers of each homologous series. Therefore, the total should also
include the 2,3,7,8-substituted congeners. The total number of GC
signals included in the homologous total concentration value must be
specified in the report.
14.5 Sample-Specific Estimated Detection Limit
The sample-specific estimated detection limit (EDL) is the concentration
of a given analyte required to produce a signal with a peak height of at
least 2.5 times the background signal level. An EDL is calculated for
each 2,3,7,8-substituted congener that is not identified, regardless of
whether or not other non-2,3,7,8-substituted isomers are present. Two
methods of calculation can be used, as follows, depending on the type of
response produced during the analysis of a particular sample.
14.5.1 Samples giving a response for both quantification ions (Tables 6 and 9)
that is less than 2.5 times the background level.
D-66
3
13
-------
14.5.1.1 Use the expression for EDL (specific 2,3,7,8-substltuted PCDD/PCDF)
below to calculate an EDL for ejach absent 2,3,7,8-substltuted PCDD/
PCDF (I.e., S/N < 2.5). The background level Is determined by
measuring the range of the noise (peak to peak) for the two quanti-
fication ions (Table 6) of a particular 2,3,7,8-substituted isomer
within an homologous series, in the region of the SICP trace
corresponding to the elution of the internal standard (if the congener
possesses an internal standard) or in the region of the SICP where
the congener is expected to elute by comparison with the routine
calibration data (for those congeners that do not have a 13c-iabeie(j
standard), multiplying that noise height by 2.5, and relating the
product to an estimated concentration that would produce that product
height.
Use the formula:
2.5 x Ax x Qj_s
EDL (specific 2,3,7,8 subst.-PCDD/PCDF) •
Als x W x RRF(n)
where
EDL = estimated detection limit for homologous 2,3,7,8-substituted
PCDDs/PCDFs.
Ax, Ajs, W, RRF(n), and Q^s retain the same meanings as defined
in Section 14.1.
D-67
314
-------
14.5.2 Samples characterized by a response above the background level with a
S/N of at least 2.5 for at least one of the quantification ions
(Tables 6 and 9).
14.5.2.1 When the response of a signal having the same retention time as a
2,3,7,8-substituted congener has a S/N in excess of 2.5 and does not
meet any of the other qualitative identification criteria listed in
Section 13.4, calculate the "Estimated Maximum Possible Concentration'
(EMPC) according to the expression shown in Section 14.1.
14.6 The relative percent difference (RPD) is calculated as follows:
I Si - S2 |
RPD = x 100
( Si + S2 ) / 2
Si and 82 represent sample and duplicate sample results.
14.7 The 2,3,7,8-TCDD toxic equivalents (TE) of PCDDs and PCDFs present in the
sample are calculated, only at the data user's request, according to the
method recommended by the Chlorinated Dioxins Workgroup (CDWG) of the EPA
and the Center for Disease Control (CDC). This method assigns a 2,3,7,8-
TCDD toxicity equivalency factor (TEF) to each of the fifteen 2,3,7,8-
substituted PCDDs and PCDFs (Table 3) and the non-2,3,7,8-substituted
. compounds as shown in Table 11. The 2,3,7,8-TCDD equivalent of the PCDDs
and PCDFs present in the sample is calculated by summing the TEF times
their concentration for each of the compounds or groups of compounds
D-68
315
-------
listed In Table 11. The exclusion of other homologous series such as
mono-, di-, tri- and octachlorinated dibenzodioxins and dibenzofurans
does not mean that they are non-toxic. Their toxicity, as known at this
time, is much less than the toxicity of the compounds listed in Table 11.
The above procedure for calculating the 2,3,7,8-TCDD toxic equivalents is
not claimed by the CDUG to be based on a thoroughly established scientific
foundation. The procedure, rather, represents a "Consensus recommendation
on science policy". Since the procedure may be changed in the future,
reporting requirements for PCDD and PCDF data would still include the
reporting of the analyte concentrations of the PCDD/PCDF congener as
calculated in Sections 14.1 and 14.4.
14.7.1 Two-GC Column TEF Determination
Isomer specificity for all 2,3,7,8-substituted PCDDs/PCDFs cannot be
achieved on the 60-m DB-5 GC column alone. In order to determine the
proper concentrations of the individual 2,3,7,8-substituted congeners,
the sample extract must be reanalyzed on a 60-m SP-2330 (or SP-2331) GC
column.
14.7.1.1 The concentrations of 2,3,7,8-TCDD (see note below), 2,3,4,7,8-PeCDF,
l,2,3,4,6,7,8HpCDD, 1,2,3,4,6,7,8-HpCDF, and 1,2,3,4,7,8,9-HpCDF are
calculated from the analysis of the sample extract on the 60-m DB-5
fused-silica column. The experimental conditions remain the same as the
conditions described previously in Section 13 (this exhibit), and the
calculations are performed as outlined in Section 14 (this exhibit).
D-69
316
-------
14.7.1.2 The concentrations of 2,3,7,8-TCDF, 1,2,3,7,8-PeCDD and -PeCDF,
1,2,3,4,7,8-HxCDD and -HxCDF, 1,2,3,6,7,8-HxCDD and -HxCDF,
1,2,3,7,8,9-HxCDD and -HxCDF, and 2,3,4,6,7,8-HxCDF are obtained from
the analysis of the sample extract on the second fused-silica capil-
lary column (confirmation GC column: 60 m SP-2330). However, the
GC/MS conditions must be altered so that: (1) only the first three
descriptors (i.e., tetra-, penta-, and hexachlorinated congeners)
of Table 6 are used; and (2) the switching time between descriptor 2
(pentachlorinated congeners) and descriptor 3 (hexachlorinated
congeners) takes place following the elution of ^C ~!»2>3,7,8-PeCDD.
The concentration calculations are performed as outlined in Section
14 (this exhibit).
NOTE: The confirmation and quantification of 2,3,7,8-TCDD (this
exhibit, Section 14.7.1.1) may be accomplished on the SP-2330 GC
column instead of the DB-5 column, provided the criteria listed in
Section 8.1.2 (this exhibit) are met and the requirements described
in Section 2.2 (Exhibit E) are followed.
14.7.1.3 For a gas chromatographic peak to be identified as a 2,3,7,8-
substituted PCDD/PCDF congener, it must meet the ion abundance and
signal-to-noise ratio criteria listed in Sections 13.4.2 and 13.4.3
(this exhibit), respectively. In addition, the retention time
identification criterion described in Section 13.4.1.1 (this exhibit)
applies here for congeners for which a carbon-labeled analogue is
available in the sample extract. However, the relative retention
D-70
317
-------
time (RRT) of the 2,3,7,8-substituted congeners for which no carbon-
labeled analogues are available must fall within 0.006 units of the
carbon-labeled standard RRT. Experimentally, this is accomplished by
using the attributions described in Table 12 and the results from the
routine calibration run on the SP-2330 column.
D-71
318
-------
APPENDICES
318
-------
APPENDIX A
Procedure for the Collection, Handling, Analysis, and Reporting
Requirements of Wipe Tests Performed within the Laboratory
D-72
320
-------
This procedure Is designed for the periodic evaluation of potential con-
tamination by 2,3,7,8-substituted PCDD/PCDF congeners of the working areas
inside the laboratory.
PERFORMING WIPE TEST
Perform the wipe tests on surface areas of two inches by one foot with
laboratory wipers saturated with distilled-in-glass acetone using a pair of
clean stainless steel forceps. Use one wiper for each of the designated areas.
Combine the wipers to one composite sample in an extraction jar containing 200
mL distilled-in-glass acetone. Place an equal number of unused wipers in 200
mL acetone and use this as a control.
COMPOSITE SAMPLE PREPARATION
Close the jar containing the wipers and 200 mL acetone and extract for 20
minutes using a wrist-action shaker. Transfer the extract into a KD apparatus
fitted with a concentration tube and a three-ball Snyder column. Add two
Teflon™ or Carborundum™ boiling chips and concentrate the extract to an apparent
volume of 1.0 mL on a steam bath. Rinse the Snyder column and the KD assembly
with two 1-mL portions of hexane into the concentrator tube. Add 100 uL of the
sample fortification solution to the concentrator tube (Section 3.8, this
exhibit), and concentrate its contents to near dryness with a gentle stream of
nitrogen. Add 1.0 mL hexane to the concentrator tube, and swirl the solvent on
the walls.
D-73
o f
9 I
>-* .L
-------
Prepare a neutral alumina column as described in Section 12.7 (this
exhibit) and follow the steps outlined in Sections 12.8 thru 12.8.2 (this
exhibit).
Add 10 uL of the recovery standard solution as described in Section
12.9.2 (this exhibit).
EXTRACT ANALYSIS
Concentrate the contents of the vial to a final volume of 10 uL (either in
a minivial or in a capillary tube). Inject two uL of each extract (wipe and
control) onto a capillary column and analyze for 2,3,7,8-substituted PCDDs/PCDFs
as specified in the analytical method Section 13 (this exhibit). Perform
calculations according to Section 14 (this exhibit).
REPORTING FORMAT
Report the presence of 2,3,7,8-substituted PCDDs and PCDFs as a quantity
(pg or ng) per wipe test experiment (WTE). Under the conditions outlined in
this analytical protocol, a lower limit of calibration of 25 pg/WTE is expected
for 2,3,7,8-TCDD. A positive response for the blank (control) is defined as a
signal in the TCDD retention time window at any of the masses monitored which
is equivalent to or above 8 pg of 2,3,7,8-TCDD per WTE. For other congeners,
use the multiplication factors listed in Table 1, footnote (a) (e.g., for OCDD,
the lower MCL is 25 x 5 = 125 pg/WTE and the positive response for the blank
D-74
322
-------
would be 8 x 5 = 40 pg). Also, report the recoveries of the internal standards
during the simplified cleanup procedure.
FREQUENCY OF WIPE TESTS
At a minimum, wipe tests should be performed when there is evidence of
contamination in the method blanks.
CORRECTIVE ACTION
An upper limit of 25 pg per TCDD isomer and per wipe test experiment is
allowed. (Use multiplication factors listed in footnote (a) from Table 1 for
other congeners.) This value corresponds to the lower calibration limit of the
analytical method. Steps to correct the contamination must be taken whenever
these levels are exceeded. To that effect, first vacuum the working places
(hoods, benches, sink) using a vacuum cleaner equipped with a high-efficiency
particulate absorbent (HEPA) filter and then wash with a detergent. A new set
of wipes should be analyzed before anyone is allowed to work in the dioxin area
of the laboratory.
D-75
323
-------
APPENDIX B
Standards Traceabillty Procedure
NOTE: The content of this appendix is based on the assumption that EPA
will have within its repository a mixture (named S2) containing known
concentrations (e.g., 100 pg/uL) of the eight 13C-labeled 2,3,7,8-substi-
tuted PCDD/PCDF congeners marked with an asterisk in Table 3 of this
exhibit, and a second solution (named SI, with the same concentration as
used for S2) containing the eight corresponding unlabeled analogues.
D-76
,24
-------
All laboratories are expected to maintain traceablllty of their standard
solutions by verifying that all standard solutions used for direct quantifica-
tion of samples agree in chemical Identity and concentration with the EPA
primary standard solutions. The specific procedures are described below:
Each time a new laboratory working standard solution (W) is prepared, the
identities and concentrations of the components of this solution must be veri-
fied. Verifications of the identities of the compounds are to be carried out
by HRGC/HRMS. The EPA reference standard (S) and the laboratory working stan-
dard (W) are to be analyzed under the instrumental conditions described in this
exhibit, which are appropriate for the analysis of PCDDs and PCDFs. Two
criteria must be satisfied to verify the identifications:
o Elution of the component(s) of the laboratory working standard must
be at the same retention time(s) as those of the component(s) of the
EPA reference standard solution.
o Concentration^ ) of the laboratory working standard component(s) must
be equal to or less than 20 percent different from the EPA reference
standard component(s).
Qualitative Characterization
Due to the complexity brought by the large number of possible PCDD and
PCDF congeners, the requirement for qualitative verification by comparison of
the retention times applies only to the eight 2,3,7,8-substituted PCDD/PCDF
D-77
^ (s
G£
-------
congeners marked with an asterisk in Table 3 and for which a carbon-labeled
analogue is available. Two situations need to be considered:
a) The laboratory Is required to trace back Its unlabeled PCDD/PCDF standards
to EPA standards. This is accomplished by adding an appropriate aliquot
of the EPA l^c-iabeied standard solution (S2) to an aliquot of the labora-
tory working solution (VI) so that the concentrations are comparable; the
new mixture is then analyzed by HRGC/HRMS. The retention times of the
eight unlabeled PCDDs/PCDFs discussed above must fall within -1 to +3
seconds of the EPA 13c_iat,eie(j analogues.
b) In addition to a), the laboratory is required to trace back its
labeled standards to EPA standards. Proceed as follows: Add an aliquot
of the laboratory working standard solution (W2) containing the carbon-
labeled compounds to an aliquot of the EPA standard solution (SI) containing
the eight unlabeled 2,3,7,8-substituted PCDD/PCDF congeners discussed
above, and analyze by HRGC/HRMS. The concentrations must be comparable.
The retention times for the eight carbon-labeled compounds must fall
within -3 to +1 seconds of the EPA unlabeled analogues.
Quantitative Characterization
To establish that the concentration of the laboratory working standard is
correct with respect to the EPA reference standard, the relative response
factors (RRFs) for the eight 2,3,7,8-substituted PCDD/PCDF congeners (marked
with asterisks In Table 3) must be determined as described in this exhibit.
D-78
326
-------
The concentrations of the EPA reference and laboratory working standards should
be approximately the same (e.g., 50 pg/uL/congener). Proceed as follows:
1) Mix equal portions of the two EPA standard solutions (SI and S2) and
analyze by HRGC/HRMS. Calculate two RRFs for each of the eight analytes
as shown below:
Response factor of unlabeled congener (i) relative to carbon-labeled
analogue (j):
AI x
RRF (Sl,i) =
Qi x AJ
Response factor of carbon-labeled congener (j) relative to unlabeled
analogue (i):
AJ x
RRF (S2,j)
Qj
where A^ and AJ represent the integrated ion abundances of, respectively,
the unlabeled congener and carbon-labeled congener, and Q^ and Q-j the
quantities of, respectively, the unlabeled congener and carbon-labeled
congener, with i = 1 to 8, j = 1 to 8.
D-79
327
-------
2) Add an appropriate aliquot of the laboratory working solution Wl (or W2)
to an aliquot of the EPA solution S2 (or SI). Analyze the mixture by
HRGC/HRMS and calculate the corresponding response factors as indicated
below:
Ai x Q-j
RRF (Wl,i) =
Qi x Aj
or
AJ x Qi
RRF (W2,j) =
Qj x Ai
A and Q have the same meanings as in (1).
3) When the percent difference between each congener relative response factor
— RRF (Sl,i) and RRF (Wl,i), and RRF (S2,j) and RRF (W2,j) — does not
exceed 20 percent, the concentration of the laboratory working standard is
correct. (RPD = relative percent difference.)
| RRF (Sl,i) - RRF (Wl,i) |
RPD = • x 100
RRF (SI, i)
and
| RRF (S2,j) - RRF(W2,j) |
RPD = • x 100
RRF (S2,j)
D-80
0
8
-------
Traceability Requirements
If any or all of the above conditions for qualitative and quantitative
verifications for the laboratory working standard are not met, the standard is
not traceable to the EPA reference standard and can therefore not be used for
the analysis of samples.
NOTE: The procedure outlined above is required for laboratories which use
different batches of analytical standard compounds in the preparation of
the sample fortification and recovery standard solutions and in the prepara-
tion of the HRCC solutions. Laboratories which use the same batch of
analytical standards during the preparation of the sample fortification
and recovery standard solutions and the HRCC solutions are exempt from
following the above procedure, provided proper traceability documentation
is available.
In addition, the records pertaining to the above qualitative and
quantitative requirements, records of all verifications, documentation of the
preparation, and all inventory must be kept for all contract laboratory pri-
mary, secondary, and working standards that are generated for the purpose of
analyzing samples for EPA. These records should include the signed and dated
logbooks containing the information pertaining to the preparation of the
laboratory standards (weight of compound(s), volume and nature of the solvent,
laboratory code name, EPA reference standard lot number) and of any modification
made to the EPA reference standard. All standards should be used on a first
in, first out basis. The raw data, quantification reports and calculations
must be kept on file.
D-81
329
-------
APPENDIX C
Signal-to-Nolse Ratio Determination
D-82
330
-------
SIGNAL-TO-NOISE RATIO DETERMINATION
MANUAL DETERMINATION
This method describes a manual determination of the signal-to-noise ratio
(S/N) from a GC/MS signal, based on the measurement of its peak height relative
to the baseline noise. The procedure is composed of four steps as outlined
below. (Refer to Figure 7 for the following discussion.)
1. Estimate the peak-to-peak noise (N) by tracing the two lines (El and E2)
defining the noise envelope. The lines should pass through the estimated
statistical mean of the positive and the negative peak excursions as shown
on Figure 7. In addition, the signal offset (0) should be set high enough
such that negative-going noise (except for spurious negative spikes) is
recorded.
2. Draw the line (C) corresponding to the mean noise between the segments
defining the noise envelope.
3. Measure the height of the GC/MS signal (S) at the apex of the peak relative
to the mean noise C. For noisy GC/MS signals, the average peak height
should be measured from the estimated mean apex signal D between E3 and
E4.
D-83
331
-------
4. Compute the S/N.
This method of S/N measurement is a conventional, accepted method of noise
measurement in analytical chemistry.
D-84
09
33
-------
FIGURES CAPTIONS
1. Method flow chart for sample extraction and cleanup as used for the
analysis of PCDDs and PCDFs In complex waste and biological samples.
2. General structures of dibenzodioxin and dibenzofuran.
3. Peak profile displays demonstrating the effect of the detector zero on the
measured resolving power. In this example, the true resolving power is
5,600.
A) The zero was set too high; no effect is observed upon the measurement
of the resolving power. (Not aesthetic.)
B) The zero was adjusted properly.
C) The zero was set too low; this results in overestimating the actual
resolving power because the peak-to-peak noise cannot be measured
accurately.
4. Typical 12-hour analysis sequence of events.
5. Selected ion current profile for m/z 322 (TCDDs) produced by MS analysis
of the GC performance check solution on a 60-m DB-5 fused-silica capillary
column under the conditions listed in Table 7.
D-85
333
-------
I DRAFT
— ' ~^Z^_-
){
6. Peak profiles representing two "ffTC'TtifUluuLe -Iwnd at m/z 305 and 381. The
resolution of the high-mass signal is 95 ppm at 5 percent of the peak
height; this corresponds to a resolving power M/AM of 10,500 (10 percent
valley definition).
7. Manual determination of S/N.
The peak height (S) is measured between the mean noise (lines C and D).
These mean signal values are obtained by tracing the line between the
baseline average noise extremes, El and E2, and between the apex average
noise extremes, E3 and E4, at the apex of the signal. Note, it is
imperative that the instrument interface amplifier electronic zero offset
be set high enough such that negative-going baseline noise is recorded.
D-86
33-1
-------
Complex
Waste
Sample
Soil/
Sediment
oistur
Fish and
Adipose
Tissues
1) Internal
Standards
2)Extraction
Sample Extract
1) Acid-Base Cleanup
2)Chromatographic Cleanup
3) Recovery Standards
HRGC/HRMS
Figure 1
D-87
335
-------
8
0
Dibenzodioxin
8
0
Diben zof ura n
Figure 2
D-88
33G
-------
M/AM
5,600
B
5,600
Moo
8,550
Figure 3
D-89
.3'
-------
Analytical Procedure
8:00 AM
I
vo
rO
CO
Oc
Mass Resolution
Mass Accuracy
Thaw Sample Extract
1
Concentrate to 10 uL
9:00 AM
initial or
Routine
Calibration
GC Column
Performance
11:00 AM
Method
Blank
8:00 PIVi
Mass
Resolution
'Routi
ne
Figure A
-------
o
CO
CO
*ta±soAv/A~*~»*-W
Figure 5
-------
O
o
i
VO
Ref. mass 304.9824
Span. 200 ppm
System file name
Data file name
Resolution
Group number
lonization mode
Switching
Ref. masses
Peak top
YVES150
A:85Z567
10000
1
EI+
VOLTAGE
304.9824
380.9260
M/M A M~10.500
Channel B 380.9260 Lock mass
Span 200 ppm
Figure 6
-------
Go
o
vo
100-n
90-
80-
70-
60-
50-
40-
30-
20-
10-
20:00
117
N
= 19.5
D
22:00
24:00
26:00
28:00
30:00
Figure 7
-------
Table 1. Types of Mat
Method Calibi
Soil Fly
Sediment Ash
' DRAFT 1
B^^"-'Sgrr^ ^iT? anflf?, ^T^-Trnn-RaaoH
ration Limits" (ParL-s---p«t Trillion)
Fish
Tissue
Sludges Still- Paper
Water Fuel Oil Bottom Pulp
Lower MCL 2.5 2.5 0.025 12.5 25 2.5
Upper MCL 200 200
Weight (g) 10 10
2 1000 2000 200
1000 2 1 10
Human
Adipose
Tissue
2.
200
10
5
IS Spiking
Levels (ppt) 100 100
Final Extr.
Vol. (uL) 10 50
10
500 1000 100 100
50 50 10 10
(a>For other congeners multiply the values by 1 for TCDF/PeCDD/
PeCDF, by 2.5 for HxCDD/HxCDF/HpCDD/HpCDF, and by 5 for OCDD/OCDF.
NOTE: Chemical reactor residues are treated as still-bottoms if
their appearances suggest so.
D-94
342
-------
Table 2. Composition of the Sample Fortification
and Recovery Standard Solutions
Analyte
Sample Fortification
Solution
Concentration
(pg/uL; Solvent:
Isooctane)
Recovery Standard
Solution
Concentration
(pg/uL; Solvent:
Tridecane)
13C12-2,3,7,8-TCDD
J3C12-2,3,7,8-TCDF
1JC12-1,2,3,4-TCDD
10
10
50
1JC12-l,2,3,7,8-PeCDD
13C12-l,2,3,7,8-PeCDF
13C,2-l,2,3,6,7,8-HxCDD
}3C12-l,2,3,4,7,8-HxCDF
1JC12-l,2,3,7,8,9-HxCDD
13C12-l,2,3,4,6,7,8-HpCDD
13C12-1,2,3,4,6,7,8-HPCDF
13C12-OCDD
10
10
25
25
25
25
50
50
D-95
o /? '1
-------
Table 3. The Fifteen 2,3 ,7 ,8-Substltuted PCDD and PCDF Congeners
PCDD PCDF
2,3,7,8-TCDD(*> 2,3,7,8-TCDF(*>
l,2,3,7,8-PeCDD(*> 1,2, 3,7,8-PeCDF<*)
l,2,3,6,7,8-HxCDD(*) 2, 3,4,7,8-PeCDF
1,2,3,4,7,8 -HxCDD 1 , 2 , 3 , 6 , 7 , 8 -HxCDF
1,2,3,7,8, 9-HxCDD ( +) 1,2,3,7,8, 9-HxCDF
l,2,3,4,6,7,8-HpCDD(*> l,2,3,4,7,8-HxCDF(*>
2,3,4,6,7,8-HxCDF
l,2,3,4,6,7,8-HpCDF<*)
1,2,3,4,7,8,9-HpCDF
(*)xhe l^c-iabeled analogue is used as an internal standard.
13c_].abej.e(j analogue is used as a recovery standard.
D-96
344
-------
Table 4. Isomers of Chlorinated Dioxlns and Furans as a
Function of the Number of Chlorine Atoms
Number of
Chlorine
Atoms
Number of
Dioxin
Isomers
Number of
2,3,7,8
Isomers
Number of
Furan
Isomers
Number of
2,3,7,8
Isomers
1
2
3
4
5
6
7
8
2
10
14
22
14
10
2
1
1
1
3
1
1
4
16
28
38
28
16
4
1
1
2
4
2
1
Total
75
135
10
D-97
45
-------
Table 5. High-Resolution Concentration Calibration Solutions
Concentration (pg/uL)
Compound HRCC
Unlabeled Analytes
2,3,7,8-TCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDD
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,4,6,7,8-HpCDD
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDD 1 ,
OCDF ' 1,
Internal Standards
J3c12-2,3,7,8-TCDD
};?C12-2,3,7,8-TCDF
|3C12-l,2,3,7,8-PeCDD
|3c12-l,2,3,7,8-PeCDF
1JC12-l,2,3,6,7,8-HxCDD
|3Cl2-l,2,3,4,7,8-HxCDF
;3Cl2-l,2,3,4,6,7,8-HpCDD
;3Cl2-l,2,3,4,6,7,8-HpCDF
C12-OCDD
Recovery Standards
13C12-l,2,3,4-TCDD(a)
13C12-1,2.3,7,8,9-
HxCDD>
7
200
200
200
200
200
500
500
500
500
500
500
500
500
500
500
000
000
50
50
50
50
125
125
125
125
250
50
125
6
100
100
100
100
100
250
250
250
250
250
250
250
250
250
250
500
500
50
50
50
50
125
125
125
125
250
50
125
5
50
50
50
50
50
125
125
125
125
125
125
125
125
125
125
250
250
50
50
50
50
125
125
125
125
250
50
125
4
25
25
25
25
25
62.5
62.5
62.5
62.5
62.5
62.5
62.5
62.5
62.5
62.5
125
125
50
50
50
50
125
125
125
125
250
50
125
3
10
10
10
10
10
25
25
25
25
25
25
25
25
25
25
50
50
50
50
50
50
125
125
125
125
250
50
125
2
5
5
5
5
5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
25
25
50
50
50
50
125
125
125
125
250
50
125
1
2.5
2.5
2.5
2.5
2.5
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
12.5
12.5
50
50
50
50
125
125
125
125
250
50
125
(a)tJsed for recovery determinations of TCDD, TCDF, PeCDD and PeCDF
internal standards.
(b)llsed for recovery determinations of HxCDD, HxCDF, HpCDD, HpCDF,
and OCDD internal standards.
D-98
346
-------
Table 6. Ions Monitored for HRGC/HRMS analysis of PCDD/PCDFs
( S = Internal/recovery standard)
Descriptor Accurate(a)
Mass
1 303.9016
305.8987
315.9419
317.9389
319.8965
321.8936
331.9368
333.9339
375.8364
[354.9792]
2 339.8597
341.8567
351.9000
353.8970
355.8546
357.8516
367.8949
369.8919
409.7974
[354.9792]
Ion
ID
M
M+2
M
M+2
M
M+2
M
M+2
M+2
LOCK
M+2
M+4
M+2
M+4
M+2
M+4
M+2
M+4
M+2
LOCK
Elemental
Composition
C12H435C140
C12H435C1337C10
13C12H435C140
13C12H435C1337C10
C12H435cl4°2
C12H435C1337C102
13C12H435C14°2
13C12H435C1337C102
C12H435C160
C9F13
C12H335C1437C10
C12H335C1337C120
13C12H335C1437C10
13r H 35r1 37ri n
o i ^rio v> J. o \s*.*t\J
C12H335C1437C102
C12H335C1337C1202
13C12H335C1437C102
13C12H335C1337C1202
C12H335cl7°
CqF13
Analyte
TCDF
TCDF
TCDF (S)
TCDF (S)
TCDD
TCDD
TCDD (S)
TCDD (S)
HxCDPE
PFK
PeCDF
PeCDF
PeCDF (S)
PeCDF (S)
PeCDD
PeCDD
PeCDD (S)
PeCDD (S)
HpCDPE
PFK
(Continued)
D-99
O
A '-)
'i i
-------
Table 6. Continued
Descriptor Accurate
Mass
3 373.8208
375.8178
383.8642
385.8610
389.8156
391.8127
401.8559
403.8529
445.7555
[354.9792]
4 407.7818
409.7789
417.8253
419.8220
423.7766
425.7737
435.8169
437.8140
479.7165
[430.9728]
Ion
ID
M+2
M+4
M
M+2
M+2
M+4
M+2
M+4
M+4
LOCK
M+2
M+4
M
M+2
M+2
M+4
M+2
M+4
M+4
LOCK
Elemental
Composition
C12H235C1537C10
C12H235C1437C120
13C12H235C160
13C12H235C1537C10
C12H235C1537C102
C12H235C1437C1202
13C12H235C1537C102
13C H 35C1 37C1 0
C12H235C1637C120
C9F13
C12H35C1637C10
C12H35C1537C120
13C12H35C170
13C12H35C1637C10
C12H35C1637C102
C12H35C1537C1202
13C12H35C1637C102
13C12H35C1537C1202
C12H35C1737C120
C9F17
Analyte
HxCDF
HxCDF
HxCDF (S)
HxCDF (S)
HxCDD
HxCDD
HxCDD (S)
HxCDD (S)
OCDPE
PFK
HpCDF
HpCDF
HpCDF (S)
HpCDF (S)
HpCDD
HpCDD
HpCDD (S)
HpCDD (S)
NCDPE
PFK
(Continued)
D-100
348
-------
Table 6. Continued
Descriptor Accurate Ion
Mass ID
Elemental
Composition
Analyte
5 441.7428
443.7399
457.7377
459.7348
469.7779
471.7750
513.6775
[430.9728 ]
M+2
M+4
M+2
M+4
M+2
M+4
M+4
LOCK
c1235ci737cio
C1235ci637ci2o
C1235ci737cio2
C1235ci637ci2o2
13c1235ci737cio2
13c1235ci637ci2o2
C1235ci837ci2o
C9Fi7
OCDF
OCDF
OCDD
OCDD
OCDD (S)
OCDD (S)
DCDPE
PFK
(a)The following nuclidic masses were used:
H = 1.007825 0 = 15'. 994915
C - 12.000000 35Cl = 34.968853
13C = 13.003355 37C1 = 36.965903
D-101
349
-------
Table 7. Recommended GC Operating Conditions
Column coating DB-5
Film thickness 0.25 urn
Column dimension 60 m x 0.32 mm
Injector temperature 270° C
Splitless valve time 45 s
Interface temperature Function of the final temperature
Temperature program
Stage
1
2
3
In it. Temp. Inlt. Hold. Temp.
(° C) Time (min) Ramp
(° C/min)
200 2 5
5
5
Total
Fin. Temp
(° c)
220
235
330
time: 60
. Fin.
Hoi.
Time
16
7
5
min
D-102
350
-------
Table 8. PCDD and PCDF Congeners Present In the GC Performance
Evaluation Solution and Used for Defining the
Homologous GC Retention Time Windows on a
60-m DB-5 Column
No. of
Chlorine
Atoms
4<*>
5
PCDD-Positional Isomer
Early Late
Eluter Eluter
1,3,6,8 1,2,8,9
1, 2,4,6,8/ 1,2,3,8,9
PCDF-Positional
Early
Eluter
1,3,6,8
1,3,4,6,8
Isomer
Late
Eluter
1,2,8
1,2,3,8
,9
,9
6
7
8
1,2,4,7,9
1,2,3,4,6,8 1,2,3,4,6,7
1,2,3,4,6,7,8 1,2,3,4,6,7,9
1,2,3,4,6,7,8,9
1,2,3,4,6,8 1,2,3,4,8,9
1,2,3,4,6,7,8 1,2,3,4,6,7,9
1,2,3,4,6,7,8,9
(a)in addition to these two PCDD isomers, the 1,2,3,4-, 1,2,3,7-,
1,2,3,8-, 2,3,7,8-, 13C12-2,3,7,8-, and 1,2,3,9-TCDD isomers
must also be present.
D-103
351
-------
Table 9. Theoretical Ion Abundance Ratios and Their
Control Limits for PCDDs and PCDFs
Number of
Chlorine
Atoms
4
5
6
6(a)
7Used only for 13C-HxCDF (IS).
(b>Used only for 13C-HpCDF (IS).
D-104
-------
Table 10. Relative Response Factor [RRF (number)] Attributions
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
2,3,7
2,3,7
1,2,3
1,2,3
2,3,4
1,2,3
1,2,3
1,2,3
1,2,3
1,2,3
1,2,3
2,3,4
1,2,3
1,2,3
1,2,3
OCDD
OCDF
l^Cj-
J7C12
*-*r
W 1 rt
it 12
UP"
17 12
Up**
** 1 O
17 1 ^
l^r1
11 12
•Up
17 12
UP**
17 12
Up"
^11
Totl?
Total
Total
Total
Specific Congener Name
,8-TCDD (and total TCDDs)
,8-TCDF (and total TCDFs)
,7,8-PeCDD (and total PeCDDs)
,7,8-PeCDF
,7,8-PeCDF
,4,7,8-HxCDD
,6,7,8-HxCDD
,7,8,9-HxCDD
,4,7,8-HxCDF
,6,7,8-HxCDF
,7,8,9-HxCDF
,6,7,8-HxCDF
,4,6,7,8-HpCDD (and total HpCDDs)
,4,6,7,8-HpCDF
,4,7,8,9-HpCDF
-2,3,7,8-TCDD
-2,3,7,8-TCDF
-1,2,3,7,8-PeCDD
-1,2,3,7,8-PeCDF
-1,2,3,6,7,8-HxCDD
-1,2,3,4,7,8-HxCDF
-1,2,3,4,6,7,8-HpCDD
-1,2,3,4,6,7,8-HpCDF
-OCDD
PeCDFs
HxCDFs
HxCDDs
HpCDFs
D-105
i O O
-------
TABLE 11. 2,3,7,8-TCDD Equivalent Factors (TEFs) for the
Polychlorinated Dlbenzodioxins and Dibenzofurans
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Compound (s)
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,6,7,8-HxCDD
1, 2,3,7,8,9-HxCDD
1, 2,3,4,7,8-HxCDD
1, 2,3,4,6,7,8-HpCDD
* Total - TCDD
* Total - PeCDD
. * Total - HxCDD
* Total - HpCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
1,2,3,4,7,8-HpCDF
2,3,4,6,7,8-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
* Total - TCDF
* Total - PeCDF
* Total - HxCDF
* Total - HpCDF
TEF
1.00
0.50
0.04
0.04
0.04
0.001
0.01
0.005
0.0004
0.00001
0.10
0.10
0.10
0.01
0.01
0.01
0.01
0.001
0.001
0.001
0.001
0.0001
0.00001
*Excluding the 2,3,7,8-substituted congeners,
D-106
Q cr ,1
OU'i
-------
Table 12. Toxiclty Equivalency Factor: Analyte Relative
Retention Time Reference Attributions
Analyte Analyte RRT Reference(a
1,2,3,4,7,8-HxCDD 13C12-l,2,3,6,7,8-HxCDD
1,2,3,6,7,8-HxCDF 13C12~1,2,3,4,7,8-HxCDF
1,2,3,7,8,9-HxCDF 13C12-l,2,3,4,7,8-HxCDF
2,3,4,6,7,8-HxCDF 13C12-l,2,3,4,7,8-HxCDF
The retention time of 2,3,4,7,8-PeCDF on the DB-5 column is
measured relative to *3Cj2-l,3,7,8-PeCDF and the retention
time of 1,2,3,4,7,8,9-HpCDF relative to 13C12-1,2,3,4,6,7,8
HpCDF.
D-107
355
-------
QUALITY ASSURANCE REQUIREMENTS
(Quality Assessment and Quality Control)
(Exhibit E)
n r-11
0Gb
-------
DRAFT
1. SUMMARY OF QA/QC ANALYSES •**—»• I
o Initial and periodic calibration and instrument performance checks.
° HRGC/HRMS method blank analysis.
0 Field blank analyses (Section 2.4.2, this exhibit); a minimum of one
fortified field blank shall be analyzed with each sample batch; an
additional fortified field blank must be analyzed when a new lot of
absorbent or solvent is used. A matrix spike may be used in place of
a fortified field blank.
° Analysis of a batch of samples with accompanying QA/QC analyses:
Sample Batch — < 24 samples, including field blank and rlnsate
sample(s).
Additional QA/QC analyses per batch:
Fortified field blank or matrix spike (MS) 1
Method blank (MB) 1
Duplicate sample or matrix spike duplicate (MSD) 1
Total 3
0 "Blind" QC samples (soil, sediment, water) may be submitted to the
laboratory as ordinary samples included in the sample batch.
E-l
-------
Blind samples include:
Uncontaminated soil, sediment, or water samples
Split samples,
Unidentified duplicates, and
Performance evaluation samples.
2. QUALITY ASSESSMENT/QUALITY CONTROL
2.1 Performance Evaluation Samples -- Included among the samples in all
batches will be samples (blind or double blind) containing known amounts
of unlabeled 2,3,7,8-substituted PCDDs/PCDFs or other PCDD/PCDF congeners.
2.2 Performance Check Solutions
2.2.1 At the beginning of each 12-hour period during which samples are to be
analyzed, an aliquot of the 1) GC column performance check solution and
2) high-resolution concentration calibration solution No. 3 (HRCC-3)
shall be analyzed to demonstrate adequate GC resolution and sensitivity,
response factor reproducibility, and mass range calibration, and to
establish the PCDD/PCDF retention time windows. A mass resolution check
shall also be performed to demonstrate adequate mass resolution using an
appropriate reference compound (PFK is recommended).
These procedures are described in Section 8 of Exhibit D. If the
required criteria are not met, remedial action must be taken before any
E-2
O KT O
ouo
-------
samples are analyzed.
2.2.2 To validate positive sample data, the routine or continuing calibration
(HRCC-3) and the mass resolution check must be performed also at the end
of each 12-hour period during which samples are analyzed. Furthermore,
an HRGC/HRMS method blank run must be recorded following a calibration
run and the first sample run.
2.2.2.1 If the laboratory operates only during one period (shift) each day
of 12 hours or less, the GC performance check solution must be
analyzed only once (at the beginning of the period) to validate the
data acquired during the period. However, the mass resolution and
continuing calibration checks must be performed at the beginning as
well as at the end of the period.
2.2.2.2 If the laboratory operates during consecutive 12-hour periods (shifts),
analysis of the GC performance check solution must be performed at the
beginning of each 12-hour period. The mass resolution and continuing
calibration checks from the previous period can be used for the
beginning of the next period.
2.2.3 Results of at least one analysis of the GC column performance check
solution and of two mass resolution and continuing calibration checks
must be reported with the sample data collected during a 12-hour period.
2.2.4 Deviations from criteria specified for the GC performance check or for
E-3
9 c r
o u
-------
the mass resolution check (Section 8, Exhibit D) invalidate all positive
sample data collected between analyses of the performance check solu-
tion, and the extracts from those positive samples shall be reanalyzed
(Exhibit C).
If the routine calibration run fails at the beginning of a 12-hour shift,
the instructions in Exhibit D, Section 9.4.4 must be followed. If the
continuing calibration check performed at the end of a 12-hour period
fails by no more than 25 percent RPD, use the mean RRFs from the two
daily routine calibration runs to compute the analyte concentrations,
instead of the RRFs obtained from the initial calibration. A new
initial calibration (new RRFs) is required immediately (within two hours)
following the analysis of the samples, whenever the RPD from the end-
of-shift routine calibration exceeds 25 percent. Failure to perform a
new initial calibration immediately following the analysis of the
samples will automatically require reanalysis of all positive sample
extracts analyzed before the failed end-of-shift continuing calibration
check.
2.3 The GC column performance check mixture, high-resolution concentration
calibration solutions, and the sample fortification solutions may be
obtained from the EMSL-LV. However, if not available from the EMSL-LV,
standards can be obtained from other sources, and solutions can be pre-
pared in the laboratory. Concentrations of all solutions containing
2,3,7,8-substituted PCDDs/PCDFs, which are not obtained from the EMSL-
LV, must be verified by comparison with the EPA standard solutions that
E-4
,60
-------
are available from the EMSL-LV. (Refer to Appendix B, Exhibit D, for
details on the recommended standards traceability procedure.)
2.4 Blanks
2.4.1 Method Blank
One method blank is required per batch of samples. To that effect,
perform all steps detailed in the analytical procedure (Section 12,
Exhibit D) using all reagents, standards, equipment, apparatus, glass
ware and solvents that would be used for a sample analysis, but omit
addition of the soil, aqueous or any other matrix sample portion.
2.4.1.1 The method blank must contain the same amount of x C
internal standards that is added to samples before extraction.
2.4.1.2 An acceptable method blank exhibits no positive response as stated in
Section 3.16, Exhibit D. If the method blank, which was extracted
along with a batch of samples, is contaminated, all positive samples
must be rerun (Exhibit C).
2.4.1.2.1 If the above criterion is not met, check solvents, reagents, forti-
fication solutions, apparatus and glassware to locate and eliminate
the source of contamination before any further samples are extracted
and analyzed.
2.4.1.2.2 If new batches of reagents or solvents contain interfering
E-5
o f> \
ool
-------
contaminants, purify or discard them.
2.4.2 Field Blanks
Each batch of samples contains a field blank sample of uncontaminated
soil, sediment or water that is to be fortified before analysis accord-
ing to Section 2.4.2.1 (this exhibit). In addition to this field blank,
a batch of samples may include a rinsate, which is a portion of the sol-
vent (usually trichloroethylene) that was used to rinse sampling equip-
ment. The rinsate is analyzed to assure that the samples were not
contaminated by the sampling equipment.
2.4.2.1 Fortified Field Blank
2.4.2.1.1 Weigh a 10-g portion or use 1 L (for aqueous samples) of the speci-
fied field blank sample and add 100 uL of the solution containing
the nine internal standards (Table 2, Exhibit D) diluted with 1.5 mL
acetone (Section 12.1, Exhibit D).
2.4.2.1.2 Extract by using the procedures beginning in Sections 12.2.5 or
12.2.6 of Exhibit D, as applicable, add 10 uL of the recovery stan-
dard solution (Section 12.9.2, Exhibit D) and analyze a 2-uL aliquot
of the concentrated extract.
2.4.2.1.3 Calculate the concentration (Section 14.1, Exhibit D) of 2,3,7,8-
substituted PCDDs/PCDFs and the percent recovery of the internal
standards "(Section 14.2, Exhibit D). If the percent recovery at the
E-6
o f*'»
-------
measured concentration of any 2,3,7,8-substltuted PCDD/PCDF congener
Is <40 percent or >120 percent, report the results to SMO before
proceeding with the samples.
2.4.2.1.4 Extract and analyze a new simulated fortified field blank whenever
new lots of solvents or reagents are used for sample extraction or
for column chromatographic procedures.
2.4.2.2 Rinsate Sample
2.4.2.2.1 The rinsate sample must be fortified like a regular sample.
2.4.2.2.2 Take a 100-mL (+ 0.5 mL) portion of the sampling equipment rinse
solvent (rinsate sample), filter, if necessary, and add 100 uL of the
solution containing the nine internal standards (Table 2, Exhibit D).
2.4.2.2.3 Using a Kuderna-Danish appparatus, concentrate to approximately
5 mL.
2.4.2.2.4 Transfer the 5-mL concentrate from the K-D concentrator tube in 1-mL
portions to a 1-mL minivial, reducing the volume in the minivial as
necessary with a gentle stream of dry nitrogen.
2.4.2.2.5 Rinse the K-D concentrator tube with two 0.5-mL portions of hexane
and transfer the rinses to the 1-mL minivial. Blow down with dry
nitrogen as necessary.
E-7
3i
-------
2.4.2.2.6 Just before analysis, add 10 uL tridecane recovery standard solution
(Table 2, Exhibit D), and reduce the volume to a final volume of 10
uL, or 50 uL, as necessary (Section 12.9.2, Exhibit D). No column
chromatography is required.
2.4.2.2.7 Analyze an aliquot following the same procedures used to analyze
samples (Section 13, Exhibit D).
2.4.2.2.8 Report percent recovery of the internal standard and the presence
of any PCDD/PCDF compounds on Form (to be determined) in pg/mL of
rinsate solvent.
2.5 Duplicate Analyses
2.5.1 In each batch of samples, locate the sample specified for duplicate
analysis, and analyze a second 10-g soil or sediment sample portion or
1-L water sample, or an appropriate amount of the type of matrix under
consideration.
2.5.1.1 The results of the laboratory duplicates (percent recovery and concen-
trations of 2,3,7,8-substituted PCDD/PCDF compounds) must agree within
25 percent relative difference (difference expressed as percentage of
the mean). If the relative difference is >25 percent for any one of
the fifteen 2,3,7,8-substituted PCDDs/PCDFs, the laboratory shall
immediately contact the Sample Management Office for resolution of the
problem. Report all results.
E-8
o n
00
-------
2.5.1.2 Recommended actions to help locate problems:
2.5.1.2.1 Verify satisfactory Instrument performance (Section 8, Exhibit D).
2.5.1.2.2 If possible, verify that no error was made while weighing the sample
portions.
2.5.1.2.3 Review the analytical procedures with the performing laboratory
personnel.
2.6 Matrix Spike and Matrix Spike Duplicate
2.6.1 Locate the sample for the MS and MSD analyses (the sample may be labeled
"double volume").
2.6.2 Add on appropriate volume of the matrix spike fortification solution
(Exhibit D, Section 3.24), adjusting the fortification level as specified
in Exhibit D, Table 1, under IS Spiking Levels.
2.6.3 Analyze the MS and MSD samples as described in Exhibit D, Section 12.
2.6.4 The results obtained from the MS and MSD samples (percent recovery and
concentrations of 2,3,7,8-substituted PCDDs/PCDFs) must agree within 20
percent relative difference.
2.7 Percent Recovery of the Internal Standards
E-9
381)
-------
For each sample, method blank and rinsate, calculate the percent recovery
(Section 14.2, Exhibit D). It is recommended that the percent recovery be
>40 percent and <120 percent for all 2,3,7,8-substituted internal standards.
NOTE: A low or high percent recovery for a blank does not require dis-
carding the analytical data but it may indicate a potential problem with
future analytical data.
2.8 Identification Criteria
2.8.1 If either one of the identification criteria appearing in Sections
13.4.1.1 through 13.4.1.4, Exhibit D, is not met for an homologous
series, it is reported that the sample does not contain unlabeled
2,3,7,8-substituted PCDD/PCDF isomers for that homologous series at
the calculated detection limit (Section 14.5, Exhibit D).
2.8.2 If the first initial identification criteria (Sections 13.4.1.1 through
13.4.1.4) are met, but the criteria appearing in Sections 13.4.1.5 and
13.4.2.1, Exhibit D, are not met, that sample is presumed to contain
interfering contaminants. This must be noted on the analytical report
form, and the sample must be rerun or the extract reanalyzed. Detailed
sample rerun and extract reanalysis requirements are presented in
Exhibit C.
2.9 Blind QA/QC Samples
Included among soil, sediment and aqueous samples may be QA/QC samples
E-10
3GG
-------
that are not specified as such to the performing laboratory. Types that
may be included are:
2.9.1 Uncontaminated soil, sediment, or water.
2.9.1.1 If a false positive is reported for such a sample, the laboratory
shall be required to rerun the entire associated batch of samples
(Section to be determined, Exhibit C).
2.9.2 Split samples — composited sample portions sent to more than one
laboratory.
2.9.3 Unlabeled field duplicates — two portions of a composited sample.
2.9.4 Performance evaluation samples — soil/sediment or water samples con-
taining a known amount of unlabeled 2,3,7,8-substituted PCDDs/PCDFs
and/or other PCDD/PCDF compounds.
2.9.4.1 If the performance evaluation sample result falls outside the accept-
ance windows established by the EPA, the laboratory shall be required
to rerun the entire associated batch of samples (Exhibit C).
NOTE: EPA acceptance windows are based on previously generated data.
2.10 Quality Control Charts
The performance of the entire measurement system (i.e., from the extraction
E-ll
-------
of the sample to the mass spectrometric determination) must be documented
by using germane control charts. The selection and design of a specific
measurement control chart must be accomplished in a rational manner so
that the measurement process can be adequately surveyed. By using the
standard deviations obtained from control samples or control runs, the
laboratory must delineate control limits, i.e., statistically congruous
extreme values, which should warn the operator of possible problems. It
is recommended to consider the values corresponding to two standard devi-
ations as warning limits and the values from three standard deviations as
control limits (i.e., corrective actions are required). For some par-
ticular applications, however, the control limits must not exceed the
limits set forth by the EPA (e.g., ion-abundance ratios). [Specific and
required QC charts, such as mass and GC resolutions, ion abundance ratios,
RRF values, etc., will be described in the final version of this protocol.]
2.11 Standard Operating Procedures (SOPs)
As part of the quality assurance program, the laboratory must use in-house
SOPs describing how the basic operations executed within the laboratory
are done.
2.12 Internal Audits
Internal audits of records, Instrumentation performances and calibration
data are highly encouraged in order to identify defects that could
compromise the quality of the results.
E-12
368
-------
2.13 Records
At each laboratory, records must be maintained on site for six months
after contract completion to document the quality of all data generated
during the contract period. Before any records are disposed, written
concurrence from the Contracting Officer must be obtained.
2.14 Unused portions of samples and sample extracts must be preserved for six
months after sample receipt; appropriate samples may be selected by EPA
personnel for further analyses.
2.15 Reuse of glassware is to be minimized to avoid the risk of contamination.
3. Laboratory Evaluation Procedures
3.1 On a quarterly basis, the EPA Project Officer or his/her designated repre-
sentatives may conduct an evaluation of the laboratory to ascertain that
the laboratory is meeting contract requirements. This section outlines
the procedures which may be used by the Project Officer or his/her author-
ized representative in order to conduct a successful evaluation of
laboratories conducting dioxin analyses according to this protocol. The
evaluation process consists of the following steps: 1) analysis of a
performance evaluation (PE) sample, and 2) on-site evaluation of the
laboratory to verify continuity of personnel, instrumentation, and quality
assurance/quality control functions. The following is a description of
these two steps.
E-13
36i)
-------
3.2 Performance Evaluation (PE) Sample Analysis
3.2.1 The PE sample set will be sent to a participating laboratory to verify
the laboratory's continuing ability to produce acceptable analytical
results. The PE sample will be representative of the types of samples
that will be analyzed under this contract.
3.2.2 When the PE sample results are received, they are scored using the PE
Sample Score Sheet shown in Figure (to be determined). If a false
positive (e.g., a PE sample not containing 2,3,7,8-TCDD or other PCDD/
PCDF but reported by the laboratory to contain it or them) is reported,
the laboratory has failed the PE analysis requirement. The Project
Officer will notify the laboratory immediately if such an event occurs.
3.2.3 As a general rule, a laboratory should achieve 75 percent or more of the
total possible points for all three categories listed on the PE Sample
Score Sheet, and 75 percent or more of the maximum possible points in
each category, to be considered acceptable for this program. However,
the Government reserves the right to accept scores of less than 75
percent.
3.2.4 If unanticipated difficulties with the PE samples are encountered, the
total points may be adjusted by the Government evaluator in an impartial
and equitable manner for all participating laboratories.
3.3 On-site Laboratory Evaluation
E-14
370
-------
3.3.1 An on-site laboratory evaluation is performed to verify that (1) the
laboratory is maintaining the. necessary minimum level in instrumentation
and levels of experience in personnel committed to the contract and (2)
i
that the necessary quality assurance activities are being carried out.
It also serves as a mechanism for discussing laboratory weaknesses
identified through routine data audits, PE sample analyses results, and
prior on-site evaluations. Photographs may be taken during the on-site
laboratory evaluation tour.
3.3.2 The sequence of events for the on-site evaluations is shown in Figure
(to be determined). A Site Evaluation Sheet (SES) is used to document
the results of the evaluation.
E-15
i-> r~t -I
O ( 1
-------
METHOD : MEASUREMENT OF 2.3.7,8-TETRACHLORINATED
DIBENZO-P-DIOXIN (TCDD) AND 2,3.7.8-TETRACHLORINATED
DIBENZOFURAN (TCDF) IN PULP. SLUDGES. PROCESS SAMPLES
AND WASTEWATERS FROM PULP AND PAPER MILLS
1. SCOPE AND APPLICATION
1.1 This method is appropriate for the determination of
2,3,7,8-tetrachlorinated dibenzo-p-dioxins (TCDD) and
2 , 3 , 7,8-dibenzofurans (TCDF) in paper mill process
samples, including paper pulp, sludge, ash, mud,
woodchips, and treated and untreated wastewaters.
Chemical Abstracts Service STORET
Analyte Registry Number (CASRN) Number
2,3,7,8-Tetrachlorodibenzofuran • 51207-31-9
2,3,7,8-Tetrachlorodibenzo-p-dioxin 1746-01-6 3475
1.2 The sensitivity of this method is dependent upon the
level of interferences within a given matrix. Target
quantification levels for the analytes are 1 ppt in
solid samples and 10 ppq in water and wastewater
samples.
1.3 Certain *• 3 Ct 2 -labelled and 3 7 Cl« -2 , 3 ,7 , 8-substituted
congeners are used to provide calibration and method
recovery information. Appropriate capillary GC columns
and reference isomer standards are used to achieve
isomer specific data. Other isotopically labelled
congeners are also used to refine method recovery data.
1.4 This method is recommended for use only by analysts
experienced with residue analysis and skilled in mass
spectral analytical techniques.
2. SUMMARY OF THE METHOD
2.1 This procedure uses a matrix-specific extraction,
analyte-specific cleanup, and high-resolution capillary
column gas chromatography/ mass spectrometry (HRGC/MS)
techniques.
2.2 If interferences are encountered, the method provides
selective cleanup procedures to aid the analyst in
their elimination. The analysis flow chart is shown in
Figure 1.
3. DEFINITIONS
Internal standard: a pure compound added to a sample in
known amounts prior to sample preparation and used to
O f-t
O (
-------
Paper/Pulp
Process
Sample(s)
(1)
(2)
Add Internal Standards: *3 Ci2-TCDD
and laCitTCDF
Perform matrix-specific extraction.
Sample
Extract
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
Wash with 40% KOH
Wash with doubly distilled water
Wash with cone. Ha SO«
Repeat wash with Cone. H2SO«
Wash with doubly distilled water
Dry extract
Solvent Cone.
Silica gel column
Solvent Cone.
Alumina column
50% CH2 C12/hexane
Fraction
(1) Concentrate eluate
(2) Repeat alumina column
50% GHz C12/hexane
Fraction
(1) Concentrate eluate
(2) Perform carbon column cleanup
(3) Repeat alumina column
50% CH2 C12/hexane
Fraction
(1) Concentrate eluate
(2) Add Total Recovery Standard(s)
Analyze by GC/MS
Figure 1. Flow chart for sample extraction and cleanup
as used for the analysis of 2,3,7,8-TCDD and 2,3,7,8-TCDF
in paper/pulp process samples.
6
-------
calculate concentration of the analyte that is a sample
component. The internal standard is used to indicate
recovery of the analyte in the method.
External standard: a known amount of a pure compound that
is added to the sample extract prior to analysis. From
measured instrument response for a known amount of the
external standard, the recovery of the internal standard is
calculated.
Surrogate standard: a compound that is not expected to be
found in the sample, is added to the environmental sample
prior to sample preparation to monitor performance, and is
measured with the same procedures used to measure sample
components. A final calculation ratio to the internal
standard indicates any interfering responses in the region
of the analyte.
Method blank: a quality assurance sample that is prepared
through sample preparation, but without any environmental
sample.
Native spike: a small volume of a solution containing the
analytes that is added to a sample and/or blank solution and
is analyzed with the procedure used for an environmental
sample. Results of analyses are used to determine
statistically the accuracy and precision that can be
expected.
Laboratory duplicate: two aliquots of the same sample that
are treated in exactly the same manner throughout laboratory
analytical procedures. Analysis of laboratory duplicates
indicate precision associated with laboratory procedures but
not with sample collection, preservation or storage
procedures.
4. INTERFERENCES
4.1 Residues or contaminants present in solvents, or
reagents, or on glassware and other sample processing
equipment may yield elevated baselines in the mass
chroroatograms which may cause misinterpretation of GC-
MS data. All of the materials and apparatus used in
the analysis must be demonstrated to be free from
interferences under the conditions of analysis by
running laboratory method blanks.
4.2 The use of high purity reagents and solvents helps to
minimize interference problems. Purification of
solvents by distillation in all glass systems may be
required.
4.3 Interferences co-extracted from the sample may vary
considerably from sample to sample, depending upon the
«
3 374
-------
process from which the samples originated. The
2,3,7.8-TCDD and 2,3,7,8-TCDF isomers are often
associated with other interfering chlorinated compounds
which may be present in the sample at concentrations
several orders of magnitude higher than those of the
analytes of interest. Retention times of target
analytes must be verified using reference standards.
These values must correspond to the retention time
windows established as described in Section
9. While certain cleanup techniques are provided as
part of this method, unique samples may require
additional cleanup techniques in order to achieve the
stated method detection limit.
4.4 High resolution capillary gas chromatography (GO
columns are used to provide optimum resolution of TCDD
and TCDF isomers as possible.
5. SAFETY AND HANDLING PROCEDURES FOR 2.3.7.8-TCDD AND 2.3.7.8-
TCDF
5.1 The 2,3,7,8-TCDD isomer has been found to be acnegenic,
carcinogenic, and teratogenic in the course of
laboratory animal studies. The 2,3,7,8-TCDD is a solid
ar room temperature, and has a relatively low vapor
pressure. The solubility of this compound in water is
only about 200 parts-per-trillion, but the solubility
in various organic solvents ranges from about 0.001
percent to 0.14 percent. The physical properties of
2,3,7,8-TCDF have not been well established, although
it is presumed that the physical properties of these
congeners are generally similar to those of the
2,3,7,8-TCDD isomer. On the basis of the available
toxicological and physical property data for TCDD, this
compound, as well as the 2,3,7,8-TCDF should be handled
only by highly trained personnel who are thoroughly
versed in the appropriate procedures, and who
understand the associated risks.
5.2 The 2,3,7,8-TCDD and 2,3,7,8-TCDF isomers, and samples
containing these, are handled using essentially the
same techniques as those employed in handling
radioactive or infectious materials. Well-ventilated,
controlled-access laboratories are required, and
laboratory personnel entering these laboratories should
wear appropriate safety clothing, including disposable
coveralls, shoe covers, gloves, and face and head
masks. During analytical operations which may give
rise to aerosols or dusts, personnel should wear
respirators equipped with activated carbon filters.
Eye protection equipment (preferably full face shields)
must be worn at all times while workin.g in the
analytical laboratory with TCDD/TCDF. Various types of
gloves can be used by personnel , depending upon the
o t-i i,;
O ( O
-------
analytical operation being accomplished. Latex gloves
are generally utilized, and when handling samples
thought to be particularly hazardous, an additional set
of gloves are also worn beneath the latex gloves (for
example, Playtex gloves supplied by American Scientific
Products, Cat. No. 67216). Bench-tops and other work
surfaces in the laboratory should be covered with
plastic-backed absorbent paper during all analytical
processing. When finely divided samples (dusts, soils,
dry chemicals) are processed, removal of these from
sample containers, as well as other operations,
including weighing, transferring, and mixing with
solvents, should all be accomplished within a glove
box. Glove boxes, hoods and the effluents from
mechanical vacuum pumps and gas chromatographs on the
mass spectrometered should be vented to the atmosphere
preferably only after passing through HEPA particulate
filters and vapor-sorbing charcoal.
5.3 All laboratory ware, safety clothing, and other items
potentially contaminated with 2,3,7,8-TCDD and 2,3,7,8-
TCDF in the course of analyses must be carefully
secured and subjected to proper disposal. When
feasible, liquid wastes are concentrated, and the
residues are placed in approved steel hazardous waste
drums fitted with heavy gauge polyethylene liners.
Glass and combustible items are compacted using a
dedicated trash compactor used only for hazardous waste
materials and then placed in the same type of disposal
drum. Disposal of accumulated wastes is periodically
accomplished by high temperature incineration at EPA-
approved facilities.
5.4 Surfaces of laboratory benches, apparatus and other
appropriate areas should be periodically subjected to
surface wipe tests using solvent-wetted filter paper
which is then analyzed to check for TCDD/TCDF
contamination in the laboratory. Typically, if the
detectable level of TCDD or TCDF from such a test is
greater than 50 ng/m2 , this indicates the need for
decontamination of the laboratory. In the event of a
spill within the laboratory, absorbent paper is used to
wipe up the spilled material and this is then placed
into a hazardous waste drum. The contaminated surface
is subsequently cleaned thoroughly by washing with
appropriate solvents (methylene chloride followed by
methanol) and laboratory detergents. This is repeated
until wipe tests indicate that the levels of surface
contamination are below the limits cited.
5.5 In the unlikely event that analytical personnel
experience skin contact with TCDD/TCDF or samples
containing these, the contaminated skin area should
immediately be thoroughly scrubbed using mild soap and
water. Personnel involved in any such accident should
O P*-I / *
o ft)
-------
subsequently be taken to the nearest medical facility,
preferably a facility whose staff is knowledgeable in
the toxicology of chlorinated hydrocarbons. Again,
disposal of contaminated clothing is accomplished by
placing it in hazardous waste drums.
5.6 It is desirable that personnel working in laboratories
where TCDD/TCDF are handled be given periodic physical
examinations (at least yearly). Such examinations
should include specialized tests, such as those for
urinary porphyrins and for certain blood parameters
which, based upon published clinical observations, are
appropriate for persons who may be exposed to
TCDD/TCDF. Periodic facial photographs to document the
onset of dermatologic problems are also advisable.
6. APPARATUS AND EQUIPMENT '
6.1 GAS CHROMATOGRAPH/MASS SPECTROMETER DATA SYSTEM:
6.1.1 Gas chromatograph: An analytical system which
incorporates a temperature-programmable gas
chromatograph and associated accessories,
including syringes, analytical columns and
gases, is required for these analyses.
6.1.2 Fused silica capillary GC columns are used. As
shown in Tables 1 and 2, two different
capillary GC columns are employed and the
performance of these is evaluated on a
continuing basis by using column performance
check mixtures containing all 22 TCDD isomers
and all 38 TCDF isomers.
The columns used include the following: (a) a
60-m DB-5 column operated at an initial
temperature of 180°C, held for 1 minute at that
temperature, and then programmed from 180° to
240°C at a rate of 2°C/minute, and held at.that
temperature for 14 minutes; (b) a hybrid column
consisting of a section of DB-5 (10-m x 0.25 mm
I.D.; 0.25 urn film thickness) coupled to a
section of DB-225 (30-m x 0.25 mm I.D.; 0.25
urn film thickness), which is operated initially
at 180°C, held at that temperature for 1
minute, then programmed from 180° to 220°C at a
rate of 2°C/minute. The column described in
Section (a) provides complete separation of
2,3,7,8-TCDD from the other 21 TCDDs, but does
not completely resolve 2,3,7,8-TCDF from all
other TCDFs. The column described in (b)
provides complete separation of 2,3,7,8-TCDF
from the other 37 TCDFs, but does not uniquely
resolve 2,3,7,8-TCDD. Other capillary columns
6 377
-------
which provide separation of 2,3,7,8-TCDD from
all of the other 21. TCDD isomers which is
equivalent to that specified in this protocol
may also be used for 2,3,7,8-TCDD analysis.
This separation must be demonstrated and
documented periodically using the performance
text mixture described in Section 7.3.5.
Moreover, any capillary column which provides
separation of 2,3,7,8-TCDF from all the other
37 TCDF isomers equivalent to that specified in
this protocol may be used for 2,3,7,8-TCDF
analysis. This separation must also be
demonstrated and documented at regular
intervals using the performance test mixture
described in Section 7.3.7.
6.1.3 Mass spectrometer: An instrument capable of
mass spectral resolution of at least 1:6500 is
desirable. This instrument should utilize the
electron impact ionization mode (70 eV nominal
electron energy). The system must be capable
of selected ion monitoring (SIM) for at least
11 ions simultaneously, with a cycle time of 1
sec or less. Minimum integration time for SIM
is 50 ms per mass peak. The use of systems not
capable of monitoring 13 ions simultaneously
will require the analyst to make multiple
injections.
6.1.4 GC/MS interface: Any GC-to-MS interface that
gives an acceptable calibration response for
each analyte of interest at the concentration
required and achieves the required tuning
performance criteria (see Section 9) may be
used. GC-to-MS interfaces constructed of all
glass or glass-lined materials are required.
6.1.5 Data system: A computer system must be
interfaced to the mass spectrometer. The
system must allow for the continuous
acquisition and storage on machine-readable
media of all data obtained throughout the
duration of the chromatographic program. The
computer must have software that can search any
GC/MS data file for ions of a specific mass and
can plot such ion abundances versus time or
scan number. This type of plot is defined as
an Selected Ion Current Profile (SICP).
Software must also be able to integrate the
abundance, in any SICP, between specified time
or scan number limits.
6.2 LABORATORY APPARATUS
6.2.1 For Standard Preparation
o (
-------
6.2.1.1 Analytical balance
6.2.1.2 Class A pipets, various sizes
6.2.1.3 Glove box
6.2.2 For Sample Homogenization
6.2.2.1 Desiccator, to dry solid samples
6.2.2.2 Blender, glass jar, for mixing dried
pulp and sludge sample
6.2.2.3 Laboratory mill, for grinding wood
chips
6.2.3 For Per Cent- Moisture Determination
6.2.3.1 Desiccator
6.2.3.2 Drying oven
6.2.4 For Sample Preparation
6.2.4.1 Filtering funnel
6.2.4.2 Round bottom flask, 5 L
6.2.4.3 Teflon magnetic stir bag, egg-shaped,
(51 mm x 19 mm)
6.2.4.4 Magnetic stir plate
6.2.4.5 Soxhlet extractor, 45 mm I.D., with
500 mL erlenmeyer flask
6.2.4.6 Snyder column, three ball macro
6.2.4.7 Extraction heater
6.2.4.8 Nitrogen blowdown apparatus, water
bath capable of maintaining
temperation ±5°
6.2.4.9 Wrist action shaker
6.2.4.10 Tube furnace, for silica activation
6.2.4.11 Muffler furnace, capable of
maintaining 600° ±10° , for alumina
activation
-------
7. REAGENTS AND CONSUMABLE MATERIALS
7.1 CONSUMABLE MATERIALS
7.1.1 Filter paper, 15 cm, 2.5 micron retention
rating, glass fiber filters are also
recommended
7.1.2 Aluminum weighing boats
7.1.3 Glass bottles, various sizes, 2L, 1L, 250 mL,
and 125 mL, equipped with Teflon-lined caps
7.1.4 Dish, petri, 150 x 15 mm
7.1.5 Silica gel (Bio-Sil A 100/200 mesh): Bio-Rad
Rockville Centre, NY. the silica gel is
conditioned prior to use by initially placing a
200 g portion of the silica in a 30 mm x 30 cm
long glass tube (the silica gel is held in
place by glass wool plugs) which is placed in a
tube furnace. The glass tube is connected to a •
pre-purified nitrogen cylinder through a series
of oxygen scrubber traps. The first step in
conditioning the silica gel entails heating the
glass tube containing the 200 g aliquot of
silica for 30 minutes at 180° C while purging
with nitrogen (flow rate 50-100 mL/minute),
subsequently the tube is removed from the
furnace and allowed to cool to room
temperature. Methanol (175 mL) is then passed
through the tube, followed by 175 mL methylene
chloride. The tube containing the silica is
then returned to the furnace, the nitrogen
purge is again established (50 to 100 mL/minute
flow), the tube is heated at 50°C for 10
minutes, then the temperature is gradually
increased to 180°C over a period of 25 minutes
and maintained at 180°C for 90 minutes.
Heating is then discontinued but the nitrogen
purge is maintained until the tube cools to
room temperature. Finally, the silica is
transferred to a clean, dry, glass bottle and
capped with a Teflon-lined screw cap for
storage in a desiccator.
7.1.6 Silica Gel Impregnated with Sulfuric Acid (30%
w/w) : Concentrated sulfuric acid (4.4 g) is
combined with 10.0 g silica gel (conditioned as
described in Section 7.1.5) in a screw capped
bottle and agitated to mix thoroughly.
Aggregates are dispersed with a stirring rod
until a uniform mixture is obtained. The
Hz SO*-silica gel is stored in a screw-capped
bottle (equipped with a Teflon-lined cap).
o c
SO
-------
7.1.7 Silica Gel Impregnated with Sodium Hydroxide:
IN Sodium hydroxide (30 g) is combined with 100
g Bio-Sil A (conditioned as described in
Section 7.1.5) in a screw capped bottle and
agitated to mix thoroughly. Aggregates are
dispersed with a stirring rod until a uniform
mixture is obtained. The NaOH-silica gel is
stored in a screw-capped bottle (Teflon-lined
cap) .
7.1.8 Carbon/Celite: A 10.7 g aliquot of PX-21
carbon (Anderson Development Co., Adrian,
Michigan) is combined with 125 g of Celite 545
in a 250 mL glass bottle, fitted with a Teflon-
lined cap, and the mixture is shaken to obtain
a uniform mixture. The Carbon/Celite mixture
is stored in' the screw-capped bottle.
7.1.9 Micro-reaction vial, 3 mL
7.1.10 Glass boiling beads, solvent washed
7.1.11 Soxhlet thimble (90 mm x 35 mm), glass
7.1.12 Glass wool, silanized
7.1.13 Nitrogen gas, prepurified
7.1.14 Hydrogen, ultra high purity
7.1.15 Syringes, various sizes
7.1.16 Capillary gas chromatography column, DB-5, 60 M
x 0.25 mm I.D.; 0.25 um fil thickness (J and W
Scientific Co.)
7.1.17 Capillary gas chromatography column, DB-225, 30
M x 0.25 mm I.D.; 0.25 um film thickness (J and
W Scientific Co.)
7.2 REAGENTS
7.2.1 Potassium Hydroxide (ACS): 40 percent (w/v) in
distilled water
7.2.2 Sulfuric acid (ACS), concentrated
7.2.3 Methylene chloride, methanol, acetone, benzene,
hexane, ethyl acetate, cyclohexane, tridecane.
Distilled in glass or highest available purity
7.2.4 Anhydrous sodium sulfate (ACS), Dry in an oven
at 400°C for 3 hours, cool in a desiccator over
self-indicating silica gel, transfer to Teflon-
10
381
-------
lined screw cap bottles and seal. Store in a
desiccator until use.
7.3 CALIBRATION AND SPIKING STANDARDS
7.3.1 Stock standard solutions of the appropriate
TCDD and TCDF isomers, and mixtures thereof,
are prepared in a glovebox, using weighed
quantities of the authentic isomers. These
stock solutions are contained in appropriate
amber bottles and are stored tightly stoppered
in a refrigerator. Aliquots of the stock
standards are removed for direct use or for
subsequent serial dilutions to prepare working
standards. These standards must be checked
regularly (by comparing instrument response
factors for them over a period of time) to
ensure that solvent evaporation or other losses
have not occurred which would alter the
standard concentration. The standard solutions
which may be required to perform the
quantitative analyses of 2,3,7,8-TCDD and
2,3,7,8-TCDF are listed below.
7.3.2 Internal Standard (IS) Solution. An aliquot
of this solution is added to samples which are
to be analyzed for 2,3,7,8-TCDD and 2,3,7,8-
TCDF. Prepare a stock solution containing the
following isotopically-labelled TCDD and TCDF
compounds in isooctane at the indicated
concentrations: 0.05 ng/uL l3Ciz-2,3,7,8-TCDD,
0.02 ng/uL 37C14-2,3,7,8-TCDD, 0.05 ng/uL
13Ci2-2,3,7,8-TCDF, and 0.02 ng/uL 37Cl«-
2,3,7,8-TCDF. Typically a twenty microliter
aliquot of this standard solution is added to
each sample aliquot prior to preparation and
the 13Ci2-labelled materials serve as internal
standards for use in quantitation. Recovery of
these standards is also used to gauge the
overall efficacy of the analytical procedure.
7.3.3 External Standard (ES) Solution. Prepare a
stock solution containing 0.05 ng of *3Ci2-
1,2,3,4-TCDD and 0.10 ng of 3 7 Cl« -1, 2 , 7 . 8-
TCDF/uL tridecane. A 10 microliter aliquot of
this standard is added to the final extract
obtained for each sample just prior to GC-MS
analysis. When the DB-5 capillary column is
employed, the r3Ci2-1,2,3,4-TCDD is used as an
external standard in the quantitation of the
i3Ci2-2,3,7,8-TCDD and the *3Ci2-2,3,7,8-TCDF
internal standards present in the final
extract, and the percent recovery of each of
these *3Ci2-labelled internal standards is
calculated on the basis of this quantitative
11 9
-------
analysis. The 37Cl«-1,2,7,8-TCDF external
standard is employed in quantitating the
concentration of 13Ci2-2,3,7,8-TCDF when the
hybrid DB-5/DB-225 capillary column is
implemented in quantitating 2,3,7,8-TCDF.
These latter results are subsequently
implemented in calculating the percent recovery
of the l3Ci2-2,3,7,8-TCDF internal standard
achieved during the analysis performed using
the hybrid column.
7.3.4 Calibration Standard Solutions (CS1, CS2, CS3,
CS4 and CSS). Prepare five separate
calibration standards as follows: (a)
Calibration Standard (CS1), 0.2 ng/uL 2,3,7,8-
TCDD, 0.2 ng/uL 2,3,7,8-TCDF, 0.05 ng/uL »3Ci2 -
2,3,7,8-TCDD, 0.05 ng/uL »3Ci2-2,3,7,8-TCDF,
0.02 ng/uL-37C14-2,3,7,8-TCDD, 0.02 ng/uL
37C1«-2,3,7,8-TCDF, 0.05 ng/uL »3Ci2-1,2,3,4-
TCDD and 0.10 ng/uL 37Cl«-1,2,7,8-TCDF; (b)
Calibration Standard (CS2), 0.05 ng/yL 2,3,7,8-
TCDD, 0.05 ng/uL 2,3,7,8-TCDF plus the same -
concentration of isotopically-labelled
standards included in CS1; (c) Calibration
Standard (CS3), 0.01 ng/uL 2,3,7,8-TCDD, 0.01
ng/uL 2,3,7,8-TCDF plus the same concentration
of isotopically-labelled standards included=s_jjiv
CS1; (d) Calibration Standard (CS4) , ^6.0025^-
ng/uL 2,3,7,8-TCDD, 0.0025 ng/uL 2 , 3,7Y8-TCDF'
plus the same concentration of isotopically-
labelled standards included in CS1; (e)
Calibration Standard (CSS), 2.0 ng/uL 2,3,7,8-
TCDD, 2.0 ng 2,3,7,8-TCDF plus the same
concentrations of isotopically-labelled
standards included in CS1. Aliquots of these
standards are injected to obtain data which is
implemented in constructing the calibration
curve used in quantitating 2,3,7,8-TCDD and
2,3,7,8-TCDF.
7.3.5 TCDD Gas Chromatographic Resolution Standard
Mixture (RM1). Prepare an isooctane solution
containing 0.05 ng/uL concentrations of each of
the following TCDD isomers: 1,3,6,8-TCDD;
1,2,3,7-TCDD; 1,2,3,9-TCDD; 2,3,7,8-TCDD;
1,2,3,4-TCDD and 1, 2 , 8 , 9-TCDD . Two of the
isomers in this mixture are used to define the
gas Chromatographic retention time window for
TCDDs (1,3,6,8-TCDD is the first eluting TCDD
isomer and 1,2,8,9-TCDD is the last eluting
TCDD isomer on the DB-5 GC column). The
remaining isomers serve to demonstrate that the
2,3,7,8-TCDD isomer is resolved from the other
nearest eluting TCDD isomers, and that the
column therefore yields quantitative data for
383
12
-------
the 2,3,7,8-TCDD isomer alone.
7.3.6 TCDF Native Standard Mixture (NS1). Prepare
a solution containing 0.250 ng/uL 2,3,7,8-TCDF
in tridecane. This standard is implemented
when it is desired to add only native 2,3,7,8-
TCDF to a sample.
7.3.7 TCDF Gas Chromatographic Resolution Standard
Mixture (RM2). Prepare a solution containing
approximately 0.250 ng/uL of each of the 37
TCDF isomers (exclusive of 2,3,7,8-TCDF) in
isooctane. This standard is used when it is
desired to add all of the TCDF isomers except
2,3,7,8-TCDF to a sample. This standard is
also implemented to determine the relative
retention times of the TCDF isomers and, when
this standard is co-injected with an aliquot of
standard NS1, the efficacy of a particular gas
Chromatographic column for separating 2,3,7,8-
TCDF from each of the 37 other TCDF isomers can
be ascertained.
7.3.8 TCDD/TCDF Native Standard Mixtures NS2 and
NS3. For Mixture NS2, prepare a solution
containing 0.025 ng 2,3,7,8-TCDD and 0.025 ng
2,3,7,8-TCDF per microliter of tridecane. For
mixture NS3, prepare a solution containing
0.005 ng 2,3,7,8-TCDD and 0.005 ng 2,3,7,8-
TCDF. These standards are employed when it is
desired to simultaneously add both native
2,3,7,8-TCDD and native 2.3,7,8-TCDF to a
sample.
8. SAMPLE COLLECTION, PRESERVATION AND HANDLING
8.1 Grab and composite samples must be collected in glass
containers. Conventional sampling practices must be
followed. The bottle must not be prewashed with sample
before collection. Composite samples should be
collected in glass containers. Sampling equipment must
be free of tygon, rubber tubing and other potential
sources of contamination which may absorb the target
analytes.
8.2 All samples are refrigerated at 4°C, and sample
preparation must commence within 30 days.
9. CALIBRATION AND STANDARDIZATION:
9.1 Calibrating the MS Mass Scale: Perfluorokerbsene,
decafluorotriphenyl phosphine, or any other accepted
mass marker compound must be introduced into the MS, in
13
-------
order to calibrate the mass scale through at least m/z
350. The procedures specified by the manufacturer for
the particular MS instrument used are to be employed
for this purpose. The mass calibration should be
rechecked at least at 8 hr. operating intervals.
9.2 Gas Chromatograph Temperature Program: Table 1 shows
the GC temperature program typically used to resolve
2,3,7,8-TCDD from each of the 21 other TCDD isomers and
indicates the ion-masses monitored and the time
analytical sequence implemented for isomer specific
quantitation of 2,3,7,8-TCDD and non-isomer specific
quantitation of 2,3,7,8-TCDF. This temperature program
and ion monitoring time cycle must be established by
each analyst for the particular instrumentation used by
injecting aliquots of Standard Mixture RM1, as well as
the calibration mixtures (CS1, CS2, CS3 and CS4) into
the GC-MS . It may be necessary to adjust the
temperature program and ion monitoring cycles slightly
based on the observations from analysis of these
mixtures.
9.3 Checking GC Column Resolution for 2,3,7,8-TCDD and
2,3,7,8-TCDF: Utilize Standard Mixture 109071-1 to
check the DB-5 column resolution for 2,3,7,8-TCDD, and
utilize a combination of Standards NS1 and RM2 to
verify that 2,3,7,8-TCDF is separated from all of the
other TCDF isomers on the hybrid DB-5/ DB-225 column.
A 25% valley or less must be obtained between the mass
chromatographic peak observed for 2,3,7,8-TCDD and
adjacent peaks arising from other TCDD isomers and
similar separation of 2,3,7,8-TCDF from other
neighboring TCDFs is required. Analyze the column
performance standards using the instrumental parameters
specified above and in Table 1 and 2. The column
performance evaluation must be accomplished each time a
new column is installed in the gas chromatograph, and
at the beginning and conclusion o.f each 8 hour
operating period. If the column resolution is found to
be insufficient to resolve 2,3,7,8-TCDD and 2,3,7,8-
TCDF from their neighboring TCDD and TCDF isomers,
respectively, (as measured on the two different columns
used for resolving these two isomers) , then a new
DB-5 and/or DB-5/DB-225 hybrid GC column must be
installed.
9.4 Calibration of the GC-MS-DS System: To accomplish
quantitative analysis of 2,3,7,8-TCDD and 2,3,7,8-TCDF
contained in the sample extract, the GC-MS system is
calibrated by analyzing a series of at least three
working calibration standards. Each of these standards
is prepared to contain the same concentration of each
of the »3Ci 2-2,3,7,8-TCDD and >3Ci 2-2,3,7,8-TCDF
internal standards used here but a different
concentration of the native 2,3,7,8-TCDD and 2,3,7,8-
14
°> O ^
ooO
-------
TCDF. Typically, mixtures will be prepared so that the
ratio of the native 2,3,7,8-TCDD and 2,3,7,8-TCDF to
the isotopically-labelled TCDD/TCDF ranges between 0.05
and 4.0 in the four working calibration mixtures.
Prior to injecting aliquots of actual sample extracts,
an aliquot of a standard containing typically 0.2 ng of
1 3Ci 2-1,2,3,4-TCDD and 0.4 ng of a 7 CI* -1 , 2 , 7 , 8-TCDF
(External Standard) is used to dilute the extract in
the sample vials and is therefore co-injected along
with the sample extract, in order to obtain data
permitting calculation of the percent recovery of the
1 3Ci 2-2,3,7,8-TCDD and *3Ci 2-2,3,7,8-TCDF internal
standards. When the analysis of the extract is
performed using the DB-5 capillary column, the x 3 Ci 2 -
1,2,3,4-TCDD standard is implemented as the external
standard in quantitating both l3Ci2-2,3,7,8-TCDD and
13Ci2-2,3,7,8-TCDF. However, when the hybrid DB-5/DB-
225 column is employed in analyzing the 2,3,7,8-TCDF,
the 3TCl«-1,2,7,8-TCDF is implemented in quanitiating
the J 3Ci 2-labelled TCDF internal standard. Equations
for calculating relative response factors from the
calibration data derived from the calibration standard
analyses, and for calculating the recovery of the
13Ci2-2,3,7,8-TCDD and »3Ci 2-2,3,7,8-TCDF, as well as
the concentration of native 2,3,7,8-TCDD and 2,3,7,8-
TCDF in the sample (from the extract analysis) , are
summarized below.
9.5 Daily Checks of the Instrument Response: These will be
accomplished using Standard CS3. This standard will be
injected at the beginning of each work-day (or the
beginning of each 8-hour shift) and RRF values for
2,3,7,8-TCDD and 2,3,7,8-TCDF will be calculated. If
either of these RRF values deviate from the values
contained in the calibration curve by more than +20%,
then a second injection will be made and RRF values for
the two compounds will be again calculated. If either
of these RRF values also fail to agree with the
calibration curve by more than ±20%, then the entire
series of calibration standards will be analyzed, new
RRF values will be calculated, and a new calibration
curve will be constructed and applied in subsequent
analyses.
10. QUALITY ASSURANCE/QUALITY CONTROL PROCEDURES
10.1 The Quality Assurance and Quality Control procedures
itemized below will be implemented throughout the
course of the 2,3,7,8-TCDD and 2,3,7,8-TCDF analyses:
10.1.1 Each sample analyzed is spiked with stable
isotopically-labelled internal standards, prior
to extraction and analysis. Recoveries
obtained for each of these standards should
15
386
-------
typically be in the range from 40-120%. Since
these compounds are used as true internal
standards however, lower recoveries do not
necessarily invalidate the analytical results
for native 2,3,7,8-TCDD and 2,3,7,8-TCDF, but
may result in higher detection limits than are
desired.
10.1.2 Processing and analysis of at least one method
blank sample is generally accomplished for each
set of samples.
10.1.3 It is desirable to analyze at least one sample
spiked with representative native TCDD/TCDF for
each set of samples. The results of this
analysis provides an indication of the efficacy
of the entire analytical procedure. The
results of this analysis will be considered
acceptable if the detected concentration of the
. native 2,3,7,8-TCDD and 2,3,7,8-TCDF added to
the sample is within +_50% of the known
concentration.
10.1.4 At least one of the samples analyzed out of
each set is usually analyzed in duplicate and
the results of the duplicate analysis are
.included in the report of data.
10.2 A report describing the results of the analyses
discussed above will, at a minimum, include copies of
original mass chromatograms obtained during analyses of
the sample extracts and associated calibration
standards, a description of the analytical methodology
employed, and tabulations of calculated results.
Calculations and manipulation of data are most
efficaciously accomplished using computerized data
reduction techniques. The tabulations of calculated
results provided in the report will include tables
showing the concentrations of 2,3,7,8-TCDD and of
2,3,7,8-TCDF which were measured in each sample. Also
typically shown in these tables are the quantity of
each sample analyzed; the detection limits for those
samples which were found to contain no 2,3,7,8-TCDD or
2,3,7,8-TCDF; the GC-MS instrument implemented in the
analysis; the date and time of the analysis; the ratio
of the intensities of m/z 320 vs. m/z 322 and m/z 332
vs. m/z 334 for TCDD and the ratio of the intensities
of m/z 304 vs. m/z 306 and m/z 316 vs. m/z 318 for
TCDF; the percent recovery of the internal standard
(i3Ci 2-2,3,7,8-TCDD or »3Ci2-2,3,7,8-TCDF); and the ion
intensities for the following m/z's: 320, 322, 257,
332 and 334 for TCDD, and 304, 306, 241, 316 and 318
for TCDF. Other tabular data which should be provided
in the report include a table showing a summary of the
calibration data obtained for 2,3,7,8-TCDD and 2,3,7,8-
16 387
-------
TCDF, which indicates the date of the calibration; the
GC-MS instrument implemented; the WSU identification
number of the calibration solution; the calculated
response factors and the mean response factors obtained
for native 2,3,7,8-TCDD, native 2,3,7,8-TCDF, *3Ci2 -
2,3,7,8-TCDD and »•Ci2-2,3.7 , 8 ; and the % valley
observed for the GC separation of the 2,3,7,8-TCDD from
adjacent-eluting TCDD isomers and of the 2,3,7,8-TCDF
from adjacent-eluting TCDF isomers. Additional tables
which present calibration data and results obtained for
each individual sample in a more detailed manner than
that given in the summary tables mentioned earlier
should also be included in the data package.
11. PROCEDURE
11.1 SAMPLE PREPARATION -
11.1.1 Sludge Samples
11.1.1.1 Open the sample container and using a
spatula, break the sludge into small
particles (about 2 cm diameter or
less) and stir the sample vigorously
to make it as homogeneous as
possible. Remove an aliquot of this
sample (approximately 5 g) for an
"oven-dried solids as-received"
determination, using the procedures
described below (Section 11.1.1.4).
Remove the remaining sample from the
container and distribute it uniformly
on a stainless steel screen which is
supported at a distance of about 1 cm
above a sheet of aluminum foil, both
the foil . and the screen being
contained within a desiccator
containing an appropriate water
sorbent. To minimize the possiblity
of contamination or cross-
contamination of the sample, only one
sample at a time is dried in a given
desiccator. Allow the sample to
remain in the desiccator until it is
essentially dry, as indicated by the
sample color, consistency, and ease
of mixing. For each group of five
sludge samples which are desiccated,
prepare a laboratory blank as
follows. Place a 15 cm paper filter
on a stainless steel screen supported
at a distance of 1 cm above a sheet
of aluminum foil contained in a
desiccator and allow the filter to
17 388
-------
remain in the desiccator for the same
period as that which was used for
that sample of the five which
required the longest drying time.
Subsequently remove the filter from
the desiccator and continue with the
homogenization, drying, and other
sample preparative steps described
below.
11.1.1.2 When the sample has been dried
sufficiently, remove it from the
desiccator and transfer it to a
laboratory blender which is housed
within a glove box or similar
enclosure. Following homogenization
in the blender, remove an aliquot
(approximately 5 g) of the blended
solids, accurately weigh this sample
aliquot, and subject it to an oven-
dried solids determination, as
described in Section 11.1.1.3 of this
protocol.
11.1.1.3 Place the remaining desiccated,
blended sample into a clean sample
bottle fitted with a Teflon-lined
screw cap, and store the bottle in a
refrigerator (5°C). An aliquot of
this desiccated, blended sample is
subsequently analyzed for 2,3,7,8-
TCDD and 2,3,7,8-TCDF by applying the
extraction and analysis procedures
which are described in Sections 11.3
and 11.4, respectively, of this
protocol.
11.1.1.4 Determination of oven-dried-solids on
a sample aliquot is accomplished by
placing the weighed aliquot into a
tared aluminum boat which is then
placed in an oven maintained at
105°C. After heating for a period of
twenty-four hours, the aluminum boat
containing the sample is removed from
the oven, allowed to cool for 30
minutes in a desiccator, and then
weighed. The boat and sample are
then returned to the oven for an
additional four-hour period, after
which, the boat is again removed from
the oven, allowed to cool and weighed
again. The latter procedure is
repeated until the weight of the
sample as indicated by two successive
380
-------
weighings is observed to be constant.
From the observed weight loss upon
drying, the percentage of oven-dried-
solids in the original sample can be
determined. This result, as
determined for an aliquot of the
sample as received, is reported as
the "initial oven-dried solids as
received." The oven-dried weight
loss, as determined for an aliquot of
the previously desiccated sample (a
separate aliquot of which is
subsequently analyzed for TCDD/TCDF)
is used only to determine the actual
weight of the sample aliquot which is
analyzed on the oven-dried solids
basis.
11.1.2 Wood Chip Samples: Samples of wood chips in
which the chips are relatively large (typically
1-1.5 inches in length) are initially reduced
to smaller particle size (2 cm diameter or
less) using a laboratory mill. This mill is
cleaned thoroughly before each sample is
introduced. The pulverized wood sample
resulting from this operation is subsampled and
dried using exactly the same procedures
described for sludge samples in the foregoing
Section 11.1.1.1.
11.1.3 Ash Samples: Ash Samples are prepared using
the same procedures as described above for
sludge (Section 11.1.1.1), with the exception
that these samples cannot be supported on a
screen to dry, and are therefore placed into a
shallow, flat dish in order to dry them in the
desiccator. The ash" is spread as a thin layer
over the bottom of the dish and is gently
stirred periodically during the drying period.
11.1.4 Paper Pulp Samples: Remove the pulp sample
from the container, and then express as much
water from the sample as possible by
compressing it with a spatula after wrapping
the sample in aluminum foil. Using the spatula,
separate the sample mass into pieces which are
about 2 cm or less in diameter, and distribute
these pieces uniformly on a stainless steel
screen supported about 1 cm above a sheet of
aluminum foil, both the screen and the foil
being placed in a desiccator. Allow the sample
to remain in the desiccator until it is
essentially dry, as gauged by color and
consistency. For each group of five pulp
samples, prepare a laboratory blank using the
19 390
-------
procedures described in 11.1.1.1. Proceed with
the subsaropling and other drying procedures, as
described for sludge samples, beginning with
Section 11.1.1.2.
11.1.5 Slurry-Type Samples (Secondary Sludge, etc.):
Shake the sample bottle vigorously so as to
obtain a uniform suspension of the sample and
when the sample is homogeneous throughout, as
judged by visual inspection, remove an aliquot
of the sample and subject it to a Total
Suspended Solids Determination, as described in
Standard Methods For the Examination of Water
and Wastewater, 17th Edition. APHA, AWWA,
WPCF, 1986, Method 209C. Allow the remainder
of the sample to stand under refrigeration and
when the solids appear to have totally settled
to the bottom of the container, filter the
supernatant using a previously desiccated and
XXX tared Gelman Type A/E filter contained in a
glass filtering funnel. Remove the solids from
the sample container using a clean spatula and
utilize three 100 mL aliquots of HPLC water to
accomplish three successive rinses of the
sample container, and to effect a quantitative
transfer of the solids from the sample
container to the filter. Following separation
of the water from the solid, remove the solid
along with the filter paper from the funnel and
distribute the solids and the filter paper on
a stainless steel screen supported about 1 cm
above a sheet of aluminum foil, both the screen
and the foil being placed in a desiccator.
Allow the sample to dry until it is friable.
For each group of five samples, prepare a
laboratory blank using the procedures described
in 11.1.1.1. Proceed with the subsampling and
other drying procedures described for sludge
samples, beginning with Section 11.1.1.2. Note
that the tare weights of all filters used in
the separation of the liquid and solid phases
of these samples must be subtracted from the
combined solids-filter weight to determine the
actual weights of the solids samples prepared,
since the filter and solids cannot be readily
separated.
11.1.6 Water and Wastewater Samples
11.1.6.1 Clean and prepare four new 2 L
bottles fitted with Teflon-lined
caps. Mark the 1 gallon bottle
containing the aqueous sample, as
received, to show the original level
of the liquid in the bottle. Shake
20
-------
the bottle vigorously until all
solids in the bottle (which may have
settled to the bottom of the bottle
if the sample was undisturbed for
some time prior to analysis) are
suspended, as visually estimated.
Pour approximately equal portions of
the resuspended aqueous sample from
the 1 gallon bottle into each of the
four 2 L bottles using a funnel. To
accomplish this transfer, pour small
portions from the 1 gallon bottle
into each of the 2 L bottles in
succession, repeating this cycle as
many times as necessary to dispense
all of the contents of the 1 gallon
container. Following each pouring
step, recap the 1 gallon bottle and
shake it vigorously to ensure that
any particulate in the liquid remains
suspended.
11.1.6.2 After all of the contents of the 1
gallon bottle have been transferred
to the four 2 L bottles, rinse the 1
gallon bottle successively with two
50 mL portions of HPLC grade water,
accumulating these rinses in a 250 mL
graduated beaker. Transfer one-
fourth of each of these accumulated
water rinses to each of the four 2 L
bottles. Rinse the 250 mL beaker
successively with two 40 mL portions
of HPLC water, transferring one-
fourth of these rinses to each of the
four 2 L bottles. Recap all four 2 L
bottles and retain for subsequent
extraction and analysis.
11.1.6.3 Rinse the original 1 gallon empty
sample bottle successively with two
50 mL portions of methylene chloride,
accumulating these in the 250 roL
beaker used earlier. Transfer these
methylene chloride rinses to a clean
250 mL bottle fitted with a Teflon-
lined cap. Rinse the 250 mL beaker
successively with two 50 mL portions
of methylene chloride, and pool these
rinses with the other accumulated
methylene chloride rinses in the 250
mL bottle. Reserve the pooled
methylene chloride rinses for later
splitting and combination with the
methylene chloride rinses collected
21
-------
as described in Section 11.2.2 below.
11.1.6.4 Select one of the 2 L bottles
containing the split water/wastewater
sample, and add to this bottle a
solution of the »3Ci2-labelled TCDD
and TCDF internal standards, prepared
by combining the 20 iiL of Internal
Standard with 1.0 mL of acetone in a
glass test tube. Rinse the test tube
with 0.5 mL acetone, followed by a
second 0.5 mL portion of acetone, and
transfer these rinses to the aqueous
sample.
11.1.6.5 Place a Teflon-coated, magnetic
stirring bar in the sample container,
and stir the aqueous sample using a
magnetic stirplate for 15 minutes to
disperse the spiking solution.
Position the stem of a glass
filtering funnel to discharge into a
pre-cleaned 5 L round bottom flask
and place a paper filter retention
rating, 2.5 u, filter, into the
funnel.
11.1.6.6 Decant and/or pour the internal
standard-spiked water sample from the
2 L bottle into the filter and
collect the filtrate in the 5 L
flask.
11.1.6.7 Rinse the empty 2 L sample container
sequentially with three 100 mL
aliquots of HPLC grade water, pouring
each rinse through the filter, and
collecting the filtrate in the 5 L
vessel. Check to ensure that all
residual particulates and sediments
are removed from the original sample
container by the aqueous rinsing
procedure. Retain this filtrate for
subsequent extraction using the
procedures described in Section 11.2.
11.1.6.8 Transfer the combined filter and
particulate to a clean Petri dish and
place them into a desiccator. Allow
these solids to dry completely (as
indicated by constant weight upon
successive weighings). Retain these
solids for subsequent extraction as
described in Section 11.3.2.
22 393
-------
11.1.6.9 Rinse the original 2 L sample
container sequentially with three 50
mL aliquots of methylene chloride.
Pour the rinsates through the empty
funnel and collect in a clean 1000 mL
glass bottle fitted with a Teflon-
lined lid. Retain this for
subsequent combination with the
methylene chloride extract of the
aqueous filtrate, obtained as
described in Section 11.2.
11.1.7 Exceptional Samples: Some of the samples
received may be too wet to dry efficiently in a
desiccator, but may still not contain
sufficient liquid to permit separation of the
phases by filtering or other such means. Such
samples will be distributed on sheets of
aluminum foil and allowed to air-dry at ambient
temperature on a bench top. For such samples,
the surface area will be recorded, so that a
correction can be made, if this is necessary,
for contamination or cross-contamination of the
samples. The extent of contamination of such
samples will be estimated by placing a filter
paper blank in the same area where these
samples are air-dried and this blank will
subsequently be analyzed for 2,3,7,8-TCDD and
2,3,7,8-TCDF. If the blank is found to be
positive for these compounds, the corresponding
levels of these compounds in the samples will
be corrected for the levels detected in the
blank.
11.2 PROCEDURES FOR EXTRACTING 2,3,7,8-TCDD AND 2,3,7,8-TCDF
FROM AQUEOUS FILTRATE
The internal-standard-spiked, aqueous filtrate
resulting from application of the procedures described
in Section 11.1.6. is extracted utilizing the following
procedures.
11.2.1 Add 400 mL of methylene chloride to the aqueous
filtrate contained in the 5 L flask (from the
step described in Section 11.1.6.7). Place a
magnetic stirring bar into the 5 L flask, place
the flask on a stir-plate, and stir the liquid
in the flask for 16 hours.
11.2.2 Discontinue stirring of the contents of the 5 L
flask, allow the aqueous and organic phases to
separate, then remove the organic layer using a
pipette, and place it in the 1000 mL bottle
containing the accumulated methylene chloride
rinsates collected as described in Section
23
0 f\ A
o d '1
-------
11.1.6.9. At this point, also add to the
contents of this same 1000 mL bottle, one-
fourth of the methylene chloride rinsates
collected as described in Section 11.1.6.3.
Retain the rest of the latter rinsate for
subsequent splitting among the extracts of the
other three splits of the original aqueous
sample, if these are subsequently analyzed.
11.2.3 Sequentially, repeat the extraction of the
aqueous filtrate two additional times using a
100 mL portion of methylene chloride each time,
and combine each extract with the original
extract in the 1000 mL bottle. Reserve this
pooled extract for later combination with the
Soxhlet extract of the particulate and filter,
as described in Section 11.3.
11.3 PROCEDURES FOR SOXHLET-EXTRACTING 2,3,7,8-TCDD AND
2,3,7,8-TCDF FROM DRIED BULK SOLIDS AND FILTERED
WASTEWATER SOLIDS
11.3.1 Dried Sludges, Ash Samples, Wood Chips and
Paper Pulp: Solid samples of these types,
prepared as described in Section 11.1, are
extracted using the following procedures.
11.3.1.1 Prepare a glass Soxhlet extraction
thimble (90 mm by 35 mm) for use by
rinsing it sequentially with
methanol, acetone and methylene
chloride. Add silica to form a 3-6
mm layer on the surface of the glass
frit at the bottom of the thimble,
and place a 10 mm layer of glass wool
over the layer of silica.
11.3.1.2 Prepare a Soxhlet extraction
apparatus, consisting of a Soxhlet
extraction tube, a 250 mL Erlenmeyer
flask and a water-cooled condenser,
for use by rinsing it sequentially
with methanol, acetone, and roethylene
chloride, and allowing it to air-dry.
Place 175 mL of a solution consisting
of 50% benzene and 50% acetone (by
volume) , along with about 10 pre-
cleaned 2 mm glass beads, into the
Erlenmeyer flask. Place the Soxhlet
thimble (prepared as described in
Step 11.2.1.1) into the Soxhlet
extraction tube, assemble the
Soxhlet-extraction apparatus , heat
the contents of the Erlenmeyer flask
to reflux temperature, and continue
24
o
O
-------
the Soxhlet extraction procedure for
a period of 3 hours.
11.3.1.3 Remove the heat source from the
Soxhlet apparatus, allow the
apparatus to cool, and then decant
the benzene/acetone solution into a
clean, 250 mL flint glass bottle and
seal the bottle with a Teflon-lined
screw cap. This solution is retained
in case additional analyses are
required to check the cleanliness of
the Soxhlet apparatus, as a QC
measure.
11.3.1.4 Place a fresh 175 mL aliquot of 50:50
(volume/volume) benzene/acetone into
the Erlenmeyer flask of the Soxhlet
extraction apparatus and re-connect
the Soxhlet extraction tube to the
Erlenmeyer flask. Remove the layer of
glass wool from the glass thimble.
Transfer an accurately weighed
aliquot (approximately 7-10 grams,
depending upon the sample type) of
the previously desiccated solid
sample, prepared as described in
Section 11.1, from the sample bottle
containing the dried sample to the
Soxhlet extraction thimble.
11.3.1.5 Using a microsyringe, add the
appropriate internal standard
solution described in Section 9.3.2
to the solid sample in the Soxhlet
extraction thimble. Place the
previously .removed glass wool (Step
11.3.1.4) on top of the sample in the
glass thimble. Place the condenser on
the Soxhlet extraction tube and heat
the solvent reservoir so that the
extraction solvent refluxes. Soxhlet
extract the sample for a period of 16
hours, then discontinue heating the
apparatus and allow it to cool to
ambient temperature.
11.3.1.6 Remove the Soxhlet extractor from the
Erlenmeyer flask reservoir and
replace the extractor with a 3-ball
Snyder column. Resume heating the
reservoir and concentrate the
benzene/acetone extract to a volume
of about 15 mL. Rinse the Snyder
column twice with small quantities of
o r\ o
25 39
-------
hexane, then continue heating and
concentrating the solution in the
reservoir with the column in place
until a final volume of 10 mL is
attained.
11.3.1.7 Using a 10 mL disposable pipette,
transfer the concentrated solution
obtained in Step 11.3.1.6 to a pre-
rinsed, 125 mL flint glass bottle
fitted with a Teflon-lined screw cap.
Rinse the Erlenmeyer flask four times
using 10 mL aliquots of hexane,
transferring each rinse solution to
the 125 mL bottle, to effect a
quantitative transfer of the
concentrate from the Erlenmeyer flask-
to. the bottle.
11.3.1.8 Proceed with the remainder of the
clean-up and analytical procedures
described in Section 11.4.
11.3.2 Soxhlet Extraction of Filtered Wastewater
Solids
11.3.2.1 Remove the desiccated filter and
associated solids resulting from
filtration of a water/wastewater
sample containing particulates, as
described in Section 11.1 from its
sample container, and immediately
place the filter and solids into a
Soxhlet extraction thimble which has
been pre-cleaned, as described above
in Step 11.3.1.1.
11.3.2.2 Pre-clean a Soxhlet extraction
apparatus as described in Steps
11.3.1.2 through 11.3.1.4.
11.3.2.3 Concentrate the methylene chloride
extract resulting from extraction of
the aqueous filtrate which has been
pooled with other methylene chloride
rinsates (obtained as described in
Section 11.2.2) by transferring about
150 mL of the methylene chloride
extract to a 250 mL Erlenmeyer flask,
attaching a 3-ball Snyder column to
the flask and heating the flask to
concentrate the methylene chloride.
Continue to transfer 150 mL aliquots
of the methylene chloride extract to
the Erlenmeyer flask as each portion
26
O f\'-,
od <
-------
is reduced in volume by
concentration, until the volume of
the extract is reduced to about 25
mL. Then add 150 mL of 50:50
volume:volume benzene-acetone to the
Erlenmeyer flask containing the
residue from the methylene chloride
concentration and reconnect the flask
to the Soxhlet extractor. Note that
it is not necessary to spike this
sample with internal standards since
the wastewater sample was previously
spiked with internal standards prior
to filtering.
11.3.2.4 Heat the Soxhlet apparatus and
extract the filter and solids for a
period of 16 hours, then discontinue
heating and allow the apparatus to
cool. Remove and concentrate the
extract as described in Section
11.2.1.6. Transfer the concentrate
to a new sample bottle, as described
in Section 11.3.1.7.
11.3.2.5 Proceed with the remainder of the
clean-up and analytical procedures
.described in Section 11.4.
11.4 PROCEDURES FOR ISOLATING AND QUANTITATING 2,3,7,8-TCDD
AND 2,3,7,8-TCDF PRESENT IN ORGANIC EXTRACTS OF PAPER
MILL PROCESS AND EFFLUENT SAMPLES
11.4.1 Preliminary Separation of 2,3,7,8-TCDD and
2,3,7,8-TCDF From Other Chemical Residues in
the Extracts Obtained As Described in Sections
11.2 and 11.3. Organic extracts obtained
utilizing the procedures described in Sections
III. and IV. are subjected to the fractionation
procedures which follow.
11.4.1.1 Add 30 mL of aqueous potassium
hydroxide (20% w/v) to the bottle
containing the sample extract, seal
the bottle and agitate it for a
period of 10 minutes on a wrist
action shaker. Aspirate and discard
the aqueous phase, retaining the
organic phase.
11.4.1.2 If the aqueous layer from the
previous step appears to be colored
following the base extraction
procedure, then repeat this operation
(Step 11.4.1.1).
27 •}<
398
-------
11.4.1.3 Add 30 mL of double-distilled water
to the organic phase from Step
11.4.1.1, seal the bottle, and agitate
the mixture for a period of 1 minute.
Again, aspirate and discard the
aqueous phase, retaining the organic
phase.
11.4.1.4 Add 30 mL of concentrated sulfuric
acid to the residual hexane extract
from the previous step, seal the
bottle, and agitate it for a period
of 10 seconds. If emulsions form,
centrifuge the bottle to achieve
separation of the organic and acidic
aqueous phases. Remove and discard
the aqueous acidic layer, retaining
the organic layer.
11.4.1.5 Repeat the concentrated sulfuric acid
wash (Section 11.4.1.4) this time .
adding 30 mL of sulfuric acid to the
sample extract, and agitating the
acidified sample for 10 minutes.
Again, aspirate and discard the
aqueous layer. Repeat this step
until the acid layer is visibly
colorless.
11.4.1.6 Repeat Step 11.4.1.3.
11.4.1.7 Add 5 g of anhydrous sodium sulfate
to the organic extract and allow the
mixture to stand for at least 15
minutes.
11.4.1.8 Quantitatively transfer the organic
extract, using hexane to rinse the
sample bottle, to a clean test tube,
and reduce the volume to
approximately 5 mL by passing a
stream of pre-purified nitrogen over
the extract, while maintaining the
test tube at 55° C in a water bath
(nitrogen slowdown apparatus).
11.4.1.9 Proceed with the liquid column
chromatographic procedures described
in Section 11.4.2.
11.4.2 Liquid Column Chromatographic Procedures for
Isolating 2,3,7,8-TCDD and 2,3,7,8-TCDF From
Extracts Previously Washed With Acids and Bases
28
-------
11.4.2.1 Fabricate a silica gel glass
chromatography-coluron. Pack the
column, in succession, with a plug of
glass wool (silanized), 1.0 g silica,
2.0 g silica containing 28% (w/w) 1 M
NaOH, 1.0 g silica, 4.0 g silica
containing 30% (w/w) sulfuric acid,
and 2.0 g silica.
11.4.2.2 Quantitatively transfer the
concentrated extract obtained in Step
11.4.1.8, along with two rinsings of
the sample container, using 1 mL
portions of hexane each time, to the
column and elute the column with 90
mL of hexane. Collect the entire
eluate and concentrate to a volume of
1-2 mL in a centrifuge tube.
11.4.2.3 If any layer of the silica gel column
implemented in Step 11.4.2.2 becomes
visibly colored as the column is
eluted, repeat Steps 11.4.2.1 and
11.4.2.2.
11.4.2.4 Prepare a liquid chromatography
column by packing the constricted end
with a plug of silanized glass wool
and then adding three grams of Woe 1m
basic alumina.
11.4.2.5 Aspirate the concentrated extract
obtained in Step 11.4.2.2 and
transfer it onto the alumina column
prepared in Step 11.4.2.4. Rinse the
test tube which contained the
concentrate successively with two 1
mL portions of hexane, each time
transferring the rinse solution to
the alumina column.
11.4.2.6 Elute the alumina column as follows:
(a) Elute the alumina column with 10
mL of 3% (v/v) methylene chloride-in-
hexane, taking care not to let the
column become completely dry during
the elution, and discard the entire
eluate. (b) Elute the column with 15
mL of 20% (v/v) methylene chloride-
in-hexane and discard the entire
eluate. (c) Elute the column with 15
mL of 50% (v/v) methylene chloride-
in-hexane, retain this entire eluate,
and reduce the volume to about 1.0 mL
by passing a stream of pre-purified
29
400
-------
nitrogen over the solution while
heating the solution in a 55° C water
bath.
11.4.2.7. Prepare a second alumina column as
described in Step 11.4.2.4, transfer
the concentrated eluate obtained in
Step 11.4.2.6 to the column, and
elute the column as described in Step
11.4.2.6. Collect the column eluate
and concentrate it to a volume of
about 1 mL.
11.4.2.8 Prepare a liquid chromatography
column by cutting off a 9-inch
disposable Pasteur pipette 1.25 cm
above the tip constriction leaving a
straight glass tube with an
indentation approximately 2.5 cm inch
from the top. Insert a filter paper
disk in the tube and position the
disk 2.5 cm below the indentation.
Add a sufficient quantity of PX-21
Carbon/Celite 545 (prepared as
described in Section 7.1.8) to the
tube to form a 2 cm length of the
Carbon/Celite. Insert a glass wool
plug on top of the Carbon/Celite.
Pre-elute the column sequentially
with 2 mL of a 50% benzene/50% ethyl
acetate solution (v/v) , 2 mL of 50%
methylene chloride/50% cyclohexane,
and 2 mL of hexane, and discard these
eluates. Transfer the residual
sample extract (in 1 roL of hexane)
resulting from the alumina column
cleanup (Step 11.4.2.7) onto the top
of the Carbon/Celite column, along
with 1 mL of a hexane rinse of the
original sample vessel. Elute the
column with 2 mL of 50% methylene
chloride/50% cyclohexane solution and
2 mL of 50% benzene/50% ethyl acetate
and discard these eluates. Invert
the column and elute it in the
reverse direction with 4 mL of
toluene, retaining this eluate.
Concentrate the collected column
effluent to a volume of about 1 mL
using a stream of pre-purified
nitrogen.
11.4.2.9 Prepare a third alumina column as
described in Step 11.4.2.4, transfer
the concentrated eluate from the
30 401
-------
second alumina column sequence to
this column, elute the column, and
collect the eluate as already
described in Step 11.4.2.6, in a test
tube. Concentrate the collected
eluate to a volume of about 1 mL,
then quantitatively transfer the
concentrate to a 3 mL micro-reaction
vessel, using two 1 mL portions of
methylene chloride to rinse the test
tube, and also transferring these to
the micro-reaction vessel.
Concentrate the solution in the
latter vessel just to dryness, using
a stream of dry N2 , as described
previously. Rinse the walls of the
micro-reaction vessel using 0.5 mL of
methylene chloride, and again
concentrate just to dryness. Seal
the vessel and store it in a freezer
(-15°C). Just prior to GC-MS
analysis, remove the vessel from the
freezer, allow it to warm to ambient
temperature, and reconstitute the
residue in the vessel by adding 10 uL
of Standard External Standard to the
vial.
11.5 GAS CHROMATOGRAPHIC - MASS SPECTROMETRIC (GC-MS)
PROCEDURES FOR QUANTITATING 2,3,7,8-TCDD AND
2,3,7,8-TCDF PRESENT IN SAMPLE EXTRACTS
Sample extracts prepared by the procedures described in
the foregoing are analyzed by GC-MS utilizing the
instrumentation and operating parameters listed below.
Typically, 1 to 5 uL portions of the extract are
injected into the GC. Sample extracts are initially
analyzed using the DB-5 capillary GC column at a mass
spectral resolution of 1:600 to obtain data on the
concentration of 2,3,7,8-TCDD and to ascertain if
2,3,7,8-TCDF or other isomers which coelute with
2,3,7,8-TCDF are present. If the latter are detected
in this analysis, then another aliquot of the sample is
analyzed in a separate run, using a"^ newly developed
hybrid column which consists of a 10 meter length of a
0.25 mm I.D. fused silica open tubular DB-5 capillary
column coupled with a 30 meter section of a 0.25 mm
I.D. DB-225 column. Again, the mass spectrometer is
operated at low resolution (1:600). The hybrid column
uniquely separates 2,3,7,8-TCDF from the other 37 TCDF
isomers and therefore yields definitive data on the
concentration of 2,3,7,8-TCDF in the extract which is
analyzed. However, in some instances compounds are
present in the sample extract which give rise to ion
masses which, at low (1:600) mass resolution, interfere
31
402
-------
with the quantitation of 2,3,7,8-TCDF. In these
instances the analysis of the sample extract can be
repeated, using the DB-5/DB-225 hybrid column, but this
time at a mass spectral resolution of 1:6,500. The
instrumentation and operating parameters utilized in
these analyses are as follows.
11.5.1 Gas Chromatograph: Carlo Erba Mega 5000 or
Varian 3740
11.5.1.1 Injector: Configured for capillary
column, splitless/split injection
(split flow on 60 seconds following
injection): injector temperature,
280°C.
11.5.1.2 Carrier gas:
For DB-5 column: Hydrogen, 30 Ib.
head pressure (MS-25 jet separator);
18 Ib. head pressure (MS-30 direct
coupled)
For DB-5/DB-225 column: Hydrogen, 30
Ib. head pressure (MS-25 jet
separator); 18 Ib. head pressure (MS-
30 direct coupled)
11.5.1.4 Capillary Column 1: For quantitation
of 2,3,7,8-TCDD (isomer specific) and
2,3,7,8-TCDF (non-isomer specific),
60 M x 0.25 mm ID fused silica coated
with a 0.25 micron film of DB-5,
temperature programmed, see Table 1
for temperature program. Capillary
Column 2: For quantitation of
2,3,7,8-TCDF (isomer specific), 10 M
x 0.25 mm ID fused silica column
coated with a 0.25 micron film of DB-
5 coupled with a 30 M x 0.25 mm I.D.
fused silica column coated with a
0.25 micron film of DB-225. This
column is temperature programmed as
indicated in Table 2.
11.5.1.4 Interface Temperature: 250°C
11.5.2 Mass Spectrometer: Kratos MS-30 or Kratos MS-25
or other suitable
instruments
11.5.2.1 lonization Mode: Electron impact (70
eV)
32
403
-------
11.5.2.2 Static Resolution: 1:3000 to 1:10,000
(10% valley) depending upon
instrumentation.
11.5.2.3 Source Temperature: 250°C
11.5.2.4 Accelerating Voltage: 2KV or 4KV,
depending upon instrument.
11.5.2.5 Ions Monitored: Computer controlled
Selected Ion Monitoring, See Tables 1
and 2 for list of ion masses
monitored and time intervals during
which ions characteristic of 2,3,7,8-
TCDD and 2,3,7,8-TCDF are monitored.
Note that in the case of quantitation
of the 2,3,7,8-TCDF, the
hexachlorinated diphenylether
molecular ion, which could give rise
to an interference at m/z 304 and
306, is also monitored as indicated
in Table 2.
12. CALCULATIONS
12.1 EQUATIONS USED FOR CALCULATING ANALYTICAL RESULTS FROM
THE GC-MS DATA
12.1.1 Equation 1: Calculation of Relative Response
Factor for native 2,3,7,8-TCDD
(RRF2) using »3Ci2-2,3,7,8-TCDD
as an internal standard.
RRF2 = (AsCia /AisCs )
where: Aa = SIM response for 2,3,7,8-
TCDD ion at m/z 320 + 322
Ai 3 = SIM response for * 3 Ci 2 -
2,3,7,8-TCDD internal
standard ion at m/z 332 + 334
Cia = Concentration of the internal
standard (pg./uL.)
C8 = Concentration of the 2,3,7,8-
TCDD (pg./uL.)
404
-------
12.1.2 Equation 2: Calculation of Relative Response
Factor for l3Ci2-1,2,3,4-TCDD
(RRFi, )
RRFb = (Ai s Ces /Acs Ci s )
where: At8 = SIM response for 13Ci2-2,3,7,8-TCDD
internal standard ion at m/z 332 + 334
Acs = SIM response for *3Ci2-1,2,3,4-TCDD
external standard at m/z 332 + 334
Cia = Concentration of the *3Ci2-2,3,7,8-
TCDD internal standard (pg./pL.)
Ces = Concentration of the *3Ci 2-1,2,3,4-
TCDD standard (pg./uL.)
12.1.3 Equation 3: Calculation of concentration of
native 2,3,7,8-TCDD using
13Ci2-2,3,7,8-TCDD as internal .
standard
Concentration, pg./g. = (As) (Is)/(At9)(RRF2)(W)
where: As = SIM response for 2,3,7,8-TCDD ion
at m/z 320 + 322
Ai s = SIM response for the *• 3 Ci 2 -2, 3 ,7 , 8-
TCDD internal standard ion at m/z
332 + 334
Is = Amount of internal standard added to
each sample (pg.)
W = Weight of sample in grams
RRF2 = Relative response factor from
Equation 1
34
405
-------
12.1.4 Equation 4: Calculation of % recovery of
13Ci2-2,3,7,8-TCDD internal
standard
% Recovery = 100 (At s ) (Es ) / (A« » ) (It ) (RRFb )
Ais = SIM response for »3Ci2-2,3,7,8-TCDD
internal standard ion at m/z 332
+ 334
Aes = SIM response for l3Ci2-1,2,3,4-TCDD
external standard ion at m/z 332 +
334
EB = Amount of 13Ci2-1,2,3,4-TCDD external
standard co-injected with sample
extract
It = Theoretical amount of l 3 Ci 2 -2, 3 , 7 , 8--
TCDD internal standard in injection
RRFb = Relative response factor from •
Equation 2
12.1.5 Equation 5: Calculation of Relative Response
Factor for native 2,3,7,8-TCDF
(RRFc) using *3Ci2-2,3,7,8-TCDF
as an internal standard.
RRFc = (As Ci s /Ai s Cs )
where: As = SIM response for 2,3,7,8-TCDF ion
at m/z 304 + 306
Ais = SIM response for *• 3 Ci 2 -2 , 3 , 7 , 8-TCDF
internal -standard ion at m/z 316 +
318
Cis = Concentration of the internal
standard (pg./uL.)
Cs = Concentration of the 2,3,7,8-TCDF
(pg./uL.)
35
4GG
-------
12.1.6 Equation 6: Calculation of Relative Response
Factor for l3Ci2-1,2,3,4-TCDF
(RRFd) (When analysis is performed
using DB-5 Column)
RRFd
where: Ai a
= (Al aCes /Aes Ct s )
= SIM response for 13Ci2-2,3,7,8-TCDF
internal standard ion at m/z 316 + 318
Acs = SIM response for »3Ci2-1,2,3,4-TCDD
external standard at m/z 332 + 334
Cj8 = Concentration of the 13Ci2-2,3,7,8-TCDF
internal standard (pg./uL.)
Ces = Concentration of the *3Ci2-1,2,3,4-TCDD
external standard (pg./uL.)
12.1.7 Equation 7:
where
RRFc
Ais
Acs
Cis
-e s =
Calculation of Relative Response
Factor for 37C«-1,2,7,8-TCDF
(RRFe) (When analysis is
performed using . DB-5/DB-225
Hybrid Column)
(Al 3 Ces /Aes Cls )
SIM response for l3d2-2,3,7,8-TCDF
internal standard ion at m/z 316 +
318
SIM response for 37Cl«-1,2,7,8-TCDF
external standard at m/z 312
Concentration of the l3Ci2-2,3,7,8-
TCDF internal standard (pg/uD
Concentration of the 37Cl«-1,2,7,8-
TCDF external standard (pg/uD
36
407
-------
12.1.8 Equation 8: Calculation of concentration of
native 2,3,7,8-TCDF using l3Ci2 -
2,3,7,8-TCDF as internal standard
Concentration, pg./g. = (As) (Is)/(Aja)(RRFC)(W)
where: A8 = SIM response for 2,3,7,8-TCDF ion at
m/z 304 + 306
Aia = SIM response for the J3Ci2-2,3,7,8-
TCDF internal standard ion at m/z 316
+ 318
Is = Amount of internal standard added to
each sample (pg.)
W = Weight of sample in grams
RRFc = Relative response factor from
Equation 5
12.1.9 Equation 9: Calculation of % recovery of
13Ci2-2,3,7,8-TCDF internal
standard (When analysis is
performed using DB-5 Column)
% Recovery = 100(Ai.) (E. )/(A..) (Ii) (RRF*)
Ais = SIM response for 13Ci2-2,3,7,8-TCDF
internal standard ion at m/z 316 +
318
Acs = SIM response for *• 3 Ci 2 -1, 2 , 3 , 4-TCDD
external standard ion at m/z 332 +
334
E8 = Amount of -1 3 Ci 2-1, 2 , 3 , 4-TCDD external
standard co-injected with sample
extract
Ii = Theoretical amount of * 3 Ci 2-2 , 3 , 7 , 8-
TCDF internal standard in injection
RRFd = Relative response factor from
Equation 6
37 408
-------
12.1.10 Equation 10: Calculation of % recovery of
13Ci2-2,3,7,8-TCDF internal
standard (when analysis is
performed using hybrid DB-5/
DB-225 column
% Recovery = 100 (Ai s ) (E8)/(Aes) (Ii) (RRFe )
Ai 3 = SIM response for 13Ci2-TCDF internal
standard ion at m/z 316 + 318
Aes = SIM response for 37Cl4-1,2,7,8-TCDF
external standard ion at m/z 312
Es = Amount of 37Cl4-1,2,7,8-TCDF external
standard in injection
RRFe = Relative response factor from
Equation 7
12.2 REPORTING RESULTS
The 2,3,7,8-TCDD and 2,3,7,8-TCDF data for solid'
samples are reported on a dry weight basis in picograms
of analyte per gram of solid material, that is, parts-
per-trillion. The 2,3,7,8-TCDD and 2,3,7,8-TCDF data
for liquid samples are reported in femtograms of
analyte per gram of liquid sample (wastewater or
water), that is, parts-per-quadrillion.
13. INTERPRETATION OF RESULTS
13.1 Spectral responses must be observed at both the
molecular and fragment ion masses corresponding to the
ions indicative of TCDD and TCDF (see Tables 1 and 2)
and intensities of these ions must maximize essentially
simultaneously (within + 1 second). In addition, the
chromatographic retention times observed for 2,3,7,8^
TCDD and 2,3,7,8-TCDF must be correct relative to the
appropriate stable-isotopically labelled internal
standard.
13.2 The ratio of the intensity of the response for the
molecular ion, [M]*, to the response for the [M+2]*
ion must be within ±15% of the theoretically expected
ratio for both the native TCDD and native TCDF signals
(for example, 0.77 in the case of TCDD and TCDF;
therefore, the acceptable range for this ratio is 0.65
to 0.89).
13.3 The intensities of the ion signals for either 2,3,7,8-
TCDD or 2,3,7,8-TCDF are considered to be detectable if
each exceeds the baseline noise by a factor of at least
2.5:1.
38
409
-------
13.4 For reliable detection and quantitation of 2,3,7,8-
TCDF, it is also necessary to monitor the molecular ion
of hexachlorinated diphenyl ether which, if present,
could give rise to fragment ions yielding ion masses
identical to those monitored as indicators of the TCDF.
Accordingly, in Tables 1 and 2, the appropriate ion-
mass for hexachlorinated diphenyl ether is specified
and this ion-mass must be monitored simultaneously with
the 2,3,7,8-TCDF ion-masses. Only when the response
for the diphenyl ether ion-mass is not detected at the
same time as the 2,3,7,8-TCDF ion mass can the signal
obtained for 2,3,7,8-TCDF be considered unique.
14. PRECISION AND ACCURACY
14.1 This analytical protocol was found to be satisfactory
for isomer-specific determinations of 2,3,7,8-TCDD and
2,3,7,8-TCDF in a variety of selected pulp and paper
mill sample matrices. Intra-laboratory method
validation experiments for pulp, sludge, and wastewater
samples indicate that the performance of the analytical
method with respect to precision and spike recovery is
demonstrably uniform. The method performance does not
appear to be sensitive to any specific matrix or
chemical effects which might be associated with the
manufacturing processes at a given mill.
14.2 Laboratory precision for the method expressed as
relative percent difference between duplicate analyses
for thirty-five 2,3,7,8-TCDD determinations was 15
percent mean (range 1-138 percent); and for thirty-
three 2,3,7,8-TCDF determinations, 16 percent mean
(range 0-62 percent).
14.3 Field precision for eight 2,3,7,8-TCDD determinations
was 14 percent mean (range 4-19 percent) ; and for nine
2,3,7,8-TCDF determinations, 22 percent mean (range 0-
99 percent).
14.4 For thirty-five 2,3,7,8-TCDD determinations, accuracy
expressed as percent spike recovery was 103 percent
mean (range 66-168 percent); and for thirty-five
2,3,7,8-TCDF determinations, 102 percent mean (range
58-153 percent).
14.5 Including results from method validation experiments,
97 percent of the analyses for a previous study met the
quality assurance objectives for laboratory precision
and accuracy. Ninety-five percent of 133
determinations for 2,3,7,8-TCDD and for 2,3,7,8-TCDF
resulted in analytical data suitable for project
objectives.
39
410
-------
14.6 Target analytical detection levels of 1 ppt for solid
samples were achieved for all but one sample for
2,3,7,8-TCDD and all but one sample for 2,3,7,8-TCDF
(different samples) in a prior study. Target
analytical detection levels of 0.01 ppt for liquid
samples were achieved for all but three samples for
2,3,7,8-TCDD and all but two samples for 2,3,7,8-TCDF
(different samples).
15. REFERENCES
1. "Total Suspended Solids Determination," Method 209C,
Standard Methods for the Examination of Water and
Wastewater, 17th Edition. APHA, AWWA, WPCF, 1986.
2. Tiernan, T. O., Garrett, J. H., Solch, J. G., Wagel, D.
J., VanNess, G. F-. and Taylor, M. L., "Improved
Separation Procedures for Isolating TCDD and TCDF from
Chemically-Complex Aqueous and Solid Sample Matrices
and for Definitive Quantitation of These Isomers at PPQ
to PPT Concentrations," Chemosphere, 1988 (In Press).
40 411
-------
TULS 1
SIQUUCl Of OPKHTIOK II 6C-K-JS QUUTITHIOI Of 2,3,1,8-TCD? 119 2,3,7,8-TCDD
I! EITI10IUIT1L SlIPLES OSIIG 1 (0 I DB-5 COIUII
Elapsed
Tiie dial
0.00
1.00
31.00
31.00
45.00
Erent
Injection, Splitless
Stirt letra Progra:
steep = 350 ppi;
tiie/iiss = 0.08 sec.
CC Coloin Teiperatnre
Teiperatare Projra lite
CO CC/iis)
180
Tin 01 split Tilre;
Bejin Teip Projrai to 240'C
Stop Tetra Projrai
Increase Colou Teip
lallisticallf Projrai
to 300'
Couence Coolinj GC Colou
Oien leiperatore to 180*
180
180
240
240
300
Ipproiinate
Ions Ionitored
by lass
Spectroieter d/z)
240.5378
256.9328
303. 90H
305.8996
311.8857
315.5418
317.5388
315.85(5
321.8535
327.8845
331.5357
333.5337
Theoritical
Identity of
fragient Ion
[I-COC1]'
[1-COC1]'
[1]'
[1*2]'
[1]'
[1]'
[1+2]'
[I]'
[1+2]'
[1]'
[1]'
[1+2]'
Coiposnds latio
Ionitored of [I]
TCDF
TCDD
TCDF
TCDf
"Cl4-2,3,7,8-TCDF
"dHJ^S-TCDr
''Cu-J.J.T.S-TCDJ
TCDD
TCDD
ITCl<.-2,3,7,8-TCDD
"Cu^^.S-TCDD,
llCu-l,2,3,4-TCDD
"Cu^.T^-TCDD,
l»Ctj-l,2,3,4-TCDD
of
!':[!+.
0.77
0.77
0.77
0.77
373.8393
lezaehlorodiphenyl Ether
412
-------
TUIE 2
SEQJEItt OF OPEHTIOIS II CC-IS-DS QDUTIT1TIOI Of 2,3,1,1-TCDF
ii iniioimm SUPLES OSIK i IHIID 01-5/01-225 coimu
Elapsed
Tiie din)
0.00
Event
Injection, Splitless
Start Tetra Prograi:
steep = 350 ppi;
tiie/iass - 0.08 sec.
6C Cohu
Teiperatnre
CO
180
Teiperatore
Projrai late
CC/iin)
Ions loaitored
by lass
Spectroieter d/i)
240.911!
303.9016
305.8936
ni.S»97
315.H18
311.935!
331.9361
333.9337
373.1393
Identity of
Fragient Ion
ll-COCl]'
111*
U+2]'
II]'
HI*
[1+2]'
til*
[1+2]'
[I]'
Coiponnds
lonitored
Ipproxiiate
Tbeoritical
latio of
of [I]':[1*2]'
TCDF
TCDF
TCDF 0.71
»Cl4-l,2,7,HCDF
"Cu^.l.t-TCDF
»»Cn-2,3,7,l-TCIIF 0.77
l'Cn-l,2,3,HCDD
l>Ctt-l,2,3r(-!CDD
lezacilorodipienrl Ether
1.00
41.00
11.00
Tin 01 split talfe; ItO
Begin Teip Projra to 220'C; 110
Stop Tetra Prograi 220
Couence Cooling Cohu 220
to 1JO«
413
-------
ncasl
««.
leal bufetin
NATIONAL COUNCIL Of TH« PAPER MOUSTRY FOR AM AND STREAM «4P*WVEi«W. •«, MO MAOaONAVOJUt NEW YORK. N.Y. 1001 e
eC£r/^D
HAY19
NCASI PROCEDURES FOR THE PREPARATION AND
ISCMER SPECIFIC ANALYSIS OF PULP AND PAPER INDUSTRY S/m£S
FOR 2,3,7,8-TCDD AND 2,37,8-TCDF
TECHNICAL BUUHIN NO, 551
WY1989
414
-------
NATIONAL COUNCIL OF THE PAPER INDUSTRY FOR AIR AND STREAM IMPROVEMENT, INC.
260 MADISON AVE. NEW YORK, N.Y. 10016 (212) 532-9000
Dr. Isaiah (tollman
May 1, 1989 President
(212)5329000
TECHNICAL BULLETIN NO. 551
NCASI PROCEDURES FOR THE PREPARATION AND
ISOMER-SPECIFIC ANALYSIS OF PULP AND PAPER INDUSTRY
SAMPLES FOR 2.3.7.8-TCDD AND 2.3.7.8-TCDF
On discovering the presence of chlorinated dioxins and
furans in pulp mill wastewater treatment plant sludges, the paper
industry embarked on a wide array of investigations into the
source and significance of these unintentional trace materials.
The pursuit of these studies required the development of
reliable, sensitive and accurate analytical procedures. To
provide the industry with the necessary procedures, NCASI
undertook a method development program with the objective of
making considerable analytical procedures capable of the isomer-
specific determination of 2,3,7,8-TCDD and 2,3,7,8-TCDF with
detection limits in the low parts per trillion for solid sample
matrices and low parts per quadrillion for aqueous matrices.
This technical bulletin describes the sample preparation and
analytical protocols that resulted from over two years of
intensive testing and evaluation. The sample preparation
protocol insures the stability and homogeneity of the samples.
The analytical procedure builds on techniques that were developed
as part of the USEPA/Paper Industry Cooperative Dioxin Screening
Study by incorporating (a) matrix-specific extraction procedures,
(b) an innovative quality control step to monitor recoveries in
the cleanup procedure, (c) an alternative method for the isomer-
specific determination of 2,3,7,8-TCDF, and (d) quantitation by
high resolution GC/MS. The procedure has been (a) demonstrated
to be applicable to a variety of pulp and paper industry matrices
and (b) shown to give comparable data to other procedures through
interlaboratory comparison studies.
Development of the analytical protocol was a collaborative
effort between NCASI and Enseco-CAL. The studies were conducted
under the direction of Lawrence E. LaFleur, Organic Analytical
Program Manager, assisted by Kenneth Ramage, Theresa Bousquet and
Robert Brunck of the NCASI West Coast Regional staff and Dr.
Michael J. Miille, Enseco-CAL Division Director, who was assisted
by William J. Luksemberg, Steve Valmores and Bob Peterson.
415
«N«BoMi Council ol 9» P»p«r Indutty tar Air and Sawn knpTMrant he. 1969
-------
- 2 -
Your comments and suggestions on this technical bulletin are
solicited and should be directed to Lawrence LaFleur, Organic
Analytical Program Manager at the West Coast Regional Center,
P.O. Box 458 Corvallis, OR 97339 (503-752-8801), or to this
office (212-532-9000).
Very truly yours,
Isaiah Gellman
416
-------
NCASI PROCEDURES FOR THE PREPARATION AND
ISOMER SPECIFIC ANALYSIS OF PULP AND PAPER INDUSTRY
SAMPLES FOR 2.3.7.8-TCDD AND 2.3.7.8-TCDF
TECHNICAL BULLETIN NO. 551
MAY 1989
ABSTRACT: This technical bulletin describes the sample
preparation and isomer specific analysis procedures
NCASI has developed in support of the industry programs
investigating the source and significance of 2,3,7,8-
TCDD and 2,3,7,8-TCDF in pulp and paper industry
matrices. Part I describes the sample preparation
procedures which insure the stability and homogeneity
of the samples. Part II describes the matrix-specific
extraction, cleanup and isomer-specific GC/MS analysis
procedures. These analytical procedures have been
carefully optimized and thoroughly tested to insure the
reliability of the results. The procedures have been
shown to be capable of achieving detection limits in
the low parts per trillion (generally around one ppt)
for solid matrices and in the low parts per quadrillion
(generally less than ten ppq) for liquid matrices.
Through extensive interlaboratory comparison studies
the NCASI procedure has been shown to give comparable
results to other procedures. In particular, USEPA has
recognized that the NCASI procedures are equivalent to
the procedures used in the USEPA/Paper Industry
Cooperative Dioxin Screening Study (the five Mill
Study).
KEYWORDS: 2,3,7,8-TCDD, 2,3,7,8-TCDF, analytical procedure,
isomer specific, extraction procedure, clean-up
procedure.
RELATED NCASI PUBLICATIONS:
(1) U.S. Environmental Protection Agency/Paper Industry
Cooperative Dioxin Screening Study. NCASI Technical
Bulletin No. 545, May, 1988
417
-------
TABLE OF CONTENTS
Page
I INTRODUCTION 1
PART I - NCASI PROCEDURES FOR THE PREPARATION
AND ISOMER-SPECIFIC ANALYSIS OF PULP
AND PAPER INDUSTRY SAMPLES FOR
2,3,7,8-TCDD AND 2,3,7,8-TCDF 3
PART II - NCASI METHOD TCDD/F - 88.01 12
II LITERATURE REFERENCES 57
418
-------
NCASI PROCEDURES FOR THE PREPARATION AND
ISOMER SPECIFIC ANALYSIS OF PULP AND PAPER INDUSTRY
SAMPLES FOR 2.3.7.8-TCDD AND 2.3.7.8-TCDF
I INTRODUCTION
Following the announcement that 2,3,7,8-TCDD had been
detected in pulp and paper mill wastewater treatment plant
sludges, NCASI undertook an evaluation of dioxin analysis
contract laboratories to determine which, if any, might be able
to provide the analytical support required by the industry as
investigations of these findings were undertaken. The evaluation
was based on experience, demonstrated performance and available
resources. Once a suitable laboratory was selected, a program
was begun to test the applicability of existing analytical
procedures for the analysis of trace levels of 2,3,7,8-TCDD and
related compounds in pulp and paper industry samples.
Quite early, it was recognized that commonly employed
cleanup and low resolution MS analysis procedures were inadequate
for the determination of 2,3,7,8-TCDD and related compounds in
pulp and paper industry matrices at the low parts per trillion
and low parts per quadrillion detection limits which were
required. An extensive analytical methods development program
was therefore undertaken jointly by NCASI and Enseco-Cal Labs.
The methods development work undertaken by NCASI/Enseco Cal
was conducted concurrently with the methods development work that
culminated in the analytical protocol used in the USEPA/Paper
Industry Cooperative Dioxin Screening Study (1). Thus, the
NCASI/Enseco Cal analytical protocol presented in Part II,
represents a combination of many features of the USEPA/Paper
Industry Dioxin Screening Study plus features which resulted
directly from the continued research conducted by NCASI and
Enseco Cal. Although the primary objective of the methods
development program was an analytical protocol capable of
providing quality analytical data which meet all the low
detection limit and QA/QC criteria, it was also realized that the
techniques and procedures used in the NCASI procedure were more
along the lines used for other analytical protocols commonly used
by contract laboratories. Thus, it is believed that other
laboratories experienced in high resolution gas chromatographic/
high resolution mass spectrometric methods for analyses for
2,3,7,8-TCDD should be readily able to perform this procedure.
The NCASI procedure is similar to the USEPA/Paper Industry
Cooperative Dioxin Screening Study procedure in that it
incorporates the techniques of (a) spiking the internal standards
into the aqueous samples using a water miscible solvent, and (b)
filtration and soxhlet extraction of the filtered solids as part
419
-------
- 2 -
of the wastewater extraction procedures. Studies conducted
parallel to, and following, the USEPA/Paper Industry Cooperative
Dioxin Screening Study developed matrix-specific extraction
procedures for solid samples to further optimize extraction
efficiencies (2). Enseco Cal was able to demonstrate an
alternative isomer-specific procedure 2,3,7,8-TCDF which utilized
a 30 M DB-225 column. The point of addition of 37Cl4-2,3,7,8-TCDD
was changed to be immediately prior to any cleanup procedures.
This .innovative use of the CI^-2,3,7,8-TCDD provides additional
insight into the extraction efficiency and is a useful QC tool
for troubleshooting analytical problems. The GC/MS quantitation
procedures were intentionally selected to follow along the lines
of generally accepted techniques for high resolution GC/MS dioxin
analysis procedures. This facilitates the use of the procedure
by other contract laboratories and assures the acceptance of the
procedure by regulatory agencies.
The objectives of this technical bulletin are to describe
(a) the NCASI procedures for the preprocessing and homogenization
of samples intended for PCDD/PCDF analysis, and (b) the
NCASI/Enseco Cal analytical procedures themselves. Thus, the
technical bulletin is organized into two parts, each presenting
the details relevant to these objectives. By making these
procedures available to other laboratories, it is hoped that this
will facilitate expansion of the number of laboratories capable
of analyzing pulp and paper industry samples. This will in turn,
enhance the ability of the pulp and paper industry to conduct its
many programs investigating the source, potential control and
significance of the trace levels of chlorinated dioxins and
furans associated with the production of bleached pulp.
A summary of the performance characteristics of the
procedure is in preparation. Included will be discussions of (a)
accuracy as demonstrated by matrix spike recovery data and
interlaboratory comparisons, and (b)" precision as measured by
routine laboratory duplicates and intra- and inter- batch
precision experiments. In the interest of making a procedure,
which has been carefully optimized, validated and thoroughly
tested, immediately available for use by the industry, it was
decided that it would be best to publish the analytical
procedures as soon as possible, to be followed by distribution of
a second report once the performance characteristics have been
summarized.
42U
-------
- 3 -
PART I - NCASI DIOXIN PROGRAM SAMPLE PREPARATION/PROCESSING
I SAFETY GUIDELINES
The analyst should be familiar with the location and proper
use of all safety equipment throughout the building (e.g. fire
extinguishers, respirators, spill kits, etc.). Recommendations
on the use of specific safety precautions are referenced, where
appropriate, in the following procedures. These recommendations
include the use of fume hoods for solvents or the processing of
samples with nuisance odors and dust masks and/or glove boxes to
prevent the inhalation of particulate matter.
II GENERAL LAB PROCEDURES
Under no circumstances should a sample be touched, stored or
in any way come in contact with any materials other than those
prescribed below and then only after they have been properly pre-
pared. Aluminum foil or unpowdered latex gloves require no pre-
treatment but fresh foil or a new pair of gloves should be used
for each situation.
Ill CLEANING PROCEDURES
A. Solvent Cleaning
All materials (except aluminum foil and latex gloves) which
come in contact with the sample (restricted to glass, stainless
steel and Teflon) shall be solvent cleaned. Only Teflon squeeze
bottles are to be used.
The following general cleaning procedure will be followed:
1. Soap and tap water wash all items using Pierce RBS-35 soap
(20 mL RBS-35 per liter of tap water). Rinse with tap water
followed by deionized water.
2. Methanol (Burdick and Jackson) rinse.
3. Acetone (Burdick and Jackson) rinse.
4 Methylene chloride (Burdick,and Jackson) rinse.
5. Air dry.
Used solvents should be stored in separate bottles marked
"Used Methanol," "Used Acetone," and "Used DCM." Conduct solvent
rinsing in a hood.
Revision 4, 6/1/88
42.1
-------
- 4 -
B. Glove Box Cleaning Procedure
The following cleaning procedure should be used prior to and
between each sample when using the glove box for sample grinding
or sample splitting:
1. Vacuum all interior surfaces of the glovebox.
2. Spray all interior surfaces with tap water and wipe down all
inside surfaces with a wet sponge or squeegee.
3. Use a sponge to remove excess H2O from floor of glove box.
If necessary an electric blow dryer can be used to speed up
the drying process.
4. The neoprene glove box sleeves will be vacuumed, wet wiped
with a sponge and air dried. A clean pair of latex gloves
will be placed over them prior to processing any sample.
C. Cleaning of Cabinets and Fume Hoods Used For Drying
Drying cabinets and fume hoods should be cleaned between
usage by vacuuming, wiping all interior surfaces with a wet
sponge or sponge mop, and then should be left to air dry. The
vent in the drying cabinet will be wiped clean with a wet sponge
monthly. More frequent cleaning of the vent is required if the
analyst observes accumulated dust or particulate between clean-
ings .
D. Cleaning of Blender Motor
The blender motor should be dismantled and cleaned monthly
or any time the blender is dismantled for maintenance.
E. Wiley Mill Cleaning Procedure
The following cleaning procedure should be used prior to and
between each sample when using the Wiley Mill for sample
grinding.
1. Vacuum all outside surfaces of the mill. Remove the glass
plate, stationary blades and rotor blade.
2. Vacuum the chamber, hopper inlet, and blade slots. Rinse
all areas with deionized water and scrub with a sponge or
sponge swab. Rinse using Methanol, Acetone and Hexane
catching waste in a 25 mL beaker.
3. Dry using a forced air heater.
Revision 4, 6/1/88
42
-------
- 5-
4. Place small metal disc and screw in a stainless steel screen
basket for cleaning. Use solvent cleaning procedure (Section
III) .
5. Reassemble the mill. Check alignment of the blades by plac-
ing a strip of aluminum foil over each stationary blade and
turning rotor shaft clockwise. Adjust the stationary blades
so they touch the foil but don't cut it.
F. Laboratory Cleaning
The analyst should observe daily the general cleanliness of
the laboratory and look for accumulations of dust or particulate
in the room. If necessary wipe surfaces with a wet sponge and
maintain an uncluttered work area. Every two weeks a more tho-
rough inspection and cleaning of the laboratory should be con-
ducted and all possible work surfaces should be cleaned with a
wet sponge.
IV RECORD KEEPING
All processing of any dioxin samples should be recorded in
an appropriate laboratory notebook in indelible ink. The proces-
sor must date and initial each entry corresponding to a
processing step.
V SAMPLE HANDLING
A fresh pair of unpowdered latex gloves should be used for
each sample and should be discarded after use. Reasonable
efforts should be taken to protect samples from direct light.
Thus the lights should be turned off in cabinets and fume hoods
used for drying samples except when required for handling and
inspection. A dust mask should be worn during any processing
where the inhalation of particulate matter is possible (e.g.
during grinding of samples). Samples producing a nuisance odor
should be handled with proper ventilation. Samples with an
obvious chlorine odor should be prepared and dried in a fume
hood.
VI SOLID/SEMI-SOLID SAMPLE PREPARATION (DRYING TECHNIQUES)
The following is a general procedure for processing samples
that require drying. All air drying of samples must be done in a
drying cabinet or fume hood. When air drying samples the cabinet
or hood should be turned on and the doors closed. During the
Revision 4, 6/1/88
4 - M
i f«*/ »_*
-------
- 6 -
evenings when the janitors are scheduled to come in or when acti-
vity in the room may increase particulate levels in the air, turn
the cabinets or hoods off with the doors closed. Samples are
placed in the cabinets or hoods beginning with the top shelf
until all shelves are full. If samples dry at varying rates no
additional samples will be added until the last sample is dry and
the cabinet or hood is cleaned. The samples on each shelf are
segregated by a physical barrier.
A. Blanks for Drying
An 8" x 10" Gelman glass fiber filter type A should be
placed in the center of the cabinets or hoods for drying prior to
placing samples in the cabinet. The filter sheet should not be
pre-treated. Place the filter sheet on a piece of aluminum foil,
edges folded up and label the foil with the date and time exposed
in the laboratory. Barriers should separate the blank from sam-
ples on the same shelf in a manner analogous to the way samples
are segregated. At the conclusion of the drying of all the sam-
ples in the cabinet or hood, the blank should be folded so as to
cover the exposed upper surface and should be wrapped in aluminum
foil until it is blended. Just prior to blending, the blank fil-
ter should be torn into small pieces and placed in the blender.
Blend as described in Section VIII. Wrap the entire blended
blank filter (i.e. do no split the blended filter) in aluminum
foil and place the foil packet into an I-Chem bottle. Do not
label the blank until it has been put into the I-Chem bottle.
Record the blank preparation, cabinet number, glove box number,
the dates exposed, blended and bottled, and sample identification
number assigned.
B. Sample Typing and Preparation
Sample types fall into the following categories as
identified on the analysis request form which should accompany
each sample set. Samples received without an analysis request
form should be referred to the supervisor for further
clarification.
(1) Total Sample - Refers to the entire sample. The liquid and
solid phases are thoroughly mixed to form a homogeneous mix-
ture. Proceed with Section VI, Part C.
(2) Solids Only - Refers to the solid phase only. The analyst
squeezes, or decants, and centrifuges the sample to remove
and discard the liquid phase.
Friable samples such as pulp mats should be removed from the
sample jar and hand squeezed to remove as much water as possible,
discarding the water.
Revision 4, 6/1/88
^v 424.' •
-------
- 7 -
For slurry type samples, decant any liquid phase observed
after settling. Use a clean stainless steel spatula to transfer
the remaining semi-solid to precleaned centrifuge tubes. Centri-
fuge the sample for three to five minutes at full speed, decant
and discard the supernatant. Transfer the solid material as per
Section VI, Part C.
C. Drvina Samples
After typing the sample and dewatering if necessary, the
analyst should visually inspect the samples moisture content,
texture, etc., and thereby determine if the sample would be amen-
able to drying on screens, pie plates or sheets of foil as des-
cribed below. If the examination reveals obvious inert materials
(e.g. rocks) the entire sample should be weighed and those mater-
ials greater than 1/4 inch diameter should be removed, record the
weight and calculate the percent inert material. Discard the
inert material.
(1) Screens (moderate moisture and large pieces or corrosive
samples) - Break the sample into small pieces (about dime
size) and lay out on a stainless steel screen supported
about 1 cm above a sheet of aluminum foil and transfer to a
drying cabinet. The size of the foil should at least equal
the area of the screen to catch any fines that may fall
through. Wooden dowels wrapped in fresh aluminum foil can
be used to support the screen over the foil. Save the
sample bottle, leaving the cap off until the inside moisture
evaporates, for NCASI sample archives.
Label the foil with the sample identification number and the
time and date the sample was laid out in the cabinet to dry.
This information and the drying cabinet number should also
be recorded.
(2) Pyrex Pie Plates (Hicrh Moisture Slurry Type Samples and/or
Fine Particles) - The sample is transferred to a properly
cleaned pyrex pie plate(s) and placed in the drying cabinet.
Label the plate(s) with the sample identification number and
the time and date. This information and the drying cabinet
number should also be recorded.
(3) Aluminum Foil (Low Moisture and/or Fine Particles) - Spread
the sample on a sheet of aluminum foil, edges folded up,
break up large pieces to at least dime size and place in the
drying cabinet. Label the foil with the sample
identification number and the time and date. This
information and the drying cabinet number should also be
recorded.
Revision 4, 6/1/88
-------
- 8 -
On a daily basis, check to see if the sample is completely
dry and if not turn the material and further break it up into
smaller pieces to facilitate drying.
When the sample is completely dry, fold aluminum foil over
the sample to cover it while waiting to grind. When screens are
used transfer the sample to the aluminum foil base and wrap for
grinding. Record the date and time the sample was wrapped up.
If the dried sample is not ground immediately store the covered
sample in the dry sample storage cabinet.
Grind, quarter and archive the dried sample following
General Procedures VIII, Part A or B, and IX.
VII PAPER PRODUCT PREPARATION
The following is a general procedure for sub-sampling and
compositing paper products. The working surfaces should be cov-
ered with aluminum foil and equipment (e.g. scissors) coming in
contact with the paper product should be properly cleaned. (See
Cleaning Procedures, Section III).
A. Paper Product Blanks
Tear an 8" x 10" Gelman Glass Fiber Filter Type A into small
pieces and grind as described in Section VIII, Part A or B.
After the filter is completely blended, empty the entire
contents of the blender onto a fresh piece of aluminum foil.
Carefully fold the foil over the blended filter material to
form a packet which will effectively enclose all the material so
none will be lost.
Place the foil packet in an I-Chem bottle, label the bottle
and seal with a signed and dated custody seal.
B. sub—Sampling of Paper Products
(1) Paper specimens (e.g. sheets or rolls) - cut out and weigh a
measured aliquot (e.g. 2" x 2" area) of paper specimen
representative of each paper product source. Calculate and
record the weight per area. Use this data to determine the
number of equal aliquots from all sources required to
prepare a composite of approximately 15 grams. If a QA/QC
duplicate and spike are required prepare an additional 30
gram composite.
Revision 4, 6/1/88
42G
-------
- 9 -
Prepare an equal number (as determined above) of measured
aliquots from each of the different sources of paper
specimens which will make up the composite. If the amount
of paper specimens provided will exceed the 15 gram aliquot
required cut and discard the outer 1/2 inch of the roll or
sheet to minimize the potential for contamination which may
have occurred from the original cutting or shipment of the
sheet or roll. The measured aliquots (areas) should be cut
randomly from various different paper specimens so as to
provide a representative sub-sample. Note that the
composite will be comprised of equal measured aliquots
(areas) not equal weights.
(2) Paper Products (e.g. cups, plates) - Weigh one unit (l cup)
of paper product representative of each paper product
source. Record the weight per unit. Use this data to
determine the number of equal units from all sources re-
quired to prepare a composite of approximately 15 grams. If
a QA/QC duplicate and spike are required prepare an
additional 30 gram composite.
Take an equal number (as determined above) of units from
each of the different sources of paper products which make
up the composite. The units should be randomly selected
from the various paper products so as to provide a
representative sub-sample of the material. Note that the
composite will be comprised of equal units, not equal
weights.
C. Compositing of Paper Products
Combine the aliquots or units of paper product sub-sampled
as described above (Section VII, Parts B., 1 and 2) and cut each
to a size appropriate for grinding. Approximately l" x 1" for
the Waring Blender procedure (Section VIII, Part A) or approxi-
mately l" strips for the Wiley Mill procedure (Section VIII, Part
B).
Grind the composited paper product sub-samples following
General Procedures Section VIII, Part A or B.
VIII SAMPLE GRINDING
The following are general procedures for grinding Paper
Specimens/Products dried pulps, sludges and compost sample.
Since each matrix presents its own grinding difficulties it is up
to the discretion of the analyst to determine the appropriate
method. Typically the Waring Blender procedure (Part A) is
amenable for grinding most dried pulp and sludge samples.
Revision 4, 6/1/88
-------
- 10 -
Some sludge samples when ground in the Waring Blender will
generate both fines and a fluff type product creating
difficulties in achieving a homogeneous mixture. Sludge samples
of this nature should be ground in the Wiley Mill.
For paper specimens/products the Wiley mill procedure is
generally more suitable for grinding.
A. Waring Blender
The grinding (or blending) of a sample should be conducted
in the glove box. The working surface of the glove box should be
covered or lined with aluminum foil. The door to the glove box
room should be closed and traffic through the room minimized.
The processes of air drying and blending will be separated by as
many physical barriers as possible (i.e. separated on different
floors).
Blend the entire sample in a properly cleaned blender (see
Section III) . Be sure not to add too much sample into the
blender at one time otherwise blending will not be uniform and
the blender motor may overheat causing fragments of the blender
to mix into the sample. To check for overheating press the
bottom of the blender assembly with gloved hands. If the
assembly feels warm discontinue grinding until cool. Place the
blended sample on a sheet of aluminum foil in the glove box.
Proceed with Section IX if an archived sample is required.
B. Wilev Mill
The grinding of a paper product sample should be conducted
in the glovebox room. The grinding of sludge samples should be
conducted in a fume hood or glovebox. The door to the glovebox
room should be closed and traffic through the room minimized.
The processes of air drying and grinding will be separated by as
many physical barriers as possible (i.e. separated on different
floors).
Blend the entire sample in a properly cleaned Wiley Mill
(see Section III). If an archived sample is required place the
blended sample on a sheet of aluminum foil in the glove box and
proceed with Section IX.
IX QUARTERING AND ARCHIVING
Thoroughly mix the blended sample, using gloved hands or a
stainless steel spoon, by turning the entire sample at least
three times, then form into a conical pile. Carefully flatten
Revision 4, 6/1/88
428
-------
-li-
the conical pile to a uniform thickness and diameter (as wide as
spatially possible) by pressing down the apex. Divide the
flattened mass into four equal quarters. Refer to ASTM "Standard
Methods for Reducing Field Samples of Aggregate to Testing Size."
An oven dried solids determination (103-105°C) is required
for solid/semi-solid samples. Subsample each quarter and place
on a small piece of foil to be transferred to a pre-tared
crucible. A minimum of .5 grams is required. Refer to Standard
Methods 209F pg. 99-100, of 16th (1985) edition.
Repeat cycle of drying, cooling, desiccating, and weighing
until a constant weight is obtained or until weight loss reflects
0.1 percent or less difference in the final percent oven dried
solids calculation. Combine the opposing wedges into separate
I-Chem jars (i.e., two opposite wedges per jar). If more than
two containers are required successively mix and quarter the
opposing wedges until the sample aliquot is reduced to the size
needed.
Label the jars with the sample code.
X DRY SAMPLE PROCESSING
Ash samples (if dry) require no processing except to split
the sample for archives. The samples will be split in the glove-
box using the quartering technique (see ASTM Standard Methods for
Reducing Field Samples of Aggregate to Testing Size). Combine
the opposing wedges into separate I-Chem jars (i.e., two opposite
wedges per jar) and label with the sample code. The sample por-
tion for archives.
XI LIQUID SAMPLE PROCESSING
For liquid samples which do not require filtering no further
preparation is required. For samples which require analysis of
solids and/or liquids only, follow the specified protocol for
waste water filtration.
XII SAMPLE STORAGE
All samples other than ash, paper products, bleached pulps
and processed blanks are refrigerated (4°C).
Revision 4, 6/1/88
-------
-12-
PART II - NCASI METHOD TCDD/F - 88.01
ANALYTICAL PROCEDURES FOR THE ANALYSIS OF PULP AND PAPER INDUSTRY
WASTEWATERS, SOLID WASTES, ASHES AND BLEACHED PULPS FOR 2,3,7,8-
TETRACHLORODIBENZO-p-DIOXIN AND 2,3,7,8-TETRACHLORODIBENZOFURAN
1. Scope and Application
1.1 This method provides procedures for the detection and
quantitative measurement of 2,3,7,8-tetrachlorodibenzo-p-
dioxin (2,3,7,8-TCDD) and 2,3,7,8-tetrachloro-dibenzofuran
(2,3,7,8-TCDF) in pulp and paper industry wastewaters, solid
wastes, ashes and bleached or partially bleached pulps.
Detection limits for the solid sample matrices are in the
low parts per trillion (ppt) and in the low parts per
quadrillion (ppq) for the wastewater matrices. The
analytical method calls for the use of high-resolution gas
chromatography and high-resolution mass spectrometry
(HRGC/HRMS) on purified sample extracts. Table 1 lists the
sample types covered by this protocol, Method Calibration
Limits (MCLs) for 2,3,7,8-TCDD and 2,3,7,8-TCDF and other
germane information.
1.2 The sensitivity of this method is dependent upon a number of
factors, including (but not limited to) the level of
interferents within a given matrix, the recovery of the
internal standard, the instrument sensitivity, etc. If an
analyte is not detected, estimated limits of detection will
be reported for each individual analysis.
1.3 This method is designed for use by analysts who are
experienced with residue analysis and skilled in high-
resolution gas chromatography/high resolution mass
spectrometry (HRGC/HRMS).
1.4 Because of the extreme toxicity of these compounds, the
analyst must take the necessary precautions to prevent
exposure to himself or herself, or to others, of materials
known or believed to contain 2,3,7,8-TCDD or 2,3,7,8-TCDF.
It is the laboratory's responsibility to ensure that safe
handling procedures are employed.
2. Summary of the Method
2.1 This procedure uses a matrix-specific extraction, analyte-
specific cleanup, and isomer specific high-resolution
capillary column gas chromatography/high-resolution mass
spectrometry (HRGC/HRMS) techniques.
NCASI Method TCDD/F - 88.01
430
-------
-13-
2.2 If interferents are encountered, the method provides
selected cleanup procedures to aid the analysts in their
elimination.
2.3 An aliquot of a blended solid waste, ash or pulp is spiked
with a solution containing specified amounts of each of the
two isotopically (13C^2> labeled TCDD/TCDFs. The sample is
then extracted according to a matrix-specific extraction
procedure. The extraction procedures are: (a) 95% ethanol
Soxhlet extraction for pulps, (b) 68:32 ethanol/toluene
Soxhlet extraction for waste treatment plant sludges, and
(c) benzene or toluene Soxhlet extraction for ashes.
Wastewaters are first spiked with a solution containing
specified amounts of each of the two isotopically (13C12)
labeled TCDD/TCDFs and are then stirred or shaken for thirty
minutes to allow equilibration of the spike. The sample is
then filtered through a glass fiber filter. The filtered
solids and filter are soxhlet extracted for 16 hours using
68:32 ethanol/toluene. The filtered wastewater is liquid-
liquid extracted first with methylene chloride then with
toluene. The extracts from the filtered solids and the
aqueous portions are then combined prior to cleanup and
analysis. Immediately following extraction and prior to any
extract cleanup procedures, the extracts are spiked with a
cleanup recovery standard (37Cl4-2,3,7,8-TCDD) to monitor
losses through the cleanup.
2.4 The extracts are then submitted to an acid-base washing
treatment and dried. Following a solvent exchange step, the
residues are cleaned up with two or more column
chromatographic procedures. Just prior to HRGC/HRMS
analysis, a recovery standard (13Ci2~l/2,3,4-TCDD) is added
to all final extracts.
2.5 Two uL of the concentrated extract are injected into an
HRGC/HRMS system capable of performing selected ion
monitoring at resolving powers of at least 10,000 (10
percent valley definition).
2.6 The identification of 2,3,7,8-TCDD and 2,3,7,8-TCDF is based
on their elution at their exact retention time (-1 to +3
seconds from the respective internal standard signal), the
simultaneous detection of the M~*" and M+2* ions, on a
comparison of the ratio of the integrated ion abundance of
these ions to their theoretical abundance ratio and peak
signal to noise ratio of >2.5:1 for both ions.
NCASI Method TCDD/F - 88.01
431
-------
-14-
3. Definitions
3.1 Internal Standards: 13c12-2,3,7,8-TCDD and 13C12-2,3,7,8-
TCDF labeled internal standards are used in this method for
the quantitation of the corresponding native analytes. The
internal standards are added to all samples including method
blanks and quality control samples prior to extraction.
3.2 Cleanup Recovery standard (Table 2); One standard (37cl4-
2,3,7,8-TCDD) is used to determine the recovery throughout
the cleanup procedure. It is added to the sample extract
immediately following extraction and prior to any cleanup
procedures.
3.3 Recovery Standard: One recovery standard (^Ci 2-1,2,3,4-
TCDO) is used to determine the percent recoveries for the
13C12-2,3,7,8-TCDD and 13C12-2,3,.7,8-TCDF internal standards
and the 37Cl4-2,3,7,8-TCDD cleanup recovery standard. It is
added to the final sample extract immediately prior to
HRGC/HRMS analysis.
3.4 Calibration Standard Solutions (CS1 to CSS, Table 4): A set
of a minimum of five solutions are prepared for instrument
calibration. Each solution containing known amounts of
2.3,7,8-TCDD and 2,3,7,8-TCDF, the internal standards
(^3Ci 2~labele<* TCDD and TCDF), the cleanup recovery standard
(37Cl4-2,3,7,8-TCDD), and the recovery standard (13C12~
1,2,3,4-TCDD). The calibration solutions are used to
determine the instrument response for the native analytes
relative to the internal standards and for the internal
standards and cleanup recovery standard relative to the
recovery standard.
3.5 Standard Sample Fortification Solution (IS, Table 2);
Solution containing the two internal standards (i;iC12-
2,3,7,8-TCDD and 13C12-2,3,7,8-TCDF), which is used to spike
all samples before extraction and cleanup.
3.6 Cleanup Recovery Standard Spiking Solution (CR, Table 2):
Solution containing the cleanup recovery standard (^'Cl^
2,3,7,8-TCDD), which is added to the sample extract
immediately following liquid-liquid extraction or soxhlet
extraction prior to any cleanup steps.
3.7 Recovery Standard Spiking Solution (Table 2); Solution
containing the recovery standard (^3Ci2-l,2,3,4-TCDD)' which
is added to the final sample extract before HRGC/HRMS
analysis.
NCASI Method TCDD/F - 88.01
432
-------
-15 -
3.8 GC Column Performance Check Mixture(s) (Table 3): A
solution (or solutions) containing a mixture of selected
TCDD/TCDF standards/ which is used to demonstrate continued
acceptable performance of the capillary column (i.e., <25
percent valley separation of 2,3,7,8-TCDD from all other 2.1
TCDD isomers).
3.9 Relative Response Factor: Response of the mass spectrometer
to a known amount of an analyte relative to a known amount
of an internal standard.
3.10 Laboratory Method Blank: This blank is prepared in exactly
the same manner as a sample, performing all analytical
procedures except the addition of a sample aliquot to the
extraction vessel.
3.11 Estimated Level of Method Blank Contamination: The response
from a signal occurring at the retention times for either
2,3,7,8-TCDD or 2,3,7,8-TCDF and at any of the masses
monitored is used, as described in Section 12, to calculate
the level of contamination in the method blank. The results
from such calculations must be reported along with the data
obtained on the samples belonging to the batch associated
with the method blank. Reporting a method blank
contamination level for 2,3,7,8-TCDD or 2,3,7,8-TCDF that
either equals or exceeds the level present in any one sample
from the batch or, which exceed the target detection limits
would invalidate the results and require sample reruns for
all positive samples found in that batch of samples. A
positive sample is defined as a sample found to contain
either 2,3,7,8-TCDD or 2,3,7,8-TCDF.
3.12 Sample Rerun: Extraction of another aliquot of the sample
followed by extract cleanup and extract analysis.
3.13 Extract Reanalysis: Analysis by HRGC/HRMS of another
aliquot of the same final extract.
3.14 Mass Resolution Check: Standard method used to demonstrate
a minimum static resolving power (10 percent valley
definition) of 10,000.
3.15 Method Calibration Limits (MCLs): For a given sample size,
the final extract volume, and the lowest and highest
concentration calibration solutions, the lower and upper
MCLs delineate the region of quantitation for which the
HRGC/HRMS system was calibrated with standard solutions.
NCASI Method TCDD/F - 88.01
M fl •)
4 o o
-------
- 16 -
3.16 HRGC/HRMS Solvent Blank: This additional QC check analysis
corresponds to a 2 uL injection of pure tetradecane into the
GC column and a complete HRGC/HRMS analysis. Such QC check
can be used following the analysis of a method blank which
has shown detectable levels of the target analytes in order
to verify that the contamination was in the extract and was
not the result of syringe, septum and/or solvent
contamination. Acceptable method blanks or HRGC/HRMS
solvent blanks (Section 9.1 for guidelines) must be obtained
before pursuing the analysis of subsequent sample extracts.
4. Interferents
4.1 Solvents, reagents, glassware, and other sample processing
hardware may yield discrete artifacts and/or elevated
baselines which may cause misinterpretation of
chromatographic data. All of these materials must be
demonstrated to be free from interferents under the
conditions of analysis by running laboratory method blanks.
4.2 The use of high purity reagents and solvents helps to
minimize interference problems. Purification of solvents by
distillation in all glass systems may be required.
4.3 Interferents co-extracted from the sample may vary
considerably from sample to sample. 2,3,7,3-TCDD and
2,3,7,8-TCDF are associated with other interfering
chlorinated compounds which may be found at concentrations
several orders of magnitude higher than that of the analytes
of interest. Retention times of the target analyte must be
verified using reference standards. These values must
correspond to the retention time criteria established in
Section 11.3.
4.4 A 60 m DB-5 fused silica capillary column is used to resolve
2,3,7,8-TCDD from all other TCDD isomers. A 30 m DB-225
fused silica capillary column is used to resolve 2,3,7,8-
TCDF from all other TCDF isomers.
5. Apparatus and Materials
5.1 Gas chromatograph/mass spectrometer data system.
5.1.1 Gas chromatograph: An analytical system with a
temperature-programmable gas chromatograph and all
required accessories including syringes, analytical
columns, and gases. A Grob type splitless injector is
recommended for use with high resolution columns.
NCASI Method TCDD/F - 88.01
434
-------
- 17 -
5.1.2 A60mx0.25mm DB-5 fused silica capillary column is
required for the analysis of 2,3,7,8-TCDD and a 30 m x
0.25 mm DB-225 fused silica capillary column is
required for the analysis of 2,3,7,8-TCDF unless the
laboratory can achieve all isomer specificity and
detection limit requirements with 0.32 mm DB-5 or DB-
225 columns.
5.1.3 Mass spectrometer: A high resolution instrument
capable of performing selected ion monitoring at
resolving powers of at least 10,000 (10 percent valley
definition) and utilizing electron impact ionization is
specified. The system must be capable of selected ion
monitoring (SIM) for at least eleven ions
simultaneously, with a total cycle time of 1 second or
less. At a minimum, the ions listed in Table 5 for
each descriptor must be monitored.
5.1.4 GC/MS interface: Any GC-to-MS interface that gives an
acceptable calibration response for each analyte of
interest at the concentration required and achieves the
required tuning performance criteria (Section 7.2) may
be used. Inserting a fused silica column directly into
the MS source is recommended, however, care must be
taken not to expose the end of the column to the
electron beam. If other GC-to-MS interfaces are
utilized, they must be constructed of all glass or
glass-lined materials. The glass can be deactivated by
silanizing with dichlorodimethylsilane.
5.1.5 Data system: A computer system must be interfaced to
the mass spectrometer. The system must allow for the
continuous acquisition and storage on machine-readable
media of all data obtained throughout the duration of
the chromatographic program. The computer must have
software that can search any GC/MS data file for ions
of a specific mass and can plot such ion abundances
versus time or scan number. This type of plot is de-
fined as a Selected Ion Current Profile (SICP). Soft-
ware must also be able to integrate the abundance, in
any SICP, between specified time or scan number limits.
5.2 Pipets-Disposable
5.2.1 Pipets-Disposable, Pasteur, 150 mm long x 5 mm ID
(Fisher Scientific Company, No. 13-678-6A, or
equivalent).
NCASI Method TCDD/F - 88.01
435
-------
- 18 -
5.2.2 Pipets, disposable, serological 5 mL (6 nun i.d.) for
preparation of the carbon column specified in Section
10.2.3.
5.3 Conical Reacti-vial, 2 mL (Pierce Chemical Company), it is
recommended that these vials be silanized with
dichlorodimethlysilane.
5.4 Soxhlet Extractors, 40 mm i.d. with 250 mL boiling flask and
appropriate thimbles.
5.5 Chromatography columns, 15 mm i.d. x 20 cm length, with 250
mL reservoir, and 11 mm i.d. x 16 cm length with 50 mL
reservoir.
5.6 125 mL and 2 L Separatory Funnels (Fisher Scientific Company,
No. 10-437-5b, or equivalent).
5.7 Nitrogen blowdown apparatus (N-Evap Analytical Evaporator
Model 111, Organomation Associates, Inc., Northborough,
Massachusetts or equivalent).
5.8 Glass fiber filter paper (Whatman No. GF/D, or equivalent).
5.9 Top loading balance capable of weighing 200 g to the nearest
0.1 g.
5.10 Roto-evapbrator, (R-110. Buchi/Brinkman, American Scientific
No. E5045-10 or equivalent).
5.11 Pressure Filtration Apparatus, (Millipore No. YT30 142 HW or
equivalent), or alternatively, 1 L glass Microanalysis (90
mm) filter holder for vacuum filtration (Fisher No. 09-753-2
or equivalent) with 1 L vacuum flask.
5.12 Glass wool, extracted with methylene chloride and stored in
a clean glass jar.
5.13 Glass funnels, sized to filter 150 to 250 mL of solvent
through a sodium sulfate bed.
5.14 Glass vials, 20 to 40 mL capacity with Teflon lined screw
caps
6. Reagents
6.1 Sodium hydroxide-(ACS), 20 percent (w/v) in distilled water.
6.2 Sulfuric acid-(ACS), concentrated.
NCASI Method TCDD/F - 88.01
43G
-------
- 19 -
6.3 Methylene chloride, n-hexane, benzene, toluene, acetone,
isooctane, methanol, tetradecane, cyclohexane. Distilled in
glass or highest available purity. 95 percent ethanol should
be prepared by mixing the appropriate quantities of absolute
ethanol (100 percent, undenatured, highest purity available)
and organic free water.
6.4 Prepare stock standards in a glovebox or hood from
concentrates. The stock solutions (50 ppm) are stored at
room temperature in the dark, and checked frequently for
signs of degradation or evaporation, especially just prior
to the preparation of working standards.
6.5 Prepurified nitrogen gas.
6.6 Anhydrous sodium sulfate (reagent grade). Pre-extract with
methylene chloride and dry at 130° C.
6.7 Silica gel - Kieselgel 60, activate for >12 hours at 130°C
before use. Store at 130°C in covered flask.
6.8 Acid alumina - Bio-Rad Ag-4, activate for 12 hours at 130°C
before use. Store at 130°C in covered flask.
6.9 Basic alumina - Bio Rad Ag-10, kiln at 600° C for >24 hours
before use. Store at 130° C in covered flask. DO NOT USE
IF OLDER THAN FIVE DAYS!
6.10 Carbopack/silica gel - Mix 3.6 g Carbopack C, 60/80 mesh
(Supelco 1-0257) and 16.4 g activated silica gel. Activate
mix for >12 hours at 130°C before use. Store at 130°C in
covered flask.
6.11 AX-21 carbon/silica - AX-21 carbon is available through
Anderson Development Company 1415 E. Michigan St., Adrian,
MI 4922. Wash 100 g of AX-21 carbon powder (as received) by
suspending in 300 mL methanol and subsequently vacuum filter
through a glass fiber filter (Gelman A/E or equivalent)
fitted in a 350 mL Buchner funnel. Follow with two 100 mL
methanol rinses and draw the excess methanol from the filter
cake by continued vacuum. Dry the washed AX-21 carbon at
130°C for a minimum of 72 hours. Combine 5 g of prepared AX-
21 carbon with 95 g of prepared silica gel (6.7) in a wide
mouth bottle with a teflon lined screw cap. Blend by shaking
until a uniform color is achieved. Activate the blended AX-
21 carbon/silica at 130°C for a minimum of 24 hours and store
in a desiccator at room temperature.
NCASI Method TCDD/F - 88.01
437
-------
- 20 -
6.12 44% I^SCM/silica gel - Mix 24 mL cone. H2SC>4 and 56 g
activated silica gel. Stir and shake until free flowing.
Store in sealed containers at room temperature.
6.13 33% NaOH/silica gel - Mix 34 mL 1 N NaOH and 67 g activated
silica gel. Stir and shake until free flowing. Store in
sealed containers at room temperature.
6.14 Calibration Solutions.
6.14.1 Calibration Standard Solutions (CS1 to CSS, Table 4) -
Five solutions containing native TCDD and TCDF,"
labeled TCDD and TCDF and 37C14 TCDD at known
concentrations used to calibrate the instrument.
6.14.2 High level stock solutions (50±5 ug/mL) obtained from
NCASI will be used to prepare the appropriate dilutions
required to make up the calibration standards described
in Table 4. Suitable 13C^2 and 37Cl4 standards must be
obtained from commercial sources. It is the
responsibility of the laboratory to ascertain that the
calibration solutions prepared are indeed at the
appropriate concentrations before they are used to
analyze samples. Store the high level stock solutions
protected from light at room temperature.
6.14.3 Store the working concentration calibration solutions
in 1 mL minivials at 4°C protected from light.
6.15 GC Column Performance Check Solutions.
These solutions contain other TCDD isomers or TCDF isomers
for the purpose of documenting the chromatographic
resolution. The 13C12-2,3,7,8-TCDD or 13C12~2'3,7,8-TCDF is
also present. The laboratory Is required to use tetradecane
as the solvent and adjust the volume so that the final
concentration does not exceed 100 pg/uL. Table 3 summarizes
the qualitative composition (minimum requirement) of these
performance evaluation solutions. The difference in
concentrations of each isomer should be less than 35 percent
of the concentration of the 2,3,1,8-substituted isomer.
NCASI Method TCDD/F - 88.01
438
-------
-21-
6.16 Sample Fortification Solution (IS, Table 2).
This solution contains the two internal standards (
2,3,7,8-TCDD and 13C12-2,3,7,8-TCDF) at the nominal
concentration that is listed -in Table 2. Ten uL of solution
IS is diluted in acetone and then is spiked into each
wastewater sample aliquot prior to filtration and extraction.
Ten uL of solution IS is spiked into each solid type sample
aliquot prior to soxhlet extraction.
6.17 Cleanup Recovery Standard Spiking Solution (CRS, Table 2).
This solution contains the cleanup recovery standard (37Cl4~
2,3,7,8-TCDD) at the nominal concentration listed in Table
2. Ten uL of this solution will be spiked into each sample
extract immediately after the extraction step and before any
cleanup procedures are started.
6.18 Recovery Standard Spiking Solution (RS, Table 2).
This solution contains the recovery standard (^C^-l* 2,3,4-
TCDD) at the nominal concentration listed in Table 2. Ten
uL of this solution will be spiked into each sample extract
before HRGC/HRMS analysis.
7. System Performance Criteria
System performance criteria are presented below. The
laboratory will use the GC columns described in Section
5.1.2. It must be documented that all applicable system
performance criteria specified in Section 7.1 were met before
analysis for any sample is performed. Table 6 provides
recommended GC conditions that can be used to satisfy the
required criteria. During a typical 12-hour analysis
sequence, the GC column performance and mass spectrometer
resolving power checks must be performed at the beginning
of the 12-hour period of operation. A routine calibration
verification is required at the beginning and end of each
12-hour period during which samples are analyzed. A method
blank or HRGC/HRMS solvent blank run is required between a
calibration run and the first sample run.
7.1 GC Column Performance.
7.1.1 Inject the column performance check solution (Section
6.14) and acquire selected ion monitoring (SIM) data as
described in Section 5.1.3.
7.1.2 The chromatographic separation between 2,3,7,8-TCDD and
NCASI Method TCDD/F - 88.01
430
-------
- 22 -
the peaks representing any other TCDD isomers or between
2,3,7,8-TCDF and the peaks representing other TCDF
isomers must be resolved with a valley of <25 percent,
where
Valley Percent = (a/b) x 100
a = height of valley measured between 2,3,7,8-
TCDD or 2,3,7,8-TCDF and the closest TCDD or
TCDF eluting isomer, and
b = the peak height of 2,3,7,8-TCDD or 2,3,7,8-
TCDF.
It is the responsibility of the laboratory to verify
the conditions suitable for the appropriate resolution
of 2,3,7,8-TCDD from all other TCDD isomers and of
2,3,7,8-TCDF from all other TCDF isomers. Any
individual selected ion current profile (SICP) (for the
tetras, this would be the SICP for m/z 322 and m/z 304)
or the reconstructed homologue ion current (for the
tetras, this would correspond to m/z 320 + m/z 322 +
m/z 304 + m/z 306) constitutes an acceptable form of
data presentation. An SICP for m/z 334 (labeled TCDD)
and m/z 318 (labeled TCDF) is also required.
7.2 Mass Spectrometer Performance.
All mass spectrometer tuning must be performed with the GC
oven equilibrated at a temperature of 260° C. This will
provide carrier gas flow into the ion source at a rate
consistent with the flow during TCDD data acquisition.
7.2.1 The mass spectrometer must be operated in the electron
ionization mode. It is recommended that the ionization
potential be set to optimize sensitivity for the given
column flow and source design. A static resolving power
of at least 10,000 (10 percent valley definition) must
be demonstrated at appropriate masses before any
analyses are performed (Section 11). Static resolving
power checks must be performed at the beginning of each
12-hour period of operation. However, it is recommended
that a visual check (i.e., documentation is not
required) of the static resolution be made by using the
peak matching unit before and after each analysis.
Corrective actions must be implemented whenever the
resolving power does not meet the requirement.
NCASI Method TCDD/F - 88.01
440
-------
- 23 -
7.2.2 Chromatography time for TCDDs and TCDFs exceeds the
long-term mass stability of the mass spectrometer.
Because the instrument is operated in the high-
resolution mode, mass drifts of a few ppm can have
serious adverse effects on the instrument performances.
Therefore, a mass-drift correction is mandatory. To
that effect, use a lock-mass ion from the reference
compound (PFK) used for tuning the mass spectrometer
and monitor and record the lock-mass ion channel during
SIM acquisitions. The level of the reference compound
metered inside the ion chamber during HRGC/HRMS analyses
should be adjusted so that the amplitude of the selected
lock-mass ion signal does not exceed 10 percent of the
full-scale deflection for a given set of detector
parameters. Under those conditions, sensitivity changes
that might occur during the analysis can be more
effectively monitored.
7.2.3 By using PFK molecular leak and an appropriate ion
within the scan window, tune the instrument to meet the
minimum required resolving power of 10,000 (10 percent
valley).
7.2.4 Documentation of the instrument resolving power must be
accomplished by recording the peak profile of the high-
mass reference signal (m/z 380.9760). The format of
the peak profile must allow manual determination of the
resolution; i.e., the horizontal axis must be a
calibrated mass scale (amu or ppm per division). The
result of the peak width measurement (performed at 5
percent of the maximum, which corresponds to the 10
percent valley definition) must appear on the hard copy
and cannot exceed 100 ppm at m/z 380.9760 (or 0.038 amu
at that particular mass).
8. Calibration
8.1 Initial Calibration.
Initial calibration is required before any samples are
analyzed for 2,3,7,8-TCDD and 2,3,7,8-TCDF. Initial
calibration is also required if any routine calibration
(Section 8.3) does not meet the required criteria listed in
Section 8.4.
8.1.1 A minimum of five high-resolution concentration
calibration solutions covering a minimum concentration
range of three orders of magnitude must be used for the
initial calibration. The maximum difference between
NCASI Method TCDD/F - 88.01
441
-------
- 24 -
the concentrations of one standard and the next higher
concentration standard will be one order of magnitude.
The lowest concentration for 2,3,7,8-TCDD and 2,3,7,8-
TCDF must be 0.5 pg/uL or less. If a second instrument
is used for high concentration confirmations only, then
the low calibration standard can be at a higher
concentration. . However, it is the responsibility of
the laboratory to analyze and report only results that
meet the target detection limits and which are within
the calibrated range of the instrument used for the
determination.
8.1.2 Tune the instrument with PFK as described in Section 7.2.
8.1.3 Inject the GC column performance check solution,
(Section 6.14) and acquire SIM mass spectral data as
described earlier in Section 5.1.3. The laboratory
must not perform any further analysis until it has
demonstrated and documented that the column performance
criterion listed in Section 7*1.2 was met.
8.1.4 By using the same GC and mass spectrometer conditions
that produced acceptable results with the column
performance check solution, analyze a 2 uL aliquot of a
minimum of five concentration calibration solutions in
duplicate with the following mass spectrometer operating
parameters.
8.1.4.1 The total cycle time for data acquisition must be
<1 second. The total cycle time includes the sum
of all the dwell times and voltage reset times.
8.1.4.2 Acquire SIM data for the ions listed in the
descriptors in Table 5.
8.1.5 Integrate peak areas for all target analyte and standard
peaks. All the following criteria must be met for all
individual injections.
8.1.5.1 The retention time (at maximum peak height) of the
native analytes (i.e., the two ions used for
quantitation purposes) must be within -1 and +3
seconds of the retention time of the peak for the
isotopically labeled internal standard at m/z
corresponding to the first characteristic ion (of
the set of two; Table 5).
8.1.5.2 The ion current responses for both ions used for
quantitative purposed (e.g., for 2,3,7,8-TCDD: m/z
319.8465 and 321.8936) must reach maximum
NCASI Method TCDD/F - 88.01
4 A O
^i<~
-------
- 25. -
simultaneously (± 1 second).
8.1.5.3 The ion current responses for both ions used for
the labeled standards (e.g., for ^C-TCDDs: m/z
331.9368 and 333.9339) must reach maximum
simultaneously (± 1 second).
8.1.5.4 The ratio of integrated ion currents for the M+ to
the M+2+ ions belonging to the native analyte, the
carbon-labeled internal standards and the recovery
standard must be between 0.65 and 0.89.
8.1.5.5 For each SIM trace, all target analyte and labelled
standard peaks must be detected with a minimum
signal to noise ratio (S/M) £ 5.
8.1.6 Calculate the relative response factors (RRF) for
unlabeled target analytes relative to the appropriate
internal standards (Table 4) according to the following
formula:
Ax x QIS
RRF(n) =
Ox x AIS
where:
Ay = the sum of the integrated ion
abundances of the quantitation ions (Table 5) for
unlabeled TCDD/TCDF.
ATS = tne sum °f t^ie integrated ion abundances of
the quantitation ions (Table 5) for the labeled
internal standards.
QIS = tne quantity of the internal standard
injected (pg) .
QX = the quantity of the unlabeled TCDD/TCDF
analyte injected (pg).
RRF(n) = relative response factor of a particular
TCDD/TCDF isomer (n) relative to the internal
standard as determined from one injection.
The RRF(n) are dimensionless quantities; the units used
to express QIS and gx and must be the same.
NCASI Method TCDD/F - 88.01
443
-------
-26 -
8.1.7 Calculate the relative response factors (RRF) for the
labeled 13C12 internal standards relative to the
recovery standard according to the following formula:
* QRS
RRF(ra) = _
Qisra * ARS
where :
AJS™ = sum of the integrated ion abundances of the
quantitation ions (Table 5) for a given internal
standard (m = 1-2).
ARS ~ sum °^ tne integrated ion abundances of the
quantitation ions (Table 5) for the recovery
standard.
QRS and Qism = quantities of the recovery standard
(RS) and a particular internal standard (IS = m)
respectively, injected (pg) .
RRF(m) = relative response factor of a particular
internal standard (m) relative to the recovery
standard as determined from one injection.
The RRF(m) is a dimensionless quantity; the units used to
express Qjsm an<* QRS must ^e tne same.
8.1.8 Calculate the relative response factors (RRF) for the
labeled ^'CI^ cleanup recovery standard relative to the
recovery standard according to the following formula:
ACS° x
RRF(o) =
x RS
where :
= the integrated ion abundances of the
quantitation ion (Table 5) .
ARS = sum °f tne integrated ion abundances of the
quantitation ions (Table 5) for the recovery
standard.
Qps and QCS° = quantities of the recovery standard
(RS) and the clean-up recovery standard injected
respectively, (pg).
NCASI Method TCDD/F - 88.01
444
-------
- 27 -
RRF(o) = relative response factor the Cleanup
recovery standard relative to the recovery standard
as determined from one injection.
The RRF(o) is a dimensionless quantity; the units used
to express Q^s and Qcs° must be the same.
8.1.9 Calculate the RRF(n)s and their respective percent
relative standard deviations (%RSD) for the five
calibration solutions.
RRF(n) = 1/10 j[ RRFj(n),
where n represents the particular TCDD/TCDF isomer
(n=l-2), and j is the injection number (j=l-10)
and RRF(n) is the relative response factor as
defined in Section 8.1.6.
RRF(rO = calculated mean relative response factor
of a TCDD/TCDF isomer (n) relative to the internal
standard as determined from the 10 initial
calibration injections (j).
8.1.10 Mean relative response factors [RRF(m)] to be used for
the determination of the percent recoveries for the
internal standards are calculated as follows:
RRF(m) = 1/10
where m represents the particular internal standard
(m=l-2), and j is the injection number (j=l-10)
and RRF(m) is the relative response factor as
defined in Section 8.1.7.
RRF(m) = calculated mean relative response factor
of a particular internal standard (m) relative to
the recovery standard as determined from the 10
initial calibration injections (j).
8.1.11 Mean relative response factors [RRF(o)] to be used for
the determination of the percent recovery for the clean-
up recovery standard are calculated as follows:
NCASI Method TCDD/F - 88.01
445
-------
- 28 -
RRF(o) = 1/10 V RRFj (o),
where o represents the clean-up recovery standard,
and j is the injection number (j=l-10) and RRF(o)
is the relative response factor as defined in
Section 8.1.8.
RRF(o) = calculated mean relative response factor
for the clean-up recovery standard relative to the
recovery standard as determined from the 10 initial
calibration injections (j).
8.2 Criteria for Acceptable Calibration.
The criteria listed below for acceptable calibration must be
met before the analysis is performed.
8.2.1 The percent relative standard deviation for each of the
mean response factors [RRF(n), RRF(m) and RRF(o)] from
each of the 50 determinations (20 for the unlabeled
standards, 20 for the labeled reference compounds and 10
for the cleanup recovery standard) must be less than 20
percent.
8.2.2 The S/N for the GC signals present in every SIM trace
(including the ones for the labeled standards) must be > 5.
8.2.3 Isotopic ratios of all the M+ to M+2+ peaks must be within
± 15 percent of the theoretical value (i.e. 0.65 to 0.89).
NOTE: If the criterion for acceptable calibration
listed in Section 8.2.1-is met, the analyte-
specific RRF can then be considered independent of
the analyte quantity for the calibration
concentration range. The mean RRFs will be used
for all calculations until the routine calibration
criteria (Section 8.4) are no longer met. At such
time, new mean RRFs will be calculated from a new
set of injections of the calibration solutions.
8.3 Routine Calibration.
Routine calibrations must be performed at the beginning of a
12-hour period after successful mass resolution and GC
resolution performance checks and at the end of a 12-hour
period following the analysis of samples.
NCASI Method TCDD/F - 88.01
:G
-------
- 29 -
8.3.1 At the beginning of the 12-hour period, inject 2 uL of
the lowest or the next to lowest concentration calibra-
tion solution (Table 4). By using the same HRGC/HRMS
conditions as used in Sections 8.1.3 and 8.1.4,
determine and document an acceptable calibration as
provided in Section 8.4.
8.3.2 At the end of the 12-hour period, inject 2 uL of the
highest concentration calibration solution (Table 4).
By using the same HRGC/HRMS conditions as used in
Sections 8.1.3 and 8.1.4, determine and document an
acceptable calibration as provided in Section 8.4.
8.4 Criteria for Acceptable Routine Calibration.
The following criteria must be met before further analyses
are performed. If these criteria are not met, corrective
action must be taken and the instrument must be recalibrated.
8.4.1 The measured RRFs [RRF(n) for the unlabeled standards]
obtained during the routine calibration runs must be
within 20 percent of the mean values established during
the initial calibration (Section 8.1.4.6).
8.4.2 The measured RRFs [RRF(m) for the labeled standards RRF
(o) for the clean-up recovery standard] obtained during
the routine calibration runs must be within 20 percent
of the mean values established during the initial
calibration (Section 8.1.4.7).
8.4.3 Isotopic ratios of all the M+ to M+2+ peaks must be within
± 15 percent of the theoretical value (i.e. 0.65 to 0.89).
8.4.4 If any one of the criteria above (Sections 8.4.1 through
8.4.3) are not satisfied, a 'second attempt can be made. If
the second attempt fails to meet all the criteria, then the
entire initialization process (Section 8.1) must be
repeated.
8.4.5 If the second analysis of the standard at the end of
the 12-hour shift fails, then all samples run during
that shift must be reanalyzed after the initialization
process has been repeated.
9. Quality Control
9.1 Method Blank.
A method blank is performed by executing all of the specified
NCASI Method TCDD/F - 88.01
4 4 7
-------
- 30 -
extraction and cleanup steps, except for the introduction of
a sample. The method blank is also dosed with the internal
standards, the cleanup standard and the recovery standard at
the appropriate stages of the analysis. For water samples,
one liter of deionized or distilled water should be used as
the method blank.
9.1.1 Before processing any samples, the analyst must
demonstrate through the analysis of a method blank that
all glassware and reagents are interferant-free at the
method detection limit of the matrix of interest. Each
time a set of samples is extracted, or there is a change
in reagents, a method blank must be processed to
document the absence of laboratory contamination.
9.1.2 A laboratory "method blank" must be run along with each
set of 24 or fewer samples.
9.1.3 The Method Blank internal standard and cleanup recovery
standard recoveries must be greater than 40 percent.
If the recoveries fall below this minimum, the source
of the problem will be identified and corrected. The
laboratory will repeat the analysis of all samples in
that set which had either internal standard or cleanup
recoveries below the 40 percent minimum objective.
9.2 Method Blank Native Spike.
A method blank native spike is performed by executing all of
the specified extraction and cleanup steps, except for the
introduction of a sample. The method blank native spike is
dosed with the native analytes at the target detection
limits, the internal standards, the cleanup standard and the
recovery standard at the appropriate stages of the analysis.
For water samples, one liter of deionized or distilled water
should be used as the method blank native spike.
9.2.1 The laboratory will conduct a Method Blank Native Spike
at the target detection limits (Section 9.4) whenever
there is a change in the lot number of the acid alumina,
Carbopack C or basic alumina to demonstrate acceptable
recoveries of the native spikes.
9.2.2 All quality control criteria must be met before the
reagent will be judged acceptable for use in the
analysis of samples.
9.2.3 The Method Blank Native Spike internal standard and
cleanup recovery standard recoveries must be greater
than 40 percent. If the recoveries fall below this
NCASI Method TCDD/F - 88.01
448
-------
- 31 -
minimum, the source of the problem will be identified
and corrected. The laboratory will repeat the analysis
of all samples in that set which had either internal
standard or cleanup recoveries below the 40 percent
minimum objective.
9.3 GC column performance must be demonstrated initially and
verified prior to analyzing any sample in a 12-hour period.
The GC column performance check solution must be analyzed
under the same chromatographic and mass spectrometric
conditions used for other samples and standards.
9.4 Detection Limit Criteria.
The target detection limits will be 1 ppt for pulps and
sludges and 10 ppq for treated effluents.
9.4.1 If the laboratory has used a sample size smaller than
the maximum recommended sample size listed in Table 1
and fails to achieve the target detection limit, the
analysis will be repeated using an appropriately larger
sample size.
9.5 Isomer Specificity.
The laboratory will only report isomer specific data when
2,3,7,8-TCDD or 2,3,7,8-TCDF is detectable.
9.5.1 Isomer specificity will be demonstrated on a daily basis
according to the procedure outlined in Section 7.1.
9.5.2 If a non-isomer specific column is used to screen for
an analyte and the analyte is not detectable above the
target detection limit, the laboratory can report the
results of this analysis and re-analysis on the isomer
specific column will not be required. If the detection
limit exceeds the target detection limit, it will be
left to the discretion of the laboratory as to whether
the extract will be re-analyzed on the isomer specific
column or a newjportion of sample will be extracted and
analyzed in order to meet the target detection limit
criteria.
9.6 Native Spikes.
In order to provide an estimate of method accuracy, the
laboratory will spike samples with known concentration of
2,3,7,8-TCDD and 2,3,7,8-TCDF and then analyze using this
procedure to determine recovery.
NCASI Method TCDD/F - 88.01
449
-------
- 32 -
9.6.1 The native spike recovery is calculated as follows:
-------
- 33 -
9.7.1 The relative percent difference (RPD) is calculated as
follows:
RPD =
Si - S2
x 100
(Si + S2)/2
where:
S^ and S2 represent sample and sample duplicate
results.
9.7.2 The client will designate the samples and frequency for
duplicate or native spike duplicate determinations. It
is recommended that the minimum frequency of duplicate
determinations is one sample in ten.
9.7.3 The relative percent difference between sample duplicate
determinations must be < 50 percent.
9.7.4 The relative percent difference between matrix spike
duplicates shall be < 50 percent.
9.8 For reliable detection and quantitation of 2,3,7,8-TCDF, the
molecular ion of hexachlorodiphenyl ether must be monitored.
The fragment ions of the hexachlorodiphenyl ether if present,
will cause interferences for the ions monitored for 2,3,7,8-
TCDF. If a positive response on the hexachlorodiphenyl ether
ion is noted at the 2,3,7,8-TCDF retention time, then the
peak should be flagged and the report should indicate "sample
not analyzable due to HCDPE interference".
10. Extraction and Cleanup Procedures
10.1 Extraction procedures.
Prior to extraction, all sludge, pulp and ash samples should
be dried and blended according to the NCASI DIOXIN PROGRAM
SAMPLE PREPARATION/PROCESSING PROTOCOL (Revision 4 or later).
10.1.1 Wastewater Samples. If an aliquot of the sample is to
be analyzed, proceed as described in step 10.1.1.1. If
the entire contents of the sample bottle are to be used
for the analysis, proceed directly to step 10.1.1.3.
10.1.1.1 Sub-sampling for analysis of an aliquot of the
sample. Mark the liquid meniscus on the side of
the sample bottle. Shake the sample bottle
vigorously and measure out the desired volume of
NCASI Method TCDD/F - 88.01
451
-------
- 34 -
sample in a graduated cylinder. Transfer the sample
to suitable bottle. Transfer the remainder of the
sample to a new bottle and refrigerate. Rinse the
original sample bottle and the graduated cylinder
used to measure the sample aliquot with three 20
mL portions of methylene chloride. Combine "the
methylene chloride rinses in a suitably labeled
bottle. Refill the original sample bottle to the
mark and transfer the liquid to a graduated
cylinder. Record the original sample volume to
the nearest 5 mL.
10.1.1.2 Calculate the percentage of the total sample which
is to be used for the analysis. Use the same
percentage of the total sample volume for the
volume of the methylene chloride rinses which
should be included with the final sample extract
(Section 10.1.1.9). Proceed to Section 10.1.1.4.
10.1.1.3 If the entire contents of the bottle are to be
used for the analysis, mark the liquid meniscus on
the side of the sample bottle for later
determination of the sample volume (Section
10.1.1.7).
10.1.1.4 Internal Standard Addition. Add 10 uL of the
internal standard spiking solution of • C]^-
2,3,7,8-TCDD and 13C12-2,3,7,8-TCDF (IS, Table 2)
to a test tube containing 1 mL of acetone. Mix
and transfer the solution into the sample bottle.
Rinse the test tube with two one mL portions of
acetone, adding the rinses to the sample. Shake
or stir the sample for a minimum of 30 minutes.
10.1.1.5 Filtration. Shake the sample vigorously and
transfer the sample to the pressure filtration
apparatus. Add 50 mL deionized water to the sample
bottle, cap, shake then filter to assure
quantitative transfer of solids. Repeat the
deionized water wash a second time. Pressure
filter through Whatman GF/D (or equivalent) glass
fiber filter paper collecting the filtrate in the
original sample bottle.
10.1.1.5.2 Filtration Alternative. If using the vacuum
filtration apparatus with the 90mm filter holder,
perform the filtration as described in 10.1.1.5
collecting the filtrate and rinsate in a 1L vacuum
flask. Subsequent methylene chloride rinses in
section 10.1.1.6 should include the vacuum flask
NCASI Method TCDD/F - 88.01
452
-------
- 35 -
in addition to the sample bottle.
10.1.1.6 Liquid-liquid Extraction. Transfer the filtered
wastewater into a 2 L separatory funnel. Add
sufficient deionized water to the filtered
wastewater to bring the total sample volume to
approximately one L. After removing the glass
fiber filter, rinse the pressure filtration
apparatus with 60 mL methylene chloride, collecting
the solvent rinses in the sample bottle. Seal the
bottle and shake 30 seconds to rinse the inner
surface. Transfer the solvent to the separatory
funnel and extract the sample by shaking the funnel
for two minutes with periodic venting. Allow the
organic layer to separate from the water phase for
a minimum of 10 minutes. If the emulsion interface
between layers is more than one-third the volume
of the organic layer, the analyst must employ
mechanical techniques to complete phase separation.
Drain the extract and repeat this extraction two
additional times with fresh 60 mL portions of
methylene chloride. Combine the methylene chloride
extracts.
Note: It is recommended that the analyst
measure the volume of the solvent recovered.
If the solvent recovery is less than 85
percent, the extraction should be repeated a
fourth time combining the fourth extract with
the original three.
Extract the sample one final time using 60 mL
toluene. Combine the toluene extract with the
methylene chloride extracts. Dry the organic layer
by pouring through a funnel containing anhydrous
sodium sulfate into a 'round bottom flask, wash
with two 15 mL portions of hexane, and concentrate
the extract solution to a volume of approximately
15 mLs with a rotary evaporator (heated water baths
required).
10.1.1.7 If the entire sample was used for the analysis,
determine the original sample volume by refilling
the sample bottle to the mark and transferring the
liquid to a graduated cylinder. Record the sample
volume to the nearest 5 mL.
10.1.1.8 Soxhlet extraction of filtered solids. Assemble
and pre-extract a soxhlet apparatus for three hours
with 68:32 ethanol/toluene. Place the filtered
NCASI Method TCDD/F - 88.01
453
-------
19,
- 36 -
solids in the pre-extracted soxhlet thimble and
break up the filter as much as possible with a
spatula. Place the thimble in a soxhlet charged
with 96 mL toluene and add 204 mL ethanol to the
soxhlet by passing it through the thimble. Extract
the filter for a minimum of 16 hours.
10.1.1.9 Combine the methylene chloride extract concentrate
from 10.1.5 with the ethanol/toluene extract. If
the sample was subdivided as described in Section
10.1.1.1, measure out a volume of the methylene
chloride sample bottle rinsate which corresponds
to the proportion of the original sample which was
used for the analysis and add it to the combined
ethanol/toluene-methylene chloride extract. Add
ul* of the cleanup recovery spiking standard of
rCl4~2,3,7,8-TCDD and ca 100 \LL of tetradecane
(as a keeper) to the extract. Concentrate to the
tetradecane to remove the acetone and methylene
chloride. Using 5 mL of ethanol, redissolve the
residue and pipet into 100 mL of hexane. Proceed
to Section 10.2.
10.1.2 Sludge Samples
10.1.2.1 Assemble and pre-extract a soxhlet apparatus,
including the thimble, for three hours with 68:32
ethanol/toluene. Discard the solvent and allow
the thimble to dry.
10.1.2.2 Weigh a representative (2g to lOg) portion of the
sample into the pre-extracted soxhlet thimble.
10.1.2.3 Internal Standard addition. Add 10 ui of the
internal standard spiking solution of C^-
2,3,7,8-TCDD and 13C12-2,3,7,8-TCDF (IS, Table 2).
10.1.2.4 Soxhlet extraction. Place the soxhlet thimble
containing the sample into the pre-extracted
Soxhlet extraction apparatus charged with 68:32
ethanol/toluene and extract for a minimum of 16
hours. Add 10 uL of the cleanup recovery standard
spiking solution (37Cl4-2,3,7,8-TCDD) and ca 100
UL of tetradecane (as a keeper) to the extract.
Concentrate to the tetradecane. Using 5 mL of
ethanol, redissolve the residue and pipet into 100
mL of hexane. Proceed to Section 10.2.
NCASI Method TCDD/F - 88.01
454
-------
- 37 -
10.1.3 Pulps
Partially bleached, fully bleached and most paper products
pulp samples are extracted as follows:
10.1.3.1 Assemble and pre-extract a soxhlet apparatus,
including the thimble, for three hours with 95
percent ethanol. Discard the solvent and allow
the thimble to dry.
10.1.3.2 Weigh a representative 10 g portion of the sample
into the pre-extracted soxhlet thimble.
10.1.3.3 Internal Standard addition. Add 10 uL of the
internal standard spiking solution of 13C^2-
2,3,7,8-TCDD and 13C12-2,3,7,8-TCDF (IS, Table 2).
10.1.3.4 Soxhlet extraction. Place the soxhlet thimble
containing the sample in the pre-extracted Soxhlet
extraction apparatus charged with 95 percent
ethanol and extract for a minimum of 16 hours.
Add 10 uL of the cleanup recovery spiking solution
of 37Cl4-2,3,7/8-TCDD and ca 100 uL of tetradecane
(as a keeper) to the extract. Concentrate to the
tetradecane. Using 5 mL of ethanol, redissolve the
residue and pipet into 100 mL of hexane. Proceed
to Section 10.2.
10.1.4 Ash Samples
10.1.4.1 Assemble and pre-extract a soxhlet apparatus,
including the thimble, for 3 hours with benzene or
toluene. Discard the solvent and allow the thimble
to dry.
10.1.4.2 Weigh a representative (5g to lOg) portion of the
sample into the pre-extracted soxhlet thimble. Add
an equivalent weight of pre-extracted anhydrous sodium
sulfate and mix into the sample.
10.1.4.3 Internal Standard addition. Add 10 uL of the internal
ftandard spiking solution of 13C^2-2,3,7,8-TCDD and
3C12-2,3,7,8-TCDF (IS, Table 2).
10.1.4.4 Soxhlet extraction. Place the soxhlet thimble contain-
ing the sample in the pre-extracted Soxhlet extraction
apparatus charged with benzene or toluene and extract
for a minimum of 16 hours. Add 10 uL of the cleanup
recovery spiking solution of 3'Cl4-2,3,7,8-TCDD to the
NCASI Method TCDD/F - 88.01
455
-------
-38 -
extract and proceed to the Section 10.2.
10.2 Extract Cleanup Procedures.
10.2.1 Sulfuric Acid/Sodium Hydroxide Partitioning.
10.2.1.1 Partition the solvent against 50 mL of concentrated
sulfuric acid. Shake for two minutes. Remove and
discard the aqueous layer (bottom). Repeat the
acid washing until no color is visible in the
bottom layer (perform acid washings a maximum of
four times).
10.2.1.2 Partition the extract against 50 mL of distilled
water. Shake for two minutes. Remove and discard
the aqueous layer (bottom).
10.2.1.3 Partition the solvent against 50 mL of 10 N NaOH.
Shake for two minutes. Remove and discard the
bottom aqueous layer (perform the base washings a
maximum of four times).
10.2.1.4 Partition the solvent against 50 mL of distilled
water. Shake for two minutes. Remove and discard
aqueous layer (bottom).
10.2.1.5 Dry the organic layer by pouring through a funnel
containing anhydrous sodium sulfate into a round
bottom flask, wash with two 15 mL portions of
hexane, add ca 100 uL of tetradecane and
concentrate the extract solution to the tetradecane
with a rotary evaporator (heated water bath),
making sure all traces of benzene (or toluene) are
removed. (Use of blowdown with an inert gas to
concentrate the extract is also permitted).
10.2.2 NaOH-silica:silica:H2SO4rsilica/Acid alumina columns
10.2.2.1 Place a glass wool plug in a 15 mm i.d.
chromotography column followed by 1 g silica gel,
4 g 33% 1 N sodium hydroxide/silica gel, 1 g silica
gel, 8 g 44% sulfuric acid/silica gel, 2 g silica
gel, and top with 1 cm sodium sulfate. Place a
glass wool plug in an 11 mm i.d. chromatography
column followed by 6 g acid alumina and top with 1
cm sodium sulfate. Pre-rinse both columns with
hexane and place the 15 mm i.d. column above the
11 mm i.d. column so that the eluant from the top
column goes directly onto the bottom column.
NCASI Method TCDD/F - 88.01
456
-------
- 39 -
10.2.2.2 Dissolve the sample residue in 2 mL of hexane and
apply the hexane solution to the top column. Elute
with 120 mL hexane. (Discard this eluate).
10.2.2.3 Remove the top column and elute the bottom column
with 20 mL hexane. (Discard this eluate).
10.2.2.4 Elute with 20 mL 20% methylene chloride/hexane
(collect this eluate). Add ca 100 uL tetradecane
to the eluate and blow down to the tetradecane
with purified nitrogen.
10.2.3 Charcoal/silica gel column.
The following cleanup procedure can be used for bleached
pulps but is not recommended for sample matrices which
are more heavily contaminated.
10.2.3.1 Cut 1 cm off the end tip of a 5 mL (7.5 mm i.d.)
disposable pipet. Place a glass wool plug at the
2.5 mL mark, add 0.65 g charcoal/silica gel packing
followed by another glass wool plug.
10.2.3.2 Pre-elute the column with 5 mL toluene and invert
the column, continue the pre-elution in the
opposite direction with 2 mL toluene, 2 mL 75:20
methylene chloride/methanol, 1 mL 1:1 methylene
chloride/cyclohexane, 5 mL hexane. (Discard
eluates).
10.2.3.3 Dissolve the sample residue in 2 mL hexane and
transfer to the top of the charcoal/silica column.
Rinse the sample vial twice with 2 mL of hexane
and transfer each rinse to the column. Elute the
column with 2 mL 1:1 methylene
chloride/cyclohexane, 2 mL 75:20 methylene
chloride/methanol. (Discard eluates).
10.2.3.4 Turn the column over and elute the column with 15
mL of toluene. (Collect the eluate).
10.2.3.5 Add 10 uL of tetradecane and concentrate the sample
with a stream of purified nitrogen or by vacuum
evaporation. Transfer the sample quantitatively
into a 2 mL cone shaped vial and evaporate to the
tetradecane with a stream of purified nitrogen. If
the optional additional cleanup procedure described
in section 10.3 is used, proceed directly to that
clean-up. If the extract is ready for analysis,
NCASI Method TCDD/F - 88.01
457
-------
- 40 -
add 10 uL of the recovery standard (^C^-l>2,3,4-
TCDD) to the extract and mix thoroughly. Store the
extracts protected from light.
10.2.4 Option to Charcoal/silica gel column 10.2.3.
The following cleanup procedure is recommended for all
wastewater, sludge and process pulp extracts. It
provides a higher capacity column for heavily
contaminated sample extracts.
10.2.4.1 Cut 2 cm off the tip end of a 10 mL (8 mm i.d.)
disposable sealogical pipet. Place a glass wool
plug at the 1.0 mL mark, add 1.0 g of the 5 percent
AX-21 carbon/silica gel packing followed by another
glass wool plug.
10.2.4.2 Pre-elute the column with 5 mL 1:1 methylene
chloride/cyclohexane and invert the column.
Continue the pre-elution in the opposite direction
with another 5 mL 1:1 methylene
chloride/cyclohexane. (Discard eluates).
10.2.4.3 Dissolve the sample residue in 1 mL hexane and
transfer to the top of the AX-21 carbon/silica
column. Rinse the sample vial twice with 2 mL of
1:1 methylene chloride/cyclohexane and transfer
each rinse to the column. Elute the column with an
additional 6 mL 1:1 methylene chloride/cyclohexane
and 5 mL 75:20:5 methylene chloride/methanol/
benzene. (Discard eluates).
10.2.4.4 Turn the column over and elute the column with 25
mL of toluene. (Collect the eluate).
10.2.4.5 Add 10 uL of tetradecane and concentrate the sample
with a stream of purified nitrogen or by vacuum
evaporation. Transfer the sample quantitatively
into a 2 mL cone shaped vial and evaporate to the
tetradecane with a stream of purified nitrogen. If
the optional additional cleanup procedure described
in section 10.3 is used, proceed directly to that
clean-up. If the extract is ready for analysis,
add 10 uL of the recovery standard (13C12-1,2,3,4-
TCDD) to the extract and mix thoroughly. Store the
extracts protected from light.
NCASI Method TCDD/F - 88.01
458
-------
- 41 -
10.3 Optional Additional Clean-up.
The following additional cleanup procedure can be used for
samples which have chemical interferences that cause
unacceptable detection limits.
10.3.1 Place a glass wool plug in an 11 mm c.d. chromatography
column followed by 5 g activated basic alumina and top
with 1 cm of sodium sulfate.
10.3.2 Dissolve the residue in 2 mL hexane and apply to the
top of the column.
10.3.3 Elute with 20 mL hexane (discard eluate). Elute with 8 mL
3% methylene chloride/hexane (save eluate). Elute with 35
mL 50% methylene chloride/hexane. Collect the eluate.
10.3.4 Transfer the extract quantitatively into a 2 mL cone
shaped vial. Add 10 uL tetradecane and concentrate to
the tetradecane with a stream of purified nitrogen.
Add 10 uL of the recovery standard (13C12-1,2,3,4-TCDD)
to the extract and mix thoroughly. Store extracts
protected from light.
11. Analytical Procedures
11.1 Inject a 2 uL aliquot of the.extract into the GC, operated
under the conditions previously used (Section 7.1) to produce
acceptable results with the performance check solution.
11.2 Acquire SIM data according to Section 5.1.3. Use the same
acquisition and mass spectrometer operating conditions
previously used to determine the relative response factors
(Sections 8.1.4 through 8.4.4).
NOTE: A selected ion current profile (SICP) for the
lock-mass ion will also be recorded. It is recommended
that the lock-mass ion SICP be examined for departure
of the instrument's basic sensitivity and stability
that could affect the measurements.
11.3 Identification Criteria.
For a gas chromatographic peak to be identified as a 2,3,7,8-
TCDD or 2,3,7,8-TCDF, it must meet all of the following
criteria:
NCASI Method TCDD/F - 88.01
4 5 3
-------
- 42 -
11.3.1 Relative Retention Times.
11.3.1.1 The retention time (at maximum peak height) of the
sample components (i.e., the two ions used for
quantitation purposes) must be within -1 and +3
seconds of the retention time of the peak for the
isotopically labeled internal standard at m/z
corresponding to the first characteristic ion (of
the set of two; Table 5 ) to obtain a positive
identification .
11.3.1.2 The ion current responses for both ions used for
quantitative purposed (e.g., for 2,3,7,8-TCDD: m/z
319.8465 and 321.8936) must reach maximum
simultaneously (± 1 second).
11.3.1.3 The ion current responses for both ions used for
the labeled standards (e.g., for 13C-TCDDs: m/z
331.9368 and 333.9339) must reach maximum
simultaneously (± 1 second).
11.3.2 Ion Abundance Ratios
11.3.2.1 The integrated ion current for the two ions of the
native analytes used for quantitation purposes
must have a ratio (M+/M+2*) within ±15 percent of
the theoretical value (i.e. 0.65 to 0.89).
11.3.2.2 The integrated ion current for the two ions of the
13- internal standards (including ^ -
2,3,7,8-TCDD, C12~1'2'3'4~TCDD and C12-
2,3,7,8-TCDF) must have a ratio (M+/M+2+) within ±
15 percent of the theoretical value (i.e. 0.65 to
0.89) .
11.3.3 Signal-to-noise Ratio. All ion current intensities
must be > 2.5 times noise level for positive identific-
ation of 2,3,7,8-TCDD or 2, 3 ,7 , 8-TCDF.
11.3.4 Internal Standard Recoveries. Internal standard
recoveries will be reported based upon the data obtained
for the analysis used for the isomer specific
determination of the analyte. The internal standard
recovery objective is 40 to 120 percent.
11.3.4.1 If the internal standard recoveries are between 40
and 120 percent, no further criteria or notations
will be required.
NCASI Method TCDD/F - 88.01
460
-------
- 43 -
11.3.4.2 If the internal standard recoveries are less than
40 percent but greater than 20 percent, the ion
current intensities for the internal standard ions
must be >10 times the noise level and the S/N
calculation must appear on the SIM trace above the
GC peak in question. If the internal standard S/N
is not greater than 10:1, corrective steps will be
taken and the sample will be rerun. All results
with internal standard recoveries between 20 and
40 percent will be reported with the internal
standard recovery given in brackets next to the
concentration detected or the detection limit (i.e.
[30%]).
11.3.4.3 If the internal standard recoveries are less than
20 percent, the source of the low recovery will be
determined, corrective steps will be taken and the
sample will be rerun. If the recoveries are still
unsatisfactory, the results will be reported as
"Detected - not quantifiable" where the
identification criteria for the native compound
were met, and as "analytical difficulties" where
the identification criteria for the native compound
were not.
12. Calculations
12.1 Quantitation of 2,3,7,8-TCDD and 2,3,7,8-TCDF detected in
samples.
For gas chromatographic peaks that have met the criteria
outlined in Section 11.3, calculate the concentration of the
2,3,7,8-TCDD or 2,3,7,8-TCDF using the formula:
Ax x QlS
AIS x W x RRF(n)
where:
Cx = concentration of unlabeled TCDD/TCDF in pg/g,
Ax = sum of the integrated ion abundances of the quantitation
ions (Table 5) for the native TCDD/TCDF
Ajg = sum of the integrated ion abundances of the quantita-
tion ions (Table 5) for the labeled internal standards,
QlS = quantity, in pg (i.e. 2000 pg), of the internal
standard added to the sample before extraction,
NCASI Method TCDD/F - 88.01
461
-------
- 44 -
W = weight, in g, of sample (solid or liquid), and
RRF(n) = calculated mean relative response factor from the
initial calibration for the analyte [RRF(n) with n = 1-2:
Section 8.1.9] .
12.2 Calculate the percent recovery of the internal standards
standard measured in the sample extract, using the formula:
AIS x
Internal standard
percent recovery = - x 100
x ARS
where:
ATS = sum of the integrated ion abundances of the quantita-
tion ions ( Table 5 ) for the labeled internal standard,
Apg = sum of the integrated ion abundances of the quantita-
tion ions (Table 5) for the labeled recovery standard;
QlS = quantity, in pg (i.e. 2000 pg), of the internal
standard added to the sample before extraction, and
QRS - quantity, in pg (i.e. 2000 pg) , of the recovery
standard added to the cleaned-up sample residue before
HRGC/HRMS analysis, and
RRF(m) = calculated mean relative response factor for the
labeled internal standard relative to the recovery standard.
This represents the mean obtained in Section 8.1.10.
12.3 Calculate the percent recovery of the cleanup recovery
standard measured in the sample extract, using the formula:
Cleanup Recovery A^s x QRS
Standard = - x 100
percent recovery Qcs x ARS x RRF(o)
where :
AQS = sum of the integrated ion abundances of the quant ita-
tion ions (Table 5) for the cleanup recovery standard,
ApS = sum °f tne integrated ion abundances of the quantita-
tion ions { Table 5 ) for the labeled recovery standard;
NCASI Method TCDD/F - 88.01
4B2
-------
- 45 -
= quantity» in P9 (i.e. 1000 pg) , of the cleanup recovery
standard added to the sample before clean up, and
= quantity, in pg (i.e. 2000 pg) , of the recovery
standard added to the cleaned-up sample residue before
HRGC/HRMS analysis, and
RRF(o) = calculated mean relative response factor for the
cleanup recovery standard relative to the recovery standard.
This represents the mean obtained in Section 8.1.11.
12.4 If the concentration in the 20 uL final extract of either
2,3,7,8-TCDD or 2,3,7,8-TCDF exceeds the upper method
calibration limits (MCL) listed in Table I, the linear range
of response versus concentration may have been exceeded.
The extract should be diluted to an appropriate volume and
reanalyzed. The report should include a notation of the
dilution used for the analysis.
12.5 Sample Specific Estimated Detection Limit.
The sample specific estimated detection limit (EDL) is the
concentration of a given analyte required to produce a signal
with a peak height of at least 2.5 times the background
signal level.
12.5.1 Samples presenting a response for either ion for native
2,3,7,8-TCDD or 2,3,7,8-TCDF that is less than 2.5 times
the background level.
12.5.1.1 By using the expressions of EDL below, calculate
an EDL for each isomer characterized by the absence
of a response (i.e., S/N < 2.5). The background
level is determined by measuring the range of the
noise (peak-to-peak) for the M+2"1" ion (Table 5) of
a particular analyte, in the region of the SICP
trace corresponding to the elution of the internal
standard, multiplying that noise height by 2.5,
. and relating the product height to an estimated
concentration that would produce that product
height.
Calculate the EDL using the following formula:
2.5 x HtxM+2 x QIS
EDL =
HtISM+2 x W x RRF(n)
NCASI Method TCDD/F - 88.01
468
-------
- 46 -
where:
EDL = estimated detection limit for 2,3,7,8-TCDD
or 2,3,7,8-TCDF.
HtxM+2 = noise level height of the M+2+ ion (Table
5.) for the native TCDD/TCDF
HtlM+2 _ noiSe level height of the M+2+ ion (Table
5_) for the labeled internal standard
and W, RRF(n), and Qjs retain the same meanings as
defined in Section 12.1.
12.5.1.2 When only one of the guantitation ion signals is
below 2=5 tiroes the background level and the other
ion signal is above, consider the positive ion
channel as an interfering peak. Calculate the
estimated detection limit based on the single
undetected ion using the corresponding M+ or M+2+
internal standard ion as follows:
2.5 x Htx x QIS
EDL =
Htls x W x RRF(n)
where:
EDL = estimated detection limit for 2,3,7,8-TCDD
or 2,3,7,8-TCDF.
Htx = the noise level height of the undetected
quantitation ion M+ or M+2+ (Table 5) for the
native TCDD/TCDF
Htjs = the noise level "height of the corresponding
M+ or M+2+ ion (Table 5) for the labeled internal
standard
and W, RRF(n), and QrS retain the same meanings as
defined in Section 12.1.
12.5.1.3 When the response of a signal having the same
retention time as the analyte has a S/N in excess
of 2.5 and does not meet any of the other
qualitative identification criteria, calculate the
estimated detection limit according to the equation
given in Section 12.1 except using peak heights.
NCASI Method TCDD/F - 88.01
464
-------
- 47 -
13. Deliverables
13.1 Initial Calibration Data. The laboratories standard
operating procedure for the preparation of standards must be
submitted prior to or simultaneously with the first
submission of initial calibration data. The SOP need.not be
submitted with subsequent initial calibration data sets
unless the procedure has been revised. The information
pertaining to the initial calibration and described in the
remainder of this Section must be submitted prior to or
simultaneously with the results of any sample analyses. If
the initiation process must be repeated, the information
pertaining to the repeat analyses must be submitted prior to
or simultaneously with the results of any sample analyses.
13.1.1 Documentation of the mass spectrometer resolution as
described in Section 7.2.4.
13.1.2 Documentation of the isomer specificity as described in
Section 7.1.2.
13.1.3 All SICP and area tables for all standard analyses
presented in the format described in Section 13.3.
13.1.4 Tabular summaries containing the following information:
13.1.4.1 The source, identification code and concentration
of all standards.
13.1.4.2 The ion ratios for native analytes and labeled
standards for all calibration standard analyses.
13.1.4.3 The native response factors relative to the
appropriate internal standard (RRF(n)) for each
calibration standard analysis.
13.1.4.4 The relative response factors for each internal
standard relative to the recovery standard (RRF(m))
for each calibration standard analysis.
13.1.4.5 The relative response factor for the cleanup
recovery standard relative to the recovery standard
(RRF(o)) for each calibration standard.
13.1.4.6 The mean relative response factors for all
determinations (RRF(n), RRF(ra), RRF(o)).
13.1.4.7 The relative standard deviation (RSD) of each
relative response factor (including RRF(n), RRF(m),
RRF(o)) for all calibration analyses.
NCASI Method TCDD/F - 88.01
4G5
-------
- 48-
13.1.4.8 The ±20 percent control limits to be used as
criteria to determine the acceptability of the
daily calibration checks.
13.2 Sample Data Packs. Each submission of sample analysis
results must include all information described in this
Section.
13.2.1 A cover letter or case narrative describing the
analytical procedure, identifying quality assurance
samples, analyses conducted, appropriate references to
Purchase Order Number, etc. should be submitted.
13.2.2 Pertinent copies of chain-of-custody records should be
included.
13.2.3 Documentation of isomer specificity for each column and
for each day analyses were performed as described in
Section 7.1.2.
13.2.4 Documentation of mass spectrometer resolution for each
instrument and day analyses were performed as described
in Section 7.1.4.
13.2.5 Daily calibration data including the dates and times
analyzed (alternatively, a copy of the appropriate page
of an analysis log book which shows the sequence and
times of injections can be submitted), instrument used,
the identification code for the standards used cross-.
referencing the initial calibration data set, all SICP
formatted as described in Section 13.3, all native and
labeled standard ion ratios, relative response factors
(RRF(n), RRF(n) and RRF(o)), and the percent relative
difference between the daily- calibration check relative
response factors and the initial mean relative response
factor (RRF(n), RRF(m) and RRF(o)).
13.2.6 Method Blank data including the dates prepared, sample
size or volume, all SICP formatted as described in
Section 13.3, tabular summaries of the estimated
detection limits, internal standard ion ratios, internal
standard recoveries and cleanup recovery standard
recoveries and, where analyte(s) are detected, the
analyte ion ratios and concentrations) detected.
13.2.7 Sample results including client sample identification
number, contract laboratory identification number,
sample weight, date sample preparation started, all
SICP formatted as described in Section 13.3, tabular
NCASI Method TCDD/F - 88.01
4GG
-------
-49 -
summaries of internal standard and analyte ion ratios,
internal standard and cleanup recovery standard
recoveries and concentrations detected or estimated
detection limits where applicable.
13.2.8 Laboratory duplicate results including client sample
identification number, contract laboratory
identification number, sample weight, date sample
preparation started, all SICP formatted as described in
Section 13.3, tabular summaries of internal standard
and analyte ion ratios, internal standard and cleanup
recovery standard recoveries and concentrations detected
or estimated detection limits where applicable and the
calculated relative percent difference.
13.2.9 Native Spike recovery results including client sample
identification number, contract laboratory
identification number, sample weight, date sample
preparation started, exact spike level expressed as a
concentration in the sample, all SICP formatted as
described in Section 13.3 tabular summaries of internal
standard and analyte ion ratios, internal standard and
cleanup recovery standard recoveries, concentrations
detected or estimated detection limits where applicable
and the native spike recovery.
13.3 Raw Data/SICP Reporting Format.
13.3.1 The report for isomer specific TCDD results should
include separate pages plotting the following SICPs:
13.3.1.1 M+ and M+2+ for native TCDD plus M+2+ for the
internal standard.
13.3.1.2 M+ and M+2+ for the TCDD internal standard plus M+
for 37Cl4-2,3,7,8-TCDP.
13.3.1.3 The lock mass check channel.
13.3.1.4 M+ for native TCDD
If the DB-5 column is used to screen for 2,3,7,8-TCDF
and it is not detectable at or below the target
detection limit, then the following additional SICPs
should be included in the report:
13.3.1.5 M+ and M+2+ for native TCDF plus M+2+ for the
internal standard.
13.3.1.6 M+ and M+2+ for the TCDF internal standard.
NCASI Method TCDD/F - 88.01
-------
- 50-
13.3.1.7 M+ for native TCDF
13.3.2 The report for TCDF results should include separate
pages plotting the following SICPs:
13.3.2.1 M+ and M+2+ for native TCDF plus M+2+ for the
internal standard.
13.3.2.2 M+ and M+2+ for the 13C12-1,2,3,4-TCDD recovery
standard.
13.3.2.3 M+2+ for native TCDF, M+ for hexachlorodiphenyl
ether and the lock mass check channel.
13.3.2.4 M+ for native TCDF
If the DB-225 column is used to screen for 2,3,7,8-TCDD
and it is not detectable at or below the target
detection limit, then the following additional SICPs
should be included in the report:
13.3.2.5 M+ and M+2+ for native TCDD plus M+2+ for the
internal standard.
13.3.2.6 M+ and M+2+ for the TCDD internal standard plus M+
for 37Cl4-2,3,7,8-TCDD.
13.3.2.7 M+ for native TCDD
13.3.3 Each SICP plot should include in the header the client
sample identification number, the contract laboratory
identification number, the date, time and the instrument
and column type (including internal diameter) used for
the analysis. If the date and time cannot be included
in the header, the report should include a copy of the
appropriate page of an analysis log book which shows
the sequence and times of injections.
13.3.4 Noise levels used in the estimation of detection limits
should be marked on all appropriate SICPs.
13.3.5 Noise levels and signal to noise level calculations
should be recorded on all internal standard and cleanup
recovery standard SICPs when the internal standard or
cleanup recovery standard recoveries are below the
target 40 percent.
13.3.6 Noise levels and signal to noise level calculations
should be recorded on all native SICPs when the signal
to noise is S 5 to 1.
NCASI Method TCDD/F - 88.01
4G8
-------
TABLE 1
c:
2.3,7,8-TCDD
Lower MCLa
Upper MCL
2,3,7,8-TCDF
Lower MCLa
Upper MCL
Recommended
Sample Size
IS Spiking Levels (ng)
13C12-2,3,7/8-TCDD
13C12-2,3,7,8-TCDF
TYPES OF MATRICES, SAMPLE SIZES AND 2,3,7,8-TCDD
AND 2,3,7,8-TCDF BASED METHOD CALIBRATION LIMITS
Wastewaters
Treated
Effluent
0.01 to 0.04
10 to 40
Process
Sewers
Sludges
Pulps
1
Ash
0.01 to 0.1 1 to 5 1 1 to 2
10 to 100 1000 to 5000 1000 1000 to 2000
0.01 to 0.04 0.01 to 0.1 1 to 5 1 1 to 2
10 to 40 10 to 100 1000 to 5000 1000 1000 to 2000
1.0 to 0.25L 1.0 to 0.10L 10 to 2g
2.0 ng
2.0 ng
Cleanup Recovery Spiking Level
37Cl4-2,3,7,8-TCDD 1.0 ng
Recovery Standard Spiking Level
13C12-1,2,3,4-TCDD 2.0 ng
Final Extract Vol. (y.L)
20
2.0 ng
2.0 ng
1.0 ng
2.0 ng
2.0 ng
1.0 ng
lOg lOg to 5g
i
M
2.0 ng 2.0 ng (
2.0 ng 2.0 ng
1.0 ng 1.0 ng
2.0 ng 2.0 ng 2.0 ng 2.0 ng
20 20 20 20
-------
- 52 -
TABLE 2 COMPOSITION OF THE SAMPLE FORTIFICATION SOLUTIONS
Analvte
13C12-2,3,7,8-TCDD
Sample Fortification Solutions (IS)
Concentration (pg/uL; Toluene)
200
200
(RS)
Analyte
13C12-1,2,3,4-TCDD
Recovery Standard Spiking Solution
Concentration
(pg/uL; Tetradecane)
200
Analyte
37Cl4-2,3,7,8-TCDD
Cleanup Recovery Spiking Solution (CRS)
Concentration
(pg/uL; Toluene)
100
NCASI Method TCDD/F - 88.01
470
-------
53 -
TABLE 3 GC COLUMN PERFORMANCE CHECK STANDARDS
2,3,7,8-TCDD COLUMN PERFORMANCE
CHECK STANDARD
(for use with 60 m DB-5 columns)
1,2,3,7-TCDD
1,2,3,8-TCDD
1,2,3,9-TCDD
2,3,7,8-TCDD
2,3,7,8-TCDF COLUMN PERFORMANCE
CHECK STANDARD
(for use with 30 m DB-225 columns)
2,3,4,7-TCDF
1,2,3,9-TCDF
2,3,7,8-TCDF
NCASI Method TCDD/F - 88.01
471
-------
- 54 -
TABLE 4 CALIBRATION STANDARD SOLUTIONS
CONCENTRATION (pg/uL)
Compound
Native
2,3,7,8-TCDD
2,3,7,8-TCDF
Internal Standards
13C12-2,3,7,8-TCDD
13C12-2,3,7,8-TCDF
Recovery Standard
13C12-1,2,3,4-TCDD
Cleanup Recovery Standard
37Cl4-2,3,7,8-TCDD
CSS
1
0.5
0.5
2 3
5 50
5 50
4
250
250
5
500
500
100
100 100 100 100
100 100 • 100
100 100 100
10
25
50
NCASI Method TCDD/F - 88.01
472
-------
- 55 -
TABLE 5 QUANTITATION, INTERFERENCE AND LOCK MASS IONS
MONITORED
Accurate
Descriptor Mass
TCDF 303.9016
305.8987
315.9419
317.9389
375.8364
318.9793
TCDD 319.8965
321.8936
331.9368
333.9339
327.8855
318.9793
FOR HRGC/HRMS ANALYSIS OF TCDD/TCDFs
Ion
ID
M
M+2
M
M+2
M+2
LOCK
M
M+2
M
M+2
M
LOCK
Elemental
Composition
C12H435C140
C12H435C1337C10
13C12H435C140
13C12H435C1337C10
C12H435C160
PFK
C12H435C1402
C12H435C1337C102
13C12H435C1402
13C12H435C1337C102
C12H437C1402
PFK
Analyte
TCDF
TCDF
TCDF ( S ) a
TCDF ( S ) a
HxCDPEb
TCDD
TCDD
TCDD(S)a
TCDD(S)a
TCDD(SS)C
alnternal and/or Recovery Standard
bRexachlorodiphenylether interference monitoring ion
GCleanup Recovery Standard
NCASI Method TCDD/F - 88.01
-4^-10
4. / O
-------
- 56 -
TABLE 6 RECOMMENDED GC OPERATING CONDITIONS
Column Coating
Film Thickness
Column Dimension
Injector Temperature
Splitless Valve Time
Interface Temperature
DB-5
.25 urn
60 m x .25 mm id
280°C
40 Sec
Function Of The Final Temp.
Temperature Program
Init. Temp.
190
Init. Hold
Time (min)
1 min.
Temp
Ramp
(°C/min)
Final
Temp
(°C)
260
Final
Hold Time
(min.)
20
Column Coating
Film Thickness
Column Dimension
Injector Temperature
Splitless Valve Time
Interface Temperature
DB-225
.25 urn
30 m x .25 mm id
2508C
40 Sec
Function Of The Final Temp.
Temperature Program
Init. Temp.
190
Init. Hold
Time (min)
Temp
Ramp
(°C/min)
Final
Temp.
(°C)
235
Final
Hold Time
(min.)
15
It is recommended that the analyst bake out the DB-225 column
between runs.
NCASI Method TCDD/F - 88.01
4
-------
(1)
- 57 -
LITERATURE REFERENCES
"U.S. EPA/Paper Industry Cooperative Dioxin Screening
Study", EPA-440/1-88-025, (March 1988).
(2) LaFleur, L.E., Ramage, K., Gillespie, W.J., Miille, M.J.,
Luksemburg, W.J., Valmores, S., "Optimization of Extraction
Procedures for the Analysis of TCDD/TCDF in Pulp, Paper Base
Stocks, and Pulp Industry Solid Wastes", presented at The
8th International Symposium on Chlorinated Dioxins and
Related Compounds, Umea, Sweden,(August 1988).
4
•*•r;;.
( U
-------
U.S. Environmental Protection Agency
National Oioxin Study - Phase II
Analytical Procedures and Quality Assurance Plan
for the Determination of PCDD/PCDF in Fish
Environmental Research Laboratory
6201 Congdon Blvd
Duluth, MN 55804
July 26, 1988
'47C
-------
INTRODUCTION
This document, "Analytical Procedures and Quality Assurance Plan fur the
Determination of PCDD/PCDF in Fish" has been drafted in response to the need
for the Environmental Research Laboratory of Duluth (ERL-0) to perform analysis
for tetrachloro- to octachloro- congeners/isomers of polychlorinated cibenzo-
p-dioxins and dlbenzofurans (PCOD/PCDF).
These analyses are limited by lack of analytical standards, however uomer
specificity may be determined using specially developed standards. Analytical
results will, therefore, be reported as concentration (pg/g) for each gas
chromatography (GC) peak in a congener class by making the assumption that
the response for the molecular ion of all isomers in that class is equal to
the response observed for the isomer for which ERL-D does have a standard.
The target minimum level of detection (HLO) for specific PCDD/PCDF isomers is
given in Table 1 below. This document is meant to be only a guideline for
analyses and may be modified as needed to satisfactorily analyze any sample.
Table 1
. PCDD/PCDF Target Minimum Level of Detection
TCDD, TCDF : 1 pg/g
PeCDD, PeCDF 2 pg/g
HxCDO, HxCDF 4 pg/g
HpCDD, HpCOF 10 pg/g
OCDD. OCDF 40 pg/g
477
-------
TABLE OF CONTENTS
I. Sample Preparation
A. Grinding
B. Extraction
C. Anthropogenic Chemical Isolation
D. Florisil Chromatography
E. PCDO/PCOF Isolation
F.' Reagents
G. Percent Lipid Determination
II. Gas Chromatography/Hass Spectrometry Analyses
III. Analytical Standard Spiking Solutions
IV. Quality Assurance/Quality Control
A. General Procedures of Operation
B. Instrumental Quality Checks
C. Evaluation of Data
1. Accuracy
2. Precision
3. Signal Quality Assurance Requirements
0. Quality Assurance Problems and Corrective Actions
V. Quantification Procedures
A. Method Efficiency
B. Quantification of PCOO/PCOF
C. Signal Quality
0. Quantification Standards
478
i i
-------
I. Sample Preparation
A. Grinding: Fish tissue is ground frozen in a stainless steel power
meat grinder. The ground tissue is stored at -20°C in solvent
rinsed glass jars with aluminum lined plastic lids.
U. Extraction: Tissue (20 g) is blended with enough anhydrous sodium
sulfate to dry the tissue (100 g). Two-thirds of the sample is
placed in a glass Soxhlet thimble, spiked with Standard Solutions A
and 8 (Appendix A) and the remainder of the sample is added to the
thimble. The sample is extracted at least eight hours with 1:1
mixture of hexane and methylene chloride in a Soxhlet extractor.
The solvent is removed with a Kuderna-Danish (KD) apparatus, and
percent lipid is determined (see I.G.) before proceeding to the
anthropogenic chemical isolation step.
C. Anthropogenic Chemical Isolation: The sample extract is quanti-
tatively transferred to a 30 cm x 2.5 cm glass chromatography
column (MACRO-columns) fitted with a 300 ml reservoir on top,
which has been packed with plug of glass wool (bottom to top) 2 g
silica gel, 2 g potassium silicate, 2 g sodium sulfate, 10 g celite/
sulfuric acid and 2 g sodium sulfate, and previously washed with
100 ml hexane. The column is eluted with 100 ml benzene/hexane
(5X) and the element is collected in a Kuderna-Oanish (KD)
apparatus. Isooctane (1.0 ml) is added and the volume is reduced
and transferred to the florisil column.
D. Florisil Chromatography: A 1.0 cm x 20.0 cm glass chromatography
column fitted with a lt)0 ml reservoir is packed with a plug of glass
-1-
473
-------
wool (bottom to top) 5.0 cm (1.5 g) activated florisil and 1.0 cm
sodium sulfate. The column is washed with 20 ml methylene chloride
followed by 10 ml hexane. Sample and hexane rinses are quantita-
tively applied in small "plugs". The column is eluted with 2C ml 2%
methylene chloride/hexane, nd the eluent discarded. This wash is
followed by 50 ml methylene chloride which flows directly onto the
micro carbon/silca gel column for PCDO/PCOF isolation.
E. PCDD/PCDF Isolation: Effluent from the previous step is passed onto
a 4 nun x 200 mm column (micro-column) containing 300 mg silica gel/
carbon (see F.I.) which was fitted with a solvent reservoir. After
the sample has almost completely eluted from the micro-column, the
reservoir is washed with 2 x 2 ml benzene/methylene chloride (25%)
and the column is finally eluted with an additional 11 ml benzene/
methylene chloride. The PCDD/PCOF are then eluted with toluene
(20 ml) in a reverse flow direction. The toluene fraction is
collected in a pear shaped flask (25 ml) and reduced in volume to
0.1 ml (60°C). The sample is transferred to a microvial using
toluene to rinse the flask. The sample is stored in a freezer
until mass spectrometry analyses.
F. Reagents:
1. Solvents: Only pesticide grade distilled in glass solvents are
to be used. They are: hexane, isooctane, methylene chloride,
benzene, toluene, acetone, and methanol (Burdick and Jackson,
. Fischer Scientific or Mallincrodt).
-2-
480
-------
2. Sodium Sulfate: Sodium sulfate (Baker Chemical Company reagent
grade anhydrous) is baked at 650°C in a furnace for 24 hours,
cooled, and stored in an empty hexane solvent bottle.
3. Silica-Gel: Silica-Gel-60 (Merck-Darmstadt), is Soxhlet
extracted eight hours with methanol , air dried for 12 hours,
and vacuum oven dried (125°C) for 24 hours. It is then stored
in an empty hexane solvent bottle. Just before use it is
activated at 105°C for 24 hours.
4. Sulfuric Acid/Celite: Sulfuric acid (Baker Chemical Company,
Ultrex) (5 ml) is blended in a 250 ml beaker with Celite 545
(Baker) (10 g).
5. Potassium Silicate: High purity potassium hydroxide (Aldridge
Chemical Company) (56 g) is dissolved in methanol (300 ml).
Silica-gel (100 g) is added to the mixture and. stirred (1 hour,
60°C). The mixture is cooled, the solvent drained, and the
solids transferred to a Soxhlet thimble. The solids are extracted
with methanol for 4 hours, cooled, air dried in a hood, and
stored in a solvent rinsed bottled until use.
6. Silica Gel/Carbon; Silica Gel-60 (100 g) (Merck-Darmstadt) is
Soxhlet extracted with methanol (200 ml) for 24 hours, air dried
in a hood, and further dried in vacuum oven for 24 hours. AMOCO
PX-21 Carbon (5 g) Js added and then blended until uniform in color.
The Silica Gel/Carbon is stored in a closed jar at room
temperature until use.
-3-
481
-------
7. Florisil 60-100 mesh (Baker Analyzed) is soxhlet extracted
with methanol for 24 hours, air dried and activated (stored)
at 120°C.
G. Percent Lipid Determination: The percent lipid will be determined
for all fish. To determine percent lipid, the sample will be
extracted as in Section B of sample preparation. After sample
concentration, the KD lower tube is placed in a 60°C water bcth
under a gentle stream of dry filtered air. After any remaining
solvent has been evaporated, the lower tube and contents are weighed.
The lipid is then quantitatively transferred to the macro column as
described in Section C of sample preparation. After transfer, the
empty lower tube and boiling chips are weighed. The percent lipid
is calculated from the weight-differences..
4
43
-------
II. GC/MS Analyses
All gas chromatography/mass spectrometry analyses (GC/MS) will be done
on a Finnigan-HAT 8230 high resolution GC/high resolution MS (HRGC/
MRMS) system. Instrumental parameters are given in Appendix 8. Prior
co analyses each sample will be spiked with Standard Solution C
(Appendix A) and the sample volume adjusted to 20 uL.
III. Analytical Standard Spiking Solution
Each sample (20 g) will be spiked with 100 uL of Internal Standard
Solution A (stable isotope labeled PCOO/PCDF) and Internal Standard
Solution B (Appendix A) (surrogate analytes) prior to extraction, and
20 uL of Internal Standard Solution C (Recovery Internal Standard) just
before MS analysis. Appendix A provides details of the spiking solutions
The surrogate analytes are used by the data reviewer to insure that
calculated MLO values are reasonable.
IV. Quality Assurance/Quality Control (QA/QC)
A. General Procedures of Operation
1. Analysis of samples; Samples will be analyzed in sets of
twelve consisting of:
a. Blank: Method Blank (extraction apparatus) is prepared in
the laboratory and subjected to the same sample preparation
procedures as environmental samples. The Method Blank will
be used in every sample set.
b. Fortified Matrix: Native analyte (100 uL) (Appendix C)
are added to a blank sample matrix. The levels of fortifi-
cation of native analytes in the matrix spike will be above
the target detection limit, to provide an estimate of the
method's sensitivity, and for determination of percent
5
488
-------
accuracy of quantification. This sample may be substituted
with a reference sample that has been analyzed by at least
three labs and a mean value of contamination has been
established.
c. Detection Limit Verification Sample: An environmental
sample with detectable amounts of native analyte (determined
from a previous analysis) will be spiked with native analytes
(Appendix C) and analyzed with the next sample set. The
addition of the QA/QC sample will be done for only the
first three sample sets of any matrix type to establish
that the calculated MID is achievable. If analytical
results show difficulty in obtaining the MLD, then this
QA/QC sample must be in each set. If no problem is
experienced, then this QA/QC sample may be dropped.
d. Duplicate Sample: Two separate portions of the same
environmental sample will be processed and analyzed.
e. Environmental Samples: The total number of environmental
samples analyzed will be eight if the Detection Limit
Verification sample is used, otherwise nine samples will
be analyzed.
2. Sample Tracking and Labeling of Samples
a. Logging Incoming Samples: ERL-D completes the chain of
custody forms and informs the Sample Control Center (SCC)
that samples arrived safely or informs SCC of any problems
with the samples. Each sample received by ERL-D had
previously been assigned two numbers by the Sample Control
Center, the sample number (SCC0) and an episode number.
6
48'<1
-------
The SCCff number is unique for each sample and provides a
means for tracking a given sample throughout its analysis
and its permanent storage at the locker plant. The samples
are placed into freezer A upon arrival at ERL-Duluth,
homogenized, and an aliquot (100-500 g) is placed into
freezer B. After the samples are extracted they are put
into freezer C. If all the data meets QA requirements after
mass spectral analysis and quantification, the samples are
transferred to a locker plant for permanent storage (-20°C).
b. logging and Labeling Samples During Preparation: A laboratory
identification code (lab 10) is randomly assigned to each
sample in a set of twelve at the start of sample preparation.
The code consists of a letter, A through L, date of
extraction, and two initials of the sample preparation
chemist, i.e. A091587ML, and is used to identify the sampl-e.^
throughout the analysis period. The SCCJif, lab ID, sample
description, weight of sample, and amount of analytical
standards added to each sample are all recorded in a sample
preparation log book at the start of extraction. The lab
ID is written on labeling tape which is transferred from
beaker to flask during sample preparation. The lab ID is
written into the MS log book along with the mass spectra
analysis number.
3. Data System Sample Tracking: ERL-0 has developed the National
Oioxin Study (NOS) Phase II: Bioaccumulative Pollutant Study
Sample Tracking Database to facilitate record keeping and
summary report generation for each sample on the DEC-VAX 11/785
7
485
-------
(Digital Equipment Corporation). For each sample, including QA
samples, information pertinent to each sample is entered by the
preparation chemist. Quantification data (final concentration,
ion ratios, percent recovery, MLDs, and signal to noise) are
automatically uploaded to the database once all QA criteria
have been met. Appendix 0 is an example of the NOS database.
The first two letters of the SCC number indicate whatl«*»r
the sample is an Environmental , Method or Matrix Blank, or
Duplicate Sample. All environmental samples begin with the
letter 0. The Blank and Duplicate samples begin with the
letter Q followed by a D or an R for duplicate or reference
fish sample, respectively. Appendix E lists the possible cades
for the SCC number, and matrix type. Episode numbers for
Blanks and Fortified Matrix samples are entered as 0000.
B. Instrumental Quality Control
1. Gas Chromatograph
a. Operation and Maintenance: Operation and maintenance of the
gas chromatograph will be done according to manufacturer's
recommendations.
b. Column Performance: GC columns performance will be evaluated
by:
i. Resolution of 1,2,3,4-TCDO from 2,3,7,8-TCDD
(Appendix F).
ii. Correlation of relative retention time of all
biosignificant PCDD/PCDF, specifically calculated by
relative retention time of 13C12l,2,3,6,7,8-HxCDD to
37
C142,3,7,8-TCDO (Appendix F).
8
486
-------
iii. Elution of all PCOD/PCDF during analysis will use a
GC window defining solution of select PCDD/PCDF
(Appendix G.).
2. Mass Spectral Performance: The performance of the mass
spectrometer will be evaluated for resolution, sensitivity and
linearity. Sens-itivity and linearity will be evaluated using
calibration standards (Appendix H). The mass resolution used
for these analyses will be a minimum of 5000 (10* valley
*
definition). The mass spectrometer will be tuned each day to
the required resolution according to the procedures established
by the instrument manufacturer. The calibration curve must be
linear over the range of concentrations used in the calibration
standards. The percent relative standard deviations for the
mean response factors must be "less than 20 percent.
C. Evaluation of Data
1. Accuracy: Accuracy, the degree to which the analytical
measurement relects the true level present will be evaluated in
two ways for each sample set. These are: the difference of
measurement of a PCDD/PCDF isomer added to a blank matrix, or
difference of measurement of a PCDO/PCDF from the level in an
established reference material; and the efficiency for recovery
of the internal standard added for each congener group. QA
requirements for accuracy and method efficiency are provided in
Appendix I.
40 >-1
C l
-------
S Accuracy and 2 Method Efficiency are defined as follows:
% Accuracy = measured value X 100
amount native isomer added to blank matrix
2 Method efficiency = measured value X 100
amount internal standard added to each sample
2. Precision: Precision, a measure of mutual agreement amcng
individual measurements of the same pollutant in replicate
samples, will be evaluated for each sample set by the ratio of
the difference of duplicate values to their mean value.
Appendix I provides QA requirements for precision.
Precision is defined as follows:
Precision = difference between duplicate" values X 100
mean value for the duplicates
3. Signal Quality: The quality of the mass spectral signals used
for qualitative and quantitative analysis will be evaluated
using two parameters, the ion intensity ratio for the two ions
monitored in each congener group, and the signal to noise (S/N)
ratio. Appendix I provides QA requirements for signal quality.
In addition, qualitative identification will be based on
coelution with the stable isotope labeled compound, or relative
retention time correlation (Appendix F).
10
488
-------
0. Quality Assurance Problems and Corrective Action:
Problem Corrective Action
MS performance outside QA
GC column performance
outside QA.
Method efficiency outside
of QA.
Accuracy outside of QA for
spiked matrix.
Precision of duplicates
outside QA.
Detection of analyte in
blank for 2.3,7,8-.TCOO,
2,3,7,8-TCOF and
1,2,3,7,8-PCOO
For other analytes in
blank
Analyte exceeds calibration
standard range.
Method efficiency for
blank outside of QA or
blank lost
Adjust MS parameters for resolution,
rerun initial curve and reanalyze
sample(s).
Reanalyze standards and samples on
modified or alternate column.
If 2378-TCDO method efficiency <40%,
reanalyze sample set. If method
efficiency <40% for analytes other
than 2378-TCOD, flag and report data.
If more than 20« of the analytes are
outside of QA for accuracy and pre-
cision, reanalyze the sample set.
Reextract and reanalyze all samples
for which the level of contamination,
or MLO, is < 2.5X blank level.
Record blank concentration in comment
field of samples.
Mea-sure method efficiency. Dilute sample
100:1 respike with each standard
solution (A and B), adjust volume
and reanalyze.
Reextract and reanalyze all positives
in set.
Note: Because of the complexity of these analyses types, it is not expected
that all analytes will meet all QA criteria. Therefore, a complete review of
the data by a chemist is essential. Responsibility for the evaluation of data
will be that of the sample preparation chemist and the mass spectrometer operator.
Review of the data, including QA, and resolution of data quality problems will
finally be the responsibility of the Principal Investigator/Program Manager.
Resolution of data questions may require reanalysis of samples to include the
addition of confirmatory ions or analysis on different types of GC columns.
11
489
-------
V. Quantification Procedures
Quantification of analytes will be accomplished by assigning isomer
identification, integrating the area of mass specific GC peaks, c.id
calculating an analyte concentration based upon an ion relative
response factor between the analyte and standard.
A. Method Efficiency: The method efficiency for the recovery of stable
isotope labeled compounds is determined by calculating the amount of
stable isotope labeled compound in the final extract and dividing by
the amount spiked into the sample at the start of the cleanup procedure.
This is done by determining the relative response factor between 13C12
1,2,3,4-TCDO and the stable isotope labeled internal standard (Eq. 1)
and using the response factor to calculate the concentration of the
internal standard in the final solution (Eq. 2)._ The concentration in
the final solution times the final volume equals the total amount present.
Example calculation shown for 37C14 2,3,7,8-TCDD:
Determine Response Factor:
RF 328/334 = A328 x C334 Eq. 1
A334 x C328
Calculate Internal Standard Concentration:
C328 = A328 x C334 Eq. 2
A334 x RF328/334
Calculate Method Efficiency:
% Recovery = C328 found X 100
C328 spiked
where: RF = response factor
A - peak area
C = concentration of chemical
1.2
400
-------
B. Quantification of PCDD/PCDF: The concentration of a natural
PCDO/PCOF is determined by calculating a response factor between
PCDD/PCOF and the stable isotope labeled PCDO/PCOF for the congener
group (see Section V.F.). As shown in Eq. 3 in the example below,
and using the response factor to calculate the level of natural
PCOO/PCOF in the sample; (Eq. 4).
Example calculation shown for 2,3,7,8-TCDO:
Standard
RF322/334 = A322 x C334
Sample
V322
A334 x C322
A322 x S334
Eq. 3
Eq. 4
C.
A334 .x RF322/334
where: RF = response factor
A = peak area
C = concentration in standard
S = spiking level in sample
V = level of analyte in sample
Signal Quality
1. Minimum Level of Detection (HID): Minimum Level of Detection is
defined as the concentration predicted from the ratio of baseline
noise area to labeled standard area, plus three times the standard
error of the estimate derived from the initial .calibration curve
for the analyte of interest.
Initial Calibration Based Method of MLO: MID will be
estimated from the ratio of the noise area to the isotopically
labeled internal standard area, plus three times the standard
13
491
-------
error of the estimate (SE) for the area ratio, or Y-axis, of the
initial calibration curve. The Y-intercept (INT) is subtracted
from this quantity, in keeping with the normal formal is.-i fcr
"inverse prediction" of a point on the X, or concentrafiuii ratio
axis, from a point on the Y, or signal ratio axis. The SE term
is derived from an analysis of variance (ANOVA) performed during
the weighted least squares fit of the initial calibration curve.
This term represents the random error in the replicate iniecilons
used to generate the calibration curve, the error not accounted
for by the linear model. The weighting is necessary because of
the relation, often observed in instrumental analysis, of increasing
variance with increasing concentration. MLO, according to this
scheme, is defined below, using 2,3,7,8-TCOD as an example.
MID 5- [(An/A334) + (3 x SE). - INT] x C334
RF322/334 x K
where: An = noise area in the 2,3,7,8-TCOD window for
ion 322
A334 = labeled internal standard peak area in the
sample
INT = the Y-axis intercept on the initial calibration
. . curve
C334 = labeled internal standard concentration
K = constant to adjust for sample size and final
volume
RF322/334 = response factor for native/labeled
2,3,7,8-TCDD, the slope of the initial
calibration curve
SE = standard error of the estimate of the initial
calibration curve
In addition, fish tissue will be spiked with surrogate
analytes (see Internal Standard Solution B, Appendix A) prior to
extraction. The surrogate analytes will serve as an added check
14
49;
-------
to Insure that HLO values calculated from the initial calibration
I
curve, as discussed above, are reasonable.
2. Signal to Noise (S/N) : The method of determining the signal to
noise ratio is shown below.
Analyt* signal
Hols* signal
Analjte Sifnal Peak Area
— __ __ «—___———— — __—
- ———————————
Noise Signal Peak Area
The noise area will be calculated by Integrating over a peak
width equivalent to the. anal yte signal, typically about 10 seconds.
0. Quantification Standards: Quantification standards were prepared by
Wright State University. The concentration of 2.3,7,8-TCOO was
checked against a primary standard obtained from the U.S. National
Bureau of Standards. A table of the concentrations of each isomer in
each standard is given in Appendix H.
E. Qualitative Standards: ERL-0 has developed two qualitative analytical
standards, one containing all 75 PCDO's and all 138 PCDF's was
developed from an extraction of municipal Incinerator fly ash
(Appendix J) and the other containing only the biosignificant isomers
was developed by exposure of fish to an extract of municipal
incinerator fly ash and processing the exposed fish for PCOD/PCOF
(Appendix F). These standards will be used to assign structures for
isomer specific analyses.
15
433
-------
F. jm'tial and Daily Calibration of the HRMS: An initial calibration of
the instrument will be performed as needed. This will include making
*
three replicate injections of each calibration standard (see Appendix
H). Weighted least-squares linear regression will be used to generate
a calibration curve for each analyte. The weighting factor will be
inversely proportional to the variance among the replicate injections
of each calibration standard. The slope of the regression line will
be the response factor used to quantify the analyte. At lesst two
calibration standards will be injected daily to insure that any
response factors used for quantification and recovery calculations do
not deviate from the initial calibration by more than 20J. If the
daily calibration generates values outside this margin, and less
drastic corrective action does not solve the problem, a new set of
initial calibration curves will be generated and the old response
factor libraries discarded. An example of a typical calibration
curve, using 2,3,7,8-TCDO as an example, Is shown in Figure 1.
G. Integration of Automated Data Processing and Quality Assurance:
QA parameters for method efficiency, ion ratios, retention time
correlations, signal/noise"ratio, accuracy and precision will be
monitored with the aid of software either developed in-house, or
modified from existing programs included with the HRMS data system.
Raw data will be sorted and edited using the mass spectrometer's
dedicated data system, transferred to the DEC-VAX system and processed
using software programs RFACTOR and DFQUANT (Figure 2.). Data will
be reviewed by the Project Director before entering into the NDS
data base.
16
494
-------
Appendix A.
Internal Standard Solution A (100 ul)
Quantification Standards
Compound
37C14 2.3,7,8-TCDD
13C12 2,3,7,8-TCOO
i:C12 2,3,7,8-TCOF
I3C1Z 1,2.3,7,8-PCDO
13C12 1,2,3,7,8-PCOF
13C12 1,2,3,4,7.8-HxCDO
13C12 1,2,3,4,7,8-HxCDF
13C12 1,2,3,4,6.7,8-HpCOO
13C12 1,2,3,4.6,7,8-HpCDF
13*C12 OCDO
37C14 2,3,7,8-TCDF
Concentration in Solution
pg/ul
2.0
5.0
5.0
5.0
5.0
12.5
12.5
12.5
12.5
25.0
2.0
Concentration in Tissue*
ppt
10.0
25.0
25.0
25.0
25.0
62.5
62.5
62.5
62.5
125.0
10.0
Surrogate Analytes Internal Standard Solution B (100 ul)
1,2,3,4-TCDO
1,2,4,7,8-PeCDO
1,2.3.4-TCOF
1,2,3,6,7-PeCOF
1.0
1.0
1.0
1.0
Recovery Internal Standard Solution C (20 ul)
5.0
5.0
5.0
5.0
13C12 1,2,3,4-TCOO
50.0
50.0
* Assuming analysis on a 20 g aliquot of tissue.
17
495
-------
Figure 1.
2,3,7,8-TCDD
WEIGHTED CALIBRATION CURVE
UJ
ec
UJ
CONCENTRATION/
2 3
CONCENTRATION
SLOPE = RESPONSE FACTOR
-------
Figure 2.
DATA REDUCTION FOR PCDO/PCDF
NATIONAL DIOXIN STUDY
DAILY
CALIBRATION
STANDARDS
RFACTOR
SOFTWARE
YES
INITIAL
CALIBRATION
LIBRARIES
DEC-VAX
OR
IBM-PC
MASS
SPECTROMETER
DATA
SYSTEM
BEGIN")
*^^^^^^^^^^
(BEGIN
'DATA
PASSES
QA?
NO
CORRECTIVE
ACTION
SAMPLES
DFOUANT
SOFTWARE
YES
DATA BASE
i
i
GENERATE
FINAL REPORT
STOP
49V
-------
Appendix B.
HRGC/HRMS Operating Parameters
Data Acquisition: Multiple Ion Selection of the Following Ions
Compounds Mass Window m/z Value
TCDF
37C14-TCDF
13C12-TCOF
TCDO
37C14-TCDO
13C12-TCDO
PCDF
13C12-PCOF
PC DO
13C12-PCDD
HxCDF
13C12-HxCDF
HxCDD
13C12-HxCDD
HpCDF
13C12-HpCDF
HpCDO
13C12-HpCdd
OCOF
13C12-OCDF
OCDO
13C12-OCDD
1
1
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4
5
5
5
5
303.9016,
311.8898
317.9389
319.8965,
327.8847
333.9338
339.8597,
351.9000
353.8576,
367.8949
373.8207,
385.8610
389.8156,
401.8559
407.7817,
419.8220
423.7766,
435.8169
443.7398,
455.7801
457.7377,
471.7750
305.8986
321.8936
341.8567
355.8546
375.8178
391.8127
409.7788
425.7737
445.7369
459.7348
Note: Nominal masses will be used for low resolution MS. .
Sample Introduction: Capillary Column directly inserted into the ionizer.
lonization: Electron Impact, 70ev, 1mA emission current
Source Pressure: 1 x 10~5 torr.
Ionizer Temperature: 250°C
Mass Resolution: 5000, 10X valley
Scan Rate: 1 MIS cycles per second
GC Column: 30 n DB-5, 60 m SP2330
Linear Velocity: 35 cm/sec Helium
Temperature Program: 180°C (2 min.), 3°/min. to 240°. Z0°/min. to
280° (12 min.)
Mass windows are monitored sequentially during the temperature programs with
the windows defined by elution of standards.
18
-------
Comnound
Appendix C.
Native PCDO and PCOF spiking solutions (100 ill)
Concentration
(pg/ul Tridecane)
2,3,7,8-TCDD
2,3,7,3-TCDF
1,2,3,7,8-PCOD
1,2,3,7,8-PCDF
2,3,4,7,8-PCOF
1,2.3.4.7.8-HxCOO
1,2,3,6,7,8-HxCOO
1,2,3,7,8,9-HxCDO
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCOF
2,3,4.6,7,8-HxCOF
1,2,3,7,8,9-HxCOF
1,2,3,4,6.7,8-HpCOO
1,2,3,4,6,7,8-HpCDF
1,2,3.4,7,8,9-HpCDF
OCOO
OCOF
c
0.5
0.5
0.5
0.5
0.5
1.25
1.25
1.25
1.25
1.25
1.25
1.25
1.25
1.25
1.25
2.5
2.5
Solutions
0
1.0
1.0
1.0
1.0
1.0
2.5
2.5
2.5
2.5'
2.5
2.5
2.5
2.5
2.5
2.5
5.0
5.0
E
1.5
1.5
1.5
1.5
1.5
3.75
3.75
3,75
3.75
3.75
3.75
2.75
3.75
3.75
3.75
7.5
7.5
19
-------
Appendix 0.
NDS PHase II: Bioaccumulative Pollutant Study:
Sample Tracking System
ERL-0 loc: 25
EPISODE 0: 0000
Sampling Information:
Sampling Office:
State & City:
Sampling Contact:
Date Sampled: O/ O/ 0
Site Location:
Latitude: N 0 0' 0"
Analysis Lab: 0
Matrix Type: R
Analytical: PCOO/PCOF
Extraction Date: 7/14/86
GC/MS ID: MAT86824
LAB ID: K071486LH
Weight: 20.00
XLipid: 5.2
SCC |: QR071486
Longitude: WOO1 0"
Date Received: O/ O/ 0
Rerun: 0
Pesticide 4 Industrial
O/ O/ 0
Chemicals *
Comments:' Reference Fish 86
Mass Lipid on GPC:
0.00
0.0
OiOOOO
Quality Assurance (QA) Information for PCODs/PCDFs
Congener Minimum QA 2 Recovery
Tetra-PCDO/PCDF 50
Penta-PCDO 35
Penta-PCDF 35
Hexa-PCDD/PCOF 35
Hepta-PCDD/PCDF 35
Theoretical
0.76
0.61
1.53
1.23
1.02-
lon Ratio
Information pertinent to a related study.
20
-------
EPISODE I: 0000
Appendix D. continued
NOS Phase II: Bioaccumulative Pollutant Study
SCC I: QR071486
DATA 1-iH BIOSIGNIFICANT POLYCHLORINATED OIBENZOOIOXINS AND FURANS:
Analvte CAS NO.: I/R S/N 2REC OL
0.0000
0.9726
0.4863
0.0000
1.0892
1.6357
2.1784
4.0729
1.4654
0.7327
1.4654
0.7327
0.7327
1.3863
0.0000
1.3863
0.0000
0.0000
0.0000
2.3,/,8-TCOF
2,3,6,7-TCDF
3,4,6.7-TCDF
2,3,7,8-TCDD
1.2,?,7,8-PeCDF
2,3,',7,8-PeCDF
2.3,4,6,7-PeCDF
1.2,3,7,8-PeCDO
(1.2,3,4,7,8-HxCDF)
(1,2.3,4,6.7-HxCDF)
1,2,3,6,7,8-HxCOF
2,3,4,6,7,8-HxCDF .
1,2,3,7,8,9-HxCOF
1,2,3,4,7,8-HxCOO
1,2,3,6,7,8-HxCOD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
1.2,3.4,6,7,8-HpCDO
51207-31-9
1746-01-6
57117-41-6
57117-31-4
40321-76-4
70648-29-9
57653-85-7
19408-74-3
67562-39-4
55673-89-7
37871-00-4
32598-13-3
0.74
1.00
1.71
0.78
1.33
1.10
0.00
0.25
0.80
0.00
0.67
1.25
0.00
0.00
1.31
0.00
0.62
0.00
1.13
55.75
8.28
16.56
40.75
16.72
11.15
8.36
4.24
28.52
57.03
28.52
57.03
57.03
29.08
4.67
29.08
18.97
37.94
10.50
62
62
62
73
54
54
54
57
47
47
47
47
47
49
49
49
39
39
39
ERL-D Loc: 25
Amount(pg/g)
5.26
NO
NO
15.63
NO
NO
NO
NO
NO
NO
NO
NO
NO
"NO
3.23
NO
NO
NO
5.93
I/R = Ion Ratio; S/N = Signal to Noise; OL
( ) = Coeluting peaks on OB-5 column
Detection Limit
21
501
-------
Appendix E.
Codes for the SCC Number and Matix Type
SCC number first letter options:
D -- Environmental Samples
Q — QA samples
Second letter options for Environmental Samples:
- Region 1
Region
Region
- Region 4
- Region 5
- Region 6
G - Region 7
H - Region 8
Y - Region 9
J - Region 10
T - All regional data
Second letter optoins for QA samples:
B - Method or matrix blank
0 - Laboratory duplicate
R - Reference fish or fortified matrix
Matrix type:
PF - Predator Fillet
MB - Whole bottom
WP - Whole predator
R - Reference
Y - Blank
502
22
-------
I.
II.
Appendix F.
GC Column Performance Quality Control
Resolution of 1,2,3,4-TCDO from 2,3,7,8-TCDO will be used to evaluate
general column performance. Resolution (R) must be 0.75 or greater.
2d
ul
Relative retention times of 13CJ2 1,2.3,6,7.8-HxCDO tci 13C12 2.3,7,8-
ROO (or C,j 2,3,7,8-TCDO) wilt be used to evaluate column performance
for PCDO/PCOF analyses. Relative retention time calculated for these
standards for an analysis set should not change by more than 51 from
that established during the initial instrument calibration. Relative
retention time of biological PCDO/PCDF are shown below.
Relative Retention Times of Biosignificant PCOO/PCDF
Compound
2378-TCDF
2367-TCOF
3467-TCOF
2378-TCDD
12378-PCDF
23478-PCDF
23467-PCDF
12378-PCOD
123478-HxCOF
123678-HxCOF
123789-HxCDF
234678-HxCOF
123467-HxCDF
123478-HxCOO
123678-HxCDO
123789-HxCDO
1234678-HpCOF
1234789-HpCDF
1234678-HpCOO
RRT DBS
0.939
0.973
0.988
1.000
1.280
1.359
1.371
1.400
1.663
1.676
1.827
1.744
1.663
764
775
802
1.954
2.043
2.023
RRT SP2330
1.263 •
1.322
1.000
1.276
1.775
1.857
1.373
1.793
1.817
2.466
2.717
1.874
1.989
015
199
360
174
2.912
23
503
-------
Appendix G.
GC Elution Window Defining Solutions For DB-5 Column
Window First Eluting Compound Last Eluting Compound
TCOO 1,3,6,8- 1,2,8,9-
TCDF 1,3,6,8- 1,2,8,9-
PeCDO 1,2,4,7,9- / 1,2,4,6,8 1,2,3,8,9-
PeCOF 1,3,4,6,8- 1,2,7,8,9-
HxCDO 1,2,4,6,7,9- /1,2,4,6,8,9- 1,2,3,4,6,7-
HxCDF 1,2,3,4,6,8- 1,2,3,4,8,9-
HpCDO 1,2.3,4,6,7,9- 1,2.3,4,6,7,8-
HpCOF 1,2,3,4,6,7,8- 1,2,3,4,7,8,9-
504
24
-------
en
o
•en
ro
en
Calibration Standard
Appendix H
CALIBRATION STANDARDS
Concentrations in Calibration Solutions in pg/uL Tridecane
Ul W2 U3 U4 US W6
H7
W8
2,3,7,8-TCDD
2,3,7,8-TCOF
1,2,3,7.8-PeCDD
1,2,3.7.8-PeCDF
2,3,4.7,8-PeCDF
1,2,3,4,7,8-HxCDD
1,2.3.6,7,8-HxCDD
1,2,3,7.8.9-HxCDD
1,2,3,4.7,8-HxCDF
1.2,3.6,7.8-HxCDF
1,2,3,7,8,9-HxCDF
2.3.4.6,7.8-HxCDF
1,2,3,4,6.7.8-HpCDD
1,2.3,4,6.7.8-HpCDF
1,2,3.4.7.8,9-HpCDF
OCDD
OCDF
|Jc12-2,3,7,8-TCDD
3C o-2,3.7.8-TCDF
13CJ2-l,2,3.7,8-PeCDD
1 3 r i *5 i T o DnT nr
.,^19-1 ,2,3,7,8-PeLUr
13C o-l, 2.3.6. 7.8-HxCDD
13C o-l, 2,3,4, 7,8-HpCDF
13C ,-l. 2, 3.4.6.7 ,8-HpCDD
13c|2-l,2,3,4,6,7,8-HpCDF.
13C,o-OCDD
37Cn-2,3,7,8-TCDD
37ClJ-2,3,7.8-TCDF
13^5-1,2,3,4-1000
200
200
200
200
200
500
500
500
500
500
500
500
500
500
500
1000
1000
50
50
50
50
*t u
125
125
125
125
250
20
20
50
100
100
100
100
100
250
250
250
250
250
250
250
250
250
250
500
500
50
50
50
50
125
125
125
125
250
20
it
50
50
50
50
50
125
125
I 125
,|125
|! 125
125
125
125
125
125
250
250
50
50
50
50
125
125
125
125
250
20
20
50
25
25
25
25
25
62.5
62.5
62.5
62.5
62.5
62.5
62.5
62.5
62.5
62.5
125
125
50
50
50
50
125
125
125
125
250
20
it
10
10
10
10
10
25
25
25
25
25
25
25
25
25
25
50
50
50
50
50
50
125
125
125
125
250
20
it
5
5
5
5
5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
25
25
50
50
50
50
125
125
125
125
250
20
it
2.5
2.5
2.5
2.5
2.5
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
12.5
12.5
50
50
50
50
125
125
125
125
250
20
18
kJ \J
1
1
1
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
5
5
50
50
50
50
125
125
125
125
250
20
it
-------
Appendix I.
Quality Assurance Parameters
Ion Ratio
TCDD
PCDD
HxCDO
HpCDD
OCDO
TCOF
PCDF
HxCDF
HpCDF
OCOF
0.
0.
1.
.76+15*
.61+151
,23+152
1. 02*152
0.
0.
1.
1.
1.
1.
88+152
76+152
53+152
23+152
02+152
53+152
Method1-
Efficiency
>402,
>40%,
>402,
>402,
>401.
>40%.
>40I,
HOI.
>40J,
>40J,
<1202
<1202
<1202
<1202
<120I
<120X
<120X
<120I
<1202
<120X
Accuracy1-
at 10 pg/g
j+50%
+;50l
+1002
+1002
j+2002
^501
+;50l
+1002
^5002
j+5001
Precision^-
at 10 pg/g
+502
+501
+1002 .
+1001
+1002
+502
+501
+1002
+5002
+5001
S/N
Minimum
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
Variance of measured value from actual.
Variance of difference of duplicates from mean.
26
-------
Appendix J.
Relative Retention Times for 4-8 Chlorine PCDD/PCDF Isomers
Compound
1368
1379
1369
1378
1469
1247
1248
1246
1249
1268
1478
1279
1234-
1236-
1269-
1237-
1238-
2378-
1239-
1278
1267-
1289-
-TCDD
-TCOD
-TCDO
-TCOO
•TCOO
•TCDO
•TCOD
•TCOO
•TCOO
•TCOO
TCDO
TCOO
TCDO
TCOO
TCDO
TCOO
TCOO
TCOO
TCOO
TCOO
TCDO
TCDO
12468-
12479-
12469-
12368-
12478-
12379-
12369-
12467-
12489-
12347-
12346-
12378-
12367-
12389-
•PCDO
•PCOO
•PCOO
•PCOO
PCOO
PCDO
PC DO
PCOO
PCDO
PCOO
PCOO
PCOO
PCOO
PCOO
RRT 085
0.814
0.838
0.861
0.912
0.912
0.912
0.912
0.921
0.921
0.934
0.940
0.960
.985
,985
,985
0.993
0.993
.000
,009
.028
.048
0.
0.
0.
1.079
1.224
1.224
1.265
1.293
1.308
1.320
1.348
1.348
1.348
1.368
1.368
1.400
1.415
1.443
RRT SP2330
0.826
0.871
0.948
0.916
1.072
0.948
0.948
1.014
1.014
0.972
0.990
1.027
1.014
1.027
1.105
1.014
1.014
1.000
088
1.072
1.130
1.216
1
1.111
1.111
1.268
1.148
1.188
1.209
1.307
1.321
1.321
1.268
1.352
1.288
1.363
1.463
27
50'
-------
Appendix J. continued
Compound
124679-HxCDO
124689-HxCDO
123468-HxCDD
123679-HxCOD
123689-HxCDO
123469-HxCDD
123478-HxCDO
123678-HxCOD
123467-HxCOD
123789-HxCDO
1234679-HpCOO
1234678-HpCOD
12346789-OCDO
RRT 085
1.620
1.620
1.673
1.700
1.700
1.700
1.764
1.775
1.802
1.802
1.976
2.023
2.234
RRT SP2330
1.473
1.473
1.473
1.546
1.546
1.681
1.604
1.618
1.789
1.721
2.135
2.297
3.225
' 1368
1468
2468
1247
1347
1378-
1346-
2368-
1367-
1348-
1379-
1268-
1248-
1467-
1478-
1369-
1237-
2467-
1234-
2349-
1236
1469-
1238-
1278-
1349-
1267-
-TCDF
-TCDF
-TCDF
-TCDF
-TCDF
-TCDF
-TCDF
-TCDF
-TCDF
-TCDF
-TCDF
•TCDF
•TCDF
TCDF
•TCDF
TCDF
TCDF
TCDF
TCDF
TCDF
TCOF
TCDF
TCDF
TCDF
TCOF
TCDF
0.
0.
0.
0.
0.730
0.752
0.763
,782
.782
,782
.782
0.782
0.801
0.801
0.801
0.835
0.835
0.853
0.853
0.863
0.863
0.863
0.380
0.880
0.880
0.880
0.880
0.902
0.920
0.920
0.777
0.875
0.989
0.885
0.865
0.853
0.919
1.071
0.881
0.900
0.853
0.943
0.919
0.989
0.943
0.943
0.943
1.109
0.977
0.977
0.989
1.061
0.989
1.017
1.013
1.049
28
508
-------
Appendix J. continued
Compound
2378-
2348-
2347-
2346'
1246-
1249
1279
2367
1239
1269
3467
1289
•TCOF
•TCDF
•TCDF
•TCOF
•TCOF
• TCDF
•TCOF
•TCOF
•TCDF
•TCDF
•TCDF
•TCDF
13468-
12468-
23479-
12368-
12473-
13467-
12467-
13478-
13479-
23469-
12479-
13469-
23468-
12469-
12347-
12346-
12348-
12378-
12367-
23489-
12379-
23478-
12489-
13489-
12369-
23467-
12349-
12389-
PCDF
PCDF
PCOF
PCDF
PCDF
PCDF.
PCOF
PCOF
PCDF
PCDF
PCDF
PCDF
PCDF
PCDF
PCDF
PCOF
PCOF
PCOF
PCDF
PCDF
PCDF
•PCDF
•PCDF
• PCOF
•PCOF
•PCOF
•PCOF
-PCDF
RRT DBS
0.939
0.939
0.939
0.939
0.939
0.939
0.939
0.973
0.988
0.988
0.988
1.071
123468-HxCDF
134678-HxCDF
124678-HxCDF
134679-HxCDF
120
120
190
202
1.202
1.202
1.202
1.202
1.217
1.217
1.233
1.253
1.253
1.253
1.253
1.253
1.280
1.280
1.295
1.309
1.309
1.359
1.359
1.359
1.359
1.371
1.392
1.446
1.556
1.570
1.570
1.570
RRT SP2330
1.169
1.175
1.140
1.193
0.940
1.071
1.049
1.206
140
1.162
1.264
1.341
1
1.008
1.028
1.065
1.103
1.121
1.142
1.160
1.083
1.103
1.173
1.142
1.204
1.278
1.278
1.173
1.231
1.216
1.216
1.252
1.388
1.237
1.557
1.446
1.350
1.373
1.612
1.420
1.590
1.336
1.370
1.348
1.348
29
503
-------
Appendix J. continued
Compound
124679
124689
123467
123478
123678-
123479-
123469-
123679-
123689-
234678-
123789-
123489-
•HxCOF
•HxCDF
HxCDF
HxCDF
HxCDF
HxCDF
HxCDF
HxCDF
HxCDF
HxCDF
HxCDF
HxCDF
1234678-HpCDF
1234679-HpCOF
1234689-HpCDF
1234789-HpCDF
12346785-OCDF
RRT DBS
1.602
1.621
1.663
1.663
1.676
1.676
1.712
1.730
1.744
1.744
1.827
1.827
1.954
1.979
2.024
2.043
2.240
RRT SP2330
1.428
1.521
1.533
1.489
1.502
1.489
1.668
1.562
1.668
2.012
1.871
1.940
1.936
2.001
2.161
2.463
3.165
30
510
-------
APPENDIX B
ITD SPECIFIC TABLES
511
-------
TABLE 1: COMPOSITION OF THF. SAMPLE FORTIFICATION AND RECOVERY
STANDARDS SOLUTIONS
(concentrations in pg/uL)
A n a 1 y t e
Sample Fortification
Solution
Recovery Standards
Solut i on
1 3Ci 2 -2,3,7 ,8-TCDD (IS)
' 3Ci 2 -1 , 2, 3 ,4-TCDD ( RS )
37Cl4 -2,3,7,8-TCDD (SS)
» 3Ci 2-1 ,2,3,7 ,8-PeCDD (IS)
i 3Ci 2 -1 ,2,3,6, 7.8-HxCDD (IS)
i 3Ci 2 -1,2,3,7,8,9-HxCDD ( RS )
1 3Ct 2 -1 ,2,3,4 ,6,7,8-HpCDD (IS)
l3Ci2-OCDD (IS)
i 3Ci 2-2,3,7,8-TCDF (IS)
i3Ci2-l,2,3,7,8-PeCDF (IS)
i3Cl2 -1 ,2,3,4, 7,8-HxCDF (IS)
i3Ci2-l,2,3,7,8,9-HxCDF (SS)
i 3Ci2'-l ,273,4 , 6,7,8-HpCDF ( IS)
100
-
100
100
100
-
100
200
100
100
100
100
100
-
500/100*
-
-
-
500/100*
-
-
_
-
-
-
~
IS = Internal Standard; SS = Surrogate Standard; RS
Standard
= Recovery
Note: The volume of the recovery standards solution added to the
final extract before GC/MS analysis is 20 uL.
* 500 pg/ul for sludge sample and 100 pg/ul for water samples
51
-------
TABLE 2: COMPOSITION OF THE INITIAL CALIBRATION SOLUTIONS
USED FOR HIGH LEVEL SAMPLES
Compound
So 1 . Number
Unlabeled
Analy tes
2,3,7,8-TCDD
2,3,7, 8-TCDF
1 ,2,3,7,8-PeCDD
1 ,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4 ,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,7,8-HxCDF
1,2,3, 6,7, 8-HxCDF
. 1,2,3,7,8,9-HxCDF
" 2 , 3 , 4', 6 , 7 ", 8-HxCDF
1,2,3,4,6,7,8-HpCDD
1 ,2,3,4,6,7,8-HpCDF
1, 2,3,4,7,8,9-HpCDF
OCDD
OCDF
Internal
Standards
13Ci 2 -2,3,7,8-TCDD
»3Ci2-l ,2,3,7,8-PeCDD
13 Ci 2-1, 2, 3, 6, 7, 8-HxCDD
13 Ci 2-1, 2, 3, 4, 6, 7, 8-HpCDD
1 3Ci2 -OCDD
13Ciz -2,3,7 , 8-TCDF
13 Ci 2-1, 2, 3,7, 8-PeCDF
1 3 Ci 2 - 1 , 2 , 3 , 4 , 7., 8-HxCDF
i3 Ci 2-1, 2, 3, 4,6, 7, 8-HpCDF
Surrogate
Standards
37C1< -2,3,7,8-TCDD
13C12-1 ,2,3,7,8,9-HxCDF
Recovery
Standard
1 3Ci 2 - 1 ,2, 3 , 4-TCDD
1 3G, 2 -1 , 2.3,7,8.9-HxCDD
1
5
5
25
25
25
25
25
25
25
25
25
25
25
25
25
50
50
100
100
100
100
200
100
100
100
100
5
5
100
100
Concen Lra t. i ons (pg/uL)
2 3 -5 5
50
50
250
250
250
250
250
250
250
250
250
250
250
250
250
1000
1000
100
100
100
100
200
100
100
100
100
50
50
100
100
100
100
500
500
500
500
500
500
500
500
500
500
500
500
500
1000
1000
100
100
100
100
200
100
100
100
100
100
100
100
100
500
500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
5000
5000
100
100
100
100
200
100
100
100
100
500
500
100
i no
1000
1000
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
10000
,10000
100
100
100
100
200
100
100
100
100
1000
1000
100
inn
513
-------
TAI1LE 3. COMPOSITION OF THE INITIAL CALIBRATION' SOLUTIONS
USED TOR LOW LEVEL SAMl'LES
Co in pound
So L . Number
Un labe Led
Analy tes
2,3,7 ,8-TCDD
2,3,7,8-TCDF
1,2,3,7 ,8-PeCDD
1 ,2., 3,7,8-PeCDF
2,3,4,7 ,8-PeCDF
1,2,3,4,7 ,8-HxCDD
1,2,3,6,7 ,8-HxCDD
1 , 2,3,7,8,9-HxCDD
1,2,3,4,7, 8-HxCDF
1 ,2,3,6, 7,8-HxCDF
1,2,3,7,8,9-HxCDF
'2,3,4', 6,7','8-HxCDF
1,2,3,4,6,7,8-HpCDD
1,2,3,4,6,7 ,8-HpCDF
1,2,3,4,7 ,8,9-HpCDF
OCDD
OCDF
Internal
Standards
l3Ciz -2,3 ,7,8-TCDD
13C12 -1,2,3,7 ,8-PeCDD
13 Ct 2-1, 2,3, 6, 7, 8-HxCDD
13Ci2 -1 ,2,3,4 ,6,7,8-HpCDD
l3Ct 2 -OCDD
1 3Ct 2 -2,3,7,8-TCDF
13d 2 -1 ,2,3,7,8-PeCDF
1 3d 2 -1 ,2,3,4 ,7,8-HxCDF
I3d 2 -1,2, 3,4, 6, 7 ,8-HpCDF
Surrogate
Standards
37C14 -2,3,7,8-TCDD
1 3C, 2 -1 ,2.3,7 ,8,9-HxCDF
Recovery
S tandard
1 3 C, 2 -1 ,2 , 3 ,4-TCDD
1 3C, 2 -1 ,2 , 3. 7 ,8 ,9-HxCDD
TAHLK •) . iOr.'-AIUJNDANCK RATIO
Conce
1
0.5
0.5
2.5
2.5
2.5
2.5
2.5
2.5
2. 5
2.5
2.5
2.5
2.5
2.5
2.5
5
5
100
100
100
100
200
100
100
100
100
0.5
0.5
100
100
ACCEPT
514
2
1
1
5
5
5
5
5
5
5
5
5
5
5
5
5
10
10
100
100
100
100
200
100
100
100
100
1
1
100
100
AHI.I-:
ntrat i
3
5
5
25
25
25
25
25
25
25
25
25
25
25
25
25
50
50
100
100
100
.100
200
100
100
100
100
5
5
100
100
RANiCI-
ons ( pg/uL)
4
50
50
250
250
250 '
250
250
250
250
250
250
250
250
250
250
500
500
100
100
100
100
200
100
100
100
100
50
50
100
100
:s FOR
5
100
100
500
500
500
500
500
500
500
500
500
500
500
500
500
1000
1000
100
100
100
100
200
100
100
100
100
100
100
100
100
I'cnn
6
500
500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
2500
5000
5000
100
100
100
100
200
100
100
100
100
500
500
100
100
s AND
-------
TABLE
Number of
Ch
At.
a)
b)
1. o r i n e
oms
4
5
6
6a
7»
7
8
Used only
Used only
I on
Typo
M/M + 2
M+2/M+4
M+2/M+4
M/M+2
M/M+2
M+2/M+4
M+2/M+4
for i^C-HxCDF
for l3C-HpCDF
Theore
Ra 1.
0.
1 ,
1 ,
0.
0.
1 .
0.
t i ca 1
i.o
77
55
2-1
51
44
04
89
Con
trol
Lowe r
0.
1 .
1 .
0.
0.
0.
0.
65
32
05
43
37
88
76
L i in
i ts
l/'ppe r
0
1
1
0
0
1
1
.89
.78
.43
.59
.51
. 20
.02
515
-------
TABLE 5. ELEMENTAL COMPOSITIONS AND EXACT MASSES OF THE IONS
MONITORED I'.Y HIGH-RESOLUTION MASS SPECTROMETRY FOR
PCDDs AND PCDFs
Descriptor Accurate
Number Mass*
1 303.9016
305.8987
315.9419
317 .9389
319.8965
321 .8936
327 .8847
331.9368
333.9339
375.8364
[354.9792]
. 2 339.8597
'" " '""' 341 .8567
351 .9000
353.8970
355 .8546
357.8516
367,8949
369.8919
409.7974
[354.9792]
3 373.8208
375.8178
383.8639
385.8610
389.8157
391.8127
401.8559
403.8529
445-7555
[430.9729]
4 407.7818
409.7789
417.8253
419.8220
423.7766
425.7737
435.8169
437.8140
479.7 165
[•130.9729]
Ion
Type
M
M + 2
M
M+2
M
M + 2
M
M
M + 2
M + 2
LOCK
M + 2
M + 4
M+2
M + 4
M + 2
M + 4
M+2
M+4
M + 2
LOCK
M+2
M+4
M
M+2
M + 2
M+4
M + 2
M + 4
M+4
LOCK
M+2
M+4
M
M + 2
M + 2
M + 4
M+2
M+4
M + 4
LOCK
Elemen ta 1
Compos i U. Lon
d 2H4 3SCl4O
C, zH,3 SC1337 CIO
1 3Ci 2H4 3SCl40
1 3Ci 2H« 3 SC1337C1O
Cl2H43SCl402
C12H43SC1337C102
C,2H437C1402
1 3Ci 2H, 3 SC14O2
1 3Ci2H4 3 5C1337C1O2
Ci2H43 5Cls37ClO
C$ Fl 3
Ci2H3 3SC1437C1O
Ci2H33SCl337Cl2O
1 3Ci2H33 SC1437C1O
l3Ci2H3 3SC1337C120
Ci2H335Cl437ClO2
Ci2H33 SC1337C12O2
13Cl2H33SCl437C102
l3Ci2H33SCl33-7Cl2O2
Ci2H33SCl637ClO
C9Fi3
Ci2H23SCls37C10
Ci2H23SCl437Cl2O
13Ci2H23SCl60
1 3Ci2H2 3SCls37ClO
Ci2H235Cls37C102
Ci2H2 3SCl4 37C12O2
1 3Cl2H23SCls37C102
1 3Ct 2H2 3 SC14 37CL2O2
C12H2 3SC1637C120
C9F, 3
Ci 2 H35C16 37C1O
Ci 2H3SCls37Cl20
1 3Ci 2H3sci7O
1 3Ci 2H3SC1637C10
Ci2H3sCl637C102
Ct 2H3 SC1S37C12O2
1 3d 2II3 SC1637C102
1 3C, 2 H3scis37Cl2O2
C, 2 H3SC17 37C120
C, F,7
Analy te
TCDF
TCDF
TCDF (S)
TCDF (S)
TCDD
TCDD
TCDD (S)
TCDD (S)
TCDD (S)
HxCDPE
PFK
PeCDF
PeCDF
PeCDF (S)
PeCDF (S)
PeCDD
PeCDD
, PeCDD (S)
PeCDD (S)
HpCDPE
PFK
HxCDF
HxCDF
HxCDF (S)
HxCDF (S)
HxCDD
HxCDD
HxCDD (S)
HxCDD (S)
OCDPE
PFK
HpCDF
HpCDF
HpCDF (S)
HpCDF (S)
HpCDD
HpCDD
HpCDD (S)
HpCDD (S)
NCDPE
PFK
516
-------
Tab.
Con t. i nued
f) 4 4 1
•143
457
459
469
471
513
(442
.7428
.7399
.7377
.7348
.7779
.7750
.6775
.9728]
M+2
M+4
M + 2
M+4
M+2
M + 4
M + 4
LOCK
C,
d
d
C,
1 3 Ci
1 3 Ci
Ci
Ci
2
2
2
2
2
2
2
0
3
3
3
3
3
3
3
F
5 C 1 7
5CU
SC17
SGU
SC17
'CU
sCle
1 7
3
3
3
3
3
3
3
7
7
7
7
7
7
7
CIO
C120
C102
CljOj
C102
C1202
C120
OCDF
OCDF
OCDD
OCDD
OCDD
OCDD
DCDPE
PFK
(S)
(S)
a) The following nuclidic masses were used:
H
C
13C
F
1 .007825
12.000000
13.003355
18.9984
- 15
- 34
= 36
994915
968853
965903
S = Labeled Standard
i '7
-L (
-------
TABLE 6: GAS CHROMATOGKAPHY CON1!) IT IONS
Column type
Length (m)
i.d. (mm)
Film Thickness (urn)
Carrier Gas
Carrier Gas Flow (mL/min)
Injection Mode
Valve Time (s)
Initial Temperature (* C)
Program Temperature
DD-5
60
0.25
0.25
He 1ium
1-2
splitless
30
150
150* C to 190° C ballistically
then 3° C/min up to 300° C.
518
-------
TABLE 7. INITIAL AND CONTINUING CALIBRATIONS
MINIMUM REQUIREMENTS
RESPONSE rACTORS
Compound
Unlabeled
Analytes
2,3,7,8-TCDD
2,3,7,8-TCDF
1,2, 3, 7,8-PeCDD
1,2,3, 7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4 ,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4, 7, 8-HxCDF
1,2,3,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
•2,3;4',6,"'7",8-HxCDF
1,2, 3,4,6,7,8-HpCDD
1,2,3,4,6,7,8-HpCDF
1,2, 3,4,7,8, 9-HpCDF
OCDD
OCDF
Internal
Standards
i s Ci 2 -2,3,7,8-TCDD
13Ci2-l,2,3,7,8-PeCDD
is Ci 2-1, 2, 3, 6, 7, 8-HxCDD
i3Ci2-l,2,3,4,6,7, 8-HpCDD
l3Ci2-OCDD
i3d 2 -2,3,7,8-TCDF
l3Ci2-l,2,3,7,8-PeCDF
i3Ci2-l,2,3,4,7,8-HxCDF
i3Ci2-l,2,3,4,6,7,8-HpCDF
Surrogate
Standards
37C14 -2,3,7,8-TCDD
13 Ci 2-1, 2, 3, 7, 8, 9-HxCDF
RelaL i ve
I-Cal %RSD
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
30
25
30
25
30
30
30
30
30
30
25
30
Response Factors
Con-Cal %Delta
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
30
t
25
30
25
30
30
30
30
30
30
25
30
519
-------
Determination of Tetra-, Hexa-, Hepta-, and
Octachlorodibenzo-p-dioxin Isomers in Particulate Samples at
ivarts per Trillion Levels
L. Lamparski* and T. J. Nestrick
Analytical Laboratories. 574 Building, Dow Chemical US.A, Midland. Michigan 48640
An analytical procedure Is presented which permits the Iso-
mer-specific determination ol tetra-, hexa-, hepta-, and octa-
chlorodlbenzo-p-dioxins simultaneously at parts per trillion
concentrations. Typical data are presented to establish Its
applicability on a variety of environmental particulate samples.
The use of a highly specific sample clean-up procedure based
on multiple chromatographies Is shown to permit the Iso-
mer-speciflc determination of 2,3,7,8-tetrachlorodibenzo-p-
dloxin (2378-TCDD) by packed-column gas chromatogra-
phy-tow-resolution mass spectrometry In the presence of any
or all other TCDD Isomers.
The determination of parts per trillion (10*12 g/g, pptr)
concentrations of chemical residues generally requires the use
of either highly selective sample purification procedures
and/or very specific detectors (1,2). As detection limits are
lowered, the number of possible interferences present at
significant concentrations increases dramatically (3). Don-
aldson (4) has surmized that every known organic chemical
\uld be detected in water at a level of 1(T15 g/g or higher.
<-u>45imuarly, considering an analysis at the 10 pptr concentration
level in a sample matrix that is 993% pure, intcrfL rt ~.ces from
kas many as 10* compounds at concentrations 10s times higher
Jthan the component of interest are possible. Naturally the
addition of interferences from sources other than'the sample
matrix can make this task formidable. Such contamination
of laboratory reagents by a multitude of compounds has been
reported (5-79). Indeed, in some cases, the controlling factor
in determining the limit of detection (LoD) for a given analysis
is not the instrumental sensitivity of the detector but the
apparent response observed in reagent blanks (20-22).
This paper reports the development of an analytical pro-
cedure which permits the isomer-specific determination of
2378-TCDD at low parts per trillion concentrations, even in
matrices that have been intentionally fortified with equivalent
amounts of each of the other 21 TCDD isomers. Higher
chlorinated dioxins, including hexachlorodibenzo-p-dioxins
(HCDDs, 10 possible isomers), heptachlorodibenzo-p-dioxins
(H7CDDs, 2 possible isomers), and octachlorodibenzo-p-dioxin
(OCDD), can also be determined at low parts per trillion levels
by using this technique. In regards to the isomer-specific
determination of 2378-TCDC, the other 21 TCDD isomers
may also be considered as possible interferences. Several
publications have recently appeared which demonstrate CDD
determination capabilities but do not provide complete TCDD
isomer specificity (23-32).
EXPERIMENTAL SECTION
. Reagents. The preparation of 44% concentrated sulfuric acid
Jon silica, 10% silver nitrate on silica, basic alumina, and purified
nitrogen (Femtogas) have been described (1).
Silica. This adsorbent is prepared from chromatographic grade
> silicic acid as described for the preparation of 44% sulfuric acid
on silica (7).
33% 1 M Sodium Hydroxide on Silica. The silica support
is prepared as described (1). Activated silica is weighed into an
appropriately sized glass bottle. On the basis of the support
weight, the amount of 1 M aqueous sodium hydroxide necessary
to yield a reagent containing 33% by weight is added in a stepwise
fashion with shaking to produce a uniformly coated, free-flowing
powder.
Chemicals and Solvents. All solvents used are Burdick and
Jackson, distuled-in-glass quality. Laboratory chemicals (H^SO^
AgNOa, NaOH) are ACS reagent grade. These materials are tested
by subjecting them to the analytical procedure to verify the
absence of contamination. Spectrophotometric grade Gold-label
n-hexadecane was obtained from Aldrich Chemical Co. (Mil-
waukee, WI) and was purified by passage through basic alumina.
Expendables. Pyrex glass wool, silica boiling stones, and
disposable pipettes are cleaned before use. Glass wool and boiling
stones are Soxhlet extracted *~ 1 h consecutively with the following
solvents: methanol, chloroform + benzene (1:1 by volume),
benzene, and methylene chloride. They are then dried in a hot
air oven at ~160 °C for ~1 h. Disposable pipettes are cleaned
ultrasonically in deionized water and then methanol and finally
methylene chloride prior to drying at ~160 °C. Final sample
residues are stored in Reacti-Vials obtained from Pierce Chemical
Co. (Rockford, EL). The vials are cleaned by washing with de-
tergent and water and then boiled sequentially in benzene +
chloroform + methanol (1:1:1 by volume), benzene + chloroform
(1:1 by volume), benzene, and finally methylene chloride. They
are air-dried and again rinsed with methylene chloride imme-
diately before use.
Dioxin Standards. The primary standard of 2378-TCDD was
prepared by W. W. Muelder (Dow Chemical Co.) and its structure
was confirmed by single-crystal X-ray diffraction techniques (33).
Purity was assessed at 98% by mass spectrometry. Standards
of other TCDD isomers were synthesized and isolated as previously
described (34). Primary standards of 1,2,3,4,6,7,8-heptacnloro-
dibenzo-p-dioxin (1234678-HrCDD) and OCDD were synthesized
by H. G. Fravei and W. W. Muelder (Dow Chemical Co.). A
standard containing two HCDD isomers was prepared by Aniline
(35). Standards of 1234679-HTCDD and the 10 HCDD isomers
were synthesized and isolated in a manner similar to that reported
for TCDDs (34). Isotope-enriched 13C-2378-TCDD and UC-
123478-HCDD were synthesized by A. S. Kende (University of
Rochester, Rochester, NY). Mass spectrometric analysis indicated
these standards to be 86 atom % and 43 atom % "C, respectively.
Perchlorination of the "C-2378-TCDD provided 13C-OCDD.
Apparatus. Reverse-Phase High-Performance Liquid Chro-
motognphy (RP-HPLO. Residues containing chlorinated dioxins
are injected into the RP-HPLC system: column, two 6.2 X 250
mm Zorbax-ODS (DuPont Instruments Division, Wilmington, DE)
columns in series; isocratic eluent, methanol at 2.0 mL/min; pump,
Altex Model 110A; column temperature, 50 °C; UV detector,
Perkin-Elmer Model LC-65T liquid chromatographic column oven
and detector operated at 0.02 aufs at 235 nm; injector, Rheodyne
Model 7120 with 50-pL sample loop.
Normal-Phase Adsorption High-Performance Liquid Chro-
matography (Silica-HPLC). Residues containing TCDDs are
injected into the sflica-HPLC system: column, two 6.2 X 250 mm
Zorbax-SIL (DuPont Instruments Division) columns in series;
isocratic eluent, hexane at 2.0 mL/min; pump, Altex Model 110A;
column temperature, ambient; UV detector. Laboratory Data
Control Model 1204 variable-wavelength detector at 0.05 aufs at
235 nm; injector, Rheodyne Model 7120 with 100-jiL sample
injection loop. The columns were activated by the procedure of
Bredeweg et aL (36).
Packed-Column Gas Chromatography--Low-Resolution Mass
Spectrometry (GC-LRMS). Chlorinated dioxin quantification
Reprinted from Analytical Chemistry, 1980, 52, 2045.
Copyright © 1980 by the American Chemical Society and reprinted by permission of the copyright owner.
520
-------
2046 • ANALYTICAL CHEMISTRY, VOL. 52. NO. 13. NOVEMBER 1980
'
—
was accomplished by GC-LRMS using a Hewlett-Packard Model
5992-A operating in the selected ion mode (SIM) at unit resolution:
column, 2 mm Ld. x 210 cm silylaced glass; packing. 0.60% OV-L7
siiicone -t- 0.40% Poly S-179 on 80/100 mesh Pennabond Methyl
Silicone-10 cycle (HNU Systems Inc, Newton. MA); injection port
temperature. 280 °C on-column; carrier gas. helium at 14 cm1/mm:
separator, single stage glass jet operating at column temperature;
electron energy, 70 eV. TCDD analyses conditions: column
temperature, 246 °C isothermal; ions monitored, native TCDDs
at mfe 320. 322. and 324, and 13C-2378-TCDD internal standard
at m/e 332. Higher chlorinated dioxin analyses conditions:
column temperature, programmed from 230 to 300 "C at 10
"C/min and hold at Tnaiimirm- ions monitored, native HCDDs
at m/e 388, 390, and 392. native H,CDDs at m/e 422. 424. and
426, and native OCDD at m/e 458. 460. and 462, "C-123478-
HCDD and ^-OCDD are monitored at m/e 398 and 470, re-
spectively.
Environmental Paniculate Samples. Industrial Dust.
Particulates were removed from the air intake filtration system
from a research building located in Midland, ML
Electrostatically Precipitated Fly Ash. Particulates were
collected from the ash-removal system associated with the elec- .
troatatic precipitator on the Nashville Thermal Transfer Corp.
refuse incinerator located in Nashville, TN.
Activated Municipal Sludge. Representative samples were
removed from the center of a commercially purchased 20-kg bag
of Milwaukee Milorganite.
Urban ParticulateJAatter. Standard Reference Material No.
1648 was obtained from the National Bureau of Standards (NBS).
European Fly Ash. Particulate emissions from a municipal
trash incinerator were collected on filter paper by a nonisokinetic
sampling procedure. The location of the sampling port was
downstream from the electrostatic precipitator. This incinerator
was not operated to recover energy for'power generation.
' Sample Preparation Prior to GC-LRMS STM quantification,
the sample is prepared by using five basic steps: (1) chlorinated
diozins removal from the matrix via hydrocarbon extraction, (2)
chemically modified adsorbent txeafinent of th* extract to remove
easily oridizable species, (3) adsorbent treatment to remove
common chemical interferences, (4) RP-HPLC residue fraction-
ation to remove residual chemically aim tin* interferences and to
separate diozins present into groups according to their degree of
chlorination, and (SXsflica-HPLC refractionation of the RP-HPLC
TCDD fractions to provide a second high-efficiency chromato-
graphic separation having different selectivity to remove residual
interferents and to permit TCDD isomer specificity.
An appropriately sized all-glass Soxhlet extraction apparatus
equipped with a water-cooled condenser, a 43 X 125 mm glass
thimble with coarse frit, a 250-mL boiling flask, and a tempera-
ture-controlled heating mantle is assembled. Each of the-parts
is thoroughly scrubbed with an aqueous detergent solution, rinsed
with deionized water followed by acetone, methanol, and meth-
ylene chloride, and finally air-dried. Depending on the particulate
sample size (larger samples require most), 5-15 g of silica is
charged into the thimble followed by a plug of glass wool large
enough to cover the silica bed completely. The assembled system
(thimble installed) is charged with benzene (~250 mL) and al-
lowed to reflux at a recycle rate of ~20 mL/min for a minimum
period of 2 b. Following this preextraction period, the system
is permitted to cool and the total benzene extract is discarded.
The extraction thimble is removed and allowed to drain completely
on a. clean wire stand in a fume hood. The glass wool plug is
removed with clean forceps while a representative particulate
sample, ranging from 50 mg for filtered airborne particulates to
100 g for heavy soils, is quickly charged on top of the silica bed.
The glass wool plug is replaced and the thimble returned to the
Soxhlet extractor body. At this Hma aliquots of isooctane internal
standard solutions containing isotopically labeled 2378-TCDD,
123478-HCDD, and OCDD are introduced directly into the
particuiatea bed. The system is recharged with fresh benzene and
exhaustively extracted at the rate previously described for a
minimum period of 16 n. Each sample or set should have at least
one system treated as described for the sample to serve as a
reagent blank.
Upon completion of the prescribed extraction period, the flask
mnfj»inint? t.h<» honTpnp »YT.rarr. is rpmovorl nnfi fittwi with a t.hrw-
r
31
I/
u
6 mm
Figure 1. Liquid chromatograprtic dean-up columns.
to six-stage Snyder distillation column. The volume of the extract
solution is then reduced by atmospheric pressure distillation of
the benzene solvent to a final volume of approximately 25 mT..
The concentrated benzene extract is then diluted with a roughly
equal volume of hexane when cooL
Bulk matrix (benzene extractables other thm^ CDDs) removal
is accomplished by passing the residue extract solution through
a Super-Macro chromatographic column (see Figure 1) prepared
as follows. The column is thoroughly washed and dried just prior
to use via the g»tn« procedure described for the Soxhlet extracted
A glass wool plug is inserted into the end of the column to se^H
as a bed support, and the following reagents are then carefiu^P
weighed directly into the column: 1.0 g of silica (bottom layer),
2.0 g of 33% 1M sodium hydroxide on silica, 1.0 g of silica, 4.0
g of 44% concentrated sulfuric acid on silica, and 2.0 g of silica
(top layer). The freshly packed column is then immediately
prewashed with 30 mL of hexane and the effluent discarded. The
residue extract is then passed through the column followed by
3 x 5-mL hexane rinses of the boiling flask vessel Following these
rinses an additional 30 mL of hexane is passed through the
column. The total effluent is collected in a 150-mL beaker and
then evaporated to dryness under a stream of Femtogas nitrogen.
A single drop of n-hexadecane (~25 mg) is added to the reagent
blank prior to its evaporation to dryness as a means of improving
internal standard recovery.
Common chemical interferences are removed by passage of the
residue through a dual column system consisting of a top Macro
chromatographic column draining into a bottom High Aspect
column. (See Figure 1.) Each of these columns is cleaned as
previously described and a glass wool bed support inserted just
prior to use. The Macro column is packed with Ld g of 10% silver
nitrate on silica and prewashed with 15 mL of hexane prior to
use. The High Aspect column is packed with 5.0 g of basic
alumina. When the top Macro column prewash has drained, it
is positioned over the High Aspect column reservoir. The sample
residue is dissolved in ~ 15 mL of hexane and introduced into
the top column followed by 3 x 5-mL hexane beaker rinses.
Following the rinses, an additional 30 mT^ of hexane is passed
through the system. When drained, the top column is discarded.
After the hexane has drained to bed level in the High Aspect
column, 50 mL of 50% (v/v) carbon tetrachloride in hexane is
passed through. The total effluent to this point can be discarded
A 25-mL glass vial (cleaned same as chromatographic columM
is used to collect the total effluent after 22.5 mL of 50% (v^
methylene chloride in hexane is introduced into the column.
When elution is complete this fraction which contains chlorinatea
diozins is evaporated to dryness under a stream of Femtogas
nitrogen (1).
RP-HPLC fractionation of the residue is initiated by calibration
of the appropriate collection zones for TCDDs. HCDDs. H-rCDDs.
and OCDD. This is accomolished bv iniectine a calibration
-------
-—-••*-'-
ANALYTICAL CHEMISTRY. VOL. 52. NO. 13. NOVEMBER 1980 « 2047
standard containing approximately 10-20 ng each: 2378-TCDD,
HCDD(s), H-rCDD(s), and OCDD in no more than 30 u.L of
chloroform. In accordance with the chromatogram obtained,
appropriate collection zones are established for each of these
species (see Discussion section). Following calibration, the injector
is rinsed with copious quantities of chloroform, to include multiple
consecutive injections of 50 *iL of chloroform into the column to
ensure that no residual chlorinated dioxins remain.
The residue is prepared for RP-HPLC fractionation by
quantitative transfer to a 0.3-mL Reacti-Vial Quantitative in-
jection requires complete residue solubility in 30 jiL or less of
chloroform. Larger injections of chloroform into this RP-HPLC
system severely reduce column efficiency. An aliquot of no more
than 30 iiL can be fractionated if the sample residue requires
greater amounts of chloroform to be dissolved. Appropriate
chlorinated dioxin fractions are collected in 25-mL volumetric
flasks, equipped with ground glass stoppers, containing ~ 1 mL
of hexane. The chlorinated dioxins are recovered by addition of
2% (w/v) aqueous sodium bicarbonate. The hexane layer is
transferred to a 5-mL gla«« vial and the aqueous phase is extracted
three additional times with — 1 mL of hexane. The combined
extracts are then evaporated to dryness under a stream of Fern-
togas nitrogen. HCDD, H,CDD, and OCDD fractions are
quantitatively transferred to 0.3-mL Reacti-Vials and diluted to
appropriate volumes for determination by GC-LRMS.
Regarding the case for an isomer-specific 2378-TCDD deter-
mination, additional silica-HPLC fractionation of the RP-HPLC
2378-TCDD fraction is required (see Discussion section). Cali-
bration of the appropriate collection zone is accomplished by
injecting approximately 10 ng of 2378-TCDD into the silica-HPLC
in 60-80 fiL of hexane and monitoring the chromatogram obtained.
Adequate isomer specificity is obtained when the silica-HPLC
columns are sufficiently dry so as to provide a 2378-TCDD re-
tention tm"» ranging from a minimum of 12J5 min to maximum
of 17 min (24). Following injection of the residue fraction, the
chromatogram is monitored and the appropriate 2378-TCDD
fraction is collected in a 5-mL glass vial This fraction is then
evaporated to dryness under a stream of Femtogas nitrogen and
diluted to appropriate volume for determination by GC-LRMS.
This procedure can also be used to collect other TCDD isomers
aa described in the Discussion section; see Figure 2.
DISCUSSION
The purpose of this paper is to demonstrate the feasibility
of using a single multiple-step procedure to accomplish the
isomer-specific determination of TCDDs, HCDDa, HtCDDs,
and OCDD at low part per trillion concentrations in a variety
of environmental particulate samples. There were two
prerequisites for our development of the methodology. First,
the sample cleanup must be capable of recovering each of the
listed chlorinated dioxin (CDD) groups from a single sample
and from a single workup. And second, all procedures must
use the least sophisticated and most reliable instrumentation
possible so that such analyses could be conducted in the
greatest number of analytical facilities. These prerequisites
have determined the means by which the described analyses
<-an be accomplished. That is, a neutral or acid extraction
procedure must be used. Any treatment of either the sample
or its extracts with strong bases is known to cause degradation
of the higher chlorinated dioxins (21,37). In accordance with
ease of handling and the general solubility characteristics of
higher chlorinated dioxins (least soluble species), continuous
benzene extraction was found to be adequate for all particulate
samples examined. The selection of packed-column gas
chromatography-low-resolution mass spectrometry as opposed
to capillary column gas chromatography-high-resolution mass
spectrometry represents our attempt to use the least so-
phisticated instrumentation for CDD determination. Because
packed-column GC-LRMS is inherently more subject to
possible interference than Capillary column GC-HRMS, a more
rigorous sample preparation is required. The approach of
combining classical extraction and adsorbent clean-up tech-
niques with consecutive RP-HPLC and silica-HPLC residue
5
[ a ICM i eri t sex H urr SXTP^CT : c :i
O IT I £O_C LASSj >A L_ "
— iTCSO lT»CtLOn»i-
fMCT:CNAT:C.1 £— tMICHCR
HP-ZJ7S TCOQs
- [SO* 1 TCSO»
ID t MATED C00«r—j
M7»*c7:s:iAT:c!i
nn-nmo
1478-TC3O
ap-isoil/siiii
1«7-TCDD
-i «--".' 1
JJ71--C50
RP-127a/5tL»L
UJ8-TCM
1247-TCIO
RP-I378/SXLI2
127B-TCOO
RP-2J7I/SXLI]
1244-TCOO
U24»-TCOOI
im-TCBO
11JJ-TC9D
RF-217B/SILI4
CU4«-TC3DI
1249-TCSO
136I-TC30
1234-TCCO |
1 «-"" ' '
Figure 2. Block diagram for COO sample preparation.
fractionations can be one solution to this problem. Under
these circumstances a significant portion of the method ca-
pabilities to prevent MS interferences during the identification
ajd. quantification of CDDs is relegated to the cleanup rather
than to the final gas chromatographic separation. This can
be advantageous when dealing with highly contaminated
samples because the chromatographic capacity of the clean-up
steps is usually much greater *nn" that of the GC column,
'especially when capillaries are used. In addition, this approach
incorporates the consecutive RP-HPLC and silica-HPLC steps
that we have published for the separation and isolation of the
22 TCDD isomers (34). Their described application in this
procedure permits the analyst to predetermine which possible
TCDD isomers can be present in a given residue fraction.
Hence, the necessity of using a capfllary GC column to obtain
improved TCDD isomer separations is eliminated. This ca-
pability may be of utmost importance as the authors are not
aware of any published date suggesting that all 22 TCDD
isomers can be separated simultaneously using a single ca-
pillary GC column. The described methodology will address
this problem.
It is to be understood that this procedure has been de-
veloped and used for survey purposes on a variety of different
environmental particulates. A complete method validation
including controls, fortifications, and replicates would be
required for each specific matrix before its absolute degree
of reliability can be established. The inclusion of isotopically
enriched TCDD, HCDD, and OCDD internal standards pro-
vide a reasonable degree of reliability under the circumstances
of its described uses.
The samples LO g of NBS urban particulate matter, 1.0 g
of industrial dust, LO g of electrostatically precipitated fly ash
from a municipal burner (fly ash), 16.7 g of Milorganite, and
0.3968 g of European flyash were Soxhlet extracted with
benzene for —16 h and the resulting residues processed
through the preliminary liquid chromatographic clean-up
steps. Each sample, to include a reagent blank, was fortified
with 5-20 ng of isotopically enriched internal standard CDDs
(^ enrichment) prior to analysis. After transfer to a 0.3-mL
Reacti-Vial and evaporation of the solvent, ail samples yielded
a visible white residue. Each of these was then quantitaritvelv
-------
2048 « ANALYTICAL CHEMISTRY, VOL. 52. NO. 13. NOVEMBER 1980
Table I. TCDD Isomer RP-HPLC Fractionation Scheme
and Specific Retention Indices
Table II. HCDDs, H,CDDs, and OCDD Retention Indices
'••'
">. --.:?!
--. .i*
t' "-s
»
-r-
$$'.
•iSife;:
to?
•fc~"
W
TCDD isomer
1269
1469
1267/1239
1268/1279
1369/1478
1246/1249
2378
1236/1239
1278
1237/1238
1247/1248
1378
1379
1368
1234
RP-HPLC
abs RT,°
min
RP-Iso No.
11.6-13.0
11.6-13.0
12.2-12.9
12.2-12.9
13.3-13.9
13.3-13.9
13.3-13.9
13.3-13.9
RP-2378
13.7-14.5
13.7-14.5
13.8-14.5
13.8-14.4
14.4-15.2
14.0-14.7
14.0-15.0
14.0-15.0
14.2-15.1
14.2-15.1
silica-
HP LC C
rel RT6
1 Fraction
1.702
1.497
1.623
1.795
1.238
1.291
1.220
1.340
Fraction
1.323
1.411
1.000
1.356
1.350
1.288
1.100
1.128
1.154
1.199
RP-Iso No. 2 Fraction
14.9-15.7 1.000
14.9-15.9 0.940
1*9-16.8 0.977
15.8-16.8 1.248
GC packed column
rel RTe
0.998
0.912
1.081
1.200
0.956
1.065
0.802
0.907
0.896
0.898
1.006d
1.037
0.969
0.893
0.979
0.990
0.854
0.857
0.858
0.771
0.729
0.960
a RP-HPLC abs RT = absolute retention time (±0.1 min)
to collect peak. * Silica-HPLC rei RT = retention time
relative to 2378-TCDD (±0.010). c GC-packed column
rel RT = retention time relative to "C:2378-TCDD
(*0.005). d Native 2378-TCDD elutes slightly later than
"C-2378-TCDD.
subjected to reverse-phase high-performance liquid chroma-
tography fractionation. The resultant liquid chromatograms
monitored by a UV detector at 235 nm ^X^ for TCDDs)
and 0.02 aufs are shown in Figure 3b-f. Shown in Figure 3a
is the cfaromatogranr obtained for a CDD calibration standard
by RP-HPLC. Although the appropriate CDD collection
zones, denoted by dotted lines, were initially established by
individual injections of 22 TCDD isomers, 10 HCDD isomers,
2 HvCDD isomers, and OCDD, we routinely compute their
location from the observed retention times of only a few se-
lected species. The specific RP-HPLC retention indices for
TCDDs are given in Table I and those for HCDDs, HrCDDs,
and OCDD are listed in Table U.
As indicated, all 22 TCDD isomers can be fractionated from
a sample residue by collecting the column effluent beginning
at — 1L5 and ending at ~ 17.0 min. The initial stage of TCDD
iaomer specificity is achieved by collecting the 22 isomers in
three separate fractions as shown. TCDD Iso No. 1 (RP-
HPLC TCDD iaomer fraction no. 1) can contain the following
isomers: 1269-, 1469-, 1267-, 1289-, 1268-, 1279-, 1369-, and
1478-TCDD. The TCDD 2378 fraction contains 1246-, 1249-,
2378-, 1236-, 1239-, 1278-, 1237-, 1238-, 1247-, and 1248-TCDD.
TCDD Iso No. 2 contains the remaining four isomers: 1378-,
1379-, 1368-, and 1234-TCDD. Preliminary evidence, gained
by fortifying samples with roughly equal amounts of all 22
TCDD isomers at approximately the 150 pptr concentration
level, has indicated that three of the possible isomers in TCDD
Iso No. 1 must be sacrificed in order to ensure quantitative
collection of 2378-TCDD in the following fraction. This
consequence will be discussed later. Its occurrence is related
to the RP-HPLC retention times for the isomers: 1369-TCDD,
1478-TCDD, and one of the pair 1268- or 1279-TCDD having
Sil rel RT 1.238 (normal-phase silica HPLC retention time
CDD isomer
HCDDs
123469-HCDD
123467-HCDD
124679/124689-HCDD
124679/124689-HCDD
123678/123789-HCDD
123679/123 639-HCDD
123679/123689-HCDD
123678/123789-HCDD
123478-HCDD
123468-HCDD
H7CDDs
1234679-H,CDD
1234678-H,CDD
OCDD
RP-
silica- HPLC GC-packe;
HPLC abs column
rel RT° RT6 rel RTC
i
1.081
1.192
0.958
0.972
1.060
0.970
1.039
0.974
0.941
0.890
19.23
19.47
19.62
19.70
20.07
20.20
20.23
20.85
21.02
21.37
0.954
1.077
0.805
0.306
1.103
0.903
0.908
1.016
1.006d
0.861
24.00
24.65
29.40
" Siiica-HPLC rel RT= retention time relative to 2378-
TCDD (±0.010). 6 RP-HPLC abs RT = absolute retention
(±0.1 min) at peak maximum. c GC packed column rel
RT = retention time relative to "C-123478-HCDD.
d Native 123478-HCDD elutes slightly later than "C-
123478-HCDD.
relative to 2378-TCDD). Their retention times are very close
to the fraction boundary separating Iso No. 1 and 2378 and
are split rather irreproducibly between these fractions. Al-
though these isomers do not necessarily interfere with the
quantitation of the isomers expected to the present in the
TCDD 2378 fraction, then1 quantitation essentially becomes
impossible. For cases where quantitation of these
TCDDs is required, a second aliquot of sample residue '
be fractionated by RP-HPLC in such a manner so as to <
pand the Iso No. 1 fraction-to ensure their collection.
The 10 HCDD isomers are collected in accordance with
Figure 3 and Table fl. Although isomer-specific HCDD de-
terminations are possible by using essentially the same
chromatography procedures described for TCDDs (Le., RP-
HPLC — sflica-HPLC — GC), we have not yet applied this
system to samples. Similarly, the two H7CDD isomers are
collected in a single fraction, as is OCDD.
The RP-HPLC residue fractionation chromatograms in
Figure 3 are typical of those associated with particulate sam-
ples. The presence of higher chlorinated species, such as
HrCDDs and OCDD, can often be observed at this point in
the analysis. Although the UV detector has been adjusted
for TTm-rJTtiiim sensitivity for TCDDs, under these conditions
a detectable response for HCDDs, H^ODs, and OCDD is
obtained for approximately 5 ng. Similarly, heptachlorodi-
benzofurans (HfCDFs) and octachlorodibenzofuran (OCDF)
may also be observed in the RP-HPLC fractionation. Because
of the lack of availability of authenticated chlorinated di-
benzofuran (CDFs) standards, we have made no attempt to
quantitate these species. Via collection of appropriate RP-
HPLC fractions, and capillary GC-EC and GC-LRMS, we
have established the possible presence of four H^CDF isomers
and OCDF in a variety of particulate samples.
Refractionation of the RP-HPLC TCDD fractions via
normal-phase HPLC (silica-HPLC) is the final stage of the
sample cleanup prior to GC-LRMS analysis. Normally
monitoring of these chromatograms with a UV detector at 0-J
aufs and 235 nm does not produce observable peaks with t^
exception of the "C-2378-TCDD internal standard. For this
reason example chromatograms are omitted. Table I lists the
individual TCDD isomera contained in each RP-HPLC TCDD
fraction. Included are the RP-HPLC, silica-HPLC, and GC
packed column retention indices for each species. By use of
-------
ANALYTICAL CHEMISTRY. VOL. 52. NO. 13. NOVEMBER 1980 • 2049
0*1,
****
\
!
rcDOi
" ~ ~~ *""""
an '*
"°\\
\ > ,
\ :A:/i
•n
4-
~-/^-
OCOO
^N.
O 1 • • t 10 II 1* I* It JO
3* RP-HPUC fractionation chroinutOQrarns:
European fly ash, (f) NBS urban partculates.
(a) calibration standard, (b) Industrial dust, (c) electrostatic fly ash, (d) municipal sludge, (e)
this information, appropriate fractions can be collected from
the sflica-HPLC which permit isomer-specific GC-LRMS
identification and quantitation.
The sflica-HPLC TCDOs fractionation scheme in Table HI
is designed to provide maximum isomer-apedfic information
when using our packed-column GC-LRMS analysis, while
minimizing the total number of fractions collected. Remem-
bering that the primary goal was to provide the highest quality
analytical data for 2378-TCDD, this scheme is adequate.
Examination of the GC packed column relative retention times
(GC rel RT, TCDD retention time relative to I3C-2378-TCDD)
for all TCDDs present in the RP-2378-TCDD fraction indi-
cates that four other TCDDs have GC rel RTs within ±0.050
(~ 12 s for 4 min absolute retention time for "C-2378-TCDD)
of 2378-TCDD. Arbitrarily defining GC rel RT ±0.050 as the
minimum GC pakced column separation for qualitative
identification of a TCDD isomer from 2378-TCDD and then
direct GC-LRMS analysis of the RP-2378-TCDD fraction
would yield a 2378-TCDD value which could include a max-
imum of four other TCDD isomers (2378-TCDD + 4). How-
ever, examination of the silica HPLC relative retention times
(Sil rel RT, TCDD retention time relative to 2378-TCDD) for
these TCDDs indicates that 2378-TCDD is the first isomer
to elute. The next isomer to elute is 1237/1238-TCDD (Sil
rel RT 1.10); however, even at the minimum acceptable sil-
ica-HPLC retention time for 2378-TCDD which is ~1£5 min,
this isomer is separated by ~ 1.75 min. The remaining nine
TCDD isomera, other than 2373-TCDD, present in the RP-
2378-TCDD fraction can be determined as single isomera with
the exception of those in Sil Fraction No. L Although 1237-,
1238-, 1247-, and 1248-TCDD are essentially baseline sepa-
rated by sflica-HPLC, attempts to collect them in individual
fractions under conditions where the species cannot be ob-
served by a UV detector would be difficult. Hence a single
^ fraction is collected for GC-LRMS analysis. As indicated by *
'the respective GC rel RTs, these isomers can be determined
as a total for 1237- and 1238-TCDD and a total for 1247- and
1248-TCDD.
Three of the TCDD isomers present in RP-Iso No. 1 are
ificed in order to ensure ma-rimnm recovery of 2378-TCDD
in the following RP-HPLC fraction. The consequence of this
situation is the possible presence of 1268/1279-TCDD (Sil rel
RT 1.238), 1369-TCDD, and 1478-TCDD in the RP-2378-
TCDD fraction. Regarding their effect upon the isomer-.
specific determination of 2378-TCDD, it can be observed that
no interference occurs by virtue of both their respective sil-
ica-HPLC rel RTs and their GC-packed column rel RTs.
However, under circumstances where the 1268/1279-TCDD
(Sil rel RT 1.238) isomer is relatively high in concentration,
it could be misidentified as 1237- and 1238-TCDD present
in Sfl Fraction No. 1 of the RP-2378-TCDD fraction. This
interference results from similar GC rel RTs for these isomers
as indicated in Table HL The 1369/1478-TCDD (Sil rel RT
1.220) will not cause any similar interference problems with
those TCDDs present in RP-2378-TCDD fraction—Sil
Fraction No. 1 because of its GC rel RT of 0.802. The re-
maining isomer, 1369/1478-TCDD (SU rel RT 1.340), if
present in high concentration may interfere with 1246/
1249-TCDD (Sil rel RT 1.411) in RP-2378-TCDD fraction-Sil
Fraction No. 3.
-------
*• •:-
• -.1
i
•ST
i>
**"*
r^«-
*«"*r.
f* 7
2050 • ANALYTICAL CHEMISTRY. VOL. 52. NO. 13. NOVEMBER 1980
(o)
(mM 3
Figure 4. Isomer-specific 2378-TCDO GC-LRMS mass cftromatograrns: (a) caffljratton standard, (b) reagent blank, (c) industrial dust, (d) electrostatic
fly ash, (e) municipal sludge, (f) European fly ash.
GC-LRMS mass chromatograms for the isomer-specific
2378-TCDD fractions of each paniculate sample analyzed are
shown in Figure 4. Native 2378-TCDD is monitored at m/e
320,322, and 324 and "C-2378-TCDD at 332. The calibration
standard (Figure 4a shown is typical for a 2-nL injection of
a reference standard containing 100 pg/^L of native 2378- .
TCDD and 500 pg//*L of l3C-2378-TCDD.
The GC-LRMS mass chromatograms in Figure 5 compare
the analysis of the RP-2378-TCDD fraction from electro-
statically precipitated fly ash for 2378-TCDD, before and after
silica-HPLC refractionation. As a means of ensuring homo-
geneity, a 2-g portion of sample was processed through the
cleanup including RP-HPLC fractionation. At this point the
RP-2378-TCDD fraction was divided into two equal portions,
each equivalent to 1 g of original sample. One portion was
analyzed directly by GC-LRMS as illustrated in Figure 5a.
The other portion was further fractionated by silica-HPLC,
the Sil Fraction 2378 collected, and this residue analyzed by
GC-LRMS (Figure 5b). Comparison of 2378-TCDD quanti-
tation for these residues yields 1500 pptr before silica-HPLC
refractionation, and 430 pptr after. The value obtained before
silica-HPLC refractionation must be qualified as being the
concentration of 2378-HPLC plus four possible unseparated
isomers (see Table IV).
Isomer-specific TCDD analysis data for each of the de-
scribed particulate samples appear in Tables IV and V.
Quantitation of TCDDs was accomplished by averaging the
observed response at m/e 320,322, and 324 for all cases except
where denoted. Instrumental calibration for all TCDD isomers
was based upon the observed responses for a primary standard
of 2378-TCDD. The listed concentrations for 2378-TCDD
have been corrected for recovery of the "C-2378-TCDD in-
ternal standard as given in Table V. Concentrations given
for all other TCDD isomers represent absolute observed
values. The limit of detection (LoD) for all species was defined
as 2.5 X peak-to-valley noise in a region nearby the expected
elution tune. Observed concentrations less than the LoD are
listed as not detected (ND).
525
FS-41
"C • 2378 • TCOD
m/i* 333
FS-18
intM TCDOt
m/«" 324
FS-28
/e' 322
m/e' 320"
FS-28
;'C-2378-TCI
ml*' 332
m/t- 324
FS-8
ntlt 372
FS-3
m/e' 320
(mini 3
(mini 3
Hgure 5. Comparative 2378-TCDD GC-LRMS mass chromati
(or electrostatic fly ash (a) after RP-HPLC (RP-2378 fraction) (b) al
silica-HPLC (silica-2378 fraction).
As a means of investigating the degree of reliability asso-
ciated with the isomer-specific determination of 237S-TCDD
in a sample containing equivalent concentrations of all 21 other
-------
ANALYTICAL CHEMISTRY, VOL. 52. NO. 13. NOVEMBER 1980 « 2051
Table III. TCDD Isomer Silica-HPLC Fractionation
Scheme and Specific Retention Indices
GC
silica- packed
HPLC Sil collection column
TCDD isomer rel RT" zone rel RT° rel RT*
RP-Isol No. 1 Fraction TCDDs
Sil fraction no. 1
1268/1279-TCDD
1369/1478-TCDD
Sil fraction no. 2
1269-TCDD
1469-TCDD
1267/12S9-TCDD
RP-2378
Sil fraction 2378
2378-TCDD
Sil fraction no. 1
1237/1238-TCDD*
1247/1248-TCDD*
Sil fraction no. 2
1278-TCDD
Sil fraction no. 3
1246/1249-TCDD
1236/1239-TCDD
Sil fraction no. 4
1246/1249-TCDD
1.455-1.850
1.180-1.370
1.238C
1.291
1.220C
1.340C
1.702
1.497
1.623
1.795
Fraction TCDDs
0.950-1.050
1.000
1.050-1.244
1.100
1.128
1.154
1.199
, 1.244-1.300
1.288
1.300-1.385
1.328
1.356
1.350
1.385-1.450
1.411
RP-Iso No. 2 Fraction TCDDs
Sil fraction no. lf 0.900-1.050
1368-TCDD 0.940
1379-TCDD 0.977
1378-TCDD 1.000
Sil fraction no. 2f 1.210-1.288
1234-TCDD 1.248
0.956
1.065
0.802
0.907
0.998
0.912
1.081
1.200
1.006d
0.979
0.990
0.854
0.857
0.893
0.896
1.037
0.969
0.898
0.729
0.771
0.858
0.960
" Silica-HPLC rel RT = retention time relative to 2378-
TCDD (±0.010). * GC packed column rel RT= retention
time relative to "C-237S-TCDD (*0.005). « See text for
recovery information. d Native 2378-TCDD elutes
slightly later than "C-2378-TCDD. * Related isomers
typically reported as a total. f Fractions typically com-
bined prior to GC-LRMS analysis.
TCDD isomers, we intentionally fortified a second portion of
muncipal sludge with each TCDD isomer at the levels shown
in Table VI Neither 1237- or 1238-TCDD was added due
to their natural presence at 230 pptr (see Table V). Analysis
of the fortified sample yielded the recovery data shown in
Table VL Regarding the 2378-TCDD data, the amount found
was corrected for the recovery of the "C-2378-TCDD and also
for the 20 pptr natural 2378-TCDD previously observed in
ft • ITS oat «SO
F5-MS m/« 4M
FS-131 TO. 424
"'COOi
FS - TOt mlf 330
I m/«"
X.
10 It 12
Figure 6. Higher chlorinated dloxin GC-LRMS mass chromatograms
for electrostatic fly ash.
the sample. These data indicate that no other TCDD isomer
interferes with the determination of 2378-TCDD. when this
analytical procedure is used. Recovery values given for all
other TCDD isomers represent absolute observed values and
were corrected for natural levels when necessary as listed in
Table VL
Typical temperature programmed GC-LRMS mass chro-
matograms for the determination of higher chlorinated dioxins
appear in Figure 6. For the analysis of electrostatically
, precipitated fly ash the RP-HPLC HCDDs, HTCDDs, and
OCDD fractions were combined prior to GC-LRMS exami-
nation (see Figure 3c). As a means of overcoming problems
associated with samples having relatively large amounts of
native chlorinated diorins compared to the 1-20 ng of fortified
internal standards, a complete method validation study was
conducted for HCDDs, HTCDDs, and dCDD ranging from 50
pptr to 10 ppm (jig/g) and from 10 pptr to 5 ppb for 2378-
TCDD. The control particulate sample used was a sandy loam
sofl, to which was added ~150 fiL of Mobile 1 synthetic engine
lubricant per 20 g, as a means of increasing the total organics
content to better simulate typical particulates. The following
native CDD standards were used for sample fortification:
2378-TCDD, 123678-HCDD, 123679/123689-HCDD (Sil rel
RT 1.039), 1234678-HTCDD, and OCDD. The results of this
Table IV. Chlorinated Dioxina Observed in Environmental Particulate Samples
parts per billion
reagent
CDDs blank, ng
2378-TCDD + 4 isomers" ND (0.06)
other TCDDs (17 isomers) ND (0.04)
HCDDsc (10 isomers) ND (0.18)
1234679-H.CDDC ND (0.14)
1234678-H,CDDe ND (0.14)
OCDD* ND(0.29)
" RP-HPLC RP-2378 fraction analyzed directly by GC-LRMS and not isomer specific as described in text. b Sample fully
fractionated for isomer-specific results given in Table V. c Observed values without correction run as part of validation
work reported in Table VII. d For "semi" isomer specific see Table VIII. e "C-2378-TCDD recovery 78% and value listed
has been corrected, see Table V for others, and ND = compound not detected at limit of detection in parentheses and no
parentheses indicates detected signal > 10X limit of detection.
industrial
dust
. . .6
. . .»
ND(14)
200
220
4000
electrostatic
flyash
1.5*
. . .ft
14
11
17
30
municipal
sludge
. . .6
. . .»
2.1
14
15
180
European
flyash
. . .6
. . .&
550d
470
570
650
NBS urban
particulates
0.12(0.12)'
0.16(0.08)
2(2)
16
• 18
210
r or
U .^u
-------
20S2 • ANALYTICAL CHEMISTRY. VOL 52. NO. 13. NOVEMBER 1980
Table V. Isotner-Specii'ic TCDD Analyses of Environmental Particulate Samples
i I
* .1
•
•-11
Sit--*
parts per trillion
TCDD isomer
2373-TCDD"
1269-TCDD
1469-TCDD
1267/12S9-TCDD Sil rel RT 1.623
1267/12S9-TCDD Sil rel RT 1.795
1268/1279-TCDD Sil rel RT 1.233
1268/1279-TCDD Sil rel RT 1.291
1369/1478-TCDD Sil rel RT 1.220
1369/1478-TCDD Sil rel RT 1.340
1273-TCDD
1236/1239-TCDD Sil rel RT 1.356
1236/1239-TCDD Sil rel RT 1.350
1237/123S-TCDD Sil rel RT 1.100
1237/123S-TCDD Sil rel RT 1.128
1246/1249-TCDD Sil rel RT 1.323
1246/1249-TCDD Sil rel RT 1.411
1247/1248-TCDD Sil rel RT 1.154
1247/1248-TCDD Sil rel RT 1.199
1378-TCDD
1379-TCDD
1368-TCDD
1234-TCDD
total TCDDs
"C-2378-TCDD %"recovery
reagent
blank, pg
ND (40)
ND(20)
ND (20)
ND (20)
ND (20)
ND (30)
ND(30)
ND(60)
ND (60)
ND (60)
[ ND (60)
}ND(60)
| ND (60)
ND(20)
ND (20)
ND(20)
ND (20)
ND
63%
industrial
dust
1100
ND(40)
ND (50)
ND (50)
ND(50)
_ . .6
ND(50)
electrostatic
flyash
430 (110)
190(60)
ND (50)
100 (60)
120(60)
190 (90)c
310(90)
municipal
sludge
20(2)
ND(2)
ND(2)
ND (2)
ND(2)
.. . .
3(3)d
European
flyash
2300
1000 (140)
250(140)
300 (140)
500 (140)
1000C
1500
ND (40)
ND(60)
ND(60)
240 (50)*
ND(60)e
140(50)
560(110)
1340
2780
180
6340
59%
ND (80)
2SO(110)
150(110)
720*
730(110)e
310(70)
1370(150)
1160(150)
1320 (150)
370 (150)
7750
54%
ND(3)
ND(3)
ND(3)
230e
ND (3)e
8(2)
23(5)
13(5)
13(5)
ND (30)
310
61%
3100
1500
800 (400)
8500*
2000*
1500
6900
13200
7000
16200
2100
69800
56%
0 Corrected for "C-2378-TCDD recovery and all other isomers are absolute observed. * • • • = not recovered as described
in text. c Observed but recovery questionable. * Detected on in/e 322 only. * Possible isomer interference as described
in text.
Table VI. Isomer-Specific TCDD Analysis of Municipal Sludge after Fortification
concn in pptr
TCDD isomer
2378-TCDD
1269-TCDD
1469-TCDD
126-f£L289-TCDD Sil rel RT 1.623
1267/1289-TCDD Sil rel RT 1.795
1268/1279-TCDD Sil rel RT 1.238
1268/1279-TCDD Sil rel RT 1.291
1369/1478-TCDD Sil rel RT 1.220
1369/1478-TCDD Sil rel RT 1.340
1278-TCDD
1236/1239-TCDD Sil rel RT 1.356
1236/1239-TCDD Sil rel RT 1.350
1237/1238-TCDD Sil rel RT 1.100
1237/1238-TCDD Sil rel RT 1.128
1246/1249-TCDD Sil rel RT 1.328
1246/1249-TCDD Sil rel RT 1.411
1247/1248-TCDD Sil rel RT 1.154
1247/1248-TCDD Sil rel RT 1.199
1378-TCDD
1379-TCDD
1368-TCDD
1234-TCDD
added
143
150
166
150
171
137
140
143
151
160
147
146
141
151
131
163
171
171
101
143
found
140
108
122
126
145
. . .*
69
104
103
80
(180)d
}220e
[203*
151
138
45
122
% recovery
98°
72
73
84
85
49
65
70
55
75
69
88
81
45
85
0 Corrected for recovery of "C-2373-TCDD (72%) and native 2378-TCDD present given in Table V, all other isomers are
absolute observed. 6 • • • = not recovered as described in text. c Total not added. High native concentration given in
Table V. d Absolute amount observed in this sample. * Total.
study appear in Table VTJ. These data indicate that the
average recoveries of HCDDs, H7CDDS, and OCDD over the
described concentrations range are reasonably constant and
are between 70 and 80%. Because typical paniculate samples
contain higher chlorinated CDDs within this range, recovery
factors derived from the validation can be used. Since 13C-
:vj«
1
labeled internal standards are added to all samples, whenev^
very low native concentrations are observed appropri
correction factors can be applied. Note that recovery valu<
reported for TCDD have been corrected for the observed
l3C-2378-TCDD internal standard recoveries after RP-HPLC
fractionation.
-------
ANALYTICAL CHEMISTRY, VOL, 52. NO. 13. NOVEMBER 1980 • 2053
Table VII. Chlorinated Dioxin Recovery and Precision Data for Fortified Sandy Loam Soil"
2373-TCDD* HCDD H.CDD
sample added, found,
no. pptr pptr
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
xallc
o_ allc
x prec0
a prec'
10
20
20
50
50
50
50
50
50
50
50
100
100
5000
5000
50
13
23
21
49
45
51
53
50
50
47
52
97
109
5350
5400
49.6
2.6
130
140
105
98
90
102
106
100
100
94
104
97
109
107
108
106
13
99.2
5.2
added.
pptr
50
100
100
250
250
250
250
250
250
250
250
500
500
1 x 10*
1 X 10*
250
found.
pptr
30
72
57
160
180
170
170
190
160
ISO
170
410
440
8.1 X 10s
9.1 X 10'
173
10.4
%
60
72
57
64
72
68
68
76
64
72
68
82
88
87
91
73
10
69
6.0
added.
pptr
50
100
100
250
250
250
250
250
250
250
250
500
500
5 x 10*
5 x 10«
250
found,
pptr
46
75
65
170
200
200
170
210
160
180
160
430
460
4.5 X 10°
4.7 X 10*
181
19.6
%
92
75
65
68
80
80
68
84
64
72
64
86
92
90
94
78
11
72
10.8
OCDD
added,
pptr
200
400
400
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
10 X 10"
10 X 10"
1000
found,
pptr
160
330
260
730
820
7SO
720
880
700
690
690
1900
2060
8.4 x 10"
9.0 x 10'
751
69.4
%
80
83
65
73
82
73
72
88
70
69
69
95
103
84
90
80
11
75
9.2
0 Data for all species obtained by GC-LRMS analysis of appropriate RP-HPLC fractions. 2378-TCDD values corrected for
"C-2378-TCDD internal standard recovery, other CDOs are absolute observed. fc Corrected for 1JC-2378-TCDD where _
average recovery was 59.8% for all samples. c x all and a all represent the mean and standard deviation of all samples. d x
prec and a prec represent the mean and standard deviation of samples 4-11 to determine precision of the analysis.
Table VTIL "Semi" Isomer-Specific HCDD Analysis Data for European Flyash, Absolute Values Reported
parts per billion
HCDD isomer"
124679/124689-HCDD Sil rel HT 0.958
124679/124689-HCDD Sil rel RT 0.972
123468-HCDD
123679/123689-HCDD Sil rel RT 0.970
123679/123689-HCDD Sil rel RT 1.039
123469-HCDD
123478-HCDD
123678/123789-HCDD Sil rel RT 0.974
123678/123789-HCDD Sil rel RT 1.060
123467-HCDD
reagent blank
}ND(0.13)6-C
ND (0.13)
\ND(0.13)C
}ND(0.13)C
}ND(0.13)C
" HCDD Sil rel RT = retention time relative to 2378-TCDD by silica-HPLC (Table II).
limit of detection in ppb based on flyash sample size. c Total.
European flyash
82C
9(9)
. 260C
110e
85(9)c
ND (0.13) is not detected with
GC-LRMS analysis data for higher chlorinated CDDs ap-
pear in Tables IV and VTU. Table VUI illustrates a format
for HCDD determination that is "semi"-isomer specific. In
this case, the total RP-HCDDs fraction was analyzed directly
by packed-column GC-LRMS. However, because GC rel RTs
have been experimentally determined (see Table II) for all
10 individual HCDD isomers, we can separate the HCDDs
observed into five distinct groups. Within each group only
a limited number of isomers are possible. These analyses are
accomplished by using isothermal column condition (~270
°C) so as to ma-rimi^g the separation power of the column and
to improve relative retention time measurements.
CONCLUSIONS
Although this paper demonstrates the applicability of a
multiple-step procedure to isomer-specifically determine a
variety of CDDs in environmental particuiate samples, we have
also applied the technique to many other matrices successfully.
Simple modification of the preliminary matrix extraction has
permitted the analysis of tissues, human milk, vegetable
matter, chemical products, and wastes without sacrificing high
sensitivity or isomer specificity. This procedure, utilizing
packed-column GC-LRMS, has provided reliable results for
several heavily contaminated matrices where the combination
of a less sophisticated cleanup followed by both packed and
capillary column GC-HR MS has failed. Interested individuals
may request a more thorough discussion of the method de-
velopment experiments from the authors.
ACKNOWLEDGMENT
The authors express their gratitude to O. Hutzinger for
graciously suppling the European fly ash sample and to R.
Bumb, W. Crummett, and V. Stenger for their help in re-
pairing this manuscript.
LITERATURE CITED
(1) Lamparskl, l_ !_; Nestrick. T. J.; Stert. H. H. Anal. Cham. 1979. St.
1453.
(2) Baugftman. R W.; Mesetson. M: 5. EHP. Environ. Health Parspact.
1973, S. 27.
(3) Widmark. G. Adv. Ctiem. Ser. 1971. No. 104. 1.
(4) Donaldson. W. T. Environ. 3d. Tgctinol. 1977. 11. 348.
(5) Lavi. I.; Nowicki. T. W. Bull. Environ. Contain. Toxicol. 1972, 7, 193.
(6) Scfiwaitz. 0. P. J. Chromatogr. 1978. 752. 514.
(7) RouKe. 0. R: Mueiter. W. ?A Yang, R S. H. J. Assoc. Off. Anal. Cham.
1977. 60. 233.
(8) Oennay. 0. W.: Karasek. F. W. J. Chromatogr. 1978. 151. 75.
(9) ASTM Method 11104-04-T3T, Health Lab. So. 1974, 11. 218.
(10) Giam. C. S.: Chan, H. S.; Nert. G. S. Anal. Cham. 1975. 47. 2225.
(11) Singmaster. J. A.; CrosSy. 0. G. Bull. Environ. Comam. Toxicol. 1976.
76. 291.
(12) Glam. C. S.: Wong, M. K. J. Chromatogr. 1972. 72. 283.
(13) Millar. J. M.: Kirchner. J. G. Anal. Cham. 1952. 24. 1480.
528
-------
Anal. Chem. 1980. 52. 2054-2057
(14) Stanley, W. L: Vanmer. S. H.; Gemilli, a. J. Assoc. Oil. Anal. Cham.
1957. 40. 282.
(15) Bowyer. C. E.: Usat W. M. ¥.; Howard. A. N.: Gresnam. G. A. BJocnam.
J. 1963. MS. 24P.
(16) Brown. T. 1_; Benjamin. J. /»na/. Cham. 1964. 36. 446.
(17) Amos. ft. J. Chromatogr. 1970. •«. 343.
(18) Bevenuo. A.: Keftoy. T. W.: Hylin. J. W. J. Chromatogr. 1971. 54. 71.
(19) Hutzingor. O.; Safo. S.: Zhko. V. "The Chemistry o( PC3s": Chemical
Ruober Co. Cross: Cleveland. OH. 1974.
(20) foots. C. F.: wajberty. 0. G. J. Chromatogr. 1977. /32. 511.
(21) Lamoarskf. l_ 1_: MaWe. N. H.; Shaaoff. l_ A. J. Agric. Food. Chem.
1978. 25. 1113.
(22) Pe«H. 0. A. Mature (London} 197S. ^54. 324.
(23) Bceman. G. A.; Clement. R £.: KasaseK. F. W. Anal. Cham. 1979. .57,
2343.
(24) CSemetn. R E; Sceman. G, A..- Karasek. F. W.; Sowars. W. O.; Parsons.
M. l_ J. Chromatogr. 1980. 189. S3.
(25) Matsumura. P.; Wart. C. T. Project No. OWRT A-058-W1S: Wisconsin
University: Maaison. WI. 1978.
(26) Buser. H. R: SossnaroX H. P. J. Chromatogr. 1974, SO. 71.
(27) Buser. H. R; Bossharot. H. P. J. Assoc. Off. Anal. Cham. 1978. •
562.
(28) Raooe. C.; Markiund. S.: Buser. H. R.; Bossnarot. H. P. Chamospne
1978. 3. 269.
(29) Busor. M. H. Anal. Cham. 1977. 43. 918.
(30) Buser. H. R. Chemosanere 1979. a. 251. j
(31) Tha Oow Chemical Co.. Trace Chemistries o« Ftre. 1978. '
(32) Blair. E. H.. Ed. Adv. Chem. Ser. 1973. No. 120.
(33) Boer. F. P.; Van Remoonaro. F.: Moewaf. W. W. J. ,4m. Cham Soc
1972. 34. 1006.
(34) Nestrick, T. J.; Lamparsfci. l_ l_: Start), R H. Anal. Cham. 1979. 5.
2273.
(35) Aniline. O. /4oV. C/wn. Ser. 1973. Mo. 720. 126.
(36) Breoeweg. R. A.; flotnman. L. O.; Prettier. C. 0. Anal. Chem. 1979. 5:
2081.
(37) Firestone. 0. J. Assoc. Off. Anal. Chem. 1977, SO, 354.
RECEIVED for review January 28,1980. Resubmitted May 14.
1980. Accepted July 31, 1980.
Secondary Ion Mass Spectra of Diquaternary Ammonium Salts
Timothy M. Ryan, Robert J. Day, and R. Graham Cooks*
Department of Chemistry. Purdue University, West Lafayette, Indiana 47907
Molecular dications emitted by momentum transfer processes
are observed In secondary (on mass spectra (SIMS) of dl-
quatemary ammonium salts. The relationship between mo-
lecular structure and the observation of dlcations Is explored.
Large (ntercharge separations, corresponding to lessened In-
tramolecular couiombfc repulsions, are observed to correlate
with dication detection. Fragmentation with charge separation
is facilitated by small (ntercharge distances and can preclude
observation of the dlcauon. Electron attachment to yield the
monocatlon is an alternative to dlcatlon emission when the
structure of the dicatton facilitates reduction. This occurs, for
example, for the herbicide dlquat (M^-etnyiene-2^-ofpyrldyl
dibromide) which Is detected as Its monocatlon. Complete
spectra of dlquaternaries can be taken with nanogram size
samples.
Secondary ion mass spectrometry (SIMS) has recently been
shown to be a sensitive method for the characterization of
organic salts (1-4). Reported here is the observation of intact
organic dications emitted from diquatemary ammonium saits
upon sputtering. This constitutes the first observation of
multiply charged organic molecular ions in SIMS. The result
is of interest with regard to both analytical applications of
SIMS and the fundamentals of ionization during sputtering.
Specifically, some biologically important compounds, such as
the herbicides paraquat and diquat and the curare alkaloids,
have the diquatemary structure, so that SIMS may facilitate
their characterization. In addition, studies on organic dications
reflect the degree to which electron attachment occurs during
sputtering. This process yields observable charged products
for dications, but neutrals are sputtered when monocations
are reduced during ion bombardment.
EXPERIMENTAL SECTION
All compounds were synthesized by using standard methods
for the preparation of quaternary ammonium salts. The organic
salts were burnished onto a 1 cm2 roughened foil of either silver
or platinum prior to SIMS analysis using argon primary ions at
5 keV and 0.3-0.5 nA primary ion current. Beam diameter was
approximately 1 mm and pressures in the ultra-high-vacuum
chamber remained below 1 X 10"* torr during the course of j
experiments.
All spectra were taken with Riber SIMS system usir'
quadrupole mass analyzer, Channeltron electron multiplier, and
pulse-counting electronics.
Ihtercharge distances were measured by using Dreiding models;
charge localization on nitrogen was assumed and the ma-rimnm
distance in the unstrained molecule is reported. Intercharge
distances (r) were used to calculate coulombic repulsive energies
(T) from T (eV) =» 14.6/r (A).
RESULTS AND DISCUSSION
The SIMS spectrum of WV/-bis(dimethyl)-4,4'-tri-.
methylenedipiperidine diiodide (1) is shown in Figure L This
spectrum provides both the molecular weight (inferred from
the highest mass doubly charged ion, 26S2'*') and structural
information on the compound. Emission of the doubly
charged species is confirmed by the observation of the 13C
isotope peak one-half mass unit above the dication peak (m/z
134.5 in Figure 2). Changing the counterion does not affect
the SIMS spectrum; for example, the dibromide and diiodide
of compound 1 gave identical SIMS spectra.
Analogous results were obtained for AfJV'-bis(ethyl-
methyl)-4,4'-trimethylenedipiperidine diiodide (2) and for the
aromatic compounds ^V^v"'-bis(trimethyl)-4,4'-methylenedi-
aniline diiodide (3) and ^^V'-bis(dimethylethyi)-4,4'-
methyienedianiiine diiodide (4). The spectrum of compound
3 is shown in Figure 3; the dication, 2842* at m/z 142 ia of
relatively low abundance, but its 13C isotope is well resolved
in high-resolution scans.
A considerable number of diquatemary salts (5-19, Table
I) did not exhibit observable dications. Compounds 18 an5
19, while they did not yield molecular dications, did show i
corresponding singly charged ions in their SIMS spectral
Compounds 5-17 may fail to exhibit dications because they
fragment by a favorable charge separation route, M'2* — MI*
+ Mo*. This is indicated by the absence of both singly and
doubly charged molecular ions for these samples.
0003-2700/80/0352-2054301.00/0 © 1980 American Chemical Society
-------
Chemosphere, Vol.19, Nos.1-6, pp 27-31, 1989
Printed in Great Britain
0045-6535/89 $3.0O +
Pergamon Press pic
.OO
asw
NOVEL EXTRACTION DEVICE FOR THE DETERMINATION OF CHLORINATED
DIBENZO-P-DIOXINS (CDDs) AND DIBENZOFURANS (CDFs)
IN MATRICES CONTAINING WATER
L. L, Lamparski and T. J. Nestrick
The Dow Chemical Company
Michigan Applied Science & Technology Laboratories
Analytical Sciences, 1602 Building
Midland, Michigan 48674 USA
ABSTRACT
The efficient extraction of CDDs I CDFs from paniculate matrices containing measurable amounts of water is compli-
cated by the need to remove the water from the particles before beginning the extraction process. Traditionally, this has
been done by pre-extracting the sample with a water miscible solvent such as isopropanol or by air drying for an ex-
tended period of time over various desiccants such as sulfuric acid or phosphorus pentoxide. These techniques cau suffer
from three possible problems: potential loss ofanalytes, accidental sample contamination, or an unacceptably long
sample drying time. In response to these drawbacks, we have developed an extraction device that combines the simulta-
neous removal and measurement of water from a particulate sample with an exhaustive Soxhlet extraction of the dried
particulars.
KEYWORDS
Chlorinated dibenzo-p-dioxin and dibenzofuran; Soxhlet-Dean-Stark extractor (SDS); wet particulate matrices; water
determination in solids; drying procedure; aromatic solvents; Soxhlet extractor; Dean-Stark receiver.
INTRODUCTION
In the field of trace-level determination of chlorinatedjdibenzo-p-dioxins and dibenzofurans (CDDs/CDFs), it is fre-
quently necessary to measure the analytes in solid matrices of environmental, biological, or industrial origin. When
the solid matrix cannot be dissolved in an appropriate solvent, optimum extraction of the CDD/CDF analytes is often
achieved by subjecting the sample to an exhaustive Soxhlet extraction using an aromatic solvent such as benzene or
toluene (1,2). While this approach is known to be reliable for relatively dry solid matrices, the often encountered
presence of a significant quantity of water in a sample can render typical benzene Soxhlet extraction procedures
ineffective for CDDs/CDFs removal. Because high and variable levels of water in samples can distort the apparent
analyte concentration, it is often desirable to report quantitative findings based upon the dry weight of the sample
matrix. An obvious solution to this problem is the elevated temperature drying of an aliquot of wet sample to provide a
reasonably quick determination of water content-by weight-loss measurement. This dried sample can then be sub-
jected to Soxhlet extraction to remove the CDD/CDF analytes. Unfortunately, this procedure can yield unreliable
results due to loss of CDDs/CDFs during the elevated temperature drying. This phenomenon (codistillation of CDDs
with water) has been described previously for the steam-stripping of CDDs from particulates (3).
The currently accepted method for determining CDDs/CDFs in solid sample matrices containing appreciable amounts
of water (e.g., sediments) requires dividing the sample into two portions of assumed identical composition. One portion
is then subjected to elevated temperature drying (typically 12 to 24 hours at 105°C) to provide apparent water content
by weight loss measurement The second fraction is subjected to ambient temperature desiccation (typically requiring
in excess of 24 hours for relatively small aliquots of 1 to 10 g), followed by particle size reduction and homogenization,
and finally Soxhlet extraction of the dried material for 12 to 24 hours with an aromatic solvent to remove and concen-
trate the CDDs/CDFs (4). Alternative methods of water removal include: lyophilization (5), batch extraction with
hexane/acetone (6), mixing with anhydrous sodium sulfate followed by column extraction (7), or Soxhlet extraction (8),
or Soxhlet extraction with a water miscible solvent followed by Soxhlet extraction with methylene chloride (9,10).
1
27
i > u
-------
28
While these previously described procedures for determining CDDs/CDFs in wet solids can provide reliable quantita-
tive results, they may suffer from the following problems. (A) Water and CDDs/CDFs content are determined on
separate aliquots of the original solid matrix which requires the assumption of sample homogeneity. (B) Desiccation
procedures for the aliquot to be subjected to Soxhlet extraction typically require extended time periods even for rela-
tively small sample sizes. (C) Desiccation procedures can lead to loss of CDD/CDF analytes. (D) Desiccation proce-
dures can lead to inadvertent sample contamination. (E) Certain extraction solvents which are suitable for removing
water from the sample matrix may not efficiently extract CDDs/CDFs from some types of particulates.
In response to these problems, we have combined two classical analytical techniques, Soxhlet extraction (11) and Dean-
Stark azeotropic distillation (12). into a single process which effectively removes both water and CDDs/CDFs from wet
particulate samples. In doing so, this Soxhlet-Dean-Stark (SDS) extraction apparatus permits the determination of
both water and CDDs/CDFs on the same sample aliquot, it eliminates the separate time period required for sample
desiccation prior to Soxhlet extraction, it eliminates the possible loss of analytes during desiccation, and it substan-
tially reduces the potential for sample contamination during the separate desiccation step prior to extraction.
In the course of submitting a patent application for the SDS extraction device we discovered a similar apparatus for
use on biological matrices described by H. E. Wistreich in US Patent 3,170,767, dated 23FEB65. Although combining
azeotropic distillation and collection of water with solvent extraction of the sample, this unit does not employ Soxhlet
conditions of matrix extraction and is therefore not expected to be as efficient an the described SDS design for inorganic
matrices.
EXPERIMENTAL
Apparatus
The SDS apparatus for the extraction of wet, solid, sample matrices is shown in Figure 1. The extractor consists of
three basic parts: the Soxhlet extractor body, including extraction thimble and reflux flask; the Dean-Stark azeotropic
distillation receiver; and the water-cooled condenser. The Soxhlet extractor (Ace Glass, Vineland, NJ 08360; Cat. No.
6730-10) is designed to hold a 43 x 125 mm glass thimble with a porosity-A glass frit (Ace Glass; Cat No. 6813-22) and
is equipped with a 250-mL boiling flask. The Dean-Stark receiver was constructed in the Dow Glass Fabrication
Laboratory with 55/50 standard-taper fittings as shown; and the
water receiver was made to hold -35 mL of water and, if necessary,
could be graduated to directly measure water volume collected.
The extractor condenser is also of standard design (Ace Glass; Cat.
No. 6740-10) for use with the Soxhlet apparatus. Heat is applied
to the reflux flask by a standard heating mantle and variable
transformer.
Extraction Procedure
Typical operation of the SDS extractor in our laboratory consists of
the following procedure. The individual components of the extrac-
tor are thoroughly cleaned and dried prior to use. The extraction
thimble is charged with 5.0 g of 100/200-mesh silica (previously
cleaned and activated as described in reference 13) and 100 g of
50/70-mesh white, quartz sand (available from Aldrieh Chemical
Company, Milwaukee, WI 53233; Cat. No. 27,473-9; used as
received). Care should be taken to avoid disrupting the layers of
the two adsorbents. The thimble is then placed in the extractor
body; 230 mL of benzene is placed into the reflux flask and 35 mL
into the Dean-Stark receiver; and the apparatus is assembled as
shown in Figure 1. Power is applied to the heating mantle, and
the entire system is pre-extracted for approximately four hours.
After the pre-extraction period, the apparatus cooled and dis-
mantled; the solvent is discarded; and the receiver, extractor
body, and reflux flask are rinsed with methylene chloride and
allowed to air dry. Excess benzene is removed from the extraction
thimble, sand, and silica gel by means of a slight vacuum applied
to the bottom of the thimble through a small teflon-covered rubber
cone. At this point, the wet sample (up to -35 g) is loaded into the
thimble and manually mixed into the sand layer with a clean
metal spatula (taking care not to disrupt the silica layer below the
sand) until a reasonably uniform, porous mixture is obtained. This
mixture is then fortified with [1SC,,]-CDD/CDF internal standards.
Figure 1. Typical operating configuration for
Soihlet-Dean-Stark (SDS) extractor apparatus.
531
-------
29
The extractor apparatus is reassembled with a fresh charge of benzene loaded into the reflux flask and Dean-Stark
receiver. Extraction of the sample is commenced by applying power to the heating mantle. (Note: If the sample has a
large amount of water, benzene flow through the sand bed may be somewhat restricted for the first few extraction
cycles. To prevent this, it is useful to have made a small opening in the bed for the solvent to flow through with rela-
tively little restriction. As the sample bed dries, this opening will gradually collapse allowing efficient extraction of the
analytes. It is also helpful to slow the solvent reflux rate to match the rate of percolation through the sand and silica
beds until water removal lessens the restriction to benzene flow, otherwise thimble overflow may occur.)
Water removal from the bed usually begins with the first pass of hot benzene through the bed. In this process, water
and extractable components removed from the wet solid sample are transported to the reflux flask along with the
warm benzene. Because water and benzene distill as an azeotrope (containing -4% water and 96% benzene) at a lower
temperature than pure benzene, the water that has been extracted from the sample and returned to the reflux flask,
will be transported overhead where it will condense and separate from the benzene and be collected as a lower phase in
the Dean-Stark receiver. The relatively dry benzene drains back into the extractor body to repeat the extraction
process. Because CDDs/CDFs do not azeotropically distill with benzene, they will remain and be concentrated in the
reflux flask. Water removal is usually complete within eight hours and for an additional eight hours, the extraction of
the dried sample is continued to assure removal of the CDD/CDF analytes.
Following extraction, the reflux flask is removed and the crude benzene extract solution is processed through the
remainder of the sample preparation procedure. The collected water can be removed from the Dean-Stark receiver and
measured volumetrically or gravimetrically to determine the dried sample weight
RESULTS
Our primary reason for developing the SDS extractor was to maximize the efficiency of extraction of CDDs/CDFs from
wet, solid matrices. Therefore, initial evaluations of the device were designed to compare results from SDS extraction
to two different, standard, Soxhlet extraction procedures using wet sediment samples collected from the Biltic Sea as
the test material. The three extraction tests can be summarized as follows:
SI'S - Wet. SDS extraction of -25 g of wet sediment using the procedure described in the experimental section.
Soxhlet - Wet. Soxhlet extraction of-25 g of wet sediment The sediment was spread as a thin film on Whatman
GF/C glass microfiber filters, fortified with internal standard solutions, and extracted for 16 hours with benzene.
Soxhlet - Dry. Soxhlet extraction of pre-dried sediment. An aliquot of the wet sediment (-35 g) was spread on
the inner surface of an 800-mL beaker and placed under a stream of purified nitrogen for -48 hours to dry. At
this time, the dried material could easily be ground into a fine powder which was fortified with internal stan-
dards and Soxhlet extracted with benzene for -16 hours.
All of the extracts were then processed through identical sample preparations using a
Table 1. Comparison of extraction techniques for TCDD and TCDFin sediment.
Sample
Identification'
Sediment # 4
Sediment # 5
Sediment # 6
Reagent Blank
—
Benzene
Extraction
Technique
Soxhlet - Wet
Soxhlet - Dry
SDS - Wet
Soxhlet - Wet
Soxhlet - Dry
SDS - Wet
Soxhlet - Wet
Soxhlet - Dry
SDS - Wet
Soxhlet - Wet
Soxhlet - Dry
SDS - Wet
PARTS PER TRILLION
(based on dry sediment weight)
2378-TCDD
ND
4.9
5.5
ND
0.8
1.0
ND
1.6
1.6
ND
ND
ND
(0.6)
(0.3)
(0.4)
(0.4)
(0.2)
(0.5)
(0.3)
(0.3)
(3pg)
(5pg)
(3pg)
%Rec
84%
86%
85%
69%
78%
64%
66%
85%
69%
67%
79%
73%
2378-TCDF
1.0
52
59
1.9
8.9
9.6
2.6
16
16
ND
ND
ND
(0.2)
(0.3)
(0.4)
(3pg)
(4pg)
(2pg)
%Rec
82%
91%
77%
91%
71%
76%
77%
69%
71%
84%
78%
74%
(* See Table 2 for water content of these samples according to number.)
modification of our previously
reported procedure (13). The
analytes and internal stan-
dard recoveries were meas-
ured by capillary column gas
chromatography-low resolu-
tion mass spectrometry using
a Hewlett-Packard Model
5987-A quadrupole instru-
ment operating in the se-
lected-ion-monitoring mode at
unit resolution: column, 0.17
mm ID x 20 m J&W DB-5
fused silica capillary with a
0.40 urn film thickness (J&W
Scientific Inc., Folsom, CA
95630); injector, splitless con-
figuration operated at 280°C
with a 1.0 minute splitless
valve time; carrier gas,
helium at -30 cm/second
linear velocity; GC/MS
interface, column coupled
directly to the ion source with
the interface operating at
250°C; ion source, electron
I
'
-------
•. ?:
.1
"•: -a
'v I
30
impact mode operated at 300°C and 70 eV ionization potential; ion masses monitored for TCDD = 320, 322,334; ion
masses monitored for TCDF = 304,306,318,376. During the examination of sample residues for 2378-TCDF, the
parent ion for hexachlorodiphenyl ether (m/z 376) was also monitored in order to demonstrate no interference from
such species.
The results of our comparison of sediment extractions for 2378-TCDD and 2378-TCDF are shown in Table 1. Three
sediment samples and a reagent blank are compared for each of the three extraction procedures studied. The concen-
trations of each analyte in parts per trillion, based on the dry sediment weight (i.e., corrected for water content), are
given. An ND signifies that the analyte was not detected at the limit of detection given in parentheses. These results
have been corrected for the internal standard recoveries reported in Table 1.
A secondary criterion for the applicability of the SDS extractor was the accuracy of its sample water content measure-
ment We compared results from a series of wet sediment samples that were analyzed for water content by the two
techniques given below. Results for these experiments are presented in Table 2.
SDS Extraction. The water content of a sample aliquot used for TCDD/TCDF determination was measured
gravimetrically after 16 hours extraction time.
Oven Drying. A separate aliquot of each sample was oven-dried at 10S°C for 16 hours and the weight loss
measured gravimetrically to calculate the water content.
DISCUSSION
As can be seen from the data in Table 1, the SDS extraction procedure yields results for TCDD and TCDF in wet
sediment samples that are comparable to those obtained using the currently accepted method of air drying followed by
Soxhlet extraction of the homogenized, dried powder. The significant difference is that the SDS procedure does not
require the -48 hour drying period prior to extraction. Both of these methods of extraction are much more effective
than a Soxhlet extraction of the wet sediment with benzene. We have used this technique (Soxhlet extraction of wet
particulate samples) successfully when the total amount of water in the matrix was less than -5 g. This amount of
water could be absorbed by the silica gel in the extraction thimble or distributed throughout the apparatus via
azeotropic distillation; hence, the sample matrix is dried and then efficiently extracted with benzene. When a greater
amount of water is present, the Soxhlet extraction system's capacity for redistributed water is easily exceeded and the
sample never becomes adequately dry to allow efficient extraction of CDDs/CDFs adsorbed on the particulates. This
occurrence may not be apparent since added internal standards, which are not actually adsorbed on the total surface of
the sample, may be recovered quantitatively (as is the case in Table 1). Because we have not attempted to determine
the exact water capacity of the Soxhlet extraction apparatus and we do not usually know the water content of a solid
sample before beginning the extraction, we have adopted the use of the SDS
= extractor for all solid samples which would otherwise be extracted by the
classical Soxhlet technique.
Table 2. Comparison of water content
determinations on wet sediment.
WEIGHT % WATER
The data in Table 2 show that the SDS extractor can be used to determine
the water content of the actual sample aliquot being extracted for CDDs/
CDFs analysis. This can be an important advantage when only a limited
amount af sample is available or there is some concern about the homogene-
ity of the water content of the bulk sample. Closer examination of the data
in Table 2 shows that the water levels determined by the SDS procedure are
consistently lower than the standard oven drying technique by -5% relative.
Although we do not believe that this is a significant difference for typical
samples, we suspect that it is due to small amounts of water that adhere to
various parts of the apparatus and are not collected in the Dean-Stark
receiver. When a greater degree of accuracy in the water determination is
required, it may be possible to dislodge these water droplets from the
glassware and collect them in the Dean-Stark receiver.
Operation of the SDS extraction system in our laboratory over the past year indicates it to be a convenient method of
efficiently extracting wet, solid samples for CDDs/CDFs determinations while eliminating the need for separate proce-
dures to dry and measure moisture content in a given matrix. In addition to the convenience, we have discovered other
advantages: (A) Wet solids can be extracted directly with benzene or other aromatic solvents without external drying
procedures that may lead to either loss of analytes or contamination of the sample. (B) Total analysis time is signifi-
cantly reduced by eliminating the usual sample drying procedure. (C) Extraction of organics entrained in the solid
sample matrix may be more efficient than with externally dried samples because the water is removed slowly by
dissolving it in warm benzene. Hence, as the water is removed it is replaced by solvent, therefore, the matrix never
becomes devoid of a solvent and does not tend to solidify into an "impenetrable* mass as it does when the original
water is permitted to evaporate directly. When combined with the dispersion of the sample as a thin film on a clean
supporting surface, this new mode of simultaneous drying and extraction should be beneficial for recovery of various
Sample
Number
1
2
3
4
5
6
Oven Drying
at 1058C
23.3
47.4
49.1
70.0
72.0
70.2
SDS
Extraction
22.2
45.0
46.6
65.9
69.5
66.6
538
-------
31
organic species from wet, solid matrices.
REFERENCES
1. R. M. M. Kooke, J. W. A. Lustenhouwer, K. Olie and 0. Hutzinger, Anal- Chem. 5JL461-463 (1981).
2. R. E. Clement, A. G. Viau and F. W. Karasek, Can, J.. Chan. 82, 2629-2633 (1984).
3. D. I. Townsend, L. L. Lamparski and T. J. Nestrick, Chemospherp Ifi. 1753-1757 (1987).
4. G. F. VanNess, J. G. Solch, M. L. Taylor and T. O. Tiernan, Chemospherg 2, 553-563 (1980).
5. H. Hagenmaier, H. Brunner, R. Haag and A. Berchtold, Chemosphere 15, 1421-1428 (1986).
6. L. A. ShadofF, R. A. Hummel, L. L. Lamparski and J. H. Davidson, BulL Environ. Contarp. and Toxieol. 12,478-
485 (1977).
7. L. M. Smith, D. L. Stalling and J. L. Johnson, Anal. Chem. 5ji, 1830-1842 (1984).
8. R. C. G. Wegman, J. Freudenthal, G. A. L. deKorte, G. S. Groenemeijer and J. Japenga, Chemosphere 15, 1107-
1112 (1986).
9. J. M. Czuczwa and R. A. Kites, Environ. Sci. ledmd. 22, 195-200 (1986).
10. I. Comoni, A. DiMuccio, D. Pontecorvo and L. Vergori, J. Chromatogr. 153. 233-238 (1978).
11. F. Soxhlet and J. Szombathy. Dingler's Polvtech. Journal 232. 461-165 (1879).
12. E. W. Dean and D. D. Stark, J, of Ind. and Eni* Chem. 12, 486-496 (1920).
13. L.L. Lamparski and T.J. Nestrick, Anal. Chfim, 52,2045-2054 (1980).
tr M *
U O 4
-------
Environmental Toxicology and Chemistry Vol. 5, pp. 355-360, 1986
Crimed in the USA. Pergamon Press Ltd.
0730-7268/86 $3.00 + .00
CONTROL OF INTERFERENCES IN THE ANALYSIS
OF HUMAN ADIPOSE TISSUE TO 2,3,7,8-
TETRACHLORODIBENZO-/7 DIOXIN (TCDD)
D. G. PATTERSON*, J. S. HOLLER, D. F. GROCE, L. R. ALEXANDER,
C. R. LAPEZA, R. C. O'CONNOR, and J. A. LIDDLE
Toxicology Branch, Clinical Chemistry Division, Center for Environmental Health,
Centers for Disease Control, Public Health Service,
U.S. Department of Health and Human Services, Atlanta, Georgia 30333
(Received 20 June 1985; Accepted 17 October 1985) .
Abstract-While developing a method to analyze human adipose tissue for 2,3,7,8-tetrachloro-
dibenzo-p-dioxin (TCDD) at the 1 ppt level, we observed several peaks with all of the analytical
characteristics of TCDDs in the analyses of wipe tests and blank, quality control, and human adi-
pose samples at concentrations equal to the low part-per-trillion level in a 10 g sample. The source
of these contaminants was established to be a cleaning solution used to wash the floors in the lab-
oratory performing the analysis.
Keywords—Dioxin 2,3,7,8-TCDD Interferences Cleaning solution
INTRODUCTION
During the past year, we have been developing
a method for the analysis of 2,3,7,8-tetrachloro-
dibenzo-p-dioxins (TCDDs) in human adipose tis-
sue down to the 1 ppt level. In addition, we have
been synthesizing milligram amounts of numerous
dioxin and furan congeners and monitoring possi-
ble laboratory contamination of these compounds
with wipe tests. Several peaks that have all of the
characteristics of TCDDs have appeared in the
analyses of these wipe tests and human adipose
samples over a period of time. Because of the very
low level of detection required for our analyses
and our desire to be able to analyze for all 22
TCDDs in biological samples, we began to system-
atically search for the source of these contami-
nants. After establishing the structural identity of
the TCDD isomers found in the wipe tests, we
discovered that several of these had never been
worked with nor synthesized in our laboratory. In
addition, the peaks maintained the same relative
ratio to each other in each sample in which they
were detected. This information suggested a sys-
tematic source of contamination from some place
"To whom correspondence may be addressed.
Use of trade names is for identification only and does
not constitute endorsement by the Public Health Service
or by the U. S. Department of Health and Human
Services.
outside the analytical and synthetic program. An
earlier report [1] suggested a possible link between
the illness of a janitress and her cleaning solution
which contained a mixture of bleach and 2-benzyI-
4-chIorophenol dissolved in acid. In this article, we
describe the results of an investigation of one of
the ingredients used in a cleaning solution for
washing the floors of the building in which the
dioxin laboratory is located.
SAMPLE PREPARATION
Enrichment procedure
The samples (spiked with isotopic marker com:
pounds) are processed in a two-part procedure
developed by Smith et al. [2] that we have adapted
for use in our laboratory. These authors have pro-
vided a schematic diagram of the cleanup appara-
tus in Figure 2 of their publication [2J. In part I,
the mixture is subjected to solvent extraction
(methylene chloride/hexane, 50:50), and the ex-
tract is, in the same process, passed through a
series of silica-based absorbents and then through
a carbon/glass fiber adsorbent. The extract passes
through the absorbents in the following order:
potassium silicate, silica gel, and finally the carbon
adsorbent. The residues of interest (polychlori-
nated dizenzo-p-dioxins (PCDDs), polychlorinated
dibenzofurans (PCDFs), and non-ortho substi-
tuted polychlorinated biphenyls (PCBs), as well as
other chemical classes such as polychlorinated
355
-------
356
D. G. PATTERSON ET AL.
naphthalenes (PCNs), polychlorinated biphenyl-
enes, and certain pplyaromatic hydrocarbons) are
retained on the carbon adsorbent and subsequently
recovered by reverse elution with toluene.
In part II of the procedure, after a change of
solvent to hexane, the sample is applied to a sec-
ond series of absorbents contained in the two tan-
dem columns. The first column contains small
amounts of cesium silicate and sulfuric acid-im-
pregnated silica gel. The effluent from this column
flows directly onto the second column containing
activated alumina on which the final fractionation
of several classes of residues is accomplished.
After the sample volume is reduced, analyses are
carried out by high-resolution gas chromatogra-
phy/high-resolution mass spectrometry.
The components of the apparatus used in part I
of the enrichment procedure are as follows: Col-
umn #1 (25 mm i.d. x 600 mm, Michel-Miller
chromatography column 5795-48) is connected to
column #2 (22 mm i.d. X 100 mm, Michel-Miller
pre-column 5796-34) and to column #3 (8 mm
i.d. x 85 mm, Michel-Miller filter column 5813-23,
all from Ace Glass, Vineland, NJ), by means of
Ace Glass #5801 Teflon end fitting adapters and
standard one-eighth inch o.d. Teflon tubing and
tube end fittings (available from most chromato-
graphic supply companies). The washing solvent
reservoir is also an Ace Glass chromatography col-
umn (25 mm i.d. x 450 mm, 5795-40).
The solvent flow switching valves are Hamilton
miniature inert valves (Hamilton Co., Reno, NV):
selector valve (#86781) and bypass and reverse
flow valves (#86779). The valving arrangement is
designed to permit the following operations to be
performed: venting of the solvent line from col-
umn #1, venting of the solvent reservoir, bypass of
column #2, delivery of the effluent from column
#1 to columns #2 and #3 sequentially, delivery of
solvent from the reservoir sequentially to columns
#2 and #3 or to column #3 only, reversal of solvent
flow in columns #2 and #3, and stoppage of sol-
vent flow in all lines. The solvent reservoir is rou-
tinely pressurized with I to 10 p.s.i. nitrogen
during column washings. Column #1 is packed in
the following sequence: one or two discs cut from
glass microfiber filters (GF/F, 4.7-cm diameter,
Whatman Inc., Clifton,-4MJ), a 2-cm depth of
anhydrous sodium sulfate, 15 g of silica gel, 15 g
of potassium silicate, another disc of glass micro-
fiber filter (GF/D), 50 g of a I to 4 (w/w) mixture
of the sample and anhydrous sodium sulfate, and,
lastly, a.2-cm depth of anhydrous sodium sulfate.
Column #2 is packed with equal volumes, 15 ml
each, of potassium silicate and silica gel bracketed
by plugs of glass wool or preferably discs of glass
fiber filters (3 urn retention GF/D, Whatman Inc.,
Clifton, NJ). Column #3 is packed with a mixture
of activated carbon (Amoco PX-21, Amoco Re-
search Corp., Chicago, IL) and glass fibers. The
apparatus for part II of the enrichment procedure
consists of two columns arranged in tandem. Col-
umn #4 is prepared from a disposable Pasteur
pipet and is packed first with a plug of glass wool,
then with 2 cm (0.50 g) of sulfuric acid-impreg-
nated silica gel, then with 3 cm (0.54 g) of cesium
silicate, and, finally, with 0.5 cm of anhydrous
sodium sulfate. Column #5 is constructed from a
225-mm length of 5 mm i.d. heavy-walled glass
tubing fitted with a 50-ml reservoir and a 24/40
ground glass joint. Column #5 is packed with a
plug of glass wool, followed by 3.50 ml (3.65 g) of
activated alumina, and then 0.5 cm of anhydrous
sodium sulfate. The alumina is packed firmly by
sharply tapping the supporting clamp. A flow
chart of the two part cleanup procedure is shown
in Figure 1 along with the function of each step in
the procedure.
Materials
All solvents are glass-distilled grades (Burdick
& Jackson, Muskegon, MI). Silica Gel 60, 70 to
230 mesh (EM Reagent, MC/B, Cincinnati, OH),
acid alumina (AG4, Bio Rad Labs, Richmond,
CA), and sodium sulfate (Mallinckrodt AR) were
used.
All glassware is washed with acetone, toluene,
and, finally, with 50/50 hexane/methylene chlo-
ride (in that order) before use.
Alumina, sodium sulfate, potassium and ce-
sium silicates, sulfuric acid-impregnated silica gel,
and carbon on glass fibers are all prepared as
described by Smith et al. [2]. The cleaning solu-
tion was Vesphene II (for descriptive purposes
only) manufactured by Vestal Labs a Division of
Chemed Corp., St. Louis, Missouri.
Sample Preparation
Before samples were prepared, a clearing run
through the apparatus was conducted that con-
sisted of 100 ml of hexane/methylene chloride
50/50 (solvent A), followed by two samples that
were carried through the sample cleanup process
described in the preceding section.
Sample I. Consisted of 100 ml of hexane/
methylene chloride 50/50 (solvent A), "spiked"
with an internal standard solution consisting of
240 pg of l3C,2-Iabeled 2,3,7,8-TCDD. The stan-
-------
Analysis of human adipose for dioxins
357
SAMPLE PREPARATION
COLUMN EXTRACTION/
PRIMARY CLEAN-UP
ADSORPTION OF
PLANAR AROMATICS
SECONDARY CLEAN-UP
FRACTIONATION OF
PLANAR AROMATICS
ANALYSIS
HOMOGENIZATION/DRYING
ADDITION OF INTERNAL STANDARD
POTASSIUM SILICATE
SILICA GEL
POTASSIUM SILICATE
SILICA GEL
CARBON/GLASS FIBERS
PART I
(AUTOMATED)
CESIUM SILICATE
H2SO4/SILICA GEL
ALUMINA
f- PART
CAPILLARY GC/HRMS
Fig. 1. Dioxin analysis flow chart.
dard is accurately measured, with an SMI "micro/
pettor." The disposable pipet is primed by dipping
the glass capillary into the standard and operating
the plunger rapidly 10 times before the standard is
dispensed. This sample is then loaded onto column
#1, by using washing of solvent A, and carried
through the entire five-column cleanup procedure
described in the preceding section. This sample
serves as a blank for the cleanup apparatus.
Sample 2. Consisted of 20 g of the cleaning
solution extracted five times with 20-ml portions
of solvent A; the combined extracts were "spiked"
with 240 pg of 13C|2-2,3,7,8-TCDD and carried
through the cleanup procedure with the same col-
umns and same absorbents used to process the
blank (Sample I).
INSTRUMENTAL ANALYSIS
The instrument system consists of a Vg ZAB-
2F high-resolution mass spectrometer with a Vg
2250 data system and a Hewlett Packard 5840 gas
chromatograph. The analyses are conducted in an
isomer-specific mode, with a 60M SP2330 capillary
column. The chromatographic conditions are in-
jection in the splitless mode while the temperature
is maintained at 100°C for 2 min, programming to
180°C at 20°C/min, programming to 220°C at
3°C/min, and holding at the final temperature for
20 min. The mass spectrometer is operated in the
high-resolution (static RP =10000 at 10% valley)
selected-ion recording mode, with perfluorotri-
butylamine providing the lock mass at 313.9839 or
254.9854. The most sensitive mode of detection and
quantitation uses a lock mass at 313.9839 and mon-
itors ions in the molecular ion cluster (319.8964,
321.8933) and the same corresponding ions in the
internal standard (331.9864, 333.9335). The less
sensitive confirmation mode monitors additional
ions due to loss of chlorine from the molecular ion
(284.9274, 286.9244) and loss of COC1 from the
molecular ion (256.9324, 258.9296). In this mode
only the most intense ion of the internal standard
(333.9335) is monitored. In the quantitation mode
1-2 pg of 2,3,7,8,-TCDD can readily be observed
at a signal-to-noise greater than 3/1. Generally, 20
pg of 2,3,7,8-TCDD is the minimum amount
required to yield a 3/1 signal-to-noise for the ions
257 and 259. Analytical standards containing 1.5,
2, 10, 20, and 50 pg/ML of 2,3,7,8-TCDD and 48
pg/AiL of I3C12-TCDD are used to establish rela-
tive response factors for quantitative calculations.
Recovery of the internal standard after sample
cleanup is estimated on the basis of absolute ion
counts of the internal standard in the sample
versus those in the standard.
r O '-;
-------
358
D. G. PATTERSON ET AL.
RESULTS AND DISCUSSION
Two peaks with all the analytical characteristics
of 1,3,6,8- and 1,3,7,9-TCDD (see experimental)
have appeared in samples of human adipose, qual-
ity control material, blanks, and wipe tests ana-
lyzed in our laboratory. The levels of these two
peaks were always in the very low part-per-trillion
in adipose, quality control, and blank samples.
These peaks have not been observed in runs of
analytical standards, indicating a source of the
peaks external to the analytical instrument system.
The levels in wipe tests varied considerably, de-
pending on the area in which the test was taken,
as well as when it was taken. The results of the
analysis on the samples described in the experi-
mental section are given in Table 1. The recon-
structed ion chromatograms for a system blank,
an extract from 20 g of the cleaning solution,
and a standard of the 22 TCDDs are shown in
Figure 2. The isomer specificity for the 1,3,6,8-
and 1,3,7,9-isomer is demonstrated in the figure by
the baseline separation of these two isomers in our
analytical method. The criteria for a positive iden-
tification of a TCDD are outlined in Table 2.
We have confirmed elevated levels of two com-
pounds that have the analytical characteristics of
1,3,6,8- and 1,3,7,9-TCDD in the cleaning solution
used to clean the floors in the building containing
SYSTEM
BLANK
CLEANING
SOLUTION
EXTRACT
|X15
1379
CHROMATOGRAPHIC
SEPARATION OF
TCOO ISOMERS
0 h«S. 24MNS. I6SECS
0 HflS. 47MINS. SOSECS.
Fig. 2. Mass chromatograms (m/z 322) for a system
blank, cleaning solution extract, and a standard mixture
of TCDD isomers.
Table 1. Summary of TCDDs' observed in Samples 1 and 2.
Retention
time
33:51
36:08
38:39
40:24
43:54
44:29
Relative
retention
timeb
0.772
0.824
0.882 •
0.922
1.001
1.015
TCDD isomer
(RRT of
standard)
1,3,6,8(0.773)
1,3,7,9(0.825)
1,3,7,8(0.882)
1,2,4,8
1,2,4,7(0.922)
1,3,6,9
2.3,7,8(1.002)
Various'
Concentration'
Blank Sample 1
ND «.l)
ND
ND
ND
NQ (0.2)'
ND
Concentration0
Sample 2
1.7d
NQ (0.7)
NQ (0.2)
NQ
NQ (0.2)'
NQ
Signal
to
noise
42
17
3.0
2.1
3.5
1.0
Isotope
ratio
(320/322)
0.73
0.86
0.78
0.69
0.84
0.70
'All concentrations are calculated assuming a response factor of 1 relative to 2,3,7,8-TCDD and are in part-per-trillion
(ppt). The recovery based upon absolute areas of the internal standard is at least 55%.
"The relative retention time (RRT) is calculated on the basis of 1.00 for 2,3,7,8-13C,2-TCDD internal standard.
'When calculated concentration is below 1.0 ppt, the determination is not quantifiable with the calculated amount
given in parentheses, for example, NQ (ppt); ND is not detected, with the limit of detection for the analysis given
in parentheses.
dThis represents a minimum amount present since the I3C,2-TCDD internal standard was spiked after the extrac-
tion of the sample. In a separate experiment, a 20-g sample was spiked with internal standard before extraction and
the overall recovery was 30%, with a calculated concentration of 2.6 ppt.
'The observed signal for 2,3,7,8-TCDD is indistinguishable from the amount present in the 2,3,7,8-l3C,2-TCDD in-
ternal standard.
rl,2,3,8-, 1,2,3,7-, 1,2,4,9-, 1,2,4,6-, and 1,2,3,4-TCDD all elute at this retention time.
-------
Analysis of human adipose for dioxins
359
Table 2. Criteria for TCDD identification
Retention Time of Each group of ions must max-
Monitored Ions
Ratio of Isotopic
Ions Monitored
Ratio of Signal
to Noise
imize within 1 s of each other.
The relative retention time (to
13C,2-TCDD) must be within 2
pans per thousand of the analyti-
cal standard.
For TCDDs, 0.77 ±0.1
(m/z 320/322)
Greater than or equal to 3.0
the dioxin laboratory. The three active ingredients
in this germicidal detergent are the sodium salts of
ortho-hydroxy biphenyl (5.5%), para-/-amylphenol
(1.1%), and ortho-benzyl-p-chlorophenol (3.3%).
This latter compound probably is prepared from
the sodium or potassium phenoxide and benzyl
chloride, followed by chlorination of the resulting
o-benzylphenol (US Paten* 1,967,825) [3].
The self condensation, along with the accom-
panying "Smiles" rearrangement [4] of either of
two different trichlorinated phenols (TCPs), could
account for the possible presence of these two
TCDDs in the cleaning solution (see Fig. 3). One
of these TCPs (2,4,6-TCP) could be formed by
chlorination of unreacted phenol in the last step
of the synthesis of ortho-benzyl-p-chlorophenol
(chlorophene).
A mixture of compounds containing, among
other substances, 2-hydroxy-2',3,4,4',5-penta-
chlorodiphenylether at the part-per-million level
was put through the cleanup procedure, and no
Ci
1,3,6,8,-TCDD
- 1.3.7.9-TCDD
Fig. 3. Formation of 1,3,6,8- and 1,3,7,9-TCDDs.
1.2.3,8-TCDD
1,2,3,7-TCDD
Fig. 4. Formation of 1,2,3,8- and 1,2,3,7-TCDDs.
.
J
-------
360
D. G. PATTERSON ET AL.
TCDDs were detected. If TCDDs were formed
during the cleanup procedure by interacting with
acidic or basic chromatographic materials, then
this ortho-hydroxydiphenyl ether should have pro-
duced measurable amounts of l-,2,3,8- and 1,2,3,7-
TCDDs (see Fig. 4).
A review of the on-line toxicological informa-
. tion data bases suggests that the 1,3,6,8-isomer is
less active in one biological test by -10~7 com-
pared with the 2,3,7,8-isomer. Due to the low
concentration and the apparent low toxicity of
1,2,3,8-TCDD, we feel that this cleaning solution
does not present a health risk to workers. The
observation of this isomer in general laboratory
contamination is analytically important in labora-
tories where ultra-trace dioxin analyses are being
conducted.
REFERENCES
1. Lynch, R.E., G.R. Lee, and J.P. Kushner. 1975. Por-
phyria cutanea tarda associated with disinfectant mis-
use. Arch. Intern. Med. 135:549-552.
2. Smith, L.M., D.L. Stalling, and J.L. Johnson. 1984.
Determination of part-per-trillion levels of polychlo-
rinated dibenzofurans and dioxins in environmental
samples. Anal. Cliem. 56:1830-1842.
3. Windholz, M. and S. Budauari. 1983. The Merck In-
dex and Encyclopedia of Chemicals, Drugs, and Bio-
logicals. Merck and Co., Inc., Rahway, NJ, p. 2364.
4. Kende, A.S. and M.R. Decamp. 1975. Smiles rear-
rangement in the synthesis of hexachlorodibenzo-p-
dioxins. Tetrahedron Letters 33:2877-2880.
r .1 r
-------
United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
P.O. Box 15027
Us Vegas NV 89114-5027
EPA 600/4-86-004
January, 1986
Research and Development
Protocol for the
Analysis of 2, 3, 7, 8-
Tetrachlorodibenzo-p-Dioxin
by High Resolution Gas
Chromatography/
High-Resolution
Mass Spectrometry
-------
PROTOCOL FOR THE ANALYSIS OF 2,3,7,8-TETRACHLORODIBENZO-£-DIOXIN BY
HIGH-RESOLUTION GAS CHROMATOGRAPHY/HIGH-RESOLUTION MASS SPECTROMETRY
by
John S. Stanley and Thomas M. Sack
Midwest Research Institute
Kansas City, Missouri 64110
Contract Number SAS 1576X
Project Officer
Werner F. Beckert
Quality Assurance Division
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada 89114
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
Office of Research and Development
U.S. Environmental Protection Agency
Las Vegas, Nevada 89114
f 4 >~)
U '-1 (L
-------
NOTICE
The information in this document has been funded wholly or in part by
the United States Environmental Protection Agency under Contract Number
SAS 1576X to the Midwest Research Institute, Kansas City, Missouri. It has
been subject to the Agency's peer and administrative review, and it has been
approved for publication as an Environmental Protection Agency document.
Mention of trade names or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
PREFACE
This report describes the activities completed as part of a single-
laboratory evaluation of a high-resolution gas chromatography/high-
resolution mass spectrometry method for the determination of tetrachloro-
dibenzo-£-dioxins in water, soil, and sediment samples. The work described
in this report was completed at the Midwest Research Institute under con-
tract to Viar and Company (Special Analytical Services SAS 1576X) for the
U.S. Environmental Protection Agency, Environmental Monitoring Systems
Laboratory, Quality Assurance Division, Las Vegas, Nevada. The revision of
the protocol to allow for lower quantitation limits for tetrachlorodibenzo-
£-dioxins was carried out at the Environmental Monitoring Systems Laboratory-
Las Vegas.
This report was prepared with assistance from M. McGrath. The authors
acknowledge the technical project monitor, W. F. Beckert, as well as R. K.
Mitchum and S. Billets of the Environmental Monitoring Systems Laboratory-
Las Vegas and, especially, Y. Tondeur of the Environmental Research Center,
University of Nevada, Las Vegas for guidance provided during this study.
111
r
-------
ABSTRACT
This report provides the results of the single-laboratory evaluation of
a high-resolution gas chromatography/high-resolution mass spectrometry
method for the determination of 2,3,7,8-tetrachlorodibenzo-£-dioxin and
total tetrachlorodibenzo-£-dioxins at concentrations ranging from 10 to
200 pg/g (ppt) in soils and 100 to 2,000 pg/L (ppq) in water. The report
summarizes the data for the precision and accuracy of triplicate measure-
ments of five solid and five aqueous samples. The results indicate that the
method is capable of generating accurate and precise data within the concen-
tration limits specified above and within absolute recoveries of 40 to 120
percent with 50 percent precision. An attempt to reach a quantitation limit
for TCDD of 2 ppt (or less) for soil and 20 ppq (or less) for aqueous sam-
ples was not successful. Based on the data generated during this study and
based on discussions at the Environmental Monitoring Systems Laboratory-
Las Vegas, the Environmental Monitoring Systems Laboratory-Las Vegas revised
certain parts of the protocol to lower the quantitation limit for tetra-
chlorodibenzo-£-dioxins to 2 ppt in soil and 20 ppq in water samples.
IV
rr
-------
CONTENTS
Preface iii
Abstract iv
Figures vi
Tables vii
1. Introduction 1
2. Conclusions 3
3. Recommendations 5
4. Experimental Procedures 7
Sample description. 7
Sample preparation 7
Reagents 9
HRGC/HRMS instrumentation 9
Mass measurement accuracy 11
Chromatographic resolution 13
Injection technique 13
5. Results and Discussion 14
Approach to cleanup column evaluation 14
Final method evaluation 22
References ^2
Appendices 43
A. Validated Analytical Protocol
B. Proposed Analytical Protocol
rr
-------
FIGURES
Number Page
1 Column cleanup procedures specified in the protocol 15
2 Column cleanup procedures proposed by the EMSL-LV 16
3 Background levels of 1,3,6,8- and 1,3,7,9-TCDD observed over
the single-laboratory evaluation study 41
vx
' -I
/
-------
TABLES
Table Page
1 Solid Samples Used for HRGC/HRMS Method Evaluation ...... 8
2 Aqueous Samples Used for HRGC/HRMS Method Evaluation ..... 8
3 TCDD Isomers Used for HRGC/HRMS Method Evaluation ....... 10
4 Composition of Concentration Calibration Solutions (pg/pL) . . 10
5 HRGC/HRMS Operating Conditions ................ 12
6 Recovery (%) of Several TCDD Isomers from Cleanup Option A . . 16
7 Recovery (%) of Several TCDD Isomers from Cleanup Option B .. . 19
8 Recovery (%) of Several TCDD Isomers from Cleanup Option C . . 20
9 Recovery (%) of Several TCDD Isomers from Cleanup Option D . . 21
10 Initial Calibration Summary .................. 23
11 HRGC and Mass Resolution Check Summary ..... ....... 24
12 TCDD Data Report Form ..................... ?.6
13 Accuracy and Precision of the HRGC/HRMS Analysis for
2,3,7,8-TCDD from Laboratory Aqueous Matrix Spikes ..... 33
14 Precision of the HRGC/HRMS Analysis for 2,3,7,8-TCDD of
Soil and Fly Ash Samples .................. 34
15 Accuracy of the HRGC/HRMS Method for the Determination of
TCDD Isomers Spiked into Aqueous Matrices .......... 35
16 Accuracy of the HRGC/HRMS Method for the Determination of
TCDD Isomers Spiked into Soil Matrices ........... 36
17 Fortified Field Blank Results ................. 37
VII
rr /? o
-------
SECTION 1
INTRODUCTION
The U.S. Environmental Protection Agency's (EPA) strategy for dealing
with dioxin requires the development and validation of an analytical method
capable of achieving detection of the tetrachlorodibenzo-£-dioxins (TCDD),
specifically 2,3,7,8-TCDD, at the parts-per-trillion (ppt) level in soil and
sediment and parts-per-quadrillion (ppq) level in water.1 This validated
method will be used by qualified contract laboratories to extend the analyt-
ical capabilities for such analyses to all EPA regional and program offices.
This report deals specifically with the single-laboratory evaluation of
a high-resolution gas chromatography/high-resolution mass spectrometry
(HRGC/HRMS) analysis method for TCDDs in soil, sediment, and water. The
method (Appendix A) is intended to provide quantitative determination of
TCDD at levels of 10 to 200 pg/g (soil and sediment) and 100 to 2,000 pg/L
(water) at a mass resolution of 10,000. This single-laboratory evaluation
has been completed as part of the validation process recommended by EPA.2
The proposed method was prepared after several candidate methods were
reviewed and their best features were selected. After peer review, the pro-
posed method was refined for completeness, technical accuracy, clarity, and
regulatory applicability. The single-laboratory evaluation of the proposed
analytical method has been accomplished through three tasks. The first task
involved preliminary performance testing of the method using TCiJD-
contaminated soils and TCDD-spiked aqueous samples. The results o* this
study indicated that the proposed method required modification to achieve
the target method detection limits and the accuracy and precision criteria.
The second task focused on ruggedness testing of the chromatographic cleanup
procedures. The results of this study were used to modify the proposed
method. This report is focused on the results of the triplicate analysis
of five solid and five aqueous samples completed under the third task of
the evaluation, using the modified method.
Section 2 of this report summarizes the conclusions based on the
single-laboratory evaluation of this method using TCDD-contaminated soils
and TCDD-spiked aqueous samples. Section 3 presents recommendations that
should be considered for inclusion in the method before proceeding with
collaborative testing. Section 4 presents some specific experimental con-
ditions, and Section 5 summarizes the analytical data for the triplicate
analysis of four soil, one fly ash, and five aqueous samples completed in
the third task of the single-laboratory evaluation. Triplicate analyses of
a l-pg/(jL calibration solution did not give satisfactory results. In order
to achieve a quantitation limit of 2 ppt for soil (using a 10-g sample) and
rr /?
-------
20 ppq for water (using a 2.0-L sample), the protocol evaluated in this
study was modified. The rationale for the modifications and the revised
protocol are included as Appendix B.
-------
SECTION 2
CONCLUSIONS
The single-laboratory evaluation of the analytical method for the
determination of 2,3,7,8-TCDD in soil and aqueous samples demonstrates that
the method as described is capable of achieving the target detection limits
of 10 pg/g (ppt) for soils and 100 pg/L (ppq) for water.
The relative response factors (RRF) determined for native 2,3,7,8-TCDD
versus the internal standard 13C12-2,3,7,8-TCDD, and the RRF of the internal
standard versus the recovery standard 13C12-1,2,3,4-TCDD over the five-point
concentration calibration curve demonstrate that the HRGC/HRMS method main-
tains a linear response for 2,3,7,8-TCDD from 10 to 200 ppt for soils and
100 to 2,000 ppq for water.
The results of the analysis of spiked aqueous samples demonstrate that
internal standard (isotope dilution) quantitation provides an accurate mea-
surement of 2,3,7,8-TCDD. The accuracy of the 2,3,7,8-TCDD measurement for
triplicate analysis of four water samples spiked at various concentrations
was quite good. The accuracy of measurement for 2,3,7,8-TCDD averaged 104
percent for three aqueous matrices prepared as laboratory matrix spikes.
The absolute recovery of the internal standard 13C12-2,3,7,8-TCDD did not
significantly affect the accuracy of the 2,3,7,8-TCDD determination. The
precision of the analyses for 2,3,7,8-TCDD ranged from 3.6 to 16 percent for
replicate analyses of the five aqueous samples. The precision of the crip-
licate analyses of the soil samples was somewhat higher than determined -"or
aqueous samples. The precision of triplicate analyses of the four soil
samples ranged from 19 to 50 percent. The difference in precision from that
of the aqueous samples may be attributable to the potential for TCDD aC£,urp-
tion on the soil samples.
The results from the analyses of soil and aqueous samples spiked with
additional TCDD isomers demonstrate that the internal standard quantitatiun
gives good estimates of total TCDD values. The accuracy of the analyses of
fortified distilled water and influent and effluent wastewaters averaged
101 ± 14 percent for five TCDD isomers. The accuracy of the measurements of
these isomers for the four fortified soil samples averaged 87 ± 24 percent.
The results of the analyses demonstrate that the requirements for abso-
lute recovery of the internal standard (40 to 120 percent) and precision of
replicate analyses (RPD < 50 percent) can be achieved for relatively clean
samples.
rr ^ \
i \ 11 i.
-------
The sample matrix can severely impact the performance of the analytical
method. This is evidenced by the consistent low recovery of the internal
standard from the fifth aqueous sample, an industrial wastewater, and from a
fly ash sample. The low recovery from the industrial wastewater is possibly
due to the effect of coextractants on the elution sequence from alumina.
The low recoveries observed for the fly ash sample, on the other hand, may
be attributed to adsorption by the sample matrix.
One of the most critical variables in the analytical method is the com-
pleteness of removal of the benzene from the extract before proceeding with
the acidic alumina column fractionation. The cleanup column ruggedness
testing experiments demonstrated that the recoveries of 2,3,7,8-TCDD and the
other TCDDs are affected by the presence of benzene in the alumina column
fractionation step.
The analyst must be aware of the potential problem of interferences
arising from background contamination. For example, the 1,3,6,8- and
1,3,7,9-TCDD isomers were present in the fortified field blanks in this
work. From other referenced activities it becomes clear that these isomers
may present problems in other laboratories as well. The fortified field
blanks are important tools in assessing the background contamination prob-
lems over time.
Although the l.O-pg/pL standard did not yield satisfactory results in
this study, due to unacceptable ion ratios, the response factors are within -
the established curve. The data for the triplicate analyses of the 1.0-|jg/|jL
standard demonstrate that the characteristic ions for TCDD were greater than
20:1 for the m/z 322 S/N and approximately 10:1 for m/z 259 S/N. Thus, it
should be possible to extend the detection limit to 1 pg/|jL if an allowance
for abundance ratios based on ion statistical errors is incorporated.
Based on the column performance and bleed characteristics, the column
of cfiolrs for the analysis for TCDD at ppt (for soils and sediments) and ppq
(for. water) levels appears to be the 50-m CP-Sil 88 with a 0.2-pm film
thickners. To preserve the performance characteristics of the HRGC columns,
an injection technique that excludes any air is highly recommended.
r-- c;
O o <
-------
SECTION 3
RECOMMENDATIONS
1. Mass measurement accuracy should properly be determined relative to the
lock mass (if any), rather than m/z 254.9856, because it is that rela-
tionship which will determine how accurately the masses of the TCDD
ions will be measured.
2. It is recommended that the chromatographic resolution check be per-
formed on the summed ion chromatograms of m/z 259 + m/z 320 + m/z 322.
This yields a chromatogram which is less noisy and more representative
of the true column performance.
3. The 5 percent peak width criterion for mass resolution should be the
selected mass/1,000 mmu rather than 31.9 mmu because the protocol
allows peaks other than m/z 319 to be used for resolution measurement
(e.g., 31.7 mmu if m/z 317 is used).
4. It is recommended that the mass measurement accuracy be recorded and
reported along with the resolution check summary table.
5. The addition of the recovery standard 13C12-1,2,3,4-TCDD should be
achieved by using a spike volume of 25 to 50 |JL rather than 5 pL to
minimize errors resulting from volume measurement.
6. The recommended temperature program settings in the method shou.M be
converted to those presented in the experimental section of this re-
port. These conditions were established for analysis with tridecane r.s
the solvent.
7. Lower limits of detection can be achieved by allowing the analyse to
concentrate the final extract to as low as 10 pL. It may be necessary
to use the smaller final volume with other HRMS instruments to achieve
the same levels of detection.
8. The method should recommend several techniques to break up emulsions
resulting from extraction of aqueous samples. In this evaluation the
emulsion phase was put through a column packed with glass wool, which
was then rinsed with additional methylene chloride. Other options
might include stirring or centrifugation of the emulsion phase.
9. The method should specify the procedure to deal with aqueous samples
containing high levels of suspended solids. In this study it was
i- c. '\
•J uO
-------
necessary to centrifuge the soil extract sample before proceeding with
the extraction.
10. It is highly recommended that the method be modified such that the ben-
zene extract is completely exchanged to hexane prior to cleanup on the
silica column since this is apparently one of the most critical factors
leading to successful sample analysis.
11. It may be worthwhile to evaluate a cleanup procedure in which the char-
coal column precedes the alumina column as a means to improve method
recovery.
-------
SECTION 4
EXPERIMENTAL PROCEDURES
SAMPLE DESCRIPTION
Five solid samples were provided by the Environmental Monitoring
Systems Laboratory-Las Vegas (EMSL-LV) to the Midwest Research Institute
(MRI) for analysis for 2,3,7,8-TCDD and total TCDD using the analytical
method in Appendix A. A description of the five solid samples and the esti-
mated 2,3,7,8-TCDD concentrations from previous analyses by an independent
laboratory are provided in Table 1. Each sample was analyzed in triplicate
as specified in the protocol. One of the triplicate samples for each soil
sample was spiked with the seven TCDD isomers (1,3,6,8-; 1,3,7,9-; 1,2,3,7-;
1,2,3,8-; 1,2,3,4-; 1,2,7,8-; and 1,2,8,9-TCDD) at approximately 10 times
the estimated level of 2,3,7,8-TCDD specified in Table 1.
Five aqueous samples were generated for the evaluation of the analyt-
ical method at the ppq detection level. Table 2 presents a description of
each water type and lists the fortification levels of 2,3,7,8-TCDD and seven
additional TCDD isomers (1,3,6,8-; 1,3,7,9; 1,2,3,7-; 1,2,3,8-; 1,2,3,4-;
1,2,7,8-; and 1,2,8,9-TCDD) in each sample.
The influent and effluent wastewater samples were collected from a
sewage treatment facility in metropolitan Kansas City, Missouri. The indus-
trial wastewater was obtained from a holding pond within a hazardous waste
area that was known to be highly contaminated with PCBs and possibly other
chlorinated aromatic compounds (chlorobenzenes). This aqueous sample was
very acidic (pH < 1) and was dark in color.
The soil extract was prepared from 30 g of a soil sample, Hyde Park 002
(H2), and 1 gallon of distilled water. The mixture was stirred constantly
(at least 24 hrs) until just prior to subsampling of 1.0-L aliquots.
SAMPLE PREPARATION
All samples listed in Tables 1 and 2 were extracted and analyzed in
triplicate according to the protocol provided in Appendix A. As indicated
in Tables 1 and 2, one aliquot of each sample matrix was fortified with
additional TCDD isomers, which represent the compounds that elute first
(1,3,6,8-TCDD), last (1,2,8,9-TCDD), and within the approximate retention
window of 2,3,7,8-TCDD (1,2,3,7-; 1,2,3,8-; and 1,2,3,4-TCDD) from the HRGC
columns used for sample analysis.
-------
TABLE 1. SOLID SAMPLES USED FOR HRGC/HRMS METHOD EVALUATION
EPA sample no.
B25-Piazza Road (B5)
Hyde Park 001 (HI)
B52-Shenandoah (Bl)
Hyde Park 003 (H3)
RRAI-5,7,8 (FA)
Matrix
Soil
Soil
Soil
Soil
Fly ash
Approximate
sample
size
10 g
10 g
1 8
1 8
10 g
Estimated
2,3,7,8-TCDD
concentration
(ppt)B
50
70
360
1,700
NRd
Spike level
(ppt.) of
TCDD isomers
100
140
720
1,700
e
.Approximate sample size of each replicate sample.
Estimated level of endogenous 2,3,7,8-TCDD reported to MRI by
W. Beckert in letters dated April 19, 1985 and August 30, 1985.
.Approximate fortification level of each of seven additional TCDD isomers.
No estimate of 2,3,7,8-TCDD concentration was reported.
Additional TCDD isomers were not spiked into this matrix.
TABLE 2. AQUEOUS SAMPLES USED FOR HRGC/HRMS METHOD EVALUATION
Fortification Fortification
Approximate level of level of
sample 2,3,7,8-TCDD TCDD isomers
Sample type size (ppq) (ppq)
Distilled water (DW)
POTW influent (IWW)
POTW effluent (EWW)
Industrial wastewater (IND)
Hyde Park 002; soil extract (H2W)
1
1
1
1
1
.0 L
.0 L
.0 L
.0 L
.0 L
250
500
1,000
500
c
500
1,000
2,000
1,000
c
.Approximate sample size of each replicate sample.
Approximate fortification level of each of seven additional TCDD isomers.
This aqueous sample was not fortified with TCDD isomers.
8
u
-------
All samples were fortified with 500 pg 13C12-2,3,7,8-TCDD in 1.5 mL
acetone. The solid samples were extracted continuously for 24 hr in a
Soxhlet apparatus with benzene and the 1.0-L aqueous samples were batch-
extracted using 2.0-L separatory funnels and three 60-mL portions of methyl-
ene chloride. The extractions of the influent wastewater (IWW) and effluent
wastewater (EWW) and the soil extract (H2W) were complicated by the forma-
tion of emulsions. In each case, the emulsion was removed by passing the
methylene chloride and emulsion layer through a column of glass wool pre-
rinsed with methylene chloride. The extract and resulting aqueous layers
were collected in a sample bottle and the glass wool plug was rinsed with an
additional 10 mL methylene chloride. Following the complete extraction of
the aqueous sample, the contents of the bottle were transferred to a clean
250-mL separatory funnel and the methylene chloride was removed from the
aqueous phase that was transferred with the emulsion. All extracts were
concentrated with Kuderna-Danish evaporators and nitrogen evaporation to
approximately 1.0 mL. Each extract was taken through the entire cleanup
procedure including the acidic silica, acidic alumina, and Carbopak C as
specified in the protocol (Appendix A). The HRGC/HRMS analysis of each ex-
tract was completed as specified below.
REAGENTS
All solvents for extraction and cleanup were obtained as "Burdick and
Jackson distilled-in-glass" quality. The tridecane (99 percent purity) was
obtained from Aldrich (TS, 740-1). The chromatographic materials, acidic
alumina (100-200 mesh AG-4, Biorad Laboratories 132-1340), silica (70-230
mesh Kieselgel 60, EM Reagent, American Scientific Products C5475-2), sodium
sulfate, potassium carbonate, Celite 545® (Fisher Scientific Company), and
the silanized glass wool and Carbopak C (80-100 mesh Supelco 1-1025) were
prepared for use as specified in Section 7 of the protocol (Appendix A).
Table 3 provides the sources of standards used to prepare the calibra-
tion solutions, sample fortification solutions, recovery standard spiking
solution, internal standard spiking solutions, field fortification solu-
tions, and TCDD isomer fortification solutions.
Table 4 is a summary of the concentration calibration standards pre-
pared for the HRGC/HRMS method evaluation. These standards were prepared as
specified in the protocol (Appendix A) . The standard HRCC6 was included in
the final evaluation of the HRGC/HRMS method as a means to demonstrate the
lower limit of detection under optimum instrumental conditions.
HRGC/HRMS INSTRUMENTATION
Sample extracts and calibration standards were analyzed using a Carlo
Erba Mega Series gas chromatograph (GC) which was coupled to a Kratos MS50
TC double-focusing mass spectrometer (MS). The GC/MS interface was simply a
direct connection of the GC column to the ion source via a heated interface
oven. A Finnigan 2300 Incos data system was used for data acquisition and
processing.
r - ,.,
U u I
-------
TABLE 3. TCDD ISOMERS USED FOR HRGC/HRMS METHOD EVALUATION
Isomer
Stock
concentration
Source
Standard code
2,3,7,8-TCDD
13C12-2,3,7,8-TCDD
1,2,3,4-TCDD
13C12-1,2,3,4-TCDD
l,3,-6,8-/l,3,7,9-TCDD
1,2,3,7-/1,2,3,8-TCDD
1,2,7,8-TCDD
1,2,8,9-TCDD
Column performance
standard
7.87 ± 0.79
50 ± 5
2.7 mg/mL
50 ± 5 Mg/mL
0.82 mg/mL
0.5 mg/mL
0.39 mg/mL
1.46 mg/mL
10 (Jg/mL
EPA QA Reference
Materials
Cambridge Isotope
Laboratories
Cambridge Isotope
Laboratories
Cambridge Isotope
Laboratories
Cambridge Isotope
Laboratories
Cambridge Isotope
Laboratories
Cambridge Isotope
Laboratories
Cambridge Isotope
Laboratories
Cambridge Isotope
Laboratories
20603
R00201 (Lot
AWN-1203-65)
ED-915C (Lot
6578)
R00212 (Lot
AWN-1203-93)
ED-913C (Lot
F2086)
ED-905C (Lot
7371)
ED-915C (Lot
7184)
ED-916C (Lot
MLB-682-26)
ED-908 (Lot
No. R00215)
a...
Mixture of TCDD isomers including 2,3,7,8-; 1,2,3,4-; 1,2,3,7-/1,2,3,8-;
1,2,7,8-; and 1,4,7,8-TCDD.
TABLE 4. COMPOSITION OF CONCENTRATION CALIBRATION SOLUTIONS (pg/ML)
HRCC1
HRCC2
HRCC3
HRCC4
HRCC5
HRCC63
Recovery standard
13C12-1,2,3,4-TCDD
2.5
5.0
10.0
20.0
40.0
1.0
Analyte
2,3,7,8-TCDD
2.5
5.0
10.0
20.0
40.0
1.0
Internal standard
13C12-2,3,7,8-TCDD
10.0
10.0
10.0
10.0
10.0
10.0
This solution is not specified in the analytical method in Appendix A.
10
JoO
-------
The HRGC/HRMS operating conditions used in the final phase of this work
are summarized in Table 5. The GC operating conditions recommended in the
protocol were not used for these analyses for three reasons. First, the
TCDDs have rather long retention times, and the solvent (tridecane) boils at
235°C. Thus no benefit could be realized with a low initial temperature.
Second, past experience at MRI has indicated that 200°C is an acceptable
starting temperature for these types of analyses when tridecane is used as
a solvent. Finally, since the CP-Sil 88 and SP-2330 phases are both very
polar and thinly coated, it has been recommended that they not be subjected
to rapid heating or cryogenic cooling to prevent thermal shock to the
column.3
The MS was tuned daily to yield a resolution of at least 10,000 (10
percent valley) and optimal response at m/z 254.986. This step was followed
by calibration of an accelerating voltage scan beginning at m/z 254 (typical
calibration range was 255 to 605 amu). Other voltage scans from the same
data file were then used to establish and document both the resolution at
m/z 316.983 and the mass measurement accuracy at m/z 330.979.
MASS MEASUREMENT ACCURACY
For this work, mass measurement accuracy was measured relative to PFK
m/z 254.986, as required by the protocol, by applying the mass correction,
Am, to the entire spectrum, which yields an error of 0 ppm at m/z 254.986.
In this way, it was possible to meet routinely the 5 ppm accuracy criterion
at m/z 330.979. However, if a lock mass other than 254.986 is used, the
mass measurement accuracy should be measured relative to that lock mass,
since it is that peak which is used to maintain magnet alignment and will
ultimately control the mass measurements during the selected ion monitoring
(SIM) experiments.
Mass Resolution
Mass resolution at m/z 316.983 was documented by an output of the Incos
PROF program. However, the computer-generated value for resolution was
found to be significantly higher than the value measured manually. Thus,
the manually determined resolution, which was nearly identical to the value
measured by using the peak matching unit, is reported. Closer inspection of
the PROF source code revealed that resolution is computed via a statistical
method, not as m/Am at 5 percent height. Incos users should therefore be
aware of this discrepancy, because the computer-generated value can be as
much as 20 percent over the proper value.
Following calibration, the SIM experiment descriptor was updated to re-
flect the new calibration. Six masses (see Table 5) were monitored by scan-
ning ^ m/10,000 amu over each mass. The total cycle time was kept to 1 sec.
The m/z 280.983 ion from PFK was used as a lock mass because it is the most
abundant PFK ion within the range of m/z 255 to 334 and therefore permits
the use of low partial pressures of PFK, which minimizes PFK interferences
at the analytical masses.
11
r c.
q
J
-------
TABLE 5. HRGC/HRMS OPERATING CONDITIONS
Mass spectrometer
Accelerating voltage:
Trap current:
Electron energy:
Electron multiplier voltage:
Source temperature:
Resolution:
Ions monitored
258.930
319.897
321.894
331.937
333.934
280.9825 (lock mass)
Overall SIM cycle time
Gas chromatograph
Column coating:
Film thickness:
Column dimensions:
Helium linear velocity:
Helium head pressure:
Injection type:
Split flow:
Purge flow:
Injector temperature:
Interface temperature:
Injection size:
Initial temperature:
Initial time:
Temperature program:
= 1 sec
8,000 V
500 pA
70 eV
2,000 V
280°C
10,000 (10% valley definition)
Nominal dwell times (sec)
0.15
0.15
0.15
0.15
0.15
0.10
CP-Sil 88
0.2 (jm
50 m x 0.22 mm ID
~ 25 cm/sec
1.75 kg/cm2 (25 psi)
Splitless, 45 sec
30 mL/min
6 mL/min
270°C
240°C
2 |JL
200°C
1 min
200°C to 240°C at 4°C/min
12
-------
CHROMATOGRAPHIC RESOLUTION
Chromatographic resolution values were measured for the SIM plot of m/z
320. However, it may be advantageous to measure chromatographic resolution
from a plot of the sum of m/z 259, 320, and 322. The sum trace has better
signal-to-noise ratio (S/N) and peak definition than the SIM plots, which
permits a more accurate measurement of resolution.
Selection of the HRGC Column
Three different HRGC columns were evaluated in the course of this proj-
ect: SP-2330 (60 m x 0.24 mm); DBS (60 m x 0.22 mm); and CP-Sil 88 (50 m x
0.22 mm). By evaluating the mass spectra of the bleed from each column at
240 to 250°C, it became apparent that the column background may be the lim-
iting factor in achieving the desired detection limit for this method. The
DBS column provided the least amount of background at 250°C, and the SP-2330
had the worst. This coincides with the fact that quantitation at the detec-
tion limit (i.e., 2.5 pg/HlO with the SP-2330 column was difficult at best.
The CP-Sil 88 column appeared to offer less bleed than the SP-2330 column
and indeed does permit more accurate quantitation due to reduced background
contribution.
The chromatographic performance afforded by these columns is a further
issue, since the column best suited for low detection limits, DB-5, cannot
meet the 25 percent valley chromatographic resolution criteria in all cases.
Both the SP-2330 and CP-Sil 88 columns can easily resolve the 2,3 , 7 ,8-TCDD.
However, based on the bleed considerations discussed above, the 50-m CP-Sil
88 column is recommended for the best combination of low bleed and good
isomer separation.
It may also be advisable that other HRGC columns (including SP-2340,
Silar IOC, and SP-2331) that have, been used for 2,3,7,8-TCDD analysis at the
1-ppb soil level be evaluated for background contribution and their applica-
tion for HRMS analysis at ppt and ppq concentrations.
INJECTION TECHNIQUE
The HRGC column performance can degrade very quickly if proper injec-
tion techniques are not used. Specifically, the SP-2330 and CP-Sil 88
phases are very sensitive to 02 and will decompose rapidly at 200°C if any
trace of 02 is present. Therefore, the common practice of using 1 |JL of air
to flush the syringe and effect reproducible injections is to be avoided,
since even that small amount of air per injection can cause column perfor-
mance to degrade in less than one week of continued use.
The following injection technique is recommended. First rinse the sy-
ringe copiously with isooctane (or other volatile solvent, such as toluene).
Dry the syringe by drawing air through it. Pull up and expel several vol-
umes of tridecane until all bubbles are gone, and leave 1 pL of tridecane in
the barrel. Finally, pull up 2 pL of the sample solution and inject. This
technique has worked very well and yields injection reproducibility compar-
able to that of the air purge method, without introducing air onto the ana-
lytical GC column.
13
r (^ i
obl
-------
SECTION 5
RESULTS AND DISCUSSION
The primary purpose of any method validation process is to assure that
the method under consideration is adequate to meet testing and monitoring
requirements.1 The single-laboratory evaluation of the analytical protocol
presented in this report has been preceded by several evaluation and im-
provement steps. These have included the preparation of a written protocol,
technical review of the protocol for completeness, technical accuracy, and
clarity; preliminary testing to evaluate performance of the analytical
method; and revision and refinement of the written protocol based on the
results of the preliminary testing.
Prior to the assessment of the refined protocol presented in Appendix A,
the proposed analytical method had been evaluated for performance through
the analysis of several duplicate samples. The results of the preliminary
evaluation indicated that problems existed in the design and approach to the
extract cleanup steps, which greatly affected the method detection limit,
accuracy, and precision.
This section presents a summary of the studies that have led to the
refinement of the analytical protocol as provided in Appendix A and also
summarizes the single-laboratory evaluation of this protocol.
APPROACH TO CLEANUP COLUMN EVALUATION
The initial method evaluation completed under the first task resulted
in very low recoveries of the internal standard, 13C12-2,3,7,8-TCDD, and the
accuracy and precision of duplicate sample analyses were poor. After re-
viewing the data, it was apparent that the problems were the result of poor
chromatographic separation in the cleanup columns. The initial protocol in-
volved reducing sample extract volumes to 1.0 mL in benzene, elution through
the acidic silica column with hexane, and collection of the total eluent
which was then added to the acidic alumina column. The alumina column was
further eluted with hexane/20-percent methylene chloride. The eluate was
concentrated and cleaned further using a Carbopak C/Celite column, and the
TCDDs were eluted with 2 mL toluene.
Column cleanup techniques were revised and further evaluated following
the procedures depicted in Figures 1 and 2. The column evaluations were
completed with triplicate measurements at'three spike levels (0.10, 1.0, and
10 ng) equivalent to 10, 100, and 1,000 ppt of TCDD in solids with several
TCDD isomers (2,3,7,8-; 1,3,6,8-; 1,3,7,9-; 1,2,3,4-; 1,4,7,8-; 1,2,3,7-;
1,2,3,8-; and 1,2,8,9-TCDD).
14
-------
OPTION A
1 ml Benzene Extract
H2SO4 - SiO2
4.0g
Si O2
l.Og
Acidic AI2O3
6.0g
30mL 20% CH2Cl2/Hexane
Concentrate to 100/it
Carbopak C/Celite
6mL Toluene
HRGC/HRMS
OPTION B
1 ml Benzene Extract
H2SO4 - SiO2
4.0g
SIO2
l.Og
Acidic A1203
6.0g
30mL20%
Carbopak C/Celite
6mL Toluene
HRGC/HRMS
Figure 1. Column cleanup procedures specified in the protocol
15
583
-------
OPTION C
1 mL Benzene Extract
H2SO4 - SiO2
Si O2
l.Og
Concentrate to O.SmL
I
Acidic AI2O3
o.Og
30mL 20% CH2Cl2/Hexane
Concentrate to 100/it.
Carbopak C/Ceiite
6mL Toluene
HRGC/HRMS
OPTION D
1 mL Benzene Extract
4.0g
- SiO2
SiO2
l.Og
Concentrate to O.SmL
I
Acidic AI2O3
o.Og
30 ml 20% CH2Cl2/Hexane
Carbopak C/Celite
6mL Toluene
HRGC/HRMS
Figure 2. Column cleanup procedures proposed by the EMSL-LV.
16
564
-------
The TCDD isomers were added to 1-mL portions of benzene and were taken
through the four sample cleanup sequences depicted in Figures 1 and 2. One
of the replicates for each procedure was also spiked with 100 ng of Aroclor
1260.
The results of the sample analyses are provided in Tables 6 through 9.
As noted in Tables 6 and 7, recoveries of the TCDD isomers were low and
quite variable for the early eluting isomers 1,3,6,8- and 1,3,7,9-TCDD as
compared to 1,2,8,9-TCDD. Recovery of 1,2,8,9-TCDD was still low and vari-
able (approximately 60 percent recovery with an RSD of -v 20 percent). These
results were generated using the procedures specified in the original proto-
col (see Figure 1). The results of the analyses following the cleanup op-
tions A and B demonstrate that accurate quantitation of all TCDD isomers is
not possible using only the 13C12-2,3,7,8-TCDD surrogate standard. The low
recoveries measured for options A and B are obviously a result of the pres-
ence of benzene in the eluent from the acid-modified silica column that is
taken directly through the acidic alumina column.
In contrast, options C and D (Tables 8 and 9) demonstrate quantitative
recovery of the TCDD isomers. Some background contamination has been noted
from the acidic alumina for the 1,3,6,8- and 1,3,7,9-TCDD isomers. This
material had previously been prepared by Soxhlet extraction with methylene
chloride and activation at 190°C prior to use. As noted in Tables 8 and 9,
the average recovery of the other spiked TCDD isomers was greater than 84
percent.
When the recoveries of the different isomers and the 13C12-2,3,7,8-TCDD
are compared, the average relative percent difference ranges from 1 percent
for 2,3,7,8-TCDD (Table 3) to 24 percent for 1,2,3,4-TCDD (Table 4). These
results demonstrate that either of these cleanup procedures (options C and
D) will provide good recovery and reliable quantitation of 2,3,7,8-TCDD and
very good estimates of the concentrations of the other TCDD isomers present
in the samples. No interferences were observed in the samples spiked with
100 ng Aroclor 1260. The lack of PCB interferences was especially noted in
the extracts of samples spiked at 0.10 ng/TCDD isomer.
In addition to the evaluations of the cleanup procedures presented above,
the acid-modified silica gel/acidic alumina columns and the Carbopak C/Celite
column were evaluated separately. Evaluation of the silica/alumina at the
0.10-ng spike level as shown in Figure 2 resulted in an average recovery of
120 percent for 1,2,3,4-, 1,2,3,7-, 1,2,3,8-, and 1,4,7,8-TCDD; 114 percent
for 2,3,7,8-TCDD; 118 percent for 13C12-2,3,7,8-TCDD; and 118 percent for
1,2,8,9-TCDD. The results for the recovery of 1,3,6,8- and 1,3,7,9-TCDD
indicated that some contamination originated from the acidic alumina.
Replicate analyses of the Carbopak C/Celite column at the 0.10-ng spike
level resulted in average recoveries of 97 percent for 1,3,6,8-TCDD; 88 per-
cent for 1,3,7,9-TCDD; 81 percent for 1,2,3,4-, 1,2,3,7-, 1,2,3,8-, and
1,4,7,8-TCDD; 75 percent for 2,3,7,8-TCDD; 96 percent for 13C12-2,3,7,8-TCDD;
and 90 percent for 1,2,8,9-TCDD. Elution of the Carbopak C/Celite column
with additional toluene beyond 6 mL did not improve recoveries even for the
samples spiked at 10 ng/TCDD isomer.
17
r~ f ~~
ubo
-------
TABLE 6. KECOVEKY (%) Of SEVERAL TCDO I.SOUtUS HiOH CI.EANUi' OPTION A
00
CT
c:
Recovery (%) ot
Spike
level
InL Bcfutnc Ealrocl
1
H2SO4 - SIO2
4.09
Si O2
I.OR
1
Acidic AljOj
10ml 20%
Concentrate to 100/iL
1
Cwbopok C/C.liu
6mLlolu«n
HRGC/HRMS
1 ng
1 ng
1 ng
10 ng
10 ng
ChjC.^...™, ^ ^
Mean
% RSI)
1,3,6.8 1,3,7,9
4.6 9.6
12 21
3.1 9.3
12 19
6.2 12
6.3 12
51 36
i.'Ols
21
40
28
41
29
31
32
24
1.2.3.7/
1,2,3.8
14
33
17
31
20
20
23
34
TCDU isomer
2.3,7,8
19
39
23
44
31
29
31
31
13C12-2,3,7.8
27
48
31
.38
36
34
36
20
1,2,8,'J
38
64
40
76
61
59
56
26
''Sample was also spiktMl will] 100 ng ol Arorlor 1260.
-------
TABLE 7. RECOVERY (%) Of SEVERAL TCDD 1SOHEKS FROM CLEANUP OPTION B
cn
cr;
_ _ _ .__ ____ __. ___ , _
Spike
level
l«Lft«n**nt Cjitiocl
1
HjSCU-SIOj
4.0,
SI02
I.Og
Acidic AI}O3
«.0g
1 ng
1 ng
1 ng
10 ng
10 ng
30«L 20% CH2Cl2/H.»on.
10 nga
Corbopol. C/C.IIlt
6nLUIixni
Mean
HRGC/HRMS
% BSD
1,3.6,8
7.6
2.4
6.4
O.'J
1.8
11
5.0
79
1,3.7.9
15
21
25
7.6
11
17
16
40
Recovery
1.2.3.4/ 1
1.4,7,8 1
35
31
34
14
36
44
32
31
(%) of
,2,3,7/
.2,3.8
14
17
22
21
25
31
22
28
TCOD isoiner
2.3,7.8
37
34
40
50
47
59
45
21
»3C12-2,3,7,8
38
40
41
44
48
52
44
12
1,2,8,9
48
52
57
70
59
78
61
19
Sample was also spiked with 100 ng of Aroclor 1260.
-------
TABLE 8. RECOVER* (%) OF SEVERAL TCDO ISOMEKS FKOH CLEANUP OPTION C
cn
a:
Cc
ro
O
Recovery (%) of TCDO isomer
1 mi ft*fu*m Extract
1
HjS04 - SiOj
4.0g
SIOJ
I.Og
, 1
Concentrate to 0. 1 »L
1
Acidic AljOj
«.08
1 30.1 30%
Canc
-------
TABLE 9. RECOVERY (%) OF SEVERAL TCDD ISOMERS FROM CLEANUP OPTION 0
cn
'
\ ml ft*n>«ni Extract
1
MjSC>4 - SiOj
4.0,
SiOz
1.0,
, 1
Cencinliiili to O.inL
1
Acidic AI2Oj
6.09
1 30ml 30%
Cwfaepol C/C.m.
6 ml Toluc
HRGC/HRMS
Spike
level 1
0.10 ng
0.10 ug
0.10 ngd
1.0 ng
1.0 ng
1.0 ngd
10.0 ng
CHjOzAu.on. 10.0 ng
10.0 ngd
K
Mean
% RSD
,3,6, 8a
147
290
260
90
180
135
81
114
126
1S8
46
l,3,7,9a
197
200
360
93
IBS
158
107
72
138
168
51
Recovery
1,2, 3,4/ 1
1,4,7,8 1
113b
,2l"
110b
106
95
85
50
105
89
97
21
(%) of TCDU
,2,3,7/
,2,3,8 2,3
c
c
c
103
121
79
104
108
95
101
14
isomer
,7,8
72
71
84
85
109
47
101
107
91
85
23
»3Cl2-2,3,7,8
84
90
97
86
90
60
76
82
92
84
13
1,2,8,9
70
71
e
53
108
63
84
118
112
85
29
The 1,3,6,8- ;m.1 1,3,7,9-TCOD isomers were also noted iti reagent blanks from Lhe acidic alumina
column. No such interferences were noted from the acidified silica gel or tlie Carbopak C/Celite
.column.
Resolution of 1,2,3,4-, 1,2,3,7-/1,2,3,8-, and 1,4,7,8-TCDD was not achieved. This value
represents recovery of the four isomers.
^Recovery reported with 1,2,3,4-/l,4,7,8-TCUU.
Sample was also spiked with 100 ng of Aroclor-1260.
IIKGC/IIRMS analysis was interrupted prior to the elution of this isomer.
-------
Three additional experiments were completed to evaluate the efficiency
of reverse elution of the carbon column. The Carbopak C/Celite was placed
in a 5-mL disposable pipette packed at both ends with glass wool plugs. The
column was eluted in one direction for the hexane, cyclohexane/methylene
chloride, and the methylene chloride/methanol/benzene mixture. The column
was then turned over and eluted with 6 ml toluene. Triplicate analyses at
the 0.10 ng/TCDD isomer spike level demonstrated average recoveries of 98
percent for 1,3,6,8-TCDD; 91 percent for 1,3,7,9-TCDD; 104 percent for
1,2,3,4-, 1,2,3,7-, 1,2.3,8-, and 1,4,7,8-TCDD; 116 percent for 2,3,7,8-
TCDD; 102 percent for 13C12-2,3,7,8-TCDD; and 93 percent for 1,2,8,9-TCDD.
FINAL METHOD EVALUATION
Based on the results of the column evaluation study, the analytical
method was revised to specify the cleanup procedure presented as Option D in
Figure 2. The final protocol, as presented in Appendix A, was then evalu-
ated as described below.
The data presented in Tables 10 through 17 are summaries of the initial
column calibration, HRGC and HRMS resolution checks, and the results of the
sample analysis.
Calibration
Table 10 summarizes the RRF data for the concentration calibration
standards from the initial calibration and the routine monitoring of the RRF
values over the time required to complete the sample analyses. The RRF(I)
as specified in the protocol is a measure of the response of 2,3,7,8-TCDD
versus the internal standard, l3C12-2,3,7,8-TCDD. The value for RRF(I) var-
ied ±9.4 percent over the five concentration levels of 2,3,7,8-TCDD ranging
from 2.5 pg/Ml- to 40 pg/pL. The RRF(II) is used to calculate the absolute
recovery of the internal standard as compared to the recovery standard
13C12-1,2,3,4-TCDD. The average RRF(II) was determined to vary by ± 19.3
percent over the calibration curve. The variability of the RRF (I) and
RKF(II) were determined to be less than ± 10 percent and ± 18 percent,
respectively, over all data points required to complete the sample analysis.
In addition to the analysis of calibration standards specified in the
protocol, solution HRCC6 was analyzed in triplicate to determine the lower
limit of sensitivity (1 pg/|jL). Although the calculated RRF(I) and RRF(II)
values and the S/N are within the specified criteria, the ion ratio for the
native compound and recovery standard indicate that these measurements fall
outside the acceptable calibration window.
22
r" r^ i\
u f(J
-------
TABLE 10. INITIAL CALIBRATION SUMMARY
ro
CO
Calibration
standard
IIRCC1
HRCCI
HRCC1
1WCC2
HRCC2
1IRCC2
HRCC2
HRCC3
I1RCC3
HRCC3
HRCC4
HRCC4
HRCC4
HRCC5
HRCC5
HRCC5
Date
09/12/8S
09/12/85
09/12/85
09/12/85
09/12/85
09/12/85
09/12/85
09/13/85
09/13/85
09/13/85
09/13/85
09/13/85
09/13/85
09/13/85
09/13/85
09/13/85
Time
09:05
09:44
12:31
13:00
13:27
13:53
15:39
10:31
10:57
11:23
13:02
13:29
13:56
14:22
14:49
15:15
m/z 320/322
0.82
0.84
0.84
0.73
0.80
0.92
0.78
0.78
0.78
0.73
0.77
0.73
0.77
0.78
0.7S
0.76
ra/z 332/334 (IS)
0.77
0.80
0.73
0.77
0.73
0.66
0.77
0.79
0.80
0.83
0.80
0.76
0.78
0.80
0.82
0.78
m/z 332/32
0.80
0.73
0.87
0.83
0.81
0.69
0.73
0.79
0.78
0.78
0.76
0.78
0.76
0.78
0.83
0.79
IIRCC6
HRCC6
HRCC6
HRCCI
HRCCI
HRCC2
HRCC2
IIRCC2
IIRCC2
HRCC2
IIRCC2
HRCC2
HRCC2
HRCC2
09/16/85
09/16/85
09/16/R5
09/16/85
09/16/85
09/20/85
09/23/85
09/23/85
09/24/85
09/25/85
09/26/85
09/27/85
09/30/85
10/03/85
10:43
11:19
13:44
12:42
14:40
10:46
08:47
10:46
10:47
08:39
08:56
09:33
09:19
08:53
1.32
1.18
0.86
0.87
0.83
0.86
0.82
0.88
0.89
0.7C
0.77
0.78
0.83
0.68
0.71
0.83
0.80
0.83
0.79
0.73
0.80
0.80
0:69
0.78
0.80
0.84
0.80
0.75
0.84
1.14
1.01
0.75
0.85
0.80
0.70
0.83
0.69
0.89
0.72
0.81
0.85
0.81
(RS) S/N 259 S/N 322 S/N 334(15)
41:1 > 65:1 > 65:1
48:1 > 65:1 > 65:1
24:1 > 65:1 > 65:1
36:1 > 65:1 > 65:1
58:1 > 65:1 > 65:1
78:1 > 65:1 > 65:1
76:1 > 65:1 > 65:1
97:1 > 65:1 > 65:1
111:1 > 65:1 > 65:1
110:1 > 65:1 > 65:1
96:1 > 65:1 > 65:1
> 144:1 > 65:1 > 65:1
> 144:1 > 65:1 > 65:1
> 144:1 > 65:1 > 65:1
> 144:1 > 65:1 > 65:1
> 144.1 > 65:1 > 65:1
Overall Hean
10: 21: > 65:1
9.6: 25: > 65:1
18: 30: > 65:1
36: 63: > 65:1
48: > 63: > 65:1
> 75: > 63: > 63:1
> 75: > 63: > 63:1
42: > 63: > 63:1
26: > 63: > 63:1
58: > 63: > 63:1
73: > 63: > 63:1
49: > 63: > 63:1
73: > 63: > 63:1
29:1 > 63:1 > 63:1
Mean
X RSD
Mean
X RSD
Mean
X RSD
Mean
X RSD
Hean
X RSD
(RRF)
X USD
RRF(l)
0.783
0.794
0.750
0.776
2.9X
0.829
0.853
0.7993
0.861
0.848
2. OX
0.974
0.965
0.972
0.970
0.5X
0.945
0.935
0.967
0.949
1-7X
0.964
1.01
0.989
0.987
2.2X
0.906
9.4X
0.917
0.878
0.935
0.876
0.850
0.835
0.832
0.941
1.01
0.949
0.941
1.04
0.854
0.955
RRF(II)
2.18
2.27
2.28
2.24
2.4X
1.76
1.93
2.033
2.27
1.99
13. OX
1.53
1.57
1.58
1.56
1.9X
1.49
1.52
1.49
1.50
1.1X
1.52
1.42
1.43
1.46
3.9X
1.75
19. 3X
1.44
1.09
1.58
1.86
2.11
1.98
2.65b
1.41
1.68
1.96
1.57
1.68
1.67
1.53
.Nol included in mean RRF computation.
Not within allowable limits for routine calibration.
-------
TABLE 11. HRGC AND MASS RESOLUTION CHECK SUMMARY
Date
9/12/85
9/12/85
9/12/85
9/12/85
9/13/85
9/13/85
9/13/85
9/13/85
9/16/85
9/16/85
9/16/85
9/16/85
9/17/85
9/17/85
9/18/85
9/18/85
9/19/85
9/19/85
9/20/85
9/20/85
9/20/85
9/20/85
9/23/85
9/23/85
9/23/85
9/23/85
9/24/85
9/24/85
9/24/85
9/24/85
last.
ID
MS50
HS50
MSSO
MS50
MS50
MSSO
MS50
MS50
HSSO
MS50
MS50
MS50
MS50
MS50
MS50
MS50
MSSO
MS50
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
Sol.
ID
.
PC
PC
-
-
PC
PC
-
-
PC
PC
-
PC
-
-
PC
-
PC
-
PC
PC
-
- '
PC
PC
-
PC
PC
Time
07:51
08:32
16:05
16:47
08:26
08:45
15:48
16:23
09:25
10:45
15:16
15:57
10:26
10:14
07:58
08:18
11:14
12:56
08:00
08:16
15:44
16:14
07:55
08:15
16:01
16:45
09:51
10:15
16:01
16:33
TCDD isomer
File resolution
name (% valley)
MID254I12X1
8367I12XQ1 5.9
8367I12XQ9 2.9
MID254I12X2
MID254I13X1
8367I13XQ1 6.9
83671 13XQ12 11.4
MID254I13X2
MID254I16X1
8367I16XQ1 11.9
83671 16XQ8 23.0
MID254I16X2
8367I17XQ1 13.3
MID254I17X1
MID254I18X1
83671 18XQ1 20
MID254I19X1
8367I19XQ1 3.5
MID254I20X1
8367I20XQ1 6.7
8367I20XQ5 4.1
MID254I20X3
MID254I23X1
8367I23XQ1 8.8
8367I23XQ6 12.5
Manual check3
HID254I24X1
8367I24XQ1 12.2
8367I24XQ3 13.1
HID254I24X2
Mass
resolution
at 10% valley
10,774
-
-
10,450
10,230
-
-
10,384
10,294
-
-
10,388
-
•10,824
11,019
-
11,679
-
12,068
-
-
10,777
10,096
-
12,500
10,374
-
-
10,567
Mass
measurement
error
5 ppm
-
-
-
0 ppm
-
-
-
4 ppm
-
-
- .
-
2 ppm
1 ppm
-
4 ppm
-
3 ppm
-
-
-
1 ppm
-
~
3 ppm
-
-
A manual resolution check was performed due to data system failure.
(continued)
24
-------
TABLE 11. (continued)
Date
9/25/85
9/25/85
9/25/85
9/25/85
9/26/85
9/26/85
9/26/85
9/26/85
9/27/85
9/27/85
9/27/85
9/27/85
9/30/85
9/30/85
9/30/85
9/30/85
10/3/85
10/3/85
10/3/85
10/3/85
Inst.
ID
MSSO
MS50
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
MSSO
Sol.
ID
-
PC
PC
-
-
PC
PC
-
-
PC
PC
-
-
PC
PC
-
-
PC
PC
'
Time
07:50
08:05
16:13
16:45
08:07
08:21
15:49
16:21
08:19
09:02
16:01
16:29
08:15
08:33
15:10
15:41
07:59
08:20
15:56
16:29
File
name
MID254I25X1
8367I25XQ1
8367I25XQ3
MID254I25X2
MID254I26X1
8367I26XQ1
8367I26XQ3
MID254I26X2
MID254I27X1
8367I27XQ1
8367I27XQ3
MID254I27X2
MID254I30X1
8367I30XQ1
8367I30XQ3
MID254I30X2
MID254J03X1
8367J03XQ1
8367J03XQ3
MID254J03X2
TCDD isoraer
resolution
(% valley)
-
11.1
6.5
-
-
8.3
13.2
-
-
11.9
11.8
-
-
< 25
< 25
-
-
17
12
"
Mass
resolution
at 10% valley
11,165
-
-
11,419
10,989
-
-
10,499
11,564
-
-
10,639
11,149
-
-
11,321
10,567
-
-
10,442
Mass
measurement
error
0 ppm
-
-
-
3 ppm
-
-
-
1 ppm
-'
-
-
5 ppm
-
-
-
0 ppm
-
-
'
A manual resolution check was performed due to data system failure.
25
-------
TABU II. TCDD DATA RWORT n»l
tactymuia AMI..?
ro
. Aliauot
Settle Air-dry wt. (|)
•o. or Vol. (L)
IM7-83-I576I-DW 1.0 L
1367-82-13761-DV 1.0 L
1367-19-13761-nM 10 L
1167-11-15761-tW 1.0 I
oM7-90-137oI-[IM 1.0 L
1367-05-I576X-IH) 1.0 L
1367-92-I576X-IWW 1.0 L
1367-14-1576X-DWN 1 0 L
?Aqueoue aeaple data reported ae pp<
'Criteria for poaltlve Identlflcatl
laoaer rould not fae identified.
ran
lao*er
2,3,7,1-
1,1,6,1-
1,1.7,9-
2.1.7,1-
1.1.6,1-
1.1.7,9-
c
2.1.7.1-
c
1,1.7.1-
1,3,6,8-
1,3,7,9-
,2,3,7/
.2.3,8-
,1,1,4-
.1.7,1-
,2.1,9-
1.3,7.1-
1,1,7,1-
1,3,6,1-
1,1,7,9-
1.1,7,1-
1,1.7,1-
1.3,6.8-
1, ,7,9-
1. ,3.7/
l! !l|4-
1. ,7,1-
1.1,1.9-
and aoil
o require
Retentio
TCM) "C|
11:15
16:4}
17:54
21:40
16:41
17:51
24:10
21:29
23:37
11:59
17:02
11:14
22:11
22:31
14:30
30:01
21:55
22:01
17:06
11:16
24:31
24:31
17:06
11:11
21:21
22:15
24:12
30:02
aaeople data
that the lo
TCDD (ppt
n tisje or peq) listr.
1-2,3,7,8 Heal. Dl ID ' Date Ti«e
21:14 221 - IG50 09/20/15 11:27
- 117
16 5 -
21:17 2,277 - HS50 09/20/15 14:41
134
212
137
21:27 1.090 - MS50 09/20/15 15:11
75.9 -
21:56 1,010 - IG50 09/23/15 11:17
502
766 -
1,160
1,140
1,430
1.330
11:55 1.290 - IG50 09/23/15 12:51
21:59 501 - 1650 09/23/13 13:29
191
201
55.2 -
'516
512
395
403
616
140
321
presented aa ppt.
l ratio! fall between 0.67 and 0.90.
(continued)
Relative loa Abundance Ralloab
320/321 111/114(1!) 111/114(RS) 1 Rec.
0.71 0.63 0.61 40
O.i9
0.10
0.51 0.69 0.72 14
0.79 0.72 0.71 96
1.02
0.14
0.90
0.77 0.71 0.71 61
0.19
0.74 0.72 0.12 91
0.15
0.13
0.11
0.71
0.71
0.10
0.11 0.90 0.11 23
0.74 0.10 0.71 75
0.15
0.75
0.19
0.71
0.10
0.74
0.72
0.72
0.10
0.79
•It 259
73:1
64:1
41:1
21:1
10:1
11:1
6:1
146:1
7:1
27:1
22:1
30:1
54:1
45:1
71:1
22:1
37:1
49:1
25:1
27:1
1:1
49: 1
21:1
23' 1
73:1
53:1
42:1
61.1
70:1
23:1
5/1
•/t 32} Wl 334(IS) Comenta
>63:l >49:l Ratio unacceptable. Rerun.
50:1
10:1
> 63:1 > 63:1 Saaa>le aplked at twice requested level.
> 3:1 Rerun.
> 11:1
63:1
> 63:1 > 63:1
11:1
34:1 > 63:1
11:1
33:1
> 61:1
60:1
> 63:1
11:1
63:1 10:1 Low recovery.
63:1 > 63:1
19:1
29:1
16:1
63:1 23:1 Low recovery.
31:1
> 63: 1
63:1
50:1
45:1
55:1
63:1
20:1
-------
ro
01
^,7
Cl
Saajple A
Ho
8367-9I-I576X-IW
8367-66-15761-iHw
8367-93-I576X-IMI
8367-70- 15 J6l-HI»
8367-65- I576K-15
8367-66-I576X-85D
B367-67-lS76X-»S»
'Aqueous aaav>le data
Aliquot
it-dry «t. (|) ICDO
or Vol. (L) IIOK
1.0 I 2.3.7,
1.3,6,
1.3.7,
1,2.7,
1.0 L 2,3,7,
1.0 L 2,3.7,
1.3.6.
1.3,7.
1.2.3,
1.2,3,
l.J.3.
1.2,7,
1,2,1.
10.01 | 2,3,7,
1.3.6.
1.3.'.
1,2.3,
1,2,3,
1.2.3,
1.2,'.
1.2.1,
10.00 | 2,3,7,
1.3,7,
9 15 i 2,3,7,
1.3,7.
10.00 | 2.3,7.
1.3,6.
1.3,7,
1.2.3,
1.2.3.
1,2,3,
1.2.7,
1 .2.8,
Reteo
TCDD '
- 22:00
• 17:03
- 18:14
- 24:10
- 21:50
- 21:39
- 16:47
- 17:57
/ 21:58
- 21:11
- 21:01
- 29:35
- 21:38
- 16:45
- 17:55
/ 21:58
- 22:10
- 24:09
- 29:36
- 2I:*7
- 18:05
- 21-44
- 18:03
- 21:47
- 16:55
- 18:05
/ 22:07
- 22:19
- 24:16
- 29:45
TABLE 12. I
HRGC/IOUIS
tCDO (ppt
tioo lite or fpqj Initr.
ICll-2,3.7.8 Mea>. DL ID Dale Tix
21:57 534 - MSSO 09/23/85 15:01
246
122
54.7 -
21:48 1,430 - HSSO 09/23/85 15:31
Jl:38 530 - I1S50 09/24/85 11:19
582
690
940
1,180
1 . 790
691
21:37 30.3 - ItSSO 09/24/65 12:55
29.0 -
Sl.l -
I2S
118
252
100
21:45 18.2 - IB50 09/24/85 13:27
8.5 -
21-43 15.1 - HS50 09/24/85 13:56
4.2 -
21:46 12.9 - KSSO 09/24/85 14:29
NO 9.2
15.2 -
61.8 -
54.1 -
147
63.9 -
ion ralloa f«ll between 0.67 and 0.90.
continued)
Analyiii
Relative Ion Abundance Ratio!
320/322 332/334(15) 332/334OS) I Kec.
0.87 0.83 0.81 77
0.71
0.78
0.70
0.80 0.76 0.79 29
0.76 0.72 0.73 71
0.82
0.80
0.72
0.79
0.82
0.77
0 85 0.78 0.77 56
0.63
0.87
0.61
0.69
0.61
0.85
0.87 0.78 0.75 73
0.85
0.67 080 0.72 85
0.87
0.66 0.73 0.71 46
0.59
0.67
0.77
0.90
0.75
0.83
•/z 2S9
72: 1
47:1
36:1
7:1
49:1
23:1
42:1
43:1
49:1
50:1
72:1
22:1
12:1
15:1
24:1
46:1
38:1
73:1
20.1
12:1
9.4:1
18:1
6.3:1
9:1
11:1
14:1
4R:1
35: 1
73:1
26:1
- -
S/«
•/z 322 ml I 334(15) Coraeoli
> 62:1 > 63:1
48:1
32:1
16:1
63:1 > 33:1 Lou recovery.
4:1 > 63:1
4 :l
6 :l
4 :1
6 :l
6 :l
2 :!
15:1 > 63:1
26:1
33:1
63:1
58:1
> 63:1
23:1
42.1 > 63:1
25:1
11:1 > 63 : 1
12:1
16:1 > 63:1
15: 1 Ratio unacceptable.
20:1
63:1
58:1
63:1
22:1
:Uo.»er could not be identified
-------
TMU II. (continued)
Saiple
He.
1167-6I-IS76»»I
I167-69-IS76I-IID
l]67-7l-IS76l-«l
I167-72-IS76.-IID
IM7-7]-IS76i-ll»
1167-74-IS76I-H1
1367-77-IS76.-FA
Aliquot
Air-dry wt. (() TCDD
or Vol. .9-
c
1.2,7,1-
e
9 94 | i,),7,l-
1.1.6.1-
1.1.7,9-
Rcteotion tiae
TCDO 13C|i-2,3,7
2I:«2 2I-.40
16;4»
11:11 21:11
16:4!
17:52
21:17 2I:1S
16:44
17:54
19:04
21:1} 21:1]
16:42
17:53
21:41 21:41
I6:4«
11:59
19:09
22:01
22: IS
24:12
29:41
21:44 21:42
16:49
11:00
19:10
24:59
26:01
21:16 21:14
16:42
I7:S1
19:02
ICDD (ppt
or PM) ln»tt
1 He.i Ot 10
14.) - HSSO
4.S -
M.6 - »SO
5.1 -
10.0 •
117 - n»o
160
1IJ
SO. 6 -
71} - 1050
201 -
101
1,210 • IBSO
111
435
51
Sll
695
1,170
461
2.020 - IBSO
164
217
70.6 -
11.7 -
27.1 -
1.720 - KSV>
I.MO
1,150
1,250
ReUlivt lor, Abundance Ritioi^
Dit< Tl» 120/122 132/334(13) 112/114(13) X Ice.
Of/24/15 15:02 0.12 0.67 0.75 7]
O.t7
09/24/15 15:12 0.70 0.74 0.11 46
0.69
0.11
09/2S/IS 10:01 0.11 0.11 0.11 95
O.IS
0.14
0.69
09/25/15 10:10 0.71 0.15 0.10 75
0.11
0.12
0,65
09/25/15 11:!7 0.77 0.11 0.11 10
0.77
O.IS
0.11
0.72
0.71
0.12
0.11
09/25/15 11:00 0.11 0.11 0.11 79
0.16
0.11
0.74
0.92
0.61
09/25/15 11:10 0.10 0.12 0.12 4
0.11
0.11
0.10
SIX
i7Tl59 •/! 122 mil liUTsT CoKntl
29:
II:
19:
II:
15:
97:
14:
44:
7.1:
71:
16:
40:
6.S:
97:
47:
79:
6.7:
57:
SI:
17:
21:
> 145:
22:
24:
9:
2.5:
1:
67:
109:
94:
60.
61: > 61:1
14:
31.
5:
10:
61:
14:
24:
4:
63:
22:
21:
21:
15:
4:
27:
29:
61:
29:
> 63:
>
> 1 :
> ;
21:
31:
14:
22:
> 54:1
> 63:1
> 61:1
> 63:1
> 61:1
6: 1 Low recovery.
.
r\>
00
oui tuple d«t» reported •• ppq a
er could not be Identified.
oil aaa^le data presented at ppt.
:>een 0.67 and 0.90
(continued)
-------
cn
•vl
ro
10
Aliquot
Stifle Air-dry « . (|) ICDD
1367-77-1576X-FA c
(concluded) c
c
I.2.3.7/
1,2,3,1-
1.2,3.4-
c
1,2,7.1-
c
c
c
c
c
1367-94-I576X-H2V 1.0 L 2,3,7,8-
1.3.6,8-
1,3,7,9-
c
c
c
1.2.7,1-
c
c
c
c
•367-13- 1576X-OM1 1.0 L 2,3.7, -
1,3,6, -
1,3,7. -
c
1.3,6. -
1.3,7. -
^Aqueoue •••pie d«ta reported •• ppq ind toil
Retention 11
20:00
20.13
21:25
21:5*
22:0>
22:35
24:05
24: 49
25:23
25:55
27:11
28:07
21:42 21:
16:47
17.51
19:01
20:06
20:49
24:13
24:55
26:01
27:25
21:11
21:11 21:
16:25
17:14
21:39
It 22
17:12
•Japlr dill pre
TCDD (ppt
•e or ppq)
1,220
27*
109
675
4,660
179
720
3,460
155
3,430
441
2,920
41 »C
IK
K
1C
K
K
»C
«C
•C
DC
HC
10 265
m 167
125
50.1 -
140
106
looted <| ppt.
TABU 12. (continued)
Instr. ReUtive loo Abundto.ce fUtloi
HS50 09/25/15 13:30 0.14
0.75
0.77
0.10
0.11
0.18
0.11
0.10
0.75
0.71
0.79
0.61
IIS50 09/25/15 15:11 0.16 1.3 0.12 ND
0.11
0.14
0.79
0.98
0.68
0.75
0.80
0.72
0.82
0.81
HS50 09/26/85 09:56 0.79 0.73 0.70 42
0.91
0.71
0.71
0.81
0.96
0.67 >nd 0.90.
44
16
6
32
146
31
23
80
7
16
14
68
> 154:
IB:
12:
20
6
}
3
17
63
9
30
11
11:
14:
1.2:
6.3:
3.»
S/
1 :
:
2. :
1 :
6 :
1 :
1 :
5 :
6: :
4 :
> 65:
55:
40:
61:
20:
11:
3:
35:
63:
8:
25:
31:
24:
21:
7:
12-
11:
4:1 332/334(13) Ratio unacceptable; no jeount
co«putationa performed.
-
-
47:1
Ratio unacceptable.
*r could not be identified.
-------
CJ
o
en
Aliquot
Saaple Air-dry vt. (|)
•o. or Vol. (t)
I16I-U-IS76X-IW 1.0 I
I167-IS-IS7II-I2H) 1.0 I
I16I-M-IS7U-UVI 1.0 L
TCDO Retention tieje
leoe*r ICDO "C, ,-2, 1,7.1
.1,1,1- 21:11 21:15
.M.I- 16:11
.1,7.1- I7:M
.1.7,1- 11:44
,1.1,1- 21:11 11:11
.1,6.1- I6:2S
.1,1.1- 17:35
11:41
11:31
20:10
11:11
,1.1,1- 21:11
24:21
25:21
U:4I
21:11
21:14
10:11
11:01
.1.1.1- 11:14 11:11
,3,8,1- 16:16
,1,7,1- 17:15
19:41
10:12
11:11
,1,7,1- 11:40
24:14
15:30
16:51
27:31
10:14
11:11
Tn»
«... Dl
,010
III
111
III
K
•C
K
K
K
K
K
K
K
K
K
K
1C
K
1C
•C
K
K
K
K
K
K
K
K
•C
•C
•c
K
•C
K
TAIL! 11. (continued)
•KGC/MmS Aoalnla
laatr. lelatlvt Inn Abundance latloi*
ID Dale Tle» 110/111 1)2/114(15) 1)1/1)4(13) 1 dec.
IBSO 01/16/IS 12:32 0.76 0.71 O.IS 66
0.11
0.41
0 76
IO50 01/16/15 11:0} O.I) I).) 0.67 1C
0.17
0.11
0.11
0.17
0.71
014
0.70
0.76
0.76
0.11
0.11
0.11
a.n
O.SI
IBSO 01/16/IS 14:21 0.11 1.47 0.17 K
.71
.71
.71
.11
.61
.S|
.11
.71
.75
.76
.70
.77
.14
.11
•/a 2S
11:
11:
II:
6:
> 146:
74:
41:
46:
IS:
1:
S:
S:
S4:
71:
II:
16:
7:
10:
4:
> I4S:
40:
11:
14:
II:
1:
4:
S:
SI:
11:
II:
11:
1:
11:
1:
s
fit 32
> 61:
> 10:
> IS:
11:
> 6):
> 26:
> 14:
> 11:
> S:
> 4:
> 1:
l.S:
19:
6):
1:
IS:
S:
IS:
l.S:
> 61:
11:
IS:
11:
6:
S:
6:
1:
41:
61:
10:
21:
6:
20:
25:
n
•/a 134(11) Coeatenta
> 63:1
.
.
•
11:1 111/1)4(1$) latin unacceptable) nn Mount
conputationa perfortwd.
.
.
-
-
•
•
•
.
-
.
.
1:1 111/1)4(18) Ratio unacceptable; no aojoojt
coatiuletiou perforaN*.
-
-
-
•
•atio unacceptable.
Katie unacceptable.
-
-
•
-
.
-
Ratio unacceptable.
Criteria for p«alti*e identification require that the ion ratioa (all between 0.67 and 0.10.
laoe»r coeild not be ideal 11 led
(continued)
-------
C1
-v!
Aliquot
Sample Air-drj M. (|) TCDD
Ho. or Vol. (1) lao<*r
8367-75-I576X-H3D 1.16 |
B367-76-I576-X-H3K 1.14 |
8367-78- I576X-FAD 10.04 |
8367-99-1576X-FA1C 9.93 |
,3,7,1-
.3.6.1-
.3,7,9-
.3.7,1-
.3,6.8-
,3.7.9-
,2.3,7/
.2,3,8-
.2,3,4-
,2,7,8-
.2,1,9-
.3,7,1-
,3,6.8-
,3,7,9-
,2.3.'/
,2,1.1-
,2.1.4-
.3.7.8-
.3.6,8-
,3,7,9-
,2.3,7/
,2,3.1-
.2,3.4-
a
Criteria for poaitive identification require
laoKr could not be identified.
Retention time
TCDD "C,,-2,3.7
21:20 21:19
16:31
17:41
11:49
21:39 21:39
16:45
17:56
21:51
22:11
24:10
29:36
21:33 21:30
16:41
17:51
19:01
19:57
20:41
21:52
22:05
22:33
24:01
24:45
25:51
27:14
28:03
30:01
21:39 21:31
16:45
17:56
19:06
20:01
20:47
2I:S8
22:11
that the ion ratios
TABLE
KRG1
TCDD (ppt
" PPQ) ln«tr.
8 Heaa. DL ID Date
2,260 - HS50 09/27/85
116
163
1,800 - HS50 09/30/85
383
367
825
B55
2,310
952
1,020 - HS50 09/30/85
926
747
610
557
146
286
2,260
KD 329
356
3,520
3,680
558
3,120
3.270
1,160 - HSSO 09/30/85
1.390
1.160
881
818
194
423
3.620
fall between 0.67 and 0.90.
1 continued)
12. (continued)
fTJjjjjjS Analyaia
Relative Ion Abundance Ratioafa
Tile 120/322 312/334(15) 332/33KHS)
15:31 0.79 0.88 0.80
0.90
0.67
09:57 0.80 0.79 0.80
0.80
O.BI
O.BI
0.13
0.15
0.77
10:29 0.73 0.71 O.M
0.77
0.80
0.17
0.19
0.71
0.67
0.79
0.91
0.74
0.71
0.79
0.11
0.76
0.79
13:00 0.80 0.84 0.88
0.78
0.79
0.85
0.74
0.76
0.80
0.86
- ... ... . . -- . .- ...
,» / 25, ,'tf 1 334..S) C
99 > 145: • 63: > 63:1
14:
20
6
86 > 143
44
41
79
60
145
46
6.7 45
55
42
32
23
8.7
17
75
16
13
92
97
17
80
83
5 56:
94:
80.
60:
35:
13:
29:
131 :
32-
' 5:
> 7:
> 2.5
> 63:
> 15:
> 17:
> 29:
•• 24:
63:
23:
32:
37:
25:
22:
14:
6:
10:
63:
11:
8:
56:
63:
10:
48:
47:
25:
40:
Ji-
ll:
17:
5:
10:
r,3:
10:
1
> 63:1
9:1 Recovery low.
Ratio unaccrptatble.
6: 1 Low recovery.
-------
TAIU 12. (concluded)
10
SMVlo Air-dry «*. (|)
•o. or Vol. (it
TCDO (t.l.otioo tin
lico*r TCOO "C, i-l.l, 7
H67-tt-l376I-r«l 22:M
(coocltJed)
6M7-IOO-I376I-OM> 1.0 L
•167-I02-I376I-IB 300 mi
6167-IOI-I376I-DM) 1.0 ml
B167-I01-I376X-IB) 300 ml
I167-I04-I}76I-UH 4N mi
•M7-I03-I376I-IW 610 mi
,1,7.1- 16:01
14:32
13:31
17:10
26:10
10:09
M:67
,1,7,1- 21:04 21:01
,1.6,6- 16: It
. 1.1,1- 17:26
,1,7,6- 21:02 21:01
,1,7,1- 20:34 20:33
,1,6,1- 16:14
,1.7,t- 17:21
. 1,7,6- 20:37 20:36
.1,6.1- 16:17
,l,J,t- 17:21
.1,7,1- 21:00 20:St
,1,6,1- 16:11
,1,7.1- 17:27
11:12
24:06
23:09
27:16
29:6*
,1,7,1- 20:36 20:36
.1,6,6- 16:13
,1.7,9- 17:22
11:10
24:0)
23:07
24:17
2t:43
or ppo.) loitr.
i He... 01 10
561 • IBIO
316
4,110
4.SM
n> 611
1,960
4,010
1,170
246 - IBM
617
tn
604 - IBM
1,030 - IBM
137
1*4
611 - IBM
D 41
17
17,100 - IBM
» 71
164 -
Ml -
ID 417
311 •
S7J •
111 -
21,100 - IB30
214 -
102 -
451 -
364 -
710 -
318
147 -
«eUti» loo Abiudioco lotloi*
0>t< Tl«e 120/121 111/114(18) 111/114(19)
09/10/13 11:00 .11
.62
.73
.71
.1$
.77
.86
•"
10/01/63 11:12 0.66 0.62 0.74
0.77
O.M
10/01/43 12:31 0.71 0.17 0.13
10/01/13 11:21 0.61 0.11 O.N
0.11
10/01/13 14:01 0.71 0.73 0.71
0.51
0.69
10/01/1] 14:11 0.74 0.14 0.74
1 OS
O.M
0.13
0.63
0.70 .
0.11
O.M
10/01/1} 11:11 O.M 0.14 0.71
0.13
O.M
0.71
0.60
0.11
0.10
0.79
S/N
I Ice. •/• 251 mil 122 •/• 114(18) CooBenti
1 12: 10:
20:
117:
146:
10:
14:
121:
26:
61.3 21:
71:
SI:
39.S 16:
6O 71:
17:
26:
37 11:
6:
I:
71 > 143:
9:
16:
»:
It:
12:
11:
II:
96 > 143:
16:
17:
10:
20:
11:
11:
10:
34:
61:
7:
46:
44:
16:
18:
61:
47:
11:
61:
11:
26:
42:
11:
17:
> 61:
II:
11:
S3:
21:
12:
26:
21:
> 61:
19:
21:
H:
14:
11:
14:
11: 10:
Kitio unacceptable.
> 61:1
> 61:1
> 61:1
> 61:1
> 61:1
title wucccptoblt.
•jtlo u»«rcepteble.
> 61:1
___ .
.Aooeove tmmflt mill reported •• ppo, tod loll •••pie dil« prevented •• ppt.
Criteria for polillee lorntiticitloe require tbil the ioo r.Uoi fill beMea 0.67 tod 0.10.
luo*r could >ot be Idratllled.
-------
TABLE 13.
ACCURACY AMD PRECISION OF THE HRGC/HRMS ANALYSIS FOR 2,3,7,8-TCDD
FROM LABORATORY AQUEOUS MATRIX SPIKES
cn
GO
co
CO
2,3,7,8-TCDD 2,3,7,8-TCDD
Sample matrix Spike level (ppq) Detected (ppq)
Distilled water (DW)
Effluent wastewater (EWW)
Influent wastewater (IWW)
Industrial wastewater (IND)
Industrial wastewater (IND)
Soil extract (H2W)
250
250
250
1,000
1,000
1,000
500 .
500
500
500
500
500
.
-
_
.
Average cone.
RPR
Average cone.
RPR
Average cone.
RPR
Average cone.
RPR
Average cone.
RPD
Average cone.
RPD
234
265
246
248
12.5
1,090, 1,030
1,010
1,050
1,050
7.6
534
508
530
524
5.0
1,290
1,520
1,430
1,410
16
604
628
616
3.9
27,100
28,100
27,600
3.6
2,3,7,8-TCDD
Recovery (1)
93.6
106
103
Average rec. 101
RPR 9.3
109, 103
101
105
Average rec. 105
RPR 7.6
107
102
106
Average rec. 105
RPR 4.8
258
304
286
Average rec. 283
RPD 16
-
-
.
-
I3C,2-2
Absolute
Average rec.
RPR
Average rec.
RPR
Average rec.
RPR
Average rec.
RPR
Average rec.
RPD
Average rec.
RPD
,3,7,8-TCDD
recovery (%)
82
42
69
64
63
61, 66
91
80
75
40
77
75
71
74
8.1
23
20
29
24
38
60
57
58
5.2
78
96
87
25
"Relative percent range (calculated from the difference of the high and low values divided by the average of all values and
.multiplied by 100 percent).
Relative percent difference.
-------
TABLE 14. PRECISION OF THE IIRGC/HRMS ANALYSIS FOR 2,3,7,8-TCDD
OF SOIL AND FLY ASH SAMPLES
C-7
U)
Sample matrix
B25-Piazza Road (B5)
Hyde Park 001 (HI)
B52-Shenandoah (Bl)
Hyde Park 003 (H3)
Fly ash - RRAI
Endogenous
2,3,7,8-TCDD 2,3,7,8-TCDD
level (ppt)a Detected (ppt)
50
Average cone.
RPR6
70
Average cone.
RPR
360
Average cone.
RPR
1,700
'Average cone.
RPR
.
Average cone.
RPR
18.2
15.1
12.9
15.4
36
34.3
36.6
30.3
33.7
19
937
785
1,280
1,000
50
2,020
2,260
1,800
2,030
23
1,720
1,020
1,160
1,300
54
13C,2-2,3,7,8-TCDD
Absolute recovery (X)
Average rec.
RPR
Average rec.
RPR
Average rec.
RPR
Average rec.
RPR
Average rec.
RPR
73
85
48
69
54
73
46
56
58
47
95
75
80
83
24
79
99
86
88
23
4
7
5
5.3
57
Estimated level of endogenous 2,3,7,8-TCDD reported to MRI by Dr. W. Beckert in letters dated
bApril 19, 1985, and August 30, 1985.
Relative percent range (calculated from the difference of the high and low values, divided by
the average of all values, and multiplied by 100 percent.
-------
TABLE 15. ACCURACY OF THE HRGC/HRMS METHOD FOR THE DETERMINATION OF TCDD 1SOHEKS SPIKED INTO AQUEOUS MATRICES
Effluent wastewater Distilled vater Influent wastewater Industrial wastewater
TCDD
analyte
1.3.6.8
1.3.7.9
1.2,3,7/1,2.3,8
1,2.3,4
1,2.7,8
1,2,8,9
2,3,7,8
l3CI2-2,3.7,8
Spike
(pg)
1,840
840
1,680
2,440
3,080
1,200
1,000
SOO
Measured
(Pg)
502
766
1,860
1,840
3,430
1,330
1,010
455
Recovery
(X)
27
91
110
75
111
111
101
91
Spike
(pg)
460
210
420
610
770
300
250
SOO
Measured
(Pg)
512
395
403
616
840
328
234
410
Recovery
(X)
111
190
96
101
no
110
94
82
Spike
(PR)
920
420
840
1,220
1,540
600
500
500
Measured
(Pg)
582
690
940
1,180
1,790
691
530
355
Recovery
(t)
63
164
112
97
116
115
106
71
Spike Measured
(pg) (PR)
920
420
840
1,220
1,540
600
500
SOO
HD«
ND
NO
ND
586
ND
904b
100
Recovery
(t)
0
0
0
0
38
0
181
20
en . —-
GO |W detected.
(Vj; Measured value corrected for endogenous 2,3,7,8-TCDD content (averaged 616 pg/L).
^^"^ CO
en
-------
TABLE 16. ACCURACY OF THE HHCC/HRMS METHOD FOR THE DETERMINATION OF TCDD ISOMERS SPIKED IHTO SOIL MATRICES
Hyde Park 001 (HI)
TCDD
analyte
1,3,6,8
>,3,7,9
1,2,3,7/1
1.2,3,4
1,2,7,8
1,2.8,9
2.3.7,8
1SC,,-2,3
*HD - oot
to
cn
Spike
(Pi)
130
60
,2,3,8 120
170
220
84
-
,7,8 SOO
detected. The
Measured Recovery
(PR) (X)
29.0
51.1
12S
118
2S2
100
30.3
280
value in
22
86
106
69
117
119
-
56
parentheses
B25-Piazza Road (85)
Spike
(PS)
92
42
84
120
ISO
60
-
SOO
reflect*
Measured
(Pt)
HD (9.2)
IS. 2
61.8
S4.I
147
63.9
12.9
240
the estia
Recovery
(I)
• o
36
74
44
95
107
-
48
BS2-Sheoandoah
Spike
(PR)
660
300
600
880
1,110
430
-
SOO
Measured
(PC)
333
635
S18
69S
1,170
463
1,280
400
(Bl)
Recovery
(X)
SO
210
87
79
106
108
-
80
Hyde Park 003
Spike
(PR)
1,560
710
1,430
2,070
2,620
1,020
-
500
Measured
(PI)
383
367
82S
8SS
2,330
952
1,800
430
(H3)
Recovery
(X)
24
51
58
41
89
93
-
86
ated detection limit.
01
CO
-------
TABU I). FOBT1F1F.D FIELD BUMK RESULTS
cn
oc
co
Aliquot
Saiple Air-dry vt. (t)
Do. or Vol. (L)
8367-62- I576X-FFVB 1.0 L
8367-64- I576X-FFSB 10 |
B367-63-1576X-FF5A 10 |
8367-61- I576X-FFWA l.OL
8367-81- I576X-FFB 10 .01 «
8367-80-I576X-FFA 10.01 |
B367-97-1S76X-FFA 10 L
8367-98- 1576X-FTB 1 0 L
Retention limt
Katiire
23:38
23:38
23:40
23:40
22:17
21:37
21:37
21:43
'
23
23
23
23
22
21
21
21
JC
:35
:38
:38
:39
:15
:35
:37
:43
Inter.
ID '
HS50
MS50
MS50
HS50
HS50
MS50
MS50
ns50
Date
09/11/85
09/11/85
09/12/85
09/12/85
09/20/85
09/20/85
09/20/65
09/20/85
Ti-e
14:08
14:42
14:47
15:14
13:33
14:09
09: 10
09:24
Relati
320/322
0.78
0.79
0.77
0.80
0.88
0.86
0.82
0.80
ve Ion Abundance Ratio!
332/334(15) 332/334OS) X >ec.
0 71 65
0.80 66
0.78 - 71
0.76 - 79
0.74 29
0.73 - 83
0.83 - 50
0.82 48
•/z 259
> 145:1
> 145:1
145: 1
145:1
144:1
145:1
23:1
145:1
S/K
*/l 322
> 63:1
> 63:1
> 63:1
> 63:1
> 63:1
> 63:1
> 63:1
> 63:1
mil 334(18) Cootenti
> 63:1
> 63:1
> 63:1
> 63:1
> 63:1 . Low reiovery
> 63:1
> 63:1
> 63:1
-------
HRGC and Mass Resolution
Table 11 presents a summary of all chromatographic and mass resolution
checks completed during the final method evaluation. As per the protocol
requirements the required mass resolution was demonstrated as the first and
last quality control requirements for each day. The column performance
check mixture was also analyzed before the first sample analysis and after
the final sample analysis each day as a QC measure to assure that speci-
ficity for 2,3,7,8-TCDD was maintained. The mass measurement accuracy at
m/z 330.979 is also included in this table, as it was verified on a daily
basis prior to any sample analyses.
Sample Analysis
The results from the analysis of the aqueous and soil samples are pro-
vided in Table 12. The data in Table 12 are presented in the format speci-
fied as Form B-l in the protocol reporting requirement. The data are re-
corded in the chronological order that they were obtained by HRGC/HRMS.
As indicated in Table 12, several samples required reanalysis due to
low recovery of the internal standard, unacceptable ion ratios for 320/322,
or the result of interferences at the internal standard. Two of the dis-
tilled water samples demonstrated responses for the characteristic ions at
m/z 259, 320, and 322 for 2,3,7,8-TCDD. However, the ion ratio for the
native 2,3,7,8-TCDD in one replicate and the ion ratio for the internal
standard in another required that both samples be reanalyzed. Although both
samples met all the qualitative criteria, recoveries were noted to be low
(< 20 percent) for one of the samples and complete reanalysis of the repli-
cate was required.
Significant problems were encountered with the aqueous soil extract,
H2W, and the fly ash sample. The problems with the soil extract resulted
from an interference at m/z 332 that coeluted with the internal standard,
13Ci2~2,3,7,8-TCDD. No accurate quantitative measurements could be achieved
for TCDD responses observed for this sample. The original sample contained
a large amount of suspended particulate in each of the three replicates.
Problems with the extraction were noted with the first portion of methylene
chloride. A large amount of particulate matter was noted at the interface
of the aqueous and organic phases. Hence, the interference at m/z 332 and
TCDD responses observed in these replicates were probably due to direct ex-
traction of the suspended soil particulate rather than the actual water-
soluble TCDD.
The remaining aqueous sample for H2W was centrifuged for 10 min at ap-
proximately 2,000 rpm, and the aqueous phase was decanted from the settled
particulate. The resulting aqueous sample was divided into duplicate 430-mL
samples and each was analyzed according to the protocol. The HRGC/HRMS
analysis was successful for both replicates with absolute recoveries of
78 percent and 96 percent of the internal standard.
The triplicate analyses of the fly ash sample resulted in absolute
recoveries less than 10 percent for the internal standard in each aliquot
38
-------
analyzed. These low recoveries may be associated with the total fixed car-
bon content of the fly ash material. Previous work in this laboratory with
fly ash from coal-fired power plants has demonstrated low recoveries of ana-
lytes from materials with high carbon content.4
The only other sample for which successful analysis was not achieved as
specified in the protocol on first analysis was the industrial wastewater
(IND). The triplicate analysis of the sample resulted in absolute internal
standard recoveries of 23, 20, and 29 percent. The criteria for successful
analysis for TCDD as discussed in the protocol require an absolute recovery
of 40 to 120 percent. In addition to the observed low recoveries, the level
of 2,3,7,8-TCDD detected in the sample averaged 1,410 ppq as compared to the
500-ppq spike level. Two 500-mL aliquots of the unspiked industrial waste-
water sample were reanalyzed to determine the background level of 2,3,7,8-
TCDD. The results of the duplicate analysis yielded an average 2,3,7,8-TCDD
concentration of approximately 620 ppq and the absolute recoveries were
noted to be 60 percent and 57 percent. The increase in absolute recovery of
the internal standard in the unspiked sample by approximately a factor of
two is possibly due to the preparation of samples one half the size of that
used for the original analysis. This suggests that the sample matrix has a
considerable impact on the effectiveness of the cleanup procedure.
Table 13 provides a summary of the accuracy and precision of the analy-
ses of the five aqueous sample types for 2,3,7,8-TCDD. Only the data points
from Table 12 that demonstrate compliance with all QC criteria (ion ratios,
absolute recovery of the internal standard, etc.) are included in Table 13.
These data demonstrate that the isotope dilution method of quantitation pro-
vides accurate and precise quantitation of 2,3,7,8-TCDD in the aqueous sam-
ples. It should be noted that even when the absolute recovery of the 13Ci2~
2,3,7,8-TCDD internal standard varies by as much as 66 percent (RPR) for the
triplicate distilled water samples, the accuracy of the measurement of the
spiked 2,3,7,8-TCDD averaged 101 percent with less than 10 percent variabil-
ity. Table 13 summarizes data for both the spiked and unspiked aliquots of
industrial wastewater. The high recovery noted for the 2,3,7,8-TCDD value
in the spiked samples is a result of the presence of this compound at ap-
proximately 620 ppq in the original matrix.
Table 14 presents a similar summary for the five solid samples ana-
lyzed. The precision of the measurements is not quite as good as noted for
the aqueous samples and may reflect the difference in adsorption of the
endogenous 2,3,7,8-TCDD and the spiked internal standard to the matrices.
Tables 15 and 16 provide data dealing with the accuracy of the HRGC/
HRMS methods for the determination of total TCDD isomers in aqueous and
solid samples. In general, the data support the use of the internal stan-
dard method of quantitation for all but the earliest eluting isomers,
1,3,6,8- and 1,3,7,9-TCDD. The accuracy for the additional isomers is very
good and more consistent than is observed for the solid samples. This may
be partially due to the differences in adsorption to the soil particles.
39
r~ o ""i
DO i
-------
Fortified Field Blanks
As part of the overall quality assurance/quality control (QA/QC) pro-
gram identified in the HRGC/HRMS protocol, the analyst is required to ana-
lyze fortified field blanks to demonstrate (a) that the extraction and
cleanup procedure will provide recovery of the 2,3,7,8-TCDD within the cri-
teria of greater than 40 percent specified in the protocol and (b) that the
reagents are free from contamination with TCDD isomers.
Table 17 provides the results of the fortified field blanks run before
proceeding with sample analysis and also those of an additional set of
blanks prepared along with the actual samples. The analyses of the forti-
fied field blanks at the outset of the study demonstrated that the recover-
ies of 2,3,7,8-TCDD and 1,2,3,4-TCDD ranged from 65 to 79 percent. No de-
tectable levels of other TCDD isomers were found in this preliminary study.
The field fortification blanks analyzed with the actual samples resulted in
recoveries of 29 percent and 83 percent. More importantly, these analyses
demonstrated some interferences arising from 1,3,6,8- and 1,3,7,9-TCDD.
Previous studies involving evaluation of the cleanup procedure indicated
that these isomers are associated with the acidic alumina cleanup.
Figure 4 is a plot of the ratio of response of 1,3,6,8- and 1,3,7,9-
TCDD and the response of the recovery standard 13C12-1,2,3,4-TCDD versus the
time elapsed since the acidic alumina was cleaned and activated at 190°C.
The results of the analyses of the fortified field blanks and the samples
not spiked with the 1,3,6,8- and 1,3,7,9-TCDD isomers are presented in Fig-
ure 4. As noted from this plot, these TCDD isomers were not initially de-
tected in the acidic alumina immediately following cleanup by Soxhlet ex-
traction. The first set of fortified field blanks was taken through the
acidic alumina column 7 days later. Although response was observed at m/z
320 and 322 at the retention time for these isomers, the ion ratios did not
indicate presence of the compounds. Since the detectable levels were well
below 10 pg/g of alumina, the sample analyses were initiated. The data for
the fortified field blanks and samples taken through alumina from 14 to 30
days from activation indicate that the contamination of the 1,3,6,8- and
1,3,7,9-TCDD isomers apparently occurs over time using this particular oven.
The background contamination of 1,3,6,8- and 1,3,7,9-TCDD isomers has also
been recently addressed by the Center for Disease Control.5
Note Added in Proof
A second magnetic sector instrument (built in 1976) from a different
manufacturer was tested and found to be incapable of achieving sufficient
sensitivity at 10,000 resolving power to be used in experiments for this
study.
40
-------
Q
Q
U
1.2
1.1
1.0
CO
*
CN
r 0.9
u
CO
2 0.8
3 0.7
o 0.6
U
Js 0.5
CO
-- 0.4
00
o- 0.3
o 0.2
o
&!.
0.1
• Fortified Field Blanks
I Aqueous and Environmental Samples
Samples
1
10 20
Time (Days) Elapsed from Cleanup and
Activation of Acidic Alumina
30
Figure 3. Background levels of 1,3,6,8- and 1,3,7,9-TCDD observed
over the single-laboratory evaluation study.
41
-------
REFERENCES
1. U.S. Environmental Protection Agency, "Dioxin Strategy," prepared by the
Office of Water Regulations and Standards and the Office of Solid Waste
and Emergency Response in conjunction with the Dioxin Strategy Task
Force, Washington, B.C., November 28, 1983.
2. L. R. Williams, Validation of Testing/Measurement Methods.
EPA 600/X-83-060, 1983.
3. GC Bulletin 793C, Supelco Inc., Beliefonte, Pennsylvania, 1983.
4. C. L. Haile, J. S. Stanley, T. Walker, G. R. Cobb, and B. A. Boomer,
"Comprehensive Assessment of the Specific Compounds Present in Combus-
tion Processes. Volume 3. National Survey of Organic Emissions from
Coal-Fired Utility Boiler Plants," EPA-560/5-83-006, September 1983.
5. J. S. Heller, D. G. Patterson, L. R. Alexander, D. F. Groce, R. P.
O'Connor, and C. R. Lapeza, "Control of Artifacts and Contamination in
the Development of a Dioxin Analytical Program," presented at the 33rd
Annual Conference on Mass Spectrometry and Allied Topics, May 26-31,
1985, San Diego, California.
42
590
-------
APPENDICES
43
531
-------
APPENDIX A
VALIDATED ANALYTICAL PROTOCOL
for the Determination of
2,3,7,8-Tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) and Total
TCDDs in Soil/Sediment and Water by High-Resolution Gas
Chromatography/High-Resolution Mass Spectrometry
September 10, 1985
This analytical protocol has been written in the format used in the
Superfund program, as "Exhibit D" of a Statement of Work which in turn is part
of an Invitation-for-Bid package under the Superfund Contract Laboratory Program.
The other exhibits of the Statement of Work, although cited in Exhibit D, do
not pertain to this method evaluation study.
c; c 9
-------
EXHIBIT D
Analytical Methods
2,3,7,8-Tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) and Total
TCDDs in Soil/Sediment and Water by High-Resolution Gas
Chromatography/High-Resolution Mass Spectrometry
r o, l'\
u
-------
EXHIBIT D
Section Subject Page
1 Scope and Application D-l
2 Summary of Method D-l
3 Definitions D-2
4 Interferences D-3
5 Safety D-3
6 Apparatus and Equipment D-3
7 Reagents and Standard Solutions D-6
8 System Performance Criteria D-8
9 Quality Control Procedures D-13
10 Sample Preservation and Handling D-13
11 Sample Extraction D-14
12 Analytical Procedures D-l 7
13 Calculations D-18
-------
1. SCOPE AND APPLICATION
1.1 This method provides procedures for Che detection and quantitative
measurement of 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD; CAS
Registry Number 1746-01-6; Storet number 3475) at concentrations of
10 pg/g (10 parts per trillion) to 200 pg/g (200 parts per trillion)
in 10-g portions of soil and sediment and at 100 pg/L (100 parts per
quadrillion) to 2000 pg/L (2 parts per trillion) in 1-L samples of
water. The use of 1-g or 100-mL portions permits measurements of
concentrations up to 2,000 pg/g (2 parts per billion) or 20 ng/L,
respectively* This method also allows the estimation of quantities
of total TCDD present in the sample. Samples containing concentrations
of 2,3,7,8-TCDD greater than 2 ppb or 20 ng/L must be analyzed by a
protocol designed for such concentration levels, with an appropriate
instrument calibration range.
1.2 The-minimum measurable concentration is estimated to be 10 pg/g (10
parts per trillion) for soil and sediment samples and 100 pg/L for
water samples, but this depends on kinds and concentrations of
interfering compounds in the sample matrix.
1.3 This method is designed for use by analysts who are experienced in
the use of high-resolution gas chromatography/high-resolution mass
spectrometry•
CAUTION: TCDDs are extremely hazardous. It is the laboratory's responsi-
bility to ensure that safe handling procedures are employed.
2. SUMMARY OF METHOD
Five hundred pg of C,2~2,3,7,8-TCDD (internal standard) are added to a
10-g portion of a soil/sediment sample (weighed to 3 significant figures)
or a 1-L aqueous sample and the sample is extracted with 200 to 250 mL
benzene using a Soxhlet apparatus with a minimum of 3 cycles per hour or a
continuous liquid-liquid extractor for 24 hours. A separatory funnel and
3 x 60 mL methylene chloride may also be used for aqueous samples. After
appropriate concentration and cleanup, 50 uL of tridecane are added to the
extract. Before HRGC-HRMS analysis, 500 pg of a recovery standard ( C^"
1,2,3,4-TCDD) are added to the extract which is then concentrated to a
final volume of 50 uL. A 2-uL aliquot of the concentrated extract is
injected into a gas chromatograph with a capillary column interfaced to a
high-resolution mass spectrometer capable of rapid multiple ion monitoring
at resolutions of at least 10,000 (10 percent valley).
Identification of 2,3,7,8-TCDD is based on the detection of the ions m/z
319.897 and 321.894 at the same GC retention time and within -1 to +3
seconds GC retention time of the internal standard masses of m/z 331.937
and 333.934. Confirmation of 2,3,7,8-TCDD (and of other TCDD isomers) is
based on the ion m/z 258.930 which results from loss of COCL by the parent
ion.
D-l
n
-------
3. DEFINITIONS
3*1 Concentration calibration solutions -- solutions containing known
amounts of the analyte (unlabeled 2,3,7,8-TCDD), the internal standard
13C12-2,3,7,8-TCDD and the recovery standard 13C,2~1,2,3,4-TCDD;
they are used to determine instrument response or the analyte
relative to the internal standard and of the internal standard
relative to the recovery standard.
3.2 Field blank — a portion of soil/sediment or water uncontaminated with
2,3,7,8-TCDD and/or other TCDDs.
3.3 Rinsate — a portion of solvent used to rinse sampling equipment; the
rinsate is analyzed to demonstrate that samples have not been contami-
nated during sampling.
3.A Internal standard — ^Cj2~2»3,7,8-TCDD, which is added to every
sample (except the blanks described in Sections 4.2.1 and 4.2.3 of
Exhibit E) and is present at the same concentration in every labora-
tory method blank, quality control sample, and concentration calibra-
tion solution. It is added to the soil/sediment or aqueous sample
before extraction and is used to measure the concentration of each
analyte. Its concentration is measured in every sample, and percent
recovery is determined using an internal standard method.
3.5 Recovery standard — C12~*,2,3,4-TCDD which is added to every sample
(except for the blanks discussed in Sections 4.2.1.A.2 and 4.2.3.6,
Exhibit E) extract just before HRGC-HRMS analysis.
3.6 Laboratory method blank — this blank is prepared in the laboratory
through performing all analytical procedures except addition of a
sample aliquot to the extraction vessel.
3.7 GC column performance check mixture —• a mixture containing known
amounts of selected standards; it is used to demonstrate continued
acceptable performance of the capillary column, i.e., separation
(jC 25% valley) of 2,3,7,8-TCDD isomer from all other 21 TCDD isomers
and to define the retention time window.
3.8 Performance evaluation sample — a soil, sediment or aqueous sample
containing a known amount of unlabeled 2,3,7,8-TCDD and/or other
TCDDs. It is distributed by EPA to potential contractor laboratories
who must analyze it and obtain acceptable results before being awarded
a contract for sample analyses (see IFB Pre-Award Bid Confirmations).
It may also be included as an unspecified ("blind") QC sample in any
sample batch submitted to a laboratory for analysis.
3.9 Relative response factor — response of the mass spectrometer to a
known amount of an analyte relative to a known amount of an internal
standard.
3.10 Mass resolution check — standard method used to demonstrate static
resolution of 10,000 minimum (102 valley definition).
D-2
r: r
-------
4. INTERFERENCES
Chemicals which elute from the GC column within ^10 scans of the internal
and/or recovery standard (ra/z 331.937 and 333.934) and which produce ions
at any of the masses used to detect or quantify TCDD are potential inter-
ferences. Most frequently encountered potential interferences are other
sample components that are extracted along with TCDD, e.g. PCBs, methoxy-
biphenyls, chlorinated hydroxydiphenylethers, benzylphenylethers, chlori-
nated naphthalenes, DDE, DDT, etc. The actual incidence of interference
by these chemicals depends also upon relative concentrations, mass spectro-
metric resolution, and chromatographic conditions. Because very low
levels of TCDD must be measured, the elimination of interferences is
essential. High-purity reagents and solvents must be used and all equip-
ment must be scrupulously cleaned. Laboratory reagent blanks (Exhibit E,
Quality Control, Section 4) must be analyzed to demonstrate absence of
contamination that would interfere with TCDD measurement. Column chromato-
graphic procedures are used to remove some coextracted sample components;
these procedures must be performed carefully to minimize loss of TCDD
during attempts to increase its concentration relative to other sample
components.
5. SAFETY
•
The toxicity or carcinogenicity of each reagent used in this method has
not been precisely defined; however, each chemical compound should be
treated as a potential health hazard. From this viewpoint, exposure to
these chemicals must be reduced to the lowest possible level by whatever
means available. The laboratory is responsible for maintaining a file of
current OSHA regulations regarding the safe handling of the chemicals
specified in this method. A reference file of material data handling
sheets should also be made available to all personnel involved in the
chemical analysis. Additional references to laboratory safety are identi-
fied d~3) (page D-38). 2,3,7,8-TCDD has been identified as a suspected
human or mammalian carcinogen. The laboratory is responsible for ensuring
that safe handling procedures are followed.
6. APPARATUS AND EQUIPMENT
6.1 High-Resolution Gas Chrotnatograph/High-Resolution Mass
Spectrometer/Data System (HRGC/HRMS/DS)
6.1.1 The GC must be equipped for temperature programming, and all
required accessories must be available, such as syringes, gases,
and a capillary column. The GC injection port must be designed
for capillary columns. The use of splitless injection tech-
niques is recommended. On-column injection techiques can be
used but this may severely reduce column lifetime for non-
chemical ly bonded columns. When using the method in this
protocol, a 2-uL injection volume is used consistently. With
some GC injection ports, however, 1-uL injections may produce
improved precision and chromatographic separation. A 1-uL
D-3
597
-------
injection volume may be used if adequate sensitivity and
precision can be achieved.
NOTE: If 1 uL is used at all as injection volume, the injection
volumes for all extracts, blanks, calibration solutions and
the performance check sample must be 1 uL.
6.1.2 Gas Chromatograph-Mass Spectrometer Interface
The GC-MS interface may include enrichment devices, such as
a glass jet separator or a silicone membrane separator, or
the gas chromatograph can be directly coupled to the mass
spectrometer source. The interface may include a diverter
valve for shunting the column effluent and isolating the mass
spectrometer source. All components of the interface should
be glass or glass-lined stainless steel. The interface com-
ponents should be compatible with 300°C temperatures. The
GC/MS interface must be appropriately designed so that the
separation of 2,3,7,8-TCDD from the other TCDD isomers which
is achieved in the gas chromatographic column is not appreci-
ably degraded. Cold spots and/or active surfaces (adsorption
sites) in the GC/MS interface can cause peak tailing and peak
broadening. It is recommended that the GC column be fitted
directly into the MS source. Graphite ferrules should be
avoided in the GC injection area since they may adsorb TCDD.
Vespel* or equivalent ferrules are recommended.
6.1.3 Mass Spectrometer
The static resolution of the instrument must be maintained at
a minimum 10,000 (10 percent valley). The mass spectrometer
must be operated in a selected ion monitoring (SIM) mode with
total cycle time (including voltage reset time) of one second
or less (Section 8.3.A.I). At a minimum, the following ions
which occur at these masses must be monitored: m/z 258.930,
319.897, 321.894, 331.937 and 333.934.
6.1 .4 Data System
A dedicated hardware or data system is employed to control the
rapid multiple ion monitoring process and to acquire the data.
Quantification data (peak areas or peak heights) and SIM traces
(displays of intensities of each m/z being monitored as a
function of time) must be acquired during the analyses.
Quantifications may be reported based upon computer-generated
peak areas or upon measured peak heights (chart recording).
NOTE: Detector zero setting must allow peak-to-peak measurement of the noise
on the base line.
6.2 GC Columns
D-4
-------
For isoner-spec ific determinations of 2,3,7,8-TCDD, the following two
fused silica capillary columns are recommended: a 60-ra SP-2330 column
and a 50-m CP-Sil 88 column. However, any capillary column which
separates 2,3,7,8-TCDD from all other TCDDs may be used for such
analyses, but this separation must be demonstrated and documented.
Minimum acceptance criteria must be determined per Section 8.1. At
the beginning of each 12-hour period (after mass resolution has been
demonstrated) during which sample extracts or concentration calibra-
tion solutions will be analyzed, column operating conditions must be
attained for the required separation on the column to be used for
samples. Operating conditions known to produce acceptable results
with the recommended columns are shown in Table 2 at the end of this
Exhibit.
6.3 Miscellaneous Equipment
6.3.1 Nitrogen evaporation apparatus with variable flow rate.
6.3.2 Balance capable of accurately weighing to 0.01 g.
6.3.3 Centrifuge capable of operating at 2,000 rpm.
6.3.4 Water bath — equipped with concentric ring cover and capable
of being temperature-controlled within j*2°C.
6.3.5 Stainless steel spatulas or spoons.
6.3.6 Stainless steel (or glass) pan large enough to hold contents
of 1-pint sample containers.
6.3.7 Glove box.
6.3.8 Drying oven.
6.4 Glassware
6.4.1 Soxhlet apparatus — all-glass, Kontes 6730-02 or equivalent;
90 mm x 35 mm glass thimble; 500-mL flask; condenser of appro-
priate size.
6.4.2 Kuderna-Danish apparatus — 500-mL evaporating flask, 10-mL
graduated concentrator tubes with ground-glass stoppers, and
3-ball macro Snyder column (Kontes K-570001-0500, K-503000-
0121 and K-569001-0219 or equivalent).
6.4.3 Mini-vials — 1-mL borosilicate glass with conical-shaped
reservoir and screw caps lined with Teflon-faced silicone disks.
6.4.4 Funnels — glass; appropriate size to accommodate filter
paper used to filter jar extract (volume of approximately 170 mL).
6.4.5 Separatory funnel -- 2000 mL with Teflon stopcock.
D-5
-------
6.4.6 Continuous liquid-liquid extractors equipped with Teflon or
glass connecting joints and stopcocks requiring no lubrication
(Hershberg-Wolf Extractor - Ace Glass Company Vineland, NJ,
P/N 6841-10 or equivalent).
6.4.7 Chromatographic columns for the silica and alumina chroma-
tography — 1 cm ID x 10 cm long and 1 cm ID x 30 cm long.
6.4.8 Chromatography column for the Carbopak cleanup — disposable
5-mL graduated glass pipets, 7 mm ID.
6.4.9 Desiccator.
6.4.10 Glass rods.
NOTE: Reuse of glassware should be minimized to avoid the risk of
cross contamination. All glassware that is reused must be
scrupulously cleaned as soon as possible after use, applying
the following procedure.
Rinse glassware with the last solvent used in it then with
high-purity acetone and hexane. Wash with hot water containing
detergent. Rinse with copious amounts of tap water and several
portions of distilled water. Drain dry and heat in a muffle
furnace at 400°C for 15 to 30 minutes. Volumetric glassware
should not be heated in a muffle furnace, and some thermally
stable materials (such as PCBs) may not be removed by heating
in a muffle furnace. In these cases, rinsing with high-purity
acetone and hexane may be substituted for muffle furnace
heating. After the glassware is dry and cool, rinse with hexane,
and store inverted or capped with solvent-rinsed aluminum foil
in a. clean environment.
7. REAGENTS AND STANDARD SOLUTIONS
7.1 Column Chromatography Reagents
7.1.1 Alumina, acidic — Extract the alumina in a Soxhlet with
methylene chloride for 6 hours (minimum of 3 cycles per hour)
and activate it by heating in a foil-covered glass container
for 24 hours at 190°C.
7.1.2 Silica gel — high-purity grade, type 60, 70-230 mesh; extract
the silica gel in a Soxhlet with methylene chloride for 6 hours
(minimum of 3 cycles per hour) and activate it by heating in a
foil-covered glass container for 24 hours at 130°C.
7.1.3 Silica gel impregnated with 40 percent (by weight) sxilfuric
acid — add two parts (by weight) concentrated sulfuric acid
to three parts (by weight) silica gel (extracted and activated),
mix with a glass rod until free of lumps, and store in a
screw-capped glass bottle.
D-6
GOO
-------
7.1.4 Sulfuric acid, concentrated — ACS grade, specific gravity 1.84.
7.1.5 Graphitized carbon black (Carbopack C or equivalent), surface
of approximately 12 m^/g, 80/100 mesh — mix thoroughly 3.6
grams Carbopak C and 16.4 grams Celite 545* in a 40-mL vial.
Activate at 130° C for six hours. Store in a desiccator.
7.1.6 Celite 545®, reagent grade, or equivalent.
7.2 Membrane filters or filter paper with pore size of <25 urn; rinse with
hexane before use.
7.3 Glass wool, silanized — extract with methylene chloride and hexane
and air-dry before use.
7.4 Desiccating Agents
7.4.1 Sodium sulfate — granular, anhydrous; before use, extract it
with methylene chloride for 6 hours (minimum of 3 cycles per
hour) and dry it for >4 hours in a shallow tray placed in an
oven operated at 120°C. Let it cool in a desiccator.
7.4.2 Potassium carbonate—anhydrous, granular; use as such.
7.5 Solvents — high purity, distilled in glass: methylene chloride,
toluene, benzene, cyclohexane, methanol, acetone, hexane; reagent
grade: tridecane.
7.6 Concentration calibration solutions (Table 1) — five tridecane
solutions containing unlabeled 2,3,7,8-TCDD and 13C,o-l,2,3,4-TCDD
(recovery standard) at varying concentrations, and Ci -y-2 ,3,7,8-TCDD
(internal standard, CASRN 80494-19-5) at a constant concentration
must be used to calibrate the instrument. These concentration calibra-
tion solutions must be obtained from the Quality Assurance Division,
US EPA Environmental Monitoring Systems Laboratory (EMSL-LV), Las Vegas,
Nevada. However, additional secondary standards may be obtained from
commercial sources, and solutions may be prepared in the contractor
laboratory. Traceability of standards must be verified against EPA-
supplied standard solutions. Such procedures will be documented by
laboratory SOPs as required in IFB Pre-award Bid Confirmations, part
2.f.(4). It is the responsibility of the laboratory to ascertain that
the calibration solutions received are indeed at the appropriate
concentrations before they are injected into the instrument. Serious
overloading of the instrument may occur if the concentration calibra-
tion solutions intended for a low-resolution MS are injected into the
high-resolution MS.
7.6.1 The five concentration calibration solutions contain unlabeled
2,3,7,8-TCDD and labeled ^C.--1»2,3,4-TCDD at nominal concen-
trations of 2.5, 5.0, 10.0, 20.0 and 40.0 pg/uL, respectively,
and labeled Cj^-2 ,3 , 7 ,8-TCDD at a constant nominal concen-
tration of 10.0 pg/uL.
D-7
601
-------
7.6.2 Store the concentration calibration solutions in 1-mL mini-
vials at 4°C.
7.7 Column performance check mixture — this solventless mixture must be
obtained fvom the Quality Assurance Division, Environmental Monitoring
Systems Laboratory, Las Vegas, Nevada, and dissolved by the Contractor
in 1 mL tridecane. This solution will then contain the following
components (including TCDDs (A) eluting closely to 2,3,7,8-TCDD, and
the first- (F) and last-eluting (L) TCDDs when using the columns
recommended in Section 6.2) at a concentration of 10 pg/uL of each of
these isomers:
Analyte Approximate Amount Per Ampule
Unlabeled 2,3,7,8-TCDD 10 ng
13C12-2,3,7,8-TCDD 10 ng
1,2,3,4-TCDD (A) 10 ng
1,4,7,8-TCDD (A) 10 ng
1,2,3,7-TCDD (A) 10 ng
1,2,3,8-TCDD (A) 10 ng
1,2,7,8-TCDD 10 ng
1,3,6,8-TCDD (F) 10 ng
1,2,8,9-TCDD (L) 10 ng
7.8 Sample fortification solution — an isooctane solution containing
the internal standard at a nominal concentration of 5 pg/uL.
7.9 Recovery standard spiking solution — an isooctane solution contain-
ing the recovery standard at a nominal concentration of 100 pg/uL.
Five uL of this solution will be spiked into the extract before
HRGC/HRMS analysis.
7.10 Internal standard spiking solution — an isooctane solution containing
the internal standard at a nominal concentration of 100 pg/uL. Five
uL of this solution will be added to a fortified field blank extract
(Section 4.2.1.A.2, Exhibit E).
8. SYSTEM PERFORMANCE CRITERIA
System performance criteria are presented in two sections. One section
deals with GC column performance criteria while the other section consists
of initial calibration criteria. The laboratory may use either of the
recommended columns described in Section 6.2. It must be documented that
D-8
-------
all applicable system performance criteria specified in Sections 8.1, 8.2
and 8.3 have been met before analysis of any sample is performed. Table 2
provides recommended conditions that can be used to satisfy the required
criteria. Table 3 provides a typical 12-hour analysis sequence.
8.1 GC Column Performance
8.1.1 Inject 2 uL (Section 6.1.1) of the column performance check
solution (Section 7.7) and acquire selected ion monitoring
(SIM) data for m/z 258.930, 319.897, 321.894, 331.937 and
333.934 within a total cycle time of ^1 second (Section
8.3.4.1).
8.1.2 The chromatographic peak separation between 2,3,7,8-TCDD and
the peaks representing any other TCDD isomers must be resolved
with a valley of <25 percent, where
Valley Percent » (x/y)(100)
x = measured as in Figures 1 and 2
y « the peak height of 2,3,7,8-TCDD.
It is the responsibility of the laboratory to verify the con-
ditions suitable for the appropriate resolution of 2,3,7,8-TCDD
from all other TCDD isomers. The column performance check
solution also contains the TCDD isomers eluting first and last
under the analytical conditions specified in this protocol
thus defining the retention time window for total TCDD determi-
nation. The peaks representing 2,3,7,8-TCDD, the first and
the last eluting TCDD isomers must be labeled and identified
as such on the chromatograms.
8.2 Mass Spectrometer Performance
8.2.1 The mass spectrometer must be operated in the electron (impact)
ionization mode. Static mass resolution of at least 10,000
(10 percent valley) must be demonstrated before any analysis
of a set of samples is performed (Section 8.2.2). Static
resolution checks must be performed at the beginning and at
the end of each 12-hour period of operation. However, it is
recommended that a visual check (i.e., not documented) of the
static resolution be made using the peak matching unit before
and after each analysis.
8.2.2 Chromatography time for TCDD may exceed the long-term mass
stability of the mass spectrometer and thus mass drift correc-
tion is mandatory. A reference compound (high boiling PFK is
recommended) is introduced into the mass spectrometer. An
acceptable lock mass ion at any mass between m/z 250 and m/z
334 (m/z 318.979 from PFK is recommended) must be used to
monitor and correct mass drifts.
D-9
603
-------
NOTE: Excessive PFK may cause background noise problems and contami-
nation of the source resulting in an increase in "downtime"
for source cleaning.
Using a PFK molecular leak, tune the instrument to meet the
minimum required mass resolution of 10,000 (10Z valley) at
m/z 254.986 (or any other mass reasonably close to m/z 259).
Calibrate the voltage sweep at least across the mass range m/z
259 to m/z 334 and verify that m/z 330.979 from PFK (or any
other mass close to m/z 334) is measured within _+5 ppm (i.e.,
1.7 mmu) using m/z 254.986 as a reference. Documentation of the
mass resolution must then be accomplished by recording the
peak profile of the PFK reference peak m/z 318.979 (or any
other reference peak at a mass close to m/z 320/322). The
format of the peak profile representation must allow manual
determination of the resolution, i.e., the horizontal axis
must be a calibrated mass scale (amu or ppm per division).
The result of the peak width measurement (performed at 5
percent of the maximum) must appear on the hard copy and
cannot exceed 31.9 mmu or 100 ppm.
8.3 Initial Calibration
Initial calibration is required before any samples are analyzed for
2,3,7,8-TCDD. Initial calibration is also required if any routine
calibration does not meet the required criteria listed in Section 8.6.
8.3.1 All concentration calibration solutions listed in Table 1 must
be utilized for the initial calibration.
8.3.2 Tune the instrument with PFK as described in Section 8.2.2.
8.3.3 Inject 2 uL of the column performance check solution (Section
7.7) and acquire SIM mass spectral data for m/z 258.930,
319.897, 321.894, 331.937 and 333.934 using a total cycle time
of £ 1 second (Section 8.3.4.1). The laboratory must not
perform any further analysis until it has been demonstrated
and documented that the criterion listed in Section 8.1.2 has
been met.
8.3.4 Using the same GC (Section 8.1) and MS (Section 8.2) conditions
that produced acceptable results with the column performance
check solution, analyze a 2-uL aliquot of each of the 5 concen-
tration calibration solutions in triplicate with the following
MS operating parameters.
8.3.4.1 Total cycle time for data acquisition must be £ 1
second. Total cycle time includes the sum of all the
dwell times and voltage reset times.
8.3.4.2 Acquire SIM data for the following selected
characteristic ions:
D-10
-------
m/z Compound
258.930 TCDD - COC1
319.897 Unlabeled TCDD
321.894 Unlabeled TCDD
331.937 13C12-2,3,7,8-TCDD, 13C12~1 ,2,3 ,4-TCDD
333. 934 13C12-2,3,7,8-TCDD, 13C12-1 ,2 ,3 ,4-TCDD
8.3.4.3 The ratio of integrated ion current for m/z 319.897 to
m/z 321.894 for 2,3,7,8-TCDD must be between 0.67 and
0.90.
8.3.4.4 The ratio of integrated ion current for m/z 331.937 to
m/z 333.934 for l JC12-2,3, 7, 8-TCDD and 13C12-1 ,2 ,3 ,4-
TCDD must be between 0.67 and 0.90.
8.3.4.5 Calculate the relative response factors for unlabeled
2,3,7,8-TCDD [RRF(I) ] relative to 13C12~2 ,3 , 7 ,8-TCDD
and for labeled C12-2 ,3 ,7, 8-TCDD [RRF(II)] relative
to 13C-1,2,3,4-TCDD as follows:
RRF(I) = —
QX ' AIS
Ais * QRS
RRF(II) =
QIS " ^S
where
AX ™ sum of the integrated ion abundances of m/z 319.897 and m/z 321.894
for unlabeled 2,3,7,8-TCDD.
AIS » sum of the integrated ion abundances of m/z 331.937 and m/z 333.934
for 13C12-2,3,7,8-TCDD.
ARs B sura of the integrated ion abundances for m/z 331.937 and m/z
333.934 for 13C12~1,2,3,4-TCDD.
QIS * Quantity of 13C^2~2,3,7,8-TCDD injected (pg).
QRS - quantity of 13Cj2~1,2,3,4-TCDD injected (pg).
Qx * quantity of unlabeled 2,3,7,8-TCDD injected (pg).
D-ll
605
-------
RRF is a dimensionless quantity; the units used to express QIS, QRS and Qx
must be the same*
8.4 Criteria for Acceptable Calibration
The criteria listed below for acceptable calibration must be net
before analysis of any sample is performed.
8.4.1 The percent relative standard deviation (RSD) for the response
factors from each of the triplicate analyses for both unlabeled
and Cj2~2»3,7,8-TCDD must be less than ^20 percent.
8.4.2 The variation of the 5 mean RRFs for unlabeled 2,3,7,8-TCDD
obtained from the triplicate analyses must be less than _+20
percent RSD.
8.4.3 SIM traces for 2,3,7,8-TCDD must present a signal-to-noise
ratio of >2.5 for m/z 258.930 and MO for m/z 321.894.
8.4.4 SIM traces for Cj2~2,3,7,8-TCDD must present a signal-to-
noise ratio 2.10 for 333.934.
8.4.5 Isotopic ratios (Sections 8.3.4.3 and 8.3.4.4) must be within
the allowed range.
f * W » ^ •• W WV»**VfcBWAWtl WWIIWCvllk^a
5 triplicate determinations for
13C12-2,3,7,8-TCDD will be used
NOTE: If the criteria for acceptable calibration listed in Sections
8.4.1 and 8.4.2 have been met, the RRF can be considered inde-
pendent of the analyte quantity for the calibration concentra-
tion range. The mean RRF from
unlabeled 2,3,7,8-TCDD and for
for all calculations until routine'calibration criteria (Section
8.6). are no longer met. At such time, new mean RRFs will be
calculated from a new set of five triplicate determinations.
8.5 Routine Calibrations
Routine calibrations must be performed at the beginning of a 12-hour
period after successful mass resolution and GC column performance
check runs.
8.5.1 Inject 2 uL of the concentration calibration solution which
contains 5.0 pg/uL of unlabeled 2,3,7.8-TCDD, 10.0 pg/uL
of 13C12-2,3,7,8-TCDD and 5.0 pg/uL C12-l,2,3,4-TCDD.
Using the same GC/MS/DS conditions as used in Sections 8.1,
8.2 and 8.3, determine and document acceptable calibration as
provided in Section 8.6.
8.6 Criteria for Acceptable Routine Calibration
The following criteria must be met before further analysis is per-
formed. If these criteria are not met, corrective action must be
taken and the instrument must be recalibrated.
D-12
-------
8.6.1 The measured RRF for unlabeled 2,3,7,8-TCDD must be within +20
percent of the mean values established (Section 8.3.4.6) by
triplicate analyses of concentration calibration solutions.
8.6.2 The measured RRF for C,2~2,3,7,8-TCDD must be within +20 per-
cent of the mean value established by triplicate analysis
of the concentration calibration solutions (Section 8.3.4.6).
8.6.3 Isotopic ratios (Sections 8.3.4.3 and 8.3.4.4) must be within
the allowed range.
8.6.4 If one of the above criteria is not satisfied, a second attempt
can be made before repeating the entire initialization process
(Section 8.3).
NOTE: An initial calibration must be carried out whenever any HRCC
solution is replaced.
9. QUALITY CONTROL PROCEDURES
See Exhibit E for QA/QC requirements.
10. SAMPLE PRESERVATION AND HANDLING
10.1 Chain-of-custody procedures — see Exhibit G.
10.2 Sample Preservation
10.2.1 When received, each soil or sediment sample.will be contained
in a 1-pint glass jar surrounded by vermiculite in a sealed
metal paint can. Until a portion is to be removed for analysis,
store the sealed paint cans in a locked limited-access area
where the temperature is maintained between 25° and 35°C.
After a portion of a sample has been removed for analysis,
return the remainder of the sample to its original container
and store as stated above.
10.2.2 Each aqueous sample will be contained in a 1-liter glass
bottle. The bottles with the samples are stored at 4°C in a
refrigerator located in a locked limited-access area.
10.2.3 To avoid photodecomposition, protect samples from light.
10.3 Sample Handling
CAUTION: Finely divided soils contaminated with 2,3,7,8-TCDD are hazardous
because of the potential for inhalation or ingestion of particles
containing 2,3,7,8-TCDD. Such samples should be handled in a
confined environment (i.e., a closed hood or a glove box).
10.3.1 Pre-extraction sample treatment
D-13
(-"! n';'
vJ U (
-------
10.3.1.1 Homogenization — Although sampling personnel will
attempt to collect homogeneous samples, the contrac-
tor shall examine each sample and judge if it needs
further mixing.
NOTE: Contractor personnel have the responsibility to take a
representative sample portion; this responsibility
entails efforts to make the sample as homogeneous as
possible. Stirring is recommended when possible.
10.3.1.2 Centrifugation — When a soil or sediment sample
contains an obvious liquid phase, it must be
centrifuged to separate the liquid from the solid
phase. Place the entire sample in a suitable centri-
fuge bottle and centrifuge for 10 minutes at 2000 rpm.
Remove the bottle from the centrifuge. With a dis-
posable pipet, remove the liquid phase and discard
it. Mix the solid phase with a stainless steel
spatula and remove a portion to be air-dried, weighed
and analyzed. Return the remaining solid portion to
the original sample bottle and store it as described
in 10.2.1.
CAUTION: The removed liquid may contain TCDD and should be
disposed as a liquid waste.
10.3.1.3 Weigh between 9.5 and 10.5 g of the air-dried soil
sample (^0.5 g) to 3 significant figures. Dry it to
constant weight at 100°C. Allow the sample to cool
in a desiccator. Weigh the dried soil to 3 signifi-
cant figures. Calculate and report percent moisture
on Form B-l.
11. SAMPLE EXTRACTION
11.1 Soil Extraction
11.1.1 Immediately before use, the Soxhlet apparatus is charged
with 200 to 250 mL benzene which is then refluxed for 2 hours.
The apparatus is allowed to cool, disassembled and the benzene
removed and retained as a blank for later analysis if required.
11.1.2 Accurately weigh to 3 significant figures a 10-g (9.50 g to
10.50 g) portion of the wet soil or sediment sample. Mix 100
uL of the sample fortification solution (Section 7.8) with
1.5 mL of acetone (500 pg of C,2~2,3,7,8-TCDD) and deposit
the entire mixture in small portions on several sites on the
surface of the soil or sediment.
11.1.3 Add 10 g anhydrous sodium sulfate and mix thoroughly using a
stainless steel spoon spatula.
D-14
-------
11.1.4 After breaking up any lumps, place the soil-sodium sulfate
mixture in the Soxhlet apparatus using a glass wool plug (the
use of an extraction thimble is optional). Add 200 to 250 mL
benzene to the Soxhlet apparatus and reflux for 24 hours. The
solvent must cycle completely through the system at least 3
times per hour.
11.1.5 Transfer the extract to a Kuderna-Danish apparatus and
concentrate to 2 to 3 uL. Rinse the column and flask with 5 mL
benzene and collect the rinsate in the concentrator tube.
Reduce the volume in the concentrator tube to 2 to 3 uL.
Repeat this rinsing and concentrating operation twice more.
Remove the concentrator tube from the K-D apparatus and care-
fully reduce the extract volume to approximately 1 mL with a
stream of nitrogen using a flow rate and distance such that
gentle solution surface rippling is observed.
NOTE: Glassware used for more than one sample must be carefully
cleaned between uses to prevent cross-contamination (Note on
page D-10).
11.2 Extraction of Aqueous Samples
11.2.1 Mark the water meniscus on the side of the 1-L sample bottle
for later determination of the exact sample volume. Pour
the entire sample (approximately 1 L) into a 2-L separatory
funnel.
11.2.2 Mix 100 uL of the sample fortification solution with 1.5 mL
of acetone (500 pg of 13C12-2,3,7,8-TCDD) and add the mixture
to the sample in the separatory funnel.
NOTE: A continuous liquid-liquid extractor may be used in place of
a separatory funnel.
11.2.3 Add 60 mL methylene chloride to the sample bottle, seal and
shake 30 seconds to rinse the inner surface. Transfer the
solvent to the separatory funnel and extract the sample by
shaking the funnel for 2 minutes with periodic venting.
Allow the organic layer to separate from the water phase for
a minimum of 10 minutes. If the emulsion interface between
layers is more than one-third the volume of the solvent
layer, the analyst must employ mechanical techniques to
complete the phase separation. Collect the methylene
chloride (3 x 60 mL) directly into a 500 mL Kuderna-Danish
concentrator (mounted with a 10 mL concentrator tube) by
passing the sample extracts through a filter funnel packed
with a glass wool plug and 5 g of anhydrous sodium sulfate.
After the third extraction, rinse the sodium sulfate with an
additional 30 mL of methylene chloride to ensure quantitiative
transfer.
D-15
609
-------
11.2.4 Attach a Snyder column and concentrate the extract until
the apparent volume of the liquid reaches 1 mL. Remove the
K-D apparatus and allow it to drain and cool for at least
10 minutes. Remove the Snyder column, add 50 mL benzene,
reattach the Snyder column and concentrate to approximately
1 mL. Rinse the flask and the lover joint with 1 to 2 mL
benzene. Concentrate the extract to 1.0 mL under a gentle
stream of nitrogen.
11.2.5 Determine the original sample volume by refilling the sample
bottle to the mark and transferring the liquid to a 1000-ml
graduated cylinder. Record the sample volume to the nearest
5 mL.
11.3 Cleanup Procedures - Column Cleanup
11.3.1 Prepare an acidic silica column as follows: Pack a 1 cm x 10
cm chromatographic column with a glass wool plug, a layer (1
cm) of Na2804/K2CC^d :1) , 1.0 g silica gel (Section 7.1.2) and
4.0 g of 40-percent w/w sulfuric acid-impregnated silica gel
(Section 7.1.3). Pack a second chromatographic column (1 cm x
30 cm) with a glass wool plug, 6.0 g acidic alumina (Section
7.1.1) and top with a 1-ctn layer of sodium sulfate (Section
7.4). Add hexane to the columns until they are free of
channels and air bubbles.
11.3.2 Quantitatively transfer the benzene extract (1 mL') from the
concentrator tube to the top of the silica gel column. Rinse
the concentrator tube with two 0.5-mL portions of hexane.
Transfer the rinses to the top of the silica gel column.
11.3.3 Elute the extract from the silica gel column with 90 mL hexane
directly into a Kuderna-Danish concentrator. Concentrate the
eluate to 0.5 mL, using nitrogen blow-down as necessary.
11.3.4 Transfer the concentrate (0.5 mL) to the top of the alumina
column. Rinse the K-D assembly with two 0.5-mL portions of
hexane and transfer the rinses to the top of the alumina
columns. Elute the alumina column with 18 mL hexane until the
hexane level is just below the top of the sodium sulfate.
Discard the eluate. Columns must not be allowed to reach
dryness (i.e., a solvent "head" must be maintained.)
11.3.5 Place 30 mL of 20-f-percent (v/v) methylene chloride in hexane
on top of the alumina and elute the TCDDs from the column.
Collect this fraction in a 50-mL Erlenmeyer flask.
11.3.6 Certain extracts, even after cleanup by column chromatography,
contain interferences which preclude determination of TCDD
at low parts-per-trillion levels. Therefore, a cleanup step
is included using activated carbon which selectively retains
planar molecules such as TCDD. The TCDDs are then removed
D-16
GIG
-------
from the carbon by elution with toluene. Proceed as follows:
Prepare a 18-percent Carbopak C/Celite 545® mixture by
thoroughly mixing 3.6 grams Carbopak C (80/100 mesh) and 16.4
grams Celite 545® in a 40-mL vial. Activate at 130°C for 6
hours. Store in a desiccator. Cut off a clean 5-mL disposable
glass pipet at the 4-mL mark. Insert a plug of glass wool
(Section 7.3) and push to the 2-mL mark. Add 340 mg of the
activated Carbopak/Celite mixture followed by another glass
wool plug. Using two glass rods, push both glass wool plugs
simultaneously towards the Carbopak/Celite mixture and gently
compress the Carbopak/Celite plug to a length of 2 to 2.5 cm.
Preelute the column with 2 mL toluene followed by 1 mL of
75:20:5 methylene chloride/raethanol/benzene, 1 mL of 1:1
cyclohexane in methylene chloride, and 2 mL hexane. The flow
rate should be less than 0.5 mL min."*. While the column is
still wet with hexane, add the entire eluate (30 mL) from the
alumina column (Section 11.3.5) to the top of the column.
Rinse the Erlenmeyer flask which contained the extract twice
with 1 mL hexane and add the rinsates to the top of the column.
Elute the column sequentially with two 1-mL aliquots hexane, 1
mL of 1:1 cyclohexane in methylene chloride, and 1 mL of
75:20:5 methylene chloride/ methanol/benzene. Turn the column
upside down and elute the TCDD fraction with 6 mL toluene into
a concentrator tube. Warm the tube to approximately 60°C and
reduce the toluene volume to approximately 1 mL using a stream
of nitrogen. Carefully transfer the residue into a 1-mL
mini-vial and again at elevated temperature, reduce the volume
to about 100 uL using a stream of nitrogen. Rinse the concen-
trator tube with 3 washings using 200 uL of 12 toluene in
CH2C12- Add 50 uL tridecane and store the sample in a refrig-
erator until GC/MS analysis is performed.
12. ANALYTICAL PROCEDURES
12.1 Remove the sample extract or blank from storage, allow it to warm to
ambient laboratory temperature and add 5 uL recovery standard solution.
With a stream of dry, purified nitrogen, reduce the extract/blank
volume to 50 uL.
12.2 Inject a 2-uL aliquot of the extract into the GC, operated under the
conditions previously used (Section 8.1) to produce acceptable results
with the performance check solution.
12.3 Acquire SIN data according to 12.3.1. Use the same acquisition time
and MS operating conditions previously used (Section 8.3.4) to deter-
mine the relative response factors.
12.3.1 Acquire SIM data for the following selected characteristic ions:
D-17
611
-------
Compound
TCDD - COC1
Unlabeled TCDD
Unlabeled TCDD
13C12-2,3,7,8-TCDD, 13C12-1,2,3,4-TCDD
13C12-2,3,7,8-TCDD, 13C12-1,2,3,4-TCDD
12.4 Identification Criteria
12.4.1 The retention time (RT) (at maximum peak height) of the sample
component m/z 319.897 must be within -1 to «3 seconds of the
retention time of the peak for the isotopically labeled internal
standard at m/z 331.937 to attain a positive identification of
2,3,7,8-TCDD. Retention times of other tentatively identified
TCDDs must fall within the RT window established by analyzing
the column performance check solution (Section 8.1). Retention
times are required for all chromatograms.
12.4.2 The ion current responses for m/z 258.930, 319.897 and 321.894
must reach maximum simultaneously (+_ 1 scan), and all ion
current intensities must be 2. 2.5 times noise level for
positive identification of a TCDD.
12.4.3 The integrated ion current at m/z 319.897 must be between 67
and 90 percent of the ion current response at m/z 321.894.
12.4.4 The integrated ion current at m/z 331.937 must be between 67
and 90 percent of the ion current response at m/z 333.934.
12.4.5 The integrated ion currents for m/z 331.937 and 333.934 must
reach their maxima within +_ 1 scan.
12.4.6 The recovery of the internal standard 13C12~2,3,7,8-TCDD must
be between 40 and 120 percent.
13. CALCULATIONS
13.1 Calculate the concentration of 2,3,7,8-TCDD (or any other TCDD isoraer)
using the formula:
Ax '
cx »
AIS • W • RRF(I)
D-18
812
-------
where:
GX * unlabeled 2,3,7,8-TCDD (or any other unlabeled TCDD isomer) concen-
tration in pg/g for soil/sediment and pg/L for aqueous samples.
AX = sum of the integrated ion abundances determined for m/z 319.897
and 321.894.
Ajg = sum of the integrated ion abundances determined for m/z 331.937
and 333.934 of C12~2,3,7,8-TCDD (IS » internal standard).
QIS - quantity (in picograms) of 13Cj2~2,3,7,8-TCDD added to the
sample before extraction (Qxs ° 500 pg).
W = weight (in grams) of dry soil or sediment sample or volume of
aqueous sample (in liters).
RRF(I) » calculated mean relative response factor for unlabeled 2,3,7,8TCDD
relative to 13C,2-2,3,7,8-TCDD. This represents the grand
mean of the RRF(I)'s obtained in Section 8.3.4.5.
13.2 Calculate the recovery of the internal standard 3Cj2~2, 3,7,8-TCDD,
measured in the sample extract, using the formula:
Internal standard A,g * Q__
percent recovery * ———^—————— x 100
ARS ' RRF(II) ' Q1S.
where Ajg and Q^g have the same definitions as above (Section 13.1)
= sum of the integrated ion abundances determined for m/z 331.937
and 333.934 of C12~l,2,3,4-TCDD (RS = recovery standard).
QRS = quantity (in picograms) of 3Cj2~1,2,3,4-TCDD added to the sample
residue before HRGC-HRMS analysis.
(QRS " 500 pg).
RRF(II) = calculated mean relative response factor for labeled 3Cj2-2,3,7,8-
TCDD relative to 13C12-1,2,3,4-TCDD. This represents the grand
mean of the RRF(II)'s calculated in Section 8.3.4.5.
13.3 If the calculated concentration of unlabeled 2,3,7,8-TCDD exceeds
200 pg/g for soils or sediments, or 2000 pg/L for aqueous samples,
the linear range of response vs. concentration may have been exceeded
and a smaller portion of that sample must be analyzed. Accurately
weigh to three significant figures a 1-g portion of the wet soil/
sediment. Add the sample fortification solution (Section 11.1.2),
extract and analyze as discussed for the 10-g sample. Similarly,
add the sample fortification solution (Section 11.2.2) to 100 mL of
the aqueous sample, extract and analyze.
D-19
G13
-------
13.4 Total TCDD concentration — all positively identified isoraers of TCDD
must be within the RT window and meet all identification criteria
listed in Sections 12.4.2, 12.4.3 and 12.4.4. Use the expression
in Section 13.1 to calculate the concentrations of the other TCDD
isomers, with C^ becoming the concentration of any unlabeled TCDD
i s one r.
c Total TCDD - Sum of the concentrations of the individual TCDDs.
13.5 Estimated Detection Limit — For samples in which no unlabeled
2,3,7,8-TCDD was detected, calculate the estimated minimum detectable
concentration. The background area is determined by integrating the
ion abundances for m/z 319.897 and 321.894 in the appropriate region
of the selected ion monitoring trace, multiplying that area by 2.5,
and relating the product area to an estimated concentration that
would produce that product area.
Use the formula:
(2.5) • (Ax) ' (QIS)
where
(AIS) • (RRF(I)) • (W)
* estimated concentration of unlabeled 2,3,7,8-TCDD required to
produce Ax.
* sum of integrated ion abundance for m/z 319.897 and 321.894 in the
same group of >5 scans used to measure
8um °f integrated ion abundance for the appropriate ion character-
istic of the internal standard, m/z 331.937 and m/z 333.934.
QjS» RRF(I), and W retain the definitions previously stated in Section 13.1.
Alternatively, if peak height measurements are used for quantification, measure
the estimated detection limit by the peak height of the noise in the TCDD RT
window.
13.6 The relative percent difference (RPD) is calculated as follows:
I Sj - S2 | | Si - S2 |
RPD - — — — — — — — - - x 100
Mean Concentration (Si + 82) 12
Si and $2 represent sample and duplicate sample results.
References
1. "Carcinogens - Working with Carcinogens", Department of Health, Education
and Welfare, Public Health Service, Center for Disease Control, National
Institute for Occupational Safety and Health, Publication No. 77-206, Aug.
1977.
D-20
G14
-------
2. "OSHA Safety and Health Standards, General Industry" (29 CFR1910),
Occupational Safety and Health Administration, OSHA 2206 (Revised January
1976).
3. "Safety in Academic Che'nistry Laboratories", American Chemical Society
Publication, Committee on Chemical Safety, 3rd Edition 1979.
D-21
Giro
-------
TABLE 1. COMPOSITION OF CONCENTRATION CALIBRATION SOLUTIONS
HRCC1
HRCC2
HRCC3
HRCC4
HRCC5
Recovery Standard
13C12-1,2,3,4-TCDD
2.5 pg/uL
5.0 pg/uL
10.0 pg/uL
20.0 pg/uL
40.0 pg/uL
Analyte
2,3,7,8-TCDD
2.5 pg/uL
5.0 pg/uL
10.0 pg/uL
20.0 pg/uL
40.0 pg/uL
Internal Standard
13C12-2,3,7,8-TCDD
10.0 pg/uL
10.0 pg/uL
10.0 pg/uL
10.0 pg/uL
10.0 pg/uL
Sample Fortification Solution
5.0 pg/uL of 13C12-2,3,7,8-TCDD
Recovery Standard Spiking Solution
100 pg/uL 13C12-1,2,3,4-TCDD
Field Blank Fortification Solutions
A) 5.0 pg/uL of unlabeled 2,3,7,8-TCDD
B) 5.0 pg/uL of unlabeled 1,2,3,4-TCDD
Internal Standard Spiking Solution
100 pg/uL of 13C1?-2,3,7,8-TCDD
(Used only in Section 4.2.1.A.2, Exhibit E)
D-22
GiG
-------
TABLE 2. RECOMMENDED GC OPERATING CONDITIONS
Column coating
Film thickness
Column dimensions
Helium linear velocity
Initial temperature
Initial time
Temperature program
2,3,7,8-TCDD retention
time
SP-2330
0.2 urn
60 m x 0.24 mm
28-29 cm/sec
at 240eC
70°C
4 min
Rapid increase to 200°C
200°C to 250°C
at 4°C/min
24 min
CP-SIL 88
0.22 urn
50 m x 0.22 mm
28-29 cm/sec
at 240°C
45°C
3 min
Rapid increase to 190°C
190°C to 240°C
at 5°C/min
26 min
D-23
6 1'V
-------
TABLE 3. TYPICAL 12-HOUR SEQUENCE FOR 2,3,7,8-TCDD ANALYSIS
1. Static mass resolution check
2. Column performance check
3. HRCC2
4. Sample 1 through Sample "N"
5. Column performance check
6. Static mass resolution check
10/20/84
10/20/84
10/20/84
10/20/84
10/20/84
10/20/84
0700 hrs.
0730 hrs.
0800 hrs.
0830 hrs.
1800 hrs.
1830 hrs.
D-24
-------
IM
r-*
I*
2J7I
CT-SIL N
11)4
127t
I2CI
IN
IM
IM
Figure 1. Selected ion current profile for ra/r 320 and 322 produced by MS analysis for
performance check solution using a 50-m CP Si 1-88 fused silica capillary
column and conditions listed in Table 2.
-------
IM
J
o>
u
u
2371
I2M
IIM
IM
» I ' '
•M
tit
440
•M
Figure 2. Selected ion current profile for m/z 320 and 322 produced by MS analysis of performance
check solution using a 60-m SP-2330 fused silica capillary column and conditions
listed in Table 2.
-------
APPENDIX B
PROPOSED ANALYTICAL PROTOCOL
for the Determination of
2,3,7,8-Tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) and Total
TCDDs in Soil/Sediment and Water by High-Resolution
Gas Chromatography/High-Resolution Mass Spectrometry
December 1, 1985
This analytical protocol has been written in the format used in the
Superfund program, as "Exhibit D" of a Statement of Work which in turn is part
of an Invitation-for-Bid package under the Superfund Contract Laboratory Program.
Also included are other exhibits listed below for the Statement of Work which
have been tailored to meet the specific requirements of this protocol:
EXHIBIT B: Reporting Requirements and Deliverables
EXHIBIT C: Sample Rerun Requirements
EXHIBIT D: Analytical Method
EXHIBIT E: Quality Assurance/Quality Control Requirements
821
-------
This protocol (Protocol B) is a modification rf the protocol presented as
Appendix A (Protocol A). Examination of the results from the single-laboratory
evaluation of Protocol A had shown that the minimum amount of 2,3,7,8-TCDD that
could be quantified under the conditions specified in Protocol A was 5 pg.
However, a requirement existed to lower the quantitat ion limits to 2 ppt for
soil and sediment samples and to 20 ppq for aqueous samples. The sample size
should stay at 10 g for soil and sediments and at 1 L for aqueous samples,
since the effect of larger sample sizes on the extract cleanup efficiencies is
not known. Also, the range of the method should overlap with the 1-ppb lower
limit of the low-resolution analytical method for TCDD used in the Superfund
Contract Laboratory Program without necessitating second extractions for samples
containing higher levels of TCDDs.
After careful evaluation by EMSL-LV of the requirements and the options,
the following protocol changes were made:
* In Protocol B, the following calibration solutions will be used:
HRCC1: 2 pg/uL 2,3,7,8-TCDD and 13C12-1,2,3,4-TCDD
10 pg/uL 13Cl2-2,3,7,8-TCDD
HRCC2: 10 pg/wL 2,3,7,8-TCDD and 13C12-1,2,3,4-TCDD
10 pg/uL 13Cl2-2,3,7,8-TCDD
HRCC3: 50 pg/pL 2,3,7,8-TCDD and 13C12-1,2,3,4-TCDD
10 pg/uL 13Cl2-2,3,7,8-TCDD
-------
HRCC4: 100 pg/yl_ 2,3,7,8-TCDD and 13C12-1,2,3,4-TCDD
10 pg/yL 13C12-2,3,7,8-TCDD
o In Protocol B, the final extract volume will be 10 yL. The decision
to select a final volume of 10 yL was necessary in order to comply
with the above requirements. It is realized that such a small volume
may pose technical difficulties for the analyst.
0 In Protocol B, the fortification level of the internal standard
13C12-2,3,7,8-TCDD was raised from 500 pg/sample to 1,000 pg/sample.
This allows analysis of soil and sediment samples containing between
100 ppt and 1.2 ppb of any TCDD isomer and of water samples containing
between 1 ppt and 12 ppt of any TCDD isomer by diluting a 2-yL aliquot
of the remaining extract concentrate by a factor of 12 with a solution
of the recovery standard (100 pg/yL of 13C12-1,2,3,4-TCDD in tridecane)
Recoveries will be reported using the data generated from the first
injection. Thus, the decision to dilute an aliquot of the 10-yL final
extract will not be based on the concentration of 2,3,7,8-TCDD or total
TCDD in the sample, but on the concentration of the most abundant TCDD
isomer in the 10-yL final extract volume. This will eliminate un-
necessary dilutions of the sample extract and analyses for samples
containing between 100 ppt and 250 ppt for soil and sediment and 1 ppt
and 2.5 ppt for water samples of a TCDD isomer, but for which the
recoveries were low.
623
-------
EXHIBIT B
Reporting Requirements and Deliverables
G24
-------
1. SCOPE AND APPLICATION
The Contractor shall provide reports and other deliverables as specified
in the Contract Reporting Schedule. These reports are described below.
All reports shall be submitted in legible form or resubmission shall be
required. All reports and documentation required, including selected ion
current profiles (also called selected ion monitoring traces), shall be
clearly labeled with the Sample Management Office Case number and associated
Sample/Traffic Report number(s). If documentation is submitted without
the required identification, as specified above, resubmission shall be
required.
The Contract Reporting Schedule (Section 2) specifies the numbers of
copies required, the delivery schedule and the distribution of all required
deliverables.
1.1 Sample data package — Hard copy analytical data and documentation
are required as described below.
NOTE: This analytical protocol is designed for the receipt and analysis
of samples by batches. Therefore, it is desired that sample data
from samples in the same batch be reported together, i.e., on the
same reporting form. However, contract accounting and billing are
based on the sample unit.
1.1.1 Case narrative: Contains the Case number, Dioxin Shipment
Record numbers, Contract number and detailed documentation of
any quality control, sample, shipment and/or analytical pro-
blems encountered in a specific Case. Also included should be
documentation of any internal decision tree process used along
with a summary of corrective actions taken. The Case narrative
must be signed in original signature by the Laboratory Manager
or his designate.
1.1.2 Results of initial triplicate analyses of four (4) concentration
calibration solutions (Form H-2), routine calibration solutions,
(Form H-3), including all selected ion current profiles or
selected ion monitoring (SIM) traces, calculated relative
response factors (RRF), and computer-generated quantification
reports (or manual calculations).
1.1.3 Completed data reporting sheets (Forms H-l, H-4, and H-5, H-8
and H-9) with appropriate SIM traces (including the lock mass
SIM traces). Data results for levels less than 10 ppt but
above the quantitation limit (Section 1.1, Exhibit D) attained
for that sample shall be reported to two (2) significant
figures; results greater than 10 ppt shall be reported to three
(3) significant figures. Apply the rounding rules found in
Section 7.2.2, "Handbook for Analytical Quality Control in Water
and Wastewater Laboratories," EPA-600/4-79-019. Each SIM trace
shall include computer-generated header information indicating
instrumental (GC and MS) operating parameters during data
B-l
G25
-------
acquisition. When samples are analyzed more Chan once, all
sample data shall be reported. Rejected sample runs must be
separated and attached to the back of the data package and
marked on the SIM trace as "Rejected," with an explanation of
the reasons for the rejection.
1.1.A SIM traces generated during each GC column performance check
analysis; peak profile outputs of the reference signal used
to document the mass resolution.
1.1.5 Documentation of acceptable MS calibration (Section 8, Exhibit
D, and Exhibit E) for each confirmatory analysis. As
applicable, submit peak matching box settings and calculations
for accurate mass assignments and any other related printouts.
State, in ppm, the level of mass accuracy achieved (Section
8.2.2, Exhibit D).
1.1.6 A chronological list of all analyses performed (Form H-6). If
more than one GC/MS system is used, a chronological list is
required for each system. The list must provide the Data
System File name, the EPA sample number, and (if appropriate)
the contractor laboratory sample number for each sample,
blank, concentration calibration solution, performance check
solution, or other pertinent analytical data. This list shall
specify date and time of beginning of analysis. All sample/
blank analyses performed during a 12-hour period, must be
accompanied by two GC column performance check solution ana-
lyses, one preceding and one following the sample/blank ana-
lyses. If multiple shifts are used, the ending GC column
performance check sample analysis from one 12-hour period
shall serve as the beginning analysis for the next 12-hour
period; see Exhibit D, Section 8, for system performance
criteria. The same schedule applies to the mass resolution
check analysis. See Section 8.2.2, Exhibit D.
1.1.7 Verification of recovery of TCDDs from cleanup columns
(Section 11.3, Exhibit D, and Section 4.2.1.2.2, Exhibit E).
1.2 Sample extracts and unused sample portions — Unused portions of
samples and sample extracts shall be retained by the Contractor for
a period of six months after receipt. When directed in writing by
the Project Officer (PO) or Sample Management Office (SMO), the
Contractor shall ship (not at Contractor's expense but in accordance
with Department of Transportation Regulations) specific samples
and/or extracts to specified locations and persons. After six months,
upon obtaining PO or SMO clearance, remaining samples and extracts
shall be disposed of by the Contractor at Contractor's expense, in
accordance with applicable regulations concerning the disposal of
such materials.
1.3 Document Control and Chain-of-Custody Package — The Document Control
and Chain-of-Custody Package includes all laboratory records received
B-2
-------
or generated for a specific case, that have not been previously
submitted to EPA as a deliverable. These items include but are not
limited to: sample tags, custody records, sample tracking records,
analysts logbook pages, bench sheets, chromatographic charts, computer
printouts, raw data summaries, instrument logbook pages, corre-
spondence, and the document inventory (Exhibit G).
NOTE: Pages from logbooks or bench sheets kept exclusively in a high-
hazard area (containment facility) need not be copied.
1.4 Monthly Sample Status Report — The Monthly Sample Status Report
shall provide the status of all samples the Contractor has received
or has had in-house during the calendar month. Required status
information includes: samples received, samples extracted, samples
analyzed, samples rerun, and samples which required special cleanup.
All samples shall be identified by the appropriate EPA sample, case
and batch/shipment numbers.
1.5 Daily Sample Status Report — In response to a verbal request from the
Sample Management Office or the Project Officer, the Contractor must
verbally provide sample status information on a same-day basis.
Should written confirmation be requested, the Contractor must send the
daily sample status information in a written form that same day using
first-class mail service. The required Daily Sample Status informa-
tion shall include the items noted for the Monthly Sample Status
Report and, in addition, shall require information on sample analysis
reports in progress and analysis reports submitted/mailed.
2. In accordance with applicable delivery requirements, the Contractor shall
deliver specified items per the following Contract Reporting Schedule
(Section 2.1). Recipients include the CLP Sample Management Office, the
EMSL/LV QA Division, the appropriate Regional Technical Officer and NEIC.
2.1 Contract Reporting Schedule
CONTRACT REPORTING SCHEDULE
Item Delivery Report Distribution
No. Report No. Copies Schedule SMO EMSL/LV Region NEIC
1 Sample Data 3 30 days after validated XX X
Package sample receipt date
-OR-
10 days after initial XXX
data due date
2 Sample Extracts Within 180 days after As directed
analysis, 7 days after
request by Project Officer
or SMO
(Continued)
B-3
627
-------
CONTRACT REPORTING SCHEDULE (Continued)
Item
No.
Report
No. Copies
Delivery
Schedule
Report Distribution
SMO
EMSL/LV
Region
NEIC
Document 1
Control & Pkg
Chain-of-
Custody Package
Monthly 2
Sample Status
Report
7 days after request by
Project Officer
or SMO
5 days following end of
each calendar month
Daily Sample
Status Report
Verbal and/or written
upon request by SMO or PO;
maximum frequency is daily.
As directed
NOTE: All results shall be reported total and complete.
2.2 Addresses for distribution
SMO
EMSL-LV
NEIC
CLP Sample Management Office
P. 0. Box 818
Alexandria, VA 22313
For overnight deliveries, use
street address:
300 N. Lee St., Suite 200
Alexandria, VA 22314
US EPA EMSL-LV QA Division
Box 15027
Las Vegas, NV 89114
Attn: Data Audit Staff
US EPA NEIC
Bldg. 53
Box 25227
Denver Fed. Center
Denver, CO 80225
For overnight deliveries, use
street address:
944 E. Harmon Ave.
Executive Center
Las Vegas, NV 89109
3.
Regional Technical Officer — Following contract award and prior to
Contractor's receipt of the first batch of samples, the Sample Manage-
ment Office will provide the Contractor with the list of Technical
Officers for the ten EPA Regions. SMO will provide the Contractor
with updated Regional address/name lists as necessary throughout the
period of the contract.
FORM INSTRUCTION GUIDE
This section includes specific instructions for the completion of all
required forms. These include instructions on header information as
well as specific details to the bodies of individual forms. Instructions
are arranged in the following order:
B-4
-------
Data Summary (Form H-l)
Initial Calibration Summary (Form H-2 ; 2 pages)
Routine Calibration Summary (Form H-3)
GC and Mass Resolution Check Summary (Form H-4)
Quality Control Summary (Form H-5)
Chronological List of All Analyses Performed (Form H-6)
GC Operating Conditions (Form H-7)
HRMS TCDD Calibration Report Form (Form H-8)
High-Resolution MS TCDD Data Report Form (Form H-9)
3.1 Data Summary (Form H-l) — This form is used for summarizing the
results from all samples in the batch. The detailed results are
available on Form H-8 for each sample.
Complete the header information at the top of the page, including
laboratory name, case number and batch/shipment number (from the
dioxin shipment record), and matrix (soil, sediment, water).
Complete the form using one horizontal row for each sample.
The SMO sample number should be suffixed with the appropriate letter
code as needed.
The TCDD retention time should be reported in minutes and seconds.
TCDD levels are reported as parts per trillion (ppt) regardless of
the matrix. Total TCDD concentration (in ppt) is the sum of the
concentrations of all TCDDs reported on Form H-9;
The S/N criteria apply to m/z 259, 320, 322 (for unlabeled TCDD)
and m/z 322 and 334 (internal and recovery standards). The symbols
used are: (+) all S/N ratios are 2.5 or greater including all TCDDs
present, (-) S/N ratio for native 2 ,3,7 ,8-TCDD, the internal or the
recovery standard are less than 2.5, (0) other suspected TCDDs are
present but did not meet the S/N criteria.
The file name is the HRGC/HRMS file name and is used for tracking
results and raw data.
The comments column should be used for any remarks specific to a
particular sample.
3.2 Initial Calibration Summary (Form H-2): Page 1 — The header infor-
mation should be filled in. The column headings are similar to those
on Form H-l.
Ax '
RRF(I) =
Q ' A
IS
B-5
G29
-------
RRF(II) » (Section 8.3.4.5, Exhibit D)
QlS ' ARS
Page 2 — The header information should be filled in. For each RRF,
the mean, percent relative standard deviation (%RSD) and number of
runs (N) are reported; N must be at least three (3) for each HRCC
solution. The grand means (RRFs) are the mean of the individual
means and are reported with their XRSD and N. The routine calibra-
tion relative response factor permissible ranges are also reported
(Section 8.3.4.8, Exhibit D).
3.3 Routine Calibration Summary (Form H-3) — The header information
Includes case and batch numbers in addition to the laboratory and
instrument identification.
The columns are the same as on Page 1 of Form H-2. The results
reported are for the routine calibration runs rather than the initial
calibration. The calculated RRF(I) and RRF(II) must be within the
routine calibration relative response factor permissible ranges
(Section 8.3.4.8, Exhibit D) and other criteria listed in Section
8.6, Exhibit D must be met before further analysis is performed.
3.4 GC and Mass Resolution Check Summary (Form H-4) — The header informa-
tion should be filled in. The TCDD isoner resolution (% valley) is
measured from the column performance check solution (Section 8.1.2,
Exhibit D). The resolving power and mass measurement error are measured
using PFK (or equivalent) (Section 8.2, Exhibit D).
3.5 Quality Control Summary (Form H-5) — The items should be completed
as indicated. The "other Interferences" should be included even if
they only occur at one mass.
Form H-5 in conjunction with Form H-9 is used to report results
relative to the fortified field blank pair and rinsate analyses.
The total TCDD retention time window is a window that includes all of
the TCDD isomers and is based on the first and last eluting isomers
in the GC column performance check solution using the conditions sum-
marized in Form H-7. All materials used should be recorded in the
standard/reagent QC table. Standards provided by EPA should be
listed, however, the QC columns may be left blank as these are refer-
ence materials.
3.6 Chronological List of All Analyses Performed (Form H-6) — The
header information should be filled in. If more than one instrument
is used, use one form per instrument.
The "Analysis Identification" column should contain enough information
for the data user to clearly identify the analysis, i.e., HRCC 2
Routine Calibration, Fortified Field Blank A, Fortified Field Blank B,
B-6
-------
Reanalysis of Sanple //I, 2, 3, 4, etc. The "SMO //" column should be
used only for samples etc. which have an assigned SMO sample number.
3.7 GC Operating Conditions (Form H-7) — This form must be filled out to
describe the GC operating conditions used to analyze a batch of
samples and to analyze the GC performance evaluation check solution.
3.8 HRMS TCDD Calibration Report (Form H-8) — This form is to be filled
in for each initial and routine calibration analysis made. It will be
the first page of the chrociatograms and calculations for that analysis.
It is suggested that this form be used as a worksheet for completing
Forms H-2 and H-3. S/N ratios greater than five (5) may be reported
with a (+); S/N ratios of five or less must have a numerical value
reported with accompanying chromatograms scaled so that the measure-
ments may be checked by the data user.
3.9 High-Resolution MS TCDD Data Report (Form H-9) — This form contains
the details of the data reported in summary on Form H-l. It will be
the first page of the chrociatograms and calculations for each sample
including the fortified field blank pair samples. All data presented
(retention times, areas, and S/N ratios) must also be available on
the accompanying chromatograms. The chromatograms must be scaled
so that the data user may check any S/N ratios that are near or below
five (5). It is suggested that this form be used as a worksheet for
completing Form H-l.
4. REPORTING REQUIREMENTS SUMMARY:
Items that must be included with the data package:
4.1 Complete identification of the samples analyzed (sample numbers and
type).
4.2 The dates and times at which all analyses were accomplished. This
information should also appear on each selected ion current profile
included with the report.
4.3 Raw mass chromatographic data which consist of the absolute peak
heights or peak areas of the signals observed for the ion masses
monitored.
4.4 The calculated ratios of the intensities of the M+° to (M+2)+°
molecular ions for all TCDD isoroers detected.
4.5 The calculated concentrations of native 2,3,7,8-TCDD and other TCDD
isomers for each sample analyzed, expressed in picograms TCDD per gram
of sample (that is, parts per trillion), as determined from the raw
data. If no TCDDs are detected, the notation "Not Detected" or
"N.D." is used, and the minimum detectable concentrations (or detection
limits) are reported.
B-7
G 31
-------
HIGH RESOLUTION
FORM H-l DATA SUMMARY
HRGC/HRMS DIOXIN ANALYSIS
Lib:
Case*
Bitcn/Shipnwiit It.
Matrix:
8MO
Number
TCOD
2.3.7.8 (IS)
PPt
2.3.7.8-TCDD
Mea*. DL
Relative Ion
opt Abundance Ratio*
Total 32fl 332 J2i %Rac. 8/N Inat. Ane»y*t* Ha
TCOD 322 334(18) 334(RS) (IS) Criteria ID Data Time Name Comment*
w
I
00
RB - H>«g«nl Blank
N- Untobetod TCDD Spfta
D - Duplicate
FB- FWd Blank
8R • Sampte Reran
ER - Extract Reanahr*!*
NO- NotDatectwl
DL- Detection Umit
MB- Method Blank
Roc • Recovery
Matrix: S - Soil '
W- Water
O- Other
S/N Criteria: report (») rf aN S/N > 2.6
report (-) If 2.3,7.8 TCDO.
"C,f 2,3.7,8 TCDO or
"C1f-1.2,3.4-TCOOS/N < 2.5
report (0) if other TCOD* are auepected
not to meet criteria
-------
HIGH RESOLUTION
FORM H-2 INITIAL CALIBRATION SUMMARY
pa«alo<2
Lab:
Contract *:
Instrument ID:
i
VO
o
Co
Calibration File m/i m/t m/i S/N
Standard Name Data Tlma 320/322* 332/334(18)* 332/334(RS)* Criteria RRF(l|b RRF|II)C Cornnenta
* Ion ratios mud be in the range of 0.67 to 0.90
b 2.3.7.8-TCOD veraut "C,j-2.3.7.8 TCDD
S/NCritaria: report |+| if greater than 2.6
report (•) rl toM than 2.6
"C,, 2.3.7.8 TCDD wriu* '
'C,2 1.2.3.4 TCOD
-------
HIGH RESOLUTION
FORM H-2 INITIAL CALIBRATION SUMMARY
paoa2o<2
lab:
Date of Inthial Calibration:
Contract #:
Instrument ff:
RRF (I) Mean %RSD
RRF (II) Maan % RSO
HRCC1
HRCC2
HRCC3
HRCC4
eo
i
RRF (I) Grand Maan:
%RSD:
N:
RRF (II) Grand MOOT:
HR80:
N:
Routine Calibration Pa
CT;
CO
RRF (l| = 2.3.7.8 TCDD v«
"€,,-2.3.7.81000
aiblaRanga:
RoutinaCattN
Ma Rang*:
RRF (111 = "C -2.3.7.B-TCDD wt
»C.,-1.2.3.4-TCOD
-------
HIGH RESOLUTION
FORM H-3 ROUTINE CALIBRATION SUMMARY
Lab:
Case ft:
Batch ff:
Instrument ID:
C73
CO
Ol
w
i
Calibration File m/t m/t m/i S/N
Standard Him* Date Time 320/322* 332/334(15)* 332/334C,, 1.2.3.4 TCOO
-------
HIGH RESOLUTION
FORM H-4 GC AND MASS RESOLUTION CHECK SUMMARY
Ub:
C«M*
Batch #
Oat*
bi«t.
ID
Sol.
IO
Ton*
TCDO Iwrrwr RMorving *MUw
File RMolution Power Measurement
Nam* (%VaU*y) atlOSValtoy Error (PPM)
•Me** ueed for HIM* menurement i
B-12
-------
HIGH RESOLUTION
FORM H-5 QUALITY CONTROL SUMMARY
Lab:
Case*
Batch #
Number of sample* in batch: _
Mean S of recovery for the I.S.:
0 of data points:
Fortified field blank A. S recovery ("C,2-2.3.7.8-TCDD):
No Yes
D D
SMO Sample »:
Contamination by 1.2.3.4-TCOO
"C12-1.2.3.4-TCDD
Estimated
Concentration (ppt)
Other interferences:
Retention times:
Estimated concentrations (ppt):
Fortified field blank B. * recovery (taC12-1.2.3.4-TCDD): -
SMO Sample «:
"C12-2.3.7.8-TCDD
BYes
[j
Estimated
Concentration (ppt)
Other interferences:
Retention times:
Estimated concentrations (ppt):
Rinsate. % recovery:
SMO sample «:
Contamination by 2.3.7.8-TCDD
Other TCOO
Duplicate anarysis. SMO sample «:
Estimated
Concentration
(pg/mL)
"C12-2.3.7.8-TCDD Mean Recovery:
Percent Relative difference "C12-2.3.7.8-TCDD (Recovery)
Percent relative difference t*C12-2.3.7.8-TCDD (Concentration) _
Percent relative difference Total TCOO (Concentration)
HRMS Lab.
Standard /Reagent Number or Origin Date of O.C File Result* of
Type Mfg. « QC Name QC
B-13
63 7
-------
HIGH RESOLUTION
FORM H-6 CHRONOLOGICAL LIST OF ALL ANALYSES PERFORMED
Lab:
Instrument ID:
Case*
Batch
Nam*
Analyci*
Identification
SMO
Number
O*tt
Time
B-14
/••> ••> o
boo
-------
HIGH RESOLUTION
FORM H-7 GC OPERATING CONDITIONS
Lab: Instrument ID:
GC Column:
Film Thickness:
Column Dimensions:
Initial Column Temperature:
Temperature Program:
Injector Temperature:
Interface Temperature:
Injection Mode:
Injection Volume:
SplMess Valve Closed Time:
Septum Purge Flow:
Injector Sweep Flow:
Carrier Gas Flow Rate (ml/min or cm/sec):
-QQ B-15
-------
HIGH RESOLUTION
FORM H-8 HRMS TCDD CALIBRATION REPORT FORM
Lab:
Case*:
Batch/Shipment #:
Instrument ID:
Calibration:
Initial
Routine
2.3.7.8-TCDD
m/z 258.930
319.897
321.894
'»C,,-2.3.7.8-TCDD
m/z 331.937
333.934
"C12-1.2.3.4-TCDD
m/z 331.937
m/z 333.934
Calibration Solution:
GC Column:
Date of Initial Calibration:
Analysis Date:
Time:
File Name
Retention
Time
Area
Ratios
322
332,
334
332
334
(•I If S/N i« flraatar than 6. antar ( +); if Ian than S. anar th« maaaurad ratio
B-16
G-iO
-------
HIGH RESOLUTION
FORM H-9 HIGH RESOLUTION MS TCDD DATA REPORT FORM
l.h
fmm* a
l^ttrunwt ID'
Aliquot
MM «,W«I
SMO Surnpta 0
FiU M.rrut-
M*tri»- lA/ffflr
Circle
One Percent
g L Moittiir*
So" <>•»»•*
Tim.-
GC Co'timp . ., ,
P.trx^inn O»f«
Muturod ppt 2.3.7.8-TCDD
Estimatad Total TCDD (ppt):
2.3.7.8-TCDD
m/z 258.930
319.897
321.894
Detection Umrt 2.3.7.8-TCDD:
Report Date:
Retention
Time
Area
Ratios
S/N'
"C12-2.3.7.8-TCDD
m/z 331.937
333.934
"C12-1.2,3.4-TCDD
m/z 331.937
m/z 333.934
Percent Recovery "C12-2.3.7.8-TCDD:
322
332
334
332
334
Other TCDD*
Estimated
Retention 320 S/N* 259 S/N* 320 S/N* 322 Concentration
Time 322 (ppt)
*H S/N is greater than 5. enter (•»); rf leu than 6 enter the measured ratio
B-17
G41
-------
4.6 The same raw and calculated data which are provided for the actual
samples will also be reported for the duplicate analyses, the method
blank analyses, the fortified field blank pair and rinsate analyses,
and any other QA or performance sample analyzed in conjunction with
the actual sample set(s).
4.7 The recoveries of the internal standard ( Cj2-2,3,7,8-TCDD) in percent.
4.8 The calibration data, including relative response factors calculated from
the calibration procedure described in Section 8.3, Exhibit D. Data
showing that these factors have been verified at least once during each
12-hour period of operation must be included (Section 8.5, Exhibit D).
Exact mass measurement error. Include peak matching box settings
and calculations as appropriate.
4.9 The calculated dry weight of the original soil or sediment sample portion
based on the dry weight determination of another sample portion of approxi-
mately equal wet weight. The exact volumes of the water and rinsate
samples analyzed.
4.10 Documentation of the source of all TCDD standards used and available
specifications on purity.
4.11 In addition, each report of analyses will include the following- selcted
ion current profiles: 1) those obtained from all samples analyzed, 2)
those from each GC colunn performance check, and 3) those from the
calibration solutions. The peak profile from each mass resolution
check must also be part of the data package.
4.12 Identify which HRGC/HRMS system was used for the analyses (manufacturer
and laboratory identification number of system -01, 02, 03, etc.).
4.13 GC operating conditions such as type of GC column, film thickness, column
dimensions, initial column temperature, temperature program, injector
temperature, interface temperature, injection mode and volume, valve time
(valve flush), septum purge flow, flow rate, and total injector flow
should be provided (Form H-7).
B-18
-------
EXHIBIT C
Sample Rerun Requirements
G43
-------
1. SCOPE AND APPLICATION
The Contractor shall be required to reextract and/or perform additional
cleanup and reanalyze certain samples or batches of samples in a variety
of situations that may occur in the process of contract performance.
(For purposes of this contract, the term "sample rerun" shall indicate
sample extraction of a fresh 10-g soil or sediment portion or 1-L aqueous
sample, followed by cleanup and analysis, and the term "extract reanalysis"
shall indicate analysis of another aliquot of the final extract.
In situations where the sample rerun is required due to matrix effects,
interferences or other problems encountered because of very complex samples,
the Government will pay the Contractor for the sample reruns. Such sample
reruns shall be billable and accountable under the specified contract
allotment of automatic reruns.
In situations where the sample rerun or the extract reanalysis is required
due to Contractor materials, equipment or instrumentation problems, or
lack of contractor's adherence to specified contract procedures, the
sample rerun or extract reanalysis shall not be billable under the terms
of the contract.
Contractor's failure to perform any of the sample reruns or extract re-
analyses specified herein, either billable or non-billable, shall be
construed as Contractor nonperf ormance and may result in termination of
the contract for default by the Contractor.
2. Required Sample Reruns and Extract Reanalyses
2.1 Automatic sample reruns and extract reanalyses that may be billable
as such under the contract.
2.1.1 If the percent recovery for the internal standard C~2, 3,7,8-
TCDD was outside of the acceptance limits of >bO percent and
^120 percent, the Contractor shall reextract and reanalyze the
sample. If the percent recovery for the sample rerun is still
outisde the acceptance limits, then both analyses can be billed
if the recoveries from both analyses are either 120%.
If, however, the percent recovery for the sample rerun is
within the acceptance limits, or if it is still outside the
acceptance limits but the percent recoveries from the original
analysis and the sample rerun are not both either <40% or
>120%, then the sample rerun may not be billed.
2.1.2 If the internal standard was not found upon monitoring m/z
331.937 and 333.934, the Contractor shall reextract and
reanalyze the sample. If the internal standard is not
found in the sample rerun, the' sample rerun is billable. If
the internal standard is found in the sample rerun, then the
sample rerun is not billable.
C-l
-------
2.1.3 If either one of the isotope abundance ratios for m/z 319.897/
321.894 or for 331.937/333.934 is less than 0.67 or greater
than 0.90 and all other criteria contained in Section 12.4 of
Exhibit D are met, then the extract shall be reanalyzed. If
both ion abundance ratios now meet the criterion, these values
shall be reported as the isotope abundance ratios, and the
Contractor shall not bill the Government for the extract
reanalysis. If the ratio in question is still outside the
criterion, the Contractor shall rerun the sample (Section 7.2,
Exhibit E). If either one of the ratios determined from the
sample rerun is still outside the acceptance limits, then
both runs and the extract reanalysis can be billed if the
corresponding isotope abundance ratios from both runs are
either <0.67 or >0.90. If, however, both isotope abundance
ratios from the sample rerun meet the criteria, or if both
corresponding isotope abundance ratios from the original run
and the sample rerun are not both either <0.67 or >0.90,
then the extract reanalysis and the sample rerun may not be
billed.
2.1.4 If the recoveries of 2,3,7,8-TCDD (Section 4.2.1.1.3.1,
Exhibit E) and/or 1,2,3,4-TCDD (Section 4.2.1.2, Exhibit E)
in the fortified field blank pair are <40% or >120%, the
Contractor shall reextract and reanalyze a second portion of
the field blank sample (Section 4.2, Exhibit E). If the
percent recoveries for the sample rerun are still outside the
acceptance limits, then both analyses can be billed as long
as the recoveries from both analyses are either <40% or >120%.
If, however, the percent recoveries for the sample rerun
are within the acceptance limits, or if they are still outside
the acceptance limits but the percent recoveries from the
original run and the sample rerun are not both either <40%
or >120%, then the sample rerun may not be billed.
NOTE: Fortified field blanks as described in Sections
4.2.1.1.4 and 4.2.1.2.2, Exhibit E, can never be billed.
2.2 Automatic sample extract dilution and HRGC/HRMS analysis, billable as
such under the Contract.
If any individual or group of coeluting TCDD isomer concentrations in
the 10-uL final extract exceeds 100 pg/uL, the analyst will perforn a
dilution as specified in Section 13.3, Exhibit D, and reanalyze the
diluted portion using HRGC/HRMS.
2.3 Sample reruns and/or extract reanalyses to be performed at Contractor's
expense (i.e., not billable under the terns of the contract).
2.3.1 If the method blank contains any signal in the TCDD retention
time window at or above the method quantitation limit (2 ppt
C-2
G45
-------
for soil and sediment and 20 ppq for aqueous samples), the
Contractor shall rerun all positive samples in the batch of
samples (Section A.1.2, Exhibit E).
2.3.2 If the system performance using the GC column performance
check (PC) solution does not meet specified criteria, the
Contractor shall take corrective action, demonstrate acceptable
GC column performance, and reanalyze the extracts from all
positive samples run during the time period between the last
acceptable PC run and the unacceptable PC run (Section 2.4,
Exhibit E).
2.3.3 If a false positive is reported for an uncontaminated soil
(blind QC) sample, upon notification by the Sample Management
Office the Contractor shall reextract and reanalyze all samples
reported as positive in the associated batch of samples
(Section 8.1.1, Exhibit E).
2.3.4 If the analysis results for a performance evaluation blind QC
sample fall outside of EPA-established acceptance windows, upon
notification of the Sample Management Office the Contractor
shall reextract and reanalyze the entire associated batch
of samples (Section 8.4.1, Exhibit E).
2.3.5 If the isotope abundance ratio for m/z 319.897/321.894 or for
331.937/333.934 is less than 0.67 or greater than 0.90, and
all other criteria contained in Section 12.4 of Exhibit D are
met, then the extract shall be reanalyzed'. If the ion abundance
ratio in question now meets the criterion, this value shall be
reported as the isotope abundance ratio, and the Contractor
shall not bill the Government for the extract reanalysis.
2.3.6 If the system performance mass resolution check does not meet
the specified criterion, the Contractor shall take corrective
action, demonstrate acceptable mass resolution and reanalyze
the extract from all positive samples analyzed during the time
period between the last acceptable mass resolution check and
the unacceptable mass resolution check (Section 2.4,
Exhibit E).
C-3
-------
EXHIBIT D
Analytical Method
2,3,7,8-Tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) and Total
TCDDs in Soil/Sediment and Water by High-Resolution Gas
Chrotnatography/High-Resolution Mass Spectrometry
P 4 7
^ i (
-------
EXHIBIT D
Section Subject Page
1 Scope and Application D-l
2 Summary of Method D-l
3 Definitions D-2
4 Interferences D-3
5 Safety D-3
6 Apparatus and Equipment D-4
7 Reagents and Standard Solutions. ...... D-6
8 System Performance Criteria D-9
9 Quality Control Procedures D-1A
10 Sample Preservation and Handling D-1A
11 Sample Extraction , D-15
12 Analytical Procedures D-18
13 Calculations D-19
-------
1. SCOPE AND APPLICATION
1.1 This method provides procedures for the detection and quantitative
measurement of 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD; CAS
Registry Number 1746-01-6; Storet number 3475) at concentrations of
2 pg/g (2 parts per trillion) to 100 pg/g (100 parts per trillion)
in 10-g portions of soil and sediment and at 20 pg/L (20 parts per
quadrillion) to 1000 pg/L (1 part per trillion) in 1-L samples of
water. Dilution of an aliquot of the final extract permits measure-
ment of concentrations up to 1.2 ng/g (1.2 parts per billion) or 12
ng/L (12 parts per trillion), respectively. This method also allows
the estimation of quantities of total TCDD present in the sample.
Samples containing concentrations of any individual TCDD isomer or
group of coeluting TCDD isomers greater than 1.2 ng/g or 12 ng/L must
be analyzed by a protocol designed for such concentration levels,
with an appropriate instrument calibration range.
1.2 The minimum measurable concentration is estimated to be 2 pg/g (2
parts per trillion) for soil and sediment samples and 20 pg/L (20
parts per quadrillion) for water samples, but this depends on kinds
and concentrations of interfering compounds in the sample matrix.
1.3 This method is designed for use by analysts who are experienced in
the use of high-resolution gas chromatography/high-resolution mass
spectrometry.
CAUTION: TCDDs are assumed to be extremely hazardous. It is the labora-
tory's responsibility to ensure that safe.handling procedures are
employed.
2. SUMMARY OF METHOD
One thousand pg of C,2-2,3,7,8-TCDD (internal standard) are added to a
10-g portion of a soil/sediment sample (weighed to 3 significant figures)
or a 1-L aqueous sample, and the sample is extracted with 200 to 250 mL
benzene using a Soxhlet apparatus for soils and sediments with a minimum
of 3 cycles per hour, or with methylene chloride using a continuous liquid-
liquid extractor for aqueous samples for 24 hours. A separatory funnel
and 3 x 60 mL methylene chloride may also be used for aqueous samples.
After appropriate cleanup, 10 uL of a tridecane solution of the recovery
standard ( C12~l»2,3,4-TCDD) are added to the extract which is then
concentrated to a final volume of 10 uL. One to three uL of the concen-
trated extract is injected into a gas chromatograph with a capillary
column interfaced to a high-resolution mass spectrometer capable of rapid
multiple ion monitoring at resolutions of at least 10,000 (10 percent
valley).
Identification of 2,3,7,8-TCDD is based on the detection of the ions n/z
319.897 and 321.894 at the same GC retention time and within -1 to +3
seconds GC retention time of the internal standard masses of m/z 331.937
and 333.934. Confirmation of 2,3,7,8-TCDD (and of other TCDD isomers) is
D-l
'
'1
•->
-------
based on the ion m/z 258.930 which results from loss of COCL by the parent
molecular ion.
3. DEFINITIONS
i
3.1 Concentration calibration solutions — solutions containing known
amounts of the analyte (unlabeled 2,3,7,8-TCDD), the internal standard
I3C12-2,3,7,8-TCDD and the recovery standard 13C,2~1.2,3,4-TCDD;
they are used to determine instrument response of the analyte
relative to the internal standard and of the internal standard
relative to the recovery standard.
3.2 Field blank — a portion of soil/sediment or water uncontaminated with
2,3,7,8-TCDD and/or other TCDDs.
3.3 Rinsate — a portion of solvent used to rinse sampling equipment; the
rinsate is analyzed to demonstrate that samples have not been contami-
nated during sampling.
3.4 Internal standard — 13C12-2,3,7,8-TCDD, which is added to every
sample (except the blank described in Sections 4.2.1 of Exhibit E)
and is present at the same concentration in every method blank and
quality control sample. It is added to the soil/sediment or aqueous
sample before extraction and is used to measure the concentration of
each analyte. Its concentration is measured in every sample, and
percent recovery is determined using an internal standard method.
3.5 Recovery standard — ^ Cl2~l»2»3,4-TCDD which is added to every sample
extract (except for the blank discussed in Sections 4.2.1, Exhibit E)
just before the final concentration step and HRGC-HRMS analysis.
3.6 Laboratory method blank — this blank is prepared in the laboratory
through performing all analytical procedures except addition of a
sample aliquot to the extraction vessel.
3.7 GC cdlumn performance check mixture — a mixture containing known
amounts of selected standards; it is used to demonstrate continued
acceptable performance of the capillary column, i.e., separation
(_< 25% valley) of 2,3,7,8-TCDD isomer from all other 21 TCDD isomers,
and to define the TCDD retention time window.
3.8 Performance evaluation sample — a soil, sediment or aqueous sample
containing a known amount of unlabeled 2,3,7,8-TCDD and/or other
TCDDs. It is distributed by the EMSL-LV to potential contractor lab-
oratories who must analyze it and obtain acceptable results before
being awarded a contract for sample analyses (see IFB Pre-Award Bid
Confirmations). It may also be included as an unspecified ("blind")
QC sample in any sample batch submitted to a laboratory for analysis.
3.9
Relative response factor — response of the mass spectrometer to a
known amount of an analyte relative to a known amount of an internal
standard.
D-2
-------
3.10 Mass resolution check — standard method used to demonstrate static
resolution of 10,000 minimum (10% valley definition).
3.11 Positive response for a blank — defined as a signal in the TCDD
retention time window, at any of the masses monitored, which is
equivalent to or above the method quantitation limit (2 ppt for soil
and sediment, and 20 ppq for aqueous samples).
3.12 Sample rerun — extraction of another 10-g soil or sediment sample
portion or 1-L aqueous sample, followed by extract cleanup and
extract analysis.
3.13 Extract reanalysis — analysis of another aliquot of th final extract.
4. INTERFERENCES
Chemicals which elute from the GC column within _+10 scans of the internal
and/or recovery standard (m/z 331.937 and 333.934) and which produce within
the TCDD retention time window ions at any of the masses used to detect or
quantify TCDD are potential interferences. Most frequently encountered
potential interferences are other sample components that are extracted
along with TCDD, e.g. PCBs, chlorinated tnethoxybiphenyls, chlorinated
hydroxydiphenylethers, chlorinated benzylphenylethers, chlorinated naphtha-
lenes, DDE, DDT, etc. The actual incidence of interference by these
chemicals depends also upon relative concentrations, mass spectrometric
resolution, and chromatographic conditions. Because very low levels of
TCDDs must be measured, the elimination of interferences is essential.
High-purity reagents and solvents must be used and all equipment must be
scrupulously cleaned. Blanks (Exhibit E, Quality Control, Section 4) must
be analyzed to demonstrate absence of contamination that would interfere
with TCDD measurement. Column chromatographic procedures are used to
remove some coextracted sample components; these procedures must be
performed carefully to minimize loss of TCDDs during attempts to increase
their concentration relative to other sample components.
5. SAFETY
The toxicity or carcinogen!city of each reagent used in this method has
not been precisely defined; however, each chemical compound should be
treated as a potential health hazard. From this viewpoint, exposure to
these chemicals must be reduced to the lowest possible level by whatever
means available. The laboratory is responsible for maintaining a file of
current OSHA regulations regarding the safe handling of the chemicals
specified in this method. A reference file of material data handling
sheets should also be made available to all personnel involved in the
chemical analysis. Additional references to laboratory safety are identi-
fied O-3) (page D-21). 2,3,7,8-TCDD has been identified as a suspected
hunan or mammalian carcinogen. The laboratory is responsible for ensuring
that safe handling procedures are followed.
D-3
P r~ 1
b u 1
-------
6. APPARATUS AND EQUIPMENT
6.1 High-Resolution Gas Chromatograph/High-Resolution Mass
Spectrometer/Data System (HRGC/HRMS/DS)
6.1.1 The GC must be equipped for temperature programming, and all
required accessories must be available, such as syringes, gases,
and a capillary column. The GC injection port must be designed
for capillary columns. The use of splitless injection tech-
niques is recommended. On-column injection techiques can be
used but this may severely reduce column lifetime for
nonchemically bonded columns. When using the method in this
protocol, a 2-uL injection volume is used consistently. With
some GC injection ports, however, 1-uL injections may produce
improved precision and chromatographic separation. A 1- to 3-uL
injection volume may be used if adequate sensitivity and
precision can be achieved.
NOTE: If 1 uL or 3 uL is used at all as injection volume, the injec-
tion volumes for all extracts, blanks, calibration solutions
and the performance check sample must be 1 uL or 3 uL.
6.1.2 Gas Chromatograph-Mass Spectrometer Interface
The GC-MS interface may include enrichment devices, such as a
glass jet separator or a silicone membrane separator, or the
gas chromatograph can be directly coupled to the mass spectrome-
ter ion source. The interface may include a diverter valve
for shunting the column effluent and isolating the mass spec-
trometer ion source. All components of the interface should
be glass or glass-lined stainless steel. The interface com-
ponents should be compatible with 300°C temperatures. The
GC/MS interface must be appropriately designed so that the
separation of 2,3,7,8-TCDD from the other TCDD isomers which
is achieved in the gas chromatographic column is not appreci-
ably degraded. Cold spots and/or active surfaces (adsorption
sites) in the GC/MS interface can cause peak tailing and peak
broadening. It is recommended that the GC column be fitted
directly into the MS ion source. Graphite ferrules should be
avoided in the GC injection port since they may adsorb TCDD.
Vespel™ or equivalent ferrules are recommended.
6.1.3 Mass Spectrometer
The static resolution of the instrument must be maintained at
a minimum 10,000 (10 percent valley). The mass spectrometer
must be operated in a selected ion monitoring (SIM) mode with
total cycle time (including voltage reset time) of one second
or less (Section 8.3.4.1). At a minimum, the following ions
which occur at these masses must be monitored: m/z 258.930,
319.897, 321.894, 331.937 and 333.934.
D-4
-------
6.1.4 Data System
A dedicated hardware or data system is employed to control the
rapid multiple ion monitoring process and to acquire the data.
Quantification data (peak areas or peak heights) and SIM traces
(displays of intensities of each m/z being monitored as a
function of time) must be acquired during the analyses.
Quantifications may be reported based upon computer-generated
peak areas or upon measured peak heights (chart recording).
NOTE: Detector zero setting must allow peak-to-peak measurement of the noise
on the base line.
6.2 GC Columns
For isomer-specific determinations of 2,3,7,8-TCDD, the following
fused silica capillary columns are recommended: a 60-m SP-2330 (SP-
2331) column and a 50-m CP-Sil 88 column. However, any capillary
column which separates 2,3,7,8-TCDD from all other TCDDs may be used
for such analyses, but this separation must be demonstrated and
documented. Minimum acceptance criteria must be determined per
Section 8.1. At the beginning of each 12-hour period (after mass
resolution has been demonstrated) during which sample extracts or
concentration calibration solutions will be analyzed, column operating
conditions must be attained for the required separation on the column
to be used for samples. Operating conditions known to produce accept-
able results with the recommended columns are shown in Table 2 at the
end of this Exhibit.
6.3 ' Miscellaneous Equipment
6.3.1 Nitrogen -evaporation apparatus with variable flow rate.
6.3.2 Balance capable of accurately weighing to +0.01 g.
6.3.3 Centrifuge capable of operating at 2,000 rpm.
6.3.4 Water bath — equipped with concentric ring cover and capable
of being temperature-controlled within ^2°C.
6.3.5 Stainless steel spatulas or spoons.
6.3.6 Stainless steel (or glass) pan large enough to hold contents
of 1-pint sample containers.
6.3.7 Glove box.
6.3.8 Drying oven.
6.4 Glassware
6.4.1 Soxhlet apparatus — all-glass, Kontes 6730-02 or equivalent;
D-5
G53
-------
90 mm x 35 ram glass thimble; 500-mL flask; condenser of appro-
priate size*
6.4.2 Kuderna-Danish apparatus — 500-mL evaporating flask, 10-mL
graduated concentrator tubes with ground-glass stoppers, and
3-ball macro Snyder column (Kontes K-570001-0500, K-503000-
0121 and K-569001-0219 or equivalent).
6.4.3 Mini-vials — 1-mL borosilicate glass with conical-shaped
reservoir and screw caps lined with Teflon-faced silicone disks.
6.4.4 Funnels — glass; appropriate size to accommodate filter
paper used to filter jar extract (volume of approximately 170 mL).
6.4.5 Separator/, funnel — 2000 mL with Teflon stopcock.
6.4.6 Continuous liquid-liquid extractors equipped with Teflon or
glass connecting joints and stopcocks requiring no lubrication
(Hershberg-Wolf Extractor - Ace Glass Company, Vineland, NJ;
P/N 6841-10 or equivalent).
6.4.7 Chromatographic columns for the silica and alumina chroma'
tography — 1 cm ID x 10 cm long and 1 cm ID x 30 cm long.
6.4.8 Chromatographic column for the Carbopak cleanup — disposable
5-raL graduated glass pipets, 6 to 7 mm ID.
6.4.9 Desiccator.
6.4.10 Glass rods.
NOTE: -Reuse of glassware should be minimized to avoid the risk of
cross contamination. All glassware that is reused must be
scrupulously cleaned as soon as possible after use, applying
the following procedure.
Rinse glassware with the last solvent used in it then with
high-purity acetone and hexane. Wash with hot water containing
detergent. Rinse with copious amounts of tap water and several
portions of distilled water. Drain, dry and heat in a muffle
furnace at 400°C for 15 to 30 minutes. Volumetric glassware
must not be heated in a muffle furnace, and some thermally
stable materials (such as PCBs) may not be removed by heating
in a muffle furnace. In these two cases, rinsing with high-
purity acetone and hexane may be substituted for muffle-furnace
heating. After the glassware is dry and cool, rinse with hexane,
and store inverted or capped with solvent-rinsed aluminum foil
in a clean environment.
7. REAGENTS AND STANDARD SOLUTIONS
7.1 Column Chromatography Reagents
D-6
-------
7.1.1 Alumina, acidic — extract the alumina in a Soxhlet with
methylene chloride for 6 hours (minimum of 3 cycles per hour)
and activate it by heating in a foil-covered glass container
for 24 hours at 190°C.
7.1.2 Silica gel — high-purity grade, type 60, 70-230 mesh; extract
the silica gel in a Soxhlet with methylene chloride for 6 hours
(minimum of 3 cycles per hour) and activate it by heating in a
foil-covered glass container for 24 hours at 130°C.
7.1.3 Silica gel impregnated with 40 percent (by weight) sulfuric
acid — add two parts (by weight) concentrated sulfuric acid
to three parts (by weight) silica gel (extracted and activated),
mix with a glass rod until free of lumps, and store in a
screw-capped glass bottle.
7.1.4 Sulfuric acid, concentrated — ACS grade, specific gravity 1.84.
7.1.5 Graphitized carbon black (Carbopack C or equivalent), surface
of approximately 12 m^/g, 80/100 mesh *-- mix thoroughly 3.6
grams Carbopak C and 16.4 grams Celite 545® in a 40-mL vial.
Activate at 130°C for six hours. Store in a desiccator.
7.1.6 Celite 545®, reagent grade, or equivalent.
7.2 Membrane filters or filter paper with pore size of ^25 urn; rinse with
hexane before use.
7.3 Glass wool, silanized — extract with methylene chloride and hexane
and air-dry before use.
7.4 Desiccating Agents
7.4.1 Sodium sulfate — granular, anhydrous; before use, extract it
with methylene chloride for 6 hours (minimum of 3 cycles per
hour) and dry it for >4 hours in a shallow tray placed in an
oven at 120°C. Let it cool in a desiccator.
7.4.2 Potassium carbonate—anhydrous, granular; use as such.
7.5 Solvents — high purity, distilled in glass: methylene chloride,
toluene, benzene, cyclohexane, methanol, acetone, hexane; reagent
grade: tridecane.
7.6 Concentration calibration solutions (Table 1) — four tridecane
solutions containing Cj2~l»2,3,4-TCDD (recovery standard) and
unlabeled 2,3,7,8-TCDD at varying concentrations, and ^Cj2-2,3,7 ,8-
TCDD (internal standard, CAS RN 80494-19-5) at a constant concentration
must be used to calibrate the instrument. These concentration calibra-
tion solutions must be obtained from the Quality Assurance Division,
US EPA, Environmental Monitoring Systems Laboratory (EMSL-LV), Las
Vegas, Nevada. However, additional secondary standards may be obtained
D-7
-------
from commercial sources, and solutions may be prepared in the con-
tractor laboratory. Traceability of standards must be verified
against EPAsupplied standard solutions. Such procedures will be
documented by laboratory SOPs as required in IFB Pre-award Bid Con-
firmations, part 2.f.(4). It is the responsibility of the laboratory
to ascertain that the calibration solutions received are indeed at the
appropriate concentrations before they are injected into the instrument,
NOTE: Serious overloading of the instrument may occur if the concentration
calibration solutions intended for a low-resolution MS are injected
into the high-resolution MS.
7.6.1 The four concentration calibration solutions contain unlabeled
2,3,7,8-TCDD and labeled * C12~l»2,3,4-TCDD at nominal concen-
trations of 2.0, 10.0, 50.0, and 100 pg/uL, respectively, and
labeled 3Cj2~2»3,7,8-TCDD at a constant nominal concentration
of 10.0 pg/uL.
7.6.2 Store the concentration calibration solutions in 1-mL mini-
vials at 4°C.
7.7 Column performance check mixture — this solvent less mixture must be
obtained from the Quality Assurance Division, Environmental Monitoring
Systems Laboratory, Las Vegas, Nevada, and dissolved by the Contractor
in 1 mL tridecane. This solution will then contain the following
components [including TCDDs (A) eluting closely to 2,3,7,8-TCDD, and
the first- (F) and last-eluting (L) TCDDs when using the columns
recommended in Section 6.2] at a concentration of 10 pg/uL of each of
these isomers:
Analyte
Unlabeled 2,3,7,8-TCDD
13C12-2,3,7,8-TCDD
1,2,3,4-TCDD (A)
1,4,7,8-TCDD (A)
1,2,3,7-TCDD (A)
1,2,3,8-TCDD (A)
1,3,6,8-TCDD (F)
1,2,8,9-TCDD (L)
Approximate Amount Per Ampule
10 ng
10 ng
10 ng
10 ng
10 ng
10 ng
10 ng
10 ng
7.8 Sample fortification solution — an isooctane solution containing
the internal standard at a nominal concentration of 10 pg/uL.
D-8
G5G
-------
7.9 Recovery standard spiking solution — a tridecane solution contain-
ing the recovery standard at a nominal concentration of 10 pg/uL.
Ten uL of this solution will be spiked into each sample extract
(except for the fortified field blank A) before the final concentration
step and HRGC/HRMS analysis. It is also used for the dilution of the
extracts from samples with high TCDD levels (Section 13.3, Exhibit D).
7.10 Internal standard spiking solution — a tridecane solution containing
the internal standard ( C^2~2,3,7,8-TCDD) at a nominal concentration of
10 pg/uL. Ten uL of this solution will be added to a fortified field
blank extract (Section 4.2.1.1, Exhibit E). This is the only case
where Cj22,3,7,8-TCDD is used for recovery purposes.
7.11 Field blank fortification solutions — isooctane solutions containing
the following TCDD isomers:
Solution A:
Solution B:
10.0 pg/uL of unlabeled 2,3,7,8-TCDD
10.0 pg/uL of unlabeled 1,2,3,4-TCDD.
8. SYSTEM PERFORMANCE CRITERIA
System performance criteria are presented below. The laboratory may use
any of the recommended columns described in Section 6.2. It must be
documented that all applicable system performance criteria specified in
Sections 8.1, 8.2, 8.3 and 8.5 have been met before analysis of any sample
is performed. Table 2 provides recommended conditions that can be used to
satisfy the required criteria. Table 3 provides a typical 12-hour analysis
sequence. The GC column performance and mass resolution checks must be
performed at the beginning and end of each 12-hour period of operation.
8.1 GC Column Performance
8.1.1 Inject 2 uL (Section 6.1.1) of the column performance check
solution (Section 7.7) and acquire selected ion monitoring
(SIM) data for m/z 258.930, 319.897, 321.894, 331.937 and
333.934 within a total cycle time of <1 second (Section
8.3.4.1).
8.1.2 The chromatographic peak separation between 2,3,7,8-TCDD and
the peaks representing any other TCDD isomers must be resolved
with a valley of <25 percent, where
Valley Percent
(x/y)(100)
x = measured as in Figure 1
y = the peak height of 2,3,7,8-TCDD.
It is the responsibility of the laboratory to verify the con-
ditions suitable for the appropriate resolution of 2,3,7,8-TCDD
D-9
657
-------
from all other TCDD isomers. The column performance check
solution also contains the TCDD isomers eluting first and last
under the analytical conditions specified in this protocol
thus defining the retention time window for total TCDD determi-
nation. The peaks representing 2,3,7,8-TCDD and the first and
the last eluting TCDD isomer must be labeled and identified as
such on the chromatograms (F and L, resp.). Any individual
selected ion current profile or the reconstructed total ion
current (m/z 259 + m/z 320 + m/z 322) constitutes an acceptable
form of data presentation.
8.2 Mass Spectrometer Performance
8.2.1 The mass spectrometer must be operated in the electron (impact)
ionization mode. Static resolving power of at least 10,000
(10 percent valley) must be demonstrated before any analysis
of a set of samples is performed (Section 8.2.2). Static
resolution checks must be performed at the beginning and at
the end of each 12-hour period of operation. However, it is
recommended that a visual check (i.e<>, not documented) of the
static resolution be made using the peak matching unit before
and after each analysis.
8.2.2 Chromatography time for TCDD may exceed the long-term mass
stability of the mass spectrometer and thus mass drift correc-
tion is mandatory. A reference compound [high-boiling
perfluorokerosene (PFK) is recommended] is introduced into the
mass spectrometer. An acceptable lock mass ion at any mass
between m/z 250 and m/z 334 (m/z 318.979 from PFK is recommended)
must be used to monitor and correct mass drifts.
NOTE: Excessive PFK may cause background noise problems and contami-
nation of the source resulting in an increase in "downtime"
for source cleaning.
Using a PFK molecular leak, tune the instrument to meet the
minimum required resolving power of 10,000 (10% valley) at
m/z 254.986 (or any other mass reasonably close to m/z 259).
Calibrate the voltage sweep at least across the mass range m/z
259 to m/z 334 and verify that m/z 330.979 from PFK (or any
other mass close to m/z 334) is measured within +5 ppn (i.e.,
1.7 mmu, if m/z 331 is chosen) using m/z 254.986 as a reference.
Documentation of the mass resolution must then be accomplished
by recording the peak profile of the PFK reference peak m/z
318.979 (or any other reference peak at a mass close to m/z
320/322). The format of the peak profile representation must
allow manual determination of the resolution, i.e., the hori-
zontal axis must be a calibrated mass scale (amu or ppm per
division). The result of the peak width measurement (performed
at 5 percent of the maximum which corresponds to the 10%
valley definition) must appear on the hard copy and cannot
exceed 100 ppn (or 31.9 mmu if m/z 319 is the chosen reference
ion).
D-10
p c: o
b u O
-------
8.3 Initial Calibration
Initial calibration is required before any samples are analyzed for
2 ,3 ,7,8-TCDD. Initial calibration is also required if any routine
calibration does not meet the required criteria listed in Section 8.6.
8.3.1 All concentration calibration solutions listed in Table 1 must
be utilized for the initial calibration.
8.3.2 Tune the instrument with PFK as described in Section 8.2.2.
8.3.3 Inject 2 uL of the column performance check solution (Section
7.7) and acquire SIM mass spectral data for m/z 258.930,
319.897, 321.894, 331.937 and 333.934 using a total cycle time
of ^ 1 second (Section 8.3.4.1). The laboratory must not
perform any further analysis until it has been demonstrated
and documented that the criterion listed in Section 8.1.2 has
been met.
8.3.4 Using the same GC (Section 8.1) and MS (Section 8.2) conditions
that produced acceptable results with the column performance
check solution, analyze a 2-uL aliquot of each of the 4 concen-
tration calibration solutions in triplicate with the following
MS operating parameters.
8.3.4.1 Total cycle time for data acquisition must be <_ 1
second. Total cycle time includes the sum of all the
dwell times and voltage reset times.
8.3.4.2 Acquire SIM data for the following selected
characteristic ions:
m/z
258.930
319.897
321.894
331.937
333.934
Compound
TCDD - COC1
Unlabeled TCDD
Unlabeled TCDD
13C12-2,3,7,8-TCDD,
13C,,-2,3,7,8-TCDD,
13C12-1,2,3,4-TCDD
13C,,-1,2,3,4-TCDD
8.3.4.3 The ratio of integrated ion current for m/z 319.897 to
m/z 321.894 for 2,3,7,8-TCDD must be between 0.67 and
0.90.
8.3.4.4 The ratio of integrated ion current for m/z 331.937 to
m/z 333.934 for 13C12-2,3 , 7 ,8-TCDD and 3CJ2-1 ,2 , 3 ,4-
TCDD must be between 0.67 and 0.90.
D-ll
P CT r\
bu.J
-------
8.3.A.5 Calculate the relative response factors for unlabeled
2,3,7,8-TCDD (RRF(I)] relative to 13C12~2,3,7,8-TCDD
and for labeled 13Cl2-2,3,7,8-TCDD [RRF(II)] relative
13C12-1,2,3,4-TCDD as follows:
to
if, - - -
Ax * <*IS
RRF(I)
* AIS
AIS " QRS
RRF(II) =
QIS * ARS
where
Ax » sun of the integrated ion abundances of m/z 319.897 and m/z 321.894
for unlabeled 2,3,7,8-TCDD.
AJS * sutn J?f tne integrated ion abundances of m/z 331.937 and m/z 333.934
for 13C17-2,3,7,8-TCDD.
ARS
sum of the integrated ion abundances for m/z 331.937 and n/z
333.934 for C12-l,2,3,4-TCDD.
QIS - quantity of 13C12-2,3,7,8-TCDD injected (pg).
QRS - quantity of Cj2-l,2,3,4-TCDD Injected (pg).
Qx - quantity of unlabeled 2,3,7,8-TCDD injected (pg).
RRF is a dimensionless quantity; the units used to express Qj$, QRS
must be the same.
8.3.4.6 Calculate the four means (RRFs) and their respective
relative standard deviations (%RSD) for the response
factors from each of the triplicate analyses for both
unlabeled and 13C12-2,3,7,8-TCDD (Form H-2).
8.3.4.7 Calculate the grand means RRF(I) and RRF(II) and their
respective relative standard deviations (%RSD) using
the four mean RRFs (Section 8.3.4.6) (Form H-2).
8.3.4.8 Calculate the routine calibration permissible range
for RRF(I) and RRF(II) using a ^20% window from the
grand means RRF(I) and RRF(II) (Section 8.3.4.7)
(Form H-2).
8.4 Criteria for Acceptable Calibration
The criteria listed below for acceptable calibration must be met
before analysis of any sample is performed.
D-12
GGO
-------
8.4.1 The percent relative standard deviation (RSD) for the response
factors from each of the triplicate analyses for both unlabeled
and Ci2-2,3,7,8-TCDD must be less than 20 percent.
8.4.2 The variation of the 4 mean RRFs for unlabeled and Ci?"
2,3,7,8-TCDD obtained from the triplicate analyses must be
less than 20 percent RSD.
8.4.3 SIM traces for 2,3,7,8-TCDD must present a signal-to-noise
ratio of ^2.5 for m/z 258.930, m/z 319.897 and, m/z 321.894.
8.4.4 SIM traces for *C12~2,3,7,8-TCDD must present a signal-to-
noise ratio 2.2-5 for m/z 331.937 and m/z 333.934.
8.4.5 Isotopic ratios (Sections 8.3.4.3 and 8.3.4.4) must be within
the allowed range.
NOTE: If the criteria for acceptable calibration listed in Sections
8.4.1 and 8.4.2 have been met, the RRF can be considered inde-
pendent of the analyte quantity for the calibration concentra-
tion range. The mean RRF from 4 triplicate determinations for
unlabeled 2,3,7,8-TCDD and for C12-2,3,7,8-TCDD will be used
for all calculations until routine calibration criteria (Section
8.6) are no longer met. At such time, new mean RRFs will be
calculated from a new set of four triplicate determinations.
8.5 Routine Calibrations
Routine calibrations must be performed at the beginning of a 12-hour
period after successful mass resolution and GC column performance
check runs.
8.5.1 Inject 2 uL of the concentration calibration solution which
contains 10 pg/uL of unlabeled 2,3,7.8-TCDD, 10.0 pg/uL
of 13C12-2,3,7,8-TCDD and 10 pg/uL C12~l,2,3,4-TCDD.
Using the same GC/MS/DS conditions as used in Sections 8.1,
8.2 and 8.3, determine and document acceptable calibration as
provided in Section 8.6.
8.6 Criteria for Acceptable Routine Calibration
The following criteria must be met before further analysis is per-
formed. If these criteria are not met, corrective action must be
taken and the instrument must be recalibrated.
8.6.1 The measured RRF for unlabeled 2,3,7,8-TCDD must be within 20
percent of the mean values established (Section 8.3.4.8) by
triplicate analyses of concentration calibration solutions.
8.6.2 The measured RRF for 13Ci2-2,3,7,8-TCDD must be within 20 per-
cent of the mean value established by triplicate analysis
of the concentration calibration solutions (Section 8.3.4.8).
D-13
661
-------
8.6.3 Isotopic ratios (Sections 8.3.4.3 and 8.3.4.4) must be within
the allowed range.
8.6.4 If one of the above criteria is not satisfied, a second attempt
can be made before repeating the entire initialization process
(Section 8.3).
NOTE: An initial calibration must be carried out whenever the HRCC 2
solution is replaced by a new one from a different lot.
9. QUALITY CONTROL PROCEDURES
See Exhibit E for QA/QC requirements.
10. SAMPLE PRESERVATION AND HANDLING
10.1 Chain-of-custody procedures — see Exhibit G.
10.2 Sample Preservation
10.2.1 When received, each soil or sediment sample will be contained
in a 1-pint glass jar surrounded by vermiculite in a sealed
metal paint can. Until a portion is to be removed for analysis,
store the sealed paint cans in a locked limited-access area
where the temperature is maintained between 25° and 35°C.
After a portion of a sample has been removed for analysis,
return the remainder of the sample to its original container
and store as stated above.
10.2.2 Each aqueous sample will be contained in a 1-liter glass
bottle. The bottles with the samples are stored at 4°C in a *
refrigerator located in a locked limited-access area.
10.2.3 To avoid photodecomposition, protect samples from light.
10.3 Sample Handling
CAUTION: Finely divided soils and sediments contaminated with 2,3,7,8-TCDD
are hazardous, because of the potential for inhalation or ingestion
of particles containing 2,3,7,8-TCDD. Such samples should be
handled in a confined environment (i.e., a closed hood or a
glove box).
10.3.1 Pre-extraction sample treatment
10.3.1*1 Homogenization — Although sampling personnel will
attempt to collect homogeneous samples, the contrac-
tor shall examine each sample and judge if it needs
further mixing.
NOTE: Contractor personnel have the responsibility to take a
representative sample portion; this responsibility
D-14
Gb'2
-------
entails efforts to make the sample as homogeneous as
possible. Stirring is recommended when possible.
10.3.1.2 Cent rifugat ion — When a soil or sediment sample
contains an obvious liquid phase, it must be
centrifuged to separate the liquid from the solid
phase. Place the entire sample in a suitable centri-
fuge bottle and centrifuge for 10 minutes at 2000 rpn.
Remove the bottle from the centrifuge. With a dis-
posable pipet, remove the liquid phase and discard
it. Mix the solid phase with a stainless steel
spatula and remove a portion to be weighed and analyzed.
Return the remaining solid portion to the original
sample bottle (which must be empty) or to a clean,
empty sample bottle which is properly labeled, and
store it as described in 10.2.1.
CAUTION: The removed liquid may contain TCDD and should be
disposed as a liquid waste.
10.3.1.3 Weigh between 9.5 and 10.5 g of the soil or sediment
sample (+0.5 g) to 3 significant figures. Dry it to
constant weight at 100°C. Allow the sample to cool
in a desiccator. Weigh the dried soil to 3 signifi-
cant figures. Calculate and report percent moisture
on Form H-9.
11. SAMPLE EXTRACTION
11.1 Soil/Sediment Extraction
11.1.1 Immediately before use, the Soxhlet apparatus is charged
with 200 to 250 mL benzene which is then refluxed for 2 hours.
The apparatus is allowed to cool, disassembled and the benzene
removed and retained as a blank for later analysis if required.
11.1.2 Accurately weigh to 3 significant figures a 10-g (9.50 g to
10.50 g) portion of the wet soil or sediment sample. Mix 100
uL of the sample f ortif ication solution (Section 7.8) with
1.5 mL acetone (1000 pg of C12~2 .3,7 ,8-TCDD) and deposit the
entire mixture in small portions on several sites on the
surface of the soil or sediment.
11.1.3 Add 10 g anhydrous sodium sulfate and mix thoroughly using a
stainless steel spoon spatula.
11.1.4 After breaking up any lumps, place the soil-sodium sulfate
mixture in the Soxhlet apparatus using a glass wool plug (the
use of an extraction thimble is optional). Add 200 to 250 mL
benzene to the Soxhlet apparatus and reflux for 24 hours. The
solvent must cycle completely through the system at least 3
times per hour.
D-15
-------
11.1.5 Transfer Che extract to a Kuderna-Danish apparatus and
concentrate to 2 to 3 mL. Rinse the column and flask with 5 mL
benzene and collect the rinsate in the concentrator tube.
Reduce the volume in the concentrator tube to 2 to 3 mL.
Repeat this rinsing and concentrating operation twice more.
Remove the concentrator tube from the K-D apparatus and care-
fully reduce the extract volume to approximately 1 mL with a
stream of nitrogen using a flow rate and distance such that
gentle solution surface rippling is observed.
NOTE: Glassware used for more than one sample must be carefully
cleaned between uses to prevent cross-contamination (Note on
page D-6) .
11.2 Extraction of Aqueous Samples
11.2.1 Nark the water meniscus on the side of the 1-L sample bottle
for later determination of the exact sample volume. Pour
the entire sample (approximately 1 L) into a 2-L separatory
f unne 1 .
11.2.2 Mix 100 uL of the sample fortification solution with 1.5 mL
acetone (1000 pg of C12-2,3 ,7,8-TCDD) and add the mixture
to the sample in the separatory funnel.
NOTE: A continuous liquid-liquid extractor may be .used in place of
a separatory funnel.
11.2.3 Add 60 mL methylene chloride to the sample bottle, seal and
shake 30 seconds to rinse the inner surface. Transfer the
solvent to the separatory funnel and extract the sample by
shaking the funnel for 2 minutes with periodic venting.
Allow the organic layer to separate from the water phase for
a minimum of 10 minutes. If an emulsion interface between
layers exists, the analyst must employ mechanical techniques
(to be described in the final report) to complete the phase
separation. Collect the methylene chloride (3 x 60 mL)
directly into a 500-mL Kuderna-Danish concentrator (mounted
with a 10 -mL concentrator tube) by passing the sample extracts
through a filter funnel packed with a glass wool plug and 5
g of anhydrous sodium sulfate. After the third extraction,
rinse the sodium sulfate with an additional 30 mL of methylene
chloride to ensure quantitative transfer.
11.2.4 Attach a Snyder column and concentrate the extract until •
the apparent volume of the liquid reaches 1 mL. Remove the
K-D apparatus and allow it to drain and cool for at least
10 minutes. Remove the Snyder column, add 50 mL benzene,
reattach the Snyder column and concentrate to approximately
1 mL. Rinse the flask and the lower joint with 1 to 2 mL
benzene. Concentrate the extract to 1.0 mL under a gentle
stream of nitrogen.
D-16
'"*
"
-------
11.2.5 Determine the original sample volume by refilling the sample
bottle to the mark and transferring the liquid to a 1000-mL
graduated cylinder. Record the sample volume to the nearest
5 mL.
11.3 Cleanup Procedures
11.3.1 Prepare an acidic silica column as follows: Pack a 1 cm x 10
cm chromatographic column with a glass wool plug, a layer (1
cm) of Na2804/K2003(1:1), 1.0 g silica gel (Section 7.1.2) and
4.0 g of 40-percent w/w sulfuric acid-impregnated silica gel
(Section 7.1.3). Pack a second chromatographic column (1 cm x
30 cm) with a glass wool plug, 6.0 g acidic alumina (Section
7.1.1) and top with a 1-cm layer of sodium sulfate (Section
7.4.1). Add hexane to the columns until they are free of
channels and air bubbles.
11.3.2 Quantitatively transfer the benzene extract (1 mL) from the
concentrator tube to the top of the silica gel column. Rinse
the concentrator tube with two 0.5-mL portions of hexane.
Transfer the rinses to the top of the silica gel column.
11.3.3 Elute the extract from the silica gel column with 90 mL hexane
directly into a Kuderna-Danish concentrator. Concentrate the
eluate to 0.5 mL, using nitrogen blow-down as necessary.
11.3.4 Transfer the concentrate (0.5 mL) to the top of the alumina
column. Rinse the K-D assembly with two 0.5-mL portions of
hexane and transfer the rinses to the top of the alumina
column. Elute the alumina column with 18 mL hexane until the
hexane level is just below the top of the sodium sulfate.
Discard the eluate. Columns must not be allowed to reach
dryness (i.e., a solvent "head" must be maintained.)
11.3.5 Place 30 mL of 20-percent (v/v) methylene chloride in hexane
on top of the alumina and elute the TCDDs from the column.
Collect this fraction in a 50-mL Erlenmeyer flask.
11.3.6 Prepare an 18-percent Carbopak C/Celite 545® mixture by thoroughly
mixing 3.6 grams Carbopak C (80/100 mesh) and 16.4 grams Celite
545® in a 40-mL vial. Activate at 130°C for 6 hours. Store
in a desiccator. Cut off a clean 5~mL disposable glass pipet
(6 to 7mm ID) at the 4-mL mark. Insert a plug of glass wool
(Section 7.3) and push to the 2-mL mark. Add 340 to 600 mg of
the activated Carbopak/Celite mixture (see NOTE) followed by
another glass wool plug. Using two glass rods, push both
glass wool plugs simultaneously towards the Carbopak/Celite
mixture and gently compress the Carbopak/Celite plug to a
length of 2 to 2.5 cm. Preelute the column with 2 mL toluene
followed by 1 mL of 75:20:5 methylene chloride/methanol/benzene,
1 mL of 1:1 cyclohexane in methylene chloride, and 2 mL hexane.
The flow rate should be less than 0-5 mL/min. While the
D-17
r> i» r."
"OOtJ
-------
column is still wet with hexane, add the entire eluate (30 mL)
from the alumina column (Section 11.3.5) to the top of the
column. Rinse the Erlenmeyer flask which contained the extract
twice with 1 mL hexane and add the rinsates to the top of the
column. Elute the column sequentially with two 1-mL aliquots
hexane, 1 mL of 1:1 cyclohexane in methylene chloride, and 1
mL of 75:20:5 raethylene chloride/ methanol/benzene. Turn the
column upside down and elute the TCDD fraction with 6 mL tolu-
ene into a concentrator tube. Warm the tube to approximately
60°C and reduce the toluene volume to approximately 1 mL using
a stream of nitrogen. Carefully transfer the concentrate into
a 1-mL mini-vial and, again at elevated temperature, reduce the
volume to about 100 uL using a stream of nitrogen. Rinse the
concentrator tube with 3 washings using 200 uL of 1% toluene
in CH2Cl2« Add 10 uL of the tridecane solution containing the
recovery standard and store the sample in a refrigerator until
HRGC/HRMS analysis is performed.
NOTE: The amount of activate Carbopak/Celite mixture required
to form a 2-to 2.5-cm plug in the column depends on the
density of the Celite being used.
12. ANALYTICAL PROCEDURES
12.1 Remove the sample extract or blank from storage and allow it to warm
to ambient laboratory temperature. With a stream of dry, purified
nitrogen, reduce the extract/blank volume to 10 uL.
12.2 Inject a 2-uL aliquot of the extract into the GC, operated under the
conditions previously used (Section 8.1) to produce acceptable results
with the performance check solution.
12.3 Acquire SIM data according to 12.3.1. Use the same acquisition and
MS operating conditions previously used (Section 8.3.4) to determine
the relative response factors.
12.3.1 Acquire SIM data for the following selected characteristic ions:
ci/z Compound
258.930 TCDD - COC1
319.897 Unlabeled TCDD
•
321.894 Unlabeled TCDD
331.937 13C12-2,3,7,8-TCDD, 13C,2-1,2,3,4-
TCDD
333.934 13C12-2,3,7,8-TCDD, 13C12-1,2,3,4-
TCDD
D-18
P ' * C'
toh
-------
NOTE: The acquisition period must at least encompass the TCDD reten-
tion time window previously determined (.Section 8.1.2, Exhibit
D).
12.4 Identification Criteria
12.4.1 The retention time (RT) (at maximum peak height) of the sample
component m/z 319.897 must be within -1 to +3 seconds of the
retention time of the peak for the isotopically labeled internal
standard at m/z 331.937 to attain a positive identification of
2,3,7,8-TCDD. Retention times of other tentatively identified
TCDDs must fall within the RT window established by analyzing
the column performance check solution (Section 8.1). Retention
times are required for all chromatograms.
' 12.4.2 The ion current responses for m/z 258.930, 319.897 and 321.894
must reach maximum simultaneously (_+ 1 sec), and all ion
current intensities must be >^ 2.5 times noise level for
positive identification of a TCDD or group of coeluting TCDD
isomers.
12.4.3 The integrated ion current at m/z 319.897 must be between 67
and 90 percent of the ion current response at m/z 321.894.
12.4.4 The integrated ion current at m/z 331.937 must be between 67
and 90 percent of the ion current response at m/z 333.934.
12.4.5 The integrated ion currents for m/z 331.937 and 333.934 must
reach their maxima within +_ 1 sec.
12.4.6 The recovery of the internal standard 13C12~2,3,7,8-TCDD must
be between 40 and 120 percent.
13. CALCULATIONS
13.1 Calculate the concentration of 2,3,7,8-TCDD (or any other TCDD isomer
or group of coeluting TCDD isomers) using the formula:
AX ' QlS
AIS * W • RRF(I)
where:
GX = unlabeled 2,3,7,8-TCDD (or any other unlabeled TCDD isomer or group of
coeluting TCDD isomers) concentration in pg/g.
Ax = sum of the integrated ion abundances determined for m/z 319.897
and 321.894.
= sum of the integrated ion abundances determined for m/z 331.937
and 333.934 of C12~2,3 , 7,8-TCDD (IS = internal standard).
D-19
G6V
-------
QIS - quantity (in picogranis) of 3C12-2,3,7 ,8-TCDD added to the
sample before extraction (Qjg = 1000 pg).
W « weight (in grams) of dry soil or sediment sample or volume of
aqueous sample converted to grams.
RRF(I) = calculated
mean relative response factor for unlabeled 2,3,7,8-TCDD
Cj2~2,3,7 ,8-TCDD. This represents the grand mean of
the RRF(I)'s obtained in Section 8.3.4.5.
relative to j2
13.2 Calculate the recovery of the internal standard C12~2 ,3,7,8-TCDD
measured in the sample extract, using the formula:
Internal standard
percent recovery - Y — - • 100
ARg * RRF(II)
Where:
= sun of the integrated ion abundances determined for m/z 331.937
and 333.934 of 1JC12-2,3,7,8-TCDD (IS = internal standard).
ARS = sum °^ tlie integrated ion abundances determined for m/z 331.937
and 333.934 of C12-l ,2,3,4-TCDD (RS = recovery standard).
Y = 0.1 for the "10-yL extract" injection (to be reported on Forms H-l,
H-5 and H-9).
and Y " 1.2 for the "24-uL extract" injection (Section 13.3) (to be reported
on Form H-9 used for reporting the diluted extract analysis).
RRF(II) » calculated mean relative response factor for labeled C|2-2,3,7,8-
TCDD relative to C12~l ,2,3,4-TCDD. This represents the grand
mean of the RRF(II)'s calculated in Section 8.3.4.5.
13.3 If the concentration of the most abundant TCDD isomer (or group of
coeluting TCDD isomers) exceeds 100 pg/uL in the 10 uL final extract,
the linear range of response vs. concentration may have been exceeded,
and a diluted aliquot of the original sample extract must be analyzed.
Accurately dilute 2 uL of the remaining original extract with 22 uL
of the tridecane solution containing 10 pg/uL of the recovery standard
(Section 7.9, Exhibit D).
13.4 Total TCDD concentration — all positively identified isomers of TCDD
must be within the RT window and meet all identification criteria
listed in Sections 12.4.2 and 12.4.3. Use the expression in Section
13.1 to calculate the concentrations of the other TCDD isomers, with
Cx becoming the concentration of any unlabeled TCDD isomer or group
of coeluting TCDD isomers.
c Total TCDD - Sum of the concentrations of the individual TCDDs including
2,3,7,8-TCDD.
D-20
-------
13.5 Estimated Detection Limit — For samples in which no unlabeled
2,3,7,8-TCDD was detected, calculate the estimated minimum detectable
concentration. The background area is determined by integrating the
ion abundances for n/z 319.897 and 321.894 in the appropriate region
of the selected ion current profiles, multiplying that area by 2.5,
and relating the product area to an estimated concentration that
would produce that product area.
Use the formula:
(2.5) ' (Ax) ' (QIS)
CE
(AIS) ' (RRF(I)) ' (W)
where
CE = estimated concentration of unlabeled 2,3,7,8-TCDD required to
produce Ax.
Ax = sum of integrated ion abundances for m/z 319.897 and 321.894 in the
same group of >5 scans used to measure Ajg.
AJS = sum °^ integrated ion abundances for the appropriate ion character-
istic of the internal standard, m/z 331.937 and m/z 333.934.
QiS» RRF(I), and W retain the definitions previously stated in Section- 13.1.
Alternatively, if peak height measurements are used for quantification, measure
the estimated detection limit by the peak height of the noise in the 2,3,7,8-
TCDD RT window.
13.6 The relative percent difference (RPD) is calculated as follows:
I S! - S2 | | S1 - S2 |
RPD = = ' x 100
Mean Concentration (Sj + S2)/2
Sj and $2 represent sample and duplicate sample results.
References
1. "Carcinogens - Working with Carcinogens", Department of Health, Education
and Welfare, Public Health Service, Center for Disease Control, National
Institute for Occupational Safety and Health, Publication No. 77-206, Aug.
1977.
2. "OSHA Safety and Health Standards, General Industry" (29 CFR1910),
Occupational Safety and Health Administration, OSHA 2206 (Revised January
1976).
3. "Safety in Academic Chemistry Laboratories", American Chemical Society
Publication, Committee on Chemical Safety, 3rd Edition 1979.
D-21
660
-------
TABLE 1 . COMPOSITION OF CONCENTRATION CALIBRATION SOLUTIONS
HRCC1
HRCC2
HRCC3
HRCC4
Recovery Standard
13C12-1,2,3,4-TCDD
2.0 pg/uL
10.0 pg/uL
50.0 pg/uL
100.0 pg/uL
Sample
Analyte
2,3,7,8-TCDD
2.0 pg/uL
10.0 pg/uL
50.0 pg/uL
100.0 pg/uL
Fortification Solution
Internal Standard
13C12-2,3,7,8-TCDD
10.0 pg/uL
10.0 pg/uL
10.0 pg/uL
10.0 pg/uL
10.0 pg/uL of 13C12-2,3,7,8-TCDD
Recovery Standard Spiking Solution
10.0 pg/uL 13C12-1,2,3,4-TCDD
Field Blank Fortification Solutions
A) 10.0 pg/uL of unlabeled 2,3,7,8-TCDD
B) 10.0 pg/uL of unlabeled 1,2,3,4-TCDD
Internal Standard Spiking Solution
10 pg/uL of 13C12-2,3,7,8-TCDD
(Used only in Section 4.2.1.1, Exhibit E)
D-22
-------
TABLE 2. RECOMMENDED GC OPERATING CONDITIONS
Column coating
Film thickness
Column dimensions
Helium linear velocity
Initial temperature
Initial time
Temperature program
Approximate 2,3,7,8-TCDD
retention time
SP-2330 (SP-2331)
0.2 urn
60 m x 0.24 mm
28-29 cm/sec
at 240°C
150°C
4 min
Rapid increase to .200°C
(15°C/min)
200°C to 2508C
at 4°C/min
27 min
CP-SIL 88
0.22 urn
50 m x 0.22 mm
28-29 cm/sec
at 240°C
2008C
1 min
Program from 200°C
to 240°C
at 4°C/min
22 min
TABLE 3. TYPICAL 12-HOUR SEQUENCE FOR 2,3,7,8-TCDD ANALYSIS
1. Static mass resolution check and mass
measurement error determination 10/20/84
2 . Co lumn
3 . HRCC2
4. Sample
5 . Co lumn
6. Static
performance check
1 through Sample "N"
performance check
mass resolution check
10/20/84
10/20/84
10/20/84
10/20/84
10/20/84
0700h
0730h
0800h
0830h
1800h
1830h
D-23
f! 7 I
w I JL-
-------
100i
o
K>
*>
80-
M
§ 60
«•*
c,
9
<0 40
"3
DC
20-
00
-------
EXHIBIT E
QA/QC Requirements
B73
-------
SUMMARY OF QC ANALYSES
o Initial and periodic calibration and instrument performance checks.
o Field blank analyses (Section A.I); a minimum of one fortified field blank
pair shall be analyzed with each sample batch; an additional fortified field
blank pair must be analyzed when a new lot of absorbent and/or solvent is used.
o Analysis of a batch of samples with accompanying QC analyses:
Sample Batch —_<24 samples, including field blank and rinsate sample(s).
Additional QC analyses per batch:
Fortified field blanks 2
Method blank (1*)
Duplicate sample 1
TOTAL 3(4)
* A method blank is required whenever a fortified field blank shows a
positive response as defined in Section 3.11, Exhibit D.
o "Blind" QC samples may be submitted to the contractor as ordinary soil,
sediment or water samples included among the batch of samples. Blind samples
include:
Uncontaminated soil, sediment and water,
Split samples,
Unidentified duplicates, and
Performance evaluation samples.
QUALITY CONTROL
1. Performance Evaluation Samples — Included among the samples in all batches
will be samples containing known amounts of unlabeled 2,3,7,8-TCDD and/or
other TCDDs that may or may not be marked as other-than-ordinary samples.
2. Performance Check Solutions
2.1 At the beginning of each 12-hour period during which samples are to
be analyzed, an aliquot each of the 1) GC column performance check
solution and 2) high-resolution concentration calibration solution
E-l
r* >"i
6 <
-------
No. 2 (HRCC2) shall be analyzed Co demonstrate adequate GC resolution
and sensitivity, response factor reproducibility, and mass range
calibration. A mass resolution check shall also be performed to
demonstrate adequate mass resolution using an appropriate
reference compound (PFK is recommended).
These procedures are described in Section 8 of Exhibit D. If the
required criteria are not met, remedial action must be taken before
any samples are analyzed.
2.2 To validate positive sample data, the GC column performance check
and the mass resolution check must be performed also at the end of
each 12-hour period during which samples are analyzed.
2.2.1 If the contractor laboratory operates only during one period
(shift) each day of 12 hours or less, the GC performance check
solution must be analyzed twice (at the beginning and end of
the period) to validate data acquired during the interim
period. This applies also to the mass resolution check.
2.2.2 If the contractor laboratory operates during consecutive
12-hour periods (shifts), analysis of the GC performance check
solution at the beginning of each 12-hour period and at the
end of the final 12-hour period is sufficient. This applies
also to the mass resolution check.
2.3 Results of at least two analyses of the GC column performance check
solution and the mass resolution check must be reported with the
sample data collected during a 12-hour period.
2.4 Deviations from criteria specified for the GC performance check or
for the mass resolution check (Section 8, Exhibit D) invalidate all
positive sanple data collected between analyses of the performance
check solution, and the extract from those positive samples shall be
reanalyzed Exhibit C).
The GC column performance check mixture, concentration calibration solu-
tions, and the sample fortification solutions are to be obtained from the
EMSL-LV. However, if not available from the EMSL-LV, standards can be
obtained from other sources, and solutions can be prepared in the contractor
laboratory. Concentrations of all solutions containing unlabeled 2,3,7,8-
TCDD which are not obtained from the EMSL-LV must be verified by comparison
with the unlabeled 2,3,7,8-TCDD standard solution (concentration of 7.87
ug/raL) that is available from the EMSL-LV. When a lower-concentration
standard solution becomes available from the EMSL-LV, it will be substituted
for the 7.87 ug/tnL standard.
E-2
7
f
-------
4. Blanks
4.1 A method blank is required whenever a positive response (Section 3.11,
Exhibit D) is obtained for a fortified field blank. To that effect,
perform all steps detailed in the analytical procedure (Section 11,
Exhibit D) using all reagents, standards, equipment, apparatus,
glassware, and solvents that would be used for a sample analysis, but
omit addition of the soil, sediment or aqueous sample portion.
13
4.1.1 The method blank must contain the same amount of Cj2~2,3,7,8-
TCDD that is added to samples before extraction.
4.1.2 An acceptable method blank exhibits no positive response (Section
3.11, Exhibit D) for any of the characteristic ions monitored.
If the method blank which was extracted along with a batch of
samples is contaminated, all positive samples must be rerun
(Exhibit C).
4.1.2.1 If the above criterion is not met, check solvents,
reagents, fortification solutions, apparatus, and
glassware to locate and eliminate the source of
contamination before any samples are extracted and
analyzed.
4.1.2.2 If new batches of reagents or solvents contain
interfering contaminants, purify or discard them.
4.2 Field blanks — Each batch of samples contains a field blank sample
of uncontaminated soil/sediment or water that is to be fortified
before analysis according to Section 4.2.1, Exhibit E. In addition
to this field blank, a batch of samples may include a rinsate, that
is a portion of solvent (usually trichloroethylene) that was used to
rinse sampling equipment. The rinsate is analyzed to assure that the
samples have not been contaminated by the sampling equipment.
4.2.1 Fortified field blank pair
4.2.1.1 Fortified field blank A: 2,3,7,8-TCDD
4.2.1.1.1 Weigh a 10-g portion or use 1 liter (for aqueous
samples) of the specified field blank sample and
add 100 uL of the solution containing 10.0 pg/uL of
2,3,7,8-TCDD (Table 1, Exhibit D) diluted in 1.5 mL
of acetone (Section 11.1.2, Exhibit D).
4.2.1.1.2 Extract using the procedures beginning in Sections
11.1 or 11.2 of Exhibit D, as applicable, add 10 uL
of the internal standard solution (Section 7.10,
Exhibit D) and analyze a 2-uL aliquot of the con-
centrated extract.
E-3
f
-------
NOTE: This is the only case where the recovery standard is
used for other than recovery purposes.
4.2.1.1.3 Calculate the concentration (Section 13.1, Exhibit
D) of 2,3,7,8-TCDD and the percent recovery of
unlabeled 2,3,7,8-TCDD. If the percent recovery at
the measured concentration of 2,3,7,8-TCDD is <40
percent or >120 percent, report the results and
repeat the fortified field blank extraction and
analysis with a second aliquot of the specified
field blank sample (Exhibit C).
4.2.1.1.4 Extract and analyze a new fortified simulated field
blank whenever new lots of solvents or reagents are
used for sample extraction or for column chromato-
graphic procedures. When a fortified simulated
field blank produces a positive response (Section
3.11, Exhibit D) for any m/z being monitored at the
retention time of 1,2,3,4-TCDD, a method blank
(Section 4.1, Exhibit E) is required.
NOTE: For this purpose only, the Contractor will simulate
field blanks by using clean sand or distilled water.
4.2.1.2 Fortified field blank B: 1,2,3,4-TCDD
4.2.1.2.1 Repeat steps 4.2.1.1.1 to 4.2.1.1.3 using unlabeled
1,2,3,4TCDD (instead of 2,3,7.8-TCDD) and 13C
12-1,2,3,4-TCDD (instead of 13C12-2,3,7,8-TCDD)
recovery standard.
as
4.2.1.2.2 Extract and analyze a new fortified simulated field
blank whenever new lots of solvents or reagents are
used for sample extraction or for column chromato-
graphic procedures. When a fortified simulated
field blank produces a positive response (Section
3.11, Exhibit D) for any m/z being monitored at the
retention time of 2,3,7,8-TCDD, a method blank
(Section 4.1, Exhibit E) is required.
4.2.2 Rinsate sample
4.2.2.1 The rinsate sample must be fortified as a regular
sample.
4.2.2.2 Take a 100-mL aliquot of sampling equipment rinse
solvent (rinsate sample), filter, if necessary, and
add 100 uL of the solution containing 10.0 pg/uL of
13C12-2,3,7,8-TCDD (Table 1, Exhibit D).
E-4
r. 7 V
o J '
-------
4.2.2.3 Using a Kuderna-Danish apparatus, concentrate to
approximately 5 mL.
4.2.2.4 Transfer the 5-mL concentrate in 1-mL portions to a 1-
mL mini-vial, reducing the volume as necessary with a
gentle stream of dry nitrogen; see Exhibit D,
Section 11.1.5 for volume reduction procedures.
4.2.2.5 Rinse the container with two 0.5-mL portions of hexane
and transfer the rinses to the 1-mL mini-vial.
4.2.2.6 Just before analysis, add 10 uL tridecane recovery
standard spiking solution (Table 1, Exhibit D), and
reduce the volume to a final volune of 10 uL (no
column chromatography is required).
4.2.2.7 Analyze an aliquot following the same procedures used
to analyze samples (Section 12, Exhibit D).
4.2.2.8 Report percent recovery of the internal standard and
the level of contamination by any TCDD isomer (or
group of coeluting TCDD isomers) on Form H-5 in pg/mL
of rinsate solvent.
5. Duplicate Analyses
5.1 Laboratory duplicates — in each batch of samples, locate the sample
specified for duplicate analysis and analyze a second 10-g soil or
sediment sample portion or 1-L water sample.
5.1.1 The results of laboratory duplicates (percent recovery and
concentrations of 2,3,7,8-TCDD and total TCDD) must agree
within 50 percent relative difference (difference expressed as
percentage of the mean). If the relative difference is >50
percent, the Contractor shall immediately contact the Sample
Management Office for resolution of the problem. Report all
results.
5.1.2 Recommended actions to help locate problems:
5.1.2.1 Verify satisfactory instrument performance
(Section 8, Exhibit D).
5.1.2.2 If possible, verify that no error was made while
weighing sample portions.
5.1.2.3 Review the analytical procedures with the performing
laboratory personnel.
6. Percent Recovery of the Internal Standard ^C,--2,3,7,8-TCDD — For each
sample, method blank and rinsate, calculate the percent recovery (Section
13.2, Exhibit D) of the measured concentration of 13C12-2,3,7,8-TCDD. If
E-5
G78
-------
the percent recovery is <40 percent or >120 percent for a sample, analyze
a second portion of that sample and report both results (Exhibit C).
NOTE: A low or high percent recovery for a blank does not require discarding
analytical data but it may indicate a potential problem with future
analytical data.
7. Identification Criteria
7.1 If either of the two identification criteria (Sections 12.4.1 and
12. A. 2, Exhibit D) is not met, it is reported that the sample does
not contain unlabeled 2,3,7,8-TCDD at the calculated detection limit
(Section 13.5, Exhibit D).
7.2 If the first two initial identification criteria are met, but the
third, fourth, fifth or sixth criterion (Sections 12.4.3 through
12.4.6, Exhibit D) is not met, that sample is presumed to contain
interfering contaminants. This must be noted on the analytical
report form and the sample must be rerun or the extract reanalyzed.
Detailed sample rerun and extract reanalysis requirements are
presented in Exhibit C.
8. Blind QC Samples — Included among soil, sediment and aqueous samples may
be QC samples that are not specified as such to the performing laboratory.
Types that may be included are:
8.1 Uncontaminated soil, sediment or water.
8.1.1 If a false positive is reported for such a sample,
the Contractor shall be required to rerun the entire
associated batch of samples (Section 2.3.3, Exhibit C).
8.2 Split samples — composited sample portions sent to more than
one laboratory.
8.3 Unlabeled field duplicates — two portions of a composited
sample.
8.4 Performance evaluation sample — soil/sediment or water sample
containing a known amount of unlabeled 2,3,7,8-TCDD and/or
other TCDDs.
8.4.1 If the performance evaluation sample result falls
outside the acceptance windows established by EPA, the
Contractor shall be required to rerun the entire associ-
ated batch of samples (Exhibit C).
NOTE: EPA acceptance windows are based on previously generated
data.
9. Records - At each contractor laboratory, records must be maintained on
E-6
r: 7
*_/ (
-------
site for six months after contract completion to document the quality of
all data generated during the contract performance. Before any records are
disposed, written concurrence from the Contracting Officer must be obtained.
10. Unused portions of samples and sample extracts must be preserved for
six months after sample receipt; appropriate samples may be selected
by EPA personnel for further analyses.
11. Reuse of glassware is to be minimized to avoid the risk of contamination.
LABORATORY EVALUATION PROCEDURES
1. On a quarterly basis, the EPA Project Officer and/or designated
representatives may conduct an evaluation of the laboratory to ascertain
that the laboratory is meeting contract requirements. This section outlines
the procedures which may be used by the Project Officer or his authorized
representative in order to conduct a successful evaluation of laboratories
conducting dioxin analyses according to this protocol. The evaluation
process consists of the following steps: 1) analysis of a performance
evaluation (PE) sample, and 2) on-site evaluation of the laboratory to
verify continuity of personnel, instrumentation, and quality assurance/
quality control functions. The following is a description of these
two steps.
2. Performance Evaluation Sample Analysis
2.1 The PE sample set will be sent to a participating laboratory to
verify the laboratory's continuing ability to produce acceptable
analytical results. The PE sample will be representative of the
types of samples that will be subject to analysis under this contract.
2.2 When the PE sample results are received, they are scored using the
PE Sample Score Sheet shown in Figure 1. If a false positive
(e.g., a PE sample not containing 2,3,7,8-TCDD and/or other TCDDs
but reported by the laboratory to contain it and/or them) is reported,
the laboratory has failed the PE analysis requirement. The Project
Officer will notify the laboratory immediately if such an event
occurs.
2.3 As a general rule, a laboratory should achieve 75 percent or more of
the total possible points for all three categories, and 75 percent or
more of the maximum possible points in each category to be considered
acceptable for this program. However, the Government reserves the
right to accept scores of less than 75 percent.
2.4 If unanticipated difficulties with the PE samples are encountered,
the total points may be adjusted by the Government evaluator in an
impartial and equitable manner for all participating laboratories.
E-7
G80
-------
Number of Maximum Possible Recommended Passing
PE Samples Score Score (75Z)
1 290 218
2 475 356
3 660 495
4 845 634
5 1030 773
On-Site Laboratory Evaluation
3.1 An on-site laboratory evaluation is performed •to verify that (1) the
laboratory is maintaining the necessary minimum level in instrumen-
tation and levels of experience in personnel committed to the con-
tract and (2) that the necessary quality control/quality assurance
activities are being carried out. It also serves as a mechanism for
discussing laboratory weaknesses identified through routine data
audits, PE sample analyses results, and prior on-site evaluation.
Photographs may be taken during the on-site laboratory evaluation
tour.
3*2 The sequence of events for the on-site evaluations is shown in
Figure 2. The Site Evaluation Sheet (SES) (Figure 3) is used to
document the results of the evaluation.
E-8
GB1
-------
Laboratory
PERFORMANCE EVALUATION SAMPLE SCORE SHEET
Date
False Positive
I. False Positive - If a laboratory reports a false
positive on any PE sample, the laboratory may be
disqualified, i.e., rendered ineligible for
contract award based on the failure to pass the
PE sample analysis requirement.
2,3,7,8-TCDD
Other TCDD(s)
( ) Yes ( ) No
( ) Yes ( ) No
Possible
Score
II. Calibration Data
1. Method Blank:
a. Results properly recorded on Forms H-l, H-5 and 5
H-9.
b. No native TCDD isomers at/or above method
quantitative limit. 5
c. Results documented by selected ion
monitoring (SIM) traces for m/z being
monitored to detect TCDDs. 5
d. Percent recovery of 13C,2-2,3,7,8-TCDD
240 and £120%. 5
2. Initial Concentration Calibration:
a. Results properly recorded on Forms H-2 5
and H-8.
b. The percent relative standard deviation
(RSD) for the response factors for each
of the triplicate analyses for both unlabeled
and 13C12-2,3,7,8-TCDD less than 20%. 5
c. The variation of the 4 mean RRFs for both
unlabeled and labeled 2,3,7,8-TCDD obtained
from the triplicate analyses less than 20% RSD. 5
d. For unlabeled 2,3,7,8-TCDD the abundance ratio
must be 20-67 and £0.90 for m/z 319.897 to
321.894. 5
Figure 1. Performance evaluation sample score sheet.
E-9^
O
-------
Possible Score
Score Achieved
e. The abundance ratios must be X).67 and <0.90
for 331.937 to 333.934 for 13C12-2,3, 7,8-TCDD
and 13C12-1,2,3,4-TCDD.
f. Results must be documented with appropriate
SIM traces, labeled with the corresponding EPA
sample numbers, and calculations.
Performance Checks:
a. GC resolution and MS resolution checks performed
at the beginning and end of each 12-hour period.
b. Results of performance checks properly recorded
on Form H-4.
c. MS Resolution: PFK (or alternate) tune shows
appropriate mass resolution (Section 8.2,
Exhibit D) with mass assignment accuracy
within +5 ppm.
d. GC Resolution: chromatograms meet the criteria
specified in Section 8.1, Exhibit D.
Routine Calibration:
a. Performed each 12 hours, after MS and GC
resolution checks, using HRCC2.
b* Results of routine calibrations properly
reported on Forms H-3 and H-8.
c. For un labeled 2,3 , 7,8-TCDD: abundance
ratio must be X).67 and £0.90 for m/z
319.897 to 321.894.
Abundance ratio correct for isotopically
labeled standards (e.g., 331.937/333.934
be >0.67 and £0
3C-1,2,3,4-TC
must be >0.67 and £0.90 for C-2 ,3 , 7,8-TCDD
and 13C12-1,2,3,4-TCDD).
Response factors [RRF(I) and RRF(II)] are
within j+20Z of the mean of the respective
initial calibration response factors. 5
Signal-to-Noise (S/N) .Ratio: SIM traces
for 2, 3, 7,8-TCDD demonstrate S/N of ^2.5. 5
Results documented with appropriate SIM
traces and calculations. 5 .
Subtotal II 105
Figure 1. (Continued).
E-10
G83
Possible Score
Score Achieved
III. Performance Evaluation (PE) Sample Data
(Scores to be determined for each sample
in the PE set)
1. Forms H-l and H-9 properly filled out for sample. 5
2. Measured concentration of unlabeled
2,3,7,8-TCDD within acceptance window
established by EPA. 40
3. Estimated concentration of total TCDDs
within acceptance window established by
EPA. 20
4. Identification Criteria for 2,3,7,8-TCDD:
a. Retention time (RT) (at maximum peak
height) of the sample component tn/z
319.897 is within -1 to +3 seconds
of the m/z 331.937 13C122,3,7,8-TCDD
internal standard peak. 10
b. The ion current responses for m/z
258.930, 319.897 and 321.894 must reach
a maximum simultaneously (_+! second)
and must be J>2*5 times noise level. 10
c. The m/z 319.897/321.894 ratio is X).67
and £0.90. 10
d. The m/z 331.937/333.934 ratio is X).67
and £0.90. 5
e. The S./N ratio for m/z 331.937 and
333.934 is ^2.5. 5
5. Identification Criteria for other TCDDs:
a. Retention time must fall into window
established by GC performance check. 5
b. The ion current responses for m/z
258.930, 319.897, and 321.894 reach
a maximum simultaneously (_+l second)
and are >2.5 times noise level. 10
c. The ra/z 319.897/321.894 ratio is
X).67 and £0.90. 5
Figure 1. (Continued).
E-ll
684
-------
Possible Score
Score Achieved
6. Concentrations of unlabeled TCDDs
are calculated according to D-13.1. 10
7. Duplicate analysis values agree within
_*502. 10
8. Estimated detection limits calculated
according to D-13.5. 10
9. Percent recovery of 13C,,-2,3,7,8-TCDD
>40 and O20%. 10
10. Results documented with appropriate
SIM traces and calculations. 20
Subtotal III 185
Total 290
Figure 1. (Continued).
E-12
f~" f*. v •
b o O
-------
EVENT SEQUENCE FOR ON-SITE LABORATORY EVALUATION
I. Meeting with Laboratory Manager and Project Manager
Introduction; discuss purpose of visit; discuss problems with data
submitted by the laboratory.
II. Verification of Personnel
Review qualification of contractor personnel in place and committed to
project (Section I, SES).
III. Verification of Instrumentation
Review equipment in place and committeed to project (Section II, SES).
The Contractor must demonstrate adequate equipment redundancy, as defined
in SES, Section II.D., to ensure his capability to perform the required
analyses in the required time.
IV. Quality Control Procedures
Walk through the laboratory to review:
1. Sample receiving and logging procedures,
2. Sample and extract storage area,
3. Procedures to prevent sample contamination,
4. Security procedures for laboratory and samples,
5. Safety procedures,
6. Conformance to written SOPs,
7. Instrument records and logbooks,
8. Sample and data control systems,
9. Procedures for handling and disposing of hazardous materials,
10. Glassware cleaning procedures,
11. Status of equipment and its availability,
12. Technical and managerial review of laboratory operations and
data package preparations,
13. Procedures for data handling, analysis, reporting and case
file preparation, and
14. Chain-of-custody procedures.
V. Review of Standard Operating Procedures (SOPs)
Review SOPs with the Project Manager to assure that the laboratory under-
stands the dimensions and requirements of the program.
VI. Identification of Needed Corrective Actions
Discuss with the Project Manager the actions needed to correct weaknesses
identified during the site inspection, PE sample analysis or production of
Figure 2. Event Sequence for On-Site Laboratory Evaluation.
E-13
686
-------
reports (hard copies and, if appropriate, manual calculations) and documen-
tation. Determine how and when corrective actions will be documented,
how and when improvements will be demonstrated, and identify the contractor
employee responsible for corrective actions.
VII. Previously Identified Problems
Check the most recent SES to verify that all previously identified
problems have been corrected.
VIII. Identification of New Problems
a. Discuss any weaknesses identified in the performance evaluation
sample analyses and reports.
b. Discuss any weaknessess identified in this site inspection.
Figure 2. (Continued)
E-14
Gr; •> ;
Cj .(
-------
SITE EVALUATION SHEET
Laboratory: Date;
Location:
EVALUATORS
Name Organization
1.
2.
3.
4.
5.
6.
7.
I. Laboratory Personnel Committed to Project:
A. Project Manager (responsible for overall technical effort)
Name:
Title:
B. GC/MS Operator:
Experience:*
(one year minumum)
C. GC/MS Data Interpreter:
Experience:*
(two year minimum)
D. Person responsible for sample exraction, column chromatography
and extract concentration:
Experience:*
(one year minumum)
E. Person(s) responsible for calculations and report preparation:
Hardcopy Reports:
F. Person responsible for handling, storage and (if appropriate)
preparation of solutions of standard compounds:
*Experience is deemed to mean "more than 50 percent of the person's productive
work time."
Figure 3. Site Evaluation Sheet.
E-15
688
-------
G. Person responsible for standards preparation/storage:
H. Person responsible for record keeping:
I. Quality Assurance Officer: _____
J. Personnel checklist
( ) Yes ( ) No
1. Do personnel assigned to this project have
the appropriate level and type of experience
to successfully accomplish the objectives of
this program?
2. Is the organization adequately staffed to ( ) Yes ( ) No
meet project requirements in a tmely
manner?
3. Does the Laboratory Quality Assurance officer ( ) Yes ( ) No
report to senior management levels?
4. Was the Quality Assurance officer available ( ) Yes ( ) No
during the evaluation?
II. Laboratory Equipment
A. Gas chromatograph(s)*
Manufacturer and Model:
Installation Date:
Type of Capillary Column Injection System: ___
Capillary Column to be used (length, ID, coating, etc.):
Necessary Ancillary Equipment (gases, syringes, etc.):
B. High Resolution Mass Spectrometer^ s)*
Static Resolution Capability (10,000 min.):
Peak matching system:
Manufacturer and Model:
Installation Date:
Pertinent Modifications: _____________
Peak Matching System/Accuracy (Mfg. spec.):
C. Data System(s)*
Manufacturer and Model:
If more than one GC/MS/DC, indicate system 1,2,3, etc., by numbering
components with 1,2,3, etc.
Figure 3. (Continued).
E-16
-------
Installation Date:
Software Version Identifier:
Appropriate selected ion monitoring software/hardware ( ) Yes ( ) No
Capability to produce hard copies of computer-
generated information ( ) Yes ( ) No
D. Evidence that at least one GC/MS/DS system can be reasonably
expected to be operating acceptably at any given time:
( ) More than one adequate GC/MS/DS system is available in-house,
(i.e.,meeting requirements specified in SOW Section 6.1,
Exhibit D).
( ) Appropriate in-house replacement parts and trained service
personnel are available.
( ) A service contract is in place with guaranteed response time
(specify type of contract and limitations).
( ) Voltage control devices are used on major instruments; isolated
circuits are used.
( ) Other (specify)
III. Facilities Checklist
A. Does the laboratory appear to have adequate ( ) Yes ( ) No
workspace (120 sq. feet, 6 linear feet of
unencumbered bench space per analyst)?
B. Does the laboratory have a source of distilled/ ( ) Yes ( ) No
demineralized water?
C. Is the analytical balance located away from ( ) Yes ( ) No
draft and areas subject to rapid temperature
changes or vibration?
D. Has the balance been calibrated within one year ( ) Yes ( ) No
by a certified technician?
E. Is the balance routinely checked with class S ( ) Yes ( ) No
weights before each use and the results recorded
in a logbook?
F. Is the laboratory maintained in a clean and ( ) Yes ( ) No
organized manner?
Figure 3. (Continued)
E-17
690
-------
G. Is the facility designed for hazardous organic ( ) Yes ( ) No
chemical analysis?
1. Is ventilation provided in the sample ( ) Yes ( ) No
preparation areas?
2. Are vented hoods available and adequately ( ) Yes ( ) No
vented in the sample preparation areas?
3. Are the hoods equipped with charcoal ( ) Yes ( ) No
and HEPA filters?
4. Are instruments, including GC/MS pumps, ( ) Yes ( ) No
vented into hoods or control devices such
as charcoal traps?
H. Are adequate secured facilities provided for ( ) Yes ( ) No
storage of samples, extracts, and calibration
standards, including cold storage?
I. Are the temperatures of the cold storage units ( ) Yes ( ) No
recorded daily in logbooks?
J. Are chemical waste disposal policies/procedures ( ) Yes ( ) No
in place?
K. Is the laboratory secure? ( ) Yes ( ) No
IV. Analysis Control Checklist
A. Do the project personnel have SOPs for the required ( ) Yes ( ) No
activities?
B. Is a logbook maintained for each instrument and ( ) Yes ( ) No
is information such as calibration data and
instrument maintenance continually recorded?
C. Do the analysts record bench data in a neat ( ) Yes ( ) No
and accurate manner?
D. Standards
1. Are fresh analytical standards prepared ( ) Yes ( ) No
at a frequency consistent with good QC?
2. Are reference materials properly labeled with ( ) Yes ( ) No
concentrations, date of preparation, and the
identity of the person preparing the sample?
3. Is a standards preparation and tracking ( ) Yes ( ) No
logbook maintained?
Figure 3. (Continued).
E-18
G31
-------
4. Are working standards traceable to EPA ( ) Yes ( ) No
standards or validated against EPA
standards?
V. Documentation/Tracking Checklist
A. Is a sample custodian designated? If yes, ( ) Yes ( ) No
name of sample custodian.
Name:
B. Are the sample custodian's procedures and ( ) Yes ( ) No
responsibilities documented? If yes, where
are these documented?
Are the chain-of-custody procedures documented? ( ) Yes ( ) No
C. Are written Standard Operating Procedures (SOPs) ( ) Yes ( ) No
developed for receipt of samples? If yes, where
are the SOPs documented (laboratory manual,
written instructions, etc.)?
D. Are quality assurance procedures documented ( ) Yes ( ) No
and available to the analysts? If yes, where
are these documented?
E. Are written Standard Operating Procedures (SOPs) ( ) Yes ( ) No
developed for compiling and maintaining sample
document files? If yes, where are the SOPs
documented (laboratory manual, written
instructions, etc.)?
F. Are the magnetic tapes stored in a secure area? ( ) Yes ( ) No
G. Are samples that require preservation stored ( ) Yes ( ) No
in such a way as to maintain their integrity?
If yes, how are the samples stored?
Documentation/Notebooks Checklist
A. Is a permanently bound notebook with preprinted, ( ) Yes ( ) No
consecutively numbered pages being used?
B. Is the type of work clearly displayed on the ( ) Yes ( ) No
notebook?
C. Is the notebook maintained in a legible manner? ( ) Yes ( ) No
D. Are entries noting anomalies routinely recorded? ( ) Yes ( ) No
Figure 3. (Continued).
E-19
^ '•} 2
\_7 «J C~*
-------
E. Has Che analyst avoided obliterating entries or the ( ) Yes ( ) No
use of a pencil?
F. Are inserts (i.e. chromatograros, computer print' ( ) Yes ( ) No
outs, etc.) permanently affixed to the notebook
and signed across insert edge and page?
G. Has the supervisor of the individual maintaining the ( ) Yes ( ) No
notebook personally examined and reviewed the notebook
periodically, and signed his/her name therein, together
with the date and appropriate comments as to whether or
not the notebook is being maintained in an appropriate
manner?
H. Where applicable, is the notebook holder ( ) Yes ( ) No
referencing reports or memoranda pertinent to
the contents of an entry?
VI. Quality Control Manual Checklist
Does the laboratory maintain a Quality Assurance/ ( ) Yes ( ) No
Quality Control (QA/QC) Manual?
Does the manual address the important elements ( ) Yes ( ) No
of a QA/QC program, including the following:
A. Personnel ( ) Yes ( ') No
B. Facilities and equipment ( ) Yes ( ) No
C. Operation of instruments ( ) Yes ( ) No
D. Documentation of Procedures ( ) Yes ( ) No
E. Procurement and inventory practices ( ) Yes ( ) No
F. Preventive maintenance ( ) Yes ( ) No
G. Reliability of data ( ) Yes ( ) No
H. Data validation ( ) Yes ( ) No
I. Feedback and corrective action ( ) Yes ( ) No
J. Instrument calibration ( ) Yes ( ) No
K. Recordkeeping ( ) Yes ( ) No
L. Internal audits ( ) Yes ( ) No
Figure 3. (Continued).
E-20
C o o
I/ J < >
-------
Are QA/QC responsibilities and reporting relationships ( ) Yes ( ) No
clearing defined?
Have standard curves been adequately documented? ( ) Yes ( ) No
Are laboratory standards traceable? ( ) Yes ( ) No
Are quality control charts maintained for each ( ) Yes ( ) No
routine analysis?
Do QC records show corrective action when ( ) Yes ( ) No
analytical results fail to meet QC criteria?
Do supervisory personnel review the data and QC results? ( ) Yes ( ) No
VII. Data Handling Checklist
Are data calculations checked by a second person? ( ) Yes ( ) No
Are data calculations documented? ( ) Yes ( ) No
Do records indicate corrective action that has ( ) Yes ( ) No
been taken on projected data?
Are limits of detection determined and reported ( ) Yes ( ) No
properly?
Are all data and records retained for the ( ) Yes ( ) No
required amount of time?
Are quality control data (e.g., standard curve ( ) Yes ( ) No
duplicates) accessible for all analytical
results?
VIII. Summary
Do responses to the evaluation indicate that ( ) Yes ( ) No
project and supervisory personnel are aware
of QA/QC and its application to the project?
Do project and supervisory personnel place ( ) Yes ( ) No
positive emphasis on QA/QC?
Have responses with respect to QA/QC aspects of ( ) Yes ( ) No
the project been open and direct?
Has a cooperative attitude been displayed by all ( ) Yes ( ) No
project and supervisory personnel?
Figure 3. (Continued)
E-21
694
-------
Does the organization place the proper emphasis ( ) Yes ( ) No
on quality assurance?
Have any QA/QC deficiencies been discussed before ( ) Yes ( ) No
leaving?
Is the overall quality assurance adequate to ( ) Yes ( ) No
accomplish the objectives of the project?
Have corrective actions recommended during ( ) Yes ( ) No
previous evaluations been implemented?
Are any corrective actions required? If so, ( ) Yes ( ) No
list the necessary actions below.
E-22
631)
-------
TECHNICAL REPORT DATA
(flettt read Instructions on lite reverse before completing)
1. REPORT NO.
3. RECIPIENT S ACCESSION NO.
4. TITLE AND SUBTITLE 8. REPORT DATE
PROTOCOL FOR THE ANALYSIS OF 2,3,7,8-TETRACHLORODIBENZOt
p-DIOXIN BY HIGH-RESOLUTION GAS CHROMATOGRAPHY/HIGH-
RESOLUTION MASS SPECTROMETRY
i. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J. S. Stanley and T. M. Sack
. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
Contract Number SAS 157 6X
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Monitoring Systems Laboratory - LV, NV
Office of Research and Development
U.S. Environmental Protection Agency
Las Vegas, NV 89114
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/07
16. SUPPLEMENTARY NOTES
Project Officer - Werner F. Beckert, Environmental Monitoring Systems Laboratory
Las Vegas, NV 89114
16. ABSTRACT
An analytical protocol for the determination of 2,3,7,8-tetrachlorodibenzo-p-
dioxin (TCDD) and total TCDDs in soil, sediment and aqueous samples using high-
resolution gas chromatography/high-resolution mass spectrometry (HRGC/HRMS) was
developed using the best features of several candidate methods and input from experts
in the field. Preliminary tests led to refinements of the chromatographic cleanup
procedures and corresponding changes in the protocol. A final single-laboratory
evaluation of the refined protocol, consisting of triplicate analyses of five solid
and five aqueous samples showed that the method is useful for the determination of
2,3,7,8-TCDD and total TCDDs at concentrations from 10 to 200 pg/g (ppt) in soils and
100 to 2,000 pg/L (ppq) in aqueous samples. Based on the data generated and on the
evaluation of several options, parts of the protocol were modified at the EMSL-LV to
lower the quantitation limit for TCDD to 2 ppt in soil/sediments and to 20 ppq in
aqueous samples.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lOENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
18. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. Of PAGES
20. SECURITY CLASS (Thispagel
UNCLASSIFIED
22. PRICE
EPA f*rm 2220-1 (••». 4-77) PMKVIOUS COITION i» OBSOLETE
-------
136
Federal Register / Vol. 49, No. 209 / Friday. October 26, 1984 / Rules and Regulations
Method 613—2,3,7,8-Tetrachlorodibenzo-p-
Dioxin
1. Scope and Application
l.l This method covers the determination
of 2,3,7.8-tetrachlorodibenzo-p-dioxin (2.3,7,8-
TCDD). The following parameter may be
determined by this method:
Parameter
237 8-TCOD
STORET
No.
34675
GAS No.
1 746-01 -6
1.2 This is a gas chromatographic/mass
spectrometer (GC/MS) method applicable to
the determination of 2,3,7.8-TCDD in
municipal and industrial discharges as
provided under 40 CFR 136.1. Method 625
may be used to screen samples for 2.3,7,8-
TCDD. When the screening test is positive.
the final qualitative confirmation and
quantification mlust be made using Method
613.
1.3 The method detection limit (MDL,
defined in Section 14.1)' for 2,3.7.8-TCDD is
listed in Table 1. The MDL for a specific
wastewater may be different from that listed.
depending upon the nature of interferences in
the sample matrix.
1.4 Because of the extreme toxicity of this
compound, the analyst must prevent
exposure to himself, of to others, by materials
knows or believed to contain 2.3,7,8-TCDD.
Section 4 of this method contains guidelines
and protocols that serve as minimum safe-
handling standards in a limited-access
laboratory.
1.5 Any modification of this method.
beyond those expressly permitted, shall be
considered as a major modification subject to
application and approval of alternate test
procedures under 40 CFR 136.4 and 136.5.
1.6 This method is restricted to use by or
under the supervision of analysts
experienced in the use of a gas
chromatograph/mass spectrometer and in the
interpretation of mass spectra. Each analyst
must demonstrate the ability to generate
acceptable results with this method using the
procedure described in Section 8.2.
2. Summary of Method
2.1 A measured volume of sample,
approximately 1-L, is spiked with an internal
standard of labeled 2.3,7.8-TCDD and
extracted with methylene chloride using a
separatory funnel. The methylene chloride
extract is exchanged to hexane during
concentration to a volume of 1.0 mL or less.
The extract is then analyzed by capillary
column CC/MS to separate and measure
2.3,7,8-TCDD."
2.2 The method provides selected column
chromatographic cleanup proceudres to aid in
the elimination of interferences that may be
encountered.
3. Interferences
3.1 Method interferences may be caused
by contaminants in solvents, reagents.
glassware, and other sample processing
hardware that lead to discrete artifacts and/
or elevated backgrounds at the masses (m/z)
monitored. All of these materials must be
routinely demonstrated to be free from
interferences under the conditions of the
analysis by running laboratory reagent
blanks as described in Section 8.1.3.
3.1.1 Glassware must be scrupulously
cleaned.4 Clean all glassware as soon as
possible after use by rinsing with the last
solvent used in it. Solvent rinsing should be
followed by detergent washing with hot
water, and rinses with tap water and distilled
water. The glassware should then be drained
dry, and heated in a muffle furnace at 400 °C
for 15 to 30 min. Some thermally stable
materials, such as PCBs, may not be
eliminated by the treatment. Solvent rinses
with acetone and pesticide quality hexane
may be substituted for the muffle furnace
heating. Thorough rinsing with such solvents
usually eliminates PCB interference.
Volumetric ware should not be heated in a
muffle furnace. After drying and cooling,
glassware should be sealed and stored in a
clean environment to prevent any
accumulation of dust or other contaminants.
Store inverted or capped with aluminum foil.
3.1.2 The use of high purity reagents and
solvents helps to mininmize interference
problems. Purification of solvents by
distillation in all-glass systems may be
required.
3.2 Matrix interferences may be caused
by contaminants that are coextracted from
the sample. The extent of matrix
interferences will vary considerably from
source to source, depending upon the nature
and diversity of the industrial complex or
municipality being sampled. 2.3,7,8-TCDD is
often associated with other interfering
chlorinated compounds which are at
concentrations several magnitudes higher
than that of 2.3,7,8-TCDD. The cleanup
producers in Section 11 can be used to
overcome many of these interferences, but
unique samples may require additional
cleanup approaches '•>1to eliminate false
positives and achieve the MDL listed in Table
1.
3.3 The primary column, SP-2330 or
equivalent, resolves 2.3,7,8-TCDD from the
other 21 TCDD insomers. Positive results
using any other gas chromatographic column
must be confirmed using the primary column.
4. Safety
4.1 The toxicity or carcinogenicity of each
reagent used in this method has not been
precisely defined: however, each chemical
compound should be treated as a potential
health hazard. From this viewpoint, exposure
to these chemicals must be reduced to the
lowest possible level by whatever means
available. The laboratory is responsible for
maintaining a current awareness file of
OSHA regulations regarding the safe
handling of the chemicals specified in this
method. A reference file of material data
handling sheets should also be made
available to all personnel involved in the
chemical analysis. Additional references to
laboratory safety are available and have
been identified •'•• for the information of the
analyst. Benzene and 2,3.7.8-TCDD have been
identified as suspected human or mammalian
carcinogens.
4.2 Each laboratory must develop a strict
safety program for handling 2,3,7,8-TCDD.
The following laboratory practices are
recommended:
4.2.1 Contamination of the laboratory will
be minimized by conducting all
manipulations in a hood.
4.2.2 The effluents of sample splitters for
the gas chromatograph and roughing pumps
on the GC/MS should pass through either a
column of activated charcoal or be bubbled
through a trap containing oil or high-boiling
alcohols.
4.2.3 Liquid waste should be dissolved in
methanol or ethanol and irradiated with
ultraviolet light with a wavelength greater
than 290 nm for several days. (Use F 40 BL
lamps or equivalent). Analyze liquid wastes
and dispose of the solutions when 2.3,7.8-
TCDD can no longer be detected.
4.3 Dow Chemical U.S.A. has issued the
following precautions (revised November
1978) for safe handling of 2,3,7,8-TCDD in the
laboratory:
4.3.1 The following statements on safe
handling are as complete as possible on the
basis of available toxicological information.
The precautions for safe handling and use an:
necessarily general in nature since detailed.
specific recommendations can be made only
for the particular exposure and circumstances
of each individual use. Inquiries about
specific operations or uses may be addressed
to the Dow Chemical Company. Assistance in
evaluating the health hazards of particular
plant conditions may be obtained from
certain consulting laboratories and from
State Departments.of Health or of Labor.
many of which have an industrial health
service. 2.3,7,8-TCDD is extremely toxic to
laboratory animals. However, it has been
handled for years without injury in analytical
and biological laboratories. Techniques used
in handling radioactive and infectious
materials are applicable to 2,3,7,8,-TCDD.
4.3.1.1 Protective equipment—Throw-
away plastic gloves, apron or lab coat, safety
glasses, and a lab hood adequate for
radioactive work.
4.3.1.2 Training—Workers must be
trained in the proper method of removing
contaminated gloves and clothing without
contacting the exterior surfaces.
4.3.1.3 Personal hygiene—Thorough
washing of hands and forearms after each
manipulation and before breaks (coffee.
lunch, and shift).
4.3.1.4 Confinement—Isolated work area.
posted with signs, segregated glassware and
tools, plastic-backed absorbent paper on
benchtops.
4.3.1.5 Waste—Good technique includes
minimizing contaminated waste. Plastic bag
liners should be used in waste cans. Janitors
must be trained in the safe handling of waste.
4.3.1.6 Disposal of wastes—2,3.7,8-TCDD
decomposes above 800 "C. Low-level waste
such as absorbent paper, tissues, animal
remains, and plastic gloves may be burned in
a good incinerator. Gross quantities
(milligrams) should be packaged securely and
disposed through commercial or
governmental channels which are capable of
handling high-level radioactive wastes or
extremely toxic wastes. Liquids should be
allowed to evaporate in a good hood and in a
disposable container. Residues may then be
handled as above.
G 9 7
-------
Federal Register / Vol. 49. No. 209 / Friday, October 26, 1984 / Rules and Regulations 137
^^^M^^B^H^HM^BIHH^B^MBB^H^^BHMiMVM^MB^HMMMM^^^^H^^^^HM^^MMi^^B^^Mi^^^^i^MHlHi^Mai^^^>M^Bi^^aHM^H^HHMH'**M^M|HM^H
V. 4.3.1.7 Decontamination—For personal
decontamination, use any mild soap with
plenty of scrubbing action. For
decontamination of glassware, tools, and
surfaces. Chlorothene NU Solvent
(Trademark of the Dow Chemical Company)
is the least toxic solvent shown to be
effective. Satisfactory cleaning may be
accomplished by rinsing with Chlorothene.
then washing with any detergent and water.
Dishwater may be disposed to the sewer. It is
prudent to minimize solvent wastes because
they may require special disposal through
commercial sources which are expensive.
4.3.1.8 Laundry—Clothing known to be
contaminated should be disposed with the
precautions described under Section 4.3.1.6.
Lab coats or other clothing worn in 2.3.7.8-
TCDD work areas may be laundered.
Clothing should be collected in plastic
bags. Persons who convey the bags and
launder the clothing should be advised of the
hazard and trained in proper handling. The
clothing may be put into a washer without
contact if the launderer knows the problem.
The washer should be run through a cycle •
before being used again for other clothing.
4.3.1.9 Wipe tests—A useful method of
determining cleanliness of work surfaces and
tools is to wipe the surface with a piece of
filter paper. Extraction and analysis by gas
chromatography can achieve a limit of
sensitivity of 0.1 fig per wipe. Less than 1 jig
of 2.3,7,8-TCDD per sample indicates
acceptable cleanliness: anything higher
warrants further cleaning. More than 10 jig
on a wipe sample constitutes an acute hazard
and requires prompt cleaning before further
use of the equipment or work space. A high
(>10 jig) 2,3,7.8-TCDD level indicates that
unacceptable work practices have been
employed in the past.
4.3.1.10 Inhalation—Any procedure that
may produce airborne contamination must be
done with good ventilation. Gross losses to a
ventilation system must not be allowed.
Handling of the dilute solutions normally
used in analytical and animal work presents
no inhalation hazards except in the case of
an accident.
4.3.1.11 Accidents—Remove
contaminated clothing immediately, taking
precautions not to contaminate skin or other
articles. Wash exposed skin vigorously and
repeatedly until medical attention is
obtained.
5. Apparatus and Materials
5.1 Sampling equipment, for discrete or
composite sampling.
5.1.1 Grab sample bottle—1-L or 1-qt,
amber glass, fitted with a screw cap lined
with Teflon. Foil may be substituted for
Teflon if the sample is not corrosive. If amber
bottles are not available, protect samples
from light. The bottle and cap liner must be
washed, rinsed with acetone or methylene
chloride, and dried before use to minimize
contamination.
5.1.2 Automatic sampler (optional)—The
sampler must incorporate glass sample
containers for the collection of a minimum of
250 mL of sample. Sample containers must be
kept refrigerated at 4 °C and protected from
light during compositing. If the sampler uses a
peristaltic pump, a minimum length of
compressible silicone rubber tubing may be
used. Before use, however, the compressible
tubing should be thoroughly rinsed with
methanol, followed by repeated rinsings with
distilled water to minimize the potential for
contamination of the sample. An integrating
flow meter is required to collect flow
proportional composites.
5.1.3 Clearly label all samples as
"POISON" and ship according to U.S.
Department of Transportation regulations.
5.2 Glassware (All specifications are
suggested. Catalog numbers are included for
illustration only.):
5.2.1 Separatory funnels—2-L and 125-mL,
with Teflon stopcock.
5.2.2 Concentrator tube, Kuderna-
Danish—lOmL graduated (Kontes K-570050-
1025 or equivalent). Calibration must be
checked at the volumes employed in the test.
Ground glass stopper is used to prevent
evaporation of extracts.
5.2.3 Evaporative flask. Kuderna-
Danish—500-mL (Kontes K-570001-0500 or
equivalent). Attach to concentrator tube with
springs.
5.2.4 Snyder column. Kuderna-Danish—
Three-ball macro (Kontes K-503000-0121 or
equivalent).
5.2,5 Snyder column. Kuderna-Danish—
Two-ball micro (Kontes K-569001-0219 or
equivalent).
5.2.6 Vials—10 to 15-mL, amber glass.
with Teflon-lined screw cap.
5.2.7 Chromatographic column—300 mm
long X 10 mm ID, with Teflon stopcock and
coarse frit filter disc at bottom.
5.2.8 Chromatographic column—400 mm
long X 11 mm ID, with Teflon stopcock and
coarse frit filter disc at bottom.
5.3 Boiling chips—Approximately 10/40
mesh. Heat to 400 °C for 30 min or Soxhlet
extract with methylene chloride.
5.4 Water bath—Heated, with concentric
ring cover, capable of temperature control
(±2 *C). The bath should be used in a hood.
5.5 GC/MS system:
5.5.1 Gas chromatograph—An analytical
system complete with a temperature
programmable gas chromatograph and all
required accessories including syringes,
analytical columns, and gases. The injection
port must be designed for capillary columns.
Either split, splitless. or on-column injection
techniques may be employed, as long as the
requirements of Section 7.1.1 are achieved.
5.5.2 Column—60 m long X 0.25 mm ID
glass or fused silica, coated with SP-2330 (or
equivalent) with a Him thickness of 0.2 u.m.
Any equivalent column must resolve 2, 3, 7.
8-TCDD from the other 21 TCDD isomers.16
5.5.3 Mass spectrometer—Either a low
resolution mass spectrometer (LRMS) or a
high resolution mass spectrometer (HRMS)
may be used. The mass spectrometer must be
equipped with a 70 V (nominal) ion source
and be capable of aquiring m/z abundance
data in real time selected ion monitoring
(SIM) for groups of four or more masses.
5.5.4 GC/MS interface—Any GC to MS
interface can be used that achieves the
requirements of Section 7.1.1. GC to MS
interfaces constructed of all glass or glass-
lined materials are recommended. Glass
surfaces can be deactivated by silanizing
with dichlorodimethylsilane. To achieve
maximum sensitivity, the exit end of the
capillary column should be placed in the ion
source. A short piece of fused silica capillary
can be used as the interface to overcome
problems associated with straightening the
exit end of glass capillary columns.
5.5.5 The SIM data acquired during the
Chromatographic program is defined as the
Selected Ion Current Profile (SICP). The SICP
can be acquired under computer control or as
a real time analog output. If computer control
is used, there must be software available to
plot the SICP and report peak height or area
data for any m/z in the SICP between
specified time or scan number limits.
5.6 Balance—Analytical, capable of
accurately weighing 0.0001 g.
6. Reagents
6.1 Reagent water—Reagent water is
defined as a water in which an interferent is
not observed at the MDL of 2. 3. 7. 8-TCDD.
6.2 Sodium hydroxide solution (10 N)—
Dissolve 40 g of NaOH (ACS) in reagent
water and dilute to 100 mL. Wash the
solution with methylene chloride and hexane
before use.
6.3 Sodium thiosulfate—(ACS) Granular.
6.4 Sulfuric acid—Concentrated (ACS. sp.
gr. 1.84).
6.5 Acetone, methylene chloride, hexane,
benzene, ortho-xylene, tetradecane—
Pesticide quality or equivalent.
6.6 Sodium sulfate—(ACS) Granular,
anhydrous. Purify by heating at 400 °C for 4 h
in a shallow tray.
6.7 Alumina—Neutral. 80/200 mesh
(Fisher Scientific Co.. No. A-540 or
equivalent). Before use, activate for 24 h at
130 'C in a foil-covered glass container.
• 6.8 Silica gel—High purity grade, 100/120
mesh (Fisher Scientific Co., No. S-679 or
equivalent).
6.9 Stock standard solutions (1.00 jig/
fiL)—Stock standard solutions can be
prepared from pure standard materials or
purchased as certified solutions. Acetone
should be used as the solvent for spiking
solutions: ortho-xylene is recommended for
calibration standards for split injectors: and
tetradecane is recommended for splitless or
on-colum injectors. Analyze stock internal
standards to verify the absence of native
2,3,7.8-TCDD.
6.9.1 Prepare stock standard solutions of
2.3,7,8-TCDD (mol wt 320) and either 37Cl4
2.3,7,8-TCDD (mol wt 328) or 13Cl,j 2.3.7.8-
TCDD (mol wt 332) in an isolated area by
accurately weighing about 0.0100 g of pure
material. Dissolve the material in pesticide
quality solvent and dilute to volume in a 10-
mL volumetric flask. When compound purity
is assayed to be 96% or greater, the weight
can be used without correction to calculate
the concentration of the stock standard.
Commercially prepared stock standards can
be used at any concentration if they are
certified by the manufacturer or by an
independent source.
6.9.2 Transfer the stock standard
solutions into Teflon-sealed screw-cap
bottles. Store in an isolated refrigerator
protected from light. Stock standard solutions
should be checked frequently for signs of
degradation or evaporation, especially just
698
-------
138 Federal Register / Vol. 49. No. 209 / Friday. October 26. 1984 / Rules and Regulations
prior to preparing calibration standards or
spiking solutions from them.
6.9.3 Stock standard solutions must be
replaced after six months, or sooner if
comparison with check standards indicates a
problem.
6.10 Internal standard spiking solution (25
ng/mL)—Using stock standard solution,
prepare a spiking solution in acetone of
either'3Cl,2 or "Cl4 2,3.7.8-TCDD at a
concentration of 25 ng/mL. (See Section 10.2)
6.11 Quality control check sample
concentrate—See Section 8.2.1.
7. Calibration
7.1 Establish gas chromatograhic
operating conditions equivalent to those
given in Table 1 and SIM conditions for the
mass spectrometer as described in Section
12.2 The GC/MS system must be calibrated
using the internal standard technique.
7.1.1 Using stock standards, prepare
calibration standards that will allow
measurement of relative response factors of
at least three concentration ratios of 2,3,7,8-
TCOO to internal standard. Each calibration
standard must be prepared to contain the
internal standard at a concentration of 25 ng/
mL. If any interferences are contributed by
the internal standard at m/z 320 and 322, its
concentration may be reduced in the
calibration standards and in the internal
standard spiking solution (Section 6.10). One
of the calibration standards should contain
2,3,7,8-TCDD at a concentration near, but
above, the MDL and the other 2.3,7,8-TCDD
concentrations should correspond to the
expected range of concentrations found in
real samples or should define the working
range of the GC/MS system.
7.1.2 Using injections of 2 to 5 fiL, analyze
each calibration standard according to
Section 12 and tabulate peak height or area
response against the concentration of 2,3,7,8-
TCDD and internal standard. Calculate
response factors (RF) for 2,3,7,8-TCDD using
Equation 1.
Equation 1.
RF="
(A.) (Cu)
(A(s) (C.)
where:
A,=SIM response for 2,3,7,8-TCDD m/z
320.
Ai5 = SIM response for the internal
standard, m/z 332 for 13C,2 2,3.7.8-TCDD
m/z 328 for "C14 2,3,7,8-TCDD.
Cu= Concentration of the internal standard
C, = Concentration of 2,3,7,8-TCDD (fig/L).
If the RF value over the working range is a
constant (< 10% relative standard deviation,
RSD). the RF can be assumed to be invariant
and the average RF can be used for
calculations. Alternatively, the results can be
used to plot a calibration curve of response
ratios, A./AU, vs. RF.
7.1.3 The working calibration curve or RF
must be verified on each working day by the
measurement of one or more 2,3.7,8-TCDD
calibration standards. If the response for
2,3,7,8-TCDD varies from the predicted
response by more than ±15%, the test must
be repeated using a fresh calibration
standard. Alternatively, a new calibration
curve must be prepared.
7.2 Before using any cleanup procedure.
the analyst must process a series of
calibration standards through the procedure
to validate elution patterns and the absence
of interferences from the reagents.
8. Quality Control
8.1 Each laboratory that uses this method
is required to operate a formal quality control
program. The minimum requirements of this
program consist of an initial demonstration of
laboratory capability and an ongoing
analysis of spiked samples to evaluate and
document data quality. The laboratory must
maintain records to document the quality of
data that is generated. Ongoing data quality
checks are compared with established
performance criteria to determine if the
results of analyses meet the performance
characteristics of the method. When results
of sample spikes indicate atypical method
performance, a quality control check .
standard must be analyzed to confirm that
the measurements were performed in an in-
control mode of operation.
8.1.1 The analyst must make an initial,
one-time, demonstration of the ability to
generate acceptable accuracy and precision
with this method. This ability is established
as described in Section 8.2.
8.1.2 In recognition of advances that are
occurring in chromatography, the analyst is
permitted certain options (detailed in
Sections 10.5,11.1, and 12.1) to improve the
separations or lower the cost of
measurements. Each time such a modification
is made to the method, the analyst is required
to repeat the procedure in Section 8.2
8.1.3 Before processing any samples, the
analyst must analyze a reagent water blank
to demonstrate that interferences from the
analytical system and glassware are under
control. Each time a set of samples is
extracted or reagents are changed, a reagent
water blank must be processed as a
safeguard against laboratory contamination.
8.1.4 The laboratory must, on an ongoing
basis, spike and analyze a minimum of 10% of
all samples with native 2.3,7,8-TCDD to
monitor and evaluate laboratory data quality.
This procedure is described in Section 8.3.
8.1.5 The laboratory must, on an ongoing
basis, demonstrate through the analyses of
quality control check standards that the
operation of the measurement system is in
control. This procedure is described in
Section 8.4. The frequency of the check
standard analyses is equivalent to 10% of all
samples analyzed but may be reduced if
spike recoveries from samples (Section 8.3)
meet all specified quality control criteria.
8.1.6 The laboratory must maintain
performance records to document the quality
of data that is generated. This procedure is
described in Section 8.5.
8.2 To establish the ability to generate
acceptable accuracy and precision, the
analyst must perform the following
operations.
8.2.1 A quality control (QC) check sample
concentrate is required containing 2,3,7,8-
TCDD at a concentration of 0.100 fig/mL in
acetone. The QC check sample concentrate
must be obtained from the U.S.
Environmental Protection Agency,
Environmental Monitoring and Support
Laboratory in Cincinnati, Ohio, if available. If
not available from that source, the QC check
sample concentrate must be obtained from
another external source. If not available from
either source above, the QC check sample
concentrate must be prepared by the
laboratory using stock standards prepared
independently from those used for
calibration.
8.2.2 Using a pipet, prepare QC check
samples at a concentration of 0.100 fig/L (100
ng/L] by adding 1.00 mL of QC check sample
concentrate to each of four 1-L aliquots of
reagent water.
8.2.3 Analyze the well-mixed QC check
samples according to the method beginning in
Section 10.
8.2.4 Calculate the average recovery (X)
in ng/L. and the standard deviation of the
recovery (s) in u£/L, for 2,3.7,8-TCDD using
the four results.
8.2.5 Compare s and (X) with the
corresponding acceptance criteria for
precision and accuracy, respectively, found in
Table 2. If s and X meet the acceptance
criteria, the system performance is
acceptable and analysis of actual samples
can begin. If s exceeds the precision limit or
X falls outside the range for accuracy, the
system performance is unacceptable for
2,3,7,8-TCDD. Locate and correct the source
of the problem and repeat the test beginning
with Section 8.2.2.
8.3 The laboratory must, on an ongoing
basis, spike at least 10% of the samples from
each sample site being monitored to assess
accuracy. For laboratories analyzing one to
ten samples per month, at least one spiked
sample per month is required.
8.3.1 The concentration of the spike in the
sample should be determined as follows:
8.3.1.1 If, as in compliance monitoring, the
concentration of 2,3,7,8-TCDD in the sample
is being checked against a regulatory
concentration limit, the spike should be at
that limit or 1 to 5 times higher than the
background concentration determined in
Section 8.3.2. whichever concentration would
be larger.
8.3.1.2 If the concentration of 2,3,7,8-
TCDD in the sample is not being checked
against a limit specific to that parameter, the
spike should be at 0.100 fig/L or 1 to 5 times
higher than the background concentration
determined in Section 8.3.2. whichever
concentration would be larger.
8.3.1.3 If it is impractical to determine
background levels before spiking (e.g.,
maximum holding times will be exceeded),
the spike concentration should be (1) the
regulatory concentration limit, if any: or, if
none (2) the larger of either 5 times higher
than the expected background concentration
or 0.100 ng/L.
8.3.2 Analyze one sample aliquot to
determine the background concentration (B)
of 2,3.7,8-TCDD. If necessary, prepare a new
QC check sample concentrate (Section 8.2.1)
appropriate for the background concentration
in the sample. Spike a second sample aliquot
with 1.0 mL of the QC check sample
concentrate and analyze it to determine the
699
-------
Federal Register / Vol. 49, No. 209 / Friday. October 26. 1984 / Rules and Regulations 139
concentration after spiking (A) of 2,3.7.8-
TCDD. Calculate percent recovery (P) as
100(A-B)%T, where T is the known true value
of the spike.
8.3.3 Compare the percent recovery (P) for
2.3.7,8-TCDD with the corresponding QC
acceptance criteria found in Table 2. These
acceptance criteria were calculated to
include an allowance for error in
measurement of both the background and
spike concentrations, assuming a spike to
background ratio of 5:1. This error will be
accounted for to the extent that the analyst's
spike to background ratio approaches 5:1."If
spiking was performed at a concentration
lower than 0.100 ftg/L, the analyst must use
either the QC acceptance criteria in Table 2.
or optional QC acceptance criteria calculated
for the specific spike concentration. To
calculate optional acceptance criteria for the
recovery of 2.3,7,8-TCDD: (1) calculate
accuracy (X') using the equation in Table 3.
substituting the spike concentration (T) for C:
(2) calculate overall precision (S') using the
equation in Table 3. substituting X' for X; (3)
calculate the range for recovery at the spike
concentration as (100 X'/T)±2.44(100 S'f
T)%. "
8.3.4 If the recovery of 2,3,7,8-TCDD falls
outside the designated range for recovery, a
check standard must be analyzed as
described in Section 8.4.
8.4 If the recovery of 2.3,7,8-TCDD fails
the acceptance criteria for recovery in
Section 8.3, a QC check standard must be
prepared and analyzed.
Note.—The frequency for the required
analysis of a QC check standard will depend
upon the complexity of the sample matrix
and the performance of the laboratory.
8.4.1 Prepare the QC check standard by
adding 1.0 mL of QC check sample
concentrate (Section 8.2.1 or 8.3.2) to 1 L of
reagent water.
8.4.2 Analyze the QC check standard to
determine the concentration measured (A) of
2,3,7,8-TCDD. Calculate the percent recovery
(P,) as 100 (A/T)%, where T is the true value
of the standard concentration.
8.4.3 Compare the percent recovery (P.)
with the corresponding QC acceptance
criteria found in Table 2. If the recovery of
2.3.7,8-TCDD falls outside the designated
range, the laboratory performance is judged
to be out of control, and the problem must be
immediately identified and corrected. The
analytical result for 2,3,7,8-TCDD in the
unspiked sample is suspect and may not be
reported for regulatory compliance purposes.
8.5 As part of the QC program for the
laboratory, method accuracy for wastewater
samples must be assessed and records must
be maintained. After the analysis of five
spiked wastewater samples as in Section 8.3,
calculate the average percent recovery (P)
and the spandard deviation of the percent
recovery (sp). Express the accuracy
assessment as a percent recovery interval
from P-2sp to P + 2sD. If P=90% and sp=10%,
for example, the accuracy interval is
expressed as 70-110%. Update the accuracy
assessment on a regular basis (e.g. after each
five to ten new accuracy measurements).
8.6 It is recommended that the
laborataory adopt additional quality
assurance practices for use with this method.
The specific practices that are most
productive depend upon the needs of the
laboratory and the nature of the samples.
Field duplicates may be analyzed to assess
the precision of the environmental
measurements. Whenever possible, the
laboratory should analyze standard reference
materials and participate in relevant
performance evaluation studies.
9. Sample Collection. Preservation, and
Handling
9.1 Grab samples must be collected in
glass containers. Conventional sampling
practices " should be followed, except that
the bottle must not be prerinsed with sample
before collection. Composite samples should
be collected in refrigerated glass containers
in accordance with the requirements of the
program. Automatic sampling equipment
must be as free as possible of Tygon tubing
and other potential sources of contamination.
9.2 All samples must be iced or
refrigerated at 4 "C and protected from light
from the time of collection until extraction.
Fill the sample bottles and. if residual
chlorine is present, add 80 mg of sodium
thiosulfate per liter of sample and mix well.
EPA Methods 330.4 and 330.5 may be used for
measurement of residual chlorine.13 Field test
kits are available for this purpose.
9.3 Label all samples and containers
"POISON" and ship according to applicable
U.S. Department of Transportation
regulations.
9.4 All samples must be extracted within
7 days of collection and completely analyzed
within 40 days of extraction.2
10. Sample Extraction
Caution: When using this method to
analyze for 2,3,7,8-TCDD, all of the following
operations must be performed in a limited-
access laboratory with the analyst wearing
full protective covering for all exposed skin
surfaces. See Section 4.2.
10.1 Mark the water meniscus on the side
of the sample bottle for later determination of
sample volume. Pour the entire sample into a
2-L separatory funnel.
10.2 Add 1.00 mL of internal standard
spiking solution to the sample in the
separatory funnel. If the final extract will be
concentrated to a fixed volume below 1.00
mL (Section 12.3), only that volume of spiking
solution should be added to the sample so
that the final extract will contain 25 ng/mL of
internal standard at the time of analysis.
10.3 Add 60 mL of methylene chloride to
the sample bottle, seal, and shake 30 s to
rinse the inner surface. Transfer the solvent
to the separatory funnel and extract the
sample by shaking the funnel for 2 min with
periodic venting to release excess pressure.
Allow the organic layer to separate from the
water phase for a minimum of 10 min. If the
emulsion interface between layers is more
than one-third the volume of the solvent
layer, the analyst must employ mechanical
techniques to complete the phase separation.
The optimum technique depends upon the
sample, but may include stirring, filtration of
the emulsion through glass wool,
centrifugation, or other physical methods.
Collect the methylene chloride extract in a
250-mL Erlenmeyer flask.
10.4 Add a second 60-mL volume of
methylene chloride to the sample bottle and
repeat the extraction procedure a second
time, combining the extracts in the
Erlenmeyer flask. Perform a third extraction
in the same manner.
10.5 Assemble a Kudema-Danish (K-D)
concentrator by attaching a 10-mL
concentrator tube to a 500-mL evaporative
flask. Other concentration devices or
techniques may be used in place of the K-D
concentrator if the requirements of Section
8.2 are met.
10.6 Pour the combined extract into the
K-D concentrator. Rinse the Erlenmeyer flask
with 20 to 30 mL of methylene chloride to
complete the quantitative transfer.
10.7 Add one or two clean boiling chips to
the evaporative flask and attach a three-ball
Snyder column. Prewet the Snyder column by
adding about 1 mL of methylene chloride to
the top. Place the K-D apparatus on a hot
water bath (60 to 65 *C) so that the
concentrator tube is partially immersed in the
hot water, and the entire lower rounded
surface of the flask is bathed with hot vapor.
Adjust the vertical position of the apparatus
and the water temperature as required to
complete the concentration in 15 to 20 min. At
the proper rate of distillation the balls of the
column will actively chatter but the chambers
will not flood with condensed solvent. When
the apparent volume of liquid reaches 1 mL,
remove the K-D apparatus and allow it to
drain and cool for at least 10 min.
10.8 Momentarily remove the Snyder
column, add 50 mL of hexane and a new
boiling chip, and reattach the Snyder column.
Raise the temperature of the water bath to 85
to 90'C. Concentrate the extract as in Section
10.7, except use hexane to prewet the column.
Remove the Snyder column and rinse the
flask and its lower joint into the concentrator
tube with 1 to 2 mL of hexane. A 5-mL syringe
is recommended for this operation. Set aside
the K-D glassware for reuse in Section 10.14.
10.9 Pour the hexane extract from the
concentrator tube into a 125-mL separatory
funnel. Rinse the concentrator tube four times
with 10-mL aliquots of hexane. Combine all
rinses in the 125-mL separatory funnel.
10.10 Add 50 mL of sodium hydroxide
solution to the funnel and shake for 30 to 60 s.
Discard the aqueous phase.
10.11 Perform a second wash of the
organic layer with 50 mL of reagent water.
Discard the aqueous phase.
10.12 Wash the hexane layer with a least
two 50-mL aliquots of concentrated sulfuric
acid. Continue washing the hexane layer with
50-mL aliquots of concentrated sulfuric acid
until the acid layer remains colorless. Discard
all acid fractions.
10.13 Wash the hexane layer with two 50-
mL aliquots of reagent water. Discard the
aqueous phases.
10.14 Transfer the hexane extract into a
125-mL Erlenmeyer flask containing 1 to 2 g
of anhydrous sodium sulfate. Swirl the flask
for 30 s and decant the hexane extract into
the reassembled K-D apparatus. Complete
the quantitative transfer with two 10-mL
hexane rinses of the Erlenmeyer flask.
700
-------
140
Federal Register / Vol. 49. No. 209 / Friday, October 26, 1984 / Rules and Regulations
10.15 Replace the one or two clean boiling
chips and concentrate the extract to 6 to 10
mL as in Section 10.8.
10.16 Add a clean boiling chip to the
concentrator tube and attach a two-ball
micro-Snyder column. Prewet the column by
adding about 1 mL of hexane to the top. Place
the micro-K-D apparatus on the water bath
so that the concentrator tube is partially
immersed in the hot water. Adjust the
vertical position of the apparatus and the
water temperature as required to complete
the concentration in 5 to 10 min. At the
proper rate of distillation the balls of the
column will actively chatter but the chambers
will not flood. When the apparent volume of
liquid reaches about 0.5 mL. remove the K-D
apparatus and allow it to drain and cool for
at least 10 min. Remove the micro-Snyder
column and rinse its lower joint into the
concentrator tube with 0.2 mL of hexane.
Adjust the extract volume to 1.0 mL with
hexane. Stopper the concentrator tube and
store refrigerated and protected from light if
further processing will not be performed
immediately. If the extract will be stored
longer than two days, it should be transferred
to a Teflon-sealed screw-cap vial. If the
sample extract requires no further cleanup,
proceed with GC/MS analysis (Section 12). If
the sample requires further cleanup, proceed
to Section 11.
10.17 Determine the original sample
volume by refilling the sample bottle to the
mark and transferring the liquid to a 1000-mL
graduated cylinder. Record the sample
volume to the nearest 5 mL.
11. Cleanup and Separation
11.1 Cleanup procedures may not be
necessary for a relatively clean sample
matrix. If particular circumstances demand
the use of a cleanup procedure, the analyst
may use either procedure below or any other
appropriate procedure.us'7However, the
analyst first must demonstrate that the
requirements of Section 8.2 can be met using
the method as revised to incorporate the
cleanup procedure. Two cleanup column
options are offered to th° analyst in this
section. The alumina column should be used
first to overcome interferences. If background
problems are still encountered, the silica gel
column may be helpful.
11.2 Alumina column cleanup for 2,3,7,8-
TCDD:
11.2.1 Fill a 300 mm long x 10 mm ID
chromatographic column with activated
alumina to the 150 mm level. Tap the column
gently to settle the alumina and add 10 mm of
anhydrous sodium sulfate to the top.
11.2.2 Preelute the column with 50 mL of
hexane. Adjust the elution rate to 1 mL/min.
Discard the eluate and just prior to exposure
of the sodium sulfate layer to the air,
quantitatively transfer the 1.0-mL sample
extract onto the column using two 2-mL
portions of hexane to complete the transfer.
11.2.3 Just prior to exposure of the sodium
sulfate layer to the air, add 50 mL of 3%
methylene chloride/97% hexane (V/V) and
continue the elution of the column. Discard
the eluate.
11.2.4 Next, elute the column with 50 mL
of 20% methylene chloride/80% hexane (V/V)
into a 500-mL K-D flask equipped with a 10-
mL concentrator tube. Concentrate the
collected fraction to 1.0 mL as in Section
10.16 and analyze by GC/MS (Section 12).
11.3 Silica gel column cleanup for 2.3.7,8-
TCDD:
11.3.1 Fill a 400 mm long x 11 mm ID
chromatographic column with silica gel to the
300 mm level. Tap the column gently to settle
the silica gel and add 10 mm of anhydrous
sodium sulfate to the top.
11.3.2 Preelute the column with 50 mL of
20% benzene/80% hexane (V/V). Adjust the
elution rate to 1 mL/min. Discard the eluate
and just prior to exposure of the sodium
sulfate layer to the air, quantitatively transfer
the 1.0-mL sample extract onto the column
using two 2-mL portions of 20% benzene/80%
hexane to complete the transfer.
11.3.3 Just prior to exposure of the sodium
sulfate layer to the air, add 40 mL of 20%
benzene/80% hexane to the column. Collect
the eluate in a clean 500-mL K-D flask
equipped with a 10-mL concentrator tube.
Concentrate the collected fraction to 1.0 mL
as in Section 10.16 and analyze by GG/MS.
12. GC/MS Analysis
12.1. Table 1 summarizes the
recommended operating conditions for the
gas chromatograph. Included in this table are
retention times and MDL that can be
achieved under these conditions. Other
capillary columns or chromatographic
conditions may be used if the requirements of
Sections 5.5.2 and 8.2 are met.
12.2 Analyze standards and samples with
the mass spectrometer operating in the
selected ion monitoring (SIM) mode using a
dwell time to give at least seven points per
peak. For LRMS, use masses at m/z 320, 322,
and 257 for 2,3,7,8-TCDD and either m/z 328
for "CU 2,3,7,8-TCDD or m/z 332 for I3C,2
2,3,7,8-TCDD. For HRMS, use masses at m/z
319.8965 and 321.8936 for 2,3,7,8-TCDD and
either m/z 327.8847 for "CU 2,3,7,8-TCDD or
m/z 331.9367 for 13Cu 2,3,7,8-TCDD.
12.3 If lower detection limits are required,
the extract may be carefully evaporated to
dryness under a gentle stream of nitrogen
with the concentrator tube in a water bath at
about 40 'C. Conduct this operation
immediately before GC/MS analysis.
Redissolve the extract in the desired final
volume of ortho-xylene or tetradecane.
12.4 Calibrate the system daily as
described in Section 7.
12.5 Inject 2 to 5 /xL of the sample extract
into the gas chromatograph. The volume of
calibration standard injected must be
measured, or be the same as all sample
injection volumes.
12.6 The presence of 2,3,7,8-TCDD is
qualitatively confirmed if all of the following
criteria are achieved:
12.6.1 The gas chromatographic column
must resolve 2,3,7,8-TCDD from the other 21
TCDD isomers.
12.6.2 The masses for native 2,3.7,8-TCDD
(LRMS-m/z 320, 322. and 257 and HRMS-m/z
320 and 322) and labeled 2,3,7,8-TCDD (m/z
328 or 332) must exhibit a simultaneous
maximum at a retention time that matches
that of native 2,3,7,8-TCDD in the calibration
standa-d, with the performance specifications
of the analytical system.
12.6.3 The chlorine isotope ratio at m/z
320 and m/z 322 must agree to within±10% of
that in the calibration standard.
12.6.4 The signal of all peaks must be
greater than 2.5 times the noise level.
12.7 For quantitation, measure the
response of the m/z 320 peak for 2.3,7.8-
TCDD and the m/z 332 peak for 13C,2 2.3.7.8-
TCDD or the m/z 328 peak for "CU 2.3.7,8-
TCDD.
12.8 Co-eluting impurities are suspected if
all criteria are achieved except those in
Section 12.6.3. In this case, another SIM
analysis using masses at m/z 257, 259, 320
and either m/a 328 or m/z 322 can be
performed. The masses at m/z 257 and m/z
259 are indicative of the loss of one chlorine
and one carbonyl group from 2,3,7,8-TCDD. If
masses m/z 257 and m/z 259 give a chlorine
isotope ratio that agrees to within ±10% of
the same cluster in the calibration standards.
then the presence of TCDD can be confirmed.
Co-eluting ODD, DDE, and PCB residues can
be confirmed, but will require another
injection using the appropriate SIM masses or
full repetitive mass scans. If the response for
"CU 2.3,7,8-TCDD at m/z 328 is too large,
PCB contamination is suspected and can be
confirmed by examining the response at both
m/z 326 and m/z 328. The "CU 2,3.7,8-TCDD
internal standard gives negligible response at
m/z 326. These pesticide residues can be
removed using the alumina column cleanup
procedure.
12.9 If broad background interference
restricts the sensitivity of the GC/MS
analysis, the analyst should employ
additional cleanup procedures and reanalyze
by GC/MS.
12.10 In those circumstances where these
procedures do not yield a definitive
conclusion, the use of high resolution mass
spectrometry is suggested.5
13. Calculations
13.1 Calculate the concentration of 2,3,7,8-
TCDD in the sample using the response factor
(RF) determined in Section 7.1.2 and Equation
2.
Equation 2:
Concentration (/ig/L) =
(AS)(IS)
(Als)(RF)(V0)
where:
A, = SIM response for 2,3.7.8-TCDD at m/z
320.
Ai»=SIM response for the internal
standard at m/z 328 or 332.
I, = Amount of internal standard added to
each extract (u.g).
V0=Volume of water extracted (L).
13.2 For each sample, calculate the
percent recovery of the internal standard by
comparing the area of the m/z peak
measured in the sample to the area of the
same peak in the calibration standard. If the
recovery is below 50%. the analyst should
review all aspects of his analytical technique.
13.3 Report results in jig/L without
correction for recovery data. All QC data
701
-------
Federal Register / Vol. 49. No. 209 / Friday. October 26, 1984 / Rules and Regulations 141
obtained should be reported with the sample
results.
14. Method Performance
14.1 The method detection limit (MDL) is
defined as the minimum concentration of a
substance that can be measured and reported
with 99% confidence that the value is above
zero.' The MDL concentration listed in Table
1 was obtained using reagent water."The
MDL actually achieved in a given analysis
will vary depending on instrument sensitivity
and matrix effects.
14.2 This method was tested by 11
laboratories using reagent water, drinking
water, surface water, and three industrial
wastewaters spiked at six concentrations
over the range 0.02 to 0.20 ng/Lls Single
operator precision, overall precision, and
method accuracy were found to be directly
related to the concentration of the parameter
and essentially independent of the sample
matrix. Linear equations to describe these
relationships are presented in Table 3.
References
1. 40 CFR Part 136, Appendix B.
2. "Determination of 2.3.7.8-TCDD in
Industrial and Municipal Wastewaters,"
EPA-600/4-S2-028. U.S. Environmental
Protection Agency. Environmental Monitoring
and Support Laboratory. Cincinnati, Ohio
45268. June 1982.
3. Buser, H.R., and Rappe, C. "High
Resolution Gas Chromatography of the 22
Tetrachlorodibenzo-p-dioxin Isomers."
Analytical Chemistry, 52. 2257 (1980).
4. ASTM Annual Book of Standards, Part
31. D3694-78. "Standard Practices for
Preparation of Sample Containers and for
Preservation of Organic Constituents,"
American Society for Testing and Materials.
Philadelphia.
5. Harless, R. L.. Oswald. E. O., and
Wilkinson, M. K. "Sample Preparation and
Gas Chromatography/Mass Spectrometry
Determination of 2.3.7,8-Tetrachlorodibenzo-
p-dioxin." Analytical Chemistry. 52. 1239
(1980).
6. Lamparski, L. L.. and Nestrick. T. J.
"Determination of Tetra-, Hepta-, and
Octachlorodibenzo-p-dioxin Isomers in
Particulate Samples at Parts per Trillion
Levels." Analytical Chemistry. 52. 2045
(1980).
7. Longhorst, M. L. and Shadoff, L. A.
"Determination of Parts-per-Trillion
Concentrations of Tetra-. Hexa-. and
Octachlorodibenzo-p-dioxins in Human
Milk," Analytical Chemistry. 52. 2037 (1980).
8. "Carcinogens—Working with
Carcinogens." Department of Health,
Education, and Welfare. Public Health
Service, Center for Disease Control, National
Institute for Occupational Safety and Health,
Publication No. 77-206, August 1977.
9. "OSHA Safety and Health Standards.
General Industry." (29 CFR 1910),
Occuptional Safety and Health
Administration, OSHA 2206 (Revised.
January 1976).
10. "Safety in Academic Chemistry
Laboratories," American Chemical Society
Publication, Committee on Chemical Safety.
3rd Edition, 1979.
11. Provost, L. P., and Elder, R. S.,
"Interpretation of Percent Recovery Data."
American Laboratory. 15, 58-63 (1983). (The
value 2.44 used in the equation in Section
8.3.3 is two times the value 1.22 derived in
this report.)
12. ASTM Annual Book of Standards. Part
31. D3370-76, "Standard Practices for
Sampling Water." American Society for
Testing and Materials, Philadelphia.
13. "Methods, 330.4 (Titrimetric, DPD-FAS)
and 330.5 (Spectrophotometric DPD) for
Chlorine, Total Residual." Methods for
Chemical Analysis of Water and Wastes.
EPA-600/4-79-020, U.S. Environmental
Protection Agency, Environmental Monitoring
and Support Laboratory. Cincinnati. Ohio
45268. March 1979.
14. Wong. A.S. et al. 'The Determination of
2.3.7.8-TCDD in Industrial and Municipal
Wastewaters. Method 613, Part 1—
Development and Detection Limits." G.
Choudhay. L. Keith, and C. Ruppe. ed..
Butterworth Inc., (1983).
15. "EPA Method Validation Study 23.
Method 613 (2.3,7.8-Tetrachlorodibenzo-p-
dioxin)," Report for EPA Contract 68-03-2863
(In preparation).
TABLE 1.—Chromatographic Conditions and
Method Detection Limit
Parameter
2.3.7.8.-TC
CO _
Retention
time
(min)
13.1
Method
detection
limit (tig/
U
0.002
Column conditions: SP-2330 coated on a 60 m long x
0.25 mm ID glass cotumn with hydrogen carrier gas at 40
cm/sec linear velocity, splitless injection using tetradecane.
Column temperature held isothermal at 200*C tor 1 min, then
programmed at 8'C/min to 250 "C and held. Use of helium
carrier gas will approximately double the retention time.
TABLE 2.—QC Acceptance Criteria—Method
613
Parameter
2,3.7.8-TCOD
Test
cone.
°3'
o.too
Limit
tors
"L?'
0.0276
Range for X
ig/L (Section 8.2.4).
X= Average recovery for four recovery measurements, in
ng/L (Section 8.2.4).
P, P1=Percent recovery measured (Section 8.3.2. Section
8.4.2).
Note.—These criteria are based directly upon the method
performance data in Table 3. Where necessary, trie limits for
recovery have been broadened to assure applicability of the
limits to concentrations below those used to develop Table
TABLE. 3.—METHOD ACCURACY AND PRECISION AS FUNCTIONS OF CONCENTRATION—METHOD 613
Parameter
237 8-TCOD
Accuracy, as
recovery. X '
Oig/U
0 86C+0 00145
Single analyst,
precision, s,'
(M/L)
0 13)? -f-000129
Overall
precision. S '
(ji/g/U
X' = Expected recovery for one or more measurements, of a sample containing a concentration of C. in ug/L
V = Expected single analyst standard deviation of measurements at an average concentration found of X, in jig/L
S = Expected interlaboratory standard deviation of measurements at an average concentration found of X. in pg/L
C = True value for the concentration, in ug/L,
X = Average recovery found for measurements of samples containing a concentration of C. in pg/L.
702
-------
By Lloyd P Provost and Robert S. Elder
fiterpretacion of oercent
Bff* J t
L ...
'COvery ciata
fit SPIKED (FORTIFIED) sample studies, known
famounts of a compound or compounds of inter-
fest are added to aliquots of a sample, and the
ntage of analyte recovered by a test method is
to evaluate the performance of that method.
Environmental Protection Agency (EPA), for
aple, uses spiking studies in method develop-
(e.g., Ref. 1) and has proposed the use of
ked samples in quality control programs under
itional Pollutant Discharge Elimination System
t>DES) permits.2 Thus, the proper conduct and
pretation of spiking programs are critical to the
Ifevelopment and implementation of the analytical
.methods upon which important environmental pro-
-ams are based.
> Spiking is particularly useful in wastewater anal-
yses because the variety of sample matrices and the
number of analytes of interest in each sample make
realistic standard reference materials difficult to
produce. Spiking permits flexibility in the choice of
sample matrix and in the combinations and levels of
.analytes that can be evaluated. The usefulness of
spiked-sample analyses is not limited to wastewater
.or environmental samples, however, and proper in-
terpretation of data from such analyses (percent re-
covery data) is important whatever the application.
This paper will describe statistical properties of
percent recovery data when analytical bias and pre-
cision are proportional to sample concentration.
The impact of the presence of the analyte of interest
in the unspiked sample (i.e., nonzero background
concentration) will be examined, and some of the
potential pitfalls in the interpretation of percent re-
covery data in method development and quality
control applications will be discussed.
Assumptions
In investigating the statistical properties of per-
cent recovery data, we assume that the expected val-
Mr. Provost is Quality Assurance Director. Raitiun Corporation.
Mr. E/rfer is Senior Statistician. JRB Associates.
ue of a concentration measurement (X) for a
sample with true concentration B is
E (X) = pB
(1)
where 100 p is the mean percent recovery of the
method. If p = 1, the method is unbiased; other-
wise, its absolute bias is proportional to true con-
centration. We also assume that the variance of a
sample with concentration B is
Var(X) =
(2)
where 100 C is the coefficient of variation of the
method. That is, analytical precision is proportion-
al to concentration; the smaller C, the more precise
the method (on an absolute basis).
It is important to keep in mind that p and C, the
parameters that characterize the bias and precision
of the analytical method, are assumed constant with
respect to concentration. This is a realistic assump-
tion for many methods within their ranges of ap-
plicability. However, the values of p and C some-
times depend on the sample matrix involved, and
the value of C sometimes increases at low concen-
trations. These possible departures from the simple
properties assumed above often are investigated in
method development studies through statistical
analyses of percent recovery data (Ref. 1, for exam-
ple).
To estimate the mean and variance of percent re-
covery for a test method at a particular concentra-
tion, one typically analyzes n aliquots of a sample
spiked at that level and computes
and
5- = (n - !)-'!(/>, -
(3)
(4)
(e.g., Ref. 2), where P, denotes the observed percent
recovery for the /th aliquot. The statistical proper-
AMERICAN LABORATORY : 57
-------
ties of P" and s* are described separately below for
the cases of zero and nonzero background concen-
tration.
Zero background
'if the sample background concentration is known
to be zero (B = 0), the percent recovery is defined
as
P = 100 Y/T
(5)
where T is the spike level and X is the measured con-
centration for a spiked aliquot. The most common
case in which background concentration is known
to be zero is when spikes are added to distilled
water. It can be shown using the assumed properties
of the analytical method that in this case the mean
and variance of 7*" [defined in Eq. (3)] are
percent recovery for the method.
Nonzero background
If the sample background concentration is non-
zero (B>0), percent recovery may be defined as
P = 100 (Y -
where T is the spike level, Y is the measured concen-
tration of a spiked aliquot, and ~X is an estimate of
the background concentration based on the mean of
measurements on m unspiked aliquots [e.g., Eq
(3)]. If n aliquots of the same sample are spiked at
level 7" and analyzed, ~F and 5: can be computed as
described in Eqs. (3) and (4) and used to estimate
the mean and variance of percent recovery for the
analytical method. It can be shown in this case that
the mean and variance of 7r are $
and
- lOOp
VarC?5") = (lOOpC)2/n
(6)
(7)
and
Thus the sample average percent recovery (7s") is an
unbiased estimator of the mean percent recovery of
the method. It also can be shown that the mean of s2
[defined in Eq. (4)] is
E(s>) = (lOOpCY (8)
that is, s' is an unbiased estimator of the variance of
QOOpCy-
n
£(7r) = \00p
Var(F) =
\
where k = T/B (k may be termed the spike/bad^
ground ratio). These results show that 7s" also is a*
unbiased estimator of mean percent recovery in
case, but that the variance of 7s" is greater in
zero-background case by a factor that depend**
•"'$
*i?j
, ''"••'*
Table 1
Impact of spike-to-background rates on variability of percent recoveries
;'.*;
M
:> Spike-to-background
ratio (k)
Zero background
100
50
10
•>" ' ., 5
'•'"" - ' "1
!-!fT-'- ' 0.5
"•'" "->• -• ->. -.0.1
-------
RECOVERY DATA continued
the numbers of spiked and unspiked aliquots ana-
lyzed (n and m) and the ratio of spike and back-
ground concentrations (k). As the spike/back-
ground ratio decreases, it can be seen from Eq. (11)
that Var(7r) increases, and T5" becomes a poorer es-
timator of analytical percent recovery.
The consequence of this result is easily illustrated
by some examples. Table 1 shows the impact of k
on Var(75") and the expected range in recoveries for
three cases with only one spiked and one unspiked
aliquot analyzed (m = n = 1). The expected range
in recovery is based on a 95% tolerance interval for
a normal distribution:
[100 p ± 1.96VVar(FT]
As can be seen from Table 1, when k = 1,
is five times the zero-background value; when k =
0.1, Va^T5") is about 221 times the zero-background
value.
It can also be shown that the mean of s2 in the
nonzero-background case is
E(s~) = (100 pC)1 (1 + \/kY
(12)
This is greater than the result for B = 0 by a factor
that depends once again on k. For example, when
spike and background levels are equal (k = l),E(s2)
in Eq. (12) is four times the zero-background value;
when k = 1/5, £ (s1) is 36 times the zero-back-
ground value. Thus, s2 is a biased estimator of the
variance of percent recovery, and s: overestimates
that variance to a greater extent the smaller the
spike/background ratio.
Alternate definitions for percent recovery
The definitions above are not the only ones used
for percent recovery. One alternative definition
(based on expressing recovery as a percentage of the
total spiked sample concentration) is
/>= 100X
T+X
Another alternative (applicable when the spike level
is a multiple, h, of the estimated background con-
centration) is
p = IOOQ-- T)
Regardless of hov\ percent recovery is defined, it
60 : DECEMBER 1983
can be shown that percent recovery data tend to be
unreliable when the spike/background ratio is
small.
.'* •- '.. *
'*;
Interpreting percent recovery data
Two issuei:were investigated in the method eval-
uation studies for all of EPA's 600 series methods:
1. Does met hod performance depend on the
sample matrix involved; e.g., do p and ~C values dif-
fer for distilled, natural water, and waisitewater sam-
ples?
2. Does performance depend on the "sample con-
centration; e.g., is C larger at lower concentrations?
These questions were investigated'by analyzing
spiked aliquots of both distilled and natural water
samples and by spiking aliquots of given samples at
different levels (e.g., Ref. 1).
We have shown that for samples with nonzero
background, such as wastewater samples, VarfP'jis
large when k is small. Thus the estimate of p fora
wastewater sample with fl>0 and k small may differ
greatly from the true mean percent recovery of the
method and, therefore, may appear to differ from;
the corresponding estimate for a distilled water.
sample. That is, the different statistical properties]
of percent recovery data in the zero- and nonzero-"
background cases may mislead one to conclude that}
matrix differences affect mean analytical percent
recovery. ' ..•*'
We also have shown that for samples with noil-
zero background, s: tends to overestimate the van-
ance of percent recovery to a greater extent the
smaller the spike/background ratio. Thus, if s' vat.
ues from different spike levels are compared whes
fi>0, it is likely to appear that relative precision ».
poor at lower concentrations even when it is not. -
The discussion above shows two misconception*.
that can arise in method development due to
terprctation of percent recovery data. The
cal properties of such data may also lead to ""
rection in analytical quality control pr
example. EPA's Handbook for Analytical
Control in Water and U'astewater
suggests that when analytical precision varies
concentration, separate control charts
kept for different concentration ranges.' I/
tionship of precision to concentration is
gated by estimating variances of percent.
data at different spike levels, one may eirc
conclude (if 5>0) that separate charts are
for low concentrations. This would-increase ^
of proems control activities unnecessarily
in ihe QA appendix to the proposed^
705
-------
M.., ''.}..' RECOVERY DATA continued
•ri.- "'•: . methods,2 EPA recommends spiking at levels equal
to 1, 9, and 99 times'background, computing T5" and
s2 at each level, and comparing these estimates for
different levels. Results at the lowest spike level
may well appear to be more variable than results at
higher levels because of the dependence of E (s2) on
.'•>.;•• .the spike/background ratio.
• An example of the importance of knowing the
" " statistical properties of percent recovery data is il-
. . lustrated in a report by the Chemical Manufacturers
-I'-; .">" Association (CMA) on results of a joint CMA/EPA
study .of the quality of wastewater from five organic
..>;-.' .•;- chemicals plants." One objective of the study was to
•v.---. ,v characterize the mean and standard deviations of
"-''•'' 1, percent recovery for the analytical methods used to
v "measure organic priority pollutants in the industry's
' wastewater. The report generally was critical of the
capabilities of the analytical methods used. How-
p••; ever, in planning the study it was decided that the
- . , spike level should approximately equal the back-
ground concentration; thus, overestimation of per-
cent recovery variability was built into the study.
Furthermore, spike/background ratios less than
one sometimes were employed. The consequences
were reflected most dramatically in results for acryl-
onitrile: the 12 influent percent recovery results for
this compound ranged from - 7000% to 400% and
had mean and standard deviation of -465 and
2060, respectively, due to the two extreme result
The estimated spike/background ratios for the ab-\
quots with -7000% and 400% recovery wotj-
10,000/890,000 = 0.011 and 10,000/210,000 "4i
0.048, respectively. By Eq. (11), it can be seen that'
the standard deviation of 7s" is about 14,200 pC fat,
k = 0.01 and m = n = 1; thus, a result of - 7000%
is not surprising when k is this small. Proper consid*'
eration of the statistical properties of percent reco*^
:ery data would have led to the choice of a higbff,
spike/background ratio in planning the study
to the exclusion from summary statistics of
values made meaningless by the use of too
spike/background ratios.
Another example of potential misinterpretati*
of data from a spiking study can be found in
As described, spiking studies were used to
performance of laboratories. The authors
eluded that overall performance by the five 1
tories in the study was poor. In one test,
known freshwater sample was analyzed
without spikes of various minerals. The
spike/background ratios for the six minerals^
as follows: 0.14, 1.1,0.21, 5.0, 0.71, and 5.0--
of the variability in recoveries attributed
laboratory performance may have been du.
statistical properties of recoveries WB»J
spike/background ratios.
62
DECEMBER 1983
-------
V It should be noted that these examples of prob-
lems in the interpretation of percent recqy.ery data
from the area of wastewater analysis were selected
because this is the application with which the au-
thors are most familiar. There is no reason to doubt
that similar examples could be found in other areas
such as clinical and agricultural chemistry..
of a method near the detection limit), it may be difr
ficult to obtain low background levels in the .sample.
matrix of interest. If it then becomes necessary to
perform spiking' studies. with a low spjkWback-
ground ratio, the statistical properties of the recov-
eries should be considered in interpreting the'results
and in comparing^them to results at other concen-
trations or in other matrices. ,'. '.'!','•
Summary . • . '.'..;
The statistical properties of percent recovery data
are important to consider when interpreting results
of analytical studies. When background quantities
of the spiked artalyte are present, percent ^recoveries
can be highly' variable a'hid estimates of analytical
precision can be b iased.
In designing method evaluation studies, spike
kvels should first be chosen to cover the range of
. concentration of interest. Once these levels are de-
termined, then sample matrices with background
kvels that are small compared to these spiking lev-
els should be chosen for the study. If this is done,
^en the statistical properties of percent, recovery
data will not affect the evaluation of method bias
and precision.
In some situations (e.g., studying the properties
References •:!•:"•"'•-' • '. 'r, •..:'1>"
1. WINTER, J.P.. BRjTJON; P.. CLEMENTS, H.,.and KRONER.
R., "EPA Method,Study 8, Total Mercury in Water," U.S.
Environmental Protection Agency, Environmental Monitor-
ing and Support Laboratory, Cincinnati (1977). "
2. Environmental Protection Agency, "Guidelines'"iEstabJishing
Test Procedures for.tJijiAnalysis of Po{\iirai\ts'i''j]federat Reg-
ister 44 (223), 694^MJ9J75,(December 3, 1979).•\'l, " '
3. Environmental Monitoring and,Support.Labpratory,, Hand-
book/or Analytical Quality Control in WaKr..qnd Waste-
water Laboratories, EPA-600/4-79^9, 'Office of Research.
and Developrnent;:u;S. -t'^A, Cincihnati', M"a'r'cn, J979.
4. Engineering-'SL-ience, Inc.,^MA/EPA Five-Planf Study, pre-
pared for Chemical Manufacturers Association, 3-20-3-22
VD-3. Austin, TX, April, 1982. ' 's'l;
5. EDWARDS, R.R., SCHILLING, D.L., and ROS'MILLER. T.L.,
"A performance evaluation of certified water analysis labora-'
tories,"y. Wat. Poll. Contr. Fed. 49, 1704(1977), :
AMERICAN LABORATORY : 63
'0.
------- |