SEFA
United States
Environmental Protection
Agency
Office c«
Toxic Substances
Washington DC 20430
EPA-56CH5-85-010
M, 1985
ubstances
Analytical Method:
The Analysis of
By-Product Chlorinated
Biphenyls in Commercial
Products and Product
Wastes, Revision 2
-------
ANALYTICAL METHOD: THE ANALYSIS OF BY-PRODUCT
CHLORINATED BIPHENYLS IN COMMERCIAL PRODUCTS
AND PRODUCT WASTES, REVISION 2
by
Mitchell D. Erickson
WORK ASSIGNMENT NO. 6
SPECIAL REPORT NO. 1
EPA Contract No. 68-02-3938
MRI Project No. 8201-A(6)
May 20, 1985
For
U.S. Environmental Protection Agency
Office of Toxic Substances
Field Studies Branch, TS-798
401 M Street, SW
Washington, DC 20460
Attn: Joseph J. Breen, Project Officer
Daniel T. Heggem, Work Assignment Manager
-------
DISCLAIMER
This document has been reviewed and approved for publication by the
Office of Toxic Substances, Office of Pesticides and Toxic Substances, U.S.
Environmental Protection Agency. The use of trade names or commercial prod-
ucts does not constitute Agency endorsement or recommendation for use.
-------
PREFACE
This report contains an analytical method for the analysis of by-
product polychlorinated biphenyls in commercial products and wastes. The
work was done on Work Assignment No. 6 on U.S. Environmental Protection Agency
Contract No. 68-02-3938. This is the second revision of the method. Previous
revisions are cited as references 3 and 4 in the method. This report was
prepared by Mitchell Erickson. The work on the previous revisions was con-
ducted by Dr. Erickson, John Stanley, Kay Turman, Gil Radolovich, Karin Bauer,
Jon Onstot, Donna Rose, Margaret Wickham, and Ruth Blair. The work for the
previous revisions was performed on Task 51 of EPA Contract No. 68-01-5915.
Two companion methods have been published which address air (Special
Report No. 2, EPA Report No. EPA-560/5-85-011) and water (Special Report No. 3,
EPA Report No. EPA-560-5-85-012).
The EPA Work Assignment Manager, Daniel T. Heggem, of Field Studies
Branch provided helpful guidance.
MIDWEST RESEARCH INSTITUTE
Clarence L. Haile
DepUtty Program Manager
5hn E. Going
'rogram Manager
Approved:
James L. Spigarelli, Director
Chemical and Biological Sciences
Department
m
-------
TABLE OF CONTENTS
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
16.0
17.0
References
Scope and Application
Summary
Interferences
Safety « . . .
Apparatus and Materials
Reagents
Calibration
Sample Collection, Handling, and Preservation
Sample Preparation
Gas Chromatographic/Electron Impact Mass Spectrometric
Determination
Qualitative Identification
Quantitative Data Reduction
Confirmation
Quality Assurance
Quality Control
Method Performance
Documentation and Records
Page
1
3
4
5
6
9
14
24
26
35
38
39
41
45
46
49
49
51
-------
LIST OF TABLES
Number Page
1 Numbering of PCB Congeners 2
2 DFTPP Key Ions and Ion Abundance Criteria 8
3 Concentrations of Congeners in PCB Calibration Standards
for Full Scan Analysis (ng/[jL) 10
4 Concentrations of Congeners in PCB Calibration Standards
for Selected Ion Monitoring and Limited Mass Scan
Analysis (pg/pL) 11
5 Composition of Internal Standard Spiking Solution (SS100)
Containing 13C-Labeled PCBs 13
6 Operating Parameters for Capillary Column Gas Chromato-
graphic System 15
7 Operating Parameters for Packed Column Gas Chromatography
System 16
8 Operating Parameters for Quadrupole Mass Spectrometer
System 17
9 Operating Parameters for Magnetic Sector Mass Spectrometer
System 18
10 Limited Mass Scanning (LMS) Ranges for PCBs 19
11 Characteristic SIM Ions for PCBs 21
12 Pairings of Analyte and Calibration Compounds 23
13 Relative Retention Time (RRT) Ranges of PCB Homologs Versus
d6-3,3' ,4,4'-Tetrachlorobiphenyl 25
14 Characteristic Ions for Internal Standards and 13C-Labeled
PCB Surrogates 40
15 Analysis Worksheet 42
16 Analysis Report 44
17 Method Performance Parameters 50
-------
LIST OF FIGURES
Number Page
1 Reconstructed ion chromatogram of calibration solution
FS100 ng PCB obtained in the full scan mode 36
VII
-------
THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN
COMMERCIAL PRODUCTS AND PRODUCT WASTES
1.0 Scope and Application
1.1 This is a gas chromatographic/electron impact mass spectrometric
(GC/EIMS) method applicable to the determination of chlorinated
biphenyls (PCBs) in commercial products and product wastes. The
PCBs present may originate either as synthetic by-products or as
contaminants derived from commercial PCB products (e.g., Aroclors).
The PCBs may be present as single isomers or complex mixtures and
may include all 209 congeners from monochlorobiphenyl through
decachlorobiphenyl listed in Table 1.
This method was prepared for use in demonstrating compliance with
the EPA rules regarding the generation of PCBs as by-products in
commercial chemical production1'2 and is based on earlier ver-
sions.3'4 This revision includes elimination of a calculation
which corrects the native PCB concentration based on the recovery
of the l3C-labeled PCB recovery surrogates. In addition, full
scan is now emphasized over the selected ion monitoring and lim-
ited mass scan mass spectrometric data collection modes. The
latter two techniques provide less qualitative information and
should be used only if needed to achieve the required sensitivity.
Additional background information on selection of the techniques
has also been published.5
1.2 The detection and quantitation limits are dependent upon the com-
plexity of the sample matrix and the ability of the analyst to
remove interferents and properly maintain the analytical system.
For low level PCBs (0.05 to 0.2 H9/9). the method precision ap-
pears to be about ± 60% based on a very limited study. The method
performance at higher levels (e.g., 2 M9/9) will be determined in
future studies.
1.3 This method is restricted to use by or under the supervision of
analysts experienced in the use of gas chromatography/mass spec-
trometry (GC/MS) and in the interpretation of gas chromatograms
and mass spectra. Prior to sample analysis, each analyst must
demonstrate the ability to generate acceptable results with this
method by following the procedures described in Section 15.2.
1.4 The validity of the results depends on equivalent recovery of the
analyte and 13C PCBs. If the 13C PCBs are not thoroughly incor-
porated in the matrix, the method is not applicable.
-------
TABLE 1. NUMBERING OF PCB CONGENERS3
HO.
1
2
3
4
5
6
7
8
9
10
11
12
11
14
15
16
17
18
19
20
21
22
71
74
?S
If,
77
28
29
30
31
32
33
14
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Structure
HB«oe>i1ore65.5'.6
2. 2', 3. 3' ,5. 6. S1
2, 2', 3, 4. 4', 5, 5'
2, 2', 3, 4, 4'. 5.6
2.2'.3,414I,5,6'
2,2',3,4.4',5-.6
2. 2'. 3. 4, 4', 6, 6'
2,2' .3. ,5.5'. 6
2, 2'. 3, ,5,6,6'
2,2',3. '.S.S'.S
2,2',3i '.5,6,6'
2.3,3', .4', 5,5'
2,3,3', .4'. 5,6
2,3,3'. ,4',S',6
2.3.3', .5, 5'. 6
2, 3, 3' ,4' ,5,5' .6
OetJCftlorabi phenyl s
2, 2'. 3, 3', 4,4', 5, 5'
2, 2', 3, 3', 4,4', 5, 6
2, 2', 3, 3' .4. 4'. 3. 5'
2, 2', 3, 3', 4,4', 5, 6'
2,2'.3,3'14,S.5'.6
2, 2'. 3,3'. 4,5,6, 6'
2, 2' ,3, 3' .4.5'. 6. 6'
2. 2'. 3, 3', 4. 5. 5'. 5'
2, 2'. 3, 3'. 5. 5', 6, 6'
2.2' .3,4. 4' .5.5' .6
2.21,3.4,4'.S.6,6'
2.3.3* .4,4',5.S'.6
NonaeMoroblonefyls
2.2'. 3, 3'. 4, 4', 5. 5'. 6
2.2', 3,3' .4, 4'. 5, 5.6'
2,2', 3,3'. 4, 5. 5', 6, 6'
Deeaetiloroblphenyl
2,2',3,3'*,4'.5,5'.6.o1
•»dopt*d froa a*ll*cMt*r. X. tfid ZeM, N.. FrtuMus Z. Anal. Chen., 302. 20-31 (1980).
-------
1.5 During the development and testing of this method, certain analyti-
cal parameters and equipment designs were found to affect the valid-
ity of the analytical results. Proper use of the method requires
that such parameters or designs must be used as specified. These
items are identified in the text by the word "must." Anyone wish-
ing to deviate from the method in areas so identified must demon-
strate that the deviation does not affect the validity of the data.
Alternative test procedure approval must be obtained from the
Agency. An experienced analyst may make modifications to param-
eters or equipment identified by the term "recommended." Each
time such modifications are made to the method, the analyst must
repeat the procedure in Section 15.2. In this case, formal ap-
proval is not required, but the documented data from Section 15.2
must be on file as part of the overall quality assurance program.
1.6 This method contains many options because of the diversity of
matrices and interferences which may be encountered. Once the
appropriate options for each sample type have been selected, each
laboratory should prepare a written step-by-step protocol for use
by the analysts. The protocol may contain verbatim sections from
this method, more detailed steps1for certain techniques, or
totally different extraction or cleanup techniques.
2.0 Summary
2.1 The process or product must be sampled such that the specimen col-
lected for analysis is representative of the whole. Statistically
designed selection of the sampling position, time, or discrete
product units should be employed. The sample must be preserved
to prevent PCB loss prior to analysis. Customary inventory stor-
age may be adequate for products. For intermediates, process
samples, and other non-product specimens, other preservation
techniques may be needed.
2.2 The sample is mechanically homogenized and subsampled if necessary.
The sample must then be spiked with four 13C PCB surrogates
(4-chloro-13C6-biphenyl; 3,3',4,4'-tetrachloro-13C12-biphenyl;
2,2' ,3,3',-5,5',6,6l-octachloro-13C12~biphenyl; and decachloro-
13C12-bipnenyl) and the surrogates incorporated by further
mechanical agitation.
2.3 The surrogate-spiked sample is extracted and cleaned up at the
discretion of the analyst. Simple dilution or direct injection
is permissible. Possible extraction techniques include liquid-
liquid partition, thermal desorption, and sorption onto resin
columns followed by solvent desorption. Cleanup techniques may
include liquid-liquid partition, sulfuric acid cleanup, saponifi-
cation, adsorption chromatography, high performance liquid chro-
matography, gel permeation chromatography, or a combination of
cleanup techniques. The sample is diluted or concentrated to a
final known volume for instrumental determination.
-------
2.4 The PCB content of the sample extract must be determined by high
resolution (preferred) or packed column gas chromatography/electron
impact mass spectrometry (HRGC/EIMS or PGC/EIMS) operated in the
full scan, selected ion monitoring (SIM), or limited mass scan
(LMS) mode.
2.5 PCBs are identified by comparison of their retention time and mass
spectral intensity ratios to those in calibration standards.
2.6 PCBs are quantitated by the internal standard technique, using
response factors for a mixture of 10 PCB congeners. The recover-
ies of four 13C surrogates are used to monitor for losses in
workup and determination.
2.7 The PCBs identified by the SIM technique may be confirmed by full
scan HRGC/EIMS, retention on alternate GC columns, other mass spec-
trometric techniques, infrared spectrometry, or other techniques,
provided that the sensitivity and selectivity of the technique
are demonstrated to be comparable or superior to GC/EIMS.
2.8 The analysis time is dependent on the extent of workup employed.
The time required for instrumental analysis of a single sample,
excluding instrumental calibration, data reduction and reporting,
is typically 30 to 45 min.
2.9 A quality assurance (QA) plan must be developed for each labora-
tory.
2.10 Quality control (QC) measures include laboratory certification
and performance check sample analysis, procedural QC (instrumental
performance, calculation checks), and sample QC (blanks, repli-
cates, and standard addition).
3.0 Interferences
3.1 Method interferences may be caused by contaminants in solvents,
reagents, glassware, and other sample processing hardware, leading
to discrete artifacts and/or elevated baselines in the total ion
current profiles. All of these materials must be routinely demon-
strated to be free from interferences by the analysis of labora-
tory reagent blanks as described in Section 15.0.
3.1.1 Glassware must be scrupulously cleaned. All glassware
should be cleaned as soon as possible after use by rins-
ing with the last solvent used. This should be followed
by detergent washing with hot water and rinses with tap
water and reagent 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 this treatment. Solvent
rinses with acetone and pesticide quality hexane may be
-------
substituted for the muffle furnace heating. Volumetric
ware should not be heated in a muffle furnace. After it
is dry and cool, glassware should be sealed and stored
in a clean environment to prevent any accumulation of
dust or other contaminants. It is stored inverted or
capped with aluminum foil.
3.1.2 The use of high purity reagents and solvents helps to
minimize interference problems. Purification of sol-
vents by distillation in all-glass systems may be re-
quired. All solvent lots must be checked for purity
prior to use.
3.2 Matrix interferences may be caused by contaminants that are coex-
tracted from the sample. The extent of matrix interferences will
vary considerably from source to source, depending upon the nature
and diversity of the sources of samples.
4.0 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 lab-
oratory 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 han-
dling sheets should also be made available to all personnel in-
volved in the chemical analysis.
4.2 Polychlorinated biphenyls have been tentatively classified as
known or suspected human or mammalian carcinogens. Primary stan-
dards of these toxic compounds should be prepared in a hood.
Personnel must wear protective equipment, including gloves and
safety glasses.
Congeners highly substituted at the meta and para positions and
unsubstituted at the ortho positions are reported to be the most
toxic. Extreme caution should be taken when handling these com-
pounds neat or in concentrated solutions. This class includes
3,3',4,4'-tetrachlorobiphenyl (both natural abundance and isotop-
ically labeled).
4.3 Diethyl ether should be monitored regularly to determine the per-
oxide content. Under no circumstances should diethyl ether be
used with a peroxide content in excess of 50 ppm, as an explosion
could result. Peroxide test strips manufactured by EM Labora-
tories (available from Scientific Products Company, Cat. No.
P1126-8 and other suppliers) are recommended for this test. Pro-
cedures for removal of peroxides from diethyl ether are included
in the instructions supplied with the peroxide test kit.
-------
4.4 Waste disposal must be in accordance with RCRA and applicable
state rules.
5.0 Apparatus and Materials
5.1 Sampling containers - Amber glass bottles, 1-L or other appropri-
ate volume, fitted with screw caps lined with Teflon are appro-
priate for liquid samples. Cleaned foil may be substituted for
Teflon if the sample is not corrosive. If amber bottles are not
available, samples should be protected from light using foil or
a light-tight outer container. The bottle must be washed, rinsed
with acetone or methylene chloride, and dried before use to mini-
mize contamination. Non-liquid samples may require other con-
tainers. Gas bags and stainless steel cannisters may be appro-
priate for gas samples. Bags or cans may be appropriate for
solid samples. Consumer products in small containers may remain
in the manufacturer's packaging.
5.2 Glassware - All specifications are suggestions only. Catalog
numbers are included for illustration only.
5.2.1 Volumetric flasks - Assorted sizes.
5.2.2 Pipets - Assorted sizes, Mohr delivery.
5.2.3 Micro syringes - 10.0 pL for packed column GC analysis,
1.0 pL for on-column GC analysis.
5.2.4 Chromatographic column - Chromaflex, 400 mm long x 19 mm
ID (Kontes K-420540-9011 or equivalent).
5.2.5 Kuderna-Danish Evaporative Concentrator Apparatus
5.2.5.1 Concentrator tube - 10 mL, graduated (Kontes
K-570050-1025 or equivalent). Calibration must
be checked. Ground glass stopper size ($19/22
joint) is used to prevent evaporation of solvent.
5.2.5.2 Evaporative flask - 500 mL (Kontes K-57001-0500
or equivalent). Attached to concentrator tube
with springs (Kontes K-662750-0012 or equivalent).
5.2.5.3 Snyder column - Three ball macro (Kontes
K-503000-0121 or equivalent).
5.3 Balance - Analytical, capable of accurately weighing 0.0001 g.
-------
5.4 Gas chromatography/mass spectrometer system.
5.4.1 Gas chromatograph - An analytical system complete with a
temperature programmable gas chromatograph and all re-
quired accessories including syringes, analytical columns,
and gases. The injection port must be designed for on-
column injection when using capillary columns or packed
columns. Other capillary injection techniques (split,
splitless, "Grob," etc.) may be used provided the per-
formance specifications stated in Section 7.1 are met.
5.4.2 High resolution (capillary) GC column - A 10-30 m long x
0.25 mm ID fused silica column with a 0.25 urn thick DB-5
bonded silicone liquid phase (J&W Scientific) is recom-
mended. Alternate liquid phases may include OV-101,
SP-2100, Apiezon L, Dexsil 300, or other liquid phases
or columns which meet the performance specifications
stated in Section 7.1.
5.4.3 Packed GC column - A 180 cm x 0.2 cm ID glass column
packed with 3% SP-2250 on 100/120 mesh Supelcoport or
equivalent is recommended. Other liquid phases or
columns which meet the performance specifications stated
in Section 7.1 may be substituted.
5.4.4 Mass spectrometer - Must be capable of scanning from m/z
150 to m/z 550 every 1.5 sec or less, collecting at least
five spectra per chromatographic peak, utilizing a 70-eV
(nominal) electron energy in the electron impact ioniza-
tion mode and producing a mass spectrum which meets all
the criteria in Table 2 when 50 ng of decafluorotriphenyl
phosphine [DFTPP, bis(perfluorophenyl)phenyl phosphine]
is injected through the GC inlet. Any GC-to-MS interface
that gives acceptable calibration points at 10 ng per
injection for each PCB isomer in the calibration stan-
dard and achieves all acceptable performance criteria
(Section 10) may be used. Direct coupling of the fused
silica column to the MS is recommended. Alternatively,
GC-to-MS interfaces constructed of all glass or glass-
lined materials are recommended. Glass can be deacti-
vated by silanizing with dichlorodimethylsilane.
5.4.5 A computer system that allows the continuous acquisition
and storage on machine-readable media of all mass spectra
obtained throughout the duration of the chromatographic
program must be interfaced to the mass spectrometer.
The data system must have the capability of integrating
the abundances of the selected ions between specified
limits and relating integrated abundances to concentra-
tions using the calibration procedures described in this
method. The computer must have software that allows
searching any GC/MS data file for ions of a specific mass
-------
rw ^^':J.::f'TabTe';2. DFTPP Key Ions and Ion Abundance Criteria
(':' mTz Ion abundance criteria
197 Less than 1% of mass 198
198 100% relative abundance
199 5-9% of mass 198
275 10-30% of mass 198
365 Greater than 1% of mass 198
441 Present, but less than mass 443
442 Greater than 40% of mass 198
443 17-23% of mass 442
-------
and plotting such ion abundances versus time or scan
number to yield an extracted ion current profile (EICP).
Software must also be available that allows integrating
the abundance in any EICP between specified time or scan
number limits.
5.5 Chromatographic systems.
5.5.1 Gel permeation chromatography.
5.5.1.1 GPC Autoprep 1002 (Analytical Biochemistry
Laboratories, Inc.) or equivalent.
5.5.1.2 A Bio-Beads SX-3 (Bio-Rad) column.
5.5.2 High performance liquid chromatography.
5.5.2.1 Waters Model 6000A pump or equivalent.
5.5.2.2 Waters Model 440 UV detector or equivalent.
5.5.2.3 Rheodyne 7125 injector or equivalent.
5.5.2.4 Amine column (Waters uBondapak, 3.9 x 300 mm)
or equivalent.
5.5.3 Gas chromatograph for GC/FID screening.
5.5.3.1 A temperature-programmable GC equipped with a
flame ionization detector. Varian 3740 or
equivalent.
5.5.3.2 A 2 m x 2 mm ID glass column packed with 3%
SP-2250 on 100/120 mesh Supelcoport or equiva-
lent. A high resolution GC column may also be
used.
6.0 Reagents
6.1 Solvents - All solvents must be pesticide residue analysis grade.
New lots should be checked for purity by concentrating an aliquot
by at least as much as is used in the procedure. HPLC solvents
should have UV cutoffs of 210 nm or less.
6.2 Calibration standard congeners - Standards of the PCB congeners
listed in Tables 3 and 4 are available from Ultra Scientific,
Hope, Rhode Island; or Analabs, North Haven, Connecticut.
-------
Table 3. Concentrations of Congeners in PCB Calibration Standards
for Full Scan Analysis (ng/pL)
Homolog
1
2
3
4
5
6
7
8
9
10
4
-
-
13C-Cli
13C-C14
13c-ci8
13c-ci10
Congener
no.
1
7
30
50
97
143
183
202
207
209
210 (IS)
C10H7I (IS)b
C18Di2 dS)C
211 (RS)
212 (RS)
213 (RS)
214 (RS)
FS100
ng PCB
100
100
150
200
200
200
300
300
450
200
250
250
250
100
250
400
500
FS050
ng PCB
50
50
75
100
100
100
150
150
225
100
250
250
250
50
125
200
250
FS010
ng PCB
10
10
15
20
20
20
30
30
45
20
250
250
250
10
25
40
50
FS005
ng PCB
5
5
7.5
10
10
10
15
15
22.5
10
250
250
250
5
12.5
20
25
FS001
ng PCB
1
1
1.5
2
2
2
3
3
4.5
2
250
250
250
1
2.5
4
5
^Concentrations given as examples only.
1-Iodonaphthalene.
d12~Chrysene.
10
-------
Table 4. Concentrations of Congeners in PCB Calibration Standards
for Selected Ion Monitoring and Limited Mass Scan
Analysis (pg/pL)
Homo log
1
2
3
4
5
6
7
8
9
10
4
-
-
"C-Clj
13C-C14
13c-ci8
13c-ci10
Congener
no.
1
7
30
50
97
143
183
202
207
209
210 (IS)
C10H7I (IS)b
Ci8D12 (IS)C
211 (RS)
212 (RS)
213 (RS)
214 (RS)
SIM1000
pg PCB
1,000
1,000
1,500
2,000
2,000
2,000
3,000
3,000
4,500
2,000
250
250
250
1,000
2,500
4,000
5,000
SIM100
pg PCB
100
100
150
200
200
200
300
300
450
200
250
250
250
100
250
400
500
SIM050
P9 PCB
50
50
75
100
100
100
150
150
225
100
250
250
250
50
125
200
250
SIM010
pg PCB
10
10
15
20
20
20
30
30
45
20
250
250
250
10
25
40
50
.Concentrations given as examples only.
1-Iodonaphthalene.
d12-Chrysene.
Jl
-------
6.3 Calibration standard stock solutions - Primary dilutions of each
of the individual PCBs listed in Tables 3 and 4 are prepared by
weighing approximately 1-10 mg of material within 1% precision.
The PCB is then dissolved and diluted to 1.0 ml with hexane. The
concentration is calculated in mg/mL. The primary dilutions are
stored at 4°C in screw-cap vials with Teflon cap liners. The
meniscus is marked on the vial wall to monitor solvent evaporation.
Primary dilutions are stable indefinitely if the seals are main-
tained. The stock solutions and dilutions should be clearly la-
beled with pertinent information such as sample code, solvent,
date prepared, initials of person preparing the solution, and
notebook reference.
6.4 Working calibration standards - Working calibration standards are
prepared that are similar in PCB composition and concentration to
the samples by mixing and diluting the individual standard stock
solutions. Example calibration solutions are shown in Tables 3
and 4. The mixture is diluted to volume with pesticide residue
analysis quality hexane. The concentration is calculated in
ng/mL as the individual PCBs. Dilutions are stored at 4°C in
narrow-mouth, screw-cap vials with Teflon cap liners. The menis-
cus is marked on the vial wall to monitor solvent evaporation.
These secondary dilutions can be stored indefinitely if the seals
are maintained.
These solutions are designated FSxxx ng PCB and SIMxxx pg PCB
where the xxx is used to encode the nominal concentration of the
lower congeners in ng/uL and pg/uL, respectively. The FS prefix
helps aid the analyst in identifying solutions which are appro-
priate for full scan analysis; the SIM prefix is for solutions to
calibrate in the selected ion monitoring and limited mass scan
acquisition modes.
6.5 Alternatively, certified stock solutions similar to those listed
in Tables 3 and 4 may be available from a supplier, in lieu of
the procedure described in Section 6.4.
6.6 DFTPP standard - A 50-ng/uL solution of decafluorotriphenylphos-
phine (DFTPP, PCR Research Chemicals, Gainesville, FL) is pre-
pared in acetone or another appropriate solvent.
6.7 Surrogate standard stock solution - The four 13C-labeled PCBs
listed in Table 5 are available as a certified solution from Toxic
and Hazardous Materials Repository, U.S. Environmental Protection
Agency, Environmental Monitoring and Support Laboratory, 26 West
St. Clair Street, Cincinnati, Ohio 45268, (513) 684-7327. This
solution may be used as received or diluted further. These solu-
tions are designated "SSxxx," where the xxx is used to encode the
nominal concentration in (jg/mL.
6.8 Internal standard solutions - Solutions of d6-3,3',4,4'-tetra-
chlorobiphenyl (KOR Isotopes, Cambridge, MA), 1-iodonaphthalene
(Aldrich Chemical Company, Milwaukee, WI) or d12-chrysene (KOR
12
-------
Table 5. Composition of Internal Standard Spiking Solution (SS100)
Containing 13C-Labeled PCBs
Congener
no.
211
212
213
214
Compound
4-Chloro-d1 ,2' ,3' ,4' ,5' ,6'-13C6)-biphenyl
3,3' ,4,4'-Tetrachloro-(13C12)-biphenyl
2,2' ,3,3' ,5,5' ,6,6'-Octachloro-(13C12)-biphenyl
Decachloro-(13C12)-biphenyl
Abbreviations
13C-Clj
13C-C14
13c-ci8
13c-ci10
Concentration
(|jg/mL)
100
250
400
500
-------
Isotopes, Cambridge, MA) are prepared at nominal concentrations
of 1-10 mg/mL in hexane. The solutions are further diluted to
give working standards.
Note: Any internal standard may be used, provided it meets the
following criteria: (a) it is not already present in the sample,
(b) it gives a strong, recognizable mass spectrum, (c) it does
not give mass spectral ions which interfere with native or 13C-
labeled PCB quantisation, (d) it is chemically stable, and (e) it
elutes in the PCB retention window. Ideally, several internal
standards are used which have retention times spanning the PCB
retention windows to improve the response factor precision.
6.9 Solution stability - The calibration standard, surrogate, and
DFTPP solutions should be checked frequently for stability. These
solutions should be replaced after 6 months, or sooner if compar-
ison with quality control check samples indicates compound degrada-
tion or concentration change.
7.0 Calibration
7.1 The gas chromatograph must meet the minimum operating parameters
shown in Tables 6 and 7, daily. If all criteria are not met, the
analyst must adjust conditions and repeat the test until all cri-
teria are met.
7.2 The mass spectrometer must meet the minimum operating parameters
shown in Tables 2, 8, and 9, daily. If all criteria are not met,
the analyst must retune the spectrometer and repeat the test un-
til all conditions are met.
7.2.1 Full scan data acquisition - Quadrupole mass spectrom-
eters must meet the tuning criteria in Table 2. The
spectrometer must scan between m/z 150-550, although
wider scan ranges are permissible.
7.2.2 Limited mass scan data acquisition - Table 10 presents
a suggested set of IMS ranges. The mass spectrometer
should be set to at least'unit resolution. The computer
acquisition parameters should utilize the minimum thres-
hold filtering necessary so as not to lose pertinent data.
Optimum acquisition parameters will vary depending on
the condition of the mass spectrometer and should be
checked daily.
14
-------
Table 6. Operating Parameters for Capillary Column Gas Chromatographic System
Parameter
Recommended
Tolerance
Gas chromatograph
Column
Liquid phase
Liquid phase thickness
Carrier gas
Carrier gas velocity
Injector
Injector temperature
Injection volume
Initial column temperature
Column temperature program
Separator
Transfer line temperature
Tailing factor
Peak width1
Finnigan 9610
15 -30 m x 0.255 mm ID
Fused silica
DB-5 (J&W)
0.25 urn
Helium
30-45 cm/sb
"Grob" (split/splitless
mode)
250-270°C
1.0-2.0 uL
60-80°C (2 min)d
70°-300°C at 10°C/mine
Nonef
280°C
0.7-1.5
7-10 s
Othera
Other
Other nonpolar
or semipolar
< 1 urn
Hydrogen
Optimum performance
Other0
Optimum performance
Other
Otherd
Other
Glass jet or other
Optimum^
0.4-3
< 15 s
Substitutions permitted with any common apparatus or technique provided
.performance criteria are met.
Measured by injection of air or methane at 270°C oven temperature.
Manufacturer's instructions should be followed regarding injection tech-
.nique.
With on-column injection, initial temperature equals boiling point of the
solvent; in this instance, hexane.
C^Clio elutes at 270°C. Programming above this temperature ensures a
fclean column and lower background on subsequent runs.
Fused silica columns may be routed directly into the ion source to prevent
separator discrimination and losses.
9High enough to elute all PCBs, but not high enough to degrade the column
.if routed through the transfer line.
Tailing factor is width of front half of peak at 10% height divided by
width of back half of peak at 10% height for single PCB congeners in solu-
•tion FSxxx ng PCB or SIMxxx pg PCB.
Peak width at 10% height for a single PCB congener in FSxxx ng PCB or
SIMxxx pg PCB.
15
-------
Table 7. Operating Parameters for Packed Column Gas Chromatography System
Parameter
Recommended
Column packing
Carrier gas
Carrier gas flow rate
Injector
Injector temperature
Injection volume
Initial column temperature
Column temperature program
Separator
Transfer line temperature
Tailing factor0
Peak widthd
glass
3% SP-2250 on 1007
120 mesh Supelcoport
Helium
30 mL/min
On-column
250°C
1.0 uL
150°C, 4 min
150°-260°C at 8°/min
Glass jet
280°C
0.7-1.5
10-20 sec
Tolerance
Gas chromatograph
Column
Finnigan 9610
180 cm x 0.2 cm ID
Other3
Other
Other nonpolar
or semipolar
.Hydrogen
Optimum performance
Other ,
Optimum
S 5 ML
Other
Other
Other
Optimum
0.4-3
< 30 sec
.Substitutions permitted if performance criteria are met.
°High enough to elute all PCBs.
Tailing factor is width of front half of peak at 10% height divided by
width of back half of peak at 10% height for single PCB congeners in solu-
.tion FSxxx ng PCB or SIMxxx pg PCB.
Peak width at 10% height for a single PCB congener in FSxxx ng PCB or
SIMxxx pg PCB.
16
-------
Table 8. Operating Parameters for Quadrupole Mass Spectrometer System
Parameter Recommended Tolerance
Mass spectrometer Finnigan 4023 Other3
Data system Incos 2400 Other
Scan range 95-550 Other
Scan time 1 sec Other
Resolution Unit Optimum performance
Ion source temperature 280°C 200°-300°C
Electron energy 70 eV 70 eV
.Substitutions permitted if performance criteria are met.
Greater than five data points over a GC peak is a minimum.
Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no solvent venting is
used.
17
-------
Table\9,^v?Operating:'P,arameiers!'for Magnetic Sector" Mass Spectrometer Sys't'em
Parameter
Mass spectrometer-/
Data system
Scan range
Scan mode
Cycle time
Resolution
Ion source temperature
Electron energy
i ^^Recommended
Finnigan MAT 311A
Incos 2400
98-550
Exponential
1.2 sec
1,000
280°C
70 eV
Tolerance
Other3 "^
Other -G
Other
Other
Otherb
> 500
250-300°
70 eV
^Substitutions permitted if performance criteria are met.
Greater than five data points over a GC peak is a minimum.
Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no solvent venting is
used.
18
-------
Table 10. Limited Mass Scanning (IMS) Ranges For PCBs
Compound
PHP! + 13r 12r u.ri
I»j2ngl> 1 ^ T l/g LgngU 1
C^2H8C12
^12''7^ '3
C12H6C14 + C12D6 C14 + 13C12H6C14
^12^5^5
Ci2H4Cle
^12^3^1?
Ci2H2Cl8
Cj2HCl g
^12^10
C10H7I
Ci8D12
13C12H2C18
13Ci2Cl10
Mass range (m/z)
186-198
220-226
254-260
288-310
322-328
356-362
390-396
426-434
460-468
496-502
254
240
440-446
508-514
19
-------
The dwell times for the mass ranges given in Table 10
will vary with instrument and should be optimized to
allow at least five data points across a chromatographic
peak. Maximum sensitivity will be achieved when utiliz-
ing maximum dwell time.
Instruments having the capability to switch mass ranges
during an analysis require particular attention to the
switching points to assure minimal data loss. Switching
points can be initially determined by analyzing a high
level congener or Aroclor mixture while in the full scan
mode.
7.2.3 Selected ion monitoring data acquisition - Table 11 pre-
sents a suggested set of characteristic ions for SIM.
The SIM program must include at least two ions for each
analyte, generally the primary and secondary ions in
Table 11. The mass spectrometer should be set to at
least unit resolution. The computer acquisition param-
eters should utilize the minimum threshold filtering
necessary so as not to lose pertinent data. Optimum
acquisition parameters will vary depending on the con-
dition of the mass spectrometer and should be checked
daily.
The dwell times for the masses given in Table 11 will
vary with instrument and should be optimized to allow at
least five data points across a chromatographic peak.
Maximum sensitivity will be achieved when utilizing max-
imum dwell time.
Instruments having the capability to switch mass ranges
during an analysis require particular attention to the
switching points to assure minimal data loss. Switching
points can be initially determined by analyzing a highly
concentrated congener or Aroclor mixture while in the
full scan mode.
7.3 The PCB response factors (RF ) must be determined in triplicate
or other replication, as discussed below, using Equation 7-1 for
the analyte homologs.
RF _
RF -
_ _
p - A,, x Mp '
where RF = response factor of a given PCB congener
A = area of the characteristic ion for the PCB congener
P peak
M = mass of PCB congener in sample (micrograms)
20
-------
Table 11. Characteristic SIM Ions for PCBs
Homolog
Cj2ngC I
^12^8^2
^12^7^13
C12H6C14
C^HsCls
^12"4^^6
Ci2HsCl7
^12^2^8
Ci2nCl 9
C12C11o
CiflHyl
Ci2D6Cl4
^18^12
13C612C6H9C1
13C12H6C14
13C12H2C18
13C12Clio
Primary
188 (100)
222 (100)
256 (100)
292 (100)
326 (100)
360 (100)
394 (100)
430 (100)
464 (100)
498 (100)
254 (100)
298 (100)
240 (100)
194 (100)
304 (100)
442 (100)
510 (100)
Ion (relative intensity^
Secondary
190 (33)
224 (66)
258 (99)
290 (76)
328 (66)
362 (82)
396 (98)
432 (66)
466 (76)
500 (87)
-
300 (49)
. -
196 (33)
306 (49)
444 (65)
512 (87)
Tertiary
-
226 (11)
260 (33)
294 (49)
324 (61)
364 (36)
398 (54)
428 (87)
462 (76)
496 (68)
-
296 (76)
-
-
302 (76)
440 (87)
508 (68)
21
-------
A. = area of the characteristic ion for the internal
15 standard peak (d6~3,3',4,4'-tetrachlorobiphenyl
1-iodonaphthalene or other compound)
M. = mass of internal standard in sample (micrograms)
Using the same conditions as for RF , the surrogate response fac-
tors (RF ) must be determined using^Equation 7-2.
A X M.
RFs = irrinr E«- 7'2
I 3 O
where A = area of the characteristic ion for the surrogate peak
M = mass of surrogate in sample (micrograms)
Other terms are the same as defined in Equation 7-1.
If specific congeners are known to be present and if standards
are available, selected RF values may be employed. For general
samples, solutions in Tables 3 and 4 or a mixture with a similar
level of internal standard added, may be used as the response
factor solution. The PCB-surrogate pairs to be used in the RF
calculation are listed in Table 12.
Generally, only the primary ions of both the analyte and surrogate
are used to determine the RF values. If alternate ions are to be
used in the quantisation, the RF must be determined using that
characteristic ion.
The RF value must be determined in a manner to assure ± 20% pre-
cision^ For instruments with good day-to-day precision, a running
mean (RF) based on seven values may be appropriate. A new value
is added each day and the oldest dropped from the mean. Other
options include, but are not limited to, triplicate determinations
of a single concentration spaced throughout a day or determination
of the RF at three different levels to establish a working curve.
If samples are being analyzed on successive days, a single RF
determination which is within ± 20% of the initial day's tripli-
cate determination may be used.
If replicate RF values differ by greater than ± 10% RSD, the sys-
tem performance should be monitored closely. If the RSD is greater
than ± 20%, the data set must be considered invalid and the RF re-
determined before further analyses are done. The analyst is re-
sponsible for maintaining records of the RF precision.
22
-------
Table 12. Pairings of Analyte and Calibration Compounds
Analyte
Congener
no. Compound
1~3 C H Cl
A-m run
^ 1282
1 C-QQ P U PI
J.D 03 I»i2n7l»l3
40-81 C12H6C14
82-127 C12H5C15
128-169 C12H4C16
170-193 C12H3C17
iqA-onci P H PI
X3H £U3 12 2 8
206-208 C12HC19
oriq P n
c.\jj 1210
Calibration standard
Congener
no.
1
7
30
50
97
143
183
202
207
209
Compound
2
2,4
2,4,6
2, 2', 4, 6
2, 2' ,3' ,4,5
2, 2' ,3,4, 5, 6'
2, 2', 3', 4,4', 5', 6
2,2',3,3')5,5',6)6'
2, 2', 3, 3', 4, 4' ,5,6,6'
C^Clio
aBallschmiter numbering system, see Table 1.
23
-------
7.4 If the GC/EIMS system has not been demonstrated to yield a linear
response or if the analyte concentrations are more than two orders
of magnitude different from those in the RF solution, a calibration
curve must be prepared. If the analyte and RF solution concentra-
tions differ by more than one order of magnitude, a calibration
curve should be prepared. A calibration curve should be estab-
lished with triplicate determinations at three or more concentra-
tions bracketing the analyte levels.
7.5 The relative retention time (RRT) windows for the 10 homologs and
surrogates must be determined unless the entire chromatogram is
scanned for each homolog. If all congeners are not available, a
mixture of available congeners or an Aroclor mixture (e.g., 1016/
1254/1260) may be used to estimate the windows. The windows must
be set wider than observed if all isomers are not determined.
Typical RRT windows for one column are listed in Table 13. The
windows may differ substantially if other GC parameters are used.
8.0 Sample Collection, Handling, and Preservation
8.1 Amber glass sample containers should have Teflon-lined screw caps.
With noncorrosive samples, methylene chloride-washed aluminum foil
liners may be substituted. The volume and configuration are deter-
mined by the amount of sample to be collected and its physical
properties. For dry powders, other containers such as wide-mouthed
ointment jars or heavy-walled polyethylene bags may be appropriate.
8.2 Sample bottle preparation
8.2.1 All sample containers and caps should be washed in deter-
gent solution, rinsed with tap water, and then with dis-
tilled water. The bottles and caps are allowed to drain
dry in a contaminant-free area. Then the caps are rinsed
with pesticide grade hexane and allowed to air dry.
8.2.2 Sample bottles are heated to 400°C for 15 to 20 min or
rinsed with pesticide grade acetone or hexane and allowed
to air dry.
8.2.3 The clean bottles are stored inverted and sealed until
use.
8.3 Sample collection - Sample collection must be designed to meet
sampling objectives.
8.3.1 The primary consideration in sample collection is that
the sample collected be representative of the whole.
Therefore, sampling plans or protocols for each individ-
ual producer's situation will have to be developed. The
recommendations presented here describe general situa-
tions. The number of replicates and sampling frequency
also must be planned prior to sampling.
24
-------
Table 13. Relative Retention Time (RRT) Ranges of PCB Homologs
Versus d6-3,3',4,4'-Tetrachlorobiphenyl
PCB
homolog
Monochloro
Dichloro
Trichloro
Tetrachloro
Pentachloro
Hexachloro
Heptachloro
Octachloro
Nonachloro
Decachloro
No. of
isomers
measured
3
10
9
16
12
13
4
6
3
1
Observed range
of RRTsa
0.40-0.50
0.52-0.69
0.62-0.79
0.72-1.01
0.82-1.08
0.93-1.20
1.09-1.30
1.19-1.36
1.31-1.42
1.44-1.45
Congener
no.
1
7
30
50
97
143
183
202
207
209
Observed
RRTa
0.43
0.58
0.65
0.75
0.98
1.05
1.15
1.19
1.33
1.44
Projected
range pf
RRTsD
0.35-0.55
0.45-0.80
0.55-1.00
0.55-1.05
0.80-1.10
0.90-1.25
1.05-1.35
1.10-1.50
1.25-1.50
1.35-1.50
The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
measured versus 3,3',4,4'-tetrach1orobiphenyl-d6 (internal standard) using
a 15-m J&W DB-5 fused silica column with a temperature program of 110°C
for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an on-
column injector. A Finnigan 4023 Incos quadrupole mass spectrometer oper-
ating with a scan range of 95-550 daltons was used to detect each PCB
congener.
The projected relative retention windows account for overlap of eluting
homologs and take into consideration differences in operating systems and
lack of all possible 209 PCB congeners.
25
-------
8.3.2 Discrete product units - If the product is small enough
that one or more discrete units would be used as the
analytical sample, a statistically random sampling ap-
proach is recommended.
8.3.3 Liquids or free-flowing solids - If possible, the source
is mixed thoroughly before collecting the sample. If
mixing is impractical, the sample should be collected
from a representative area of the source. If the liquid
is flowing through an enclosed system, sampling through
a valve should be randomly timed.
8.3.4 Solids - Larger bulk solids which must be subsampled to
get a reasonably sized analytical sample must be treated
on a case-by-case basis. A representative sample should
be obtained by designing a sampling location selection
scheme such that all parts of the whole have a finite,
known probability of inclusion. Based on such a scheme,
the PCB content of the sample can be used to extrapolate
to the content of the whole.
8.4 Sample preservation - Product samples should be stored as the bulk
or packaged product inventory would be stored, or in a cool, dry,
dark area. Intermediates, process samples, or other non-product
specimens should be stored at 4°C. If there is a possibility of
microbial degradation, addition of H2S04 during collection to a
pH < 2 is recommended. A test strip is used to monitor pH. Stor-
age times in excess of 4 weeks are not recommended.
If residual chlorine is present in the sample, it should be
quenched with sodium thiosulfate. EPA Methods 330.4 and 330.5
may be used to measure the residual chlorine.6 Field test kits
are available for this purpose.
9.0 Sample Preparation
Since a wide variety of matrices may be subjected to analysis by this
method, the extraction/cleanup procedure cannot be specified. This
section describes general guidelines for subsampling, addition of 13C
surrogates, dilution, extraction, cleanup, extract concentration, and
other sample preparation procedures.
9.1 Sample homogenization and subsampling - The sample is homogenized
by shaking, blending, shredding, crushing, or other appropriate
mechanical technique. A representative subsample of 100 g or other
known mass is then taken. The sample size is dependent upon the
anticipated PCB levels and difficulty of the subsequent extraction/
cleanup steps.
26
-------
Note: The precision of the mass determination at this step will
be reflected in the overall method precision. Therefore, an an-
alytical balance must be used to assure that the weight is accu-
rate to ± 1% or better.
9.2 Surrogate addition - An appropriate volume of surrogate solution
SSxxx must be pipetted into the sample. The final concentration
of the surrogates must be in the working range of the calibration
and well above the matrix background. The surrogates are thor-
oughly incorporated by further mechanical agitation. For nonvis-
cous liquids, shaking for 30 sec should be sufficient. For vis-
cous liquids or free-flowing solids, 10-min tumbling is recommended.
In cases where inadequate incorporation may be expected, such as
solids, overnight equilibration with agitation is recommended.
Note: The volume measurement of the spiking solution is critical
to the overall method precision. The analyst must exercise cau-
tion that the volume is known to ± 1% or better. Where necessary,
calibration of the pi pet is recommended.
9.3 Sample preparation (extraction/cleanup) - After addition of the
surrogates, the sample is further treated at the discretion of
the analyst, provided that the GC/EIMS response of the four sur-
rogates are sufficient for reliable quantitation. The literature
pertaining to these techniques has been reviewed.5 Several pos-
sible techniques are presented below for guidance only. The ap-
plicability of any of these techniques to a specific sample ma-
trix must be determined by the precision and accuracy of the 13C
PCB surrogate recoveries, as discussed in Section 15.4.1.
9.3.1 Extraction
9.3.1.1 Dilution - In some cases, where the PCB concen-
tration is high, a simple volumetric dilution
with an appropriate solvent may be sufficient
sample preparation.
9.3.1.2 Direct injection - If sample viscosity permits,
direct injection with no dilution is permissible.
9.3.1.3 Kuderna-Danish concentration - If the sample is
a solvent with low PCB concentration, reduction
of sample volume may be sufficient preparation.
In this case, the sample is placed in the
Kuderna-Danish apparatus and concentrated over
steam to an appropriate volume. (Alternatives
to Kuderna-Danish, such as rotary evaporation
techniques may also be used.)
9.3.1.4 Evaporative concentration using nitrogen - For
smaller volume (5-50 mL) solvent samples concen-
tration may be achieved by blowing sample with
27
-------
a gentle stream of pre-purified nitrogen. Con-
centrate sample to final volume, rinse sides of
vial with small aliquots of solvent, and again
concentrate to final volume. In some cases,
where background is a problem, concentration to
dryness will eliminate interference by low-boil-
ing compounds, e.g., chloroform, without remov-
ing the more volatile PCBs. After blowing to
dryness, reconstitute with solvent and sonicate
for 5 min. Care must be taken to avoid evapora-
tive loss of the more volatile PCBs. Do not
heat a dry sample for extended periods.
9.3.1.5 Liquid-liquid extraction - If the matrix is
aqueous (or another solvent in which PCBs are
only slightly soluble), a liquid-liquid parti-
tion may be effective. The solvent, number of
extractions, solvent-to-sample ratio, and other
parameters are chosen at the analyst's discretion.
9.3.1.6 Sorbent column extraction - PCBs may be isolated
from free-flowing liquids onto sorbent columns.
The selection of sorbent (XAD, Porapak, carbon-
polyurethane foam, etc.) will depend on the na-
ture of the matrix. The available methods have
been reviewed.5
9.3.1.7 Thermal desorption - If the matrix is nonvol-
atile, thermal desorption of the PCBs onto a
sorbent column, filter, or cold trap may be an
effective extraction/cleanup method.
9.3.1.8 Matrix destruction - Some matrices may be easily
degraded to a volatile or extractable compound.
Once the matrix is degraded, the PCBs can be
isolated, either by evaporation of the matrix
or partitions of the PCBs into a nonpolar sol-
vent and the matrix into a polar solvent (e.g.,
water). Examples include: (1) esters which
may be saponified with base to the acid anion
and then extracted with water, and (2) acid
chlorides which may be hydrolyzed to the acid
anion with water or base and then extracted
with water.
9.3.2 Cleanup - Several tested cleanup techniques are described
below. All but the base cleanup (9.3.2.8) were previously
validated for PCBs in transformer fluids.7 Depending
upon the complexity of the sample, one or more of the
techniques may be required to fractionate the PCBs from
interferences. For most cleanups a concentrated (1-5 ml)
extract should be used.
28
-------
9.3.2.1 Acid cleanup
9.3.2.1.1 Place 5 mL of concentrated sulfuric
acid into a 40-mL narrow-mouth screw-
cap bottle. Add the sample extract.
Seal the bottle with a Teflon-lined
screw cap and shake for 1 min.
9.3.2.1.2 Allow the phases to separate, transfer
the sample (upper phase) with three
rinses of 1-2 mL solvent to a clean
container.
9.3.2.1.3 Back-extract sample extract with 5-
10 drops of distilled water. Pass
through a short column of anhydrous
sodium sulfate and concentrate to an
appropriate volume.
9.3.2.1.4 Analyze as described in Section 10.0.
9.3.2.1.5 If the sample is highly contaminated,
a second or third acid cleanup may
be employed.
9.3.2.2 Florisil column cleanup
9.3.2.2.1 Variations among batches of Florisil
(PR grade or equivalent) may affect
the elution volume of the various
PCBs. For this reason, the volume
of solvent required to completely
elute all PCBs must be verified by
the analyst. The weight of Florisil
can then be adjusted accordingly.
9.3.2.2.2 Place a 20-g charge of Florisil,
activated overnight at 130°C, into a
Chromaflex column. Settle the Flor-
isil by tapping the column. Add about
1 cm of anhydrous sodium sulfate to
the top of the Florisil. Pre-elute
the column with 70-80 ml of hexane.
Just before the exposure of the sodium
sulfate layer to air, stop the flow.
Discard the eluate.
9.3.2.2.3 Add the sample extract to the column.
9.3.2.2.4 Carefully wash down the inner wall
of the column with 5 ml of hexane.
29
-------
9.3.2.2.5 Add 200 ml of 6% ethyl ether/hexane
and set the flow to about 5 mL/min.
9.3.2.2.6 Collect 200 mL of eluate in a Kuderna-
Danish flask. All the PCBs should be
in this fraction. Concentrate to an
appropriate volume.
9.3.2.2.7 Analyze the sample as described in
Section 10.0.
9.3.2.3 Alumina column cleanup
9.3.2.3.1 Adjust the activity of the alumina
(Fisher A540 or equivalent) by heat-
ing to 200°C for at least 2 hr. When
cool, add 3% water (wt:wt) and mix
until uniform. Allow the deactivated
alumina to equilibrate at least 1/2 hr
before use. Store in a tightly sealed
bottle.
9.3.2.3.2 Variations between batches of alumina
may affect the elution volume of the
various PCBs. For this reason, the
volume of solvent required to com-
pletely elute all of the PCBs must
be verified by the analyst. The
weight of alumina can then be ad-
justed accordingly.
9.3.2.3.3 Place a 50-g charge of alumina into
a Chromaflex column. Settle the
alumina by tapping. Add about 1 cm
of anhydrous sodium sulfate. Pre-
elute the column with 70-80 mL of
hexane. Just before exposure of the
sodium sulfate layer to air, stop
the flow. Discard the eluate.
9.3.2.3.4 Add the sample extract to the column.
9.3.2.3.5 Carefully wash down the inner wall
of the column with 5 mL of hexane.
9.3.2.3.6 Add 295 ml of hexane to the column.
9.3.2.3.7 Discard the first 50 ml.
9.3.2.3.8 Collect 250 mL of the hexane in a
Kuderna-Danish flask. All of the
PCBs should be in this fraction.
Concentrate to an appropriate volume.
30
-------
9.3.2.3.9 Analyze the sample as described in
Section 10.0.
9.3.2.4 Silica gel column cleanup
9.3.2.4.1 Activate silica gel (Davison Grade
950 or equivalent) at 135°C overnight.
9.3.2.4.2 Variations between batches of silica
gel may affect the elution volume of
the various PCBs. For this reason,
the volume of solvent required to
completely elute all of the PCBs must
be verified by the analyst. The
weight of silica gel can then be ad-
justed accordingly.
9.3.2.4.3 Place a 25-g charge of activated
silica gel into a Chromaflex column.
Settle the silica gel by tapping the
column. Add about 1 cm of anhydrous
sodium sulfate to the top of the
silica gel.
9.3.2.4.4 Pre-elute the column with 70-80 mL
of hexane. Discard the eluate. Just
before exposing the sodium sulfate
layer to air, stop the flow.
9.3.2.4.5 Add the sample extract to the column.
9.3.2.4.6 Wash down the inner wall of the column
with 5 mL of hexane.
9.3.2.4.7 Elute the PCBs with 195 mL of 10%
diethyl ether in hexane (v:v).
9.3.2.4.8 Collect 200 mL of the eluate in a
Kuderna-Danish flask. All of the
PCBs should be in this fraction.
Concentrate to an appropriate volume.
9.3.2.4.9 Analyze the sample as described in
Section 10.0.
9.3.2.5 Gel permeation cleanup
9.3.2.5.1 Set up and calibrate the gel permea-
tion chromatograph with an SX-3 column
according to the Autoprep instruction
manual. Use 15% methylene chloride
in cyclohexane (v:v) as the mobile
phase.
31
-------
9.3.2.5.2 Inject 5.0 mL of the sample extract
into the instrument. Collect the
fraction containing the PCBs (see
Autoprep operator's manual) in a
Kuderna-Danish flask equipped with
a 10-mL ampul.
9.3.2.5.3 Concentrate the PCB fraction to an
appropriate volume.
9.3.2.5.4 Analyze the sample as described in
Section 10.0.
9.3.2.6 Acetonitrile partition
9.3.2.6.1 Place the sample extract into a 125-mL
separatory funnel with enough hexane
to bring the final volume to 15 mL.
Extract the sample four times by shak-
ing vigorously for 1 min with 30-mL
portions of hexane-saturated acetoni-
trile. Retain hexane layer for com-
bination with other hexane extracts
in 9.3.2.6.3.
9.3.2.6.2 Combine and transfer the acetonitrile
phases to a 1-L separatory funnel
and add 650 mL of distilled water and
40 mL of saturated sodium chloride
solution. Mix thoroughly for about
30 sec. Extract with two 100-mL por-
tions of hexane by vigorously shaking
about 15 sec.
9.3.2.6.3 Combine the hexane extracts in a
1-L separatory funnel and wash with
two 100-mL portions of distilled
water. Discard the water layer and
pour the hexane layer through an 8-
to 10-cm anhydrous sodium sulfate
column into a 500-mL Kuderna-Danish
flask equipped with a 10-mL ampul.
Rinse the separatory funnel and
column with three 10-mL portions of
hexane.
9.3.2.6.4 Concentrate the extracts to an ap-
propriate volume.
9.3.2.6.5 Analyze as described in Section 10.0.
32
-------
9.3.2.7 Florisil slurry cleanup
9.3.2.7.1 Place the sample extract into a 20-mL
narrow-mouth screw-cap container.
Add 0.25 g of Florisil (PR grade or
equivalent). Seal with a Teflon-lined
screw cap and shake for 1 min.
9.3.2.7.2 Allow the Florisil to settle; then
decant the treated solution into a
second container with rinsing. Con-
centrate the sample to an appropriate
volume. Analyze as described in Sec-
tion 10.0.
9.3.2.8 Base cleanup8
9.3.2.8.1 Quantitatively transfer the concen-
trated extract to a 125-mL extraction
flask with the aid of several small
portions of solvent.
9.3.2.8.2 Evaporate the extract just to dryness
with a gentle stream of dry filtered
nitrogen, and add 25 ml of 2.5% alco-
holic KOH.
9.3.2.8.3 Add a boiling chip, put a water con-
denser in place, and allow the solu-
tion to reflux on a hot plate for 45
min.
9.3.2.8.4 After cooling, transfer the solution
to a 250-mL separatory funnel with
25 ml of distilled water.
9.3.2.8.5 Rinse the extraction flask with 25
mL of hexane and add it to the
separatory funnel.
9.3.2.8.6 Stopper the separatory funnel and
shake vigorously for at least 1 min.
Allow the layers to separate, and
transfer the lower aqueous phase to
a second separatory funnel.
9.3.2.8.7 Extract the saponification solution
with a second 25-mL portion of hexane.
After the layers have separated, add
the first hexane extract to the sec-
ond separatory funnel and transfer
the aqueous alcohol layer to the
original separatory funnel.
33
-------
9.3.2.8.8 Repeat the extraction with a third
25-mL portion of hexane. Discard
the saponification solution, and com-
bine the hexane extracts.
9.3.2.8.9 Concentrate the hexane layer to an
appropriate volume, and analyze as
described in Section 10.0.
9.3.2.9 High performance liquid chromatographic cleanup
9.3.2.9.1 Quantitatively transfer the concen-
trated extract into the sample loop
or the barrel of a syringe. Rinse
the vial with several small portions
of solvent. It may be necessary to
inject several fractions.
9.3.2.9.2 Inject the extract and washes onto
the amine column (Waters pBondapak
3.9 x 300 mm or equivalent) and elute
the PCBs with 1.0 mL/min hexane.
The UV at 254 nm or lower should be
monitored.
9.3.2.9.3 Collect the eluent from 3 min to 9.5
min as it exits UV cell. The elution
time should be verified using PCB
standards covering a range from mono-
chlorobiphenyls to decachlorobiphenyls.
9.3.2.9.4 After collection, wash the column by
eluting with methylene chloride until
the absorbance attains a stable mini-
mum. Return the system to hexane.
9.3.2.9.5 Concentrate the hexane eluate under
a gentle stream of purified nitrogen
to an appropriate volume and analyze
as described in Section 10.0.
9.4 Optional screening for interferences using GC/FID
Note: Since many sample matrices are one of a kind or infrequently
encountered by the analyst, the effectiveness of the extraction
and cleanup for a matrix may be unknown. A simple screen to as-
sess whether the interferences have been reduced to a tolerable
level can both save GC/MS time and prevent contamination of the
GC/MS instrument with very dirty samples. This screen should
not be used to determine PCB levels under this analytical method.
34
-------
9.4.1 Using a GC system as described in Section 5.5.3, analyze
for background interferences.
9.4.2 A 2 m x 2 mm glass column packed with 3% SP-2250 on 1007
120 Supelcoport or equivalent is suggested. A flow rate
of 40 mL/min 95% air/5% methane or nitrogen is recommended.
The air and hydrogen flow rates should be sufficient to
keep the flame lit and to burn efficiently, e.g., 300
mL/min air and 30 mL/min H2.
9.4.3 The recommended temperature program is from 50°C to 250°C
at 20°C/min with an initial hold of 3 min and a final
hold of 10 min. The injector temperature should be 200°C
and the detector 300°C.
9.4.4 Set instrumental sensitivity comparable to the antici-
pated mass spectral sensitivity. It is advisable to
establish criteria for rejection of samples at a given
attenuation such as (1) any off-scale peaks in PCS elu-
tion window, (2) a baseline rise of over 40% full-scale,
and (3) other criteria which are indicative of "problem"
samples.
9.4.5 If the FID screen suggests that the sample is not amen-
able to analysis by GC/EIMS, the analyst may either (1)
cycle the sample through the same cleanup again if it
appears that the cleanup technique was overloaded by the
matrix the first time, (2) submit the extract to another
cleanup technique which may remove more interferences,
or (3) analyze a new aliquot of sample by another extrac-
tion or cleanup technique.
10.0 Gas Chromatographic/Electron Impact Mass Spectrometric Determination
10.1 Internal standard addition - An appropriate volume of the internal
standard solution is pipetted into the sample. The final concen-
tration of the internal standard must be in the working range of
the calibration and well above the matrix background. The inter-
nal standard is thoroughly incorporated by mechanical agitation.
Note: The volumetric measurement of the internal standard solu-
tion is critical to the overall method precision. The analyst
must exercise caution that the volume is known to be ± 1% or bet-
ter. Where necessary, calibration of the pipet is recommended.
10.2 Tables 2, and 6 through 11 summarize the recommended operating
conditions for analysis. Figure 1 presents an example of a
chromatogram.
35
-------
109.8-1
RIC
oo
13
C]2D6C14
C12H6C14
13C612C6H9C1
C1QH7I
50
30
97
183
143
88960.
202
C18D12
13
C12H2C18
209
13
207
2C110
13:20
1000
16:40
1200
20:00
1400
23:20
1600
26:40
1800
30:00
SCAN
TIME
Figure 1. Reconstructed ion chromatogram of calibration solution FS100 ng PCB obtained in the full scan mode. The
concentration of the 10 PCB calibration congeners, the four 13C-labeled PCB recovery surrogates, and the three inter-
nal standards are in Table 3. See Table 1 for PCB numbering system, Table 6 for capillary GC parameters, and Table 8
for mass spectrometer operating parameters.
-------
The analyst may choose to operate the mass spectrometer at any
appropriate sensitivity, using either full scan, limited mass
scanning or selected ion monitoring acquisition. The sensitivity
selected will depend on anticipated PCB levels and the instrumen-
tal LOQ needed to meet the required method LOQ. In general, the
more concentrated the PCBs, the greater the precision, accuracy,
and qualitative data confidence. Thus, if possible, the amount
of sample and the concentration factor should be scaled so that
full scan acquisition may be utilized.
10.3 While the highest available chromatographic resolution is not a
necessary objective of this method, good chromatographic per-
formance is recommended. With the high resolution of HRGC, the
probability that the chromatographic peaks consist of single com-
pounds is higher than with PGC. Thus, qualitative and quantita-
tive data reduction should be more reliable.
10.4 After performance of the system has been certified for the day
and all instrument conditions set according to Tables 2, and 6
through 11, inject an aliquot of the sample onto the GC column.
If the response for any ion, including surrogates and internal
standards, exceeds the working range of the system, dilute the
sample and reanalyze. If the responses of surrogates, internal
standards, or analytes are below the working range, recheck the
system performance. If necessary, concentrate the sample and re-
analyze.
10.5 Record all data on a digital storage device (magnetic disk, tape,
etc.) for qualitative and quantitative data reduction as discussed
below.
10.6 The instrumental performance must be monitored from run-to-run.
The areas of internal standards must be consistent (e.g., ± 20%).
If a low area is encountered, the injection may be suspect.
The resolution and peak shape of the internal standards, surro-
gates, and other peaks should be monitored during or immediately
after data acquisition. Poor chromatography may indicate a bad
injection, matrix interferences, or column degradation.
10.7 If a "dirty" sample is encountered, the analyst must employ appro-
priate measures to demonstrate that there is no memory or carry-
ons to subsequent samples. To assess the system cleanliness, a
standard, blank sample, or solvent blank may be run.
If the system is contaminated, remedial efforts may include (1)
changing or cleaning the syringe, (2) cleaning the injector, (3)
baking out the column at its maximum temperature, (4) changing
to a new column, or (5) cleaning the ion source.
37
-------
11.0 Qualitative Identification
11.1 Full scan data
11.1.1 The peak must elute within the retention time windows
set for that homolog (as described in Section 7.5).
11.1.2 The unknown spectrum should be compared to that of an
authentic PCB. The intensity of the three largest ions
in the molecular cluster (two largest for monochlorobi-
phenyls) must match the ratio observed for a standard
within ± 20%. Fragment clusters with proper intensity
ratios should also be present. System noise at low con-
centration or interferences may skew the ion ratio be-
yond the ± 20% criteria. If the analyst's best judge-
ment is that a peak, which does not meet the qualitative
criteria, is a PCB, the peak may be included in the cal-
culation, with a footnote explaining the data and the
reason for relaxing the criteria.
11.1.3 Alternatively, a spectral search may be used to auto-
matically reduce the data. The criteria for acceptable
identification include a high index of similarity.
11.2 Selected ion monitoring (SIM) or limited mass scan (LMS) data -
The identification of a compound as a given PCB homolog requires
that two criteria be met:
11.2.1 (1) The peak must elute within the retention time window
set for that homolog (Section 7.5); and (2) the ratio of
two ions obtained by LMS (Table 10) or by SIM (Table 11)
must match the ratio observed for a standard within ± 20%.
The analyst should search the higher mass windows, in
particular M+70, to prevent misidentification of a PCB
fragment ion cluster as the parent. System noise at low
concentration or interferences may skew the ion ratio
beyond the ± 20% criteria. If the analyst's best judge-
ment is that a peak, which does not meet the qualitative
criteria, is a PCB, the peak may be included in the cal-
culation, with a footnote explaining the data and the
reason for relaxing the criteria.
11.2.2 If one or the other of these criteria is not met, inter-
ferences may have affected the results, and a reanalysis
using full scan EIMS conditions is recommended.
11.3 Disputes in interpretation - Where there is reasonable doubt as
to the identity of a peak as a PCB, the analyst must either iden-
tify the peak as a PCB or proceed to a confirmational analysis
(see Section 13.0).
38
-------
12.0 Quantitative Data Reduction
12.1 After a chromatographic peak has been identified as a PCB, the
compound is quantitated based on the integrated abundance of
either the EICP or the SIM data for the primary characteristic
ion in Tables 10 and 11. If interferences are observed for the
primary ion, use the secondary and then tertiary ion for quanti-
tation. If interferences in the parent cluster prevent quantita-
tion, an ion from a fragment cluster (e.g., M-70) may be used.
Whichever ion is used, the RF must be determined using that ion.
The same criteria should be applied to the surrogate compounds
(Table 14).
Note: With the higher homologs, the mass defect from unity is
significant. For instance, the mass of the most intense peak for
decachlorobiphenyl is 497.6830. Areas, EICPs, etc. must be based
on the true mass, not the nominal mass, or erroneous results may
be obtained. In addition, the tuning of some quadrupoles may be
less stable at high masses. The data quality must be monitored
especially carefully for the higher homologs.
12.2 Using the appropriate analyte-internal standard pair and response
factor (RF ) as determined in Section 7.3, calculate the concen-
tration oreach peak using Equation 12-1.
A -. M.
Concentration ((jg/g) = -f- ' np— * M— Eq. 12-1
is p e
where A = area of the characteristic ion for the analyte PCB
P peak
A. = area of the characteristic ion for the internal
standard peak
RF = response factor of a given PCB congener
M. = mass of internal standard added to sample extract
(micrograms)
M = mass of sample extracted (grams)
12.3 If a peak appears to contain non-PCB interferences, which cannot
be circumvented by a secondary or tertiary ion, either:
12.3.1 Reanalyze the sample on a different column which sepa-
rates the PCB and interferents;
12.3.2 Perform additional chemical cleanup (Section 9) and then
reanalyze the sample; or
12.3.3 Quantitate the entire peak as PCB.
39
-------
Table 14. Characteristic Ions for Internal Standards
and 13C-Labeled PCB Surrogates
Ion (relative intensity)
Compound
chlorobiphenyl
1-Iodonaphthalene
d12-Chrysene
13C612C6H9C1
13C12H6C14
13C12H2C18
13C12C110
Abbreviation
d6-C!4
INAP
DCRY
"c-cu
13C-C14
13c-ci8
13c-ci10
Primary
298 (100)
254
240
194 (100)
304 (100)
442 (100)
510 (100)
Secondary
300 (49)
127
-
196 (33)
306 (49)
444 (65)
512 (87)
Tertiary
296 (78)
-
-
-
302 (78)
440 (89)
514 (50)
40
-------
12.4 Calculate the recovery of the four 13C surrogates using the ap-
propriate surrogate-internal standard pair and response factor
(RF. ) as determined in Section 7.4 using Equation 12-2.
I o
Recovery (%) = ^- • ^- • ^ • 100 Eq. 12-2
where A = area of the characteristic ion for the surrogate peak
o
RF = response factor for the surrogate compound with respect
to the internal standard (Equation 7-2)
M = mass of surrogate added to original sample (micrograms)
Other terms are the same as defined in Equation 12.1.
12.5 Sum all of the peaks for each homolog, and then sum those to yield
the total PCB concentration in the sample. Report all numbers in
|jg/g. The worksheet in Table 15 and reporting form in Table 16
may be used. The concentrations and percent recovery of the four
surrogates are to be reported.
If an alternate reporting format (e.g., concentration per peak)
is desired, a different report form may be used. For example,
if the PCB analysis is being conducted to satisfy regulatory re-
quirements for by-product PCBs, results may need be reported on
a per resolvable chromatographic peak basis. One rule1 states
that PCBs in products or wastes must be below the practical limit
of quantitation, defined as "2 micrograms per gram (2 ppm) per
resolvable chromatographic peak." For regulatory purposes, only
peaks greater than the 2 ug/g cutoff may need to be reported. A
second rule2 requires reporting total PCBs, with a regulatory
cutoff at an annual average of 25 ppm with a 50 ppm maximum.
12.6 Round off all numbers reported to two significant figures.
13.0 Confirmation
If there is significant reason to question the qualitative identifica-
tion (Section 11), the analyst may choose to confirm that a peak is not
a PCB. Any technique may be chosen provided that it is validated as
having equivalent or superior selectivity and sensitivity to GC/EIMS.
Some candidate techniques include alternate GC columns (with EIMS detec-
tion), GC/CIMS, GC/NCIMS, high resolution EIMS, and MS/MS techniques.
Each laboratory must validate confirmation techniques to show equivalent
or superior selectivity between PCBs and interferences and sensitivity
(limit of quantitation, LOQ).
41
-------
Table 15. Analysis Worksheet
BY-PRODUCT PCBs IN COMMERCIAL PRODUCTS OR PRODUCT WASTES
Internal Sample No.
Notebook No.
Data File Code
External Sample No.
Date Prepared
Date Analyzed
Amount Extracted
Final Volume
Internal
standards Cone, (ug/9) Quant, ion Ratio Scan no.
d6Cl4
INAP
DCRY
298,300
254
240
100/49
ml
Check std. to
be used for
Area calculations
Surrogate
compounds Cone, (ug/g) Ions
Ratio
Scan no.
Area
% Recovery
-c-cu
13C-C14
13c-ci8
13c-ci10
Analyte
homolog
194,196
304,306
442,444
510,512
Ions
100/33
100/49
100/65
100/87
Analyte
Ratio homolog Ions Ratio
1-C1
2-C1
3-C1
4-C1
5-C1
188,190
222,224
256,258
292,290
326,328
100/33
100/66
100/99
100/76
100/66
6-C1
7-C1
8-C1
9-C1
10-C1
360,362
394,396
430,432
464,466
498,500
100/82
100/98
100/66
100/76
100/87
Homolog Scan no. (s) Area(s) Ratio Total area Ion used RF Cone. (|jg/g)
42
-------
Table 15 (continued)
BY-PRODUCT PCBs IN COMMERCIAL PRODUCTS OR PRODUCT WASTES
Internal Sample No. External Sample No.
Data File Code
Homolog Scan no.(s) Area(s) Ratio Total area Ion used RF Cone. (M9/9)
43
-------
Table 16. Analysis Report
BY-PRODUCT PCBs IN COMMERCIAL PRODUCTS OR PRODUCT WASTES
Internal Sample No.
Notebook No.
Data File Code
Amount Extracted c
Extraction/Cleanup Procedure
External Sample No.
Sample Source
Sample Matrix
Final Volume
Analyte
homolog
1-C1
Concentration
Analyte
homolog
6-C1
Concentration
2-C1
7-C1
3-C1
8-C1
4-C1
9-C1
5-C1
10-C1
Surrogate
compound Concentration (|jg/g)
Total
Surrogate
compound
13C-C18
Concentration (|jg/g)
13C-C1.
13c-ci10
Estimated Method LOQ
Highest concentration per resolvable chromatographic peak
Reported by:
Internal Audit:
EPA Audit:
Name
Name
Name
Signature/Date
Signature/Date
Signature/Date
Organization
Organization
Organization
44
-------
If a peak is confirmed as being a non-PCB, it may be deleted from the
calculation (Section 12). If a peak is confirmed as containing both
PCB and non-PCB components, it must be quantitated according to Section
12.3.
14.0 Quality Assurance
Each participating laboratory must develop a quality assurance plan
(QAP) according to EPA guidelines.9 Additional guidance is also avail-
able.10 The quality assurance plan must be submitted to the Agency (re-
gional QA officer) for approval prior to analysis of samples.
The elements of a QAP include:
• Title Page
• Table of Contents
• Project Description
Project Organization and Responsibility
• QA Objectives for Measurement Data in Terms of Precision, Accuracy,
Completeness, Representativeness, and Comparability
• Sampling Procedures
• Sample Custody
• Calibration Procedures and Frequency
• Analytical Procedures
• Data Reduction, Validation and Reporting
Internal Quality Control Checks
• Performance and System Audits
• Preventive Maintenance
• Specific Routine Procedures Used to Assess Data Precision, Accuracy
and Completeness
• Corrective Action
• Quality Assurance Reports to Management
45
-------
15.0 Quality Control
15.1 Each laboratory that uses this method must operate a formal qual-
ity control (QC) program. The minimum requirements of this pro-
gram consist of an initial and continuing demonstration of lab-
oratory capability by the analysis of check samples. The labora-
tory must maintain performance records to define the quality of
data that are generated.
15.2 Certification and performance checks - Prior to the analysis of
samples, the laboratory must define its routine performance. At
a minimum, this must include demonstration of acceptable response
factor precision with at least three replicate analyses; and anal-
ysis of a blind QC check sample (e.g., the response factor cali-
bration solution at unknown concentration submitted by the QA
officer). Acceptable criteria for the response factor precision
and the accuracy of the QC check sample analysis must be pre-
sented in the QA plan.
Ongoing performance checks should consist of periodic repetition
of the initial demonstration or more elaborate measures. More
elaborate measures may include control charts and analysis of QC
check samples consisting of other congeners or with matrix inter-
ferences.
15.3 Procedural QC - The various steps of the analytical procedure
should have quality control measures. These include but are not
limited to:
15.3.1 GC performance - See Section 7.1 for performance criteria.
15.3.2 MS performance - See Section 7.2 for performance criteria.
15.3.3 Qualitative identification - At least 10% of the PCB
identifications, as well as any questionable results,
should be confirmed by a second mass spectrometrist.
15.3.4 Quantisation - At least 10% of all manual calculations,
including peak area calculations, must be checked. After
changes in computer quantisation routines, the results
should be manually checked.
15.4 Sample QC - Each sample and each sample set must have QC measures
applied to it to establish the data quality for each analysis re-
sult. The recoveries of the surrogates, general spectral data
quality, and consistency of the internal standard area are all
measures of the data quality on individual samples. Within a
sample set, analysis of replicates and standard addition samples
are measures of the precision and accuracy, respectively.
46
-------
15.4.1 The surrogate recoveries must be within an acceptable
range, as defined in the QAP, for each sample. Typically,
recoveries below 50% are unacceptable, and indicate that
the sample(s) must be reanalyzed. Recoveries much greater
than 100% indicate interferences, improper tuning, a
problem with response factors, or an error in the concen-
tration of the surrogate compound. The trend of low re-
coveries can be indicative of the cause of the loss. If
the lower homologs are poorly recovered, but the higher
homologs are recovered quantitatively, volatility or
chemical degradation losses may be suspected. If higher
homologs are selectively lost, or if the losses are ir-
regular, a fractionation cut on a chromatographic clean-
up would be a likely suspect to account for the loss.
15.4.2 The general spectral data quality is indicative of the
overall reliability of the data for a sample. The levels
of the background, intensity ratios within chlorine clus-
ters, etc., must all be evaluated. If the data quality
is marginal, the analyst may footnote results with an
explanation regarding any doubts about the data quality.
If the data are unacceptable (see Section 11.0), either
the GC/MS determination or the entire analysis must be
repeated.
15.4.3 An easy and significant assessment of the data quality
is the consistency of the internal standard areas. If
the internal standard area is consistent, the injection
volume was correct and the system is operating within
general tolerances (i.e., the chromatography column is
transmitting compounds and the spectrometer is detecting
them). If the internal standard area does not meet the
criteria specified in the QAP, e.g., ± 20% of other in-
jections, the data must be reviewed. If the injection
or the GC/MS performance is suspect, the sample should
be reanalyzed, or other corrective action taken.
15.4.4 QC for small sample sets - For small sample sets (1-10
samples), the minimum QC requirements can be a heavy bur-
den. Analysts are encouraged to be efficient and group
similar samples to increase the size of a set. A set is
defined as a group of samples analyzed together by the
same extraction/cleanup technique and determined on the
GC/MS system on the same day or successive days under
the same conditions.
At least one method blank must be run. The blank must
be exposed to the same sources of contamination—solvent,
glassware, etc.--as the samples. If conditions change,
additional blanks must be generated. An example would
be a new lot of solvent, or a change in dishwashing
protocol.
47
-------
At least one sample must be run in replicate. Tripli-
cates are preferable, but duplicates may be acceptable.
The acceptable precision among replicates must be speci-
fied in the QAP.
At least one sample must be analyzed by the standard
addition technique. The analyst may select the most
difficult sample, based on prior knowledge of the sample
set, or a random sample. Two aliquots of the sample are
analyzed, one "as is" and one spiked (surrogate spiking
and equilibration techniques are described in Section
9.2) with Solution "FSxxx ng PCB" or "SIMxxx pg PCB."
If the analyst has no prior knowledge of the sample, the
spiking level should be in the middle of the calibrated
range for the mass spectrometer. If the concentrations
of PCBs are known to be high or low, the amount added
should be adjusted so that the spiking level is 1.5 to
4 times the measured PCB level in the unspiked sample.
The samples should be analyzed together and the quanti-
tative results calculated. The recovery of the spiked
compounds (calculated by difference) must be 70-130%.
If the sample is known to contain specific PCB isomers,
these isomers may be substituted for solution FSxxx ng
PCB or SIMxxx pg PCB.
15.4.5 QC for intermediate sample sets - With intermediate
(approximately 10-100 samples) sample sets, the number
of method blanks, replicates, and standard addition sam-
ples must comprise at least 10% each.
15.4.6 QC for large sample sets - When a large sample analysis
program is being planned, the QA plan may propose spe-
cific QC measures. If none are proposed, the guidelines
for intermediate sets may be followed. One QC measure
which may increase efficiency is the use of control
charts. If, for example, the control charts establish
that there is no blank problem over the long term, the
percent of blanks may be reduced. Any changes in the
procedure (e.g., a new lot of solvent) will still, of
course, require a blank.
15.5 It is recommended that the participating laboratory adopt addi-
tional QC practices for use with this method. The specific prac-
tices that are most productive depend upon the needs of the lab-
oratory and the nature of the samples. Field duplicates or trip-
licates may be analyzed to monitor the precision of the sampling
technique. Whenever possible, the laboratory should perform
analysis of standard reference materials and participate in rele-
vant performance evaluation studies.
48
-------
16.0 Method Performance
The method has been evaluated using both intra- and inter!aboratory
studies.3'11 Both studies must be considered preliminary; only a
limited number of matrices and method options have been investigated.
The preliminary inter!aboratory study involved only four participants
and was conducted before the method was ruggedized, so data from that
study may not be representative of the potential method performance.
Further intra- and interlaboratory validation is anticipated. Prelimi-
nary values for limits of quantisation; intralaboratory recoveries, pre-
cision, and accuracy; and interlaboratory recoveries, precision, and
accuracy are presented in Table 17. The values in the "best" and "worst"
columns represent typical extremes of the measurement, ignoring excep-
tional cases. For example, some mass spectrometers, when running under
optimal conditions, can probably quantitate less than 0.5 ng/uL in the
full scan mode for some matrices and PCB congeners. On the other hand,
the method LOQ will be much greater than the stated value if a high
concentration coeluting interference completely obscures any PCB signal,
even at the percent level.
The performance values in Table 17 were derived using either the SIM or
LMS options of the method and at concentrations at the lower end of the
working range. Working in the middle-to-upper concentration ranges with
full scan data collection, the precision, recovery, and accuracy should
all improve considerably.
The values in Table 17 represent best estimates of the parameters and
will be refined as additional intra- and interlaboratory studies pro-
duce more data. Performance better than or worse than any of these
parameters cannot, of itself, be construed as grounds for acceptable or
unacceptable data quality. Performance criteria should be stipulated
in the QAP.
17.0 Documentation and Records
Each laboratory is responsible for maintaining full records of the
analysis. A detailed documentation plan should be prepared as part of
the QAP. Laboratory notebooks should be used for handwritten records.
GC/ MS data must be archived on magnetic tape, disk, or a similar device.
Hard copy printouts may be kept in addition if desired. QC records
should be maintained separately from sample analysis records.
The documentation must describe completely how the analysis was performed.
Any variances from the protocol must be noted and fully described. Where
the protocol lists options (e.g., sample cleanup), the option used and
specifics (solvent volumes, digestion times, etc.) must be stated.
The remaining samples and extracts should be archived for at least 2
months or until the analysis report is approved, whichever is longer,
and then disposed unless other arrangements are made. The magnetic
tapes of the analysis and hardcopy spectra, quantitation reports, work
sheets, etc., must be archived for at least 3 years.
49
-------
Table 17. Method Performance Parameters
Measurement
Best case
Worst case Typical value
Instrumental LOQ (ng/ul)
Full scan
Limited mass scan
Selected ion monitoring
Sample concentration factor
Matrix interference level (ng/ul)
Sample injection volume (ul)
Method LOQ (ug/g)
b
0.5
0.05
0.01
1,000+1
< 0.01
10
io"5
10
0.5
0.2
1*100
> 1,000
0.1
500
1
0.1
0.05
_a
_a
1
1
Recovery (%)
Precision (%)
Accuracy
b e
Interlaboratory '
Recovery (%)
Precision (%)
Accuracy
90 ± 15C < 50, > 200d 70-130
± 34%C
22-690
± 60'
± 629
.Varies widely from matrix-to-matrix, no "typical value."
Data for preliminary validation at low (< I ug/g) levels using selected ion
monitoring mass spectrometry.
"Methods of Analysis for By-Product PCBs—Preliminary Validation and Interim
Methods," M. D. Erickson, J. S. Stanley, G. Radolovich, K. Turman, K. Bauer,
J. Onstot, D. Rose, and M. Wickham, Interim Report No. 4, Washington D.C.:
Office of Toxic Substances, EPA-500/5-82-006, October 1982, 243 pp. NTIS No.
RB83 127 696.
Values outside the 50-200% range are generally considered unacceptable and
the analysis must be repeated.
Preliminary interlaboratory study involving four participants: "Analytical
Methods for By-Product and Destruction Derived PCBs—Interlaboratory Valida-
tion A," M. D. Erickson, K. M. Bauer, and F. J. Bergman, Draft Interim Report
No. 6, Washington, D.C.: Office of Toxic Substances, U.S. Environmental Pro-
tection Agency, Draft Interim Report No. 6, Task 51, EPA Contract No. 68-01-
5.915, August 1983.
Based on a mean of the 10 homolog values from the analysis of two samples by
four laboratories.
%ased on the deviation of the reported values from the prepared value for 10
homologs x 2 samples x 4 laboratories.
50
-------
REFERENCES
1. USEPA. 1982. 40 CRF 761, Polychlorinated biphenyls (PCBs); manufactur-
ing, processing, distribution in commerce, and use prohibitions; use in
closed and controlled waste manufacturing processes. 47 FR 46980-46996.
2. USEPA. 1984. 40 CFR Part 761, Polychlorinated biphenyls (PCBs); manu-
facturing, processing, distribution commerce and use prohibitions; re-
sponse to individual and class petitions for exemptions. 49 FR 28154-28209.
3. Erickson MD, Stanley JS, Radolovich G, Turman K, Bauer K, Onstot J,
Rose D, Wickham M. 1982. Analytical methods for by-product PCBs--
preliminary validation and interim methods. Washington, DC: Office of
Toxic Substances, U.S. Environmental Protection Agency. EPA-500/5-82-006;
NTIS No. PB83 127 696.
4. Erickson MD, Stanley JS, Radolovich G, Blair RB. 1983. Analytical method:
the analysis of by-product chlorinated biphenyls in commercial products
and product wastes. Revision 1, Prepared by Midwest Research Institute
for Office of Toxic Substances, U.S. Environmental Protection Agency,
Washington, DC, under Subcontract No. A-3044(8149)-271, Work Assignment
No. 17 to Battelle, Washington, DC, August 15, 1983.
5. Erickson MD, Stanley JS. 1982. Methods of analysis for incidentally
generated PCBs—literature review and preliminary recommendations.
Washington, DC: Office of Toxic Substances, U.S. Environmental Pro-
tection Agency. EPA-560/5-82-005; NTIS No. PB83 126573.
6. USEPA. 1979. Methods 330.4 (titrimetric, DPD-FAS) and 330.5 (spec-
trophotometric, DPD) for chlorine, total residual. Methods for Chemical
Analysis of Water and Wastes, U.S. Environmental Protection Agency,
Environmental Monitoring and Support Laboratory, Cincinnati, OH, EPA
600-4/79-020.
7. Bellar TA, Lichtenberg JJ. 1981. The determination of polychlorinated
biphenyls in transformer fluid and waste oils. Prepared for U.S. Environ-
mental Protection Agency, EPA-600/4-81-045.
8. American Society for Testing and Materials. 1980. Standard method for
analysis of environmental materials for polychlorinated biphenyls. In:
Annual book of ASTM standards, 1980. Philadelphia, PA: ANSI/ASTM D
3304 - 77.
9. USEPA. 1980. Guidelines and specifications for preparing quality as-
surance project plans. Office of Monitoring Systems and Quality Assur-
ance, QAMS-005/80.
10. USEPA. 1983. Quality assurance program plan for the Office of Toxic
Substances. Office of Pesticides and Toxic Substances, U.S. Environ-
mental Protection Agency, Washington, D.C.
51
-------
11. Erickson MD, Bauer KM, Bergman FJ. 1983. Analytical methods for by-
product and destruction derived PCBs--interlaboratory validation A.
Draft Interim Report No. 6, Task 51, EPA Contract No. 68-01-5915, Office
of Toxic Substances, U.S. Environmental Protection Agency, Washington,
DC.
52
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-560/5-85-010
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Analytical Method: The Analysis of By-product Chlorin-
ated Biphenyls in Commercial Products and Product
Wastes, Revision 2
5. REPORT DATE
May 1985
6. PERFORMING ORGANIZATION CODE
8201A06
7. AUTHOR(S)
Mitchell D. Erickson
8. PERFORMING ORGANIZATION REPORT NO.
Special Report No. 1
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Midwest Research Institute
425 Volker Boulevard
Kansas City, MO 64110
10. PROGRAM ELEMENT NO.
Work Assignment No. 6
11. CONTRACT/GRANT NO.
68-02-3938
12. SPONSORING AGENCY NAME AND ADDRESS
Field Studies Branch (TS-798), Office of Toxic Substances
U.S. Environmental Protection Agency
401 M Street, SW
Washington, DC 20460 '
13. TYPE OF REPORT AND PERIOD COVERED
Special (September 84 - May 85]
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
The EPA Work Assignment Manager is Daniel T. Heggem (202) 382-3990.
The EPA Project Officer is Joseph J. Breen (202) 382-3569.
16. ABSTRACT
This is a gas chromatographic/electron impact mass spectrometric (GC/EIMS) method ap-
plicable to the determination of chlorinated biphenyls (PCBs) in commercial products
and product wastes. The PCBs present may originate either as synthetic by-products or
as contaminants derived from commercial PCB products (e.g., Aroclors). The PCBs may be
present as single isomers or complex mixtures and may include all 209 congeners from
monochlorobiphenyl through decachlorobiphenyl.
A variety of general and specific sample preparation options are presented in this
method. This method takes a different approach from those which rely on Aroclor mix-
tures for calibration and quantisation. In this method PCBs are detected and quanti-
tated by homolog group. The results can be summed to give a total PCB value comparable
to results generated by other methods or they may be presented as 10 individual homolog
values. This homolog distribution can provide additional quantitative information on
the composition and source of the PCBs.
The method performance is assessed for each sample. A set of four 13C-labeled PCBs is
employed as recovery surrogates. If the surrogates are recovered and other QC param-
eters are within acceptable limits, then the data may be considered valid.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI r'ield/Group
PCBs
Polychlorinated biphenyls
Chlorinated biphenyls
Analytical methods
Determination
By-products
GC/MS
Commercial products
Wastes
8. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
58
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |