United States
Environmental Protection
Agency
Environmental Research
Laboratory
Duluth MN 55840
EPA/600/3-85/019
May 1986
Research and Development
National Dioxin Study
Analytical Procedures and
Quality Assurance
Plan for the Analysis of
2,3,7,8-TCDD in Tier 3-7
Samples of the U.S.
Environmental Protection
Agency National Dioxin
Study
-------
EPA/600/3-85/01 9
May 1986
National Dioxin Study
Analytical Procedures and Quality
Assurance Plan for the Analysis of
2,3,7,8-TCDD in Tier 3-7 Samples of the
U.S. Environmental Protection Agency
National Dioxin Strategy
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Environmental Chemistry Laboratory
U.S. Environmental Protection Agency
NASA/NSTL, Building 1105
Bay St. Louis, MS 39529
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, MN 55804
-------
Notice
This document has been reviewed in accordance with U.S. Environmental
Protection Agency policy and approved for publication. Mention of trade names
or commercial products does not constitute endorsement or recommendation
for use.
Analytical methodology used by EMSL-RTP, ECL-BSL, and ERL-D in the analysis
of samples for the study of Tiers 3-7 in the U.S. EPA National Dioxin Strategy is
provided. The methods are based on low and high resolution mass spectrometry
using stable isotope labeled internal standards. The methods were peer
reviewed by the Quality Assurance Committee for the National Dioxin Strategy,
as revised in November 1984'. The methods are to serve only as a guideline for
these analyses and may be modified as required to successfully meet target
detection limits.
-------
Contents
Section I
Analytical Procedures and Quality Assurance Plan for
Tiers 3 through 7 of the National Dioxin Study.
Environmental Monitoring Systems Laboratory,
Environmental Chemistry Laboratory
Page
I. Introduction 1-1
II. Objective 1-2
III. Description of Organizational Structure and Route of Samples,
Analysis and Data Flow 1-2
IV. Block Diagram of Organizational Structure 1-2
V. Safety 1-2
VI. Background Information Regarding EMSL-RTP and
ECL-Bay St. Louis, Mississippi Methods of Analyses for
CDDs and CDFS 1-4
VII. Sample Preparation Procedures for 2378-TCDD,
Other CDDs and CDFs 1-4
A. Grinding, Blending, and Storage of Samples 1-4
B. Extraction Procedures 1-5
C. Cleanup Procedures 1-5
D. Glassware Cleanup Procedures 1-6
E. Soil Dry Weight Determination 1-7
VIM. EMSL-RTP HRGC-HRMS Methods of Analysis for 2378-TCDD,
Other CDDs and CDFs 1-7
A. Instrumentation 1-7
B. Calibration 1-7
C. Multiple Ion Monitoring Analyses 1-7
D. HRGS-HRMS MIS Operation Parameters 1-7
E. CDD and CDF Analytical Standards 1-8
F. Analysis of Sample Extracts for CDDS and CDFs 1-9
G. Analytical Criteria Used for Confirmation of 2378-TCDD 1-9
H. Analytical Criteria Used for Confirmation of
Other CDDs and CDFs 1-10
I. Quantification Procedures 1-10
J. Quantification of TCDFs, Penta through Octa CDDs find
CDFs and Determination of Minimum Limits of Detection 1-11
K. HRGC-HRMS Analytical Data 1-12
IX. Internal Quality Control/Quality Assurance 1-12
X. Quality Assurance/Quality Control Requirements for
Analytical Data 1-13
-------
Contents (Continued)
Page
XI. Capabilities, Qualifications, and Limitations of Methods of
Analysis for Tetra through Octa CDDs and CDFs 1-15
XII. External Quality Assurance/Multiple Laboratory Participation 1-15
XIII. Analytical Report Format, Reporting of Data and Review Panel .... 1-15
XIV. Sampling Procedures 1-16
XV. Sample Custody 1-16
Section II
Analytical Procedures and Quality Assurance Plan for the
National Dioxin Study.
Environmental Research Laboratory-Duluth
Page
Introduction 2-1
I. Sample Preparation for 2,3,7,8-TCDD Analysis 2-1
A. Grinding, Blending, and Storage of Samples 2-1
B. Extraction Procedures 2-1
C. Isolation Procedures 2-2
II. GC/MS Parameters 2-2
A. Quantification Standards 2-2
B. GC/MS Column Performance Standards 2-3
C. GC/MS Instrumentation Parameters 2-3
D. Quantification Procedure 2-3
III. Quality Assurance/Quality Control 2-5
A. General Procedures of Operation 2-5
B. Instrumentation 2-6
C. Evaluation of Data 2-7
IV. Percentage Lipid of Fish Tissue 2-8
V. Age of Fish 2-9
VI. Sampling Procedures 2-9
VII. Sample Custody 2-9
VIM. Data Reporting Format 2-9
IX. Dioxin Isostereomer Analysis 2-9
X. Validation of Methodology and Evaluation of Data Comparability 29
A. PCDD/PCDF Congeners 2-9
B. Other Polychlorinated Planar Molecules 2-9
fv
-------
Contents (Continued)
Page
Appendix A: Capillary GC Conditions A-1
Appendix B: Mass Spectrometric Conditions B-1
Appendix C: National Dioxin Study Data Hardcopy Reporting Form C-1
Appendix D: National Dioxin Study Computer File D-1
Appendix E: ERL-Duluth Sample Analysis Tracking Form E-1
-------
Section I
Analytical Procedures and Quality Assurance Plan for Tiers 3 Through 7 of the
National Dioxin Study
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Environmental Chemistry Laboratory
U.S. Environmental Protection Agency
NASA/NSTL, Building 1105
Bay St. Louis, Mississippi 39529
I. Introduction
The identification and quantitative measurement of
part-per-trillion (ppt) to part-per-quadrillion (ppq)
levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin (2378-
TCDD) TCDD isomers, other CDDs and CDFs in
various sample matrices is currently performed by
only a limited number of laboratories in the U.S. and
in foreign countries. Efficient sample preparation pro-
cedures and sensitive and specific detection tech-
niques are required for the analytical determinations.
Credibility and validity of results at this level have
been the subject of intense scrutiny by the scientific
community. Therefore, stringent measures have been
used by EPA since 1975 to validate the analytical
methodology/results derived for quantitative meas-
urement of 2378-TCDD by participating laboratories.
Four basic principles used for this purpose were:
comprehensive studies were performed to demon-
strate and validate the methods; analysis of blind
quality assurance QA samples during the analysis of
test samples; multiple laboratory participation; and
defined analytical criteria used for confirmation of
2378-TCDD. Participants have included two EPA
laboratories. Environmental Monitoring Systems
Laboratory(EMSL-RTP) and the Environmental Chem-
istry Laboratory-Bay St. Louis, Mississippi (ECL);
Wright State University; University of Nebraska;
Harvard University; University of Utah and Dow
Chemical Company. Sample preparation procedures
and detection techniques utilized by the various
laboratories are quite different, yet the analytical
results have been in remarkably close agreement as a
result of the application of these principles. Therefore,
the credibility and validity of each laboratory's method
and the results are enhanced. These four principles
will also be used in the recently created National
Dioxin Study.
In order to satisfy the urgent need for special
analytical services, existing facilities, equipment and
personnel of ORD and OPP, which are currently
involved in 2378-TCDD analysis are being detailed
and expanded into a consolidated two year effort by
establishment of a National Trace Analysis TROIKA.
TheTROIKA consists of components from the Environ-
mental Research Laboratory-Duluth (ERL-D) and the
Environmental Monitoring Systems Laboratory-
Research Triangle Park (EMSL-RTP) of the Office of
Research and Development (ORD) and the Environ-
mental Chemistry Laboratory-Bay St. Louis (ECL) of
the Office of Pesticide Programs (OPP). These three
laboratories have two major functions: (1) provide
immediate and valid ppt and ppq level analytical
results for the National Study and (2) serve as the
central nucleus for the incorporation of laboratories
from the private sector into the national program after
they have demonstrated and validatedtheir analytical
capabilities in the EMSL-LV certification program. It
is estimated that direct TROIKA involvement will be
minimal by October 1, 1985 and the analysis there-
after will be handled directly by the State or Regional
Offices and contract laboratories.
The primary efforts of the TROIKA will be devoted to
analytical determinations for 2378-TCDD. It is esti-
mated that 100 to 150 analyses for 2378-TCDD per
month can be generated. Analyses will also be
performed for determination of penta- through octa-
chlorinated dibenzo-p-dioxins (CDDs) and tetra-
through octa-chlorinated dibenzofurans (CDFs) in
samples from special studies. These type of analyses
require additional time and effort and pose some
problems due to unavailability of sufficient reference
standards.
The portion of this plan concerned with TCDD isomers,
other CDDs and CDFs represents a unique endeavor
for the Agency. The complexity of the analytical task
to extract, separate and identify the many CDDs and
CDFs combined with detection and quantification at
parts per trillion levels yields a study of such a scope
as to defy comparison. No reference point exists for so
many analyses at such low levels. Compounding the
complexity and adding to the uncertainty is the lack of
certified reference standards for many of the CDDs
1-1
-------
and CDFs and the unknown extraction properties of
the matrices to be tested. For these reasons, portions
of this document pertaining to CDDs and CDFs can
only be considered an "expectation plan." It defines
the quality of work expected. Only the assessment of
the actual analytical data will determine if the
expectations exceeded the current analytical capabil-
ities.
Responsibility for performing the Tier 4 analytical
work was assigned to the ECL-BSL and EMSL-RTP
laboratories during the last year of the study. Sample
preparations are performed by the ECL laboratory and
the HRGC-HRMS analyses for tetra through octa
CDDs and CDFs are performed by the EMSL-RTP
laboratory. The sample preparation procedures and
HRGC-HRMS methods of analyses are fully described
in the Tier 4 attachment to this document.
II. Objective
The objective of this specific portion of the National
Study is to determine the absence and/or degree of
environmental contamination by ppt levels of 2378-
TCDD in soil, sediment, fish and ppq levels in water.
Analytical data generated in this study must be
scientifically sound and valid in order for it to be used
for meaningful assessment purposes. Although a
brief description of the organization and responsibil-
ities of various laboratories and offices is presented,
the primary purpose of this document is to present the
analytical methodology and the quality assurance
procedures that will be used to insure that the quality
of data is sufficient to satisfy the study objective.
This QA plan for the analytical portion of the work is
designed for use by experienced EPA laboratories for
a period of two years. The analytical methods that will
be used by the EPA laboratories are described fully
and those used by specific universities are described
in peer reviewed journals. The described methods
will be used for the major part of the work. However,
based on experience, it is emphasized that many
"troublesome" samples will be encountered during
this two-year study. Therefore, the laboratories will
make any necessary modifications and/or develop
the methods required to provide valid analytical
results for the Agency. At the conclusion of this effort
in October 1985, a specific QA plan for use by private
contract laboratories will be prepared by appropriate
EPA personnel.
Collection, storage, shipment and prioritization of
samples are described in a separate document
prepared by OW.
III. Description of Organizational
Structure and Route of Samples,
Analysis and Data Flow
The organizational structure that describes this
portion of the National Study is shown on page 1 -3.
Briefly, requests for analysis are submitted to the OW
sample management office for prioritization. The
requestor is then notified about the status of the
request, the date and the analytical laboratory that
the samples should be sent to. Simultaneously, the
information is also transmitted to the analytical
laboratory, i.e., the Environmental Chemistry Labora-
tory (ECL), located at Bay St. Louis, Mississippi. After
receipt of samples, appropriate quality assurance
(QA) samples and test samples are fortified with
labeled compounds of interest and subjected to
specific extraction and clean-up procedures. Extracts
from the sample preparation procedures (60 fj\
extracts) are then shipped by Federal Express to the
EMSL-RTP for HRGC-HRMS analysis. Analytical
results for the set of samples are verbally transmitted
to the ECL. Any problems that may be encountered
must be resolved before proceeding to the next
sample set. If QA results are within acceptable limits
the EMSL-RTP submits a memorandum describing
analytical results to the ECL. The ECL inserts theoret-
ical QA values and a signature and submits the
memorandum to ERL-D. This can now be considered
valid data upon which tentative actions may be
formulated, i.e., confirmation analysis by a referee
laboratory or collection of more samples to determine
extent of contamination. Results generated in each
quarter will be subjected to a peer review prior to
releasing them to EPA-HQ.
IV. Sample and Data Flow
(see diagram on page 1 -3)
V. Safety
2378-TCDD is toxic and can pose health hazards if
handled improperly. Techniques used for handling
radioactive and infectious materials are applicable to
2378-TCDD. Only highly trained individuals who are
thoroughly versed in appropriate laboratory proce-
dures and familiar with the hazards of 2378-TCDD
should handle this substance. A good laboratory
practice involves routine physical examinations and
blood checks of employees working with 2378-TCDD.
Also, facial photographs using oblique photoflood
lighting should be periodically taken to detect
chloracne, which is an early indication of 2378-TCDD
exposure.
Safety and health protocols and a well ventilated,
controlled access laboratory are required for work
involving 2378-TCDD. Persons working with 2378-
TCDD should be enrolled in a health monitoring
program and subjected to specific medical examina-
tions on a semi-annual or annual basis. Other
requirements are: appropriate safety clothing and eye
1-2
-------
Slate
State
1
State
State
Regional Office
Collect Samples
Sample Control Center
ERL-Duluth Sample Tracking
Issue SCC Numbers
Distribute Sample
Receive Data
Sample
ECL-BSL Sample Preparation
HRGC-HRMS Analysis
Data
Sample Extracts
EMSL-RTP
HRGC-HRMS Analysis
Data
Data
ERL-D Sample Preparation
HRGC-LRMS Analysis
HRGC-HRMS Analysis
Compiled Data
TROIKA Coordinator
Data Review
Electronic Transmission
Office of Water
Regional Office
T
State
Distribution of Data
State
State
State
protection equipment should be provided; exhausts
from vacuum pumps should be vented through
appropriate filters to the atmosphere; analytical
standards and sample extracts should be stored in
locked refrigerators; work with 2378-TCDD should be
performed in the hood; bench tops and floor of hood
should be covered with teflon sheets and/or plastic
backed adsorbent paper; the surfaces of bench tops,
1-3
-------
hood floor, laboratory floor, and apparatus should be
periodically subjected to wipe tests utilizing filter
paper wetted with appropriate solvent, such as
hexane; filter papers should be subjected to sample
preparation prior to HRGC-HRMS analysis for 2378-
TCDD (filter paper extracts may be analyzed without
clean-up to determine the extent of contamination if
an accident should occur); procedures to deal with
accidents should be clearly defined in the safety and
health protocols; inventories of 2378-TCDD and other
toxic compounds in the laboratory should be per-
formed on a semi-annual basis to insure that large
amounts are not allowed to accumulate (analytical
standards that cannot be used and the remainder of
sample extracts should be removed from the labora-
tory and stored in appropriate room or building to
await disposal); liquids should be concentrated to
dryness and all disposable items such as absorbent
paper and towels that may be contaminated with low
levels of 2378-TCDD should be stored in steel drums
fitted with heavy gauge polyethylene liners to await
proper disposal. In summary, careful attention to
details, log books, inventories, analytical reports, safe
laboratory practices and good judgment are required
in order to document and perform the necessary work
in a safe and acceptable manner.
VI. Background Information Regarding
EMSL-RTP and ECL-Bay St. Louis,
Mississippi Methods of Analyses for
CDDs and CDFs
The EMSL-RTP and ECL-Bay St. Louis, Mississippi
laboratories have performed analyses for 2378-TCDD
for EPA since 1973. Basic procedures and methods of
analysis used today were developed in early 1977 and
are described fully in Analytical Chemistry, pp. 1239-
1245, 1980. They have been applied for the deter-
mination of 2378-TCDD in a wide variety of sample
matrices. For example, human milk, human adipose
tissue, beef, deer and elk adipose tissue and muscle,
fish, water, soil, sediment, fly ash, stack gas emis-
sions, chemical products, etc. The credibility and
validity of these methods have been demonstrated
and established by participation in numerous national
and international method validation studies such as
the Canadian/American Study regarding 2378-TCDD
in fish from the Great Lakes.
These basic sample preparation procedures and
HRGC-HRMS methods of analysis have been modified
over the years to perform analyses for tetra- through
octa-chlorinated CDDs and CDFs and to incorporate
knowledge, wisdom and advances such as fused
silica SP-2330 capillary columns as they become
available. The major modifications are: 13Ci2-2378-
TCDD and 13Ci2-OCDD are used as internal standards
to determine method efficiency achieved for each
sample. These two internal standards are also used in
quantification of CDDs and CDFs and for determining
the minimum limits of detection for each series of
tetra through octa CDD and CDF isomers; mild
extraction procedures are used because hot KOH
digestion will destroy hexa through octa CDDs; fused
silica polar HRGC capillary columns are used for
resolution of components; analytical criteria for
confirmation of 2378-TCDD have been expanded
and/or modified to use for other CDDs and CDFs as
described in specific portions of the analytical
methods. The credibility and validity of these pro-
cedures for tetra through octa CDDs and CDFs have
been demonstrated and validated by participation in
method validation studies with Wright State
University—water samples, Dow Chemical—soil
samples, and an international study—human adipose
tissue samples. These analyses for tetra through octa
CDDs and CDFs are complex, time consuming and
have specific limitations due to unavailability of
sufficient labeled and native reference standards.
Analytical standards of available CDDs and CDFs are
prepared by ECL-Bay St. Louis, Mississippi and
submitted to EMSL-RTP for HRGC-HRMS analyses
and certification for use in the National Dioxin Study.
The analytical capabilities and the limitations of these
analyses are described in specific portions of the
following analytical methods.
The QA program is instituted simultaneously with
sample preparation. Aliquots (specific grams) of QA
and test sample matrices are fortified with 2.5, 5.0 or
10 ng 13Ci2-2378-TCDD in order to determine and
document the efficiency achieved in sample prepa-
ration (NOTE: 20 ng 13Ci2-OCDD are also added at this
point if tetra through octa CDD and CDF analyses are
required). QA samples are also fortified with various
levels of native 2378-TCDD and various levels of
2378-TCDF and penta through octa CDDs and CDFs if
these analyses are also required. Extracts of QA and
test samples are submitted in a "blind" fashion to the
analytical laboratory for HRGC-HRMS analysis, i.e.,
the identity of the QA samples in the set of extracts is
unknown to the mass spectroscopist.
VII. Sample Preparation Procedures for
2378-TCDD, Other CDDs and CDFs
A. Grinding, Blending and Storage of Samples
Samples received at ECL-Bay St. Louis will be
pre-ground and/or blended at the region or state
laboratories. If not, homogeneous samples will
be prepared at ECL. Fillet, whole fish or specific
organs of animal analyses will be determined by
the region or state submitting the sample. The
major portion of all samples will be stored for
reference. Fish and animal tissue will be stored
at -15°C. Water, sediment and soil samples will
be stored in the dark at 4°C.
1-4
-------
B. Extraction Procedures
1. Soil
Weigh 10.0 g of soil into a 33 mm x 80 mm
cellulose thimble. Add 5.0 ng of 13Ci2-2378-
TCDD directly onto the soil (NOTE: add 20 ng
13Ci2-OCDD if other CDD and CDF analysis
required), cover with a plug of glass wool and
place the thimble into a 40 mm soxhlet extraction
apparatus fitted with a 250 ml_ Erlenmeyer flask.
Add 175 mL of benzene and place the apparatus
on a hot plate. Adjust the heat until the benzene
drips at a rate of two drops per second. Extract for
1 6 hours. Allow the apparatus to cool. Remove
the extractor and place a three-bulb Snyder
column onto the flask containing the benzene
extract. Place on a hot plate and concentrate the
benzene to 10 mL (do not let go to dryness). Add
100 mL of hexane and again concentrate to 10
mL. Add a second 100 mL portion hexane and
concentrate to 10 mL. Let cool and add 25 mL
hexane. The extract is ready for the acid/base
cleanup at this point.
2. Sediment
Place a 11.0 cm round medium porosity filter
paper on a 15 cm x 15 cm piece of screen wire,
weigh and record the weight. Weigh 10.0 g of
wet sediment onto the filter paper and spread
evenly into a 1-2 mm layer and record the weight.
Add 5.0 ng of 13Ci2-TCDD directly onto the
sediment layer (NOTE: add 20 ng 13Ci2-OCDD if
other CDD and CDF analysis required). Let the
sediment dry Jn a chemical fume hood at room
temperature for 24 hours. Re-weigh the filter
paper, dry sediment and the screen wire. Transfer
the filter paper and dry sediment into a 33 mm x
80 mm cellulose thimble and cover with a plug of
glass wool. Proceed with the soil procedure at
the point where the thimble is placed in the 40
mm soxhlet extraction apparatus.
3. Water
Measure exactly 1000 mL of water and pour into
a 2 L separatory funnel. Add 2.5 ng of 13Ci2-
2378-TCDD directly to the water (NOTE: add 20
ng 13Cis-OCDD if other CDD and CDF analysis
required). Add 100 mL of methylene chloride
(CH2CL2) to the separatory funnel and shake
vigorously for about one minute. Allow the layers
to separate and drain the lower layer (CH2CL2)
into a 500 mL Erlenmeyer flask. Reextract the
water with 2 more 100 mL portions of CH2CL2
and combine all the CH2CL2 extracts in the 500
mL Erlenmeyer flask. Place a three-bulb Snyder
column on the 500 mL Erlenmeyer flask, add a
few large granules of sodium sulfate and concen-
trate to ~10 mL on a steam bath. Add 100 mL of
hexane and again concentrate to 10 mL. Allow
the flask to cool and add 25 mL hexane. The
extract is ready for the acid/base cleanup at this
point.
4. Fish and Tissue
Weigh 10.0 g of a homogeneous fish or tissue
sample into a 10OmL round bottom boiling flask.
Add 5.0 ng of 13C12-2378-TCDD directly to the
sample (NOTE: add 20 ng 13Ci2-OCDD if other
CDD and CDF analysis required). Add 20 mL of
ethyl alcohol, 40 mL of a 45% potassium
hydroxide solution, a teflon-coated stir bar and
stir at room temperature for 1 6 hours. Transfer
the alcoholic-base solution to a 250 mL separa-
tory funnel. Rinse the boiling flask with 10 mL of
ethanol and add the rinse to the separatory
funnel. Rinse the boiling flask with 25 mL of
hexane and transfer the hexane wash to the
separatory funnel. Shake the separatory funnel
vigorously for one minute. Allow the hexane
layer to separate and drain the aqueous layer
(lower layer) into a second 250 mL separatory
funnel. Drain the hexane extract from the first
separatory funnel into a 125 mL Erlenmeyer
flask and save. Add 25 mL of hexane to the
second separatory funnel and shake vigorously
for one minute. Allow the layers to separate and
drain the lower layer back into the first separatory
funnel. Drain the hexane into the 125 mL
Erlenmeyer flask. Repeat the extraction step
with two more 25 mL portions of hexane.
Combine the four 25 mL hexane extracts into
one of the separatory funnels. Proceed with the
acid/base cleanup at the point where the
combined hexane is washed with 30 mL of 2 N
potassium hydroxide.
C. Cleanup Procedures
1. Acid/Base Cleanup
Transfer the hexane extract to a 250 mL separa-
tory funnel with 2-25 mL portions of hexane.
Wash the combined hexane with 30 mL of 2 N
potassium hydroxide. Allow layers to separate
and discard the aqueous layer. Carefully add 50
mL of concentrated sulfuric acid. Shake vigor-
ously for 1 minute, allow layers to separate and
discard the acid layer. Repeat the acid wash with
two more 50 mL portions of concentrated sulfuric
acid. Carefully add 25 mL of distilled water,
shake, allow layers to separate and discard the
water. Transfer the hexane through a 42 mm x
160 mm filter funnel containing a plug of glass
wool and 3 cm of sodium carbonate into a 250 mL
Kuderna-Danish (KD) concentrator fitted with a
15 mL catch tube. Rinse the filter funnel with
2-25 mL portions of hexane. Place a three-bulb
Snyder column on the KD concentrator and
1-5
-------
concentrate on a steam bath to 1 -2 ml_. The
extract is ready for the alumina column cleanup
at this point, but it can be stoppered and stored in
the dark if necessary.
2. Alumina Column Preparation
Gently tamp a plug of glass wool into the bottom
of a 5-3/4 inch disposable Pasteur pipet. Pour
Woelm neutral alumina into the pipet while
tapping the column with a pencil until a height of
4.5 cm of alumina is packed into the column. Top
the alumina with 0.5 cm of anhydrous granular
NazS04. Prewash the column with 3 ml CH2CL2.
Allow the CH2CI_2 to drain from the column, then
force the remaining CH2CL2 from the column
with a stream of dry nitrogen. Place prepared
columns in an oven set at 225°C. Store columns
in the oven until ready for use, at least overnight.
Remove only the columns needed and place
them in a dessicator over Drierite until they have
equilibrated to room temperature. Use immedi-
ately.
3. Alumina Column Cleanup
Prewet the alumina column with 1 mL of hexane.
Transfer the 2 mL hexane extract from acid/base
cleanup onto the column. Wash the column with
6.0 ml of carbon tetrachloride and discard. Elute
the column with 4.0 mL of methylene chloride
and catch the eluate in a 12 mL distillation
receiver. Place a micro-Snyder column on the
receiver and evaporate the methylene chloride
just to dryness by means of a hot water bath. Add
2 mL of hexane to the receiver and evaporate just
to dryness. Add another 2 mL portion of hexane
and evaporate to 0.5 mL. The extract is ready for
the carbon column cleanup at this point.
4. Carbon Column Preparation
Weigh 9.5 g of Bio-Sil-A (100-200 mesh) silica
gel, which has been previously heated at 225°C
for 24 hours, into a 50 mL screw cap container.
Weigh 0.50 g of Amoco PX-21 carbon onto the
silica gel, cap and shake vigorously for one hour.
Just before use, rotate the container by hand for
at least 1 minute. Break a glass graduated 2.0
mL disposable pipet at the 1.8 mL mark and fire
polish the end. Place a small plug of glass wool in
the pipet and pack it at the 0.0 mL mark using
two small solid glass rods. Add 0.1 mLof Bio-Sil-
A 100-200 mesh silica gel. If more than one
column is made at one time it is best to add the
silica gel to all the columns and then add the
carbon-silica gel mixture to all the columns. Add
0.45 mL of the carbon-silica gel mixture to the
column. The top of this mixture will be at the
0.55 mL mark on the pipet. Top the column with
a small plug of glass wool.
5. Carbon Column Cleanup
Place the carbon column in a suitable clamp with
the silica gel plug up. Add approximately 0.5 mL
of 50% benzene-methylene chloride (v/v) to the
column. Fita 10mL disposable pipet on the top of
the carbon column with a short piece of extruded
teflon tubing. Add an additional 9.5 mL of the
50% benzene-methylene chloride. When approx-
imately 0.5 mL of this solvent remains, add 10
mL of toluene. After all the toluene has gone
onto the column, remove the 10 mL reservoir
and add a total of no less than 2.0 mL of hexane
to the column. When approximately 0.1 mL of
the hexane is left on the top of the column,
transfer the sample extract (0.2 mL to 0.4 mL in
hexane) onto the column with a Pasteur pipet.
Rinse the distillation receiver that contained the
extract with two separate 0.2 mL portions of
hexane and transfer each rinse onto the column.
Allow the top of each transfer layer to enter the
glass wool before adding the next one. When the
last of the transfer solvent enters the glass wool,
add 0.5 mL of methylene chloride, replace the 10
mL reservoir and add 4.5 mL of methylene
chlorideto it. When approximately 0.5 mLof this
solvent remains, add 10 mL of 50% benzene-
methylene chloride. When all this solvent has
gone onto the column, remove the reservoir,
take the column out of the holder, rinse each end
with toluene, turn it over and put it back in the
holder. All previous elution solvents are dis-
carded. Place a suitable receiver tube under the
column and add 0.5 mL of toluene to the top of
the column. Fit the 10 mL reservoir on the
column and add 9.5 mL of toluene to it.
When all the toluene has eluted through the
column and has been collected in the receiving
tube, add 3 microliters of tetradecane and
concentrate just to dryness using a stream of
nitrogen and a heating block of 60°C. Just before
the sample reaches dryness, it should be put into
a solvent such as benzene or iso-octane which is
suitable for GC/MS analysis.
D. Glassware Cleanup Procedures
In this procedure each piece of glassware is
taken through the cleaning separately except in
the oven baking process. The 100 mL round
bottom flasks, the 250 mL separatory funnels,
the KD concentrators, etc., that are used in the
digestion and extraction procedures are washed
three times with hot water, two times with
acetone and two times with hexane. This glass-
ware is then baked in a forced air oven which is
vented to the outside for 16 hours at 450°C. The
teflon stopcocks are cleaned as above except for
the oven baking step. All glassware is rinsed with
acetone and hexane immediately before use.
1-6
-------
E. Soil Dry Weight Determination
Weigh a 15 cm diameter aluminum pie pan and
record the weight. Weigh 20.0 g of soil into the
pan and spread evenly in the bottom and record
the weight. Place the pan containing the soil in a
chemical fume hood and dry at room temperature
for 48 hours. Re-weigh the dry soil and pan and
record the weight.
% Dry Weight Calculation
wt. dry soil x 100 = Dry Weight
wt. wet soil
VIM. HRGC-HRMS Methods of
Analysis for 2378-TCDD, Other
CDDs and CDFs
A. Instrumentation
1. EMSL-RTP—A Varian/MAT 311A mass spec-
trometer (MS) directly coupled (open split inter-
face)toa Varian Model 2700gaschromatograph
(GC) is utilized for these analyses. The GC is
equipped with polar or non-polar WCOT glass or
fused silica capillary columns. The MS is equip-
ped with an eight-channel hardwired multiple
ion selection (MIS) device. The MIS output signals
are recorded on an eight-channel Soltec strip
chart recorder. This HRGC-HRMS manual control
MIS device is fully described in the described
Analytical Chemistry publication. NOTE: Com-
puter controlled MIS acquisition and processing
programs will be installed in October 1984. This
will hopefully provide for much more efficient
analysis for tetra through octa CDDs and CDFs.
For example, the analysis for tetra through octa
CDDs and CDFs can be performed on one sample
injection.
2. ECL-BSL—A Varian/MAT 312 mass spectrom-
eter (MS) directly coupled (modified open split
coupler) to a Varian Model 3700 gas chromato-
graph (GC) is utilized for these analyses. The GC
is equipped with polar or non-polar WCOT fused
silica capillary columns. The MS is equipped with
an eight-channel hardwired multiple ion selec-
tion (MIS) device. The MIS output signals are
recorded on an eight-channel Rikadenki strip
chart recorder.
B. Calibration of HRGC-HRMS Instrumentation
Perfluorokerosene (PFK), m/z 318.9793, is used
in daily operations to establish mass resolution
and initial sensitivity. 250 pg 13Ci2-2378-TCDD
and 10 pg native 2378-TCDD are injected at the
beginning of each day of operation. Parameters
are optimized to provide approximately 60% full
scale deflection on the strip chart recorder for the
10 pg 2378-TCDD. The m/z 320/322 chlorine
isotope ratio must be between 0.67 and 0.87:1.0
before initiation of analysis of test samples.
Standards are analyzed at intervals during daily
operation to provide the data required for quanti-
fication of 2378-TCDD in test samples.
The instrumentation is calibrated for analysis of
other CDDs and CDFs in a similar manner. For
example, PFK m/z 330.9793 is used as the
reference mass to establish mass resolution and
initial sensitivity. 250 pg 13Ci2-2378-TCDD and
50 pg each of penta and hexa CDF are injected at
the beginning of each day of operation. Param-
eters are optimized to provide approximately 60%
full scale deflection on the strip chart recorder for
the 50 pg each of penta and hexa CDF. The
chlorine isotope ratio achieved for the respective
compounds must be ±20% of their theoretical
values. Standards are analyzed at intervals
during daily operation to provide the data required
for quantification of penta and hexa CDFs in test
samples.
A diagram of the HRGC-HRMS MIS analysis for
CDDs is shown on the next page.
C. Multiple Ion Monitoring Analysis
(see diagram on page 1-8)
D. HRGC-HRMS MIS Operation Parameters
EMSL-RTP ECL-BSL
1. HRGC Parameters
60mSP-2330fusedsilica
capillary column
Helium carrier gas—
1.2 ml/min
Injection port
temperature—275°C
Injection technique—
splitless
Same
Same
Same
Same
Initial column
temperature—100°C 140°C
Hold for seven minutes at
100°C after injection Hold for 2 min/140°C
Program to 240°C at Program to 240°C at
25°C/min 20°C/min
Hold at 240°C for
20 to 40 minutes
Same
1-7
-------
's.
/ \
PFK
REF
/ \
TCDDs
r L_ }
1 '3C,2-TCDD
PENTA CDDs
HEXA-CDDs
318.9793
319.8965
321.8935
333.9338 355 8546
3578517
389.8156
391 8127
v/
PFK
REF
HEPTA CDDs
\/
\/
OCDD
13C12-OCDD
4189729 4237766 4257737 457.7377 4597347
INTEGRATION TIME 30 msec/mass
469.7780
or
471 7749
GC transfer lines to ion
source—270°C Same
2. HRMS Parameters
Electron energy, 70 eV ~40 eV
Filament emission, 1mA 2mA
Acceleration voltage,
3 kV maximum Same
Mass resolution,
8000 to 10,000 Same
Multiplier gain greater
thanIO6 Same
Ion source temperature,
270°C Same
3. Hardwired MIS Parameters
Repetitive integration
rate, 30 milliseconds
per mass Same
Jump time between
masses, 12 milliseconds Same
E. COD and CDF Analytical Standards
Analytical standards of CDD and CDF reference
isomers described below were prepared by ECL-
Bay St. Louis, Mississippi. An aliquot of each
standard was submitted to EMSL-RTP for HRGC-
HRMS analysis and certification for use in the
National Dioxin Study. These analytical standards
were subjected to comprehensive HRGC-HRMS
analysis to determine: the presence or absence
of pg amounts of native isomers in ng amounts of
the labeled standards; the presence of more than
one isomer and lower and higher chlorinated
isomers; the concentrations of specific analytical
standards were compared with those prepared
by FDA, Dow Chemical, Wright State University
and the University of Nebraska. The analytical
standards and several qualitative reference
standards that are used in the National Dioxin
Study are listed as follows:
13C12-2378-TCDD
13C12-2378-TCDF
13C12-OCDD
37CL4-2378-TCDD
2378-TCDD
1234-TCDD
1368-TCDD
2378-TCDF
13478-penta-CDF
124679-hexa-CDF
123478-hexa-CDD
1234689-hepta-CDF
37CI4hepta-CDD
1234678-hepta-CDD
OCDD
OCDF
Qualitative standards used for establishing re-
tention time windows and for identification
purposes:
1-8
-------
Mixture of 22 TCDD isomers
2:2 type TCDD isomers
3:1 type TCDD isomers and 1234-TCDD
1368/1379-TCDDs
1469/1269/2348/1289/2468-TCDFs
12378-penta CDD
Mixture of tetra through hexa-CDF isomers
Extracts of municipal incinerator fly ash that
contains all tetra through octa'CDDs and CDFs
F. Analyses of Sample Extracts for CDDs and
CDFs
Each sample extract may be subjected to one or
all of the following six analyses. The reference
mass, exact masses of respective CDD and CDF
compounds and the type of capillary column that
is typically used are listed as follows:
1. PFK, 318.9793; TCDDs, 319.8965/321.8936;
13C12-TCDD, 333.9338.
60 m fused silica SP-2330 capillary column.
2. PFK, 292.9825; TCDFs, 303.9016/305.8987;
TCDDs, 319.8965/321.8936; 13Ci2-TCDD,
333.9338.
60 m fused silica SP-2330 capillary column.
3. PFK, 330.9793; 13d2-TCDD, 333.9338; PCDDs,
355.8546/357.8517; HxCDDs, 389.8156/
391.8127.
30 m fused silica SP-2330 capillary column.
4. PFK, 330.9793; 13C12-TCDD, 333.9338; PCDFs,
339.8597/341.8567; HxCDFs, 373.8207/
375.8178.
30 m fused silica SP-2330 capillary column.
5. PFK, 418.9729; Hepta CDDs, 42S.7766/
425.7737; OCDD, 457.7377/459.7348; 13C12-
OCDD, 471.7750.
20 m fused silica SP-2330 or 15 m SE-54 WCOT
glass capillary column.
6. PFK, 404.9761; Hepta CDFs, 407.7817/
409.7788; OCDF, 441.7428/443.7398; 13C12-
OCDD, 471.7750.
20 m fused silica SP-2330 or 15 m SE-54 WCOT
glass capillary column.
HRGC-HRMS Methods of Analysis
Utilizing a Digital POP-11 /34
Computer and Associated MIS
Acquisition and Processing Programs.
The computer based MIS programs were installed
and put into service during the last year of the
study. This concept is similar but more efficient
than the previously described hardwired MIS
technique. For example, the analysis for total
tetra through octa CDDs and CDFs are performed
on one sample injection utilizing a 30m or 60m
OV-101 or DB-5 fused silica capillary column. A
standard or standards containing all of the
previously described labeled and native tetra
through octa CDDs and CDFs are used for
determination of response factors and for quanti-
fication purposes. The peak area of each tetra,
penta, and hexa CDD and CDF is used to deter-
mine the response factor of each compound
relative to the labeled 37CL4 or13Ci2-2378-TCDD.
Response factors for the hepta and octa CDDs
and CDFs are developed in a similar manner
relative to the peak area of 13Ci2-OCDD. This
method of analysis is described fully in the
attachment regarding Tier 4.
Aliquots, 1 to 3 /ul, of 60 //I extracts are injected
for analysis. Typical a mounts of analytical stand-
ards injected for quantification purposes are:
• 100 pg 13C,2-TCDD, 2.5 pg 2378-TCDD
• 250 pg 13Ci2-TCDD, 10 pg 2378-TCDD
• 250 pg 13C12-TCDD, 10 pg 2378-TCDD, 16 pg
2378-TCDF
• 250 pg 13Ci2-TCDD, 50 pg each of penta and
hexa CDF
• 250 pg 13Ci2-TCDD, 50 pg of hexa CDD
• 1000 pg 13Ci2-OCDD, 100 pg each of hepta
CDD and OCDD
• 1000 pg 13Ci2-OCDD, 100 pg each of hepta
CDD and OCDF
A standard containing all of these compounds in
similar concentrations/amounts is used in com-
puter MIS analysis.
• NOTE: nanogram amounts used for determi-
nation of high concentrations
G. Analytical Criteria Used for Confirmation of
2378-TCDD
Since many compounds can interfere with the
determination of CDDs and CDFs, it is of the
utmost importance that positive identifications
be made. Compounds such as PCBs (hepta and
nonachlorobiphenyl), chlorinated methoxybi-
phenyls, DDT, DDE, chlorinated phenyl-benzo-
quinones, chlorinated xanthenes and hydroxy-
chlorinated dibenzofurans interfere at the m/z's
of interest and necessitate a resolution of at least
8,000. Chromatographic separation of the 22
TCDD isomers must be achieved and verified by
injection of isomer mixes to confirm resolution of
2378-TCDD. It is crucial that the 2378-TCDD
isomer be separated from all other isomers. The
criteria for positive 2378-TCDD confirmation are
listed below. Criteria 1 -4 must be met before the
presence of 2378-TCDD can be considered valid.
These criteria will be met for all positive samples
1-9
-------
reported by the EPA laboratories performing the
analyses. Supplemental criteria, a or b are
applied to a representative number of positive
samples as needed for conclusive confirmation
purposes. The mass spectroscopist will use the
appropriate and necessary technique and criteria
to provide the required valid data for the Agency.
1. Correct HRGC-HRMS retention time of 2378-
TCDD (±3 sec) on a 2378-TCDD isomer specific
column relative to the labeled internal standard.
GC peak resolution should not exceed 35% valley
for TCDD isomers of equal concentrations eluting
before and after 2378-TCDD.
2. Correct chlorine isotope ratio of molecular ion,
m/z 320/322 (0.67-0.87).
3. Correct HRGC-HRMS multiple ion monitoring
response for 2378-TCDD masses and 13Ci2 or
37CI4-TCDD mass (simultaneous response, ±3
sec, for elemental compositions m/z 320, 322,
and 334 or 328).
4. Response of m/z 320/322 must be greater than
2.5 times the noise level.
Supplemental Criteria:
a. COCL loss indicative of TCDD structure.
b. HRGC-HRMS peak matching analysis of m/z
320/322 in real time to confirm exact masses
that correspond to TCDD elemental compositions.
H. Analytical Criteria Used for Confirmation of
Other CDDs and CDFs
1. HRGC-HRMS retention time of specific CDD or
CDF isomers available relative to the labeled
internal standards.
2. HRGC-HRMS retention time window of respec-
tive CDD or CDF series of isomers based on
reference fly ash extract.
3. Chlorine isotope ratio of previously described
molecular ions of respective CDD or CDF isomers
within ±20% of theoretical values:
penta—1.54 octa—0.88
hexa—1.23 hepta—1.03
Supplemental criteria: a representative number
of samples may be subjected to the following
analyses if needed for confirmation purposes:
a. HRGC-HRMS peak matching analysis of molecu-
lar ions in real time to confirm masses that
correspond to the elemental compositions of
respective CDDs or CDFs.
b. Comparison of sample analysis to analysis of
reference fly ash sample that contains all of the
tetra through octa CDDs and CDFs.
c. HRGC-HRMS analysis to confirm the absence of
specific chlorinated diphenylethers at appropriate
retention times.
/. Quantification Procedures
The manual control MIS has specific limitations
in regard to quantification over a wide range.
Therefore, standards in a very narrow range of
interest are used for quantification purposes.
1. Method Efficiency, percent recovery of 13Ci2-
2378-TCDD and 13Ci2-OCDD.
Aliquots of external standards containing specific
amounts of 13Ci2-2378-TCDD and native 2378-
TCDD are analyzed periodically during the anal-
ysis of sample extracts in daily operations. The
peak height response produced by the specific
amount of 13Ci2-2378-TCDD in the standard is
used to calculate the amount of 13Ci2-2378-
TCDD in the aliquot of sample extract analyzed.
The amount present in the aliquot is used to
determine the amount of 13d2-2378-TCDD in
the total volume of extract. Typical responses
produced by sample and standard are shown
along with other pertinent data:
Sample
m/z 334
m/z 322
m/z 320
Peak
Height
Response
195 mm
170 mm
133 mm
Standard
m/z 334
m/z 322
m/z 320
Peak
Height
Response
200 mm for
250pgl3C12-
2378-TCDD
150 mm for
1 0 pg native
2378-TCDD
1 1 7 mm for
10 pg native
2378-TCDD
Aliquot analyzed, 2 /ul from 60 fj\
10 ng 13Ci2-2378-TCDD added to 5 gram
sample prior to extraction
Example of calculation to determine % recovery-
Standard 200 mm _ 250 pi
Sample 195mm x
x = 244 pg
244 pg _
x
7.32 ng =
60 Ail 10 ng
x = 7.32 ng 13Ci2-2378-TCCD
This type of calculation is also used to determine
the % recovery of 13Ci2-OCDD.
2. Quantification of 2378-TCDD and TCDD
Isomers
EMSL-RTP
The concentrations of 2378-TCDD and TCDD
isomers in a sample extract are determined by
1-10
-------
comparing the labeled and native 2378-TCDD
peak height responses to those produced by the
external standard (previously shown in calcula-
tion of % recovery).
Example of calculation to determine ppt levels of
2378-TCDD in sample extract-
Standard 117 mm _ 1Qpg 2378-TCDD
Sample 133 mm x
x=11.4pg
11.4pg = 244_pg_l3C12-TCDD
x 10 ng 13Ci2-TCDD
x = 467 pg native 2378-TCDD
= 93 ppt
5 grams
This calculation and respective peak heights are
also used to determine the concentration of
TCDD isomers.
ECL-BSL
Equation 1: Response Factor (RRF) for native
2378-TCDD using 13Ci2-2378-TCDD as an inter-
nal standard.
RRFd = (A8C,8/A,,C,)
where: A8 = SIM response for 2378-TCDD ion
at m/z +322
Ais = SIM response for 13Ci2-2378-
TCDD internal ion at m/z 334
C,8 = Concentration of the internal
standard (pg//ul)
Cs = Concentration of the 2378-TCDD
Equation 2: Calculation of concentration of native
2378-TCDD using 13Ci2-2378-TCDD as internal
standard .
Concentration, pg/g - (A8) (l8)/(A,8)(RRFd)(W)
where: A8 = SIM response for 2378-TCDD ion
at m/z + 322
Ai8 = SIM response for the 13Ci2-2378-
TCDD internal standard ion at
m/z 334
ls = Amount of internal standard added
to each sample (pg)
W = Weight of sample in grams
RRFd = Relative response factor from
Equation 1
3. Calculation of Minimum Limit of Detection
(MLD) for 2378-TCDD
The minimum limit of detection is defined as the
amount of 2378-TCDD that will produce clearly
defined peak shapes for the masses m/z
320/322 in the proper isotope ratio and with a
signal to noise ratio greater than 2.5:1.0. The
sample weight, aliquot of extract used in analysis,
sample preparation efficiency, sample matrix
effects and noise, and the HRGC-HRMS sensitiv-
ity are variables that influence and determine the
MLD generated for each sample extract.
Example of calculation to determine the minimum
limit of detection utilizing the peak height
responses and pertinent data shown in calcula-
tion of % recovery.
2.5 x 2 mm baseline noise = 5.0 mm
instrument sensitivity for standard
10 pg- 0.09 pg/mm
117 mm
5.0 x 0.09 pg/mm = 0.45 pg 2378-TCDD
0.45 pg _ 244 pg 13Ci2-2378-TCDD
x 10ng 13C12-2378-TCDD
x = 18 pg 2378-TCDD
18pg
5 gram
sample
= 4 ppt MLD
J. Quantification of TCDFs, Penta Through Octa
CDDs andCDFs and Determination of Minimum
Limits of Detection
The calculations previously described are also
used in quantification of the other CDDs and
CDFs. An example is shown below:
Sample
m/z 334
m/z 340
m/z 342
m/z 374
m/z 376
Peak Peak
Height Height
Response Standard Response
200mm m/z 334 185 mm for
250 pg "C,r
2378-TCDD
190mm m/z 340 175 mm for
100 pg Penta-
CDF
125mm m/z 342 115 mm for
100 pg Penta-
CDF
4mm m/z 374 125 mm for
90 pg Hexa-CDF
2mm m/z 376 113 mm for
90 pg Hexa-CDF
aliquot analyzed, 2 fj\ from 60 fj\
10 ng 13Ci2-2378-TCDD added to 5 gram sample
prior to extraction.
amount of 13Ci2-TCDD in 2 fj\ aliquot analyzed
-------
185mm_ 250 pg 13Ci2-TCDD
200 mm x
x = 270 pg 13Ciz-TCDD
Penta - CDF
175 mm _ 100 pg penta-CDF x=109pg
190mm x
Figure 1.
Analysis of analytical standard (simultaneous
responses).
109pg _ 270pg13Ci2-TCDD
x = 4.04 ng
x 10 ng
4-04 ng - 808 ppt penta CDF
5 gm
Hexa-CDF
2.5 x 2 mm baseline noise = 5 mm
90 Pg = 0.72 pg/mm instrument
125mm sensitivity
5 x 0.72 pg/mm = 3.6 pg hexa-CDF
3.6 pg _ 270pg13Ci2-TCDD
x 10ng 13Cia-TCDD
x = 133 pg hexa-CDF
133pg_ =
5 grams
K. HRGC-HRMS Analytical Data
Some examples of typical analytical data are
shown in Figures 1 , 2 and 3.
1. Analysis of analytical standard.
2. Analysis of TCDDs in soil extract.
3. Analysis for hepta-CDDs and OCDD in soil
extract.
IX. Internal Quality Control/Quality
Assurance
Extraction, clean-up and analysis of samples will
be done in sets of twelve.
1. Labeled 13Ci2 or 37CL4-2378-TCDD in known
amo.unts, 2.5 to 10 ng, is added to each test
sample, QC sample and method blank prior to
extraction and clean-up in order to determine the
analytical methodology efficiency and for quanti-
fication purposes. 20 ng 13Ci2-OCDD will be
added to those samples that require analysis for
other CDDs "and CDFs.
Frequency: every sample.
2.
a. Method Blank: A blank extraction apparatus is
prepared in the laboratory and subjected to same
10
m/z320
m/z322
m/z 334
Figure 2. Analysis for TCDDs in Soil.
2378-TCDD
I
m/z 320
m/z 322
- m/z334
1-12
-------
Figure 3. Analysis for hepta CDDs and OCDD in soil.
m/z424
m/z426
m/z458
m/z460
m/z 470
10
sample preparation procedures as test samples.
Frequency: one every other extraction set, alter-
nate with matrix blank.
b. Matrix Blank: Portions of respective sample
matrix previously analyzed and known to be free
of TCDD contamination will be subjected to
sample preparation and HRGC-HRMS analysis.
Frequency: one every other extraction set, alter-
nate with method blank.
3.
a. Fortified Matrix Blank: Exact amounts of native
2378-TCDD and/or other compounds are added
to a known amount of the respective matrix
described above or to a specific environmental
sample previously analyzed.
Frequency: one every other extraction set, alter-
nate with control sample.
b. Control Sample: Portions of a respective sample
matrix previously analyzed and known to contain
2378-TCDD and/or TCDD isomers will be sub-
jected to sample preparation and HRGC-HRMS
analysis.
Frequency: one every other extraction set, alter-
nate with fortified matrix blank.
4.
a. Duplicate sample (lab): Two separate portions of
the same sample will be subjected to respective
sample preparation procedures and HRGC-HRMS
analysis.
Frequency: one every other extraction set, alter-
nate with composite duplicate(field).
b. Composite duplicate (field): Two separate por-
tions of the same field sample will be subjected to
respective sample preparation procedures and
HRGC-HRMS analysis.
Frequency: one every other extraction set, alter-
nate with duplicate sample (lab).
5. Solvent Blank: Aliquots of solvent will be injected
into HRGC-HRMS and analyzed.
Frequency: after every high level positive sample
analysis and/or as needed.
6. Analytical Standards: Concentration of working
standards will be compared with a primary
standard and fresh standards will be prepared as
needed.
7. Aliquots of analytical standards (13d2-TCDD,
2378-TCDD, etc.) used in fortification of test
samples and QA samples will be submitted along
with the extracts to the respective analytical
laboratories by the ECL. This concept provides a
common quantification base for all participating
laboratories.
8. Analytical standards used by ECL in the National
Dioxin Program will be validated and certified by
EMSL-RTP.
9. All sample extracts will be submitted to the mass
spectrometry laboratory in a "blind" fashion.
X. Quality Assurance/Quality Control
Requirements for Analytical Data
1. Each set of twelve samples must satisfy the
following criteria in order for it to be considered
valid data:
1-13
-------
Criteria
• Method efficiency achieved
for 13Ci2-2378-TCDD
• Method efficiency achieved
for 13Ci2-OCDD
• Analytical criteria used for
confirmation of 2378-TCDD
and other CDDs and CDFs
• Accuracy and precision
for 2378-TCDD, and other
specific CDDs and CDFs in
laboratory fortified sample
or control sample
• Method blank and matrix
blank free of CDD and
CDF contamination at
Requirements
50 to 1 20%
40 to 1 20%
Satisfies the
criteria
previously
stated
50 to 1 50%
@>6xMDL.
Accuracy
below6xMDL
may have
greater
variability
Described
below
and the corrective action that will be applied are
shown:
Problems
• If method blank or
matrix blank is
positive
• If detection limit for
blank is >the level of
2,3,7,8-TCDD in a
sample
• If spiked matrix or
reference sample is
outside QA for
2,3,7,8-TCDD
accuracy
• If method efficiency is
outside QA for sample
• If method efficiency is
outside QA for blank
Action
Reextract and reanalyze
blank and all positives
Reextract and reanalyze
blank and all positives
which are at levels below
the mid for the blank
Reextract and reanalyze
all positives
Reextract and reanalyze
sample
Reextract and reanalyze
blank and all positives
detection
Target Minimum
Limits of Detection'
2378-TCDD
Compounds
2378-TCDF
21 TCDD Isomers 37 TCDF Isomers
14Penta CDDs
10HexaCDDs
2 Hepta CDDs
OCDD
28 Penta CDFs
1 6 Hexa CDFs
4 Hepta CDFs
OCDF
Soil,
Sediments,
Fish, etc.
(ppt)
1-5
1-5
20-40
30-60
40-80
50-100
Water
(ppq)
10-40
10-40
20-40
30-60
40-80
50-100
'per isomer, which is based on previously described isomers
available for coinjection purposes.
2. Accuracy is defined as the degree to which the
analytical measurement reflects the true level of
2378-TCDD present in the sample. Laboratory
fortified matrices are used to determine accuracy.
3. Precision is defined as the measure of mutual
agreement among individual measurements for
a specific compound in a sample. Control
samples, field duplicates, duplicate and triplicate
of test samples split in the laboratory will be used
to determine precision.
4. Method efficiency, percent recovery, is deter-
mined by dividing the amount of labeled TCDD
and/or OCDD present in the extract by the
amount that was added prior to extraction.
5. Corrective Actions
Certain types of problems can be encountered in
the evaluation of analytical data achieved for a
set of twelve sample extracts. Potential problems
• If method efficiency is
outside QA for
fortified matrix or
reference sample
• If 2,3,7,8-TCDD level
is outside QA for
duplicates
• If 2,3,7,8-TCDD level
exceeds calibration
standard range
• If GC resolution is
outside QA for isomer
specific analysis
• If GC relative
retention time
between 13Ci22,3,78-
TCDD and/or "CU
2,3,7,8 and 2,3,7,8-
TCDD are outside QA
for fortified matrix
or reference sample
Reextract and reanalyze
spiked matrix
Reanalyze duplicates
Reextract and reanalyze
with smaller portion of
sample, or extend
calibration range with
additional standard
Reanalyze complete set on
alternate column
Reanalyze complete set on
alternative column
6. Completeness is defined as the percentage of
valid (meets all quality assurance requirements)
data compared to the total number of samples
analyzed. For the National Dioxin Study the
"TROIKA" will not release to the Office of Water
any data that has not been reviewed by the
"TROIKA" to assure validity of data. Therefore,
completeness is defined as 100% for this study
and will not otherwise be calculated.
7. Representativeness is dependent on the sam-
pling plan, and, therefore, is not covered in this
plan. Assessment of representativeness by eval-
1-14
-------
uation of blind coded analyses of field duplicates
is the charge of the regional laboratory dioxin
coordinators and not the "TROIKA."
8. Comparability is defined as the extent to which
the sample results can be verified or duplicated
by another independent laboratory or compared
against results previously found. No interlabora-
tory studies are currently planned by the Office of
Water for the TROIKA labs to participate in.
However, the TROIKA labs will routinely ex-
change blind coded samples between themselves
and Dr. Mike Gross, University of Nebraska,
Lincoln, Nebraska, who is currently working with
the TROIKA under an EPA cooperative agree-
ment.
XI. Capabilities, Qualifications, and
Limitations of Methods of Analysis for
Tetra Through Octa CDDs and CDFs
The previously described limited supply of labeled
and native CDD and CDF analytical standards
and the qualitative standards provide the base
required for:
• Determination of method efficiency for tetra
through octa CDDs and CDFs.
• Identification and quantification of 2378-TCDD
in the presence of the other 21 TCDD isomers.
• Identification and quantification of each of
the other 21 TCDD isomers.
• Identification and quantification of 2378-TCDF
in the presence of the other 37 TCDF isomers.
• Identification and quantification of the 2 hepta
CDD isomers, 4 hepta CDF isomers, OCDD
andOCDF.
• Establishing the specific retention time
windows and identification and quantification
of TCDFs, penta and hexa CDDs and CDFs.
Qualifications—
• The 2378-TCDD analytical standard is used
for quantification of the other 21 TCDD
isomers.
• 2378-TCDF and the specific penta through
octa CDD and CDF isomers previously de-
scribed are used for quantification of the
respective CDD and CDF isomer series.
• The response factor, sensitivity of other
isomers and the extraction and clean-up
efficiency for other isomers may not be identi-
cal to those reference isomers on hand.
• Therefore, the concentration of TCDD, TCDF,
penta through hexa CDD and CDF isomers
should be considered to be a "semi-quanti-
tative" estimate. Additional labeled and native
CDD and CDF isomers are needed in order to
provide more accurate and valid quantification
of these compounds.
Limitations—
• Conclusive and valid isomer assignment of all
TCDFs, penta through hexa CDDs and CDFs
cannot be made with the limited supply of
reference compounds available today.
• Concentrations are reported as total TCDFs,
penta CDDs and CDFs, hexa-CDDs and CDFs.
Most of these qualifications and limitations will
be removed after additional reference standards
are made available.
XII. External Quality Assurance/Multiple
Laboratory Precipitation
1. Specific test samples and QA samples 2A, 2B,
3A, 3B, 4A and 4B described under the Internal
QC/QA will be submitted to other validated
laboratories such as the ERL-D, University of
Nebraska, Wright State University and University
of Umea for sample preparation and analysis.
Frequency: multiple laboratory participation,
numbers, etc. will be worked out and
determined at a later date.
2. Extracts and/or "split extracts" from the labora-
tories above will be submitted to EMSL-RTP for
HRGC-HRMS analysis.
Frequency: as needed for confirmation/valida-
tion purposes.
3. Private sector laboratories from the EPA certifi-
cation program will initially receive a very large
percentage (30 to 50%) of QA samples in their
shipment of test samples in order to demonstrate
and validate their analytical capabilities and
results.
4. Extracts and/or "split extracts" from 3 above will
be submitted to EMSL-RTP for HRGC-HRMS
analysis.
Frequency: as needed for confirmation/valida-
tion purposes.
5. Analytical standards will be shipped along with
extracts to insurethat all laboratories will have a
common base for quantification purposes.
XIII. Analytical Report Format,
Reporting of Data and Review Panel
Method efficiency will be reported as the percent
recovery of 13Ci2-2378-TCDD and also 13d2-
OCDD if it was used. CDDs and CDFs will be
reported in ppt on weight basis. Compounds not
detected will be reported as ND with minimum
limit of detection in parenthesis, for example ND
(2). A text that describes and summarizes the
analysis and analytical results will accompany
each report that is submitted to the Director,
1-15
-------
ERL-D. ERL-D will submit the complete report,
text and results, to the region or state at a later
date. Examples of typical reports and a table of
results for 2378-TCDD and for tetra through octa
CDDs and CDFs are shown in Figures 4, 5 and 6.
The data and other pertinent information will be
subjected to a review by the participants and
experienced scientists at the end of each quarter.
Dr. Jaworski will submit the review and validated
data to HQ by 20 April, 20 July, 20 October and
20 January following each quarter.
The raw data, mass spectrum, chromatograms
and calculations remain in the analytical labora-
tory that performed the analyses and are filed for
reference.
Review Panel: An independent review panel
consisting of recognized experts in HRGC-HRMS
analysis of dioxins will be formed to evaluate the
effectiveness of the analytical method and QA
program detailed in this plan. Members of this
panel will include representatives from EPA ERL-
Duluth, an independent commercial laboratory,
and a third party, yet to be determined.
XIV. Sampling Procedures
Sampling protocol and field QC procedures are
detailed in a separate document prepared by OW.
All field sample containers will be pre-cleaned
with alconox, denatured alcohol and trichloro-
ethylene. Samples will be kept iced at 4°C during
storage and transport to the ECL. No holding time
has been established for soils, but samples in
this study will be sent to the laboratory as soon as
possible after homogenization and extracted
within 30 days. Samples will be kept refrigerated
during storage and extracts will be stored in
capillary tubes and kept frozen. Once soil samples
have been homogenized and split, CDD/CDF
aliquots will be kept at ambient temperature in
the dark. Others will be refrigerated.
XV. Sample Custody
Samples and accompanying information under
chain of custody are submitted to the ECL or
ERL-D laboratory by the regions for sample
preparation and analysis. Extracts of test samples
and QA samples are then coded and shipped by
Federal Express to respective analytical labora-
tories under a chain of custody that has been
used since 1975. An example of the chain of
custody form is shown below. The form is
completed, copied and the original is returned to
the submitting laboratory. Pertinent information
required by the analytical laboratory (sample
weights, type, labeled TCDD fortification level,
etc.) are also provided in an accompanying
memorandum.
NOTE: All shipments of extracts must conform to
DOT specifications and are shipped by commer-
cial cargo planes, such as Federal Express.
CHAIN OF CUSTODY RECORD
Date Samples Shipped-
Samples Shipped From
(Signature, Affiliation)
Samples Received by
(Signature, Affiliation)
Date Samples Received.
Condition of Samples
How Samples Transferred-
Samples Transferred:
1-16
-------
Figure 4. Analytical Report Form.
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Research Triangle Park, NC 27711
Date October 5, 1 983
Subject Analyses for 2,3,7,8-TCDD Residues
From:
To: Director, ERL-D
Text: Description and summary of results, TCDD isomers, qualifications, etc
Sample
ID
Sample
TYPE
Sample
Weight
(9)
13Cii-TCDD
Fortification
Level (ng)
13C12-TCDD
% recovery
ppt Levels of
2,3,7,8-TCDD
Detected/Minimum
Limit of Detection
(MLD)
ECL
Entry
QA
PPT
RTP-1"
soil8
10a
2.5"
Example
88"
ND(1)b
"information supplied with extract
experimental results
Column Description
Retention Time of 2,3,7,8-TCDD.
Figure 5.
Table 1 (Set 2 thru Set 6)
Analytical Results for 2378-TCDD in Soil Samples
Sample
ID
13312
13313
13314
13315
13316
13317
13318
13319
13320
13321
13322
13323
Sample
Weight
(g)
10
10
10
10
10
10
10
10
10
10
10
10
Set 2"
Method Efficiency.
% Recovery of
5 ng 37CI4-TCDD
96
92
108
72
100
86
94
92
50
80
100
78
2378-TCDD
Detected (ppt)
26
13
12
17
18
26
ND
28
12
2
ND
29
2378-TCDD
Minimum Limit
of Detection (ppt)
3
1
1
2
1
1
1
1
3
1
1
2
ECL
Entry
QA
PPT
ND - not detected at specified minimum limit of detection.
•received at EMSL-RTP, June 22, 1984.
NOTE: 13C12-TCDD has replaced "C4-TCDD.
1-17
-------
Figure 6.
Table 2. Analytical Results for CDDs and CDFs in Soil Samples (Continuation)
Compounds
% Recovery of
5 ng 13Ci2-TCDD
% Recovery of
20ng'3Ci2-OCDD
1368-TCDD
1379-TCDD
1378-TCDD
1369-TCDD
1 247-TCDD
1248-TCDD
1268-TCDD
1478-TCDD
2378-TCDD
1234-TCDD
1237-TCDD
1238-TCDD
1246-TCDD
1249-TCDD
1236-TCDD
1279-TCDD
1278-TCDD
1279-TCDD
1289-TCDD
2378-TCDF
Penta-CDDs
Penta-CDFs
Hexa-CDDs
Hexa-CDFs
Hepta-CDDs
Hepta-CDFs
OCDD
OCDF
Sample I.D., Method Efficiency, Concentrations Detected and Minimum Limits of Detection (ppt)°
I.D.s 13398
75
89
ND
ND
ND
ND
ND
ND
ND(4)
ND
ND
ND
ND
ND(6)
ND(27)
ND(8)
31(27)
35(20)
68(21)
ND(27)
126(22)
ND(44)
13399
80
90
95
62
25
19
9
7
103(18)
19
14
ND
ND
34(7)
268(91)
321(42)
2400(91)
939(40)
38000(1600)
5600(240)
120000(1900)
5600(780)
13400
98
95
ND
ND
ND
ND
ND
ND
4(2)
ND
ND
ND
ND
ND(4)
ND(44)
ND(1.6)
72(44)
ND(36)
200(167)
ND(23)
10600(233)
ND(56)
13401
90
72
ND
ND
ND
ND
ND
ND
ND(4)
ND
ND
ND
ND
ND(4)
ND(24)
ND(8)
ND(24)
ND(22)
1 50(24)
ND(31)
340(26)
ND(51)
13406 13412
100 100
70
105
76
40
20
13
ND
3500(39) 271(7)
26
39
ND
ND
450(60) 27(7)
236(67)
895(137)
4040(67)
3092(130)
75000(900)
1 5400(383)
375000(1300)
8580(475)
'Qualifications stated in text.
1-18
-------
Section II
Analytical Procedures and Quality Assurance Plan for the National Dioxin Study
Environmental Research Laboratory
U.S. Environmental Protection Agency
6201 Congdon Boulevard
Duluth, Minnesota 55804
Introduction
This section contains the analytical methods and
quality assurance procedures to be used at ERL-D for
theanalysisof2,3,7,8-TCDD in environmental tissue
and water samples. The samples will be a part of the
National Dioxin Study, Tiers 3-7 of the National
Dioxin Strategy. ERL-D will join EMSL-RTP and ECL-
BSL Labs of the U.S. EPA to form a National Dioxin
Study Trace Analysis Group, referred to as the
TROIKA.
Although specific analytical procedures will vary
somewhat between the three labs, the quality
assurance protocol will be consistent and controlled
by internal checks. A separate Analytical Procedures
and Quality Assurance Plan has been developed for
EMSL-RTP and ECL-BSL. Our entire QA/QC plan
(December 1983) has been peer reviewed by scien-
tists within and without the Agency. Their comments,
along with information developed by the TROIKA
through further methods development work, have
been reviewed and incorporated into this draft when
appropriate. A list of TROIKA responses to reviewer
comments have been forwarded to both the review-
ers and the Office of Water.
The methodology described in this section is to serve
only as a guideline for the methods of analysis.
Changes in specific methods may be required due to
changes in reagents, instrumental parameters, sam-
ple matrix, etc., however, all guidelines for quality
assurance will be met before data is released.
A diagram showing the involvement of the TROIKA in
the development of sampling/analytical protocols,
sample traffic flow, analyses of samples, data review,
and data reporting is presented in the section
described by ECL/BSL and EMSL/RTP.
I. Sample Preparation for 2,3,7,8-TCDD
Analyses
A. Grinding, Blending, and Storage of Samples
1. Biological Tissue
In some instances, samples received at ERL-D
will be pre-ground at the Region or state labs.
Samples to be ground at ERL-D will be ground in
a stainless steel meat grinder by putting the
sample through three times. Specific organ or
whole animal analysis will be determined by the
Region or state submitting the sample. After
grinding, all samples will be stored in a solvent-
rinsed glass jar fitted with an aluminum seal at
2. Water
Water will be stored, in the container in which it
is received, in the dark at 4°C.
B. Extraction Procedures
1 . Biological Tissue
Subsamples (20g) of previously ground, homog-
enized, and frozen biological tissue are blended
in a 400 mL beaker with enough anhydrous
sodium sulfate (~ 60g) to dry the sample. Two-
thirds of the dried sample is transferred to a
course-frit Soxhlet extraction thimble. The sam-
ple is spiked with 100 /uL each of 5 pg///L 13Ce
1,2,3,4-TCDD and 1 pg/yuL 1,2,3,4-TCDD in iso-
octane and the remaining sample is added to the
thimble. The sample is extracted for 24 hours in
a Soxhlet extraction apparatus fitted with a 500
mL round bottom flask using 250 mL of a 1:1
mixture of hexane and methylene chloride. The
sample is then quantitatively transferred to a
500 mL Kuderna-Danish (KD) apparatus and 5
mL of iso-octane is added. The apparatus is then
fitted with a 3-ball Snyder column and the
volume is reduced to 5 mL on a steam bath. The
extract is sealed and stored in a freezer.
2. Water
Water (900 mL) is added to a solvent rinsed
separatoryfunnel(IOOOmL), spiked with 100/yL
of 5 pg/fjL 13C6 1,2,3,4-TCDD in dimethylform-
amide and shaken for 5 min. Methylene chloride
(1 20 mL) is then added to the separatory and the
water is vigorously extracted. The phases are
allowed to separate and the methylene chloride
2-1
-------
(at least 100 ml_) is drained through a funnel of
anhydrous sodium sulfate into a Kuderna-Danish
apparatus. The water is extracted twice more
with methylene chloride (2 x 100 mL), and the
extracts along with iso-octane (5 mL) are added
to the Kuderna-Danish apparatus. The extract is
reduced in volume to 5 mL, sealed, and stored in
a freezer.
C. Isolation Procedures
1. Reagents
a. Solvents
Only pesticide grade distilled-in-glass solvents
are to be used. They are: hexane, iso-octane,
methylene chloride, benzene, toluene, acetone,
and methanol (Burdick and Jackson).
b. Sodium Sulfate
Sodium sulfate (Baker reagent grade anhydrous)
is Soxhlet extracted eight hours with 1:1 hexane/
acetone, air dried for 12 hours, and vacuum oven
dried (100°C) for 24 hours. The sodium sulfate is
then baked at 600°C in a furnace for 24 hours,
cooled, and stored in an empty hexane solvent
bottle.
c. Alumina
Basic alumina (Merck-Darmstadt) is Soxhlet
extracted eight hours with methanol, air dried for
12 hours, and vacuum oven dried (100°C) for 24
hours. It is then stored in an empty hexane
solvent bottle. It is then activated at 225°C for 24
hours and stored at 105°C until used.
d. Silica-Gel
Silica-Gel-60 (Merck-Darmstadt), is Soxhlet ex-
tracted eight hours with methanol, air dried for
12 hours, and vacuum oven dried (105°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.
e. Sulfuric Acid/Celite
Sulfuric acid (Baker Chemical Company, Ultrex)
(5 mL) is blended in a 250 mL beaker with Celite
545 (Baker) 10g.
f. Potassium Silicate
High purity potassium hydroxide (Aldridge Chem-
ical Co.) (56 g) is dissolved in 500 mL methanol.
Silica-Gel (100 g) is added to the mixture and
refluxed with stirring for 1 hour. Cool the
mixture, drain the solvent, and transfer the solids
to a Soxhlet thimble. Extract with methanol for 4
hours, cool, air dry, and place in a hexane solvent
bottle until use.
g. Silica Gel/Carbon
Silica gel (100 g) (Merck-Darmstadt) is Soxhlet
extracted with methanol (200 mL) for 24 hours,
air dried, and activated at 1 50°C for 24 hours. To
a portion of silica gel (20 g), add Amaco PX-21
carbon (500 mg) (carbon courtesy of Dr. David
Stalling, U.S. FWS, Columbia, MO) and blend
until uniform color.
2. Xenobiotic-Biological Molecule Separation
The sample extract in 5 mL iso-octane is quanti-
tatively transferred to a 30 cm x 2.5 cm glass
chromatography column fitted with a 300 mL
reservoir on top and a course glass frit on the
bottom, which has been packed with (bottom to
top) 2 g silica gel, 2 g potassium silicate, 2 g
sodium sulfate, 10 g celite/sulfuricacid and 2 g
sodium sulfate, and washed with 100 mL hexane.
The column is eluted with 200 mL hexane into a
KD apparatus. Iso-octane (1 mL) is added to the
KD and the sample is concentrated to 0.5 mL for
PCDD separation.
3. Isolation of 2,3,7,8-TCDD from other Xeno-
biotics
a. Alumina Chromatography
Alumina (1.5 g) is packed into a disposable
pipette, washed with methylene chloride (4 mL),
allowed to drip dry, and activated at 225°C for 24
hours prior to use. The sample extract in 0.5 mL
iso-octane is transferred to the cooled alumina
column previously washed with hexane (1 mil).
The column is eluted with 6 mL carbon tetra-
chloride and a fraction containing 2,3,7,8-TCDD
is eluted with methylene chloride (4 mL).
b. Silica Gel/Carbon Chromatography
The methylene chloride fraction from the alu-
mina column procedure is transferred to a 4 mm
x 200 mm column containing 200 mg silica
gel/carbon. The column is eluted with methylene
chloride (15 mL) and benzene/methylene chlor-
ide (1:3, 1 5 mL) in the forward direction of flow,
and toluene (7 mL) in a reverse direction flow.
The toluene fraction is collected and reduced in
volume toO.5 ml usinga micro-Snyder apparatus
on a heating block. The final extract is stored in a
freezer until mass spec analysis.
II. GC/MS Parameters
A. Quantification Standards
Standards of 2,3,7,8-TCDD are to be provided by
ECL-BSL. Standards of 13C6 1,2,3,4-TCDD and
13Ci2 1,2,3,4-TCDD are purchased from Cam-
bridge Isotope Inc., Cambridge, MA, in 50/ug/mL
2-2
-------
solutions. All standards are verified for use by
EMSL-RTP. 13C6 1,2,3,4-TCDD will be used as
the internal standard spiked into the sample at
the beginning of extraction. 13Cia 1,2,3,4 will be
used as the internal standard spiked into all
samples at the end of the clean-up procedure. All
analytical standards contain an equal amount of
13Ci2 1,2,3,4-TCDD with varying amounts of 13C6
1,2,3,4-TCDD and 2,3,7,8-TCDD. By using re-
sponse factor ratios to calculate concentrations
(see section of quantification procedures), the
percent recovery of the overall clean-up proce-
dure, and the level of 2,3,7,8-TCDD can be
determined.
Standard
3C,2 1,2,3,4 13C6 1,2,3,4
2,3,7,8
A
B
C
D
25 pg/fjL
25 pg//jL
25 pg///L
25 pg/AiL
2.5 pg/fj\.
12.5pg//uL
25 pg///L
25 pg/fjL
2.5 pg///L
1 2.5 pg/fjiL
25 pg/AiL
50 pg//J\-
B. GC/MS Column Performance Standard
A column performance dioxin mixture has been
prepared by ERL-Duluth with all 22 TCDD
isomers. This standard contains equal amounts
of all TCDD isomers except 2,3,7,8-TCDD but
includes 37CU 2,3,7,8-TCDD. The column per-
formance standard is evaluated by MIS monitor-
ing of ions m/z 320 and 328.
C. GC/MS Instrumentation Parameters
Two GC/MS systems will be used for dioxin
analysis at ERL-D, a Finnigan-MAT 4500 MS/
INCOS DS high resolution GC/low resolution
MS system, and a Finnigan-MAT 8200 MS/
SS300 DS high resolution GC-high resolution
MS system. Both instruments will be used in the
multiple-ion-selection mode (MIS). Tuning of
these instruments will, therefore, optimize the
signals for 2,3,7,8-TCDD (m/z 320 and 322).
Ions to be monitored are shown in the table
below.
Mode
13C12 1,2,3,4 13C6 1,2,3,4
2,3,7,8
Low Resolution 334 328 320/322
High Resolution 333.9338 3279166 319.8965/321.8936
Confirmation of 2,3,7,8-TCDD by high resolution
MS using five ions (shown in the table below) will
be done for 10% of all positive analyses, whether
originally analyzed by low or high resolution
mass spectrometry.
Ion
M
M+2
M+4
M-CICO M+2-CICO
Mass 319.8965
Ratio .67 - .87
321.8936 323.
1 .00 .39 -
89
.59
256.9327
At least
25% M
258.9298
At least
25% M
The 4500 MS should be tuned and mass
calibrated each day, and the 8230 MS should be
tuned each day. A complete set of analytical
standards will be run when necessary to estab-
lish sensitivity and linearity. Subsequently, a
limited subset of standards will be run each day
to establish instrument performance. Immed-
iately before analysis, each sample is spiked with
10 /uL of 50 pg/M- 13Ci2 1,2,3,4-TCDD, and the
volume adjusted to 20 /uL. Typical instrument
operating parameters are given in Appendices A
and B.
D. Quantification Procedure
1
Percent Recovery 13C6 1,2,3,4-TCDD
The percent recovery of 13C6 1,2,3,4-TCDD is
determined by calculating the amount of 13Ce
1,2,3,4 present in the final sample extract and
dividing by the amount spiked into the sample at
the start of the clean-up procedure. This is done
by using the mean (3 values) relative response
factor for the ion response ratios between 13Ci2
1,2,3,4 and 13C6 1,2,3,4 (Eq 1) and using this
response factor to calculate the concentration in
the final solution (Eq 2). Concentration in the
final solution times the final volume equals the
total amount present.
Standard
RF 328/334 =
Sample
C328
A328 x C334
A334 x C328
A328 x C334
A334xRF 328/334
Eq1
Eq2
Where: RF 328/334 = response factor for ions
328 and 334.
A328 = Area for peak of m/z 328
A334 = Area for peak of m/z 334
C328 = Concentration of 13CB 1,2,3,4-TCDD
C334 = Concentration of 13Ciz 1,2,3,4-TCDD
2. Quantification of 2,3,7,8-TCDD
The quantification of 2,3,7,8-TCDD is determined
by calculating the mean response factor between
2,3,7,8-TCDD and 13C6 1,2,3,4-TCDD from each
standard (Eq 3), and using the response factor to
calculate the level of 2,3,7,8-TCDD (Eq 4).
Standards
RF322/328 =
Sample
V322
A322 x C328
A328 x C322
A322 x S328
A328xRF 322/328
Eq3
Eq4
2-3
-------
Where: RF 322/328 = response factor for ions
322 and 328.
A322
A328
C322
C328
S328
Area for peak of m/z 322
Area for peak of m/z 328
Concentration of 2,3,7,8-TCDD
Concentration of 13Ce 1,2,3,4-TCDD
Spike level of 13C6 1,2,3,4-TCDD
(pg/g)
V322 = Final level of 2,3,7,8-TCDD in original
sample (pg/g)
3. Determination of Minimum Detection Limit
(MDL)
The MDL for each sample will be statistically
calculated according to procedures outlined in*
Anal. Chem. 55, 842-847 (1983) and "Confi-
dence Limits for Isotope Dilution-Gas Chroma-
tography/Mass Spectrometric Determination of
2,3,7,8-Tetrachlorodibenzo-p-dioxin in Environ-
mental Samples" in Chlorinated Dioxins and
Dibenzofurans in the Total Environment, Eds. G.
Choudhary, L. Keith, and C. Rappe, Butterworth
Press, 1983. Those techniques are not currently
available to the TROIKA, however, OW has
contracted with Finnigan Corporation to imple-
ment the appropriate software into the SS-300
and Incos data systems used by TROIKA labs. In
the interim, all data will be generated in a format
appropriate for reanalysis when the statistical
package is implemented. In addition, the MDLfor
biological tissue will be determined by two
methods. First, a MDL will be established for
each sample using a "surrogate analyte present"
technique; and second a "verified" MDL will be
established for a representative subset of each
matrix type (muscle, walleye, duck, etc.).
4. Surrogate Analyte Present Technique
The target detection limit for biological tissue is 1
pg/g. Since it has been established that biolog-
ical tissue contains only 2,3,7,8-TCDD and not
the other 21 TCDD isomers, a "surrogate ana-
lyte" can be added to each sample near the target
detection limit to aid in establishing the 2,3,7,8-
TCDD MDL. The method used is presented below
If 2,3,7,8-TCDD is Not Detected Then:
Measure the peak height of the 5 pg/g surrogate
analyte, 1,2,3,4-TCDD. Determine 1 standard
deviation of noise through the retention time
window of 2,3,7,8-TCDD, and calculate a signal
level 2.5X noise. (Both of these measurements
are computer controlled commands.) Using a
response factor between 1,2,3,4-TCDD and
2,3,7,8-TCDD of 1.0 (previously determined)
calculate MDL by:
MDL =
1250 counts peak height
(5000 counts peak height)
(5 pg/g) = 1.3 pg/g
If 2,3,7,8-TCDD is Present in the Sample Then:
Measure the peak height of the quantified
2,3,7,8-TCDD in the sample.
5 pg/g 1234-TCDD Surrogate Analyte
Example: 2378-TCDD quantified
at 4.5 pg/g
peak height = 1000
Next, measure the standard deviation of noise
through the retention time window of 2,3,7,8-
TCDD in the "instrument blank" solvent injection.
5 pg/g 1234-TCDD Surrogate Analyte
standard deviation of noise = 150
mean noise
elution window 2378-TCDD
elution window
1234-TCDD
5 pg/g 1234-TCDD Surrogate Analyte
peak height 5000 counts
2.5 x noise = 1250 counts
standard deviation of noise = 500
/mean noise "
elution window 2378-TCDD
The MDL is then calculated by the following
equation:
Standard Deviation of Noise Peak Height
MDL = 2.5 x ~
peak height of quantified 2,3,7,8-TCDD
x level of 2,3,7,8-TCDD
Example:
2.5 [ 150/1000 ] 4.5 = 1.7 pg/g
2-4
-------
5. Representative Matrix Technique:
As one of the QA samples analyzed with each
sample set, one sample previously analyzed as
"not detected" will be spiked at 5X, the target
detection limit, with 2,3,7,8-TCDD before being
routinely processed for 2,3,7,8-TCDD quantifi-
cation. This sample will not be used to evaluate
QA/QC for each sample set, however, a record of
quantification of 2,3,7,8-TCDD in the spiked
sample, and MDL in each of the spiked and non-
spiked samples will be kept at ERL-D for evalua-
tion during the course of the study.
III. Quality Assurance/Quality Control
A. General Procedures of Operation
Analysis of samples will be done in sets of
twelve.
1. Method Blank: A blank extraction apparatus is
prepared in the laboratory and subjected to the
same sample preparation procedures as test
samples; or a Matrix Blank: A sample previously
analyzed and known to be free of TCDD contam-
ination (at a level below the accepted detection
limit for that sample matrix). The Method Blank
and Matrix Blank will be used in alternate sample
sets.
2. Fortified Matrix: Exact amounts of native 2,
3,7,8-TCDD and/or other compounds to be
analyzed for are added to a sample previously
shown not to be contaminated. This sample may
be substituted for a reference sample that has
been analyzed by at least three labs and a mean
value of contamination has been established.
The level of contamination of 2,3,7,8-TCDD in
the matrix spike will be at least 10X the target
detection limit.
3. Detection Limit Verification Sample: One sam-
ple from a previously analyzed sample set will be
spiked with 2,3,7,8-TCDD at 2.5X the target
detection limit and analyzed with the next
sample set. The addition of this QA/QC sample
will be done for only the first three sample sets of
any matrix type to establish that the calculated
MDL is achievable. If analytical results show
difficulty in obtaining the MDL, then this QA/QC
sample must be in each set. If no problem is
experienced, then this QA/QC sample may be
dropped.
4. Duplicate Sample: Two separate portions of the
same sample will be processed and analyzed.
5. 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.
6. Sequence of Logging and Labeling Samples:
a. Logging Incoming Samples:
The Sample Control Center (SCC) notifies ERL-D
when samples have been shipped. Upon arrival,
the samples are checked to make sure they are in
good condition and the Dioxin Shipment Records
are complete. ERL-D informs SCC that samples
arrived safely or if there were any problems with
the samples (example: mislabeled, no species
identification). The samples are then put in a
freezer until they are homogenized. After a
sample is homogenized, it is put into the double-
door freezer. Samples that have been extracted
are put into the single-door freezer. After accept-
able data is generated for a sample, it is brought
to a locker plant for permanent storage. A Locker
Plant Log is kept with episode # and SCC # of all
National Dioxin Study samples stored there.
A computer program has been developed for
sample tracking and data storage (see Appendix
D). The episode #, SCC #, date sample received,
and matrix type are entered into the Dioxin
Survey Sample Log and the computer. The
following additional information is also entered
into the computer: analysis lab, latitude, long-
itude, and site location.
b. Logging and Labeling Samples During Prepara-
tion:
Since only one set of 9 samples and 3 QA
samples is started each day, a lab ID number is
assigned so that the samples go to the MS lab
blind coded. A letter, A through L, is used for the
individual samples followed by the date the set
was started and the initials of the person doing
the prep work, e.g., A053084CS.
As each sample is weighed, the assigned lab ID
#, episode #, SCC #, matrix description, weight of
sample, and amounts of the surrogate spike and
internal standard are entered into the log book
and onto the Dioxin Study data sheet (see
Appendix E). The lab ID # is written on label tape
and placed on the beaker containing the appro-
priate sample before weighing another. Soxhlet
thimbles with respective beakers are placed in a
row. Transfer from the beaker into the thimble
occurs in alphabetical order and the empty
beaker is placed behind the thimble. After the
thimble is placed in the Soxhlet, the label tape is
immediately put on the round bottom flask.
During each transfer throughout the procedure,
the label is transferred to the receiving vessel.
Before the final transfer, the lab ID # is written on
the microvial with a permanent marker. The lab
ID # is recorded in the GC/MS sample log book
along with the GC/MS file number at the time of
analysis.
2-5
-------
7. Format: All samples will be blind coded by the
sample preparation chemist and analyzed as
such by the mass spectrometry lab. Once during
each working day, acquire data on an injection of
solvent to obtain an "instrument" blank. ECL-
BSL and ERL-D will also routinely exchange blind
coded samples to verify quality of data. The
schedule for this will be approximately every six
months.
B. Instrumentation
Typical instrumental parameters are shown in
Appendices A and B.
1. Maintenance
A rigorous maintenance schedule of the gas
chromatograph, mass spectrometer, and com-
puter maintenance has been established for all
GC/ MS/COM systems at ERL-D. This schedule
is closely followed and all maintenance is logged
into proper log books. Corrective measures are
taken immediately upon identification of an
instrument malfunction.
2. Quality Control Parameters
a. GC Column Performance
The ability of the GC column to resolve 2,3,7,8-
TCDD from the other TCDD isomers will be
evaluated on an SP2230 column using the GC
Column Performance Standard when 2,3,7,8
needs to be confirmed. The ion current profile of
m/z 328 must be resolved from non-2,3,7,8-
TCDD isomers(m/z322) eluting both before and
after the 328 peak by a resolution of 0.75 (87.5%
resolved) or greater.
Method:
W1
Quality A ssurance:
R>0.75
Quality Control:
If resolution fails, the column must be either
cleaned or replaced, and samples reanalyzed
with a column that meets resolution require-
ments.
In addition, the absolute retention time difference
between 13C6 1,2,3,4-TCDD and 2,3,7,8-TCDD
must be checked throughout the analysis set. For
2,3,7,8-TCDD retention time confirmation, 37CU
2,3,7,8-TCDD will be added to the sample at 25
ppt.
Method:
Calculate the absolute time (sec.) difference
between 13C6 1,2,3,4 and 2,3,7,8, or 37C4
2,3,7,8 and 2,3,7,8 for standards and samples.
Quality A ssurance:
The retention time change between standard
and sample should not exceed 3 sec.
Quality Control:
Examine GC for leaks, perform routine main-
tenance on GC, etc., to correct the problem.
Verify proper performance before starting
sample analysis.
b. Mass Spectral Qualitative Identification of
2,3,7,8-TCDD
Initial MIS analysis will be done by monitoring
m/z 320 and 322 (M and M + 2 respectively) and
confirmation will be done by also monitoring m/z
324, 257, and 259 (M + 4, M - COCI, and M + 2
-COCI respectively) at or above 10,000 resolution.
Method:
Multiple Ion Selection Gas Chromatography-
Mass Spectrometry.
Quality Assurance:
Ion intensity ratios for the molecular ion is M =
0.67 - 0.87, M + 2 = 1.0, M+4 = 0.39 - 59. M
-COCI and M - COCI must be 25% < M <
100%.
Quality Control:
If all ions do not meet intensity ratio require-
ments, identification is rejected.
c. Mass Spectral Quantitative Analysis
After 2,3,7,8-TCDD has been qualitatively ident-
ified in a sample, it must be quantified. The
method of quantification is the method of using
peak areas and response factors. The analytical
2-6
-------
standards used allow the calculation of a re-
sponse factor for 2,3,7,8-TCDD: 13C6 1,2,3,4-
TCDD at a high, medium, and low signal level.
The response factor used for the quantification of
2,3,7,8-TCDD in environmental samples will be
the mean value of all measurements at these
three levels. Peak areas will be computer soft-
ware determined.
Method: See Section II.D.
Quality Assurance: See Section III.B.
Quality Control: See Section III.B. In addition,
samples in which the level of 2,3,7,8-TCDD is
greater than the highest calibration standard
will be reanalyzed with a smaller sample size.
d. Evaluation of Detection Limits for 2,3,7,8-TCDD
The method to be used by the TROIKA for the
evaluation of the limits of detection is a statistical
evaluation of signal levels described by Dr.
Ronald Mitchum, U.S. FDA, Jefferson, AK. The
software to do this will be provided to the TROIKA
by Finnigan Corp. through a contract with U.S.
EPA, Office of Water. The QA/QC for the M.D.L.
for biological tissue used at ERL-D is described.
Method: See II.D.3
Quality Assurance Requirements: The signal
level for 2,3,7,8-TCDD in all positive samples
must be at least 2.5X the background noise
level. The matrix blank and lab blank shall not
contain detectable levels of 2,3,7,8-TCDD.
Quality Control: If the matrix blanker laboratory
blank contain detectable levels of 2,3,7,8-
TCDD, all samples in that set that contain
detectable levels of 2,3,7,8-TCDD will be
reanalyzed in the next sample set. In addition,
appropriate measures must be taken to identify
the source of contamination and eliminate the
contamination.
C. Evaluation of Data
All data generated by the TROIKA will meet all
quality assurance requirements, and be reviewed
by the TROIKA before being released to the
Office of Water. In addition to all QA require-
ments listed in sections above, data must also
meet guidelines of accuracy and precision dis-
cussed below. A summary is shown in the table
below.
Parameter Target Detection Limit Accuracy* Precision*
Biological Tissue 1 pg/g 50-150% 50-150%
Water 10fg/g 50-150% 50-150%
"at >6 target detection limit.
1. Accuracy of 2,3,7,8-TCDD Quantification
Accuracy, the degree to which the analytical
measurement reflects the true level present, will
be evaluated in two ways for each sample set (8
samples and 4 QA). These are: (1) the variance of
2,3,7,8-TCDD spiked into a blank sample matrix,
and (2) the recovery of the internal standard 13Ce
1,2,3,4-TCDD for each sample.
Method:
% Accuracy 2,3,7,8 =
% Recovery..
13C6 1,2,3,4
Measured
2,3,7,8
Actual
2,3,7,8
x100
x100
Measured concentration of C6 1,2,3,4
level of 13C6 1,2,3,4 spike
Quality Assurance Requirements:
50% <% accuracy 2,3,7,8 < 150%
50% < % recovery 13C6 1,2,3,4 < 120%
Quality Control Action:
If % accuracy is out of range, whole sample set
will be reanalyzed. If 13C6 1,2,3,4-TCDD is out
of range for any individual sample, then that
sample must be reanalyzed in the next set.
Note: The spike level of 2,3,7,8-TCDD will be
10X the expected detected limit, and the spike
level of 13C6 1,2,3,7-TCDD will be 25X the
expected detection limit.
2. Precision of 2,3,7,8-TCDD Quantification
Precision, a measure of mutual agreement
among individual measurements of the same
pollutant in a sample, will be evaluated for each
sample set as the relative percent difference
between labduplicate measurements of 2,3,7,8-
TCDD.
Method:
I. If 2378-TCDD is measured at >6X the
detection limit, then:
relative percent difference =
I QAdup 1 -QAdup 2 I v inn
1 ^ | \j\J
QAdup
QA dup 1 and 2 = levels of 2378-TCDD in
duplicates.
2-7
-------
QA dup = mean value of 2378-TCDC
measurements.
If 2378-TCDD is measured at <6X the
detection limit, then:
relative percent difference =
detection limit
X 100
QAdup
III. If 2,3,7,8-TCDD is reported as "ND" then
precision is recorded as 0%.
IV. If 2,3,7,8-TCDD is ND for one sample, and
positive for duplicate, then calculate precision
using detection limit for ND sample.
Quality Assurance Requirements:
I. <50% if measured >6X detection limit
II. <100% if measured >6X detection limit
Quality Control Action:
If relative percent difference is out of range,
reextract and reanalyze duplicates to meet QA.
3. Completeness
Completeness is defined as the percentage of
valid (meets all quality assurance requirements)
data compared to the total number of samples
analyzed. For the National Dioxin Study the
"TROIKA" will not release to the Office of Water
any data that has not been reviewed by the
"TROIKA" to assure validity of data. Therefore,
Completeness is defined as 100% for this study
and will not otherwise be calculated.
4. Representativeness
Representativeness is dependent on the sam-
pling plan, and, therefore, is not covered in this
plan. Assessment of representativeness by eval-
uation of blind coded analyses of field duplicates
is the charge of the regional laboratory dioxin
coordinators and not the "TROIKA."
5. Comparability
Comparability is defined as the extent to which
the sample results can be verified or duplicated
by another independent laboratory or compared
against results previously found. No interlabora-
tory studies are currently planned by the Office of
Water for the TROIKA labs to participate in.
However, the TROIKA labs will routinely ex-
change blind coded samples between themselves
and Dr. Mike Gross, University of Nebraska,
Lincoln, Nebraska, who is currently working with
the TROIKA under an EPA cooperative agree-
ment.
6. Additional Potential Problems and Corrective
Action:
Additional problems with a sample set not
specifically identified in the above discussions
are described below along with the corrective
action. No data will be reported until corrective
action satisified QA.
Problem
Corrective Action
If method blank or
matrix blank is
positive
If detection limit for
blank is >the level of
2,3,7,8 in a sample
If spiked matrix or
reference sample is
outside QA for 2,3,7,8-
TCDD accuracy
If method efficiency is
outside QA for sample
If method efficiency is
outside QA for blank
If method efficiency is
outside QA for fortified
matrix or reference
sample
If 2,3,7,8-TCDD level is
outside QA for
duplicates
If 2,3,7,8-TCDD level
exceeds calibration
standard range
If GC resolution is
outside QA for isomer
specific analysis
If GC relative retention
time between 37CU
2,3,7,8 and 2,3,7,8-
TCDD are outside QA
for fortified matrix
or reference sample
Reextract and reanalyze
blank and all positives
Reextract and reanalyze
blank and all positives
which are at levels below
the mid for the blank
Reextract and reanalyze all
positives and matrix spike
sample
Reextract and reanalyze
sample
Reextract and reanalyze
blank and all positives
Reextract and reanalyze
spiked matrix
Reanalyze duplicates
Reextract and reanalyze
with smaller portion of
sample, or extend calibra-
tion range with additional
standard.
Reanalyze complete set on
alternate column
Reanalyze complete set on
alternative column
IV. Percentage Lipid of Fish Tissue
The percent lipid will be determined for all fish
that show a detectable level of 2378-TCDD. To
determine percent lipid, a separate aliquot of
fish (2 g) will be blended with 6 g anhydrous
sodium sulfate and extracted with methylene
chloride (25 ml_). The extract will be reduced to
volume(2 mL) with an air line, and transferred
to a previously weighed, dried aluminum boat.
The remainder of the methylene chloride is
allowed to evaporate in a hood. The aluminum
2-8
-------
boat is placed in an oven at 50°C for 2 hours,
cooled and weighed. The percent lipid is
calculated from the difference in weights.
V. Age of Fish
Aging of fish will be done only for samples with
detectable levels of 2378-TCDD. Scales will be
sampled before fish are ground and preserved.
Determination of age will be done by qualified
aquatic biologists at the University of
Wisconsin-Superior through a cooperative
agreement with ERL-D.
VI. Sampling Procedures
Fish and water sampling procedures will be
developed by the Office of Water, EPA Head-
quarters and reviewed by each of the TROIKA
labs.
VII. Sample Custody
Chain-of-custody procedures will be estab-
lished by the Office of Water, implemented by
Sample Control Center, Headquarters, and
used by the TROIKA when required.
VIII. Data Reporting Format
A standardized reporting form will be used at
ERL-D to report quantitative results for fish,
water and sediments. See Appendix C.
IX. Dioxin Isostereomer Analysis
A. PCDD/PCDF Congeners
The TROIKA will perform isomer specific
analysis for all tetra-octa PCDDs and tetra-
octa PCDFs when agreed upon with the Office
of Water. These analyses are limited, however,
to only congeners available to the TROIKA.
Moreover, these analyses are limited by the
assumption that all congeners reported behave
similar during sample preparation to the stable
isotope labeled congeners available for inter-
nal standards. Analytical methodology will
basically followthat described for 2378-TCDD,
but has been modified as described in the
ECL/EMSL section.
B. Other Polychlorinated Planar
Molecules
The TROIKA will perform qualitative analyses
for other TCDD isostereomers when agreed
upon with the Office of Water. These molecules
will be defined as those co-eluting with TCDD
in the sample preparation scheme. These mol-
ecules may include naphthalenes, biphenyl-
enes, acenaphthalenes, anthracenes, phen-
anthrenes, f luorenes, carbazoles, pyrenes and
chrycenes.
Analytical methodology will basically follow
that described for 2378-TCDD, but may need
to be modified. Basically this will involve the
elimination of the alumina column (PCBs/
PIMCs separation from PCDDs/PCDFs) and the
use of an alternate 2nd internal standard to
replace 13Ci2 2378-TCDD.
X. Validation of Methodology and
Evaluation of Data Comparability
The validation of the methodology used at
ERL-Duluth for biological tissue and water
was extensive and followed the outline given
below:
A. Evaluation of concentration of quantification
standards: A solution of 2,3,7,8-TCDD was
received from ECL-Bay St. Louis and quanti-
fied.
Result: Expected 26pg, found 25pg.
B. Evaluation of sample preparation biasbetween
internal standard and analyte: The ratio of
analyte (2,3,7,8-TCDD) and internal standard
(13C6 1,2,3,4-TCDD) was compared before
and after sample preparation when spiked into
fish at 5, 10 and 25pg/g.
Result: Before (m/z 322/328) 1.0 ± .08
After (m/z 322/328) 1.0 ± .08
C. Demonstrate that calculated minimum level of
detection was actually achievable: Samples of
fish were spiked at 2.0, 0.5 and 0.1 pg/g and
analyzed:
Result: All samples were quantified within
the QA/QC criteria for the National
Dioxin Study.
D. Demonstrate accuracy and precision with
2,3,7,8-TCDD spiked fish: Spike six (6) fish at
5.0pg/g and analyze.
Result: mean7.0pg/g, range3.7to 11pg/g,
s.d. 2.40
E. Demonstrate accuracy and precision with
environmental samples contaminated with
2,3,7,8-TCDD: Analyze one fish 15 times.
Result: mean 14.5pg/g, range 11-18pg/g,
s.d. 1.85
F. Demonstrate credibility of results for 2,3,7,8-
TCDD reported as "not detected": Spike seven
different environmental samples previously
analyzed as N.D. with 2,3,7,8-TCDD at 5pg/g.
Result: The 5pg/g spike was quantified with
sufficient S/N to show that the
previously reported level of detection
for the N.D. determination was
achievable.
G. Evaluate data comparability: Exchange a blind
coded environmental sample used as a 2,3,7,8-
2-9
-------
TCDD contaminated reference fish. (The fish
exchanged was ERL-Duluth reference fish
#3).
Result: ERL-Dululth -15pg/g
ECL-Bay St. Louis/
EMSL-RTP -12pg/g
University of Nebraska
(through EPA Region VII)
- 16pg/g
Dow Chemical Company - 18pg/g
California Analytical
(through EPA Region VII) -11 pg/g
The mean of the labs without ERL-Duluth is
14.2pg/g.
The mean of the labs with ERL-Duluth is
14.5pg/g.
Exchange blind coded reference fish and fish
from the National Dioxin Study with a highly
respected lab outside the USEPA. The lab
selected was Dow Chemical Company,
Midland, Ml.
During the course of the past year, it has been
shown that the hazards of exposure to 2,3,7,8-
TCDD in the lab are extremely small. Because
of this, and also partly because of improved
instrumental sensitivity, ERL-Duluth will now
be allowed to spike each sample with 25pg/g
37Cn 2,3,7,8-TCDD. Procedures for quantifica-
tion of 2,3,7,8-TCDD will remain the same for
consistency of data from start to finish of this
study. However a comparison of the ratio of
13C61,2,3,4,-TCDDto37Ci42,3,7,8-TCDD(m/z
328) will be made for each analyses to
determine if a bias has developed between
1,2,3,4-TCDD and 2,3,7,8-TCDD. As a QA
requirement the peak area of 37C14 2,3,7,8-
TCDD (m/z 328) must be within ± 25% of the
peak area of 13C6 1,2,3,4-TCDD. The QC
requirement is that if the peak area is outside
of the QA requirement, the analyses must be
done over. All other QA/QC requirements will
also be maintained.
Result:
Sample Description
2,3,7,8-TCDD found (pg/g)
Dow Chemical ERL-Duluth
Reference fish
Reference fish 2
Reference fish 3
Blank fish
EPA 13239
EPA 13273*
EPA 13272
EPA 13271
EPA 13243
13
32
18
ND(1.4)
2.2(1.2)
170
16
76
4.1 (1.4)
14
35
15
ND(1.0)
ND(1.8)
190
17
70
12 (.5)
* mean of duplicate
A review of the analytical methodology used
by the TROIKA labs for the National Dioxin
Study was held May 18 and 19,1985 in Bay St.
Louis, MS. Review comments specific to the
procedures used at ERL-Duluth primarily
concerned the identification of 2,3,7,8-TCDD
by relative GC retention time (12 ±2 seconds
past 13C6-1,2,3,4-TCDD), and quantification of
2,3,7,8-TCDD relative to 13C6 1234-TCDD
instead of using 13Ci2 2378-TCDD. For several
reasons, primarily Health and Safety related,
restrictions placed upon ERL-Duluth did not
allow for routine use of 2,3,7,8-TCDD in
sample preparation areas. This in turn promp-
ted the use of 13Ci2 and 13C6 labeled 1,2,3,4-
TCDD. As a result the very extensive validation
of the methodology and a thorough evaluation
of data comparability was necessary. As can
be seen from the above discussion, 2,3,7,8-
TCDD data generated by ERL-Duluth will meet
all QA/QC requirements established for the
Dioxin Study.
2-70
-------
Appendix A
Capillary GC Conditions
Column
Helium Carrier
linear velocity
Capillary head
pressure
Injection***
Septum purge
flow
Split flow
Injector temp.
Splitless time
Initial temp.
Program rate
Final temp.
Hold time
A(30m DB5)*
40 cm/sec
at 1 00°C
30psi
Grob type
splitless
5 mL/min
25 mL/min
300°C
1 .0 min
80°C
(hold 2 min)
4°C/min
250°C
30 min
B(30m SP2340)**
40 cm/sec
at100°C
30 psi
Grob type.
splitless
5 mL/min
25 mL/min
300°C
1 .0 min
80°C
4°C/min
250°C
30 min
*Column used for identification and quantification.
**Column used for isomer specific confirmation.
***GC will have both splitless and on-column capability.
A-1
-------
Appendix B
Mass Spectrometric Conditions
LRMS
Instrumentation: Finnigan Mat 4500 w/INCOS
data system.
Inlet System: Capillary direct to ionizer.
lonization Mode: Electron impact at 70 eV, 150°C,
2 x 10"8 torr.
Scan Mode: Computer controlled selected ion
monitoring.
Scan Time: 1.0 sec (0.25 sec/each of 4 mass
intervals).
Mass Intervals:
319.771-320.021 m+ TCDD
321.769-322.019 [m+2]+ TCDD
327.790-328.040 [m+2]+ 13C6 TCDD
333.809-334.059 [m+2]+ 13Ci2TCDD
HRMS
Instrumentation: Finnigan Mat 8230/SS300
data system.
Inlet: Capillary column directly inserted into
ionizer.
lonization: Electron impact, 70eV, 1 mA emission
current.
Source Pressure: 1 x 10"5torr.
Ionizer Temperature: 250°C.
Resolution: 5000, 10% valley.
Data Acquisition: Multiple Ion Selection (MIS) of
the following ions:
319.8964 [m]+ TCDD
321.8934 [m+2]+ TCDD
323.8904 [m+4]+ TCDD (optional)
327.9138 [m+2f 13C6 TCDD
333.9338 [m+2]+13d2TCDD
330.9793 PFK Lock Mass
Scan Rate: 15 MIS cycles per 10 sec.
Supplemental criteria for HRMS confirmation
Resolution: 8000-10000, 10% valley
Additionally, monitor COCI loss ions, 256.9327
[COCIf and 258.9298 [(m+2)-COCI]+
B-1
-------
Appendix C
Date:.
National Dioxm Study
United States Environmental Protection Agency
Environmental Research Laboratory
6201 Congdon Blvd.
Duluth, MN 55804
2378-TCOD Analysis
Episode Number
scc#
Matrix
2378-TCDD
(d1.)
pg/g
% Rec.
Int. Std.
pg/g
% Accuracy
pg/g
% Precision
pg/g
%
Moisture
Lipid
KEY
SCC # = Sample Control Center sample identification number.
d.L. = detection limit.
% Rec. Int. Std. = Percent recovery of internal standard . pg/g = level of spike.
Accuracy = measured/actual]! 00 : pg/g = level at which accuracy determined.
Precision =[( [sample A-duplicate A| (/mean of A values]100 : pg/g = level at which precision determined.
P2-NA = Priority two sample not analyzed because priority one below detection limit.
C-1
-------
Appendix D
NATIONAL DIOXIN STUDY: Sample Tracking System
1 -EPISODE*.
4-Date Received Troika: O/ O/ 0
7-2378-TCDD:
10-LablD-
12-MatrixType:
Spike Levels 14-NATURAL 1234: 100
16-320/322 Ratio: .00
18-% Lipid or Moisture: 0 0
21-Latitude: 00' 0"
23-Location:
2-SCC #:
5-Ana. Lab:
8-d.l.: 0.0
11-GC/MSID:
13-WT: 0.00
ERLD Loc = 158
3-Tier#: 0
6-Extraction Date: O/ O/ 0
9-% Recovery (INT. STD.): 0
15-13C61234: 500
17-S/N Ratio: 0.0
19-% Anal. Done: 1
22-Longitude: 0 0' 0"
20-Other TCDD: 0
24-Coment:
D-1
-------
Appendix E
oc
oc
O >.
o
en
0>
X 00
a x
a.
n
O
3 "
"7 S
* S
CO (-
CM 4
•- co"
J N-
*~ <0 tD
\ CJ
Q «
U Q
^— Q
4 o
co" 'r
CM °>
o CO"
V CM"
OC " II
0 So,
»- to *
o - 5.
< a w
S ° o
Z "o -fc
1 8 |
uj » £
OC -
*• o ^
05 — <
05 W DC
1 1 1
5 a. <
(A
C
a>
E
E
o
o
x £
"2 D
.9-5
-1 6
CM
CM
CO
X
CM
CM
CO
X
o
CM
O
Q
I"-
x co. -a
CO
CM"
a
O
aj *7
8 CO
CJ
o
Q
Q
— (J
9? ^
* «t
~" co" g
^ CM ~
W 1,
O
_ Q
ffl Q
Q) j_ .
0) ^ Q.
NX ^y^ "— "
°TM"
•-"
*- ">
"° "3.—
«- E °*
> to
> w
j,.
g> tt X
M O m
Q. C/) 2
LU
tt
W
5
o
C3
o
.a
ra
E-1
SUSOOVERNMENTPflimiNQ OFFICE 19» - 646-116/20818
------- |