SEFA
           United States
           Environmental Protection
           Agency
Office c«
Toxic Substances
Washington DC 20430
EPA-56CH5-85-010
 M, 1985
             ubstances
           Analytical Method:
           The Analysis of
           By-Product Chlorinated
           Biphenyls in Commercial
           Products and Product
           Wastes, Revision 2

-------
ANALYTICAL METHOD:  THE ANALYSIS OF BY-PRODUCT
 CHLORINATED BIPHENYLS IN COMMERCIAL PRODUCTS
        AND PRODUCT WASTES, REVISION 2
                      by

             Mitchell D. Erickson
             WORK ASSIGNMENT NO. 6

             SPECIAL REPORT NO. 1

          EPA Contract No. 68-02-3938
           MRI Project No. 8201-A(6)

                 May 20, 1985
                      For

     U.S. Environmental Protection Agency
          Office of Toxic Substances
         Field Studies Branch, TS-798
               401 M Street, SW
             Washington, DC  20460

Attn:  Joseph J.  Breen, Project Officer
       Daniel T.  Heggem, Work Assignment Manager

-------
                                 DISCLAIMER
          This document has been reviewed and approved for publication by the
Office of Toxic Substances, Office of Pesticides and Toxic Substances, U.S.
Environmental Protection Agency.  The use of trade names or commercial prod-
ucts does not constitute Agency endorsement or recommendation for use.

-------
                                   PREFACE
          This report contains an analytical method for the analysis of by-
product polychlorinated biphenyls in commercial products and wastes.  The
work was done on Work Assignment No. 6 on U.S. Environmental Protection Agency
Contract No. 68-02-3938.  This is the second revision of the method.  Previous
revisions are cited as references 3 and 4 in the method.  This report was
prepared by Mitchell Erickson.  The work on the previous revisions was con-
ducted by Dr. Erickson, John Stanley, Kay Turman, Gil Radolovich, Karin Bauer,
Jon Onstot, Donna Rose, Margaret Wickham, and Ruth Blair.  The work for the
previous revisions was performed on Task 51 of EPA Contract No. 68-01-5915.

          Two companion methods have been published which address air (Special
Report No.  2, EPA Report No. EPA-560/5-85-011) and water (Special Report No. 3,
EPA Report No. EPA-560-5-85-012).

          The EPA Work Assignment Manager, Daniel T.  Heggem, of Field Studies
Branch provided helpful guidance.

                                        MIDWEST RESEARCH INSTITUTE
                                        Clarence L. Haile
                                        DepUtty Program Manager
                                          5hn E. Going
                                         'rogram Manager
Approved:
James L.  Spigarelli, Director
Chemical  and Biological Sciences
  Department
                                     m

-------
TABLE OF CONTENTS

1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0

11.0
12.0
13.0
14.0
15.0
16.0
17.0
References

Scope and Application 	
Summary 	
Interferences 	
Safety 	 « . . .
Apparatus and Materials 	
Reagents 	
Calibration 	
Sample Collection, Handling, and Preservation 	
Sample Preparation 	
Gas Chromatographic/Electron Impact Mass Spectrometric
Determination 	
Qualitative Identification 	
Quantitative Data Reduction 	
Confirmation 	
Quality Assurance 	
Quality Control 	
Method Performance 	
Documentation and Records 	

Page
1
3
4
5
6
9
14
24
26

35
38
39
41
45
46
49
49
51

-------
                               LIST OF TABLES

Number                                                                Page

   1      Numbering of PCB Congeners	     2

   2      DFTPP Key Ions and Ion Abundance Criteria	     8

   3      Concentrations of Congeners in PCB Calibration Standards
            for Full Scan Analysis (ng/[jL)	    10

   4      Concentrations of Congeners in PCB Calibration Standards
            for Selected Ion Monitoring and Limited Mass Scan
            Analysis (pg/pL)	    11

   5      Composition of Internal Standard Spiking Solution (SS100)
            Containing 13C-Labeled PCBs 	    13

   6      Operating Parameters for Capillary Column Gas Chromato-
            graphic System	    15

   7      Operating Parameters for Packed Column Gas Chromatography
            System	    16

   8      Operating Parameters for Quadrupole Mass Spectrometer
            System	    17

   9      Operating Parameters for Magnetic Sector Mass Spectrometer
            System	    18

  10      Limited Mass Scanning (LMS) Ranges for PCBs	    19

  11      Characteristic SIM Ions for PCBs	    21

  12      Pairings of Analyte and Calibration Compounds 	    23

  13      Relative Retention Time (RRT) Ranges of PCB Homologs Versus
            d6-3,3' ,4,4'-Tetrachlorobiphenyl	    25

  14      Characteristic Ions for Internal Standards and 13C-Labeled
            PCB Surrogates	    40

  15      Analysis Worksheet	    42

  16      Analysis Report	    44

  17      Method Performance Parameters 	    50

-------
                               LIST OF FIGURES

Number                                                                Page

   1      Reconstructed ion chromatogram of calibration solution
            FS100 ng PCB obtained in the full  scan mode	    36
                                     VII

-------
             THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN
                   COMMERCIAL PRODUCTS AND PRODUCT WASTES
1.0   Scope and Application

      1.1   This is a gas chromatographic/electron impact mass spectrometric
            (GC/EIMS) method applicable to the determination of chlorinated
            biphenyls (PCBs) in commercial products and product wastes.   The
            PCBs present may originate either as synthetic by-products or as
            contaminants derived from commercial PCB products (e.g., Aroclors).
            The PCBs may be present as single isomers or complex mixtures and
            may include all 209 congeners from monochlorobiphenyl through
            decachlorobiphenyl  listed in Table 1.

            This method was prepared for use in demonstrating compliance with
            the EPA rules regarding the generation of PCBs as by-products in
            commercial chemical production1'2 and is based on earlier ver-
            sions.3'4  This revision includes elimination of a calculation
            which corrects the native PCB concentration based on the recovery
            of the l3C-labeled PCB recovery surrogates.   In addition, full
            scan is now emphasized over the selected ion monitoring and lim-
            ited mass scan mass spectrometric data collection modes.  The
            latter two techniques provide less qualitative information and
            should be used only if needed to achieve the required sensitivity.
            Additional background information on selection of the techniques
            has also been published.5

      1.2   The detection and quantitation limits are dependent upon the com-
            plexity of the sample matrix and the ability of the analyst to
            remove interferents and properly maintain the analytical system.
            For low level PCBs (0.05 to 0.2 H9/9). the method precision ap-
            pears to be about ± 60% based on a very limited study.   The method
            performance at higher levels (e.g., 2 M9/9) will be determined in
            future studies.

      1.3   This method is restricted to use by or under the supervision of
            analysts experienced in the use of gas chromatography/mass spec-
            trometry (GC/MS) and in the interpretation of gas chromatograms
            and mass spectra.   Prior to sample analysis, each analyst must
            demonstrate the ability to generate acceptable results with this
            method by following the procedures described in Section 15.2.

      1.4   The validity of the results depends on equivalent recovery of the
            analyte and 13C PCBs.   If the 13C PCBs are not thoroughly incor-
            porated in the matrix, the method is not applicable.

-------
                                TABLE  1.   NUMBERING OF  PCB CONGENERS3
HO.
1
2
3
4
5
6
7
8
9
10
11
12
11
14
15
16
17
18
19
20
21
22
71
74
?S
If,
77
28
29
30
31
32
33
14
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Structure
HB«oe>i1ore65.5'.6
2. 2', 3. 3' ,5. 6. S1
2, 2', 3, 4. 4', 5, 5'
2, 2', 3, 4, 4'. 5.6
2.2'.3,414I,5,6'
2,2',3,4.4',5-.6
2. 2'. 3. 4, 4', 6, 6'
2,2' .3. ,5.5'. 6
2, 2'. 3, ,5,6,6'
2,2',3. '.S.S'.S
2,2',3i '.5,6,6'
2.3,3', .4', 5,5'
2,3,3', .4'. 5,6
2,3,3'. ,4',S',6
2.3.3', .5, 5'. 6
2, 3, 3' ,4' ,5,5' .6
OetJCftlorabi phenyl s
2, 2'. 3, 3', 4,4', 5, 5'
2, 2', 3, 3', 4,4', 5, 6
2, 2', 3, 3' .4. 4'. 3. 5'
2, 2', 3, 3', 4,4', 5, 6'
2,2'.3,3'14,S.5'.6
2, 2'. 3,3'. 4,5,6, 6'
2, 2' ,3, 3' .4.5'. 6. 6'
2. 2'. 3, 3', 4. 5. 5'. 5'
2, 2'. 3, 3'. 5. 5', 6, 6'
2.2' .3,4. 4' .5.5' .6
2.21,3.4,4'.S.6,6'
2.3.3* .4,4',5.S'.6
NonaeMoroblonefyls
2.2'. 3, 3'. 4, 4', 5. 5'. 6
2.2', 3,3' .4, 4'. 5, 5.6'
2,2', 3,3'. 4, 5. 5', 6, 6'
Deeaetiloroblphenyl
2,2',3,3'*,4'.5,5'.6.o1
•»dopt*d froa a*ll*cMt*r. X.  tfid ZeM, N.. FrtuMus Z.  Anal. Chen., 302. 20-31  (1980).

-------
      1.5   During the development and testing of this  method,  certain analyti-
            cal  parameters  and equipment designs  were found to  affect the valid-
            ity  of the analytical  results.   Proper use  of the method requires
            that such parameters or designs  must  be used as specified.   These
            items are identified in the text by the word "must."   Anyone wish-
            ing  to deviate  from the method in areas so  identified must demon-
            strate that the deviation does not affect the validity of the data.
            Alternative test procedure approval must be obtained  from the
            Agency.   An experienced analyst  may make modifications to param-
            eters or equipment identified by the  term "recommended."  Each
            time such modifications are made to the method, the analyst must
            repeat the procedure in Section  15.2.  In this case,  formal ap-
            proval is not required, but the  documented  data from  Section 15.2
            must be on file as part of the overall quality assurance program.

      1.6   This method contains many options because of the diversity of
            matrices and interferences which may  be encountered.   Once the
            appropriate options for each sample type have been  selected, each
            laboratory should prepare a written step-by-step protocol for use
            by the analysts.  The protocol may contain  verbatim sections from
            this method, more detailed steps1for  certain techniques, or
            totally different extraction or  cleanup techniques.
2.0   Summary
      2.1   The process or product must be sampled such that the specimen col-
            lected for analysis is representative of the whole.   Statistically
            designed selection of the sampling position, time, or discrete
            product units should be employed.   The sample must be preserved
            to prevent PCB loss prior to analysis.   Customary inventory stor-
            age may be adequate for products.   For intermediates, process
            samples, and other non-product specimens, other preservation
            techniques may be needed.

      2.2   The sample is mechanically homogenized and subsampled if necessary.
            The sample must then be spiked with four 13C PCB surrogates
            (4-chloro-13C6-biphenyl;  3,3',4,4'-tetrachloro-13C12-biphenyl;
            2,2' ,3,3',-5,5',6,6l-octachloro-13C12~biphenyl; and decachloro-
            13C12-bipnenyl) and the surrogates incorporated by further
            mechanical agitation.

      2.3   The surrogate-spiked sample is extracted and cleaned up at the
            discretion of the analyst.  Simple dilution or direct injection
            is permissible.  Possible extraction techniques include liquid-
            liquid partition, thermal desorption, and sorption onto resin
            columns followed by solvent desorption.   Cleanup techniques may
            include liquid-liquid partition,  sulfuric acid cleanup, saponifi-
            cation, adsorption chromatography, high performance liquid chro-
            matography, gel permeation chromatography, or a combination of
            cleanup techniques.  The  sample is diluted or concentrated to a
            final  known volume for instrumental determination.

-------
      2.4   The PCB content of the sample extract must be determined by high
            resolution (preferred) or packed column gas chromatography/electron
            impact mass spectrometry (HRGC/EIMS or PGC/EIMS) operated in the
            full scan, selected ion monitoring (SIM), or limited mass scan
            (LMS) mode.

      2.5   PCBs are identified by comparison of their retention time and mass
            spectral intensity ratios to those in calibration standards.

      2.6   PCBs are quantitated by the internal standard technique, using
            response factors for a mixture of 10 PCB congeners.   The recover-
            ies of four 13C surrogates are used to monitor for losses in
            workup and determination.

      2.7   The PCBs identified by the SIM technique may be confirmed by full
            scan HRGC/EIMS, retention on alternate GC columns, other mass spec-
            trometric techniques, infrared spectrometry, or other techniques,
            provided that the sensitivity and selectivity of the technique
            are demonstrated to be comparable or superior to GC/EIMS.

      2.8   The analysis time is dependent on the extent of workup employed.
            The time required for instrumental analysis of a single sample,
            excluding instrumental calibration, data reduction and reporting,
            is typically 30 to 45 min.

      2.9   A quality assurance (QA) plan must be developed for each labora-
            tory.

      2.10  Quality control (QC) measures include laboratory certification
            and performance check sample analysis, procedural QC (instrumental
            performance, calculation checks), and sample QC (blanks, repli-
            cates, and standard addition).


3.0   Interferences

      3.1   Method interferences may be caused by contaminants in solvents,
            reagents, glassware, and other sample processing hardware, leading
            to discrete artifacts and/or elevated baselines in the total  ion
            current profiles.   All of these materials must be routinely demon-
            strated to be free from interferences by the analysis of labora-
            tory reagent blanks as described in Section 15.0.

            3.1.1    Glassware must be scrupulously cleaned.   All glassware
                     should be cleaned as soon as possible after use by rins-
                     ing with the last solvent used.   This should be followed
                     by detergent washing with hot water and rinses with tap
                     water and reagent water.  The glassware should then be
                     drained dry and heated in a muffle furnace at 400°C for
                     15 to 30 min.  Some thermally stable materials, such as
                     PCBs, may not be eliminated by this treatment.   Solvent
                     rinses with acetone and pesticide quality hexane may be

-------
                     substituted for the muffle furnace heating.   Volumetric
                     ware should not be heated in a muffle furnace.   After it
                     is dry and cool, glassware should be sealed and stored
                     in a clean environment to prevent any accumulation of
                     dust or other contaminants.   It is stored inverted or
                     capped with aluminum foil.

            3.1.2    The use of high purity reagents and solvents helps to
                     minimize interference problems.  Purification of sol-
                     vents by distillation in all-glass systems may be re-
                     quired.  All solvent lots must be checked for purity
                     prior to use.

      3.2   Matrix interferences may be caused by contaminants that are coex-
            tracted from the sample.  The extent of matrix interferences will
            vary considerably from source to source, depending upon the nature
            and diversity of the sources of samples.
4.0   Safety
      4.1   The toxicity or carcinogenicity of each reagent used in this
            method has not been precisely defined; however, each chemical
            compound should be treated as a potential health hazard.   From
            this viewpoint, exposure to these chemicals must be reduced to
            the lowest possible level by whatever means available.   The lab-
            oratory is responsible for maintaining a current awareness file
            of OSHA regulations regarding the safe handling of the chemicals
            specified in this method.  A reference file of material data han-
            dling sheets should also be made available to all  personnel in-
            volved in the chemical analysis.

      4.2   Polychlorinated biphenyls have been tentatively classified as
            known or suspected human or mammalian carcinogens.   Primary stan-
            dards of these toxic compounds should be prepared in a hood.
            Personnel must wear protective equipment, including gloves and
            safety glasses.

            Congeners highly substituted at the meta and para positions and
            unsubstituted at the ortho positions are reported to be the most
            toxic.   Extreme caution should be taken when handling these com-
            pounds neat or in concentrated solutions.  This class includes
            3,3',4,4'-tetrachlorobiphenyl (both natural abundance and isotop-
            ically labeled).

      4.3   Diethyl ether should be monitored regularly to determine the per-
            oxide content.   Under no circumstances should diethyl ether be
            used with a peroxide content in excess of 50 ppm,  as an explosion
            could result.   Peroxide test strips manufactured by EM Labora-
            tories (available from Scientific Products Company, Cat.  No.
            P1126-8 and other suppliers) are recommended for this test.   Pro-
            cedures for removal of peroxides from diethyl ether are included
            in the instructions supplied with the peroxide test kit.

-------
      4.4   Waste disposal  must be in accordance with RCRA and applicable
            state rules.
5.0   Apparatus and Materials

      5.1   Sampling containers - Amber glass bottles,  1-L or other appropri-
            ate volume,  fitted with screw caps lined with Teflon are appro-
            priate for liquid samples.   Cleaned foil may be substituted for
            Teflon if the sample is not corrosive.   If  amber bottles are not
            available, samples should be protected from light using foil or
            a light-tight outer container.   The bottle  must be washed,  rinsed
            with acetone or methylene chloride, and dried before use to mini-
            mize contamination.  Non-liquid samples may require other con-
            tainers.   Gas bags and stainless steel  cannisters may be appro-
            priate for gas samples.  Bags or cans may be appropriate for
            solid samples.  Consumer products in small  containers may remain
            in the manufacturer's packaging.

      5.2   Glassware -  All specifications are suggestions only.   Catalog
            numbers are  included for illustration only.

            5.2.1    Volumetric flasks - Assorted sizes.

            5.2.2    Pipets - Assorted sizes, Mohr delivery.

            5.2.3    Micro syringes - 10.0 pL for packed column GC analysis,
                     1.0 pL for on-column GC analysis.

            5.2.4    Chromatographic column - Chromaflex, 400 mm long x 19 mm
                     ID  (Kontes K-420540-9011 or equivalent).

            5.2.5    Kuderna-Danish Evaporative Concentrator Apparatus

                     5.2.5.1  Concentrator tube - 10 mL,  graduated (Kontes
                              K-570050-1025 or equivalent).   Calibration must
                              be checked.  Ground glass stopper size ($19/22
                              joint) is used to prevent evaporation of  solvent.

                     5.2.5.2  Evaporative flask - 500 mL (Kontes K-57001-0500
                              or equivalent).  Attached to concentrator tube
                              with springs (Kontes K-662750-0012 or equivalent).

                     5.2.5.3  Snyder column - Three ball  macro (Kontes
                              K-503000-0121 or equivalent).

      5.3   Balance - Analytical, capable of accurately weighing 0.0001 g.

-------
5.4   Gas chromatography/mass spectrometer system.

      5.4.1    Gas chromatograph - An analytical  system complete with a
               temperature programmable gas chromatograph and all re-
               quired accessories including syringes,  analytical columns,
               and gases.   The injection port must be  designed for on-
               column injection when using capillary columns or packed
               columns.   Other capillary injection techniques (split,
               splitless,  "Grob," etc.) may be used provided the per-
               formance specifications stated in  Section 7.1 are met.

      5.4.2    High resolution (capillary) GC column - A 10-30 m long x
               0.25 mm ID fused silica column with a 0.25 urn thick DB-5
               bonded silicone liquid phase (J&W  Scientific) is recom-
               mended.  Alternate liquid phases may include OV-101,
               SP-2100, Apiezon L, Dexsil 300, or other liquid phases
               or columns which meet the performance specifications
               stated in Section 7.1.

      5.4.3    Packed GC column - A 180 cm x 0.2  cm ID glass column
               packed with 3% SP-2250 on 100/120  mesh  Supelcoport or
               equivalent is recommended.  Other  liquid phases or
               columns which meet the performance specifications stated
               in Section 7.1 may be substituted.

      5.4.4    Mass spectrometer - Must be capable of  scanning from m/z
               150 to m/z 550 every 1.5 sec or less, collecting at least
               five spectra per chromatographic peak,  utilizing a 70-eV
               (nominal) electron energy in the electron impact ioniza-
               tion mode and producing a mass spectrum which meets all
               the criteria in Table 2 when 50 ng of decafluorotriphenyl
               phosphine [DFTPP, bis(perfluorophenyl)phenyl phosphine]
               is injected through the GC inlet.   Any  GC-to-MS interface
               that gives acceptable calibration  points at 10 ng per
               injection for each PCB isomer in the calibration stan-
               dard and achieves all acceptable performance criteria
               (Section 10) may be used.  Direct  coupling of the fused
               silica column to the MS is recommended.  Alternatively,
               GC-to-MS interfaces constructed of all  glass or glass-
               lined materials are recommended.  Glass can be deacti-
               vated by silanizing with dichlorodimethylsilane.

      5.4.5    A computer system that allows the  continuous acquisition
               and storage on machine-readable media of all mass spectra
               obtained throughout the duration of the chromatographic
               program must be interfaced to the  mass  spectrometer.
               The data system must have the capability of integrating
               the abundances of the selected ions between specified
               limits and relating integrated abundances to concentra-
               tions using the calibration procedures  described in this
               method.  The computer must have software that allows
               searching any GC/MS data file for  ions  of a specific mass

-------
rw ^^':J.::f'TabTe';2.   DFTPP Key Ions and Ion Abundance  Criteria

(':'    mTz                            Ion abundance criteria


      197                            Less than 1% of mass  198
      198                            100% relative abundance
      199                            5-9% of mass 198

      275                            10-30% of mass  198

      365                            Greater than 1% of  mass 198

      441                            Present, but less than mass 443
      442                            Greater than 40% of mass 198
      443                            17-23% of mass  442

-------
                     and plotting  such  ion  abundances  versus  time  or  scan
                     number to yield an extracted ion  current profile (EICP).
                     Software must also be  available that allows  integrating
                     the abundance in any EICP between specified  time or scan
                     number limits.

      5.5   Chromatographic systems.

            5.5.1    Gel permeation chromatography.

                     5.5.1.1  GPC  Autoprep  1002 (Analytical Biochemistry
                              Laboratories, Inc.) or equivalent.

                     5.5.1.2  A Bio-Beads SX-3 (Bio-Rad)  column.

            5.5.2    High performance liquid chromatography.

                     5.5.2.1  Waters Model  6000A pump  or  equivalent.

                     5.5.2.2  Waters Model  440 UV detector or equivalent.

                     5.5.2.3  Rheodyne  7125 injector or equivalent.

                     5.5.2.4  Amine column  (Waters uBondapak, 3.9 x 300 mm)
                              or equivalent.

            5.5.3    Gas chromatograph  for  GC/FID screening.

                     5.5.3.1  A temperature-programmable  GC equipped  with  a
                              flame ionization detector.   Varian  3740 or
                              equivalent.

                     5.5.3.2  A 2  m x 2 mm  ID glass column packed with 3%
                              SP-2250 on 100/120 mesh  Supelcoport or  equiva-
                              lent.   A  high resolution GC column  may  also  be
                              used.
6.0   Reagents
      6.1   Solvents - All solvents must be pesticide residue analysis grade.
            New lots should be checked for purity by concentrating an aliquot
            by at least as much as is used in the procedure.   HPLC solvents
            should have UV cutoffs of 210 nm or less.

      6.2   Calibration standard congeners - Standards of the PCB congeners
            listed in Tables 3 and 4 are available from Ultra Scientific,
            Hope, Rhode Island; or Analabs, North Haven, Connecticut.

-------
     Table 3.   Concentrations of Congeners in PCB Calibration Standards
                       for Full  Scan Analysis (ng/pL)
Homolog
1
2
3
4
5
6
7
8
9
10
4
-
-
13C-Cli
13C-C14
13c-ci8
13c-ci10
Congener
no.
1
7
30
50
97
143
183
202
207
209
210 (IS)
C10H7I (IS)b
C18Di2 dS)C
211 (RS)
212 (RS)
213 (RS)
214 (RS)
FS100
ng PCB
100
100
150
200
200
200
300
300
450
200
250
250
250
100
250
400
500
FS050
ng PCB
50
50
75
100
100
100
150
150
225
100
250
250
250
50
125
200
250
FS010
ng PCB
10
10
15
20
20
20
30
30
45
20
250
250
250
10
25
40
50
FS005
ng PCB
5
5
7.5
10
10
10
15
15
22.5
10
250
250
250
5
12.5
20
25
FS001
ng PCB
1
1
1.5
2
2
2
3
3
4.5
2
250
250
250
1
2.5
4
5
^Concentrations given as examples only.
 1-Iodonaphthalene.
 d12~Chrysene.
                                     10

-------
     Table  4.   Concentrations  of  Congeners  in  PCB Calibration Standards
              for Selected  Ion Monitoring and  Limited Mass  Scan
                              Analysis  (pg/pL)
Homo log
1
2
3
4
5
6
7
8
9
10
4
-
-
"C-Clj
13C-C14
13c-ci8
13c-ci10
Congener
no.
1
7
30
50
97
143
183
202
207
209
210 (IS)
C10H7I (IS)b
Ci8D12 (IS)C
211 (RS)
212 (RS)
213 (RS)
214 (RS)
SIM1000
pg PCB
1,000
1,000
1,500
2,000
2,000
2,000
3,000
3,000
4,500
2,000
250
250
250
1,000
2,500
4,000
5,000
SIM100
pg PCB
100
100
150
200
200
200
300
300
450
200
250
250
250
100
250
400
500
SIM050
P9 PCB
50
50
75
100
100
100
150
150
225
100
250
250
250
50
125
200
250
SIM010
pg PCB
10
10
15
20
20
20
30
30
45
20
250
250
250
10
25
40
50
.Concentrations given as examples only.
 1-Iodonaphthalene.
 d12-Chrysene.
                                     Jl

-------
6.3   Calibration standard stock solutions - Primary dilutions of each
      of the individual PCBs listed in Tables 3 and 4 are prepared by
      weighing approximately 1-10 mg of material within 1% precision.
      The PCB is then dissolved and diluted to 1.0 ml with hexane.   The
      concentration is calculated in mg/mL.   The primary dilutions are
      stored at 4°C in screw-cap vials with Teflon cap liners.   The
      meniscus is marked on the vial wall  to monitor solvent evaporation.
      Primary dilutions are stable indefinitely if the seals are main-
      tained.  The stock solutions and dilutions should be clearly la-
      beled with pertinent information such as sample code,  solvent,
      date prepared, initials of person preparing the solution,  and
      notebook reference.

6.4   Working calibration standards - Working calibration standards are
      prepared that are similar in PCB composition and concentration to
      the samples by mixing and diluting the individual standard stock
      solutions.  Example calibration solutions are shown in Tables 3
      and 4.  The mixture is diluted to volume with pesticide residue
      analysis quality hexane.   The concentration is calculated  in
      ng/mL as the individual PCBs.   Dilutions are stored at 4°C in
      narrow-mouth, screw-cap vials with Teflon cap liners.   The menis-
      cus is marked on the vial wall to monitor solvent evaporation.
      These secondary dilutions can be stored indefinitely if the seals
      are maintained.

      These solutions are designated FSxxx ng PCB and SIMxxx pg  PCB
      where the xxx is used to encode the  nominal concentration  of the
      lower congeners in ng/uL and pg/uL,  respectively.  The FS  prefix
      helps aid the analyst in identifying solutions which are appro-
      priate for full  scan analysis; the SIM prefix is for solutions  to
      calibrate in the selected ion monitoring and limited mass  scan
      acquisition modes.

6.5   Alternatively, certified stock solutions similar to those  listed
      in Tables 3 and 4 may be available from a supplier, in lieu of
      the procedure described in Section 6.4.

6.6   DFTPP standard - A 50-ng/uL solution of decafluorotriphenylphos-
      phine (DFTPP, PCR Research Chemicals,  Gainesville, FL) is  pre-
      pared in acetone or another appropriate solvent.

6.7   Surrogate standard stock solution -  The four 13C-labeled PCBs
      listed in Table 5 are available as a certified solution from Toxic
      and Hazardous Materials Repository,  U.S.  Environmental Protection
      Agency, Environmental Monitoring and Support Laboratory, 26 West
      St. Clair Street, Cincinnati,  Ohio 45268, (513) 684-7327.   This
      solution may be used as received or  diluted further.   These solu-
      tions are designated "SSxxx,"  where  the xxx is used to encode the
      nominal concentration in (jg/mL.

6.8   Internal standard solutions -  Solutions of d6-3,3',4,4'-tetra-
      chlorobiphenyl (KOR Isotopes,  Cambridge,  MA), 1-iodonaphthalene
      (Aldrich Chemical Company, Milwaukee,  WI) or d12-chrysene  (KOR

                               12

-------
Table 5.   Composition of Internal Standard Spiking Solution (SS100)
                   Containing 13C-Labeled PCBs
Congener
no.
211
212
213
214
Compound
4-Chloro-d1 ,2' ,3' ,4' ,5' ,6'-13C6)-biphenyl
3,3' ,4,4'-Tetrachloro-(13C12)-biphenyl
2,2' ,3,3' ,5,5' ,6,6'-Octachloro-(13C12)-biphenyl
Decachloro-(13C12)-biphenyl
Abbreviations
13C-Clj
13C-C14
13c-ci8
13c-ci10
Concentration
(|jg/mL)
100
250
400
500

-------
            Isotopes, Cambridge, MA) are prepared at nominal concentrations
            of 1-10 mg/mL in hexane.  The solutions are further diluted to
            give working standards.

            Note:   Any internal  standard may be used,  provided it meets the
            following criteria:   (a) it is not already present in the sample,
            (b) it gives a strong, recognizable mass spectrum, (c) it does
            not give mass spectral ions which interfere with native or 13C-
            labeled PCB quantisation, (d) it is chemically stable, and (e) it
            elutes in the PCB retention window.   Ideally,  several internal
            standards are used which have retention times  spanning the PCB
            retention windows to improve the response  factor precision.

      6.9   Solution stability - The calibration standard, surrogate, and
            DFTPP solutions should be checked frequently for stability.   These
            solutions should be replaced after 6 months, or sooner if compar-
            ison with quality control check samples indicates compound degrada-
            tion or concentration change.


7.0   Calibration

      7.1   The gas chromatograph must meet the minimum operating parameters
            shown in Tables 6 and 7, daily.   If all criteria are not met, the
            analyst must adjust conditions and repeat  the  test until all  cri-
            teria are met.

      7.2   The mass spectrometer must meet the minimum operating parameters
            shown in Tables 2, 8, and 9, daily.   If all criteria are not  met,
            the analyst must retune  the spectrometer and repeat the test  un-
            til all conditions are met.

            7.2.1    Full scan data  acquisition - Quadrupole mass spectrom-
                     eters must meet the tuning criteria in Table 2.  The
                     spectrometer must scan between m/z 150-550, although
                     wider scan ranges are permissible.

            7.2.2    Limited mass scan data acquisition -  Table 10 presents
                     a suggested set of IMS ranges. The mass spectrometer
                     should be set to at least'unit resolution.   The computer
                     acquisition parameters should utilize the minimum thres-
                     hold filtering  necessary so as not to lose pertinent data.
                     Optimum acquisition parameters will vary depending on
                     the condition of the mass spectrometer and should be
                     checked daily.
                                     14

-------
Table 6.   Operating Parameters for Capillary Column Gas Chromatographic System
        Parameter
      Recommended
    Tolerance
Gas chromatograph
Column

Liquid phase

Liquid phase thickness
Carrier gas
Carrier gas velocity
Injector

Injector temperature
Injection volume
Initial column temperature
Column temperature program
Separator
Transfer line temperature
Tailing factor
Peak width1
Finnigan 9610
15 -30 m x 0.255 mm ID
Fused silica
DB-5 (J&W)
0.25 urn
Helium
30-45 cm/sb
"Grob" (split/splitless
mode)
250-270°C
1.0-2.0 uL
60-80°C (2 min)d
70°-300°C at 10°C/mine
Nonef
280°C
0.7-1.5
7-10 s
Othera
Other

Other nonpolar
or semipolar
< 1 urn
Hydrogen
Optimum performance
Other0

Optimum performance
Other
Otherd
Other
Glass jet or other
Optimum^
0.4-3
< 15 s
 Substitutions permitted with any common apparatus or technique provided
.performance criteria are met.
 Measured by injection of air or methane at 270°C oven temperature.
 Manufacturer's instructions should be followed regarding injection tech-
 .nique.
 With on-column injection, initial temperature equals boiling point of the
 solvent; in this instance, hexane.
 C^Clio elutes at 270°C.  Programming above this temperature ensures a
fclean column and lower background on subsequent runs.
 Fused silica columns may be routed directly into the ion source to prevent
 separator discrimination and losses.
9High enough to elute all PCBs, but not high enough to degrade the column
.if routed through the transfer line.
 Tailing factor is width of front half of peak at 10% height divided by
 width of back half of peak at 10% height for single PCB congeners in solu-
•tion FSxxx ng PCB or SIMxxx pg PCB.
 Peak width at 10% height for a single PCB congener in FSxxx ng PCB or
 SIMxxx pg PCB.
                                       15

-------
  Table 7.  Operating Parameters for Packed Column Gas Chromatography System
      Parameter
   Recommended
Column packing


Carrier gas

Carrier gas flow rate

Injector

Injector temperature

Injection volume

Initial column temperature

Column temperature program

Separator

Transfer line temperature

Tailing factor0

Peak widthd
glass

3% SP-2250 on 1007
120 mesh Supelcoport

Helium

30 mL/min

On-column

250°C

1.0 uL

150°C, 4 min

150°-260°C at 8°/min

Glass jet

280°C

0.7-1.5

10-20 sec
   Tolerance
Gas chromatograph
Column
Finnigan 9610
180 cm x 0.2 cm ID
Other3
Other
Other nonpolar
or semipolar

.Hydrogen

Optimum performance

Other  ,

Optimum

S 5 ML

Other

Other

Other

Optimum

0.4-3

< 30 sec
.Substitutions permitted if performance criteria are met.
°High enough to elute all PCBs.
 Tailing factor is width of front half of peak at 10% height divided by
 width of back half of peak at 10% height for single PCB congeners  in solu-
 .tion FSxxx ng PCB or SIMxxx pg PCB.
 Peak width at 10% height for a single PCB congener in FSxxx ng PCB or
 SIMxxx pg PCB.
                                       16

-------
   Table 8.   Operating Parameters for Quadrupole  Mass  Spectrometer System

      Parameter                    Recommended                Tolerance


Mass spectrometer               Finnigan 4023              Other3

Data system                     Incos 2400                 Other

Scan range                      95-550                     Other

Scan time                       1 sec                      Other

Resolution                      Unit                       Optimum performance

Ion source temperature          280°C                      200°-300°C

Electron energy                 70 eV                      70 eV


.Substitutions permitted if performance criteria  are met.
 Greater than five data points over a GC peak is  a minimum.
 Filaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no solvent venting is
used.
                                      17

-------
 Table\9,^v?Operating:'P,arameiers!'for Magnetic Sector" Mass Spectrometer  Sys't'em
Parameter
Mass spectrometer-/
Data system
Scan range
Scan mode
Cycle time
Resolution
Ion source temperature
Electron energy
i ^^Recommended
Finnigan MAT 311A
Incos 2400
98-550
Exponential
1.2 sec
1,000
280°C
70 eV
Tolerance
Other3 "^
Other -G
Other
Other
Otherb
> 500
250-300°
70 eV
^Substitutions permitted if performance criteria are met.
 Greater than five data points over a GC peak is a minimum.
 Filaments should be shut off during solvent elution to  improve instrument
stability and prolong filament life, especially if no solvent venting  is
used.
                                      18

-------
Table 10.   Limited Mass Scanning (IMS) Ranges For PCBs
Compound
PHP! + 13r 12r u.ri
I»j2ngl> 1 ^ T l/g LgngU 1
C^2H8C12
^12''7^ '3
C12H6C14 + C12D6 C14 + 13C12H6C14
^12^5^5
Ci2H4Cle
^12^3^1?
Ci2H2Cl8
Cj2HCl g
^12^10
C10H7I
Ci8D12
13C12H2C18
13Ci2Cl10
Mass range (m/z)
186-198
220-226
254-260
288-310
322-328
356-362
390-396
426-434
460-468
496-502
254
240
440-446
508-514
                           19

-------
               The dwell times for the mass ranges given in Table 10
               will vary with instrument and should be optimized to
               allow at least five data points across a chromatographic
               peak.   Maximum sensitivity will be achieved when utiliz-
               ing maximum dwell  time.

               Instruments having the capability to switch mass ranges
               during an analysis require particular attention to the
               switching points to assure minimal data loss.   Switching
               points can be initially determined by analyzing a high
               level  congener or Aroclor mixture while in the full  scan
               mode.

      7.2.3    Selected ion monitoring data acquisition - Table 11 pre-
               sents a suggested set of characteristic ions for SIM.
               The SIM program must include at least two ions for each
               analyte, generally the primary and secondary ions in
               Table 11.  The mass spectrometer should be set to at
               least unit resolution.   The computer acquisition param-
               eters should utilize the minimum threshold filtering
               necessary so as not to lose pertinent data.   Optimum
               acquisition parameters will vary depending on the con-
               dition of the mass spectrometer and should be checked
               daily.

               The dwell times for the masses given in Table 11 will
               vary with instrument and should be optimized to allow at
               least five data points across a chromatographic peak.
               Maximum sensitivity will be achieved when utilizing max-
               imum dwell time.

               Instruments having the capability to switch mass ranges
               during an analysis require particular attention to the
               switching points to assure minimal data loss.   Switching
               points can be initially determined by analyzing a highly
               concentrated congener or Aroclor mixture while in the
               full scan mode.

7.3   The PCB response factors (RF ) must be determined in triplicate
      or other replication, as discussed below, using Equation 7-1 for
      the analyte homologs.
                RF  _
                RF  -
                      _ _
                  p - A,, x Mp                                  '

      where    RF  = response factor of a given PCB congener

                A  = area of the characteristic ion for the PCB congener
                 P     peak

                M  = mass of PCB congener in sample (micrograms)
                               20

-------
Table 11.  Characteristic SIM Ions for PCBs
Homolog
Cj2ngC I
^12^8^2
^12^7^13
C12H6C14
C^HsCls
^12"4^^6
Ci2HsCl7
^12^2^8
Ci2nCl 9
C12C11o
CiflHyl
Ci2D6Cl4
^18^12
13C612C6H9C1
13C12H6C14
13C12H2C18
13C12Clio

Primary
188 (100)
222 (100)
256 (100)
292 (100)
326 (100)
360 (100)
394 (100)
430 (100)
464 (100)
498 (100)
254 (100)
298 (100)
240 (100)
194 (100)
304 (100)
442 (100)
510 (100)
Ion (relative intensity^
Secondary
190 (33)
224 (66)
258 (99)
290 (76)
328 (66)
362 (82)
396 (98)
432 (66)
466 (76)
500 (87)
-
300 (49)
. -
196 (33)
306 (49)
444 (65)
512 (87)

Tertiary
-
226 (11)
260 (33)
294 (49)
324 (61)
364 (36)
398 (54)
428 (87)
462 (76)
496 (68)
-
296 (76)
-
-
302 (76)
440 (87)
508 (68)
                     21

-------
         A.  = area of the characteristic ion for the internal
          15     standard peak (d6~3,3',4,4'-tetrachlorobiphenyl
                 1-iodonaphthalene or other compound)

         M.  = mass of internal standard in sample (micrograms)

Using the same conditions as for RF ,  the surrogate response fac-
tors (RF ) must be determined using^Equation 7-2.

                  A  X M.
            RFs = irrinr                              E«-  7'2
                   I 3    O

where A  = area of the characteristic ion for the surrogate  peak

      M  = mass of surrogate in sample (micrograms)

Other terms are the same as defined in Equation 7-1.

If specific congeners are known to be present and if standards
are available, selected RF values may be employed.   For general
samples, solutions in Tables 3 and 4 or a mixture with a similar
level of internal standard added, may be used as the response
factor solution.   The PCB-surrogate pairs to be used in the  RF
calculation are listed in Table 12.

Generally, only the primary ions of both the analyte and surrogate
are used to determine the RF values.   If alternate ions are  to  be
used in the quantisation, the RF must be determined using that
characteristic ion.

The RF value must be  determined in a manner to assure ± 20%  pre-
cision^ For instruments with good day-to-day precision, a running
mean (RF) based on seven values may be appropriate.   A new value
is added each day and the oldest dropped from the mean.   Other
options include,  but  are not limited to,  triplicate determinations
of a single concentration spaced throughout a day or determination
of the RF at three different levels to establish a working curve.
If samples are being  analyzed on successive days,  a single RF
determination which is within ± 20% of the initial  day's tripli-
cate determination may be used.

If replicate RF values differ by greater than ± 10% RSD, the sys-
tem performance should be monitored closely.   If the RSD is  greater
than ± 20%, the data  set must be considered invalid and the  RF  re-
determined before further analyses are done.   The analyst is re-
sponsible for maintaining records of the RF precision.
                         22

-------
          Table 12.   Pairings of Analyte and Calibration Compounds
Analyte
Congener
no. Compound
1~3 C H Cl
A-m run
^ 1282
1 C-QQ P U PI
J.D 03 I»i2n7l»l3
40-81 C12H6C14
82-127 C12H5C15
128-169 C12H4C16
170-193 C12H3C17
iqA-onci P H PI
X3H £U3 12 2 8
206-208 C12HC19
oriq P n
c.\jj 1210
Calibration standard
Congener
no.
1
7
30
50
97
143
183
202
207
209
Compound
2
2,4
2,4,6
2, 2', 4, 6
2, 2' ,3' ,4,5
2, 2' ,3,4, 5, 6'
2, 2', 3', 4,4', 5', 6
2,2',3,3')5,5',6)6'
2, 2', 3, 3', 4, 4' ,5,6,6'
C^Clio
aBallschmiter numbering system, see Table 1.
                                     23

-------
      7.4   If the GC/EIMS system has not been demonstrated to yield a linear
            response or if the analyte concentrations are more than two orders
            of magnitude different from those in the RF solution, a calibration
            curve must be prepared.   If the analyte and RF solution concentra-
            tions differ by more than one order of magnitude, a calibration
            curve should be prepared.  A calibration curve should be estab-
            lished with triplicate determinations at three or more concentra-
            tions bracketing the analyte levels.

      7.5   The relative retention time (RRT) windows for the 10 homologs and
            surrogates must be determined unless the entire chromatogram is
            scanned for each homolog.  If all congeners are not available, a
            mixture of available congeners or an Aroclor mixture (e.g., 1016/
            1254/1260) may be used to estimate the windows.   The windows must
            be set wider than observed if all isomers are not determined.
            Typical RRT windows for one column are listed in Table 13.   The
            windows may differ substantially if other GC parameters are used.


8.0   Sample Collection, Handling, and Preservation

      8.1   Amber glass sample containers should have Teflon-lined screw caps.
            With noncorrosive samples, methylene chloride-washed aluminum foil
            liners may be substituted.  The volume and configuration are deter-
            mined by the amount of sample to be collected and its physical
            properties.  For dry powders, other containers such as wide-mouthed
            ointment jars or heavy-walled polyethylene bags may be appropriate.

      8.2   Sample bottle preparation

            8.2.1    All sample containers and caps should be washed in deter-
                     gent solution,  rinsed with tap water, and then with dis-
                     tilled water.  The bottles and caps are allowed to drain
                     dry in a contaminant-free area.   Then the caps are rinsed
                     with pesticide grade hexane and allowed to air dry.

            8.2.2    Sample bottles are heated to 400°C for 15 to 20 min or
                     rinsed with pesticide grade acetone or hexane and allowed
                     to air dry.

            8.2.3    The clean bottles are stored inverted and sealed until
                     use.

      8.3   Sample collection - Sample collection must be designed to meet
            sampling objectives.

            8.3.1    The primary consideration in sample collection is that
                     the sample collected be representative of the whole.
                     Therefore, sampling plans or protocols for each individ-
                     ual producer's situation will have to be developed.  The
                     recommendations presented here describe general situa-
                     tions.  The number of replicates and sampling frequency
                     also must be planned prior to sampling.

                                     24

-------
       Table 13.   Relative Retention Time (RRT) Ranges of PCB Homologs
                   Versus d6-3,3',4,4'-Tetrachlorobiphenyl
PCB
homolog
Monochloro
Dichloro
Trichloro
Tetrachloro
Pentachloro
Hexachloro
Heptachloro
Octachloro
Nonachloro
Decachloro
No. of
isomers
measured
3
10
9
16
12
13
4
6
3
1
Observed range
of RRTsa
0.40-0.50
0.52-0.69
0.62-0.79
0.72-1.01
0.82-1.08
0.93-1.20
1.09-1.30
1.19-1.36
1.31-1.42
1.44-1.45
Congener
no.
1
7
30
50
97
143
183
202
207
209
Observed
RRTa
0.43
0.58
0.65
0.75
0.98
1.05
1.15
1.19
1.33
1.44
Projected
range pf
RRTsD
0.35-0.55
0.45-0.80
0.55-1.00
0.55-1.05
0.80-1.10
0.90-1.25
1.05-1.35
1.10-1.50
1.25-1.50
1.35-1.50
 The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were
measured versus 3,3',4,4'-tetrach1orobiphenyl-d6 (internal standard) using
a 15-m J&W DB-5 fused silica column with a temperature program of 110°C
for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an on-
column injector.  A Finnigan 4023 Incos quadrupole mass spectrometer oper-
ating with a scan range of 95-550 daltons was used to detect each PCB
congener.
 The projected relative retention windows account for overlap of eluting
homologs and take into consideration differences in operating systems and
lack of all possible 209 PCB congeners.
                                     25

-------
            8.3.2    Discrete product units - If the product is small  enough
                     that one or more discrete units would be used as  the
                     analytical  sample,  a statistically random sampling ap-
                     proach is recommended.

            8.3.3    Liquids or free-flowing solids - If possible, the source
                     is mixed thoroughly before collecting the sample.   If
                     mixing is impractical,  the sample should be collected
                     from a representative area of the source.   If the liquid
                     is flowing through  an enclosed system, sampling through
                     a valve should be randomly timed.

            8.3.4    Solids - Larger bulk solids which must be subsampled to
                     get a reasonably sized analytical  sample must be  treated
                     on a case-by-case basis.  A representative sample should
                     be obtained by designing a sampling location selection
                     scheme such that all parts of the whole have a finite,
                     known probability of inclusion.   Based on such a  scheme,
                     the PCB content of  the sample can be used to extrapolate
                     to the content of the whole.

      8.4   Sample preservation - Product samples should be stored as  the bulk
            or packaged product inventory would be stored, or in a cool,  dry,
            dark area.   Intermediates, process samples, or other non-product
            specimens should be stored at 4°C.  If there is a possibility of
            microbial degradation, addition of H2S04 during collection to a
            pH < 2 is recommended.  A test strip is used to monitor pH.  Stor-
            age times in excess of 4 weeks are not recommended.

            If residual chlorine is present in the sample, it should be
            quenched with sodium thiosulfate.  EPA Methods 330.4 and 330.5
            may be used to measure the residual chlorine.6  Field test kits
            are available for this purpose.


9.0   Sample Preparation

      Since a wide variety of matrices may be subjected to analysis by this
      method, the extraction/cleanup procedure cannot be specified.   This
      section describes general  guidelines for subsampling, addition of 13C
      surrogates, dilution, extraction,  cleanup, extract concentration, and
      other sample preparation procedures.

      9.1   Sample homogenization and subsampling - The sample is homogenized
            by shaking, blending, shredding, crushing,  or other appropriate
            mechanical  technique.  A representative subsample of 100 g or other
            known mass is then taken.   The sample size is dependent upon  the
            anticipated PCB levels and difficulty of the subsequent extraction/
            cleanup steps.
                                     26

-------
      Note:   The precision of the mass determination at this step will
      be reflected in the overall method precision.   Therefore, an an-
      alytical  balance must be used to assure that the weight is accu-
      rate to ± 1% or better.

9.2   Surrogate addition - An appropriate volume of surrogate solution
      SSxxx must be pipetted into the sample.   The final  concentration
      of the surrogates must be in the working range of the calibration
      and well  above the matrix background.   The surrogates are thor-
      oughly incorporated by further mechanical agitation.   For nonvis-
      cous liquids, shaking for 30 sec should be sufficient.  For vis-
      cous liquids or free-flowing solids, 10-min tumbling is recommended.
      In cases where inadequate incorporation may be expected, such as
      solids, overnight equilibration with agitation is recommended.

      Note:   The volume measurement of the spiking solution is critical
      to the overall method precision.  The analyst must exercise cau-
      tion that the volume is known to ± 1% or better.  Where necessary,
      calibration of the pi pet is recommended.

9.3   Sample preparation (extraction/cleanup) - After addition of the
      surrogates, the sample is further treated at the discretion of
      the analyst, provided that the GC/EIMS response of the four sur-
      rogates are sufficient for reliable quantitation.  The literature
      pertaining to these techniques has been reviewed.5  Several pos-
      sible techniques are presented below for guidance only.  The ap-
      plicability of any of these techniques to a specific sample ma-
      trix must be determined by the precision and accuracy of the 13C
      PCB surrogate recoveries, as discussed in Section 15.4.1.

      9.3.1    Extraction

               9.3.1.1  Dilution - In some cases, where the PCB concen-
                        tration is high, a simple volumetric dilution
                        with an appropriate solvent may be sufficient
                        sample preparation.

               9.3.1.2  Direct injection - If sample viscosity permits,
                        direct injection with no dilution is permissible.

               9.3.1.3  Kuderna-Danish concentration - If the sample is
                        a solvent with low PCB concentration, reduction
                        of sample volume may be sufficient preparation.
                        In this case, the sample is placed in the
                        Kuderna-Danish apparatus and concentrated over
                        steam to an appropriate volume.  (Alternatives
                        to Kuderna-Danish, such as rotary evaporation
                        techniques may also be used.)

               9.3.1.4  Evaporative concentration using nitrogen - For
                        smaller volume (5-50 mL) solvent samples concen-
                        tration may be achieved by blowing sample with
                               27

-------
                  a gentle stream of pre-purified nitrogen.   Con-
                  centrate sample to final  volume,  rinse sides of
                  vial  with small aliquots  of solvent,  and again
                  concentrate to final  volume.   In  some cases,
                  where background is a problem,  concentration to
                  dryness will eliminate interference by low-boil-
                  ing compounds, e.g.,  chloroform,  without remov-
                  ing the more volatile PCBs.   After blowing to
                  dryness, reconstitute with solvent and sonicate
                  for 5 min.   Care must be  taken  to avoid evapora-
                  tive loss of the more volatile  PCBs.   Do not
                  heat a dry sample for extended  periods.

         9.3.1.5  Liquid-liquid extraction  - If the matrix is
                  aqueous (or another solvent in  which  PCBs  are
                  only slightly soluble), a liquid-liquid parti-
                  tion may be effective.  The solvent,  number of
                  extractions, solvent-to-sample  ratio, and  other
                  parameters are chosen at  the analyst's discretion.

         9.3.1.6  Sorbent column extraction - PCBs  may  be isolated
                  from free-flowing liquids onto  sorbent columns.
                  The selection of sorbent  (XAD,  Porapak,  carbon-
                  polyurethane foam, etc.)  will depend  on the na-
                  ture of the matrix.   The  available methods have
                  been reviewed.5

         9.3.1.7  Thermal desorption -  If the matrix is nonvol-
                  atile, thermal desorption of the  PCBs onto a
                  sorbent column, filter, or cold trap  may be an
                  effective extraction/cleanup method.

         9.3.1.8  Matrix destruction -  Some matrices may be  easily
                  degraded to a volatile or extractable compound.
                  Once the matrix is degraded,  the  PCBs can  be
                  isolated, either by evaporation of the matrix
                  or partitions of the  PCBs into  a  nonpolar  sol-
                  vent and the matrix into  a polar  solvent (e.g.,
                  water).  Examples include:  (1) esters which
                  may be saponified with base to  the acid anion
                  and then extracted with water,  and (2) acid
                  chlorides which may be hydrolyzed to  the acid
                  anion with water or base  and then extracted
                  with water.

9.3.2    Cleanup - Several tested cleanup techniques are described
         below.   All but the base cleanup  (9.3.2.8) were previously
         validated for PCBs in transformer  fluids.7  Depending
         upon the complexity of the sample, one or  more of the
         techniques may be required to  fractionate  the  PCBs  from
         interferences.   For most cleanups  a concentrated (1-5 ml)
         extract should be used.
                         28

-------
9.3.2.1  Acid cleanup

         9.3.2.1.1  Place 5 mL of concentrated sulfuric
                    acid into a 40-mL narrow-mouth screw-
                    cap bottle.   Add the sample extract.
                    Seal the bottle with a Teflon-lined
                    screw cap and shake for 1 min.

         9.3.2.1.2  Allow the phases to separate, transfer
                    the sample (upper phase) with three
                    rinses of 1-2 mL solvent to a clean
                    container.

         9.3.2.1.3  Back-extract sample extract with 5-
                    10 drops of distilled water.  Pass
                    through a short column of anhydrous
                    sodium sulfate and concentrate to an
                    appropriate volume.

         9.3.2.1.4  Analyze as described in Section 10.0.

         9.3.2.1.5  If the sample is highly contaminated,
                    a second or third acid cleanup may
                    be employed.

9.3.2.2  Florisil column cleanup

         9.3.2.2.1  Variations among batches of Florisil
                    (PR grade or equivalent) may affect
                    the elution volume of the various
                    PCBs.  For this reason, the volume
                    of solvent required to completely
                    elute all PCBs must be verified by
                    the analyst.  The weight of Florisil
                    can then be adjusted accordingly.

         9.3.2.2.2  Place a 20-g charge of Florisil,
                    activated overnight at 130°C, into a
                    Chromaflex column.  Settle the Flor-
                    isil by tapping the column.  Add about
                    1 cm of anhydrous sodium sulfate to
                    the top of the Florisil.  Pre-elute
                    the column with 70-80 ml of hexane.
                    Just before the exposure of the sodium
                    sulfate layer to air, stop the flow.
                    Discard the eluate.

         9.3.2.2.3  Add the sample extract to the column.

         9.3.2.2.4  Carefully wash down the inner wall
                    of the column with 5 ml of hexane.
                29

-------
         9.3.2.2.5  Add 200 ml of 6% ethyl ether/hexane
                    and set the flow to about 5 mL/min.

         9.3.2.2.6  Collect 200 mL of eluate in a Kuderna-
                    Danish flask.   All the PCBs should be
                    in this fraction.   Concentrate to an
                    appropriate volume.

         9.3.2.2.7  Analyze the sample as described in
                    Section 10.0.

9.3.2.3  Alumina column cleanup

         9.3.2.3.1  Adjust the activity of the alumina
                    (Fisher A540 or equivalent) by heat-
                    ing to 200°C for at least 2 hr.   When
                    cool,  add 3% water (wt:wt) and mix
                    until  uniform.   Allow the deactivated
                    alumina to equilibrate at least 1/2 hr
                    before use.   Store in a tightly sealed
                    bottle.

         9.3.2.3.2  Variations between batches of alumina
                    may affect the elution volume of the
                    various PCBs.   For this reason,  the
                    volume of solvent required to com-
                    pletely elute all  of the PCBs must
                    be verified by the analyst.   The
                    weight of alumina can then be ad-
                    justed accordingly.

         9.3.2.3.3  Place  a 50-g charge of alumina into
                    a Chromaflex column.   Settle the
                    alumina by tapping.   Add about 1 cm
                    of anhydrous sodium sulfate.   Pre-
                    elute  the column with 70-80 mL of
                    hexane.   Just before exposure of the
                    sodium sulfate layer to air,  stop
                    the flow.   Discard the eluate.

         9.3.2.3.4  Add the sample extract to the column.

         9.3.2.3.5  Carefully wash down the inner wall
                    of the column with 5 mL of hexane.

         9.3.2.3.6  Add 295 ml of hexane to the column.

         9.3.2.3.7  Discard the first 50 ml.

         9.3.2.3.8  Collect 250 mL of the hexane in a
                    Kuderna-Danish flask.   All of the
                    PCBs should be in this fraction.
                    Concentrate to an appropriate volume.

                30

-------
         9.3.2.3.9  Analyze the sample as described in
                    Section 10.0.

9.3.2.4  Silica gel  column cleanup

         9.3.2.4.1  Activate silica gel (Davison Grade
                    950 or equivalent) at 135°C overnight.

         9.3.2.4.2  Variations between batches of silica
                    gel may affect the elution volume of
                    the various PCBs.   For this reason,
                    the volume of solvent required to
                    completely elute all of the PCBs must
                    be verified by the analyst.  The
                    weight of silica gel can then be ad-
                    justed accordingly.

         9.3.2.4.3  Place a 25-g charge of activated
                    silica gel into a Chromaflex column.
                    Settle the silica gel by tapping the
                    column.  Add about 1 cm of anhydrous
                    sodium sulfate to the top of the
                    silica gel.

         9.3.2.4.4  Pre-elute the column with 70-80 mL
                    of hexane.  Discard the eluate.  Just
                    before exposing the sodium sulfate
                    layer to air, stop the flow.

         9.3.2.4.5  Add the sample extract to the column.

         9.3.2.4.6  Wash down the inner wall of the column
                    with 5 mL of hexane.

         9.3.2.4.7  Elute the PCBs with 195 mL of 10%
                    diethyl ether in hexane (v:v).

         9.3.2.4.8  Collect 200 mL of the eluate in a
                    Kuderna-Danish flask.  All of the
                    PCBs should be in this fraction.
                    Concentrate to an appropriate volume.

         9.3.2.4.9  Analyze the sample as described in
                    Section 10.0.

9.3.2.5  Gel permeation cleanup

         9.3.2.5.1  Set up and calibrate the gel permea-
                    tion chromatograph with an SX-3 column
                    according to the Autoprep instruction
                    manual.  Use 15% methylene chloride
                    in cyclohexane (v:v) as the mobile
                    phase.

                31

-------
         9.3.2.5.2  Inject 5.0 mL of the sample extract
                    into the instrument.  Collect the
                    fraction containing the PCBs (see
                    Autoprep operator's manual) in a
                    Kuderna-Danish flask equipped with
                    a 10-mL ampul.

         9.3.2.5.3  Concentrate the PCB fraction to an
                    appropriate volume.

         9.3.2.5.4  Analyze the sample as described in
                    Section 10.0.

9.3.2.6  Acetonitrile partition

         9.3.2.6.1  Place the sample extract into a 125-mL
                    separatory funnel  with enough hexane
                    to bring the final volume to 15 mL.
                    Extract the sample four times by shak-
                    ing vigorously for 1 min with 30-mL
                    portions of hexane-saturated acetoni-
                    trile.   Retain hexane layer for com-
                    bination with other hexane extracts
                    in 9.3.2.6.3.

         9.3.2.6.2  Combine and transfer the acetonitrile
                    phases to a 1-L separatory funnel
                    and add 650 mL of distilled water and
                    40 mL of saturated sodium chloride
                    solution.   Mix thoroughly for about
                    30 sec.   Extract with two 100-mL por-
                    tions of hexane by vigorously shaking
                    about 15 sec.

         9.3.2.6.3  Combine the hexane extracts in a
                    1-L separatory funnel and wash with
                    two 100-mL portions of distilled
                    water.   Discard the water layer and
                    pour the hexane layer through an 8-
                    to 10-cm anhydrous sodium sulfate
                    column into a 500-mL Kuderna-Danish
                    flask equipped with a 10-mL ampul.
                    Rinse the separatory funnel and
                    column with three 10-mL portions of
                    hexane.

         9.3.2.6.4  Concentrate the extracts to an ap-
                    propriate volume.

         9.3.2.6.5  Analyze as described in Section 10.0.
                32

-------
9.3.2.7  Florisil  slurry cleanup

         9.3.2.7.1  Place the sample extract into a 20-mL
                    narrow-mouth screw-cap container.
                    Add 0.25 g of Florisil (PR grade or
                    equivalent).  Seal with a Teflon-lined
                    screw cap and shake for 1 min.

         9.3.2.7.2  Allow the Florisil to settle; then
                    decant the treated solution into a
                    second container with rinsing.   Con-
                    centrate the sample to an appropriate
                    volume.   Analyze as described in Sec-
                    tion 10.0.

9.3.2.8  Base cleanup8

         9.3.2.8.1  Quantitatively transfer the concen-
                    trated extract to a 125-mL extraction
                    flask with the aid of several small
                    portions of solvent.

         9.3.2.8.2  Evaporate the extract just to dryness
                    with a gentle stream of dry filtered
                    nitrogen, and add 25 ml of 2.5% alco-
                    holic KOH.

         9.3.2.8.3  Add a boiling chip, put a water con-
                    denser in place, and allow the solu-
                    tion to reflux on a hot plate for 45
                    min.

         9.3.2.8.4  After cooling, transfer the solution
                    to a 250-mL separatory funnel with
                    25 ml of distilled water.

         9.3.2.8.5  Rinse the extraction flask with 25
                    mL of hexane and add it to the
                    separatory funnel.

         9.3.2.8.6  Stopper the separatory funnel and
                    shake vigorously for at least 1 min.
                    Allow the layers to separate, and
                    transfer the lower aqueous phase to
                    a second separatory funnel.

         9.3.2.8.7  Extract the saponification solution
                    with a second 25-mL portion of hexane.
                    After the layers have separated, add
                    the first hexane extract to the sec-
                    ond separatory funnel and transfer
                    the aqueous alcohol layer to the
                    original separatory funnel.

                33

-------
                        9.3.2.8.8  Repeat the extraction with a third
                                   25-mL portion of hexane.   Discard
                                   the saponification solution, and com-
                                   bine the hexane extracts.

                        9.3.2.8.9  Concentrate the hexane layer to an
                                   appropriate volume,  and analyze as
                                   described in Section 10.0.

               9.3.2.9  High performance liquid chromatographic cleanup

                        9.3.2.9.1  Quantitatively transfer the concen-
                                   trated extract into  the sample loop
                                   or the barrel of a syringe.   Rinse
                                   the vial with several small  portions
                                   of solvent.   It may  be necessary to
                                   inject several fractions.

                        9.3.2.9.2  Inject the extract and washes onto
                                   the amine column (Waters  pBondapak
                                   3.9 x 300 mm or equivalent) and elute
                                   the PCBs with 1.0 mL/min  hexane.
                                   The UV at 254 nm or  lower should be
                                   monitored.

                        9.3.2.9.3  Collect the eluent from 3 min to 9.5
                                   min as it exits UV cell.   The elution
                                   time should be verified using PCB
                                   standards covering a range from mono-
                                   chlorobiphenyls to decachlorobiphenyls.

                        9.3.2.9.4  After collection, wash the column by
                                   eluting with methylene chloride until
                                   the absorbance attains a  stable mini-
                                   mum.   Return the system to hexane.

                        9.3.2.9.5  Concentrate the hexane eluate under
                                   a gentle stream of purified nitrogen
                                   to an appropriate volume  and analyze
                                   as described in Section 10.0.

9.4   Optional screening for interferences using GC/FID

      Note:   Since many sample matrices are one of a kind or infrequently
      encountered by the analyst, the effectiveness of  the extraction
      and cleanup for a matrix may be unknown.   A simple screen to as-
      sess whether the interferences have been reduced  to a  tolerable
      level  can both save GC/MS time and prevent contamination of the
      GC/MS instrument with very dirty samples.   This screen should
      not be used to determine PCB levels under this analytical method.
                               34

-------
            9.4.1    Using a GC system as described in Section 5.5.3,  analyze
                     for background interferences.

            9.4.2    A 2 m x 2 mm glass column packed with 3% SP-2250  on 1007
                     120 Supelcoport or equivalent is suggested.   A flow rate
                     of 40 mL/min 95% air/5% methane or nitrogen  is recommended.
                     The air and hydrogen flow rates should be sufficient to
                     keep the flame lit and to burn efficiently,  e.g.,  300
                     mL/min air and 30 mL/min H2.

            9.4.3    The recommended temperature program is from  50°C  to 250°C
                     at 20°C/min with an initial hold of 3 min and a final
                     hold of 10 min.   The injector temperature should  be 200°C
                     and the detector 300°C.

            9.4.4    Set instrumental sensitivity comparable to the antici-
                     pated mass spectral sensitivity.  It is advisable to
                     establish criteria for rejection of samples  at a  given
                     attenuation such as (1) any off-scale peaks  in PCS elu-
                     tion window, (2) a baseline rise of over 40% full-scale,
                     and (3) other criteria which are indicative  of "problem"
                     samples.

            9.4.5    If the FID screen suggests that the sample is not amen-
                     able to analysis by GC/EIMS,  the analyst may either (1)
                     cycle the sample through the same cleanup again if it
                     appears that the cleanup technique was overloaded by the
                     matrix the first time, (2) submit the extract to  another
                     cleanup technique which may remove more interferences,
                     or (3) analyze a new aliquot of sample by another extrac-
                     tion or cleanup technique.


10.0  Gas Chromatographic/Electron Impact Mass Spectrometric Determination

      10.1  Internal  standard addition - An appropriate volume of the  internal
            standard solution is pipetted into the sample.  The final  concen-
            tration of the internal standard must be in the working range of
            the calibration and well  above the matrix background.   The inter-
            nal standard is thoroughly incorporated by mechanical  agitation.

            Note:  The volumetric measurement of the internal standard solu-
            tion is critical to the overall method precision.  The analyst
            must exercise caution that the volume is known to be  ± 1%  or bet-
            ter.   Where necessary, calibration of the pipet is recommended.

      10.2  Tables 2, and 6 through 11 summarize the recommended  operating
            conditions for analysis.   Figure 1 presents an example of  a
            chromatogram.
                                     35

-------
109.8-1
  RIC
 oo
                                                            13
                                   C]2D6C14


                                   C12H6C14
                13C612C6H9C1
                      C1QH7I
               50
                                 30
                                  97
                              183
                                         143
                                                                                        88960.
                                                                                   202
                                                                                   C18D12
                                                                                   13
                                                                                     C12H2C18
                                                                                                     209

                                                                                                    13
                                                                                           207
                                                                                                        2C110
         13:20
1000
16:40
1200
20:00
1400
23:20
1600
26:40
1800
30:00
SCAN
TIME
    Figure 1.  Reconstructed ion chromatogram of calibration  solution FS100 ng PCB obtained in the full scan mode.   The
    concentration of the 10 PCB calibration  congeners,  the four 13C-labeled PCB recovery surrogates, and the three  inter-
    nal standards are in Table 3.   See Table 1  for  PCB  numbering system, Table 6 for capillary GC parameters, and Table  8
    for mass spectrometer operating parameters.

-------
      The analyst may choose to operate the mass spectrometer at any
      appropriate sensitivity,  using either full scan,  limited mass
      scanning or selected ion  monitoring acquisition.   The sensitivity
      selected will  depend on anticipated PCB levels and the instrumen-
      tal LOQ needed to meet the required method LOQ.   In general,  the
      more concentrated the PCBs,  the greater the precision, accuracy,
      and qualitative data confidence.   Thus, if possible,  the amount
      of sample and the concentration factor should be  scaled so that
      full scan acquisition may be utilized.

10.3  While the highest available chromatographic resolution is not a
      necessary objective of this method, good chromatographic per-
      formance is recommended.   With the high resolution of HRGC, the
      probability that the chromatographic peaks consist of single  com-
      pounds is higher than with PGC.  Thus,  qualitative and quantita-
      tive data reduction should be more reliable.

10.4  After performance of the  system has been certified for the day
      and all instrument conditions set according to Tables 2, and  6
      through 11, inject an aliquot of the sample onto  the GC column.
      If the response for any ion, including surrogates and internal
      standards, exceeds the working range of the system, dilute the
      sample and reanalyze.  If the responses of surrogates, internal
      standards, or analytes are below the working range, recheck the
      system performance.  If necessary, concentrate the sample and re-
      analyze.

10.5  Record all data on a digital storage device (magnetic disk, tape,
      etc.) for qualitative and quantitative data reduction as discussed
      below.

10.6  The instrumental performance must be monitored from run-to-run.
      The areas of internal standards must be consistent (e.g., ± 20%).
      If a low area is encountered, the injection may be suspect.

      The resolution and peak shape of the internal standards, surro-
      gates, and other peaks should be monitored during or immediately
      after data acquisition.  Poor chromatography may  indicate a bad
      injection, matrix interferences, or column degradation.

10.7  If a "dirty" sample is encountered, the analyst must employ appro-
      priate measures to demonstrate that there is no memory or carry-
      ons to subsequent samples.  To assess the system  cleanliness, a
      standard, blank sample, or solvent blank may be run.

      If the system is contaminated, remedial efforts may include (1)
      changing or cleaning the  syringe, (2) cleaning the injector,  (3)
      baking out the column at  its maximum temperature, (4) changing
      to a new column, or (5) cleaning the ion source.
                               37

-------
11.0  Qualitative Identification

      11.1  Full  scan data

            11.1.1   The peak must elute within the retention time windows
                     set for that homolog (as described in Section 7.5).

            11.1.2   The unknown spectrum should be compared to that of an
                     authentic PCB.   The intensity of the three largest ions
                     in the molecular cluster (two largest for monochlorobi-
                     phenyls) must match the ratio observed for a standard
                     within ± 20%.   Fragment clusters with proper intensity
                     ratios should also be present.   System noise at low con-
                     centration or interferences may skew the ion ratio be-
                     yond the ± 20% criteria.   If the analyst's best judge-
                     ment is that a peak, which does not meet the qualitative
                     criteria, is a PCB, the peak may be included in the cal-
                     culation, with a footnote explaining the data and the
                     reason for relaxing the criteria.

            11.1.3   Alternatively,  a spectral search may be used to auto-
                     matically reduce the data.   The criteria for acceptable
                     identification include a high index of similarity.

      11.2  Selected ion monitoring (SIM) or limited mass scan (LMS) data -
            The identification of a compound as a given PCB homolog requires
            that two criteria be met:

            11.2.1   (1) The peak must elute within the retention time window
                     set for that homolog (Section 7.5); and (2) the ratio of
                     two ions obtained by LMS (Table 10) or by SIM (Table 11)
                     must match the ratio observed for a standard within ± 20%.
                     The analyst should search the higher mass windows,  in
                     particular M+70,  to prevent misidentification of a PCB
                     fragment ion cluster as the parent.  System noise at low
                     concentration or interferences may skew the ion ratio
                     beyond the ± 20% criteria.   If the analyst's best judge-
                     ment is that a peak, which does not meet the qualitative
                     criteria, is a PCB, the peak may be included in the cal-
                     culation, with a footnote explaining the data and the
                     reason for relaxing the criteria.

            11.2.2   If one or the other of these criteria is not met, inter-
                     ferences may have affected the results, and a reanalysis
                     using full scan EIMS conditions is recommended.

      11.3  Disputes in interpretation - Where there is reasonable doubt as
            to the identity of a peak as a PCB,  the analyst must either iden-
            tify the peak as a PCB or proceed to a confirmational analysis
            (see Section 13.0).
                                     38

-------
12.0  Quantitative Data Reduction

      12.1  After a chromatographic peak has been identified as a PCB,  the
            compound is quantitated based on the integrated abundance of
            either the EICP or the SIM data for the primary characteristic
            ion in Tables 10 and 11.   If interferences are observed for the
            primary ion, use the secondary and then tertiary ion for quanti-
            tation.  If interferences in the parent cluster prevent quantita-
            tion, an ion from a fragment cluster (e.g., M-70) may be used.
            Whichever ion is used, the RF must be determined using that ion.
            The same criteria should be applied to the surrogate compounds
            (Table 14).

            Note:  With the higher homologs, the mass defect from unity is
            significant.  For instance, the mass of the most intense peak for
            decachlorobiphenyl is 497.6830.  Areas, EICPs, etc. must be based
            on the true mass, not the nominal mass, or erroneous results may
            be obtained.  In addition, the tuning of some quadrupoles may be
            less stable at high masses.  The data quality must be monitored
            especially carefully for the higher homologs.

      12.2  Using the appropriate analyte-internal standard pair and response
            factor (RF ) as determined in Section 7.3, calculate the concen-
            tration oreach peak using Equation 12-1.

                                         A      -.    M.
                  Concentration ((jg/g) = -f- ' np— * M—            Eq. 12-1
                                          is     p    e

            where     A  = area of the characteristic ion for the analyte PCB
                       P     peak

                     A.  = area of the characteristic ion for the internal
                             standard peak

                     RF  = response factor of a given PCB congener

                     M.  = mass of internal standard added to sample extract
                             (micrograms)

                      M  = mass of sample extracted (grams)

      12.3  If a peak appears to contain non-PCB interferences, which cannot
            be circumvented by a secondary or tertiary ion, either:

            12.3.1   Reanalyze the sample on a different column which sepa-
                     rates the PCB and interferents;

            12.3.2   Perform additional chemical cleanup (Section 9) and then
                     reanalyze the sample; or

            12.3.3   Quantitate the entire peak as PCB.
                                     39

-------
Table 14.  Characteristic Ions for Internal Standards
           and 13C-Labeled PCB Surrogates
Ion (relative intensity)
Compound
chlorobiphenyl
1-Iodonaphthalene
d12-Chrysene
13C612C6H9C1
13C12H6C14
13C12H2C18
13C12C110
Abbreviation
d6-C!4
INAP
DCRY
"c-cu
13C-C14
13c-ci8
13c-ci10
Primary
298 (100)
254
240
194 (100)
304 (100)
442 (100)
510 (100)
Secondary
300 (49)
127
-
196 (33)
306 (49)
444 (65)
512 (87)
Tertiary
296 (78)
-
-
-
302 (78)
440 (89)
514 (50)
                         40

-------
      12.4  Calculate the recovery of the four 13C surrogates using the ap-
            propriate surrogate-internal  standard pair and response factor
            (RF. ) as determined in Section 7.4 using Equation 12-2.
                I o


                   Recovery (%) = ^- •  ^- •  ^ •  100             Eq. 12-2


            where A  = area of the characteristic ion for the surrogate peak
                   o

                 RF  = response factor for the surrogate compound with respect
                         to the internal  standard (Equation 7-2)

                  M  = mass of surrogate added to original sample (micrograms)

            Other terms are the same as  defined in Equation 12.1.

      12.5  Sum all of the peaks for each homolog, and then sum those to yield
            the total PCB concentration  in the sample.  Report all numbers in
            |jg/g.  The worksheet in Table 15 and reporting form in Table 16
            may be used.   The concentrations and percent recovery of the four
            surrogates are to be reported.

            If an alternate reporting format (e.g.,  concentration per peak)
            is desired, a different report form may be used.   For example,
            if the PCB analysis is being conducted to satisfy regulatory re-
            quirements for by-product PCBs, results  may need be reported on
            a per resolvable chromatographic peak basis.   One rule1 states
            that PCBs in products or wastes must be  below the practical limit
            of quantitation, defined as  "2 micrograms per gram (2 ppm) per
            resolvable chromatographic peak."   For regulatory purposes, only
            peaks greater than the 2 ug/g cutoff may need to be reported.   A
            second rule2 requires reporting total PCBs, with a regulatory
            cutoff at an annual average  of 25  ppm with a 50 ppm maximum.

      12.6  Round off all numbers reported to  two significant figures.


13.0  Confirmation

      If there is significant reason to  question the qualitative identifica-
      tion (Section 11),  the analyst may choose to confirm that a peak is not
      a PCB.   Any technique may be chosen provided that it is validated as
      having equivalent or superior selectivity and  sensitivity to GC/EIMS.
      Some candidate techniques include  alternate GC columns (with EIMS detec-
      tion),  GC/CIMS, GC/NCIMS, high resolution EIMS, and MS/MS techniques.
      Each laboratory must validate confirmation techniques to show equivalent
      or superior selectivity between PCBs and interferences and sensitivity
      (limit of quantitation, LOQ).
                                     41

-------
                        Table 15.  Analysis Worksheet
          BY-PRODUCT PCBs IN COMMERCIAL PRODUCTS OR PRODUCT WASTES
Internal Sample No.
Notebook No. 	
Data File Code
            External Sample No.
            Date Prepared 	
            Date Analyzed 	
Amount Extracted
                                        Final Volume
Internal
standards   Cone, (ug/9)   Quant, ion   Ratio    Scan no.
d6Cl4
INAP
DCRY
298,300
  254
  240
                                        100/49
                                       ml
                                       Check std. to
                                        be used for
                                Area   calculations
Surrogate
compounds Cone,  (ug/g)    Ions
                                   Ratio
                 Scan no.
                                                          Area
% Recovery
-c-cu
13C-C14
13c-ci8
13c-ci10
Analyte
homolog
194,196
304,306
442,444
510,512
Ions
100/33
100/49
100/65
100/87
Analyte
Ratio homolog Ions Ratio
1-C1
2-C1
3-C1
4-C1
5-C1
188,190
222,224
256,258
292,290
326,328
100/33
100/66
100/99
100/76
100/66
6-C1
7-C1
8-C1
9-C1
10-C1
360,362
394,396
430,432
464,466
498,500
100/82
100/98
100/66
100/76
100/87
Homolog   Scan no. (s)   Area(s)   Ratio   Total area    Ion  used    RF    Cone.  (|jg/g)
                                      42

-------
                            Table 15 (continued)
          BY-PRODUCT PCBs IN COMMERCIAL PRODUCTS OR PRODUCT WASTES

Internal Sample No.  	    External Sample No. 	
Data File Code
Homolog   Scan no.(s)   Area(s)   Ratio   Total area   Ion used   RF   Cone. (M9/9)
                                     43

-------
                         Table 16.  Analysis Report
          BY-PRODUCT PCBs IN COMMERCIAL PRODUCTS OR PRODUCT WASTES
Internal Sample No.
Notebook No. 	
Data File Code
Amount Extracted 	 c
Extraction/Cleanup Procedure
                            External Sample No.
                            Sample Source 	
                            Sample Matrix 	
                            Final Volume
Analyte
homolog

 1-C1
Concentration
                                          Analyte
                                          homolog

                                           6-C1
Concentration
 2-C1
                               7-C1
 3-C1
                               8-C1
 4-C1
                               9-C1
 5-C1
                              10-C1
Surrogate
compound      Concentration (|jg/g)
                                          Total


                                          Surrogate
                                          compound

                                          13C-C18
                                            Concentration (|jg/g)
13C-C1.
                              13c-ci10
Estimated Method LOQ
Highest concentration per resolvable chromatographic peak
Reported by:
              Internal Audit:
                                                      EPA Audit:
        Name
                         Name
                                                                Name
    Signature/Date
                   Signature/Date
                                                           Signature/Date
     Organization
                    Organization
                                                            Organization
                                      44

-------
      If a peak is confirmed as being a non-PCB, it may be deleted from the
      calculation (Section 12).  If a peak is confirmed as containing both
      PCB and non-PCB components,  it must be quantitated according to Section
      12.3.
14.0  Quality Assurance

      Each participating laboratory must develop a quality assurance plan
      (QAP) according to EPA guidelines.9  Additional  guidance is also avail-
      able.10  The quality assurance plan must be submitted to the Agency (re-
      gional QA officer) for approval prior to analysis of samples.

      The elements of a QAP include:

      •   Title Page

      •   Table of Contents

      •   Project Description

         Project Organization and Responsibility

      •   QA Objectives for Measurement Data in Terms of Precision, Accuracy,
         Completeness, Representativeness, and Comparability

      •   Sampling Procedures

      •   Sample Custody

      •   Calibration Procedures and Frequency

      •   Analytical Procedures

      •   Data Reduction, Validation and Reporting

         Internal Quality Control Checks

      •   Performance and System Audits

      •   Preventive Maintenance

      •   Specific Routine Procedures Used to Assess Data Precision,  Accuracy
         and Completeness

      •   Corrective Action

      •   Quality Assurance Reports to Management
                                     45

-------
15.0  Quality Control

      15.1  Each laboratory that uses this method must operate a formal  qual-
            ity control  (QC) program.  The minimum requirements of this  pro-
            gram consist of an initial  and continuing demonstration of lab-
            oratory capability by the analysis of check samples.   The labora-
            tory must maintain performance records to define the quality of
            data that are generated.

      15.2  Certification and performance checks - Prior to the analysis of
            samples, the laboratory must define its routine performance.   At
            a minimum, this must include demonstration of acceptable response
            factor precision with at least three replicate analyses; and anal-
            ysis of a blind QC check sample (e.g., the response factor cali-
            bration solution at unknown concentration submitted by the QA
            officer).  Acceptable criteria for the response factor precision
            and the accuracy of the QC check sample analysis must be pre-
            sented in the QA plan.

            Ongoing performance checks should consist of periodic repetition
            of the initial demonstration or more elaborate measures.  More
            elaborate measures may include control charts and analysis of QC
            check samples consisting of other congeners or with matrix inter-
            ferences.

      15.3  Procedural QC - The various steps of the analytical procedure
            should have quality control measures.   These include but are not
            limited to:

            15.3.1   GC performance - See Section 7.1 for performance criteria.

            15.3.2   MS performance - See Section 7.2 for performance criteria.

            15.3.3   Qualitative identification - At least 10% of the PCB
                     identifications, as well as any questionable results,
                     should be confirmed by a second mass spectrometrist.

            15.3.4   Quantisation - At least 10% of all manual calculations,
                     including peak area calculations, must be checked.   After
                     changes in computer quantisation routines, the results
                     should be manually checked.

      15.4  Sample QC - Each sample and each sample set must have QC measures
            applied to it to establish the data quality for each analysis re-
            sult.   The recoveries of the surrogates, general spectral data
            quality, and consistency of the internal standard area are all
            measures of the data quality on individual samples.  Within  a
            sample set,  analysis of replicates and standard addition samples
            are measures of the precision and accuracy, respectively.
                                     46

-------
15.4.1   The surrogate recoveries must be within an acceptable
         range, as defined in the QAP, for each sample.   Typically,
         recoveries below 50% are unacceptable, and indicate that
         the sample(s) must be reanalyzed.   Recoveries much greater
         than 100% indicate interferences,  improper tuning, a
         problem with response factors, or an error in the concen-
         tration of the surrogate compound.   The trend of low re-
         coveries can be indicative of the cause of the loss.   If
         the lower homologs are poorly recovered, but the higher
         homologs are recovered quantitatively, volatility or
         chemical degradation losses may be suspected.   If higher
         homologs are selectively lost, or if the losses are ir-
         regular, a fractionation cut on a chromatographic clean-
         up would be a likely suspect to account for the loss.

15.4.2   The general spectral data quality is indicative of the
         overall reliability of the data for a sample.   The levels
         of the background, intensity ratios within chlorine clus-
         ters, etc., must all be evaluated.   If the data quality
         is marginal, the analyst may footnote results with an
         explanation regarding any doubts about the data quality.
         If the data are unacceptable (see Section 11.0), either
         the GC/MS determination or the entire analysis must be
         repeated.

15.4.3   An easy and significant assessment of the data quality
         is the consistency of the internal  standard areas.  If
         the internal standard area is consistent, the injection
         volume was correct and the system is operating within
         general tolerances (i.e., the chromatography column is
         transmitting compounds and the spectrometer is detecting
         them).  If the internal standard area does not meet the
         criteria specified in the QAP, e.g., ± 20% of other in-
         jections, the data must be reviewed.  If the injection
         or the GC/MS performance is suspect, the sample should
         be reanalyzed, or other corrective action taken.

15.4.4   QC for small sample sets - For small sample sets (1-10
         samples), the minimum QC requirements can be a heavy bur-
         den.  Analysts are encouraged to be efficient and group
         similar samples to increase the size of a set.   A set is
         defined as a group of samples analyzed together by the
         same extraction/cleanup technique and determined on the
         GC/MS system on the same day or successive days under
         the same conditions.

         At least one method blank must be run.   The blank must
         be exposed to the same sources of contamination—solvent,
         glassware, etc.--as the samples.   If conditions change,
         additional blanks must be generated.  An example would
         be a new lot of solvent, or a change in dishwashing
         protocol.
                         47

-------
               At least one sample must be run in replicate.   Tripli-
               cates are preferable, but duplicates may be acceptable.
               The acceptable precision among replicates must be speci-
               fied in the QAP.

               At least one sample must be analyzed by the standard
               addition technique.  The analyst may select the most
               difficult sample, based on prior knowledge of the sample
               set, or a random sample.   Two aliquots of the sample are
               analyzed, one "as is" and one spiked (surrogate spiking
               and equilibration techniques are described in Section
               9.2) with Solution "FSxxx ng PCB" or "SIMxxx pg PCB."
               If the analyst has no prior knowledge of the sample, the
               spiking level should be in the middle of the calibrated
               range for the mass spectrometer.   If the concentrations
               of PCBs are known to be high or low, the amount added
               should be adjusted so that the spiking level is 1.5 to
               4 times the measured PCB level in the unspiked sample.
               The samples should be analyzed together and the quanti-
               tative results calculated.   The recovery of the spiked
               compounds (calculated by difference) must be 70-130%.
               If the sample is known to contain specific PCB isomers,
               these isomers may be substituted for solution FSxxx ng
               PCB or SIMxxx pg PCB.

      15.4.5   QC for intermediate sample sets - With intermediate
               (approximately 10-100 samples) sample sets, the number
               of method blanks, replicates, and standard addition sam-
               ples must comprise at least 10% each.

      15.4.6   QC for large sample sets - When a large sample analysis
               program is being planned, the QA plan may propose spe-
               cific QC measures.   If none are proposed, the guidelines
               for intermediate sets may be followed.   One QC measure
               which may increase efficiency is the use of control
               charts.   If, for example, the control  charts establish
               that there is no blank problem over the long term,  the
               percent of blanks may be reduced.   Any changes in the
               procedure (e.g.,  a new lot of solvent) will still,  of
               course,  require a blank.

15.5  It is recommended that the participating laboratory adopt addi-
      tional QC practices for use with this method.   The specific  prac-
      tices that are most productive depend upon the needs of the  lab-
      oratory and the nature of the samples.  Field duplicates or  trip-
      licates may be analyzed to monitor the precision of the sampling
      technique.   Whenever possible, the laboratory should perform
      analysis of standard reference materials and participate in  rele-
      vant performance evaluation studies.
                               48

-------
16.0  Method Performance

      The method has been evaluated using both intra- and inter!aboratory
      studies.3'11  Both studies must be considered preliminary;  only a
      limited number of matrices and method options have been investigated.
      The preliminary inter!aboratory study involved only four participants
      and was conducted before the method was ruggedized, so data from that
      study may not be representative of the potential method performance.
      Further intra- and interlaboratory validation is anticipated.   Prelimi-
      nary values for limits of quantisation; intralaboratory recoveries, pre-
      cision, and accuracy; and interlaboratory recoveries, precision, and
      accuracy are presented in Table 17.  The values in the "best"  and "worst"
      columns represent typical extremes of the measurement, ignoring excep-
      tional cases.   For example, some mass spectrometers, when running under
      optimal conditions, can probably quantitate less than 0.5 ng/uL in the
      full scan mode for some matrices and PCB congeners.  On the other hand,
      the method LOQ will be much greater than the stated value if a high
      concentration coeluting interference completely obscures any PCB signal,
      even at the percent level.

      The performance values in Table 17 were derived using either the SIM or
      LMS options of the method and at concentrations at the lower end of the
      working range.  Working in the middle-to-upper concentration ranges with
      full scan data collection, the precision, recovery, and accuracy should
      all improve considerably.

      The values in Table 17 represent best estimates of the parameters and
      will be refined as additional intra- and interlaboratory studies pro-
      duce more data.  Performance better than or worse than any of these
      parameters cannot, of itself, be construed as grounds for acceptable or
      unacceptable data quality.  Performance criteria should be stipulated
      in the QAP.


17.0  Documentation and Records

      Each laboratory is responsible for maintaining full records of the
      analysis.   A detailed documentation plan should be prepared as part of
      the QAP.   Laboratory notebooks should be used for handwritten  records.
      GC/ MS data must be archived on magnetic tape, disk, or a similar device.
      Hard copy printouts may be kept in addition if desired.   QC records
      should be maintained separately from sample analysis records.

      The documentation must describe completely how the analysis was performed.
      Any variances from the protocol must be noted and fully described.   Where
      the protocol lists options (e.g.,  sample cleanup), the option  used and
      specifics (solvent volumes, digestion times, etc.) must be stated.

      The remaining samples and extracts should be archived for at least 2
      months or until the analysis report is approved, whichever is  longer,
      and then disposed unless other arrangements are made.   The  magnetic
      tapes of the analysis and hardcopy spectra, quantitation reports,  work
      sheets, etc.,  must be archived for at least 3 years.


                                     49

-------
                  Table 17.   Method Performance Parameters
Measurement
Best case
Worst case   Typical value
Instrumental LOQ (ng/ul)
Full scan
Limited mass scan
Selected ion monitoring
Sample concentration factor
Matrix interference level (ng/ul)
Sample injection volume (ul)
Method LOQ (ug/g)
b

0.5
0.05
0.01
1,000+1
< 0.01
10
io"5


10
0.5
0.2
1*100
> 1,000
0.1
500


1
0.1
0.05
_a
_a
1
1

     Recovery (%)
     Precision (%)
     Accuracy
               b e
Interlaboratory '
     Recovery (%)
     Precision (%)
     Accuracy
 90 ± 15C     < 50, > 200d    70-130
                              ± 34%C
                 22-690
                              ± 60'
                              ± 629
.Varies widely from matrix-to-matrix, no "typical value."
 Data for preliminary validation at low (< I ug/g) levels using selected ion
monitoring mass spectrometry.
 "Methods of Analysis for By-Product PCBs—Preliminary Validation and Interim
Methods," M. D. Erickson, J. S. Stanley, G. Radolovich, K. Turman, K. Bauer,
J. Onstot, D. Rose, and M. Wickham, Interim Report No. 4, Washington D.C.:
Office of Toxic Substances, EPA-500/5-82-006, October 1982, 243 pp.  NTIS No.
RB83 127 696.
 Values outside the 50-200% range are generally considered unacceptable and
the analysis must be repeated.
 Preliminary interlaboratory study involving four participants:  "Analytical
Methods for By-Product and Destruction Derived PCBs—Interlaboratory Valida-
tion A," M. D.  Erickson, K. M. Bauer, and F.  J. Bergman, Draft Interim Report
No. 6, Washington, D.C.:  Office of Toxic Substances, U.S. Environmental Pro-
tection Agency, Draft Interim Report No. 6, Task 51, EPA Contract No. 68-01-
5.915, August 1983.
 Based on a mean of the 10 homolog values from the analysis of two samples by
four laboratories.
%ased on the deviation of the reported values from the prepared value for 10
homologs x 2 samples x 4 laboratories.
                                     50

-------
                                 REFERENCES

 1.  USEPA.  1982.  40 CRF 761, Polychlorinated biphenyls (PCBs); manufactur-
     ing, processing, distribution in commerce, and use prohibitions; use in
     closed and controlled waste manufacturing processes.  47 FR 46980-46996.

 2.  USEPA.  1984.  40 CFR Part 761, Polychlorinated biphenyls (PCBs); manu-
     facturing, processing, distribution commerce and use prohibitions; re-
     sponse to individual and class petitions for exemptions.  49 FR 28154-28209.

 3.  Erickson MD, Stanley JS, Radolovich G, Turman K, Bauer K, Onstot J,
     Rose D, Wickham M.  1982.  Analytical methods for by-product PCBs--
     preliminary validation and interim methods.  Washington, DC:  Office of
     Toxic Substances, U.S. Environmental Protection Agency.  EPA-500/5-82-006;
     NTIS No. PB83 127 696.

 4.  Erickson MD, Stanley JS, Radolovich G, Blair RB.  1983.  Analytical method:
     the analysis of by-product chlorinated biphenyls in commercial products
     and product wastes.  Revision 1, Prepared by Midwest Research Institute
     for Office of Toxic Substances, U.S. Environmental Protection Agency,
     Washington, DC, under Subcontract No. A-3044(8149)-271, Work Assignment
     No. 17 to Battelle, Washington, DC, August 15, 1983.

 5.  Erickson MD, Stanley JS.  1982.  Methods of analysis for incidentally
     generated PCBs—literature review and preliminary recommendations.
     Washington, DC:  Office of Toxic Substances, U.S. Environmental Pro-
     tection Agency.  EPA-560/5-82-005; NTIS No. PB83 126573.

 6.  USEPA.  1979.  Methods 330.4 (titrimetric, DPD-FAS) and 330.5 (spec-
     trophotometric, DPD) for chlorine, total residual.  Methods for Chemical
     Analysis of Water and Wastes, U.S. Environmental Protection Agency,
     Environmental Monitoring and Support Laboratory, Cincinnati, OH, EPA
     600-4/79-020.

 7.  Bellar TA, Lichtenberg JJ.  1981.   The determination of polychlorinated
     biphenyls in transformer fluid and waste oils.  Prepared for U.S. Environ-
     mental Protection Agency, EPA-600/4-81-045.

 8.  American Society for Testing and Materials.  1980.  Standard method for
     analysis of environmental materials for polychlorinated biphenyls.  In:
     Annual book of ASTM standards, 1980.  Philadelphia, PA:  ANSI/ASTM D
     3304 - 77.

 9.  USEPA.  1980.  Guidelines and specifications for preparing quality as-
     surance project plans. Office of Monitoring Systems and Quality Assur-
     ance, QAMS-005/80.

10.  USEPA.  1983.  Quality assurance program plan for the Office of Toxic
     Substances.   Office of Pesticides  and Toxic Substances, U.S. Environ-
     mental Protection Agency, Washington, D.C.
                                     51

-------
11.   Erickson MD, Bauer KM, Bergman FJ.   1983.   Analytical  methods for by-
     product and destruction derived PCBs--interlaboratory validation A.
     Draft Interim Report No.  6,  Task 51, EPA Contract No.  68-01-5915, Office
     of Toxic Substances, U.S.  Environmental  Protection Agency,  Washington,
     DC.
                                     52

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
 EPA-560/5-85-010
                                                           3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
 Analytical  Method:   The Analysis of By-product  Chlorin-
 ated Biphenyls  in  Commercial Products and Product
 Wastes, Revision  2                          	
                                  5. REPORT DATE
                                    May  1985
                                  6. PERFORMING ORGANIZATION CODE

                                    8201A06
7. AUTHOR(S)

 Mitchell  D.  Erickson
                                                           8. PERFORMING ORGANIZATION REPORT NO.
                                    Special  Report No. 1
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Midwest  Research  Institute
 425 Volker  Boulevard
 Kansas City,  MO  64110
                                   10. PROGRAM ELEMENT NO.

                                    Work  Assignment  No. 6
                                   11. CONTRACT/GRANT NO.

                                    68-02-3938
 12. SPONSORING AGENCY NAME AND ADDRESS
 Field Studies  Branch (TS-798), Office of Toxic  Substances
 U.S. Environmental  Protection Agency
 401 M Street,  SW
 Washington,  DC  20460	 '
                                   13. TYPE OF REPORT AND PERIOD COVERED
                                   Special  (September 84 - May 85]
                                   14. SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES
 The EPA Work  Assignment Manager is Daniel T.  Heggem (202) 382-3990.
 The EPA Project  Officer is Joseph J. Breen  (202)  382-3569.
 16. ABSTRACT
 This  is a gas chromatographic/electron impact mass spectrometric  (GC/EIMS) method ap-
 plicable to the determination of  chlorinated biphenyls (PCBs)  in  commercial products
 and product wastes.  The PCBs present may originate either as  synthetic by-products or
 as contaminants derived from commercial  PCB products (e.g., Aroclors).   The PCBs may be
 present as single isomers or complex mixtures and may include  all  209 congeners from
 monochlorobiphenyl through decachlorobiphenyl.

 A variety of general and specific sample preparation options are  presented in this
 method.   This method takes a different approach from those which  rely on Aroclor mix-
 tures for calibration and quantisation.   In this method PCBs are  detected and quanti-
 tated by homolog group.  The results can be summed to give a total  PCB value comparable
 to results generated by other methods or they may be presented as  10  individual homolog
 values.   This homolog distribution can provide additional quantitative information on
 the composition and source of the PCBs.

 The method performance is assessed for each sample.   A set of  four 13C-labeled PCBs is
 employed as recovery surrogates.   If the surrogates are recovered  and other QC param-
 eters are within acceptable limits,  then the data may be considered valid.
 7.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                                                c. COSATI r'ield/Group
 PCBs
 Polychlorinated biphenyls
 Chlorinated biphenyls
 Analytical methods
 Determination
 By-products
 GC/MS
Commercial products
Wastes
 8. DISTRIBUTION STATEMENT
       Unlimited
                                              19. SECURITY CLASS (This Report)
                                                UNCLASSIFIED
                                                21. NO. OF PAGES
                                                     58
                                              20. SECURITY CLASS (This page)
                                                UNCLASSIFIED
                                                22. PRICE
EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE

-------