v>EPA
            United Stales
            Environmental Protection
            Agency
Office of
Toxic Substances
Washington DC 20460
EPA-560/5-85-012
April. 1985
            Toxic Substances
            Analytical Method:
            The Analysis of
            By-Product Chlorinated
            Biphenyls in Water,
            Revision 2

-------
ANALYTICAL METHOD:   THE ANALYSIS OF BY-PRODUCT
  CHLORINATED BIPHENYLS IN WATER, REVISION 2
                      by

             Mitchell D. Erickson
             WORK ASSIGNMENT NO. 6

             SPECIAL REPORT NO. 3

          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 products
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 water.  The work was done on Work Assign-
ment No. 6 on US 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 conducted 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 commercial
products and product wastes (Special Report No. 1, EPA Report No. EPA-560/5-
85-010) and air (Special Report No. 2, EPA Report No. EPA-560/5-85-011).

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

                                        MIDWEST RESEARCH INSTITUTE
                                                 L. Haile
                                                 ogram Manager
                                        Jfchn E. Going
                                        Program Manager
Approved:
James L. Spigarelli, Director
Chemical and Biological Sciences
  Department

-------
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

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 	
References 	
Page
1
3
5
6
6
10
13
24
26

35
37
38
44
44
45
48
48
49

-------
                               LIST OF TABLES
Number
Page
   1      Numbering of PCB Congeners	    2

   2      DFTPP Key Ions and Ion Abundance Criteria	    9

   3      Concentrations of Congeners in PCB Calibration Standards
            for Full Scan Analysis (ng/mL)	11

   4      Concentrations of Congeners in PCB Calibration Standards
            for Selected Ion Monitoring and Limited Mass Scan
            Analysis (pg/uL)	12

   5      Composition of Internal Standard Spiking Solution (SS100)
            Containing 13C-Labeled PCBs 	   14

   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	20

  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	39

  15      Analysis Worksheet	41

  16      Analysis Report	43


                               LIST OF FIGURES

Number                                                                Page

   1      Reconstructed ion chromatogram of calibration solution
            CS100 obtained in the full scan mode	36

-------
              THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS
                          IN INDUSTRIAL WASTEWATER
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 industrial wastewater.   The PCBs present may
            originate either as synthetic by-products or as contaminants de-
            rived from commercial  PCB products (e.g., Aroclors).  The PCBs
            may be present as single isomers or complex mixtures and may in-
            clude all 209 congeners from monochlorobiphenyl through deca-
            chlorobiphenyl listed in Table 1.

            This method was prepared for use in demonstrating compliance with
            the EPA rules regarding the generation of PCBs as byproducts 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 13C-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 vol-
            ume of sample extracted the complexity of the sample matrix and
            the ability of the analyst to remove interferents and properly
            maintain the analytical system.  The method accuracy and preci-
            sion 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.

      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.

-------
                                TABLE 1.   NUMBERING  OF PCB CONGENERS3
No.

1
2
3



4
S
6
7
8
9
10
n
12
13
14
15



IS
17
18
19
20
21
22
23
24
25
26
27
28
29
20
31
32
33
34
35
36
37
38
39



40
41
42
43
44
45
46
47
48
49
50
51

Structure
Honocft 1 ant 1 eheny 1 s
2
3
4

0
-------
            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 steps for certain techniques, or
            totally different extraction or cleanup techniques.
2.0   Summary
      2.1   The wastewater must be sampled such that the specimen collected
            for analysis is representative of the whole.   Statistically
            designed selection of the sampling position (valve,  port,  outfall,
            etc.) or time should be employed.   The sample must be preserved to
            prevent PCB loss prior to analysis.   Storage at 4°C  with optional
            preservation at low pH is recommended.

      2.2   The sample is mechanically homogenized and subsampled if necessary.
            The sample must then be spiked with four 13C PCB surrogates 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.  Possible extraction techniques include
            liquid-liquid partition and sorption onto resin columns followed
            by solvent elution.  Cleanup techniques may include  liquid-liquid
            partition, sulfuric acid cleanup,  saponification, adsorption 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.  The EPA Method
            6086 and 6257 extraction and cleanup procedures may  be used.

      2.4   The PCB content of the sample extract must be determined by high
            resolution (preferred) or packed column gas chromatography/elec-
            tron impact mass spectrometry (HRGC/EIMS or PGC/EIMS) operated in
            the full scan, selected ion monitoring (SIM), or limited mass scan
            (IMS) 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 against the response factors for a mixture
      of 10 PCB congeners, using the response of the 13C surrogate to
      compensate for losses in workup and instrument variability.

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 is
      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, excluding instru-
      mental  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 (instru-
      mental  performance, calculation checks), and sample QC (blanks,
      replicates, and standard addition).

2.11  While several  options are available throughout this method,  the
      recommended procedure to be followed is:

      2.11.1   The sample is collected according to a scheme which per-
               mits extrapolation of the sample data to the body or con-
               tainers of water being sampled.

      2.11.2   The sample is preserved at low pH and at 4°C to prevent
               any loss of PCBs or changes in matrix which may adversely
               affect recovery.

      2.11.3   The sample is mechanically homogenized and subsampled if
               necessary.

      2.11.4   The sample is spiked with four 13C-PCB surrogates
               (4-chlorobiphenyl; 3,3',4,4'-tetrachlorobiphenyl;
               2,2',3,3',5,5',6,6'-octachlorobiphenyl;  and decachloro-
               biphenyl).

      2.11.5   The sample is extracted.

      2.11.6   The extract is  cleaned up, if necessary, and concentrated
               to an appropriate  volume.

-------
            2.11.7   An aliquot of the extract is analyzed by HRGC/EIMS oper-
                     ated in the SIM mode.   On-column injections onto a 15-m
                     DB-5 capillary column,  programmed (for toluene solutions)
                     from 110° to 325°C at 10°/min after a 2 min hold is used.
                     Helium at 45-cm/sec linear velocity is used as the carrier
                     gas.

            2.11.8   PCBs are identified by retention time and mass spectral
                     intensities.

            2.11.9   PCBs are quantitated against the response factors for a
                     mixture of 10 PCB congeners.

            2.11.10  The total PCBs are obtained by summing the amounts for
                     each homolog found and the concentration is reported as
                     micrograms per liter.
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 laboratory
            reagent blanks as  described in Section 15.

            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 solvents
                     by distillation in all-glass systems may be required.
                     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 cardnogenicity 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 labor-
            atory is responsible for maintaining a current awareness file of
            OSHA regulations regarding the safe handling of the chemicals spe-
            cified in this method.   A reference file of material data handling
            sheets should also be made available to all personnel involved in
            the chemical analysis.

      4.2   Polychlorinated biphenyls have been tentatively classified as known
            or suspected human or mammalian carcinogens.   Primary standards
            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 concentration solution.   The 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 perox-
            ide 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 Laboratories
            (available from Scientific Products Company,  Cat.  No. P1126-8 and
            other suppliers) are recommended for this test.  Procedures for
            removal of peroxides from diethyl ether are included in the in-
            structions 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-liter or other ap-
            propriate volume, fitted with screw caps lined with Teflon.
            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 con-
            tainer.  The bottle must be washed, rinsed with acetone or methy-
            lene chloride, and dried before use to minimize contamination.

      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  |jL  for  packed  column  GC  analysis,
               1.0 uL  for  on-column HRGC 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  (J19/22
                        joint) is used to prevent  evaporation of solvent.

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

               5.2.5.3  Snyder column -  Three ball  macro (Kontes K503000-
                        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    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 recommended.  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 GOto-MS interface
               that gives acceptable calibration points at 10 ng per
               injection for each PCB isomer in the calibration standard
               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 deactivated 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
               and plotting such ion abundances versus time  or scan num-
               ber 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 Biobeads  SX-3 (Biorad) 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 (jBondapak, 3.9  x 300 mm)
                        or equivalent.

                              8

-------
   Table 2.   DFTPP Key Ions and Ion Abundance Criteria

m/z                 "~~~~~      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

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

      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.
            Calculate the concentration 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 evapora-
            tion.   Primary dilutions are stable indefinitely if  the seals are
            maintained.  The stock solutions and dilutions should be  clearly
            labeled 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 Table  3.
            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 meniscus 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.

                                    10

-------
     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
C18D12 (IS)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.
Cd12-Chrysene.
                                   11

-------
     Table 4.   Concentrations of Congeners in PCB Calibration Standards
              for Selected Ion Monitoring and Limited Mass  Scan
                              Analysis (pg/pL)
Homolog
I
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
Ci8Di2 (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
pg 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.
                                   12

-------
      6.5   Alternatively,  certified stock solutions similar to those listed
            in Table 3 may  be available from a supplier,  in lieu of the pro-
            cedures 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)  and 1-iodonaphthalene
            (Aldrich Chemical Company, Milwaukee,  WI)  or  d12-chrysene (KOR
            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 quantitation, (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 so-
            lutions should  be replaced after 6 months, or sooner if comparison
            with quality control check samples indicates  compound degradation
            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  of the criteria are not
            met, the analyst must adjust conditions  and repeat the test until
            all criteria 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.
                                    13

-------
Table 5.   Composition of Internal  Standard Spiking Solution (SS100)
                   Containing 13C-Labe1ed PCBs
Congener
no.
211
212
213
214
Compound
4-Chloro-d' ,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-Cli
13C-C14
13c-ci8
13c-ci10
Concentration
((jg/mL)
100
250
400
500

-------
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
None
280°C
0.7-1.5
7-10 s
Other0
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.
 C12C110 elutes at 270°C.  Programming above this temperature ensures a
-rdean column and lower background on subsequent runs.
 Fused silica columns may be routed directly into the ion source to  prevent
 separator discrimination and losses.
^High 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-
.jtion 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
Gas chromatograph
Column
Recommended
Finnigan 9610
180 cm x 0.2 cm ID
Tolerance
Other3
Other
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 100/
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
Other nonpolar
or semipolar

Hydrogen

Optimum performance

Other

Optimum

^ 5 |A

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 energy0                70 eV                      70 eV


^Substitutions permitted if performance criteria are met.
 Greater than five data points over a GC peak is a minimum.
GFilaments 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.   Operating Parameters for Magnetic Sector Mass Spectrometer System

      Parameter                    Recommended                Tolerance


Mass spectrometer               Finnigan MAT 311A          Other3

Data system                     Incos 2400                 Other

Scan range                      98-550                     Other

Scan mode                       Exponential                 Other

Cycle time                      1.2 sec                    Other

Resolution                      1,000                      >  500

Ion source temperature          280°C                      250-300°

Electron energy0                70 eV                      70 eV


^Substitutions permitted if performance criteria are met.
 Greater than five data points over a GC peak is a minimum.
cFilaments should be shut off during solvent elution to improve instrument
stability and prolong filament life, especially if no solvent venting is
used.
                                    18

-------
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 LMS ranges.   The mass spectrometer
         should be set to at least unit resolution.  The com-
         puter acquisition parameters should utilize the minimum
         threshold 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.

         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 required particular attention to the
         switching points to assure minimal data loss.  Switch-
         ing points can be initially determined by analyzing a
         highly concentrated 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 spectrometer should be set to at least
         unit resolution.  The computer acquisition parameters
         should utilize the minimum threshold 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.
                        19

-------
Table 10.   Limited Mass Scanning (LMS) Ranges For PCBs
Compound
r n~n + 13r 12r w n
129 1 6 ^6''9*' '
Ci2H8Cl2
Cl2H7Cl3
C12H6C14 + C12D6 C14 + 13C12H6C14
C12H5C15
C12H4C16
Ci2H3Cl 7
Ci2H2Cl8
C12HClg
^12^10
C10H7I
Ci8Di2
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
                         20

-------
Table 11.  Characteristic SIM Ions for PCBs
Homolog
Ci2HgCl
^12^8^2
Cj^HyCls
C12H6C14
Cia^sCls
Ci2H4Cle
Cia^sCl 7
^12^2^8
Ci2HCl 9
Ci2^1 10
CioHyl
C12D6C14
^18^12
13C612C6H9C1
13C12H6C14
13C12H2C18
13Ci2Cl10

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

-------
7.3   The PCB response factor (RF ) must be determined in triplicate
      or other replicate,  as discussed below,  using Equation  7-1  for
      the analyte homologs.

                      An x M.

                Rpp= t^C

      where    RF  = response factor of a given  PCB isomer

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

                M  = mass of PCB congener in sample (micrograms)

               A.  = area of the characteristic  ion for the internal
                       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
      factors (RF ) must be determined using Equation  7-2.


                  RFs = ST^Aj                          •    E«' 7'2

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

            M  = mass of surrogate  in  sample (micrograms)

      Other items 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, the solutions in Tables 3 and 4 or  a mixture with  a  sim-
      ilar 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%
      precision.  FojMnstruments with good day-to-day precision, a
      running mean (RF) based on seven values  may  be appropriate.
                              22

-------
         Table  12.   Pairings  of  Analyte  and  Calibration  Compounds
Analyte
Congener
no. Compound
1 _ o P H P 1
4-15 C12H8C12
i c.- "3o r u n
ID 03 L^2ri70i3
40-81 C12H6C14
82-127 C12H5C15
128-169 C12H4C16
170-193 C12H3C17
194-205 C12H2C18
206-208 C12HC19
onQ r r~\
^uy ^12^ ' 10
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
22' 3456'
f. ,(. ,0 , t , J ,O
2,2',3I,4,4',5I,6
2,2')3,3')5,5I,6,6I
2,21,3,3',4,41,5,6,61
C12C110
Ballschmiter numbering system, see Table 1.
                                    23

-------
            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 replicate RF values differ by greater than ±10% RSD, the system
            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.

      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.   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 is determined by the amount
            of sample to be collected but will usually be 1 liter or 1 qt.
            The sample size is dependent on the anticipated PCB levels and
            difficulty of the subsequent extraction/cleanup steps.

      8.2   Sample bottle preparation

            8.2.1    All sample bottles and caps should be washed in detergent
                     solution, rinsed with tap water and then with distilled
                     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 allow 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.


                                    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 of
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'-tetrachlorobiphenyl-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   Sample collection

            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.

            8.3.2    If possible, mix the source 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,  sam-
                     pling through a valve should be randomly timed.

            8.3.3    Fill the bottle with water, add preservative (Section
                     8.4), cap tightly, and shake well.  To prevent the caps
                     from working loose during storage tape the caps on with
                     a water-insoluble tape.

      8.4   Sample preservation - Samples should be stored at 4°C.  Since
            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 the pH.  Storage 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.8  Field test kits
            are available for this purpose.


9.0   Sample Preparation

      9.1   Sample homogenization and subsampling - The sample is homogenized
            by shaking, blending, or other appropriate mechanical technique,
            if necessary.  If the density of the sample is not between 0.9
            and 1.1, it should be determined and reported.  Consideration
            should be given to treating the sample as a product waste (see
            separate method)  when the density is outside the range of 0.9
            to 1.1.

            Note:     The precision of the mass determination at this step will
                     be reflected in the overall method precision.  Therefore,
                     an analytical balance must be used to assure that the
                     weight is accurate to ±1% or better.
                                    26

-------
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.

      Note:    The volume measurement of the spiking solution is criti-
               cal to the overall method precision.  The analyst must
               exercise caution that the volume is known to ±1% or
               better.  Where necessary, calibration of the pipet 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 quantisation.  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 matrix
      must be determined by the precision and accuracy of the 13C PCB
      surrogate recoveries, as discussed in Section 15.

      9.3.1    Extraction - The entire sample must be transferred to the
               extraction vessel with PCB-free water washing, if neces-
               sary, to transfer all solids.  The container is then
               rinsed with the extraction solvent to recover any PCBs
               adhering to the container wall.  The solvent rinses are
               combined with the extracts from below.  Measure the sam-
               ple volume to the nearest 0.5%.

               9.3.1.1  Liquid-liquid extraction - The solvent, number
                        of extractions, solvent-to-sample ratio, and
                        other parameters are chosen at the analyst's
                        discretion.  A suggested extraction from water
                        is presented in EPA Methods 60S6 and 625.7

               9.3.1.2  Sorbent column extraction - PCBs may be isolated
                        from water onto sorbent columns, although these
                        techniques are not as widely used or thoroughly
                        validated as liquid-liquid extraction.  The
                        selection of sorbent (XAD, Porapak, carbon-
                        polyurethane foam, etc.) will depend on the
                        nature of the matrix.  The available methods
                        have been reviewed.8

      9.3.2    Cleanup - Several optional tested cleanup techniques are
               described below.  All but the base cleanup (9.3.2.8) were
               previously validated for PCBs in transformer fluids.9
               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.
                              27

-------
With many water samples, the extract is clean enough
for analyis without cleanup.  The analyst may decide
to either go directly to the GC/EIMS determination
(Section 10) or screen the sample for interferences
(Section 9.4).

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, trans-
                    fer 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 of the PCBs must be veri-
                    fied by the analyst.   The weight of
                    Florisil can then be adjusted accor-
                    dingly.

         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 sul-
                    fate 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.

               28

-------
         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 the hexane.

         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 of 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 (wtrwt) and
                    mix until uniform.  Allow the  de-
                    activated 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 alu-
                    mina 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 volume of
                    hexane.
               29

-------
         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.

         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.
               30

-------
         9.3.2.4.9  Analyze the sample according to Sec-
                    tion 10.0.

9.3.2.5  Gel  permeation cleanup

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

         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 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 for combination
                    with other extracts in Section 9.3.2.6.3.

         9.3.2.6.2  Combine and transfer the acetonitrile
                    phases to a 1-liter 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 from Sec-
                    tions 9.3.2.6.1 and 9.3.2.6.2 in a
                    1-liter separatory funnel and wash
                    with two 100-mL portions of distilled
                    water.  Discard the water layer and
                    pour the hexane layer through a 8-10
                    cm anhydrous sodium sulfate column
               31

-------
                    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
                    appropriate volume.

         9.3.2.6.5  Analyze as described in Section 10.0.

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 cleanup10

         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 dry-
                    ness with a gentle  stream of dry
                    filtered nitrogen,  and add 25 ml of
                    2.5% alcoholic 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.
               32

-------
         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.

         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 ac-
                    cording to 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 uBondapak
                    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.
               33

-------
                        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 in-
      frequently encountered by the analyst, the effectiveness of the
      extraction and cleanup for a matrix may be unknown.   A simple
      screen to assess 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.

      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 100/
               120 Supelcoport or equivalent is suggested.   A flow rate
               of 40 mL/min 95% air/5% methane or nitrogen is recom-
               mended.  The air and hydrogen flow rates should be suf-
               ficient 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 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 sample at a given
               attenuation such as (a) any off-scale peaks in PCB
               elution window, (b) a baseline rise of 40% full scale,
               (c) 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 (a)
               cycle the sample through the same cleanup again if it
               appears that the cleanup technique was overloaded by the
               matrix the first time, (b) submit the extract to another
               and cleanup techniques which may remove more interfer-
               ences, or (c) analyze a new aliquot of sample by another
               extraction or cleanup technique.
                              34

-------
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 better.
            Where necessary, calibration of the pipet is recommended.

      10.2  Tables 2, and 6 through 11 summarize the recommended operating con-
            ditions for analysis.  Figure 1 presents an example of a chromato-
            gram.

            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 instru-
            mental LOQ needed meet the required method LOQ.  In genreal, the
            more concentrated the PCBs, the greater the precision, accuracy,
            and qualitative data confidence.  Thus, if possible, the amount
            of sample and 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, analyte,
            or internal standard are below the working range, recheck the
            system performance.  If necessary, concentrate the sample and
            reanalyze.

      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.


                                    35

-------
 100.0-1
   RIC
OJ
                                                             13
                                   C12D6C14


                                   C12H6C14
                       3

                 13C612C6H9C1
                       I NAP
               50
                                  30
                                  97
                              183
                                         143
                                                                                        88960.
                                                                                    202
                                                                                    DCRY

                                                                                    13
                                                                                      C12H2C18
                                                                                                     209

                                                                                                     13
                                                                                            207
         13:29
1000
16i40
1200
20:00
  I
1400
23:20
                                                                                                         2C110
1600
26:40
1800
30:00
SCAN
TIHE
     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 in-
     ternal 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 resolution and peak shape of the internal  standards,  surrogates,
            and other peaks should be monitored during or  immediately after
            data acquisition.   Poor chromatography may indicate a bad injec-
            tion, matrix interferences,  or column degradation.

      10.7  If a "dirty" sample is encountered, the analyst must employ ap-
            propriate measures to demonstrate that there is no  memory or
            carry-over to subsequent samples.   To assess the system cleanli-
            ness, a standard,  blank sample,  or solvent blank may be run.

            If the system is contaminated, remedial efforts may include (a)
            changing or cleaning the syringe,  (b) cleaning the  injector,  (c)
            baking out the column at its maximum temperature, (d) changing
            to a new column, or (e) cleaning the ion source.


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 judgement
                     is that a peak, which does not meet the qualitative cri-
                     teria, is a PCB, the peak may be included in the calcu-
                     lation, with a footnote explaining the data and the rea-
                     son 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 (IMS) data -
            The identification of a compound as a given PCB homolog requires
            that two criteira be met:

            11.2.1   (1) The peak must elute within the retention time window
                     set for that homolog (Seciton 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%.
                                    37

-------
                     The analyst must search the higher mass windows, in
                     particular M+70, to prevent mi sidentification 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 judg-
                     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
            identify the peak as a PCB or proceed to a confirmational analysis
            (see Section 13.0).


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 quadru-
            poles 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 of^each peak using Equation 12-1.

                                        A      ,    M.
                 Concentration (pg/L) = ^- • HF- • rj1^             Eq.  12-1
                                        Ais   %   Vs

            where     A  = area of the characteristic ion for the analyte PCB
                       p     peak
                                    38

-------
Table 14.  Characteristic Ions for Internal Standards
           and 13C-Labeled PCB Surrogates
Ion (relative intensity)
Compound
dg-3,4,31 ,4'-Tetra-
chlorobiphenyl
1-Iodonaphthalene
d12~Chrysene
13C612C6H9C1
13C12H6C14
13C12H2C18
13C12C110
Abbreviation
d6-Cl4
INAP
DCRY
13C-Cl!
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)
                        39

-------
               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)

                V  = volume of sample extracted (liters)
                 o

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.

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 7.3 using Equation 12-2.

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


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

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

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

    and 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/L.   The worksheet in Table 15 and reporting form in Table 16
      may be used.   The uncorrected concentrations and percent recovery
      of 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
      requirements for by-product PCBs, results should be reported
      on a per resolvable chromatographic peak basis.   The rules1'2
      state that PCBs in products or wastes must be  below the  practical
      limit of quantitation, defined as 100 [jg/L (roughly 0.1  ppm) per
      resolvable chromatographic peak."  For regulatory purposes,  only
      peaks greater than the 100 ng/L cutoff may need  to be reported.

                              40

-------
                        Table 15.   Analysis Worksheet
Internal Sample No.
Notebook No.
Data File Code
Amount Extracted
Internal
standards Cone, (ng/g)
d6Cl4
INAP
DCRY
Surrogate
compounds Cone, (ng/g)
13C-Cli 194,196
13C-C14 304,306
13C-C18 442,444
13C-C110 510,512
Analyte
homolog Ions
1-C1 188,190
2-C1 222,224
3-C1 256,258
4-C1 292,290
5-C1 326,328
BY-PRODUCT


ml
Quant, ion
298,300
254
240
PCBs IN WATER
External Sample No.
Date Prepared
Date Analyzed
Final Volume ml
Check std. to
be used for
Ratio Scan no. Area calculations

100/49


Ions Ratio Scan no. Area % Recovery
100/33
100/49
100/65
100/87
Ratio
100/33
100/66
100/99
100/76
100/66




Analyte
homolog Ions Ratio
6-C1 360,362 100/82
7-C1 394,396 100/98
8-C1 430,432 100/66
9-C1 464,466 100/76
10-C1 498,500 100/87
Homolog   Scan no.(s)   Area(s)   Ratio   Total area   Ion used   RF   Cone, (ug/g)
                                    41

-------
                            Table 15 (continued)
                          BY-PRODUCT PCBs  IN WATER

Internal Sample No. 	    External Sample No. 	
Data File Code 	


Homolog   Scan no. (s)   Area(s)   Ratio    Total area   Ion  used    F!F    Cone,  (ug/g)
                                     42

-------
                         Table 16.   Analysis Report
                          BY-PRODUCT PCBs IN WATER
Internal Sample No.
Notebook No.  	
Data File Code
Amount Extracted 	 ml_
Extraction/Cleanup Procedure
                            External Sample No.
                            Sample Source 	
                            Sample Matrix 	
                            Final Volume
                                                                   mL
Analyte
homolog

 1-C1
Concentration (ug/L)
                                          Analyte
                                          homolog

                                           6-C1
Concentration (pg/L)
 2-C1
                               7-C1
 3-C1
                               8-C1
 4-C1
                               9-C1
 5-C1
                              10-C1
                                          Total
                                                            M9/L|
Surrogate
compound      Concentration (ug/g)
                                          Surrogate
                                          compound

                                          13C-C1«
                                            Concentration
13C-C1.
                              13C-C1
                                                10
Estimated Method LOQ
Highest concentration per resolvable chromatographic peak
                                                             M9/L
Reported by:
              Internal Audit:
                                                      EPA Audit:
        Name
                         Name
                                                                Name
    Signature/Date
                   Signature/Date
                                                           Signature/Date
     Organization
                    Organization
                                                            Organization
                                    43

-------
      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 de-
      tection), 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 quantisation, LOQ).

      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.11  Additional  guidance is also avail-
      able.12  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
                                    44

-------
      •   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


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 7.2 for performance criteria.

            15.3.2   MS performance - See Section 7.3 for performance criteria.

            15.3.3   Qualitative identification - At least 10% of the PCB
                     identifications, as wel.l as any questionable results,
                     should be confirmed by a second mass spectrometrist.
                                    45

-------
      15.3.4   Quantitation - 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 responses of the internal  standards, 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.

      15.4.1   The surrogate recoveries must be within an acceptable
               range, as defined in the QAP, for each sample.  Typi-
               cally, recoveries below 50% are unacceptable, and indi-
               cate 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 concentration of the surrogate compound.   The
               trend of low recoveries 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 irregular, a fractionation cut on a chromatographic
               cleanup 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
               injections, the data must be reviewed.   If the injection
               or GC/MS performance is suspect, the  sample must be re-
               analyzed or other corrective action taken.
                              46

-------
15.4.4   QC for small  sample sets - For small  sample sets (1-10
         samples), the minimum QC requirements can be a heavy
         burden.   Analysts are encouraged to be efficient and
         bunch 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 pro-
         tocol .

         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 addi-
         tion 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 equili-
         bration 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 ad-
         justed so that the spiking level is 1.5 to 4 times the
         measured PCB level in the unspiked samp.le.  The samples
         should be analyzed together and the quantitative results
         calculated.  The recovery of the spiked compounds (calcu-
         lated 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 (ap-
         proximately 10-100 samples) sample sets, the number of
         method blanks, replicates, and standard addition samples
         must comprise at least 10% each.  For example, if 23 sam-
         ples are to be analyzed as a set, 3 blanks, 3 duplicates,
         and 3 standard addition samples would be added in to give
         a total of 32 samples, at a minimum.
                        47

-------
            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.


16.0  Method Performance

      The method performance has not been evaluated.   Limits of quantisation;
      average intralaboratory recoveries, precision,  and accuracy; and inter-
      laboratory recoveries, precision, and accuracy will  be presented when
      available.
17.0  Documentation and Records

      Each laboratory is responsible for maintaining full  records of the analy-
      sis.  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
      specifies (solvent volumes, digestion times,  etc.)  must be stated.

      The remining 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.
                                    48

-------
                                 REFERENCES

1.  USEPA.  1982.  40 CRF 761,  Polychlorinated biphenyls  (PCBs);  manufactur-
    ing, precision, 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 As-
     signment 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.   Longbottom  JE, Lichtenberg, JJ.   1982.   Organochlorine pesticides  and
     PCBs--method 608.   In:  Methods  for  organic  analysis of municipal  and
     industrial  wastewater.  U.S. Environmental  Protection Agency, Report No.
     EPA-600/4-82-057.

 7.   Longbottom  JE, Lichtenberg JJ.   1982.   Base/neutrals, acids,  and
     pesticides—method  625.  In:  Methods for  organic  chemical analysis  of
     municipal  and industiral  wastewater.   U.S.  Environmental  Protection  Agency.
     Report  No.  EPA-600/4-82-057.

 8.   USEPA.   March 1979.   U.S. Environmental  Protection Agency.  Methods  330.4
     (titrimetric, DPD-FAS) and 330.5  (spectrophotometric, DPD) for chlorine,
     total  residual.   Methods  for chemcial analysis  of  water and wastes.
     Cincinnati,  OH:   Environmental  Monitoring  and Support Laboratory.   EPA
     600-4/79-020.

 9.   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.

10.   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.

                                    49

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

12.   USEPA.   1983.   Quality assurance program plan for the Office of Toxic
     Substances.   Office of Pesticides and Toxic Substances, U.S.  Environmental
     Protection Agency,  Washington, D.C.
                                    50

-------
                                  TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
1. REPORT NO.
 EPA-560/5-85-012
                                                           3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
 Analytical Method:  The  Analysis of By-product
 Chlorinated Biphenyls  in Water,  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. 3
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 industrial  wastewater.
The PCBs present may originate either as  synthetic by-products or as contaminants de-
rived  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.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
            b.lDENTIFIERS/OPEN ENDED TERMS  C. COSATI Held/Group
 PCBs
 Polychlorinated biphenyls
 Chlorinated biphenyls
 Analytical methods
 Determination
 By-products
 GC/MS
Water
Wastewater
18. DISTRIBUTION STATEMENT
        Unlimited
                                              19. SECURITY CLASS (This Report)
                                                UNCLASSIFIED
                                       21. NO. OF PAGES
                                            55
                                              20. SECURITY CLASS (This page)
                                                UNCLASSIFIED
                                       22. PRICE
EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE

-------