United States
        Environmental Protection
        Agency
Office of Water
(4303)
EPA-821-R-97-001
March 1997
DRAFT
&EPA   Method1668
        Toxic Polychlorinated Biphenyls by
        Isotope Dilution High Resolution
        Gas Chromatography/High Resolution
        Mass Spectrometry

-------

-------
&EPA  Method1668
       Toxic Polychlorinated Biphenyls by
       Isotope Dilution High Resolution
       Gas Chromatography/High Resolution
       Mass Spectrometry

-------
Method 1668
                                       Acknowledgments

Method 1668 was prepared under the direction of William A. Telliard of the Engineering and Analysis
Division within the EPA Office of Water (OW). This document was prepared under EPA Contract No.
68-C3-0337 by DynCorp Environmental, Energy & National Security Programs with assistance from its
subcontractor Interface, Inc.
                                          Disclaimer

This method has been reviewed by the Engineering and Analysis Division, U.S. Environmental Protection
Agency, and approved for publication. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
   Note:  This method is a draft based on preliminary validation in a single laboratory.  In surveys
   of several laboratories using this method, EPA has found that it is normal for background levels of
   certain congeners that are found in high concentrations  in PCBs to be present in the analytical
   systems in this method.  Therefore, the concentrations of certain congeners in calibration and other
   solutions have been adjusted for these backgrounds.  EPA welcomes constructive suggestions for
   improvement of this method.
                                                                                   Draft, March 1997

-------
                                                                                      Method 1668
                                          Introduction


Method 1668 was developed by the United States Environmental Protection Agency's Office of Science
and Technology for congener-specific determination of the toxic co-planar and mono-ortho-substituted
polychlorinated biphenyls (PCBs) in aqueous, solid, and tissue matrices by isotope dilution, high resolution
capillary column .gas chromatography (HRGQ/high resolution mass  spectrometry (HRMS).


Questions and comments concerning this method or its application should be addressed to:

William A. Telliard,  Director
Analytical Methods Staff (4303)   /           ;
U.S. Environmental Protection Agency               .
401 M Street, SW
Washington, DC 20460
Phone:202-260-7120
Fax: 202-260-7134             '
 Draft, March 1997
                                                                                               in

-------
Method 1668
                                                                                                       Draft, March 1997

-------
                                                                                 Method 1668
                               Method  1668


  Toxic Polychlorinated Biphenyls by Isotope Dilution High Resolution

         Gas  Chromatography/High Resolution Mass  Spectrometry


1.0  Scope and Application

 1.1  This method is for determination of the toxic polychlorinated biphenyls (PCBs) in water, soil,
     sediment, sludge, tissue, and other sample matrices by high resolution gas chromatography/high
     resolution mass spectrometry (HRGC/HRMS). The method is for use in EPA's data gathering and
     monitoring programs associated with the Clean Water Act, the Resource Conservation and
     Recovery Act, the Comprehensive Environmental Response, Compensation and Liability Act, and
     the Safe Drinking Water Act. The method is based on a compilation of methods from the
     technical literature (References 1-3) and on EPA Method 1613.
 1.2  The toxic PCBs listed in Table 1  (Reference 4) and other  specific congeners may be determined
     by this method.
 1.3  The detection limits and quantitation levels in this method are usually dependent on the level of
     interferences rather than instrumental limitations. The minimum levels (MLs) in Table 2 are the
     levels at which the PCBs can be determined with only common laboratory interferences present.
     The Method Detection Limit (MDL) for PCB #126 has been determined as 40 pg/L
     (picograms/Liter; parts-per-quadrillion) in water using this method.
 1.4  The GC/MS portions of this method are for use only by analysts experienced with HRGC/HRMS
     or under the close supervision of such qualified persons. Each laboratory that uses this method
     must demonstrate the ability to generate acceptable results using the procedure in Section 9.2.
 1.5  This method is performance-based. The analyst is permitted to modify the method to overcome
     interferences or lower the cost of measurements, provided that all performance criteria in this
     method are met. The requirements for establishing method equivalency are given in Section
     9.1.2.
 1.6  Any modification of this method, beyond those expressly  permitted, shall be considered a major
     modification subject to application and approval of alternate test procedures under 40 CFR. Parts
     136.4 and 136.5.

 2.0  Summary of Method
     Flow charts that summarize procedures for sample preparation, extraction, and analysis are given
     in Figure 1 for aqueous and solid samples, Figure 2 for multi-phase samples, and Figure 3 for
     tissue samples.
 Draft, March 1997

-------
 Method 1668
 2.1   Extraction.
    2.1.1   Aqueous samples (samples containing less than 1% solids)—Stable isotopically labeled
           analogs of the toxic PCBs are spiked into a 1-L sample, and the sample is vacuum-filtered
           through a glass-fiber filter on top of a solid-phase extraction (SPE) disk. Sample
           components on the filter and disk are eluted with methylene chloride and the eluant is
           concentrated for cleanup.
    2.1.2   Solid, semi-solid, and multi-phase samples (but not tissue)—The labeled compounds are
           spiked into a sample containing 10 g (dry weight) of solids. Samples containing multiple
           phases are pressure filtered and any aqueous liquid is discarded. Coarse solids  are ground or
           homogenized. Any non-aqueous liquid from multi-phase samples is combined with the
           solids and extracted in an SDS extractor. The extract is concentrated for cleanup.
    2.1.3   Fish and other tissue—A 20-g aliquot of sample is homogenized, and a 10-g aliquot is
           spiked with the labeled compounds. The sample is mixed with sodium sulfate,  allowed to
           dry for 12-24 hours, and extracted for 18-24 hours using methylene chloridern-hexane (1:1)
           in a Soxhlet extractor.  The extract is evaporated to dryness, and the lipid content is
           determined.
 2.2   After extraction, samples are cleaned up using back-extraction with sulfuric acid and/or base, and
      gel permeation, silica gel, Florisil and activated carbon chromatography. High-performance liquid
      chromatography (HPLC) can be used for further isolation of specific isomers or congeners. Prior
      to the cleanup procedures cited above, tissue extracts are cleaned up using an anthropogenic
      isolation column.
 2.3   After cleanup, the extract is  concentrated to near dryness. Immediately prior to injection, internal
      standards are added to each  extract, and an aliquot of the extract is injected into the  gas
      chromatograph. The analytes are separated by the GC and detected by a high-resolution
      (£10,000) mass spectrometer. Two exact m/z's are monitored for each analyte.
 2.4   An individual PCB congener is identified by comparing the GC retention time and ion-abundance
      ratio of two exact m/z's  with the corresponding retention time of an authentic standard and the
      theoretical or acquired ion-abundance ratio of the two exact m/z's. Isomer specificity for the toxic
      PCBs is achieved using  GC  columns that resolve these congeners from  the other PCBs.
 2.5   Quantitative analysis is performed using selected ion current profile (SICP) areas in  one of two
      ways:
    2.5.1   For PCBs with  labeled analogs (see Table 1), the GC/MS system is calibrated,  and the
           concentration of each compound is determined using the isotope dilution technique.
    2.5.2   For PCBs without  labeled compounds, the GC/MS system is calibrated, and the
           concentration of each compound is determined using the internal standard technique.
 2.6   The quality of the analysis is assured through reproducible calibration and testing of the
      extraction, cleanup, and  GC/MS systems.

3.0   Definitions
 Definitions are given in the glossary at the end of this method.
 2
                                                                                    Draft, March 1997

-------
                                                                                        Method 1668
4.0   Contamination and Interferences

 4.1   Solvents, reagents, glassware, and other sample processing hardware may yield artifacts and/or
      elevated baselines causing misinterpretation of chromatograms. Specific selection of reagents and
      purification of solvents by distillation in all-glass systems may be required. Where possible,
      reagents are cleaned, by extraction or solvent rinse. The non-coplanar PCB congeners 105, 114,
      118, 123, 156, 157, 167, and  180 have  been shown to be  very difficult to completely eliminate
      from the laboratory at the MDLs in this method, and baking of glassware in a kiln or furnace  at
      450-500 °C may be necessary to remove these and other contaminants.
 4.2   Proper cleaning of glassware  is extremely important because glassware may not only contaminate
      the samples but may also remove the analytes of interest by adsorption on the glass surface.
    4.2.1   Glassware should be rinsed with solvent and washed with a detergent solution as. soon after
           use as is practical. Sonication of glassware containing a  detergent solution for
           approximately 30 seconds may aid in  cleaning.  Glassware with removable parts, particularly
           separatory funnels with  fluoropolymer stopcocks, must be disassembled prior to detergent
           washing.
    4.2.2  After detergent washing, glassware should be rinsed immediately; first with methanol, then
           with hot tap water. The tap water rinse is followed by another methanol rinse, then acetone,
           and then methylene chloride.
    4.2.3  Baking  of glassware in  kiln or other high temperature furnace (450-500 °C) may be
           warranted after particularly dirty samples are encountered.  However, baking should be
           minimized, as repeated  baking of glassware may cause active sites on the glass surface that
           may irreversibly adsorb PCBs.
    4.2.4  Immediately prior to use, the Soxhlet apparatus should be pre-extracted with toluene for
           approximately 3 hours (see Sections 12.3.1-12.3.3).  The solid-phase extraction apparatus
           (Section 6.4.1.5) should be rinsed with methylene chloride/toluene  (80/20 mixture).
 4.3  All materials used in the analysis shall be demonstrated to be free from  interferences by running
      reference matrix method blanks (Section 9.5) initially and with each.sample batch (samples
      started through the extraction process on a  given 12-hour shift to a maximum of 20 samples).
    4.3.1  The reference matrix must simulate as closely as possible the sample matrix under test.
           Ideally, the reference matrix should not contain the  PCBs in detectable amounts, but should
           contain potential interferents in the concentrations expected to be found in the samples to
           be analyzed.
     4.3.2  When a reference matrix that simulates the sample  matrix under test is not available,
           reagent water (Section 7.6.1) can be used to simulate water samples; playground sand
           (Section 7.6.2) or white quartz sand (Section 7.3.2) can  be used to simulate soils; filter
           paper (Section 7.6.3) can be used to simulate papers and similar materials; and corn oil
           (Section 7.6.4) can be used to simulate tissues.
 4.4  Interferences coextracted from samples will vary considerably from source to source, depending
      on the diversity of the site being sampled.  Interfering compounds may be present at
      concentrations several orders of magnitude higher than the PCBs. The most frequently encoun-
      tered interferences are chlorinated dioxins  and dibenzofurans, methoxy biphenyls, hydroxy-
      diphenyl ethers, benzylphenyl ethers, polynuclear aromatics, and pesticides. Because very low
  Draft, March 7997
                                                                                                 3

-------
 Method 1668
      levels of PCBs are measured by this method, the elimination of interferences is essential. The
      cleanup steps given in Section 13 can be used to reduce or eliminate these interferences and
      thereby permit reliable determination of the PCBs at the levels shown in Table 2.
 4.5   Each piece of reusable glassware should be numbered to associate that glassware with the
      processing of a particular sample. This will assist the laboratory in tracking possible sources of
      contamination for individual samples, identifying glassware associated with highly contaminated
      samples that may require extra cleaning, and determining when glassware should be discarded.
 4.6   Cleanup of tissue—The natural  lipid content of tissue can interfere in the analysis of tissue
      samples for the PCBs. The lipid contents of different species and portions of tissue can vary
      widely. Lipids are soluble to varying degrees in various organic solvents and may be present in
      sufficient quantity to overwhelm the column chromatographic cleanup procedures used for
      cleanup of sample extracts. Lipids must be removed by the lipid removal procedures in Section
      13.6, followed by Florisil (Section 13.7), and carbon (Section 13.4) as minimum additional
      cleanup steps.

5.0   Safety

 5.1   The toxicity or carcinogenicity of each chemical used in this method has not been precisely
      determined; however, each compound should be treated as a potential health hazard. Exposure to
      these compounds should be reduced to the lowest possible level.
    5.1.1   The PCBs have been tentatively classified as known or suspected human  or mammalian
           carcinogens. On the basis  of the available lexicological and physical properties of the
           PCBs, pure standards should be handled only by highly trained personnel thoroughly
           familiar with handling and cautionary procedures and the associated risks.
    5.1.2   It is recommended that the laboratory purchase dilute standard solutions of the analytes in
           this method. However,  if primary solutions are prepared, they shall be prepared in a hood,
           and a NIOSH/MESA approved toxic gas respirator shall be worn when high concentrations
           are handled.
 5.2   The laboratory is responsible for maintaining a current awareness file of OSHA regulations
      regarding the safe handling of the chemicals specified in this method. A reference file of material
      safety data sheets (MSDSs) should also be made available to all personnel involved in these
      analyses. It is also suggested that the laboratory perform personal hygiene monitoring of each
      analyst who uses this method and that the results of this monitoring be made available to the
      analyst. Additional information  on laboratory  safety can be found in References 5-8. The
      references and bibliography at the end of Reference 8 are particularly comprehensive in dealing
      with the general subject of laboratory safety.                              '
 5.3   The pure PCBs and samples  suspected to contain these compounds are handled using essentially
      the same techniques employed in handling radioactive or infectious materials. Well-ventilated,
      controlled access laboratories are required. Assistance in evaluating the health hazards of
      particular laboratory conditions  may be obtained from certain consulting laboratories and from
      State Departments of Health  or  Labor, many of which have an industrial health service. Each
      laboratory must develop a strict safety program for handling these compounds.  The practices in
      Reference 11  for handling chlorinated dibenzo-/?-dioxins and dibenzofurans (CDDs/CDFs) are
      also recommended for handling the toxic PCBs.
                                                                                     Draft, March 1997

-------
                                                                                       Method 1868
   5.3.1   Facility—-When finely divided samples (dusts, soils, dry chemicals) are handled, all
          operations (including removal of samples from sample containers, weighing, transferring,
          and mixing) should be performed in a glove box demonstrated to be leak tight or in a fume
          hood demonstrated to have adequate air flow. Gross losses to the laboratory ventilation
          system must not be allowed. Handling of the dilute solutions normally used in analytical
          and animal work presents no inhalation hazards except in the case of an accident.
   5.3.2   Protective equipment—Disposable plastic gloves, apron or lab coat, safety glasses or mask,
          and a glove box or fume hood adequate for radioactive work should be used. During
          analytical operations that may give rise to aerosols or dusts, personnel should wear
          respirators equipped with activated carbon filters. Eye protection equipment (preferably full
        1  face shields) must be worn while working with exposed samples or pure analytical
          standards. Latex gloves are commonly used to reduce exposure of the hands. When
          handling samples suspected or known to contain high concentrations of the PCBs, an
          additional set of gloves can also be worn beneath the latex gloves.
   5.3.3   Training—Workers must be trained in the proper method of removing contaminated gloves
          and clothing without contacting the exterior surfaces.
   5.3.4   Personal hygiene—Hands and forearms should be washed thoroughly after each
          manipulation and before breaks (coffee, lunch, and shift).
   5.3.5   Confinement—Isolated work areas posted with signs, segregated glassware and tools, and
          plastic absorbent paper on bench tops will aid in confining contamination.
   5.3.6   Effluent vapors—The effluents of sample splitters from the gas chromatograph (GC) and
          from roughing pumps on the mass spectrometer (MS) should pass through either a column
          of activated charcoal or be bubbled through a trap containing oil or high-boiling alcohols to
          condense PCB vapors.
   5.3.7   Waste Handling—Good technique  includes minimizing contaminated waste. Plastic bag
          liners should be used in waste cans. Janitors and other personnel must be trained in the safe
          handling of waste.
   5.3.8   Decontamination.                                                       .
       5.3.8.1   Decontamination of personnel—Use any mild soap with plenty  of scrubbing action.
       5.3.8.2  Glassware, tools, and surfaces—Chlorothene NU Solvent is a less toxic  solvent that  .
               should be effective in removing PCBs. Satisfactory cleaning may be accomplished by
               rinsing with Chlorothene, then washing with any detergent and  water.  If glassware is
               first rinsed with solvent, then the dish water may be disposed of in the sewer. Given
               the cost of disposal,  it is prudent to minimize solvent wastes.    ,
   5.3.9   Laundry—Clothing known to be contaminated should be collected in plastic bags. Persons
          who convey the bags and  launder the clothing should be advised of the hazard and trained
          in proper handling. The clothing may be  put into a .washer without contact if the launderer
          knows of the potential problem. The washer should be run through a cycle before being
          used again for other clothing.
  5.3.10   Wipe tests—A useful method of determining cleanliness of work  surfaces and tools is to
          wipe the surface with a piece of filter paper. Extraction and analysis  by GC with  an
          electron capture detector (ECD) can achieve a limit of detection of 0.1 g per wipe; analysis
Draft, March 1997

-------
 Method 1668
           using this method can achieve an even lower detection limit. Less than 0.1 ug per wipe
           indicates acceptable cleanliness; anything higher warrants further cleaning. More than 10 ug
           on a wipe constitutes an acute hazard and requires prompt cleaning before further use of the
           equipment or work space, and indicates that unacceptable work practices have been
           employed.

6.0   Apparatus, Equipment and Supplies
   Note:  Brand names, suppliers, and part numbers are for illustration purposes only and no
   endorsement is implied. Equivalent performance may be achieved using apparatus and materials
   other than those specified here. Meeting the performance requirements of this method is the
   responsibility of the laboratory.         	

 6.1   Sampling equipment for discrete or composite sampling.
    6.1.1   Sample bottles and caps.
        6.1.1.1   Liquid samples (waters, sludges and similar materials containing 5% solids or less)—
                Sample bottle, amber glass, 1.1-L minimum, with screw cap.
        6.1.1.2   Solid samples (soils, sediments, sludges, paper pulps, filter cake, compost, and similar
                materials that contain more than 5% solids)—Sample bottle, wide mouth, amber glass,
                500-mL minimum.
        6.1.1.3   If amber bottles are not available, samples shall be protected from light.
        6.1.1.4   Bottle caps—Threaded to fit sample bottles. Caps shall be lined with fluoropolymer.
        6.1.1.5   Cleaning.
           6.1.1.5.1   Bottles are detergent water washed, then solvent rinsed before use.
           6.1.1.5.2  Liners .are detergent water washed and rinsed with reagent water (Section 7.6.1).
    6.1.2  Compositing equipment—Automatic or manual compositing system incorporating glass
           containers cleaned per bottle cleaning procedure above. Only glass  or fluoropolymer tubing
           shall be used. If the sampler uses a peristaltic pump, a minimum length of compressible
           silicone rubber tubing may be used in the pump only. Before use, the tubing shall be
           thoroughly rinsed with methanol, followed by repeated rinsing with reagent water to
           minimize sample contamination.' An integrating flow meter is used  to collect proportional
           composite samples.
 6.2   Equipment for glassware cleaning—Laboratory sink with overhead  fume hood.
 6.3   Equipment for sample preparation.
           Laboratory fume hood of sufficient size to contain the sample preparation equipment listed
           below.
6.3.1

6.3.2
6.3.3

6.3.4
           Glove box (optional).
           Tissue homogenizer—-VirTis Model 45 Macro homogenizer (American Scientific Products
           H-3515, or equivalent) with stainless steel Macro-shaft and Turbo-shear blade.
           Meat grinder—Hobart, or equivalent, with 3- to 5-mm holes in inner plate.
                                                                                    Draft, March 1997

-------
                                                                                        Method 1668
   6.3.5   Equipment for determining percent moisture.
       6.3.5.1   Oven—Capable of maintaining a temperature of 110 ±5°C.
       6.3.5.2   Desiccator.
   6.3.6   Balances.
       6.3.6.1   Analytical—Capable of weighing 0.1 mg.
       6.3.6.2   Top loading—Capable of weighing 10 mg.
6.4  Extraction apparatus.
   6.4.1   Water samples.
       6.4.1.1   pH meter, with combination glass electrode.  '  -,
       6.4.1.2  pH paper, wide range (Hydrion Papers, or equivalent).
       6.4.1.3  Graduated cylinder, 1-L capacity.               ,
       6.4.1.4  Liquid/liquid extraction—Separatory funnels, 250-, 500-, and 2000-mL, with
               fluoropolymer stopcocks.
       6.4.1.5  Solid-phase extraction.
           6.4.1.5.1   1-L filtration apparatus, including glass funnel, frit support, clamp, adapter,
                     stopper, filtration flask, and vacuum tubing (Figure 4). For wastewater samples,
                     the apparatus should accept 90 or 144 mm disks. For drinking water or other
                     samples containing low solids, smaller disks may be used.
           6.4.1.5.2   Vacuum source capable of maintaining  25 in.  Hg, equipped with shutoff valve
                     and vacuum gauge.
           6.4.1.5.3   Glass-fiber filter—Whatman GMF 150  (or equivalent),  1 micron pore size, to fit
                     filtration apparatus in Section 6.4.1.5.1.
           6.4.1.5.4   Solid-phase extraction disk containing octadecyl (Clg) bonded silica uniformly
                     enmeshed in an inert matrix—Fisher Scientific 14-378F (or equivalent), to fit
                     filtration apparatus in Section  6.4.1.5.1.
    6.4.2   Soxhlet/Dean-Stark (SDS) extractor (Figure 5 and  Reference 12) for filters and solid/sludge
           samples.
        6.4.2.1   Soxhlet—50-mm ID, 200-mL capacity with 500-mL flask (Cal-Glass LG-6900, or
                equivalent, except substitute 500-mL round-bottom flask for  300-mL flat-bottom
                flask).
        6.4.2.2   Thimble-^3 x 123 to fit Soxhlet (Cal-Glass LG-6901-122, or equivalent).
        6.4.2.3   Moisture trap—Dean Stark or Barret with fluoropolymer stopcock, to fit Soxhlet.
        6.4.2.4   Heating mantle—Hemispherical, to fit 500-mL round-bottom flask (Cal-Glass LG-
                8801-112, or equivalent).
        6.4.2.5   Variable transformer—Powerstat (or equivalent), 110-volt, 10-amp.
    6.4.3   Beakers—400-to 500-mL.
    6.4.4   Spatulas—Stainless steel.
 6.5   Filtration apparatus.                                           .

 Draft, March 1997    :           •       '    :   :   ''      ~~~~'•    '"~~     ~      *

-------
Method 1668
    6.5.1   Pyrex glass wool—Solvent-extracted by SDS for 3 hours minimum.
    6.5.2   Glass funnel—125- to 250-mL.
    6.5.3   Glass-fiber filter paper—Whatman GF/D (or equivalent), to fit glass funnel in Section 6.5.2.
    6.5.4   Drying column—15- to 20-mm ID Pyrex chromatographic column equipped with coarse-
          glass frit or glass-wool plug.
    6.5.5   Buchner funnel—15-cm.
    6.5.6   Glass-fiber filter paper for Buchner funnel above.
    6.5.7   Filtration flasks—1.5- to 2.0-L, with side arm.
    6.5.8   Pressure filtration apparatus—Millipore YT30 142 HW, or equivalent.
6.6  Centrifuge apparatus.
    6.6.1   Centrifuge—Capable of rotating 500-mL centrifuge  bottles or 15-mL centrifuge tubes at
          5,000 rpm minimum.
    6.6.2   Centrifuge bottles—500-mL, with screw-caps, to fit centrifuge.
    6.6.3   Centrifuge tubes—12- to 15-mL, with screw-caps, to fit centrifuge.
6.7  Cleanup apparatus.
    6.7.1   Automated gel permeation chromatograph (Analytical Biochemical Labs, Inc, Columbia,
          MO, Model GPC Autoprep 1002, or equivalent).
       6.7.1.1   Column—600-700 mm long x 25 mm ID,  packed with 70 g of SX-3 Bio-beads (Bio-
               Rad Laboratories, Richmond, CA, or equivalent).
       6.7.1.2   Syringe—10-mL, with Luer fitting.
       6.7.1.3   Syringe filter holder—stainless steel, and glass- fiber or fluoropolymer filters (Gelman
               4310, or equivalent).
       6.7.1.4   UV detectors—254-nm, preparative or semi-preparative flow cell (Isco, Inc., Type 6;
               Schmadzu, 5-mm path length; Beckman-Altex 152W, 8-uL micro-prep flow cell, 2-
               mm path; Pharmacia UV-1, 3-mm flow cell; LDC Milton-Roy UV-3, monitor #1203;
               or equivalent).
    6.7.2   Reverse-phase high-performance liquid chromatograph.
       6.7.2.1   Column oven and detector—Perkin-Elmer Model LC-65T (or equivalent) operated at
               0.02 AUFS at 235 nm.
       6.7.2.2   Injector—Rheodyne 7120 (or equivalent) with 50-uL sample loop.
       6.7.2.3   Column—Two 6.2 mm x 250 mm Zqrbax-ODS columns in series (DuPont
               Instruments Division, Wilmington, DE, or equivalent), operated at 30°C and 2.0
               mL/min with  gradient from TBD percent methanol:acetonitrile to 100 percent
               acetonitrile in TBD minutes.
       6.7.2.4   Pump—Altex 11OA (or equivalent).
    6.7.3   Pipets.
       6.7.3.1   Disposable, Pasteur,  150-mm long x 5-mm ID (Fisher Scientific 13-678-6A, or
               equivalent).
                                                                                   Draft, March 1997

-------
                                                                                       Method 1668
       6.7.3.2   Disposable, serological, 50-mL (8-to 10-mm ID).
   6.7.4   Glass chromatographic columns.
       6.7.4.1   150-mm long x 8-mm ED, (Kontes K-420155, or equivalent) with coarse-glass frit or
               glass-wool plug and 250-mL reservoir.
       6.7.4.2   200-mm long x 15-mm ID, with coarse-glass frit or glass-wool plug and 250-mL
               reservoir.
       6.7.4.3   300-mm long x 22-mm ID, with coarse-glass frit, 300-mL reservoir, and glass or,
               fluoropolymer stopcock.
   6.7.5   Stirring apparatus for batch silica cleanup of tissue extracts.
       6.7.5.1   Mechanical stirrer—Corning Model 320, or equivalent.
       6.7.5.2   Bottle—500- to 600-mL wide-mouth clear glass.
   6.7.6   Oven—For baking and storage of adsorbents, capable of maintaining a constant temperature
          (±5°C) in the range of 105-250°C.   ,
6.8   Concentration apparatus.
   6.8.1   Rotary evaporator—Buchi/Brinkman-American Scientific No. E5045-10 or equivalent,
          equipped with a variable temperature water bath.                            ,
       6.8.1.1   Vacuum source for rotary evaporator equipped with shutoff valve at the evaporator
               and vacuum gauge.
       6.8.1.2   A recirculating water pump and chiller are recommended, as use of tap water for cool-
               ing the evaporator wastes large volumes of water and can lead to inconsistent
               performance as water temperatures  and pressures vary.
       6.8.1.3   Round-bottom flask—100-mL and 500-mL or larger, with ground-glass fitting
               compatible with the rotary evaporator.
   6.8.2   Kuderna-Danish (K-D) concentrator.
       6.8.2.1   Concentrator tube—10-mL, graduated (Kontes K-570050-1025, or equivalent) with
               calibration verified. Ground-glass stopper (size  19/22 joint) is  used to prevent
               evaporation of extracts.
       6.8.2.2  Evaporation flask—500-mL (Kontes K-570001-0500,  or equivalent),  attached to
               concentrator tube with springs (Kontes K-662750-0012 or equivalent).
       6.8.2.3  Snyder column—Three-ball macro  (Kontes K-503000-0232, or equivalent).
       6.8.2.4  Boiling chips.
           6.8.2.4.1   Glass or silicon carbide—Approximately 10/40 mesh, extracted with methylene
                     chloride and baked at 450°C for 1 hour minimum.
           6.8.2.4.2   Fluoropolymer (optional)—Extracted with methylene chloride.
       6.8.2.5  Water bath—Heated, with concentric ring cover, capable of maintaining a temperature
                within ±2°C, installed in a fume hood.
    6.8.3  Nitrogen blowdown apparatus—Equipped with water bath controlled in the range of 30 -
          60°C (N-Evap, Organomation Associates, Inc., South Berlin,  MA,, or equivalent), installed
          in a fume  hood.       '
Draft, March 1997

-------
 Method 1668
     6.8.4  Sample vials.
        6.8.4.1   Amber glass, 2- to 5-mL with fluoropolymer-lined screw-cap.
        6.8.4.2  Glass, 0.3-mL, conical, with fluoropolymer-lined screw or crimp cap.
 6.9  Gas chromatograph—Shall have splitless or on-column injection port for capillary column,
      temperature program with isothermal hold, -and shall meet all of the performance  specifications in
      Section 10.
     6.9.1.  GC columns—The pair of GC columns listed below are capable of resolving all 209 PCB
           congeners. Other GC columns may be used so long as PCBs 126 and 169 are each resolved
           from their respective most closely eluted leading and trailing congeners. The valley height
           between PCB  126 or 169  and its respective most closely eluted leading and trailing
           congeners must be less than 10 percent of the height of the shorter of the pair.
     6.9.2  Column #1—30±5-m long x 0.25±0.02-mm ID; 0.25-um film SPB-Octyl (Supelco 2-4218,
           or equivalent).
     6.9.3  Column #2—30±5-m long x 0.25±0.02-mm ID; 0.25-um film DB-1 (J&W, or equivalent).
6,10  Mass spectrometer—28- to 40-eV electron impact ionization, shall be capable of repetitively
      selectively monitoring 12 exact m/z's minimum at high resolution (>10,000) during a period less
      than 1.5 seconds, and shall meet all of the performance specifications in Section 10.
6.11  GC/MS interface—The mass spectrometer (MS) shall be interfaced to the GC such that the end
      of the capillary column terminates within 1 cm of the ion source but  does not intercept the
      electron or ion beams.
6.12  Data system—Capable of collecting, recording,  and storing MS data.

7.0  Reagents  and Standards

 7.1  pH adjustment and back-extraction.
     7.1.1   Potassium hydroxide—Dissolve 20 g reagent grade KOH in 100 mL reagent water.
     7.1.2  Sulfuric acid—Reagent grade (specific gravity 1.84).
     7.1.3  Hydrochloric acid—Reagent grade, 6N.
     7.1.4  Sodium chloride—Reagent grade, prepare  at 5% (w/v) solution  in reagent water.
 7.2  Solution drying and evaporation.
     7.2.1   Solution drying—Sodium  sulfate, reagent grade, granular, anhydrous (Baker 3375, or
           equivalent), rinsed with methylene chloride (20 mL/g), baked at 400°C for 1 hour
           minimum, cooled in a desiccator, and stored in a pre-cleaned glass bottle with screw-cap
           that prevents moisture from entering. If, after heating, the sodium sulfate develops a
           noticeable grayish cast (due to the presence of carbon in the crystal matrix), that batch of
           reagent is not  suitable for use and should be discarded. Extraction with methylene chloride
           (as opposed to simple rinsing) and baking at a lower temperature may produce  sodium
           sulfate that is  suitable for  use.
     7.2.2  Tissue drying—Sodium sulfate, reagent grade, powdered, treated and  stored as above.
     7.2.3  Prepurified nitrogen.
 10
Draft, March 1997

-------
                                                                                        Method 1668
7.3   Extraction.                            •
    7.3.1   Solvents— Acetone, toluene, n-hexane, methanol, methylene chloride, and nonane; distilled
          in glass, pesticide quality, lot-certified to be free of interferences.
    7.3.2   White quartz sand, 60/70 mesh— For Soxhlet/Dean-Stark extraction (Aldrich Chemical/Cat.
          No. 27-437-9, or equivalent). Bake at 450°C for 4 hours minimum.
7.4   GPC calibration solution — Prepare a solution containing 300 mg/mL com oil, TBD mg/mL PCB
     209, 1.4 mg/mL pentachlorophenol, 0.1  mg/mL perylene, and 0.5 mg/mL sulfur. [To be modified
     if necessary.]                                                                 ..
7.5   Adsorbents for sample cleanup.                                                      .
    7.5.1   Silica gel.
       7.5.1.1   Activated silica gel— 100-200 mesh, Supelco 1-3651  (or equivalent), rinsed with
                methylene chloride, baked at 180°C for a minimum of 1 hour, cooled in a desiccator,
                and stored in a precleaned glass bottle with screw-cap that prevents  moisture from
                entering.
       7.5.1.2   Acid silica gel (30% w/w) — Thoroughly mix 44.0 g of concentrated sulfuric acid with
                100.0 g of activated silica gel in a clean container. Break up aggregates  with a stirring
                rod until a uniform mixture is  obtained. Store in a screw-capped bottle with
                fluoropolymer-lined cap.
       7.5.1.3   Basic silica gel — Thoroughly mix 30 g of IN sodium hydroxide with 100 g of
                activated silica gel in a clean container. Break up aggregates with a stirring rod until a
                uniform mixture is obtained. Store in a screw-capped bottle with fluoropolymer-lined
                cap.
       7.5.1.4   Potassium silicate.
           7.5.1.4.1   Dissolve 56 g of high purity potassium hydroxide (Aldrich, or equivalent) in 300
                     mL of methanol in a 750- to 1000-mL flat-bottom flask.
           7.5.1.4.2  Add 100 g of activated silica gel (Section 7.5.1.1) and a stirring bar, and stir on
                     a hot plate at 60-70°C  for 1-2 hours.
           7.5.1.4.3  Decant the liquid and rinse the potassium silicate twice with 100-mL portions of
                     methanol, followed by a single rinse with 100 mL of methylene chloride.
           7.5.1.4.4  Spread the potassium silicate on solvent-rinsed aluminum foil  and dry for 2-4
                     hours  in a hood.
           7.5.1.4.5  Activate overnight at 200-25Q°C.
    7.5.2   Carbon.
        7.5.2.1   Carbopak C— -(Supelco 1-0258, or equivalent).                 •
        7.5.2.2   Celite 545— (Supelco 2-0199, or equivalent).                  ,
        7.5.2.3   Thoroughly mix 18.0 g Carbopak C and 18.0 g Celite 545 to produce a 50% w/w
                mixture. Activate the mixture  at 130°C for a minimum of 6 hours. Store in a
                desiccator.
    7.5.3   Anthropogenic isolation column— Pack the column in  Section 6.7.4.3 from bottom to top
           with the following:
 Draft, March 1997

-------
Method 1668
       7.5.3.1   2 g activated silica gel (Section 7.5.1.1).
       7.5.3.2   2 g potassium silicate (Section 7.5.1.4).
       7.5.3.3   2 g granular anhydrous sodium sulfate (Section 7.2.1).
       7.5.3.4   10 g acid silica gel (Section 7.5.1.2).
       7.5.3.5   2 g granular anhydrous sodium sulfate.
    7.5.4  Florisil column.                                       •
       7.5.4.1   Florisil—PR grade, 60-100 mesh (U.S. Silica Corp, Berkeley Springs, WV, or
                equivalent). Fill a clean 1- to 2-L bottle 1/2 to 2/3 full with Florisil and place in an
                oven at 130-150 °C for a minimum of three days.
       7.5.4.2   Immediately prior to use, dry pack a 300-mm x 22-mm ID glass column (Section
                6.7.4.3) bottom to top with 0.5-1.0 cm of anhydrous sodium sulfate (Section 7.2.1),
                10-10.5 cm of warm to hot activated Florisil (Section 7.5.4.1), and 1-2 cm of warm to
                hot anhydrous sodium sulfate. Allow the column to cool and wet immediately with
                100 mL of n-hexane to prevent water from entering.
       7.5.4.3   Using the procedure  in Section  13.7, establish the elution pattern for each carton of
                Florisil received.
7.6  Reference matrices—Matrices in which the PCBs and interfering compounds are not detected by
     this method.
    7.6.1
    7.6.2
          Reagent water—Bottled water purchased locally or prepared by passage through activated
          carbon.
          High-solids reference matrix—Playground sand or similar material. Prepared by extraction
          with methylene chloride and/or baking at 450°C for a minimum of 4 hours.
    7.6.3  Paper reference matrix—Glass-fiber filter, Gelman type A, or equivalent. Cut paper to
          simulate the surface area of the paper sample being  tested.
    7.6.4  Tissue reference matrix—Com or other vegetable oil. May be prepared by extraction with
          methylene chloride.
    7.6.5  Other matrices—This method may be verified on  any reference matrix by performing the
          tests given in Section 9.2. Ideally, the matrix should be free of the PCBs, but in no case
          shall the background level of the PCBs in the reference matrix exceed three times the
          minimum levels in Table 2. If low background levels of the PCBs are present in the
          reference matrix, the spike  level of the analytes used in Section 9.2 should be increased to
          provide a spike-to-background ratio in the range of  1:1 to 5:1 (Reference 11).
7.7  Standard solutions—Purchased as solutions or mixtures with certification to their purity,
     concentration, and authenticity, or prepared from materials of known purity and composition. If
     the chemical purity is 98% or greater, the weight may be used without  correction to compute the
     concentration of the standard. When not being  used, standards  are stored in the dark at room
     temperature in screw-capped vials with fluoropolymer-lined caps. A mark is placed on the vial at
     the level of the solution so that solvent loss by evaporation can be detected. If solvent loss has
     occurred, the solution should be replaced.
12
                                                                                     Draft, March 1997

-------
                                                                                       Method'1668
 7.8   Stock solutions.
    7.8.1   Preparation—Prepare in nonane per the steps below or purchase as dilute solutions
           (Cambridge Isotope Laboratories (CIL), Wobum, MA, or equivalent). Observe the safety
           precautions in Section 5 and the recommendation in Section 5.1.2.
    7.8.2   Dissolve an appropriate amount of assayed reference material in solvent. For example,
           weigh 1 to 2 mg of PCB 126 to three significant figures in a  10-mL ground-glass-stoppered
           volumetric flask and fill to the mark with nonane. After the PCB is completely dissolved,
           transfer the solution to a clean 15-mL vial with fluoropolymer-lined cap.
    7.8.3   Stock standard solutions should be checked for signs of degradation prior to the preparation
           of calibration or performance test standards.  Reference standards that can be used to
           determine the accuracy of calibration standards are available from several vendors.
 7.9   PAR stock solution.
    7.9.1   All PCBs—Busing the solutions in Section 7.8, prepare the PAR stock solution to contain
           the PCBs of interest at the concentrations shown in Table 3. When diluted,  the solution will
           become the PAR (Section 7.14).                                     ,
    7.9.2   If the toxic, non-ortho, co-planar PCBs (PCBs 77,  126, and 169) only are to be determined,
           prepare the PAR stock solution to contain these compounds only.
7.10   Labeled-compound spiking solution.
   7.10.1   All toxic PCBs—From stock solutions, or from purchased mixtures, prepare this solution to
           contain the labeled compounds in nonane at  the concentrations shown in Table 3. This
           solution is diluted with acetone prior to use (Section 7.10.3).
   7.10.2   If PCBs 77, 126, and  169 only are to be determined, prepare  the labeled-compound solution
           to contain these compounds only. This solution is diluted with acetone prior to use (Section
           7.10.3).
   7.10.3   Dilute  a sufficient volume of the labeled compound solution (Section 7.10.1 or 7.10.2) by a
           factor of 500 with acetone to prepare a diluted  spiking solution. Each sample  requires 1.0
           mL of the diluted solution, but no more solution should be prepared than can  be used in
           one day.
7.11   Cleanup standard—Prepare PCBs 81 and 111 in nonane at the concentration shown in Table  3.
      The cleanup standard is added to all extracts prior to cleanup to measure the efficiency of the
      cleanup process.
7.12   Internal standard(s).                                                      .
   7.12.1   All toxic PCBs—Prepare the internal standard solution to contain labeled PCBs 52, 101,
           138, and 178 in nonane at the concentration shown in Table  3.
   7.12.2   If PCBs 77, 126, and 169 only  are to be determined, the internal standard solution may be
           prepared to contain PCBs 52, 101, and 138 only.             ,
7.13   Calibration standards (CS1  through CS5)---Combine the solutions in Sections  7.9-7.12 to produce
      the five calibration solutions shown in Table 4 in  nonane. These solutions permit the relative
      response (labeled to native) and response factor to be measured as  a function of concentration.
      The CS3 standard is used for calibration verification (VER). If the  PCBs 77, 126, and 169 only
      are to be determined, combine the solutions appropriate  to these compounds.
 Draft, March 1997
                                                                                               13

-------
 Method 1668
7.14  Precision and recovery (PAR) standard—Used for determination of initial (Section 9.2) and
      ongoing (Section 15.5) precision and recovery (See Table 3). Dilute 200 uL of the PAR stock
      solution (Section 7.9.1  or 7.9.2) to 10 mL with acetone for each sample matrix for each sample
      batch. One mL of each are required for the blank and OPR with each matrix in each batch.
7.15  GC retention time window defining solution and isomer specificity test standard—Used to define
      the beginning and ending retention times for the PCB congeners and to demonstrate isomer
      specificity of the GC columns employed for determination of PCB 126. The standard must
      contain the compounds listed in Table 5 (CIL   , or equivalent),  at a minimum. It is not
      necessary to monitor all of the window-defining compounds if PCBs 77, 126, and  169 only are
      to be determined. In this case, a congener-specificity test standard containing the most closely
      eluted isomers listed in Table 5
      (CIL   , or equivalent) may be used.
7.16  QC Check Sample—A QC Check Sample should be obtained from a source independent of the
      calibration standards. Ideally, this check sample would be a certified standard reference material
      (SRM) containing the PCBs  in known concentrations in a sample  matrix similar to the matrix
      being analyzed. The National Institute of Standards and Technology (NIST) in Gaithersburg,
      Maryland has SRMs for several individual PCB congeners, and as Aroclors in transformer and
      motor oil, in combination with pesticides in cod liver oil, and in combination with 2,3,7,8-TCDD
      in human serum.
7.17  Stability of solutions—Standard solutions used for quantitative purposes (Sections 7.9 through
      7.15) should be analyzed periodically, and should be assayed against reference standards before
      further use.

8.0  Sample Collection, Preservation, Storage, and Holding Times.

 8.1  Collect  samples in amber glass containers following conventional  sampling practices (Reference
      12).
 8.2  Aqueous samples.
     8.2.1   Samples that flow freely are collected as grab samples or in  refrigerated bottles using
           automatic sampling equipment.
     8.2.2  If residual chlorine is present, add 80 mg sodium thiosulfate  per liter of water. EPA
           Methods  330.4 and 330.5 may be used to measure residual chlorine (Reference 13).
     8.2.3  Adjust sample pH 2-3 with sulfuric acid.
     8.2.4  Maintain aqueous samples in the dark at 0-4°C from the time of collection until receipt at
           the laboratory. Store in the dark at 0-4°C.
 8.3  Solid samples.
     8.3.4  Solid samples  are collected as grab samples using wide-mouth jars.
     8.3.4  Maintain solid, semi-solid, oily, and mixed-phase samples in the dark at <4°C from the
           time of collection until  receipt at the laboratory. Store solid,  semi-solid, oily, and mixed-
           phase samples in the dark at <-10°C.
 8.4  Fish  and tissue samples.
 14
                                                                                   Draft, March 1997

-------
                                                                                        Method 766S
    8.4.1  Fish may be cleaned, filleted, or processed in other ways in the field, such that the
          laboratory may expect to receive whole fish, fish fillets, or other tissues for analysis.
    8.4.2  Fish collected in the field should be wrapped in aluminum foil and must be maintained at a
         . temperature less than 4°C from the time of collection until receipt at the laboratory.
    8.4.3  Samples must be frozen upon receipt at the laboratory and maintained in the  dark at <-10°C
          until prepared. Maintain unused sample in the dark at <-10°C.
8.5  Holding times.
    8.5.1  There are no demonstrated maximum holding times associated with the PCBs in aqueous,
          solid,  semi-solid, tissues, or other sample matrices. If stored in the dark at 0-4°C and
          preserved as given above (if required), aqueous samples may be stored for up to one year.
          Similarly, if stored in the dark at <-10°C, solid, semi-solid, multi-phase, and  tissue samples
          may be stored for up to one  year.
    8.5.2  Store  sample extracts in the dark at <-10°C until analyzed. If stored in the dark at <-10°C,
          sample extracts may  be stored for up to one year.

9.0  Quality Assurance/Quality Control

9.1  Each laboratory that uses this method is required to  operate a formal quality assurance program
     (Reference  14). The minimum requirements of this program consist of an initial demonstration of
     laboratory capability, analysis of samples spiked with labeled compounds to evaluate and
     document data quality, and analysis of standards and blanks as tests of continued performance.
     Laboratory  performance is compared to established performance criteria to determine if the
     results of analyses meet the performance characteristics of the method.
     If the  method  is to be applied to a sample matrix  other than water (e,g., soils, filter cake,
     compost, tissue) the most appropriate alternate matrix (Sections 7.6.2-7.6.5 and 7.16) is sub-
     stituted for the reagent water matrix (Section 7.6.1)  in all performance tests.
    9.1.1   The analyst shall make an initial demonstration of the ability to generate acceptable
           accuracy and precision with this method. This  ability is established as described in Section
        '   9-2. '    '       •••.'..'...••-..'"      -
    9.1.2   In recognition of advances that, are  occurring in analytical technology and to allow the
           analyst to overcome  sample matrix interferences, the analyst is permitted certain options to
           improve separations  or lower the costs of measurements. These options include alternate
           extraction, concentration, cleanup procedures,  and changes in columns and detectors.
           Alternate determinative techniques, such as the substitution of spectroscopic  or immuno-
           assay techniques, and changes that degrade method performance, are not allowed. If an
           analytical technique  other than the techniques  specified in this method is used, that
           technique must have a specificity equal to or better than the specificity of the techniques in
           this method for the analytes of interest.  .
        9.1.2.1   Each time a modification is made to this method, the analyst is required to repeat the
                 procedure in Section 9,2. If the detection limit of the method will be affected by the
                 change, the laboratory is required to demonstrate that-'the.MDL (40 CFR Part  136,
                 Appendix B) is lower  than one-third the  regulatory compliance level or one-third the
 Draft, March 1997
                                                                                                15

-------
Method 1668
               ML in this method, whichever is higher. If calibration will be affected by the change,
               the analyst must recalibrate the instrument per Section  10.
       9.1.2.2  The laboratory is required to maintain records of modifications  made to this method.
               These records include the following at a minimum:
           9.1.2.2.1   The names, titles, addresses, and telephone numbers of the analyst(s) who
                     performed the analyses and modification, and of the quality control officer who
                     witnessed and will verify the analyses and modifications.
           9.1.2.2.2   A listing of pollutant(s) measured, by name and CAS Registry number.
           9.1.2.2.3   A narrative stating reason(s) for the modifications.
           9.1.2.2.4   Results from  all quality control (QC)  tests comparing the modified method to
                     this method.  These results are to include the following:
                     a)        Calibration (Section 10.5-10.7).
                     b)        Calibration verification (Section 15.3).
                     c)        Initial precision and recovery (Section 9.2).
                     d)        Labeled compound recovery (Section 9.3).
                     e)        Analysis of blanks (Section 9.5).
                     f)        Accuracy assessment (Section 9.4).
           9.1.2.2.5   Data that will allow an independent reviewer to validate each determination by
                     tracing the instrument output (peak height,  area, or other signal) to the final
                     result. These  data are to include the following:
                     a)        Sample numbers and other identifiers.
                     b)        Extraction dates.
                     c)        Analysis dates and times.
                     d)        Analysis sequence/run chronology.
                     e)        Sample weight or volume (Section  11).
                     f)        Extract volume prior to each cleanup step (Section 13).
                     g)        Extract volume after each cleanup step (Section 13).
                     h)        Final extract volume prior to injection (Section 14).
                     i)        Injection volume (Section 14.3).
                     j)        Dilution  data, differentiating between dilution of a sample or extract
                              (Section  17.5).
                     k)        Instrument and operating conditions.
                     1)        Column  (dimensions, liquid phase,  solid support, film thickness, etc).
                     m)       Operating conditions (temperatures, temperature program, flow rates).
                     n)        Detector (type, operating conditions, etc).
                     o)        Chromatograms, printer tapes, and other recordings of raw data.
                     p)        Quantitation reports, data system outputs, and other data to link the
                              raw data to the results reported.
    9.1.3   Analyses of method blanks are required to demonstrate freedom from contamination
           (Section 4.3). The procedures and criteria for analysis of a method blank are described in
           Sections 9.5 and 15.6.
    9.1.4   The laboratory shall spike all samples with labeled compounds to monitor method
           performance. This test is described in Section 9.3. When results of these spikes indicate
16
Draft, March 1997

-------
                                                                                       Method 1668
          atypical method performance for samples, the samples are diluted to bring method
          performance within acceptable limits. Procedures for dilution are given in Section 17.5.
    9.1.5  The laboratory shall, on an ongoing basis, demonstrate through calibration verification and
          the analysis of the ongoing precision and recovery aliquot that the analytical system is in
          control. These procedures are described in Sections 15.1 through 15.5.
    9.1.6  The laboratory shall maintain records to define the quality of data that is generated.
          Development of accuracy statements is described in Section 9.4.
9.2  Initial precision and recovery (IPR)—To establish the ability to generate acceptable precision and
     recovery, the analyst shall perform the following operations.
    9.2.1  For low solids (aqueous, < 1% solids) samples, extract, concentrate, and analyze four 1-L
          aliquots of reagent water spiked with the diluted labeled compound spiking solution
          (Section 7.10.3) and the precision and recovery standard (Section 7.14) according to the
          procedures in  Sections 11 through  18. For an alternative sample matrix, four aliquots of the
          alternative  reference matrix (Section 7.6) are used. All sample processing steps that are to
          be used for processing samples, including preparation (Section 11), extraction (Section 12),
          and cleanup (Section 13), shall be  included in this test.
    9.2.2  Using results of the set of four analyses, compute the average concentration (X) of the
          extracts in  ng/mL and the standard deviation of the concentration (s) in ng/mL for each
          compound, by isotope dilution for  PCBs with a labeled analog, and by internal standard for
          the PCBs without a labeled analog, and the labeled compounds.
    9.2.3  For each PCB and labeled compound, compare s and X with the corresponding limits for
          initial precision and recovery in Table 6. If PCBs 77, 126, and 169 only are to be
          determined, compare s and X with the corresponding limits for initial precision and
          recovery in Table 6a. If s and X for all  compounds meet the acceptance criteria, system
          performance is acceptable and analysis of blanks and samples may begin. If, however, any
          individual s exceeds the precision limit  or any individual X falls outside the range for
          accuracy, system performance is unacceptable for that compound. Correct the problem and
          repeat the test (Section 9.2).
9.3  The laboratory shall spike all samples with the diluted labeled compound spiking solution
     (Section 7.10.3)  to assess method performance on the sample matrix.
    9.3.1  Analyze each  sample according to the procedures in Sections 11  through 18.
    9.3.2  Compute the percent recovery of the labeled compounds and the cleanup standard using the
          internal standard method (Section  17.2).
    9.3.3  The recovery  of each labeled compound must be within the limits in Table 7 when all of
          the toxic PCBs are determined, and within the limits in Table 7a when PCBs 77, 126, and
          169 only are determined. If the recovery of any compound falls outside of these limits,
          method performance is unacceptable for that compound in that sample. Additional cleanup
          procedures must then be employed to attempt to bring the recovery within the normal
          range. If the recovery cannot be brought within the normal range after all cleanup
          procedures have been employed, water samples are diluted and smaller amounts of soils,
          sludges, sediments, and other matrices are analyzed per Section 18.4.
Draft, March 1997
                                                                                              17

-------
Method 1668
9.4  Recovery of labeled compounds from samples should be assessed and records should be
     maintained.
    9.4.1  After the analysis of five samples of a given matrix type (water, soil, sludge, pulp, etc.) for
          which the labeled compounds pass the tests in Section 9.3, compute the average percent
          recovery (R) and the standard deviation of the percent recovery (SR) for the labeled
          compounds only. Express the assessment as a percent recovery interval from R - 2SR to R
          + 2SR for each matrix. For example, if R = 90% andRS  = 10% for five analyses of pulp,
          the recovery interval is expressed as 70 to  110%.
    9.4.2  Update the accuracy assessment for each labeled compound  in each matrix on a regular
          basis (e.g., after each five to ten new measurements).
9.5  Method blanks—Reference matrix method blanks are analyzed to  demonstrate freedom from
     contamination (Section 4.3).
    9.5.1  Prepare, extract, clean up, and concentrate  a method blank with each sample batch (samples
          of tide same matrix started through the extraction process on the same 12-hour shift, to a
          maximum of 20 samples). The  matrix for the method blank  shall be similar to sample
          matrix for the batch; e.g., a 1-L reagent water blank (Section 7.6.1), high-solids reference
          matrix blank (Section 7.6.2), paper matrix blank (Section 7.6.3), tissue blank (Section
          7.6.4), or alternative reference matrix blank (Section 7.6.5).  Analyze the blank immediately
          after analysis of the OPR (Section 15.5) to demonstrate freedom from contamination.
    9.5.2  If any PCB (Table 1) is found in the blank at greater than the minimum level (Table 2) or
          one-third the regulatory compliance level whichever is greater, or if any potentially
          interfering compound is found in the blank at the minimum  level for each level of
          chlorination given in Table 2 (assuming a response factor of 1 relative to the internal
          standard at that level  of chlorination for compounds not listed in Table 1), analysis of
          samples is halted until the blank associated with the sample batch shows no evidence of
          contamination at this  level. All samples must be associated with an uncontaminated method
          blank before the  results for those samples may be  reported for regulatory compliance
          purposes.
9.6  QC Check Sample—Analyze the QC Check Sample (Section 7.16) periodically to assure the
     accuracy of calibration standards and the overall reliability of the analytical process. It is
     suggested that the QC Check Sample be analyzed at least quarterly.
9.7  The specifications contained in this method can  be met  if the apparatus used is calibrated
     properly and then maintained in a calibrated state. The standards  used for calibration (Section
      10), calibration verification (Section 15.3), and for initial  (Section 9.2) and ongoing (Section
      15.5) precision and recovery should be identical, so that the most precise results will be obtained.
     A GC/MS instrument will provide the most reproducible results if dedicated to the settings and
     conditions required for the analyses of PCBs by this method.
9.8  Depending  on specific program requirements, field replicates may be collected to determine the
     precision of the sampling technique, and spiked samples may be required to determine the
      accuracy of the analysis when the internal standard method is used.
                                                                                    Draft, March 1997

-------
                                                                                        Method 1668,
10.0   Calibration
 10.1   Establish the operating conditions necessary to meet the minimum retention times for the internal
       standards in Section 10.2.4 and the relative retention times for the PCBs in Table 2.
    .10.1.1   Suggested GC operating conditions:
                       Injector temperature:
                       Interface temperature:
                       Initial temperature:
                       Initial time:
                       Temperature program:
290°C
290°C
150°C
2 minutes
150 to 200°C at 10°C/minute
200 to 280°C at 2.0°C/minute
     Note:  All portions of the column that connect the GC to the ion source shall remain at or
     above the interface temperature specified above during analysis to preclude condensation of less
     volatile compounds.                     •	_	

            The GC conditions may be optimized for compound separation and sensitivity. Once
            optimized, the same GC conditions must be used for the analysis of all standards, blanks,
            IPR and OPR aliquots, and samples.
     10.1.2  Mass spectrometer (MS) resolution—Obtain a selected ion current profile (SICP) of each
            analyte in Table 3 at the two exact m/z's specified in Table 8 and at >10,000 resolving
            power by injecting an authentic standard of the PCBs either singly or as  part of  a mixture
            in which there is no interference between closely eluted components.
        10.1.2.1   The analysis time for PCBs may exceed the long-term mass stability of the mass
                  spectrometer. Because the instrument is  operated in  the high-resolution mode, mass
                  drifts of a few ppm (e.g., 5 ppm in mass) can have serious adverse effects  on
                  instrument performance. Therefore, a mass-drift correction is mandatory and a lock-
                  mass m/z from PFK is used for  drift correction. The lock-mass m/z is dependent on
                  the exact m/z's monitored within each descriptor, as shown in Table 8. The level of
                  PFK metered into the HRMS during analyses should be adjusted so that the amplitude
                  of the most intense selected lock-mass m/z signal (regardless of the descriptor
                  number) does not exceed 10% of the full-scale deflection for a given set of detector
                  parameters. Under those conditions, sensitivity changes that might occur during  the
                  analysis can be more effectively monitored.                                 .       .

     Note:  Excessive PFK (or any other reference substance) may cause noise problems and
     contamination of the ion source necessitating  increased frequency of source cleaning.       	

        10.1.2.2   If the HRMS has the capability  to monitor resolution during the  analysis, it is
                  acceptable to terminate the analysis when the resolution falls below 10,000 to save
                  reanalysis time.
   Draff, March 1997
                                                                                                19

-------
 Method 1668
       10.1.2.3   Using a PFK molecular leak, tune the instrument to.meet the minimum required
                resolving power of 10,000 (10% valley) at m/z 304.9824 or any other reference signal
                close to m/z 305 (from PeCB). For each descriptor (Table 8), monitor and record the
                resolution and exact m/z's of three to five reference peaks covering the mass range of
                the descriptor. The resolution must be greater than or equal to 10,000, and the
                deviation between the exact m/z and the theoretical m/z (Table 8) for each exact m/z
                monitored must be less than 5 ppm.
10.2  Ion abundance ratios, minimum levels, signal-to-noise ratios, and absolute retention  times—
      Choose an injection volume of either 1- or 2-uL, consistent with the capability of the
      HRGC/HRMS instrument. Inject a 1 or 2 uL aliquot of the CS1  calibration solution (Table 4)
      using the GC conditions from Section 10.1.1. If PCBs 77, 126, and 169 only are to  be
      determined, the operating conditions and specifications below apply to analysis of those
      compounds only.
    10.2.1   Measure the SICP areas for each analyte,. and compute  the ion abundance ratios at the exact
           m/z's specified in Table 8. Compare the computed ratio to the theoretical ratio given in
           Table 9.
       105.1.1   The exact m/z's to be monitored in each descriptor are shown in Table 8. Each group
                or descriptor shall be monitored in succession as a function of GC retention time to
                ensure that all of the toxic  PCBs are detected. Additional m/z's may be monitored in
                each descriptor, and the m/z's may be divided among more than the descriptors listed
                in Table 8, provided that the laboratory is able to monitor the m/z's of all the PCBs
                that may elute from the GC in a given retention-time .window. If PCBs 77,  126, and
                169 only are to be determined, the descriptors may be modified to include only the
                exact m/z's for the tetra-, penta-, and hexa-,  congeners, and the lock m/z's.
       10.2.1.2  The mass spectrometer shall be operated in a mass-drift correction mode, using PFK
                to provide lock m/z's. The  lock mass for each group of m/z's is shown in Table 8.
                Each lock mass shall be monitored and shall not vary by more than ±20% throughout
                its respective retention time window. Variations of the lock mass by more than 20%
                indicate the presence of coeluting interferences that may significantly reduce the
                sensitivity of the mass spectrometer. Reinjection of another aliquot of the sample
                extract will not resolve  the problem. Additional cleanup of the extract may be required
                to remove the interferences.
    105.2   All PCBs and labeled compounds in the CS1 standard  shall be within the QC limits in
            Table 9 for their respective ion abundance ratios;  otherwise, the mass spectrometer shall be
            adjusted and this test repeated until the m/z ratios fall within the limits specified. If the
            adjustment alters the resolution of the mass spectrometer, resolution shall be verified
            (Section 10.1.2) prior to repeat of the test.
    10.2.3   Verify that the HRGC/HRMS instrument meets the minimum levels in Table 2. The peaks
            representing the PCBs and labeled compounds in  the CS1 calibration standard must have
            signal-to-noise ratios (S/N) greater than or equal to 10.0. Otherwise, the mass spectrometer
            shall be adjusted and this test repeated until the minimum levels in Table 2 are met.
                                                                                     Draft, March 1997

-------
                                                                                       Method 1668
    10.2.4   The absolute retention time of PCB  169, (Section 7.12) shall exceed 20 minutes on the
           SPB-Octyl column, and the retention time of PCB 157 shall exceed 25 minutes on the DB-
           1 column; otherwise, the GC temperature program shall be adjusted and this test repeated
           until the above-stated minimum retention time criteria are met.
10.3  Retention- time windows — Analyze the window defining mixtures (Section 7.15) using the
      optimized temperature program in Section 10.1. Table 5 gives the elution order (first/last) of the
      window-defining compounds. If PCBs 77, 126, and 169 only are to be determined, the window-
      defining tetra-, penta-, and hepta-PCBs are the only compounds that need to be tested.
10.4  Isomer specificity.
    10.4.1   Analyze the isomer specificity test standards (Section 7.15) using the procedure in Section
           14 and the optimized conditions for sample  analysis (Section 10.1.1).
    10.4.2   Compute the percent valley between the GC peaks that elute most closely to PCB  126 and
           169 on the SPB-Octyl column and to PCB 156/157 on the DB-1 column, per Figures 6 and
           7.                  '   .    ..'           '                 '    •
    10.4.3   Verify that the height of the valley between the most closely eluted isomers and the PCBs
           given in Section 10.4.2 is less than 25% (computed as  100 x/y in Figures 6 and 7). If the
           valley exceeds 25%,  adjust the analytical conditions and repeat the test or replace the GC
           column  and  recalibrate  (Sections 10.1.2 through  10.7).
10.5  Calibration by isotope dilution — Isotope dilution calibration is used for the native PCBs for
      which labeled compounds  are added to samples prior to extraction. The reference compound for
      each native compound is shown in Table  2.          .
    10.5.1   A calibration curve encompassing the concentration range is prepared for each compound to
           be determined. The relative response (RR) (labeled to native) vs. concentration in  standard
           solutions is plotted or computed using a linear regression. Relative response is determined
           according to the procedures described below. Five calibration points are employed.
    10.5.2   The response of each native PCB relative to its labeled analog is determined using the area
           responses of both the primary and secondary exact m/z's specified in Table 8, for each
           calibration standard as follows:
                                      E>D _
                                          ~
(Al
                                                  + A2) C
                                                  ______  n
                                            (Al + A2) C
                                                *      *   .I*
  Where:
  Al and A2  = The areas of the primary and secondary m/z's for the PCB.
     n        n                            _           .                           *
   Al  and A2  = The areas of the primary and secondary m/z's for the labeled compound.
            C  - The concentration of the labeled compound in the calibration standard (Table 4).
            C  = The concentration of the native compound in the calibration standard (Table 4).
             n


    10.5.3   To calibrate the analytical system by isotope dilution, inject  a volume of calibration
           standards CS1 through CSS (Section 7:13 and Table 4) identical to the volume chosen in
 Draft, March 1997
                                                 21

-------
 Method 1668
           Section 10.2, using the procedure in Section  14 and the conditions in Section 10.1.1 and
           Table 2. Compute the relative, response (RR) at each concentration.
    10.5.4  Linearity—If the relative response for any compound is constant (less than 20% coefficient
           of variation) over the five-point calibration range, an averaged relative response may be
           used for that compound; otherwise, the complete calibration curve for that compound shall
           be used over the five-point calibration range.
10.6  Calibration by internal standard—The internal standard method is applied to determination of the
      native PCBs for which a labeled compound is not available and to the determination of labeled
      compounds for intralaboratory statistics (Sections 9.4 and 15.5.4).
    10.6.1  Response factors—Calibration requires the determination of response factors (RF) defined
           by the following equation:
                                             (Als + A2s)
                                      RF =
      Where:
       Als and A2S = The areas of the primary and secondary m/z?s for the PCB.
      Alh and A2^ = The areas of the primary and secondary mlz?s for the internal standard.
                C^ = The concentration of the internal standard (Table 4).
                C  = The concentration of the compound in the  calibration standard (Table 4).
    Note:  There is only one m/zfor PCBs 81 and 111 (see Table 8).
    10.6.2   To calibrate the analytical system by internal standard, inject 1.0 or 2.0 uL of calibration
            standards CS1 through CSS (Section 7.13 and Table 4) using the procedure in Section 14
            and the conditions in Section 10.1.1 and Table 2. Compute the response factor (RF) at each
            concentration.
    10.6.3   Linearity—If the response factor (RF) for any compound is constant (less than 35%
            coefficient of variation) over the five-point calibration range, an averaged response factor
            may be used for that compound; otherwise, the complete calibration curve for that
            compound shall be used over the five-point range.
10.7  Combined calibration—By using calibration solutions (Section 7.13 and Table 4) containing the
      native PCBs, labeled compounds, and the internal standards, a single set of analyses can be used
      to produce calibration curves for the isotope dilution and internal standard methods. These curves
      are verified (Section 15.3) each shift by analyzing the calibration verification  standard (VER,
      Table 4). Recalibration is required if any of the calibration verification criteria (Section 15.3)
      cannot be met.
10.8  Data storage—MS  data shall be collected, recorded, and stored.
    10.8.1   Data acquisition—The signal at each exact m/z shall be collected repetitively throughout the
            monitoring period and stored on a mass storage device.
    10.8.2   Response factors and multipoint calibrations—The data system shall be used to record and
            maintain lists of response factors (response ratios for isotope dilution) and multipoint

 25	"   ~         :                       Draft, March 1997

-------
                                                                                         Method 1668
            calibration curves. Computations of relative standard deviation (coefficient of variation)
            shall be used to test calibration linearity. Statistics on initial performance (Section 9.2) and
            ongoing performance (Section 15.5) should be computed and ^maintained, either on the
            instrument data system or on a separate computer system.

11.0   Sample Preparation

 11.1   Sample preparation involves modifying the physical form of the sample so that the PCBs can be
       extracted efficiently. In general, the samples must be in a liquid form or in the form of finely
       divided solids in order for efficient extraction to take place. Table 10 lists the phases and
       suggested quantities for extraction of various sample matrices.
       For samples known or expected to contain high levels of the PCBs, the smallest sample size
       representative of the entire sample should be used (see Section 17.5).
       For all samples, the blank and EPR/OPR aliquots must be processed through the same steps as
       the sample to check for contamination and losses in the preparation processes.
    11.1.1   For samples that  contain particles, percent solids and particle size are determined using the
            procedures in Sections 11.2 and 11.3, respectively.
    11.1.2  Aqueous samples—Because PCBs-may be bound to suspended particles, the preparation of
            aqueous samples  is dependent on the solids content of the sample.
        11.1.2.1   Aqueous samples containing 1%  solids or less are prepared  per Section 11,4 and
                 extracted directly using the SPE technique in 12.2.
        11.1.2.2   For aqueous samples containing greater than 1% solids, a sample aliquot sufficient to
                 provide  10  g of dry solids is used as-described  in Section 11.5..
    11.1.3  Solid samples are prepared using the procedure described in Section 11.5 followed by
            extraction via the SDS procedure in Section 12.3.
    11.1.4  Multiphase samples—The phase(s) containing the  PCBs is separated from the non-PCB
            phase using pressure filtration and centrifugation as described in Section 11.6. The PCBs
            will be in the organic phase in a multiphase sample in which an organic phase exists.
    11.1.5  Procedures for grinding, homogenization, and blending of various sample phases are given
            in Section 11.7.
    11.1.6  Tissue samples—Preparation procedures for fish and other tissues are given in Section 11.8.
 11.2   Determination of percent suspended  solids.

    Note:  This aliquot is used for determining the solids content of the sample, not for
    determination of PCBs.

    11.2.1   Aqueous liquids and multi-phase samples consisting  of mainly an aqueous phase.
        11.2.1.1   Desiccate and weigh a GF/D filter (Section  6.5.3)  to three significant figures.
        11.2.1.2  Filter 10.0 ± 0.02 mL of well-mixed sample through the filter.
        11.2.1.3  Dry the filter a minimum of 12 hours at 110 ± 5°C and cool in a desiccator.
  Draft, March 1997
                                                                                                 23

-------
 Method 1668
       11.2.1.4  Calculate percent solids as follows:

    n.   ,.,    weight of sample aliquot plus filter after drying (g) - weight of filter (g)
    % solids  = 	—	-	-	
                                                  10 g
   11.2.2  Non-aqueous liquids, solids, semi-solid samples, and multi-phase samples in which the
           main phase is not aqueous, but not tissues.
       11.2.2.1   Weigh 5 to 10 g of sample to three significant figures in a tared beaker.
       115.2.2   Dry a minimum of 12 hours  at ll'O ± 5°C, and cool in a desiccator.
       11.2*2.3   Calculate percent solids as follows:

                     % solids  =  wei8ht °fsamPle aliquot after drying  x
                                 •weight of sample aliquot before drying
11.3   Determination of particle size.
   11.3.1   Spread the dried sample from Section 11.2.2.2 on a piece of filter paper or aluminum foil
           in a fume hood or glove box.
   11.3.2   Estimate the size of the particles in the sample. If the size of the largest particles is greater
           than 1 mm, the particle size must be reduced to 1 mm or less prior to extraction using the
           procedures in Section 11.7.
11.4   Preparation of aqueous samples containing 1% suspended solids or less.
   11.4.1   Aqueous samples containing 1% suspended solids or less are prepared using the procedure
           below and extracted using the SPE technique in Section  12.2.
   11.4.2   Preparation of sample and QC aliquots.
       11.4.2.1   Mark the original level of the sample on the sample bottle for reference. Weigh the
                 sample-plus bottle to ±1 g.
       11.4.2.2   Spike 1.0 mL of the diluted labeled-compound spiking solution (Section 7.10.3) into
     .            the sample bottle. Cap the bottle and mix the sample by careful shaking. Allow the
                 sample to equilibrate for 1 to 2 hours, with occasional shaking.
       11.4.2.3   For each sample or  sample batch (to a maximum of 20 samples) to be extracted
                 during the same 12-hour shift, place two 1.0-L aliquots of reagent water in clean
                 sample bottles or flasks.
       11.4.2.4   Spike 1.0 mL of the diluted labeled-compound spiking solution (Section 7.10.3) into
                 both  reagent water aliquots. One of these aliquots will serve as the method blank.
       11.4.2.5   Spike 1.0 mL of the PAR standard (Section 7.14) into the remaining reagent water
                 aliquot. This aliquot will serve as the OPR (Section 15.5).
       11.4.2.6   Add  5 mL of methanol to the sample and QC aliquots. Cap and shake the sample and
                 QC aliquots to mix  thoroughly and proceed to Section 12.2 for extraction.
                                                                                     Draft, March 1997
24"

-------
                                                                                        Method1668
11.5   Preparation of samples containing greater than 1 % solids.
   11.5.1   Weigh a well-mixed aliquot of each sample (of the same matrix type) sufficient to provide
           10 g of dry solids (based on the solids determination in Section 11.2) into a clean beaker or
           glass jar.
   11.5.2   Spike 1.0 mL of the diluted labeled compound spiking solution (Section 7,10.3) into the
           sample.
   11.5.3   For each sample or sample batch (to a maximum of 20 samples) to be extracted during the
           same 12 hour shift, weigh two 10-g aliquots of the appropriate reference matrix (Section
           7.6) into clean beakers or glass jars.
   11.5.4   Spike 1.0 mL of the diluted labeled compound spiking solution (Section 7.10.3) into each
           reference matrix aliquot. One  aliquot will serve as the method blank. Spike 1.0 mL of the
           PAR standard (Section 7.14) into the other reference matrix aliquot. This aliquot will serve
           as the OPR (Section 15.5).
   11.5.5   Stir or tumble and equilibrate  the aliquots for 1 to 2 hours.
   11.5.6   Decant  excess water. If necessary to remove water, filter the sample through a glass-fiber
           filter (Section 6.5.6) and discard the aqueous liquid.
   11.5.7   If particles >1 mm are present in the sample (as determined in Section 11.3.2), spread the
           sample  on clean aluminum foil in a hood. After the sample is dry, grind to reduce the
           particle size (Section 11.7).
   11.5.8   Extract the sample and QC aliquots using the SDS procedure in Section 12.3.
11.6   Multiphase samples.
   11.6.1   Using the percent solids determined in Section 11.2.1 or 11.2.2, determine the volume of
           sample  that will provide 10 g of solids, up to 1 L of sample.
   11.6.2   Pressure filter the amount of sample determined in Section 11.6.1 through Whatman GF/D
           glass-fiber filter paper (Section 6.5.3). Pressure filter the blank and OPR aliquots through
           GF/D papers also. If necessary to separate the phases and/or settle the solids, centrifuge
           these aliquots prior to filtration.;
   11.6.3   Discard any aqueous phase (if present). Remove any non-aqueous liquid present and reserve
           the maximum amount filtered from the sample (Section 11.6.1) or 10 g, whichever-is less,
           for combination with the solid phase (Section 12.3.5).
   11.6.4   If particles >1 mm are present in the sample (as determined in Section  11.3.2) and  the
           sample is capable of being dried, spread the sample and QC  aliquots on clean aluminum
           foil in a hood. After the aliquots are dry, or if the sample cannot be dried, reduce the
           particle size using the procedures in Section 11.7 and extract the reduced particles using the
           SDS procedure  in Section  12.3. If particles  >1 mm are not present, extract the particles and
           filter in the sample and QC aliquots directly using the SDS procedure in Section  12.3.
11.7   Sample grinding, homogenization, or blending—Samples with particle sizes greater than  1  mm
      (as determined in Section 11.3.2) are subjected to grinding, homogenization, or blending. The
      method of reducing particle size to less than 1 mm is matrix-dependent. In general, hard particles
      can be reduced by grinding with a mortar and pestle. Softer particles can be reduced by grinding
      in a Wiley mill or meat grinder, by homogenization, or in a blender.
 Draft, March 1997.
                                                                                                25

-------
 Method 1668
   11.7.1   Each size-reducing preparation procedure on each matrix shall be verified by running the
           tests in Section 9.2 before the procedure is employed routinely.
   11.7.2   The grinding, homogenization, or blending procedures shall be carried out in a glove box  or
           fume hood to prevent particles from contaminating the work environment.
   11.7.3   Grinding—Certain papers and pulps, slurries, and amorphous solids can be ground in a
           Wiley  mill or heavy duty meat grinder. In some cases, reducing the temperature of the
           sample to freezing or to dry ice or liquid nitrogen temperatures can aid in the grinding
           process. Grind the sample aliquots from Sections  11.5.7 or 11.6.4 in a clean grinder. Do not
           allow the sample temperature to exceed 50°C. Grind the blank and reference matrix aliquots
           using a clean grinder.
   11.7.4   Homogenization or blending—Particles that are not ground effectively, or particles greater
           than 1 mm in size after grinding,  can often be reduced in size by high speed
           homogenization or blending. Homogenize and/or blend the particles or filter from Sections
           11.5.7 or 11.6.4 for the sample, blank, and OPR aliquots.
   11.7.5   Extract the aliquots using the SDS procedure in Section 12.3.
11.8   Fish and other tissues—Prior to processing tissue samples, the laboratory must determine the
      exact tissue to be analyzed. Common requests  for analysis of fish tissue include whole fish-skin
      on, whole fish-skin removed, edible fish fillets (filleted in the field or by the laboratory), specific
      organs, and other portions. Once the  appropriate tissue has been determined, the sample must be
      homogenized.
   11.8.1   Homogenization.
       11.8.1.1  Samples are homogenized while still frozen, where practical. If the laboratory must
                dissect the whole fish to obtain the appropriate tissue for analysis, the unused tissues
                may be rapidly refrozen and stored in a clean glass jar for subsequent use.
       11.8.1.2  Each analysis requires 10 g  of tissue (wet weight). Therefore, the laboratory should
                homogenize at least 20 g of tissue to allow for re-extraction of a second aliquot of the
                same homogenized sample,  if re-analysis is required. When whole fish analysis is
                necessary, the entire fish is  homogenized.
       11.8.1.3  Homogenize the sample in a tissue homogenizer (Section 6.3.3)  or grind in a meat
                grinder (Section 6.3.4). Cut  tissue too large to feed into the grinder into smaller
                pieces. To assure homogeneity, grind three times.
       11.8.1.4  Transfer approximately 10 g (wet weight) of homogenized  tissue to a clean, tared,
                400- to 500-mL beaker.
       11.8.1.5  Transfer the remaining homogenized tissue to a  clean jar with a fluoropolymer-lined
                lid. Seal the jar and store  the tissue at <-10°C. Return  any  tissue that was not
                homogenized to its original  container and store at <-10°C.
    11.8.2  QC aliquots.
       11.8.2.1  Prepare a method blank by  adding approximately 10 g of the oily liquid reference
                matrix (Section 7.6.4) to a 400- to 500-mL  beaker.
       11.8.2.2  Prepare a precision and recovery aliquot by adding approximately 10 g of the oily
                 liquid reference matrix (Section 7.6.4) to a separate 400- to 500-mL beaker. Record


 JS	;                           ~:                    Draft, March  1997

-------
                                                                                          Method 1668
                  the .weight to the nearest 10 mg. If the initial precision and recovery test is to be
                  performed, use four aliquots; if the ongoing precision and recovery test is to be
                  performed, use a single aliquot.
     11.8.3   Spiking.
         11.8.3.1   Spike 1.0 mL of the labeled compound spiking solution (Section 7.10.3) into the
                  sample, blank, and OPR aliquot.
         11.8.3.2   Spike 1.0 mL of the PAR standard (Section 7,14) into the OPR aliquot.
     11.8.4   Extract the aliquots using the procedures in Section 12.4.

 12.0   Extraction and Concentration

  12.1   Extraction procedures include solid phase (Section 12.2) for aqueous liquids;  Soxhlet/Dean-Stark
        (Section 12.3) for solids  and filters; and Soxhlet extraction (Section 12.4) for tissues. Acid/base
        back-extraction (Section  12.5) is used for initial cleanup of extracts.
        Macro-concentration procedures include rotary evaporation (Section 12.6.1), heating mantle
        (Section 12.6.2), and Kudema-Danish (K-D) evaporation (Section 12.6.3). Micro-concentration
        uses nitrogen blowdown  (Section 12.7).
  12.2   SPEof samples containing less than 1% solids.
     12.2.1   Disk preparation.
         12.2.1.1   Remove the test tube from the suction flask (Figure 4). Place an SPE disk on the base
.                  of the filter holder and wet with methylene chloride. While holding a GMF 150 filter
                  above the SPE disk with tweezers, wet the filter with methylene chloride and lay the
                  filter on the SPE disk, making sure that air is not trapped between the filter and disk.
                  Clamp the filter and SPE disk between the  1-L glass  reservoir and the vacuum
                  filtration flask.
         12.2.1.2  Rinse the sides of the reservoir with approximately 15 mL of methylene chloride
                  using a squeeze bottle or pipet. Apply vacuum momentarily until a few drops appear
                  at the drip tip. Release the vacuum and allow the filter/disk to soak for approximately
                  one minute. Apply vacuum and draw all of the methylene chloride through the
                  filter/disk. Repeat the wash step with approximately 15 mL of acetone and allow the
                  filter/disk to air dry.
     12.2.2  Sample extraction.
         12.2.2.1  Pre-wet the disk by adding approximately 20 mL of methahol to the reservoir. Pull
                  most of the methanol through the filter/disk, retaining a layer of methanol
                  approximately 2 mm thick on the filter. Do not allow the filter/disk to go dry from
                  this point until the extraction is completed.
         12.2.2.2  Add approximately  20 mL of reagent water to the reservoir and pull most through,
                  leaving a layer approximately 2 mm thick on the filter/disk.
         12.2.2.3  Allow the sample (Section 11.4.2.2) to stand for 1-2  hours, if necessary, to settle the
                  suspended particles. Decant the clear layer  of the sample, the blank (Section 11.4.2.4),
                  or IPR/OPR aliquot (Section  11.4.2.5) into  the reservoir and turn  on the vacuum to
                  begin the extraction. Adjust the vacuum to  complete  the extraction in  no less than  10
   Draft, March 1997

-------
 Method 1668
                 minutes. For samples containing a high concentration of particles (suspended solids),
                 the extraction time may be one hour or longer.
       12.2.2.4   Before all of the sample has been pulled through the filter/disk, add approximately 50
                 mL of reagent water to the sample bottle, swirl to suspend the solids (if present), and
                 pour into the reservoir. Pull through the filter/disk. Use additional reagent water rinses
                 until all solids are removed.
       12.2.2.5   Before all of the sample and rinses have been pulled through the filter/disk, rinse the
                 sides of the reservoir with small portions of reagent water.
       12.2.2.6   Partially dry the filter/disk under vacuum for approximately  3 minutes.
    125.3  Elution of the filter/disk.
       12.2.3.1   Release the vacuum, remove the entire filter/disk/reservoir assembly from the vacuum
                 flask, and empty the flask. Insert a test tube for eluant collection into the flask. The
                 test tube should have sufficient capacity to contain the total  volume of the elution
                 solvent (approximately 50 mL) and should fit around the drip tip. The drip tip should
                 protrude into the test tube to preclude loss of sample from spattering when vacuum is
                 applied. Reassemble the filter/disk/reservoir assembly on the vacuum flask.
       125.3.2   Wet the filter/disk with 4-5 mL of acetone. Allow the acetone to spread evenly across
                 the disk and soak for 15-20 seconds. Pull the acetone through the disk, releasing the
                 vacuum when approximately 1  mm thickness remains on the filter.
       12.2.3.3   Rinse the sample  bottle with approximately 20 mL of methylene chloride and transfer
                 to the reservoir. Pull approximately half of the solvent through the filter/disk and
                 release the vacuum. Allow the filter/disk to soak for approximately 1 minute. Pull all
                 of the solvent through the disk. Repeat the bottle rinsing and elution step with another
                 20 mL of methylene chloride. Pull all of the solvent through the disk.
       12.2.3.4   Release the vacuum, remove the filter/disk/reservoir assembly, and remove the test
                 tube containing the  sample solution. Quantitatively transfer the solution to a 250-mL
                 separatory funnel  and proceed to Section 12.5 for back-extraction.
12.3  SDS extraction of samples containing particles.
    12.3.1  Charge a clean extraction thimble (Section 6.4.2.2) with 5.0 g of  100/200 mesh silica
           (Section 7.5.1.1) topped with 100 g of quartz sand (Section 7.3.2).

    Note:  Do not disturb the silica  layer throughout the  extraction process.

                                                    l1    •    ' •                 '             :  ' '
    12.3.2  Place the thimble in a clean extractor. Place 30 to 40 mL of toluene in the receiver and 200
           to 250 mL of toluene in the flask.
    12.3.3  Pre-extract the glassware by heating  the flask until the toluene is  boiling. When properly
           adjusted, 1 to 2 drops of  toluene will fall per second from the condenser tip into the
           receiver. Extract the apparatus for a minimum of 3 hours.
    12.3.4  After pre-extraction,  cool and disassemble the apparatus.  Rinse the thimble with toluene and
           allow to air dry.
 28
Draft, March 1997

-------
                                                                                        Method 16S8
   12.3.5   Load the wet sample and/or filter from Sections 11.5.8, 11.6.4, 11.7.3, or 11.7.4 and any
           nonaqueous liquid from Section 11.6.3 into the thimble and manually mix into the sand
           layer with a clean metal spatula, carefully breaking up any large lumps of sample.
   12.3.6   Reassemble the pre-extracted SDS apparatus, and add a fresh charge of toluene to the
           receiver and reflux flask. Apply power to the heating mantle to begin refluxing. Adjust the
           reflux rate to match the rate of percolation through the sand and silica beds until water
           removal lessens the restriction to toluene flow. Frequently check the apparatus for foaming
           during the first 2 hours of extraction. If foaming occurs, reduce the reflux rate until
           foaming subsides.
   12.3.7   Drain the water from the receiver at 1 to 2 hours and 8 to 9 hours, or sooner if the receiver
          - fills with water. Reflux the sample for a total of 16 to 24 hours. Cool and disassemble the
           apparatus. Record the total volume of water collected.
   12.3.8   Remove the distilling flask. Drain the water from the Dean-Stark receiver and add any
           toluene in the receiver to the extract hi the flask.
   12.3.9   Concentrate the extracts from particles (Sections 11.5-11.7) to approximately 10 mL using
           the rotary evaporator or heating mantle (Section 12.6.1 or 12.6.2), transfer to a 250-mL
           separatory  runnel, and proceed with back-extraction (Section 12.5).
12.4   Extraction of tissue.                          .                          .
   12.4.1   Add 30 to  40 g of powdered anhydrous sodium sulfate to each of the beakers  (Section
           •11.8.1,4, 11.8.2.1, and 11.8.2.2) and mix thoroughly. Cover the beakers with aluminum foil
           and allow to equilibrate for 12-24 hours. Remix prior to extraction to prevent clumping.
    12.4.2  Assemble and pre-extract the Soxhlet apparatus per Sections 12.3-1-12.3.4, except use the
           methylene chloride:n-hexane (1:1) mixture for the pre-extraction and rinsing and omit the
           quartz sand. The Dean-Stark moisture trap may also be omitted, if desired.  .
    12.4.3  Reassemble the pre-extracted Soxhlet apparatus and add a fresh charge of methylene
           chloridern-hexane to the reflux flask.
    12.4.4  Transfer the sample/sodium sulfate mixture (Section 12.4.1) to the Soxhlet thimble, and,
           install the  thimble in the Soxhlet apparatus.       .
    12.4.5  Rinse the beaker with several portions of solvent mixture and add to the thimble. Fill the
           thimble/receiver with solvent. Extract for 18 to 24 hours.
    12.4.6  After extraction, cool  and disassemble the apparatus.
    12.4.7  Quantitatively transfer the extract to a macro-concentration device (Section 12.6), and
           concentrate to near dryness. Set aside the concentration apparatus for re-use.
    12.4.8  Complete  the removal of the solvent using the nitrogen blowdown procedure (Section  12.7)
           and a water bath temperature of 60°C.  Weigh the receiver,, record the weight, and return the
           receiver to the blowdown apparatus, concentrating the residue until a constant weight is
           obtained.                                                                   :
    12.4.9   Percent lipid determination—The lipid content is  determined by extraction of tissue with the
            same  solvent system (methylene chloridern-hexane) that was used in EPA's  National Dioxin
            Study (Reference 15) so that lipid contents are consistent with that study.
  Draft, March 1997
                                                                                                 29

-------
 Method 1668
       12.4.9.1   Redissolve the residue in the receiver in n-hexane and spike 1.0 mL of the cleanup
                 standard (Section 7.11) into the solution.
       12.4.9.2   Transfer the residue/n-hexane to the anthropogenic isolation column (Section 13.6.1),
                 retaining the boiling chips in the concentration apparatus. Use several rinses to assure
                 that all material is transferred. If necessary, sonicate or heat the receiver slightly to
                 assure that all material is re-dissolved. Allow the receiver to dry. Weigh the receiver
                 and boiling chips.
       12.4.9.3   Calculate the lipid content to the nearest three significant figures as follows:

                        D     * r- -j     *   *   Weight of residue (g)   ,nn
                        Percent lipid  content =	*££ * 100
                                                 Weight of tissue (g)
       12.4.9.4   It is not necessary to determine the lipid content of the blank, IPR, or OPR aliquots.
12.5  Back-extraction with base and acid.
    12.5.1  Spike 1.0 mL of the cleanup standard (Section 7.11) into the separatory funnels containing
           the sample and QC extracts from Section 12.2.3.4 or 12.3.9.
    12.5.2  Partition the extract against 50 mL of potassium hydroxide solution (Section 7.1.1). Shake
           for 2 minutes with periodic venting into a hood. Remove and discard the aqueous layer.
           Repeat the base washing until no color is visible in the aqueous layer to, a maximum of
           four washings. Minimize contact time between the extract and the base to prevent degrada-
           tion of the PCBs. Stronger potassium hydroxide solutions may be employed for back-
           extraction, provided that the laboratory meets the specifications for labeled compound
           recovery and demonstrates acceptable performance using the procedure in Section 9.2.
    12.5.3  Partition the extract against 50 mL of sodium chloride solution (Section 7.1.4) in the same
           way is with base. Discard the aqueous layer.
    12.5.4  Partition the extract against 50 mL of sulfuric acid (Section 7.1.2) in the same way as with
           base. Repeat the acid washing until no color is visible in the aqueous layer to a maximum
           of four washings.
    12.5.5  Repeat the partitioning against sodium chloride solution and discard the aqueous layer.
    12.5.6  Pour each extract through a drying column containing 7 to 10 cm of granular anhydrous
           sodium sulfate (Section 7.2.1). Rinse the separatory funnel with 30 to 50 mL of solvent,
           and pour through the drying column. Collect each extract in a round-bottom flask. Re-
           concentrate the sample and QC aliquots per Sections 12.6-12.7, and clean up the samples
           and QC aliquots per Section 13.
12.6  Macro-concentration—Extracts in toluene are concentrated using a rotary evaporator or a heating
      mantle; extracts in methylene chloride or n-hexane are concentrated using a rotary evaporator,
      heating mantle, or Kudema-Danish apparatus.
    12.6.1  Rotary evaporation—Concentrate the extracts in separate round-bottom flasks.
       12.6.1.1   Assemble the rotary evaporator according to manufacturer's instructions, and warm the
                 water bath to 45°C. On a daily basis, preclean the rotary evaporator by concentrating
 30
Draft, March 1997

-------
                                                                                       Method 1668
               100 mL of clean extraction solvent through the system. Archive both the concentrated
               solvent and the solvent in the catch flask for a contamination check if necessary.
               Between samples, three 2- to 3-mL aliquots of solvent should be rinsed down the feed
               tube into a waste beaker.
     12.6.1.2   Attach the round-bottom  flask containing the sample extract to the rotary evaporator.
               Slowly apply vacuum to  the system, and begin rotating the sample flask.
     12.6.1.3   Lower the flask into the water bath, and adjust the speed of rotation and the
               temperature as required to complete concentration in 15 to 20 minutes. At  the proper
               rate of concentration, the flow of solvent into the receiving flask will be steady, but
               no bumping or visible boiling of the extract will occur.

  Note:  If the rate of concentration is too fast, analyte loss may  occur.	    '

     12.6.1.4   When the  liquid in the concentration flask has reached  an apparent volume of
               approximately 2 mL, remove the flask from the water bath and stop the  rotation.
               Slowly and carefully, admit air into the system. Be sure not to open the valve so
               quickly  that the sample is blown out of the" flask. Rinse the feed tube with
               approximately 2 mL of solvent.
     12.6.1.5   Proceed to Section 12.6.4 for preparation for back-extraction or micro-concentration
               and solvent exchange.
  12.6.2   Heating mantle—Concentrate the extracts in separate round-bottom flasks.
     12.6.2.1   Add one or two clean boiling chips to the round-bottom flask, and attach a three-ball
               macro-Snyder column. Pre-wet the column by adding approximately 1 mL of solvent
               through the top. Place the round-bottom flask in a  heating mantle, and apply heat as
               required to complete the concentration in 15 to 20 minutes.  At the proper rate of
               distillation, the balls of the column will actively chatter, but the chambers  will not
               flood.
     12.6.2.2   When the liquid has reached an apparent volume of approximately 10 mL, remove the
               round-bottom flask from the heating mantle and allow  the solvent to drain and cool
               for at least 10 minutes. Remove the Snyder column and rinse the glass joint into the
               receiver with small portions of solvent.
     12.6.2.3   Proceed to Section  12.6.4 for preparation for back-extraction or micro-concentration
               and solvent exchange.
  12.6.3  Kuderna-Danish (K-D)—Concentrate the extracts in separate 500-mL  K-D flasks equipped
          with 10-mL concentrator tubes. The K-D technique is used for solvents such as methylene
          chloride and  n-hexane. Toluene is difficult to concentrate using the K-D technique unless a
          water bath fed by a steam generator is used.
      12.6.3.1   Add 1 to 2 clean boiling chips to the receiver. Attach a three-ball macro-Snyder
               column. Pre-wet the column by adding approximately  1 mL of solvent through the
               top. Place the K-D apparatus in a hot water bath so that the entire lower rounded
               surface of the flask is bathed with steam.
Draft, March 1997
                                                                                              31

-------
 Method 1668
       12.6.3.2   Adjust the vertical position of the apparatus and the water temperature as required to
                 complete the concentration in 15 to 20 minutes. At the proper rate of distillation, the
                 balls of the column will actively chatter but the chambers will not flood.
       12.6.3.3   When the liquid has reached an apparent volume of 1  mL, remove the K-D apparatus
                 from the bath and allow the solvent to drain and cool  for at least 10 minutes. Remove
                 the Snyder column and rinse the flask and its lower joint into the concentrator tube
                 with 1 to 2 mL of solvent. A 5-mL syringe is recommended for this operation.
       12.6.3.4   Remove the three-ball Snyder column, add a fresh boiling chip, and attach a two-ball
                 micro-Snyder column to the concentrator tube. Pre-wet the column by adding
                 approximately 0.5 mL of solvent through the top. Place the apparatus in the hot water
                 bath.
       12.6.3.5 •  Adjust the vertical position and the water temperature as required to complete the
                 concentration in 5 to 10 minutes. At the proper rate of distillation, the balls of the
                 column will actively chatter but the chambers will not flood.  .
       12.6.3.6   When the liquid reaches an apparent volume of 0.5 mL, remove the apparatus from
                 the water bath and allow to drain and  cool for at least 10 minutes.
       12.6.3.7   Proceed to 12.6.4 for preparation for back-extraction or micro-concentration and
                 solvent exchange.
    12.6.4  Preparation for back-extraction or micro-concentration and solvent exchange.
       12.6.4.1   For back-extraction (Section 12.5), transfer the extract to a 250-mL separatory funnel.
                 Rinse the concentration vessel with small portions of n-hexane, adjust the n-hexane
                 volume in the separatory funnel to 10  to 20 mL, and proceed to back-extraction
                 (Section 12.5).
       12.6.4.2   For determination of the weight of residue in  the extract or for clean-up procedures
                 other than back-extraction, transfer the extract to a blowdown vial using 2-3 rinses of
                 solvent. Proceed with micro-concentration and solvent exchange (Section 12.7).
12.7  Micro-concentration and solvent exchange.
    12.7.1  Extracts to be subjected to GPC or HPLC cleanup are exchanged into methylene chloride.
           Extracts to be cleaned up using silica gel, carbon, and/or Florisil are exchanged into n-
           hexane.
    12.7.2  Transfer the vial containing the sample extract to a nitrogen blowdown device. Adjust the
           flow of nitrogen so that the surface of the solvent is just visibly disturbed.

    Note:  A large vortex in the solvent may cause  analyte loss.

    12.7.3  Lower the vial into a 45°C water bath and continue concentrating.
       12.7.3.1   If the extract is to be concentrated to dryness for weight determination (Sections
                 12.4.8 and 13.6.4), blow dry until a constant weight is obtained.
       12.7.3.2   If the extract is to be concentrated for injection into the GC/MS or the solvent is to  be
                 exchanged for extract cleanup, proceed as follows:
 32
Draft, March 1997

-------
                                                                                        Method 1668
    12.7.4   When the volume of the liquid is approximately 100 jiL, add 2 to 3 mL of the desired
            solvent (methylene chloride for GPC and HPLC, or n-hexane for the other cleanups) and
            continue concentration to approximately 100 uL. Repeat the addition of solvent and
            concentrate once more.                                      .
    12.7.5   If the extract is to be cleaned up by GPC, adjust the volume of the extract to 5.0 mL with
            methylene chloride. If the extract is to be cleaned up by HPLC, further concentrate the
            extract to 30 uL. Proceed with GPC or HPLC cleanup (Section 13.2 or 13.5, respectively).
    12.7.6   If the extract is to be cleaned up by column chromatography (silica gel, Carbopak/Celite,  or
            Florisil), bring the final volume to 1.0 mL with n-hexane. Proceed with column cleanups
            (Sections 13.3  - 13.4 and 13.7).
    12.7.7   If the extract is to be concentrated for injection into the GC/MS (Section  14), quantitatively
            transfer the extract to a 0.3-mL conical vial for final concentration, rinsing the larger vial
            with n-hexane  and adding the rinse to the conical vial. Reduce the volume to approximately
            100 |jL. Add 10 |jL of nonane to the vial, and evaporate the solvent to the level of the
            nonane. Seal the vial and label with the sample number. Store in the dark at room temper-
            ature until ready for GC/MS analysis.  If GC/MS analysis will not be performed on the same
            day, store the vial at <-10°C.

13.0   Extract Cleanup

 13.1   Cleanup may not be necessary for relatively clean samples (e.g., treated effluents, groundwater,
       drinking water). If particular circumstances  require the use of a cleanup procedure, the analyst
       may use any or all of the procedures below or any other appropriate procedure. Before using a
       cleanup procedure, the analyst must demonstrate that the requirements of Section 9.2 can be met
       using the cleanup procedure. If PCBs 77, 126, and 169 only are to be determined, the cleanup
       procedures may be optimized for isolation  of these compounds.
    13.1.1   Gel permeation chromatography (Section  13.2) removes high molecular weight interferences
            that cause GC column performance to degrade. It should be used for all soil and sediment
            extracts. It may be used for water extracts that are expected to contain high molecular
            weight organic compounds (e.g., polymeric materials, humic acids). It may also be used for
            tissue extracts after initial  cleanup on  the anthropogenic isolation column (Section 13.6).
    13.1.2   Acid, neutral,  and basic silica gel (Section 13.3) and Florisil (Section  13.7) are used to
            remove nonpolar and polar interferences.
    13.1.3   Carbopak/Celite (Section 13.4) can be used to separate PCBs 77, 126, and 169 from the
            mono-and di-ortho-substituted PCBs, if desired.
    13.1.4   HPLC (Section 13.5) is used to provide specificity for certain congeners  and congener
            groups.
    13.1.5   The anthropogenic isolation column (Section 13.6) is used for removal of lipids from tissue
            samples.
 13.2   Gel permeation chromatography (GPC).
    13.2.1   Column packing.
        13.2.1.1  Place 70 to 75 g of SX-3 Bio-beads (Section 6.7.1.1) in a 400- to 500-mL beaker.
  Draft, March 1997
                                                                                                33

-------
 Method 1668
       135.1.2  Cover the beads with methylene chloride and allow to swell overnight (a minimum of
                12 hours).
       13.2.1.3  Transfer the swelled beads to the column (Section 6.7.1.1) and pump solvent through
                the column, from bottom to top, at 4.5 to 5.5 mL/minute prior to connecting the
                column to the detector.
       13.2.1.4  After purging the column with solvent for 1  to 2 hours, adjust the column head
                pressure to 7 to 10 psig and purge'for 4 to 5 hours to remove air. Maintain  a head
                pressure of 7 to 10 psig. Connect the column to the detector (Section 6.7.1.4).
    13.2.2  Column calibration.
       13.2.2.1  Load 5 mL of the calibration  solution (Section 7.4) into the sample loop.
       13.2.2.2  Inject the calibration solution and record the signal from the detector. The elution
                pattern will be com oil, PCB  209, pentachlorophenol, perylene, and sulfur.
       13.2.2.3  Set the "dump time" to allow >85% removal of the corn oil and >85% collection of
                PCB 209.
       135.2.4  Set the "collect time" to the peak minimum between perylene and sulfur.
       13.2.2.5  Verify the calibration with the calibration solution after every 20 extracts. Calibration
                is verified if the recovery of the pentachlorophenol is greater than 85%. If calibration
                is not verified, the system shall be  recalibrated using the calibration solution, and the
                previous 20 samples shall be re-extracted and cleaned up using the calibrated GPC
                system.
    13.2.3  Extract cleanup—GPC requires that the column not be overloaded.  The column specified in
           this method is designed to handle a maximum of 0.5 g of high molecular weight  material in
           a 5-mL extract. If the extract is known or expected to contain more than 0.5 g, the extract
           is split into aliquots for GPC, and the aliquots are combined after elution from the column.
           The residue content of the extract may be obtained gravimetrically  by evaporating the
           solvent from a 50-uL aliquot.
       135.3.1  Filter the extract or load through the filter holder (Section 6.7.1.3) to remove the
                particles. Load the  5.0-mL extract onto the column.
       13.2.3.2  Elute the extract using  the calibration data determined in Section 13.2.2. Collect the
                eluate in a clean 400- to 500-mL beaker.
       135.3.3  Rinse the sample loading tube thoroughly with methylene  chloride between  extracts to
                prepare for the next sample.
       135.3.4  If a particularly dirty extract is encountered, a 5.0-mL methylene chloride blank shall
                be run through the  system to check for carry-over.
       135.3.5  Concentrate the eluate per Section  12.6 and Section 12.7 for further cleanup or
                injection into the GC/MS.
13.3  Silica gel cleanup.
    13.3.1  Place a glass-wool plug in a 15-mm ID chromatography column (Section 6.7.4.2). Pack the
           column bottom to top with 1 g silica gel (Section 7.5.1.1), 4 g basic silica  gel (Section
           7.5-1.3), 1 g silica gel, 8 g acid silica gel (Section 7.5.1.2), 2 g  silica gel, and 4 g granular
           anhydrous sodium sulfate (Section 7.2.1). Tap the column  to settle  the adsorbents.

 "54                                      '           '      "    !                     Draft, March 1997

-------
                                                                                        Method 1668
   13.3.2   Pre-elute the column with 50 to 100 mL of n-hexane. Close the stopcock when the n-
           hexane is within 1 mm of the sodium sulfate. Discard the eluate. Check the column for
           channeling. If channeling is present, discard the column and prepare another.
   13.3.3   Apply the concentrated extract to the column. Open the stopcock until the extract is within
           1 mm of the sodium sulfate.                                                    .
   13.3.4   Rinse the receiver twice with 1-mL portions of n-hexane, and apply separately to the
           column. Elute the PCBs with 25 mL of n-hexane and collect the eluate.
   13.3.5   Concentrate the eluate per Section 12.6 and 12.7 for further cleanup or injection into the
           HPLC or GC/MS.
   13.3.6   For extracts of samples known to contain large quantities of other organic compounds  (such
           as paper mill effluents), it may be advisable to increase the capacity of the silica gel
           column. This may be accomplished by increasing the strengths of the acid and basic silica
           gels. The acid silica gel (Section 7.5.1.2) may be increased in strength to as much as 44%
           w/w (7.9 g sulfuric acid added to 10 g silica gel). The basic silica gel (Section 7.5.1.3) may
           be increased in  strength to as much as 33% w/w (50 mL IN NaOH added to  100 g silica
           gel), or the potassium silicate (Section 7.5.1.4) may be  used.

   Note:  The use of stronger acid silica gel (44% w/w) may lead to charring of organic
   compounds in some  extracts. The charred material may retain some of the analytes and lead to
   lower recoveries of the PCBs. Increasing  the strengths of the acid and basic silica gel may also
   require different volumes of n-hexane than those specified above to elute the analytes from the
   column. Therefore, the performance of the method after such modifications must be verified by
   the procedure in Section 9.2.

13.4   Carbon column (Reference 16).
   13.4.1   Cut both ends from a 50-mL disposable serological pipet (Section 6.7.3.2) to produce a 20-
           cm column. Fire-polish both  ends and flare both ends if desired. Insert a glass-wool plug at
           one end, and pack the column with 3.6 g of Carbopak/Celite (Section 7.5.2.3) to form an
           adsorbent bed 20 cm long. Insert a glass-wool plug on  top of the bed to hold the adsorbent
           in place.
   13.4.2   Pre-elute the column with 20 mL each in succession of toluene, methylene chloride, and n-
           hexane.
   13.4.3   When the solvent is within 1 mm of the column packing, apply the n-hexane sample extract
           to the column. Rinse the sample container twice with 1-mL portions of n-hexane and apply
           separately to the column. Apply 2 mL of n-hexane to complete the transfer.
   13.4.4   Elute the column with 25 mL of n-hexane and collect the eluate. This fraction will contain
           the mono- and di-ortho PCBs. If carbon particles are present in the eluate, filter through
           glass-fiber filter paper.
   13.4.5   Elute the column with 15 mL of methanol and discard  the eluate. The fraction discarded
           will contain residual lipids and other potential interferents, if present.
 Draft, March 1997
                                                                                               35

-------
 Method 1668
   13.4.6  Elute the column with 15 mL of toluene and collect the eluate. This fraction will contain
           PCBs 77, 126, and 169. If carbon particles are present in the eluate, filter through glass-
           fiber filter paper.
   13.4.7  Concentrate the fractions per Section 12.6 and 12.7 for further cleanup or injection into the
           HPLC or GC/MS.
13.5   HPLC (Reference 17).
   13.5.1  Column calibration.
       13.5.1.1  Prepare a calibration standard containing the toxic congeners and other congeners of
                interest at a concentration of approximately TBD  pg/uL in methylene chloride.
       13.5.1.2  Inject 30 pL of the calibration solution into the HPLC and record the signal from the
                detector. Collect the eluant for reuse. The elution  order will  be the mono- through
                deca- congeners.
       13.5.1.3  Establish the collection time for the congeners of interest. Following calibration, flush
                the injection system with copious quantities of methylene chloride, including  a
                minimum of five 50-uL injections while the detector is monitored, to ensure that
                residual PCBs are removed from the system.
       13.5.1.4  Verify the calibration with the calibration solution after every 20 extracts. Calibration
                is verified if the recovery of the PCBs from the calibration standard  is 75 to  125%
                compared to the calibration (Section 13.5.1.1). If calibration is not verified, the system
                shall be recalibrated using the calibration solution, and the previous 20 samples  shall
                be re-extracted and cleaned up using the calibrated system.
    13.5.2  Extract cleanup—HPLC requires that the column not be overloaded. The column specified
           in this method is designed to handle a maximum of 30 uL of extract.  If the extract cannot
           be concentrated to less than 30 uL,  it is split into fractions and the fractions are combined
           after elution from the column.
       13.5.2.1   Rinse the sides of the vial twice with 30 uL of methylene chloride and reduce to 30
                 uL with the evaporation apparatus (Section 6.8.3).
       13.5.2.2   Inject the 30 uL extract into the HPLC.
       13.5.2.3   Elute the extract using the calibration data determined in Section 13.5.1. Collect the
                 fraction(s) in a clean 20-mL concentrator tube containing 5  mL  of n-hexane:acetone
                   (1:1 v/v).
       13.5.2.4   If an extract containing greater than TBD ng/mL of total PCBs is encountered, a 30-
                 pL methylene chloride blank shall be run through the system to  check for carry-over.
       13.5.2.5   Concentrate the eluate per Section 12.7 for injection into the GC/MS.
13.6   Anthropogenic isolation column (References 1-2)—Used for removal of lipids from tissue
       extracts!        .
    13.6.1   Prepare the column as given in Section 7.5.3.
    13.6.2   Pre-elute the column with 100  mL of n-hexane. Drain  the n-hexane layer to-the top of the
            column, but do not expose the  sodium sulfate.
                                                                                       Draft, March 1997

-------
                                                                                         Method1668
    13.6.3   Load the sample and rinses (Section 12.4.9.2) onto the column by draining each portion to
            the top of the bed. Elute the PGBs from the column into the apparatus used for
            concentration (Section 12.4.7) using 200 mL of n-hexane.
    13.6.4   Concentrate the cleaned up extract (Sections 12.6-12.7) to constant weight per Section
            12.7.3.1. If more than 500 mg of material remains, repeat the cleanup using a fresh
            anthropogenic isolation column.
    13.6.5   Redissolve  the extract in a solvent suitable for the additional cleanups to be used (Section
            13.2-13.5 and 13.7).
    13.6.6   Spike 1.0 mL of the cleanup standard (Section 7.11) into the residue/solvent.
    13.6.7   Clean up the  extract using the procedures in Sections 13.2-13.5 and 13.7. Florisil (Section
            13.7) and carbon (Section  13.4) are recommended as minimum additional cleanup steps.
    13.6.8   Following cleanup, concentrate the^ extract to 10 uL as described in Section 12.7 and
            proceed with  the analysis in Section 14.
 13.7   Florisil cleanup (Reference 18).                        '
    13.7.1   Begin to drain the n-hexane from the column (Section 7.5.4). Adjust the flow rate of eluant
            to 4.5-5.0 mL/min.
    13.7.2   When the n-hexane is within 1 mm of the sodium sulfate, apply the sample extract (in n-
            hexane) to  the column. Rinse the sample container twice with 1-mL portions of n-hexane
            and apply to the column.
   , 13.7.3   Elute the mono-ortho and di-ortho PCBs with approximately  165 mL of n-hexane and
            collect the  eluate. Elute the non-ortho co-planar PCBs with approximately 100 mL of 6%
            ethenn-hexane and collect the eluate. The exact volumes of solvents will need to be
            determined for each batch of Florisil. If the mono/di-ortho PCBs are hot to  be separated
            from the non-ortho co-planar PCBs, elute all PCBs with 6% ethenn-hexane.
    13.7.4   Concentrate the eluate(s) per Sections 12.6-12.7 for further cleanup or for injection into the
            HPLC or GC/MS.

14.0   HRGC/HRMS  Analysis

 14.1   Establish the operating conditions given in Section 10.1.
 14.2   Add 10 uL of the appropriate internal standard solution (Section 7.12) to the sample extract
       immediately prior to injection to minimize the possibility of loss by evaporation, adsorption, or
       reaction. If an extract is to be reanalyzed and evaporation has occurred, do not add more
       instrument internal standard solution. Rather, bring the extract back to its previous volume (e.g.,
       19 uL) with pure nonane only (18 uL if 2 uL injections are used).
 14.3   Inject 1.0 or 2.0 uL of the concentrated extract containing the internal standard solution, using
       on-column or splitless injection. The volume injected must be identical to the volume used for
       calibration (Section 10). Start the GC column initial  isothermal hold upon injection. Start MS
       data collection after the solvent peak elutes. Stop data collection after the 13C12-PCB 209 has
       eluted. If PCBs 77, 126, and 169 only are to be determined, stop data collection  after 13CI2-PCB
       169 has eluted. Return the column to the initial temperature for analysis of the next extract or
       standard.
  Draft, March 1997
                                                                                                37

-------
  Method 1668
15.0   System and Laboratory Performance

 15.1   At the beginning of each 12-hour shift during which analyses are performed, GC/MS system
       performance and calibration are verified for all native PCBs and labeled compounds. For these
       tests, analysis of the CSS calibration verification (VER) standard (Section 7.13 and Table 4) and
       the congener specificity test standards (Section 7.15 and Table 5) shall be used to verify all
       performance criteria. Adjustment and/or recalibration (Section  10) shall be performed until all
       performance criteria are met. Only after all performance criteria are met may samples, blanks,
       IPRs, and OPRs be analyzed.
 15.2   MS resolution—A static resolving power of at least 10,000 (10% valley definition) must be
       demonstrated at the appropriate m/z before any analysis is performed. Static resolving power
       checks must be performed at the beginning and at the end of each 12-hour shift according to
       procedures in Section  10.1.2. Corrective actions must be implemented whenever the resolving
       power does not meet the requirement.
 15.3   Calibration verification.
    15.3.1   Inject the VER standard using the procedure in Section 14.
    15.3.2   The m/z abundance ratios for all PCBs shall be within the limits in Table 9; otherwise, the
            mass spectrometer shall be adjusted until the m/z abundance ratios fall within the limits
            specified, and the verification test shall be repeated. If the adjustment alters the resolution
            of the mass  spectrometer, resolution shall be verified (Section 10.1.2) prior to repeat of the
            verification test.
    15.3.3   The peaks representing each native PCB and labeled compound in the VER standard must
            be present with a S/N of at least 10; otherwise, the mass spectrometer shall be adjusted and
            the verification test repeated.
    15.3.4   Compute the concentration of each native PCB compound by isotope dilution (Section 17.1)
            for those compounds that have labeled analogs (Table  1). Compute the concentration of
            each native compound that does not have a labeled analog and of each labeled compound
            by the internal standard method (Section 17.2). These concentrations are computed based on
            the calibration data in Section 10.
    15.3.5   For each compound, compare the concentration with the  calibration verification limit in
            Table 6. If PCBs 77, 126, and 169 only are to be determined, compare the concentration to
            the limit in Table 6a. If all compounds meet the acceptance criteria, calibration has been
            verified and analysis of standards and sample extracts may proceed. If, however, any
            compound fails its respective limit, the measurement system is not performing properly for
            that compound. In this event, prepare a fresh calibration  standard or correct the problem
            causing the failure and repeat the resolution (Section 15.2) and verification (Section 15.3)
            tests, or recalibrate (Section 10).
 15.4   Retention times and GC resolution.
    15.4.1   Retention times.
        15.4.1.1   Absolute—The absolute retention times of the GC/MS internal standards in the
                 verification test (Section 15.3) shall be within ±15  seconds of the retention times
                  obtained during calibration (Section 10.2.4).
  38
uratt, Marcn iyy/

-------
                                                                                       Method 1668
       15.4.1.2  Relative—The relative retention times of native PCBs and labeled compounds in the
                verification test (Section 15.3) shall be within 5 percent of the relative retention times
                given in Table 2.
   15.4.2   GC resolution.
       15.4.2.1  Inject the-isomer specificity standards (Section 7.15) on their respective columns.
       15.4.2.2  The valley height between PCBs 123 and 118 at m/z 325.8804 shall not exceed 10
                percent on the SPB-Octyl column, and the valley height between PCBs 156 and  157
                shall not exceed 10 percent at m/z 359.8415 on the DB-1 column (Figures 6 and 7).
   15.4.3   If the absolute retention time of any compound is not within the limits specified or if "the
           congeners are not resolved, the GC is not performing properly. In this event, adjust the GC
           and repeat the verification test (Section 15.3) or recalibrate (Section 10), or replace the GC
           column and either verify calibration or recalibrate.                    .
15.5   Ongoing precision and recovery.
   15.5.1   Analyze the extract of the ongoing precision and recovery (OPR) aliquot (Section 11.4.2.5,
           11.5.4,  11.6.2, 11.7.4, or 11.8.3.2) prior to analysis of samples  from the same batch.
   15.5.2   Compute the concentration of each native PCB by isotope dilution for those compounds
           that have labeled analogs (Section  17.1). Compute the concentration of the native PCBs that
           have no labeled analog and each labeled compound by the internal standard method
           (Section 17.2).
   15.5.3   For each PCB and labeled compound, compare the concentration to the OPR limits given in
           Table 6. If PCBs 77, 126, and 169 only are to be determined, compare the concentration to
           the limits in Table 6a. If all compounds meet the acceptance criteria,  system performance is
           acceptable and analysis of blanks and  samples may proceed. If, however, any individual
           concentration falls outside of the range given, the extraction/concentration processes are not
           being performed properly for that compound. In this event, correct the problem, re-prepare,
           extract, and clean up the sample batch  and repeat the ongoing precision and recovery test
           (Section 15.5).
   15.5.4   Add results that pass the specifications in Section 15.5.3 to initial and previous ongoing
           data for each compound in each matrix. Update QC charts to form a graphic representation
           of continued laboratory performance. Develop a statement of laboratory accuracy for each
           congener in each matrix type by calculating the average percent recovery (R) and the
           standard deviation of percent recovery  (SR). Express the accuracy as a recovery interval
           from R - 2SR to R + 2SR. For example, if R = 95% and § = 5%, the accuracy is 85 to
           105%.
15.6   Blank—Analyze the method blank extracted with each sample batch immediately following
      analysis of the OPR aliquot to demonstrate  freedom from contamination and freedom from
      carryover from the OPR analysis. The results of the analysis of the blank must meet the
      specifications in Section 9.5.2 before sample analyses  may proceed.
 Draft, March 1997
                                                                                               39

-------
  Method 1668
16.0   Qualitative Determination
  A PCB or labeled compound is identified in a standard, blank, or sample when all of the criteria in
  Sections 16.1 through 16.4 are met.
 16.1   The signals for the two exact m/z's in Table 8 must be present and must maximize within the
       same two seconds.
 16.2   The signal-to-noise ratio (S/N) for the GC peak at each exact m/z must be greater than or equal
       to 2.5 for each PCB detected in,a sample extract, and greater than or equal to 10 for all PCBs in
       the calibration standard (Sections 10.2.3 and 15.3.3).
 16.3   The ratio of the integrated areas of the two exact m/z's specified in Table 8 must be within the
       limit in Table 9, or within ±10 percent of the ratio in the midpoint (CSS) calibration or
       calibration verification (VER), whichever is most recent.
 16.4   The relative retention time of the peak for a toxic PCB must be within 5 percent of the relative
       retention times listed in Table 2. The retention time of peaks representing PCBs  other than the
       toxic PCBs must be within the retention time windows established in Section 10.3.
 16.5   Confirmatory analysis—Isomer specificity for PCBs 156 and  157 cannot be achieved on the
       SPB-Octyl column. Therefore, any  sample in which these PCBs are tentatively identified by
       analysis on the SPB-Octyl column and when rigorous identification is required must have a
       confirmatory analysis performed on a DB-1 or equivalent GC column. The operating conditions
       in Section 10.1.1 may be adjusted to optimize the analysis on the  second GC column, but the
       GC/MS must meet the mass resolution and calibration specifications in Section 10.
 16.6   If the criteria for identification in Sections  16.1-16.5 are not met, the PCB has not been identified
       and the results may not be reported for regulatory compliance purposes. If interferences preclude
       identification, a new aliquot of sample must be extracted, further cleaned up, and analyzed.

17.0   Quantitative Determination

 17.1   Isotope dilution quantitation—By adding a known amount of a labeled compound to every
       sample prior to extraction, correction for recovery of the PCB can be made because the native
       compound and its labeled analog exhibit similar effects upon extraction, concentration, and gas
       chromatography. Relative response (RR) values are used in conjunction with the initial cali-
       bration data described in Section 10.5 to determine  concentrations directly, so long as labeled
       compound spiking levels are constant, using the following equation:
                                       (ng/mL) =
                                                        A2n)
                                                 (Al, + A2,) RR
                        where:
                        Ca = The concentration  of the PCB in the extract.
                             The other terms are as  defined in Section 10.5.2

       Any peaks representing the other congeners are quantitated using an average of the response
       factors from all of the labeled PCBs isomers at the same level of chlorination.
                                                                                     Draft, March 1997
"40-

-------
                                                                                       Method 1668
17.2   Internal standard quantitation and labeled compound recovery.
   17.2.1   Compute the concentrations of labeled analogs (including the cleanup standard) in the
           extract using the response factors determined from the initial calibration data (Section 10.6)
           and the following equation:
                                                (Al  + A2t) C.
                                 C   (nglmL) =     s
                                  -'>*
                  where:
                  C^ = The concentration of the labeled compound in the extract.
                        The other terms are  as  defined in Section  10.6.1
    17.2.2   Using the concentration in the extract determined above, compute the percent recovery of
           the labeled compounds (including the cleanup standard) using the following equation:

                        r.        fn,\    Concentration found (lig/mL) „ 1An
                       Recovery (%) = ——	^-^	-x 100
                                        Concentration spiked (ug/wzL)
17.3  The concentration of a native PCB in the solid phase of the sample is computed using the
      concentration of the compound in the extract and the weight of the solids (Section 11.2.2.3), as
      follows:
                           Concentration in solid (nglkg)  =
                                                               w.
                      where:
                      C^ = The concentration of the compound in the extract.
                      V^ = The extract -volume  in mL.
                      Ws = The sample weight (dry weight)  in kg.
17.4  The concentration of a native PCB in the aqueous phase of the sample is computed using the
      concentration of the compound in the extract and the volume of water extracted (Section 11.4),
      as follows:
                       Concentration in aqueous phase (pgIL) =
                      where:
                      C^. = The concentration of the compound in the extract.
                      V^ = The extract volume  in mL.
                       V  = The sample volume  in liters.
 Draft, March 1997
                                                                                              41

-------
 Method 1668
17.5  If the SICP area at either quantitation m/z for any compound exceeds the calibration range of the
      system, a smaller sample aliquot is extracted.
    17.5.1   For aqueous  samples containing 1% solids or less, dilute 100 mL, 10 mL, etc., of sample to
           1 L with reagent water and re-prepare, extract, clean up, and analyze per Sections 11 -  14.
    17.5.2  For samples  containing greater than 1%  solids, extract an amount of sample equal to 1/10,
           1/100, etc., of the amount used in Section  11.5.1. Re-prepare, extract, clean up, and analyze
           per Sections  11-14.
    17.5.3  If a smaller sample size will not be representative of the entire sample, dilute the sample
           extract by a factor of 10, adjust the concentration of the instrument internal standard to 100
           pg/uL in the extract,  and analyze an aliquot of this diluted extract by the internal standard
           method.
17.6  Results are reported.to three significant figures for the PCBs and labeled compounds found in all
      standards, blanks,  and samples.
    17.6.1   Reporting units and levels.
       17.6.1.1   Aqueous  samples—Report results in  pg/L (parts-per-quadrillion).
       17.6.1.2   Samples containing greater than 1% solids (soils, sediments, filter cake, compost)—
                 Report results in ng/kg based on the  dry weight of the sample. Report the percent
                 soli'ds so  that the result may be corrected.
       17.6.1.3   Tissues—Report results in ng/kg of wet tissue, not on the basis of the lipid content of
                 the sample. Report the percent lipid content, so that the data user can calculate the
                 concentration on a lipid basis if desired.
       17.6.1.4   Reporting level.
           17.6.1.4.1   Standards (VER, IPR, OPR) and samples—Report results at or above the
                      minimum level (Table 2). Report results below the minimum level as not
                      detected or as required by the regulatory authority.
           17.6.1.4.2   Blanks—Report results above the MDL or as required by the regulatory
                      authority. Do not blank-correct results. If a blank accompanying a sample result
                      shows contamination above the MDL for the congener, flag the sample result
                      and report the results for the sample and the accompanying blank.
    17.6.2  Results for PCBs in  samples that have been  diluted are reported  at the least dilute level at
           which the areas  at the quantitation m/z's are within the calibration range (Section 17.5).
    17.6.3  For PCBs having a labeled analog, results are reported at the least dilute level at which the
           area at the quantitation m/z is within the calibration range (Section 17.5) and the labeled
           compound recovery is within the normal range for the method (Section  9.3 and Tables 6,
           6a, 7, and 7a).
    17.6.4  Additionally, if requested, the total concentration of all congeners at a given level of
           chlorination  (i.e., total TCB, total PeCB, total HxCB, etc.) may be reported by summing the
           concentrations of all congeners identified in that level of chlorination, including both the
           toxic and other congeners.
                                                                                      Draft, March 1997

-------
                                                                                       Method 1668
18.0   Analysis of Complex Samples

 18.1   Some samples may contain high levels (>10 ng/L; >1000 ng/kg) of the compounds of interest,
       interfering compounds, and/or polymeric materials. Some extracts  will not concentrate to 10 uL
       (Section 12.7); others may overload the GC column and/or mass spectrometer.
 18.2   Analyze a smaller aliquot of the sample (Section 17.5) when the extract will not concentrate to
       10 uL after all cleanup procedures have been exhausted.
 18.3   Several laboratories have reported that elimination of several of the toxic PCBs, particularly
       non-coplanar congeners 105, 114, 118, 123, 156, 157, and 167 is difficult.  Backgrounds of these
       congeners can therefore interfere with the determination of these congeners in environmental
       samples.  Care should therefore be exercised  in the determination  of these congeners.
 18.4   Recovery of labeled compounds—In most samples, recoveries of the labeled compounds will be
       similar to those from reagent water or from the alternate matrix (Section 7.6).
    18.4.1   If the recovery of any of the labeled compounds is outside of the normal range (Table 7), a
            diluted sample shall be analyzed (Section 17.5).
    18.4.2   If the recovery of any of the labeled compounds in the diluted sample is outside of normal
            range, the calibration verification standard (Section 7.13) shall be analyzed and calibration
            verified (Section 15.3).
    18.4.3   If the calibration cannot be verified, a new calibration rnust be performed and the original
            sample extract reanalyzed.
    18.4.4   If the calibration is verified and the diluted sample does not meet the limits for labeled
            compound recovery, the method does not apply to the sample being analyzed and the result
            may not be reported for regulatory compliance purposes. In this case, alternate extraction
            and cleanup procedures in this method must be employed  to resolve the interference. If all
            cleanup procedures in this method have been employed and  labeled compound recovery
            remains outside of the normal range, extraction and/or cleanup procedures that are beyond
            this scope of this method will be required to analyze these samples.

19.0   Method Performance

  For this draft version of Method 1668, performance was validated and preliminary data were collected
  in a single laboratory.                            .'•••.

20.0   Pollution Prevention


 20.1   The solvents used in this method pose little threat to the environment  when managed properly.
       The solvent evaporation techniques used in this method are amenable  to solvent recovery, and it
       is recommended that the laboratory recover solvents wherever feasible.
 20.2   Standards should be prepared in volumes consistent with laboratory use to minimize disposal of
       standards.
  Draft, March 1997
                                                                                               43

-------
  Method 1668
21.0   Waste Management

 21.1   It is the laboratory's responsibility to comply with all federal, state, and local regulations
       governing waste management, particularly the hazardous waste identification rules and land
       disposal restrictions, and to protect the air, water, and land by minimizing and controlling all
       releases from fume hoods and bench operations. Compliance is also required with any sewage
       discharge permits and regulations.
 21.2   Samples containing HC1 or H2SO4 to pH <2 are hazardous and must be neutralized before being
       poured down a drain or must be handled as hazardous waste.
 21.3   The PCBs decompose above 800°C. Low-level waste such as absorbent paper, tissues, animal
       remains, and plastic gloves may be burned in an appropriate incinerator. Gross quantities
       (milligrams) should be packaged securely and disposed of through commercial or governmental
       channels that are capable of handling extremely toxic wastes.
 21.4   [This section may need to be modified to accommodate the PCBs: Liquid or soluble waste
       should be dissolved in methanol or ethanol and irradiated with ultraviolet light with a wavelength
       shorter than 290 mri for several days. Use F40 BL or equivalent lamps. Analyze liquid wastes,
       and dispose of the solutions when the PCBs can no longer be detected.]
 21.5   For further information on waste management, consult  "The Waste Management Manual for
       Laboratory Personnel" and "Less is Better-Laboratory Chemical Management for Waste
       Reduction," available from the American Chemical Society's Department of Government
       Relations and Science Policy, 1155 16th Street N.W., Washington, D.C. 20036.

22.0   References

    1   Kuehl, D.W., B.C. Butterworth, J. Libal, and P. Marquis, "An Isotope Dilution High Resolution
       Gas Chromatographic-High Resolution Mass Spectrometric Method for the Determination of
       Coplanar Polychlorinated Biphenyls: Application to Fish and Marine Mammals," Chemosphere
       22:9-10, 849-858, 1991.
    2   Feltz, Kevin P., Donald E. Tillitt, Robert W. Gale, Paul H. Peterman, Environ. Sci. Tech. TBD.
    3   "Analysis of Coplanar PCBs," Axys Environmental Systems Ltd., Fax from Mary McFarland to
       Dale Rushneck dated November 25,  1994, available from the EPA Sample Control Center,
       operated by DynCorp I&ET, 300 N.  Lee St, Alexandria, VA 22314.
    4   Ahlborg, U.G., G.C. Becking, L.S. Birnbaum, A. Brouwer, H.J.G.M. Derks, M. Feeley, G. Golor,
       A. Hanberg, J.C. Larsen, A.K.D. Liem, S.H. Safe, C. Schlatter, F. Waern, M. Younes, and E.
       Yrjanheikki, "Toxic Equivalency Factors for Dioxin-like PCBs,"  Chemosphere 28:6, 1049-1067,
       1994.
    5   "Working with Carcinogens," Department of Health, Education, & Welfare, Public Health
       Service, Centers for Disease Control, NIOSH, Publication 77-206, August 1977, NTIS PB-
       277256.                         .
    6   "OSHA Safety and Health Standards, General Industry," OSHA 2206, 29  CFR 1910.
    7   "Safety in Academic Chemistry Laboratories," ACS Committee on Chemical Safety,  1979.
                                                                                   Draft, March 1997
  44

-------
                                                                                   Method 1668
 8   "Standard Methods for the Examination of Water and Wastewater,"  18th edition and later
     revisions, American Public Health Association, 1015 15th St, N.W., Washington, DC 20005, 1-
     35: Section 1090 (Safety), 1992.
 9   "Method 613—2,3,7,8-Tetrachlorodibenzo-/?-dioxin," 40 CFR 136 (49 FR 43234), October 26,
     1984, Section 4.1.
10   Lamparski, L.L., and Nestrick, T.J., "Novel Extraction Device for the Determination of
     Chlorinated Dibenzo-p-dioxins (PCDDs) and Dibenzofurans (PCDFs) in Matrices Containing
     Water," Chemosphere, 19:27-31,  1989.
11   Provost, L.P., and Elder, R.S.,  "Interpretation of Percent Recovery Data," American Laboratory,
     15:56-83, 1983.                         '._.-'_
12   "Standard Practice for Sampling Water," ASTM Annual Book of Standards, ASTM, 1916 Race
     Street, Philadelphia, PA 19103-1187, 1980.
13   "Methods 330.4 and 330.5 for  Total Residual Chlorine," USEPA, EMSL, Cincinnati, OH 45268,
     EPA 600/4-70-020, March 1979.
14   "Handbook of Analytical Quality Control in Water and Wastewater  Laboratories," USEPA
     EMSL, Cincinnati, OH 45268, EPA-600/4-79-019, March 1979.
15   "Analytical Procedures and Quality Assurance Plan for the Determination of PCDD/PCDF in
     Fish",  U.S. Environmental Protection Agency, Environmental Research Laboratory, Duluth  MN
     55804, EPA/600/3-90/022, March 1990.
16   Storr-Hansen, E. and T. Cederberg, "Determination of Coplanar Polychlorinated Biphenyl (CB)
     Congeners in Seal Tissues by Chromatography on Active Carbon, Dual-Column High Resolution
     GC/ECD and High Resolution GC/High Resolution MS" Chemosphere 24:9, 1181-1196, 1992.
17   Workman, S.M., TBD [or Peterman in Reference 2]
18   Tessari, J.D., Personal communication  with Dale Rushneck, available from the EPA Sample
     Control Center,  operated by DynCorp I&ET,  300 N. Lee St., Alexandria, VA 22314 (703-519-
     1140).
Draft, March 1997
                                                                                           45

-------
  Method 1668
23.0  Tables and Figures

  Table 1.  Toxic Polychlorinated Biphenyls Determined by Isotope Dilution and Internal Standard High
  Resolution Gas Chromatography (HRGC)/High Resolution Mass Spectrometry (HRMS)
PCS'
3,3',4,4'-TCB
2)3,31I4,41-PeCB
2,3,4,4',5-PeCB
2,3',4,4',5-PeCB
2',3,4,4',5-PeCB
3,3',4,4',5-PeCB
2,3,3',4,4',5-HxCB
213,31I4I4')51-HxCB
2,3'14,4',5I5I-HxCB
3,31,4,41,5,5'-HxCB
2,21,3,31)4>4')5-HpCB
212',3,4,4',5,51-HpCB
2,3,3l)4I4'I5,51-HpCB
Cleanup standards
13C12-3,4I4',5-TCB
13C12-2,3,3',5)5l-PeCB
Internal standards
"C^^'.S.S'-TCB
13C -2 2',4|4 5'-PeCB
"C^^SA^S'-HxCB
13C12-2,2',3,3l,5,5l,6-HpCB
Final eluter standard
Native compound
CAS Registry No.
32598-13-3
32598-14-4
74472-37-0
31508-00-6
65510-44-3
57465-28-8
38380-08-4
69782-90-7
52663-72-6
32774-16-6
35065-30-6
35065-29-3
39635-31-9









IUPAC
No.
77
105
114
118
123
126
156
157
167
169
170
180
189

81
111

52
101
138
178

13C12 analog
CAS Registry No.
160901-67-7
160901-70-2
160901-72-4
160901-73-5
160901-74-6
160901-75-7
160901-77-9
160901-78-0
161627-18-5
160901-79-1
160901-80-4
160901-82-6
. 160901-83-7

160901-68-8
160901-71-3

160901-66-6
160901-69-9
160901-76-8
160901-81-5 .

    13C12-DCB
209
160901-84-8
    Polychlorinated biphenyls:
      TCB     =   Tetrachlorobiphenyl
      PeCB    =   Pentachlorobiphenyl
      HxCB    =   Hexachlorobiphenyl
      HpCB    =   Heptachlorobiphenyl
      DCS     =   Decachlorobiphenyl
                                                                                        Draft, March 1997

-------
           O

           2
           UJ -iS
 CO

LJJ

•a


 TO
 CO


Q

S

UJ,


 co
 £Z
.O

"o
-a
 o

 co

 E

'•55
UJ
cc

DC.


 co
 CD

    ,_
o  5l\

"

                 111
           co
          a


          UJ
                 cc
                 cc
          t-  =

          °=  E
a:


.i
 co
EC

 en
 cu
 o


 I
.co

 a>
CC
 co
 S'8
 g^
 CD  O

    "
                  CO  O>


                  2  2
                 •££  co
                  a>  s
          O

          <  d

    _o

    ^—^
    CO
 CD
.

-9  "=
 CO  ;=
                 m
                 o
                 a.
          CO


          o

          •o
          to
          cu


          2





CD l^-
CD CM
CD in
f- in
co co
-

co m
o o

In In
iff iff
CM CM
c\T c\F
CM CM
O O
co co
"*" "*""



-j _i
fo c^1




















o 59
IT "
CM" "sf
CM CO
CM CM
O O
CO CO




CM i—
•





CD
CD
in
in

CD
CM

CQ
O
H-
In
uf
CM
CM"
CM
O
CO
"f—



CM
in



















rn

£
.co
cp
CM
O
CO




	 1

"-
CM
in
o
CD
CM
in
CM
CD
CD
CO

CD
CM

Q
K-
^3-
•^
co
erf
CM
O
CO




1 —






















m
_>
^J_
co
co"



i —






CD
CD
CD
CO
CO


CQ
O
In
iff
•*€
CM
CM"
-o ***
co O
-a co
CO
CO
"co
CO 1
.
o
••g
' *•".
.£=
CD
•^z
CO
CO
	 1

CD
=**=

CD
O
CD
O_
In
in" m
•* o

CH Q_
CM _'
52 CM
cn CM"
-— CM
= 0
CO CO

CD
1
§
CD 5






CO
CM

CM

0
CM

CQ
O
CU
In
iff
•sF
CM
CM"
CM
o
CO




5
















^
rr\
O

Q.
In
iff
CO
c\T
CM
0
^—



_j

in
CD
-
CD
CD
CD
•SJ-
t-~
CO
CO
co
en
CM

CQ
0
£
in_
^r
•^
CO
CM"
CM
O
CO




^





















ca
o

i^-
^^
S:
CM



CO
CM

X
S
CD
CD
CD
CM
§
35
CM CD
1^- CD
CM CO
CM CM
CM CM

CQ CO
a. a!
In "n.
iff ^
"^ ^C
CM CO
CM" CM
CM CM
o o
CO CO






















CQ.
o
CD
a.
^1 O

CM" 15^
CM -
g c>r
i— CM"



	 i
CO CO

i
CD
CD :.-.
CM
CD
CD
CD '
g
CM
CD
CD
•«3- co
en co

CO ^f"
CM CO
CM CM

CO CQ
<£<£
^3- In
•^ iff
co •*!
crT CM
CM" CM"
CM CM
O O
CO CO




in i—
CD CD

















CQ
O
CD
Q.
g S:
g. "I
Jfi CM"
^ CM
2-0
CM" i—



••3- in

in
CD
CD
*
o
CD
CD
CD
CD
•Si-
CD cn
CD CO
^ "t
co in
••a- in
CO CD
CM CM

m m
O O
-*r m
^t iff
co •*!
co" CM
CM" CM"
CM CM
££




in T—
CD CD

















m
O
CD
0_
CQ i°-
o •*
CD •<3-
Q_ _ -
_• CO
^ co"
^ CM
'%$
CM i—



in co
CD CM

in
-
CD
CD
CD
O
CD
CD
CO
in
CD
CM

m
o
in
^j-
•^
co
co"
CM
%




CO
CM





















m
o
CD
a.
^f-
^J_
co
co"



CO
CM

0 O O CD
O O O CD
CM CM CM CM
CO CO CD CD
CD CD CD CD
CD CD CD CD
CM CM CM CM
CD CD CD CD
CD CO CO CO
CM CM CD 1 — t — CD CD •<— CD
"t'~:c:?'^:'l~:c-!<-?cocD
in CD CM h- l~~ CD CD CO CM
co in in r*; i^> co co co "^
m co co' cn cn 03 o> co co
CMCMCMCMCMCMCMCOCO
, CQ CO CQ CD rn CQ CQ CQ
moofooSoocj
In ^T '"? V ^ ^ ^r V "C
iff •**•" ^J- ^F •<

LO
^T CQ CO • CQ CQ • . . CQ
^r Cj (^ (^ ^~£ ^~t
CO" "T* T" ^ T" T

CM Io LO rn LO to -— rn LO rn
gj ^ !°- S? ^ ^: o o h°- ^
co-rT^r^nS-S-'n^n^r10
_ J CM CO ^^ CO CO _ ~ _ ~ CO ,-r
jjj c\T c\T _ - c\T CM" ^ ^1 erf _ -
d CM CM - CM CM - ^«. CM -
™ O O ^" O O *°~ ^- O -^
CO ^__ Y^. ^yj ^«' ^«- CM CM ^^ CO
CD
C
CD
C?
1
CO 1 	 | 	 | 	 | 	 |
>
-------
CO
Ol

3'
CD

cp_
cz











































































^
O
CD
CD

C
-a


=3
a.
CO
o.



















r



















































Cd
CO
o
3
CD

m


CO
03
13
O,
m
S
i —
CO
I
CT
CD
CD
Cn
2-
£0
cr
0
g
5
CD
o"
S
CD
£2-
o
22.
5=
§.
o'
rs
•^j
o
Z3"
"^j

CD
^
cr
CD
£0
s
CD
cr
0
SKT
CO
0

o_
s.
Sr
cn
CD
~O

O*
i"
i
*j
CO
CD
CD

CO
CD
0.
M
o"
c:
ct
CD_
3'
Z3"
CO
S"
o
3
co~
cr
o
CO
o"
«<
CO
Z5
*<"

o
S.

en
CD"
en


10 •*
O3 CO
£2. S.
CD S1
13 X
=: =
° rr
— 3 —
f-*- S="
— • -J
CD O*
a. Si.
J13 CD
03 55"
-^ O*
 CD
•Ej* CD
° 0.
c?"o
t|
o
o
cz
3
=3
cn
o
CD
o.
5*
CO
cn
o
CD
CL
ca"
CD
CD"
Z3
5"
^

i""
CD
0
O-
~*
s.
3;
13
CO
«
^J
o
o
*§
3
CO
3
pz
*~*

































poocn— JOOCD^ICD
CO <0 CO 0 0 0 00-^


•=}
a.
o.
CD
— 1.JO— "-JOJO— i— >-CO
l».co_4.jojo_i._i.--,
Cj 5*" rSI02-^" K rb^
^t^^^'o^^^i™
01 "en ^- " ~ "en "oo CO co
i ~p^ ^ ^ -^ - - cn
•o "^ "D "D - O1 rs"
O Oi O O 01 01 ea
CO "01 CO CO '01 'a, — ^
T3 "O "O — »•
O O O 10
co co co ro
"ro
to
to
tji
Vl
03'
3C
T3
o
CO
^^.
=«=

-j oo ~j cs> oo —j o r^
OO CD CO O O CO — *•
r— r— r— r— r—r-r— g

CD

"CD"
OOOOOOO3"
rororororororo —
rororororororoS
"ioto'Kiioro'ro'ro™
to °5 to to co to.tti. to
co jC>. co ^b. jt>- co 5Ji ff,
tn -*i "01 -^i -^ "or PI S*
01 01 01 01 01 01 ~n 3

O3 O1 C33 ^^ ^2. O> ^-s CJ3
-7- — r" •T* nr T" T" CD ^~
-a -a -a -a -a -a
O O O O O O
CD CO CO CO CO CO








Sco co' co co co ro
cn o> ro -j- — ^ 01
.&. CO CO ~J CO CO -^1
oo cn ro 01 ro CD ao



CO O O -C*. O — *• «C»-
^J ^j. CO C73 — *• ^^ CJ3


§ 0 0



ro ro ro
O O d
O O C3



O3 O3 O3

ro . ro ro
o o o



o o o



Z

























p




•V
§-
m
3
^f

















•g-
^
ro


























_
[5
^>


"*



03
2.
CD
r^
o
-Tt
S-
CD
-o
o
CD










C
5
o^






— <

2i
o"
3




3d




30
30
—1

m
a ^
r~~ ^5* ^?

Q.
0
3
O
CD

3

o



m
CO
3"


CD
o_

;^
^»
3"
3"

1
r—
2
CD
CO
Tn
S
~J£-
o
« i
CD
^"
X
o"

c^
CD
CO
























































—1

CD"


»


O
§
-J-.
CD.
CD"
t—+_
id
CD
•KM-
00
2,
S'
CD
O
CD

£~)
r—
co
^3
sT
o"

33
CD^
CD"
f^
CD
v>
— _
CD
SI

CD

«m
CD
yi-
CD
3
o'
— |

3
CD
CO
"53
3
Jg
m
CO

i"
CD
a.
s:
22.
o
Q.
0
S-
CD
CT_
Q"

| —
3"


m
S
o
ET

o>
^3
a.




-------
                                                                                                                Method 1SS8
Table 3. Concentrations of Stock and Spiking Solutions Containing the Native PCBs and Labeled Compounds
PCB congener
3,3',4,4'-TCB
2,3,3',4,4'-PeCB
2,3,4,4',5-PeCB
2,3',4,4',5-PeCB
2',3,4,4',5-PeCB
3,3',4,4',5-PeCB
2,3,3',4,4',5-HxCB
2,3,31,4,41,5'-HxCB
2,3',4A',5,5'-HxCB
3,3',4,4',5,5'-HxCB
2,2',3,3',4,4',5-HpCB
2,2',3,4>4'>5,5'-HpCB
2,3,3',4,4I,5,5'-HpCB
1 3C1 2-3,3',4,4'-TCB
13C12-2!3,3'',4,4'-PeCB
13C12-2,3',4,4l,5-PeCB
1 3C1 2-3,3',4,4',5-PeCB
13C12-2!3,3\4,4',5-HxCB
13C12-2,3,3',4,4',5'-HxCB
1 3C1 2-2,3',4,4',5,5-HxCB
1 3C1 2-3i3'X4',5,5'-HxCB
1 3C1 2-2,2',3,4,4',5,5'-HpCB
1 3C1 2-2,3,3' ,4,4',5,5'-HpCB
13C12-DCB
Cleanup standards5
13C12-3,4,4',5-TCB
1 3C1 2-2,3,3' ,5,5'-PeCB
Internal standards6
13C12-2,2',5,5'-TCB
1 3C1 2-2I2',4,5,5'-PeCB
13C12-2,2',3,4,4',5'-HxCB
13C12-2,2',3,3',5,5',6-HPCB
IUPAC
No.
77
105
114
118
123
126
156
157
167
169
170
180
189-
77L
105L
118L
126L
156L
157L
167L
169L
180L
189L
209L

81L
111L

52L
101L
138L
178L
Labeled
Stock'
(ng/mL)
_
--
.
-
- . .
. -
-
- •
; • -
- •
'-
.'
-
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000

200
1000

1000
1000
1000
1000
compound
Spiking2
(ng/mL)
•
-
-
-
-
-
-
.
•- •

-
• •
- •
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
,2.0
2.0
4.0

1.0
' 5.0

- -
: -
-

Precision
Stock3
(ng/mL)
220
1000
1000
1000
1000
100
1000
1000
1000
200
200
1000
200
. -
•
-
'
-
-
.
-

• • • - • .
-

•
-

-
-.
. •

and Recovery
Spiking4
(ng/mL)
0.4
, . 20.0
20.0 •
20.0
20.0
2.0
20.0
20.0
20.0
4.0
4.0 .
20.0
4.0
. . •

.
•
- •
- .
-
-• -
,
' -
'

-
• -, - •'. .

• - ' .
• -
-

 1  Section 7.10-prepared in nonane and diluted to prepare spiking solution
 2  Section 7.10.3-prepared in acetone from stock solution daily
   Section 7.9-prepared in nonane and diluted to prepare spiking solution. Concentrations are adjusted for expected background levels.
   Section 7.14-prepared  in acetone from stock solution daily. Concentrations are adjusted for expected background levels.
    Section 7.11-prepared in nonane and added to extract prior to cleanup
    Section 7.12-prepared in nonane and added to concentrated extract prior to injection
           Draft, March 1997
                                                                                                                         49

-------
           Method 1668
Table 4. Concentrations of
Toxfc PCB congener
3,3',4,4'-TCB
2,3,3',4,41-PeCB
2,3,4,4',5-PeCB
2,3',4,4',5-PeCB
2',3,4,4',5-PeCB
3,3',4,4',5-PeCB
2,3,3',4,4',5-HxCB
2,313',4141,5l-HxCB
2131,4,4'I5,5'-HxCB
3,3'l4,4',5,5I-HxCB
2,2',3,3',4,4',5-HpCB
2,2t13,414',515'-HpCB
2,3,3'I4141,5,51-HpCB
Labeled congener
13C12-3,3',4,4'-TCB
13C12-2,3l3lI4I4'-PeCB
13C12-2I3l,4,4',5-PeCB
13C12-3,3l,4,4l,5-PeCB
13C12-2,3I3l,4,4l,5-HxCB
13C12-2,3,3',4,4l)5l-HxCB
13C12-3l3Il4I4')5I5'-HxCB
13C12-2,2ll3)4,4',5,5l-HpCB
13C12-2,3,3l,4,4ll5l5l-HpCB
13C12-DCB
Cleanup standards
13G12-3,4,4',5-TCB
13C12-2,3,3',5,5'-PeCB
Internal standards
13C12-2,2',5,51-TCB
1 3C1 2-2,2',4,5,5'-PeCB
ISCte-^'isX^S'-HxCB
13C12-2l2'l3)3lI5,5l,6-HpCB
PCBs in Calibration
IUPAC'
No.
77
105
114
118
123
126
156
157
167
169
170
180
189

77L
105L
118L
126L
156L
157L
169L
180L
189L
209L

81 L
111L

52L
101L
138L
178L
and Calibration Verification Solutions
CS1 CS2 CS32
(ng/mL) (ng/mL) (ng/mL)
0.5
2.5
2.5
2.5
2.5
2.5
5
5
5
5
5
5
5

100
100
100
100
100
100
100
100
100
200

0.5
2.5

100
100
100
100
2
10
10
10
.10
10
20
20
20
20
20 '
20
20

100
100
100
100
100
100
100
100
100
200

2
10

100
100
100
100
10
50
50
50
50
50
100
100
100
100
100
100
100

100
100
100
100
100
100
100
100
100
200

10
50

100
100
100
100
CS4
(ng/mL)
40
200
200
200
200
200
400
400
400
400
400
400
400

100
100
100
100
100
100
100
100
100
200

40
200

100
100
100
100
CSS
(ng/mL)
200
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000

100
100
100
100
100
100
100
100
100
200

200
1000

100
100
100
100
* Suffix "L" indicates labeled compound
2 Section 15.3, calibration verification solution
                                                                                                                    Draft, March 1997
50

-------
                                                                                                 Method fffffff
Table 5.  GC Retention Time Window Defining Solution and Congener Specificity Test Standard1'2
(Section 7.15)
Congener
Group
TCB
PeCB
HxCB
HpCB
First eluted3
54
104
155
188
2,2',6,6'
2,2,4,6,6'
2,2'A4',6,6'
2,21,3I41I5,616I
Last eluted
77
126
169
189 '
3,3,4,4'
3,3,4,4,5
3,3,4,4,5,5'
2,3,3,4,4,5,5'.
 SPB-Octvl resolution test compounds

 123      2',3,4,4',5-PeCB

 118      2,3',4,4',5-PeCB

 114      2,3,4,4',5-PeCB

DB-1 column resolution test compounds

 156 '     2,3,3',4,4',5-HxCB

 157      2,3,3',4,4',5'-HxCB
* All compounds are at a concentration of 100 ng/mL in nonane.       •
2 It is not necessary to monitor for the earliest eluted compounds if the toxic PCBs only are to be determined. If the co-planar
  PCBs (77,126,169)only are to be determined, it is necessary to resolve these-co-planar PCBs and potentially interfering
  compounds only; i.e., use of the compounds listed in this. Table is not required.
3 The earliest eluted compound in each congener group is provided for those instances in which all PCBs in that congener
  group are to be determined.  If the toxic PCBs only (Table 1) are to be determined, use of the first eluted compounds is not
  required.
Draft, March 1997
51

-------
Method 1668
Table 6. Preliminary Acceptance Criteria for Performance Tests When All Toxic PCBs are Tested*
IPR
Congener

3,3",4,4'-TCB
2&3',4,4'-PeCB
2,3,4,4',5-PeCB
2,3',4,4',5-PeCB
2',3,4,4',5-PeCB
3,3',4,4',5-PeCB
2,3,3',4,4',5-HxCB
2,3,3',4,41,5'-HxCB
2,3l,4,4I,5,5>-HxCB
3,3',4,41,5,5'-HxCB
2,2',3,3',4,4',5-HpCB
2,2'l3,4,4115,5'-HpCB
2,313',4,41,5,5'-HpCB
13C12-3,3',4,4'-TCB
13C12-2,3,3ll4,4'-PeCB
13C12-2,3',4,4l,5-PeCB
13C12-3,3',4,4',5-PeCB
13C12-2,3,3ll4,4',5-HxCB
13C12-2l3,3'l4,4l,5'-HxCB
13C12-3,3',4,4l,5,5'-HxCB
13C12-2,2',3,4,4l,5,5l-HpCB
13C12-2l3,3',4,4l,5,5l-HpCB
Cleanup standards
13C12-3,4,4')5-TCB
13C12-2,3l3',5,5'-PeCB
Table 6a. Preliminary Acceptance
IUPAC
No.
77
105
114
118
123
126
156
157
167
169
170
180
189
77L
105L
118L
126L
156L
157L
169L
180L
189L

81 L
111L
Criteria for
Test
(ng/mL)
20
1000
1000
1000
1000
100
1000
1000
1000
200
200
1000
.200
100
100
100
100
100
100
100
, 100
200

20
100
Performance
s*
(ng/mL)
5.6
172
390
172
390
17
222
222
222
45
33
165
33
37
39
39
39
34
43
43
41
82

7.2
36
Tests when
X"5
(ng/mL)
16-26
720-1500
160-2800
720-1500
160-2800
72-150
740-1600
740-1600
740-1600
148-320
152-260
760-1300
152-260
28-134
16-279
16-279
16-279
24-157
24-157
24-157
28-141
56-282

8-31
39-154
PCBs 77,
OPR
(ng/mL)
14-32
680-1600
130-3300
680-1600
130-3300
68-160
640-1700
640-1700
640-1700
128-340
140-280
700-1400
140-280
20-175
13-328
13-328
13-328
17-205
17-205
17-205
20-186
40-372

6-38
31-191
126, and 169
VER
(ng/mL)
16-26
780-1300
770-1300
780-1300
770-1300
78-130
780-1300
780-1300
780-1300
156-260
172-232
860-1160
172-232
71-140
77-130
77-130
77-130
70-143
70-143
70-143
72-138
144-276

15-26
79-127
only are Tested 1
IPR
Native PCB

S^'A^-TCB
3,3',4,4',5-PeCB
3,3l,4,4',5,5l-HxCB
13C12-3,3',4,4'-TCB
13C12-3,3',4,4',5-PeCB
13C12-3,3',4l4l,5,5'-HxCB
Cleanup standards
13C12-3)4,4',5-TCB
13C12-2,3,3l,5I5l-PeCB
IUPAC

77
126
169
77L
126L
169L

81 L
111L
Test
(ng/mL)
20
100
200
100
100
200

20
100
s*
(ng/mL)
5.4
17
45
34
17
45

7.2
36
x°
(ng/mL)
17-25
72-150
148-320
32-115
72-150
148-320

8-31
39-154
OPR
(ng/mL)
15-30
68-160
128-340
25-141
68-160
128-340

6-38
31-191
VER
(ng/mL)
16-25
78-130
156-260
76-131
78-130
156-260

15-26
79-127
f Preliminary criteria transferred from Method 1613. All criteria given as concentration in the final extract,
  assuming a 20-pL volume.
2 s=standard deviation
3 X=average concentration
                                                                                                      uran, Marcn

-------
                                                                                         Method 1668
Table 7. Labeled Compound
Labeled PCB


13C12-3,3',4,4'-TCB
13C12-2>3,3l,4l4l-PeCB
13C12-2,3',4,4',5-PeCB
1 3C1 2-3,3',4,4',5-PeCB
13C12-2!3,3'A4l,5-HxCB
13C12-2,3,3',4,4',5'-HxCB
1 3C1 2-2,3',4,4',5,5'-HxCB
'l3C12-3!3'X4l',5,5'-HxGB
13C12-2,2l,3f4f4l,5,5l^pCB
1 3C1 2-2,3,3',4,4',5,5'-HpCB
Cleanup standards
13C12-3,4,4',5-TCB
13C12-2,3,3',5,5'-PeCB
Recovery in Samples When All
IUPAC
No.

77
105
118
126
156
157
167
169
180
,189

81
111
PCBs are
Test
cone
(ng/mL)
100
100
100
100
100
100
100
100
100
100

20
100
Tested

Labeled compound

(ng/mL)
24-169
' 21-178
21-178
21-178
26-152
26-152
26-152
26-152
23-143
. 23-143

7-40
: 35-197
recovery
(%)
24-169
21-178
21-178
21-178
26-152
26-152
' 26-152
26-152
23-143
23-143

35-197
35-197
  Table 7a. Labeled Compound Recovery When PCBs 77,126, and 169 Only are Tested'
  Labeled PCB
IUPAC
 No.
 Test
 cone
(ng/mL)
Labeled compound Recovery

  (ng/mL)           (%)
  13C12-3,3',4,4'-TCB
  13C12-3)3',4,4l,5-PeCB
  13C12-3,3',4)4',5,5'-HxGB
  Cleanup standards
  13C12-3,4,4',5-TCB
  13C12-2,3,3',5,5'-PeCB
  77
  126
  169

  81
  111
  100
  100
  100

  20
  100
  29-140
  21-178
  26-152

   8-33
  .35-197
29-140
21-178
26-152

42-164
35-197
   Preliminary criteria transferred from Method 1613. Criteria are given as concentration in the final extract, assuming a
   20 uL volume.
Draft, March 1997
                                                                                                 53

-------
Method 1668
Table 8. Descriptors, Exact m/z's,
Descriptor Exact
mlz1
1. 289.9224
291.9194
292.9825
301.9626
303.9597
325.8804
327.8775
330.9792
2. 325.8804
327.8775
337.9207
339.9178
354.9892
359.8415
361.8385
371.8817
373.8788
393.8025
395.7996
405.8428
407.8398
3. 442.9728
509.7229
511.7199
513.7170
* Nuclidic masses used were:
m/z Types,
mlz
type
M
M+2
Lock
M
M+2
M+2
M+4
QC
M+2
M+4
M+2
M+4
Lock
M+2
M+4
M+2
M+4
M+2
M+4
M+2
M+4 '
Lock
M+4
M+6
M+8

and Elemental Compositions
Elemental composition

C12 H6 35CI4
C12 H6 35CI3 37CI4
C7 F11
13C12 H6 35CI4
13C12 H6 35CI3 37CI
C12 H5 35CI4 37CI
C12 H5 35CI3 37CI2
C7 F13
C12 H5 35CI4 37CI
C12 H5 35CI3 37CI2
13C12 H5 35CI4 37CI
13C12 H5 35CI3 37CI2
C9 F13
C12 H4 35CI5 37CI
C12 H4 35CI4 37CI2
13C12 H4 35CI5 37CI
13C12 H4 35CI4 37CI2
C12 H3 35CI6 37CI
C12 H3 35CI5 37CI2
13C12 H3 35CI6 37CI
13C12 H3 35CI5 37CI2
C10 F17
13C12 35CI10 37CI2
13C12 35CI9 37CI3
13C12 35CI8 37CI4

of 'the RGBs
Substance2

TCB
TCB
PFK
PeCB3
PeCB3
PeCB
PeCB
PFK
PeCB
PeCB
PeCB3
PeCB3
PFK
HxCB
HxCB
HxCB3
HxCB3
HpCB
HpCB
HpCB3
HpCB3
PFK
DCB3
DCB3
DCB3

H = 1.007825 0=12.00000
130 = 13.003355 35CI =
2 TCB = Tetrachlorobiphenyl
PeCB = Pentachlorobiphenyl
HxCB = Hexachlorobiphenyl
HpCB = Heptachlorobiphenyl
DOB = Decachlorobiphenyl
34.968853





37CI = 36.965903











   130 labeled compound
  54
Draft, March 1997

-------
                                                                                                 Method tffffff
 Table 9.  Theoretical Ion Abundance Ratios and QC Limits
Clorine
atoms
4
5
6
6
7
7
8
m/z's forming
ratio
M/(M+2)
(M+2)/(M+4)
M/(M+2)
(M+2)/(M+4)
M/(M+2)
(M+2)/(M+4)
(M+2)/(M+4)
Theoretical
ratio
0.77
1.55
0.51
1.24
0.44
1.05
0.89
QC Limit'
Lower
0.65
1.32
0.43
1.05
0.37
0.88
0.76
Upper
0.89
1.78
0.59
1.43
0.51
1.20
1.02
 * QC limits represent +/-15 % windows around the theoretical ion abundance ratio.  These limits are preliminary.
Draft, March 1997
                                                                                                         55

-------
Method 1668
 Table 10.  Suggested Sample Quantities to be Extracted for Various Matrices'

                                                                  Percent
  Sample matrix2             Example                           solids          Phase
                                                                     Quantity
                                                                    extracted
  Single-phase
  Aqueous


  Solid


  Organic


  Tissue


  Multi-phase
  Liquid/Solid
  Aqueous/Solid
  Organic/solid

  Liquid/Liquid
  Aqueous/organic
  Aqueous/organic/solid
Drinking water
Groundwater
Treated wastewater
Dry soil
Compost
Ash
Waste solvent
Waste oil
Organic polymer
Fish
Human adipose
Wet soil
Untreated effluent
Digested municipal sludge
Filter cake
Paper pulp
Industrial sludge
Oily waste


In-process effluent
Untreated effluent
Drum waste
Untreated effluent
Drum waste
 <1
 >20
    Solid


   Organic


   Organic
1000mL



  10 g



  10g



  10g
1-30
    Solid
  10g
1-100
 <1
  >1
     Both
   Organic
Organic & solid
  10 g
  10 g
  10 g
  1   The quantity of sample to be extracted is adjusted to provide 10 g of solids (dry weight). One liter of aqueous samples
      containing one percent solids will contain 10 grams of solids. For aqueous samples containing greater than one percent
      solids, a lesser volume is used so that 10 grams of solids (dry weight) will be extracted.
  2   The sample matrix may be amorphous for some samples. In general, when the PCBs are in contact with a multiphase
      system in which one of the phases is water, they will be preferentially dispersed in or adsorbed on the alternate phase
      because of their low solubility in water.
  3   Aqueous samples are filtered after spiking with the labeled compounds. The filtrate and the materials trapped on the
      filter are extracted separately, and the extracts are combined for cleanup and analysis.
                                                                                                Draft, March 1997

-------
                                                                                                 Method 1668


Determine % solids
§11.2




Determine particle size
§1.1.3



                                                                                            Back-extract
                                                                                             per §12.5
                                                                                             Transfer
                                                                                             through
                                                                                             Na2S04
                                                                                         Reconcentrate per
                                                                                         §12.6.1,  §12.6.21
                                                                                             or 12.6.3
   1  The K-D concentration procedure in
   • §12.6.3 can be used it the water bath
     is fed by a steam generator.
   Figure 1   Flow Chart for Analysis of Aqueous and Solid Samples
                                                                                          Microconcentrate
                                                                                              per §12.7
                                                                                          Extract cleanup
                                                                                            per §13.2-
                                                                                           §13.5, §13.7
                                                                                                 52-001-126
Draft, March 1997
                                                                                                         57

-------
Method 1668
                         Aqueous
           Discard
    The K-D concentration procedure in
    § 12,6.3 can be used if the water bath
    & fed by a steam generator
                                             Determine % solids
                                                 per §11.2
                                           Determine particle size
                                                per §11.3
                                            Pressure filter aliquot
                                                per §11.6.2
     I
                                                               Yes
                                                       No
                                              Reduce per §11.7
                                      Concentrate per §
                                         12.6-§12.7
    Clean up per
§13.2-§13.5, §13.7
                                      Reconceritrate per
                                        §12.6 -§12.7
Non-aqueous
                                                                         SDS extract per
                                                                            §12.3
                                                                          Transfer thru
                                                                            Na2S04
                                                                      Macro-concentrate per
                                                                       § 12.6.1-§12.6.21
                                                                     Back-extract per §12.5
                                     Transfer thru
                                       Na2S04
                                       Analysis
                                    per§14-§18
  Figure 2  Flow Chart for Analysis of Multi-Phase Samples
                                                                                                     52-001-127
  58
                                                          Draft, March 1997

-------
                                                                                               Method 1668
                                                     Homogenize tissue
                                                        per §11.8
                              Soxhlet Extraction
Mix with r>
2
^4
                                                       Remove 10g
                       Extract per §12.4
                          Decant thru,
                           Na2S04
                     Concentrate to dryness
                       per §12.6-§12.7
                             J_
                     Determine % lipids per
                           §12.4.9
             Column cleanup    |      Batch cleanup
                  •           j=fa-              |
      Remove lipids per
          §13.6
       Concentrate per
       §12.6-§12.7
           1.
     Redissolve and spike
        cleanup std
Redissolve and spike
    cleanup std
 Remove lipids per
      §13.6    ,
   Transfer thru
     Na2S04
                                               T
                                          Concentrate per
                                           §12.6-§12.7
                                                                                   Clean up per
                                                                               §13.2-§13.5, §1.3.7
                                                                                 Concentrate per
                                                                                  §12.6-§12.7
     Figure  3 Flow Chart for Analysis of Tissue Samples
                                                                              Analysis per §14-§'18
                                                                                                  52-001-128
Draft, March 1997
                                                                                                       59

-------
Method 1668
                                     1000
                                     750
                                     500
                                     250
                         r
                                                                        GMF150 Filter
                                                                           SPEDisk
                                                       Test Tube. 25-mm x 200-mtn
                                                        1 -Liter Suction Flask
  Figure 4         Solid-phase Extraction Apparatus
  60
Draft, March 1997

-------
                                                                                                  Method 1568
                                                                                                      £2-027-02
 Figure 5        Soxhlet/Dean-Stark Extractor
Draft, March 1997
                                                                                                            61

-------



ScoHeomcoor^Hvoo t— uico'
a
X) GO
3 ca
1
en
• • CO
CPU
*>J ID
«.H
I'll
•* ®.
V E1* fQ
8
JO H
w r^i
a
O f-l *^
CJKfcO
H O CO
2 8*10 0 "o 0
•4 H e>i o °> °°
fen M en H 	 •
X




,
_
~
-
•
— ^ 	

X — ^ 	 ^




	 	 ^-» ^






















O
o
CS| p| 	 _^ '
o
O
• ••
CM
to 1

eo T
C3 T? 	 	 	 i ' 	 	 -O
_O • CM 	 • — 3

CM

• o
o
— ••
^^
•CM



1
••-1

-O
O
^•*
»••
CM
CM

O
O
" **
o
•CM



-O
0
CO
•H
;
.0
o
~v> I **
'" ' i
• o
0
~v
•I-i


"
• o
o
CM

•

r-
f ^
OOOO'OOOO f^OOOOOOO'OOO<
• o
CD
if
• rH



•O
O
CM
•r<
•
•
•
.
3
CO rH
' 	 ' '
Figure 6 Congener-Specific Separation of Resolution Test Compounds on SPB-Octyl Column
62

-------
                         vo  vo  u>  vo  vo  on
               W  H  W  W  «  H  W  W
                  O  r-
         CM
04 fa to BJ '2
S^io 8 i «
O"»«H n f -S •
H 0 "» W h2
psi"s s "S s s 5 3 s s J gl's 5 ? s s
UCM
..,;..., 	 ,j
00 0 0 0
*» CO CM H
Figure 7       Congener-Specific Separation of PCBs 156 and 157 on DB-1 Column
                                            63

-------
Method 1668
24.0
 24.1
 24.2
 Glossary of Definitions and Purposes
 These definitions and purposes are specific to this method but have been conformed to common usage
 as much as possible.
 Units of weight and measure and their abbreviations

24.1.1     Symbols
 °C       degrees Celsius
 uL       microliter
 urn      micrometer
 <        less than
 >        greater than
 %       percent

24.1.2    Alphabetical abbreviations
 amp     ampere
 cm      centimeter
 9        gram
 h        hour
 ID       inside diameter
 in.       inch
 L        liter
 M       Molecular ion
 m       meter
 mg      milligram
 min      minute
 ml      milliliter
 mm      millimeter
 m/z      mass-to-charge ratio
 N       normal;  gram molecular weight of solute divided by hydrogen equivalent of solute, per liter of
          solution
 OD      outside diameter
 pg       picogram
 ppb      part-per-billion
 ppm     part-per-million
 ppq      part-per-quadrillion
 ppt      part-per-trillion
 psig     pounds-per-square inch gauge
 v/v      volume per unit volume
 w/v      weight per unit volume

 Definitions and acronyms (in alphabetical order).

 Analvte: A PCB tested for by this method. The analytes are listed in Table 1.

 Calibration standard (GAL): A solution prepared from a secondary standard and/or stock solutions and
 used to calibrate the response of the instrument with respect to analyte concentration.
     *  ''' i1                              ,      ,   '      i.           :    .       II „
    ' ,H' III1  !                       ,1                      .   ,'                   "'ill "
 Calibration verification standard (VER): The mid-point calibration standard (CSS) that is used in to verify
 calibration. See Table 4.
                                                                                           Draft, March 1997

-------
                                                                                             Method 16B8
         CS1. CS2. CS3. CS4. CSS: See Calibration standards and Table 4.

         DCB: Decachlorobiphenyl (PCB 209)          -..-'.•

         Field blank: An aliquot of reagent water or other reference matrix that is placed in a sample container in
         the laboratory or the field, and treated as a sample in all respects, including exposure to sampling site
         conditions, storage, preservation, and all analytical procedures. The purpose of the field blank is to
         determine if the field or sample transporting procedures and environments have contaminated the
         sample.                                                                •

         GJ^ Gas chromatograph or gas chromatography.

         GPC: Gel permeation chromatograph or gel permeation chromatography.

         HpCB: Heptachlorobiphenyl

         HPLC: High performance liquid chromatograph or high performance liquid chromatography.

         HRGC: High resolution GC.

         HRMS: High resolution MS.

         HxCB: Hexachlorobiphenyl

         I PR: initial precision and recovery; four aliquots of the diluted PAR standard analyzed to establish the
         ability to generate acceptable precision and accuracy. An IPR is performed prior to the first time this
         method is used and any time the method or instrumentation is modified.

         K-D: Kudema-Danish concentrator; a device used to concentrate the analytes in a solvent.

         Laboratory blank: See Method blank.

         Laboratory control sample (LCS): See Ongoing precision and recovery standard (OPR).

         Laboratory reagent blank: See Method blank.

          May: This action, activity, or procedural step is neither required nor prohibited.

          May not: This action, activity, or procedural step is prohibited.

          Method blank: An aliquot of reagent water that is treated exactly as a sample including exposure to all
          glassware, equipment, solvents, reagents, internal standards,  and surrogates that are used with
          samples. The method blank is used to determine if analytes or interferences are present in the
          laboratory environment, the reagents, or the apparatus.

          Minimum level  (MU: The level at which the entire analytical system must give a recognizable signal and
          acceptable calibration point for the analyte. It is equivalent to the concentration of the lowest calibration
          standard, assuming that all method-specified sample weights, volumes,  and cleanup procedures have
          been employed.

          MS: Mass spectrometer or mass spectrometry.                •


Draft, March 1997   '.               ~~               :   '                  :      '           -   :             §5

-------
Method 1668
          Must: This action, activity, or procedural step is required.

          OPR: Ongoing precision and recovery standard (OPR); a laboratory blank spiked with known quantities
          of analytes. The OPR is analyzed exactly like a sample. Its purpose is to assure that the results
          produced by the laboratory remain within the limits specified in this method for precision and recovery.

          PAR: Precision and recovery standard; secondary standard that is diluted and spiked to form the IPR
          and OPR.

          PFK: Perfluorokerosene; the mixture of compounds used to calibrate the exact m/z scale in the HRMS.

          Preparation blank: See Method blank.

          Primary dilution standard: A solution containing the specified analytes that is purchased or prepared
          from stock solutions and diluted as needed to prepare calibration solutions and other solutions.

          Quality control check sample (PCS): A sample containing all or a subset of the analytes at known
          concentrations. The QCS is obtained from a source external to the laboratory or is prepared from a
          source of standards different from the source of calibration standards. It is used to check laboratory
          performance with test materials prepared external to the normal preparation process.

          PeCB: Pentachlorobiphenyl

          PCB: Polychlorinated biphenyl

          Reagent water water demonstrated to be free from the analytes of interest and potentially interfering
          substances at the method detection limit for the analyte.

          Relative standard deviation (RSD): The standard  deviation times 100 divided by the  mean. Also termed
          "coefficient of variation."                                   •

          RR Response factor. See Section 10.6.1.

          RR: Relative response. See Section 10.5.2.

          RSD: See Relative standard deviation.

          SDS: Soxhlet/Dean-Stark extractor;  an extraction  device applied to the extraction of solid and semi-solid
          materials (Reference 12 and Figure 5).

          Should: This action, activity, or procedural step is suggested but not required.

          SICP: Selected ion current profile; the line described by the signal at an exact m/z.

          SPE: Solid-phase extraction; an extraction technique in  which an analyte is extracted from an aqueous
          sample by passage over or through a material capable of reversibly adsorbing the analyte. Also termed
          liquid-solid extraction.

          Specificity: The ability to measure an analyte of interest in the presence of interferences and other
          analytes of interest encountered in a sample.
 66
Draft, March 1997

-------
                                                                                                 Method rffffff
          Stock solution: A solution containing an analyte that is prepared using a reference material traceable to
          EPA, the National Institute of Science and Technology (NIST), or a source that will attest to the purity
          and authenticity of the reference material.

          TCB: Tetrachlorobiphenyl.

          VER:  See Calibration verification standard.                       .                 •
Draft, March 1997
                                                                                                          67

-------

-------