United States
Environmental Protection Agency
Office of Water
Washington, D.C. 20460
EPA 503/6-90-002
May 1986
BIOACCUMULATION
MONITORING GUIDANCE:
4. ANALYTICAL METHODS FOR U.S. EPA
PRIORITY POLLUTANTS AND 301 (h)
PESTICIDES IN TISSUES FROM
ESTUARINE AND MARINE ORGANISMS
     Cl
     ci
                              CCI3
       ill
        JW

-------

-------
EPA 503'6-9OOC2
May 1966
BIOACCUMULATION
MONITORING GUIDANCE:

ANALYTICAL METHODS FOR U.S. EPA
PRIORITY POLLUTANTS AND 301 (h)
PESTICIDES IN TISSUE FROM
ESTUARINE AND MARINE ORGANISMS
Prepared by:
Tetra Tech, Inc.
11820 Northup Way, Suite 100
Beltevue, Washington 98005
Prepared for:
Marine Operations Division: 301 (h) Program
Office of Marine and Estuarine Protection
U.S. Environmental Protection Agency
401 M Street SW
Washington, D.C. 20460

-------
     This report  is  one element cf the  Bioaccumulation Monitoring Guidance
Series.   Tne  purpose of this series is to  provide guidance  for monitoring
of priority  pollutant residues in  tissues  of resident marine organisms.
Trsese guidance  documents  were  prepared  for  the sewage  discharge program
of Section 301(h)  of the Clean Water Act  unoer  the U.S. EPA Office of  Marine
end Estuarine Protection, Marine Operations  Division.  Two kinds of monitoring
guicance  are  provided  in  this series;  recommendations  for  sampling  ana
analysis assigns,  and aids for interpretation of monitoring c'ata,

     Some oasic  assumptions were made in developing  tne guidance presentee
in these documents:  1)  each bioaccumulation  monitoring  program will be
designed  to  meet the requirements  of the  30i(h)  regulations, 2)  tissue
samples i-.il! be collected from appropriate  locations near the sewage discharge
and  from  en  unpolluted reference site,  3) the  initial chemicals of concern
are the  U.S.  EPA  priority  pollutants ana 301{h) pesticides, and 4)  the
monitoring cata  should be suitable for a  meaningful evaluation of the potential
hazards  to living marine  resources  as  well  as human health.  It should
be recognized  that the design of a monitoring  program reflects tne  site-
specific characteristics of the pollutant discharge  and  the receiving environ-
ment.   Thus,  site-specific considerations may lead to a moaification of
the generic recarmendations  herein.  Finally, although these guidance documents
were  prepared  specifically for monitoring of sewage discharges under  the
301(h) program,  their potential use extends  to assessment  and monitoring
cf bioaccunul ation resulting from otner  kinds  of pollutant discharges into
marine and estuarine environments.
                                     11

-------
                                    I ' ' — ' ' v
                                  L, J i \ . _ . 1 I J




        PREFACE                                                             ii

        LIST  OF  FIGURES

        LIST  OF  TABLES                                                      vi

        'ir*|/M-'^t'ir''-\p*"i**~»i"f'c                                                   • • ^ 1
        H'-IVlUrt utl JbCi'iwFi t b                                                   ' ' I

        INTRODUCTION                                                        ix

SECTION I.   ANALYSIS OF  EXTRACTABLE  ORGANIC COMPOUNDS
             IN  ESTUARINE AND  MARINE  TISSUES

 1.0    SCO^E AND APPLICATION                                              1-1

 2.*0    SUMMARY  OF METHOD                                                 1-4

 3.0    INTERFERENCES                                                      1-5

 4.0    SAFETY                                                             1-6

 5.0    APPARATUS AND EQUIPMENT                                            1-6

 6.0    REAGENTS AND CONSUMABLE  MATERIALS                               1-10

 7.0    SAMPLE COLLECTION,  PREPARATION,  AND  STORAGE                     1-14

 8.0    CALIBRATION AND STANDARDIZATION                                  1-16

 9.0    QUALITY  ASSURANCE/QUALITY  CONTROL                               1-21

10.0    PROCEDURE                                                        1-24

11.0    QUANTITATIVE DETERMINATION (CALCULATIONS)                        1-35

12.0    PRECISION AND ACCURACY                                           1-42

13.0    REFERENCES                                                       1-43


SECTION II.  ANALYSIS OF VOLATILE OR6ANIC COMPOUNDS
             IN ESTUARINE AND MARINE  TISSUES

 1.0    SCOPE AND APPLICATION                                            11-1

 2.0   SUMMARY OF METHOD                                               11-2

                                     iii

-------
 5.0   AP3AF,ATL3 A!,3 £3^I?/!E,'iT

 O , 0   ~~r-.G;.i»>5  "ti'J L ->. iS U;'l--.u L.t i'l*- i trt .. .-M_5

 7.0   SAI-.PLE  C;LLE:TI::;, PREPARATION,  ;;;D  STORAGE

 s.c   CALIBRATION AND STANDARDIZATION

 or,   ,•>• i • i • ^ v r "' "* "5""'                                                   ""_".".
 j , J   J J r,:_» i I L ^. i i '. _ *.                                                  - -  —•

10. C   PROCEDURE                                                        ":-2C

11.0   QUANTITATIVE DETERMINATION (CALCULATIONS)                       11-26

1 "> Pi   n->^.-T-i«(i,nii'i«^/>!i'Kf-v                                           TT-?7
i 4. . U   rr.l^^iu.i n!>J nowUriiOT                                           - *  ^'
SECTION III.  ANALYSIS OF METALS AND METALLOIDS
               IN  ESTUARINE AND MARINE  TISSUES

 1.0   SCOPE  AND  APPLICATION                                            III-i

 2.0   SUMMARY  OF METHOD                                                HI-1

 3.0   DEFINITION'S                                                      ^11-^
 4.0    INTERFERENCES                                                    iii-ii

 5.0    SAFETY                                                            III-4

 6.0    APPARATUS  AND EQUIPMENT                                          III-4

 7.0    REAGENTS AND CONSUMABLE MATERIALS                                III-6

 8.0    SAMPLE  COLLECTION, PREPARATION,  AND STORAGE                     III-7

 9.0    CALIBRATION AND STANDARDIZATION                                  III-9

10.0    QUALITY CONTROL                                                 III-11

11.0    PROCEDURE                                                        II1-13

12.0    CALCULATIONS                                                    III-21

13.0    PRECISION  AND ACCURACY                                          111-21

14.0    REFERENCES                                                      111-21
                                       iv

-------
                                  FIGURES


_N umber                                                                  Page

  1-1   Relative response calibration curve                             1-45

  1-2   Extracted ion current profiles for chrornatographically
        resolved labeled (^2/z) anc* unlabeled  (mi/z) pairs              1-45

  1-3   Extracted ion current profiles for (3A)  unlabeled compound,
        (3B) labeled compound, and  (3C) equal  mixture of  unlaoeled
        and labeled compounds                                           1-45

  1-4   Flow chart for sample preparation                               1-46

 II-l   Apparatus for vacuum distillation and  cryogenic concen-
        tration                                                        II-29

 II-2   Relative response calibration curve                            11-30

 II-3   Extracted ion current profiles for (A) the  unlabeled pol-
        lutant, (B)  the laPeled analog, and  (C)  a mixture of the
        labeled and unlabeled compounds                                11-30

III-l   Quality control chart                                         111-24

-------
                                      C C
                                      _t J
  1-1    Gas chromatography of extractable compounds                    i-47



  1-2    OFTPP nass- intensi ty specifications                            1-50



  1-3    Precision and accuracy of method blanks                        1-51



 II-l    Volatile organic analytes                                     11-31



 II-2    BF3 mass-intensity specification                              11-32



 II-3    Percent spike recoveries for volatile priority pollutants

        using vacuum distillation                                     11-33



Ill-i    General information for each priority pol 1 utant metal         111-25



III-2    Typical data obtained on a certified reference material

        (N3S oyster tissue)                                           111-26
                                     v1

-------
     Tnis document has been  reviewed  by  the 301 (h) Task Force of tre Environ-
mental Protection Agency, which  includes  representatives  from the Water
Management Divisions of U.S. EPA  Regions  I,  II,  HI, IV, IX, and X;  tne
Office of Research and Development - Environmental  Research Laboratory  -
Narragansett (located in  Narragansett,  RI and Newport, OR), and the Marine
Operations Division in the  Office  of  Marine and Estuarine Protection, Office
of Water.

     This technical guidance document was  produced for the U.S.  Environmental
Protection Agency under the  301(h) post-decision  technical support  contract
No.- 63-01-6938,  Allison J.  Duryee, Project  Officer.  This report was prepared
by Tetra Tech, Inc., under  the direction  of Dr. Thcnas C. Ginn.

c r " T »m'  T
O _ o I 1 • J . 1  i

     The primary authors were Mr.  Robert  C. Barrick and Mr. Harry R. Seller.
The assistance of Mr.  Raleigh C. Farlow  is  appreciated.

     Existing  U.S. EPA  analytical methods were  incorporated into Section  I
whenever possible.  Specifically, many  sections were adapted from  the  Contract
Laboratory Program for Qrganics Analysis  (Section I,  reference 2) and U.S.  EPA
Method 1625 Revision B (Section  I,  reference  3),  which was developed by
the  Industrial  Technology  Division of the Office of Water Regulation  and
Standards.

     Validation data  presented  in Section  I  (Precision and Accuracy) were
generated by  California Analytical Laboratories and Weyerhaeuser Technology
Center .
                                    vn

-------
errT'H" T T
-JwU 1 lUH 1 1

     Tne primary  c'jtr.crs .-.ere Xr.  ROD en C.  Barri>, enc !!r. -erry R.
The assistance  cf Mr.  Raleigh C. Farlow is appreciated.

     The procedure  described  in  Section II  is  largely a compilation  cf
methods cevelooec bv U.S. EPA.  Specifically, tne rr.etnods  were ceve'cc-c
by tr.e  Environmental Monitoring Systems  Laboratory (EMSL)  in  Las  Vecas
{Section II,  references  1  and 2)  and  the Industrial Technology  Division
of the  Office  of Water Regulation anc Stancaras  (Section II, refc-*-enc~ 2''.
Dr. M.  Hiatt  (Analytical Technologies,  Inc.,  national  City,  CA, previously
at EMSL Las  Vegas) was a valuable source of  technical information  presenter
in tnis document.

SECTION III

     The primary authors  were Mr.  Robert  C.  Barries, Mr. Harry R.  Seller,
and Mr.  Robert  V.  Several 1.   The assistance of Dr. Charles  R.  Lytle  is
apprec iatec.
     Validation cata  presented  in  Section III  (Precision ana Accuracy)
were generated by Analytical Service  Laboratories, Ltd.

     Mention  of trace names or commercial  products nerein coes not constitjte
endorsement  for use 3y U.S. EPA or Tetra  Tech,  Inc.
                                   vi n

-------
     The tnree  analytical metnocs in this  document  have been designed to
be consistent  with  probable uses of 501 (h)  bioaccumulation monitoring  cats.
Comparison  of  tissue  c en tarn in ant concentrations from contaminated  and relative'_,•
uncontarr.inatec areas  ana  estimation  cf  the potential  health effects  cf
b i OcCC'jnul £tc-G  Substances often  require  sensitive  analytical tecr.nic^es
for a wide  range of chemically diverse pollutants.  The  recommenced 301(h)
procedures  allow for  sensitive analyses  of the target  compounds *itr; a
reasonacle  amount of  laboratory effort.  Organopnospnate  201(n) pesticices
have net yet  been tested  with the recommended  techniques (i.e., Malathicn,
Paratnicn,  jernetor, Guth ion).   Analyses for  2 , 5 ,7,3-TCuD rtith appropriate
detection  limits  will require the dedicated  U.S.  EPA  Contract Laboratory
Program procedure for dioxin analysis  (9/15/33), which  involves selected
ion monitoring (SIX) GC/MS  analysis.

     Tnere  are currently no formally approved  U.S. EPA procedures  for analyzing
priority pollutants and 301(h) pesticides  in biological  tissue.  However,
various  U.S.  EPA procedures were reviewed during  development of tnis report
fe.g.,  Interim Methods  for  the Sampling and  Analysis  of Priority Pollutants
in Sediments  and  Fish Tissue (1977, revised  1980); Contract Laboratory
Program procedures  for  organics analysis  and inorganics  analysis].  Conse-
quently, the  recommended  301 (h)  procedures include portions of U.S.  EPA
analytical  and quality  assurance procedures  that were considered appropriate
for  sensitive,  full-scan analyses  (e.g.,  at the  low  parts per billion  level
for organics analysis).  The 301(h) methods  have  been  assembled accorcirg
to gjidei ines for EMSL  (Environmental  Monitoring  and  Support Laboratory,
Cincinnati)  analytical  methods (as specified in  EPA-600/8-83-020).
                                    ix

-------
                SECTION  I
ANALYSIS OF EXTRACTA3LE ORGANIC COMPOUNDS
                  AND MARINE TISSUES

-------
                                 CONTENTS
 1.0   SCOPE AND APPLICATION                                            1-1


 2.0   SUMMARY OF METHOD                                                1-4


 3.0   INTERFERENCES                                                    1-5


 4.0   SAFETY                                                           1-6


 5.0   APPARATUS AND EQUIPMENT                                           1-6


 6.0   REAGENTS AND CONSUMABLE  MATERIALS                                1-10


 7.0   SAMPLE COLLECTION,  PREPARATION,  AND  STORAGE                      1-14


 8.0   CALIBRATION AND STANDARDIZATION                                  1-16
  *

 9.0   QUALITY ASSURANCE/QUALITY CONTROL                                1-21


10.0   PROCEDURE                                                       1-24


11.0   QUANTITATIVE DETERMINATION (CALCULATIONS)                        1-35


12.0   PRECISION AND ACCURACY                                          1-42


13.0   REFERENCES                                                      1-43

-------
                 ANALYSIS  OF EXTRACTA3LE ORGANIC COMPOUNDS
                      IN ESTUARINE AND MARINE TISSUES
1.0  SCOPE AND APPLICATION

1.1  This method  is  designed to determine the semivolatile priority pollutants
(Tab'e 1-1)  associated with the Clean Water Act Section 301(h)  regulation
[40 CFR  125.5S(k)  and  (v)].   Additional  compounds  anenaole  to extraction
and analysis  by capillary column gas chrcmatography-mass  spectrometry (GC/KS)
and/or gas chromatography-electron capture detection  (GC/ECD) may be suitable
for analysis,  subject  to testing.

     These  procedures are applicaole when  low  part  per  billion analyses
are required to monitor differences between oody burdens in organisms  from
relatively  uncontaminated  reference  areas and  contaminated  estuarine and
marine environments.   In addition, the procedures  are applicable when low
detection limits  are required  for the estimation of potential health effects
of bioeccxiulated substances.

     Two  GC/MS options included  in the methods are  analyses  by  isotopa
dilution GC/MS  (preferred) or by a GC/KS internal standard  technique (minimum
required).   In both  cases, the laboratory procedures for  sample extraction
and concentration of the resulting extract are identical.   Compound-specific
recovery corrections  used  in the  isotope dilution  technique are designed
to increase the accuracy of the analysis and the  comparability  of  results
among  laboratories.   In  addition,  use  of the multiple  recovery stencrds
in each analysis  increases confidence in the validity of detection  limits
reported for  undetected  target compounds.  By  forcing  a  search for  every
recovery standard  in the sample  extract (over 50 are available),  the technique
also increases the efficiency  of detection and reporting frequency of target
compounds that otherwise may be overlooked in complex  extracts.
                                   1-1

-------
 1.2  Tne cornpounas listed  in Table I-i  incline pesticides suoject to  regulation
 under Section 301 (h;  of the Clean water  Act.   However, the  epplic=Dility
 of  this Tiethcd to non-chlorinated  organopnosphorous pesticides  (Malatrnor,,
 Parathion, Demeton, and Guthion) has not  been  demonstrated.   Chemists  at
 the  Food and Drug Administration  have  recently published  a technique for
 determining organophosphate pesticides  of  wide ranging polarity in  matrices
 including fatty  animal tissue (J.J.  Blaha and P.J. Jackson,  J.  Assoc. Anal.
 Chem., Vol. 68,  pp. 1095-1099, 1985).   The technique involves 1iquid-1icuid
 partitioning and gel permeation chromatography [both are included in this
 recomenced 301(h)  procedure] as well  as  N/P  alkali thermionic  or  flame
 photometric detection.  Further work  is required to determine the  suitability
 of the recommended 301(h)   procedure  for organophosphorous pesticides.

 1.3  The  detection  limit  of  this method  is usually dependent on  the level
 of interferences  rather than instrumental  limitations.  The  limits  listed
 in Table  1-1 represent  the  minimum quantity that can be detected with no
 interferences present.

     Lower  limits of detection  (LLD)  are established  by  analysts basea
 on their experience  with  the instrumentation  and with  interferences  in
 the  sample matrix being  analyzed.   LLD are greater than the instrumental
 detection limits  in Table 1-1  because they take into account  sample  inter-
 ferences.   To estimate LLD,  the noise   level should  be determined in the
 retention window  for the  quantitation mass of  representative  analytes.
 These determinations should  be made  for at least  three field  samples in
 the sample set under  analysis.   The signal  required to exceed  the average
 noise level  by at least  a factor  of  two should  then be estimated.  This
 signal  is the minimum response required  to  identify a  potential  signal
 for  quantification.   The  LLD is the  concentration  corresponding to tr,e
 level of this signal  based on  calibrated response  factors.   Based on  best
 professional  judgment, this  LLD would   then be  applied  to samples in the
 set with comparable or lower interference.  Samples with much higher  inter-
 ferences  (e.g.,  at  least  a  factor  of two higher)  should  be assigned LLD
at a multiple of the  original  LLO.
                                  1-2

-------
     These LLO values may be less than the  rigorously defined  metricc
Tin-its specified  in  the  revised  "Guidelines  EstaDlishing  Test Procedures
for  the  Analysis cf Pollutants" (£0 CFR  Part  136, 10/26/84).  This latter
proceaure requires  tr.e analysis of seven replicate  samples anc a statistical
determination of the method  detection limit  with 99 percent confidence.
Data quantified between  the LLD  and  the rigorous method  detection  limit
are valid and  useful  in  environmental  investigations of  Tow-level contamination,
but have a lower statistical  confidence  associated  witn  them than data
quantified aDove  the method detection limit.

     LLD  for  the described method  on  a wet weight basis are 10 ug/kg  for
aromatic hydrocaroons and pnthaTates (GC/MS  analysis)   and  10-20 ug/kg  for
chlorinated  hydrocarbons and halogenated ethers  (GC/MS  analysis).  Detection
limits for GC/MS analyses  of  pesticides  are 50  ug/kg  (wet weignt);  the
GC/ECD  detection limits for pesticides  are 0.1-5 ug/kg.  A method cetect^on
limit of 20  ug/kg is attainable for GC/ECD  analysis of  total  PCBs.

1.4  The GC/MS portions  of this method are  for use  only  by analysts experienced
with  GC/MS or  under  the close supervision of such  qualified persons.   Labora-
tories unfamiliar with the analyses of environmental samples by GC/MS should
run the performance  tests in reference 1 before  beginning.

1.5   This procedure has been designed to  analyze for a Targe number of
organic compounds with wide-ranging chemical  properties  (e.g., polarity,
molecular weight)  while minimizing procedural  complexity.  The accuracy
and sensitivity that could be attained by conducting a dedicated analysis
for  only one  of the compounds  in Table  1-1  cannot  be attained in such a
comprehensive  analysis.

1.6   Several  analytes  are particularly susceptible to decomposition during
analysis.  Benzidine can  be subject  to oxidative Tosses during  solvent
extraction.   Hexachl orocyclopentadiene is  subject to  thermal decomposition
in the  inlet  of the  gas chromatograph, chemical reaction in acetone solution,
and  photochemical  decomposition.  The polycyclic aromatic  hydrocarbons
are also photosensitive, especially benzo(a)pyrene.
                                  1-3

-------
2.C  SUM'-^RY OF METHOD

2.1  A  maceratcG  hcrrocemzeG 20 to 25  g  tissue  sar.p'e  is Scxnlet extrict = c
end the extract is dried over sodium sjlfate.   Biological mecromolecules are
removed from  the  extract by  gel  permeation  chromatogrepny (GPC) v,-ith 3"'o
Beads S-X3 (or equivalent)  (reference 2).  A portion  of the  extract (20 percent;
is  subjected  to  alumina cnromatogrephy  to  separate  polar conpojncs fror,
pesticides end PCBs prior to  capillary column GC/ECD analysis (reference 2).
The remaining SO  percer.t of  the  extract  is  analyzes  for acid, base, ana
neutral  compounds by capillary column GC/MS.  An  isotope  cilution  technique
(EPA Metnod 1625  Revision  B, reference 3)  is  recommended for all compounds
analyzed by GC/MS.   This technique  involves  spiking tne  homogenized  tissue
sample  with the  stable isotope labeled  analogs of most of the  pollutants
to be  en^yzec by GC/f-'S.   The advantage  of  isotope cilution is that re" iaDie-
recovery  corrections can be  made for each analyte with a labeled analog
or a chemically similar analog.

     2.1.1  Mjch of the text  of EPA Met hoc 1625  Revision 3 has  been  incorporated
into this method in mocified  forn.  The modifications were necessary because,
in relation to Methoc  1625  Revision B, the  present method involves different
sample matrices (biological  tissues), different calibration  requirements,
and cGditicnal  analytes (pesticides and  PCBs, both requiring  GC/ECD analysis].

2.2  Identification cf compounds is performed by  comparing the GC  retention
times and background-corrected characteristic spectral masses with those
of authentic  standards.   Tentative identifications  of low  levels of  pesticides
and PCEs ere maae by comparing GC retention  times to  standards.  The identities
cf pesticides end PCBs ere  confirmed by  GC/ECD  analysis  on  an  alternative
column phase or by GC/MS when sufficient concentrations occur.

2.3  Quantitative analysis  is performed  by GC/MS  using extracted  ion current
profile  (EICP) areas.  Isotope dilution, with  labeled analogs of pollutants
acting  as  recovery standards, is the method  of  quantification  when  labeled
compounds are available.   When the isotope dilution technique  is usea anc
                                   1-4

-------
r'v^i  C'~  « N^ C* ' i ~ •  • C vv  t, /  d'l  I •  « " T   j   ^ v ^> • ^ u • xJ I VT .. I • X^ 'u *   -rf i^' M s, ^ . . i» [ ^ „ . !„,' i ^-
C'Onpcuncs  Gjantifiea  by  GC/MS fi^e  >"9porteG after correcting  for  metnoa
recoveries wricn  t°.e isotope ai'jfion tecr.nicue  is  usea.
3.1   Solvents, reagents,  glcss^are,  ~rz  ctr.er  sample processing r.arc-.-.ere
nay yield a^ti^acts  anc/Dr  elevctec  case" ir,es, causing  -i s intercretat ior,
of cnrcncto:'"c~5 anc  suectrj.  All materials snou^c ?e  cer'icrstratec  tc
oe free from interferences  unoer tr.e conditions  of  thie  analysis c-y running
netn^G  bla^KS initially  cnc with  each sample lot (Sect. 9.4).   Specific
selection of reagents  and purification of solvents  by distillation in  all-
glass  systems a^e  recuired.  Hic^-purity ,  distil led- in.-gl ess solvents are
corviercTally available  (e.g., Bjrdick end Jackson  Laboratories,  Muskegon,
MI).   An  effective way  of cleaning laboratory  glassware  is covering  with
cluniT'jm  foil, r.eating at 450C C fcr severs;  hours, and rinsing  wit P. polar
ana ncr-cclar solvents  oefcre use.   Note tnat r.eatin; '.'.itno^t sjssecuent
sc'ven*  ri.-sir.g may not  eliminate laboratory  resiGu-es of  PCBs ana certain
otner crlorinateo  rydrccarbons.

3.2   Phtralates are common  laboratory contaminants that are usea  widely
as plasticizers.   Phthalates can derive from plastic  labware, plastic tubing,
plastic  cloves, plastic  coated glassware  clamps, and have been found  as
a contaminant in  Na2SQ4.  Polytetraf1uoroethylene  (PTFE) can be usea instead
of polypropylene or polyethylene  to  minimize tnis  potential  source  of
contamination.  However,  use of PTFE labware will  not necessarily precluce
all phthalate contamination.

3.3  Interferences coextracted from tissue  samples  limit the method detection
and quantitation limits.  For this reason, sample  extract cleanup is  necessary
to yield  reoroducible and reliable analyses  of  low  level contaminants  in
tissue samples.
                                  1-5

-------
4.0  SAFETY

4.1  The  toxicity or care inogen ic i ty of  each compound or reagent  used  in
this method has not  been  precisely determined.   However,  each chemical
compound  should be treated as a potential  health hazard.  Exposure  to  these
compounds  should be reduced to the lowest  possible level.   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 data  handling sheets  should also be made available
to all  personnel  involved  in  these  analyses.   Additional  information  on
laboratory safety  can be found in references  4-6.

4.2  The following compounds covered by this method have been  tentatively
classified  as  known or suspected human or  mammalian carcinogens:  benzene,
benzo(a) anthracene,  3,3'-dichlorobenzidine , benzo( a) pyrene , d ibenzo(a ,h) anthra-
cene,  N-nitrosod imethyl anrine,  4,4'-DDT,  alpha-, beta-, delta-,  and gamma-
hexachlorocycl ohexane,  and PCBs.  Primary standards  of  these  compounds
should  be  prepared in a hood,  and a NIOSH/MESA  approved  toxic gas  respirator
should  be  worn when high concentrations are  handled.

5.0  APPARATUS AND EQUIPMENT

5.1  Soxhlet  Extractor -  50  ml extractor  (Corning 3740-S, or equivalent),
with 250 ml flask (Corning 4320-250,  or  equivalent)  and  condenser with
34/45  joint.   Cellulose thimbles of the  appropriate size should  be  cleaned
with the extraction solvent mixture for at  least 30 cycles.

5.2  Drying  Column - 30 cm x 2 on borosilicate glass chromatography column
with glass wool  plug.  Glass wool is extracted  with the appropriate  solvents
and allowed  to dry before use.
                                  1-6

-------
5.3  Kutierna-Danish  (K-D)  Apparatus  -

     5.3.1  Concentrator  tuDe - 10 ml,  graduated  (Kontes K-570050-1G25,
or equivalent.   A ground glass  stopper (19/22  joint) is  used  to prevent
evaporation of extracts.

     5.3.2  Evaporation  flask  -  500  ml  (Kontes K-570050-0500, or  equivalent),
attached to concentrator  tube  with springs (Kontes K-662750-0012).

     5.3.3  Snyder  column  -  three ball  macro  (Kontes  K-503000-0232,  or
equivalent).

     5.3.4  Snyder column  -  two  ball micro (Kontes K-469002-0219,  or  equiva-
lent).

     5.3.5  Silicon  carbide  boiling  chips  - approximately 10/40 mesh, extracted
with methylene chloride  and heated  at 4500 C for 1 h minimum.   Note  that
boiling chips  can be a  significant source  of contamination  if  not  properly
cleaned.

5.4  Separatory Funnel  -  500 ml,  borosilicate glass with PTFE stopcock.

5.5  Borosilicate Glass  Beaker -  400 ml and 100 ml.

5.6  Water  Bath -  heated,  with concentric ring cover,  capable  of  temper-
ature control  (+2° C) ,  installed  in  a fume hood.

5.7  Sample Vials -  amber  glass,  2-5 ml with PTFE-lined screw cap.

5.8  Analytical  Balance  -  capable of weighing 0.1 mg.

5.9  Micro-grinder (e.g.,  Tekmar Tissuemizer, Tekmar Company,  Cincinnati,  OH).
                                  1-7

-------
5.10   Nitrogen evaporation  cevice equipped  «itri a water Datn  tnet  can z±
naintained  at 35-40°  C.  Tr.e N-Evap by Organonation Associates,  Inc.,  ScJtn
Berlin, XA is  suitable.

5.11  Balance  - capable of 100 g to tne nearest  0.01 g.

5.12  Disposable Pasteur Pipets -  sealed  with  aluminum  foil  and annealed
at 450° C fcr  several  h, and rinsed witn solvents before use.

5.13  Drying Oven.

5.14  Annealing Oven - capable of reaching 45QO  C.

5.15  Dessicator.

5.16   Chromatograpny Column  for Alumina  - 5  ml  disposable Dorosilicate
glass serological  pipet  with borosilicate glass wool  plug.   (Glass wool
must  be  extracted with the  appropriate  solvents and allowed  to dry  oefore
use.)

5.17  Gel  Permeation Chromatography Cleanup Device -

     5.17.1   Automated  system:   gel  permeation chromatograph (GPC) ,  e.g.,
Analytical  Biochemical Labs, Inc.  GPC  Autoprep  1002, including

     •    25 mm ID x  600-700  mm  glass column  packed with 70  g  of Bio
          Beads S-X3.

     •    Syringe,  10  ml with Luer Lok  fitting.

     •    Syringe Filter Holder  and  Fitters  - stainless steel  and
          TFE, Gelman  4310 or equivalent.

     5.17.2   Manual  system assembled from parts (Wise, R.H., D.F.  Bishop,
R.I. Williams,  and  B.M.  Austern.  Gel  permeation chroniatography  in tne
                                  1-8

-------
/MS ana
           ysis of  organ -ics  in  sieges.  U.S. EP^., K-niciscl  Err, ire ".rental
Researcn Laboratory, Cincinnati,  CH.  ^5268) (see reference 2, p.  D-25).

5.18  Gas  Chronatog raph  - 1) one  equipped with electron capture detector
(ECO)  and 2)  one  interfaced  to  trie  mass spectrometer  (Sect.  5.19).  Both
snould  have splitless  injection  ports for  capillary column, temperature
programs with 30°  C  hold,  and  should meet all the performance specifications
in Sect. 9.9.

     5.18.1  Column  - 30^5  n  x 0. 25 +_ 0.02 rrm  i .d .  5 percent pnen yl ,  9-1 percent
methyl,  1 percent  vinyl  silicone  bonded  phase  (0.25 in film thickness)  fused
silica capillary column  (J  & W D8-5, or  equivalent).

5.19   Mass Spectrometer -  70  eV  electron  impact  ionization, should  repeatedly
scan  fron  35 to 450 a^u  in 0.95  to  1.00 second and snould produce a unit
resolution (valleys  between  m/z 441-442  less than 10 percent  of the height
of the 441  peak) ,  background-corrected mass spectrum from 20 ng decaf luorotri-
phenyl phosphine (DFTPP)  introduced through  the  GC  inlet.   The  spectrum
shoulc  meet the mass- inten si ty criteria in Table 1-2 (reference 7).  The
mass spectrometer  should be  coupled  with the  GC  such that  the end of  the
capillary  column terminates  within  one centimeter of the ion  source  but
does net intercept  the electron or  ion beans.   All  portions  of the column
which connect the  GC to  the ion  source  should remain at or above  the column
temperature during analysis to preclude  condensation of less volatile  compounds.

5.20  Data System  - should  collect  and  record MS data, store mass  intensity
data in  spectral  libraries,  process  GC/MS data, generate reports,  and should
compute and record  response  factors.

     5.20.1  Data  acquisition  - mass spectra should be collected continuously
throughout the analysis  and  stored on a mass storage device.

     5.20.2   Mass  spectral  libraries  - user created libraries  containing
mass spectra obtained from  analysis of authentic standards should be employed
to reverse search  GC/MS  runs for  the compounds of interest (Sect.  8.2).
                                   1-9

-------
     5.20.3  Data  processing - the data  system should  be  used to sesrcr ,
locate, identify,  and  quantify the compounds  of interest  in each  GC/'-'S
analysis.   Software routines should be employed to compute  retention times
and peak areas.   Displays of spectra, mass chromatograms,  and  library compar-
isons are required  to verify results.

     5.20.4  Response  factors and  multipoint calibrations -  the data  system
should be used  to record and maintain lists of  response  factors  (response
ratios  for  isotope dilution)  and multipoint calibration curves (Sect. £;.
Computations of relative  standard deviation  (coefficient  of variation)
are useful  for  testing  calibration  linearity.

6.0  REAGENTS AND CONSUMABLE MATERIALS  (partially adapted from references
2  and 3)

6.1  Reagents

     6.1.1  Acetic  acid,  acetone, benzene, n_-hexane,  isooctane, rnethanol ,
and methylene chloride  (CH2Cl2)  (pesticide quality, distiHed-in-glass).

     6.1.2  Alumina - neutral, super Woelm or equivalent (Universal Scientific,
Atlanta,  GA).   Extract  alumina  with methylene  chloride  for 30-40  cycles
in a Soxhlet extractor  to remove contamination.  Allow solvent to evaporate.
Prepare activity  III alumina  by adding 7 percent (v/w)   reagent  water to
neutral  alumina that  has  been  activated at approximately 225° C  for at
least 2 h or preferably overnight.   Store in  tightly sealed, clean  glass
container.

     6.1.3  Hydrochloric  acid -  concentrated, make 2N HC1 with reagent
water.   Solvent  clean in a separatory funnel  with methylene chloride.

     6.1.4  Potassium  hydroxide  - reagent  grade, 5N  in  reagent  water.
Solvent clean in a  separatory funnel with methylene chloride.
                                  1-10

-------
     6.1.5  Sodium  sulfete - reagent  grade,  granular  anhydrous, rinsed
with Ch^Cl?  (20 mL/g) and conditioned  at  450° C for 1  h minimum.

     6.1.6  Reagent  weter  - water  in  whicn  the compounds  of  interest en.c
interfering  compounds are not detected by this method.

6.2  GPC Calioration Solutions:

     6.2.1   Corn  oil - 200 mg/mL  in  CH2C12-

     6.2.2  Bis(2-ethylhexyl)phthslate  and  pentachlorophenol  -  4 mg/mL
in CH2Cl2.

6.3  Stock  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  compound purity
is 96 percent  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  -20  to  -10° C  in  screwcapped vials with PTFE-
linec lids.   A mark is  placed  on  the vial at  the  level of the solution
so that  solvent  evaporation  loss  can  be detected.   The vials are brought
to roan temperature prior to  use.   Any precipitate is  redissolved and solvent
is added if  solvent loss has occurred.

     6.3.1  Preparation of  stock solutions - prepare  in  methylene chloride,
benzene, isooctane, or a mixture  of  these solvents  according to the  steps
below.   Observe  the safety precautions given in Sect. 4.  The large number
of labeled  and unlabeled acid and base/neutral  compounds used for combined
calibration  (Sect. 8) and  calibration verification  (9.9.1.3) require high
concentrations (approximately 40 mg/mL) when  individual stock solutions
are  prepared, so that dilutions  of mixtures will  permit calibration with
all compounds  in  a single set of solutions.   The  working  range for  most
compounds is 1-50 ug/mL.  Compounds  with a reduced MS  response may be prepared
at higher concentrations.
                                  1-11

-------
     Siancarcs for GC/ECD  nave  lower .vorking ranges  (e.g.,  0.04 to 2.G jc.<~i_
for single component  pesticides)  than GC/i-'S  standards.   Ho^e.-er,  GC/'E-D
stock  solutions sr.C'jia  De preoerec ,-,itn 51 least 10 rg  of  tr.e pure r-.2-.er~i"
(e.g., in iO ~L of solvent)  10  reduce  potential weighing er^or.

     6.3.2   Dissolve an  appropriate amount  of assayed  reference material
in a suitable solvent.   For  example,  weigh 400  mg  naphtha! ene  in  a  10 rrL
ground glass stoppered volumetric flask and fill  to the mark with benzene.
After the naphthalene is completely  dissolved, transfer the solution to
a 15 ml vial  with PTFE-lined  cap.

     6.3.3  Stock standard  solutions should be checked for  signs of degradation
prior to  the preparation  of calibration  or  performance test  stancarcs.
Quality  control  check  samples that  can  be used  to determine the accuracy
of calibration stancarcs  are  available  from  the U.S.  Environmental Protection
Agency,  Environmental  Monitoring and Support Laboratory,  Cincinnati, Ohio
45268.

     6.3.4   Stock stancaro  solutions should  be  replaced after  6 no, or
sooner  if comparison  with  quality control  check samples  indicates  a  cr.ange
in concentration.

6.4  Injection  Internal  Standard Solutions

     6.4.1  GC/MS  internal standard solution -  prepare 2,2'-difluoroDiDhenyi
(DFB)  at  a concentration of 2 mg/mL in benzene.

     6.4.2   GC/ECD internal  standard solution -  prepare  decafluorobenzo-
phenone (DF8P)  at  a concentration of 2.5 ug/mL in isooctane.

6.5  GC/MS  Secondary Dilution Standards -  using  stocx  solutions (Sect.
6.3),  prepare a  secondary  standard containing  each of the  unlabeled priority
pollutants  in  Table  1-1  at a concentration of 100 ug/mL, or  at a higher
concentration appropriate  to the MS response of the  compounds.
                                  1-12

-------
b.6  Laoe'ea  CornDOunG  SDiKing So'uticn  -  prepare a spiking sc1 -tier  frc~
stock star.carc  solutions prepared es  ir  Sec*.  6.3, or  from mixtures, at
a ccncar.tr a:i en of  iCO ug/rrL, or at a  concentration  appropriate  to  t'r.e
MS response of each  compound.  The deuterium  and 1->C-1 aDel eu  conpourics
listed In Taole  1-1  are  commercially available  individually  or as r.-ixtures
(e.g., :-,erck Sharp  i  Dohrne/Isotcpes, Montreal, Canada),

6.7  Solutions  for  obtaining authentic  mass  spectra (Sect.  8.2)  -  preparc-
;nixtures  of labeled  ana  unlaoeled compounds at  concentrations tnat will
assure that  authentic  spectra are obtained  for storage in libraries.

6.8  Calibration Solutions  -  the concentrations of calibration solutions
suggested in the  following sections are  intended to bracket  concentrations
that will  be  encountered  curing sample  analysis without  overloading GC
columns or saturating  detection systems.

     6.8.1  GC/MS  calibration  solutions - comoine 0.1  ml of the spiking
solution (Sect.  6.6)  with  10, 50, 100, 200,  ana  500 uL of  the  secondary
dilution solution  (Sect.  6.5) and bring  to  1.00 ml total  volume  each.
This  v.'ill produce  calibration solutions  of  nominal 1, 5,  10,  20, and  50 ug/nL
of the  pollutants  end a constant nominal  10 ug/mL of the labele::  corcpouncs.
Spike each solution with 10 uL of the GC/MS  internal  standard  solution,
yielding 20  ug/nL.

     6.8.2  PCS  calibration solutions  -

          6.8.2.1  Aroclor  stock solution  for GC/MS  - prepare  a  solution
in hexane with 250 ng/uL  of each of three PCB mixtures, Aroclor 1016, Aroclor
1254, and Aroclor 1260.

          6.S.2.2  Aroclor  standard  solution for GC/ECO - dilute the stock
solution (Sect.  6.8.2.1) to one-tenth  its  original concentration.   It is
essential  that this  solution be prepared  directly from the batch  used  for
Sect. 6.8.2.1.  Combine  20, 50, 250, 500,  and  1,000 uL of the diluted  standard
                                  1-13

-------
               the GC/ECD intGrr.a*.  s: = r.c = rc s:-l-ticn arc  Dr'r-g  eacr so".,:::"
 to  a  fine' volume cf 5.0 r.L.

      This  will  proauce  calibration solutions  of nominel  concentrations
 of  100, 250,  1,253,  2,5CC, ana  5,000 ng/nL cf  the i:l:l Aroclor mixture
 and a constant nominal  concentration  of 50 ng/rr.L  of internal  standard.

      C.S.3   Pesticide  calioration  solution  -  combine  20 uL  of the GC/ECD
 internal standard  solution with  2,  5,  10, 50,  and  100 uL of a 20  ug/r,L
 stock solution cf  all  chlorinated  pesticides  listed  in Table 1-1 (excect
 toxapnene) and bring  to  a 1.0 ml  total volume.   This will  produce calibration
 solutions  of 40, 100,  200,  1,000,  and  2,OCO  ng/mL of eacn  pesticice arc
 a constant  internal  stancara concentration of  50  ng/mL.

     6.8.4   Tcxapnene calibration solution  -  prepare  toxephene solutions
of 100,  250,  1,250,  2,500,  and 5,000  ng/mL with constant  internal standard
concentration of  50  ng/mL.

     6.8.5  DFTPP  solution  - prepare  at 20 ug/mL  in acetone.

6.9  Stability of Solutions -  all  standard   solutions  (Sect, 6.4-6.8.4)
should be analyzed within 43 h of preparation  and on a monthly basis thereafter
for signs of  degradation.   Standards will  remain  acceptable if response
factors  relative to the  internal standard  correspond within +_! 5 percent
to those obtained  in the  initial  analysis  of the  standard.

7.0  SAMPLE COLLECTION,  PREPARATION, AND  STORAGE

7.1  In  the  field, sources  of contamination  include  sampling gear, grease
from  ship  winches or  cables,   ship  engine exhaust, dust, and ice used for
cooling.  Efforts  should be made to minimize handling  and to  avoid sources
of contamination.   This will usually require  that resection  (i.e.,  surgical
removal)  cf tissue be performed  in a controlled environment (e.g., a labora-
tory).   For  example, to  avoid  contamination from ice, the samples  snould
be wrapped in  aluminum foil, placed  in watertight  plastic bags, and immediate!y
                                  1-14

-------
cooled  In  a covered  ice  cnest.  Organises  snculc not be frozen  prior  to
resection if analyses  will  only be conducted on selected  tissues (e.g.,
internal  organs), because  freezing may  cause  internal organs to rupture
and contaminate  other  tissue.  If organisms  are eviscerated  on board the
survey  vessel,  the remaining  tissue  may  be  wrapped as described above ana
frozen.

7.2   To  avoid  cross-contamination, all  equipment used in sample  handling
should be thoroughly cleaned before each  sample  is  processed.   All  instruments
must  be  of a material  that can be easily cleaned (e.g.,  stainless  steel,
anodized alur.inun, or borosilicate glass).   Before  the next sample  is processed,
instruments snould be  washed  with a detergent solution, rinsed  with tap
water, soaked in  high-purity acetone and methylene chloride,  and  finally
rinsed  with reagent water.  Work surfaces  should be cleaned with 95 percent
ethanol  and allowed to dry completely.

7.3  Resection should  be carried  out by or  under  the supervision of a  competent
biologist.   Each  organism should be  handled with clean  stainless steel
or quartz  instruments (except for external surfaces).  The  specimens  should
come into contact  with precleaned glass surfaces only.   Polypropylene and
polyethylene surfaces  are  a potential source  of contamination and  should
not be used.  To  control  contamination when resecting tissue, separate
sets  of  utensils  should be  used  for removing  outer  tissue and  for resecting
tissue for  analysis.   For fish samples, special care must be taken to avoid
contaminating target  tissues (especial ly muscle) with slime and/or  adhering
sediment from the  fish  exterior (skin) during resection.   The  incision
"troughs" are subject  to such contamination; thus, they should not be included
in the sample.   In the case  of muscle,  a  "core" of tissue  is  taken from
within  the c"ea boarded  by the incision  troughs, without  contacting  then.
Unless specifically sought as a sample, the  dark muscle  tissue that may
exist  in  the vicinity  of the  lateral  line  should not be mixed  with the
light muscle tissue that constitutes the rest  of  the muscle tissue mass.

7.4  The  resected tissue  sample should  be  placed in a clean glass  or TFE
container which  has been washed with detergent, rinsed twice with tap  water,
                                  1-15

-------
 rinsea  once -with distilled  water, rinsed witn acetone, end, finally,  rinses
 with  hign-purity nethylene chloride.  Firing of the glass  jar  at  450°  C
 may be substituted  for  the  final  solvent rinse, but  precautions must be
 taken to avoid contamination as the container  is  cooled after drying.

 7.5  The U.S. EPA  and  other federal  agencies (e.g.,  National  Bureau of
 Standards)  have not yet  provided specific  guidance regarding holding tines
 and temperatures  for  tissue  samples to be analyzed  for  seni-volatile organic
 compounds.   Until U.S.  EPA develops definitive guidance,  the  following
 holding conditions  should  be observed.  Resected tissue  samples should be
maintained  at -20° C  and extracted as soon as possible, but witnin ID days
 of sample receipt.  Complete analyses should be performed  within 40  days.
 These holding  times are based on the Contract  Laboratory Program requirements
 for sediments  (reference  10).   Liquid associated with  the tissue  sample
ttust De  maintained as part of trie  sample (the liquid will contain lipic
material).

8.0  CALIBRATION  AND STANDARDIZATION (adapted  from  reference 2)

 8.1  Estaolish  the  GC/MS operating conditions  in Table 1-1.  Analyze  stancards
per the  procedure in  Sect. 10.2 to demonstrate that  the analytical  system
meets tne detection limits in  Table  1-1  and  the mass-intensity criteria
 in Table 1-2  for  20 ng DFTPP.

8.2  Mass Spectral  Libraries  -  detection and  identification of  compounds
of interest  are dependent  upon spectra stored  in  user created libraries.

     8.2.1  Obtain  a mass  spectrum  of  each  pollutant,  labeled  compound,
end the  internal  standard by analyzing an  authentic standard either  singly
or as part  of a  mixture  in which there  is  not interference between closely
eluting  components.  Confirmation that only a single  compound  is  present
is attained  by  examination of the spectrum.  Fragments not attributable
to the compound under study indicate the presence of an  interfering  compound.
                                  1-16

-------
     8.2.2  Adjust  the  analytical conditions and scan  rate  (for tnis test
only)  to  produce an  undi storied  spectrum at  GC  peak maximum.  An  undis-
torted spectrum will be oDtained  if five complete  spectra  are collected
across tr.e upper half of the  GC peak.  Software algorithms designer  to
"enhance"  the  spectrum may eliminate distortion,  but may  also eliminate
authentic masses or  introduce other distortion.

     8.2.3  The authentic  reference spectrum  is obtained under DFTPP tuning
conditions (Sect. 8.1 and Table 1-2) to normalize it to spectra  from  other
i n s t r u.i en t s .

     8.2.4  The spectrum  is edited  for entry in the  library  by saving the
five most intense mass spectral  peaks and  all  other mass spectral  peaks
greater  than  10 percent of the  base  peak.  This edited spectrum is stored
for reverse  searcn ana for  compound confirmation.

8.3  Polar Compound  Detection - demonstrate that unlabeled  pentachlorophenol
and benzldine are detectable  at  the  10 ug/mL  level  (per  all criteria  in
Sect.  10.4).   The  10 ug/mL calibration standard (Sect. 6.8.1)  can be used
to demonstrate  this  performance.

8.4  Calibration  with the  isotope dilution  technique -  isotope dilution
technique is  used when labeled  compounds are  available  and  interferences
do  not preclude its use.   If  either  of these conditions  precludes isotope
dilution, the  internal standard  method (Sect.  8.5)  is  used  and noted  as
such in the  report.

     8.4.1  A  calibration  curve encompassing the concentration range is
prepared  for  each compound  determined.  The relative response  (pollutant
to  labeled)  vs. concentration in standard solutions is  plotted  or computed
using  a linear  regression.   The example in Figure 1-1 shows a calibration
curve  for an  unlabeled compound and its labeled  analog.   Also shown are
the ^10 percent error  limits (dotted   lines).   Relative Response  (RR)  is
determined  according to the procedures described below.   A minimum of five
data points  are employed for calibration.
                                  1-17

-------
      3.4.2  The relative response  of  a pollutant  to its labeled  analog
 is determined from isotope ratio values  conputed  from  acquired data.   Three
 isotope ratios are usea in tnis process:

      RX =  the isotope ratio measured  for  the  pure  pollutant
      Ry =  the isotope ratio measured  for  the  labeled compound
      RTC =  the isotope ratio of an analytical  mixture of pollutant and labeled
           compounds.

 The m/zs are selected such that Rx>Ry.   If  Fh is  not oetween 2 Ry and 0.5RX,
 the method does not apply and the sample  is  analyzed by the  internal standard
method (Sect. 8.5).

      8.4.3  Capillary  colors usually  separate  the pel 1 utant- labe- ed pair,
with  the labeled conpound  eluting first  (Figure 1-2).   For  this case,

                         Rx =  (area mi/z)/l

at the retention tirce of the pollutant  (RT2)  and

                         Ry =  l/(area m2/z)

at the retention time of the labeled  compound  (RT^).   Also,

            Rm = [area mi/z (at RT2)]/[area m2/z  (at RTi)]

as measured in the mixture of the pollutant  and  labeled compounds (Figure 1-2),
and RR = Rm.

     8.4.4  Special  precautions are taken when the pollutant and its labeled
analog are not chromatographically separated  and  have  overlapping  spectra,
or when  another  labeled  compound with  interfering spectral masses overlaps
the pollutant (a case that  can occur with  isomeric  compounds).   In  such
cases,  it  is necessary to determine  the respective  contributions of the
                                  1-18

-------
pollutant  and  labeled  compounds tc the respective EICP areas.   If  the  pea.
-------
no  label 53 analog,  and  to measure  labeled  compouncs  for iritral cDcra^o1' >
statistics (Sect.  9.5.1) .

     8.5.1   Response factors  -  calioration requires the determination of
response factors (RF)  ^hich  are defined by the  following  equation:

                    RF  =  (As x  Cis)/(Ais x Cs)

where:

      AS = the  area  of  the  target peak in the daily  stancard
     ATS = the  area  of  the  internal  standard  peak
     Cis = the  concentration of the internal  standard  (ug/mL)
      Cs = the  concentration of the compound  in  the  daily standard (jg/nL).

          8.5.1.1   The  response factor is  determined  over  the range of
concentrations described in Sect. 6.8.1, 6.8.2,  6.8.3,  and 6.8.4.   The
amount  of internal  standard  added to each  extract  is the same so that C-;s
remains constant.   The  RF  is plotted versus concentration for each compound
{or class  of compounds  in  the case  of toxaphene)  in the standard  (Cs) to
produce a calibration curve.

          8.5.1.2  Linearity - if the response  factor (RF)  for any compound
is constant (less than 35 percent coefficient of  variation) over the calibration
range, an  averaged  response factor may be used for that compound;  otherwise,
the complete  calibration curve  for  that  compound should be  used over  the
range.

8.6  Combined  Calibration  -  by using  calibration  solutions (Sect.  6.8.1)
containing the  pollutants, labeled  compounds,  and the  internal  standard,
a  single  set of analyses  can  be  used  to  produce calibration  curves for
the  isotope dilution  and  internal  standard methods.  These  curves are verified
each  shift  by  analyzing  the  10 ug/mL  calibration  standard (Sect.  9.9.1).
Pesticide  and PCB calibration  standards must  be  analyzed  separately by
                                  1-20

-------

r. D '  "•  ^fj-ar-j:  cAnno*  b ~ r-r
£.7  Ongoing  CiV:orafion  (see  Stct. 9.9)

9.0  CL'ALITY  ASSURANCE/QUALITY CONTROL  Tor further g uicance,  see  Quality
Ass Jrar.c e/Qual ity  Control  (QA/QC) for 301 (n)  Monitoring  Programs:  Guicancc-
on Field ana  Laooratory Methods  (Tetra Tech 19S6).j

9.1  Eacn  laoorstcrv  tr.at  uses  znis :r,etncc is require-  to  operate 2  ~cr~;".
quality assurance  progra.r.  The  requirements of this  program  consist cf
an initial  demonstration of laboratory capability,  analyses  of replicates
ar.a n.atrix  spikes  usea  to  e/aKete ana cocument aate quality,  ana  analysis
of stanaarcs  ana  blanks usea to  test continued performance.

9.2  Initial  Demonstration  of  GC/MS  Capability  - the  analyst  should make
an initial  demonstration of the ability to generate  acceptaole accuracy
and precision witn  the  GC/MS  component of this  method.  This  ability is
estaolishea  as describee  in  reference i.

9.3  The analyst  is  permitted  to  modify this method  to  improve  separations
or lo,-/er the  costs  of measurements, provided that the new met nod is aenonstratea
to perform  comparably to  the present  method  (i.e.,  with comparable  spike
recoveries  ana precision).

9.4  Blanks  - method blanks  snould be analyzed by GC/MS and  GC/ECD to  demon-
strate freedom from contamination.

     9.4.1   At least one  method  blank must be induced  with eacn batch cf sam-
ples; method blanks must constitute  at least 5 percent  of  all  samples analyzed.

     9.4.2   Method blank  concentrations  of  compounds of  interest  and cf
potentially   interfering  compounds  should be  less than 5  percent  of the
expected values for the corresponding  analytes  in  samples  anc below trie
LLD,  if possible.  It is  recommended that if blank concentrations of compounds
                                  1-21

-------
 cf  interest fc-xce"  "ntr;^ 1 ates; -••'- creator  tr.an 1C percent cf tr.e ccrreszcr.; :r,;
 anslyte conccntrat icns  in  samples, sa^p'ie analysis  snoulc  ce ra'tea  ^r. r-i'
 lie contamination sojrce  is Gnnir-£t=G.

 9.5  S^ircec S3i."."";-:-s  ='"£ rec-'irea to assess metnoc  performance on tne  sa~^'i-
 matr ix .

     9.5.1  For  samples analyzed by isotope dilution, the percent recovery
 (p) of lebelec conpouncs  can be computed by the  internal  standard  met roc.
 (Sect.  £.5; a°c  serves  as  an indication  of analytical accuracy (cut r:-t
 necessarily of extraction efficiency).   After the  analysis of five samples,
 compute  the average percent  recovery  (P) and the  standard  Deviation of
 the percent recovery  ^s->)  for the laoeled ccrrpouncs  only.   Express ir-.e
 accuracy  assessment as a  percent recovery interval  from  P - 2sp to P + 2sp.
 For example,  if  P  = 90  percent and sp = 10 percent, the accuracy interval
 is expressed as  70-110 percent.  Upcate  the accuracy assessment for  each
 conpojna  en a  regular basis (e.g.,  after each 5-10  new accuracy mea

     9.5.2  Laboratories  unaole to use isotope  dilution must analyze matrix
spikes of pollutants  (other than  pesticides and PCBs)  at a frequency  cf
5 percent  of  all samples analyzed or once with  each  sample set, whicneve'-
is more frequent.  Compounds should be added at  concentrations one  to  five
times those in  the sample.

     9.5.3  All  laboratories are recuired  to spike  samples with PCBs and/or
pesticiaes  at  a  frequency of 5 percent  of all  samples  analyzed  or  once
per  sample set, whichever is more  frequent.  The  spike  can be a standard
pesticide mixture or  an  Aroclor mixture,  whichever  is  considered to  be
more  representative  of  the sample.  The  mixture  should be  added at one
to five times the sample concentration of  these  compounds.

9.6   Replicates (i.e.,  analyses of at  least  two  separate  aliquots  from
a  tissue homogenate) must  be analyzed  by  GC/MS and  GC/ECD to monitor the
precision  of  laboratory  analyses.  At a  minimum, 5  percent of the analyses
should be laboratory replicates.   A triplicate  analysis should be performed
with each sample  batch of  over 40 samples.

                                  1-22

-------
9.7  The  Icooratory  snoulc maintain recoras to aefine tne quality  cf  csta
tnat is generates.   These  records  induce documentation of Panics  arc  resorts
cf IcDel-EG compouna  recovery  (5e:t.  5.5.ij, if tre latter is app"ic=o"e.

9.£  Tne laooratory snoulc,  on  an ongoing oasis, cerncnstrate tnrougn  cal iora tic-"
verification that tne  analysis system is in control  (Sect.  9.9.1.3).

9.9  System and Laboratory Performance

     9.9.1  At  tne Beginning  anc  enc of eacn 12-h srnft during  w
are performed,  GC/I-'S system  performance  and calibration are  verifiec
all  pollutants and  labeled compour.cs.   For these  tests, analysis  cf tre
10 uc/mL calibration standards (Sect. 6.8.1) should  be used to  verify all
performance criteria.   The GC/ECD performance is  checkea  at  the  beginning
anc enc cf each shift  or  at  least every 6 h by analyses of  250  ana  100  ug/mL
solutions  of the PC3 and  pesticide  stancarcs (Sect.  6.8.2.Z  anc 6.6.2).

          9.9.1.1   Retention times -  the absolute  GC/MS  retention time
of 2,2'-difluorobipheny 1  should be  within  the range  of 1078  to  1248 sec.
Tne  aosoiute GC/ECG  retention time of  4,4'-DDT should be  within  the  range
cf 1C-50 and 1200 sec.

          9.9.1.2   GC  resolution  fcr  GC/MS analysis - the  valley neight
between anthracene and  phenanthrene  at m/z 178 {or the analogs at  m/z 188}
should  not exceed  10 percent  of the  taller of the two  peaks.

     GC resolution for GC/ECD  analysis  -  the  valley  height between two
peaks should not exceed 25 percent  of the taller  of  the two peaks for the
following  pairs:   beta-  and delta-HCH,  dieldrin and 4,4'-DDE, 4,4'-DDD
and endrin aldehyde, and  endosulfan  sulfate and 4,4'-DDT.

          9.9.1.3  Ongoing calibration verification -  conpute the concentration
of each pollutant (Table  1-1) by  isotope dilution (Sect. 8.4) for  those  com-
pounds  tnat have labeled  analogs.   Compute the concentration of each pollutant
                                   1-23

-------
      :' s s no  1 e D e "; e G  & n a } c j v, i * ,-,  t r, e  n e a r e s ; e "i u t i P. g  1 a o e 1 c c s t a r c a
 Commute :ne concentration  cf the  U^eleci compouncs  Dy  the  internal star.carc
 r~5t"oc.   -'-.150 COT. DJte  inc: i .' icje'i :e st "' c ice- concer tr=tiors   c tct= 1  22E
 snc  i ox a 3 n. -"e  concentrations oy  the  interne! standard  ~ethoc  (GC''EC3-).
 Tnese  cere en -ra 1 1 or-s  are carpu-ec casec cr, the cal iSraticn cate je'e-'rir.s-
 in Sect. S.  Preparations of new  calibration standards or revisions  of
 calioration curves  ere  required  if  observed responses cf analytes  vary
 from predicted responses Dy more  than +_20 percent.   Samples and blanks
 may be run only after calibration performance meets this control li-rit.

          9.9.1.4   Multiple peaks -  escn  ccmpcjnd  injeciec shou'c  give
 a single,  distinct GC peak.

     9.9.2   DFTPP  spectrum validity  - inject 1  uL  of the DFTPP solution
 (Sect.  6.8.5  jither  separately  cr within  a few  seconcs of injection  of
 the  standard (Sect. 9.9.1)  analyzed  at  the beginning of eacn shift.   The
 criteria in Table 1-2 should be met.

 10.0  PROCEDURE  (see Figure 1-4)

 10.1  Sample Processing and Extraction

     10.1.1  Mince  tissue sample  with a  scalpel  and homogenize tne sanole
 to a uniform consistency with a micro-grinder.  Care must  be taken to ensure
 that the  micro-grinder  is thoroughly  cleaned after each  use.  This usually
 entails disassembly  of the unit.   Devices with large  surface areas (e.g.,
 blenders,  meat grinders) should be avoided,  as they  are  difficult to  clean
 ana a small  sample is difficult to remove after grinding.

     10.1.2  Dry weight  determination -   if  sample  size permits  and  dry
weight concentrations  are required, dry weight determinations may be performed
 as follows:   transfer an aliquot of approximately 3 g  (weighed to the nearest
 0.1  g)  to  a preweighed  dish.   Allow  the  sample  to dry  in  a  1050 c  oven
 overnight,  and  determine the solid residue weight   (to the nearest  0.1  c) .
Calculate  ana report  the percent  solids (Ts)  as:
                                  1-24

-------
          Ts  =  [dry  residue wt (g)]/~wet sample  wt  (g}[ x 100^

     10.1.3  Weigh  a 20 to 25-g (wet weight)  tissue  aliquot to the nearest
0.1 g.   Add  any excess  liquid from the sample  directly  to  the round bottci
flas<  to  be  used  for Soxhlet extraction;  this  will  help to avoid saturating
the cellulose  extraction thimble with tissue fluids,  which  can lessen extraction
efficiency.   Also  add  approximately  50 ml  methanol and 140 ml CH2C12  to
the flask.

     Spike the  tissue  aliquot  with  10 ug of  each stable  isotope labeled
base/neutral  compound and 15 ug of each  labeled  acid compound.  Mix the
spiked,  homogenized  tissue sample with approximately  30 g precleaned  Na2$04
in a beaker  to  improve  the  texture of  the  sample.   Place  the mixture  into
a precleaned  Soxhlet thimble for  extraction.   Add  20  ml of methanol directly
to the  thimble  and stir  to  enhance the  removal  of  water  from the sample.
Cover the  sample with a  thin layer of  solvent-cleaned  glass wool.

          10.1.3.1   An  as  yet  unvalidated  procedure may enhance extraction
efficiency for  phenolic materials.   This  procedure  entails adding 20  ml
of acetic  acid directly to the thimble (rather  than methanol, as specified
in 1C.1.3).   The purpose of adding acetic  acid  is to reduce  the extraction
pH and  thus  to increase  the  solvent  solubility  of  phenols.  Acetic  acid,
1 ike methanol ,  would  facilitate water  removal.   The  acetic acid step  is
recommended  only if  laboratory validation  data  is made  available.
     10.1.4  Soxhlet extract  the  ti ssue/Na2S04 mixture with
(2/1)  for 24 h  (60-90  cycles).  Stir  the  sample in the  thimble at least
twice (after the  first cycle and after approximately 12 h) to prevent  solvent
channeling (replace the glass wool  cover).   The Soxhlet  apparatus  should
be wrapped  up  to the condenser with aluminum  foil to ensure even heating
during cycl ing.
                                  1-25

-------
      10.1.6  Liquid-1iquid  extraction -

          10.1.6.1   After  Soxhlet  extraction, transfer  me  extrac* to a
 500 ml separatory funnel.   Rinse  the Soxhlet flask  with  clean extraction
 solvent  anc add  this  rinse to  the extract in  trie  separatory funnel.  Wash
 the solvent extract  with  approximately 100 ml of pH  2,  50 percent saturated
 Na2S04 organic-free  water; the pH of  the water  should  be adjusted with
 solvent-cleaned  HC1.   Collect  and store  the  CH2Cl2  layer.   The purpose
 of washing the extract  with  an acidic, aqueous solution  is  to remove  water
 and nethanol  from  the C^Cl? and  :o enhance the  partitioning of acidic,
 organic  compounds  into  the CH2C12 layer.  Re-extract the  acidic, aqueous
 phase twice with 60  ml of  clean  CH2C12 and  add  both extracts to the initial
 CH2Cl2 fraction.

          10.1,6.2   Adjust  the pH of the aqueous  phase  to >_12 with solvent-
 cleaned,  6 N KOH.   Back-extract the base  compounds  three  times with  60  ml
 CH2C12-   The  pH adjustment to alkaline  conditions  ehances the partitioning
 of basic  compounds  into  the  CH2Cl2 layer.  Combine all CH2C12 layers from
 Sect.  10.1.6.1  and 10.1.6.2.

          10.1.6.3   Formations  of emulsions or precipitates during liquid-
 liquid extraction  should be  noted and  considered when  reviewing  results.
 The addition  of  N32S04  may reduce emulsions.  However,  if  the  emulsion
 interface between  layers  is more  than one-third the  volume of the solvent
 layer, the  analyst must  emloy mechanical  techniques to complete the phase
 separation.   The  optimum technique depends  on the  sample  and may include
 stirring, filtration  of the  emulsion through pre-cleaned glass wool,  centrifu-
gation, or other  physical methods (reference 2).

     10.1,7  Dry  the total combined  solvent extract by pouring it through
an anhydrous Na2S04 drying column (approximately 30  cm x 2 on).  Use approxi-
mately 30 ml  of  CH2C12  to  rinse the drying column and combine this with
the dried  extract.   Collect the extract in  a Kuderna-Danish (K-D)  concentrator
containing 1 to 2  clean boiling chips.
                                  1-26

-------
     10.1.8   Attacn a  3-oall  nacro Snyaer  column to the K-D concentrator
and concentrate the extract  on  an  80° C water bath.  Place tr.e K-D apparatus
on the hot water oath so that  the  concentrator  tune is partially immersed
in the hot water and the entire lower rounded surface of the flask is  bathed
with hot vapor.  At the proper  rate  of distillation, the balls of the  column
will actively chatter but the  chambers will not flood with condensed solvent.
When the apparent volume reaches  5 ml,  remove  the K-D apparatus from  the
water bath and rinse the flask  with  3 ml CH2C12 draining into  the concentrator
tube.   Reduce the contents of  the  concentrator tube to 3 ml using a stream
of purified N2 gas, never allowing  tre extract to go to cryness.

     10.1.9   Alternative methods of extract  concentration  may be used  if
evidence of acceptaole performance [i.e., retention of more  volatile compounds
(e.g., naphthalene) comparable  to  that of K-D concentration]  is provided.

     10.1.10   Accurately measure the extract volume,  Vj,  remove a  small
aliquot ( 100 uL)  and apply  directly to a preweighed microbalance weighing
boat.   Allow the  solvent to   evaporate  on a hotplate at low  temperature
(less  than 40° C).   Quickly  remove  the  boat when it is dry and weigh  the
residue and boat  to the nearest  0.1  mg.  Calculate the  total  extracted
residue, R-r,  as:

               *T (mg}  - [Vt (ml)  x  W (mg) x 103]/Va (uL)

where:

     Vy = the extract volume (  3 ml) in mL
     Va = the aliquot volume transferred to the weighing boat  in  uL
      W = the  residue  weight  in mg  as determined with the analytical micro-
          balance.

Report the total extracted residue  as Rj (mg)/sample wet weight (g).

     10.1.11   Extract cleanup  - GPC  cleanup is required to separate biological
macromolecules from the analytes.
                                   1-27

-------
          10.1.11.1  GPC  set-JD and calibration  (reference 2).

               10.1.11.1.1   Place 70 g of Bio Beads  S-X3 in a 400-mL beaker.
Cover  the beads with Cr^Cl?.  Allow the beads  to  swell overnight.  Transfer
the swelled beads to the column and start pumping  solvent through the column,
from  bottom to  top,  at  5  ml/min .  After  1 h,  adjust the pressure on the
column to 7-10 psi  and pump  for an  additional  4 h  to  remove  air  from  the
column.  Adjust the colunn  pressure as required  to maintain 7-10 psi.

               10,1.11.1.2   Calibration of the  column  -  Load 5 ml of tne
corn oil  solution  into  sample loop  No.  1  and 5 mL of  the  phthalate-PCP
solution  into loop  Ho.  2.   Inject the corn oil  solution and collect 10 rnL
fractions  for 36  min.  Determine the corn oil elution pattern by evaporation
of each  fraction  to  dryness followed  by  a gravimetric determination of
the residue.   Analyze  the phthalate PCP fractions by GC/FID,  a UV-spectro-
photometer  at 254 nm, or  a GC/MS system.   Plot the  concentration cf each
component in eacn fraction versus total el uant  volume.  Choose a "durnptirne"
that  allows >_85 percent  removal of  the corn oil  and ^_85 percent recovery
of the phthalate.   Select  the "collect  time" to  extend  at  least  10  min
after  the  elution of pentachloropienol .   Wash the column at least 15 min
between samples.   Typical parameters are:  dumptime,  30 min (150 ml); collect
time,  36  min (ISO ml);  and wash time, 15  min (75  ml).  The S-X3 Bio Beads
column may be reused  for several months, but  should  be  checked by  frequent
system recalibration  for every 20 extracts loaded onto  the GPC.

          10.1.11.2   Extract cleanup  - Prefilter  or  load all extracts via
the filter holder to  avoid particulates that might  cause system blockage.
Load  the  3-mL extract onto  the GPC column.   Do not  apply excessive pressure
when loading the GPC.   Purge the  sample  loading  tubing  thoroughly  with
solvent between extracts.   Process  the extracts  using  the dump, collect,
and wash  parameters as selected from the calibration and collect the cleaned
extracts  in  400-mL  beakers.
                                  1-28

-------
ar.c recalibrate  the  system ones for every 20 extracts  loaaec onto  tr:e  GDC.
Tr.s: recoveries and  eiutien profiles are reocrtec  as  del 1 veraol es.

     10.1.12  Transfer tr.e extract to e Kucernc-Oonisr,  (<-D)  ccr,: entratc-r
consisting  of a  10-mL concentrator tube, a 500-mL  evaporative flasK, boiling
cnips,  and a  Snycer column.   Carefully  concentrate  the extract to 2.5 nL
using  metnocs  previously  cescribec  and the f<2 gas  Dl Givccwn technique.
Nitrogsn blowdown  should  be  performed  at approximately  353 C.   A gentle
stream  of  clean, cry  N2  (filterec through a column  cf activated carocn;
should  oe  used.   The ir.side walls cf the tube containing the extract  shculc
DC- rinsed ccwn with  trie  appropriate solvent several  tires Curing  concentration.
The extract must  not be  allowed to go to dryness.

     10.1.13  Use  a 20 percent aliquot (500 uL)  cf the extract for alpine
cotuirn  cleanup  ana  subsequent GC/ECD  analysis  for  pesticides and PC3s.
Use the remaining  80  percent (2 ml)  for GC/MS  analysis.   Carefully reduce
the 2  ml extract  to  400  uL using the N2 blowdown  technique.

          1C. 1.13.1   Solvent excnange of extract  for al'jnina cleanup (reference
2) - transfer 0.5 ml of  the extract to  a separate concentrator tube.   Add
5 ml of hexane  and  a  boiling chip and  mix using a  vortex mixer.  Attacn
a twc-Dali  micro-Snyder  column.  Pre-wet the Snyder  column by adding  0.5 ml
of hexane  to the top of the column.  Place the K-D  apparatus on a hot water
bath (80-900  C) so  that  the  concentrator tube is  partially immersed in
tne rot water.   Adjust  the vertical position of tne  apparatus and the water
temperature as required  to complete  the concentration  in 5 to 10  min.
Concentrate the  extract  to an apparent volume  of  approximately 1  ml.   Use
N£ bl owe own to  reduce the volume to 0.5 ml.  Dilute  to  1 nL  by adding C',5 mL
of acetone. Proceed with alumina column cleanup.

     10.1.14  Alumina  column  setup and  use  -  The alumina column cleanup
is required  to remove  polar interferences prior to GC/ECD  analysis of pesticices
ano FCBs (reference  2).
                                  1-29

-------
          10.1.14.1   Add  3  3 of activity  III  neutral alumina to tns  clean
5 ml dispcsaole serological pipet (with glass wool  plug).   Tap trie column
to settle the alumina.   Do not pre-wet the alumina  with solvent.

          10.1.14.2   Transfer  the  1.0  ml  hexane/acetone extract (Sect.
10.1.13.1) to the top of the  alumina column  with a  disposable Pasteur pipet.
Collect the eluate  in  a  10  rnL K-D concentrating  tube.  Add 1 ml of hexane
to the original  extract concentrator tube to  rinse  it.   Transfer  these
rinsings to the alumina  column.  El ute the column with an additional 9 ml
of hexane.  Do not allow  the column to go dry during the addition and eluticn
of the sample.

          10.1.14.3   Note  that batches  of alumina may differ and storage
may alter the water  content of deactivated alumina.   Thus, column performance
must  be  checked  regularly and for each batch of alumina.   PCB and pesticide
standards (e.g., from  Sect.  6.8.2.2,  6.8.3)   and  a  suitable model  polar
compound (e.g.,  tribromophenol) should be used  to  determine the appropriate
elution volumes for these pollutants.  Recovery  of  single  PCB or pesticide
components  should be greater than 85 percent and  tne tribromophenol should
not oe detected.

          10.1.14.4   Concentrate the eluate to 500 uL using a nicro-Snyder
column and the  N2 blowdown technique {e.g.,  Sect. 10.1.13.1).

          10.1.14.5   Care must be taken to allow  the N£ gas to create only
a small dimple on  the  surface of the solvent  and  prevent blowdown to dryness.
Submit extract  for  GC/ECD analyses.

10.2  GC/MS Analysis

     10.2.1  Establish  the following operating  conditions  for the GC (Table
1-1):   5 min at 300 C; 30-280° C at 80 C/min; isothermal at 280° C  until
benzo(g,h,i)perylene)  elutes.   Make certain  that the  concentrated  extract
or standard  is at room temperature  and make  note of any precipitate that
does not redissolve.
                                  1-30

-------
     1C. 2.2  ncd 5 .L c* tr.e GC/'-'S "interne',  star.car-; sc'^ticn tc t~e  4'1D  -._
extract to yield a 10  ug  spike.  Add  the  solution i~ned iatel y prior  to
injection  tc  minimize  tr.e  possibility of  less oy evaporation, acsc^Dtio^,
or reaction.   Mix thoroughly.

          10.2.2.1  It  is  advised that a  late  eljting internal  injection
standard  (e.g.,  5-alpha-cholestane} oe  used in audition  to  DF3.   The ^se
of early  eluting  (DF3) and  late elating  injection stanaards «ill  allow
the analyst  to detect and compensate  for problems in the GC  injection port
related to differential   loading of ana'ytes  onto the GC column.

     10.2.3  Inject  1.0-1.5 uL  and  start  the GC column initial isothermal
hold.   Start MS  data collection after  the solvent peak elutes.    Stop cata
collection after the benzo(g,h ,i)perylene elutes.

          10.2.3.1  Dilution and re-injection  are required  for samples
that exceed  the  upper concentration  limit  of the calibration standards.
Data for  compounds witnin  the  calibration range should be acquired  in the
inif'el run.  Data for compounds exceeding the  calibration  range should
be acquired  after  dilution.   Respike  the  sample  with labeled  compounds
and assume  100 percent recovery.   This assumption is net unreasonaDle con-
sidering  the  high concentrations of  native  compounds involved when dilution
is necessary.

10.3  GC/ECD Analysis

     10.3.1  The recommended GC conditions are modified from those  specified
in reference  2:

          Helium carrier gas:  4 mL/rr.in at 280° C and 25 psi
          Septum purge:  15 mL/min
          Split  vent:  none
          Initial temperature:  60° C, initial  hold - 2 min
          Program at 25° C/min to 160° C
                                  1-31

-------
          Program at 50 c/~in  from 160° C
          Final temperature:   2700 C;
                              hold until decachlorobiphenyl elutes
          Injection port temperature:  225° C

     10.3.2   Add  10 uL of the  GC/ECD internal  standard solution  to the
500 uL extract to  yield a 25 ng  spike.  Add the solution  immediately  prior
to injection to minimize the possibility of loss by  evaporation, adsorption,
or reaction.   Mix  thoroughly.  Inject 1.0-1.5 uL.

     10.3.3   Dilution  and  re-injection are required for  samples that exceec
the upper concentration limit  of  the calibration standards.   Data for compounds
within  the calibration range  should  be retained  from the initial  run.
Data for compounds  exceeding the calibration range  should be  acquired  after
d ilution.

     Column overloading can result in abnormal  peak shape, which can  reduce
the accuracy  of quantification.  It may  also result  in  a marked increase
in the  retention  time of  the  peak  maximum, which can  displace an  analyte
frcm  the retention  time  window  established  with standards  at lower concentra-
tions .

10.4  Qualitative  Determination

     10.4.1  Qualitative determination  is  accomplished by comparison  of
data  from  analysis  of  a sample or blank with data  from  analysis of the
shift standard (Sect.  9.9.1) and,  for  GC/MS analyses,  with data  stored
in the  spectral libraries (Sect.  8.2.4).   Identification  is confirmed  when
spectra  and  retention times agree per the  criteria below.

     10.4.2   Labeled  compounds and  pollutants having  no labeled analog:

          10.4.2.1  The signals for all  characteristic  masses stored  in
the spectral  library  (Sect.  8.2.4)  should be present  and  should maximize
within  the  same two  consecutive  scans.
                                  1-32

-------
          10.4.2.2   Either l) tne background corrected  extracted  ion current
Defile (EICP)  areas, or 2) the corrected relative intensities of  tne r.iss
spectra]  peaks at  the  GC  peak maximum snould  agree  within  a  factor  of two
(0.5 to 2 times)  for all masses stored in the spectral  library.

          10.4.2.3   The  retention  time difference  between an  analyte ana
the nearest eluting internal standard  during  sample analysis  should be
within  +_5  scans or  +_5  sec  (whichever  is  greater)  of  this difference in
the shift standard  (Sect. 9.9.1).

10.5  Pollutants  Having  a Labeled Analog:

     10.5.1  The  signals for  all characteristic  masses  stored in  the spectral
library (Sect.  8.2.4)  should be present and  should  maximize  within tne
same two consecutive scans.

     10.5.2  Either  1) the background corrected  EICP areas, or 2) the corrected
relative intensities of the mass  spectral peaks  at  the GC  peak maxirnur,
should  agree within a  factor of two for all  masses  stored  in the spectral
1 ibrary.

     10,5.3  The  retention  time difference between  the  pollutant and its
labeled analog  should agree within ^2 scans  or +_2 sec  (whichever  is greater)
of this difference  in the shift standard (Sect.  9.9.1).

10.6  If an experimental mass spectrum contains  masses  that  are not  present
in the reference  mass spectrum, an experienced  spectrometrist is  to determine
the presence or absence  of the compound.

10.7  Chlorinated  Pesticides  and PCBs

     10.7.1  Single  component chlorinated pesticides are tentatively identified
by comparison of  sample  peak  relative retention  times  to  those  of authentic
standards  (Sect. 6.8.3).   Three  times the standard deviation  of relative
                                   1-33

-------
retention  times established  from calibration standarcs (Sect.  6.8.2; cin
be used to calculate  relative retention  tine  window boundaries.  Confirm
the  identities of pesticides  by comparing  the  relative retention  t:r,es
of sample and  standard  peaks on another  column pnase (e.g.,  865  dimetnyl-
[14%]-cyanopropyl phenyl  polysiloxane or J&W DB-1701).  Confirmation by
GC/MS is required  when  concentrations  are sufficient.

     10.7.2  Peaks  of multi-component  mixtures (PCBs  and  toxaphene) are
tentatively identified  in samples by comparison  of relative retention times
to those  of authentic  standards (Sect. 6.8.2.2  and  6.8.4).   Three  times
the standard deviation of relative retention times established  fron standards
can be  used to  calculate relative retention time window boundaries.  Choose
as many peaks as possible while avoiding  those with potential  interferences
(e.g.,  PCBs co-eluting with  DDT and DDE isomers).  Label  on  all sample
chroma tog rams the  peaks identified as  PC8  and  toxaphene congeners.   All
GC/ECD chromatograms are part of the del iverables.  Interpretation of chrofna-
tograms requires  the  attention of an  experienced  analyst.   Confirm the
identities of  all selected  congeners by injection on an alternate colunn
phase  (e.g., J&W 03-1701). Confirmation  by  GC/MS is required if  concentrations
are sufficient.

10.8   Tentatively  Identified Compounds (GC/MS Analysis)  - The  ten  non-target
peaks of greatest  area  in the RIC (reconstructed ion chromatogram)  should
be identified and  quantified, if possible.

     10.8.1 Guidelines for making tentative identification (reference  10):

     1)   Tentative identifications should be based on a forward search
          of the EPA/NIH mass spectral library.  Sample  spectra  should
          be visually compared  with the most similar  library match.

     2)   Relative intensities of major  ions  in the  reference spectrum
          (ions greater than  10 percent of  the  most  abundant  ion)
          should be present in  the sample spectrum.
                                  1-34

-------
     3)   The  relative  intensities of the major  ions should  d;ree
          within +20 percent.   (Example:   For  an  ion with an abundance
          of 50 percent  in the standard  spectra, trie  corresponding
          sample ion abundance must be between  30  and  70  percent.)

     4)   Molecular  ions  present in reference  spectrum should be present
          in sample  spectrum.

     5)   Ions  present in  the reference  spectrum  but not in the sample
          spectrm  should be  reviewed for possible  subtraction  frcm
          the  sample spectrum  because  of background  contamination
          or co-eluting  compounds.  Data  system library  reduction
          programs can sometimes create  these discrepancies.

          10.S.1.1   If,  in the opinon  of  the mass  spectral  specialist,
        tentative identification can be made, the compound should be reported
as jjnjqiown.   The mass spectral specialist should give additional  classification
of the unknown compound if  possible (e.g.,  unknown phthalate, unknown  hydro-
carbon, unknown aromatic  compound,  unknown  chlorinated compound).  If probable
molecular weights can be  distinguished,  include then.

     10.8.2  Tentative  quantification  -  quantification  of TIOs will  be
based on the internal standard technique  and  an  assumed  response factor
of one  (in  the absence  of data from authentic  standards).  The uncertain
nature of this  quantification should be clearly noted  in the data report.

11.0  QUANTITATIVE DETERMINATION (CALCULATIONS)

11.1  Isotope Dilution -  by adding a known  amount of a labeled compound
to every  sample prior to extraction, correction for losses  of the pollutant
during the analysis  can be made because the  pollutant and its labeled  analog
exhibit  similar behavior during extraction, concentration, and gas chroma-
togrephy.   Note that pollutants and their  labeled analogs  are not always
retained  identically by complex  matrices,  so  their behavior during the
extraction step may  differ.
                                  1-35

-------
     11.1.1   Relative response  (RR) values  for  sample mixtures are use:!
in conjunction wiih calibration curves cescriced in Sect.  9.4 tc determine
concentrations directly,  so long  as  labeled compound  spiking levels are
constant.

     11.1.2  Specifically,  the concentration, C(in  ug/kg), can be determines
as:

                    C( ug/kg)  = CA( ug/kg) x RR x n
where:

     CA = the  concentration of the stable isotope-labeled  compound as spiked
          into the  sample, wet weight
     RR = relative  response of unlabeled pollutant  to isotope labeled surrogate
          in  the  sa/nple
    RRi = relative  response of ith point in calibration
     Zi = absolute  amount of  unlabeled compound of  ith  point  in  calibration
    ^Ai = absolute  amount of  labeled compound of i^h point in  calibration
      n = number  of calibration points.

11.2   Internal  Standard  -  all  data  reported as  determined  by this method
are uncorrected  for method recoveries.

     11.2.1   GC/MS internal  standards method  -  Compute the concentration
in the  sample, C {in  ug/kg), using  the response  factor, RF, determined
from calibration  data, and the following equation:

               C(ug/kg)  =  (Ax x Zis x 1.25 x 103)/(S x  Ais x  RF)
                                  1-36

-------
where:
      Ax = the  area  at the  characteristic  mass for  the  compouna  in  the
           sample
     A-js = the area of the  characteristic mass  for the  internal standard
     Z-js = the  absolute amount,  in  ug,  of  the  GC/MS  internal  standard added
           to the final extract  prior  to instrumental analysis
       S = sample wet weight  (g)  that  was extracted.

           11.2.1.1   The  stable  isotope  labeled  compound recovery, X, is
determined and reported for  each  sample  in  the  following manner:

                    X = [C(ug/kg)/CA{ug/kg)j  x  100%

where:

     CA = the concentration  of the stable isotope labeled compound as spiked
           into the sample

     11.2.2  GC/ECD internal  standard  method  -

           11.2.2.1   Pesticides  -  compute the  concentration  in the sample,
C (in ug/kg), using the response  factor  (RF, determined  from calibration
data) and the following equation:

               C(ug/kg) =  (Ax x  Zis  x  5  x 103)/(S x A1s x RF)

where:

      Ax = the area of the  integrated  GC peak for the compound  in the sample
            (Ax represents  the summation of areas for a group  of GC peaks
            if toxaphene is  being  quantified)
     A-js = the area of the  integrated  GC peak for the  internal  standard
                                   1-37

-------
     Z^s ~ the absolute amount,  In  ug,  of  the GC/ECD Internal  standard
            added to the final  extract  prior to  instrumental  analysis
       S = the sa.-nple wet weight  (g) that  was extracted.

           11.2.2.2  Accurate PCB  quantification  is difficult to achieve
 in routine full-scan analyses.   It  has been  common  practice  to quantify
 PCBs with packed-column GC/ECD by comparing several selected  peaks in samples
 to corresponding peaks in commercial Aroclor formulations that most closely
 resemble the sample.   Shortcomings  of this technique have been described
 elsewhere (e.g., references 8  and 9).   The critical difficulties with  this
 procedure relate  to two  factors:  (1) environmental  PCB assemblages  often
 differ considerably from comnercial  Aroclor mixtures because of the variable
 properties of PCB congeners (e.g., aqueous solubility,  volatility, suscepti-
 bility to biodegradation)  and  (2) the ECD  has a markedly variable response
 to the  209  PCB  congeners depending on the numoer and position of chlorine
 atoms on the biphenyl nucleus  (e.g., reference 10).

     It has  been suggested that "the least systematic error [in  PCB quantifi-
 cation] will  be  given by the summation  of  all or at least  nearly all  areas
 of PCB  peaks corrected by their  individual  ECD-response factor and  their
 biphenyl  content"  (reference  11).   Another  alternative is  to use GC/MS
 instead of GC/ECD.   However, GC/MS analysis is relatively insensitive  unless
 selected ion monitoring (SIM)  is  used,  which can involve considerable  effort
and expense.

     The quantification technique recommended in this 301 (h)  document  relies
 on high resolution  (capillary column) GC/ECO and a determination of response
 factors  for resolvable PCB peaks (as  suggested in the previous paragraph).
 The technique is modified  from the Webb  and McCall  technique  (reference  12),
which  has  been   widely used for  packed-column PCB quantification.  Briefly,
 the resolved  peaks  in a PCB standard (Sect. 6.8.2.1,  6.8.2.2)  are quantified
by GC/MS and  GC/ECD.   The  GC/MS results are used to correct  for the variability
of ECD  response.  Samples  are analyzed and quantified by GC/ECD.  Total
PCBs are  calculated  as the sum of all resolved,  response factor-corrected
PCB peaks.
                                  1-38

-------
               11.2.2.2.1  GC/MS analysis of PCB  standard - Each resolvable
peak in a PCB calibration standard (Sect.  6.3.2.1)  is  quantified by  GC/MS,
which  can  identify the  chlorine content of biphenyls  in each peak and can
quantify PCBs based  on their chlorine  content.  This  quantification  does
not require that  the analyst know the exact identity of  the congeners  consti-
tuting a peak, only  the chlorine content must be determined.

     Another  GC/MS calibration standard is necessary to  perform this quanti-
fication.  An MS  response factor standard consisting  of representatives
of all  the congener groups (mono-  through decachlorobiphenyl) is necessary
to convert  areas  of  peaks in ion plots  to  the appropriate masses of chloro-
biphenyls.  A standard  solution should be made with  approximately 10 ng/uL
of each of  the following  congeners  (see  reference 13  for  an explanation
of these choices):

           2
           2,3
           2,4,5
           2,2',4,6
           2,2',3,4,5'
           2,2', 4,4', 5,6'
           2,2', 3,4',5,6,6'
           2,2', 3,3',4,5',6,6'
           2,2',3,3',4,4',5,5',6,6'  - (used for nona-  and deca-congeners).

These congeners are  available  from Ultra Scientific, Inc.  (Hope, RI)  except
the  heptachloro-congener, which is available from Wellington Environmental
Consultants,  Inc. (Guelph, Ontario,  Canada).  The primary  quantification
ions  used  for mono- through  deca-chlorobiphenyl  are:  188, 222, 256,  292,
326, 360, 394, 430,  464,  and  498.   The spectrum  for  each  peak should be
manually confirmed at least once for the Aroclor standard.
                                  1-39

-------
     Relative  amounts of cc-e'iuting  congeners of different chlorine  conta-
in a giver peak  in the stancard can oe aetermined  raring  GC/MS  analysis.
Co-e'uticn  c^n t? accojr.tsc  for .-,itn appropriate  response  factors.  Fcr
example,  if  a peak is composed of tetracnloro-  and pentachlorc-isomers
as determined  by  ion plots of n/z 292  and  32G, tne 2,2',4,6 response  f-ictc'
is used for  tne  m/z 292 area and the 2,2',3,4,5'  response  factor  for tne
m/z 326 area.   Care muse be taken to  ensure  that M-70 ions are not  interprets;:
as M+ ions if congeners differing by  two chlorine atoms co-elute.

               11.2.2.2.2  GC/ECD analysis  of PCB standard (Sect.  6.3.2.2)  -

     Each resolvable  peak in  the PCB calibration  stances  is  quantifier;
by GC/ECD according to the internal standard  technique  (Sect.  8.5).  The
GC/ECD araTysis is performed with the same GC column phas-3 and temperature
program used  for GC/MS analysis of the standard.   An ECO  response  factor
(RF) is  establisnea  for  each peak  based  on the GC/MS analysis  of  tne PCB
standard  using  tne equation defined :n  Section 8.5.1

where:

     ^s     = the  mean  concentration  of  the peak  in  the PCB stancard as
             determined by GC/MS (determined with at least three  replicate
             analvses).
               11.2.2.2.3  GC/ECD  PCB  quantification  in  samples - Total
PCBs are calculated as  the  sum of all PCB  peaks identified  in  a sample
(Sect.  10.7.2):
         C  (ug/kg, wet wt)  =
                  n
                 £  [(Ax x Z-s  x  5  x  500)/(S x Ais x RF)]i
                                  1-40

-------
      i - escn identif;ec  PCS peak, ,vitn n total  peaks
     Ax = area c* the integrated GC Deak for tne  compound in the sample
     -JS = area of the integrated GC peak for the  internal standard
     TS = the  absolute amount,  in  ug, ox ths GC/ECD  internal  stancar;
         acced to the final  extract prior to instrumental analysis
      S = tne sample wet weight  (g) extracted
     RF = calibration response factor (Sect. 8.5.1).
             11.2.2.2.4  This quantification  method  involves two nctewcrtny
imitations:

    (1)  Interferences cen be a significant  problem in ECO analyses.
        PCS  peaks  co-eluting with  interferents may be  neglected
        or quantified, in either  case  resulting  in a decrease  in
        accuracy.   It is  essential that experienced  analysts evaluate
        chromatograms to determine the  presence  of suspected inter-
        ferents.  Interferents suspected of  overwhelming PCS  peaks
        should  be  neglectec.  Trie  alumina  column cleanup  step  is
        designed to preclude major interferences.  High resolution
        capillary  columns also  reduce  the  potential for co-eluting
        interferences.

    (2)  When  two  or more congeners  have  identical retention times
        on a given  column  phase, it is impossible  to determine  their
        relative  concentrations in a peak when using GC/ECD.  Thus,
        it is not  possible  to determine  whether sample peaks are
        composed  of the  same relative  combination of congeners  as
        corresponding standard  peaks.  Thus,  the response  factor
        for  a  peak may be different during calibration and sample
        analysis.   The potential  error in  assigning  appropriate
        reponse factors  has been minimized  in  this technique  by
        the use of  high resolution capillary  columns.
                                1-41

-------
               11.2.2.2.5   Alternative techniques  of detection [e.g.,  Hall
electrolytic conductivity detector  {HECD) or MS (with  selected  ion monitoring}]
can  provide comparable or superior  PCS  identification and quantification
relative to ECD (e.g.,  references 13 and 14} and  are acceptable substitutes
for  ECD detection.  Although ECD is widely available and  is more sensitive
for PCBs than HECD or MS, HECO has  a  linear response to  chlorine content
and is more specific  to  chlorinated compounds,  and  MS offers more definitive
compound identification  than ECD.

          11.2.2.3  Quantify PCBs by summing  the  response  factor-corrected
areas of the characteristic  PCB  peaks identified in Sect.  10.7.2.   Report
the results as  total  PCSs.

11.3   Report results  for  all  pollutants  and labeled compounds found in
all standards  and  samples, in  ug/kg,  to  two significant figures.   Note
in the  report  all compounds that have not  been  recovery corrected.   Report
results for blanks  as total  ng/sa-nple.

12,0  PRECISION AND ACCURACY

     Multiple  laboratory comparison studies  of the precision and accuracy
attainable  with this  technique are not available.   Limited precision  data
for environmental  samples are available; they derive from replicate analyses
of English  sole muscle  and liver tissue (reference  15).  Mean coefficients
of variation for  total  PCBs  were 11 percent  for muscle tissue (calculated
from six duplicates)  and 15  percent for liver tissue (calculated from one
duplicate  and  one triplicate).  These PCB  results  were generated by packed
colimn  analysis  and  comparisons of sample chromatograms to Aroclor standards.
The accuracy of these  PCB analyses was not  assessed.  Validation data using
the PCB quantification  procedure in this 301(h) document  has not yet  been
generated.

     A method  test was recently  conducted with  spiked blanks.  Replicate
blanks were spiked  with  known amounts of labeled  and unlabeled compounds.
                                  1-42

-------
The  amounts of the  unlabelec  co~ipojnds were calculates jsinc  tie  isc-tooe

dilution technique (i.e.,  the  calculatec  amounts of the unlabeled  compouncs
were  adjusted for  the  recovery  of the labeled  compounds).   The  ratio  of

the calculated amount of the  unlabeleci compounds  relative  to  tneir actual
spiked  amount (expressed  as  percent) is given in Table 1-3.   The  procedure

for this method test included  two  additional cleanup steps (metallic mercury
and  reverse phase  column chromatography) that are not part of  this  301 (h)

tissue protocol.   Therefore results  for  the 301(h) tissue procedure should

be comparable or  superior  to  those of the method test.


13.0  REFERENCES
 1.  "Performance Tests  for  the  Evaluation of Computerized Gas Chronatography/-
     Mass Spectrcmetry Equipment  anci  Laboratories," USEPA,  EMSL/Cincinnati,  OH
     45268,  EPA-600/4-80-025  (April 1980).

 2.  U.S. Environmental  Protection Agency.  1984 (revised  January,  1985).
     U.S.  EPA Contract  Laboratory  Program - statement of work  for organics
     analysis,  multi-media,  multi-concentration .   IFB  WA 85-J176,  J177,
     J178.

 3.  Ped.  Register,  Vol.  49,  No. 209, October 26, 1984, pp. 43416-43429.

 4.  "Carcinogens -  Working  with Carcinogens," DREW, PHS, CDC, NIOSH, Publica-
     tion 77-206 {Aug  1977).

 5.  "OSHA  Safety and  Health  Standards, General  Industry,"  OSHA  2206,  29 CFR
     1910  (revised Jan  1976).

 6.  "Safety in  Academic  Chemistry  Laboratories," ACS Publications, Committee
     on  Chemical Safety,  3rd  Edition  (1979).

 7.  Eichel berger,  J.W., I.E.  Harris, and W.L. Budde,  "Reference compound
     to  calibrate  ion  abundance  measurement  in  gas chromatography-mass
     spectrometry,"  Anal.  Chem.  Vol.  47, 1975, pp.  995-1000.

 8.  Duinker, J.C., M.T.J.  Hillebrand,  K.H.  Palmork, and  S. Wilhelmsen,
     "An evaluation  of  existing methods for quantitation  of PCBs  in environ-
     mental  samples  and  suggestions for an improved  method based on  measurement
     of  individual  components,"  Bull. Environm.  Contam.  Toxicol. Vol. 25,
     1980,  pp. 956-964.

 9.  AT ford-Stevens,  A.L.,  W.L.  Budde,  and  T.A.  Bellar, "Interlaboratory
     study on determination of PCBs in environmentally contaminated  sediments,"
     Anal.  Chem. Vol.  57,  1985,  pp. 2452-2457.


                                   1-43

-------
10.   Mullin,  M.D.,  C.M. PDchini,  S.  McCrindle,  K.  Ronkes, S.H. Safe, ana
     L.M. Safe, "High-resolution PCB analysis:   synthesis and chromatograpnic
     properties  of  all  209 PCB congeners,"  Environ.  Sc i . Techno!. Vol. 18,
     1984,  pp.  463-476.

11.   Bal1schmiter ,  K.,  and M.  Zell ,  "Analysis  of PCB by glass capillary
     gas chronatography," Fresenius Z.  Anal.  Chem.  Vol. 302, 1980, pp. 20-31.

12.   Webb, R.G.,  and A.C. McCall, "Quantitative  PCB  standards for electron
     capture  gas chromatography," J.  Chromatographic  Science, Vol.  11,
     1973,  pp.  366-373.

13.   Gebhart,  J.E.,  T.L.  Hayes, A.L.  Al ford-Stevens, and W.L. Budde, "Mass
     spectrometric determination of polychlorinated biphenyls as  isomer
     groups,"  Anal. Chem. Vol.  57, 1985, pp.  2458-2463.

14.   Sonchik,  S, ,  D. Madeleine,  P.  Macek, and  J.  Longbottom, "Evaluation
     of sample  preparation  techniques  for  the analysis  of  PCBs in  oil,"
     J.  Chronatograpnic  Science, Vol.  22,  1984, pp.  265-271.

15.   Tetra  Tech,  Inc.  Comm en cement Bay nearshore/tideflats remedial  investi-
     gation.  Vol. 1.   Final  report prepared for  the Washington State Department
     of Ecology and U.S. Environmental  Protection  Agency.  1985.
                                  1-44

-------
I
*»
tn
               50 -
            01
            O
            D.
            at
            Ut 1.0
            oc
            UJ
UJ
QC
               0.1 -
                                       I
                                      10
                    5     10     20

              CONCENTRATION (Mg/mL)
                                        I
                                        50
             ADAPTED FROM REFERENCE 3
                                                              1.
                                                                                                AREA AT
                                                                                                AREA AT
                                                                                                  M,IZ
                                                                                                  ADAPT ID H«iM HI f!«( Nil 3
                                                                              (3A)
                                                                 M,/Z
                                                                 M./Z
                                                                              (30)
                                                                              M,IZ
                                                                              M,IZ
                                                                              (3C)
                                                                              M,/Z
                                                                              M,/Z
                                                                                    AREA = 46100

                                                                                    AREA = 4780
                                                                                    AREA = 43600


                                                                                    AREA = 2650
AREA = 48300


AREA = 49200
                                                                                      ADAPT (0 FHOM BIFIKINf! 3
                                                                                                                       3.
                 Figure  1-1.   Relative response calibration  curve.
                 Figure  1-2.   Extracted  ion current profiles for chromatographically resolved labeled
                               (m?/z) and  unlabeled (m,/z)  pairs.
                 Figure  1-3.   Extracted  ion current profiles for (3A) unlabeled compound,  (3tt) labeled
                               compound,  and (3C) equal mixture of unlabeled  and labeled  compounds.

-------
                                -25 gm HOMOGENIZED
                                  T.SSUE SAMPLE
                                  S?IK= 10 -j.c EACH
                                     ISOTOPE
                                   MIX SAMP_E
                                   WITH NiA
                                 SOXHLET EXTRACT

                                       I
                                  WASH EXTRACT
                                     ATPH<2

                               ORGANIC   AQUEOUS
                                                    I
                                           ADJUST pH>i2
                                                    I
                                           SOLVENT EXTRACT
                                                	I
                                       I
                                   DRV EXTRACT
                               CONCENTRATE TO 3 ml
                                      GPC
                                       I
                              CONCENTRATE TO 2.5 ml
                                  80V.
                                                  20%
                                  CONCENTRATE
                                   ADD GGVS
                                INTERNAL STANDARD
                                   INJECT ON
                                     GC/MS
                                                                  RESIDUE WEIGHT
                                                                   DETERM.'NATlON
    EXCHANGE
 AND CONCENTRATE
                                                                  ALUMIMA COLUMN
                                                                     CLEANUP
                                                                   CONCENTRATE
   ADDGC.ECD
INTERNAL STANDARD
FOR ISOTOPE DILUTION TECHNIQUE • RECOMMENDED BUT NOT REQUIRED
                                                                    INJECT ON
                                                                     GC/ECD
    Figure  1-4.   Flow  chart for sample preparation.
                                   1-46

-------
GAS CHROMATOGRAPHY OF EXTRACTA3LE  COMPOUNDS
Compcy-c
2,2' -cifluoroDicnenyi (DFB)
N-nitrosc5;Tetny 1 an me
pnef>o" -^5>* '
pr.enoi
ais(2-cnloroethyl letner-ag
Dis(Z-cnloroetny 1 ) ether
1 ,3-dichlorobenzene-C4
1 ,3-dichlorooenzene
l,4-dicnlorcoenzene-d4
1,4-oicnlorobenzene
1 ,2-dlcPi1. oroDenzene-o<;
1 , 2-dichlorocenzene
bis(2-ch!oroisopropyl )ether-di2
bis(2-chloroisopropylj ether
nexacnloroethar.e-i3c.(''
hexachloroetnane
N-nitrosocn-n-propylamine
nitrooenzene-Oj
nitrooenzene
isophorone- Og
isophorone
2,4-dimEtny 1 phenol-d3
2, 4 -dimethyl phenol
bis (2-cnloroethoxy) methane
1.2,4-trichlcrooenzene-d3
1 , 2,4-tr ichlorosenzene
naphtha lene-dg
naphthalene . ,
hexach' aroDutadiene- ^4
hexacn: orooutacnene
hexacnlorocycl open tad iene-^c^
hexacnl orocyc 1 open taa iene
2-ch1oronaDhtnalene-C7
2-chloronapnthaiene
biphenyl-aiQ
biphenyl
acenaprthylene-ds
acenaphtnylene
d imethy 1 phthal ate-Q4
dimethylphthalate
2,6-dinUrotoluene-d3
2,6-dinurotoiuene
acenapntnene-diQ
acenaphthene
dibenzofuran-dg
dibenzofuran
fluorene-oiQ
f luorene
4-chlorophenylphenyl ether-ds
4-chlorophenylphenyl ether
diethyl phthal ate-d4
diethyl pntnalate
2, 4-din itrotol uene-d3
2,4-dinitrotoluene
1,2-diphenylhydrazine-da
1,2-dipnenylhydrazine^ '
Re tent
Sec
1163
385
596
700
696
704
722
724
737
740
758
760
788
799
819
823
830
845
849
881
889
921
924
939
955
958
963
967
1005
1006
1147
1142
1185
1200
1205
1211
1245
1247
1269
1273
1283
1300
1298
1304
1331
1335
1395
1401
1406
1409
1409
1414
1344
1359
1433
1439
Quant i tatior
Ion Tir.e (Prirr.aryj
Sei4:''
-------
  "A3LH
                                  1437
179
C tpner-vl ar. me
N-nitrosoc * o^ery i am m e- j*
N-nurcsod iD^er.y'i a/nine ' •
4-brorcDieryi pneny", efer
nexacnlcroDenzene-'^CK
nexacnlorobenzene
pnenantnrene-c- G
Dr.enan^nrene
antnracene-a^g
anthracene
oioenzothiopnene-dg
dibenzothiopnene
carbazole
di-n-outyl phthalate-d4
di-n-Dutyl pntna'ate
f luoranthene-ain.
f luoranthene
pyrene-cio
pyrene
benzidine-dg
benzicme
butylbenzyl phthalate
cnrysene-d^2
cnrysene
ben20(a)ar,thracene-di2
benzo(a)anthracene
3 , 3 ' - (5 i c r, 1 or ober. 2 1 d i n e- dg
3,3' -dichlorooenzicine
bis(2-ethylhexyl)phthalate-d4
bis(2-etnylnexyl] phthalate
di-n-octyl phthalate-d4
di-n-octyl pntnalate
benzo(b) fluoranthene-di2
benzo(b)f1uoranthene
benzo(ic)fluoranthene-di2
benzo(K') f Igor an then e
benzo( a) pyrene-di2
benzo(a) pyrer.e
benzo(g,n,i)perylene-di2
benzo(g,h,i jperylene
indeno U, 2, 3-c,d) pyrene
d ibenzo{ a, h) anthracene
2-chlorophenol-d4
2-cnlorophenol
2-nitrophenol-d4
2-nitrophenol
2,4-dichlorophenol-d3
2,4-dichlorophenol
4 -chloro-3 -methyl phenol -03
4-chloro-3-methylpnenol
2,4,6-trichlorophenol-d2
2,4,6-trichlorophenol
2,4,5-tr ichloropnenol
2,3,6-trichloropnenol
2,4-dinitrophenol-d3
2,4-dinitropnenol
4-nitrophenol-d4
4-nitrophenol
2-methyl-4,6-dinitropheno1-d2
2-methyl-4,6-dinitrophenol
pehtachloropnenol-13Cg
ptntachl orophenol
1435
1438
1439
1493
1521
1522
1573
15£3
1538
1592
1559
1564
1650
1719
1723
1613
1817
1844
1852
1854
1855
2060
20Si
2083
2082
2090
2086
2088
2123
2124
2239
Z240
2281
2286
2287
2293
2350
2352
2741
2750
2650
2660
701
705
898
900
944
947
1086
1091
1162
1165
1170
1195
1323
1325
1349
1354
1433
1435
1559
1561
. £JO
.226
.CC1
.283
.3C8
.OCO
.357
.OC3
.365
.003
1.340
1.0C3
1.419
1.473
1.002
1.559
1 . 002
1.556
1.004
1.594
1.000
1.771
1.789
1.001
1.790
1.004
1.794
1.001
1.825
1.000
1.9Z5
1.000
1.961
1.002
1.966
1.003
2.021
1.001
2.357
1.003
0.967
0.970
0.603
1.006
0.772
1.002
0.812
1.003
0.934
1.005
0.999
1.003
1.007
1.028
1.138
1.002
1.160
1.004
1.232
1.001
1.340
1.001
A t 7
175
169
248
292
2S4
188
173
133
173
192
184
167
153
149
212
202
212
202
192
134
149
240
223
240
228
253
252
153
149
153
149
264
252
264
252
264
252
288
276
276
278
132
128
143
139
167
162
109
107
200
196
196
196
187
184
143
139
200
198
272
266
I
1
';
i
1
i
1
;
1
1
1
1
3
1
1
1
1
1
1
5
5
1
1
1
1
1
5
5
1
1
1
1
1
1
1
1
1
1
2
2
2
2
1
1
2
2
1
1
2
2
2
2
2
2
20
20
6
6
13
13
5
5
:?;

NN?
4-3PPE

6CLBNZ

PHENANTHRN

ANTHRACENE

DIBNZTHIO
CARBAZOLE

OINBP

FLUQRANTHN

PYRENE

3ZID
BJTBNZPHT

CHRYSE'tE

SAA

33-2CLBZID

B2ETHXPHTH

2NOCTP

BBF

BK.F

BAP

BGHIP
INDENO-PYR
OB An A

2-CLPHK

2-NPHN

24-2CLPHN

4-CL2-MPHN

246-3CLPHN

236-3CLPHN

24-2NPHN

4-NPHN

46-2NOCRES

5CLPHN
i:"-is-4

S6-20-5
1C1 -55 -5

i:3-74-i

so -C 1-3

120-12-?

132-75-0
86-74-2

S4-74-2

206-44-0

129-00-0

92 -67 -5
55-66-7

213-01-9

56-55-3

91-94-1

117-61-7

117-84-0

205-99-2

207-08-9

50-32-8

191-24-2
193-39-5
53-70-3

95-57-8

88-75-5

120-83-2

59-50-7

88-06-2

93-37-55

51-28-5

100-02-7

534-52-1

87-86-5
a Relative retention  times for labeled compounds are referenced to DF8.   Relative retention
for unlabeled compounds are referenced to their  labeled analogs or to the most  chemically similar
most closely eluting  labeled compounds if labeled  analogs are not listed.

                                          1-48

-------
TA3LE
          -i
{Continued;

C3T50 jrc
2ecsf'uC'=3e'>zc3"e'1cre ;DP3P>
tciis-.e'e nit^e
ircc^or 1242 (PCS flixtjre)
Arccior 1254 (PCS mixture)
Arccior 1260 (PCS mixture)
a-HCH
C -HCH
y-nCH
5-HCH
a i d r : n
'••eotachlor
"estacr'ar eoox :de
y-cn!orsar.e
a-»ncosui fsn
a-cfilorcare
c-eicrin
4,4'-OOE
C -end os u If an
enar in
e<-drin aldehyde
4,4'-OOD
enoosulfan sulfate
y-Cf.lordene
4.4'-OOT
2,3,7,6-TCDD'6'
Additional 301 (h) Pesticides:
Oemeton
Gutnicn
Malatnion
Parathioi
Methcxycnlor
Mire*
Relative
Rj? ten 1 1 on
Time
to DF3
0.736
1.2-1.9



1.32
1.36
1.41
1.43
1.64
1.70
1.83
1.85
1.88
1.91
1.98
2.00
Z.02
2.02
2.07
2.10
2.13
2.13
2.17
2.01



1.19
1.19
1.51
1.52

*•' Z
.
231, 233



183, 181
183. 181
183, 181
183, 161
263, 265
100, 272
353, 255
373, 375
195, 207
373, 375
241, 263
246, 24S
207, 195
263, 277

235. 165
272, 387
336, 333
235, 237
320, 322



127, 99, 174
291, 109, 139
238, 227, 274
272, 237, 274

SC/WS
.
10
:s
10
10
2
3
2
3
2
2
2
2
2
2
2
2
2
2

2
5
2
2
3








SC .' E - D
5
100
100
100
100
20
5
5
20
5
5
15
10
5
5
5
30
5
5
30
30
30
10.-
.(5)
40





50
100

O:ES

TCXAPii;SE
?C = S
PCBS
PCBS
6CI-CHX-A
6CL-CKX-3
LISDANE
6CL-CHX-D
-L3RIN
HE°TACHLDR
"ttPCb. E?CX

ENDOSULFAN

DIELOR.'N
ODE

ENDRIN
ENDRIN-iLD
ODD
ENDOSLFN-S

DOT
DIOXIN

SYSTOX
GUTHSON
MALATHION
PARATHION
METHOXYCL
MlREX

CASSS

5:O35-2
53469-21-?
11C97-69-1
11096-B2-5
319-34-6
319-34-7
319-86-3
5S-39-9
3:9-00-2
76-44-S
1C24-57-3

115-29-7

60-57-1
72-55-9
115-29-7
72-20-8
'421-93-4
72-54-8
1031 -07-8

50-29-3
1746-C1-6

3065-48-3
86-50-0
121-75-5
56-38-2
72-43-5
2385-85-5
1 Deuterium labeled recovery  (surrogate) standard;  isotopically labeled  surrogates oo not have
ODES CDCes.

2 *JC-1abelea  recovery (surrogate)  standard; isocopically  laoeled surrogates do  not have ODES cooes.

^ Detected  as  azobenzene.

  Detected  as  diphenylanine.

^ Low  level  amounts (<2 ng) of  DDT  are denydrohalogenated and converted to DOE at variable rates
en tne GC system.

6 Acceptable  detection  limits will  be attainable  with the  U.S.  EPA Contract Laboratory Program
Oioxin Analysis procedure [SoU/S«diment Matru, Multi-Concentration, Selectefl  Ion Monitoring (SiK)
GC/MS Analysis; 9/15/83].

Column:  30 m x 0.25 mm i.d., 94S methyl, 41 phenyl,  IS vinyl bonded  phase  fused silica capillary
(J4« 08-5,  or  equivalent)

Tefflperature program (GC/HS):   5 «in at 30° C;  30-280° C at 8° C per  min; isothermal at 280° C until
benzoig,  h,i)perylene elutes.

Carrier gas  linear velocity:  30 OT/sec, helium.
                                                 1-49

-------
TABLE 1-2.  DFTP? MASS-INTENSITY SPECIFICATIC


Mass       Intensity  Recuireo

   51       30-60fo of  mass  198
   68       Less  than  2'=  of mass  69
   70       Less  than  2;=  of mass  69
 127       40-60% of  mass  193
 197       Less  than  1°;  of mass  193
 193       Base  peak,  100% relative  abuncance
 199       5-9^  of mass  198
 275       10-3C;o of  mass  19S
 565       1:,  of  mass  193
 441       Less  than  mess  443
 442       Greater tr.an  40;»  of mass  193
 443       17-23% of mass  442
                   1-50

-------
TABLE 1-3.  PRECISION AND ACCURACY OF METHOD BLANKS
Percent Recoverya
EPA Priority Pollutants
Phenols
phenol
2, 4-dimethyl phenol
2-chlorophenol
2,4-dichlorophenol
4-chl oro-3-methyl phenol
2,4,6-trichlorophenol
pentachlorophenol
2-nitrophenol
4-nitrophenol
2,4-dinitrophenol
4, 6-dinitro-2-methyl phenol
Aromatic Hydrocarbons
naphthalene
acenaphthene
acenaphthylene
fluorene
phenanthrene
anthracene
fluoranthene
pyrene
benz(a)anthracene
chrysene
benzo ( b) f 1 uoranthene
benzo ( k) f 1 uoranthene
benzo (a) pyrene
indeno(l, 2, 3-cd) pyrene
dibenzo( a, h) anthracene
benzo (ghi)perylene
Chlorinated Hydrocarbons
1,2-dichlorobenzene
1,3-dichlorobenzene
1,4-dichlorobenzene
1,2,4-trichlorobenzene
2-chloronaphthalene
hexachlorobenzene
hexachloroethane
hexachlorobutadiene
hexachlorocyclopentadiene
Blank
1

96
92
100
98
99
98
110
96
110
99
110

110
120
110
120
120
120
120
120
120
120
100
110
130
160
150
120

100
87
110
93
110
160
73
98
23
Blank
2

102
120
102
101
97
100
110
99
110
100
100

110
110
110
120
120
120
120
110
110
120
102
113
125
161
189
120

110
120
120
98
110
110
104
105
25
Blank
3

91
110
110
100
110
140
120
120
110
98
110

120
120
120
130
120
110
130
150
95
150
110
120
130
170
180
120

no
94
120
130
120
120
69
130
24
Blank
4

96
97
104
110
100
100
110
110
110
110
no

120
120
120
120
130
130
130
130
110
130
110
110
130
160
190
120

no
120
130
100
120
no
77
110
25
Mean

96
105
104
102
102
110
113
106
no
102
108

115
118
115
123
123
120
125
129
109
130
106
113
129
163
177
120

108
105
120
105
115
125
81
111
24
Coeff. of
Variation

4.7
5.0
4.2
5.2
5,7
18.6
4.4
10.3
0.0
5.5
4.7

5.0
4.3
5.0
4.1
4.1
6.8
4.6
13.4
9.5
10.9
5.0
4.2
1.9
3.0
10.6
0.0

4.7
16.4
6.8
15.9
5.0
19.0
19.6
12.4
3.9
                       1-51

-------
TABLE 1-3.  (Continued)
Phthalates

bis(2-ethy1hexyl)phtha1ate   120     130     120      130     125     4.6
benzylbutylphthalateb        270     298     120      310     250    35.2
di-n-butylphthalate          120     120     130      120     123     4.1
di-n-octylphthalate          120     120     120      120     120     0.0
dlethylphthalate             120     120     130      120     123     4.1
dimethylphthalate            110     120     140      120     123    10.3

Halogenated Ethers

bis(2-ch1oroethy1)ether       91      93      91      100      94     4.6
bis(2-ch1oroisopropy1)ether   c       c       c        61     —     —
bis(2-ch1oroethoxy)methane   200     110     190      140     160    26.5
4-chlorophenylphenylether    120     120     140      120     1Z5     8.0
4-bromophenylphenylether     150     140     120      150     140    10.1

Qrganonitrogen Compounds

nitrobenzene                  37      21      19       c       19    38.4
N-nitrosodipropylamine        26     135     150       73      96    59.7
N-nitrosodimethylamine       110     120     110      120     115     5.0
N-nitrosodiphenylamine        76      89     173       77     104    44.9
2,4-dinitrotoluene           120     120     120      100     115     8.7
2,6-dinitrotoluene           130     74      130      100     109    24.9
benzidine                      0000
3,3'-dichlorobenzidine       120     170     140      120     138    17.2
1,2-diphenylhydrazine        120     110     140      100     118    14.5

Miscellaneous

isophorone                   130     120      72      120     111    23.6


a   Metnod blanks were processed after spiking with  known amounts of unlabeled
and labeled compounds.  Recovery-corrected concentration of unlabeled priority
"poTlutants was  calculated  using the  recovery  of  labeled  analogs for each
compound.  The final percent recovery for the unlabeled  compounds  was then
computed as  the  ratio of the calculated  concentration to the known spike
level of each expound.

b    Benzyl butyl phthalate results are anomolously high because of laboratory
contamination  traced to  mercury used  in the sulfur-removal step  of the
procedure.  The contamination was subsequently eliminated.

c   Spectral  interferences precluded quantification.
                                  1-52

-------
              SECTION II
ANALYSIS OF VOLATILE ORGANIC COMPOUNDS
    IN  ESTUARINE  AND MARINE  TISSUES

-------
                                 CONTENTS





                                                                       Pace



 1.0   SCOPE AND APPLICATION                                           Ii-1



 2.0   SUMMARY OF METHOD                                               11-2



 3.0   INTERFERENCES                                                   II-3



 4.0   SAFETY                                                          11-5



 5.0   APPARATUS AND EQUIPMENT                                         11-6



 6.0   REAGENTS AND CONSUMABLE MATERIALS                                51-9



 7.0   SAMPLE COLLECTION, PREPARATION,  AND  STORAGE                     II-12



 8.0   CALIBRATION AND STANDARDIZATION                                 11-14



 9.0   QUALITY CONTROL                                                11-18



10.0   PROCEDURE                                                      11-20



11.0   QUANTITATIVE DETERMINATION (CALCULATIONS)                       11-26



12.0   PRECISION AND ACCURACY                                         11-27



13.0   REFERENCES                                                     11-28

-------
                   ANALYSIS OF VOLATILE ORGANIC COMPOUNDS
                      IN ESTUARINE  AND  MARINE  TISSUES
i.O  SCOPE AND APPLICATION

1.1  This method is designed  to  determine  the volatile toxic organic pollutants
associated with Clean Water Act  Section  3Cl(h)  regulation  [40 CFR 125.53(k)
and (v)] and additional  compounds amenable to purge and trap gas chronatography-
mass spectrometry (GC/MS)  (Table II-l).

1.2  The chemical  compounds listed  in  Table II-l can be determined  in biological
tissue samples collected  from estuarine  and marine  environments by  this
method.

1.3  The detection  limit of this method  is usually  dependent on the level
of interferences rather  than  instrumental  limitations.

     Lower  limits of detection (LLD)  are established  by analysts based
on their experience  with  the  instrumentation and  with  interferences  in
the sample matrix being  analyzed.  LLD are greater than instrumental  detection
limits because they take into account sample  interferences.  To estimate
LLD,  the  noise  level  should be determined in the retention window for the
quantitation mass of representative analytes.   These determinations  should
be made for at  least  three  field  samples  in the  sample set under analysis.
The signal  required to  exceed the average  noise level  by at least a  factor
of two should then be estimated.  This signal is the minimum response required
to identify a potential signal for  quantification.  The LLD is  the concentration
corresponding  to  the  level  of this signal based on calibrated response
factors.  Based on best  professional judgment, this LLD would then be applied
to samples in  the  set  with  comparable  or  lower interference.  Samples with
much higher interferences  (e.g., at least  a  factor  of two higher)  should
be assigned LLD at a  multiple of the  original LLD.
                                   II-l

-------
     These LLD values may  be  less than the rigorously  defined method detection
 limits specified in the revised  "Guidelines Establishing Test Procedures
 for  the  Analysis  of Pollutants"  (40 CFR Part 136,  10/26/84).  This latter
 procedure requires the analysis  of  seven replicate samples and a statistical
 determination  of  th emethod detection  limit with  99  percent confidence.
 Data quantified between the  LLD and  the  rigorous method detection  limit
 are valid and useful  in environmental  investigations of low-level  contamination,
 but have a lower  statistical confidence associated with them  than  data
 quantified above the  method detection limit.

     The LLD are roughly 5-10 ppb (wet weight)  with the exception of acrolein
 and acrylonitrile, which have not been thoroughly  tested  on  tissue matrices
 with this method.

 2.0  SUMMARY OF METHOD

 2.1  Volatile organic  compounds  are vacuum  extracted  from  a macerated,
 5-g (wet wt) tissue  sample  and  concentrated  in a cryogenic  trap  cooled
with liquid nitrogen  (references 1 and  2).   The cryogenic  trap is then
 transferred  to a conventional purge-and-trap device.   The  extract is diluted
 to a 5  mL volume with water and  treated  as an aqueous  sample.  In the purge-
 and-trap device, the  volatile organic  compounds are purged  from the  aqueous
 phase into a gaseous  phase with  an  inert carrier gas.  The  volatile compounds
are passed into a  sorbent  column  and trapped.  After purging is completed,
 the  trap is back'flushed  and heated rapidly  to  desorb  the compounds into
 a gas chrcmatograph (GC).  The compounds are separated by GC and detected
with a  mass  spectrometer (MS).

     Analysis is carried out  by  GC/MS  either according to  the isotope dilution
 technique (U.S.  EPA Method 1624  Revision  B; reference  3) or  U.S. EPA  Method
624 (reference 4).  Both of these methods  were  developed  for water/wastewater
sample  matrices.   The isotope dilution  technique, which requires  spiking
the sample with a  mixture  of  stable isotope labeled analogs  of the analytes,
is the  technique of choice because  it  provides  reliable  recovery data  for
                                   11-2

-------
each analyte.  Method  624  requires spiking samples with  only three surrogate
compounds and does not allow  for  recovery corrections.   If uniformly  high
recoveries can be attained  with Method 624, then addition of numerous labeled
compounds (Method 1624B)  and  recovery corrections are  unnecessary.   However,
until such performance can  be demonstrated, Method 1624B provides a detailed
and valuable assessment of  analytical performance.

     Hiatt  (reference 5)  proposed  another  vacuum distillation procedure
that did not include a purge-and-trap device.   In this  technique, volatile
organic  compounds are transferred directly from a cryogenically cooled
trap to a fused-silica capillary  column for GC/MS analysis.  This capillary
column technique  allows for optimum resolution and rapid conditioning between
samples.   However, the performance of the technique has  not been thoroughly
tested  (reference 6).  A  potential  problem is  that  water  can enter the
capillary column  and cause  chromatographic problems or  freeze, effectively
plugging the column.   Thus, Hiatt1 s original  procedure  (references 1 and 2),
which has been tested  more  thoroughly, is recommended  here.

     Vacuum  distillation  is recommended rather than direct  purge-and-trap
extraction (i.e., without vacuum  distillation)  because the  former technique
has been  demonstrated to  allow for  better  recoveries of spiked compounds
than the latter technique (reference 1; comparisons based on similar spiking
level s).

2.2  Laboratories may use alternative analytical  procedures provided  that
evidence of  performance comparable to the recommended  procedure is provided.

3.0  INTERFERENCES

3.1  Impurities  in the  purge  gas, organic  compounds out-gassing from the
plumbing  upstream of the trap, and solvent vapors in the laboratory  account
for the majority  of contamination problems.  The analytical system is demon-
strated to be  free from interferences  under conditions  of the analysis
by analyzing blanks  initially  and with each sample lot  (samples analyzed
                                  II-3

-------
on the same 8 h shift),  as described in Sect.  9.3.   Common laboratory solvents
(e.g., methylene chloride) are often contaminants  in volatiles analyses.

     3.1.1  Newly packed  traps should be conditioned overnight at 1700-180° C
by backflushing with  an  inert gas  at a flow rate of 20-30  ml/min.   Traps
must be conditioned daily for a minimum of 10  min  before use.

3.2  There is potential  for ambient contamination of samples and extracts
when using vacuum and  cryogenic concentration techniques.   Care must  be
taken  to  eliminate any leaks  in  the  vacuum extraction and concentration
device.  A critical source  of potential  contamination  is  pump oil vapor
and exhaust from the vacuum pump; this should  not  be a problem if the  system
is properly sealed.   A  cold  trap  is  placed  between the vacuum pump and
concentration trap to  prevent condensation of  pump oil vapors in the  concen-
tration trap (Figure  II-l).   All materials in the vacuum  extraction and
concentration device that  contact the  sample and its vapors must  be made
of stainless steel and/or borosilicate glass.  All connections and seals
must be free of elastomers or grease that either outgas or allow penetration
of ambient contaminant vapors.

3.3  Samples can be contaminated by diffusion of  volatile organic compounds
(particularly methylene  chloride) through the  bottle seal during shipment
and storage.   A field blank prepared from reagent water and  carried  through
the sampling and  handling protocol  serves as a check on such  contamination.

3.4  Contamination by  carry-over can occur  when  high level and low level
samples are analyzed sequentially.   When an unusually contaminated  sample
is analyzed, it should  be  followed  by  analysis of a reagent water blank
to check  for carry-over.  Because the transfer  lines, trap, and other parts
of the system can retain  contaminants and interferences, frequent  bakeout
and purging of the entire system may be required.
                                  II-4

-------
4.0  SAFETY

4.1  The  toxicity or  care inogen ic ity of  each  compound or reagent  used  in
this method has not been  precisely determined; however, each chemical  compound
should be treated  as  a  potential health hazard.   Exposure to these  compounds
should be reduced to the  lowest possible level.   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 data handling  sheets should also  be made available to  all  personnel
involved in these  analyses.  Additional  information on  laboratory safety
can be found in  references 7-9.

4.2  The  following  compounds  covered  by this method have been tentatively
classified as known or  suspected human or  mammalian carcinogens:   benzene,
carbon  tetracnl oride, chloroform, and  vinyl chloride.  Primary  standards
of these toxic  compounds should  be prepared in a hood,  and a  NIQSH/MESA-
approved  toxic gas  respirator should  be worn when high concentrations are
handled.

4.3  The  following safety measures  must be  employed when handling  cryogenic
and vacuum systems:

     4.3.1   Liquid nitrogen (LN2) must  not be allowed  to  contact flesh
since it will  cause extreme frostbite  and  deaden  (kill)  tissues.

     4.3.2  The concentrator and  cold traps  must never  be  closed off or
sealed after allowing  any concentration of liquid air.  The liquid  air
will vaporize, resulting  in tremendous pressure build up and  explosive
damage to  the vacuum system.   Always  vent any  vessel  imred iately after
removing  the cryogenic or LN2  bath.   Wear  safety goggles when  working with
cryogenic  and vacuum  systems.
                                  II-5

-------
5.0  APPARATUS AND EQUIPMENT

5.1  Sample Handling  Equipment

     5.1.1   Stainless steel  spatula, rinsed  wi th methanol  and  oven-dried
at 1500 c.

     5.1.2   Sample vessel  -  Pyrex  flask with 15 mm 0-ring connector,  washed
with detergent and rinsed  with distilled water and oven-dried  at 4500  C.

     5.1.3   0-ring,  Buna  N,  sonicated with 50 percent methanol/water  then
dried by vacuum at 600  c.

     5.1.4   Tissue homogenizer  (e.g., Tekmar Tissuemizer, Tekmar Co.,  Cin-
cinnati, OH)  - must be  free of volatiles and solvents before use.

5.2  Apparatus for  Vacuum Distillation and Cryogenic Concentration  {Figure
II-l).

     5.2.1  Vacuum pump, capable  of achieving 10-3 Torr and 25 L/min.

     5.2.2   Vaccum/pressure  gauge - with a range of subatmospheric pressure
to 10 psi.

     5.2.3   Concentrator  trap  or purge flask, 25 ml capacity (Tekmar  Part
No. 14-0957-024 or equivalent) modified with 9 mm 0-ring connectors.

     5.2.4   Cold  trap  - glass trap (easily produced by glassblowing,  Figure
11-1} with 0-ring  fittings  (e.g., Kontes 671750-009).

     5.2.5   Transfer line,  1/4  in o.d. glass-lined stainless steel  tubing.
Lines should  be kept  as short as  possible to minimize sample carryover.

     5.2.6  Vacuum valves,  Nupro  B-4BKT or equivalent.
                                   11-6

-------
     5.2.7  Dewar  flasks,  665 ml or 1,000 nL, for liquid  nitrogen D2th.

     5.2.8  Assorted  compression fittings ancj graphite ferrules (Figure II-l).

     5.2.9  Ultrasonic  bath, Branisonic 12 or equivalent.

     5.2.10   Heater  tape - to  wrap around stainless steel  lines and valve
bodies to maintain  a  temperature of 60° C.

     5.2.11  Pinch  clasps, Thomas - to secure 0-ring  connections.

5.3  Purge-ana-Trap Device - capable  of meeting  specifications  listed in
U.S. EPA Method  1624  B  (see below).  Complete devices consisting of a purging
device (the concentrator trap,  Sect.  5.2.3), a  Tenax/silica  trap, and a
desorber  are commercially available (e.g., Tekmar Model  LSC-2, Tekmar  Co.,
Cincinnati, OH).

     5.3.1  Trap -  25 to 30 cm x 2.5 mm i.d. minimum, containing the following:

          5.3.1.1  Methyl  silicons  packing - one  +_0.2  cm,  3  percent  OV-1
on 60/80 mesh Chromosorb W, or equivalent.

          5.3.1.2  Porous  polymer -  15  +1.0 on, Tenax  GC  (2,6-diphenylene
oxide polymer),  60/80 mesh, chrcmatographic grade, or equivalent.

          5.3.1.3  Silica  gel  - 8 +1.0  cm, Davison  Chemical, 35/60 mesh,
grade 15, or equivalent.

     5.3.2   Desorber -  should heat the trap to 175 +5° C  in  45  sec or less.
The  polymer section  of  the trap should not exceed 180° C, and the remaining
sections should  not exceed 220° C.

     5.3.3   Commercial  purge and  trap devices  are easily  coupled to GC
systems.
                                  n-7

-------
5.4  GC/KS (Gas Chromatograph-Mass Spectrometer)  System.

     5.4.1   GC  -  should be  linearly temperature  programmable with initial
and final temperature  holds.

     5.4.2   GC  column - 6 ft long x 0.1 in i.d.  {stainless steel or glass)
packed with 1 percent  SP-1000 on Carbopak B, 60/80  mesh  or equivalent.

     5.4.3   MS  -  70 eV electron  impact ionization; capable of repeatedly
scanning from 20 to 250  amu every 2 to 3 sec.

     5.4.4   GC/MS interface - GC to MS interfaces constructed of all-glass
or glass-lined materials are recommended.  Glass  can be deactivated  by
silanizing with dichloro-dimethyl silane.

5.5  Data System - should collect and record MS  data, store mass  intensity
data  in  spectral  libraries, process  GC/MS data  and generate reports, and
should calculate  and record response factors.

     5.5.1  Data  acquisition - mass spectra  should  be collected continuously
throughout the analysis  and stored on a mass storage device.

     5.5.2   Mass  spectral  libraries - user created libraries containing
mass spectra obtained from analysis of authentic  standards  should be employed
to reverse search  GC/MS  runs for the compounds of interest.

     5.5.3   Data  processing -  the  data system  should be used to search,
locate, identify,  and quantify the  compounds of  interest in each  GC/MS
analysis.   Software  routines should be employed  to  compute retention times
and extracted ion  current profile (EICP) areas.   Displays  of spectra,  mass
chromatograms, and library comparisons are required  to verify results.

     5.5.4   Response  factors and multipoint calibrations  -- the data system
should be used to  record  and maintain lists  of response factors  (response
ratios  for isotope dilution) and generate  multi-point  calibration curves.
                                  II-8

-------
Computations  of relative  standard deviation  (coefficient of  variation)
are useful  for  testing calibration linearity.

5.6  Other  Materials

     5.6.1   Syringe,  10 uL _* 1 percent  of  volume.

     5.6.2   Syringe,  50 uL +_ 1 percent  of  volume.

     5.6.3   Syringe,  5 ml +_ 1 percent of volume, gas-tight with  shut-off.

     5.6.4   Bubble  flowneter.

6.0  REAGENTS AND CONSUMABLE MATERIALS

6.1  Reagent Water

     6.1.1  Reagent  water is defined  as  water free of interferences  (i.e.,
interferents are not  observed at the detection limits of  the  compounds
of interest).

     6.1.2  Prepare  water by boiling  1  L of freshly distilled  water down
to 900 mL and  transferring  the water  to  a  l-l  volumetric flask that has
been modified  by replacing the ground  glass joint with a 15-mm  i.d.,  Buna-N
0-ring connector.

     6.1.3  Connect  the  flask to  the distillation apparatus at  the  sample
chamber site and evacuate for 15 min while continuously agitating  the flask
in an ultrasonic cleaner.

     6.1.4  After  evacuation, release an  inert gas (^2 or He  can  be used)
into the flask  until  equilibrium is obtained,  then seal  with a cap made
from a Buna-N 0-ring  connector.

6.2  Methanol -  pesticide quality or equivalent.
                                  II-9

-------
6.3  Stancard Solutions  -  purchased  as solutions or mixtures with certification
as to  their purity, concentration,  and  authenticity, or  prepared  from materials
of known purity and composition.   If compound purity  is 96 percent or greater,
the weight may be used  without correction to calculate the  concentration
of the standard.

6.4   Preparation  of  Stock Solutions  -  prepare in  methanol  using  liquid
or gaseous standards per the  steps below.  Observe  the safety  precautions
given in Sect. 4.

     6.4.1   Place approximately 9.8 ml of methanol  in a 10  ml ground glass
stoppered volumetric flask.   Allow the flask to stand  unstoppered  for  approxi-
mately  10 min or  until all  methanol-wetted surfaces have  dried.   In  each
case, weigh the flask,  inmediately add the compound,  then immediately reweigh
to prevent evaporation  losses from affecting the measurement.

          6.4.1.1   Liquids - using  a  100 uL syringe, permit  two  drops of
liquid to fall into the  methanol without contacting  the neck  of the flask.
Alternatively,  inject  a known volume  of the  compound into  the  methanol
in the flask using a micro-syringe.   With the exception of 2-chloroethylvinyl
ether,  stock standards of compounds  that boil above room  temperature are
generally stable  for at  least 4  wk when stored at 4° C.

          6.4.1.2   Gases ( chloromethane, bromomethane, chloroethane, vinyl
chloride) - fill  a valved  5 mL gas-tight syringe with the compound.  Lower
the needle to approximately 5 mm above the methanol meniscus.   Slowly  introduce
the compound  above the surface of  the meniscus.   The gas  will  dissolve
rapidly in the methanol.

     6.4.2  Fill  the flask to volume, stopper, then  mix by  inverting several
times.   Calculate the  concentration  in mg/mL  (ug/uL)  from the  weight gain
(or density if a  known  volume was injected).
                                   11-10

-------
     6,4.3  Transfer tne  stock  solution to a Teflon sealed  screw-cap  bottle.
Store, with minimal  heaaspace,  in  trie dark at -10 to -20° C.

     6.4.4  Prepare fresh  standards  weekly for the gases and 2-chloroethylvinyl
ether.  All other standards  are replaced after 1 mo, or sooner  if  comparison
with check standards indicates  a change in concentration of over 10 percent.
Quality control  check standards that can be used to determine  the  accuracy
of calibration standards  are available from the U.S. Environmental Protection
Agency, Environmental Monitoring and Support Laboratory, Cincinnati,  Ohio.

6.5   Labeled Compound  Spiking Solution  -  from  stock standard solutions
prepared as above,  or from mixtures, prepare the spiking solution  to  contain
a concentration such  that  a 5-10 uL spike into each  5 tnL sample  "extract",
blank,  or aqueous standard  analyzed will result in a concentration of 10 ng/mL
of each labeled  compound.  For  the gases and for the water soluble compounds
(acrolein, acrylonitrile) , a concentration of 50 ng/mL may be  used.   Include
the internal  standards  (Sect. 8.1.2) in this solution  so that a concentration
of 10 ng/mL in each  sample,  blank, or aqueous standard will be  produced.

6.6  Secondary Standards  - using stock solutions, prepare a secondary standard
in methanol to contain  each pollutant  at a concentration of 250  ug/mL.
For  the  gases and  water  soluble compounds (Sect.  6.5),  a concentration
of 1.25 mg/mL may be used.

6.7  Aqueous  Calibration  Standards -  the  concentrations  of calibration
solutions suggested  in this  section are intended to bracket  concentrations
that  will  be encountered  during sample  analysis that will  not overload
the analytical  system.  Use  sufficient  amounts  of  the secondary standard
(Sect.  6,6)  and reagent  water to produce concentrations of 5, 10,  20, 50,
and 100 ug/L  in  the aqueous calibration standards.  The  concentrations
of gases  and water  soluble compounds  will  be higher (i.e., 25, 50, 100,
250,  and 500 ug/L).  Analysts  may use a wider range of standard concentrations
if linearity  can  be  demonstrated.
                                  11-11

-------
6.8   Aqueous Performance Standard  -  an aqueous  stancard containing a" 1
pollutants, internal  standards,  labeled compounds, and BFB  (4-bromofl uoro-
benzene) Is prepared  daily,  and  analyzed  each shift to demonstrate  performance
(Sect.  11).   This standard  should  contain  either 10 or 50 ug/L of the  labeled
and  pollutant gases  and  water  solub'e compounds, 5 ug/L of BFB, and  10 ug/L
of all other pollutants,  labeled compounds,  and internal  standards.   It
may be the  nominal  10 ug/L aqueous calibration standard (Sect.  6.7).

6.9  A methanol ic standard containing all pollutants and internal  standards
is prepared to demonstrate recovery of these compounds when syringe injection
and purge-and-trap analyses  are compared.  This standard  should  contain
either  10  ug/mL or  50  ug/mL of the gases and water soluble compounds, and
10 ug/mL of the remaining pollutants and  internal  standards  (consistent
with the amounts in the  aqueous  performance standard in Sect. 6.8).

6.10  Other standards that may  be needed  are those for testing  of BFB per-
formance (Sect.  8.2.1)  and for  collecting mass spectra for storage  in  spectral
libraries (Sect.  8.1.4).

6.11  High  Purity Helium  - 99.999 percent.

6.12  Liquid Nitrogen (LN£).

7.0  SAMPLE COLLECTION,  PREPARATION. AND  STORAGE

7.1  In  the  field,  sources  of contamination include sampling  gear, grease
from ship winches or  cables,  ship engine exhaust, dust,  and  ice used for
cooling.  Efforts should be made to minimize handling and to avoid  sources
of contamination.  This  will  usually require that resection (i.e., surgical
removal) of tissue  be performed  in a controlled environment (e.g., a  labora-
tory).  For example to  avoid contamination from ice, the  samples  should
be wrapped  in aluminum  foil,  placed in watertight plastic  bags and  immediately
iced in a covered ice chest.  Aluminum  foil should be cleaned by  heating
at over  105° C before  use.   Solvent cleaning is  unacceptable unless  heating
is performed afterward.   Organisms should not be  frozen prior  to  resection
                                  11-12

-------
if analyses  will  only be  conducted on  selected  tissues, Because freezing
may cause internal organs to rupture  and  contaminate other  tissue (e.g.,
muscle).   If  organisms are  eviscerated  on  boara the survey  vessel,  the
remaining tissue may be wrapped as described  above  and frozen.

7.2  To  avoid  cross-contamination,  all  equipment used in sample handling
should be thoroughly cleaned before each sample  is  processed.   All  instruments
must be  of a  material that  can be  easily  cleaned (e.g.,  stainless  steel,
anodized aluminum, or borosilicate glass).  Before  the  next sample is processed,
instruments  should  be washed with  a detergent solution, rinsed with  tap
water,  soaked  in high-purity acetone  and  methylene chloride,  and finally
rinsed  with reagent water.

7.3  Resection  should be carried out by  or under  the supervision of a competent
biologist.   Each organism  should be handled  with clean  stainless steel
or quartz  instruments (except for external  surfaces).  The specimens  should
come  into contact with precleaned glass  surfaces only.  Polypropylene  and
polyethylene  surfaces are  a potential  source of contaminatin and  should
not be  used.   To control  contamination  when  resecting tissue, separate
sets of  utensil should be used for removing  outer  tissue and  for dissecting
tissue  for  analysis.  For fish samples,  special  care must be taken to avoid
contaminating  target tissues  (especial ly  muscle) with slime and/or adhering
sediment  from the fish  exterior  (skin)  during resection.   The incision
"troughs" are  subject to such contamination;  thus,  they should not be included
in the  sample.   In the case of muscle, a "core" of tissue  is  taken from
within the area boarded  by  the incision troughs, without contacting  them.
Unless  specifically sought  as a sample,  the dark muscle  tissue that may
exist  in  the  vicinity of  the lateral line  should not be  mixed with  the
light muscle  tissue that constitutes the rest of the muscle tissue mass.

7.4  The  resected  tissue sample  should  be placed in a clean glass  or  TFE
container which  has been washed with detergent,  rinsed twice with tap  water,
rinsed once  with  distilled  water, and heated  at  105° C for several  hours.
Jars should  be heated at  1050 c and  allowed to cool immediately  before
use.
                                  11-13

-------
7.5  The U.S.  EPA and other  federal  agencies  (e.g.,  National  Bureau  of
Standards)  have not  yet provided specific guidance  regarding  holding  times
and  temperatures for  tissue samples  to  be analyzed  for volatile organic
compounds.   Until U.S.  EPA develops  definitive guidance,  the  following
holding  conditions  should be observed.   Resected tissue  samples should
be maintained  at -200 C and analyzed as soon as  possible, but within 10 days
of sample receipt.  The 10 day holding time is based on  the Contract Laboratory
Program regulations  for sediments to be analyzed  for  voUtiles (reference 10).

8,0  CALIBRATION AND STANDARDIZATION

8.1  Initial  Calibration

     8.1.1   Calibration  by the  isotope dilution  technique -- the isotope
diTution technique is  used for the purgeable organic compounds when appropriate
labeled  compounds  are available and  when interferences  do  not preclude
the analysis.    If labeled  compounds  are  not available  or  interferences
are  present,  the internal  standard  technique (Sect. 8.1.2) is used.   A
calibration curve encompassing the concentration range  of  interest is prepared
for  each compound determined.  The relative response (RR) vs. concentration
(ug/L)  is plotted  or computed using a linear regression.   An example  of
a calibration curve  for  a pollutant  and  its  labeled analog is given  in
Figure  II-2.   Also shown are the _+10 percent error  limits  (dotted  1   es) .
Relative  response  is determined according to the  procedures described below.
A minimum of  five  data  points is required for calibration (Sect. 6.7).

         8.1.1.1  The  relative response (RR) of pollutant to  labeled compound
is determined  from  isotope ratio values  calculated  from  acquired  data.
Three isotope  ratios are used in this process:

     Rx = the  isotope ratio measured in the pure  pollutant (Figure II-3A)
     Ry = the  isotope ratio of pure labeled compound  (Figure  II-3B)
     Rfji = the  isotope ratio measured in the analytical mixture of the pollutant
         and  labeled compounds (Figure II-3C).
                                  11-14

-------
     The correct way to  calculate  RR  is:

                    RR = (Ry  -  Rm)(Rx + l)/(Rm - Rx)(Ry
If Rfn is not between  2Ry  and  0.5RX,  the method does not apply and the sample
is analyzed by the internal  standard  technique (Sect. 8.1.2).

          8.1.1.2  In most  cases,  the retention times  of the pollutant
and labeled compound  are  the  sane  and  isotope ratios (R's)  can be calculated
from the EICP areas,  where:

                         R =  (area at m^zj/farea  at m2/z)
If either of the areas  is  zero,  it  is assigned a value of one in the calcu-
lations; that is, if:   area  of mi/2=50,721, and area of m2/z=0, then R=50721/l =
50720.   The m/z's are  always  selected  such that Rx>Ry.  When there is a
difference in retention times (RT)  between the pollutant and labeled compounds,
special  precautions are  required  to determine the isotope ratios.

     Rx , Ry, and R^ are  defined  as  follows:
          Rx =  [area  mi/z  (at
          Ry =  l/[area m2/z  (at  RTi)]
          Km =  [area  m\/z  (at  RT2)]/[area m2/z (at RTi)]
          8.1.1.3  An  example of  the  above calculations can be taken from
the data plotted  in  Figure II-3  for a  pollutant and its  labeled analog.
For these data, Rx=168920/l=  168920,  Ry=l/60960=0. 00001640,  and Rnf% 868/82 508
= 1.174.   The RR for  the  above data  is then calculated using the equation
given  in  Sect. 8.1.1.1.   For the example, RR=1.174.  Note:   Not all  labeled
compounds el ute before  their  pollutant analogs.

          8.1.1.4  To  calibrate the analytical  system by isotope dilution,
analyze a 5 ml aliquot of each of the  aqueous calibration standards (Sect. 6.7}
                                   11-15

-------
spiked  with an appropriate constant amount  of  the  labeled compound spiking
solution  (Sect. 6.5),  using  the purge  and  trap procedure  in  Sect. 10.
Compute tne RR at  each concentration.

          8.1.1.5  Linearity - if the ratio  of  relative response  to concen-
tration for any compound is constant  (less than 20  percent coefficient
of variation) over  the  five  point calibration range, an averaged  relative
response/concentration  ratio may be  used  for  that compound; otherwise,
the  complete  calibration  curve for that compound should be used  over  the
5 point calibration range.

     8.1.2  Calibration by internal  standard  - used when criteria  for  isotope
dilution (Sect.  8.1.1} cannot be met.  The method  is applied  to pollutants
having  no  labeled analog  and  to the  labeled  compounds themselves.  The
internal standards used  for volatiles  analyses are  bromochloromethane,
2-bromo-l-chloropropane,  and  1,4-dichlorobutane.   Concentrations  of  the
labeled compounds and pollutants  without labeled analogs  are computed  relative
to the nearest eluted internal  standard.

          8.1.2.1  Response factors  - calibration  requires the determination
of response factors (RF), which are  defined by the following equation:

                         RF = (As x  Cis)/{A1s x  Cs)

where:

     As  =  the  EICP area  at  the characteristic m/z for the compound  in  the
          daily standard
    A-JS  =  the  EICP area at the  characteristic m/z  for the internal  standard
    Cis  =  the  concentration (ug/L) of the internal standard
     C$  -  the  concentration of the pollutant  in  the daily standard.

          8.1.2.2  The  response factor  is  determined  at  5, 10,  20,  50,
and 100 ug/L for the pollutants (optionally at  five  times these concentrations
for gases  and water soluble  pollutants - see  Sect. 6.6 and 6.7),  in a  way
                                  11-16

-------
analogous  to that  for  calibration by  isotope  dilution  (Sect,  £.1.1.4}.
The RF is plotted  against concentration  for  eacn  compound  in  tne  stancarc
(Cs) to produce  a  calibration curve.

          8.1.2.3   Linearity - if tne  response factor (RF)  for any compound
is constant (less  than 35 percent 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  should
be used over the five point range.

     8.1,3  Combined calibration - by adding the  isotopically labeled  compounds
and internal standards  (Sect.  6.5)  to  the  aqueous calibration  standards
(Sect.  6.7), a  single  set  of  analyses can be  used to produce calibration
curves for  the  isotope dilution  and internal standard methods.

     8,1.4  Mass  spectral  libraries -  detection and identification  of the
compound of interest during calibration  and  sample analysis  are dependent
upon the spectra stored  in user  created  libraries.

          8.1.4.1   Obtain  a mass spectrum of each pollutant and  labeled
compound and each  internal  standard by analyzing an authentic  standard
either  singly  or as part  of a mixture in which there is no interference
between closely  el uted components.  That  only a single compound is  present
is determined  by examination  of the spectrum.   Fragments  not attributable
to the compound under  study indicate the  presence of an interfering  conpound.
Adjust  the  analytical  conditions and   scan rate (for this  test  only)  to
produce an  undistorted spectrum at the  GC  peak  maximum.   An undistorted
spectrum  will  usually  be  obtained if  five complete spectra  are  collected
across the  upper  half of  the  GC  peak.   Software algorithms designed  to
"enhance"  the  spectrum may eliminate  distortion,  but may also  eliminate
authentic  ions or  introduce other distortion.

          8.1.4.2  Obtain the  authentic  reference spectrum  under BFB  tuning
conditions  {Table  II-2) to normalize  it  to spectra from other instruments.
                                  11-17

-------
          8.1.4.3   The  spectrum  is eaitec  by  saving me five nest  intense
mass spectral peaks  and  all other mass spectral  peaks greater tran 10  percent
of  the base peak.   This  spectrum is stereo  for  reverse search ana for  car.pcuna
confirrnation.

8.2  Ongoing Calibration

     8,2.1  The  8F3 stana5"d must  be analyzed at the  beginning  of  eacn
8-h shift.   The tuning  criteria in Table  11-2  must be  met before blenKs
and samples may be  analyzed.

     8.2.2  At  the beginning anc end  of each 8-h shift, system calibration
should be verified  by purging the aqueous performance standard  (Sect. 6.8).

          8.2.2.1   Calibration is tested  by  computing the  concentration
of unlebeled compounds  by the isotope  dilution technique  (Sect.  8.1.1)
for compounds with labeled analogs.   Concentrations of unlabeled  compounds
without labeled  analogs are calculated  according to the  internal  standard
technique (Sect. 8.1.2).

     A complete  (five-point) recalibration  should be performed when  results
vary from predicted concentrations by  more  than _+25 percent.  The  last
sample  analyzed before  failing criteria  should  then be reanalyzed.   If
the results differ  by more than +20 percent  (i.e., twice  the median repro-
ducibility for replicate  analysis of  tissue  samples, Table  II-3),  then
it is to be  assumed that the instrument was out of control during the original
analysis  and  the earlier  data should  be rejected.  Reanalysis of  samples
should progress in  reverse order until  it is  determined  that  there is _<20
percent difference  between initial  and  reanalysis results.

9.0  QUALITY CONTROL  [For further guidance, see Quality  Assurance/Quality
Control  (QA/QC) for 301(h)  Monitoring Programs:   Guidance  on  Field and
Laboratory  Methods  (Tetra Tech 1986).]

9.1  Each  laboratory that uses this method  is  required to operate a formal
quality assurance program.  The minimum requirements of this program  consist

                                  11-18

-------
of  an  initial cemor.stration of laboratory  Capability, analysis of  samples
spiked with labelec  conoouncs to evaluate  anc document data  quality, ana
analysis of stancaras  ana Dianas as tests  of  continues performance.

9.2  initial  Demonstration of Laooratory  Capaoility

     9.2.1  Analyze the aqueous performance stendara (Sect.  6.3)  according
to the purge-anc-trap procedure  in Sect.  10.   Compute the area  at  tne  primary
m/z  (Taole II-l) for  each compound.   Compare  these areas  to  those o&tairsec
by  injecting  one  uL  cf the  methanoiic  standarc (Sect. 6.9)  to ceterr.ine
compound  recovery.   The recovery should be  greater than 20  percent for
the water  soluble compounds (acrolein  and  acrylonitrile), and  60-110  percent
for  all  other compounds.   This recovery snculd be cemonstratec  initially
for each purge-and-trap GC/MS system.   The  test should be repeated only
if  tne  purge  and trap  or  GC/MS systems  are modified in any  way that  might
result in  a change  in  recovery.

9.3  Blanks

     9.3.1  Reagent water  blanks must  be analyzed to demonstrate  freeaorn
from carry-over  (Sect.  3) and contamination.

          9.3.1.1   The  level  at which  the purge-and-trap  system will  carry
greater  than  5 ug/L of a  pollutant of  interest into a succeeding blank
should  be  determined  by analyzing  successively  larger  concentrations of
these compounds.  When  a  sample contains this concentration  or more,  a
blank should  be analyzed  immediately  following this sample  to  demonstrate
no carry-over  at  the 5  ug/L level.

          9.3.1.2   With each sample lot  (samples analyzed on the same 8-h
shift),  a  blank  should  be analyzed  immediately  after analysis of  the  aqueous
performance standard (Sect.  8.2.2)  to  demonstrate freedom from  contamination.
If any of  the  compounds of interest,  except common laboratory  contaminants
(e.g., methyiene chloride  and toluene), or  any  potentially  interfering
compound  is found  in a blank at  greater than  10 ug/L (assuming  a response
                                  11-19

-------
factor of 1 relative to the nearest  eluted  internal stancara for canpouncs
not listed in Table II-l), analysis  of samples  is halted  until  the source
of contamination is eliminated anc  a clank  snows  no  evicence of contaT;inaticr-.
at this level.   This control  action   also  applies if rr.ethylene cnlorice
or toluene "is cetectec in a blank at greater  than 50 ug/L.

9.4  Sample Spiking

     9.4.1   The laboratory snould  spike all samples 'with labeled cornpouncs
to assess metnod performance on  the  sample matrix.

     9.4.2   Spike anc analyze each  sample  according to the method beginning
in Sect. 10.

     9.4.3   Compute the  percent recovery (P) of  the labeled compounds  using
the internal  standard technique  (Sect.  6.1.2).

9.5  Replicates

     9.5.1   Replicate analyses (i.e., analyses of two subsamples from the
same tissue homogenate) must be  performed to  monitor laboratory precision.

     9.5.2   At  least one laboratory duplicate should be run for cases  of up
to 20 samples.   For cases of over 20 samples, one blind triplicate and  addi-
tional  duplicates must be run for a  minimum of 5 percent replication  overall.

10.0  PROCEDURE

10.1  Sample Processing

     10.1.1  Mince tissue  sample  with a  scalpel and homogenize the sample
to a uniform consistency  with a  micro-grinder.  Care must be taken to ensure
that the  micro-grinder  is  thoroughly cleaned after each use.  This usually
entails disassembly of the unit.  Devices with  large surface areas (e.g.,
blenders,  meat  grinders)  should  not  be used,  as they are difficult to clean
                                   11-20

-------
and a. small  sample  Is difficult to remove after grinding.   Liquia associated
with the sample  should be retained throughout the  procedure.

     10.1.2  Dry  weight determination -  if  sample size permits and dry-
weight concentrations  are  required, dry weight determinations may be performed
as fellows:   transfer an aliquot of approximately  3  g (weighed to the  nearest
0.1 g)  to  a  preweighed dish.  Allow the sample to  dry  in  an oven at 1050 r
overnight and determine the solid  residue weight  to  the nearest  0.1  g.
The percent  total  solids is calculated as:

              Ts =  [dry residue wt (g)]/[wet sample  wt  (g)]

Dry weight determinations  should not be made at the cost of  having insufficient
sample for volatiles analysis.  Significant decreases in  the size of samples
used for extraction will decrease attainable detection limits.

     10.1.3   Immediately after homogenization,  use a  stainless steel  spatula
to transfer  a 5-g aliquot  to a preweighed sample  vessel (Sect. 5.1.2).
Weigh the transferred portion to the nearest 0.1 g.

     10.1.4  Spike  50  nanograms of each labeled  compound  (or 250 nanograms
of gaseous and water soluble compounds)  into 2 ml of reagent water  and
add to  the  sample matrix.  Seal the sample vessel  with  an 0-ring connector
and clamp and sonicate  for 10 min.   After sonication, store the  sample
contained  in the sample vessel overnight in a refrigerator/freezer  and
analyze  the next day.

10.2  Vaccim  Distillation and Concentration (Reference 2)

     10.2.1  The  vacuum extractor  must be airtight and free of moisture
before an extraction can be started.

     10.2.2   A clean 100 mL pyrex flask is connected  to the vacuum distillation
apparatus at  the sample vessel site  (see Figure  II-l), the vacuum pump
started,  and V2-V4 opened  to evacuate  the apparatus.   Line condensation
                                  11-21

-------
 is  prevented by warming  the  transfer lines while  evacuating the systerr.
 Heating tape is  effective  in creating  even transfer  line temperatures  and
 can be used continuously during the  procedure.

     10.2.3  The  vacuum apparatus is  pressurized  with helium by closing
 ₯3  and  opening  Vi.  The  apparatus is tested for leaks  with a helium leak
 detector or Snoop  ,  and  appropriate  adjustments  are made as necessary.
 When  the  apparatus has  been  found to be airtight,  close Vj, open V3 and
 then  heat the  transfer  lines and  concentrator  trap to 1000 C for 5 min
 to  eliminate any residual contamination.

     10.2.4  The  flask  containing the sample should  be  immersed in liquid
 nitrogen,  before  the flask is uncapped.  To begin  the  distillation,  close
 V;?  (with V3 and  V4  remaining open),  cool the concentrator  trap with a liquid
 nitrogen bath, and  replace the empty sample vessel with the cooled  sample
 flask.  Disconnect the vacuum  source by closing ₯3.   Open  V2 to permit
 vapors from the sample  vessel to reach the  concentrator trap.   Immerse
 the sample vessel  in a 500 C water bath and sonicate  for 5 min.

     10.2.5  Connect  the vacuum source  to  the  sample  vessel  by  opening
 V3«  The  lower  pressure  hastens the transfer of volatile compounds from
 the sample to the cooled concentrator  trap.  After  15 min of vacuum,  close
 V3  and  open  Vi  to fill  the  system with helium  to  atmospheric pressure.
 Close Vi and  V2  to  isolate the condensate.  The distillation is now completed
 and the condensate  is ready for transfer to a  purge-and-trap device.   The
 condensate can  be held  in  the  liquid nitrogen bath for  up  to 1 h prior
 to analysis.  Care  should be taken to  ensure that moisture does not  freeze
 in  the  narrow glass tubing  in  the concentrator trap.  Careful drying of
 the system prior  to analysis  and maintenance of  an  airtight  system  will
 preclude this problem.

     10.2.6  Disconnect the sample concentrator trap  from  the vacuum apparatus
 and connect it to the purge-and-trap device.   Some outgassing  is observed
when the  sample condensate  is  melted;  therefore, the  condensate should
be kept  frozen until the concentrator  trap is attached tc  the purge-and-
                                 11-22

-------
trap device.   After attacr.nent, warm the concentrator  trap walls to loosen
the concensate and  allow the ring  of  ice formed during condensation  to
crop to  the  bottom of the trap.  To tnis partially melted extract add 3  ~L
of reagent  water  containing 50 ng of each of the  internal  standards  (bromo-
cnloromethane, Z-bromo-l-chloropropane,  and 1,4-dichlorobutane).  The internal
stancards are  added  after vacuum extraction to  allow the analyst tne  assess
analytical  losses of labeled compounds  during  the  extraction/ concentration
procedure.

10.3  Purge-and-Trap Procedure

     10.3.1   Because  commercial  purge flasks  must  be  slightly  Codified
(with 0-ring  fittings) to be attached to the vacuum distillation apparatus,
a sinple 0-ring  adapter  is  necessary  to connect the  purge flask to the
commercial  device for  which it was  designed.  The  modified purge  flask
(Sect.  5.2.3) used in this  procedure  can be  fitted  to a commercial purge-
and-trap device (e.g., a Tekmar ALS  interfaced  with  a  Tekmar LSC-2)  with
9 mm 0-ring  fittings fused  to short sections of  glass tubing.  Commercial
purge-and-trap devices are almost entirely automated and are easy to operate
with manufacturer's  instructions.

     10.3.2   Purge  the extract solution with  the concentrator trap immersed
in an  ice-water  bath for 5 min followed by immersion in  a 550 C-water bath
for an  additional 7 min.  This provides  conditions  for reproducibly melting
the frozen  extracts  in order to obtain reproducible purging efficiencies.

     10.3.3 The  GC  conditions for analysis are as  follows:

               Injector zone temp.               2250 c
               Initial GC oven temp.               60° C
               Final GC temp.                    1750 c
               Initial hold time                 3 min
               Ramp  rate                         80 C/min
               Final hold time                   24 min
               Jet separator oven temp.           2250 c
                                  11-23

-------
 10.4   Qualitative  Determination - accompl isned  by comparison of data
 analysis of a sample or  blank with data from  analysis of the shift standard
 (Sect. 8.2.2).   Identification is confirmed  when spectra  and retention
 times  agree according to  the criteria below.

     10.4.1  Labeled compounds and pollutants having no labeled analog:

          10.4.1.1   The  signals for  all  characteristic  masses stored  in
 the spectral  library (Sect. 8.1.4.3) should be present  and  should maximize
 within the same two  consecutive scans.

          10.4.1.2   Either 1) the background corrected EICP areas or 2)  the
 corrected relative intensities of the  mass spectral peaks at the GC peak
maximum  should  agree within a factor of two  (0.5  to 2 times) for all  masses
 stored in the library.

          10.4.1,3  The retention time relative  to  the nearest  eluted  internal
 standard should  be within +7  scans  or +_20 sec, whichever is greater,  of
 this difference in the shift standard.

     10.4.2  Pollutants having a labeled analog:

          10.4.2.1   The  signals for  all  characteristic  masses stored  in
the spectral  library should be present  and  should maximi ze  within the same
two consecutive  scans.

          10.4.2.2   Either 1) the background  corrected EICP areas or 2}  the
corrected relative intensities of the  mass spectral peaks at the GC peak
maximum  should agree within  a factor of two  for all masses stored  in  the
spectral  library.

          10.4.2.3   The  retention  time difference between  the pollutant
and its labeled  analog should agree  within +2 scans or +6 sec, whichever
is greater, of this  difference in the shift standard.
                                  11-24

-------
          10.4.2.4  If  the  experimental  mass  spectrum contains masses  that
are not present  in the reference spectrum,  an  experienced  spectronetri st
is to determine  the  presence or absence of the  compound.

10.5  Tentatively  Identified Compounds  (GC/MS Analysis) - The ten non-target
peaks  of  greatest area  in  the RIC (reconstructed ion chromatogram)  should
be identified  ana  quantified, if possible.

     10.5.1  Guidelines  for making  tentative  identification (reference 10):

        (1)   Tentative  identifications  should be based on a forward
     search of  the  EPA/NIH mass  spectral  library.   Sample spectra
     should  be visually compared with the  most  similar library match.

        (2)  Relative intensities of major ions  in the reference spectrun
     (ions greater than  10 percent  of  the most  abundant ion)  should
     be present  in the sample spectrum.

        (3)   The relative  intensities of the major ions should agree
     within +_20  percent.  (Example:   For  an  ion with  an  abundance
     of 50 percent  in  the standard spectra,  the corresponding sample
     ion  abundance must be between 30 and  70  percent.)

        (4)   Molecular  ions  present  in  reference spectrum should  be
     present  in  sample spectrum.

        (5)   Ions present  in  the  reference spectrum but  not in  the
     sample  spectrum should be reviewed for possible  subtraction from
     the  sample  spectrum because of background contamination or co-eluting
     compounds.   Data system library reduction  programs can  sometimes
     create  these discrepancies.

          10.5.1.1  If,  in  the opinion  of  the mass  spectral  specialist,
no valid tentative identification can be  made,  the  compound should be  reported
                                  11-25

-------
as unknown.   The mass spectral  specialist should give additional  classification
of the unknown  compound  if  possible (e.g., unknown  hydrocarbon,  unknown
aromatic  compound, unknown chlorinated compound).   If  probable molecular
weights can  be  distinguished, include them.

     10.5.2  Tentative  quantification  -  quantification of  TIOs  will be
based on the internal standard  technique and  an  assumed response  factor
of one  (in  the absence of  data  from  authentic standards).   The uncertain
nature of this  quantification should be clearly noted  in the data report.

11.0  QUANTITATIVE  DETERMINATION (CALCULATIONS)

11.1  Isotope Dilution - by  adding a known amount  of a  labeled compound
to every  sample  prior  to vacuum  distillation, correction for recovery of
the pollutant can be made because the  pollutant and its labeled  analog
exhibit similar behavior during  purging,  desorption,  and gas chromatography.
Note that  pollutants and their labeled analogs  are not always retained
identically by complex matrices, so  their behavior during the extraction
step may differ.   Use of this  technique is  to enable correction for analytical
losses after extraction, not for matrix recovery.

     11.1.1  Relative  response (RR)  values  for sample mixtures are used
in conjunction  with calibration  curves described in Sect.  8.1.1 to determine
concentrations directly, so long as  labeled  compound  spiking levels are
constant.

     11.1.2   For the isotope dilution technique, concentration  is calculated
as follows:

     C {ug/kg,  wet  wt tissue) =

     CA (ug/kg) x RR x n

           RRi  * ZAi
                                  11-26

-------
where
      CA =  the concentration  of the stable  isotope  labeled compound  as
           spiked  into the sample
      RR =  relative response  of unlabeled pollutant to isotope  labeled
           surrogate in the sample
      Ri =  relative response at itn  point in calibration
      Z-j =  absolute amount of unlabeled compound at i^n point of calibration
         =  absolute amount of labeled  compound at i^1 point in calibration
       n =  number  of calibration points.

11.2   Internal  Standard -  calculate the  concentration  using the response
factor determined  from calibration data (Sect.  8.1.2)  and  the  following
equation:
     Concentration  =  (As  x CiS)/(Ais * RF) where the  terms are as defined
in Sect. 8.1.2.1, except  that  C-js is in ug/kg (wet tissue) and As is the
EICP area at  the characteristic m/z  for the analyte in the  sample.

11.3  If the  EICP area at the  quantitation mass for  any compound exceeds
the calibration range of the  system, a smaller  sample aliquot should  be
analyzed if  possible.  However,  sample sizes of  less than 0.5 g are not
recommended because such small  samples may not be representative.

11.4  Report  results for  all  pollutants and  labeled  compounds found in
samples,  in  ug/kg (wet weight,  unless dry  weight  is  required)  to  three
significant  figures.  Pollutants  and labeled compounds in blanks should
be reported  in ng/sample.

12.0  PRECISION AND ACCURACY

12.1  Recoveries from replicate spiked  water  and tissue  analyses  are presented
in Table  II-3 (references  1  and 2).  These analyses  were not performed
with the  isotope dilution technique  and recovery results are  uncorrected.
                                  11-27

-------
13.0  REFERENCES


 1.  Hiatt, M.H.,  "Analysis  of  Fish and Sediment  for Volatile  Priority
     Pollutants,"   Anal. Chem. Vol. 53, 1981,  pp.  1541-1543.

 2.  Hiatt, M.H.,  and T.L.  Jones.   Isolation of Purgeable Organics from
     Solid  Matrices by  Vacuum Distillation.   U.S.  Environmental  Protection
     Agency,  Region IX, Las Vegas Laboratory,  1984.

 3.  Fed. Register,  Volume 49, No.  209,  October  26,  1984, pp. 43407-43415.

 4.  Fed. Register, Volume 49, No. 209, October  26, 1984, pp. 43373-43384.

 5.  Hiatt, M.H.,  "Determination of Volatile Organic Compounds  in Fish
     Samples by Vacuum  Distillation and Fused  Silica Capillary Gas Chrorna-
     tography/Mass  Spectrometry," Anal. Chem.  Vol. 55,  1983, pp.  506-516.

 6.  Hiatt, M.H.  4 November 1985.  Personal  Communication (phone by Mr. Harry
     Beller).  Analytical Technologies, Incorporated,  National City,  CA.

 7.  "Working with  Carcinogens," DHEW, PHS,  NIOSH, Publication 77-206  (1977).

 8.  "OSHA Safety  and Health  Standards, General  Industry," 29 CFR 1910,
     OSHA 2206, (1976).

 9.  "Safety in Academic Chemistry Laboratories," American Chemical  Society
     Publication, Corrniittee on Chemical Safety (1979).

10.  U.S. Environmental  Protection  Agency.   1984 (revised January,  1985).
     U.S. EPA Contract  Laboratory Program  -  statement  of  work for organ ics
     analysis, multi-media, multi-concentration.   IFB WA 85-J176, J177,
     J178.
                                  11-28

-------
                        VACUUM
                         GAUGE
                V4" S.S.
             TEE UNION
i
INJ
10
                                                       25mL
                                                      PURGE
                                                      FLASK
                                             {CONCENTRATOR
                                                       TRAP)
NUPRO B-4BKT
                                                                         LIQUID
                                                                         NITROGEN
                                                                         BATH
                                                                                                                VACUUM
                                                                                                                PUMP
                                                                                                COLD TRAP
          LIQUID
          NITROGEN
          BATH
                                                                                                        AMI'11 II I HUM Kill HI Nil ,'

                                                                                   NOTf: PURtU Aim IHAP OlVICf IS NOT IfKllinin IH I II.Illll
                  Figure II-l.  Apparatus  for vacuum  distillation and cryogenic  concentration.

-------
                10 -
I
CJ
o
           01
           U>
           Z
           O
           GL
           CO
           IU
           DC
           Ul
Ul
a:
               1.0 -
               01 —
                                    I
                                   to
                             I
                             20
I
SO
                            CONCENTRATION
         ADAPTED FROM R[F{RtNCE 3
T
100
                                                             2.
                                                                              (A)
                                                                                                AREA 168920
                                                                                                        M./Z
                                                                                                        M,/Z
                                                                   (B)
                                                                                      AREA 60960
                                                                                                        M ,/Z
                                                         M,/Z
(C)
 M ,/Z 96868

 M?/Z 82508

	M../Z

      M,/Z
                                                       AOWUD FROM RfFIRCHCt J
                    Figure M-2.   Relative response  calibration  curve.
                    Figure 11-3.   Extracted ion current profiles  for (A)  the unlabeled pollutant,  (B)  the
                                   labeled analog,  and  (C) a mixture  of the labeled  and the unlabeled  compounds.

-------
                                       TABLE  II-l.   VOLATILE  ORGANIC ANALYTES
I
CO

Analyte
Acrolein
Acrylonitrile
Benzene
Bromod ichl oromethane
Bromoform
Bromomethane
Carbon tetrachloride
Chlorobenzene
Chloroethane
2-chloroethylvinyl ether
Chloroform
Chi oromethane
Dibromochl oromethane
1,1-dichloroethane
1,2-dichloroethane
1,1-dichloroethene
trans-1 ,2-dichloroethene
1,2-dichl oropropane
cis-lt3-dichloropropene
trans-1 ,3-dichloropropene
Ethyl benzene
Methylene chloride
1,1,2,2-tetrachloroethane
Tetrachloroethene
Toluene
1 ,1,1-trichloroethane
1,1,2-trichloroethane
Trichloroethene
Vinyl chloride
CASRN
107-02-8
107-13-1
71-43-2
75-27-4
75-25-2
74-83-9
56-23-5
108-90-7
75-00-3
110-75-8
67-66-1
74-87-3
124-48-1
75-34-3
107-06-2
75-35-4
156-60-5
78-87-5
10061-01-5
10061-02-6
100-41-4
75-09-2
79-34-5
127-18-4
108-88-3
71-55-6
79-00-5
79-01-6
75-01-4
ODES
ACROLEIN
ACRYLNTRLE
BENZENE
2CLBRMETHA
BROMOFORM
METHYLBR
CARBON TET
CLBNZ
ETHYL CL
2-CLEVE
CHLOROFORM
METHYL CL
2BRCLMETH
11-2CLETH
12-2CLETH
11-2CLETHE
12-2CLETHE
12-2CLPRP
C13-2CLPRE
T13-2CLPRP
ETHYLBENZ
METHYLE CL
4CLETHAN
4CLETHE
TOLUENE
111-3CLETH
112-3CLE1H
3C LETHE
VINYL CL
Quantitation
Ion (m/z)
56
53
78
83
173
94
117
112
64
63
83
50
129
63
62
96
96
63
75
75
106
84
83
164
92
97
97
130
62
Secondary
Ion(s)
55
51, 52
--
85, 129
171, 175
96
119, 121
114
66
65, 106
85
52
206, 208, 127
65, 83
64, 98
61, 98
61, 98
65, 114
77
77
91
86
85, 168
129, 131, 166
91
99, 117, 119
83, 85, 99
95, 97, 132
64

-------
fASLE  II-2.   3FB  MASS-INTENSITY  SPEC:
       Xass          Intensity  Required

        50           15-40%  of  mass 95
        75           30-60%  of  mass 95
        95           Base  peak,
                    100%  relative abundance
        96           5-9%  of mass 95
       173           <2% of nass 174
       174           >50S  of mass 95
       175           5-9%  of mass 174
       176           >95%  but <101S of mass 174
       177           5-9%  of mass 176.
                 11-32

-------
             TABLE  II-3.   PERCENT  SPIKE  RECOVERIES  FOR  VOLATILE
               PRIORITY  POLLUTANTS USING VACUUM  DISTILLATIONS
Spiking Compound
 Average Percent
Recovery (Water)b
 Average Percent
Recovery (Tissue)0
Chloromethane
Sromomethane
Vinyl chloride
Chloroethane
Methylene chloride
1 , 1-dichloroethene
1 ,1-dichloroethane
trans-1 ,2-dichloroethene
Chi oroform
1,2-dichloroethane
1,1,1-trichloroethane
Carbon tetrachloride
Acryl oni tr il e
Bromod ic hi or om ethane
1,2-dichloropropane
J.ran_s-l,3-dichloropropene
Tnchloroethene
Benzene
Dibromochloromethane
1 , 1, 2 -tri chloroethane
cis-1 ,3-dichloropropene
Bromofom
Tetrachl oroetnene
1,1,2,2-tetrachloroetnane
Tol uene
Chlorobenzene
Ethyl benzene
2-chloroethyl vinyl ether
Acrolein
Average compound recovery
105 + 22
110 + 23
83 + 12
103 + 16
126 + 22
98+5
96 + 5
98+5
93+8
98 + 10
104 + 9
102 + 10
85 + 13
108 + 10
104 + 7
109 + 9
105 + 9
106 + 7
102 + 11
95+8
109 + 9
104 + 14
105 + 9
90+9
106 + 7
101 + 7
103 + 5
94 + 50
113 1 76
102 i 8
85 + 22
126 + 75
64 + 11
69 + 22
LCd
74+8
90+6
86+9
107 + 31
92+5
92+8
91+9
NAe
64 + 11
54+7
52 + 9
65 + 11
57 + 10
56+9
66 + 7
54 9
NDf
NO
61 + 10
NO
64 + 15
ND
NA
76 +_ 20
a From references 1 and 2.
b Reagent water  was spiked  with  25  ug/L of each compound except acrolein
and  acrylonitrile, which were added  at 100 ug/L.  The recoveries are averaged
from 9 analyses and were calculated by comparing vacuum  extracted determinations
to determinations for which spikes  were added directly  to  a  purge-and-trap
device.
c Ten-gram fish samples were spiked at 25 ppb.  The recoveries were averaged
from 12 analyses.
d Laboratory contamination of fish  prevented the generation of valid data.
e Compound was not added to this matrix.
f Not determined.
                                  11-33

-------
           SECTION III
ANALYSIS OF METALS AND METALLOIDS
 IN ESTUARINE AND MARINE TISSUES

-------
                                 CONTENTS





                                                                       Page




 1.0   SCOPE AND APPLICATION                                          III-l



 2.0   SUMMARY OF METHOD                                              III-l



 3.0   DEFINITIONS                                                    III-2



 4.0   INTERFERENCES                                                  III-2



 5.0   SAFETY                                                         III-4



 6.0   APPARATUS AND EQUIPMENT                                        III-4



 7.0   REAGENTS AND CONSUMABLE MATERIALS                              III-6



 8.0   SAMPLE COLLECTION, PREPARATION,  AND STORAGE                     II1-7



 9.0   CALIBRATION AND STANDARDIZATION                                 III-9



10.0   QUALITY CONTROL                                               III-ll



11.0   PROCEDURE                                                     111-18



12.0   CALCULATIONS                                                  111-21



13.0   PRECISION AND ACCURACY                                        111-21



14.0   REFERENCES                                                    III-Z1

-------
                   ANALYSIS  OF  METALS AND METALLOIDS IN
                       ESTUARINE AND MARINE TISSUES
1.0  SCOPE AND APPLICATION

1.1  This  metnod is designed  to determine  antimony, arsenic, beryllium,
catfrnin, chromium,  copper, lead, mercury,  nickel,  selenium, silver, thallium,
and  zinc  concentrations  in  biological  tissue  samples.  The method  may  be
used for the analysis  of varying tissue  types such as  edible muscle and
livers of estuarine  and marine organisms.

1.2  A universal  wet oxidation (acid digestion)  procedure  is  recommended
that is capable of providing a clean extract  suitable for analysis by  atomic
absorption spectrophotmetry (AAS).  This  digestion has  proven effective
when determining  the priority pollutant metals  listed above {with the possible
exception of beryllium  and  thallium)  (e.g.,  Table III-2).   Because of  a
lack of  reference materials  certified  for beryllium and thallium,  little
is known regarding method suitability for  these elements.

1.3  Limits of detection  (LOD)  are listed  in Table III-l.   These may vary
depending on the element being measured,  method of detection, and  instrumental
sensitivity.

2.0  SUMMARY OF METHOD

2.1  A macerated 5-g sample  of  tissue  is homogenized wet, subsampled and
digested using a  wet oxidation method.  The  resulting extract  is analyzed
for  the metals of interest using various  atomic absorption (AA) techniques
such as:
                                 III-l

-------
          direct aspiration (DFAA)    =   for  higher concentration metals
          graphite furnace (GFAA)     =   for  lower concentration metals
          hydriae generation (HYDAA)  =   for  hydride forming elements (antimony,
                                        arsenic , selenium)
          cold vapor (CVAA)          =   for  mercury.

2.2  Al ternative methods of  detection  (e.g.,  inductively  coupled  plasma
emission spectrometry)  may be used  providing  their performance and limitations
have been establ ished.

3.0  DEFINITIONS

     Certified  Reference  Materials  (CRM):  A homogeneous sample that has
been analyzed a sufficient number of  times by  numerous qualified laboratories.
The data ere compiled and certified values are determined through statistical
analysis.  A number of  CRM are commercially available  in  a  wide range  of
matrices for metals analyses  (e.g., NBS  Oyster Tissue,  SRM 1566) (Reference 1).

     Control Standard:   A solution, independent of the calibration standards
whose analyte concentration is  known.   These are often analyzed as an external
check after calibration.

     Limit  of  Detection  (LOD):  The LOD is the lowest concentration level
that can be determined   to  be  statistically different  from  a blank.  The
recommended value  for LOD is  3cr, where a  is the standard deviation of the
blank in replicate  analyses (reference 2).

     Matrix  Modifier:   A  reagent  added to  a sample that alters some aspect
of its  composition  (references  3-5).

4.0  INTERFERENCES

4.1  Interferences  should be  considered  to be  any chemical  or physical
phenomenon  that  can influence  the accuracy of the data during an analytical
                                  III-2

-------
operation.   These  can have eitner  a  positive or  a  negative effect  on  the
result depending on their nature.

4.2  Contamination of the sample  caT occur during  any  stage of collection,
handling,  storage, or  analysis.   Potential contaminant  sources must be
known  and  steps  should be  taken  to minimize or  eliminate them.  Some of
the most commor) sources of  contamination  include  prolonged  exposure of
the tissue  to metal-containing  fumes and dust; insufficiently clean  sample
containers,  storage facilities  and  testing apparatus;  and the use of contam-
inated reagents during analysis (reference 6).

     In general, clean laboratory  procedures are  extremely important when
performing  trace metal analysis.

4.3  Most  instrumental methods are prone to matrix  interferences, which
can either  suppress or enhance  the  analyte signal.   If a matrix interference
is suspected, its  effect should be  determined and  corrective action  taken.
Some common  matrix interferences  are listed  below  along with  suggested
corrective measures (references 7,  8).

     4.3.1  High sample viscosity  - usually  due  to dissolved solids  and
high acid  content.   Match  the matrix of the  calibration standards with
the samples  where possible.

     4.3.2   Non-specific absorption (light scatter)  -  usually due to dissolved
solids or suspended  particulates, which absorb analyte  radiation.  Background
correction  (see  instrument manufacturer's instructions)  should be  used
whenever this  occurs.

4.4  Many  chemical interferences,  some of which are poorly understood,
can occur during  instrumental  analysis of  the sample  extracts.  A great
many  of these interferences  have been addressed  in the  literature  and in
most cases  a sample  pretreatment  or instrumental  modification  has been
proposed as  a  remedy  (e.g., reference 9).
                                 III-3

-------
5.0  SAFETY

     Laboratory  personnel should  be  well  versed  in  standard laboratory
safety practices.   It  is  the  responsibility  of all staff  and management
to ensure  that  safety training is mandatory.  The laboratory is responsible
for maintain ing a  current awareness  file  of OSHA regulations  regarding
the  safe  handling of the  chemicals specified  in this method.  A reference
file of data handling  sheets should also  be made available to all  personnel
involved  in these analyses.   Additional  information on laboratory safety
can be found in  references 10-12.

5.1  Chemicals  and reagents  should be  properly labelled and  stored  in an
area appropriate  to their properties.  Any  reagents whose  composition or
properties may  change with  time must be  dated  and  properly disposed of
on or before the  expiration date.

5.2  Areas  where  strong oxidizing  agents  and flammable or explosive materials
are used  should  be  well labeled and the necessary restrictions imposed.

5.3  Where laboratory apparatus and instrumentation are used, the manu-
facturer's  safety  precautions should be strictly followed.

6.0  APPARATUS AND  EQUIPMENT

6,1  Sample Containers  - wide-mouth screw  cap  jars made of  either glass
or non-contaminating plastic (linear or high density polyethylene  or equiva-
lent).  All  containers  should be pre-rinsed  with dilute acid and distilled
deionized  water  (DOW)  as  described in Sect.  10.6.

6.2  Dissection  Tools -  scalpels  should  be made of high-quality,  corrosion-
resistant  stainless steel, while tweezers and cutting surfaces  should be
plastic or teflon.  All  tools should be  thoroughly rinsed with DOW prior
to use and  between  samples.
                                 III-4

-------
6.3   Tissue Gr inder/Homogen i zer  -  (e.g.,  Tekmar  Ti ssuemizer, Tekrnar Co.,
Cincinnati, OH) a standard tissue  homogenizer can  be  used with ninor modifi-
cations.   If  the  apparatus  contains stainless  steel  parts, they snould
be replaced with tantalum or titanium.   Stainless  steel  blades used during
homogenization have been found to  be  a  source of nickel  and chromium contam-
ination .

6.4  Digestion Vessels - 125 ml borosilicate  glass Erlenmeyer flasks equipped
with all glass reflux caps (Tuttle covers).   Tuttle  covers  or equivalent
reflux caps are essential  for preventing  evaporative loss of volatile compounds
or elements during high temperature digestion.   They  are  commercially available
(Fisher Scientific)  or are easily  produced  from borosilicate test tubes.

6.5  Hot Plate - a thermostatically controlled plate  with  a range  of  75  to
4000 C.

6.6   Fume Hood -  a  properly  constructed  hood capable  of withstanding acid
fumes.  It must be equipped  with an exhaust  fan having  sufficient capacity
to remove all  fumes.

6.7  Atomic Absorption Spectrophotometer  (AAS).

     6.7.1  The AAS must have sufficient  sensitivity and stability to perform
within the specifications  required  by the method (Sect.  11).  The instrument
should  have automatic  background correction, direct  aspiration flame,  as
well  as flameless capabilities.   The  instrument  must  have a routine maintenance
program to ensure proper performance  and  trouble-free operation.  All source
la/nps should be handled  with care  and the exit windows kept free of dust and
fingerprints.  Periodic  intensity and stability  checks of the lamps should
be made.   Replace any lamps  showing signs of  deterioration (reference 13).

     6.7.2  A  graphite furnace (also called carbon  rod) attachment for
the AAS is recomierKJed  when  determining most  elements  in the low concentration
ranges.   Most, if not  all, AAS manufacturers offer  this equipment  as  an
                                  III-5

-------
accessory.  The stability  and  sensitivity afforded  by the furnace is typically
one to two orders of magnitude better than direct aspiration (reference  14).

     6.7.3  In addition  to  the graphite furnace, another flameless attachment
can be used in conjunction  with the  AAS  to determine  the  hydride-form ing
elements (arsenic,  antimony, and  selenium).  Most  such  attachments may
also be used to analyze  for mercury using the cold  vapor  technique.  These
methods  are preferable to the graphite  furnace since they vaporize the
analyte from the sample  matrix prior to detection.

7.0  REAGENTS  AND CONSUMABLE MATERIALS

     The  purity of all  reagents  used for  trace  metal  determinations is
extremely important.   Reagents should  be checked  for  purity prior to  use
to confirm the absence of contamination (reference 6).  Low level analyses
wiTI  require Ultrex grade  acids (J.T. Baker)  or equivalent.   Instra-Analyzed
grade  acids {J.T.  Baker)  or equivalent may be  suitable for less sensitive
analyses.  Copper contamination may be particularly troublesome when Instra-
Analyzed  acids are used.

7.1  Distilled Deionized  Water  (DDW)  -  a  water  purified by distillation
(or equivalent)  followed by conditioning  with a mixed bed  ion exchanger.
Such units are commercially available  and yield a  water  with a typical
resistivity of 18  megohms/cm.

7.2  Hydrochloric  Acid - concentrated (35 percent).

7.3  Hydroxylamine Hydrochloride [20  percent (w/v)]: - dissolve 20 g  of
American  Chemical  Society (ACS)  grade  NH20H-HC1 in  100 mL of ODW.   Store
in a  precleaned glass or plastic bottle - prepare weekly.

7.4  Nitric Acid - concentrated (70 percent).

7.5  Perchloric  Acid - concentrated (70 percent).
                                 III-6

-------
7.6  Sodium Borohydride, ACS Grade Granular  or  Powder.

7.7  Sodium Hydroxide,  ACS Grade - pellets or  flakes,

7.8  Stannous Chloride [20  percent (w/v}]  - dissolve 20  g  of ACS grade
SnCl2 in 20 ml of concentrated hydrochloric  acid.  Warm gently until  solution
clears, cool and  add  DDW until the solution  reaches a 100 ml volume.   Store
in a precleaned glass or plastic bottle  -  prepare  fresh daily.

7.9  Stock Standard Solutions - These standards (typically  1,000 ppm) can
be purchased as  certified solutions or prepared  from ACS grade metal  salts
and pure compounds.  Suitable procedures  for preparing  stock solutions
are well documented (e.g., reference 15) and  include the steps below.

     7.9.1   Accurately weigh  1,000  mg of pure metal  or  metal  equivalent
of the  salt and dissolve in a minimum amount (usually about 20 ml)  of an
appropriate acid.  Once  the  reagent  is dissolved, dilute the solution to
1,000 ml with DDW and store in a precleaned  plastic bottle.   The solution
is usually stable for  at  least  1 year  but  must be checked  periodically
against an in-house control standard (Sect.  10).

8.0  SAMPLE COLLECTION, PREPARATION, AND STORAGE

8.1  The major difficulty in  trace metal  analyses  of  tissue  samples is
controlling contamination  of the sample.  In the  field, sources of contamination
include sampling gear, winches  or steel  cables, engine exhaust,  dust, or
ice used for cooling  (reference 16).  Care  must be taken during handling
to avoid these and any other possible sources of contamination.   For example,
stainless steel collection  and handling  devices (e.g., grab  samplers or
sieves  used for  infaunal  collection  from  sediments) are  suitable. The
ship should be positioned  such that the engine exhausts  do  not  fall on
deck during sampling.  To avoid contamination  from melting ice, the  samples
should  be wrapped in  aluminum foil and placed  in watertight  plastic  bags.
The ojter skin of the fish or shell of the  shellfish is protection against
metals  contamination  from the aluminum foil.
                                 III-7

-------
8.2  Sample resection  (i.e., surgical  removal  of tissue)  and  any  subsampling
of the organisms  should be carried out in  a  controlled environment (e.g.,
a dust-free room).   In most  cases, this  requires  that the organisms be
transported on  ice  to  a  laboratory,  rather than  being resected  on board
the  sampling vessel.  It is recommended  that whole organisms not  be  frozen
prior to resection  if  analyses will be conducted only on selected  tissues,
because  freezing may  cause internal  organs  to rupture and  contaminate other
tissue.  If organisms are eviscerated on board  the survey vessel, the  remaining
tissue (e.g., muscle)  may be wrapped  as described above and frozen.

8.3  Resection  is  best performed under "clean room" conditions.  The  "clean
room" should have positive  pressure  and  filtered  air.   The "clean room"
should  also be entirely metal-free  and  isolated from all  samples high in
contaminants (e.g.,  hazardous waste).   At  a  minimum, care should  be taken
to avoid  contamination from  dust,  instruments, and all  materials that may
contact the samples.   The best equipment  to use  for trace  metal  analyses
is made of quartz, TFE, polypropylene, or  polyethylene.  Quartz utensils
are ideal  but expensive.  To control  contamination  when resecting tissue,
separate  sets  of  utensils  should be  used  for  removing outer tissue and
for removing tissue  for analysis.  For bench  liners and bottles, borosilicate
glass  would be preferred  over plastic  if trace  organic  analyses  are to
be performed on the same sample.

8.4  Resection  should  be conducted by or under the supervision of a competent
biologist.   For fish  samples, special care must be taken to avoid contaminating
target  tissues (especially  muscle) with  slime  and/or  adhering sediment
from the fish exterior (skin)  during  resection.   The incision "troughs"
are  subject to such  contamination; thus,  they  should  not be  included in
the sample.  In the case of muscle, a "core"  of tissue is  taken from within
the  area  boarded  by  the incision troughs,  without contacting them.  Unless
specifically sought as a  sample, the  dark muscle tissue  that may exist
in the  vicinity  of the lateral line  should  not  be mixed  with  the light
muscle tissue that  constitutes the rest of the muscle tissue  mass.
                                 II1-8

-------
3.5  After  the appropriate  tissues  are  resected, samples should  De  store-
in suitable containers  (Sect.  6.1)  and  frozen  at  -20° C until  analysis.
Although specific  holding times have  not  been  recommended  by  U.S.  EPA or
other  federal  agencies  (e.g.,  National  Bureau  of  Standards),  a holding
time of 6 mo (except  for mercury samples,  which  should be  held  no  longer
than 28 days)  would be consistent with that recommended for water  samples.

9.0  CALIBRATION AND  STANDARDIZATION

9.1  Calibration  standards  are prepared by serial  dilutions  of  the stock
solutions.   The acid matrix of the  standrds  should be as  closely matched
to the  samples as possible  (i.e., approximately 1  percent (v/v) HN03 and
4 percent (v/v) HClO/i).

     Mixed  standards of more than one  element may  be prepared only after
their compatibility has been determined.   Some  common mixed  standards  are
as follows:

          Cd,  Cu,  Pb, Ni, and  Zn
          As,  Se,  and Sb

     9.1.1  Do not add  an  incompatible  anion to a mixed or single element
standard.   For  example,  adding chloride  to  a  silver standard could form
a precipitate  of silver chloride  (AgCl).

     9.1.2   Do  not mix metals  that are incompatible in solution.  For example,
lead  and chromium may form a precipitate of lead chromate (PbCr04).

9.2  Concentration ranges of  the standards  should bracket  those for  the
samples to  be  analyzed.  At least four analyses (one blank and three standards
of increasing  concentration)  should  be used  to  calibrate  the  instrument
at the  beg in ing of each shift.

9.3  Stability of a  calibration standard  varies with element,  acid matrix,
concentration,  and presence of other elements.  As a general rule,  standards
                                 III-9

-------
 should  be continuously monitored and replaced  when  necessary.  As a matter
 of protocol, the following can be used as a guideline:

          less than  0.1 ppm -  prepare daily
          0.1 to 1 ppm      -  prepare weekly
          1.0 to 10  ppm     -  prepare monthly
          10 to 100  ppm     -  prepare  quarterly
          100+ ppm          -  prepare yearly (at a minimum)

 9.4  Initial  Standardization - follow manufacturer1s suggestions for standard-
 izing the instrument and check sensitivity performance with specifications.
 If performance is acceptable, proceed with analysis;  if not, refer to manu-
 facturer's troubleshooting guide.

 9.5  After  standardizing  the  instrument, analyze an  independent control
 standard as a  check.   If  the result  is acceptable,  proceed; otherwise,
 troubleshoot calibration standards, control standard, or  instrument.

 9.6  Ongoing  Calibration (reference 17) - the  instrument should be tested
 with a single  point  calibration every  2 h during  an  analysis run or  at
 a  frequency of  10 percent  of  the  analyses, whichever  is  more frequent.
 A calibration check must also be run after the last sample  in a laboratory
 shift.   A standard  concentration  in the middle of  the  initial calibration
 range should be used.

     If  the difference between the ongoing calibration result and the known
 standard concentration  is  greater  than ^10 percent  (or  +2Q percent  for
mercury  analysis), the  instrument must  be recalibrated and the preceding
 10 samples reanalyzed for the analytes affected.

9.7  In  the event  that a  sample is outside of the linear response of the
 instrument,  it must be diluted to within  range or reanalyzed using a  less
 sensitive setup.  This is commonly accomplished  by calibrating the instrument
with higher  concentration standards using a secondary or  tertiary wavelength
having  less  sensitivity.
                                 111-10

-------
 10.0   QUA LI TY CGNTRCL   ,see  reference 17 ana Quality  Assurance/Qua! it./
 Control tQA/QC)  for  301(h)  Monitoring  Programs:   Guidance on Field  arc
 Lcooretcry Metnoos  (Terra Teen 1936].]

     A quality  control  program  enables trie  assessment of  the  precision
 and accuracy of  data.  Precision is estimated  by  analysis  of replicates.
 Accuracy  is estimated  by the analysis of blanks,  spiked samples,  and/or
 laboratory control  samples (reference 18).

 10.1   Replicates can be cnosen  to  reflect the precision  of most  stages
 of the overall analytical metnod.  Replicates can  consist of different
 subsamples of a tissue  homogenate  or  replicate  instrumental analyses  of
 the same digestion  extract.

     10.1.1   Replicate  analyses  of tissue subsamples are important Decause
 "the greatest  potential for  sample deterioration and/or contamination  occurs
 during  preanalysis steps of  sample  collection,  handling,  preservation,
 and storage"  (reference 19).

     10.1.2   Replicate  analyses of a digestate  focus  only on the  bench
 chemistry  and/or  instrumental  variability of  the  method.   Together with
 replicate  analysis of tissue  subsamples, they can  be used to assess  the
 impact of  each stage on the  overall precision of the analytical result.

     10.1.3   At  least one  replicate  (a subsample of a tissue homogenate)
must be analyzed  from each group of samples of a similar matrix type and
 concentration for each batch  of samples or for each 20 samples, whichever
 is more frequent.   If two analytical methods are used  for  the  same element
 in a batch  of  samples, duplicates  must  be run by each  method used.

     The relative percent differences (RPD) for each component are calculated
 as follows:
                 D  - D
          RPD  =
                                 III-ll

-------
where
     DI = first  sample value.
     $2 = second  sample value.

10.2   As  in  the case of  replicates, blanks can be chosen to  address most
stages of the overall  analytical method.   They include  transportation,
dissection,  reagent,  and calibration blanks.

     10.2.1  Transportation blanks  are  derived from empty containers that
have been stored  with samples in the  field  and  carried with  them to the
laboratory.   A small amount of 5 percent  (v/v) HNQ3 is used  to  rinse the
inside of the container.   The acid rinse  is then  retained  for  analysis.
Transportation blanks serve as estimates of contamination during  preanalysis
steps (Sect.  10.1.1).
  *
     10.2.2   Resection blanks are used  to  estimate concentration from resection
utensils that may carry over from one  sample  to the next.  They  are prepared
by collecting  a final  rinse after cleaning  utensils that have  been used
for resection.   The final  rinse should be  performed with  a known volume
of 5 percent (v/v) HN03.   One resection blank should be analyzed for each
batcn of samples.

     10.2.3  Reagent (preparation)  blanks  are al iquots of 5 percent (v/v)
HN03 that are processed through each  sample  preparation step (e.g., reagent
addition, digestion, dilution).   At least  one reagent blank  must  be prepared
for each batch of samples or for every 20  samples, whichever is more frequent.
Reagent blanks serve as estimators of contamination resulting  from the
chemical analysis steps.

     All samples with  at  least one analyte  concentration that  is  less than
10 times the  corresponding  concentration  in  the  associated  reagent  blank
must be redigested and reanalyzed.

     10.2.4  Calibration  blanks consist of 5 percent  (v/v) HN03 and are
analyzed each time the instrument is calibrated,  at  the beginning of each
                                 111-12

-------
analysis  run,  and at a  frequency of 10 percent during a run.  Calibration
blanks are used  to  ensure that the analytical  instrument is not introducing
false positive results during analysis.   (Ongoing calibration quality assurance
is discussed  in  Sect. 9.6.)

10.3  The  results  obtained from the reagent  blanks can be used to calculate
the LOO  (Sect. 3}  for the method.  This  is the  assigned minimum value above
which reliable  data can be  reported.   Results  for at least  the reagent
blank should  be  reported with the final  data  set.

10.4  Spiked  samples are samples to which small  volumes of standard solutions
of the elements of  interest have  been  added.  Spiked  samples provide a
means of assessing  losses during digestion, distillation, or other pretreatment
steps.  The spike  is added  before the  pretreatment  steps  and should be
0.5  to  2.0 times the concentrations of the  elements  in  the  sample.  At
least one  spiked  sample must be analyzed for  each batch  of samples of a
similar matrix  type and concentration  or  for each 20 samples, whichever
is more  frequent.

     10.4.1  The  percent recovery for each element  is calculated  as follows:

          % Recovery = (spike + sample result)  -  (sample result)  x 1QO
                   J                  (spike  added)

Spike percent recoveries should not be used to  determine a correction factor
to compensate for  losses.

     10.4.2  If  graphite furnace  AA is  used, a  single analytical  spike
is required after  any digestion steps to determine  if the method of standard
additions  (MSA)  is  required (reference 17 was used  to develop this section).

     The spike  should  be added  at a concentration  (in the sample)  that
is twice the LOD.  The  unspiked  sample aliquot must  be  compensated  for
any  volume change in the spiked samples by  addition of DDW to the unspiked
                                 111-13

-------
sample aliquot.  The  percent  recovery of the  spike  should be calculated
as in Sect.  10.4.1

          10.4.2.1   If  the sample absorbance or concentration  is >50 percent
of the  spikel ana  the  spike recovery  is between 85 percent  and 115 percent,
the sample should be  quantified directly from the calibration  curve.

          10.4.2.2   If the  spike recovery  is  less than 40 percent,  the
sample must  be diluted  and  rerun with another  spike.   Dilute the  sample
by a  factor of 5 to 10  and  rerun.   This step must only be performed once.
If after dilution the spike recovery  is still <40 percent, there are  inter-
ferences associated with the instrumental technique that prevent GFAA analysis
of the sample.

          10.4.2.3   If the  spike recovery is >40  percent and the sample
absorbance or  concentration is <50 percent of the spikel, report the analyte
as less than the LOD or  less than the LOD  times the  dilution factor if
the sample was diluted.

          10.4.2.4  If  the sample absorbance or concentration  is >50 percent
of the  spikel and  the spike  recovery is <85 percent  or >115 percent, the
sample must  be quantified by MSA.

          10.4.2.5  The  following procedures should  be incorporated into
MSA analyses.

     a)    Data from MSA  calculations must be within the linear range
          as determined  by  the calibration curve  generated at  the
          beginning of  the analytical  run.
l[Note  that  spikel is defined  throughout Sect.  10.4.2  as  (absorbance or
concentration of spike sample)  minus (absorbance  or concentration  of  the
sample.]
                                 111-14

-------
     b)   The  sample and  three  spikes must be analyzed  consecutively
          for MSA quantitation (the  "initial" spike run data  is speci-
          fically excluded from use  in  the MSA quantitation).

     c)   Spikes (post-digestion,  as  for the "initial"  spike  in Sect.
          10.4.2) should be prepared  such that:

              Spike 1  is approximately  50 percent of  the sample
              absorbance.

              Spike 2  is approximately  100 percent of  the sample
              absorbance.

              Spike 3  is approximately  150 percent of  the sample
              aosorbance.

     d)   The  data  for  each  MSA  analysis should be clearly  identified
          in the raw data documentation along with the slope,  intercept
         and  correlation coefficient (r)   for  the least  squares fit
         of the data.

10.5  Laboratory  control samples  are certified reference  material s (CRM)
submitted  blind  to the laboratory.   CRM provide an estimate of  the  accuracy
of the overall  method.   A CRM must be  chosen  that has a similar matrix
to samples  and  contains all  the  analytes.   CRM can be purchased  from a
number of  agencies and  are available in a range of matrices  (e.g., U.S. EPA
Trace Metals in  Fish Tissue or NBS Oyster Tissue).

     10.5.1 Unlike an analyte spike  (Sect.  10.4), a CRM tests the dissolution
technique as well as instrument calibration  and matrix interferences.

     10.5.2  A minimum  of one CRM  should  be analyzed  for each  survey or
2 percent of the total number of samples  (i.e., 1 per 50 samples),  whichever
is more  frequent).
                                 111-15

-------
      10.5.3  The percent  recovery  for  eacn element for the overall  method
 is calculated as follows:
          % Recovery = -^ x 100
where
       x = the analytical result for the  element
     REF = the certified result for the element.

The data obtained for each reference material  should be used to troubleshoot
the method if results fall outside the  acceptable range (i.e., the 95  percent
confidence interval).   Percent recovery values  should not be used to determine
a correction factor  to compensate  for apparent  procedural  losses,

10.6  Maintenance of Records -  the data  obtained from any  QC work  should
be recorded in an organized  manner to allow for easy retrieval and reviewing.
If sufficient data have  been collected,  it  is recommended  that these  be
plotted on a control chart for  a quick  visual  assessment.  A typical  control
chart for CRM results is presented in Figure  III-l.

     10.6.1   The quality control  chart can be used  to  determine  if the
following recommended guidelines are met:

     10.6.1.1  Not more than 5  percent  of the  results lie outside two  standard
deviations (warning  limit).  A result outside three  standard deviations
requires action.

     10.6.1.2  There are no regular periodic  variations.

10.7  Cleaning and preparation  of  labware  is  an integral part of  a quality
assurance/quality  control  (QA/QC)  program.   Many cleaning procedures have
been proposed in the literature that are  suitable for decontaminating  equip-
ment.   The  main  concerns with cleaning are  removing elements of interest

                                  111-16

-------
from labware while maintaining an inactive  surface.  Sons cleaning  procedures
tend to be too  harsh,  producing an surface  with an  ion  exchange capacity.
In  this  case a solution  could  partially  or  completely "lose" an  analyte
to the container walls  {references 20,  21).

     10.7.1  If at  all possible, labware  should  not be used for work  where
analyte concentrations  vary by more than  ten  times.   For  example, never
use glassware  for tissue  analysis that  has  also been used for sediments.
If one can use  dedicated glassware,  the cleaning  requirements  are greatly
simp! if led.

     10.7.2  A good  universal cleaning procedure for glass and plasticware
is out! ined  below.

          10.7.2.1  Wash labware with a metal-free detergent and warm water.

          10.7.2.2  Rinse  at least three  times with  tap  water  followed
by distilled deionized  water  (DDW).

          10.7.2.3  Soak  equipment or labware in a dilute acid (25  percent
HN03)  bath  for 24 h.   If  possible, the bath should be maintained  at  an
elevated  temperature {70° C).

          10.7.2.4  Rinse  labware with large volumes of DDW and  use  imme-
diately.   If a  time lapse must exist, the apparatus should be stored under
dust-free conditions and rinsed further with DDW prior to use.

NOTES:     -     Change  the  acid  batch  periodically such that no significant
               buildup of metals occurs.
               At  no  time  should a  metal containing reagent such as  chromic
               acid be  used.

10.8   Round  Robin or  Interl aboratory Check  Programs - In addition to the
quality control measures discussed above, all  laboratories should participate
in interlaboratory check programs.
                                 111-17

-------
11.0  PROCEDURE

11.1   Homogenize samples prior to analysis  to  ensure that a representative
aliquot is taken.    Any grinder or  homogeni zer  that has  been  found to be
free  of  contamination may  be used (Sect.  6.3).  Samples should  be ground
wet to avoid losses  of  volatile elements  (e.g., Hg,  Se)  during drying.
The liquid  associated with a sample after thawing should be retained as
part of the  sample.

11.2   Transfer the  sample  paste  to a container suitable for  storage.  If
not immediately analyzed, the samples should  be  frozen (-200 C) until  required.
Containers should be  tightly sealed to  prevent moisture loss or  gain during
storage.
   *
11.3   Dry Weight Determination  - if sample  size  permits and dry-weight
concentrations  are required, dry weight  determinations  may be performed
as follows:   transfer an aliquot of approximately 3 g (weighed  to the nearest
0.1 g) to a preweighed dish.  Allow the  sample to dry in an oven  at 1050 c
overnight, and  determine  the solid  residue  weight  to  the  nearest 0.1 g.
The percent  total solids is calculated  as:

          Ts =  [dry residue wt (g)]/[wet  sample wt (g)]

Dry weight determinations  should not be made  at  the cost of having  insufficient
sample for  metals  analysis.   Significant decreases in the size of samples
used  for  extraction will decrease attainable  detection limits.

11.4   Accurately weigh representative al iquots  of homogenized  tissue to
the nearest  0.1  mg.   If sample size permits,  approximately  5 g  is  required
to maintain optimum detection limits.  Transfer  the weighed tissue  to a
precleaned 125-mL Erlenmeyer flask  equipped with an  all-glass  reflux cap.
Analyze  a sufficient number  of reagent  blanks, sample duplicates, analyte
spikes, and  certified reference materials  concurrently (Sect. 10).
                                 111-18

-------
11.5  Add 10.0 ml of concentrated nitric acid  (ACS grade or better), replace
cap and swirl.  Allow flask  to  stand at room temperature for about  15  hours
in  a  dust-free  ventilated  environment.   Periodically swirl the contents
to help solubilize the  tissue.

11.6   After 15  hours,  gently  heat the flask  to  approximately 1000 C - hold
at this temperature for 1  hour.   Gradually increase  the  temperature  in
50° C  increments to  a  maximum of 2500 c.   Continue digesting  until  all
tissue has been solubilized.   This  usually takes  about  4  hours.  Do  not
rush the initial  digestion as losses of volatile  elements will  likely occur.
Once digestion is complete,  cool flasks to  room temperature  and  add 4.0  ml
of perchloric  acid.

     CAUTION:   Perchloric acid  is a strong  oxidizing agent.  The analyst
               must  be fully aware of the  precautions  assod a ted  with
               Its use.  This  procedure  (i.e.,  use of perchloric  acid
               at sub-fuming temperatures)   has been safely performed
               without a perchloric  hood, but  a  perchloric hood  is
               strongly  recommended nonetheless.   Laboratories  that
               do not carefully  monitor perchloric  acid digestions
               will  be endangered  without  perchloric hoods.  Safety
               precautions  and background information pertaining to
               perchloric acid  can be  found  in reference 22.

11.7   Return  flasks to the hotplate  which  has been cooled to about 200° C.
Continue  heating for 1 hour,  then  increase  plate temperature  to 300° C.
Hold at this  temperature until  all  traces of nitric  acid fumes have disappeared
and the solutions have become clear.  Do not overheat flasks  or allow perchloric
fumes  (dense white)  to  appear.   If perchloric  fumes appear, reduce heat
immediately.  Remove the extracts when clear and  cool them to room temperature.

11.8  When the digestion is  complete,  rinse the  caps into the  flasks  and
transfer  the  extract to  a  precleaned 100-mL volumetric  flask.   Rinse  the
Erlermeyer flask three  times with DDW and combine  with the extract  previously
                                 111-19

-------
added  to  the volumetric  flask.   Adjust the volume with  DDW  and  transfer
to a precleaned  plastic bottle.

     NOTE:   Some elements are not as stable  as others  in solution and  therefore
should be  analyzed  first.  Stability can  be determined by  daily analysis
of the extracts, however, the following  can  be used as a guideline:

          Sb, Pb, Hg, Se and Ag - analyze  within one day
          As  and Cd - analyze within two days
          Cr, Cu, Ni and Zn - analyze within  one week
          Be  and Tl - to be determined.

11.9  Instrumental Analysis - The extracts will  be analyzed  using  various
techniques  of atomic absorption spectrophotometry (AAS).  The method of
choice (i.e., GFAA vs HYDAA) depends on instrument availability,  analyte
concentration and  sample  matrix.  In some  instances it  may be useful to
use more than one AAS method to confirm  a  result.

     11.9.1   Follow  the  manufacturer's  instructions for initial  setup and
calibrate as outlined in Sect. 9 of this method.  As every instrument  responds
uniquely to  a given set  of  conditions,  it  is the analyst's responsibility
to develop  the optimum set of parameters.   Use  calibration  standards and
CRM to ensure that optimum conditions exist.

     11.9.2   Table  III-l  lists  some general  information  for each of the
priority pollutant metals.

     11.9.3   It  is possible to use alternate methods of detection providing
they have been validated using a sufficient  number of previously analyzed
samples or  CRM.

11.10  All  data  generated must be clearly  recorded on a strip chart, printer
or manually  logged  in prepared tables.   The order in which  the extracts
are analyzed should be the  same as it appears  in the records. The data,
when assembled,  should be  reported in consistent units (i.e., mg/L)  to
                                 111-20

-------
avoid  errors when calculating the final  results  (ug/g).  The final  report

should contain  all  necessary methodology,  results,  quality control data

(e.g.,  reagent blank  values)  and  limits  of detection  for each element.

The report must clearly state if any data were blank-corrected.


12.0  CALCULATIONS


12.1   All  results are reported  as micrograms of element per wet gram  of

tissue:
          ug/g  ELEMENT
          (wet  weight basis)
where:
     C =  concentration (may be blank corrected)  of element in final  extract
          (ug/mL)

     V =  volume  of  final extract (mL)
     W =  weight  of  wet tissue (g)


Reagent blank  corrections may be made and  blank  values must always be reported.


13.0  PRECISION AND  ACCURACY


     In order  to  estimate precision and  accuracy  (single lab, multi-operator),

a number of  CRM and  analyte spikes were  analyzed  using  this method.   Table

III-2  summarizes typical  data  obtained.   No data  are currently available
for beryllium  or  thai!inn.


14.0  REFERENCES
 1.   Taylor, O.K.   1985.   Standard  reference material s:  handbook for  SRM
     users.  National  Bureau of  Standards Special  Publicaton 260-100.
     National Bureau of Standards, Washington, DC.
                                 111-21

-------
 2.  Keith,  L.J., W.  Grummet, J. Deegan,  Jr., R.A.  Libby, J.K.  Taylor,
     and  G. Wentler,   1983.   Principles of environmental  analysis.  Anal.
     Chem.  55:2210-2218.

 3.  Manning, D.C., and W.  Slav in.   1983.  The determinatin of trace elements
     in natural  waters using  the  stabilized temperature  platform furnace.
     Applied Spectroscopy  37:1-11.

 4.  Hinderberger, E.J.,  M.L.  Kaiser, and S.R. Koirtyohann.  1981.   Furnace
     atonic absorption analysis of  biological samples using the L'vov  platform
     and matrix  modification.  Atomic Spectroscopy 2:1-7.

 5.  Sturgeon,  R.E.,  S.N.  Willie,  and S.S. Berman.  1985.  Preconcentration
     of selenium and  antimony from  seawater for  determination by graphite
     furance atomic absorption spectrometry.  Anal. Chem. 57:6-9.

 6,  Murphy,  T.J.  1976.  The  role of the  analytical  blank in  accurate
     trace analysis,  pp. 509-539.   In:  Accuracy in Trace Analysis:   Sampling,
     Sample  Handling, and Analysis.  National  Bureau of Standards Special
     Publication 422.   National Bureau of Standards, Washington,  DC.

 7.  Skoog,  D.A.   1985.   Principles of  Instrumental Analysis.   Saunders,
     Philadelphia, PA.   pp. 270-279.

 8.  Veil Ion, C.   1976. Optical  atomic spectroscopic methods,  pp.  123-181.
     In:   Trace Analysis: Spectroscopic Methods for Elements.  D.  Winefordner
     (ed).  Wiley, New York.

 9.  Slav in,  W., G.R.  Carnrick, and D.C.  Manning.   1984.  Chloride interferences
     in graphite  furnace atomic absorption spectrometry.   Anal. Chem 56:163-8.

10.  Carcinogens - working  with carcinogens.  DHEW, PHS, CDC,  NIOSH.   Publica-
     tion  77-206 (Aug.  1977).

11.  OSHA safety  and  health standards,  general  industry.  OSHA  2206,  29
     CFR 1910 (revised  Jan. 1976).

12.  Safety  in  academic chemistry laboratories.  ACS Publications, Committee
     on Chemical  Safety, 3rd  Edition  (1979).

13.  Cantle,  J.E. (ed).   1982.  Atomic absorption spectrometry.   Elsevier,
     New York.

14.  Fuller,  C.W.  1978.   Electrothermal  atomization  for atomic  absorption
     Spectroscopy.  The Chemical  Society, London.

15.  U.S.  Environmental Protection Agency.   1979.   Methods for  chemical
     analysis of water  and  wastes,   pp.  202.1-289.2.   U.S.  Environmental
     Protection  Agency Environmental Monitoring and  Support  Laboratory.
     Cincinnati,  OH.
                                  111-22

-------
16.   Mart, L.   1979,   Prevention of contamination and other accuracy risks
     in voltammetric  trace metal  analysis  of  natural waters.   Part  II:
     Collection  of  surface water  samples.  Fresenius Z. Anal.  Chem.  299:97-102.

17.   U.S.  Environmental  Protection Agency.  1985.  U.S. EPA  Contract Laboratory
     Program  -  statement  of work  for inorganic analyses, multi-media multi-
     concentration.

18.   U.S. Environmental  Protection Agency.   1983.  Guidance for preparation
     of combined work/quality assurance project plans  for water monitoring.
     Office  of  Water Regulations and Standards, U.S. EPA,  Washington,  DC.
     33 pp.
19,
     Plumb,  R.H.,  Jr.   1981.   Procedures for handling and chemical  analysis
     of sediment  and  water  samples.  Technical Report EPA/CE-81-1.   Environ-
     mental  Protection Agency/Corps of  Engineers  Technical  Committee  on
     Criteria for Dredged and  Fill Material, U.S. Army Waterways Experiment
     Station, Vicksburg, MS.   471 pp.

20.   Batley,  G.E., and D.  Gardner.  1977.  Sampling and storage of natural
     waters for  trace metal analysis.  Water Res. 44:745-756.

21.   Laxen , D.P.H., and  R.M. Harrison.  1981.  Cleaning methods  for  polythene
     containers  prior to the determination  of  trace metals  in freshwater
     samples. Anal.  Chem.  53:345-350.

22.   Schilt,  A.A.   1979.   Perchloric  acid and perchlorates.  G.  Frederick
     Smith  Chemical Company, Columbus, OH.
                                  111-23

-------