EPA-R2-73-277
August 1973               Environmental Protection Technology Series
CURRENT PRACTICE  IN GC-MS
ANALYSIS  OF ORGANICS  IN  WATER
                                National Environmental Research  Center
                                Office of Research and Development
                                U. S. Environmental Protection Agency
                                Corvallis, Oregon  97330

-------
            RESEARCH REPORTING SERIES
Research reports of the  Office  of  Research  and
Monitoring,  Environmental Protection Agency, have
been grouped into five series.  These  five  broad
categories  were established to facilitate further
development  and  application   of   environmental
technology.   Elimination  of traditional grouping
was  consciously  planned  to  foster   technology
transfer   and  a  maximum  interface  in  related
fields.  The five series are:

   1.  Environmental Health Effects Research
   2.  Environmental Protection Technology
   3.  Ecological Research
   
-------
                                                 EPA-R2-73-277
                                                 August 1973
         CURRENT  PRACTICE  IN  GC-MS ANALYSIS

                OF ORGANICS IN WATER
                          by
                    Ronald  G.  Webb
                  Arthur W.  Garrison
                   Lawrence  H.  Keith
                    John M.  McGuire
     Southeast  Environmental Research Laboratory
                 College  Station Road
                Athens, Georgia  30601
                   Project  16020 GHP
                Program Element 1B1027
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.25
       NATIONAL ENVIRONMENTAL RESEARCH CENTER
          OFFICE OF RESEARCH AND MONITORING
    UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
               CORVALLIS,  OREGON  97330

-------
               EPA Review Notice
Mention of commercial products and trade names
is for information only and does not constitute
endorsement.

-------
                        ABSTRACT

Experiences during five years of evaluating the appli-
cation of gas chromatography-mass spectrometry (GC-MS)
to wastewater analysis at the Southeast Environmental
Research Laboratory have resulted in the selection of
recommended practices for such applications.  Liquid-
liquid extraction with solvents such as methylene
chloride and chloroform removed greater than 50 percent
of compounds found in pulp mill and petrochemical waste
at concentrations of 2 yg/£ to 20 ug/£.  The Kuderna-
Danish evaporator was the most effective means of con-
centration after extraction.  Diazomethane and dimethyl
sulfate proved to be the most effective of five
methylation reagents studied.  Packed columns were
effective for gas chromatography of simple mixtures
and SCOT columns provided better overall performance
for complex mixtures.  Computerized data reduction was
essential for practical use of GC-MS for samples
containing many compounds.  A computerized spectra
matching program proved highly effective in identifying
compounds contained in the computer library.  The system
was shown to be effective in solving problems related
to fishkills caused by pesticides, confirmation of
polychlorinated biphenyl residues in water and identi-
fication of compounds discharged by over a dozen indus-
tries.  Over two hundred compounds were identified in
industrial effluents.

-------
                         CONTENTS

Section                                       Page No.

I.     RECOMMENDATIONS                             1

II.   INTRODUCTION                                3

III.  SAMPLE HANDLING                             5
        Collection                                5
        Extraction                                6
        Kuderna-Danish Evaporation                9
        Clean-up Techniques                      11

IV.   DERIVATIVE FORMATION                       15
        Trimethylsilyl Derivatives               15
        Methyl Derivatives                       16
          Diazomethane                           16
          Dimethyl sulfate                       18
          Others                                 18
        Ozonolysis                               20

V.     GAS CHROMATOGRAPHY                         21
        Columns                                  21
          Packed                                 21
          Capillary and SCOT                     21
          Coatings                               22
        Flow Rate                                22
        Solvents                                 23
        Temperature Programming                  23

VI.   MASS SPECTROMETRY                          25
        Instrumental                             25
        Operation                                28
        Specialized Techniques                   31
          LMRGC                                  31
          Solids probe                           34

VII.  INTERPRETATION AND COMPUTER MATCHING       37

VIII. CONFIRMATORY TECHNIQUES                    41

IX.   CASE HISTORIES                             43
        Black Warrior River and Locust Fork      43
          Branch Fish Kills

-------
Section                                       Page No.

        Polychlorinated Biphenyls (PCB's)        43
        Industrial Effluent Characterization     45
        Correlation of Compounds in Natural      48
          Waters with Industrial Wastewater
          Discharges
        Waste Treatment                          50

X.    ACKNOWLEDGEMENTS                           55

XI.   BIBLIOGRAPHY                               57

XII.  APPENDICES                                 61
                            VI

-------
                            FIGURES

 NO.                                                PAGE

 1    KUDERNA DANISH EVAPORATOR                     10

 2    OUTLINE OF FINNIGAN  GC-MS-COMPUTER SYSTEM     27

 3    RECONSTRUCTED GAS  CHROMATOGRAM (RGC) OF A     30
        PETROCHEMICAL  PLANT EFFLUENT EXTRACT

 4    MASS SPECTRUM OF FLUORENE                     32

 5    RECONSTRUCTED GAS  CHROMATOGRAM AND LIMITED    33
        MASS RECONSTRUCTED GAS CHROMATOGRAM

 6    COMPLICATED  RGC  FROM AN INDUSTRIAL            35
        EFFLUENT EXTRACT

 7    LIMITED MASS RGC OF  AN INDUSTRIAL EFFLUENT    36
        EXTRACT

 8    MASS SPECTRA OF  EXTRACTS AND STANDARDS        44
        CONTAINING MALATHION

 9    FLAME IONIZATION DETECTOR GAS CHROMATOGRAMS   46
        OF PCB'S AND KNOWNS

10    PARTIAL MASS SPECTRA OF PCB'S FROM            47
        FIGURE 9

11    COMPUTER RECONSTRUCTED GAS CHROMATOGRAM OF    51
        CHLOROFORM BLANK

12    RGC OF UPRIVER CONTROL SAMPLE EXTRACT         52

13    RGC OF PAPER MILL  EFFLUENT EXTRACT            53

14    RGC OF DOWNSTREAM  SAMPLE EXTRACT              54

15    APPARATUS FOR DIAZOMETHANE METHYLATION        89

16    APPARATUS FOR DIMETHYL SULFATE METHYLATION    91
                              VI l

-------
                          TABLES

No.                                              Page

1    Methylation Efficiencies of Four Reagents    17

2    Qualitative Comparison of Methylation        19
       Reagents

3    Minimum Acceptable Criteria for Mass         26
       Spectrometer System

4    Steps in GC-MS-Computer Data Reduction       29

5    Comparison of Three Spectrum Matching        39
       Programs

6    Compounds Identified in Wastewater of        49
       Petrochemical Plant
                            Vlll

-------
                          SECTION I

                       RECOMMENDATIONS

Gas chromatography-mass spectrometry should be used as
the first approach to analysis of organic water pollu-
tants when the identities of the organic constituents
in the sample are in doubt.  Computerized data
reduction and data interpretation are recommended to
improve efficiency and to eliminate the necessity for
an interpretation specialist at every laboratory.

The procedures described in this report are recommended
as a first step toward eventual development of a method
suitable for corroborative testing and consideration
as a standard method.

-------
                       SECTION II

                      INTRODUCTION

Currently, the most effective method for identification
of organic pollutants in water is gas chromatography-
mass spectrometry (GC-MS).

Any compound that can be gas chromatographed without
decomposition can be analyzed by GC-MS.  Appendix 1,
containing over two hundred compounds identified in
industrial wastes, illustrates the wide applicability
of GC-MS.  However, some classes of compounds are better
suited to other methods.  Polymers, sugars, free amino
acids and many other biologically derived compounds are
not directly amenable to GC-MS because of their
non-volatility.

One evidence of the recent popularity of GC-MS is the
purchase of nineteen instruments within the last two
years by EPA laboratories throughout the country.
Significant time will be saved if these new users have
the benefit of five years experience at the Southeast
Environmental Research Laboratory in sample preparation,
GC conditions, data reduction, interpretation of spectra,
and all the other facets of this type of analysis.
This report describes the most successful current
practice in these areas  and illustrates the power of
GC-MS in pollution analysis.

-------
                         SECTION III

                       SAMPLE HANDLING

                         Collection

Lake and river waters are not homogeneous.  Industrial
effluents also change with time.  Some investigators
prefer to take subsamples from various locations and
at different times and to combine them for analysis.
Many chemists avoid filtering the sample because the
colloidial and suspended particulates frequently have
organics absorbed on their surfaces.

Glass jars closed with Teflon-lined caps are preferred
for sample containers.  The jars and caps should be
thoroughly cleaned with soap and water, rinsed with
distilled water, and rinsed with the solvent that will
later be used for extraction.  After thorough drying,
the jars should be tightly capped until the sample is
taken.  Additional jars should be prepared for
distilled water blanks.

For qualitative analysis, plastic bottles can be used
to collect samples that contain high concentrations of
organics.  For these samples, e.g., paper mill efflu-
ents and municipal sewage, the adsorption of organics
on the walls of the bottle and the release of plasti-
cizers  (phthaiate esters) into the sample are insigni-
ficant for short periods of storage.

The volume of sample collected is usually one or two
liters.  Compounds present in a one-liter sample at
concentrations of 2 yg/fi, or greater will generally give
good quality spectra when processed by extraction,
concentration, and GC-MS techniques.  For greater
sensitivity larger samples are required.  For example,
twenty-liter samples of municipal sewage have been
processed with an apparent detection limit of 0.1 yg/Jl
based on an internal standard added to the extract
before concentration.

Because samples may have to be transported or stored
for considerable time before analysis, some means for
stopping biological action and preventing chemical

-------
changes is needed.  Freezing is the preferred method.
The addition of strong acid as a preservative is not
recommended because it degrades some sensitive compounds
such as geosmin and 2-methyl isoborneol (taste and odor
causing compounds).  Some space must be allowed in glass
containers for the expansion of water as it freezes.
Some workers find such a high proportion of breakage
on freezing that they prefer to pack the bottles in
ice for shipment and store them in a refrigerator at
about 4° C.  Plastic containers, particularly the
popular cubitainers, do not usually rupture on freezing.

Samples are conveniently shipped by air freight.  They
are usually frozen, placed in styrofoam containers for
insulation and protection against breakage, and
further packed in dry ice in an inexpensive styrofoam
chest.
                       Extraction

The first step in GC-MS analysis is extraction of the
organic compounds from the bulk of the water.  Commonly,
a liter sample of water is extracted with two or three
50 to 100-ml portions of solvent.  The combined
extracts are dried and the solvent is evaporated to 1
ml or less for GC analysis.  Recovery studies on
compounds found in petrochemical refinery wastes and
paper mill effluents show the effectiveness of this
procedure.  A liter of water, spiked with knowns, was
extracted with three portions (100/50/50 ml) of chloro-
form (duplicates were extracted with methylene chloride),
dried, and evaporated to 500 to 200 microliters for
injection and quantitation by flame ionizetion GC.  The
recoveries were 50 to 100% at the 20 and 2 yg/& levels.
No significant differences were noted between methylene
chloride recovery and chloroform recovery; however,
methylene chloride is more likely to form emulsions.
Compounds with high vapor pressures (methyl styrene,
cymene, indene, methyl indene, terpinolene, camphor)
gave recoveries of 50-70%, whereas low vapor pressure
organics  (indole, naphthalene, quinoline, biphenyl,
dimethyl naphthalene, diphenyl methane, acenaphthene,
bibenzyl, carbazole, benzophenone, phenanthrene,
fenchone, triphenylmethane, fenchyl alcohol,

-------
a-terpineol, and guaiacol) gave 72-100% recoveries.

Other solvents in common use are petroleum ether,
hexane (also 15% diethyl ether in hexane), diethyl
ether, and carbon tetrachloride.  Petroleum ether and
hexane are so non-polar that they do not effectively
extract polar materials such as fatty acids and phenols.
Ether is not recommended  (except in very small quanti-
ties for diazomethane methylations) because of its
flammability and inefficiency as a solvent, the danger
of explosion from peroxide impurities, and the
presence of additives.  Various preservatives are added
to ether to inhibit peroxide formation but are not
usually listed on the container label.  One peroxide
inhibitor identified by GC-MS and later confirmed by
the manufacturer was 2,6-ditertiarybutyl-p-cresol.
This compound was recognized as an artifact because
it also occurred in the blank.  Finally, in solvent
comparison studies, ether was found to be only half as
effective as chloroform in extracting a paper mill
effluent.

Carbon tetrachloride is sometimes preferred because
it gives a narrow solvent peak on some GC columns;
however, it is less efficient than chloroform for
extraction of polar compounds.  Carbon disulfide and
various freons have the advantage of very high vola-
tility but are difficuT'-  'zo purify and usually show
numerous impurity peaki,.   .uditionally, carbon disul-
fide is very flammable.

Extraction is affected by the pH of the sample.
Organic acids are ionized salts in basic solution and
will not extract into organic solvercs.  Therefore
samples containing fatty  acids  (municipal sewage)
or phenols  (forest runoff) are best extracted at a
pH <5 to insure maximum covalent character.
Similarly, basic materials are best extracted at high
pH.  A complete, but somewhat involved, procedure for
these "solubility class" separations was given by
Braus, Middleton and Walton (1).  We recommend
extracting samples whose original pH is 5-7.5 with
solvent to isolate the neutral compounds.  Then the
sample is acidified to pH 3 or less with hydrochloric
or sulfuric acid and extracted to isolate the acids

-------
and phenols.  To isolate basic compounds/ the acid
solution is adjusted to pH 10 and extracted.  Usually
very few basic compounds are found.  Many variations
in procedure are possible and each problem requires an
individual judgment.

Several operational details help make extraction
simpler or more effective.  Pesticide quality (dis-
tilled-in-glass) solvents are recommended.  The
separatory funnel should have a Teflon stopcock and
the glass stopper should not be lubricated.  If the
sample contains suspended matter, a representative
portion of it should be included in the sample for
extraction.  The first portion of extracting solvent
should be used to rinse the sample container and then
added to the separatory funnel.  Some samples, parti-
cularly those from anaerobic sites, will form an
emulsion or deposit a sludge in the solvent layer.
One of the main components of this material is elemental
sulfur.  Some of the sulfur dissolves in the solvent
and precipitates as a whitish solid when the extract is
concentrated by evaporation.  It usually chromatographs
as a single peak during GC-MS analysis and gives a
molecular ion at m/e 256  (Sg) and fragments from
successive losses of 32 mass units.

To separate the sludge from the solvent or to break an
emulsion, place a 2.5 cm ball of glass wool in a 2.5
cm diameter plain chromatography column—no frit—
and, after wetting the wool with fresh extraction
solvent, pour the emulsion through the glass wool.
Ignore the debris that remains.  It may be necessary
to force the emulsion through with a little air
pressure.

After a sample is extracted, a little water may remain
dissolved in the solvent.  The analyst can either dry
the extract before concentration or concentrate the
solvent directly.  No studies have been done to show
which is best.  Drying is usually accomplished by
filtering the extract through a column of anhydrous
sodium sulfate  (200 ml of solvent through a 5 cm X 2.5
cm diameter column is common).  The sodium sulfate
should first be heated to 500° C for 2 hours to remove
organic impurities, usually phthalate esters.  Another

                             8

-------
method for drying is to pour the extract through a
pre-wet glass plug as described earlier.

Some analysts simply concentrate the extract and
inject it on the GC.  With non-chlorinated solvents
even the ultra-water-sensitive electron capture
detector will accept this concentrate.

Numerous other methods have been tried to separate and
concentrate organic materials.  In our experience
freeze concentration and steam distillation are vastly
inferior to extraction.  Carbon adsorption has limited
use but the labor involved in processing makes it
unpopular.  Polyurethane foams are now being used to
extract polychlorinated biphenyls directly from water.
These foams, coated lightly with DC-200, also extract
organochlorine pesticides.  Our tests show that
neither material is effective for extracting dissolved
oil, phenols, or terpenes.  Organic resins, in
particular Rhom and Haas' XAD series, show much more
promise but applications are still in the developmental
stages.
               Kuderna-Danish Evaporation

The best method  for concentrating the organic extract
is distillation  with a Kuderna-Danish evaporator
(Figure 1), available from Kontes.  The bottom and
middle flasks are filled to about half the nominal
flask capacity.  Boiling chips are added and the whole
assembly is placed on a boiling water bath or a steam
bath in a hood.  The water level should be maintained
just below the bottom joint and steam should bathe the
rounded bottom of the middle flask.  Some people let
distillation become just vigorous enough to maintain
liquid in the upper bubble joint; others feel that it
is equally efficient, as well as quicker, to allow
moderate splashing out the top.  Distillation is
usually continued until about one ml of liquid remains
in the concentrator tube; the whole assembly is then
removed and allowed to cool.  As the liquid drains
from the condenser it rinses the residue from the walls
of the middle flask.  The final volume is about five ml.

-------
FIGURE 1.  KUDERNA-DANISH EVAPORATOR




                  10

-------
The calibrated concentrator tube is removed and placed
in a beaker of warm water (<70°) and the volume reduced
to any desired level, usually one ml, by directing a
stream of clean dry air or other gas over the surface.
A fresh glass disposable pipette is used as an air jet
for each sample.  The pipettes are connected to the
air supply with Tygon tubing.  Analysis of blanks so
treated have not indicated the presence of volatile
plasticizers from the tubing.  If the extracting
solvent is not compatible with the GC system (if it
tails badly and obscures parts of the peaks), four or
five volumes of a more desirable solvent are added
and the volume is again reduced to the volume necessary
to give adequate GC sensitivity.

Several studies  (2-4) have shown that considerable
loss of the extracted compounds takes place during
air evaporation.  For very volatile compounds, 50%
loss on evaporation to about  1 ml is common; for less
volatile materials such as DDT-type pesticides, losses
are only a few percent.  Fortunately the concentration
factor increases faster than  the loss factor.  There-
fore, enough compound for a good mass spectrum is
concentrated into a solvent volume acceptable to both
the GC and the MS.  If the sample must be  stored over-
night or longer  the tube should be closed  with a glass
stopper and placed in the refrigerator.  Some prefer
to store extracts in Reacti-Vials or other screw-cap
vials fitted with a Teflon cap  liner.

The Kuderna-Danish method is  preferred to  a rotary
evaporator because there is less danger of contamina-
tion in the all-glass apparatus, less danger of
bumping with loss of sample,  and less loss due to
handling since transfers do not have to be made from
large round bottomed flasks to  a calibrated container.
                    Clean-Up  Techniques

When  extracts  of water  are analyzed by GC  for  specific
organic  compounds,  e.g.,  pesticides,  the extract
usually  must be  "cleaned-up"  to remove interferences.
The most widely  used  technique is  Florisil or  silica  gel
column chromatography,  sometimes in combination with
acetonitrile partitioning.   Thin-layer chromatography,

                           11

-------
preparative GC, and liquid chromatography are used less
frequently.  Fortunately, since the mass spectrum of a
compound is much more definitive than its GC retention
time/ extracts for GC-MS analysis generally need not be
cleaned up, particularly when solubility class separa-
tion has been used during sample extraction.

If preliminary GC runs show that clean-up of the
extract is necessary, the concentrated extract is
chromatographed on activated florisil.  In the original
technique  (5), increments of solvent containing
increasing proportions of ethyl ether in petroleum ether
are used to elute fractions containing increasingly
polar organics.  We use methylene chloride rather than
ethyl ether.  This technique was used with a textile
plant effluent being analyzed for dieldrin by GC-MS (6).

As another approach, after solubility separation of
the crude extract, the neutral fraction may be further
separated on a silica gel column.  Elution with
isooctane, benzene, 1:1 chloroform:methanol separates
organics into aliphatic, aromatic, and oxygenated
organic fractions, respectively  (7).

Thin layer chromatography is sometimes used as a
preliminary separation technique for GC-MS analysis.
Samples of industrial and municipal effluents contain
so many components that direct TLC of extracts usually
gives only unresolved streaks.  In a few case inter-
mediate clean-up by TLC is useful, resulting in the
separation of bands of compounds of the same type.
This type of sample clean-up has been used in the
identification of pesticide metabolites by GC-MS  (8).

Preparative gas chromatography may be used as a
preliminary separation technique.  In many cases
initial GC runs of crude sample extracts result in
several overlapping peaks, even under the best of
chromatographic conditions.  These overlapping
portions of the effluent are collected in capillary
tubes, cooled if necessary by dry ice or a tissue
soaked in acetone, and rechromatographed on a high
resolution (usually capillary) column.  With this
less complex mixture, GC parameters may be varied
more freely to obtain better separation.  Preparative

                           12

-------
GC was used in the separation and identification of
polychlorinated biphenyl (PCB)  isomers (9).   In this
case, as with some water samples, several sharp pre-
parative chromatographic peaks that appeared to be one
component each were shown by capillary GC to consist of
two or more components.

Liquid chromatography (LC)  seldom resolves sample
components as well as GC, but may be considered as
a preliminary separation tool in some cases.  Volatile
compounds collected from a LC eluate may be analyzed
directly by GC-MS if the LC eluting solvent is
compatible with the GC-MS system.

One advantage of LC is its amenability to chromato-
graphy of aqueous samples.  Furthermore, ion-exchange
LC columns provide for the separation of water soluble,
non-extractable organic compounds.  One established
procedure  (10) calls for concentration of the original
water sample by direct evaporation and/or freeze drying.
The concentrated sample is dissolved in a buffer solution
and chromatographed on a high pressure anion-exchange
column.  Fractions, usually representing several over-
lapping peaks, are collected in  their buffered eluting
solution and the solution is freeze-dried.  The solid
residue is derivatized to make volatile components
that may be analyzed directly by GC-MS.
                            13

-------
                      SECTION IV

                 DERIVATIVE FORMATION

Many compounds are not volatile enough to be trans-
formed to the gas phase without decomposition.  Such
polar compounds as carbohydrates, amino acids,
sulfonic acids, and nucleic acids are not directly
gas chromatographable.  Some steroids, carboxylic
acids, and phenols can be directly chromatographed in
their free form, but require special columns and
techniques not readily available.  Phenols "tail" on
most GC columns, steroids elute too slowly, and only
recently have columns become available for the
efficient separation of free long-chain carboxylic
acids.

Derivatization reagents are now available  for all
these classes of compounds.  To be useful  for GC-MS
analysis, a  derivative must  (1) be formed  quantita-
tively  from  the  free precursor by a rapid  reaction
with  a  readily obtainable reagent,  (2) be  volatile
enough  for vaporization in the GC inlet,  (3) be
thermally stable in the GC-MS  system,  and  (4) be of
a chemical class that has been studied extensively by
mass  spectrometry  so  that a  spectral  library  is
available for matching.  Methyl  derivatives,  in the
form  of methyl esters of carboxylic acids  or  methyl
ethers  of phenols, and  trimethylsilyl (TMS) derivatives
of carbohydrates,  steroids,  phenols,  amines,  carboxylic
acids, and amino  acids are the  most common.


               Trimethylsilyl Derivatives

A variety of TMS reagents  are  available,  some of which
are tailored for specific  functional  groups.   These
reagents generally react with  groups  containing active
hydrogen atoms  (-PH,  -SH,  -NH,  -COOH)  to  replace  the
hydrogen with a  trimethylsilyl [-Si(CH3)3]  moiety.
Suggestions  for  the  use of  TMS reagents  and information
on their chromatographic  and mass  spectrometric
properties have  been  collected by  Pierce  (11).

TMS derivatives  of about  fifty urinary constituents
separated by high-pressure  anion-exchange liquid

                            15

-------
chromatography were prepared to increase their vola-
tility for identification by GC and MS  (12).  TMS
derivatives are also being used in the identification
of municipal waste components separated by liquid
chromatography (10).
                   Methyl Derivatives

Both diazomethane and dimethyl sulfate are used
routinely to prepare the methyl esters of carboxylic
acids and the methyl ethers of phenols.

Diazomethane.  The diazomethane methylation method is
simple, fast, and almost quantitative for carboxylic
acids  (Table 1).  The procedure for diazomethane
methylation is given in Appendix 2.  It is the method
of choice for extracts in which carboxylic acids are
the most important compounds to be studied.  It has
been applied very successfully to the methylation of
acid fractions of municipal wastes.  However, diazo-
methane has two important disadvantages as a methy-
lating reagent:  it does not react quantitatively with
all phenols, and some undesirable side reactions
occur  (13) .

Although pentachlorophenol reacts quantitatively and
other chlorinated phenols have been observed to
produce methyl ethers with diazomethane, guaiacol,
vanillin, and other phenols found in paper mill wastes
give poor yields (Table 1).

Samples that have been extracted with chloroform,
evaporated to nominal dryness and then redissolved in
ether and methylated, frequently contain half a dozen
or more di- and trichloroalkanes (C2 through Cy).
These are formed by the reaction of the residual
chloroform and diazomethane, a reaction enhanced by
exposure to uv light and excess diazomethane (14).
Since methylene chloride does not react to form
interfering peaks,  it is now routinely used to extract
samples that will be methylated with diazomethane.
Another advantage of methylene chloride is that it does
not have to be completely evaporated before methylation;
loss of volatile phenols is thereby decreased.

                           16

-------
                                     Table 1

                 Methylation Efficiencies of Four Reagents
Compound
Guaiacol
Vanillin
Palmitic acid
Dehydroabietic acid
% Yield*
Diazo-
me thane
15
40
80
95
Dimethyl
sulfate
90
70
60
60
HC1-CH3OH
0
0
90
10
"Me thy 1-8"
0
0
50
85
"MethElute"
85
55
70
70
     *Yield of methyl ether or ester, based on GC peak heights.   These yields
also represent losses occurring upon evaporation of solvents and other
procedural errors.

-------
Dimethyl Sulfate.  In a comparison of methylation
techniques using the acid fractions of kraft paper
mill effluents, dimethyl sulfate gave much better
yields of phenolic methyl ethers than did diazomethane
(Tables 1 and 2).  Dimethyl sulfate also methylates
the resin acids and fatty acids found in paper mill
wastes fairly well, but was not successful in methy-
lating the acid fraction of municipal wastes  (Table 2).

Dimethyl sulfate is recommended for methylation of
extracts containing phenols.  The procedure for
dimethyl sulfate methylation is given in Appendix 3.
An advantage of this method is that the aqueous sample
is used directly, after extracting bases and neutrals.
An important disadvantage is the time required—about
four hours.

Others.  Three other methylation reagents tried are
compared with diazomethane and dimethyl sulfate in
Tables 1 and 2.  Methanolic hydrochloric acid  (6N)
reagent (15) (Table 2) was used with municipal waste
with results comparable to those obtained with diazo-
methane.  However, it does not methylate pure vanillin
or guaiacol, and the esterification yield with
dehydroabietic acid was only 10%  (Table 1).  The
reaction is slower than diazomethane methylations, and
the solvent is restricted to methanol.  There are also
problems with neutralization of the HCl.

"Methyl-8", dimethylformamide dimethyl acetal  (16) was
found to methylate palmitic acid in fair yields and
dehydroabietic acid in good yields, but it did not
methylate the phenols  (vanillin and guaiacol).  It
methylated the municipal waste fraction nearly as well
as diazomethane and is therefore recommended  for
samples containing only carboxylic acids.

The most convenient reagent tested is "MethElute", a
0.2 M solution of trimethyl anilinium hydroxide in
methanol  (17).  The sample extract is dissolved in
this reagent, and the solution is injected into the
GC inlet where the methylation reaction occurs.
By-products and excess starting materials appear  near
the solvent peak, followed by the later-eluting
ethers and esters.  Methylation of municipal waste and
paper mill acid fractions gave results comparable to

                           18

-------
                                     Table 2

                 Qualitative Comparison of Methylation Reagents
Reagent
Diazo-
methane
Dimethyl
sulfate
"MethElute"
"Me thy 1-8"
HCl-MeOH
Phenols
Paper mill
waste
P
B
G
P
—
Municipal
waste
F*
P
—
—
—
Fatty Acids
Paper mill
waste
G
B
G
P
--
Municipal
waste
B
P
G**
G
E
Resin Acids
Paper mill
waste
G
G
G
B
—
Municipal
waste
__
--
--
—
—
B=Best, E=Excellent, G=Good, F=Fair, P=Poor

     *0nly one sample tested; chlorinated phenol methyl ethers were detected
but yields are unknown.  Cresol was not methylated.
    **Two or three reagent peaks formed—they did not interfere in this
analysis.

-------
those using diazomethane.  Because this reagent methy-
lates both carboxylic acids and phenols in fairly good
yields, as seen in Table 1, it has great potential but
should be tested on extracts from other sources.
                      Ozonolysis

Ozonolysis can be used as a derivatization technique.
Many unsaturated compounds can be converted to alde-
hydes or ketones through formation and cleavage of
ozonides.  The aldehyde and ketone products may be
identified, leading to deduction of the original
position of unsaturation.  The ozone for the reaction
can be generated using commonly available laboratory
apparatus  (18).  This technique was used to differen-
tiate the unsaturated fatty acids from their saturated
analogs in municipal sewage by comparison of a sample's
gas chromatogram before and after ozonolysis.
                            20

-------
                      SECTION V

                  GAS CHROMATOGRAPHY

Most samples require some gas chromatographic experi-
mentation to work out the optimum, or at least a
useful set of chromatographic conditions prior to GC-MS
analysis.  Therefore, an additional gas chromatograph
is useful to avoid wasting MS instrument time while
gas chromatographic conditions are being determined.
Glass-lined injectors should be used in both GC's to
avoid decomposition of the sample by initial contact
with hot metal.  The detector of the auxiliary GC
should be flame ionization to achieve sensitivity
similar to that of the MS.  The additional gas chromato-
graph should be of the same type as the one interfaced
with the mass spectrometer because, after optimum
chromatographic conditions are determined, the smallest
change will affect the peak separation of a complex
mixture.  Often the column from the auxiliary GC is
transferred to the GC-MS for the final analysis.
                        Columns

Packed Columns.  The 1/8 inch diameter packed columns
commonly used in GC laboratories are inexpensive and
readily available.  Most mass spectrometers will
accept them.  The analyst must be careful not to
inject the large amounts of solvents (5-10 microliters)
traditionally used because this raises the gas pressure
in the MS past acceptable limits.  For routine surveys
in which only the more concentrated pollutants are of
interest, packed columns are most often used.

Capillary Columns.  Although capillary columns offer
the ultimate in peak resolution they are expensive.
They accept only small amounts of sample and require
long separation times.

Support-coated-open-tube (SCOT) columns 0.02 inches
in inside diameter, are a compromise between the tra-
ditional packed column and the true capillary column.
As such, they offer some advantages of both:
                           21

-------
     •  better peak resolution than packed columns
        with complex mixtures

     •  moderate size sample injections (usually
        1-2 y£)

     •  chromatographic times not much longer than
        with packed columns  (usually 20-40 minutes)

     •  broad choice of support-coating combinations

The technology involved in preparing them is so complex
and involved that it is cheaper to buy them commercially
than to develop the expertise and equipment to make
them.  Perkin-Elmer SCOT columns are normally terminated
with a male nut that fits the PE 900 series gas
chromatographs.  To render these columns useful with
other gas chromatographs, these male nuts need only be
pushed back and replaced with 1/16-inch female slotted
nuts in which the slot is made just large enough to
accept the capillary tubing.  Although these columns
are relatively expensive, they can be returned to the
factory and recoated when column deterioration becomes
evident (broadened and/or tailing peaks).  Some of our
columns have been recoated 4 or 5 times without
apparent effect on their separation efficiencies.

Column Coatings.  Many liquid phases and supports are
on the market.  The two we have found most useful are
SE-30 (non-polar)  and Carbowax 20M terminated with
terphthalic acid (polar).  We also keep on hand DECS,
OV-17, OV-101, OV-225, DC-200, DX-300, QF-1, LB-550X,
and Apiezon-L for special needs.
                       Flow Rate

Helium is the usual carrier gas for GC-MS systems and
should also be used as the carrier gas in auxiliary gas
chromatographs.  The optimum carrier gas flow for GC-MS
systems operating under vacuum and using a Gohlke
separator is 16-18 ml/min.  If about half this helium
flow is used for optimizing conditions with the auxiliary
GC (operating under atmospheric pressure), the other
chromatographic conditions (program rate, initial
temperature, initial hold) will hold when the column is

                           22

-------
transferred to the GC-MS.  The resulting chromatogram
(or computer-reconstructed chromatogram) will closely
approximate the chromatogram obtained on the auxiliary
GC.
                       Solvents

Chloroform and methylene chloride are excellent
solvents for extraction.  They give smaller and
narrower solvent peaks than hydrocarbon solvents in
GC using flame ionization detection.  Unfortunately in
the inlet of a MS their ionization cross-sections are
such that they easily trip the electronic protection
circuit and cause automatic machine shutdown.  As a
result, no more than 1.8 yJl of chloroform or methylene
chloride solutions can be injected into our GC-MS
system without taking special precautions.  With
solvents such as toluene, hexane, and other hydro-
carbons , which do not cause this problem, up to 3 y£
samples can be injected.
                Temperature Programming

Temperature-programmed GC is nearly always used for
GC-MS analysis.  A medium program rate of 4° per
minute is common.  Frequently the program is extended
to the upper limit recommended for the liquid phase.
At the resulting high temperatures, considerable bleed
occurs with many columns.  For example, MS fragments
at m/e 73, 147, and 221 resulting from dimethyl
silicones appear when the usually stable silicone
columns are overheated.  The program should therefore
be terminated short of these high temperatures.

When the initial temperature is too high, many
components of interest elute with or close to the
solvent.  This may explain the absence of low
molecular weight compounds from Appendix I.
Limited experience with sub-ambient temperature
programming indicates that this approach is feasible
for the analysis of such compounds.
                           23

-------
                       SECTION VI

                    MASS SPECTROMETRY

                      Instrumental

The first GC-MS system used at SERL in 1968 was a
Perkin-Elmer/Hitachi RMU-7 double-focusing mass spectro-
meter connected to a Perkin-Elmer 900 gas chromatograph
through a Watson-Biemann separator.  The data output
was to be collected and processed by computer but this
portion of the system was never fully developed.
Without a workable computer system, a typical twenty
peak chromatogram required approximately twelve hours
of manual data handling before interpretation could
begin.  Despite these limitations, the first system
demonstrated the power of GC-MS by providing answers
to problems considered too complex for other types of
instrumental analysis (6, 19).  Based on almost three
years'experience with this system, a set of MS system
criteria was prepared for use in testing and evaluating
the GC-MS-computer systems available in 1971.  These
criteria, modified slightly  (Table 3), are still valid.

Two GC-MS-computer systems (the Varian CH-7 and the
Finnigan 1015) were shown to satisfy these criteria
(20).  Later, during 1972, an improved version of the
DuPont 21-490 system was also shown to satisfy the
criteria.  The less expensive Finnigan system, outlined
in Figure 2, was chosen and installed during 1971.

The GC, a vendor-modified Varian Model 1400 with no
independent detector, serves as a specialized inlet to
the mass spectrometer.

The Gohlke separator is an all-glass jet separator
that separates organic samples from the helium carrier
gas based on differences in the diffusion rates of the
gases in a turbulent jet.  Its small surface area
eliminates chemisorption and catalytic degradation and
thus overcomes the chief disadvantage of the Watson-
Biemann fritted glass separator.
                            25

-------
                        Table 3

            Minimum Acceptable Criteria for
                Mass Spectrometer System
Mass Spectrometer

  Scan rate

  Mass range

  Resolution

  Minimum identifiable
    level of pesticide

  Ionizing voltage range
2 sec/decade or 230 amu/sec

12-650 amu

1/400 or unit throughout range

10 ng in GC inlet


10-100 volts
Data System

  Field experience


  High speed storage
    and retrieval

  Data acquisition mode

  Calibration method

  Minimal software
Both software and hardware
  proven in the field

Magnetic tape
Continuous cyclic scan

Semi-automatic

Computer reconstructed
  chromatogram, background
  correction, and spectrum
  plotting routines
                           26

-------
ro
Vac ion
mnn
Chromatograph
Digital Equip.
Dec Tapes ""




-^
Diablo Disk



Gohlke
Separator

^>^
^s^*~- PDP fi/« —

p 1
^^-^^, vOmpUlvr -^
" |
Reduced- Data

Interpretation
B nf MantaH 7j\nae
Finnigan :-*i — Direct Probe Inlet
Moss Spectrometer -^ 	 Liquid Inlet •
•
\
System

* maUSUIGS 	 Plnttnr


ADC a DAC ^ I
Reconstructed
Chromotogroms
ond Moss Spectra

              FIGURE 2.  OUTLINE  OF FINNIGAN  GC-MS-COMPUTER  SYSTEM

-------
The Finnigan 1015 mass spectrometer is a quadrupole
instrument having a range of 1-750 amu.  Resolution is
one mass unit throughout the range (e.g., 1/20 at mass
20, but 1/625 at mass 625).  As a consequence, instru-
mental sensitivity at low mass is much higher than in
a magnetic instrument.  At a scan speed of 120 amu/sec
the sensitivity is adequate to give identifiable
spectra from 20 ng of material introduced into the GC
inlet.

The liquid inlet is used for introduction of calibrating
compounds, the direct probe for solid materials.

The System Industries interface and the analog-to-
digital and digital-to-analog converters are used to
permit the Digital Equipment Corporation (DEC) computer
to control the mass spectrometer during calibration,
data acquisition, and checking; to accept data from the
mass spectrometer; and to control the Houston plotter
during data reduction.

The DEC PDF 8/e computer, the heart of the data system,
has a 4096 word core and an ASR33 teletypewriter.
Programs, raw data, and reduced data are stored on
either the two DECtape units or the Diablo disk.
Output of reduced data is achieved under computer
control via the plotter, the teletypewriter, or a
coupling device.  The coupling device connects the
POP8 via telephone lines to a large computer and permits
semi-automatic spectrum identification by a matching
procedure described in Section VII.
                       Operation

The basic steps in processing a sample by GC-MS computer
are listed in Table 4 and can best be illustrated by
reference to an example.

The mass spectrometer is calibrated with perfluoro-
n-tributyl amine, which gives a number of characteristic
fragments in the mass range 50-614.  The computer uses
the signals from these fragments to establish the
relationship between the digital-to-analog converter
values and the m/e values.
                            28

-------
                        Table 4

        Steps in GC-MS-Computer Data Reduction


1.  Formation of calibration reference file

2.  Data acquisition

3.  Reconstructed gas chromatogram plot

4.  Manual selection of GC peak and background spectra

5.  Creation of background corrected spectra files

6.  Output of plotted spectra

7.  Interpretation of spectra


After the sample is injected into the GC,  the mass
spectrometer automatically scans its pre-set mass
range every five seconds  (or other pre-set interval).
As it does so,  the plotter draws a trace that is equi-
valent  to a GC  signal  (lower trace in Figure  3).    If
the sample was  pre-analyzed on an auxiliary GC, this
trace can be used to determine when to terminate the
run.

After the run is complete, the computer plots a recon-
structed gas chromatogram (RGC)  like that  shown in  the
upper trace in  Figure  3.  The  largest peak is automa-
tically plotted at amplitude 100 and the other peaks
are shown proportional  to it.   Each point  on  the
spectrum number scale  under  the RGC represents a
complete mass spectrum collected on magnetic  tape.
These spectra can be displayed individually and thus
the analyst can see the composition of each peak.

Consider the peak at spectrum  number 287  (marked with
an arrow in Figure 3).   The  spectrum stored on the  tape
will be that of one of the  sample compounds plus
various fragments from column  bleed, traces of air, oil,
moisture, and perhaps  a small  residual amount of material
from the peak that eluted just prior to the one at  287.
These background fragments may mask some of the peaks  of

                           29

-------
U)
O
                             109 118 129 130 1W ISO ISO 170 ISO IS! 289 210 220 230 2*1 2SI 200 2TO 280 JSO 300 410
    FIGURE 3.
RECONSTRUCTED GAS  CHROMATOGRAM  (RGC) OF A  PETROCHEMICAL PLANT
EFFLUENT  EXTRACT

-------
the sample.  To overcome this problem, the operator
selects a spectrum that contains nearly the same
background and none of the compound fragments and has
the computer subtract the background from the sample.
In this example a logical choice is the chromato-
graphic valley at spectrum number 284.  The resulting
mass spectrum plotted by the computer is shown in
Figure 4 (the compound structure and M"1" were added
manually).

The processing of the MS data that formerly required
many hours of hand calculations and graph plotting
now takes less than three hours, and most of this time
does not require operator attention.  Only 30 minutes
of operator time are required to enter into the
computer the instructions to output the reduced data
for a 20 peak chromatogram.  Elapsed instrument time
to reduce the data ranges from slightly more than one
hour for the disk system to more than two-and-a-
quarter hours for the tape system.  The time difference
is due to the time lost by the computer while searching
the tapes for the appropriate data.

The overall GC-MS-computer system as outlined in Table
4 works well; however, two obvious improvements can
be made in the computer programs.  The first is faster
data output utilizing a cathode ray tube with a hard
copy device and the second, a modification to permit
time-shared acquisition and processing of data.  With
these modifications overall data reduction time should
be reduced by half.
                 Specialized Techniques

LMRGC.  Specialized techniques of MS or data reduction
can be used to detect a specific material or class of
materials in a mixture.  The most common technique is
the generation of the limited mass reconstructed gas
chromatogram (LMRGC).  For example, in Figure 5, the
RGC shows as peaks those spectra that contain signifi-
cant numbers of any ion fragments of m/e 35 to 250.
Above the RGC is the LMRGC in which the computer was
instructed to respond only to those spectra that
contain the m/e 149 fragment.  This fragment is often

                            31

-------
SPETTHJ1
                                   207

                                 carntf
                                                                                   M
tu
                                                                                                      MODE! OUTP
                                                                                                      PMNT71
                                                                                                      PLOT?:  r
                                                                                                      FILE! AGO
                                                                                                      SPEC NO I   £87
                                                                                                      AMP EXP7I Y
                                                                                                      THRESHOLD t:  I
                                                                                                      EXPAND BY:   5
                                                                                                      HIM Xl
                                                                                                      SUBT BKGD?:  r
                                                                                                      BKGD FILEI   AGO
                                                                                                      SPEC NO I   284
                                                                                                      NOHM ON:
                                                                                                      SAVE RESULT7I Y
                                                                                     170  ISO 190  200  2VO 228  230  210 235
                                     FIGURE  4.    MASS  SPECTRUM  OF  FLUORENE

-------
                100
                 80
               §60
                 20
                         LMRGC  m/e!49
                                                                                 3561
    0
(Phthalate)
292
A
                     I  I I I  I  I  i  i i  i  i  i  i i  i  i
                   0  20  40  60  80  100  120  140 160  180 200  220 240  260  280 300  320  340  360 380
                                               SPECTRUM NUMBER
CO
u>
                    0  20  40   60  80  100  120 140  160 180  200 220 240  260  280 • 300  320 349  360 380
                                                SPECTRUM NUMBER
FIGURE  5.   LOWER:   RECONSTRUCTED  GAS  CHROMATOGRAM  OF A PETROCHEMICAL  COMPANY
              EFFLUENT.   UPPER:   LIMITED MASS  RECONSTRUCTED  GAS  CHROMATOGRAM
              INDICATES  ONLY SPECTRA 292 AND 356 CONTAIN  SIGNIFICANT  FRAGMENTS
              OF  M/E  149,  TYPICAL OF PHTHALATE ESTERS

-------
found in the spectra of phthalate esters.   The results
indicate that two of the sample peaks may be phthalates.
Similar characteristic fragments exist for several other
classes of compounds.

Limited mass monitoring is also possible in the data
acquisition step.  In this mode, the mass spectrometer
is stepped sequentially through a series of limited
mass ranges.  Longer detection times can be spent on
each individual mass than in normal runs;  hence, the
sensitivity for a given peak can be enhanced.  The
technique can also be used to eliminate peaks that
might otherwise interfere with the spectrum of
interest.  Figure 6 shows a complicated RGC from an
industrial effluent.  In Figure 7, the sample was
rerun using the limited mass range technique.  Based
on the peak patterns and the particular masses
monitored, the impurity was identified as a PCB
mixture.

Solids Probe.  Samples not amenable to GC can sometimes
be introduced without decomposition into the MS through
the solid probe inlet.  Most GC-MS instruments have
this accessary.  In one example, a blue material
isolated from a municipal sewage plant would not
chromatograph.  When introduced into the MS by the
solids probe, it was identified as copper phthalocyanine,
Since the uses of this compound are limited, the source
was soon identified as a paint factory.
                            34

-------
00
          8
          «— »

          5L

          s..
ftl
        I*.,
         o
           '?   )0   20   33  40   SO   60  73   83  93   133  110  120  133  140  150  163  173  18C
               SPECTRUM MJMBER
          FIGURE  6.   COMPLICATED  RGC FROM AN INDUSTRIAL EFFLUENT  EXTRACT

-------
OJ
        e
        8.
        Rj
iM|fiM|Ti?i|Miim
10   20   30   1C
SPECTHLH NU«R
                             70
                                                       >m^^
                                                   80   90   100  110  120  130
FIGURE  7.   RGC OF EFFLUENT EXTRACT FROM  FIGURE 6 USING LIMITED MASS TECHNIQUE
            TO ENHANCE SENSITIVITY AND REDUCE INTERFERENCES

-------
                       SECTION VII

          INTERPRETATION AND COMPUTER MATCHING

The mass spectrum provides a chemical fingerprint that
is characteristic of a material and can be interpreted
to give the structure.  The majority of organic spectra
can be interpreted through a combination of experience
and a knowledge of chemistry and mass spectral fragmen-
tation theory.  Depending on the complexity and unique-
ness of the spectrum and the analyst's familiarity with
the compound or class of compounds, interpretation can
be made in a matter of seconds or it can require hours
or even days.  Tables that list the eight or ten most
intense peaks  (21-23) enable the analyst to search for
compounds at a rate of about 10 minutes per search.
This becomes tedious if more than one or two spectra
are analyzed.  Computerized empirical spectra matching
program can identify spectra at a much faster rate.

Many matching schemes using computers have been
described in the literature.  After a thorough survey
of the literature and of EPA's needs, a research grant
was awarded to Battelle Columbus Laboratories to
modify the spectra matching system of Hertz, Hites,
and Biemann  (24).  The modifications included
increasing the speed of matching and constructing  a
data base suitable for pollutant identification.

Matching, taking advantage of the information
redundancy of mass spectra, is based only on the two
most intense peaks in every 14 amu slot.  There are
four main steps in the matching process:

     1) Screening based on molecular weight range

     2) Screening based on most intense peak of the
        unknown spectrum

     3) Pre-searching based on spectrum family

     4) Calculation  of the similarity index and
        ordering of  best matches based on peak-by-
        peak comparison of unknown to those reference
        spectra that passed the pre-search.

                            37

-------
This modified program, SEWL3P, has been evaluated and
improved during the past year.  Matching the spectrum
of an unknown against the present data base of 11,000
spectra (10,600 general spectra from the Aldermaston
collection and 400 pollutant spectra from Southeast
Environmental Research Laboratory and Battelle)
requires approximately 45 seconds.  The similarity
index gives the user an immediate indication of whether
the "best hit" is a poor match (<0.2 if no closely
related compound is in data base), one of several fair
matches (0.2-0.35 if the correct compound is not in
data base) , or a good match (>0.35 where the second
best hit is significantly lower).

In cases where the correct answer is not included in
the data base, related compounds are usually found as
the best hits.  This can lead to a rapid identification
by other means.  In one study made at the Southeast
Environmental Research Laboratory, 50% of the unknowns
present in an environmental sample were found correctly
as the best hit; 8%, as the second best; and 2% as the
third best.

Addition of new pollutant spectra to the library takes
place continuously.  As the data base grows, the success
of the system is expected to improve.

To reduce  the operator time involved in utilizing this
system, PDP8 utility programs were developed with
Battelle that enable a laboratory to transfer all of
the necessary data from the laboratory computer through
a telephone coupling device to the central matching
computer.  These programs also eliminate human errors
and prejudices in selection and  transmission of data.

At the time the National Water Contaminants Characteri-
zation Research Program and Battelle were developing
this system, a somewhat less sophisticated matching
program was developed at the National Institutes of
Health  (25) and a simple one at  Finnigan Corporation
and System Industries  (26).  A comparison of these
programs is given in Table 5.  Each has certain
strengths  that adapt it to different phases of pollu-
tant identification.
                            38

-------
                                     Table S

                 Comparison of Three Spectra  Matching Programs
Par JUTE ter
System
EPA/Battelle NIH
Finnigan/
System Industries
Basis for match
Average number of
input peaks used

Data selection
and entry
Average natch
time for best
fit  (based on
present library
size)
Output ordered
and ranked by
Effect of an
extraneous peak
Number of spectra
in data base

Expansion of  data
base

Present availa-
bility to Agency
users
Minimum hardware
required
Masses and intensi-
ties of the two
most intense peaks
in all 14-amu slots
                     -25
Semi-automatic or
manual
                     45 seconds
Similarity  index
(high-low)
Slight reduction in
similarity  index
11.000
Masses and inten-
sities selected
from the two most
intense peaks in
14-amu slots
                      Manual
Screens only.
Does not choose a
•best fit-
Screening tine
(including dialog)
approximately 2
minutes for 4
mass peaks.

File number
(low-high)
Eliminates  the
correct match
8,800
EPA program manager   User
Nationwide on
Battello time-
shared computer
Acousticoupler
S. TTY
Preferred  hardware*  Coupler,  KLBE/c,
                      and  CRT
 Estimated  cost
 per  match  (exclu-
 ding library.
 storage  costs)
                      $1.50
Nationwide on
NIH  tine-shared
computer
Acousticoupler
& TTY

Coupler  t CRT


$1.00
Masses of the
most intense peak
in all 14-amu
slots
                                                               -15
Manual (semi-
automatic version
has been written)

<15 seconds for
System 250 disk

<3 minutes for
System 150
DECtape
Number of mis-
matched peaks
 (low-high)

 Increases the
mismatch  index
by 1

 300
                                           User
 System  Industries
 PDP8 routine—
 also available on
 GE  time-shared
 computer

 TTY
                                                                CRT
                                                                Free
      •Comparisons  of  match time  and costs  made on  this  basis.
                                       39

-------
The EPA/Battelle program is the most useful because it
quickly provides ordered listings of the best matches
with minimal operator handling.  These listings
are based on detailed comparison of both masses and
intensities for all peaks in the reduced unknown
spectrum with those of the 11/000 reduced spectra of
the library.  When the correct compound is not in the
data base, the best matches have low similarity indices
but generally are related to the correct compound.

The NIH system provides sequential screenings of the
data base through comparison of the intensities of
user-selected masses from the unknown with those of
the corresponding masses in reference spectra.  This
system has been useful in suggesting possible
structural fragments associated with the selected
masses.

The Finnigan/System Industries program appears well
suited for low-cost, rapid, in-house matching of
spectra against small data bases.  The library
furnished with the system was developed for drug
enforcement use (26) and is not applicable to
environmental pollutants.  Work is underway to develop
a similar library of the 240 industrial organic
pollutants listed in Appendix 1.  When this is
completed, a better assessment of the Finnigan/System
Industries program can be made.
                           40

-------
                      SECTION VIII

                 CONFIRMATORY TECHNIQUES

Although the mass spectrum of a compound can be
described as a fingerprint, it is not always completely
unique.  The spectra of different members of homologous
series (hydrocarbons and fatty acids) and of isomers
(polychlorobiphenyls and alkanes) are usually indistin-
guishable.  In the case of m- and p-cresol, not only
are the spectra indistinguishable, but on many GC
columns the retention times are the same.

A broad range of confidence levels exist for an iden-
tification by GC-MS.  The lowest level is identifica-
tion based entirely on a chemist's interpretation of
the spectrum without any reference to a standard or
known spectrum.  Next is a poor match with a standard
spectrum or a low similarity index (<0.2) from a
machine search.  Computer matches with similarity
indices above 0.4 or close resemblance with spectra
from the literature are termed tentative identifications.
An even higher level of confidence is placed in a good
computer or literature MS match combined with a GC
retention time match with a known measured in the
laboratory.  This level is accepted as a confirmed match
in our laboratory, particularly when the compound has
been run here earlier and its spectrum placed in our
files and in the computer library.  The best identifi-
cation based solely on GC-MS is a match of both the GC
and MS data with a known run under the same conditions.
The ultimate in identification is agreement between
GC-MS and other independent methods, e.g., NMR or IR.
Unfortunately, the amounts of sample required usually
preclude the use of conventional NMR and IR in water
analysis.
                           41

-------
                      SECTION IX

                    CASE HISTORIES

Black Warrior River and Locust Fork Branch Fish Kills

Fish kills in the Black Warrior River occurred at the
Locust Fork Branch, near Birmingport, Alabama/ in
October, 1969, and again near Demopolis, Alabama, in
September, 1970.  The 1969 kill involved 750 thousand
fish and the 1970 incident killed 8 thousand fish.
Both kills were suspected to have been caused by the
spraying of malathion in conjunction with a U. S.
Corps of Engineers mosquito control program in the
area.  The presence of malathion was confirmed by GC-MS
in extracts of both the Locust Fork Branch  (Figure 8-A)
and the Black Warrior River area near Demopolis
(Figure 8-B).

The molecular ion (m/e 330) was not observed in spectra
of either of the samples or of the standard when it was
introduced from the gas chromatographic column (Figure
8-C).  Instead, the fragment (m/e 173) resulting from
cleavage of (CH-jO^PSo* was the largest significant
ion.  A small parent ion was found in a sample intro-
duced by the direct probe  (Figure 8-D).  The fragmen-
tation pattern of 8-D is significantly different from
8-C, primarily because of different sample pressures
in the ion source, illustrating the necessity of
comparing mass spectra of samples with standards under
identical conditions.  The major fragmentations of
malathion are indicated in Figure 8-D.

The results of these analyses were used in a brief
prepared by the Alabama Department of Conservation and
submitted to the Alabama State Attorney.


           Polychlorinated Biphenyls  (PCB's)

Aroclors, manufactured by the chlorination of biphenyl,
are distillation fractions containing 20 or more PCB
isomers.  They are used widely in industry and have
become ubiquitous pollutants of the aquatic environment.
We have used mass spectrometry several times to confirm
the presence of PCB's, identified tentatively by

                           43

-------
    B
                       »3


                                      MAIATHION - LOCUST CREEK



                        «   m


                                     173
                                  US
                     'ar"'''te • 'ao'" v>a—ws > •
                                      MALATHION - (LACK WARRIOR

                                                 CREiK
                                      MMATHtON STANDARD- GC/MS
                                 ''
                             0 " l*f	BS^^^S	30iS 7!B  SB"
                                        MALATHION STANDARD • DIRECT PROBE


                                  ,*$ \°
                                                               330

           "To ; «  w-^Tir—KB ' of—as—to—BS—SB—ao • a«i—ato acr  so  j»  s
FIGURE  8.    MASS SPECTRA  OF EXTRACTS AND STANDARDS

                CONTAINING  MALATHION


                                        44

-------
electron capture GC, in water and mud samples.  The
FID chromatograms of the extracts are invariably
complicated, but the specificity of the mass spectro-
meter allows unequivocable identifications of submicro-
gram amounts of PCB's,

In one case, a Florida Bay sediment extract, cleaned
on a florisil column, was analyzed for Aroclor 1254.
The GC effluent was split 1:1 between the flame detec-
tor and the MS.  The GC flame detector pattern is shown
in Figure 9.  Eleven of these peaks were shown by MS to
be PCB's.  Their retention times and chlorine numbers
(number of chlorine atoms per PCB molecule, as
determined by MS) correspond with those of an Aroclor
1254 standard, also shown in Figure 9.

Mass spectra showing several chlorine isotope clusters
for some of the PCB's from the sediment are compared
with corresponding spectra from an Aroclor 1254
standard in Figure 10.  The parent ion was observed in
all the PCB's.  The major fragmentation path is loss
of successive chlorine atoms from the parent molecule.


         Industrial Effluent Characterization

In April, 1971, EPA Division of Field Investigations,
Denver, Colorado, requested chemical characterization
of the wastewaters from seven different companies in
an industrial area of western Louisiana.

One liter of the effluent from a petrochemical company
was extracted with chloroform; the organic layer was
dried over anhydrous sodium sulfate and concentrated
in a Kuderna-Danish apparatus to about 1 ml.  Further
concentration  (to 0.25 ml) was effected by carefully
blowing dry nitrogen over the sample, which was warmed
only by body heat from the fingers.  The concentrated
sample was chromatographed on a 50 ft SCOT column
coated with carbowax 20M-TPA.

Figure 3 shows the ion current summation (ICS) and,
plotted above it, the computer-reconstructed gas
chromatogram (RGC).  Although many of the RGC peaks
exhibited a variation in height or area relative to
the FID chromatogram obtained prior to the GC-MS run,

                           45

-------
                 PCB's- SEDIMENT EXTRACT

                  6        CHLORINE
                  0        NUMBERS
                                            1
                                               •50
                                                40
                                                20
                                                  WTO
                                                  m
                                    JO
                                    Q
                                  0  3
          14
12
10
 864
-TIME,min.
FIGURE  9.   FLAME IONIZATION DETECTOR GAS CHROMATOGRAMS
            OF PCB'S FROM THE ENVIRONMENT AND IN AN
            AROCLOR STANDARD.  THE  NUMBERS ARE CHLORINE
            ATOMS PER PCS MOLECULE  DETERMINED BY MS

                             46

-------
                      PCB's-SEDIMENT EXTRACT
. . M*,4CI
*«' too


. SO

SCI





(M*-CI>




— ~TOJ 370 --V/B




BO 580
£
ftt
z
1 «
i
Jo

300 -J
4Cl M*«
100

3CI 3CI

M',3CI go
(









ji






L~4r^r4^-V






l"3DD 	 TBt
s
§60
Z
C 4O
4
1 S

^'
'



4CI

5°
3CI ;; i

!
' i i
!<
C" TZ60 28B 300 3JO HO X
PCB's-AROCLOR 1254

wo
DO


60

40

20

M**CI
?CI
(M*->CI)







~H» f.


3CI




D 	 SO """


•ins — 280
no
80
g

| 60
: z
2 40
2
20
Too
4C1




2CI
SCI




|


!
KX)
M»,5CI j. K





j


60

4C

M
! [l
. 550 240 KO 280 J ! 300 320' 340
M»



4CI
5CI


3CI ,!|
,i

~54ff 2«) 280^ 500 320 W J(
FIGURE 10.   PARTIAL MASS SPECTRA OF SOME OF THE PCB'S SHOWN IN FIGURE  9

-------
the resemblance was sufficient to relate the FID GC
peaks to their RGC counterparts.

After the RGC and desired LMRGC's were plotted,
requests for spectra, with background or interfering
peak subtraction, were logged into the computer.  When
the spectra of the individual peaks were produced by
the computer, they were identified by the spectra
matching program through the central data bank located
at the Battelle Research Institute, Columbus, Ohio.
Table 6 shows the list of compounds identified in the
plant wastewater.
       Correlation of Compounds in Natural Waters
          with Industrial Wastewater Discharges

In the summer of 1972 the Organic Analysis Unit of the
Lower Mississippi River Field Facility requested our
assistance in determining whether paper mills in that
region contributed measurably to organic pollution of
the Mississippi River.  Only one mill will be discussed
as an example.

The mill wastewater  (1.8 Si) was extracted with two
500 ml portions of chloroform.  Combined extracts were
dried over sodium sulfate and concentrated to 1 ml in
a Kuderna-Danish apparatus.  The extract was further
concentrated to 0.3 ml by blowing dry nitrogen over
the sample.

Six liters of Mississippi River water sampled 0.25 mi
above the mill  (control sample) and 6 Jl sampled 0.25
mi below the mill were each extracted with two 1 &
portions of chloroform.  Combined extracts of each
sample were dried over sodixun sulfate, concentrated to
1 ml in a Kuderna-Danish apparatus, and further concen-
trated to 0.25 ml with dry nitrogen,  in addition to
the upriver control  sample, a solvent blank was
necessary because of  the very large concentration
factors (8,OOOX) required.  Two liters of the same lot
of chloroform were therefore treated similarly.
                            48

-------
                      Table 6

        Compounds Identified in Wastewater
             of Petrochemical Company
RGC Spectrum #
(from Figure 3)             Compound Name
       2              m-xylene*
       4              p-xylene*
      10              1,5-cyclooctadiene
      16              o-xylene*
      29              isopropylbenzene  (cumene)
      36              styrene*
      47              o-ethy1toluene
      65              o-methylstyrene*
      70              diacetone alcohol
      75              indan*
      86              2-butoxyethanol
      89              3-methylstyrene
     109              indene*
     121              dimethylfuran isomer
     129              n-pentadecane
     140              1-methylindene*
     145              3-methylindene
     156              acetophenone
     160              n-hexadecane
     168              a-terpineol
     177              naphthalene*
     193              a-methylbenzyl alcohol
     202              2-methylnaphthalene*
     206              benzyl alcohol
     2io              l-methylnaphthalene*
     221              ethylnaphthalene  isomer
     233              2,6-dimethylnaphthalene*
     233              phenol*
     244              methyl ethyl  naphthalene isomer
     249              cresol isomer
     256              acenaphthene
     265              acenaphthalene
     278              methylbiphenyl isomer
     287               fluorene
     292              phthalate diester
     296               3,3-diphenylpropanol
     356              phthalate diester	

    identification was  confirmed with  a standard.

                          49

-------
The results of the GC-MS analysis are summarized in
Figures 11-14.  The solvent contained the impurities
shown in Figure 11.  Figure 12 shows that the upriver
control sample contains only one compound (guaiacol)
that is commonly found in paper mill effluents.  The
paper mill effluent contained eleven compounds  (Figure
13).  Five of these compounds were confirmed in the
downriver sample  (camphor, fenchyl alcohol, terpinene-
4-ol, alpha-terpineol and anethole isomer "C") as shown
in Figure 14.
                     Waste Treatment

The cases already discussed are examples of typical
short-term problems that confront most service labora-
tories .  They usually require a few man-days to
several man-months effort.  In addition to these,
several extensive projects based on GC-MS analysis are
in progress at SERL.  GC-MS is invaluable in studying
the effects of waste treatment practices in areas such
as municipal sewage, paper mills, textile mills, and
refineries.  In these studies, GC-MS determines

     •  the chemicals present in the raw wastewaters,

     •  which chemicals are reduced or removed by the
        treatment,

     •  which chemicals resist the treatment and are
        discharged,

     •  which new chemicals are produced by the
        treatment, and

     •  which treatment is effective with each
        compound.

Each of these studies is a major effort that requires
development of specific techniques for sampling,
extraction, cleanup, concentration, and GC-MS.  In
addition to providing answers to specific pollution
problems, these studies are also demonstrating that
GC-MS is an indispensable tool in water pollution
studies.

                            50

-------
Ui
       FIGURE  11.   COMPUTER RECONSTRUCTED  CHROMATOGRAM  OF CHLOROFORM  BLANK

-------
(Ji
                                                             ao '» JT15 ab 'M jaTSe 120 aw
                  FIGURE 12.   RGC OF UPRIVER CONTROL  SAMPLE  EXTRACT

-------
             o-t
          B.
          B
          U.
          I.
U)
          8.
          JL
          S.
              to * w  « SB
                                to  » wo iia UB ue i-w  ias iw lie
to VB «B to
                      FIGURE  13.    RGC OP PAPER MILL EFFLUENT  EXTRACT

-------
                                                   2- PHOPIWIYLTHIOIMENE
R.

9.

V
                     HEXACHLOflOCtHANE




                 N-NONYL ALDEHYW
                                             2-AcET¥LTHIOPHEHE



                                         ALPHA-TERPINEOL
    I I I  I  I  I I '" P '  I " I' I'
e   ID  28  an «  a  so  TB
    BPETTHJI MMEH
tee  UB 'i2B  lac  110 'IBB  'i» tie ri
aob
                                                                               zib
                                                                                                    ZIB
»  »(i MB  »• W W  3" '">  •"
                        FIGURE  14.     RGC  OF  DOWNSTREAM  SAMPLE  EXTRACT

-------
                        SECTION X

                    ACKNOWLEDGEMENTS

Mary Walker, Ann Alford and Mike Carter, mass spectro-
metrists at the Southeast Environmental Research
Laboratory, have made substantial contributions to
this report and to the application of GC-MS to water
pollution problems.

The data on extraction efficiency were contributed by
A. D. Thruston, Jr.

William Loy, Surveillance and Analysis Division,
Region IV, has made extensive application of GC-MS
and his contribution to Appendix One is appreciated.
                            55

-------
                         SECTION  XI


                        BIBLIOGRAPHY


 1.  Brans, H., Middleton,  F.  M., and Walton,  W.,
    "Organic  Chemical Compounds  in Raw and Filtered
    Surface Waters/1  Anal.  Chem., 23, 1160 (1951).

 2.  Burke, J. A.,  Mills, P.  A.,  and Bostwick, D.  C. ,
    "Experiments with Evaporation of Solutions of
    Chlorinated Pesticides,"  J.  Assn. of Official
    Anal.  Chemists,  49, 999 (1966).

 3.  Chiba, M.,  and Morley, H. V., "Studies of Losses  of
    Pesticides  During Sample Preparation," Ibid., 51,
    56  (1968).

 4.  Goldberg, M.  C.,  DeLong,  L., and Sinclair, M.,
    "Extraction and Concentration of Organic Solutes
    from Water,"  Anal. Chem., 45, 89  (1973).

 5.  Mills, P. A.,  Onley, J. H.,  and Gaither,  R. A.,
    "Rapid Method for Chlorinated Pesticide Residues
    in  Nonfatty Foods," J^ Assn. Official Anal. Chem.,
    46.,  186  (1963).

 6.  Garrison, A.  W., and Hill, D. H., "Organic Pollu-
    tants from  Mill Persist in Downstream Waters,"
    American Dyestuff Reporter,  61, 21  (1972).

 7.  Rosen, A. A.,  and Middleton, P. M. , "Identification
    of Petroleum Refinery Wastes in Surface Waters,"
    Anal. Chem. ,  27.,  790  (1955)  .

 8.  Paris, D.,  Personal Communication,  1972.

 9.  Webb, R.  G., and McCall, A.  C.,  "Identities of
    Polychlorinated Biphenyl Isomers  in Aroclors,
    J._ Assn.  Official Anal. Chem. ,  55,  746 (1972) .

10.   Katz, S., Pitt, W. W., Jr., Scott,  C- D.  and Rosen,
     A  A., "The Determination of Stable Organic Com-
     pounds in Waste Effluents at Microgram/Liter Levels
    by Automatic High-Resolution Ion  Exchange Chromato-
     graphy," Water Research, 6_,  1029  (1972).

                            57

-------
11.  Pierce, A. E. , Silylatign of Organic Compounds,
     Pierce Chemical Company, Rockford, 111.  (1968).

12.  Mrochek, J. E., Butts, W. C., Rainey, W. T., Jr.,
     and Burtis, C. A., "Separation and Identification
     of Urinary Constituents by Use of Multiple-
     Analytical Techniques," Clinical Chemistry, 17,
     72 (1971).

13.  Hopps, H. B.,  "Preparation and Reactions of Diazo-
     methane," Aldrichimica Acta, 3_, 9  (1970), published
     by Aldrich Chemical Company, Milwaukee, Wisconsin.

14.  Green, C. R.,  "The Reaction of Diazomethane with
     Polyhalomethanes," University Microfilms,  Incor-
     porated, Ann Arbor, Michigan (1970).

15.  Crowell, E. P., Arnovic, S. M., and Burnett, B. B.,
     "Gas Chromatographic Determination of Mono- and
     Dibasic Acids," £._ Chromatog. Sci. , 9_, 296  (1971).

16.  "Alkyl-8 Reagents" from Pierce Previews, November,
     1972, published by Pierce Chemical Co., Rockford,
     Illinois.

17.  "MethElute," Product Bulletin No.  49300  (1971) by
     Pierce Chemical Co., Rockford, Illinois.

18.  Beroza, M. and Bierl, B. A., "Rapid Determination
     of Olefin Position in Organic Compounds in Micro-
     gram Range by Ozonolysis and Gas Chromatography,"
     Anal. Chem. , 39^ 1131  (1967) .

19.  Garrison, A. W., Keith, L. H. , and Alford, A.  L.,
     "Confirmation  of Pesticide Residues by Mass
     Spectrometry and NMR Techniques," Advances  in.
     Chemistry Series, No. Ill, 26-54  (1972).

20.  Neher, Maynard B., "Evaluation of  Gas Chromatograph/
     Mass Spectrometer/Computer Systems," Battelle
     Columbus Laboratories Research Report to the Water
     Quality Office of the Environmental Protection
     Agency  (1971), Battelle Laboratories, Columbus,
     Ohio.

21.  Cornu, A., and Massat, R., Compilation of Mass
     Spectral Data, Heyden & Son Limited, London (1966).

                           58

-------
22.  Atlas of Mass Spectral Data, Vols. 1-3, Edited by
     Stenhagen, E., Abrahamsson, S., and McLafferty,
     F. W., Interscience, New York  (1969).

23.  Eight Peak Index of Mass Spectra, Vols. 1&2, Mass
     Spectrometry Data Centre, AWRE, Aldermaston,
     England (1970).

24.  Hertz/ H.  S., Kites, R. A., and Biemann, K.,
     "Identification of Mass Spectra by Computer-
     Searching a File of Known Spectra," Anal. Chem.,
     4_3, 681 (1971) .

25.  Heller, S. R. , "Conversational Mass Spectral
     Retrieval System and Its Use as an Aid in Struc-
     ture Determination," Anal. Chem., 44, 1951  (1972).

26.  Finkle, B. S., Taylor, D. M. ,  and Bonelli,  E. J. ,
     "A GC/MS Reference Data System for the Identifi-
     cation of Drugs of Abuse," J.  Chromatog. Sci. ,
     10, 312 (1972).
                           59

-------
                      SECTION XII

                      APPENDICES

No.                                                Page

1     Organic Compounds Identified in Water by      62
        Gas Chromatography-Mass Spectrometry

2     Procedure for Diazomethane Methylation        88

3     Procedure for Dimethyl Sulfate Methylation    90
                            61

-------
                                               APPENDIX ONE

                                 ORGANIC COMPOUNDS IDENTIFIED IN WATER BY
                                   GAS CHROMATOGRAPHY-MASS SPECTROMETRY
           Compound (1)
         Sample source
  Concentrations, toxicities  (4),
  	comments, analyst	
cr\
    6,8 ,11,13-Abietatetraen-18-
      oic acid (2,C) (3)

    13~Abieten-18-oic acid (3)
    Abietic acid (C) (3)
    Acenaphthalene
    Acenaphthene
                 (C)
                 (C)
    Acetophenone (C)
Paper mill's raw waste and trick-
  ling filter effluent

Paper mill's raw waste and trick-
  ling filter effluent

Paper mill's raw waste and lagoon
Petrochemical plant's five-day
  lagoon effluent

Petrochemical plant's five-day
  lagoon effluent

Wood preserving plant's lagoon
  effluent

Wood preserving plant's settling
  pond

Pesticide plant's raw effluent

Chlorinated paraffin plant's
  lagoon

Petrochemical plant's five-day
  lagoon effluent
Toxic to salmon at 2-5 mg/1  (10).
  Keith.

Toxic to salmon at 2-5 mg/1  (10).
  Keith.

Toxic to salmon at 5 mg/1  (10).
  Keith, Loy.

Keith.
Caused skin tumors in mice  (9)
  Keith.

Loy.
0.2 mg/1.  Loy.
McGuire.

0.29 mg/1; LD50 in rats is 3g/kg
  (7).  Loy.

Keith.

-------
a\
   Acetosyringone (C)


   Acetovanillone (C)


                   (C)

   2-Ac e t yl th io phe ne

   Acrylonitrile  (C)
Adiplc acid (C)


Adiponitrile (C)


Aldrin


m-Anethole

o-Anethole

p-Anethole

Anthraquinone  (C)
    Anteisomargaric acid (C)
      (3)
                              Gulf coast paper mill's settling
                                pond

                              Gulf coast paper mill's settling
                                pond

                              Paper mill's raw waste and lagoon

                              Paper mill's raw waste

                              Acrylic fiber plant's settling
                                pond
Nylon plant's raw waste


Nylon plant's raw waste


Pesticide plant's raw effluent


Paper mill's raw waste

Paper mill's raw waste

Paper mill's raw waste

Wood preserving plant's settling
  pond

Paper mill's raw waste and five-
  day lagoon
0.14 mg/1.  Loy.


Estimated at  0.025  mg/1.   Loy.


Keith, Loy.

Keith.

100 mg/1; 24  hr. TLm  to bluegllls
  is 22.5 mg/1 arid  30 mg/1  caused
  100% mortality to pin perch  (8).
  Garrison.

3.7 mg/1; 24  hr. TLm for bluegills
  is 330 mg/1 (11).  Loy.

320 mg/1; 24  hr. TLm for bluegills
  is 815 mg/1 (8).  Loy.

96 hr. TLm at 25° C is 0.01 mg/1
  for bluegill (8).  McGuire.

Keith.

Keith.

Keith.

0.05 mg/1.   Loy.


Keith.

-------
Anteisopentadecanoic acid
  (C) (3)

Arachidic acid (3)
Arachidonic acid (C) (3)

Behenic acid (C) (3)


Benzaldehyde (C)

Benzyl alcohol


2-Benzothiazole (C)


                (C)


Biphenyl


Borneo1


1-Butanol (C)


2-Butoxyethanol


n-Butylisothiocyanate (C)
Paper mill's five-day lagoon
Paper mill's raw vaste
Paper mill's five-day lagoon

Paper mill's raw effluent and
  five-day lagoon

Paper mill's raw waste

Petrochemical plant's five-day
  lagoon effluent

Latex accelerators and thickeners
  plant's holding pond

Synthetic rubber plant's aerated
  lagoon

River below textile finishing
  plant

Paper mill's raw waste and trick-
  ling filter effluent

Petrochemical (alcohols) plant's
  raw effluent

Petrochemical plant's five-day
  lagoon effluent

Latex accelerators and thickeners
  plant's holding pond
Keith


MLD for fish is 5 mg/1 of the
  sodium salt (8).  Garrison.

Keith.

Keith.


Keith.

Keith.
0.16 mg/1; LD50 for mice is 100
  mg/kg (7).  Loy.

Very disagreeable odor.  Loy.
ID50 in rats is 2.2 g/kg  (7).  Loy.
Toxic—probable human lethal dose
  is 50-500 mg (5).  Keith, Loy.

16.0 mg/1; 90 Ib/day discharge.
  Keith.

Keith.
Raw effluent 0.5 mg/1 and holding
  pond 0.1 mg/1.  Loy.

-------
a\
ui
Camphor (C)


        (C)


Caproic acid (C)


Carbazole  (C)


Chlordane


Chlordene

o-Chlorobenzoic acid  (C)
   (3)

bis-(2-Chloroethoxy)
  methane  (C)

bis-2-Chloroethyl ether
   (C)

bis-2-Chloro isopropyl
  ether (C)

trans-Communic acid  (3)


o-Cresol  (C)
                                  Paper mill's raw waste and trick-
                                    ling filter effluent

                                  Gulf  coast paper mill's settling
                                    pond

                                  Nylon plant's raw waste
                                  Wood preserving plant's settling
                                    pond

                                  Pesticide plant's raw effluent
Pesticide plant's raw waste

Chlorinated paraffin plant's
  lagoon

Synthetic rubber plant's treated
  waste

Synthetic rubber plant's treated
  waste

Glycol plant's thickening and
  sedimentation pond

Paper mill's raw waste and
  trickling filter effluent

Wood preserving plant's settling
  pond
                                    Minimum detectable  taste  is  2 mg/1
                                       (8).  Keith, Loy.

                                    0.031 mg/1.  Loy.
220 mg/1; 24 hr. TLm  for bluegills
  is 200 mg/1  (11).   Loy.

0.27 mg/1; intraperitoneal LD50  for
  rats is 200 mg/kg (9).  Loy.

96 hr. TLM at 25° C is 0.02 mg/1
  for bluegill  (8).  McGuire.

McGuire.

0.24 mg/1.  Loy.


140 mg/1.  Loy.


0.16 mg/1.  Loy


Loy.
                                                                      Toxic  to  salmon at  2-5 mg/1  (10),
                                                                        Keith.

                                                                      1.4 mg/1; 48 hr. TLm  for  fathead
                                                                        minnows is 24 mg/1; odor thres-
                                                                        hold is 0.07 mg/1 at 30° C  (8).
                                                                        Loy.

-------
    o-Cresol  (C)


    m-Cresol  (C)



    p-Cresol  (C)


    Cumene (isopropylbenzene)


    Cyclohexano1  (C)


    1,5-Cyclooc tad iene


o\   p-Cymene  (C)

        it

    Decane (C)

    1-Decanol (C)


    Dehydroabietic acid  (C)  (3)



                        (C)  (3)
Petroreflnery's eight-hour
  lagoon effluent

Wood preserving plant's settling
  pond
Paper mill's raw waste and lagoon
Petrochemical plant's five-day
  lagoon effluent

Nylon plant's raw waste
Petrochemical plant's five-day
  lagoon effluent

Paper mill's raw waste and trick-
  ling filter effluent

Pesticide plant's raw waste

Polyolefin plant's lagoon

Petrochemical (alcohols) plant's
  raw effluent

Wood preserving plant's settling
  pond
Paper mill's raw waste and trick-
  ling filter effluent
0.120 mg/1; 300 Ib/day discharge.
  Keith.

2.5 mg/1; 24 hr. TLm for carp is
  24 mg/1; odor threshold is 0.33
  mg/1 at 30° C (8).  Loy.

More toxic than phenol (5); odor
  threshold 0.05 mg/1 (8).  Loy.

Keith.
LD50 to rats is 1-10 g/kg  (12).
  Loy.

Keith.
Keith.


McGuire.

0.03 mg/1.  Keith.

2.5 mg/1; 15 Ib/day discharge.
  Keith.

0.02 mg/1; the  sodium  salt  is  toxic
  to salmon at  5 mg/1  (10);  LD50
  in rats is Ig/kg  (7).   Loy.

Keith.

-------
Dehydroabietic acid (C) (3)
Gulf coast paper mill's settling
  pond
                                                                  0.47 mg/1.  Loy.
                    (C) (3)   Tall oil refinery's settling pond   Loy.
Diacetone alcohol
4,4'-Diamino-dicyclohexyl
  methane
Dibenzofuran (C)


             (C)


             (C)

2,3-Dibromo-l-propanol (C)


Dibromopropene isomer


Dibutylamine (C)


Dieldrin  (C)
Petrochemical plant's five-day
  lagoon effluent

Nylon and polyester plant's
  effluent after neutralization
  and sedimentation

Wood preserving plant's settling
  pond

Wood preserving plant's lagoon
  effluent

Nylon plant's settling pond

Acrylic fibers plant's settling
  pond

Acrylic fibers plant's settling
  pond

Latex accelerators and thickeners
  plant's raw effluent

Anaerobic lagoon of yarn finish-
  ing mill
                                                                  Causes liver damage and anemia in
                                                                    animals (5).  Keith.

                                                                  0.4 mg/1.  Loy.
                                                                  0.12 mg/1.  Loy.


                                                                  Loy.


                                                                  Garrison.

                                                                  0.5 mg/1.  Loy.


                                                                  Garr ison.
                                                                  Less than 1 mg/1; LD50 in rats is
                                                                    550 mg/kg (7).  Loy.

                                                                  48 hr. TLm for bluegill is 3.4 ug/1
                                                                    and 0.3 yg/1 for marine shrimp;
                                                                    lethal to rainbow trout after
                                                                    three months' exposure to 1 yg/1
                                                                    (5).  Garrison.

-------
CP>
00
    Dieldrin
    N,N-Diethylformamide  (C)
    Diethy1 phthalate  (C)
    3,4-D ihydroxyacetophenone
      (pungenin)  (3)

    3,5-Dimethoxy-4-hydroxy-
      acetophenone  (C)  (3)
Pesticide plant's raw effluent
96 hr. TLm values for several fish
  species are 0.005-0.05 mg/1 (8).
  McGuire.
Latex accelerators and thickeners   Less than 1 mg/1.  Loy.
  plant's raw effluent

Synthetic rubber plant's settling   Loy.
  pond
Paper mill's trickling filter
  effluent

Paper mill's raw effluent and
  five-day lagoon
    2,4-Dimethyldiphenylsulfone   Nylon plant's settling pond
    Dimethyl furan isomer


    2,6-Dimethyl naphthalene
       (C)

    Dimethyl naphthalene isomer

    Dimethyl phthalate  (C)


                        (C)
Acrylic fibers plant's settling
  pond

Petrochemical plant's five-day
  lagoon effluent

Petrochemical plant's five-day
  lagoon effluent

Pesticide plant's raw effluent

Plastic (PVA) plant's settling
  pond
Probably low T&O threshold.  Keith,


Keith.


Garrison.

Garr ison.


Keith.


0.015 mg/1.  Keith


McGuire.

LD50 to rats is 8400 mg/kg (7).
  Loy.
Synthetic rubber plant's settling   Loy.
  pond

-------
VD
    Dimethyl pyridine isomer
    Dimethyl quinolirie isomers
    Dimethyl sulfone (C)
    Dimethyl sulfoxide (C)
    10,12-Dimethyl tridecanoic
      acid (3)

    4,6-Dinitro-o-cresol (C)
      (2-methyl-4,6-dinitro-
      phenol)

    2,4-Dinitrotoluene (C)
    2,6-Dinitrotoluene (C)
                       (C)

    3,4-Dinitrotoluene (C)
    Diphenylene sulfide (C)
Wood preserving plant's settling
  pond

Wood preserving plant's settling
  pond

Paper mill's raw waste and trick-
  ling filter effluent

Paper mill's raw waste and trick-
  ling filter effluent

Paper mill's five-day lagoon
Specialty chemical plant's
  effluent
Explosives (DNT) plant's raw
  waste and settling pond
  effluent

Explosives (DNT) plant's raw
  waste and settling pond
  effluent

TNT plant's raw effluent

Explosives (DNT) plant's raw
  waste and settling pond
  effluent

Wood preserving plant's settling
  pond
0.1-0.2 mg/1.  Loy.
0.1 mg/1.  Loy.
Keith.
Keith.
Keith.
18 mg/1; minimum lethal dose at
  6 hr.  was 3-4 mg/1 for minnows
  (8).  Loy.

190 mg/1 in raw waste.  Loy.
150 mg/1 in raw waste and 0.02 mg/1
  in pond effluent.  Loy.
0.68 mg/1.  Loy.

40 mg/1 in raw waste.  Loy, Garrison.



0.1 mg/1.  Loy.

-------
Diphenyl ether
3,3-Diphenylpropanol
2,6-Di-t-butyl-p-benzo-
  qu inone (C)
p-Dlthlane (C)
Dodecane  (C)
          (C)
Eicosane  (C20)  (c)
 End r in
 Ethyl  carbamate  (C)
 2-Ethyl-l-hexanol (C)
Pesticide plant's raw effluent
Petrochemical plant's five-day
  lagoon effluent

Surface drainage from closed
  waste treatment system of
  particle board plant

Synthetic rubber plant's treated
  waste

Petrorefinery's lagoon effluent
  after activated sludge treat-
  ment

Petrorefinery's eight-hour
  lagoon effluent

Paper mill's raw effluent

Petrorefinery's lagoon effluent
  after activated sludge treat-
  ment

Pesticide plant's raw effluent
Paper mill's  trickling  filter  and
   aerated lagoon

Gulf coast  paper mill's settling
   pond
Threshold odor and taste level is
  0.013 mg/1 (8).  McGuire.

Keith.


Estimated at 0.01 mg/1.  Loy.
0.12 mg/1; strong offensive odor.
  Loy.

0.22 mg/1; 0.4 Ib/day discharge.
  Keith.
0.031 mg/1; 79 Ib/day discharge.
  Keith.

Kerosene based defearner.  Keith.

0.30 mg/1; 2.9 Ib/day discharge.
  Keith.
TLm values are less  than  O.OOS mg/1
  for  six fish species  (8).
  McGuire.

Keith.
 0.019 mg/1.   LD50  to  rats is  3200
  mg/kg  (7).   Loy.

-------
2-Ethyl-l-hexanol (C)

                  (C)
Ethylidenecyclopentane

Ethyl isothiocyanate (C)


Ethyl naphthalene isomer


Ethyl naphthalene isomer

m-Ethyl phenol (C)

Ethyl phenylacetate (C)


o-Ethyl toluene


Eugenol

Fenchyl alcohol (C)


Fenchone (C)


Fluoranthene (C)
Laboratory sewage

Plastic  (PVA) plant's settling
  pond

River below textile finishing
  plant

Paper mill's raw waste

Latex accelerators & thickeners
  plant's raw effluent

Petrochemical plant's five-day
  lagoon effluent

Pesticide plant's raw effluent

Paper mill's raw waste and lagoon

Resin plant's lime treated hold-
  ing pond effluent

Petrochemical plant's five-day
  lagoon effluent
Webb.

Loy.


Loy.


Keith.

Less than 1.5 mg/1; used as a
  military poison gas (7).  Loy.

Keith.


McGuire.

Loy.

Loy.


Keith.
Paper mill's raw waste and lagoon   LD50 in rats is 2 g/kg (7).  Keith.
Paper mill's raw waste and trick-
  ling filter effluent

Paper mill's raw waste and trick-
  ling filter effluent

Wood preserving plant's settling
  pond
Taste threshold 2 mg/1 (5).   Keith,
  Loy.

Keith.
0.6 mg/1; oral LD50 for rats is
  2000 mg/kg (9).   Loy.

-------
    Fluorene (C)
to
2-Formy 1 1 hio phene

Furfural (3)


         (C)


Guaiacol (C)



         CC) (3)


Heneicosane (C2l) (C)
    Heptachlor

    Heptachloronorbornene
      isomers

    Heptadecane
Wood preserving plant's settling
  pond

Petrochemical plant's five-day
  lagoon effluent

Paper mill's raw waste

Paper mill's raw waste
                                  Synthetic rubber plant's settling
                                    pond

                                  Gulf coast paper mill's settling
                                    pond
Paper mill's raw waste and trick-
  ling filter effluent

Petrorefinery's lagoon effluent
  after activated sludge treat-
  ment

Pesticide plant's raw waste

Pesticide plant's raw effluent


Nylon plant's settling pond
                                                                  0.17 mg/1.  Loy.
                                                                      Keith.
Keith.

Ingestion of 0.06g produces per-
  sistent headache (7).  Keith.

0.002 mg/1.  Keith.
0.43 mg/1; toxic to perch at 70-80
  mg/1; odor threshold is 0.002
  mg/1 (8).  Loy.

Keith, Loy.
                                                                      0.19 mg/1; 1.8 Ib/day discharge.
                                                                        Keith.
                                                                  McGuire.

                                                                  McGuire.


                                                                  Garrison.
                (C)
                              Petrorefinery's eight-hour lagoon   0.022 mg/1;  53 Ib/day discharge.
                                effluent                            Keith.

-------
U)
Heptadecane (C)



Hexachlor epoxide

Hexachlorobenzene (C)


Hexachlorobutadiene

Hexachlorocyclopentadiene

Hexachloronorbornadiene
  isomers

Hexadecane (C)

           (C)


           (C)
    Hexadieneal

    1-Hexanol  (C)
Petrorefinery's lagoon effluent
  after activated sludge treat-
  ment

Pesticide plant's raw waste

Chlorinated solvents plant's raw
  effluent

Pesticide plant's raw effluent

Pesticide plant's raw waste

Pesticide plant's raw effluent


Nylon plant's settling pond

Petrorefinery's eight-hour
  lagoon effluent

Petrorefinery's lagoon effluent
  after activated sludge treat-
  ment

Paper mill's raw waste

Petrochemical plant's five-day
  lagoon effluent

Pesticide plant's raw effluent

Petrochemical (alcohols) plant's
  raw effluent
                                                                     0.34 mg/1; 3.3 Ib/day discharge.
                                                                       Keith.
McGuire.

Loy.


McGuire.

McGuire.

McGuire.


Garrison.

0.026 mg/1; 66 Ib/day discharge.
  Keith.

0.42 mg/1; 4.0 Ib/day discharge.
  Keith.


Kerosene based defoamer.  Keith.

Keith.


McGuire.

65.0 mg/1; 375 Ib/day discharge.
  Keith.

-------
Homovanlllic acid (3)


p-Hydroxyacetophenone (C)

p-Hydroxybenzaldehyde (C)
  (3)

o-Hydroxybenzoic acid (3)

Hydroxybiphenyl isomer

4~Hydroxy-3 methoxypropio-
  phenone  (C)  (3)

p-Hydroxythiophenol  (3)

Indan  (C)


Indene (C)


Isodrin


Isoeugenol

Isopalmitic acid  (C)  (3)

Isopentyl  alcohol  (C)

Isooctyl  phthalate  (C)

Isopimaric acid (C)  (3)
Paper mill's raw waste and five-
  day lagoon

Paper mill's raw waste and lagoon

Paper mill's raw waste and lagoon


Paper mill's raw waste

Pesticide plant's raw effluent

Paper mill's raw effluent


Paper mill's raw waste

Petrochemical plant's five-day
  lagoon effluent

Petrochemical plant's five-day
  lagoon effluent

Pesticide plant's raw effluent


Paper mill's raw waste and lagoon

Paper mill's five-day lagoon

Laboratory  sewage

Nylon plant's raw waste

Paper mill's raw waste and trick-
   ling  filter effluent
Keith.


Keith.

Keith.


Keith,

McGuire.

Keith.


Keith.

0.007 mg/1.  Keith.


0.026 mg/1.  Keith.


TLm  for bluegill  is  0.006 mg/1.
   (8).  McGuire.

Keith.

Keith.

0.17 mg/1.   Webb.

Loy.

Toxic to  salmon at 2-5 mg/1 (10).
   Keith.

-------
   Jasmone

   Lignoceric acid  (3)


   Limonene  (C)


   Linoleic  acid  (C)  (3)


   Handelic  acid  (3)

   Margaric  acid  (C)  (3)


   2-Mercaptobenzothiazole  (C)
ui
    alpha-Methylbenzyl  alcohol
    Methyl biphenyl  isomer
    Methyl 3,4-Dimethoxybenzyl
      ether (3)

    2-Methyl-4-ethyl dioxolane
      (C)

    Methyl ethyl naphthalene
      isomer
Pesticide plant's raw effluent

Paper mill's raw waste
Paper mill's raw waste and trick-
  ling filter effluent

Paper mill's raw waste and lagoon
Paper mill's raw waste

Paper mill's raw waste


Synthetic rubber plant's aerated
  lagoon
                            (C)    Paper  mill's  raw waste and lagoon
Petrochemical plant's five-day
  lagoon effluent

Petrochemical plant's five-day
  lagoon effluent

Paper mill's raw waste
Fiberglass plant's  effluent


Petrochemical  plant's  five-day
   lagoon  effluent
McGuire.

MID for fish is 5 mg/1 of  the
  sodium salt  (8).  Garrison

Keith.
MID for fish is 10 mg/1 of  the
  sodium salt  (8).  Keith.

Keith.

MLD for fish is 5 mg/1 of the
  sodium salt  (8).  Keith.

Very disagreeable odor.  Loy.
Probably highly toxic.  Keith.

Keith.


Keith.


Keith.


Distinct odor of black walnuts.
  Loy.

Keith.

-------
1-Methyl indene (C)
3-Methyl indene
1-Methyl naphthalene
                     (C)
                     (C)
                     (C)
2-Methyl naphthalene (C)
                      (C)
Methyl naphthalene  isomer
Methyl naphthalene  isomers

13-Methyl pentadecanoic
   acid  (3)
Petrochemical plant's five-day
  lagoon effluent

Petrochemical plant's five-day
  lagoon effluent

River below textile finishing
  plant

Petrorefinery'a eight-hour
  lagoon effluent

Petrochemical plant's five-day
  lagoon effluent
0.002 mg/1.  Keith.


0.003 mg/1.  Keith.


Probable lethal human dose 500-5000
  mg/kg (5).   Loy.

0.005 mg/1; 12 Ib/day discharge.
  Keith.

0.02 mg/1.  Keith.
Synthetic rubber plant's settling   0.002 mg/1.  Keith.
  pond

Petrorefinery's eight-hour lagoon   0.013 mg/1; 33 Ib/day discharge.
  effluent                            Keith.
Petrochemical plant's five-day
  lagoon effluent

Wood preserving plant's lagoon
  effluent

Pesticide plant's raw effluent

Paper mill's five-day lagoon
0.03 mg/1.  Keith.


Loy.


McGuire.

Keith.
Methyl  phenanthrene
Wood  preserving  plant's  lagoon      Loy,
   effluent

-------
-J
-J
Methyl quinoline isomers


o-Methylstyrene (C)


beta-Methylstyrene


Methyl trisulfide

Myristic acid  (C)  (3)


Naphthalene  (C)


             (C)



             (C)
   2-Naphthoic acid  (C)
   Neoabietic acid  (C)  (3)
   Nitrobenzene  (C)
                                 Wood preserving plant's settling
                                   pond

                                 Petrochemical plant's five-day
                                   effluent

                                 Petrochemical plant's five-day
                                   lagoon effluent

                                 Paper mill's raw waste

                                 Paper mill's raw waste


                                 Nylon plant's settling pond
Surface drainage from closed
  treatment of system of
  particle board plant

Petrochemical plant's five-day
  lagoon effluent

Pesticide plant's raw waste

Wood preserving plant's settling
  pond

Paper mill's raw waste
                               Chemical company's lagoon after
                                 steam stripping
                                    0.5 mg/1.   Loy.
                                    0.001 mg/1.   Keith.
                                    Keith.
Keith.

MLD for fish is 5 mg/1 of the
  sodium salt (8).  Keith.

MLD to minnows for 6 hrs. is 15
  mg/1 (8).  Garrison.

Less than 0.01 mg/1.  Loy.
0.05 mg/1.  Keith.


McGuire.

0.16 mg/1.  Loy.
Toxic  to salmon at 2-5 mg/1  (10).
  Loy, Keith.

0.11 mg/1; minimum lethal dose at
  6 hr. was 90-100 mg/1 for minnows
  (8); approximate cone, in water
  causing faint odor is 0.03 mg/1
  (8).  Loy.

-------
-J
00
2-Nitro-p-cresol (C)


o-Nitrophenol (C)



o-Nitrotoluene (C)

               (C)

               (C)


m-Nitrotoluene

p-Nitrotoluene (C)



               (C)


Nonachlor

Nonadecane  (C)



            (C)


Nonylphenol (C)
Chemical company's lagoon after
  steam stripping

Chemical company's lagoon after
  steam stripping


Paper mill's five-day lagoon

TNT plant's raw effluent

DNT plant's raw effluent


DNT plant's raw effluent

Chemical company's lagoon after
  steam stripping


DNT plant's raw effluent


Pesticide plant's raw effluent

Petrorefinery's lagoon  effluent
  after activated sludge treat-
  ment

Petrorefinery's eight-hour  lagoon
  effluent

Anaerobic  lagoon  of  yarn finishing
  mill
                                                                     9.3 mg/1.  Loy.
1.4 mg/1; minimum lethal dose at
  6 hr. was 125-130 mg/1 for
  minnows (8).  Loy.

Keith.

0.15 mg/1.  Loy.

7.8 mg/1 in raw waste and 0.012 in
  pond effluent.  Loy.

Garrison.

0.04 mg/1; minimum lethal dose at
  6 hr. was 45-50 mg/1 for minnows
  (8).  Loy.

8.8 mg/1 in raw waste.  Loy,
  Garrison.

McGuire

0.31 mg/1; 3.0 Ib/day discharge.
  Keith.
                                                                     0.013 mg/1; 33 Ib/day discharge.
                                                                       Keith.

                                                                     LD50 for rats  (orally)  is 400-1400
                                                                       mg/fcg  (6); estimated  human  lethal
                                                                       dose 500-5000 mg/kg  (5).  Garrison.

-------
Nonylphenol


Norcamphor

beta-Oclmene (C)

1-Octanol (C)


Octachlorocyclopentene

Octadecane (C)


           (C)

Oleic acid (C)


           (C)  (3)


Octylphenol


Palmitic acid  (3)


               (C)

               (C)  (3)


               (C)  (3)
River below textile finishing
  plant

Paper mill's raw waste

Paper mill's raw waste

Petrochemical (alcohols) plant's
  raw effluent

Pesticide plant's raw effluent

Petrorefinery's eight-hour lagoon
  effluent

Nylon plant's settling pond

Tall oil refinery's settling pond
Paper mill's raw waste and trick-
  ling filter effluent

River below textile finishing
  plant

Textile chemical plant's raw
  effluent

Tall oil refinery's settling pond

Paper mill's raw waste and trick-
  ling filter effluent

Gulf coast paper mill's settling
  pond
Loy.


Keith.

Loy.

19.0 mg/1; 110 Ib/day discharge.
  Keith.

McGuire.

0.017 mg/1; 43 Ib/day discharge.
  Keith.

Garrison.

MLD for fish is 5 mg/1 of the
  sodium salt (8).  Loy.

Keith, Loy.


LD50 to mice is 25-50 mg/kg  (6).
  Loy.

MLB for fish is 5 mg/1 of the
  sodium salt (8).  Garrison.

Loy.

Keith, Loy.
0.013 mg/1.  Loy.

-------
    Palmitoleic acid (C) (3)

    Pentachlorocyclopentadiene
      isomers

    Pentachloronorbornadiene
      isomer

    Pentachloronorbornene
      Isomer

    Pentachloronorbornene
      isomer

    Pentachloronorbornadiene
      epoxide isomer

    Pentachlorophenol (C)
oo
o
                      (C)  (3)


                      (C)


                      (C)


                      (C)
Paper mill's five-day lagoon        Keith.

Pesticide plant's raw effluent      McGuire.


Pesticide plant's raw effluent      McGuire.


Pesticide plant's raw effluent      McGuire.
    Pentadecane (C)
Pesticide plant's raw waste
Pesticide plant's raw waste
Latex accelerators and thickeners
  plant's holding pond
Wood preserving plant's raw
  effluent

Resin plant's lime treated
  holding pond effluent

Synthetic rubber plant's aerated
  lagoon

Wood preserving plant's lagoon
  effluent

Petrorefineryfs eight-hour
  lagoon effluent
McGuire.
McGuire.
0.4 mg/1; 0.2 to 0.6 mg/1 toxic to
  19 varieties of fish; odor thres-
  hold of 0.86 mg/1 at 30° C (8).
  Loy.

Garrison.
Loy.
Loy.
Loy.
0.030 mg/1; 76 Ib/day discharge.
  Keith.

-------
   Pentadecane  (C)
oo
                              Petrorefineryls lagoon effluent
                                after activated sludge treat-
                                ment

                              Paper mill's raw waste

                              Petrochemical plant's five-day
                                lagoon effluent
   Pentadecanoic acid  (C)  (3)    Paper mill's lagoon
Phenanthrene (C)


             (C)


Phenol (C)



   "   (C)


   11   (C)


   11   (C)


   "   (C)  (3)

Phenyl ether (C)
                                 Wood  preserving plant's  lagoon
                                    effluent

                                 Wood  preserving plant's  settling
                                    pond

                                 Laboratory  sewage
                                  Petrorefinery's  eight-hour
                                    lagoon effluent

                                  Wood  preserving  plant's  settling
                                    pond

                                  Petrochemical plant's five-day
                                    lagoon effluent

                                  Paper mill's raw waste

                                  Nylon plant's settling pond
0.49 mg/1; 4.8 Ib/day discharge.
  Keith.
                                                                     Kerosene based defoamer.   Keith.

                                                                     Keith.
MLD for fish is 5 mg/1 of the
  sodium salt (8).  Keith.

5 mg/1 killed rainbow trout and
  bluegills in 24 hrs. C8).  Loy.

1.4 mg/1.  Loy.
96 hr. TLm for bluegills is 13.5
  mg/1 and odor threshold is
  0.02-0.03 mg/1 (8).  Webb.

0.2 mg/1; 510 Ib/day discharge.
  Keith.

0.66 mg/1.  Loy.
0.06 mg/1.  Keith.
Keith.

0.05 mg/1; odor threshold  is 0.013
  mg/1  (8).  Garrison.

-------
00
o-Phenylphenol


Pimaric acid (C) (3)


             (C) (3)


beta-Pinene (C)

Pinene isomer


Polychlorinated biphenyls
   (Arochlor 1254)  (C)

2-Propionylthiophene

4-n-Propylphenol (C)

Pyrene (C)


Quinoline  (C)
Sandaracopimeric acid
  (C) (3)

Stearic acid (3)
                              River  below textile finishing
                                plant

                              Paper  mill's raw waste and  trick-
                                ling filter effluent

                              Gulf coast paper mill's settling
                                pond

                              Paper  mill's raw waste

                              Gulf coast paper mill's settling
                                pond

                              Nylon  plant's raw waste
                              Paper mill's raw waste

                              Paper mill's raw waste and lagoon

                              Wood preserving plant's settling
                                pond

                              Wood preserving plant's settling
                                pond
Paper mill's raw waste and lagoon
Textile chemical plant's raw
  effluent
                                    Probable lethal human dose 500-
                                      5000  mg/kg (5).   Loy.

                                    Toxic to salmon at 2-5 mg/1 (10).
                                      Keith, Loy.

                                    0.12 mg/1.   Loy.
                                    Loy.

                                    0.008 mg/1.   Loy.


                                    0.2 vig/1.  Loy.


                                    Keith.

                                    Loy.

                                    0.33 mg/1.  Loy.
                                                                      1.5 mg/1; oral LD50 for rats is
                                                                        460 mg/kg (9); 5 mg/1 was lethal
                                                                        to bluegills in 4 hrs. at 13° C
                                                                        (8).  Loy.

                                                                      Toxic to salmon at 2-5 mg/1 (10).
                                                                        Keith.

                                                                      MLD for fish is 5 mg/1 of the
                                                                        sodium salt (8).  Garrison.

-------
00
CO
    Stearic acid  (C)  (3)


    Styrene  (C)


            (C)


    Syringaldehyde  (C)


                   (C)  (3)

    Terpinene-4-ol

    alpha-Terpineol (C)

                    (C)
    Terpineol isomer
    Terpinolene

    1,1,2,2-Tetrachloroethane
      (C)

    Tetrachlorophenol isomer
      (3)
Gulf coast paper mill's settling
  pond

Petrochemical plant's five-day
  lagoon effluent

Synthetic rubber plant's settling
  pond

Gulf coast paper mill's settling
  pond

Paper mill's lagoon

Paper mill's raw waste

Nylon plant's settling pond

Paper mill's raw waste and trick-
  ling filter effluent

Petrochemical plant's five-day
  lagoon effluent

Gulf coast paper mill's settling
  pond

Paper mill's raw waste

Chlorinated solvents plant's
  raw effluent

Wood preserving plant's raw
  effluent
0.02 mg/1.  Loy.
0.03 mg/1; ID50 in rats is 5 g/kg
  (5).  Keith.

0.003 mg/1.  Keith.
Estimated at 0.01 mg/1.  Loy.


Keith, Loy.

Keith.

Garrison.

Keith, Loy.


Keith.


0.200 mg/1.  Loy.


Kerosene based defoamer.  Keith.

2.2 mg/1; LD50 intravenous in
  rabbits is 80 mg/kg  (7).  Keith.

Odor threshold of 0.9 mg/1 at 30° C
  (8); probable lethal human dose
  is 50-500 mg/kg (5).  Garrison.

-------
   Tetradecane (C)
               (C)
Petrorefinery's lagoon effluent
  after activated sludge treat-
  ment

Petrorefinery's eight-hour lagoon
  effluent
   Tetratuethylbenzene isomer     Pesticide plant's raw waste
oo
   2,2'-Thiodiethanol  (C)
      (Thiodiglycol)

   Toluic acid  (C)  (3)
   Trichlorobenzene  isomer
    Trichlorobenzene  isomer
    Trichlorocyclopentene
      isomers

    1,1,2-Trichloroethane  (C)
Synthetic rubber plant's treated
  waste

Chlorinated paraffin plant's
  lagoon

River below textile finishing
  plant
Textile chemical plant's raw
   effluent
 Chlorinated  solvents  plant's
   raw effluent
    Trichloroguaiacol (C)  (3)      Paper mill's raw waste
    n-Tridecane (C)
                (C)
 Petrorefinery's eight-hour
   lagoon effluent

 Petrorefinery's lagoon effluent
   after activated sludge treat-
   ment.
0.58 mg/1; 5.6 Ib/day discharge.
  Keith.
0.039 mg/1; 99 Ib/day discharge.
  Keith.

McGuire.

Estimated at 2 mg/1.  Loy.


0.24 mg/1.  Loy.
May cause liver damage; estimated
  human lethal dose is 50-500
  mg/kg (5).  Loy.

Estimated lethal human dose  is
  50-500 mg/kg  (5).  Garrison.
 Pesticide  plant's  raw  effluent      McGuire.
 5.4 mg/1;  TLm  for marine  pinperch
   is  150-175 mg/1  (5).  Loy.

 Present  in sample  toxic to  salmon
   (10).  Keith.

 0.042 mg/1; 107  Ib/day discharge.
   Keith.

 0.39  mg/1; 3.8 Ib/day discharge.
   Keith.

-------
oo
   n-Tridecane (C)

   Triethylurea (C)
   3,4,5-Trimethoxyaceto-
     phenone (C)  (3)

   2,4,6-Trimethylpyridine
     (C)

   2,4,6-Trinitrotoluene  (C)
n-Undecane

           (C)


           (C)

           (C)



Valeric acid  (C)


Vanillin  (C)


          (O


Veratraldehyde  (C)
                              Paper  mill's raw waste

                              Latex  accelerators & thickeners
                                plant's raw effluent

                              Paper  mill's raw waste and trick-
                                ling filter effluent

                              Wood preserving plant's settling
                                pond

                              TNT plant's raw effluent
Paper mill's raw waste

Petroref iriery 's eight-hour
  lagoon effluent

Polyolefin plant's lagoon

Petrorefinery's lagoon effluent
  after activated sludge treat-
  ment

Nylon plant's raw waste
                                    Kerosene based defoamer.   Keith.

                                    6.4 mg/1.  Loy.


                                    Keith.


                                    0.3 mg/1.  Loy.
0.7 mg/1; MLD for minnows over 6
  hrs. is 4 mg/1 (8).  Loy.

Keith.

0.027 mg/1; 69 Ib/day discharge.
  Keith.

0.02 mg/1.  Keith.

0.05 mg/1; 0.4 Ib/day discharge.
  Keith.
                                 Paper mill's raw waste and trick-
                                    ling  filter effluent

                                 Gulf coast paper mill's  settling
                                    pond

                                 Paper mill's raw waste &  lagoon     Keith.
500 mg/1; 48 hr. TLm for daphnia
  magna is 4-5 mg/1 (11).  Loy.

Odor threshold 0.15 mg/1 (8).
  Keith, Loy.

Estimated at 0.02 mg/1.  Loy.

-------
o-Xylene (C)


         (C)


m-Xylene (C)



p-Xylene (C)


2,5-Xylenol (C)


3,4-Xylenol (C)



3,5-Xylenol (C)
                             Synthetic resin  plant's  settling
                               pond

                             Petrochemical  plant's  five-day
                               lagoon effluent

                             Petrochemical  plant's  five-day
                               lagoon effluent
                              Petrochemical  plant's  five-day
                                lagoon  effluent

                              Wood  preserving  plant's  settling
                                pond

                              Wood  preserving  plant's  settling
                                pond
                             Wood  preserving  plant's  settling
                                pond
Loy.
0.006 mg/1.  Keith.
0.008 mg/1; MLD is 10-90 mg/1;
  taste and odor threshold 0.3-1.0
  mg/1 (8).  Keith.

0.002 mg/1.  Keith.
0.82 mg/1; behavior similar to
  3,4-xylenol (8).  Loy.

0.5 mg/1; 24 hr. TLm for carp is
  30 mg/1; odor threshold similar
  to phenol (8).  Loy.

1.5 mg/1; 24 hr. TLm for carp is
  53 mg/1; odor threshold similar
  to phenol (8).  Loy.
(1)   Arranged  alphabetically.   Prefixes  showing position (e.g. : p-, alpha-, bis-, trans-)  are not
     considered part of  the  name for  this  purpose.

(2)   Confirmed identifications  are marked  (C).   All others are  to be regarded as tentative.   Using
     only GC-MS data, identifications can  be confirmed by concurrent examination of the sample and
     a known compound for duplication of GC retention time and  MS fragmentation pattern.  As a
     substitute for the  concurrent mass  spectrum, the MS comparison can be with data from the
     literature, from computer  files, or from other reference collections.
(3)  Converted to the methyl ester or ether (usually by dlazomethane or dimethylsulfate)  before analysis.

-------
    (4)  Brief discussions of how toxicity data are obtained for water pollutants are found in
        references (8) and (11).  LD50 (lethal dose, 50%) is the amount of compound administered
        by direct feeding or injection that kills 50 percent of the test animals.  The term TLm
        (tolerance limit, median) designates the exposure concentration required to kill 50 percent
        of the test organisms within a specified time period, e.g. 96 hrs.  MID (minimum lethal dose)
        is the minimum concentration required to kill one or more of the test species.

    (5)  Gleason, M. N., Gassel, R. E., Hodge, H. C., and Smith, R. R., Chemical Toxicology of
        Commercial Products, The Williams & Wilkens Co., Baltimore, Md., 3rd Edition (1969).

    (6)  Garrison, A. W. and Hill, D. H. , American Dyestuff Reporter, 61^, 21 (1972).

    (7)  The Merck Index of Chemicals and Drugs, Merck & Co., Inc., Rahway, N.J., 7th Ed (1960).

    <8)  Water Quality Criteria, Edited by McKee, J. E. and Wolf, H. W., California State Water
        Resources Control Board, 2nd Edition (1963).

    (9)  Survey of_ Compounds Which Have Been Tested for Carcinogenic Activity, Supplement 2_, USDHEW,
        Public Health Service, Bethesda, MD. (1969).
CO
-J  (10)  Rogers, I. H., "Isolation and Chemical Identification of Toxic Components of Kraft Mill Wastes.
        presented at the joint annual meeting of the Pacific Coast and Western Technical Sections of
        the Canadian Pulp and Paper Association, Jasper, Alberta, May 25-27, 1972,

   (11)  Water Quality Criteria Data Book, Volume 3 , "Effects of Chemicals on Aquatic Life," by
        Battelle's Columbus Laboratories for the Environmental Protection Agency, Project No.
        18050 GWV (1971).

   (12)  Handbook of Analytical Toxicology, The Chemical Rubber Co., Cleveland, Ohio (1969).

-------
                      APPENDIX TWO
         PROCEDURE FOR DIAZOMETHANE METHYLATION
The apparatus is shown in Figure 15.

1.  Evaporate the sample extract just to dryness with a
stream of nitrogen in a centrifuge tube, the bottom
tube of a Kuderna-Danish apparatus/ or the sample
storage vial.  A small amount of methylene chloride may
be retained, but the presence of chloroform may produce
artifacts.  Dissolve the extract in one-half to one
milliliter of distilled-in-glass ethyl ether.

2.  Add about 5 ml of distilled-in-glass ether to the
first tube of the apparatus to saturate the nitrogen
carrier gas with ether.  Add 0.7 ml of ether, 0.7 ml of
carbitol, 2-(2-ethoxyethoxy)ethanol, 1.0 ml of 37%
aqueous KOH  (not over 2 days old), and 0.1-0.2 g of
N-methyl-N-nitros o - p-toluenesulfonamide  ("Diazald,"
Aldrich Chemical Co.) to the second tube.  The base
immediately begins to release diazomethane from the
sulfonamide.

3.  Immediately position the second test tube and adjust
the nitrogen flow to about 10 ml per minute.  Caution I
Diazomethane is an extremely toxic and explosive gas.
A good fume hood and safety glasses are mandatory.  No
chipped glassware should be used, as rough glass
surfaces catalyze decomposition of diazomethane.

4.  Position the third tube (a safety trap to prevent
reagent carry-over) and the sample tube to bubble the
nitrogen and diazomethane gas mixture through the
sample.  Continue the reaction until the  slight yellow
color of diazomethane persists in the sample solution
 (from a few  seconds to 30 minutes, depending upon the
sample concentration).  In the case of dark colored
extracts in which the diazomethane is not visible, a
reaction time of 30 minutes is recommended.

5.  Allow the esterified sample to stand  unstoppered  in
the hood for 15 to  30 minutes to allow excess diazo-
methane to escape from the ether solution.  Discard all
waste from the reaction with care  and rinse the apparatus
with acetone.  Evaporate the sample to the volume
necessary for gas chromatography.
                           88

-------
  REDUCTION VALVE
           METERING
            VALVE
   NITROGEN TANK
      GLASS OR
   STAINLESS STEEL-
     0.7mm I D
  RUBBER STOPPER



0.7cm 00
                    TUBE I
                  (I6x 150mm)
              RUBBER STOPPER


            0.7cm 0 D
            O.I cm 0 0
       TUBE 2
      (13 x 85mm)
      GENERATOR
                                                   RUBBER STOPPER
                                                 0.7cm 0 0
O.I cm 00
                                             TUBE 3
                                           (15 x 85mm)
                                             TRAP
                                      0.1 cm 0 0
                                                        TUBE 4
                                                       (I5x 85mm) or
                                                     KD CONCENTRATOR TUBE
                                                     SAMPLE
FIGURE  15.   APPARATUS FOR DIAZOMETHANE METHYLATION

                                    89

-------
                    APPENDIX THREE

      PROCEDURE FOR DIMETHYL SULFATE METHYLATION

The apparatus is shown in Figure 16.

1.  Bring the original sample (300 ml) to pH 11 and
extract with chloroform to remove neutral and basic
compounds.

2.  A 500-ml 3-neck (standard taper 24/40) round
bottom flask, equipped with a fourth neck for a ther-
mometer, is fitted with two pressure-equalizing
addition funnels, the probe of a single-probe pH meter,
and a magnetic stirrer.

3.  Nitrogen is introduced into the top of the first
addition funnel and exits from the top of the second
one.  Place forty ml of Eastman reagent grade dimethyl
sulfate into the first addition funnel and a 50%
solution of sodium hydroxide  (80 ml) into the second.

4.  Pour the sample into the  flask and flush the system
with nitrogen.

5.  After raising the temperature to  85° C, begin
dropwise addition of both the dimethylsulfate and
the sodium hydroxide solution.  Maintain temperature
between 80-90°  (Caution—exothermic reaction.  Have
ice available to add to water bath.)  and the pH
between 10.5-11.  Since diraethylsulfate is not
readily soluble in water vigorous stirring must be
used.  The addition time is about 1 hour.

6.  After all the dimethylsulfate is  added, maintain
the reaction vessel at 85-90° for an  additional 15-20
minutes and then cool to room temperature.

7.  Add 5 ml concentrated ammonium hydroxide to destroy
excess dimethylsulfate and  re-extract the reaction
mixture with chloroform  to  remove the  methyl esters of
acids  and the methyl ethers of  phenols.

8.  Dry the chloroform extract  and  evaporate to the
appropriate volume  for GC analysis.
                            90

-------
                            Nitrogen Out
                              50% Sodium Hydroxide
FIGURE  16.   APPARATUS  FOR  DIMETHYL  SULFATE METHYLATION



                                  91

-------
  SELECTED WATER
  RESOURCES ABSTRACTS
  INPUT TRANSACTION FORM
                                   w
  4.  Title   CURRENT PRACTICE  IN GC-MS ANALYSIS OF
 ORGANICS IN WATER
  ;, Authors)   Webb,  R.G., Garrison, A.W.,  Keith, L.H.,
 and McGuire,  J.M.
  9. Organization  Environmental Protection Agency
              National Environmental  Research Center—
                Corvallis
              Southeast Environmental Research Lab.
              Athens, Georgia  30601
  1?  Sponsoring O
                                    3. R

                                    6.
                                                         8. Pc.-'ormit- Qrgar at/on
                                                           Rt ortffu.
                                   10.
                                                           16020 GHP
                                   11, Contract/Z.-dfrt sir).
                                   ]13. Type- ' Repe.  "
                                      Perron Covered
  is. supplementary Notes  Environmental Protection Agency Report No.
                    EPA-R2-73-277 , August 1973.
  16.  Abstract Experiences during five years  of evaluating the application of
 gas chromatography-mass spectrometry  (GC-MS)  to wastewater analysis  at the
 Southeast Environmental Research Laboratory have resulted in the  selection
 of recommended practices for such applications.  Liquid-liquid extraction
 with solvents  such as methylene ch^ride and chloroform removed greater
 than 50 percent of compounds found in pulp mill and petrochemical waste
 at concentrations of 2 yg/& to 20 jag A.   The Kuderna-Danish evaporator was
 the most effective means of concentration after extraction.  Diazomethane
 and dimethyl sulfate proved to be the most effective of five methylation
 reagents studied.  Packed columns were effective for gas chromatography of
 simple mixtures and SCOT columns provided better overall performance for
 complex mixtures.  Computerized data reduction was essential for  practical
 use of GC-MS for samples containing many compounds.  A computerized  spec-
 tra matching program proved highly effective in identifying compounds
 contained in the computer library.  The  system was shown to be effective
 in solving  problems related to fishkills caused by pesticides, confirma-
 tion of polychlorinated biphenyl residues in water and identification of
 compounds discharged by over a dozen industries.  Over two hundred
 compounds were identified in industrial  effluents.
  i?a. Descriptors  *pollutant Identification,  *Water Sampling, *Solvent
 Extractions,  *Gas Chromatography, *Mass  Spectrometry, Industrial
 Wastes, Organic  Wastes, Data Storage and Retrieval, Water Pollution
 Sources, Water  Chemistry


  ub. identifiers   *GC~MS,  GC/MS, *Computer-aided  Organic Compound Identifica-
 tion, derivative Formation, Clean-up, Case  Histories, Kuderna-Danish
 Evaporator
  17c. COWRR Field & Group  05A
  18.  Availability
! 19.  St -taity C-ass,
,    (Report)

, 20.  Sfcur, t.y Class.
21. £j:of
   Pages

22. Price
                                              Send To:
                                              WATER RESOURCES SCIENTIFIC INFORMATION CENTER
                                              U.S. DEPARTMENT OF THE INTERIOR
                                              WASHINGTON, O. C. 2O24O
  Abstractor  R« G. Webb
           | institution S.E.  Environmental Research Lab,
\VRS1C IO2 (RE.V JUNK l«7i'.

-------