United States
Environmental Protection
Agency
Environmental Research
Laboratory
Athens G A 30613
Research and Development
EPA/600/S4-85/008 July 1985
&ERA Project Summary
Master Analytical Scheme for
Organic Compounds in Water
E. D. Pellizzari, L. S. Sheldon, J. T. Bursey, L. C. Michael, R. A. Zweidinger, and
A. W. Garrison
A Master Analytical Scheme (MAS)
has been developed for the analysis of
volatile (gas chromatographable) organ-
ic compounds in water. In developing
the MAS, it was necessary to evaluate
and modify existing analysis procedures
and develop new techniques to produce
protocols that provide for the compre-
hensive qualitative-quantitative analysis
of almost all volatile organics in many
types of water. The MAS provides for
analysis of purgeable and extractable,
as well a neutral and ionic water soluble
organics in surface and drinking waters
and in leachates and various effluents.
Nominal lower quantifiable limits range
from 0.1 /ug/L to 100 /ug/L, depending
on the chemical and physical class of
the analyte and the complexity of the
aqueous matrix. Recoveries are reported
for about 260 model compounds of a
wide variety of chemical classes and
physical properties dosed into represen-
tative samples of several types of water.
This Project Summary was developed
by EPA's Environmental Research Labo-
ratory. Athens, GA, to announce key
findings of the research project that is
fully documented in five separate re-
ports (see Project Report ordering
information at back).
Introduction
The MAS represents the first effort to
develop a comprehensive qualitative-
quantitative scheme for the analysis of
organic compounds in water. It is a set of
analytical protocols for a broad range of
organics with a wide variety of functional
groups and physical properties. These
protocols provide for the gas chroma-
tography-mass spectrometry-computer
(GC-MS-COMP) analysis of the usual
purgeable and extractable compounds,
with the addition of various neutral and
ionic water soluble compounds. In fact,
any compounds that can pass unchanged
through a gas chromatograph, or can be
derivatized to do so, are amenable to
analysis by the procedures. Recoveries
have been determined from distilled and
drinking water, industrial and municipal
effluents, and, in some cases, surface
water and energy effluents, so the proto-
cols are expected to be applicable to most
water types. One unique feature of the
MAS is its comprehensiveness. Another
is its quantitative aspect: an extensive
data base of mass spectrometer detector
responses and recovery factors for model
compounds allows computer estimation
of concentration without recourse to
standards for each compound.
Tables 1-3 of this summary provide
summarized recovery data for the chem-
ical classes applicable to each protocol.
During MAS development, recovery data
were generated for approximately 260
different model compounds of a wide
variety of chemical classes and physical
properties dosed into representative
samples of several major types of water.
Complete recovery data for the individual
analytes are given in the MAS protocols
(Volume I, Part 1).
Although designed to span the com-
plexity encountered in a variety of water
types, procedures are included in the
MAS protocols that define the water
quality and allow for optimal detection
limits for that water sample. If the nominal
detection limit for qualitative GC-MS
analysis is assumed to be 10 ng for an
organic compound, then the limits for the
MAS range from 0.1 /ug/L (e.g., volatile
organics in drinking water) to 100 /ug/L
(e.g., nonvolatile strong acids in energy
-------
effluents) depending upon the physical/
chemical class of the analyte and com-
plexity of the matrix.
The prospective user has the latitude of
applying all the protocols or just those
that cover organic group types of interest.
Thus, each protocol stands alone, con-
taining the elements for determining
water quality, collecting the sample,
adding internal standards and processing
the sample with subsequent analysis
according to prescribed GC-MS-COMP
conditions.
In developing the MAS, existing analyt-
ical techniques were evaluated and modi-
fied and new techniques were developed
to produce the comprehensive protocols.
Development was in two stages. An
interim set of protocols was developed by
October 1980; analysis of environmental
samples by these protocols revealed
several important deficiencies that were
subsequently corrected by additional ex-
perimental work. The final result is this
edition of MAS protocols: Master Analyt-
ical Scheme for Organic Compounds in
Water; Part 1, Protocols, and Part 2,
Appendices.
Two companion reports resulted from
MAS development: (1) Experimental
Development of the Master Analytical
Scheme for Organic Compounds in Water
and (2) Literature Review for Develop-
ment of the Master Analytical Scheme for
Organic Compounds in Water. The user
can refer to the experimental report for
information on techniques considered
and studied for MAS incorporation, and
on experiments dealing with technique
optimization and recovery studies. The
other report is essentially a literature
review through June 1982 on techniques
for analysis of organics in water; in an
earlier version, it was the starting point
for experimental development and will
also be of interest to many users. Neither
report is essential to MAS use, however;
the protocols report stands alone as the
handbook for implementation. Part 2 of
the protocols report ("Appendices to
Protocols") includes: Appendix A-specific
instructions on fabrication of the purge
and trap apparatus and ancillary devices
for purgeable organics; Appendix B - hard
copy of computerized relative molar
response and recovery data for standards
and analytes; and Appendix C - documen-
tation of MASQUANT computer program
for quantification of MAS data.
MAS Overview
Figure 1 depicts a flow diagram of the
procedures for implementation of the
MAS. Each step is summarized below.
Sample Handling
Seven sub-samples (one for each pro-
tocol class) are required for a comprehen-
sive sample analysis. Procedures are
prescribed in the protocols for sample
collection, storage, and preservation.
Volatile organic(VO fraction) samples are
collected in septum-capped bottles with
no headspace. Methylene chloride is
added to all extractable and ionic com-
pound samples as a bacteriocide; hexane
is used as "keeper" solvent layer for
extractable compounds. Chlorine deter-
mination indicates the level of sodium
thiosulfate necessary to stoichiometrically
reduce any residual chlorine left from
water treatment. All samples are stored
at ~4°C in the dark.
Various water quality scouting meas-
urements help in the selection of appro-
priate analytical procedures, which are
optimized according to water quality
rather than sample "type" (e.g., drinking
water or municipal effluent). Headspace
gas analysis by GC of a separate small
sample is employed to determine the
dilution necessary for VO purge and trap
analysis. Atrial shake-out with methylene
chloride of a small aliquot of the extrac-
table (WABN) sample shows whether
emulsion formation is a problem, and
thus whether the flow-under extractor
must be used. Conductivity measure-
ments indicate maximum sample volume
allowable for isolation of ionic compounds
by ion-exchange resin without exceeding
resin capacity.
Internal Standards
Prescribed deuterated internal stand-
ards (included in Table 1) are added to
each sub-sample, preferably in the field,
before processing or storage. Selection of
packaging assures that from one to nine
standards of the total of 20 will appear in
each extract for GC-MS analysis; reten-
tion times are such that the standards
span the chromatographic window in
most cases. These standards are used for
monitoring recovery during analysis, for
quantifying sample components, and for
calculating relative retention times.
The initial sets of MAS standards were
prepared by the National Bureau of
Standards. Purgeable standards, in
methanol, are packaged in glass capillary
ampoules that are placed in an empty
sample bottle, then crushed with a mag-
netic stirbar after the water sample has
been collected. For other sample aliquots,
internal standards are packaged in vials
in methanol or water solution such that
emptying the entire content of the vial
into the prescribed sample volume pro-
duces the optimum concentrations of |
standards.
Isolation of Organics
After addition of internal standards, the
seven subsamples are processed as fol-
lows. (Protocol symbols are in parenthe-
sis.)
Volatile Organics (VO)—Highly volatile
(purgeable) organics (Table 1) are ana-
lyzed by a modification of the Bellar-
Lichtenberg method (EPA's Method 624),
using a custom built purge and trap
system designed especially for capillary
GC columns. (Fabrication of this system is
described in Appendix A to the protocols.)
Sodium sulfate is used to "salt out" the
organics in a 200 mL sample, which is
purged at 30°C. More concentrated sam-
ples are first diluted to 200 mL in accord-
ance with the total concentration of
purgeable organics as indicated by GC
scouting of the separate headspace
sample. Dilution prevents saturation of
the GC-MS-COMP and decreases foam-
ing potential. Purged organic vapors are
collected on a Tenax GC sorbent trap,
from which they are thermally desorbed
into a liquid nitrogen cold trap. An
"external" standard, perfluorotoluene, is
added to the cold trap from an injection |
port system, which is installed between
the sorbent trap and the cold trap, before
desorption of the purged sample compo-
nents into the cold trap. The total conden-
sate is then flash evaporated into a fused
silica capillary for analysis by GC-MS-
COMP. Comparison of MAS signals for
the external standard with those for the
internal standards purged from the sample
allows calculation of recoveries of the
internal standards, thus monitoring per-
formance of the entire analytical opera-
tion.
Neutral Water Soluble Compounds
(NEWS)—Low molecular weight, water
soluble, non-extractable compounds
(Table 1) are purged from a 10-mL water
sample containing 20% sodium chloride
at 80°C and trapped on Tenax, using the
same equipment as for the VO fraction.
To achieve lower limits of detection for
drinking water, a 200-mL sample is con-
centrated by azeotropic distillation to
produce a 3-mL aqueous condensate
enriched in neutral organics. This con-
densate is then purged as above.
Organics Extracted at pH 8 (WABN)—
Compounds of intermediate volatility,
most of which are water insoluble (Table
1), are analyzed by batch liquid-liquid
extraction of 1 L of water sample with A
-------
Sample
Handling ~
\
Addition of
Internal
Standards
Isolation from
Aqueous Matrix
Collection (7 Sub-samples) I
Storage/preservation I
Water quality scouting measurements
(conductivity, headspace gas analysis,
emulsion index. pH. and chlorine determination)
Purgeables 12)
Extractables (2)
Other Ionic Compounds (31
Volatiles
Neutral, Water Soluble, Low -
Molecular Weight Compounds
-To Purge and Trap on Tenax GC (VO)
-[o Heated Purge and Trap (NEWS)
• Extractables
o pH 1.0-
-Tsem/Vo/af/te Strong Acids (ESSA)
~ Batch Liquid-Liquid (WABN-BL)
(separatory funnel)
o pHS.O —
13 alterna-
tive tech-
niques)
^Continuous Flow-under (WABN-FU
(emulsion prone samples)
*Sorbent Accumulator (WABN-SC)
(drinking water only)
Other Ionic Compounds
(4 fractions from _
ion-exchange resins)
o Volatile Strong Acids (VOSA)
o Nonvolatile Strong Acids (NOVA)
o Primary and Tertiary Amines (SAM-PT)
o Secondary Amines (SAM-S)
Extract
Processing'
GC-MS-COMP
Analysis —
(10 maximum
fractions)
• Derivatization ofS Fractions
o ESSA methyl esters/ethers
o VOSA benzyl esters
o NOVA methyl esters/ethers
o SAM-PT Schiff bases
o SAM-S pentafluorobenzyl amines
• Clean-up of pH 8 extractables (Silica column)
o WABN-BL 3 subtractions (WABN-BL1,
WABN-BL2.
WABN-BL3)
• Evaporation/concentration of 8 fractions
• Addition of external standard
VO thermal desorption into CGC
NEWS thermal desorption into CGC
ESSA CGC
VOSA CGC
NOVA CGC
SAM-PT CGC
SAM-S CGC
WABN CGC
o WABN-BL
(Silica
subtractions
\*WABN-BL1 ... CGC
"WABN-BL2 ... CGC
I *WABN-BL3 ... CGC
Qualitative^
Analysis
T
Quantitative
Analysis
Hi:
Computer Searches
Manual Interpretation
Manual Calculations
Operator Interactive Computer Program (MASQUANT)
Figure 1. Master analytical scheme flow diagram.
-------
Table 1 . Summary of Master Analytical Scheme Recovery Data (Including Deuterated Internal Standards)
Protocol Class Recovery Mean CV
Chemical Class Compounds Range, Recovery, Range,
(Examples) Studied % % %
Volatile (Purgeable) Organics (VO)
Aromatic Hydrocarbons
(benzene, naphthalene)
Halogenated Aromatics
(chlorobenzene; 1 ,2,4-tri-chlorobemene)
Misc. Aromatic Compounds
(aniso/e)
Aliphatic and Alicyclic Hydrocarbons
(cyclohexane; r\-tridecane)
Halogenated A liphatic Hydrocarbons
(chloroform; 1 ,4-dibromobutane)
Miscellaneous Oxygen & Sulfur Compounds
(diethyl ether; hexyl ether)
Deuterated Standards
(ds-bromoethane; 2,4,6-dyanisole;
ds-chlorobenzene; da-naphthalene)
Neutral Water Soluble Organics (NEWS)
Alcohols
(1 -propanol; 1 -heptanol)
Aldehydes
(n -butyraldehyde; crotonaldehyde)
Esters
(methyl formate; ethyl butyrate)
Ethers
(tetrahydrofuran; dioxane)
Ketones
(methyl ethyl ketone; cyclohexanone)
Nitrites
(acrylonitrile; benzonitrile)
Nitro Compounds
(nitromethane; nitrobenzene)
Deuterated Standards
(da-\-butanol; ds-nitrobenzene)
Weak Acids, Bases, and Neutrals
(WABN-SC and WABN-BL)
Weak Acids
(phenol, 2,4-dichlorophenol)
accumulator column
batch liquid-liquid
Weak Bases
(aniline; carbazole)
accumulator column
batch liquid-liquid
Alkanes
(n-decane; n-tridecane)
accumulator column
batch liquid-liquid
Aliphatic Ketones, Alcohols, and Esters
(fenchone; methyl stearate)
accumulator column
batch liquid-liquid
Misc. Aliphatic Compounds
(di-\-butyldisulfide; tributylphosphate)
accumulator column
batch liquid-liquid
9
7
1
11
7
5
4
4
3
7
2
2
4
3
2
6
12
16
15
8
11
7
9
6
4
59-113
91-106
-
44-120
77-118
70-115
57-120
106-131
72-83
45-93
50-79
65-69
57-95
70-96
93-98
55-95
49-118
53-59
40-86
42-66
45-82
49-94
49-111
40-92
57-104
85
100
68
82
90
93
90
115
79
74
65
67
83
86
96
77
71
82
64
52
62
75
71
72
75
3-35
2-30
-
1-40
4-16
3-26
10-25
18-25
25-32
4-17
14-31
2-32
6-14
5-13
7-10
5-23
5-27
0-16
6-40
6-20
13-31
1-22
10-43
2-36
9-36
Mean
CV,
%
11
15
12
16
9
11
18
22
29
7
23
17
9
9
9
11
13
7
24
15
21
10
25
18
21
Footnotes
a
a
a
a
a
a
a
b
b
c
b
b
b
b
b
d
e
d
e
d
f
d
e
d
e
-------
Table 1. (Continued)
Protocol Class
Chemical Class
(Examples)
Aromatic Hydrocarbons
(2-methylnaphthalene, pyrene)
accumulator column
batch liquid-liquid
Halogenated Aromatics
(o-chloroanisole; hexachlorobenzene)
accumulator column
batch liquid-liquid
Aromatic Aldehydes and Ketones
(o-tolualdehyde; acetophenone)
accumulator column
batch liquid-liquid
Aromatic Esters and Sulfonates
(benzylacetate. ethyl-p-to/uene-
sulfonate)
accumulator column
batch liquid-liquid
Misc. Aromatic Compounds
(nitrobenzene; tetraphenyltin)
accumulator column
batch liquid-liquid
Deuterated Standards
(dio-xylene; ds- phenol; d$- acetophenone;
d5-phenylethanol; t/s- nitrobenzene;
ds-propiophenone; da- naphthalene;
dg-acridine; d^-perylene)
accumulator column
batch liquid-liquid
Compounds
Studied
10
7
10
9
3
4
6
7
6
10
9
8
Recovery
flange.
%
60-87
48-118
56-100
43-107
88-96
43-105
46-87
55-138
47-89
48-105
55-93
40-78
Mean
Recovery,
%
74
79
73
68
92
69
69
84
74
68
78
58
CV
Range.
%
1-27
13-41
1-42
13-33
2-17
6-19
7-17
3-19
6-24
8-33
4-21
11-40
Mean
CV,
%
16
24
16
20
12
13
12
11
13
17
10
26
Footnotes
d
e
d
e
d
e
d
e
d
e
d
e
Extract able Semivolatile Strong Acids (ESSA)
Carboxylic Acids
(benzoic acid; palmitic acid)
Phenols
(2-nitro-p-cresol; pentachlorophenol)
Deuterated Standards
(di3-heptanoic acid; dy benzole acid)
Volatile Strong Acids IVOSA)
Volatile Carboxylic Acids
(acrylic acid; n-octanoic acid)
Deuterated Standards
(dj-butyric acid)
Nonvolatile Acids (NOVA)
Carboxylic Acids
(succinic acid; 2,4,5-trichloro-
phenoxyacetic acid)
Sulfonic Acids
(benzenesulfonic acid; 2-naphthalene-
sulfonic acid)
Misc. Nonvolatile Acids
(benzenephosphoric acid;
pentachlorophenol)
Deuterated Standards
(2 -naphthalenesulfonic acid- dr-HiO)
Strong Amines (SAM-PT and SAM-S)
Primary and Tertiary Amines
(n-butylamine; tri-n-butylamine)
17
5
2
63-110
88-100
65-92
89
94
79
16
1
46-90
42-87
84-110
62-140
11
58-86
65
85
64
96
102
110
72
2-20
6-12
4-34
2-45
7-45
11-50
12-41
9
4-19
9
19
14
18
23
25
14
24
9
8
9
h
h
-------
Table 1. (Continued)
Protocol Class
Chemical Class
(Examples)
Secondary Amines
(diallylamine; 2-methylpiperidine)
Deuterated Standards
(dg-butylamine; dt-phenylethylamine;
N-ethyl-2-fluorobenzylamine)
Compounds
Studied
6
1
Recovery
Range,
40-98
Mean
Recovery,
63
75
CV
Range,
20-53
Mean
CV,
36
27
Footnotes
j
/*
"Mean recoveries are for triplicate determinations from drinking water, spiked at 0.2 to 1.8 ppb (nominally 1 ppb), plus triplicate determinations from a
60/40 industrial/municipal wastewater, spiked at 30 to 87 ppb (nominally 50 ppb).
''Mean recoveries are for triplicate determinations from drinking water, spiked at 0.8 to 1.2 ppb (nominally 1 ppb), plus triplicate determinations from a
60/40 industrial/municipal wastewater, spiked at 40 to 63 ppb (nominally 50 ppb).
cMean recoveries are for triplicate determinations from 60/40 industrial/municipal wastewater only, spiked at 40 to 63 ppb (nominally 50 ppb).
aMean recoveries are for triplicate determinations from drinking water, spiked at 0.5 to 5 ppb (nominally 1 ppb), using XAD-4 resin sorbent columns.
"Mean recoveries are for triplicate determinations from a 60/40 industrial/municipal wastewater or, for about 1/4 of the total compounds, from
reagent water spiked at 15 to 50 ppb (nominally 25 ppb), using batch liquid-liquid extraction, with clean-up included.
'Mean recoveries are for triplicate determinations from reagent water only, with clean-up step included. (Interferences prevented recovery determina-
tions from wastewater.)
gMean recoveries are from triplicate deteminations from drinking water only, spiked at 50-100 ppb (nominally 55 ppb). Recoveries were not determined
from more complex waters.
nMean recoveries are for triplicate determinations from drinking water, spiked at 0.3 ppb, plus triplicate determinations from a 60/40
industrial/municipal wastewater, spiked at 120 ppb.
'Mean recoveries are for triplicate determinations from several industrial and municipal effluents.
'Mean recoveries are for triplicate determinations from three industrial and two municipal effluents spiked at 110 ppb, and including, in some cases,
triplicate determinations from drinking water spiked at 35 ppb.
^Recoveries determined for only one (dg-butylamine) of the three internal standards.
methylene chloride in a separatoryfunnel.
Adjustment of sample to pH 8.0 allows
reproducible extraction of the weak acids,
(e.g., most phenols) and weak bases (e.g.,
most anilines) as well as neutral com-
pounds.
For some samples, however, batch
liquid-liquid extraction is not suitable.
Initial trial solvent extraction in a stop-
pered graduated cylinder indicates wheth-
er emulsion formation is likely to be a
problem. For emulsion-prone samples,
continuous liquid-liquid extraction with
methylene chloride in a flow-under ex-
tractor should be used. For samples in
which the extractable organic concentra-
tion is expected to be low, such as drinking
water and some surface waters, XAD-4
resin sorbent accumulator columns are
used for sorption/concentration from 10-
15 L of water. The organics are desorbed
using methanol followed by methylene
chloride.
Organics Extracted at pH 1 (ESSA)—
Extraction at pH 8 does not efficiently
recover strong acids or bases. Strong
bases are extracted on ion-exchange
resins, but a new procedure has been
developed for semivolatile strong acids
(Table 1). This involves batch liquid-liquid
extraction of a 1-L sample with methylene
chloride at pH 10 to remove most neutrals
and bases (discarded), after which the
sample is made to pH 1.0 with HCI and the
semivolatile strong acids are extracted
with methyl-f-butyl ether. This procedure
includes most carboxylic acids and strong-
ly acidic phenols. The lower molecule
weight carboxylic acids, however, are
included in a separate volatile acids
protocol (VOSA); they are too volatile to
be efficiently recovered during liquid-
liquid extraction and subsequent extract
processing. In addition, some acids, e.g.,
sulfonic acids, are too ionic to be extracted
under these conditions and are included
in the nonvolatile strong acid analytical
protocol (NOVA).
Semivolatile strong acids are deriva-
tized with diazomethane to form the
corresponding methyl esters or ethers
before GC/MS analysis.
Other Ionic Compounds (VOSA, NO VA.
SAM)—Compounds that are easily disso-
ciated in water have not previously been
included in analytical schemes because
of difficulties with extraction and chroma-
tography. Newtechniques, however, were
developed to allow inclusion of most of
these compounds in the MAS. Ion-
exchange resins are used to separate four
classes of ionic compounds from the
sample matrix using three separate ali-
quot s of the sample.
"Volatile" strong acids (VOSA), such as
acrylic acid, octanoic acid, and other
volatile carboxylic acids (Table 1) are
separated from the water on Biorad AG
1-X8 anion exchange resin, then eluted
with sodium bisulfate in an acetone:water
solution. The volatile acids are distilled
from the eluate, converted to nonvolatile
salts, then derivatized with benzylbromide
to form benzyl esters.
"Nonvolatile" strong acids (NOVA), e.g.,
naphthalene sulfonic acids, (Table 1) are
also separated from water on Biorad AG
1-X8 resin. They are eluted with HCI in
methanol, the solvent is evaporated, and
the acids are methylated with diazo-
methane.
Strong amines (SAM), such as butyl-
amine and diallylamine (Table 1) are
isolated from the water sample on Biorad
AG 50W-X8 cation exchange resin, then
eluted with sodium hydroxide in aceto-
nitrile:water solution. The eluate is acidi-
fied, the solution is evaporated to dryness,
and the amine hydrochloride salts are
dissolved in base and extracted with
methyl-f-butyl ether. The extract is split,
half is derivatized with pentafluorobenzyl
bromide to make the pentafluorobenzyl
tertiary amines from the secondary
amines (SAM-S), and half is derivatized
with pentafluorobenzaldehyde to make
Schiff bases of the primary amines (SAM-
PT).
Tertiary amines are also separated by
this protocol and quantified (underiva-
tized) in the primary amine function.
Certain other weak bases may also be
detected in these fractions, but are
measured in the pH 8 extractable fraction
(WABN), where they are extracted more
-------
efficiently. (Quarternary amines are not
addressed by the MAS.)
Extract Processing
Extractable and ionic fractions require
further processing before GC/MS. The
necessary derivatization steps, for ex-
ample, are mentioned above and sum-
marized in Figure 1.
The pH 8 extractable fraction (WABN-
BL) of many industrial effluents will
require clean-up and sub-fractionation
before effective separation can be
achieved, even on capillary columns. First,
however, a scouting procedure is imple-
mented to determine whether clean-up is
necessary. The crude extract is analyzed
by GC using a packed column and flame
ionization detection; baseline rise relative
to a separately run standard is the
evaluation criterion. Clean-up, if neces-
sary, is on a silica gel column from which
three fractions (WABN-BL1, -BL2, and
-BL3) are eluted using pentane, meth-
ylene chloride, methanol, and their
mixtures.
Concentration of extracts for GC-MS
analysis is by Kuderna-Danish evapora-
tion down to 4 ml, followed by nitrogen
blowdown to 0.5 mL or 1.0 mL using a
modified Snyder column. "External"
standards are added to each final extract
just before GC-MS analysis to monitor
recovery of the deuterated internal stand-
ards that were added to the original water
samples. The external standard for the
purgeable fractions (VO and NEWS) is
perfluorotoluene. External standards for
all the other fractions are 2-fluorobi-
phenyl and/or 4-fluoro-2-iodotoluene.
Gas Chromatography
As shown in Figure 1, as many as 10
extracts or fractions may be obtained
from one sample if the entire MAS
protocol is applied (this may be reduced to
7 if cleanup of the pH 8 (WABN) extract is
not necessary, and if the primary and
secondary amine fractions can be mixed
for a single GC-MS analysis). Glass or
fused silica capillaries are prescribed.
Bonded phase (e.g., Durabond DB-1 or
DB-5), wide-bore, thick film (1.0 /urn), 30-
or 60-m fused silica columns are recom-
mended for inertness, stability, and
sample capacity. In all cases performance
standards (see Quality Assurance) rather
than specific columns are specified in the
analytical protocol. No more than four
different GC columns should be necessary
for the entire MAS. The analytical protocol
for each fraction prescribes optimum GC
^conditionsfor the GC-MS-COMP system.
Qualitative Analysis
Sample components are identified by
established GC-MS-COMP techniques.
No research was conducted on MAS
identification procedures. GC-MS data
are stored on tape or disk. Internal
standards in each extract are used as
reference points for retention time meas-
urements as well as for quantification.
Compounds are identified by computer
searching of mass spectra data banks or
by manual interpretation.
Quantitative Analysis
Extensive recovery studies were con-
ducted during development of the MAS.
Approximately 260 different model com-
pounds from a wide variety of chemical
classes and physical property groups were
dosed into representative samples of
several major types of water (distilled and
drinking water, and municipal and/or
industrial effluents). Recoveries were
determined and average recovery factors
were stored in a computer data bank.
Relative molar response (RMR) factors
(relative to the deuterated internal stand-
ards), based on MAS selected ion peak
areas, were also determined and stored
in the data bank. (Appendix B to the
protocols is a hard copy of these data.)
The MAS user can use these data banks
and a computer program developed for
the MAS (MASQUANT, which is docu-
mented in Appendix C to the protocols) to
calculate the concentration in the original
water sample for these model compounds
as they are identified.
If the identified sample component is
not a model compound, RMR and recovery
factors of a structurally similar compound
in the data bank may be used for an
estimate of concentration. In addition to
this obvious source of error, an additional
error is involved in using any recovery
factors from the data bank. Because
sample matrices used for recovery studies
were only representative of the various
water types, and because all recovery
values for all matrices studied were
averaged to give the factor in the data
bank, errors will occur in applying these
factors to other samples. This error is
dependent on the matrix differences
between the sample being analyzed and
the representative recovery samples. On
the other hand, separate recovery values
are given for drinking water for all MAS
fractions except nonvolatile acids. These
are more accurate than MASQUANT data
for drinking water analysis because the
MASQUANT data bank (Appendix B)
values are averages over all water types
studied. Recoveries using the WABN-FU
protocols (for industrial effluents) are so
matrix dependent that there is no repre-
sentative sample matrix for recovery
studies. The user must generate his own
recoveries for his own sample matrix in
this case.
Quality Assurance/
Quality Control
Complete quality control steps are
prescribed for the user in each MAS
protocol. Some of these steps are outlined
below.
System Performance Solution—Sta nd-
ard performance solutions are used to
check performance of the GC-MS-COMP
each day. The MAS prescribes a standard
performance solution and corresponding
criteria of acceptance for each analytical
protocol. Solutions include compounds to
measure GC peak asymmetry, separation
number, resolution, polarity, and column
acidity and basicity; capacity of the capil-
lary column; inertness of the GC to MS
transfer line; limits of detection of the
MS-COMP system; and tune of the MS.
These solutions also contain deuterated
internal standards and the external
standard appropriate to each protocol for
periodic RMR verification and, if neces-
sary, determination of the RMR correction
factor by linear regression.
Internal and External Standards—Com-
parison of the recovered quantity of
deuterated internal standards to the
quantity of external standard added to the
extract just before GC/MS analysis
reveals recovery deficiencies, thus serv-
ing as a check to indicate malfunctions of
the MAS analytical procedure. The
primary use of deuterated internal stand-
ards is for quantification; reference to an
internal standard is generally accepted as
the most accurate quantitation technique
available for the broad spectrum GC-MS
analysis of organics in water. Internal
standards may also be used as retention
time indices to aid in compound identifi-
cation.
Sample Scouting—As described earlier,
several sample scouting measurements
are prescribed to characterize water
quality and in turn allow selection of the
appropriate and optimal analytical tech-
niques for a particular water sample.
Scouting measurements also help in
determining optimum sample sizes and
dilution factors for certain protocols.
Blanks, Controls. Duplicates, and Sur-
rogate Samples—Requirements and pro-
cedures for field and laboratory blanks,
spiked field and laboratory controls, and
-------
duplicate and surrogate samples are
specified in each sampling protocol. Pro-
cedures for cleaning glassware and
apparatus and other steps to assure
quality of measurement are also specified
throughout the MAS.
MAS Test Samples—for each MAS
protocol, instructions are given for pre-
paring control samples (for quality assur-
ance) by dosing known amounts of
analytes into reagent water. Test samples
for practicing and checking MAS proce-
dures may be prepared in the same way.
Resource Requirements for the
MAS
A very preliminary estimate of time per
comprehensive MAS sample, or a corre-
sponding quality assurance sample, is 80
hours. This is for a laboratory analyzing
only a few, say 10 to 50, samples per year,
using personnel who are experienced in
trace organic analysis of water and set up
with the equipment and techniques used
for the MAS.
It should be remembered that the MAS
protocols were developed and designed
as separate entities so that a laboratory
could analyze only the fractions appropri-
ate to its mission. The cost for analyzing
pH 8 extractables, for example, might be
only 10% of that for a comprehensive
MAS analysis.
Recovery and Precision
Tables 1-3 provide summarized recov-
ery data for the chemical classes corre-
sponding to each MAS protocol. Footnotes
to Table 1 give informaiton on sample
matrices and spiking levels used for
recovery studies. Several observations
can be made regarding these data (see
Table 2): (1) recoveries for volatile (purge-
able) organics are highest (these purge-
ables data are for a wide variety of
compound classes from several types of
water); (2) neutral water soluble organics,
a new class of organic analytes, are
recovered adequately with adequate pre-
cision; and (3) two classes of ionic com-
pounds, volatile strong acids and strong
amines, are characterized by relatively
low recoveries and poor precision. It is
also seen from Table 2 that recovery of
organics using accumulator columns is
better than with batch liquid-liquid
extraction in a separatory funnel, and that
precision is also better. Matrix effects
may be more important than the extrac-
tion technique, however; only drinking
water was extracted by accumulator
column, whereas more complex matrices
were extracted by the batch technique.
Table 2.
Summary of All MAS Recovery Data by Protocol Class
Protocol Class
Volatile (Purgeable) Organics (VO)
Neutral Water Soluble Organics (NEWS)
Weak Acids, Bases, and Neutrals (WABN)
accumulator column (WABN-SC)
batch liquid-liquid fWABN-BLl
Extract able Semivolatile Strong Acids (ESS A)
Volatile Strong Acids (VOSA)
Nonvolatile Acids (NOVA}
Strong Amines (SAM-PT and SAM-S)
No.
Compounds
44
27
87bf
95"
24°
17
14"
18
Mean
Recovery,*%
89
84
74C
69
89C
66
85"
69
Mean
CV."%
13
14
12°
20
9C
19
20"
28
327"
76°
16"
"Unweighted means are given, i. e., the n value for each chemical class within a protocol was not
included in the calculations.
^Sixty-nine compounds were used for both accumulator column and batch liquid-liquid recovery
studies; the total number of different compounds in this table is therefore 258.
CESSA and WABN-SC recovery data are for drinking water only.
"NOVA recovery data are for industrial and municipal effluents only.
"Overall mean recoveries and mean CVs were calculated from values for the 327 individual
compounds.
Table 3. Summary of MAS Recovery Data for Organics in Drinking Water by Protocol Class'
Nominal
Spiking Spiking Mean Mean
No. Range Level Recovery* CV°
Protocol Class Compounds (ppbf (ppbf % %
Volatile (Purgeable) Organics (VO)
Neutral Water Soluble Organics (NEWS)
Weak Acids, Bases, and Neutrals
52
25
87
0.2-1.8 1
0.8-1.2 1
0.5-5 1
90
84
74
10
16
12
(WABN-SC, accumulator column)
Extractable Semivolatile Strong
Acids (ESSA)
Volatile Strong Acids (VOSA)
Strong Amines (SAM-PT and SAM-S)
24
18
11
50-100 55
0.3
35
217
89
82
81
82"
10
1JL
12"
aFor triplicate determinations from drinking water. Non volatile A cids (NO VA) were not determined
in drinking water.
''Level spiked into water sample
^Unweighted means are given, i. e.. the n value for each chemical class within a protocol was not
included in the calculations.
"Overall mean recoveries and mean CVs were calculated from values for the 217 individual
compounds.
Table 3 shows summarized recovery
data for organics in drinking water only,
by MAS protocol. (These data were inte-
grated into the total recovery data of
Tables 1 and 2.) Spiking ranges and
nominal spiking levels are significantly
lower than those used for recoveries from
industrial and municipal effluents (see
footnotes to Table 1). Recoveries for VO
and NEWS compounds are practically the
same as those given in Table 2; i.e.,
matrix effects or spiking levels did not
make a significant difference in the
summarized data. Extractable Semivolatile
strong acids were recovered well from
drinking water, with good precision, but.
8
-------
the spiking level was relatively high. The
other ionic classes of organics(VOSA and
SAM) were recovered at significantly
higher levels and with better precision
from drinking water than from other
matrices (cf. Table 2).
Mean recoveries over all protocols for
water types studied (Table 2) for 327
determinations (258 different compounds)
was 76% with a mean relative standard
deviation (for 3 or more measurements)
of 16%. For drinking water (Table 3), the
mean recovery for 217 spiked compounds
was 82%, with a mean RSD of 12%.
Chapter 1 of the MAS protocols includes
much more recovery data than is given in
Tables 1 -3. Recovery values are given for
each individual analyte, and separate
recovery values are given for drinking
water for all MAS fractions except non-
volatile acids. As mentioned above, re-
coveries using the WABN-FU protocols
(for industrial effluents) are so matrix
dependent that there is no representative
sampling matrix for recovery studies, and
no recovery data are provided.
Conclusions
The Master Analytical Scheme proto-
cols for analysis of volatile organic
compounds in water have been developed
and recoveries have been established.
The MAS is unique in its comprehensive
scope—no other collection of protocols
exists that includes such a broad spectrum
of organic compounds. This complete set
of protocols may be applied for a survey
analysis, or each protocol may be used as
a separate entity for analysis of organic
fractions of particular interest. The main
application of the MAS will be for the
analysis of carefully selected samples to
answer the question, "What compounds
are present above detection limits and
approximately how much of each is
present?" Applying the MAS should be
cost effective in areas such as:
• Drinking Water—In epidemiological
studies and as early warning for toxic
pollutants below acutely toxic levels.
• Industrial/Municipal Wastewaters—In
wasteload allocations, permit applica-
tion evaluation and long-range projec-
tions for the state of the environment.
• Surface Waters—In trends analysis,
assessments of abatement program
effectiveness, watershed management
(including exposure assessment), and
incident investigation.
• Landfill Leachates—In exposure as-
sessment, evaluation of landfill per-
formance, and diagnosis of problems.
• Environmental Processes—In chemical
characterization of aqueous sources
and discharges from natural processes
and treatment systems.
E. D. Pellizzari, L S. Sheldon. J. T. Bursey, L C. Michael, andR. A. Zweidinger are
with Research Triangle Institute, Research Triangle Park, NC 27709.
A. W. Garrison is the EPA Project Officer (see below).
This Project Summary covers the following reports:
"Master Analytical Scheme for Organic Compounds in Water: Part 1.
Protocols," (Order No. PB 85-154 367/AS; Cost: $28.00. subject to change).
"Master Analytical Scheme for Organic Compounds in Water: Part 2.
Appendices to Protocols," (Order No. PB 85-204 360/AS; Cost: $20.50,
subject to change).
"Literature Review for Development of the Master Analytical Scheme for
Organic Compounds in Water, "(Order No. PB 85-152 874/AS; Cost: $26.50,
subject to change).
"Experimental Development of the Master Analytical Scheme for Organic
Compounds in Water: Part 1," (Order No. PB 85-153 096/AS; Cost: $56.50,
subject to change).
"Experimental Development of the Master Analytical Scheme for Organic
Compounds in Water: Part 2," (Order No. PB 85-153 088/AS; Cost: 53.50,
subject to change).
The above reports will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Research Laboratory
U.S. Environmental Protection Agency
Athens, GA 30613
U. S. GOVERNMENT PRINTING OFFICE:1985/559-l 11/20604
-------
=•
oo
O)
o
o
CO
o
m
8
00
-D C/
m -
3D >
>
c
------- |