United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-82-048 Sept. 1982
Project Summary
Laboratory Evaluation of
Level 1 Organic Analysis
Procedures
Judith C. Harris, Zoe A. Grosser, Philip L Levins, Debra J. Sorlin, and Clifford
H. Summers
Level 1 is the first stage in a three-
tiered approach to performing an
environmental source assessment.
Level 1 is designed to provide enough
information about the composition of
effluent and process streams to permit
them to be ranked in order of priority
for probable environmental hazard. A
set of sampling and analysis procedures
designed to achieve these objectives
was developed and published in EPA
report EPA-600/2-76-160a in June
1976. That methodology was based
on previously available laboratory
procedures, which had not, however,
been specifically tested to determine
their suitability for this particular
purpose. The overall objective of the
work described in this report was to
evaluate the Level 1 organic sampling
and analysis procedures as proposed
in the June 1976 manual. Priorities of
this study included: development of a
resource of information concerning
the behavior of compounds and
classes of compounds when subjected
to Level 1 procedures, and identifica-
tion of problems and limitations of the
proposed procedures that might require
revisions in methodology. This report
gives results of a series of experimental
studies of the organic analysis proce-
dures as proposed and as eventually
modified. The report also includes
several examples of Level 1 organic
analysis data for samples analyzed
according to the revised procedures.
The results for coal, fuel oil, and SASS
train samples of an actual emission
source are presented in the Level 1 re-
port format.
This Project Summary was developed
by EPA's Industrial Environmental
Research Laboratory. Research Triangle
Park, NC. to announce key findings of
the research project that is fully
documented in a separate report of
the same title (see Project Report
ordering information at back).
Introduction
EPA/IERL-RTP's Technical Support
Staff has developed a three-tiered or
phased approach to performing an
environmental source assessment. The
first phase. Level 1, is designed to
provide enough information about the
composition of effluent and process
streams to permit them to be ranked in
order of priority for probable environ-
mental hazard. The Level 1 assessment
is intended to: 1) provide preliminary
environmental assessment data, 2)
identify principal problem areas, and 3)
provide the data needed for prioritiza-
tion of energy and industrial processes,
streams within a.process, components
within a stream, and classes of materials,
for further consideration in the overall
assessment. The second phase of
sampling and analysis, Level 2, is
designed to provide additional informa-
tion that will confirm and expand the
information gathered in Level 1. The
third phase, Level 3, will permit quanti-
-------
tative monitoring of specific pollutants
identified in Level 2.
A set of sampling and analysis
procedures designed to achieve the
objectives of Level 1 environmental
assessment was developed and published
in an EPA report in June 1976.1 The
Level 1 procedures are comprehensive,
covering inorganic and organic chemical
analysis, and biological testing. The
overall objective of the work described
in this report was to evaluate the Level 1
organic sampling and analysis proce-
dures as proposed in the June 1976
manual. This methodology was designed
to identify the major classes of organic
compounds present in a process or
effluent stream and to estimate their
concentrations. Priorities of this study
included: development of a resource of
information concerning the behavior of
compounds and classes of compounds
when subjected to Level 1 procedures,
and identification of problems and
limitations of the proposed procedures
that might require revision in meth-
odology.
The report gives results of a series of
experimental studies of the organic
analysis procedures as proposed and as
eventually modified. The report also
includes several examples of Level 1
organic analysis data for samples
analyzed according to the revised
procedures. The results for coal, fuel oil,
and SASS train samples of an actual
emission source are presented in the
full Level 1 format in the complete
report; selected examples are included
in this summary.
An integrated overview of the evaluated
and revised procedures are in Chapter 9
of the second edition (October 1978) of
the Level 1 procedures manual.2 The
Level 1 organic analysis methodology
includes procedures for sample prep-
aration, an open-column liquid chro-
matographic (LC) procedure to separate
the extract into fractions prior to
analysis, quantitative analysis of total
organic content by gas chromatography
(TCO) and gravimetry (GRAV), and
qualitative analysis of compound types
by infrared (IR) and low resolution mass
spectrometry (LRMS).
Summary of Experimental
Studies
The following paragraphs briefly
summarize each experimental study
performed in this work.
Field GC Analysis of Organic
Gases
The original Level 1 procedures'
recommended that organic gases (species
with boiling points <100°C) be collected
in glass sampling bulbs and analyzed
on-site by gas chromatography (Field
GC procedure). Although the possibility
of substituting polymeric plastic sam-
pling bags was advocated, systematic
studies3 indicated that glass bulbs were
the sampling system of choice for Level
1. Time-integrated sampling of gases
with evacuated glass bulbs can be
achieved if a glass fiber filter and a 0.2
Lpm critical flow orifice are used
upstream of the bulb. The original GC
procedures for gas chromatographic
analysis of sulfur gases and gaseous
hydrocarbons were reviewed and found
to be unsatisfactory for Level 1. Revised
procedures using alternative columns
and GC temperature programs were
tested and recommended as replace-
ments.
Preparation and
Characterization of XAD-2
Resin
A macroreticular crosslinked polysty-
rene resin, XAD-2, was selected for the
SASS train organic vapor sampling
module because of its high collection
efficiency for a variety of organic
compounds. The optimum procedures
for preparation (cleanup) and analysisof
the blank for XAD-2 resin were not
specified in the June 1976 manual1.
Two different resin preparation proce-
dures (water + methanol + pentane and
water + methanol + methylene chloride)
were compared to the original 4-step
procedure (water + methanol + diethyl
ether + pentane). Parameters investigated
include the surface properties of resin
(pore volume and pore size distribution)
and the quantity and nature of residual
solvent extractable material. Further
experiments examined the possibility of
resin self-contamination by thermal
processes (at 20°, 40°, and 60°C) and
the efficiency of recovery of spiked
materials. Based on these studies, the
new recommended procedure is to
clean the XAD-2 resin with the water+
methanol+ methylene chloride overnight
extraction sequence and to extract the
collected SASS sample with methylene
chloride.
Extraction of Aqueous
Samples
The June 1976 manual1 suggested
that aqueous samples be adjusted to
neutral pH and then sequentially
extracted with three 500-mL portions of
methylene chloride for a 10-L water
sample. A literature study of the
efficiency to be expected from methylene
chloride solvent extraction of aqueous
samples indicated that the original
Level 1 procedure1, which called for
three extractions at neutral pH, was
inadequate for Level 1, since moderately
strong organic acids and bases would be
<0.1% extracted. A revised procedure
for Level 1 extractions at both acidic and
basic pH was proposed and evaluated.
Neutral species, whose recovery is
independent of pH, are essentially
extracted four times by this procedure;
acidic and basic species are each
extracted twice. It had been suggested
that ether may be superior to methylene
chloride for extracting certain acidic
compounds, such as phenols or car-
boxylic acids. This was tested experi-
mentally to see if the advantage was
significant. Four of the five model
compounds tested were extracted
efficiently with either methylene chloride
or ether. One compound, phenol, was
extracted four times more efficiently by
ether than by methylene chloride
(yielding an 82% recovery versus only
20%), although higher homotogs of
phenol (such as cresols) are adequately
extractable with methylene chloride.
The improvement in efficiency for a few
specific compounds did not appear to
warrant a change to the less convenient
(lighter than water) solvent; methylene
chloride remained the recommended
Level 1 solvent.
Extraction of Sludge/Slurry
Samples
The June 1976 manual1 included
no explicit Level 1 procedures for
preparation of sludge/slurry samples
prior to organic analysis. This sample
category can span a broad range,
including slurries and solids or semisolid
sludges containing 95% water. Some of
these materials are very difficult to
handle and no one procedure will work
for all of them. Nevertheless, it was
desirable to define a sample preparation
protocol that could be applied consistently
and that would minimize variability. A
protocol was tested using a variety of
complex sludge/slurry samples. It
involves, in most cases, tests on small
portions of the sample to determine the
best procedure prior to committing the
entire sample. The basic approach is to
determine whether the sample is best
-------
treated as a solid, as a liquid, or by a
combination of procedures.
Analyses of Volatile Species
in Organic Extracts: TOO and
Solvent Exchange
The original Level 1 organic analysis
procedure, specified in the June 1976
manual1, was designed to provide
qualitative compound class identification
and quantitative (gravimetric) data for
components of a sample that are
retained when an organic sample
extract is evaporated to dryness. In the
experimental investigation of the LC-IR-
LRMS organic analysis procedure, it
became apparent that the range of
material lost when the sample extract
was evaporated to dryness for the
gravimetric analysis and preparation of
the LC sample was considerably higher
than expected. Quantitative retention of
the model compounds appeared to be
achievable only for species with boiling
points of about 300°C and above. At
about the same time, it was recognized
that many, if not most, of the organic
compounds that were considered to be
of primary concern in environmental
assessment [e.g., the compounds on the
Multimedia Environmental Goals (MEG)
list4], were in the <300°C boiling point
range. Preliminary studies indicated
that the dramatic loss of moderately
volatile material occurs only after the
sample has been evaporated to dryness;
concentration of an extract by a factor of
10to 100(1 orO.1 mL final volume from
10 mL extract) produces less drastic
losses.
It was decided to modify the
procedures for preparation of sample
extracts to include a solvent exchange
step prior to LC. This avoids evaporating
the sample to dryness and thus allows
acquisition of qualitative chemical
information about species in the
<300°C boiling point range. It was also
necessary to specify an alternative
procedure for quantitative analysis' of
materials in the 100-300°C range; a
procedure based on gas chromatography
with a flame ionization detector (FID)
was investigated. A satisfactory column
for this Total Chromatographable
Organics (TCO) was found to be 1.8 m x
3mm(6ftxy8-in.)O.D. 10%OV-101 on
100/120 mesh Supelcoport. The GC
was operated isothermally at about
30°C — or room temperature — for 5
minutes after sample injection and then
programmed rapidly to 250°C and held
as long as necessary. From a calibration
curve based on n-hvdrocarbons and the
total integrated area of a GC trace, the
TCO of the example environmental
sample extract is estimated.
Elution Patterns in Level 1
LC
Several sets of mixtures of pure
compounds were prepared to evaluate
the elution patterns of the Level 1 LC
separation.1 The mixtures were
generally applied to the columns at
levels corresponding to <25 mg of each
compound to avoid overloading the
column and therefore distorting the
separation.
A table, showing results of LC elution
pattern studies of 17 model compounds,
is given in the full report. The data
illustrate a phenomenon borne out in
subsequent work: it is uncommon to
find any given material isolated in one
LC fraction. Several years of cumulative
experience with the Level 1 LC
separation have led to the inference that
the band-broadening in this low
resolution chromatographic method
leads to elution peaks that are about 1
LC fraction wide.
There was concern that the presence
of water in Level 1 organic extracts
might lead to irreproducible deactivation
of the silica gel and, therefore, to
irreproducible results of the LC
separation. Experimental results
indicated that sodium sulfate may be
used as a dryer for methylene chloride
extracts without causing unacceptable
losses of sample components.
It was also suspected that the most
polar solvents used in the LC scheme
might dissolve some of the silica gel
from the column bed, giving rise to
spurious high values in the gravimetric
analysis. The data were obtained for
blank columns with no sample added
indicating that 0.9 ±0.2 mg of silica gel
was found in the eluant of Fraction 7; no
weighable material was found in
Fractions 1 through 6. It appeared to be
necessary to correct the apparent
weights of Fraction 7 arithmetically, by
subtracting the appropriate blank value,
rather than by actually separating the
sample from the silica gel.
Ruggedness Testing of Level
1 LC Procedure
It was desirable to determine the
change in results of the LC separation
that might occur if a slight deviation in
procedure was made. This margin of
error within which an experiment can
be varied without changing the results
is commonly referred to as the
"ruggedness" of the experiment. If the
procedure can be done slightly
differently each time and the same
results obtained, chances are increased
that intralaboratory or even interlabora-
tory results will be consistent and
comparable.
The Level 1 LC procedure was studied
for ruggedness in three aspects:
continuity of elution, exact composition
of eluants, and activity of silica gel.
In all cases a solution of model
compounds with a convenient elution
pattern was used to test the procedures.
The fractions were analyzed by GC
(using the Level 1 TCO program) and
compared to controls. No significant
change in the elutionpattern, compared
to the control LC procedure, was
observed under any of the following
conditions: 0.5-hr interruption in LC
elution; use of month-old vs fresh
elution solvents; activation of silica gel
at times (2-24 hrs) and temperatures
(90-200°C) other than those specified in
the procedure; and LC column
temperature varied from 14.5 to 22°C.
Report Formats for Level 1
Organic Analysis Results
A particularly important aspect of the
evaluation and evolution of the Level 1
organic analysis procedures was the
development of systematic reporting
formats that summarize and integrate
the data usefully. The report formats
developed in this study are incorporated
in the October 1978 manual.2 They are
described briefly in this summary;
complete examples are given in the full
report.
In interpreting and integrating the
Level 1 organic analysis data and
reporting the results, a list of organic
compound categories, based on the
Multimedia Environmental Goals (MEG)4
categorization scheme (slightly
modified), was used to organize the
Level 1 data. It should be emphasized
that it is the list of MEG categories, and
not the list of specific MEG compounds,
that is referred to in the context of Level
1 organic analysis. With the addition of
a very few non-MEG compound classes,
this list of categories represents almost
all of the organic chemistry likely to be
encountered in most sources. In
developing this report format for Level 1
organic analysis results, it was
assumed that users of Level 1 EA data
would be interested in comparing
estimated mass loadings in various
streams with decision criteria
expressed as concentrations. A
* US. OOVSRNMENTPRINTING OFFICE: 1M2 -559-017/0809
-------
requirement of the approach is that all
data reduction steps described in this
document be performed by personnel
with access to the original Level 1
organic analysis data and/or direct
communication with the original
analyst(s). The reason for this is that the
organic analysis data outputs (e.g.,
spectra) contain much information (e.g.,
presence or absence of particular peaks)
that is not easily reduced to tables or
other forms intelligible to nonchemists,
but is very valuable in assessing the
Level 1 sampling and analysis results.
Results of the LC fractionation
procedure include quantitative
estimates of TCO and GRAV range
materials in each of seven fractions.
GRAV analyses involve weighing to the
nearest 0.1 mg. TCO values and all
Level 1 concentration data are reported
to 1 or 2 significant figures.
The total sample extract, or neat
organic liquid, and the seven LC
fractions are analyzed by IR spectroscopy.
Spectra are interpreted in terms of the
functional group types present in the
major components of the sample or LC
fraction. Interpretation of the spectrum
is guided by consideration of the LC
fractionation scheme and the LRMS
results. Components amounting to <5%
of the total sample will not contribute
significantly to the IR spectrum and
cannot be detected by this technique.
The report format forthe results of the
IR analysis includes specification of
frequencies, intensities, and
assignments for all major peaks in the
spectrum.
LRMS are obtained on each LC
fraction which has sufficient quantity to
potentially exceed decision criteria
levels of organic pollutants. Samples
with significant quantities of TCO range
material are analyzed by insertion in the
batch inlet; all samples require analysis
by direct insertion probe.
At the end of the Level 1 organic
analysis procedure, there will be an LC
report, eight IR reports, and up to seven
LRMS reports for each organic extract
or neat organic sample. This is an
unwieldy body of data from which to
make a decision. These data are
reduced to a workable form by preparing
a single table that summarizes the
organic analysis results for each
extract.
References
(1) Hamersma, J.W., S.L. Reynolds, and
R.F. Maddalone, "IERL-RTP
Procedures Manual: Level 1
Environmental Assessment," EPA-
600/2-76-160a, NTIS No. PB 257-
850, June 1976.
(2) Lentzen, D.E., D.E. Wagoner, E.D.
Estes, and W.F. Gutknecht,
"IERL-RTP Procedures Manual:
Level 1 Environmental Assessment
(Second Edition)," EPA-600/7-78-
201, NTIS No. PB 293-795, October
1978.
(3) Thrun, K.E., J.C. Harris, and K.
Beltis, "Gas Sample Storage," EPA-
600/7-79-095, NTIS No. PB 298-
350, April 1979.
(4) Cleland, J.G. and G.L Kingsbury,
"Multimedia Environmental Goals
for Environmental Assessment,
Volume I," EPA-600/7-77-136a,
NTIS No. PB 276-919, Noverrber
1977.
Judith C. Harris, ZoeA. Grosser, Philip L Levins, DebraJ. Sorlin, and Clifford H.
Summers are with Arthur D. Little, Inc., Cambridge, MA 02140.
Larry D. Johnson is the EPA Project Officer (see below/.
The complete report, entitled "Laboratory Evaluation of Level 1 Organic Analysis
Procedures." (Order No. PB 82-239 294; Cost: $ 18.00, subject to change) 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:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
0000329
------- |