v/EPA
United States Industrial Environmental Research EPA-600/7-79-191
Environmental Protection Laboratory August 1979
Agency Research Triangle Park NC 27711
Measurement of
Polycyclic Organic
for Environmental
Assessment
Interagency
Energy/Environment
R&D Program Report
ENViHT -ME
XAS
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports m this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide'range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-191
August 1979
Measurement of Polycyclic Organic Matter
for Environmental Assessment
by
P.L Levins, C.E. Rechsteiner Jr., and J.L Stauffer
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts 02140
Contract No. 68-02-2150
Task No. 10202
Program Element No. INE624
EPA Project Officer: Larry D. Johnson
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ACKNOWLEDGEMENT
The authors wish to acknowledge the contributions
of groups at Battelle Columbus Laboratories and
Monsanto Research Corporation for their comments
concerning methods of data acquisition on various
instruments and analytical procedures for POM
analysis.
ii
-------
TABLE OF CONTENTS
PAGE
ACKNOWLEDGEMENT ii
LIST OF FIGURES v
LIST OF TABLES vi
ABSTRACT vii
I. INTRODUCTION 1
A. Historical Perspective 1
B. Analytical Methodology 2
1. Fluorescence Methods 3
2. Ultraviolet Absorption Spectroscopy 5
3. Gas Chromatographic Methods 6
4. Miscellaneous Methods 8
II. SCREENING METHODS 9
A. Total POM by Fluorescence 9
B. Sensitized Fluorescence Spot Test 10
III. SPECIFIC POM DETERMINATION BY GC/MS 12
A. Mass Spectrometer Conditions 12
B. Packed Column GC Procedures 17
1. General Procedures 17
2. Liquid Crystal GC Procedures 17
C. Capillary GC Column Procedures 19
IV. VERIFICATION STUDIES 24
A. POM Analysis with Packed Column GC/MS 24
B. Selected POM Analysis with Liquid Crystal
GC Procedure 32
C. Capillary Column Procedures for POMs 36
continued ....
iii
-------
TABLE OF CONTENTS (continued)
PAGE
V. METHODS 48
A. Total POM Measurement by Solution Fluorescence . . 48
B. Sensitized Fluorescence for Total POM 49
C. POM Measurement by Packed Column GC/MS ..... 51
D. Measurement of Selected POMs by Liquid Crystal
Column GC/MS 53
E. POM Measurement by Capillary Column GC/MS .... 55
REFERENCES 58
APPENDIX A - EPA Level 1 - Liquid Chromatographic (LC)
Separation ..... 61
iv
-------
LIST OF FIGURES
Figure PAGE
1 Selected Ion Chromatograms for m/z 228 (a) and
252 (b) from Packed Column GC/MS Run Using
Dexsil 400 GC Phase 18
2 Comparison of m/z 252 Selected Ion Chromatograms
for Packed Column, Dexsil 400 (a) and Liquid
Crystal Column, SP-301 (b) 20
3 Sum Ion Chromatogram for PAH Standard Using OV-17
Capillary GC Column 22
4 Selected Ion Chromatograms for m/z 178 (a) and
m/z 228 (b) of Figure 3 Expanded 23
5 Reconstructed Ion Chromatogram for Clean Water
Sample Spiked with Priority Pollutants 28
6 Reconstructed Ion Chromatogram for Raw Wastewater
Spiked with Priority Pollutants 29
7 Reconstructed Ion Chromatogram for Raw Wastewater
Spiked with Priority Pollutants (Replicate) ... 30
8 Benzo(a)pyrene Calibration Curve Used with Liquid
Crystal GC/MS Analysis 37
9 Reconstructed Ion Chromatogram for Sample of Jet
Engine Exhaust 41
10 Reconstructed Ion Chromatogram for Sample of
Fumes from a Low Burn Pitch 45
11 Sum Ion Chromatogram for Sample of Fumes from a
Low Burn Pitch for Analytical Ions Listed in
Table 13 46
v
-------
LIST OF TABLES
Table PAGE
1 Key POMs of Interest in Environmental Samples .... 4
2 Fluorescence/Concentration Information for POM
Spot Test Procedure using Naphthalene as
Sensitizer 11
3 GC/MS Conditions for Publicly-Owned Treatment
Works Sample 25
4 POMs Specified in the Priority Pollutant Protocol . . 27
5 Average Recoveries of POMs from Aqueous Samples ... 31
6 GC/MS Operating Conditions for Coke Oven Quench
Sample Analysis 33
7 Selected POMs from Atmospheric Emissions of Coke
Oven Quench Samples Using Clean or Recycled
Water 34
8 GC/MS Conditions for BaP Analysis with Liquid
Crystal GC Phase 35
9 Comparison of Benzo(a)pyrene Levels from Liquid
Crystal to Benzopyrene Levels from Packed Column
GC/MS 38
10 GC/MS Conditions for Jet Engine Exhaust Emissions . . 40
11 POM Concentrations in Jet Engine Exhaust 42
12 GC/MS Conditions for Low Burn Pitch Fumes 43
13 Concentrations of Selected POMs in Low Burn Pitch
Fume Sample 44
A-l Liquid Chromatography Elution Sequence 64
vi
-------
SUMMARY
This report discusses methods for the measurement of Poly-
cyclic Organic Matter (POM) for environmental assessment.
Two fluorescence methods are described for the estimation
of total POM levels in samples. Either of these methods
may be used to screen samples for further specific analysis.
Three gas chromatography-mass spectrometry (GC/MS) methods
are described for the measurement of specific POM compounds.
The use of liquid crystal chromatographic phases is recom-
mended for the measurement of a few POMs, i.e., specifically
for benzo(a)pyrene. GC/MS methods for a wide range of POMs
are discussed for both capillary and packed GC columns. The
three methods for specific POM identifications are verified
with collected environmental samples of different kinds.
vii
-------
I. INTRODUCTION
There is a great deal of interest in the reliable analysis of
polycyclic organic matter (POM). POM has been defined by the National
Academy of Sciences as the polycyclic aromatic hydrocarbons (PAH) plus
their heterocyclic analogs. POM species are often found in combustion
2
effluents and are widely distributed at varying levels throughout the
O
environment. Since specific POM compounds have been identified as
causal agents in the development of cancers, analytical methods for
specific POMs are needed to accurately assess potential health hazards.
In this document, five areas of effort are covered: 1) pertinent back-
ground literature is reviewed to illuminate current activities in the
field of POM analysis, 2) several general methods for detection of POM
are discussed for use as screening methods, 3) the development of
analytical procedures for analyzing specific POM compounds are presented,
A) the verification data for the procedures developed for specific
POMs using real samples are presented, and 5) a detailed description
of each method is furnished.
A. Historical Perspective
The earliest association of certain cancers with combustion ef-
fluents dates to 1776 with the report of high incidence of scrotal
cancers in chimney sweeps. By the early 1900*s, various workers
had succeeded in producing skin tumors in experimental animals by
fi 7 ft
direct application of tar. Work by Dreifus, Kennaway, and Cook,
and others established a link between tumor formation and specific
POMs (i.e., 1,2,5,6-dibenzanthracene, 1,2-benzopyrene and substituted
benzanthracenes). Since these early studies, a great deal of effort
has been spent in determining the carcinogenic potential of specific
compounds. By 1953, 450 compounds were determined to be carcinogenic
of which over 200 were POMs, specifically polycyclic aromatic hydro-
9
carbons, their derivatives and analogs. The carcinogenic activity
of POM species changes markedly with relatively minor structural
changes in ways that are difficult to predict. Whereas the addition
of methyl groups frequently enhances carcinogenicity, addition of a
-------
methyl group at the 5 position of the potent dibenzo (ah or ai) pyrenes
reduces their activity and introduction of two methyl groups eliminates
9
the activity completely. The large variation in carcinogenic activity
of POMs with slight variations in structure makes the accurate analysis
of specific POMs an important factor in environmental monitoring and
environmental impact assessment.
Information concerning POMs, their occurrence, health effects, and
chemistry is available in many sources. A thorough review of the re- *
lationships between POMs and carcinogenesis is found in a document pre-
pared by the Committee on Biologic Effects of Atmospheric Pollutants of
the National Academy of Sciences (USA) entitled, "Particulate Polycyclic
Organic Matter." This document includes discussion of such topics as
the sources of POMs, the physics and chemistry of POMs distributed
throughout the environment, testing procedures for carcinogenic, muta-
genic and/or teratogenic properties, and clinical studies on exposure
to POMs, as well as a set of appendices dealing with the collection
and analysis of samples for POMs. This document provides a general
overview of the POM problem with respect to health effects.
A series of symposia (International Symposia on the Analysis,
Chemistry, and Biology of Polynuclear Aromatic Hydrocarbons) published
under the title "Carcinogenesis," ' offers one method of maintaining
current awareness of the status of chemical and biological investigations
concerning POMs. The intent of this symposium series, as stated by the
editors, is to "provide a valuable forum for discussion and examination
of the most recent research findings in the area of analysis, chemistry,
and biology of PAHs." ' The types of papers presented at these
symposia cover sampling and analysis techniques for POMs, studies of
POM formation mechanisms, studies of POM metabolic pathways, and bio-
testing procedures applicable to the study of POMs.
B. Analytical Methodology
The interest in POMs has given rise to a host of qualitative and
quantitative procedures for POM analysis. Of the various procedures,
some are intended to give a single quantitative value for a specific
-------
compound, such as for benzo(a)pyrene, while others attempt to give a
composite value indicative of the total quantity of POM present in a
sample. The main difference between most specific and general analytical
techniques is a separation of the species of interest prior to detection.
In the following discussion, the reported methods will be grouped
by the general analytical techniques used. Application of each technique
to either specific or general analyses will be discussed within that
section. For reference, Table 1 contains a list of specific POMs
presently of interest in most environmental analyses.
Prior to the detection of POMs, some suitable procedure should be
used to isolate POM from interferences and to concentrate the sample.
Historically, the techniques of thin layer chromatography and paper
chromatography were used to isolate specific POMs for subsequent
identification. Current practice utilizes column chromatography to
separate POMs from non-POMs (see Appendix A) followed by either high-
pressure liquid chromatography or gas chromatography for resolving spe-
cific POM isomers from one another. The use of these procedures, in
12
conjunction with POM analysis, is reviewed extensively elsewhere.
The emphasis in this report is on the use of gas chromatography (GC)
which when combined with mass spectrometry is a powerful tool for the
detection and quantification of POMs ranging up to the dibenzopyrenes in
molecular weight.
1. Fluorescence Methods
Fluorescence methods were among the very first methods used for the
detection of POMs. The work in the early 1930's by J. W. Cook and co-
8 13
workers ' made extensive use of fluorescence spectroscopy to identify
benzo(a)pyrene as a carcinogenic constituent of coal tar. Various
workers ' ' have employed fluorescence spectroscopy to follow the
metabolic reactions of specific POM compounds during hydroxylation. The
analysis of specific POMs has also been undertaken with fluorescence
17 18
methods. Benz(a)anthracene, benzo(a)pyrene, and indeno(l,2,3-cd)-
19
pyrene are just a few examples of specific POMs which have been analyzed
with fluorescence spectroscopy. Fluorescence techniques have been used
20
by Sawicki and others to examine a wide range of POM containing samples.
-------
TABLE 1
Key POMs of Interest in Environmental Samples
*
NAS Rating
Fluorene -
Anthracene -
Phenanthrene -
Fluoranthene -
Pyrene
Benzo(c)phenanthrene + + +
Benz(a)anthracene +
Chrysene i
Triphenylene (not rated)
Benzo(b,k or j) fluoranthene + +
Benzo(e)pyrene
Benzo(a)pyrene + + +
Perylene
7,12-dimethyl benz(a)anthracene + + + +
3-methyl cholanthrene + + + +
Indeno(l,2,3-cd)pyrene +
Benzo(g,h,i) perylene
Dibenz(a,h) anthracene + + +
Dibenz(a,i or a,j) acridine + +
Dibenzo(c,g) carbazole + + +
Dibenzo(a,i or a,h) pyrene + + +
Coronene -
*
Reference 1 .
-------
Fluorescence methods are capable of measuring subnanogram quantities
of individual POM species, but tend to be fairly non-selective. The
normal spectra obtained from POM and related species tend to be intense
and lack resolution in fluid media. To determine levels of specific
compounds, thorough separation of the sample constituents is needed to
minimize spectral interferences. Efforts to overcome this difficulty
have been directed towards low temperature techniques. The fluorescence
of various POMs at temperatures of 20 K or less no longer consists of a
series of broad, smooth peaks found in room temperature fluid media, but
contains many discrete peaks of varying intensity. The unique nature
of the spectra obtained with this method is demonstrated by the fact
that the six isomeric mono-methylchrysenes can be distinguished from
21
one another by their low temperature fluorescence spectra. However,
identification of an unknown species from its spectrum is not an easy
task. Compilation of reference spectra and rapid methods of searching
reference spectra are not as well developed or as widespread as those
often used in infrared or mass spectroscopy. Furthermore, it may be
difficult to recognize spectra, even of pure compounds, because of
phenomena such as saturation at high concentrations. Also, certain
classes of POMs, such as nitrogen containing heterocyclics, exhibit low
efficiency towards fluorescence and may not be detected in small quantities.
22
Work at Oak Ridge National Laboratories has applied synchronous
fluorescence and phosphorescence to PAH analysis. With the synchronous
technique, the normally broad fluorescence spectra of the PAH species
is simplified to a series of distinct and better resolved peaks. The
inprovement in the distinguishing features of the different PAH spectra
enhances the specificity of these luminescence techniques, and the
potential application of these methods to provide more specific infor-
mation than total POM measurements needs to be further explored.
2. Ultraviolet Absorption Spectroscopy
Ultraviolet absorption spectroscopy (UV) is another analytical
detection method which has been widely used. For many years, the pre-
ferred approach for POM analysis was to use a combination of liquid
chromatography (LC) and thin layer chromatography (TLC) to isolate
-------
specific POMs followed by UV techniques for species identification and
0 1 O/ *? *N
quantification. Benz(a)anthracene, benzo(a)pyrene, and indeno
7 f\
(1,2,3-cd) pyrene are typical examples of POMs which may be detected.
Compilations of UV data for numerous POMs are found in a number of
27 28
sources. ' As with fluorescence measurements, the individual spectra
for various POMs are unique, although portions of spectra for different
compounds may be the same; for example, both chrysene and 4,5-methylene
chrysene have similar absorption bands at 361 nm and both benzo(a)pyrene
and benzo(ghi)perylene have similar bands at 382 nm. The possibility
of spectral overlap requires complete separation of sample components
to insure accurate measurement of component levels, as with fluorescence
techniques. Also, the overall sensitivity of measurements by UV
methods is somewhat lower than for fluorescence methods (one-tenth to
one-thousandth that of fluorescence). Hence, the use of UV measurements
for POM analysis has declined, being replaced with the more sensitive
fluorescence methods and the highly sensitive and specific GC/MS methods.
3. Gas Chromatographic Methods
Gas chromatographic methods have shown the most rapid development
for POM analyses in recent years. The first useful application of GC
techniques to POM analysis dates to 1965, with methods being found in
29 30
the Journal of Chromatography and Analytical Chemistry. The
instrumentation for a GC separation is relatively inexpensive and samples
may be analyzed conveniently with high speed and good reproducibility.
Detection of the GC effluent may be by any of a wide range of analytical
procedures. Since UV methods have been widely used for POM detection
and the UV spectra are specific for different POMs, methods which couple
GC for separation and UV for detection have been developed. Other detec-
tors, such as flame ionization detection (FID), electron capture detection
(BCD), mass spectrometer (MS) detection or infrared (IR) detection have
been coupled with GC separation for sensitive and specific analyses.
Numerous GC column packings have been developed to enhance the
separation of specific POMs. Methyl phenyl silicones, such as SE-52,
OV-17, or SP2250,and carborane methyl- and methyl phenyl silicones, such
as Dexsil 300 or Dexsil 400, have been widely applied to POM separations.
-------
These GC phases will resolve groups of POMs from one another, 3-ring
POMs from 4-ring, 4-ring POMs from 5-ring, etc., but the resolution of
specific isomers is not possible in every case.
Two developments greatly improved the specificity of GC analyses
of POMs. Capillary columns have been known for their inherently high
31
resolution since the work of Golay in 1958. However, extensive use
of capillary columns was delayed until the development of efficient
injection systems and reproducible coating of the stationary phase on
the capillary tube. The high resolution obtainable by capillary
GC techniques allows separation of selected isomers unresolved with
32
normal packed columns (see e.g., Lee, et al. for an extensive list of
POM relative retention indices). However, capillary columns will only
tolerate small sample volumes. Thus, the detector for capillary GC
methods must have high sensitivity such as flame ionization detectors
(FID), electron capture detector (BCD), or a mass spectrometer (MS).
Of these, the last detector system, a mass spectrometer, allows confirmation
of species identification and, through the use of computerized data handling
procedures, greatly enhances the specificity of the analytical measurement.
The other advance in GC technique which has affected POM analysis
is the liquid crystal GC column phase [for example, N,N'-bis(p-phenyl-
33
benzylidene) a,a'-bi-p-toluidine (BPhBT)]. In 1975, Janini and co-
workers first applied nematic liquid crystals to the separation of
polycyclic aromatic hydrocarbons. These columns resolve geometric
isomers of different POMs, but the observed elution order of the isomers
is frequently different than that observed for the general purpose sili-
cone columns. For example, the normal elution order — benzo(e)pyrene,
benzo(a)pyrene, pyrene — observed on OV-17 and Dexsil-300 and -400
columns is different from that of a liquid crystal column such as
SP-301 (BPhBT) — benzo(e)pyrene, perylene, and benzo(a)pyrene. To
obtain the best separation of POMs, isothermal column operation is
required which limits the use of these columns with complex samples.
When coupled to a selective detector, e.g., a mass spectrometer, which
may be adjusted to minimize the interferences due to other non-POM
species in a complex sample, liquid crystal columns are highly useful
-------
for determining quantities of narrow groups of specific POMs.
The power of the combined GC/MS system for POM analyses makes such
systems the method of choice for specific POM analyses. Detailed
procedures are presented later in this report.
4. Miscellaneous Methods
The intense interest in POM analysis has brought forth a large
34
variety of special techniques for specific analyses. Infrared and
35
Raman spectroscopic techniques, nuclear magnetic resonance methods,
T £ r\ ~j
polarography, and potentiometric titration are examples of the types
of alternative techniques which have been applied to POM analysis. For
the most part, these methods tend to be prone to numerous interferences
in real samples, and offer no significant advantage over UV, fluorescence
or GC methods discussed previously.
In the following chapters, specific analytical techniques are
presented for the analysis of POMs. In Chapter 2, two survey methods
for POM analysis are presented. These methods are designed to give
rapid order of magnitude data on total POM presence in environmental
samples. In Chapter 3, three general GC/MS methods for specific POM
analysis are presented. The three methods involve current technology
and are chosen such that at least one of these methods will be appli-
cable to GC/MS systems which have been or are being produced. In
Chapter 4, experimental verification of the three methods described
in Chapter 3 is given to provide practical information on the application
of these methods to real samples. In Chapter 5, the methods presented
in Chapters 2 and 3 are described in further detail for users of these
methods.
-------
II. SCREENING METHODS
An initial step in the analysis of samples for POMs may be to
estimate the level of POM present. The method used should be rapid,
low-cost and simple to perform. Methods which meet these requirements
can be used to screen samples to point out those samples which need
further specific analysis. If the level of POM present is very low,
there may not be sufficient justification to go to the expense of
performing specific POM analyses. Two methods are recommended here
for use as low-cost screening procedures; both use luminescence spec-
troscopic techniques.
A. Total POM by Fluorescence
The procedure described in this section is based on the investiga-
38 39
tions of Sawicki, as adapted for use by Battelle Columbus Laboratories.
This procedure makes use of the similarity of the fluorescence spectra
for many POMs. In normal fluorescence analyses for POM, specific
identifications can only be made for pure POM species. This is due to
POMs having overlapping excitation and emission wavelengths. If instead
of isolating specific POMs, the sample is fractionated so that one
fraction contains only the POMs, then use of a broad excitation source
and a broad emission detector will give data representative of the sum
of the POMs present.
The first step in determining total POM is to fractionate the
sample to isolate the POMs. Use of the Level 1 liquid chromatography
procedure (Appendix A) results in the POMs eluting in fractions 2, 3,
and 4. These fractions are combined, reduced in volume from 30 mL to a
convenient small volume, such as 2 mL, and measured in a spectrofluori-
meter set for 350 1 5 nm wavelength excitation and 410 1 5 nm wavelength
emission. After each fluorescence measurement, the sample is diluted
and remeasured until the sample is in the linear calibration region.
Calibration is conveniently referenced against anthracene to obtain
concentration values. Use of this POM measurement procedure yields
answers which agree within a factor of three with those found by GC/MS
39
for the same sample. The complete procedure is found in section A
of Chapter 5.
-------
B. Sensitized Fluorescence Spot Test
The second method recommended for screening samples for POMs was
40
developed by Arthur D. Little, Inc. and involves the use of sensitized
fluorescence. Sensitized fluorescence occurs when two or more fluores-
cent compounds are present in a solid or crystalline mixture with one
being at a much higher concentration than the others and when these
compounds are able to have vibrational coupling between their excited
singlet energy states (i.e., the compounds have at least one vibrational
level frequency in common in the excited state). In such cases, if the
mixture is excited and the compounds absorb energy, the fluorescence
emission will occur preferentially from the compound which has the
lowest vibrational energy level of the excited state. For example,
sensitized fluorescence of naphthacene occurs in mixtures of naphthacene
:ene
41
and anthracene with the naphthacene level about 10 that for the
anthracene.
The spot test procedure involves drawing three spots on a piece
of filter paper, applying sample and/or naphthalene to each spot such
that one spot contains naphthalene, a second spot contains the sample
of interest, and the third spot contains both naphthalene and the
sample, exciting the sample with 254 nm radiation and visually observing
the fluorescence. The observed fluorescence of the spots gives an
order of magnitude estimate of the total POM level found in a sample.
Table 2 shows possible observations and the resulting POM concentration
level. For complete details, see reference 40 and section B of Chapter
V.
10
-------
TABLE 2
Fluorescence/Concentration Information for POM Spot Test
Procedure* using Naphthalene as Sensitizer
POM
Observation Concentration
a. Non-fluorescent when treated with sensitizer < 1 pg/yL
b. Weakly fluorescent when treated with
sensitizer 1-10 pg/yL
c. Strongly fluorescent when treated with
sensitizer but not fluorescent alone > 100 pg/yL
d. Fluorescent without sensitizer ^ 10,000 pg/yL
*
from Reference 40.
11
-------
III. SPECIFIC POM DETERMINATION BY GC/MS
The choice of GC/MS as the recommended procedure for specific POM
analysis is influenced by several factors. The sensitivity and selectiv-
ity of the GC/MS combination is well known. After separation of a
complex sample into a large number of chromatographic peaks, the identity
of those peaks suspected from their retention time to be POMs can be
confirmed from the mass spectra. The use of GC/MS for specific POM
determinations is widely accepted and implemented. Due to the instru-
mental differences between the various GC/MS combinations available and
the manner in which different instruments acquire and process sample
data, no single combination of instrumental parameters will be directly
applicable to all GC/MS systems.
It is not the intent of this document to specify a single instru-
ment for POM analysis by GC/MS to the exclusion of the wide range of
other instruments capable of obtaining high quality analyses. For
these reasons, the methods outlined in this chapter are somewhat general
and intended to be as widely applicable as possible. In the remainder
of this chapter, analytical details and procedures for the analysis
of POMs are discussed. The general mass spectrometer operating parameters
are discussed first,with the discussion of gas chromatography conditions
following. The latter discussion is divided into three sections to
cover the three major choices for GC separation of POM: packed columns,
capillary columns, and liquid crystal phases for GC column packing. The
verification data resulting from application of the methods in this
section to environmental sample are given in Chapter 4. Detailed
descriptions of methods for GC/MS determination of POMs using packed GC
columns, capillary GC column and liquid crystal phases in GC columns
are given in Chapter 5.
A. Mass Spectrometer Conditions
The exact mass spectrometer conditions chosen for a single analysis
depend on many things: the specific manner in which the mass spectrometer
acquires and processes data, and instrumental sensitivity required, as
well as the reason for the analysis. Although the specific conditions
12
-------
are highly variable, a number of common conditions are present in any
analysis. The discussion in this section focuses on the instrumental
options available and their application. The instrumental variables
discussed include:
• general instrument set-up
• mode of ionization
• data acquisition/processing
• use of internal standards
• quantitative calibration
The overall instrument operating condition obviously has a great
effect on the reproducibility of the data obtained. What is sometimes
less obvious, however, is that fully automated GC-MS-DS systems may
make it difficult to recognize the overall instrument operating condition.
The condition of the ion source affects sample data collected. A source
which is dirty shows altered focusing conditions, distorted ion peak
profiles, and lower transmission efficiency compared to a clean source.
For quadrupole instruments, which are quite sensitive to the transmission
of high mass ions, this can lead to loss of sensitivity for high molecular
weight POMs. Leaks in the vacuum system will affect instrument performance
through ion-molecule reactions and/or depletion of the quantity of the
ionizing species (electrons for El or charged reagent ions for CI). For
some automated GC-MS-DS systems, it is difficult to monitor the ion peak
profiles to assess the condition of the ion source. For such systems,
routine periodic monitoring of the absolute intensities of sets of ions
spanning the mass range of interest from a calibration mixture, for
example, provides a method of checking instrument performance in lieu
of monitoring ion peak profiles directly. A continuous drop in ion in-
tensities (most apparent for higher mass ions) for a constant multiplier
voltage, etc., is indicative of a problem in the ion source optics such
as a dirty source.
The daily tune of the mass spectrometer is essential in obtaining
reproducible data. This is especially critical for quadrupole spectro-
meters where small changes in lens potentials can drastically change the
13
-------
ion transmission at higher masses. One may tune, for example, on a
bleed peak of the GC column being used (m/z 207, 253, 281, or 315 for
many silicone columns) or to a specific compound spectrum such as deca-
41
fluorotriphenylphosphine specified in the Priority Pollutant Protocol.
For GC-MS analysis, tuning the instrument to some component of the GC
effluent insures that the spectrometer is correctly focussed for the
GC-MS analysis, since for some systems proper focus is dependent upon
the location and orientation at which compounds used for tuning and
analysis are introduced.
The mode of ionization chosen for POM analysis, either electron
impact ionization (El) or chemical ionization (CI), depends upon the
instrument sensitivity as well as sample-related characteristics. Of
the two ionization modes, CI tends to be more sensitive for POMs, although
the differences are highly instrument dependent. With either mode, the
mass spectra for the POMs are simple. For El, the mass spectrum consists
+ 2+
of the molecular ion M , the doubly-charged molecular ion M , and the
loss ions (M-2) and (M-l) . For CI using methane as the reagent gas,
the mass spectrum consists of the protonated molecular ion (M+l) and
the adduct ions at (M+29) and (M+43) due to the addition of C^H- and
C-H . The sensitivity gain with CI is in part due to background noise
reduction. Fragment peaks and bleed background tend to be reduced with
CI, yielding clean spectra, but identification of other compounds in
the sample is somewhat more difficult. For the POMs listed previously
in Table 1, either El or CI may be used for ionization, contingent upon
the sensitivity requirements for the analysis. It is recommended that
El ionization be used whenever possible to allow identification of
potential interferences by comparison to library reference spectra and,
thus, minimize their effect on the analysis.
It is recommended that an internal standard be used in POM analyses
to improve the retention time accuracy for qualitative POM identification
and provide a quantitative standard. The specific internal standard or
standards used for an analysis depends upon the method of data acquisition.
d.-anthracene, 9-methyl anthracene, 9-phenyl anthracene, and 9,10-diphenyl
anthracene are popular choices as internal standards, since these four
14
-------
compounds are not normally seen in environmental samples. The choice
of using one or more of these compounds for internal standard is up
to the analyst. Several cautions should be noted. In acidic media,
d -anthracene may slowly react to replace deuterium with hydrogen. If
the internal standard measurement is based on the specific ion of mass
188, artifically high concentration values will be reported unless the
exchange is taken into account. Although 9-methyl anthracene is not
found in environmental samples, background interferences due to fragment
ions at m/z 192 can be substantial with El ionization, changing the
accuracy of the internal standard measurement. The 9,10-diphenyl anthra-
cene also has some problems when used with electron impact ionization.
In addition to the molecular ion (m/z 330), 9,10-diphenyl anthracene has
a strong fragment ion (about 40% to 50% of the intensity of the parent
ion) at m/z 252. Since this compound elutes in the region of the m/z 252
POMs (e.g., benzo(a)pyrene), it represents a potential interference
problem. For many POM analyses, 9-phenyl anthracene represents a good
choice for use as a single internal standard unless multiple internal
standards are required due to unique requirements of the analysis or
data acquisition system.
The specific method of data acquisition used for POM analysis de-
pends upon the instrumental capabilities and sensitivity. Data may be
acquired with either selected ion monitoring (SIM), selected mass range
scanning, or full mass range scanning. In SIM, only a few ions are
monitored during a sample run allowing long integration times for each
ion. In selected mass range scanning, small portions of the mass range
are sampled allowing longer integration times for each mass than in full
mass range scanning where each of the masses is sampled. An example of
the use of selected mass range scanning is in the measurement of poly- r'
chlorinated biphenyl where the mass ranges from 254 to 260 and from 288
43
to 294 are sampled for tri- and tetrachlorobiphenyl. Since SIM and
selected mass range scanning are more sensitive than full mass range
scanning (due to enhanced signal-to-noise ratio resulting from increased
integration time per mass), one of these methods often will be used to
enhance the instrumental sensitivity. The disadvantage associated with
15
-------
using either selected ion monitoring or limited mass range scanning is
the limited amount of information acquired. With SIM, there is no way
to distinguish between a POM of m/z 252 and, for example, a coeluting
silicon bleed peak ion at m/z 252. With limited mass range scanning,
this problem is diminished but, with either method, other species which
may become of interest would be impossible to assay from the acquired
data at a later date. Therefore, we recommend acquisition of full mass
scan data for POM analysis whenever possible for detection of a wide
range of POMs or selected mass range scanning for detection of a limited
number of isomers.
Quantitative data for a specific POM should be determined from
calibration curves prepared for that POM from data obtained on the
GC-MS system to be used. The calibration curve prepared relates the
instrumental response for that POM to its concentration. All work
should be based upon response relative to that for an internal standard.
For some POMs, reference standards may not be available for preparation
of calibration curves and some method, such as the following, would be
needed to estimate their response vs. concentration curve. Since the
slopes of the response curves for many POMs show a roughly linear
decrease with molecular weight, estimation of the response curves for
POMs, as needed, can be made by interpolation from the response curve
slopes found for POMs of higher and lower molecular weight. The ab-
solute levels of POMs required for use in calibration mixtures will
depend on the instrument sensitivity and the samples to be studied. In
general, there should be at least four calibration mixtures to yield
at least four calibration points for each species, spanning the linear
region of the spectrometer with the lowest concentration standard at
2 to 5 times the instrument detection limit for that species.
Since the mass spectra of the individual POMs are dominated by the
molecular ion, calibration is conveniently obtained on a relative basis
from the area of the POM molecular ion alone vs. that for the internal
standard. Variations of this procedure, where several representative
ions for each species are summed before obtaining the area> are also
16
-------
possible but, for whatever procedure is chosen, it must be consistent
throughout the analyses.
B. Packed Column GC Procedures
1. General Procedures
Most of the work reported on POM analysis by GC or GC-MS has been
carried out on packed columns. Packed columns do not have the high
resolution of capillary columns coated with the same phases; however,
packed columns have longer lifetimes, will tolerate larger sample
sizes, and can give good results for many specific POMs. The methyl
silicone, (e.g., OV-101), methyl phenyl silicone (e.g., OV-17), and the
carborane methyl- and methyl phenyl silicone (e.g., Dexsil 300 and Dexsil
400) GC phases separate at least some of the specific POM isomers, such
as fluoranthene and pyrene, and perylene from benzo(a)pyrene or benzo(e)
pyrene. Some non-isomer POM species may not be chromatographically re-
solved from one another but will be differentiated by molecular weight.
To illustrate this point, Figure 1 shows several ion chromatograms for
some incompletely resolved POMs obtained from a packed column GC/MS run.
Despite the fact that anthracene and phenanthrene or chrysene, triphenylene,
and benz(a)anthracene cannot be individually identified, the sum concen-
tration of these non-resolved species is readily determined. For many
purposes, the sum concentration value is adequate. Comparison of the
sum concentration levels of different samples is one method for assessing
the impact on total POM load of different facility-operating conditions.
2. Liquid Crystal GC Procedures
Liquid crystal stationary phases, such as BPhBT (N,N'-bis(p-phenyl-
benzylidine)a,a'-bi-p-toluidine) allow specific POM identifications to
be made on a packed column, in some cases separating isomers which are
not resolved even on capillary GC columns. The liquid crystal phases
separate geometric isomers of different POMs quite easily. For example,
44
Strand and Andren have compared the separation of benz(a)anthracene
and chrysene on Dexsil 300 and two liquid crystal phases. The liquid
crystal phases resolve these two isomers which are not resolved on the
17
-------
100.0-1
CHRVSENG
TRIPHENYLENE,
BENZ< A >ANTHRACENE
1868
34s 86
1880
34128
1908
34150
1920
35il2
1940
35:34
160.0-1
9,10-DIPHENYL ANTHRACENE
3-METHYL CHOLANTHRENE
2200
40:20
2250
41:15
2300
42:10
2350
43:05
2400
44:00
2450
44:55
2500
45:5C
FIGURE 1. Selected Ion Chromatograms for m/2 228 (a) and 252 (b)
from Packed Column GC-MS Run Using Dexsil 400
GC Phase.
18
-------
silicone columns. Another example is the separation of benzo(a)pyrene
and benzo(e)pyrene as shown in Figure 2. In part A, the separation is
shown for a normal GC phase compared to that found with a liquid crystal
GC phase in part B. Unfortunately, the liquid crystal phases are not
well suited to temperature programmed use and should be used isothermally
slightly above their transition point. When used with temperature pro-
gramming, two factors become important. First, the column bleed is
substantial and, second, at temperatures below the transition point,
chromatographic peak shapes are poor and often not resolved from one
another. Even under isothermal conditions, the liquid crystal structure
continues to undergo phase transition. The effect of this on a benzo(a)-
pyrene analysis is to continually alter the benzo(a)pyrene retention
time. At the recommended isothermal temperature above the transition
point, i.e., 260 C, the retention time continually decreases. Variations
in the retention time for benzo(a)pyrene can be from 30 min. or more to
10 min. or less, over a four-hour period. Judicious changes in the
temperature can be used to maintain the benzo(a)pyrene retention time
within a reasonable range. A quick cooling below the transition point
will lengthen the retention time, and a quick heating above the transi-
tion point will shorten it. Operation at temperatures close to the
transition point extends the limited lifetime of these phases.
An additional point of interest to the analyst is that the elution
order is substantially different compared to that from normal GC columns.
For most methyl silicone GC phases, the relative retention time for
benzo(a)pyrene is smaller than that for perylene, but for liquid crystal
phases like SP-301 (BPhBT), the elution order is reversed. Use of liquid
crystal GC columns for a few specific POMs, e.g., for benzo(a)pyrene
alone, is an attractive procedure but caution must be exercised is using
relative retention times from other GC columns to determine POM species
elution order.
C. Capillary GC Column Procedures
Of the available chromatographic procedures, capillary GC procedures
are best suited to routine determination of a wide range of POM compounds.
19
-------
106.1
BEPfc
BAP
9,10-OIPHENYL ANTHRACENE
3-METHVL CHOLANTHRENE
2280
40120
2290
41il5
2300
42tl0
2350
43105
2450
44s 55
2500
45:5Q
100.0-1
JYLENE
200
4i00
250
5>00
300
6:00
350
7:00
400
8:00
FIGURE 2. Comparison of m/z 252 Selected Ion Chromatograms for
Packed Column, Dexsil 400 (a) and Liquid Crystal
Column, SP-301 (b). *
20
-------
Of the key POMs listed in Table 1, all but a few are at least partially
resolved. Figure 3 shows a mixture of POMs selected from those listed
in Table 1 run on an OV-17 capillary using the conditions described in
Chapter 4. For this column, phenanthrene and anthracene and chrysene
and triphenylene show some separation, which is not the case for packed
columns. Figure 4 shows the separation in higher display resolution for
these particular pairs of isomers for the same sample displayed in
Figure 3. The high resolution of capillary GC columns not only separates
POM isomers from one another but also separates potential background
interferences from the species of interest. Unlike liquid crystal columns
where temperature programming degrades the GC performance, temperature
programming may be adjusted to enhance species resolution while simul-
taneously decreasing analysis time. The prime disadvantage of capillary
GC techniques is the small sample size permissible, requiring high
sensitivity enhancement procedures discussed in Section A of Chapter 3.
21
-------
f
t
s
; ""I
Ilj
if
•i*
Spiv)
_ ITJ ..
k.7
_S ..
L?:
rs?
h „""
s
v f"
60
C
•rl
W
•u
CO
§
n
o
<4-l
IS
i
t_
8
T_
I
-8-
M g
oo 3
O iH
4J O
a u
o o
n o
5 >,
1-1
C n)
O iH
§&
co u
22
-------
100.0-
PHENANTHRENE ANTHRACENE
258
8:22
260
8:43
270
9:03
290 300 310 320
9:43 10103 10:23 10:43
CHRYSENE,
TKIPHENYLENE
BEN2-
ANTHRACENE
560
18146
620
20(46
640
21:26
FIGURE 4. Selected Ion Chromatograms for m/z 178 Ca) and
m/z 228 (b) of Figure 3 Expanded.
23
-------
IV. VERIFICATION STUDIES
The purpose of this section is to illustrate the use of the general
procedures outlined in Chapter 3 on collected environmental samples.
The instrument used for all of these studies was a Finnigan Model 4000
GC/MS with either a Finnigan 6110 Data System or an Incos 2300 Data
System. Due to the high sensitivity of this instrument in the electron
impact mode (limit of detection for Benzo(a)pyrene -2 ng injected with
full mass range scanning for packed GC columns), it was not necessary to
use chemical ionization techniques to increase instrument sensitivity
or to use selected ion monitoring or limited mass range scanning pro-
cedures except as noted.
A. POM Analysis with Packed Column GC/MS
Two types of samples were chosen to demonstrate the capabilities
of packed column GC/MS. The first was from a study of the influent to
a publicly-owned treatment works, the other was from a study of the
emissions from a coke oven quench facility. Specific operating details
are included for the analyses described (see i.e., Tables 3,5 etc.).
The first example of samples analyzed by packed column GC/MS tech-
niques consists of methylene chloride extracts from basic solutions of
sample influent to a publicly-owned treatment work. The samples of
interest were collected, extracted and analyzed under conditions
42
specified in the priority pollutant protocol. Since the conditions
specified in the priority pollutant protocol were being followed, the
mass spectrometer was tuned to give a mass spectrum for decafluorotri-
phenyl phosphene with ion intensities within the limits specified.
Other GC/MS operating parameters used for these analyses are listed in
Table 3. In addition to demonstrating the use of packed column GC/MS
procedures for POM analysis, the samples discussed here provide some
data on the precision and accuracy which are possible in routine analyses
for POM.
The conditions specified in Table 3 were not optimized for POM
analyses alone but were selected for a much wider range of species.
However, the POM related data, such as calibration data, relative
24
-------
TABLE 3
GC-MS Conditions for Publicly Owned Treatment
Works Sample
GC Conditions
Temperature program
He flow rate
Sample Size
Internal standard
MS Conditions
Mass scanning mode
Mass ranges
Integration times
EM Voltage
Electron energy
Filament emission
Scan rate
Data system
50°C Isothermal for 4 min, linear
heating at 88C/min to 265°C.,
265°C isothermal for 30 min.
30 mL/min
2 pL
d.. ^-anthracene
full mass range scanning
40-220, 221-425
3, 5 msec
1800 V
50 V
45 ma
3 sec/spectrum
Finnigan 6110
25
-------
retention time, etc., obtained under these conditions is similar to those
obtained under the conditions used for the analyses of coke oven quench
samples reported later. The priority pollutant protocol specifies
a small set of POMs to be of interest which are listed in Table 4.
The actual samples were collected over a one-week period. During
the analyses of the samples, selected samples were subject to quality
assurance (QA) procedures. For each QA sample, the collected wastewater
was split into three portions. One portion was extracted in the usual
fashion, the remaining two were spiked with reference standards to a
level at least five times that of the minimum detection threshold prior
to extraction. Two additional samples are prepared to complete the QA
sample set. Two high purity water samples are put through the extraction
procedure, one as a blank and the other spiked with the same reference
standard mixture as was used for the spiked raw wastewater samples.
In the raw wastewater samples, no POMs were found with molecular
weights greater than naphthalene (naphthalene is not generally considered
to be a problem POM). The spiked raw wastewater and spiked clean water
samples provide a means of assessing the precision and accuracy of
routine POM measurement. Figures 5 through 7 show typical chromatograms
for the three spiked samples associated with a single collected sample.
For the POMs listed in Table 4, Table 5 lists the recoveries and standard
deviations for seven sets of spiked replicates of the wastewater and
spiked clean water including the sample shown in Figures 6-7. The average
recoveries for the set of POMs tabulated are 81% and 71% for the clean
water and dirty water samples, respectively, with corresponding relative
standard deviations of 20% and 30%.
In the steel industry, water from different sources is commonly used
to quench the incandescent coke. The water used for quenching may be
either clean water, water recycled from other processing procedures, or
a mixture of the two. Assessment of the levels of POMs associated with
the quench operation was of interest,and samples of the atmospheric
emissions under different quenching conditions were collected and analyzed
for POM by GC/MS.
The analysis of the coke oven quench emissions for a general list
26
-------
TABLE 4
POMs Specified in the Priority Pollutant Protocol
NAS Rating
Naphthalene
Acenaphthylene -
Acenaphthene
Fluorene -
Anthracene/Phenanthrene*
Fluoranthene
Pyrene
Benz(a)anthracene/Chrysene* +
Benzo(b or k) fluoranthene* + +
Benzo(a) pyrene + + +
Dibenz(a, h) anthracene + + +
Indeno(l,2,3-cd) pyrene** +
Benzo(g,h,i) perylene
*Isomers reported together.
**Not available in this Quality Assurance set.
27
-------
F-o
LJJ
C)
PD
LLO
rcu
pn
b
J8
•H
CU
CO
p.
CO
n
CU
0)
t-1
o
oo
o
o to
4-1
g §
4-1 O
O PM
M >•>
4J 4-1
CO i-l
C M
O O
O i-l
0) J-<
PJ AH
m
O
H
fe
tfl
H
i
4J
3
>.
4-1
•H
M
O
•rl
CO
3
O
•H
(U
T3
CO
•a
CU
ex
g
•H
4J
0)
u
o
o
d
cu
co
•H
0)
4-1
co
28
-------
A
4J
•H
>
•O
3
•H
ex
CO
u
CU
4J
5
§
4J
CO
s
1
M
O
1-1
t>0
O
4-1
ja .
O CO
4-1
C C
o cd
H 4J
-a ,3
CU i-f
4J O
O PH
CO -H
C *
O O
O -H
0) tJ
W
3
4J
3
o
ex
O
•H
O
•H
M
01
0)
^J
cd
0)
ex
M-l
O
cd
o
o
13
0)
4J
CU
HI
4-1
O .
C •
CU
W H
cu co
w S3
co o
<) PU
29
-------
KJ
LO
CO
hen
Pcsi
M
cu
-i
CO
C-0
FLO
Jlo v o
LD
1.0
too
Un
rs
HOD
FLO
t-o
Ho
tea
L. >
b'J
4-1
•H
13
0)
Q)
4J
tfl
co
I
14 CO
00 O
O -H
4J i-i
§O.
-------
TABLE 5
Average Recoveries of POMs from Aqueous Samples
Clean Water
Species****
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Anthracene/Phenanthrene
Fluoranthene
Pyrene
Chrysene/Benz (a) -
anthracene
Benzof luoranthenes
Benzo (a) Pyrene
Dibenz (a , h) anthracene
Benzo(g,h,i) perylene
n*
7
7
7
7
7
7
7
7
7
7
7
7
P**
78.6
84.6
89.7
90.4
106.8
86.3
86.8
77.1
57.1
67.1
69.1
75.4
% Sp***
31.6
19.8
20.5
15.8
10.1
2.5
5.1
14.5
69.3
10.5
24.5
20.2
Waste Water
n*
14
14
14
14
14
14
14
14
14
14
14
14
p **
85.5
80.4
81.6
73.9
85.6
62.6
64.3
69.6
52.0
62.8
64.0
66.8
% Sp***
36.3
21.4
23.6
21.2
23.2
21.5
25.3
18.8
93.0
23.3
29.0
23.8
Average ior all POMs
examined
80.8
20.4
70.8
30.0
*n = Number of samples
** P = Average % recovery
***%Sp = Relative standard deviation
**** = pQMs listed by increasing elution order
31
-------
of POMs was performed under the conditions shown in Table 6. Although
some POM isomers could not be specifically identified, a wide range of
POMs were readily identified. For the purposes of this study, combined
values for several isomers were reported together as requested. The
POM components were first separated from the other organic constituents
in the sample by the Level 1 LC procedure (Appendix A) with the fractions
containing the POMs of interest (fractions 2, 3, and 4) being combined
and concentrated prior to GC/MS analysis.
The results in Table 7 compare the POM data obtained for samples
with clean water or recycled water used for quenching the coke. The
samples also contain many other organic species which are not shown in
this table. Since full mass range scanning was used during data ac-
quisition, levels for the non-reported species could have been obtained
but were not within the scope of the project. The instrumental limit
of detection for all of the species reported was <1 ng injected.
B. Selected POM Analysis with Liquid Crystal GC Procedure
In Chapter 3, it was recommended that the nematic liquid crystal
GC phases such as SP 301 (N,N - bis(p-phenyl-benzylidine)a,a'-bi-p-
toluidine or BPhBT) be used for the analysis of a narrow range of POMs.
As part of the study of the emissions from a coke oven quench operation,
it was requested that the concentration of benzo(a)pyrene be determined
exclusive of the other POM isomers with identical molecular weights
(i.e., molecular ions of m/z 252) by a second analysis method of high
sensitivity. To accomplish this analysis, the samples collected from
the coke quench facility were analyzed by liquid crystal GC/MS procedures.
The GC/MS conditions shown in Table 8 were used for the specific
analysis of benzo(a)pyrene. 9,10-Diphenyl anthracene was chosen for
use as the internal standard for these measurements. 9-Phenyl anthracene,
which was used previously as an internal standard in these samples,
elutes with the solvent under the GC conditions used, is obscured by the
solvent and early eluting POMs, and could not be used as an internal
standard for the benzo(a)pyrene measurements. To enhance the sensitivity
32
-------
TABLE 6
GC-MS Operating Conditions for Coke Oven
Quench Sample Analysis
GC CONDITIONS
Column
Temperature Program
Helium Flow rate
Sample Size
Internal Std
MS CONDITIONS
Mass Scanning Mode
Mass Ranges
Integration Times
Electron Multiplier
Voltage
Electron Energy
Filament Emission
Scan Rate
Dexsil 400
Isothermal operation at 170°C
for 1 min. Linear operation
to 300°C at 15°C/min,
isothermal operation at
300°C for 30 min.
30 mL/min
2 - 4 yL
9-phenyl anthracene
Full mass range scanning
70-210, 211-270, 271-350
2, 5, 13 msec
1800 V
50 V
45 ma
1 sec/spectrum
33
-------
TABLE 7
Selected POMs from Atmospheric Emissions of Coke Oven Quench
Samples Using Clean or Recycled Water*
Concentration yg/m3
Recycled
Species Clean Water Samples Water Samples
Naphthalene 0.39 170
Fluorene 0.09 23.4
Carbazole - 3.6
Anthracene/Phenanthrene 0.80 33
Fluoranthene 0.24 7.6
Pyrene 0.16 8.0
Methyl-Fluoranthene and
-Pyrene 0.09 0.96
Chrysene, Benz(a)-
anthracene, etc. - 0.54
Benzofluoranthenes,
Benzopyrene - 0.59
* Concentrations on column range from 0.1 ng up for species
detected.
34
-------
Response in Arbitrary Units Relative to 9 , 10-Diphenylanthracene
2 — b c
/
/
/
f
/
/
f.
t
J
•
f
,
«
/
<
/
/
»
/
/
m
f
/
*f
•
/
'
m
m
«
>
/
• f
m
/
/
/
/
A
f
f
.1 1 10 10
BaP Concentrations ng/yL
FIGURE 8. Benzo(a)pyrene Calibration Curve Used with Liquid
Crystal GC/MS Analysis. Horizontal lines indicate
range of responses found for set of five determina-
tions at each concentration, 1 pL injection.
37
-------
TABLE 9
Comparison of Benzo(a)pyrene Levels from Liquid Crystal
to Benzopyrene Levels from Packed Column GC/MS *
3
Concentration yg/m
Sample Description
Clean H20, green coke**
Clean H20, cured coke**
Recycled H20, cured coke**
Recycled H20, cured coke**
Recycled H20, cured coke***
Benzo(a)pyrene
Liquid Crystal
SP-301
38
66
98
72
300
Benzopyrene
Packed Column****
Dexsil-400
120
89
400
* - on column levels of benzo(a)pyrene injected rangedfrom <0.1 to 10 ng
** - atmospheric emissions collected
*** _ sample of recycled water from settling pond prior to dilution
with clean water and use
**** - packed column data for both benzo(e)-arid benzo(a)pyrene
38
-------
minute from a low sulfur fuel with the engine idling. For the sample of
interest, the chromosorb trap was extracted with hexane, 9-phenyl anthra-
cene added as the internal standard and the sample run under the con-
ditions listed in Table 10.
Figure 9 shows the chromatogram for this sample with chromatogram
peaks labeled as to their identity. A large variety of aliphatic com-
pounds present in these samples were not of interest to the study of
jet engine exhaust or the present report. They are not detailed in the
following. This sample contained POMs at relatively low levels, as
shown in Table 11. The primary interest in showing this example is the
resolution of a variety of polar and nonpolar compounds with ease and
the detection of various POMs in a different type of sample matrix than
was presented previously.
The second sample was collected as part of an occupational exposure
study. In this study, a number of roofing pitches and asphalts were
heated under laboratory control to simulate the fumes present during a
commercial roofing operation. The resultant fumes were collected and
concentrated for later use in animal exposure tests. The sample of
interest was collected in cyclohexane/acetone from a low burn pitch
sample heated to 450 C and was chromatographed with an OV-17 capillary
GC column and analyzed using the conditions listed in Table 12.
This sample was rich in POMs as can be seen from Table 13 and
Figures 10 and 11. Figure 10 shows the reconstructed gas chromatogram
for this sample and the sum ion chromatogram (sum of the individual ion
chromatograms) for the specific ions of interest. As can be seen from
Figure lOa, the sample of fume emissions is a complex mixture containing
many POMs and their alkyl substituents. An estimate of the amount of
sample represented by the POMs of interest (found in Table 13) can be
made by comparing the area of the two chromatograms in Figures 10 and 11.
Using these two areas, the concentrations reported in Table 13 account
for about 15%'of the total POM in the sample.
The remaining POMs could be quantified at any time using appropriate
calibration standards and/or the methods described in Chapter 3 but for
the specific program requiring analysis of this sample, only the POMs
39
-------
TABLE 10
GC-MS Conditions for Jet Engine Exhaust Emissions
Gas Chromatographic Conditions
20 meter glass capillary column coated with OV-101
Grob type - splitless inj ection
Multilinear temperature program
1) 55° isothermal program for 1.1 min
2) 55°C - 150°C linear program at 25.5°C/min
3) 150°C - 250°C linear program at 4°C/min
4) 260°C isothermal program for 10 min
1 yL sample injections
Mass Spectrometric Conditions
Finnigan Model 4000 mass spectrometer
Mass scanning mode full mass range scanning
100 - 310 amu
10 ms/amu
1800V
70 eV
30 ma
Mass range
Integration
Electron multiplier
Electron energy
Filament emission
Scan rate
1 sec spectrum
40
-------
auaijiuex
aui|OUUiD02uaq |Ai4iaw
Lor)
to
ZtH8lO.
01H810.
pjepueis |BuJ3jui -»•«
i
0)
l->
>4-l
o
0)
^
&.
CO
n
o
60
O
4J
M
J2
CJ
c
o
H •
4->
•d W
0) 3
•U (0
O .£
=J X
M W
•U
CO 0)
c a
O «H
U 00
-------
TABLE 11
POM Concentrations in Jet Engine Exhaust
Species
Fluorene
Anthracene
Phenanthrene
Methyl fluorene
Fluoranthene
Pyrene
Aceanthralyene
Benzofluorene
Benz ophenanthr ene
Chrysene
Naphthacene
Benzopyrenes
Perylene
Totals
Composition
C16H10
C17H12
C18H12
C20H12*
m/z
166
178
228
252
252
Run #1*** Run #2***
3.85
106,0
48.68
37.62
12.11
517.0
7.52
223,8
180
192
202
202
204
216
2.57
27.85
133.6
46.79
13.96
16.19
ND
76.19
232.3
1195.2
29.10
30.67
86.66
43.19
19.67
2038.1
Sum of signals for both benzo(a)- and benzo(e)pyrene, with benzo(e)
pyrene contributing more to the signal than benzo(a)pyrene.
**ND = below instrumental detection limit of 0.010 yg/mL, or 0.001 yg/mL
for 10X concentrated samples (Total Sample).
*** Different engine operating conditions .
42
-------
TABLE 12
GC-MS Conditions for Low Burn Pitch Fumes
GC Conditions
a) Temp. Program 60°C isothermal for 1 min
to 200°C at 20°C/min
to 285°C at 3°C/min
285°C isothermal for 45 min.
b) Split type injection
c) Helium carrier gas @ 2 mL/min through column
d) Sample Size 1 yL
e) Internal Standard 9-phenylanthracene
MS Conditions
a) mass scanning mode
b) mass range
c) EM voltage
d) Electron Energy
e) Filament Emission
f) Scan Rate
g) Data System
full mass range scanning
125 to 310
1600V
50V
40 ma
2 sec/spectrum
Incos 2300
43
-------
TABLE 13
Concentrations of Selected POMs in Low Burn
Species
Naphthalene
Fluorene
Anthracene
Fluoranthene
Pyrene
Benzo(e)pyrene
Benzo(a)pyrene
Perylene
Pitch Fume Sample
i Phenanthrene
k
anthrene +
thracene
riphenylene
ranthene
ranthene
me
;ne
-cd)pyrene
perylene
inthracene
Analytical m/z
128
166
178
202
202
228
228
252
252
252
252
252
276
276
278
Cone. (yg/mL)*
233
278
>1150**
>936**
>823**
244
267
57
75
35
57
15
13
11
2.5
* Concentration prior to dilution of 20X.
** Indicates saturated analytical ion.
44
-------
ring POMs
—
400
13:24
608
20(06
800
26:48
1000
33:30
EXPANDED BY 10X AT SCAN 1126
6+ ring POMs
12fJ8
40: 12
1400
46:54
1600
53:36
1800
60:18
2000
67i00
SCAN
TIME
FIGURE 10 . Reconstructed Ion Chromatogram for Sample of Fumes
from a Low Burn Pitch.
45
-------
[-100.9
3 ring POMs
i '
4 ring POMs
5 ring POMs
Ll.
200
6:42
400
13:24
£30
26:06
300
26:48
1003
33:30
EXPANDED 10X Al SCAN 1126
^J\jjw,
1 •
1200
48:12
1
^_
1400
46:54
1600
53 1 36
r~
1800
60:18
6+ ring POMs
2000
67:00
1
SCAN
TIME
FIGURE 11. Sum Ion Chromatogram for Sample of Fumes from a
Low Burn Pitch for Analytical Ions Listed in
Table 13.
46
-------
listed in Table 13 were reported.
To illustrate the resolving power attained under routine operating
conditions, Figure 3 (given earlier on p. 22) shows the typical selected
ion plots for the molecular ion of some common POMs from a calibration
run. Figure 3A shows the separation of equal concentrations of phenanthrene
and anthracene under the conditions of Table 12. The valley between the
peaks is about 20% of the peak height. Figure 3B shows the partial sepa-
ration between chrysene and triphenylene which are baseline resolved from
benz(a)anthracene. Typically, the former two isomers will not show any
separation when analyzed with packed column GC procedures. Figure 3C
shows the separation between benzo(e)pyrene, benzo(a)pyrene, and perylene
for these conditions. Since this type of separation is typical on glass
capillary columns, benzo(a)pyrene may be selectively analyzed with results
similar to those obtained using liquid crystal GC procedures.
47
-------
V. METHODS
This section presents detailed descriptions of the five procedures
presented for POM analysis. The first two procedures describe measure-
ment of total POMs, the remaining three describe measurement of specific
POM by GC/MS.
A. Total POM Measurement by Solution Fluorescence
1. Abstract
POMs are inherently flourescent materials with fairly broad and
intense emission and excitation bands. Since fluorescence emission
measurements are linear over a wide range of species concentrations,
the fluorescence emission for a sample may be used to estimate total
POM for that sample.
2. Interferences
The most prominent interferences on the fluorescence of POMs is
due to quenching of the emission by either high POM concentration or
by quenching agents such as highly-nitrated aromatics. These will be
minimized in dilute solution.
3. Sample Extraction
Sample extractions should be done using distilled-in-glass pentane
or methylene chloride. Samples should be concentrated to 1 to 10 mL
using a Kuderna-Danish evaporator. If necessary, samples may be further
concentrated to .1 mL using a gentle stream of nitrogen.
4. Sample Clean-up
The concentrated extract should be cleaned up using the EPA Level 1
Liquid Chromatography procedures (Appendix A). After clean-up, combine
fractions 2, 3, and 4 and reduce in volume to a small, convenient level
such as 2 mL using nitrogen.
5. Analysis
Measure the sample for total POM using a spectrofluorimeter with
350 ±5 nm excitation and 410 ±5 nm emission wavelengths. After any
48
-------
measurement, dilute sample by a factor of 2 to 5 and remeasure until the
emission decreases by the same factor; i.e., sample is in linear working
region where no significant concentration quenching is occurring. Cali-
brate the spectrometer emission against standard anthracene solutions
ranging in concentration from about 5 to 500 ng/mL of POM as anthracene.
B. Sensitized Fluorescence for Total POM
1. Principle; The fluorescence of a polycyclic aromatic hydro-
-4
carbon is greatly enhanced when it is present in trace quantities (10
to 10 mole ratio) in a solid aromatic hydrocarbon of lower molecular
weight, e.g., anthracene in naphthalene. In the case of isomers, the
less linearly conjugated one is a sensitizer for the more linearly
conjugated one(s), e.g., phenanthrene for anthracene.
2. Range and Sensitivity; Many PAH can be detected at 10 pg in
the presence of 6 to 60 yg of naphthalene. Benzo(a)pyrene has been
detected at 1 pg.
3. Interferences; Highly-nitrated aromatic compounds are known to
quench fluorescence of PAH. At low levels, however, that effect is
probably less likely than the transfer of energy to PAH.
4. Precision and Accuracy; Concentrations of PAH can be estimated
within a factor of 10 in the sensitized fluorescent spot test by 1:10
serial dilutions of the sample.
5. Apparatus; Sources are those used during study, and equivalent
sources are acceptable.
5.1 Ultraviolet source, 254 nm (Chromatovue Model C5)
5.2 Filter Paper (Whatman #42)
5.3 Pipets (Drummond Microcaps, 1 yL)
6. Reagents;
6.1 Naphthalene (Fisher Scientific Catalog #N-134, "Certified")
60 yg/yL.
6.2 Benzene or methylene chloride (Fisher Scientific,
Spectroanalyzed Grade).
49
-------
7. Procedure; This sensitized fluorescence spot test presupposes
that the sample has been obtained in an organic solvent either by direct
solution, extraction, or a separation procedure such as liquid chroma-
tography.
7.1 With pencil, mark three circles on filter paper$ each
approximately 0.25 cm in diameter.
7.2 With the paper supported so that marked spots are not in
contact with any other surface, apply 1 yL of sample
solution to central portion of each of two marked spots.
Allow to air dry, keeping spots from contacting other
surfaces.
7.3 Similarly apply 1 uL of naphthalene reagent solution to
remaining blank circle and to a spot containing sample.
7.4 Observe spots under 254 nm, viewing either side of
substrate. Note whether differences in intensity or
color exist between sample-reagent spot and either spot
alone, since naphthalene impurities may fluoresce. Any
difference indicates sensitized fluorescence. (At 1 ng
PAH the fluorescence of the sample spot itself should not
be evident.
Since the limits of detection are 1 to 10 pg PAH/spot for sensitized
fluorescence and approximately 10 ng/spot for non-sensitized fluorescence,
the results of the spot test procedure can be used to make the following
estimates of PAH contents in the 1 yL of sample.
a) non-fluorescent when treated with sensitizer : <1 pg
b) weakly fluorescent when treated with sensitizer : 1-10 pg
c) strongly fluorescent when treated with sensitizer
but not fluorescent alone : ^100 pg
d) fluorescent without sensitizer : >10,000 pg
From such estimates, the decision to proceed with further analyses can
be made. In the case of strong sensitized fluorescence, or fluorescence
without sensitization, a better estimate of concentration may be made
by directly testing dilutions of the sample solution.
50
-------
8. Calculations: In order to determine the PAH level in the
sample solution, the observation should be made on successive 1:10 dilu-
tions until the sensitized fluorescence is no longer observed. Under
the conditions for the test—a 1 uL sample volume and 10 pg as the lowest
detectable amount of PAH—the concentration of PAH in the solution can
be calculated with a factor of 10 as follows:
C = 1 x 10(n"6) g/L
where n = number of 1:10 dilutions.
The above formula is derived from the more explicit one:
C =io *io:12* xio11-1
1 x 10 L
For example, a solution sample that was diluted eight (8) times to reach
the point of no recognizable sensitized fluorescence would contain
1 x 10(8"6) = 100 g/L.
Since the sample solution used in this test may be an extract, LC
fraction, aliquot of another solution, or derived on some other way from
an original environmental assessment sample, the appropriate factors
must then be applied to compute the PAH content of that original sample.
9. Stability; The naphthalene sensitizer solution, kept in a
tightly-stoppered dark brown bottle, has been found to be stable over a
one-year period, but should be checked periodically.
C. POM Measurement by Packed Column GC/MS
1. Abstract of Method
The method is designed to measure POMs in environmental samples
using packed column GC/MS. Data are acquired to include the molecular
ions of all of the POM species of interest. Specific POMs are identi-
fied from their retention time relative to an internal standard and con-
firmed from their mass spectra. Data are reported either for specific
POMs or for combinations of specific POMs which are not resolved.
2. Interferences
Interferences in the POM measurements are due to coeluting species
51
-------
with mass spectra containing the POM molecular ion. Identity confirma-
tion by the mass spectra minimizes the effect of interferences on the
POM measurement.
3. Sample Extraction
Sample extractions should be done using distilled-in-glass pentane
or methylene chloride. Samples should be concentrated to 1 to 10 mL
using a Kuderna-Danish evaporator. If necessary, samples may be further
concentrated to 0.1 mL using a gentle stream of nitrogen.
4. Sample Clean-up
The concentrated extract should be cleaned up using the EPA Level 1
Liquid Chromatography procedure (Appendix A). After clean-up, combine
fractions 2, 3, and 4 and reduce in volume to a small, convenient level
such as 2 mL using nitrogen.
5. Analysis
a. GC Conditions
Use a2mx2mmI.D. glass column containing any of several
phases, i.e., OV-1, OV-101, OV-17, Dexsil 300 or Dexsil 400 at
1% to 3% loading on 80/100 or 100/120 Chromosorb. Temperature
program from about 150 C to about 300 C or to the upper temperature
limit of the column. An injection of 2—5 yL of sample is made onto
the column with the GC gas stream diverted from the mass spectro-
meter inlet line. After the solvent has eluted, the diverter is
closed and data acquisition initiated.
b. MS Conditions
Exact conditions will depend on the spectrometer type and con-
dition and the sensitivity required for the analysis. The spectro-
meter should collect data for the analytical ions of interest such
as the following:
POM Analytical m/z
Fluorene 166
Anthracene or Phenanthrene 178
52
-------
POM Analytical m/z
Fluoranthene or Pyrene 202
Chrysene, Triphenylene, etc. 228
Benzo-pyrenes or -fluoranthenes 252
Indeno pyrene or benzo(ghi)perylene 276
Dibenz-anthracenes, etc. 278
Dibenz pyrenes 302
Full mass scanning over the range of 125-310 is recommended.
6. Qualitative Identification
The relative retention time for the POM of interest, compared to the
internal standard present, is used to select the peak from the analytical
ion plot containing the POM of interest. The mass spectra of that peak
are examined to confirm the identity of the POM.
7. Quantitative Measurement
When the species are confirmed as POMs, individual mass chromato-
grams for the analytical m/e's are obtained. The peak areas for the
POMs of interest and that for the interanl standard are ratioed and com-
pared to a calibration curve for each POM (or POM group). The calibra-
tion standards and the samples should have the same amounts of internal
standard(s).
D. Measurement of Selected POMs by Liquid Crystal Column GC/MS
1. Abstract of Method
The method is designed to measure a few specific POMs in environ-
mental samples using liquid crystal phases for GC-MS. Data for limited
mass regions containing the POM and internal standard analytical ions
are acquired. Specific POM isomers are identified from their retention
time relative to the internal standard.
2. Interferences
Interferences in the POM measurements are due to coeluting species
with mass spectra containing the POM molecular ion. Identity confirma-
tion by the mass spectra minimizes the effect of interferences on the
53
-------
POM measurement.
3. Sample Extraction
Sample extractions should be done using distilled-in-glass pentane
or methylene chloride. Samples should be concentrated to 1 to 10 mL
using a Kuderna-Danish evaporator. If necessary, samples may be further
concentrated to 0.1 mL using a gentle stream of nitrogen.
4. Sample Cleanup
The concentrated extract should be cleaned up using the EPA Level 1
Liquid Chromatography procedure (Appendix A). After cleanup, combine
fractions 2, 3, and 4 and reduce in volume to a small convenient levels
such as 2 mL, using nitrogen.
5. Analysis
a. GC Conditions
Use a 2 m x 2 mm ID glass column containing a nematic liquid
crystal phase, such as SP-301 coated on 80/100 Supelcoport. Temperature
programming has been used but isothermal operation gives best results.
For example, for benzo(a)pyrene use isothermal operation at 260 C.
Inject a 2-5 yL sample onto the column, and collect data following the
elution of the solvent.
b. MS Conditions
Exact conditions will depend on spectrometer type and condition.
Either selected ion monitoring or selected mass range scanning should
be used to minimize background and enhance the signal-to-noise ratio.
The spectrometer should collect data to include the analytical ions of
interest. For example, to analyze benzo(a)pyrene with 9,10-diphenyl
anthracene as the internal standard, collect data in the following
ranges:
Species Mass Range Analytical m/z
Benzo(a)pyrene 240 - 260 252
9,10-Diphenyl anthracene 320 - 340 330
54
-------
6. Qualitative Identification
The relative retention time for the POM of interest, compared to the
internal standard present, is used to select the peak from the analytical
ion plot containing the POM of interest. The mass spectra of that peak
are examined to confirm the identity of the POM.
7. Quantitative Measurement
When the species are confirmed as POMs, individual mass chromatograms
for the analytical m/z's are obtained. The peak area for the POMs of
interest and that for the internal standard are ratioed and compared to a
calibration curve for each POM (or POM group). The calibration standards
and the samples should have the same amount of internal standard, such
as 9,10-diphenyl anthracene, added to them.
E. POM Measurement by Capillary Column GC-MS
1. Abstract of Method
The method is designed to measure POMs in environmental samples using
capillary column GC-MS. Data are acquired to include the molecular ions
of all of the POMs of interest. Specific POMs are identified from their
retention time relative to an internal standard and confirmed from their
mass spectra. Data are reported either for specific POMs or for combi-
nations of specific POMs which are not resolved.
2. Interferences
Interferences in the POM measurements are due to coeluting species
with mass spectra containing the POM molecular ion. Identity confirma-
tion by the mass spectra minimizes the effect of interferences on the
POM measurement.
3. Sample Extraction
Sample extractions should be done using distilled-in-glass pentane
or methylene chloride. Samples should be concentrated to 1 to 10 ml
using a Kuderna-Danish evaporator. If necessary, samples may be further
concentrated to 0.1 mL using a gentle stream of nitrogen.
55
-------
4. Sample Cleanup
The concentrated extract should be cleaned up using the EPA Level 1
Liquid Chromatography procedure (Appendix A). After cleanup, combine
fractions 2, 3, and 4 and reduce in volume to a small convenient level,
such as 2 mL» using nitrogen.
5. Analysis
a. GC Conditions
Use a 20-30 m glass capillary column coated with a silicone phase,
e.g., OV-101, OV-17, or SE-52. Depending upon sample concentration,
use either split or splitless GC injection of 1 or 2 yL of sample, with
splitless on-column injection preferred. Temperature-program the column
to separate the various POMs. For example, with a splitless injection
of sample in methylene chloride, a typical program might be: 35 C iso-
thermal for 1-4 min., heat at 20°C/min. to 200°C, then heat at 2 or 3°C/min.
to just below the upper temperature limit of the column, and hold as
necessary to completely elute the sample. Data acquisition should start
after the elution of the solvent.
b. MS Conditions
Exact conditions will depend on the spectrometer type and condition
and the sensitivity required for the analysis. The spectrometer should
collect data for the analytical ions of interest such as the following:
POM Analytical m/z
Fluorene 166
Anthracene or phenanthrene 178
Fluoranthene or pyrene 202
Chrysene, Triphenylene, etc. 228
Benzo-pyrenes or -fluoranthenes 252
Indeno pyrene or benzo(ghi)perylene 276
Dibenz-anthracenes, etc. 278
Dibenz pyrenes 302
Full mass scanning over the range of 125-310 is recommended.
56
-------
6. Qualitative Identification
The relative retention time for the POM of interest, compared to the
internal standard present, is used to select the peak from the analytical
ion plot containing the POM of interest. The mass spectra of that peak
are examined to confirm the identity of the POM.
7. Quantitative Measurement
When the species are confirmed as POMs, individual mass chromatograms
for the analytical m/e's are obtained. The peak areas for the POMs of
interest and that for the internal standard are ratioed and compared to
a calibration curve for each POM (or POM group). The calibration stan-
dards and the samples should have the same amount of internal standard(s).
Use of multiple internal standards, such as d -anthracene, 9-phenyl-
anthracene, and 9,10-diphenylanthracene, which have boiling points covering
the temperature range of the analysis, will minimize errors that might
arise due to variation of the split ratio with respect to molecular weight
and/or boiling point for the compounds of interest.
57
-------
REFERENCES
1. Committee on Biologic Effects of Atmospheric Pollutants, Div. of
Medical Sciences, National Research Council, "Particulate Poly-
cyclic Organic Matter." National Academy of Sciences, Washington,
DC, 1972.
2. Hangebrauck, R.P., D.J. von Lehmden and J.E. Meeker, "Sources of
Polynuclear Hydrocarbons in the Atmosphere." EPA No. AP-33, NTISPB
174-706.
3. Blumer, M., "Polycyclic Aromatic Compounds in Nature," Scientific
American, March 1976, p. 35.
4. Potts, P., "Chirurgical Observations," (1775), reprinted in
National Cancer Institute Monograph jj), 7 (1963).
5. Yamagiwa, K., and K. Ichikawa, Mitt, med Fak. Tokio 15. 295 (1915).
6. Block, B., and H. Dreifus, Schweiz. med. Wschr. 51, 1033 (1921).
7. Kennaway, E.L., Brit, med. J;. ii. 1 (1925).
8. Cook, J.W., C.L. Hewett, and I. Heiger, _J. Chem. Soc., 395 (1933).
9. Clar, E., "Polycyclic Hydrocarbons, ^," London, Academic Press,
1964, p. 139.
10. Freudenthal, R.I., and P. Jones, (ed.) "Carcinogenesis. A Compre-
hensive Survey, Vol. 1." Raven Press, New York, NY, 1976.
11. Jones, P.W., and R.I. Freudenthal, "Carcinogenesis, A Comprehensive
Survey. Vol. 3." Raven Press, New York, NY, 1978.
12. Schaad, R., Chromatography Reviews 13, 61 (1970).
13. Cook, J.W., and C.L. Hewett, J. Chem. Soc., 398 (1933).
14. La Budde, J.A., and C.J. Heidelberger, JAGS 80. 1225 (1958).
15. Berenblum. I., and R. Schoental, Cancer Research 6, 699, (1946).
16. Cook, J.W., R.S. Ludwiczak, and R. Schoenthal, J_. Chem. Soc.,
1112 (1950).
17. Hood, L.V., and J.D. Winefordner, Anal. Chem. Acta 42, 199 (1968).
18. Sawicki, E., R.C. Corey, A.E. Dooley, J.B. Gisclard, J.L. Monkman,
J.B. Neligan, and L.P. Ripperton, Health Lab. Sci. ]_, 56 (1970),
60 (1970), 68 (1970).
58
-------
19. Matsushita, H., U. Esumi, and K. Yamada, Bunseki Kagaku 19, 951,
(1970).
20. Sawicki, E., Talanta 16. 1231 (1969).
21. Wehry, E.L., G. Mamantov, R.R. Kemmerer, R.C. Stroupe,
P.T. Tokousbalides, E.R. Hinton, D.M. Hembree, R.B. Dickinson, Jr.,
A.A. Garrison, P.V. Bilotta, and R.R. Gore, "Carcinogenesis, Vol 3:
Polynuclear Aromatic Hydrocarbons," P.W. Jones and R.I. Fredenthal
(eds.), Raven Press, New York, NY, pp. 193-202 (1978).
22. VoDinh, T., Anal. Chem. 50, 396 (1978).
23. Sawicki, E., T.W. Stanley, and W.C. Elbert, J. Chromatog. 20,
348 (1965).
24. Sawicki, E., R.C. Corey, A.E. Dooley, J.B. Gisclard, J.L. Monkman,
R.E. Neligan, and L.A. Ripperton, Health Lab. Scj. _7, Supplement 3,
31 (1970).
25. Searl, T.D., F.J. Cassidy, W.H. King, and R.A. Brown, Anal. Chem. 42,
954 (1970).
26. Matsushite, H., Y. Esumi, and K. Yamada, Bunseki Kagaku 19, 951,
(1970).
27. "Ultraviolet Atlas of Organic Compounds," 5 Vols., Plenum Press,
New York, NY, 1967-1968.
28. Ungnade, H.E. (ed.), "Organic Electronic Spectral Data,"
Interscience Publishers, Inc., New York, NY, 1960-present.
29. Wilmhurst, S.R., j;. Chromatog. 17, 50 (1965).
30. Deamio, L., and M. Corn, Anal. Chen. ^8, 131 (1966).
31. Golay, J.J.E., "Theory and Practice of Gas Liquid Partition Chroma-
tography with Coated Capillaries in Gas Chromatography," V.J. Coates,
H.J. Noebels, and I.S. Ragerson (eds.), Academic Press, New York,
NY, (1958).
32. Lee, M.L., D.L. Vassilaros, C.M. White, and M. Novotny, Anal. Chem.
_51, 768 (1979).
33. Janini, G.M., K. Johnston, and W.L. Zielinski, Jr., Anal. Chem. 47,
670 (1975).
34. Oi, W., and E. Inabe, Yakugaku Zasshi. 87, 741 (1967).
35. Harmon, D.M., J_. Ass. Off. Analytical Chem., 57, 495 (1974).
59
-------
36. Nagy, F., and J. Sohar, Chem. Zvesti. 16, 389 (1962).
37. Martrka, M., J. Kroupa, and A. Spevak, Coll. Cs. Chem. Cominun. _36,
1379 (1971).
38. Sawicki, E., R.C. Corey, A.E. Dooley, J.B. Gisclard, J.L. Monkman,
R.E. Neligan, and L.A. Ripperton, Health Lab. Sci., _7» 45 (1970).
39. Jones, P.W., personal communication.
40. Smith, E.M., and P.L. Levins, "Sensitized Fluorescence for the
Detection of Polycyclic Aromatic Hydrocarbons," EPA 60017-78-182,
NTISPB 287-182/AS, (1978).
41. Lipsett, F.R., and A. Dekker, Can. J. Phys.. 30, 165 (1951).
42. "Sampling and Analysis Procedures for Screening of Industrial
Effluents for Priority Pollutants," (Draft) EPA/EMSL, Cincinnati,
Ohio, April, 1977.
43. Levins, P.L., C.E. Rechsteiner, and J.L. Stauffer, "Measurement of
PCB Emissions from Combustion Sources," EPA 600/7-79-047, NTISPB
293-360, (1979).
44. Strand, J.W., and A.W. Andren, Anal. Chem.. 50, 1508 (1978).
45. Robertson, D.J., J.H. Elwood, and R.H. Broth, "Chemical Composition
of Exhaust Particles from Gas Turbine Engines," EPA-600/2-79-041,
NTISPB 292 380/3SL.
46. Lentzen, D.E., D.E. Wagoner, E.D. Ester, and W.F. Gutknecht,
"IERL-RTP Procedures Manual: Level 1 Environmental Assessment
(2nd ed.)", EPA 600/7-78-201, NTISPB 293-795, (1978).
60
-------
APPENDIX A*
EPA Level 1 - Liquid Chromatographic (LC) Separation
All sample extracts, neat organic liquids, and SASS-train-dried
probe/cyclone rinse extracts are subjected to LC separation if sample
quantity is adequate. An aliquot of the concentrated extract containing
100 rag of organic matter is preferred for the LC>but smaller quantities
down into a lower limit of about 15 mg may be used. The sample components
are separated according to polarity on silica gel using a step gradient
elution technique. The detailed procedure for the LC separation is given
below:
Column: 200 mm x 10.5 mm I.D., glass with Telfon stopcock,
waterjacketed with inlet water temperature in the
range of 18° to 22°C and sufficient flow to maintain
this temperature through to the outlet.
Adsorbent: Davison, Silica Gel, 60-200 mesh, Grade 950 (avail-
able from Fisher Scientific Company) is to be used;
no other types or grades of silica gel can be
substituted. This material should be cleaned prior
to use by sequential Soxhlet extractions with methanol,
methylene chloride, and pentane. This adsorbent is
then activated at 110 C for at least two hours just
prior to use and cooled in a desiccator.
Drying Agent: Sodium Sulfate (Anhydrous, Reagent Grade). Clean
by sequential Soxhlet extraction for 24 hours each
with methanol, methylene chloride, and pentane. Dry
for at least two hours at 110 C just prior to use
and cool in a desiccator.
9.4.4.1 Procedure for Column Preparation
The chromatographic column, plugged at one end with a small portion
*Reference 46, Section 9.4.4
61
-------
of glass wool, should be slurry packed with 6.0 g of freshly-activated
silica gel in n-pentane. A portion of properly-activated silica gel
weighing 6.0 ±0.2 g occupies 9 mL in a 10 mL graduated cylinder. The
total height of the silica gel in this packed column is 10 cm. The
solvent void volume of the column is 2 to 4 mL. When the column is fully
prepared, allow the pentane level in the column to drop to the top of
the silica bed so that the sample can be loaded for subsequent chromato-
graphic elution.
After packaging the silica gel column, add 3 g ±0.2 g clean sodium
sulfate to the top of the column. Vibrate for 1 min. to compact. The
sodium sulfate should occupy 2 mL in a 10 mL graduated cylinder. The
sodium sulfate will remove small quantities of water from the organic
extract; however, appreciable quantities of water will solidify the
sodium sulfate, inhibiting proper flow through the column. Therefore,
it is advisable that if enough water is present in the sample to form
two layers, it should be removed by another method—pipet or separatory
funnel.
9.4.4.2 Evaporation of Sample Extracts with Low Total Chromatographicable
Organics (TCO) (<2 mg original sample)—
For these samples, the aliquot of extract containing 15 mg (minimum)
to 100 mg (preferred) of material is added to a small amount of silica
gel, the solvent is allowed to evaporate, and the residue plus silica
gel is transferred to the LC column with the aid of a microspatula. The
container is rinsed as described in Section 9.4.4.5.
9.4.4.3 Solvent Exchange of Sample Extract with High Total Chromatograph-
icable Organics (TCO) (>2 mg original sample)—
An aliquot of methylene chloride extract containing 15 mg (minimum)
to 100 mg (preferred) of material is added to 200 mg of silica gel in a
graduated receiver. The volume of extract is carefully reduced to 1 mL
at ambient temperature under a gentle stream of nitrogen (tapped from a
liquid nitrogen cylinder, if possible, to minimize impurities). The
solvent evaporates rapidly, so it is important that this operation be done
under constant surveillance to insure that the volume is not reduced
below 1 mL. It is also necessary to warm the samples slightly, either
62
-------
by hand or water bath, at <40 C, to prevent condensation of atmospheric
moisture in the sample due to evaporative cooling. One milliliter of
cyclopentane is added and mixed by gentle agitation. The volume is
reduced to a total of 1 mL as before. A second milliliter of cyclopentane
is added, mixed, and the volume is again reduced to 1 mL. The exchange
is repeated with a third milliliter of cyclopentane. After the volume
has been reduced to 1 mL for this last time, the solvent mixture will be
<5 percent methylene chloride. This is sufficiently low to prevent
breakthrough of aromatic sample components into the aliphatic hydrocarbon
fraction, LCI.
The cyclopentane and silica gel are transferred to the top of the
previously prepared LC column using a Pasteur pipet. The container is
rinsed as described in Section 9.4.4.5.
9.4.4.4 Neat Organic Liquids—
A 100 mg sample is weighed into a tared glass weighing funnel and
mixed with about 200 mg of silica gel using a microspatula. The sample
is then transferred to the top of the column. The container is rinsed
as described in Section 9.4.4.5.
When neat organic liquids are fractionated by the liquid chromato-
graphy scheme, they have the same theoretical gravimetric detection
limitations as other samples separated by this means, 0.1 mg/100 mg or
0.1 percent of the sample applied. Since these aliquots are neat samples
and do not have concentration factors as multipliers, the resultant
detection limits for minor components are 1 g/kg at best.
9.4.4.5 Chromatographic Separation into Seven Fractions—
Table A-l shows the sequence for the Chromatographic elution. In
order to insure adequate resolution and reproducibility, the column
elution rate is maintained at 1 mL/min.
The volume of solvents shown in Table A-l represents the solvent
volume added to the column for that fraction. If the volume of solvent
collected is less than the volume actually added due to evaporation,
restore the fraction volume to the proper level with fresh solvent. In
all cases, the solvent level in the column should be at the top of the
gel bed, i.e., the sample-containing zone, at the end of the collection
63
-------
TABLE A-l LIQUID CHROMATOGRAPHY ELUTION SEQUENCE
Fraction
1
2
3
4
5
6
7
*POM
Solvent Composition
Pentane
20% Methylene chloride in pentane
50% Methylene chloride in pentane
Methylene chloride
5% Methanol in methylene chloride
20% Methanol in methylene chloride
50% Methanol in methylene chloride
containing fraction
Volume
25
10*
10*
10*
10
10
10
64
-------
of any sample fraction. The fractions are retained as solutions for TCO
analyses.
After the first fraction is collected, rinse the original sample
container or weighing funnel with a few milliliters of Fraction 2 solvent
(20 percent methylene chloride/pentane) and carefully transfer this
rinsing into the column. Repeat with each successive solvent mixture in
turn.
Add each new solvent to the column slowly to minimize disturbing
the gel bed and eliminate the trapped air bubbles, particularly in the
zone of the sample-containing silica gel.
After each sample is collected, an aliquot (1 to 5 yL) is taken for
TCO analysis of each fraction (unless the sample taken for LC had a TCO
of <2 mg). Also, an aliquot (10 mL for Fraction 1 and 5 mL for Fractions
2-7) is transferred to a tared aluminum micro weighing dish for evapora-
tion and gravimetric analysis. The GRAV data for Fraction 7 must be
corrected for a blank contributed by a small quantity of silica gel that
dissolves in the highly polar eluent. The blank value is determined by
running an LC column in which no sample is added; it is on the order of
0.9 ±0.1 mg in LC7 (10 mL). After TCO and GRAV determinations, the
fractions are analyzed by IR and, when the quantity is defficient, by
LRMS.
The objective of the LC procedure is to separate the sample into
fractions of varying chemical class type to facilitate subsequent analyses.
The LC separation procedure is not a high resolution technique and,
consequently, there is overlap in class type between many of the fractions.
With respect to POM analysis, the three fractions (2, 3, and 4) should
be combined and the resulting solution reduced in volume to 2 mL prior
to analysis.
65
-------
TECHNICAL REPORT DATA
(Please read Inunctions on the reverse before completing}
1 REPORT NO
EPA-600/7-79-191
4. TITLE AND SUBTITLE
Measurement of Polycyclic
2.
Organic Matter for
Environmental Assessment
7 AUTHOR(S)
P. L. Levins , C. E. Rechsteiner Jr. , and J. L. Stauffer
9 PERFORMING ORGANIZATION NAME Alv
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts
JD ADDRESS
02140
12 SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
August 1979
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
INE624
11. CONTRACT/GRANT NO.
68-02-2150, Task 10202
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 12/76 - 3/79
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES JERL-RTP project officer is Larry D. Johnson, Mail Drop 62,
919/541-2557.
16 ABSTRACTThe report discusses methods of measuring polycyclic organic matter (POM)
for environmental assessment. It describes two fluorescence methods of estimating
total POM levels in samples. Either method may be used to screen samples for
further specific analyses. It also describes three gas chromatography/mass spectro-
metry (GC/MS) methods for measuring specific POM compounds. The use of liquid
crystal chromatographic phases is recommended for measuring a few POMs; e.g. ,
specifically for benzo(a)pyrene. It discusses GC/MS methods for a wide range of
POMs for both capillary and packed GC columns. The three methods for specific
POM identification have been verified with collected environmental samples for
different kinds .
17.
a. DESCRIPTORS
KEY WORDS AND DOCUMENT ANALYSIS
b. IDENTIFIERS/OPEN ENDED TERMS
Pollution Gas Chromato- Pollution Control
Assessments graphy Stationary Sources
Measurement Mass Sprectro- Environmental Ass ess -
Polyclic Compounds scopy ment
Organic Compounds Polycyclic Organic Mat-
Fluorescence ter
Benzo(a)pyrene
13. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COS ATI Field/Group
1313
14B 07D
07C
20F
21. NO. OF PAGES
73
22. PRICE
EPA Form 2220-1 (9-73)
66
------- |