PB83-156935
Measurements of Hazardous Organic
Chemicals in the Ambient Atmosphere
SRI International
Menlo Park, CA
Prepared for
Environmental Sciences Research Lab.
Research Triangle Park, NC
Jan 83
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
-------
EPA-600/3-83-002
January 1983
MEASUREMENTS OF HAZARDOUS ORGANIC CHEMICALS
IN THE AMBIENT ATMOSPHERE
by
H.B. Singh, LJ. Salas
R. Stiles, and H. Shigeishi
Atmospheric Science Center
SRI International
Menlo Park, California 94025
Cooperative Agreement
805990
Project Officer
L. Cupitt
Atmospheric Chemistry and Physics Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK. NORTH CAROLINA 27711
-------
TECHNICAL REPORT DATA
(Please read Imtrucf.em on the reverse before completing)
1. REPORT NO.
EPA-600/3-83-002
3. RECIPJJ=NJ.'S ACCESSION NO.
1569? ?
4. TITLE AND SUBTITLE
MEASUREMENTS OF HAZARDOUS ORGANIC CHEMICALS IN THE
AMBIENT ATMOSPHERE
5. REPORT DATE
January 1983
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
H. B. Singh, L. J. Salas, R. Stiles, and H. Shigeish
8. PERFORMING ORGANIZATION REPORT NO.
i SRI Project 7774
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Atmospheric Sciences Center
SRI International
333 Ravenswood Avenue
Menlo Park, California 94025
10. PROGRAM ELEMENT NO.
C9TA1B /01-0352 (FY-83)
11. CONTRACT/GRANT NO.
CA 805990
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory-RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final 9/78-10/81
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Analytical methods were refined and applied to the ambient analysis of 44
organic .chemicals, many of which are bacterial mutagens or suspected carcinogens.
On-site field collection programs, based on single site studies of 9 to 11 days
duration each, were conducted in 10 U.S. cities. Field studies were performed
with an instrumented mobile laboratory. A round-the-clock measurement schedule
was followed at all sites. The field measurements allowed a determination of
atmospheric concentrations, variabilities, and mean diurnal behaviors of the
chemicals. The data were analyzed relative to theoretically estimated removal
rates. Typical diurnal profiles show highest concentrations of the primary
pollutants during nighttime or early morning hours, with minimum concentrations
in the afternoon hours. Chemistry plays only a nominal role in defining this
diurnal behavior in most cases. Except for aromatic hydrocarbons and aldehydes,
average concentrations of the measured species were in the 0- to 5-ppb range.
The average concentration range observed for aromatics and aldehydes was 0- to
20-ppb. '
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report I
UNCLASSIFIED
21. NO. OF PAGES
99
20. SECURITY CLASS (This page I
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (R«v. 4-77) PREVIOUS EDITION is OBSOLETE
-------
NOTICE
THIS DOCUMENT HAS BEEN REPRODUCED
FROM THE BEST 'COPY FURNISHED US BY
THE SPONSORING AGENCY. ALTHOUGH IT
IS RECOGNIZED THAT.CERTAIN PORTIONS
ARE ILLEGIBLE, IT IS BEING RELEASED
IN THE INTEREST OF MAKING'-A VAI LA B L E
AS MUCH INFORMATION AS POSSIBLE.
LIBRARY
ENVIRONMENTAL RESEARCH IAPORATCR;
PUI.UTH, MINNESOTA
-------
DISCLAIMER
This report has been reviewed by the Office of Research and
Development, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the con-
tents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or recom-
mendation for use.
11
-------
ABSTRACT
Analytical methods were refined and applied to the ambient analysis of
44 organic chemicals, many of which are bacterial mutagens or suspected
carcinogens. On-site field collection programs, based on single site
studies of 9 to 11 days duration each, were conducted in 10 U.S. cities.
Field studies were performed with an instrumented mobile laboratory.
A round-the-clock measurement schedule was followed at all sites. The
field measurements allowed a determination of atmospheric concentrations,
variabilities, and mean diurnal behaviors of the chemicals. The data were
analyzed relative to theoretically estimated removal rates. Typical
diurnal profiles show highest concentrations of the primary pollutants
during nighttime or early morning hours, with minimum concentrations in the
afternoon hours. Chemistry plays only a nominal role in defining this
diurnal behavior in most cases. Except for aromatic hydrocarbons and
aldehydes, average concentrations of the measured species were in the 0-
to 5-ppb range. The average concentration range observed for aromatics .and
aldehydes was 0- to 20-ppb.
-------
CONTENTS
Abstract ..... ill
Figures vi
Tables. vlii
1. Introduction.. 1
2. Overall Objectives 2
3. Analytical Methodology 3
Trace Constituents of Interest 3
Field Instrumentation... 5
Experimental Procedures 5
Calibrations 8
Quality Control 22
4. Plan of Field Measurements. 26
5. . Estimated Loss Rates of Measured Chemicals
in Polluted Atmospheres..... 29
6. Analysis and Interpretation of Field Data,
Results, and Discussion 32
Atmospheric Concentrations and Variabilities >
of Measured Species... 32
Interpretation of Field Data
by Chemical Category 37
7. Summary and Conclusions .... 80
8. Recommendations for Future Research ' 83
References 84
Preceding page blank
-------
FIGURES
Number Page
1 Chromatogram showing ambient analysis of selected
chlorinated and brominated toxic chemicals 10
2 Chromatogram showing separation of methyl halides
and chlorofluorocarbons from ambient air 11
3 Analysis of aromatic hydrocarbons in ambient air 12
4 Halogenated aroma tics in the air 13
5 Analysis of 1,2 dichloroethane 14
6 Chromatogram showing methylene chloride separations
from other halocarbons 15
7 PAN and PPN separation from the air 16
8 Separation of dinitrophenyl hydrazones of formaldehyde
and acetaldehyde on HPLC 17
9 Permeation tube holder 17
10 Permeation tube weight-time relationship for selected
halogenated chemicals • 20
11 Permeation tube weight loss of selected oxygenated
chemicals 20
12 Secondary standard response for 1,1,1 trichloroethane
and benzene in Phoenix—Site 2 23
13 Calibration curve for dinitrophenyl hydrazones
of acetaldehyde and its repeatability 24
14 Methyl chloride in the ambient air of selected cities 40
15 Atmospheric concentrations of methyl bromide 41
16 Mean diurnal variation of methyl iodide 43
17 Mean diurnal variation of methylene chloride
at selected sites.... 44
18 Mean diurnal variation of chloroform at Phoenix, AZ 45
19 Atmospheric concentrations of chloroform
at Staten Island, NY 46
vi
-------
Number Page
20 Atmospheric concentrations of carbon tetrachloride
at selected sites 47
21 Mean diurnal variations of carbon tetrachloride
at Staten Island, NY 48
22 Atmospheric concentrations of ethyl chloride 49
23 Mean diurnal variation of 1,1 dichloroethane 50
24 Atmospheric concentration of 1,2 dichloroethane 51
25 Mean diurnal variation of 1,2 dichloroethane. 52
26 Mean diurnal variation of 1,2 dibromoethane 54
27 Atmospheric, concentrations of 1,2 dibromoethane 55
28 Mean diurnal variation of 1,1,1 trichloroethane. 56
29 1,1,1 trichloroethane behavior at Staten Island, NY............ 57
30 Mean diurnal variation of 1,2 dichloropropane. 59
31 Mean diurnal variation of trichloroethylene 61
32 Trichloroethylene behavior at Pittsburgh, PA 62
33 Mean diurnal variation of tetrachloroethylene. 63
34 Tetrachloroethylene behavior at Pittsburgh, PA 64
35 Mean diurnal variation of monochlorobenzene
at Denver, CO 66
36 Mean diurnal variation of o-dichlorobenzene
at Phoenix, AZ 67
37 Mean diurnal variation of m-dichlorobenzene 68
38 Mean diurnal variation of 1,2,4 trichlorobenzene
at Riverside, CA . 69
39 Mean diurnal variation* of benzene 70
40 Mean diurnal vairation of toluene. 71
41 Mean diurnal variation of m/p-xylene 72
42 Aromatic hydrocarbons at Pittsburgh, PA 73
43 Comparison of formaldehyde concentrations as measured
by the chromotropic acid and the DNPH-HPLC procedure 76
44 Average diurnal variation of PAN and PPN
At Phoenix—Site 2 78
vii
-------
TABLES
Number Page
1 List of Target Chemicals 4
2 Environmental Mobile Laboratory Instrumentation 6
3 Analytical Conditions for the Analysis
of Selected Toxic Chemicals 9
4 Permeation Rate Data for Generating Primary Standards 19
5 PPM Level Primary Standards in Air 21
6 Interlaboratory Comparison of Selected Trace Constituents 25
7 Field Sites and Measurement Schedule 27
8 Estimated Daily Loss Rates (%) of Selected Trace Chemicals 30
9 Production, Emission and Usages of Selected Chemicals 33
10 Atmospheric Concentrations of Measured Chemicals (Site 1-3).... 34
11 Atmospheric Concentrations of Measured Chemicals (Site 4-7).... 35
12 Atmospheric Concentrations of Measured Chemicals (Site 8-10)... 36
13 Average Background Concentration of Trace Species
at 40°N for Year 1981 38
14 Ambient Formaldehyde Levels in Selected Locations
as Measured with the Chromotropic-Acid Procedure 74
15 Comparison of Formaldehyde and Acetaldehyde Data 75
viii
-------
SECTION 1
INTRODUCTION
A recent report from the Office of the U.S. Surgeon General concluded
that "toxic chemicals are adding to the disease burden of the United States in
a significant, although as yet not precisely defined way" (U.S. SG, 1980):
Estimates suggesting that 50 to 90 percent of human cancer is of chemical ori-
gin continue to persist (LaFond, 1978; U.S. SG, 1980). The degree to which
the general ambient environment contributes to human cancer is a matter of
both active research and debate (Peto, 1980). There is little doubt, however,
that over the last three decades large amounts of a growing number of syn-
thetic organic chemicals have been released into the ambient environment. In
many cases, virtually the entire quantity of synthetic organic chemicals
manufactured is released into the environment as a necessary outcome of use
(ADL, 1975; Singh et al., 1979a). Urban atmospheres contain a complex mixture
of a.large number of chemicals, many of which are known to be toxic at concen-
trations significantly higher than those encountered in typical ambient atmo-
spheres. The process of understanding the risks associated with exposure to
potentially hazardous chemicals requires a determination of the ranges of con-
centrations that can be found in the ambient air.
This study was initiated primarily to examine the range of concentrations
of a variety of potentially hazardous gaseous organic chemicals* at selected
urban locations under varying meteorological and source-strength conditions.
These chemicals were measured and analyzed on-site in ambient air using a
suitably outfitted mobile laboratory. The overall program of analytical
methods development, field measurements, data collection, and data analysis is
expected to provide information that will permit a better assessment of the
atmospheric abundance and chemistry of this potentially harmful group of
chemicals.
*The term "hazardous chemicals" as used here is not intended to imply that a
proven human health hazard exists. In most cases toxicity studies are incom-
plete or inconclusive and involve extrapolation of animal data to humans.
1
-------
SECTION 2
OVERALL OBJECTIVES
The overall objective of the study was a survey of the ranges of concen-
trations of selected hazardous organic chemicals which may be found in urban
atmospheres within the United States.
To achieve this general objective, the following approach was used:
• Develop procedures for the sampling and analysis of selected organic
chemicals, at expected ambient concentrations.
• Equip and prepare a mobile environmental laboratory to conduct on-site
and around-the-clock measurements of chemical species of interest.
• Conduct field measurements at several locations with the primary pur-
pose of developing a reliable data base that could be used to better
understand the concentrations and diurnal behavior of these chemicals.
• Develop and synthesize information from the literature on sources,
fates, and effects of these potentially hazardous chemicals.
• Prepare a final report that combines information developed from the
preceding tasks.
-------
SECTION 3
ANALYTICAL METHODOLOGY
TRACE CONSTITUENTS OF INTEREST
The chemicals targeted for this study were those suspected of being
hazardous and chemicals structurally similar to these. Although toxicity data
in many cases were insufficient to prove a human health hazard, they were ade-
quate to merit inclusion of the chemical in our measurement plan. Many
selected chemicals were either bacterial mutagens or suspected of being carci-
nogens. Some nontoxic chemicals, such as chlorofluorocarbons, were measured
primarily because of their ability to act indicators of anthropogenic pollu-
tion. The final list of chemicals to be measured was based on further dis-
cussions with the project officer; our ability to satisfactorily measure a
trace constituent at its expected ambient concentration was an essential
requirement.
A total of 45 trace chemicals were targeted and are,categorized in Table
l...The categories include chlorofluorocarbons, halomethanes, haloethanes,
halopropanes, chloroalkenes, chloroaromatics, and oxygenated species. In
addition to the chemicals of Table 1, other important meteorological parame-
ters (wind speed, wind direction, temperature, pressure, relative humidity,
and solar flux) were also measured.
Table 1 also identifies more than two dozen chemicals-as bacterial
mutagens (BM) or suspected carcinogens (SC)i This information is obtained
from literature and studies.that have evaluated large bodies of available data
(Helmes et al., 1980; Albert, 1980; U.S. SG, 1980). Information about bac-
terial mutagenicity is based largely on the "Ames Salmonella Microsome Assay"
(McCann-and Ames, 1977). In some cases other bacterial tests have also been
utilized (BM(0)]. It is relevant to add that nearly 90 percent of tested
animal chemical carcinogens are also found to be mutagens in the
"salmonella/microsome" test, while an equal percentage of tested noncarcino-
gens are found to be nonmutagens (McCann and Ames, 1977). Mutagenic tests are
direct and simple, but the carcinogenicity information is based on epidemiol-
ogy, animal tests, and a critical and a comprehensive evaluation of carcino-
genic, mutagenic, and other toxicological data (Albert, 1980; U.S. SG, 1980).
Evidence for the mutagenicity of toluene (U.S. SG, 1980; Albert, 1980) and
carcinogenicity of trichlorethylene (Albert, 1980) is currently in some
dispute for lack of sufficient data.
-------
TABLE 1. LIST OF TARGET CHEMICALS
Chemical Name*
Chloro-Fluorocarbons
Trichloromonof luorome thane (Fll)
Dichlorodifluoromethane (F12)
Trichlorotrifluoroethane (F113)
Dlchlorotecraf luoroethane (F114)
Hal ome thanes
Methyl chloride
Methyl bromide
Methyl iodide
Methylene chloride
Chloroform
Carbon tetrachlorf.de
Haloe thanes and halopropanes
Ethyl chloride
1,1 Dichloroethane
1,2 Dichloroethane
1,2 Dlbronoethane
1,1,1 Trlchloroethane
1,1,2 Trlchloroethane
1,1,2,2 Tetrachloroethane
1,2 Dichloropropane
Chloroalkenes
Vinylidene chloride
(els) 1,2 Dichloroethylene
Trlchloroethylene
Tetrachloroethylene
Allyl chloride
Hexachloro-1 ,3 butadiene
Chi oroaroma tics
Monochlorobenzene
a-Chlorotoluene
o-Dlchlorobenzene
m-Dichlorobenzene
p-Oic hi oro benzene
Ir2,4 Tricltlorobenzene
Aromatic hydrocarbons
Benzene
Toluene
Ethyl benzene
m/p-Xylene
o-Xylene
4-Ethyl toluene
1,2,4 Trimethyl benzene
1,3,5 Trimethyl benzene
Oxygenated and nitrogenated species
Formaldehyde
Ace caldehyde
Phosgene
Peroxyacetyl nitrate (PAN)
Peroxypropionyl nitrate (PPN)
Chemical Formula
CC13F
CC12F2
CC12FCC1F2
CC1F2CC1F2
,CH3C1
CH3Br
CH3I
CII2C12
CHC13
CC14
C2H5C1
CIIC12CII3
CH2C1CH2C1
CH2BrCH2Br
CII3CC13
CH2C1CHC12
CIIC12CHC12
CH2C1CHC1QI3
C1!2=CC12
CHC1=CHC1
CIIC1-CC12
CC12=CC12
Clai2CH=CH2
C12C-CC1-CC1=CC12
C6H5C1
C6H5C1I2C1
0-C6H4C12
m-CgIl4Cl2
P-C6II4C12
1,2,4 C6H3C13
C6H6
C6H5CII3
C6H5C2H5
m/p-C6H4(ClI3)2
o-C6H4(CH3)2
4-C6H4C2H5CH3
1,2,4 C6H3(CH3)3
1,3,5 C6H3(C]I3)3
HCHO
C1I3C1IO
COC12
CII3COOON02
C113CH2COOON02
Toxicityt
These chlorof luorocarbons are
nontoxic but have excellent
properties as tracers of urban
air masses
BM?
BM
Scf.BM
BM
SC.BM
SC.NBM?
-
BM(0)f
SC.BM
SC.BM
Weak BM
SC.NBM
SC.BM
BM
SC.BM
NBM
SC.BM
SC
SC
BM
BM(0)
BM
BM(0)
BM(0)
BM(0)
—
SC
BM(0)
—
—
—
-
-
—
SC.BM
-
-
Phytotoxic
Phytotoxic
In addition to chemical species, meteorological parameters were measured.
These were: wind speed, wind direction, temperature, pressure, relative'
humidity, and solar flux.
tToxicity information obtained from reviews by Helmes et al. (1980);
Albert (1980); U.S. SC (1980). Additional references are contained within
these reviews.
'BM: Positive mutagenic activity based on Ames salmonella mutagenicity
test (Bacterial Mutagens).
NBM: Not found to be mutagens in the Ames salmonella test
(Not Bacterial Matagens).
SC: Suspected Carcinogens.
BM(0): Bacterial Mutagen (Other microbial tests).
-------
FIELD INSTRUMENTATION
One primary motivation of our study was to conduct on-site analysis to
minimize the many problems that are encountered when air samples are collected
in bags, vessels, or in tubes filled with solid sorbents.
It is widely recognized that the integrity of an air sample is best main-
tained when:
• Nominal amounts of air samples are collected on inert surfaces
• Time between collection and analysis is kept to an absolute minimum
• Prior to analysis, trace volatile chemicals are exposed to as low a
temperature as possible.
Our on-site field analysis program was devised to take maximum advantage of
these desirable features. All field work was conducted using a suitably
instrumented mobile environmental laboratory. Table 2 summarizes the equip-
ment that was available on our mobile laboratory for the conduct of this
study. This laboratory was air-conditioned for temperature control and
operated on a 220-V, 80-A circuit. Provision was also devised for operating
on 110-V input. A 200-m electrical cord was always used to station the
laboratory away from the electrical source or a power pole. The sampling man-
ifold was all stainless steel with a variable inlet height. (In all cases,
the sampling manifold was adjusted to be higher than nearby structures: A
typical manifold inlet height was 5 m above ground.) For pumping and pressur-
ing air samples, a special stainless-steel metal bellows compression pump
(Model MB 158) was always used. For the analysis of aldehydes, surface air
was sampled in an all-glass apparatus.
EXPERIMENTAL PROCEDURES
For all the haiogenated species and organic nitrogen compounds shown in
Table 1, electron-capture detector (ECD) gas chromatography (GC) was the pri-
mary means of analysis. The aromatic hydrocarbons were measured using flame-
ionizatibn detector (FID) gas chromatography. Formaldehyde was measured by
the spectrographic chemical analysis technique utilizing the chromotropic acid
procedure (U.S. Public Health Service, 1965). In the third year of this
research, formaldehyde and acetaldehyde were also measured by analyzing the
2,4 dinitrophenylhydrazine derivatives, formed by reaction of 2,4 dinitro-
phenylhydrazlne (DNPH) with aldehydes, with high-performance liquid chromato-
graphic (HPLC) methods (Kuwata et al., 1979; Hull, 1980; Fung and Grosjean,
1981; Salas and Singh, 1981).
For the aldehyde DNPH-HPLC analysis, the sampling reagent was prepared by
dissolving 0.25 g of purified DNPH in 1.0 liter of HPLC-grade acetonitrile and
adding 0.2 ml of concentrated sulfuric acid to this solution. DNPH was puri-
fied by repeated recrystallization (at least three times) from HPLC-grade
acetonitrile. A 7-ml aliquot of this reagent solution was transferred into a
bubble and cooled with the help of an icewater Dewar flask. An air flow rate
of 0.5 1/min was maintained for a typical sampling period of 2 hours. After
-------
TABLE 2. ENVIRONMENTAL MOBILE LABORATORY INSTRUMENTATION*
Instrument
Features
Analysis
Perkin Elmer 3920 CC1
Perkin Elmer 3920 CC2
Perkin Elmer 3920 CC3
(capillary column CC)
Coulometric dual EC-Gc
Spectraphysics HPLC 8700
(variable wavelength
SP3400 detector)
Beckman 6800
Horiba AIA-24
Bendix 8101-B
Monitor Labs Model 8440E
Dasibi Model 1003 AH
AID Model 560
Bendix 8002
Eppley pyranometer
Eppley UV radiometer
Miscellaneous meteorological
equipment
Auto Lab IV Data System (No. 1)
SP-4000 Multichannel Data System
(No. 2)
HP-3390 printer plotter
Digitera Data System (Ho. 3)
2 ECDt, 1 dual FID?
2 ECD, 1 dual FID
2 ECD, 1 dual FID
Coulometric ECD
HPLC8
FID
NDIR**
Chemiluminescent
Chemiluminescent
Photometric principle
Chemiluminescent
Chemiluminescent
Trace constituents
Trace constituents
Trace constituents
Halocarbons, PAN, PPN,
COC12; calibration
Aldehydes
CO-CHA-THC
CO, C02
1JO, N02
NO and 1102
°3
°3
03
Solar flux
Ultraviolet radiative flux
Wind speed, wind direction,
temp, pressure, dew point,
relative humidity
CC data
CC data
All cortinuous air quality
and meteorological data
Note: Sampling of all trace organics is performed from a stainless-steel
manifold. A Teflon® manifold is used for inorganics (e.g., 03, NO,
Finnitjan 3200 CC/MS available to this project at SRI.
'Electron capture detector.
•Flame ionization detector.
High performance liquid chromatograph.
Nondispersive infrared.
-------
sampling, a 2.0 ml aliquot of the exposed reagent solution was transferred
into a heavy-walled reaction flask with Teflon® cap, warmed at 75°C for 20
minutes, and subsequently cooled to room temperature. The DNP hydrazone
derivatives were analyzed with a Spectra Physics HPLC (Model 8700) equipped
with a variable wavelength detector (Model 3400) set at 360-nm wavelength.
The HPLC was used in an isocratic mode with a solvent flow rate of 1.5 ml/min.
A 36 percent H^O; 64 percent acetonitrile solvent gave the most desired reso-
lution of the two hydrazones of Interest. A 10 microliter sampling loop was
used for HPLC analysis. Typical analysis time was less than 10 minutes.
Under normal operating conditions, five GC channels were operated with
ECDs and only one with FID. Although the exquisite sensitivity of the ECD
would allow the determination of several species in Table 1 with a direct 5-ml
injection of air, preconcentration of air samples was necessary for efficient
operation. All six GC channels were equipped with stainless-steel sampling
valves and could be operated either with a direct sampling loop or with a
preconcentration trap. In no instance was a sample size of greater than 1
liter used: In most cases, sample volumes of 500 ml or less were satisfac-
tory. Sample preconcentration was conducted on a 4-inch-long bed of 100/120
mesh glass beads packed in 1/16-inch diameter stainless-steel tubing main-
tained at liquid oxygen temperature. The glass beads could be replaced with
an equivalent length of SE-30 packing (3 percent SE-30 on 100/120 mesh acid-
washed chromosorb W) or glass wool with completely satisfactory results.
Desorption of chemicals from the preconcentration traps was accomplished by
holding the trap at boiling-water temperature and purging with carrier gas.
Additional details have also been earlier provided by Singh et al. (1979a,b;
1980). • . ' .- '
Because the use of liquid oxygen is tedious at best, we attempted to
preconcentrate air samples on Tenax® traps at room temperature. Considerable
testing indicated that Tenax® suffers from serious artifact problems. A
number of "ghost peaks" were seen, particularly on our ECD systems. In addi-
tion, we encountered serious difficulties in quantitatively absorbing and
desorbing specific species that were tested. Because of the possibility of
confusing artifacts (sometimes present in significant amounts) with real pol-
lutants, we have discontinued the use of Tenax® as a column pretrap. It
appears that oxygen or ozone can.oxidize Tenax® monomer to produce electron-
absorbing oxygenated species; therefore, all preconcentrations in this study
have been performed on glass beads, glass wool, or SE-30 packing surfaces.
These artifact problems have also been observed by other investigators
(Sievers, 1981).
The sampling for GC analysis was achieved by pressurizing a 1-liter
SUMMA® polished stainless-steel canister to 32 psi. The sampling line and the
pretrap (maintained at 90°C) were flushed with ambient air and the canister
pressure brought to 30 psi. Sampling then began. The preconcentration trap
was immersed in liquid oxygen and an air volume sampled from pressure PI to
P2» A high-precision pressure gauge (±0.05 psi) was used to measure the can-
ister pressure. A typical setting was pj = 30.0 psi and p£ = 24.0 psi. Ideal
gas laws were found to hold excellently at these pressures and were used to
estimate sample volumes. The pressure range of 30 to 20 psi assured smooth
flow through the preconcentration traps without problems of plugging. All
-------
other sampling was accomplished by using sampling loops that were flushed with
all-glass syringes of 100-ml volume. A 10-(jLl direct syringe injection of the
sampled DNPH solution was injected into the HPLC for aldehyde analysis.
Table 3 summarizes methods used for the analysis of trace species. The
GC and the HPLC conditions used are also stated. Because of the dominant
water response of the ECD, a post-column Ascarite trap was inserted to remove
water from halocarbon analysis. No water trap was used for the analysis of
aromatic hydrocarbons, PAN, PPN, and phosgene. The latter three did not
require any preconcentration step and were measured with a direct 5-ml air
injection.
The identity of trace constituents was established by using the following
criteria:
• Retention times on multiple GC columns (minimum of two columns)
• EC thermal response
• EC ionization efficiency
Details of these comparisons for halocarbon species, organic nitrogen com-
pounds, and aromatic hydrocarbons have already been published (Singh et al.,
1979a,c; 1980). Figures 1-8 provide representative chromatograms of the
atmospheric analysis of selected trace chemicals.
CALIBRATIONS
Primary Standards
Calibrations for all species were performed using three basic methods:
• Permeation tubes
• Multiple dilutions
• Gas-phase coulometry.
As reported earlier (Singh et al., 1977b; 1980), permeation tubes provide a
reliable means to generate low-ppb primary standards for a significant number
of chemicals listed, in Table 1. Permeation tubes (8- to 10-cm long) for many
trace constituents of interest, constructed from standard FEP or TFE Teflon®
tubing of varying thicknesses, were obtained commercially. Each permeation
tube was contained in a specialized glass holder (Figure 9). Based on our
previous experience, we concluded that some permeation tubes could operate
satisfactorily only at high temperatures. Therefore, two temperature baths
maintained at 30.0°C ±0.05°C and 70.0°C ±0.1°C were installed. The 30°C bath
was a water bath, and the 70°C bath was an oil bath. All permeation tubes
were contained in specialized holders and were purged continuously with a
prepurified gas (helium, air, or nitrogen) flowing at 50 to 80 ml/min. Per-
meation tubes were weighed roughly once a week on a semimicro (10~^g) balance.
These weighings were done before, during, and after the field experiments.
The constancy of permeation rate over a period of many months could be
8
-------
TABLE 3. ANALYTICAL CONDITIONS FOR THE ANALYSIS OF SELECTED TOXIC CHEMICALS
;;o.
i
^
3
U
5
6
7
8
1
CC or HPLC Coluon
Description
6 ft x 1/8 in, SS,* 21);; SP2100,
O.i:: DC 1500 on 100/120 nesh
Suptlcoport •
33 ft x 1/8 in, :;i , 20:: DC 200 -
on 8CV 100 nesh Supelcoport
fi ft x 1/8 in, SS, I0» ", ", -bis
( 2-cyanoeth/l ) Fomanitie on
Chromosorb P (A/W)
.3 ft x 1/8 in. Hi, 5" SP 1200
-5:; Bentunu on 100/120
nesh Supelcoport
15 ft x 1/8 in, SS, 10.-: SP 1000
on K'JlVl-M nesn Supelcoport
Hi ft x 1/8 in, SS, 0.2S CK 1500
80/100 nesli on oirbopacK C
10 in x 1/4 in, Teflon, y. CS
KA/u)
5 It l-'i in, JS, 30', didecyl
piitlia ate, 100/120 nesh,
Ui.-ori sorb P (A/W) '
1 ft 1,'.. in SS, sphersoru .
OliS-l'
Imp.
45.
45
65
65
45
30
30
.30
Species Measured
CHC13; CH3CCl3; CC14; cis-QI
C1CI1C1, C2ncl3; CI12C1CHC12;
CHiBrCH2Br; C2C14; ClljCHClj;
CII2ClCCl3[ CHC12CHC121
CI12C1C1IC1CI13
ClljCl; CI^Br; CH2CCl2i Cl^l ;
CCljF; CC12F2; CC1F2CC12F[
CC1F2CC1F2" .
C6116J ^^113013; n/p/o-C^H^ (^13)2;
4— CtllAC^HtL' 135 Ct.ll'ifCHi ) i •
1,2,4 C6ll3( 013)3
C6II5C1; n-C(,H/iCl2; o-CjHiCli;
1,2,4 C6ll3Cl3; C6lljCII?Cl;
CC12CC1CC1CC12
CU2C121 CC13F; cIS-QiClCHCl;
013!; CC12FCC1F2; QI3CC13;
CC14; C2lljCl; CiliCMCI^Cl
PAN', ??;;
COCl-j
~
iicno, CH3ciin
Detector
' Type
Elect ron
capture
Electron
captur.e
Flaoe
ionizatlon
I'.lect ron
capture
Ulect ron
capture
Elect ron
capture '
Electron
capture
Electron
capture
Variable wave-
length detector
sut at 360 nm
Temp.
275
275
275
275
265
265
30
30
30
Typical
Ca rr ler
Flow Kate
(inl/nln)
40
25 •
45
45
25
40
60
70
1 .5
Typical
Saople
Size
(ml)
500
500
500
750
100
10
5
5
0.01
Remarks
No water trap
No water trap
x
No water trap
also used for Cl^CC^
measurement with
injection
No water trap; direct
Inject ion
Isocratic node, 362 I^O,
6AX acetonltrile SP node
8700. SP 3400 detector
-------
c2ci4.
1 i i r
CHCI3
CH3CCI3-
CCI4
C2HCI3—
CH2BrCH2Br
CH2CICHCI2
J I
I I I
Note: Menlo Park air; Attenuation 16; 30O-ml sample; Column No. 1.
Figure 1. Chromatogram showing ambient analysis of selected chlorinated
and brominated toxic chemicals.
-------
CCI2FCCIF
CCI3F
(F11)
CH2CI.
CCI2F2
(F12)
Note: Chicago air; Attenuation 8; 300-ml sample; Column No. 6.
Figure 2. Chromatogram showing separation of methyl halides
and chlorofluorocarbons from ambient air.
11
-------
m/p-C6H4(CH3)2
C6H5C2HS
o-C6H4(CH3)2
4-C6H4CH3C2H5
1, 3. 5. C6H3(CH3)3
1, 2, 4, C6H3(CH3)3 V
I
Note: Menlo Park air; Attenuation 8; 400-ml sample; Column No. 3.
Figure 3. Analysis of aromatic hydrocarbons in ambient air.
12
-------
UNKNOWN
C6H5CI
. UNKNOWN
nvCgH
Note: Menlo Park air: Attenuation 2; 800-ml sample; Column No. 4.
Figure 4. Halogenated aromatics in the air.
13
-------
i r
CH2CICH2CI
i r
i i i i
Note: Menlo Park air; Attenuation 4; 50-ml sample; Column No. 5.
Figure 5. Analysis of 1,2 dichloroethane.
14
-------
CD
Note: Chicago air; Attenuation 2; 300-ml sample; Column No. 2.
Figure 6. Chromatogram showing methylene chloride separations from other halocarbons.
15
-------
10
TIME — minutes
Figure 7. PAN and PPN separation from the air.
16
-------
HCHO
Figure 8. Separation of dinitrophenyl hydrazones
of formaldehyde and acetaldehyde on HPLC.
9-1/4"
MIXTURE PURIFIED
OUT HELIUM IN
MALE BALL JOINTS
GLASS SIEVE PLATE
(Permeation tube stands
vertically on this plate)
Figure 9. Permeation tube holder.
17
-------
established. A large-volume mixing chamber was installed at the permeation
tube exit to allow for complete mixing. Syringe samples were withdrawn from
the mixing chamber using all-glass syringes.
With the installation of the 70°C bath, all permeation tubes performed
excellently. Table 4 reports the measured permeation-rate data for each of
the chemical constituents of interest. It is clear from Table 4 that many
species for which permeation tubes could not be used earlier (Singh et al.,
1979c) are now giving excellent results. Figures 10 and 11 demonstrate the
excellent, linearity of the permeation rate for some of these chemicals. It is
also clear from Table 4 and Figure H that the formaldehyde permeation tube
could be further improved. Overall, we believe that this method is a reliable
means to generate primary standards.
It is also clear from Table 4 that most of these permeation tubes can be
used to prepared standards directly at parts per billion (ppb) concentration
levels. Batch dilutions were carried out to reduce these concentrations by a
factor of 10^ to 10-*. These were performed by injecting a known volume (typi-
cally 10.0 ml) of the high concentration mixture into an evacuated precleaned
electropolished stainless steel container of 1- to 5-liter size, followed by
pressurization with diluent gas to 40 psi. Over a wide range of concentration
levels of low ppb's and low ppt's (parts per trillion), the frequency-
modulated ECDs that we used were linear (Singh, et al., 1977b). The linearity
of the FID over a much larger concentration range is well known.
In addition to permeation tubes, standards were obtained from Scott-
Marrin (Riverside, California). These were obtained at higher concentrations
(5 to 10 ppm) for long-term stability. Table 5 lists the chemicals, the stan-
dard concentrations, and the cylinder materials. All of the chemicals were
stored in aluminum cylinders except those containing Ct^Cl, which were con-
tained in stainless-steel cylinders. Extreme care was required in selecting
cylinder'materials; some of the chemicals (e.g., methyl chloride) form unknown
chemical complexes that might react explosively with aluminum (Private
Communication—Scott-Marrin Inc.).
All of the commercially obtained standards were rechecked with our
permeation-tube standards when this was possible.. The comparisons were found
to yield excellent results (±10 percent). The aromatic hydrocarbon standards
were checked for carbon response against those available from the National
Bureau of Standards (NBS) and found to agree within ±5 percent. For other
aromatic hydrocarbons, carbon response derived from benzene and toluene
responses was used.
For the chlorinated aromatics, the Scott-Marrin standards were found to
deteriorate over a period of several months. In the case of PAN and PPN, only
gas-phase coulometry was used, and the data must be considered preliminary
until the confirmation of the reliability of PAN and PPN determination using
gas-phase coulometry can be established.
18
-------
TABLE 4. PERMEATION RATE DATA FOR GENERATING PRIMARY STANDARDS
Compound
HCHO
CH2-CHCHO
CH2OCH2
CH3CHO
CH3COCH3
C2H5CHO
C6H5CHO
CC12F2 (F12)
CC13F (Fll)
CHC12F (F21)
CHC1F2 (F22)
CC12FCC1F2 (Fll 3)
CC1F2CC1F2 (F114)
CH3C1
C2H5C1
CH2CHC1
C1CH2CH=CH2
CH3Br
CH3I
CH2CI2
(els) CHC1CHC1
(trans) CHC1CHC1
CC12CH2
C»2C1CH2C1
CH2C1CH2C1
CHC12CH3
CH2C1CHC1CI13
(trans) CHC1=CHCH2C1
COC12
CHC13
C2HC13
CC13CH3
CC13CH3
aici2ai2ci
CC14
C2Clit
C2C14
CH2BrClI2Br
CllBr3
C6H5C1
C6H5CH2C1 .
0-C61IAC12
ra-C6HACl2
P-C6H4C12
Permeation
Tube Number
or I.D.
MET8
2356
1908
MET9
MET10
MET 11
MET12
6138
1911
2347
2348
1238
2345
2355
2350
2352
7497
1893
1239
2354
1939
1898
1897
1907
1899
2353
MET1
MET2
2351.
1229
1253
1896
1589
1901
1894
1902
1590
1237
1895
.MET3
MET4
MET5
MET6
MET7
Temperature
(°c*)
70.0
' 30.0
30.0
„ 30.0
30.0
30.0
30.0
30.0
30.0
30.0
30.0
30.0
30.0 .
30.0
30.0
30.0
30.0
30.0
30.0
30.0
30.0
. 30.0
30.0
70.0
30.0
30.0
70.0
70.0 .
30.0
30.0
30.0
70.0
70.0
30.0
70.0
70.0
. 30.0
70.0
70.0
70.0
70.0
. 70.0
70.0
70.0
Permeation Rate
ng/min
326
969
1120
3300
615
2604
95
615
1680
942
80
715
6254
1915
480
1270
142
2477
109
523
2564
1696
731
2622
125
71
2456
7806
942
174
314
980
3450
129
1983
3352
706
1220
1316
4507
1528
1359
2515
3571
ppb/1/min
(25°C, 1 atm)
266
423
618
1837
260
1100
23
123
299
224
23
93
894
927
182
497
45
. 638
19
150
646
428
184
648
31
18
531
.1720
233
36
58
179
632
24
315
494
104
160
127
.980
295 .
226
418
593
Quality!
F
E
E
E
C
E
F
E
E
E
E
E
E
E
E
E
E
S
G
E
E
E
C
E
G
C
E
E
E
G
E
E
E
C
E
E
E
E
E
E
E
E
E
E
Note: All permeation tubcs'were given a 2 week or longer conditioning time.
Temperature maintained to ±0.05°C.
TE=Excellent (errors in permeation rate < ±10%); C=Cood (errors in perr.iention rate
< ±15%); F=Fair (errors in permeation rate < ±25%).
19
-------
0.06
0.20
0.18
0.16
o 0.14
0.12
0.10
O 0.08
UJ
| 0.06
ui
Q.
0.04
0.02
0.00
12.8)
= 7.2)
\\N-
CCI4 (w0 = 7.8)
0 2 4 6 8 10 12
TIME — hundreds of hours
Figure 10. Permeation tube weight-time relationship
for selected halogenated chemicals.
ACETALDEHYDE (WQ = 10.0)
6 8 10
TIME — hundreds of hours
12
0.30
16
Figure 11. Permeation tube weight loss of selected oxygenated chemicals.
20
-------
TABLE 5. PPM LEVEL PRIMARY STANDARDS IN AIR*
Standard and Compoundf
SI
1,1,1 Trlchloroethane
Carbon tetrachloride
1,2 Dibromoethane
Hexachloroe thane
S2
Monochlorobenzene
o-Dichlorobenzene '
S3
Benzene
Toluene
S4
Methyl chloride
Methylene chloride
1,2 Dichloroethane
S5
Trichloroethylene
Tetrachloroethylene
Chloroform
S6 '
Ethane
Propane
.n-Butane
S7
Methyl chloride
Methyl bromide
Methyl Iodide
Concentration
(ppm)
5.0,
5.2
5.0
0.8
5.0
5.0
5.0
5.0
10.0
10.0. .
10.0
10.0
10.0
10.0
4.07
5.03
4.95
10.0
10.0
5.0
Long-term
Stability^
(2-year period)
E
. E
E
U
P
P
E
E
E
' E
E
E
E
E
E
E
E
E/S
E/S
E/S
Cylinder
Type
Aluminum
Aluminum
Aluminum
Stainless
steel
Aluminum
Alumi num
Stainless
s tecl
Size
(ft3)
30
150
150
30
30
150
30
*0btained on order from Scott-Marrin, Inc., Riverside, California.
'For all of these chemicals (except C^Hg and C^IljCH-j) satisfactory permeation
tubes were also operational. Therefore, a majority of these standards
were used more as secondary standards than as primary ones. For aromatic
hydrocarbons, the Scott-Marrin standards were used as primary standards. •
~E: Excellent; P: poor; U: unknown; E/S: Excellent over the short term;
long-term tests have not been made.
21
-------
Secondary Standards
Except for the aromatic hydrocarbons, it was not possible to use primary
standards during field operation. Therefore, an optimal scheme that depended
on the use of secondary standards was devised.
A 35-liter and several 5-liter (as backups) SUMMA® polished stainless-
steel samplers were filled with urban air samples to a pressure of 35 to 40
psi. These were allowed to stabilize for one to two days and then analyzed by
comparing them against the primary standards. The 35-liter pressurized secon-
dary standard was then used for field operation: Each GC channel was cali-
brated about three times a day with'this secondary standard. The stability of
nearly all species over a period of several days was found to be excellent.
Figure 12 shows an example of the stability of a secondary standard during the
course of a field experiment. Some species, such as PAN, PPN, or COC12, could
not be stored for any reasonable length of time. This was not a serious hin-
drance since other chemicals could be used to ascertain the constancy of the
ECD and the FID responses during field operations. All of the Scott-Marrin
standards were also carried on board after these had been diluted to low-ppb
levels. These were also used as secondary standards (in addition to the col-
lected air samples). The stability of the diluted Scott-Marrin cylinders (in
polished 1- to 5-liter stainless-steel vessels) was found to be excellent.
Analysis of these before field experimentation, during field studies, and
after the completion of field studies did not show a change from the measure-
ment precision under field conditions.
In the case of aldehydes, all calibrations had to be performed in the
field. The repeatability of the acetaldehyde-hydrozone calibration with HPLC
is clearly shown in Figure 13.
QUALITY CONTROL
Two major factors were critical in establishing the quality of the
acquired data: the accuracy of primary standards and the precision and repea-
tability of measurements. As stated earlier in this section, the primary
standards commercially obtained were compared with our permeation tubes, which
can be routinely used to obtain reliable standards within errors of ±5 to 10
percent. The aromatic hydrocarbon standards were compared with NBS propane
standards and found to be accurate to within ±5 percent. The cross-
calibrations between SRI-generated standards and Scott-Marrin standards typi-
cally resulted in differences of about ±10 percent or less. The use of secon-
dary standards nearly three time's a day clearly demonstrated the excellent
precision that was obtainable during field studies: The precision of reported
field measurements is estimated to be better than ±15 percent. In order'to
assess the overall accuracy of field measurements further, we conducted two
programs of interlaboratory comparisons. The results from one of these pro-
grams conducted in early 1981 are shown in Table 6. The measurement errors
were found to be less than ±10 percent. A similar program was also conducted
under the auspices of the U.S. Environmental Protection Agency (EPA). Four
laboratories participated in this program. The results of this intercom-
parison were inconclusive because of uncertainties associated with the
22
-------
5000
4000
3000
J? 2000
O
o
U
1000
> 1
50
40
a.
I
05
30
20
10
+' +
00
o o o
o o
4 68
NUMBER OF DAYS
(a) CH3CCI3
10
O
o-
o o
o
q, oo
+• ..•*- +
4 6 8
NUMBER OF DAYS
10
O AMBIENT DATA
•f SECONDARY STANDARD
(b) C6H6
Figure 12. Secondary standard response for 1,1,1 trichloroethane and benzene
in Phoenix — Site 2.
23
-------
1400
1200
E
01
1000
z
g
<
cc
z
LU 800
O
O
UJ
O
N
<
a
600
UJ
O
I
ui 400
LU
200
PITTSBURGH CALIBRATION
CHICAGO CALIBRATION
I
16 APRIL 1981 (O)
29 APRIL 1981 (•)
I
I
I
I
1234
INTEGRATOR COUNTS — hundred thousand
Figure 13. Calibration curve for dinitrophenyl hydrazones of acetaldehyde
and its repeatability.
procedures used to prepare standard mixtures that were sent for analysis to
the four laboratories (Arnts, 1980). For the analysis of aldehydes, the
methods are still in a development stage, and accuracies of ±30 percent can be
expected. (It should be possible to reduce these errors in the near future.)
Further details on the quality control aspects of the research plan have been
presented separately (Singh et al., 1981). It is noted that interlaboratory
comparisons provide one of the best means to test the quality of new data.
These comparisons to date have been performed only on an extremely limited
basis.
-------
TABLE 6. INTERLABORATORY COMPARISON OF SELECTED TRACE CONSTITUENTS
Chemicals
N20
CC12F2
CC13F
CH3CC13
Sample 1
SRI
312 ppb
315 ppt
280 ppt
180 ppt
OGC
336 ppb
330 ppt
278 ppt
183 ppt
Sample 2
SRI
309 ppb
285 ppt
175 ppt
131 ppt
OGC
335 ppb
298 ppt
175 ppt
138 ppt
g of unknown composition ("blind samples") analyzed
at SRI; same samples analyzed by Oregon Graduate Center (OGC).
25
-------
SECTION 4
PLAN OF FIELD MEASUREMENTS
After the measurement methodology was developed, field studies were con-
ducted at ten selected urban sites in the continental United States. In con-
sultation with the project officer, the following cities were chosen:
• Los Angeles, California
• Phoenix, Arizona
• Oakland, California
• Houston, Texas
• St. Louis, Missouri
• Denver, Colorado
• Riverside, California
• Staten .Island, New York
• Pittsburgh, Pennsylvania
• Chicago, Illinois.
Within the above cities, specific sites were chosen that represented an open
urban area. Large point sources or topographical features that could affect
the representativeness of the measurements were avoided. Every attempt was
made to select sites that can be expected to be indicative of general pollu-
tion levels prevalent in the area. Practical constraints such as power and
shelter availability also played a role in the selection of sites. It must be
emphasized that only one site within each of the selected cities was moni-
tored. The data collected here, while perhaps typical of general ambient
environment, are truly representative only of the specific site monitored.
The site locations, and the periods of field measurements, are shown in
Table 7. Each field study was of roughly two weeks' duration. Actual field
data was collected from 9 to 11 days on an around-the-clock basis (Table 7).
Preliminary literature search clearly indicated that available data on hazar-
dous organic chemicals were highly limited and virtually all were obtained
during daytime hours. Based on our past experience (Singh et al., 1979a), we
believed that significant night and daytime differences in the abundance of
organic chemicals were likely. Thus we concluded that despite the logistical
difficulty, a 24-hour measurement schedule offered the most efficient means to
collect the maximum amount of data to characterize the burden of toxic organic
chemicals in the ambient air. In addition, night abundances of trace
26
-------
TABLE 7. FIELD SITES AND MEASUREMENT SCHEDULE
;jo.
Data
1
2
3
4
5
6
7
8
9
10
Field Site
City Jianu
Los Angeles, CA
Phoenix, AZ
Oakland, CA
Houston, TX
St. Louis, MO
Denver, CO
Riverside, CA
Staten Island, NY
Pittsburgh, PA
Chicago, IL
Latitude
. (°U)
34°04'
33°28'
37°45'
29°47"
38°46'
39°45'
3 3° 59'
40°35'
40°26*
41°45' •
Longitude
(°W)
118°09'
112°06'
122°11'
95°15'
90°17'
104°59'
117°18'
7,4°12'
79°56"
87°42'
Experiment Period
9 Apr 79 - 21 Apr 79
23 Apr 79 - 6 May 79
28 Jun 79 - 10 Jul 79
14 May 80 - 25 May 80
29 May 80-9 Jun 80
15 Jun 80 - 28 Jun 80
1 Jul 80 - 13 Jul 80
26 Mar 81 - 5 Apr 81
7 Apr 81 - 17 Apr 81
20 Apr 81-2 May 81
Days of
Actual Data
Collection
10
11
9
10
10
11
11
9
9
9
Site Address
Los Angeles State
University
19th and Adam St.
at state capitol
Hegenberger and 14th
St.
Mae St. and 1-10 Front-
age Road (Florissant
Valley College)
3400 Pershall Rd.
(Florissant Valley
College)
Marion St. and E. 51st
Big Spring Rd. and
Perimeter Koad (U.C.
Riverside campus)
Wild Ave. and Victory
Blvd. (Consolidated
Edison Power Plant)
Carnegie Mellon
Institute (campus)
79th St. and Lawndale
to
-------
chemicals were likely to provide important information about the sources and
sinks of measured species. Therefore, a 24-hour-per-day, seven-days-a-week
measurement schedule was followed during all field programs.
28
-------
SECTION 5 .
ESTIMATED LOSS RATES OF MEASURED CHEMICALS
IN POLLUTED ATMOSPHERES
A knowledge of the atmospheric loss rate of a chemical is essential for
the interpretation of field data and for the eventual prediction of the fate
of an organic chemical. It is well known that hydroxyl radical (HO) plays a
central role in depleting atmospheric organics, both in the polluted and the
.clean atmospheres. For the halogenated (except methyl iodide) and aromatic
hydrocarbons of interest here, we have concluded that no significant error in
loss rates is Incurred when reactions with species other than HO, such as
0(-*P), Og, HOj, are neglected. In the case of the aldehydes and methyl iodide
both reaction with HO and photolysis are important.
The residence times of PAN and PPN are largely controlled by their ther-
mal decomposition and are estimated from mechanisms suggested by Hendry and
Kenley (1979). For the purposes of these calculations an average daytime HO
radical abundance of 2 x 10° molecule/cm^ is assumed. [These HO levels are
well supported by HO estimates from available field data (Calvert, 1976; Singh
et al., 1981a) and are probably typical of summer months within the boundary
layer of polluted urban environments. In winter months the HO levels can be
lower by a factor of about two, but no direct wintertime estimates are avail-
able.] The kinetic and photolytic data utilized in Table 8 are taken from
Atkinson et al. (1979), Hampson (1980), Hudson and Reed (1980), and estimated
from Hendry and Kenley (1979) and Hendry et al. (1980).
Table 8 provides these data and estimates the percentage loss due to
chemical reaction in one day (12 sunlit hours). For virtually all species of
Table 8, nighttime loss rates are negligibly small. This percentage loss is
defined as:
. ' • ' ' 4
percent loss = [1 - exp (-4.32 x 10 K)] x 100 ,
where K = kHO(HO) + khv + kthermal.
It is clear from Table 8 that the daily loss rate of HOCs ranges from
near-zero to 100 percent per day. Chioromethanes
and chloroethanes, collectively a dominant group, are relatively unreactive,
and a daily loss rate of 0 to 3 percent per day is estimated. In the entire
haloalkane group, methyl iodide is the only species that is relatively rapidly
removed by photolytic decomposition (=12 percent/day loss rate). The daily
loss rate of aromatic hydrocarbons, aldehydes, and PAN is quite substantial.
29
-------
TABLE 8. ESTIMATED DAILY LOSS RATES (%) OF SELECTED TRACK CHEMICALS
Chemical Name
Methyl chloride
Methyl bromide
Methyl iodide
Methylene chloride
Chloroform
Carbon tet rachloride
Ethyl chloride
1,1 Dtchloroethane
1,2 Dicliloruethane
1,2 Dibromoetliane
1,1,1 Trichloroethane
1,1,2 Trichloroethane
Tetrachloroethane (isom)
1,2 Dlchloropropane
Vinylidene chloride
(cis) 1,2 Dichloroettiylene
Trichloroethylene
Tetrachloroethylene
Allyl chloride
Hexachl oro- 1,3 bu tadi ene
• Chlorobenzene
a-chlorotoluene
Uichlorobenzene (o,m,p)
Trichlorobenzene (isom)
Benzene
Toluene
Ethyl benzene
m-Xylene
p-Xylene
o-Xylene
4-Ethyl toluene
1,2,4 Trimethyl benzene
1,3,5 Trimethyl benzene
Formaldehyde
Acetaldeliydc
Phosgene
Peroxyacetyl nitrate (PAN)
Pcroxypropionyl nitrate (PPII)
Dominant Removal .
Mechanism
(reaction with)
HO
110
photolysis
HO
HO
Strat. photolysis
HO
110
HO
HO
110
HO
HO
HO
110
HO
ilO
HO
110
HO (?)
HO
110
110
HO
110
HO
HO
HO
110
HO
HO
110
110
HO, photolysis
110, photolysis
-
The ma I
Thermal
Rate Constant
at 300K*
(k x 1012)
0.05
0.04
: 3 x io61=
0.15
0.10
<0.0001
0.39
0.26
0.22
0.25
0.01
0.33
<0.01
1.3s
4.0s
4.0
2.2
0,17
28*
—
0.9s
3.0*
0.3s
0.1s
1.4
6.0
8.0
23.4
12.3
13.9
12.9
33.2
49.2
11.0, 2.8 x 107f
15.0, 1.4 x 10??
-
—
—
Percentage Loss
in One Day
( 12 sunlit hours)t
0.4
0.4
12.2
1.3
0.9
= 0.0
3.3
2.3
1.9
2.2
<0.1
. 2.8
<0.1
10.2
29.2
29.2
17.2
1.5
91.1
' —
7.4
22.8
2.6
0.9
11.4
40.9 .
51.0
86.5
67.0
71.3
67.0
96.4
99.1
88.2
85.1
-
99.9**
99.9**
Rate constant with hydroxyl radical (110) in units of cn'molec~'s~'.
'Calculated baser! on un estimated daytime (12 hour) average 110 abundance of 2 x 10*> noloc/ci;-'.
Rate constant due to photolysis in units of s~'.
^Estimated from SlCiidry and IConliiy (1979).
TliiTii.il degradation in the jirusiMice uf NO ami i.'Oi.
30
-------
In all cases listed in Table 8, the loss rate should be reduced in winter
months because of reduced temperatures and solar flux; In the case of PAN the
effect of temperature is extremely dominant and could very substantially
reduce its loss rate (Hendry and Kenley, 1977; Cox and Roffey, 1977).
31
-------
SECTION 6
ANALYSIS AND INTERPRETATION OF FIELD DATA
The field operations, which entailed onsite measurement studies, were
conducted around-the-clock on a seven-day-per-week basis, allowing the collec-
tion of a large body of ambient data on hazardous organic chemicals. This
body of data supplements and considerably expands the highly limited data pre-
viously available.
The data collected during these studies have been compiled, validated,
and statistically treated. A computer-compatible master data file has been
created. In addition to computer processing of field data, we have also
analyzed data to study the diurnal variations that are typically observed.
Further interpretation of these data was beyond the available resources of
this study because of the lack of daily city-based emissions information for
these chemicals. The data generated in this study, however, when further
analyzed in the context of prevailing meteorology and source inventories, have
the potential to add significantly to our knowledge of urban atmospheric
chemistry.
Table 9 is presented to give a general idea of the yearly U.S. produc-
tion, average emissions, and typical use patterns of important chemicals. A
major source term for each of the chemicals has also been assigned, based on
available information. Table 9 provides a preliminary basis for comparing the
relative abundance of chemicals in the ambient atmospheres.
ATMOSPHERIC CONCENTRATIONS AND VARIABILITIES
OF MEASURED SPECIES
Tables 10, 11, and 12 summarize the ambient data collected at ten sites.
Average concentrations* (arithmetic averages) and standard deviations (one
sigma) in units of parts per trillion (ppt = 10"^ v/v) and ng/rn^ are pro-
vided. This redundancy is often convenient, because exposures are invariably
expressed in mass concentration units. Maximum and minimum concentration lev-
els are also provided. A dash is used to show instances where no data were
obtained or data obtained were such that standard deviations could not be com-
puted. In addition, mean diurnal profiles have been plotted. These are based
*"Concentration" as used here includes "volumetric mixing ratio" (e.g. concen-
tration in parts per trillion).
32
-------
TABLE 9. • PRODUCTION, EMISSION AND USAGES OF SELECTED CHEMICALS
• Compound
Methyl chloride
Methyl bromide
Methyl Iodide
Methylene chloride
Chloroform
Carbon . tetrachlorlde
Ethyl chloride
1.2 Dlchloroethane
1,2 Dlbromoethane
1,1,1 Trlchlorbethane
Vinyl chloride
Vlnylldene chloride
Trlchloroethy lene'
Tetrachloroethylene
Monochlorobenzene
o-Dlchlorobe_nzene
p-Dlchlorobenzene
1,2,6 Trlchlorobenzene
Benzene
Toluene
o-m-p Xylenes
Formaldehyde
Source
A; N(0)
A; N(0)
N(0)
A '
A
A
A
A
. A
A
A
A
A
A"
A
A
A
A
A
A
A
A, N
U.S. Production
(million metric tons)
0.20 (1978)
6.02 (1977)
:o.o
0.30 (1978)
0.16 (1978)
0.34 (1978)
0.3 (1978)
4.8 (1978)
0.1 (1976)
0.3 (1978)
3.2 (1978)
0.2 (1974)
0.14 (1978)
0.33 (1978)
1.5 (1978)
0.04 (1976)
0.03 (1976)
0.01 (1973)
5.0 (1978)
4.2 (1978)
2.9 (1978)
2.6 (1979)
Emissions '
(percent)
5 - 10
:so
80 - 90
5-10.
5 - 10
25
=2
' 5 - 25
' =95
2-5
2-5
>90
>90
>50
>25
>90
No data
2-5
2 - 5 .
2-5
Usage and Remarks
85
of slllcones and tetramethyl lead intermedi-
ates; large natural source (=3 million
tons/yr) Identified In the ocean
Soil fumlgant; oceanic source significantly
8 P
tons/yr IB estimated
removing and solvent degress Ing
85 percent consumed In the manufacture of
tetraethyl lead
87 percent used for vinyl chloride monomer
synthesis; also used for the production of
chlorethylenes and chloroethanes ; about 0. 4
Ing In automobiles
Major gasoline additive for lead scavenging;
also used as a fuoigant
cleaning; most other applications also result
Used for polymer synthesis
Used for polymer synthesis
cleaning
remainder In the production of nitrobenzene,
DDT» dlphenyl oxide
vent applications
90 percent of production used for space deo-
dorizing and moth control
70 percent used as a dye carrier and a herbi-
cide Intermediate
Auto exhaust Is the major emission source;
estimated U.S. total emissions approach 0.5
million metric tons per year
Auto exhaust Is the major emission source;
estimated U.S. total emissions approach 1
Auto exhaust Is the major emission source;
estimated total U.S. emissions approach 0.5
million metric tons per year
Auto exhaust Is a major primary source; sig-
nificant secondary photochemical sources also
exist. A major natural source from methane
oxidation Is 1 tkely •
A"anthropogenic, N-natural, 0«oceanlc.
^Source: Singh et al. (1979a, 1980); U.S. Tariff Commission reports.
33
-------
TARI.K IO. ATMIISPHER
U)
Helonethinea
Methyl chloride
Hethyl bromide
Methyl Iodide
Methylene chloride
Chloroforn
Haloethanee end halopropanea
Ethyl chloride
1.1 Dlchloroethane
1,2 Dlchloroethane
1,2 Dtbromoethnne
1,1.1 Trlchloroethene
1,2 Dlchloropropane
Chloroalkenes
Vlnylldene chloride
(da) 1.2 Dlchloroechylene
Trlchloroethylene
Allyl chloride
Hexachloro-l , 3 butadiene
Chloroaromatlca
Monochlorobenzene
o-Chlorotoluene
D'Dlchlorobenzene
p-DlchlorobenEene
Benzene
Toluene .
Ethyl benzene
m/p-Xylene
0-Xylene
4-Elhyl toluene
1,3,% Trlmethyl benzene
Oxygenated species
Formaldehyde
Acetaldehyde .
Phoagene
Peroxyacetylnltrete (PAH)
Peroxyproplonylnttr«te (PPN)
Loa Angelea — Site 1
(9-21 April 1979)
47)
-?
105
3001
244
3751
It
215
519
33
1029
9
4
17
5
399
1480
3
.200
13
6
7
6040
11720
2250
4610
1930
1510
1880
380
722
197
667
1759
174
2620
40
107
233
26
646
6
2
11
3
302
446
2
10
6
5
4580
9070
4470
6140
1830
1450
2380
680
673
ppt
1070
4160
7761
894
12029
995
1353
187
3144
45
12
96
10
1702
2065
8
-500
50
25
34
27870
53380
271,60
49960
12740
10150
13290
5020
2740
221
49
1018
11
601
97
171
5
224
4
-------
T*J!ir II. ATKltSPKEHIC (
(HEMIIIAI.S (SITF 4-7J
fhr.tr. 1 Croup (nd Sprel**
TrlrhlnroUui>rra»tK*n> (Til)
DlchloroMuori«th*n» (Til)
Trlrhtorof ririunrorthBn* mill
N.l«»th*n*i
Mtthjl rhlortd*
W*ChyI lodld*
0.1 orator.
Hi|o»lti«nri «nd hilnpropm**
I.I r>lchl«r(*rh*n*
1,1 OlrMnrnrthin*
I.I mbrrw»ih«™
I.I, 1 Tr1ch|.«rntih*n*
I.I Plrhlaroprnpifl*
Allyl chlnrld*
riBi>orhlorob*n(*n*
n-tMrhlarnbrni'n*
•-nichloratirnirn*
1,:,* Trtehlorohtnitnc
Arautlc hydrocarbon*
Rrhrl b»ni*n*
4-e«hrl int..,..
1,1.1 Trlnrthrl tw«i*n*
r™ld,h,d.
Act(ild*hyd*
r>r<"7j>ro»tonf Inlir.t* (PPN)
M...,,,'
474
199
SS
1.6
74
11
117
mi
jj
ii
ii
is
71
«s
;
7
7
1(7 HO
D'O
1»40
1)17
870
4JI
"°
P
S.D.'
us
190
40)
I.I
1S1
749
IM
i8 )
i
•
9
- )'
16
S9
1"
-»
10
„,
,
f
',
""
100
•>0
**0
i'O
.-
-
,»
...
Routton--
(i4-i> ruT
Pi
H..IM.
lint
166*
I:B»
it. i
JlOi
SMI
7100
)*B
II«
111
l)t>
t:o
114
I!
13
17700
6^S10
1)70(1
9740
7170
9IbO
-
4 ISO
6)0
It* 4
I9BO)
HI.|M
lOi
17
til
0.6
44
i»
'"
50
in
.1
j
U
;i
1
040
1040
HO
s
.."..
-
•
.10
. -1°
"I/
tto.n
lb-,9
1)21
1*61
II
1991
10 i\
6110
410
174
7S
)7«
»1
S.D.
»•
I4S1
JO
1)1
)
19 9
H 1
71 0
61
171
U)
:i)
14
IS
H'ln
eii7
S04S
>?nn
- •
4114
in
htm
)T4
I)J
I)
11?
I.i
411
J)
60
17
I)
s
«i
)
6
I
1 10
M>
40
;o
UNO
Jit
<»
p
s...
10}
171
6
MS
l.t
11)
)0
14
101
1)6
'5
I
II
1
:™
70)
KM)
IHO
170
4SOO
10)
"
I. LM.I.-
H.T - 9 /
pt
tollBUB
90S
1791
31
lOli
.1
640
19
10
60
:
\i
61
10
;:
ino
?in
**n
140
1*0
IS 700
'
m
ISO
Slta S
unf 1980)
mm~.
117
II
1)
SJI
O.I
•I
IS
10
J4
41
I
117
6
It
;
i
no
in
MO
60
no
60
1100
to
««/
*„
1097
174
ISO9
I.M
"!
lit
14}
1ft 1
III
11*1
«:
•i;
)6
14
;
4A9
110
141
176
BII
I1B16
1)61
.>
....
IK*
Mil
14*
S7
<0»
31
741
))
10
II
S
7
:::
0*1
I9H
MI
*II
illO
1* ;
inn)
S,I
«..„
617
76)
«6I
!"
6)
141
M)
17
)l
"
6
4)40
?|*0
linn
900
1410
11100
44)
p
S.D."
IIS
1)1
„;••
106
Jl
197
IS)
10
)ll
)no
))IO
IIIO
IbO
DID
S900
1!;;
Ora»r~S
(IS-IB Jun
Pt
N..1M.
4114
16)6
MI
ION9
1699
S6
114
14)1)
7
1114
16
)S
11910
;:,'"
fiino
41KO
IUSO
If 700
-;::
It* 6
1980)
Hlnlmim
17
108
19
M
f4
IM
J
<4
7
<)
0.4
1)
110
MO
**/
N**n
11S5
•99
974
18IS
106)
II
1)31
lit I*
1 174
111
401
906
IW6I
II8B
1
I.D.
)II1
1001
1100
3013
II
1610
996
30
IIS44
1)S
>!I
1 )f4
7114
61 S4
M*«n
1949
10)
))7
\ 11
'»'
60
III
-------
TABLE 12. ATNOSHIF.RIC CONCENTRATIONS I1F MF.ASl'ltF.n CHEMICALS (SITE 0-10)
Chtorof luorocarbons
Ha ora* thanes
Methyl chloride
Methyl hrotnlde
Methyl Iodide
Chloroform
Carbon tetrachlorlde
thyl chloride
,1 Olchloroetha e
.2 Dlchloroetha e
,2 Dtbronoethan
, 1, 1 Trlchloroe hane
,1,2 Trichloroe hane
Chloroalkitnes
Vlnylldune chloride
(els) 1,2 Dichloroethylene
Trlchloroe thy ten*
ChloroaronutlcB
a-Chlorototuene
o*Dlchlorobenzene
D-Dlchlorobenzene
p-Dlchlorobenzene
Aromatic hydrocarbon*
Benzene
Toluene
Ethyl benzene
/p-Xylene
-Xylene .
-Ethyl toluene
,2.4 Trlraethyt benzene
.3.5 Trlwethyl benzene
Formaldehyde
Acetaldehyde
Phosgene
Staten Island— SI te 8 (26 March-5 April 1981)
360
519
129
39
701
84
1609
309
110
13
256
20
468
26
18
167
292
4204
8975
1742
4088
1288
411
831
210
14300
7*7
204
143
190
78
33
106
10A
1
2947
202
64
5
520
6
24B
2
15
6
199
200
4287
10638
2472
8352
468
911
27)
9100
7IS
527
ppr
909
1028
359
204
1208
671
*
18176
1200
312
)7
4312
36
1427
11
79
1005
10)4
_
19034
67304
17230
54f>3B
277«
4682
1621
45900
3888
3110
.
175
318
59
21
446
27
226
125
10
3
55
12
221
3
10
",:
79
-
82
623
9
170
13
62
51
7000
65
32
«B/
2019
2563
987
272
1445
326
12
5S46
1942
290
53
1034
153
2550
38
120
71
896
1978
_
13384
3)707
7537
17687
201)
4070
1029
17510
3689
1124
.'
80!
9)8
597
2)0
383
419
6
10224
1270
168
20
46
1)51
1 1
69
24
1068
1)55
-
13648
39947
10695
36135
2292
4451
l))7
1114)
)546
2904
Pittsburgh — Site 9 (7-17 April 1981)
33)
496
68
)0
665
41
390
97
331
84
12
16
486
6
4
23
"|J
96
409
45
178
27
5
105
6
1
244
41
107
45
15
,0
272
2
1
<>
5
9)
)57
7
I
i 1 _
5003
3928
765
,55,
)09
,034
,21
20600
1400
266
45
~
9818
7286
1564
2357
828
416
3349
128
5200
600
121
8
ppt
486
976
,62
43
852
62
3
,308
238
691
229
,05
59
,595
II
5
4
50
25
420
1657
19
-
64619
46313
10465
1078)
3787
2881
24772
797
35100
2600
648
65
279
306
42
22
450
27
0
152
31
131
42
3
6
158
4
)
'
4
13
80
-------
on two-hour averages and one sigma (cr) standard deviation. Where raw data are
plotted, the day of the week is also shown.
Because many of the chemicals measured here are also ubiquitous com-
ponents of the global troposphere, Table 13 has been prepared to define this
background (Singh et al., 1981c; Singh and Hanst, 1981). To facilitate dis-
cussion, chemicals in Tables 10-12 have been divided into seven categories.
Much of the information presented in Tables 10-12 is self-explanatory, so only
the salient features will be discussed in the next section. Some additional
details can also be found in interim reports (Singh et al., 1979c; 1980).
INTERPRETATION OF FIELD DATA
BY CHEMICAL CATEGORY
Chlorofluorocarbons (CFCs)
As indicated earlier, CFCs are not considered or expected to be toxic.
They can, however, act as useful indicators of'polluted air masses, and their
possible involvement in stratospheric ozone destruction (Molina and Rowland,
1974) is well known. A maximum of four CFCs (fluorocarbons 11, 12, 113, and
114) were measured. Fluorocarbons 12 and 11 are the dominant CFCs. The aver-
age F12 and Fll concentrations were typically in the 0.5 to 1 ppb and 0.3 to
0.7 ppb range respectively (Tables 10 through 12). Maximum concentrations of
3.2 ppb for F12 and 1.2 ppb for Fll were measured. Average concentrations of
F113 and F114 were significantly lower, although a high F113 level of 4.2 ppb
was measured in Los Angeles (Site 1). Typical average concentrations at all
sites were 2 to 10 times the geochemical background concentrations (Table 13).
The mean F12/F11 concentration ratio at all sites (Sites 4-10) was
between 1.5 and 1.9. The highest value of 1.9, measured at Houston (Site 4),
may be due to a greater use of air conditioned automobiles, which use F12 as a
refrigerant. The F12 and Fll emission ratio for the United States is not
available for 1981, but is determined to be approximately between 1.5 to 1.7
for the previous two years (CMA, 1981; CIS, 1981). In the clean troposphere
and at midlatitudes, an F12/F11 concentration ratio of about 1.6 has been
measured (Table 13). It is clear, therefore, that the F12/F11 ratio measured
in the urban environment is consistent with the expected values. A much
greater variability in the F12/F113 ratio (3 to 9) is observed probably indi-
cative of the difference in use patterns of these fluorocarbons.
Although a considerable data base on the background concentrations of
these fluorocarbons is available, relatively little urban data have been pub-
lished. Typical F12 and FIT levels from several cities (Simmonds et al.,
1973; Lillian et al., 1975; Singh et al., 1979a) are not inconsistent with
measurements in this study. Because of the rapidly changing use patterns of
fluorocarbons in recent years (CMA, 1981) urban measurements obtained at dif-
ferent times cannot be quantitatively compared.
37
-------
TABLE 13. AVERAGE BACKGROUND CONCENTRATION
OF TRACE SPECIES AT 40°N FOR YEAR 1981
Chemical Group and Species
Chlorofluorocarbons
Trlchlorof luoromethane (Fll)
Dlchlorof luoromethane (F12)
Trlchlorotrlfluoroethane (F113)
Dlchlorotetrafluoroethane (F114)
Halomethanes
Methyl chloride
Methyl bromide
Methyl iodide
Methylene chloride
Chloroform
Carbon tetrachlorlde
Haloethanes and halopropanes
Ethyl chloride
1,1 Dichloroethane
1,2 Dichloroethane
1,2 Dlbromoethane
,1,1 Trichloroethane
,1,2 Trichloroethane
,1,1,2 Tetrachloroethane
,1,2,2 -Tetrachloroethane
,2 Dlchloropropane
Chloroalkenes
Vinylidene chloride
(els') 1,2 Dichloroethylene
Trichloroethylene
Tetrachloroethylene
Allyl chloride
Hexachloro-1, 3 butadiene
Chi oroaroma tics
Monochlorobenzene
a-Chlorotoluene
o-Dichlorobenzene
m-Dichlorobenzene
p-Dichlorobenzene
1,2,4 Trichlorobenzene
Aromatic hydrocarbons
Benzene
Toluene
Ethyl benzene
m/p-Xylene
o-Xylene
4-Ethyl toluene
1,2,4 Trinethyl benzene
1,3,5 Trimethyl benzene
Oxygenated species
Formaldehyde
Acetaldeliyde
Phosgene
Peroxyacetylnitrate (PAN)
Peroxyprop tonylnltr.ite (PPH)
Concentration
ppt
190
300
25
15
650
10
2
50
20
135
10
_*
40
2
180
-
-
-
—
-
-
15
50
-
—
-
—
-
-
—
—
-
-
-
-
-
-
—
400
40^
i
2ot
—
ng/m3
1066
1481
191
105
1340
39
12
173
97
848
26
-
162
15
981
-
-
-
—
-
-
80
337
—
—
-
—
-
—
-
—
—
-
-
-
-
-
-
—
490
72t
—
99t
*"
*Dashes indicate absence of available data.
'Estimated from mechanistic models.
38
-------
Halomethanes
Six halomethanes—methyl chloride, methyl bromide, methyl iodide,
methylene chloride, chloroform, and carbon tetrachloride—were measured. As
can be seen from Table 1, all six of these are bacterial mutagens or suspected
carcinogens. Three methyl halides have dominant natural (oceanic) sources
(Table 9).
Methyl Chloride—The dominant natural halocarbon in the atmosphere is
methyl chloride. A nearly uniform background concentration of 0.6 to 0.7 ppb
has been measured around the globe (Table 13), and no temporal trend is evi-
dent (Singh et al., 1981c). Urban levels of methyl chloride have not been
reported in the literature, although Lovelock (Watson et al., 1979) did report
methyl chloride levels of over 2 ppb in Kenya. The methyl chloride urban data
base presented here is the most extensive available to date, because solid
sorbents such as Tenax®, which are used for routine data collection, do not
appear to collect methyl halides (Pellizzari and Bunch, 1979;* Bozzelli et
al., 1980). Based on our measurements, it appears that although typical
methyl chloride levels in urban areas are close to or only slightly elevated
above background levels, concentrations an order of magnitude higher than the
background can be encountered. Thus average methyl chloride levels of 0.60 to
0.85 ppb measured at Sites 5-10 are indistinguishable from the background.
Average levels approaching 3 ppb at Los Angeles and 1 ppb at Houston are
clearly elevated. As is clear from Table 9, methyl chloride from man-made
sources is extremely small compared with its natural (oceanic) source. How-
ever, unidentified primary or secondary sources of methyl chloride must exist
at least in some of the cities. Methyl chloride is slowly decomposed (Table
8) in the atmosphere (daily loss rate <0.5%), and its global residence time is
estimated to be between 1 and 2 years. Figure 14 shows the actual measured
concentrations of methyl chloride at six selected urban sites.
Methyl Bromide—Methyl bromide levels as high as 1 ppb were measured in
Los Angeles (Site 1). At Staten Island (Site 8), a concentration as high as
0.67 ppb was measured—but only once. The highest average levels of 0.25 ppb
were measured in southern California (Sites 1 and 8), perhaps because of the
application of methyl bromide as a fumigant in this area (Table 9). At all
other sites, average levels were approximately of 0.1 ppb or less. In almost
all cities, however, methyl, bromide levels were significantly elevated above
the background of 10 to 15 ppt (Table 13). Substantial natural sources con-
tributing to this background must exist (Lovelock, 1975). Suggestions have
been made that such gasoline additives as 1,2 dibromoethane could be decom-
posed to form methyl bromide (NAS, 1978). Data to support this suggestion are
currently limited: other than data reported by us and a few isolated measure-
ments reported in NAS (1978), very few methyl bromide measurements exist.
Pellizzari and Bunch (1979) noted traces of methyl bromide on some of their
Tenax® cartridges but did not quantify these. Figure 15 shows the variability
of methyl bromide at three selected sites. Both primary and secondary sources
of methyl bromide may exist in urban areas, but the actual nature of these
*This reference summarizes data from a number of studies conducted over a
period of several years under several different contracts from EPA.
39
-------
•>wu
o 2500
=
£2000
1500
I
D 1000
r>
I
U
500
o
Thu Fri Sat Sun Mon Tue Wed Thu Fri SatJ
_,
o .
0 . -
o
§ «ooo
fc
8 3000
- . 4 a
°o* o „
D
n O n ^Pft a
0 0 »°T! o ° 0 .
00 000
. 0 B 0°* 0 0 0 _
a
zooo
o
n
u 1000
I
I . I . I . I . 1 . I . I . I . I . I
Fri Sat Sun Mon Tue Wed Thu Fri Sat Sun
-
_ _
— —
•
° -0 °° 0 ~
0 o0 ° 0D°o OOD o DO OD oo
1 . 1 . 1 . 1
0123456789 1C
TIME — days
(a) HOUSTON. TX
TIME — days
(d) STATEN ISLAND. NY
3000
o 2500
| 2000
5
£ 1500
D 1000
n
U
500
I F I II I I 1 I
Fri Sat Sun Mon Tue Wed Thu Fri Sat Sun
o-oo
° 0*0 o o .o
O 0 • .
OD
«,„„ •O0o_
4000
a 3000
2000
O
O
I
1000
0123456789 10
TIME — days
(b) ST. LOUIS. MO
Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri
0123456789 10
TIME — days
(e) PITTSBURGH. PA
3000
e 2500
2000
1 1500
1000
Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri Sat
sooo
§. 3000
2000
O
o
J o IOOO
1 . I
0123456789 10
TIME — days
Tue Wed Thu Fri Sat Sun Mon Tue Wed Thu
Q o
012
3 4.5 6 7
TIME — days
8 9 10
(c) RIVERSIDE. CA «> CHICAGO. IL
Figure 14. Methyl chloride in the .ambient air of selected cities.
40
-------
1400
I 120°
f 1000
g 800
&
i 600
o 400
f
0 200
^* 1 ^ 1 " 1 * 1 ' I ~* 1^* 1 " I * I ~*~~
-Fri Sat Sun Mon Tue Wed Thu Fri Sat Sun-
.•"/•.'IT? -
I ..-•... .-.a—. •, i a , . ,,,i
0 12 3 45 67 8 9 10
. TIME — days
(a) ST. LOUIS, MO
1400
§ 1200
Is
* 1000
i
g 800
I 600
1
o 400
i"
0 200
0
Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri Sat-
. .
o
-
o
.' •• • o"
0 ° ° -
0 ° 0 "
0 0°0 •' . • 0
°° .. ~.° c on CB DO _,
0 0oaBaOB • 0
0 0° ° ° « •
T I . i . i . I . i
0 1 2 3 4 5 6 7 8 9 10 11
TIME — days
(b) RIVERSIDE, CA
TOO r-^ 1 1 1 1 . 1 1 f .
. Fri Sat Sun Mon Tue Wed Thu ° Sat Sun.
aoo
joo
J, "»
U
KM
. oo „ - o
00° „,«
01 23456789 10
TIME — days
(c) STATEN ISLAND, NY
Figure 15. Atmospheric concentrations of methyl bromide.
41
-------
sources has not yet been characterized. Methyl bromide is chemically removed
from the atmosphere at a slow rate (Table 8) that is comparable to the removal
rate of methyl chloride.
Methyl Iodide—Methyl iodide was carefully measured to avoid interference
or contamination from other pollutants. It was resolved on two different CC
columns (columns 2 and 6 in Table 3): The results were very nearly identical.
As indicated in Table 1, methyl iodide is both a suspected carcinogen and a
mutagen. It is different from methyl chloride and methyl bromide primarily
because of its rather exclusive natural source. Our measurements at all urban
sites point to average methyl iodide levels in the 1- to 4-ppt range, with 2
ppt perhaps a typical average value. These concentrations are very nearly the
same as or slightly less than those measured near marine environments (Table
13 and Singh et al. , 1981). This is not surprising, because methyl iodide is
primarily of oceanic origin (Lovelock, 1975) and is significantly less stable
than either methyl chloride or methyl bromide. Methyl iodide is decomposed by
sunlight in the troposphere, and a daily loss rate of approximately 12 percent
is estimated (Table 8). Limited atmospheric concentration data from clean as
well as polluted environments, as summarized by Chameides and Davis (1980),
point to a considerable variability. A substantial part of this variability,
we believe, is associated with earlier measurement problems. Significantly
elevated methyl iodide levels at Bayonne, New Jersey, reported by Lillian et
al. (1975), appear anomalous. Measurements at Bayonne should be repeated.
Our findings, based on measurements conducted at several cities, point to very
low methyl iodide levels, with an average value of about 2 ppt and a maximum
value of less than 11 ppt. Figure 16 shows the diurnal variation of methyl
iodide at two selected sites. A slight dip in the afternoon is indicative of
photochemical loss as well as of possible vertical gradients in the concentra-
tion of methyl iodide.
Although methyl iodide is a major carrier of organic iodine in the bio-
sphere, its sources and its atmospheric role are currently not well under-
stood. A certain marine algae known to concentrate .iodine from sea water has
been identified as one source of methyl iodide (Lovelock, 1975; Watson et al.,
1980). It has also been postulated that methyl iodide could react with
chloride ions in sea water to form methyl chloride (Zafiriou, 1975).
Chameides and Davis (1980) have postulated that methyl iodide could photolyze
to form iodine atoms, which could lead to some ozone destruction within the
boundary layer. However, their calculations are based on average methyl
iodide levels of 10 to 50 ppt, rather than 1 to 4 ppt.
Methylene Chloride—Methylene chloride is clearly a large-volume chemical
of exclusively anthropogenic origin (Table 9). To the best of our knowledge
no natural sources of methylene chloride have been identified. Maximum con-
centrations of 12 ppb, 9.ppb, 18 ppb, and 56 ppb were measured at Los Angeles
(Site 1), Riverside (Site 7), Staten Island (Site 8) and Chicago (Site 10),
respectively. Average levels were highest in southern California, with con-
centrations of 3.7 ppb and 1.9 ppb at Los Angeles and Riverside, respectively.
Concentration levels at Staten Island and Chicago were about 1.7 ppb each; at
all other sites, average concentrations ranged between 0.4 and 1 ppb. Back-
ground levels of methylene chloride are about 50 ppt at 40°N (Table 13); thus
a significant elevation above background levels is evident. Average urban
42
-------
.16
i
C 8
l
o
I"
10 15
TIME — hour
(a) HOUSTON, TX
20
25
16
i
£ 8
'"A
0
ill ' .if m M I
10 15
TIME — hour
20
25
(b) RIVERSIDE. CA
Figure 16. Mean diurnal variation of methyl iodide.
concentrations are one or two orders of magnitude higher than the background
environment. The highest average levels were typically encountered at night.
Figure 17 shows typical diurnal variations at six selected sites.
Methylene chloride is relatively unreactive, and a daily chemical loss
rate of less than 2 percent is estimated (Table 8). The diurnal variation of
methylene chloride is therefore primarily determined by its source strength
and the atmospheric mixing processes. The afternoon minimum observed at
several sites (e.g. Figure 17) can only be attributed to dilution caused by
deep vertical mixing. In the absence of local emissions inventories and
detailed meteorological analysis, further conclusions would be premature.
Pellizzari and Bunch (1979) have also reported methylene chloride concen-
trations from several locations within the United States. Although our data
are not necessarily inconsistent with their results, certain discrepancies are
43
-------
4UUU
3500 •
3000 .
2500
2000 •
1500 •
N
N 1000 -
500 •
0
i ' i '.i ' i
• |2o
T T
• 1 1 1 ' m i h
-!• . -v i . 4- . i
0 5 10 15 20
TIME — hour
(a) HOUSTON, TX
4000
3500
g 3000
f 2500
0)
Q.
. 2000
r
* 1500
N 1000
o
CM
Z 500
U
0
1 " I ' 1 • 1
-
T '
1 T
i
1 T
i
" X • T
1 . ^ I . 1 . * 1
0 5 10 15 20-
TIME — hour
(b) DENVER, CO
| 3500
= 3000
i
2 2500
i
2000
r1 1500
N
• 1000
500
1 ' 1 "1
T ' ' '
T
1 n
:im ' 1 ''
.
'
i . i . i . i
0 5 10 15 20
TIME — hour
(c) RIVERSIDE, CA
-
-
-
-
_
I-
aiAMj
C
.2 4000
!3000
Jo
Q.
2000
N
CN
(J 1000
1 . 1 , , .
•
T
-
- 1 T T
LI M,J. ,' :
, -1 I ¥ •
25 o s 10 is . 20 zs
TIME — hour
(d) PHOENIX, AZ
_
-
_
_
-
_
8000
C
g
£
•- 6000
a
*^
4000
I
1
N
0MZOOO
I
o
A
1 ' 1 ' 1
- -
— _
_
-
•
' I ' '
'.. . ^ J.J.I I -J,- -
25 "o s to is 20 z.
TIME — hour
(e) STATEN ISLAND, NY
-
-
-
?:
_
TWVW
O 3500
*" 3000
o
a
» zsoo
ra
a
2000
CM
cj 1000
500
25 °e
i ' i ' i
T
: ' I I I I '^
IJ n "
T ' X J- -
•^ J.
I i
S 10 IS 20 ZS
TIME — hour
(1) CHICAGO. IL
Figure 17. Mean diurnal variation of methylene chloride at selected sites.
-------
evident. For example, they report many concentrations significantly below the
geochemical background that has been relatively well characterized (Cox et
al., 1976; Singh et al., 1979a, 1981c; Cronn et al., 1977). A maximum concen-
tration of slightly over 300 ppb was also reported from an industrial site in
Edison, New Jersey.
Chloroform—Chloroform, a rautagen and a suspected carcinogen (Table 1),
has received a great deal of attention in recent years because of high concen-
trations that have been found to be present in drinking water (Symons et al.,
1975). The background of chloroform is in the 10- to 20-ppt range (Table 13).
Tables 10 through 12 clearly show that urban levels are significantly
elevated. Highest concentration levels approaching 5 ppb were measured at
more than one site. At Houston and Riverside, average concentrations were
found to be 0.4 ppb and 0.7 ppb respectively. At most other sites typical
average concentrations were in the vicinity of 0.1 ppb.
The direct emissions of chloroform appear to be too small (Table 9) to
account for its pervasiveness in urban environments. Figure 18, showing the
diurnal variation of chloroform at Phoenix (Site 2), and Figure 19, showing
raw data from Staten Island (Site 8), leave little doubt that :high chloroform
levels are typical for geographically widely separated areas. The urban
sources of chloroform are probably secondary in nature and also complex. In a
recent review (Batjer et al., 1980), chlorination of water and possibly auto-
mobile exhaust were identified as two important sources. The reactivity of
chloroform is comparable to that of methylene chloride (Table 8), and its
diurnal variation is therefore not chemically controlled.
Although chloroform levels in several urban environments have been
reported by Pellizzari and Bunch (1979), a wide variability does not allow
useful comparisons. Pellizzari and Bunch (1979) report concentrations that
ow
400
c
g •
'5
£ 300
M
I
1 200
o
0
X
0
100
o
•
.
— '
•
_
1
.
I" '.
'
1
.
' 1
T T I
1 i i . - J
•
—
-
_
i
•
.
i.i.i.
10 IS
TIME — hourt
20
25
Figure 18. Mean diurnal variation of chloroform at Phoenix,
45
AZ.
-------
ouu
400
c
g
5 300
0.
I
a
1 ZOO
n
O
O
100
0
(
1 ' 1 ' 1 ' 1
Fri Sat Sun Mon Tue Wed Thu Fri Sat Sun
- —
.
O
o
— o -
o o
•
- . - • •
-•.-•'• •. .-». •
o _ _ "8
° ° On0 ^ 0 0
oao a°qj d" Qjpli o°_ o _a
d> o °a
•
1 . 1 . 1 . 1
) 1 Z 3 4 56789 1C
TIME — day»
Figure 19. Atmospheric concentrations of chloroform at Staten Island, NY.
vary from unquantifiable levels to about 7 ppb, a range comparable to that
found during this study.
Carbon Tetrachloride—Carbon tetrachloride, a suspected carcinogen, has
been found to be-nearly uniformly distributed around the globe at a background
concentration of about 125 to 150 ppt (Table 13 and Singh et al., 1979a).
Considerable evidence suggests that this background reservoir of carbon tetra-
chloride is of man-made origin (Singh et al., 1976; Altshuller, 1976). Urban
carbon tetrachloride levels are higher than the background levels by about a
factor two or three. The highest levels, approaching 3 ppb, were measured in
Houston (Site 4), but the average concentration was still only 0.4 ppb. At
most other sites, average carbon tetrachloride levels were between 0.2 to 0.3
ppb.
Figure 20 shows the kind of scatter typically observed at three different
sites. The mean diurnal variation at Staten Island (Site 8) is shown in Fig-
ure 21. The afternoon minimum is comparable to the background carbon tetra-
chloride levels, a condition accomplished by deep vertical mixing during the
afternoon hours. In a manner somewhat similar to that observed for chloro-
form, the highest levels are encountered during stagnant night hours.
The carbon tetrachloride levels measured here are not only in good agree-
ment with our earlier published data (e.g. Singh et al., 1977a, 1979a), but
also agree well with a three-day study conducted in Los Angeles by Simmonds et
al. (1974). They measured carbon tetrachloride levels in the range of 0.1 to
2 ppb, with an average of 0.22 ppb [compared with our results of 0.1 to 1 ppb
and an average of 0.22 ppb at Los Angeles (Site 1)]. In addition, Pellizzari
and Bunch (1979) and Bozzelli et al. (1980) have both reported carbon tetra-
chloride data from several locations using the Tenax® collection process. (It
appears that there are serious problems associated with the use of Tenax® for
carbon tetrachloride measurement. The bulk of the data presented by these two
investigators is almost a factor of 10 lower than the geochemical background
46
-------
1500
w 1000
u
a
-------
2000
1500
-ooo
0
0
500
I"
10 IS
TIME — hours
20
25
Figure 21. Mean diurnal variations of carbon tetrachloride
at Staten Island, NY.
of carbon tetrachloride.) Ohta et al. (1976) report unusually high carbon
tetrachloride concentrations from Tokyo: The average concentration level of
1.4 ppb is a factor of three to six higher than typical averages found during
this study.
Haloethanes and Halopropanes
Nine important chemicals in this category were measured: ethyl chloride;
1,1 dichloroethane; 1,2 dichloroethane; 1,2 dibromoethane (or ethylene
dibromide); 1,1,1 trichloroethane; 1,1,2 trichloroethane; 1,1,1,2 tetra-
chloroethane; 1,1,2,2 tetrachoroethane; and 1,2 dichloropropane. Seven of
these (excluding ethyl chloride and 1,1,1,2 tetrachloroethane) are either bac-
terial mutagens or suspected carcinogens (Table 1).
Ethyl Chloride—Ethyl chloride is a commonly used chemical intermediate
(Table 9). It is estimated that about 0.01 million tons of ethyl chloride are
released into the atmosphere every year in the United States. A daily chemi-
cal loss rate of about 3 percent is estimated (Table 8). Average ethyl
chloride concentrations at all sites were 0.1 ppb or less. [Houston (Site 4),
whose average and maximum levels were 0.23 ppb and 1.3 ppb, respectively was
an exception (Figure 22). At no other site did the maximum concentration ever
exceed 0.32 ppb.] Background concentration of ethyl chloride (at 40°N) is
found to lie between 10 to 15 ppt (Table 13), clearly suggesting that signifi-
cant urban sources exist. No atmospheric data on ethyl chloride could be
found in the literature. Part of the reason may be its poor collection effi-
ciency on Tenax®, which has been frequently used for urban monitoring of toxic
chemicals.
48
-------
1000
800
£ 600
£
O
in
I
CM
O
400
200
Th
1 1 ' 1 ' lo ' J, ' 1 ' 1 w'^ 1 ' 1 ' 1 '
hu Fri Sat Sun ° Mon Tue We2 Thu Fri Sat
DO
a
°a o
o-a
o 0
o a
t*a
I 2 3 4 S 6 7
NO. OF DAYS
(a) HOUSTON, TX
8 9 10
3UU
400
c
o
SL 300
V)
Q.
i
1 200
O
in
0*
100
o
1 1 ' 1 ' 1 ' 1 - '
Wad Thu Fri Sat Sun Mon TUB Wed Thu Fri
_ • . .
.
_ • ' w
'
a •
- J? —
O OJ
r o 8
00
fa
a o_ °
"W11 °\aJ9> "on? ^a ^° ° ^>°aa>
.1.1.1.1,
01 23456789
TIME — d»yi
(b) PITTSBURGH, PA
Figure 22. Atmospheric concentrations of ethyl chloride.
10
-------
Dichloroethanes—At all sites monitored, 1,1 dichloroethane was present
in relatively low concentrations. Average concentrations at any of the sites
did not exceed 0.07 ppb; the maximum measured concentration did not exceed
0.15 ppb. At Sites 8-10, extremely low average levels (10 to 15 ppt) were
encountered (Table 12); comparable concentration levels have been reported by
Pellizzari and Bunch (1979) from parts of New Jersey and Los Angeles. Figure
23 shows the mean diurnal variation of 1,1 dichloroethane at Denver (Site 6)
and Riverside (Site 7). A daily loss rate of only about 2 percent is
estimated (Table 8).
2UU
c
o
5 150
&
S 100
1
M
O
5 50
£
0
0
1
I"
-
_
.
•
1
1
1 T
1 f
i J
11
1
1
-
1
1
1
1
A
1
10 15
TIME — hour
(a) DENVER. CO
20
25
c
_o
1 iso
I
£
S 100
CM
O
u
£
u
50
10 15
TIME — hour
(b) RIVERSIDE. CA
20
25
Figure 23. Mean diurnal variation of 1,1 dichloroethane.
1,2 dichloroethane is a large-volume chemical (Table 9) that is a bac-
terial mutagen and a suspected carcinogen (Table 1). Estimated yearly U.S.
emissions exceed 0.2 million tons. It is an exclusively man-made chemical
that has also become a part of our global environment. Singh et al. (1981c)
report a background concentration of about 40 ppt at 40°N (Table 13). It is
obvious from Tables 10-12 that urban levels are significantly elevated. The
50
-------
highest average concentration (1.5 ppb) was measured in Houston, compared with
average levels of 0.1 ppb to 0.5 ppb at all other sites. The maximum concen-
tration (7.3 ppb) was also measured at Houston (Site 4), followed by Staten
Island (Site 8), where a maximum of 4.3 ppb was measured. Figure 24 provides
a comparison of the atmospheric concentrations of 1,2 dichloroethane at Hous-
ton (Site 4) and Pittsburgh (Site 9). Figure 25 shows the mean diurnal varia-
tion at Houston and Riverside.
CM
O
O 2000
CM
O
Thu Fri Sat Sun Man Tue Wed Thu Fri Sat
O D
]
°o° o
°°
01 23456789 10
TIME days
. (a) HOUSTON. TX
•Wed Thu Fri Sat .Sun Mon Tue Wed Thu Fri -
** MOO
I
£ ZMO
O 1500
I
C 100°
CM.
I ...
01 23456 789 10
TIME — days
(b) PITTSBURGH, PA
Figure 24. Atmospheric concentration of 1,2 dichloroethane.
The measurements of 1,2 dichloroethane conducted by Bozzelli et al.
(1979) and Pellizzari and Bunch (1979) provide a great deal of data that
appear to be well below the measured (as well as estimated) background of this
chemical (Singh et al., 1981c; Altshuller, 1980). While higher numbers
51
-------
4000
c
.9 3500
* 3000
&
g 2500
&
1 2000
O 1500
IM
| 1000
M
5 500
0
i
-
^
-
-
"l
-
1
1
I
i
i
i
i . i
0-^5 10
i • i
i
i
•
i
<
15
(
•
_
-
-
-
I
1
20 2
TIME — hour
(a) HOUSTON, TX — 15-24 MAY 1980
-------
Although it is highly toxic, 1,2 dibromoethane is present in urban atmo-
sphere at relatively low concentrations. The average'concentration at none of
the sites exceeded 0.06 ppb. The typical range of average concentrations was
from 0.015 ppb to 0.06 ppb. The highest concentrations of 0.37 ppb were meas-
ured in Houston (Site 4); highest levels in Los Angeles, Phoenix, and Chicago
were in the vicinity of 0.2 ppb. Figures 26 and 27 show the distribution and
the mean diurnal variation of ethylene dibromide at selected sites.
\
Limited ethylene dibromide data from U.S. cities is available from the
literature. Ambient levels in the 15- to 30-ppt range were reported by Going
and Spigarelli (1976). Leinster et al. (1978) have reported ambient levels of
between 0 and 20 ppt in London air. Bozzelli et al. (1980) report rather high
levels from several New Jersey cities, with maximum concentrations of about 6
ppb. [Quantifiable samples indicate average concentrations of 0.5 to 1 ppb.
These results are clearly in disagreement with ours.] Data reported by Pel-
lizzari and Bunch (1979) away from the highways are also in the 0 to 0.3 ppb
range, although highway air concentrations as high as 8 ppb are reported. The
discrepancies that currently appear to exist should be resolved, because 1,2
dibromoethane is expected to be a potent carcinogen.
Trichloroethanes—1,1,1 trichloroethane is another large-volume chemical
that is released in significant quantities to the atmosphere (Table 9). The
chemical has a long atmospheric lifetime and is globally distributed (Singh et
al., 1979a). Its atmospheric residence time is estimated to be about eight
years (Singh, 1979b); thus about 15 percent of all 1,1,1 trichloroethane
released at ground level enters the stra-tosphere, where it can interact with
ozone in a way similar to fluorocarbons. 1,1,1 trichloroethane is suspected
to be weakly mutagenic (Table 1), although considerable disagreement on its
mutagenic and carcinogenic potential persists (Farber, 1979; Lapp et al.,
1979). The background burden of this chemical is constantly increasing; back-
ground concentration is now reported to be about 0.18 ppb (Table 13).
The highest 1,1,1 trichloroethane concentration, 5.1 ppb, was measured in
Los Angeles. Average concentration at all sites ranged between 0.2 and 1 ppb.
Typical diurnal variations at selected sites are shown in Figure 28. Figure
29 shows the raw data.as well as the mean diurnal profile at Staten Island:
Although nighttime averages are typically high, the associated higher standard
deviations at Staten Island are easily explained from raw data. Indeed, high
midnight values were encountered on only three of the days monitored (Figure
29).
Because of its potential stratospheric significance, a great deal of data
on 1,1,1 trichloroethane have been collected in clean environments around the
globe. Once again the urban data base has been limited. Simmonds et al.
(1974) conducted limited measurements in Los Angeles in 1973 and reported a
concentration range of 0.01 ppb to 2.3 ppb with an average of 0.37 ppb. The
absolute coulometric technique utilized by these authors is known to underes-
timate the actual concentrations, especially for relatively inefficient elec-
tron absorbers (Lillian and Singh, 1974). As a comparison, our average con-
centrations are 1 ppb at Los Angeles (Site 1) and 0.7 ppb at Riverside (Site
7). The agreement is quite good if one recognizes that the emissions of 1,1,1
53
-------
cw
j
: ISO
I
»
i
1 100
5
N
5 so
5
N
3
i • i • i • i
I"
-
-..
i
i
I
i
\ I
I.I.I.
I
i
-
-
•
10 IS
TIME — tain
(a) PHOENIX, AZ
CQ 100
(M
I
U
i so
I
U
10 IS
TIME — hour.
20 25
(b) HOUSTON, TX
^ 200
I
r iso
w 100
m
CM
o
J »
u
i ii 11
. 11
10 IS
TIME — hour«
(e) DENVER, CO
Figure 26. Mean diurnal variation of 1,2 dibromoethane.
54
-------
I
o
I I 1 I 1 f 1 I II
Tue Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri
a o
o ",
o a
o"*
_n° 1 " ° ~~«D
*. I V "•! . I" 1° ~ " In"
f
01 23456 789 10 11
TIME — days
(a) PHOENIX. AZ
•= too
o
u
T -j -i j T j- -i | i -j —r | -i -f r I T I 1
Thu a Sat Sun Mon Tue o Wed Thu Fri Sat
F
O 01 „ o
-o o -"'"S-'1,,*0"00
. i . i . i . r . i °. r . i . i . i .
01 23456789 10
TIME — days
(b) HOUSTON. TX
290
c
o
!g 200
i
S ISO
m
(N
I
o
m so
N
o
Mon Tue Wed Thu Fri Sat Sun Mon Tue Wed Thu
o o
o
O " Q O
H-ff on " "o
a of o "t^ff
I . I
og
- °J°0
n on ° O
. "\
0123456789 10 11
TIME — days
(c) DENVER. CO
Figure 27. Atmospheric concentrations of 1,2 dibromoethane.
55
-------
auwu
| 4000
a
i
1 2000
n
U
U
X 1000
O
0
c
,
-,
I-
.
. «
1
5
1 ' 1 ' 1
1
t .
Illl.l
• i . I . I
10 IS 20
TIME K/t.ir
.
'
_
-
«JUO
c
o
| 1500
i
w
« 1000
a.
I
n
8 500
n
O
-
-
"'
25 U
1 * 1 ' 1 * 1
~
l
'
"
' ' { I I I « { I 1
.1.1.1.
5 10 15 20 2!
TIME — hour
(a) LOS ANGELES, CA
(c) HOUSTON, TX
•WOO
c 3500
O
% MOO
a
0- ZSOO
lit
a. 2000
nisoo
O
O 1000
U
soo
-
-
-
i
1 ' 1 ' 1 ' 1
"
-
•
1
1
,1,
'iM.,1
'
i
-
^uuu
1
=c 1500
I
5
a 1000
I
U 500
1°
U
n
1
-
I
"
-ll
1
1
i i i •
-
i
•111.
1
I
•
-
10 13
TIME — hour
(b) PHOENIX. AZ
25
TIME — hour
(dl DENVER. CO
Figure 28. Mean diurnal variation of 1,1,1 trichloroethane.
trichloroethane have nearly doubled in the last six to eight years. A sub-
stantial amount of data have also been reported by Pellizzari and Bunch
(1979). Once again they report a significant number of measurements that are
inconsistent with measured as well as estimated background levels (Table 13).
1,1,2 trichloroethane, a suspected carcinogen, was measured at extremely
low concentration levels at all sites. Typical average concentration was
around 0.01 ppb, except, at Houston (Site 4) and Riverside, where average con-
centrations of 0.03 ppb and 0.04 ppb, respectively, were measured. A maximum
concentration of 0.15 ppb was measured at Houston; at all other sites maximum
measured levels were below 0.1 ppb. Bozzelli et al. (1980) report a concen-
tration range in New Jersey that is qualitatively estimated to be between 0-
0.01 ppb, although they report one data point as high as 11 ppb in Elizabeth,
New Jersey. Pellizzari and Bunch (1979) also report levels of <0.01 ppb to
56
-------
2000
0 1500
1000
_r>
U
5 500
-r
I 1 ,
S 10 IS
TIME — hours
(a) MEAN DIURNAL VARIATION
*
20
25
. ISOO
S. 1000
O
U
r>
U 500
1 . 1 ' 1 . 1
Frl Sat Sun Mon Tue Wed Thu Fri Sat Sun
0
°
°°aa
a
o
3456
TIME — days
7.8 9 10
(bl ATMOSPHERIC CONCENTRATIONS
Figure 29. 1,1,1 trichloroethane behavior at Staten Island, NY.
57
-------
2 ppb at sites in New Jersey, Texas, and Louisiana. Despite its structure
similarity to 1,1,1 trichloroethane, the 1,1,2 isomer is 30 times more reac-
tive (Table 8). A daily chemical loss rate of about 3 percent is estimated
(Table 8).
Tetrachloroethanes—Two tetrachloroethane isomers (1,1,1,2 and 1,1,2,2)
were measured, both of which were present at extremely low concentrations.
The two isomers were present together at an average concentration of less than
0.02 ppb. At no time did the concentration of either one of these chemicals
exceed 0.1 ppb. The symmetric isomer (1,1,2,2) is found to be a bacterial
mutagen and a suspected carcinogen (Table 1). The asymmetric isomer (1,1,1,2)
has been tested for mutagenicity with negative results. Tetrachloroethanes
are virtually inert (Table 8) in the atmosphere.
Bozzelli et al. (1980) analyzed nearly 209 Tenax® cartridge samples of
air collected from New Jersey for 1,1,2,2 tetrochloroethane and were able to
quantify only six. The average concentration in these six samples was 3 ppb.
Pellizzari and Bunch (1979) have also reported data for 1,1,2,2 isomer in the
0.01 ppb range although levels as high as 3 ppb were also measured near indus-
trial sites.
1,2 dichloropropane—1,2 dichloropropane, a bacterial mutagen (Table 1),
was the only chlorinated propane measured. Its average measured concentration
was in the range of 0.02 ppb to 0.05 ppb at all sites except Houston (Site 4),
where an average concentration of 0.08 ppb and a maximum concentration of 0.25
ppb were measured. At no other site did the maximum concentration ever exceed
0.01 ppb. We expect 1,2 dichloropropane to be fairly reactive and estimate a
daily loss rate of about 10 percent (Table 8). Figure 30 shows the mean diur-
nal variations of 1,2 dichloropropane at Riverside and Staten Island. Seven
samples from Louisiana (Tenax® trapped) were analyzed by Pellizzari and Bunch
(1979). In five out of seven, an average 1,2 dichloropropane concentration of
0.02 ppb was measured. One sample was measured at the 1-ppb level.
Chloroalkenes
Six chloroalkenes were sought: vinylidene chloride (1,1 dichloroethy-
lene), (cis) 1,2 dichloroethylene, trichloroethylene, tetrachloroethylene,
allyl chloride (3 chloro-1-propene), and hexachloro-1,3 butadiene. Of these,
allyl chloride, a suspected carcinogen, was never detected at our measurement
sensitivity of 5 ppt (Table 10-12). Similarly, vinylidene chloride (a ba.c-
terial mutagen and suspected carcinogen) was never measured at an average con-
centration exceeding 0.03 ppb. Approximately 30 to 50 percent of the time,
vinylidene chloride was below the limit of detection of about 5 ppt. Concen-
tration as high as 0.22 ppb was detected in Denver (Site 5), but maximum con-
centrations were typically in the order of 0.1 ppb. The low abundance of
vinylidene chloride is related to its relatively low emission levels (Table 9)
as well as to its high reactivity. A 30 percent daily chemical loss rate
(Table 8) could prevent any atmospheric accumulation of vinylidene chloride.
Another equally reactive dichloroethylene (cis-1,2) was found to. be somewhat
more ubiquitous. Average concentrations at all sites varied between 0.02 ppb
and 0.08 ppb. A concentration as high as 0.6 ppb was measured in Denver
58
-------
c
o
= 200
1
£ iso
&
1
111
f too
a.
0
C
0.
S 50
O
Z
u
0 0
.1-1 1 1
T*°
-
T T — —
ll T T T 1
TJvlTT«l •
IjJlif'I-
i 1 11
1 . 'l . 1 . 1
10 15
TIME — hour
(a) RIVERSIDE, CA
20
25
a
~ iso
uj
Z 100
a.
O
CE
a.
° ^
OC »
O
X
o
i 10 IS
TIME — hour
(bl STATEN ISLAND, NY
Figure 30. Mean diurnal variation of 1,2 dichloropropane.
(Site 6). Unlike vinylidene chloride, the symmetric isomer is not found to be
a mutagen. No carcinogenicity data on 1,2 dichloroethylene are currently
available.
Vinylidene chloride has also been measured by Pellizzari and Bunch (1979)
at several U.S. locations. Trace quantities were detected but quantification
at below 0.05 to 0.1 ppb level was generally not possible. Occasionally, how-
ever, scattered data between 0.01 ppb and 0.6 ppb have been reported. A
nearly identical situation exists for 1,2 dichloroethylene, with extremely
limited data reported in the range of 0.01 ppb to 1 ppb. In both cases, it is
evident that the measurement sensitivity was inadequate during much of the
experimentation of Pellizzari and Bunch (1979).
Tri- and Tetra-Chloroethylenes—One of the two dominant chloroethylenes
is trichloroethylene.It is a large-volume chemical (annual U.S. emissions =
0.15 million tons) that is also a bacterial mutagen and a suspect carcinogen
59
-------
(Table 1). As stated earlier, disagreement on the carcinogenicity of tri-
chloroethylene still persists (Albert, 1980; Demopoulos et al., 1980). The
highest concentration of 3 ppb was measured in Los Angeles. At most other
sites (except Sites 7 and 9) the highest measured concentration was between 1
and 2 ppb. The maximum and minimum concentration levels in Tables 10-12
clearly show a wide variability. Average concentrations at all sites were
between 0.1 ppb and 0.6 ppb, compared with a geochemical background of about
0.01 ppb to 0.02 ppb (Table 13) at midlatitudes. Part of the atmospheric
variability of trichloroethylene is due to its relatively fast atmospheric
removal rate. We estimate a daily chemical loss rate of about 17 percent
(Table 8).
Figure 31 shows the mean diurnal variation of trichloroethylene at
Phoenix (Site 2), Houston (Site 4), and Denver (Site 6). In all cases the
highest averages are encountered during the stagnant nighttime hours. The
afternoon minimum is due to a superimposition of dilution caused by deep vert-
ical mixing and to the compound's substantial reactivity. The high nighttime
values are also often associated with increased variability. This is perhaps
best illustrated in Figure 32, where both the mean diurnal profile and the raw
data are shown for Pittsburgh (Site 9). A look at the raw data clearly shows
the high nighttime averages are greatly influenced by the unusually high con-
centrations measured on a few nights.
Trichloroethylene has been'measured by a number of investigators in urban
environments (Singh et al., 1977c, 1979a,b; Lillian et al., 1975; Bozzelli et
al., 1980; Pellizzari and Bunch, 1979). Lillian et al. (1975) reported aver-
age concentrations of 0.1 ppb to 0.9 ppb at several east coast locations. A
concentration as high as 18 ppb was reported from Bayonne, New Jersey. Boz-
zelli et al. (1980) were able to quantify less than 50 percent of the col-
lected samples and report average concentrations of 1 to 2 ppb from six New
Jersey sites. Ohta et al. (1976) reported high concentrations (average = 1.2
ppb) of trichloroethylene from Tokyo. Pellizzari and Bunch (1979) show a wide
variability with concentrations as high as 17 ppb from industrial sites in
Edison, New Jersey.
The second large-volume chloroethylene that is also a suspected carcino-
gen (Albert, 1980; Greenberg and Parker, 1979) is tetrachloroethylene.
Approximately 0.3 million tons of this chemical are emitted annually in the
United States (Table 9): Unlike trichloroethylene, the reactivity of tetra-
chloroethylene is modest. We estimate that slightly less than 2 percent is
depleted daily (Table 8). Because of its larger emissions and its reduced
reactivity, tetrachloroethylene is present at a background concentration of
approximately 50 ppt (Table 13). At all sites except Los Angeles (Site 1) and
Phoenix (Site 2), the average measured tetrachloroethylene concentration was
between 0.3 ppb and 0.6 ppb. At Los Angeles and Phoenix, considerably higher
average concentrations (1.5 ppb and 1.0 ppb, respectively) were measured. The
highest measured concentration was 7.6 ppb at St. Louis (Site 5); at all other
sites maximum tetrachloroethylene levels were typically between 1 ppb and 3
ppb. Average concentration of tetrachloroethylene was higher than
60
-------
3500
c
g
= 3000
I. 2SOO
S
a 2000
_n
O
X 1000
M
O
soo
I I
10 IS
TIME — hour
(a) PHOENIX. AZ
1000
£ 800
1 .
I
600
400
n
O
Lil
10 15
TIME — hour
(b) HOUSTON. TX
.11
•••20
25
1 www
| 800
& 600
«t'
I
| 400
n
0
5,200
.
.
_
-
1
,
_'
1 J
,
0 5
jj,]
10 15
-
•
-
T
i
20 •"• 2!
TIME — hour
(e) DENVER. CO
Figure 31. Mean diurnal variation of trichloroethylene.
61
-------
1000
800
| 600
_n
X
u
CM
U
400
200
1
5 10 IS
TIME — hour
(i) MEAN DIURNAL VARIATION
20
25
1000
800
600
a
o.
_r>
X
o
u
400
zoo
Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri
a o
o
_ a
DO
o °° a
%Qoa °/ ° "
Q I g i I
1 2 3 4 5 6. 7
TIME — days
8 9 10
(b) CONCENTRATION
Figure 32. Trichloroethylene behavior at Pittsburgh, PA.
62
-------
trichloroethylene at all sites. Typically this ratio was between two and
four. At remote sites, a similar ratio is also encountered (Table 13; Singh
et al., 1981c).
The diurnal behavior of tetrachloroethylene was very similar to that of
trichloroethylene. Figure 33 shows the mean diurnal behavior of tetra-
chloroethylene at Phoenix (Site 2) and Denver (Site 6). Once again high
nighttime values are encountered. The mean diurnal variation of tetra-
chloroethylene at Pittsburgh, with the raw data, are shown in Figure 34. Once
again the reasons for the higher standard deviations observed at nighttime are
clear.
A number of studies dealing with the atmospheric abundance of tetra-
chloroethylene were recently reviewed by Greenberg and Parker (1979). Addi-
tional data have also been provided by Singh et al. (1980). Lillian
sooo
c
o
- 4000
WOO
; i--
u
o .000
10 IS
TIME — hour
(a) PHOENIX. AZ
2000
§ .
i 1500
a 1000
b 500
1 1 n
0 5 10 15 20 25
TIME — hour
(b) DENVER, CO
Figure 33. Mean diurnal variation of tetrachloroethylene.
63
-------
ZSOO
- aooo
a 1500
u 1000
(M
O
o
500
o i—±—4t
1
5 10 IS
TIME — hours
(a) MEAN DIURNAL VARIATION
ZO
ZS
2500
.c
.9
'£ ZOOQ
9}
a.
M
5 1500
n
O
U
u
o
1000
500
• I ' I ' I ' I '
Wed Thu Fri Sat Sun Won Tue Wod Thu Fri
ffi
0 1 23456789 10
TIME — day*
(b) CONCENTRATION
Figure 34. Tetrachloroethylene behavior at Pittsburgh, PA.
64
-------
et al. (1975) reported average concentrations of 0.1 ppb to 4.5 ppb, with a
maximum value of 8.2 ppb reported from Bayonne, New Jersey. They reported the
highest average concentration, 4.5 ppb, from New York City.
Bozzelll et al. (1980) were able to quantify only a small fraction of the
samples collected. An average concentration range of 0.3 ppb to 4 ppb was
found at six New Jersey sites. Contrary to our findings, their average tetra-
chloroethylene levels are higher than the trichloroethylene levels in only
three of the six sites. Although Pellizzari and Bunch (1979) have detected
tetrachloroethylene in the ambient air, much of the data could not be quanti-
fied. Ohta et al. (1976) also reported an average tetrachloroethylene concen-
tration of 1.2 ppb from Tokyo.
Hexachloro-1,3 butadiene—From the best information we could obtain from
private sources, hexachloro-1,3 butadiene (a bacterial mutagen) is no longer
manufactured in the United States, but it has been identified in the effluents
of sewage treatment plants; thus secondary sources may exist. It may also, be
formed as a byproduct during the combustion of plastics. It was measured at
an average concentration of 1 to 11 ppt; at no time did its concentration
exceed 0.15 ppb. No information is available on the reactivity of this chemi-
cal, but its structure would suggest that it is likely to be highly reactive.
Pellizzari and Bunch (1979) have also reported measuring hexachloro-1,3 buta-
diene at Niagara Falls (New York) between 4 and 100 ppt. At selected sites in
Louisiana and Texas, they reported concentration of <1 ppt to 50 ppt. A sin-
gle measurement showing a concentration as high as 0.3 ppb was also made in
Deer Park, Texas.
Chloroaromatics
Six chloroaromatics were sought: monochlorobenzene, a-chlorotoluene, o-
m-p dichlorabenzes, and 1,2,4 trichlorobenzene. No data on p- dichlorobenzene
could be collected because of unknown interferences. Because of an apparently
malfunctioning detector, all chloroaromatic data from Sites 8-10 had to be
discarded. The production, estimated emissions, and use patterns of the dom-
inant chlorobenzenes were shown in Table 9. It is obvious from the table that
modest amounts of these materials are released into the environment. [Of the
six chlorobenzenes sought, a-chlorotoluene (also known as benzyl chloride) is
the only species that is a clear bacterial mutagen and a suspected carcinogen
(Table 1). Bacterial tests other than the salmonella-typhimurium ("Ames .
Assay") have indicated positive bacterial mutagenicity for dichlorobenzenes.]
Once in the atmosphere, chlorobenzenes are moderately reactive. A chemical
loss rate of approximately 7 percent per day for monochlorobenzene and less
than 3 percent per day for di- and tri-chlorobenzenes is estimated. Q-
chlorotoluene is somewhat more reactive, and approximately 23 percent of its
atmospheric burden could be chemically depleted in one day (Table 8).
Monochlorobenzene was found to be the most abundant of the chloroben-
zenes. Its average concentration at all of the sites monitored was between
0.1 ppb to 0.3 ppb, although concentrations as high as 2.8 ppb were encoun-
tered. These levels are not inconsistent with its .relatively large source
65
-------
(yearly U.S. emissions of about 0.08 million tons) and a moderate removal rate
(7 percent/day). Figure 35 shows the mean diurnal variation of monochloroben-
zene at Denver, which is typical of all chlorobenzenes at this site.
1OOO
c
| 800
1
I
. 600
t
&
1 400
O
in
JJD 200
Q
i • i * i * i
I2o
-
•
-
k-
•
I
1
—
" T T
i —
111, In-
• I.I. I.I.'
0 5 10 15 20 2!
TIME — hour
Figure 35. Mean diurnal variation of monochlorobenzene at Denver, CO.
Data on chlorobenzenes was generally only sparsely available in litera-
ture. Monochlorobenzene was measured by Bozzelli et al. (1980): Data from
the six sites in New Jersey indicated average levels that were between 0.5 and
1.0 ppb. Pellizzari and Bunch (1979), who use very similar sampling tech-
niques, report a large body of data below their limit of detectability of
approximately 0.05 ppb. Occasionally, however, they report concentrations as
high as 1 ppb.
a-chlorptoluene was frequently below our limit of detection of 5 ppt, and
its concentration never exceeded 0.11 ppb. We can only attribute this
behavior to relatively fast removal rate (23 percent per day) and low emis-
sions. The absence of a-chlorotoluene is not inconsistent with its estimated
yearly U.S. emissions of only about 45 tons. No data that were representative
of open ambient atmospheres could be found. Pellizzari and Bunch (1979) meas-
ured this chemical near a Stauffer plant site in Edison, New Jersey, at
between 1-ppb and 2-ppb concentration levels.
Both of the dichlorobenzenes (o- and m-) together were present at an
average concentration of between 10 ppt and 40 ppt at all of the sites moni-
tored. Ortho-dichlorobenzene was measured at the highest concentration of
0.24 ppb in Phoenix (Site 2) and a very comparable maximum concentration of
0.23 ppb at Denver (Site 6). At all other sites the maximum concentration of
o-dichlorobenzene was well below 0.1 ppb. The meta-isomer was frequently less
abundant, and its maximum concentration never exceed 0.06 ppb. Given the
order of magnitude lower emissions of o-dichlorobenzene when compared to mono-
chlorobenzene (Table 9), these levels do not appear unreasonable. Figures 36
and 37 show the mean diurnal variations off o- and m-dichlorobenzene at
selected sites.
66
-------
£UUU
c
g
'C
Z 1500
a
VI
1
1
" 1000
z
UJ
N
Z
m
O
CC
0 500
I
O
5
0
0
(
1
• I"
_
» 1
.
1 a
.
i
> 5
,.,.,.
'
_
'
i
i
i y
,s .T T I. t ,i .
10 """ is "*" -1- 20
.
—
i
,
K
TIME — hour*
Figure 36. Mean diurnal variation of o-dichlorobenzene at Phoenix, AZ.
In ambient surveys conducted by Pellizzari and Bunch (1979), a bulk of
the data base reports trace quantities or nondetectable levels of these
dichlorobenzenes. Part of the reason for this is inadequate measurement sen-
sitivity. The measurement sensitivity as reported by these authors varies
from sample to sample but for o- m-dichlorobenzes was typically in the 0.01 to
0;05 ppb range. Much of the data reported in this study would also be unquan-
tifiable at these sensitivity levels. Pellizzari and Bunch (1979) do occa-
sionally report concentrations in the range of 0.005 ppb to 3 ppb for these
chemicals, their data, however, are obtained in the vicinity of industrial
sources.
1,2,4 trichlorobenzene was ubiquitously present, but its ambient concen-
tration never exceeded 40 ppt. Typical average concentrations were in the 1
to 10 ppt range. Figure 38 shows the mean diurnal behavior of 1,2,4 tri-
chlorobenzene at Riverside. This diurnal pattern was typical of other pollu-
tants at this site. As discussed earlier, trichlorobenzenes are highly
unreactive and a daily loss rate of less than one percent is computed (Table
8).
Aromatic Hydrocarbons
Eight important aromatic hydrocarbons were sought. Benzene is a
suspected human carcinogen (Table 1). Carcinogenicity as well as mutagenicity
information on toluene is disputed (Albert, 1980), although the compound has
been classified as a potential mutagen (U.S. SG, 1980). In most other cases,
toxicity data are currently highly uncertain. Although aromatic hydrocarbons
are manufactured in large quantities, direct releases constitute a minor part
of the atmospheric emissions. Both from the ambient measurements as well as
the emissions data (NAS, 1976; Mayrsohn et al., 1976) it is clear that the
67
-------
w •««»
2L 400
V)
I
| 300
01
Z
UJ
N ZOO
Z
111
m
O
§ K»
_l
I
U
5 o
I ' I ' i . i
11-
-
'
i— —
T 1-
.
-
• ^" IT! T T
T T
. . !-• ,X - . . I
g 0 5 10 IS 20 Z!
TIME — hour
(a) PHOENIX. AZ
§
= 50
40
30
UJ
N -xi
Z ^
UJ
CD
O
§10
' I
U
O 0
U
20
25
Figure
0 5 10 15
TIME — hour
(b) DENVER, CO
37. Mean diurnal variation of m-dichlorobenzene.
dominant ambient source is automobile exhaust. Although a considerable body
of data on aromatic hydrocarbons is available (e.g. Mayrsohn et'al., 1976),
nearly all of those data were collected during daytime.
The two dominant aromatic hydrocarbons are benzene and toluene.. The
average benzene concentration at all sites ranged between 1.5 and 6 ppb,
although concentrations as high as 65 ppb were measured. The average toluene
levels were in the 1.5-ppb to 12-ppb range, and concentrations as high as 67
ppb. were measured. At all sites except Pittsburgh (Site 9), toluene average
concentration was higher than benzene. The average toluene/benzene concentra-
tion ratio at all ten sites was 1.6, with a range of 0.8 to 2.1. In the only
place where average benzene levels were higher than toluene (Pittsburgh) con-
siderable stationary sources of benzene (coke ovens) are known to exist (Mara
and Lee, 1977). Both benzene and toluene are photochemically reactive, and
daily loss rates of 1 1 percent and 41 percent respectively can be computed
68
-------
* 100
80
. I"
ID
N
60
O
ec
O
u
£
20
•I in.'...,. i.I i
10 IS
TIME — hour
20
25
Figure 38. Mean diurnal variation of 1,2,4 trichlorobenzene at Riverside, CA.
(Table 8). This greater reactivity of toluene cause toluene/benzene ratio to
decline as the air masses age. This is partly the reason for a below average
ratio of 1.1 at St. Louis (Site 5), where relatively clean atmospheric condi-
tions were encountered. The mean diurnal behavior for all aromatic hydrocar-
bons measured at a given site was virtually identical, which is suggestive of
a common source. Figures 39 and 40 show the mean diurnal profile for benzene
and toluene at selected sites. It is clear from these two figures that dis-
tinct diurnal patterns exist. Although exceptions can be found, a fairly com-
mon feature is the high nighttime and low afternoon concentrations. Figure 41
shows the same observation for m/p xylenes. The ambient levels of xylenes
(o,m/p) and other aromatic hydrocarbons are shown in Tables 10-12. The xylene
isomers collectively can approach or exceed the concentration levels of ben-
zene, despite the xylenes' extremely high reactivity (Table 8). As a group
aromatic hydrocarbons are important because of their high abundance, potential
toxicity, and high reactivity. Many of the products of oxidation of aromatic
hydrocarbons (e.g. cresols, aromatic aldehydes, phenols) may have toxic
effects that exceed those associated with the parent molecule (Helmes et al.,
1980).
As is typical with virtually all our diurnal-variation findings, we were
unable to find any data taken over the last decade to verify the patterns
observed here. Much of the hydrocarbon data collected to date was obtained
primarily to study photochemical air pollution, which is driven by sunlight
and therefore essentially ceases at night. Comparisons of average concentra-
tions are still possible, however. Mayrsohn et al. (1976) report data from
several locations in the California South Coast Air Basin. Both the measured
levels of aromatics as well as the ratios (e.g. toluene/benzene ratio varies
from 1 to 3, with an average of about 2) are consistent with our results from
Los Angeles (Site 1). Similarly, early morning (6:00 am - 9:00 am) data col-
lected by Westberg et al. (1978) from a site in Houston are compatible with
our Houston (Site 4) data. The average toluene/benzene ratio of 1.8 measured
69
-------
*.«
IS
XI
a
a
1
1 10
10
I
ID
o
s
0
(
1
1 ' 1 ' 1
- 1" 1
—
— 1
.1 .
.
)
'
i
1
6
1
,
i
L ill'
1
' ;
• •
—
L -
I.I.I,
10 16 20 _. Z
«a
20
a
a.
? 15
1
(0
I
-------
so
40
st
a
a
1
f)
I
J"
10
0
(
1
. I"
_
;. i :.
,
) S
1 1 ' 1 ' . 1 '
"
llllll1
1
_
_
-
I.I.I.
10 IS 20 Z
MJ
40
I
\ 30
n
Z
U
s-20
0°
10
1 ' 1 ' 1 ' 1
_
• . -
-
-llljljj>5?n-
, 0 5 10 15 20 2
TIME -hour TIME — hour
(a) LOS ANGELES, CA
(c) DENVER, CO
50
40
a
t 30
Z°
U
•c" 20
cf
10
0
_
-
-
-I
0
1 ' 1 ' 1 ' 1
• '
1
.
-
y -
|T T
, in n:
.1.1.1.
3U
40
1^
n
Z
^
to 20
ID
O
10
1
_.
-
_
I
-
T '
11
1 ' 1 ' 1
_
-
. •'
§ 10 15 20 .25 °o * ' s J
.'.. i
i t ?
_
T
X
10 IS ZO 2.
TIME — hour
(b) HOUSTON, TX
TIME — hour
(d) STATEN ISLAND, NY
Figure 40. Mean diurnal variation of toluene.
Formaldehyde, a suspected carcinogen and a bacterial mutagen (Table 1)
was measured at relatively high concentrations that varied between 6 ppb to 46
ppb. Considerable ambient data on formaldehyde have been collected over the
last two decades by using the chromotropic-acid procedure. Table 14 summar-
izes similar data collected by SRI at several cities during short-term field
studies. Typically, we encountered formaldehyde concentrations averaging 10
to 20 ppb. Table 15 shows formaldehyde concentrations at sites in Pittsburgh
and Chicago, where concurrent measurements with the chromotropic-acid and the
more specific DNPH-HPLC procedures were made. A comparison between formal-
dehyde data collected by these two methods at Pittsburgh and Chicago is shown
in Figure 43. The three data points shown with asterisks in Table 15 are
excluded because sampling problems were encountered during the DNPH-HPLC col-
lection process. The corresponding chromotropic-acid data are, however,
valid. A linear regression analysis (Figure 43) of formaldehyde-concentration
data (ppb) by the DNPH-HPLC method (Y) and the chromotropic-acid method (X) is
best represented by the fit Y = 0.95X - 0.04 with a regression coefficient of
71
-------
25
I20
ui 15
z
UI
x 10
i
E
5
n
- I"
.
T
_
i:
i
T
i
.ihi..lr
10 15
TIME — hour
(a) HOUSTON, TX
20
25
111
ui
QL
1
i T: T.T.i-. i. «.T
10 •>• IS
TIME — hour
(b) STATEN ISLAND, NY
Figure 41. Mean diurnal variation of m/p-xylene.
0.71. The intercept is not significantly different from zero. The error on
the slope is 30 percent. Averages computed from the Pittsburgh and the Chi-
cago data individually are very nearly identical (Table 15). The variability
in the fractional differences [Y - 0.95X/(Y + X)0.5] is computed to be less
than ±30 percent. This disagreement is not considered unreasonable since the
overall accuracy of either of these methods, in their present state of
development, is expected to be comparable to these differences. A substantial
part of the uncertainty associated with the DNPH-HPLC method is caused by
impurities in solvent solutions and can be eliminated or further reduced in
the future. Although additional studies under atmospheric conditions should
be made, it does appear that past formaldehyde data, collected by and large by
the chromotropic-acid procedure, represents a valid data base.
Altshuller and McPherson (1963) measured an average of 40 ppb formal-
dehyde compared to 19 ppb measured in this study at Riverside (table 15).
Cleveland et al. (1977) have reported extensive measurements from New Jersey
with average concentration of about 10 ppb. The peak concentrations, as
72
-------
JD
a
<0
Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri
23456789 10
TIME — days
t
n
U
1 J— -, I . I I
Wed Thu Fri Sat Sun Mon Tue Wed oThu Fri
0123456789 10
TIME — days
40
a
a.
a
Ul
U)
> 20
X
E
10
Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri
0123456789 10
TIME — days
Figure 42. Aromatic hydrocarbons at Pittsburgh, PA.
73
-------
TABLE 14. AMBIENT FORMALDEHYDE LEVELS IN SELECTED LOCATIONS
AS MEASURED WITH THE CHROMOTROPIC-ACID PROCEDURE
Field Site
City
St. Louis, MA
Denver, CO
Riverside, CA
Staten Island, NY
Pittsburgh, PA
Chicago, IL
Latitude
(°N)
38°46'
39045'
33°59'
40°35'
40°26'
41°45'
Longitude
(°W)
90°17'
104°59'
117°18'
74°12'
79°56'
87°42'
Experiment
Period
5-7 June 80
23-24 June 80
8-10 July 80
3-4 April 81
15-16 April 81
27-28 April 81
Number
of Data
Points*
11
18
18
17
21
8
Concentration
(ppb)
Max
18.7
28.7
41.0
45.9
35.1
17.2
Average ±o-
11.3 ± 4.5
12.3 ± 5.9
19.0 ± 7.6
14.3 ± 9.1
20.6 ± 5.2
12.8 ± 3.3
Each data point represents approximately a 2-hour average concentration.
-------
TABLE 15. COMPARISON OF FORMALDEHYDE AND ACETALDEHYDE DATA
Sampling Site
and Date
Pittsburg
15 April 1981
16 April 1981
Average (all data)
Average (excluding
asterisked data)t
Chicago
27 April 1981
28 April 1981
Average (all data)
Average (excluding •
asterisked data)T
Sampling
Period
(hours)
1655-1830
1645-1845
1830-2000
1950-2050
2000-2130
1950-2150
2140-2310
2150-2250
2310-0030
2250-0100
0040-0200
0100-0200
0200-0340
0200-0400
0340-0510
0319-0500
0510-0700
0500-0700
0910-1050
0920-1020
1830-2000
1800-2000
2030-2200
2030-2200
2200-2400
2200-2400
0003-0200
0002-0200
0200-0400
0200-0400
0400-0600
0400-0600
,
Formaldehyde Concentration (ppb)
DNPH-HPLC
Method
28.5
16.4
10.3
19.7
25.7
12.4
22.9
12.8*
12.4
10.1*
17.2 ± 6.7
18.5 ± 6.7
13.1
5.6
15.6
9.9
12.1
6.6*
10.5 ± 3.9
11.3 ± 3.8
Chromo tropic-Acid
Method
22.8
15.3
16.4
16.0
21.2
19.1
25.0
28.5
19.5
27.2
21.1 ± 4.7
19.4 ± 3.5
13.0
9.1
9.9
10.5
17.2
17.2
12.8 ± 3.6
11.9 ± 3.3
Acetaldehyde
Concentration (ppb)
DNPH-HPLC
Method
0.2
1.0
1.1
2.1
1.0
1.3
2.6
0.8*
2.1
1.8*
1.4 1 0.7
1.4 ± 0.8
2.4
2.2 .
3.4
1.7
0.9
0.3*
1.8 ± 1.1
2.1 ± 0.9
Sampling problems encountered—outlier data.
'Although the corresponding chromotropic-acid data are reliable, they are excluded
for consistent comparisons.
75
-------
30
28
26
§24
!| 22
U
a! 20
i 18
Q
z- 16
O
c
5 12
o
8 10
UJ
> 8
•I
UJ
O 6
5
cc
o
I I I I I I I I 7 I I
BEST FIT
Y = 0.95 X-0.04
(R = 0.71)
Y- X
(45° LINE) Vv
I
I
I
I
I
I
I
I
0 2 4 6 8 10 12 14 16 18 20 22 . 24 26 28
FORMALDEHYDE CONCENTRATION, CHROMOTROPIC-ACID METHOD — ppb
Figure 43. Comparison of formaldehyde concentrations as measured by the
chromotropic acid and the DNPH-HPLC procedure.
measured by the upper decile, were in the 14- to 20-ppb range. These are
similar to our average Staten Island levels of 14 ppb. Similarly, Joshi
(1977) reported average concentrations of about 10 ppb from Houston, although
maximum concentrations as high as 27 ppb were measured. Kok (1980), using a
chemiluminescent technique, reports formaldehyde levels of 8 ppb to 38 ppb
(average 19 ppb) during a pollution episode (September 13-14, 1979) in Los
Angeles. Kuwata et al. (1979) used the DNPH-HPLC procedure to report average
concentrations of 27 ppb from limited measurements in Osaka, Japan. In clean
background locations formaldehyde levels of about 0.4 ppb have been measured
and computed from mechanisms involving methane oxidation (Table 13; Ehhalt and
Tonnissen, 1980).
Acetaldehyde data are significantly more sparse than formaldehyde data.
Hoshika (1977) and Kuwata et al. (1979) provide a limited number of measure-
ments. From these data, acetaldehyde levels of 1 to 10 ppb in the ambient air
have been reported. Kuwata et al. (1S79), who measured both formaldehyde and
acetaldehyde, report average acetaldehyde concentrations of about 4.8 ppb.
(On the average, formaldehyde is about six times more abundant than acetal-
dehyde in Osaka.) Table 15 shows the concentrations of formaldehyde and
76
-------
acetaldehyde as measured by the DNPH-HPLC procedure during our study. The
average acetaldehyde concentrations at Pittsburgh and Chicago are in the 1- to
2-ppb range, and therefore are significantly lower than corresponding formal-
dehyde levels. The average f ormaldehyde/acetaldehyde ratio of 12 and 6 at
Sites 9 and 10 respectively is comparable to a ratio of 6 reported from Osaka
by Kuwata et al. (1979).
The reactivity of acetaldehyde (due to photolysis and reaction with OH
radicals) is comparable to that of formaldehyde, and a daily loss rate of
about 80 to 95 percent respectively can be computed (Table 8). When one con-
siders automobile exhaust as a major emission source, 65 to 75 percent (by
volume) of all aldehydes is formaldehyde, while 7 to 10 percent is acetal-
dehyde (NAS, 1976). If we assume equal reactivity, a formaldehyde/
acetaldehyde ratio of 6 to 11 is entirely consistent with an automobile
source .
Singh and Hanst (1981) estimate that approximately 40 ppt of acetaldehyde
is present in the lower troposphere as an intermediate photochemical product
of nonmethane hydrocarbons. No information on the carcinogenity of acetal-
dehyde could be found.
It appears that the DNPH-HPLC method may provide a technique for the
ambient analysis of a wide variety of carbonyl compounds. A comparison of
emission levels (NAS, 1976) and acetaldehyde field data would suggest that
higher aldehydes (C, - Dy) are likely to be present at even lower concentra-
tions (sub-ppb). In our sampling protocol, a measurement sensitivity of 0.05
ppb is feasible. Both the measurement sensitivity and the accuracy of data
collected by the DNPH-HPLC method could be further improved by reducing or
eliminating solvent impurities. While this study was limited to formaldehyde
and acetaldehyde, the presence of other carbonyls was evident. There is some
evidence that acetaldehyde may not be rautagenic (U.S. SG, 1980; Sasaki and
Endo, 1978).
Phosgene was not detected at most sites, largely because the coulometer
was also used for analysis of PAN and PPN. Extensive column conditioning is
also required for phosgene analysis, which could only be done with great dif-
ficulty in the field. Average phosgene levels as high as 50 ppt (but often
below 20 ppt) were encountered. Phosgene is expected to be a photochemical
product of the oxidation of chlorinated ethylenes (Singh, 1976; Gay et al.,
1976).
As is clear from Tables 10-12, PAN and PPN average levels were quite low.
Highest PAN and PPN concentrations of 16.8 ppb and 2.7 ppb respectively were
measured in Los Angeles (Site 1). The maximum PAN levels of 16.8 ppb at Los
Angeles (Site 1) can be compared with a maximum of 4.4 ppb in Houston (Site
4), 5.8 ppb in Riverside (Site 7) and 3.9 ppb in Staten Island (Site 8).
Daily average PAN levels were in the 0.3 ppb to 5 ppb range, with signifi-
cantly reduced values at night. The PAN/PPN average ratio varied from 4 to
10. PPN was nondetectable a significant fraction of the time (30 to 60 per-
cent). Figure 44 shows the mean diurnal variation of PAN and PPN at Phoenix
(Site 2). This diurnal behavior of PAN is fairly typical of all urban sites
77
-------
g 6000
2000
JLt
10 15
NUMBER OF HOURS
(a) PAN
20
25
800
c 600
o
a
» 400
2
a.
200
10 15
NUMBER OF HOURS
(b) PPN
20
25
Figure 44. Average diurnal variation of PAN and PPN at Phoenix — Site 2.
(e.g. Nieboer and Von Ham, 1976). "It is pertinent to repeat here that the
absolute coulometric analysis was used for PAN and PPN measurements; a method
that has yet to be rigorously tested.
PAN has been measured by a number of investigators (EPA, 1978). From Los
Angeles, Hoboken (New Jersey), and St. Louis, average daytime PAN levels of 18
ppb (0 to >70 ppb), 4 ppb (0 to 10 ppb) , and 6 ppb (0 to >12 ppb) respectively
have been measured. Measurements from the Houston area, as reported by Ludwig
and Martinez (1979), indicate significantly lower PAN levels between 0 and 16
ppb, with about 70 percent of the data reported as 0 (less than 0.2 ppb).
Little data from clean background locations are available, but measurements
78
-------
from rural sites suggest PAN levels in the 0.1- to 0.5-ppb range (Singh et
al., 1979a; Lonneman et al., 1978). Singh and Hanst (1981) estimate that at
midlatitudes the lower troposphere contains about 10 to 30 ppt of PAN.
Both PAN and PPN are rapidly removed from the atmosphere, and a daily
loss rate of 99 percent is computed (Table 8). Because this loss rate is
highly temperature dependent, PAN is nearly infinitely stable at upper levels
of the troposphere (Singh and Hanst, 1981). PAN has not been tested for
mutagenicity or carcinogenicity, but it is a well-known eye irritant and is
known to cause visible damage to agricultural crops (EPA, 1978).
79
-------
SECTION 7
SUMMARY AND CONCLUSIONS
This three—year research effort comprised a program of analytical methods
development, field-data collection, data processing, and data interpretation
for a group of 44 organic chemicals, of which 29 are bacterial mutagens and
more than a dozen are suspected carcinogens. All field measurements were con-
ducted on-site with the help of an instrumented mobile environmental labora-
tory. The chemical categories targeted for field measurements included
chlorofluoromethanes, halomethanes (nonfluorinated), haloethanes,
chloroethylenes, chloroaromatics, aromatic hydrocarbons, and oxygenated
species. The ambient analysis of these species was possible with the help of
electron capture gas chromatography for the halogenated and nitrogenated
species, flame ionization gas chromatography for hydrocarbons, and high-
performance liquid chromatography for aldehydes. After the analytical methods
development was completed, a total of ten field studies were' conducted at a
selected site within the following cities:
• Los Angeles, California
• Phoenix, Arizona
• Oakland, California
• Houston, Texas
• St. Louis, Missouri
• Denver, Colorado
• Riverside, California
• Staten Island, New York
• Pittsburgh, Pennsylvania
• Chicago, Illinois.
Although these studies were of short term-duration, our practice of round-
the-clock operation allowed for extensive data collection. The degree of tem-
poral and spatial variability in the atmospheric abundance of toxic chemicals
is clear from data presented. Typical concentrations of most chemicals meas-
ured were in the sub-ppb range, with the exception of aromatic hydrocarbons
and formaldehyde (where average concentrations in the 5 to 20 ppb range were
frequently encountered). For most predominantly man-made chemicals, average
concentrations in urban atmospheres were one to two orders of magnitude higher
than in clean remote atmospheres.
80
-------
Distinct mean diurnal variations in the concentrations of these atmos-
pheric chemicals exist. For most chemicals, the mean .diurnal variations are
determined by source strength and prevailing meteorology, with chemistry play-
ing only a nominal role. Chemical loss rates for a majority of species were
shown to be <10 percent/day. For several primary pollutants, afternoon mixing
leads to sufficient dilution to cause an afternoon minimum in concentrations;
secondary photochemical pollutants, however, show a clear afternoon maxima
(e.g. PAN, PPN). Thus for many of the toxic chemicals the highest concentra-
tions in the ambient air are encountered at nighttime or early morning hours.
There is abundant evidence that most of the chemicals measured here (except
methyl halides and aldehydes) have nearly exclusive man-rmade origin. The sig-
nificant elevation in concentration'above background in urban areas points to
large sources associated with man-made activities: Methyl iodide was the only
chemical that appears to have an exclusive natural source.
The chemicals measured in this study are important not only for their
potential toxicity but also for their role as indicators of urban photochemis-
try. The many chemicals with a range of removal rates (lifetimes) provide an
ideal opportunity for studying the chemistry, of the urban atmosphere. Such
analysis, however, must wait until accurate emissions information becomes
available.
Some man-made chemicals are sufficiently stable and are released in large
enough quantities to have become a part of the global environment. Carbon
tetrachloride is one such chemical, which is nearly uniformly distributed over
the globe as a result of slow accumulation and a lack of rapid removal mechan-
isms (Singh et al., 1976; 1979a,b). Methylene chloride, 1,2 dichloroethane,
and tetrachloroethylene, however, are emitted in such large quantities (global
release rates of 0.4 to 0.6 million tons per year for each) that even a rela-
tively fast atmospheric removal rate (atmospheric lifetime of two to eight
months) does not prevent their spread and accumulation.
An investigation of the mutagenicity of chemicals clearly showed that
methyl chloride, methyl bromide, methyl iodide, and formaldehyde are mutagens.
These chemicals are known to be a ubiquitous part of our natural atmosphere
(and oceans, in the case.of methyl halides).
The total exposure to mutagens and carcinogens from the urban ambient air
.is of course much higher than measured here because of nongaseous species.
(e.g. polyaromatic hydrocarbons) as well as other gaseous species for which
either toxicity studies are inconclusive or measurement methods are inadequate
(e.g. oxygenated chemicals). Most synthetic chemicals in this study came into
major use after 1950. Since then, their production and release have continued
to grow exponentially, with a doubling time of about six years (Bauer, 1978).
Because of the long time lag (10 to 50 years) associated with the onset of
cancer (LaFond, 1978), a significant risk may not be identified until a future
date. It is also possible that continuous exposure to low levels of such
chemicals may erode any human threshold that may exist or enhance the fre-
quency of cancer occurring from other primary causes, such as cigarette smok-
ing (Albert and Burns, 1977).
81
-------
On the whole, we conclude that typical urban atmospheres contain chemi-
cals that are known to be toxic at much higher concentrations. The risks
associated with exposure to ambient levels of these species are highly uncer-
tain. The task of characterizing the atmosphere, with which this study is
most concerned, is itself at best highly incomplete. Much more atmospheric
and toxicity data will be needed to determine the risks associated with long-
term exposures to low levels of toxic species.
82
-------
SECTION 8
RECOMMENDATIONS FOR FUTURE RESEARCH
One of Che primary functions of this study was to develop techniques for
the atmospheric measurement of important organic chemicals and to apply these
under field conditions. Both of these objectives were partially achieved, but
the analytical methods for chloroaromatics and (especially) for oxygenated
species must be further improved.
The DNPH-HPLC technique was applied to the measurement of formaldehyde
and acetaldehyde, but the presence of other aldehyde was evident. In addi-
tion, only ten sites could be studied each for a period of about 9 to 11 days
each. The data base must be expanded to include other sites and other chemi-
cals, and researchers must conduct studies during several seasons. Because of
the complexity of the mix of ambient chemicals, it is possible that signifi-
cant spatial and temporal'variations exist. The interpretation of data
collected to date is incomplete, at least in part due to lack of emissions
inventories of these chemicals. In many cases complex secondary sources are
evident. .
Interlaboratory comparisons of field data should be rigorously pursued.
Preliminary comparisons with other data, the bulk of which are collected by
using solid "sorbents (mostly Tenax®) followed by GC-MS analysis, suggest some
inconsistencies.that could be resolved with additional research and interla-
boratory comparisons.
83
-------
REFERENCES
ADL, 1975: "Preliminary Economic Impact Assessment of Possible Regulatory
Action to Control Atmospheric Emissions of Selected Halocarbons," EPA
450/3-75-073.
Albert, R.E. , 1980: "The Carcinogen Assessment Group's Assessment of
1 ,2-Dichloroethane, 1 ,2-Dibromoethane, Vinylidene Chloride, Tri-
chloroethylene, Tetrachloroethylene, Methylene Chloride, Benzene,
Toluene, and Formaldehyde for Carcinogenicity ," submitted to U.S. EPA
(Group chairman, R.E. Albert).
Albert, R.E. , and F.J. Burns, 1977: "Carcinogenic Atmospheric Pollutants and
the Nature of Low-Level Risks," in Origins of Human Cancer — Book A, pp.
289-292, Cold Spring Harbor Laboratory,
Altshuller, A. P., 1976: "Average Tropospheric Concentration of Carbon Tetra-
chloride Based on Industrial Production, Usage, and Emissions," Environ.
Soi. Technol. , Vol. 10, pp. 596-598.
Altshuller, A. P. , 1980: "Lifetime of Organic Molecules in the Troposphere and
Lower Stratosphere," Advances in Environmental Science and Technology,
Vol. 10, pp. 181-215.
Altshuller, A. P. and S.P. McPherson, 1963: "Spectrophotometric Analysis of
Aldehydes in the Los Angeles Atmosphere," J_. Air Poll. Control Assoc. ,
Vol. 13, pp. 109-111.
Arnts, R. , 1980: .Private communication, ESRL, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina.
Atkinson, R. , K.R. Darnall, A.C. Lloyd, A.M. Winer, and J.N. Pitts, Jr.,
1979: "Kinetics and Mechanisms of the Reaction of the Hydroxyl Radical
With Organic Compounds in the Gas Phase," Adv. Photochem. , Vol. 11, pp.
375-488.
Batjer, K. , M. Cetin Kaya, J. Duszeln, B. Gabel, U. Lahl, B. Stachel, and W.
Thiemann, 1980: "Chloroform Emissions Into Urban Atmospheres," Chemo-
sphere, Vol. 9, pp. 311-316.
Bauer, E. , 1979: "A Catalog of Perturbing Influences on Stratospheric Ozone,
1955-1975," J. Geophys. Res., Vol. 84, pp. 6929-2940.
84
-------
Bozzelli, J.W., B.B. Kebbekus, and A. Greenberg, 1980: "Analysis of Selected
toxic and Carcinogenic Substances in Ambient Air in New Jersey," Final
Report submitted to New Jersey Department of Environmental Protection by
New Jersey Institute of Technology.
CIS, 1981: Chemical Information Service, SRI International, Menlo Park,
California.
Calvert, J.G. (1976): "Hydrocarbon Involvement in Photochemical Smog,"
Environ. Sci. Technol., Vol. 10, pp. 256-262.
Chameides, W.L., and D.D. Davis, 1980: "Iodine: Its Possible Role in Tropo-
spheric Photochemistry,"^, of Geophys. Res., Vol. 85, pp. 7383-7398.
Cleveland, W.S., T.E. Graedel, and B. Kleiner, 1977: "Urban Formaldehyde:
Observed Correlation With Source Emissions and Photochemistry," Atm.
Env., Vol. 11, pp. 357-360.
CMA, 1981: "World Production and Release of Chlorofluorocarbons 11 and 12
Through 1980, July 29, 1981," Chemical Manufacturers Association, Wash-
ington, D.C.
Cox, R.A., R.G. Derwent, A.E.G. Eggleton, and J.E. Lovelock, 1976: "Photo-
chemical Oxidation of Halocarbons in the Troposphere," Atm. Env. , Vol.
10, pp. 305-308.
Cox, R.A., and M.J. Roffey, 1977: "Thermal Decomposition of Peroxyacetyl
Nitrate in the Presence of Nitric Oxide," Environ. Sci. Technol., Vol.
11, pp. 900-906.
Cronn, D.R., R.A. Rasmussen, and E. Robinson, 1977: "Measurements of Tropo-
spheric Halocarbons by Gas Chromatography/Mass Spectrometry," Report for
Phase II, EPA Grant R0804033-02.
Detnopoulos, H.B., B. Wagner, and J. Ciraino, 1980: "An Academic Review of the
Hazards Posed by Trichloroethylene," New York University Medical Center,
unpublished manuscript.
Ehhalt, D.H. and A. Tonnissen, 1980: "Hydrogen and Carbon Compounds in the
Stratosphere," Report FAA-EE-80-20, pp. 129-151.
EPA, 1978: "Air Quality Criteria for Ozone and Other Photochemical Oxidants,"
EPA-600/8-78-004, NTIS PB80-124753.
Farber, H.P., 1979:- "1,1,1-Trichloroethane As An Industrial Solvent: A
Review of Current Health and Environmental Knowledge," Dow Chemicals
U.S.A., Midland, Michigan.
Ferraan, M.A., R.S. Eisinger, P.R. Monson, 1977: "Characterization of Denver
Air Quality," Denver Air Pollution Study—1973, Proceedings of a
Symposium—Vol. II, EPA 600/9-77-001.
85
-------
Fung, K., and D. Grosjean, 1981: "Determination of Nanogram Amounts of Car-
bonyls as 2,4-Dinitrophenlhydrazones by High Performance Liquid Chroma-
tography," Anal. Chem., 53, pp. 168-171.
Gay, B.W., P.L. Hanst, J.J. Bufalini, and R.C. Noonan, 1976: "Atmospheric
Oxidation of Chlorinated Ethylenes," Env. Sci. Technol., Vol. 10, pp.
58-67.
Going, J.E. and J.L. Spigarelli, 1976: "Sampling and Analysis of Selected
Toxic Substances Task IV - Ethylene Dibromide," EPA 560/6-76-021.
Greenberg, M.M., and J.C. Parker, 1979: "Health Assessment Document for
Tetrachlorethylene," U.S. Environmental Protection Agency, External
Review Draft No. 1, Research Triangle Park, North Carolina.
Hampson, R.F., 1980: "Chemical Kinetics and Photochemical Data Sheets for
Atmospheric Reactions," Report FAA-EE-80-17, National Bureau of Stan-
dards, Washington, D.C.
Helmes, C.T., C.C. Sigman, K.L. Thompson, J. Jaffer, D.L. Atkinson, and P.A.
Sullivan, 1980: "Evaluation and Classification of the Potential
Carcinogenicity of Air Pollutants," SRI International, NCI Contracts
N01-CP-33285 and 95607, Menlo Park, California.
Hendry, D.G., A.C. Baldwin, and D.M. Golden, 1980: "Computer Modeling of
Simulated Photochemical Smog," EPA 600/3-80-029.
Hendry, D.G. and R.A. Kenley, 1979: "Atmospheric Chemistry of Proxynitrates,"
in Nitrogenous Air Pollutants (D. Grosjean, ed.), Ann Arbor Science,
Michigan, pp. 127^1^S~.
Hoshika, Y., 1977: "Simple and Rapid Gas-Liquid-Solid Chromatographic
Analysis of Trace Concentrations of Acetaldehyde in Urban Air," J.
Chromatogr. , Vol; 137, pp. 455-460. ~
Hudson, R.D., and E.I. Reed, 1979: "The Stratosphere: Present and Future,"
NASA Reference Publication 1049.
Hull, L.A., 1980: "Procedure for 2,4-Dinitrophenylhydrazone Aldehyde-Ketone
Air Analysis," Dept. of Chemistry, Union College, New York, unpublished
report.
Joshi, S.B., 1979: "Houston Field Study—1978 Formaldehyde and Total
Aldehydes Monitoring Program," EPA Contract 68-0*2-2566, Northrop Services
Inc. Report ESC-TR-79-22.
Kok, G.L., 1980: "Atmospheric Measurements of Formaldehyde and Hydrogen
Oxide," EPA Grant R80662910, Progress Reports 1 and 2.
86
-------
Kuwata, K., M. Uebori, and Y. Yamasaki, 1979: "Determination of Aliphatic and
Aromatic Aldehydes in Polluted Airs as their 2,4-Dinitrophenylhydrazones
by High Performance Liquid Chromatography," ^J. of Chr. Sci., Vol. 7, pp.
264-268.
LaFond, R.E. (ed.), 1978: "Cancer The Outlaw Cell," American Chemical
Society, Washington, D.C., p. 192.
Lapp, W.T., B.L. Herndon, C.E. Murama, A.D. Tippit, and R.P. Reisdorf, 1979:
"An Assessment of the Need for Limitations on Trichloroethylene, Methyl
Chloroform, and Perchloroethylene," EPA Contract 68-01-4121, MRI Project
No. 4276-L, Draft Final Report.
Leinster, P., R. Perry, and P.J. Young, 1978: "Ethylene Dibromide in Urban
Air," Atm. Env., Vol. 12, pp. 2383-2387.
Lillian, D., H.B. Singh, A. Appleby, L. Lobban, et al., 1975: Atmospheric
Fates of Halogenated Compounds," Env. Sci. Technol., Vol. 9, pp. 1042-
1048.
Lillian, D., and H.B. Singh, 1974: "Absolute Determination of Atmospheric
Halocarbons by Gas Phase Coulometry," Anal. Chem. , Vol. 46, pp. 1060-
1063.
Lonneman, W.A., J.J. Bufallni, and R.L. Seila, 1976: "PAN and Oxidant Mea-
surement in the Ambient Atmosphere," Environ. Sci. Technol., Vol. 10, pp.
374-380.
Lovelock, J.E., 1975: "Natural Halocarbons in Air and in the Sea," Nature,
Vol. 256, pp. 193-194.
Ludwig, F.L., and J.R. Martinez, 1979: "Aerometric Data Analysis for the
Houston Area Oxidant Study," Final Report for Houston Area Oxidant Study,
Contract DA-1, Project 7106, SRI International, Menlo Park, California.
Mara, S.J., and S.S. Lee, 1977: "Human Exposure to Atmospheric Benzene,"
Contract 68-01-4314, Project EGU-5794, SRI International, Menlo Park,
California.
Mayrsohn, H., M. Kuramoto, J.H. Crabtree, R.D. Sothern and S.H. Mano, 1976:
"Atmospheric Hydrocarbon Concentrations, June - September 1975,"
DTS-76-15, California Air Resources Board.
McCann, J., and B.N. Ames, 1977: "The Salmonella/Microwave Mutagenicity
Test: Predictive Value for Animal Carcinogenicity," in Origins of Human
Cancer, Cold Spring Conference on Cell Proliferation, Vol. 4, pp.
1431-1450.
Molina, M.J. and F.S. Rowland, 1974: "Stratosphere Sink for Chloro-
fluororaethanes: Chlorine Atom Catalyzed Destruction of Ozone," Nature,
Vol. 249, pp. 810-812.
87
-------
NAS, 1976: "Vaporphase Organic Pollutants," National Academy of Sciences,
Washington, D.C.
NAS, 1978: "Chloroform, Carbon Tetrachloride, and Other Halomethanes: An
Environmental Assessment," National Academy of Sciences, Washington, D.C.
Nieboer, H., and J. Van Ham, 1976: "Peroxyacetyl Nitrate (PAN) in Relation to
Ozone and Some Meteorological Parameters at Delft in the Netherlands,"
Atm. Env. , Vol. 10, pp. 115-120.
Ohta, T., M. Morita, and I. Mizoguchi, 1976: "Local Distribution of Chlori-
nated Hydrocarbons in the Ambient Air in Tokyo," Atm. Env., Vol. 10, pp.
557-560.
Pellizzari, E.D., and J.E. Bunch, 1979: "Ambient Air Carcinogenic Vapors -
Improved Sampling and Analytical Techniques and Field Studies,"
EPA-600/2-79-081, NTIS PB297932, Final Report submitted to U.S. EPA by
Research Triangle Institute.
Peto, R., 1980: "Distorting the Epidemiology of Cancer: The Need for a More
Balanced Overview," Nature, Vol. 284, p. 297.
Salas, L., and H. Singh, 1981: "Analysis of Formaldehyde and Acetaldehyde in
Ambient Air Using the Dinitrophenyl-Hydrazone-HPLC Procedure and Its Par-
tial Verification," Atm. Env. , in preparation.
Sasaki, Y. , and Endo, R., 1978: "Mutagenicity of Aldehydes in Salmonella,"
Mut. Res., Vol. 54, pp. 251-252. .
Sievers, R.E., 1981: "Artifact Problems in Atmospheric Analysis of Organic
Compounds and Strategies for Minimization," National Symposium on Moni-
toring Hazardous Organic Pollutants in Air, Raleigh, North Carolina
(April 28 - May 1).
Simmonds, P.G., S.L. Kerrin, J.E. Lovelock, and F.H. Shair, 1974: "Distribu-
tion of Atmospheric Halocarbons in the Air Over the Los Angeles Basin,"
Atm. Env. , Vol. 8, pp. 209-216.
Singh, H.B., 1976: "Phosgene in the Ambient Air," Nature, Vol. 264, pp.
428-429.
Singh, H.B., and P.L. Hanst, 1981: "Peroxyacetyl Nitrate (PAN) in the Unpol-
luted Atmosphere: An Important Reservoir for Nitrogen Oxides," Geophys.
Res. Lett., Vol. 8, pp. 941-944.
Singh, H.B., D. Lillian, and A. Appleby, 1975: "Absolute Determination of
Phosgene: Pulse Flow Coulometry," Anal. Chem., Vol. 47, pp. 860-864.
Singh, H.B., D.P. Fowler, and T.O. Peyton, 1976: "Atmospheric Carbon Tetra-
chloride: Another Man-Made Pollutant," Science, Vol. 192, pp. 1231-1234.
88
-------
Singh, H.B., L. Salas, and L.A. Cavanagh, 1977a: "Distribution, Sources, and
Sinks of Atmospheric Halogenated Compounds," J_. Air Poll. Contr. Assoc. ,
Vol. 27, pp. 332-376. ~
Singh, H.B., L. Salas, D. Lillian, R. Arnts, and A. Appleby, 1977b: "Genera-
tion of Accurate Halocarbons Primary Standards Using Permeation Tubes,"
Environ. Sci. Techno1., Vol. 11, pp. 511-513.
Singh, H.B., L. Salas, H. Shigeishi, and A. Crawford, 1977c: "Urban-Nonurban
Relationships of Halocarbons, SF,, NnO, and Other Atmospheric Trace Con-
stituents," Atm. Env. , Vol. 11., pp. 819-828.
Singh, H.B., L.J. Salas, H. Shigeishi, A.J. Smith, E. Scribner, and L.A.
Cavanagh, 1979a: "Atmospheric Distribution Sources and Sinks of Selected
Halocarbons, Hydrocarbons, SF& and N20," EPA-600/3-79-107.
Singh, H.B., L. Salas, H. Shigeishi, and E. Scribner, 1979b: "Atmospheric
Halocarbons, Hydrocarbons and SF^: Global Distributions, Sources and
Sinks," Science, Vol. 203, pp. 899-903.
Singh, H.B., L.J. Salas, A. Smith, H. Shigeishi, 1979c: "Atmospheric Measure-
ments of Selected Toxic Organic Chemicals," EPA-600/3-80-072.
Singh, H.B., L.J. Salas, A. Smith, R. Stiles, H. Shigeishi, 1980: "Atmos-
pheric Measurements of Selected Hazardous Organic Chemicals,"
EPA-600/3-81-032.
Singh, H.B., J.R. Martinez, D.G. Hendry, R.J. Jaffe, and W.B. Johnson, 1981a:
"Assessment of the Oxidant-Forming Potential of Light Saturated Hydrocar-
bons in the Atmosphere," Environ. Scj^. Technol., Vol. 15, pp. 113-119.
Singh H.B., L.J. Salas, and R. Stiles, 1981b: "Quality Assurance Plan.
Atmospheric Measurements of Toxic Organic Chemicals—Cooperative Agree-
ment 805990," Project 7774, prepared for U.S. Environmental Protection
Agency, SRI International, Menlo Park, California.
Singh, H.B., L.J. Salas, and R. Stiles, 1981c: "Trace Chemicals in the Clean
Troposphere," EPA 600/3-81-055.
Symons, J.M., T.A. Bellar, J.K. Carswell, J. DeMarco, et al., 1975: "National
Organics Reconnaissance for Halogenated Organics," J_. Amer. Wat. Wks.
Assoc., pp. 634-637.
U.S. Public Health Service, 1965: "Selected Methods for the Measurement of
Air Pollutants," Publication 999-AP-ll, Cincinnati, Ohio.
U.S. Surgeon General, 1980: "Health Effects of Toxic Pollutants: A Report
from the Surgeon General, Department of Health and Human Services,"
Report prepared for U.S. Senate, Serial No. 96-15 (August).
89
-------
Watson, A.J., J.E. Lovelock, and D.H. Stedway, 1979: "The Problem of Methyl
Chloride," FAA-EE-80-20, U.S. Department of Transportation, Washington,
D.C.
Westberg, H., K. Allwine, and E. Robinson, 1978: "Measurement of Light Hydro-
carbons and Oxidant Transport," EPA-600/3-78-062.
Zafiriou, 0., 1975: "Reaction of Methyl Halide with Seawater and Marine Aero-
sols," J. Marine Res., Vol. 33, pp. 73-49.
90
------- |