EPA-600/3-76-085
July 1976
257 512
Ecological Research Series
CHEMICAL CHARACTER
OF MODEL AEROSOLS
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional giouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are.
1 Environmental Health Effects Research
2 Environmental Protection Technology
3 Ecological Research
4 Environmental Monitoring
5 Socioeconomic Environmental Studies
This report has been assigned to the ECOLOGICAL RESEARCH series This series
describes research on the effects of pollution on humans, plant and animal
species, and materials Problems are assessed for their long- and short-term
influences. Investigations include formation, transport, and pathway studies to
determine the fate of pollutants and their effects This work provides the technical
basis for setting standards to minimize undesirable changes in living organisms
in the aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-76-085
July 1976
CHEMICAL CHARACTERIZATION
OF MODEL AEROSOLS
by
W. E. Schwartz, G. D. Mendenhall, P. W. Jones,
C. J. Riggle, A. P. Graffeo, and D. F. Miller
Battelle Memorial Institute
Columbus, Ohio 43201
Grant No. 801174
Project Officer
Ronald K. Patterson
Atmospheric Chemistry and Physics Division
Environmental Sciences Research 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
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences
Research Laboratory, U.S. Environmental Protection Agency, and approved
for publication. Approval does not signify that the contents 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 recommendation for use.
ii
-------
ABSTRACT
Model aerosols were generated from the individual hydrocarbons, toluene
and 1-heptene, by irradiation under simulated atmospheric conditions in
the presence of NO , or NO /S09. The reactions were carried out in a
X X £
3
17.3 m environmental chamber. The collected aerosols were subjected to
analysis by mass spectrometry and chromatographic techniques, both with
and without chemical derivatization. Polyfunctional oxidation products,
including quinones and carboxylic acid, were tentatively identified in the
toluene aerosol. The 1-heptene filtered aerosol was shown to contain con-
densation products from different 1-heptene molecules. Tentative identifi-
cation of a number of vapor-phase species was accomplished in both systems.
The health effects of the atmospheric oxidation products from hydrocarbons
is discussed.
iii
-------
CONTENTS
ABSTRACT iii
LIST OF FIGURES v
LIST OF TABLES vi
SECTIONS
I SUMMARY AND CONCLUSIONS 1
II INTRODUCTION 2
III BACKGROUND 3
IV METHODS AND PROCEDURES 7
A. Smog Chamber Characteristics 7
B. Smog-Chamber Cleanup 8
C. Aerosol Generation, Monitoring and Collection. . 9
D. Collection and Analysis of Organic Vapor Samples 11
E. Preparation of Methyl Esters with Diazomethane . 11
V RESULTS AND DISCUSSION 12
A. Organic Characterization of Model Aerosols ... 12
B. Gas Chromatographic Analysis of Model Aerosols . 17
C. High-Pressure Liquid Chromatography of
Derivatized Aerosols 28
D. Direct Mass Spectral Analysis of Aerosol
without Prior Chromatography 29
E. Infrared Spectrum of 1-Heptene Aerosol 36
F. Analysis of Vapor Phase Organic Species
Associated with Model Aerosols 38
G. Reaction Types and Mechanisms Inferred from
Tentatively Identified Reaction Products in
Model Aerosols 44
H. Health Effects 48
I. Conclusions 52
REFERENCES 53
APPENDIX A. METHANE IONIZATION SPECTRA OF COMPOUNDS OB-
SERVED ON THE GC-MS ANALYSIS (FIGURE 10) OF
TOLUENE AEROSOL, DERIVATIZED WITH BSTFA. . . 54
APPENDIX B. METHANE IONIZATION SPECTRA OF COMPOUNDS TENTA-
TIVELY IDENTIFIED ON THE GC-MS ANALYSIS
(FIGURE 15) OF VAPORS ASSOCIATED WITH TOLUENE/
NO /S00 AEROSOLS 61
x 2
APPENDIX C. METHANE IONIZATION SPECTRA OF COMPOUNDS OB-
SERVED ON THE GC-MS ANALYSIS (FIGURE 16) OF
VAPORS ASSOCIATED WITH l-HEPTENE/NOx/S02
AEROSOLS 73
-------
No.
FIGURES
Page
1 Smog Profile, Toluene/NO Aerosol 14
X
2 Smog Profile, Toluene/NO /SO Aerosol 15
X £.
3 Smog Profile, 1-Heptene/NO /S09 Aerosol 16
X ^
4 Gas Chromatograms of Methanol Extract of
Toluene Aerosol, after Treatment with Diazo-
methane (Toluene/NO ) 19
x
5 Gas Chromatograms of Methanol Extract of
Toluene Aerosol, after Treatment with Diazo-
methane (Toluene/NO /S0?) 20
X 6-
6 Gas Chromatograms of Methanol Extract of
Toluene Aerosol, after Treatment with Diazo-
methane (Toluene/NO /S02> exposed to H2SO,) 21
7 Gas Chromatograms of Methylene Chloride Extract
of Toluene Aerosol, after Treatment with Diazo-
methene (Toluene/JO ) 22
X
8 Gas Chromatograms of Methylene Chloride Extract
of Toluene Aerosol, after Treatment with Diazo-
methane (Toluene/NO /S0~) 23
X £r
9 Gas Chromatograms of Methylene Chloride Extract
of Toluene Aerosol, after Treatment with Diazo-
methane (Toluene/NO /S0?, exposed to H^SO.) 24
X £* £ Q
10 Reconstructed Gas Chromatograms of Toluene
Aerosols Derivatized with BSTFA 26
11 Electron Impact Mass Spectrum of 1-Heptene
Aerosol (Sample M599). Successive Spectra
Recorded as Probe Temperature Raised from 25 to 200 C 30-32
12 Chemical lonization Mass Spectrum of 1-Heptene
Aerosol (Sample M599). Successive Spectra
Recorded as Probe Temperature Raised from 25 to 200 C 33-35
13 Infrared Spectrum of Products in Kethylene Chloride. . 37
14 Methane lonization GC-MS Analysis of Vapor Phase
Organic Compounds Associated with Toluene/NO
Aerosol 41
15 Methane lonization GC-MS Analyais of Vapor Phase
Organic Compounds Associated with Toluene/NO /S02
Aerosol 42
16 Methane lonization GC-MS Analysis of Vapor Phase
Organic Compounds Associated with 1-Heptene/NO /S0~
Aerosol 43
v i
-------
TABLES
No. Page
1 Summary of Irradiation Conditions and Smog
Parameters 10
2 Solvent Extraction of Model Aerosol Products . 17
3 Sulfur Content of Aerosol Extractable Matter . 27
4 Toxicology of Oxidation Products Obtained
from Toluene and 1-Heptene in Smog Chamber Runs 49
5 Components of Embalming Fluid Preparations. . . 51
vii
-------
SECTION 1
SUMMARY AND CONCLUSIONS
This report constitutes the second year's study of the composition
of model aerosols from individual hydrocarbon precursors under simulated
atmospheric conditions.
In the previous year's work, aerosols derived from toluene, cyclo-
hexene, and a-pinene were examined. Because of the complexity of the products,
from a-pinene, research during the present year has concentrated on toluene
and the terminal olefin 1-heptene. This pair constitutes a representative
aromatic and aliphatic species of the sort found in polluted urban atmospheres.
Aerosols from the systems toluene/NO , toluene/NO /S0_, and 1-heptene
X. X £*
NO /SO were prepared, collected, and subjected to analysis by mass spectral
X £,
techniques, chemical derivatization, and gas and liquid chromatography.
The gas-phase organics associated with the aerosols were also studied.
Individual species identified tentatively in toluene aerosol include isomeric
hydroxybenzoic acids and hydroxy-p-benzoquinones. Among the gas-phase toluene
products were identified tentatively nitrotoluenes, nitrocresols, benzaldehyde,
and methyl-p-benzoquinone. The 1-heptene aerosol was examined by mass
spectromety; the results suggested that a number of the components are con-
densation products of species arising from different 1-heptene molecules.
n-Hexaldehyde was tentatively identified among the vapor-phase products from
1-heptene. The possible adverse health effects of aerosol and gas-phase
products are suggested to be those arising from a general degradative action
toward cellular proteins and amino acids.
-------
SECTION II
INTRODUCTION
This report describes research conducted on "Chemical
Characterization Model Aerosols" during the period June 11, 1974 to
November 10, 1975. This period constitutes the second project-year
under EPA Grant No. 801174. The report detailing results of the first
year's study, "Chemical Characterization of Model Aerosols" is avail-
able through the National Technical Information Service under NTIS
No. PB238557/3.
During the current project period organic analysis of
model and atmospheric aerosols was conducted. Model aerosols were
generated from individual hydrocarbon precursors under atmospheric
simulation in the Battelle-Columbus 17 m environmental chamber.
Under the heading below, Background, this year's aerosol
characterization studies are placed in perspective with respect
to results of the first year's study. Key differences in emphasis
and approach are described.
-------
SECTION III
BACKGROUND
During the first year of this program, research was limited
to characterization of model aerosols. Three systems were studied; a-
pinene/NO , toluene/NO , and 1-heptene/NO . Aerosol generated under
X X X.
atmospheric simulation was collected on glass fiber filters, and organic
matter was obtained by solvent extraction. In order to simplify pro-
duct identification, reaction mixtures were fractionated into defined
acid, neutral, and basic components. Products were then analyzed by
gas chromatography (GC) and gas chromatography combined with mass
spectrometry (GC-MS). Tentative identification of a variety of
interesting aerosol products was accomplished. The results of the
first year's study are best summarized by reference to these aerosol
products. In considering the structures shown below, it should be
emphasized that these are tentative identifications.
Cyclohexene aerosol.
COOH
TENTATIVE
0 0
CH3-CH=Ch-C -C -CH3 HO-CH2-CH2-CH=CH-CH=CH,
CrC_OEXENE
or
Or
-------
o-Pinene aerosol.
DOH
TENTATIVE
a-PINENE
COOH
CHO
Toluene aerosol.
COOH
COOH
"CLLlENE
CH2OH
-------
During the current year's study, we have moved to investigation of
simulations including both NO and S09; specifically, toluene/NO /S09
X * X £
and 1-heptene/NO /SO . In characterizing these systems, the scope of
X £-
organic analysis has also been broadened. During the first year's effort,
the research emphasized identification of aerosol products that had
not undergone extensive oxidation or degradation. Identification of
such products is profitable in that the structures can be more easily
related to the parent hydrocarbon. On the basis of this relationship
we may infer the types of reactions occuring under atmospheric simula-
tion. Indeed, in this report we present a discussion of the reaction
types and mechanisms suggested by the products identified during the
first year's study.
Although identification of the relatively less-degraded
aerosol products is helpful in elucidating the nature of aerosol re-
actions, we have found that such products constitute a rather small
fraction of the total organic constituency. Data concerning material
distribution was obtained during earlier work on extraction and
fractionation of aerosol products. It will be helpful to review this
data as well as aspects of the rationale and methods used during the
first year of the program.
Collected aerosol was subjected to Soxhlet extraction first
with methylene chloride and then with either acetone or methanol.
The sequential extraction procedure was used to obtain a measure of the
total extractable matter. The methylene chloride extracts were
expected to contain products less oxidized than those obtained in
methanol or acetone. Hence, consistent with the analytical emphasis
and rationale, the methylene chloride extractable matter was used for
subsequent fractionation and analysis. Fractionation was conducted
according to the scheme below.
Methylene Chloride Extractable Matter
\ 1
Water-Soluble Fraction Water-Insoluble Fraction
Acid Neutral Basic
Fraction Fraction Fraction
-------
The water-insoluble fraction can be expected to contain the
relatively less-oxidized, less-degraded components of the reaction
mixture. Hence, the water-insoluble material was used for detailed
analysis.
Further fractionation into acid, basic, and neutral components
was performed by partitioning a methylene chloride solution of the water-
insoluble fraction against dilute aqueous acid and then dilute aqueous
base. Subsequent analyses could then be conducted on mixtures partly
defined according to chemical class. For example, in analyzing a neutral
fraction, one might anticipate the presence of aldehydes, ketones, or
alcohols, but not carboxylic acids or phenols. This feature of the
fractionation approach proved extremely helpful in interpretation of mass
spectra of the reaction products.
Data concerning material distribution obtained during the
extraction and fractionation steps is shown in Tables B-l and B-2
(Ref. Report I, p. 26 and p. 29). From the data, it is clear that the
water-insoluble fraction of the methylene chloride extractable matter
represents only a small fraction of the total solvent extractable
matter. In view of this fact, and having achieved some success in
characterization of the relatively less-oxidized components of the
aerosol products, it was decided to direct future efforts toward a
more comprehensive analysis, specifically including the more oxidized
components of the reaction mixture.
This decision has had serious consequences as to the experi-
mental approach that may be applied. In particular, the water solu-
bility of the more oxidized products precluded use of the previously
applied acid/base/neutral fractionation procedure. Moreover, the
extreme polarity of such products severely limited the utility of gas
chromatographic analysis. Alternative approaches that have been investi-
gated during the current project period have included selective deri-
vatization followed by high pressuie liquid chromatography, and
direct examination of aerosols by mass spectral techniques without
prior chromatography. Research in these areas will be detailed in
sections below.
-------
Finally, mention should be made of an additional area in
which the research effort has been broadened. During the current
project period collection and analysis of gas-phase organics associated
with model aerosols has been conducted. Samples were collected by
drawing filtered air through adsorbant traps. Collected organics
were later desorbed by heating in a stream of carrier gas and analyzed
by GC and GC-MS.
Overall, the program has grown in both the types of aerosol
samples included for analysis, and the analytical methods brought to
bear on the complex problem of aerosol characterization.
SECTION IV
METHODS AND PROCEDURES
The methods of procedure are those described in the August,
1974, Annual Report on Chemical Characterization of Model Aerosols.
The old procedures are partially summarized below along with a des-
cription of the new methods.
A. jmog__Chamb_er_ _C_har_ac t er is tic s
The smog chamber utilized in these studies is approximately
2.4 m high, 4.9 m long, and 1.5 m wide, having a volume of approximately
3
17.3 m . The surface-to-volume ratio is 0.78. The inside surface
2 2
consists of 35.3 m of polished aluminum and 9.2 m of FEP-Teflon
windows (5 mil thick) through which the reaction mixture is irradiated
using an external bank of lamps. The lamp bank consists of 96
fluorescent "black" lamps and 15 fluorescent sunlamps. The spectral
distribution of the black lamps peaks in intensity at 370 nm; the sun-
lamps peak intensity occurs at 310 nm. Light intensity generated
corresponds to a k, of 0.45 min as determined by N0? photolysis and
o-nitrobenzaldehyde actinometry. This intensity is comparable to that
of noonday sun over the wavelength integral of N0~ absorption.
-------
Air supplied to tne chamber IF Laker1 in through a j C-r staci aior
a tnree-story building, and is passed througr a purification s\-sterr
including a permanganate filter bed, a charcoal filter system, an
absolute filter, and a humidification unit. After purification,
background total-hydrocarbon is generally 1-3 ppmC with the majority
of this being methane. A trace of background ethane is occasionally,
observed by gas chromatography. Olefins and aromatics have
always been below gas chromatographic detection limits. Sulfur
dioxide is always oelow the detection limit of a flame photometric
detection ('"10 ppb) and carbon monoxide background levels are
generally between 1 and 3 ppm as measured by nondispersive infrared
spectroscopy. Light scattering readings are essentially the same
as for clean air (Raleigh scattering) and the condensation nuclei
count is generally less than 200 particles per cubic centimeter.
B. Smog-Chamber Cleanup
Prior to each series of experiments with a different model hydro-
carbon, it was necessary to thoroughly clean the chamber's surface
to prevent cross contamination of aerosol products. Such cross
contamination had been observed during previous studies . Cleanup
was accomplished by washing down chamber surfaces with a spray of
4:1 isopropanol:water, Prewashed cloths were used to scrub the sur-
faces and surgical scrub suits were worn to prevent contamination
from street clothing. Throughout the procedure, the chamber was main-
_ i
tained on purge (8500 1 min """) . Nevertheless, the high concentration
of solvent vapor that develops, requires that the chamber be sealed
and that the worker be provided with an auxiliary breathing apparatus.
After cleanup, the chamber was purged overnight and was then per-
mitted to stand with 1 ppm ozone for several hours. Before genera-
tion of aerosol for analytical use, a "conditioning" run was conducted
using the model system to be studied.
-------
C. Aerosol Generation, Monitoring, and Collection
Detailed data concerning, reactant concentrations and conditions are
shown in Table 1. After initial concentrations were estab-
lished, the mixing fan was turned off and irradiation was begun.
Parameters monitored throughout the course of the experiment included
NO and NOi concentration using a continuous Saltzman analyzer, total
hydrocarbon concentration using a flame ionization detector, ozone
concentration using a chemiluminescence analyzer, and light scatter-
ing using an integrating nephelometer.
Ir general, irradiation was continued until light scattering measure-
ments indicated that maximum aerosol growth had occurred. After
this maximum was observed, irradiation was halted and the aerosol
was collected by evacuating the chamber contents through tared
10-cirdiameter glass-fiber filters (Gelroan Type A) using a high-
volume type sampler. Typically, about 11 m of chamber volume
was sampled over a 30-ir.inute period. Before use, the filters were
washed repeatedly, first with distilled-in-glass methanol, and then
with methylene chloride. They were then equilibrated at 50 percent
R.H. ana tared. After collection, the filters were reequi1ibrated,
weighed, and placed ir. glass containers for storage at -60 C. After
a series of runs was completed, collected aerosol was suDiected to
Soxhlet extraction, as described in the previous report.
-------
TABLE 1. SUMMARY OF IRRADIATION CONDITIONS AND SMOG PARAMETERS
(a)
Aerosol System
Number of runs
Irradiation period, min.
Relative Humidity
Initial, percent
Final, percent
Hydrocarbon
Mass irradiated
Initial concentration, ppm C
Final concertration, ppm c(b)
Nitric oxide
Initial concentration, ppm
Nitrogen dioxide
Initial concentration, ppm
Maximum concentration, ppm
Sulfur dioxide
Initial concentration, ppm
Final concentration, ppm
Ozone
Maximum concentration, ppm
tmax, min.
Light scattering
Maximum, 10~^m~-'-
Total aerosol mass collected, mg
Toluene/NO/
NO 2
6
240
50
73
45
0.97
0.94
1.2
0
0
1.1
198
76
11.5(0
Toluene/NO/
N02/S02
4
240
50
71
41
0.89
1.09
1,5
1.08
0.82
0.32
200
120
16.3
1-Heptene/NO/
12
265
50
38
69
6
1.25
1.26
1.7
1.0
0.44
0.42
200
188
38.6
(a) Data shown are averages for the series of runs, except for values of total
aerosol mass collected, which represent the total mass collected during an
entire series of runs.
(b) Includes contribution from products formation.
(c) After treatment with H^SO, aerosol the average weight was 12.9 mg.
10
-------
D. Collection and Analysis of Organic Vapor Samples
Organic vapor samples associated with model aerosols were
collected and subjected to GC-MS analysis. The filtered (pre-extracted
tissue quartz) gas sample was drawn through an ambient temperature
chromosorb 102 chromatographic trap (15 cm long, 3 cm diameter) at
approximately 0.5 cfm by a Cast 0522 pump. Prior to analysis, the
traps were sealed with Teflon fittings and protected from light.
The collected samples were subjected to GC-MS analysis by
thermally desorbing the collected species at 180°C using a helium
flow of approximately 500 ml min . The desorbed material was pre-
liminarily transferred to a cooled (-196°C) stainless steel sample
loop. Chromatographic injection of the collected sample was subse-
quently achieved by diverting the GC carrier gas through the cooled
sample loop, and simultaneously flash-heating the loop to 250°C,
(1) *
which immediately sweeps the sample to the head of the GC column.
The chromatographic column is routinely interfaced with either a Finni-
gan 1050 (electron impact) or a Finnigan 3200 (chemical ionization)
mass spectrometer; interpretation of electron impact mass spectra is
facilitated with a mass spectral matching routine with a data bank of
27,000 spectra.
E. Preparation of Methyl Esters with Diazomethane
An ether solution of diazomethane was prepared from Diazala ,
a diazomethane precursor, according to directions given by the supplier
(Aldrich). About 1 jog of aerosol, prepared by evaporation of the re-
quisite amount of solution, was dissolved in 1 ml of diazomethane solu-
tion and allowed to stand 2-3 hours. The solution was concentrated
to a volume of about 0.1 ml. Portions of the concentrate were injected
into the gas chromatograph for examination.
^References are listed on page 53.
11
-------
SECTION V
RESULTS AND DISCUSSION
A. Organic Characterization of Model Aerosols
During the current year's study, the following model aerosol
systems have been investigated:
o Toluene/NO
X
o Toluene/NO
^
X
o 1-Heptene/NO /SO
X jL
Typical aerosol profiles are shown in Figures 1-3. Irradia-
tion conditions and smog manifestations characterizing the aerosols
are further described in Table 1. The values shown in the table rep-
resent averages for a series of runs. There was very little run-to-
run variation in the parameters; it is considered that the magnitude
of such variation was insufficient to be of consequence to the com-
posite aerosol composition.
In addition to generation of the above aerosols, an addi-
tional experiment was conducted to determine if a 7 organo-sulfur
products are formed after collection of the aerosol by reaction of
sulfuric acid with organics residing on the filter. In this experi-
ment, a series of five toluene/NO aerosols were generated under con-
X
ditions described in Table 1. Collected aerosol from four such runs
was exposed to sulfuric acid aerosol; one toluene/NO sample was re-
X
tained as a control.
Sulfuric acid aerosol was generated by evaporating a quantity
of 98 percent sulfuric acid into the chamber which had been previously
humidified to ~~50 percent RH. Upon reaching the saturation concentra-
_3
tion of sulfuric acid under these conditions ("^10 ppb), nucleation
commenced. Aerosol concentration and growth were monitored with a con-
densation nuclei counter (Environment-One Inc.) and an electrical par-
ticle-mobility analyzer (Thermo Systems, Inc. ). Aerosol generation was
3 3
terminated upon reaching a volume concentration of^-700 yrn /cm . This
concentration is approximately equivalent to that expected upon irradi-
ation of 1 ppm S0_ in a typical smog environment. The mean particle
diameter in the volumetric distribution was ^0.2 ym.
To further simulate experimental conditions employed in
aerosol generation from toluene/NO /SO., 1 ppm SO,, was included during
X
-------
were exposed to both sulfuric acid and S0_. Exposure was conducted
as follows. After the sulfuric acid aerosol was fully developed, as
described above, portions of the chamber contents were exhausted
through filters containing previously collected toluene/NO aerosol.
X
Each filter was exposed individually to 15 percent of the chamber
contents over a period of 15 minutes.
13
-------
2.0
1.8
1.6
e
o.
Q.
c
o
1.2
£ 1.0
c
0)
o
§ 0.8
(A
3
0.4
0.2
0
Initial Cone.
Toluene
NO
N02
10 ppm
0.97 ppm
0.94 ppm
60 120 180 240
Irradiation Time, minutes
200
180
160 _
'e
140 ^
X
120 "5
o
ioot
c.
o
o
en
80 2>
60
40
20
300
FIGURE 1. SMOG PROFILE, TOLUENE/NO AEROSOL
X
14
-------
200
Toluene 10 ppm
NO 0.89 ppm
N02 1.09 ppm
1.08 ppm
0
60
120 180 240
Irradiation Time, minutes
FIGURE 2. SMOG PROFILE, TOLUENE/NO /S00 AEROSOL
x 2.
15
-------
200
l-Heptene 10 ppm
NO 1.25 ppm
1-26 ppm
SO? I Ppm
60 120 180 240
Irradiation Time, minutes
300
FIGURE 3. SMOG PROFILE, 1-HEPTENE/NO /SO AEROSOL
X Z.
16
-------
Aerosol products were collected by filtration as described
in Methods of Procedure and were then subjected to sequential Soxhlet
extraction first with methylene chloride, and then methanol. Values
for weight percent solvent extractable are shown in Table 2. Gas-
phase organics associated with the model aerosols were collected at
the same time as filtration samples by passing filtered chamber con-
tents through Chromsorb 102 adsorbent traps. Aerosol extracts and
gas phase organics were analyzed in detail as described under the
headings below.
TABLE 2. SOLVENT EXTRACTION OF MODEL AEROSOL PRODUCTS
Aerosol System
Weight-Percent Solvent Extractable
First Extraction, Second Extraction,
Methylene Chloride Methanol
Toluene/NO aerosol
X
Toluene/NO aerosol exposed
to sulfuric acid aerosol
Toluene/NO /S09 aerosol
X £,
1-Heptene/NO /SO aerosol
X £*
18
29
21
22
63
55
75
49
B. Gas Chromatographic Analysis of Model Aerosols
As described in the Background section of this report, the
analytical objectives of this year's research emphasize characteriza-
tion of the abundant, highly-oxidized components of model aerosol.
These polar, water-soluble organics cannot be treated using the pre-
viously applied fractionation procedure. Fractionation had proven
useful in that particular compound types were isolated in defined
fractions. An alternative strategy based on selective derivatization
was attempted to distinguish between different compound types. Initial
17
-------
gas chromatographic analyses were performed on unfractionated, underi-
vated aerosol extracts. It was reasoned that free carboxylic acids,
phenols, and other highly polar compounds such as polyols would be
entirely retained on the column; only neutral products of moderate
polarity would be eluted. A short pre-column can be used to prevent
buildup of retained material on the analytical column. The strategy
then called for treatment of parallel portions of aerosol extract with:
(a) Diazomethane
(b) 0,N-Bis(trimethylsilyl)trifluoroacetamide (BSTFA)
Treatment with diazomethane yields relatively volatile methyl
esters from carboxylic and sulfonic acids, and methyl-aryl ethers from
phenols. BSTFA reacts with acids and alcohols to form trimethylsilyl
(TMS) esters and ethers. Derivatization was first attempted with the
aerosol from the toluene runs. Analyses of underivatized methylene
chloride and methanol extracts were performed with a 10 ft x 2 mm
column of 3-percent OV-17 and a 14 ft x 2 mm column of 3-percent Dexil.
The temperature was programmed from 100°C to 275°C at 6 degrees/min.
Dual flame-ionization and flame-photometric (sulfur-specific) detectors
were employed. The chromatograms obtained showed only a few poorly
resolved peaks of negligible intensity. Total detector response sug-
gests that no more than 1 percent of injected matter was eluted
(injected samples = 50 yg each). The results suggest that even the
non-acid constituents of the aerosol consisted of highly polar products.
Figures 4-9 show several chromatograms obtained after the
toluene aerosol extracts were treated with diazomethane. A small
increase in total detectoL response was noted compared to the underi-
vated samples. Nevertheless, the overall response account for less
than 10-percent of injected methylene chloride extractable matter and less
than 5-percent of injected methanol extractable matter (50 yg injec-
tions). The components showing flame-photometric (sulfur) response
have been shown in most cases to correspond to impurities in several
batches of diazomethane reagent. In the case of the methanol extract
of toluene/NO /S0_, two sulfur-containing components appear which may
X «£
be genuine reaction products. They do not appear in toluene/NO aero-
X
sol, or toluene/NO aerosol exposed to H?SO. aerosol. They account for
18
-------
c
'i
o
ro
C\J
CJ
_ 00
w
is
H
W
a
o
H
M
DS
H
W
g
W
Pi
H
Pi
W u
H
fn O
-------
c
i
O
ro
CJ
C\J
- ro -
w
53
O
N
LO
w
O
HOI33I3Q
-------
-8-
(\J
CVJ
OJ'
-00-
r
OJ-
u
-ro -
w
tu
NOIIVZINOI-aWVli
21
-------
in
O)
CM
c\i
- CM -
_ CO _
_ in _
- cr> -
-ro-
O
r^
a
HOID3I3Q
NOIIVZINOI-aHVld
22
-------
in
o>
3
C
i
8
r-
C\J
CM
_ CO -
-ro-
u
td O
I
O
I
w
J
H
a
o a,
o w
H S
^ <
2 33
O H
oi w
N:
CO <
NOIIVZINOI-3HV1J
DIHX3WOIOHd-3WVld
-------
o _
ro
CM
CM
33
H
w
g
H
<
W
ed
H
ft* O
-------
less than 1-percent of the injected sample, however.
Comparison of chromatograms for diazomethane treated extracts
of the three toluene aerosols suggest that exposure of collected
toluene/NO aerosol products to sulfuric acid aerosol does not lead
to formation of new products (detectable by this approach). In the
diazomethane-treated, methylene chloride extractable matter from toluene/
x 2 aerosol there is only one moderate peak not observed in the cor-
responding sample of toluene/NO aerosol (peak at 25 minutes).
X
As described above, the next step in the analytical approach
called for derivatization with BSTFA. It was reasoned, that if the
uneluted material consisted of non-volatile alcohols or polyols, BSTFA
should generate volatile IMS ethers. Gas chromatographic analysis of
toluene aerosol products indicates that a significantly larger fraction
of subjected material is being eluted after BSTFA treatment (^30-per-
cent eluted). Thus, these samples were analyzed by GC-MS. Reconstructed
gas chromatograms are shown in Figure 10.
The CI-MS spectra of the BSTFA-derivatized samples are shown
in Appendix A. Most of the spectra, even after background subtraction,
showed numerous peaks, and only the relatively simple spectra that
appeared to contain only one major component have been included. In
those cases the parent ions were identified and structures were assigned
by selection of reasonable structures with the correct molecular
weights. The compounds tentatively identified are shown in the figure.
We emphasize the tentative nature of these assignments.
The methylene chloride extract of 1-heptene aerosol was simi-
larly treated with BSTFA and analyzed gas chromatographically. Total
detector response corresponded to only 5-percent of injected material.
Thus, the sample was not analyzed by GC-MS.
For both 1-heptene and toluene aerosols analysis by flame
photometric gas chromatography revealed negligible sulfur response
above blank. In view of the low sulfur response for all aerosol samples,
extracts were subjected to analyses for total sulfur. The data are
shown in Table 3. Samples taken for analysis varied somewhat in size,
leading to some variation in the lower limit of detection. Thus, any
25
-------
d P flj 11 (j LU \/
26
-------
values below 1-percent sulfur should be viewed as approximately equi-
valent. The methanol extracts for toluene/NO /SO aerosol and toluene/
X £*
NOX exposed to sulfuric acid aerosol show significantly higher concen-
trations of sulfur, which probably arise from inorganic sulfate sol-
uble in the methanol. This was indicated by a sulfate analysis of
one collected aerosol from a 1-heptene/NO /SO, run, which showed 40%
, X ^
of SO" by weight of total particulates on the filter. These data in-
dicate that the low concentrations of organo-sulfur compounds observed
by flame-photometric gas chromatography was not due to retention of
abundant organo-sulfur compounds on the column.
TABLE 3 . SULFUR CONTENT OF AEROSOL EXTRACTABLE MATTER
Weight-Percent Sulfur
Methylene Chloride Methanol
Aerosol System Extractable Matter Extractable Matter
Toluene/NO aerosol
X
Toluene/NO /SO aerosol
Toluene/NO aerosol exposed
<0.5
<0.8
<0.8
<0.3
3.9
1.2
to sulfuric acid aerosol
1-Heptene/NO /S09 aerosol 0.1 17.9
X ^
One additional derivatization approach was attempted to
facilitate gas chromatographic analysis. It has been observed that low
molecular weight aldehydes and ketones tend to polymerize at the ele-
vated temperatures of gas chromatographic analysis. In order to deter-
mine whether such compounds correspond to the uneluted fraction of
aerosol products, extractable matter was treated with ethanedithiol
to stabilize such carbonyls as dithiolane derivatives.
\ \ RF, ^ /* CH2
C ' 0 + CH2 ^ -^ C
/ 1 / S CH2
CH2
HS
27
-------
Preparation of dithiolane derivatives is additionally useful
in that the derivatives give both flame-ionization and flame-photometric
response. In these analyses, a rather high blank was obtained. Never-
theless, for both toluene and 1-heptene aerosols no significant compon-
ents above blank were observed. Based on the sample size taken for
derivatization, injected material should correspond to 20 yg. Virtually
complete loss of sample to the aqueous phase during workup could account
for chromatographic results observed.
Results of the series of gas chromatographic analyses suggests
that the aerosol products consist principally of highly oxidized, highly
polar compounds. These have not been rendered volatile by generally
employed derivatization techniques. In view of the failure of this
approach, two different techniques have been investigated to determine
feasibility. These are:
(1) High-pressure liquid chromatography followed
by mass spectrometric analyses of isolated products.
(2) Direct mass spectral analysis of aerosol without
chromatography.
These studies are described under subsequent headings.
C. High-Pressure Liquid Chromatography of Derivatized Aerosols.
Since the functional groups expected in the components of the
aerosol do not absorb light strongly, the UV-visible detection system
currently in use with the HPLC instrument was not suitable for analysis
of small amounts of atmospheric aerosols. Simple derivatives of the
components are suitable, however, if the derivatizing reagent contains
an absorbing group.
Prliminary experiments were carried out with a sample of the
1-heptene aerosol. The derivatizing reagent was a,p-dibromoaceto-
phenone, which reacts with acids as follows:
R COO
-I- Br
CH2-02CR
-------
This reagent was repeatedly reacted with the methanol extract of the 1-
heptene aerosol and chromatographed using HPLC with UV detection. A
large peak corresponding to the derivatizing reagent was always observed.
Other peaks were also observed but did not appear on duplicate runs.
This may be due to impurities in the derivatizing reagent as well as
to interferences by some component of the sample. Due to the unknown
nature of the sample and the experimental nature of derivatization re-
actions in HPLC, more development in optimizing derivatization procedures
is needed. These include methods for derivatization of alcohol, ketone,
as well as acid functionalities, and of compounds containing multiple
functionalities.
D. Direct Mass Spectral Analysis of Aerosol without
Prior Chromatography.
Since the bulk of the 1-heptene aerosol products did not pass
through a chromatographic column under our conditions, even when deri-
vatized, we examined a sample of 1-heptene aerosol that was evaporated
from a solution placed directly on the probe of the mass spectrometer.
The probe was placed in the instrument and heated while successive
spectra were run. The procedure was carried out twice, first with con-
ventional electron impact (El) and then with chemical ionization (CI)
by proton transfer from C,H n ion. Under these conditions, we expect
low-molecular weight species to volatilize first, followed by higher
molecular weight species with lower vapor pressures. The resulting
spectra (Figures 11 and 12) are, of course, complicated because the
method does not cleanly separate related compounds, and because a signal
at any given m/e value can result from several isomeric compounds pre-
sent in the sample.
The El spectra (Figure 11) show fragments up to about 260,
which corresponds approximately to two heptene molecules plus four
oxygens (C H 0 ) = 260. The CI spectra (Figure 12) show more dis-
tinctive features, in part because CI usually gives simpler fragmenta-
tion patterns and larger parent ions. Spectrum No. 4 shows two groups
of peaks with a minimum around m/e 180, while spectrum No. 7 has a
second minimum around m/e 300. Each group of peaks consists of triplets
of quartets separated by 14-16 mass units. This mass difference is equal
29
-------
90-
>- 80-
i
co 70-
JJJ 60-
550-
^40-
,r 30-
'CE
_J
UJ
20-
10-
04
20
CL
cc:
40
,imtL.^.imtliH--UmllU-,..tMmHJlt-i y"1)" . t"h' , i "'!* | i1'. i 'i- i -i 1 ) i j- -i | i i i'
60 6'0 100 20 4'0 60 80 200 20
M/E
40
SPECTRUM NO. 2
uu-
90-
80-
70-
60-
40-
30-
10-
n
i
ii
L
i,
i!'
Illnil
1 1 .IJlilL Jliii, Ji ,1,1,11,1 ,1,. .,!,., ,i,.. ,li ill .1, ... .
I UU UU
-------
100
90
>~ 80-
co 70^
50-j
40^
30-1
C£ ..
20 40 60 80 lOO^'O^O'e'O'S'O 200'20 40 60 80 300 20 40 60 80 400 20
M/E
SPECTRUM NO. 4
>: eo
ri /,
CO /U
LU A;
30
CL
i >-,-
LU -U
ct: . -
LJOliL^
[ llt|(»j)H| iilllllli. .lilllltlln I "IHHyiiii lill1'11!'!-- MP'lf-1 .U.'ll'. i.ililHI. .ljUlllt. '''Mi I1'1!1 1. II I . i1'1!' 'r'l .Ml i.'vni i. Irlii ," , .'. ilii. i /'....'... 1 .... i ..
0 ' 20 ' 40 6D 80 130 20 40 60 ' 8'0 '200' 20 ' 40 ' 6'0 ' SO '300' 2'Q ' 4'0 ' 5'0 8'0 400' 20
M/E
SPECTRUM NO. 5
FIGURE 11. ELECTRON IMPACT MASS SPECTRUM OF 1-HEPTENE AEROSOL (SAMPLE M599).
SUCCESSIVE SPECTRA RECORDED AS PROBE TEMPERATURE RAISED FROM 25
TO 200 C.
31
-------
CE
I
LU
lOOn
90-
80
1 60-
i 50-
1 40-
' 30-
-i
20-
10-
OH
j.i|l,..lii,iJlili ,iil., ,i,.
(lilliilin II .tiViH"' ji "l1 ''i li" \' ' 1 | --'T"1", "^"'Y T~T"1 't",i-i.'jfa' i.uMii.......^.!. ..u^.jJU, ^.u ...I.M...,.iJ^...P-....^, , .. i, .j-.f*^..p^...,
4'0 ' 6'0 ' 8'0 200' 2'0 ' 4'd ' 6'0 8'0 300 20 40 60 80 400 20 4'0
M/E
SPECTRUM NO. 7
2'ti ' 4D
100-j
90-
>: 80-
^o 70-
z:
LU 60-
250:
LL; 40-
^ 30-
cr
H 2CH
^ i
101
2'0 ' 4'0 ' 6'0 ' 8'0 'lOO 20 40 60 80 200 2D 40 bd 8'Cl 3'od'"2'0"' 4'0 ' 6'd"'' 3U 400'
M/E
SPECTRUM NO. 8
FIGURE 11. ELECTRON IMPACT MASS SPECTRUM OF 1-HEPTENE AEROSOL (SAMPLE M599)
SUCCESSIVE SPECTRA RECORDED AS PROBE TEMPERATURE RAISED FROM 25
TO 200 C.
32
-------
lOOn
90-
>: so-
co 70-
|^ 60-
2 50-
^J 40-
i i
h- 30-
cr
td20;
^ 10-
n.
,11
,. U.L1
Jj
.
h i it
.
|
lHiLlJj JL,|lL,dI,l|L lliiL 1 ,.,f ,
M/E
SPECTRUM No. 1
20 40 60 80
4'0 ' 6'0 80 400 2'0 ' 4'0 ' 60
SPECTRU1"! NO. 4
FIGURE 12. CHEMICAL IONIZATION MASS SPECTRUM OF 1-HEPTENE AEROSOL (SAMPLE M599)
SUCCESSIVE SPECTRA RECORDED AS PROBE TEMPERATURE RAISED FROM 25 TO
200 C.
33
-------
.11 I
0 2'0 <0 60 80 200 20 40 60 80 300' 2'ti ' 4'0 ' 60 80 400 20 40 60 8'0 500
M/E
SPECTRUM NO. 7
100
90
^ 80
CO 70
^ 60
iso
Uj 40
Hi
MMM^MMimAJt
300
20 40 60 80 100 20 40 60 80 200 20 40 60 80 300 20 40 f'J 9G 400' 20 ' 4ii' 60 ' 8'0 500' 20"' 4u ' 60 ' 8'0 '5(50'
M/E
SPECTRUM NO. 10
FIGURE 12. CHEMICAL IONIZATION MASS SPECTRUM OF 1-HEPTENE AEROSOL (SAMPLE M599)
SUCCESSIVE SPECTRA RECORDED AS PROBE TEMPERATURE RAISED FROM 25 TO
200 C.
34
-------
100-]
90-
>: so-
u^ 10-
~z_
^ 60-
550-
UJ 40-
~ 30-
LU
20-
10-
0
ii jL(iiiLUii,,i!iik,iiiiJk
ill, .,
2'0 ' 4'0 ' 6'0 ' 8'0 '100' 2'0 ' 4'0 ' 6'0 ' 8'0 200 Z'O 4'0 ' 6'0 ' 8'0 '300' 2'0 ' 4'0 ' 6'0 ' 00 400' 20 ' 4'0 ' 6'0 ' 8'0 500' 2'0 ' 4'0 ' 6'0 ' 8'0
M/E
SPECTRUM NO. 13
FIGURE 12. CHEMICAL IONIZATION MASS SPECTRUM OF 1-HEPTENE AEROSOL (SAMPLE M599).
SUCCESSIVE SPECTRA RECORDED AS PROBE TEMPERATURE RAISED FROM 25 TO
200 C.
35
-------
to one oxygen atom, a methylene unit, or to one oxygen atom less two
hydrogen atoms. It is reasonable that components in the aerosol mix-
ture would differ in this way, since photodegradation of organics under
smog-chamber conditions involves oxidative removal of H-atoms, chain
scission reactions, and incorporation of atmospheric oxygen.
The signals with the highest m/e appear in No. 10 with m/e
436. It is not known whether compounds of higher molecular weight were
absent from the sample or simply were not sufficiently volatile to
appear on our spectra. Another possible explanation for their absence
is that such high-molecular weight species were not removed from the
filter by our extraction procedure. To discount this alternative, we
examined a filter sample containing 1-heptene aerosol in the mass
spectrometer probe. Although signals with m/e up to several hundred
were observed, the spectrum was very weak and could not be ascribed
with certainty to aerosol alone.
To our knowledge, the presence of polymeric substances in
hydrocarbon aerosols has not been demonstrated previously, although
there has been some discussion in this regard.
Further work will be necessary to determine the nature of the
cross-links in the polymers, which will provide a clue to the mechanism
for formation of the aerosol.
E. Infrared Spectrum of 1-Heptene Aerosol.
The infrared spectrum of the methylene chloride extract of
the aerosol from the 1-heptene/NO /S0_ experiment was recorded as a
film on sodium chloride plates. The spectrum is shown in Figure 13,
and reveals strong absorption in the regions corresponding to frequencies
where hydroxy, carbonyl groups, and C-0 single bonds absorb. The only
distinctive features are strong absorptions at 1620 and 1280 m which
correspond to strong bonds associated with organic nitrates (general
structure RON09). Such compounds have been postulated to be present
(2)
in urban atmospheric aerosols. Although this result was interesting,
36
-------
w
p
M
Pi
O
£
O
w
a
H
E-i
W
H
O
3
O
O
Pi
pu
w
AJ
en
O
w
pi
fn
-------
we have not recorded other infrared spectra of our aerosols because the
general features usually are not very informative. Similar spectra have
been described in detail in an earlier report on model aerosols.
F. Analysis of Vapor Phase Organic Species
Associated with Model Aerosols
Vapor phase species were collected using Chromosorb 102 chromato-
graphic traps, following filtration by a pre-extracted quartz fiber
filter. Approximately 300 liters of gas were sampled from the smog-
chamber in each instance. The collected samples were recovered by
thermal desorption, followed by freeze-out in a stainless steel sample
loop; chromatographic injection was subsequently achieved through flash-
heating the loop to 250 C and diverting the GC carrier gas through it,
as described earlier. The three model aerosol systems chosen for this
study were toluene/NO , toluane/NO /SO-, and 1-heptene/NO /S09/N0 . GC
X X £. X t- £
separation was achieved using a Silar SCP chromatographic column; GC-MS
analysis was carried out by methane chemical ionization.
1. Analysis of Volatile Species Associated with Toluene Aerosol. Two
experiments were carried out, one utilizing toluene/NO /S09, and the
X /.
other in the absence of S0_. The objectives were to determine what
volatile products were formed, and how these differed from the species
associated with filterable particulate, and also to determine whether
sulfur-containing species were associated with the more volatile aerosol
products.
Striking differences from the previous analysis of the par-
ticulate aerosol material were immediately evident in both experiments.
The products observed had undergone little or no oxidative degradation,
and a considerable degree of ring substitution had taken place. Rela-
tively minor differences were observed between the two experiments,
and no organic sulfur-containing compounds were detected. All major
products were common to both experiments, except one, and eight of the
nine major products contained a nitro group.
As might be expected on the basis of free radical oxidation,
benzaldehyde, phenol, and three isomers of nitrotoluene have been
38
-------
tentatively identified on the basis of their methane ionization mass
spectra. Additionally, one isomer of nitrophenol has been tentatively
identified in each reaction system. We are presently unable to assign
a structure to three further products which are common to each experi-
ment, these compounds most probably have molecular weights of 122, 112,
and 90. One major peak which appears to be a polar compound and to
have a molecular weight of 122 was found only in the toluene/NO SO-
X ^
system, although this material does not appear to contain sulfur. The
mass spectrum of this latter compound is consistent with methyl-p-ben-
zoquinone, which will be identified conclusively when an authentic
sample is received for comparison.
While four isomeric nitrocresols were found in both of these
systems, only two were previously found in filterable particulate from
a similar toluene/NO system^' However, the vapor sampling/analysis
X
system permits significantly higher sensitivity which could account
for this apparent discrepancy, since two of the isomers were of rela-
tively minor concentration. Surprisingly, nitrotoluene was not pre-
viously observed in filtered aerosol particulate, yet all three possible
isomers were observed in this study. We are unable to explain the
presence of styrene and ethyl styrene, other than to attribute them to
contamination or impurities in the original toluene.
In summary, the following species have been determined on
both toluene/NO and toluene/NO S0? systems. No sulfur organics were
X X £-
detected.
CHO
-------
The reconstructed gas chromatograms for the respective GC-MS analysis
are shown in Figures 14 and 15; tentatively identified compounds are in-
dicated on the figures, and the mass spectra for peaks in the toluene/NO /
X
SO experiment are included in Appendix B.
2. Analysis of Volatile Species Associated with 1-Heptene Aerosol. Gas
chromatographic analysis of the organic vapors associated with 1-heptene/
NO /S09 aerosol provides a remarkably simple well resolved chromatogram.
X ^
Unfortuantely, the subsequent GC-MS analysis of this sample by methane
ionization has presented some difficulty. The six major products give
high quality spectra, but since they are all presumably aliphatic, the
spectra leave oper. a very wide range of possibilities with regard to
interpretation. The spectral assignments indicated on Figure 16 must
at this stage be regarded as speculative, although they are not incon-
sistent with the mass spectra of the major peaks which are presented in
Appendix C.
40
-------
X
O
LJ
2
UJ
te
IS
w
M
O
GO
to
u
M
a
PH
o
to
M
CO
00
S
a o
O C/3
M O
E-i BJ
<: w
SI -^
o o
W
W
O
H
o
001
-------
TOLUENE/NOX/S02
VAPORS
13
23 33 13 S3
HTBER
69 70 68 S3 103 113 129 139 119 ISO 163 173 180 133 203 2SO 223 Z33 2tO 2SS
FIGURE 15 METHANE IONIZATION GC-MS ANALYSIS OF VAPOR PHASE ORGANIC'
COMPOUNDS ASSOCIATED WITH TOLUENE/NO /S09 AEROSOL
X £
42
-------
CO
\
x
O
00
O
O
u
O
Pi
O
w
00
fe W
O
-------
G. Reaction Types and Mechanisms Infrared from Tentatively Identified
Reaction Products in Model Aerosols
On the basis of reactions which are known to occur during
smog formation, it is possible to write reasonable mechanisms leading
to the products that we have tentatively identified from different aero-
sol precursors (see Background section). Some mechanisms we favor are
as follows.
-------
Cyclohexene products
OH
o2
OH
OOO- NO
OH^
CHO
P«
OH
N0
ix^/CHOH
00-
2-
k ./CHOH
^^ I
0-
CCHO 02a>
0-
N0
2ON02
CH2OH
COOH
RCHO
(-RCOOH)
45
-------
a-Pinene products
CH,
0-0
6
,0
COOH
2.
H0«
OH
00-
H
NO
0«
H
OH
CHO °2
0
CHO
OX
3.
CHO
0 °2
CO
0 NO
-C02
5-00.
0
0 °2
o 3
,CH20«
0
CHO
46
-------
Toluene products
CH
HO
N02
2. O
H0«or
hv/02
,0.
CHOH
CH3
CH3
02
0
00»
H
0
NO
CH:
H
CH3 CH3
CHO Og r^^CHO NO
CH3
CHO CHO
H
H
47
-------
The above reactions are rather obvious ones, and we have not
tried to include every plausible route to each compound. The most in-
teresting feature in the products, from a mechanistic point of view,
are the large number of terminal alcohol groups. These represent a
relatively reduced functionality, and it is difficult to write many
homogeneous gas-phase mechanisms from the starting materials we studied,
which give alcohols as products. We have accordingly considered many
redox reactions in our mechanisms involving aldehyde disproportionations:
RCHO + R'CHO + RCH OH + R'COOH.
This type of reaction is given as "ox" in the mechanisms, and
presumably would take place within the aerosol droplets. If the reduc-
ing aldehyde is formaldehyde, the product is formic acid. Both of the
latter compounds are known constituents of atmospheric smog.
The ring-operated products from toluene can be explained in a
simple fashion if we invoke 3-scission in alkoxy radicals, which we postu-
late as intermediate species. Nitration and hydroxylation of toluene
probably occur by a free-radical mechanism as depicted above, followed
by rearrangements of Che radical adducts that probably take place in
the condensed phase. On the other hand, cresols (hydroxytoluenes) are
quite reactive toward nitration in solution, and it is therefore also
reasonable that the nitrocresols may be formed from nitric acid and
cresols within the aerosol drops or at surfaces, rather than as shown
above.
It should be emphasized that neither the mechanisms nor the
nature of the products are firmly established. Many products can be
generated on paper by slight variations of the above reaction paths,
that are not observed (or have not been identified) by our analytical
procedures. Future work will render the nature of the reactions more
certain.
H. Health Effects
The toxic doses and safety limits for concentrations of the
compounds tentatively identified in the vapor and particulate phases in
this work are given in Table 4. The compounds that appear in brackets
48
-------
TABLE 4. TOXICOLOGY OF COMPONENTS IN TOLUENE
AND 1-HEPTENE-DERIVED AEROSOLS
Compound Toxicology
4-Nitro-m-cresol ipr-mus LDLo:500 mg/kg
2-Nitro-p-c.resol orl-rat LD50:3360 mg/kg
m-,o-, and p-Nitrotoluene USOS-air: TWA 5 ppm (skin)
[Nitroglycerin USOS 2 mg/m3 (0.22 ppm) (skin)]
[p-Benzoquinone USOS air: TWA 0.1 ppm]
2-Methyl-p-benzoquinone orl-rat LDLo 250 mg/kg
Benzaldehyde orl-rat LD50: 1300 mg/kg
n-Hexaldehyde ihl-rat LCLo: 2000 ppm/4H
[1,2-Epoxybutane ihl-rat LCLo: 4000 ppm]
Source: NIOSH Toxic Substances List, 1974
49
-------
are related to ones tentatively identified, since data on few of the
more oxidized toluene products are available. (Nitroglycerin has been
included because two strong bands in the IR spectrum of the crude 1-
heptene aerosol supported the presence of organic nitrates.)
In polluted atmospheres, the concentration of particulate is
3
usually below 100 yg/m , of which about one quarter is organic. If
an average molecular weight of 200 in the organic portion is assumed,
the upper limit of organic pollutant concentration in the aerosol is
about 0.01 ppm. The toxic levels and safety thresholds in Table 4
are all considerably above this value. Although (1) much of the data
in Table 4 pertains to animal studies, (2) many of the figures refer
to doses administered by means other than inhalation, and (3) amounts
less than the lethal doses cited may cause tissue damage, it does not
appear that any of the organic compounds tentatively identified poses
an immediate health hazard. On the other hand, we cannot discount
long-term effects on humans of inhalation of such compounds in small
quantities, either alone or in combination with other pollutants.
In particular, many of the compounds identified in our study
are present in, or related to components of embalming preparations
(Table 5). The function of the aldehydes in such preparations is to
coagulate and harden protein, while the alcoholic constituents lower
the surface tension of the liquid and promote diffusion of the aldehydes
into tissue.' ' A different cellular phenomenon for which aldehydes (in
particular, malondialdehyde) are held responsible, is the appearance of
fluorescent pigments (age pigments) in living tissues in amounts that
increase with the age of the organism. Such pigments are thought merely
to accompany oxidative reactions of lipids/ 'although their formation
involves cross-linking reactions that may cause loss in tissue resiliency.
Compounds containing multiple aldehyde groups, which may react in the
same manner to induce cross-linking, were identified tentatively in
earlier work in these laboratories^ ' and elsewhere.' '
50
-------
TABLE 5. COMPONENTS OF EMBALMING FLUID
PREPARATIONS
Formaldehyde Phenol
Glutaraldehyde Glycerin
Benzaldehyde Salicylic Acid
Furfural tert-Butanol
Water Methanol
Hydroxymethylcellulose
Sources: Bennet, "The Chemical Formulary";
Kirk-Othmer, "Encyclopedia of
Chemical Technology"
51
-------
I. Conclusions
Although product identifications for many of the oxidized
species are tentative, the overall picture of the fate of toluene and
1-heptene under simulated urban atmospheric conditions is what we would
reasonably expect. The initial oxidation processes give relatively
simple, identifiable derivatives that are found on gas phase sampling.
The particulate aerosol products from both hydrocarbons are more highly
oxidized, and hence more polar and less volatile. For 1-heptene the aero-
sol consists in part of compounds of high molecular weight that are
formed by condensation of species arising from different heptene mole-
cules. The fact that conventional derivatization fails to render either
aerosol more than about 30% volatile is consistent with the presence
of several functional groups in the molecules or with a high molecular
weight. Sulfur dioxide and sulfuric acid aerosol do not give appreci-
able quantities of new compounds detectable by our approach when in-
cluded in the toluene runs-.
All of the products identified from the vapor or aerosol
phase in toluene and 1-heptene are present in concentrations that do not
appear to present a direct health hazard in the quantities expected in
urban atmospheres. The composition of the model aerosols and oxidized
species in the vapor, however, resembles that of several embalming
fluids that owe their preservative action in part to reactions of aldehyde
components with tissue protein. The effects on humans of organic aerosols
over extended periods, either alone or in synergism with other pollutants,
therefore, may be significant.
52
-------
REFERENCES
(1) Jones, P.W., Report No. 74-265, 67th APCA Meeting, Denver, Colorado,
(June 9-13, 1974).
(2) Schuetzle, D., Cronn, D., Crittenden, A. L., and Charlson, R. J.,
Environ. Sci. Tech. , _9_, 838 (1975).
(3) Miller, D. F., Schwartz, W. E., Gemma, J. L., and Levy, A. "Haze
Formation - Its Nature and Origin", Battelle Columbus Laboratories,
Final Report to the U.S. Environmental Protection Agency (Contract
No. 68-02-0792) and the Coordination Research Council (March, 1975).
(4) Kirk-Othmer, Encyclopedia of Chemical Technology, 2nd ed., Interscience,
19, Col. 8, p. 100.
(5) See e.g., Chio, K. S., and Tappel, A. L. Biochemistry, 8, 2821 (1969).
(6) Schwartz, W. E., "Chemical Characterization of Model Aerosols",
Battelle-CoLumbus Laboratories, Final Report to the U.S. Environmental
Protection Agency, Grant No. 801174, (August, 1974).
53
-------
APPENDIX A
METHANE IONIZATION SPECTRA OF
COMPOUNDS OBSERVED ON THE GC-MS
ANALYSIS (FIGURE 10) OF TOLUENE AEROSOL,
DERIVATIZED WITH BSTFA
54
-------
IK
IS
i T 1 1 1 1 1 1 1 r
goteeoeD6fieosoveeozot
UGd 3SHB JO 39UlKXftOd
g
I?
- o
"l8
= lfi
-P
o
_ro
IB
i r
MGd 3SH8 JO 39UUOUSd
et
IB
IB
o
8
'8
'S
8
"S
15
'§
- D
8
18
55
-------
e>
-
~~.
_
-
_
-
31
§ :
K - ^
^ t
e '
" ix
o o o ~
3 o o
O XX
f j
I
b
r-*
S
^
₯ ^
u
IB
"8
_(N
IB
IE
18
-N
18
Ifi
o
"s
-IN
'ft
_
'S
~
18
.?
1?
._
IS
o
L"
iR
Is
LsS
f
S
( ( i i I i ( l l i
^Qd 3SH8 JO 3SHiK33U3d
(VJ
§
E
O
g
t
f
D
8 i
1 U.
S! r.
w s
5=
*J jP
D) G5
=S
SS5
(
!
I ~~
8 g
°\=/ ° 2
_ »»
< o
o
=
ft>
.^
;
J
J
J
0
ff)
s
J
-"
g
E>
t-
"o
c*
' S
g
0
.
-
-"-
'
19
001
3SUB JO 33bOQ3H3d
56
-------
,
§
UI
Q 0 x
C
I
U.
^
I
^
« '
..... !'
.TSS
^i
-.
0
IS
- ID
_**)
"H
_r*
A g
S-r1 a
w
E
-2 g
3 I
-; I
« P f
K i
' ? ' h
-" C °
s §
D) ID
ID
_- r
g oot
L»
rr-
_v c
*n
8
--8
o
;8
__
IS
-5.^
_-'«
_,__^- * 2
- c
w O
5 - ~"
o ;
LJ
(\J C
x _w
u
.
_8
'15
I i i i f i i i i
Q608Q/.QSQSQt0COZ&l 0
\4j"Tj "acyfl JTJ *ynHlMT!tf;Lj
.
MbBd aSUB JO 39blN33kOd
57
-------
IB
18
18
o
.N
. t
i
m
s -
-8
= 'R
= '8
_,.«
001 06 08 OL 03 OS 0V BE OZ 01
MdBd 3SH8 JO 3SUlN3>Gd
e
a
y
18
-------
APPENDIX B
METHANE IONIZATION SPECTRA OF
COMPOUNDS TENTATIVELY IDENTIFIED
ON THE GC-MS ANALYSIS (FIGURE 15) OF
VAPORS ASSOCIATED WITH
TOLUENE/NO /SO AEROSOLS
X £-
61
-------
SPECTFU1 H - 8
GC-tIS COW OF TQUBC NOX SOX flEBOSOL JZS-J
8J
feSJ
60 70 88 30 100 110 120 130 110 1S0 160 170
62
-------
SPECTRUM 116 - 112
GC-*S COW OF TOJUENE NOX SOX flEROSQL J2S-1
8.
S.
0
o
T'"I'"Tmi
70 80
M/E
"iT"n""|»»i""|" i»»|»»i»"|»»i»"|""i"»|"'T*"j
90 100 110 120 130 110 1S0 160 170 1
60
63
-------
g
8.
?
co ^
8.
SPECTRUM 11S - 137
GCHS COW OF TO.UENE NOX SOX flEBDSQL J2S-1
W
4r
tMi
4r
"'""I
60 70
80 90 100 110 120 130
150 160 170 1
tire
64
-------
SPECTRUM 2S1 - 2S2
g
GO-MS COW OF TOLUENE NOX SOX flEROSOL J2S-1
O
mrnTrmn*
60 70 80
M/E
t^
h
tt
LmrmUl
|""|"M ^JTTTJ.T.IJTTHJir.TJTTTTJt.r, r.,.fnTTpTTjnTTIFmiTTlTJTTTTyn
90 100 110 120 130 ItO ISO 160 170
65
-------
SPECTflUh 267 - 261
g
GCH6
OF TOLUENE NOX SOX flEROSQL J2S-1
w «^«
8_
8_
L
g8_
5?"
1s.
S.
0
^ ~
o _
mi uiiiltriniiliiiinlll
llniniiiiilHtiMimihiliiiiiiuhin!
60 70 80
M/E
30 100 110 120 130 IfO ISO 160 170 1
66
-------
as£CTRUM 283 - 276
GO-flS COW) OF TOLUENE NOX SO* flEBDSOL J2S-1
;£j
feS
Sp
8
8J
63 TO 80 SO KB 110 120 130 110 ISO 160 170 180 130 2C
M/E
67
-------
SPECTRUM 292 - 287
OC-MS COW OF TOLUENE N9X SOX flERQSQL J2S-I
8.
§*-
ss_
8u
o
M.l.MH.mpM.pMIMM,!!,.
60 70 80 30
M/E
L,.n,r,|,..,r ,f
IMI|WfltTttf^TtttTtHl|*?ltJI1JI|Tirrii
100 110 120 130 110 150 160 170 1
63
-------
18
8
U)
T.
T
O
O
1
(
I
i 1 1 I (
001 06 06 06 09 05 &V 06
MOd 3Sye JO 39yjX33b3d
§
e
O
s
to
5
:
01 0
00!
i r
3SH8 JO
59
-------
:se
18
18
'g - e>
~ C1"
.IP
jr
18
.8'
o
_m
LS
ill
o
_K
T8
g
_N
'8
ti
GE
a
\=/
te
r
eoi
i i i i i r i r
oeoew.Q9osDVQez
3S«8 JO 3SWJK»Gd
P
01
jfi
E
Q
S
a
c
§ I
' b
S 5
^ 5
~r
~r
001 06 06
i i n~ n
OL 09 os w ee oz ci
MbQd 3SH8 JO 39HiK33k£3d
60
-------
SPECTRUM 336 - 301
8
OF TOLUENE NOX SOX BEROSOL
;-i
67
l
Tfd
1-rmf
I
Ti..,,., |,,M,.,..j»..,,n.r..., ftmiv,.,,,,, j,^,,.,,.,,.,.,,,,,.,.,,.,,,,.,.,,,.,,,...,
100 110 120 130 110 ISO 160 170 180 190
T"'|""I""J""I""|
60 70 80 30
M/E
69
-------
aPECTHUH 311 - 311
60*15 COtt3 OF TOLUENE NO* SOX BERDQOL J2S-1
8_
.J
R.
1
Hi
69 70 80 90 100 110 120 130 110 150 160 170 180 ISO 200
70
-------
8
SPECTRUM 331-330
GC-HS COW OF TCLUOC NOX SOX PEPOSOL J2S-1
^
8.
8.
feS.
GCO
g<*5 *
8-
O
o .
mimiumiiimtiiliiimiiliiiiiMi
,...,t,ii|»...,.inH,i,,i..,,Mii,i.i.j
IM J. Jl, mil h 1
..|....,....|... (....i....,...^....,.!.!^!!!^..^.!!^!!.^.
60 70 80 90 100 110 120 13(3 110 150 160 170 180 190 20
71
-------
368 - 363
Gc-tis co-tn OF TQLUQC NO* sox POCSO. 025-1
!£J
60 70 80 30 100 110 120 130 110 ISO 160 170 1
72
-------
APPENDIX C
METHANE IONIZATION SPECTRA OF
COMPOUNDS OBSERVED ON THE GC-MS ANALYSIS
(FIGURE 16 ) OF VAPORS ASSOCIATED
WITH 1-HEPTENE/NO /SO, AEROSOLS
X £,
73
-------
SPECTRUM 17 - 9
1-4GPTDE NOX SOX Hi
"
8.
-
EP.
feS.
I'-
8-
o
7 .
, 1
ff|f>1*|l ffl VIKVIIVI
III
10 SO 60
H/E
70
30 100 110 120
74
-------
SPECTRUM 21 - 21
8
1-H3TBC NOX SOX HI
8.
8.
j£O
EL r ^
tiJO
8 -
io
i/> -
l?_
|y* 5?
raco _
8.
C
** «
o
-nnJ
UUmi ll 1 1 1 n.,1 1 .Ml...
10 SO 60 70
M/E
80 30 108 110 120
75
-------
SPECTRUM 32-27
8
I*.
8
O
O
1-HEPTQC NOX SOX HI
llLpnJfl,
IIf IVI r*»|IfI*V
SO 63 70
M/E
J4
fW
4
,,.,,.l,,,.|,,,.,, ,.j.,.,,,,,,,.,.!,,,.,.,,..,,,,,..,.,.,..,,..M
80 90 100 110 120 130 110 ISO 160
76
-------
1S3 - H7
B
v-p
CC^^
i«ig>
s -
feS.
18 _
»
R.
O
1-tCFTBC NOX SOX Ml
T
10 SO GO 70 80 90
M/E
.,...T,.,,,,,.r,M,,,T,,M,,,,r..,.,,....,....,
100 110 120 130 110 ISO 160 170
77
-------
SP8CTHUM 178 - 171
8
8.
1-UEFTOC NOX SOX Ml
feS _
bj
S.
o
(t
Vhmmtml
tttrnnrnifi
,WH
.,...,|,.»J»° .,,...J.,..,.,,, ..M..,.|..,,r...|,.rM.rTT|.,.,|frf.fTTTTTT..F|..'.|..
10 SO GO 70 80 30 100 110 120 130 110 ISO
M/E
78
-------
SPECTRUM 218 - 209
1-HEPTQC NCW SOX HI
8.
feS.
rl
fc
,...,,,.,.,..,....,.
10 SO 60 70
MAE
III .J|l
rfr
*
.|i..ir...|ft..|i.i.|f...|.ii.|rr*.|i.»|fiffp»T|impTTi|rirTpnrTj
80 30 100 110 120 130 110 ISO 160
79
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/3-76-085
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
CHEMICAL CHARACTERIZATION OF MODEL AEROSOLS
5. REPORT DATE
Iuly__1916_
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
W.E. Schwartz, G.D. Mendenhall, P.W.Jones,
C. J. Riggle, A.P. Graffeo, and D.F. Miller
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Battelle-Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
R801174
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final 6/74 4/76
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Model aerosols were generated from the individual hydrocarbons,
toluene and 1-heptene, by irradiation under simulated atmospheric
conditions in the presence of NO , or NO /SO . The reactions were
carried out in a 17.3 m^ environmental chamber. The collected
aerosols were subjected to analysis by mass spectrometry and chroma-
tographic techniques, both with and without chemical derivatization.
Polyfunctional oxidation products, including quinones and carboxylic
acid, were tentatively identified in the toluene aerosol. The 1-hep-
tene filtered aerosol was shown to contain condensation products from
different 1-heptene molecules. Tentative identification of a number
of vapor-phase species was accomplished in both systems. The health
effects of the atmospheric oxidation products from hydrocarbons is
discussed.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Air pollution
*Toluene
*Heptenes
*Aerosols
*Environment Simulation
*Photochemical
react ions
*Chemical analysis
13B
07C
07D
14B
07E
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO OF PAGES
88
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22 PRICE
EPA Form 2220-1 (9-73)
80
------- |