Ecological Research Series
SMOG CHAMBER STUDIES ON PHOTOCHEMICAL
AEROSOL-PRECURSOR RELATIONSHIPS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
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 grouping was consciously
planned to foster technology transfer and a maximum inte-lace 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.
-------
SMOG-CHAMBER STUDIES OK
PHOTOCHEMICAL AEROSOL PRECURSOR RELATIONSHIPS
by
David F. Miller and Darrell W. Joseph
BATTELLE - Columbus Laboratories
Columbus, Ohio 43201
Contract No. 68-02-1718
Marijon Bufalini
Technical Planning and Review Office
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
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 endorse-
ment or recommendation for use.
-------
ABSTRACT
An experimental program was conducted in which controlled
atmospheres containing water vapor, CO, NO (NO + NO,), and a constant
2£ £
distribution of 17 hydrocarbons (NMHC) were irradiated in a smog chamber.
The principal experimental variables were the initial concentrations of
NMHC and NO . Complete smog profiles were developed for NO and hydro-
X
carbon photooxidation and aerosol and ozone formation over 10-hour
irradiation periods. The dependence of photochemical aerosol formation
goes through a maximum with respect to the initial NO concentrations,
X
and it is an ever-increasing function of the initial NMHC concentrations.
The precursor relationships vary with irradiation time. As the irradiation
period increases from 2 to 6 to 10 hours, peak aerosol concentrations relate to
initial NMHC/NO ratios of 15/1, 13/1, and 10/1, respectively. At NMHC/NOx
ratios <10/1 there are maxima in the relationships between photochemical
aerosol concentration and the initial pollutant concentrations. These
maxima generally occur for initial NMHC concentrations in the 2-3 ppmC
range. The relationships between aerosol formation and their precursors
(NMHC and NO ) were found to be qualitatively similar to those for ozone
X,
formation, and thus NMHC and NO control strategies for limiting ozone
X
are mutually beneficial in reducing photochemical aerosols.
iii
-------
CONTENTS
Abstract iii
Figures v
Tables viii
Acknowledgments ix
1. Introduction. . . , . , . 1
2. Summary , . . , 3
3. Review of Aerosol Formation in Smog Chambers 7
Reactivity Studies . . . . , 7
Aerosol Studies With SC>2 7
Aerosol Studies Without S02 8
4. Current Interpretation of Organic Aerosol Formation ...... 20
Precursor Characteristics of Organic Aerosol Formation . . 20
Interlaboratory Comparisons, 26
5. Experimental Approach 32
6. Experimental Methods 38
Smog-Chamber Description and Operation 38
Analytical 39
7. Results 43
8. Discussion , 47
Overall Reactivity 47
Hydrocarbon Oxidation 49
Aerosol Precursor Relationships. , 55
Ozone Precursor Relationships. . 65
Aerosol and OzoneMutual Benefits From Precursor
Controls ..... 75
References 81
Appendices
A. Smog Profiles 86
B. Summary of Hydrocarbon Data Determined by Gas Chromatography. . 96
iv
-------
FIGURES
Number
la-c Surface Projections Representing Instantaneous
Aerosol Volume Concentrations as Functions of the
Initial Concentrations of NMHC and NOX At Irradiation
Times of 2, 6, and 10 Hours ..............
2 Relative Reactivity of Exhaust Hydrocarbons in Forming
Light-Scattering Aerosols in Simulated Smog ...... 1"
3 Regression Relationship of Aerosol Formation (Light
Scattering) With Hydrocarbon Reactivity For Auto Exhaust
Derived From a Linear Summation of Individual
Reactivities ...................... 12
4 Profile of Aerosol Formation During Irradiation of
Filtered and Diluted Auto Exhaust (16 ppmC Hydro-
carbons) ....................... 16
5 Reproduction of Smog Profile From the Photooxidation
of 1-Hexene and NO in the Calspan Chamber ....... 18
6 Photochemical Aerosol Formation During a Smog-Chamber
Irradiation of a Toluene-NO -Air Mixture ....... 21
X
7 Photochemical Aerosol Formation During a Smog-Chamber
Irradiation of a 1-Heptene-NO -Air Mixture ...... 22
8 Photochemical Aerosol Formation During a Smog-Chamber
Irradiation of a Surrogate Hydrocarbon Mixture and
N0x .......................... 25
9 Effect of N02/N0 Ratio on Photooxidation Rate
Parameters in the 1-Butene-NO -System ......... 28
X
10 The Effect of Primary Auto Exhaust Aerosol on
Photochemical Aerosol Growth and Light Scattering. . . 33
11 Evidence of Preferential Homogeneous Nucleation
of Photochemically Derived Aerosol in Air Con-
taining Primary Nuclei ................ 34
12 Initial Hydrocarbon and Nitrogen Oxide Concen-
tration Coordinates in the Experimental Program. ... 37
-------
FIGURES (Continued)
Number
13 Representative Chromatogram Showing Resolution
of the Surrogate Hydrocarbon Mixture Obtained
With Two Gas Chromatographs 40
14 Computer-Generated Graphs of the Changes in the
Aerosol Surface-Area and Volume-Size Distri-
bution that Occur as a Function of Irradiation
Time 42
15 Fractional Hydrocarbon Decay Rates at 9.1/1 NMHC/NOX
Ratio, Run No. 1 50
16 Effect of NMHC/NO Ratio on the Rate of Aromatic
Hydrocarbon Decay 53
17 Isopleths of Maximum Rates of Aerosol Formation
as a Function of the Initial Concentrations of
NMHC and NOX (Isopleths correspond to intervals
of volume production rates of 2 pm3/cm3/hr.) 57
18 A Surface Projection Representing Maximum Rates
of Aerosol Formation as Functions of the Initial
Concentrations of NMHC and NOX 57
19a-c Surface Projections Representing Aerosol Volume
Concentrations as Functions of the Initial Con-
centrations of NMHC and NO at Irradiation Times
of 2, 6, and 10 Hours. . .X 59
20a-c Isopleths of Aerosol Volume Concentration as
Functions of Initial Concentrations of NMHC and
NOX at Irradiation Times of 2, 6, and 10 Hours
(Isopleths correspond to volume concentration
intervals of 2 ym3/cm3/hr.) 60
21a-c Surface Projections Representing Aerosol Volume
Concentrations as Functions of Initial Pollutant
Concentrations at a Constant NMHC/NOX Ratio of
10/1 and Irradiation Times of 2,6, and 10 Hours. ... 63
22a-c Surface Projections Representing Aerosol Volume
Concentrations as Functions of Initial Pollutant
Concentrations at a Constant NMHC/NOX Ratio of
5/1 and Irradiation Times of 2, 6, and 10 Hours. ... 64
vi
-------
FIGURES (Continued)
Number
23 Isopleths of Constant Ozone Concentration (ppm)
Developed From Peak Ozone Concentrations in an
Earlier Smog-Chamber Study 66
24 Isopleths of Constant Ozone Concentration (ppm)
Based on 5-Hr Data Predicted by a Kinetic Smog
Model 66
25 Isopleths of Constant Ozone Concentrations (ppm)
Derived From the LACAPCD Smog-Chamber Studies 67
26 Isopleths of Constant Ozone Concentrations (ppm)
Derived From Instantaneous Ozone Concentrations
at 6-Hr of Irradiation 67
27a-c Surface Projections Representing Ozone Concentrations
as Functions of Initial Concentrations of NMHC and NOX
at Irradiation Times of 2, 6, and 10 Hours 79
28a-c Isopleths of Ozone Concentrations as Functions
of Initial Concentrations of NMHC and NOX at
Irradiation Times of 2, 6, and 10 Hours
(Isopleths correspond to concentration intervals
of 0.05 ppm 03.) ................... 71
29a-c Surface Projections Representing Ozone Concen-
trations as Functions of Initial Pollutant
Concentrations at a Constant NMHC/NOX Ratio of
10/1 and Irradiation Times of 2, 6, and 10 Hours ... 73
30a-c Surface Projections Representing Ozone Concen-
trations as Functions of Initial Pollutant
Concentrations at a Constant NMHC/NOX Ratio
of 5/1 and Irradiation Times of 2, 6, and 10 Hours . . 74
31a-f Comparisons of the Concentration Dependence of
Aerosol (a,c,e) and Ozone (b,d,f) Volume on the
Initial Concentrations of NMHC and NO at
Irradiation Times of 2, 6, and 10 Hours
32a-f Comparisons of the Concentration Dependence of
Aerosol (a,c,e) and Ozone (b,d,f) Volume on the
Initial Concentrations of Pollutants at a Constant
NMHC/NOX Ratio of 10/1 and Irradiation Times of
2, 6, and 10 Hours .................. ,
33a-f Comparisons of the Concentration Dependence of
Aerosol (a,c,e) and Ozone (b,d,f) Volume on the
Initial Concentrations of Pollutants at a Constant
NMHC/NOX Ratio of 5/1 and Irradiation Times of
2, 6, and 10 Hours .................. 79
vii
-------
TABLES
Number
1 Aerosol Formation From Selected Hydrocarbons 11
2 Comparisons of Aerosol Formation and Reactivity for
Smog Chambers at Calspan and the University of
Minnesota 17
3 Estimated Aerosol Conversion Efficiencies for a Few
Familiar Hydrocarbons 23
4 Comparisons of Smog-Chamber Conditions at Calspan and
Battelle and Some Reactivity Results of Olefin Photo-
Oxidation 27
5 Reference Atmosphere 36
6 Initial Pollutant Concentrations 44
7 Summary of Experimental Results 46
8 Correlation Coefficients Among Measured Reactivities. ... 47
9 Correlation Coefficients Between Aerosol Concentration
and the Time Integrals of Hydrocarbon Decay 49
10 Hydrocarbon Oxidation Rates in Polluted Air and in
Smog-Chamber Simulations 52
11 Average Hydrocarbon Loss Rates Under Natural and
Simulated Irradiation Conditions 54
12 Selected Data on the NMHC and NO Distribution in
Urban Areas 61
13 Worst-Case Ozone Episodes in Pasadena (1969-1970) and
the Precursor Hydrocarbon and NOX Concentrations 69
viii
-------
ACKNOWLEDGMENTS - '
The generation, compilation, and presentation of data on which
this report rests required the expertise of many coworkers. The authors
are especially grateful to the contributions made by Fred Blakeslee,
James Hoyland, George Keigley, Barbara Levine, Joseph Miller, Philip
Schumacher, and Gerald Ward.
ix
-------
SECTION 1
INTRODUCTION
The photochemical conversion of gases to aerosols in the
atmosphere results primarily in the formation of sulfate, nitrate, and
oxygenated organic compounds. In submicron-size aerosol samples
collected at the fringes of some eastern and midwestern cities, the
mass concentrations (24-hour avg.) of these compounds fall into the
following ranges :
Sulfates (2-25 yg/m3)
1 3
Nitrates (0.2-4 yg/m )
2
Oxygenated organics (2-40 yg/m ).
Stationary and mobile combustion sources also make contributions to these aerosol
compounds, and from chemical analyses of the samples alone it is not possible
to specify the respective sources. In the Los Angeles basin where photo-
chemical smog prevails, a recent study reports maximum concentrations of
these compounds ranging 2-5 times as great as the maxima indicated above'^).
Such high concentrations illustrate the tremendous potential of our
polluted atmosphere to produce aerosols via photochemical reactions.
Reduction of the photochemically derived aerosols can best be
achieved by control of the gaseous precursors. Therefore, to develop an
effective emission control strategy, it is necessary to quantitate the
dependence of photochemically derived aerosols on the controllable gaseous
precursors. In this program, a smog-chamber approach is taken in seeking
these relationships. Emphasis is placed on measuring aerosol formation
in complex but controllable experimental atmospheres whose composition
closely resembles that of our polluted urban air. The principal experi-
mental variables studied thus far are the concentrations of total non-
methane hydrocarbons and nitrogen oxides.
The major findings of the experimental study and some thoughts
on control strategies are presented in the Summary section of the report.
The scope of the study and the experimental details are presented in
the sections entitled Experimental Approach, Experimental Methods,
Results, and Discussion.
-------
In addition to the laboratory investigation, we have been
requested to provide some overall interpretations of photochemical
aerosol formation in light of other smog-chamber research in this area.
Those discussions are contained in the sections entitled Review of
Aerosol Formation in Smog Chambers and Current Interpretations of
Organic Aerosol formation.
-------
SECTION 2
SUMMARY
An experimental program was conducted in which controlled
atmospheres containing water vapor, CO, NO (NO + N0_), and a constant
X ^
distribution of 17 hydrocarbons (NMHC) were irradiated in a smog chamber.
Complete smog profiles were developed for NO and hydrocarbon photooxidation
and aerosol and ozone formation over 10-hour irradiation periods. Com-
parisons of the smog-chamber results with data on hydrocarbon oxidation
rates observed in the Los Angeles area and with worst-case ozone episodes
in that area suggest that the models (precursor relationships) developed
here for photochemical aerosol formation are highly relevant to the smog
problems in polluted atmospheres.
The simultaneous dependence of aerosol formation on the initial
NMHC and NO concentrations is summarized in Figure 1. In these graphs,
X
aerosol formation is represented by a response surface in perspective
while NMHC and NO are the abscissa and ordinate, respectively. In all
X
regions of the graphs dependence of aerosol formation on NMHC is always
positive, but the dependence with respect to NO is both positive and
X
negative, i.e., the latter dependence goes through a maximum. In effect
the initial NO concentration controls the extent to which hydrocarbon
X
vapor is converted to organic aerosol, and the aerosol response surface
can be thought of as a gas-to-aerosol conversion efficiency. The crest in
the response surface thus corresponds to conditions for maximum conversion.
One of the most important features of aerosol precursor relation-
ships is the time dependency. At 2 hours (Figure la) aerosol formation is
strongly suppressed by high NO concentrations, and the crest corresponding
X
to maximum conversion efficiency follows a NMHC/NO ratio of 15/1 in the
X
region of lower pollutant concentrations. The ridge bends and follows a
course of higher NMHC/NO ratios in the region of higher pollutant concen-
X
trations. By 6 hours (Figure Ib) the response surface has swelled up in
the NO region of the grapha result indicative of the diminishing suppres-
X
sion of NO as the irradiation time increases. This trend is further
X
-------
a. 2 hours
N0x(ppm)
14.25
NMHC(ppmC)
b. 6 hours
NMHC(ppmC)
14.25
c. 10 hours
N0x(ppm)
NMHC(ppmC)
14.25
FIGURE 1. SURFACE PROJECTIONS REPRESENTING INSTANTANEOUS AEROSOL VOLUME
CONCENTRATIONS AS FUNCTIONS OF THE INITIAL CONCENTRATIONS OF
NMHC AND NO AT IRRADIATION TIMES OF 2, 6, AND 10 HOURS
X
-------
illustrated by the response surface at 10 hours (Figure Ic). By this
time the maximum aerosol concentrations correspond to a NMHC/NO ratio
3C
of 10/1 in the region of common atmospheric pollutant concentrations.
Another interesting feature of the data pertains to the
dependency of aerosol formation on the initial pollutant concentrations,
i.e., for varying NMHC and NO concentrations but constant NMHC/NO ratios
x x
(e.g., see Figures 21 and 22). For all irradiation periods there appears
to be a region of initial pollutant concentrations where the aerosol con-
centration becomes constant; pollutant concentrations above the region
do not increase the aerosol concentration. This phenomenon is interpreted
to mean that, at limiting pollutant levels, gas-to-aerosol conversion
efficiency diminishes, and the atmosphere is overburdened in its effort
to oxidize primary pollutants. In general the point of limiting pollutant
concentrations increases with increasing NMHC/NO ratios.
X
If the models developed can be trusted quantitatively, it appears
that for NMHC/NO ratios of 10/1 and 5/1 the peak efficiency in aerosol
X
production occurs near pollutant concentrations of 2 ppmC NMHC and 0.2
and 0.4 ppm NO , respectively, with this knowledge, there are two
X -
plausible approaches for reducing the concentrations of photochemically
derived aerosols: (1) lower the overall primary pollutant (NMHC and NO,.)
X
concentrations in the region of maximum conversion efficiency, or (2)
cause a shift in the distribution of NMHC and NO in a direction which
X
lowers the efficiency. The former approach is more asthetically appeal-
ing in that it permits the hydrocarbon degradation to proceed most
efficiently while still maintaining acceptable concentrations of secondary
pollutants. If we assume a starting point of 3.5 ppmC NMHC and 0.35 ppm
NO and apply the 6-hour-irradiation model to obtain an 80 percent reduction
A
in aerosol via the former strategy, a concomitant reduction in NMHC and NO
X
concentrations of 74 percent would be required (NMHC =0.9 ppmC and NO =
X
0.09 ppm). If,the latter strategy of unilateral NMHC control was invoked
at the same starting point, an 80 percent reduction in aerosol could be
obtained by reducing the NMHC level about 83 percent (NMHC =0.6 ppmC).
From a practical standpoint the latter approach is more attractive. A draw-
back of this approach, however, is the pitfall of "hydrocarbon ruts" which
-------
occurs when limitations of hydrocarbon emissions are reached before a
standard is met, At this point of the hypothetical condition, large
reductions in NO would be required before any further improvement in
X *
the level of photochemical aerosols would be realized.
Comparisons between precursor relationships for aerosol formation
and ozone formation are brought out in the text of the report. Although
there are some substantial differences in their relationships to NMHC and
NO , it was quite satisfying to find that NMHC and NO control strategies
X X
follow mutually beneficial paths. However, in following a course of
unilateral NMHC control, the benefit with respect to aerosol formation
is predicted to be less than that for ozone.
-------
SECTION 3
REVIEW OF AEROSOL FORMATION
IN SMOG CHAMBERS
REACTIVITY STUDIES
Numerous smog-chamber studies have been conducted to assess the
functions of hydrocarbons and nitrogen oxides in photochemical smog. Most
of the studies have focussed on the reactivity of individual hydrocarbons
or organics based on smog-chamber manifestations other than aerosol formation;
i.e., the reactivities have been based on NO photooxidation rates, ozone
formation, hydrocarbon depletion rates, aldehyde and PAN production, and
(3-17)
eye irritation . Insofar as this is an aerosol study, we have not
attempted to review those reports with the objective of comparing our
data on those reactivity bases. However, since aerosol formation, at least
under the conditions investigated, is strongly linked to hydrocarbon oxidation
some pertinent comparisons of this parameter are made here and in the text
of the report. Other general comparisons of reactivity data are also inter<-
spersed in the report.
AEROSOL STUDIES WITH S02
Most of the smog-chamber studies on aerosol formation have
involved irradiations of individual hydrocarbons with NO ; often with S09
(18-33) X
added . Results of studies conducted with S02 generally concur with
the following summary. When SO, is added to either aromatic-NO -air or
£, X
alkane-NO -air mixtures, the total aerosol produced is approximately that
X
predicted by an additive model; i.e., the sum of the organic aerosol pro-
duced when the respective hydrocarbon is irradiated with NO , and the
X
sulfate aerosol produced when SO- is irradiated in the absence of the
reactive hydrocarbon. With the more reactive aromatic and alkane hydro-
carbons and SO2 there may be some enhancement of total aerosol formation
over the linear combination of the individual systems.. With the olefins,
the levels of aerosol obtained with added SO- show a definite synergistic
effect. The enhancement of total aerosol production is greatest for the
C2-C, olefins which make little organic aerosol.
-------
Data are also available showing the enhancement of photochemical
aerosol formation when S02 is added to auto exhaust^ 9»3^~36)_ As expected,
this effect is greatest for exhaust compositions highest in olefinic content.
While these generalities on the involvement of S02 in aerosol formation may
hold true, many quantitative and mechanistic aspects of S02 oxidation are
not understood and are the subject of other investigations. In this study S0_
was deliberately excluded from the experiments, and it will not therefore be
considered further in our discussions.
AEROSOL STUDIES WITHOUT S02
In reviewing the aerosol studies conducted without S02, there
appears to be some controversy over the relative importance of two hydro-
carbon types, aromatics and olefins, in their roles as organic aerosol
precursors. There is unanimous agreement that common alkanes and aldehydes
play little part as precursors in photochemical aerosol production.
Several studies(20>25'29>32'37)38) have pointed out the tremendous
propensity of some diolefins, cyclic olefins, and terpenes to form organic
aerosols. In addition to being highly reactive with ozone, these hydro-
carbons appear to be unusually prolific aerosol precursors by providing two
sites for oxidation and thereby readily acquiring the low vapor pressures
needed for condensation. A few examples of the oxidized compounds of aerosol
produced in Battelle's smog chamber from cyclohexene and o-pinene are shown
below. Identification was made by gas chromatography/chemical ionization
(38)
mass spectrometry
-------
CCOOH
CH2OH \^
TENTATIVE
0 0
// //
CHj-CH=CH-C -C-
r CHO
CYCLOHEXENE
CHO
HO-CHj-CH2-CH=CH-CH=CH2
CHO
or
CH, CH, CH,
COOH X^*° x^*D CHO
COOH
TENTATIVE
a-PINENE
CHO
CH, CH, CH,
CHO
Apart from forested areas where terpenes are prevalent, the rather
exotic olefins mentioned above are rarely found. Thus In polluted urban
atmospheres we need be concerned about the more familiar olefins and the
aromatic hydrocarbons typical of combustion and evaporative emissions. Here
the distinction between the importance of olefins and aromatics in aerosol
formation is 'not so clear.
In the early work of Stevenson, et al. , photochemical aerosol
formation (measured by light scattering) was observed upon irradiation of
1,3,5-trimethylbenzene-NO -air mixtures as well as NO -air mixtures with
x x
-------
1-hexene, 1-heptene, 3-heptene, and cyclohexene. With trans-2-butene only
smaller particles were produced as evidenced by condensation nuclei counts.
The results of our more detailed study of aerosol formation from
hydrocarbon-^NOx mixtures showed that aromatic hydrocarbons were more
reactive in aerosol production than the olefins and alkanes typical of
urban pollution^29). The relative reactivities of these classes are
summarized in Figure 2.
100
10
S.
0.1
(6),
(26)1
Aromatics
(alkylbenzenesj
Olefins
Paraffins
- (alkanes)
FIGURE 2. RELATIVE REACTIVITY OF EXHAUST
HYDROCARBONS IN FORMING LIGHT-
SCATTERING AEROSOLS IN SIMULTATED
SMOG
The vertical bars in Figure 2 indicate the reactivity range
within each structural class. Based on the average light scattering in
each class, the relative reactivity ranking of the three classes was
aromatic > olefins > alkanes in the ratio 26/6/1.
Results of a more recent study of organic aerosol formation by
O'Brien, et al. 3 are shown in Table 1. With the exception of o-xylene,
these data indicate that aromatic compounds and tnonoolefins with carbon
10
-------
TABLE 1. AEROSOL FORMATION FROM SELECTED
HYDROCARBONS 0'Brien(33).
Maximum
/ >. Light Scattering
Hydrocarbon^ } bscat x 10^ nT1
Glutaraldehyde 0
Ethylbenzene 1
Mesitylene 1
2,6-Octadiene 1
1-Octene 1
trans-4-Octene 1
5-Methyl-l-hexene 1
2,6-Dimethylheptane ' 1
1-Heptene 1
o-Xylene 8
1,5-Hexadiene 40
Cyclohexene 90
2-Methyl-l,5-hexadiene 110
1,6-Heptadiene 160
1,7-Octadiene 180
a-Pinene 180
(a) Each hydrocarbon (2.0 ppm) irradiated
with 1.0 ppm nitric oxide and 70% RH
measured at 22 C.
chains >_ C, produce aerosols corresponding to similar light-scattering levels,
A rather reactive alkane, 2,6-dimethylheptane, was also found to be in this
dategory. Olefins of carbon length <5 were found to produce no light
scattering in'accord with the other studies.
Even more relevant to the atmospheric situation, we want to know
the reactivity of these hydrocarbons behaving in complex mixtures. Studies
of secondary aerosol formation from auto exhaust have helped in this respect.
11
-------
( 35)
In some of the early work, Schuck, et al. observed that the higher
olefinic exhausts produced the most aerosol (measured by light scattering),
but the sulfur content of the fuels employed was quite high (up to 0.22
/og\
weight percent). Data from a study by Hamming, et al. , where fuel
sulfur levels were only 0.01-0.04 weight percent, indicated that fuels
high in aromatic content produced more aerosol than other compositions.
Selected results of several years' work at Battelle on auto
(39)
exhaust are summarized in Figure 3 . The smog-chamber experiments
(replicate experiments shown as averages) were conducted with 8 ppmC
exhaust hydrocarbons generated from low-sulfur (<0.02 weight percent)
nonleaded fuels. Figure 3 depicts an implied linear relationship between
- 3.0
'£
*Q2.5
0
2.0
1.0
CO
0.5
Q
8 ppm C Exhaust Hydrocarbons
0 5 10 15 20 25
Relative Hydrocarbon Reactivity
FIGURE 3. REGRESSION RELATIONSHIP OF AEROSOL
FORMATION (LIGHT SCATTERING) WITH A
HYDROCARBON REACTIVITY FOR AUTO EXHAUST
DERIVED FROM A LINEAR SUMMATION OF
INDIVIDUAL REACTIVITIES
peak light scattering and a normalized hydrocarbon reactivity parameter computed
for the exhaust composition on the basis of linear summation using a 26/6/1 weight
ing for aromatic/olefin/alkane hydrocarbons. While there is considerable scatter
12
-------
in the mid-region of the reactivity scale the correlation is reasonably good
(0.91) especially realizing that innumerable factors (slight variations in
HC/NO ratios, light intensity, CO concentration, etc.) are not taken into
A
account. Actually, an improvement in the correlation coefficient for this
type regression was obtained if the aromatic/olefin/paraffin weighting
factors were changed from 26/6/1 to 10/6/1. This analysis would suggest
that, for exhaust mixtures, aromatic hydrocarbons (on the average) are
only 2 rather than 4 times as reactive as olefins in promoting photochemical
aerosol formation. A reduction in the relative reactivity for aromatic
hydrocarbons in extrapolating reactivity data from experiments with single
hydrocarbons to those with mixtures of hydrocarbons is also consistent with
the results obtained when binary hydrocarbon mixtures containing aromatics
are irradiated. That is, the peak light scattering observed with a binary
hydrocarbon mixture (including one or two aromatics) of different reactivity
is consistently less than (and often half) that predicted by a linear summation
of peak light scattering derived from experiments with the individual hydro-
carbons ^39).
Results from two other laboratories conducting similar research
with auto exhaust seem to support, at least qualitatively, our findings
regarding the relationship between exhaust aromatic content and peak light
(40 41)
scattering from secondary aerosol formation '
One final pointbshould be made here about the relevance of
aromatics. Several investigators have dismissed aromatics as major
participants in aerosol formation on the basis that there is little aromatic
character associated with the organic extracts of atmospheric aerosols.
Indeed, we found that the ratio of aliphatic to aromatic protons was >10/1
for most samples . The explanation, we feel, is that the oxygenated
compounds emanating from aromatic degradation loose their aromatic
character via ring cleavage. Evidence of this is demonstrated by the
structures shown below. These compounds were identified as major aerosol
/og\
products when toluene and NO were irradiated in a smog chamber . Note
X
that with the exception of the nitrated compound, all products have lost
their ring structure.
13
-------
COOH
COOH
CHZOH
.CHzOH
TENTATIVE
TOLUENE
(TW(T"TSOMERS)
CH
CHO
CHO 1 CH0H
I CH2OH
In nearly all the aforementioned studies, aerosol formation has
(42)
been determined by light-scattering principles or condensation nuclei
(43)
counters . For reasons that we will not attempt to detail here, neither
the light-scattering methods nor the nuclei counters is completely satis-
factory for quantitating the concentration of aerosol produced in units of
volume or mass. The volume (or mass) of aerosol formed is the most important
quantity to determine in this work because it is the only quantity con-
served during the experiments*: Since aerosol volume is conserved, the rate
of its formation is directly proportional to the rate of gas-to-aerosol
conversion (whether via nucleation or condensation). How the photochemical
aerosols manifest in the real atmosphere (i.e., how the aerosol mass is
eventually distributed with respect to aerosol size) will depend in large
part on the nature of the aerosol environment in which they are generated.
*0ther quantities, such as the number concentration (CNC), the total surface
area concentration or the surface cross section in a particular size range
(light scattering) cannot be interpreted in terms of volume or mass without
additional information.
-------
The most serious limitations of the light-scattering instruments
are the strong dependence of light scattering on particle size in the
diameter range 0.1 to 0.5 urn and their insensitivity to aerosols <0.1 pm.
Since most smog-chamber experiments are conducted without primary aerosols
in the light-scattering range, secondary aerosol growth begins from
nucleation, and the subsequent growth via condensation (and, in some cases,
continued nucleation) requires a substantial degree of gas-to-aerosol con-
version before many of the aerosols approach the light-scattering size
range<39>.
(44) (45) (46)
With the development^ ', refinement*" ', and calibration^ of
the Electrical Aerosol Analyzer over the past several years, it is now
possible to monitor, in real time, the size distribution of secondary
aerosol growth over a considerable size range (0.005 to 0.3 pm)* and
thereby infer by integration the aerosol volume concentration.
An example of the information obtained by this new monitoring
technique is shown in Figure 4 where filtered auto exhaust (16 ppmC hydro-'
carbons) was irradiated for 6 hours. The aerosol number, surface and
volume concentrations were derived from the electrical aerosol analyzer.
The light-scattering curve was determined by an integrating nephelometer.
For the last few years electrical aerosol analyzers have been
utilized in smog-chamber research at the University of Minnesota (U of M)
and Calspan, as well as at Battelle. Currently, these instruments are
operating at several other laboratories including EPA-RTP, University of.
North Carolina, California Institute of Technology, Science Center at
Rockwell International, and General Motors Research.
In spite of the vast improvements in aerosol monitoring, there
remain major differences in the rates of aerosol production for
seemingly similar experiments conducted in different smog chambers. Table 2
summarizes results from a series of experiments conducted at Calspan and
the U of M in which the initial concentrations of hydrocarbons and nitric
oxides were nearly identical^ . Looking at the rate parameter for NO
*A-range which includes >90 percent of all the aerosol volume observed in
the experiments conducted in this study.
15
-------
Number (8 xlO5 cm*3)
Surface (4000 umz cm-')
(2 ppm)
Volume (20Qu.tr?
^^ r-03(2ppm)
N- \-~-~
234
Irradiation Time, hr
FIGURE 4. PROFILE OF AEROSOL FORMATION DURING
IRRADIATION OF FILTERED AND DILUTED
AUTO EXHAUST (16 ppmC HYDROCARBONS)
photooxidation (Table 2, N00-t ) it appears that for the more slowly
4f Iua,fC
reacting hydrocarbons (toluene and 1-hexene) reactivity in the Calspan
chamber is considerably less than in the U of M chamber, in spite of the
fact that the light intensity of the Calspan chamber was substantially
greater than that of the U of M chamber (kd =0.23 rain" and 0.15 min ,
respectively). The differences in reactivity in terms of NO photooxidation
are less in the cases of m-xylene and cyclohexene, but again the rates
are highest for Calspan. Differences in the maximum aerosol formation rate
(Table 2) for "replicate" experiments are quite diverse; a factor of 10
for toluene, 20 for 1-hexene, 5 for m-xylene, and 2 for cyclohexene. Most
disconcerting, perhaps, is the fact that the higher formation rate in each
case is not occurring in the same chamber. The higher rates of aerosol
16
-------
TABLE 2. COMPARISONS OF AEROSOL FORMATION AND REACTIVITY FOR
SMOG CHAMBERS AT CALSPAN AND THE UNIVERSITY OF
MINNESOTA
Reactivity Parameters
Run
No.
6
76
5
92
15
81
10
83
Laboratory (a)
Calspan
U. of Minn.
Calspan
U. of Minn.
Calspan
U. of Minn.
Calspan
U. of Minn.
Initial
Hydrocarbon ppm
toluene
toluene
1-hexene
1-hexene
m-xylene
m-xylene
cyclohexene
cyclohexene
Conditions
(vol/vol)
0.35
0.35
0.33
0.35
0.34
0.35
0.33
0.35
[NO],
ppm
0.17
0.15
0.15
0.12
0.15
0.15
0.14
0.13
A
N02-tmax,
min
400
210
420
280
100
80
120
90
&im CO "1 \7rt*l iimo TnV^i^iTiEi t~ "i rtm
Rate (dV/dt),
ym3/cm3/hr
2.2
24.5
2.1
0.09
14.1
73
110
50
(a) Laboratory Conditions: Calspan chamber volume = 20,800 ft3, k, = 0.23 min~l;
University of Minnesota chamber volume = 600 ft3, k, = 0.15 min~l.
-------
production with aromatic hydrocarbons (toluene and m-xylene) were observed
in the U of M chamber while the higher aerosol rates for olefins (1-hexene
and cyclohexene) were observed at Calspan.
There appears to be another peculiarity in the above study that
needs attention. Figure 5 is a time-concentration profile of Calspan
Run No. 5. Considering the normalcy of the HC/NO ratios in the experiment
X
(ppmC/ppmV = 13/1 in the case of 1-hexene), the rate of NO photooxidation
0,3
e
ex
ex
10
O
CM
i
0.2
-i 1 r
[Aerosol]
1 1 1 1 1 1
- CALSPAN Run No. 5, 22 February, 1974
o [N02]ppm N0= 0.152 ppm; N02= 0.014 ppm
A [NO] ppm
a [03]ppm
xene-
I
I
0.3 -
0.2^-
o>
O.I g-
x
o>
120 240 360 "480 600 720 840 960 1080 1200
Time, min
0 -
20
12
FIGURE 5. REPRODUCTION OF SMOG PROFILE
FROM THE PHOTOOXIDATION OF
1-HEXENE AND NO IN THE CALSPAN
CHAMBER
is unusually slow, and there is an unusually long induction period to
aerosol formation (5 hours before any appreciable aerosol volume was
observed in Calspan Run No. 5). As pointed out by comparisons made
later in the report, these rates of oxidation are much slower than observed
in the real atmosphere. The low light intensity in the Calspan chamber
cannot be entirely responsible for the apparent lack of reactivity; *ln
the real atmosphere, Jefferies, et al.(48) report an average k value of
18
-------
0.28 min" for the 5 hours between 0800 and 1300-EDT (latitude 35.72°,
September 19, 1974) which is only twice as great as the k.. value
estimated for the Calspan chamber.
It is not likely that the peculiarities among smog-chamber
results can be rationalized satisfactorily at this time. We would like
to believe, however, that the major differences are related primarily
to our insufficient knowledge of the chemical reactions taking place
rather than to some mysterious artifacts involving "dirty chamber walls".
As we attempt to provide a "current interpretation" of organic aerosol
formation in the succeeding section of the report, we will try to explain
some of the peculiarities and inconsistencies described above.
19
-------
SECTION 4
CURRENT INTERPRETATION OF ORGANIC
AEROSOL FORMATION
PRECURSOR CHARACTERISTICS OF
ORGANIC AEROSOL FORMATION
With the history of earlier work and the data emanating from
ongoing studies at several laboratories, it is possible to present an
updated overview of the formation of photochemically derived organic
aerosols.
In polluted urban atmospheres, the aromatics and the higher
molecular-weight olefins have been shown to be the most important types
of hydrocarbons in the formation of organic aerosols. It appears that
these hydrocarbons react by different mechanisms in initiating the
oxidation steps leading to condensable matter, and it is instructive to
recognize these differences.
Differences in the simulated reaction profiles give evidence
of the mechanistic differences. When tolulene is photooxidized with NO,
NO,, and water vapor in the Battelle smog chamber, aerosol formation
results as shown in Figure 6. Toluene is oxidized more slowly than
other aromatics (alkylbenzenes). The slower oxidation of toluene
serves to illustrate an important characteristic we have seen with all
aromatic hydrocarbons investigated; namely that, under proper conditions,
toluene and the other alkylbenzenesproduce aerosol during the period of
NO oxidation and before appreciable 0_ formation. This, of course, is
due to the fact that the aromatics react most exclusively with HO radicals
under these conditions. The important role played by HO will be dis-
cussed shortly.
20
-------
280
0 60 120 180 240 300
Irradiation Time, min
FIGURE 6. PHOTOCHEMICAL AEROSOL FORMATION DURING A
CHAMBER IRRADIATION OF A TOLUENE-NO -AIR MIXTURE
Figure 7 is a smog profile resulting from irra4iation of 1-heptene,
NO, N02» and water vapor. In this case, aerosol formation seems to be
delayed until 03 is formed, and we believe that in the case of olefins,
the olefin-Og reaction may be more important overall to aerosol production
than the olefin-HO reactions. Notice here that during the first 30 minutes
of the irradiation, before appreciable 0« is formed, a substantial amount
of 1-heptene has been oxidized via 1-heptene-HO reactions, yet only a
very small volume of aerosol was produced in this period. With the
i
appearance of 0^, the rates of aerosol formation and 1-heptene oxidation
increase .
21
-------
320
-240«g
l-HeptenetxICT) _
0 60 120 180 240 300
Irradiation Time, min
FIGURE 7. PHOTOCHEMICAL AEROSOL FORMATION DURING A SMOG-
CHAMBER IRRADIATION OF A 1-HEPTENE-NO -AIR MIXTURE
Because we are now able to measure the volume of aerosol pro-
duced with irradiation time (i.e., the gas-to-aerosol conversion rate)
it seems important to attempt to provide a quantitative (or at least
semiquantitative) measure of reactivity of hydrocarbons with respect
to aerosol formation. In the past (e.g., references 29 and 33) reacti-
vities have been expressed on relative scales, and there has been no
basis for assigning an absolute measure of aerosol production to any
of the hydrocarbons. In an effort to provide quantification, we have
defined a rather simple relationship called conversion efficiency
(relationship A):
Percent Conversion Efficiency _ Maximum Aerosol Formation Rate
(gas-to-aerosol conversion) = Maximum HC Oxidation Rate : X 100' '
22
-------
We define conversion efficiency for a particular hydrocarbon as the maximum
aerosol formation rate divided by the maximum oxidation rate of the hydro-
carbon during this occurrence; or the fraction of the hydrocarbon consumed
which results in condensable matter. We have used mass as the basic unit
of comparison. If the efficiency terms can be trusted (and there will no
doubt be some variations of the values for different smog conditions;
HC/NOx ratios, etc.), then to predict aerosol formation under normal smog
circumstances one might only have to multiply the observed hydrocarbon
depletion rate by the appropriate conversion-efficiency value.
Table 3 shows some estimates of the efficiencies of a few hydro-
carbons .
TABLE 3. ESTIMATED AEROSOL CONVERSION EFFICIENCIES
FOR A FEW FAMILIAR HYDROCARBONS
Hydrocarbon
Toluene . .
** + HO - 6 * 10 * >
Efficiency, percent
HC + HO HC + 0.
7 very small
Butane
HO
Propylene
10
*H - 2'5 * 10
C + HO
very small very small
(<1 x 10-2)
very small
(<3 x 10-3)
0.1
1-Heptene
"H
C + HO
Cyclohexene
lkHC H- 0,
10J
- 5 * 10
~2
r»-l
0.15
(a) Rate units = ppm~^ min~l.
1.6
28
23
-------
For toluene we estimate a fairly substantial conversion efficiency of
7 percent. This value was established by both electrical aerosol analyzer
data and gravimetric determinations of aerosols collected during smog-
chamber experiments. Since toluene reacts almost exclusively with HO,
the efficiency term is indicated under the HC+ HO column in Table 3.
Under toluene, as well as under the other hydrocarbons listed, we have
indicated a rate constant for the precursor reactions thought to be
significant in each case. Thus by comparing the rate constants for two
hydrocarbons (with a particular radical) coupled to the respective
efficiency factor, one can appreciate the relative importance of the
hydrocarbons to produce organic aerosols.
For butane, the efficiency factor is estimated to be very small,
and its participation (as well as that of many alkanes) in aerosol production
can be neglected. For propylene, which reacts much faster with HO, we
nonetheless estimate a very small efficiency value for HO reactions, and
this process can certainly be neglected. With ozone, however, a small
but measurable ability to make aerosol is observed. Overall though,
propylene can make only a very small contribution to the organic aerosol
problem.
Because 1-heptene is a larger molecule than propylene it is
appreciably more efficient in aerosol formation. Here again there is
experimental evidence that the efficiency is greater upon reaction with
0, compared to OH, but the distinction is not clear-cut since the
reaction profiles are not clearly separable. By coupling the efficiencies
with the respective rate constants for the 0- reactions, 1-heptene is
estimated to produce about 25 times as much aerosol as propylene. Cyclo-
hexene is very unusual in its efficiency in producing aerosol for reasons
discussed earlier. Here the efficiency value is a fairly crude estimate
based on light scattering and gravimetric measurements.
As a final example of this analysis let us look at aerosol
production in Figure 8 where a representative urban mixture of 17 hydro-
carbons is irradiated at near-ambient conditions. Here too we arrive
at an approximate efficiency value for the hydrocarbon mixture by
24
-------
\
I
« 1.0
0.8
|
0.6
I
*»
o"
S 0.4
z
o*
Alkanes
. [NMHC]Q=3.04 X I03/ig/m3; 5.34 ppm .,35
[N0x]o=0.63 ppmV
NMHC/NOX = 8.5 (ppmC/ppmV)
34567
Irradiation Time, hr
25 .i
20
8
o
O
15 I
.0 I
O
10
FIGURE 8. PHOTOCHEMICAL AEROSOL FORMATION DURING A
SMOG-CHAMBER IRRADIATION OF A SURROGATE
HYDROCARBON MIXTURE AND NO
X
averaging the hydrocarbon oxidation rates and dividing it into the aerosol
formation rate. These quantities are summarized below.
Smog-Chamber Simulation
q
Maximum total HC oxidation rate - 400 yg/m /hr (13% hr)
3
Maximum aerosol formation rate - 10 yg/m /hr
Maximum conversion efficiency = 2.5 percent
Urban Conditions
3 3
Assume [total HC] = 3 ppmC (1.7 x 10 yg/m )
Assume rvn = 4 hr
no o
Maximum aerosol concentration after 4 hr ~ 20 yg/m
25
-------
Dividing the maximum aerosol formation rate by the maximum total hydro-
carbon oxidation rate results in an overall maximum conversion efficiency
for the hydrocarbon mixture of 2.5 percent. If we extrapolate these
findings to polluted urban conditions, as indicated above, and' assume
a typical mean hydrocarbon lifetime of 4 hours, we would predict a
2
maximum organic aerosol concentration after 4 hours of 20 yg/m . Admittedly,
much of this analysis is handwaving, but judging from the fairly good agree-
ment between the simulated production of organic aerosols and the actual
concentrations observed in urban areas it seems reasonable to conclude as
follows:
(1) On the average, only a very small percentage
(2-3) of the hydrocarbon that gets oxidized
in the urban atmosphere ends up as aerosol,
and it is possible to estimate the efficiency
of certain hydrocarbons to make organic
aerosols.
(2) Higher molecular weight olefins and aromatics
are principally responsible for organic aerosol
formation. Olefin-0^ and aromatic-HO reactions
appear to be the important precursor reactions
in each case.
(3) The smog chamber appears reliable in simulating
aerosol formation in photochemical smog, i.e.,
the rates of aerosol formation in the chamber
are in accord with our expectations of the
polluted atmosphere.
INTERLABORATORY COMPARISONS
If smog chambers are reliable in simulating photochemical aerosol
formation then why do we see such divergent results among the different
chambers? And why, in some cases, do we see rates of oxidation so
26
-------
markedly different from the polluted atmosphere? In Los Angeles
for example, 55 percent of the olefins are consumed in a 4-hour period
(0800-1200) while in the Calspan and U of Minn, chambers only a few
percent of 1-hexene was oxidized in 4 hours. We feel that the answers
to these questions are attributable to actual differences in chemical
composition and other conditions of the simulated atmospheres compared to
the real atmosphere. These differences may sometimes appear to be subtle,
but they are believed to have rather profound effects on the radical
concentrations responsible for aerosol production.
Let us examine some of the differences in reaction conditions
which might explain the divergent rates of photooxidation observed
between Calspan Run No. 5 (Figure 5) with 1-hexene and a Battelle
experiment (Figure 7) with 1-heptene. Comparisons of some initial
conditions and a few reactivity results are tabulated in Table 4.
TABLE 4. COMPARISONS OF SMOG-CHAMBER CONDITIONS AT CALSPAN AND
BATTELLE AND SOME REACTIVITY RESULTS OF OLEFIN PHOTO-
OXIDATION
Calapan
Battelle
Initial Conditions
Chamber volume, ft3
Lamps
Light intensity (kd), min
Hydrocarbon, ppmV
NO, ppmV
N02, ppmV
N02/NOX
HC/NOX, ppmV/ppmV
RH, percent
Photooxidation Parameters
v> min
-1
Appearance time for aerosol
vol. min
Aerosol production rate
(dv/dt) max, um3/cm3/hr
Aerosol conversion rate
normalized to Calspan 's
pollutant concentrations
20,800 610
blacklamps, sunlamps, blacklamps, sunlamps
whitelamps
0.23 0.47
0.33 5.1
0.152 0.72
0.014 0.76
0.08
1.98
41
42°
300
2.1
2.1
0.51
3.44
66
36
15
160
10.4
27
-------
It is apparent in this case that there are large differences in initial
reactant conditions as well as in experimental conditions. In the
categories of light intensity and HC/NOx ratio, the Battelle conditons
are more favorable toward reactivity than Calspan's. The Battelle-to-
Calspan ratio in these categories is 2/1 and 1.7/1, respectively.
Another important factor here is the N0,,/N0x ratio. The effect of this
parameter on photooxidation rates is shown in Figure 9 for the 1-butene-
NO systetn(27). On the basis of the data in Figure 9, the difference
between 8 percent N02 for Calspan and 50 percent N02 for Battelle could
result in an additional factor of 2 difference for the ^O^max. rate.
Coupling these three terms results in a predicted photooxidation rate
6.8 times greater for the Battelle conditions versus the Calspan con-
ditions.
tzo
so
60
40
20
- o
_L
I
10 20 30 40
Percent N02 in NO, (Total NO, ~l ppm)
SO
EFFECT OF N02/NOX RATIO ON RATE PARAMETERS
1-Butane lyttem.
FIGURE 9. EFFECT OF N02/NOX RATIO ON PHOTOOXIDATION RATE
PARAMETERS IN THE 1-BUTENE-NO -SYSTEM
X
28
-------
In addition to the difference in the appearance time of aerosol
in the two experiments, there is a tremendous difference in the maximum
aerosol production rate. Differences in this rate should, however, be
viewed after normalizing for the pollutant concentrations, and having
done so, we see the aerosol formation rate is about 5 times greater for
the Battelle experiment, in accord with the other differences in reactivity.
The use of a higher molecular weight olefin (1-heptene) in the Battelle
experiment is yet another reason to expect higher aerosol concentrations
compared to the Calspan results. Thus, a large number of factors are
important in attempting to conpare secondary aerosol results from different
laboratories. If we had a better understanding of the chemical processes
involved in aerosol formation, we might be able to provide a more accurate
accounting for the differences.
One factor, namely nitrous acid (whose initial concentration
is related to background air purification and chamber surfaces) is
believed to be highly variable from chamber to chamber and may well account
for some reactivity differences observed between seemingly similar experi-
mental conditions. In the absence of light, nitrous acid (HONO) forms
in the atmosphere and in smog chambers according to reaction la
la h
NO. + NO + H,0 2 HONO . (1)
2 2 ID
Decomposition of HONO, reaction Ib, limits its concentration.
There is considerable evidence ~ that equilibrium concentrations of
nitrous acid exist in the Battelle chamber prior to irradiation. If so,
HO will be generated from HONO photolysis, and hydrocarbon oxidation is
expected to occur at the moment of irradiation. The high rate of HO
attack causes hydrocarbons to become radicals which both oxidize NO and
regenerate HO to continue the chain sequence. This pattern of immediate
oxidation is in contrast to that where no HONO exists as the irradiation
29
-------
begins. In this case, N0_ photodissociates, and nearly all of the 0 atoms
produced serve to oxidize NO back to NO,,. If the hydrocarbons in the
system are not successfully attacked by very low concentrations of 0
atoms or 0- the generation of HO radicals (which are much more likely
to react with hydrocarbons) proceeds rather slowly*". In instances
where aromatic hydrocarbons are involved, it appears that a very long
induction period to hydrocarbon oxidation might result where no HO
source (e.g., HONO) was present initially. Indeed this seems to be the
case with both the Calspan experiments and the U of M experiments.
Because of the heterogenebus nature of reaction la** and the very large
volume of the Calspan chamber, it is easy to understand why appreciable
HONO is not formed prior to irradiation. In the U of M chamber where
*There are usually two reactions which predominate in HO production:
HONO + hv > HO + NO (2)
H02 + NO > HO + N02 (3)
Therefore, aside from nitrous acid formation via reaction (la), we
must look for other sources of HONO and those for H02- The only
important HONO sources are (4)
HO + NO(-ttl) > HONO(+M) (4)
which leads to no net increase in HO radicals, and (5)
H02 + N02 > HONO + 02 (5)
which requires H02 radicals. The principal primary source of H02 is
aldehyde photolysis, for example (6)
CH20 + hv > H + HCO, (6)
followed by reactions (7)
H + 02(+M) -* H02(+M) + HCO (7)
and (8)
HCO + 02 > H02 + CO . (8)
Likewise, the reaction of alkoxy radicals with oxygen produces H02;
CH30 + 02 > H02 + CH20 . (9)
However, since the secondary reactions of CH30 and H usually require
an earlier reaction of HO (H from HO + CO > H + C02 and RO from the
sequence HO + RH * H20 + R, R + 02 > R02, R02 + NO > RO + N02)
we must look back at the sources of HO and stress the importance of the
initial [HONO] in initiating photooxidation reactions in relatively
unreactive systems.
**The heterogeneous nature of reaction (la) to form HONO is suspected on
the basis of kinetic data which show the reaction to be progressively
slower as the reaction vessel is enlarged. The original rate constant
obtained by Wayne and Yost for reaction (la) is 4.3 x 10~6 ppm~2min~l(53).
Using a chamber 40 times larger Graham and Tyler(54) obtained a much
smaller value; 1.2 x 10~9 ppm~2min~l. Recently Calvert and associates
at O.S.U. have observed consistent homogeneous behavior corresponding
to a rate constant of 2.1 x 10-9 ppm~2min~l(55).
30
-------
the S/V ratio is much greater, one would expect appreciable HONO formation
unless mixing is poor or the Teflon surfaces are not conducive to the
heterogeneous reaction. While there seems to be some basis for expecting
different HONO concentrations in the Calspan and University of Minnesota
chambers, that alone would not seem to adequately account for the
differences observed in reactivity for toluene and m-xylene in the experi-
ments compared earlier (Table 2). In the U of M chamber the initial
[HONO] was perhaps significant enough to overcome the light intensity
advantage of the Calspan chamber. For the experiments with olefins, light
intensity was probably the dominate factor in accounting for the reactivity
differences in these two chambers.
In a final analysis of smog-chamber performances, an important
question to be addressed is, how do the rates of hydrocarbon oxidation
in the polluted atmosphere compare with the smog-chamber results? This
subject is treated in detail in the Discussion section of the report.
The limited data from hydrocarbon mixtures indicate that the rates
of oxidation of alkanes in the Battelle chamber are nearly identical
(49) 3
to those found in the Los Angeles atmosphere and in the 6-m glass
chamber at Riverside . Compared to these two sources of rate data,
olefin oxidation may be 50 percent greater in the Battelle chamber.
On the average, the oxidation rates for aromatic hydrocarbons are
about twice as great in the Battelle chamber as they are in the Los
Angeles atmosphere (avg. rate, 0800-1200) and perhaps 50-100 percent
greater than those observed in the Riverside chamber. Thus, the
photooxidation rate data from the Battelle chamber, which have been
emphasized in much of this discussion, are thought to be somewhat greater
than the rates in the real atmosphere.
31
-------
SECTION 5
EXPERIMENTAL APPROACH
In the planning of this program, much consideration .was given
to the appropriateness of including S02 and primary aerosols in the
initial experimental program. Arguments were presented that nuclei,
either those provided as primary aerosols or as secondary sulfuric-acid
aerosols, might be necessary to cause nucleation of the organic vapors
at low concentrations, and that experimental simulations of authentic
organic aerosol formation might be meaningful only if these important
variables were included. On the other hand, inclusion of S02 in the
reactant mixture would not permit an accurate assessment of organic
aerosol formation because analytical methods were not sophisticated
enough to distinguish quantitatively between organic and sulfate aerosols.
It was also well known that, although the HC-NO constituency of smog
X
profoundly effects the rate of SO- oxidation, the corollary is not true;
i.e., the presence of SO- in a HC-NO mixture has virtually no effect on
A, X
the rate of hydrocarbon oxidation or even NO oxidation*. This is primarily
due to the fact that hydrocarbons, particularly those involved in organic
aerosol formation, out compete S02 for reactions with radicals by factors
of 10 to 100. Furthermore, unpublished data from our laboratory indicate
that sulfuric acid aerosols and organic aerosols are formed independently
when mixtures of hydrocarbons, NO and SO- are irradiated in a smog chamber.
X £,
For these reasons, S02 was excluded from the reaction mixture.
Primary aerosols were also excluded from the initial program for
several reasons. First, the generation and control of primary aerosols
is difficulty particularly where contamination from gases must be precluded.
Secondly, the additional concentration of surface area provided by primary
aerosol (generally 1-5 x 10^ ym2/cm3 for aged aerosol) is very small
relative to the surface/volume ratio of the smog chamber (2.6 x 106
ym2/cm3 for our 17.3 m3 chamber). Thirdly, and most importantly, :
*For example, confer reference 52 in which experimental evidence is
presented showing insignificant effect of 0.5-3 ppm SO, on propylene
and NO photooxidation rates.
32
-------
results of our studies of secondary aerosol formation from auto exhaust
teach that while primary aerosols do provide the surface upon which
secondary aerosol preferentially condenses they do not significantly
effect the degree of secondary organic aerosol formation, and they are
not necessary to cause nucleation of organic vapors.
The principal effect of including primary aerosol is to alter
the amount of light scattering attributable to secondary aerosol formation.
An example is shown in Figure 10 in which auto exhaust was irradiated in
the absence and presence of primary aerosol. With the exception of
100
to 80
«»
i6°
u
£ 40
20
Filtered auto exhaust
Unfiltered auto exhaust
Surface (4000/zm2 cm'3)
Light Scattering (10 XIO"4 rrr1)
I 23456
Irradiation Time, hr
FIGURE 10. THE EFFECT OF PRIMARY AUTO EXHAUST AEROSOLS
ON PHOTOCHEMICAL AEROSOL GROWTH AND LIGHT
SCATTERING
the differences in primary aerosol concentration, the experiments were
essentially identical, and the ensuing gas-phase reactions were also
similar. The volume-concentration curves in Figure 10 indicate that
the volumetric gas-to-aerosol conversion rates are fairly similar during
the two experiments, and that at the end of the irradiations the difference
33
-------
in total volume is nearly equal to the initial difference, i.e., the
volume of primary aerosol. The difference in light scattering caused by
primary aerosol can be rationalized on the basis of the consequent
differences in the size distribution of the secondary aerosols and the
strong dependence of light scattering on this parameter. Interpretative
(30 39 50)
details of these data have been presented elsewherev ' ' ' and will
not be repeated here.
The propensity of organic vapor to nucleate under smog conditions
is further illustrated in Figure 11 where auto exhaust (low sulfur fuel)
was irradiated in a smog chamber. In this case, the surface distribution
of aged primary-exhaust aerosol is represented by the hatched area. At
O.I
Particle Diameter (Dp),
FIGURE 11. EVIDENCE OF PREFERENTIAL HOMOGENEOUS
NUCLEATION OF PHOTOCHEMICALLY DERIVED
AEROSOL IN AIR CONTAINING PRIMARY NUCLEI
34
-------
the onset of irradiation we see that, in spite of the presence of primary
aerosol surface, homogeneous nucleation of new aerosol occurred as is
evident by the additional mode in the surface distribution at 0.03 ym. The
aerosol formed in this "nucleation mode" is soon consumed by collisions
with aerosols in the "accumulation mode" (0.1 to 1 pm) and thereafter it
appears that all new aerosol growth occurs by condensation of vapor on
the aerosol existing in the accumulation size region.
In summary, it was felt that the initial goal of establishing
definitive relationships among the hydrocarbon and NO precursors of
X
organic aerosols could best be achieved by irradiating pollutant
mixtures of hydrocarbons and NO in otherwise clean air. Because the
X
behavior of hydrocarbons in photochemical smog cannot be adequately
simulated by a single hydrocarbon, a surrogate mixture of 17 hydrocarbons
was used to simulate polluted urban atmospheres. Water vapor and CO
were also added at constant levels to constitute what is referred to as
a "reference atmosphere". The distribution of the pollutants in the
reference atmosphere, -including the hydrocarbons employed and the atmo-
spheric hydrocarbons they represent, are indicated in Table 5. The hydro-
carbon mixture was formulated mainly from the atmospheric data of
Stephens ^57\
Seventeen experiments were conducted varying the total hydro-
carbon and NO concentration. The experimental design is shown in
X
Figure 12. All irradiations were conducted for 10 hours.
35
-------
TABLE 5. REFERENCE ATMOSPHERE
Carbon Monoxide » 2.5 ppm
Nitrogen Oxides (total) = 0.100 ppra
Nitric oxide - 0.083 ppm
Nitrogen dioxide - 0.017 ppm
Nonmethane Hydrocarbons =1.00 ppm as
Hydrocarbons
Represented
acetylene
ethane
propane
2-methylpropane
n-butane
2-methylbutane
n-pentane
2 , 2-dimethylbutane "|
2-methylpentane I
2,3-dimethylbutane [
n-hexane J
ethylene
propylene
1,3-butadiene "1
1-butene 1
trans-2-butene >
cis-2-butene I
2-methylpropene J
2-methylbutene-l "^
2-methylbutene-2 >
trans-2-pentene J
benzene
toluene
ethylbenzene "1
p-xylene 1
m-xylene |
o-xylene J
isopropylbenzene "^
n-propylbenzene 1
p-ethyltoluene >
m-ethyltoluene I
o-ethyltoluene J
1,3,5-trimethylbenzene "1
1,2,4-trimethylbenzene >
1,2,3-trimethylbenzene J
Reference
Hydrocarbons
acetylene
ethane
propane
2-methylpropane
n-butane
2-methylbutane
n-pentane
2-methylpentane
ethylene
propylene
trans-2-butene
2-methylbutene-2
benzene
toluene
m xylene
p-ethyltoluene
1 , 2 ,4-trimethylbenzene
Molar
Los Angeles Air
.177
.087
.036
.024
.100
.066
.036
.063(C)
.129
.039
.033
i018(c)
.024
.054
.069(c>
.024(c)
.021(c>
Concentrations Relative
to Total NMHC
(a) Experimental Air(b)
.136
.100
.040
.023
.099
.070
.037
.044
.162
.035
.043
.013
.029
.061
.069
.025
.013
1
(a) Composition derived from that reported by E.R. Stephens
(b) Initial concentrations from Run No. 8.
(c). The sum of the concentrations for groups of similar hydrocarbons in Los Angeles air
are indicated opposite the reference hydrocarbon.
36
-------
0 2 4 6 8 10 \Z 14 16
Initial Nonmethane Hydrocarbon Concentrations, ppmC
FIGURE 12. INITIAL HYDROCARBON AND NITROGEN OXIDE
CONCENTRATION COORDINATES IN THE
EXPERIMENTAL PROGRAM
37
-------
SECTION 6
EXPERIMENTAL METHODS
SMOG-CHAMBER DESCRIPTION
AND OPERATION
All irradiation experiments were conducted in Battelle-Columbus'
3 -1
17.3-m smog chamber having a surface-to-volume ratio of 2.6 m ; the
surface is polished aluminum and FEP Teflon^ Direct irradiation through
5-mil Teflon windows is provided by a bank of 95 fluorescent blacklamps
and 15 fluorescent sunlamps. The photon flux of the blacklamps is dis-
tributed unimodally in the uv region, with peak intensity at 370 mm; the
sunlamp peak intensity occurs at 310 nm. Light-intensity measurements by
/co\ (59}
NO photolysis^ ' and o-nitrobenzaldehyde photolysisv ' agree quite
well, as described by Gordon^ . Prior to the first series of experiments
new fluorescent blacklamps were installed, and the k, value was 0.48 min .
Four months later when the second series of experiments was conducted, the
light intensity had diminished to a k, value of 0.41 min .
Background air supplied to the chamber is taken through a 10-m
stack atop a three-story building and is passed through a purification
system which includes a permanganate filter bed, a charcoal filter system,
an absolute filter, and a humidification unit. After purification, back-
ground total hydrocarbon is generally 2-3 ppmC, with the majority being
methane. Nonmethane hydrocarbons (relatively unreactive) are <0.2 ppmC,
j _-j
NO <0.02 ppm, CO <4 ppm, and particles <10 cm .
2£
Prior to each series of experiments, the chamber surfaces were
thoroughly cleaned by washing with water. After cleaning, the chamber
was dried by continuous purging with purified air, and then conditioned
by prolonged irradiation of background air.
All experiments were conducted for about 10 hours. Typically,
the chamber was first humidified with deionized, double-distilled water
vapor followed by consecutive injections of NO, NO-, CO, a low molecular-
weight hydrocarbon mixture (C2-C,), a high molecular-weight hydrocarbon
mixture (C^-Cg), and tracer (SF-). The inert tracer was added to determine
the dilution rate. Continuous and intermittent sampling of the chamber
38
-------
air together with a small unavoidable leak rate results in dilution of
the original air volume. Makeup air is the same as the purified back-
ground air. For experiment Nos. 1-12, the dilution rate averaged about
8 percent/hr; for experiments 13-19, the rate was near 13 percent/hr.
The chamber air is well mixed with a stirring fan during the injection
period. The stirring fan is turned off when irradiation begins.
ANALYTICAL
The gas-phase chemistry of the smog experiments was monitored
with conventional instrumentation. Carbon dioxide was determined by non-
dispersive IR, 0 by chemiluminescence with ethylene, NO and NO- by
automated Saltzman using a dichromate oxidizer for NO oxidation, CH, and
total NMHC by flame ionization using a dual-flame analyzer. The latter
analysis was used primarily to indicate the approximate hydrocarbon con-
centratins during chamber loading.
Detailed hydrocarbon analyses were obtained hourly with two
flame ionization gas chromatographs. The C, to C- hydrocarbons and 2-
methylpropane were chromatographed on a Duropatc^phenylisocyanate column
(10-ft long, 0.06-in. i.d. aluminum tubing) immersed in a wet ice bath.
The sample size was 5 cc. The other C, hydrocarbons and all those >C,
were chromatographed on a capillary column (300-ft long, 0.01 in. i.d.
stainless steel tubing) with programmed temperatures from -100 to 136 C.
The sample size was 20 cc. SF, was determined by electron-capture gas
chromatography. The analysis was performed on a silica gel and carbosieve
column (3-ft long, 0.06 in i.d. stainless steel tubing) maintained at 120 C.
The sample size was 1 cc. Figure 13 is a reproduction of a typical
chromatogram showing good resolution. Only the propylene peak was
troublesome in that integration was sometimes inaccurate at low concen-
trations. Typically, unknown hydrocarbon concentrations (excluding the
impurity in the helium carrier gas) were in the range 0.01-0.5 percent
by weight.
The ozone instrument was calibrated by the neutral-buffered-KI
method. The N0-N0_ analyzer was calibrated by an Og-NO titration
procedure^61). The chromatographs were calibrated each day from a NBS
certified bottle of propane in nitrogen.
39
-------
Capillary Column (DC-ZOO)
20 cc Sample
Ouropak (phenylisocyonote)
5cc Sample ""
100
zoo
300
400
600
700
800
900
1000
1100
1500
1600
1700
1800
1900
Approximate Retention Time, sec
FIGURE 13. REPRESENTATIVE CHROMATOGRAM SHOWING RESOLUTION OF THE
SURROGATE HYDROCARBON MIXUTRE OBTAINED WITH TWO GAS
CHROMATOGRAPHS
-------
Aerosol measurements were made with an integrating nephelometer
and an electrical aerosol analyzer (EAA). Aerosol growth into the light-
scattering size range occurred in only a couple of experiments, so the
EAA data was the principal method of aerosol analysis. The EAA measures
in situ the size distribution of aerosols in the 0.005 to 0.3-ym diameter
size range. The instrument operates on the principle of unipolar electric
diffusion charging followed by mobility analysis. All data are based on
the recent calibration data reported by Liu and Pui^ . The application
of this instrument in numerous atmospheric aerosol studies has been reviewed
by Willeke and Whitby^62\
Data from the EAA were examined to determine if substantial
truncation errors existed due to the analyzer's cut-off size at 0.3 ym
diameter. Assuming a log-normal distribution of aerosol volume in the
0.03 to 0.3 ym-diameter range, the projected aerosol volume extending
beyond the 0.3-ym size range was always< 10 percent of the total integrated
volume and thus no corrections for truncation were necessary.
Examples of the development of photochemical aerosol under the
conditions employed are shown in Figure 14 where the changes in the
surface-area and volume-concentration size distributions are plotted
against irradiation time.
41
-------
AP-OOH 1 1 -25-TH
a
IS
TSO.
500.
250
0.001
0.010 0.100
DPI , UP)
SURFACE DISTRIBUTION
I . 000
AP-001 11-25-71
to. o
0. 001
0 . 0 I 0 0.100
DPI, un
VOLUME DISTRIBUTION
1 .000
FIGURE 14. COMPUTER-GENERATED GRAPHS OF THE CHANGES IN THE AEROSOL SURFACE-
AREA AND VOLUME-SIZE DISTRIBUTION THAT OCCUR AS A FUNCTION OF
IRRADIATION TIME
42 .
-------
SECTION 7
RESULTS
At the request of EPA, a comprehensive file of data was prepared
as a supplement to this report*. In this section of the report we have
included summary tables of initial experimental conditions and results
pertinent to our discussions. In addition, smog profiles (continuous
time-concentration profiles of NO, N02> 03> and aerosol) are presented
in Appendix A, and cumulative hourly profiles of hydrocarbon depletion
are presented in Appendix B.
The relative composition of the atmosphere irradiated in each
experiment was approximately that described in Table 5. The only intended
variables in the experiments were the total NMHC and the total NO con-
x
centrations. Efforts were made to maintain constant distributions among
the hydrocarbon and N0x (NO and N0£) mixtures. The measured initial
concentrations of the reactants are presented in Table 6. There was
some inadvertent variation in [CO] , but this should not have had a
substantial effect on the results of interest. According to the data
in Table 6, there are also slight variations in the initial hydrocarbon
distributions, but in view of the calibrated volume injection procedure
employed, these variations may reflect analytical inaccuracies as much as
actual discrepencies. Here, too, small variations in the relative dis-
tributions of these reactants are thought to have been inconsequential.
*The data file consists of both magnetic tapes and conventional computer
printouts. One magnetic tape (No. 230) contains all the gas-phase data,
with the exception of the gas chromatographic data, acquired during the
course of the smog experiments. A second magnetic tape (No. 268) con-
tains the aerosol-size-distribution data for all samples taken during
the experiments. This record includes tabulations of the surface-area
and volume size distribution of each sampling and an integrated value
for the total number, total volume, and total surface area concentrations.
As requested, 9-track tapes were prepared at 800 BPI, odd parity in EBCDIC
code with no labels. Instructions for reading the tapes were included in
the package. ' The gas chromatographic data is complied entirely as printed
output and bound separately. The output includes the concentration of
each hydrocarbon (expressed as ppmC, percent carbon, ppmV, and percent
volume) for every hour of the irradiation. For each experiment there
are summary tables of the average rates of decay fitted to first-order
kinetics.
43
-------
TABLE 6. INITIAL POLLUTANT CONCENTRATIONS
(a)
Carbon monoxide
nitrogen oxides
Nitric oxlda
Nitrogen dioxide
Honmethane hydrocarbons (as CH.)
acetylene
ethylene
propylene
trans- 2-butene
2-methyl-2-butene
ethane
propane
n-butane
2-aethylpropane
n-pentane
2-aechylbutane
2-methylpentane
benzene
toluene
m-xylene
p-ethyltoluene
1,2,4-trimethylbenzene
' ' " "" .mi..
1
16
0.63
0.50
0.13
5.742
0.401
0.528
0.125
0.211
0.091
0.314
0.195
0.563
0.110
0.249
0.477
0.340
0.406
0.575
0.753
0.249
0.146
IWBM*H^"V*^V^^^MV
2
10
0.28
0.23
0.05
5.642
0.335
0.520
0.109
0.202
0.088
0.310
0.196
0.541
0.128
0.257
0.489
0.352
0.237
0.594
0.857
0.279
0.140
3
14
1.75
1.43
0.32
7.221
0.555
0.683
0.224
0.313
0.129
0.414
0.253
0.756
0.181
0.270
0.514
0.385
0.544
0.631
0.898
0.304
0.159
^^^^HI»^H^^to-V
4
14
1.18
0.99
0.19
6.244
0.452
0.526
0.112
0.236
0.099
0.278
0.192
0.610
0.125
0.297
0.561
0.414
0.273
0.715
0.883
0.294
0.168
»^^MIMV*-w4f>hv
5
14
0.30
0.25
0.05
3.731
0.247
0.290
0.123
0.126
0.057
0.174
0.110
0.358
0.072
0.175
0.326
0'.248
0.163
0.419
0.552
0.185
0.097
PWH*^B^^M-l«^-W-
6
15
0.60
0.51
0.09
3.482
0.266
0.300
0.078
0.117
0.052
0.194
0.126
0.318
0.079
0.148
0.276
0.207
0.149
0.364
0.507
0.187
0.105
H^UHfumftm m mim
7
15
1.19
0.99
0.20
3.957
0.265
0.299
0.094
0.134
0.058
0.179
0.113
0.416
0.093
0.181
0.353
0.255
0.177
0.420
0.573
0.214
0.126
MVVVMI^BBMM
Run
8
14
0.58
0.48
0.10
14.294
0.964
1.150
0.370
0.606
0.238
0.715
0.426
1.409
0.324
0.658
1.252
0.933
0.628
1.529
1.944
0.711
0.429
^VMI^»^MB^
Number
9
14
1.16
0.96
0.20
13.922
1.047
1.153
0.372
0.589
0.228
0.699
0.431
1.349
0.344
0.628
1.185
0.889
0.600
1.478
1.825
0.679
0.420
M~^W^^I^
10
14
0.30
0.25
0.05
14.136
1.039
1.193
0.406
0.576
0.227
0.727
0.471
1.329
0.353
0.614
1.175
0.874
0.587
1.459
1.881
0.741
0.475
13
10
1.93
1.52
0.41
14.352
0.932
1.105
0.333
0.609
0.226
0.665
0.433
1.390
0.327
0.622
1.213
0.864
0.591
1.508
1.994
0.901
0.631
14
12
0.99
0.83
0.16
3.826
0.216
0.317
0.088
0.152
0.060
0.222
0.131
0.432
0.107
0.179
0.332
0.228
0.148
0.358
0.486
0.215
0.146
15
10
0.60
0.50
0.10
1.766
0.110
0.148
0.029
0.036
0.027
0.099
0.060
0.189
0.040
0.079
0.154
0.109
0.072
0.170
0.245
0.109
0.080
16
12
0.29
0.25
0.04
1.819
0.119
0.146
0.032
0.046
0.026
0.100
0.056
0.179
0.034
0.078
0.155
0.115
0.081
0.200
0.263
0.107
0.075
17
11
0.16
0.13
0.03
1.847
0.131
0.152
0.025
0.044
0.027
0.111
0.063
0.189
0.041
0.077
0.155
0.114
0.075
0.212
0.252
0.104
0.068
18
15
0.15
0.13
0.02
3.457
0.221
0.273
0.080
0.119
0.048
0.179
0.106
0.336
0.082
0.151
0.293
0.211
0.144
0.371
0.486
0.204
0.142
19
10
0.15
0.13
0.02
7.600
0.520
0.588
0.176
0.348
0.121
0.356
0.246
0.801
0.178
0.326
0.640
0.444
0.299
0.774
1.000
0.451
0.324
(a) All concentration units are ppra (vol/vol); hydrocarbon units expressed as ppm CH, or pp«C.
-------
Experimental results are summarized in Table 7. The reactivity
parameters are defined by footnotes. Three principal manifestations, the
concentrations of ozone and aerosol and the hydrocarbon depletion rates
were corrected for dilution of the smog chamber. This was necessary
because the dilution rate varied somewhat from run-to-run. The dilution
rate was particularly great (~13%/hr) for Run Nos. 13-19 due to a small
leak in a Teflon window that went undetected.
The units of aerosol volume concentration used throughout this
3 3
report, urn /cm , are convenient in that they correspond to familiar mass
o
concentration units of pg/m if the density of the aerosols is unity.
Unless otherwise specified units of ppm and ppb refer to parts-per-million
or parts-per-billion by volume (ppmV and ppbV) while ppmC refers to hydro-
carbon concentrations of parts-per-million equivalent in carbon atoms to
methane; e.g., 1 ppm propane = 3 ppm as CH^ or 3 ppmC,
45
-------
TABLE 7 . SUMMARY OF EXPERIMENTAL RESULTS
Conditions
NMHC, NOX, NOfc-tmax. N02
Smog Reactivity Parameters
rate.<">
Run No. ppmC ppm min ppb/min~l
1 5.74 0.63 74 '
2 5.64 0.28 38
3 . 7.22 1.75 210
4 6.24 1.18 150
5 3.73 0.30 45
6 3.48 0.60 112
7 3.96 1.19 285
8 14.29 0.58 38
9 13.92 1.16 85
10 14.14 0.30 22
13 14.35 1.93 150
14 3.83 0.99 157
15 1.77 0.60 180
16 1.82 0.29 90
17 1.85 0.16 65
18 3.46 0.15 43
19 7.60 0.15 23
(a) Time to reach the maximum [NC^J .
(b) t[N02]fflax - [N02li}/time to lN02]max.
(c) Maximum [0^] corrected for the smog-chamber
(d) NMHC depletion rate corrected for dilution;
7
7
5
5
8
3
3
13
12
11
6
2
1
2
2
4
5
dilution
the data
[o ]
ppIB
0.60
0.42
0.16
0.48
0.51
0.53
0.14
0.50
0.82
0.45
0.92
0.30
0.30
0.43
0.39
0.35
0.40
rate.
fitted to
(e) Aerosol volume inferred from the size-frequency distribution of
, HC rate,
Z/hr
13.4
11.2
11.4
12.5
15.8
13.8
12.6
10.2
17.5
5.2
11.7
12.3
8.7
12.7
12.3
12.8
5.2
a first-order
Aerosol Volume
(d) Concentration, pm3/cm3
I
2 hr
9.9
14.6
2.6
5.2
11.8
6.8
2.6
29.3
12.5
24.6
6.2
2.2
1.8
5.8
10.1
9.9
12.6
4 hr
16.6
17
6.6
9.9
15.0
10.2
4.4
30.5
19.8
23.6
12.8
5.6
5.4
9.1
13.5
13.1
12.6
8 hr
20.3
17
10.1
14.3
15.8
13.3
5.8
29.7
24.3
23.3
19.0
8.7
9.1
14.3
14.2
13.6
12.6
Aerosol*
rate
lim'/cm'/hr
6.2
11.5
2.7
4.2
6.5
4.2
2.1
32.0
10.0
25.5
5.1
2.2
2.5
3.5
3.2
8.5
14.4
decay expression.
aerosols assuming spherical
shape; the volume concentrations corrected for dilution.
(f) Maximum aerosol volume formation rate during the 10-hour irradiation; the rates corrected
for dilution.
-------
SECTION 8
DISCUSSION
OVERALL REACTIVITY
Before turning to the discussion of aerosol precursor relation-
ships it is of interest to comment on some measures of reactivity in
general and make some comparisons with other smog-chamber and atmospheric
results.
Linear regressions of two-variable combinations of various
reactivity parameters were performed on our data, and the results appear
in Table 8. It is noted that the rate parameters designated (A), (G), (H),
and (I) are concentration normalized, i.e., the dimensions do not contain
a concentration term. The other rate parameters, (B) and (F), do include
a concentration term, and, of course, parameters (C), (D), and (E) have
concentration units.
TABLE 8. CORRELATION COEFFICIENTS AMONG MEASURED REACTIVITIES
Parameters
(A)
(B)
(C)
(D)
(E)
(F)
(G)
(H)
(D
N0--t
2 max
N0_-rate
0--max
Aerosol-4 hr
Aerosol-8 hr
Aerosol-rate
NMHC-rate
Olefln-rate
Aromatic-rate
(A) (B) (C) (D) (E) (F) (G)
1.0 -0.82 -0.36 -0.73 -0.63 -0.61 0.
1.0 0.04 0.64 0.44 0.71 -0.
1.0 0.47 0.67 0-19 0.
1.0 0.95 0.89 -0.
1.0 0.77 0.
1.0 -0.
1.
18
44
29
12
02
47
0
(H)
0.
-0.
0.
o.
0.
-0.
0.
1.
06
14
25
19
24
03
77
0
(I)
0.
r-0.
0.
-0.
-0.
-0.
0.
0.
1.
05
38
18
32
22
64
78
26
0
(a) Dimensions are: (A), (G), (H), and (I) time; (B) and (F) concentration/time;
(C), (D), and (E) concentration.
47
-------
A normalized rate parameter for NO photooxidation (N0,-t)
uiclX
correlates fairly well with the absolute NO-photooxidation rate (K02~rate)
and the aerosol parameters (D, E, and F), but rather poorly with peak
ozone (C) and the fractional rates of hydrocarbon decay (G, H, and I).
Rather surprisingly, NO--rate shows no improvement in correlating with
maximum 00 concentrations, and the correlations with the rates of hydro-
(6)
carbon decay were low and negative. In a study by Heuss and Glasson
with individual hydrocarbons, the correlation of N02~rate with peak 0^
was 0.61, and between N0_-rate and percent hydrocarbon reacted (parameter G)
it was 0.56. Perhaps the higher correlations in their work are related to
constant initial concentrations of reactants.
Maximum ozone concentration does not correlate highly with the
other measures of reactivity, although there is fairly good correlation
with aerosol concentrations at 8 hours irradiation. One sees from the
smog profiles (Appendix A) that aerosol formation often precedes 0_
formation, and even less correlation between these dependent variables
would be expected for instantaneous data. These relationships will be
discussed more fully in the sections to follow.
Aerosol concentrations at 4 and 8 hours are well correlated
with each other and with the maximum rate of aerosol formation, but
they are not correlated with the fractional rates of hydrocarbon decay.
The latter finding is not surprising since the hydrocarbon parameters are
normalized. A more meaningful comparison can be made on the basis of the
total amount of hydrocarbon reacted and the amount of aerosol produced.
Therefore, linear regressions were performed between time integrals of
hydrocarbon decay and the concentrations of aerosols at the respective
4 hr
time limits (e.g., / d(HC)/dt vs aerosol volume concentration at 4 hr).
0 hr
The results are shown in Table 9. While none of the correlations are
especially good there is considerable improvement over the previous analysis.
In all categories the correlation coefficients increase with irradiation
time.
48
-------
TABLE 9. CORRELATION COEFFICIENTS BETWEEN AEROSOL
CONCENTRATION AND THE TIME INTEGRALS OF
HYDROCARBON DECAY
Integral Integral Period
Variable 2 hr 4 hr 8 hr
Olefins
Aromatics
NMHC
0.50 0.65 0.76
0.33 0.53 0.70
0.27 0.48 0.68
HYDROCARBON OXIDATION
In nearly all of the experiments conducted, the observed hydro-
carbon depletion rates could be fitted satisfactorily to first-order
kinetics. Therefore, fractional first-order rates are used throughout
the report in summarizing the hydrocarbon data. Examples of the decay
rates are shown in Figure 15. The rates were corrected for the first-
order dilution of the chamber air and they therefore can be no more accurate
than the determinations of the dilution rate. As indicated in Figure 15,
benzene, acetylene, ethane, and propane are oxidized very slowlygenerally
at rates <1 percent/hr. Other alkanes, ethylene, and toluene decayed at
rates in the range of 2-10 percent/hr. The ethylene and toluene rates
were often quite similar. The other olefins and aromatics disappeared
at substantially higher rates.
49
-------
Benzene (k=0.004hr'')
'Acetylene*! .
Ethane Kk=O.OI2 hr"1)
.Propane J
n-Butane(k=OX3l7hr"')
-I,
2-Methylbutane(k=0.029 hr'}
2-Methylpentane(k=0.052 hf1)
Ethylene(k=0.057hr"')
Toluene (k=0.067hr~')
2-Methylpropane(k=0.082 hr*1)
p~Ethyltoluene(k=O.I99 hr'1)
1,2,4-Trimethylbenzene (k=0.4l hr")
Trans-2-butene(k=Q69hr"')
/Propylene(k=0.76hr-l) .
ll2-Methyl-2-butene(k=0.79hr~')
I I I I I
4 68
Irradiation Time, hr
10
FIGURE 15. FRACTIONAL HYDROCARBON DECAY RATES AT
9.1/1 NMHC/NO RATIO, RUN NO. 1
There were two peculiar results which were consistently observed.
(1) Propylene disappeared at unusually rapid rates. In many cases this
appeared to be the result of peak broadening and inaccurate electronic
integration at low concentrations. (2) At initial NMHC concentrations
<4 ppaC, the ethylene concentration actually increased late in the
irradiations. We confirmed that this anomaly was not due to ethylene
leaks to the chamber or to ethylene in the make-up air. The possibility
of ethylene being produced via aldehyde photolysis has been discussed
50
-------
by Altshuller, et al. ', but its presence was not detected in their
work. Other explanations appear equally speculative.
Table 10 summarizes some of the pertinent data on hydrocarbon
disappearance rates when typical urban mixtures of hydrocarbons are
irradiated naturally or with artificial sunlight. Footnotes (a)-(c)
describe three studies cited for comparison with our smog-chamber results.
In the study with Los Angeles air, the average NMHC/NO ratio was 8.8/1.
JL
In the other study with actual urban air [footnote (b)], the ratio was
not stated. Smog-chamber experiments at NMHC/NO ratios of 9.1/1, 4.1/1,
X
and 24/1 were selected for comparison. All of the rate data in Table 10
are normalized with respect to the rate of n-butane decay. Measured rates
for n-butane are given in footnote (d).
In comparing first the rates of hydrocarbon decay in the three
smog-chamber experiments of this study, there is remarkable similarity
in the overall rates in view of the wide range of NMHC/NO ratios. It
*x
is important to note that the absolute rates for butane are also very
similar, as are the relative rates for the other alkanes (propane and
2-methylbutane) and acetylene, ethylene, and benzene. Because the
propylene data is questionable it will not be compared. 2-methyl-2-
butene shows slightly increased decay rates with increasing NMHC/NOx
ratios. Rather interestingly, the alkylbenzenes all show a maximum
rate of decay for the NMHC/NO condition of 9.1/1. This trend is further
X
illustrated in Figure 16 where the average fractional decay rate of all
aromatics is plotted against the NMHC/NOx ratio for 17 experiments.
Although there is considerable scatter because of the dependency on the
absolute hydrocarbon and N0x levels, there appears to be a trend of
maximum decay rate near the NMHC/NO ratio of 10/1. The same type plot
X
for total olefins did not reveal a reasonable trend.
51
-------
TABLE 10. HYDROCARBON OXIDATION RATES IN POLLUTED AIR
AND IN SMOG-CHAMBER SIMULATIONS
Oi
N>
(a)
Los Angeles
Air
NMHC /NO (ppmC/ppmV ) : 8.8
X
Acetylene
Ethylene
Propylene
2-Methyl-2-Butene
Propane
n-Butane'd'
2-Methylbutane
Benzene
Toluene
m-Xylene
1,2, 4-Trimethylbenzene
0.5
3.7
8.6
34.7
0.6
1.00
1.6
-
2.1
4.2
Oxidation Rates Relative to n-Butane
Cb\ (c)
Riverside vy Riverside v ' Battelle Chamber This Study
Air Smog Chamber Run No. 1 Run No. 3 Run No. 8
7.7 9.1
0.7
3.8 - 3.3
16.1 - 44
46
0.7
1.00 1.00 1.00
1.8^ - 1.7
<1 0.23
1.4 3.9
7.5 13.8
11 24
4.1 24
0.7 0.5
4.0
26 43
44 75
0.4
1.00 1.00
1.6 1.6
<0.2 0.23
2.5 1.9
10.1 10.3
20 18
(a) Downtown L.A. air
collected at 0800 and
4-hr avg (0800-1200) (49) .
(b) Central Riverside
rates are 8-hr avg
air collected at 0630
(0730-1530) (57).
irradiated naturally in Tedlar bags. Oxidation rates are
and irradiated naturally in borosilicate
1 3
(c) Surrogate HC mixture irradiated with blacklamps (k^ = 0.20 min ) in 6-m chamber
carboys . Oxidation
at U of Calif.
(Riverside). Oxidation rates are 2-hr avg(56).
(d) Rates normalized to n-butane. Absolute rates for n-butane are: ref. (a), k = 0.023 hr (6-hr avg);
ref. (b), not given; ref. (c), k = 0.023 hr"1 (estimated from published data, 2-hr avg); run no. 1,
k = 0.017 hr"-1- (10-hr avg); run no. 3, k = O.013 hr"1 (lO-hr avg); run no. 8, k = O.O17 hr"1 (10-hr avg).
-------
o
25
20
.a 15
"5
I I0
?£
J L
10
_L
NMHC(ppmC)/NOx
100
FIGURE 16. EFFECT OF NMHC/NO RATIO ON THE RATE
OF AROMATIC HYDROCARBON DECAY
With Run No. 1 data (NMHC/NO = 9.1) as the comparable Battelle-
X
chamber data, it appears that the absolute rate for butane decay is in
satisfactory agreement with the results of the Los Angeles air study,
with which the Riverside-chamber rate is in perfect agreement. It must
be kept in mind here that the rates reported with natural irradiation
[studies footnoted (a) and (b)] are averages over a period of variable
irradiation intensity while the rates reported from the Riverside and
Battelle chambers are average rates over periods of constant irradiation
intensity. For acetylene, ethylene, 2-methyl-2-butene, propane, and
2-methylbutane there is good agreement between the Battelle-chamber data
and the Los Angeles air data. However, at the 9.1/1 NMHC/NO ratio, the
Jb
rate of toluene disappearance was nearly twice as large in the Battelle
chamber, and the rate for m-xylene was about 3 times larger than the rate
observed in Los Angeles air. At the NMHC/NOx ratios of 4.1/1 and 24/1,
the rates were more comparable.
The Riverside air study [footnote (b)] showed good agreement
with the Los Angeles air study for the limited data. In the Riverside
53
-------
smog-chamber study [footnote (c)] the decay rate reported for toluene is
somewhat less than that measured in Los Angeles air, but the rate for
m-xylene is nearly a factor of 2 larger. The ratio of the rates
of m-xylene to toluene are about 5/1 in the Riverside chamber,' and they
ranged from 3.5/1 to 5.5/1 in the Battelle chamber. In Los Angeles
air the ratio was only 2/1.
As a final indication of the comparability of the smog-chamber
data with the atmospheric data, averages of the decay rates of paraffins,
olefins, and aromatics are shown in Table 11. Based on the'se averages,
there is good agreement between the decay rates for paraffins, the olefiri
rate is somewhat higher (perhaps inaccurately higher because of propylene
uncertainties) in the Battelle chamber, and the aromatic rate is about a
factor of 2 greater in the Battelle chamber when the data are compared to
the Los Angeles atmospheric rates averaged over the period of 0800-1200
hours at full sunlight intensity.
TABLE 11. AVERAGE HYDROCARBON LOSS RATES UNDER NATURAL AND
SIMULATED IRRADIATION CONDITIONS
Hydrocarbon
Class
Paraffins
Olefins
Aromatics
Hydrocarbon Decay
Los Angeles AirW
Natural Irradiation
2.8
20
8.5
Rate, percent/hr
Battelle Smog Chamber (b)
Blacklamp Irradiation
3.0
37
15
(a) Reference No. 49
(b) Run No. 1, this study.
54
-------
In conclusion, we feel that the data obtained in this smog-
chamber program are highly representative of that associated with intense
photochemical smog conditions which manifest in some urban areas. Since
the photochemically induced rates of organic aerosol formation have
never been measured in polluted atmospheres direct comparisons of the
smog chamber's aerosol data are not possible. Although the correlation
results are not impressive, it nonetheless seems reasonable to presume
that organic aerosol formation in the smog chamber is closely related
to the rates of hydrocarbon oxidation. By deduction then it would appear
that the harmony observed between the hydrocarbon rate data in the atmosphere
and in the smog chamber lends credence to the relevancy of the aerosol
data presented next.
AEROSOL PRECURSOR
RELATIONSHIPS
The principal objective of this study was to establish the
relationships that exist between nonmethane hydrocarbon, nitrogen
oxides concentrations,and the subsequent development of photochemically
related aerosols. As discussed earlier, the relationships sought thus
far relate to the formation of organic aerosols and not to sulfate
aerosols. Experiments were conducted for 10 hours, and it is apparent
from the data that there are significant changes in the aerosol growth
dependency on NMHC and NO concentrations as the irradiations progress,
<2t
While irradiation time normally indicates the duration of a smog reaction
at constant irradiation intensity, it is possibly justifiable to think
of the irradiation period as the density or total flux of irradiation.
In other words, the results of a 2-hour simulation in a smog chamber may
be representative of smog conditions that would result at some reduced
level of irradiation on a cloudy day.
Two methods were adopted for illustrating the simultaneous
effects of the independent variables (NMHC and N0x) on the dependent
aerosol variables. In one case, 2-dimensional contour diagrams are
drawn depicting isopleths of the response surface (dependent variable)
as a function of the independent variables. An example of this analysis
55
-------
is shown in Figure 17 where the maximum rate of formation of aerosol
volume is represented by the contour lines (isopleths) and NMHC and NO
X
concentrations are represented on the abscissa and ordinate, respectively.
Each contour line represents intervals of the rate of aerosol volume
3 3
formation in units of 2 ym /cm /hr. The second graphical method
involves making projections of the response surface as it would appear
in 3 dimensions. With this method a realistic impression of the response
is conveyed but at some sacrifice of the numerial value of the surface
height. However, since the emphasis in our interpretations is on relative
functions and values, the 3-dimensional projections seem to be particularly
descriptive. Figure 18, for example, shows the aerosol rate data (same
as Figure 17) as a surface projection. In viewing these illustrations
it is important to establish the proper orientation. In Figures 17 and 18,
[NMHC] and [NO ] both increase in the directions away from the 0 point.
X
In all projections, the response surface is in a positive-Z orientation*.
Figure 18 shows that the maximum rate of aerosol formation lies
along a NMHC/NO line of about 30/1 for [NMHC] <9 ppmC. Above 9 ppmC,
X
the crest in the surface shifts to a NMHC/NO ratio of 18/1. Although
X
the maximum aerosol formation rate is nearly linear with respect to[NMHC],
the [NMHC] regions of 0-3.5 ppmC and 9.0-14.25 ppmC appear to have
slightly greater inclinations. In the [NMHC] range 0-9 ppmC, the
o o
maximum aerosol formation rate normalized to NMHC is about 1.9 pm /cm /hr
per unit ppmC hydrocarbon.
Figure 18 also shows that the maximum aerosol formation rate
goes through a maximum with respect to initial [NO ], with NO showing
X X
a strong inhibition effect at the higher pollutant concentrations. In
the NMHC region between 0 and 3 ppmC, however, the maximum rate dependence
*The graphic surfaces in perspective were generated on a CDC 6400 computer
and are the culmination of a 3-step process. First, a triangulation program
(CNTOUR) was used to generate isopleths of the dependent variables from the
original data. Smoothing programs were used to smooth the data and generate
a symetric data base. The dimension of surface was then produced with a
computer program (SRFACE) obtained from the National Center for Atmospheric
Research.
56
-------
14.25
N0x(ppm)
NMHC(ppmC)
FIGURE 17;. ISOPLETHS OF MAXIMUM RATES OF AEROSOL FORMATION AS A FUNCTION
OF THE INITIAL CONCENTRATIONS OF NMHC AND NOX (Isopleths
correspond to intervals of volume production rates of
2 }ra3/cm3/hr.)
14.25
N0x(ppm)
NMHC(ppmC)
FIGURE 18 . A SURFACE PROJECTION REPRESENTING MAXIMUM RATES OF AEROSOL
FORMATION AS FUNCTIONS OF THE INITIAL CONCENTRATIONS OF
NMHC AND NOX
57
-------
on [NO ] is not so pronounced. These precursor trends will be examined
X
in more detail as we look at the relationships between the instantaneous
aerosol concentations as functions of [NMHC], [NO], and irradiation time.
^v
Surface diagrams and contour plots of aerosol concentrations
at 2-hour, 6-hour, and 10-hour irradiation periods are shown in Figure 19a,b,c
and Figure 20a,b,c, respectively. (Our discussions relate primarily to
the surface projections; the contours plots are included to provide
quantitative intervals of the dependent variables.) The surface depicting
aerosol concentrations at 2 hours shows relationships with [NMHC] and [NO ]
X
which are similar to those for the maximum aerosol formation rate (Figure 18).
This is because the maximum formation rate usually occurred during the
first 2 hours of the experiments. There are some subtle differences
however. Most noteworthy is the relatively greater dependence of the
2-hour aerosol concentration on the initial concentrations of both NMHC
and NO in the lower concentration regions. Here the NO dependence is
x x
particularly striking.
Figures 19b and 20b show the aerosol concentrations at 6 hours.
Compared to the 2-hour situation a much expanded surface area has emerged
corresponding to N0x dependence. In other words, the inhibiting effect
of NO on aerosol formation becomes less significant as the irradiation
^v
time increases. It is also interesting to compare the position of the
ridges of maximum aerosol concentration at 6 hours and 2 hours. At 2 hours,
the ridge follows a NMHC/NO line of 15/1 up to [NMHC] of about 3 ppmC.
J±
The ridge then flattens out and turns toward a much higher NMHC/NO
X
ratio (34/1). At NMHC concentrations of about 7.5 ppmC, the ridge rises
again to a peak. At 6 hours, the ridge follows a NMHC/NO line of 13/1
X
up to [NMHC] of 5.5 ppmC, and then it turns and follows a ratio line of
44/1.
At 10 hours, (Figures 19c and 20c) the aerosol "mound" fills out
further than at 6 hours, but the surface and ridge seem to maintain the
shapes established at the 6-hour period. The NMHC/NO ratio at peak
X
aerosol concentration is 10/1 for [NMHC] <7.5 ppmC, and it approaches
infinity for higher pollutant concentrations.
58
-------
a. 2 hours
N0x(ppm)
14.25
NMHC(ppmC)
b. 6 hours
N0x(ppm)
14.25
NMHC(ppmC)
c. 10 hours
N0x(ppm)
14.25
NMHC(ppmC)
FIGURE 19. SURFACE PROJECTIONS REPRESENTING AEROSOL VOLUME CONCENTRATIONS
AS FUNCTIONS OF THE INITIAL CONCENTRATIONS OF NMHC AND NO AT
IRRADIATION TIMES OF 2, 6, AND 10 HOURS
59
-------
a. 2 hours
NOK(ppfn
NMHC(ppmC)
b. 6 hours
N0x{ppm)
NMHC(ppmC)
c. 10 hours
N0x(ppm)
NMHC(ppmC)
FIGURE 20. ISOPLETHS OF AEROSOL VOLUME CONCENTRATION AS FUNCTIONS OF INITIAL
CONCENTRATIONS OF NMHC AND N0x AT IRRADIATION TIMES OF 2, 6,
AND 10 HOURS (Isbpleths correspond to volume concentration
intervals of 2 ynH/cm3.)
60
-------
Aside from concentrating on the overall structure of the response
surfaces and the crests of maximum response it is important to examine
trends corresponding to distributions of NMHC and N0x that currently
exist in our polluted atmospheres .and to the distributions predicted for
future years. Unfortunately it appears that there is no typical NMHC/NO
distribution among major urban areas where smog is a problem. The
reliability of much atmospheric data has been questioned, and the reasons
given for the apparent wide ranges of NMHC/NO ratios are controversial
A
and will not be dealt with here. A few examples of atmospheric data with
which we are familiar are shown in Table 12. References of data sources
are indicated next to the sampling year.
TABLE 12. SELECTED DATA ON THE NMHC AND NO
DISTRIBUTIONS IN URBAN AREAS ?
Sampling
Site
Welfare Is. (NY)
St. Louis
South Coast Basin
Dayton, Ohio
(downtown)
New Carlisle, Ohio(b)
Year
1972<1)
1973<64>
1973(65)
1974<66)
1974<66)
Averaging
Period
20 days
5 days
90 days
(many stations)
30 days
30 days
Average
NMHC,
ppmC
2.6
0.62
3.9(1.7)(a>
1.76
0.67
Average
N0x,
ppm
0.098
0.055
0.14
0.105
0.022
Average
NMHC/NO
2C
26/1
11.3/1
12.1/1
16.7/1
30.4/1
(a) Total hydrocarbon reported at 3.9 ppmC. NMHC estimated at 1.7 ppmC.
(b) Semirural area 30 miles NE of Dayton, Ohio.
During the past couple of years it appears that NMHC/NO ratios
* X
have been >10/1 in most areas. Let's arbitrarily select 10/1 as a ratio
to examine the NMHC and NO effects on aerosol formation. To do this
we have "sliced" the response surfaces corresponding to the 10/1 NMHC/NO
61
-------
section, and we have removed part of the mound to expose the section face*.
The sections are shown in Figure 21a,b,c as a function of time from two
vantage points; nearly normal to the ordinate (NO ) and nearly normal to
X k
the abscissa (NMHC). Judging from the 2-hour data, the response surface
(aerosol concentration) at the 10/1 ratio is constant over a large range
of NMHC and NO concentrations. Only when the NMHC and N0x concentrations
become small is any reduction in the aerosol concentration noticeable.
For 6-hour irradiations the trend changes (Figure 21b). Here the aerosol
concentration is also constant at high pollutant concentrations, but there
is a gradual reduction in aerosol concentration for NMHC and N0x levels
<5 and 0.5 ppm, respectively. However, the reduction in aerosol concen-
trations becomes precipitous only where NMHC and NOX concentrations become
<2 and 0.2 ppm, respectively. Similar trends are obvious for the 10-hour
irradiation periods. Again only moderate reduction in aerosol concentration
occurs until low pollutant concentrations are reached. At 2 ppmC NMHC
levels, the aerosol concentration increases 60 percent during the period
from 2 hours to 6 hours, and 90 percent during the period from 2 hours to
10 hours of irradiation. Thus most of the organic aerosol is formed
during the more typical irradiation period of 6 hours.
All predictions of the direction of future NMHC/NO ratios are
X
toward lower values due primarily to emphasis on hydrocarbon emission
controls. To estimate the effect of these atmospheric trends, we have
rather arbitrarily sliced the response surface to reveal the 5/1 NMHC/NO
X
section. The results are shown in Figure 22a,b,c. For the 2-hour
irradiation period at the 5/1 ratio, Figure 22a indicates that aerosol
concentrations actually increase slightly with decreasing NMHC and NO
*The computer program does not permit perfect slicing rather only sectioning
of the smallest dimensions of the array. Thus where truncation is used to
reveal a particular edge of the response surface, array points are accepted
or rejected based on integral values, and a jagged edge results. The
heavy lines outlying the sections are interpolations between the array
points, and they are particularly necessary in interpreting the data at
very low concentrations where the number of significant figures becomes
seriously limited. Sometimes the interpolation curve lies across the
peaks of bisection edges and sometimes is passes through the valleys.
62
-------
14.25
u
E
a
a
14.25
N0x(ppm)
2 hours
1.9.
NMHC(ppmC)
N0x(ppm)
NMHC(ppmC)
14.25
E
a,
14.25
b. 6 hours
14.25
u
£
2
w
u
5
N0x(ppm)
NMHC(ppmC)
14.25
c. 10 hours
FIGURE 21. SURFACE PROJECTIONS REPRESENTING AEROSOL VOLUME CONCENTRATIONS
AS FUNCTIONS OF INITIAL POLLUTANT CONCENTRATIONS AT A CONSTANT
NMHC/NOX RATIO OF 10/1 AND IRRADIATION TIMES OF 2, 6, AND 10 HOURS
63
-------
14.25
NMHC(ppmC)
a. 2 hours
14.25
o
o
X
s
NMHC(ppmC)
14.25
N0x(ppm)
b. 6 hours
14.25
o
E
a.
a.
O
I
NMHC(ppmC)
14.25
c. 10 hours
FIGURE 22. SURFACE PROJECTIONS REPRESENTING AEROSOL VOLUME CONCENTRATIONS
AS FUNCTIONS OF INITIAL POLLUTANT CONCENTRATIONS AT A CONSTANT
NMHC/NOX RATIO OF 5/1 AND IRRADIATION TIMES OF 2, 6, AND 10 HOURS
64
-------
concentrations down to a point of maximum aerosol concentration near
the region of 2 ppmC NMHC and 0.4 ppm N0x> Then there is a nearly
linear reduction in aerosol concentration as zero pollutant concentrations
are approached. At 6 hours (Figure 22b), the plateau of maximum aerosol
concentration still persists until NMHC and NO are reduced below 2.0 and
X
0.4 ppm, respectively. At the 10-hour irradiation period (Figure 22c),
the picture is unchanged except that the aerosol concentrations have
increased slightly.
In the plateau regions of constant aerosol concentrations along
the specified NMHC/NO ratios, the relative reduction in aerosol concen-
X
tration in going from 10/1 to 5/1 ratios is only about 25 percent.
In conclusion, we see that the dependence of aerosol concentration
goes through a maximum with respect to initial NO concentrations, partict-
X
ularly at low NO concentrations or high NMHC/NO ratios. As the irradiation
X X
exposure increases from 2 hours to 6 hours and 10 hours, the NMHC/NO
X
ratios corresponding to peak aerosol concentrations change from 15/1 to 13/1
to 10/1, respectively, in the pollutant concentration ranges common to our
atmosphere. At higher pollutant concentrations the ratios at peak aerosol
concentrations are much higher. At NMHC/NO ratios >10/1, there is a
X
strong dependence of aerosol formation, on the initial pollutant levels.
In general, the pollutant level effect is more pronounced as the NMHC/NOX
ratio increases. At ratios <10/1, the pollutant loading effect is slight
except at [NMHC] <2 ppmC. Looking back at Figure 19b and c, we see that the
pollutant concentrations in the atmosphere must get into the regions of
NMHC <2 ppmC and NO <0.2 or >0.6 ppm before photochemical aerosol
X
formation is greatly suppressed.
OZONE PRECURSOR
RELATIONSHIPS
The results of several smog-chamber studies have provided
guidance in establishing the relationships of hydrocarbons and nitrogen
oxides in the formation of ozone in smog. The results of a study by
65
-------
Romanovsky, et al. , reproduced in Figure 23, well established the
relative roles of hydrocarbon and nitric oxide with respect to peak 0^
concentration. Computer simulations of 03 formation in smog have also
been useful. Simulation results of Hecht are reproduced in Figure 24.
N-butane (75%) and propylene (25%) were used in the computer simulation.
Propylene was the hydrocarbon employed in the Romanovsky study. There
are similarities in the trends of 0 dependency shown by the data in the
two studies, but, owing to different conditions and assumptions, there are
major differences in the quantitative results.
2 I -
3456
Propylene, ppm
0.8
O)
T3
'x
O
0.4
0.2
0 0.4 0.8 1.2 1.6 2.0
Total Hydrocarbon (Butane + Propylene), ppm
FIGURE 23. ISOPLETHS OF CONSTANT
OZONE CONCENTRATION
(ppm) DEVELOPED FROM
PEAK OZONE CONCENTRA-
TIONS IN AN EARLIER
SMOG-CHAMBER STUDY
Romanovsky, et a
FIGURE 24. ISOPLETHS OF CONSTANT
OZONE CONCENTRATION
(ppm) BASED ON 5-HR
DATA PREDICTED BY A
KINETIC SMOG MODEL
Hecht(67).
Several studies have demonstrated that simplified smog systems
containing only one or two hydrocarbons do not adequately simulate the
smog manifestations representative of actual urban conditions. Thus,
66
-------
more realistic smog-chamber experiments have been conducted with auto
exhaust emissions or surrogate mixtures of typical 6-9 a.m. hydrocarbon
distributions in cities. The smog-chamber results used by the Los Angeles
County Air Pollution Control District (LACAPCD) to predict future trends
/go\
in peak ozonev are reproduced in Figure 25 alongside a drawing (Figure 26)
of our results. Direct comparisons of the LACAPCD results with the history
of smog episodes in that area have shown that their smog-chamber model
underestimates actual peak ozone concentrations. Efforts to adjust the
model to fit atmospheric data have met with criticism^ .
2.0
I I / I
I .2 .3 .4 .5
5 10
NMHC.ppmC
5 10
NMHC.ppmC
FIGURE 25. ISOPLETHS OF CONSTANT
OZONE CONCENTRATIONS
(ppm) DERIVED FROM THE
LACAPCD SMOG-CHAMBER
STUDIES
Hamming, et al.
(68)
FIGURE 26. ISOPLETHS OF CONSTANT
OZONE CONCENTRATIONS
(ppm) DERIVED FROM
INSTANTANEOUS OZONE
CONCENTRATIONS AT 6-
HR OF IRRADIATION
This Study.
67
-------
Our model shows higher 0_ concentrations for the corresponding
LACAPCD conditions, but it too is undoubtedly imperfect (no model eo
simple is expected to be extremely accurate). The "D" line in Figure 26,
a boundary established in a study by Dimitriades^ , represents a NMHC/NOx
ratio required to meet the present air quality standard for ozone.
Dimitriades' smog-chamber study utilized auto exhaust mixtures. His results,
at least those defining the "D" lines, are in accord with the results of
this study.
Presumably, atmospheric conditions resulting in worst-case
incidents of ozone occurrence are those where a highly polluted air mass
is confined in space throughout a day-long irradiation period. An air
mass stagnant over Los Angeles, for example, does not necessarily meet this
criteria because in the late afternoon automotive emissions are added to the
stagnant atmosphere under attentuated irradiation, and the additional NO
emission effectively reduces the afternoon 0., level. A condition more
nearly representative of a worst case occurs when a highly polluted air
mass from an urban area like Los Angeles travels into a more remote area
(like Riverside or Azusa), and the full ozone-forming potential of the
air mass is realized. This situation is akin to the smog-chamber conditions
where a static or moderately diluted condition is simulated over prolonged
irradiation periods.
Accepting the hypothetical similarity between the smog chamber
and atmospheric conditions we can compare to the model a few data points
that were reported as "worst case" incidents of ozone for the Pasadena
area in 1969 and 1970^ '. The data are shown in Table 13. In most
cases, the atmospheric data points are quite close to the 0~ concentrations
predicted by the smog-chamber model. Again, these data are not convincing
that the smog model is always accurate, but the agreement does provide
an additional element of confidence.
68
-------
TABLE 13. WORST-CASE OZONE EPISODES IN PASADENA (1969-1970) AND
THE PRECURSOR HYDROCARBON AND NO CONCENTRATIONS (a)
X
Initial Concentrations
Date
9-10-69
9-29-69
8-6-70
8-31-70
10-1-70
NMHC, ppmC
4.0
4.5
3.9
3.0
4.3
NOX, ppm
0.43
0.75
0.32
0.31
0.75
Ozone Maximum, ppm
Pasadena
0.60
0.59
0.56
0.51
0.52
6-hr Model Prediction
0.55
0.42
0.54
0.47
0.40
(a) Data are 6-9 a.m. NMHC and NOX concentrations measured in downtown
Los Angeles and maximum hourly average 03 measured in Pasadena on
days when the airflow trajectory was predominantly from Los Angeles
to Pasadena.
A more thorough appreciation of the ozone precursor model can be
gained by viewing 3-dimensional graphs as we did for the aerosol model.
One of the most interesting features of the data is the irradiation-time
effect on the NMHC/NO ratios corresponding to maximum ozone concentrations.
A
This is depicted in Figure 27a,b,c in which the response surfaces represent
the instantaneous 0- concentrations for all initial concentrations of NMHC
and NO . For quantitative reference isopleths of constant 0^ concentration
are presented in Figure 28a,b,c; each isopleth represents 0.05 ppm 0^. At
the 2-hour irradiation interval the ozone concentration is nearly zero for
low NMHC/NO ratios. The NMHC/NO ratio at peak ozone concentrations lies
x x
along the 28/1 plane over the entire range of precursor concentrations.
Thus there is a pronounced effect of N0x inhibiting 03 formation over this
irradiation period. By 6 hours, there are striking differences. In addition
to the response surface "swelling up" in the N0x region of the diagram, the
NMHC/NO ratio of the ridge (maximum 03> shifts to 11.5/1. At 10 hours, the
69
-------
a. 2 hours
N0x(ppm)
NMHC(ppmC)
14.25
b. 6 hours
14.25
c. 10 hours
NOx(ppm)
14.25
NMHC(ppmC)
FIGURE 27. SURFACE PROJECTIONS REPRESENTING OZONE CONCENTRATIONS AS
FUNCTIONS OF INITIAL CONCENTRATIONS OF NMHC AND NOX AT
IRRADIATION TIMES OF 2, 6, AND 10 HOURS
70
-------
a. 2 hours
N0x(ppm)
14.25
NMHC(ppmC)
b. 6 hours
N0x(ppm)
14.25
NMHC(ppmC)
c. 10 hours
N0«(ppm)
NMHC(ppmC)
FIGURE 28. ISOPLETHS OF OZONE CONCENTRATIONS AS FUNCTIONS OF INITIAL
CONCENTRATIONS OF NMHC AND NOX AT IRRADIATION TIMES OF
2, 6, AND 10 HOURS (Isopleths correspond to concentration
intervals of 0.05 ppm 03.)
71
-------
ridge swings further in the NO direction to a NMHC/NO ratio of 8/1. Thus
X X
as the irradiation time is extended, the inhibiting effect of NO on peak 0-
(67)
continually diminishes in agreement with the modeling results of Hecht
Hecht points out that true suppression of 03 occurs only when all the
reactive hydrocarbon is consumed without complete conversion of NO to ^2-
Although somewhat academic it is interesting to note that at
very high NMHC/NO ratios the 0, dependency on NMHC goes through a maximum
X j
at all irradiation periods. At the 2-hour, 6-hour, and 10-hour periods the
maximum occurs near [NMHC] of 7 ppmC, 4.5 ppmC, and 2.3 ppmC, respectively.
Such conditions are possibly relevant to rural situations where high HC/NOx
ratios may be encountered. The dependency of ozone on NO also goes through
X
a maximum, as it did for aerosol, with the functionality broadening with
increasing irradiation time and increasing pollutant concentrations.
To illustrate the changes in the 0- precursor relationships at
constant NMHC/NO ratios, ozone response surfaces were "sliced and exposed"
X
at 10/1 and 5/1 ratios. The results are presented in Figures 29a,b,c and
30a,b,c; the ordinate (NO ) and the abscissa (NMHC) are the vantage points
X
in each pair of graphs. Presumably, this type of illustration is becoming
more familiar and self explanatory.
Looking first at the 10/1 data, one sees that at 2 hours there
is a 0,. plateau which does not decline until NMHC and NO concentrations
J X
< 3 ppm and < 0.3 ppm are reached. At 6 hours, there is a gradual
dependence of 0» on the pollutant concentration corresponding to [NO ]
j X
< 0.8 ppm and [NMHC] < 8 ppmC. At 10 hours, increasing O-j concentration
occurs with increasing pollutant concentrations over the entire range of
initial concentrations, but the slope is steep only for NMHC < 2 ppmC and
NO < 0.2 ppm.
X
At the 5/1 ratio, almost no 0_ is present at 2 hours of irradiation,
but the small peak which does exist occurs at relatively low pollutant
concentrations. At 6 hours, the ozone maximum is still below 0.3 ppm for
all pollutant concentrations. The peak ozone concentration goes through a
maximum with respect to the initial pollutant concentrationthe maximum 0,
occurring in the pollutant concentration range from 2 ppmC NMHC and 0.4 ppm
NO to 5 ppmC NMHC and 1 ppm NO .
X X
72
-------
14.25
o
E
a
a
o
X
s
N0x(ppm)
NMHC(ppmC)
14.25
2 hours
NMHC(ppmC)
14.25
b. 6 hours
14.25
o
e
a.
a.
^»*
O
X
5
z
NMHC(ppmC)
14.25
N0x(ppin)
c. 10 hours
FIGURE 29. SURFACE PROJECTIONS REPRESENTING OZONE CONCENTRATIONS AS FUNCTIONS
OF INITIAL POLLUTANT CONCENTRATIONS AT A CONSTANT NMHC/NOX RATIO
OF 10/1 AND IRRADIATION TIMES OF 2, 6, AND 10 HOURS
73
-------
14.25
o
E
a.
O.
O
X
s
NMHC('PP
-------
Results over prolonged irradiations of 10 hours are similar to
those at 6 hours except that an even greater range of constant maximum
ozone concentrations and a 25 percent greater maximum value are evident.
A precipitous decline in the ozone concentration does not occur until
pollutant levels < 1.5 ppmC NMHC and < 0.3 ppm NO are reached.
X
AEROSOL AND OZONEMUTUAL BENEFITS
FROM PRECURSOR CONTROLS
At first glance, the precursor relationships of aerosol formation
with NMHC and N0x may appear similar to those for ozone formation. In many
respects they are, however, the fact that peak aerosol and ozone concen-
trations do not correlate well is a clue that there must be substantial
differences. Side-by-side comparisons of the response surfaces of aerosol
and ozone for identical precursor conditions will be used to identify the
differences as well as the many similarities in the relationships, and
they will likewise be useful in estimating benefits anticipated from
precursor controls.
Figure 31a-f shows in parallel the overall aerosol and ozone
relationships to NMHC and NO at progressive irradiation times. The
relationships for aerosol and 0« are similar at 2 hours. High concen-
tration of NO show strong inhibition effects at this period, more so
X
for ozone than for aerosol. In both cases, the crest of maximum concen-
trations falls along a NMHC/NO section near 25/1.
* X
At 6 hours, substantial differences are apparent. The crest in
the ozone surface sweeps dramatically toward lower NMHC/NOx ratios, and
lower 0- concentrations appear where the crest was oriented at 2 hours.
By 10 hours, the crest has swept to a NMHC/NOx ratio of 8/1.
The crests in the response surfaces of aerosol concentrations
each contain bends over the range of pollutant concentrations studied.
At high concentrations, the crests are relatively invariant with respect
to irradiation time, but, at more common concentrations, the initial N0x
concentration becomes increasingly crucial with time, as is the case with
0_. The effect of irradiation time is less pronounced than for 03, however.
75
-------
14.25
NMHC(ppmC)
N0x(ppm)
NMHC(ppmC)
2 hours
b.
N0x(ppm)
14.25
NMHC(ppmC)
N0x(ppm)
NMHCtppmC
N0x(ppm)
6 hours
d.
14.25
NMHC(ppmC)
10 hours
NMHCtppmC)
FIGURE 31. COMPARISONS OF THE CONCENTRATION DEPENDENCE OF AEROSOL (a,c,e) AND
OZONE (b,d,f) VOLUME ON THE INITIAL CONCENTRATIONS OF NMHC AND NOX
AT IRRADIATION TIMES OF 2, 6, AND 10 HOURS
76
-------
The crest of peak aerosol concentration changes from a NMHC/NO ratio of
15/1 at 2 hours to 13/1 and 10/1 at 6 hours and 10 hours, respectively.
Smog profiles show that aerosol formation often precedes ozone
formation and that later in the irradiations the rate of aerosol formation
often diminishes markedly while that for 03 remains appreciable. For these
reasons, the maximum aerosol concentration at 2 hours is 93 percent of
the maximum at 10 hours, while for ozone the 2-hour maximum is only 65 percent
of the 10 hour maximum concentration.
Additional comparisons of the precursor relationships are made
by inspecting models at constant NMHC/NO ratios. Figure 32a-f shows
j\,
the relationships at a 10/1 ratio, and Figure 33a-f shows them at a 5/1
ratio. At 2 hours and at 10/1 ratio, both the aerosol and 0_ relationships
are nearly constant over a wide range of initial pollutant (NMHC and NO )
JL
concentrations, except at the relatively low concentrations. At 6 hours
and 10 hours (NMHC/NO of 10/1), the initial pollutant concentration is
X
somewhat more influential on the aerosol and ozone levels and to similar
degrees.
Looking at the data at 5/1 NMHC/NO ratios one sees that little
X
0- has formed compared to aerosol at 2 hours. Aside from the inverse
relationship between 0« concentrations and the initial pollutant levels
(NO range > 0.4 ppm) at the 2-hour irradiation period, both aerosol and
X
0_ are essentially insensitive to the initial pollutant concentrations
until relatively low pollutant concentrations are attained. Thus at 5/1
ratios little improvement in either aerosol or 03 concentrations is
realized until NMHC and NO concentrations are < 2 ppmC and 0.4 ppm,
X
respectively, and this condition holds over a wide range of irradiation
periods.
There are many ways of looking at precursor-control strategies,
and we will not attempt to discuss the ramifications of all possible
maneuvers. An approach considered by many as both practical and prudent
is one based on unilateral control of NMHC after achieving some reasonably
safe level of NO . For mean yearly NOX concentrations of 0.05 ppm, hourly-
average maximum concentrations of 0.35 ppm are often equated, and we will adopt
this NO concentration for purposes of assessing the effect of unilaterial
x
77
-------
14.25
o
o
2
14.25
o
E
Q.
_a
CJ
x
Z
14.25
o
I
*>^
o
a.
N0x(ppm)
c.
N0x(ppm)
e.
14.25
2 hours
6 hours
14.25
o
E
a
a.
O
X
5
10 hours
FIGURE 32. COMPARISONS OF THE CONCENTRATION DEPENDENCE OF
OZONE (b,d,f) VOLUME ON THE INITTAT. rrarnrMTDAT
78
-------
14.25
u
ex
a
**
u
z
NOx(ppmC)
a.
14.25
u
E
a
a
**
u
z
z
z
14.25
2 hours
14.25
o
E
o
X
s
N0x(ppm)
c.
6 hours
14.25
14.25
u
E
a
a
U
I
S
Z
U
£
o
X
s
z
e.
10 hours
FIGURE 33. COMPARISONS OF THE CONCENTRATION DEPENDENCE OF AEROSOL (a.c.e) AND
OZONE (b.d.f) VOLUME ON THE INITIAL CONCENTRATIONS OF POLLUTANTS
AT A CONSTANT NMHC/NOX RATIO OF 5/1 AND IRRADIATION TIMES OF 2, 6,
AND 10 HOURS
79
-------
NMHC control on both aerosol and 0, concentrations. To further limit the
discussion, only the data corresponding to 6-hour irradiations is selected.
(Presumably the models presented will permit the reader to make analyses
of additional control strategies, if desired). It should also be pointed
out that the smog-chamber models approximate worst-case conditions with
respect to both the initial pollutant concentrations and the smog
manifestations.
We begin the analysis by recording 0» data corresponding to 0.35 ppm
NO and 3.5 ppmC NMHC; i.e., at a 10/1 NMHC/NO ratio. At this point the
X X
[0_] * 0.5 ppm. If NMHC is reduced 50 percent ([NMHC] =1.75 ppmC and
NMHC/NO = 5/1) the model predicts a 50 percent reduction in 0. (0.25 ppm).
4V **
A 70 percent reduction in NMHC (NMHC/NO = 3/1) results in a 70 percent
X
reduction in 0_ (0.15 ppm), and an 80 percent reduction in NMHC (NMHC/NO -
j *
2/1) results in an 84 percent reduction in 0. which meets the 0.08 ppm
standard.
With the above control scheme applied, a 50 percent reduction in
NMHC (NMHC/NO = 5/1 at 0.35 ppm NO ) results in only a 28 percent reduction
" X
in aerosol concentration. Further reduction to 70 percent (NMHC = 1.05 ppmC)
results in a 57 percent decrease in aerosol, and an 80 percent control of
NMHC (NMHC =0.7 ppmC) reduces the aerosol concentration 71 percent.
In conclusion, it is satisfying to find that control strategies
designed to limit the photochemical formation of 0, are mutually beneficial
in limiting the formation of aerosols. Unfortunately, the model predicts
that the degree of benefit for aerosols will be less than that for 0-.
80
-------
REFERENCES
1. Miller, D.F.,,Schwartz, W.E., Gemma, J.L., and Levy, A., "Haze
Formation: Its Nature and Origin-1975", EPA-650/3-75-010, NERC,
Research Triangle Park, N.C. (1975).
2. Hidy, G.M., et al., "Characterization of Aerosols in Los Angeles",
Report on the ACHEX Study, Volumes I-IV, prepared for the California
Air Resources Board (1975).
3. Hamming, W.J. and Dickinson, J.E., J. Air Poll. Control Assoa. , 16,
317 (1966).
4. Brunelle, M.F., Dickinson, J.E., and Hamming, W.J., "Effectiveness
of Organic Solvents in Photochemical Smog Formation", Air Poll.
Control Office, Los Angeles County (1966).
5. Romanovsky, J.C., Ingels, R.M., and Gordon, R.J., J. Air Poll.
Control Assoa., 17, 454 (1967).
6. Heuss, J.M. and Glasson, W.A., Environ. Soi. and Tedhnol. , 2^
1109 (1968).
7. Altshiiller, A.P., Kopczynski, S.C., Wilson, D., Lonneman, W.A., and
Sutterfield, F.D., J. Air Poll. Control Assoa., 19, 791 (1969).
8. Dimitriades, B., "On the Function of Hydrocarbons and Nitrogen
Oxides in Photochemical Smog Formation", U.S. Bur. Mines Rep.
Invest. 7433 (1970).
9. Wilson, K.W. and Doyle, G.J., "Final Report on Investigation of
Reactivities of Organic Solvents", Contract No. CPA 22-69-125,
(September, 1970).
10. Levy, A., Miller, S.E., and Scofield, F., "The Photochemical Smog
Reactivity of Solvents", Second Int. Clean Air Congress, Academic
Press, New York, 1970, p. 305.
11. Altshuller, A.P. and Bufalini, J.J., Environ. Sci. and Teehnol. ,
.5, 39 (1971).
12. Glasson, W.A. and Tuesday, C.S., Environ. Sci. and Teehnol. ,
5_, 151, (1971).
13. Laity, J.L., Burstain, I.G., and Appel, B.R., "Photochemical Smog
and Atmospheric Reactions of Solvents", Adv. in Chemistry Series
124, Washington, D.C., 1973, p. 95.
81
-------
14. Dimitriades, B. and Wesson T.C., "Reactivity of Exhaust Aldehydes",
U.S. Bur. Mines Rept. Invest. 7527, 1971.
15. Dimitriades, B., Environ. Sai. and Technol., j3, 253 (1972).
16. Kopczynski, S.L., Altshuller, A.P., and Sutterfield, F.D., Environ.
Sai. and Technol. , JJ, 909 (1974).
17. Kopczynski, S.L., Kuntz, R.L., and Bufalini, J.J., Environ. Sai.
and Technol. , JJ, 648 (1975).
18. Prager, M.J., Stephens, E.R., and Scott, W.E., "Aerosol Formation
from Gaseous Air Pollutants", Ind. and Eng. Chem. , 52, 521 (1960).
19. Renzetti, N.A. and Doyle, G.J., "Photochemical Aerosol Formation
in Sulfur Dioxide-Hydrocarbon Systems", Int. J. Air Poll. , 2_,
327 (1960).
20. Stevenson, H.J.R., Sanderson, D.E., and Altshuller, A.P., Int. J.
Air Wat. Poll. , 9», 367 (1965).
21. Goetz, A. and Pueschel, R., J. Air Poll. Control Assoo., 15, 90 (1965).
22. Goetz, A. and Pueschel, R., Atmos. Environ., .1, 287 (1967).
23. Orr, C., Jr., Hard, F.K., and Corbett, W.J., J. Colloid Sai. , 13,
472 (1968).
24. Wilson, W.E., Merryman, E.L., Levy, A., and Taliaferro, H.R., "Aerosol
Formation in Photochemical Smog - I. The Effect of Stirring", J. Air
Poll. Control Assoc. , 21., 128 (1971).
25. Groblicki, P.J. and Nebel, G.J., "The Photochemical Formation of
Aerosols in Urban Atmospheres", in Chemical Reactions in Urban
Atmospheres, (C.S. Tuesday, Ed.) Elsevier, N.Y. (1971).
26. Wilson, W.E., "Aerosol Formation in Photochemical Smog - II. The
Role of Sulfur Dioxide", presented at the 161st National ACS Meeting,
Los Angeles, Calif., (1971).
27. Wilson, W.E., Miller, D.F., and Levy, A., "A Study of S02 in
Photochemical Smog - III". Battelle-Columbus Final Report (third year)
to the American Petroleum Institute, Committee for Air and Water
Conservation (Project 5-11), 1971.
28. Stephens, E.R. and Price, M.A., J. Colloid and Interface Sai.,
_39_, 272 (1972).
29. Wilson, W.E., Miller, D.F., and Levy, A., J. Air Poll. Control Assoc.,
23, 949 (1973).
82
-------
30. Miller, D.F. and Levy, A., "Environmental Chamber Studies of
Atmospheric Aerosols", EPA-650/4-74-009, NERC, Research Triangle
Park, N.C. (1973). S
31. Smith, J.P. and Urone, P., Environ. Sai. and Teohnol. , j$, 742 (1974).
32. Kocmond, W.C., Kittelson, D.B., Yang, J.Y., and Demerjian, K.L.,
Determination of the Formation Mechanisms and Composition of
Photochemical Aerosols", Calspan Corp., Report No. NA-5365-M-1,
Buffalo, New York (1973).
33. O'Brien, R.J., Holmes, J.R., and Bockian, A.H., Environ. Soi. and
Teohnol. , 9^ 568 (1975).
34. Doyle, G.J. and Renzetti, N.A., J. Air Poll. Control Assoc. , 8,
23 (1958).
35. Shuck, E.A., Ford, H.W., and Stephens, E.R., "Air Pollution
Effects of Irradiated Automobile Exhaust as Related to Fuel
Composition", Air Pollution Foundation Report 26, Oct., 1958).
36. Hamming, W.J., Mader, P.P., Nicksic, S.W., Romanovsky, J.C.,
and Wayne, L.G., "Gasoline Composition and the Control of Smog",
Los Angeles County Air Pollution Control District Report,
Sept., 1961.
37. Ripperton, L.A. and Jeffries, H.E., and White, 0., "Formation of Aerosols
by Reactions of Ozone with Selected Hydrocarbons", paper presented at
the 161st National ACS Meeting, Los Angeles, Calif. (1971).
38. Schwartz, W.E., Jones, P.W., Riggle, C.J., and Miller, D.F., "Chemical
Characterization of Model Aerosols", EPA-650/3-74-011, NERC, Research
Triangle Park, N.C. (1974).
39. Miller, D.F..and Levy, A., "Exhaust Hydrocarbon Relationships with
Photochemical Aerosol Formation", Paper No. 75-16.3 presented at the
68th Annual Meeting of the Air Poll. Control Assoc., Boston, Mass.
(1975).
40. Smith, J.H. and Wilson, K.W., "Motor Fuel Composition and Photo-
chemistry", Stanford Research Institute Final Report to the American
Petroleum Institute (1971).
41. Vardi, J., "Selected Fuel Factors and the Formation of Automotive
Photochemical Serosols and Oxidants", presented at the 66th Annual
Meeting of the Air Poll. Control Assoc., Paper No. 73-71 (June, 1973).
42. Kerker, M., The Scattering of Light and Other Electromagnetic Radiation,
Academic Press, New York, 1969-
83
-------
43. Liu, B.Y.H. and Pui, D.Y.H., J. Colloid and Interface Soi. , 47,
155 (1974).
44. Whitby, K.T. and Clark, W.E., Tellus, 18, 573 (1966).
45. Liu, B.Y.H., Whitby, K.T., and Pui, D.Y.H., J. Air Poll. Control
Assoo., 24, 1067 (1974).
46. Liu, B.Y.H. and Pui, D.Y.H., J. of Aerosol Sci., .6, 249 (1975).
47. Kocmond, W.C., Kittelson, D.B., Yang, J.Y., and Demerjian, K.L.,
"Study of Aerosol Formation in Photochemical Air Pollution",
EPA-650/3-75-007, NERC, Research Triangle Park, N.C. (1975).
48. Jefferies, H., Fox, D., and Kamens, R., "Outdoor Smog Chambers"
presented at the EPA Smog Chamber Conference, Research Triangle
Park, N.C. (1974).
49. Kopczynskl, S.L., Lonneman, W.A., Sutterfield, D.F., and Barley, P.E.,
Environ. Soi. Teohnol. , £, 342 (1972).
50. Miller, D.F. and Levy, A., "Aerosol Formation in Photochemical
Smog. The Effect of Humidity and Small Particles", Proceedings
of the Third International Clean Air Congress, Verein Deutscher
Ingenieure, Dusseldorf, Germany (1973).
51. Demerjian, K.L., Kerr, J.A., and Calvert, J.G., "The Mechanism of
Photochemical Smog Formation", in Adv. Environ. Sci. Technol.,
(J.N. Pitts and R.L. Metcalf, Eds.) John Wiley and Sons, New York
(1974) pp. 1-262.
52. Miller, D.F., "A Smog Chamber Study of the Rate of Conversion of
S02 as a Function of Reactant Concentrations", Annual Progress
Report from Battelle-Columbus to the EPA, in preparation (1976).
53. Wayne, L.G. and Yost, D.M., J. Chem. Phys., 19, 41 (1951).
54. Graham, R.F. and Tyler, B.J., J. Chem. Soo. Faraday I, 68, 683 (1972).
55. Calvert, J.G., Private communication (1975).
56. Doyle, G.J., Lloyd, A.C., Darnall, K.R., Winer, A.M., and Pitts, J.N.,
Environ. Soi. Teohnol. » JJ, 237 (1975).
57. Stephens, E.R., "Hydrocarbons in Polluted Air", Summary Report to the
Coordinating Research Council (Project CAPA-5-68), Statewide Mr
Pollution Research Center at Riverside, Calif. (1973).
58. Tuesday, C.S., Chemical Reactions in the Upper and Lower Atmospheret
Interscience, New York, New York(1961).'
84
-------
59. Pitts, J.N., Jr., Vernon J.M,, and Wan, J.K.S., Intern. J. Air and
Water Poll., 9., 595-600 (1965).
60. Gordon, R.J., 'Pilot Study of Ultraviolet Radiation in Los Angeles.
J.S. Nader (Ed.), National Air Pollution Control Administration
Durham, North Carolina (1965).
61. Hodgeson, J.A., Baumgardner, R.E., Martin, B.E., and Rehme, K.A.,
Anal. Chem. , 43, 1123 (July,1971).
62. Willeke, K. and Whitby, K.T., J. Air Poll. Control Assoo. , 25, 529
(1975).
63. Altshuller, A.P., Cohen, I.R., and Purcell, T.C., Can. J. Chem.,
44, 2973 (1966).
64. Spicer, C.W., "The Fate of Nitrogen Oxides in the Atmosphere",
Battelle-Columbus Report to the Coordinating Research Council
(CAPA-9-71) and the EPA, September, 1974.
65. Paskind, J. and Kinosian, J.R. , "Hydrocarbon, Oxides of Nitrogen and
Oxidant Pollutant Relationships in the Atmosphere Over California
Cities", paper presented at the 67th Annual Meeting of the Air Poll.
Control Assoc., Denver, Colo. (1974).
66. Spicer, C.W., Gemma, J.L., Joseph, D.W., Sticksel, P.R., and Ward,
G.F., "The Transport of Oxidant Beyond Urban Areas", Battelle-
Columbus Report to EPA, draft submitted May, 1975.
67. Hecht, T.A., "Smog Simulation Models and Their Use in Evaluating
Air Quality Control Strategies", paper presented at the Scientific
Seminar on Automotive Pollutants, EPA-600/9-75-003, Washington, D.C.
68. Hamming, W., Chass, R., Dickinson, J., and MacBeth, W., "Motor Vehicle
Control and Air Quality. The Path to Clean Air for Los Angeles",
presented at the 66th Annual Meeting of the Air Pollution Control
Assoc., Chicago, 111. (1973).
69. Souten, D.R., Hopper, C.J., and Mueller, R.Ii., "A Critical Review
of the Los Angeles County APCD Method for Simulating Atmospheric
Oxidant Based on Smog Chamber Irradiation Experiments", paper
presented at the 68th Annual Meeting of the Air Poll. Control
Assoc., Boston, Mass. (1975).
70. Kinosian, J.R., "Ambient Air Quality Trends in the South Coast Air
Basin", paper presented at the Scientific Seminar on Automotive
Pollutants, EPA-600/9-75-003, Washington, D.C. (1975).
85
-------
APPENDIX A
SMOG PROFILES
The profiles were drawn from the original data and are not
corrected for dilution or analytical errors. The initial NMHC values
indicated at the top of each profile are nonmethane readings from a
total hydrocarbon analyzer. The more precise concentrations determined
by gas chromatography are presented in Table 6 of the text.
86
-------
1.6
1.4
1.2
1.0
.1
s
NO
0.4
N02
Run No. AP-OOI
Initial Concentration (ppm):
NMHC as CH4 6.0
NO 0.5
NO, QI2
Aerosol
eo 120 iso
240 300 360 420
Irradiation Time, minutes
40
35
30
»
25 I
20 o
15 £
10 <
480 540 .
so8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
02
Run No. AP-002
Initial Concentration (ppm):
NMHC as CH4 6.2
NO 0.23
N02 0.05
Aerosol
60 i?0
240 300 360 420
Irradiation Time, minutes
40
35
30
25
20
O
(U
,5
10
87
-------
1.6
1.4
1.2
Run No. AP-OQ3
Initial Concentration (ppm)
NMHC as CH4 6.5
NO 1.43
NOz 0.33
.9
P
B OK
240 300 360 420
Irrodialion Time, minutes
1.6
Run No. AP-004
Initial Concentration (ppm)
NMHC as CH4 5.9
NO 0.98
N02 0.19
60
120
240 300 360 420
Irradiation Time, minutes
88
-------
1.6
40
.1
s
o
1.4
1.2
1.0
0.6
0.6
0.4
NO
Q2
Run No. AP-005
Initial Concentration (ppm):
NMHC as CH« 3.3
NO 0.25
N02 0.27
Aerosol
60 120 180 240 300 360 420
Irradiation Time, minutes
35
30 "fe
%
25 I
a
w.
1
20
8
i
15 J
10
480 540
T5?
o
I
1.6
1.4
1.2
1.0
0.8
40
s
8 0.6
0.4
Q2
NO
Run No. AP-006
Initial Concentration (ppm):
NMHC as CH4 3.3
NO 0.51
N02 0.10
60 120
"180 2<0 300 360
Irradiation Time, minutes
35
30 E
\
25 c
20
15
10
420 480 540
89
-------
1.6
.1
s
Run No. AP-007
Initial Concentration (ppm):
NMHC as CH4 3.35
NO a99
NOz 0.20
40
35
so "6
25
20
,5
10
.9
60 120 180 240 300 360 420 480 540 .
Irradiation Time, minutes
.1
S
Run No. AP-008
Initial Concentration (ppm)
NMHC as CH4 11.3
NO 0.48
N02 0.10
60
120 180 240 300 i60 420
Irradiation Time, minutes
480 540
90
-------
Run No. AP-009
Initial Concentration (ppm):
NMHC as CH4 11.3
NO 0.96
NOz 0.20
60
120
ISO
240 300 360
Irradiation Time, minutes
420
480
540
Run No. AP-OIO
Initial Concentration (ppm)
NMHC as CH4 11.8
NO 0.25
NOz 0.06
240 300 360
Irradiation Time, minutes
91
-------
1.6
1.4
1.2
a
£0.8
o
c
o
o
S0.6
O
0.4
0.2
Run No. AP-013
Initial Concentration (ppm):
NMHC as CH4 12,5
NO 1.52
N0? 0.41
40
35
60
120
180
240 300 360 420
Irradiation Time, minutes
480
S40
1.4
1.2
1.0
a.
a.
w
o
o 0.6
O
in
O
0.4
02
Run No. AP-014
Initial Concentration (ppm):
NMHC asCH4 3.1
NO 0.84
N02 0.18
35
30
25
.1
20 £
«
o
S
-------
1.4
1.2
IX)
08
O.Z
Run No. AP-015
Initial Concentration, ppm
NMHC as CH4 1.5
NO 0.5
N02 a I
"0 60
120
ISO 240 300 360 420
Irradiation Time, minutes
480
540
35
30
10
25 1
20 I
8
a
15 «
.1-6
1.4
1.2
1.0
£08
u
04
0.2
NO
Run No. AP-016
Initial Concentration,ppm
NMHCasCH4 1.5
NO 0.25
0.04
Aerosol
180 240 300 360
Irradiation Time,minutes
420
480
540
40
35
10
30 6
25
c
20 8
o
U
-------
1.4
1.2
1.0
a.
a.
3 0.6
o
0.6
w
o
O
O.4
0.2
Run No. AP-017
Initial Concentration (ppm):
NMHC as CH4 1.6
NO 0.14
N02 0.03
35
30
25 I
20 £
IS
o
c
o
o
3
10 I
Aerosol
60
120
ISO
240 300 360
Irradiation Time, minutes
420
480
540
1.4
1.2
1.0
a
a.
08
o
o
§0.6
O
0.4
0.2
Run No. AP-018
Initial Concentration (ppm):
NMHC as CH4 3.0
NO 0.13
N02 0.02
30
n
I
25 1
£
o
20 1
I
o
O
a>
3
IS
Aerosol
°8
o
IZO
ISO 240 300 360 42O
Irradiation Time, minutes
480
540
0
600
94
-------
1.6
0.2
Run No AP-019
Initial Concentration ,ppm
NMHC as CH4 6.0
NO 0.13
N02 0.02
40
35
m
30S
E
25
20 S
o
O
v
15 2
10
120
ISO 240 300 360 420
Irradiation Time,minutes
480
540
0
600
95
-------
APPENDIX B
SUMMARY OF HYDROCARBON DATA
DETERMINED BY GAS CHROMATOGRAPHY
The computer-generated summaries are designed to show the
consumption of hydrocarbons after correcting for the chamber dilution
rate. Each successive asterisk represents the hourly cumulative consumption
(by percent) of the respective hydrocarbon. The first asterisk opposite a
hydrocarbon corresponds to the initial concentration, and the second
asterisk represents the percent loss of that hydrocarbon after the first
hour of irradiation; the third asterisk is the cumulative loss (by percent)
after the second hour, etc. Where less than 11 asterisks are present,
either the hydrocarbon concentration became undetectably small or the
rate of decay became indistinguishable from the dilution rate.
96
-------
Surf***f PoH KJN AH-QOj li»i9»7» cOK^eCTEU ^
OF
ftHCt.nl
70 BO
2-rifclrtTl.
*
« « »
* » « « *
« » * » « «
*"*** ** ** *
Eln«"t
»««« *»««
3U
60
60
?-iit 1 1 TI.-C-BU f t'
»Lt.NC.
J-e-bu J t.j\t _
»
tr
»
» » ~~ «
«*« *«««*»«*
» «
* *
"Utf
» »««»«««
SO 6U 70 BO
10V
J ,2,<»-|Kli«it IrirL
»
»»**
*
* «
"* "
»
__ _ ^ V"~"» '
* * * **
* *
«
fe.HCC.nl
ALL
30
* «*«»»««
60 70
-------
HYDROCARBON SUMMARY FOR RUN AP-002 ll-20-7<» CORRECTED FOR DILUTION
sjcs-ssivEm ASTERISKS, RE^RES^NT_J^HULATIVE PERCC.NT CJNSUHPTION.PER.HOUR^QJLJRRADIATIQN
» PERCENT CONSUMED 0 Id 20 30 *tO 50 60 ffl 80 90 100
PARAFFINS
_2-H£THYL PROPANE *_ _ f * ? * * * * * * »
2-HETHYL PENTANE ** »».»*»*»«
_N-P£NTANE »_ »_ * * * * * * * * *
3-N£fHYL BUTANE " * *'»"» » » » » »
N-3UTANE *» **** »**»»
PROPANE
_EJHANE
ALL PARAFFINS
PERCENT CONSUMED 0 10 20 30 40 50 60 70 60 90 100
.QLEFINS AND ACEJYLEN& ...
TRANS-2-BUTENE.
PR3PVLENE
* * * ***
THYLENE * * * * * * * * * * *
ACETYLENE ».*»*
_AL U J3.LEF INS 1
SO
oo-
PERCENT CONSUMED 0 10 20 30 f»0 50 60 70 80 90 100
AROMA TICS
1,2«V-TRIHETHYL BENZENE * * » » » , **.*
_M-XYLENE * » * ___^ *_ * *_ _*_ * A_* .
P-ETHYL TOLUENE "* * » » * » " »" » ~*
TOLUENE *» *********
9ENZENE
ALL_AROMATICS
PERCENT COXSUMc0 0 10 20 30 VO 50 60 70 . £0 90 100
GR4NO SUM
ALL HYDROCARBONS
*******
-------
HYDROCARBON SUMMARY FOR RUN AP-003 ll-21-7<» CORRECTED FOR DILUTION
SUCCESSIVE ASTERISKS REPRESENT CUMULATIVE PERCENT CONSUMPTION PER HOUR OF IRRADIATION
I " PERCENT CONSUMED 0 10 30 30 <»0 50 60" 70~ "80 90 100
PARAFFINS
2-1ETHYL PROPANE *
Z-METHYL PENTANE » »
N->ENTANE * _»
Z-1ETHYL BUTANE * * * * * «
N-9UTANE »»»*»» »» »* *
PROPANE
ETHANE ******
ALL PARAFFINS »»»»»,»»»»»
PERCENT SONSUHEO 0 10 ZO 30 <»0 50 60 70 SO 90 100
_OL;FINS ANO_ACE.T.r.L£N£. . ... ......
3-1£THYL-Z-BUT£N;
TR4NS-2-8UTENE
PRD»YLENE
ACETYLENE
ETHYLENE
ALL OLEFINS
* * »*
» » » »*
» » » » » *«»»
* ***»* »»»»
* » »»*
PERCENT CONSUMED 0 10 20 30
-------
, .HY-OROCARaiN.--SUMMARY. FOR.RUN AP-QC"» 11-25-71* CORRECTED FOR DILUTION _.-
SUCCESSIVE ASTERIS
-------
HYDROCARBON SUMMARY FOR RUN AP-005 11-26-74 CORRECTED FOR DILUTION
PERCENT CONSUMED 0
iy.C5£_Sj»IVE_ASTERISKS REPRESENT CUMULATIVE PERCENT CONSUNPTION PER H3UR OF IRRAOIATJON_
_ ___ _ __ _.
10
20
30
90
100
2-1ETHYL PROPANE
2-1ETHYL PENTANE
N-°ENTANt
2-1ETHYL BUTANE
N3 1 1 T A fci C
-3UT ANt
*
*
*
*
* * * *******
* * ** ** ****
PROPANE
_SJH*N.E
ALL PARAFFINS
»» *** **** **
*******
* * * * »»»»»*»
Tor
"9F
TOF
PERCEMT CONSUMED 0
OLEFI NS ANO_ ACcTYLENt
TR4NS-2-BUTENE _
"2-H£THYL-2-BUTEN£
PR33YLENE
10
20
30
50
60
* *
* ***
ACETYLENE
_ET*rLENE _
ALL OLEFINS
* * * »» ******
****
"V
* ***
o
PERCENT CONSUHEO 0
10
20
30
50
60
TOTT
A301ATICS
BENZENE_
P-ETHYL TOLUENE
J» * * **»
* * * * * *
TOLUENE
BENZENE
ALL AROMATICS
* * *
***********
* * * * >
PERCEST CONSUMED 0
10
20
30
50
60
60
90
J»RA_ND_SUML
TBT
* * ******
-------
HYDROCARBON SUMHARY FOR RUN AP-006' 11-27-74 CORRiCTtO FOR DILUTION
_aaa3.£.SSIVE ASTERISKS REPRESENT CUMULATIVE PERCENT CONSUMPTION £ER_H3UR_Of-
PARAFFINS
PERCENT CONSUMED
10
30
50
60
70
90
100
_2-HETHYL PENTANE
2-METHYL PROPANE
JN-!»ENTANE
2-METHYL BUTANE
N-BUTANE
ETHANE
_PR3PANE __
ALL PARAFFINS
PERCEST SONSUHSD
.OLIFINS AND ACETYLENE.
10
20
30
40
50
60
70
60
90
100
TRANS-2-3UTENE
2 1ETHYL-2-BUTENE
PRS'YLENE
*
*
*
*
*
*
* * *»
* * »»
* * ***
ACETYLENE
_£THYLENE
ALL OLEFINS
o
KJ
*»«*******
PERCENT CONSUMED
.AR3MATICS ._
1.2,4-TRIMETHYL _BENZ£ME
H-XYLtNE
P-ETHYL TOLUENE
10
20
30
40
50
60
70
SO
90
100
*
«
»
»
*
*
* * * * »»*
9 f w 9 9^ -
* *****
TOLUENE
BENZENE
ALL AROMATICS
* * *
*» »» »*» »»»
~» ~~
**-*-«*
PERCENT CONSUMED 0
..6R4NO..SUN
10
20
30
40
50
60
70
80
90
100
ALL HYDROCARBONS
*******
-------
HYDROCARBON SUMMARY FOR RUN AP-007 12-02-74 CORRECTED FOR DILUTION
SU:;ESSIVE ASTERISKS REPRESENTCUMULATIVE PERCENT CONSUMPTION PER HOUR_OF IRRAPIATION
PERCENT CONSUMED 0 10 20 30 40 50 60 70 80 90 100
PARAFFINS
J-1ETHYI,
2-1ETHYL PENTANE * * "* * * * * * » » «
N-»ENTANE * * * * * * * * * * *
2-METHYL BUTANE »"» »* * * »»»» »
PR3PANE ** »»»»»»»»
N-9JTANE * »»»»»»»»
ALL PARAFFINS * » » »*.*»*»**
PERCENT CONSUMED 0 10 20 30 40 50 60 70 80 . 90 100
..OLEFINS AND ACETYLENE ____________________________ __ . __________ _______________ . __ __________________ _ ...... ____________ ___________________ ........
TR»SS-2-9UTENE _____ * ________ _____ ____________ _______ _ ____ * _ _____ *
-------
HYDROCARBON SUMMARY FOR RON AP-008 12-03-7<» CORRECTED FOR DILUTION
_SUC1ES5IV£ ASTERISKS REPRESENT CUMULATIVE PERCENT CONSUMPTION PER HOUR_QF, IRRADIATION
PERCENT CONSUMED 0 10 20 30 <»0 50 60 70 80 "90 100
PARAFFINS ..
PENTAME_
* ** ** ** **»*
J2-1ETHYL_ PROPANE *__*_.* *» »*** * *
2-METHYL BUTANE ,*»-»»*»,»,»
N-8UTANE »» *»*»» ***»
PRDPANE »
JETHANE *
ALL PARAFFINS **»»»»»»»»
PERCENT CONSUMED 0 10 20 30 <*0 50 60 70 80 90 100
.OLJFINS AND. ACETYLENE . .._
*
PRD'YLENE *
ETHYL ENE * * * *******
ACETYLENE »»»»»,,
* »*»»*»*»*»
O
-P-
PERSEST CONSUMED 0 10 20 30 f»0 SO 60 70 80 90 100
AR31ATICS
l,H,f»-TRIHETHYL 3ENZENE * * * .»»*»»
M-XYLENE » __* * * _^ *__ * * ^__*_*
P-ETHVL TOLUENE » » * * »»**»»-- -
TOLUENE ** ** **»*,,,
BENZENE **»»
ALL_AROMATICS _» * * * * ****** _^
PERCENT CONSUMED 0 10 20 30 Ml 50 60 70 80 90 100
GRAND SUM
* *********
-------
HYDROCARBON SUMMARY FOR RUN AP-009 12-0<*-7i» CORRECTED FOR DILUTION
_SOCG EJS SIVE ASTERISKS REPRESENT CUMULATIVE PERCENT CON SUM PTI ON_£ER _ Hj>jUR_qF IR RA 0IIAJ_ION_.
PERCENT CONSUMED o 10 20 30 40 so &o 70 so 90 100
PARAFFINS
_2-1STHYL PENTANE » » _» _» 1 » * * * * ,
N-3iNTANE » »»» » * » » »
_PRQP_ANE _ V * ....» » * * » * »
BUTANE »»»»»»»»
N-3UT4NE * * « »»*»»»»»
ETHANE »»»»»»»
ALL PARAFFINS » »
PERCENT CONSUMED 0 10 20 30 <»0 50 60 70 80 90 100
.OLiFINS ANO_.ACETYLENs . _
TR4MS-2-8UTENE
2-1ETHYL-a-6UT£NE
» «»»
ETHYLENE » * * .»»»**»»
ACETYLENE *** ****** *» , .
"ALL OLEFINS * *
* *******
o
Ul
PERCENT CONSUMED 0 10 20 30 40 50 60 70 80 90 100
i,2,
-------
HYDROCARBON SUMMARY FOR RUN AP-D10 12-35-7** CORRECTED FOR DILUTION
SUCCESSIVE ASTERISKS REPRESENT CUMULATIVE PERCcNT_CQNSUMPTION_PEfv HOUROF IRRADIATION
PARAFFINS
Z-METMYL "»fNTAHE
2-METHVL PROPANE
N-PENTANE
2-METHYL "IUTAN'C
N-BUTANE
P90PANE"
ETHAN- .
ALL PARAFFINS
10
26
30
«to
d
60
70
80
90
100
*» *
* *» * »» »»
«»**** ***
PERCENT CONSUMER 3
OLEFINS AND ACHTVLENc
13
50
70
60
90
100
PR'OPYLENE
ETHYLENE "
APETVLENE
ALL O'LEFINS
₯ » * » »
* « »»
*******
PEPCcNT CONSUMED 0
10
30
50
60
80
90
100
1 2» if-TP IM£THYL 3ENZE*I£ <
M-XYL2NE
P-ETHYL TOLUENE
TOLUENE
8ENZFNE
ALL AROHATICS
t * * * *******
* * * *******
»****»****
»#»**»**»*
CONSUMED 0
10
50
60
70
80
GP.AND SUM
90
100
ALL _HYOPOC A*9ONS
* * * ***** ** *
-------
HYDROCARBON SUMMARY FOR RUN AP-013 3-18-75 CORRECTED FOR DILUTION
SUCCESSIVE ASTERISKS REPRESENT CUMULATIVE PERCENT CONSUMPTION PER HOUR OF IRRADIATION
PERCENT CONSUMED 0 10 20 30 <»0 50 60 70 80 90 100
.PARAFFINS :
2-*ETMYL PROPANE _ »
2-METHYL PENTANE » » » » »
N-PENTANE. ... » » » ».»....«
2-METHYL BUTANE »»»»»» »»
MrSUTANE . »*.*f_J!.*.*?*»?
ETHANE »»»»* »»»*»
PROPANE _ »*»»»»
ALL PARAFFINS « »»»»»
PERCENT CONSUMED 0 10 20 30 <»0 50 60 70 80 90 100
OLEFINS AND ACETYLENE _ ._.. . _ __
2-1ETHYL-2-BUTENE . . * .... - .. ._ _ * _ **
TRftNS-2-
-------
SUMMARY FftP RljM AP-014 3-20-75 CORRECTED FOR DILUTION
SUCCESSIVE aSTFWISKS REPRESENT CI|Mlll_ATIVF PERCENT CONSUMPTION PtH HOUR OF IRRADIATION
g.MPTHYl PROPANF
2-MFTHYi RUTAME
N-PCWTAvF.
PoopaK'F
~~5i -[OSKTSrr fxrs-
20
*0
50
60 70
80
90 100
« « « *
* «
~» « 5 *~
-* *-
-»»*» »» »g
-------
HYDROCARBON SUMMARY FOR RUN AP-OIS 3-25-75 CORRECTED FOR DILUTION _
PARAFFINS
SUCCESSIVE ASTERISKS REPRESENT CUMULATIVE PERCENT CONSUMPTION PER HOUR OF IRRADIATION
PERCENT CONSUMED 6 16 20 30 40 50 60 70 80 «0 100
2-HETHYL PROPANE » * » * **» «
2-HETHYL PENTANE **»«»*«
?-METHYL BUTANE ***«»*« ,4*
N-PENTANE """ *»*
N-3UTANE " "
PROPANE "~ ~
ETHANE
ALL PARAFFINS *» ** »*
PERCENT CONSUMED 6 T6" 20- 36" *0 50 60" 76" 80 90 100"
OLEFINS AND ACETYLENE
»
" " "
_ * _*__*_*
«**»* **« ' " " """ " " -"'
TRANS-2-8UTENE
2-METHYL-2-BUTENE
pROPYLENE
ACETYLENE
ETHYLENE
ALL OLEFINS
(-
o
40 PERCENT CONSUMED 6 10 20 30 *0 SO 60 70 80 90 Tb6~
AROMATICS ......
lt2i4-TRIMETHYL 8ENZEN& « » *
P-ETHYL TOLUENE * * * *.-
M-XYLENE _JL * __* _ _ * *
TOLUENE '"" "~ * * *»* " " "
BENZENE »»*
ALL AROMATICS * »«»»
PERCENT CONSUMED 6 10 20 30 *0" 50 60" 76 80 9fl
GRAND SUM
ALL HYDROCARBONS ....... * * - .* -* - * * ** *'- - -
-------
HVDKOCARBCN SUMMARYFOR RUN AP-016 .3-2.6-75. .CORR£CTJD..FOR_J)_LLyTIpN_
.ASTERISl^REPREMj^CUHULATiyg._PEgCENrr CONSUMPTION PER HO MA-OP...IRg.AlMAT.IQN.-
""""" "PERCENT CON'SUHtb " 0 "" "~ 10 20 30 " 4b 50" "60 " " 70 80 90 100
PARAFFINS
2-METHYL PROPANE
2-METHYL PENTANE
N-f»ENTAN£
2-METHYL BUTANE
N-3UTANE
PRO PA ME
ET4ANE
ALL PARAFFINS
*
*
*
*
»»
*
*
» * * * ******
*****
PERCENT CONSUMED 0 10 20 30 «*0 50 60 70 60 90 100
OLEFINS AND ACETYLENE
TR4NS-2-BUTENE
2-1ETHYL-2-BUTEne
PROPYLENE
"ACETYLENE
ETHYL EME
"ALL OLEFINS »
* * »*
* * * **
* *****
^ *****
PERCENT CONSUMED 0 10 20 30 <»0 50 60 70 80 90 100
ARDMATICS -
1,2,«»-TRIMETHYL 3E.NZ.cNE * » _*I * * » ****
"P-ETHYL TOLUENE" " *'" " " " "» " ' * "» * * »"""» *~i
H-XYLENE * ^ * * » » » * * * * *
TOLUENE * < * * » » * » , » » »
^BENZENE . ****** .
«LL AROMATiCS * * »»-»-»» »- » » »
PERCENT COJ4SUWEO 0 10 20 30 ^0 50 60 70 8^ 90
GRAND SUM
ALL HYDROCARBONS
* ******
-------
APPQN SUMMARY FOR RUN AP-017 3-27-75 CORRECTED FOR DILUTION
SUCCESSJLV£_ASlER.ISKS_REPRESEWI_GUJHULAlItfE_P£SC£JOJIQNSUMPnOM.PER HOUR _QF_IRRAOIATIOfL
PERCENT CONSUMED C
PARIFF.INS
10
30
ENTANE
N-FENTANE :
2-MFTHYL -3UTANE
-5UTA.NE
PROPANE
. ETHAME
ALL PARflFFTNS
* * * * ******
****»* * »*»
*» * »
»« *«*»***
PERCENT CONSUMED C
-OLEFINS AND ACE.T.YJ.ENE
10
30
60
70
8C
90
100
TRANS-2-3lJT£.^4E,
2-METHVL-3-8UTENE
PPOPYLFN?
ACETYLENE
ETHYLFNF
ALL OLEFINS
*
*
* * » » »
*
*
* » * »»»»
PERCENT CONSUMED 0
10
30
60
70
80
90
100
1 . 9. U-TPTMr THYL 9EN7ENE
P-ETMYL TOLUENE
TOtUFNE
ft PM7PNP
ALL ARCHATICS
» »
* * * *
* * * * *
** **»*»***
* * * ***** **
_ * . .» »»».
» « « « «
« * *
PERCENT CONSUMED 0
G"?ANO SUM
.JkLL HYDROCARBONS *-
10
20
30
50
60
80
90
100
* * *******
-------
.HYDROCARBON SUMMARY_FOS.HUN AR-018 4-01-75 CORRECTED FOR DILUTION
-.-... ... - SUCCESSIVE ASTERISKS.REPRESENT CUMULATIVE_PERCENT_ CONSUMPTION PER HOUR OF IRRADIATION
PERCENT CONSUMER 0 i6203040 5060 70 BO
90
100
PENTANE
2. METHYL PROPANE
2-MfTHYi BUTANE
* * *.*__**
*«» ** *» **
«*«*«
PROPANE
ETHANE ......
AIL PARAFFINS
***»***
PERCENT CONSUMER 0
OI.EFINS AND ACETYLENE.._ .
10
20
30
60
70
80
90
TOANS-2-PUTENE
2-MPTHYL-2-RUTENE
POOPYLENE __________
ACEYLEME
ALL
100
* _«*__
* * « **
PERCENT COKSUMEO 0
10
20
30
50
60
70
90
100
1. 2. A-TOIMETHYL BENZENE
P-ETHYL TOLUENE
BFN7ENE
ALL APOMATICS
NZENE
* *
* * » *
* » «
* » ***»»
*
» *
* * '
PERCENT
00ANO SUM
ALL. HYDROCARBONS -
10
40
50
60
70
80
90
100
* * * *«
-------
SUMMARY F«H KUN AP-Oli 4-02.75 CORRECTED FOR nJLUTXON
SUCCESSIVE ASTERISKS REPRESENT CUMULATIVE PERCENT CONSUMPTION P£« HOUR OF IRRADIATION
PAMAFpINS
PERCENT CONSUMED 0
SO
TO
ao
2-MElHYL PROPANE
2-HfcTnYU PfcHTANE
2-ntlHYU BUTANE
£ I HAM£
'
*
PERCENT CONSUMED0
ACETYLENE
TO"
'80
so
70
"id"
*
»
'*"
PERCENT CONSUMED
BENZENE
30
so
60
70
ao
iOv
P*feTnyU TOLUENE
TWUUtNE
*
*
«««*«
«*
*
*
*
* *
»
«
* *
*
*«
« *
9KANU SUM
AM.
PERCENT CONSUMED 0
10 20
*
30
SO
60
70
80"
-------
TECHNICAL REPORT DATA
(Plratt rrael Instructions on the reverse before completing/
1. REPORT NO.
EPA-600/3-76-080
2. 3. RECIF
4. TITLE AND SUBTITLE 5. REPO
SMOG CHAMBER STUDIES ON PHOTOCHEMICAL AEROSOL- J
PRECURSOR RELATIONSHIPS B.PERF
7. AUTHOR(S)
David F. Miller and
9. PERFORMING ORG "\NIZATION NAME At
Battell e Columbus L<
505 King Avenue
Columbus, Ohio 432(
12. SPONSORING AGENCY NAME AND ADC
Environmental Scieni
Office of Research c
U.S. Environmental 1
Research Triangle P<
8. PERF
Darrell W. Joseph
go ADDRESS 10. PRO
iboratories
11. CON
31
3RESS 13.JYP
*pc Rpsparrh 1 aboratorv
ind Development 14
------- |