EPA-650/3-75-007
August 1975
Ecological Research Series
STUDY OF AEROSOL FORMATION
IN PHOTOCHEMICAL AIR POLLUTION
01
a
-------
EPA-650/3-75-007
STUDY OF AEROSOL FORMATION
IN PHOTOCHEMICAL AIR POLLUTION
by
W. C. Kocmond, D. B. Kittelson,
J. Y. Yang, andK. L. Demerjian
Calspan Corporation
4455 Genesee Street
Buffalo, New York 14221
Contract No. 68-01-1231
ROAPNo. 21AKB-02
Program Element No. 1A1008
EPA Project Officer: Marijon Bufalini
Chemistry and Physics Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
COORDINATING RESEARCH COUNCIL, INC.
30 Rockefeller Plaza
New York, New York 10020
CAP A-8-71
and
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
August 1975
-------
EPA REVIEW NOTICE
This report has been reviewed by the National Environmental Research
Center - Research Triangle Park, Office of Research and Development,
EPA, and approved for publication. Approval does not ignify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environ-
mental Protection Agency, have been grouped into series. These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These series are:
1. ENVIRONMENTAL HEALTH EFFECTS RESEARCH
2. ENVIRONMENTAL PROTECTION TECHNOLOGY
3. ECOLOGICAL RESEARCH
4. ENVIRONMENTAL MONITORING
5. SOCIOECONOMIC ENVIRONMENTAL STUDIES
6. SCIENTIFIC AND TECHNICAL ASSESSMENT REPORTS
9. MISCELLANEOUS
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 for sale through the National
Technical Information Service, Springfield, Virginia 22161.
Publication No. EPA-650/3- 75-007
ii
-------
ABSTRACT
Photochemical aerosol production in several 502 + clean air (fil-
tered air), HC + NO and BC + NO + 502 systems has been examined using the
smog chamber approach. The reaction vessels used in this study were the
20,800 ft 3 Calspan chamber and the 600 ft 3 University of Minnesota chamber.
Aerosol formation, growth, and decay mechanisms are described for each of
the systems studied. It has been possible in this investigation to charac-
terize system reactivity in terms of aerosol behavior. The most important
variables are maximum number concentration, equilibrium surface conèentration,
and particle volumetric growth rate. Measurements of these variables are
made for several systems and are discussed within the text.
Of the hydrocarbons studied, cyclohexene was the most reactive in
terms of aerosol production and rate of NO oxidation followed by m-xylene,
hexene, and toluene. For the simple BC + NO system, each experiment can be
divided into two phases. During the initial phase, NO is converted to NO 2
and some oxidation of hydrocarbon occurs. No appreciable aerosol is formed
during this phase, but ozone starts to appear near the end of this period.
The second phase, accompanied by substantial aerosol formation, begins as
soon as NO is oxidized out of the system and NO 2 reaches a maximum; ozone
levels rise rapidly during this phase and approach a maximum. The addition
of SO 2 to the 1 - IC + NO system leads to some aerosol formation during the first
phase and was generally found to exert a synergistic effect on aerosol forma-
tion in the second phase. The addition of SO 2 also led to a marked decrease
in the diameter of the particles ultimately formed. This results from the
formation of very high concentrations of nuclei during the initial stages of
the experiment.
For the SO 2 + clean air system, photooxidation rates of a few tenths
of a percent per hour are typically observed for a light intensity of 50% noon
day sun. In the presence of hydrocarbons and NO, accelerated rates are gen-
erally observed.
The data show that aerosol formation rates are enhanced at high
relative humidities, probably as a result of the higher water content of the
aerosols.
in
-------
ACKNOWLEDGMENTS
The authors wish to express their thanks to three University of
Minnesota graduate students: David Cress, Kui Chiu Kwok, and Ping Auw for
performing the experiments in and reducing the data from the University of
Minnesota Smog Chamber.
Special thanks are also due Mrs. Joyce Terrano for her skillful
typing of this manuscript.
iv
-------
TABLE OF CONTENTS
Section Page
1.0 INTRODUCTION . 1
2.0 EXPERIMENTAL FACILITIES 4
2.1 Ca lspan 4
2.1.1 Instrumentation 4
2.2 University of Minnesota 6
2.2.1 Instrumentation 7
3.0 CHAMBER LIGHT INTENSITY MEASUREMENTS i i
3.1 Calspan Chamber 11
3.2 University of Minnesota Chamber 15
4.0 RESULTS AND DISCUSSION 19
4.1 November Workshop 19
4.2 Duplicate Experiments to the November Workshop - Univer-
sity of Minnesota 27
4.3 Conclusions from the November Workshop 34
4.4 March 1974 Workshop 35
4.4.1 SO 2 Experiments 36
4.4.2 Hydrocarbon Experiments 45
4.4.3 Hydrocarbon + NO Experiments 46
4.4.4 Hydrocarbon + SO 2 Experiments 54
4.4.5 Hydrocarbon + NO + SO 2 Experiments 58
4.5 Conclusions from the March Workshop 74
5.0 CHAMBER CHARACTERIZATION TESTS
5.1 Dark Reaction Tests -- University of Minnesota 77
5.2 Coagulation Experiments -- Caispan 83
5.3 NO Photolysis Experiments - - Ca lspan and University
of Minnesota 85
6.0 REFERENCES 89
v
-------
TABLE OF CONTENTS (Cont’d)
Page
APPENDIX A - - AEROSOL AND CHEMISTRY DATA FROM NOVEMBER
1973 WORKSHOP WITH DUPLICATE UNIVERSITY
OFMINNESOTAEXPERIMENTS A-i
APPENDIX B - - AEROSOL AND CHEMISTRY DATA FROM MARCH 1974
WORKSHOP WITH DUPLICATE UNIVERSITY OF
MINNESOTAEXPERIMENTS B-i
vi
-------
LIST OF TABLES
Table No. Page
I Modeled Results for NO 2 Photolysis in Air with 1420
and CO present 13
II Calspan Chamber - NO 2 Photolysis in Air 14
I II Light Intensity Measurements 17
IV Volume Average Light Intensities - U of M 18
V Summary of Data from November 1973 Workshop 21
VI Hydrocarbon Reactivity 26
VII U of M Duplicate Tests Subsequent to November Workshop:
Summary of SO 2 Experiments 30
VIII U of M Duplicate Tests Subsequent to November Workshop:
Summary of Chemical and Aerosol Data for Toluene Experiments. 31
IX Summary of Aerosol Data from March Workshop 38
X U of M Duplicate Tests Subsequent to March Workshop:
Summary of Aerosol Data for Hydrocarbon Experiments 39
XI Summary of Chemistry Data from March Workshop 40
XII U of M Duplicate Tests Subsequent to March Workshop:
Summary of Chemical Data for Hydrocarbon Experiments 41
XIII U of M Duplicate Tests Subsequent to March Workshop:
Summary of SO 2 Experiments 42
XIV Influence of Bag Size on Aerosol Production 43
X V Aerosol Reactivity Experiments - U of M 78
XVI NO Oxidation Experiments 86
vii
-------
7 Hexene + NO
8 Xylene + NO
9 Xylene + NO
10 Cyclohexene
11 Cyclohexene
12 Hexene + NO
13 Hexene + NO
14 Xylene + NO
15 Xylene + NO
16 Cyclohexene
17 Cyclohexene
- Caispan
System - U of M
System - Caispan
System - U of
+ NO System -
+ NO System -
+ S02 System -
+ SO 2 System -
+ SO 2 System -
+ SO 2 System
+ NO + SO 2 -
Page
8
23
23
28
29
48
49
52
53
56
57
62
63
66
67
‘70
71
72
73
79
80
84
LIST OF FIGURES
Figure No .
1 E lectrica lAeroso lAna lyzer
2 SO 2 Experiment Showing Linear Volumetric Growth
3 SO Experiment Showing Upward Curvature of Aerosol Volume....
4 Toluene + NO 2 + SO 2 System - Calspan
5 Toluene + NO 2 + SO 2 System - U of M
6 Hexene + NO System
S
M
Caispan..
U of M...
Caispan.
U of M..
Calspan.
— U of N
Caispan
+ NO + SO 2 - U of M
18 Mean Surface Diameter vs Time for Several HC + NO, HC + NO +
SO 2 , and SO 2 Experiments
19 Light Scattering Coefficient 8 5CAT of Photochemical
Aerosols vs Time
20 Time Histories of Aerosol Coagulation
21 Time Histories of Aerosol Coagulation
22 Aerosol Coagulation Data - Auto Emission Test Series
viii
-------
Section 1
INTRODUCTION AND SUMMARY
Caispan Corporation, in collaboration with the Particle Technology
Laboratory of the University of Minnesota, has been engaged in a laboratory
study of the formation mechanisms and growth processes of photochemical aero-
sols. The primary objective of the investigation is to improve our under-
standing of photochemical aerosol behavior in urban environments by conducting
studies in carefully controlled environments, simulating those occurring in
the real atmosphere.
During the first year program, emphasis was placed on preparing
the Caispan 20,800 ft 3 reaction chamber for photochemical aerosol studies
and in examining aerosol behavior in SO 2 and ProPylene_NO systems. At the
University of Minnesota, studies in 600 ft 3 and 90 ft 3 chambers were con-
ducted to help determine the effects of bag size on aerosol production and
also to study the influence of varying relative humidities on aerosol
behavior. - Results of the first year program have been presented in an
earlier report, Kocmond et al. (1973).
During this second year program, greater emphasis was placed on
understanding aerosol behavior in representative HC+NO +SO 2 systems. Two
collaborative workshops were held at Caispan. During the November 1973
workshop, aerosol behavior was studied in SO 2 + clean air, toluene + NO 2 ,
and hexene + NO 2 + SO 2 systems. The data and experience gained from these
tests were applied to designing experiments for a second and perhaps more
productive workshop held during March 1974. These experiments involved HC-
NO _SO 2 systems using toluene, hexene, m-xylene, and cyclohexene as repre-
sentative hydrocarbons. For each of the test series performed in Calspan’s
20,800 ft 3 chamber, a comparative set of experiments were conducted in the
3 . .
600 ft chamber at the University of Minnesota.
1
-------
A total of 145 smog chamber experiments were performed. For the
most part, good agreement was found in the experimental data obtained at
Calspan and the University of Minnesota; and a reasonably good understanding
of the kinetics of gas to aerosol conversion in the polluted atmosphere was
achieved. From the data generated in these experiments, the following points
can be made:
(1) SO 2 photooxidation rates of a few tenths of a percent per hour
are typically observed in clean, filtered air for a light intensity of about
50% noon day sun. In the presence of hydrocarbon contamination, accelerated
rates are generally observed.
(2) Each HC + NO experiment can be divided into two phases. In the
first phase, NO is converted to NO 2 and some oxidation of hydrocarbon occurs.
Ozone starts to appear near the end of this period. The second phase, accom-
panied by substantial aerosol formation, begins as soon as NO is oxidized out
of the system and NO 2 reaches its maximum; ozone grows rapidly and approaches
a maximum.
(3) The addition of 502 to the HC + NO system was generally found
to exert a synergistic effect on aerosol surface and volume production. At
Calspan the effect was greatest for m-xylene, while at the University of
Minnesota the largest effect on aerosol behavior was observed in the hexene +
NO + 502 system. Possible synergistic effects in the cyclohexene system were
masked by the explosive growth of aerosol with and without the addition of SO 2 .
(4) The addition of SO 2 to the HC + NO system produces a dramatic
decrease in the mean particle diameter. This results from the initial forma-
tion of very high concentrations of H 2 S0 4 nuclei during the initial stages of
the experiment. During the second stage of aerosol growth, condensation pro-
ceeds on the existing particles. In the HC + NO system alone, fewer but larger
particles are produced.
(5) Of the hydrocarbons studied, cyclohexene was the most reactive,
both in terms of aerosol and chemical behavior, followed by m -xylene, hexene,
2
-------
and toluene. The main difference observed in the duplicate experiments at
the University of Minnesota was that hexene was the least reactive hydrocarbon.
(6) It is possible to characterize system reactivity in terms of
aerosol behavior. The most important variables are maximum number concentra-
tion, equilibrium surface concentration, and volumetric growth rate. These
aerosol measures of reactivity have been found to correlate well with other
conventional parameters, such as time to [ NO ] and [ 0 1
2max 3max
(7) Increasing relative humidity was found to significantly increase
aerosol surface and volume production.
Because of the large body of experimental data generated during the
program, graphical presentations of both the chemical conversion and the aero-
sol growth data for the individual experiments are given in Appendices A and
B at the end of this report. Data summations and discussions are provided in
Section 4, Results and Discussion. Since detailed descriptions of experimental
facilities were provided previously in the first year report, only brief des-
criptions highlighting some recent facilities improvements are given in Section
2. Detailed discussions of light intensity measurement techniques employed at
Calspan and the University of Minnesota are therefore provided in Section 3.
Chamber characterization tests designed to assess the effects of chamber con-
tamination and to assure validity of experimental data presented in this report
are described in Section 5.
3
-------
Section 2
EXPERIMENTAL FACILITIES
2.1 Caispan
The smog chambers used at Calspan and the University of Minnesota
have been discussed elsewhere (Kocinond et al., 1973; Clark, 1972) and will
not be treated in detail here. Briefly, however, the Caispan chamber con-
sists of a cylindrical chamber 30 feet in diameter and 30 feet high, enclosing
a volume of 20,800 ft 3 . The chamber walls are coated with a specially formu-
lated fluoroepoxy, which has surface adhesion characteristics very similar to
those of FEP teflon. Illumination within the chamber is provided by a combina-
tion of fluorescent daylight and blacklight lamps installed inside 24 lighting
modules and arranged in eight vertical channels attached to the wall of the
chamber. Light intensity has been increased to give kd [ NO ]_O.23 min during
this second year by installing two 215-watt fluorescent daylight lamps, eight
85-watt high output blacklamps, and two 40-watt sunlainps in each module.
(Further modifications have been made since the end of the program to give
kd [ No 2 ] O. 33 min ). The lighting modules are covered with 1/4” Pyrex glass
and are thus sealed from the chamber.
Air purification is provided by a recirculation system which can
continuously filter the air through a series of absolute and activated char-
coal filters. Experiments show that nearly all gaseous contaminants and
particulate matter (<200 nuclei cc ) can be removed from the chamber air in
about four hours of filtration.
2.1.1 Instrumentation
Instrumentation used to monitor aerosol behavior and reactant con-
centrations within the chamber included the following:
4
-------
(1) Bendix Model 8002 Ozone Analyzer -- The instrument uses photo-
metric detection of chemiluminescence resulting from the reaction of ozone
with ethylene to determine ozone level. The minimum detectable sensitivity
is reported to be 0.001 ppm. Good reliability and reproducibility of data
was achieved using this instrument.
(2) Bendix Model 8101-B Nitrogen Oxides Analyzer - - Detection is
based on chemiluminescent reaction between nitric oxide and ozone. The
detection limit for each of the nitrogen oxides is 0.005 ppm. Periodic
maintenance, as well as frequent calibration of this instrument, was often
necessary.
(3) Bendix Model 8300 Sulfur Analyzer -- Operation of this instru-
ment is based on the photometric detection of sulfur atoms excited in a
hydrogen-rich flame. A set of filters is used for selective monitoring of
sulfur dioxide and hydrogen sulfide. The minimum detectable sensitivity
is 0.005 ppm. This instrument was found to be excessively sensitive to
pressure changes within the smog chamber and required frequent adjustments
to the sample flow.
(4) Hewlett-Packard Model 5750 Gas Chroinatograph - - The chromato-
graph is equipped with dual column and dual flame ionization detectors.
Depending on the column in use, either total hydrocarbon or individual com-
ponents can be analyzed.
(5) Bendix Model 820 Reactive Hydrocarbon Analyzer -- This instru-
ment uses flame ionization detection to provide quantitative analysis of
methane [ CH 4 ), total hydrocarbons (THC), and reactive hydrocarbons (THC-CH 4 ).
The instrument was received late in the contract period and provided only
limited use during this program.
In addition to the gas analyzers, a number of aerosol measuring
instruments were used on the program. These included the following:
S
-------
(6) Electrical Aerosol Analyzer (EAA) -- This instrument (described
in the next section) was provided by the University of Minnesota during our
joint workshops held at Calspan in November 1973 and March 1974. The instru-
ment gives size distribution data of the photochemically-produced aerosols.
Caispan obtained a Model 3030 Electrical Aerosol Analyzer of its own during
the summer of 1974.
(7) An MRI integrating nephelometer for aerosol light scattering
and visual range measurements.
(8) A Gardner Associates Small Particle Detector -- This manually
operated instrument is used to measure total particle concentration. It has
a range of sensitivity from 200 to 1O 7 nuclei/cc and can detect particles
as small as 0.002 pm.
(9) An Environment-One Model 100 Condensation Nucleus Monitor - -
This instrument is reported to measure particle sizes down to 0.0025 pm and
have a range of sensitivity from 50 to l0 nuclei/cc . Considerable diffi-
culty was experienced in maintaining continuous operation of this instrument.
There also appeared to be some lack in sensitivity to photochemically-produced
aerosols during the initial homogeneous nucleation stage. An earlier stage
of GE model has been acquired and modified for current usage with more reliable
performance.
2.2 University of Minnesota
The University of Minnesota smog chamber is a cylindrical vessel
fabricated of 0.01 in. DuPont FEP Teflon and encompassing a volume of 625 ft 3 .
For a complete description of the chamber and supporting facilities, see
Clark (1972). The illumination system consists of 72 GE F4OBL fluorescent
lamps mounted in vertical pairs on 36 evenly-spaced supports. Aluminum foil
has been attached behind the lamps to increase the uniformity and intensity
of the light. Light intensity (see next section) is measured to be kd [ N0 2 ]
-0.20 min for the U of M chamber.
6
-------
The air purification system consists of an absolute particle filter,
an activated charcoal scrubber, silica gel dryer, humidifier, and final filter.
Ambient laboratory air is purified by pumping it through the purification sys-
tern at about 15 CFM. Air passing through the purification system is exposed
to only non-reactive metal, glass and Teflon duct surfaces in order to minimize
sources of contamination.
2. 2 . 1 Instrumentation
(1) Electrical Aerosol Analyzer - - Two versions of a portable
electrical aerosol analyzer were used for the experiments. The “laboratory
prototype” analyzer, used for the joint Caispan-University of Minnesota
workshops and the first 45 experiments at the University of Minnesota, has
been described by Liu et al. (1974). A second version, the “commercial pro-
totype” (Thermo-Systems Model 3030), was used for the remainder of the experi-
ments. A description of this second instrument is in Liu and Pui (1975).
Figure 1 shows a schematic of the comnerical instrument. This latter instru-
ment requires 4 l/m of aerosol-laden air as a sample and 46 1/rn “clean” sheath
air. Usually, the entire 50 1 1 T h air supply was taken from the chamber, and
the sheath air portion was filtered with an absolute filter. The sheath air
then had the same relative humidity and contained the same trace gases as the
s amp 1 e.
Both analyzers are based on the “diffusion charging-mobility analysis”
principle described by Whitby and Clark (1966). The aerosol-laden air flows
through the charger, a region containing unipolar ions which have been pro-
duced by a corona discharge. The aerosol particles emerge from the charger
carrying a negative charge and are introduced into the mobility analyzer.
In this section, a positive voltage on a collection rod causes all particles
with electrical mobilities greater than a certain critical value to be pre-
cipitated. Those particles with smaller mobilities flow past this section
and are collected by an absolute filter. An electrometer, which is con-
nected to the filter, measures the current carried by the charged particles.
The mobility spectrum, and, therefore, the size distribution can be inferred
7
-------
BALL
VALVE
AEROSOL.
SHEATH AI
METAL
VACUUM
B ALL
VALVE
- - — - — CORONA VOLTAGE
—---.CHARGING CURRENT
———-SCREEN VOLTAGE
————-.ROD VOLTAGE
ELECTROMETER
CURRENT
Figure 1 Electrical Aerosol Analyzer
MASS-FLOW
TAAN SL)UCER
‘1
AEROSOL I
ELECTRiCAL SIGNAL
L_ .._.AEROSOL FLOWRATE
8
o c i.
FILTER
PLASTiC
SENSOR
— - TOTAL FLOW
S
-------
from the electrometer current as a function of collecting rod voltage.
A complete set of readings takes about 2 1/2 minutes; each current is
measured at a different time. The aerosol is dynamic, however, and the
size distribution may change substantially during the course of one set of
data. To compensate for the error that this time lag introduces, an inter-
polation program was used to generate corrected currents for a given time
in the experiment. All data were analyzed using the calibration of Liu et al.
(1974). These constants are in error when used with the TSI 3030. However,
the errors associated with their use should not be large. The use of these
constants results in higher measured volume and surface concentrations.
(2) Condensation Nuclei Counter -- A General Electric Condensation
Nuclei Counter (CNC) was used to measure total particle concentrations. All
particles larger than about 0.002 pm should be detected. Details of the
instrument have been given by Skala (1963). It has a range of sensitivity
from about 50 to Ia 7 particles/cm 3 . For this study, the instrument was used
primarily on the 100,000 particles/cm 3 scale. This scale was calibrated as
described by Liu and Pui (1974). If concentrations greater than 100K were
encountered, diluters were used in the sampling line. They could be used
to give 394,000 or 1.3 x i0 6 particles/cm 3 full scale deflection. The dilu-
ters have been described by Whitby et al. (1972).
(3) Gas Analysis - - SO 2 concentration was measured with a Meloy
Model SA 160-2 flame photometric total sulfur analyzer. This instrument
was calibrated with a span gas produced with an 502 permeation tube. A
Bendix Model 8101-B N0+N0 2 +N0 analyzer was used to measure oxides of nitro-
gen. A span gas for calibrating the NO 2 scale was again produced using a
permeation tube,while a commercial 20S ppm NO gas was diluted to provide
calibration points for the NO scale. The Bendix instrument was zeroed with
“boil off’ gas from liquid nitrogen. The gas analysis equipment included
an REM Model 612B chemiluminescent ozone analyzer. An ozone generator and
wet analysis tests were used for its calibrations. A Cambridge Instrument
Model 880 hygrometer indicated the dew point temperature, while chamber
temperature was measured with a copper-constantan thermocouple. The chamber
air and dew point temperatures were then used to derive relative humidity.
9
-------
Hydrocarbon concentration was measured with a Hewlett-Packard Model
5700 gas chromatograph. Separation of components was made using a column
packed with SE-30 silicon rubber on chroinasorb. The column was maintained
at 80°C for l-hexene and cyclohexene, while a temperature of 90°C was used
for toluene and m-xylene. The output from the gas chromatograph’s flame
ionization detector was recorded, and peak areas were measured with a plani-
meter for quantitative computations. Hydrocarbon span gases were made by
evaporating small, measured volumes of the liquid hydrocarbons into known
volumes of air.
10
-------
Section 3
CHAMBER LIGHT INTENSITY MEASUREMENTS
3. 1 Calspan thamber
Light source improvements made in the Calspan chamber were completed
by September 1973. The modifications involved the installation of six
additional 85-watt blacklights and two 40-watt sunlainps in each of the 24
lighting modules. The two 215-watt fluorescent white lamps in each module
remained unchanged. This mix of lamps increased the measured kd NO 1 from
-0.05 min to a new level of -0.23 m m . (Since completion of the experi-
ments reported here, all white lamps have been replaced by blacklamps giving
rise to a new kdINo of -0.33 min .)
Initial light intensity measurements were attempted by measuring
the rate constant, kd) for NO 2 photolysis in nitrogen by irradiating a 15 ft 3
nitrogen-filled Teflon bag in the chamber center. The experiments gave
results which suggested that the inner surface of the teflon bag was contami-
nated, since during the experimental period a rise in both the NO and NO 2 was
observed suggesting surface absorption and desorption of NO 2 .
An alternate scheme was adopted following a method reported by
Stedman and Niki (1973) which was found to be fairly repeatable and well
suited to a large chamber such as Calspan’s. The method gives a measured
value of k 1 4 of NO 2 from the ratio of the initial production rates of 03 or
NO by photolysis of NO 2 in clean chamber air. Before reporting the experi-
mental results, however, it is instructive to recognize a few of the inherent
errors regarding the method. Some effects of normal background contaminants
on the accuracy of the method were studied, as well as the conditions needed
for valid interpretation.
1].
-------
A computer model for the N0 +CO+H 2 O_air system was used to assess
possible contaminant effects. This model includes all the reactions consid-
ered by Stedman and Niki, as well as those reactions due to the presence of
CO and H 2 0. Several computer runs have been made for varying NO 2 , NO, CO,
HONO, and k 1 . The results are reported in Table I. The theoretical k 1 1 s
in Table I were calculated from the ratio of the 03 production rates (modeled)
and [ NO 2 ] at 6, 10, and 15 seconds. Changes in [ Ca], [ H 2 0], and [ HONO] have
no appreciable effect on the calculated k 1 ’s, but the accuracy of the calculated
is very much dependent on the [ NO]/ [ NO 2 ] ratio. The accuracy of the
method is also dependent on the response times of the 03 and NO analytical
instruments. For example, commercial ozone chemiluminescent devices have
response times of the order of 1 second; therefore, accurate initial rates
will require at least 5 seconds to be established and probably more realis-
tically 10 to 15 seconds. The error introduced in d [ 0 3 ]/dt (that is, in
terms of its representing 0-atom production from NO 2 photolysis) is due to
ozone loss reactions.
NO +0 3 ÷N0 2 +0 2 (1)
NO 2 + 03 4- NO 3 + 02 (2)
A simple calculation is sufficient to illustrate the error involved in the
method due to ozone loss reactions. Assume we are using the method to
measure a k 1 + of .15 min , and that the initial concentrations are [ NO 2 ]
5.0 ppm and [ NO] = 0.0. In a five-second photolysis period, the following
will have happened.
First, .0625 ppm NO 2 will be lost and .0625 ppm of NO and 03 will
be formed. Assuming a mean concentration of .0312 ppm for NO and 03 in this
5 sec period, ozone loss due to reactions (1) and (2) is then given by,
• 0625 ppm - (.0312 ppm) (5 ppm) (.078 ppm min 1 ) mm
-l • -l 5
- (.0312 ppm)(.03l2 ppm)(23 ppm mm ) mmn
= - .00101 - .00187 = - .00288 ppm
12
-------
Table I.
MODELED RESULTS FOR NO 2 PHOTOLYSIS IN AIR WITh H 2 0 AND CO PRESENT
NO 2
NO
i !2O
CO
HONO
actual
k 1 theoretical
Percentage
ERROR
5 sec
10
sec
15 sec
ppm
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
2.45
2.45
2.45
2.45
2.45
2.45
2.45
2.45
3.00
1.00
1.00
ppm
.01
.01
.01
.01
.01
.01
0.1
0.1
.01
.01
.01
.01
.01
0.1
0.1
0.1
.01
0.1
.01
.01
0.1
ppm
1.5x10 4
1.5x10 4
1.5x10 4
1.5x10 4
1.5x10 2
1.5x10 2
1.5x10 2
1.5x10 2
1.5x10 4
1.5x10 4
1 . 5x10 4
1.5x10 4
1.5x10 4
1. 5x10 4
1. 5x1
1.5x10 4
1.5x10 4
1.5x1O 4
l.5x10 4
1. 5x1 0
1.5x10 4
ppm
0
2.0
10.0
10.0
10.0
10.0
0
10.0
10.0
10.0
0
2.0
10.0
0
2.0
10.0
10.0
0
0
0
0
ppm
0
0
0
0.1
0
0.1
0
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
m m ’
.15
• 15
.15
.15
.15
.15
15
.15
.27
.27
.27
.27
• 27
.27
• 27
.27
15
15
.15
.15
15
mm ’
.136
136
.136
135
136
.135
.123
• 123
• 233
.233
.245
.245
.245
226
.226
.226
• 141
.119
139
• 146
134
mm
.120
.120
.120
.119
120
.119
.103
.103
.220
.195
.220
.220
.220
.192
.192
.192
.131
.112
.130
.139
.116
mm
103
103
103
• 102
103
• 102
.086
.086
• 196
158
• 192
.192
.192
.161
.161
.161
.119
.098
.118
.132
.105
@ 5 sec
9.3%
9.3%
9 .3%
10.0%
9.3%
10.0%
18.0%
18.0%
13.7%
13.7%
9.3%
9.3%
9.3%
16.3%
16.3%
16.3%
6.0%
20.7%
8.0%
2.7%
10.7%
-------
which represents a 5% error. If there were 1% NO initially present, the
error becomes 10%. A similar calculation as above except with a 10 second
interval for initial rate determination gives an inherent error of 20%. For
any given time period at which an initial rate is determined, the inherent
error will therefore be directly related to the k 1 x (NO 2 ) and the initial NO
present. It should be noted that in all cases the error introduced in calcu-
lating k 1 represents a lower value than actually present.
The experimental results for NO 2 photolysis in air carried out in
the Calspan chamber are shown in Table II. The samples were taken from a
distance of about 2 meters from the chamber wall so that the data represent
approximate average light intensity levels in the smog chamber. The initial
ozone production rates were determined in the first five seconds of the run
after switching on the lights. The lights were not temperature stabilized
prior to a run. The simple calculation mentioned above would suggest that
the experimental k’s are -7% too low.
Table II. CALSPAN CHAMBER - NO 2 PHOTOLYSIS IN AIR
[ NO 2 ] [ NO] d [ 0 3 ]/dt k 1 kd
ppm/mm min’ min 1
3.70 .04 .528 .143 .220
3.70 .04 .552 .149 .230
Although it has been shown that there are a number of possible errors
inherent in the use of this method, the results obtained are in close agreement
with the calculated kd expected from the number and type of lights that are
now used in the Caispari chamber.
14
-------
3.2 University of Minnesota Chamber Light Intensity Measurements
For the University of Minnesota smog chamber, a slightly different
procedure was followed using the Stedman and Niki (1973) method. After
admitting approximately 5 ppm NO 2 into the chamber, the 72 F4OBL blacklights
were turned on, and the NO and 03 concentrations were recorded as a function
of time. Very early (during the first few seconds) in such an experiment,
the only important reactions are:
N0 2 +hv÷N0÷O (1)
M+O+0 2 - ’ -0 3 +M (2)
where M is any third body. The 0 quickly achieves a stationary state so that
d [ NO ] - d [ O ] = k 1 N0
dt - dt ii 2
Hence, the initial rate of production of either NO or 03 may be used along
with [ NO 2 ] to determine k 1 . Later in the experiment, reactions (3) and (4)
-become important in removing 03.
NO+ 0 3 ÷N0 2 + 0 2 (3)
NO 2 ÷ 03 ÷ NO 3 + 02 (4)
When [ 031 reaches the maximum
k 1 [ N0 2 1 = k 3 [ NO} [ 03] + k 4 [ N0 2 ] [ 03].
This is called the photostationary state; and under these conditions, k 1 caxi
be calcuLated from measurements of [ NO], [ NO 2 ], and [ 03]. provided k 3 and
are known.
15
-------
Reactions (3) and (4) are fast enough to cause significant losses
of NO and 03 as they pass from the chamber to the measuring instruments.
Consequently, a correction for losses in the sampling lines had to be made.
The residence times in the sampling lines were 0.108 and 0.159 mm for the
ozone and NO instruments, respectively. Corrections for line losses were
made using the following expressions:
0 3’chamber = °3 1 measured + [ O3hline 3 [ N0] 1 + k 4 [ NO 2 ] 1 .)r 0
[ NO] = [ No] + k [ NO] . [ 0 1
chamber measured 3 line 3 line NO
The average values of [ NO] and [ 03] in the lines are not known and were deter-
mined by an iterative procedure. The NO and 03 data obtained were all corrected
using this method.
Values of k 1 were determined using both the initial rate of NO for-
mation and measurements of photostationary state. Measurements of k 1 were
made at the beginning and end of our entire experimental program in order
to determine if deterioration of the lights was significant. The results
of these measurements are listed in Table III. It may be seen that the two
methods are in good agreement, and that the light intensity fell by about
20% during the course of this program. In order to make data comparisons with
results reported in a large body of smog chamber experiments which have already
been performed, light intensities are also expressed in terms of kd values,
where kd is the first order NO 2 photolysis rate which would be observed if
NO 2 were photolyzed in N 2 under the same lights. Under such conditions, the
reactions
0+N0 2 ÷NO÷0 2 (5)
O+N0 2 +M+N0 3 +M (6)
are also important and kd becomes:
kd = d( ln [ N0 2 ])/dt = k 1 k 5 /(k 5 ÷ k 6 [ M]).
16
-------
Inserting the rate constants given by Stedman and Niki gives
kd = k 1 /0.64
This relationship and the average value of for each experiment has been
used to calculate the kd values given in Table III.
TABLE [ II. LIGHT INTENSITY MEASUREMENTS
Run k 1 (min ) kd(min ) Comments
NO 5 lope stationary state
A 0.087 0.084 0.13 Before November
B 0.094 0.097 0.15 Workshop
C 0.073 0.074 0.12 After March
Workshop
The method used above gives local values of light intensity at the
point of sampling from within the chamber which is on the chamber centerline.
Clark (1972) measured the variation of light intensity with radial position in
the chamber with a chemical actinometer utilizing the photoisomerization of
o-nitrobenzaldehyde to o-nitrobenzoic acid. He found that the light intensity
increased significantly as the walls were approached. His results have been
used to calculate the ratio of chamber volume average light intensity to cen-
terline intensity. This ratio is 1.04 for the small bag and 1.45 for the large
bag. Volume average values of kd have been calculated by multiplying these
ratios by the measured values on the chamber axis. The results are presented
in Table IV. Note that the average light intensity in the large bag is about
40% greater than in the small bag because the small bag only contains volume
near the chamber centerline where the low light intensity is lowest. From
these data, kd [ NO for the large U of M bag is 0.20 min .
2 avg
17
-------
Table IV. VOLUME AVERAGE LIGHT INTENSITIES
kd(mln )
Averages
Run Centerline Large Bag Small Bag
A 0.13 0.19 0.14
B 0.15 0.21 0.16
C 0.11 0.15 0.12
18
-------
Section 4
RESULTS AND DISCUSSION
In this section, results of the Calspan and University of Minnesota
experiments are discussed in the order the experiments were performed. Two
joint workshops were held at Calspan during the project year--one in November
1973 and another in March 1974. The purpose in performing the experiments
cooperatively was to take advantage of the additional aerosol measuring capa-
bility at the University of Minnesota. (Calspan did not acquire its own
electrical aerosol analyzer until the summer of 1974.) In addition to the
joint experiments at Caispan, more tests were matched as closely as possible
at the University of Minnesota in order to allow additional interchaniber
comparisons of the aerosol and chemistry data.
In the discussion which follows, results of the workshop data are
first summarized in tables and then treated individually when it is instruc-
tive to do so. Because of the large number of experiments performed during
the year, graphs of the aerosol and chemical data have been combined and
placed in an appropriate appendix. Wherever possible, data from duplicate
experiments performed at the University of Minnesota are matched with the
corresponding Calspan experiment. Within the text, however, only representa-
tive cases from each of the test systems are compared and discussed in detail.
4.1 November Workshop
During the first joint workshop, the experimental schedule was
divided into three phases: (1) SO 2 experiments, (2) toluene, toluene + NO 2
and toluene + NO 2 + SO 2 tests, and (3) hexene, hexene + NO 2 and hexene + NO 2 +
SO 2 experiments. The initial SO 2 experiments were performed to compare the
effects of light intensity and chamber contamination on aerosol production
and SO 2 photooxidation rates in the S0 2 -clean air system. The experiments
included irradiations of SO using the old lighting configuration (k -0.05
min ) in a “dirty’ t chamber (contamination on the walls due to auto exhaust
19
-------
irradiations performed on another contract was not cleaned prior to the first
test) followed by several experiments after washing the chamber with distilled
water. A summary of all experiments performed during the workshop is given in
Table V. Graphs of the aerosol and chemistry data are shown in Appendix A and
matched with the comparable University of Minnesota test where possible.
As in the past, computations of SO 2 oxidation rate were made from
the aerosol volun e data generated by the EAA. The procedure is described
by Clark (1972); briefly, however, the rate of production of sulfuric acid
aerosol, corrected for molecular weight change, and water concentration is
assumed to be equal to the rate of photooxidation of SO 2 which is constant
for the linear growth portion of the experiment. Thus, the slope of the
straight line volume growth curve may be related directly to the rate of
photooxidation of SO 2 . The governing equation is:
d [ SO] d
- dt = f (p)(P)( 7 !) (7)
where p is the density of the sulfuric acid droplet, P is the weight fraction
of H 2 S0 4 in the droplet, is the molecular weight of SO 2 , and 2 is the
molecular weight of 11 2 S0 4 . The quantities p and P can be determined from
data given by Bray (1970) assuming that water vapor in the gas phase is in
equilibrium with water in the aerosol droplets.
• SO 2 Experiments
The first six experiments at Calspan were SO 2 irradiations under
either partial lighting or full light intensity. After experiment 2, the
chamber walls were cleaned using a triplicate rinsing with distilled water.
The water washing did not appear to have affected the SO 2 photooxidation
rate as shown by the data in Table V. Note that an initial and final photo-
oxidation rate has been computed for some cases based on the form of the
volumetric growth curve (i.e., for the first 30 to 60 mm, a slower initial
rate was often observed followed by somewhat faster growth, probably due to
20
-------
Table V. SUMMARY OF DATA FROM NOVEMBER 1973 WORKSHOP
Run
No.
System
R h
%
HC
ppm
y volume
NO 2
ppm
SO 2
ppm
dv
Nniax I s (502)
xl0cC m 2 c l_h .]1 pm -ca’-h
SO-,
photox.
%-hr
1
2
3
4
5
6
SO 2
So 2
SO 2
SO 2
SO 2
SO 2
30
33
34
31
32
b
b
b
b
b
b
b
b
b
b
b
0.63
0.70
0.58
0.55
0.52
0.55
560
700
525
570
960
750
470
not
reached)
>600
>600
>1100
1600
1.71
6.l5i
13.75f**)
l.60i )
3.60f .1
1.20i
2.83f .3
7.33
6.0 i
16.7 f )
.03
.17 ‘
.37 3
.O5
.iii
.O4
.ogj
.27
.2l
.57.)
7
8
9
10
11
12
toluene
toluene + NO 2
toluene + NO 2
toluene + NO 2 + SO 2
toluene + NO 2 + SO 2
toluene + NO 2 + So 2
35
37
34
45
36
33
0.8
0.8
0.8
0.8
0.8
0.8
b
5.0
3.3
1.95
3.55
1.45
}
.19
.07
.01
NO AEROSO
1000
800
500
L
1000
then less
700
then less
280
then less
7.50i
l.90f .3
5.OOi
0.41f .3
0.88i
0.28f .3
.64 1
.18 .3
1.30
0.10 j
1.65 ‘
.53.3
13
14
15
hexene
hexene + NO 2
hexene + NO 2 + SO 2
38
34
32
0.6
0.6
0.6
b
3.35
1.64
---
---
.01
95
46
580
>50
---
400
---
---
2.4 i
0.9 f .3
---
---
4.571
1.71.3
lights used on experiments 1, 3 and 4 to duplicate previous year’s light intensity, i.e., kd [ NO ] .O.OS mm 1
i and £ refer to initial (usually first 30 mm) and final growth rates.
= background
-------
contributions from contaminants). It is interesting to note that the average
initial SO 2 photoox]dation rate for the two lighting configurations reflects
very nearly the measured kd ratio:
kd new lights 0.22 SO 2 ox. rate Cay, new) 0.22
kd old lights = 0.05 = 4.6 so 2 ox. rate (av. old) = 0.04 =
This relation is al.so true when comparing the average final SO 2 oxidation
rates for the two light intensities.
Figures 2 and 3 show typical SO 2 aerosol behavior for linear growth
and also for a system in which there is upward curvature in the volumetric
growth rate. Number concentration grows rapidly after the lights are switched
on as a result of homogeneous nucleation processes. As the experiment pro-
gresses, the rate of nucleation drops until the production of particles by
nucleation is balanced by the removal of particles by coagulation. The sur-
face concentration grows rapidly initially as surface is formed by the
nucleation of new particles and growth of existing particles. The rate of
growth of surface concentration slowly decreases as the nucleation rate
drops and the removal of surface by coagulation becomes more important. The
surface concentration then slowly tends toward a steady-state value where
the rate of production of new surface by nucleation and condensation is
exactly balanced by the rate of removal of surface by coagulation. The
volume concentration frequently grows slowly at first and then grows linearly.
This results from a constant rate of oxidation of SO 2 which leads to a con-
stant rate of production of sulfuric acid mist. On occasion, there is upward
curvature to the volumetric growth rate probably due to trace contaminants
within the chamber air which contribute to aerosol formation. Substantial
contamination (due mainly to inadequate filtration or the previous history
of experiments) manifests itself in the form of a very pronounced increase
in the conversion rate or large upward curvature or both. This behavior can
he seen by comparing the data shown in Figures 2 and 3.
22
-------
10
8
6
4
2
0
CALSPAN
o AEROSOL VOLUME CONC. p m 3 /cm 3
A AEROSOL SURFACE CONC. pm 2 /cm 3
O AEROSOL NUMBER
12— 12
x io2
10 —
8— E
U
E
6—
uJ
C.,
I L .
4- D
(I)
2—
TIME (mm)
FIGURE 2. SO 2 EXPERIMENT SHOWING LINEAR VOLUMETRIC GROWTH.
x10 5 x10 3 x10 1
30
10 —
E
U
E 8— E
.9
LU
2
U-
z
U)
4
2—
0 -.
2
1
0 40 80
120
TIME (mm)
E
20
E
UJ
2
-J
0
>
10
0
160 200 240
x .101
20
C d )
E
U
Cd)
12 E
LU
2
8D
-J
0
>
4
FIGURE 3.
SO 2 EXPERIMENT SHOWING UPWARD CURVATURE OF AEROSOL VOLUME.
RUN NO. 5 NOV. 13. 1973
SO 2 = 0.52 ppm
STIRRING
= 7.33 p ms/CC -hr
dt
io 5
Cd)
E
U
LU
2
z
0
0
20 40 60 80 100 120 140 160
23
-------
• HC + NO 2 Experiments
Experiments 7-15 were carried out to study the effects of toluene
and hexene on aerosol formation. The irradiation of these hydrocarbons with
background levels of NO does not lead to appreciable aerosol formation, a
fact that was also observed in the comparable U of M experiments for toluene.
The toluene and hexene systems with added NO 2 also showed very little aerosol
formation; however, the ratios of HC:N0 2 were always much less than one, and
complete domination of the chemistry by the excess NO 2 present has resulted.
The relatively small concentration of toluene present in these runs had only
a minor effect in converting NO to NO 2 . Some additional discussion relative
to the mechanisms at work in the HC÷N0 2 system during the photolysis period
is given below.
The photolysis of NO 2 in clean air proceeds through a rather complex
reaction mechanism, the most important steps of which are given below.
NO 2 + hv - NO + 0
0 + 02 + 1sf 03 + M
03 + NO •9• 02 + NO 2
03 + NO 2 -3. NO 3 + 02
NO 3 + NO 2 ÷ NO + NO 2 + 02
NO 3 + NO 2 N 2 0 5
N 2 0 5 + H 2 0 - 2H0N0 2
It has been observed, both experimentally and theoretically, through modeling
techniques that during NO 2 photolysis a gradual formation of NO occurs. The
rate of NO production is a function of light intensity, initial NO 2 concentra-
tion and relative humidity.
24
-------
The addition of hydrocarbon to the NO 2 system introduces free radi-
cal species formed as a result of the hydrocarbon photooxidation. The major
hydrocarbon photooxidation processes are thought to proceed via attack by one
or more of the following species: 0-atom, O3 HO, F10 2 , NO 3 , and CH 3 O. The
first three species arethought to contribute most significantly, though
their order of importancewill vary with respect to the structure of the hydro-
carbon being oxidized. Peroxy radicals, a transient species formed in the
hydrocarbon oxidation process, are of major importance in the nitric oxide
oxidation process. Reactivity scales have been formulated based on the rate
at which hydrocarbons catalyze the oxidation of nitric oxide, but unfortunately
these scales can be ambiguous. For instance, in any given light condition,
the rate of nitric oxide oxidation is dependent upon the hydrocarbon to nitric
oxide ratio, the initial NO 2 concentration, as well as the experimental system
itself. Table VI contains reactivities for several hydrocarbons and their
reaction rate constants with 0-atom, 031 and HO.
It appears from the toluene-N0 2 experiments, runs 8 through 12, that
the toluene-N0 2 ratios chosen were such that the rate of NO oxidation, due to
the oxidation of toluene, was too slow to compensate for the NO format]on
rate resulting from the NO 2 photolysis mechanism mentioned earlier. There-
fore, in all of the toluene studies, NO was never oxidized out of the system.
Also, any observable loss of toluene was within the experimental error of the
analytical system.
In the hexene-N0 2 studies, run 14 (see Appendix A) shows an effect
similar to that observed in the toluene system, while run 15 with a higher
HC/NO ratio shows oxidation of virtually all the hexene-1 and conversion of
all the NO formed in the system to NO 2 . For the comparable toluene experiment,
oxidation of the HC was very slow and incomplete. These results suggest a
greater difference in reactivity between toluerie and hexene-l than ‘fable VT
would indicate. One possible explanation is that aromatics may vary in
reactivity with respect to the HC/NO ratio much differently than the olefins.
For example, toluene’s reactivity is mostly dependent on the HO-toluene
25
-------
reaction, while hexene-l reactivity receives comparable contributions from
0-atom, 03. and HO reactions. Under the NO 2 -toluene conditions in these
experiments, the 1-10-NO 2 reaction dominates, virtually eliminating a HO-toluene
chain mechanism and thus exhibiting the very low reactivity observed.
In the cases of the hexene-NO 2 studies, the results suggest that
the conditions for short and long chain length reactions involving HO were
achieved respectively in Runs 14 and 15.
Table VI. HYDROCARBON REACTIVITY
Compound
- -l - -l
k in ppm nan
- (4)
Reactivity
O-atom 1
o ( )
rj( 3 )
ethylene
propylene
hexene-1
toluene
NO 2
7.7 x io2
4.4 x 1O 3
3
5.0 x 10
2
1.7 x 10
3(6)
8. lxlO
3.8 x l0
1.6 x l02
2
1.5 x 10
..5 )
1.8 x 10
7.8x10
2.5 x 1O 3
2.5 x l0 3.5
4(**)
6.0 x 10 ‘ 1.7
4(***5
2.0 x 10 1.3
4(6)
1.5x10
1 Cvetanovic, R.J., Adv. in Photochemistry 1, 115 (1963).
2 Wei, Y.K. and Cvetanovic, R.J., Can. J. Chem. 41, 913 (1963).
3 Morris, E.D., Jr. and Niki, H., J. Phys. Chem. 75, 3640 (1971).
4 Glasson, W.A. and Tuesday, C.S., Environ. Sci. Technol. 4, 916 (1970).
5 Stedman, D.H. and Niki, H., Environ. Letters, 4, 303 (1973).
6 Demerjian, K.L., Kerr, J.A. and Calvert, J.G., Adv. in Environ. Sci. Technol.
Vol. 3, Wiley-Interscience, New York (1973).
Based on the average rate of NO photooxidation (ppb/min); kd = 0.29 min and
I - IC to NO ratio of 2.5.
**
Rate constant based on reaction of HO with pentene-l.
Upper limit based on rate constant of HO-xylene reaction.
26
-------
• HG + NO 2 + SO 2 Experiments
The addition of SO 2 to the HG -NO 2 systems had a profound effect on
aerosol formation. In reviewing the data in Table V, the toluene + NO 2 + SO 2
systems (runs 10, 11, and 12) and the hexene + NO 2 + SO 2 experiment (run 15),
show rapid initial aerosol growth followed by somewhat slower volume production.
Similar characteristics were found in the U of M experiments. The apparent
SO 2 oxidation rates for these systems were quite high, initially up to several
percent per hour, followed by a slower final rate that is very similar to that
of SO 2 alone. Matched aerosol and chemistry data from a representative Caispan
and U of M experiment for the toluene + NO 2 + SO 2 system can be seen in Figures
4 and 5.
In view of the chemical profiles for these experiments, we believe
the following processes can account for the observed aerosol behavior. In
the early stages of the HC+N0 2 +S0 2 runs, there are two major sources of
aerosol formation, that resulting from SO 2 oxidation and the other from the
reaction of 03 with the hydrocarbon. Rapid volumetric growth occurs and,
therefore, the apparent initial SO 2 photooxidation rate is quite high. As
irradiation continues, ozone levels quickly decline (see Figure 4) resulting
in the loss of one important source of aerosol. This, in turn, causes a
change in the observed volumetric aerosol growth rate of the system. In the
absence of SO 2 , the HC+N0 2 system shows only one growth mode which is con-
sistent with the above explanation of aerosol behavior.
4.2 Duplicate Experiments to the November Workshop - University of
Minnesota
After completing the joint workshop at Calspan during November,
a similar set of experiments was performed at the U of M. Some additional
SO 2 irradiations were performed after HG-NO 2 experiments in order to observe
the effect of possible chamber contamination on the SO 2 oxidation rate. The
results of the Minnesota experiments are summarized in Tables VII and VIII.
Three systems were investigated: SO 2 , toluene + NO 2 , and toluene + NO 2 + SO 2 .
27
-------
FIGURE 4
CALSPAN 1N0 2 + SO 2 + TOLUENE SYSTEMI
RUN NO. 11 16 NOVEMBER 1973
NO 2 = 3.6 ppm; NO = 0.025; so 2 = 0.07 ppm
STIRRING
R.H. = 37%; TOLUENE = 0.8 ppm;
100
io2
C?
E
U
c J
E
uJ
C.,
U.
D
C l )
8
6
2
0
C ’)
E
U
C?
E
‘U
D
-I
0
>
250 300
TIME (mm)
4
100
16
X 101
3.
-a
0.
C d )
20
0
1
E
0.
0.
N
0
z
0 50 100 150 oo 250 300 so 400 450 500 550
TIME (mm)
28
-------
RUN 15 O TE 28—DEC—73 SYSTEM TOLUENE S02 . H02
0 HUM. PART. ‘riL)
SURF. ( pri 2 ’ML
U of M.
0 UOL.c JM 3 /ML.
N
LI -
h
B
E
Ni
-S
• II
-R
:F
4—
3—
1—
0—
U
0
L
IJ
2 r1
E
x l
TI tIE
I U I
1
10+’ r1IHIJTE)
I I
I I I I I
N
I
T
R
0
G
E
H
0
x
Ii
0
E
$
x1
•LTS. OFF
2
D HO (PPM) HO., (PPM)
1
TIME (10+2 rIINUTE)
FIGURE 5
29
-------
C,1
0
TABLE VII. UNIVERSITY OF MINNESOTA DUPLICATE TESTS SUBSEQUENT TO NOVEMBER WORKSHOP:
SUMMARY OF SO 2 EXPERIMENTS
RUI
No.
Concentration ppm
RH
%
1
#t-cc
SE
pm 2 -cd
c i. ’ ,
cit SO 2
3 -1
urn -cc -hr
1 civ
SO 2
3 1
tim -dd-hr -ppm
°2
photox
%-hr 1
1
.51
13
90 K
420*
2.3
4.5
.12
2
.53
9
150 K
420*
2.6
4.9
.13
3
.64
64
320 K
660
3.1
4.8
.057
4
.59
36
170 K
340
.98
1.7
.030
5
.55
35
160 K
380*
.93
1.7
.032
6
.55
29
110 K
210
.49
.89
.018
11
.55
30
600 K
1300
12.0
22
.42
12
.47
49
330 K
1400
9.8
21
.35
17
.54
27
220 K
420
1.1
2.0
.041
18
.012
17
54 K
72*
.12
10
.23
19
.012
36
35 K
25*
---
---
20
.54
29
280 K
620*
2.2
4.1
.079
Equilibrium surface not reached.
-------
have been responsible for the observation because number concentration con-
tinued to decay normally, and also a second phase of growth was not apparent
in the comparable Calspan experiment. For Run 14, the EAA data proved to
be unreliable and only number against time data were obtained. Following
EAA repair,a complete set of aerosol data was obtained for the remainder of
the experiments in the series. Runs 14 and 16 were made with the toluene +
NO 2 + SO 2 system. The aerosol growth behavior of these systems is different
from the S0 2 —pure air system in that two distinct growth phases are evident:
rapid initial volume production for about the first hour, as observed in the
Caispan experiments, followed by a slower but essentially linear increase
with time. The volume production rates during the linear rate periods are
tabulated in Table VII, as well as the corresponding apparent SO 2 photooxi-
dation rates. Comparison of these results with Calspan’s for the same chemi-
cal system reveals very similar behavior. The shapes of the plots of N, S,
and V against time, shown in Figures 4 and 5 and also Appendix A, are the
same except that the surface concentration obtained in the Calspan experiments
is not as constant as in the Minnesota tests. The Calspan and Minnesota
experiments were performed at different SO 2 concentrations and, consequently,
the resulting aerosol concentrations were different. However, when the volu-
metric production rates during the linear growth phase are normalized by
converting them into apparent SO 2 photooxidation rates, good qualitative
agreement is obtained between the two sets of experiments. Thus, on the
average, Caispan SO 2 oxidation rates with full light intensity are about
O.2%/hr, while in the U of M chamber, values less than 0.1%/hr are usually
noted.
The last group of experiments in these series, 17-20. were all SO 2
experiments. The usual SO 2 photooxidation behavior is evident. Apparent
photooxidation rates in these experiments are nearly a factor of two higher
than before the toluene experiments were performed. This suggests that some
chamber contamination was produced in the toluene experiments.
33
-------
4.3 Conclusions from the November Workshop
From a complete review of the Calspan and University of Minnesota
duplicate experiments involving 502 + clean air, I-IC + NO 2 , and HC + NO 2 + SO 2
system, the following points emerge:
(1) 502 photooxidation rates of -0.2%/hr are typically observed in
the Calspan chamber and rates generally less than -0.1% /hr and occasionally a
few hundreths of a percent hr are found in the U of H chamber.
(2) A conditioning effect in which each subsequent SO 2 irradiation
produces slightly less aerosol is normal in both chambers. The effect is
more pronounced in University of Minnesota tests. Generally, the history of
previous experiments does not appear to have as large an effect on the SO 2
oxidation rates in the Calspan chamber as it does in the U of N chamber.
This could be due to the very large size of the Caispan chamber and also
to the method of air purification (i.e., recirculation through charcoal fil-
ters until the desired level of cleanliness is achieved).
(3) lrradiations of toluene or hexene + clean air produce no
aerosol in either the Calspan or U of M chambers. The addition of excess
NO 2 does not cause appreciable additional aerosol formation in either
chamber.
(4) The addition of cvcn small amounts of SO 2 (-0.01 ppm) to the
HC ÷ NO 2 systems had a profound effect on aerosol production. Rapid initial
aerosol formation appears to be due to a combination of SO 2 oxidation (and
subsequent formation of 112504 particles) and ozone reaction with HG. As
the ozone levels decline, only the SO 2 photooxidation mechanism remains
which leads to final volume production rates similar to that for SO 2 alone.
34
-------
4.4 March 1974 Workshop
A second joint workshop was held for a three-week period in March
1974. The test series was designed to investigate chemical conversion and
aerosol behavior of various HC+NO and RC÷NO+S0 2 systems using realistic con-
centrations of reactants, The hydrocarbons chosen for study were toluene,
hexene-l, m-xylene, and cyclohexene. NO was used in place of NO 2 in order
to more closely simulate photochemical processes responsible for aerosol for-
mation in an urban environment. The normal concentrations of reactants used
in the experiments were 0.35 ppm 1 - iC, 0.15 ppm NO, and 0.05 ppm SO 2 . As in
the November workshop, a number of SO 2 + clean air experiments were performed
as part of normal chamber characterization and contaminant monitoring proce-
dures.
Two series of experiments are reported in this section. The first
represents data obtained during the joint Calspan/University of Minnesota
workshop held at Caispan in March 1974, and the other consists of duplicate
experiments performed at the U of M following the workshop. Since Caispan
did not yet have its own aerosol analyzer at the time of the March workshop,
there was no opportunity to repeat any of the experiments once the workshop
was over. In several instances, it would have been instructive to do so.
In spite of this limitation, there is generally very good agreement of the
data in both a qualitative and quantitative sense.
A total of 31 experiments were performed during the workshop using
various 1- IC, HC4-50 2 , HC4-NO, and HC+NO+50 2 mixtures. Following the workshop,
additional experiments were performed at the U of M. Many extra 502 experi-
ments were performed to study the effects of chamber contamination on SO 2
oxidation rate. All chemical and aerosol conversion data for these experi-
ments are presented in Appendix B, together with the duplicate U of M experi-
ments. Several specific examples are discussed within the text. In order to
preserve continuity, the experiments are discussed in the following sequence:
SO 2 experiments, HC+NO tests, NC + SO2 and I-IC+NO+S0 2 experiments. Four
experiments were also performed at Calspan using various concentrations of
35
-------
NaCl particles. The aerosol and chemistry data are shown in Appendix B but
not discussed in the text, since the NaC1 aerosol generally masked the impor-
tant features of the aerosol surface and volume behavior.
4.4.1 SO 2 Experiments and Influence of Bag Size on Aerosol Production
In order to allow direct comparison of the pertinent chemical and
aerosol data, summary tables from both the Caispan workshop and U of M
experiments are shown in Tables IX through XIII. At the beginning of the joint
workshop, four SO 2 experiments were performed using concentrations of either
0.50 ppm or 0.05 ppm 50 The data are consistent in that each experiment
shows an initial SO 2 photooxidation rate of approximately 0.2% hr , followed
by a somewhat accelerated rate. As previously discussed, the accelerated
or ‘final’ rate is thought to be the result of contributions to aerosol growth
by background contamination in the chamber air. Note that the final photooxida-
tion rate for the lower concentration runs (numbers 2 and 3) is somewhat ele-
vated compared to the high concentration experiments. It appears that when
the 502 concentration is fairly high, i.e., 0.50 ppn the contributions to
aerosol growth from trace contaminarits,is quite small. Thus, for a slightly
contaminated chamber, the higher the °2 concentration the more nearly the
data represent 502 photooxidation alone. From the large number of SO 2
irradiations performed in the Calspan chamber over the past two years,
it appears that the normal oxidation rate of SO 2 in filtered air is approxi-
mately 0.2% hr .
A substantial number of additional experiments were performed by
the i i of H to test the effects of bag size and condition. The dimensions of the
smog chamber bags used in these experiments are as follows:
Dimensions of Smog Chamber Bags
Large Bags Small Bags
Diameter 3.05 m 1.16 m
Height 2.44 m 2.16 in
Volume 17.8 i n 3 2.28 in 3
Surface/Volume 2.13 in 4.38 in
36
-------
The results of the experiments are presented in plots of aerosol number, sur-
face and volume concentrations in Appendix B and summarized in Table XIII.
The first experiments, numbers 45, 46, and 49, were performed in large bag #3.
This bag was fabricated shortly after the November workshop was completed and
had been used for a number of SO 2 oxidation and NO photolysis experiments.
It proved to be impossible to eliminate dark reactions in this bag, and con-
sequently, it was only used for system characterization studies.
The SO 2 photooxidation rates calculated from the aerosol volume
production rates are 0.07, 0.06, and 0.10 per hour for runs 45, 46, and 49,
respectively. The latter rate is believed to be higher because chamber
contamination was produced by runs 47 and 48, which were NO photolysis experi-
ments. These rates of SO 2 photooxidation do not relate directly to that
expected in a clean S0 2 -air system but rather give a measure of the contami-
nation present in a particular experiment.
The next group of SO 2 experiments 52, 53, 54, 57, and 66 were per-
formed in the small bag. Data showing the influence of bag size on aerosol
production are shown in Table XIV. Only seven experiments were performed in
the small bag because its volume was inadequate to allow a complete set of
data to be obtained during long experiments. The SO 2 photooxidation rates with
pure SO 2 and in the early phases of the SO 2 + NO + HC experiments were found to
be about a factor of two lower in the small bag compared to the large bag. This
is believed to be due to three factors:
(a) The average light intensity in the U of M small bag is lower
because it is located at the center of the chamber and contains none of the
high intensity regions near the walls,
(b) The chamber itself is less contaminated, and
(c) The walls of the small bag may act as a sink for photochemically-
produced species.
37
-------
TABLE IX. SUMMARY OF AEROSOL DATA FROM MARCH WORKSHOP
Run
No. System
RH
%
Nmax
SE
dv
SO2
m 3 -cc -hr’
dv SO 2
t [ max] Photox
um3_cc_1_hr_1%_hrl
Comments
6 toluene + NO
30 toluene + NO
29 toluene ÷ NO + SO 2
7 toluene + SO 2
30
20
30
20
3.1x10
l.3xl0
l.6x10
2.lxlO
640
750
>750
800
--
--
0.78
1.17
2.2
2.6
1.5
3.2
--
--
0.32
0.45
no vol. first 4 hrs
no vol. first 4 hrs
1st 4 hrs*
1st 50 min*
5 hexene + NO
21 hexene + NO
18 hexene + NO + SO 2
20 hexene + SO 2
40
37
37
35
1.4xl0
..2x10
1.4xl0
3.6x10
610
215
>1500
950
--
--
0.61
0.75
2.1
0.5
5.8
3.2
--
--
0.16
0.25
no vol. first 5 hrs
no vol. first 6 hrs
1st 6 hrs*
1st 60 min*
15 m-xylene + NO
14 m-xylene + NO + SO 2
17 m-xylene + SO 2
38
29
35
8.4x10
2.6xl0 5
2.8x10
1150
2700
384
--
0.92
0.84
14.1
25.0
1.6
--
0.32
0.23
no vol. first 60 mm.
1st 60 min*
1st 60 min*
10 cyclohexene ÷ NO
12 cyclohexene + NO
9 cyclohexene + NO + SO 2
13 cyclohexene + SO 2
1 0.52 ppm SO 2
4 0.55 ppm 502
38
30
30
35
25
30
3.6x10
4.2x10 5
l.7x10 5
2.7x10
5.5x10 5
3.9x10 5
2.3x10 5
2.9x10 5
3500
2450
4200
1300
>1450
4400
--
--
0.74
1.20
5.61
10.60
4.49
13.60
110
75
105
10
--
--
--
--
0.28
0.49
0.21
0.39
0.16
0.48
no vol. first 90 mm
no vol. first 3 hrs
first 2 hrs
first 30 mm
1st 30 min*
1st 2 hrs
1st 30 min*
1st 2 hrs
2 0.05 SO
2
3 0.05 SO
2
37
40
>575
>675
0.65
2.35
1.04
2.31
--
--
0.23
0.79
0.36
0.79
1st 30 min*
1st 2 hrs
1st 40 min*
1st 2 hrs
*Time over which aerosol growth rate was used in computing 502 photooxidation.
-------
Table X . UNIVERSITY OF MINNESOTA DUPLICATE TESTS SUBSEQUENT TO THE MARCH WORKSHOP:
SUMMARY OF AEROSOL DATA FOR HYDROCARBON EXPERIMENTS
dv 1 dv
dv SO
Run RH SE ir [ max] ph to
No. System #-c ’ um 2 -cc jim 3 -cc - - jm 3 _cc_l_hr_1_ppm_1 cc -hr oxidation
-
65 toluene + NO + SO 28 185 K 330* 1.23 11.3 same .23
76 toluene + NO 2 47 4.2 K 550* -- -- 24.5
77 toluene NO + SO 57 170 K 1850 1.17 29.9 27.4 .39
87 toluene + NO 2 30 10 K 340 -- -- 8.6
88 toluene + NO + SO 2 24 160 K 1600* 1.22 30.5 16.6 .67
0
60 hexene + NO + So 28 74 K 1200* .40 5.7 21.2 .12
78 hexene + NO + SO 2 55 230 K 1530 .53 13.9 19.1 .18
92 hexene + NO 2 33 8.8 K 31 -- -- .09
93 hexene + NO + SO 2 32 150 K 1300 .29 8.5 18.0 .16
81 m-xylene + NO 75 23 K 1600 -- -- 73
82 m-xylene + NO + SO 54 230 K 2800 .49 10.4 67 .14
89 m-xylene + NO 2 26 21 K 1000* - - - - 38 - - -
91 m-xylene + NO + So 2 26 230 K 1600 .46 10.0 31 .21
83 cyclohexene + NO 51 .9 K 320 -- -- 50
94 cyclohexene + NO 31 2.7 K 510 -- -- 65
95 cyclohexene + NO 29 1.9 K 620 -- -- 190
96 cyclohexene + NO + SO 2 28 280 K 5400 .43 9.7 250 .20
*
Equilibrium surface not reached.
-------
TABLE XI.
SUMMARY OF CHEMISTRY DATA FROM MARCH WORKSHOP
Run RH HC SO 21 NO 1 t(NO )
4 max
No. System % ppm ppm ppm m m ppm
0
6
30
29
7
toluene + NO
toluene + NO
toluene + NO +
toluene + SO 2
502
30
20
30
20
0.35
1.17
0.35
0.35
--
--
0.05
0.05
0.170
0.530
0.146
b
400
480
330
b
>
0.285
0.380
0.225
0.047
5
21
18
20
hexene + NO
hexene + NO
hexene + NO +
hexene + SO 2
SO 2
40
37
37
35
0.33
0.33
0.33
0.33
--
--
0.07
0.055
0.150
0.180
0.178
b
420
420
430
b
>
0.200
0.275
0.052
15
14
17
m-xylene + NO
m-xylene + NO
m-xylene + SO 2
+ SO 2
38
29
35
0.34
0.34
0.34
--
0.055
0.07
0.150
0.150
b
100
105
b
0.222
0.305
0.030
10
12
9
13
cyclohexene +
cyclohexene ÷
cyclohexene +
cyclohexerie +
NO
NO
NO +
SO 2
SO 2
38
30
30
35
0.33
0.33
0.33
0.33
--
0.05
0.06
0.138
0.140
0.220
b
120
190
180
b
0.190
0.192
0.325
0.011
-------
Table XII. UNIVERSITY OF MINNESOTA DUPLICATE TESTS SUBSEQUENT TO THE MARCH WORKSHOP:
SUMMARY OF CHEMICAL DATA FOR HYDROCARBON EXPERIMENTS
Run RH HC SO 21 NO 1 [ NO 2 ] t [ NO 2 ]m [ 0 3’max
No. System ppm ppm ppm ppm mm ppm
65 toluene + NO + SO 2 28 .35 .108 .30 .145* 460 .2*
76 toluene + NO 47 .35 -- .152 .095 210 .30
77 toluene + NO + SO 2 57 .38 .039 .155 .115 160 .362
87 toluene + NO 30 .35 -- .155 .140 130 .402
88 toluene + NO + SO 2 24 .35 .040 .17 .122 155 .315*
60 hexene + NO + SO 2 28 .35 .07 .16 .123 395 .162*
78 hexene + NO + SO 2 55 .35 .038 .165 .130 255 .438
92 hexene + NO 33 .35 -- .12 .104 280 .290
93 hexene + NO + SO 2 32 .35 .034 .122 .125 350 .302*
81 m-xylene + NO 75 .35 -- .155 .144 80 .343
82 m-xylene + NO + SO 2 54 .35 .047 .151 .130 94 .361
89 m-xylene + NO 26 .35 -- .132 .142 68 .379
91 m-xylene + NO + SO 2 26 .35 .046 .117 .115 70 .262
83 cyclohexene + NO 51 .35 -- .13 .101 90 .32*
94 cyclohexene + NO 31 .35 -- .103 .108 60 .20
95 cyclohexene + NO 29 .35 -- .124 .128 103 .254
96 cyclohexene + NO ÷ SO 2 28 .35 .045 .133 .130 85 .241
*max not reached by end of irradiation period.
-------
THE MARCH WORKSHOP:
MINNESOTA DUPLICATE TESTS SUBSEQUENT TO
SUMMARY OF SO 2 EXPERIMENTS
TABLE XIII. UNIVERSITY OF
Run
No.
Concentration ppm
RB
%
N
-1
cc
dv 1
SE dt SO
2 -1 um3_cc_1 hr_1
iim -cc
i.im
1 dv
502 dt -
3 1 1 1
cc hr -ppm
SO 2
photox
-hi ’
145
.59
28
260 K
1470
1.9
3.2
.07
146
.59
140
3140 K
570
2.2
3.8
.065
149
.59
67
590 K
800
5.6
9.4
.10
52
.56
36
90 K
190
.46
.83
.015
53
.56
30
670 K
2100
20
3.7
.07
514
.56
25
100 K
190
.28
.50
.01
57
.56
12
39 K
210
.78
1.4
.04
66
.61
27
100 K
680*
2.7
4•5
.10
68
.62
39
914 K
310
1.1
1.8
.03
69
.053
140
280 K
270
.84
16
.27
70
.54
57
210K
*
1400
1.2
2.2
.03
71
.048
ss
95 K
*
200
.32
6.6
.09
72
.50
5 14
220 K
8140*
4.2
8.5
.12
79
.038
70
68 K
*
110
.21
5.6
.06
80
.40
78
250 K
550
2.2
5.5
.05
65
.38
314
2145 K
6140
2.5
6.6
.13
86
.051
27
82 K
190
.35
6.8
.14
97
.035
27
100 K
*
270
1.1
30
.63
Equ. 1 .1ibri surface not reached, maximum surface.
-------
If (c) is indeed an important factor, it may help to explain the slightly
higher (—0.2%) SO 2 oxidation rates noted in the Calspan chamber under clean
conditions and similar lighting intensities. No clear trend is apparent when
the maximum volumetric aerosol production rates observed with SO 2 + NO + IC
in large and small bags are compared. in the systems studied, the growth
phase corresponding to the most rapid rate of aerosol production came so late
in the experiment that the bag was nearly collapsed, and so it is difficult
to point to any clear-cut effects. Additional experiments should be performed
with more reactive hydrocarbons in order that the times to NO and 0
2 [ max] 3 [ max]
are achieved earlier in the experiment.
TABLE XIV. INFLUENCE OF BAG SIZE ON AEROSOL PRODUCTION
Run
System No.
Small Bag
Large Bags
dv
a [ max]
3 -l -l
pm -cc -hr
Apparent SO
Photooxidat?on
Rate %-hr
Run
No.
dv
a [ max]
3 —l -l
pm -cc hr
Apparent SO 2
Photooxi dat ion
Rate %-hr 1
S0 2 +clean air (52,53,
54,57,
60)
5O 2 +NO+1-fexene (69)
SO 2 +NO+toluene (65)
--
21.2
1.4
0.05
0.12
0.23
11
runs
78
77,
88
19.1
22.0
0.10
0.18
0.53
In reviewing the data in Table XIII, U of M experiments 52 and 54
show low apparent SO 2 photooxidation rates of 0.016 and 0.01% hr , respectively.
Experiment 57 was performed after two NO photolysis experiments and shows only
slight contaminatiom in its higher 502 photooxidation rate of 0.04%/hr. Experi-
ment 66 shows more evidence of contamination with an apparent 502 photooxidation
rate of 0.1% hr . Since run 65 was a toluene + NO + 502 experiment, the higher
rate of photooxidation in run 66 is possibly a result of contaminants remaining
on the bag surface after the previous run.
43
-------
The remaining SO 2 experiments in this series were performed in a
new large bag (LB-4). Run 68 was the first SO 2 experiment in the large bag;
no dark growth was observed, and the apparent rate of SO 2 photooxidation was
0.03%/hr. Runs 69-72 were all SO 2 experiments and showed a considerable
variation in the apparent SO 2 photooxidation rate. This type of variation
has been observed previously with new bags and is believed to be related to
a conditioning process in the chamber. (The SO 2 oxidation rate tends toward
lower values but will vary in both directions depending on the purity of the
chamber air.) After run 72, a series of hydrocarbon experiments were started.
The next SO 2 experiments were runs 79 and 80 which immediately followed a
1-hexene + NO + SO 2 experiment. The apparent SO 2 photooxidation rates in
these experiments are fairly low, 0.06 and 0.05%/hr, respectively, but higher
than the lowest levels attained in the first large bag (0.02%/hr in run 6)
used after the November workshop. It is of interest to note that the apparent
photooxidation rates in runs 79 and 80 are nearly equal, despite the fact that
the SO 2 concentration in run 79 is 0.038 ppm, whereas in run 80 it is 0.40 ppm.
This suggests that either all or a constant fraction of the aerosol formed in
these experiments results from SO 2 photooxidation.
Runs 85 and 86 are another pair of SO 2 photooxidation experiments
at high and low SO 2 levels. These runs followed another series of experiments
involving hydrocarbons and NO. The presence of contaii ination is indicated
by the rather high apparent photooxidation rates, 0.13 and 0.14%/hr for these
runs. The last SO 2 experiment in this series, run 97, shows even more con-
tamination with an apparent photooxidation rate of 0.63%/hr.
It is evident from these results that the apparent SO 2 photooxida-
tion rate, based on measurements of aerosol formation in the SO 2 pure air
system, is quite dependent upon the experiments performed prior to making a
run. This is especially true in the U of M system. Previous exposure of
the bag to hydrocarbons and NO leads to enhanced aerosol formation. Apparently,
44
-------
reactive materials are either adsorbed or absorbed by the Teflon bag and
only slowly desorb when the bag is flushed with clean air.
In the Caispan chamber, there is somewhat less of a conditioning
effect on SO 2 oxidation and also less influence from the history of previous
experiments. It is possible that this is due to the very low surface-to-
volume ratio of the Caispan chamber and the fact that the walls are less of
an influence on the behavior of aerosol and gaseous species.
4.4.2 Hydrocarbon Experiments
Several types of hydrocarbon experiments were performed at the work-
shop and later at the University of Minnesota. At Caispan the three main sys-
tems studied were: I- IC + NO, HC + SO 2 , and HC + NO + SO 2 . Two basic types of
experiments were performed at the U of M: HC + NO and HC + NO + 502. Initial
reactant concentrations in each case were set as closely as possible to 0.35
ppm (volume) HC, 0.15 ppm NO, and 0.05 ppm SO 2 . Again, the main experimental
conditions and results are listed in Tables IX through XII.
Plots of aerosol number, surface and volume concentrations against
time and NO, NO 2 , 03. and hydrocarbon concentrations against time are given
in Appendix B. The plots are grouped together according to hydrocarbon type
and are in the same order as the runs listed in Tables IX through XI. The
chemical reactivity of these systems is characterized by listing the time to
the NO 2 peak and the maximum ozone concentration produced. In the U of M
tests, considerable lack of reproducibility existed in the ozone measurements,
apparently as a result of day-to-day variations in the sensitivity of the
ozone instrument. Some run-to-run variation is also evident in the NO-NO 2
results; however, this variation is believed to be real because both NO-NO 2
analyzers were stable and readily calibrated. Consequently, the main measure
of chemical reactivity of both the hydrocarbon + NO and the hydrocarbon + NO +
SO 2 will be taken as the time to maximum NO 2 concentration, t [ NO2]max• A
shorter time implies a more reactive system.
45
-------
In addition to these conventional measures, HC reactivity can also
be described in terms of aerosol behavior. The parameters used are: maximum
number concentration, maximum surface concentration, and two rates of volume
production. These parameters were chosen for the following reasons:
(1) Some description of number concentration was considered necessary
because CNC data are easily obtained and other investigators have presented
such data for various photochemical systems.
(2) Maximum surface concentration, SM, is listed because many photo-
chemical systems tend to establish an equilibrium surface concentration (Clark
Whitby, 1975) which is related to the rate of aerosol volume (or mass) production.
(3) The volume of aerosol formed multiplied by its density gives
the mass of aerosol formed. The density of these particles should not be
very different from unity. (Most organic acids and H 2 S0 4 lie in the range of
0.8 to 1.2 gm/cm 3 .) Consequently, the aerosol volume is nearly proportional
to aerosol mass. Since mass is a conserved quantity, the rate of mass forma-
tion in the aerosol phase must equal the rate of mass removal from the gas
phase. The rate of volume production is, therefore, directly proportional to
the rate of condensation of slightly volatile species formed by photochemical
reactions. Volume against time plots for these experiments have a rather com-
plex shape. Consequently, two volumetric production rates are defined and tabu-
lated in Tables IX and X, namely (dv/dt)S0 2 , the slope of the essentially linear
volume against time curve which is established early in HC + NO + SO 2 or HC +
SO 2 experiments and (dV/dt)maxj the maximum rate of volume production.
4.4.3 Hydrocarbon + NO Experiments
The HC + NO systems were found to behave quite differently from the
simple SO 2 + clean air experiments. Each hydrocarbon + NO experiment can be
divided into two phases. In the first phase, NO is converted to NO 2 , and
some oxidation of the hydrocarbon occurs. 03 starts to appear near the end
46
-------
of this phase as NO concentrations become very low. The second phase begins
as soon as the initial NO has been oxidized out of the system and NO 2 reaches
its maximum concentration. During this phase [ NO] remains low, and [ NO 2 ]
gradually decreases as NO 2 is converted to higher oxides, acids, peroxyacyl
nitrates, and other nitrogen compounds; ozone grows rapidly and approaches
a maximum; and aerosol formation takes place.
Typical examples of this two-phase behavior is shown in Figures 6
and 7 in which Caispan and U of N hexene-1 + NO experiments are compared.
it may be seen from the data that the NO disappears and NO 2 maximizes in
about 420 minutes in the Caispan experiment and about 280 minutes in the
U of N test. These times are considered as the duration of the first phase
for each case. The plots also show that by the end of this phase same ozone
has started to appear and the hydrocarbon concentration has started to decrease
rapidly. The aerosol data show no measurable aerosol production during this
phase in the U of N experiment, but some particles in the Calspan experiment
can be noted. The particles are so small, however, that the EAA does not
*
detect their presence.
Early in the second phase, growth proceeds rapidly. The same physi-
cal mechanisms control aerosol formation in this system as in the 502 + clean
air system, i.e., nucleation, condensation, and coagulation. In the Calspan
experiment, note the sharp inflection at this point as the concentration
increases and in the corresponding U of N test the sudden appearance of aero-
sol. In both experiments, there is a corresponding increase in the surface
and volume concentration of aerosol. For all of the hydrocarbon + NO systems
studied, the number of particles formed was less then in the SO 2 + clean air
case, but in every case (except for hexene + NO) the volume was always
This is a particularly useful illustration in that it shows the various
detection limits of CN counters used in these tests. For this experiment
the Environment-One CN counter did not detect particles early in the first
phase of the Calspan test, even though the more sensitive Gardner small
particle detector measured the concentrations shown. This situation was
found to occur frequently in other experiments.
47
-------
CALSPAN
RUN NO. 5 22 FEBRUARY 1974 [ HEXENE.1 - NO.FILTERED AIR SYSTEMJ R.H. = 41%; HEXENE-1 = 0.33 ppm;
NO = 0.152 ppm; NO 2 = 0.014 ppm
12 -
x10 4 x10 2
10
8
6
4
2
0
C d )
E
.2
N
E
LU
C -,
U-
C / )
E
Cd)
0
+
N
0
2
0
2
16 E
-2
E
uJ
-I
0
>
E
0.
0.
I-
LU
z
0.1
LU
I
E
U
LU
2
TIME (mm)
0
0 120 240 360 480 600 720 840 960 1080 1200
TIME (mm)
FIGURE 6
48
-------
RUN 92 0 TE r—SEPT-?4
HLIM (P RT. . ML)
1
10
‘1
H -S
LI . 1J 1
: -
6 F
E :
F -
tP :E
4—
2
0—
I
I
0
G
E
U
icr 1
4—
:o
—z
:o
• ti
1—
0
0
E
S
0
SYSTEM HE Et1E—1 HO
SLIPF . (IJM 2 ’ML)
FIGURE 7
HE> ENE—1 = 0 35 PPM
NO = 8.12 PPM
U of M.
0 UC1L (IJM 3 ’ML)
U
0
L
U
2 r•i
E
49
-------
larger, implying the production of fewer but larger particles. Another
important difference between the aerosol behavior of the hydrocarbon + NO
systems studied and that of the SO 2 + clean air systems is the shape of the
volume against time curves. They are no longer linear. Volume grows rapidly
early in the second phase; later the rate decreases and eventually volume
becomes essentially constant or even drops as the largest particles are lost
from sedimentation and fall out. In these cases, the species driving aerosol
growth becomes depleted and, consequently, new aerosol volume production stops.
The chemical behavior of the system in the second growth phase is
also shown in Figures 6 and 7. Production of ozone proceeds rapidly and
ozone concentration maximizes at about the same time that the aerosol volume
curve reaches its plateau; NO 2 decreases continuously while hexene is oxidized
out of the system more rapidly as the ozone concentration increases. In the
Caispan experiment, the ozone monitor was set for 0.2 ppm full scale so that
the ozone max was not observed. Another nearly identical case, however,
(run no. 21) produced an ozone max of 0.275 ppm.
The aerosol and chemical behavior shown in Figures 6 and 7 is quite
typical of all HC + NO experiments performed. The most pronounced difference
between the four hydrocarbons studied was in the rates of NO oxidation and
the rates of aerosol formation. Each of the systems are individually charac-
terized below.
In terms of NO oxidation and aerosol formation rates, toluene was
found to be the least reactive hydrocarbon in the Caispan chamber, followed
closely by hexene. This is, perhaps, the most significant difference between
the sets of data generated by Caispan and the U of M. In the Minnesota
studies, 1-hexene proved to be less reactive than toluene. For example,
in the l-hexene + NO experiment shown in Figure 7, the maximum aerosol volu-
3 -1 -l
metric production rate (dv/dt)axwas only 0.09 i.im icc . hr and the maximum
aerosol surface area SM was 3 iiin 2 /cc , both lower than any other hydrocarbon
system investigated. By contrast, in the toluene + NO case (run #76), a
50
-------
volumetric rate of (dV/dt)max 24.5 pm 3 /cc-hi’was measured with a maximum
surface area of 550 pm 2 / cc . The effect of relative humidity on aerosol
behavior in the toluene + NO system can be seen by comparing runs 76 and 87
in Table X. In run 87, the relative humidity was lower and aerosol was
formed less rapidly with (dV/dt)max= 8.6 jjm 3 /cc-hf and = 340 jim 2 ! cc.
The change in relative humidity produces no change in the observed chemical
behavior of the system, however. The large difference in aerosol production
would then seem to be due solely to the formation of particles having a
higher water content. Evidently, rather hygroscopic products were being
formed even in the toluene + NO case, since the relative humidity was only
50%.
The m-xylene + NO data generated by Calspan and the U of M compare
very favorably. Both aerosol and chemical measurements showed the m-xylene
+ NO systems to be considerably more reactive than either 1-hexene + NO or
toluene + NO. Times for NO 2 max were only 80 and 68 minutes for the U of M
experiments and 100 minutes for the Caispan experiment (see Tables XI and XII).
Aerosol and chemistry data for these experiments are shown in Figures 8 and 9.
The data show similar results: No aerosol is produced until approximately
the time of NO disappearance and rapid formation of ozone. The U of M data
show a slight (10 mm) lag between disappearance of NO and the onset of
aerosol formation. The volume and surface concentration for these tests is
substantially higher than for the toluene + NO and hexene + NO cases. An
expected humidity effect was noted in the U of M experiments; run 81 @ 75%
RI-I produced about twice as much aerosol volume as run 89 performed at 26%
RI-I.
Three cyclohexene + NO runs were performed at the U of M and two
at Caispan during the March workshop. The data compare very well showing
cyclohexene to have about the same reactivity as m-xylene in terms of NO
oxidation rate but much greater reactivity than any of the other hydrocarbons
tested in terms of aerosol production. In the cyclohexene cases, the aerosol
growth was almost explosive once oxidation of NO was complete. Both the
51
-------
8
4
E 0.2
0.
0.
0
0
2
c .1
0
2 0.1
0
FIGURE 8
CALSPAN _________________________
RUN NO. 15 MARCH 4. 1974 XYLENE.NO FILTERED AIR SYSTEM
m-XYLENE = 0.34 ppm; NO = 0.150 ppm; NO 2 = 0.014 ppm; RH = 38%
AEROSOL CONCENTRATION 25
20
U
c )
E
10 w
-I
0
5 >
0
480
0.3
E
0.
0.2
uJ
2
w
-J
E
0.1
0 120 240 360
TIME (mm)
480
x 102
12
10
10
8
E 6
U
IL l
z
2
0
E
--
c 1
E
ILl
C.,
4
U-
C d ,
0
0 120 240 360
TIME (mm)
52
-------
FIGURE 9
RUN Si O TE —JLIL—74
S?STEM:
I1—XYLENEMO
0 N
LIM C PART. ‘ML)
SURF C
IJM
-------
surface and volume concentrations of aerosol were much higher than any other
system, even though the number of particles was actually less. This implies
the presence of extremely large particles and, indeed, this is the case, as
in the Caispan experiments where nephelometer measurements were made, and
substantial visibility losses were noted after only 2.5 hours of irradiation.
Aerosol and chemistry plots from a Caispan and Li of M experiment are shown
in Figures 10 and 11. The volume production rates (dV/dt)maxOf 75 to 110
pm 3 ! cc—hr for the Caispan chamber and between 50 and 190 pm 3 / cc —hr 4
for the U of M are the highest we have measured and substantially higher
than that observed in typical urban polluted atmospheres. As well as oxi-
dizing NO quickly and leading to rapid aerosol formation, the cyclohexene
itself was quickly oxidized out of the system, as the data in Figures 10
and 11 show.
From the chemistry data for all the hydrocarbon experiments, a
family effect is apparent. Cyclohexene and hexene, both olefins, have
curves of basically the same shape: a gradual decrease with time until
after the NO is oxidized out of the system, followed by rapid decay as
ozone builds and aerosol forms. Toluene and m-xylene, both aromatics,
have a different shape. In both cases, hydrocarbon concentration decays
gradually and at a more or less constant rate. No change is evident in
the decay curve once ozone appears and aerosol formation begins.
4.4.4 Hydrocarbon + SO 2 Experiments
During the March workshop, four hydrocarbon + SO 2 experiments were
performed at Calspan. The purpose of these experiments was to examine SO 2
oxidation and subsequent aerosol behavior in the presence of hydrocarbons,
but without the addition of any NO beyond that normally found in clean coun-
try air (i.e., NO <0.01 ppm). As in the other experiments from the March
workshop, the chemistry and aerosol data for these tests are summarized in
Tables IX and X. Time histories of the data are shown in Appendix B.
54
-------
By-and-large the data show that the SO 2 oxidation rate for the first
hour or so is similar to that for SO 2 alone, followed by a period of acceler-
ated growth. The accelerated growth portion of the experiment is listed as
(dv/dt)maxin Table IX. Note from the table that, even in the 502 + clean air
experiments, an initial and final volumetric growth rate was observed; how-
ever, the difference in rates was not nearly as large as in the hydrocarbon-
polluted SO 2 systems. For example, the average initial rate of SO 2 oxidation
for the two Calspan experiments using a 0.05 ppm so2 + clean air (runs 2
and 3) was about 0.29% hr 4 with a final rate of approximately 0.79% hr 4 .
For the four hydrocarbon-polluted systems, the average initial 502 oxidation
rate was -0.35% hr 4 with a final rate, based on (dv/dt)max of about 1.6% hr 4
Since the shape of the aerosol curves for the HC + SO 2 systems is so similar
to the SO 2 + clean air systems, specific data plots are not presented here.
Very little chemical changes occurred since only background levels of NOx
were present and only modest levels of ozone were produced. The main differ-
ences were in the amount of aerosol produced by the hydrocarbon contaminant,
and this can be seen from the data in Table IX.
Toluene and hexene were found to have about the same effect on
accelerating apparent SO 2 oxidation rate, while cyclohexene was observed
to have a much greater impact on aerosol behavior. In the cyclohexene +
SO 2 experiment, run 13, volume production (dv/dt) 50 was 1.20 d l cc —hr 4
for the first 30 mm, followed by an increased rate of 10 pm / cc hr for
the remainder of the experiment (an additional two hours). In the toluene
and hexene + 502 runs, the final volumetric production rates were three to
four times greater than the initial rate. M-xylene, on the other hand, was
found to have the smallest effect with a final rate only about twice as
great as the initial growth stage. Since only one experiment of each type
could be run during the workshop, the magnitude of the increase cannot be
considered conclusive yet. It seems apparent, however, that the presence
of hydrocarbon contamination during SO 2 irradiations, even in the absence
of appreciable NO, substantially increases the overall production of
aerosol.
55
-------
25
)
I
FIGURE 10
Caispan
C1 1OHE) NE + NO SYSTEII
- RW # - -
AEI JS0LC0N ENTPATT0N
- - - “ - - - SURFAcE
6O
L d
100
5
LI
3
2
1
I
HENISTRY DATA
TIME (MIN)
56
-------
FIGURE 11
RUN 95 DATE 10—SEPT—74 S’ISTEM CYCL0HE EUE,l1O U of M.
D ttIM. (PART. ‘ML) SURF (IJM 2 ’ML) 0 UOL ((JH ML)
CYCLOHEXEHE = 0.35 PPM
X l 0. 124 PPM
-6
u -s
U _U
N F:
B F
E - i
C
E
1—
Xi 0
4—
H :0
I —2
i :0
R -N
07 ..E
E :
N —
0-
-
12
o
E -
s
1—
U
0
J IJ
ri
E
TIME (10+2 MINUTE)
H
V
D
R
,0
,JC
A
R
B
0
N
TIME j +2 MINUTE)
57
-------
4.4.5 Hydrocarbon + NO + SO 2 Experiments
An additional degree of complexity is introduced when SO 2 is added
to the HC + NO system. In terms of the chemical species which we monitor,
NO, NO 2 , O3 and hydrocarbon, the experiments with and without SO 2 are vir-
tually identical. The aerosol behavior, however, is quite different when
SO 2 is added. During the first phase of the experiment, aerosol growth
occurs in a manner very similar to that of a simple SO 2 + clean air experiment;
i.e., an essentially linear curve of volume against time. As the second phase
of the experiment begins, the rate of volume production increases markedly.
During this phase the growth curves are more like the hydrocarbon + NO experi-
ments, except that the aerosol number concentrations are higher. In terms
of qualitative behavior, a hydrocarbon + NO + SO 2 experiment behaves almost
as though it resulted from a linear combination of the hydrocarbon + NO sys-
tem and the SO 2 + clean air system. Quantitatively, however, this is not the
case. In both growth phases, more aerosol is usually formed than would be
predicted by a simple linear combination of the hydrocarbon + NO system and
the SO 2 + clean air system. The details of these interactions are discussed
more fully below as each hydrocarbon system is treated individually.
As previously stated, the greatest differences between the Caispan
and U of M data were found in the toluene systems. At the U of M, toluene
was observed to be substantially more reactive (in terms of aerosol behavior)
than hexene but less than m-xylene or cyclohexene. At Calspan toluene was the
least reactive, followed by hexene, m-xylene and cyclohexene.
For the toluene ÷ NO + SO 2 system, Caispan experiment #29, very
little effect was observed over that produced by toluene + NO alone. There
was, of course, initial aerosol growth soon after the start of irradiation
due to SO 2 oxidation and the formation of H 2 S0 4 particles. However, once
oxidation of NO was complete, more than six hours later, only a slight
increase in aerosol production was detected. Chemical and aerosol data for
this experiment are shown in Appendix B and compared with the U of M experi-
ments.
58
-------
At the U of M the initial phase of aerosol growth in the toluene +
NO + SO 2 system was very similar to Caispan’s; i.e., before the oxidation of
NO was complete, aerosol growth was like that of a SO 2 + clean air system,
but higher. Apparent SO 2 oxidation rates for the first 70-100 minutes
ranged from 0.23 to 0.67% hr at the U of M, compared to a rate of 0.32%
hr for the first 240 minutes of the Calspan experiment. Compared to all
of the other computed SO 2 oxidation rates for the U of M tests the toluene
+ NO + SO 2 system was the highest. This is rather surprising because toluene
is not normally believed to be a highly reactive hydrocarbon.
During the second stage of aerosol growth, appreciably greater volu-
metric production rates, (dv/dt) axwere observed in the U of M tests, compared
to the corresponding Caispan experiment. The values of (dV/dt)max fl U of M
runs 77 and 88 were 27.4 and 16.6 pm 3 / cc-hr , respectively, compared to
3 -l
about 1.5 pm /cc -hr in the Calspan experiment. The rates in both cases
are similar to those observed in the corresponding hexene + NO + SO 2 experi-
ments.
In comparing the toluene + NO + SO 2 experiments with toluene ÷ NO,
it can be seen that the nature of aerosol growth is quite different in the
second phase. Especially in the U of M cases (e.g., runs 76 and 77 and 87
and 88), the experiments reveal that, although aerosol volume production in
the second growth phase is similar with and without the addition of SO 2 ,
much larger surface and number concentrations are produced when SO 2 is present.
Thus, the particle diameters must be much smaller. (This effect is discussed
in more detail in the next section.) Similarly in the Caispan experiment,
the number concentration in the SO 2 case is more than an order of magnitude
greater than in the toluene + NO case. The surface, however, is not any
greater, but this is deceiving since the surface concentration was still
increasing at the end of the experiment (No. 29).
Run 65, the only toluene experiment performed in the small bag at
the U of M, shows much slower aerosol and chemistry progress than the other
59
-------
toluene + NO + SO 2 experiments. This probably results from two factors:
lower mean light intensity in the small bag and the inadvertent high NO
concentration (0.30 ppm) used in this experiment. Bag loss effects could
also be involved, but it is impossible to separate the factors leading to
the observed behavior, and the two reasons given above are believed to be
the most likely explanations for the decreased reactivity of the system.
Three hexene + NO + SO 2 experiments were run at the U of M (runs
60, 78, and 93) and one at Caispan during the March workshop (run 18). The
addition of SO 2 was found to appreciably increase aerosol formation in both
the Caispan and U of M tests but, as expected, had little effect on the chemi-
cal changes. Aerosol and chemistry data for the two sets of data are compared
in Figures 12 and 13. Two distinct aerosol growth phases are apparent.
During the initial phase, while NO is being converted to NO 2 , there is a
rise in number concentration due to SO 2 oxidation and H 2 S0 4 aerosol formation,
but the surface and volume production is quite low and proceeds almost as
if only SO 2 were present. For the Caispan and U of M experiments, the appar-
ent SO 2 oxidation rates during this period were 0.16 and 0.18% hr , respec-
tively. This is higher than that typically observed in the U of M tests for
SO 2 alone, but about the same as the initial rate of 502 oxidation in other
Caispan experiments with SO 2 + filtered air. The slightly increased rate
of aerosol volume production in the U of M case possibly results from an
acceleration of the SO 2 photooxidation process by reactive intermediates
formed by the NO to NO 2 conversion and 1-hexene photooxidation. However, in
the absence of data on aerosol chemical composition, the mechanisms for
increased aerosol production cannot be verified.
During the second phase of growth, at about the same time that
ozone begins to appear, the rate of aerosol production again increases.
Number concentration begins to grow and the rates of surface and volume
production are greatly enhanced. The shapes of the surface and volume con-
centration curves in this growth phase are quite similar to those obtained
in the pure 1-hexene + NO system, but the actual concentrations are higher
(see Tables IX and X for comparisons). The increases, especially in the
60
-------
L I of M data, are larger than would be expected from a simple linear com-
bination of the aerosol formed by the SO 2 pure air system (for comparable
SO 2 concentrations) and that from the hexene + NO system. Synergistic inter-
actions must take place which lead to enhanced aerosol formation in the com-
bined system. Again, chemical composition data would be particularly useful
in helping to explain these results. In the absence of thesedata, some
possible explanation may be:
I. The sulfuric acid aerosol formed during the first growth phase
might act to catalyze the formation of aerosol from l-hexene + NO reaction
products.
2. Gas phase products of SO 2 photooxidation might act to accelerate
the formation of aerosol from l-hexene + NO reaction products.
3. Gas phase 1-hexene + NO products might act to greatly increase
the rate of SO 2 oxidation and thus aerosol formation.
Schemes I and 2 both depend upon the interaction of SO 2 photooxida-
tion products with the 1-hexene + NO system. The interaction could lead to
an acceleration of the rate of formation of whatever hydrocarbon-related
species àondensed in the second growth phase or to an alteration of the gas
phase reaction paths leading to the formation of a nonvolatile, more readily
condensible species. A combination of these two effects could also occur.
Only a tiny fraction of the 1-hexene which is oxidized would have to be con-
verted to nonvolatile species to produce a significant aerosol yield. In the
first 500 mm of U of M experiment 92, about .24 ppm or 890 pg/rn 3 of 1-hexene
disappeared from the gas phase, during the same period about 0.18 pm 3 ! CC, or
assuming an aerosol density of unity, 9.18 pg/m 3 of aerosol is formed. This
represents only about 0.02% of the mass of l-hexene removed.
61
-------
CALSPAN FIGURE 12
RUN NO.18 MARCH 6, 1974 HEXENE-1-S0 2 -NO SYSTEM
HEXENE .1 = 0.33 ppm; NO = 0.178 ppm; NO 2 = 0.008 ppm; SO 2 = 0.07 ppm
RH = 37%
HEXENE-1-S0 2 -NO .FILTE RED AIR SYSTEM
— 25
x 102
20
15
C ,)
E
w
C-,
U-
x 1o 4
C ’)
E
U
U i
2
0
10
5
0
0 120 240 360 480
TIME (miii)
600
C’)
E
U
E
a
a
U i
z
Ui
x
Ui
I
E
a.
a.
C v )
0
0
N
0
12
8
4
0
0.3
0.2
0.1
600
0.2
0.1
0
0 120 240 360 480
TIME (miri)
62
-------
FIGURE 13
SYSTEMS HD(ENE— 1 , SO2.HO
a SURF.(lJM 2 ’ML) 0 UOL.q3M 3 ’ML
U. of M.
NO = 0.122 PPM
V
0
L
U
2 M
E
1
0
4
H
V
0
R
20
A
R
B
0
H
2
RUN 93 OATE 8-SEPT-74
fl NUM.(PART.’ML)
HEXENE—1 =
- s
— u
P
-A
- c
E
1 —
11
U
(‘1
B
E
P
H
I
I
P
0
C
E
H
0
I ,
1
12
C
E
S
0
2
: 0
— z
:o
1
—1
HO (PPM)
NO 2 (PPM)
03 (PPM)
HC (PPM)
1
U
I
TIME (
4
MINUTE)
U
63
-------
In run 93 with SO 2 present, the aerosol yield is raised to about
3.6% of the mass dfT-hexerie i emoved. Such áchange could.easily àccur as
a result of either scheme 1 or 2. If scheme 3 is important, it should be
easy to verify experimentally, because in order to explain the observed
aerosol formation, a significant fraction, i.e., more than 20% of the SO 2
initially present would have to be removed from the gas phase. This quantity
could be detected by gas phase sulfur monitoring. Analysis of aeroscl sulfate
content would provide more definitive results.
Before concluding the discussion of the 1-hexene system, two other
experimental variables I ust be mentioned: humidity and bag size. The effect
of humidity may be seen by comparing lJ of M experiments 78 and 93. oth of
these experiment were for the 1-hexene ÷ NO + SO 2 system; however, 78 was
at 55% relative humidity and 93 at 32%. As in the case of the SO 2 + clean
air system, aerosol formation is enhanGed by increased humidity. In the
first phase of growth, the volume production rates are .53 and .29 pm / cc -
hr for the high and low humidi,ty cases, respectively. In the second growth
phase, the effect is less apparent but still present with volume producti,on
rates of 19.1 and 18 pm 3 ! cc-hr for the high and low humidit.y cases, respec-
tively. The difference between the two cases during the second grow th.phase
is more apparent when, the maximum ,aerosol surface areas of 1530 and 1300
CC for the high and low humidity cases are considered.
Run 60 was a 1-hexene + NO + SO 2 experiment done in the small bag.
The qualitative behavior of the system is exactly like the large bag with
two-phase growth observed. Quantitatively, however, evertyhing seems to’take
place more slowly in the small bag; the ti nes to NO 2 peak, to’ the onset of
ozone formation, and to the onset of rapid aerosol growth are longer. As
previously suggested, the slowing effect on the chemistry in this and other
systems in the small bag may be d e tolower average ,light intensity,, since
the bag is mounted in the center of the cylindrical illumination housing
where the light intensity is lowest. It is not yet clear whether wall effects
also play a role in the observed results.
64
-------
The m-xylene + NO + SO 2 systems investigated at Calspan and the
Ii of M were very reactive in terms of chemical and aerosol behavior.
The times to reach [ NO2ImaX were shorter in every case than the analogous
toluene or hexene experiments. As is the case of the other hydrocarbons,
two phases of aerosol growth are evident. Comparisons between a Calspan
and U of M test are shown in Figures 14 and 15 . For the U of M test,
the SO 2 photooxidation rates during the initial growth phase were 0.14 and
0.21% hr . The Caispan experiment produced a slightly higher rate of 0.32%
hr . These rates are about the same as for the other hydrocarbons and also
for a slightly contaminated SO 2 + clean air system but lower than the Ii of M
toluene + NO + SO 2 cases. In both the Calspan and U of M tests, m-xylene
appeared to interact only weakly during the initial phase of aerosol growth.
By contrast, during the second growth phase, very rapid aerosol
production was observed as the data in Figures 14 and 15 show. The syner-
gistic interaction of the hydrocarbon + NO and SO 2 ÷ clean air system is
more obvious in the Caispan experiments. Here the addition of SO 2 to the
m-xylene + NO produces a great deal more aerosol surface and substantial
additional volume. For the corresponding U of M experiment, the large
increase in surface area takes place at essentially the same volume concen-
tration. The excess surface area produced for the measured volume increase
implies the presence of many small particles (compared to m-xylene + NO)
and, indeed, this is the case as the data in other HC + NO vs. HC + NO + SO 2
systems show. Smaller particles are present with the added SO 2 , since the
SO 2 generates a large concentration of tiny nuclei early in the experiment.
These particles then serve as sites for subsequent condensation in the
second growth phase.
A final feature of the m-xylene + NO and m-xylene + NO + SO 2 experi-
ments investigated is the relative humidity effect. U of M runs 81 and 82
were run at high humidity and runs 89 and 91 at low humidity. As usual, the
change in humidity leads to no change in the chemical data but produces
significant changes in the aerosol data. Both surface and volume concentration
65
-------
FIGURE 14
Caispan M-XYLEUE + NO + SO 2 SYSTEM
F JN#1L4
AEF JSOL CONCENTRATION
-----H------ --*4 - - - --H--- - ---- i -- -- - ------1-------
:
L
“O’UME
I
-I
I
I • ---1-- - --
-
10 i---
1 frI SURFACE ‘ TIO
__ 1
QEMISTRY L \TA
I IO 3
F -o -
‘ *i __
-- L. M-XYLENE
O 2
- t - --- - - - -----H
NO 2
--±--- - --- -i Os1
F
NO----
H : I
0
TI ’E (MIN)
60
120
180
2 O
300
360
L12 0
0
a-
5J
30
20
100
3Q1 ’r\
25
2O
I” _
CD
><
0 L
(N
0
0 3
O 2
0.1
0
66
-------
FIGURE 15
RUN 91 Oi TE
6—SEPT-’4
SYSTEM:
M—YLENE $02.. tb
HUM
PART. ‘ML)
A
SURF
(PM 2 ’lIL)
0
VOL . ( JM 3 ’IiL)
c ” . 5
A £ —
3—
- LI
-R
1 - 4
-c
1—
0—
xig-l
4—
N 0
I —z
r :o
R -N
G -
E
N-
1—
0
12
0
E
0
U. of M.
M-XVLENE 0.35 PPM
$02 = 8 046 PPM NO = 0 117 PPM
5
H
LI
f1
R
4 U
0
L
U
M
H
Y
D
R
C
A
R
B
0
t,1
TIME uo 2 MINUTE)
67
-------
markedly increase with relative humidity, implying deliquescence of the
hygroscopic fraction of the aerosol.
One cyclohexene + NO + SO 2 system was investigated at Caispan (run
9) and one at the U of M (run 96). The Caispan and U of M data for this system
are very similar and are shown in Figures 16 and 17 . The addition of SO 2
to the cyclohexene + NO system produced effects qualitatively similar to
those produced by addition of SO 2 to the other hydrocarbon + NO systems
studied. Thus, the NO, NO 2 , 031 and hydrocarbon profiles were essentially
unchanged by SO 2 addition. At the same time, aerosol behavior was markedly
changed, with two-phase aerosol growth resulting.
The changes in aerosol behavior produced by SO 2 addition to this
system were dramatic and informative. During the first phase of growth,
aerosol pr6duction occurred as though a slightly contaminated SO 2 + clean
air system was being irradiated. The apparent 502 photooxidation rate was
0.20%/hr in the U of M case and 0.28%/hr in the Calspan experiment, about
the same as that observed for other hydrocarbon + NO + SO 2 systems. With
this system, evidence of the second phase of aerosol growth appears before
NO has been completely oxidized out of the system. Especially in the U of
M case, the volume against time data becomes nonlinear with marked upward
curvature appearing as early as 50 minutes after lights on. However, in
both cases, rapid aerosol volume production does not occur until about the
time of complete NO oxidation and rapid formation of ozone.
In terms of volume production, the second phase growth is quite
similar to the cyclohexene + NO system in the absence of SO 2 . The differences
that are observed are not considered significant because of the difficulties
associated in measuring (dv/dt)maxfor the cyclohexene + NO system, i.e.,
rapid growth beyond the upper particle size limit of the EAA. The change
in surface and number concentrations associated with the addition of SO 2 ,
however, are very significant. Both number and maximum surface concentration
are much larger. However, since the total aerosol volumes produced in all
68
-------
cyclohexene cases are about the same, the addition of SO 2 must lead to a
dramatic decrease in particle size.
The effect of SO 2 on particle size can be seen in Figure 18 , in
which mean surface diameter has been plotted as a function of time for
several HC + NO and HC ÷ NO + SO 2 systems studied at the U of M. Data for
a typical SO 2 irradiation are also shown in the figure. The almost explosive
growth in the cyclohexene + NO case occurs at about the time of complete oxi-
dation of NO and appearance of appreciable ozone. In the case shown, particle
diameters of nearly 0.5 iim were produced in less than an hour from the time
of initial nucleation. With SO 2 in the system, the particle sizes are much
smaller, tinder these conditions, initial nuclei formation results from SO 2
oxidation followed by additional condensation on existing particles during
the second phase of growth. Although the particles are smaller, the number
concentration is higher.
m-Xylene + NO behaves like the cyclohexene + NO system but to a
lesser extent. The addition of SO 2 results in the same effect; initial
particle formation and growth at a rate similar to that for SO 2 alone,
followed by accelerated growth during the second phase.
By comparison, the much less reactive hexene + NO + SO 2 system
was observed to essentially follow the SO 2 particle growth curve for the
duration of the experiment.
The implications of these data relative to the production of light
scattering aerosol can be seen in Figure 19. In the figure the light scat-
tering function b(SCAT) is plotted as a function of time for several HC + NO,
HC + NO SO 2 systems in the Calspan chamber. The production of very large
particles in the cyclohexene + NO case results in the rapid formation of
light scattering aerosol. Although the particle diameters are much larger
than in the comparable cyclohexene + NO + SO 2 experiment, b(SCAT) is nearly
the same because fewer particles are produced. The addition of SO 2 results
69
-------
FIGURE 16
CALSPAN
RUN NO. 9 27 FEBRUARY 1974 I cYcLoHExENE4JOSO 2 FILTERED AIR SYSTEM ] R.H. = 36%;
CYCLOHEXENE = 0.33 ppm; NO = 0.220 ppm; NO 2 = 0.020 ppm; SO 2 = 0.05 ppm
x10 5 x10 2
2
E
I ,
E E
. 9
—
uJ uJ
D
a 0
‘ p
E a
a a
-
uJ
a
U i
S
U I
a i
A 0
0 -J
a C)
>.
C-)
C I ,
E
U
( 1
E
U I
U
C
U-
D
U ,
600
TIME (m m)
0 120 240 360 480 600 720 840 960 1080 1200
TIME (m m)
70
-------
RUN 9G 0 TE il—SEPT—?
Q HUM (PART. ‘ML)
—5
• IJ
•R
-c
-E
1—
0
xl .1
4—
N
I —z
I :o
R N
0., _E
E
N
0
x :
12
o :
E -
s
1—
a
N
U
M
B
E 2
FIGURE 17
SYSTEM 4 CYCLOHEXENE 602 HO
SURF. (jJM 2 ’ML) 0 (JUL (IJM 3 /ML)
= 0.35 PPM
U. of M.
U
0
L
U
M
E
TIME (10+2 MINUTE)
H
V
0
R
. 70
- Ic
A
R
B
0
N
TIME jg+2 MINUTE)
71
-------
0.5
CYCL0HE) ENE ÷ NO
3L4
03
02
cYCLOHDKENE + NO + SO 2
rM-)CYLENE+No
M-XYLENE + NO + SO 2
0]
502
+ HE) NE + NO + SO 2
GO 20 30
hIVE (MIN)
FIGURE 18. MEAN SURFACE DIAMETER VS TIME FOR SEVERAL HC+ NO, HC + NO + SO 2 ,
AND SO 2 EXPERIMENTS.
72
-------
FIGURE 19.
LIGHT SCATTERING COEFFICIENT (BSCAT) OF PHOTOCHEMICAL AEF 3S0LS ‘ TIl’E .
5.0
3.0-
t I
L
r
-L
4 1 i
2.0-
)
so 2 -
I-
1.5
1,0
0.7
0.5
1O
‘15
20
•50
•70
90
T I TI
H : :‘
:t: : r : :! 1
! ::: i: .
T r
1 t. I
: t
: 1=t E
tr
::::::i::j::::: ::: I :
: : ::: ::1::::::::::
Et E
0
1
2 3 4
5
6 7 8 9 10
TINE (HRs)
I
1112
73
-------
in smaller but more numerous particles whose effect on light scattering is
about the same.
The addition of SO 2 to the m-xylene + NO case can also be seen
to produce substantial light scattering aerosol during the second phase of
aerosol growth. For several hours there is no effect, but once NO oxidation
is complete and second stage growth begins, large visibility losses are
observed. The fact that the concentration of particles is much smaller in
the m-xylene + NO case (compared to m-xylene + NO + SO 2 ) accounts for the
lack of appreciable light scattering aerosol.
By contrast, the hexene system, even in the presence of SO , did
not produce significant light scattering aerosol or visibility losses. The
same result was obtained in the Calspan tests for the toluene + NO + SO 2
system. The somewhat accelerated particle growth observed in the U of M
tests may have resulted in light scattering aerosol; however, there was no
opportunity to make this measurement in their chamber. Finally, the SO 2
alone, while producing very large concentrations of particles, did not pro-
duce significant light scattering aerosol over the duration of these tests.
4.5 Conclusions from the March Workshop
From the data generated during the March workshop and the duplicate
experiments performed at the University of Minnesota, the following points
can be made:
(1) Each HC + NO experiment can be divided into two phases. In
the first phase, NO is converted to NO 2 and some oxidation of hydrocarbon
occurs. Ozone starts to appear near the end of this period. The second
phase, accompanied by substantial aerosol formation, begins as soon as NO
is completely oxidized and NO 2 reaches its maximum; ozone increases rapidly
and approaches a maximum.
74
-------
(2) Qualitatively the HC + SO 2 system behaves in a manner similar
to SO 2 alone. Quantitatively, the presence of hydrocarbon contamination, even
in the absence of appreciable NO, substantially increases the overall produc-
tion of aerosol. Average apparent SO 2 oxidation rates exceeding 1.6% hr
were observed for the final stage of growth in the HC-enriched atmospheres in
the Caispan experiments.
(3) The addition of SO 2 to the HC + NO system was generally found
to exert a synergistic effect on aerosol surface and volume formation. At
Calspan the effect on aerosol behavior was greatest for m-xylene, while at
the U of M the largest effect was observed in the hexene + NO + SO 2 system.
Possible synergistic effects in the cyclohexene system were masked by the
explosive growth of aerosol with and without the addition of SO 2 .
(4) The addition of SO 2 to the HC + NO system produces a dramatic
decrease in the mean particle diameter. This results from the initial forma-
tion of very high concentrations of nuclei in the presence of SO 2 . During
the second stage of aerosol growth, condensation proceeds on the existing
particles rather than forming fewer but larger particles typical of the
HC + NO system.
(5) Of the hydrocarbons studied, cyclohexene was the most reactive,
both in terms of aerosol and chemical behavior, followed by m-xylene, hexene,
and toluene. The main difference observed in the duplicate experiments at
the U of M was that hexene was the least reactive hydrocarbon.
(6) It has been possible from these tests to characterize system
reactivity in terms of aerosol behavior. The most important variables are
maximum number concentration, equilibrium surface concentration, and volumetric
growth rate. These aerosol measures of reactivity have been found to corre-
late well with other conventional parameters, such as time to [ N02}max and
[ 0 ]
3 max.
75
-------
(7) Aerosol formation rates were enhanced at high relative humidi-
ties, probably as a result of the higher water content in the aerosols.
(8) The data generated in these experiments in chambers of widely
different physical dimensions show a high degree of correlation. The main
difference in the results is indicated in the significantly greater aerosol
and chemical reactivity of toluene found in the U of M chamber. The reasons
for this difference have not been resolved.
76
-------
Section 5
CHAMBER CHARACTERIZATION TESTS
During the November and March workshops, a number of experiments
were performed to characterize chamber performance and test for contamination.
The photooxidation of SO 2 in clean air was one common test of chamber condi-
tion that has already been described. In addition to these experiments, how-
ever, were a number of dark reaction tests at the University of Minnesota
and also NO photolysis experiments and aerosol coagulation tests performed
both at Calspan and the U of M. Although the importance of chamber testing
is recognized, there has not yet been an established format for intercomparing
chamber performance. In retrospect, it may well be that the similarity in
test results at Calspan and the U of M during the past year provides the
best indicator of comparable chamber performance.
5.1 Dark Reaction Tests -- University of Minnesota
After the November workshop, a number of aerosol experiments were
performed to determine the degree of chamber contamination in the U of M
chamber. The experiments are summarized in Table XV. Runs 21-24 were SO 2
photooxidation and decay experiments. Note that the peak CNC is lower for
run 22 than for run 21 and lower for run 24 than for run 23. This suggests
that gradual conditioning of the smog chamber is occurring as a result of
performing 502 photooxidation experiments. The same type of conditioning
effect was observed in the November workshop after several SO 2 photooxidation
experiments were performed in a given bag.
Conditioning effects show up in another way in runs 21-24. Figures 20
and 21 show plots of NC and 1/CNC against time for these four experiments. It
is assumed that nucleation of new particles ceases after the lights are turned
off, plots of l/CNC against time during this decay period should be essentially
straight lines with a slope equal to the coagulation coefficient. Upward
77
-------
TABLE XV. AEROSOL REACTIVITY EXPERIMENTS
SO 2 Relative Time N Time N Coagulation
Run Conc. Humidity Dark max dark Light max light Rate Constant
No. (ppm) (mm) (part/mi) (mm) (part/mi) (x 1010 ml/sec) Comments
21 0.49 25 180* 30 241K 21
22 0.53 29 180* --- 30 112K 27
23 0.54 29 180* --- 15 90K 26
24 0.54 25 180* -—- 15 66K 21
27 0.54 52 97** Not measured -- -- --
28 0.54 55 42** Not measured -- -- -- New charcoal
30 0.59 55 300 52K -- -- -- New bag (LB-3)
31 0.59 50 240 53K 120 280K --
32 0.59 52 180 30K 70 330K --
33 0.59 55 11 --- 960 240K -- çLights on before
34 0.59 55 45 48K 945 220K -- Nmax dark obtained
35 0.59 48 100 30K 18 -- -- Very high peak CNC
36 b 41 --- -—- 820 8.1K --
37 b 52 --- --- 250 b --
38 b SO --- 265 b --
39 0.59 37 120 14K --- -- --
40 b 45 --- --- 328 b --
43 0.59 16 --- 6.2K --- -- --
*
Dark decay time after lights turned off.
**
Time to onset of aerosol growth.
-------
RUN 1 O TE 29JMN?4 SYSTEM 502 U of M
Q NUMBER CONC (PART ‘ML) A I ‘NUMBER CONC (ML’PAPT
RUN 22 DATE 31-..R1—?4
SYSTEM 502
FIGURE 20. TIME HISTORIES OF AEROSOL COAGULATION.
N
U
M
B
E
R
C
0
IN
C
U of M
N
U
N
B
E
P
C
0
N
C
xie-;
N
4U
N
B
E
R
3 C
0
N
C
2
a
TIME xir 2 (MINUTE)
79
-------
RUN 23 DATE 2-FEB-74 SYSTEM S02 U of M
0 NUM8ER CONC (PART ‘ML) 1 ‘NUMBER CONC (ML’PART )
SYSTEM $02
FIGURE 21. TIME HISTORIES OF AEROSOL COAGULATION.
11
U
I,
B
E
C
C
‘4
C
/
N
4U
M
B
E
P
.7
0
N
C
RUN 24 DATE 2-FEB—74
U. of M
TIME (MINUTE)
80
-------
curvature of this line suggests wall losses; downward curvature suggests
particle generation. The plot of 1/CNC against time presented in Figure 20a
is not a simple straight line but rather two straight line segments. The
slope of the line during the first 60 minutes after lights off is 2.1 x l0
cm 3 /sec, only slightly greater than the 1.5 x l0 cm 3 /sec predicted by
theory for a monodisperse system having the same mean particle size. The
theoretical monodisperse coagulation rate would be expected to be lower
than the observed coefficient for two reasons: (1) Polydisperse aerosols
should have higher coagulation rates than monodisperse aerosols of the
same mean size, and (2) losses to the wall of the chamber can only act to
increase the rate of particle disappearance and thus raise the apparent
coagulation coefficient. The difference between theory and experiment is
thus in the correct direction. For the second portion of the curve, however,
-9 3
this is not true. Here the slope is about 0.7 x 10 cm /sec which is sig-
nificantly less than predicted by theory. The only reasonable explanation
for such a low slope is the generation of new particles through a dark reac-
tion mechanism. Secondary particle production is, therefore, taken as evi-
dence of contamination. Run 22 also shows this type of behavior; however,
in this case the difference in slope between the apparently uncontaminated
and the contaminated decay period is less pronounced. In runs 23 and 24, Figure 21,
evidence of contamination has disappeared altogether, and only simple second
order aerosol decay is evident. The series of SO 2 experiments has, therefore,
led to a conditioning of the chamber and the disappearance of effects which
could only be attributed to contamination.
The coagulation coefficients measured in runs 21 through 24 varied
-9 3 . -g 3
from 2.1 to 2.7 x 10 cm/sec with an average value of 2.4 +0.3 x 10 cm /
sec.
Dark reaction experiments were also performed in order to measure
chamber contamination. In these experiments the bag was filled with clean
air and humidified to 50%. About 0.5 ppm of SO 2 was then added to the system.
The bag was left in the dark and the CNC was monitored. In experiments 27
81
-------
and 28 the time to onset of aerosol growth in the dark was measured to be
97 and 42 minutes, respectively. Dark growth in the SO 2 system is considered
an indication of chamber contamination. The short time to the onset of dark
growth in run 28 possibly indicates more chamber contamination. This seems
rather surprising because as is noted in Table XV, the charcoal in the air
purification scrubbers was changed before run 28.
The persistence of dark reactions, even after the charcoal was
changed, was taken as evidence of chamber contamination. A new large bag
(LB-3) was, therefore, constructed. Run 30 was the first dark reactivity
test performed in this bag. Aerosol growth began almost immediately after
SO 2 was introduced, and eventually a peak CNC of 52,000 particles/cm 3 was
attained. This significant dark growth was not unexpected, because past
experience has indicated that new Teflon bags require a conditioning period
to remove or deactivate contaminants on the bag surface.
Consequently, several experiments to condition the bag were per-
formed. In these experiments the normal procedure used in dark reactivity
tests were followed, and the aerosol was allowed to grow in the dark until
a maximum CNC, (N ) was obtained. The lights were then turned on and the
max
chamber was irradiated for up to 16 hours to condition the chamber. Runs
31-35 were of this type. Table XV shows the light and dark times and values
of N obtained. The steady reduction of N achieved both in the dark and
max max -
during irradiation suggests a conditioning effect. However, even after 5
runs, dark reactions were still very important. Several clean air irradia-
tions, runs 36-38, were then performed to further condition the bag. Run 36
showed slight aerosol growth with N reaching 8100 particles/cm 3 , whereas
runs 37 and 38 resulted in no aerosol formation. The aerosol produced in
run 36 probably resulted from residual so2 which remained attached to the
walls of the smog chamber after the previous SO 2 experiments.
82
-------
When SO 2 was again added in run 39, dark reaction occurred, although
N was only 14K. Run 40 was another clean air irradiation and showed no
aerosol formation. The last simple dark growth experiment was run 43; here
Nm was only 6200 particles/cm 3 , but this run was performed at a low
relative humidity of 16% due to a failure of the humidification system.
This fairly low value of N is at least partly due to the decrease in
relative humidity.
5.2 Coagulation Experiments -- Calspan
A good opportunity for studying Caispan chamber performance was
provided by data generated from a series of auto exhaust irradiations corn-
*
pleted on another EPA-sponsored program. Coagulation data were analyzed
from the tests to try to establish information about wall losses in the
large chamber. As previously stated, if wall losses are not important,
a plot of (where N is the particle concentration) vs time should give a
straight line of slope k. On the other hand, as wall losses begin to dominate,
the apparent value of k will increase as the particle concentration decreases.
Plots of particle decay for seven exhaust emissions tests are shown
in Figure 22. The data show a break in the 1/N vs time plot at 13 hours
in each case. The average k for the seven experiments was 2.65 x l0 cm 3 /
sec which is in good agreement with theory for a polydisperse aerosol in the
size range of 0.0075 pm to 0.133 pm. After 13 hours, there is upward curva-
ture in each of the auto exhaust coagulation tests, indicating that wall
losses or settling or both are beginning to influence the results. These
data suggest that working times of at least 13 hours are possible in the
large Caispan Chamber.
Longer working times are likely with photochemical aerosols
of smaller size since sedimentation is less of a factor. These tests,
however, have not yet been performed.
Contract No. 68-02-1231
83
-------
CALSPAN
ALL EXPERIMENTS
LIGHTS ON, NO STIRRING
r
•EXP# 1
24 --- (EXPT’S 11, 12,13 & 15) ARE NOT SHOWN BUT 4
)( )( 5 L EWITHINTHEEXTREMESPLOTTEDHERE
8
D—D 9 :
22 10 I I I i
20
•10
U -- -1
13 hours
16 j •- - -:
E IA I P
- I’. - -
E
I P :
12 :-
-iz
10 -j - -i -
8 I
6 k 1.59x10 7 min 1 ____j
r .
2---
0 2 4 6 8 10 12 14 16 18 20 22
TIME (hrs)
FIGURE 22. AEROSOL COAGULATION DATA - AUTO EMISSION TEST SERIES
84
-------
5.3 NO Photolysis Experiments -- Caispan and University of Minnesota
The measurement of NO photolysis rates has been suggested as a
means of measuring smog chamber contamination (Bufalini, 1972). Consequently,
a number of NO photolysis experiments were performed at Caispan and the
University of Minnesota in order to compare chamber performance.
The thermal oxidation of NO in clean air occurs by the overall
react ion
NO+NO+0 2 ÷2NO 2 . (7)
The rate of this reaction is proportional to [ NO] 2 . If contaminants
are present, irradiation of the NO + air system will result in the formation
of addition species which react with NO and produce much more rapid removal
of NO than predicted by reaction (7). Comparison of the NO removal rate,
calculated from kinetic data available for reaction (7) and the experimentally
determined removal rate, gives a measure of chamber contamination.
Table XVI summarizes the NO photolysis experiments performed in
the Calspan and Minnesota chambers. Here, initial NO and NO 2 concentrations 1
theoretical and experimental NO disappearance rates, ratio of experimental
to theoretical NO removal rates, and the percent NO loss per hour are pre-
sented. Before further discussion of these results, it should be noted that
NO photolysis experiments are more sensitive to contamination if the initial
NO concentration is low. This is because at high NO concentrations the rate
of reaction (7), which is second order in NO, is large enough to mask any
contamination effects. Another important factor lies in the fact that any
hydrocarbon contamination effect would be more pronounced at higher HC/NO ratios.
It is probable that reactions between nitrogen oxide and contaminants
present in the chaEber are first order in NO. The apparent first order dissoci-
ation rate, expressed in terms of percent NO removal per hour, provides a
useful measure for intercomparing experiments and is also shown in Table XVI•
85
-------
TABLE XVI. NO OXIDATION E) ERI [ VEf ffS
ALSPPN
ff NO NO 2 m o (A) E)TAL
(L.ARGE CHNV1BER)
DEC. 15, 1973 1 0.610 0.08 6.3xiCi 3 1 .5x10 2 2, i 9.3
F . 28, 197/4 II 0./471 0.07 3.7x10 1. t ixiO 2 6.1
M . 6, 197/4 19 0.150 0.01 3.8x1O l.9x10 2 12.6
MAR.15 , 1974 27 0.1/47 0.01 3.6x1O i.2x10 2 33 8.2
MAR. 16, 19714 28 0.1/45 0.050 3.6xlO 2.7x10 2 18.6
MAR. 21, 1974 31 0.498 0.08 4.2x10 1 .lxlo 2 2.6 2.2
UNIVERSITY OF MINNESCJFA
LARGE BAG //2 25 0.56 0.01 5 ,14x10 3 5.2x10 2 9.6 9.3
29 0.56 -- 5 .4x lO 3.4xlO 2 6.3 6,1
LARGE BAG #3 47 0.305 0.02 1.6x10 3 1.6x10 2 10 5.3
48 0.67 0.02 7.5x1(1 3 1 .7x1(V 2 2.2 2.5
S1v .u.. BAG #1 % 0.11 0.01 2,Ox10 4.lxlO 21 3.7
55 0.36 0.01 1 2.2x10 5 .6x10 2 356
56 0.575 0.012 5 .6x10 9.6x10 3 1.7 1.7
58 0.51 0.01 1 4.!4x10 7.6x10 3 1.7 1,5
(A) BUFALINI, J,J, AND STEPHENS, E.R., 1965: “THE THERIV1AL OXIDATION OF NITRIc OXIDE IN
ThE PRESENCE OF ULTRAVIOLET LIGHT”, INm. J. AIR WAT. POLL., VOL. 9, pp, 123-128.
86
-------
The initial Caispan NO photolysis experiment was conducted after
the November workshop using an initial NO concentration of 0.61 ppm. As
shown, the observed oxidation rate is only 2.4 times greater than theory,
suggesting that at higher concentrations of NO the presence of wall contami-
nants in the large chamber do not greatly affect the NO photolysis rate.
The experiments performed during the March workshop showed greater variation
depending on the NO concentration used. Generally the lower the concentra-
tion the larger the effect of contamination on test results. The very high
rate for run no. 28 is largely due to the initial conditions of the experi-
ment. In this run, a substantial amount of NO 2 was inadvertently introduced
into the chamber along with the NO resulting in accelerated disappearance of
the nitrogen oxide. At the completion of the March test series, another NO
photolysis experiment was performed using a higher concentration of NO
(0.498 ppm). As before, the presence of wall contamination (based on the
history of previous experiments without cleaning of the chamber walls) did
not appreciably affect the oxidation rate at these higher concentrations of
NO.
In the U of M tests, the first two NO photolysis experiments, runs
25 and 29, were performed in LB-2 after a long series of photochemical aero-
sol experiments. The main difference is in the 30-40% lower oxidation rate
in run 29 compared to run 25. The reduced rate probably resulted from a
reduction of background contamination when the charcoal in the scrubbers
was changed between runs 28 and 29. As was noted above, however, this new
charcoal actually led to more rapid formation of an aerosol by dark reactions.
Runs 47 and 48 were done in large bag 3 after it had been subjected
to a long series of conditioning experiments, including two SO 2 photooxidation
experiments performed as part of the duplicate March workshop. Both of these
experiments showed dark growth and, thus, contamination was indicated. Run
47 showed NO oxidation rates well above theoretical, but run 48 was only
2.2 times as great. Thus, the larger concentration of NO used in the latter
experiment helped mask the effects of contamination.
87
-------
Runs 54b, 55, 56, and 58 were all done in the small bag. Prior to
these runs, the small bag had only been used for SO 2 photooxidation experi-
ments. Only one of these experiments, run 53, showed any dark growth and
that resulted primarily from the extremely high (5.6 ppm) SO 2 concentration
used. The low reactivity of the small bag as indicated by lack of dark
growth was partly supported by the low NO photolysis rates observed in runs
54b, 56, and 58. Run 55, however, exhibited greater NO photolysis. The
reason for anomalous behavior in run 55 is not clear; however, it is possible
that a contaminant may have been present for this experiment only, since the
other rates of NO disappearance for the small bag were much lower.
Chamber contamination can thus be monitored using either dark reac-
tivitytests or NO photolysis experiments. The results discussed above
indicate a qualitative correlation exists between the two methods. Thus,
dark growth and high NO photolysis rates, in general, occur together. It
would appear, however, that in order to provide a sensitive measure of
chamber contamination using the NO photolysis method that relatively low
initial concentrations of NO must be used. While such experiments may be
used to give an indication of chamber contamination, they are neither the
only nor necessarily the best experiments for this purpose. More work on
this problem area is necessary.
88
-------
Section 6
RE FERENCES
Bray, W.H., 1970: “Water Vapor Pressure Control at Aqueous Solutions of Sul-
furic Acid, J. of Materials,”Vol. 5, No. 1, p. 233-248.
Bufalini, J.J., S.L. Kopczynski, and M.C. Dodge, 1972: “Contaminated Smog
Chambers in Air Pollution Research’ t , Environmental Letters, Vol. 3,
No. 2, pp. 101-109.
Bufalini, J.J. and E.R. Stephens, 1965: “The Thermal Oxidation of Nitric
Oxide in the Presence of Ultraviolet Light”, Intl. J. Air Wat. Poll.,
Vol. 9, pp. 123-128.
Clark, W.E., 1972 Ph.D. thesis, University of Minnesota, Measurement of Aerosol
Produced by the Photooxidation of SO 2 in Air.
Clark, W.E. and K.T. itby, 1972: “Measurement of Aerosols Produced by the
Photochemical Oxidation of SO 2 in Air”, J. Coil. Interface Sci., Vol.
51, No. 3, p. 477.
Cvetanovic, R.J., 1963: Adv. in Photochemistry, Vol. 1, p. 115.
Demerjian, K.L., J.A. Kerr, and J.G. Calvert, 1973: Adv. in Environ. Sci.
Technol., Vol. 3, Wiley-Interscience, New York.
Glasson, W.A. and C.S. Tuesday, 1970: Environ. Sci. Technol., Vol. 4, p. 916.
Kocmond, W.C., D.B. Kittelson, J.Y. Yang, and K.L. Demerjian, 1973: “Deter-
mination of the Formation Mechanisms and Composition of Photochemical
Aerosols”, First Annual Summary Report, Calspan Report No. NA-5365-M-l,
Caispan Corporation, Buffalo, New York 14221.
Liu, B.Y.H. and D.Y.H. Pui, 1975: “On the Performance of the Electrical Aerosol
Analyzer”, to be published in J. Aerosol Science.
Liu, B.Y.H. and D.Y.H. Pui, 1974: “A Submicron Aerosol Standard and the Pri-
mary, Absolute Calibration of the Condensation Nuclei Counter”, J.
Coiloid and Interface Science, Vol. 47, p. 155.
Liu, B.Y.H., K.T. Whitby, and D.Y.H. Pui, 1974: “Size Distribution Measurement
of Submicron Aerosols by a Portable Electrical Aerosol Analyzer”,
J. of the Air Poll. Control Assoc., Vol. 24, p. 1067.
Morris, E.D., Jr. and H. Niki, 1971: J. Phys. Chem., Vol. 75, p. 3640.
Skala, G.F., 1963: “A New Instrument for the Continuous Measurement of Con-
densation Nuclei”, Anal. Chem., Vol. 35, p. 702.
89
-------
Stedinan, D.H. and H. Niki, 1973: Environ. Letters, Vol. 4, p. 303.
Wei, Y.K. and R.J. Cvetanovic, 1963: Can. J. Chem., Vol. 41, p. 913.
Whitby, K.T. and W.E. Clark, 1966: “Electrical Aerosol Particle Counting
and Size Distribution Measuring System for the 0.015 to 1 im Size
Range”, Teilus, Vol. 18, p. 573.
Whitby, K.T., B Y.H. Liu, R.B. Husar, and N.J. Barsic, 1972: “The Minnesota
Aerosol Analyzing System Used in the Los Angeles Smog Project”, J.’ Coil.
and Interface Science, Vol. 39, p. 136.
90
-------
TECHNICAL REPORT DATA
(Please read in ictloas on the reveg’Ee befc,-e compieil gJ
1. REPORT NO. 2.
EPA- 5OJ3-75-OO7
3. RECIPIENTS ACCESSION NO.
4. TITLE AND SUBTITLE
STUDY OF AEROSOL FORMATION IN PHOTOCHEMICAL AIR POLLUTION
5. REPORT DATE
S. PERFORMING ORGANIZATION CODE
2. AUTHOR(S)
W.C. Kocmond, D.B. Kittelson, J.Y. Yang, and
K.L. Demerjian
I. PERFORMING ORGANIZATION REPORT NO.
NA-5365-M-2
L PERFORMING ORGANIZATION NAME AND ADDRESS
Calspan Corporation
P. 0. Box 235
Buffalo, New York 14221
10. PROGRAM ELEMENT NO.
1A1008
11. CONTRACT/GRANT NO.
68-01—1231
12. SPONSORING AGENCY NAME AND ADDRESS
Coordinating Research Council, Inc./Environmental Protection
30 Rockefeller Plaza Agency
New York, New York 10020 Office of Res. & Devel.
CAPA-8-71 Washington, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Annual
14.SPONSORINGAGENCYCOOE
15. SUPPLEMENTARY NOTES
Prepared in cooperation with the Particle Technology Laboratory, University of Minnesota
— A flfl fl fl
,e. .
Photochemical aerosol production in several SO 2 + clean air (filtered air), HC+NO and
HC+NO-l-S0 2 systems has been examined u ing the smog chamber approach. The reaction vessels
used in this study were the 20,800 ft Calspan chamber and the 600 ft 3 University of Minnesota
chamber. Aerosol formation, growth, and decay mechanisms are described for each of the sys-
tems studied. It has been possible in this investigation to characterize system reactivity
in terms of aerosol behavior. The most important variables are maximum number concentration,
equilibrium surface concentration, and particle volumetric growth rate. Measurements of
these variables are made for several systems and are discussed within the text.
Of the hydrocarbons studied, cyclohexene was the most reactive in terms of aerosol pro-
duction and rate of NO oxidation followed by m—xylene, hexene, and toluene. For the simple
HC+NO system, each experiment can be divided into two phases. During the initial phase, NO
is converted to NO 2 and some oxidation of hydrocarbon occurs. No appreciable aerosol is
formed during this phase, but ozone starts to appear near the end of this period. The second
phase, accompanied by substantial aerosol formation, begins as soon as NO is oxidized out of
the system and NO 2 reaches a maximum; ozone levels rise rapidly during this phase and approact
a maximum. The addition of S02 to the HC+NO system leads to some aerosol formation during th
first phase and was generally found to exert a synergistic effect on aerosol formation in the
second phase. The addition of SO 2 also led to a marked decrease in the diameter of the par-
ticles ultimately formed. This results from the formation of very high concentrations
7. KEY WORDS AND DOCUMENT ANALYSIS
I. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
C. COSATI Field/Group
Photochemi s try
Photochemical Air Pollution
Aerosol Formation
Smog Chamber
IS. DISTRIBUTION STATEMENT
Unlimited Distribution
19. SECURITY CLASS (ThIs Repo,r [
Unclassified
21. NO. OF PAGES
185
20 SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220.1 (0.73)
91
-------
—
Unclassified
SECURITY CLASSIFICATION OF THIS PAGEOThCn DaIa Enl.r.d )
16.
of nuclei during the initial stages of the experiment.
For the SO 2 + clean air system, photooxidation rates of a few tenths
of a percent per hour are typically observed for a light intensity of 50%
noon day sun. In the presence of hydrocarbons and NO, accelerated rates
are generally observed.
The data show that aerosol formation rates are enhanced at high rela-
tive humidities, probably as a result of the higher water content of the
aerosols.
Unclassified
SECURITY CLASSIFICATION OF THIS PAGE( When Del Ente,d)
92
-------
APPENDIX A
AEROSOL AND CHEMISTRY DATA FROM NOVEMBER 1973 WORKSHOP
WITH DUPLICATE UNIVERSITY OF MINNESOTA EXPERIMENTS
ORDER OF PRESENTATION:
(1) SO 2 Experiments
(2) Toluene 1 Toluene + NO 2 , Toluene + NO 2 + SO 2
(3) Hexene, Hexene + NO 2 , Hexene + NO 2 + SO 2
A-i
-------
Table V. SUMMARY OF DATA FROM NOVEMBER 1973 WORKSHOP
Rj .in
No.
System
RH
%
HC
ppm
NO 2
ppm
SO 2
ppm
Nm x
x10 crii 3
SE
ii 2 /cc/hr
dv
(SO 2 )
um 3 /cc/hr
SO 2
photox.
%hr 1
1
2
3
4
5
6
SO 2
so 2
SO 2
SO 2
SO 2
SO 2
30
33
34
31
32
b*
b
b
b
b
b
b
b
b
b
b
b
0.63
0.70
0.58
0.55
0.52
0.55
560
700
525
570
960
750
470
not
reached)
>600
>600
>1100
1600
1.71
6. 15i**1
13.75f**)
1.60i
3.60f .3
1.201
2.83f .3
7.33
6.0 i
16.7 f )
.03
.17 1
.37 j
.051
.ii .i
.041
.ogj
.27
.211
.57)
7
8
9
10
11
12
toluene
toluene + NO 2
toluene + NO 2
toluene + NO 2 + SO 2
toluene + NO 2 + SO 2
toluene + NO 2 + SO 2
35
37
34
45
36
33
0.8
0.8
0.8
0.8
0.8
0.8
b
5.0
3.3
1.95
3.55
1.45
--
.19
.07
.01
NO AEROSO
1000
800
500
L
1000
then less
700
then less
280
then less
7.50i 1
1.90f 3
5.OOi 1
0.41f )
0.88i 1
0.28f j
.64 )
.18 )
1.30 1
0.10 )
1.65
.53 3
13
14
15
hexene
hexene + NO 2
hexene + NO 2 + SO 2
38
34
32
0.6
0.6
0.6
b
3.35
1.64
---
---
.01
95
46
580
>50
---
400
---
2.4 i
0.9 f 3
---
4.57 1
1.71.)
lights used on experiments 1, 3 and 4 to duplicate previous year’s light intensity, i.e., kd [ NO 2 J O.OS min’
i and f refer to initial (usually first 30 mm) and final growth rates.
b = background
-------
cd
TABLE VII. UNIVERSITY OF MINNESOTA DUPLICATE TESTS SUBSEQUENT TO NOVEMBER WORKSHOP:
SUMMARY OF SO 2 EXPERIMENTS
“
No.
Concentration ppm
RH
%
N
#-co 4
SE
jm2_cci’
dv
dt SO
m 3 ccL1 j
1 dv
SO dj
p 3 -c6 -k1i ppiI
photox
%.hr’
1
.51
13
90 K
1420*
2.3
14.5
.12
2
.53
9
].50 K
1420
2.6
14.9
.13
3
.64
614
320 K
660
3.1
14.8
.057
24
.59
36
170 K
3140
.98
1.7
.030
5
.55
35
160 K
360*
.93
1.7
.032
6
.55
29
110 K
210
.149
.89
.018
11
.55
30
600 K
1300
12.0
22
.142
12
.47
I .9
330 K
11400
9.8
21
.35
17
.54
27
220 K
1420
1.1
2.0
.0141
18
.012
17
524. K
72
.12
10
.23
19
.012
36
35 IC
25
--_
---
---
20
. ij
29
280 K
620*
2.2
14.1
.079
Equilibrium surface not reached.
-------
TABLE VIII . UNIVERSITY OF MINNESOTA DUPLICATE TESTS SUBSEQUENT TO NOVEMBER WORKSHOP:
SUMMARY OF CHEMICAL AND AEROSOL DATA FOR TOLUENE EXPERIMENTS
Run
No.
System
RH
%
HC
ppm
NO
pp
SO
PP
Nmax
#- 1
SE
, fl_ C l
dv
SO 2
n 3 -cc 1 -hr 1
SO 2
photox
%-hr’
7
Toluene
24
2.0
b
---
---
---
---
---
8
9
Toluene
Toluene + NO 2
24
33
2.0
1.5
b
3.90
---
---
---
2.2 K
---
NO AEROSOL
--- ---
---
---
10
Toluene + NO 2
32
1.0
3.34
---
.25K
---
---
---
13
Toluene + NO 2 + SO 2
27
1.0
.26
.01
250 K
1200*
14
Toluene ÷ NO 2 + SO 2
36
1.0
4.19
.01
50 K
---
---
---
15
Toluene + NO 2 + SO 2
25
1.0
1.23
.54
660 K
1200
4.7
.35
16
Toluene + NO 2 + SO 2
26
1.0
1.95
.54
740 K
1300
3.6
.28
Equilibrium surface not reached.
-------
(1) SO 2 EXPERIMENTS
A-S
-------
600
500 —
400 —
E
U
300 —
2
2
200 —
100 —
0—
x 1 0 5 — x10 3
12 —
10 — 1250
8 — 1000
-2
N
U
6 — _ 750
LU
2 C.)
4
2
4 — 500
C l )
2
—
250
0—
0
600
500
400
300
LU
C.)
4
200
100
CALSPAN
RUN NO. 1 NOV. 11, 1973
SO 2 = 0.63 ppm
STIRRING
- = 1.71 pm 3 /cm 3 -hr
dt _______________
5
C )
4E
U
C d )
E
3 —
LU
2
-J
0
2>
1
0
0
0 20 40 60 80 100 120 140 160
TIME (mm)
x 101
30
c v ,
20 E
U
C d,
E
uJ
2
-J
0
>
0 20 40 60 80 100 120 140 160 180 200
TIME (mm)
A-6
-------
RUN 1 DATE. 12—DEC—73
NUM.(PART./ML)
S02 0.51
SYSTEM: S02
, SURF.( 11 2 /ML)
PPM
0 UOL,qJM 3 .’ML)
N8
U
E
R 6
X l gI4
l0
9—
:u
:R
:A
—C
:E
3—
2—
5
4
0
TIME (10+2 MINUTE)
e l i ,
_p —
(IJM 2 ’IIL)
SYSTEM.
A SURF
PPM
0 (JOL . ( iJt’1 3 ’ML)
M
B
E
R
RUN 2 DATE 14—DEC-73
o NUM . ( Pi RT ‘ML)
502 0.53
1 g1 5
-S
—U
p
-
-c
E
1—
0
4 )
0
L
U
71
—
TIME (jØ+2 MINUTE)
A-7
-------
6— 12
io2
5— 10
6-6
x i0 2
5— 5
4—
NE
3— —3
I L ’
C.)
2—
1— 1
0— 0
CALSPAN
0 AEROSOL VOLUME CONC. pm 3 1cm 3
AEROSOL SURFACE CONC. p m 2 /cm 3
D AEROSOL NUMBER
6
x 100
5
4(
E
E
3
ILl
S
1
20
>
1
0
6
x io
5
I ,,
4 E
E
3—.
w
S
-J
20
>
1
0 20 40 60 80
TIME (mm)
RUN NO. 3 NOV. 13, 1973
SO 2 = 0.58 ppm
STIRRING
dV = 1.60i pm 3 lcm 3 .hr
. j. 3.6O
10
c v ,
E
U
U i
S
io
E
U
Ui
2
0 20 40 60 80 100 120
TIME (mm)
0
100 120 140
A-8
-------
SYSTEM’ S02
A SURF.(IJM 2 /ML)
PPM
SYSTEM: 502
I SURF.(1JM 2 ’ML)
PPM
0 UOL.( M 3 /ML)
0 U0L.( JM 3 /ML)
RUN 3 DATE. 18—DEC—74
0 NUM.(PART.’ML)
S02 0.64
8U
0
7b
M
N
U
M
B
E
R
H
U
‘.1
B
E
R
TIME j +2 MINUTE)
RUN 4 DMTE 13—DEC—’3
NLIM (PI RT . ‘ML)
S02 = 0.59
‘uo ‘ io 2
2 — 4
1
—IJ
R
-A
-C
E
Ii
0
L
U
E
1
0
1
0
I
TIME rie+ 2 MINUTE)
A-9
-------
E
U
LU
2
z
12 —
10 —
8—
6—
4—
2—
x 10
(I
E
U
LU
S
z
10 —
8—
6—
4
2
0
x 10
2
C?
E
U
N
E
LU
C.,
4
U.
0 40
CALSPAN
x 101
20
C.)
E
U
12
IiJ
S
8
-j
0
>
4
x 101.
30
C .,
E
20
E
U i
2
-I
0
>
10
RUN NO. 5 NOV. 13. 1973
SO 2 = 0.52 ppm
STIRRING
= 7.33 p m 3 Icm 3 -hr
dt
O AEROSOL VOLUME CONC. p m 3 Icm 3
A AEROSOL SURFACE CONC. pm 2 /cm 3
D AEROSOL NUMBER
o 20 40 60 80
TIME (mm)
0
ioo 120 140 160
80 120
TIME (mm)
0
160 200 240
A- 10
-------
SYSTEM: S 02
A SUPF.(LrM 2 /ML)
PPM
0 VOL UM 3 ’ 1L)
RUN 6 DATE 19-OEC-?3
0 HUM PART /ML)
S02 = 55
H
U
I.’
B
E
R
-S
—U
R
-A
-c
E
1
Q UOL ( J 1 ‘ ML ’)
RUN 5 DATE: 19—DEC—73
NUM.(PART ‘ML)
502 = 0 55
N
U
M
B
E
R
. 5
—U
R
-A
-C 3
E
4t)
0
L
U
M
3 E
1
S
TIME (1e MINUTE)
SYSTEM 602
SURF (IJtl 2 ’ML)
PPM
0
L
M
I
S
A-li
-------
RUN 11 DATE
22—DEC—73
SYSTEI
1: S02
D N
UM.(PART.’ML)
SU
RF.qJM 2 ’ML) 0 ‘JUL ((JM 3 FiL)
U
a
L
‘J
E
N
U
B
E
R
H
LI
1 1
E:
E
6
-s
-U
-F
-A
:c
4 —E
3—
1—
0—
RUN 12 O TE 26—DEC—?’3
D NUM (PART . ‘ML)
1
4—
-5
-
-R
-
-E
1—
0—
TIME (10+2 MINUTE)
S STEf1 S02
SURF ((JFI ‘ML)
S02 = 0.45 PPM
0 UOL.( JM 3 ’ML)
U
70
Li
I 1
E
A- 12
-------
SYSTEM: S02
SURF (PM 2 /IIL)
PPM
SYSTEM 502
, SURF (jJM 2 ’ML)
Q LJOL. ( (ill 3 /ML)
0 UOL.((.1M 3 ,ML)
RUN ir DATE 3—JAN-74
D HUM (PART . ‘ML)
S02 = 0.54
U
LI
1•1
B
E 2
R
i
3—
-s
-U
1—
U
0
L
U
2 ri
E
TIME
1
(10+2 MINUTE
RUN 18 DATE 3-JAN-74
0 HUM (PART./ML)
:
-------
RLIN 19 DATE 4—J N-7’4
0 HLIM (PA T./ML)
502 = 0.012
SYSTEM S02
SURF (JM 2 .’ML)
PPM
QLJOL ( jii 3 /ML)
N
I-I
ri
B
E
R
i
5—
4 —9
:
:
A—
-r
:E
2
-
5
4 H
0
L
U
M
1
0
0
RLIN 20 DATE
0 NUN
1
TIME (10 f1INUTE)
5—J AN—74 SYSTEMS 502
(PART.’ML) A SURF.(FJM 2 /ML)
502 = 0.54 PPM
0 UOL.(IJM 3 ’NL)
ri
E:
E
R
.< i
-S
• IJ
,
A
-C
•E
1—
U
0
‘-‘U
M
E
0
TIME (10+2 MINUTE)
A- 14
-------
(2) TOLUENE, TOLUENE + NO 2 , TOLUENE + NO 2 + SO 2 EXPERIMENTS
A- 15
-------
CALSPAN
RUN NO.7 14 NOVEMBER 1973 I TOLUENE + BACKGROUND NO SYSTEM R.H. = 35%;
TOLUENE = 0.8 ppm; NO 2 = 0.00 ppm; NO = 0.004 ppm
2
LIGHTS OFF
‘150 mm
o [ NO 2 ]
z (NOl ppb
o [ 031 ppb
E
a
a
(,)
01
0
2
(‘1
0
2
0
0 50 100
TIME (mm)
150
A- 16
-------
8
CALSPAN __________________
RUN NO. 8 14 NOVEMBER 1973 ITOLUENE + NO 2 SYSTEMI R.H. = 37%; TOLUENE = 0.98 ppm;
NO 2 = 5.0 ppm; NO = 0.03 ppb
x lO° — xlO 1
5— 10
4—
3-
2-
1— 2
0— 0
E
a.
a.
(‘1
0
z
a.
a.
c v ,
0
0
z
6
4
0 100 200 300 400 500 600 700 800
TIME (mm)
OFF
A- 17
-------
CAISPAN _________________
RUN NO.9 15 NOVEMBER 1973 JTOLUENE + NO 2 SYSTEM ] R.H. = 34%;
TOLUENE = 0.8 ppm; NO 2 = 3.30 ppm; NO = 0.03 ppm
5
io
4
x 101
E
a
a
0
ft
a
0
0
z
3
2
1
0
0 50 100
TIME (mm)
150
A- 18
-------
RUN 9 DATE 21-DEC-73 SYSTEM TOLUENENO2
0 NUMBER CONC (PART ‘ML)
TOLUENE = 1 ,5 PPM
N02 - = 3.9 PPM
U. of M.
x i
N
U
M
B
E
R
C 2
U
N
C
1
2
1
a
N
I
T
R
G
E
N
0
x
I a
0
E
S
a 1 NO 2 (PPM)
4
3.
1
2
1
TIME (
MINUTE)
0
A-19
-------
RUN 10 DATE 22—DEC—73 S’rSTEM
o NUMBER CONC. (PART . ‘ML)
TOLUENE = 1.04 PPM
N02 3.34 PPM
TOLUEHE N02
U.ofM.
N
U
N
B
E
R
U
TIME ( io 2 MIHIJIE)
x l’
4
N
I
T
R
0-
G
E
N
0
x
12
D
E
S
1
TIME jg+2 MINIJTE)
A- 20
-------
CALSPAN ______________________
RUN NO. 10 15 NOVEMBER 1973 LNO 2 + SO 2 + TOLUENE SYSTEM R.H. 45%; TOLUENE 0.8 ppm;
NO 2 = 1.95 ppm; NO = 0.022 ppm; SO 2 = O 19 ppm
STIRRING
10 xlO°
x10 5 x10 2
- 8
I , )
E
( ) U
— E
-6 6
- 4 4
I L
- 2 2
- 0
800
2-
xlO° x10
.0
E
0 .
N 0
o
z
0-
1000
TIME (mm)
0 100 200 300 400 500 600 700 800 900
TIME (mm)
A- 21
-------
RUN 13 DT TE 27—DEC—73
SYSTEM: TOLUENE, NO2 S02
0 N
LIM (PART 4’1L)
SLIRF C JM 2
“ML) 0
VOL C Jf1 3 /ML)
<1Li 3
- LI
-C
-E
1—
1
U. of M.
N
LI
ri
E
R
U
0
L
U
E
TIME
A- 22
-------
CALSPAN [ NO 2 + S0 2 + TOLUENE SYSTEM
RUN NO. 11 16 NOVEMBER 1973
NO 2 = 3.6 ppm; NO = 0.025; $02 = 0.07 ppm
STIRRING
x 1o 2
E
U
(N
E
‘U
C.,
U.
(I,
RH. = 37%; TOLUENE = 0.8 ppm;
x 100
x io
8-
6-
(p)
E
U
‘U
m
z
2-
Op
E
U
c v,
E
‘ U
-j
0
>
250 300
TIME (mm)
4
K 100
16
X 101
0
0
3.
2
E
(N
0
z
1
0
o 50 100 150 200 250 300 350 400 450 500 550
TIME (mm)
A- 23
-------
RUN 14 O TE 27-OEC—73
0 NUM EP. C:QNC (PART
S STEh TOLUENE H02, E;C’2
A 1 ‘NUMBER COHC C ML/FART
TOLUENE 1.04 PPM
S02 = 0.012 PPM
U. of M.
I.’
E:
E
F.:
fl
C
9 ,1..
N
8U
t 1
B
P
0•
5N
C
4
TI 1E
-------
CALSPAN
RUN NO. 12 17 NOVEMBER 1973 [ NO 2 + TOLUENE SYSTEM ] R.H . = 33%; TOLUENE = 0.8 ppm;
NO 2 = 1.45; NO 0.01 ppm SO 2 = o.bi ppnil
STIRRING
500
x io2
E
N
E
uJ
U
4
U.
C ,,
100
5
io
4
C
Lu
1
100
1
0.5
0
E
U
E
uJ
D
-t
0
>
2
TIME (mm)
E
N
0
2
0 50 100 150 200 250 300 350 400 450
TIME (mm)
A- 25
-------
RUN 15 DATE: 28—DEC—73
0 HUM.(P1 RT.41L)
SYSTEM TOLUENE SO2.NO2
SURF (jJt1 2 ’PIL)
U. of M.
0 UOL.qJM 3 /ML)
N
B
-ç
—
-R
-E
4—
—
U
0
L
IJ
2 ri
E
0
xl
TIME
I I I I
MINUTE)
I I
I I
0 NO (PPM)
N
I
I
R
0
G
E
N
0
x
Ii
0
E
S
xl i
+0
2
:LTS. OFF
NO 2 (PPM)
I
1
TIME (10+2 MINUTE)
A-26
-------
0 NLI 1i.P RT./NL)
SYSTEIl: TOLUENE.. 502. NC’2
SURF.(jJII 4 /ML)
U. of M.
0 ‘JUL. (lJt1 3 /ML)
N
E;
RLIN 16 D TE 28-DEC—73
7—
-R
:F
-
-E
4—
U
0
L
U
2
E
2
1
U
N
I
T
R
U
C
E
N
0
1
Ii
0
E
S
TIME (10+2 MINUTE)
A- 27
-------
A- 28
-------
(3) HEXENE, HEXENE + NO 2 , HEXENE + NO 2 +S0 2 EXPERIMENTS
A- 29
-------
CALSPAN
RUN NO. 14 19 NOVEMBER 1973 {iIEXENE-1 + NO 2 SVSTEMJ R.H. = 34%; HEXENE-1 = 0.6 ppm;
NO 2 = 3.35 ppm; NO = 0.012 ppm
5
100 X
6
5 —
E
E 3
a a
a 4
N
0
22 32
‘ C
‘ U
2
1
1
10
io2
I
0—
0
TIME (m m)
0
100 200 300 400 500 600 700 800 900 1000
A- 30
-------
CALSPAN
RUN NO. 13 19 NOVEMBER
STIRRING
io
10 —
8—
c )
E
U
LU
2
24
2—
0—
100
16
12 -
a
(‘1
0
2
0
2
8—
4-
0-
101
4
E
N
E
LU
U
U.
U)
1
1973 IHEXENE1 = 0.6 ppm JR.H . = 38%; NO 2 = 0.006 ppm; NO = 0.011 ppm
I I I I I I
AEROSOL COMBINATION 0 NUMBER
t SURFACE
OVOLUME
0
0
0
______ LIGHTS OFF 200 mm
Ipu 10101 • i i
(1 20 40 60 80 100 120
2
1
TIME (mm)
140 160 180 200
x io 2
10
8
c )
E
E
LU
2
4
0
>
2
—0
220
10.1
E
a.
a
I-
th
2
LU
x
LU
X 101
4
3
a.
a
( )
0
0
TIME (mm)
140 160 180 200 220
A- 31
-------
c )
E
U
-N
E
‘ L i
U
4
U.
C ,)
I HEXENE-1 + SO 2 + NO 2 SYSTEM
CALSPAN
RUN NO. 15 20 NOVEMBER 1973
NO 2 = 1.64 ppm; NO = 0.018 ppm; SO 2 = 0.01 ppm
x io
8-
p)
‘ L i
4-
2
2-
E
a.
a.
N
0
2
0-
1.6 -
1.2 -
0.8 —
0.4 -
0—
R.H. = 32%; HEXENE-1 = 0.6 ppm;
02
1000
io 2
x io
0 100 200 300 400 500 600 700 800 900 1000
TIME (mm)
C v ,
E
U
E
‘ LI
-I
0
>
E
a.
U.’, a.
• UJ
2
0.4
.ft 100
a.
C v ,
0
0
z 80
0.8
TIME (mm)
A- 32
-------
APPENDIX B
AEROSOL AND CHEMISTRY DATA FROM MARCH 1974 WORKSHOP
WITH DUPLICATE UNIVERSITY OF MINNESOTA EXPERIMENTS
ORDER OF PRESENTATION:
(1) SO 2 Experiments
(2) Toluene Experiments
(3) Hexene Experiments
(4) M-xylene Experiments
(5) Cyclohexene Experiments
(6) NaC1 Experiments
B-i
-------
TABLE IX. SUMMARY OF AEROSOL DATA FROM MARCH WORKSHOP
Run
No. System
RH
%
N ,
#- c
SE
im 2 -cd
dv
dti 0 )
iim3_c -hrZ
dv
dt(HC)
m 3 -cd hi
502
Photox
%_hrM
Comments
6 toluene + NO
30 toluene + NO
29 toluene ÷ NO + SO
7 toluene + SO 2 2
30
20
30
20
3.1x10
l.3x10
1.6x10 5
2.1x10 5
640
750
>750
800
--
--
0.78
1.17
2.2
2.6
1.5
3.2
--
--
0.32
0.45
no vol. first 4 hrs
no vol. first 4 hrs
1st 4 hrs*
1st 50 inin
S hexene + NO
21 hexene + NO
18 hexene + NO + SO
20 hexene + SO 2 2
40
37
37
35
1.4x1O
l.2x10
1.4x10 5
3.6x10 5
610
215
>1500
950
--
--
0.61
0.75
2.1
0.5
5.8
3.2
--
--
0.16
0.25
no vol. first 5 hrs
no vol. first 6 hrs
1st 6 hrs*
1st 60 min*
15 m-xylene + NO
14 m-xylene + NO + SO 2
17 m-xylene + SO 2
38
29
35
8.4x1O
2.6x10 5
2.8x10
1150
2700
384
--
0.92
0.84
14.1
25.0
1.6
--
0.32
0.23
no vol. first 60 mm.
1st 60 mini
1st 60 min*
10 cyclohexene + NO
12 cyclohexene + NO
9 cyclohexene + NO + 2
13 cyclohexene ÷ SO 2
38
30
30
35
3.6xl0
4.2x10 5
l.7x10 5
2.7x10
3500
2450
4200
1300
--
--
0.74
1.20
110
75
105
10
--
--
0.28
0.49
no vol. first 90 mm
no vol. first 3 hrs
first 2 hrs
first 30 mm
1 0.52 ppm SO 2
4 0.55 ppm SO 2
2 0.05 SO 2
3 0.05 SO 2
25
30
37
40
5.5x10 5
3.9x10 5
2.3x10 5
2.9x10 5
>1450
4400
>575
>675
5.61
10.60
4.49
13.60
0.65
2.35
1.04
2. 3l
--
--
--
0.21
0.39
0.16
0.48
0.23
0.79
0.36
0.79
1st 30 min*
1st 2 hrs
1st 30 min*
1st 2 hrs
1st 30 min*
1st 2 hrs
1st 40 min*
1st 2 hrs
Time over which aerosol growth rate was used in computing SO 2 photooxidation.
-------
Table X . UNIVERSITY OF MINNESOTA DUPLICATE TESTS SUBSEQUENT TO THE MARCH WORKSHOP:
SUMMARY OF AEROSOL DATA FOR HYDROCARBON EXPERIMENTS
Run
No. System
65 toluene + NO + SO 2 28 185 IC 330* 1.23 11.3 same .23
76 toluene + NO 47 4.2 K 550* -- -- 24.5
77 toluene + NO + SO 2 57 170 K 1850 1.17 29.9 27.4 .39
87 toluene + NO 30 10 K 340 -- -- 8.6
88 toluene + NO + SO 2 24 160 K 1600* 1.22 30.5 16.6 .67
60 hexene + NO + SO 2 28 74 K 1200* .40 5.7 21.2 .12
78 hexene + NO SO 2 55 230 K 1530 .53 13.9 19.1 .18
92 hexene + NO 33 8.8 K 31 -- -- .09
93 hexene + NO + SO 2 32 150 K 1300 .29 8.5 18.0 .16
81 m-xylene NO 75 23 K 1600 -- -- 73
82 m-xylene + NO SO 2 54 230 K 2800 .49 10.4 67 .14
89 m-xylene + NO 26 21 K 1000* -- -- 38
91 m-xylene + NO + SO 2 26 230 K 1600 .46 10.0 31 .21
83 cyclohexene + NO 51 .9 K 320 -- -- 50
94 cyclohexene + NO 31 2.7 K 510 -- -- 65
95 cyclohexene + NO 29 1.9 K 620 -- -- 190
96 cyclohexene + NO + SO 2 28 280 K 5400 .43 9.7 250 .20
*
Equilibrium surface not reached.
-------
TABLE XI. SUMMARY OF CHEMISTRY DATA FROM MARCH WORKSHOP
Run RH HC SO 2 NO 1 t(NO 2 ) °3m
No. System % ppm ppm ppm mm ppm
6
toluene + NO
30
0.35
--
0.170
400
0.285
30
toluene + NO
20
1.17
--
0.530
480
0.380
29
7
toluene + NO + SO 2
toluene + SO 2
30
20
0.35
0.35
0.05
0.05
0.146
b
330
b
>
0.225
0.047
5
hexene + NO
40
0.33
--
0.150
420
>
0.200
21
hexene + NO
37
0.33
--
0.180
420
0.275
18
20
hexene + NO + SO
hexene + SO 2 2
37
35
0.33
0.33
0.07
0.055
0.178
b
430
b
--
0.052
15
m-xylene + NO
38
0.34
--
0.150
100
0.222
14
17
m-xylene + NO + SO 2
m-xylene + SO 2
29
35
0.34
0.34
0.055
0.07
0.150
b
105
b
0.305
0.030
10
cyclohexene + NO
38
0.33
0.138
120
0.190
12
cyclohexene + NO
30
0.33
--
0.140
190
0.192
9
13
cyclohexene + NO +
cyclohexene + SO 2
SO 2
30
35
0.33
0.33
0.05
0.06
0.220
b
180
b
0.325
0.011
-------
Table XII. UNIVERSITY OF MINNESOTA DUPLICATE TESTS SUBSEQUENT TO THE MARCH WORKSHOP:
SUMMARY OF CHEMICAL DATA FOR HYDROCARBON EXPERIMENTS
Run RH HC SO 21 NO [ NO 2 ] t [ NO ] [ 0 3]max
No. System ppm ppm ppm ppm mm ppm
65 toluene + NO + SO 2 28 .35 .108 .30 .145* 460 .2*
76 toluene + NO 47 .35 -- .152 .095 210 .30
77 toluene NO + SO 2 57 .38 .039 .155 .115 160 .362
87 toluene + NO 30 .35 -- .155 .140 130 .402
88 toluene NO + SO 2 24 .35 .040 .17 .122 155 .315*
60 hexene + NO + SO 2 28 .35 .07 .16 .123 395 .162*
78 hexene + NO + SO 2 55 .35 .038 .165 .130 255 .438
92 hexene ÷ NO 33 .35 -- .12 .104 280 .290
93 hexene + NO + SO 2 32 .35 .034 .122 .125 350 .302*
81 m-xylene + NO 75 .35 -- .155 .144 80 .343
82 m-xylene + NO + SO 2 54 .35 .047 .151 .130 94 .361
89 m-xylene + NO 26 .35 -- .132 .142 68 .379
91 m-xylene + NO + SO 2 26 .35 .046 .117 .115 70 .262
83 cyclohexene + NO 51 .35 -- .13 .101 90 .32*
94 cyclohexene + NO 31 .35 -- .103 .108 60 .20
95 cyclohexene + NO 29 .35 -- .124 .128 103 .254
96 cyclohexene + NO + So 2 28 .35 .045 .133 .130 85 .241
*max not reached by end of irradiation period.
-------
MINNESOTA DUPLICATE TESTS SUBSEQUENT TO THE
SUMMARY OF 502 EXPERIMENTS
TABLE XIII. UNIVERSITY OF
MARCH WORKSHOP:
j j
No.
Concentration ppm
RH
%
N
#-cc
SI!,
24
WI - CC
dv 1 dv
dt SO °2
3 -
TJBI -CC-hr -c6 .hr’.ppm’
SO 2
photox
tr
t 5
.59
28
260 K
L 7o
1.9
3.2
.07
L 6
.59
L O
3L 0 K
570
2.2
3.8
.065
L 1 9
.59
67
590 K
Boo
5.6
9.4
.10
52
.56
36
90 K
190
.46
.83
.015
53
.56
30
670 K
2100
20
3.7
.07
5I
.56
25
100 K
190
.28
.50
.01
57
.56
12
39 K
210
.78
1.4
.04
66
.61
27
100 K
680*
2.7
4,5
.10
68
.62
39
9L .K
310
1.1
1.8
.03
69
.053
L 0
280 K
270
.84
16
.27
70
.54
57
210 K
11.00*
1.2
2.2
.03
71
.048
55
95 IC
*
200
.32
6.6
.09
72
.50
5 1 1
220 X C
811.0*
4.2
8.5
.12
79
.038
70
68 K
*
110
.21
5.6
.06
80
.40
78
250 K
550
2.2
5.5
.05
85
.38
3L 1 .
21. 5 K
611.0
2.5
6.6
.13
86
.051
27
82 K
190
.35
6.8
.14
97
.035
27
100 K
*
270
1.1
30
.63
Equ.Jibrium surface not reached, maximum surface.
-------
(1) SO 2 EXPERIMENTS
B- 7
-------
CALSPAN
RUN NO. 1 20 FEBRUARY
4
1974 J SO 2 FILTERED AIR SYSTEM R.H. 25%; SO 2 = 0.52 ppm
CALSPAN __________________
RUN NO.4 21 FEBRUARY 1974 JSO 2 FILTERED AIR SYSTEMI RH. = 35%; So 2 = 0.55 ppm;
NO = 0.010 ppm; NO 2 = 0.004 ppm x 10
C .,
U
U i
2
io x i0 2
C ’)
E
U
N
E
Ui
U
4
I i .
U)
C.)
E
U
‘Ii
2
1
0
C.)
20 E
c v ,
E
Ui
2
-I
0
>
TIME (mm)
x
10
4
C ’)
S
NE
uJ
4
U.
U)
0•
TIME (mm)
B-S
-------
CALSPAN
RUN NO. 2
6
—
12
io
,cl
5
—
10
(1
E
C ? -2
r.1
8
U
.
3-
6
I I I
2 0
<
2
2—
Q,
4
1—
2
0—
0
20 FEBRUARY 1974 SO 2 FILTERED AIR SYSTEM; R.H. = 37%; SO 2 = 0.05 ppm
12
10
8E
-2
E
6 .
Lu
S
D
4
>
2
0
0
20 40 60 80
100 120 140 160
TIME
(mm)
CALSPAN
RUN NO. 3 21 FEBRUARY 1974 SO 2 FILTERED AIR SYSTEM; R.H. = 40%; SO 2 = 0.05 ppm
5— 10
x10 5 x10 2
4— 8
C ? E
uJ
52— LU 4
0
2
U-
I
I— LI)
0— 0
E
U
E
‘U
S
-J
0
>
80 100
TIME (mm)
B- 9
-------
RUN 45 DATE: i5—APR—74
SYSTEM 802
0 N
UM. ( PART. ‘ML)
A
SURF (PM 2 ’ML) 0 ‘ )0L. (PM 3 /ML)
-S
-u
Th
- C:
-
1—j-
0
RLIN 45 DATE 16—APR-74
SYSTEM 502
0 N
UM.(PART.’ML)
, SURF.( M 2 ’ML)
x l i’
4—
-S
L.U
-R
:F
-A
-E
7—
1—
0
B-1O
0 UOL . (pM 3 ’ML)
£02 t 0.59 PPM
H
L i
P1
E’-
R
TIME (10+2 MINUTE)
502 = 0.59
H
B
E
R
8
7
U
Co
U
M
4
3
2
1
0
TIME (10+2 MINUTE)
-------
RUN 49 DATE 20—APR—74
SYSTEM: $02
0 N
LIM. (PART . ‘ML)
SURF ( M 2 ’ML) 0 UOL. (PM 3 ’ML)
x i V2
9
$02 = 0.59 PPM
N
U
N
B
E
R
2
:37
-F -
- A
- c
4 :E
.3 —
4
RLIN 52 DATE: 30—APR—74
V
0
L
U
N
E
1
U
TIME ua 2 MINUTE)
0 HUM (PART . ‘ML)
SYSTEM: $02
t SURF. ( f’1 2 ’ML)
802 = 0.56 PPM
0 VOL ( t iM vNL>
H
B
E
R
R
A ’-
C
E
1
)< 10-1
to
9
gi.)
0
L
7 j
I ’ l
C
3
4
2
1
3-11
-------
RUN 5 DATE
: 1 —MAY—74
SY3TEI
1: 302
0 N
UM (PART. /HL)
SLI
RF. (L’M 2 .’ML)
302 = 5.6 PPM
, LJOL.
-------
SYSTEM: 502
SURF. c: JM 2 ’ML)
PPM
0 VOL. ( IJM 3 /ML
RUN 57 DATE: 8—M AY—74
0 NUM,(PMRT.’ML)
902 = rj.56
X1 4 X:
. LI
:F.
- A 2
-E
1—
Cj
1
U
0
L
LI
F l
E
1
TIME C i +2 MINUTE)
I. ’
U
Fl
B
E
R
RUN 65 DATE: 2 —MAY—74
U MUM . ( PART . ‘ML)
602 = 0.605
-c
—U
P
-A
C
E
1—
SYSTEM 602
SURF. (LIM 2 ’FIL)
PPM
0 VOL. ( JM 3 /ML)
I l
LI
t,1
B
E
F:
xl
I..,
0
L
U
Fl
E
1
0
0
TIME (10+2 MINUTE)
B- 13
-------
RLIN 58 DATE: 25—MAY—74
0 HUM. (PART. /ML)
802 = 0.62
SYSTEM £02
t i SURF. (jJM 2 /ML)
PR I1
0 VOL. (t iM 3 /ML)
< it
10
9—
U -L I -.
M :p
6 7
E :A
P.- LC
:E
4
2
I..J
.70
L
LI
1 1
E
0
RUN 69 DATE: 3—JUN-74
C HUM. (PART. /ML)
1
TIME
(10+2 MINUTE)
x10+ 2
I
SYSTEM: 802
A SURF ( JM 2 /ML)
802 = 0,053 PPM
0 UOL.(IJM 3 /ML)
ti
6 ’
E
F . :
x l
4
X I
4—
L 1
- R
ro
-E
U
U
1 1
E
1
1
U
00
2
1
0
B- 14
-------
SYSTEI1 802
ta SURF ,(IJM 2 ’ML)
0 UOL.(LJM 3 ’ML)
RUN 70 DATE 7—JUN—74
0 NLIM. ( PART . ‘ML)
802 = 0.54 PPM
SYSTEM 802
SURF . ( IJM 2 ”ML)
802 = 0.048 PPM
0 UOL. ((iN 3 41L>
xi d 4
10-
Hg —S
U -U
M :
g7 F.,
E :A
R,ç :..C
3:
2:
i
a—
I . )
0
L
U
N
E
TIME (10+2 MINUTE .)
B-iS
-------
RUN P2 DATEt 25—JUN—74
0 NUM.(PART.41L)
602 = 0.50 PPM
xid 5 xie
3 :10
-9
.U
— p4
• C
.E
—5
.4
1—
1.
0
N
U
11
R
8—
r —
- S
-
6
-A
-E
4—
1 —
0 —
SYSTEM: $02
SURF.(JJM 2 4IL)
0 UOL.(L.JM 3 /ML)
RUN 79 DATE 2—JUL—74
$YiTE1 802
0 N
UM (PART. ‘ML>
A SU
RF. ( M 2 ’ML) 0 VOL. C PM ‘PL)
*1
B
E
R 5
x l
I )
0
L
U
N
E
1
0
7
TIME ( 10+2 MINUTE)
$02 0.038 PPM
4 U
0
L
U
I I
B- 16
-------
SYSTEM 802
A SURF : M 2 ’ML)
0 UOL.(iit’1 3 ’ML)
RUN 95 DATE 22—AUG—74
U NLIM(PART.(ML)
0
-
A
- R -
-E
1—
U
SYSTEM: 802
A SURF . ( 4JM 2 ’IIL)
B- 17
RLtN SO DATE: 2—JUL—74
U NUM.(PART.’ML)
802 = 0.4 PPM
>: 1 U
6
l 1
U U
M •R
,
E •
R .6
1—
5 ( 3
0
L
IJ
E
3
a
1
1
TIME (10 MINUTE)
0
Q U0L.(WM’ /ML)
r i
6 ,
P
0
L
P1
1
TIME (10+2 MINUTE)
-------
RUN 86 DATE: 23—AUG-74 SYSTEM: 802
o MUM , (PART. ‘ML) SURF ( JJM 2 ’ML)
$02 = 0.051 PPM
N10 4 N :
10 r 3
N
:u
H
6 7 : F•
E :A
P -
4r
tr
I
TIME ( 10 MINUTE)
Q U0L,( M 3 ’ML)
902
, l’MUMBER CONC . ( tiL /PART.)
c i
0
L
U
H
E
RUN 97 DATE: 12-SEPT—74 SYSTEM
a NUMBER CONC.(PART.’ML)
902 = 0.035 PPM
I
U
N
E
P
4
t 1
C
3.
B- 18
-------
(2) TOLtJENE EXPERIMENTS
B- 19
-------
io
4—
a
U
‘U
2
CALSPAN
RUN NO. 6 24 FEBRUARY 1974 1T0LUENE.N0FILT E0 AIR SYSTEM ] R.H. = 30%;
TOLUENE 0.35 ppm; NO = 0.170 ppm; NO 2 = 0.01 ppm
8
E
N
3— E 6
‘U
C.)
U.
C.,
0— 0
1
E
0.
0.
( 1
0
+
N
0
2
0
2
2
0.3
0.2
0.1
0
0
2—
4
C ,)
E
U
C v)
‘U
0
0 120 240 360 480 600 720 840 960 1080 1200
TIME (mm)
E
0.
0.
‘U
2
IU
0
I-
120 240 360 480 600 720 840 960 1080
TIME (mm)
0
1200
B- 20
-------
RUN 75 DATE: 28—JUN-74
0 HUM . (PART. ‘ML)
SYSTEM: TULUEHE HO
h
TOLLIENE =
E s SLIF.:F ( M 2 ’ML)
0.35 PPM
Uof M.
0 JUL (jJM 3 ’ML)
5
F 1
B
E
R -
4
xl 1 t
S r
4 Hs
I_ I
P
rF
:A
ir
2
1
X 10+1
6
5 LJ
U
L
IJ
4 ii
E
: 3
2
1
0
N
I
T
R
0
G
E
N
NO (PPM)
4
U
x
12
0
E
S
2
:o
—z
:o
-N
3
.3
1
I
2
0
1
TIME (1042 MINUTE)
0
B- 21
-------
x 102
16
E
N
E
12
w
Ca
U.
C .,
4
CALSPAN
RUN NO. 30; MARCH 18, 1974;
x
0.8 —
0.6 —
0.4 —
0.2 —
E
0.
0
z
0
N
0
2
0—
0.5
0.4
0.3
0.2
0.1
0
I TOLUENE-NO-FILTERED AIR SYSTEM
TOLUENE = 1.17 ppm; NO = 0.53 ppm; NO 2 = 0.044 ppm; RH = 29%
C?
E
U
U i
2
2
0
0 240 480 720 960 1200 1440
TIME (mm)
0 240 480 720 960
TIME (mm)
1200 1440
B- 22
-------
RLIN S ’ DATE: 26—AUG—74 SYSTEM: T0LUEHE HO
NLIM . (PART. ‘ML) SLIRF . (jJM 2 /ML)
TOLUENE = 0.35 PPM
HO
U. of M.
0 UC’L,(1J11 3 /ML)
F l
B
E
F:
I)
0
L
(U
H
-s
R
-A
-c
E
1—
B —
H Lo
I.:Z
T :0
P4 N4
o •E
G
E —
H :
—
I
0
E
—
1
0 —
H
V
0
4R
0
C
A
R
N
4
TIME (10+2 MINUTE)
B- 23
-------
CALSPAN _____________________________
RUN NO. 8 26 FEBRUARY 1974 rTOLUENE.NO .80 2 -FILTERED AIR SYSTEM;IR.II . = 26%;
TOLUENE = 0.35 ppm; NO = 0.138 ppm; NO 2 = 0.01 ppm; SO = 0.055 ppm
2
1
0.
x
( ‘3
E
U
l U
z
C I ,
E
N
E
U i
‘3
U.
(I )
C ’)
E
80
E
I
TIME (mm)
0.2
(d )
0
0
z
N
0
z 0.1
E
0.
Ui
U i
-l
0
I-
TIME (mm)
B- 24
-------
CALSPAN
RUN NO. 29 MARCH 17. 1974
I TOLUENE-NO-S0 2 -FILTERED AIR SYSTEM
TOLUENE = 0.35 ppm; NO = 0.146 ppm; NO 2 = 0.009 ppm; SO 2 = 0.05 ppm; RH = 30%
20— 10
io 2
16 — 8
12 — 6
‘U
C.,
8—44
U-
C l ,
4— 2
— 0
E 0.2
a.
a.
0
0
2
N
0
z 0.1
0
TOLUENE.NO-S0 2 IN FILTERED AIR
0 120 240 360 480
TIME (mm)
10
8
E
U
(V)
UJ
4
1
0
>
2
720
600
C?
E
U
‘ U
2
120 240 360 480 600
TIME (mm)
B- 25
-------
RUN DATE: 29—JUN—74
SYSTEM: T0LUENE 502ND U. of M.
0 HUM i r 1
4—
:0
:0
—
1—
B- 26
0 UOL.(PM 3 ’ML)
H
U
B
£
V
0
U
N
N
I
T
R
0-.
G
E
H
0
‘ S
12
0
E
S
0
-------
RUN 55 O TE
22—Mrn’—74
$YE
TEM
TOLUENE. S
O2 NO U. of M.
0 N
UM.(P RT..’ML)
SURF.(
JM 2 ’ML)
0 UCL.(IJM 3 /ML)
4—
N 0
I •-Z
i :o
P -II
E :
N -
0
x :
12
o :
E -
s
1—
TOLUENE = 0.35 PPM
=
N
U
N
B
E
R
-s
R
- C
E
1
3
U
0
L
U
2
E
1
0
s—i
4
2
0
1
B— 27
-------
RLIN 88 O TE 28—AUG—74 SYSTEM
0 NLIM (PART . ‘ML) SURF . ( PM 2 /ML)
TOLUENE 035 PPM
= 0.040 PPM
TOLUE lE. 802 NO
U. of M.
0 VOL
-------
10
0-
CALSPAN _________________________
RUN NO. 7 25 FEBRUARY 1974 IT0I 09 1 LTERED AIR SYSTEM R.H. = 33%;
TOLUENE = 0.35 ppm; 502 = 0.54 ppm
1000
800
600
400
200
0
(V)
E
U
N
E
L u
C.,
C
L I-
en
x io 4
20
E
U
Lu
5
z
c v ,
E
2
cv )
E
Lu
2
- I
0
>
0 120 240 360 480 600 720 840 960 1080 1200
TIME (m m)
B- 29
-------
8-30
-------
(3) HEXENE EXPERIMENTS
B- 31
-------
CALSPAN __________________________
RUN NO. 5 22 FEBRUARY 1974 IHEXENE-1 - NO-FILTERED AIR SYSTEM ] R.H. = 41%; HEXENE-1 0.33 ppm;
NO 0.152 ppm; NO 2 = 0.014 ppm
12
io’
10
a
tUE
E
U C ’,
E
UI
UI
2
4. -I
0
>
E
a.
(d)
0
+
N
0
2 -
2
I-
UI
2
UI
x
UI
I
6
2
TIME (mm)
0 120 240 360 480 600 720 840 960 1080 1200
TIME (mm)
B- 32
-------
RLIN 92 O TE
—SEPT —74
SYSTEM
HE :ENE—
1,110 U of M.
0
HLIM
P R1. / ML)
A
SURF
(IJM 2 /P1L)
0 UOL ( M 3 /ML)
x l
10
9
NB -
U
ti
B ?’ F
E :
-
:E
5—
4—
3
2
0—
u :o
1 —2
r :o
R -U
U
E
N
0 -
‘I
1 —’ —
c i :
E
s
1—
B- 33
H
V
0
R
,0
‘ -‘C
A
R
B
0
H
U
0
L
U
E
-------
CALSPAN _________________________
RUN NO.21 MARCH 8, 1974 IHEXENEl-NO-FILTERED AIR SYSTEM ]
HEXENE-1 = 0.33 ppm; NO = 0.180 ppm; NO 2 = 0.010 ppm
RH = 37%
HEXENE.1.NO IN FILTERED AIR
E
U
E
(Ii
-I
0
720 840 960 1080 1200
3
0 ,
11
0
0.3
TIME (mm)
E 0.2
C ’)
0
0
z
J0.1
0
0 120 240 360 480 600
TIME (mm)
B- 34
-------
B- 35
-------
- 25
io 2
- 20
C .,
E
• ! N
- E 15
U i U i
U
I - ‘°
2
- 5
0 . 0
E 0.2
a.
a.
C l
0
0
2
N
0
20.1
0
CALSPAN
RUN NO. 18 MARCH 6. 1974 HEXENE-1-S0 2 NO SYSTEM
HEXENE .1 = 0.33 ppm; NO = 0.178 ppm; NO 2 = 0.008 ppm; SO 2 = 0.07 ppm
RH = 37%
HEXENE .1•S0 2 -NOFILTERED AIR SYSTEM
12
TIME (mini
Cl
E
U
I;-
E
‘U
-J
0
E
ft
‘ U
2
Ui
x
‘U
I
0.3
0.2
0.1
600
0 120 240 360 480
TIME (mm)
B-36
-------
RLIH 93 DiflE s-SEPT—r4 SYSTErI HE EI1E—1 .502.110
0 NLIM PART - ML)
SURF (IJM 2 ’ML)
HENEP1E—I = 0 35 PPM
= fl 034 PPM
U of M.
Q VOL (1JM 3 ’ML)
11
LI
I . ’
B
E
P
- s
I I
P
-A
- c
E
I
V
0
L
U
E
0
i i
I
I
R
E
i i
4 —
:0
—7
:o
• 11
LE
1
0 —
0
1 %
12
0
E
S
H
V
D
R
-20
- Pc
A
R
B
0
14
B- 37
-------
RLIN 75 0 TE
1JLIL?4
S’i
STEM
HEXENE1
S02
HO U. of M..
fl N
LIM, (PART ‘ML)
SURF (
jJN 2 /IIL)
0
UOL ( JM 3 /ML)
xi
-S
- U
-R
A
-c
-E
1—
0
t ’l
I
T
R
0
C
E
t.1
:0
-7
•0
4 —N
0
x
I
0
E
0
B- 38
ri
E’-
R
4U
0
L
U
M
TIME (jg+2 1INUTE)
-------
0 HUM. PART. ‘ML) t SURF ( jJM 2 4IL) 0 LJOL ( IJM 3 /ML)
RLIN 60 DATE 13—MAY—74 SYSTEM HEXEHE—1.. $02. NO
U. of M.
B
E
R 5
—
-
b -R
:F
-E
4
5 v
0
L
U
E
1
.0
-N
_E
N
I
I
R
0
G
E
H
0
‘Ii
0
E
S
0
3
TIME (10+2 MINUTE)
B- 39
-------
CALSPAN
RUN NO. 20 MARCH 7, 1974 [ HExENE-1-S0 2 SYSTEM HEXENE-1 0.33 ppm; SO 2 = 0.055 ppm RH = 40%
x10
4.
c ;r
I I
z1
0• 0
TIME (mm)
o 60 120 180 240 300
B- 40
-------
(4) M-XYLENE EXPERIMENTS
B- 41
-------
10 —
x10 4 x10 2
8—
(‘r E
E 6 — - 12
8
2
2 u3 4
0— 0
E 0.2
0.
ft
C ,)
0
0
2
N
0
20.1
0
CALSPAN _________________________
RUN NO. 15 MARCH 4. 1974 XYLENE.NOFILTERED AIR SYSTEM
m .XYLENE = 0.34 ppm; NO = 0.150 ppm; NO 2 = 0.014 ppm; RH = 38%
AEROSOL COF CENTRATION 25
20
C,)
E
I U
E
10 w
-i
0
>
5
0
480
0.3
E
0.
0.2 •
‘U
2
‘ U
-i
x
E
0.1
480
0 120 240 360
TIME (mm)
0 120 240 360
TIME (mm)
B-4 2
-------
RLIN 81 D TE
0 NLIr
—JUL—74
1.F RT.’
1L
S STEr1
SLIRF,(
1— rLENE,I 1O U. of M.
IJM
-------
RUN 89 O TE
29 UG74
SYSTEM M
YLEI4E.. NO U. of M
a
NIJM
PART. . ML)
SURF. ( JM 2
‘ML) 0 UOL ( M 3 /ML)
-S
• IJ
•F•
H
•E
1—
0
B- 44
II
I I
t 1
E
I;,
lJ
LI
(1
E
-------
B- 45
-------
50
30
20
10
0’
0 3
Caispan M-XYLE E + NO + S02 SYSTEM
I iN #1/4
AE JS0LCONcENTRATI0N
--- -4 100
L
___i
--V--- - :- -\-F-- SURFACE
:L O
I :
i :
NU ’4BER
1
j - -- - $ V..
-10
G [ MISTRY IY TA
I t I I-- -- - - -- 0 3
0:
- - I -Q r
M-XYLENE
1 . L 0 2
• -
t-r- ---
0
0
60 120 180 2 / 40 J0 360 1420
TIf E (MIN)
>(
251--
20
15 1--
10 -
5
0 2
0 ,1
B-46
-------
RUN 91 D TE G—SEPT—?4
SYSTEM: ri— LENE SO . NO
U. of M.
NUM(P RT.’NL)
t S&JRF.qJ(1 2 ’ML)
0 UOL,(tJM 3 /ML)
M-XYLENE = 0.35 PPM
= 0.046 PPM
11
R
41)
0
L
U
ii
B- 47
-------
RLIN S2 C1ATE
4JUL74
SYSTE 1 1
-------
CALSPAN ______________
RUN NO. 17 MARCH 5, 1974 [ YLENE.SO 2 SVSTEMI XYLENE = 0.34 ppm; SO 2 = 0.07 ppm; RH = 41%
30
io&
25
20
1
E
q
LU 15 E
C ,,
E
2
10 3
LU
-I
5 0
>
0.
xi i
C,)
V
I 1
E
LU
U
U.
(I)
TIME (mm)
B- 49
-------
B-SO
-------
(5) CYCLOHEXENE EXPERIMENTS
B-Si
-------
5
><
3
L
I
( LOHE) NE ÷ NO SYSTEJI
• ------ R # • - - • .4
E JS0LCON ENTRATIc ’1 1!
25 ‘1- r’-ri suRFAcE’1lOO
- F
20 ±
L LI NU ER
15 1
1 I - •“.— i •“ - . -i I
1
10”---”t VOLUME L
L I
5 L 1 20
TINE (MIN)
B-52
.1
I
300
-------
RLIN 95 DATE 1e —SEPT—r4 EXSTEM CVCLOHE> EHE.WJ U of M
0 HUM (Pe RT ‘tilL) SLIRF L’M 2 ’ ML) 0 UOL ‘ iM ML)
CYCLOHE
-------
CALSPAN
RUN NO. 10 28 FEBRUARY 1974 L CYCLOHEXENE-NO-FILTERED AIR SYSTE! 4JR.H . = 38%;
CYCLOHEXENE = 0.33 ppm; NO = 0.138 ppm; NO 2 = 0.026 ppm
NO STIRRING
200
x
160
E 3 120
C ’,
E
LU
w
z
-J
0.
>
40
0—
E —
0. E
a. a.
0
LU
d z
z LU
C; w
0 I
2 0
0.1 -j
0
>-
0
io 2
C’)
E
C)
- S.-
N
E
LU
0
U-
U)
80
TIME (mm)
0
0.2
120
TIME (mm)
B— 54
-------
CYC:L0HE ENE = 0. 35 PPM
- = 0.103 PPM
0 UCiL. < JM 3 iML)
U. of M.
5
J .1 I I I I I
:LTS. OFF
4 1’J
U
L
U
N
2
1
2
B- 55
4
H
‘ I’
D
R
70
R
B
0
N
RUN 94 O TE 9—SEPT—74
0 NUM.(P lRT.’r1L)
S’STErl C CL0HE :EHE NO
-------
RUN 83 OATE
: 6—JUL—74
SY
STEM:
CYCLOHE
ENE.ND U. of M.
0 N
UM.(P RT ’ML)
SURF.(
IJM 2 ’ML)
0
UOL.( M 3 /ML)
CYCLOHE ENE = 0.35 PPM
NO = 0.13 PPM
5
41 )
0
L
Ii
3 E
I 1
LTS. OFF-
I i t , i—
2
X1 i•—
10 : -
9 : -
N8 —s
U -U
i :R 3
B? F
E
R . LC
: E
5 :2
4:
3:
:
2:
1—
:o
—2
:o
N
:.E
1—
i i I I I I I I i i i t I i i i I i i I I
1
I
TIME
c:ie+2 MINUTE)
2
N
I
T
R
0
G
E
N
CI
x
12
0
E
S
•1
4
3
2
U
1
TIME
-------
B- 57
-------
CALSPAN
RUN NO. 9 27 FEBRUARY 19741 CYCLOHEXENENO SO 2 F1LTERED MR SYSTEMIRH = 36%;
CYCLOHEXENE = 0.33 ppm; NO 0.220 ppm; NO 2 0.020 ppm; SO 2 = 0.05 ppm
x10 5 io2
2
( )
{I ) C’)
E E
.9
Lu lu
-I
2 0
>
E
a.
a.
0
0
2
2
E
a.
a.
Lu
2
LU
x
Lu
I
0
-I
C.,
C.,
E
0
NE
U I
U
U-
U,
1•
0 120 240 360 480 600
TIME (mini
0
720 840 960 1080 1200
0 120 240 360 480 600 720 840 960 1080 1200
TIME (mm)
B-58
-------
H
II
1 1
B
E
RUN 9G 0 TE Il—SEPT—?
0 HUM PART ‘ML
-s
• IJ
-R
•F
-
-c
0—
N
I
T
R
0
G
E
H
ig-1
4—
:0
—z
:0
-N
1—
0
x
12
0
E
S
0
SYSTEM 4 CYCL0HE EHE S02d10
SURF (1JM 2 ’ML)
U of M.
0 UOL ( JM 3 /ML
U
0
L
U
‘
E
H
0
R
,0
-‘C
p
R
B
0
N
TIME O+2 MINUTE)
B-59
-------
CALSPAN ___________________________
RUN NO. 13; MARCH 2. 1974 LCVCLOHEXENE-SOZ.FILTERED AIR SYSTEMJ RH =
CYCLOHEXENE = 0.33 ppm; NO = 0.012; NO 2 0.003 ppm; SO 2 0.06 ppm
x
25
20
E
102
t )
20 E
E
‘U
0
>
0 60 120 180 240
TIME (mm)
B—60
-------
NaC1 EXPERIMENTS
B- 61
-------
CALSPAN
B- 62
Time (mm)
0
U
k
C
0
Time (mm)
-------
CALSPAN
XYLEP4E—NO-S0 2 -NaC1 NUCLEI-FILTERED AIR SYSTEM
U
U
0
TTTT
I
:: p u
• H
- .. - .-.—— -- - ,-- 1-
:t I
t-
r
- - - :4 —
—— ————— -—-— — f r
UI I::::IIT:::: r::
-- ---- ---- ----- ---dr- I --
I !. - L : t : : L J:
L
I -- --- 4 I I I -
S : I - E I I E I I : I 1T E I I
- - I IL - 1 f L :t I Ii
--- m - -- -- I
1080 12
0 120 360 4S( bOO 720 840
Tj e (un)
XYLENE-NO-S0 2 -NaC1 NUCLEI IN FILTERED AIR.
z
2
4 : .
T
if
1:
1 -
if
r
e44 a
4
UkU- 1i ’ :::
!4 .
— -. r.
H:__
— -.- •4. 1:
1
d
-
-
I
1:I -
‘T’
4$!.
- ..
-k
-
;
- - - -H-- fL - -
-H H -4-
- - - - - -
L I
0.2
-
r
N
.
0.3
- : : I in
R 1 1 i
- - - - - - -i -
I_
‘
NL .
1-wiI
___L-
-- -•- -—
i 1-
L_. J
iw-
0 120 240
360
,
T
480 600
Time (un)
B- 63
720 840 960 1080 1200
-------
CALSPAN
CYCLOHEXENE-N0-S0 2 -NaCJ NUCLEI-FILTERED AIR SYSTEM
th14 17 .
NO
a
21.
N . O Oi .p .
T ’
r
-
-
‘
- ------
-—
-r-
‘
---- -— --r---
— - —-i--
J
— —
T
--f—
-
- . 1.
- — —
— —
— .--
T ’
H
- 1 - - I -
T 1
—-i
Time (mm)
CYCLOIIEXENE-NO-S0 2 -NaC1 NUCLEI IN FILTERED AIR
-h
-
-
0
Surf & i
ec
•.j
H
.:
- . !
_ tiiLLt
420 480
L T
T.T
T .T .
JI
T
E
I E i
II
iII.
J TIT T TI
IT
*
. 4
L L4
-
4
.N.S p U0
x L - . ..
OL
L. L
- -
£ Lt
.
j
-
-
-
-k
.
---
-
-r ---
r--t-- - L
- _
4-
-
..
.
- -
1 irt l
-4-
1•
1I
I HI
tv
I !
i!
II
III
T
I T TJ T
i ± II
6 4
Cy lobex ne * •3 p
-- _ O_
So
40
U
30
U
0
>
20
1O
c
: f t++
8
0
3
C
0.
0
C
0
P.’
0.:
c
0
z
oj
-j
L
360
Time (nUn)
B-64
480
------- |