EPA-460/3-76-005
April 1975
OZONE AND PARTICULATE
FORMATION
IN PHOTOCHEMICAL
REACTIONS
l.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Mobile Source Air Pollution Control
Emission Control Technology Division
Ann Arbor. Michigan 18105
-------
EPA-460/3-76-005
OZONE AND PARTICULATE
FORMATION
IN PHOTOCHEMICAL
REACTIONS
Chi-Hunp Shcn and George S. Springer
The University of Michigan
Ann Arbor, Michigan 48104
Grant No. R-801476
EPA Project Officer: Gordon J. Kennedy
Prepared for
l'.S. EMIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Manpemcnt
Office of Mobile Source Air Pollution Control
Kmi.-.-ion Control Technology Division
Vim Arbor. Michigan 18101^
Xpril I«»7S
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - as supplies permit - from the
Air Pollution Technical Information Center, Environmental Protection
Agency, Research Triangle Park, North Carolina 27711; or, for a fee,
from the National Technical Information Service, 5285 Port Royal Road,
Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by
The University of Michigan, Ann Arbor, Michigan 48104, in fulfillment
of Grant No. R-801476. The contents of this report are reproduced herein
as received from The University of Michigan. The opinions, findings,
and conclusions expressed are those of the author and not necessarily
those of the Environmental Protection Agency. Mention of company or
product names is not to be considered as an endorsement by the Environmen-
tal Protection Agency.
Publication No. EPA-460/3-76-005
11
-------
ABSTRACT
Experiments were performed to investigate the photochemical forma-
tion of nuclei and aerosols in sulfur dioxide-air mixtures and the photo-
chemical formation of ozone, nuclei, and aerosols in cyclohexene-nitrogen
dioxide-air mixtures. The mixtures were irradiated by ultraviolet fluor-
escent lamps while flowing at a steady speed through a 9.12 m long and
15.2 cm i.d. Pyrex tube. The amounts of ozone, nuclei, and aerosols gen-
erated in the gas-air mixtures were measured at various points along the
tube by a chemiluminescent analyzer, a condensation nuclei counter, and
an aerosol counter, respectively.
Photooxidation experiments with sulfur dioxide-air mixtures were
performed for five different initial sulfur dioxide concentrations (0.2,
0.3, 0.5, 0.7, and 1 ppm) and for four relative humidities (25, 50, 75,
and 100%). The data showed that the maximum number of nuclei produced
is directly proportional to the initial sulfur dioxide concentration and
to the square of the relative humidity. There were no measurable amounts
of aerosols observed during the tests.
Experiments with cyclohexene-nitrogen dioxide-air mixtures were
performed with the initial cyclohexene concentration ranging from 0.5
to 50 ppm and the initial nitrogen dioxide concentration from 0.6 to
10 ppm. The relative humidity was kept constant at 50%. The amount of
ozone produced depended on the initial concentrations of cyclohexene
and nitrogen dioxide as well as on the light intensity. On the basis
of the present measurements and the data obtained by previous investiga-
tors, the maximum amount of ozone was found to be given by the expressions
m
-------
and
Here, k, represents the nitrogen dioxide photodissociation rate constant
in s and is proportional to the light intensity. R is the concentration
ratio of the hydrocarbon to nitrogen dioxide at which the ozone content
is maximum. The value of R was found to be related to the relative
A nr*
hydroxyl rate constant kg by R = 5.1/|kg .
The maximum number of nuclei produced varied directly with the
square root of both the cyclohexene concentration and the nitrogen di-
oxide concentration.
The maximum number of aerosols produced appeared to depend only on
the concentration ratio of cyclohexene to nitrogen dioxide and seemed to
be insensitive to the initial cyclohexene and nitrogen dioxide concen-
trations. The aerosol size distributions were also measured and followed
closely Friedlander's universal self-preserving curve.
-------
TABLE OF CONTENTS
Page
LIST OF FIGURES vi
L 1ST OF TABL ES i x
LIST OF APPENDICES x
CHAPTER
I. INTRODUCTION 1
1.1 Literature Survey 2
1.2 Objective 5
II. EXPERIMENTAL APPARATUS 7
2.t Test Section 7
2,2. Gas Supply System 14
2.3 Measuring Instruments 18
III. EXPERIMENTAL PROCEDURE 28
IV. PHOTOCHEMICAL PARTICULATE FORMATION IN SULFUR
DIOXIDE-AIR MIXTURES 40
V. OZONE AND PARTICULATE FORMATION IN CYCLOHEXENE-
NITROGEN DIOXIDE-AIR MIXTURES 52
5.1 The Effects of Mixture Composition on Ozone
Formation 53
5.2 The effects of Mixture Composition on Particulate
Formation 77
VI. CONCLUDING REMARKS 94
APPENDIX 96
REFERENCES 122
-------
LIST OF FIGURES
Figure Page
2.1 Schematic of experimental apparatus 8
2.2 Schemati c of the sampl i ng probes 9
2.3 Spectral energy of the ultraviolet light as specified
by the manufacturer 13
2.4 Correspondence between the aerosol diameter and the
channel number on the multichannel analyzer 20
2.5 Typical aerosol size distribution in room air and
i n the tunnel 21
2.6 Size distribution of aerosols in room air 25
3.1 Variation of ozone and nuclei contents with time
along the tunnel 29
3.2 Variation of ozone, nuclei, and aerosol contents
with time along the tunnel 30
3.3 Spectral energy of the light inside the tunnel 32
3.4 Ozone content along the tunnel in various nitrogen
di oxi de-ai r mi xtures 35
3.5 Relationship between the photostationary content of
ozone and the initial concentration of nitrogen
dioxide 36
4.1 Number of nuclei formed as a function of irradiation
time in sulfur dioxide-air mixtures 41
4.2 Number of nuclei formed in sulfur dioxide-air mixtures. 43
4.3 Number of nuclei formed in sulfur dioxide-air mixtures. 44
4.4 Number of nuclei formed in sulfur dioxide-air mixtures.
Correlation of the data for various initial sulfur
dioxide concentrations [S02]0 and initial relative
humidities 0 46
4.5 Formation of sulfuric acid in sulfur dioxide-air
mi xtures 49
4.6 Amount of light scattered from sulfur dioxide-air
mi xtures 51
VI
-------
LIST OF FIGURES (continued)
Figure Page
5.1 Typical variations of ozone, nuclei, and aerosol
contents along the tunnel in cyclohexene-nitrogen
dioxide-air mixtures 54
5.2 Ozone content in cyclohexene-nitrogen dioxide-air
mi xtures 58
5.3 Ozone content in cyclohexene-nitrogen dioxide-air
mixtures 59
5.4 Normalized ozone content in cyclohexene-nitrogen dioxide-
air mixtures 64
5.5 Normalized ozone content in hydrocarbon-nitrogen oxides-
ai r mi xtures 65
5.6 Maximum amount of ozone M as a function of the con-
centration ratio R at which M occurs 69
5.7 Maximum amount of ozone as a function of the initial
hydrocarbon concentration 70
5.8 Maximum amount of ozone as a function of the initial
nitrogen oxides concentration 72
5.9 Variation in the concentration ratio R (at which the
ozone content is maximum) with the reactivity kg 75
5.10 Ozone content in cyclohexene-nitrogen dioxide-air
mi xtures 76
5.11 (a) Oxidant content in the atmosphere, and (b) ozone
content in propylene-nitric oxide-air mixtures 78
5.12 Nuclei content in cyclohexene-nitrogen dioxide-air
mixtures 80
5.13 Nuclei content as a function of the initial concentra-
tion ratio in cyclohexene-nitrogen dioxide-air mixture. 82
5.14 Aerosol content in cyclohexene-nitrogen dioxide-air
mixtures 85
5.15 Aerosol content as a function of the initial concen-
tration ratio in cyclohexene-nitrogen dioxide-air
mixtures 86
vii
-------
LIST OF FIGURES (continued)
Figure Page
5.16 Light scattering from aerosols produced in cyclo-
hexene-nitrogen oxides-air mixtures 90
5.17 Aerosol size distribution obtained in the tunnel for
cyclohexene-nitrogen dioxide-air mixtures 92
vm
-------
LIST OF TABLES
Table Page
2.1 Flowrates of the FTowmeters used in the Experiments 16
3.1 Rates of Photodissociation of Nitrogen Dioxide in the
Lower Atmosphere in the Wavelength Range 29° to 430 nm... 38
3.2 Summary of Values of kj used in Differerent Experiments.. 39
4.1 Summary of Experimental Conditions used by Different
Investigators 47
5.1 The Reaction Mechanism Proposed by Eschenroeder and
Martinez [19] 56
A
5.2 Reactivities of Different Hydrocarbons (k = k/kc3H6) 74
-------
LIST OF APPENDICES
Appendix Page
A. LITERATURE SURVEY 96
B. CERTIFIED CONCENTRATION OF TEST GASES 106
C. ESTIMATION OF PARTICLE LOSSES DUE TO SETTLING, COAGULA-
TION, AND DIFFUSION 107
C.I Loss due to Settling 107
C.2 Loss due to Coagulation 109
C.3 Loss due to Diffusion 109
D. TABULATION OF EXPERIMENTAL DATA 112
D.I Ozone Content along the Tunnel 112
D.2. Nuclei and Aerosol Contents along the Tunnel 116
E. ESTIMATED PEAK VALUES OF OZONE, NUCLEI.. AND AEROSOLS
FOR DIFFERENT INITIAL MIXTURE COMPOSITIONS 120
-------
-1-
CHAPTER I
INTRODUCTION
It is well recognized that ozone and particulate matter contribute
to the problem of air pollution. Excessive amounts of ozone in the atmos-
phere accelerate the degradation processes of elastomers and textiles, and
adversely affect the growth of plants [69]. High concentrations of par-
ticulate matter in the atmosphere reduce visibility [66]. Ozone and par-
ticulate matter are also major constituents of smog [66,69]. In order
to determine the amounts of ozone and particulates in the atmosphere,
and to establish effective means for controlling the production of these
pollutants, it is necessary to understand in detail the processes by
which ozone and particulates are generated.. This study is addressed to
one aspect of the problem, namely, the formation of ozone and particu-
lates due to photochemical reactions.
Ozone is formed photochemically in the atmosphere when a mixture
of nitrogen oxides (nitrogen dioxide or nitric oxide) and certain types
of hydrocarbons is irradiated by sunlight [3,8,31,53,69]. Nitrogen oxides
and hydrocarbons are usually introduced into the atmosphere by automo-
tivevehicles [68,70]. Hence, large amounts of ozone are produced in
areas where the traffic is heavy and where the weather conditions are
such that there is much sunshine but little chance for the automobile
exhaust to disperse rapidly.
Particulates are formed photochemically when air containing sulfur
dioxide, or nitrogen oxides and hydrocarbons is irradiated [3,8,28,53,63,66].
In reference [66], HEW has suggested that "particle" or "particu-
late" be defined as any dispersed matter, solid or liquid (except
uncombined water), in which the individual aggregates are larger
than single molecules ( 0.2 n'"dia.) but smaller than about 500 yw-m
in diameter. '
-------
-2-
Since automotive vehicles may emit considerable quantities of sulfur
dioxide, nitrogen oxides and hydrocarbons, photochemical particulates
may form whenever there is heavy traffic. Sulfur dioxide is also in-
troduced into the atmosphere by power plants burning high sulfur content
fuels [67]. In the vicinity of such plants, the amount of particulates
produced due to photooxidation of sulfur dioxide may also be significant.
In view of the importance of the ozone and the particulates formed
by photochemical reactions, the mechanisms by which they are produced
need to be understood and the amounts generated need to he kno/\m. Al-
though this probleir has been investigated widely, many facets of the
problem still remain unclear. The overall objective of this study was
to perform experiments with the aim of obtaining information which will
contribute to the general understanding of the photochemical formation
of ozone and particulates, and provide quantitative data on the amounts
of ozone and particulate matter formed. The specific goals of this
study are listed at the end of this section. First, however, a brief
survey is given of the previous relevant investigations. A detailed
survey of past research efforts is presented in Appendix A.
1.1 Literature Survey
In most previous experiments studying photochemical reactions, the
procedure normally employed was the following. The gas-air mixture to
be studied was introduced into a chamber. The chamber was then irrad-
iated by sunlight or artificial ultraviolet light, and selected reaction
products were monitored.
The production of ozone during the photooxidation of hydrocarbon-
nitrogen oxides-air mixtures has been studied by several investigators
-------
-3-
[2,5,6,7,10,25,31,41,50,51,58]. The data are generally presented in one
of three ways: 1) the initial ozone formation rate [25], 2) the ozone
dosage (i.e. the integral of the ozone content versus time curve [6,7,10],
and 3) the amount of ozone produced [2,5,31,41,50,51]. Of these three
types of measurements, the last one is of the greatest interest to us,
because only this type of measurement provides data which gives directly
the maximum amount of the ozone generated.
The production of ozone in the presence of nitrogen oxides and
several different types of hydrocarbons has been studied. Ozone mea-
surements with paraffins were reported in references [6,31,50], with ole-
fins in references [2,5,10,25,31,51,58], with aromatics in references
[7,25,31], and with aldehydes in reference [41]. Unfortunately, the in-
formation provided by many of the available data were limited, because
most measurements were made only for a few selected initial concentra-
tions of the hydrocarbon and the nitrogen oxides. Consequently, it is
difficult to extend the applicability of the data to conditions beyond
which they were obtained, to correlate the results of the various ex-
periments, and to draw general conclusions from the existing data. In
order to provide additional insight into the photochemical formation
of ozone, a series of tests would be needed in which the initial con-
centrations of the hydrocarbon and the nitrogen oxides were varied sys-
tematically over a wide range.
The particulate formation in hydrocarbon-nitrogen oxides-air mix-
tures and hydrocarbon-nitrogen oxides-sulfur dioxide-air mixtures has
also been investigated [27,29,47,49,53,59]. In most of the previous
particulate experiments, the initial concentrations of the hydrocarbon
and the nitrogen oxides were kept constant and only the type of hydro-
-------
-4-
carbon was varied. Experiments have been performed with the following
types of hydrocarbons: paraffins [29,47,49], olefins [27,29,47,49,53,59],
cyclo-olefins [28,29,47,49,59], diolefins [29,47,49,53], and aromatics
[28,29,49]. The results of these investigations indicate that certain
types of hydrocarbons produce particles under certain experimental con-
ditions. The results, however, do not provide information for those
concentrations which have not been tested. In order to obtain a better
understanding of the particulate formation process, a comprehensive ex-
periment would be necessary in which the initial concentrations of the
hydrocarbon and the nitrogen oxides were varied over a range.
Several experiments have been performed in the past for measuring
particulates in sulfur dioxide-air mixtures [13,14,17,24,40,48,49].
The major goal of these experiments was to determine the effects of the
initial sulfur dioxide concentration and the relative humidity on the
number of particles formed. Most results indicate that the number of
particles formed increased with increasing amounts of initial sulfur
dioxide concentration. However, the relationship between the amount of
particulate matter and the initial sulfur dioxide concentration is not
yet well established. The effect of humidity on the amount of particu-
late matter formed is also subject to debate. For example, Gerhard and
Johnstone [24] and Quon et. al. [48] found that the humidity had no
effect on the amount of particles formed; Renzetti and Doyle [49], demon-
strated that humidity influenced significantly the formation of par-
ticles. Thus, a need for further study of the effects of the initial
sulfur dioxide concentration and the relative humidity on particulate
formation is evident.
-------
-5-
1.2 Objective
The foregoing survey indicates the need for further work on the
photochemical formation of ozone and particulates in mixtures of hydro-
carbon-nitrogen oxides-air mixtures in which the initial concentrations
of the hydrocarbon and the nitrogen oxides are varied over a wide range.
Particulate formation in sulfur dioxide-air mixtures also require fur-
ther investigation. Therefore, the major goals of this study were:
1) to design and build an apparatus suitable for studying photochemical
reactions in gas mixtures, and 2) to determine the amounts of ozone and
particulates formed in selected mixtures. In particular, experiments
were performed to investigate the following problems.
1) Pa-rbiculate formation in sulfur dioxide-air mixtures and the
evaluation of the amount of particulates produced as a function of
a) the length of time the mixture is irradiated, b) the initial sulfur
dioxide concentration, and c) the relative humidity.
2) Ozone formation and particulate formation in a hydrocrabon-
nitrogen dioxide-air mixture, and the determination of the amounts of
ozone and particulates formed for different initial concentrations of
the hydrocarbon and the nitrogen dioxide. In the present study,
cyclohexene was selected as the hydrocarbon because it has been shown
to produce large quantities of particulates in photochemical reactions
[29,47,49,59].
The ozone concentration was measured by a chemiluminescent analyzer.
The particles were examined in two different categories. The first
category contained those particles whose diameters were between 2 nm
and 0.1 yu m. These particles were designated as condensation nuclei and
were measured by a condensation nuclei counter. In the second category
-------
-6-
were those particles whose diameters were greater than about 0.3^-m.
These particles were designated as aerosols, and were measured by a
Royco Particle Counter. The number concentrations (no. per cur) of
the condensation nuclei and of the aerosols, as well as the size dis-
tribution of the aerosols were all measured in the present experiment.
-------
-7-
CHAPTER II
EXPERIMENTAL APPARATUS
The experiment was performed in a flow type apparatus which con-
sisted of: 1) the test section, 2) the gas supply system, and 3) various
measuring instruments (Fig. 2.1).
2.1 Test Section
The test section had three major components, the irradiation tunnel,
the gas mixing chamber, and the lighting system (Fig. 2.1).
The photochemical reactions were generated in the irradiation
tunnel. This tunnel was made from six identical Pyrex glass tubes joined
together. The tubes were arranged horizontally in a straight line.
Pyrex glass was selected because of its excellent chemical resistance
and constant ultraviolet transmissivity [39,58]. Each tube was 1.52 m
long and 15.2 cm i.d. with a wall thickness of 0.84 cm. The tubes were
connected by compression couplings lined with Teflon gaskets. The six
sections together were 9.12 m long and had a total volume of approxi-
mately 0.166 m3. Two openings were made in the wall of each tube,
through which the sampling probes were inserted. These openings were
located on top of the irradiation tunnel. The first opening was 23 cm
from the entrance of the tunnel. The distance between any two successive
openings was 76 cm. A 5 cm long standard 10/30 outer ground glass joint
was welded to each opening (Fig. 2.2).
Two types of sampling probes were used in the experiments, one for
sampling particulates and one for sampling ozone (Fig. 2.2). The probes
for the particulates were made from 0.636 cm o.d. stainless steel
tubing with a wall thickness of 0.075 cm. Each probe was 30 cm long.
-------
Mixer
Fluorescent
Lamps t-Photodiode
Manometer [
Flowmeters
-*To Exhaust ^- Shell
i—Filter
^- Humidifier
To
Exhaust
• Thermocouples
o Sampling Holes
Test Gases
Filter
— Air Inlet
Compressor
Storage Tank
00
Fig.
-------
(a)
(b)
Stoinless
Steel Probe
Sampling
Holes
Pyrex
Tube
Tygon Connecting
Tube
Outer
Shell
Teflon
Joint
Polyethylene
Connecting
Tube
Outer
Shell
Teflon
Joint
Needle
Pyrex
Tube
-------
-10-
One end of each probe was sealed. Twelve 0.223 cm diameter holes were
drilled into each probe. The holes were arranged in pairs, six holes
being on one side of the probe and six on the opposite side. The first
two holes were 1.27 cm from the sealed tip of the probe. The distance
between each subsequent pair of holes was 2.54 cm. The probes were in-
serted into the irradiation tunnel through the ground glass openings,
and were held in place by standard 10/18 Teflon adaptor joints fitted
with Viton-A 0-rings. The probes were mounted vertically so that the
closed tips were touching the bottom of the tunnel. The axes of the
holes were normal tj the direction of the flow, i.e., the holes faced
the walls of the tunnel. The open end of each probe was outside the
tunnel and was connected to a straight bore Teflon stopcock by a 15 cm
long tygon tube.
The ozone sampling probes were 26 gauge (0.02 cm i.d. and 1.27 cm
2
long) hypodermic needles. The needles were connected to 1 cm nominal
capacity syringes (9 cm long and 0.48 cm i.d.). The stems of the sy-
ringes were mounted in the glass joints using sv.andard 10/18 Teflon
adaptors fitted with Viton-A 0-rings. The needles were placed ver-
tically, with the openings about 4 cm deep inside the irradiation tunnel
The plungers inside the syringes were removed, and one end of a 60 cm
long polyethylene intramedic tubing (0.376 cm i.d.) was inserted into
each syringe. The other end of the tubing was either connected dir-
ectly to the ozone analyzer (during measurements) or was sealed with
a C-clamp.
The mixing chamber, made from a 30.4 cm long, 15.2 cm i.d. and
15.46 cm o.d. stainless steel tube, was connected to the upstream end
of the irradiation tunnel (Fig. 2.1). The upstream end of the chamber
-------
-11-
was covered by a stainless, steel plate fitted with three 0.636 cm Swage-
lok connectors. One of the connectors was connected to the gas supply
system and one to a mercury manometer. The third connector served as a
thermocouple feedthrough for a copper-constantan thermocouple placed
about 2 cm inside the mixing chamber.
The gas-air mixture inside the chamber was stirred by a 12.7 cm
diameter fan rotating at 500 rpm. The fan was located 6 cm from the
cover plate with its axis coinciding with the axis of the chamber. The
fan was driven by a variable speed motor placed outside the mixing
chamber and connected to the fan shaft by a rotary feedthrough.
Six stainless steel screens, serving as flow straighteners, were
placed downstream of the fan. These screens were made of 0.08 cm thick
perforated plates. The perforations were 0.318 cm diameter holes ar-
ranged uniformly in square arrays 0.48 cm apart from each other. The
plates were cut into 15.2 cm diameter circles, and were fitted inside
the mixing chamber so that the holes faced the flow. The screens were
placed side by side with 2.5 cm between each screen. The last screen
was at the end of the chamber.
The downstream end of the irradiation tunnel was connected to a
15.2 cm long hollow stainless steel cone (15.2 cm i.d. at upstream end,
10.2 cm i.d. at downstream end and 0.13 cm wall thickness). The down-
stream end of the cone was covered by a stainless steel plate, which
had three openings in it. The first was 2.54 cm in diameter and was
connected either to an exhaust vent or a vacuum pump. The second served
as a feedthrough for a copper-constantan thermocouple probe placed 5 cm
inside the tunnel. The third was used to insert a light intensity de-
tector into the tunnel.
-------
-12-
The ultraviolet radiation was produced by forty-eight F40BL (General
Electric) fluorescent lamps. The initial spectral distribution of the
lamps, provided by the manufacturer, is shown in Fig. 2.3. Each lamp
was 1.22 m long and 3.8 cm in diameter. The lamps were arranged in six
clusters along the tunnel, each cluster illuminating one section of the
tunnel. Each cluster contained eight lamps placed around the tunnel
systematically on a 31.7 cm diameter circle. The lamps in each cluster
were mounted on two semi-circular (18.5 cm radius) galvanized metal
shells. The insides of the shells were covered with aluminum foils.
The shells thus se;ved both as light fixtures and reflectors. The shells
also provided ducting for the cooling air as explained below. The
shells were hinged together and could be opened to install the lamps
and the sampling probes. There were small holes along the upper edges
of the shells to allow the sampling probes to pass through.
The ultraviolet lamps generated a large amount of heat. To mini-
mize the temperature rise of the gas-air mixture inside the tunnel [9],
room air was blown through the annul us between the irradiation tunnel
and the sheTT. This air was blown by a 45 cm fan in the direction oppo-
site to the flow inside the tunnel. The cooling air flowed along the
entire length of the tunnel and was then discharged into an exhaust
vent. The temperature of the cooling air was monitored by four copper-
constantan thermocouples placed inside the annulus at equal intervals
along the tunnel. The maximum temperature rise of the cooling air was
7°C. The temperature rise of the gas-air mixture in the tunnel (i.e.
the temperature difference between the gas-air mixture in the mixing
chamber and at the exhaust port) was always less than 4°C.
-------
-13-
1.5
E
c
O
~ i.o
o
LJ
<0.5
I-
o
UJ
0.
n
300
400 500
WAVELENGTH, nm
600
Fig. 2.3. Spectral energy of the ultraviolet light as specified
by the manufacturer (General Electric Fluorescent Lamp
F40BL)
-------
-14-
2.2 Gas Supply System
The air used in the experiments was supplied from an air-conditioned
room kept at a constant temperature (20°C) and at a constant relative
humidity (75%). The air was compressed by an oil-free, diaphragm type
compressor (Sears-Model No. 15122) into a 0.03 m3 storage tank. The air
2 2
pressure inside the tank was maintained between 110 kN m and 165 kN m
gauge pressures. The air stream from the tank passed through four traps
connected in series. The first trap, consisting of 0.45 kg of activated
carbon in a Plexiglas container, was installed to remove tract; gases pre-
sent in the air. The second trap was a 1.5 m long, 1.9 cm i.d. tygon
tubing filled with anhydrous calcium chloride. The third trap was 0.45kg
of "indicating" anhydrous calcium sulfate packed in a glass jar. These
latter two traps were intended to absorb the water vapor in the air
stream. A change in color in the calcium sulfate from blue to pink in-
dicated that the calcium chloride had been saturated and needed replace-
ment. The fourth trap was a Gel man type A glass fiber filter for re-
moving atmospheric aerosols from the air stream. After these four stages
of purification, the air was branched into two streams. One stream,
designated as "dry air", passed through a flowmeter and a needle valve.
The other stream, which was to be saturated with water vapor, passed
through a flowmeter, a needle valve, and a humidif;er. The humidifier
was made from a 19 cm high and 12.7 cm i.d. cylindrical Plexiglas chamber.
The chamber was packed three-quarters full with 0.4 cm diameter glass
spheres and filled with double-distilled water so that the water level
was about 1 cm above the glass spheres. The temperature of the humidi-
fier was kept constant (~'250C) by electrical heating tapes surrounded
-------
-15-
by thermal insulation. The air was bubbled through the humidifier. The
air, thus saturated with water vapor (100% relative humidity), was led
through a water trap consisting of a 10 cm high and 4 cm i.d. cylindrical
Plexiglas chamber. After passing through the trap and a needle valve,
the saturated air was reunited with the dry air stream. The relative
humidity of the test air was controlled by adjusting the flowrates of the
two streams of air. For example, equal flowrates of the dry and sat-
urated stream would result in an air of 50% relative humidity. The air
stream was led through a 0.636 cm o.d., 0.486 cm i.d., and 60 cm long
stainless steel tube until it was mixed with the test gases.
The test gases were supplied as special blends and were pre-diluted
with pure dry nitrogen to approximately 200 ppm by the manufacturers.
Nitrogen dioxide was obtained from the Air Products and Chemicals Com-
pany. Sulfur dioxide and cyclohexene were supplied by the Matheson Gas
Company. The composition of each gas was individually analyzed and
certified by the supplier (see Appendix B). The supply line of the
test gases were made of 0.636 cm o.d. and 0.486 cm i.d. stainless steel
tubing. The flowrates of each gas were regulated by stainless steel
needle valves and were measured by two flowmeters connected in parallel.
These flowmeters operated at different ranges of flowrates (see Table
2.1), allowing accurate measurements of the flowrates (and consequently
the concentration of the test gases) over a wide range of values. After
the flowmeters, the gas lines were joined with the line containing the
humidity-controlled air. The entire gas-air mixture then passed through
another Gelman type A glass fiber filter and a three-way stainless
steel ball valve. This valve was used to admit the mixture either
directly into the mixing chamber or into an exhaust vent.
-------
-16-
Table 2.1 Flowrates of the Flowmeters Used in the Experiments
(all flowmeters were manufactured by Kontes Glass Company)
Flowmeter
Gas
Dry Air
Moist Air
Sulfur Dioxide
Nitrogen Dioxide
Cyclohexene
Size No.
1221
1221
1123
1123
1423
1123
1423
Flowrate, cnrs
12-130
12-130
0.08-1.75
0.08-1.75
0.33-17.9
0.08-1.75
0.33-17.9
-------
-17-
It is noted here that the lengths of the gas supply lines from the
tanks to the flowmeters were about 4 m, from the flowmeters to the con-
nection with the air 0.6 m, and from this connection to the mixing cham-
ber 1.5 m.
After the apparatus was assembled, the downstream end of the tunnel
was connected to a mechanical vacuum pump, and the entire system was
tested for leakage. The lowest pressure obtained in the system (in-
cluding the gas lines, the mixing chamber, and the irradiation tunnel)
2 21
was 66 N m . The leak rate of the system was 3 mN m s .
During the experiments, the tunnel was not connected to the vacuum
pump, but to an exhaust vent. The pressure in the vent was atmospheric.
The pressure in the air supply line ahead of the traps was kept at 35
2
kN m gauge. The pressure decreased considerably along the supply line
till it became practically atmospheric in the mixing chamber, as indi-
cated by the mercury manometer. The pressure drop inside the tunnel
2
was less than 0.7 mN m , so that the pressure throughout the tunnel
was practically atmospheric during the tests.
The flowrate in the tunnel could be varied from 46 cnrs to
184 crrr^s . The average velocities corresponding to these flowrates
were 0.26 cm s and 1.04 cm s"1. The Reynolds numbers, based on these
velocities, were approximately 24 and 96, respectively. At such low
•Reynolds numbers the flow was expected to be laminar in the tunnel.
An estimate of the number of particles lost along the tunnel due
to various mechanisms is presented in Appendix C. It was found that,
in general, the loss of particles in the tunnel was negligible except
when the particle diameter was either smaller than 0.01 M m or greater
than 1 u m.
-------
-18-
2.3 Measuring Instruments
In this investigation, measurements of the following parameters were
of major concern: nuclei concentration, aerosol concentration, ozone
concentration, total light intensity, and spectral distribution of the
light.
The concentration of the condensation nuclei was measured by a
"Small-Particle Detector" manufactured by Gardner Associates, Inc. In
this instrument, the sample was drawn into a chamber lined with wet
blotting paper. The pressure was then suddenly reduced in the chamber,
resulting in water condensing on the nuclei present in the sample. A
light beam was directed through the sample, and the decrease in light
intensity (wftfch is proportional to the number of nuclei) was measured
by a photomultiplier. The output of the photomultiplier was displayed
on a meter. The reading on this meter, together with a calibration chart
supplied by the manufacturer, gave the number of nuclei per cm . The
instrument could measure nuclei whose diameters were between 2 nm and
0.1 yum. The measurable concentration range was from 200 to 1x10
nuclei cm~ . The fluctuation of the instrument was within -10% when
the nuclei count was above 1000 per cm3. The fluctuation became higher
(±20%) when the nuclei count was smaller than 1000 per cm3. The cali-
bration chart provided by the manufacturer is accurate within -20%.
Thus the maximum percentage error in the nuclei measurement was about
- 40%. All condensation nuclei measurements presented in sections IV
and V were the averages of at least three readings.
The concentrations of the aerosols (i.e. particles having diameters
greater than 0.3 um) were measured by a Royco Particle Counter (Model
225) coupled with a Royco plug-in unit (Module 507). This instrument
-------
-19-
operated as follows: The sample was drawn into a chamber and a light
was passed through it. The light scattered in the near-forward direc-
tion by the aerosols in the sample was measured by a photomultiplier.
The output of this photomultiplier was calibrated by the manufacturer
to indicate directly the number of aerosols in a certain size range.
Although the unit could be used to measure the aerosol content in five
different size ranges, in the present experiment, it was always set to
measure the concentration of all the aerosols larger than 0.3 urn in
diameter. The measured particle concentration was displayed digitally
and was also recorded by a five-digit printer.
In order to measure the size distribution of the aerosols, a multi-
channel (128 channels) analyzer (Nuclear Data Inc., Model ND 555) was
attached to the Royco Particle Counter. The output of the Royco Instru-
ments was fed directly into the analyzer with each channel corresponding
to a given particle size. The particle size corresponding to a parti-
cular channel was obtained by comparing the particle count on the
Royco Instrument with the particle count given by the analyzer. This
calibration was performed at the five particle size ranges available on
the Royco Instruments (particle diameter greater than 0.3, 0.7, 1.4, 3,
and 5yum) (Fig. 2.4). The calibration could not be done accurately for
particles between 0.7 and 1.4 Mm because, in this size range, the output
of the Royco Instruments varied with the particle size in a complex
manner.
During the test, the output of the analyzer was displayed on a
Tektronix oscilloscope (type 565 Dual-Beam). A typical result obtained
with room air is shown in Fig. 2.5. Each bright dot in the picture
-------
-20-
10.0
5.0
o:
LJ
h-
LJ
5
O
CO
o
cr
UJ
r 2.0
1.0
0.5
0.2
I
1
1
20 40 60 80 100
CHANNEL NUMBER
120
Fie. .-.4. Corresponaence between the aerosol diameter and the channel
number on tne n.ultichannel analyzer. O calibration data,
—- fit to data
-------
-21-
(b)
Fig. 2.5. Typical aerosol size distributions in a) room air and b) in
the tunnel. Vertical axis represents the voltage output from
the multichannel analyzer (0.1 volt per division): Voltage is
proportional to the number of particles (eq. 2.1). Horizontal
•axis represents channel number and, correspondingly, particle
size
-------
-22-
corresponds to a channel, with the first dot on the right hand side
corresponding to channel #128. The correspondence between the channel
number and the aerosol size is given in Fig. 2.4. The vertical axis in
Fig. 2.5 has the units of volts and is proportional to the number of
aerosols counted in a particular channel (i.e. in a given size range)
during a given period of time. The relationship between the voltage
and the number of aerosols counted was determined by a selector switch
on the multichannel analyzer. The selector switch has 8 positions.
At positions 10 , 10 , 10 , or 10 the output of 1 volt per channel
corresponded to 10 , 10 , 10 , or 10 counts, respectively. When the
selector switch was at positions 4x10 , 4x10^, or 4x10 , the output of
23 4
2.5 volts per channel corresponded to 4x10 , 4x10 , or 4x10 counts,
respectively. For these seven selector switch settings, the voltage
was linearly proportional to the number of aerosols counted. There was
an eighth setting in which the output voltage for each channel, v, was
related to the number of aerosols counted, n, by the expression
v - c log1Qn + VQ (2.1)
where c and v were undetermined constants. These constants were evalu-
ated by sampling a mixture, and by measuring the number of aerosols in
a particular channel using one of the seven settings for which n was
linearly proportional to v. The selector switch was then turned to the
logarithmic scale and the voltage was measured for the same channel.
Repeating this procedure twice, two sets of v and n values for the log-
arithmic setting were obtained. Substituting these values into eq. 2.1,
c and v could be calculated. In practice, the procedure was repeated
for several sets of v and n values in order to obtain representative
-------
-23-
average values for c and v . The c and v values were found to be 0.14
and 0.06 volts, respectively.
From the results displayed in pictures such as Fig. 2.5, the number
of aerosols per channel could be determined to within 20% accuracy. How-
ever, the inaccuracy in the aerosol results (both the size distribution
and the total number count) was considerably larger due to the statis-
tical nature of the measurement. The actual number of aerosols fluc-
tuated and this was manifested in the fluctuations in the outputs of the
Royco Particle Counter and the multichannel analyzer. Characteristically,
the fluctuation was of the order of ±30%. All the results presented in
Section V for the total number of aerosol count and for the size dis-
tribution were the averages of at least three measurements.
The Royco Particle Counter could be operated at sampling rates of
ol O 1
47 cm s and 4.7 cm s . The lower sampling rate (which was about
one tenth of the flowrate in the tunnel) was used in the experiments.
At this flowrate the lowest and highest detectable aerosol concentrations
o 3
were 0 and 3.5x10 particles cm , respectively. It is noted here, that
neither the-Small-Particle Detector nor the Royco Particle Counter was
calibrated in our laboratory. The results reported were based on the
calibrations provided by the maunfacturers.
In order to test the instruments, the size distribution of the at-
mospheric aerosols in the room were determined. The data were presented
in terms of the dimensionless parameters
A N/Ar
(2.3)
N/(V/N)1/3
-------
-24-
AN is the number per unit volume of aerosols whose radii lie within the
range of r and r + Ar. N is the total number of the aerosols per unit
volume
(2.4)
and V is the total aerosol volume per unit sample volume
N (2.5)
The parameters^ and •& were used in presenting the results because, as
has been shown by Friedlander [21], and by Liu and Whitby [45], with
the use of these parameters, all the data reduce to a common curve.
Moreover, it" toss- been demonstrated both experimentally [11,22,38,45] and
analytically [45] that in the size range where the aerosol is in a state
of dynamic equilibrium (i.e. where the rate of gain of aerosols due to
coagulation is equal to the rate of loss of aerosols due to coagulation,
diffusion and sedimentation) \{r varies inversely with the fourth power
of n , i.e.
(2.6)
The results of the measurements are given in Fig. 2.6. The data
follow quite closely a line whose slope equals -4. The present results
also fall within the size range observed by other investigators [11,22,
38,45].
The ozone content was measured by a Thermo Electron Corporation
Model 12A Chemiluminescent Analyzer. This instrument operated on the
principle that light was produced during the reaction of ozone with
nitric oxide [56], Therefore, the reaction
-------
-25-
10
o
10°
ID'1
-2
10
.o3
•€
y \
v\
•\
" \ \ "
\ \
\\
0 1
\ -\
\ • \
\ v \
\B
0 \
\-. \
__ Line of \ V • \
\ \ %
Slope -4— — Ji \ ° \
\ * 0 \
\ \ .-\
\-\
33 O.I 1.0 10
r
T\~ ». .,*m
/ v V'**
IN")
Fig. £.b. Size distribution of aerosols in room air. o present data,
range of data from Cartwright et. al. [H], Friedlanrier
and Pasceri [22], Junge [38], and Liu and Whitby j.*&J
-------
-26-
NO
was generated inside the instrument and the emitted light was measured by
a photomultiplier. The photo current produced by the photomultiplier was
a measure of the ozone content in the sample. A linear, full scale meter
on the instrument indicated directly the ozone concentration in the
sampled gas stream. Eight concentration ranges were provided with the
instrument: 0-0.01, 0-0.025, 0-0.1, 0-0.25, 0-1, 0-10, 0-100, and 0-1000
ppm. Within each of these ranges the ozone concentration could be assessed
to within two percent precision, while the fluctuations in the measure-
ments were less than ±5%.
The total intensity of the ultraviolet light inside the irradiation
tunnel was monitored by a silicon Schottky PIN-5 photodiode (United
Detector Technology, Inc.). This photodiode was 0.915 cm in diameter
and had an active area of 0.04 cm2. The photodiode was cemented onto
one end of a 40 cm long, 0.636 cm o.d. stainless steel tube. The tube
(with the photodiode at its tip) was then inserted about 30 cm deep inside
the irradiation tunnel through one of the Swagelok connectors at the
downstream end of the tunnel. The electrical leads of the photodiode
were led through the tube to the outside of the tunnel. The signal from
the photodiode was processed by an electrical circuit recommended by the
manufacturer, and was displayed on a digital voltmeter. The relation-
ship between the output voltage and the light intensity was not de-
termined directly. Nevertheless, the voltage output indicated any
change in the light intensity in the tunnel, and any possible deteriora-
tion in the intensity of the fluorescent lamps. It is noted, however,
-------
-27-
that a measure of the total light intensity in the tunnel was obtained
as described in Section III.
The spectral distribution of the ultraviolet radiation was measured
with a CGA/McPherson Monochromator and an RCA photomultiplier tube
1P28-A. The light was collected by a 30 cm long, 2.5 cm i.d. and 3 cm
o.d. brass tube which passed through the shell surrounding the tunnel.
One end of the tube was attached to the monochromator. The other end,
machined to the shape of the outer surface of the tunnel, was attached
to the outer wall of the tunnel. All openings were carefully sealed so
that only light from the tunnel could reach the monochromator. The
spectral distribution of the ultraviolet light, as given by the output
of the monochromator, was recorded on a strip chart recorder. The
output was not calibrated to give the value of the light intensity at
different wavelengths. However, the output indicated the magnitudes of
the light intensities and showed any change in these magnitudes. There-
fore, this measurement sufficed in the present experiments where the
only concern was to ascertain that the spectral distribution remained
constant throughout the tests.
-------
-28-
CHAPTER III
EXPERIMENTAL PROCEDURE
The experimental procedures used in studying the photochemical reac-
tions in sulfur dioxide-air and in cyclohexene-nitrogen dioxide-air mix-
tures were similar.
At the start of each experiment, the ultraviolet lamps were turned
on, the flow of the cooling air was started, and 50% relative humidity
air (without any of the test gases) was fed through the system contin-
0 _1
uously at a flowrate of 46 cnr s . The temperatures inside the tunnel
and in the cooling air were recorded. The system required three to four
hours to reach thermal equilibrium. The particulate contents were also
measured along the entire length of the tunnel. This procedure contin-
ued until both the nuclei and aerosol counters indicated zero readings
(i.e., a "clean" system) at every sampling station. This condition was
usually reached in four to seven days. Once the system was thus
"cleaned", the required gas-air mixture was introduced into the tunnel.
After about four hours, both the ozone and particulate contents reached
a constant value ("steady state") in the entire tunnel. This is demon-
strated in Figs. 3.1 and 3.2 where the ozone, nuclei and aerosol con-
tents are shown as a function of time at different sampling stations.
When the conditions had reached steady state, the ozone and particulate
contents were measured, and the gas-air mixture was changed to a new
value.
The composition of the gas-air mixture was not measured along the
tunnel. The mixture composition was known only at the entrance of
the tunnel. Therefore, in the subsequent discussions, the specified
-------
0.6
Stotion 10
Station I
-a.
ro
vo
•8 8 8 a.
Station 3
Station 9
I
I
8 12 16
TIME, HOUR
20
24
Fig. 3.1.
Variation of ozone and nuclei contents with time along the tunnel.
Mixture composition: cyclohexerie 1.25 ppni, nitrogen dioxide 0.5 ppm,
relative humidity 50%. Mean flow velocity 0.255 cm s-1
-------
NUCLEI AND
AEROSOLS OZONE
CONTENT, no cm3 CONTENT, ppm
I.O
1.2
0.8
0.4
0
I03
IOE
1 1 1 1 1 1
o o o
^^~** w o
_ y^ Station 10 _
-If Station 1
f 1 1 1 1 1
1 1 1 1 _ 1
ff " Nuclei \ /
If Station 3 /
_ . , /Stations. __
Aerosol / \
^^o « a O/ a\ P
468
TIME, HOUR
10
12
OJ
o
Fig. 3.
Variation of ozone, nuclei, and aerosol contents with time along the tunnel
Mixture composition: cyclohexene 1U ppm, nitrogen dioxide i; ppn,, relative
humidity bU%. Mean flow velocity 0.255 cm s~l
-------
-31-
test gas concentration refers to the amount of test gas in the air
prior to exposure to the ultraviolet light. This condition is denoted
by the subscript o.
The spectral distribution of the ultraviolet light and the total
light intensity was monitored throughout the tests. The spectral dis-
tribution of the light did not vary regardless of the compositions or
concentrations of" the mixture in the tunnel. Typical spectra taken with
just air in the tunnel (i.e. no photochemical reactions) and with a cyclo-
hexene-nitrogen dioxide-air mixture (in which photochemical reactions
were taking place) are shown in Fig. 3.3.
The total light intensity also remained constant throughout the
tests, as indicated by the steady voltage output of the photodiode.
As mentioned fn Section II, the output of the photodiode was not cali-
brated. Thus, while the output indicated that the intensity remained
constant, it did not provide the actual value of the light intensity
in the tunnel.
The light intensity, J, may be related to the photolysis rate con-
stant (k-j) of nitrogen dioxide by the expression [43]
kl-Z 2.303 £^JA (3-D
where
-------
-32-
O
or —
Ul »
«w ^-
il
ui io
AIR
(50% Relative Humidity)
C6H|0-N02-AIR MIXTURE
(5O%Relative Humidity) I
325 350 375
WAVE LENGTH, nm
400
Ma. 3.3. jpecLral Lnercy of the liciht inside the tunnel.
Top: enertjy in air. BotLon: energy in cycloiiexene-
nitroqen dioxide-air mixture
-------
-33-
The dominant reactions at the initial stage of the photochemical
reactions in a nitrogen dioxide-dry air mixture are [8,43,54,57]
ki (3 2)
N02 + \\V - *-»• NO + 0 l '
k
0 + 02 + M - -- + Og + M (3.3)
k
0 + NO - l - + N02 + 02 (3.4)
where k^ , k2 and k3 are the respective rate constants. The rate equa-
tion for the nitrogen dioxide may be written as
~^~ = -kjNO^ k^M (3'5)
A nitrogen mass balance at any instant gives
(3.6)
where [N02J0 is the initial nitrogen dioxide concentration (i.e. the
concentration supplied to the system). Some time after the start of
the photochemical reaction, the "photostationary state" is reached in
the system, at which the concentrations of nitrogen dioxide, nitric
oxide, ozone and oxygen atom do not change with time, i.e.,
• o
Depending upon the initial nitrogen dioxide concentration, this state
was reached in the tunnel in 120-300 seconds. Equations (3.5) and (3.7)
yield
(3.8)
-------
-34-
where the subscript pss denotes photostationary state. For low initial
nitrogen dioxide concentrations ([NOg] less than 2 ppm) , [N0j
v TjS 5
may be approximated by [03] [54,57]
(3.9)
From eqs. (3.6), (3.8) and (3.9), we obtain
3lpss
l°3lpss - iElo "Npss (3.10)
According to the above equation, a plot of [°3]LS against •[ [N02] -
[03] \ should result in a straight line, the value of k^/k- being
given by the slope of this line. Knowing the slope of the [OslLc
versus {[NQ^L - ^3^DSS} curve and tne value of k3, the parameter k,
can be calculated readily. The value of k3 has been determined by sev-
eral investigators, the reported values of k3 ranging from 0.35 to 1.23
ppm" s [12,20,37,52]. The value k3 = 0.425 ppm s has been
adopted here as recommended by Stedman and Niki [57].
In order to generate the [°3]DSS versus {
curve, the ozone concentration was measured at different points inside
the tunnel using dry air and three different initial nitrogen dioxide
concentrations. Since the flowrate in the tunnel was constant, the
distance along the tunnel could be related to a time scale, as shown
on the top of Fig. 3.4. The maximum values of the ozone concentration
curves were taken as the photostationary state values as these maximum
points satisfy the condition d[03]/dt = 0 (see eq. 3.7). The data thus
generated was used to plot a [03]pSS versus f [N02]Q - [03]DSSJ curve
(Fig. 3.5). From the slope of this plot, and for k3 = 0.425 ppm" s~ ,
the value of k, was found to be 6.33 ks . Typical values of k1 for the
-------
0.16-
10
TIME, MINUTE
20 30
40
23456
DISTANCE ALONG TUNNEL, m
50
2ppm Initial NO? Concentration
CO
in
i
Fig. 3.4. Ozone content along the tunnel in various nitrogen dioxide-air mixtures.
Top axis represents irradiation time
-------
0035
O.O3O
0.025
-36-
CSJ
0.020
o.
O
O.OIO
0.005
0.5
1.0
[N02]o- [03]oss, ppm
I
1.5
'pss
2.0
2.5
Fig. 3.5. Relationship between the photostationary content of
ozone and the initial concentration of nitrogen
dioxide, o data, fit to data
-------
-37-
photodissociation of nitrogen dioxide in the lower atmosphere are pre-
sented in Table 3.1. As can be seen, the k1 values of the present ex-
periment compare well with those which exist in the atmosphere. It is
noted that k-, values obtained in other experimental apparatus range from
0.78 ks"1 to 9.17 ks"1 (Table 3.2). The present k] = 6.33 ks"1 falls in
this range.
-------
-38-
Table 3.1 Rates of Photodissociation of Nitrogen Dioxide in
the Lower Atmosphere in the Wavelength Range 290 to 430 nm
Solar zenith angle Photodissociation Rate
Z, degree k, = £ 2.303 4> <. J. , ks"1
1 X 'A A A
0
20
40
60
80
9.51
9.17
8.02
5.60
1.68
Values of <> from Table A-II, reference [15]
Values of
-------
Table 3.2 Summary of Values of k-. Used in Different Experiments
Volume, nr
0.166
0.15
0.044
0.022
0.141
8.49
1.87
0.09
1.811
9.48
0.09
0.2
Material
Pyrex tube
Teflon bag
Pyrex jar
Pyrex flask
Pyrex tube
Stainless steel chamber
Aluminum, Pyrex, Teflon and
stainless steel removable surfaces
Pyrex flask
Aluminum chamber, Teflon window
Aluminum chamber, poly vinyl fluoride
wi ndow
Mylar and Teflon bags
Pyrex jar
3_m HA CD n^fh miiltinlo K*of 1 or t i nn
in L/uoc pa in multiple i c i ICI*LIUM
infrared cell
Type of Experiment
F
S
S
S
F
S
S
S
S
F
S
F
S
«J
k^ ks'1
6.33
0.78
2.33
3.67
3.83
4.17
5.0
5.33
6.33
6.67
6.67 (Teflon)
5.0 (Mylar)
8.33
Q 17
y • i i
Reference
Present Investigation
Lilian [44]
Stedman & Niki [57]
Laity [42]
Stephens & Price [58]
Groblicki & Nebel [29]
Jaffe & Smith [36]
Holmes et. al . [34]
Dimitriades [16]
Altshuller et. al . [5]
Altshuller & Cohen [1]
Wilson et. al . [65]
TiiP^dav FfiOl
i ucoviujr L^^J
oo
vo
i
S Static type experiment
F Flow type experiment
-------
-40-
CHAPTER IV
PHOTOCHEMICAL PARTICULATE FORMATION IN SOg-AIR MIXTURES
It has been demonstrated by several investigators that participates
form when a mixture of sulfur dioxide and air is exposed to ultraviolet
light [13,17,24,48,49]. However, the number of particulates formed and
the dependence of this number on the initial sulfur dioxide concentra-
tion and the humidity of the air are not well established. The objec-
tives of the present experiments were, therefore, to measure the number
of particulates formed in sulfur dioxide-air mixtures, and to determine
the relationships between the number of particulates formed, the initial
sulfur dioxide concentration and the humidity of the air.
When a sulfur dioxide-air mixture was exposed in the tunnel to
ultraviolet radiation, the number of particulates formed increased with
time, reaching a maximum value. The number of particulates then re-
mained nearly constant for some time at the maximum value (see Fig. 4.1)
before it decreased due to coagulation, diffusion, and deposition on
the walls. Here we were interested in the maximum number of particulates
formed in the mixture. In order to determine the exposure time required
to reach the maximum number of particulates, the number of nuclei and
aerosols were measured at sampling station #1 at different flowrates
3-1 3-1
ranging from 46 cm s to 184 cm s . Since sampling station #1 was
23 cm from the entrance to the tunnel, the exposure times corresponding
to these flowrates ranged from 90 s to 22.7 s. The experiments were
repeated for different sulfur dioxide concentrations and different
humidities (Fig. 4.1). There were no measurable amounts of aerosols in
any of the experiments. There were, however, nuclei formed in the mix-
-------
10
UJ
_i
o
D
UJ
m
I I
-41-
I I
I I I
.Oppm, 50%
0.5ppm,50%
0.2ppm,50%
0.2ppm,25%
Initial S02
Concentration, ppm
Relative Humidity, %
1111 I I
I
0 10 20 30 40 50 60 70 80 90 100
IRRADIATION TIME, sec
Fiq. 4.1. Number of nuclei formed as a function of irradiation
time in sulfur dioxide-air mixtures
-------
-42-
ture. The amount of nuclei formed depended on the exposure time, as
shown in Fig. 4.1. The time required to reach the maximum number of
nuclei increased as both the initial sulfur dioxide concentration and
the humidity in the air decreased.
The lowest initial sulfur dioxide concentration used in the ex-
periment was 0.2 ppm, and the lowest relative humidity was 25%. For
this condition, the tests showed that the time required to form the
maximum number of nuclei was about 90 s. Therefore, to ensure that the
maximum number of nuclei formed was achieved in all the tests, the 90 s
exposure time was used throughout the subsequent experiments. The
number of nuclei per cm given below correspond to the maximum value
obtained in the manner just discussed.
The effects of the initial sulfur dioxide concentration and the
relative humidity on the nuclei formation were measured for five differ-
ent sulfur dioxide concentrations ([S09] = 0.2, 0.3, 0.5, 0.7, and 1 ppir),
£ o
and four different relative humidities ([^]Q = 25, 50, 75, and 100%).
The data are presented in Fig. 4.2, and indicats that the number of nuclei
formed vary linearly with the initial sulfur dioxide concentration in
the air.
The results in Fig. 4.2 were cross plotted in order to obtain dir-
ectly the variation of the number of nuclei with relative humidity. As
shown in Fig. 4.3, the number of nuclei varies with the square of the
relative humidity. Thus, according to the results of Fig. 4.2 and 4.3,
the maximum number of nuclei formed may be expressed as
N = ACS0][ (4.!)
-------
-43-
5x10*
u
•»
Ul
CJ
u.
o
QC
Id
00
I I
oDATA
— FIT TO DATA
(Slope =1)
Relative
Humidity
25%
i i i I
I I
I
O.I 1.0 3.0
INITIAL SULFUR DIOXIDE CONCENTRATION, ppm
Fig. 4.2. number of nuclei formed in sulfur dioxide-air mixtures
-------
-44-
5x10'
UJ
_l
(J
u.
o
a:
UJ
CD
i i i i
i i i i i i i i
Initial S02
Concentration
i i
0.75
0.5
0.25 ppm
I I
10 100
RELATIVE HUMIDITY, %
Fia. 4.3. Number of nuclei formed in sulfur dioxide-air mixtures
-------
-45-
where A is a constant. In order to test further the validity of eq. (4.1)
and to obtain the value of A, all the data were compiled on an N versus
^Fl^' 4-4). The data correlated well with the above
expression for A = 3.33x10 no. cm ppm . Thus, the maximum number of
nuclei is given by the expression
where [S02]Q is in ppm and []_ is in percent.
It should be noted that eq. (4.2) was determined from the data
obtained under specific experimental conditions. As the light inten-
sity inside the tunnel was not varied during these experiments, eq. (4.2)
does not indicate the effect of the light intensity on nuclei production.
However, Groblicki and Nebel [29] have found that the maximum amount of
aerosols produced in a propylene-nitric oxide-sulfur dioxide-air mixture
was not affected by a change in the light intensity. The light intensity
merely affected the time required to produce the maximum number of nuclei.
It is expected, therefore, that a change in the light intensity would
not alter eq. (4.2), but would change the irradiation time required to
produce the maximum number of nuclei.
It would be of interest now to compare the present data with re-
sults of other investigations. Results of direct measurements of the
number of nuclei formed in sulfur dioxide-air mixtures were reported in
references [13,14,40,48,49]. The conditions of these experiments are
listed in Table 4.1. The present results can be compared with those
obtained by Renzetti and Doyle [49] and Kocmond et. al . [40]. The
number of nuclei obtained by Renzetti and Doyle for 0.24 ppm sulfur
45 -3
dioxide at 50% relative humidity ranged from 5x10 to 8x10 nuclei cm .
-------
-46-
io7cr
'S 6
. 10
UJ
Li.
O
cr
LU 5
£ 10
ICT
I I I I I 11 I I I I I I I I
FIT TO DATA
N = 3-33xl06[S02]0[*]|
i i i i i i 111 i i i i i i i i
0.01
O.I
1.0 2.0
Fig. 4.4. Number of nuclei formed in sulfur dioxide-air mixtures.
Correlation of the data for various initial sulfur
dioxide concentrations [SO^j and initial relative
humidities fl 0 °
-------
Table 4.1 Summary of Experimental Conditions Used by Different Investigators
Reference
Volume, m
3
Experimental Apparatus
Material Light Source
Type
Cox and Penkett [13]
Cox [14]
Gerhard and Johnstone [24]
Kocmond et. al. [40]
Quon et. al . [48]
Renzetti and Doyle [49]
0.216
0.005
0.008
0.25
1.7
600
0.014
0.05
Aluminum Chamber
Pyrex Flask
Lucite Chamber
Teflon Bags
Steel Chamber
Saran Bags
Pyrex Flask
Sunlight
Mercury Lamp
Sun Lamp
Fluorescent Lamps
Fluorescent Lamps
Fluorescent Lamps
Mercury Lamps
S
S
S
S
S
S
F
S Static type experiment
F Flow type experiment
-------
-48-
For this sulfur dioxide concentration and relative humidity, the number
of nuclei calculated from eq. (4.2) is 2x10 nuclei cm" . Hence, the
value obtained in the present study lies within the spread of Renzetti
and Doyle's data. The amounts of nuclei measured by Kocmond et. al. were
two to four times higher than the values obtained in this investigation.
The reason for this is unclear. The discrepancy between the two re-
sults might possibly be due to the fact that the present data were ob-
tained from flow type experiments while Kocmond et. al. performed their
measurements in static type experiments.
The results of Cox and Penkett [13], Quon et. al. [48], and Cox [14]
cannot be compared with the present data. Cox and Penkett did not re-
port the humidity in their results; Quon et. al. performed measurements
before the maximum amounts of nuclei were produced. The data obtained by
Cox were obtained for much higher initial sulfur dioxide concentrations
(5 to 1000 ppm) than those used in the present study (0.2 to 1 ppm).
Some indirect comparisons with other existing data can be made to
confirm the present results. Gerhard and Johnstone [24] performed static
type experiments in a 0.008 m Lucite chamber irradiated by a sun lamp,
and collected and measured the amount of sulfuric acid formed in sulfur
dioxide-air mixtures. If it is assumed that most of the nuclei formed
in the present study were composed of sulfuric acid, then the number of
nuclei should be proportional to the amount of sulfuric acid collected
by Gerhard and Johnstone. The variation of sulfuric acid with the ini-
tial sulfur dioxide concentration, as reported in reference [24], is
presented in Fig. 4.5. Similar to the number of nuclei formed (see
eq. 4.2), the amount of sulfuric acid collected varies linearly with the
initial sulfur dioxide concentration.
-------
-49-
10
I
E |
o»
E
0.5
o
(T
o.2
O.I
o DATA
— FIT TO DATA (Slope = I)
Irradiation Time I20min
0.05
2 5 10 20 50
INITIAL SULFUR DIOXIDE CONCENTRATION, ppm
Fie. 4.3. Formation of iulfuric acid in sulfur dioxide-air mixtures.
Ucta fron, Gernard and Johnstone [24]
-------
-50-
Renzetti and Doyle [49] measured the amount of light scattered from
a sulfur dioxide-air mixture undergoing photochemical reaction. Their
data are reproduced in Fig. 4.6. Although there is considerable spread
in Renzetti and Doyle's data, the amount of light scattered (which is
proportional to the number of nuclei formed) appears to vary linearly
with the initial sulfur dioxide concentration, further supporting the
validity of eq. (4.2).
The information available on the effect of the relative humidity on
the number of nuclei formed is conflicting. Renzetti and Doy'e observed
that the amount of light scattered increased with relative humidity.
Cox [14] also reported that the number of nuclei formed increased with
humidity. Katz and Gale [39] observed an increase in the conversion
rate of sulfur dioxide to sulfur trioxide (or su'lfuric acid) as the
humidity was increased. These results thus support the present data
which show an increase in nucleation rate as well as the number of nuclei
formed with relative humidity. In contrast to these results, Gerhard
and Johnstone [24] found that the sulfur dioxide conversion rate did not
change with relative humidity. The results of Quon et. al. [48] indicated
that the number of nuclei formed was independent of relative humidity.
The data obtained by Kocmond et. al. [40] did not show any dependence
of the conversion rate of sulfur dioxide on relative humidity. However,
Kocmond et. al. found that the rate of change of the total volume of
the aerosols depended on relative humidity. The effect of humidity on
the nuclei formation in sulfur dioxide-air mixtures would thus merit
further studies.
-------
-01-
100
o:
<
cr
H
CD
o 10
a:
UJ
u
X
Ij
o DATA
FIT TO DATA (Slope«I)
I 1 I I I I I
0.2 1.0 10
INITIAL SULFUR DIOXIDE CONCENTRATION, ppm
Fia. 4.6. Amount of light scattered from sulfur dioxide-air
mixtures. Data from Renzetti and Doyle 1.49]
-------
-52-
CHAPTER V
OZONE AND PARTICULATE" FORMATION IN CgH10-N02-AIR MIXTURES
Photochemical reactions in a cyclohexene-nitrogen dioxide-air mix-
ture depend on several parameters, including a) the initial concentra-
tion of cyclohexene, b) the initial concentration of nitrogen dioxide,
c) the ultraviolet light intensity, and d) the length of time the mix-
ture is irradiated. The major goal of this study was to evaluate the
first two of these effects, namely, the influence of the initial con-
centrations of cyc'ohexene and nitrogen dioxide on the amounts of ozone,
condensation nuclei and aerosols formed. Therefore, throughout the tests,
the light intensity was kept at the constant value of k-,=6.33 ks ,
o ]
and the flowrate was maintained at 46 cm s . All the results dis-
cussed below apply to these conditions.
One might also expect the relative humidity to affect the formation
of ozone, nuclei and aerosols. Bufalini and Altshuller [10] and Stephens
and Price [58] have found, however, that the amount of ozone formed was
not influenced by the relative humidity. The effects of the humidity on
nuclei and aerosol formation have not yet been established. In the pre-
sent study, a relative humidity of 50% was used. According to the re-
sults of Bufalini and Altshuller and of Stepehns and Price, the use of
a constant relative humidity would not affect the amount of ozone formed.
The relative humidity might influence the nuclei and aerosol formation
but this effect was not explored here.
The initial concentrations of cyclohexene and nitrogen dioxide were
varied in a systematic manner. Initially, five concentrations of cyclo-
hexene (0.5, 1, 2, 3, and 5 ppm) and five concentrations of nitrogen
-------
-53-
dioxide (0.6, 1, 2, 3, and 5 ppm) were permutated to provide twenty-five
different mixture composition combinations. As the measurements progressed,
it became evident that additional mixture compositions were needed to pro-
vide more information in certain areas of particular interest. Hence,
measurements were also made with additional mixtures. A complete set of
the experimental conditions and of the data obtained are given in Appen-
dix D. The data from a typical series of tests are presented in Fig.
5.1. The results in this figure show the amounts of ozone, nuclei and
aerosols formed along the irradiation tunnel. Since the flowrate in the
tunnel was constant, the horizontal axis in Fig. 5.1 represents both the
distance along the tunnel and the length of time the mixture has been
irradiated.
As can be seen from Fig. 5.1, the amounts of ozone, nuclei and aero-
sols first increase, reach a peak and then decrease along the tunnel.
Here, we shall be concerned with the "peaks" which represent the maximum
amounts of ozone, nuclei and aerosols formed. Our interest in the maxi-
mum values is motivated by the general concern about the maximum amounts
of ozone, nuclei and aerosols formed in the atmosphere due to photochem-
ical reactions. In all subsequent discussions, the term "amount" was
used to denote the peak values, which were determined by plotting all the
data in a manner similar to those shown in Fig. 5.1, and by estimating
the peaks from these plots. The results thus obtained are tabulated in
Appendix E.
5.1 The Effects of Mixture Composition on Ozone Formation
The typical behavior of ozone along the tunnel is illustrated in
Fi;. 5.1. Although this figure represents only one set of data, similar
-------
TIME, mmult
10 ZO 30 40
90
TIME, minute
IP 20 30 4O 50 60 |08
O 2 4 6 8
DISTANCE ALONG TUNNEL. m
O 2 4 6 6
DISTANCE ALONG TUNNEL, m
(O
Fig. b.l. Typical variations of ozone, nuclei, and aerosol contents along the tunnel in
cyclohexene-nitrogen dioxide-air mixtures. Initial nitrogen dioxide concentration:
3 ppm. Relative humidity: 50%. Top axis represents irradiation time
-------
-55-
trends were observed with other initial concentrations of cyclohexene and
nitrogen idoxide. For all mixture combinations, the amount of ozone along
the tunnel first increased, reached a peak and then decreased. In order
to describe these results quantitatively, it would be necessary to know
in detail the reaction mechanism in the mixture and the corresponding re-
action equations. To date, many different sets of reaction equations have
been proposed [15,19,32,46,61,62], but the validity and accuracy of the
various sets are still of considerable debate. Nevertheless, it is
possible to explain, at least qualitatively, the observed behavior of the
ozone formation by the relatively simple kinetic scheme introduced by
Eschenroeder and Martinez [19]. These investigators suggested twelve
reactions for describing the photochemical processes occurring in hydro-
carbon-nitrogen oxides-air mixtures. The reaction equations proposed by
Eschenroeder and Martinez are listed in Table 5.1 (eqs. 5.1-5.12).
According to eqs. (5.1) and (5.2), ozone and nitric oxide are formed
by the photodissociation of nitrogen dioxide. The ozone reacts with the
nitric oxide and the hydrocarbon (eqs. 5.3 and 5.10). These reactions
tend to deplete the ozone. However, simultaneously, the nitric oxide
reacts with various free radicals(denoted by R02, eq. 5.11), and this re-
action proceeds at a much faster rate than the reactions of ozone with
the nitric oxide and the hydrocarbon. The net result is an excess amount
of ozone in the system; hence, the observed increase in the ozone con-
tent in the first part of the tunnel. As the photochemical reaction pro-
gresses, the nitrogen dioxide is depleted by reacting with the free
radicals (eqs. 5.4 and 5.12). As the amount of nitrogen dioxide dim-
inishes, so does the amount of ozone formed resulting in an eventual de-
crease in the ozone content along the tunnel.
-------
-56-
Table 5.1 The Reaction Mechanism Proposed by
Eschenroeder and Martinez [19]
OH
1MU- T \\f " HVJ T \J
+ n a. M > n a. M
U_ f 1"! - U— T I'l
_ + WU •*•••• *• NU- •+ U_
k4
\tn i jjri ^ t~ ouMn
wu T wu_ ,. ;i ^ zin«u_
?. H20 2
HMO _L v> ±y - — — — *• ni i • 4- MO
iiiiU— T n " • uji T ri\j
k?
:. J. tin -k M -'- - *• HMO 4- M
ka
ko
nu. -L. 11^" 7 ^- •u /nrv ^
Ull- T Ho ^ Dp^«*J_;
• T no ^ u»j^inj_y
R02 + NO ^> N02 + d(OH)
DA -L. MA _Jl£_» M^T>AM^
^.A;
(5.2)
(5.3)
(5.4)
(5.5)
(5.6)
(5.7)
(5.8)
(5.9)
(5.10)
(5.11)
fc 10\
-------
-57-
It is now possible to evaluate the maximum amount (referred to sim-
ply as the "amount") of ozone produced for different initial concentra-
tions of cyclohexene ([C5H]Q]0) and nitrogen dioxide ([N02]0)« The amount
of ozone produced as a function of the initial concentrations of cyclo-
hexene and nitrogen dioxide is shown in Figs. 5.2 and 5.3. For a given
initial nitrogen dioxide concentration, the amount of ozone first in-
creased with increasing cyclohexene concentration. An increase in the
cyclohexene concentration beyond a certain value resulted in a decrease
in the amount of ozone (Fig. 5.2). Similar behavior was observed when
the initial cyclohexene concentration was kept constant and the concen-
tration of nitrogen dioxide was varied (Fig. 5.3). Such an interrelation
between the amount of ozone formed and the initial concentrations of the
hydrocarbon and the nitrogen oxides has also been noticed by other in-
vestigators [2,5,6,7,25,41,50], and plausible qualitative explanations
of these phenomena have been offered in references [25,50,55]. A brief
explanation of these phenomena may be given by referring again to eqs.
(5.1)-(5.12). First, we consider the case when the initial nitrogen
dioxide concentration is kept constant and the initial concentration of
cyclohexene is varied (Fig. 5.2). As discussed above, the photodissocia-
tion of nitrogen dioxide yields ozone and nitric oxide (eqs. 5.1 and 5.2).
Ozone is depleted by reactions with the nitric oxide and the hydrocarbon
(eqs. 5.3 and 5.10). In addition to reacting with ozone, the nitric
oxide also reacts with the free radicals (eq. 5.11), and this reaction
proceeds faster than the reactions depleting ozone. Thus, an increase in
the hydrocarbon concentration results in a decrease in the amount of nit-
ric oxide available for the depletion of ozone. Consequently, the amount
-------
5.0
E
a.
o.
._• 1.0
UJ
o
o
o
IM
O
I I
I I I I
I I 1 ITT T
Initial N02
Concentration
5ppm
0.1
0.3
Fig. 5.
I I I ! I I I
I I I I I I I
I I I I I I I
1.0 10 100
INITIAL CYCLOHEXENE CONCENTRATION (C6H|0]0 , ppm
Ozone content in cyclohexene-nitrogen dioxide-air mixtures. Relative
humidity: 50%. Light intensity Iq = 6.33 ks-1
-------
-59-
5.0
i
1.0
o
o
UJ
z
o
N
O
O.I
I I I I
T I I I I i I
2ppm
0 ^ Initial C6HIO
Concentration
i i i i i i
"0 1.0 10
INITIAL NITROGEN DIOXIDE CONC. [N02]Q , ppm
20
Fig. b.3. Ozone content in cyclohexene-nitrogen dioxide-air mixtures.
Relative humidity: bOi. Light intensity k] = 6.33 ks'l
-------
-60-
of ozone formed increases with initial increase in the hydrocarbon concen-
tration. However, once there is a sufficient amount of hydrocarbon pre-
sent in the mixture to react with all the available nitric oxide, then
any additional hydrocarbon will react with ozone. Under these condi-
tions, an increase in the hydrocarbon concentration results in a decrease
in the ozone content.
The formation of ozone can be explained similarly when the initial
concentration of the hydrocarbon is kept constant while the initial con-
centration of nitrogen dioxide is varied (Fig. 5.3). Ar the nitrogen
dioxide concentration is increased, both the amounts of ozone and nitric
oxide are Increased (eqs. 5.1 and 5.2). Because the reaction of nitric
oxide with- the free radicals (eq. 5.11) proceeds at a faster rate than
the reactions depleting ozone (eqs. 5.3 and 5.10), ozone accumulates in
the mixture. This explains the increase in the ozone content with the
initial increase in the nitrogen dioxide concentration. When the ini-
tial nitrogen dioxide concentration is increased tc the point where all
the hydrocarbon present in the mixture reacts with all the nitric oxide
produced by the nitrogen dioxide, then any further increase in the nit-
rogen dioxide concentration provides some excess nitric oxide to react
with the ozone. More ozone is depleted due to this reaction than is
formed due to the increase in the nitrogen dioxide concentration. Hence,
the ozone content is reduced when the nitrogen dioxide concentration is
increased.
The results presented in Fig. 5.2 and 5.3 would only apply to the
formation of ozone in cyclohexene-nitrogen dioxide-air mixtures. An
attempt is now made to generalize these results so as to gain additional
information regarding the ozone production by other types of hydrocarbons.
-------
-61-
It is recognized [15,46,55] that, in a reaction described by N reaction
equations, the amount of ozone formed depends on: a) the initial con-
centration of nitrogen dioxide [N02] , b) the initial concentration of
the hydrocarbon [HC] , and c) the rate constants k.(i=l to N), i.e.
fo3> f(fN02!0, rHC]0, kj_ , k2 .....kj, ) (5.13)
The terms in eq. (5.13) may be nondimensionalized, and eq. (5.13) may be
expressed as
* * *
(5.14)
where k-j , k2 ...... k^_, represent N-l dimensionless groups formed from the
different rate constants, and [63] is some characteristic ozone concen-
tration in the mixture. To select [03lc, it is recalled that in a nitro-
gen dioxide-air mixture the amount of ozone formed is (Section III)
(3.10)
In most practical situations, [Os]-.. is negligible compared to [N02]0>
in which case eq. (3.10) becomes
l°3JPss - ^l"Vo (5.15)
Representing [03] by this expression, eq. (5.14) can be rewritten as
k*_< ) (5-16)
V O
For a hydrocarbon^nitrogen dioxide-air mixture which contains just one
type of hydrocarbon, the values of the k^s are constants, and
\lIT [N02L depends only on the ratio of the initial concentrations of
the hydrocarbon and the nitrogen dioxide, i.e.,
-------
-62-
Coj
/JL
V k3
Equation (5.17) relates the amount of ozone formed to the initial
nitrogen dioxide concentration and the initial concentration ratio.
In many previous experiments, instead of nitrogen dioxide, nitric
oxide was used in the initial gas-air mixture. With nitric oxide in the
initial gas-air mixture, the production of ozone occurred at a later time
than with nitrogen dioxide, but the maximum amount of oznne generated was
found to be the sar.e for both nitric oxide and nitrogen dioxide. It is
noted, further, that some investigators reported measurements of "oxidant"
instead of ozone. The term oxidant designates certain pollutants formed
during the photochemical reaction. These pollutants are so named be-
cause they will oxidize some specific chemical reagents (e.g. a solution
of neutral-phosphate-buffered potassium iodide) which are not readily
oxidized by oxygen [69]. Although the oxidant measured includes other
chemical compounds as well as ozone (e.g. peroxyacyl nitrates), ozone
is generally the largest amount (by volume) present in the measured oxi-
dant, and the data reported for ozone and oxidant were usually very sim-
ilar [8,69].
For the reasons outlined above, results obtained with nitrogen
dioxide or with nitric oxide, and results reported in terms of ozone or
oxidant may be discussed together, and compared directly.
According to eq. (5.17), for a given type of hydrocarbon and fixed
initial concentrations of the hydrocarbon and the nitrogen dioxide, the
amount of ozone varies directly with the square root of k,
(5-18)
-------
-63-
It is recalled from Section III that k, is proportional to the light in-
tensity (eq. 3.1). Hence, the ozone content varies with the square root
of the light intensity. Harton and Bolze [31], Altshuller et. al. [5],
and Groblicki and Nebel [29] measured the effect of light intensity on
the amount of ozone (or oxidant) formed in various hydrocarbon-nitrogen
oxide-air mixtures. The results reported by these investigators also
confirm that the amount of ozone (or oxidant) varies with the square root
of the light intensity. The computer solution obtained by Niki et. al.
[46] for the photochemical reactions in a propylene-nitrogen oxide-air
mixture also confirms the validity of eq. (5.18).
Equation (5.17) also shows that, for a given type of hydrocarbon,
Hrj
the data points should fall on a common curve on a [Q^]/ / jj—• [N02]Q
versus [HCL/[N09] plot. Such a plot for the present data is given in
0 ' o
Fig. 5.4 . it is seen that, within experimental error, the data follows
a single curve. This lends support to the validity of eq. (5.17). In
order to test further the validity of this equation, the data reported
by other investigators have been plotted in a graph similar to Fig. 5.4.
These results, presented in Fig. 5.5, were for eight different types of
hydrocarbons. The data given in Fig. 5.5 were all obtained in chambers
irradiated artificially by either fluorescent lights [2,5,6,41,58] or by
mercury lamps [50]. The measurements were made either in a static type
chamber [2,6,41,50] or in a flow type tube [5,58]. Scott [50] irradiated
3-methylheptane-nitrogen dioxide-air mixtures in a long path infrared ab-
sorption cell. Altshuller and Cohen [2] tested ethylene-nitrie oxide-
air mixtures in Teflon bags. Altshuller et. al. [6] and Kopczynski et.
al. [41] used the same aluminum irradiation chamber and tested n-butane-
nitric oxide-air mixtures [6] and aldehyde-nitric oxide-air mixtures [41].
-------
10
8
-I 1*1
I I I I I I I
i i i i i i i
I II I I I I I |
o DATA
FIT TO DATA
i ill
I I i
0.2 1.0 10
INITIAL CONCENTRATION RATIO, "[ C6Hlo]o/[ N0z]o
Fig. b.4. Normalized ozone content in cyclohexene-nitrogon dioxide-air mixtures.
Relative humidity: 50%. Light intensity kj = 6.33 ks'1
50
-------
-6b-
2
0
10
^ 5
X
O
-I 10
H«
0
10
0
10
O.I
Fig. b.b.
fo I
(a)
(b)
(c)
(d)
1.0
10
100
10
0
10
0
10
(e)
(g)
10
100
INITIAL CONCENTRATION RATIO, [HC]0/tN°x]o
Normalized ozone content in hydrocarbon-nitrogen oxides-
air mixtures, (a) Ethylene, Altshuller and Cohen [2].
(b) 3-methylheptane, Scott [50]. (c) Cis-2-butene,
Stephens and Price [58]. (d) Propionaldehyde, Kopczynski,
et. al. [41]. (e) Acetaldehyde, Kopczynski, et. al. [41].
(f) Propylene, Altshuller et. al. [5]. (g) Formaldehyde,
Kopczynski, et. al. [41]. (h) N-butane, Altshuller et. al.
[6]
-------
-66-
Flow type experiments were performed by Altshuller et. al. [5] for propy-
lene-nitric oxide-air mixtures in an aluminum chamber, and by Stephens
and Price [58] for cis-2-butene-nitric oxide-air mixtures in a Pyrex
tube. Generally, fewer data points were taken in the previous experi-
ments than in the present study. Nevertheless, the previous data, sum-
marized in Fig. 5.5 also indicate that the data can be correlated ac-
cording to eq. (5.17).
In addition to the investigations quoted above, ozone (or oxidant)
measurements were reported in [7,10,25,31,50,51]. The results of these
investigations [7/0,25,31,50,51] could not be included in Fig. 5.5 be-
cause either there was not enough information given to plot the graph
[10,31], or the amount of ozone (or oxidant) measured was not given
directly [7,25,50,51]. The amount of ozone (or oxidant) was measured by
Bufalini and Altshuller [10] and by Harton and Bolze [31]. However,
Bufalini and Altshuller reported the nitrogen dioxide dosage instead of
the initial concentration of nitrogen dioxide. Harton and Bolze did not
report the maximum amount of the oxidant generated. The amount of ozone
(or oxidant) was reported indirectly by Altshuller et. al. [7], Glasson
and Tuesday [25], Haagen-Smit [50], and by Romanovsky et. al. [51].
Altshuller et. al. reported the oxidant dosage, Glasson and Tuesday the
initial formation rate of ozone. Haagen-Smit measured the crack depths
of rubber strips. Romanovsky et. al. presented their results in the
form of ozone contour maps.
An attempt can now be made to estimate a) the maximum possible
amount of ozone formed in a hydrocarbon-nitrogen dioxide-air mixture,
and b) the concentration ratio at which this maximum occurs. Let the
-------
-67-
ki
- [N02]Q versus the concentration ratio curve be
maximum of the
M, and the concentration ratio corresponding to this point be R (see
Fig. 5.4 and 5.5), i.e.
Co.]
'max
M =
and
R s.
2J0
(5.19)
at f°31
(5.20)
In principle, the values of M and R can be obtained from the appropriate
reaction equations (e.g. eqs. 5.1-5.12). In practice, this is a formid-
able, if not an impossible task owing to the complexity of the equations.
The data summarized here, however, provide some information about M and
R. Equations (5.19) and (5.20) may be rearranged in the form
(5.21)
(5-22)
M
The maximum amounts of ozone formed ([03!,,,) in Figs. 5.2 and 5.3
met A
are represented by the envelopes of the ozone versus cyclohexene and the
ozone versus nitrogen dioxide curves (dotted lines in Figs. 5.2 and 5.3).
From Figs. 5.2 and 5.3, the equations of these curves (envelopes) may be
determined, and are
(5.23)
(5.24)
-------
-68-
1/2
Comparing eqs. (5.21) and (5.23), and noting that/k^/k-j' = 0.122 ppm
(Section III), the following expression is obtained
M-5.3/T (5.25)
Tne above relationship between M and R was derived entirely from the
present data using cyclohexene. In order to estimate the relationship
between M and R for other types of hydrocarbons, M was plotted as a func-
tion of R for the nine different hydrocarbons for which values of M and
R could be estimated (Figs. 5.4 and 5.5). The results, given in Fig. 5.6,
show that in all c
-------
70
x
O
10
>
I i i i i
I I TIT
I I I I I
I I I I I I I
I
er>
0.4
I.O
R =
[HC](
10
30
Fig. 5.6. Maximum amount of ozone M as a function of the concentration ratio R
at which M occurs. 0 cyclohexene, present study; o ethylene, Altshullpr
and Cohen [2]; a 3-methylheptane, Scott [50]; t> Cis-2-butene, Stephens
and Price [58]; 7 propionaldehyde, Kopczynski et. al. [41]; < acetaldehyde,
Kopczynski et. al. [41]; * propylene, Altshuller et. al. [5]; A formaldehyde,
Kopczynski et. al. [41]; O n-butane, Altshuller et. al. f6l
-------
-70-
CM
Fig. 5.7.
INITIAL HYDROCARBON CONG. [HC]O , ppm
Maximum amount of ozone as a function of the initial
hydrocarbon concentration. present result;
«» ethylene, Altshuller and Cohen [2];
-------
-71-
were also included. These atmospheric measurements were reported by the
U.S. Department of Health, Education and Welfare [70] and gave the oxi-
dant content as a function of non-methane hydrocarbon concentration in
ppm C. In order to include the atmospheric data in Fig. 5.7, an average
carbon number of five was assumed for these non-methane hydrocarbon mix-
tures and k, was taken to be 8 ks (see Table 3.1). It is interesting
to note that eq. (5.27), which was based on the data obtained from mix-
tures containing only one type of hydrocarbon, also yields reasonable
results for mixtures containing several types of hydrocarbons such as
those present in the atmosphere.
The maximum amount of ozone that can be formed for a- given concen-
tration of rrftrogen dioxide depends on M (eq. 5.22). Since it is easier
to determine.R than M, eq. (5.22) is written as (see eqs. 5.22 and 5.25)
8.13 /k/TyrN0l (5.28)
or
fo.l
'max
,r w J2'0 (5.29)
Therefore, the maximum amount of ozone depends on R which is different
for different types of hydrocarbons. However, the ratio [03]max//k-|R
is independent of the type of the hydrocarbon and depends only on the
initial concentration of the nitrogen dioxide. Equation (5.29) is pre-
sented in Fig. 5.8 together with the available data deduced from Figs.
5.4 and 5.5. It is seen that there is reasonable agreement between the
curve given by eq. (5.29) and the existing data.
The foregoing discussion illustrated that in order to determine the
maximum amounts of ozone produced in a given hydrocarbon-nitrogen dioxide-
-------
-72-
-------
-73-
air mixture, the values of~M and R for that hydrocarbon are needed. It
would be desirable, therefore, to be able to estimate the values of M and
R for different types of hydrocarbons. The values of M and R depend on
all the rate constants in a complex manner. An inspection of the reac-
tion equations (eqs. 5.1 to 5.12), indicated that the rate constants which
would most likely influence M and R were kg, kg, and k1Q. In addition,
M and R might also be related to the conversion rate, kNQ, of nitric oxide
to nitrogen dioxide [1,3,4,26,33]. Therefore, an attempt was made to
correlate R with each of these rate constants. To facilitate this cor-
relation, the suggestion of Glasson and Tuesday [26] was followed and a
A
dimensionless relative rate constant called "reactivity", k, is defined
as
*
k =
(5.30)
where ^CHK 1S tne rate constant of propylene. The results, presented
in Table 5.2 show only weak correlations between R and the reactivities
A A A A
kg, k,Q, and kNQ, but indicate a strong correlation between R and kg.
A
Therefore, a graph of R versus kg was plotted (Fig. 5.9), and the follow-
A
ing approximate relationship between R and kg was obtained
R = FIT (5-31)
/ K9
The value of R can be found from this expression for any given hydrocar-
bon. This R value may then be used in eqs. (5.26) and (5.28) to calcu-
late the maximum amount of ozone formed.
Finally, in Fig. 5.10, an ozone contour map is presented on the
basis of the cyclohexene data. The general features of this map were
similar to those reported for propylene-nitrie oxide-air mixtures [51]
-------
-74-
Table 5.2 Reactivities of Different Hydrocarbons (k = k/kC3Hg)
Hydrocarbon
Initial
Reactivity
n-Butane
Formal dehyde
Propylene
Acetal dehyde
Propional dehyde
Cycl ohexene
Cis-2-Butene
3-Methyl heptane
Ethyl ene
Concentration
Ratio, Ra
10.5
8
5
4
3.5
2.4
2.1
2.0
1.3, 17f
A
k8
0.008C
0.05C
1.0
0.15C
0.2C
3.6e
0.08e
0.2C
— - - -^
A
kg
0.24C
0.9C
1.0
0.9C
1.8C
0.1C
/y . A
I, b i,
K10 KNO
0.2C
.... 0.7C
1.0 1.0
— - 0.7C
2.0C
5.2 1.0d
2.0 2.0d
— - 0.27d
0.33 0.3C
a
b
c
d
e
f
From Figs. 5.4 and 5.5
From Table 46 reference [43]
From reference [46]
From reference [26]
From reference [54]
From reference [25]
-------
30
10
o
x
O
I I I I I
1.0
0.4
T 1—J I I I I
FIT TO DATA
R=5.l /
I I I I I
_LL
on
t
O.I
1.0
4.0
REACTIVITY BASED ON HYDROXYL RATE CONSTANT,
9
Fig. 5.9. Variation ij the concentration ratio R (at which the ozone content is maximum) with the
reactivity kg. <* ethylene, Altshuller and Cohen [2]; O n-butane, Altshuller et. al. [6];
A formaldehyde, Kopczynski et. al. [41]; x propylene, Altshuller et. al. [5];
« acetaldehyde, Kopczynski et. al. [41]; 7 propionaldehyde, Kopczynski, et. al. [41]
-------
E
S 10
-------
-77-
and for mixtures of non-methane hydrocarbons in the atmosphere [70] (Figs,
5.10 and 5.11). There was one notable difference between the maps ob-
tained with one single hydrocarbon (e.g. cyclohexene or propylene) and
with mixtures of hydrocarbons. When only a single hydrocarbon was used,
each curve had reasonably well defined "knolls", which could be connected
by a straight line. For the atmospheric data, a straight line could not
be drawn through the knolls either because the measurements were inac-
curate, or because the many different types of hydrocarbons in the at-
mosphere distorted the slope of the curve.
5.2 The Effects of Mixture Composition on Particulate Formation
The nuclei and the aerosol contents in the tunnel are given in
Fig. 5.1 for s typical set of initial cyclohexene and nitrogen dioxide
concentrations. For other initial concentrations of nitrogen dioxide and
cyclohexene, the number distribution of the nuclei and the aerosols along
the tunnel exhibited similar trends, i.e. the numbers first increased,
reached maximum values and then decreased slightly. Thus, in general,
the nuclei and the aerosol formation followed a pattern similar to the
process of ozone formation. However, the nuclei and aerosol contents
changed at slower rates than the ozone content; in fact, the number of
nuclei and aerosols varied only slightly past the third sampling station
which was 1.75 m from the entrance of the tunnel. Since the nuclei and
the aerosol counts at the tunnel entrance were nearly zero, most of the
nuclei and the aerosols were formed in the region between the entrance
and the third sampling station. According to the data in Fig. 5.1, ozone
was also generated at the fastest rate in this region.
-------
E
Q.
Q.
o
o
en
UJ
o
X
o
UJ
o
o
-------
-79-
Im'tially, nuclei were produced by the interaction of some of the
gaseous reactants which were the byproducts of the photochemical reac-
tions [18,49,53,63]. The nuclei then interacted both with gaseous mole-
cules and with each other, resulting in an increase in the size of the
nuclei. Eventually, the nuclei reached a size which became detectable
on the aerosol counter. These "nuclei" were then designated as aerosols.
The number of nuclei and aerosols decreased along the tunnel due to co-
agulation, diffusion, and deposition. Since the number of nuclei and
aerosols remained reasonably constant along the tunnel, it appeared that
in the present system the formation and the depletion of nuclei and aero-
sols proceeded at nearly the same rates. Here, we were not concerned
with the details of these two processes but were only interested in the
maximum number of nuclei and aerosols generated for each combination of
nitrogen dioxide and cyclohexene concentrations. The maximum numbers of
nuclei and aerosols deduced from all the data are listed in Appendix E.
The maximum number (referred to simply as the "number") of nuclei
produced for different initial concentrations of cyclohexene and nitro-
gen dioxide is given in Fig. 5.12. In comparing this figure with Fig.
5.2, it can be seen that the variation in the nuclei content with the
initial concentrations of nitrogen dioxide and cyclohexene is remarkably
similar to the variation of the ozone content with the initial concen-
trations of nitrogen dioxide and cyclohexene. For example, similar to
ozone, for a given initial nitrogen dioxide concentration, the number
of nuclei produced first increased with the cyclohexene concentration,
reached a peak and then decreased with further increase in the initial
cyclohexene concentration. Moreover, for a given initial nitrogen diox-
ide concentration, the amounts of the ozone and the nuclei produced
-------
I04
10
'E
u
LJ
_l
O
or
u
m
[N]mox= 540
I T
I IT
lOppm
5ppm
O.Sppm
Initial N02 Concentration
i i i i i i
I
i i i i i i i
I i i i i i i
0.3 I 10
INITIAL CYCLOHEXENE CONCENTRATION
, ppm
Fig. 5.12. Nuclei content in cyclohexene-nitrogen dioxide-air mixtures
humidity: 50%. Light intensity kj = 6.33ks-l
100
Relative
-------
-81-
reached their respective maximum values at approximately the same initial
cyclohexene concentration (e.g. for [N02] =2 ppm, [03] occurs at
U Illu A
[C6H10]0=4'6 ppm and the maximum number of nuclei [N]max at [CsHio]0 =
4.4 ppm).
The similarities in the results of the nuclei and the ozone become
even more evident if the number of nuclei is plotted against the ini-
tial concentration ratio ([C6Hio]0/[N02]0) (Fig. 5.13). It is seen that
the maximum number of nuclei always occurs at a concentration ratio of
about 2.4, which is the same ratio at which the ozone content is a maxi-
mum in the mixture (Fig. 5.4).
The maximum number of nuclei produced for a given initial cyclohex-
ene concentration is represented by the dotted line in Fig. 5.12. The
equation of this line is
Hlo (5.32)
-3 -1/2
A, is a numerical constant having the value of 540 cm ppm ' . Again,
similar to ozone, a) the maximum number of nuclei may be expressed inde-
pendently of the initial nitrogen dioxide concentration, and b) the maxi-
mum number of nuclei varies with the square root of the initial cyclo-
hexene concentration. The maximum number of nuclei produced for a given
initial nitrogen dioxide concentration may also be evaluated from the
data (Fig. 5.12). The results show that
o (5.33)
-3 -1/2
where A~ = 835 cm ppnf . As in the case of ozone (see eq. 5.24),
the maximum number of nuclei varies directly with the square root of the
initial nitrogen dioxide concentration.
-------
ICT
10
I
E
u
UJ
or
UJ
O>
10'
i i i
Initial
Concentration
10 ppm
5ppm
I I I i I I
I
i i i i i i
CO
.N5
I
0.3 I 10
INITIAL CONCENTRATION RATIO, [C6H,0]0/ [N02]0
100
Fig. 5.13. Nuclei content as a function of the initial concentration ratio in c/clohexene-
nitrogen dioxide-air mixtures. Relative humidity: 50%. Light intensity
k = 6.33ks-l
-------
-83-
It should be noted that eqs. (5.32) and (5.33) were determined from
the data obtained under specific experimental conditions, i.e., k^=6.33
ks and ^ =50%. The constants A, and Ap, therefore, may depend on the
light intensity and the relative humidity. Data are as yet unavailable
which would indicate the dependence of A-, and Ap on the light intensity
and the relative humidity.
It would be of interest now to assess whether or not eqs. (5.32) and
(5.33) could be applied to other types of hydrocarbons. In order to ac-
complish these goals, data such as reported here, but obtained with
different types of hydrocarbons, would be needed. Unfortunately, such
data are unavailable at the present time. Only in a few investigations
were the nuclei number counts measured, and most of the existing data
were obtained under selected conditions. For example, in most previous
experiments, the initial hydrocarbon and nitrogen oxide concentrations
were fixed and only the type of the hydrocarbon used in the mixture was
varied. Consequently, the results of these tests do not give a range of
concentration ratios but apply only to one specific value. Furthermore,
in most of the past tests the number of nuclei was not determined dir-
ectly; instead, the relative number concentration of the nuclei (or
aerosols) was evaluated by measuring the light scattered from a light
beam passing through the mixture. Hence, we were unable to fit similar
expressions like eqs. (5.32) and (5.33) to the data obtained with other
types of hydrocarbons.
There appears to be only one nuclei measurement which can be com-
pared directly with the present results. Stevenson et. al. [59] per-
formed flow type experiments in a 0.125 m3 Pyrex tube which was irradiated
-------
-84-
by fluorescent lamps. They reported that for a mixture containing 2 ppm
of nitrogen dioxide and 3.5 ppm of cyclohexene, the number of nuclei pro-
duced was about 700 cm" . They did not report either the light intensity
or the relative humidity inside their Pyrex tube. However, for the cor-
responding mixture composition, the present data indicate a nuclei count
•3
of 800 cm (Fig. 5.12) which is in good agreement with the value obtained
by Stevenson, et. al.
The number of aerosols as a function of the initial cyclohexene and
nitrogen dioxide concentrations is shown in Fig. 5.14. For a given ini-
tial nitrogen dioxide concentration, the aerosol content at first in-
creased sharply with increasing concentrations of cyclohexene. At about
•3
400 aerosols per cm , the aerosol content gradually levelled off with
further increase in the initial cyclohexene concentration. Figure 5.14
does not provide much information on the number of aerosols formed. When
the aerosol content is plotted against the initial concentration ratio
([CsHio] /[N02]Q), the data points all collapse onto one common curve
(Fig. 5.15). In contrast to the nuclei (Fig. 5.13) and ozone (Fig. 5.4)
formation, the number of aerosols produced appear to depend only on the
initial concentration ratio. The effects of the initial nitrogen dioxide
concentration on the aerosol content seem to be small. The apparent lack
of influence of the nitrogen dioxide concentration on the aerosol pro-
duction may be due to the fact that the fluctuations in the aerosol mea-
surements (-30%) may be larger than the effects of the changes in the
nitrogen dioxide concentrations on the number of aerosols produced.
It is interesting to note that the aerosol content becomes appre-
ciable only when the initial concentration ratio ([C6HioL/[N02]0) is
larger than about 2.4, the same value at which the amounts of ozone and
-------
1000
to
I
u
en
O
o
100
oc
§
Initial N02
Concentration 0.6 ppm
1ppm
10
0.3
i i i i i i I
I
i i
I I i Mi1-
c:
in
i
lOppm
l i i i i i i
1.0 10
INITIAL CYCUOHEXENE CONCENTRATION, ppm
100
Fig. 5.14. Aerosol content in cyclohexene-nitrogen dioxide-air mixtures,
humidity: 50%. Light intensity k] = 6.33 ks-1
Relative
-------
-86-
10"
10
u
3
o
o:
UJ
<
u.
o
or
UJ
CD
IO
10
I I T_
Initial N02
Concentration ppm
x 0.6
o i
* 2
a 3
A 5
010
o
1
J I
10 60
INITIAL CONCENTRATION RATIO, [C6H|0]0/[N02]0
Fig. 5.15. Aerosol content as a function of the initial concentration
ratio in cyclohexene-nitrogen dioxide-air mixtures.
Relative humidity: 50%. Light intensity kj = 6.33 ks'1
-------
-87-
nuclei reach their maximums (see Figs. 5.4 and 5.13). When the ratio is
below 2.4, only about 50 aerosols cm are generated in the mixture.
This is a small number compared to the nearly 600 aerosols cm generated
when the concentration ratio is ~10.
It would be desirable now to compare the present aerosol data with
the results of other investigators. Direct comparison between the re-
sults of the present and previous investigations is not possible because
different light intensities and different relative humidities were used
in different experiments. The effects of the light intensity and the
relative humidity on aerosol formation in hydrocarbon-nitrogen oxides-
air mixtures have not been studied in the past. Some indications of these
two effects may be deduced from the aerosol studies performed with hydro-
carbon-nitrogen oxides-sulfur dioxide-air mixtures. Groblicki and Nebel
[29] investigated the effect of the light intensity on aerosol formation
in propylene-nitric oxide-sulfur dioxide-air mixtures. They found that
the light intensity had no effect on the maximum amount of aerosols pro-
duced. The light intensity merely affected the time required to produce
the maximum amount of aerosols.
The effect of the relative humidity on the amount of aerosols gen-
erated in hydrocarbon-nitrogen oxides-sulfur dioxide-air mixtures has been
studied by Groblicki and Nebel [29], Harkins and Nicksic [30], Prager et.
al. [47], Schuck et. al. [53], and by Wilson and Levy [64]. Some of the
results showed an increase in the amount of aerosols with increasing
humidity [53], while some showed a decrease [29,30,64]. Furthermore,
some of the data showed no apparent relationship between the amount of
aerosols generated and the relative humidity [47].
-------
-88-
Comparisons between the present and existing aerosol data are not
possible not only because the light intensities and the relative humi-
dities were different in the tests, but also because the number count of
the aerosols was not reported in most previous investigations.
In most of the past measurements, the light scattering method was
employed which provided the amount of light scattered by the aerosols
in some arbitrary unit, but did not yield directly the number of aerosols
per unit volume [29,47,49,59]. One notable exception was the test by
Renzetti and Doyle [49] who performed flow type experiments in a 0.05 m
Pyrex flask irradiated by mercury arc lamps. They recorded the number
_3
of aerosols produced to be 290 cm in a mixture containing 1 ppm nitric
oxide and 2 ppnr cyclohexene at a relative humidity of 50%. The present
data, shown in Fig. 5.15, indicate that the number of aerosols is in the
range of 100 to 300 cm"3. Renzetti and Doyle's data fall within the
spread of the present results.
A direct quantitative comparison is not feasible between the pre-
sent data and the data obtained by light scattering methods. A quali-
tative comparison between the present results and the data obtained by
light scattering methods is possible, because the amount of light
scattered is proportional to the number of aerosols in the mixture. Mea-
surements of light scattering in hydrocarbon-nitrogen oxides-air mixtures
were reported in references [29,47,49,59]. Renzetti and Doyle [49]
o
performed flow type experiments in a 0.05 m Pyrex flask irradiated by
mercury lamps, using 3 ppm hydrocarbon, 1 ppm nitric oxide, and 50%
relative humidity in all their tests. Prager et. al. [47] performed
static type experiments in a long-path infrared absorption cell irradiated
by mercury lamps. Prager et. al. kept the mixture compositions at
-------
-89-
10 ppm hydrocarbon, 5 ppm nitric oxide, and 100 ppm water vapor (~ 0.3%
relative humidity). Groblicki and Nebel [29] tested gas-air mixtures in
a 8.5 nr* stainless steel chamber irradiated by fluorescent lamps. The
mixture composition was fixed at 4 ppm hydrocarbon and 2 ppm nitric oxide.
Stevenson et. al. [59] performed flow type experiments in a 0.125 m
Pyrex tube irradiated by fluorescent lamps and measured the aerosol light
scattering in cyclohexene-nitrogen dioxide-air mixtures (at 3.5 ppm cyclo-
hexene and 2 ppm nitrogen dioxide, and at 1.83 ppm cyclohexene and 1 ppm
nitrogen dioxide). Altogether, over 40 different types of hydrocarbons
were tested by these previous investigators [29,47,49,59]. However, in
each experiment, only one set of initial hydrocarbon and nitrogen oxides
concentrations, was. tested. Hence, an aerosol light scattering versus
initial concentration ratio curve could not be plotted for any of these
hydrocarbons, with the exception of cyclohexene for which light scatter-
ing was measured by four different investigators [29,47,49,59]. For cy-
clohexene, the aerosol light scattering against the concentration ratio
could be plotted. This plot, given in Fig. 5.16, may be compared with the
present data shown in Fig. 5.15. It is seen that these two figures ex-
hibit certain common features. First, both the aerosol light scattering
and the aerosol number count indicate that when the initial concentration
ratio is increased from 2 to 5 the amount of aerosols produced is in-
creased by a factor of ten (i.e. the aerosol content is increased from
40 to 400 cm" ; the light scattering from 3 to 30). Second, the light
scattering measurements support the present observations that: a) few
aerosols are produced when the concentration ratio is below 2.4, and
b) the number of aerosols produced defends mainly on the concentration
ratio and very little on the initial nitrogen dioxide concentration.
-------
-90-
100
o
!o
10
o:
LJ
<
o
III I I L-
I I I I I I 1
I
10
INITIAL
CONCENTRATION RATIO
[C6HJ0/[NOX]0
Fig. 5.16. Light scattering from aerosols produced in cyclohexene-
nitrogen oxides-air mixtures, a Stevenson et. al. [59];
4 Groblicki and fJebel [29]; S Prager et. al. [47];
{ Renzetti and Doyle [49]
-------
-91-
In addition to the total number of aerosols, the size distributions
of the aerosols were also determined for several mixture compositions.
The results of these aerosol size measurements obtained at two different
sampling stations (#3 and #11) are shown in Fig. 5.17. The data are
expressed in terms of the dimensionless parameters (see Section II)
7
J,
'
A NMr
(2.3)
With the use of these paramp.ters, all aerosol size measurements should
fall on one common curve [21, 45]. Moreover, in the size range where
the aerosol is irt a ai.ate of dynamic equilibrium (i.e. the rate of gain
of aerosols by ro&gulation is equal to the rate of loss of aerosols due
to coagulation, diffusion and sedimentation), ^- should vary inversely
with the fourth power of *) [11,22,38,45], i.e.,
Y* <^ n-4 (2.6)
The results in Fig. 5.17 show that, on a graph of/" versus ^ , the
present data follow a common curve and the slope of this curve is nearly
-4. The slope of the curve becomes somewhat steeper for larger values
of H (i.e., larger aerosols). These results agree very closely with the
results of Husar and Whitby [35] (solid line in Fig. 5.17) who monitored
the growth of aerosols due to photochemical reactions in atmospheric air
contained in a 77.5 m polyethylene bag. Husar and Whitby did not re-
port the composition of the atmospheric air.
-------
-92-
iov
10
4
o In Tunnel
Room Air
Atm.Air
\
A
Present
00,0
Husor a \ \
Whllby V\
•\
• \
0.03 0.1
1.0
10
Aerosol size distribution obtained in the tunnel for
cyclohexene-nitrogen dioxide-air mixtures, o in the tunnel,
present study; room air, present study; atmospheric
air, Husar and Whitby [35]
-------
-93-
The size distribution of the aerosols in the room where the pre-
sent tests were performed (see Fig. 2.6) is also included in Fig. 5.17
(dotted line). It appears that the size distributions (expressed in
terms of ijr and n ) of the aerosols produced by the photochemical reac-
tions, either in a tunnel or in a bag, do not differ markedly from the
size distribution of the aerosols in the natural atomsphere. The aero-
sol size distributions in the present tunnel, in Husar and Whitby's test,
and in the atmosphere all agree very well with each other when presented
on a\t- versus rj plot.
The size distributions (expressed in terms of ^ and rj ) seem to be
unaffected by the mixture composition, the light intensity or the rela-
tive humidity.
-------
-94-
CHAPTER VI
CONCLUDING REMARKS
The experiments performed in this investigation provide new informa-
tion on the participate formation in sulfur dioxide-air mixtures and on the
ozone and particulate formation in cyclohexene-nitrogen dioxide-air mix-
tures.
The results obtained during the photooxidation of sulfur dioxide-air
mixtures showed that, for initial sulfur dioxide concentrations ranging
from 0.2 to 1 ppm and relative humidities ranging from 25 to 100%, the
maximum number of nuclei varied directly with the sulfur dioxide concen-
tration and with tie square of the relative humidity. The present results
obtained at sulfur dioxide concentrations less than 1 ppm are in general
agreement with the data reported by earlier investigators.
Aerosol measurements were also made in sulfur dioxide-air mixtures;
the amounts of aerosols produced in the mixture were found to be negligible.
The maximum amount of ozone formed in cyclohexene-nitrogen dioxide-
air mixtures was shown to be directly proportional to a) the square root
of the initial cyclohexene concentration, b) the square root of the ini-
tial nitrogen dioxide concentration, and c) the square root of the light
intensity. Using the results of the present study and the data obtained
by previous investigators for different types of hydrocarbons, expres-
sions were developed which indicated quantitatively the maximum amount
of ozone which can be formed with different types of hydrocarbons.
Nuclei measurements in cyclohexene-nitrogen dioxide-air mixtures
showed that the maximum number of nuclei varied directly with the square
root of both the cyclohexene and the nitrogen dioxide concentrations.
The production of aerosols in cyclohexene-nitrogen dioxide-air mix-
tures appeared to depend only on the ratio of the initial concentrations
-------
-95-
of cyclohexene and nitrogen dioxide. Within the accuracy of the measure-
ments, the amount of aerosols formed seemed to be insensitive to the nit-
rogen dioxide concentration.
The aerosol size distributions in cyclohexene-nitrogen dioxide-air
mixtures were also measured for several mixture compositions. The size
distributions were found to follow closely Friedlander's universal self-
preserving curve.
During the experiments with cyclohexene-nitrogen dioxide-air mixtures,
the light intensity and the relative humidity were kept constant. The
effects of these two parameters on the amount of ozone formed could be
assessed by utilizing the results of previous investigations. The effects
of the light intensity and the humidity on nuclei and aerosol formation
could not be evaluated because of the lack of relevant data. Hence, the
influence of light intensity and relative humidity on particulate forma-
tion merit further investigations.
Finally, it is noted that static type experiments were also performed
in a 0.6 m i.d. and 1.2 m tall Plexiglas irradiation chamber. The re-
sults obtained from these tests were unrepeatable and were scattered widely.
The exact reasons for the scatter in the data are unknown. It is strongly
suggested, however, that the scatter was caused by the fact that the cham-
ber could not be cleaned thoroughly before each test. Although the chamber
was flushed with clean air several times prior to each test, such flushing
did not seem to clean the system adequately. This conclusion is supported
by the observation that in the flow type experiments, up to seven days of
continuous flushing was needed to achieve a "clean" chamber (Section III).
-------
APPENDIX A
LITERATURE SURVEY
In the experiments performed in the past to study photochemical re-
actions, the following procedure was generally used. The gas mixture was
introduced into a container. The mixture was then irradiated and some
products of the reaction were monitored. The source of irradiation was
either natural sunlight or artificial ultraviolet light. The gas mix-
ture either stayed inside the container during the entire experiment
(static type experiment) or flowed through the container at a steady
rate (flow type experiment).
In this survey, only those experiments are summarized in which the
measurements of ozone (or oxidant, see below) and particulates were of
primary interest. Accordingly, this survey is divided into two parts.
The first part describes experiments in which the measurement of ozone
(or oxidant) was the major concern. The second part describes experi-
ments in which the production of particulates wpre investigated. The
results of the various studies are presented in chronological order.
Ozone
Here the results of those previous investigations are included in
which the amount of either ozone or oxidant was measured. This is be-
cause the formation processes of the oxidant and of the ozone are very
similar [8,69]. Moreover, in a major portion of the existing literature
measurements of the oxidant and not ozone were reported. It is, there-
fore, both desirable and advantageous to include these reports in the
survey.
-96-
-------
-97-
Haagen-Smit [50] was among the first to study the photochemical
reaction in hydrocarbon-nitrogen oxides-air mixtures. He investigated
the formation of ozone in a series of static type experiments in which
3
the gas mixture was placed in a 0.005 m Pyrex flask irradiated by flu-
orescent lights. The gas mixture tested was 3-methyheptane and nitro-
gen dioxide diluted in dry oxygen. Haagen-Smit varied the initial con-
centrations of 3-methyheptane (0.1 to 10,000 ppm) and of nitrogen dioxide
(0.04 to 400 ppm). In the report, he also presented measurements in
which natural sunlight and air were used instead of the fluorescent
lights and oxygen. In all these experiments, the amounts of ozone gen-
erated were inferred from the crack depths of the rubber strips placed
inside the flask and irradiated for ten hours.
Scott [47] performed static type experiments in which the gas mix-
ture was placed in a long-path infrared absorption cell irradiated by a
mercury lamp. The experiments were performed for two initial concentra-
tions of 3-methyheptane (3 and 40 ppm). The initial nitrogen dioxide
concentration was varied from 0.1 to 8 ppm. For each mixture composi-
tion, the amount of ozone generated was monitored as a function of the
irradiation time (up to six hours).
Harton and Bolze [29] performed a series of static type experiments
in which the hydrocarbon-nitrogen dioxide-air mixture was placed in a
62 m glass chamber irradiated by natural sunlight. The initial con-
centrations were usually 0.3 and 0.6 ppm for the hydrocarbon and 0.6 and
1.2 ppm for the nitrogen dioxide. Harton and Bolze tested over thirty
different types of hydrocarbons which included paraffins, straight and
branched olefins, cyclo-olefins, diolefins, aromatics as well as commer-
-------
-98-
cial solvents. Besides measuring the oxidant and the nitrogen oxides
concentrations during the tests, eye irritation measurements were also
taken.
Altshuller and Cohen [2] performed static type experiments in 0.09
3
m Teflon bags irradiated by fluorescent lamps. They studied the photo-
chemical reactions in ethylene-nitrie oxide-air mixtures by varying the
initial concentrations of ethylene (0.1 to 5 ppm) and of nitric oxide
(0.2 to 10 ppm). In addition to measuring the amount of oxidant gen-
erated, Altshuller and Cohen also monitored the concentrations of ethy-
lene, nitrogen dioxide, carbon monoxide, and formaldehyde during the
tests.
Glasson and Tuesday [25] performed static type experiments. The
hydrocarbon-nitric oxide-air mixture was placed in a long-path infrared
absorption cell irradiated by fluorescent lights. Glasson and Tuesday
varied the initial concentration of nitric oxide from 0.05 to 3 ppm.
For the five hydrocarbon tested (propylene, ethylene, trans-2-butene,
isobutene, and m-xylene), the initial concentrations most often used
were 1 and 2 ppm. In their experiments, Glasson and Tuesday measured the
formation rate of ozone, the oxidation rate of hydrocarbon and the con-
centration of peroxyacetyl nitrate.
Romanovsky et. al. [51] performed static type experiments in a
32 m glass aluminum chamber irradiated by fluorescent lamps. The mix-
ture investigated was propylene and nitric oxide in air. The initial
concentration of propylene was varied from 0 to 8 ppm and that of nitric
oxide was varied from 0 to 4 ppm. During the experiments, the amounts
of oxidant, nitrogen oxides, propylene, and formaldehyde were monitored
and eye irritation was also studied.
-------
-99-
Altshuller et. al. [5] performed both static type and flow type
3
experiments in a 95 m aluminum chamber irradiated by fluorescent lamps.
In the propylene-nitric oxide-air mixture studied, the initial concen-
tration of propylene was varied from 0.25 to 3 ppm and the nitric oxide
initial concentration was varied from 0.125 to 4 ppm. The concentra-
tions of oxidant, nitrogen dioxide, carbon monoxide, methyl nitrate, and
the aldehydes were all monitored.
Stephens and Price [58] performed flow type experiments in a 0.141
m Pyrex tube irradiated by fluorescent lamps. The initial concentration
of cis-2-butene was varied from 0 to 2.4 ppm while the nitric oxide con-
centrations used were 0.05, 0.125, 0.25, and 0.5 ppm. Stephens and Price
measured the amounts of oxidant, nitrogen dioxide, methyl nitrate, and
peroxyacetyl nitrate during these tests.
Bufalini and Altshuller [10] performed static type experiments in
3
which hydrocarbon-nitric oxide-air mixtures were placed in a 0.072 m
Pyrex flask irradiated by fluorescent lamps. The hydrocarbons tested
were 2-3-dimetnyl-2-butene, 1-butene, n-butane, and 1,3,5-trimethylben-
zene. The initial concentrations of the hydrocarbon were varied such that
the ratio of the hydrocarbon concentration to the nitric oxide concen-
tration ranged from 0 to 5. The parameters measured included the amount
of oxidant formed, the percent of hydrocarbon reacted, and the nitrogen
dioxide dosage produced.
Altshuller et. al. [6] performed static type experiments in which
paraffinic hydrocarbon-nitric oxide-air mixtures were placed in a 95 m
aluminum chamber irradiated by fluorescent lamps. Of the eight paraffins
studied, n-butane was most frequently tested. The initial concentra-
tion of n-butane was kept at 3 and 6 ppm while the nitric oxide concen-
-------
-100-
tration was varied from 0 to 2.4 ppm. In their experiments Altshuller
et. al. monitored the amounts of oxidant, nitrogen dioxide, aldehydes,
peroxyacetyl nitrate, and methyl nitrate, and also studied the eye irri-
tation effect as well as the damage incurred in Pinto Beam plants.
Using the apparatus in reference [6], Altshuller et. al. [7] also
studied the photochemical reaction in aromatic hydrocarbon-nitric oxide-
air mixtures. The initial concentrations of the two aromatics studied
(toluene and m-xylene) were kept at 1.5, 3, and 4.5 ppm while the nitric
oxide initial concentration was varied from 0.2 to 1.5 ppm. Altshuller
et. al. measured the dosages of the oxidant, nitrogen dioxide and per-
oxyacetyl nitrate and evaluated eye irritation effects.
Kopczynski et. al. [41] used the same apparatus as Altshuller et. al.
[6,7]. Kopczynski et. al. investigated the photochemical reaction in
aldehyde-nitric oxide-air mixtures. The initial concentrations of the
three aldehydes tested (formaldehyde, acetaldehyde and propionaldehyde)
were kept at 3 ppm while the initial nitric oxide concentration was varied
from 0.1 to 7.6 ppm. In this investigation the concentrations of the
oxidant, hydrocarbon, nitrogen dioxide and peroxyacetyl nitrate were
monitored and the eye irritation effect was studied.
In general, the results of all the previous investigations showed
that, for a given initial hydrocarbon concentration, the amount of ozone
produced first increased with the initial nitrogen oxides concentration,
reached a peak, and then decreased with further increase in the nitro-
gen oxides concentration. When the initial nitrogen oxides concentra-
tion was kept fixed and the initial hydrocarbon concentration was varied,
the amount of ozone produced also first increased with the initial
-------
-101-
hydrocarbon concentration, reached a peak, and then decreased with
further increase in the initial hydrocarbon concentration. Compre-
hensive quantitative comparisons and correlations of the results were
not attempted in these investigations.
Particulates
Particulate formation has been studied in many different types of
gas-air mixtures and the particles formed have been generally classi-
fied under two categories. "Condensation nuclei" are particles whose
diameters lie between 2 nm and 0.1 Mm and "aerosols" are particles
whose diameters are greater than 0.3yum. Condensation nuclei were
usually detected and measured by condensation nuclei counters (e.g.
General Electrtc Condensation Nuclei Counter). Aerosols were frequently
detected and" measured by aerosol light scattering photometers (e.g.
Royco Particle Counter).
The formation of particles during the photooxidation of sulfur
dioxide-air mixtures were studied by several investigators [13,14,17,
24,40,48,49]. Gerhard and Johnstone [24] performed static type experi-
ments in a 0.008 cnr* Lucite chamber irradiated by a sun lamp. The ini-
tial sulfur dioxide concentrations were 5, 10, 20, and 30 ppm. The
relative humidity was varied from 32 to 91% and the length of irradia-
tion time from 15 minutes to 6 hours. Gerhard and Johnstone measured
the amount of sulfuric acid drops collected on a high velocity impactor.
Dunham [17] performed two static type tests to demonstrate the
pohtooxidation mechanism in sulfur dioxide-air mixtures. In the first
test, sulfur dioxide was introduced into a quartz tube containing ozone-
rich air, while the ultraviolet light was turned off. In the second
test, a sulfur dioxide-air mixture in the quartz tube was irradiated by
-------
-102-
an ultraviolet light. Dunham found that nuclei was formed only in the
second test and concluded that nuclei formation was due to the photo-
oxidation of sulfur dioxide.
Renzetti and Doyle [49] performed flow type experiments in a 0.05
m Pyrex flask irradiated by mercury lamps. The initial sulfur dioxide
concentration was varied from 0 to 0.5 ppm. Relative humidities of 0
and 50% were used. The residence time of the gas-air mixture inside the
flask was kept at either 17 or 27 minutes. Particulate formation was
measured by an aerosol counter photometer and a nuclei counter.
Quon et. al. [48] performed a series of static type experiments in
0.014 m Saran bags irradiated by fluorescent lamps. The initial sulfur
dioxide concentration was varied from 0.202 to 0.647 ppm, the relative
humidity from 13 to 77%, and the length of the irradiation time from 30
to 240 seconds. The amount of nuclei formed was measured by a conden-
sation nuclei counter.
T
Cox and Penkett [13] performed static type experiments in a 0.216 m
aluminum chamber irradiated by natural sunlight. Sulfur dioxide at var-
ious initial concentrations (0.08 to 0.76 ppm) were tested. Cox and
Penkett did not report the relative humidity in the tests.
o
Cox [14] performed flow type experiments in a 0.005 m Pyrex flask
irradiated by a mercury lamp. The initial sulfur dioxide concentration
was varied from 5 to 1000 ppm. The relative humidities tested were 1,
5, 24, and 80%. The residence time of the gas mixture inside the flask
was fixed at 150 seconds. The number of nuclei formed was measured
during the tests.
-------
-103-
Kocmond et. al. [40] performed a series of static type experi-
ments in three different chambers irradiated by fluorescent lamps. Two
of the chambers were 0.25 and 1.7 m Teflon bags, the third one was a
600 m steel chamber whose interior surfaces were lined with a spec-
ially formulated fluoro-epoxy polymer. Various initial sulfur dioxide
concentrations (from 0.049 to 2.88 ppm) and relative humidities (from 11
to 77%) were used in the experiments. The particulates produced were
measured by a Whitby Aerosol Analyzer and a General Electric Condensa-
tion Nuclei Counter.
The results from the investigations described above all agreed that
the number of nuclei produced increased with the initial sulfur dioxide
concentration. However, the exact relationship between the number of
nuclei produced and the initial sulfur dioxide concentration still re-
mains unclear. There is also disagreement among the reports on the
effect of the relative humidity on the production of the condensation
nuclei. Some studies showed that the number of nuclei produced in-
creased with humidity [14,49]; others found that there was no humidity
effect [24,48].
Experiments on particulate formation in hydrocarbon-nitrogen
oxides-air mixtures were reported in references [29,47,49,59]. Using
the apparatus described before, Renzetti and Doyle [49] studied aerosol
production in 18 different types of hydrocarbons. The same initial
concentration (3 ppm hydrocarbon, 1 ppm nitric oxide and 50% relative
humidity) was used in each test for each hydrocarbon. Renzetti and
Doyle used an aerosol counter photometer to measure the amount of light
scattered by the aerosols.
-------
-104-
Prager et. al. [47] performed static type experiments in a long-
path infrared absorption cell irradiated by mercury lamps, with 10 ppm
hydrocarbon, 5 ppm nitrogen dioxide and 100 ppm water vapor. Altogether
Prager et. al. tested 17 different types of hydrocarbons for aerosol
production. They used a Sinclair-Phoenix smoke photometer to measure
the amount of light scattered by the aerosols.
Stevenson et. al. [59] performed flow type experiments in a 0.125 m3
Pyrex tube irradiated by fluorescent lamps. Selected initial concentra-
tions of five hydrocarbons and nitrogen dioxide were tested. The
amounts of particulates formed were measured by a nuclei counter and a
Sinclair-Phoenix smoke photometer.
Groblicki and Nebel [29] performed static type experiments in a
3
8.5 m stainless steel chamber irradiated by fluorescent lamps. Five
different hydrocarbons were used in the hydrocarbon-nitric oxide-air
mixture for the production of particulates. The initial hydrocarbon
and nitric oxide concentrations were always 4 ppm and 2 ppm respectively.
Groblicki and Nebel used a condensation nuclei counter and a Sinclair-
Phoenix smoke photometer in their particulate measurements.
A total of over 40 different types of hydrocarbons were studied for
particulate formation in the investigations listed above [29,47,49,59].
The general conclusion reached by the previous studies was that only
highly branched olefins, cyclo-olefins and diolefins were capable of
producing large quantities of particles.
Investigations were made to study the effect of adding sulfur diox-
ide to hydrocarbon-nitrogen oxides-air mixtures [27,29,47,49,53,59].
Groblicki and Nebel [29], Prager et. al. [47], Renzetti and Doyle [49],
and Stevenson et.al. [59] all introduced known concentrations of sulfur
-------
-105-
dioxide into the hydrocarbon-nitrogen oxides-air mixtures and measured
the amounts of particulates formed. Schuck et. al. [53] performed static
3
type experiments in a 14.7 m glass chamber irradiated by a combination
of mercury lamps and fluorescent lamps. Schuck et. al. measured the
aerosol light scattering produced in an air mixture containing 3 ppm
2-methyl-2-butene, 1 ppm nitric oxide, and 0.1 ppm sulfur dioxide at
various relative humidities. Other types of hydrocarbons (ethylene,
acetylene, isobutylene, and formaldehyde) were also tested for particu-
late formation. Goetz and Pueschel [27] performed flow type experi-
ments in which 1-octene-nitrogen dioxide-sulfur dioxide-air mixtures
•5
were placed in a 0.019 m glass tube irradiated by fluorescent lamps..
The initial concentrations of 1-octene and nitrogen dioxide were kept at
80 and 60 ppm respectively. The initial sulfur dioxide concentrations
used were i* 15, and 60 ppm and the relative humidity was changed from
15 to 70%. Goetz and Pueschel measured the aerosols by means of a
Royco Particle Counter and an aerosol spectrometer.
The results of these investigations showed that, in general, sulfur
dioxide would enhance the production of particulates in a hydrocarbon-
nitrogen oxides-air mixture.
There have been very few experiments performed to determine the
chemical compositions of the photochemical particles. The particulates
formed in sulfur dioxide-air mixtures have been found to be sulfuric
acid [13,14,24,48]. The exact chemical compositions of the particulates
formed in hydrocarbon-nitrogen oxides-air mixtures and in hydrocarbon-
nitrogen oxides-sulfur dioxide-air mixtures are not well known, although
a few measurements have indicated that the particulates are predominantly
nitrates and sulfates [18,29,47].
-------
-106-
APPENDIX B
CERTIFIED CONCENTRATION OF TEST GASES
Test Gas Concentration, ppm Supplier
Nitrogen Dioxide 205 Air Products & Chemicals
Company
Cyclohexene 219 Matheson Gas Company
Sulfur Dioxide 208 Matheson Gas Company
-------
-107-
APPENDIX C
ESTIMATION OF PARTICLE LOSSES DUE TO
SETTLING, COAGULATION AND DIFFUSION
Particles may be lost along the irradiation tunnel and along the
sampling line between the sampling stations and the particle counters.
Table C.I lists the flow conditions in the tunnel and in the sampling
line. The "residence time", (defined as the average time during which
the gas-air mixture stays in the tunnel or in the sampling line) is ob-
tained by dividing the length of the tunnel or the sampling line by the
corresponding mean flow velocity. The Reynolds numbers in both flow
situations are low and the flows are essentially laminar (Table C.I).
C.I Loss Due to Settling
The settling velocities of very small particles settling under the
action of gravity are given in reference [23]. The distance travelled
by a particle falling under the action of gravity in a certain time is
equal to the product of the settling velocity and the time. Table C.2
lists the settling velocities and the vertical distances travelled by
different size particles during a time interval equal to the residence
time both in the tunnel and in the sampling line. Particles deposit on
the lower part of the inside surface of the tube when the distance fallen
is larger than the inside diameter of the tube. The inside diameters of
the tunnel and the sampling line are 15.2 and 0.635 cm respectively.
These distances are larger than the distances travelled by particles
less than ~ 1 yu m in diameter. Thus, losses due to settling is neglig-
ible for particles less than -I *-ni in diameter.
-------
-108-
Table C.I Flow Conditions in the Tunnel and in the Sampling Line
Tunnel
Sampling
Line
Diameter
cm
15.2
0.635
Length
cm
912
120
Flowrate
cnrs"'
46
4.72
Mean Flow
Velocity
cm s-T
0.255
14.9
Residence
Time
3,598
8.1
Reynolds
Number
24
59
Table C.2 Estimation of the Vertical Distance Travelled by Different
Size Particles Due to Gravity During Residence Time in the
Tunnel and in the Sampling Line
Particle
Diameter
;A m
2xlO"3
4xlO"3
IxlO"2
Settling
Velocity*
nm s-1
1.31x10
2.62x10
6.63x10
Distance
Tunnel
4.7xlO"3
9.4x!0"3
2.4xlO"2
Travelled
Sampling
IxlO"5
2xlO"5
5x1 O"5
2x10
1x10
2x10
1
2
-2
-1
-1
1.37x10'
8.64xl02
2.24xlO:
3.47x10^
1.28x10*
4.9x10"
3.1x10"
8.1x10"
1.2x10
4.6x10
1.1x10"
_4
7x10
1.8x10"
2.8x10"
IxlO"1
Tunnel Diameter=15.2 cm
Sampling Line Diameter=0.635 cm
From Table 13 in reference [23]
-------
-109-
C.2 Loss Due to Coagulation
The number of particles also decreases due to coagulation with the
passage of time. For particles initially of the same size, it can be
shown that the number of particles n, after time t, is related to the
initial number nQ by the expression
~n^ = ~"l + KnQt (C.I)
where K is the coagulation constant [23]. Table C.3 lists the coagula-
tion constants for different particle sizes. In order to compute the
value n/n0, t is assumed to be equal to the residence time and n0 is
•s 3
assumed to be 10 particles cm . It is found that, for every particle
size, the value n/n calculated from eq. (C.I) is very close to one.
Hence, the Toss due to coagulation is negligible for all sizes of par-
ticles that may be present in the gas -air mixture.
C.3 Loss Due to Diffusion
Particles may also be lost due to diffusion to the inside surface
of a tube. An expression was given in reference [23] for the diffusion
of particles in a laminar flow inside a circular tube. If nQ is the
initial particle concentration, n the particle concentration after passing
through a tube of radius R and length L, U the mean flow velocity in the
tube, and D the diffusion coefficient of the particle, then n/nQ is
given by
-- = 1 - 2.56u.2/3 + 1.?, + 0.177v1/3
yu. + 1.?^, + 0.177/v (c.2)
where JU is a dimensionless parameter equal to DL/UR^ . Calculations
have been made for various particle sizes using eq. (C.2) and the re-
-------
-110-
Table C.3 Calculation of Particle Loss Due to Coagulation
3 3
-------
suits are presented in Table C.4. It is seen that the loss of aerosols
(particle diameter > 0.3 urn) due to diffusion is negligible. However,
for the very small particles (particle diameter < 0.01 yum) which fall
into the condensation nuclei category, the loss due to diffusion is ap-
preciable. Nearly all the particles whose diameters are about 0.002yum
appear to be lost due to diffusion. A correction for this effect was
not made to the present data since the size distribution of the conden-
sation nuclei was not measured. The measurable range of the condensation
nuclei counter is from 0.002 u m to 0.1 M-m. The loss due to diffusion
of the larger size particles recorded by the nuclei counter is seen to
be small (Table C.4).
-------
Appendix D.I Ozone Content Along the Tunnel
[N02]
ppm
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
1.0
1.0
1.0
1.0
1.0
1.0
1.0
jre Composition
[C6H10]0
ppm
0.5
1.0
1.25
2.0
3.0
5.0
10.0
20.0
0.5
1.0
2.0
3.0
5.0
10.0
20.0
1
0.09
0.155
0.22
0.23
0.265
0.2
0.13
0.072
0.12
0.15
0.22
0.375
0.29
0.185
0.105
2
0.185
0.345
0.62
0.47
0.46
0.325
0.188
0.102
0.17
0.32
0.76
1.0
0.67
0.37
0.188
Ozone Content (ppm) at Different Sampling Stations
4 6 8 10 12
0.255
0.49
0.70
0.575
0.54
0.31
0.14
0.068
0.157
0.36
0.825
1.0
0.605
0.265
0.124
0^32
0.585
0.75
0.66
0.60
0.305
0.115
0.504
0.157
0.395
0.91
1.0
0.635
0.215
0.088
0.33
0.595
0.71
0.67
0.59
0.28
0.088
0.041
0.145
0.37
0.87
1.0
0.62
0.17
0.058
0.41
0.695
0.82
0.81
0.73
0.33
0.086
0.039
0.158
0.395
0.98
1.1
0.77
0.19
0.058
0.385
0.61
0.42
0.715
0.66
0.30
0.083
0.037
0.145
0.315
0.86
1.0
0.70
0.175
0.055
ro
i
-------
Appendix 0.1 (continued)
Mixture Composition Ozone Content (ppm) at Different Sampling Stations
ppm ppm 1 2 4 6 8 10 12
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
0.5
1.0
2.0
3.0
5.0
10.0
20.0
50.0
0.5
1.0
2.0
3.0
5.0
7.5
10.0
20.0
30.0
0.115
0.115
0.27
0.41
0.65
0.38
0.215
0.078
0.112
0.118
0.168
0.28
0.55
0.85
0.65
0.35
0.24
0.115
0.15
0.585
0.98
1.3
0.81
0.415
0.133
0.081
0.092
0.167
0.52
1.2
1.65
1.3
0.71
0.44
0.071
0.11
0.545
0.91
1.2
0.69
0.29
0.09
0.041
0.047
0.088
0.38
1.05
1.4
1.15
0.56
0.32
0.054
0.092
0.50
0.90
1.2
0.695
0.225
0.068
0.029
0.032
0.062
0.26
0.95
1.2
1.2
0.5
0.23
0.041
0.070
0.41
0.8
1.1
0.66
0.16
0.046
0.0215
0.026
0.046
0.16
0.85
1.0
1.1
0.42
0.15
0.042
0.072
0.41
0.86
1.2
0.82
0.165
0.043
0.022
0:026
0.044
0.13
0.77
1.0
1.05
0.47
0.145
0.037
0.063
0.33
0.73
1.1
0.71
0.135
0.032
0.017
0.02
0.033
0.09
0.73
0.60
1.0
0.32
0.095
U)
I
-------
Appendix D.I (continued)
Mixture Composition Ozone Content (ppm) at Different Sampling Stations
ppm ppm 1 2 4 6 8 10 12
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
0.5
1.0
2.0
3.0
5.0
10.0
12.5
20.0
30.0
50.0
0.07
0.12
0.143
0.17
0.34
0.8
1.8
0.56
0.4
0.23
0.035
0.061
0.074
0.106
0.35
1.75
1.9
1.4
0.84
0.42
0.015
0.027
0.035
0.049
0.148
1.55
1.55
1.25
0.67
0.33
0.01
0.018
0.024
0.034
0.095
1.45
1.45
1.25
0.59
0.25
0.008
0.013
0.018
0.024
0.068
1.3
1.3
1.2
0.46
0.17
0.007
0.012
0.017
0.025
0.068
1.4
1.4
1.35
0.50
0.15
0.007
0.009
0.012
0.018
0.042
0.90
0.80
0.90
0.30
0.09
-p.
I
-------
Appendix D.I (continued)
Mixture Composition Ozone Content (ppm) at Different Sampling Stations
L ~£JQ
ppm
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
L O 1 \J J/v
ppm
0.5
1.0
2.0
3.0
5.0
10.0
15.0
20.0
25.0
30.0
50.0
1
0.054
0.039
0.039
0.036
0.037
0.125
0.35
0.70
0.95
0.70
0.50
2
0.014
0.01
0.009
0.015
0.01
0.055
0.68
1.50
1.70
1.60
1.05
4.
0.0055
0.004
0.004
0.009
0.004
0.027
0.62
1.25
1.4
1.5
0.97
6
0.0035
0.001
0.002
0.007
0.002
0.022
0.53
1.05
1.1
1.4
0.95
8
0.003
0.0
0.001
0.006
0.001
0.017
0.51
0.85
0.9
1.3
0.87
10
0.0025
0.0
0.001
0.005
0.001
0.017
0.55
0.90
0.95
1.4
0.90
12
0.002
0.0
0.001
0.005
0.001
0.012
0.37
0.60
0.70
0.90
0.60
-------
Q
Appendix 0.2 Nuclei (N) and Aerosol (A) Contents (no. cm ) Along the Tunnel
Mixture Composition
Sampling Station
[N02]
ppm
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
1.0
1.0
1.0
1.0
1.0
1.0
1.0
ppm
0.5
1.0
1.25
2.0
3.0
5.0
10.0
20.0
0.5
1.0
2.0
3.0
5.0
10.0
20.0
3
N
250
350
650
400
450
400
< 200
< 200
< 200
250
480
550
450
400
300
A
0.0
0.0
0.0
17
95
170
230
152
0.0
0.0
1
159
389
512
424
5
N
250
400
650
500
500
400
350
^200
<200
300
650
700
550
400
300
A
0.0
0.0
0.0
71
240
356
374
272
0.0
0.0
3
243
555
664
565
7
N
250
550
600
700
550
450
300
<200
<200
350
700
750
520
350
250
A
0.0
0.0
0.0
92
314
424
388
300
0.0
0.0
6
216
572
505
424
9
N
250
550
500
650
650
400
250
<200
300
300
600
650
450
350
350
A
0.0
0.0
0.0
99
339
424
360
293
0.0
0.0
7
219
551
576
530
11
N
300
600
800
600
600
500
250
250
<200
<200
700
700
450
400
300
A
0.0
0.0
0.0
85
293
424
353
279
0.0
0.0
14
184
459
530
671
-------
Appendix D.2 (continued)
Mixture Composition
Sampling Station
[N02]Q
ppm
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
[C6H10]0
ppm
0.5
1.0
2.0
3.0
5.0
10.0
20.0
50.0
0.5
1.0
2.0
3.0
5.0
7.5
10.0
20.0
30.0
3
N
< 200
-------
Appendix D.2 (continued)
Mixture
[N02]0
ppm
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
Composition
[C6H10]0
ppm
0.5
1.0
2.0
3.0
5.0
10.0
12.5
20.0
30.0
50.0
3
< 200
* 200
<200
250
500
1400
2000
1600
1000
650
A
0.0
0.0
0.0
0.0
0.0
155
152
290
300
212
5
N
<2QQ
<200
<200
300
400
1300
1500
1500
800
705
A
0.0
0.0
0.0
0.0
0.0
166
145
240
261
237
*>OIII|J 1 II
N
<200
< 200
<200
300
500
1400
1200
1100
800
680
iy j ia i. n
A
0.0
0.0
0.0
0.0
0.0
124
95
120
198
198
9
N
<200
< 200
<200
300
400
1500
1200
900
850
500
A
0.0
0.0
0.0
0.0
0.0
138
99
134
346
565
11
N
<200
<200
<200
300
500
1600
800
1200
900
600
A
0.0
0.0
0.0
0.0
0.0
124
81
124
636
813
00
-------
Appendix 0.2 (continued)
Mixture Composition
[N02]Q [C6Hlo]0
ppm ppm
Sampling Station
9
A N
11
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
0.5
1.0
2.0
3.0
5.0
10.0
15.0
20.0
25.0
30.0
50.0
<200
< 200
< 200
< 200
< 200
500
1500
2000
2600
2200
1400
0.0
0.0
0.0
0.0
0.0
0.0
0.0
81
60
187
212
<200
<200
<200
<200
<200
500
1300
1600
2400
2200
1200
0.0
0.0
0.0
0.0
0.0
0.0
0.0
74
53
130
191
<200
<200
<200
<200
<200
350
1400
1600
2000
1700
1000
0.0
0.0
0.0
0.0
0.0
0.0
0.0
49
34
53
95
<200
<200
^200
<200
<200
250
1300
1500
2100
1700
1000
0.0
0.0
0.0
0.0
0.0
0.0
0.0
56
35
64
173
<200
< 200
<200
<200
<200
400
1200
1400
2200
1600
1300
0.0
0.0
0.0
0.0
0.0
0.0
0.0
49
31
60
460
I
«J
vo
-------
-120-
Appendix E Estimated Peak Values of Ozone, Nuclei and
Aerosols for Different Initial Mixture Compositions
[N02]Q
ppm
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
1.0
1.0
1.0
1.0
1.0
1.0
1.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
ppm
0.5
1.0
1.25
2.0
3.0
5.0
10.0
20.0
0.5
1.0
2.0
3.0
5.0
10.0
20.0
0.5
1.0
2.0
3.0
5.0
10.0
20.0
50.0
[03]
ppm
0.3
0.6
0.7
0.65
0.50
0.33
0.185
0.1
0.17
0.375
0.9
1.0
0.68
0.38
0.19
0.0
0.15
0.585
0.98
1.3
0.81
0.42
0.16
Nuclei
_3
no. cm
250
550
650
580
450
350
250
<200
< 200
350
700
750
550
400
300
<200
< 200
450
700
900
650
450
<: 200
Aerosols
no. cm
0.0
0.0
0.0
99
318
424
389
300
0.0
0.0
14
247
565
636
565
0.0
0.0
0.0
2.0
166
495
565
353
-------
-121-
Appendix E (continued)
ppm °
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
ppm °
0.5
1.0
2.0
3.0
5.0
7.5
10.0
20.0
30.0
0.5
1.0
2.0
3.0
5.0
10.0
12.5
20.0
30.0
50.0
0.5
1.0
2.0
3.0
5.0
10.0
15.0
20.0
25.0
30.0
50.0
[03]
ppm
0.0
0.0
0.2
0.52
1.2
1.67
1.33
0.72
0.46
0.0
0.0
0.0
0.0
0.48
1.87
2.0
1.42
0.88
0.46
0.0
0.0
0.0
0.0
0.0
0.0
0.70
1.55
1.8
1.6
1.05
Nuclei
no. cm"-'
<200
< 200
350
600
1000
1200
1100
600
500
< 200
< 200
< 200
400
700
1500
1800
1600
1000
650
< 200
<200
< 200
< 200
400
800
1500
2200
2600
2200
1400
Aerosols
no. cm- 3
0.0
0.0
0.0
0.0
10
46
343
424
406
0.0
0.0
0.0
0.0
0.0
166
152
290
636
707
0.0
0.0
0.0
0.0
0.0
0.0
0.0
81
60
187
240
-------
-122-
REFERENCES
1. Altshuller, A.P., and Cohen, I.R., "Structural Effects on the
Rate of Nitrogen Dioxide Formation in the Photo-Oxidation of Or-
ganic Compound-Nitric Oxide Mixtures in Air," International
Journal of Air and Water Pollution, ]_, 787-797 (1963).
2. Altshuller, A.P., and Cohen, I.R., "Atmospheric Photo-oxidation
of the Ethylene-Nitric Oxide System," International Journal of
Air and Water Pollution, 8, 611-632 (1964).
3. Altshuller, A.P., Klosterman, D.L., Leach, P.W., Hindawi, I.J.,
and Sigsby, J.E., Jr., "Products and Biological Effects from
Irradiation of Nitrogen Oxides with Hydrocarbons or Aldehydes
Under Dynamic Conditions," International Journal of Air and
Water Pollution, 1_0, 81-98 (1966).
4. Altshuller, A.P., "Reactivity of Organic Substances in Atmospheric
Photooxidatior1 Reactions," International Journal of Air and Water
Pollution, 1£, 713-733 (1966).
5. Altshuller, A.P., Kopczynski, S.L., Lonneman, W.A., Becker, T.L.,
and Slater, R., "Chemical Aspects of the Photooxidation of the
Propylene-Nitrogen Oxide System," Environmental Science and
Technology, 1, 899-914 (1967).
6. Altshuller, A.P., Kopczynski, S.L., Wilson, D., Lonneman, W.,
and Sutterfield, F.D., "Photochemical Reactivities of N-Butane
and Other Paraffinic Hydrocarbons," Journal of the Air Pollution
Control Association, J9.' 787-790 (1969).
7. Altshuller, A.P., Kopczynski, S.L., Lonneman, W.A., Sutterfield,
F.D., and Wilson, D.L., "Photochemical Reactivities of Aromatic
Hydrocarbon-Nitrogen Oxide and Related Systems," Environmental
Science and Technology, 4_, 44-49 (1970).
8. Altshuller, A.P., and Bufalini, J.J., "Photochemical Aspects of
Air Pollution: A Review," Environmental Science and Technology,
5_, 39-64 (1971).
9. Bufalini, J.J., and Altshuller, A.P., "The Effect of Temperature
on Photochemical Smog Reactions," International Journal of Air
and Water Pollution, £, 769-771 (1963).
10. Bufalini, J.J,, and Altshuller, A.P., "Oxidation of Nitric Oxide
in the Presence of Ultraviolet Light and Hydrocarbons," Environ-
mental Science and Technology, 3, 469-472 (1969).
11. Cartwright, J., Nagelschmidt, G., and Skidmore, J.W., "The Study
of Air Pollution with the Electron Microscope," Quarterly Journal
of the Royal Meteorology Society, 82., 82-86 (1956).
-------
-123-
12. Clyne, M.A.A., Thrush, B.A., and Nayne, R.P., "Kinetics of the
Chemiluminescent Reaction between Nitric Oxide and Ozone," Trans-
action of the Faraday Society, 60, 359-370 (1964).
13. Cox, R.A., and Penkett, S.A., "The Photooxidation of Sulfur Diox-
ide in Sunlight," Atmospheric Environment, 4^ 425-433 (1970).
14. Cox, R.A., "Particle Formation from Homogeneous Reactions of Sulfur-
Dioxide and Nitrogen Dioxide," Tullus, 26, 235-240 (1974).
lb. Demerjian, K.L., Kerr, J.A., and Calvert, J.G., "The Mechanism of
Photochemical Smog Formation," in Advances in Environmental Science
and Technology, edited by Pitts, J.N., Metcalf, R.L., and Lloyd,
A.C., 4_, 1-262 (1974), John Wiley & Sons, New York.
16. Dimitriades, B., "Methodology in Air Pollution Studies Using Ir-
radiation Chambers," Journal of the Air Pollution Control Assoc-
iation, ]_7, 460-466 (1967).
17. Dunham, S.B., "Detection of Photochemical Oxidation of Sulfur
Dioxide by Condensation Nuclei Techniques," Nature, 188, 51-52
(1960).
18. Endow, tCr Doyle, G.J., and Jones, J.L., "The Nature of Some Model
Photochemical Aerosols," Journal of the Air Pollution Control
Association, ]_3, 141-147 (1963).
19. Eschenroeder, A.Q., Martinez, J.R., "Concepts and Applications
of Photochemical Smog Models," in "Photochemical Smog and Ozone
Reactions," edited by Gould, R.F., Advances in Chemistry Series.
113, 107-168 (1972), American Chemical Society, Washington, D.C.
20. Ford, H.W., Doyle, G.J., and Endon, N., "Rate Constants at Low
Concentrations. II Reaction between Nitric Oxide and Ozone in Air
at Room Temperature," Journal of Chemical Physics, .26, 1337 (1957).
21. Friedlander, S.K., "Similarity Considerations for the Particle-
Size Spectrum of a Coagulating, Sedimenting Aerosol," Journal of
Meterology, V7, 479-483 (I960).
22. Friedlander, S.K., and Pasceri, R.E., "Measurement of the Particle
Size Distribution of the Atmospheric Aerosol: I. Introduction and
Experimental Methods," Journal of Atmospheric Science, 22, 571-
576 (1965).
23. Fuchs, N.A., "The Mechanics of Aerosols," The Macmillan Company,
New York (1964).
24. Gerhard, E.R., and Johnstone, H.F., "Photochemical Oxidation of
Sulfur Dioxide in Air," Industrial and Engineering Chemistry,
47, 972-976 (1955).
-------
-124-
25. Glasson, W.A., and Tuesday, C.S., "Inhibition of the Atmospheric
Photooxidation of Hydrocarbons by Nitric Oxide," Research Publi-
cation GMR-475, General Motors Research Laboratories, Warren,
Michigan (1965).
26. Glasson, W.A., and Tuesday, C.S., "Hydrocarbon Reactivities in the
Atmospheric Photooxidation of Nitric Oxide," Environmental Science
and Technology, 4_, 916-924 (1970).
27. Goetz, A., and Pueschel, R., "Basic Mechanisms of Photochemical
Aerosol Formation," Atmospheric Environment, ]_, 287-306 (1967).
28. Goetz, A., and Klejnst, O.J., "Formation and Degradation of Aero-
colloids by Ultraviolet Radiation," Environmental Science and
Technology, 6_, 143-151 (1972).
29. Groblicki, P.O., and Nebel, G.J., "The Photochemical Formation of
Aerosols in Urban Atmospheres," in "Chemical Reactions in Urban
Atmospheres," edited by Tuesday, C.S., 241-264 (1971), American
Elsevier Publishing Company, New York.
30. Harkins, J., and Nicksic, S.W., "Studies on the Role of Sulfur
Dioxide in Visibility Reduction," Journal of the Air Pollution
Control Association, 1_5, 218-221 (1965).
31. Harton, E.E., Jr., and Bolze, C.C., "Eye Irritation from Solar
Radiatton of Organic Compounds and Nitrogen Dioxide," Report No.
23, Air Pollution Foundation, San Marino, California (1958).
32. Hecht, T.A., and Seinfeld, J.H., "Development and Validation of a
Generalized Mechanism for Photochemical Smog," Environmental Science
and Technology, 6, 47-57 (1972).
33. Heuss, J.M., and Glasson, W.A., "Hydrocarbon Reactivity and Eye
Irritation," Environmental Science and Technology, 2, 1109-1116
(1968).
34. Holmes, J.R., O'Brien, R.J., Crabtree, J.H., Hecht, T.A., and
Seinfeld, J.H., "Measurement of Ultraviolet Radiation Intensity
in Photochemical Smog Studies," Environmental Science and Technology,
7_, b!9-523 (1973).
35. Husar, R.B., and Whitby, K.T., "Growth Mechanisms and Size Spectra
of Photochemical Aerosols," Environmental Science and Technology,
7., 241-247 (1973).
36. Jaffe, R.J., and Smith, F.C., Jr., "Factors Affecting Reactions in
Smog Chambers," Lockheed Missiles & Space Company, Inc., Sunnyvale,
California. Presented at'the 67th Annual Meeting of the Air Pollu-
tion Control Association, Denver, Colorado, June 13, 1974.
37. Johnston, ri.S., and Crosby, H.J., "Kinetics of the Fast Gas Phase
Reaction between Ozone and Nitric Oxide," Journal of Chemical
Physics, 22, 689-692 (1954).
-------
-125-
36. Junge, C.E., "Atomspheric Chemistry," in Advances In Geophysics.
edited by Lansberg, H.E., and Van Kieghem, J., 4_, 1-101 (1958),
Academic Press, Mew York.
39. Katz, M., and Gale, S.B., "f'.echanism of Photooxidation of Sulfur
Dioxide in Atmosphere," Proceedings of the Second International
Clean Air Congress, edited by Englund, H.f'., and Beery, W.T.,
336-343 (1971), Academic Press, New York.
40. Kocmond, W.C., Kittelson, D.B., Yang, J.Y., arid Denerjian, K.L.,
"Determination of the Formation Mechanisms and Compositions of
Photochemical Aerosols," Calspan Corporation, Buffalo, Hew York.
First Annual Summary Report, 31 August 1973. Prepared for Environ-
nental Protection Agency, Durham, North Carolina and Coordinating
Research Council, Mew York, NewYork.
41. Kopczynski, S.L., Altshuller, A.P., and Sutterfield, F.U., "Photo-
chemical Reactivities of Aldehyde - Nitrogen Oxide Systems,"
Environmental Science anc Technology, J3, 909-918 (1974).
42. Laity, J.L., "A Smog Chamber Study Comparing Blocklight Fluroes-
cent Lamps with Natural Sunlight," Environmental Science and
Technology, b_, 1218-1220 (1971).
43. Leighton, P.A., "Photochemistry of Air Pollution," Academic
Press, New York (19bl).
44. Lillian, D., "Formation and Destruction of Ozone in a Simulated
natural System (Nitrogen Dioxide + «* -Pinene + hv)," in "Photo-
chemical Smog and Ozone Reactions," edited by Gould, R.F., Advances
in Chemistry Series. 113, 211-218 (1972), American Chemical Society,
Washington, D.C.
45. Liu, B.Y.H., and Whitby, K.T., "Dynamic Equilibrium in Self-Pre-
serving Aerosols," Journal of Colloid and Interface Science, 26,
lfal-165 (1968). ~"~
4b. iiiki, H., Daby, E.E., and Weinstock, B., "Mechanisms of Smog
Reactions," in "Photochemical Smog and Ozone Reactions," edited
by Gould, R.F., Advances in Chemistry Series. 113, 16-57 (1972),
American Chemical Society, Washington, D.C.
47. Prager, M.J., Stephens, E.R., and Scott, W.E., "Aerosol Formation
from Gaseous Air Pollutants," Industrial and Engineering Chem-
istry, 52_, 521-524 (1960).
48. Quon, J.E., Siegel, R.P., and Hulburt, H.M., "Particle Formation
from Photooxidation of Sulfur Dioxide in Air," Proceedings of the
Second International Clean Air Congress, edited by Englund, H.M.,
and Keery, W.T., 330-335 (1971), Academic Press, Mew York.
4$. Renzetti, N.A., and Doyle, G.J., "Photochemical Aerosol Formation
in Sulfur Dioxide-Hydrocarbon Systems," International Journal of
Air Pollution, 2, 327-345 (1960).
-------
-126-
bG. Rogers, L.H., ed. , "Proceedings of the Conference on Chemical
Reactions in Urban Atmospheres," Report No. lb, Air Pollution
Foundation, Los Angeles, California, 1956.
bl. Romanovsky, J.C., Ingels, R.M. , and Gordon, R.J., "Estimation of
Smog Effects in the Hydrocarbon-Nitric Oxide Systeir," Journal of
the Air Pollution Control Association, 1_7_, 454-45S (1967).
62. Schofield, K. , "An Evaluation of Kinetic Rate Uata for Reactions
of Neutrals of Atmospheric Interest," Planetary and Space Science,
]5., 643-670 (1967).
b3. Schuck, E.A. , Doyle, G.J., and Endow, N. , "A Progress Report on
the Photochemistry of Polluted Atmospheres," Report No. 31, Air
Pollution Foundation, San Marino, California, 1960.
b4. Schuck, E.A. , and Stephens, E.R., "Oxides of Nitrogen," in
Advances in Environmental Science and Technology, edited by Pitts,
J.i, ..and Metcalf, R.L., 1, 73-117 (1969), John Wiley i Sons,
Mew York.
b5. Seinfeld, J.H., Hecht, T.A. , and Roth, P.M., "Existing fieeds in
the Experimental and Observational Study of Atmospheric Chen.ical
Reactions," Report Mo. EPA-R4-73-031, Systems Applications, Inc.,
deverly hills, California, Prepared for U.S. Environmental Protec-
tion Agency, 1973.
56. Stedman, D.H. , Daby, E.E., Stuhl , F., and Niki, H., "Analysis of
Ozone and Nitric Oxide by a Cliemi luminescent Method in Laboratory
and Atmospheric Studies of Photochemical Smog," Journal of the Air
Pollution Control Association, 22_, 260-263 (1972).
b7. Stedman, D.H. , and l.iki, H. , "Photolysis of NO? in Air as Measure-
ment Method for Light Intensity," Environmental Science and
Technology, 7., 735-739 (1973).
58. Stephens, E.R. , and Price, H.A., "Atmospheric Photochemical Reac-
tions in a Tube Flow Reactor," Atmospheric Environment, 3, 573-
(1969).
b9. Stevenson, H.J.R. , Sanderson, D.E., and Altshuller, A. P., "Forma-
tion of Photochemical Aerosols," International Journal of Air and
Water Pollution, £, 3C7-375 (1965).
60. Tuesday, C.S., "The Atmosphere Photooxidation of Trans-Butene-2
and Nitric Oxide," in "Chemical Reactions in the Lower and Upper
Atmosphere," edited by Cadle, R.D. , 15-49 (1961), Interscience,
Mew York.
Ol. Wayne, L.G. , Weisburd, II. , Danchick, R., and Kokin, A., "Final
Report- Development of a Simulation Model for Estinating Grounc
Level Concentrations of Photochemical Pollutants," Technical
Memorandum, System Development Corporation, Santa Monica, California,
(1971).
-------
-127-
62. Westberg, K., and Cohen, N., "The Chemical Kinetics of Photo-
chemical Smog as Analyzed by Computer," ATR-70(8107)-1, The Aero-
space Corporation, LI Segundo, California (1969).
03. Wilson, Wm. E., Jr., Merryrcan, E.L., and Levy, A., "A Litera-
ture Survey of Aerosol Formation and Visibility Reduction in Photo-
chemical Smog," Battelle Memorial Institute, Columbus, Ohio,
Project EF-2, August 1, 1969, Prepared for the American Petroleum
Institute, Committee for Air and Water Conservation.
64. Wilson, W.E., Jr., Levy, A., and Wimmer, D.B., "A Study of Sulfur
Dioxide in Photochemical Smog II. Effect of Sulfur Dioxide on
Oxidant Formation in Photochemical Smog," Journal of the Air
Pollution Control Association, 22_, 27-32 (1972).
65. Wilson, W.E., Jr., Merryman, h.L., Levy, A., and Taliaferro, H.R.,
"Aerosol Formation in Photochemical Smog I. Effect of Stirring,"
Journal of the Air Pollution Control Association, 21, 128-132
(1971).
66. "Air Quality Criteria for Particulate Matter," National Air Pollu-
tion Control Administration, Publication No. AP-49, 1969, U.S.
Department of Health, Education, and Welfare, Washington, D.C.
67. "Air Quality Criteria for Sulfur Oxides," National Air Pollution
Control Administration Publication No. AP-50, January 1969,
U.S. Department of Health, Education, and Welfare, Washington, D.C.
68. "Air Quality Criteria for iJitrogen Oxides," Pollution Control
Uffice Publication ho. AP-G4, 1971, U.S. Environmental Protection
Agency, Washington, D.C.
69. "Air Quality Criteria for Photochemical Oxidants," National Air
Pollution Control Administration Publication No. AP-63, 1970, U.S.
Department of Health, Education and Welfare, Washington, D.C.
70. "Air Quality Criteria for Hydrocarbons," National Air Pollution
Control Administration Publication No. AP-64, 1970, U.S. Depart-
ment of Health, Education and Welfare, Washington, D.C.
-------
128
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO
EPA-460/3-76-Q05
3 RECIPIENT'S ACCESSION-NO
4 TITLE AND SUBTITLE
Ozone and Particulate Formation In Photochemical
Reactions
5 REPORT DATE
April 1Q75
6 PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
Chi-Hung Shen
George S. Springer
8 PERFORMING ORGANIZATION REPORT NO
9 PERFORMING ORGANIZATION NAME AND ADDRESS
The University of Michigan
Ann Arbor, Michigan 48104
10 PROGRAM ELEMENT NO.
11 CONTRACT/GRANT NO
R-801476
12 SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air and Waste Management
Office of Mobile Source Air Pollution Control
2565 Plymouth Road
Ann Arbor. Michigan 48105
13 TYPE OF REPORT AND PERIOD COVERED
TOPIC Final
14 SPONSORING AGENCY CODE
15 SUPPLEMENTARY NOTES
16. ABSTRACT
Experiments were performed to investigate the photochemical formation of
nuclei and aerosols in sulfur dioxide and air mixtures; and also the photo-
chemical formation of ozona,. nuclei, and aerosols in cyclohexene, nitrogen
dioxide, and air mixtures. These mixtures were irradiated with ultraviolet
radiation while they flowed at a constant speed through a pyrex tubulation
measuring 15.2 cm inside diameter by 9.12 meters long. The quantities of ozone,
nuclei, and aerosols generated in the gas - air mixtures were measured at
various points along the tubulation by means of a chemiluminescent analyzer, a
condensation nuclei counter, and an aerosol counter.
Photo-oxidation experiments with sulfur dioxide and air mixtures were
performed for five different initial sulfur dioxide concentrations of 0.2, 0.3,
0.5, 0.7, and 1.0 ppm; and for four different relative humidities of 25, 50, 75,
and 100%. Test results showed that the maximum number of nuclei produced was
directly proportional to the initial sulfur dioxide concentration, and to the
square of the relative humidity. There were no measurable amounts of aerosols
observed during the tests.
Experiments with cyclohexene, nitrogen dioxide, and air mixtures were
performed with the initial cyclohexene concentration ranging from 0.5 to 50 ppm;
and the initial nitrogen dioxide concentratinn fmm n fi tn in m
17
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air pollution
Particulates
Ozone
Photochemical
Aerosols
Nuclei
b IDENTIFIERS/OPEN ENDED TERMS
L COSATI I ield/Gr. up
13B
n DISTRIBUTION STATEM&M
Unlimited
19 SECURITY CLASS (//in Kipar
Unclassified
20 SECURITY CLASS \flinpaKf)
Unclassified
21 NO OF PAGES
135
22 PRICE
EPA Form 2220-1 (9-73)
------- |