United States
Environmental Protection
Agency
Environmental Sciences Research EPA 600 3-79-006
Laboratory January 1979
Research Triangle Park NC 27711
Research and Development
Laboratory
Investigation of the
Photooxidation and
Catalytic
Oxidation of SCb
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U S Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields
The nine series are
1 Environmental Health Effects Research
2 Environmental Protection Technology
3 Ecological Research
4 Environmental Monitoring
5 Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8 "Special" Reports
9 Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials Problems are assessed for their long- and short-term influ-
ences Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161
-------
EPA-600/3-79-006
January 1979
LABORATORY INVESTIGATION OF -THE PKOTOOXIDATION
AND CATALYTIC OXIDATION OF S00
by
R.J. Anderson, R.J. Pilie
E.J. Mack, and W.C. Kocmond
Calspan Corporation
Advanced Technology Center
Buffalo, New York 14225
Contract No. 68-02-1785
Project Officer
B. Dimitriades and J.L. Durham
Atmospheric Chemistry and Physics Division
National Environmental Research Center
Research Triangle Park, N.C. 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, N.C. 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute endorsement
or recommendation for use.
ii
-------
ABSTRACT
Experimental studies of SO. oxidation performed in Calspan's 590 m smog
chamber involved the study of the photooxidation of SO- in irradiated auto
exhaust + SO- systems, the catalytic oxidation of SO- in the solution drop-
lets of hazes, clouds and fogs containing several concentrations of heavy
metals (Mn and Fe ), and SO. oxidation in irradiated hazes and fogs con-
taining only trace quantities of heavy metals.
The results of the irradiated auto exhaust + SO. systems revealed that
the use of leaded fuels produced increased numbers of primary particulates
and reduced visibilities as compared to unleaded fuels. Systems containing
high hydrocarbon concentrations (HC/NO >50/1) revealed an apparent enhanced
-1 x
rate of S02 photooxidation (>5% hr ) during about the first two hours of the
irradiation. Following this initial period of enhanced SO. photooxidation,
little or no further SO- conversion was observed. The primary particulates
of the auto exhaust produced no discernible effect on the oxidation of SO-.
The catalytic oxidation of SO- in fogs and hazes was found to be
significant under certain conditions. In nonirradiated dense hazes (b
-2 -1 3 +2 scat
on the order of 10 m ) containing about 1 yg/m each of Mn and
Fe+ , as much as 80 yg/m of sulfate was formed within 30 minutes after the
introduction of S02< This quantity of sulfate corresponds to the oxidation
of almost 3% of the available SO- (0.75 ppm). Fog water samples OlO ml)
^ 3
from simulated natural fogs (LWC ^0.3 g/m ) contained about 360 yg/ml of
+2 +3
sulfate for measured concentrations of Mn and Fe of about 2 yg/ml each
and an ambient SO concentration of about 0.4 ppm. This sulfate concentra-
tion corresponds to over 6% conversion of the available SO-.
No synergistic effect on the rate of S02 oxidation was observed from
the simultaneous presence of irradiation and a fog or haze. The measured
1.11
-------
-4 -1
rate of SO. oxidation in irradiated hazes (b <3 x 10 m ) containing
I scat
only trace concentrations of heavy metals agreed well with the observed
rate of SO. photooxidation in irradiated clean air systems (^0.2% hr ). In
subsequent irradiated fog experiments also containing only trace concentra-
tions of heavy metals, the observed sulfate formation appeared to be con-
sistent with that expected from summing the sulfate formation found in
similar nonirradiated fogs with that measured in irradiated clean air
experiments.
This report was submitted in fulfillment of 68-02-1785'by Calspan
Corporation under the sponsorship of the U.S. Environmental Protection
Agency. This report covers a period from July 1, 1975 to January 1, 1978,
and work was completed as of January 30, 1978.
iv
-------
CONTENTS
Abstract iii
Figures vi
Tables vii
Abbreviations and Symbols viii
Acknowledgments ix
1. Introduction 1
2. Conclusions 4
3. Recommendations 8
4. Experimental Facilities 10
5. Aerosol Sampling and Sulfate Analysis 15
6. SO- Photooxidation in HC+NOx+S02 Systems 19
The HC+NO +S0~ Systems 19
Jt M
Auto Exhaust Experiments 25
7. Catalytic SO- Oxidation in Clouds, Fogs and Hazes 34
Haze and Fog Generation..«4...«.*««..«...«^.**k<4>««.36
Haze Experiments 37
Nonirradiated Haze Experiments 37
Irradiated Haze Experiments 46
Discussion of Results 49
In-Fog Experiments 51
Results 61
Discussion of Results 67
References 70
Appendix A Microphysical Characteristics of Selected Laboratory Fogs..73
-------
FIGURES
Number Page
1 Cut -Away View of Calspan Environmental Simulation Facility ......... 11
2 Inside View of Calspan 's 600 m Photochemical Aerosol Chamber ...... 12
3 Schematic Drawing of Heated Sampling Inlet ......................... 18
4 Sulfate Concentrations Versus Time for the Indicated Irradiated
Auto Exhaust + S02 Experiments ......................... ". ........ 29
5 Rate of S02 Oxidation Versus Time for Indicated Irradiated Auto
Exhaust + SO Experiments ....................................... 30
6 Rate of SO- Oxidation Versus Time for the Indicated Irradiated
Auto Exhaust + SO- Experiments ................................ ... 31
7 Rate of SO- Oxidation Versus Time for the Irradiated Auto
Exhaust + SO- Experiments
8 Parameters for Hazes Nebulized from Aqueous Solutions Containing
10~3M and 10~2M NaCl ............................................ 40
9 Cumulative Size Distributions for Hazes Nebulized from Aqueous
Solutions Containing 10~3M and 10~2M NaCl for Specified
Relative Humidities ............................................. 41
10 Typical Diffusion Chamber Data Showing CCN Concentration for
0.3% Supersaturation from Aerosols Produced by the Nebulization
of an Aqueous Solution Containing 10~3M and 10~2M NaCl .......... 44
11 Microphysical Characteristics of Fog #100/1 as Functions of Time... 58
12 Dependence of Sulfate Content of Fog Water on SO- Concentration
and Heavy Metals .................................. . ............. 64
-------
TABLES
Number Page
1 Complete Log of Calspan Smog Chamber Experiments 20
2 Summary of Chemistry Data from HC+NO and HC+NO +S0_ Irradiations... 23
3 Summary of Aerosol Data from HC+NO and HC+NO +SO- 24
X 2C £
4 Log of Auto Exhaust Experiments 26
5 Auto Exhaust Experiments - Chemistry Data 28
6 Haze Experiments 39
7 Sulfate Formation in Pre-fog Hazes Indicating Dependence on
Heavy Metals 45
8 Irradiated Haze Experiments 47
9 In-Fog Experiments 52
10 Irradiated In-Fog Experiments 54
11 Mean Fog Microphysical Properties 59
12 Chemistry of/ In-Fog Experiments 62
VII
-------
ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
AE auto exhaust
CID cubic inch displacement
CCN cloud condensation nuclei
CYCLO - cyclohexene
EAA electrical aerosol analyzer
HC hydrocarbon
LWC liquid water content
N maximum observed number of aitken nuclei
PROP propylene
RH relative humidity
Rg02 rate of oxidation of sulfur dioxide
r reaction
A
w/ with
w/o without
XRF x-ray fluorescence
Vlll
-------
ACKNOWLEDGMENT
The authors wish to acknowledge the many helpful discussions with and
the support of Dr. Jack L. Durham, Technical Project Monitor. Dr. Richard
Keppler, Director of ORD in EPA Region 1, also provided information and
guidance in the performance of the project. Appreciation is expressed for
the X-ray fluorescence analysis provided by R. B. Kellogg and J. M. Lang of
Northrop Services, Inc. and ion chromatography performed by Melville Richards
of Northrop Services, Inc. and Kevin Peterson of the Environmental Protection
Agency at Research Triangle Park, North Carolina. Many individuals at
Calspan unselfishly provided time and effort during the performance of this
project. They include Thomas Niziol, John Michalovic, C. William Rogers,
Dr. Ulrich Katz, and Dr. John Yang. The text was typed by Mrs. Marilyn T.
Handley.
-------
SECTION 1
INTRODUCTION
Atmospheric gas to particulate conversion processes are widely recognized
as complicating factors in the development of air quality criteria and asso-
ciated emission control strategies. The ubiquitous presence of SO. in the
urban atmosphere together with the effects of sulfuric acid and sulfate aero-
sols make the understanding of S0_ oxidation in the ambient atmosphere
especially important. As a part of a broad base program under the auspices of
the EPA, Calspan performed smog chamber simulation studies to assess the S0_
oxidation rates and mechanisms in varying ambient environments. Early in the
program, emphasis was placed on investigating chemical and aerosol behavior
under homogeneous photooxidation conditions. In the latter stages of the
program, experiments were directed toward studies of solution droplet cata-
lytic effects on SO- oxidation as might be expected under in-cloud trans-
formation and power plant plume conditions. In addition, considerable effort
was expended to develop and refine reliable sulfate sampling and analysis
procedures.
The S0_ photooxidation rate in clean filtered air as determined by smog
chamber (Kocmond and Yang, 1975) and flow reactor (Kasahara and Takahashi,
1976) measurements appears to be about a few tenths of a percent per hour
under noontime sunlight irradiation at middle latitudes. In polluted atmo-
spheres with the presence of nitrogen oxides and reactive hydrocarbons, the
homogeneous photooxidation process may, however, result in SO- to aerosol
conversion rates of several percent per hour CHidy and Burton, 1975). A
mechanism of aerosol formation by OH radical reaction with SO- has been
proposed by Wood, Castleman and Tang (1975). Calvert and McQuigg (1975)
demonstrated by computer simulation that enhanced photooxidation of SO- in
-------
nitrogen oxide and hydrocarbon polluted atmospheres would be consistent with
such a proposed mechanism. The formation of intermediate molecular complexes
of S0_ with hydrocarbon and nitrogen oxides contaminants has been proposed as
another possible mechanism (Richards, Fox, and Reist, 1976) giving rise to
enhanced SO. oxidation via homogeneous photochemical processes. In view of
the current lack of understanding of the complex mechanisms involved in SO
photooxidation under polluted urban atmospheres, smog chamber determinations
of the oxidation rate under a variety of contaminant conditions are needed.
In-cloud transformation of SO to sulfate is considered as an important
route responsible for sulfur balance in the environment. One likely
transport mechanism of sulfur pollutants from the atmosphere to the earth
is the scavenging of SO during cloud droplet formation and subsequent
precipitation. During nonprecipitation events, the cloud droplets form
aloft and evaporate prior to reaching the earth. As a result, the in-cloud
SO- to sulfate transformation preceding droplet evaporation- produces
residual sulfate aerosols in the atmosphere. Junge and Ryan (1958) were
among the first to investigate the rate of S0_ oxidation in dilute aqueous
solutions and found rapid conversion in the presence of trace metal ions
which appeared to serve as catalysts. Radke (1970) apparently observed
residual sulfate aerosols in conjunction with wave clouds over Mt. Ranier,
Washington. In several instances he noted cloud condensation nuclei con-
centrations in the air leaving the wave clouds to be significantly greater
than those entering the cloud.
Another likely environment favorable to aqueous droplet SO to sulfate
conversion is that existing within the power plant stack emission plume.
A stack plume from a combustion process is typically a continuous point
source of hot gases rising vertically due to a buoyancy effect. The rising
air mass is cooled due to the expansion and mixing with cold ambient air,
which under proper conditions results in condensation. The droplets thus
formed are transported within the plume and present a reaction medium
conducive to S0_ oxidation. To date, there are no convincing atmospheric
studies that support this mechanism.
-------
Experiments were performed in the Calspan chamber to compile S0_
oxidation kinetic data under a wide range of environmental conditions likely
to be encountered in the real atmosphere. In order to ascertain the effects
of individual environmental variables on S0_ oxidation, initial efforts
emphasized the establishment of base line kinetic data under clean air
conditions for both the homogeneous gas phase and the in-cloud transformation
processes. A major effort was expended in refining procedures for collection
of sulfate aerosols free of artifact conversion effects and for analysis of
sulfate at the low concentration ranges often observed in clean air.
This report covers the entire 30 months of the current program. Con-
clusions drawn during the program are presented in Section 2 and recommenda-
tions for future related work are given in Section 3. A brief description
of experimental facilities is provided in Section 4, while a discussion of
aerosol sampling techniques and sulfate analysis procedures is presented in
Section 5. Data summaries in Section 6 cover the series of experiments on
SO. photooxidation in the presence of propylene and NO and in the presence
£f «
of auto exhaust. In Section 7, the results of the investigations of the
oxidation of SO- in haze and cloud droplets are presented. Literature
references are presented in Section 8.
-------
SECTION 2
CONCLUSIONS
Three extensive sets of experiments were performed on this program to
investigate oxidation processes of SO- and associated changes in aerosol
characteristics in the atmosphere. These included photooxidation in
HC+NO +SO systems (primarily auto exhaust + S0_ systems), catalytic oxida-
X ^ &
tion in haze and fog aerosols, and combined catalytic and photooxidation in
haze and fog systems.
Early experiments demonstrated that conventional methods for sample
collection and chemical analysis were not adequate for direct, quantitative
measurements of sulfate production under the required experimental conditions.
In related internally-supported programs, therefore, two new sampling systems
were developed at Calspan for providing samples that could be analyzed with
the most modern and sophisticated analytical equipment available at the EPA's
Research Triangle Park (RTP).
One sampling system consisted of a filter holder with a heated air inlet
that evaporates all liquid water of wet aerosols before collection on teflon
fluoropore membrane filters. This system minimized artifact sulfate formation
stimulated by wet catalytic processes on the filter surfaces. Filter samples
were subsequently analyzed for total sulfur content by RTP personnel using
energy dispersive X-ray fluorescence techniques. Occasionally, extractions
were prepared for independent analysis of soluble sulfate content by ion
chromatography, also at RTP.
The second sampling system consisted of a cloud water collector in which
cloud droplets, accelerated through an inlet of suitable aerodynamic design,
were impacted on a sintered glass rod where they coalesced and dripped into
a glass container. Cloud water samples were analyzed on site for pH and
mailed to RTP for ion chromatographic analysis of soluble sulfate.
-------
HC+NO +SO SYSTEMS
The auto exhaust experiments showed that the use of leaded fuels produced
greater quantities of particulates, both in number and volume, and con-
sequently, greater reductions in visibility than the use of unleaded fuels.
The primary particulates of auto exhaust did not measurably influence the
oxidation of SO . The measured loss of gaseous SO in these experiments was
not sufficient to account for the measured rate of production of aerosol
volume. A substantial fraction of aerosol growth, therefore, appeared to be
due to; species other than sulfates.
Nevertheless, the volumetric production of aerosol provided the only
measure of sulfate formation available prior to incorporation of the new
sampling techniques described above. The data can be correctly interpreted
as upper limits of sulfate production during those early experiments and
showed that the average rate of S0_ oxidation over the approximate 10 hour
^ _i
duration of each auto exhaust experiment (HC/NO -v-6.1) was <=-!% hour .
A
Using upgraded sampling and analysis procedures in later experiments, with
HC/NO ^50/1 and substantially better time resolution in data collection,
x -1
enhanced rates of SO. oxidation of ^5% hour were observed during the first
two hours of irradiation. Subsequently, little or no S02 oxidation or
volumetric aerosol growth was observed through the remainder of each
experiment (up to four additional hours).
HAZE AND FOG + S02 SYSTEMS
This initial laboratory investigation of the importance of catalytic
oxidation of SO. in typical wet industrial haze and urban fog demonstrated
the importance of the phenomenon on local conditions and suggests potential
significance on a broad, synoptic scale. Of equal importance, methods for
detailed investigation of the phenomenon were developed. Specific results
may be summarized as follows:
1. No discernible catalytic oxidation of SO was observed in haze with
-4 -1
bscat °f about 2*3 x 10 m and catalyst concentrations (V, Mn or
Fe) in solution droplets of 10~4M or less. Whether or not this
-------
result pertains to more dense haze (i.e., greater liquid water
content) with the same catalyst concentrations was not investigated.
Substantially enhanced oxidation rates were observed in nonirradia-
-4 -1
ted dense haze fb . from 23 to 11.5 x 10 m ) with Fe and Mn
scat ' __
catalyst in solution concentrations of the order of 10~ M. These
concentrations are equivalent to M. yg m~ , which is typical of
urban plumes, and resulted in sulfate concentrations of up to
80 iig m in the air mass within 30 minutes. Average oxidation
rates up to 5% hour~ were observed, but we suspect that the actual
rate exceeded that value in the very early portion of the experiment
and approached zero by the end of the 30 minute interval (see #5
below).
As a continuation of each dense haze experiment, fog was formed on
the haze aerosol and maintained for an additional 25 minutes.
Catalytic oxidation persisted during this period at approximately
the same average rate to yield up to 360 ug ml of sulfate in the
^ __ _ _
liquid water, corresponding to M.40 yg m~ in the air mass. The
measured concentrations of Mn and Fe in the fog water were each
^2.5 yg ml" , corresponding to typical urban plume values of
_3
1 yg m . Typical pH of the liquid water was about 3,
The haze and fog experiments demonstrated a direct dependence of
sulfate production on concentration of catalyst in the solution
droplets and on liquid water content of the aerosol. The direct
dependence of sulfate production on SO. concentration was demon-
strated in fog experiments. While data are not available for haze
experiments, a similar dependence probably exists. These data are
not sufficient to allow a determination of reaction rates.
From measurements in successive fogs produced in the same air mass,
it is apparent that all of the catalytic oxidation of SO. that does
occur in the haze-fog water occurs prior to the end of the first
25 minute-long fog. It is suspected, therefore, that the actual
rate of oxidation is more rapid than the observed average rate of
-------
per hour but that this rate did not persist even for 25 minutes.
It is further suspected that the total production of sulfate is
limited by the decrease of pH of the solution droplets which
accompanies the increasing concentration of SO*.
6. There appears to be no synergistic effect on the rate of SO
oxidation due to irradiation of fog or haze composed of catalytic
solution droplets. The amount of sulfate produced in each
experiment of this type was about equivalent to the sum of what
would be expected from independent photooxidation and catalytic
oxidation experiments in the same conditions.
This preliminary investigation of the role of the in-fog or in-cloud
oxidation of SO confirmed that the catalytic oxidation of SO. in the solution
droplets of clouds, fogs, and hazes may contribute significantly to the total
oxidation of SO. in the atmosphere. The observed dependence of the catalytic
oxidation of SO. on SO. concentration and type as well as concentration of
catalysts demonstrated that this mechanism may be responsible for the major
amount of SO. oxidation observed in urban plumes and especially in the plumes
of specific heavy industries. In such plumes, typical conditions of high SO.
and abundant potential catalysts as well as high LWC creat ideal conditions
for the catalytic oxidation of SO.. In remote areas where low levels of S0_
and potential catalytic materials are common, the photooxidation of SO. may
be the dominant mechanism for the conversion of SO. to sulfate.
-------
SECTION 3
RECOMMENDATIONS
Discrepancies between the rate of SO- oxidation as determined by XRF
analysis of the sulfur content of aerosol samples and the SO- oxidation rate
calculated from EM aerosol size distribution data indicate the need for
direct determinations of aerosol composition in future auto exhaust experi-
ments. This need is apparently the result of a mixed aerosol composition.
It is also recommended that future experiments be performed using short
time resolution (i.eJ, ^15 min) to investigate the influence of HC/NO ratio
A
on the rate of production of sulfate and aerosol as a function of time after
SO introduction into the system. These experiments should be designed to
reveal the proportion of the total aerosol produced which is attributable to
sulfate, and if possible, to investigate the nature of the nonsulfate com-
pounds contributing to aerosol growth. Pertinent related measurements should
be made to help elucidate information on the effect of the ozone peak on
photooxidation of SO .
This initial set of wet haze and fog experiments was heavily involved
with development of experimental techniques, and with attempting to determine
which experimental parameters were important by scanning catalytic oxidation
processes under widely different conditions. It is now necessary to examine
the functional relationships of sulfate production and the variables that
were demonstrated to be important on this program. It is recommended that
additional experimentation be conducted to generate these functional relation-
ships .
As with all smog chamber experiments, initial effort must be devoted
to the demonstration of repeatability of results. Based on current results,
it is very important in the future to use the shortest available time
resolution in acquisition of data so that temporal variations in the rate
-------
of catalytic oxidation of SO in wet aerosols can be determined. The greatest
potential for achieving high temporal resolution appears to rest with the use
of the fog water collector which could be modified to provide independent 1 ml
samples at approximately one minute intervals in dense fog (b >460 x
4 -1 scat
10~ m~ ) and perhaps 5 minute intervals in light fog (e.g., b ^100 x
-4 -1 -4 -1
10 m ). Under dense haze conditions (b ^23 x 10 m ), it appears
s cat
appropriate to reduce sampling time of the heated air inlet system by a factor
of at least three, providing temporal resolution of approximately 5 minutes.
Future experiments should be designed to separate some of the variables
that appear, from current results, to be important. In the broadest sense,
by the simple expedient of introducing SO. into the system after the pre-fog
haze has come to equilibrium but before expansion to produce fog, a much
better definition of oxidation rates in fog could be obtained, expecially
with the recommended improved temporal resolution.
Some of the data to date indicating the importance of catalyst concentra-
tion in solution droplets also illustrate the joint effect of two different
catalysts, which probably are synergistically more effective than either alone.
Careful consideration should be given in the future to isolating potentially
catalytic materials as well as investigating synergistic effects of mixed materials,
The existing haze data correspond to conditions in clean urban air and
contaminated urban plumes. To investigate these specifics, both liquid water
content and catalyst concentration in solution droplets were altered simul-
taneously. Future experiments investigating these variables independently
and at intermediate ranges are desirable. Again, improved temporal resolution
is required to determine the importance of duration of haze on total sulfate
production.
With the quantitative importance of individual variables established,
improved numerical modeling can be expected. Further chamber experiments in
which real world industrial and urban plumes are simulated will be important
for verification of modeling results.
-------
SECTION 4
EXPERIMENTAL FACILITIES
The Calspan smog chamber has been described in substantial detail else-
where (e.g., Kocmond et. al., 1973) and will only briefly be reviewed here.
The facility itself consists of a cylindrical chamber 9.14 meters (30 feet)
in diameter and 9.14 meters (30 feet) high enclosing a volume of 590 m
(20,800 ft ). A cut-away view of the entire facility is illustrated in
Figure 1. The 1.25 cm thick steel wall of the chamber is designed to with-
stand pressure differentials of 60 kilopascals (6.1 m of water); however, the
Pyrex cover plates of the irradiation system currently limit pressure differ-
entials to about 3.9 kilopascals (40 cm of water). The inner chamber surface
is coated with a fluoroepoxy-type urethane (developed at the Naval Research
Laboratory, Washington, D.C.) which has surface energy and reactivity proper-
ties comparable to those of FEP Teflon. Illumination within the chamber is
provided by 28.6 kw of fluorescent blacklight and sun lamps installed inside
24 lighting modules and arranged in eight vertical channels attached to the
wall of the chamber. Each lighting module contains two 40-watt sun lamps,
eight 85-watt high output black lamps, and two 215-watt specially-produced
black lamps. The lighting modules are covered with 0.5 cm thick
Pyrex glass and are air cooled and sealed from the chamber working volume in
order to minimize sources of contamination inside the chamber. Measured light
intensity using the k, [NO ] method reported by Stedman § Niki (1973) is
k, M).35 min"1. A variety of light intensities can be provided by selectively
d
turning off some of the light modules and lamps. A photograph of the interior
of the chamber with the lighting system turned on is shown in Figure 2.
Air purification for the chamber is accomplished by recirculating the air
through a series of absolute and activated carbon filters. Nearly all gaseous
contaminants and particulate matter can be removed from the chamber air in
10
-------
Figure 1. Cut-away view of Calspan Environmental Simulation Facility.
11
-------
Figure 2. Inside view of Calspan's 600 m Photochemical Aerosol Chamber.
12
-------
about four to five hours of filtration. Filtered air generally contains no
measurable Aitken particles, less than 0.1 ppm NO , 0.2 ppmC non-methane HC,
and no measurable SO or ozone.
Chamber washdown, humidification, and dehumidification systems were in
frequent use during this program's tests. The recirculating washdown system
consists of a stainless steel spray head which rotates on two axes and can
wet all of the chamber surfaces with distilled water or cleaning solution.
Generally, the procedure is to first wash the chamber surfaces with a 5%
solution of a laboratory glass cleaning agent followed by two or three
rinsings with tap water and two final rinsings with distilled water. Drying
is accomplished by fresh air flushing followed by air filtration. The wash-
down system is also used to wet the chamber wall prior to a cloud forming
experiment. Chamber humidity is increased by spraying distilled water into
the chamber from a remotely-operated spray nozzle near the chamber top.
Nuclei which are introduced by the evaporating spray droplets are removed by
absolute particle filters during the air filtration cycle. Chamber dehumidi-
fication, when needed, is accomplished by passing the chamber air over
refrigeration coils to remove excess water. The system, designed and fabri-
cated at Calspan, is capable of controlling humidity down to about 20% RH.
Instrumentation used to monitor aerosol behavior and react ant concentra-
tions within the chamber includes a Bendix Model 8002 chemiluminescent ozone
analyzer, Model 8101-B nitrogen oxides analyzer, Model 8300 total sulfur
analyzer, and the Model 8201 reactive hydrocarbon analyzer, a Hewlett-Packard
5750 gas chromatograph, a Thermo Systems Model 3030 Electrical Aerosol Analyzer
(EAA), an MRI Integrating Nephelometer, a Gardner Associates small particle
detector, a GE condensation nucleus counter, a Royco Model 220 Optical
Particle Counter, and one of Calspan's static diffusion chambers (used to
measure cloud condensation nuclei concentrations). A Meloy flame photometric
total sulfur analyzer on loan from the EPA was also used for these experiments.
A Royco Model 225 Optical Particle Counter was available during the last half
of the current program. An isokinetic sampling inlet was designed and
employed for particle sampling of chamber aerosols with minimal losses.
13
-------
During the first half of this project, aerosol collection for determina-
tions of sulfate (and occasionally nitrate and ammonium) contents was provided
by filtering selected amounts of chamber air through Pallflex 2500 QAO
Tissuqiiartz filters. A Millipore filter system (142 mm) and a Gelman hi-vol
sampler (105 mm) were used for the sample collection. As a result of incon-
sistencies in the measured concentration of sulfate from samples collected in
this manner, an alternative system was designed to reduce the possibility of
artifact sulfate formation. The new system was designed around a temperature-
controlled heated sampling inlet. For a detailed discussion of this sampling
system, refer to Section 5 entitled Aerosol Sampling and Sulfate Analysis.
During in-fog experiments, bulk samples of fog water suitable for chemi-
cal analysis were collected with Calspan's Fog Water Collector. This device
as well as several other unique fog characterization instruments and
techniques were developed at Calspan under various internal research programs
to provide means for the accurate measurements of the microphysical character-
istics of fogs. The use of the Fog Water Collector during the current program
is discussed in Section 7.
14
-------
SECTION 5
AEROSOL SAMPLING AND SULFATE ANALYSIS
Early in the project, the inconsistency of aerosol sulfate data indicated
that representative determinations of the sulfate content of the aerosols were
not being obtained. These inconsistencies were especially obvious in the
preliminary in-cloud experiments in which relatively large quantities of sul-
fates were observed in background experiments where little if any sulfate
should have been found.
Aerosol collections were initially made using a 142mm Millipore in-line
filter holder. Aerosol samples on Tissuquartz filters were taken over a 30-
minute period encompassing sample volumes of approximately 3.2 m . Subse-
quently, a Gelman hi-vol sampler was also used. With the hi-vol system,
sample volumes of approximately 10 m were passed through the filter over a
30-minute period. The barium perchlorate titration method (Fielder and
Morgan, 1960) was used for sulfate analysis.
Numerous experiments were performed to test aerosol collection methods
and microanalytical procedures. These culminated in the demonstration that
filtered S0_ injected into particle free air and drawn through Tissuquartz
filters moistened with doubly distilled water was oxidized sufficiently to
produce serious error in the experiments that we were attempting. These
experiments precipitated an internal research and development project at
Calspan which resulted in what we believe to be an acceptable aerosol sampling
technique. In addition, the wet chemical technique for sulfate analysis was
replaced by energy dispersive X-ray fluorescence analysis of filter samples
and frequent analysis of filter extractions by ion chromatography. Both of
these latter efforts were provided by the EPA at Research Triangle Park,
North Carolina and are gratefully acknowledged. Since XRF analysis provides
a measure of total elemental sulfur present on the sample, careful attention
15
-------
was required to minimize sorption of SO. and artifact sulfate formation on the
filter surfaces. Previous investigations by Piersen et. al.,(1976) and Byers
and Davis (1970) had shown that teflon membrane filters satisfied these
requirements. Thus 37mm diameter fluoropore membrane filters (0.5 ym pore
diameter) which could be analyzed directly in the XRF device were chosen for
use on the program. Use of these filters also provided samples of sufficient
quantity to permit extraction and independent analysis of soluble sulfate
content by ion chromatography for occasional verification of the interpreta-
tion of the elemental sulfur measurement as sulfate.
Additional precautions were required to minimize catalytic oxidation of
S0_ on the filters during sampling of in-cloud or haze experiments. At the
relative humidities of >90% required in the experiments, it was anticipated
that significant amounts of liquid water could potentially collect on the
filter. Two choices were considered to reduce the potential for the catalytic
oxidation of S0_ on the filter surfaces during sample collection: (1) remove
the S02 from the sample prior to its exposure to the filter; or (2) prevent
liquid water from accumulating on the filter.
The feasibility of applying an S0_ denuder such as the one described
by Durham, Wilson, and Bailey (1976) was studied. The denuder selectively
removes S0_ by diffusion to the walls of a tube coated with lead dioxide.
The residence time of the sample in the tube is chosen such that only a mini-
mal amount of the aerosols (which have much smaller diffusion coefficients
than gaseous S02) is lost to the walls while the major portion of the S02 is
scavenged. Because of flow rate requirements of the sampling system, no
acceptable denuders system could be devised.
The alternative would be to devise a method for elimination of the
collection of liquid water. In a typical worse case experimental condition,
chamber air might be slightly supersaturated at perhaps 20°C. By heating this
air to 70°C prior to filtration, the relative humidity would be reduced to
7% and no substantial quantities of liquid water would be collected. The 70°C
temperature would not produce a discernible loss of sulfuric acid.
16
-------
Following heat transfer calculations which showed that these conditions
could be met in a laminar flow inlet of reasonable configuration, a heated
sampling inlet was designed, constructed and installed in the chamber
(Anderson et. al., 1977) under Calspan Internal Research and Development
funding. A schematic diagram of the heated sampling inlet is presented in
Figure 3.
The 30mm ID stainless steel tube through which the sample is drawn is
located in the center of the vertical configuration in Figure 3. Concentric
with the stainless steel sampling tube is a teflon sleeve with 1/4" thick
walls. This sleeve serves two purposes; first, it acts as an insulator to
prevent significant heat losses into the chamber; and, secondly, it serves as
an inert outer material which will not contribute to the catalytic or photo-
chemical reactions under investigation. Heating tape is located between these
two concentric tubes. Thermocouples are located at the heater and inside the
sampling tube near the filter to monitor temperature. The temperature within
the tube is "controlled by~ari automatic temperature control circuit which is
capable of holding the tube temperature to within + 1°C following an initial
warm-up period.
For the latter stages of the program involving in-cloud (in-fog) experi-
ments, samples of cloud water (^10 ml) were obtained with Calspan's Fog Water
Collector.* Samples of fog water were collected in addition to filter
samples. The liquid water samples were analyzed by ion chromatography and
provided some of the most significant data from the in-cloud experiments (see
Section 7).
*The" Fog Water Collector draws air through four narrow vertical openings
located symmetrically about a vertical glass column. The sample velocity
corresponding to a flow rate of about 1.9 m /min is sufficient to result in
impaction upon the glass column of most of the fog droplets greater than one
micron diameter. The fog water then drips from the glass column into a glass
sample bottle.
17
-------
TO VACUUM
THERMOCOUPLE
. STAINLESS STEEL
FILTER
. TEFLON
HEATER
STAINLESS STEEL
SAMPLE
Figure 3. Schematic Drawing of Heated Sampling Inlet
18
-------
SECTION 6
SO PHOTOOXIDATION IN HC + NO + SO SYSTEMS
£* Jv £»
The first series of tests performed under this contract were devoted to
studies of the photooxidation of S02< Of particular interest were the
possible influences of specific hydrocarbons as well as particulates to the
rate of S0_ photooxidation. A complete log of the 65 tests performed as part
of this group of experiments is given in Table 1. As has been our practice
in the past, many experiments involved clean air + SO. irradiations to
establish chamber reactivity and to assure the absence of memory effects of
preceding experiments on H_SO. formation.
THE HC + NO + S02 SYSTEM
During an earlier program, a number of smog chamber experiments were
performed at Calspan using various combinations of hydrocarbon systems. The
most frequently employed test mixtures (propylene + NO systems with and with-
out added S0_) were selected on the basis that such reaction systems have
been extensively modeled and comparisons of model predictions with chamber
results would be possible. It was found from these previous experiments
(Kocmond and Yang, 1975) that apparent synergistic effects in aerosol produc-
tion and growth occur with the addition of SO- to the propylene + NO system.
4* A
The maximum aerosol volumetric growth rate [dv/dt] has always been found
max
to be greater for the mixed system than the sum of the rates for propylene +
NO and SO. alone. In the late stages of the experiment, after NO has been
X ft
completely oxidized, abundant aerosol was formed but it has not been deter-
mined explicitly what fraction of the aerosol was actually sulfate, An
additional series of tests was performed during the current phase program
using the propylene + NO system with added SO,. The cyclohexene + NO +
X £t X
S0_ system is also very reactive in terms of aerosol production and was also
tested.
19
-------
TABLE 1
COMPLETE LOG OF CALSPAN SMOG CHAMBER EXPERIMENTS
RUN NO.
A
1
2
3
4
5
6
7
8
g
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
36a
DATE
11/10/75
11/18/75
12/8/75
12/8/75
12/10/75
12/10/75
12/12/75
12/15/75
12/16/75
12/18/75
12/30/75
1/8/76
1/12/76
1/13/76
1/15/76
1/16/76
1/20/76
1/30/76
2/3/76
2/4/76
2/5/76
3/9/76
3/10/76
3/11/76
3/16/76
3/16/76
3/22/76
4/5/76
4/27/76
5/3/76
5/4/76
5/4/76
5/4/76
5/6/76
5/6/76
5/11/76
5/11/76
5/13/76
SYSTEM
LIGHT INTENSITY
SO2 + CLEAN AIR
S02 + CLEAN AIR
S02 + CLEAN AIR
SO2 + CLEAN AIR
SO2 + 60 ppm CO
C3H6 + NOX
C3H6 + NOX
C3H6 + NOx
C3H6 + NOx + S02
C3H6 + NOX + S02
CYCLOHEXENE + NOX + SO2
CYCLOHEXENE + NOX + SOj
CYCLO + PROP + NOX + SO2
AUTO EXHAUST + NO2 W/NUCLEI
AUTO EXHAUST W/NUCLEI
AUTO EXHAUST W/O NUCLEI
AUTO EXHAUST W/O NUCLEI
AUTO EXHAUST DARK rx
AUTO EXHAUST W/O NUCLEI
SO2 + CLEAN AIR
S02 + CLEAN AIR
AUTO EXHAUST (Pb): DARK rx
AUTO EXHAUST (Pb): W/NUCLEI
AUTO EXHAUST (Pb): W/NUCLEI
AUTO EXHAUST (Pb): W/O NUCLEI
AUTO EXHAUST (Pb)- W/O NUCLEI
S02+1.1 ppmNOx
SO2 + CLEAN AIR
SO2 + CLOUD (DARK ONLY)
S02 + CLOUD (DARK ONLY)
SO2 + CLOUD (DARK ONLY)
S02 + CLOUD (DARK ONLY)
S02 + CLOUD (DARK ONLY)
S02 OVER WET FILTER
SO2 ! NaCI CLOUD
SO2 + NaCI CLOUD
LIGHT INTENSITY
SO,
(ppm)
CHECK
0.15
0.38
0.31
028
0.19
0
0
0
031
0.28
0.40
034
040
0.50
0.38
0.50
0.30
0.44
034
0.55
0.65
0.78
0.52
0.62
0.44
OAO
0.50
038
0
0.18
0.25
0.50
0.70
0.95
0.45
028
CHECK
COMMENTS
k.[NO,l 0.41 mm"1 NEW LAMPS
d z
EXPERIMENTS 1-4; BACKGROUND REACTIVITY
TESTS
TEST OF OH SCAVENGING
EXPERIMENTS 6-13: TEST OF HC EFFECTS ON
X
EXPERIMENTS 14-19: AUTO EXHAUST IRRADIA-
TION WITH AND WITHOUT COMBUSTION NUCLEI -
LEADED VS. UNLEADED FUEL
BACKGROUND REACTIVITY EXPERIMENTS
EXPERIMENTS 22-26: AUTO EXHAUST IRRADIA-
TION WITH AND WITHOUT COMBUSTION NUCLEI -
LEADED FUEL
TEST N0x EFFECT ON RSO2
BACKGROUND rx ABORT EXPERIMENTS 29-40
INVOLVE IN-CLOUDSO_ ADDITIONS
CHECK OF ON-FILTER OXIDATION
IRRADIATE 50 min
IRRADIATE 70 min
kd [NO2' ~°'36 min"1
(continued)
20
-------
TABLE 1 (continued)
RUN NO.
37
38
39
40
40a
41
42
43
44
45
46
47
48
49
50
50a
50b
51
52
53
54
55
56
57
58
59
60
XI
X2
X3
X4
X5
DATE
5/16/76
5/18/76
5/18/76
5/18/76
5/28/76
5/28/76
6/9/76
6/9/76
6/10/76
6/10/76
7/16/76
7/20/76
7/23/76
7/29/76
7/29/76
7/29/76
7/29/76
8/17/76
8/17/76
8/18/76
8/18/76
8/18/76
8/19/76
8/20/76
8/20/76
8/20/76
8/25/76
1/25/77
3/15/77
3/16/77
10/18/77
10/19/77
SYSTEM
CLOUD ONLY
SO2 CLOUD
S02 + CLEAN AIR
SO2 + CLOUD (DARK)
CLEAN AIR IRRADIATION
SO2 + CLEAN AIR
SO2 + CLEAN AIR
SO2 + CLEAN AIR
SO, + CLEAN AIR
SO2 + CLEAN AIR
SO2 + CLEAN AIR
SO2 + CLEAN AIR
SO2 + CLEAN AIR
SO2 + CLEAN AIR
S02 + CLEAN AIR
0.4M H2SO4 + CLEAN AIR
0.4M H2SO4 -I- CLEAN AIR
0.4M H2S04 + CLEAN AIR
0.4M H2SO4 + CLEAN AIR
0.4M (NH4)2S04 + CLEAN AIR
0.4M (NH4)2SO4 + CLEAN AIR
S02 + CLEAN AIR
SO2 + CLEAN AIR
CLEAN AIR ONLY
RURAL AIR ONLY
SO2 + RURAL AIR
SO2 + RURAL AIR (DARK rx)
AUTO EXHAUST (Pb): W/NUCLEI
AUTO EXHAUST (Pb). W/NUCLEI
AUTO EXHAUST (Pb): W/O NUCLEI
AUTO EXHAUST (Pb): W/O NUCLEI
AUTO EXHAUST (Pb). W/O NUCLEI
so2
(ppm)
0
0.55
0.90
1.30
0
1.30
1.10
1.00
1.10
1.10
1.00
030
0.92
1.18
036
<0.01
<0.01
< 0.01
< 0.01
<0.01
<0.01
0.95
1.04
0
-------
Chemical and aerosol data for these experiments are summarized in Tables
2 and 3. In Table 2, the run number, date, relative humidity, reactant con-
centration and times to ozone and NO. max are shown. Table 3 summarizes the
maximum particle concentration for each system, the initial and maximum
volumetric production rates of aerosol (i.e., [dv/dt], and [dv/dt] ) and the
computed initial SO photooxidation rate as determined from electrical aerosol
analyzer (EM) data.
As in the past, values of the volumetric growth of the aerosols as
measured by the EAA are used to estimate the rate of SO oxidation. The basis
for the calculation is the assumption that all of the aerosol growth is the
result of the formation of sulfuric acid and its subsequent absorption of
water from the air. The governing equation is:
d[SO ] MW
where p is the density of the sulfuric acid droplet, P is the weight fraction
of H SO in the droplet, MW is the molecular weight of S02, and MW2 is the
molecular weight of H_SO..
The data in Table 2 show that for average propylene concentrations of
about 3.1 ppmC and NO concentrations of about 0.5 ppm, the time to ozone
A
peak is about 270 minutes and the corresponding time to N0_ max is about 140
minutes .
The summary of aerosol data in Table 3 shows that the initial rate of
S0_ oxidation (Rso-Ji in tne propylene + NO + SO- system is about 0.9% hr~
£ £ j\ £ 1
and that for the cyclohexene + NO + SO- system was only 0.4% hr" . Pre-
Jv «
sumably, the rate of S02 oxidation is relatively slow initially because of
the competition for OH by reactions with hydrocarbon and NO species.
Estimates made by investigators at Battelle suggest that only 1-2% of the OH
radicals produced during this period is consumed by the reaction with S0_.
Later, when abundant aerosol is formed, the rate of S0_ oxidation also
increases. If all of the aerosol produced during this period were aqueous
22
-------
N)
CM
TABLE 2|
SUMMARY OF CHEMISTRY DATA FROM HC + NOX AND HC + NOX + SO2 IRRADIATIONS
RUN NO.
6
7
8
9
10
11
12
13
DATE
12/12/75
12/15/75
12/16/75
12/18/75
12/30/75
1/8/76
1/12/76
1/13/76
SYSTEM
PROP + NOX
PROP + NOX
PROP + NOX
PROP + NOX + S02
PROP + NOX + SO2
CYCLO + NOX + S02
CYCLO + NOX + SO2
CYCLO + PROP + NOX
+ so2
RHi
%'
25
35
45
40
45
30
30
35
HC:
ppmC
2.9
3.2
3.0
3.0
3.1
-
3.2
3.3
IMOji
ppm
.45
.45
.45
.45
.46
.44
.45
.47
N°2ij
ppm
.035
.09
.06
.07
.12
.03
.04
.04
[O3l maxj
ppm
.620
.640
.530
.420
.495
.385
.400
.535
TIME TO
[03] max
mm
345
260
270
240
235
310
320
330
TIME TO
[NO2] max
mm
210
120
140
140
140
180
190
180
-------
TABLE 3
SUMMARY OF AEROSOL DATA FROM HC + IMOX & HC + NOX
+ SO
RUN NO.
6
7
8
9
10
11
12
13
DATE
12/12/75
12/15/75
12/16/75
12/18/75
12/30/75
1/8/76
1/12/76
1/13/76
SYSTEM
PROP + NOX
PROP + NOX
PROP + NOX
PROP + NOX + SO2
PROP + NOX + SO2
CYCLO + NOX + SO2
CYCLO + NOX + S02
CYCLO + NOX + SO2
Nmax
cm'3
1.3 xlO5
9.0 x 104
1.1 x 105
>107
9.0 x 106
7.0 x 105
2.0 x 105
3.5 x 105
dv
dt(i)
/inAr^enT3
0.1
0.1
0.1
16.5
16.1
9.2
8.4
8.2
^vj"
dt (max)
Mm3hr"1cm'3
10.6
10.2
2.7
206.1
201.8
680.0
720.0
570.0
Ks%,
EAA
%hr'1
-
-
-
0.8
0.9
0.4
0.4
0.4
* INITIAL AEROSOL VOLUME FORMATION RATE
»* MAXIMUM AEROSOL VOLUME FORMATION RATE
'"INITIAL RATE OF SULFUR DIOXIDE OXIDATION CALCULATED FROM EAA DATA,
USING EQUATION 1
24
-------
sulfate, (note [dv/dt] values in Table 3) equivalent S0_ photooxidation
rates of 30-50% hr~ would be derived. Since no such loss of SO- was observed,
it appears that most of the aerosol produced is nonsulfate in nature; however,
the exact composition of the aerosol must still be determined. Consequently,
Rso2 values calculated from EAA data using Equation 1 represent upper, bounds
for the actual RSO? due to aerosol growth from components other than sulfuric
acid. It is instructive to note that the initial aerosol volumetric produc-
tion rate dv/dt is approximately two orders of magnitude lower in the HC + NO
A
systems not containnng SO-.
AUTO EXHAUST EXPERIMENTS
A number of auto exhaust irradiation experiments were performed in an
effort to study and compare the rates of S0~ photooxidation in the presence or
absence of combustion nuclei. Both leaded and unleaded fuels were used in
this series of tests. The experiments consisted of operating a 1972 Chevrolet
with 350 CID engine at a cruise speed of 40 mph on a chassis dynamometer while
a fixed portion of the auto exhaust was introduced into the chamber. Evapora-
tive emissions were added to the system to supplement the HC content and to
give initial conditions of 3.0 ppmC HC, 0.5 ppm NO and 0.5 ppm S09. Relative
X fc
humidity was adjusted to 50%. For experiments directed toward studies of S02
oxidation in the absence of primary combustion nuclei, the air mixture follow-
ing introduction of auto exhaust was first passed through the chamber absolute
filter system. After achieving the desired initial conditions, with or with-
out the presence of primary auto exhaust nuclei, the system was irradiated
for periods of up to 11 hours.
Aerosol data from the auto exhaust irradiation experiments are summarized
in Table 4. The table shows: the particle concentration at t ~ 0, i.e.,
time of lights on; the maximum particle concentration observed during irradia-
tion^ (max); the t = 30 minutes and maximum aerosol volumetric production
rates; the initial, maximum, and average S0_ photooxidation rates as deter-
mined from electrical aerosol analyzer (EAA) data; and the average SO- photo-
oxidation rates as determined from the aerosol sulfate analysis (XRF) data.
25
-------
TABLE 4
LOG OF AUTO EXHAUST EXPERIMENTS*
EXPERIMENT
14 AE W/NUCLEI
IB AE W/NUCLEI
16 AE W/O NUCLEI
17 AE W/O NUCLEI
18 AE: DaricRx
19 AE W/O NUCLEI
22 AE (Pb): Dark Rx
23 AE (Pb) W/NUCLEI
24 AE (Pb) W/NUCLEI
25 AE (Pb) W/O NUCLEI
26 AE (Pb) W/O NUCLEI
XI AE (Pb) W/NUCLEI"
X2 AE (Pb) W/NUCLEI
X3 AE (Pb) W/O NUCLEI**
X4 AE (Pb) W/O NUCLEI***
X5 AE (Pb) W/O NUCLEI***
DATE
1/15/76
1/16/76
1/20/76
1/30/76
2/3/76
2/4/76
3/10/76
3/11/76
3/15/76
3/16/76
3/22/76
1/25/77
3/15/77
3/17/77
10/18/77
10/19/77
N It - 0)
cm"3
3.7 x 104
4:0 x 104
<103
<103
2.0 x 104
<103
1. Ox 10S
1.1 x 105
15 x 106
<103
<103
-
2.0 xlO5
<103
<103
<103
N(max)
cm"3
5.0 x 106
6X1 x 10S
>107
>107
2.0 x 104
>107
IX) x 105
2.0 x 106
5.0 x 106
>107
>107
-
2.0 x 10*
>107
>107
>107
dt (t - 30)
jlitrhr em-
ISA
12.8
21.2
113
1.0
3.8
1.2
9.0
18.8
5.2
7.0
-
1200
370
-
-
dv
dt[max)
EAA/lmV'cm-3
30JB
33.8
37.0
25.2
2.7
26.6
2.4
58.0
73.0
56.4
464
-
1200
711
-
-
Rs°2(0
EAA*h,-\
0.4
0.4
OA
0.7
<0.1
0.2
<0.1
0.3
0.6
0.2
0.3
-
34.6
11.0
-
-
Rs°2(m.x)
^Kh,'1
03
1.1
1.1
14
<0.1
1.3
<0.1
14
2.3
2.0
2.0
-
34.6
2A3
-
-
Rs°2(a.-)
"^hr-1
0.7
0.6
0.8
1X1
<0.1
03
<0.1
03
1.3
1.0
1X>
-
11X1
9.8
-
-
«so2
***%*'
-
-
-
-
-
-
-
-
-
-
-
0.8
4J
4.4
2.2
OS
HC - 2X> ppmC (EVAPORATIVE EMISSION ADDED
**HC-25ppmC
**HC->50ppmC
SO2~ 0.5 ppm (0.3 to OXi ppm)
RH,~60W
-------
Table 5 summarizes the time events of chemical and visibility changes
during the course of the auto exhaust irradiation experiments. In this table
are shown the duration of each experiment, the initial SO. and NO concentra-
£* Jt
tions, time to N0_ maximum, time to 0_ max (not reached in any of the experi-
ments), and the initial and final b values observed after 400 minutes of
scat
irradiation.
Using the modified aerosol sampling procedure with filter analysis by
energy dispersive X-ray fluorescence (see Section 5), five additional auto
exhaust + SO- irradiation experiments were subsequently performed and are
included in Tables 4 and 5 as XI through X5. These additional experiments
were performed to determine the rate of S02 photooxidation before and after
0, (max) is reached. In each of these experiments, large HC/NO ratios were
j X
obtained by introducing the exhaust emissions of an engine at fast idle under
no load conditions. In two of the five experiments, additional HC was intro-
duced from accidentally contaminated water used in chamber humidity control.
Leaded fuel was used in all five tests.
During each test, filter samples were obtained at approximately 30-minute
intervals. The quantity of data obtained as well as its consistent nature
allow much more accurate interpretations of the experiments than the previous
data analyzed by the aforementioned wet chemical technique. Two sets of plots
have been generated from this more complete data: (1) measured concentration
of sulfate in units of yg/m versus time after the initiation of irradiation
shown in Figure 4; and (2) calculated rate of S02 oxidation in units of %/hour
versus time after the initiation of irradiation displayed in Figures 5 through
7.
After about two hours each of the experiments with the exception of X3
exhibit a decay in the rate of SO- oxidation to extremely low values, in some
cases zero. In one experiment, the complete lack of oxidation was tracked for
four hours using independent sample collections and analyses.
Referring only to experiments XI and X2, the reduced rate of SO- oxidation
appears to be correlated with the 0 (max). Experiments X4 and X5 with larger
HC/NO ratios which did not allow the appearance of 0, (max) during the
X «J
27
-------
TABLE 5
AUTO EXHAUST EXPERIMENTS- CHEMISTRY DATA
EXPERIMENT
14 AE W/NUCLEI
15 AE W/NUCLEI
16 AE W/O NUCLEI
17 AE W/O NUCLEI
18 AE:DARK Rx
19 AE W/O NUCLEI
22 AE (Pb): DARK Rx
23 AE (Pb) W/NUCLEI
24 AE (Pb) W/NUCLEI
25 AE (Pb) W/O NUCLEI
26 AE (Pb) W/O NUCLEI
X1 AE (Pb) W/NUCLEI
X2 AE (Pb) W/NUCLEI
X3 AE (Pb) W/O NUCLEI
X4 AE (Pb) W/O NUCLEI
X5 AE (Pb) W/O NUCLEI
DATE
1/15/76
1/16/76
1/26/76
1/30/76
2/3/76
2/4/76
3/10/76
3/11/76
3/15/76
3/16/76
3/22/76
1/25/77
3/15/77
3/17/77
10/18/77
10/19/77
LENGTH
min
330
420
690
510
180
586
240
420
400
420
400
400
180
150
215
250
so2
ppm
.50
.38
.50
.30
.44
.34
.78
.52
.62
.44
.40
.10
.15
.25
.30
.70
NOX
ppm
.50
.47
.50
.45
.55
.44
.69
.50
.51
.40
.53
.45
.40
.35
.55
.40
tNO2 (max)
min
330
370
420
500
...
570
...
330
350
400
350
30
30
30
...
t03 (max)
mm
>330
>420
>690
>510
>580
>420
>400
>420
>400
100
90
90
...
"scat^o
10-V1
0.34
0.411
0.24!
0.24'
0.72 i
0.24 i
0.521
0.481
0.411
0.241
0.24 1
...
0.641
0.57 i
0.291
0.291
bscatt=400l
lO^m-1!
1.191
1.19
0.82
0.72'
0.64!
0.72i
0.481
2.60
3.57;
1.43!
2.39 I
...
>|40i
25.58 i
35.42 1
4.61!
K)
00
-------
3B0
I2B
IBB 2H0
TIME CHIN}
3BB
3BB
H20
Figure 4. Sulfate Concentrations Versus Time for the Indicated Irradiated
Auto Exhaust + SO Experiments.
experiment, reveal the same decay in the rate of S02 oxidation after about the
first two hours. Due to the redirection of the program towards in-cloud con-
ditions, no effort was made to determine the behavior of the rate of S02 oxi-
dation in systems containing a low HC/NOx ratio.
The plots of Rso2 shown in Figures 5, 6, and 7 indicate that there is no
discernible difference in terms of the rate of S02 photooxidation between the
systems containing the primary aerosol of the auto exhaust and the systems
from which this primary aerosol has been removed. These findings corroborate
the results of the previous set of experiments based only on EAA data and are
exhibited in Table 4. In addition, the resolution resulting from obtaining
filter samples so frequently allows us to observe that an apparent enhanced
29
-------
H '
§
XI HE H/NIKUEI
§
I
i
EB
X2 RE H/NUCLE1
IBB
2MB
TIME CHIN)
Figure 5. Rate of S02 Oxidation Versus Time for Indicated Irradiated Auto
Exhaust + SC>2 Experiments
30
-------
I.
3
i
i "
X3 RE H/D NUCLEI
i
i
g
H
i
XH RE U/Q NUCLEI
IZB
TIME CHIN)
IBB
2MB
Figure 6. Rate of S02 Oxidation Versus Time for the Indicated Irradiated
Auto Exhaust + S02 Experiments
31
-------
a
u, a t
XS HE W/D NUCLEI
L
EB
I2B
TIME CHIN)
IBB
2MB
Figure 7. Rate of SC>2 Oxidation Versus Time For The Indicated Irradiated Auto
Exhaust + SO- Experiments.
S0_ oxidation rate occurs only during about the first two hours of the experi-
ments. Following this period of high S0_ oxidation, the rate as well as
volumetric aerosol growth falls to very low values. It is not known if these
results are also applicable to the earlier experiments using a lower HC/NO
A
ratio. It should be recognized that five tests are not sufficient to make
unqualified judgements as to the mechanism or even the qualitative form of the
rate of SO- oxidation in the auto exhaust + S02 irradiated system. As a
result, these findings should be considered tentative.
From the data summaries in the tables, as well as the tentative results
shown in Figures 5, 6 and 7, the following points can be made concerning
irradiated HC + NO + SO, systems.
32
-------
1. Initial particle number concentrations were an order of magnitude
higher for the auto exhaust tests in which leaded fuel was used.
After lights were turned on, the maximum particle concentration
and the rates of production of additional aerosol were also higher
by a factor of about 2 for the leaded fuel cases. Maximum produc-
tion of new particles was greater in experiments in which primary
particulates were removed, confirming that significant amounts of
volumetric growth occurs on existing primary aerosols.
2. For the first series of auto exhaust experiments, the average photo-
oxidation rate of S09 as determined from measured volumetric aerosol
^ _i
growth was about the same (M.% hr ) with or without nuclei. In
these tests, the HC/NO ratio was about 6:1 and the reactivity of
A
the HC was not sufficiently high to allow complete oxidation of NO
in less than about six hours. CBy contrast NO dissappearance
normally occurred in slightly more than two hours when propylene or
or cyclohexene were used under equivalent conditions). In similar
auto exhaust tests containing HC/NO ratios greater than 50:1, an
* 1
apparent enhanced rate of photooxidation of SO (^5% hr j
was observed during the first two hours of the experiments from XRF
analysis of filter samples. Following this period of apparent
enhanced S02 oxidation there was little if any further oxidation of
S02 or volumetric aerosol growth. This behavior appears to be
independent of the presence of the primary aerosol or the occurrence
of 0- Onax). It is not known if the systems containing HC/NO
O Jv
ratios of 6:1 display the same behavior of SO. photooxidation.
3. In available data, the photooxidation rates for SO, as determined by
XRF analysis of filter samples were less than the SO- oxidation
rates determined from EM data. This discrepancy suggests that the
aerosol composition in the experiments was of a mixed nature {i.e.,
organics, nitrates, plus sulfates), assuming the EAA data is reliable.
4. The largest b values (at t - 400 min) were observed for the
SCcLl
leaded fuel experiments.
5. No appreciable dark reaction was observed for either the leaded or
unleaded auto exhaust cases; i.e., no new aerosol was formed in the
dark prior to irradiation.
33
-------
SECTION 7
CATALYTIC S02 OXIDATION IN CLOUDS, FOGS AND HAZES
The measured rates of S0_ photooxidation of a few tenths of a percent per
hour (Gerhard and Johnstone, 1955; and Kocmond et. al., 1975) are not suffi-
cient to explain the approximate four day lifetime of S0_ in the atmosphere
(Junge, 1960). In order to produce an atmospheric lifetime of four days, the
actual rate of SO- oxidation must be almost an order of magnitude greater than
that apparently due to photooxidation. One likely mechanism responsible for
higher rates of atmospheric S02 oxidation is the catalytic oxidation of S0_
in cloud, fog, or haze droplets containing high concentrations of heavy metals
(>10"4M).
Noted investigations of the catalytic oxidation of S0_ in aqueous solu-
tions containing heavy metals have been performed by Junge and Ryan (1958) and
Johnstone and Coughanowr (1958). While the experimental techniques used by
the former group have questionable pertinence to actual conditions occurring
in the atmosphere, qualitatively, the results are of interest. They found the
rate of catalytic oxidation of SO. in an aqueous solution to be dependent upon
the SO^ concentration, pH of the solution, and the type and concentration of
heavy metal catalyst. Suspending 300 ym diameter droplets from a glass fiber,
Johnstone and Coughanowr (1958) observed rates of SO, oxidation within the
_1 ^
droplets corresponding to about 60% hr when extrapolated to liquid water con-
tents of 0.2 g/m and ambient SO- concentrations of 1 ppm. In a more recent
investigation, Barrie and Georgii (1976) essentially repeated the work of
Johnstone et. al., (1958) using refined techniques and SO. concentrations more
likely to be found in clouds. The results of the Barrie and Georgii (.1976)
investigation were extrapolated to clouds containing 0.1 g/m of liquid water,
heavy metals Mn+ and Fe at concentrations of 10 M each, and 20 ppb of
34
-------
ambient S0_. Their estimated in-cloud oxidation rate of S0_ varied from 0.08
-1
to 2.2% hr . They found the rate of SO- oxidation to be dependent upon tem-
perature, SO concentration, pH, and type as well as concentration of catalyst.
Junge and Ryan (1958) also proposed that the presence of NH_ in the solu-
tion droplets containing heavy metal ions might produce a more active system
in terms of S02 oxidation. The "catalytic" oxidation of S02 in aqueous systems
by NH3 has since been expounded by Scott and Hobbs (1967), McKay (1971), Miller
and DePena (1972), and Easter and Hobbs (1974). However, recent reviews of
the proposed mechanism (Barrie et. al., 1974; and Beilke, et. al., 1975) sug-
gest that it will not contribute significantly to S0_ oxidation in the atmo-
sphere for values of pH normally observed in cloud droplets (typically between
4 and 5). Barrie et. al.,(1974) conclude that the major function of the SO -
NH_-liquid water system is not the conversion of SO, to sulfate but rather the
conversion of sulfuric acid aerosol and ammonia gas to ammonium sulfate.
At Calspan, unique chamber facilities allow the repeatable generation of
fogs, clouds, or hazes which have characteristics similar to those of the
respective phenomena occuring in nature. As a result of the ability to accu-
rately simulate and measure the microphysical features of these natural phe-
nomena, we can determine, without the need for questionable extrapolations,
the rate of S0_ oxidation for typical cloud and haze conditions, and because
^ 3
of the large size of the chamber (590 m ), with minimum wall effects.
Calspan performed two separate series of chamber experiments to investi-
gate the role of the catalytic oxidation of S02 in the atmosphere. In the
first series of experiments, hazes were produced in the chamber and SO.
oxidation was monitored as concentrations of SO- and heavy metals (potentially
catalytic materials) were varied. The influence of hazes upon S02 oxidation
was investigated under irradiated as well as dark conditions. The other
series of experiments involved the production of clouds or fogs with natural
microphysical features. The oxidation of S0_ was observed in this system for
irradiated and dark conditions and the influences of SO- concentration and
heavy metal concentration were studied.
35
-------
HAZE AND FOG GENERATION
The production of light hazes was accomplished by humidification of the
chamber to near 100% relative humidity using a commercial nebulizer located
at the top of the chamber; after humidification, all particulates were removed
by absolute filters. The filtration process usually resulted in a decrease of
relative humidity to about 90%. Subsequently, an aqueous solution containing
given amounts of NaCl and heavy metals was nebulized by a Collison nebulizer
using particle-free nitrogen as a carrier gas. Approximately one hour of
nebulization was required to produce the desired concentrations of aerosol
within the chamber. Following characterization of the aerosol, S02 (M. ppm)
and eventually NH_ were added to the system. Filter samples were taken prior
to the addition of SO- and regularly thereafter to determine the rate of
catalytic oxidation of SO.. The heated sampling inlet at a temperature of
709C was used to eliminate the liquid water from the filter and, consequently,
reduce artifact sulfate formation resulting from catalytic oxidation of S02 on
the filter surface. Denser hazes were formed during pre-fog activities dis-
cussed below.
The fogs required for in-fog experiments were produced in the smog chamber
by expansion of near-saturated air. Initially, the chamber was humidified to
near-saturated conditions by wetting the entire inner surfaces with distilled
water using a remotely-controlled spray nozzle. With the relative humidity in
the chamber typically in excess of 95% following the washdown, the desired
nuclei were nebulized into the chamber to serve as condensation sites for fog
formation.* As an inevitable consequence of the presence of large concentrations
of hygroscopic nuclei and high pre-fog humidities, dense hazes (b «. of about
_4 _i scat
4 x 10 m ) were formed during the pre-fog experiments. The measurements of
catalytic oxidation of SO- in dense haze were made under these coaditloas.
*An absolute filtration system as well as an activated charcoal filter system
are available to remove ambient aerosols or gases from the chamber prior to
the preparation of the chamber for fog formation.
36
-------
Subsequently, the chamber was pressurized to about 2.9 kilopascals (30 cm
of water) relative to atmospheric pressure. The pressurized state was main-
tained for about 15 minutes to restore equilibrium temperature conditions and
permit further evaporation from the wet chamber walls. Immediately after this
waiting, pressure was decreased within the chamber by withdrawing air at a
metered rate. Initially, the pressure was dropped abruptly to produce a state
of supersaturation which activated the fog nuclei and secured the desired
growth of the fog droplets. A continuous pressure reduction was required to
maintain the fog by negating the effects of heating from the warm chamber walls.
While fogs have been sustained for periods longer than one hour in the
chamber, their microphysical characteristics (i.e., liquid water content, drop
sizes, and visibility) change appreciably over such long intervals. The fogs
produced for these experiments were maintained for only 25 minutes. At the end
of the 25-minute expansion period the pressure inside the chamber was generally
about 2.9 kilopascals (30 cm of water) less than atmospheric pressure. At this
point the fog was rapidly dissipated by allowing outside air to pass through
the absolute filters and into the chamber, restoring atmospheric pressure.
The resulting dilution from the increase in pressure was less than 3%. This
procedure resulted in a fog of relatively uniform microphysical characteristics
throughout its abbreviated duration. (The exception was the irradiated fogs.
Due to heating effects, irradiated fogs tended to dissipate much sooner than
fogs which were not irradiated.)
Measured haze, pre-fog haze, and fog characteristics are presented along
with SO- oxidation data in the appropriate sections.
HAZE EXPERIMENTS
The catalytic oxidation of S02 was studied in two separate groups of
light haze experiments as well as one series of dense pre-fog hazes. Each
group of experiments is discussed separately below.
Nonirradiated Haze Experiments
A matrix of experiments was performed to determine the rate of catalytic
oxidation of SO in solution droplets smaller than 2 pm diameter with a
37
-------
-4
catalyst concentration of about 10 M or less. SO,,, concentrations of about
1.0 ppm were used for these experiments. Specific catalytic materials used
included ferric chloride hexahydrate (FeCl_*6H_0), manganese dichloride tetra-
hydrate (MnCl2«4H20), and vanadyl sulfate dihydrate (VOS042H20). These
materials were introduced into the chamber by nebulizing an aqueous solution
of the catalytic material and a supporting electrolyte (NaCl) with a Collison
nebulizer. The aerosol formed by these hygroscopic materials is composed of
solution droplets smaller than 2.0 ym in diameter for relative humidities fcn
the chamber of about 90% (see Figure 9).
The complete log of the 22 light haze experiments performed during this
phase of the effort is presented in Table 6. The parameters shown for each
haze experiment in Table 6 include: (1) The experiment title; (2) The concen-
tration of S02 in parts per million; (3) Relative humidities, RH; (4) The
Aitken nucleus concentration (per cubic centimeter); (5) Maximum b in
_4 _i scat
10 m ; and (6) The rate of SO- oxidation, RsOo* *-n percent per hour as
determined by XRF analysis of filter samples acquired at about 30-minute
intervals during the experiments. Except for experiments 1 and 2, the tests
are classified in Table 6 according to the concentration of NaCl and heavy
metal catalyst contained in the aqueous solution before nebulization. Experi-
ment 1 involved only natural rural air and nuclei; experiment 2 involved only
NaCl nuclei in the absence of catalyst.
Two distinct "haze" aerosols were generated during the series of experi-
ments. Aerosol characteristics were controlled by altering the concentration
-2 -3
of the supporting electrolyte (10 and 10 M NaCl) in the aqueous solution
which was nebulized. Typical parameters of the two aerosol distributions
responsible for the two haze types are exhibited in Figure 8. The resulting
cumulative size distributions are shown in Figure 9. The two aerosol spectra
presented in Figure 9 correspond to the aerosols observed early in experiments
#8 and #16 listed in Table 6. The figure reveals the resulting increase in
mean particle size due to the increased concentration of NaCl in the nebulizer
solution. Values of b ... as determined by an integrating nephelometer were
-4 -1
found to be about 2.3 and 1.2 x 10 m for the hazes nebulized from given
-2 -3
quantities of aqueous solutions of 10 and 10 M NaCl, respectively.
38
-------
TABLE 6
HAZE EXPERIMENTS
EXPERIMENT
1 NATURAL NUCLEI
2. 10'2M NaCI
3. 10'2M NaCI. 10'7M FeCI3
4 10"2M NaCI, 10"7M FeCU
+ NH3 MO ppb)
5. 10"2M NaCI, 10 7M MnCI2
6. 10'2M NaCI, 10'7M MnCI,
+ NH3 (-10 ppb)
7. 10'2M NaCI, 10'7M VOSO4
8. 10"2M NaCI, 10"7M VOSOd
+ NH3 (-10 ppb)
9. 10'2M NaCI, 10'7M VOSO4
10 10'2M NaCI, 10'7M VOSO,
+ NH3 (-10 ppb)
11 10'2M NaCI. 10'7M FeClj
12. 10"2M NaCI. 10*7M FeCI,
+ NH3 (~10 ppb)
13. 10"2M NaCI, 10'7M MnClj
14. 10" 2M NaCI, 10'7M MnCI.,
+ NH3 (~1 ppm)
15. 10"3M NaCI, 10"7M MnClj
16. 10 3M NaCI, 10'7M MnCI,
+ NH3 (~1 ppm)
17 10"3M NaCI. 10'7M FeCI3
18. 10'3M NaCI. 10'7M FeCU
+ NH3 M ppm)
1 9- 1 0"2M NaCI. 1 0'7M MnCI2
20. 10'2M NaCI, 10"7M MnCU
+ NH3 (~1 ppm)
21. 10'2M NaCI
22. 10'2M NaCI + NH3 ( -1 ppm)
so2
(ppm)
10
0.7
0.5
0.4
0.6
0.4
0.5
0.3
0.5
0.3
0.8
0.5
0.5
0.3
0.5
07
1.0
0.8
0.6
0.5
0.6
0.3
RH
89%
90%
87%
86%
87%
85%
93%
90%
89%
92%
91%
93%
90%
90%
95%
96%
93%
93%
86%
88%
91%
91%
AITKEN NUCLEI
per cm3
2.3 x 104
5.7 x 104
4.3 x 104
3.7 x 104
5.2 x 104
3.5 x 104
6.0 x 104
3.5 x 104
5.0 x 104
3.2 x 104
5.1 x 104
3.1 x 104
5.2 x 104
3.3 x 104
2.4 x 104
2.0 x 104
2.4 x 104
2.0 x 104
6.0 x 104
4.0 x 104
5.2 x 104
3.9 x 104
MAX. b^
(lO^m'1)
2.42
2.09
1.92
2.30
1.77
277~1
2.56
2.09
2.56
2.88
2788
2.71
2.88
1.15
1.59
1.10
2.09
2.88
2.88
3~84
3.^9
Rso2
(%hr'1)
»
*
*
»
*
*
*
*
*
~0.01
*
*
~0.09
*
*
0.04
*
*
*
*
~0.10
NOT DETECTABLE
39
-------
a
i
b
3 1
J
B
1
U
>
I
s
B
i
5
X
1
NUMBER CIBtH CM+3J IB-^H NHCL
X SURFtKE tlB*3 UM+2/Ol+3> I0-"(1 MN<3._
0 VOLUME (IB UM»3/<:Mt3) RH , 9S%
(
.
^
"^f fl ^x x -x"
[ X X x^__x ^
*- » * -
"" B^. IB~'M MNCL-
"~-~-^H RH » 87%
' B B B B-
\.
~*^rj x
XVs^.t X -*---. x x x x-
^s**^
NUMBER
-------
I SKI
N
H
(D
>
H
0
0\=
>
rt
ti
iI
3
0.BI
X 10"2M NaCl
RH = 93%
10"3M Na(
RH = 95%
B 10"3M NaCl
B.IB
Particle Diameter (microns)
I.BB
Figure 9. Cumulative Size Distributions for Hazes Nebulized
from Aqueous Solutions Containing 10 M and 10~ M
NaCl for Specified Relative Humidities .
The concentration of catalyst in the haze droplets is dependent upon the
initial concentrations of the catalyst and the supporting electrolytes in the
nebulized droplets and upon ambient relative humidity. With the procedure
used, droplets nebulized from the initial solution evaporate until the
increasing concentration of the supporting electrolyte CNaCl) decreases the
equilibrium vapor pressure of the droplet to the same value as that of its
41
-------
immediate surroundings.* The equilibrium concentration of the supporting
electrolyte is readily calculated. (See for example Low (1969) for a discus-
sion of the calculation of the vapor pressure above a solution droplet).
Results of such calculations show that, for initial concentrations of 10~ M
NaCl and 10~ M catalyst, the expected equilibrium concentration of the catalyst
-5 -4
in the aerosol is between 10 and 10 M at 90% relative humidity. The alter-
native mixture used, containing initial concentrations of 10~ M NaCl and 10~ M
catalyst, produces equilibrium concentrations of catalysts of approximately
-4 -4
10 M within the solution droplets. Concentrations of heavy metal ions <10 M
were target values since they appear to be typical of concentrations occurring
within clouds formed in parcels of urban air.**
In an effort to obtain qualitative, in situ information of the catalytic
oxidation of SO., one of Calspan's diffusion chambers was used to monitor the
cloud condensation nuclei (CCN)*** concentration. The reason for employing the
measurement is the dependence of the critical supersaturation of a particle
(supersaturation at which a particle attains free growth) on the dry size of
the particle. As a result of catalytic oxidation of S0_ within the solution
droplets, the dry sizes of the particles are expected to increase with time.
Therefore, an increase in the number of particles activated at a particular
supersaturation is expected; this increase is attributable to those particles
whose dry size was increased beyond the critical size by the addition of sul-
*For radii <0.1 ym, the increase in vapor pressure due to the decreased radius
of curvature of the droplet may become significant. This effect is described
by the Kelvin equation.
**Typical concentrations of Fe and Mn appear to be near 1 and 0.1 yg/m3,
respectively, in urban plumes (Georgii et al, 1974 and Flocchini et al, 1976).
When these values are associated with LWC of about 0.1 g/m3 as expected in
clouds, the result is concentrations of about 10 M for Fe and 10~5M for Mn.
***CCN are those particles whose unbounded growth due to the condensation of water
vapor is initiated by low supersaturations (typically less than about 1%). As
a result, CCN are the particles which are responsible for cloud and fog formation
in the atmosphere.
42
-------
furic acid. By setting the diffusion chamber at a fixed super-saturation and
monitoring the number of nuclei activated as the oxidation process proceeds,
the increase in the number concentration of CCN of a given critical size is
obtained.
The CCN concentration measurements obtained from the diffusion chamber as
functions of time for two typical experiments are shown in Figure 10. These
data exhibit no consistent increase in the number of nuclei activated for the
preset supersaturation level. The observed variation in concentration is
within the approximate uncertainty range of the instrument. Calculations
suggest that a rate of SO- oxidation on the order of 1%/hour would be necessary
to produce clearly discernible results for the diffusion chamber technique. No
such results were obtained.
Gases other than SO were present at background levels only, and typical
measured values were NO <0.1 ppm, 0_ < 1 ppb, and THC ^5 ppm. Ammonia was
X «j
added during some of the experiments to assure that the acidity of the haze
droplets was not inhibiting droplet absorption of S0_. Since we presently
have no convenient method for analyzing the amount of ammonia in the system,
specific quantities of NH_ were added volumetrically. During the first por-
tion of the experiments, the quantity of NH_ used was very small ('v/lO ppb).
In the later stages, sufficient NH, was added to achieve a 1 ppm concentration
in the chamber. To obtain these concentrations, 6 and 600 ml, respectively,
3
of gaseous NH, were introduced into the 590 m chamber.
XRF analysis of filter samples acquired at regular intervals throughout
these experiments reveal only minimal rates of S02 oxidation (see Table 6).
Many of the samples contain no detectable levels of elemental sulfur, despite
the fact that the samples were taken over a period of 2-1/2 hours. Sampling
procedures used in this phase of the experiments, coupled with the detection
limit of the XRF device, require quantities of about 0.4 yg/m of sulfate for
detection. For a 2-1/2 hour period, the collection of 1.2 yg/m of sulfate
(three times the detection limit) would represent a rate of about 0.02% hr~
for an S0_ concentration of 0.5 ppm. The results of observed Rso2 listed in
Table 6 indicate that maximum sulfate concentrations of about 1 yg/m were
only occasionally measured by the end of the experiments. Consequently, the
43
-------
I
I
a si
SB 121 IS
I in
BB
N
S
i
121 151 IBB
Tine
Figure 10. Typical Diffusion Chamber Data Showing CCN Concentration for 0.3%
Supersaturation from Aerosols Produced by the Nebulization of an
Aqueous Solution Containing 10-% and 10'^M NaCl
44
-------
results of these light haze experiments reveal consistent rates of SO- oxida-
_1 *
tion of <0.1% hr , and in many tests there was no measurable sulfate forma-
tion, suggesting rates <0.01% hr~ .
The dense haze measurements performed in conjunction with in-fog experi-
ments #114, 115 and 116 present a somewhat different story. Data are
presented in Table 7, which includes the following information: (1) The
-4 -1
number of the pre-fog haze; (2) b in 10 m ; (3) Median S0_ concentration
SC3.L. £.
in the haze in ppm; (4) The sulfate concentrations as measured prior to the
addition of S0_ and during the interval 15 to 30 minutes after SO, addition;
^ 3
C5) the concentrations within the chamber of Mn and Fe in ng/m as determined
by subsequent measurements of fog water samples; and, (6) The percent of
available S02 apparently oxidized to produce the observed sulfate concentra-
tions.
The concentrations of heavy metals contained in the solution droplets of
these pre-fog hazes are conservatively estimated at values greater than 10 M
and correspond to concentrations per unit volume of air (see Table 7) which
are typical of the concentrations encountered in urban plumes CGeorgii, et. al.,
1974; and Flocchini, et. al., 1976). In contrast, concentrations of heavy
metals in the previous haze experiments were less than 10~ M.
TABLE 7
SULFATE FORMATION IN PRE-FOG HAZES
INDICATING DEPENDENCE ON HEAVY METALS
EXPT.
NO.
114/1
115/1
116/1
b *
scat
do-V1)
12.45
19.19
24.24
SO2
(ppm)
1.0
0.5
0.75
SO. CONC
BACKGROUND
(pg/m3)
3.09
4.63
3.47
AFTER S02
ADDITION
(fig/m3)
11.19
31.64
81.02
Mn
to :% i'c-i- hour c.m occur in dense haze conditions. Because of the sparcity
of available data, care must be taken in interpretation. The apparent depen-
dence of Rso2 on metal content of the haze is sufficiently large to be con-
sidered demonstrated. The functional dependence is hidden, however, by
45
-------
changes in other parameters from experiment to experiment. The most important
of these is probably the addition of the second metal catalyst in experiment
#116. From data presented in the discussion of in-fog S02 oxidation in the
next section, it is suspected that the change in liquid water content by a
factor of two in the experiments may also be important.
Because of the known dependence of dissolution of SO- on pH of the solute
it is suspected that the observed sulfate may have been produced very quickly
after introduction of SO- and that Rso2 was substantially diminished there-
after. If this apparent saturation of sulfate is correct, the most important
data presented may be the total concentration of sulfate measured rather than
the rate of production. Meaningful rate data, for wet plume model applications
for example, require better time resolution in the experiments. The available
data suggest that further experiments are warranted.
Irradiated Haze Experiments
A group of 12 haze experiments were performed to investigate the influence
of hazes containing only trace quantities of heavy metals on the photooxidation
of S0_. These experiments are listed in Table 8. The parameters presented for
each experiment include: (1) The experiment title; (2) An identifying number;
(3) The concentrations in parts per million of sulfur dioxide (SO-)*, total
hydrocarbons (THC)**, methane CCHA3» total oxides of nitrogen CNO )***, nitric
*r X
acid CNO)***, and ozone (P3); (4) The median temperature in degrees Celsius;
(5) The median relative humidity; (6) The number of Aitken nuclei per cubic
centimeter at the beginning and end of each experiment; (7) The observed mean
oxidation rate of S0_ in percent per hour; and, (8) maximum b in 10" m .
£* S C 3. c
*Mean values
**Units are in parts per million carbon-ppmC
***Due to a malfunction of the NOx~N02-NO analyzer, the lower recorded values
of N0x and NO are qualitative. To obtain the higher concentrations, NO was
metered into the chamber, thus eliminating the reliance on the faulty gas
analyzer.
46
-------
Typically, each of the irradiated haze experiments exhibits a mean
of a few tenths of a percent per hour which is in good agreement with the rate
of S02 photooxidation in irradiated clean air systems observed by others
(Kocmond and Yang, 1975; Kasahara and Takahashi, 1976). As noted earlier,
this agreement between the rates of S02 oxidation observed in the irradiated
haze experiments and the rates of SO- oxidation for irradiated clean air
systems suggests that the total sulfate formed during the irradiated haze
experiments was due to the same photooxidation mechanism responsible for sul-
fate formation in irradiated clean air tests. These results can be applied
only to hazes containing trace concentrations of heavy metals. Other experi-
mental data concerning fogs suggest similar conclusions.
TABLE 8
IRRADIATED HAZE EXPERIMENTS
EXPERIMENT
RURAL AIR (W/O NATURAL NUCLEI) + SO2
RURAL AIR (W/O NATURAL NUCLEI) +
SEA SALT NUCLEI + SO2 + NH3 ( ~1 ppm)
RURAL AIR
RURAL AIR + SO2
RURAL AIR + SEA SALT NUCLEI
RURAL AIR + SEA SALT NUCLEI + SO2
RURAL AIR
RURAL AIR + SO2
RURAL AIR + SO2 + NH3 (~1 ppm)
RURAL AIR (W/O NATURAL NUCLEI)
RURAL AIR (W/O NATURAL NUCLEI) +
PHOTOCHEMICAL AEROSOL + SO2
RURAL AIR (W/O NATURAL NUCLEI) +
PHOTOCHEMICAL AEROSOL + SO2 + NO
EXP.
#
101/3
101/4
102/1
102/2
103/1
103/2
105/1
105/2
105/3
111/1
111/2
111/3
GAS CONCENTRATION - ppm
so2
0.45
0.30
0X11
1.00
0.01
1.00
0.01
0.50
0.20
0.04
1.00
5.00
THC*
2.6
2.6
2.8
23
2.2
22
2.6
2.6
2.6
5.8
5.8
5.8
CH4
1.9
1.9
2A
1.8
1.9
13
2.0
2.0
2.0
3.9
3.9
3.9
N0x
^<0.1
^<0.1
<0.1
<^0.1
-------
TABLE 8 (continued)
EXP. 1?
101/3
101/4
102/1
102/2
103/1
103/2
105/1
105/2
105/3
111/1
111/2
111/3
T
I°C)
22
24
21
24
23
26
24
26
26
22
24
26
RH
93%
85%
93%
94%
95%
91%
94%
91%
90%
95%
91%
86%
AITKEN NUCLEI
lC/cm3
BEFORE/AFTER
2.2x104/9.5x104
9.5x104/9.0x104
1.4x104/1.2x104
2.0x105/9.5x104
5.0x104/3.0x104
7.5x105/1.0x105
~ 700/8 .OxIO5
8.0x105/1.5x105
1.3x105/9.0x104
~700/4.9x104
4.6x104/1.8x105
1.3x105/9Jx104
CCN"
#/em3
BEFORE/AFTER
1505/1700
1700/1747
516/576
576/624
960/1344
1344/1404
20/40
40/1596
1596/>3000
64/140
140/164
164/2840
"SO,
%hr"1
02
1.3
»
02
*
02
*
02
02
0.3
0.1
0.1
MAX.
w
wV1)
32
3.2
2.4
4.1
192
12.1
<0.5
13
3.9
1.5
23
12.1
* NOT DETECTABLE
"SUPERSATURATION = 0/45%
48
-------
The one exception to the noted rates of SO- oxidation of a few tenths of
a percent for the irradiated haze experiments are the results listed in exper-
iment #101/4 in which a rate of oxidation greater than 1% hr~ was observed.
The results of this experiment, which is merely a continuation of experiment
#101/3 with the exception that about 1 ppm of NH, has been added to the
chamber, have not been substantiated with further experiments.
It is of interest to compare the observed SO- oxidation in the dense
* _i
hazes to the photooxidation of S02 in clean air (/v0.2% hr ). The sulfate
formed in the #114/1 haze is about the same amount as would be produced in an
irradiated clean air system during a one-hour period. An irradiated clean
air system would require about 10 hours to produce the levels of sulfate
observed in hazes #115/1 and 116/1. Thus, it appears that concentrations of
heavy metals such as were used in these experiments and which are commonly
found in urban plumes can potentially produce significant levels of S02
oxidation in dense haze situations.
Discussion of Results
Barrie and Georgii (1976) found the rate of SO- absorption in an aqueous
_c ^
system containing 10 M MnCl2 to be given by
R = 1.11 x 10'7 [S02] °'727 (2)
where [SO-] is the gaseous concentration of SO- in ppb. The units of the
rate are M/min. Based on one of our typical light haze experiments with an
SO- concentration of 1 ppm and a volume of solution droplets of 8 x 10~
3 3
ml/m , their expression results in 0.015 ug/m of sulfate being produced
in a 120 minute period (assuming that all of the absorbed SO- was oxidized
A 1
to sulfate). The resultant S02 oxidation rate of about 10~ % hr is two
orders of magnitude lower than the rates we are capable of resolving with
our aerosol sampling-XRF analysis procedures. Consequently, unless SO-
oxidation rates at least two orders of magnitude greater than those observed
by Barrie et. al. were encountered in the nonirradiated hazes of this series
49
-------
of experiments, we should not have expected to observe sulfate formation. To
the point that we could detect little if any sulfate in these hazes, our
results are consistent with those of Barrie et. al.
Cautiously, we can conjecture that the catalytic oxidation of SO- in non-
i *
irradiated light hazes such as these will be much less than 0.1% hr and
perhaps lower than 0.01% hr and consequently will not significantly con-
tribute to the total atmospheric oxidation of S0_. However, it is possible
that higher concentrations of these heavy metals (i.e., V, Mn, and Fe) or
other catalytic materials of higher catalytic activity may produce substan-
tially greater oxidation rates.
The following conclusions appear warranted:
1. The rate of catalytic oxidation of SO- in persistent hazes (droplet
f
diameters < 2 ym) in the presence of Ions of Mn or Fe at < 10~ M is
not significant (i.e., < 0.01% hr~ ) unless haze density is such as
to reduce visibility to substantially less than 20 km.
2. No consistent discernible effect on SO oxidation was observed by
the addition of approximately 1 ppm NH to light haze situations.
This implies that the catalytic oxidation of S02 within the solution
droplets was not quenched by the lack of dissolved S02 in the solu-
tion droplets.
3. Filter samples from dense pre-fog hazes (b on the order of
y 1
10" m~ ) revealed considerable sulfate formation. Within a 30-
minute period, sulfate levels as high as 80 yg/m were observed,
corresponding to the oxidation of about 3% of the available SO- or
_i ^
an average Rso2 of over 5% hr . The sulfate formation appeared
to be dependent upon the concentration of heavy metals.
4. Measured rates of SO- oxidation for irradiated hazes containing only
trace concentrations of heavy metals agree well with rates of S0_
oxidation observed for irradiated clean air systems. This agreement
suggests that the observed sulfate formation in these irradiated
haze experiments is due to the same photooxidation mechanism
responsible for sulfate formation in irradiated clean air tests. No
additional contribution due to the presence of the haze is discernible.
50
-------
IN-FOG* EXPERIMENTS
A series of experiments aimed at assessing the oxidation of SO- in fogs
with known microphysical characteristics, known concentrations of S0_, and
known concentrations of heavy metals was performed. A log of these experiments
is shown in Tables 9 and 10. Some 28 separate experiments are displayed in
this log along with significant experimental parameters. The two distinct
groups within this matrix of experiments are (1) In-fog without irradiation
CTable 9); and (2) In-fog with irradiation (Table 10). This grouping of
experiments was chosen to elicit possible synergistic effects on the oxidation
rate of S0_ as a result of the simultaneous influence of the fog and irradia-
tion.
The range of SO. concentrations used in this series of experiments was
chosen to encompass the spectrum of conditions expected in such drastically
different situations as clean remote locations (^10 ppb) and heavily polluted
industrial plumes (>10 ppm). Specific interest was expressed by the EPA in
the refinery to be built at Eastport, Maine, and the interaction of its
stack emissions with the frequent coastal fogs experienced in that area.
Based on expected emissions of 153.7 g/sec of S0_ and 217 g/sec of NO **, the
& J^
atmospheric diffusion of the plume under stable conditions and a wind speed
of 1 m/sec was modeled. The resultant centerline concentrations in the
plume 1 km downwind of the stack were selected as upper limit values for the
in-fog experiments. The particular values for SO- and NO were about 10 ppm
and 20 ppm, respectively.
*To avoid repetition, the term fog is applied to our chamber clouds exhibiting
characteristics similar to natural fogs and clouds, the only difference between
fogs and clouds being their proximity to the ground.
**These values of gaseous emissions are estimates furnished by Dr. Richard
Keppler, Director or ORD of the EPA in Region 1, and Mr. Warren Peters also
of EPA, Region 1.
51
-------
TABLE 9
IN-FOG EXPERIMENTS
EXPERIMENT
RURAL AIR + SEA SALT NUCLEI
RURAL AIR + SEA SALT NUCLEI
+ SO2
RURAL AIR (W/O NATURAL
NUCLEI) + SEA SALT NUCLEI
RURAL AIR (W/O NATURAL
NUCLEI) + SEA SALT NUCLEI
+ so2
RURAL AIR + SEA SALT
NUCLEI + SO2
RURAL AIR + SEA SALT
NUCLEI* SO 2
RURAL AIR + PHOTOCHEMICAL
AEROSOL + SO2
RURAL AIR (W/O NATURAL
NUCLEI) + SEA SALT
NUCLEI + SO2
RURAL AIR (W/O NATURAL
NUCLEI) + SEA SALT
NUCLEI* SO 2
RURAL AIR + SEA SALT
NUCLEI + 10'2 M MnCI2 + SO2
RURAL AIR + SEA SALT
NUCLEI + 10-2 M MnCI2 + SOj
RURAL AIR + SEA SALT
NUCLEI + 10-1 M MnCI2 + SO2
RURAL AIR + SEA SALT
NUCLEI + 10-1 M MnCI2 + SO2
RURAL AIR + SEA SALT
NUCLEI + 10-1 MMnCI.
+SO2 + SOOT *
RURAL AIR + SEA SALT
NUCLEI + 10-1 M FeCI- + 10'1
M MnCI2 + SO2
RURAL AIR + SEA SALT
NUCLEI + 10-1 M FeCI,+ 10-1
M MnCI2 + S02 J
EXP.#
100/1
100/2
101/1
101/2
106/1
106/2
110/1
112/1
112/3
114/1
114/2
115/1
115/2
115/4
116/1
116/2
T
I°C)
19
19
19
1«T
22
22
22"
22
22"
17
18"
IT
1"8
19
18
TB
GAS CONCENTRATIONS PPM
so2
0.04
0.40
0.02
0.60
6.50
5.00
0.80
1.10
6.00
0.60
0.90
0.30
0.80
0.70
0.40
1.0
THC*
2.6
3.2
2.5
2.5
2.6
2.6
2.6
6.1
6.1
3.2
3.2
15.0
15.0
25.0
13.0
13.0
CH4
2.3
2.6
1.9
1.9
1.7
1.7
1.8
3.8
3.8
2.2
2.2
>10
>10
>10
>10
>10
N0x
<0.1
<0.1
<0.1
<0.1
<0.1
20.0
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
NO
<0.1
<0.1
<0.1
<0.1
<0.1
20.0
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
°3
0.001
0.001
<0.001
0.001
0.002
0.001
0.010
0.001
<0.001
0.001
0.001
< 0.001
<0.001
< 0.001
<0.001
<0.001
ppmC
(continued)
52
-------
TABLE 9 (continued)
EXP.#
100/1
100/2
101/T
101/2
i067i
10672
iioTi
112/1
112/3"
114/i
1 14/2
iTsZi
iH/2
115/4
116/1
AITKEN NUCLEI
#/cm3
BEFORE/AFTER
-
-
3.2x104/2.4x104
2.2x1 O4/ -
4.1x104/3.1x104
2.9x1 04/2.3x104
9.4x104/9.2x104
5.3x 104/3.6x104
2.6x104/2.3x104
4.8x104/3.1x104
3.0x1 04/2.4x104
4.0x104/3.2x104
2.6x104/2.4x104
2.2x104/1.8x104
5.4x1 04/3.7x104
116/2 || 3.1x104/2.5x104
I
CCN"
*/em3
PRE FOG
1824
1552
2136
1260
1152
1044
2712
1456
1480
2136
1632
1896
1944
2340
2272
2040
POST FOG
1608
1152
1260
984
996
780
2832 ~
1464 '-
1376
1536
1092
1644
1788
1536
1908
1904
ml OF
FOG WATER
COLLECTED
iolo
i
-------
TABLE 10 IRRADIATED IN-FOG EXPERIMENTS
EXPERIMENT
RURAL AIR + SEA SALT NUCLEI
RURAL AIR + SEA SALT NUCLEI
+soz
RURAL AIR + SEA SALT NUCLEI
+ so2
RURAL AIR + SEA SALT NUCLEI
+ SO2 + NO
RURAL AIR + SEA SALT NUCLEI
+ so2
RURAL AIR + SEA SALT NUCLEI
+ so2
RURAL AIR + SEA SALT NUCLEI
+ so2
RURAL AIR + SEA SALT NUCLEI
+ SO2 + NO
RURAL AIR (W/0 NATURAL
NUCLEI) + SEA SALT NUCLEI
+ S02
RURAL AIR (W/O NATURAL
NUCLEI) + SEA SALT NUCLEI
+ so2
RURAL AIR + SEA SALT NUCLEI
+ 10-2MMnCI2 + SO2
RURAL AIR + SEA SALT NUCLEI
10-2 M MnCI2 + SO2
EXP.jf
104/1
104/2
107/1
107/2
108/1
108/2
108/3
108/4
112/2
112/4
114/3
115/3
T
<°C)
24
26
21
22
19
20
2>
20
21
2T
19
19
GAS CONCENTRATIONS PPM
so2
0.04
0.70
5.80
4.80
5.30
3.50
2.50
4.50
0.40
4.40
1.10
1.50
THC-
2.8
2.8
2.0
2.0
4.6
4.6
4.6
4.6
6.1
6.1
3.2
15.0
CH4
2.4
2.4
1.7
1.7
3.9
3.9
3.9
3.9
3.8
3.B
2.2
>10
NOX
<0.1
<0.1
<0.1
20.0
<0.1
<0.1
<0.1
20
<0.1
<0.1
<0.1
<0.1
NO
<0.1
<0.1
<0.1
20.0
<0.1
<0.1
<0.1
20
<0.1
<0.1
<0.1
<0.1
°3
0.025
0.015
0.001
< 0.001
0.001
0.001
0.004
-------
TABLE 10 (continued)
EXP.#
104/1
104/2
107/1
107/2
108/1
108/2
108/3
108/4
112/2
112/4
114/3
115/3
AITKEN NUCLEI
f/cm3
BEFORE/AFTER
7.4x104/5.0x104
4.4x104/1.0x105
5.5x104/1.0x105
9.0x104/8.2x104
5.2x104/8.5x104
5.4x104/7.5x104
6.5x104/6.2x104
4.2x104/4.0x104
3.0x104/2.6x104
2.8x1 04/4.0x1 05
2.4x1 04/7.4x104
2.4x104/2.6x104
CCN-
#/cm3
PRE FOG
1464
1188
1088
2840
1200
1272
1656
2568
1824
1008
1188
1308
POST FOG
1128
1020
2730
4000
1248
1416
2304
2832
1760
2264
1800
1760
ml OF
FOG WATER
COLLECTED
4.5
3.3
5.8
4.5
5.5
0.8
2.3
7.5
5.4
3.1
3.0
43
MAX.
"scat
no"4!"1)
781
279
1123
1096
853
59
512
1096
622
79
V71
658
SUPERSATURATION = 0.45%
55
-------
The parameters for each in-fog experiment listed in Tables 9 and 10 are:
An identifying number; (2) The median temperature in degrees Celsius;
(3) The concentrations in parts per million of sulfur dioxide (S02)* , total
hydrocarbons (THC) **, methane (CH.) total oxides of nitrogen (NO )*** ,
nitric oxide (NO)*** , and ozone CO,): (4) The number of Aitken nuclei and
cloud condensation nuclei (CCN) per cubic centimeter before and after each
fog event; (5) The volume of liquid water in milliliters col ected during
-4 -1
the fog; and (6) The maximum b in 10 m as determined from measurements
SCcLu
of visibility with an optical transmissometer (b = 3.9I/visibility (meters)).
S CcL t
The aerosols used in these experiments as condensation nuclei for the fogs
consisted primarily of NaCl. They were generated by the nebulization of a com-
mercial mixture of sea water**** which contained even the trace elements normally
found in the sea. In specific experiments, the aerosol normally found in the
rural, western New York air were also allowed to remain during the fog.
For nine of the in-fog experiments (#114-116 series) the nebulizer solu-
tion used to generate nuclei for the experiments was doped with manganese
dichloride tetrahydrate (MnCl *4H20). In two of these nine experiments (#116
series) ferric chloride hexahydrate (FeCl,-6H2)) was also added to the nebu-
lizer solution. The concentration of these compounds in the nebulizer solution
is presented in Tables 9 and 10 for each experiment. In experiment #115/4,
about 10 grams of soot scraped from the flue of a local wood burning fireplace
was disseminated as a fine powder into the chamber. These nine experiments
were aimed at assessing the role of increased concentrations of heavy metals
upon the in-fog oxidation of S0_.
*Mean values
**Units are parts -per million carbon - ppmC
***Due to a malfunction of our NO -N0_-N0 gas analyzer, the low reported values
X ^
(i.e., <0.1) of NO and NO are only qualitative. The higher concentrations
A
were obtained when NO was actually metered into the chamber, thus eliminating
the reliance upon the faulty gas analyzer.
****The mixture is Instant Ocean and is distributed by Aquariums Systems, Inc.
33208 Lakeland Blvd., Eastlake, OH 44094.
56
-------
Fogs in Tables 9 and 10 which exhibit the same series number (i.e.,
#112/1, 112/2, and 112/3 - where 112 is the series number) are consecutive
fogs produced in the same air and aerosol sample at roughly two hour intervals.
During the interim period, two filter samples were obtained and the previously
collected fog water removed from the chamber. Typically, only the SO- concen-
tration was altered between fogs, but occasionally nitric oxide was added. In
one instance (between fogs #115/4 and 115/3) about 10 g of dry soot nuclei
were added to the chamber.
The fogs were characterized by measurements of visibility, liquid water
content (LWC), and number and size distributions of fog droplets. An example
of the time history of the microphysical characteristics on one of the non-
irradiated fogs (#100/1) is shown in Figure 11. Droplet spectra are shown
for selected times during the fog. The legend of each size distribution
includes: (1) The age of the fog, T, in minutes: (2) Mean radius of the fog
droplets, R, in microns; (3) Total number of fog droplets per cm , N; (4)
Liquid water content, W, in grams per m ; and (5) Visibility in meters, V, as
calculated from the measured drop distribution. All of these values are
based upon the measured size distribution of the fog droplets discussed below.
Common trends in microphysical properties of these fogs can be noted in
the plots in the middle and at the bottom of Figure 11. As shown by the data,
soon after initiation of the fog, visibility rapidly drops to' its lowest
value. The decrease in visibility is the result of the appearance of a large
number of small droplets. With age, visibility improves due to the selective
growth of the large water droplets (at the expense of smaller droplets) and
their eventual loss from the system due to sedimentation, thus decreasing the
total amount of light scattered. The loss of droplets by sedimentation is
also indicated by the decreasing LWC of the fog as. well as total number of
fog droplets, even though the mean size of the droplets is increasing.
The fog microphysics data were obtained by collection of fog droplets by
impaction on a glass slide coated with a thin film of gelatin. The resulting
craters in the gelatin have been found to be almost exactly twice the diameter
of the initial droplet, thus providing a measure of true drop diameter. In
57
-------
10
8
gio
T- 10
N-S.S
1*1077.
1*0.250 -=
100/1
T- IS
K-3.9
N-818.
H.0.1W -g
V-83
'\oan
10 II1III III11I IIIIII
0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 IS 20 25 0 S 10 15 20 25 0 S 10 15 20 25
RADIUS (MICRONS)
MEAN DROP RADIUS
D LWC
0.3
0.2
M |
o.o 3
200
1400
O VEASURED VIS.
D CALCULATED VIS.-|-120C
A DROP CONC.
10 15
TI^e (min.)
25
Figure 11. Microphysical Characteristics of Fog #100/1 As Functions of Time
58
-------
addition, the characteristics of the impactor are such that the collection
efficiency is acceptable even for one micron diameter droplets. As a result
of known exposure time of the slide, measurement of the size spectra includes
a determination of concentration as well as sizes of the fog droplets from
which LWC can be computed. Due to the effort required to determine the drop-
let spectra [i.e., counting and sizing of ^200 craters per spectra) of the
fogs, the droplet size distributions of only eight of the fogs are presented.
These data are summarized in Table 11 and additional details may be found in
Appendix A. As shown by Table 11, fog characteristics are quite consistent
within the two groupings and comparable to those of natural fogs. The para-
meters presented in Table 11 for each fog are the fog number or description,
the mean values of fog droplet radius (vim), droplet concentration (#/cm ),
and LWC Cg/m )» and the maximum b
scat
(lO'V1).
TABLE 11 MEAN FOG MICROPHYSICAL PROPERTIES
W/O IRRADIATION
W/IRRADIATION
tfATURATFOGS
FOG#
100/1
100/2
101/1
101/2
116/1
116/2
107/1
107/2
RADIATION FOG
(INLAND)*
ADVECTION FOG
(COASTAL)*
MEAN
RADIUS
(Urn)
3.9
3.6
4.0
3.7
3.8
3.2
3.3
2.7
5.0
"1OO
DROP
CONCENTRATION
(#/cm3)
700
750
600
650
600
500
900
800
200
40
LWC
(g/m3)
0.20
0.16
0.21
0.14
0.19
0.10
(TT9~
0.08
0.11
OT7~
MAX.
bscat
(lO'V1)
IOTP
869
1001
755
959
485
T023f
755
461
~154~
JIUSTO, 1964
59
-------
Visibility determinations are obtained from measurements of the extinc-
tion of a beam of white light traversing twice the diameter of the chamber
(M.8 m) and calculated from measured drop size distributions. Both the
measured and calculated visibilities are displayed in the characterizations
presented in Figure 11. The visibilities presented in Appendix A are cal-
culated values based upon measured size distributions.
The LWC of a fog was determined by two methods. For the collection of
samples of fog water, the Calspan Fog Water Collector is placed in the chamber
and removes the droplets from a known volume of air. The droplets are removed
by impaction upon a glass column from which they drip into a glass collection
bottle. Comparison of the volume of liquid water collected to the volume of
air sampled results in a measure of the average LWC of the fog as well as a
sample of fog water suitable for chemical analysis. Another determination of
LWC may be obtained by integration of the droplet size distributions. The
latter method was used to provide the values of LWC displayed in Figure 11 and
in Appendix A.
Determinations of SO- oxidation occurring within the fogs were accomplished
by two independent means. The aforementioned Fog Water Collector was operated
during the last 15 minutes of each fog, collecting a sample of fog water.
These fog water samples represented an average over the last 15 minutes of
the fog and generally consisted of about 10 ml of fog water. The pH of the
fog water samples was measured immediately after each fog with a microelectrode.
Subsequently, portions of the samples were sent to RTF for analysis by ion
chromatography, and portions were retained for analysis Cat Calspan) of heavy
metal content by atomic absorption. The sulfate content of the fog water as
measured by ion chromatography provided a means for the determination of S0_
oxidation.
In addition to samples of fog water, filter samples were acquired before
and after each fog using the technique described at the conclusion of Section
5. The sulfuric acid present on these samples was then fixed in an atmosphere
of NH, (for stabilization), and the samples were mailed to RTF for subsequent
analysis by XRF. Assuming that all of the elemental sulfur measured by XRF
analysis existed on the filter in the form of a sulfate, a second, independent
determination of SO- oxidation was available.
60
-------
Results
The results of various analyses of the filter and water samples collected
for each fog event are presented in Table 12. This table displays: (1) The
identifying number of the fog; (2) The measured pH of the fog water sample;
(3) The average liquid water content of the fog in g/m as determined by the
fog water collector; (4) Status of irradiation (I implies an irradiated fog;
_ _ __ _A 1
- implies a nonirradiated fog); (5) Pre-fog b in 10 m ; (6) Median SO,
SCcL t ff
concentration in ppm; (7) Nitrate and sulfate concentrations of the collected
fog water in yg/ml from analysis by ion chromatography; (8) Sulfate concentra-
tions in yg/m as determined by XRF analysis of filter samples acquired before
and after the fog; (9) The concentrations of Na and Cl in yg/m as determined
by XRF analysis of filter samples obtained before and after the fog; and (10)
The concentrations in yg/ml of Mn, Mg, Cu, and Fe as well as their total con-
centration in the collected fog water as determined by atomic absorption.
Missing data points generally indicate an insufficient quantity of sample for
a complete analysis.
Although the potential importance of the nitrate content of the collected
fog water is recognized, this variable was not emphasized during the experi-
ments and hence discussions of the data will concentrate on sulfates.
The most significant data from the in-cloud experiments stems from
analyses of the collected fog water. Measured values of the pH of the col-
lected fog water were typically about 3. Such values are in fair agreement
but generally lower than field measurements of the pH of precipitation events
in the northeastern United States (e.g., Likens, 1976).
Comparison of the sulfate content of the collected fog water from non-
irradiated experiments suggests a dependence of the sulfate concentration in
the fog water on the S0_ concentration during the experiment. The correlation
of sulfate formation with SO- concentration is most clearly demonstrated by
the data shown in Figure 12 where the observed fog water sulfate concentrations
are plotted as functions of median S02 concentrations for the nonirradiated
fog experiments. The two types of data points (solid and open symbols) repre-
sent data from experiments where additional heavy metals (Mn and Fe) were and
61
-------
TABLE 12 CHEMISTRY OF IN-FOG EXPERIMENTS
FOG*
100/1
100/2
101/1
101/2
104/1
104/2
106/1
106/2
107/1
107/2
108/1
108/2
108/3
108/4
110/1
112/1
112/2
112/3
112/4
114/1
114/2
114/3
115/1
115/2
115/3
115/4
116/1
116/2
PH
4.3
4.3
3.0
1.5'
3.5
-
2.2
2.8
2A
2.1
1.0*
1.5*
1.0*
2A
2.8
3.1
2.7
2.6
2.6
3.6
3.6
335
3.9
3A
3.15
3.9
2.6
2.65
LWC
(g/m3)
0.35
0.35
0.35
0.35
0.16
0.12
0.38
0.36
0.21
0.16
0.20
0.03
0.08
0.27
0.26
0.37
0.19
0.23
0.11
0.32
0.32
0.11
030
0.29
0.15
0.30
0.29
0.26
IRRAD.
OF
FOG
1
1
-
-
-
1
1
-
-
1
-
-
1
-
PRE-FOG
bscat
(10"V)
ISA
4.1
15.9
82
12.1
17.1
15.9
32.9
135
9.0
4.8
9.6
3.7
32.9
153
6.9
2.6
12.4
3.6
2.9
19.2
9.2
2.4
8.7
24.2
144
(ppm)
0.04
0.40
0.02
0.60
0.04
0.70
6.50
5.00
5.80
4.80
5.30
3.50
2.50
4.50
0.80
1.10
0.40
6.00
4.40
0.60
0.90
1.10
0.30
0.80
1.50
0.70
0.40
1.00
NITRATE
CONC.
(jUg/ml)
0.84
0.73
0.004
0.045
6.48
-
7.83
219.60
14.23
3.69
2.07
54.31
7.01
7.81
15.37
11.33
037
0.33
1.99
-
-
0.648
-
SI
(^g/mj)
FOG WATER
14.16
31.70
4.49
15.45
24.26
42.76
68.83
70.61
_
102.10
142.80
251.70
232.80
49.83
76.54
43.85
190.70
9.43
35.99
28.09
37.75
85.02
72.72
82.98
73.92
359.32
284.83
(LFATE CONC.
UlB/m3)
PRE-FOG
8.88
4.63
1.93
_
10.04
4.63
7.33
5.79
8.88
50.94
13.12
10.81
15.05
20.84
22.38
13.89
3.09
15.82
2.70
11.19
10.03
5.79
31.64
14.28
10.42
44.76
81.04
57.89
POST-FOG
3.86
4.63
0.39
_
4.25
6.17
4.25
6.56
62.90
62.52
10.42
15.05
20.84
27.40
29.33
2.32
3.86
2.70
20.07
8.49
6.95
8.88
18.14
10.81
13.51
9.65
65.60
50.17
POSSIBLE INSTRUMENT ERROR
(continued)
62
-------
TABLE 12 (continued)
FOGiC
1007T
100/2
101/1
101/2
104/1
104/2
106/1
106/2
107/1
107/2
108/1
108/2
108/3
108/4
110/1
112/1
112/2
112/3
112/4
114/1
114/2
114/3
115/1
115/2
115/3
115/4
116/1
116/2
SUPPORTING
(Ul
Na:PRE
4.24
1.54
5.27
2.19
5.27
1.80
2.70
1.29
3.09
0.90
3.86
1X1
=
0.64
4.50
1.54
1.29
5.66
0.77
2.70
0.64
_
2.96
0..51
Na:POST
1.16
1.03
2.19
1.16
1.41
1.93
1.54
1.03
1.24
-
1.54
1.00
-
1.67
0.90
2.06
0.51
0.64
-
_
0.64
-
ELECTROLYTE
/m3)
CI:PRE
12.47
2.96
18.00
4.12
19.03
7.20
2.96
2.31
12.22
-
14.40
1.67
-
-
19.29
4.24
2.06
19.30
2.32
0.39
9.26
0.39
_
12.86
57.89
CI:POST
2.96
1.41
5.14
1.93
6.59
3.73
2.31
0.39
-
2.06
10.39
~
-
4.89
2.19
0.64
-
2.83
0.77
0.51
1.41
0.26
1.67
0.39
-
FOG WATER
(/Ug/ml )
Mn
< 0.004
< 0.004
< 0.004
< 0.004
< 0.004
< 0.004
0.006
0.011
0.006
-
-
0.005
< 0.004
< 0.004
< 0.004
0.009
0.368
0.131
2.340
0.928
0.195
2.550
0.858
Mg
0.403
0.224
0.525
0.242
0.513
0.166
-
-
-
-
0.124
0.666
0.145
0.720
0.221
-
0.630
0.183
0.120
0.783
0.172
Cu
0.025
0.035
0.004
0.053
0.190
0.627
0.118
-
-
-
-
0.148
0.018
0.026
0.120
0.099
0.106
-
0.021
0.197
0.071
0.206
0.122
Fa
0.046
0.007
< 0.004
0.004
-
0.057
0.334
0.294
-
0.160
-
-
0.567
0.003
0.015
0.004
0.048
< 0.003
0.010
-
0.018
0.136
0.095
2.730
0.832
TOTAL
0.48
0.27
0.54
0.30
-
0.76
1.13
0.42
0.01
0.17
-
-
0.84
0.02
0.71
0.01
0.37
-
1.19
0.47
-
3.01
1.44
0.48
6.27
1.98
63
-------
10.0
0.01
1.1
2.4
5.3
D Background Heavy Metals
Additional Heavy Metals
. . I » «
100 200 300
Sulfate Content of Fog Water (yg/ml)
Figure 12 Dependence of Sulfate Content of Fog Water
On SO- Concentration and Heavy Metals
400
64
-------
were not introduced into the chamber, respectively. For those experiments
where heavy metals were added (solid symbols), each data point is annotated
with the total fog water concentration of Mn and Fe (wg/ml). In addition, the
data from two consecutive fogs of a given series are connected by lines in
Figure 12. It is clear from this presentation that at trace concentrations of
heavy metals (open symbols), sulfate formation in the fog water is correlated
with the ambient S0_ concentration.
The SO- concentrations and experimental conditions for the #114, 115 and 116
series of fogs (nonirradiated) were similar to those of the previous fogs with the
exception of the type and concentrations of the potentially catalytic heavy metals.
Apparently as a result of increasing the concentration of Mn in the #115 series
fogs as compared to the #114 series, the sulfate content of the collected fog
water increased from about 30 pg/ml (in the #114 fogs) to about 80 yg/ml (for
the #115 series of fogs). The #116 series of fogs were essentially a repeat
of the #115 series of fogs with the exception that equal quantities of FeCl-
and MnCl- were added to the nebulizer solution for production of condensation
nuclei. The resultant increase in the sulfate content of the fog water of the
#116 as compared to the #115 series of fogs points to enhanced catalytic
activity within the solution droplets. The level of this increase suggests
either that Fe is a much more active catalyst than Mn or, as other investiga-
tors have found, that a mixture of Mn and Fe is more active catalytically than
would be expected from observing the two in separate tests.
The concentrations of Mn and Fe of approximately 0.01, 1.0, and 5.0 yg/ml
found in the collected fog water (Table 12) correspond to atmospheric concen-
3
trations of about 0.003, 0.3, and 1.5 yg/m , respectively. The former value
approximates the heavy metal concentrations expected in very clean air*, while
*Typical values of Fe and Mn observed at distances of several hundred kilometers
off either the eastern or western coasts of the continental United States are
about 0.01 yg/m3 (Mack, 1978).
65
-------
typical urban plumes may contain concentrations of heavy metals comparable to
the latter values. The data, therefore, suggest that high conversion rates
of S0» may occur in the atmosphere in the presence of aqueous solution drop-
lets (i.e., clouds or fogs) for concentrations of heavy metals commonly found
in urban plumes.
The fact that the sulfate content of the fog water collected from con-
secutive nonirradiated fogs in each of the series #114, 115, and 116 is
virtually constant appears also to be a significant result. This apparent
saturation of sulfate concentration implies that further oxidation of SO- is
inhibited after a specific concentration of sulfate has been reached. As
previously cited, this level seems to be dependent upon the concentration of
S02 and the type and concentration of heavy metals (specifically Mn and Fe).
Such behavior is typical of the results found by previous investigators using
very different experimental techniques. This saturation effect appears to be
related to the dependence of the solubility of S0_ on the pH of the solution.
The data from the irradiated fogs represent quite a menagerie of SO- con-
version "rates" (see Table 12). For these experiments, an irradiated fog is
defined as a typical fog experiment in which the irradiation system was turned
on only for the 25-minute duration of the fog. Inspection of data from
individual fogs, however, reveals no systematic increases in sulfate and hence
no apparent synergistic effect from the combination of irradiation and fog.
For example, while the sulfate content of the collected fog water of irradiated
fog #104/1 is slightly higher than that of the similar non-irradiated fogs
#100/1 and 101/1, previously observed photooxidation rates for S0_ (^.2% hr~ )
in the absence of the fog can explain the difference to within the scatter of
the data. A similar statement can be made about the differences between
irradiated fog #104/2 and the similar nonirradiated fogs #100/2 and 101/2.
The filter samples for the #108 series of irradiated fogs reveal sulfate con-
centrations which are adequately described by a 0.2% photooxidation of SO.,
providing the sulfate levels are corrected by referencing to Na losses. For
the most part, the sulfate contents of the other irradiated fogs fall within
the range anticipated by merely adding the expected sulfate from catalytic
oxidation and that due to photooxidation of SO-.
66
-------
Discussion of Results
One major difference between the results of these experiments and those
of previous investigations is the time required for the saturation value of
sulfate to form. In the investigations of Junge and Ryan (1958) and Terraglio
and Manganelli (1967) periods of about two hours were required before a final
sulfate concentration was reached. Barrie and Georgii C1976) found a decreas-
ing rate of S02 absorption by 2mm droplets until some three to five hours into
the experiment when SO- absorption essentially came to a halt. Our fog water
data suggest that the final sulfate concentrations have been attained in less
time than the 25-minute period of a fog. Since the duration of the fogs was
not varied and only one water sample was obtained during each fog, there is
not sufficient information to determine precisely how much time was required
to reach the final concentration of sulfate in the fog water. The long times
required to reach the saturation of sulfate by other investigators suggest
that the observed reaction rates may have been influenced by diffusional
processes within their solutions for large solution droplets. Tlie catalytic
oxidation of S02 within the much smaller droplets of the current experiments
apparently were not so strongly influenced by diffusional processes within the
droplets.
While the oxidation of 5% of the available SOj in the one-hour duration
of the combined fog and pre-fog haze experiment is an impressive rate, it can-
not be interpreted so simply. From data obtained in subsequent portions of
each series (i.e., the second and third fogs formed in each air mass) it
appears that SO- was oxidized until the total sulfate concentration in the
solution droplets reached some equilibrium value and there oxidation ceased.
This leads us to suspect that the observed total amount of sulfate formed was
the upper limit of that which could be produced by catalytic oxidation in each
specific air mass. Furthermore, the establishment of equilibrium within the
time frame of the experiment suggests that actual oxidation rates in the early
portion of exposure may have been greater than the maximum 5% observed.
67
-------
Fog series #116 shows the greatest formation of sulfate observed in our
experiments. The final concentration of about 300 yg/ml of sulfate in the fog
water represents an oxidation of about 5% of the available S02. It must be
recognized that a portion of the sulfate was definitely produced during the
dense pre-fog haze and the remainder formed during the fog. While the oxida-
tion of 5% of the available SCL in the one hour duration of the combined pre-
fog haze and fog experiment is an impressive rate, it appears to represent
an upper limit of the catalytic oxidation of S02 for this system. The data
suggest that even with additional time no further significant amounts of SO-
would be oxidized. Consequently, for studies of the catalytic oxidation of
SO. in fogs and hazes the rate of S02 oxidation may be less meaningful than
the amount of sulfate formed.
In many instances the post-fog sulfate values were unexpectedly lower
than the pre-fog or in-fog samples. These losses are probably the result of
fallout or sedimentation of fog droplets during the fog event. An estimate
of the loss of sulfate occurring during a fog may be obtained by assuming
that the loss is proportional to the loss of Na observed during the fog.
Referring to Table 12, such losses are generally greater than 50% for the
first fog of each series but not quite as large for subsequent fogs of that
series. When the sulfate values are corrected by the ratio of the post-fog
Na values, most experiments exhibit larger sulfate concentrations, as
expected, in the post-fog measurements than in the corrected pre-fog values.
In the case of many of the irradiated fogs, the fallout of fog droplets
is reduced due to the fact that heat from the lamps inhibits the growth of the
fog droplets. Referring to Table 11 or Appendix A, the mean size of the drop-
lets of irradiated fogs is seen to be smaller than that of the droplets of the
nonirradiated fogs. Since the sedimentation velocity is proportional to the
square of the radius, the larger fog droplets of the nonirradiated fogs fall
out faster than the droplets of the irradiated fogs.
Although the fallout of fog droplets during the fog complicates the
interpretations of the experiments, this is precisely the mechanism which is
responsible for the major portion of S02 scavenging and removal from the
68
-------
atmosphere. While no effort per se was devoted to investigating the fallout
during the fogs of these experiments, future experiments should study this
phenomenon.
In summary, the following conclusions concerning the catalytic oxidation
of SO- for in- fog experiments appear warranted.
1. Analysis of collected fog water reveals sulfate concentrations which
appear to be dependent upon S02 concentration and type as well as
concentrations of heavy metals Cspecif ically Mn and Fe) . Sulfate
corresponding to the oxidation of more than 5% of the available SCL
was found in given samples of fog water. This sulfate was produced
during some undetermined interval within the one-hour fog and pre-
fog haze experiments. The average RSQ? °^ >% hr~ may be misleading
since there was an apparent saturation in the amount of sulfate
formed.
2. There is no obvious synergistic effect on sulfate formation from
the simultaneous presence of irradiation and a fog containing only
trace concentrations of heavy metals. It appears that the observed
rates of photooxidation of SO. in irradiated clean air systems
_i ^
O0.2% hr ) are sufficient to describe most increases in the
content of irradiated fogs over those of nonirradiated fogs.
69
-------
REFERENCES
Anderson, R.J., R.J. Pilie, and J.L. Durham, 1977: "An Evaluation of a
Technique for the Improved Collection of Sulfate Aerosols," presented at
51st Colloid and Interface Science Symp., Grand Island, NY, June.
Barrie, L.A., S. Beilke, and H.W. Georgii, 1974: "S02 Removal by Cloud
and Fog Drops as Affected by Ammonia and Heavy Metals," Proc. Precip.
Scavenging Symp., Champaign, IL, U.S. Atomic Energy Commission.
Barrie, L.A., and H.W. Georgii, 1976: MAn Experimental Investigation of
the Absorption of Sulfur Dioxide by Water Drops Containing Heavy Metal
Ions,'' Atmos. Environ., 10, 743.
Beilke, S., D. Lamb, and J. Muller, 1975: "On the Uncatalyzed Oxidation
of Atmospheric S02 by Oxygen in Aqueous Systems," Atmos. Environ., 9,
1083.
Byers, R.L., and J.W. Davis, 1970: "Sulfur Dioxide Adsorption and Desorp-
tion on Various Filter Media," J. Air Pollut. Control Assoc., 20, 236.
Calvert, J.G. and R.D. McQuigg, 1975: "The Computer Simulation of the
Rates and Mechanisms of Photochemical Smog Formation," Int. J. Chem.
Kinetics, Symp. No. 1, 113.
Durham, J.L., W.E. Wilson and E.B. Bailey, 1976: "Continuous Measurement
of Sulfur in Submicrometric Aerosols," EPA-600/3-088.
Easter, R.C. and P.V. Hobbs, 1974: "The Formation of Sulfates and the
Enhancement of Cloud Condensation Nuclei in Clouds," J. Atmos. Sci., 31,
1586.
Fielder, R.S. and C.H. Morgan, 1960: "An Improved Titrimetric Method for
Determining Sulfur Trioxide in Flue Gas," Anal. Chem. Acta, 23, 538.
70
-------
Flocchini, R.G., T.A. Cahill, D.J. Shadoan, S.J. Lange, R.A. Eldred,
P.J. Feeney, G.W. Wolfe, D.C. Sinuneroth, and J.K. Suder, 1976: "Monitor-
ing California's Aerosols by Size and Elemental Composition," Environ.
Sci. Tech., 10, 76.
Georgii, H.W., U. Jendricke, D. Jost, and J. Midler, 1974: "The Distribution of
Heavy Metals in Clean and Polluted Atmospheres," Staub 34, 16 (English trans.)
Gerhard, E.R. and H.F. Johnstone, 1955: "Photochemical Oxidation of SO-
in Air," Ind. Eng. Chem., 47, 972.
Hidy, G.M. and C.S. Burton, 1975: "Atmospheric Aerosol Formation by
Chemical Reactions," Int. J. Chem. Kinetics. Symp. No. 1, 509.
Husar, J.D., R.B. Husar, E.S. Macias, W.E. Wilson, J.L. Durham,
W.K. Shepherd, and J.A. Anderson, 1976: "Particulate Sulfur Analysis:
Application to High Time Resolution Aircraft Sampling in Plumes,"
Atmos. Envir., 10, 591.
Jiusto, J.E., 1964: Investigation of Warm Fog Properties and Fog Modifi-
cation Concepts, NASA Contractor Report CR-72.
Johnstone, H.F. and D.R. Coughanowr, 1958: "Absorption of Sulfur Dioxide
from Air," Ind. Eng. Chem., 50, 1169.
Junge, C.E. and T.G. Ryan, 1958: "Study of the SO- Oxidation in Solution
and its Role in Atmospheric Chemistry," Quart. J. Roy, Meteorol. Soc.»
84, 46.
Junge, C.E., 1960: "Sulfur in the Atmosphere/1' J. Geophys. Res., 65, 227^
Kasahara, M. and K. Takahashi, 1976: "Experimental Studies on Aerosol
Particle Formation by Sulfur Dioxide," Atmos. Environ., 10, 475.
Kocmond, W.C., D.B. Kittelson, J.Y. Yang, and K.L. Demerjian, 1973:
"Determination of the Formation Mechanisms and Composition of Photo-
chemical Aerosols," EPA-650/3-73-002, December.
Kocmond, W.C. and J.Y. Yang, 1975: "A Smog Chamber Study of S02 Photo-
oxidation Rates and Aerosol Formation Mechanisms," Calspan Report No.
NA-5365-M-3.
71
-------
Likens, G.E., 1976: "Acid Precipitation," CSEN, Nov. 22, 29.
Low, R.D.H., 1969: "A Generalized Equation for the Solution Effect in
Droplet Growth," J. Atmos. Sci., 26, 608.
Mack, E.J. 1978: Private Communication
McKay, H.A.C., 1971: "The Atmospheric Oxidation of Sulfur Dioxide in
Water Droplets in Presence of Ammonia," Atmos. Environ., 5, 7.
Miller, J.M. and R.G. DePena, 1972: "Contribution of Scavenged Sulfur
Dioxide to the Sulfate Content of Rain Water," J. Geophys. Res., 77,
5905.
Piersen, W.R., R.H. Hammerle, and W.W. Brachaczek, 1976: "Sulfate Formed by Inter-
action of Sulfur Dioxide with Filters and Aerosol Deposits," Anal. Chem.,
48, 1808.
Radke, L.F., 1970: "Field and Laboratory Measurements with an Improved
Automatic Cloud Condensation Nucleus Counter," Preprints Conference
on Cloud Physics, Fort Collins, Colo., Amer. Meteor. Soc., 7-8.
Richards, J.R., D.L. Fox and P.C. Reist^ 1976: "The Influence of Molecular
Complexes on the Photooxidation of Sulfur Dioxide," Atmos. Environ., 10,
211.
Scott, W.D. and P.V. Hobbs, 1967: "The Formation of Sulfates in Water"
Droplets," J. Atmos. Sci., 24, 54.
Stedman, D.H. and H. Niki, 1973: "Photolysis of NO- in Air as Measurement
Method for Light Intensity, Environ. Sci. Tech., 7, 735.
Terraglio, P.P. and R.M. Manganelli, 1967: "the Absorption of Atmospheric
Sulfur Dioxide by Water Solutions," J. Air. Pollot. Control Assoc.. 17,
403.
Wood, W.P., A.W. Castlemen, Jr., and I.N. Tang, 1975: "Mechanism of
Aerosol Formation from S02." J. Aerosol. Sci., 6, 367.
72
-------
APPENDIX A
MICROPHYSICAL CHARACTERISTICS OF SELECTED LABORATORY FOGS
Included in this Appendix are collections of the characterizations of
selected fogs which were produced as a part of the program to investigate the
catalytic oxidation of SO- in fogs and clouds. Several size distributions of
fog droplets are presented for each of the fogs. A legend with each size
distribution describes: (1) The age of the fog, T, in minutes; (2) The mean
radius of the fog droplets, R, in microns; (3) The total number of fog drop-
lets per cm , N; (4) The liquid water content of the fog, W, in grams per
cubic meter; and, (5) The visibility in meters, V. All of these values are
based upon the measured size distribution of the fog droplets. Located
immediately below the legend of each size distribution is the identifying
number of the fog. For more information concerning the parameters of each
fog, refer to Table 12.
73
-------
10 ~3
. :
Z
t i ~~
CO ~
s10 2 ~=
cr :
cc
o
o
CC 10 l -q
£ :
to I
o
en
« ._ 0
10
.l°3 1
2
*
S
-*10 ~s
<=>
a: ~
cc.
o
o
cc ,n l -=,
UJ lu
Q- :
to "
0
cc
£3 0
10
R-3.3
1*1078.
U-0.23S -
100/1 -3
:
I
-
i i i i i
f:
T- S
R-2.8
1*1371.
H- 0.150 -=
V-55
S -
»100/2
-=
:
L '
i i i i i
~s
Jl,
\\
T- 10
R-3.5
N-1077.
H-0.2SO -=
V-43
=
» 100/1
;
_
t 1 1 1 1
-S
T» 10 -
J R-3.0
fl N-1128.
H- 0.187 -=
V-53
1100/2
-=
5
-
1 l l l l
~s
~
T- 15
R-3.8
j\
N-818.
H- 0.189 -=
V-63
'100/1
_
.
l i i i i
a
T- IS
R-3.7
f-t N-548.
H-0.157 -g
/
J
V-74 =
-
»100/2
-=
~
~
l l l l I
-s
"
T- 20
R-4.2
(I
N-4SB.
1*0.185 -q
V-73
tlOO/1
;
_
-
i l i i l
T" 20
R-4.1
N-478.
fl k*-0.l78 -3
I
f
1
V-70 :
100/2
-:
;
l l i l l
T- 25
R-4.5
N-316.
AH- 0.153
V-89
* 100/1
J 1
1
n
1
T- 25
R«H.H
N-298.
H- 0.110
V-87
n
1100/2
1 1 1 1 1
0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 IS 20 25 0 S 10 15 20 25
RRDIUS (MICRONS)
-------
1
IE
f i ~
tn
I10' 1
8
(Tin 1 -a
1 1 1 1U
Q- :
s10 2 1
0
fiw1 1
e :
g o
0
3
T- 5
R-4.2
H-0.153 -g
L V-80 :
L 1101/1
l
1 .- s 1
R-2.7
N-12B4.
H- 0.136 -5
V-81
.101/2 -
S s
-
:
-------
3 .
en
8
J
-
3
1
T- 5 H
R-3.3
M-1187.
M-0.222 -a
V-45 :
1107/1
Z ~
§,n* -=
S*~ 1
CE
E
0
o
a. :
to I
Q_
a 0
« |n u _
T- 10
[I, R-3.8
N-847.
W» 0.249 -3
1 "" 1
1
V.
»107/1
1 1 1 1
T- 10
R-2.9
^527. ~
H- 0.081 -3
T- 15
R-3.5
rJ] N-857.
I
l
V-138 4J
I
«107/2 ~]
1
3
\
H-0.223 -3
V-52
1107/1
I 1 1 1
T- 15
R-2.8
N-1178.
H-0.148 -
V-61 ~-
~
"
1107/2
:
T- 20
R-2.9
N-7S8.
H- 0.131 -3
I
1
1
'
i
V=80 :
1107/1
1 '
T» 20
It- 2.8
N-616.
H- 0.043
T- 25
fl R-2.9
N-838.
M- 0.111
V-B2
1107/1
V-170
1107/2
1 1 1 1
0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25
RflDlUS (MICRONS)
-------
10 a -
g 1
CC :
oc.
8 :
C3 ._ 0
10
10S 1
| =
|«21
H
0
^s
T« 10
R-3.7
J | N-819.
' L M-0.2S8
V-48 =
116/1
L ]
1 k
T- 10
R-3M ~
H- 0.1 13 -
V-95 |
j }
5 10 15 20 25 0
^|
T- 15
R-3.7 ]
J~| ^0.153 -
V«60
116/1
3 ]
-i
T- is :
R-3.2
A M-S33.
W-0.099 -
V-104 :
i
1
S 10 15 20 25 0
T- 20
R-3.9
f
H-457.
H- 0.166
V-T?
1 I116/1
L
1 1 1 1 1
rt
T« 20
R-3.1
N-554.
M-0.093
V-106
116/2
S 10 IS 20 25
RflDIUS (MICRONS)
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO
EPA-600/3-79-006
3 RECIPIENT'S ACCESSION"NO.
4 TITLE AND SUBTITLE
LABORATORY INVESTIGATION OF THE PHOTOOXIDATION AND
CATALYTIC OXIDATION OF SO-
5 REPORT DATE
January 1979
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
R.J. Anderson, R.J. Pilie, E.J. Mack, W.C. Kocmond
8. PERFORMING ORGANIZATION REPORT NO.
NA-5781-M-1
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Calspan Corporation
Advances Technology Center
4455 Genesee Street
Buffalo, NY 14225
10. PROGRAM ELEMENT NO.
1AA603 AC-13 FY77
11. CONTRACT/GRANT NO.
68-02-1785
12 SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory - RTF, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/09
15 SUPPLEMENTARY NOTES
16 ABSTRACT
The photooxidation of S02 in irradiated auto exhaust + S02 systems, the
catalytic oxidation of S02 in the solution droplets of hazes, clouds and fogs
containing several concentrations of heavy metals (Mn+^ and Fe+3)9 and the
oxidation of S02 in irradiated hazes and fogs containing only trace quantities of
heavy metals were studied in a 590 m-* indoor smog chamber. For the irradiated
auto exhaust + S02 systems, S02 oxidation rates as high as 5% hr~l were observed.
Primary particulates from auto exhaust produced no discernible effect on S02
oxidation. Catalytic oxidation of S02 in fogs and hazes was significant under
certain conditions, with as much as 6% of the available S02 being oxidized in a
30-minute period. No synergistic effect on the rate of S02 oxidation was observed
from the simultaneous presence of irradiation and a fog or haze.
17
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
*Air pollution
*Sulfur dioxide
*Aerosols
*Photochemical reactions
*Catalysis
*Sulfates
*Test chambers
13B
07B
07D
07E
14B
IS DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19 SECURITY CLASS (This Report)
UNCLASSIFIED
21 NO. OF PAGES
88
20 SECURITY CLASS (Thispage)
UNCLASSIFIED
22 PRICE
EPA Form 2220-1 (9-73)
78
------- |