EPA-600/3-77-132
December 1977
Ecological Research Series
THE ROLE OF SOLID-GAS INTERACTIONS
IN AIR POLLUTION
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
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 ENVIRONMENTAL HEALTH EFFECTS RE-
SEARCH series. This series describes projects and studies relating to the toler-
ances of man for unhealthful substances or conditions. This work is generally
assessed from a medical viewpoint, including physiological or psychological
studies. In addition to toxicology and other medical specialities, study areas in-
clude biomedical instrumentation and health research techniques utilizing ani-
mals — but always with intended application to human health measures.
This document is available to the public (iirough the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-77-132
December 1977
THE ROLE OF SOLID-GAS INTERACTIONS IN AIR POLLUTION
by
H.S. Judeikis, T.B. Stewart, A.G. Wren, and J.E. Foster
The Ivan A. Getting Laboratories
The Aerospace Corporation
El Segundo, California 90245
Grant Number R-802687
Project Officer
Jack L. Durham
Atmospheric Chemistry and Physics Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Office of Research and Develop-
ment, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the Environmental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or recommendation
for use. This project has been financed in part with Federal funds from the
Environmental Protection Agency under Grant Number R-802687.
11
-------
ABSTRACT
Sulfur dioxide and other sulfur-containing gases have been studied to
evaluate their interaction with solids likely to be found in urban aerosol
and on ground-level surfaces in the urban environment. The results of this
study indicate that sulfur dioxide readily reacts with most of these materials
by capacity-limited reactions, particularly at high relative humidities.
Removal of hydrogen sulfide and dimethylsulfide over ground-level surfaces
is a slow process and largely reversible. The implications of these results
with regard to air pollution chemistry and sulfur control strategies are
discussed. Publications, reports, and presentations that resulted from this
work are listed.
This report was submitted in fulfillment of Grant No. R-802687 by
The Aerospace Corporation under the sponsorship of the Environmental
Protection Agency. This report covers the period November 1973 to February
1977.
iii
-------
CONTENTS
Abstract iii
Figures vi
Tables vi
Acknowledgments vii
1. Introduction 1
2. Conclusions 3
3. SO2-Aerosol Interactions 4
Experimental results 4
Environmental implications 7
4. Interaction of Sulfur-Containing Gases With Ground-
Level Surfaces 9
Experimental results 9
Environmental implications 11
5. Publications and Presentations 13
Publications 13
Reports 13
Presentations 14
References 15
Appendices
A. Laboratory studies of heterogeneous reactions of
S02 17
B. Laboratory measurements of SC>2 deposition
velocities on selected building materials and soils 34
C. Deposition of H?S and dimethylsulfide on selected
soil materials 53 •
v
-------
FIGURES
Number Page
1 Tubular flow reactor 5
TABLES
Number Page
1 Heterogeneous Removal of SC>2 6
2 Deposition of Sulfur-Containing Gases Onto Ground-
Level Surfaces 10
VI
-------
ACKNOWLEDGMENTS
The authors acknowledge helpful discussions with Dr. Jack L. Durham
of the Environmental Protection Agency during the course of this work.
They also acknowledge the assistance of Herbert R. Hedgpeth in the design
and construction of the apparatus, and that of Dr. C. R. Ginnard,
Dr. Roger W. Phillips, and Lucio U. Tolentino for the conduct and analyses
of ESC A experiments.
VII
-------
SECTION 1
INTRODUCTION
The atmospheric chemistry of sulfur-containing compounds is of con-
siderable interest because of potential adverse health effects attributable to
these species as well as acidic rainfalls and haze formation (1). A process
of importance in all of these phenomena is the atmospheric transformation
of gaseous sulfur dioxide (SO2) to sulfate aerosol. Several mechanisms,
all of which may be operable, have been suggested to account for this trans-
formation (1, 2). These include gas-phase oxidation of SC>2 by direct and
indirect photolysis, oxidation in liquid droplets, and oxidation on the sur-
face of atmospheric aerosols.
The latter type of process, involving gas-solid interactions in the
atmosphere, is poorly understood and generally neglected by air pollution
modelers. Such processes are also important in the removal of pollutant
gases from the atmosphere by interaction with ground-level surfaces (1).
For these reasons, laboratory studies of SC>2 interactions with solids likely
to be found in urban aerosols, and at ground levels in the urban environment,
were carried out.
Also of interest were gas-solid interactions, primarily with ground-
level surfaces, of biogenically emitted sulfur-containing gases. Emissions
of the latter species, on a global scale, are estimated to be comparable to
anthropogenic SC>2 emissions (3-5) and consequently are important consti-
tuents in atmospheric sulfur budgets. The biogenic sulfur emissions are
believed to arise from hydrogen sulfide (H2S) and dimethylsulfide (3-5),
although the relative contributions of these two species are uncertain (4-7).
In this work, studies of interactions of both species with selected ground-
level surfaces were carried out.
The results of our studies indicate that SC>2 interactions with repre-
sentative aerosol materials can initially occur quite rapidly. In most cases,
this takes place with a near quantitative conversion to adsorbed sulfate.
With time (SC>2 exposure), the reactivities of the solids investigated gradu-
ally diminish and ultimately approach zero because of the capacity-limited
nature of these reactions. Atmospheric projections of our results with the
use of simple models for gas-solid interactions indicate that these processes
will be most important at or near emission sources, e.g., in power plant
plumes. Nonsource interactions, such as with atmospheric ammonia, may
-------
also occur, as indicated by additional results obtained in this study. Quan-
titative estimates, based on our results, of SO2 to sulfate conversion in the
atmosphere by gas-solid reactions indicate that the amount of conversion
that occurs by this process will be primarily governed by aerosol burdens
rather than SC>2 levels.
Initial deposition of SC>2 on ground-level surfaces was also found to be
rapid. The surfaces investigated included selected soils and cements
commonly found in urban environments. The cements, on average, were
found to be more effective in removing SC>2 than the soils that were exam-
ined. The latter results indicate that certain construction materials widely
used in urban areas may be helpful in removing atmospheric SC>2- Experi-
mentally, we found that SC>2 deposition over both the soils and cements
occurs by capacity-limited reactions, which indicates that these materials
would lose their ability to remove SO2 after prolonged environmental ex-
posures. However, potential regenerative processes to rejuvenate surface
activity may be operational in the environment. Laboratory experiments to
examine these possibilities indicate that such processes do indeed exist.
In the case of H2S and DMS deposition on ground-level surfaces, the
experimental results indicate that these processes are not likely to be en-
vironmentally important. This conclusion, in turn, suggests possible
long-range transport of these species in the environment, such that they
could contribute to the sulfur-containing gas burden in urban atmospheres.
However, results of work carried out in other laboratories (8-10) indicate
that gas-phase oxidation of H2S and DMS will limit their atmospheric life-
time to a few days or less.
-------
SECTION 2
CONCLUSIONS
Results of laboratory investigation of interactions of sulfur dioxide
(SO2) and other sulfur-containing gases with solids representative of urban
aerosols and environmental ground-level surfaces have indicated high initial
reaction rates that gradually decrease with time (SO2 exposure) owing to the
observed capacity-limited nature of these reactions. Relative humidity was
found to be very important in determining the capacity for, but not the rate
of, SO2 uptake. To within experimental error, the SO2 was quantitatively
converted to adsorbed sulfate over most of the solids studied. Atmospheric
projections of these results indicated that SO2 can be converted to sulfate
at a rate as high as 32 percent/hr, with the reactions likely to be most im-
portant at or near emission sources. However, nonsource interactions with
atmospheric ammonia could be important, as indicated by additional results
obtained.
Studies of SO2 deposition over selected soils and building surfaces
yielded results qualitatively similar to those described above. Thus, initial
reactivities were high but gradually diminished with SO2 exposure; SO2 re-
moval was irreversible; and relative humidity had a significant effect on
capacities for SO2 uptake. Interestingly, various cements were found to be
even more effective than soils for SO2 removal. As in the case of SO2
interactions with aerosol-like materials, we found that interaction with
ammonia can be important in reactivating saturated surfaces. Additionally,
precipitation washing away soluble surface reaction products was shown to
be another potential surface reactivation process in the environment.
Results of studies of hydrogen sulfide (H2S) and dimethylsulfide (DMS),
biogenically emitted into the atmosphere in quantities comparable to anthro-
pogenic SO2 emissions, indicated that depositions of these species onto
selected soils are not environmentally important.
-------
SECTION 3
SO2-AEROSOL INTERACTIONS
EXPERIMENTAL RESULTS
The detailed technical results of this study are given in Appendix A.
The study involved laboratory measurements of the rates of reaction of
SO2 with solids likely to be found in urban aerosols. These included pri-
marily metal oxides, selected on the basis of their abundance in urban
aerosols and their likelihood to catalyze SO2 oxidation, as well as fly ash
from five different power plants. The latter, supplied in part by the
Environmental Protection Agency (EPA), were from coal-fired plants
(Appendix A).
Experiments were carried out in the tubular flow reactor illustrated
in block diagram form in Figure 1. The reactor contained an inner, con-
centric cylinder that was coated with the solid of interest. As a gas mix-
ture containing trace amounts of SO2 passed through the reactor, the SO2
diffused to the walls of the coated cylinder, where it was removed by
heterogeneous reaction. This resulted in a decrease in SO2 concentration
as a function of distance down the tube. The SO2 concentration gradient
was measured by means of a system of small probes, whose intakes were
centered along the cylinder axis, that were connected by means of a rotary
valve to a mass spectrometer. Results from these experiments were
analyzed in terms of -values or reactivities, which are approximately the
fraction of SO2~solid collisions leading to SO2 removal. The measured
0-values were then used in conjunction with simple atmospheric models (11)
to estimate SO2 removal rates by SOo-aerosol reactions under conditions
representative of urban atmospheres.
Measured reactivities for freshly prepared solid coatings ranged
from approximately 10"^ to less than 10~" for the materials studied or
from about 1 in 1000 SO2-solid collisions being effective in removing SO2
to less than 1 in 1, 000, 000. These results are given in Table 1 together
with projected atmospheric removal rates for SO2- The latter were
calculated as described earlier (11), assuming that an atmospheric aerosol
burden of 100 jig/m^ had the same reactivity as the indicated solid, e.g., if
100 fig/m^ of urban aerosol had the same reactivity as MgO, the SO2 re-
moval rate would be 32 percent/hr. Of course, urban aerosols would be
composites of the materials given hi Table 1, and many others, and actual
-------
|-*PUMP
PRESSURE
GAUGE
FLOW-
METER
— GAS SAMPLING PROBES
x-
/ ^4-WAY VALVE
REACTION CHAMBER
r^PUMP
VALVE
FLOWMETER
S02/N2
MASS SPECTROMETER (gas analysis)
VALVE
GAS MIXER
AIR
Figure 1. Tubular flow reactor.
-------
TABLE 1. HETEROGENEOUS REMOVAL OF SO2
Material
MgO
Fe O
Mohave fly ash
A12°3
MnO?
bi
Cholla fly ash
River Bend fly ash
Shawnee fly ash (M)
Louisville fly ash
PbO
Shawnee fly ash (M)
Charcoal
Shawnee fly ash (E)
Shawnee fly ash (E)
NaCl
Louisville fly ash
River Bend fly ash
105 x 0
100
55
50a'b
40
30
30a,b
30b
iob
7b
7
5a
3
2b
0.4a
0.3
0.2
-------
removal rates would vary in proportion to the abundance of these materials
in atmospheric aerosols. The type of study described here serves to
identify the more reactive aerosol components.
Analysis of the surfaces after SO2 exposure by x-ray photoelectron
spectroscopy (ESCA) and wet chemical techniques indicated that, to within the
accuracy of the methods employed (about a factor of 2), the SC>2 was quanti-
tatively converted to adsorbed sulfate. An exception to this finding was seen
for A^Oj, and possibly charcoal, where the experimental evidence indicated
that SO*2 removal occurred by reversible physical adsorption, with little
sulfate formation.
These removal rates are quite high and indicate the potential environ-
mental importance of SC>2-aerosol reactions in SC>2 removal and, particu-
larly, sulfate formation. However, we found that the high initial reactivi-
ties invariably decreased with time (SC>2 exposure) until, ultimately, the
solids became totally unreactive toward SC>2 removal. This result indicates
that the SO? -solid reaction is a capacity-limited process. (In the case of
the River Bend fly ash, ESCA analysis of the as-received material indicated
an already high sulfate content. Washing the material with distilled water
to remove soluble sulfates yielded the significantly enhanced reactivity
given in Table 1. ) Quantitatively, the solids investigated can remove j.rom
about 0. 1 to greater than 50 percent of their weight of SC^- Relative humi-
dity was found to be very important in most cases in determining the amount
of SC>2 that could be removed, with SOo removal increasing by up to two
orders of magnitude, in some cases, with increasing humidity. The mois-
ture content of the reaction mixture did not, however, affect the SC>2 removal
rates to within experimental error.
The high initial reactivity, coupled with the limited capacity for SC>2
removal, indicates that freshly emitted aerosols will be active toward SC>2
for about 10 hr under typical urban conditions. In many instances, however
(e.g., power plant stack emissions), SC>2 levels are much higher than those
in the average urban atmosphere. Thus, at or near emission sources,
aerosols may only be active for about 1 hr or less.
It has been suggested (1, 2, and references therein) that interaction of
atmospheric ammonia with aerosols can be important in the heterogeneous
oxidation of SC^- This is believed to result from neutralization of sulfuric
acid formed, which permits further reaction to occur. In order to examine
this possibility, we exposed a Mohave fly ash sample to SO2 until it would
no longer remove this species. The sample was then sequentially exposed
to ammonia and reexposed to SC>2 • Results indicated that the reactivity of
the fly ash to SO2 removal was substantially restored (to about 50 percent
of its original value).
ENVIRONMENTAL IMPLICATIONS
The results of these studies indicate the environmental importance of
SO2 -aerosol reactions in particulate sulfate formation. The high initial
oxidation rates observed indicate that these reactions can make an important
contribution to secondary sulfate formation near emission sources. Beyond
-------
the source region, the data indicate that their importance will diminish
because of the capacity-limited nature of the reactions. However, interac-
tion with atmospheric ammonia could promote further reaction in nonsource
areas, as suggested by the experiment with Mohave fly ash rejuvenation by
exposure to ammonia.
Measured capacities for SO2 removal differed significantly from solid
to solid and ranged from about 0. 001 to 0. 5 g SC>2 removed per gram of
solid. The capacity of actual urban aerosols for SC>2 removal could have
significant implications in control strategies for secondary sulfate formed
by gas-solid interactions. For example, if capacities for SC>2 uptake were
on the order of a few tenths of a gram of SC>2 removed per gram of solid
(or less), present SC>2 and particulate levels (1) indicate that atmospheric
sulfate originating from gas-solid interactions would be determined primari-
ly by atmospheric aerosol levels. On the other hand, considerably higher
capacities for SC>2 removal indicate that sulfate formation by this process
would be controlled primarily by SC>2 levels. Our limited results support
the former possibility.
An added result of interest in these studies is the significant increase
in capacity for SO2 uptake with increasing relative humidity. These results
support suggestions (based in part on previous experimental work) that SO2
oxidation in adsorbed water films, or water droplets, may well be one of
the most important heterogeneous processes for SO2 to sulfate conversion.
-------
SECTION 4
INTERACTION OF SULFUR-CONTAINING GASES WITH GROUND-
LEVEL SURFACES
EXPERIMENTAL RESULTS
The detailed technical descriptions of these laboratory studies are
given in Appendices B and C. They were carried out in the tubular flow
reactor used in the studies described in Section 3 and Appendix A. Data
were analyzed in terms of the deposition velocity Vg, a pseudo -heteroge-
neous rate constant for removal of the species of interest at a ground-level
surface. As indicated in Appendix B, the deposition velocity is the product
of the reactivity , which was described in the preceding section, and the
gas -solid collision frequency, which can readily be calculated from simple
kinetic theory (12). The atmospheric flux of a trace species to an environ-
mental surface can be determined by multiplying the deposition velocity of
the trace species by its atmospheric concentration.
In the case of sulfur dioxide, we measured deposition velocities over
selected soils as well as construction surfaces commonly found in urban
environments. The results of these measurements are given in Table 2.
The deposition velocities of SOŁ over soils, which agree well with other
measurements (13), indicate that these materials are effective in the re-
moval of atmospheric SO2- In addition, it is seen from Table 2 that the
cements investigated were even more effective than the soils for removing
SO2 • The average deposition velocity for the soils was 0.71 cm/sec, com-
pared with the average value for the cements of 1.8 cm/sec.
The removal of SOo was irreversible and was found to occur by
capacity -limited reactions. Presumably, adsorbed sulfates were formed
by surface reactions, although wet chemical analyses were largely unsuc-
cessful because of interferences by various species present in the unex-
posed samples. Measured capacities for SOo removal from humidified
reaction mixtures were in the range of 0.4-2.8 g SO2 removed per square
meter of surface. Capacities for removal from dry mixtures were factors
of 3-10 lower, depending on the solid.
The possible rejuvenation of the reactivity of surfaces whose capacity
for SO2 uptake had been completely expended by prolonged exposure to SO2
was also examined in these studies. Potential environmental rejuvenation
mechanisms include precipitation washing away soluble surface reaction
-------
TABLE 2. DEPOSITION OF SULFUR-CONTAINING GASES
ONTO GROUND-LEVEL SURFACES
Surface
Cement (1)
Ready- mix cement
Exterior stucco (1)
Cement (2)
Adobe clay soil (1)
Exterior stucco (2)
Adobe clay soil (2)
Sandy loam soil (1)
Sandy loam soil (2)
Asphalt
so2
2.5
2.0
1.8
1.6
0.92
0.86
0.66
0.65
0.60
0.04
Deposition velocity (cm/sec)
H2S DMSb .
0.016 0.28
0.015 0.064
(1) and (2) refer to different material sources within the Los Angeles area.
Dimethyl sulfide.
°Cured.-
10
-------
products, restoring reactivity, and interaction with atmospheric ammonia.
These possibilities were examined in the laboratory by exposing selected
surfaces to SC>2 until they were no longer active in removing this species.
The exposed surfaces were then rinsed with distilled water (to simulate
rain washing away soluble surface reaction products, i.e., sulfates) or
exposed to ammonia. When these surfaces were subsequently reexposed to
SC>2, the reactivities were restored to those of the freshly prepared sur-
faces, which supported the ideas discussed above as viable environmental
rejuvenation processes.
We also measured deposition velocities for t^S and dimethylsulfide
(DMS) over selected soil samples, as indicated in Table 2. Here the depos-
ition velocities for DMS, and especially t^S, were quite low compared with
those observed for SO2- Moreover, in the case of t^S and DMS, we found
that deposition occurred by means of reversible processes (presumably
physical adsorption) and that irreversible removal processes occurred at
rates at least a factor of 5 lower than those given in Table 2 for these two
species. Our results indicate that deposition processes of t^S and DMS
onto ground -level surfaces do not appear to be environmentally important.
ENVIRONMENTAL IMPLICATIONS
The results for SOŁ deposition over selected soils and building mater-
ials indicate that these surfaces can be effective in the removal of atmos-
pheric SO2- Of particular interest are the results for SO2 removal over
various types of cements. Not only do the latter imply that such materials
can be helpful in removing SO2 from urban atmospheres, but they also indi-
cate additional strategies that could be used for passive SO2 control to com-
plement emission source control measures.
Thus, specific concrete formulations in widespread use could be
examined for SOo uptake rates to determine which are more effective in
SO2 removal. Design of exterior surfaces could be carried out in such a
manner as to maximize available surface area for SO2 removal. We sus-
pect paint would be much less effective for SO2 removal, indicating that
these surfaces should not be painted. Sandblasting of older surfaces might
also be helpful. Many of these criteria could be applied to interior surfaces
as well. Designing interior surfaces to maximize SO2 uptake would be
particularly beneficial to individuals who may be especially sensitive to
SO2 exposure. Spedding et al. (14-17) have already done much work on.
SO2 uptake by interior surfaces.
Of course, the capacity -limited nature of the SO2 uptake indicates that
additional measures would have to be considered. Experimentally, we
found that the cements studied lose their ability to remove SO2 when expo-
sures reach the order of 1 g of SO2 removed per square meter of surface in
humidified gas mixtures. For an atmospheric SO2 concentration of 50 (j.g/m
and a deposition velocity of 1. 8 cm/sec, these results indicate that satura-
tion would occur in approximately two weeks. (Actually, a somewhat longer
period would be required because reactivity decreases with exposure, as
indicated in Appendix B. ) From our results on rejuvenation of activity by
washing surfaces with water, a weekly hosing down of concrete surfaces in
11
-------
a dry area such as Los Angeles in the summer might be an effective way of
maintaining the activity of these surfaces for SO 2 removal. Natural preci-
pitation could serve the same purpose in wetter parts of the country. Of
course, care would have to be exercised in handling wash water in order to
minimize sulfate pollution in runoff waters.
Although the results for t^S and DMS deposition over selected soils
indicate that these are not likely to be environmentally important processes,
they also indicate the possibility of long-range transport of these species
in the atmosphere. However, results of recent work (8-10) have indicated
that both of these species can be readily oxidized by homogeneous reactions
in the atmosphere that would limit their lifetime to about one day.
12
-------
SECTION 5
PUBLICATIONS AND PRESENTATIONS
Publications, reports, and presentations that resulted from this work
are given here. The first item listed in each section represents work
carried out on our previous EPA grant (Grant No. 801340, Final Report No.
EPA-650/3-74-007, August 1974) but reported on during the initial time
period of this grant.
PUBLICATIONS
Stewart, T. B., and H. S. Judeikis. Measurements of Spatial Reactant
and Product Concentrations in a Flow Reactor Using Laser -Induced
Fluorescence. Rev. Sci. Instrum. 45:1542-1545, 1974.
Judeikis, H. S. , and T. B. Stewart. Laboratory Measurement of SO2
Deposition Velocities on Selected Building Materials and Soils.
Atmos. Environ. 10:769, 1976.
Judeikis, H. S. , and A. G. Wren. Deposition of ^S and Dimethylsulfide
on Selected Soil Materials. Accepted for publication, Atmos. Environ.
May 1977.
Judeikis, H. S. Heterogeneous Reactions of Gaseous Air Pollutants. To
be published, Calif. Air Environ. 1977.
Judeikis, H. S. , T. B. Stewart, and A. G. Wren. Heterogeneous Removal
of Atmospheric SO~. Submitted for publication, Atmos. Environ.
May 1977.
REPORTS
Stewart, T. B., and H. S. Judeikis. Measurements of Spatial Reactant
and Product Concentrations in a Flow Reactor Using Laser -Induced
Fluorescence. ATR-74(7441)- 1, The Aerospace Corp. , 1 April 1974.
Judeikis, H. S., and T. B. Stewart. Laboratory Measurements of SO?
Deposition Velocities. ATR-76(7498)- 1, The Aerospace Corp. ,
19 February 1976.
13
-------
Judeikis, H. S. , and A. G. Wren. Deposition of H2$ and Dimethylsulfide
on Selected Soil Materials. ATR-77(7498)-l, The Aerospace Corp.,
June 1977.
Judeikis, H. S. , T. B. Stewart, A. G. Wren, and J. E. Foster. The
Role of Solid-Gas Interactions in Air Pollution. ATR-77(7498)-2, The
Aerospace Corp., 15 July 1977.
PRESENTATIONS
Stewart, T. B., S. Siegel, H. S. Judeikis, and H. R. Hedgpeth. Reaction
of NOX on Particle Surfaces. American Chemical Society, 167th
National Meeting, Los Angeles, 31 March-5 April 1974.
Judeikis, H. S. Heterogeneous Removal of SC>2 From the Atmosphere.
8th Aerosol Technology Meeting, University of North Carolina,
Chapel Hill, North Carolina, 6-8 October 1975.
Judeikis, H. S. Heterogeneous Removal of SOŁ From the Atmosphere.
Workshop on the Chemistry of Atmospheric Sulfur, Drexel University,
Philadelphia, 12-14 October 1976.
Judeikis, H. S. Heterogeneous Reactions of Gaseous Air Pollutants, SO2,
NOX, Freon Derived Species. California Institute of Technology,
Environmental Engineering Science Seminar, 23 February 1977.
Judeikis, H. S. Heterogeneous Removal of SO2 From the Atmosphere.
American Chemical Society, 173rd National Meeting, New Orleans,
20-25 March 1977.
Judeikis, H. S. Heterogeneous Reactions of Sulfur- and Nitrogen-Contain-
ing Pollutant Gases. Particulate Pollutant Workshop, University of
California, Riverside, 21-22 April 1977.
Judeikis, H. S. Heterogeneous Reactions of Gaseous Air Pollutants.
Gordon Conference on Chemistry at Interfaces, Meriden,
New Hampshire, 18-22 July 1977.
14
-------
REFERENCES
1. U. S. Environmental Protection Agency. Position Paper on Regula-
tion of Atmospheric Sulfates. EPA-450/2-75-007, Research Tri-
angle Park, North Carolina, 1975.
2. Hidy, G. M. , and C. S. Burton. Atmospheric Aerosol Formation by
Chemical Reactions. In: Proceedings of First Symposium on
Chemical Kinetics Data for the Upper and Lower Atmosphere, Intl.
J. Chem. Kinet., Warrenton, Virginia, 1975. pp. 509-542.
3. Friend, J. P. The Global Sulfur Cycle. Chemistry of the Lower
Atmosphere. S. I. Rasool, ed., Plenum Press, New York, 1973.
pp. 177-201.
4. Lovelock, J. E., R. J. Maggs, and R. A. Rasmussen. Atmospheric
Dimethylsulfide and the Natural Sulfur Cycle. Nature 237:452-453,
1972.
5. Rasmussen, R. A. Emission of Biogenic Hydrogen Sulfide. Tellus
26:254-260, 1974.
6. Hitchcock, D. R. Dimethyl Sulfide Emissions to the Global Atmos-
phere. Chemosphere 4:137-138, 1975.
7. Hitchcock, D. R. Microbiological Contributions to the Atmospheric
Load of Particulate Sulf ate. Environmental Biogeochemistry, Vol.
I, J. O. Nriagu, ed. , Ann Arbor Science Publishers, Inc., Ann
Arbor, Michigan, 1976. pp. 351-367.
8. Westenberg, A. A., and N. deHaas. Rate of the Reaction OH + H2S
-• SH + H?O Over an Extended Temperature Range. J. Chem. Phys.
59:6685-6686, 1973.
9. Perry, R. A., R. Atkinson, and J. N. Pitts, Jr. Rate Constants for
the Reactions OH + H2S -* SH and OH + NH, -» H2O + NH? Over the
Temperature Range 2*97-427 °K. J. Chem. Phys. 64:3237-3239, 1976.
10. Cox, R. A., and F. J. Sandalls. The Photo-Oxidation of Hydrogen
Sulfide and Dimethylsulfide in Air. Atmos. Environ. 8: 1269-1281,
1974. Also, R. D. Cadle, The Photo-Oxidation of Hydrogen Sulfide
and Dimethylsulfide in Air. Atmos. Environ. 10:417, 1976.
15
-------
11. Judeikis, H. S., and S. Siegel. Par tide-Catalyzed Oxidation of
Atmospheric Pollutants. Atmos. Environ. 7:619-631, 1973.
12. Moelwyn-Hughes, E. A. Physical Chemistry. Pergamon Press, New
York, Second Revised Edition, 1961.
13. Payrissat, M. , and S. Beilke. Laboratory Measurements of the Uptake
of Sulfur Dioxide by Different European Soils. Atmos. Environ.
9:211-217, 1975.
14. Spedding, D. J. , and R. P. Rowlands. Sorption of Sulfur Dioxide by
Indoor Surfaces. I. Wallpapers. J. Appl. Chem. 20:143-146, 1970.
15. Spedding, D. J. Sorption of Sulfur Dioxide by Indoor Surfaces. II.
Wood. J. Appl. Chem. 20:226-228, 1970.
16. Spedding, D. J., R. P. Rowlands, and J. E. Taylor. Sorption of
Sulfur Dioxide by Indoor Surfaces. III. Leather. J. Appl. Chem.
Biotechnol. 21:68-70, 1971.
17. Spedding, D. J. The Sorption of Sulfur Dioxide by Indoor Surfaces.
IV. Flooring Materials. J. Appl. Chem. Biotechnol. 22:1-8, 1972.
16
-------
APPENDIX A
LABORATORY STUDIES OF
HETEROGENEOUS REACTIONS OF SO,
INTRODUCTION
Interest in gas-aerosol reactions in the atmosphere stems from a need
to understand such reactions and their impact on atmospheric chemistry, as
well as their contribution to atmospheric haze formation and health effects
attributable to aerosols in the respirable size range. Of particular current
interest are heterogeneous reactions of SO2 and the contribution of the^e
reactions to atmospheric sulfate aerosol burdens [Environmental Protection
Agency (1975)]. Potential heterogeneous processes for SO? to sulfate con-
version, involving solid and liquid aerosols, have recently Deen reviewed
[Environmental Protection Agency (1975), Hidy and Burton (1975), and Brock
and Durham (1977)] .
Reactions between gaseous SO2 and solids have been investigated for
years, since these reactions are used extensively in the industrial production
of sulfuric acid, the leading chemical commodity. The data gleaned from
these studies, however, are of limited use in predicting atmospheric con-
version rates because of the high temperatures and reactant pressures used
in industrial processes.
In the past decade, laboratory studies conducted under conditions more
nearly approximating those of the ambient atmosphere have demonstrated the
potential importance of SOŁ-solid reactions to sulfate aerosol burdens. Ex-
amples of this work include that of Okita (1967) and Urone et al. (1968), who
found that SO2 removal from gas mixtures was accelerated in the presence of
selected solids that are of atmospheric interest. Similar results were ob-
tained by Smith, Wagman, and Fish (1969)S who used a novel exploding wire
technique to generate aerosol particles, and by workers at the University of
Pittsburgh [Cheng, Frohliger, and Corn (1971); Cheng, Corn, and Frohliger
(1971); and Corn and Cheng (1972)], who used a flow reactor in which aerosol
particles were suspended on Teflon beads. Chun and Quon (1973) studied the
oxidation of SO2 on ferric oxide particles, whereas Okita (1967); Devito-
francesco, Panke, and Petronio (1972); and Burke, Baker, and Moyers (1973)
observed SO2 removal by collected atmospheric particles. Some of these
studies demonstrated the capacity-limited nature of SO2 uptake, attributed
to a lowering of the surface pH by sulfuric acid formation [e. g. , Junge and
17
-------
Ryan (1958), Van den Heuvel and Mason (1963), Scott and Hobbs (1967),
Foster (1969), and McKay (1971)].
A number of studies have also been carried out, by various experimental
techniques, to identify adsorbed sulfur compounds. Examples include mea-
surements by x-ray photoelectron spectroscopy (ESCA or XPS) conducted by
Novakov, Chang, and Harker (1974) and Barbaray, Contour, and Mouvier
(1977). Electron paramagnetic resonance spectroscopy has been used exten-
sively by Lunsford and co-workers, as well as others [Lin and Lunsford
(1975) and references therein] . Many measurements have also been made
with infrared spectroscopic techniques [e.g., Goodsel, Low, and Takezawa
(1972); Lin and Lunsford (1975); and references therein].
A number of the earlier studies yielded only minimum rates for SC>2
uptake and sulfate formation, because measurements were limited by gas-
phase diffusion to the solid surface or by the detection of SC>2 in the effluent
from a laboratory reactor. In most of the latter cases, SC>2 was detected in
the effluent stream only after partial saturation of the solid surface; conse-
quently, initial reaction rates could not be determined. Moreover, only a
few of the earlier studies reported on capacities for SC>2 removal.
In this work, we report on the rates and capacities of heterogeneous
reactions of SC>2 with a number of solids likely to be found in urban aerosols.
The rates were measured in the laboratory by means of a cylindrical flow
reactor in which the walls were coated with the solid of interest. This type
of configuration permitted us to measure initial rates as well as rates as a
function of time (SC>2 exposure). Analysis of the experimental results speci-
ficially accounts for mass transport in the reactor, yielding rates that depend
only on the surface processes responsible for SC>2 uptake.
These results indicate that SC>2 uptake, in most cases, occurs through
capacity-limited reactions that convert SC>2 to adsorbed sulfate. Initial up-
take rates are quite high. With time, however, the measured rates decrease
until, with prolonged exposure to SO2, the solids completely lose the ability
to remove this species from the gas stream. Quantitative projections of
these results to the atmosphere, by use of a model previously described
[Judeikis and Siegel (1973)], indicate that SO2~aerosol reactions are likely to
be most important at or near the emission source and that, after a short time
(~ 1 hr) in the atmosphere, they will lose their ability to remove SO2. However,
interaction with atmospheric ammonia could promote further reaction, as
discussed under Results.
We find that the relative humidity of the gas mixture is important in the
SC>2 reaction scheme. Although moisture does not alter reaction rates
appreciably for SC>2 uptake, in most cases SC>2 capacities increase significantly
with increasing relative humidity. The results obtained with humidified re-
action mixtures indicate that reactions taking place in adsorbed surface water
films may well be one of the most important factors in SC>2 uptake and adsorbed
sulfate production in aerosols.
18
-------
EXPERIMENTAL
A detailed description of the apparatus used in these experiments can
be found elsewhere [Judeikis and Stewart (1976)]. The apparatus was a flow
reactor consisting of two concentric Pyrex cylinders, the inner one coated
with the solid of interest. The leading 15 cm of the inner cylinder was left
uncoated in order to permit full development of laminar flow. A homogeneous
gas-phase mixture containing trace amounts of the SC>2 was allowed to flow
through the reactor, where this species could diffuse to the walls for removal
by heterogeneous reaction. This led to both axial (flow direction) and radial
concentration gradients for the trace species. The axial concentration
gradient was measured with a mass spectrometer, coupled to the cylinder by
means of a multiport rotary valve, and a series of small (0. 15-cm o.d. ) probes
whose intakes were centered along the axis of the inner cylinder. The results
were analyzed by use of a model that specifically accounted for mass trans-
port by diffusion and laminar flow [Judeikis and Stewart (1976)]. The analysis
yielded heterogeneous reactivities in terms of 0-values, the fraction of SC^-
solid collisions that are effective in removing SC>2. Runs were also done on
blank (uncoated) cylinders; these gave no indication of reaction between SC>2
and the cylinder walls (0< 10-7).
The coated cylinders were generally prepared from water-ethanoi (1:1)
slurries of the appropriate solid, except for several fly ash samples (the
Shawnee and Louisville ashes described in the following paragraph) that were
prepared with v/ater as the slurry medium. In addition, several samples of
MnO2 were saturated with dilute acid or base solutions before being prepared
as water-ethanol slurries. The slurries were deposited onto the Pyrex
cylinder and the coated cylinders allowed to air dry. They were subsequently
vacuum-dried at 10-4 Torr overnight in the reactor.
Except for fly ash samples, all gases, liquids, and solids used in these
studies, whether for sample preparation, experiments, or analyses, were
reagent grade materials. Two different forms of aluminum oxide were used:
Al2C>3 and a mixed oxide A^Og-AUOHJO. These materials gave similar re-
sults, which are combined herein under Al^C^. The fly ash samples, all
from coal-fired plants and supplied in part by the EPA, were from the Mohave
power plant on the Colorado River near Hoover Dam, the Cholla power plant
in Arizona, the River Bend power plant at Charlotte, North Carolina, the
Shawnee steam plant at Paducah, Kentucky, and Combustion Engineering
Louisville Gas and Electric, Louisville, Kentucky. Ashes from the Shawnee
facility were obtained from both mechanical and electrostatic precipitators.
Typical operating conditions used in the experiments included pressures
from 10-700 Torr, flow velocities of 1-30 cm s (average linear veloc-
ities of 0.05-1.5 cm s"1), and ambient temperatures (Reynplds numbers
< 50). Depending upon the reactivity of SC>2 toward a particular surface, sub-
ambient pressures were often required to measure nondiffusion-limited re-
activities [Judeikis and Stewart (1976)]. The concentration of SC>2 was varied
from 3-100 ppm, with occasional excursions up to 1000 ppm; the mass
spectrometer sensitivity toward SC>2 detection was~0. 3 ppm. Oxygen con-
centrations were varied from 0-10 percent. Higher oxygen concentrations
could not be used since they led to oxidation of the ionization filament in the
mass spectrometer.
19
-------
In several experiments, Mohave fly ash samples were exposed
sequentially to SC>2> NHg, and SO2- The ammonia exposures were done with
gaseous NH3. On occasion, the ammonia exposures interfered with the
operation of the reactor, possibly because of NH3 adsorption on tubing, valves,
etc. In such instan.ces, purging the system with NC>2 completely eradicated
the deleterious effects.
Selected solids were analyzed for their BET surface areas [Brunauer,
Emmett, and Teller (1938)] so that capacities for reaction could be related
to the solid's active surface area. These types of measurements are well
known, and an apparatus was built based upon the design found in a familiar
physical chemistry laboratory text [Shoemaker and Garland (1967)]. These
BET surface areas can be found in Table A-l.
Wet-chemical analyses for sulfate were performed on metal oxide
samples after exposure to SO-,. The procedure involved removing the ex-
posed, coated cylinder from the reactor, separating the solid from the
cylinder, washing the solid with distilled water, and analyzing the wash water
for soluble sulfate. Analyses were carried out by precipitating barium sul-
fate from the wash water by adding a dilute barium chloride-nitric acid
solution. Nitric acid was required to prevent coprecipitation of carbonate
ion; an excess of nitric acid was avoided in order to reduce the probability
of dissolving the barium sulfate.
In addition to the wet chemical analysis, selected SC>2-exposed metal
oxides and fly ashes were examined by means of x-ray photoelectron spec-
troscopy (ESCA). Two instruments were used. MnC>2 and fly ash samples
were analyzed on a Du Pont 650 B electron spectrometer that used a
magnesium x-ray source and was operated at about 350 watts. Quoted in-
strument resolution at the time of the experiment was 1.05 eV FWHM on an
Au4f (7/2) peak. These samples were prepared by dusting the powdered
oxides on double-sided tape. The experiments were performed on three
samples of each substance; one sample was a blank, and the other two had
been exposed to SO2 in the reactor. The second ESCA instrument was a
GCA/McPherson ESCA 36 photoelectron spectrometer equipped with a
cryopump that allowed pressures of 10~° Torr to be attained. This spectro-
meter also used a magnesium anode that emitted Ka x-rays at an energy of
1253.6 eV. Resolution of this instrument was 0.2 eV. Various samples of
the metal oxides were prepared on glass slides by the method that was used
in preparation of the flow tube samples. The coated slides were exposed to
SC>2 in the tubular reactor. After exposure, samples were carried in air to
the ESCA and analyzed in the usual manner. Metal oxides and salts examined
this way included MnO2, MnSC>4, MgO, MgSC>4, Fe2O3, Fe2 (804)3, Al^O^,
A\2 (304)3, Na2SO3, and combinations of these substances such as Al2 (304)
and Al2O3« All binding energies were referenced to the C , peak in order
to compensate for charging effects.
20
-------
TABLE A-l. HETEROGENEOUS REMOVAL OF SO,
OVER VARIOUS MATERIALS '
Material
MgO
Fe2°3
Mohave fly ash
A12°3
MnO3
Cholla fly ash
River Bend fly ash
e f
Shawnee fly ash (M) '
Louisville fly ash
PbO
e f
Shawnee fly ash (M) '
Charcoal
Shawnee fly ash (E)6)f
Shawnee fly ash (E)e'f
NaCl
Louisville fly ash
River Bend fly ash
o
BET surface area
(mY1) 105x0b
100
27.3 55
15. 2C 50c'd
215 40
109 30
30c,d
30d
10d
7d
7
5C
40. 7 3
2d
0.4C
0. 3
0.2C
-------
RESULTS
Results from a representative experiment for SOo removal over
Mohave fly ash are shown in Figure A-l. The triangles represent the ex-
perimentally measu'red SC>2 concentration gradient. The solid curve was
calculated from the laminar flow model [Judeikis and Stewart (1976)] for a
best-fit reactivity (0-value) of 4.4 X 10" % which represents the fraction of
gas-solid collisions that were effective in removing SC^. Reactivities de-
termined in this manner for a number of different solids are listed in
Table A-l, together with several measured BET surface areas and projected
atmospheric removal rates for SO2 (the latter are discussed in the following
section). The reactivities in Table A-l are averages of initial values deter-
mined for the most part from five or more separate samples. Uncertainties
in reactivities (standard deviations) are~30 percent and result primarily
from variation in S(>2 uptake from sample to sample.
The reactivities given in Table A-l were found to be independent of SC>2
and O2 concentrations, as well as relative humidity and total pressure to
within a factor of 2. Representative data illustrating this point are shown in
Table A-Z. These data were obtained on sequential runs on the same samples
in order to minimize sample-to-sample variations in reactivities. Thus, the
data indicate that heterogeneous removal of SO2 occurs through first-order or
psuedo-fir'st-order kinetics.
Reactivities were also found to be independent of the thickness of the
solid coatings used in these experiments. For example, MnC>2 coatings with
average thicknesses of 0.48, 0.96, 8. 1, and 64 (jtm all gave the same initial
reactivity to within 20 percent. These results indicate that only the outer
layer of particles in the film are effective in SC>2 removal since particle
diameters, as determined by scanning electron microscopy, ranged from a
few tenths of a micrometer to~0. 5 |o.m.
The results discussed thus far are for freshly prepared solid coatings.
With continued exposure to SC^, reactivities gradually diminish until, with
prolonged exposures, the solids become unreactive toward removal of SC>2.
This effect is illustrated in Figure A-2 for SC>2 removal over MnC>2f where
the reactivity relative to the initial reactivity 0Q is plotted as a function of
time (SO2 exposure). Similar effects were found for all of the materials
investigated.
Experiments such as that illustrated in Figure A-2 gave additional
evidence that only the outermost layer of particles were involved in SC>2
removal. The data in Figure A-2 are for an MnC>2 coating with an average
thickness of 0.48 |im. Data from an experiment conducted under virtually
identical conditions, except with an average film thickness of 8. 1 |o.m, are
essentially superimposable on those illustrated in Figure A-2.
Results such as those in Figure A-2 indicate SO? removal occurs
through capacity-limited reactions. Capacities for SC>2 removal can be-de-
termined from such experiments by measuring total SO-, uptake. Results
for various solids are given in Table A-3. It will be noted that, particularly
for MgO and MnC>2, capacities increase significantly with relative humidity.
22
-------
1.0
.0.01
1
1
2 4 6
DISTANCE ALONG CYLINDER AXIS (FLOW DIRECTION), cm
Figure A-l. Removal of SO2 by Mohave fly ash. Triangles represent
experimentally measured SC>2 concentrations. Solid
curve was calculated from the laminar flow model for
0 = 4. 4 X 1C)4. Experimental conditions: total pressure,
55 Torr; O? pressure, 6 Torr; partial pressure of SC>2
in the influent gas stream, 9 mTorr; relative humidity,
0 percent.
23
-------
TABLE A-2. EFFECTS OF SO2
RELATIVE HUMIDITY, AND
TOTAL PRESSURE ON SO2 REMOVAL RATES
• Solid
MgO
p
2
(mTorr)
2.6
19.5
P0
°2
(Torr)
5.7
5.9
Relative
humidity
(percent)
48
48
P
total
(Torr)
57
59
5
10 x 0
95
102
Mohave fly ash
1.6
2.0
0.0
4.9
48
52
48
52
52
54
4.0
3.9
11.0
11.0
0
95
102
106
42
51
MnO,
9.5
6.4
9.0
0.0
0.0
0.0
0
0
0
51
103
300
27
24
27
24
-------
0.8
0.6
0.4
0.2
K
20
S02 EXPOSURE, hr
40
60
"T
\
\
\
\
\
\
\
200 400
S09 EXPOSURE,
600
T
800
g
Figure A-2. Reactivity as a function of SC>2 exposure for SO2 removal
over MnO2. SO2 in nitrogen, 95-percent relative humidity.
The arrow indicates the stoichiometric point for the reac-
MnSCT
tion MnO2 + SO-
4'
25
-------
TABLE A-3. CAPACITIES FOR SO2 REMOVAL
Solid
MgO
Fe-O.
2 3
Mohave fly ash
AUG.
/ <
{-, -J
MnO0
2
Charcoal
Relative humidity
(percent)
0
50
95
0
50
0
50
95
0
53
95
0
25
58
50-95
95
0
56
95
Capacity .
(mg S02) (g solid)"
4
12
400
0.6
1.2
0.5
0.2
1.4
25
5
17
4
78
320
210
>530
1.3
0.8
5. 7
3,' •
As received. '*•• ••"
Probably minimum values (see text).
26
-------
This contrasts to reactivities that did not change, to within experimental
error, with relative humidity. However, capacities, like reactivities, were
found to be independent of SC>2 and C>2 concentrations, as well as total pres-
sure. The latter conclusions for capacities are based on a more limited
number of experiments.
Quantitatively, we can combine the capacity data in Table A- 3 with the
BET surface areas in Table A- 1 in order to determine the surface coverage
of these materials by SC>2. For the materials in Table A- 3, except MgO and
MnO2, this amounts to ~0.03-0.2 monolayer, if absorbed SC>2 is assumed
to .occupy ISA^. For MnO2 at 0-percent relative humidity, we find a com-
parable value of 0.05 monolayer. However, in the latter case, coverage
increases substantially with increasing relative humidity and is about seven
monolayers at 95-percent relative humidity. Similar conclusions probably
also apply for MgO, although we have not measured the BET surface area
of this material.
The capacities and surface coverages given in Table A- 3 and the pre-
ceding paragraph are based on the total weight of solids used in the experi-
ments. As such, they are probably minimum values, since capacity experi-
ments were generally carried out with coatings that consisted of multiparticle
layers. As noted above, the experimental evidence indicates that only the
outermost layer of particles participates in SO2 removal. Capacities and
surface coverages, therefore, are probably an order of magnitude greater
than those given above.
Wet chemical and ESCA analyses of these materials indicate SC>2 is
quantitatively (to within a factor of 2) converted to adsorbed sulfate. (Because
of the broad nature of the ESCA sulfate peak, we would not have been able to
detect a 5-10-percent contribution by sulfite or similar species.) One excep-
tion to this result was the case of A12O3 (and possibly charcoal), where both
types of analyses indicated that little, if any, sulfate was formed. Further,
the ESCA analyses revealed no detectable amounts of any sulfur -containing
species. The latter results indicate that SO2 uptake on AloO^ occurred by
reversible physical adsorption, the SO2 desorbing during trie evacuation to
10~" Torr prior to ESCA analysis. This interpretation is consistent with
results from the flow reactor, where we found that the reactivity of
exposed to SO2 until saturated could be restored by evacuating the sample at
10-4 for ~1 hr. The latter did not occur to any appreciable extent for the
other materials in Table A-l, except for charcoal.
The capacity- limited nature of the reaction, accompanied, in most
cases, by sulfate formation, suggested the possibility that the fly ash materi-
als as received may already have undergone substantial reaction with SO2.
Indeed, ESCA analysis of the Mohave fly ash, as received, indicated a strong
sulfate signal. For this reason, we examined the fly ash materials both as
received and after they had been washed with distilled water for removal of
soluble sulfates. As indicated in Table A-l, in most cases the washing led
to substantial increases in the fly ash reactivity.
In addition to washing materials with distilled water, we also examined
the effects of pretreatment with dilute acids or bases, since it has been
27
-------
suggested that ammonia plays an important role in the heterogeneous
oxidation of SC^, primarily through neutralization of H2SO4 [Junge and Ryan
(1958), Van den Heuvel and Mason (1963), Scott and Hobbs (1967), and McKay
(1971)]. We pretreated MnC>2 with a dilute NH4OH solution (0. 1 N), as well
as 0. 1 N solutions of NaOH, HC1, and H2SC>4. The results of these studies,
illustrated in Table A-4, indicate that the basic pretreatments substantially
accelerate initial reactivities toward SC>2 removal, whereas the acidic pre-
treatments have the opposite effect. ^
TABLE A-4. EFFECT OF BASIC AND ACIDIC TREATMENT OF
MnO2 ON INITIAL SO2 REMOVAL RATES
Pretreatment
NH4OH
NaOH
None
HC1
H2S04
Pso2
(mTorr)
0.9
17.0
M
1.3
0.8
P°2
(Torr)
0
1.4
0
0
0
Relative
humidity
(percent)
0
50
0
0
0
P
total
(Torr)
51
59
50
50
51
10 x 0
240 »
85^
30
5
2
MnO_ films prepared from distilled water slurries or 0. 1 N solutions
of base or acid.
In the case of the Mohave fly ash, we conducted an experiment to test
the possible rejuvenation of reactivity of spent material by exposure to
ammonia. In that experiment, the Mohave ash was exposed to SO? until it
was totally nonreactive toward this species. The SO? exposure was then
terminated and the sample exposed to ammonia (total ammonia exposure on
'a molar basis was <10 percent of the SO? exposure required to poison the
ash). The material was then reexposed to SO2 with the result that the re-
activity was restored to~50 percent of its initial value.
DISCUSSION
The results in Table A-1 indicate that a number of materials exhibit
substantial reactivity toward SO2- These reactivities may be used to esti-
mate atmospheric removal rates of SO2 through gas-aerosol reactions. Here
we use a previously derived model for gas-aerosol reactions [Judeikis and
Siege! (1973)], wherein the SO2 removal rate is given by
d(S02).
dt
= 0kc(A)(S02).
(A-l)
28
-------
In Eq. (A-l), kc is the average SO2 velocity in one dimension; (A), the
aerosol surface area per unit volume; (SO2Ja> the atmospheric sulfur dioxide
concentration, and 0, the fraction of SC>2-aerosol collisions that are effective
in removing SO2- The value of (A) can be calculated from an expression
given by Mottershead (1970) or by integration of actual aerosol distributions
[e.g., Heisler, Friedlander, and Husar (1973)], if particles are assumed to
be spherical. In either case, for an atmospheric aerosol loading of 100 fig
m-3, we estimate (A) ^ 1.5 y. 10~^ cmr cm-3. Using this value, and kc
calculated from simple kinetic theory [Present (1958)], we obtain
-d ln(SO ) i
_±_± = 0.120 sec (A-2)
From Eq. (A-2) and the 0-values in Table A-l, we obtained the projected
atmospheric removal rates given in the last column of Table A-l. It should
be noted that calculation of these rates was based on the assumption that the
total atmospheric aerosol burden had the same reactivity as the indicated
solid. Thus, for example, if the total atmospheric aerosol burden had the
same reactivity as MgO, the SC>2 removal rate would be 35 percent/hr~ *.
The above calculations estimate SC>2 removal rates based on simple
kinetic theory, i.e. , treatment of aerosol particles as large molecules. Al-
though this is appropriate for small aerosol particles (<~0.01 (Jim), it over-
estimates rates for larger particles because of a transition from free-
molecular flow to aerodynamic flow [Hidy and Brock (1970) and Fuchs and
Sutugin (1971)]. Using approximations for mass transfer given in the latter
references, and integrating over the aerosol distribution used above [Heisler,
Friedlander, and Husar (1973)], we estimate that these effects could reduce
the projected SOo removal rates given in Table A-l by approximately a factor
of 2.
In addition, the capacity-limited nature of the removal process
indicates that these reactions will be most important at or near the emission
source. If we assume that 0 -» 0 as the SC>2 exposure approaches 0. 1 g of SC>2
removed per gram of solid, and a linear relationship between 0 and the SO-
removed, we may write
0 = 01 1 - 10
r
(P)
(A-3)
where 0O is the initial reactivity and (SC>2)r and (P) are the concentrations
(in jig m~3) of SO2 removed and particles, respectively. If we let 0O = 1 x 10
and (P) = 100 fj.g m "3, Eq. (A-3) becomes
0 = 1 X 10"4 [l - 0. l(S02)r] (A-4)
But the rate of SO? removal from Eq. (A-2) is
d(SO )
= 43° * ahr
29
-------
for a particle loading of 100 (j.g m~ 3, where (802) is the atmospheric
concentration of SC>2, which we shall take as 50 p.g m-3. Substituting the
latter value and Eq. (A-4) into Eq. (A-5) and integrating yield
(SO.) = 10(1 - e-°-22t) nig m-3 (A-6)
C* i
[Note that substitution of Eq. (A-6) into Eq. (A-3) would give the exponential
type of decay in 0/0o indicated in Figure A-2.] Substitution of t = 1, 3, and
10 hr into Eq. (A-6) gives (SO;?)r = 2.8, 4.8, and 8.9 |J.g m~3, respectively.
Thus, since (SO?) -» 10 |j,g m-3 as t-* ®, fresh aerosols would lose 90 per-
cent of their activity toward removing SC^ from the ambient urban environ-
ment in only~ 10 hr.
At emission sources, however, SC>2 loadings can typically be an order
of magnitude or more greater than those in the surrounding urban environ-
ment [Newman, Forrest, and Manowitz (1975a, 1975b)]. Under these con-
ditions, most of the heterogenous interactions would take place on a time
scale of ~ 1 hr. Thus., the results of this study indicate that heterogeneous
removal of SOo (and conversion to sulfate) will occur primarily at, or near,
emission sources, in agreement with other recent conclusions [e.g., Foster
(1969); Newman, Forrest, and Manowitz (1975a, 1975b); Freiberg (1976);
and Lusis and Phillips (1977)]. However, the possibility for further reaction
exists as ambient air begins to mix with the plume from the source. The
latter conclusion is based on the experimental results involving the rejuvena-
tion of the reactivity of expended Mohave fly ash after exposure to ammonia,
an event that would occur on mixing of ambient air with an emission plume.
The significant increases in capacities found at higher relative
humidities for selected solids in these studies indicate that reactions taking
place in adsorbed water films are likely to be of primary importance in
atmospheric SO2~solid interactions. Additionally, although sulfate aerosol
production by gas-phase processes and reactions in liquid droplets can occur,
the results of these studies indicate that contributions to atmospheric sulfate
burdens by gas-solid reactions will be limited by atmospheric particle
burdens rather than SC>2 concentrations. This is the result of the capacity-
limited nature of the reactions, which, considering atmospheric SO2 and
aerosol burdens, indicates that only a fraction of the gaseous SC>2 in the
atmosphere can be converted to sulfate by these processes. This conclusion
could have a serious impact on source emission control strategies.
30
-------
BIBLIOGRAPHY
Barbaray, B. , J. P. Contour, and G. Mouvier. 1977. Sulfur Dioxide
Oxidation Over Atmospheric Aerosol-X-Ray Photoelectron Spectra of
Sulfur Dioxide Adsorbed on V-,0^ and Carbon. Atmos. Environ.
11:351-356. *" b
Brock, J. R. , and J. L. Durham. 1977. Review of the Heterogeneous
Oxidation of SO_. To appear.
Brunauer, S. , P. H. Emmett, and E. Teller. 1938. Adsorption of Gases
in Multimolecular Layers. J. Am. Chem. Soc. 60:309-319.
Burke, M. F. , R. K. Baker, and J. L. Moyers. 1973. The Interaction
of SO~ With Airborne Particulate Matter. J. Chromatogr. Sci.
11:575-578.
Cheng, R. T. , M. Corn, and J. O. Frohliger. 1971. Contribution to the
Reaction Kinetics of Water Soluble Aerosols and SO_ in Air at ppm
Concentrations. Atmos. Environ. 5:987-1008. See also Atmos.
Environ. 6:369-370 (1972).
Cheng, R. T. , J. O. Frohliger, and M. Corn. 1971. Aerosol Stablization
for Laboratory Studies of Aerosol-Gas Interactions. J. Air Pollut.
Control Assoc. 21:138-142.
Chun, K. C. , and J. E. Quon. 1973. Capacity of Ferric Oxide Particles
to Oxidize Sulfur Dioxide in Air. Environ. Sci. Technol. 7:532-538.
Corn, M. , and R. T. Cheng. 1972. Interaction of Sulfur Dioxide With
Insoluble Suspended Particulate Matter. J. Air Pollut. Control
Assoc. 22:870-875.
Devitofrancesco, G. , F. Panke, and B. M. Petronio. 1972. On the
Behavior of Some Gases During Adsorption on Dusts. Staub-Reinhalt.
Luft 32:21-23.
Environmental Protection Agency. 1975. Position Paper on Regulation of
Atmospheric Sulfates. PB-245 760, September, U.S. Department
of Commerce, National Technical Information Service.
31
-------
Foster, P. M. 1969. The Oxidation Sulfur Dioxide in Power Station Plumes.
Atmos. Environ. 3:157-175.
Freiberg, J. 1976. The Iron Catalyzed Oxidation of SO2 to Acid Sulfate
Mist in Dispersing Plumes. Atmos. Environ. 10:121-130.
Fuchs, N. A. , and A. G. Sutugin. 1971. High-Dispersed Aerosols. In:
Aerosol Physics and Chemistry, Vol. 2, G. M. Hidy and J. R. Brock,
eds. Pergamon Press, New York. pp. 1-60.
Goodsel, A. J. , M. J. D. Low, and N. Takezawa. 1972. Reactions of
Gaseous Pollutants With Solids. II. Infrared Study of Sorption of
SO- on MgO. Environ. Sci. Technol. 6:268-273.
Heisler, S. L. , S. K. Friedlander, and R. B. Husar. 1973. The Relation-
ship of Smog Aerosol Size and Chemical Element Distributions to
Source Characteristics. Atmos. Environ. 7:633-649.
Hidy, G. M. , and C. S. Burton. 1975. Atmospheric Aerosol Formation by
Chemical Reactions. Int. J. Chem. Kinet. , Symp. No. 1:509-541.
Hidy, G. M. , and J. R. Brock. 1970. The Dynamics of Aerocolloidal
Systems. Aerosol Physics and Chemistry, Vol. 1. Pergamon
Press, New York.
Judeikis, H. S. , and S. Siegel. 1973. Particle-Catalyzed Oxidation of
Atmospheric Pollutants. Atmos. Environ. 7:619-631.
Judeikis, H. S. , and T. B. Stewart. 1976. Laboratory Measurement of
SO? Deposition Velocities on Selected Building Materials and Soils.
Atmos. Environ. 10:769-776.
Junge, C. E. , and T. G. Ryan. 1958. Study of the SO2 Oxidation in
Solution and Its Role in Atmospheric Chemistry. Q. J. R. Meterol.
Soc. 84:46-55.
Lin, M. J. , and J. H. Lunsford. 1975. Photooxidation of Sulfur Dioxide
on the Surface of Magnesium Oxide. J. Phys. Chem. 79:892-897.
Lusis, M. A., andC. R. Phillips. 1977. The Oxidation of SO2 to Sulfates
in Dispersing Plumes. Atmos. Environ. 11:239-241.
McKay, H. A. C. 1971. The Atmospheric Oxidation of Sulfur Dioxide
in Water Droplets in the Presence of Ammonia. Atmos. Environ.
5:7-14.
Mottershead, C. T. 1970. Collision Rates Between Gas Molecules and
Aerosol Particles. Project Clean Air, Vol. 4, Task Force 7,
Appendix A, University of California.
32
-------
Newman, L. , J. Forrest, and B. Manowitz. 1975a. The Application
of an Isotopic Ratio Technique to a Study of the Atmospheric
Oxidation of Sulfur Dioxide in the Plume From an Oil-Fired Power
Plant. Atmos. Environ. 9:959-968. See also Atmos. Environ.
10:671-673 (1976).
Newman, L. , J. Forrest, and B. Manowitz. 1975b. The Application of an
Isotopic Ratio Technique to a Study of the Atmospheric Oxidation of
Sulfur Dioxide in the Plume From a Coal-Fired Power Plant. Atmos.
Environ. 9:969-977. See also Atmos. Environ. 10:671-673(1976).
Novakov, T., S. G. Chang, and A. B. Harker. 1974. Sulfates as Pollution
Particulates: Catalytic Formation on Carbon (Soot) Particles.
Science 186:259-261.
Okita, T. 1967. Adsorption and Oxidation of Sulfur Dioxide at Ordinary
Temperature 1. Measurement of Atmospheric Acid Particles and
Laboratory Experiments on the Adsorption and Oxidation of Sulfur
Dioxide on the Surface of Particles at Room Temperature. Inst.
Public Health Res. Rep. 16:52-58.
Present, R. D. 1958. Kinetic Theory of Gases. McGraw-Hill, New York.
p. 33.
Scott, W. D. , and P. V. Hobbs. 1967. The Formation of Sulfate in Water
Droplets. J. Atmos. Sci. 24:54-57.
Shoemaker, D. P. , and C. W. Garland. 1967. Experiments in Physical
Chemistry, ed. 2. , McGraw-Hill, New York, pp. 262-271.
Smith, B. M. , J. Wagman, andB. Fish. 1969. Interaction of Airborne
Particles With Gases. Environ. Sci. Technol. 6:558-562.
Urone, P. , H. Lutsep, C. M. Noyes, and J. F. Parchev. 1968. Static
Studies of Sulfur Dioxide Reactions in Air. Environ. Sci. Technol.
2:611-618.
Van den Heuvel, A. R. , and B. J. Mason. 1963. The Formation of
Ammonium Sulfate in Water Droplets Exposed to Gaseous Sulfur
Dioxide and Ammonia. Q. J. R. Meterol. Soc. 89:271-275.
33
-------
APPENDIX B
LABORATORY MEASUREMENT OF SO2 DEPOSITION
VELOCITIES ON SELECTED BUILDING
MATERIALS AND SOILS
INTRODUCTION
Deposition velocities of pollutant gases are used extensively in calculat-
ing atmospheric budgets for these species [e.g., Robinson and Robbins (1970)
and Kellogg .et al. (1972)]. Both field and laboratory measurements of these
quantities have been made. Field measurements generally employ one of two
methods for determination of deposition velocities. The first involves simul-
taneous measurements of wind velocity, temperature, and pollutant gas con-
centration profiles above the surface [e.g. , Garland et al. (1973, 1974);
Shepherd (1974); Dannevik, Frisella, and Fishman (1974); and Whelpdale and
Shaw (1974)]. The vertical atmospheric diffusity K(z) is estimated from the
former two quantities, and the deposition velocities V calculated from
o
F = -K(z)^ = Vgc (B-l)
relating the downward flux (F) of the pollutant gas to K(z) and the concentra-
tion gradient. In Eq. (B-l), it is assumed that the downward flux of the
pollutant gas may be treated as diffusive transfer. The concentration c
[actually lim c(z)] is generally measured at some fixed height above (but near)
the surface.
The second method, which is also used extensively in the laboratory, is
based on total uptake of SO2 [Braun and Wilson (1970); Seim (1970); Hill
(1971); Abeles et al. (1971); and Cox and Penkett (1972)], frequently employ-
ing 35s labeled SO2 [Spedding et al. (1969a, 1970a, 1970b, 1971, 1972a,
1972b); and Owers and Powell (1974)]. In the latter case, the total uptake of
35sOŁ is measured, as well as its concentration just above the surface. The
deposition velocity is then readily calculated from Eq. (B-l). In some cases,
flow systems are also used for laboratory measurements [Chamberlain
(1966); Spedding (1969b, 1972c); Brimblecombe and Spedding (1972); and
Payrissat and Bielke (1975)].
Measured deposition velocities typically range from a few tenths of a
centimeter per second or less to several centimeters per second [e.g. ,
Spedding (1972b)]. Substantial variations in the magnitude of the deposition
34
-------
velocity determined at a given field site or for a given material in the
laboratory are common [Garland et al. (1973, 1974); Shepherd (1974); and
Whelpdale and Shaw (1974)]. These variations may be related in part to the
failure of the assumptions inherent in Eq. (B-l) as well as to surface changes
that are dependent on environmental conditions. For example, SC>2 uptake by
leaves is controlled largely by the stornata [Meidner and Mansfield (1968)].
The opening and closing of the stomata depend on a number of environmental
factors such as daylight, relative humidity, and season. In the laboratory,
deposition velocities have been found to depend on the degree of gas phase
mixing employed in static systems [Spedding (1972b)] and on gas flow rates
in dynamic systems [Lawrence (1964) and Spedding (1972c)], the measured
deposition velocities increasing with higher degrees of mixing or flow rates.
The latter results indicate that measured deposition velocities, in many
cases, represent values that are limited by mass transport to the surface.
Questions then arise as to the limits of deposition velocities imposed by
physical and chemical processes related to the actual removal of the pollutant
gas at the surface.
In this work, we present a method for laboratory measurement of de-
position velocities independent of mass transport phenomena, together with
experimental results for SC>2 removal on several environmental surfaces.
The values obtained in this manner represent the maximum deposition
velocities that would be encountered in the open atmosphere, particularly
when turbulent mixing is sufficiently high to remove mass transport
limitations.
EXPERIMENTAL
Apparatus
. A block diagram of the apparatus used in these experiments is illus-
trated in Figure B-l. This system, which is basically a cylindrical flow
reactor, is similar to systems previously described [Hedgpeth et al. (1974)
and Stewart and Judeikis (1974)]. The major difference between the present
system and the ones previously described is the method of analyzing gases
flowing through the reactor. In addition, system components that were found
to be reactive toward SC>2 were replaced. Virtually all components in the
final version of the modified system consisted of Pyrex glass, 316 stainless
steel, and Teflon-coated aluminum.
In the system shown in Figure B-l, a carrier gas stream was initially
split into two streams; one of the streams passed through a humidifer, where
it was saturated with water vapor, and the two streams were subsequently
recombined. (The ratio of flow rates of the split streams determined the
relative humidity of the carrier gas. ) The carrier gas stream was then
mixed with a small amount of nitrogen that contained traces of SC^, and the
mixture was fed into the cylindrical flow reactor (2. 5-cm radius) that con-
tained a concentric Pyrex cylinder (2. 1-cm radius) coated with the solid of
interest. (The choice of a cylinder for a substrate was not unique, and other
geometries, such as parallel plates, could have been used. ) The latter
35
-------
PRESSURE
GAUGE
FLOW-
METER
VALVE
^4-WAY VALVE
X REACTION CHAMBER
r^PUMP
VALVE
PRESSURE
GAUGE
FLOWMETER
SO2/N2
GAS SAMPLING PROBES
PUMP
MASS SPECTROMETER (gas analysis)
VALVE
"=MI
HUMIDIFIER
GAS MIXER
AIR
Figure B-l. Block diagram of cylindrical reactor.
-------
cylinder was coated by preparing a slurry of the solid of interest, coating the
blank Pyrex cylinder (outside the reactor), and permitting the coating to air
dry and then dry overnight in vacuum in the tubular reactor. Surface rough-
nesses of the dried films were typically < ~ 1 mm.
Reaction of SC>2 with the coated walls led to a concentration gradient for
SC>2 along the axial (as well as radial) directions. (In the absence of a solid
coating, there was no change in the SOŁ concentration on passage through the
reactor. ) For measurement of the axial concentration gradient, the gas
mixture in the reaction chamber was sampled by means of a set of small
probes (connected by a 16-port rotary valve to a mass spectrometer), whose
intakes were center-ed along the axis of the coated cylinder. The outside and
inside diameters of the probes were nominally 0. 15- and 0.08-cm, respec-
tively. Flow through the sampling system was sufficiently slow that the flow
pattern in the reaction chamber was essentially undisturbed [Westenberg,
Raezer, and Fristrom (1957)], yet sufficiently fast that transit time through
the sampling system was minimal (~ 3-4 sec).
Typical operating conditions employed were pressures of 10-700 Torr,
flow velocities of 1-30 cm^/sec (average linear velocities of 0.05-1.5 cm/
sec), and ambient temperatures (Reynolds numbers <50). Subambient pres-
sures were frequently required to measure nondiffusion-limited deposition
velocities. (This point is discussed more fully in subsequent sections of this
appendix.) The flow rates chosen gave a sufficiently high axial SC>2 concentra-
tion gradient to permit accurate measurements of this quantity.
Gases sampled by means of the probes were analyzed with a mass
spectrometer. The 0. 15-cm-o.d. tubing continued into the mass spectrometer
chamber, terminating just before the ionizer. Thus, effluent gases from the
probe were injected directly into the ionizer. The sensitivity of the mass
spectrometer for SC>2 detection was~0. 3 ppm. Consequently, experiments
were conducted with initial SC>2 concentrations >3 ppm. In addition, high
concentrations of oxygen in the reaction mixture tended to oxidize the fila-
ments in the mass spectrometer. For this reason, oxygen concentrations
were limited to~ 10 percent or less.
Materials
Solids examined in this study consisted of commercial formulations of
cement, ready-mix cement (cement containing sand and gravel), asphalt, and
exterior stucco. In the case of cement and exterior stucco, samples from
two different sources of each material were used. Soil samples of sandy
loam and adobe clay taken from the Los Angeles area were also examined.
In most cases, these materials were sifted through a screen in order to
eliminate particles > 1 mm in diameter. Water-based slurries of these
materials were employed in preparing the coated Pyrex cylinders (except for
asphalt, where a trichloroethylene slurry was used). Consequently, the
cement, ready-mix cement, and exterior stucco were cured during the pro-
cess of preparing the coatings. Surface roughnesses were typically
<~ 1 mm.
37
-------
Gases used in this study were reagent grade gases obtained from
Matheson and were used as received. Two specially prepared mixtures were
used for SO2 and oxygen in order to achieve the desired concentrations of
these gases in the reaction mixture. These were 1000-ppm SO2 in N2 and
20-percent 03 in -N;>. In addition, distilled water was used for humidifying
gas mixtures.
Data Analysis
Mass transport in a cylindrical flow tube, under conditions of non-
turbulent flow and at steady state, is described by [e.g. , Walker (1961),
Stewart and Judeikis (1974), and references therein]
= 0 (B-2)
subject to the boundary conditions
c = CQ at r, x = 0 (B-3) .
•|Ł = 0 at r = 0, x (B-4)
and
- D-g- = 0 k catr = R, x> 0 (B-5)
In Eqs. (B-2) through (B-5), r and x are the radial and axial coordinates; c
is the concentration of the reacting species (initial concentration of co); D is
the diffusion coefficient of the reacting gas in the mixture; Vx is the linear gas
flow velocity in the axial direction; kr ( = \B?T /2irM, where 3tf, T, and M are
the gas constant, absolute temperature, and molecular weight of the diffusing
gas, respectively) is the molecular velocity of the reacting species in the
radial direction; and R is the cylinder radius. [Throughout this report,
binary diffusion coefficients were calculated for SO2 in nitrogen by the use of
expressions given by Present (1958). The presence of oxygen or water vapor
in the reaction mixture would lead to diffusion coefficients slightly different
than the calculated values. The uncertainties arising from these differences
are less than the uncertainties arising from other sources.]
Equation (B-5) expresses the condition that the diffusion of the reacting
species to the walls is equal to its removal by heterogeneous reaction. In
that equation, krc is the gas-solid collision frequency, and the reactivity Q,
a dimensionless parameter, is the fraction of collisions that lead to removal
38
-------
of the reacting species from the gas phase. Actually c1, the concentration
at one mean free path away from the walls, should be used in Eq. (B-5) in
place of c [Paneth and Herzfeld (1937) and Stewart and Judeikis (1974)].
However, except for 0^1, the two are essentially equal.
The deposition velocity Vg is related to 0 as can be seen by comparing
Eqs. (B-l) and (B-5). Equating the right-hand sides of those equations yields
the result
V = 0k (B-6)
g Y r
Thus, the deposition velocity over a given material can be obtained from
laboratory determinations of 0-values. Note also that the deposition velocities
determined in this fashion correspond to values at one mean free path above
the .surface.
Solution of Eq. (B-2) is generally accomplished by making several
simplifying assumptions. One of these is the assumption of plug flow
For first-order or psuedo-first-order processes, 0 is actually composed
of a collection of constants, including the sticking coefficient, as well as
the rate constants for adsorption, desorption, and surface reaction. Con-
sider, for example, a reaction scheme involving first-order adsorption,
desorption, and surface reaction processes. Equation (B-5) would then be
rewritten as
r)r
- D SŁ = k c - k ,c = v(l - f)k c - k ,c
9r a d a YV ' r da
where k and kj are the rate constants for adsorption and desorption, ca is
the concentration of adsorbed c, and -y and f are the sticking coefficient and
fractional surface coverage, respectively. If we assume a steady state for
c , we may write
3,
= v(l - f)k c - k,c - k c = 0
IV r da
where ksca is the rate of surface reaction of adsorbed c. Solving the latter
equation for ca and substituting the result into the former equation, we find
Comparison of this expression with Eq. (B-5) indicates that for the case
discussed here
39
-------
(Vx = constant). The solution in this case is [e.g. , Walker (1961), Stewart
and Judeikis (19?4), and references therein]
r r
F0Qi R
P.x
e i
(B-7)
where Jn(Qf. -^) and 3 Aot.) are Bessel functions of the first kind
U 1 ix 1 1
D
6 =
(B-8)
V
1 -
(B-9)
and a. is the i root of
(B-10)
[V = 2Vaverage (1 - r2/R2)], solutions to Eq. (B-
livalent heat transfer problems by Sideman, Luss,
2)
In the case of laminar flow
have been obtained [for equiva
and Peck (1965) and references therein] where axial diffusion can be neglected
[D (32c/3x2) «=> 0]. [Criteria necessary for this assumption were delineated
in an analogous heat transfer problem by Singh (1958)]. In general, we find
that the conditions for which these solutions apply in our experiments are of
limited use in the determination of values for 0. The reason for this is that
reactions tend to become diffusion limited under experimental conditions
where axial diffusion can be neglected, particularly for high reactivities.
Examples of this point are illustrated below. Consequently, numerical
solutions of Eq. (B-2), with laminar flow, were required for the cases of
interest here. These were obtained by using a modification of the method of
finite differences [Jenson and Jeffreys (1963)].
The geometry of our system is such that laminar flow is not fully
developed at the entry to the coated cylinder. [Laminar flow is fully developed
after entry to the reaction chamber. However, the flow pattern is disrupted
when the gas stream encounters the leading edge of the coated cylinder
(Figure B-l). For our typical operating conditions, several centimeters
would be required for laminar flow to be reestablished (Betz (1966)]. This
generally presents no problem, however, since we find that, under most
experimental conditions, either the plug or laminar flow models adequately
describe our experimental results (in fact, in many cases, concentration
profiles calculated from either model are indistinguishable) and yield 0-values
that agree to within a few to 20 percent. (Values derived from the plug flow
model are always lower than those derived from the laminar flow model. )
40
-------
The major discrepancy between 0-values derived from the two models
occurs at high pressures (~ 700 Torr) and high reactivities (0> 10~'*). Under
these conditions, SC>2 removal tends to become diffusion limited. Although
these conditions are avoided in most experiments (see below), they do pro-
vide an opportunity to distinguish between the two models (since the solutions
become independent of 0). In such cases, we generally find that SO? concent-
ration gradients calculated from the plug flow model are more consistent
with experimentally measured values. Consequently, the plug flow model
was used for the analysis of data reported here. As noted above, any
deviations from this model would result in slightly higher values for 0 (or Ve)
than are reported below, by anywhere from a few to 20 percent.
In practice, data were analyzed by one of two methods. The first con-
sisted of comparing experimental SC>2 concentration gradients to those cal-
culated from Eq. (B-7) for the given experimental conditions and various
values of 0 until the best fit was obtained. The second, shorter method made
use of the fact that only the first term in Eq. (B-7) contributes to the con-
centration at large axial distances [Stewart and Judeikis (1974)]. Thus, the
SC>2 concentration gradient becomes exponential for large x. In this case,
the limiting exponential slope from the experimental concentration profile
was compared to those calculated from the first term of Eq. (B-7) for the
given experimental conditions and various 0-values until the best fit was
obtained.
RESULTS
Values of 0 derived from a number of measurements of SC>2 removal
over various solids are given in Table B-l, together with deposition velocities
calculated from Eq. (B-l) for a temperature of 25°C. The values reported
in Table B-l represent averages from three to six experiments on each
material investigated. (Each experiment was conducted with a fresh sample
of the given material. ) In the case of cement and exterior stucco, data on
the material from different sources are reported individually.
The 0-values determined from consecutive measurements of SC>2 con-
centration gradients on a given sample usually agreed to within 20-30 percent.
There were comparable variations in 0-values from sample to sample of the
same material (for an equivalent SO2 exposure, where exposure is defined
as the time-integrated quantity of SC>2 to which the solid was exposed). (The
effects of SC>2 exposure are discussed below.) Overall, the standard devia-
tions for the values reported in Table B-l are about 40 percent.
The values reported in Table B-l were generally found to be independent
of SC>2 concentrations over variations of one to two orders of magnitude.
(The minimum partial pressure of SC>2 used in these experiments was
~ 0. 15 mTorr. ) Representative data illustrating this point are showp in A of
Table B-2 for exterior stucco-I. (Here, as in the other data presented in
Table B-2, the comparisons were made in sequential runs on the same sample
of a given material in order to minimize uncertainties arising frprn sample
variations. ) Thus, SC>2 removal over freshly prepared samples of these
solids follows apparent or psuedo-first-order kinetics.
41
-------
TABLE B-l. EXPERIMENTAL RESULTS FOR SC>2 REMOVAL
Material
Cement-Ia
Ready-mix cement
Exterior stucco-I
Cement-H
Exterior stucco-II
Adobe clay soil
Sandy loam soil
Asphalt
0
3.2 x 10"4
2.6 x 10"4
2. 3 x 10"4
2.0 x 10"4
1. 1 X 10"4
8.4x 10"5
8.3x 10"5
5. 1 x 10"6
Vg
(cm/sec).
2.5
2.0
1.8
1.6
0.86
0. 66
0.65
0.04
Cured.
Reactivities as a function of oxygen concentration and relative humidity
were also examined, and 0 was found to be independent of these parameters
to within experimental error. In Table B-2, representative results from
these experiments over ready-mix cement and sandy loam soil are presented
in B and C, respectively. In the case of oxygen, problems with oxidation of
the mass spectrometer filaments limited the oxygen concentrations used to
<~ 10 percent. However, even with those limitations, the oxygen concentra-
tion exceeded that of SO? by factors ranging from~ 10^ to 104 (except, of
course, for experiments conducted in the absence of oxygen).
For materials exhibiting reactivities of ~ 10"^ or greater, measure-
ments made at atmospheric pressure yielded SO2 concentration gradients
approaching the diffusion-limited value. Consequently, values derived from
such measurements were subject to large uncertainties. To illustrate this
point, we show several concentration profiles that were calculated from our
model for typical experimental conditions at atmospheric pressure. It will
be seen that a reactivity > 10-3 results in a diffusion-limited SO2 concentra-
tion gradient, whereas the gradient for 0 = 10~^ differs by only 10 percent
from the diffusion-limited gradient.
The large uncertainties resulting from experiments conducted near the
diffusion limit we.re avoided by altering the experimental conditions for those
materials that exhibited reactivities approaching 10 . In principle, several
parameters could he varied; in practice, however, the total pressure was the
most sensitive and the most easily varied parameter. The effects of reducing
the total pressure can be seen by comparing Figures B-2(a) and B-2(b). In
the latter case (for 0. 1-atm total pressure), the concentration gradients
differ by approximately a factor of 2 for reactivities of 10-3 and 10-4.
42
-------
TABLE B-2. REACTIVITIES AS A FUNCTION OF SO2 AND O2
CONCENTRATIONS, RELATIVE HUMIDITY, AND
TOTAL PRESSURE3-
Pressure, Torr
Parameter . 3
varied Material Total 2 2
A. SO- concentration Exterior stucco-I 55
55
B. O concentration Ready-mix cement 58
58
C. Relative humidity Sandy loam soil 100
100
UJ
D. Total pressure Sandy loam soil 50
400
2.6
2.6
0.0
6.2
4.4
4. 3
0.0
0.0
1. 1
13.2
1.6
1.6
3.7
4.2
4.2
4.8
Relative
humidity
(percent) 0
28
28
57
57
0
100
50
50
2.4x 10'4
2.2 x 10"4
2.0 x 10"4
2.4 x 10"4
6. 1 x 10'5
5.9 x 10"5
8. 3 x 10"5
7.4 x 10"5
Flow rates in all of these experiments were nominally 10 cm /sec.
b_ ,
Cu reel.
-------
H
Not
0.5
Figure B-2. Calculated SO? concentration gradients. Gradients calculated for total
pressures of (a) 1.0 and (b) 0. 1 atm. In both cases, R = 2. 1 cm,
T = 25° C, and V = 1 cm/sec.
-------
Since subambient pressures were frequently used in these experiments,
the effects of total pressure on measured reactivities were examined. In
general, the 0-values were independent of total pressure, to within experi-
mental uncertainties, for pressures ranging from~ 50-500 Torr. This
point is illustrated for SO- removal over sandy loam soil in D of Table B-2.
An attempt was also made to analyze solids after reaction for sulfate
formed. Wet chemical methods were employed. These attempts were
largely unsuccessful because of interferences by various species present in
the unexposed samples. However, in a related study on SO? removal by
metal oxides and other materials (to be reported later), wet chemical and
photoelectron spectroscopy methods indicate a near quantitative conversion
of SC>2 to sulfate. [Seim (1970) obtained similar results upon exposing
various soils to SO2- j
The reactivities and deposition velocities reported above are for
removal over freshly prepared coatings. With time (SCK exposure), these
reactivities diminish as the capacity to remove SO-, is expended. This
saturation effect is illustrated in Figure B-3 for adobe clay soil. In general,
this type of behavior was found with all of the solids investigated in this
study.
The capacities for SC>2 removal (total SCU removed from the initial
exposure until complete saturation) can be determined from experiments
such as that illustrated in Figure B-3. Values measured for several of the
solids listed in Table B-l (the adobe clay and sandy loam soils, ready-mix
cement, and exterior stucco-I) range from 0.04-0.6 g SO2/m2 of solid sur-
face for dry reaction mixtures and from 0. 4-2. 8 g SOz/rrw of solid surface
for humidified reaction mixtures (50-95-percent relative humidity). Typi-
cally, we found that capacities for humidified reaction mixtures were a factor
of 3-10 higher than those for dry mixtures. The number of experiments
conducted to measure capacities is not sufficient for an accurate determina-
tion of the relationship between the capacity and relative humidity. The
limited data do indicate, however, that the capacity for SC>2 removal from
humidified reaction mixtures does not depend on relative humidity as long as
the latter is>~ 30-40 percent. Other than the relative humidity, parameters
such as the SO2 and O2 concentrations and the total pressure did not appear
to have any significant effect on capacities for SO2 uptake.
Although the experimental results suggest only a limited capacity for
SO2 removal by the ground-level surfaces examined here, several possibilities
exist for continued removal in the open atmosphere. For example, rain could
wash away soluble surfaces (or other products), rejuvenating the surfaces for
further SO2 uptake [Braun and Wilson (1970) and references therein]. Several
authors [Spedding (1972b) and Payrissat and Beilke (1975)] have suggested that
SO? removal may be pH limited (e.g., sulfuric acid is formed from SO2 taken
up oy the surface, with the reaction gradually decreasing as the acid concentra-
tion builds up). Interaction with atmospheric ammonia could diminish such an
effect. Of course, sulfates are nutrients for plant growth, and sulfates formed
on soils could be removed by this process.
45
-------
r=o
X, cm
Figure B-3. Measured SO2 concentration gradients as a function of time
(SO2 exposure). Experimental parameters for SO2 removal
over adobe clay soil: P(total) = 300 Torr; P(C>2) = 19 Torr;
P(SO2) = 22 mTorr; T = 24° C; Vx = 0 5 cm/sec. Gradients
after exposures to SO2 of 3. 6 min (-A-), 2. 7 hr (—O-), and
7. 7 hr (- D-), or 0.009, 0. 39, and 1. 1 g SO2/m2 of solid
surface, respectively. Data points are irom experimental
measurements. Solid curves are calculated for 0 = 1.0 /
10-4 (-A-), 1.2 x 10-5 (-O-), and 2.2 y 10'6 (-O-).
46
-------
Several additional experiments were carried out in order to examine
these possibilities. In one experiment, a sample of ready-mix cement was
exposed to SO2 at 95-percent relative humidity until the capacity of this mate-
rial for SC>2 removal was completely expended. (In general, we found that
SC>2 removal was an irreversible process. Thus, termination of SO2 expo-
sures and evacuation of solid samples did not lead to any desorption of SC>2 or
restoration of the ability of the solid to remove SC^.) The coated cylinder was
then removed from the reactor, and the coating was rinsed with distilled water
and allowed to air dry. The coated cylindj»^aa|then replaced into the reactor,
dried overnight in a vacuum, and subsequently reexposed to SO2 at 95-percent
relative humidity. Experimental measurements indicated a complete restora-
tion of the ability of ready-mix cement to remove SCs (e. g. , to within the
experimental uncertainties noted above, the ready-mix cement exhibited the
same reactivity as a freshly prepared, previously unexposed sample).
In another experiment, adobe clay soil was exposed to SO? in a dry
reaction mixture until completely saturated. The gas mixture was then
humidified (95-percent relative humidity), and, again, the reactivity toward
SC>2 removal was completely restored.
The effects of ammonia were examined in an experiment with a sample
sandy loam soil. The sample was exposed to SC>2 (95-percent relative hu-
midity) until completely saturated. The SC>2 exposure was then terminated,
and the sample was exposed to NH3 (the total NH3 exposure was only ~ 20 per-
cent of the SC>2 exposure required to initially saturate the sample). Following
the exposure to ammonia, the system was purged with nitrogen and then re-
exposed to SOo. The result, again, was a complete restoration of the activity
of the sandy loam soil toward SO2 removal.
DISCUSSION
In the analysis of data obtained from these experiments, we specifically
account for transport-related phenomena. Thus, the deposition velocities
given in Table B-l represent values that are limited only by the adsorption
and chemical processes leading to SC^ removal from the gas phase. These
values, then, represent the maximum deposition velocities that would be
encountered over the materials listed in Table B-l under turbulent atmospheric
conditions.
_4
Experimentally, for materials with reactivities > 10 , such as exterior
stucco or cement, we found it necessary to conduct experiments at subambient
pressures in order to obtain nondiffusion-limited reactivities. Although such
conditions deviate from the ambient atmosphere, the results are more
applicable than would be results obtained from experiments conducted at
atmospheric pressures. The reason for this is that diffusivities for SC^ in
our experiments, which are conducted under nonturbulent conditions, are
~ 0. 1 cm2/sec at atmospheric pressure [Fish and Durham (1971)]. Generally,
however, turbulent atmospheric eddys are consideredably higher than this,
by factors of ~ 103-105 [Csanady (1973) and Heines and Peters (1974)]. Thus,
47
-------
a process that would be diffusion limited in our laboratory experiments would
more likely be limited by the adsorption and chemical processes responsible
for uptake than by transport to the surface in the open atmosphere, particularly
under turbulent conditions.
Of course, a viscous sublayer, whose thickness depends on a number of
environmental factors (e.g. , surface roughness), exists near the surface
where diffusivity approaches molecular diffusion [Csanady (19*73)]. As the
thickness of this boundry layer increases, the SC>2 uptake will tend to become
more diffusion limited. Thus, the actual deposition velocities in the environ-
ment will range from the maximum values reported here in turbulent atmo-
spheres to those determined by molecular diffusion in quiet atmospheres.
An added feature of the type of experiment reported herein is the
ability to measure changes in deposition velocities with time (SO2 exposure).
In a number of measurements reported in the literature, materials are
exposed for a fixed period of time and total SC>2 uptake determined. Such
measurements can only give an average value for the deposition velocity,
the magnitude of which will depend upon the degree of saturation of the solid
under study.
It is instructive to compare our results with other data reported in the
literature for related materials. In a study conducted on seven European
soils in a system in which a fan was used to mix the air above the soil,
Payrissat and Beilke (1975) reported deposition velocities of 0. 19-0.60 cm/
sec. These authors also observed first-order kinetics for SO^ removal, saw
evidence of saturation, and measured a slight dependence of removal rates on
relative humidity. In an additional study on five soils from the midwestern
United States, Seim (1970) measured average deposition velocities of 0.2 cm/
sec. He also found that deposition velocities were relatively independent of
SC>2 concentrations (first-order kinetics) and moisture levels.
Measurements of deposition velocities over building materials, notably
limestone, have been reported in the literature [Spedding (1969a, 1969b) and
Braun and Wilson (1970)]. Reported values are in the range of 0.03-0. 3 cm/
sec, which are considerably lower than the values we find for cements and
stuccos. ' These differences may result, in part, from differences in the
materials used and, in part, from values derived from diffusion-limited
experiments in some of the earlier work. However, Braun and Wilson (1970)
obtained values of 2.4-2. 6 g/m^ for the sulfur content of limestone exposed
to atmospheric SO^; these values compare favorably with the higher capacities
for SC>2 uptake we have measured in humidified reaction mixtures.
Several interesting possibilities are indicated by the deposition velocities
and capacities for SC>2 uptake measured here and in other laboratories. If
we assume an average deposition velocity of 1 cm/sec and an atmospheric
SC>2 concentration of 0. 1 ppm, from Eq. (B-l), we calculate a deposition
rate of 2. 6 x 10"" g/m^ sec. If we further assume a capacity of 2. 5 g SC>2/
m2 of solid surface, we conclude that the ability of a solid surface to remove
SO2 from the atmosphere will be expended in 11 days, in the absence of any
48
-------
processes such as precipitation thatmight act to rejuvenate the surface
activity for SO? removal. In an urban area such as Los Angeles, where
midsummer precipitation is negligible, this could result in higher SOŁ
concentrations than would otherwise be experienced. Of course, this
type of calculation and conclusion is greatly oversimplified for a number
of reasons.
Other variables enter into application of the data in Table B-l to the
environment, such as surface roughness and total areas as well as source
strengths. In general, our samples had surface roughnesses <~ 1 mm.
Surface roughnesses in the environment are usually greater than this, in
some cases by large factors. In the environment, therefore, the actual
surface area available for uptake could be significantly greater than that
available in our reactor. Of course, vegetation would have a very high ratio
of actual to ground-level surface areas; however, as noted above, the uptake
by vegetation, which is largely controlled by the stomata, would be sensitive
to environmental factors [Meidner and Mansfield (1968)].
In addition, we indicated several possibilities above for rejuvenating
saturated surfaces. The few experiments we conducted to explore these
possibilities supported those suggestions. Thus, in the atmosphere, uptake
of SO? may be determined by the balance of rates of saturation and rejuvena-
tion o± the active surface.
49
-------
BIBLIOGRAPHY
Abeles, F. B. , L. E. Craker, L. E. Forrence, and G. R. Leather.
1971. Fate of Air Pollutants: Removal of Ethylene, Sulfur
Dioxide, and Nitrogen Dioxide by Soil. Science 173:914-916.
Betz, A. 1966. Introduction to the Theory of Flow Machines.
Pergamon Press, London.
Braun, R. C. , and M. J. G. Wilson. 1970. The Removal of Atmospheric
Sulphur by Building Stones. Atmos. Environ. 4:371-378.
Brimblecombe, P. , and D. J. Spedding. 1972. Rate of Solution of
Gaseous Sulphur Dioxide at Atmospheric Concentrations.
Nature 236:225.
Chamberlain, A. C. 1966. Transport of Gases to and From Grass and
Grass-Like Surfaces. Proc. R. Soc. London A290:236-265.
Cox, R. A. , and S. A. Penkett. 1972. Effect of Relative Humidity on
the Disappearance of Ozone and Sulphur Dioxide in Contained
Systems. Atmos. Environ. 6:365-368.
Csanady, G. T. 1973. Turbulent Diffusion in the Environment, Vol. 3.
D. Reidel Publishing Company, Boston, Massachusetts.
Dannevik, W. P. , S. Frisella, and J. Fishman. 1974. RAPS 1974 2
Surface Flux Measurement Program, Final Report. Environmental
Quality Research, Inc. , Clayton, Missouri.
Fish, B. R. , and J. L. Durham. 1971. Diffusion Coefficient of SO2
in Air. Environ. Lett. 2:13-21.
Garland, J. A., W. S. Clough, and D. Fowler. 1973. Deposition of
Sulphur Dioxide on Grass. Nature 242:256-257.
Garland, J. A., D. H. F. Atkins, C. J. Readings, and S. J. Caughey.
1974. Deposition of Gaseous Sulphur Dioxide to the Ground.
Atmos. Environ. 8:75-79.
50
-------
Hedgpeth, H. , H. S. Judeikis, S. Siegel, and T. Stewart. 1974.
Cylindrical Flow Reactor for the Study of Heterogeneous Reactions
of Possible Importance in Polluted Atmospheres. Rev. Sci.
Instrum. 45:344-347.
Heines, T. S. , and L. K. Peters. 1974. The Effect of Ground Level
Absorption on the Dispersion of Pollutants in the Atmosphere.
Atmos. Environ. 8:1143-1153.
Hill, C. A. 1971. Vegetation: A Sink for Atmospheric Pollutants.
J. Air Pollut. Control Assoc. 21:341-346.
Jenson, V. G. , and G. V. Jeffreys. 1963. Mathematical Methods in
Chemical Engineering, Academic Press, New York.
Kellogg, W. W. , R. D. Cadle, E. R. Allen, A. L. Lazrus, and E. A.
Martell. 1972. The Sulfur Cycle. Science 175:587-596.
Lawrence, E. N. 1964. The Measurement of Atmospheric Sulphur
Compounds Using Lead Dioxide. Int. J. Air Water Pollut.
8:381-388.
Meidner, H. , andT.A. Mansfield. 1968. Physiology of Stomata.
McGraw-Hill, New York.
Owers, M. J. , and A. W. Powell. 1974. Deposition Velocity of Sulphur
Dioxide on Land and Water Surfaces Using a -"S Tracer Method.
Atmos. Environ. 8:63-67.
Paneth, F. , and K. Herzfeld. 1931. Uber Freies Methyl und Freies
Athyl. Z. Elektrochem. 37:577-582.
Payrissat, M. , and S. Beilke. 1975. Laboratory Measurements of the
Uptake of Sulphur Dioxide by Different European Soils. Atmos.
Environ. 9:211-217.
Present, R. D. 1958. Kinetic Theory of Gases. McGraw-Hill,
New York. .. .. ;
Robinson, E. , and R. C. Robbins. 1970. Global Effects of Environmental
Pollution. Singer, F. S. , ed. Springer-Verlag New York Inc.
pp. 50-64.
Seim, C. E. 1970. Sulfur Dioxide Absorption by Soil. Doctoral Thesis,
University of Minnesota.
Shepherd, J. G. 1974. Measurements of the Direct Deposition of Sulphur
Dioxide on Grass and Water by the Profile Method. Atmos. Environ.
8:69-74.
51
-------
Sideman, S. , D. Luss, and R. E. Peck. 1965. Heat Transfer in
Laminar Flow in Circular and Flat Conduits With (Constant)
Surface Resistance. Appl. Sci. Res. A14:157-171.
Singh, S. N. 1958. Heat Transfer by Laminar Flow in a Cylindrical
Tube. Appl. Sci. Res. A7:325-340.
Spedding, D. J. 1969a. Sulphur Dioxide Uptake by Limestone. Atmos.
Environ. 3:683.
Spedding, D. J. 1969b. Uptake of Sulphur Dioxide by Barley Leaves
at Low Sulphur Dioxide Concentrations. Nature 224:1229-1231.
Spedding, D. J. , and R. P. Rowlands. 1970a. Sorption of Sulphur
Dioxide by Indoor Surfaces. I. Wallpapers. J. Appl. Chem.
20:143-146.
Spedding, D. J. 1970b. Sorption of Sulphur Dioxide by Indoor Surfaces.
II. Wood. J. Appl. Chem. 20:226-228.
Spedding, D. J. , R. P. Rowlands, "and J. E. Taylor. 1971. Sorption
of Sulphur Dioxide by Indoor surfaces, III. Leather. J. Appl.
Chem. Biotechnol. 21:68-70.
Spedding, D, J. 1972a. The Sorption of Sulphur Dioxide by Indoor
Surfaces. IV. Flooring Materials. J. Appl. Chem. Biotechnol.
22:1-8.
Spedding, D. J. 1972b. Investigations into the Interactions of Gas Phase
Sulphur Dioxide With Solid and Liquid Surfaces. International
Clear Air Conference, Melbourne, Australia, Paper No. 4. 7.
p. 187-191.
Spedding, D. J. 1972c. Sulphur Dioxide Absorption by Sea Water.
Atmos. Environ. 6:583-586.
Stewart, T. B. , and H. S. Judeikis. 1974. Measurements of Spatial
Reactant and Product Concentrations in a Flow Reactor Using
Laser-Induced Fluorescence. Rev. Sci. Instrum. 45:1542-1545.
Walker, R. E. 1961. Chemical Reaction and Diffusion in a Catalytic
Tubular Reactor. Phys. Fluids 4:1211-1216.
Westenberg, A. A. , S. D. Raezer, and R. M. Fristrom. 1957.
Interpretation of the Sample Taken by a Probe in a Laminar
Concentration Gradient. Combust. Flame 1:467-478.
Whelpdale, D. M. , and R. W. Shaw. 1974. Sulphur Dioxide Removal
by Turbulent Transfer Over Grass, Snow, and Water Surfaces.
Tellus 26:196-205.
52
-------
APPENDIX C
DEPOSITION OF H2S AND
DIMETHYLSULFIDE ON
SELECTED SOIL MATERIALS
INTRODUCTION
A knowledge of deposition rates of trace gases onto ground-level sur-
faces is essential to calculating atmospheric budgets for these species as
well as their transport in the environment. The interest in sulfur-containing
species over the past two decades has prompted a number of measurements
of deposition of oxidized sulfur-containing species over a wide variety of
ground-level surfaces [e.g., Judeikis and Stewart (1976) and references
therein]. However, little is known of the deposition of reduced sulfur com-
pounds. Significant quantities of the latter species (~90-280 Tg S yr )
result from biogenic emissions in the environment, as inferred from analy-
ses of the global sulfur cycle [Friend (1973) and references therein].
Originally, it was thought that biogenic sulfur emissions were com-
posed primarily of H2S. Some recent studies indicate significant contribu-
tions of organic sulfides, such as dimethylsulfide (DMS), to the biogenic
sulfur emissions [Lovelock, Maggs, and Rasmussen (1972) and Rasmussen
(1974)]. On the basis of seawater measurements, laboratory experiments,
and qualitative conclusions regarding the fate of H2S in oxidizing fresh and
ocean surface waters, the latter authors suggest that organic sulfur com-
pounds may dominate biogenic sulfur emissions. Liss and Slater (1974),
however, estimate a biogenic flux of DMS from ocean surfaces of
3.7 Tg S yr"1. Similarly, Hitchcock (1975, 1976) concludes that DMS emis-
sions can contribute only ~ 2-5 Tg S yr to biogenic sulfur emissions . She
also cites the work of Chen and Morris (1972) on the aqueous oxidation of
sulfide to conclude that H2S is the dominant form of biogenic sulfur emis-
sions, in agreement with earlier suggestions. Both H2S and DMS have been
detected in field measurements in the United States. Natusch et al. (1972)
measured H?S concentrations of ~ 0.05 ppb in Colorado. Maroulis and Bandy
(1977) found similar levels of DMS on the Atlantic Coast, although lower
levels, typically < 0. 03 ppb, were found at a site near Norfolk, Virginia.
Here we report on the laboratory measurement of the velocities of H2S
and DMS deposition over selected soil samples. We find that the measured
deposition velocities for these species are lower, in some cases by almost
53
-------
two orders of magnitude, than those observed for SO2 deposition over the
same materials. These and other considerations discussed herein lead us to
conclude that dry deposition of t^S and DMS cannot be important processes
in the environment.
EXPERIMENTAL,
The apparatus and procedures used in these experiments have been de-
scribed elsewhere [judeikis and Stewart (1976)]. The apparatus consisted of
a cylindrical flow reactor in which the walls of an inner concentric cylinder
were coated with the material of interest. As a homogeneous gas-phase mix-
ture containing trace amounts of the reagent gas of interest flowed through
the cylinder, the trace species diffused to the walls where it was removed
by deposition on the solid surface. This resulted in concentration gradients
for the trace species along the cylinder axis (flow direction) as well as the
radial dimension. The axial concentration gradient was measured by means
of a system of small probes whose intakes were centered along the cylinder
axis. These probes were coupled, by means of a multiport rotary valve, to
a mass spectrometer. The model used in the analysis of data from these
experiments specifically accounted for mass transport by diffusion and flow.
For determination of the deposition velocity, use was made of the boundary
condition (at the inner cylinder wall)
-D |Ł = V c
9r g
that equates the diffusive flux of the trace species to the wall with its heter-
ogeneous removal rate, where D is the (molecular) diffusivity of the trace
species in the gas mixture, c and 9c/9r its concentration and radial concen-
tration gradient at the wall, and Vg the deposition velocity [judeikis and
Stewart (1976)]. Deposition velocities obtained in this manner are indepen-
dent of mass transport and reflect the rates of heterogeneous interactions at
the surface that are responsible for removal of the trace species. These,
then, represent the maximum deposition velocities that would be encountered
in the environment under turbulent atmospheric conditions.
In the studies described here, the reactor was slightly modified so
that the inner cylinder consisted of an uncoated portion on the inlet side of
the reactor, sufficiently long to permit full development of laminar flow,
with the remainder of the cylinder coated with the solid of interest. This
permitted use of fully developed laminar flow models for data analysis
[Judeikis and Stewart (1976)].
Experiments were carried out at room temperature (20-25 °C) and total
pressures of ~ 500 Torr, except in the case of SO2- Experiments with the
latter species were conducted at ~ 100 Torr, for the reasons delineated by
Judeikis and Stewart (1976). Flow rates were in the range of 1-10 cm^ see"*.
Reynolds numbers were < 50. Gas mixtures containing traces of H2S, DMS,
or SC>2 generally consisted of ~3-15-percent C>2, with the balance made up of
nitrogen and water vapor (0- or 95-percent relative humidity). Trace gas
concentrations used were ~ 15-150 ppm for DMS and SC>2 and ~15-1000 ppm
for H2S. The higher concentrations in the case of t^S were necessary when
54
-------
concentration, as well as relative humidity to within a factor of 2 (except
for DMS where relative humidity did have an effect — see below), indicating
that surface deposition occurred through first-order or pseudo-first-order
processes.
TABLE C-l. MEASURED DEPOSITION VELOCITIES21
Deposition velocity (cm sec"*)
Solid
Adobe clay soil
Sandy loam soil
Fe_O,
2 3
H2S
0.016
0.015
0.38
DMS
0.28
0.064
S02
0.92
0.60
3.9
Values are averages from at least three separate determinations on each of
three separate, freshly prepared samples for each material. Uncertainties
(resulting from sample-to-sample variations and reproducibilities within
samples) are ~±20 percent.
The data given in Table C-l are for freshly prepared samples. We
found that on prolonged exposure to these species, deposition rates gradually
decreased and ultimately approached zero. This type of saturation behavior
was previously observed for SO2 [judeikis and Stewart (1976)] and occurred
at ~0.4-2. 8 g SO? m of surface in humidified systems. The behavior is
probably the result of depletion of available surface sites for uptake of the
trace species because of adsorption of reactant or surface reaction products.
Possible mechanisms for regeneration of active surfaces in the envi-
ronment, such as interaction with atmospheric ammonia and washing away of
soluble surface reaction products by precipitation, were considered in the
latter reference and supported by selected experiments carried out in that
study. In the case of H^S, interactions with atmospheric ammonia could
also be important. However, surface reactionproducts of the latter species
such as sulfides [see below and Kanivets (1970)] are water insoluble, which
precludes washing away by rain water, except as a slurry of particles con-
taining adsorbed surface reaction products.
Ferric oxide was of interest here because of its common occurrence
and reported reactivity toward F^S in soils [Kanivets (1970)]. There was
also interest in subjecting Fe2O3 samples exposed to t^S to ESCA analysis.
Soil samples could not readily be used for such experiments because of a
number of broad background peaks that complicated the signal interpretation.
The results of the ESCA experiments are summarized in Table C-2.
Samples of reagent grade Fe2(SO4)3, FeS, Na2SO3, Fe2O3 (blank), and
Fe-O., exposed to H^S were analyzed for the Fe_ , S- , and C. peaks.
56
-------
high oxygen concentrations were used because of interference with the
mass spectral peak by the tail of the C>2 peak. All gases were reagent grade
quality. Solids used were representative sandy loam and adobe clay soils,
taken from the Los Angeles area, and reagent grade Fe2O3- Average thick-
nesses of the solid coatings were ~ 1 mm, with surface roughnesses of a few
tenths of a millimeter.
Diffusion coefficients for SC>2, for use in analysis of the experimental
data, were taken from the literature [Fish and Durham (1971)]. Diffusion
coefficients for H^S and DMS were estimated from hard sphere models for
binary diffusion [Present (1958)]. In the latter case, we estimate values
of 0.268 and 0. 180 cm2/sec, respectively, for H2S and DMS at 25° C and
500 Torr. Comparison of a large number of values calculated in this
fashion for other species, with experimentally measured values for those
species, leads us to believe that this method of estimating diffusion coeffi-
cients is accurate to better than 10-15 percent. In the experiments de-
scribed here, the deposition velocities are approximately inversely propor-
tional to the diffusion coefficient. Therefore, uncertainties on the order of
10-15 percent or less in the diffusion coefficients will result in uncertainties
of similar magnitude in the deposition velocities.
Background runs were made by passing the appropriate gas mixture
through the chamber containing an uncoated cylinder. These runs indicated
very small losses of the trace species, probably caused by wall reactions with
the Pyrex cylinder. For SC>2 and DMS, and for H2S removal over Fe2C>3,
these losses were negligible (< 1 percent) compared with those observed
when coated cylinders were used. For H2S removal over soil samples, these
losses were on the order of 20-30 percent of those observed when soil sam-
ples were present because of the low reactivity of the soils, as discussed in
the following section of this appendix. However, since the Pyrex cylinder is
coated during an actual experiment, no background corrections were made
to the observed data.
Samples for ESCA analysis were prepared by depositing slurries of the
appropriate solid in deionized water or reagent grade ethanol onto clean
glass slides and evaporating to dryness under vacuum (~10"* Torr). Samples
to be exposed to HgS were then inserted into the cylindrical flow reactor, ex-
posed to H2S, and subsequently transported in air to the sample chamber of
a GCA/McPherson ESCA36 photoelectron spectrometer. Control samples
and reference standards were used without H^S exposure. The ESCA was
equipped with a cyropump that allowed spectra to be taken at a pressure of
10~9 Torr. The x-ray source for the instrument was a magnesium anode,
which emitted electrons at an energy of 1253. 6 eV. All peaks were refe-
renced to the carbon Is peak at 284.6 eV.
RESULTS
Experimental results for t^S and DMS are given in Table C-l. In-
cluded for comparison are deposition velocities for SC*2 measured in this
study; the latter values are in good agreement with previous measurements
on similar materials [judeikis and Stewart (1976)]. The deposition velocities
reported in Table C-l were found to be independent of trace gas and oxygen
55
-------
The weak sulfur peaks observed in the ESCA analyses of Fe2C>3 ex-
posed to H2S were in sharp contrast to the strong sulfur (sulfate) peaks
observed for various metal oxides exposed to SOo. These results indicate
that much of the I^S deposited on the Fe2C>3 surface is weakly adsorbed and
is removed during the evacuation of samples to ~ 10~9 Torr prior to ESCA
analysis. This is not the case for 803. where irreversible conversion to
adsorbed sulfate appears to be much more extensive.
To test the possibility that deposition of J^S occurred largely through
physical adsorption, we exposed an Fe2C>3 sample to P^S until it was nearly
saturated, e.g., until the rate of removal of H2S from the gas stream had
decreased by over an order of magnitude. The sample was then evacuated
overnight at~ iO~^ Torr and reexposed to F^S. The result of this experi-
ment indicated that the evacuation restored the reactivity of Fe2C>3 toward
H2S removal to its initial value, within experimental error (~20 percent).
(This was not the case for SO2 exposed samples, where reactivity remained
low. ) From this, we estimate that the deposition velocity for irreversible
H->S removal over Fe2Oo is a factor of 5 or more lower than that given in
TibleC-1.
Similar observations were made for experiments in which the two soils
were exposed to DMS until saturated, evacuated overnight (~10~^ Torr), and
reexposed to DMS. This indicates that here, also, the deposition velocities
given in Table C-l for DMS are largely the result of reversible physical
adsorption and that irreversible removal processes occur at much slower
rates. Further, we found that DMS deposition from humidified reaction
mixtures occurred with velocities <0. 003 cm/sec. The latter results indi-
cate that H2O competes effectively with DMS for available adsorption sites
on the soil surfaces. Additional evidence for this was found when soil sam-
ples were exposed to dry DMS reaction mixtures until saturated and subse-
quently exposed to humidified mixtures. The latter exposure indicated that
adsorbed DMS was displaced from the surface by H2O, as evidenced by an
increase of DMS in the gas stream over the input concentration.
DISCUSSION
Results for the two types of soil yield very low deposition rates for
H2S and DMS on these materials and indicate that such materials would be
very poor sinks for these reduced sulfur species. [Similarly, kiss and
Slater (1974) have estimated a transfer rate of 0. 005 cm sec" for DMS
across the air -sea interface, although their estimate is for biogenic emis-
sion into the atmosphere.] For example, if an average global atmospheric
concentration of 0.2 ppb is assumed for reduced sulfur compounds [Friend
(1973) and references therein], deposition velocities of 0.015-0.28 cm sec"
over land areas would result in removal of only ~ 0.3-5.6 Tg S yr~ , com-
pared with the estimated emission strengths noted above. Although the de-
position velocity measured for DMS over adobe clay soil appears to be suffi-
ciently high to be of interest, as noted in the preceding section, this rate
appears to be largely the result of reversible adsorption. The deposition
velocity for irreversible removal of DMS over this material is considerably
lower, perhaps by an order of magnitude or more.
58
-------
TABLE C-2. MEASURED BINDING ENERGIESa
Binding energies (eV)
Sample Fe3p S2p
Fe_O, (exposed to H_S) 55.5 169.0
Li J L*
160.0
Fe2O3 (blank) 56. 0
FeS 55.5 168.4
161.2b
Fe2(S04)3 57-58° 169.1
Na2S03 167.6b
166.2
Adobe clay soil 57. 1
Sandy loam soil 56.6
All binding energies are referenced to C.
i is.
More intense peak.
Q
Weak, broad peak.
Both the Fe2<33 sample exposed to t^S and the FeS sample have a
binding energy (B.E. ) of 55. 5 eV for the Fe3p peak; this is 0. 5 eV less than .
the B.E. found for the Fe2C>3 alone. Such a result would not, by itself, in-
dicate that a reaction between the HoS and Fe2C>3 had taken place; however,
the results of the sulfur analyses are definitive. The unreacted Fe2C>3 shows
no sulfur peak, whereas the Fe2C>3 sample exposed to f^S shows two weak
peaks at B.E. s of 169. 0 and 160 eV. These peak positions are in good agree-
ment with our own measurements for the S2p peaks from Fe2(SC>4)3 and FeS
and with literature values [Craig, Marker, and Novakov (1974)J, which indi-
cates that sulfur is present in both the +6 and -2 oxidation states. Such re-
sults are consistent with the formation of FeSO4 on the sample surface (for
example, from the reduction of Fe+3 by t^S) as well as an indication of the
presence of strongly adsorbed sulfide.
It will be noted that the Fe3 peak from Fe2 (SO^H does not agree with
that found for FeoC^, although the sulfate S2p peak we have measured agrees
with literature values [Craig, Marker, and Novakov (1974)]. Although the
Fe3p peak from the ferric sulfate was weak and broad, its position at a B.E.
of 57-58 eV was reproducible. At present, we have no explanation for this
apparent anomaly.
57
-------
The results obtained for iron oxide are consistent with earlier findings
[Kanivets (1970) and references therein] that indicate the latter material is
one of the soil components most reactive toward H2S. Although we did not
quantitatively assay our soil samples for Fe2C>3, ESCA examination of these
materials did indicate the presence of small quantities of iron.
Minimal deposition of HoS or DMS on ground-level surfaces does, of
course, indicate the possibility of long-range transport of these species in
the atmosphere. In the case of I^S, however, laboratory measurements in-
dicate a rapid oxidation of this species by OH [Westenberg and de Haas (1973)
and Perry, Atkinson, and Pitts (1976)]. Assuming an OH concentration of
~3 x 10" molecules cm , the latter authors estimate a lifetime of ~0.5 day
for H2§ oxidation in the environment through this reaction.
Similarly, Cox and Sandalls (1974) conclude that the nitric-oxide sensi-
tized photooxidation of DMS would limit the atmospheric lifetime of the latter
compound to a few hours; however, as Cadle (1976) points out, the experi-
ments of Cox and Sandalls were conducted with nitrogen oxide concentrations
more representative of photochemical smog conditions than those found in
the ambient atmosphere.
59
-------
BIBLIOGRAPHY
Cadle, R. D. 1976. The Photo -Oxidation of Hydrogen Sulfide and Dimethyl
Sulfide in Air. Atmos. Environ. 10:417.
Chen, K. Y. , and J. C. Morris. 1972. Kinetics of Oxidation of Aqueous
Sulfide by O2. Environ. Sci. Technol. 6:529-537.
Cox, R. A., and F. J. Sandalls. 1974. The Photo-Oxidation of Hydrogen
Sulfide and Dimethyl Sulfide in Air. Atmos. Environ. 8:1269-1281.
Craig, N. L.., A. B. Harker, and T. Novakov. 1974. Determination of
the Chemical States of Sulfur in Ambient Pollution Aerosols by X-Ray
Photoelectron Spectres copy. Atmos. Environ. 8:15-21.
Fish, B. R., and J. L. Durham. 1971. Diffusion Coefficient of SO2 in Air.
Environ. Lett. 2:13-21.
Friend, J. P. 1973. The Global Sulfur Cycle. Chemistry of the Lower
Atmosphere, S. I. Rasool, ed. Plenum Press, New York. pp. 177-201.
Hitchcock, D. R. 1975. Dimethyl Sulfide Emissions to the Global Atmo-
sphere. Chemosphere 4:137-138.
Hitchcock, D. R. 1976. Microbiological Contributions to the Atmospheric
Load of Particulate Sulfate. Environmental Biogeochemistry, Vol. I,
J. O. Nriagu, ed. Ann Arbor Science Publishers, Inc., Ann Arbor,
Michigan, pp. 351-367.
Judeikis, H. S., and T. B. Stewart. 1976. Laboratory Measurement of
SO2 Deposition Velocities on Selected Building Materials and Soils.
Atmos. Environ. 10:769-776.
Kanivets, V. I. 1970. Reaction of Hydrogen, Methane, and Hydrogen
Sulfide With the Mineral Part of the Soil. Sov. Soil Sci. 294-301.
Liss, P. S., and P. G. Slater. 1974. Flux of Gases Across the Air-Sea
Interface. Nature 247:181-184.
Lovelock, J. E., R. J. Maggs, and R. A. Rasmussen. 1972. Atmospheric
Dimethyl Sulphide and the Natural Sulphur Cycle. Nature 237:452-453.
60
-------
Maroulis, P. J., and A. R. Bandy. 1977. Estimate of the Contribution of
Biologically Produced Dimethyl Sulfide to the Global Sulfur Cycle.
Science 196:647-648.
Natusch, D. F. S., H. B. Klonis, H. D. Axelrod, R. J. Teck, and
J. P. Lodge, Jr. 1972. Sensitive Method for Measurement of
Atmospheric Hydrogen Sulfide. Anal. Chem. 44:2067-2070.
Perry, R. A., R. Atkinson, and J. N. Pitts, Jr. 1976. Rate Constants
for the Reactions OH + H2S -»H2O + SH and OH + NH3 -» H2O + NHo
Over the Temperature Range 297-427 °K. J. Chem. Phys. 64:3237-
3239.
Present, R. D. 1958. Kinetic Theory of Gases, McGraw-Hill, New York.
Rasmussen, R. A. 1974. Emission of Biogenic Hydrogen Sulfide. Tellus
26:254-260.
Westenberg, A. A., and N. de Haas. 1973. Rate of the Reaction OH + H2S
-• SH + H2O Over an Extended Temperature Range. J. Chem. Phys.
59:6685-6686.
61
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/3-77-132
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
THE ROLE OF SOLID-GAS INTERACTIONS IN AIR POLLUTION
5. REPORT DATE
December 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Henry S. Judeikis, Thomas B. Stewart,
Anthony G. Wren, and James E. Foster
8. PERFORMING ORGANIZATION REPORT NO
^PERFORMING ORGANIZATION NAME AND ADDRESS
The Aerospace Corporation .
2350 E. El Segundo Blvd.
El Segundo, California 90245
10. PROGRAM ELEMENT NO.
1AA603 AH-04(FY-77)
11. CONTRACT/GRANT NO.
R-802687
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory - HTP, 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 11/73 - 2/77
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Sulfur dioxide and other sulfur-containing gases have been studied to
evaluate their interactions with solids likely to be found in urban aerosols
and on ground-level surfaces in the urban environment. Results indicate
that sulfur dioxide readily reacts with most of these materials by capacity-
limited reactions, particularly at high relative humidities. Removal of
hydrogen sulfide and dimethylsulfide over ground-level surfaces is a slow
process and largely reversible. The implications of these results with
regard to air pollution chemistry and sulfur control strategies are discussed,
Publications, reports, and presentations that resulted from this work are
listed.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b-IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
*Air pollution
*Interactions
*Sulfur dioxide
*Hydrogen sulfide
*Aerosols
Surfaces
Deposition
13B
07B
07D
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
70
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
62
------- |