EPA-600/2-76-092
April 1976
Environmental Protection Technology Series
TOTAL SUSPENDED PARTICULATES:
Review and Analysis
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research pertormed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards
EPA RE VIEW NOTICE
This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield. Virginia 22161.
-------
EPA-600/2-76-092
April 1976
TOTAL SUSPENDED PARTICULATES
REVIEW AND ANALYSIS
by
R. Murray Wells
Radian Corporation
8500 Shoal Creek Boulevard
Austin, Texas 78766
Contract No. 68-02-1319, Task 27
ROAP No. 21ADK-002
Program Element No. 1AB012
EPA Task Officer: J.A. McSorley
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1
1.1 Background 1
1. 2 Obj ectives 2
1.3 Contents of Report 3
2 . 0 SUMMARY 5
2.1 Hidy and Mueller 5
2.2 Lodge 6
2.3 Brock 7
2.4 Babcock : 8
2.5 Gordon Research Conference 8
3.0 CONCLUSIONS AND RECOMMENDATIONS 11
4.0 BIBLIOGRAPHY 15
APPENDIX A
"Control Technology and Aerosols", G. M. Hidy
and P. K. Mueller, May, 1975 32
APPENDIX B
"Particulate Matter In The Atmosphere",
J. P. Lodge, May, 1975 . . i 114
APPENDIX C
"Review Of Suspended Particulate Matter",
J. R. Brock, June, 1975 130
APPENDIX D
"Particulate Matter: Relationships Between
Emissions And Ambient Air Quality",
L. R. Babcock, Jr., August, 1975 196
-------
i-Q INTRODUCTION
This report summarizes work done by Radian Corporation
under EPA Contract No. 68-02-1319, Task 27. The study involved
a review and analysis of the readily available information on
total suspended particulates in the atmosphere. The purpose of
this review was to determine the relative contribution of primary
and secondary particulate matter to the total aerosol mass sus-
pended in the atmosphere and to identify where the available
information is insufficient to determine the needs for future
control technology development. In addition, if possible, the
fraction of total suspended particulates attributable to mobile
and to stationary sources was to be identified. The work was
performed from March through August 1975,
1•1 Background
The continental tropospheric aerosol consists of primary
and secondary particulate matter which is emitted or is the result
of emission from numerous natural and antropogenic sources.
Primary particulate matter is any particulate matter that is
emitted directly from a source and remains relatively unchanged
chemically in the atmosphere, Secondary particulate matter is
formed in the atmosphere from gaseous precursors. These gaseous
precursors are generally "pollutant gases" (S02, N0x, HC) emitted
from many of the same sources as the primary particulate matter and
also the gaseous constituents (NH3) from natural sources, i.e.,
vegetation.
The aerosol mass suspended in the atmosphere varies in
concentration, composition, and size distribution both as a function
of time and location. This aerosol is normally divided into two
mass fractions as a function of particle size. The "fine" fraction
is usually defined as being the mass of particulates that have an
aerodynamics diameter between 0.01 and 1.0 micron. The large
particulate fraction roughly encompasses those particulates with an
aerodynamic diameter of 1.0 to 100 microns. Existing ambient air
-1-
-------
quality standards do not distinguish between these two fractions
although it is the "fine" fraction which is known to have the
greatest impact on visibility in a region and is thought to cause
adverse health effects. The visibility reduction is due to the
scattering of light by particles in the subtnicron range while
the potentially adverse health effects are due to the respirable
nature of particles in this size range.
As much as three-fourths of the population of the
United States is living in areas in which the levels of suspen-
ded particulate matters exceed the ambient air quality standards.
These areas include both urban and rural environments. Current
control strategies are directed at reduction of antropogenic
primary source emissions through the installation of mechanical
collectors, wet scrubbers, fabric filters, and electrostatic
precipitators. The most efficient of these devices remove nearly
all of the large primary particulates but still allow fines to be
emitted to the atmosphere. With the exception of wet scrubbers,
these control devices are incapable of achieving any control of
secondary particulates.
Evidence is now accumulating which indicates that
significant amounts of fine particulate matter may be secondary
in nature. In some locations this secondary particulate matter
may be the dominant fine particulate matter.
1.2 Objectives
The Industrial and Environmental Research Laboratory
(IERL, formerly the Control Systems Laboratory) of EPA is re-
ponsible for the development of technology for the control of
pollutant emissions from stationary sources. Major research
areas funded by IERL involve the development of improved control
technology for sulfur dioxide, nitrogen oxides, hydrocarbon, and
particulate emissions. In the area of particulate emission
-2-
-------
control, emphasis is being placed by IERL on the control of
emissions of fine primary particulates. The direction of future
particulate control technology development could be greatly
impacted by evidence being accumulated for various regions by
EPA and others, which indicates that the majority of the fine
particulate matter suspended in the atmosphere is secondary. It
is vital to the particulate control technology development program
to determine if this evidence is accurate and applicable on a
national scale. The objective of this short term study was to
determine what data were available that could be used to assess
the relative contributions of primary and secondary particulate
matter to the atmospheric aerosol and to assess whether or not
sufficient data existed to define future control technology needs
for stationary sources.
In order to accomplish the objective stated above,
Radian reviewed a small body of literature made available by
EPA and contacted several experts in the field of atmospheric
science to determine the current state of the art of their
knowledge in the areas of interest.
1.3 Contents of the Report
The consultants used by Radian in the course of this
study included: (1) G. M. Hidy and P. K. Mueller, Environmental
Research and Technology, Inc., Westlake Village, California;
(2) J. P. Lodge, Consultant in Atmospheric Chemistry, Boulder,
Colorado; (3) J. R. Brock, University of Texas, Austin, Texas.
These experts in the fields of aerosol characterization and
atmospheric chemistry independently prepared reports
incorporating their own knowledge and experience into inter-
pretations of the data provided by EPA. A fourth consultant -
L. R. Babcock, Jr., School of Public Health, University of
Illinois, Chicago, Illinois - took these three reports plus a
recent paper by Gartrell and Friedlander and interpreted each
expert's findings in light of the others.
-3-
-------
Radian utilized the output of each of the consultants,
information gathered at a recent Gordon Research Conference and
Radian in-house expertise to prepare this report. A brief summary
of the results of the overall program is given in Section 2.0.
The conclusions and recommendations are discussed in Section 3.0.
Section 4.0 is the bibliography and contains the references
provided by EPA as well as those used by each consultant. Section
5.0, the Appendix, contains the four reports generated by the
consultants mentioned above. These reports constitute the body
of this report. The interested reader should study these reports
in their entirety.
-4-
-------
2.0 SUMMARY
The results and discussion presented in this report are
based upon the four input reports (included in the Appendix)
discussed in the introduction and information obtained by Radian
at a recent Gordon Research Conference. The significant results
of each report and the information presented at the Gordon Research
Conference are summarized below.
Each of the consultants reports prepared for this study
approach the problem in a unique but meaningful way. Each report
should be read in its entirety in order to gain each author's
insight and perspective regarding atmospheric aerosol characteri-
zation and control.
2.1 Hidy and Mueller
This report emphasizes the California studies with which
the authors are most familiar. Data on both chemical composition
and particle size of atmospheric particulates is presented. The
particle size data presented confirms the existence in the Los
Angeles area aerosol of a bimodal mass distribution with respect
to particle size. The chemical composition data presented repre-
sents an attempt to trace the source of atmospheric particulates.
The chemical tracer method of S. K. Friedlander and
others is proposed by the authors as being the best available
technique for tracing the sources of atmospheric aerosols. The
results of Gartrell and Friedlander (later discussed in detail by
Babcock) were discussed. Source species data for Pasadena and
Pomona are presented and natural and antroprogenic sources of
primary particulates are identified.
-5-
-------
Material resulting from the oxidation of S02 and NO
and condensation of non-methane hydrocarbon vapor was shown to
contribute a substantial fraction of the total particulate mass
concentration occurring in urban and non-urban air. While the
authors indicated that this secondary aerosol could represent as
much as 40-50% of the total aerosol in these areas, no attempt was
made to assign sources for the secondary particulates. Hidy did
append an earlier work, for which he was coauthor, which examined
possible mechanisms for secondary particulate formation in the
atmosphere.
2.2
While Hidy and Mueller confined themselves mainly to
the California experience, Lodge provided a wider viewpoint by
discussing data obtained from urban Southern California, rural
and urban Colorado, Chicago, and Milan, Italy. Based on the
results of studies performed in each of these areas, Lodge agreed
with Hidy and Mueller that secondary particulates could comprise
as much as 40 percent of the total aerosol. However, Lodge
warned against confusing the terms primary and secondary partic-
ulates with the terms large and fine size fraction, Lodge identi-
fied the material and antropogenic sources of both primary and
secondary particulates and pointed out that primary particulates
could be both large and fine and that while most secondary
particles are fine the existence of large secondary particulates
was possible. Lodge described conditions in Denver and rural
Colorado during which nearly 90% of the atmospheric aerosol was
observed to be larger than one micron and almost no secondary
particulates could be identified. These results, obtained during
the winter are completely counter to results obtained for Denver
during the summer when the Denver aerosol closely resembles that
found in Southern California.
-6-
-------
Perhaps Lodge's most important contribution to this
study was the important differentiation he made between con-
temporaneous carbon (that derived from recent biota) and carbon
derived from fossil fuels. He concluded that "an important portion
of the organic matter (in the ambient aerosol) consisted of spores,
microorganisms, pollens and contaminated plant material". He did
not assign a size fraction to this contemporaneous carbon. It
should be noted however, if "contemporaneous carbon" were a
generally present pollutant it would most likely all be in the
large particle mode and would not be very soluble. As such, it
would be observed in total C, H, and N analysis but not as a
soluble organic. According to William Wilson of EPA most recent
work has determined that organic material present in total
particulate samples is soluble in various solvents.
2.3 Brock
This report provides a comprehensive review of both
primary and secondary particulate matter, including both theore-
tical and empirical information. Brock basically supports the
conclusions of Lodge as well as those of Hidy and Mueller. He
states that on a global basis as much as 40% of the suspended
particulate matter is secondary in nature (mostly sulfates,
nitrates, and condensed hydrocarbons). -Like the other consultants
Brock makes an attempt to assess the sources of primary particu-
lates. Brock's major contribution to this study was to provide
estimates for 1970, 1980, 1990, and 2000 of the mass emission,
particle size, and residence time in the atmosphere for emissions
from 23 different source categories. Brock uses these data to
show how particle size affects the relative contribution to
aerosol mass. That is, that large sources of large particulates,
such as crushed stone, contribute very little to aerosol mass,
and that control of such sources will have very little effect on
ambient aerosol mass.
-7-
-------
2.4 Babcock
Babcock took the output of the three previously dis-
cussed reports plus a recent paper by Gartrell and Friedlander
and attempted to determine source-characterization relationships
for atmospheric aerosols. Two detailed source species size
characterizations were performed, one for a composite of five
California cities and a second that was chosen to be representa-
tive of nationwide emissions. Secondary particulates comprised
a significant fraction in both characterizations.
Babcock's chief contribution to this program was to
attempt, by extending the work of Gartrell and Friedlander and
Brock, to determine what fraction of the fine particulates were
primary particulates. Babcock had to make a large number of
assumptions to accomplish this goal, however the results are
most interesting as they indicate that as much as 16% of the
total atmospheric aerosol may consist of fine primary particulates
from stationary sources. Babcock also indicated that 26-49 per-
cent of the total suspended particulates in the atmosphere were
from stationary combustion and other industrial sources.
2.5 Gordon Research Conference
An Environmental Sciences: Air conference was held
August 18-22, 1975 at the Gordon Research Center, New Hampton,
New Hampshire. The topic of this conference was fine particulate
matter (0.01-l.Oy) in the troposphere. The majority of the
contributors to this study as well as other experts in this field
and other interested parties attended. The areas of interest to
this study — mainly the size, mass, and composition of fine
particles in the troposphere - were discussed at length at this
conference.
-8-
-------
No data counter to that presented by the four con-
sultants utilized in this study was presented. However, indica-
tions were given that sampling and analytical procedures used to
date to gather data on the size distribution and chemical compo-
sition of atmospheric aerosol may be biased. Much of the sulfate
and nitrate data gathered to date from hi-vol filter samples may
be erroneously high. This would mean that the contribution of
secondary particulates to the total aerosol mass may be much less
than previously suspected. This problem is mainly one of
conversion of gases in the atmosphere to particulates on the
filter due to the chemical and physical properties of the filter
media used and the meteorological conditions (particularly humidity)
at the time the sample was taken.
While no one at the conference seriously questioned
the multimodel nature of the mass distribution of suspended
particulates with respect to size, there was no agreement on a
good method of determining particle size distribution. Each
method has its problems. When it is desired to separate size frac-
tions for chemical analysis, the problem becomes compounded
because of such diverse factors as particle bounce and condensa-
tion. It was pointed out that large particles have been observed
to bounce from one stage of an inertial impactor to another thus
giving distributions unfairly biased toward the fine fraction.
Condensation and or crystallization of new particulates from
gaseous material passing through particle sizing devices was
demonstrated to give significantly erroneous chemical compositions.
-9-
-------
Elemental analysis tracer studies were also discussed.
These studies utilized techniques that were much the same as
those discussed by Hidy and Mueller. The drawbacks associated
with this type approach were discussed and the need for chemical
compound identification in place of elemental identification was
demonstrated. Without such techniques it is difficult to
distinguish between particulate matter such as coal ash and soil.
Two of the most important things noted by this author
at the conference were the extent to which the whole body of
knowledge in this area is built upon the California experience
and the in-breeding among the experts in the field. There seems
to be a definite need for additional data on other regions of the
country and for fresh ideas and inputs into studies in this area.
-10-
-------
3.0 CONCLUSIONS AND RECOMMENDATIONS
The California data currently available indicates that
material resulting from the oxidation of S02 and N0x and condensa-
tion of non-methane hydrocarbon vapor appears to contribute a
substantial fraction of the total particulate mass concentration
in both urban and non-urban air. A few cases have been observed
in Southern California where more than half the sampled particu-
lates were secondary. These secondary particles were heavily
concentrated in the submicron particle range.
For Denver and rural Colorado, during a significant
fraction of the year, nearly 90% of the atmospheric aerosol was
primary and concentrated in the large fraction (greater than 1m).
This was true even though the total mass concentration of suspended
particulates was approximately equal to that observed in Southern
California.
Source characterization studies performed or cited by
the various consultants do a credible job of identifying the
sources' of primary particulates larger than one micron. From
these studies it is evident that natural or quasi-natural sources
are significant contributors to the primary aerosol mass (primarily
fugitive dust, sea salt, and pollen).
The identification of the sources of secondary particu-
lates was not attempted by most of the experts involved in the
study. Current knowledge of the mechanisms and rates of conver-
sion of gases and vapors to aerosols was believed by the majority
of the consultants to be inadequate at the present time. The
one consultant, L. R. Babcock, who attempted to trace sources of
secondary particulates through a series of assumptions, found
that stationary combustion and industrial sources were the major
-11-
-------
anthropogenic sources of atmospheric aerosol. While mobile
sources were a significant problem in Babcock's source-species-
size characterizations they were not the dominant category even
for California. Babcock felt that the ranges presented in his
study might bound the problem, thus assessing the contribution
of fine primary particulate emissions at somewhere between 4 and
16 percent of the total suspended aerosol mass on a national basis
Better definition of real problem areas such as the
concentration of particulates in each size range and better
analytical data on the compositions of particles in those size
ranges is definitely required before the true nature of the
atmospheric aerosol problem can be understood. It is clear that
there is no single generalized source-species-size characteriza-
tion which is applicable throughout the nation. The various
distributions may all tend toward bimodal, but the sources and
species appear to vary significantly from location to location,
depending on local sources. Adding further complexity, the rela-
tionship between sources and ambient air quality seems to vary
significantly from season to season and even hour to hour. The
incomplete distributions published to date are probably not
representative of their own regions, much less of the nation as
a whole.
The fact that significant amounts of aerosol may be
derived ultimately from pollutant gases places potentially impor-
tant and new constraints on control strategy for aerosols. It is
now recognized that control of fine particle emissions in the 0.1
to 1 ym diameter range at the source is crucial for achieving im-
proved air quality for public health. Even though these particles
may include only a part of the total mass concentration, they are
important for development of respiratory ailments and remain in
the atmosphere the longest. The fine particle control problem
is compounded by the addition of condensable material formed
-12-
-------
following atmospheric chemical reactions. On the basis of this
study, it would appear that the research programs sponsored by
IERL which are being directed toward improved control of sulfur
oxide, nitrogen oxides, and hydrocarbon emissions, may play a
large role in helping to control the levels of submicron suspended
particulates in the atmosphere. In addition, it would appear
that a continued program to find methods of improving control of
fine primary particulates from stationary sources is justified.
As a result of the work performed by Radian and each of
the expert consultants employed in this study, it is recommended
that:
1) The chemical element balance method for
identifying the origins of aerosols be
applied to several cases in diverse areas
of the United States to confirm the
importance of secondary processes in
aerosol production and to separate man-
made from natural secondary aerosol sources.
2) Better means of chemical analysis be developed
so that chemical compounds rather than elements
can.be used to identify major sources.
3) Research into secondary aerosol formation to
determine the mechanisms and rates of conver-
sion of gases and vapors to aerosol should
receive high priority. This information,
which is currently inadequate, will be
required to determine the degree of control
of pollutant gas emissions that will be
required for secondary particulate control.
-13-
-------
4) Existing data on nitrate and sulfate
particulates should be reassessed in
light of recent work done by EPA and the
Electric Power Research Institute on the
formation of artifact sulfates and nitrates
during filtering of atmospheric aerosol.
5) The present programs underway at IERL to
improve control of sulfur oxides, nitrogen
oxides, and hydrocarbons should be continued
as these pollutants can play a large role in
the formation of fine secondary particulates
in the atmosphere.
-14-
-------
4-0 BIBLIOGRAPHY
References in common body of pertinent literature:
1. D. L. Blumenthal, J. A. Anderson and G. J. Sem,
. "Characterisation of Denver's Urban Plume Using an
Instrumented Aircraft," Paper 74-266, Air Pollution
Control Assn., Denver (June 1974).
2. C. Brosset and A. Akerstrom, "Long Distance Transport
of Air Pollutants - Measurements of Black Air-Borne
Particulate Matter (Soot) and Particle-Borne Sulphur
in Sweden During the Period of September-December 1969,"
6:661-673 (1972).
3. R. J. Charlson and A. P. Waggoner, "Visibility, Aerosol,
and Colored Haze," Paper 74-261, Air Pollution Control
Assn, Denver (June 1974).
4. M. T. Dana and others, "Natural Precipitation Washout
of Sulfur Compounds from.Plumes," (EPA-R3-73-047),
prepared by Battelle Memorial Institute, Richland,
Washington (June 1973).
5. M. T. Dana and others, "Precipitation Scavenging of
Inorganic Pollutants from Metropolitan Sources,"
(EPA-650/3-74-005), prepared by Battelle Memorial
Institute, Richland, Washington (June 1974).
6. R. G. Draftz, "Analysis of Philadelphia Suspended
Dusts Sampled at Street Level," IITRI-C9914 (date
unknown).
-15-
-------
7. R. G. Draftz, "Analysis of 25 Ambient Dust Samples from
Philadelphia," IITRI-C9915-1 (July 20, 1973).
8. R. G. Draftz and J. Durham, "Identification and Sources
of Denver Aerosol," Paper 74-263, Air Pollution Control
Assn., Denver (June 1974).
9. J. L. Durham and others, "Comparison of Volume and Mass
Distributions for Denver Aerosols," American Chemical
Society presentation, Los Angeles (April 1974).
10. S. K. Friedlander, "Chemical Element Balances and
Identification of Air Pollution Sources," Environmental
Science and Technology. 7_: 3 235-240 (March 1973).
11. S. K. Friedlander, "Small Particles in Air Pose a Big
Control Problem," Environmental Science and Technology,
2,: 13, 1115-1118 (December 1973).
12. D. F. Gatz, "St. Louis Air Pollution: Estimates of
Aerosol Source Coefficients and Elemental Emission
Rates," published by American Meteorological Society
(1974).
13. D. Grosjean and S. K. Friedlander, "Gas-Particle
Distribution Factors for Organic Pollutants in the
Los Angeles Atmosphere," Paper 74-154, Air Pollution
Control Assn., Denver (June 1974).
14. P. R. Harrison, R. Draftz, and W. H. Murphy, "Identi-
fication and Impact of Chicago's Ambient Suspended
Dust," (source and date unknown).
-16-
-------
15. J. M. Hales, J. M. Thorp, and M. A. Wolf, "Field
Investigation of Sulfur Dioxide Washout from the Plume
of a Large. Coal-Fired Power Plant by Natural Precipi-
tation," Prepared for EPA by Battelle Memorial Institute,
Richland, Washington (March 1971).
16. S. L. Heisler, S. K. Friedlander, and R. B. Husar,
"The Relationship of Smog Aerosol Size and Chemical
Element Distributions to Source Characteristics,"
Atmospheric Environment, 7,:633-649 (1973).
17. G. M. Hidy and S. K. Friedlander, "The Nature of the
Los Angeles Aerosol," in H. M. Englund and W. T. Beery
(ed) Proceedings of _ the Second International Clean Air
Congress, Academic Press, New York (1971).
18. P. W. Jones, "Analysis of Non-Particulate Organic
Compounds in Ambient Atmospheres," Paper 74-265, Air
Pollution Control Assn., Denver (June 1974).
19. R. E. Lee, "The Size of Suspended Particulate Matter
in Air," Science, 178: 4061 567-575 (November 1972).
20. D. F. Miller and others, "Haze Formation: Its Nature
and Origin," (EPA-650/3-74-002), prepared by Battelle
Memorial Institute, .Columbus (June 1973).
21. M. S. Miller, S. K. Friedlander, and G. M. Hidy, "A
Chemical Element Balance for the Pasadena Aerosol,"
J. Colloid and Interface Science, 39: 1 165-176 (April
1972).
-17-
-------
22. P. K. Mueller, R. W. Mosley, and L. B. Pierce, "Chemical
Composition of Pasadena Aerosol by Particle Size and
Time of Day: IV. Carbonate and Noncarbonate Carbon
Content," J Colloid and Interface Science 39:1, 235-239
(April 1972).
23. T. Novakov, and others, "Chemical Composition of Pasadena
Aerosol by Particle Size and Time of Day: III. Chemical
States of Nitrogen and Sulfur by Photoelectron Spec-
troscopy," J Colloid and Interface Science 39:1 225-
234 (April 1972).
24. J. W. Roberts, A. T. Rossano, H. A. Watters, "Dirty
Roads Equal Dirty Air," APWA Reporter, 10-12 (November
1973).
25. H. Rodhe, C. Persson, and 0. Akesson, "An Investigation
into Regional Transport of Soot and Sulfate Aerosols,"
Atmospheric Environment, 6:675-693 (1972).
26. D. Schuetzle, A. L. Crittenden, and R. J. Charlson,
"Application of Computer Controlled High Resolution
Mass Spectrometry to Analysis of Air Pollutants,"
J Air Pollution Control Assn.. 23_:8 704-709 (August 1973).
27. W. Schwartz, "Characterization of Model Aerosols,"
(EPA-650/3-74-011), prepared by Battelle Memorial
Institute, Columbus (August 1974).
28. G. A. Sehmel, "Particle Resuspension from an Asphalt
Road Caused by Car and Truck Traffic," Atmospheric
Environment 7:291-309 (1973).
-18-
-------
29. J. W. Winchester and G. D. Nifong, "Water Pollution
in Lake Michigan by Trace Elements from Pollution
Aerosol Fallout," Water. Air, and Soil Pollution, 1=50-
64 (1971).
References cited by the input consultants:
On the following lists, "*" indicates the reference
was included in the common body of pertinent literature.
References cited by Hidy and Mueller:
1. Akselsson, K. R., J. W. Nelson, and J. W. Winchester,
1975: "Proton Scattering for Analyses of Atmospheric
Particulate Matter", Bull. Am. Phys. Soc. II, 2_0, p. 155
2. Barone, J. B., T. A. Cahill, R. G. Flocehini, D. J.
Shedoan, 1975: "Visibility Reduction: A Characteriza-
tion of Three Urban Sites in California", Science, in
manuscript, Feb. 12.
*3. Draftz, R. G. and J. Durham, 1975: "Identification and
Sources of Denver Aerosol". Unpubl. report to U. S.
Environmental Protection Agency; also Harrison, P. W.
et__al, "Identification and Impact of Chicago's Ambient
Suspended Dust". Unpubl. report for U. S. Environ.
Protection Agency.
*4. Durham, J. L., W. E. Wilson, T. C. Ellestod, K. Willeke
and K. T. Whitby, 1975: "Comparison of Volume-and Mass
Distribution of Denver Aerosols", Atmos. Environment,
in press.
-19-
-------
5. Durham, J. L., R. K. Patterson, J. J. VanEe and W. E.
Wilson, 1975: "The Chemical Composition of the Denver
Aerosol", Atmos. Environment, in press.
6. Frank, E. and J. P. Lodge, Jr., 1967: "Morphological
Identification of Airborne Particles with the Electron
Microscope". J. Microscopic, 6_, 449-456.
*7. Friedlander, S. K., 1973: "Chemical Element Balances
and Identification of Pollution Sources". Environ.
Sci. and Technol. , 1_, 235-240.
8. Gatz, D. F., 1975: "Relative Contributions of Different
Sources of Urban Aerosols: Application of a New Estima-
tion Method to Multiple Sites in Chicago". Atmos.
Environ. I, 1-18.
9. Gartrell, G., Jr., and S. K. Friedlander, 1975:
"Relating Particulate Pollution to Sources: the 1972
Calif. Aerosol Charact. Study," Atmos. Environ., £,
279-300.
10. Harrison, P. R., and J. W. Winchester, 1971: "Area-
wide Distributions of Lead, Copper and Cadmium in
Air Pollutants from Chicago and Northwest Indiana".
Atmos. Environ. 5, 863-880.
*11. Heisler, S. et al, 1973: "The Relationship of Smog
Aerosol Size and Chemical Element Distributions to
Source Characteristics". Atmos. Environment 7,
633-649.
12. Hidy, G. M., 1973: "Removal Processes of Gaseous and
Particulate Pollutants" in Chemistry of the Lower Atmo-
sphere, (S.I. Rasool, ed.), Plenum Press, N. Y., Chap. 3.
-20-
-------
13. Hidy, G. M. et al, 1974: "Characterization of Aerosols
in California (ACHEX)". Final Report Volumes 1-4;
Rockwell Science Center, Report #SC524.25FR, Thousand
Oaks, CA 91360.
14. Hidy, G. M. and J. R. Brock, 1971: "An Assessment of
the Global Sources of Tropospheric Aerosols" in Proc^
2nd IUAPPA Clean Air Congr. (H. W. Englund and W. T.
Berry, ed.), Academic Press, N. Y., p. 1088.
15. Hidy, G. M. and C. S. Burton, 1975: "Atmospheric
Aerosol Formation by Chemical Reactions" to be publ.
in Int'l. J. of Chem. Kinetics.
*16. Hidy, G. M. and S. K. Friedlander, 1971: "The Nature
of the Los Angeles Aerosol". Proc. 2nd IUAPPA Clean
Air Congr. (H. M. Englund and W. T. Berry, ed.),
Academic Press, N. Y., p. 391.
17. Hidy, G. M. et al, 1974: "Observations of Aerosols
over Southern California Coastal Waters", J. of
Applied Meteorology, Vol. 13, No. 1, pp. 96-107.
*18. Miller, M. S. et al, 1972: "A Chemical Element Balance
for the Pasadena Aerosol" in Aerosols and Atm. Chem.
(G. M. Hidy, ed.), Academic Press, N. Y., p. 301.
19. Trijonis, J., 1974: "A Particulate Implementation Plan
for the Los Angeles Region". TRW Report for EPA.
20. Whitby, K. T., R. B. Husar and B. Y. H. Liu, 1972:
"The Aerosol Size Distribution of Los Angeles Smog"
in Aerosols and Atmos. Chem. (G. M. Hidy, ed.),
Academic Press, N. Y., p. 237.
-21-
-------
21. Willeke, K., K. T. Whitby, W. E. Clark, V. A. Marple,
1974: "Size Distribution of Denver Aerosols - A
Comparison of Two Sites", Atmos. Environment, (3, pp.
609-633.
References cited by Lodge:
1. Colorado Air Pollution Control Program. Report to the
Public, 34-37 (1972).
2. Colorado Air Pollution Control Program. Report to the
Public, 56-59 (1974).
3. Dams, R. , J. Billiet, C. Block, M. Demuynck, and M.
Janssens. Atmospheric Environment, in press (1975).
*4. Friedlander, S. K. Environ. Science and Technol. 7,
235-240 (1973).
5. Goetz, A. and R. F. Pueschel. J. Air Pollution Control
Assoc. 15. 90-95 (1965).
6. Hagen, L. J., and N. P. Woodruff. Atmospheric Environ-
ment 7, 323-332 (1973).
*7. Harrison, P. R., R. Draftz, and W. H. Murphy. Manu-
script, source unknown.
8. Lodge, J. P., Jr., G. S. Bien and H. E. Suess. Int.
J. Air Pollution 2, 309-312 (1960).
9. Sverdrup, G. M., K. T. Whitby and W. E. Clark. Atmo-
spheric Environment 9, 483-494 (1975).
-22-
-------
10. Whitby, K. T., W. E. Clark, V. A. Marple, G. M. Sverdrup,
G. J. Sem, K. Willeke, B. Y. H. Liu, and D. Y. H. Pui.
Atmospheric Environment 9, 463-482 (1975).
11. Willeke, K., K. T. Whitby, W. E. Clark and V. A. Marple.
Atmospheric Environment 8, 609-633 (1974).
References cited by Brock:
Environmental Protection Agency: National Primary
and Secondary Ambient Air Quality Standards, Federal
Register, 36, 8186(1971).
Anderson, D. 0., "The Effects of Air Contamination on
Health" Canad. Med. Assoc. J 97 528, 585, 802 (1967).
Amdur, M. 0. "Toxicoloical Appraisal of Particulate
Matter, Oxides of Sulfur and Sulfuric Acid". Paper 69-
68, Proceedings Air Pollution Control Association,
New York, New York, June 22-26, 1969.
Task Group on Lung Dynamics, Deposition, and Retention
Models for Internal Dosimetry of the Human Respiratory
Tract, Health Physics 12 173 (1966).
Winkelstein, W. "The Relationship of Air Pollution and
Economic Status to Total Mortality and Selected Respira-
tory System Mortality in Man", Arch. Environ Health 14 .
162 (1967).
-23-
-------
6. Douglas, J. W. B. and Booras, S. G. "Air Pollution and
Respiratory Infection in Children." Brit. J. Prevt.
Social Med. 20, 1 (1966).
7. Lunn, J. E., Knowelden, J. and Handyside, A. J.,
"Patterns of Respiratory Illness in Sheffield Infant
School Children", Brit. J. Prev. Soc. Med. 21 (1967).
8. Petrilli, R. L., Agrese, G. and Kanitz, S., "Epidemi-
dogy Studies of Air Pollution Effects in Genoa,
Italy" Arch. Environ. Health 12 733 (1966).
9. Carnow, B. W., Lepper, M. H. Shebelle, R. B. and Stamler,
J. "The Chicago Air Pollution Study: S02 Levels and
Acute Illness in Patients with Chronic Broncho Pulmonary
Disease" Arch. Environ. Health 18 768 (1969).
10. Brasser, L. G., Joosting, P. E., and Von Zuelen, D.
"Sulfur Oxide - to What Level is it Acceptable?"
Report G-300, Research Institute for Public Health
Engineering, Delft, Netherlands, July, 1967.
11. Lawther, P. J., "Climate, Air Pollution and Chronic
Bronchitis," Proc. Roy. Soc. Med. 51. 262 (1958).
12. Lave, L. B. and Seskin "Air Pollution and Human Health"
Science 169 723 (1970).
13. Environmental Protection Agency "Health Consequences
of Sulfur Oxides: A Report from CHESS, 1970-1971."
Report EPA-650/1-74-004, May 1974.
-24-
-------
14. Corn, M. "Measurement of Air Pollution Dosage to Human
Receptors in the Community" Environ. Res. 3_ 218 (1970).
15. Timbrell, V. "Inhalation and Biological Effects of
Asbestos" in T. T. Mercer Stal. "Assessment of Airborne
Particles" p. 427, C. C. Thomas, Springfield, M. 1972.
16. Corn, M. "Urban Aerosols: Problems Associated with
Evaluation of Inhalation Risk" in T. T. Mercer, et al.
"Assessment of Airborne Particles" p. 465 C. C. Thomas,
Springfield, 111., 1972.
17. Corn, M., Montgomery, T. L. and Reitz, R. "Atmospheric
Particulates: Specific Surfaces and Densities"
Science 159 1350 (1968).
18. Air Quality Criteria for Particulate Matter, U..S. Dep.
H. E. W. Publ. AP-49, 1969.
19. Green, H. L. and Lane, W. R., "Particulate Clouds:
• Dusts, Smokes and Mists" Second Edition, E. and F. N.
Spon. Ltd., London, 1964.
*20. Hidy, G. M. and Friedlander, S. K., "The Nature of Los
Angeles Aerosol" in H. M. Englund and W. T. Beery (ed.)
"Proceedings of the Second International Clean Air
Congress", Academic Press, New York 1971.
21. Ensor, D. S., Charlson, R. J., Ahlquist, N. C., Whitby,
K. T., Husar, R. B. and Liu, B. Y. H., "Multiwavelength
Nephelometer Measurements in Los Angeles Smog Aerosol,"
in G. M. Hidy (ed.) "Aerosols and Atmospheric Chemistry",
Academic Press, N. Y., 1972.
-25-
-------
22, Ridker, R. G., "Economic Costs of Air Pollution",
New York, Prager, 1967.
23. Barrett, L. B. and Waddell, T. E., "Cost of Air Pollu-
tion Damage," EPA Report AP-85, February 1973.
24. Hidy, G. M. and Brock, J. R., Proceedings of 2nd Clean
Air Congress, IUAPPA, Washington, D. C., Dec. 1970.
25. "Compilation of Air Pollutant Emission Factors,"
Second Edition EPA Report AP-42, April 1973.
26. Vandegrift, A. E. et al., "Particulate Air Pollution in
the U. S.," J. Air Pollution Control Association, 2_1
321(1971).
*27. Sehmel, G. A., "Particle Resuspension From an Asphalt
Road Caused by Car and Truck Traffic," Atmos. Environ.
1 291 (1973).
28. Gatz, D. F., "Relative Contributions of Different
Sources of Urban Aerosols: Application of a New Esti-
mation Method to Multiple Sites in Chicago," Atmos.
Environ. £ 1 (1975).
*29. Miller, et. al., "A Chemical Element Balance for the
Pasadena Aerosol," J. Colloid Interface Sci. 39_ 165
(1972).
30. R. Drake in "Topics in Current Aerosol Research,"
Pergamon, Oxford, 1972.
31. M. Lee, R. et al., Atmos. Environ. 5 275 (1971).
-26-
-------
32. Pich, J., et al., Aerosol Sci. 1 115 (1970).
33. G. Hidy and J. R. Brock, "The Dynamics of Aerocolloidal
Systems," Pergamon, Oxford, 1970.
34. Kolmogorov, A., Akad, Nak SSSR, 3_1 99 (1941).
36. "Particulate Pollutant System Study," MRI Contract No.
CPA 2269104, EPA, 1971.
37. Schulz, E. J., et al., "Submicron Particles from a
Pulverized Coal Fired oiler," Atmos. Environ. 9^ 111
(1975).
38. Harrington, W., "Fine Particles", J. Air Pollution
Control Association, 1974.
r39. Winchester, J. W. and Nifong, G. D., "Water Pollution
in Lake Michigan by Trace Elements from Pollution Aerosol
Fallout," Water, Air, and Soil Pollution 1 50 (1971).
40. Natusch, D. F. S., et al., Science 183, 202 (1974).
41. Lee, R. E. and Von Lehmden, D. J., J. Air Pollution
Control Assoc. 23 853 (1973).
42. Toca, F, M., Thesis, University of Iowa, 1972.
43. Ruud, C. 0. and Williams, R. E., "X-Ray and Microscopic
Characterizations of Denver (1973) Aerosols," preprint,
Report Denver Research Institute, 1974.
44. Draftz, R. G. and Blakeslee, H. W., "Identification of
Ambient Suspended Particles from Philadelphia," preprint,
I.I.T.R.I. Report, 1974.
-27-
-------
*45. Draftz, R. G., "Analysis of Philadelphia Suspended
Dusts Sampled at Street Level," I.I.T.R.I. Report No.
C9915-1, 1974.
*46. Harrison, P., Draftz, R., and Murphy, W. H., "Identifi-
cation and Impact of Chicago's Ambient Suspended Dust,"
preprint, I.I.T.R.I., 1974.
*47. Draftz, R. G. and Durham, J., "Identification and
Sources of Denver Aerosol," preprint, Paper #74-263,
Air Pollution Control Association Meeting, Denver, 1974.
48. Whitby, K. T., "Modelling of Atmospheric Aerosol Particle
Size Distributions," EPA Progress Report, R800971.
49. Brock, J. R. in G. M. Hidy, Ed., "Aerosols and Atmo-
spheric Chemistry," Academic Press, New York, 1972.
50. Cox, R. A., "Particle Formation from Homogeneous Reac-
tions of Sulphur Dioxide and Nitrogen Dioxide," Tellus
XXVI, 235 (1974).
51. Van Luik, F. W. and Rippere, F. E. Annl. Chem., 34 1617
(1962).
*52. Miller, D. F. et al., "Haze Formation, Its Nature and
Origin," Final Report to C.R.C. and EPA, March, 1975.
53. Durham, J., Brock, J. R., Judeikis, H., and Lunsford,
J., "Review of Sulfate Aerosols," EPA Report, In Prep-
aration.
-28-
-------
54. "Proceedings of the 7th International Conference on
Condensation and Ice Nuclei," K. Spurny, Ed., Academica,
Prague, 1969.
55. Brock, J. R. and Marlow, W. A., "Charged Aerosols and
Air Pollution," Environ. Letters, To Appear, 1975.
56. Gartrell, G. and Friedlander, S. K., Atmos. Environ.
9 279 (1975).
57. Middleton, P. and Brock, J. R., "Atmospheric Aerosol
Dynamics: the Denver Brown Cloud," EPA Report, to
Appear.
58. Tuesday, C. S., Ed. "Chemical Reactions in Urban Atmo-
spheres," New York, Elsevier, 1971.
59. Altshuler, A. P. and Bufalini, J. J., Photochem. Photo-
biology, 4 97 (1965).
60. Air Quality Criteria for Photochemical Oxidants, N.A.P.
C.A. Publication No. AP-63, March 1970.
61 Alley, F. C. and Ripperton, L. A., "The Effect of
Temperature on Photochemical Oxidant Production in a
Bench Scale Reaction System," J. Air Poll. Cont. Assoc.,
11, 581 (1961).
62. Brock, J. R., Faraday Symposia .of the Chemical Society
No. 7, "Fogs and Smokes," The Chemical Society, London,
1973.
63. Lundgren, D. A., "Atmospheric Aerosol Composition and
Concentration as a Function of Particle Size and Time,"
J. Air Pollution Control Assoc. 2_0 603 (1970).
-29-
-------
64. Esmen, N. and Corn, M. "Residence Time of Particles in
the Atmosphere," Atmos. Environ. 55_ 71 (1971).
65. Hidy, G. M. and Brock, J. R., "An Assessment of the
Global Sources of Tropospheric Aerosols" Proc. of 2nd
Clean Air Congress, IUAPPA, Washington, D. C., December
1970.
References Cited by Babcock:
1. G. M. Hidy and P. K. Mueller, "Control Technology and
Aerosols," Environmental Research and Technology, Inc.,
741 Lakefield Road, Westlake Village, California 91361
(May 1975).
2. J, P^,Lodge, "Particulate Matter in the Atmosphere,"
385 Broadway, Boulder, Colorado 80303 (1975).
3. J. R. Brock, "Review of Suspended Particulate Matter,"
Chemical Engineering Department, University of Texas,
Austin, Texas (June 1975).
4. G. Gartrell and S. K. Friedlander, "Relating Particu-
late Pollution to Sources: the 1972 California Aerosol
Characterization Study," Atmospheric Environment, 9:
279-299 (1975). (This paper, although not prepared as
a part of the CSL-Radian study, provides a central input
to this summary report.)
5. L. R. Babcock and N. L. Nagda, "Indices of Air Quality,"
in W. A. Thomas (ed.), Indicators of Environmental
Quality, Plenum Press, New York (1972). pp. 183-197.
-30-
-------
6. F. F. Fennelly, "Primary and Secondary Particulates :."
as Pollutants, a Literature Review," J. Air Pollution
Control Assn., 25. (7): 697-704 (July 1975).
-31-
-------
APPENDIX A
"CONTROL TECHNOLOGY AND AEROSOLS"
G.M. HIDY AND P.K. MUELLER
MAY 1975
-------
CONTROL TECHNOLOGY AND AEROSOLS
ERT REPORT P-1588.F
MAY 1975
by
G. M. Hidy
&
P. K. Mueller
prepared for
RADIAN CORPORATION
Austin, Texas
ENVIRONMENTAL RESEARCH & TECHNOLOGY, INC
WESTERN TECHNICAL CENTER
741 Lakefield Road
Westlake Village, CA 91391
-32-
-------
SUMMARY AND RECOMMENDATIONS
This report gives a brief literature survey of the work
undertaken recently to characterize atmospheric aerosols in terms
of their origins. Particles come directly from stationary or mobil
sources, or from the production of condensed material by chemical
reactions taking place in the ambient air. Material resulting from
the oxidation of SO2, NOX and non-methane hydrocarbon vapor appears
to contribute a substantial fraction of the total particulate mass
concentration occurring in urban and non-urban air. A few cases
have been observed in Southern California when more than half of the
sampled particles stemmed from atmospheric reactions. This material
is concentrated in the submicron particle size range. The formation
of particles in the atmosphere from pollutant gases requires new
perspectives for controlling such materials, which must be taken
into account in future strategic planning.
It is recommended that:
1. The chemical element balance method for identifying the
origins of aerosols be applied to several cases in
diverse areas of the United States to confirm the
importance of secondary processes in aerosol pro-
duction and to separate man-made from natural secondary
aerosol sources.
2. Present control strategies for particulates be re-
considered to include the impact of the conversion of
SO2, NOV and non-methane hydrocarbon vapors into
!
condensed material.
3. To achieve the above recommendation, a substantial
amount of observational data is still needed for the
major urban regions. Studies are also needed to test
the evaluations made so far for the Los Angeles region.
-33-
-------
1.0 INTRODUCTION
This report has been prepared in response to a request from
the Environmental Protection Agency's (EPA) Control Systems Laboratory.
According to evidence from recent work characterizing aerosols, a
major fraction of the particles originates from atmospheric chemical
processes. Some studies suggest that these "secondary" generation
mechanisms dominate the particulate mass concentration compared with
"primary" emissions from sources. This conclusion is of considerable
importance in determining the future requirements of control technology
to meet present ambient air quality standards.
To assist EPA in the assessment of current control pro-
cedures, Radian Corporation has solicited information from several
investigators on the available studies of major atmospheric particle
sources. Findings of recent studies are discussed in this report.
Some of the work was sponsored partially or entirely by EPA. This
effort is evaluated with respect to future needs in particulate
control technology at the source, and recommendations are presented
that take into account atmospheric chemistry.
2.0 CHARACTERIZATION OF PARTICLES
During the last decade, considerable information has been
collected on the physical and chemical properties of urban and non-
urban aerosol particles for urban and non-urban air. The suspended
particles are composed of a wide variety of solid and liquid material.
Some of the particles are polycrystalline and appear to be soil dust
or other kinds of solids resulting from comminution processes. Other
solid particles are amorphous spheroids which are identified with
combustion processes or with condensation of vapor in the atmosphere.
Liquids are frequently seen in atmospheric aerosol samples, especially
in the submicron particle size range.
-34-
-------
Investigations such as those of Whitby et a_l (1972)
have established that aerosol particles range in size from less
than 100 & to several 10's of micrometers (ytn) in diameter.
The size distribution of particulate surface area and volume is
multimodal in character; the most commonly observed volume dis-
tributions are bimodal, with one mode peaking in the range between
1 and 10 vm diameter. An example of such distributions is shown
for Los Angeles air in Fig. 1. Evidence from prior observations
showing similar distributions are shown in Fig. 2. Here V is
particle volume and D is particle diameter. The volume distribution
is approximately proportional to the mass distribution.
Recent experiments such as those in the California Aerosol
Characterization Experiment (ACHEX) (e.g., Hidy e_t a_l, 1974) have
shown the behavior of the submicron and supermicron modes to be
essentially independent. The supermicron particle grouping appears
to be dominated largely by wind-blown dust, sea salt, road dust,
construction dust, and other debris from comminution processes.
The submicron range appears to be related principally to primary
emissions from the burning of fossil fuels and to the secondary
particulate production of condensable material in the atmosphere.
(See, for example, Heisler e_t a_l, 1973; Hidy and Burton, 1975;
Gartrell e_t al, 1975). The latter process is especially dramatic
in cities such as Los Angeles, where chemical reactions are a
dominant force in the evolution of smog aerosol (Hidy e_t al, 1974) .
An example of the changes due to particle growth during smog re-
actions and air mass transport is shown for Pomona, CA in Fig. 3.
The fraction of aerosol volume increased substantially between
early morning and mid-morning. The major accumulation of aerosol
volume in air moving over the fixed station takes place in the
submicron size range, while the supermicron range remains
essentially the same. Up to now, this phenomenon is poorly documented
except partially for Denver (Willeke, Whitby e_t al, 1974; Durham,
Wilson et a_l, 1975; Durham, Patterson et_ al, 1975). Currently
active EPA sponsored studies (W.E. Wilson) for the St. Louis region
-35-
-------
Lo
ON
e
W
H
W
O
O
a
o
H
3
W
O
53
O
0
60
50
40
30
20
10
1 T
1 T
o
a
*
•
Location
Hunter Liggett
Harbor freeway
Golds tone
Pt. Argue 1 lo
Time(PST)
1500
1950
1500
1830
Dace
9/13
9/27
11/1
11/9
V (M»Vc»J)
39.8
52.3
12.4
53.7
(84)
(MOTOR VEHICLE EMISSION) — J
/
/
/o—
-O D
(DESERT BACKGROUND)
.01
Figure 1.
Source:
.03 .06 0.1 0.3 0.6 1.0
Particle Diameter, pm
3.0
6.0 10
30
50
Comparison of Volume Distributions for Background and Motor Vehicle Source
Enriched (Near Harbor Freeway) Sites. V is the Total Volume Concentration
of Aerosol, as estimated from the particle counters. Data taken in 1972.
Hidy, e_t ^1, 1975: Characterization of Aerosols in California, Vol. 1,
California Air Resources Board Contract No. 358, Rockwell International
Science Center, Thousand Oaks, CA.
-------
e
m
e
a
oo
o
100
90
80
70
60
50
40
30
20
10
0
] I I II ' I I I I I I I III'! I
L.A..1969, 342 RUNS
MPLS (CLARK,1965) 56 RUNS
MPLS (PETERSON,1967) 45' 7 RUNS
COLORADO, 1970, 3 RUNS
SEATTLE (NOLL)
JAENICKE & JUNGE (1967)
IKITA (1955)
PARTICLE DIAMETER, D ,
Figure 2. Comparison of volume distributions measured by several
investigators in different locations. Note the universal
bimodal nature of all of these data and that the data
obtained by Clark, Peterson, and from the more recent
Los Angeles and Colorado studies, all fall into the
Jaenicke and Junge and Noll data at about 7 LOB. Also
note that the Colorado data which were obtained under
pollution-free conditions such that it may be assumed
that a background aerosol was being measured, is
rising sharply at 10 ym.
Source: Whitby et al, "The Aerosol Size Distribution in
Los Angeles Smog", Aerosols & Atmos. Chem. (G.M.
Hidy, ed.)- Academic press, N.Y., p. 260.
-37-
-------
I I I >lil
I I i I I I
100
80
E
60
o
r-i
<3
>
40
20
POMONA 8/17/73
A 0600-0800 PST
O 0800-1000 PST
03 MAX 28.6 PPHM
MAJOR GROWTH OF PHOTOCHEMICAL AEROSOL
I
.01 0.1 1.0 10
Figure 3. Evolution of the Volume Distribution of Smog Aerosol Taken at Pomona, CA 8/17/73.
Source: Hidy ejt ^,1975: "Characterization of Aerosols in California, Vol. 1, California Air Resources
Board Contract No. 358; Rockwell International Science Center, Thousand Oaks, CA.
-------
will test the occurrence of these types of changes further. No
preliminary data from these studies have as yet been available to
ERT.
The data shown in Figs. 1 to 3 indicate the great importance
of taking into account the wide range of particle size and the
behavior of certain parts of the size spectrum in response to primary
sources as well as atmospheric chemical processes. Since both
human respiratory responses and visibility are linked primarily
with the submicron or fine particle size mode, control efforts
should be focussed on those materials.
The chemical properties of aerosols provide important
additional information about the nature and origins of the particles.
The atmospheric aerosol is composed of a very wide range of both
inorganic and organic material.
Examples of the chemical composition of urban and non-
urban samples taken in California are shown in Table 1. Although
several elements and species are identified in these samples, the
material is dominated by carbon (water soluble sulfate), water
soluble nitrate, and ammonium. There is a large body of data on
aerosol composition available from the National Air Surveillance
Network and special studies which suggest that these four components
make up a major fraction of aerosols sampled in the United States.
Thus, the California experience should be applicable qualitatively
to other locations.
Attempts have been made recently to establish a total
material balance on filter collected aerosol particle samples
(e.g., Hidy and Friedlander, 1971; Miller et al, 1972; Heisler e_t a 1,
1973; and Gatz, 1975). Using the chemical analysis of samples and
certain assumptions about the composition, all except 20 to 3070 of
the total collected mass can be accounted for. Such calculations
have confirmed that sulfate, nitrate and non-carbonate carbon are
major contributors to the mass of aerosol. These materials must have
originated from atmospheric chemical processes since they are not
accounted for by emissions from known primary sources.
-39-
-------
MASS CONCENTRATION
PER UNIT
DIAMETER RANGE
(D , D + AD )
t
-O
o
TOTAL MASS MONITOR
AITKEN
NUCLEI
COUNTER
PRIMARY NATURAL
OR QUASI-NATURAL
PRIMARY ANTHRO-
POGENIC 6 SECONDARY
0.5
1 pm
DIAMETER
10 urn
Figure 4.
Source:
Hypothetical Presentation of Bimodal Mass Distribution
Based on Work on Urban Aerosols through 1971.
Hidy, e£_al, 1974: "Characterization of Aerosols in California1
Volume 4, Calif. Air Resources Board Contract No. 358, Rockwell
International Science Center Report SC524.25FR, Thousand Oaks,
California 91360.
-------
Table 1
Chemical Composition of Filter Collected Aerosol Samples
WeiQht Percent
Non-Urban
Urban
Constituent
Al
Si
Ma
Cl
K
Ca
Ti
V
Cr
Mn
Fe
Cu
Zn
Br '.
PB
I
S04=
N03"
NH4+
Non Carb. C
Total measured ( g/m3) 30
Pacific Coast*
Offshore
2.8
5.0
3.6
9.9
1.3
1.9
0.15
N.D.
0.02
0.04
1.6
0.6
0.2
0.2
0.4
Desert
2.4
5.2
0.71
0.32
0.39
2.1
0.13
0.004
N.D.
0.028
1.4
0.028
0.035
0.059
0.017
Fresno
2.4
6.3
0.40
0.19
0.33
0.47
0.09
N.D.
0.0058
0.025
03
0058
0.055
0.098
0.42
Pomona
0.99
2.2
1.1
0.51
0.24
1.0
0.11
0.008
0.013
0.025
0.86
0.014
0.13
0.36
1.6
17.5
5.7
2.2
3.0
2.8
3.5
1.7
N.D.**
2.0
3.8
1.5
S.G*
10.6
20.2
9.1
24. 0^
42.4
207
180
•Average of 20 samples at San Nicolas Island (Hidy et_al_., 1973)
+Single 24-hour sample (Hidy et_ al_., 1974)
•^-Estimated as 3 x cyclohexane extractable organics
**Not detected
Source: Hidy, G. M. 5 C. S. Burton, 1975: "Atmospheric Aerosol
Formation by Chemical Reactions", presented at Sytn. Chem. Kinetcis
for Lower § Upper Atmos., Warrenton, VA., Sept. 1974.
-41-
-------
Microscopic analysis has been developed through the work of
McCrone and others and has been used recently by Draftz and
colleagues to identify the origin of large, solid particles sampled
in Denver and in the Chicago area. Filter and impactor samples
have been examined by polarized light microscopy after treating
the filter substrate with immersion oil. This method is useful
in characterizing particles but suffers from the severe limitation
that it can be used only for solid, insoluble (in immersion oil)
particles larger than 0.5 um diameter. As a result, the discussions
of source characterization by Draftz and Durham (.1975) (also Harrison
e_t al, 1975) do not include the fine particle range where secondary
material may be expected. Their conclusions that most of the
sampled material is primary in origin is based on a partial
"material balance" and cannot be considered to provide a complete
picture of the contribution of primary or secondary sources to
atmospheric aerosol.
Methods for electron microscopic identification of
particles were developed by Frank & Lodge (1967) . Their work was
particularly useful in characterizing particles such as sulfuric
acid and sodium chloride. Such investigations have provided useful
information about the origins of aerosols, especially from volcanoes.
But this technique is difficult to implement quantitatively.
The second method is to "fingerprint" sources of aerosols
by interpretation of changes in size distributions with time. The
group at the University of Minnesota has catalogued a large number
of size distributions taken near individual sources in an attempt
to find unique patterns. So far the work has indicated that
blowing soil dust and strong combustion sources can be identified.
Air mass trajectory and chemical information is required in addition
to relate the source contributions to various size fractions observed
in the ambient air.
-42-
-------
The third method utilizes the chemical composition of
aerosol particles collected on a substrate by relating the com-
position to components which are virtually unique for major sources.
(See, for example, Hidy & Friedlander, 1971; Miller et al, 1972;
Friedlander, 1972; Heisler e_t al, 1973; and Gartrell & Friedlander,
1975.) Examples of chemical tracers include lead for automobiles,
sodium for sea salt and soil dust, silicon for soil dust, vanadium
for fuel oil, etc. The source identification can be made in broad
classes using such tracers and an emission inventory for a city.
The method has been applied to the particulate matter found in
Los Angeles. A typical breakdown by sources is shown for non-
smoggy and smoggy conditions in Figs. 5A and 5B. The primary emissionsj
account for less than half of the total mass concentration in air
over this city. The secondary contributors of sulfate, nitrate,
organic carbon, ammonium and water encompass the remainder of the
sample. These components are substantially increased during the
smoggy conditions.
A similar analysis has been made for aerosols sampled at
a remote island site approximately 90 miles west of Los Angeles.
The estimated breakdown by source for non-urban material offshore
is listed in Table II. Even in this case the major part of the
material appears to be attributable to secondary reactions of the
pollutant or quasi-pollutant gases, S02, N0x, NH3 and hydrocarbon
vapors. Analogous preliminary chemical information has become
available for a Bermuda aerosol sample (Akselsson ejt al, 1975.)
Utilizing a novel approach, particles are collected as a function
of size on five cascade impactor stages and are analyzed by proton
elastic scattering (PESA) and proton induced X-ray emissions
(PIXE). PESA obtains C, 0, N and S in the sample while PIXE obtains
the higher atomic number elements. The carbon was most abundant,
25 to 30 yg/m3. The ratios of S, N, and 0 were consistent with that
expected for (NH<*)2 S0i» . The data published to date is not
sufficient for detailed interpretation. An estimated excess observed
in the nitrogen fraction suggests also the presence of some nitrate.
In view of these findings at a remote location, it becomes important
to find ways for allocating separately the anthropogenic and non-
man made contributions to the secondary aerosol.
-43-
-------
NO 3
(26%)
WATER
(16%)
(13%)
ORGANICS
(43%)
ORGANICS
(24%)
NATURAL (8%)
NATURAL (SEA
SALT,SOIL)
(11%)
TRANS-
PORTATION
(12%)
INDUSTRIAL
(7%)
CHEMICAL DUST (2%)
SECONDARY
;;;\ui^INDUSTRIAL &
Mi/ CEMENT DUST (3%)
TRANSPORTATION
(6%)
A. PASADENA 9/20/72
1200-1400 PST
LOW OXIDENT, TOTAL MASS
CONCENTRATION, 79 ug/m3
B. POMONA 10/24/72
1200-1400 PST
MODERATE OXIDANT, TOTAL
MASS CONCENTRATION, 178
Figure 5. Distribution by Source of Aerosol Mass Concentration for Filter Samples
Collected Over Two-Hour Periods, and Equilibrated to Air at Less Than
50% Relative Humidity.
Source: Hidy e± al^, 1975: "Characterization of Aerosols in California, Vol. 1,
California Air Resources Board Contract No. 358, Rockwell International
Science Center, Thousand Oaks, CA.
-------
TABLE II
CONTRIBUTION BY SOURCE TO SN'I AEROSOLS (WEIGHT. PERCENT)
1. Primary Soil (based on 25% Si in soil) 20.0
Sea salt (based on 30.6% N'a in sea salt) 11.1
Automobile lead (based on 40% lead in 1.0
auto emitted aerosol)
Carbon* 3
2. Secondary Sulfate 17.5
Ammonium 2.2
Nitrate 5.7
61
3. Total volatiles other than carbon: 100 - (Total 22
ash carbon) —
Total accounted for 83
*Some of this may be produced in the atmosphere by chemical
reaction; however, it is identified with primary origin for this
estimate.
Source: Hidy e_t_ al_, 1974: "Observations of Aerosols over Southern
California Coastal Waters", J. of Appl. Meteror.Vol. 13,
pp. 96-107.
-45-
-------
The chemical element balance approach has been used by
Harrizon & Winchester (1971) for analysis of sources for the
Chicago area. Sources were identified, but the chemical analysis
was insufficient to deduce the significance of secondary processes
in that city. Gatz (1974) has recently reported a similar analysis
for aerosols sampled in St. Louis. Although the data again suggest
the important contribution of sulfate and nitrate to the aerosol,
the non-carbon constituent was unavailable to complete the material-
source balance.
Several years ago, the calculation method of Hidy &
Friedlander (1971) was applied to National Air Surveillance Network
data for several cities, including Philadelphia; New York;
Washington, D.C. and St. Louis. These calculations were undertaken
by EPA staff and resulted in an internal Agency document which is
not currently available. Friedlander was involved in this study
and may have a copy. In any case, the results of such calculations
appeared to show the importance of the "secondary" constituents,
sulfate, nitrate and organic carbon, as a major contributor to
aerosol mass concentration in the cities considered.
On a global basis, Hidy & Brock (1971) estimated that the
secondary particles formed by atmospheric chemical reactions of
SO2, NOX and hydrocarbon vapor would be dominant in the worldwide
aerosol. The calculations were based on considerations similar to
those introduced by Hidy and Friedlander (1971).
The chemical element balance approach was extended recently
to estimate the relative contributions of various sources to the
secondary part of the ambient aerosol. White et_ al (see also
Hidy e_t al, 1974) segregated two major categories -- stationary
(fuel oil) and mobile (gasoline) combustion sources in the Los Angeles
Air Basin. From the emission inventory for the city and atmospheric
data obtained in the Aerosol Characterization Experiment, the
contribution from these sources to the aerosol mass concentrations
were calculated. The results are shown graphically for different
locations in Fig. 6. Stationary sources were more important total
contributors to the aerosol in the western and central part of
the Basin than in the eastern area.
-46-
-------
DOMINGUEZ HILLS
118
Figure 6.
Source:
Distribution of Sources for Aerosol Mass Concentration in the South Coast Air Basin.
Numbers by Discs are Averaged Mass Concentrations of 2-hour 1973 samples. The Contour
is the 1000 ft. height level.
Hidy, et al , 1975: "Characterization of Aerosols in California, Vol. I, California Air
Resources Board Contract No. 358, Rockwell Intl. Science Center, Thousand Oaks, Ca .
-------
The contributions to particulate loadings from the oxidation
of S02, NOX and hydrocarbon vapors is expected from their chemistry.
The known chemistry about aerosol evolution was recently reviewed
by Hidy & Burton (1975, Appendix). They showed several plausible
pathways for the formation of condensed material by atmospheric
chemical reactions. These reactions may be homogeneous-gas phase
or heterogeneous; the latter involve particles and hydrometeors.
All of the discussion dealing with the filter or impactor
sample collection and subsequent chemical analysis assume that current
methods reflect "real" particle properties as they exist suspended in
the atmosphere. There are significant uncertainties in current
sampling methods which have been documented recently. Some of these,
such as particle bouncing in impactors and loss of water, can be
minimized. Other effects such as the influence of reactive gases,
NOX or S02, on particles collected on a substrate, or on the substrate
itself, are currently being investigated. Some of the methods used
in the past may have yielded suspect data and interpretations because
the magnitude of possible collection and chemical analysis errors
remain uncertain. The ACHEX work was conducted with such problems
in mind (see Hidy et al, 1974). However, further developments in
sampling and chemical analysis methods are needed to obtain a much
improved data base.
Of the three methods for identifying particulate sources,
the chemical element balance appears to offer the most promising
basis for clues to the origins of the atmospheric aerosol. So far,
it has been applied in complete form only to samples collected in
Southern California. The method requires (1) detailed knowledge of
the particulate chemical composition, (2) emissions inventories,
and (3) some knowledge of chemical species from primary and
secondary sources. Further use of the method should be made to
investigate aerosol behavior in diverse areas of the United States
to check the universality of the conclusion that secondary particle
production can account for a large fraction of the particulate mass.
-48-
-------
4.0 IMPLICATION TO CONTROL TECHNOLOGY
Even with uncertainties, it is evident circumstantially
that chemical processes generating condensed material in the
atmosphere provide a major fraction of the ambient aerosol con-
centration. In some cases, it appears that such material can
account for much more than half of the total mass concentration
of particles.
The material from secondary production includes water
soluble sulfate and nitrate, non-carbonate (organic) carbon and
ammonium. These are derived from atmospheric reactions involving
the gaseous precursors, S02, N0x, (non-methane) hydrocarbon vapor
and ammonia.
Because significant amounts of aerosol are derived ultimately
from pollutant gases, it places potentially important and new con-
straints on control strategy for aerosols. It is now recognized that
control of fine particle emissions in the 0.1 to 1 ym diameter range
at the source is crucial for achieving improved air quality for
public health. Even though these particles may include only a part
of the total mass concentration, they are important for development
of respiratory ailments and remain in the atmosphere the longest
(see also Hidy, 1973). The fine particle control problem is
compounded by the addition of condensable material formed following
atmospheric chemical reactions. The natural background component of
the secondary aerosol remains to be assessed. For the Los Angeles
Region, the annual background for all suspended particulate matter
has been estimated at about 4.0 pg/m3 (Trijonis, 1974). S02, N0x,
and hydrocarbon vapor from stationary and mobile sources in the light
of expected aerosol production.
Proposed control strategies for the anthropogenic aerosol
were calculated by roll-back from a base year and assumed a linear
function to relate S02, N0x and reactive hydrocarbon emissions to
sulfates, nitrates and secondary organics respectively (Trijonis,
-49-
-------
1974). Apparently, this approach is being extended to other
regions with current EPA sponsorship. Hidy e_t a_l (1974) made
similar assumptions but modulated the impact on both total mass
(TSP) and visibility ( scat) by considering the interactions which
were observed during relatively severe pollution episodes. The
following equations were derived statistically:
TSP (yg/m3) = 26 + 1.56 (N03~) + (1.28 - O.S8 w2) (S04=) + 2.55 (organics) * 21,
and
bscat x 104(m-l) = -l.i + 0.074 ($04=) + (0.025 + 0.049 y2) (NOs") + 0.025 (organics)
+ 0.025 (Mass - S04= - N03' - organics) ±0.9
where constituent concentrations are given' in ug/m3 and u is the
relative humidity.
An analogous correlation was obtained by Barone, Cahill
e_t al (1975) for a more extensive set of measurements. Their
correlation relates composition to prevailing visibility. Elements
and gas phase components including water vapor and wind speed were
unfortunately omitted. While this type of correlative analysis
of available data has the potential for more realistic evaluation
of control strategies than linear roll-back alone, their sensitivities
and applicability for different regions and time spans remain to be
evaluated. A thorough analysis and substantially more observational
data are still needed to estimate the required reductions in gaseous
emissions to achieve the ambient aerosol standard. With such
information as a base, strategies for control of major sources could
be developed for given AQCR's.
-50-
-------
REFERENCES
Akselsson, K.R., J.W. Nelson, & J.W. Winchester, 1975: "Proton
Scattering for Analyses of Atmospheric Particulate Matter", Bull.
Am. Phys. Soc. II, 20, p. 155.
Barone, J.B., T.A. Cahill, R.G. Flocehini, D.J. Shedoan, 1975:
"Visibility Reduction: A Characterization of Three Urban Sites in
California", Science, in manuscript, Feb. 12.
Draftz, R.G. & J. Durham, 1975: "Identification & Sources of Denver
Aerosol". Unpubl. report to U.S. Environmental Protection Agency;
also Harrison, P.W. e_t al, "Identification & Impact of Chicago's
Ambient Suspended Dust". Unpubl. report for U.S. Environ. Protection
Agency.
Durham, J.L., W.E. Wilson, T.C. Ellestod, K. Willeke and K.T. Whitby,
1975: "Comparison of Volume and Mass Distribution of Denver Aerosols",
Atmos. Environment, in press.
Durham, J.L., R. K. Patterson, J.J. Vanee and W. E. Wilson, 1975:
"The Chemical Composition of the Denver Aerosol", Atmos. Environment,
in press.
Frank, E. & J.P. Lodge, Jr., 1967: "Morphological identification of
airborne particles with the electron microscope". J. Microscopic, 6,
449-456.
Friedlander, S.K., 1973: "Chemical Element Balances & Identification of
Pollution Sources". Environ. Sci. & Technol., 7_, 235-240.
Gatz, D.F., 1975: "Relative contributions of different sources of
urban aerosols: application of a new estimation method to multiple
sites in Chicago". Atmos. Environ. I, 1-18.
Gartrell, G., Jr., and S.K. Friedlander, 1975: "Relating particulate
pollution to sources: the 1972 Calif. Aerosol Charact. Study", Atmos
Environ., 9, 279-300.
Harrison, P.R. & J.W. Winchester, 1971: "Areawide distributions of
lead, copper and cadmium in air pollutants from Chicago and Northwest
Indiana". Atmos. Environ. 5, 863-880.
Heisler, S., et al, 1973: "The Relationship of Smog Aerosol Size &
Chemical Element Distributions to Source Characteristics". Atmos.
Environment 7. 633-649.
-51-
-------
Hidy, G.M., 1973: "Removal Processes of Gaseous & Particulate
Pollutants" in Chemistry of the Lower Atmosphere, (S.I. Rasool,
ed.), Plenum Press, N.Y., Chap. 3.
Hidy, G.M., e_t al. , 1974: "Characterization of Aerosols in
California (ACHEX)". Final Report Volumes 1-4; "Rockwell Science
Center, Report #SC524.25FR, Thousand Oaks, CA 91360.
Hidy, G.M. & J.R. Brock, 1971: "An Assessment of the Global Sources
of Tropospheric Aerosols" in Proc. 2nd IUAPPA Clean Air Congr.
(H.W. Englund & W.T. Berry, ed.), Academic Press, N.Y., p. 1088.
Hidy, G.M. & C.S. Burton, 1975: "Atmospheric Aerosol Formation by
Chemical Reactions: to be publ. in Int'l. J. of Chem. Kinetics.
Hidy, G.M. & S.K. Friedlander, 1971: "The Nature of Los Angeles
Aerosol". Proc. 2nd IUAPPA Clean Air Congr. (H.M. Englund & W.T.
Berry, ed.), Academic Press, N.Y., p.391.
Hidy, G.M. e_t al, 1974: "Observations of Aerosols over Southern
California Coastal Waters", J. of Applied Meteorology, Vol. 13, No. 1,
pp. 96-107.
Miller, M.S. e_t al, 1972: "A Chemical Element Balance for the
Pasadena Aerosol: in Aerosols & Atm. Chem. (G.M. Hidy, ed.), Academic
Press, N.Y., p. 301.
Trijonis, J., 1974: "A Particulate Implementation Plan for the Los
Angeles Region". TRW Report for EPA.
Whitby, K.T., R.B. Husar & B.Y.H. Liu, 1972: "The Aerosol Size
Distribution of Los Angeles Smog" in Aerosols & Atmos. Chem. (G.M.
Hidy, ed), Academic Press, N.Y., p. 2TT
Willeke, K., K.T. Whitby, W.E. Clark, V.A. Marple, 1974: "Size
Distribution of Denver Aerosols -- A Comparison of Two Sites", Atmos.
Environment, 8_, pp. 609-633.
-52-
-------
APPENDIX FOR
ERT REPORT P-1588.F
-------
ATMOSPHERIC AEROSOL FORMATION
BY CHEMICAL REACTIONS
by
G. M. Hidy *
California Institute of Technology
Pasadena, California
and
C. S. Burton
Science Center, Rockwell International
Thousand Oaks, California
* Also affiliated with Environmental Research & Technology,
Thousand Oaks, California
-53-
-------
ABSTRACT
This paper is a review of several aspects of aerosol
formation processes that occur in the Earth's atmosphere.
Important contributors to atmospheric aerosol chemistry are the
sulfates, nitrates, and organic compounds, which are considered
to be formed mainly by reactions of trace reactive gases in air.
The phenomenology of the evolution of aerosols by such chemical
processes is illustrated by recent observations taken in
Los Angeles smog. The results of this program show the strong
relationships between sulfate, nitrate, and organic carbon
formation and gas phase processes of photochemical smog.
Suspected mechanisms of secondary aerosol production
are summarized, with consideration for both homogeneous and
heterogeneous processes. These mechanisms are examined in the
light of laboratory simulations and the knowledge of atmospheric
behavior to deduce the potential importance of certain classes of
reactions for explaining aerosol evolution. Such considerations
illustrate well the complexities of gas-particle interactions in
atmospheric chemistry.
•-54-
-------
1. INTRODUCTION
Over the past century, hundreds of studies have been made
on visibility reduction in the atmosphere which are associated with
aerosols making up haze. It has been known for sometime that haze
involves the light scattering from tiny submicron particles suspended
in air. Yet only recently has there been widespread recognition that
such particles play an active role in air chemistry. Through their
production and growth, workers have come to link them with the
atmospheric removal of certain reactive gases such as the sulphur
and nitrogen oxides and hydrocarbon vapors. The chemical mechanisms
of aerosol formation in the atmosphere remain uncertain, but it is
likely that many processes play a role, including important photo-
chemical reactions.
Since Tyndall's 19th century classical experiments on
aerosol optical behavior, it has been known that exposure of reactive
gases to light can produce copious quantities of airborne particles.
More recent observations have led investigators to believe that sun-
light induced chemical reactions are responsible for a substantial
fraction of the intense haze observed over cities such as those
in Southern California. The relation between haze and photochemical
reactions in polluted air was deduced many years ago from the work
of Haagen-Smit and co-workers, and photochemical aerosol production
was explored by many investigators including Stephens, Hanst,
Renzetti, and Doyle (e.g., Leighton (1)). The possibility of wide-
spread haze formation by sunlight induced reaction of background
nitrogen oxides and natural terpenoid compounds from vegetation
was suggested several years ago by Went (2). Recent research on
the stratospheric aerosol has indicated that the major constituent
by mass or airborne particles at high altitude is sulfur. One
explanation for the presence of such material is via oxidation of
traces of sulfur dioxide above the troposphere (e.g., Friend et_ al.
(3)).
-55-
-------
With increasing concern for the influence of man's
activities on the atmosphere, there is interest in improving the
air quality levels for aerosols generated from atmospheric pollutants,
Thus there has been a substantial increase in research activity
devoted to atmospheric aerosol chemistry, particularly photochemistry,
At the same time that a variety of new laboratory experiments have
been updated, major new field studies have been implemented to
characterize atmospheric haze formation in more detail. Perhaps
best documented of these is the 1969 Pasadena Smog Aerosol Study,
whose methods and results were reported respectively by Whitby et al.
(4) and in the volume, Aerosols & Atmospheric Chemistry, edited by
Hidy (5). This project was an exploratory one, but recently much
larger investigations have been initiated, including the California
Aerosol Characterization Experiment (ACHEX) (Hidy ejt al. (6)), and
a portion of the Regional Air Pollution Study (RAPS).
The key to improved characterization of photochemical
aerosol behavior in the atmosphere is the quantitative description
of the physical and chemical properties of this material as well
as its kinetics of formation. A combination of observations,
atmospheric simulation, and laboratory experiments is required to
elucidate the complexities of particle evolution. The purpose of
this paper is to review the current knowledge in these areas and to
synthesize them into suggested key processes of interest in the
atmosphere. First several important new results from the study of
photochemical aerosols are outlined. Then conclusions from
simulation experiments are related to recent basic investigations
to elucidate relevant physico-chemical mechanisms applicable to the
troposphere and the stratosphere. Finally a current picture of
aerosol formation in the atmosphere is discussed in relation to
the combination of evidence from field observations and knowledge
of physical chemistry.
-56-
-------
2. ATMOSPHERIC OBSERVATIONS
A. The Troposphere
Photochemical aerosols in the atmosphere require an
operational or phenomenological definition. One definition might
come from the comparison of the nature of suspended material from
atmospheric samples and that generated in controlled laboratory
experiments involving irradiated mixtures of reactive gases in
air or filtered ambient air (e.g. Husar et al. (7)). In the
atmosphere along, a definition becomes difficult because of the
complexities of air chemistry involving a series of processes that
are both sunlight induced and thermal in nature. A traditional
operational definition of photochemical aerosols in the troposphere
are those clouds of particles in air that are identified with hazes
(a) where ozone is present, and (b) are formed and dissipated on
a time scale of hours with maximum visibility degradation at mid-
day. Perhaps the best known example is the intense haze formation
associated with photochemical smog over the Los Angeles area.
For the details of aerosol photochemistry one must rely heavily
on observations taken in Los Angeles for they are far better documented
than elsewhere.
One semi-quantitative measure of the concentration of
haze is the extinction coefficient, bscat» f°r visible radiation
resulting from the scattering of light from airborne particles.
Although bgcac has been found not to correlate uniquely with
oxidant behavior, an important correspondence in the relationship
appears to exist for b_oat. taken when ozone is measured at its
S C a C
maximum on a given day in .the Los angeles area. This relation
is shown in Figure 1. Here the strong increase in light scatter-
ing (visibility reduction) with intensity of photochemical smog
is demonstrated. The relationship is supported further by
observations taken by blimp traveling at mid-day on a north-south
path across zones of increasingly high ozone concentration over
central Los Angeles.
-57-
-------
Physical Properties
The scattering coefficient, b _,„_,_, depends on the particle
S Ccl u
size-number distribution of the haze, the particle index of
refraction as well as the wavelength of incident light. The number-
size distribution and its moments can be considered the key physical
properties of concern in characterizing atmospheric aerosol
behavior. This distribution is often defined in terms of volume v,
or the distribution function n (v,t,x) dv refers to the number of
particles per unit volume in the particle volume range v and v +
dv at any time t, and any point in space given by x. The particle
size distribution is difficult to measure so that only moments of
the distribution are monitored routinely. The zeroth moment of
the distribution function corresponds to the total number concen-
tration of particles, N, the two-thirds moment is proportional to
the total surface area S per unit volume, and the first moment is
proportional to the volume fraction, V. For particles of an average
mass density p, the product of p V is the total mass concentration;
b is a product of the Mie scattering function and S summed
over all particle sizes. Evidence has accumulated from several
experiments (e.g., Charlson et al.(8)), that b__Qt. is dominated
- - •^— SCclt
by light scattering in the particle diameter range 0.1 ym to
1.0 urn. Indeed, observations taken in the ACHEX (6), as well as
the 1969 Pasadena study show that particle evolution in smog is
strongly concentrated in this size range. This is illustrated
readily by measurements of the change in volume-diameter distribution
from early to mid-morning in smog. An example of data taken in
Pomona in 1973 is shown in Figure 2. These and other data also
illustrate that few new nuclei smaller than 0.1 ym are produced in
smog, in contrast to intense growth of existing particles during
the evolution reactions of smog, which are believed to produce
the condensable precursors.
-58-
-------
14
12
10
E
O
-------
RADIAN
CORPORATION
Based on thertnodynamic considerations Kelvin effect,
there is reason to believe that the stability of aerosols is
strongly influenced by the change in vapor pressure of materials
with particle size. For most condensable species, the surface
energies dictate that the stable particle is approximately 0.1 ym
diameter, suppressing growth and new particle formation on ranges
less than the size (see also Heisler e_t a_l. (9) ; Hidy (10)).
The fraction of particles smaller than 5-10 pm diameter
is maintained at a low concentration by several removal processes
including gravitational fallout (for details, see Hidy (11)).
Haze evolution is most spectacular over cities with polluted air
like Los Angeles because of high concentration of precursors combined
with intense sunlight. However, it is believed that similar
reactions must take place almost universally in the troposphere at
slower rates, modulated by repeated irradiation by sunlight and
intermittent removal processes involving clouds.
Chemical Properties. Chemical analysis of airborne particles
provides the most revealing information about the nature and
significance of aerosol formation by atmospheric chemical processes.
There is an accumulation of information from many studies, as
reviewed, for example, by Junge (12) and from more recent reports
such as those of the National Air Surveillance Network (13), that
water soluble sulfate, nitrate and organics make a substantial
fraction of the tropospheric aerosol. The importance of these
constituents is illustrated in Table 1 for urban and non-urban
aerosol samples in California. Examination of such samples also
suggests that the sulfate and nitrate exist as ammonium salts
through observations in other geographical areas may indicate the
presence of sulfuric acid. For example, Charlson e_t al. (14) have
reported evidence of free sulfuric acid from non-urban air near
St. Louis, and Cunningham (15) has identified by infrared analysis
sulfuric acid in urban samples of aerosol collected near Chicago.
-60-
-------
120
I I I I I I I I
I I I I I I
iir»
100
80
E
60
I 3.
CL
o
en
O
40
20
I
I I
POMONA 8/17/73
A 0600-0800 PST
O 0800-1000 PST
O3 MAX 28.6 PPHM
MAJOR GROWTH OF PHOTOCHEMICAL AEROSOL
0.1 1.0
D (pm)
Figure 2. Evolution of the volume distribution of smog aerosol taken at Pomona, California
8/17/73 (from Hidy ejt al. (6)).
-------
Table 1
Chemical Composition of Filter Collected Aerosol Samples
Weight Percent
Non-Urban
Urban
Constituent
Al
Si
Na
Cl
K
Ca
Ti
V
Cr
Mn
Fe
Cu
Zn
Br '
PB
I
S04=
N03"
NH4+
Non Carb.
Total measured
.Pacific Coast*
Offshore
2.8
5.0
3.6
9.9
1.3
1.9
0.15
N.D.
0.02
0.04
1.6
0.6
0.2
0.2
0.4
Desert
2.4
5.2
0.71
0.32
0.39
2.1
0.13
0.004
N.D.
0.028
1.4
0.028
0.035
0.059
0.017
Fresno
2.4
6.3
0.40
0.19
0.33
0.47
0.09
N.D.
0.0058
0.025
1.03
0.0058
0.055
0.098
0.42
Pomona
0.99
2.2
1.1
0.51
0.24
1.0
0.11
0.008
0.013
0.025
0.86
0.014
0.13
0.36
1.6
17.5
5.7
2.2
3.0
2.8
3.5
1.7
N.D.**
2.0
3.8
1.5
5.0+-"
10.6
20.2
9.1
24.0++
30
42.4
207
180
*Average of 20 samples at San Nicolas Island (Hidy eit_ a_K (64))
^Single 24 hour sample (Hidy et_ a]_. (6))
"^Estimated as 3 x cyclohexane extractable organics
**Not detected
-62-
-------
The mass concentration of sulfate and nitrate cannot be
accounted for from either natural or man made primary sources
(Robinson and Robbins (16); Hidy and Brock (17)). Thus the pro-
duction of aerosols in the atmosphere is believed to involve
mainly the highly complex chemistry of sulfur, nitrogen, and organic
carbon. These constituents come from gaseous emissions of both
natural and anthropogenic origin and are key ingredients in the
"classical" hierarchy of gas reactions in photochemical smog.
Most of the nitrate appears to be inorganic, with only a minor
organonitrate contribution, evidently linked with photo-chemical
processes (O'Brien e_t al. (18) and Hidy (10)).
The organic fraction is characterized by more oxygenated
material than accounted for in the vapor. High resolution mass
spectroscopy and solvent extraction methods, combined with infrared
analysis, suggest that more than half the organic aerosol os
oxygenated, with a large carboxylic acid component (e.g., Schuetzle
(19); Hidy e_t al. (6) ; and Grosjean and Friedlander (20)).
The size distribution of "secondary" aerosol constituents
is of interest for checking the evolutionary processes indicated
by physical measurements. Observations of the mass concentration
of sulfate over two hour intervals in smoggy air over Los Angeles
and other cities indicates that at least half of this material is
found in the fraction of aerosol less than 1.0 urn diameter. Under
conditions of heavy smog, most of the sulfate in Los Angeles air is
concentrated in the small particle fraction during the day, but
at night and in the early morning the sulfate shifts to larger
particles (Hidy, et al. (6)). In contrast, nitrate appears to be
present in somewhat larger particles in the Los Angeles aerosol
except in areas such as Riverside which is far downwind of the city
under the most common daytime conditions.
-63-
-------
The distinction between the size distribution of sulfate
and nitrate in smog aerosol is illustrated in the data shown in
Figure 3. The case shown in Figure 3A corresponds to a 24 hour
sample in West Covina, California during which the maximum ozone
concentration exceeded 0.4 ppm in the east central part of the
Los Angeles Basin. Another case from the Riverside area farther
east is shown in Figure 3B in which the maximum ozone concentration
was above 0.3 ppm. These samples illustrate that the mass dis-
tribution of nitrate and sulfate constituents in more aged smog
than that at West Covina is almost identical, but much more
nitrate is present downwind, farther to the east.
Diurnal Patterns of Evolution. An important description
of the evolution of photochemical smog comes from examination of
the changes during the day of the reactive gases in the air. The
classical picture of smog formation is revealed in the set of data
in Figure 4. Here the familiar pattern of early morning depletion
in NO and hydrocarbons takes place with a mid-morning maximum in
NC>2 and a mid-day maximum in ozone concentration. Such behavior
has been simulated successfully many times in smog chamber
experiments. Accompanying the increase in ozone is the increase in
bgc , reflecting the aerosol growth and subsequent visibility
reduction in smog through the mid-day. Changes in total number
concentration are less well correlated with the reactives
gas behavior, and are more closely related to local combustion
sources.
The diurnal changes in sulfate, nitrate, and carbon con-
stituents accompanying the gas chemistry are shown in Figure 5. In
many (but not all) cases studied in ACHEX (Hidy e_t al. (6)) ,
observations showed that sulfate and non-carbonate carbon follow
closely the ozone development with maxima in mid-afternoon. S0<» =
-64-
-------
\
cr
2:
LU
O
IS)
50
-------
cr>
Figure A. Diurnal patterns for Weo t Covina, July 23-24, 1974.
Vj - total particle volume,.V(< 1 pm) •« particle volutr.c
leas than 1 \tm diameter, V(0.01 > Dp > 0.1) «• particle
volume in the range 0.01 - 0.1 ura diameter (Dp), and
CNC •• condensation nuclei count (from Hidy ct nl. (6)).
-------
also follows S02 changes during the day. In contrast, nitrate
was found to develop a maximum concentration by mid-morning with
the N02 pattern. Thus the components of the aerosol behave in
different ways, and evidently sulfate and organics are generated
by distinctly different chemical mechanisms from nitrate.
Conversion Ratios. A useful measure of the extent of
conversion of an aerosol precursor with condensed material is the
so-called conversion ratio as follows:*
r _ mass of particulate sulfur based on SOd"
S ~ mass of particulate sulfur + mass of gaseous sulfur based on SO.
, _ mass of particulate nitrogen based on NOi"
N ~ mass of particulate N + mass of gaseous N based on NO-
r _ mass of particulate non-carbonate C - primary non-carbonate C
C " mass of particulate non-carb. C - primary non-carb. C + mass of
reactive gaseous C from NMHC
In the troposphere, the data on these ratios are limited, but a few
values are given in Table 2 for comparison. Depending on the location
*The secondary organic carbon is defined as the total non-carbonate
carbon less the carbon identified with motor vehicle sources. For Los
Angeles, the latter is accounted for approximately as a factor«: pro-
portional to the airborne lead concentration. <* is taken as unity for
the purposes of this study, after Friedlander (62). NMHC = non methane
hydrocarbon vapor concentration in/<.g/m3.
-67-
-------
Table 2
- Some Examples of Conversion Ratios from Atmospheric Observations
(Numbers in parentheses are heights above the ground)
Investigator
Location
Georgii (63)
Rhode (67)
Cuong, et al (66)
Grosjean &
Friedlander (20)
ACHEX (6)
Colorado
Germany
Sweden
Antarctic
South Pacific
Mediterranean
Pasadena
East/Central Los Angales
Basin
Eastern Los Angeles Basin
19 (5 km)
17 (0.8 to)
77 (2.8 ton)
6.7-14.3(0.4-2.8 km)
90
86
72
29
21
27
5.8 3.3*
6.6 0.44
31 0.60
*defined without correction for primary C contribution.
-68-
-------
of sampling, the values of the conversion ratios vary widely, but
air masses containing aged aerosol or intense photochemical smog
conditions appear to give the highest values.
The diurnal changes in fq and fN corresponding the
behavior of the reactants and products in West Covina are shown
in Fig. 5. Typically, fg remains roughly constant during the day
with a value of approximately 0.2 while fN is more variable with
a maximum in the night through mid-morning in this case. In the
Los Angeles area, f M ~ 0.06 in central areas, while it reaches
more than 0.30 farther downwind in Riverside.
The conversion ratio for carbon, f^, is indicated accord-
ing to the definition above in Fig. 5. In this case, it displays
a maximum in mid-afternoon accompanying the ozone peak, with
average values equal to or less than 0.01.
The conversion ratio f~ can be defined in a somewhat
better way without assuming arbitrarily that the organic aerosol
precursors are uniquely propertional to the non-methane hydrocarbon
concentration. In some cases, the NMHC concentration has been
measured through Cg with chromatographic analysis (6). In view
of certain later discussion, a better normalization for f^ might
be in terms of the mass of non-carbonate carbon plus the mass of
total NMHC minus the NMHC <_ Cg, fCg. Using this definition, fcg
is the same for f,,, but the former ratio is somewhat smaller
because of the conversion factor from ppm HC to ygC/m3 used.*
*For fr, NMHC are determined using NMHC as propane, or the gas density
is 2xTQ-3 gm/cm3. For hydrocarbons and vapors < Cg) a vapor density
of 4xlO~3 gm/cm3 has been assigned for purposes of the calculation.
-69-
-------
i
=L
•5.
30
zo
to
0
/
60
Z
2.Q -
J3 1.0
PZ^—
i
» I
I I 1 I I t It < I I L
1 1
/
7
22 24 . 2 4 6 8 10 12 14 16 18 20 22
TIME IPST)
Figursi 5.
Aarosol Diurnal
Pattern for W«at
Covlna» July 23-
24, 1974 (Data
frcm Hidy «it al.
(6)).
-70-
-------
Comparison between fg and fN between West Covina and
Rubidoux (Riverside) much farther eastward (downwind) of the
Los Angeles area shows that fg is roughly the same in both locations
Yet fN was found to be significantly higher at the downwind site.
Thus qualitatively, the sulfate production process appears to
generate a relatively uniform conversion over Los Angeles even
though the principal sources of SOa are on the West and Southwest
portion of the Basin. In contrast, there appears to be two distinct
modes of nitrate production, one a rapid, localized process peak-
ing with morning NOX or NOz, and one that is more slow, causing
daily accumulation of NOl far downwind of major NOX sources in
Los Angeles.
Comparison of the variation in conversion ratios with key
changes in trace gases provides additional insight into the
mechanisms of aerosol formation in the atmosphere. For example,
analysis of the ACHEX data taken in 1973 shows that fg is not
significantly dependent on relative humidity. But fg for
particles less than 0.5 ym diameter is negatively correlated
with relative humidity. Furthermore,'a strong apparent influence
of photochemistry is demonstrated for submicron sulfate formation
by the systematic increase in fg (<_ 0.5 ym) with ozone concentration,
The consistency of this trend is shown in Fig. 6.
Further clues of the sulfate formation mechanism come from
comparison of sulfate conversion and the behavior of particulate non-
carbonate carbon. The conversion ratio fg (<_ 0.5 ym) is correlated
with changes in particulate carbon, as indicated in Fig. 7. Thus
the formation of aerosols involving sulfate and organic carbon are
closely coupled in Los Angeles smog.
*For fc, NMHC are determined using NMHC as propane, or the gas density
is 2xTO~3 gm/cm3. For hydrocarbons and vapors < Cg, a vapor density
of 4x10~3 gm/cm^ has been assigned for purposes of the calculation.
-71-
-------
e
a.
o
v
0.4
0.3
0.2
0.1
0.1
WEST COVINA 7/23/73
WEST COVINA 7/25/73
WEST COVINA 7/26/73
POMONA 8/17/73
RUBIDOUX 9/19/73
O
0.2
0.3
0.4
0.5
0.6
Figure 6. Scatter diagram of the conversion ratio -f- based on particles
less than 0.5 pm vs. 2 hour averaged ozone concentration
(data from Hidy _e_t al. (6)) .
-72-
-------
30
20
E
a.
10
»
o
V
10
WEST COVINA 7/24/74
POMONA 9/16/74
O
o
o
o
o
o
O -
o
10
15 20 25
TOTAL CARBON <0.5 \im (pg/m3)
30
35
40
Figure 7. Scatter diagram of the conversion ratio f based on particles less than 0.5 pm vs.
total carbon less than 0.5 Mm diameter. (Based on 2 hour averaged filter data.)
(Data from Hidy et al. (6)).
-------
The behavior of particulate nitrate is difficult to explain
phenomenologically. It is puzzling, for example, that fN for either
the total particle mass concentration or that confined to 0.5 pm is
poorly correlated with humidity as well as N02, N0x, and ozone con-
centration. The apparent independence of f« or N02 is illustrated
by the scatter diagram shown in Fig. 8. Such results suggest that
high nitrate content of the aerosol cannot be the result of a spurious
absorption of gaseous N02 at the moist filter medium. For if this
were the case, one would expect a proportionality between fN and N02
concentration. By the same arguement, NO I does not appear to form
by a gas diffusion limited absorption of NOa in wet particles. On
the other hand, the data suggest that the conversion process to
nitrate may be dominated by other nitrogen oxide intermediates such
as NOs, N205, or HN03 which build up in the morning with the photo-
chemical activity. However, these intermediates are more likely to
be in high concentration at mid-day rather than mid-morning in smog.
The rapid transient increase in nitrate in the mid-morning does not
seem to be related closely to the systematic increase in nitrate
eastward across the Los Angeles Basin.
Estimated Rates of Conversion. The information presently
available is very limited on the rates of conversion of aerosol
precursor gases in the atmosphere. There are reports in the
literature that tropospheric residence times of S02 and N0x are
the order of a few days, while reactive hydrocarbon vapors of high
molecular weight evidently will not survive more than a day.
Some estimates of the quasi-first order rates of S02
oxidation have been reported that range from 0.1% to ^ 1070 hour
depending on many factors, including relative humidity (e.g. Calvert
(21)). Recent work, for example, of Husar e_t al. (22) near St. Louis
suggests an oxidation rate of less than 1% in the absence of photo-
chemical reactions. However, in the presence of photochemical
activity S02 oxidation rates varying from 2 to 13% hr~ have been
estimated from Los Angeles data by Roberts and Friedlander (23).
-74-
-------
60
50
40
WEST COVINA 7/14/73
WEST COVINA 7/25/73
WEST COVINA 7/26/73
POMONA 8/17/73
RUBIDOUX 9/6/73
30
20
10
O A O
A
A
0°0
A
A
A D
H
Po
D
O
17725
0
Figure 8.
0.05
0.10
O.li
u.zu
Scatter diagram of the nitrogen oxide ratio
f vs. two hour N02 concentration (data from Hidy et al. (6).
N
-75-
-------
Crude calculations based on data from the ACHEX (6) indicate similar
rates of oxidation of SOa in the Los Angeles Basin.
There is virtually no quantitative information on the
conversion rates of N0x to nitrate available. However, the Los
Angeles experience previously discussed suggests that two extremes
may be present. A rapid transient conversion may exist as a part
of the morning photochemical nitrogen oxide cycle, giving a formation
at high humidity of some nitrate at a local production rate exeeed-
ing SOa oxidation. In contrast, there appears to be a second nitrate
forming process that may be similar to the sulfate conversion rate,
or slower depending on the location of the principal sources of N0x.
This conclusion comes from the marked differences downwind in nitrate
and sulfate distribution in the Los Angeles Basin.
The rate of organic aerosol production in the troposphere is
essentially unknown. On the basis of the Los Angeles data, however,
the conversion rate should be about the same as the sulfate formation
rate because of the close diurnal correspondence between the two.
The conversion rate of certain hydrocarbons may be even faster than
sulfate since the hydrocarbon vapor sources are more evenly dis-
tributed over the city than the sulfur oxide sources. Yet the
conversion ratio for organics as presently defined is substantially
less than that for sulfate or nitrate.
Summary
The existing knowledge of the phenomenology of aerosol
formation in the troposphere relies heavily on the photochemical
smog observations in Los Angeles. A summarizing list of key
observations is given in Table 3.
-76-
-------
Table 3
Some Properties of Tropospheric Aerosols Relevant to Their Formation Chemistry
Phenomenon
1- Production of new
particles
2- Formation of
nuclei in filtered,
irradiated air
3- Haze formation &
optical scattering
copious nuclei generated
strong diurnal changes accom-
pany smog evolution (Los
Angeles); correlation of
scattering with maximum daily
ozone
4- Particle growth predominant
in 0.1-1.0 m diameter
range
5- Role of liquid
water in particle
growth
strong evidence supporting
role in growth - urban
aerosols generally hygro-
scopic
Non-Urban
1imi ted
uncertain
widespread regional
increase of haze over
continents below 3-5 km
altitude
predominant
suspected; not proven
universally
6- Key Chemical Components
.Carbon-oxygenated
organics
.Sulfate
.Nitrate
highly variable constituent,
but normally present
universally present
major constituent, sometimes
produced by reactions; con-
fined to submicron particles
universally present; large
fraction in submicron range;
submicron particle production
depends on ozone; production
correlated with organics; not
correlated with humidity
universally present; highest normally present
concentration in photochemical
smog or in high humidity with
N02 (not correlated with N02
or S0,? concentration); some-
times in larger particles than
sulfate; spatial distribution
in Los Angeles differs from
sulfate
-77-
-------
Table 3 (continued)
Phenomenon
...Ammonium
7- Conversion Ratios
Urban
normally present; often
insufficient to "neutralize"
acid anions
f<. sulfate sulfur to * 0.1 (within hours of S0?
total sulfur
ft. Inorganic ni-
trate to total
fc
oxidized nitrogen
organic carbon to
carbon in non-
methane hydro-
carbons
source); observed 2 hr.
average maximum ~ 0.6 in
aged air without new sources
of S02 rate of formation 1*-
15% hr-1
«-0.1 (hours from sources);
observed 2 hr. averaged maxi-
mum /« 0.4 in smog
0.005-0.020 (based on non-
methane hydrocarbons) in
Los Angeles
Non-Urban
normally present; usually
insufficient to "neutra-
lize" acid anions
sometimes approaches unit
in upper troposphere
uncertain
unknown
-78-
-------
B. The Stratosphere
The principal evidence for aerosol formation in the strato-
sphere comes from the discovery in the early 60's and the subsequent
confirmation of the dominance of sulfate in aerosols samples taken
at high altitude. Observations have suggested the presence of a
maximum concentration of particles greater than ^ 0.1 ym diameter
at altitudes of about 20 km. The "model" composition for the
stratosphere aerosol is shown in Table 4. These data, based on the
work of several investigators including Junge e_t al. (24) and Lazrus
e_t al. (25), show that sulfate is by far the major component of
aerosol at altitudes of ^ 20 km.
In contrast to the tropospheric material, both nitrate and
organics, as presently reported, are essentially absent from the
stratospheric particles. In the latter case, this is not surprising
because it is expected that condensable organic vapors would have
to be high in molecular weight and probably would not survive sub-
stantial degradation above the tropopause. The absence of nitrate
in the aerosol evidently is related to the high vapor pressure of
HN03 relative to H2SOi, (see below); NHt, or cations other than N
in large quantities are not expected to be present at high altitude.
Apparently, the most acceptable steady-stage theory for the presence
and profile of sulfate comes from (photochemically related) oxidation
of S02 in situ at high altitude (e.g., Friend e_t al. (3) , Junge (26),
Harrison and Larson (27)).
The details of the physical mechanism of sulfur oxide
transition into particles are unknown in the stratosphere. The
evidence from size distribution observations combined with nuclei
and optical data suggests, however, that stable sulfate accumulation
takes place mainly in existing particles rather than through formation
of new nuclei. There is a general depletion in nuclei concentration
with altitude, but the particle number in the range ^ 0.1 ym diameter
varies a much smaller amount with height (e.g., Hidy e_t al. (28)).
-79-
-------
Table 4
Model Chemical Composition of Stratospheric Particles at 20 km Altitude
During the Period 1969-1973. Source: Hidy et al. (28).
Concentration
in air
><.g/m3
Substance (ambient)
Sulfate
BasaHa
(calculated from Si)
NHj
NO' b
NOj
Na
C1
8r
0.6C
0.05
0.005
0
0
0.01
0.04
0.002
Observed concentration range
^9/m3
(ambient)
0.01 - 4 (Agung peak)c
0 - 0.7
0 - 0.01
0
0
0.001 - 0.05
0.002 - 0.09
0 - 0.003
Total
0.71
0.01 - 1
aThis includes all other components of the basalt, e.g., Al, Ca, Mg.
Particulate NOZ as contrasted with HNO, vapor.
cFor eight-year averaged data at 18 km from Castelman e_t a]_. (65).
-80-
-------
Based on the few existing measurements, the number mean diameter
for stratospheric aerosols is approximately 0.5 ym, and the total
mass concentration is estimated to be 0.1-1.0 yg/m3 at 20 km.
The rate of oxidation of SOa above the tropopause is un-
known and very difficult to estimate because no data on S02 con-
centrations above 10 km are available. Rough estimates of the
reaction time for S02 based on known S02 oxidation chemistry now
range from months (Harker (29)) to many years (Friend e_t al. (3)).
The former is accepted by most chemists as the more likely value.
3. MECHANISMS FOR AEROSOL FORMATION
The chemical and physical processes that lead to aerosol
formation in the atmosphere are probably very complex, with no
single mechanism dominating the phenomenon. Unlike the chemical
laboratory, the atmosphere is very "dirty" in that it contains sub-
stantial amounts of airborne particles which are continuously inter-
acting with trace gases. Water as vapor and in the condensed
phase likely plays a critical role in aerosol evolution, both in
growth and removal from the atmosphere. Thus, in elucidating
atmospheric aerosol behavior, one must account for both homogeneous
processes of precursor formation and heterogeneous interactions.
A. Physical Constraints.
Regardless of the chemistry, certain physical constraints
exist on aerosol-gas interactions. These restrictions relate mainly
to (a) the thermodynamic stability of condensed particles of small
diameter, (b) particle nucleation and condensation processes, and
(c) gas phase diffusion limited rates combined with absorption or
adsorption on particles.
-81-
-------
Particles must be close to or at equilibrium with respect
to the surrounding vapor to exist in air for any length of time.
Thus, the partial pressure of condensed species on particles
essentially must be less than or equal the saturation vapor pressure
at atmospheric temperature for stability. This presents no great
problem for most inorganic salts or for sulfuric acid, even at parts
per billion concentration in the gas phase. However, it places a
severe constraint on the ability of HN03 to exist as a pure compound
or as an acid diluted in water.* The vapor pressure data reviewed
by Toon and Pollack (30), for example, indicates that HN03 has a
partial pressure over aqueous solutions more than 100 times higher
than concentrated H2SOit making nitric acid far less stable in the
atmosphere than sulfuric acid.
The requirements of low vapor pressure are particularly
important to the stability of organic aerosols. The bulk of the
organic vapors in the atmosphere that have been identified are in
range of carbon number less than four. Even if such materials
reacted to form oxygenated materials, review of vapor pressure data
(Hidy (10)) suggests that only material of carbon number much
greater than Ce would be thermodynamically stable in the condensed
phase at the concentrations of ^ 1 ppb.
Accumulation of condensed material as aerosols in the atmos-
phere may take place by two basic processes: first by condensation
of super-saturated vapor or by chemical reaction leading to spontaneous
formation of new particles; second, by condensation, absorption or
reaction on existing particles. In the latter case, the chemical
reactions actually may take place on the surface of, or within
existing particles.
* Vapor pressure data for HN02 or HNOs over solutions of ammonium
sulfate or sulfuric acid are not readily available so that the
volatility of these acids over multicomponent aqueous media is
unknown.
-82-
-------
For condensable precursors, the rate of particle formation
may occur by homogeneous nucleation or by hetrogeneous nucleation.
It is generally accepted, but not proven, that heterogeneous pro-
cesses are most likely in the atmosphere because of the large
number of existing nuclei. One can readily see, however, that growth
by condensation is limited by the rate of diffusion of vapor to
the surfaces of nuclei. If conditions exist in which aerosol pre-
cursors evolve chemically at a rate exceeding the diffusional
transfer, then supersaturation could build up to high enough levels
to permit heteromolecular (homogeneous) nucleation of new particles.
It can be shown that HzSOi, can undergo heteromolecular nucleation
at atmospheric concentrations in the absence of nuclei, but it is
unlikely that HN03 could nucleate because of its relatively high
vapor pressure. So little is known about the products of aerosol
forming organic reactions and their relevant physical properties
that nothing can be said about the importance of homogeneous
nucleation in this case.
The fact that few new particles are observed in cities away
from combustion sources or in rural areas, particularly in highly
reactive atmospheres like Los Angeles make it unlikely that
heteromolecular nucleation is a widely important formation process
in the troposphere. The general decrease in nuclei concentration
with altitude with more nearly constant large particle concentration
suggests that new particle formation normally plays a small role
at high altitude, too. Yet, this conclusion is by no means well
established in the case of sulfuric acid behavior above 10 km
altitude.
Growth of particles by accumulation on existing particles
can be classed as two broad processes. If the precursor is super-
saturated, growth will occur at a vapor diffusion limited rate,
which depends on the supersaturation, the temperature, the particle
size, and the accommodation coefficient at the surface. The
proportionality of particle size changes with the ratio of particle
diameter and mean free path of the suspending gas. At one extreme,
the growth depends on volume to the 2/3 power; at the other, growth
is proportional to volume to the 1/3 power. When the precursor is
-83-
-------
unsaturated, growth still may take place by irreversible absorption,
or by chemical reactions in the particle. In this case, the rate
law should be proportional to the particle volume if the reaction
is uniform throughout the particle. If the formation of material
is limited by reactions in the particle, then the conversion ratio
should not be dependent on the gaseous precursor concentration.
There is insufficient data available to determine the rate
law or physical mechanism most likely to predominate in atmospheric
aerosol growth. However, there are clues to differences in the pro-
cesses from the Los Angeles data. The shape of the particle volume
number distribution of tropospheric aerosol is such that the 1/3
(diameter) and 2/3 (surface) moments are concentrated in the sub-
micron fraction, while the first moment (volume) is weighted toward
larger particles. Thus the observed accumulation of sulfate and
organic carbon on the small particles in smog suggests a surface or
vapor diffusion controlled process. In contrast, the collection of
NO 3 in larger particles and independence of N02 may indicate a
volume controlled reaction in particles. The shift in NOa to
smaller particles between the east central and the eastern parts
of the Los Angeles Basin is of interest in this case and remains
unexplained.
It is of interest that the influence of thermodynamic
equilibrium must enter the growth process of particles. If the
radius of the particles is too small, the partial pressure condensable
species can increase significantly by the influence of radius of
curvature. From Kelvin's relation, examination of values of surface
tension for a range of materials suggests that this effect will
constrain growth to particles greater than 0.05 - 0.1 ym diameter.
This appears to be consistent with available observations of
atmospheric growth and the distribution of secondary chemical
components.
-84-
-------
B. Chemical Reactions for Precursors.
Physical processes of phase change, absorption and
stabilization restrict the maximum rate of formation and the
classes of materials expected in aerosols generated in the atmos-
phere. Yet in many instances, the rate controlling process in the
production mechanism may be the precursor formation. In the case
of any of the three secondary constituents, there are a variety of
possible chemical mechanisms that may occur in the atmosphere.
It is likely, in fact, that different mechanisms take place in
parallel, or complement one another so that no one process is the
dominant one. One example of this is the difference in behavior of
sulfate in Los Angeles air, as indicated by the changes in sulfate-
particle size distributions.
Sulfate Reactions
The oxidation reactions of SOa have been reviewed by several
investigators including Bufalini (.31) and Calvert (21) . There are
more than a dozen sulfate forming reactions that may be relevant
to atmospheric processes. These can be grouped in terms of homo-
geneous gas phase reactions and heterogeneous reactions involving
suspended particles. A summary of the two groups is listed in Tables
5 and 6. The homogeneous reactions are broken down into sub-
categories whose end products are S03 or R02S02, and ROS02. Although
the rates of reaction of these species with water or other species
have not been reported, it is assumed that the reactions with
water are fast to form HaSd* and not the rate determining step in
SOi* production.
Listed in Table 5 are .three classes of homogeneous reactions.
The first consists of inorganic oxidation mechanisms to form S03,
while the second involves organic radical oxidation agents generating
S03 or R02S02- The third group of reactions forms ROS02 species.
All of the reactions listed are exothermic and are favored thermo-
dynamically. However, the first five reactions have been considered
severely rate-limited on the basis of available rate data (e.g.
Calvert (21)). The remaining reactions listed appear to have rates
-85-
-------
Table 5
Escinatad Rates of Theoretically Possible Homogeneous Renoval Paths for
SO? in a Simulated Polluted Atnoaohere (froa _Calvarc
A. Inorganic Reactions Forming SO^
(1) SOj + 1/2 02 + Sunlight * S03
(2) 0(3P) + S02 -t- K * SOi + H
(3) QJ + S02 •* S03 + 02
(4) NOj -1- SOj * SOj + NO
(5) NO, + S0: •» SO, •»• N02
(6] N205 4- SOj * SOs + N20%
B. Organic Reactions Forming 30^
(7) CHi— CH2 + S02 * SOj -h 2CH20
(8) »CHiOO- 4- S02 * SOj + CH20
C5j-0 +• 0 + SOz * SO, + CH20
(9) H02 + S02 * HO + SOj . (a)
* H02saj (b)
(10) GJ302 -t- S02 * CHjO + SO, (a)
* CHi02SOj (b)
C. Reactions Forniag HOS02 or ROS02 Radical
(11) HO + S02 * HOSOJ
(12) CH30 + S02 •* CHjOSOj
kcal/mole
-24
-83
-58
-10
-33
-24
-81
- -117
- -85
-19
< -25
-30
< -25
r -82
- -73
Approx.
Rate, Z
oer hr.
< 0.021
0.014
o.oo •
0.00
0.00
0.00
< 0.4-3.0
< 0.4-3.0
0.35
1
- 0.16
7
- 0.23-1.4
- 0.48
TOTAL POTENTIAL RATE OF CONVERSION OF S02 TO S03 (OR
SULTATES) IN MODERATE..S1S3G - 1.7 - 5.5Z per hour
-86-
-------
Tab la 6
Heterogeneous ReaeCiena -to Forn SulCace
A*
(13) S02 +• H20(i) ^H
5=S H -f HS03
(13A) .2S03 * 02(aq) * 2S04
(13B)
Won-Aoueou3
(14) S02 (ads) -)
HZ0 (ada) C * Carboa(s) * HjSO,, (ads)
Oj (ads) J
-87-
-------
that are sufficiently rapid to be of importance in the atmosphere,
at least for polluted air with active photochemical processes.
For comparison, estimates of the fractional conversion rate of S02
to SCK~ is given in the Table. These results are based on Calvert's
computer simulations. These were based on an estimation of reactive
intermediates after a 30-minute period of sunlight irradiation at
zenith angle of 40°, with initial concentrations in ppm of N02 -
0.025, NO = 0.075, C^Hg = 0.10, CH20 = 0.10, CH3CHO = 0.06, CO =
10, CtU - 1.5, R.H. » 50% at 25°C, and S02 = 0.05-0.1.
The reactions (7) and (8) correspond to the interpretation
of Cox and Penkett's (32) observations that S02 is oxidized at
appreciation rates in the dark in ozone-olefin-air mixtures. The
higher rate of 3% per hour was found for cis-2-pentene, while the
lower rate of 0.470 hr.~ was found for propylene. Cox and Penkett
suggested that either of two intermediates were involved in the S02
oxidation reaction. These are the ozonide reaction (7) or the
zwitterion intermediates from the 03- olefin reaction. The
zwitterion has a diradical character and may be illustrated as
reaction (8). Calvert's calculation of olefin-Oa intermediates such
as those .in reactions (7) and (8) do not favor their importance
as oxidizing agents. However, other radical species from the ozonide
or zwitterion intermediates may be of interest, including those
summarized by reactions (9 - 12). These classes of reactions may
well account for S02 oxidation in the ozone-olefin mixtures.
Recently, the potential importance of radical addition
reactions in the third class listed in Table 5 for S02 oxidation
become more fully appreciated. Such reactions are exemplified in the
series (9) to (12).
The rate of the H02 reaction has been made by Davis e_t al.
(33). With such data, the fractional rate of S02 disappearance may
reach ^ 170 in a moderate photochemical smog. Assuming that the rate
of the CH302-S02 reaction is the same as H02-S02 reaction, Calvert has
-88-
-------
estimated that the former will contribute a fractional oxidation
rate of 0.2% hr"1 in moderate smog.
The OH radical-S02 reaction (11) appears to be of particular
importance in the lower stratosphere where OH concentrations are
estimated to be high (27,29). Calvert (21) has estimated the typical
reaction rates for OH or CHa radical S02 reactions on the basis of
analogies to reaction rates of CH3, C2HS and S02 reactions on the
basis of analogies to reaction rates of CH3, C2HS, and CFH2CH2. These
rates are listed as 0.237,, hr"1 and 0.5% hr"1 respectively at the
lower range.
Measurements of the rate constant for the OH + S02 + M •>
HOS02 + M reaction are emerging from recent fundamental studies.
Hamilton (34) has summarized the preliminary values of the rate con-
stant for this reaction. At ^ 300°K the rate constant for this re-
action ranges from Wayne's value of ^ 7 x 109 M"1 for 1 atm Ar to
(35) value of 1.1 x 108 M"1 sec"1 for 18 torr N2 and 20 torr H20.
Calvert (21) has used 1.1 x 109 M"1 sec"1 to estimate the upper
limit of 1.4% hr"1 listed in Table 4.
This radical addition products, such as HOS02, should react
rapidly with other species to generate sulfuric acid, peroxysulfuric
acid, alkylsulfates and mixed intermediates such as HOS02ON02. Any
of these ultimately should lead to sulfate in the presence of water.
Summing all of the known homogeneous reactions for S02
oxidation, it is possible to rationalize a theoretical rate of sulfate
production in the range 1.7-5.5% hr" l for moderate photochemical smog
conditions. However, such rates are clearly highly dependent on the
presence of unstable intermediates at relatively high concentrations.
This is by no means a universal condition in non-urban air or in
cities with minimal photochemical activity, as measured for example
by ozone levels.
-89-
-------
For conditions where photochemically induced homogeneous
reactions cannot be important, the heterogeneous processes must
be considered. The known reactions are listed in Table 6. These
have been categorized as aqueous and non-aqueous reactions. The
class of reactions that have been used most frequently to explain
high S02 rates in the presence of liquid water containing aerosols
in the system involving S02 absorption in water followed by oxidation
by dissolved 02 to form sulfate. Catalysis of the oxidation by
heavy metal salts such as Mn++ ion can realize rates of oxidation in
excess of 170 hr"1 in clean water solutions (e.g., Johnstone and
Coughanowr (36) , Matteson e_t al. (37) . The absorption of S02 can
be promoted by the buffering effect of simultaneous absorption of
ammonia. Scott and Hobbs (38) have shown that the aqueous S02
oxidation process can be enhanced significantly by ammonium ion.
Indeed the estimates and experiments of Miller and dePena (39) and
Corn and Cheng (40) suggest that rates of S02 oxidation can be
achieved in fog approaching 10% hr~l.
It is well known that ozone is quite soluble in water.
Therefore, one expects that ozone absorption with S02 would con-
tribute to significant oxidation of S02. Experiments of Penkett
(41) have shown that oxidation of S02 in air at 7 ppb absorbed in
water droplets with ozone, present in surrounding air at 5 pphm, can
be as large as 13% hr'1. Thus, foggy or cloudy air mixed with
photochemical smog, such as sometimes observed along the Pacific
Coast, could well be an important medium for S0i»~ formation.
Furthermore, such an aqueous mechanism could be significant at middle
altitudes over continents even at background ozone levels.
The reported rates of S02 oxidation in clean water droplets
must be considered maximum. It is questionable whether they ever
can be achieved in the atmosphere since such aqueous reactions
have been shown to be suppressed significantly by organic contaminants
The work of Fuller and Christ (43) and later of Schroeter (44) has
indicated the aqueous absorption of S02 and its subsequent oxidation
is reduced as much as an order of magnitude by dissolved organic
acids or alcohols. This type of material is known to be present in
-90-
-------
the atmospheric aerosol sampled at the ground, so one can expect that
the aqueous oxidation will probably be most efficient in relatively
clean conditions of clouds well away from the earth's surface.
The heterogeneous mechanisms of S02 oxidation in the absence
of liquid water are poorly understood. However, the recent work of
Novakov et al.(42) has shown that SCU~ can be produced rapidly on
the surface of carbon particles suspended in water vapor and air.
These workers have observed that significant amounts of SCU~ can
be found on carbon particles generated by combustion of hydrocarbons
in ppm level SOa enriched air.
It is difficult to assess the significance of carbon or
organic particles for the S02 oxidation in the free atmosphere.
There is little doubt that absorption of S02 on carbon particles
freshly generated by combustion can provide a surface catalyzed
oxidation medium. Indeed experiments such as those of Yamatnoto et al,
(45) have shown that SO2 oxidation can be as high as 30% hr l on
activated charcoal particles 5 mm in diameter. Their work also
indicates that this rate is strongly reduced by sulfuric acid
collection in the mocropores of the charcoal. The work of Yamamoto
et: al. further emphasizes that such a heterogeneous oxidation
mechanism depends on a variety of factors, ranging from grain size
of the carbon, temperature, concentration of S02, H20 vapor, and
oxygen, as well as the micropore structure of the particle surface.
It would seem that oily, gummy, wet particles collected from the
atmosphere would be poorly suited for non-aqueous reactions to
form sulfate since their micropore structure would be minimal.
Yet such a mechanism cannot be ruled out from consideration at this
time.
-91-
-------
Nitrate Forming Reactions
Like the production of sulfates, nitrates in atmospheric
aerosols can be formed by a wide variety of homogeneous reactions
as well as heterogeneous reactions. The pathways of nitrate
generation are less well understood than sulfate, but it is likely
that they are interrelated at least in some circumstances.
Since both nitrous acid and nitric acid are much more
volatile than sulfuric acid, it does not appear possible that
particulate nitrogen oxide species will exist in the atmosphere
in pure form as acids. Thus the production of nitrate must involve
formation of a condensable species such as NHitN03 in the gas phase,
or the absorption of a nitrogen oxide constituent for particles
followed by stabilization through chemical reaction. Such hetero-
geneous processes, of course, can take place in an aqueous or non-
aqueous medium.
The precursors for particulate nitrate formation are
summarized in Table 7. These are classified in terms of (a)
important nitrogen oxides, (b) volatile acids HONO, and HOMO2, and
(c) gaseous nitrates.
Because nitric oxide is relatively insoluble and nonreactive
with water, the important nitrate forming atmospheric oxides of
nitrogen are believed to be N02, N03, and N205. These species are
formed mainly in the atmosphere by the well known "smog" reactions
(16 - 20) and are not emitted primarily from material or anthropogenic
sources.
With the nitrogen oxides coexisting with water vapor and
sulfuric acid, the volatile nitrous and nitric acids can be formed
via the reactions (21 - 24). The mixed intermediates involving sulfuric
-92-
-------
Table 7
Reactions Potentially Involved In Nitrate Formation
Species
A. Nitrogen Oxides
(15) Oj +• NO * N02 + 02
(16) 0 + M + NO ••• H02 + M
(17) ROt +• HO * N02 + R1OH
(18) 0, * NO* * NO, 4- 02
(19) NOs +
B. Volatile Acids
(20) N205 + HjO •* 2 HONOz
(21) HO •»• N02 +• M •» HON02 +
(22) NO •»• N02 +• HjO * 2 HOfO
(23)
(24)
HOS020 + NO -» HOS02 ONO
-»• H20 •»
HOS020
* HOS02 OM02
•«• H20 -
C. Caseous Nitrate
(23) NH, + HONOj
(26)
(N20S
(N02"
Rate Constants (ppri"
21.8
2X10-5
TO'5 (M - 1 atn
HONO
-f HON02
- 10
10
,-6
ONO
*Typical for smog reactant 'concentrations.
-93-
-------
acid are of interest as they link the N0x and S0x chemistry. Once
the volatile acids are formed, they may react with ammonia in the
gas phase to form, for example, NH^NOs (reaction 25). The nitrogen
oxides also react with radicals such as RO to form organic nitrates
and nitrites including peroxy acetylnitrate (PAN). Reaction (26)
has been hypothesized by Calvert (21) on the basis of analogy to
the NH3-HC1 reaction. For concentrations of NH3 and HON02 at ^
1 pphm, Calvert's (21) work suggests that the ammonia reaction may
represent a significant removal path of HON02 to form aerosols.
It is possible that condensable organic nitrates are formed
via reaction (26) in the gas phase. Certainly the observation of
such materials from smog chamber experiments (O'Brien e_t_ al. (46))
would provide some evidence for such cases. However, it is known
that volatile organic nitrates such as PAN readily hydrolyze in an
aqueous medium to form nitrite ion (Reaction 31). Thus the presence
of such compounds resulting from gas phase reactions could lead to
particulate nitrate after stabilization with ammonium ion or another
cation.
Of the gas phase reactions involved to form nitric acid
vapor, (20) and (21) are believed to be most important. Using
currently accepted rate constraints for these reactions, Calvert
has estimated that
R2Q - 0.5-2 x 10"5 ppm min"1
&2i ~ 2-6 x 10"5 ppm min"1
in moderate photochemical smog. Thus, it would appear that reaction
(21) is of principal importance for HON02 formation in smog. The
rate of conversion of N02 to HON02 by this reaction should be in
the range of 2-8% hr~1 for the conditions used in the calculations
in Table 5 (21). With absorption in wet aerosols and neutralization
with ammonium ion, this is not unreasonable for the upper limit of
estimated nitrate formation rate in Los Angeles smog in the morning
"peak" condition. Once the nitrogen oxides are present or the acids
-94-
-------
begin to form, the interaction of these species with moist aerosols
can take place. Some potentially important reactions are given in
Table 8. Here the complexities of the aqueous reactions of N0x
become apparent. All of the nitrogen oxides contained in the atmos-
phere will react with liquid water to form traces of nitrate and
nitrite. These reactions are generally reversible, however, so that
the anions must be stabilized by a base such as NHi* . In the presence
of dissolved oxygen or ozone, nitrite ion can be oxidized to nitrate
in analogy to the aqueous sulfate formation reactions.
None of the nitrate reactions has been studied with atmospheric
application in mind. However, there is a variety of information in
the chemical literature dealing with NOX absorption in water. There
is ample evidence, for example, from studies reported by Nash (47)
and Borok (48) that N02 at trace levels in air is readily absorbed
in aqueous solutions. The efficiency of absorption varies widely,
however, depending on the acid-base content of the solution.
If significant quantities of concentrated sulfuric acid are
formed in atmospheric aerosols, nitrate formation via absorption of
N02 to form nitrosylsulfuric acid, HNOSOi*, may be of interest.
The rate of absorption of N02 in sulfuric acid is fast, but
the efficiency appears to be relatively low, according to Baranov et:
al. (49). Again it appears that ammonia absorption or another base
has to be involved to drive the equilibria to nitrate production.
Thus, in all of these aqueous reactions, nitrate formation may be
limited by the concentration of ammonia rather than by any of the
nitrogen oxide species.
Organic Particle Formation
Of the three major contributors to secondary aerosol pro-
duction, the least is known about mechanisms for the organics. These
processes have not been identified with any certainty yet. It is
-95-
-------
Tab la 8
Aquaoua Reactions of Nitrogen Oxidea
(27) N20S + H20(i) •* 2H* 4- NO,"
(23) NO 4- N02 + H20(£) * U+ + N02~
(29) 2N02 -f H20(A) ^H"1" + NO," + HONO
HOMO + OH" •*• H20 -f N02~
(29») 2H02" 4- 02 (aq) * 2 N03"
(29b) N02" 4- Oj (aq) -»• N03~ + 02
ENOSO* 4- H20(A) ^^EN02 4-
3HN02^±HN03 4- 2NO 4- H20
<31) RON02 4- HzO(I) •* H4" 4- N02" 4- I^OH
-96-
-------
possible that organic materials can be polymerized in sulfuric
acid solutions, for example. However, it is known that such re-
actions generally take place in very strong acid above 90% con-
centration. At equilibrium with water vapor in the lower atmosphere,
HaSOi* should not exceed 4070 concentration in water. Another possible
class of important organic aerosol reaction involves the attack of
ozone on olefins.
The ozone-olefin reaction mechanism remains a subject of
controversy. One approach that can serve to illustrate the broad
features of the reaction as related to aerosol formation is Story's
model as shown below (see next page). The ozone adds to the olefinic
band forming zwitterion, which can result in an epoxide or a molozonide
These species can follow one of three added paths to generate con-
densable peroxides or polymeric material. Addition of water in the
system can provide added paths for decomposition of the peroxides
to form acids or other oxygenated species including alcohols and
aldehydes. All of these materials have been identified in atmos-
pheric aerosols.
The intermediates identified by Cox and Penkett (32) appear
in the early stages of the ozone-olefin reaction and offer a
potential link between the sulfate chemistry and the organic aerosol
chemistry.
IM X
It R U R * 0
/A / (STAUOINGER
EMXIOE * 02 / KOLOZONIDE}
. I1
POLYMERIC MATERIAL
0
+ OIPEROXIOE
CHjCHO OR CYCLIC PEROXIDE-
-97-
-------
The recent experiments of Burton ejt al. (50) have shown
that the ozone- (terminal) olefin mechanism for aerosol production
is initia-ly first order in ozone and olefin. The constant relat-
ing the rate of maxx concentration change increases with carbon
number of the olefin. The significance of this effect is illustrated
in the data shown in Fig. 9. Here the experimental variation in
generation of condensable material from the reaction between ozone
and several terminal olefins is shown for an ozone concentration of
12 ppm. Significant production of aerosol is observed over a re-
action time of 54 seconds mainly for the larger molecular weight
species.
The empirical rates of production of aerosol from dry air
ozone-olefin mixtures are shown in Table 9. Extrapolation of these
data to olefins at the ppb level and ozone of 10 pphm suggests a
production rate of tenth's of ug/m3-hr. of organic material, which
is an order of magnitude lower than projected from atmospheric
observations in Los Angeles. Thus, either the estimate of total
precursor concentration is too low, or the ozone-olefin mechanism
cannot explain the atmospheric processes. It is also interesting
that the experiments of Burton e_t al. (50) have demonstrated that
aerosol formation in such mixtures using 10 ppm 03 and 10 ppm hexene
in dry air is heavily influenced by the addition of water vapor.
Furthermore, traces of butyraldehyde strongly inhibit the formation
of aerosols in this system.
This work suggests that the ozone-olefin reactions or oxygen
atomolefin reactions may be important. These classes of reactions
involving high molecular weight olefins are known to produce poly-
merized oxygenated species that will condense out at ppb vapor
concentrations. For carbon number greater than six, yields of
polymerized material should exceed several percent of the vapor
concentration. Such reactions are likely to be promoted on particle
surfaces when the surface of large particles could act as a pre-
concentrator for the olefin or air intermediate species.
-98-
-------
r » 54 sec.
03 » 12 ppm
O » Propylene C;
• * 1-Butene
° » 1-Hexene
& * Cyclohexene
o = 1-Octene
= 1-Decene
1 1-Dodecene
ppm OLEFIN
Figure 9. Mass Concentration of Aerosol Formation for Olefin-Ozone
Reaction as a Function Olefin Concentration.
-99-
-------
Table 9
Empirical Rate Relationships for Ozone-Olefin Reactions
Qlefin Rate (/*q/cn3-sec-ppm2)
Propylene 1.0 * 10~2 (- 0.3 x 10"2)
1 . butene 1.7 x 10'2 (- 0.5 x 10"2)
1 - hexene 6.0 x 10'2 (- 2.o-x 10'2)
1 - octane 18.0 x 10"2 (- 6.0 x 10"2)
.1 - decene 37 x TO"2 (-18 x 10"2)
1 - dodecene 4.0 (- 1.0)
cyclohexene 18.0 x 10'2 (- 6.0 x 10'2)
-100-
-------
C. Application of Laboratory Simulation.
The review of potentially significant reaction mechanisms
to form atmospheric aerosols could be an endless challenge for the
chemist. Basic laboratory experiments to determine rate constants
and important reaction steps play a key role in eliminating many
candidates. However, the complexities of the chemical interactions
in atmospheric processes cannot be elucidated by fundamental
experiments alone. The predictions of behavior of reactive mixtures
should be consistent with atmospheric processes or with atmospheric
simulation experiments. Simulation of atmospheric phenomena using
the study of controlled laboratory prototypes is a well known
technique, particularly as applied to fluid dynamic processes.
However, the simulation of atmospheric chemical phenomena is less
well established. With the possible exception of the work of Friend
e_t aJL (3), only simulations of the lower atmosphere, more specifically
urban air, have been attempted. The principal method that has been
used by many investigators is the static reactor or smog chamber
approach, where air mixtures containing reactive contaminants are
studied over a several hour period; Such studies are known to
duplicate qualitatively, at least, many of the features of urban
photochemistry to form ozone and other smog products. The principal
deficiencies as simulators include the low volume to surface ratio
compared with the atmosphere, and the uncertainty in the roll of
wall reactions.
A limited number of smog chamber experiments have been under-
taken to investigate aerosol formation in smog. Aside from the
early studies reviewed by Leighton (1), these include those of
Groblicki and Nebel (51), Wilson and colleagues (52, 53, and 54),
O'Brien e_t a_l. (46), and the CALSPAN group (53). The results of
these studies generally confirm that significant aerosol formation
in simulated smog requires an induction time until ozone begins to
build up, reflecting the time required for build up of reactive
intermediates such as HO, H02, R02, HON02, etc. from chain reactions
-101-
-------
involving olefinic hydrocarbons. Aerosols found in chamber
experiments are composed of oxygenated organic material including
carboxylic acids and organo-nitrates. If S02 is present, sulfate
also is found. Much of the nitrate generated appears to be
accumulated on the walls of the chambers.
The presence of S02 in irradiated N0x-hydrocarbon-air mixtures
enhances aerosol formation, but does not significantly influence the
ultimate ozone level realized in the chamber. With a suitable
choice of rate constants for reactions of HO and S02, and HS03 to
form S0n~, Calvert (21) has indicated that computer modeling can
duplicate many features of the Battelle chamber experiments (e.g.,
Wilson and Levy (52)).
The smog chamber studies have been useful in classifying
the reactivity of hydrocarbons for photochemical aerosol production.
With S02, the experiments generally indicate that the terpeniod
compounds such as °<-pinene are highly prolific aerosol formers
followed in rough order by di-olefins, cyclic olefins, high molecular
weight terminal olefins and aromatics. Without S02, the work of
O'Brien e_t aJL. (46) suggests that only diolefins and cyclic olefins
of carbon number six or greater will generate significant quantities
of aerosol at atmospheric pollutant concentrations for NO .
Unfortunately such compounds have not been identified in air or in
sources like motor vehicle exhaust. Their work also suggested
that N02 tended to suppress aerosol formation. In another study
using a smog chamber, Ripperton e_t al. (56) has given evidence for
the importance of ozone attack on an olefinic bond in aerosol
formation, particularly in "natural" production from terpenoid
compounds.
To minimize the interactions between gases and the contain-
ing walls, another approach to simulation of aerosol forming re-
actions has been attempted recently. Lipeles e_t a_]^. (57) have reported
a steady flow reactor technique in which the initial reactions of
aerosol formation in irradiated gas mixtures are studied at reactant
-102-
-------
concentrations approaching those in the atmosphere. The initial
experiments employing the flow method have demonstrated the
nuclei can be generated in irradiated gas mixtures containing traces
of NOz and olefins (
-------
4. INTERPRETATION OF PRESENT KNOWLEDGE
The results of a large number of laboratory experiments
and their interpretation provide strong evidence for the potential
importance of several mechanisms to produce atmospheric aerosols.
The processes are clearly complicated with many possible interacting
avenues for formation of sulfate, nitrate, and organic material.
However, the results of smog chamber simulations leave little doubt
that photochemistry must play a role in aerosol evolution. At the
same time, heterogeneous processes involving liquid water droplets
or moist aerosols certainly are key factors, at least in the tropo-
sphere. The evidence for non-aqueous heterogeneous reactions is
less compelling at this time, but they cannot be ruled out as factors
in the particle-gas interactions expected to take place in the
atmosphere.
Until recently, the details of the changes in atmospheric
aerosols over short time periods were unknown so that meaningful
intercomparisons between the atmospheric prototype and our mechanistic
models could not be undertaken. With the results of new field
experiments such as the ACHEX in hand, the significance of different
mechanisms can now be tested, at least for urban air.
A. Sulfate Formation.
In the case of photochemical smog in Los Angeles, comparison
of the observations with expectations of chemistry suggest
qualitatively that several mechanisms probably play a role in sulfate
and nitrate formation. The evidence is strong that submicron
sulfate particles are produced photochemically as a part of the smog
mechanism. Such formation is coupled closely with the changes in
organic aerosol during the day and is well correlated with the delay
required to produce the high levels of oxidant.
-104-
-------
The reactions that appear to be most significant for
submicron sulfate are homogeneous in nature, either involving a
dark reaction of ozone and olefins, or the groups of free radical
reactions such as OH attack on SOa• The accumulation of sulfate on
the submicron particle fraction.during the day is consistent with
a vapor diffusion limited process with a condensable or reactive
precursor such as HaSOi, or HSOs radical. An alternate mechanism
of catalytic SOz oxidation on carbon particles may be significant,
but in our opinion it is a less likely mechanism than the photo-
chemically related processes, or reactions involving an aqueous medium.
The behavior of sulfate in the larger particles shows a
poorer correlation with ozone and carbon in Los Angeles air, but the
correlation with relative humidity change is also weak. It is likely
that aqueous sulfate formation mechanisms play some role in total
sulfate behavior in Los Angeles, particularly at night or in
conditions of fog or haze with high liquid.water content. Incidentally,
the liquid water content of the Los Angeles aerosol varies with
total mass concentration more strongly than relative humidity (61).
The inhibiting influence of organics in aqueous urban aerosols
is uncertain, but may be a factor in the apparent variability of
sulfate generation from day to day in-Los Angeles.
Extrapolation of the experience in Los Angeles to other
situations in the troposphere is difficult because a lack of suitable
observational data. However, it is likely that the photochemical
formation of sulfate will begin to be important whenever ozone exceeds
0.1-0.2 ppm in air and reactive hydrocarbons are present. At the
ground, the reactions in aqueous media are likely to be important
in both urban and non-urban air if photochemically induced processes
are absent. Away from the ground in the middle and upper troposphere,
reactive hydrocarbon and free radical concentrations should be
rather low under most circumstances. The presence of organics in
aerosols at cloud level and above should be minimal so that the
aqueous oxidation processes globally should play a dominant role
between 2 km and 10 km altitude.
-105.
-------
In the lower stratosphere, the reactions of principal interest
for sulfate formation appear to be photochemically induced. In
particular, the homogeneous reaction of OH and SOa seems to be the
leading candidate for SOk formation at this time.
B. Nitrate Formation.
The knowledge of nitrate forming processes applicable to
the atmosphere provide a less convincing basis for explaining the
behavior of this ion than that of sulfate. The key difference in
mechanisms evidently is centered around the volatility of nitrous
and nitric acid and their equilibria in aqueous solution. The
Los Angeles experience emphasizes the distinct difference in
nitrate evolution as compared with sulfate. There appear to be two
extremes of behavior, first a short term transient production
accompanying the NOX peak in the morning and second a much slower,
but systematic production such that high concentrations of N03~
appear after two to three hours of air mass transport across the
Los Angeles Basin. The overall N03" production correlates poorly
with ozone, N0x, and NOa, but the transient peak in the morning
takes place at high relativel humidity.
The evidence suggests that nitrate could be generated via
homogeneous gas phase reactions or by hetrogeneous, aqueous pro-
cesses. It seems that intermediate species other than N02 are
involved at least during the morning transient period. The
independence of the nitrate conversion ratio on NOa or NOX may
suggest a droplet rate controlled N03~ formation after absorption.
The complicated equilibria involved in nitrogen oxide ion solutions
underscores the potential importance of basic substances such as
NH^ necessary to neutralize acidic particles. It is likely that
NH3 plays a critical role in stabilizing nitrate in aerosol particles.
This conclusion can be drawn from the Los Angeles experience since
there are relatively high ammonia concentrations in air over the
eastern parts of the Basin where the highest nitrate levels are
observed.
-106-
-------
Interpretation of nitrate formation in the troposphere is
difficult at present because of the paucity of atmospheric aerosol
data. However, it is likely that photochemical processes are
important since the highest nitrates are generally observed in
cities with high oxidant levels, particularly in Los Angeles. Again
it would seem that NH3 plays a crucial role in nitrate generation
if acids are involved. Thus one would expect that less nitrate would
be present at higher altitude away from the principal sources of
NH3.
The absence of NOs in stratospheric aerosols appears to be
related to the high volatility of HNO since this species has been
observed as a gas at altitudes above 15 km.
C. Organic Carbon.
Organic material is almost universally present in tropospheric
aerosol samples but is probably absent from stratospheric material.
The presence of organics in aerosols can be attributed to primary
natural and anthropogenic sources, as well as to secondary processes
in the atmosphere. The details of the secondary reactions are
essentially unknown, but an important candidate based on laboratory
studies is the ozone reaction with cyclic or linear olefins of carbon
number larger than approximately six. The principal limitation in
producing organic aerosols in the atmosphere appears to be the
vapor pressure of condensable species.
The data taken in Los Angeles indicate that most of the non-
carbonate carbon found in smog aerosol is secondary in nature. It
is made up of a substantial amount of oxygenated material.with some
organic nitrate present. The organic reactions yield mainly sub-
micron particles; production seems to correlate well with changes
in ozone as well as with sulfate. In Los Angeles the organic
fraction is normally the largest contributor by mass to filter
collected aerosol samples.
-107-
-------
The concentration of organics in the submicron particle
fraction provides circumstantial evidence for a gas phase formation
reaction followed by vapor diffusion controlled growth on existing
particles. The presence of significant quantities of oxygenated
material and the correlation with ozone buildup are consistent with
the expectation that ozone-olefin reaction is important.
At this time, there is virtually no information on the com-
position of olefins greater than C6 in the ambient atmosphere so
that the organic aerosol precursors cannot be considered as known in
urban air. However, taking a difference between the non-methane
hydrocarbon vapor concentration and the
-------
REFERENCES
1. Leighton, P., Photochemistry of AirPollution, Academic Press,
New York, 1961.
2. Went, F., Proc. Natl. Acad. Sci. U.S. 46, 212, 1960.
3. Friend, J.P., R. Leifer, and M. Trichan, J. Atmos. Sci. 30,
465, 1973.
4. Whitby, K.T. (ed.), "Aerosol Measurements in Los Angeles
Smog." Air Pollution Contr. Office Publ. No. APTD-0630,
U.S. Environ. Protection Agency, Research Triangle Park,
North Carolina.
5. Hidy, G.M., (ed.), Aerosols and AtmosphericChemistry, Academic
Press, New York, 1972.
6. Hidy, G.M. ejt al./'Characterization of Aerosols in California".
Report SC524.25FR, Science Center, Rockwell International,
Thousand Oaks, Calfironai, 1974.
7. Husar, R.B. and K.T. Whitby, Environ. Sci. & Technol. 7,
241, 1973.
8. Charlson, R.J., N.C. Ahlquist, H. Selvidge, and P.B. MacCready,
Jr., J. Air Poll. Contr. Assc. 19,, 937,1969.
9. Heisler, S., S.K. Friedlander, and R. B. Husar, Atmos. Environ.
7, 633, 1973.
10. Hidy, G.M., "Theory of Formation and Properties of Photochemical
Aerosols", presented at Fund. Chemical Basis of Reactions in a
Polluted Atmosphere, Battelle Summer Inst., Seattle, Wash.,
June 18, 1973.
11. Hidy, G.M. in Assessment of Airborne Particles, T. Mercer (ed.),
C. C. Thomas Co., 1972, p. 81.
12. Junge, C.E., Atmospheric Chemistry & Radioactivity, Academic
Press, New York, 1973.
13. U.S. Environmental Protection Agency, "Air Quality Data from
the National Air Surveillance Network, 1968," APTD, 1970.
14. Charlson, R.J. et al., Science 184, 156, 1974.
-109-
-------
REFERENCES (CONTINUED)
15. Cunningham, P., private communcation.
16. Robinson, G. and R. Robbins, J. Air Poll. Contr. Assoc. 20,
233, 1970.
17. Hidy, G.M. and J. R. Brock in Proc. 2nd IUAPPA Clean Air Cong.,
Academic Press, New York, 1971, p. 1088.
18. O'Brien, R. e_t al. , "Organic Photochemical Aerosols, II:
Atmospheric Analysis," submitted to Environ. Sci. & Technol.
19. Schuetzle, D., Ph.D. Thesis, University of Washington, 1972.
20. Grosjean, D. and S.K. Friedlander, "Gas-Particle Distribution
Factors for Organic Pollutants in the Los Angeles Atmosphere,"
submitted to J. Air Poll. Contr. Assoc.
21. Calvert, J.G., "Modes of Formation of the Salts of Sulfur
and Nitrogen in an N0x-S0a-Hydrocarbon-Polluted Atmosphere,"
presented at Conf. on Atmos. Salts & Gases of Sulfur &
Nitrogen in Assoc. with Photochem. Oxidant, University of Calif.
Irvine, Jan. 7-9, 1974; see also Proc. of Conf. on Health
Effects of Pollutants, Assembly of Life Sci., National Acad.
Sci., U.S. Govt. Printing Office No. 93-15, 1973, p. 19.
22. Husar, R.B. and D. Blumenthal, unpublished report, 1974.
23. Roberts, P. and S.K. Friedlander, "Conversion of S02 to
Particulate Sulfates in the Los Angeles Atmosphere, presented
at Conf., Health Consequences of Environ. Controls: Impact
of Mobile Emissions Controls. U.S. Environ. Protection
Agency, Durham, N.C., April 17-19, 1974.
24. Junge, C.E., C.W. Chagnon, and J.E. Manson, J. Meterol.
18, 81, 1961.
25. Lazrus, A.L. et_ al. , J. Geophys. Res. 76., 8083, 1971.
26. Junge, C.E., "Sulfur Budget of the Stratospheric Aerosol Layer",
in Proc. of the Intl. Conf. on the Structure, Composition, and
General Circulation of the Upper and Lower Atmosphere, Vol. 1,
IAMAP/IAPSO 1st Joint Assembly, Melbourne, Australia, Jan. 14-25
1974, p. 85.
-110-
-------
REFERENCES (CONTINUED)
27. Harrison, H. and T. Larson, J. Geophys. Res. 7_9, 3095, 1974.
28. Hidy, G.M. et al., "Aerosols from Engine Effluents," Chap. 6
Monograph III, Climatic Impact Assessment Program Dept. of
Transportation, to be published (1974).
29. Marker, A., "The Formation of Sulfate in the Stratosphere
through the Gas Phase Oxidationof Sulfur Dioxide',' submitted
to J_, Geophys.. .Res .
30. Toon, O.B. and J.B. Pollack, J. Geophys. Res. 7£, 7051, 1973.
31. Bufalini, M?J? .Environ. Sci. & Technol. 5_, 685, 1971.
32. Cox, R.A. and S.A. Penkett, J. Chem. Soc. , Faraday Soc. 68,
1735, 1972.
33. Davis, D.D., W.A. Payne, andL.J. Stief, Science 179,280, 1973.
34. Hamilton, E.J., Jr., private communication.
35. Payne, W.A., L.J. Stief, and D.D. Davis, J. Amer. Chem. Soc.
95, 7614, 1973.
36. Johnstone, H.F. and D.R. Coughanowr, Ind. Eng. Chem. 50, 1169,
1968.
37. Matteson, M., S. Stober, and H. Luther, Ind. Eng. Chem.
Fundamentals 8,, 677, 1969.
38. Scott, W.D. and P.V.. Hobbs, J. Atmos. Sci. 24, 54, 1967.
39. Miller, J.M. and R. G. dePena, Proc. 2nd IUAPPA Congr., Academic
Press, New York, 1971, p. 375.
40. Corn, M. and R.T. Cheng., J. Air Poll. Contr. Assoc.22, 871, 1972
41. Novakov, T., S.G. Chang, and A.B. Barker, "Sulfates in Pollution
Particulates: Catalytic Formaltion on Carbon (Soot) Particles,"
unpublished manuscript, Lawrence Berkeley Laboratory.
42. Penkett, S.A., Nature. Phys. Sci. 240, 105, 1972.
43. Fuller, E.G. and R.H. Christ, J. Amer. Chem. Soc. 6_3, 1644, 1941.
44. Schroeter, L.C., J. Pharm Sci. 52., 559, 1963.
45. Yamamoto, K., S. Michiham, and K. Kunitaro, Nippon Kaguku
Kaiski 7, 1268, 1973.
-Ill-
-------
REFERENCES (CONTINUED)
46. O'Brien, R. et al., "Organic Photochemical Aerosol 1:
Environmental Chamber Experiments," submitted to Environ.
Sci. & Technol.
47. Nash, T. , J. Chem. Soc. (A) 18., 3023, 1970.
48. Borok, M.T., Zhurn, Prikladnoi Khimii 33, 1761, 1960.
49. Baranov, A.V., E.A. Liberzu, and T. I. Popova, Trudy
Sibiskogo Techologicheskogo Instituta 38, 77, 1966.
50. Burton, C.S., G.M. Hidy, and E. Franzblau, "Aerosol Formation
from Ozone-Olefin Reactions," in preparation.
51. Groblicki, P. and G. J. Nebel, "The Photochemical Formation
of Aerosols in Urban Atm." in Chemical Reactions in Urban
Atmospheres, American-Elsevier, New York, 1971.
52. Wilson, W.E., Jr., and A. Levy, J. Air Poll. Contr. Assoc. 20,
385, 1970.
53. Wilson, W.E., Jr., and A. Levy, Current Research 6,
54. Wilson, W.E., Jr. A. Levy, and D.B. Wimmer, J. Air Poll.
Contr. Assoc. 22, 27, 1972.
55. Kocmond, W.C., "Determination of the Formation Mechanism and
Composition of Photochemical Aerosols," Final Report CAPA-8
Contract, CALSPAN Corp., Buffalo, N.Y., Aug. 1973.
56. Ripperton, L.A., H.E. Jeffries, and 0. White, "Formation of
Aerosols by Reaction of Ozone with Selected Hydrocarbons," in
Photochem. Smog & Ozone Reactions Adv. in Chem., No. 113,
57. Lipeles, M., D. Landis, G.M. Hidy, "Study of Formation of
Aerosols in Gas Reactions in a Flowing Stream," Report
#SC551, 15F, Science Cetner, Rockwell International, Thousand
Oaks, California, 1974.
58. Lipeles, M. et al., "Mechanism of Formation & Composition of
Photochem. Aerosols, Final Rept. EPA-RS-73-036, July 1973.
59. Heisler, S., Ph.D. Thesis, Calfornia Institute of Technology,
Pasadena, California, in preparation.
60. Wilson, W.E., Jr., discussion in Chemical Reactions in Urban
Atm., C.S. Tuesday (ed.), American-Elsevier, New York, 1969,
p. 264.
61. Ho, W.W., G.M. Hidy, and R. Govan, J. Appl. Meteor. (1974)
to be published.
-112-
-------
REFERENCES (CONTINUED)
62. Friedlander, S.K., Environ. Sci. & Technol. 1_, 235, 1972
63. Georgii, H.W. , J. Geophys. Res. 75_, 2365, 1970.
64. Hidy, G.M..et al., J. Applied Meteorol. 13, 96, 1974.
65. Castleman, A.W. , Jr. ejt al, Tellus. 26, 250, 1974.
'66. Cuong, N.B., et ajL. , Tellus 26, 241, 1974.
67. Rhode, H., J. Geophys. Res. 77, 4494, 1972.
-113-
-------
APPENDIX B
"PARTICULATE MATTER IN THE ATMOSPHERE"
J.P. LODGE •
MAY 1975
-------
PARTICIPATE MATTER IN THE ATMOSPHERE
by
James P. Lodge, Ph.D.
Consultant in Atmospheric Chemistry
385 Broadway
Boulder, Colorado 80303
Prepared For:
RADIAN CORPORATION
8500 Shoal Creek Blvd.
Austin, Texas 78766
-114-
-------
PARTICULATE MATTER IN THE ATMOSPHERE
Irrespective of the very real hazards posed by gaseous
pollutants, it is clear that, from the public viewpoint, the most
evident air pollutant is suspended particulate matter. Airborne
particles obscure visibility and change the color of the sky. The
unaided sense of sight thus documents the presence of particulate
pollutants wherever they occur. It is probably that only malodors
among the other pollutant species are as clearly detectable or as
likely to produce public outcry.
For example, it is probably safe to assert that the entire
present pressure to remove sulfur dioxide from power plant stack
gases, almost irrespective of emitted concentrations, was triggered
by the highly visible particulate plume from the Four Corners power
plant. The truth of this statement is difficult to document pre-
cisely, but a general reading of the popular news media strongly
suggests its truth. Had the Four Corners plant been originally
constructed with adequate air cleaning equipment so that visible
emissions were minimized, it is highly unlikely that the entire
present pressure to remove sulfur dioxide as well could ever have
been mounted. In fact, the public is eager to confuse pollution
with sensory impressions. Local newspapers have received quantities
of letters protesting "the unbearable order of carbon monoxide in
the streets," and "the poisonous air in the tunnel; you could
hardly see for 200 feet.
In addition to adverse public reactions, effects of
suspended particulate matter fall into the following categories.
(1). So called "nuisance dust." Generally speaking, this
is used to refer to the actual soiling of surfaces, primarily by
settlable dust. These are very large particles with significant
falling velocities in still air. The name is to some extent a misnomer,
-115-
-------
since settled dust imposes a specific economic impact by increasing
the necessary frequency of cleaning whatever surface it lands on.
(2). Decreased Visibility. Two cases must be distinguished,
The acute visibility decrease characteristic of duststortns and
extremely high pollutant levels can have an immediate impact through
increased accident rates and airport closings. At far smaller
particulate concentrations, the inability to see distant objects
and the sky discoloration has an esthetic impact that may well trans-
late into economics. That is to say, there are indirect impacts
including psychological depresssion that may decrease individual
productivity, depreciation of real estate values, or, in the case
of tourist areas, a major disruption of tourism because of the
impairment of particular scenic vistas. Values of this sort can well
change with time; what is today a tolerable degree of visibility of
impairment may become intolerable when highlighted by the erection
of a highly visible landmark building, an elevated roadway that
allows longer views than are possible in street canyons, or even
the establishment of the new park that affords previously unavailable
view over the city.
(3). Pneumoconioses. Certain specific particulate species
cause specific health effects that are adequately documented as to
etiology. Obvious examples are silica, asbestos and beryllium.
There are certainly other specific physiological responses that will
also be documented in the near future.
(4). Epidemiological Effects. There is a sizable body
of information showing a coincidence between high particulate
concentrations and morbidity. There is at the moment no etiological
connection between the two. Taking the data at face value, it is
possible to hypothesize (a) that high particulate concentrations
cause illness; (b) that some fairly ubiquitous component in the
particulate matter causes illness; (c) that some non-particulate
species that tends to vary in the same sense as particulate matter
-116-
-------
causes illness; or (d) that sick people emit particulate matter.
While external physical considerations made (d) above unlikely,
there is insufficient basis for choosing among the other three
alternatives.
Despite the uncertainties involved, the National Air
Quality Standards for particulate matter were based on the
epidemiological relationships cited above, and the secondary
standards primarily on the esthetic effects; a partial concentration
of 60 micrograms per cubic meter corresponds, on an average, to
a visibility of approximately 20 kilometers.
It should be noted that the different effects generally
derive from different portions of the size spectrum of particles.
Particles larger than a few microns are inhaled to a negligible extent,
and are hence of minor concern as lung irritants. On the other hand,
very large particles can settle on vegetation and cause toxic effects
by subsequent ingestion. The most efficient sizes for obscuration
of visibility are in the range between 0.1 and 1 micrometer, but very
large particles, if present in sufficient concentrations, can obscure
visibility very effectively. Particle size range during duststonns
is very large indeed. Particles below 0.1 micrometer are least
effective in decreasing visibility. Nuisance dust, as noted
previously, is nearly all larger than about 5 micrometers.
Generally speaking, particles larger than one micrometer
are produced by mechanical processes, while particles smaller than
that limit are produced by condensation from the gas phase. This
means, among other things, that there are significant differences
in particle composition, with the break point for most of these
occurring somewhere near one micrometer. Since a number of the effects
also differ for particles above and below that limit, there has
been pressure to create different and separate standards for particles
above and below that limit.
-117-
-------
Still more recently, with the availability of increasing
complete analyses of atmospheric particulate matter, it has become
clear that, in some cases, a significant portion of the total
particulate loading of the atmosphere is caused by conversion of
gaseous pollutants to particles. Species that enter the atmosphere
in the particulate state are generally referred to as primary
particulate pollutants; those resulting from gas to particle con-
version subsequent to emission are called secondary. The rather
high fraction of secondary pollutants found in some atmospheres
has raised the question as to whether greater gains may be made in
reducing ambient particulate concentrations by controlling the
primary emissions, or by controlling those gases that create the
secondary pollutants.
In this report, primary emphasis will be given to two
questions: (1) the relative effectiveness, on a national basis, of
controlling primary particulate emissions as against controlling
those gases that produce secondary particulate matter; and (2)
the possible restructuring of the National Air Quality Standards
on the basis of particle size or of other considerations.
SOURCES OF ATMOSPHERIC PARTICLES
Before proceeding to answer these questions, it is necessary
to understand the source of the particulate matter found in cities,
and for that matter, in remote locations. This has been nicely
schematized by Friedlander (1973). By a series of regression
formulas, Friedlander was able to account for a very large fraction
of the particulate matter collected at a site in Pasadena,
California, by considering only 12 sources, although not all of these
could be completely resolved. Harrison, et al. (manuscript, source
unknown) apparently accounted for a very large fraction of suspended
particles in Chicago and an even smaller number of categories.
Unfortunately, similar regressions or other exhaustive treatments
have not been undertaken for other cities.
-118-
-------
Sources of Particles Larger Than One Micrometer. As noted
previously, the bulk of the larger particles are produced mechanically
In most areas, the major single source of such particles is the soil,
with the mineral industries and other grinding processes generally
supplying most of the balance. Depending on location, there may
also be a significant component of marine aerosol in this size range.
However, numerous studies have shown that the surface concentration
of marine aerosol decreases rapidly with distance inland, and this
should be an important component only in coastal cities. In general,
the industrial component of this size fraction contains few surprises.
Whenever material is pulverized, some fraction of it generally escapes
to the atmosphere and will be found in samples taken at a suitable
location. A flour mill will produce starch particles, a cement plant
will produce particles of cement, an iron foundry bits of rust,
and so on through the list. In addition, the movement of traffic
will provide particles of both the road surface and of rubber tires.
One of the most generally underrated components of this
largest size fraction is soil and associated materials, including
bits of pulverized vegetation in the appropriate seasons. Under
many circumstances, this can make up more than half of the total mass
of particulate matter. Lodge, et al. (196) showed that roughly one-
fifth of the carbon in a series of large samples from St. Louis, and
nearly twice that amount in a single Los Angeles sample were
contemporaneous carbon, based on carbon 14 content. The bulk of this
was in the portion of the particulate matter that was insoluble
in ether. Unfortunately, there was no separation by size in the
samples, but the most likely interpretation would be that an important
portion of the organic matter consisted of spores, micro-organisms,
pollens, and comminuted plant material. Unfortunately, this work
seems not to come to the notice of those presently attempting to
make elementary balances on more recently collected samples.
-119-
-------
There is a strong temptation to regard soil particles as
"natural" pollutants. In fact, this is rather infrequently the
case. There are very few situations in which quantities of dust become
i
airborne from soil that is in the totally undisturbed natural state.
However, in many situations, the natural vegetation cover is extremely
fragile, and the passage of only a few people or vehicles is enough
to destroy it. At this point, then the soil becomes very susceptible
to wind erosion. There is good evidence that even as vast an area
of denuded soil at the Sahara Desert is probably originally
anthropogenic. Unfortunately, the recovery rate of such soil, once
disturbed (in this case by over-grazing) is on the time scale of
millenia.
Hagen and Woodruff (1973), among others, have shown that wind
entrainment of soil has a rather abrupt onset at approximately 7
meters per second wind speed. This offers the possibility of
differentiating aeolian dust from other components by segregating
measurements made at higher wind speeds. That even this strategy is
not completely successful will be shown later. In addition, of
course there is undoubtedly a significant portion of dust stirred up
by vehicle motion on unpaved roads and streets, in unpaved parking
lots, and by other related mobile activities.
ParticlesSmaller Than One Micrometer. Whitby and his group (see,
for example, Willeke, et al. (1974) have shown that the submicron
aerosol is generally bimodal in size distrubiton. One mode, usually
in the vicinity of 0.02 micrometers, represents the initial size
distribution of particles formed by rapid condensation from the
vapor state in thermal processes. For example, the bulk of the
lead compounds from leaded gasoline combustion enter the exhaust
system in the vapor state and condense during the transit through
the exhaust system, or within a very short distance of the end of
the auto tailpipe. Metallurgical fumes and soot from incomplete
combustion probably also begin in this size range. However, at this
very small size, the amplitude of brownian motion is large and the
concentrations at the source are rather high. As a result, rapid
coagulation is apt to occur, giving rise to the second mode at
-120-
-------
about the point at which brownian motion becomes unimportant—at
ordinary atmospheric pressures, somewhere in the decade of size
immediately below one micrometer. Whitby's group refers to this
second mode as the "accumulation size range," for fairly obvious
reasons, and considers its predominance to indicate an aged aerosol
mass. While interactions between this size range and the large
particles unquestionably occur, their rates of diffusion are
sufficiently small that it an infrequent affair, and does not con-
stitute, on the time scale of air travel over a moderate sized
city, any significant cause for depletion of the accumulation mode,
nor for particle growth in the large size range. One exception
is the case in which the fume material is temporarily deposited on
the surface and then later reentrained. For example, an automobile
that has been operated for a protracted period at low speed
accumulates a certain amount of lead in its exhaust system. If
it is then taken up to highway speed, this deposited lead is
mechanically discharged in large particles. Similarly, soot
deposited in chimneys can be blown loose in floes that are large
enough to be visible by the naked eye.
This size range will also include organic materials such
as lubricating oils that are vaporized at high temperature and
condense on subsequent cooling, precisely analogously to the
inorganic species as described above. Such materials are generally
high in carbon and low in hydrogen and, at least when fairly
recently emitted, contain negligible amounts of oxygen.
Secondary AerosoIs. The work of Goetz and Pueschel (1965) suggests
that the formation of secondary materials by the photochemical re-
action of organics and nitrogen oxides ("The Los Angeles Smog"
reaction) occurs preferentially on surfaces of pre-existing particles
It further suggests that the extent of reaction is rather pro-
portional to the available surface. Whitby, et al. (1975) have
shown that nearly all of the surface in a typical Los Angeles air
sample is in the submicrometer size range. Growth on the very
smallest particles may be somewhat inhibited by Kelvin effect, but
-121-
-------
it is still to be expected that most of the growth of organic
secondary aerosols will occur on the submicron particles.
Inorganic secondary aerosols are predominately sulfates
and nitrates; they also seem to occur predominately in the sub-
micrometer size range. Less is known concerning their precise
distribution, particularly in the very smallest sizes. However,
it should be noted that the converse statement cannot be made--
that all sulfates and nitrates are secondary. In the first place,
some sulfate and nitrate are probably formed by interaction of the
corresponding gases with existing particles to displace such
appropriate anions as halide. Since there is very little change in
particle mass or other characteristics, it is probably not
completely proper to refer to such metamorphosed particles as
secondary. In the second place, soils in many parts of the country
contain gypsum and other sulfate minerals, as do a number of
commer-ial fertilizers and other industrial materials. Thus some
fraction of many sulfate samples may be primary. The matter will
be further discussed below.
RELATIVE CONTRIBUTION OF PRIMARY AND SECONDARY AEROSOLS
Friedlander (1973) is a brilliant synthesis, managed to
resolve the primary and secondary components of a group of samples
collected in the Los Angeles area. He estimated that roughly 40 per-
cent of the total mass was secondary, with organic photochemical
aerosol comprising roughly 24 percent. Other secondary materials
were thus approximately 15 percent, predominately sulfates. Dams,
et al. (1975) carried out an exhaustive analysis of a single large
sample of particulate matter from Milan, Italy. They attempted no
interpretation, but the data are sufficiently complete to permit
some rough calculations. If it is assumed that the secondary organic
material has the same elementary composition as assumed by Friedlander,
and that all of the hydrogen in the sample is contained in the
secondary organics, then such materials make up somewhat in excess
-122-
-------
of 30 percent of the total, while the sum of sulfate and nitrate
is roughly 10 percent. Recognizing that these are all upper limits,
it can be said that the Milan sample probably does not differ
seriously from the samples examined by Friedlander. Hence, if one
assumes the experience of Los Angeles and Milan to be universal,
it would appear that the suppression of secondary aerosol formation
would decrease airborne particle loadings by nearly one-half.
On the other hand, Harrison, et al. (manuscript, source
unknown) seemed able to account for a very large percentage of the
total particulate matter collected in Chicago as primary particles.
If the data of Lodge, et al. (1960) can be taken as even remotely
representative, a large fraction of the organics in the Los Angeles
air appear to contain contemporaneous carbon. It is to be
anticipated that these would also be highly oxygenated species, and
quite conceivably indistinguishable from smog aerosol by the mass
balance technique employed by Friedlander. A similar argument can
be applied to the Milanese data as well. These findings certainly
do not invalidate the previous argument, but could open it to question
until further data, based on size fractionated aerosols, can be
adduced.
Now let us examine the data of Willeke, et al. (1974) for
two sites in and near Denver, Colorado during the late Autumn. At
a site slightly to the North of the City, in the middle of the
drainage plume, with largely cloudy skies and a mean temperature over
a period of almost a week of 9°C, they measured a mean concentration
of the order of 400 micrograms per cubic meter, with nearly 90
percent of the particles being above one micrometer in size. At
the same time, nearly 90 percent of the particle surfaces was in
the range below one micrometer.
Considering the low insolation, the low temperature, the
particle size distribution and the particle surface distribution,
it is extremely unlikely that secondary aerosols played any role in
this set of aerosol data. Interestingly enough, Sverdrup, et al.
(1975) cite a similar set of data collected during a summer night
-123-
-------
near Fort Collins, Colorado in a very clean area. Although the
numbers are far smaller, the relative size distribution is very
similar. They comment that the mode at a size around 19 micro-
meters or larger was an invariable constituent of Colorado
samples. It might be noted that, out of more than 60 particulate
monitoring stations in the State of Colorado, no more than 12
appeared to meet the Federal Secondary Standards during 1973
(Colorado State Air Pollution Control Program, 1974). Uncompiled
data from the particulate sampling network in Colorado strongly
suggest that it is difficult consistently to meet the National Air
Quality Standards in any community of greater than approximately
500 persons. There is in fact a reasonable indication that there
are portions of Colorado with a population density of fewer than
10 persons per square mile in which the Federal Primary and/or
Secondary Standards are frequently exceeded. (The previously cited
work of Harrison and his colleagues in Chicago also showed soil
materials to be a major component in the air of Chicago.)
STRATEGIES FOR DECREASING TOTAL PARTICULATE LOADINGS
Very evidently the foregoing data do not completely
catalogue the variations possible among cities and, in fact, sites
within cities. No attempt has been made to explore some fascinating
deviations from the norm, such as a few periods of data from Whitby's
group, both in Los Angeles and in Denver, when the normal relation-
ships between visibility and particle loading broke down totally.
There are also certainly other that could be adduced to prove
miscellaneous other points. However, these cases appear to represent
extremes, and thus to define the field within which control strategies
must operate. Typical aerosol levels for Los Angeles are somewhat
in excess of 100 micrograms per cubic meter. Hence a total control
of all substances leading to secondary aerosol formation would
presumably bring Los Angeles into compliance with the Federal Primary
Standards, assuming that the days measured there are typical. Periods
of measurement for Denver were unquestionably atypical; some photo-
chemical pollution does occur during the summer-time. Nevertheless
-124-
-------
they comprised a sufficiently long period to demonstrate that the
air quality standards would be frequently exceeded in the total
absence of secondary reactions. (It should be noted that sulfate and
nitrate comprise something of the order of 3 percent and one percent,
respectively, of the total Colorado aerosol, virtually independent
of the part of the state in which the measurements are made.)
(Colorado State Air Pollution Control Program, 1972). It is
difficult to interpret these as secondary aerosols. It should also
be noted that, if these compositions are fairly constant, there will
be numerous days when the total sulfate content will exceed the limits
delineated by the CHESS study of 10-15 micrograms per cubic meter
sulfate as the threshold of physiological effect. Making the
reasonable assumption that these species are contained in the large
particle fraction, it may be necessary to use a size-discriminating
collector to separate the effects of natural sulfate from the pre-
sumably toxic anthropogenic component.
Realistically, total control of substances forming secondary
aerosols is as improbable as total control of any other pollutant.
Even for the clearest case, the Los Angeles Basin, a practical degree
of control would only marginally allow the area to meet the ambient
standards. To the extent that the estimates of secondary aerosol
comprise maximum values, this strategy would fall short. Control
of primary particulate pollutants, on the other hand, controls the
larger component even for Los Angeles. In addition, if the deductions
of Goetz and Pueschel (1965) have any validity, control of primary
pollutants would also result in a decrease in secondary pollutant
deposition as well. Clearly a major effort at secondary pollutant
control would have little effect on excessive particulate loadings
during a major portion of the year in Denver. Contrasting these
two cities, Los Angeles and Denver, it becomes clear, first, that
the question of control of secondary particles is almost meaningless.
There is good reason to control hydrocarbons and nitrogen oxides in
Denver because photochemical smog, with high oxidant concentrations,
occurs all too frequently during the summer months. Should a health
-125-
-------
effect of sulfates be demonstrated, there is good reason to control
sulfur oxides, the undoubted cause of secondary sulfates. It should
be noted that primary sulfates, where they exist, are not in the
respirable size range and hence would be unlikely to cause health
effects. With regard to Los Angeles, to the degree that these sub-
stances are controlled as a means of decreasing oxidant and sulfate
concentrations, a more or less corresponding reduction in secondary
aerosols can be expected. However, it is clearly necessary also to
control the primary aerosols that make up, even in Los Angeles, more
than half of the total mass, and that appear to be the nuclei on
which the secondary aerosols form. In Denver, such control of
primary aerosols is mandatory and central. Furthermore, since the
primary mass is in the large size range, it appears that fugitive
dust control could well accomplish more than further imposition on
emitting industries.
POSSIBLE REVISION OF STANDARDS
The data base for possible standards revision appear
Totally inadequate. The available epidemiological data are all
based on total loadings without regard to size distribution. It
will require a number of years of work, starting now, to test
correlations of community health with particular portions of the
total particle size distribution. While it seems rational to
anticipate that the primary health effects will be associated with
particles below one micrometer, this must (a) be demonstrated and
(b) be connected with actual particle concentrations. Obviously,
a particulate loading of, say, 100 micrograms per cubic meter, 90
percent of which is below one micrometer typical of Los Angeles,
can headly be expected to compare with the same loading in Denver
with 90 percent of the mass above one micrometer. Farther than
this it is difficult to go. It is also necessary to weigh the
esthetic/economic effect of such a high concentration of settlable
particles as appears to be typical of the Denver area. Once again,
however, inadequate data exist to turn this into a standard, since
data of this sort have been measured predominately in cities with
a moister climate, and therefore probably lower concentrations of
large particles, although very possibly similar total airborne
-126-
-------
concentrations. The best summary that it is possible to give of
the present situation is that there is probably inadequate evidence
seriously to change the present standard. However, it should be
recognized that this standard is itself a very shaky figure, and
probably not of equal applicability in all places. Particular
problems in meeting it will exist in the more arid portions of
the country, yet the respirable fraction of the particles will be
far smaller in such areas than in moister climates at the same total
particle concentration. There also appears to be some basis for
allowing the deletion of high particle concentrations that are
coincidence with winds in excess of approximately 15 or 20 miles
per hour.
NEEDED RESEARCH
Obviously, the primary need is to repeat the previous
types of epidemiological studies using both improved gas analysis
techniques and the determination of particle size distributions as
well as total loadings. New standards are needed at the earliest
possible date to permit a realistic assessment of the true hazard
of particulate matter. In view of control cost, a major expenditure
to solve this problem seems amply warranted. Simultaneously, far
better elemental and chemical analyses must be made, including, as
appropriate isotopic analysis, to identify the true major sources of
particulate matter in a number of cities.
For the drier climates, control priority must be given
to fugitive dust. Certainly unpaved parking lots can be paved or
sprayed with dust control agents, and more care may be taken to
minimize the area denuded of vegetation at building sites. However,
these considerations are of little use where the source of dust is
agricultural activities. There appears to be real merit in research
on, and a revival of interest in, shelter belts and other means of
controlling wind erosion. Wind tunnel studies and field experiments
need to be supported. There really appears to be a basis to believe
that Denver might solve some of its particulate pollution problem
by planting more trees'. Meanwhile, there should be no cessation of
work on the specific toxic effects of individual pollutant' species.
-127-
-------
SUMMARY
This study has highlighted, although it has not been
limited to, a comparison of particulate pollutant data from
Los Angeles and Denver, which apparently constitute two fairly
extreme cases with regard to their particulate pollution problems,
although mass concentrations are not wildly different on the
average. Slightly less than half of the Los Angeles aerosol
appears to be secondary. During a significant fraction of the
year, virtually all of Denver's particulate pollution is primary.
Study of these two cases leads to the rejection of
simplistic approaches on a nationwide basis. No single strategy
will solve the problems of either city. All gains are likely to
be incremental, and the total deployment of all reasonable strategies
could well fail to solve either problem completely.
There is acute need for a better definition of the real
problem areas in terms of concentration in each of the major size
ranges, and for more and better analytical data on the ultimate
composition of the particles, especially as a function of particle
size. From health considerations, a high particulate concentration
in Denver is probably less deleterious than a similar concentration
in Los Angeles. However, high concentrations are not esthetically
pleasing in either city, and every possible means should be deployed
to control the respective sources of pollution.
-128-
-------
REFERENCES
1. Colorado Air Pollution Control Program, Report to the
Public, 34-37 (1972).
2. Colorado Air Pollution Control Program. Report to the
Public, 56-59 (1974).
3. Dams, R., J. Billiet, C. Block, M. Demuynck, and M.
Janssens. Atmospheric Environment, in press (1975).
4. Friedlander, S.K., Environ. Science & Techno 1. 1_, 235-240
(1973).
5. Goetz, A. and R.F. Pueschel. J. Air Pollution Control Ass oc,
15, 90-95 (1965).
6. Hagen, L.J., and N.P. Woodruff. Atmospheric Environment
I, 323-332 (1973).
7. Harrison, P.R., R. Draftz, and W.H. Murphy. Manuscript,
source unknown.
8. Lodge, J.P., Jr., G.S. Bien and H.E. Suess. Int. J. Air
Pollution 2, 309-312 (1960).
9. Sverdrup, G.M., K.T. Whitby and W.E. Clark. Atmospheric
Environment 9_, 483-494 (1975) .
10. Whitby, K.T., W.E. Clark, V.A. Marple, G.M. Sverdrup,
G.J. Sem, K. Willeke, B.Y.H. Liu, and D.Y.H. Pui.
Atmospheric Environment 9, 463-482 (1975).
11. • Willeke, K., K.T. Whitby, W.E. Clark and V.A. Marple.
Atmospheric Environment 8, 609-633 (1974).
-129-
-------
APPENDIX C
"REVIEW OF SUSPENDED PARTICULATE MATTER"
J.R. BROCK
JUNE 1975
-------
REVIEW
OF
SUSPENDED PARTICULATE MATTER
JUNE 1975
by:
' J.R..BROCK, PH.D.
Chemical Engineering Department
The University of Texas
for:
RADIAN CORPORATION
AND
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
-130-
-------
SUMMARY RECOMMENDATIONS AND CONCLUSIONS
Projections for the U.S.A. made in this review (Fig. 11)
indicate that the suspended particulate mass concentration in the
atmosphere owing to anthropogenic secondary sources is at present
three times that due to anthropogenic primary sources. By the
year 2000, the secondary source contribution to the ambient atmos-
pheric aerosol will be four times that from primary sources. These
projections assume no positive control of emissions of the secondary
source gases and vapors.
Effective national control of suspended particulate matter
requires control of secondary source gases and vapors. The toxicity
associated currently with secondary source aerosols (sulfate particulate
for example) reinforces this recommendation.
Projections for the U.S.A. made in this review (Fig. 6)
indicate that the suspended particulate mass concentration in the
atmosphere owing to anthropogenic primary sources will nearly double
within the next thirty years. This is based on the critical
assumption that the product, (efficiency of control) x (application
of control), is constant.
An aid in formulation of more effective particulate control
strategies would be a systematic program to classify the various
urban regions of the U.S.A. as to the relative contributions of
primary and secondary particulate sources for each area.
Future particulate control strategies must be based at a
minimum on knowledge of particle size and composition distributions
of the ambient aerosol.
The research program of the atmospheric aerosol research
section, Chemistry and Physics laboratory, National Environmental
Research Center, Environmental Protection Agency has, over the past
five years, greatly expanded our knowledge of the characteristics and
dynamics of the urban aerosol.
-131-
-------
Research into secondary aerosol formation should receive
high priority. Current knowledge is inadequate of the mechanisms and
rates of conversion of gases and vapors to aerosol.
Determination of the composition distributions of the
major anthropogenic primary particulate sources should be assigned
high priority. Insufficient information on these composition dis-
tributions is available at present.
Application of microscopy and source coefficient cal-
culations for tracing ambient particulate matter back to sources
appears to be feasible for the larger particles (% 2pm). More
knowledge of atmospheric aerosol dynamics appears to be necessary to
identify finer particles with specific sources, owning to pronounced
physicochemical alterations experienced by these particles in the
atmosphere.
-132-
-------
BACKGROUND AND INTRODUCTION
Under the provisions of the Clean Air Act we are not
past the deadline for the status to protect public health by
achieving required reductions in the ambient concentrations of the
six major air pollutants. Yet, with the deadline just past, it is
estimated that 75% of the population of the United States are still
experiencing levels of suspended particulate matter and sulfur dioxide
above ceilings established by the U.S. Environmental Protection Agency
as requisite to protect the public health. It is clear that the
air pollution control strategies formulated to date have not been
successful.
This review looks at some of the deficiencies of current
control strategies for suspended particulate matter and suggests
alternatives together with forecasts of future trends.
Of the various air pollutants, particulate matter poses
perhaps the greatest difficulty in the formulation of effective
air pollution control strategies. From the outset, the term
"suspended particulate matter" is almost hopelessly imprecise.
In the atmospheric context, this term refers to an aerocolloidal
suspension of particles of widely varying shapes, ranging in size
from the molecular to the macroscopic and consisting of a bewilder-
ing variety of chemical species and physicochemical states. At the
same time these particles in the atmosphere are participants in
kinetic processes involving interactions between themselves,
various reactive trace gases, and atmospheric dynamical processes.
While a control strategy may be proposed for particulate
matter which parallels those for the various gaseous air pollutants,
it is clear that, owing to the complexities noted, a true parallel
does not exist and such a strategy is almost certain to be seriously
flawed.
-133-
-------
It is the purpose of this review to present an analysis
of the complexities of atmospheric particulate matter and to
indicate difficulties with current control strategies. Areas in
which additional research is essential will be identified and
suggestions will be given for future approaches to particulate
control.
Particulate Regulation and Control
The incentive for control of suspended particulate matter
stems from what are believed to be its adverse effects which arise
at concentrations which presumably are above atmospheric "background"
These include toxic effects arising from inhalation of particulate
matter, reduced visibility, economic damage through soiling,
corrosion and plant damage, and inadvertent weather and climate
modification.
Of these adverse effects, particulate toxicity is the para-
mount concern and the current primary national ambient air quality
standard for total suspended particulate matter for the U.S.A.
(75 yg/m3-- annual geometric mean; 260 yg/m3--maximum 24 hr. concen-
tration not to be exceeded more than once per year) is designed to
protect the public health with an adequate margin of safety (1).
The secondary standard (60 yg/m3-- annual geometric mean; 150 yg/m3--
maximum 24 hr. concentration not to be exceeded more than once per
year) specifies upper limits of suspended particulate matter requisite
to protect the public welfare from any known or anticipated adverse
effects (1).
To meet these specified ambient air concentration levels
for total suspended particulate matter, individual states have
promulgated emission standards which either are estimates of the effect
that emission reductions might have on ambient particulate concen-
trations or are based on current technological and economic feasi-
bility for reduction in particulate emissions from particular processe
or sources. The notable feature is, of course, that the emissions
-134-
-------
controls are directed toward ensuring compliance with a given total
mass concentration of ambient suspended particulate matter.
The present national primary air quality standard for
suspended particulate matter represents then at best, a current
estimate in relation to elimination of associated adverse health
effects. This estimate has been arrived at principally through
attempts at correlation of population morbidity and mortality data
with total mass concentration of suspended particulate matter (or
a measure thereof) together with, in some cases, sulfur or sulfate
concentrations.
Of course, suspended particulate matter is a very imprecise
term as has been noted above. The choice of total mass concentration
as the index of the complex physicochemical system reflects in the main
the historical circumstance that total mass concentration (or
measure thereof) has been one of the indexes most conveniently
measurable and was therefore the index most generally available when
correlations with adverse effects were sought. Other parameters
describing suspended particulate matter in the atmosphere are now
becoming available on a routine basis and one can anticipate the
alterations in control strategy will take place as a result.
Adverse Effects of Suspended Particulate Matter
As has been noted, many adverse effects have been traced
to the presence of suspended particulate matter including increased
morbidity and mortality rates, degradation of visibility, economic
damage, and inadvertent weather and climate modification. Very
little is known at present about the relation of suspended particulate
matter to inadvertent weather and climate modification and such
effects do not yet play an important role in shaping ambient air
quality regulations. An exhaustive survey of the remaining adverse
effects will not be presented, but a brief discussion will serve
to indicate the importance of some of the important characteristics
of suspended particulate matter.-
-135-
-------
Our knowledge of the inhalation risk from the urban aerosol
or suspended particulate matter particularly in terms of long term,
low level exposure is woefully inadequate. As Anderson (2) has
pointed out, "The most important constituent currently measured
in the air is dust, whether measured as suspended particulate, soiling
index, or crudely (and less validly) as dustfall. It may be that
dust per se seriously impars lung function, or that it is merely
an integrated index of the presence of some as yet undetermined
hazardous substance found in the ambient air". In fact, Amdur
(3) has asserted that it would be toxicologically "astounding" if any
of the effects observed in the laboratory or in ambient exposures
could be caxisally related to an individual pollutant at observed
ambient concentration levels. Thus, the safest conclusion is that
adverse health effects probably arise as a result of highly complex
interactions between suspended particulate matter, trace gases and
concomitant meteorological phenomena.
Whatever the case, it seems clear that particulate health
effects (ruling out of the discussion here deposition and subsequent
absorption of particles from the body surface) must at the outset
bear some relation to the probability of deposition of particles in
the respiratory system, which is in turn dependent on the aerodynamic
characteristics of the particles. Particles larger than a few
micrometers are filtered from the incoming air in the nasel passage
and in the remainder of the upper respiratory system. Particles in
the size range 0.1 to 1 ym are deposited with the poorest efficiency
and therefore are capable of deep penetration into the lung to the
region the alveoli where clearance mechanisms are extremely inefficient^
This dependence of deposition efficiency on particles size is
summarized in Fig. 1 for particles of unit density (4) . The depositiorf
curves are modified for particles of different density, although
for reasonable values of density the pronounced minimum in the range
0.1 - 1 ym remains.
-136-
-------
The principal correlations of adverse health effects
generally link particulate matter and sulfur oxides. Table 1
summarizes some of these correlative studies, upon which current
regulations discussed above are based (5-11). A more recent review
by Lave and Seskin (12) implicate particulate matter more definitely.
The very recent reported results of the CHESS program (13) indicate
that particulate sulfate may in some cases, be the principal
determinant of adverse effects as indicated in Tables 3 and 4. Sulfate
particulate is generally found in the submicrometer range where as
indicated in Fig. 1 such particles penetrate deeply into the lung.
The CHESS and other more recent studies indicate that the
composition of the particle may be the critical factor. Toxic
effects of particulate matter, must, of course, be related to the
interaction of deposited particles with material at the site of
deposition. Possible effects produced by inhaled particulate matter
after deposition are summarized in Table 5 (14) where particles are
divided into the broad categories "soluble" and "insoluble" (in
water). A primary example of a soluble toxicant would be sulfate
particulate matter while asbestos represents a well recognized
insoluble toxicant (whose toxic action, however, may possibly be
linked to the slow solution of silica in the lung fluids).
Composition, then, also may be a critical factor in
particulate health effects. Particulate components known to be"
systemic poisons, such as lead, mercury, beryllium, etc. should be,
of course, subject to specific regulation as to their allowable
concentrations. Only recently have comprehensive programs been
initiated to identify the particulate organic and inorganic fractions
in detail. This information together with a knowledge of the
distribution of composition with particle size should prove to be
invaluable in assessment of possible toxic effects.
It is clear that particle size and density play a primary
role in determining the region of deposition in the lung. Once
deposited, particle compositions should be essential in considera-
tions of possible toxicity. Of course, other characteristics of
particles may be of importance.
-137-
-------
Particle morphology may be of importance in clearance
efficiency of deposited particles from the lung; the morphology of
asbestos is strongly implicated in asbestosis, for example (15).
Particle shape, for example, is receiving increasing attention. The
urban aerosol has been shown (16) to consist of a great variety of
shapes classified into such generic shape categories as: spherical,
crystalline, platelet, fibrous, amorphous, bulky agglomerate, and
flat agglomerate.
There have been suggestions (17) that particulate matter
may combine with trace gaseous compounds to produce synergistic
effects. This view advances the hypothesis that particles may
adsorb or absorb gaseous constituents and facilitate the transfer
of these gases deep into the lung. According to this, then, the
particulate specific surface area for adsorption, such as may be
determined for example by the BET method, would be an important
characteristic of the urban aerosol in relation to adverse health
effects.
There are, of course, other parameters of suspended particu-
late matter which may eventually be relatable to possible health
effects. These include such characteristics as electrical charge,
radioactivity, chemical states (free radical density, for example),
etc. One can only state that we have insufficient information at
present to take us past speculation on possible roles of such
characteristics.
That urban aerosol is the principal agent of degradation
of urban visibility is generally accepted today (18). Furthermore,
the mechanisms by which visibility is degraded are well understood
in terms of characteristics of an aerosol particle.
The particles composing an aerosol decrease visibility by
scattering and absorbing visible incident light. The total extinction
b t of light of wavelength X by atmospheric aerosol may be calculated
easily if one assumes that the particles are all spherical and of the
same composition by the equation:
-138-
-------
N 2 Q n
bext = £ "^ r i ^i i (2.1)
Where Q. is the extinction efficiency of a particle of size class
i and is a function of the parameter 2lTr./X and the refractive
index, m, of the particles, n. is the number concentration of
particles of that size class, there being N such size classes
accounted, r, is the particle radius.
Equation (2.1) serves to suggest the various characteristics
of atmospheric aerosols which have to do with visibility. The
number concentrations of particles in given size classes (in other
words, the size distribution) is clearly of importance. Since Q.
is a function of refractive index which is in turn a characteristic
of the chemical nature of the particles, particle composition is of
importance. As we shall see, particle composition (as for example
the water soluble components) may also play a role in determining the
n. and m for the atmospheric aerosol.
The relation of total light scattering to particle size
and refractive index is indicated in Fig. 2 (19) where K , the total
s
scattering coefficient (Q. = K + ^ hs> ^ b t^ie total absorption
coefficient), is presented as a function of particle radius, r,
at various values of the real part of the particulate refractive
index. The notable feature of Fig. 2 is the occurrence of large
modes in KS in the region of visible light at values of particle
radius in the range 0.1 - 0.5 ym for a wide range of refractive
index. That is, particles of radius near the wavelength of visible
light scatter that light with greatest efficiency. This fact
coupled with the universal observation that particles in this same
size range are very numerous in the atmosphere as shown in Fig. 3
(20) , means that most of the light scattering in the atmosphere
is due to particles in the size range 0.1 - 1.0 ym. The few measure-
ments available at present tend to indicate that the contribution of
-139-
-------
absorption to total extinction is small. Fig. 4 (21) exhibits
explicitly for the Los Angeles aerosol the overwhelming dominance
of the size range 0.1 - 1.0 urn to light scattering and hence
degradation of visibility in this case.
Of course, as has already been remarked, atmospheric
aerosols are of varied shape, not all spherical as suggested by
equation (2.1), and may have widely differing chemical composition
(and hence refractive index) within and between size classes. Thus,
even though the physics of light scattering by particles is well
understood, it is not simple to propose a detailed model for de-
gradation of visibility due to the atmospheric aerosol. A field
observation such as indicated by Fig. 4 serves to suggest that cal-
culations based on the concept of particles as spherical and of a
given refractive index do not seem to give rise to results out of line
with expectations. The uncertainties regarding degradation of
visibility seem with present knowledge, therefore, to reflect merely
the general uncertainty in the factors which determine n. the
size distribution of the atmospheric aerosol.
Several studies, e.g. (22, 23) have focussed on the economic
loss (in addition to that from adverse health affects) resulting
from particulate matter. Specific categories of loss have been
identified including corrosion and degradation of materials,
damage to vegetation and lowered property values. Most of the studies
in this category have dealt with correlations between some gross
index of suspended particulate matter and the particular category
of damage.
The relation of specific characteristics of the urban
aerosol to economic loss has not been fully explored. There is,
of course, the obvious relation of degraded visibility to general
perception by a population of air pollution with consequent lowering
of property values. Damage to materials and vegetation from
particulate matter must entail as a first step, particle deposition
followed by chemical and/or physical interaction. Hence, particle
-140-
-------
size and composition are both important characteristics in
relation to the potential for economic loss due to particulate
matter.
Processes and Characteristics of the Urban Aerosol
As the cursory review of adverse particulate effects
has indicated, various characteristics of the urban aerosol must
be central in the formulation of effective control regulations.
These characteristics may be summariced:
(1) Particle size distribution
(2) Particle composition distribution
(3) Particle specific surface area distribution
(4) Particle shape
(5) Particle electrical charge distribution
(6) Particle radioactivity distribution
For the urban aerosol, these various characteristics are
the result of a number of complex processes occurring in the atmos-
phere. These are indicated schematically in Fig. 5 which indicates
also the relation of the size and composition distribution (the
remaining characteristics are implied) to the adverse effects which
in turn, are the basis for ambient air quality regulations.
As indicated in Fig. 5, particle size and composition
distributions in urban atmospheres arise as a result of generation
processes, growth processes, and removal processes. Aerosols are
generated in urban atmospheres through injection from primary sources,
which are those sources which introduce particles directly into
the atmosphere, and creation by secondary sources which are associated
with atmospheric chemical reactions between various trace gases.
Particles in the atmosphere may undergo various growth processes
such as coagulation or agglomeration and condensation or accretion.
Finally, particles are removed from the urban atmosphere by several
-141-
-------
processes including advection, convection, dispersion, scavenging by
hydrometeors and water clouds and dry deposition (particles removed
at the earth's surface in the absence of primary and secondary
sources are generating suspended particulate matter in the atmos-
phere where the various physico-chemical growth and removal processes
act to produce finally an urban aerosol with certain size, composition
specific surface area, etc. distributions.
Of course, no feasible methods have ever been proposed to
regulate and control the natural processes of particle growth and
removal. Therefore, as indicated in Fig. 5, ambient air quality
regulations must be achieved through control of the primary and
secondary sources of urban aerosols - that is, by reducing the emissi
of particulate matter and trace gaseous species.
3. Primary Sources
For the continental tropospheric aerosol a rough estimate
(24) attributes 50 to 6070 of the total aerosol mass to primary
sources. These are the sources which inject particles directly
into the atmosphere. Current control strategies are directed
toward reduction of anthropogenic primary source emissions through
imposition of control technology in the familiar forms of mechanical
collectors, scrubbers, filters and electrostatic precipitators.
An answer to the question of whether or not this strategy is cost
effective requires a detailed understanding of the atmospheric
aerosol growth processes and a detailed knowledge of the important
characteristics of primary source aerosols.
This section examines the current state of knowledge of
primary source emissions in terms of the particle size and com-
position distributions.
-142-
-------
As shown in Table 6, on a world-wide basis, aerosol sources
may be divided into the two general categories, primary and secondary.
The secondary sources arise from formation of condensable species
as a result of atmospheric chemical reactions between various gaseous
species. One further distinguishes within these categories the
anthropogeric and natural sources. This distinction for primary
sources may be in some cases rather unclear. For example, dust rise
by wind may be owing in some regions to certain agricultural operations
or poor agricultural management. Dust rise by wind action is a natural
process, but man may in some regions, as noted, increase this process
by his actions. The extent of this increase is difficult to document
so that the distinction between anthropogenic and natural primary
sources in this instance is blurred.
In the U.S. much of our knowledge of anthropogenic primary
sources stems from the use of emission factors which attributes, in
the case of particulate matter, a certain mass emission from a given
consumption or production in a process (.25) .
These anthropogenic primary source emissions are also
conveniently divided into two categories, stationary and mobile sources.
Table 7 (26) presents the various categories of stationary
primary sources together with projections of particulate emission
levels calculated by use of the emission factor concept. The three
different cases in each year correspond to:
(1) Net control of particulate emissions is assumed to be constant
for each industry at the 1970 value.
(2) Application of control is increased to 100% over the IRS
plant lifetime and efficiency of new controls is increased
based on an estimated technological growth forecast. It
anticipates regulatory activity over all stationary primary
anthropogenic sources.
(3) Same as case (2) plus increasingly efficient emission
standards.
-143-
-------
A similar estimate may be made for mobile sources. The
most important of these are highway vehicles which are projected to
have a constant emission factor of 0.38 g/mi for exhaust particulate
and 0.20 g/mi for tire wear particulate. For highway vehicles, one
must also consider an emission factor for highway dust (27) arising
from surface stresses from vehicular traffic on roads; the emission
factor is dependent on vehicle speed and road surface and has not
yet been simplified into a single factor by means of a generalized
test cycle.
While the concept of emission factors has been a valuable
method in the past, it is now widely recognized to be too inaccurate
for formulation of the more exacting strategies now recognized as
necessary to meet ambient air quality standards. The source coeffi-
cient method, to be discussed below, offers the possibility of an
important advance in knowledge of the contribution of primary sources
to the ambient aerosol at a receptor point. Unfortunately, at
present, this method is likely to be as inaccurate as estimates based
on emission factors.
Particle size distribution
In terms of meeting current ambient air quality regulations
for particulate matter, the particle size distribution of aerosol
from anthropogenic primary sources is perhaps the most important
factor. This aerosol characteristic is the determinant of particle
residence time in the atmosphere and hence of the mass of suspended
particulate due to a given primary source.
It is of consequence, therefore, to investigate the nature
of possible processes generating particles from primary sources.
Particles from a given primary source may be generated by the processes
of nucleation, comminution, or by combinations of these processes.
-144-
-------
Nucleation may be homogeneous or heterogeneous. The term
homogeneous nucleation embodies all those processes in which vapour
molecules interact physically or chemically to form particles; the
particle growth process begins from particle sizes of molecular
order and may proceed by coagulation, condensation or a combination
of these. In heterogeneous nucleation, new particles are not formed;
vapour molecules condense physically or chemically onto existing
particles, and, primary, one is dealing with a condensational growth
process.
Particle generation by comminution involves successive,
usually mechanical, subdivisions of liquids or solids to the fine
particle state. Aerosol generation at the air-sea interface and
dust rise by wind action at the air-land interface are important
examples of natural primary sources of particles formed by comminution.
These sources, in fact, are estimated (24) to constitute the two
largest contributors of aerosol mass on a world wide basis.
With these definitions, characteristics of the particle
size distribution produced by the particle generation processes of
nucleation and comminution will now be examined.
Homogeneous nucleation
Automobile exhaust represents perhaps one of the important
examples of an anthropogenic primary source in which particles are
apparently generated principally by homogeneous nucleation as defined
here. The residence time of the aerosol, before injection and
subsequent fairly rapid dilution in the atmosphere, in this and
other important! industrial combustion sources, is usually of the
order 0.1 ~ 1 s. Therefore, the particulate emissions from such
sources, are comparatively well-aged aerosols, for which the
particle size distribution has had sufficient time to reach a "self-
preserving" form by coagulation (30). The term "self-preserving"
refers to the tendency, of aerosols coagulating with the same collision
-145-
-------
parameter, b (u,uf), to achieve similar particle size distributions
after succifient time of coagulation. Also, a simple calculation
is sufficient to show that usually aerosols formed by homogeneous
nucleation will, in the time period. 0.1 ~ls, have an average size
which is the submicrometre range. If M is the total mass concen-
tration of condensed material formed by nucleation, the order of the
mean radius of the coagulated aerosol should be:
(1 + bN0t)/4irpN0) 1/3 (3.1)
where N0 is the initial embryo concentration and p is the particle
density .
Eqn. (3.1) becomes for N t» 1:
r ~(3Mbt/4Trp)1/3 (3.2>
and the order of the mean radius becomes independent of N . As
an example, for an automobile using leaded gasoline, the undiluted
exhaust has a total particulate mass concentration, M, of 10" 7 -10" 8
g/cm3. Eqn. (3.2) indicates, as do measurements, (31) that most of
the aerosol is certainly in the submicrometer range .
For anthropogenic primary sources of aerosol formed
principally by homogeneous nucleation and in which subsequent
particulate growth is by coagulation, one might infer that particle
size distributions from all such sources are of the "self-preserving"
form. The inference is the same if simultaneous condensation occurs
(32).
Published studies (30) of the numerical solution of the
coagulation equation in the free molecule and continuum regimes
support the foregoing conclusion. However, experimental measure-
ments of coagulating aerosols reveal that generally the aerosol is
more polydisperse than predicted by the self-preserving functional
form. The explanation for this behavior is a very familiar one to
statisticians (32). The aerosol measured in the coagulation
experiments does not represent a single population, but instead
a mixture usually in random proportion of a heterogeneous population.
-146-
-------
In other words, the aerosol actually measured is a composite
of many different aerosol populations, each with a different history.
Thus, the particle distribution function realized in an experiment,
G(X) is:
G(X) - E P F (X) (3.3)
i
where the p. are random weights attached to the various members
of the heterogeneous population each with distribution functions
For an infinitely composite population:
G(X) * / F(X,a)dU(a) . (3.4)
In an experimental realization of coagulation, not only
systematic spatial or time variations or random experimentar error
serve to create a composite population but, for dilute systems, un-
avoidable random fluctuations also contribute.
As a result of the effect of heterogeneity of population,
the aerosol formed by homogeneous nucleation from a given primary
source will always be more polydisperse than predicted by the
increase in polydispersity for a given primary source will probably
depend on. the details of that source such as geometry, flow dynamics,
etc. As a result, it remains to be determined whether or not for
various primary sources of this type generalizations are possible.
Certainly the fact of the "self-preserving" form provides a useful
base from which to proceed in the inquiry.
Heterogeneous nucleation
The dense hygroscopie plumes emitted from various industrial
processes are examples of aerosols formed by heterogeneous nucleation.
In these instances, water vapor has condensed on an existing hygroscopic
aerosol, which may itself in turn have been generated by homogeneous
nucleation or comminution.
-147-
-------
Without consideration of the complex initiation process ,
for the simple process of condensation of a pure substance on an
aerosol of some given initial density function n (r)dr of particle
radius, r, it is a simple matter to examine the development in time
of n(r,t). In this case the volution of n(r,t) is:
(3.5)
where f(r) is the growth law for a particle of radius r. For
example, in the continuum region Kn ->• 0, neglecting the Kelvin
effect, f(r) = a/r, where a is a constant (33). Similarly, in
i
the free molecule region, f(r) = a , a constant (33), if the Kelvin
effect is neglected.
Eqn. (3.5) is a first order equation for which solutions
may readily be found for arbitrary initial conditions. However,
perhaps the most interesting feature of the pure condensation
process is the tendency of condensation to produce a less poly-
disperse aerosol in the continuous region, when the Kelvin effect
can be neglected. In this case, it is a simple matter to show
the ratio of the standard deviation a to the mean radius YI»
approaches zero with increasing time:
°° . (3.6)
where the subscript - designates initial conditions.
Similarly, in the free molecule regine:
(3.7)
°
-148-
-------
N being the initial total particle concentration. This character-
istic of pure condensational growth has been utilized for the pro-
duction of approximately monodisperse aerosols in variations of the
original Sinclair-La Her aerosol generator (33).
If, as in the previous examples, the concentration of
condensing vapor is held fixed by the Kelvin effe.ct is included
in the term f(r), it can be shown that O/YI ->• 0 as a result of
condensation. However, if the quantity of condensing vapor is
limited, one finds that in an initially polydisperse aerosol after
condensation has proceeded, the smaller particles will begin to
evaporate while the larger ones continue to grow.
Also, additional complication beyond the scope of this
discussion arises in consideration of a hygroscopic aerosol which
grows at humidities below the critical supersaturation of some of
the particles. In such cases, n(r) can become bimodal and very
polydisperse.
More general condensation processes, including stochastic
effects, have been examined elsewhere (34). Such processes, as
well as randomization indicated in eqn (3.3) and (3.4), usually
act to increase the polydispersity of an aerosol. Additional
complication can be.introduced by considering as well simultaneous
/
coagulation and condensation.
When the deterministic condensational growth described
by eqn. (3.5) is the only process altering the aerosol size
distribution, the final distribution clearly will be determined by
the initial size distribution. This initial size distribution will
be that owing either to homogeneous nucleation or comminution or both,
When the condensation process is stochastic and/or randomization
occurs, the final particle size distribution resulting from
condensational growth will become asymptotically independent of the
-149-
-------
initial distribution, the equivalent of the "slef-preserving"
behavior for a coagulating aerosol.
However, in general, it is much more difficult to draw
conclusions concerning the nature of the particle size distribution
resulting from condensation in these cases than for coagulation.
Unlike the coagulation equation, the condensational growth equation
is coupled to the conservation equation of the condensing vapor; the
state of the suspending gas usually plays a secondary role in the
coagulation of fine particles. Furthermore, the ability of particles
to grow by condensation depends in detail on particle composition
or surface properties; such characteristics are usually not con-
sidered to be of great importance in coagulation. Therefore, for
sources in which particle generation by heterogeneous nucleation
plays an important role, detailed examination of the process dynamics
will be necessary to characterize the particle size distribution.
Comminution
Important natural sources of aerosol particles generated
by comminution have been cited at the beginning of this section.
Anthropogenic sources of aerosol generated by comminution are also
of common occurrence and include emissions from industrial operations
such as mineral, rock and gravel processing, sand blasting, cement
manufacture, etc., as well as inadvertent emissions resulting from
farming operations, etc.
The process of comminution begins with a body of macroscopic
size and by successive subdivisions or splittings, liquid or solid
particles capable of aerosolization are formed. It is therefore
the inverse operation to homogeneous nucleation and subsequent
coagulation. The evolution for comminution may be represented by
the relation:
(3.8)
-150-
-------
where c (y/y1) dt is the probability that a particle of mass y'
will split in time dt to form 1,2,3... particles of mass y and
c (y) dt is the probability that in the same time a particle of mass
undergoes splitting. The basic assumption of eqn. (3.8) is that each
particle splits with a probability independent of the presence of
other particles. Clearly additional detail can be introduced.
It is possible to show that the splitting process approaches
asymptotically a limit distribution (35) which, for certain assump-
tions concerning the splitting probabilities, can be approximated
by the log normal distribution. 'Just as the coagulation process
has for certain assumptions concerning collisions as asymptotic
limit distribution, so too does the process of comminution.
A common assumption in the discussion of the splitting
process (35) is that the probability of splitting is proportional
to some power of the mass of a particle. Clearly, if a comminution
process is carried out so that a particle of, say 1000 ym is split
with unit probability, a particle of 1 ym radius will be split with
a probability orders of magnitude less (10~9 in fact, if splitting
is directly proportional to particle mass for particles of unit density)
For this reason, many large sources of particles produced by
comminution, such as those cited above, will produce particles
in the range of larger particle sizes.
Although asymptotic limit distributions may exist for a
given comminution process, randomization can be expected to be
important, owing to the comparatively small number of particles
per unit volume in typical comminution processes. However, very
large particles are not important in the consideration of sources
of air pollution, so that the distribution produced by a comminution
process can be truncated at the order of 100 ym radius. Therefore,
the particle size variation of interest will generally be over only
one'or two orders of magnitude of particle radius. As a result
•the range of polydispersity which might arise from randomization
is restricted.
-151-
-------
While this theoretical discussion may provide some
qualitative insight into the particle size distributions from
primary sources, mechanisms of particle formation in many primary
sources are too complex for the simple generalizations suggested
above to be of practical use. Hence, one must measure experimentally
particle size distributions from particular primary sources. Even
here, of course, generalization may not be possible. In any event,
we do not know. Only limited studies have been published on the
particle size distribution from primary sources (36, 37). At the
present time there is indication (38) that more extensive experimental
investigations are under way of particle size distributions from
primary sources. It is suggested that any such experimental pro-
gram should be sufficiently extensive that useful generalizations,
even if a statistical nature, may be possible.
Particle composition distributions
As noted earlier, particle composition distributions
provide critical information in regard to assessment of adverse
effects of suspended particulate matter. While much information
is being accumulated on the composition distribution of ambient
aerosol, very little is known at present about the composition dis-
tribution of anthropogenic primary sources.
Some first steps have been taken to identify trace elements
with certain primary sources (29,39). The variations of trace
elements with particle size has been studied for fly ash from coal
field power plants (40,41,42).
Table 8 (39) gives estimates of composition of emissions
from some Chicago primary sources. Overall estimates such as
these for various trace elem-nts form the basis for the method of
source coefficients for tracing these elements in the ambient
aerosol back to their sources. The technique is straight forward.
Essentially, if P. is the percentage of any element i in the aerosol
(which is to be measured experimentally) then
Pi " I "id "ij °j (3.9)
-152-
-------
where p.. is the percentage of element i in the particulate matter
emitted from source j, C. is the fraction of the aerosol sample
contributed by source j, and a., is a coefficient allowing for
changes in the proportion of that element which may take place
as a result of various atmospheric processes. The source coefficient
C. is to be calculated from eqn (3.9) given knowledge of P. and the
a.., p.., which is, of course, precisely where there are the largest
gaps in our current knowledge.
It is strongly recommended that studies be carried out of
the composition distributions from primary sources and of the
atmospheric processes which may alter the relative amounts of the
various elements in the aerosol.
An additional promising method for tracing aerosol back
to the primary source is now under active development. This method
employs optical and electron microscopy together with electron and
x-ray microanalysis to. attempt identification of particulate samples
collected on filters and impaction stages (43, 44, 45, 47, 48).
For large particles, amenable to optical microscopy, direct
identification of some types of particles appears to be possible
from particle morphology and properties. Rubber tire dust, mica,
quartz, cornstarch, paper fibers, plant parts, fly ash, coal fragments,
limestone, etc. have all been identified as constituents of the urban
aerosol and are traceable back to their particular sources. For
smaller particles which must be examined by electron microscopy, x--
ray microanalysis coupled with microscopy has aided in identification
of specific fine particles such as automobile exhaust aerosol
particles. For larger particles, the use of microscopy appears to
be a feasible technique for identification of specific types of
particles arising from particular primary sources. For sub-micron
particles,'however, in active chemical environments, there is
sufficient mixing due to chemical reactions and accretion that this
technique may have limited value.
-153-
-------
Projections of atmospheric contributions from various
anthropogenic primary sources.
Up to the present time controls on emissions of particulate
matter from sources have been directed primarily toward reduction
of total mass emissions. From a regional or international stand-
point this may not be sufficient. Consider that the rate of change
of total suspended particulate mass, m, owing to a primary source
emission is a function of the primary source emission rate, a, and
the residence time, b l, of the aerosols from that source. As is
well known, these simple assumptions are not generally correct,
but they are useful for this rough estimate.
The functional relationship described above may be expressed
in the form:
= a-bm (3.10)
where b"1 is a function of particle size, which for a particular
source category will be chosen as the residence time of the mass
median diameter. For short time periods:
2£ ~ 0 = a-bm (3.11)
so that m = a/b gives the total mass of atmospheric aerosol at any
time owing to a particular source category. Table 9 presents some
comparisons for anthropogenic primary source categories in the U.S.A.
In these examples, the largest tonnage sources of particulate
emissions are not necessarily the largest contributors to the
inventory of suspended particulate matter.
The data of Table 9 from Ref. (26) represent the assumption
that net control of primary particulate emissions is assumed to
be constant for each industry at the 1970 value through the year 2000.
Production capacity will vary, but the product (efficiency of control)
X (Application of control) is constant. The estimates of particle
-154-
-------
residence time as a function of particle diameter in Table 9 are
owing to Estnen and Corn (64) . These investigators attempted to
establish these values through a series of indirect field measure-
ments and their accuracy is difficult to assess.
It is notable that some of the largest source emissions
in Table 9 (Rock Crushing, for example) are among the smallest contri-
butors to the ambient aerosol mass. In fact, even if perfect control
of the large particle sources were achieved, it would have little
impact on the ambient aerosol mass, according to these estimates.
A further comparison is possible from the data of Table 9.
If it is assumed that the diurnal average mixing height for the
contiguous continental U.S. is 3000 feet, one may calculate'from
these data the average ambient aerosol concentration for the con-
tiguous continental U.S. The results of this admittedly very crude
estimation are displayed in Fig. 6 along with the total U.S.
anthropogenic primary source emissions. It must be noted that the
average ambient aerosol concentration calculated is only that part
of the aerosol concentration due to anthropogenic primary sources.
It does not include natural primary sources or anthropogenic and
natural secondary sources. The calculation' for Fig. 6 represents
certainly an overestimation inasmuch as it does not allow for
advection out of the contiguous continental U.S. nor does it allow
for convective transport into the upper troposphere. Nevertheless,
regarded as an average for the entire continguous continental U.S.,
the figures do not appear to be entirely unreasonable.'
As seen in Fig. 6, the rate of increase of the average
ambient aerosol concentration is larger than that of the primary source
emissions themselves. This is owing to the relatively large residence
time of the fine combustion particles from power generation and
transportation sources which show considerable increases over the
period 1970-2000. If lead is removed from gasoline, of course, the
-155-
-------
rates of increase shown will not be as large as has been indicated.
According to Fig. 6, by the year 2000 the average ambient
aerosol concentration due to anthropogenic primary sources will be
in excess of 80 ug/tn3, almost double the value of 1970. Of course,
this represents an average for the U.S. and some areas would have
much smaller and others much larger concentrations. If one adds
in the natural background particulate concentration of 20-40 ug/tn3 (51) ,
and further adds in a value for secondary aerosol production (to
be discussed in the next section) one would obtain values of 180-200
ug/m3 for 1970 and of the order of 400 ug/m3 for the year 2000. These
large values result from the very large secondary source contributions
as estimated in the next section.
4. Secondary Sources
The naive view of the atmospheric aerosol supposes that if
one knows the size and composition distributions of particulate matter
emitted by primary sources then the ambient aerosol is readily com-
prehended through consideration of the mixing and coagulation of these
primary emissions together with alterations resulting from dispersion,
advection and convection. That such is not the case has been known now
for some time (20).
Subsequent to their injection into the atmosphere, primary
source aerosols may be subject to a large number of alterations produced
by various physicochemical processes including photochemical and thermo-
chemical reactions. When they occur, these processes lead to growth
of and substantial alteration of the chemical composition of the
ambient aerosol. In those circumstances, without current incomplete
knowledge, attempts to trace the ambient aerosol back to primary
sources through use of the source coefficient concept are fraught
with difficulties.
At present, in the U.S. the Atmospheric Aerosol Research
Section of the Chemistry and Physics Laboratory, National Environmental
Research Center, Research Triangle Park, Environmental Protection Agency
has underway and is sponsoring effective research programs to determine
-156-
-------
the atmospheric aerosol growth processes, including secondary sources.
To date, these efforts have been most fruitful and over the past five
years our knowledge of the urban aerosol has been enormously enhanced.
It is strongly recommended that this type of effort be continued and
increased, particularly in the area of aerosol chemistry which com-
prises knowledge of secondary sources and aerosol growth processes.
This section reviews current knowledge of those atmospheric
transformations leading to production of aerosol from gaseous chemical
species. Sources of secondary aerosol are summarized. Resultant
size and composition distributions are discussed. Finally, some
projections are made for the ambient aerosol.
Sources of secondary aerosol
Many gases and vapors are known to enter into homogeneous
chemical reactions which give rise to products capable of forming
relatively involatile species and hence (under Special conditions)
condensation aerosol under tropospheric conditions.
In the urban atmosphere, homogeneous nucleation is believed
not to occur if the aerosol surface concentration is higher than
approximately 500 urn /cc (63). At high surface concentrations
heterogeneous nucleation instead occurs. One possible exception
to this could be found in the nucleation of sulfuric acid aerosol:
Free radicals
S02 -I- %02 ^ S03
• -157-
-------
which may interact rapidly with ambient water vapor:
H-SO, -nH 0 + HO
*• "T £ £,
to form an aerosol by homogeneous nucleation. In more general case,
relatively involatile products of homogeneous gas phase chemical
reactions will condense onto the ambient aerosol (49).
Examples of reactions of possible importance in production
of substances promoting growth of atmospheric aerosol by condensation
are summarized in Table 10.
In addition, various trace gases and vapors are believed
to enter into heterogeneous chemical reactions involving ambient
aerosol. Details of these reactions have been studied in only a
few cases such as in the oxidation reactions of SOa in aqueous
solution aerosols. However, for many of the observed growth pro-
cesses of suspended particulate matter, the details are as yet unknown
Table 11 lists some of the heterogeneous reactions which are believed
to contribute to oxidation of atmospheric SOz and hence to aerosol
growth.
The sorption of water vapor (the most abundant of the
trace atmospheric gases and vapors) by the water soluble components
of the atmospheric aerosol constitutes an interaction of very special
importance. This process is associated with a large part of the
visibility reduction from the soluble fractions of the atmospheric
aerosol and, of course, plays the central role in the formation of
clouds.
-158-
-------
It is possible that combustion and other energetic processes
may inject into the atmosphere vapors of relatively involatile sub-
stances which may subsequently condense (physically) onto the ambient
aerosol. This has not been studied, evidently. However, it would
be expected that such processes would provide only minor contributions
to the total aerosol mass.
The condensation onto the ambient aerosol of atmospheric
small ions and radioactive species (such as the radon daughters) plays
a very special role in air pollution and toxicology (54, 55) but
again, is not believed to be of central importance in the evolution
of the atmospheric aerosol.
Much work remains to be done to elucidate the various
mecahnisms leading to formation of secondary aerosol. It must not
.be supposed that current knowledge is adequate. Particularly complex
are the formation and atmospheric transformations of various organic
vapors to the particulate phase. It appears that in photochemically
active environments organic gases and vapors may be the principal
source of secondary aerosol (52). Yet, our knowledge of aerosol form-
ation in this case is very sketchy in spite already of extensive smog
chamber and atmospheric investigations (52). This subject should
receive high priority in future research programming.
Secondary and primary urban aerosols
For the U.S. and for most cities it is very difficult to
generalize concerning the relative importance of primary and secondary
sources in particulate pollution. In a well documented case, we
i
know that the Los Angeles aerosol may owe as much as 70 to 8070 of its
mass to secondary sources at some times (56). On the other hand,
during winter pollution episodes in Denver, the aerosol appears to
be almost entirely dominated by primary sources (52, 57).
Many factors may be responsible for dominance of secondary
or primary sources. In the case of Los Angeles, climatological
and meteorological conditions combine with strong sources of NOX and
-159-
-------
organic compounds to produce an aerosol evidently dominated by
secondary sources. In northern cities in winter time conditions,
it is suggested here that the low ambient temperatures may ensure
that the aerosol of these cities is dominated by primary sources.
It has been shown that photochemical reactions and subsequent
aerosol growth are reduced significantly at low temperatures
(58-61). Other as yet unspecified conditions may cause some urban
areas to be dominated by either primary or secondary aerosol sources.
Again, there may be also a diurnal variation in addition to seasonal
variations of secondary/primary sources.
In brief, based on current knowledge it is not possible to
classify, except in some cases, major urban areas as dominated by
primary or secondary sources at particular times. Means for arriving
at such classifications should certainly be sought on a priority
basis. Such knowledge is essential, as has been discussed, for
formulation of rational particulate control strategies.
Aerosol size and composition distributions
As has already been noted, aerosols observed in the atmos-
phere appear to form and evolve by a complex combination of physico-
chemical processes. These processes may involve chemical reactions
of gaseous species, homogeneous and heterogeneous nucleation,
condensation, coagulation, input of particles from primary sources,
sedimentation and deposition processes, and various fluid dynamical
phenomena such as convection, dispersion and mixing. The particle size
distribution of an aerosol observed at a given time is the result of
the competition between and the action of these various processes.
Owing to the complex, non-linear nature of the operation of all these
processes, it is clear that statistical correlations of field data
will in general be inadequate for providing a basis for particulate
control strategies or specific control decisions and that explicit
knowledge of the dynamics of the processes shaping the particle size
distribution is required.
-160-
-------
Just as for investigations of complex homogeneous gas
phase chemical reactions, detailed mathematical models have proven
to be invaluable, so too far an understanding of aerosol dynamics
in the atmosphere or in smog chambers are mathematical models use-
ful. Whether or not such a program can succeed for complex
atmospheric aerosols remains to be seen.
Complete knowledge of all processes shaping the particle
size distribution implies also knowledge of the distribution of
the innumerable chemical species making up the particles—that is,
knowledge of the multivariate particle composition distribution:
n (xi.xa, X3,...x ; r, t) which gives the number of particles at a
point in space r at time t having masses of chemical species 1, 2,
3..., m in the range x, dx ; X2, dxa ; ...; x , dx . v x.= s.
. . r* IT u. -L
i=l
The evolution of n (xi , x2, x3, , x r,t) in time
is extremely complex and numerical simulation appears prohibitive at
present. However, one does not require usually such complete
information for the aerosol. In that which concerns some of the .
principal adverse effects of particulate matter, it is probably
possible to follow the evolution of much less complex distributions.
n2(y, z; r, t). n2(y,z;r,t) is the distribution function for an
aerosol composed of two classes of material, A and B, or alter-
natively, composed of two different chemical species, A and B.
Thus naCy.z; r,t) is the number of particles at r,t having masses
of species A in the range y.dy and masses of species B in the
range z,dz. A and B could be, for example, respectively the
hygroscopic and non-hygroscopic components of an aerosol particle.
This is an important division in practice for it is the amount of
water soluble hygroscopic material in a particle which determines
size of the particle at a given relative humdity. Also significant
adverse health effects have been traced to the mass of sulfate in
particular matter; sulfate of course would represent , the important
hygroscopic component of the particulate matter in an atmosphere
rich in SOa.
-161-
-------
With certain assumptions (62), one can write down the
conservation equations for nz(y,z; r,t) in the atmosphere as well
as the conservation equations for those chemical species of mass
concentration s- participating in the growth processes of the
particles. The evolution equations of this coupled system have
the form:
ri:L(y-yl,z-z/)irii(y/,%/)
o o
_L_
for the particulate phase. The left hand side of the equation
represents the usual convective transport of na in the atmosphere
with mean velocity V. The first term on the right hand side gives
the dispersion of n2 due to eddy diffusion where K is the eddy
diffusivity tensor. The second and third terms represent the change
in na owing to coagulation and the fourth and fifth terms the change
due to deterministic condensation of trace gaseous substances. The
sixth term accounts for particle removal by gravitational settling
where G(X) is the settling velocity of a particle of mass x.
v (y,z;r,t) of particles from primary source p with soluble and
insoluble components. Similarly y^ (y,z; rt) is the rate of input
-162-
-------
of nuclei at r,t from homogeneous nucleation processes N.
Extensive discussions of the coagulation and condensation
rate coefficients b and if; can be found in several references (e.g. 33)
V*VSo V- K- \/Sj + 2 5jp 4- Z jr (4_2)
f
J
where s is the mass concentration of the jc chemical species.
The terms s. and s. represent respectively the rate of input
and removal of 1 by chemical reactions r and r1. s. is the
J J jp
rate of input of j from primary source p. This set of equations
should then detail the evolution in time of na.
If the particle growth processes are not dependent on the
soluble fraction, one may then study the evolution in time of the
particle size distribution n (x; r,t). Detailed discussions of
the derivation of the evolution of n has been given (62) . The
form of the coupled equations is analogous to that for n2.
(4-3)
CO
.nit)
for the particulate phase, and
^/^t 4- V- VS> =? • 7- K- ys^ 4-
(4-4)
-u
for the chemical species undergoing condensation,
-163-
-------
Detailed analysis of the urban aerosol along the lines
suggested above is, of course, a most difficult task. The description
given here serves more as a framework for consideration of all
relevant parameters which shape the urban particle size and com-
position distributions. Basic information on all the aerosol kinetic
processes for various urban regions is now becoming available through
the research programs of A.A.R.S., E.P.A.
In a few years the analysis suggested here should be
possible in an approximate manner. Some first steps along these
lines have already been taken (56,57), but much more remains to
be done.
As a result of field studies undertaken since 1969 under
sponsorship of A.A.R.S., E.P.A., a clearer, but still approximate
picture of the particle size distribution of the urban aerosol
has emerged. Figs. 7 and 8 reflect the current view (48) of the
size distribution of the urban aerosol.
In this view, as indicated by Fig. 8, the mass distribution
of the urban aerosol is a bimodal function of the particle diameter,
reflecting, it is thought, the mixture of aerosols from the basic
primary generation processes discussed above — that is, comminution
and nucleation. One estimate (48) places approximately 1/3 of the
total mass concentration of the urban aerosol in the size range
below 2 ym diameter. There will be certainly considerable
variation in this fraction as the complete aerosol dynamical situation
is altered.
Figure 7 (48) indicates that the surface area size dis-
tribution may be trimodal near sources of combustion nuclei. Also
this figure indicates schematically that this trimodal distribution
is the result of the dynamical processes discussed in detail above.
-164-
-------
The detailed picture of the processes shaping the urban
particle size distributions are inferred by Figs. 7 and 8 will no
doubt undergo modification as our knowledge of the urban aerosol
becomes more fully developed. At present, the scheme given here seems
to correspond well with our intuitive understanding of aerosol
dynamics.
Fig. 9 (63) presents some results of partial composition
distribution obtained by cascade impactor in Riverside, California.
Shown, of course, is a partial distribution including sulfates,
nitrates, lead, iron and total particle mass distribution. It may
be noted that secondary aerosol, sulfates and nitrates, are
predominantly found in particles smaller than 1 ym and that iron,
associated with comminutive sources, is found in the particles
larger than 1 ym.
Fig. 10 gives an example of particle size and composition
distributions in photochemically active urban air. The indicated
distributions correspond approximately with present limited
knowledge (52) . The particle composition distributions are not
meant to be precise but indicate only possible ranges of relative
abundance, according to the mass concentration, of the classifications
water soluble/water insoluble and organic/inorganic. Water soluble
organic species in the urban aerosol are believed to be concentrated
in particle size below 1 ym. Water soluble inorganic species,
as indicated in Fig. 9, tend to be found below 1 ym as sulfates,
nitrates, etc. but in regions invaded by marine air, for example,
the large particles may contain some sea salt constituents. Water
soluble contributions to the larger particle mode may also be
traced to soluble salts in soils or in dust from wintertime street'
and highway maintenance. Water insoluble organic species appear
to be associated mainly with combustion and other high temperature
processes, and therefore, they are found predominantly in the
fine particles, particularly where particulate control technology
is in effect. Water insoluble inorganic species are dominant
-165-
-------
constituents of the large particles from comminution sources.
Figure 11 presents the results of projections of the
average U.S. ambient aerosol mass concentration due to secondary
sources. It was assumed in this projection that secondary aerosol
was distributed with a mass average diameter of 0.5 ym corresponding
to in inverse residence time in the atmosphere of 27 yr"1 (64).
Sources of secondary aerosol were identified as the oxidation
reactions of S02, NO , reactive hydrocarbons, and NH3 which yield
X
particulate material. Projections of emission rates for these gases
are from Hidy and Brock (65) . It was assumed that 60% of the
sulfur oxide emissions would be converted to aerosol, 40% of the
nitrogen oxides, 30% of ammonia, and 2% of the reactive hydrocarbons.
Of the total hydrocarbon emissions, 30% was assumed to be reactive.
There is no question that this type of projection rests
on gross assumptions which will no doubt be refined as our knowledge
of secondary aerosol increases. According to this projection, the
anthropogenic ambient aerosol mass concentration due to secondary
sources is now three times that due to primary sources. Owing
in part to the assumption of better control of primary source
emissions in the future (increased emissions of sulfur oxides owing
to increased usage of coal is another factor), by the year 2000
the secondary source mass contributions will be nearly four
times the primary source contribution.
These figures are overestimates for urban regions. The
conversion rates of the trace gaseous species giving rise to aerosol
are thought at present not to be too rapid, Hence, advection and
convection tend to diminish greatly the importance of secondary
contributions for most urban aerosols. However, on the average
one would expect to see the average mass concentration over rural
regions and over the oceans of the northern hemisphere rise as a
result of the U.S. secondary aerosol contributions.
-166-
-------
From this projection, it is discouraging to reflect that
even if primary source emissions were completely eliminated, the
nationwide average ambient mass concentration would only be reduced
by 25-30%. Of course this is on a nation wide average. As noted
above, many urban regions appear to have ambient aerosol dominated
by primary aerosol in which case, of course, primary particulate
control is very cost effective.
One can state safely that effective control of suspended
particulate matter as a national policy requires control of
secondary source gases and vapors. This view is considerably
reinforced when possible health benefits arising from elimination
of harmful aerosols is considered inasmuch as toxicity appears
to be associated importantly with the secondary particulate sources
-167-
-------
REFERENCES
1. Environmental Protection Agency: National Primary and Secondary
Ambient Air Quality Standards, Federal Register, 36,:' 8186 (1971).
2. Anderson, D.O., "The effects of air contamination on health"
Canad. Med. Assoc. J. 9_7 528, 802 (1967)
3. Amdur, M.O. "lexicological Appraisal of Particulate Matter,
Oxides of Sulfur and Sulfuric Acid". Paper 69-68, Proceedings
Air Pollution Control Association, New York, New York, June 22-
26, 1969.
4. Task Group on Lung Dynamics, Deposition, and Retentio- Models for
Internal Dosimetry of the Human Respiratory Tract, Health Physics
12 1973 (1966).
5. Winkelstein, W. "The relationship of air pollution and economic
status to total mortality and selected respiratory system
mortality in man", Arch. Environ Health 14 162 (1967).
6. Douglas, J.W.B. and Booras, S.G. "Air Pollution and Respiratory
Infection in Children:. Brit. J. Prevt. Social Med. 20, 1 (1966).
7. Lunn, J.E., Knowelden, J. and Handyside, A.J., "Patterns of
respiratory illness in Sheffield infant school children", Brit.
J. Prev. Soc. Med. 21 (1967).
8. Petrilli, R.L., Agrese, G. and Kanitz, S., "Epidemidogy studies
of air pollution effects in Genoa, Italy" Arch. Environ. Health
12_ 733 (1966) .
9. Carnow, B.W., Lepper, M.H. Shebelle, R.B. and Stamler, J.,
"The Chicago Air Pollution Study: S02 Levels and Acute Illness
In Patients With Chronic Broncho Pulmonary Disease" Arch.
Environ. Health 18 768 (1969).
10. Brasser, L.G., Joosting, P.E., and Von Zuelen, D., "Sulfur oxide -
to what level is it acceptable?" Report G-300, Research Institute
for Public Health Engineering, Delft, Netherlands, July, 1967.
11. Lawther, P.J., "Climate, air pollution and chronic bronchitis,"
Proc. Roy. Soc. Med. 5_1 262 (1958).
12. Lave, L.B. and Seskin "Air pollution and human health", Science
169 723 (1970).
13. Environmental Protection Agency "Health Consequences of Sulfur
Oxides: A Report from CHESS, 1970-1971. Report EPA-650/1-74-004,
May 1974.
14. Corn, M. "Measurement of air pollution dosage to human receptors
in the community" Environ. Res. _3 218 (1970).
-168-
-------
REFERENCES (CONTINUED)
15. Timbrell, V. "Inhalation and biological effects of asbestos"
in T.T. Mercer stal. "Assessment of Airborne Particles" p. 427,
C.C. Thomas, Springfield, M. 1972.
16. Corn, M. "Urban aerosols: problems associated with evaluation
of inhalation risk" in T.T. Mercer, et. al. "Assessment of
Airborne Particles", p. 465 C.C. Thomas, Springfield, 111., 1972
17. Corn, M. Montgomery, T.L. and Reitz, R. "Atmospheric
particulates: Specific surfaces and densities Science 159 1350
(1968).
18. Air Quality Criteria for Particulate Matter, U.S. Dept. H.E.W.
Publ. AP-49, 1969.
19. Green, H.L. and Lane, W.R. "Particulate Clouds: Dusts, Smokes
and Mists" Secon Edition, E. & F.N. Spon. Ltd., London, 1964.
20. Hidy, G.M. and Friedlander, S.K., "The nature of Los Angeles
Aerosol" in H.M. Englund and W.T. Beery (ed.) "Proceedings of
the Second International Clean Air Congress", Academic Press,
New York 1971.
21. Ensor, D.S., Charlson, R.J., Ahlquist, N.C., Whitby, K.T., Husar,
R.B. and Liu, B.Y.H., "Multiwavelength nephelometer Measurements
in Los Angeles smog Aerosol", in G.M. Hidy (ed). "Aerosols and
Atmospheric Chemistry", Academic Press, N.Y., 1972.
22. Ridker, R.G. "Economic Costs of Air Pollution" New York, Prager,
1967.
23. Barrett, L.B. and Waddell, T.E. "Cost of Air Pollution Damage"
EPA Report AP-85, February 1973.
24. Hidy, G.M. and Brock, J.R., Proceedings of 2nd Clean Air Congress
IUAPPA, Washington, D.C., Dec. 1970.
25. "Compilation of Air Pollutant Emission Factors," Second Edition
E.P.A. Report AP-42, April 1973.
26. Vandegrift, A.E. et al, "Particulate Air Pollution in the U.S."
J. Air Pollution Control Association, 2_1 (1971)
27. Sehmel, G.A., "Particle resuspension from an asphalt road
caused by car and truck traffic," Atmos. Environ. 1_ 291 (1973)
28. Gatz, D.F., "Relative contributions of different sources of urban
aerosols: application of a new estimation method to multiple
sites in Chicago", Atmos. Environ. £ 1 (1975).
29. Miller, et. al., "A chemical element balance for the Pasadena
Aerosol", J. Colloid Interface Sci. 3£ 165 (1972).
30. R. Drake in "Topics in Current Aerosol REsearch", Pergamon,
Oxford, 1972.
-169-
-------
REFERENCES (CONTINUED
31. M. Lee, R. et. al., ACmos. Environ. .5 275 (1971).
32. Pich, J., et. al., Aerosol Sci. 1 115 (1970).
33. G. Hidy and J.R. Brock, "The Dynamics of Aerocolloidal
Systems", Pergamon, Oxford, 1970.
34. Brock, J.R., J. Colloid Interface Sci., 3j? 32 (1972).
35. Kolmogorov, A., Akad. Nank SSSR, 31 99 (1941).
36. "Particulate Pollutant System Study", MRI contract No. CPA
2269104, EPA, 1971.
37. Schulz, E.J., et. al. "Submicron particles from a pulverized
coal fired boiler", Atmos. Environ. 9_ 111 (1975).
38. Harrington, W., "Fine Particles", J. Air Pollution Control
Association, 1974.
39. Winchester, J.W. and Nifong, G.D., "Water pollution in Lake
Michigan by Trace elements from pollution aerosol fallout",
Water, Air, and Soil Pollution 1 (1971).
40. Natusch, D.F.S., et. al., Science 183. 202 (1974).
41. Lee, R.E. and Von Lehmden, D.J., J. Air Pollution Control Assoc.
23 853 (1973).
42. Toca, P.M., Thesis, University of Iowa, 1972.
43. 'Ruud, C.O. and Williams, R.E. "X-ray and microscopic
characterizations of Denver (1973) Aerosols," preprint, Report
Denver Research Institute, 1974.
44. Draftz, R.G., "Analysis of Philadelphia suspended dusts sampled
at street level:, I.I.T.R.I. Report No. C9915-1, 1974.
45. Draftz, R.G. and Blackeslee, H.W. "Identification of ambient
suspended particles from Philadelphia" preprint, I.I.T.R.I.
Report, 1974.
46. Harrison, P. Draftz, R. and Murphy, W.H. "Identification and
Impact of Chicago's Ambient Suspended Dust", preprint, I.I.T.R.I
1974.
47. Draftz, R.G. and Durham, J., "Identification and sources of
Denver Aerosol", preprint, paper #74-263, Air Pollution Control
Association Meeting, Denver, 1974.
48. Whitby, K.T., "Modelling of Atmospheric Aerosol Particle Size
Distributions", E.P.A. Progress Report, R800971.
-170-
-------
REFERENCES (CONTINUED)
49. Brock, J.R. in G.M. Hidy, ed., "Aerosols and Atmospheric
Chemistry", Academic Press, New York, 1972.
50. Cox, R.A., "Particle formation from homogeneous reactions of
sulphur dioxide and nitrogen dioxide", Tellus XXVI, 235 (1974).
51. Van Luik, F.W. and Rippere, F.E. Annl. Chem., 34 1617 (1962).
52. Miller, D.F. et. al., "Haze Formation Its Nature and Origin",
Final REport to C.R.C. and E.P.A., March, 1975.
53. Durham, J., Brock, J.R., Judeikis, H., and Lunsford, J.,
"Review of Sulfate Aerosols", EPA Report, In Preparation.
54. "Proceedings of the 7th International Conference on Condensation
and Ice Nuclei", K. Spurny, ed., Academica, Progue, 1969.
55. Brock, J.R. and Marlow, W.A. "Charged Aerosols and Air
Pollution", Environ. Letters, To Appear, 1975.
56. Gartrell, G. and Friedlander, S.K., Atmos. Environ. 9 279 (1975).
57. Middleton, P. and Brock, J.R. "Atmospheric Aerosol Dynamics:
the Denver Brown Cloud", EPA Report, To Appear.
58. Tuesday, C.S., ed. "Chemical Reactions In Urban Atmospheres",
New York, Elsevier, 1971.
59. Altshuler, A.P. and Bufalini, J.J., Photochem. Photobiology,
4 97 (1965).
60. Air Quality Criteria for Photochemical Oxidants, N.A.P.C.A.
Publication No. AP-63, March 1970.
61. Alley, F.C. and Ripperton, L.A. "The effect of temperature on
photochemical oxidant production in a bench scale reaction
system", J. Air Poll. Cont. Assoc., 11, 581 (1961).
62. Brock. J.R., Faraday Symposia of the Chemical Society No. 7,
"Fogs and Smokes", The Chemical Society, London, 1973.
63. Lundgren, D.A., "Atmospheric aerosol composition and con-
centration as a function of particle size and time", J. Air
Pollution Control Assoc. 2_0 603 (1970).
64. Esmen, N. and Corn, M. "Residence time of particles in the
atmosphere", Atmos. Environ. 5 571 (1971).
65. Hidy, G.M. , and Brock, J. R. , "An Assessment of the Global Sources'
of Tropospheric Aerosols" Proc. of 2nd Clean Air Congress,
IUAPPA, Washington, D.C., December, 1970.
-171-
-------
FIGURES
-172-
-------
1.0
c
o
c
o
CO
o
o.
OJ
o
15 rpm
1450 cm3 TV
Tracheo-
bronchial
0.01
0.1 1.0 10.0
Aerodynamic diameter (urn)
100.0
Figure 1. Deposition of monodisperse aerosols of various diameters in
the respiratory tract of man (assuming a respiratory rate of
15 resyirations per minute and a tidal volume of 1450 cm3). (4)
(5C 72 Q4 . 9* /OS 120 132
A //) w/crons
Flo. 2 2'o^ii ewtlcrlny evident Jor spl.rcrica} particles
plotted c'jairji r, a uiwi A ( l y ^
-173-
-------
10
e
c
z a
-a -a
10*
10:
10 4
103
Figure 3. Comparison of
Urban and Continental
Aerosol Size Distribution (20).
101
10°
o
o
Aerosol samples
Light to moderate
smog, Pasadena,
California
Sept. 3, 1969
0400
0820
V 1240
O 1900
10 2
10-l
-174-
10* 101
D£ Particle diameter (ym)
-------
5.00
4.00
3.00
E
j-
O
.LJ
SI
CJ
en
_0
<
a.
Q
ao
o
<
2.00
1.00
REFRACTIVE INDEX =1.50
RUN 11.10 A.M. AUG. 20,1969
Q 365 mn
O 436 nm
£> 546 nm
^ 675 nm
10
13' C
10 l 2 5 10° 2
PARTICLE DIAMETER, Dp (microns)
10'
Figure 4. Ab /Alog D , as a function of particle diameter for a particle
S C3. L p
refractive index of 1.50.
-175-
-------
GENERATION
PRIMARY
SECONDARY
•
t
I
M
CTi
REMOVAL
ATMOSPHERIC
AEROSOL
DYNAMICS
AEROSOL SIZE
AND
COMPOSITION
DISTRIBUTION
VISIBILITY
CLIMATE
HEALTH
ECONOMICS
AMBIENT
AIR
QUALITY
REQULATIONS
GROWTH
FIGURE 5. RELATIONSHIP OF CONTROL STRATEGIES TO ATMOSPHERIC AEROSOL DYNAMICS AND CHARACTERISTICS
-------
100
00
3.
z
o
H
Z
w
o
o
o
ij
o
00
§
ClJ
<
H
Z
W
co
a
W
o
w
>
<
80
60
40
60
40
20
c
oo
z
H
X
70
O
T3
O
O
PI
o
13
M
on
O
G
73
O
M
cn
oo
00
M
i
00
Z
00
1970
1980
1990
2000
2010
FIGURE 6. PROJECTION OF AVERAGE U.S. AMBIENT AEROSOL CONCENTRATION
DUE TO ANTHROPOGENIC PRIMARY SOURCE EMISSIONS.
-177-
-------
Chemical Conversion
of gases to low
volatility vapors
Condensation
Volatility
Vapor
/Primary Particles
I • .•_ . « •
Homogeneous
Nucleation
Coagulation
Condensation growth
of Nuclei
Chain Aggregates
•5
Wind blown dust
+
Emissions
+
Sea Spray
+
Volcanos
+
Plant particles
Transient Nuclei or
Aitken Nuclei Range"
.1 12
Particle diameter, Micrometer
Accumulation
Range ""
100
FINE PARTICLES
Mechanically Generated
Aerosol Range
COURSE PARTICLES
FIGURE 7. SCHEMATIC OF AN ATMOSPHERIC AEROSOL SURFACE AREA DISTRIBUTION SHOWING PRINCIPAL
MODES, MAIN SOURCES OF MASS FOR EACH MODE, AND THE PRINCIPAL PROCESSES INVOLVED
IN INSERTING MASS IN EACH MODE, AND THE-PRINCIPAL REMOVAL MECHANISMS. (48).
-178-
-------
Mass
Coarse
Particle Diameter,
10
Figure 8. Atmospheric aerosol mass size distributions are usually bimodal
with about one-third of the total mass in the fine particle ....
(or submicron) mode and two-thirds in the coarse particle mode.
1-1
0)
u
to
01
s-l
oo
oo
•H
01
C 4-1
OJ CD
0 E
^ «
0) -H
PH "O
Lead
Total
Particle
Iron
10.0 30.0
100.0
Particle diameter, microns
Figure 9. Average Size Distribution for 10 Impactor Samples.
(63)
-179-
-------
2
O
2
O
o
>>!
•H
C M •
O CJ-
•H 01 O
4J O r-l
a
c >, x
•H 4J
U -H Q
X rH 00
W -H O
JJ cfl
4T 43 t3
60 O --.
•H M M
i-3 CU T3
100
' • . /."••
/ ; dB i^xXX -•-,
10f
10"
/ .' dlogD
0.01 y -0.1 1.0 10
10
10
-it
o 2
o> c
3 3
en o"
H- fD
rr f-1
Qj O
2 a-
a cr
H-
M I—1
O H-
TO rr
a ^
o
o
u u
J M
os -J M
« 2
W O at
HMO
< 2 2
3 M M
Relative
Abundance
Sources
Relative
Abundance
Sources
Relative
Abundance
Sources
Relative
Abundance
Sources
1 D, Particle Diameter, urn
I I I i
100
' i I
Organic secondary sources
Combustion prjimary soutces
100
I
I
Inorganiif secondary sources. ...•'' I Dusts
ComBustiin Primary! ..•'!' Sea Aejosol
I i . i
100
I
I
I
LOO
I I 1
Combustion primary sources
I I
Combustion primary
I I
I _.-• I
j. •' Soil primary
, •{' .Dusts'
Figure 10. EXAMPLE OF PARTICLE SIZE AND COMPOSITION DISTRIBUTIONS IN
PHOTOCHEMICALLY ACTIVE URBAN AIR.
-180-
-------
o
00
O
at
W H
< Z
W
>• M
Oi 03
z
O CJ
CJ M
M Z
CO W
CJ
H O
2 P-
W
Z
OO
M Pi M
P3 SC EH
S H <
< 2 M
z
w
CJ
o
a. o
o H
• w
• z
D O
CJ
oo
00
H O 2!
Z M
< H nJ
< o
.-_ OO
H O
Z Pi
^ [V| (V]
CJ <
fa Z
o o >>
CJ OS
o <
M CO S
H CO M
-------
TABLES
-182-
-------
Table 1. OBSERVED RELATIONS BETWEEN POLLUTANT LEVELS AND HEALTH EFFECTS
Pollutant
Concentration level producing
adverse health effects
Adverse health effects
Reference
Pacclculate SO-100 ug/m1 participates (annual
matter and geometric mean)
sulfur
oxides 130 -jg/m3 (0.046 ppra) of S02 (annual
mean) accompanied by particulate
concentrations of 130 ug/m3
190 ug/m3 (0.068 ppm) of S02 (annual
mean) accompanied by particulate
concentrations of about 177 ug/m3
105-265 ug/m3(0.037-0.092 ppm) of S02
(annual mean) accompanied by
particulate concentrations of 185
Ug/m3
140-260 ug/m3 (0.05-0.09 ppm) of
SOa (24-hr, average).
300-500 ug/m3 (0.11-0.19 ppm) of S02
(24-hr mean) with low particulate
levels
300 ug/m3 particulates for 24 hr.
accompanied by SOa concentrations of
630 ug/m3 (0.22 ppm)
Increased death rates for persons
over 50 years of age.
Increased frequency and severity
of respiratory diseases in school-
children
Increased frequency and severity of
respiratory diseases in school-
children
Increased frequency o.f respiratory
symptoms and lung disease
Increased Illness rate of older
persons with severe bronchitis
Increased hospital admissions for
respiratory disease and absenteeism
from work of older persons
Chronic bronchitis patients suffer-
ing acute worsening of symptoms
Uinkelstein (5)
Douglas and
Wallter (6)
Lunn, et al. (7)
Petrilli, et al. (8)
Carnow, et al. (9)
Brasserl et al. (10)
Lewther (11)
Table 2. RELATION BETWEEN MORTALITY OR MORBIDITY AND AIR POLLUTION INDICES
AS DETERMINED BY LAVE AND SESKIN (12).
Disease
Air pollution Index
Relation
Bronchitis
Lung cancer
Nonrespiratory
tract cancer
Cardiovascular
'disease
Total respiratory
disease
Total mortality
Sulfur concentration;
total concentration of
solids
Urban versus rural areas
Suspended particles;
smoke density
Urban versus rural areas
Sulfates; particles.
Sulfates; particles
Strong relation between bronchitis mortality and
several indices of air pollution. Bronchitis mor-
tality could be reduced from 25 to 50 percent,
depending on the location and pollution index, by
reducing pollution to the lowest level currently
prevailing in the region.
With adjustments made for age and smoking history,
incidence of lung cancer is about 1.5 times as
great in urban as in rural areas.
Stomach cancer significantly related to a particulate
deposit index.
A substantial abatement of air pollution would lead
to 10 to 15 percent reduction in the mortality and
morbidity rates for heart disease. -.
Strong relation between incidences of emphysema and
bronchitis and air pollution. Also relations for
pneumonia and influenza cited.
A 10 percent decrease in the minimum concentration
of particles would decrease the total death rate by
0.5 percent. A 10 percent decrease in the minimum
concentration of sulfates would decrease the total
death rate by 0.5 percent.
-183-
-------
Table 3. BEST JUDGEMENT ESTIMATES OF POLLUTANT THRESHOLDS FOR ADVERSE
EFFECTS OF LONG-TERM EXPOSURES (13)
Threhold (annual average), Mg/m
Effect
Increased prevalence of
bronchitis in adults
Increased acute lower
respiratory disease in
children
Increased frequency of acute
respiratory disease in
families
Decreased lung function of
children
Sulfur
dioxide
(80)a
95
95
106
200
Total
suspended
particulates
(75)a
100
102
151
100
Suspended
sulfates
(no
standard)3
15
15
15
13
national Primary Air Quality Standard. The particulate standard is a geometric
mean; the equivalent arithmetic mean would be about 85 ug/m3.
Table, 4 . BEST JUDGMENT ESTIMATES OF POLLUTANT THRESHOLDS
«,
FOR ADVERSE EFFECTS OF SHORT-TERM EXPOSURES (13)
Threshold,
Effect
Aggravation of cardiopulmonary
symptoms in elderly •
Aggravation of asthma
Sulfur
dioxide
(365}a
>365
180-250
Total
suspended
•paniculate
(260)a
80-100
70
Suspended
sulfates
(no standard)3
8-10
8-10
aNational Primary Air Quality Standard.
-184-
-------
Table 5. POSSIBLE EFFECTS PRODUCED BY INHALED PARTICULATE MATTER AFTER
DEPOSITION IN RESPIRATORY TRACT COMPARTMENTS (16)
Compartment in which
Deposition Occurs
Nasopharyngeal
Tracheobronchial
Pulmonary
"Soluble" Particle
1. Damage mucosa and
cilia
2. Alergic response
1. Reflex bronchocon-
striction
2. Alergic response
3. Damage to mucosa and
cilia
1. Damage alveolar epi-
thelium
2. Peripheral respiratory
unit constriction
"Insoluble" Particle
1. Transferred to G.I. tract
2. Removed with sputum
3. Alergic response
1. Short-term clearance to
G.I. tract.
2. Removed with sputum
Long-term Retention
Peripheral 1.
airway and
alveolar
2.
constric-
tion
React with tissue
to cause local
effects
Remain in tissue
(inert)
Transported to
lymph nodes
Short-term Retention
Peripheral Phagocytized and transpo
airway and ed to terminal bronchiol
alveolar with subsequent clearance
cons trie- from T-B Compartment.
tion
I
-185-
-------
Table 6: Speculated Global Inventory of M.-ijor Sources after
Adjustment to Known
1. Primary
Oust Rise by Wind
Sea Spray
(Sea SO?-)
Extra terrestrial
Volcanic Dust
Forest Fires
Combustion 4 Industrial
!. Secondary
Vegetation
Anthropogenic Hydrocarbons
Sulfur Cycle
Nitrogen Cycle
Ammonia
NO •* NOT
x 3
Volcano (Volatiles)
ADJUSTED TOTAL 4
ANTHROPOGENIC 4
Compostion of Aerosols (65)
Total Production
Rate (tons day"1-) (% by We.)
106 21.
2 x 106 42.
(3.)
550 0.01
104 0.2
4 x 105 8.8
3 x 105 6.
2 x 105 4.
104 0.2
6 x 105 13.
105 2.
2 x 10 4.
103 .02
.7 x 106 100.
.3 x 105 9.
-186-
-------
Table 7
PrniiTtions of pnrtlcul.ltc emission Irvnls (cmi'.'.ions in I"'1 ton-./yrnr). \ ^ ° /
Industry
Elrrlnc irlililicS
induM/i.ti power generation
Crushed stono
Agricultural operation
Iron ,ind Steel
Cement
Pulp mills
Lime
Prim.vy nonferTOUS metals
Aluminum
Copper
Zinc
Le.id
Cl.iv
Fertilizers
Phosphate rock
Aspfi.ilt
Ferro.illoys
Iron foundries
Secondary nonfernxjs matali
Copper
Aluminum
Lcod
Zinc '
Coal cleaning
Carbon black
Petroleum
Acids
1
3. 36
3.22
5.71
1.8-1
1.31
0.9G
0.56
0.05
0.17
0.34
0.05
0.03
0.54
0.30
O.OG
0.22
0.16
0.15
O.OG
0.07
0.00
0.01
0.11
0.09
0.05
0.02
19/0
2
.1.11!
.1. on
4.72
1.67
i.?4
0.87
0.50
0.59
0.14
0.29
0.0-1
0.03
0.43
0.26
0.05
0.20
0.15
0.14
O.OG
O.OG
0.00
0.01
0.10
ft
0.04
0.02
3
.'..'"I
2.(.1
4.70
1.58
1.19
O.fiO
0.47
0.56
0.12
0.2R
0.0-1
0.03
0.41
n.,-3
0.05
0.17
0.14
0.13
0.05
0.06
0.00
0.00
0.09
»
0.04
0.01
)
1. •"->•)
•t.i;1
')./',
r;.-;r,
.1.41
O.i7
0.25
O.i9
0.13
o.;s
0.07
0.01
0.01
0.00
0.05
0.04
0.01
0.10
0.02
0.02
0.00
0.00
0.00
0.00
0.09
•
0.01
0.00
4. I
2',.\-,
.?.'' 7
i . '•'->
?..•••!
:.2--.
1.33
0.47
O.SO
O.C5
O.C4
1.77
0.50'-
0.03
0.73
0.25
0.21
0.09
0.12
O.Oi
0.0!
0.23
o.o:
O.i-5 •
0.07
'/'i'i'i
'•'
..':
'i. i
0.;' '
o.?/
''* / "•
0.2-3
0.17
0.22
o.;o
0.01
0.01
3.09
0.06
0.03
o.o;
0.17
0.03
0.03
0.00
•0.00
o.co
O.OJ
0.12
•
0.02
0.01
''
f, •'. -',
"I. ''.'.'
'j. •'•''!
f>.\2
o.;?
0..70
0.12
0.15
O.iJij
0.01
o.o;
0.00
J.&--.
0.04
0.01
0.14
o.o:
0.02
o.co •
G.GG
o.co
O.GO
0.11
»
0.02
0.03
.* Net calculated because Channel Proca** la being phased out
Table 8 COMPOSITION'S UF EMISSIONS FOR SOME CHICAGO POLLUTION SOURCES,
PERCENT (39)
Auto
Al
As
Br 7.9
Ca
Cd
Cl 6.8
Co
Cr
Cu
Fe 0.4
HS
K
La
Mg
Mn
Na
Ni
Pb AO
Sc
Tl
V
Zn 0.14
Ccxi L and
Canon t coke
2.5 14.0
0.016
44 4.0
0.004
0.009
0.03
0.04
'i.i 7.0
0.0000,;
1.2 0.8
0.024
0.4
0.04
0.12
0.9
0.08
0.09
Fuel
oil
5.0
0.4
0.15
0.12
0.16
2.5
0.10
0.3
0.03
1.5
6.0
0.18
0.03
2.5
0.05
Iron and
steel Soil
2.4 5
5.4 0.8
0.002
0.005
1.6 0.003
33.7 3
O.OOC04
2
0.004
1.6 0.7
2.4 0.03
0.6
0.005
0.005
O.OOJ5
0.3
0.007
1.8 0.01
"-187-
-------
Primary Sources,
106 Tons/vr
Indus trv
To
Acbitin t
Aorosol
Mass
Mass Mod inn
1970
1980
L990
t
2000 1
Dinr,;ctor, ^n
1970 CoiKi-oL
.•'.^oidl.i-.Oo li.T.t
(Rcf.)
Coal Fired Elect. Ucilicy
Ind. Power
Crushed Scone
Agric. Opn.
Iron i Steel
Cement
Pulpmills
Lime
Non Ferrous Metals
Clay
Fertilizers
Phosphate rock
Asphalt
Ferro alloys
Iron Foundries
Coal 'Cleaning
Petroleum
Acids
Solid. Waste Disposal
Construction Dust
Transportation-
Combustion
Transporta tion-
Tire Dust
Transportation-
Road Dust
3.36
3.22
5.71
1.84
1.31
0.96
0.56
0.65
0.73
0.54
0.30
0.06
0.22
0.16
0.15
0.11
0.05
0.02
1.4
0.8
1.2
0.3
1.0
4.40
4.12
9.66
2.05
1.38
1.39
0.81
1.04
0.97
0.79
0.35
0.07
0.32
0.19
0.18
0.16
0.07
0.03
1.4
1.1
1.6
0.4
1.34
5.98
4.12
15.64
2.26
1.40
1.90
0.93
1.22
1.18
1.17
0.42
0.08
0.48
0.22
0.19
0.22
0.11
0.05
1.5
1.4
2.0
0.5
1.68
7.65
4.00
25.16
2.47
1.48
2.43
1.24
1.38
1.36
1.77
0.50
0.09
0.73
0.25
0.21
0.28
0.16
0.07
1.6
1.7
2.4
0.6
2.0
2
2
10
10
0.4
3.0
1.0
2.0
1.0
3.0
1.0
2.0
1.0
0.4
0.4
2.0
0.5
0.4
0.3
5
0.4
3..0
5.0
(37)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(36)
(57)
(57)
(57)
(57)
(57.)
115
115
625
625
20
175
54
115
54
175
54
115
54
20
20 •
115
27
20
15
300
20
175
300
TOTAL
24.65
33.82 44.65
59.53
-188-
-------
Table 9: Continued
Elect. Utility
Industrial Power
Crushed Stone
A;-.ricultural Operations
Iron & Steel
Cement
Pulp Mills
Lime
Non Ferrous Metals
Clay.
Fertilizers
Phosphate Rock
Asphalt
Ferro alloys
Iron Foundries
Coal Cleaning
Petroleum
Acids
Solid Waste Disposal
Construction Dust
Transportation-Combustion
Transportation-Tire Dust
Transportation-Road Dust
TOTAL
Total Mass or Ambient
Suspended Particulace "attcr, 10 Tor.s
1970
0.029
0.028
0.0091
0.003
0.065
0.005
0.010
0.006
0.013
0.003
0.006
0.004
0.008
0.007
0.002
0.001
0.093
0.003
0.06
0.002
0.003
19 SO
O.Q38
0.036
0.015
0.003
0.069
0.008
0.015
0.009
0.018
0.004
0.006
0.006
0.009
0.009
0.001
0.003
0.001
0.093
0:003
0.08
0.002
0.004
1990
0.052
0.05
0.025
0.004
0.070
0.011
0.017
0.010
0.022
0.007
0.008
0.009
0.011
0.009
0.002
0.004
0.002
0.100
0.004
0.10
0.003
0.006
20.-,;
0.066
G . ui<-
0.0 •'; "'
' i , • !'• ' ••
0.074
O.Oi-
0.023
0.012
0.025
0.010
0.009
O.OOi
0.013
0.012
0.010
0.003
0.006
0.003
0.1C6
0.006
0.12
0.003
0.007
0.36
0.432
0.526
0.634
-189-
-------
Table 10. Examples of Gas Phase Reactions of Possible
Importance in Condensational Growth of
Atmospheric Aerosols.
Reaction . Reference
Free
S02 4- Js02 S03 (50)
S03 + H20 -»• H2S04
HNO + HN02 (51)
HNO
RHR1 4- 0- •*• involatile oxygenated (52)
products
-190-
-------
Table 11. Possible SO. Oxidation
Reactions in Aqueous Solution Atmospheric AerosoIs,(53)
Postulated Reactions (In Aqueous Phase)
Cationic
Catalys t
2S02 + 2H20 + 02 •* 2H2S04
Cationic
Catalyst
2S00 + 2H,0 + 00 + 4NH. + 2(NH.)SO,
2223 44
2S02
H2S04 + °2
2H20(ads) + S02 ->• H2S°4 + 2HN02
H2S°3 + H2°2 * H2S°4 + H2°
Cationic
Catalyst
^
Water Soluble
Organic Acid
-191-
-------
ADDENDUM TO SUSPENDED PARTICULATE MATTER REVIEW MADE
AT REQUEST OF ENVIRONMENTAL PROTECTION AGENCY
J. R. Brock
Chemical Engineering Department
The University of Texas
Austin, Texas 78712
August 5, 1975
-192-
-------
PROJECTION OF AMBIENT ATMOSPHERIC PRIMARY AEROSOL
MASS CONCENTRATION WITH 100% CONTROL BY 1980
This projection of ambient atmospheric primary aerosol
mass concentration assumes the control strategy proposed by Ander-
grift e£ al. In this method application of control is increased
to 100% by 1980. Replacement period for existing plants is taken
as % the IRS lifetime. Efficiency of new controls is increased
based on a technological forecast curve for each industry. This
equipment will be installed each year only on replaced capacity
(replaced following % normal IRS lifetime), new capacity, and on
some presently uncontrolled capacity so that the amount of control
corresponds to 10070 control by 1980.
With the assumption that the average efficiency of control
will increase according to a technological forecast curve, it
becomes a problem to forecast the corresponding change in the mass
median diameters of the aerosol from the: various categories of
primary emissions. From the inadequate data which do exist, it
appears that application of control to primary sources results in
substantial decrease in the mass median diameter of a controlled
emission. For example, application of current control of electric
utility power plants by electrostatic precipitators results in
reduction of mass edian diameter in emissions by a factor of 10.
Control of cement kilns by electrostatic precipitators result
in reduction of the mass median diameter by a factor of 5. Control
of limestone kilns by wet scrubbers results in reduction of mass
median diameter by a factor of 10. Increases in efficiency of
control devices would, of course, presumably further lower the
mass median diameters in these cases.
It will be assumed that by the year 2000 sources of
condensation aerosols will be controlled down to a mass median
diameter of 0.2 um corresponding approximately to the minima in
-193-
-------
the basic particle collection mechanisms of diffusion and impaction
and of electrical mobility. Comminution sources will be assumed
to have their mass median diameter reduced by a factor of 5, as
a conservative estimate. The change in mass median diameter for
each source category is assumed to occur linearly between 1970
and 2000, reflecting the assumptions of this projection.
The results of this projection are presented in the
attached figure. The somewhat surprising feature of this pro-
jection is that although total anthropogenic primary emissions
decrease, the ambient concentration due to anthropogenic primary
emissions first decreases and then increases so that by the year
2000 this concentration is approximately 85% of the value in 1970.
The explanation, of course, is that increased application and
efficiency of control lowers the average mass median diameter of
the various primary sources, thereby increasing the residence
time of the particles in the atmosphere (particle residence time
is believed to be a very sensitive function of the mass median
diameter). At the same time, economic growth implies increases
in numbers of sources to that after 1980 there is very little
decrease in the mass emission rates of the primary sources, although
control efficiency is increasing.
Clearly this attempt to predict the future course of
ambient atmospheric particulate levels is fraught with difficulty.
However, this projection together with the initial projection of
the review (based on the assumption that (efficiency of control) x
(application of control) is constant, will serve to set some
bounds on possible future trends of ambient atmospheric aerosols.
-194-
-------
O
M
H
Z
O
H
Z
W
u
z
O
CJ
hJ
O
C/5
O
oi
W
H
Z
CO
3
w
1
w
1.0
0.8 -
0.6 -
0.4 -
0.2 -
c
•
C/3
I
H
I
*d
O
cn
10
n
ra
co
M
O
z
H
O
Z
w
po
1970
1980
1990
2000
2010
YEAR
PROJECTION OF AVERAGE U.S. AMBIENT AEROSOL CONCENTRATION
RATIO DUE TO ANTHROPOGENIC PRIMARY SOURCE EMISSIONS UNDER
ASSUMPTION THAT APPLICATION BASE YEAR OF PROJECTION IS 1970.
-195-
-------
APPENDIX D
"PARTICULATE MATTER: RELATIONSHIPS BETWEEN
EMISSIONS AND AMBIENT AIR QUALITY"
L.R. BABCOCK, JR.
AUGUST 1975
-------
PARTICULATE MATTER: RELATIONSHIPS BETWEEN
EMISSIONS AND AMBIENT AIR QUALITY
Prepared for:
Radian Corporation
Austin, Texas
and
Control Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
by:
Lyndon R. Babcock, Jr.
Professor
University of Illinois
P.O. Box 6998
Chicago, Illinois 60680
August 1975
-196-
-------
EXHIBITS
1. Air Quality System
2. Idealized Particle-Size Distributions
3. Primary Emission Distribution
4. Preliminary Characterizations
a. California Composite (Before Distribution
of Secondary Species)
b. Aggregated California Composite and Nationwide
Inventory
5. Detailed Characterizations
a. California Composite
b. Nationwide (Unnormalized)
c. Nationwide (Normalized)
6. California-Composite Characterization Diagram
7. Nationwide Characterization Diagram
8. Summary Comparison
-197-
-------
SUMMARY
This report addresses source-characterization relationships
of atmospheric aerosols; fine particulate matter and secondary
pollutants receive emphasis. The report is directed to individuals
involved in defining aerosol-related research priorities and
control strategies.
Existing published information has been used to generate
the central results, which consist of two detailed source-species-
size characterizations, one for a composite of five California
cities and a second, more representative of nationwide emissions.
A summary of the characterizations follows on Page 199.
Significant differences between the two characterizations
can be noted, but in each case, combustion-and-industry and natural-
and-miscellaneous have been identified as the largest source
categories. Secondary pollutants comprise a significant fraction
of both totals, and primary fine particulate matter from stationary
sources comprises a sizeable constituent in the nationwide
characterization (16 percent of the total).
In each characterization, the total aerosol mass has
been specified with regard to source, species, and particle size.
This "closure" was accomplished only after incorporating many
assumptions into the calculations. The methodology and assumptions
are clearly defined, and readers are encouraged to critically
review the work and to suggest alternative approaches and
conclusions.
The literature review was performed in an unusual manner,
in that experts in the fields of atmospheric chemistry and aerosol
characterization reviewed a common body of pertinent literature
from which three separate reports were individually prepared. These
reports served as inputs from which the writer arrived at the
discussion and conclusions presented herein.
-198-
-------
SUMMARY COMPARISON
Category
California Nationwide
Combustion and Industry
Transportation
Natural and Miscellaneous
Water
26
23
39
12
49
15
24
12
TOTAL:
Secondary PM
Primary fine PM (from
stationary sources)
Other Primary fine PM
Large PM
Water
100
32
4
8
44
12
100
25
16
5
42
12
TOTAL:
100
100
-199-
-------
INTRODUCTION
The Control Systems Laboratory (CSL) of the U.S. Environmental
Protection Agency (EPA) has a responsibility for the development
of technology for the control of air pollutant emissions from
stationary sources. Major research programs sponsored by CSL
are directed toward improved control of sulfur dioxide, nitrogen
oxides, hydrocarbons, and particulate matter (PM), as well as
toward control problems associated with chemical, metallurgical,
and energy processes.
Technology for the control of large-size PM (above two
micron in diameter) has been available for many years; therefore, in
addition to conventional particulate control, a CSL program
emphasizes novel approaches for control of fine PM. The smaller-
sized particles are much more difficult to control, remain
suspended in the ambient atmosphere for longer periods of time, are
largely responsible for visibility degradation, and are thought to
be largely responsible for the adverse health effects associated
with PM.
Further, small particles have a high surface area per
unit mass, and this surface may catalyze gas-to-particulate re-
actions. Also, several individual toxic trace metal species such
as lead, zinc, vanadium, etc., appear to be concentrated in the
smaller particles. Another cause for concern may be related to
small particles of combustion carbon which have been found to
remain for long periods at high altitudes in the troposphere. It
has been postulated that these particles absorb solar radiation
with the resulting layer of warmer air causing a semipermanent
increase in atmospheric stability.
Clearly, there are health, esthetic, and possibly even
climatic, incentives to reduce ambient levesl of fine PM at many
locations throughout the nation, and the CSL fine PM program is
directed toward that goal. However evidence is now accumulating
-200-
-------
which indicates that significant amounts of fine PM can be
formed in the atmosphere. In some locations, this secondary PM,
formed from gaseous precursors, may be the dominant PM species.
Few deny the presence and significance of secondary PM constituents,
but we seem to be quite a ways away from having usable quantitative
source-ambient relationships, even for relatively heavily studied
areas such as California.
In the midst of this uncertainty, CSL seeks to allocate
its resources in an optimal manner. Some have suggested that,
in light of the secondary-PM evidence, the CSL fine particulate
program should be deemphasized. However, certain questions should
be answered before decisions on CSL priorities can be meaningfully
made: What are the quantitative size and species characterizations
of PM? What are the original sources of this PM? How much is
primary; how much is secondary? How much is derived from stationary
sources versus mobile and the so-called "natural" sources? How
do the findings vary from location to location?
This report summarizes an unusual short-term study which
has attempted to address these questions. Elsewhere, an exciting
long-range diverse research program funded by the Atmospheric Aerosol
Research Section (AARS) of the Chemistry and Physics Laboratory of
EPA is continuing. In contrast, the short-term study described
herein is confined to examination, interpretation, and integration
of existing results. Our short-term study may serve to suggest
additional experimental research to AARS.
Method
A common body of pertinent literature was identified by
personnel of CSL, AARS, and Radian Corporation. (These references
are listed on pages 241-246. Experts in the fields of aerosol
characterization and atmospheric chemistry agreed to independently
review this literature, and to independently prepare three reports,
-201-
-------
incorporating their own knowledge and experiences into inter-
pretations of the literature information. The consultants included:
(1) G.M. Hidy and P.K. Mueller, Environmental Research and Technology,
Inc., Westlake Village, California; (2) J.P. Lodge, Consultant in
Atmospheric Chemistry, Boulder, Colorado; (3) J.R. Brock, University
of Texas, Austin, Texas.
More recently, at the request of CSL personnel, a fourth
central input item, a paper by Gartrell and Friedlander was added
to this study.
I agreed to study the common body of literature and the
four input reports (citations listed on page 37) and to summarize and
interpret the findings for use by CSL personnel. Such is the purpose
of this summary report. It is suggested that it be read with close
referral to the four input reports. Duplication of material
included in the input reports has been kept to a minimum in this
summary report.
Source-to Ambient System
Before discussing specific findings, it seems desirable
to define some terminology and relationships.
Some basic aspects of the air quality system can be
related as shown on Exhibit 1. Flows of pollutants are indicated
by solid lines, while other influences are indicated by dashed
lines. Our concern is with the air quality experienced by receptors.
The presence of meteorological variables, transformations, and-
decay and deposition mechanisms multiplies the difficulties in
relating emissions to ambient air quality. ' Transformations of
many types are feasible; this report is largely concerned with the
transformation of gaseous pollutants, such as sulfur dioxide and
hydrocarbons, into solid and liquid fine PM, such as acid sulfates
and oxidized organic matter.
-202-
-------
On Exhibit 1, the two areas of greatest uncertainty
are indicated by the ellipses. Toward the left, four paths are
indicated for emissions: emissions may reach human receptors
directly; emissions may decay or otherwise remove themselves from
the atmosphere without reaching human receptors; emissions may
undergo various kinds of transformations; or certain species
of emissions may provide catalysts which promote the extent of
transformations. Similarly, on the right-hand side of Exhibit 1,
transformed pollutants may either influence the air quality
experienced by receptors, or can bypass and go directly to a decay
or deposition sink. Note that transformations are influenced not
only by airborne catalysts, but by solar radiation and meteorology.
In fact, meteorology influences almost every aspect of the source-
to-sink process. Also note again that there are two feasible paths
by which pollutants can circumvent human receptors. Quite clearly,
all emissions need not affect all receptors.
The paths to decay can be seen to be significant aspects
of the air quality management system. Yet many control policies
seem to ignore these aspects, while other control policies generate
interesting questions. For example, tall stacks are employed to
reduce adverse effects of pollutants such as sulfur dioxide upon
nearby receptors. Yet tall stacks reduce the likelihood of direct
sulfur dioxide deposition (a relatively rapid process compared to
sulfate deposition). By increasing atmospheric residence time,
the tall stack provides the opportunity for airborne transformation
as well as for eventual decay. A major concern then is extent
to which toxic decay intermediates, such as acid sulfates, affect
receptors prior to eventual ultimate decay and deposition.
Most workers in air-quality-management-related fields
are attempting to define some aspect of Exhibit 1; our source-
receptor study is no exception. The large number of workers and
the remaining large number of unanswered questions are indicative
of the complexity of the task. The limited effort summarized in
this report is directed at integrating existing information on
transformations, etc., toward development of some presently-useful
source-receptor relationships.
-203-
-------
The conventional approach has been to start with emissions,
add in some meteorological modeling and arrive at estimates of effects
of various sources upon receptors. This "source-to-receptor" approach
is clearly inadequate if complex transformations and decay mechanisms
are involved. Thus, more recently, "receptor-to-source"
methodologies have begun to evolve. Comprehensive characterization
of ambient particulate matter can supply evidence as to the extent
of transformation and removal as well as to the original sources.
Most workers have tended to emphasize one approach or the
other. There have been a few notable exceptions, and the goal of
this study is to explore and extend efforts to integrate the two
approaches. It is evident that neither approach alone can define
the complex source-receptor system described on Exhibit 1.
Other Definitions
Several types of PM are discussed in the input reports
and in this summary report. Some definitions and comments follow:
Bimodal distribution. Most aerosol scientists now agree
that ambient aerosols tend to be bimodal, with the minimum occurring
between one and two micron. An idealized bimodal distribution is
shown on Exhibit 2. The Brownian-motion mechanism seems to be in
part responsible for the minimum. That is, small particles can
grown by coagulation (impacting and adhering to each other), but
the Brownian motion which generates coagulation, decreases as the
particles grow larger. It seems likely that the coagulation process
functions for primary as well as secondary PM. Several shapes can
be postulated for the primary contribution to the fine PM mode
(dashed curves 1, 2 and 3 on Exhbiti 2). The term "fine PM", in
contrast to "large PM', as used herein refers to PM smaller than
two micron in diameter. Fine PM may be either primary or
secondary.
-204-
-------
Primary and secondary particulate matter. PM emitted
directly from sources and remaining relatively unchanges chemically
is termed primary. However, it is possible to postulate that these
particles, if small enough, grow through coagulation, or that
these small particles grow when gaseous species react on their
surfaces. A major question is the size distribution of the
primary fraction, particularly at the fine end of the distribution.
(See Exhibit 2.)- Combustion processes are major contributors to
primary fine PM, and comminution processes may make a limited con-
tribution.
Those particles formed in the atmosphere from gaseous
precursors are termed secondary. Even here the definition is
uncertain. Certain species may condense shortly after leaving a
stack, undetected by a stack emission test, but soon existing as
PM, virtually at the point of emission. It is generally agreed
that secondary particles are generated in very small, almost
molecular sizes, and they grow via coagulation until the Brownian
motion mechanism ceases. However, it is also possible to postulate
gaseous precursors adding to existing fine or even to large PM.
Thus, an unknown amount of the large-size fraction could be of
secondary origin. .
Combustion, comminution, and natural sources. Combustion
processes produce small particles, less than one micron in size
(Fennelly, 1975). A major species is the relatively inert com-
bustion carbon. These particles are not included in a benzene-
soluble or other "organic" characterization. Similarly, fine PM
other than carbon can be formed during high temperature condensation
or combustion. Many varieties of trace metals are likely. Vanadium
is of particular interest because of its presence in fuel oil and
its possible activity as a catalyst which promotes the oxidation
of sulfur dioxide.
-205-
-------
Comminution processes (such as crushing and grinding) tend
to produce large particles. Unlike coagulation, during comminution,
probabilities are higher that large particles will impact on each
other. It is often assumed that comminution processes produce no
fine PM; yet there seems no theoretical reason for an abrupt trun-
cation in a comminution distribution. Rather, one would expect
an asymptotic decay as particle size decreases. Lack of definition
may result from inadequate fine-size characterization techniques
used to monitor comminution processes.
Natural or "quasi-natural" (as used by Brock) represents
a third major source of particulate matter. Many "natural" sources
are actually anthropogenic (manmade): for example, dust which
would not be present except for intensive agriculture or other man-
made alteration of the environment. The natural category as
defined herein also includes reentrained or fugitive dust, that
PM which becomes re-airborne after initial deposition. Such sources
can sometimes be the major source in a region, and such sources can
be difficult to control. It is essential that the natural con-
tribution to ambient levels be accurately defined. Otherwise,
control of anthropogenic sources may result in disappointingly-
small improvements in overall ambient air quality.
Review of InputReports
The results and discussion presented in this report are
based largely upon four input reports (citations listed on page
37), three of which were prepared specifically for this study.
Each report is summarized briefly below. Each report describes
PM characterization and control in an individualistic, interesting
way, and each should be studied carefully in its entirety to gain
each author's overall message. Only those aspects closely
pertaining to the source-species-size characterizations are
emphasized in this summary report.
-206-
-------
Hidy and Mueller. This report emphasizes California-based
photochemical studies. Several useful experimental ambient particle
size distributions are presented along with composition distributions
In these distributions unknown components varied from 17 to 70 per-
cent of the total masses.
The chemical tracer method of Friedlander and others is
proposed as the best available technique for relating ambient
pollution levels back to sources. Results of the Gartrell and
Friedlander paper (1975) are discussed and mixed source-species
diagrams for Pasadena and Pomona are presented. Like Gartrell
and Friedlander, Hidy and Mueller did not attempt to allocate the
secondary species (organic, S0t» , etc.) back to sources.
Hidy and Mueller conclude by discussing the need for gas
emission control in order to reduce ambient levels of secondary PM.
A second report, by Hidy and Burton, "Atmospheric Aerosol
Formation by Chemical Reactions" (1974) is appended to the input
report. This latter report examines possible mechanisms for
secondary pollutant formation.
Lodge. After an interesting historical-philosophical
introduction, Lodge qualitatively discussed the sources of large
particles, emphasizing the likely presence at some locations of
significant or even dominant amounts of soil dust and pulverized
vegetation. He made the important differentiation between
"contemporaneous" carbon (that derived from recent biota) and carbon
derived from fossil fuels. He concluded that "an important portion
of the organic matter (in ambient PM) consisted of spores , micro-
organisms, pollens and comminuted plant material". Unfortunately
no size distribution data were taken in the described study.
A discussion of fine and secondary PM is also presented.
This point is made that species such as sulfates need not be
exclusively secondary (e.g., sulfates are contained in soil and
also are emitted from fertilizer and other industrial operations).
-207-
-------
Based upon results for Los Angeles and Milan, Lodge con-
cluded that the secondary components could comprise as much as 40
percent of the total aerosol. However, he found the works of
Harrison et al.* and Lodge (1960) to support the contention that
the fraction of secondary PM could be much smaller, even in Los
Angeles where Friedlander's methodology wouldn't distinguish between
photochemical organic secondary aerosol and primary organic
aerosol comprised largely of contemporaneous carbon.
Lodge went on to describe a Denver characterization in
which "nearly 90 percent of the particles" were above one micron
in size. Lodge concluded by discussing current and future standards
and the need for different control strategies in different areas.
The most significant aspect of Lodge's paper applicable
to this report was his insistence on consideration of the so-called
natural sources and fugitive dust, and that such sources might even
include significant quantities of organic matter. He concluded
that such sources could dominate in areas of low humidity. Also
well conveyed, was the principle that the source-species-size
characterization should be expected to vary extremely in different
regions of the nation, that no single characterization could be
expected to be applicable across the nation.
Brock. This report provides a comprehensive review of both
primary and secondary PM, including both theoretical and empirical
information. Brock supports conclusions of Lodge, as well as of
Hidy and Mueller, but Brock goes beyond the other consultants in
'' Note that . the other input consultants felt that the microscopy
techniques employed by Harrison, Draftz, and other investigators
did not adequately characterize the fine PM portion of the aerosol
mass.
-208-
-------
addressing primary PM emissions. Estimates for 1970, 1980, 1990,
and 2000 of the mass emission, particle size, and residence time
in the atmosphere are given for 23 source categories. Brock uses
these data to show how particle size affects the relative con-
tribution to aerosol mass. That is, that large sources of large
particles, such as "crushed stone", contribute very little effect
on ambient aerosol mass.
Gartrell and Friedlander. This paper was included belatedly
in the study at the request of CSL personnel. It was not included
in the original package of reports to be reviewed by the input
consultants. However, Hidy and Mueller and Brock discussed the
paper in their reports, and Lodge discussed an earlier related paper
by Friedlander.
The paper is the most-recent published report by Friedlander
and co-workers of their efforts to characterize ambient aerosols and
to relate them back to sources. This paper comes closer than some
of its predecessors toward obtaining closure (ie, specifying 100
percent of the aerosol mass).
The methodology involves use of chemical elements as tracers.
For example, if there is lead in the aerosol, most of that lead must
have originated in car exhaust. Further, proportional amounts of other
species in car exhaust which also appear in the aerosol must likewise
have originated as automobile emissions. In this manner, the total
aerosol mass can be distributed among sources. One also can deduce
that large anomalous concentrations of species such as sulfates must
have been created in the atmosphere from gaseous precursors.
Gartrell and Friedlander applied their method to several
California cities and these results form a basic part of the results
contained in this summary report.
-209-
-------
The methodology, still in its formative stages, has several
shortcomings. First, there is no distribution between fine and large
PM. Thus, one can't address Lodge's point of allocation of sulfate
between primary and secondary categories. Second, Gartrell and
Friedlander, in attempting to obtain 100 percent closure, allocate
some of the species to sources, but for organics, sulfates, nitrates,
and ammonium, they report only the percentage of each species.
Apparently, the authors felt that insufficient evidence was available
to make a meaningful distribution of these largely-secondary species.
Thirdly, application of the methodology has been largely restricted
to California cities, cities which depart considerably from the
nationwide average with regard to heavy industry, coal combustion,
and photochemistry.
Each of the input authors seems to support the results
obtained to date from application of the Friedlander methodology,
although Lodge questions the allocation of organics between secondary
photochemical and contemporaneous carbon "natural" (vegetation, etc.)
sources. As mentioned earlier, Gartrell and Friedlander did not
attempt to distribute this category amongst sources, but apparently
others have inferred that this organic category is largely
attributable to the automobile.
Results
The results consist of supporting data tables (Exhibits 3-5)
which lead to two characterization diagrams, one for a California
composite (Exhibit 6), and the second for a more-representative
nationwide approximation (Exhibit 7).
The characterizations are extensions of the results of
Gartrell and Friedlander. In each case, 100 percent closure for
source, species, and size is obtained by using data from Brock to
define the primary fine fraction and by incorporating a multitude
of assumptions.
-210-
-------
Exhibits 6 and 7 each consist of three bar graphs. The
upper bar describes the composition of the large-size particles,
while the lower bar describes the fine fraction. These two bars
are scaled such that when totaled together with the water, they
represent 100 percent of the total suspended particulate matter (TSP)
on a mass basis. The central bar represents the sources responsible
for the TSP. Similarly, the length of the central source bar
represents the total of all the sources of PM (including water, from
whatever source). Note that the sum of the lengths of the two
outer species-characterization bars is exactly the length of the
central source distribution bar. On Exhibits 6 and 7, one inch is
equivalent to ten percent of the TSP total. Also note that the
species are related to sources by the connecting lines on the diagrams
The reader may wish to refer directly to Exhibits 6 and 7
prior to reviewing the ensuing details as to how the exhibits were
arrived at. Additional interpretation and criticism of the results
is presented in the "Discussion" section of this report.
Definition of primary fine category. Primary-fine was one
of the categories left undefined by Gartrell and Friedlander.
Clearly, primary-fine PM is a constituent of most polluted ambient air,
but little useful quantitative documentation is available; none of the
input consultants estimated an amount. Recent emphases seem to have
been largely directed toward definition of secondary rather than
primary PM.
Brock did report the mass median diameters (MMD) for 23
source categories (his Table 9, derived largely from a 1971 Midwest-
Research-Institute study). Hidy and Mueller supplemented this listing
with measured distributions for marine, desert, and freeway environ-
ments. I have combined these data to arrive at size estimates for
the various source categories of interest.
-211-
-------
The fine fraction (amount below two micron) may be obtained
directly by comparing areas on the graphs of Hidy and Mueller (their
Figure 1). Thus:
Location Amount below
2 micron,
percent
Point Arguello (sea spray) .... 10
Harbor Freeway (motor vehicle). 66
Goldstone (desert) 50
An extrapolation method is required for the analogous
interpretation of Brock's MMD data, except where HMD equals two
micron. (For this exception, the fraction less than two micron
is obviously 0.50). -Brock (his page 15) implies that the particle
size distribution, for comminution processes at least, can be
approximated by a log-normal relationship. He doesn't specify a
slope, and it would be unlikely that one slope would characterize
all emission distributions. Nevertheless, I applied a single log-
normal extrapolation to Brock's emission data such that:
Particle size Amount below
MMD, micron two micron,
percent
10 6
2 50
0.2 99*
* Some feeling as to the reality of the extrapolation method can be
obtained by comparing some extrapolated fractions with the distri-
butions reported by hidy and Mueller (see my page 19).
For Point Arguello sea spray, the extrapolation method
(based upon an area-derived seven micron MMD) yields 11 percent fine,
while direct area measurement yields 10 percent fine.
-212-
-------
For transportation, the direct area measurement of the
Harbor Freeway size distribution yields 66 percent fine, while the
weighted average of Brock's three transportation categories as
extrapolated yields 38 percent (Exhibit 3). Even less satisfying,
direct application of the extrapolation method to the Harbor Free-
way distribution (using an area-derived 0.4 micron MMD) yields
94 percent less than two micron. The very irregular shape of the
curve, caused by a mixture of sources, apparently causes the extra-
polation method to yield the erroneous results.
The Goldstone desert comparison can not be used, since the
area-derived two micron MMD must and does agree with the area-
derived 50 percent below-two-micron fraction.
(End of footnote)
-213-
-------
Results of the extrapolations are summarized on Exhibit 3.
Addition of a 24th category, "miscellaneous, forest and agricultural
fires", expands the Brock listing into an emission inventory which
closely resembles that published by EPA and CEQ for 1969 or 1970
(Babcock and Nagda, 1972). This additional category, with an
assumed one-micron HMD, is included on Exhibit 3. The revised
inventory totals indicate that the fine fraction comprises 53 per-
cent of the primary-PM emission total. Thus, at a given ambient
location, the primary fine fraction could comprise a substantial
part of the TSP total. For example, one could postulate a total
particulate level of 100 pgram/meter3 with 60 ygram/meter3 in the
fine and 40 ygram/meter3 in the large fractions. The fine fraction
could be made up almost exclusively of primary material, with only
7 percent secondary material needed to obtain closure (60 - 53 - 7).
The percentages below and above two micron are needed to
allocate source emissions into size categories. These allocations
can be derived directly as weighted averages using data on Exhibit 3,
or by combining the percentages from Exhibit 3 with other summaries
of source data (such as on Exhibit 4).
One result may be of interest here. The data on Exhibit 3
indicate that nationwide, fine primary emissions derived from
stationary point sources comprise 25 percent of all the primary
emissions listed on Exhibit 3 (8.91/35.20 = 0.25).
California composite characterization. Each of the input
consultants referred to the work of Friedlander and co-workers, in
which ambient PM was allocated amongst various source and species
categories. Several related papers have been published describing
and applying this characterization method. Gartrell and Friedlander
(1975) has been utilized herein because it describes the most
recent and comprehensive application of the methodology.
-214-
-------
One of the significant features of their 1975 work is the
attempt to obtain closure. Both measured and calculated TSP masses
are presented for each of the five cities studied.
The following procedure was used to convert the Gartrell
and Friedlander results (their Table 4) into Exhibit 6:
a. The Pasadena and Pomona data were complete (no
unknowns) and, thus, the species and sources
constituents were easily converted into percentages
of the calculated total masses.
b. For Riverside, Fresno, and San Jose; organics and/
or water were listed as unknowns. These unknowns
were assumed equal to the same percentages as for
the known categories for Pasadena and Pomona.
c. After estimation of the unknowns, the percentages
of source and species constituents were determined
for Riverside, Fresno, and San Jose.
d. The percentage results for each city were averaged
together to arrive at a single California composite
(Exhibit 4a).
e. Categories were rearranged to simplify reporting and
to emphasize the point sources of interest to CSL:
"industry and agriculture" was first disaggregated,
with argiculture assumed to be 40 percent of the
total (from San Francisco, Riverside and Fresno data
reported by Gartrell and Friedlander, their Table 2);
agriculture, sea salt, and soil dust were combined
into a single so-called "natural" category; trans-
portation sources, including auto exhaust, diesel
exhaust, tire dust, and aircraft, were combined;
cement dust was added to the industry category
(although some cement dust might be generated from
erosion, construction and demolition of concrete
structures).
-215-
-------
Distribution of the chemical species proved
more difficult and subjective: of the 6 percent
SOi*, 0.5 percent was allocated to the natural
category and 5.5 percent to combustion; the 9.7
percent NOs, was distributed equally between
transportation and combustion.
The allocation of NHi» is extremely subjective, but
at the same time, quite essential to meaningful
source assessment. Various ammonium salts have
been defined as major constituents of secondary
acid sulfates. Yet few investigators have even
attempted to relate the ammonium back to sources,
beyond the relatively large natural emissions which
are known to exist. Photochemical reactions might
create ammonia from precursors emanating from other
kinds of sources. Thus, of the 4.5 percent NHi*,
1.5 percent was allocated to the natural category
and one percent each to transportation, industry,
and combustion.
Organics pose similar difficulties, since only rarely
are individual species defined, and measurement
involving benzene or other solvent extractions may
not adequately describe the total organic content.
Adding to the confusion, organics emanate from a
diversity of natural and anthropogenic sources and
are participants in a variety of ill-defined photo-
chemical reactions. Lodge makes the point that
emission inventories and other source assessments
have tended to underestimate the significance of
hydrocarbons and other organics which are derived
from natural sources. Thus, the 23.7 percent for
organics was distributed with four percent to
industry and the remainder split equally between
-216-
-------
the natural and transportation categories.
Water was included in the characterization because
it seems to comprise a significant percentage of
the TSP total mass. However, no attempt was made
at source allocation, since water usually isn't
considered to be a pollutant, and there seems to
be no meaningful way to differentiate between
natural and emitted water.
The aggregrated source categories, with species
allocated are summarized on Exhibit 4b.
g. At this point, it seemed illuminating to compare this
distribution (Exhibit 4b) based upon data of Gartrell
and Friedlander, with the nationwide emission
distribution (Babcock and Nagda, 1972), with all
pollutants summed except carbon monoxide. (Carbon
monoxide was assumed not to contribute to primary
or secondary particulate matter.). The results, with
water factored in, are also shown on Exhibit 4b.
There appear to be two notable discrepancies between
the two distributions. First, the natural-and-
miscellaneous category is much larger when based
upon California ambient measurements. This outcome
seems reasonable, since several natural and fugitive
sources are not included in the nationwide
"miscellaneous" emission category; also, the low
humidity common in California should cause relatively
high levels of reentrained dust. The second
discrepancy involves stationary combustion, with
the California result being only 60 percent of that
based on nationwide emissions. Two explanations
come to mind. California does not use much coal
-217-
-------
as a fuel, and not all the emitted sulfur oxides
and nitrogen oxides are transformed into
particulate matter. No doubt both explanations
have some validity.
In any event, the exercise of comparing California
ambient results with nationwide emission averages,
should instill some caution into those tempted
to extend California results to other regions in
the nation.
h. The next task was to distribute the species between
the fine-size and large-size categories. Here,
the previous assumptions as to species, results from
Hidy and Mueller (my page 19), and the extrapolated
data of Brock (Exhibit 3) were employed. Perhaps
the most subjective allocation was for organics.
The large number of sources and species and primary-
secondary uncertainty seemed to preclude a meaningful
distribution, so a 50-50 fine-large distribution was
assumed.
i. Next, the distinction between primary and secondary
was made. Again, previous assumptions with regard
to SCK , N03, NH4, and organics were utilized.
j. The results, derived as indicated above, are summarized
in tabular form on Exhibit 5a and were used to create
Exhibit 6, the characterization diagram.
-218-
-------
Nationwide characterization. A few attempts have been made
to apply the methodology used by Gartrell and Friedlander to other
parts of the country. However, none of these, as well as others
attempts at characterization, seem to have been as comprehensive as
the California characterizations. Particularly needed are complete
characterizations of those regions with heavy industry, more
coal combustion, and with meteorology less favorable to photochemical
reactions. Much of the nation's urban population is exposed to such
conditions, rather than to the relatively unusual California situations
studied by Gartrell and Friedlander.
Rather than extrapolate from one of the less-comprehensive
studies of a non-California region, it was decided to extrapolate the
California data itself--to-more-closely approximate the emission
conditions prevalent outside California.
The California composite was extrapolated into a nationwide
characterization by applying the following conditions, solving
algebraically, and then normalizing to 100 percent:
a. SCX/NOa = 1.4 (from nationwide emissions inventory:
S02/N0x = 1.4, Babcock and Nagda, 1972).
b. Water = 12 percent (from California composite, Exhibit
5a).
c. Fine PM from industry and stationary combustion = n 25
Total PM
(from Exhibit 3 as discussed on my page 21).
d. Industry fine PM =17
Stationary combustion fine PM
(from Exhibit 3).
e. Fine PM = large PM, for primary stationary combustion
(from Exhibit 3).
-219-
-------
f. Industry inorganic large PM 0 _
= /. j
Industry organic large PM
(from California composite, Exhibit 5a).
g. Industry total PM
» 2
Stationary combustion total PM
(from nationwide inventory, Babcock and Nagda, 1972).
The unnormalized results of the algebraic exercise are
shown on Exhibit 5b, where the seven calculated quantities (algebraic
unknowns) are .shown in parentheses. The other quantities are
duplicates of those for the California composite (Exhibit 5a). Note
that the calculated nationwide TSP level is 59 percent greater than
for the California composite. To enable comparison of the distributions,
the results were normalized to 100 percent and are shown on Exhibit
5c.
Several interesting contrasts are apparent. In the nation-
wide inventory, primary fine PM and sulfates, are the cause of
significant increases (above California) in the industry and
stationary combustion categories. Although the absolute amounts in
the natural and transportation categories remain unchanges, the
relative contributions are reduced significantly in the normalized
nationwide comparison.
The ratio of primary fine PM/large PM increases from 27
percent in the California composite to 50 percent for the nationwide
characterization. The calculated 50 percent approaches the 53 per-
cent result derived exclusively from Exhibit 3.
Additional discussion of the results is presented in the
following section.
-220-
-------
Discussion
This section contains some odds and ends which didn't seem
to fit anywhere else. Considerable expansion of this section might
be desirable, but time constraints dictated otherwise.
Residence time. Brock (his Table 9), based in part upon
the work of Esmen and Corn emphasized the effect of particle size
on atmospheric residence time:
Diameter Residence time
(Microns) (hours)
0.1 1622
0.5 324 (assumed HMD for fine PM)
1 162
10 14 (assumed HMD for large PM)
15 10
20 7
40 4
100 1.4
A large variation is apparent, but note that even the 100 micron
particles would be expected to stay airborne for a distance of 14
miles in a 10 mile/hour wind. Thus, it seems reasonable to assume that
all-sized particles stay airborne, at least within the immediate urban
region. This conclusion seems to support some of the assumptions
leading toward the characterization diagrams. However, Brock's
opposite conclusion also seems well justfied when applied to the
nationwide airshed. The small particles are the ones which increasingly
comprise the bulk of the aerosol, as distances from large sources
increases.
Limits. On a nationwide basis, it seems to be possible to
use the extrapolated Brock data to define "floating" limits for the
maximum percentages of primary.fine particulates from the various
-221-
-------
categories. For example, if there were no secondary particulates,
and if all particles remained airborne for the same length of time,
Exhibit 3 could be used to define the TSP problem. Calculations based
on Exhibit 3 indicate:
Percentage of primary emissions
below two micron
Natural and miscellaneous 24
Transportation 4
Stationary combustion 13 ) 257
Industrial 12 (
Total below two micron:
This tabulation indicates that 25 percent of the ambient fine PM in
TSP could be derived from anthropogenic stationary sources if there
were no contribution from secondary PM pollutants. Any presence of
secondary pollutants would decrease the percentage of primary fine
PM emissions from anthropogenic stationary sources. Conversely,
at greater distances from sources, fall out of the large particles
as discussed above, could increase the percentage of primary
anthropogenic fine PM.
Extrapolation of Friedlander results. My extension of
Friedlander results to a nationwide characterization is certainly
open to criticism. It might now be useful to compare some of the
existing non-California studies with the nationwide characterization
described in this report. Of course, most needed are comprehensive
experimental characterizations of several types of non-California
cities. Also, more thought as well as experiments are warranted
with regard to how best to expand the Friedlander approach to a
total source-species-size characterization. In fact, Prof. Friedlander,
himself, might choose to extrapolate his results in a manner such as to
arrive at conclusions quite different from those presented herein.
-222-
-------
Definition of particle-size distribution. There seems to
be sufficient experimental data to justify the conclusion that ambient
particle size distributions tend to be bimodal on a mass or volume
basis. There is less agreement as to where the minimum occurs. Two
mocrons was selected for use in this report and represents a sort of
maximum size for this minimum point in the distribution.
There seems to be least justification for assuming that the
entire fine PM fraction consists of secondary PM. There must be
primary emissions of fine PM, and in this report, I have tried to extend
Brock's data to define the amounts emitted from different source
categories. Combustion sources appear to be the largest source of
the primary fine PM (probably as uncombined carbon).
While examining the relative significance of primary fine
PM, the question arose as to the hypothetical shape of an ambient
distribution, if there were no transformations, and all ambient PM
were derived from primary emissions. The ambient distribution still
could be bimodal if there were sufficient fine (or sub-fine) PM
generated. (See Exhibit 2, curve 3). Such primary fine material
would be expected to coagulate in the same manner as secondary
material, until constrained by the termination of Brownian motion
mechanism at the larger particle sizes (one to two micron). Such
reasoning might partially explain the bimodal distribution reported
by Hidy and Mueller for the relatively non-anthropogenic Goldstone
location.
Primary versus secondary. Much of recent literature has
stressed the presence of the secondary constituent of TSP. The
emphasis has led some to assume that the secondary constituent is
usually dominant in ambient air, and the question is confused if
one also assumes that secondary and fine PM are synonymous. This
report identifies a sizable primary-fine constituent which reduces
the fraction of fine PM which can be allocated to secondary con-
stituents. Further, the input reports all present information
describing situations where primary and/or large PM was found to
be dominant.
-223-
-------
The characterization results, as summarized on Exhibit 8,
can also be directed to the question. The percentages of secondary
are 32 and 25 percent for the California composite and nationwide,
respectively. Both are significant fractions of the total, but neither
is dominant. Also note that the total fine fraction is larger
for the nationwide characterization, but the increase is due to the
primary fine fraction rather than to secondary pollutants.
At this point, the reader might review Exhibit 2. The
California composite (curve 1 on Exhibit 2) and nationwide
characterizations (curve 2 on Exhibit 2) seem to have roughly equal
fractions in each size category (as shown on Exhibit 2), but the
composition within the fine mode may be quite different. Further
resolution awaits more rigorous characterization of the fine mode.
Quantification of the combustion carbon fraction would be particularly
enlightening.
Impact of water and natural species. The results, summarized
on Exhibit 8, indicate a sizable contribution (36-51 percent) from
water and so-called "natural" sources. (Recall that the "natural"
category includes fugitive dusts and emissions from forest and
agricultural fires.) Information in the input reports and elsewhere
indicates that these categories often comprise highly significant
fractions in ambient aerosols. Policy makers should be more aware
of their presence, since these categories appear to pose quite a
constraint in meeting ambient air quality standards.
Note how Exhibit 8 would change if these "natural" sources
were removed. For example, in the nationwide characterization, primary
fine PM from stationary sources would increase from 16 to 25 percent.
Definition of TSP. Much of the difficulty in defining 100
percent of the constituents of TSP lies in our inability to accurately
measure the total mass of TSP itself. Presently, we rely almost
exclusively on the high vol, but two of the input consultants mentioned
the problems associates with reactions taking place upon the mat (mass
is increased) and with volatile particulate matter evaporating from
-224-
-------
the map (mass is decreased). Water was included in the presented
characterizations, in part to bring attention to the problem of
quantification of volatile particulate matter.
Rather than TSP (total suspended particulates), it has
been suggested that PM might be better defined as TFP (total
filterable particulates). Considerable research on these basic mass
measurement methods seems desirable. The discrepancy between some
cascade-type samplers (for particle size distribution) and the high
vol (for total particulate mass) is particularly disturbing.
One must question not only the characterizations but the
epidemiological studies derived from the characterizations. (Lodge
makes the point well on his pages 2-3.)
Emission monitoring. Concern with regard to measurement
adequacy should also extend to emission monitoring. Many assumptions
were required herein because of the lack of adequate emission
characterization. Unfortunately, controlversy shrouds even the
definition of what is a particle, not to mention the much needed
emission size distributions. Of particular interest would be better
definition of that PM which is formed almost at the stack exit.
Should such PM be considered primary, secondary, or is a third
category needed?
Relative toxicity. This report emphasizes the relative
masses associated with various sources, sizes and species. It should
be noted that 'it is unlikely that all the species and sizes have the
same toxicity. Yet such is at least tacitly assumed when making mass
comparisons. It seems likely that incorporation of pindex-type
toxicity-based weighting factors (Babcock and Nagda, 1972) might
increase the significance of some of the primary stationary source
PM (trace metals, asbestos, specific-organics), at least over that of
some of the natural and secondary pollutants. The reader should
consider such possibilities, but because of data limitations,
quantitative treatment presently seems unattainable.
-225-
-------
Conclusions
It seems clear that there is no single generalized source-
species-size characterization which is applicable throughout the nation.
The various distributions may all tend toward bimodal, but the sources
and species appear to vary significantly from location to location,
depending on local sources. Adding further complexity, the relation-
ship between source and ambient air quality seems to vary significantly
from season to season and even hour to hour. The incomplete distributions
published to date are probably not representative of their own regions,
much less of the nation as a whole. The uncertain results presented
herein are summarized on Exhibit 8 and might best be interpreted
as ranges, that most of the nation's characterizations would lie some-
where in between the California and nationwide results.
Both characterizations indicate significant contributions
from water and from natural-and-miscellaneous sources. These find-
ings support the contentions of Lodge and define a significant problem
to control officials. Much more effort should be directed toward
these sources (other than water) which are largely uncontrolled at
present.
Transportation remains a significant problem in both
characterizations, but it is not the dominant category, even in
California.
Finally, even when transformations and secondary pollutants
are considered, stationary combustion and industrial sources remain
the major anthropogenic sources of TSP (.26-49 percent of TSP total) .
These same stationary combustion and industrial sources contribute a
not-dominant, but still-significant fraction of TSP in the form of
fine primary PM emissions (4-16 percent of the TSP total). Continued
research directed toward control of these fine-size emissions seem
well justified.-
-226-
-------
Postscript
This report has attempted to integrate some existing
information in an effort to better relate emissions with ambient PM
characterization and levels. By inferences, the report should also
serve to help define the considerable gaps in our present knowledge.
The two characterizations presented on Exhibits 6 and 7
certainly contain a large number of questionable assumptions, but
these results do seem to be logically derived from information
contained in the input reports. Perhaps I succumbed to over use or
over interpretation of the meager data available, but it seems that
pursuit of this temptation was a fundamental purpose of the study--
to draw conclusions based upon available information. Hopefully,
these conclusions will serve to generate discussion and further study,
as well as generation of more complete experimental characterizations.
Two deficient areas are readily apparent. First, much
increased effort should be directed toward size-specific and species-
specific emission inventories. It should be unnecessary to have to
exclusively work backwards from ambient data. Better characterization
of primary emissions would multiply the utility of comprehensive
characterization of ambient air environments. Characterizations
should be conducted on a size basis, not only in California, but in
other regions.
As new information becomes available, the characterizations
presented on Exhibits 6 and 7 may warrant significant revisions. I
hope my pleas will be heeded, that future characterizations will be
reported as percentages of TSP total, even if an unknown category
must be included. Characterizations published as percentage of known,
of secondary, of fine PM, etc., tend to be confusing, if not misleading
The characterization diagram is proposed as a straightforward means for
presenting results on a TSP basis.
-227-
-------
Some readers may already be eager to draw conclusions
contrary to those presented herein. Toward this end, a blank
characterization diagram is enclosed (Exhibit 9). Generation,
comparison, and evaluation of alternative conclusions in itself
seems a worthwhile outcome of this study.
Acknowledgment
This report was prepared at the request of the Control
Systems Laboratory of the U.S. Environmental Protection Agency and
the Radian Corporation. Gratefully acknowledged is their financial
support,'in part through EPA Grant No. R-802111, "Engineering Analysis
Methodologies for Air Resource Management".
-228-
-------
References
Input Reports:
G. M. Hidy and P. K. Mueller, "Control Technology and
Aerosols", Environmental Research and Technology, Inc.,
741 Lakefield Road, Westlake Village, California 91361
(May 1975).
J. P. Lodge, "Particulate Matter in the Atmosphere",
385 Broadway, Boulder, Colorado 80303 (1975).
J. R. Brock, "Review of Suspended Particulate Matter",
Chemical Engineering Department, University of Texas,
Austin, Texas (June 1975).
G. Gartrell and S. K. Friedlander, "Relating Particulate
Pollution to Sources: the 1972 California Aerosol
Characterization Study", Atmospheric Environment, 9_:
279-299 (1975). (This paper, although not prepared
as part of the CSL-Radian study, provides a central input
to this summary report.)
Other references cited in this summary report:
L. R. Babcock and N. L. Nagda, "Indices of Air Quality",
in W. A. Thomas (ed.), Indicators of Environmental Quality
Plenum Press, New York (1972) pp. 183-197.
P. F. Fennelly, "Primary and Secondary Particulates as
Pollutants, a Literature Review," J. Air Pollution Control
Assn., 25 (7): 697-704 (July 1975).
References cited by the input consultants:
On the following lists "*" indicates the reference was
included in the common body of pertinent literature; see
page 6.
-229-
-------
REFERENCES
Cited by Hidy and Mueller.
Akselsson, K. R., J. W. Nelson & J. W. Winchester, 1975: "Proton
Scattering for Analyses of Atmospheric Particulate Matter", Bull.
Am. Phys. Soc. II, 2Q_, p. 155.
Barone, J. B., T. A. Cahill, R. G. Flocehini, D. J. Shedoan, 1975:
"Visibility Reduction: A Characterization of Three Urban Sites in
California", Science, in manuscript, Feb. 12.
Draftz, R. G. & J. Durham, 1975: "Identification & Sources of Denver
Aerosol". Unpubl. report to U.S. Environmental Protection Agency;
also Harrison, P. W. et al, "Identification & Impact of Chicago's
Ambient Suspended Dust^.Tfnpubl. report for U.S. Environ. Protection
Agency.
Durham, J.L., W. E. Wilson, T. C. Ellestod, K. Willeke and K. T.
Whitby, 1975: "Comparison of Volume and Mass Distribution of Denver
Aerosols", Atmos. Environment, in press.
Durham, J. L., R. K. Patterson, J. J. Vanee and W. E. Wilson, 1975:
"The Chemical Composition of the Denver Aerosol", Atmos. Environment,
in press.
Frank, E. & J. P. Lodge, Jr., 1967: "Morphological identification of
airborne particles with the electron microscope". J. Microscopic, 6,
449-456.
Friedlander, S.K., 1973: "Chemical Element Balances & Identification
of Pollution Sources". Environ. Sci. & Technol. 7_, 235-240.
Gatz, D.F., 1975: "Relative contributions of different sources of.
urban aerosols: application of a new estimation method to multiple
sites in Chicago". Atmos. Environ. I, 1-18.
Gartrell, G., Jr., and S. K. Friedlander, 1975, "Relating particulate
pollution to sources: the 1972 Calif. Aerosol Charact. Study",
Atmos Environ. , 9_, 279-300.
Harrison, P.R., & J. W. Winchester, 1971: "Areawide distributions of
lead, copper and cadmium in air pollutants from Chicago and Northwest
Indiana". Atmos. Environ. 5, 863-880.
Heisler, S., e_t al, 1973: "The Relationship of Smog Aerosol Size &
Chemical Element Distributions to Source Characteristics". Atmos.
Environment 7, 633-649.
Hidy, G.M., 1973: "Removal Processes of Gaseous & Particulate
Pollutants" in Chemistry of the Lower Atmosphere, (S.I. Rasool, ed.),
Plenum Press, N.Y. , Chap. 3~.
-230-
-------
REFERENCES
(Continued)
Hidy, G.M. et al, 1974: "Characterization of Aerosols in California
(ACHEX)". Final Report Volumes 1-4; Rockwell Science Center, Report
#SC524.25FR, Thousand Oaks, CA 91360.
Hidy, G.M. & J.R. Brock, 1971: "An Assessment of the Global Sources
of Tropospheric Aerosols" in Proc. 2nd IUAPPA Clean Air Congr.
(H.W. Englund & W.T. Berry, ed.), Academic Press, N.Y., p. 1088.
Hidy, G.M. & C.S. Burton, 1975: "Atmospheric Aerosol Formation by
Chemical Reactions" to be publ. in Int'l J. of Chem. Kinetics.
Hidy, G.M. & S.K. Friedlander, 1971: "The Nature of the Los Angeles
Aerosol". Proc. 2nd IUAPPA Clean Air Congr. (H.M. Englund & W.T.
Berry, ed.), Academic Press, N.Y., p.391.
Hidy, G.M. et al_,1974: "Observations of Aerosols over Southern
California Coastal Waters", J. of Applied Meteorology, Vol. 13,
No. 1, pp. 96-107.
Miller, M.S. et al, 1972: "A Chemical Element Balance for the
Pasadena Aerosolrr~in Aerosols & Atm. Chem. (G.M. Hidy, ed.), Acedemic
Press, N.Y., p. 301.
Trijonis, J., 1974: "A Particulate Implementation Plan for the
Los Angeles Region". TRW Report for EPA.
Whitby, K.T., R.B. Husar & B.Y.H. Liu, 1972: "The Aerosol Size
Distribution of Los Angeles Smog" in Aerosols & Atmos. Chem.
(G.M. Hidy, ed.), Academic Press, N.Y., p. 237.
Willeke, K., K.T. Whitby, W.E. Clark, V.A. Marple, 1974: "Size
Distribution of Denver Aerosols -- A Comparison of Two Sites",
Atmos. Environment, 8_, pp. 609-633.
-231-
-------
REFERENCES
Cited by Lodge
Colorado Air Pollution Control Program, Report to the Public,
34-37 (1972).
Colorado Air Pollution Control Program. Report to the Public,
56-59 (1974).
Dams, R., J. Billiet, C. Block, M. Demuynk, and M. Janssens.
Atmospheric Environment, in press (1975).
Friedlander, S.K. Environ. Science & Technol., 7 235-240 (1973).
Goetz, A. and R.F. Pueschel. J. Air Pollution Control Assoc. 15,
90-95 (1965). " '
Hagen, L.J., and N.P. Woodruff. Atmospheric Environment 1_,
323-332 (1973).
Harrison, P.R., R. Draftz, and W.H. Murphy. Manuscript, source
unknown.
Lodge, J.P. , Jr. , G.S. Bien and H.E. Suess. Int. J. Air Pollution 2_,
309-312 (1960).
Sverdrup, G.M. , K.T. Whitby and W.E. Clark. Atmospheric Environment 9_,
483-494 (1975).
Whitby, K.T., W.E. Clark, V.A. Marple, G.M. Sverdrup, G.J. Sem, K.
Willeke, B.Y.H. Liu, and D.Y.H. Pui. Atmospheric Environment 9,
463-482 (1975).
Willeke, K., K.T. Whitby, W.E. Clark and V.A. Marple. Atmospheric
Environment 8, 609-633 (1974).
-232-
-------
REFERENCES
Cited by Brock (1 of 4)
1. Environmental Protection Agency: National Primary and Secondary
Ambient Air Quality Standards, Federal Register, 36: 8186 (1971)
2. Anderson, D.O., "The effects of air contamination on health"
Canad. Med. Assoc. J. 9_7, 528, 585, 802 (1967).
3. Amdur, M.O. "lexicological Appraisal of Particulate Matter,
Oxides ns Sulfur and Sulfuric Acid". Paper 69-68, Proceedings
Air Pollution Control Association, New York, New Yrok, June
22-26, 1969.
4. Task Group on Lung Dynamics, Deposition, and Retention Models
for Internal Dosimetry of the Human Respiratory Tract, Health
Physics 12 173 (1966).
5. Winkelstein, W. "The relationship of air pollution and
economic status to total mortality and selected respiratory
system mortality in man", Arch. Environ. Health 14_ 162 (1967).
6. Douglas, J.W.B. and Booras, S.G., "Air pollution and respiratory
infection in Children". Brit. J. Prevt. Social Med. 20, 1
(1966). ~~
7. Lunn, J.E. Knowelden, J. and Handyside, A. J. "Patterns of
respiratory illness in Sheffield infant school children",
Brit. J. Prev. Soc. Med. 21 (1967).
8. Petrilli, R.L., Agrese, G. and Kanitz, S., "Epidemiology
studies of air pollution effects in Genoa, Italy" Arch. Environ.
Health 12 (1966).
9. Carnow, B.W., Lepper, M.H. Shebelle, R.B. and Stamler, J.
"The Chicago Air Pollution Study: S02 levels and acute illness
in patients with chronic broncho pulmonary disease". ARch.
Environ. Health 18 768 (1969).
10. Brasser, L.G., Joosting, P.E., and Von Zuelen, D. "Sulfur oxide
to what level is it acceptable?" Report G-300, Research
Institute for Public Health Engineering, Delft, Netherlands,
July, 1967.
11. Lawther, P.J., "Climate, air pollution and chronic bronchitis,
Proc. Roy. Soc. Med. 51 262 (.1958).
12. Lave, L.B. and Seskin "Air pollution and human health", Science
169 723 (1970).
13. Environmental Protection Agency "Health Consequences of Sulfur
Oxides: A Report from CHESS, 1970-1971". Report EPA-650/1-74-
004, May 1974.
-233-
-------
REFERENCES
(Continued)
(2 of 4)
14. Corn, M. "Measurement of Air Pollution Dosage to Human
Receptors in the Community". Environ. Res. 3_ 218 (1970).
15. Timbrell, V. "Inhalation and Biological Effects of Asbestos"
in T.T. Mercer Stal. "Assessment of Airborne Particles", p.
427, C.C. Thomas, Springfield, M. 1972.
16. Corn, M. "Urban aerosols: Problems Associated with
Evaluation of Inhalation Risk" in T. T. Mercer, et. al.
"Assessment of Airborne Particles, p. 465, C.C. Thomas,
Springfield, 111., 1972.
17. Corn, M. Montgomery, T.L. and Reitz, R. "Atmospheric
particulates: Specific surfaces and densities" Science
159 1350 (1968).
18. Air Quality Criteria for Particulate Matter, U.S. Dept.
H.E.W. Publ. AP-49, 1969.
19. Green, H.L. and Lane, W.R. "Particulate Clouds:' Dusts, Smokes
and Mists" Secon Edition, E. & F.N. Spon. Ltd., London,
1964.
20. Hidy, G.M. and Friedlander, S.K., "The Nature of Los Angeles
Aerosol" in H.M. Englund and W.T. Beery (ed.) "Proceedings
of the Second International Clean Air Congress", Academic
Press, New York, 1971.
21. Ensor, D.S., Charlson, R.J., Ahlquist, N.C., Whitby, K.T.
Husar, R.B. and Liu, B.Y.H., "Multiwavelength nephelometer
measurements in Los Angeles Smog Aerosol", in G. M. Hidy
(ed) "Aerosols and Atmospheric Chemistry", Academic Press,
N.Y., 1972.
22. Ridker, R.G. "Economic Costs of Air Pollution", New York.
Prager, 1967.
23. Barrett, L.B., and Waddell, T.E. "Cost of Air Pollution Damage",
EPA Report AP-85, February 1973.
24. Hidy, G.M. and Brock, J.R., "Proceedings of 2nd Air Congress",
IUAPPA, Washington, D.C., Dec. 1970.
25. "Compilation of Air Pollutant Emission Factors", Second Edition
E.P.A. Report AP-42, April 1973.
26. Vandegrift, A.E., e_t al, "Particulate Air Pollution in the U.S.",
J. Air Pollution Control Association, 2_1 321 (1971).
27. Sehmel, G.A., "Particle resuspension from an asphalt road
caused by car and truck traffic", Atmos. Environ. 1 291 (1973).
-234-
-------
<3 of 4) REFERENCES
(Continued)
29. Miller, et. al, "A Chemical Element Balance for the Pasadena
Aerosol", J. Colloid Interface Sci. 39_ 165 (1972).
30. R. Drake in "Topics in Current Aerosol Research", Pergamon,
Oxford, 1972.
31. M. Lee, R. et. al, Atmos. Environ. 5 275 (1971).
32. Pich, J., et. al, Aerosol Sci. 1 115 (1970).
33. G. Hidy and J.R. Brock, "The Dynamics of Aerocolloidal Systems",
Pergamon, Oxford, 1970.
34. Brock, J. R., J. Colloid Interface Sci., 39 32 (1972).
35. Kolmogorov, A., Akad. Nank SSSR, 3_1 (1941).
36. "Particulate Pollutant System Study", MRI Contract No.
CPA 2269104, EPA, 1971.
37. Schulz, E.J., et. al., "Submicron particles from a pulverized
coal fired boiler", Atmos. Environ. £ 111 (1975).
38. Harrington, W. "Fine Particles", J. Air Pollution Control
Association, 1974.
39. Natusch, D.F.S., et. al., Science 183, 202 (1974).
41. Lee, R.E. and Von Lehmden, D.J., J. Air Pollution Control Assoc.
23_ 853 (1973).
42. Toca, F.M., Thesis, University of Iowa, 1972.
43. Ruud, C.O. and Williams, R.E. "X-ray and microscopie
characterizations of Denver (1973) Aerosols", preprint,
Report Denver Research Institute, 1974.
44. Draftz, R.G. and Blakeslee, H.W., "Identification of ambient
suspended particles from Philadelphia", preprint., I.I.T.R.I..
Report, 1974.
45. Draftz, R.G., "Analysis of Philadelphia suspended dusts sampled
at street level, " I.I.T.R.I. Report No. C9915-1, 1974.
46. Harrison, P. Draftz, R. and Murphy, W.H., "Identification and
impact of Chicago's ambient suspended dust", pre-rpint, I.I.T.R.I
1974.
47. Draftz, R.G. and Durham, J., "Identification and Sources of
Denver Aerosol", preprint, paper #74-263, Air Pollution
Control Association Meeting, Denver, 1974.
48. Whitby, K.T., "Modelling of Atmospheric Aerosol Particle Size
Distributions", E.P.A. Progress Report, R800971.
-235-
-------
REFERENCES
(continued)
(4 of 4)
49. Brock, J.R. and G. M. Hidy, ed., "Aerosols and Atmospheric
Chemistry", Academic Press, New York, 1972.
50. Cox, R.A., "Particle formation from homogeneous reactions of
sulphur dioxide and nitrogen dioxide", Tellus XXVI, 235
(1974).
51. Van Luik, F.W. and Rippere, F.E. Annl. Chem., 34 1617 (1962).
52. Miller, D.F., et. al., "Haze Formation Its Nature and Origin",
Final Report to C.R.C. and E.P.A., March, 1975.
53. Durham, J., Brock, J.R. Judeikis, H., and Lunsford, J.,
"Review of Sulfate Aerosols", EPA Report, In Preparation.
54. "Proceedings of the 7th International Conference on Condensation
and Ice Nuclei", K. Spurney, ed., Academica, Progue, 1969.
55. Brock, J.R. and Marlow, W.A., "Charged Aerosols and Air
Pollution", Environ. Letters, To Appear, 1975.
56. Gartrell, G. and Friedlander, S.K., Atmos. Environ. 9, 279
(1975).
57. Middleton, P. and Brock, J.R. "Atmospheric Aerosol Dynamics:
the Denver Brown Cloud", E.P.A. Report, to Appear.
58. Tuesday, C.S., ed., "Chemical Reactions in Urban Atmospheres",
New York, Elsevier, 1971.
59. Altshuler, A.P. and Bufalini, J.J. Photochem. Photobiology,
4 97 (1965).
60. Air Quality Criteria for Photochemical Oxidants, N.A.P.C.A.
Publication No. AP-63, March 1970.
61. Alley, F.C. and Ripperton, L.A., "The effect of temperature on
photochemical oxidant production in a bench scale reaction
system", J. Air Poll. Cont. Assoc., 11, 581 (1961).
62. Brock, J. R., Faraday Symposia of the Chemical Society No. 7,
"Fogs and Smokes", The Chemical Society, London, 1973.
63. Lundgren, D.A., "Atmospheric aerosol composition and concen-
tration as a function of particle size and time", J. Air
Pollution Control Assoc. 20 603 (1970).
64. Esmen, N. and Corn, M. "Resident time of particles in the
atmosphere", Atmos. Environ. 5 571 (1971).
65. Hidy, G.M. and Brock, J. R., "An Assessment of the global
sources of tropospheric aerosols" Proc. of 2nd Clean Air
Congress, IUAPPA, Washington, D.C., December, 1970.
-236-
-------
REFERENCES IN COMMON BODY OF PERTINENT LITERATURE:
ACHEX, "Freeway Aerosol", (September 20, 1972).
D. L. Blumenthal, J. A. Anderson and G. J. Sem, "Characterization
of Denver's Urban Plume Using an Instrumented Aircraft", Paper
74-266, Air Pollution Control Assn., Denver (June 1974).
C. Brosset and A. Akerstrom, "Long Distance Transport of Air
Pollutants—Measurements of Black Air-Borne Particulate Matter
(Soot) and Particle-Borne Sulphur in Sweden During the Period of
September-December 1969", 6:661-673 (1972).
R. J. Charlson and A. P. Waggoner, "Visibility, Aerosol, and Colored
Haze", Paper 74-261, Air Pollution Control Assn, Denver (June 1974).
M. T. Dana and others, "Natural Precipitation Washout of Sulfur
Compounds from Plumes", (EPA-R3-73-047), prepared by Battelle
Memorial Institute, Richland, Washington (June 1973).
M. T. Dana and others, "Precipitation Scavenging of Inorganic
Pollutants from Metropolitan Sources", (EPA-650/3-74-005), prepared
by Battelle Memorial Institute, Richland, Washington (June 1974).
R. G. Draftz, "Analysis of Philadelphia Suspended Dusts Sampled
at Street Level", IITRI-C9914 (date unknown).
R. G. Draftz, "Analysis of 25 Ambient Dust Samples from Philadelphia",
IITRI-C9915-1 (July 20, 1973).
R. G. Draftz and J. Durham, "Identification and Sources of Denver
Aerosol", Paper 74-263, Air Pollution Control Assn., Denver (June
1974).
J. L. Durham and others, "Comparison of Volume and Mass Distributions
for Denver Aerosols", American Chemical Society presentation,
Los Angeles (April 1974).
S. K. Friedlander, "Chemical Element Balances and Identification of
Air Pollution Sources", Environmental Science and Technology, 7_:3
235-240 (March 1973). .
S. K. Friedlander "Small Particles in Air Pose a Big Control Problem",
Environmental Science and Technology, 7_:13, 1115-1118 (December 1973).
D. F. Gatz "St. Louis Air Pollution: Estimates of Aerosol Source Co-
efficients and Elemental Emission Rates", published by American
Meteorological Society (1974).
D. Grosjean and S. K. Friedlander, "Gas-Particle Distribution Factors
for Organic Pollutants in the Los Angeles Atmosphere", Paper 74-154,
Air Pollution Control Assn., Denver (June 1974).
-237-
-------
P. R. Harrison, R. Draftz, and W. H. Murphy, "Identification and
Impact of Chicago's Ambient Suspended Dust", (source and date
unknown).
J. M. Hales, J. M. Thorp, and M.A. Wolf, "Field Investigation of
Sulfur Dioxide Washout from the Plume of a Large coal-Fired Power
Plant by Natural Precipitation", Prepared for EPA by Battelle Memorial
Institute, Richland, Washington (March 1971).
S. L. Heisler, S.K. Friedlander, and R.B. Husar, "The Relationship
of Smog Aerosol Size and Chemical Element Distributions to Source
Characteristics", Atmospheric Environment 7:633-649 (1973).
G. M. Hidy and S.K. Friedlander, "The Nature of the Los Angeles
Aerosol", in H. M. Englund and W. T. Beery (ed) Proceedings of
the Second International Clean Air Congress, Academic Press, New
York (1971)
P. W. Jones, "Analysis of Non-Particulate Organic Compounds in Ambient
Atmospheres", Paper 74-265, Air Pollution Control Assn., Denver
(June 1974).
R.E. Lee, "The Size of Suspended Particulate Matter in Air", Science
178:4061 567-575 (November 1972).
D. F. Miller and others, "Haze Formation: Its Nature and Origin",
(EPA-650/3-74-002), prepared by Battelle Memorial Institute,
Columbus (June 1973).
M.S. Miller, S.K. Friedlander, and G.M. Hidy, "A Chemical Element
Balance for the Pasadena Aerosol:, J. Colloid and Interface Science,
39:1 165-176 (April 1972).
P. K. Mueller, R. W. Mosley, and L.B. Pierce, "Chemical Composition
of Pasadena Aerosol by Particle Size and Time of Day:IV. Carbonate
and Noncarbonate Carbon Content:, "J. Colloid and Interface Science
39:1, 235-239 (April 1972).
T. Novakov, and others, "Chemical Composition of Pasadena Aerosol by
Particle Size and Time of Day: LLL. Chemical States of Nitrogen and
Sulfur by Photoelectron Spectroscopy", J. Colloid and Interface Science
39:1 225-234 (April 1972).
J. W. Roberts, A.T. Rossano, H.A. Watters, "Dirty Roads Equal Dirty
Air", APWA Reporter, 10-12 (November 1973).
H. Rodhe, C. Persson, and 0. Akesson, "An Investigation into Regional
Transport of Soot and Sulfate Aerosols", Atmospheric Environment,
6:675-693 (1972).
D. Schuetzle, A.L. Crittenden, and R. J. 'Charlson, "Application of
Computer Controlled High REsolution Mass Spectrometry to Analysis
of Air Pollutants", J. Air Pollution Control Assn., 23:8 704-709
-238-
-------
W. Schwartz, "Characterization of Model Aerosols", (EPA-650/3-74-
001), prepared by Battelle Memorial Institute, Columbus (August
1974).
G. A. Sehmel, "Particulate Resuspension from an Asphalt Road Caused
by Car and Truck Traffic, "Atmospheric Environment ]_-.291-309 (1973),
J. W. Winchester and G.D. Nifong, "Water Pollution in Lake Michigan
by Trace Elements from Pollution Aerosol Fallout", Water, Air,
and Soil Pollution, 1:50-64 (1971).
-239-
-------
Exhibit 3. Primary Emission Distribution
Ind power
Crushed stone
Agric opn
Iron & Steel
Cement
Pulpmills
Time
Non-ferrous metals
.Clay
Fertilizers
Phosphate rock
Asphalt
Ferro alloys
Iron foundaries
Coal cleaning
Petroleum
Acids
Solid waste disposal
Construction dust
Transportatio
Transportatio
Transportatio
Miscellaneous
tural fires
Emissions, 10 ton/year
utility
:als
posal
ist
comb us tion
tire dust
road dust
:orest & Agricul-
Mass Median
Diameter.
' «a
Micron
2.0
2.0
10.0
10.0
0.4
3.0
1.0
2.0
1.0
3.0
1.0 '
2.0
1.0
0.4
0.4
2.0
0.5
0.4
0. 3
BIO
0.4
3.0
5.0
1.0
Below
Two Micron,
Percent
50
50
6
6
94
36
76
50
76
36
76
50
76
94
94
50
92
94
97
18
94
36
18
76 .
Total3
3.36
3.22
5.71
1.84
1.31
0.96
0.56
0.65
0.73
0.54
0.30
0.06
0.22
0.16
0.15
0.11
0.05
0.02
1.40
0.80
1.20
0.30
1.00
*
10.55
Below Two
Micron
1.68
1.61
0.34
0.11
1.23
0.35
0.43
0.32
0.55
0.19
0.23
0.03
0.17
0.15
0.14
0.06
0.05
0.02
1.36
0.14
1.13
0.11
0.30
8.02
TOTAL
35.20
18.72
from Brock, Table 9
-240-
-------
Exhibit 4. Preliminary Characterizations
Exhibit 4a California composite (before distribution of secondary species)
Category
Sea salt
Soil dust
Auto exhaust
Cement dust
Fly ash
Oiesil exhausts
Tire dust
Indus t. and Agric.
Aircraft
SO.
NO*
NH3
4
Organic s
Water
Mass/
percent
3.0
19.5
3.7
1.6
0.1
1.0
0.4
12.3
2.2
6.0
9.7
4.5
23.7
12.3
Total mass 100.0
Exhibit 4b Aggregated California composite and nationwide inventory
Category
Natural and Miscellaneous
Transportation
Industry0
Stationary combustion
Water
California
Composite3
35
19
16
18
12
Nationwide
Inventory*3
14
23
22
29
12
100 100
a - derived from Exhibit 4 <*-
b - from Babcock and Nagda: summation of all pollutants except carbon monoxide;
normalized to include 12 percent water
c - includes solid waste
-241-
-------
Exhibit 5. Detailed characterizations
5a: California composite
Natural & miscellaneous
Transportation
Industry
Stationary combustion
Water
TOTAL
to 5b: Nationwide (unnormalized:
j^ California composite= 100%)
i
Natural & miscellaneous
Transportation
Industry
Stationary combustion
Water
TOTAL
5c: Nationwide (normalized)
Natural & miscellaneous
Transportation
Industry
Stationary combustion
Water
Above two micron
Primary
Inorganic incl
SO. & Carbon
26
1
5
0
32
26
1
(15)
( 9)
51
16
1
9
6
Organic
5
5
2
0
12
5
5
(6)
0
16
3
3
4
Below two micron
Other inorganic
& Carbon
2
6
4
12
2
6
(16)
( 9)
33
' 1
4
10
6
Organic
5
5
2
0
12
5
5
2
0
12
3
3
1
0
Secondary
NH4 N03 S04
1
1 5
1
1 56
4 10 6
1
1 5
1
1 5 (14)
4 10 14
1
1 3
1
1 38
Water Total
39
23
14
12
12 12
12 100 .
39
23
40
38
(19) 19
19 159
24
15
25
24
12 12
TOTAL
32
10
21
12
100
-------
organic
inorganic incl 804 + carbon
large particles
water
sources
CO
I
stationary
coirbustion
industry
transportation
natural and miscellaneous
water
SO4
N03
NH>
organic
secondary
other
inorganic
incl carbon
fine particles
Exhibit 6. California composite characterization diagram
-------
< organic >< inorganic incl 804 + carbon
I I
I I
I
I I
J i
large particles
water
sources
stationary combustion
industry
transportation
natural and miscellaneous
water
<-S04->
NO3
NH4
1 I
I I
organic
secondary
<- other inorganic
incl carbon
fine particles
Exhibit 7. Nationwide characterization diagram
-------
Exhibit 8. Summary Comparison
Category
California Nationwide
Combustion and industry
Transportation
Natural and miscellaneous
Water
26
23
39
12
49
15
24
12
TOTAL
100
100
Secondary PM
Primary fine PM (from
stationary sources)
Other Primary fine PM
Large PM
Water
32
4
8
44
12
25
16
5
42
12
TOTAL
100
100
-245-
-------
i
ho
cr>
i
Source
Emissions
Anthropogenic
and
Natural
Control
Equipment
(if any)
METEOROLOGY
Catalysts
including
Moisture
Solar
Radiation
I
4
transformations
Air quality experienced by
receptors
Decay or Deposition
-------
C
o
Primary
and
Reentrained
0.1 i
Particle size, micron
Exhibit 2. IDEALIZED PARTICLE-SIZE DISTRIBUTIONS
10
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-092
2.
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
Total Suspended Particulates: Review and Analysis
5. REPORT DATE
April 1976
6. PERFORMING ORGANIZATION CODE
7. AUTHOH(S)
8. PERFORMING ORGANIZATION REPORT NO.
R. Murray Wells
200-045-27
9. PERFORMING OROANIZATION NAME AND ADDRESS
Radian Corporation
8500 Shoal Creek Boulevard
Austin, Texas 78766
10. PROGRAM ELEMENT NO.
1AB012; ROAP 21ADK-002
11. CONTRACT/GRANT NO.
68-02-1319, Task 27
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park. NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final: 3-8/75
14. SPONSORING AGENCY CODE
EPA-ORD
15.SUPPLEMENTARY NOTES Task officer for this report is J. A. McSorley, Mail Drop 63,
Ext 2745.
is. ASSTRAC,
report gives results of a review and analysis of the readily available
information on total suspended particulates in the atmosphere. The purpose of the
review was to determine the relative contribution of primary and secondary particu-
late matter to the total aerosol mass suspended in the atmosphere and to identify
where, the available information is insufficient to determine the needs for future
control technology development. The report does not identify the fraction of total
.suspended particulates attributable to mobile and to stationary sources.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Dust
Aerosols
Air Pollution Control
Stationary Sources
Primary Particulates
Secondary Particulates
Particulates
13B
11G
07D
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
255
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
"PA Form 2220-1 (9-73)
-248-
------- |