excerpted from
a draft research strategy
to apportion biogenic/
anthropogenic sources of
Secondary Organic Aerosols
March 2004
Presentation
Summaries and
Research
Recommendations
from the
Secondary Organic Aerosols Workshop
February 5-7, 2002
Reno, Nevada
Desert Research Institute
Compiled for
the APACE Workshop Series
Tim Richard, MA, Sr. Associate
Office of Community Services
Fort Lewis College
Durango, Colorado

-------
Findings & Research Recommendations
OF
The Secondary Organic Aerosols Workshop
February 4-5, 2002
Reno, Nevada
Hosted by the Desert Research Institute
Sponsors: US Environmental Protection Agency and National Park Service
Contents


EXECUTIVE SUMMARY			3
TOPIC SUMMARIES AND FINDINGS	..........	....................	8
•	Where and when does the IMPROVE and other chemically-spedated particle data bases show
high OC/EC ratios that might indicate large contributions of secondary oiganic aerosol to light
extinction?			8
What organic particles should be included in the definition of secondary organic aerosol?
Condensation of hot exhaust? Condensation of vapors on particles in the atmosphere? Equilib-
rium changes for volatile particles? Gas-phase transformations? Aqueous-phase transforma-
tions? Organic particle reactions with inorganic gases? 			10
What are the chemical mechanisms that create secondary organic aerosols, what are their
precursors, what are the environmental conditions needed to create and sustain particles, and
what are the organic substances in the particles?							11
•	Which gas and particle end-products can best distinguish secondary oiganic aerosol from
primary oiganic particles at receptor locations? How stable are these components and how
consistent are their ratios to other components in the secondary organic aerosol "source
profile?"					......							„			16
•	What are the size, composition, and hygroscopic properties of secondary oiganic particles that
are most likely to affect light extinction? Which end-products and formation mechanisms are
likely to cause the largest and smallest effects cm regional haze? .,.....,	.			19
•	What are the precursor compounds for secondary oiganic aerosols? What are the types of
vegetation, vehicle exhaust, and burning that emit these precursors and under what conditions?
											21
•	What current sampling and measurement technologies are available to measure marker compo-
nents? How can they be practically applied at urban and remote locations?					28
RESEARCH RECOMMENDATIONS.....		........	32
Near-Term Recommendations							...		32
Middle-Term Recommendations											35
Long-Tom Recommendations 				„	...»							40

-------
2 SOA Workshop Presentation Summaries & Research Recommendations
For more details about information contained in
this reference guide, visit the APACE website at:
http://ocs.fortlewis.edu/aerosols/default.htm
There you will find full topic presentations, as
well as the summaries included in this booklet.
Also a comprehensive bibliography. You may also
link to two other workshops in the APACE series
of symposia on Organic and Elemental Carbon and
Organic Speciation.

-------
SOA Workshop Presentation Summaries & Research Recommendations 3
Executive Summary
The Secondary Organic Aerosols Workshop, held February 4-5, 2002 in
Reno, Nevada, was unique because it was premised on what we don't know
rather that what we do know about SOA. The workshop bears implications
for future dialogue on research and policy development related to not only
air quality in the United States, but significant to climate change, public
health, and visibility and haze.
A major goal of the first Secondary Organic Aerosols Workshop was to pro-
vide opportunities for key researchers across the United States to contribute
to developing a Research Strategy for the three to five years following the
workshop that identifies where study is needed to advance knowledge associ-
ated with secondary organics aerosols. Bringing these key researchers together
to discuss their work and share their recommendations about where research
should be focused was a primary desired outcome of the workshop. Both the
workshop and any strategies that emerge will contribute to enhancing inter-
action among researchers and federal agencies.
Rationale and need for studying SOAs
Secondary Organic Aerosol is important because:
1.	It may contribute to large fraction of organic carbon, which in turn constitutes
a large fraction of PM2 that causes haze and adverse health effects;
2.	It contains many water soluble compounds that may enhance light scattering at
high humidities;
3.	It is an end-product of the photo chemical process that also creates sulfate,
nitrate, and ozone;
4.	It is the least well-qualified and understood of the processes that form particles
in the atmosphere;
5.	Some of it is natural and some is manmade; and
6.	We may come to a point where it is a limiting unknown in achieving PM2 5
standards and visibility goals.

-------
SOA Workshop Presentation Summaries & Research Recommendations
The need for a rationale for studying secondary organic aerosols goes be-
yond the fact that very little is known about secondary organic aerosols and
their precursors. A practical research plan and an underlying rationale is
needed to justify studying secondary organic aerosols and to identify the
benefits of characterizing ambient concentrations of secondary organic aero-
sols and their precursors. For example, what would be done with the infor-
mation if it was discovered that secondary organic aerosols constitute five
percent of the fine aerosol mass and 10 percent of the visibility extinction
on an annual basis?
The importance of secondary organic aerosols related to haze, visibility, cli-
mate, and health has been recognized since the 1950 s. A small community of
researchers in the United States, as well as Europe, have persisted in sam-
pling, measuring, improving the use of available technologies, analyzing and
interpreting data. The problem is complex because there are hundreds, possi-
bly thousands, of SOA precursors and end-products. Many of these remain
to be identified and measured. Sampling, measuring, and defining SOAs is
also recognized as being extremely difficult and challenging.
The Secondary Organic Aerosols Workshop articulated the current context
of SOA research and where knowledge gaps exist. Fifty-four workshop par-
ticipants brainstormed and began to strategize research needs and identify
funding sources for activating research of SOAs.
Lead workshop presenters and guest presenters made general recommenda-
tions about the directions in which research could and should go. Recom-
mendations were drawn from the workshop's rather comprehensive articula-
tion of the body of knowledge to enable us to begin categorizing areas for
potential research.
Actual projects were not outlined. Some funding sources were listed.

-------
SQA Workshop Presentation Summaries & Rfl.qaarch Recommendations
Introduction
This booklet is a condensed reference of findings and recommendations of
the proceedings of the Secondary Organic Aerosols Workshop, held Feb-
ruary 4-5, 2002 in Reno, Nevada at the Desert Research Institute.
The workshop, of about 60 participants, culminated years of efforts among
several representatives of the EPA, NOAA, CIRA, NPS, Desert Research
Institute, and Fort Lewis College in Durango, Colorado to initiate a dialogue
among atmospheric chemistry scientists about ways to advance their fields of
research, and improve communication among researchers and sponsors of
air pollution research.
Background
As early as 1999, an ad hoc committee had formed to put on a "specialty
workshop" that could outline a "road map" that would identify research needs
for distinguishing anthropogenic and biogenic sources of organic aerosols.
Many lively phone conference calls took place, during which several individu-
als debated the best topic for the workshop. It was a tough decision, because
so many pressing issues existed. One person noted that it was strange that a
group of experts coming together to discern the most important research
needs could not agree on the best topic for workshop.
The group had only $30,000 of EPA funds to hold the workshop. In the
end, and for the amount of money available the workshop effectively brought
together key researchers to focus on what is needed in organic aerosols re-
search; as facilitator John Watson said: "to talk about what is needed in aero-
sols research, what is not known and how to close the gap."
The Outcomes
Many participants came away from the workshop pleased that this initial dis-
cussion specific to secondary organic aerosols took place and that they were
able to list general and specific short-, mid-, and long-term research needs.
Some expressed that while a full-fledged research "strategy" would not be
realistic at the time, the potential for increased cooperative dialogue among
researchers and sponsors was strong.

-------
5 SQA Workshop Presentation Summaries & Research Recommendations
The SOA Workshop was a success in that it initiated the development of
two more workshops to identify needs in organic and elemental carbon in
2003, and organic speciation in 2004. A fourth workshop is envisioned for
2005.
In 2003, the International Workshop on the Development of Research Strategies for
the Sampling and Analysis of Organic and Elemental Carbon I Tactions in Atmospheric
Aerosols was held in Durango, Colorado. More than 107 participants from 17
countries and 17 of the Unites States attended. A strategy was begun and
articles were being outlined and written following that meeting.
For 2004, the International State of the Science Workshop on Organic Speciation
in Atmospheric Aerosols Research, was held April 5-7, 2004 in Las Vegas, Nevada
at the Desert Research Institute. This one attracted more than 130 partici-
pants and is expected to produce a "State of the Science Summary Report"
on several topics related to organic speciation.
The Future
In the spirit of perpetuating the concept of supporting cooperative dialogue
begun by the original ad hoc committee, in 2003 the Atmospheric Particulate
Carbon Exchange came into being. Members of this rather informal network
of researchers, agency members, and others primarily lend support towards
continuing the workshop series. Their interest is in building bridges between
sponsors and researchers, and between air pollution researchers in comple-
mentary disciplines.
Power Point presentations, findings and recommendations, and other de-
tails are available on the Internet at the workshop series website:
http://ocs.fortlewis.edu/aerosols/default.htm.

-------
SOA Workshop Presentation Summaries & Research Recommendations 7
Ad hoc committee members at the time of the SOA Workshop:
•	Scott Copeland, USDA-Forest Service, Fort Collins, Colorado
•	Robert Edgar, EPA Region 8, Denver
•	Doug Johnson, EPA Region 8, Denver
•	Doug Latimer, EPA Region 8, Denver
•	Charles Lewis, EPA, NERL, Research Triangle Park
•	Joellen Lewtas, NERL ORD, Seattle
•	Bill Malm, Cooperative Institute for Research in the Atmosphere, Fort Collins,
Colorado
•	Marc Pitchford, NOAA, Las Vegas, Nevada
•	John Reber, National Park Service-!ritermountain Region
•	Tim Richard, Fort Lewis College, Durango, Colorado
•	Mark Scruggs, National Park Service, Denver
•	John Watson, Desert Research Institute, Reno, Nevada
More	info:
•	John Watson, (775) 674-7046
•	Barbara Zielinska, (775) 674-7066
•	Marc Pitchford, (702) 862-5432
•	Joellen Lewtas, (206) 553-1605
•	Douglas Johnson, (303) 312-6834
•	Robert Edgar, (303) 312-6669
•	Tim Richard, (970) 247-7066
This publication was made possible with funding from the US Environmen-
tal Protection Agency. This material is also based upon work supported by
the National Science Foundation/Atmospheric Chemistry Division under
Grant No. 0233861, Any opinions, findings, and conclusions or recom-
mendations expressed in this material are those of the author(s) and do
not necessarily reflect the views of the National Science Foundation.

-------
SOA Workshop Presentation Summaries & Research Recommendations
Topic Findings and Recommendations
Where and when does the IMPROVE and other chemical ly-speci-
ated particle data bases show high OC/EC ratios that might indi-
cate large contributions of secondary organic aerosol to light
extinction?
William C. Malm, Ph. D.
National Park Service-Air Resources Division
CIRA/Colorado State University
Fort Collins, Colorado 80523-1375
Introduction
Samples collected by the IMPROVE network at about 70 sites throughout
the US for roughly the last 14 years have been analyzed by the IMPROVE
thermal optical reflectance protocol. The IMPROVE protocol provides bulk
carbon speciation based on the temperature and atmospheric conditions un-
der which a species volatilizes. The carbon fractions which volatilize between
25°C and 550°C in a helium atmosphere are considered organic, and those
which volatilize between 550°C and 800°C in a 98% helium/2% oxygen at-
mosphere are considered elemental. After an adjustment for charring that is
measured by a reflected laser beam, a "backup," "artifact," or secondary filter
is placed behind some of the primary filters to estimate the effects of gas
sorption on the measurement.
Findings
•	The highest fine organic mass concentrations are measured in the southeastern
US, southern Sierra Nevadas and Pacific Northwest.
•	The ratio of OC to EC at the Washington, D.C. urban site ranged from about 2
to 4. Non-urban eastern US sites had OC/EC ratios which varied between
about 4 and about 10. Sites in the western US generally had OC/EC ratios
between 4 and 10 for low concentration days, but high concentration days
tended to have OC/EC ratios closer to 10. There were anomalous sites, like
Yosemite, with most of the OC/EC ratios near 10. Some published source
profiles for OC/EC ratios include 2.66 for an urban tunnel and 11 for hard-
wood combustion, but these ratios are variable depending on the test.

-------
SOA Workshop Presentation Summaries & Research Recommendations
•	Organic mass accounts for a
larger fraction of fine mass on
high fine mass days in the Pacific
Northwest, compared with
median days. In the eastern US,
organic mass accounts for a
significantly lower fraction of fine
mass on high fine mass days,
compared with median days.
•	OC measurements generally track
wildfire emissions data for
California and the Colorado
Plateau.
•	There is a noticeable discontinuity
in the EC concentrations during
1995 at some sites. The disconti-
nuity is also evident in some of
the "backup" or secondary filter
concentrations.
•	80th percentile OC concentrations have been increasing in the Inter-Mountain
West, and generally decreasing elsewhere.
•	Secondary filter or "artifact" values are similar across at least four sites. The
secondary filter values are also very similar to field blank values. Field blanks
are filters which are sent to the field with the regular samples, but have no air
drawn through them.

-------
SOA Workshop Presentation Summaries & Research Recommendations
What organic particles should be included in the definition of
secondary organic aerosol? Condensation of hot exhaust? Con-
densation of vapors on particles in the atmosphere? Equilibrium
changes for volatile particles? Gas-phase transformations? Aque-
ous-phase transformations? Organic particle reactions with inor-
ganic gases?
James Pankow, Ph. D.
Dept. of Env. Sci. & Eng.
Oregon Graduate Institute, School of Science & Engineering
Portland, OR
Introduction
Definition of Secondary Organic Aerosol — For our purposes, organic aerosols are
solid or liquid particles suspended in the atmosphere containing organic car-
bon. Secondary organics are the fraction of the organics which are not originally
present in the aerosol soon after achieving equilibrium at atmospheric tem-
peratures.
Secondary organic aerosols form through a variety of mechanisms.
•	VOCs (precursors) react with 03 and sunlight to form new compounds.
•	For SOA formation to occur efficiently, high vapor pressure compounds need
to convert to low vapor pressure compounds.
•	The partitioning of these compounds between solid, liquid, and vapor state is
driven by their activity coefficients and vapor pressures.
•	Secondary compounds may be formed in liquid or solid particles.
•	The mass concentration of particle-phase secondary compounds (mg/m1) will
depend on temperature, RH and water content, dilution, and age of the air
parcel.

-------
SOA Workshop Presentation Summaries & Research Recommendations \\
Findings
Partitioning matrix can be solved given activity coefficients and vapor pres-
sures. The solution to this matrix would tell us the partitioning of each com-
pound present.
Conclusions
1.	SOA formation mechanics are coupled for multiple compounds.
2.	Dilution tends to reduce the amount of condensation (non-linearly).
3.	Presence of primary emitted compounds probably increases the condensation
of secondary compounds.
4.	Presence of water increases the condensation of primary and secondary
compounds, though the opposite can also occur.
5.	The thermodynamics of these aerosols is not adequately understood.
What are the chemical mechanisms that create secondary organic
aerosols, what are their precursors, what are the environmental
conditions needed to create and sustain particles, and what are
the organic substances in the particles?
Spyros Pandis, Ph. D.
Chemical Engineering and Engineering and Public Policy
Carnegie Mellon University
Pittsburgh, Pennsylvania
Introduction
The ability of a given volatile organic compound (VOC) to produce SOA
during atmospheric oxidation depends on three factors:
1.	The volatility of its products;
2.	Its emission rate (atmospheric abundance);
3.	Its chemical reactivity.

-------
SOA Workshop Presentation Summaries & Research Recommendations
•	Aromatics are by far the most significant anthropogenic S( )A precursors.
•	Biogenic hydrocarbons emitted by trees are the most important natural source
of SOA.
•	The incremental aerosol reactivity (IAR) is one approach to quantify the ability
of a given precursor to the ambient organic aerosol concentration.
•	Formation—There are two important steps in the formation of SOA. The first
is the chemical reactions in the gas phase leading to the production of the SOA
compounds, These reactions involve the parent VOC, its products, and O the
OH radical, the M03 radical, NO , etc. The second step is the reversible
partitioning of the produced SOA compounds between the gas and particulate
phases.
•	Both the chemical mechanisms leading to the formation of SOA compounds
and their partitioning have been studied in the laboratory for 20 years. Most of
these have focused on the SOA production during the a-pincne reaction with
ozone,
•	Developing detailed chemical mechanisms explaining the production of the
SOA compounds is an ongoing process, but because of the chemical complex-
ity, progress has been slow.
•	The partitioning of the SOA compounds between the gas and particulate
phases is a critical step in the overall SOA formation process. Most of the SOA
compounds have saturation vapor mixing ratios of the order of 1 ppb and are
therefore semi-volatile. Quantifying the fraction of these compounds in the
particulate phase under given conditions is a major challenge. The current
approach suggested first by Pankow and co-workers and refined for SOA by
Odum and co-workers assumes the formation of a solution by the SOA
compounds.
•	There are several compounds that could be used to relate SOA compounds to
precursor gases. Some (pinic acid, pinonic acid, norpinic acid, norpinonic acid,
etc.) result from the oxidation of multiple precursors, but others
(hydroxypinonaldehydes, hydroxypinaketones, sabinic acid, etc.) are unique
products of a given precursor.
•	Several SOA compounds have been identified in laboratory studies but only a
few (mostly biogenics) have been measured in ambient air.

-------
SOA Workshop Presentation Summaries & Research Recommendations J3
•	SOA concentrations are
expected to be sensitive to the
ambient temperature, which
will affect both the rates of
the gas-phase reactions and
the partitioning of the SOA
compounds.
Findings
•	Anthropogenic and biogenic
SOA compounds may interact.
Increases in the production of
one may lead to increased
concentrations of the other by shifting their partitioning towards the particu-
late phase.
•	A 10°C change in the ambient temperature can change SOA concentration in
the a-pinene/ozone system by as much as a factor of two (Kamens and Jaoui
2001). Higher temperatures are expected to decrease the SOA yields in this
system.
•	For the ambient atmosphere the increase in production rates with increasing
temperature would partially offset the evaporation of SOA (Strader and Pandis
1999). Intermediate temperatures (around 20"C in their case) could be optimal
for the SOA production. The change in partidoning of SOA at lower tempera-
tures could lead to counterintuitive behavior of the SOA concentration during
the day (Bowman and collaborators).
•	Relative humidity is a secondary factor to temperature to the formation of
SOA, but its role in SOA production has yet to be elucidated. The lower the
RH, the higher the OC/EC ratio.
•	The presence of aerosol liquid water does not significandy increase or decrease
SOA yields during the photo-oxidation of toluene in the presence of NO^
(Edney et al. 2000). The same lack of sensitivity to relative humidity was
reported for the a-pinene/ozone system (Kamens and Jaoui 2001). The
partitioning of semi-volatile organic compounds (alkanes, alkanoic acids,

-------
SOA Workshop Presentation Summaries & Research Recommendations
PAHs, etc.) on secondary
organic aerosol (formed from
the a-pinene reaction with
ozone) was not sensitive to
relative humidity (Jang and
Kamens 1998).
• We don't see the amounts
of SOA in forests that we
expected to see. We are
missing something: either the
amount of time of the
compounds in the atmosphere
is significant factor and/or
there are so many tracers that
the amount of a single one is
minimal.
Conclusions
•	Our understanding of the ability of individual VOCs, at least at semi-quantita-
tive level, to serve as SOA precursors appears to be satisfactory.
•	Our understanding of the physicochemical processes leading to the SOA
formation has improved dramatically during the last decade. A number of
models of variable complexity are now available.
•	The challenges in the effort to identify SOA compounds in laboratory studies
are significant:
1)	The difficulty in measuring the concentrations of these compounds—They are polar
and their concentrations appear to be only a few nanograms per cubic
meter.
2)	The partitioning of these compounds between the gas and particulate phases—Most
are semi-volatile and their aerosol fingerprints change continuously.
3)	The reactivity of these compounds—Some of the most volatile ones have brief
atmospheric lifetimes. The reactivity of the less volatile ones has not been
investigated in any detail.

-------
SOA Workshop Presentation Summaries & Research Recommendations
4) An additional analytical pitfall—Tobias et ah (2000) discovered that aerosol
produced from the reactions of 1-tetradecene and ozone the hydroperox-
ides, peroxides, and secondary ozonides observed by thermal desorption
particle beam mass spectrometry (TDPBMS) thermally decomposed to
more volatile compounds including tridecanal, tridecanoic acid, and few
unidentified products.
•	Nucleation of SOA could be a potentially important process for the aerosol
number concentration, but is of negligible importance for the SOA mass
concentration.
•	OC/EC ratio gives us a clue about the formation of SOA. When OC/EC rado
goes up, we can be sure that SOAs are formed. The ratio goes up the higher
the temperature. "Photochemical activity drives the whole process."
References
Edney E. O., D. J. Driscotl, R. E. Speer, W. S. Weathers, T. E. Kleindienst, W, Li, and
D. E Smith (2000) Impact of aerosol liquid water on secondary organic
aerosol yields of irradiated toluene/propylene/NOx/(NH4)2S04/air mixtures,
Aimos. Environ., 34, 3907-3919.
Jang M. and R. M. Kamens (1998) A thermodynamic approach for modeling partition-
ing of semivolatile organic compounds on atmospheric particulate matter:
Humidity effects, Environ. Sci. TechnoL, 32, 1237-1243.
Kamens R. M. and M. Jaoui (2001) Modeling aerosol formation from a-pinene and
NOx in the presence of natural sunlight using gas-phase kinetics and gas-
particle partitioning theory, Environ. Sci. TechnoL, 35, 1394-1405.

-------
| SOA Workshop Presentation Summaries & Research Recommendations
Which gas and particle end-products can best distinguish second-
ary organic aerosol from primary organic particles at receptor
locations? How stable are these components and how consistent
are their ratios to other components in the secondary organic
aerosol "source profile?"
James Schauer, Ph.D., PE
Assistant Professor
Civil and Environmental Engineering
Wisconsin State Laboratory of Hygiene
Introduction
Speciation of the organic compounds present in the carbonaceous fraction
of atmospheric particulate matter samples has been shown to provide power-
ful insight into the impact of primary air pollution sources on particulate
matter concentrations in both the urban and remote locations.
Using gas chromatography mass spectrometry (GCMS) techniques, along
with the rich knowledge of organic compound molecular markers, molecular
marker chemical mass balance (CMB) models have been developed and ap-
plied to apportion the source contributions of direct primary sources of at-
mospheric particulate matter.
CMB models have been used to apportion the primary source contribu-
tions from diesel engine, gasoline-powered motor vehicles, hardwood com-
bustion, softwood combustion, meat cooking operations, road dust, tire wear,
vegetative detritus, natural gas combustion and coal combustion. These models
are reasonably well developed and are currendy being employed in a broad
range of air quality studies.
Findings
• Molecular marker CMB models that explicitly apportion the direct primary
source contributions to particulate organic carbon can be used to quantify an
upper limit on secondary organic aerosols.

-------
SOA Workshop Presentation Summaries & Research Recommendations yj
•	Under conditions where there exists high confidence that all important direct
primary sources of particulate matter are incorporated into the molecular
marker CMB models, estimates of secondary organic aerosol contribution to
particulate matter concentrations can be obtained.
•	The direct apportionment of secondary organic aerosols would allow a direct
mass balance check on particulate matter organic carbon concentrations.
•	Three chemical analysis approaches have been used in the past to identify the
organic constituents of secondary organic aerosols (SOA):
1)	chemical analysis of laboratory generated SOA formed in smog
chamber experiments;
2)	analysis of atmospheric particulate matter samples collected in
locations where high levels of SOA are expected;
3)	statistical analysis and chemical structural interpretation of organic
compounds speciation measurements of atmospheric particulate
matter.
•	From a practical perspective, there are three criteria necessary for organic
compounds to be useful as tracers for SOA:
1)	the source of these compounds in the atmosphere must be
dominated by atmospheric chemical reactions and not primary
source emissions;
2)	the compounds must be reasonably stable in the atmosphere after
production;
3)	there must exist a quantitative relationship between SOA produc-
tion from at least a class of SOA precursors and the tracer com-
pound.
•	The aliphatic and aromatic organic diacids are of specific interest in tracking
SOA since these compounds have been identified in atmospheric particulate
matter samples in both the urban atmosphere and remote-location marine
environments.
•	The multifunctional substituted carbonyls have been shown to be major
components of SOA.

-------
Ijj SOA Workshop Presentation Summaries & Research Recommendations
•	Both nitrated mono-aromatic and nitrated polycyclic aromatic hydrocarbons
have been identified as products of atmospheric chemical reactions that are
contributors to SOA.
•	Some of the potential tracers for SOA are semi-volatile organic compounds,
which exist in both the gas and particle-phase at atmospheric conditions, while
some have very low vapor pressures and reside almost entirely in the particle-
phase.
•	The development of source apportionment techniques that only utilize par-
ticle-phase measurements will need to exclusively utilize non-volatile tracers,
such as the aromatic, aliphatic, and cycloalkyl diacids.
•	Source apportionment strategies that take advantage of both non-volatile and
semi-volatile tracers are still needed to face the challenge of accurately measur-
ing or effectively expressing the concentration of SOA in the particle-phase^
Recommendations
•	Two different strategies for source apportionment of SOA are needed: One
that can be applied to routine filter-based samples; and, one that can fully
exploit advanced sampling techniques that are likely to be employed in selected
field studies,
•	Identify molecular markers for SOA and develop source profiles for secondary
organic aerosols that can be incorporated in CMB models.
•	Field studies should be conducted that make adequate measurements of likely
SOA tracers and known primary source tracers that can be used in an advanced
positive matrix factorization (Ramadan et al,, 2001) such that effective profiles
for different sources of SOA can be identified. Field studies should be imple-
mented in both regions that are likely impacted by SOA originating from both
biogenic and anthropogenic sources.
•	Conduct smog chamber studies to identify the atmospheric reactions and
precursors for secondary aromatic diacids.
•	Additional smog chamber experiments should be conducted to further investi-
gate the stability of potential tracers for SOA, with specific attention to diacids,
multifunctional substituted carbonyls, and nitrated aromatics.

-------
SOA Workshop Presentation Summaries & Research Recommendations J9
•	Better integration of analytical techniques among atmospheric measurements,
primary source measurements and laboratory smog chamber experiments to
help assess the uniqueness of proposed SOA tracers.
•	More advanced analytical techniques need to be tested to identify novel SOA
tracers including isotopic fractionation during SOA production, physical
properties, nitrogen containing compounds, and polar organic compounds.
What are the size, composition, and hygroscopic properties of
secondary organic particles that are most likely to affect light
extinction? Which end-products and formation mechanisms are
likely to cause the largest and smallest effects on regional haze?
Lynn Hlldemann
Associate Professor
Civil Environmental Engineering Dept.
Stanford University
Findings
•	Secondary organic compounds condense on the surface of suspended par-
ticles, thereby increasing their size. This growth is usually toward particle sizes
(~0.5 nm) that scatter light more efficiently than the original particles.
•	Many secondary organic compounds absorb water as relative humidity in-
creases. This causes particles to grow toward sizes that scatter light more
efficiently than the original particles. This growth is not as much as that
observed for inorganic compounds such as ammonium sulfate and ammonium
nitrate at RH>70%. This effect has been observed in non-urban areas.
•	Some secondary organic compounds may cause particles containing ammo-
nium sulfate, ammonium nitrate, and other water-absorbing inorganic com-
pounds to grow less than expected as humidity increases. This may be due to
organic surface films that inhibit the interaction of water vapor with the
inorganic salt. This effect has been observed in urban areas.
•	Organic compounds must have a low enough vapor pressure to condense at
ambient temperatures of 10 to 30° C. They must be water-soluble to affect

-------
20 SOA Workshop Presentation Summaries & Research Recommendations
light extinction. Potential compound groups include diacids, polyols, and
amino acids. These are rarely quantified in ambient air and can arise from both
primary emissions and secondary formation.
•	Particles that include secondary organic compounds may be non-spherical and
of inhomogeneous composition. Extinction efficiencies determined by models
of uniform spherical particles may misrepresent actual efficiencies. Descrip-
tions of particle structures and models to approximate their effects on light
extinction are currently unavailable.
•	For the few water-soluble organic compounds that have been examined, a
widely-used chemical model (UNAFAC, not intended for water mixtures, but
the only model available) does not estimate water activities that are consistent
with measurement results. This lack of information frustrates efforts to
mathematically model water uptake.
•	Typical urban mitigation that strategies focus on very small or vary large
particles may increase the effect of condensed secondary organic compounds
because the remaining particles (0.3-0.7 fim) will collect more of the condens-
able secondary organic material, thereby enhancing its light scattering efficiency.
•	Secondary organic compounds will have the largest visibility effect when sulfate
levels are low and relative humidities are 50% to 70%. Sulfate dominates light
extinction at higher humidities and when it dominates PM2 5 mass.

-------
SOA Workshop Presentation Summaries & Research Recommendations 21
What are the precursor compounds for secondary organic aero-
sols? What are the types of vegetation, vehicle exhaust, and
burning that emit these precursors and under what conditions?
Richard Kamens, M. Jang, S. Lee, and M. Jaoui
Department of Environmental Sciences and Engineering
University of North Carolina-Chapel Hill
Introduction
The environmental chamber work of many investigators clearly demonstrates
that aromatics and naturally emitted terpenes have the potential to generate
secondary aerosol material. The primary atmospheric reactions of these com-
pound classes involve the hydroxyl (OH) radical, ozone, and the nitrate radi-
cal (NOj). These reactions produce a host of low volatility dicarbonyls, car-
boxylic acids, hydroxy carbonyl and organic nitrate compounds that can exist
both in the gas and aerosol phase.
As reported by Went[1Leonardo Da Vinci described haze over cities
and thought that water emissions from plants were its source. We know today
that the relative importance of precursors to secondary aerosol formation
will depend on their overall aerosol potential, atmospheric emissions, and the
presence of other initiating reactants (03, OH, NOa, sunlight, acid catalysts).
Over 40 years ago, Went posed the question, "What happens to 17,5x107
tons of terpene-like hydrocarbons or slighdy oxygenated hydrocarbons once
they are in the atmosphere?" Went*11 suggested that terpenes are removed
from the atmosphere by reaction with ozone and demonstrated "blue haze"
formation by adding crushed pine or fir needles to a jar with dilute ozone.
Monoterpenes (C|()H|6) represent about 10% of the natural non-methane hy-
drocarbon (NMHC) emitted by vegetation to the atmosphere121. A-pinene
tends to be the most ubiquitous terpene, and may account for about 20-25%
of the potential secondary aerosol mass from terpenoid type compounds'3 •4
d-limonene may be as high as 20% and b-pinene from 7-15%.
Susquiterpenes (Cl5H24) are also released from vegetation and they may
contribute as much as 9% to the total biogenic emissions from plants 15 K

-------
22 SQA Workshop Presentation Summaries & Research Recommendations
Other estimates are both higher and lower. The sesquiterpenes b-caryophyllene
and a-humulene have very short life times in the presence of representative
global average concentrations of 03 or the nitrate radical (NO^), which have
life times on the order of minutes. These two compounds on a reacted mass
basis have 3-5 times the aerosol potential of either a- or b-pinene. In addition
to monoterpenes and sesquiterpenes, a number of oxygenates are emitted by
vegetation. These include alcohols, carbonyls, acetates and organics acids. Of
significance is that in many instances the oxygenate emissions may be higher,
depending on the plant species, than monoterpene emissions.
Some recent examples of ambient terpene concentrations are given in
Table 2. Yu et al'61 reported terpene concentrations for San Bernardino Na-
tional Forest, California, USA. Sampling was for an evening to the next mid-
day. On average, terpene concentrations ranged from 10 to 63 pptV Hannele
et al171 have reported similar concentrations in an open field near a forested
area in Finland.
Globally, about 25 Tg yr1 of toluene and benzene and are emitted with
fossil fuels contributing ~80% and biomass burning another 20 %'8'. Volatile
aromatic compounds comprise a significant part of the urban hydrocarbon
mixture in the atmosphere, up to 45% in urban US and European loca-
tions'9,10 •" I Toluene, m-and p-xylenes, benzene and 1,2,4-trimethyl benzene,
o-xylene and ethylbenzene make up 60-75% of this load.
In the rural setting, the picture is quite different. At a rural site in Ala-
bama in the summer of 1990, aromatics contributed ~1.7 % to the overall
VOCs[12'. Alkenes were the major category, with isoprene and a-pinene and
b-pinene making up 37, 3.5, and 2% of the VOCs. Alkanes made up 9% and
oxygenates 46%.
Hydrocarbon emissions from two tunnels in the US showed that aro-
matic emissions comprised 40-48% of the total nonmethane hydrocarbon
emissions for light and heavy duty vehicles 1131 On a per mile basis heavy-duty
trucks emit more than twice the aromatic mass, than light-duty vehicles emit,
and the distribution of aromatic is different between these two classes. The
six aromatic compounds mentioned above comprised ~60% of the light-
duty emissions, but only about 27% of the heavy-duty emissions.

-------
SOA Workshop Presentation Summaries & Research Recommendations 23
Definitions
For this discussion secondary organic aerosol (SOA) material is defined as
organic compounds that reside in the aerosol phase as a function of atmo-
spheric reactions that occur in either the gas or particle phases.
Findings
•	Natural and anthropogenic fine aerosol emissions to the atmosphere are on the
order of 200 to 300 Tg yrBiogenic aerosols represent ~ 10% of this figure. A
modeling estimate by Griffin et al. for biogenic aerosols emissions is 13 to 24
Tg y"1; on the same order of magnitude for predictions of anthropogenic soot
and natural or anthropogenic nitrates, but much less than sea salt or natural or
anthropogenic sulfate aerosols. If more realistic, lower average global tempera-
tures were used, other existing aerosol surfaces were considered and possible
particle reactions proposed by Jang and Kamens'14', this emission rate may be
much higher.
•	The chamber work of many investigators clearly demonstrate that terpenoid
and aromatics have potential to generate secondary aerosol mate-
^[15,16,17,18,19,20,,21.22,23^ por aromatic systems with TSP concentration of
100mg/m3, and using the Pankow relationship for absorptive partitioning'24
0.1% of a multi carbonyl-OH product, 0.06% of a buten-al-oic product, and
15% of a dicarbonyl-alcohol-carboxylic acid product would be in the aerosol
phase. A host of new ring-opening products, which include oxo-butenoic,
dioxo-pentenoic, methyl-oxo-hexendienoic, oxo-heptadienoic and
trioxyohexanoic carboxylic acids, as well as similar analog aldehdyes were
recently identified by Jang and Kamens'1"^, along with chemical mechanisms to
explain their formation have recendy been reported. Many of these products
were major components in the particle phase. Very few of these products have
been observed in ambient samples, although the under predicted receptor
modeling of dicarboxylic acid aerosol content by Schauer el al.125' may be a
result of these processes.
•	Of the major SOA products observed in toluene and a-pinene outdoor smog
chamber experiments1,uv•26experimental partitioning coefficients between the
gas and the particle phases of aldehyde products were much higher and

-------
SO A Workshop Presentation Summaries & Research Recommendations
deviated more from predicted 'K. This is an extremely important result, because
it suggests that aldehyde products can further react through heterogeneous
processes and may be a very significant SOA generation mechanism for the
oxidation of aromatics in the atmosphere. As product, aldehydes become
incorporated into larger molecules in the particle phase, more parent aldehdyes
partition from the gas to the particle phase. A recent study reported that inert
particles acidified with sulfuric acid can promote these reactions and form
much higher yields of secondary products than when acid is not present'27'.
This study also shows that dialdehydes such as glyoxal, as well as hexanal and
octanal can directly participate in secondary aerosol formation, but this process
is significantly enhanced by the presence of an acid seed aerosol. The same
phenomena was observed for the reaction of aldehydes and alcohols. The
products of particle phase aldehyde reactions that lead to this SOA increase are
probably thermally unstable and do not usually survive the workup procedure
for traditional analysis techniques.
Conclusions
•	Sesquiterpenes are important to some currently unknown level in SOA forma-
tion.
•	The description and understanding of chemical mechanisms for the production
of SOAs from biogenics and aromatic precursors are critical.
•	Many of the techniques used to detect and quantify particle phase reactions are
too harsh. We are destroying, or at least changing, the compounds that we are
attempting to find.
•	Some uncertainty exists in the area of impacts caused by humidity and drought
stress on biogenic emissions, such that emission models can accurately reflect
their input.
•	Carboxylic acids may be very good candidates to look at as tracer compounds
(agreeing with J. Schauer), while aldehydes may not, because of their reactivity.
Recommendations
•	Determine the importance of particle phase reactions as a source of SOA.

-------
SQA Workshop Presentation Summaries & Research Recommendations
25
•	Determine the importance of sesquiterpenes in SOA formation.
•	Clarify the impact of drought and relative humidity on biogenic emissions so
that these factors can be incorporated into emission models.
•	Define the integrated chemical mechanisms necessary to predict SOA from
biogenics and aromatic precursors
•	Develop new analytical techniques to detect and quantify particle phase reac-
tions. These need to be non-invasive or "chemically soft" so that complex
particle phase reactions products are not decomposed.
Update
As a result of the workshop, given that Dr. Richard Kamens researched the
importance of sesquiterpenes for his workshop presentation, he and colleagues
went on to investigate the role of sesquiterpenes in SOA formation. Since the
workshop, they have identified for the first time (Jaoui et al.) many of the
reactions products for different sesquiterpenes and these compounds may
prove to be an even more important source of SOA than terpenes them-
selves.
References
1 Went, F.W, Organic Matter in the Atmosphere, and its possible Relation to Petroleum
Formation, Proc National Academy of Sciences, Botany, 212-221, 1959.
2Guenther, A.B.; Monson, R.K.; Fall, R.; Isoprene and Monoterpene Emission Variabil-
ity: Observation with Eucalyptus and Emission rate Algorithum Development,
J. Geophys. Res., 96, 10799-10808, 1991
3Griffin, R. J.; Cocker, D. R,, III; Flagan, R. C.; Seinfeld, J. H. Organic Aerosol Forma-
tion from the Oxidation of Biogenic Hudrocarbons. J. Geophys. Res., 104,
3555-3567. 1999
4 Andersson-Skold Y. and Simpson D., Secondary Organic Aerosol Formation in
Northern Europe: a Model Study, J. Geophys. Res., 106, 7357-7374, 2001
6Helmig D., L ; Klinger, L.F.; Greenberg, J.; Zimmermann, P.; Emissions and Identifica-
tion of Individual Organic Compounds From Vegetation in Three Ecosystems

-------
SQA Workshop Presentation Summaries & Research Recommendations
in the U.S., 207 ACS Meeting Am, Chem. Soc., San Diego, California, 13-18,
March 1994.
8 Yu, J.; Griffin, R.J.; Cooker III, D.R.; Flagan, R.C.; Sienfeld, J.H.; Observations of
Gaseous and Particulate Products of MonoTerpene Oxidation in Forested
Atmospheres, J. Geophys. Res. Let. , 26, 1145-1146, 1999
7Hannele, H.; Lauhla, T.; Rinne, L.; Puhto, K.; The Ambinet Concentratino fo Biogenic
Hydrocarbons at a Northern European, Boreal Site, Atmos. Environ, 4971-
4982, 2000.
B Ehhalt, D.H. Gas phase chemistry of the troposphere, pp. 21-109, in R. Zellner (edit.),
Global Aspects of Atmospheric Chemistry, Steinkopff-Verlag, Darmstadt, and
Springer-Verlag, New York, 1999.
"Kurtenbach, R.; K.J. Brockmann, J. ; J. Lorzer; A Niedojadlo; K. H. Becker |VOV-
Measurempits in Urban Air of the City of Wuppertal, in TFS-LT3 annual Report
(German), K.H. Becker editor, 1998.
10 Fujita, E.M; Z. Lu; L. Sheetz; G. Harshfeld; B. Zilinska Determination of Mobile
Source Emissions Fraction Using Ambient Field Measurements, Report to the
Coordinating Research Council (CRC), 215 Perimeter Center Parkway,
Atlanta, GA, 1997.
12Ciccioli, R.; E. Brancaleoni; M. Frattoni The Reactive Hydrocarbons in the Atmo-
sphere at Urban and Regional Scales, in Reactive Hydrocarbons in the
Atmosphere. C.N. Hewitt(editor) Academic Press, San Diego, 1999.
13Goldan, P.D., Kuster, W,C. ; Fehsenfeld, F.C.; Montzka, S.A.; Hydrocarbon Measure-
ments in the Southeast United States: the Rural Oxidants in the Southern
Environment (ROSE) Program 1990, J. Geophys. Res., 100, 259454-
25963,1995.
14Sagebiel, J.C.; Zielinska, B.; Rierson, W.P.;Gertler, A.W.: Real-World Emissions and
Calculated Reactiviteis of Organic Species from Motor Vehicles, Atmos.
Environ, 30, 2287-2296, 1996.
15 Jang, M.; Kamens, R.M.; Characterization of Secondary Aerosol from the Photooxida-
tion of Toluene in the Presence of NOx and 1-Propene, Environ. Sci. Technol,.,
35, 3625-3639, 2001
18 Prager, M. J.; Stephens, E. R.; Scott, W. E.; Aerosol Formation for Gaseous Air
Pollutants, Ind. Eng. Chem. 52,521-524, 1960
17	Izumi, K.; Fukuyama, T.; Photochemical Aerosol Formation From Aromatic
Hydricarbons in the Presence o f NOx, Atmos. Environ. 241,1433-1441, 1990
18	Hatakeyama, S.; Izumi, K.; Fukuyama, T.; Akimoto, H. ; Reactions of ozone with a-
pinene and b-pinene in air: yield of gaseous and particulate products. J.
Geophys. Res. 20, 13013-13024,1989.
19	Yu, J., Flagan R. C.; Seinfeld, J.; Identification of products Containing -COOH, -OH,

-------
SOA Workshop Presentation Summaries & Research Recommendations 27
-C=0 in Atmospheric Oxidation of Hydrocarbons, Environ. Sci. and Technoi,
32, 2357-2370, 1998.
20	Jang, M., and R. M. Kamens, Newly characterized products and composition of
secondary aerosols from reaction of a-pinene with ozone, Atmos Environ., 33,
459-474, 1999.
21	Odum, J. R.; Hoffmann, T.; Bowman, F.; Collins, D.; Flagan, R. C.; Seinfeld, J. H. Gas/
Particle Partitioning and Secondary Organic Aerosol Yields, Environ. Sci.
Technoi., 30, 2580-2585, 1996.
22Glasuis, M., Lahaniati, M,; Caligirou, A.; Di Bella, D.; Jensen, N. R.; Hjorth,.J.; Duane
M. J.; Kotzias, D.; Larsen B. R.; Carboxylic Acids ir Secondary Aerosols from
Oxidation of Cyclic Monoterpenes by Ozone, Environ. Sci. and Techno!. 34,
1001-1010, 2000.
"Hull, L. A. Terpene Ozonalysis Products. In Atmospheric Biogenic Hydrocarbons:
Butalini, J. J.; Arnts, R. R., Eds.; Ann Arbor Science: Ann Arbor, Ml, 1981; Vol.
2, 161-186, 1981
24Griffin, R. J.; Cocker, D. R., Ill; Flagan, R. C.; Seinfeld, J. H. Organic Aerosol Forma-
tion from the Oxidation of Biogenic Hudrocarbons. J. Geophys. ties., 104,
3555-3567. 1999
25Pankow, J. F. An absorption model of gas/particle partitioning of organic compounds
in the atmosphere Atmos. Environ. 28, 185-188, 1994.
2eSchauer, J.J.;Rogge, W.F.; Hildemann. L >M.; Mazurek, M.; Cass, G.R.: Source
Apportionment of Airborn Particulate Matter Using organic Compounds as
Tracers, Atmos. Environ., 30, 3837-3855,1996.
27	Kamens, R. M., and Jaoui, M., Modeling aerosol formation from a-pinene + NOx in the
presence of natural sunlight using gas phase kinetics and gas-particle
partitioning theory, Environ. Sci. Technoi. 35, 1394-1405, 2001.
28	Jang, M., Kamens; R.M, Atmospheric Secondary Aerosol Formation by Heteroge-
neous Reactions of Aldehydes in the Presence of a Sulfuric Acid Aerosol
Catalyst Environ. Sci. Technoi, 35, 4758-4766, 2001

-------
SO A Workshop Presentation Summaries & Research Recommendations
What current sampling and measurement technologies are avail-
able to measure marker components? How can they be practically
applied at urban and remote locations?
Barbara Zielinska, PhD.
Research Professor
Desert Research Institute
Reno, Nevada
Sampling
At present the most commonly used method is filter collection of ambient
aerosol, followed by laboratory analyses. Since organic compounds, includ-
ing secondary organics, are associated with fine particles (i.e. below 2.5 mm
aerodynamic diameter), the use of an appropriate cut-off inlet is necessary.
From the point of view of a sample size, a cyclone, which allows for higher
sampling flow, is recommended. The selection of a filter depends on the
type of analysis that will be run on the sample later. For thermal carbon
analysis, a quartz fiber filter is appropriate, since it withstands temperatures
up to 1000° C. However, due to the large specific surface area, a quartz filter
is prone to positive sampling artifact; i.e. adsorption of organic gases during
sample collection. Teflon membrane filters have much smaller exposed sur-
face area and are thought not to adsorb organic gases, but they are not ther-
mally stable and not easy to use for extraction. Teflon-coated glass fiber
filters (TIGF) are not stable enough in high temperatures to be use for car-
bon analyses, but they are an excellent choice for the collection of samples to
be used for organic solvent extraction. The effectiveness of Teflon coating
in reducing adsorption has not been studied; however, our data (B. Zielinska,
unpublished results) indicate that the adsorption is not significant for these
types of filters.
Impactors can be used to obtain size-segregated samples of organic aero-
sol; however due to the small sample sizes, their application to the detailed
chemical speciation of organic aerosol is still limited. Particles collected in
impactors are usually subjected to smaller pressure drops than filter-collected

-------
SOA Workshop Presentation Summaries & Research Recommendations 29
samples, resulting in lower losses due to volatilization. Particle bouncing may
be a problem, especially at low humidity, since organic analysis excludes the
use of grease.
The denuder strips the gas-phase species from the air stream by diffusion
to an adsorbent surface (e.g. activated carbon, XAD resins, etc.) before col-
lection of the particles on a filter. Since the removal of gas-phase organics
disturbs the gas-particle equilibrium and drives the volatilization of the par-
ticulate material from the filter, an adsorbent bed (such as polyurethane foam,
XAD resins, etc) should be used downstream of the filter to capture any
particle-phase organics volatilized from the filter. To obtain meaningful data
from the denuder sampling, the collection efficiency of the denuder should
be either 100% or be accurately known for the species to be measured under
variety of ambient conditions. It has been shown (R. Rasmussen, private com-
munication) that the efficiency of activated charcoal denuders is greatly in-
fluenced by ambient humidity.
Analyses
A variety of methods are used to characterize organic carbon in atmospheric
PM samples. The methods may be divided into "total" or "bulk" methods
that characterize only certain properties of organic PM (such as organic car-
bon content, functional groups, isotope ratios, etc.) and molecular-level meth-
ods that characterize individual organic compounds.
"Bulk" Methods
• The "bulk" methods include thermal/optical carbon analysis (TOC) and
various spectroscopic methods. TOC allows for measuring and separating total
amount of organic and elemental carbon (OC/EC). The definition of OC and
EC is operational only and it is tied to the method of carbon measurement and
do not necessarily correspond to a physical meaning of "organic" or "elemen-
tal" carbon. For obtaining the estimation of organic compound mass concen-
tration, the OC concentration is generally multiplied by values ranging from 1.2
to approximately 1,8 to account for hydrogen, oxygen and other elements that
constitute organic molecules. However, this factor itself is a source of uncer-

-------
SOA Workshop Presentation Summaries & Research Recommendations
tainty, since it depends on organic compound composition, which may be
different in different locations. In remote locations, the higher contribution of
secondary organic aerosol, which contains higher proportions of oxygenated
(oxidized) compounds, would result in a higher average molecular weight per
carbon weight ratio.
•	Fourier trans for... infrared (FTIR), Raman, nuclear magnetic resonance (NMR)
and other spectroscopic methods provide functional group and bond informa-
tion. FTIR spectra can be obtained directly from ZnSe impactor substrates,
without extraction. The methods do not provide quantitative information, or
the information concerning individual compounds.
Molecular Level Methods
•	Organic compound speciation provides the most valuable information about
organic aerosol composition, sources, and atmospheric transformation pro-
cesses. Presendy it is not possible to completely resolve all organic carbon
mass into concentrations of specific organic compounds and no single analyti-
cal technique is capable of analyzing the entire range of organics. The molecu-
lar level methods usually require extraction of a sample with organic solvent(s),
followed by analysis by gas chromatography/mass spectrometry (GC/MS),
GC/FTTR/MS, GC with various detectors, HPLC/MS and other methods.
•	The most widely used analysis method for complex mixtures of organic
compounds is high-resolution capillary gas chromatography with mass spectro-
metric detection (GC/MS). However, GC/MS methods have typically resolved
only 10-15% of the organic mass into specific compounds. This is because
high-molecular organics (>C40) and highly polar compounds (especially
multifunctional) do not elute through a GC column. Polar organic compounds
require derivatization prior to analysis, to convert them into less polar and
more volatile derivatives that will elute through a GC column. However, the
derivatization techniques are compound-class specific and thus several different
methods may be required for a comprehensive analysis of one ambient sample.
The derivatization reagent by-products, the complexity of derivatization
products, lack of standards, and limited mass spectral libraries makes these
analyses difficult and time consuming.

-------
SOA Workshop Presentation Summaries & Research Recommendations 3 J
•	HPLC coupled with a mass spectrometer or a photodiode array detector seems
to be especially suitable for the analysis of polar organic compounds. Aqueous
solutions can be injected into reverse-phase columns, and polar compounds do
not need a derivatizadon step in order to elute from most of the LC columns.
However, LC columns offer less resolving power than GC columns and are
usually designed for a narrower compound class. In addition, although several
LC/MS systems are commercially available, they are not necessary optimized
for atmospheric research. Further development of separation methods and
mass spectral libraries is also needed.
•	Several new and promising methods have recently been proposed for a molecu-
lar-level organic aerosol characterization. For example, flash evaporation by
Curie point pyrolysis coupled with GC/MS (CPP-GC/MS) was used for direct
analysis of atmospheric semi-volatile organic compounds (Neususs et al.,
2000). The advantage of this method is that only a few micrograms of sample
is needed (thus it could be used with size-segregated sampling) and no sample
preparation is necessary. The disadvantage is that very polar compounds may
either not elute from a GC column, or be destroyed during a flash evaporation
process.
•	Capillary electrophoresis (CE) was recendy used (Neususs et al., 2000) for
analysis of dicarboxylic and hydroxy dicarboxylic acids, as well as the common
inorganic ions and methanesulfonate. In CE, ions are separated in a strong
electric field, because of their different electrophoretic mobilities. The advan-
tage of this method over ion chromatography and GC or HPLC is that
inorganic and organic ions can be analyzed in a single run. Also, the separation
efficiency is higher than in LC and the required sample amount is very low.
In situ analysis techniques
•	An automated carbon analyzer with a one-hour resolution time is now com-
mercially available from Sunset Laboratory, Inc. However, since it uses a quartz
filter as a substrate, it does not resolve the problem of positive/negative filter
artifacts.

-------
32 SOA Workshop Presentation Summaries & Research Recommendations
•	Single particle mass spectrometry is a promising technique for a real time
characterization of individual particles. Although there are some differences
between various : istruments, the principle of operation is to fragment each
particle into positive and/or negative ions using either a high-power laser or a
heated surface and to measure the ions by a time-of-flight mass spectrometer.
At present, quantitative determinations are difficult (or not possible) for this
technique, and the instruments are generally more suitable for inorganic than
organic species, but the future development of this technique could overcome
these challenges.
Summary and Recommendations
Although size-selective and denuder sampling methods are certainly very useful
for investigating the property of organic aerosol, a filter sampling method is
presendy the main method for ambient PM sample collection, due to its sim-
plicity, relatively low cost and a large sample size. To account for semi-vola-
tile organic compounds (SVOC), the filter should be followed by a solid ad-
sorbent, such as PUF plugs, XAD resins, or "sandwich" type PUF/XAD/
PUF cartridges. Use of a cut-off inlet (e.g. 2.5 mm) is also recommended.
Research is needed to:
•	simplify and standardize derivatization procedures;
•	develop more universal derivatization reagents, standards, and MS libraries;
•	further develop HPLC methods, especially LC/MS;
•	develop methods that do not require extraction, can be used on-site, and offer
better temporal and particle size resolution.
References
Neususs, C., Pelzing, M., Plewka, A., and Herrmann, H. (2000): A new analytical
approach for size-resolved speciation of organic compounds in atmospheric
aerosol particles: Methods and first results. Journal of Geophysical Research,
105, 4513-4527.

-------
SOA Workshop Presentation Summaries & Research Recommendations 33
Research Recommendations
1
The following 20 research rec-
ommendations are divided into
near- (1 to 2 years), middle- (3
to 5 years), and long- (5 to 10
years) time periods. Most rec-
ommendations cut across the
different topics addressed at the
workshop. Most recommenda-
tions emerged out of the plenary brainstorming session at the end of the
workshop in which each participant present shared their thoughts on what
needs to be done.
Ideally, each recommendation specifies an expected product, an approach
to obtaining that product, and a summary of how the product might be used
to support other research recommendations and practical applications.
NEAR-TERM RECOMMENDATIONS
Identify primary and secondary organic compounds and their proper-
ties.
This project would produce a database of specific organic compounds and
compound groups along with important properties. The data base would
include Chemical Abstract Service and common names for identified com-
pounds, references to reports of their detection, reported concentration ranges,
water activities, melting point, boiling point, vapor pressures, codes indicat-
ing primary or secondary or both, codes indicating potential sources of pre-
cursors, potential quantification methods, and detection limits. The data base
would be updatable as new information became available and downloadable
from a central location. Queries would allow users to extract data and to
place it into usable formats. This data base would be assembled from existing
tables created by atmospheric organic chemistry researchers via a survey of

-------
SOA Workshop Presentation Summaries & Research Recommendations
these researchers. It would be used to identify which compounds are lacking
data that need to be quantified in subsequent experiments. It could also be
used by decision-makers to determine organic compounds that might result
from different source emissions.
Specify thermal evolution carbon temperature fractions that separate
organic compounds into more logical groupings than currently applied
carbon fractions.
Review, evaluate, and compare light scattering and absorption models,
Document and evaluate procedures for detection of secondary organic
compound quantification.
Define reporting conventions, database, and priorities for aerosol
smog chamber experiments and results.
This project would provide a consistent set of reporting conventions for smog
chamber secondary organic aerosol experiments. Currently smog chamber
experiments tend to fall into two groups, those characterizing the dynamics
of aerosol formation and those emphasizing aerosol chemical speciation. Data
acquired during these experiments are not always presented in a consistent
format. Possible commonly reported data might include the following infor-
mation: temperature; type of lightsource; N02 photolysis rate; humidity; seed
particle concentration and type; chamber volume, material and surface/vol-
ume ratio; initial and final concentrations of VOC, NO, N02,03 and aerosol.
If a public database becomes available, it could also include detailed particle
distribution and speciation data and intermediate data in addition to the ini-
tial and final values- The establishment of smog chamber research priorities
would provide direction for experiments leading to a better understanding of

-------
SOA Workshop Presentation Summaries & Research Recommendations 35
the origins of secondary organic aerosol formation and the atmospheric con-
ditions that affect aerosol growth. The goals of this project could be accom-
plished by surveying current investigators in the field; however, a meeting of
these individuals would also prove valuable.
Evaluate methods to measure black carbon as a normalization for
primary and secondary organic carbon.
Define and organize follow-on topical workshops on organic aerosol
issues.
Evaluate national networks for optimal resource allegation.
MIDDLE-TERM RECOMMENDATIONS
Develop improved information extraction methods for current analyti-
cal methods.
Field and laboratory measurements of particulate hygroscopic proper-
ties.
Determine shapes, sizes, and surface reaction properties of particles.
Create and disseminate calibration and performance testing standards.
Measure and tabulate vapor pressures and water activities~

-------
SOA Workshop Presentation Summaries & Research Recommendations
Most of the SOA compounds have intermediate volatilities and there-
fore exist in both the gas and particulate phases in the atmosphere.
Their fraction in ute particulate phase depends strongly on tempera-
ture and on the concentrations of other organic PM components,
and also somewhat on relative humidity. While the framework for
understanding these partitioning processes exists, there is littie infor-
mation about the physical properties of the SOA compounds (vola-
tility, behavior in organic and aqueous solutions, etc.). We recom-
mend the measurement of these parameters and their dependence
on temperature and composition. A variety of approaches can be
used including the investigation of individual compounds, or the analy-
sis of appropriate smog chamber measurements.
Develop and apply extraction and derivatization procedures
that optimize organic aerosol recovery and quantification.
Organic compound speciation provides the most valuable informa-
tion about organic aerosol composition, sources, and atmospheric
transformation processes. The molecular level methods usually re-
quire extraction of a sample with organic solvent(s), followed by analy-
sis by gas chromatography/mass spectrometry (GC/MS), GC/FTIR/
MS, GC with various detectors, HPLC/MS and other methods. Se-
quential extractions with solvents of increasing polarity and liquid
chromatographic separations are frequendy used prior to GC/MS
analysis to simplify complex organic mixtures. There is a need to
optimize the selection of solvents and extraction procedures to as-
sure the integrity of less stable organic compounds, as well as a need
for development of more selective separation methods (particularly
solid phase extraction methods).
Highly polar compounds (especially multifunctional) do not elute
through a GC column. They require derivatization prior to analysis,
to be converted into less polar and more volatile derivatives that will

-------
SOA Workshop Presentation Summaries & Research Recommendations 37
elute through a GC column. The derivatization techniques are compound-
class specific and thus several different methods may be required for a com-
prehensive analysis of one ambient sample. The derivatization reagent by-
products, the complexity of derivatization products, lack of standards, and
limited mass spectral libraries makes these analyses difficult and time con-
suming. Since the derivatization methods are currently the main tool for po-
lar compound analysis, research is needed to simplify and standardize the
derivatization procedures. There is a need for better and more universal
derivatization reagents and less laborious procedures.
Field measurements of secondary precursors and end-products in
locations with contrasting source emissions and meteorology.
Monoterpenes, which are emitted by vegetation, and aromatic compounds,
originating from the production and consumption of petroleum fuels, are
two classes of gas-phase organic compounds that have been found to pro-
duce high aerosol yields in chamber experiments. Aerosol formation events
should be investigated in locations where emission rates of these precursors
and levels of atmospheric oxidants are expected to be large. Monoterpene
emission rates in the United States are greatest in the southeast, northeast
Texas, central and northern California, the Pacific Northwest, and high el-
evations of the Southwest. Forested ecosystems in the southeast, northeast
Texas, central California, and the Southwest are likely receptors of a complex
mixture of atmospheric oxidants from major urban art*s. Aromatic com-
pound emissions in Houston and Mexico City are large and produce ambient
levels in air that frequently exceed 5 ppbv. These urban areas would be good
locations for studies of aerosol formation from anthropogenic precursors.
Field experiments should focus on measuring precursors, oxidants, and the
likely products of the chemical oxidations to generate data for aerosol model
evaluation. Meteorological measurements to support the chemical measure-
ments are essential. It would be desirable to have surface sites established at
the candidate locations for long-term monitoring and for conducting inten-

-------
SOA Workshop Presentation Summaries & Research Recommendations
sive field campaigns when; e.g., measurements from aircraft could supple-
ment the surface observations.
Characterize primary emissions of secondary organic precursors and
primary oxygenated compounds.
Source types of particulate carbon in the majority of field studies have not
been characterized very well. For example, in one recent source apportion-
ment study from Los Angeles, only two medium-duty (rather low mileage)
diesel vehicles were used to collect source samples to construct a source sig-
nature to represent diesel exhaust.
It is critical to adequately sample the most important source types of
primary (and emissions of secondary organic) precursors from the sources
thought to be most important contributors to ambient PM2 5. To do so, ex-
amine the data from the NFRAQS (http://www.nfraqs.colostate.edu) for
mobile sources, and the new Gasoline/Diesel PM Split Study in Los Angeles,
to get an idea of what sample sizes and characteristics are needed to repre-
sent the on-road mobile fleet. In addition, samples from important off-road
sources are needed; for example, from locomotives and ship emissions, as
well as maybe some data from other source types. We need to understand the
relative importance of on- and off-road mobile source contributions to am-
bient data. This previous discussion covered only mobile sources; what's also
important are the other contributors, which may already be sufficiendy char-
acterized. One could examine the source profiles from the most recent PM
blame apportionment studies to see whether there are chemical differences
between similar sources from different studies-
Regarding oxygenated compounds, this whole issue became much more
complicated once the regulators started mandating oxygen contents of fuels.
We know already, for example, that diesels are important sources of primary
formaldehyde, and once oxygenates were added to gasoline, formaldehyde
and acetaldehyde became important emissions from spark ignition vehicles.
From Eric Fujita's work in the NFRAQS, we learned that along with PM,
we need to pay special attention to SVOC measurement and characterization,

-------
SOA Workshop Presentation Summaries & Research Recommendations 39
not only because these are important emission species, but we don't know
how to apportion them between particle and vapor phase; therefore, we don't
know well how to apportion them in ambient studies. So the best approach
is to collect the total exhaust as the sum of PM and SVOCs.
IMPROVE EC/OC data hinted that SOA might not be very important
once one understands the importance of smoke and other primary emission
sources at the regional sites.
Because the source selection and exhaust collection is so cosdy, we should
consider sacrificing some of the ambient measurements in favor of adequate/
sufficient source collection for development of source profiles. We will not
be able to do apportionment properly, and thereby properly characterize the
relative importance of primary and secondary organic aerosols, until this is
done correcdy.
Develop and test operational mechanisms for secondary aerosol
formation for forecasting models.
Aerosol extinction and other optical properties affect the photolysis rate pa-
rameters. For example, the presence of soot particles may decrease photoly-
sis rates by absorbing solar radiation. On the other hand, the presence of
other types of particles, particularly fine particles, could increase photolysis
rates through increases in radiation scattering. The effect of aerosol particles
on photolysis rates is especially important with respect to the formation of
secondary organic aerosols, because these aerosols are formed through pho-
tochemically driven processes. Measurements of spectrally resolved actinic
flux should be made in conjunction with other aerosol optical properties.
Photolysis rate parameters could then be directly calculated under different
aerosol containing atmospheric conditions. Comparison of these kinds of
measurements made in urban areas such as Los Angeles in the United States
than those made in much more polluted regions such as Mexico City or Shang-
hai would be helpful in determining the effects of aerosols on photolysis
rates.

-------
40
SOA Workshop Presentation Summaries & Research Recommendations
aerosol formation.
Develop detailed mechanisms and
models for secondary organic
While the description of the forma-
tion of the SOA compounds with a
LONG-TERM
RECOMMENDATIONS
constant yield (e.g., 2% of the oxidized precursor) is a first step it is an over-
simplification of the chemical processes leading to these products. Several
reactions steps are, in general, needed for the formation of the SOA species.
Development and testing of the chemical mechanisms leading to these com-
pounds based on smog chamber work will be necessary for the efficient con-
trol of these compounds (for example, for understanding the effect of NOs
on the formation rates of SOA). The corresponding detailed models should
be evaluated against field measurements of the concentrations of these com-
pounds.
Develop measurement methods that minimize changes in organic
composition from that in the atmosphere.
Develop and apply analytical methods to identify and quantify a larger
fraction of organic compounds and groups of compounds in sus-
pended particles.

-------