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. FormationThere 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 compoundsThey 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 phasesMost are semi-volatile and their aerosol fingerprints change continuously. 3) The reactivity of these compoundsSome 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 pitfallTobias 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,uv26experimental 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. ------- |