B. Research Activities
 
 

B.1 Scientific Background

Aerosols (suspensions of particles with diameters ranging from 0.001 to 10 m m) are important in many areas of environmental science. Their impact on visibility is well known [Malm et al., 1994]. They influence the thermal balance of the planet directly through absorption and scattering of solar radiation and indirectly by serving as nuclei for cloud droplets, thereby influencing the formation processes, lifetimes, and radiative properties of clouds. Changes in tropospheric aerosols are believed to result in a net decrease in radiative forcing, the magnitude of which is highly uncertain for a number of reasons, including their spatial and chemical heterogeneity and the lack of good observations constraining these properties [IPCC, 1995].

On an urban scale, aerosols adversely affect human health. Extreme air pollution episodes demonstrate that particulate-based air pollution could produce large increases in the daily mortality rate, for example, the "London Smog" of 1952 when nearly ten thousand deaths occurred immediately following a three-fold increase in particulate concentrations [Schwartz, 1994]. Recent epidemiological evidence, mostly gathered during the 1990s, reveals an association between ambient levels of airborne particles, common to many US cities, and morbidity and mortality. Observed health effects include increased respiratory symptoms, decreased lung function, increased hospitalizations, increased absenteeism from work or school, and increased cardiopulmonary disease mortality [Pope et al., 1995]. Fine particles, with diameters less than 2.5 um, are believed to be the more important health-related fraction of urban aerosols [Wilson and Spengler, 1996]. These particles can remain suspended in air for many hours or longer and have a high probability of deposition in the deep lung upon inhalation. Fine particles are also a major carrier of airborne infectious diseases, such as tuberculosis [Wells, 1955].

The distribution of trace gases in the atmosphere is influenced by aerosol particles that provide a catalytic medium for chemical reactions that otherwise are negligible in the gas phase. This "heterogeneous chemistry" is responsible for the majority of ozone destruction that has been observed in the polar regions in the winter and in the lowermost stratosphere throughout the year [WMO, 1995] and has been invoked to explain anomalous chemistry activation in the arctic boundary layer [Fan and Jacob, 1992]. Depletions of inorganic halides from the particulate phase have long been observed in the marine boundary layer [Keene et al., 1990] and it has been inferred that halogen radicals formed photochemically can alter the abundances of long-lived trace gases [Jobson et al., 1994]. If such processes occur on a global scale, the natural destruction of ozone near the earth's surface could proceed more rapidly than is currently believed.
 
 

B.2 Studies of Aerosol Chemistry and Composition

At the center of nearly all problems in atmospheric science that involve aerosols are the chemical composition and chemical reactivity of individual particles. Understanding of the current impact of aerosols and prediction of their possible health, environmental, and industrial consequences requires the development of models that accurately describe particulate formation, growth, evolution, and reactivity. Therefore, for many decades the major thrusts in the studies of aerosols have been primarily in the following areas: (1) studies of production and evolution of model aerosol particles, (2) characterization of the size distribution and phase of ambient aerosols, (3) sampling and analysis of the chemical composition of ambient particles, (4) determination of the chemical reactivity of surrogate particulate materials, and (5) prediction of the impact of aerosols in real systems using models that incorporate this information.

Unlike gases, however, whose atmospheric impact can be modeled confidently because their behavior in the atmosphere very likely mimics that in the laboratory, aerosols in real systems are far more complex and heterogeneous than their laboratory counterparts. Therefore, for many decades the Holy Grail for aerosol research has been the development of analytical techniques that can simultaneously characterize the size, morphology, and composition of individual aerosols without bias and without altering the particle during sampling [McMurry, 1998]. Although there is still a long way to go to attain this goal, there have been recent technological developments that show great promise [Johnston and Wexler, 1995].

Typically, the reactivity of atmospheric aerosol particles is assessed in detailed laboratory experiments where bulk materials are deposited on surfaces and the reactive gases and condensed-phase composition are studied with techniques such as Fourier transform infrared spectroscopy, mass spectrometry, and tunable-diode laser spectroscopy. It is assumed that the condensed phase is a good surrogate for ambient particles, but different experimental techniques have produced different results for the same chemical reactions. Thus, questions remain as to whether these model surfaces share the properties of microscopic particles. In addition, recent measurements of the composition of particles in "clean" regions of the atmosphere [Murphy et al., 1998] indicate that aerosols are a mixture of particles, the simplest of which possess a rich suite of compounds. Therefore, it is not clear to what extent the laboratory measurements on surrogates can be applied to the real atmosphere. Although techniques for analyzing individual particles have been available for decades, they involve expensive post-capture analysis that is costly and time consuming, and most are not suited to the study of volatile materials that are likely responsible for the reactivity of the condensed phase.

To address these issues there has been a recent concerted effort to develop new techniques for real-time characterization of individual atmospheric aerosol particles with minimum perturbation [Johnston and Wexler, 1995]. Ideally, these techniques must be highly sensitive to, and be able to rapidly differentiate between, thousands of elements and compounds. In addition, they must sample large numbers of particles over a multitude of sizes without bias, and they must be reliable and portable for deployment to remote locations without the need for constant supervision. Although advances in mass spectrometric and laser analytical techniques have fostered the development of very powerful instruments, they are still far from satisfying all these needs. They are, however, providing important new insights into the nature and diversity of atmospheric aerosols that will spawn a new generation of laboratory experiments.

Studies of ambient particles have received much attention and acclaim for the past two decades, however there has been little exploitation of field techniques in laboratory studies of the chemical reactivity of aerosol particles. This is due to the relatively high cost of the equipment necessary to carry out these studies and technical difficulties in building and operating these systems. It is pertinent to note that the data sets generated by these instruments can be extraordinarily large, and it is only within the past few years that technologies to process and store all this information have been readily affordable to the average laboratory experimentalist. Consequently, there are few laboratories that are equipped to study both the composition and chemical reactivity of aerosol particles or to validate single-particle composition measurement techniques.

This proposal is to develop a comprehensive system for laboratory studies of aerosol composition, chemistry, and measurement technologies using some of the most advanced techniques available for characterizing the composition of particles and for monitoring important correlative trace gases. Our objectives are to develop a versatile system for a new generation of laboratory experiments of the reactivity of aerosols, one that will also serve as a tool for evaluating and calibrating the techniques that are being developed for single particle analysis.
 
 

B.2 Proposed Studies

We propose to develop and build an Integrated System for Analysis of Aerosol Composition and Chemistry (ISAACC) at the University of Colorado, Boulder, in the heart of a region where atmospheric aerosols are a high research priority and where measurement technologies have been developed and commercialized. The three-year development of ISAACC will be carried out by students at CU who will be supervised by members from CU Boulder (Program in Atmospheric and Oceanic Sciences, Department of Chemistry, and College of Engineering), the NOAA Aeronomy Laboratory, the Cooperative Institute for Research in Environmental Sciences (CIRES) and the National Center for Atmospheric Research (NCAR). Examples of specific projects that will be conducted with ISAACC over the near term are outlined below. Our long-term objective is to bring together a group of experts in aerosol generation, measurement technology, heterogeneous chemistry, instrument design and fabrication, and atmospheric modeling from the greater Boulder/Denver community to address the most urgent issues in environmental aerosol science. Specific projects that would benefit from ISAACC are outlined in the next section. Others are discussed below in Section D.
 
 

B.2.1 Particles in the Upper Troposphere

Ice particles in the upper troposphere can be in the form of cirrus clouds, subvisible cirrus or aircraft contrails. Cirrus and subvisible cirrus are highly variable, with particles ranging in size from 2-3000 mm [Dowling and Radke, 1990] and surface area densities ranging from 20-20,000 mm2 cm-3 [Jensen et al., 1996]. Cirrus occur over ~30% of the earth at any instant, and subvisible cirrus may be present in the tropics 75% of the time [Jensen et al., 1996]. Typically in aircraft contrails, particles are smaller and number densities are higher than in natural cirrus [Gayet et al., 1996]. A study using satellite data showed that contrails may be present in the sky 60% of the time over Europe, with an areal coverage ranging from < 0.1% to 8% [Strauss et al., 1997]. These particles can provide ample sites for heterogeneous chemical reactions in the upper troposphere, yet their impact is hard to predict because the exact conditions necessary for particle nucleation are not well understood.

Aerosol particles in the upper troposphere are also composed of sulfate, which may be in the form of sulfuric acid, ammonium sulfate, ammonium bisulfate or letovicite. Such aerosols can possess ample surface area for heterogeneous reactions [Jensen et al., 1996]. It is thought that these particles are the nuclei on which cirrus clouds form, although by mechanisms is not well understood. To complicate matters, the composition of the particles has been shown to be highly variable and includes refractory and organic materials in addition to sulfate [Murphy et al., 1998, Twohy and Gandrud, 1998]. In addition, crustal and carbonaceous aerosols were found to dominate the ice nuclei in the upper troposphere [Chen et al., 1998].

There is also an abundance of chemical species in the troposphere that may undergo heterogeneous reaction. For example, oxygenated hydrocarbons such as acetone have only limited solubility in water and have been measured to be present in significant levels in the upper troposphere [Singh et al., 1995]. Reactive nitrogen (NOy) is also present in this region of the atmosphere, with recent measurements showing 100 pptv [Singh et al., 1996] in the background to 460 pptv in a young contrail [Arnold et al., 1992]. The oxidation/ reduction reaction of CH2O with HNO3 on sulfuric acid has already been studied in the laboratory [Jayne et al., 1996; Iraci and Tolbert, 1997]. This reaction could not only alter the ratio of NOy/NOx, but could also have a profound effect on HOx levels in the troposphere. Reactions that activate halogen species from reservoir forms can also impact tropospheric chemistry.
 
 

ISAACC will be ideal for studying the types of particles that occur in the upper troposphere. The Tolbert group is conducting experiments that focus on the composition, nucleation and chemical reactivity of aerosols, cirrus clouds and contrails in the upper troposphere. This work will be expanded to include studies of the chemistry and composition of cirrus-like particles at temperatures and pressures characteristic of the upper troposphere. In particular, we are interested in trying to reproduce the range of particles detected by PALMS (Particle Analysis by Laser Mass Spectrometry) on the NASA WB-57 during the WAM mission [Murphy et al., 1998] in an effort to assess the importance of particle heterogeneity in upper tropospheric chemistry and particle formation processes. Special focus will be on materials that are being studied in the Tolbert laboratory, especially particles composed of sulfate, soot, and metal-oxides. These studies can be extended to polar stratospheric clouds, a topic that remains a high priority in the stratospheric chemistry.
 
 

B.2.2 Chemistry of the high-latitude boundary layer

The phenomenon of sudden and complete boundary-layer ozone loss has been observed at many northern high-latitude sites over the past decade [Barrie et al., 1988; Oltmans et al., 1989; Bottenheim et al., 1990; Solberg et al., 1996; Miller et al., 1997] and more recently at sites in Antarctica [Kreher et al., 1997; Wessel et al., 1998]. Simultaneous observations of other species indicated that the ozone loss events were often coincident with increases in pollutants, such as NOx and sulfate aerosol (arising from SO2), suggesting a relationship to "Arctic Haze" - the transport of polluted air from the northern continents into the more pristine arctic environment. However, upon further investigation, a more solid correlation between ozone loss and enhancement of so-called "filterable bromine" was discovered [Barrie et al., 1988]. Springtime increases of filterable bromine (above that expected from dust, sea salt, and automobile exhaust) had been observed in a multi-year series of samples taken at Barrow, Alaska. These excesses typically occurred during spring and they were matched by a similar seasonal cycle in the abundances of gas-phase bromine-containing species such as bromoform (CHBr3) [Berg et al., 1983; 1984]. This fueled speculation that springtime algal blooms or decay of benthic algae might be emitting enormous quantities of these gases to the atmosphere through breaks in the sea ice [Barrie et al., 1988].

Subsequently, it was hypothesized, these brominated gases were photodissociated, releasing Br atoms, which could catalytically destroy ozone. Laboratory studies and modeling showed that the photolysis rate of CHBr3 is too slow to account for the observed time-dependence of the ozone loss (i.e., up to 90% losses in 3-5 days). Instead, attention turned to heterogeneous release of photolabile bromine species, such as BrNO2 or Br2, from sea-salt particles [Finlayson-Pitts et al., 1990; McConnell et al., 1992]. Flaws have been found with these mechanisms, such as the necessity of relatively large amounts of NOx, which are not always seen in association with the ozone loss events [for example, Beine et al., 1997; Kreher et al., 1997]. Nonetheless, the focus remains on a heterogeneous source for the initial active bromine [Curry and Radke, 1993; Mozurkewich, 1995] and for recycling of HBr (from reactions of Br with hydrocarbons) back to active forms [Fan and Jacob, 1992; Sander et al., 1997]. The most recent observations and models suggest that the ultimate source of active bromine is sea-salt that accumulates on the snow pack during winter. Bromine is released autocatalytically after polar sunrise based on a "seed" of active bromine provided by photolysis of biogenic bromocarbons [Tang and McConnell, 1996; Lehrer et al., 1997; Platt and Lehrer, 1997].

Such a mechanism could operate wherever there are significant sources of inorganic halogens and aerosols, such as in the marine boundary layer and over large, salt-rich continental seas. Tens of parts per trillion (ppt) of BrO, an ozone destroying form of bromine that can only be sustained by heterogeneous reactions, have been observed in the arctic boundary layer [Hausmann and Platt, 1994] over large geographic regions. Even larger abundances were observed recently in association with ozone decreases over the Black Sea [Hebestreit et al., 1999].
 
 

B.2.3 Boundary layer and urban aerosol chemistry

In the lower stratosphere and the high-latitude boundary layer, heterogeneous reactions play a key role in the balance of trace gas species. In relative terms, at lower latitudes the tropospheric boundary layer contains a large number of particles and high surface area. The average number concentration for the continental background is about 5000 cm-3 with a surface area density of 10-6 cm2 cm-3 (100 times the area in the lower stratosphere under non-volcanic conditions) [Finlayson-Pitts and Pitts, 1986]. Hence, there is a significant potential for heterogeneous uptake and chemistry to occur in the mid-latitude troposphere. Thus far, however, very few heterogeneous reactions are thought to affect the chemistry there [Ravishankara, 1997; Finlayson-Pitts and Pitts, 1997].

Although much of the particle mass in the troposphere consists of inorganic species such as ammonium nitrate, ammonium sulfate, sulfuric acid, sea salt, soot, and soil minerals, a substantial fraction can be attributed to organic compounds. For example, organic species comprise 15-30% of the total fine particle mass in urban environments [Rogge et al., 1993; Saxena et al., 1995], and 20-60% of the mass in rural areas of the United States [Duce et al., 1983; Saxena et al., 1995]. These organic compounds are introduced to the atmosphere by natural and anthropogenic sources and may be formed by primary emissions or after secondary chemistry. In addition, the organic components in particles can be from a wide range of classifications and are not well characterized. At best, less than 5 % of the particle mass has been identified as over 80 different organic compounds [Rogge et al., 1993].

Two categories of studies related to organic material in aerosols may be explored with the proposed ISAACC equipment. One is to understand how inorganic aerosols take up by organic compounds. Indeed, sulfuric acid particles are known to become easily contaminated with gas-phase organic species [Middlebrook et al., 1997]. In fact, incorporation of organic compounds by sulfate aerosols may explain discrepancies between measured nanoparticle growth rates and those modeled assuming that growth is limited by the sulfuric acid condensation rate [Marti et al., 1997]. Physical uptake of gas-phase organic molecules by atmospheric aerosols can be comparable to other loss processes and may be significant for species with long chemical lifetimes. For example, assuming an uptake coefficient of g = 0.01 and an aerosol surface area density of A = 10-6 cm2 cm-3, the lifetime of toluene with respect to heterogeneous loss is 4.3 hours, compared to 4.5 hours by gas-phase reaction with hydroxyl radical [Finlayson-Pitts and Pitts, 1986]. Furthermore, secondary chemical reactions, primarily oxidation of gas-phase organic compounds, contribute to condensed-phase organic material [Kavouras et al., 1999 and references therein]. Nighttime oxidation by nitrate radical is also thought to produce organo-nitrate species with low vapor pressures which can condense onto preexisting aerosols [Klotz et al., 1998; Martínez et al., 1998]. Therefore, we propose to study the uptake of unreacted and oxidized organic compounds by inorganic aerosols.

Once organic compounds condense onto particles, they may undergo further chemical reactions. With the proposed ISAACC instrumentation, we may examine organic aerosols after exposure to oxidants such as ozone and hydroxyl, halogen, and nitrate radicals. Although the condensed phase was not probed, liquid unsaturated alkanes have been found to be more reactive to ozone than other liquid hydrocarbons [de Gouw and Lovejoy, 1998]. Reactions with oxidants may have a profound effect on the composition and physical properties of the condensed phase. Indeed, aging of urban aerosols may render them more hygroscopic [Saxena et al., 1995]. Oxidation by hydroxyl radicals has also been proposed to alter the hygroscopic growth properties of sea salt aerosols [Ellison et al., 1999]. In some instances, the presence of condensed-phase organic compounds may increase the reactivity of aerosols with respect to gas-phase oxidants. In fact, Cooper and Abbatt [1996] found that a surface coating of 1-hexanol increased the reactive uptake of gas-phase hydroxyl radical by ammonium sulfate surfaces. However, more experimental studies in conjunction with modeling are needed to complete the picture of how organic material in tropospheric aerosols might be changing the aerosol composition and properties as well as the balance of gas-phase species.
 
 

B.2.4 Evaluation and calibration of single-particle measurements

Early measurements of aerosol composition used microanalytical techniques, where particles are collected on solid substrates for subsequent analysis, each technique revealing an important aspect of individual particles. Lasers have been used to analyze the composition of single particles collected in a similar manner. In laser microprobe mass spectrometry (LMMS), the substrate is heated rapidly by a pulsed ultraviolet laser (typically a frequency-quadrupled Nd:YAG) to evaporate the particle and ionize a small fraction of the vapor, which is subsequently detected by time-of-flight mass spectrometry. Generally, several techniques in tandem are needed to adequately characterize the important properties of particles; however, such analyses are limited to particles with low volatility whereas atmospheric particulates are often volatile and their composition and size can adjust rapidly to changes in humidity and temperature.

Recent advances in ultraviolet laser and mass spectrometric technologies have facilitated the development of new instruments for rapid in situ analysis of individual particles under ambient conditions. One noteworthy example, particle analysis by laser mass spectrometry (PALMS), was deployed recently on the NASA WB-57 high-altitude aircraft [Murphy et al., 1998]. Tens of thousands of particles in the upper troposphere and lower stratosphere were analyzed and found to be composed of a wide variety of compounds. The particles were found to contain compounds such as refractory minerals, metals, oxidized nitrogen and sulfur compounds, carbonates, soot, carboxylic acids, and inorganic halogens. Related techniques employed at the ground identified a wide range of condensable material that could be related to local sources such as mineral dust, soot, and salts from oceanic and anthropogenic salts [Middlebrook, 1997 and Gard et al., 1998.]

Although laser-ionization, mass-spectrometric techniques can be quite sensitive to trace materials, because the plasma ionization process is complex and highly non-linear, they share the feature that the sensitivity differs for different elements and compounds. Species that do not readily form ions in the plasma created by the laser are difficult, and in some cases impossible, to detect. Thus, to date such measurements have provided qualitative information, for example, identification of certain elements and statistical distributions of types of particles and substrates, rather than concentrations of trace elements. However, the technique shows great promise, and there are a number of groups that are using variants on laser-ablation mass spectrometry of single particles to address a number of urgent problems in atmospheric chemistry [Johnston and Wexler, 1995]. Accordingly, a recent NRC panel’s highest-priority research recommendation was to "Improve the ability of continuous particle-by-particle analysis systems to quantitatively determine the composition of individual particles and to quantitatively characterize their ambient distribution" [NRC, 1998]

As more stringent regulations on exposure to particulate matter (PM) are developed [EPA, 1997], instruments to measure and monitor the composition of aerosol particles in real time will most certainly be commercialized and be frequently deployed. It is highly likely that such systems will be set up to permanently monitor particulates in major urban areas. Therefore, over the next two decades there will be a critical need for facilities that test new aerosol instruments and maintain their calibrations. These will require the skills of professionals professionals trained in the most advanced particle measurement techniques. Therefore, a major goal of this project is to work closely with colleagues who have developed the first generation of field-deployable single-particle instruments to examine their performance under carefully controlled laboratory conditions.
 
 

B.3 Senior Personnel
 
 

B.3.1 Darin Toohey (Program in Atmospheric and Oceanic Sciences)

Prof. Toohey conducts laboratory and field experiments that address the photochemistry and kinetics of trace gases in the atmosphere. He has experience with production and calibration of highly reactive radicals and spectroscopic detection and quantification at part-per-trillion abundances in the lab and in situ. He has employed various analytical techniques in his work, including laser magnetic resonance, resonance fluorescence, and laser-induced fluorescence. His group conducts its own mechanical design (using AutoCad and SolidWorks), layout and assembly of electronic circuits and multi-layer printed circuit boards (using PSpice and PADS), and software control (QuickBasic, C++, and Fortran). They have designed and built lightweight, autonomous instruments that have been deployed on the ground and on high-altitude aircraft and balloons from remote locations, including Spitzbergen and Antarctica, with an excellent record of success.

Prof. Toohey’s current projects include studies of halogen activation by heterogeneous reactions on background aerosols in the lower stratosphere and upper troposphere using small balloons (with Prof. Terry Deshler at the University of Wyoming), measurements of halogen chemistry and particle morphology in plumes of solid rocket motors (with Dr. William Bowers of Femtometrics Inc.), and measurements of halogen radicals in the marine boundary layer (with Prof. Linnea Avallone at CU Boulder) and the middle-to-upper troposphere and (with Prof. William Brune at Penn State).
 
 

B.3.2 Margaret Tolbert (Department of Chemistry/CIRES)

Prof. Tolbert and her research group conduct laboratory experiments to probe the chemical composition, formation mechanisms and chemical reactivity of atmospheric clouds and aerosols. Her early laboratory work demonstrated that heterogeneous chlorine activation reactions on polar stratospheric clouds (PSCs) played a key role in polar ozone depletion. Her subsequent studies of reactions on sulfuric acid surfaces showed that chlorine activation on the global sulfate aerosol layer could be important for global ozone loss. Although the importance of this heterogeneous chemistry is now well established, there are still major uncertainties about the chemical composition, phase and formation mechanisms of the clouds themselves.

Prof. Tolbert and her students perform several types of experiments to unravel the composition, phase and formation mechanisms of PSCs. Laboratory studies on thin film samples modeling PSCs and using freely floating aerosols with compositions mimicking PSCs elucidate the mechanisms for stratospheric particle growth, composition, and chemical reactivity. These studies use infrared extinction to determine PSC composition and mass spectrometry to measure the gas phase composition. Additional measurements of the infrared optical constants for PSC materials are being made for use in remote detection of the cloud composition. Finally, to determine the overall importance of heterogeneous chemistry in the troposphere, about which much less is known, ongoing studies in the Tolbert group are probing the formation mechanisms and chemical reactivity of cirrus clouds and other model tropospheric aerosols such as salts, soot and mineral aerosols.
 
 

B.3.3 Shelly Miller (Department of Mechanical Engineering)

Prof. Shelly Miller is an expert in indoor air quality, a field in which she has been actively engaged in research since 1991. Her group assesses exposures to indoor air pollutants and develops and evaluates indoor air quality control measures. She has extensive experience conducting chamber experiments, generating and measuring aerosols and bioaerosols, and modeling indoor environments. Dr. Miller's current research projects include modeling studies of personal exposure to toxic air contaminants from environmental tobacco smoke, chamber experiments that assess technological controls to reduce infectious disease transmission, and field studies of the relationship between penetration, deposition, and the concentrations of airborne particles indoors.
 
 

B.3.4 Ann Middlebrook (Univ. Colorado Cooperative Institute for Research in Environmental Sciences and NOAA Aeronomy Laboratory)

Dr. Ann Middlebrook's principal research focus is on the development and application of new methods for real-time composition measurements of atmospheric aerosols. She assisted with the development of NOAA's particle analysis by laser mass spectrometry (PALMS) instrument with which she conducted field measurements of aerosols in the marine boundary layer and urban environments. She carried out extensive laboratory work to characterize this instrument with a variety of aerosol analysis tools. Currently she is developing a new technique to remove particles from ambient air for subsequent real-time composition analysis (with Dan Murphy of NOAA and John Birks of CU Boulder).

Dr. Middlebrook has extensive experience with particle generation, physical and chemical characterization of particles, and heterogeneous chemistry of atmospheric particles. She has performed laboratory studies that involved preparation of a wide variety of particles using an aerosol nebulizer with a diffusion dryer and charge neutralizer, a custom sulfuric acid aerosol generator, a custom differential mobility analyzer, a custom relative humidity control system, optical particle counters, and condensation particle counters. She has also examined uptake of organic compounds and radical species by several typical atmospheric aerosol surfaces.
 
 

B.3.5 Brian Toon (Program in Atmospheric and Oceanic Sciences/LASP)

Prof. Toon studies the radiative and chemical influences of aerosols on planetary atmospheres. Dr. Toon is active in research areas ranging from theoretical and experimental studies of clouds and aerosols in the earth's stratosphere and troposphere to investigations of planetary atmospheres with the goal of understanding the climates of the terrestrial planets. He has been a leading figure in studies of the antarctic ozone hole, including organizing and leading a number of NASA-sponsored aircraft missions to study chemistry, dynamics, and radiation balance in the upper troposphere and lower stratosphere. His theoretical investigations of particle formation processes and transition behavior have shed important light on natural and anthropogenic influences to earth's ozone layer and climate. They have also revealed a number of issues that raise questions about classical particle physics.

Prof. Toon is a co-organizer and co-mission scientist of the 1999-2000 NASA SOLVE mission which will use the NASA ER-2 and DC-8 aircraft and balloons to study the processes that lead to ozone loss in the arctic lower stratosphere.
 
 

B.3.6 Jana Milford (Department of Mechanical Engineering)

Jana Milford, Associate Professor of Mechanical Engineering, is an expert on urban-scale air pollution. Her research interests focus on mathematical modeling of photochemical air quality. She is currently conducting research on chemical reactivities of volatile organic compounds with respect to formation of ozone and secondary organic aerosols; application of receptor models for determining source contributions to human exposure; and uncertainty analysis and parameter estimation for urban-scale air quality models. She has extensive experience in model development, application and evaluation, and in data analysis techniques including time series and factor analysis and neural network modeling.

Prof. Milford received her Ph.D. in Engineering and Public Policy from Carnegie Mellon University in 1988. She also holds a M.S. in Civil Engineering from Carnegie Mellon, and a B.S. in Engineering Science from Iowa State University. Prior to joining the University of Colorado she was an Assistant Professor of Civil and Environmental Engineering at the University of Connecticut, and a Congressional Fellow with the Office of Technology Assessment of the U.S. Congress.
 
 
 
 

B.3.7 John Birks (Department of Chemistry/CIRES)

John Birks' research involves the development and deployment of analytical techniques for measurements of trace gases in the earth's atmosphere and elemental analysis of aerosol particles. His group has recently demonstrated that atomic force microscopy can be used to observe the detailed kinetic behavior of erosion of a single particle by ozone. He has conducted studies as diverse as laboratory kinetics of stratospheric radicals and the theory of nuclear winter. He is presently collaborating with Ann Middlebrook and Dan Murphy at the NOAA Aeronomy Laboratory to develop new sampling inlets for studying the improving the ability to sample the chemical composition of aerosols.
 
 

B.3.8 Robert Sievers (Department of Chemistry and

Director, Global Change and Environmental Quality Program)

Prof. Sievers is interested in environmental, analytical, and inorganic chemistry; materials science; trace analysis; chromatography; mass spectrometry; metal chelate chemistry; aerosols; supercritical fluids; thin film deposition; and atmospheric and aquatic chemistry. His group is actively involved in fundamental and applied studies of the formation of fine aerosol particles using supercritical fluid techniques. In addition, they are developing new approaches to ultra-trace analysis at picogram levels based on gas, liquid, and supercritical fluid chromatography. Much of the research in trace analysis is aimed at problems in environmental chemistry; where high sensitivity and selectivity are the principal objectives. They are invesitgating the distribution, effects, transport, sources, sinks, and chemistry of trace species and organic compounds in the environment are being investigated. Prof. Sievers also has extensive experience with private industry in the Boulder area, where he previously founded the company Sievers Instruments.
 
 

B.4 Other contributors

We have assembled a group of collaborators from nearby laboratories and industry, including Dr. Darrel Baumgardner and Bruce Gandrud from NCAR, Dr. Dan Murphy and Charles Brock from NOAA, Dr. John Jayne from Aerodyne Corp., and Dr. William Bowers from Femtometrics. These investigators have pledged their support of this project. We plan on consulting with them extensively as we design and build the instruments. They have included letters of support for this project. They bring expertise that is too extensive to go into here, but most importantly they are all involved in development of specialized and field measurements that will provide an important synergy with this project. In addition, we hope to collaborate with them as much as possible by sharing students and by offering our facilities for their use, such as explicitly outlined above with respect to testing of the PALMS instrument.

In addition, we have received an enthusiastic response from our colleagues at CU, including Prof. Barney Ellison, Prof. Veronica Vaida, and Prof. Steve George. One of their ongoing projects is described in Section D.
 
 

C. Description of Research Instrumentation
 
 

C.1 ISAACC - An overview

The apparatus we will develop will have instruments to produce and characterize aerosol particles of known composition, a reaction chamber in which the particles will be exposed to various reactive gases under atmospheric conditions, techniques for monitoring the concentrations of reactive species in the gas-phase, optical particle counters and sizers to monitor the changes in the particles within the reactor in real time, and finally a real-time particle composition detector. These systems are described briefly below.
 
 

C.1.1 Aerosol generation

Model aerosol particles with well established compositions will be generated using several methods. Salt aerosols will be generated by atomization. In this technique, a solution of known composition is atomized using a TSI 3076 constant output atomizer with accompanying syringe pump. This method produces a polydisperse distribution of salt solution droplets with a number density >107 particles cm-3. The droplets can then be dried using a TSI diffusion drier. To re-humidify the particles in a controlled fashion, the particles are passed over a well mixed, temperature controlled sulfuric acid/water bath. The sulfuric acid concentration and bath temperature determine the resulting humidity in the particle flow. The humidity level, in turn, determines the phase of the salt particle. The relative humidity of the aerosol flow is measured in several ways. First a dew point hygrometer (EdgeTech 2002-S2) is used to determine the total water content of the flow. If the particles are dry, the total water content is equal to the gas phase water. If the particles are wet, the relative humidity must be determined by another method. We have used FTIR spectroscopy to determine the gas phase water content in the presence of condensed aerosols. Diode laser spectroscopy can also be used in situ to determine gaseous water. This technique of aerosol generation also works well for preparing sulfuric acid particles, either in a pure state or with dissolved substituents.

While the above system works extremely well for generating pure and mixed salt aerosol particles, it is limited to chemical compositions that are soluble. To generate dry particles, a fluidized bed aerosol generator (TSI 3400) that de-agglomerates powder samples is used. This source yields a constant output with respect to aerosol size distribution and number concentration. For this source, the powder of interest is mixed with bronze beads that are systematically fed into the bed. Compressed nitrogen is passed through the bed from below until the pressure drop across the bed creates a force that exceeds the force on the bed due to gravity, causing fluidization. The fluidizing action of the bed material de-agglomerates the powder that is then entrained in the air stream forming an aerosol. The aerosol then passes through a polonium neutralizer before flowing into the aerosol flow tube. The bed material is not eluted in this process due to the relatively larger mass of the bed particles.

The above sources are able to generate a variety of solid and liquid aerosol particles. However, it is also of interest to generate mixed aerosols with well known compositions. Solid particles are readily coated with sulfuric acid using a linear temperature drop oven [Scot Martin, personal communication]. Using this method, solid non-volatile particles and sulfuric acid particles pass through the flow tube oven where high temperatures vaporize the sulfuric acid but not the non-volatiles. As the aerosol cools, a linear temperature gradient across the oven ensures that each solid particle becomes coated with sulfuric acid and that sulfuric acid does not homogeneously nucleate. Other volatile species such as nitric acid can also be readily added to the particles using the oven.

Knowing the aerosol surface area is important for quantifying uptake of trace gas species and it is therefore critical to measure both the size and number concentration of the particles being generated. The size distribution of the particles will be measured at ambient temperature and pressure conditions with a Scanning Mobility Particle Sizer (TSI 3934L) upstream of the flow reactor. This equipment combines an electrostatic classifyier with a condensation particle counter and can measure particles from 10 nm to 1 micron in diameter. The advantage of using an electrostatic classifyier with a condensation particle counter instead of an aerodynamic particle sizer or an optical particle counter is that experiments could be performed as a function of particle size.
 
 

C.1.2 Aerosol reaction chamber

Following generation and characterization, aerosol particles will be injected into a temperature and pressure controlled, slow-flow (~1 m/s) reaction tube where they will be exposed to various reactive gases. For reaction probabilities typical of the atmosphere, for chemical reactivity measurements the particle number densities will have to be substantially higher than ambient. This poses no serious problem to the detection methods we have chosen, except that it may mean that the aerodynamic inlet of the aerosol mass spectrometer will have to be cleaned more than usual. However, because the gas-phase detection techniques will require signal averaging to achieve adequate signal-to-noise ratios at ambient concentrations, the number of particles sampled by the particle mass spectrometer will have to be averaged as well. So long as the chemical reactivities of the particles are similar, this will not pose a problem for interpretation. We will carry out detailed studies where the chemical composition of the particles is varied to examine this issue.

Prof. Toohey has experience in the design of "wall-less" reactors for laboratory and field studies of highly reactive radicals. In the case of the aerosol studies, we will develop a special chamber that uses injection of nitrogen (or air) into the boundary layer near the wall to reduce the rate of loss (or reaction) of particles at the walls. In addition, the walls will have to be free from large obstructions that create flow perturbations. Thus, the chemical sampling techniques will require specialized optical coupling and baffling to the system, similar to what is required for the measurements of highly reactive free radicals on fast moving platforms. This work will be challenging, but preliminary tests in our lab indicate that the system required for aerosol studies will look similar to those we use in our laboratory gas-phase kinetics systems.

For studies of aerosol deliquescence and uptake, we can use smaller abundances of particles, because our main focus will be on the equilibrium and non-equilibrium physical nature and chemical composition of the particles, rather than on concentrations of trace reactive compounds in the gas phase. Initial studies will be carried out at room temperature and moderate pressures, to mimic the boundary layer. Subsequent experiments will explore the behavior of particles at lower pressures and lower temperatures. This work will include the validation of the PALMS instrument. As we gain more experience, we will begin low temperature, lower pressure kinetics experiments. These are likely to begin only at the end of this three-year development project.
 
 

C.1.3 In situ gas and particle analysis

Abundances of gas-phase species, like H2O, HNO3, organics, and inorganic acids, will be monitored by one of three techniques: chemical ionization mass spectrometry (CIMS), tunable diode laser spectroscopy (TDLS), or fourier transform infrared spectroscopy (FTIR). The FTIR system will be purchased directly from Mattson. Prof. Tolbert's group has extensive experience with this system, including its use in detecting the phase and composition of particles containing nitric acid. The TDLS will be designed and built by the students, with some help anticipated from colleagues at the NOAA Aeronomy Laboratory and Prof. Linnea Avallone, who measure CH4 and H2O with a new generation of systems that can be operated in either open- or closed-path modes. Prof. Avallone's instrument, built by Randy May of JPL has flown successfully recently on the NCAR C-130, and is both lightweight, and extremely sensitive to water vapor under dry conditions typical of the upper troposphere. This type of instrument (purchase price of ~ $30,000) could be used to monitor total water vapor. For other species, with smaller concentrations (like HNO3), a longer path will be required. We will build a Herriot absorption cell with approximately 100 passes, for a path length of ~100 meters. This should be sufficient to measure in situ species such as HCl and HNO3 (and of course, water) at abundances typical of ambient. The number of passes will have to be restricted so that the particles to not alter with the path of the beam.

The CIMS instrument will be designed and built at CU, to specifications typical of those instruments that are used in the laboratory heterogeneous reaction studies. This system will consist of a low pressure (a few torr) secondary flow tube attached to the main aerosol chamber, through which a small flow of sample will pass. Through the use of various types of inlets (to be designed by Prof. Birks' group, in collaboration with Ann Middlebrook), this system can sample the gas phase composition, or the sum of gas and particulate phase (for example, by passing the sample over a flow heater). There are a number of ion exchange reactions that can be used to detect a wide range of species, including reactions of halide ions [Amelynck et al., 1998], SF6-, and water clusters [Arijs et al., 1998]. Once again, there is a strong heritage of similar measurements at the NOAA Aeronomy Laboratory, including in the laboratory as well as on research aircraft. Details of the technique are described elsewhere [Mohler et al., 1993, Viidanoja et al., 1998].

Aerosol size and number will be monitored in situ with a modified optical particle counter system from Droplet Measurement Technologies, Inc. of Boulder Colorado. Darrel Baumgardner of NCAR will assist in the modification, integration, and use of this system at CU. He is presently involved in a collaboration with Prof. Tolbert's group. These measurements will allow us to track the evolution of the particles after they have been injected into the chamber and have been exposed to different reactive gases.
 
 

C.1.4 Aerosol Mass Spectrometer

This project apparatus will utilize an aerosol mass spectrometer from Aerodyne Research Inc. [Jayne et al., 1996]. We have chosen this system for the following reasons: (1) measurements are obtained in situ and in real time, (2) it detects a wide variety of compounds, (3) it is quantitative for inorganic and organic compounds, (4) it obtains size-dependent composition, (5) it is sensitive to single particles. With this instrument, particles are sampled from ambient conditions into a high vacuum chamber through an aerosol focusing inlet [Liu et al., 1995a,b; Schreiner et al., 1998]. This inlet forms a particle beam that is chopped at a fixed frequency. Particles passing though the chopper are impinged onto a resisitively-heated filament (temperature ~1500 °C) where the volatile and semi-volatile components are flash vaporized. The vapors from the particle are ionized by electron impact and the ions are detected by a standard quadrupole mass spectrometer, yielding molecular composition information.

In addition to generating a particle beam, the aerosol focusing inlet allows a high fraction of particles to be analyzed. In fact, the measured particle collection efficiency is approximately unity for most particle sizes, dropping off due to impaction for larger particles and due to weak focusing of smaller particles [Jayne et al., 1998].

Chopping the particle beam allows the aerodynamic particle size to be determined by the aerosol time-of-flight to the vaporization filament. In tests of the aerosol time-of-flight response, the particle velocity was found to be proportional to the aerodynamic diameter for a variety of different aerosol compositions [Jayne et al., 1998]. Furthermore, the time scale for the aerosol time-of-flight is on the order of milliseconds, whereas the time scale for analysis is approximately 10s of microseconds. Thus, ion signals in the mass spectrometer may be obtained for individual particles when the particle concentration is low.

The instrument may be operated in one of two modes: scanning the entire spectrum from several particles or monitoring a few selected ion peaks with single particle sensitivity. In laboratory experiments, the first mode may be used to determine which peaks should be monitored in the second mode. Because the chopper stops the particle beam from impinging onto the vaporization filament, the gas-phase background signal can easily be removed from particle signals.

The clear advantage of choosing Aerodyne’s aerosol mass spectrometer over others using laser ionization is that the ion signals are quantitative with electron impact ionization. Instruments using electron impact ionization have been shown to be quantitative for a wide variety of inorganic and organic species [Allen and Gould, 1981; Sinha et al., 1982; Sinha and Friedlander, 1985]. Indeed, the Aerodyne instrument is quantitative over two orders of magnitude in volume for dioctyl sebacate ions and has single particle sensitivity for particles as small as 110 nm in diameter [Jayne et al., 1996]. In the future, intercomparisons of our aerosol mass spectrometer with NOAA’s particle analysis by laser mass spectrometry (PALMS) instrument will be possible due to the close proximity of the University of Colorado to the NOAA Aeronomy Laboratory.

One disadvantage of this technique is that it may not be very sensitive to ammonium ions [Allen and Gould, 1981]. Another potential disadvantage is that non-volatile particles will be difficult, perhaps impossible, to analyze. Although the temperature of the vaporization filament could be increased, a higher temperature could cause ionization to occur at the vaporization filament and may necessitate recalibration of the ion pattern with the hotter filament. Furthermore, if the filament is operated at a specified temperature and the particle contains less volatile components, the ion signals detected by the mass spectrometer from those components would be formed over longer time scales and may overlap with signals from other particles.

If the mass spectrometer is changed from a quadrupole to a time-of-flight, the entire mass spectrum may be obtained for ions produced from a single particle. However, slow volatilization rates of less volatile species may limit the ability of an ion burst being formed for time-of-flight analysis. A better solution could be to switch to an ion trap mass spectrometer. Ion traps have a distinct advantage over both quadrupole and time-of-flight mass spectrometers because ions may be collected for each mass spectrum. Thus, particle volatilization can occur over time scales longer than the 10s of microseconds with the current aerosol mass spectrometer. Furthermore, ion traps operate at higher than standard quadrupole or time-of-flight mass spectrometers.
 
 

D. Impact on Infrastructure Projects (Maximum length 2 pages)

This project has the potential for a strong influence on the CU atmospheric sciences community. Presently, research on aerosols is conducted in at least three separate departments on campus, as well as by affiliates at NOAA. This project will bring these researchers together in a pro-active way, and will likely result in a number of new collaborations that will address some of the most pressing issues in aerosol science. In addition, it will bring much needed equipment to the CU campus where students will learn a number of important analytical techniques, rather than just a few. We expect the collaborations with our colleagues in industry and at the local laboratories like NOAA and NCAR to have a strong impact on the communication and infrastructure of not only CU, but the labs and instrustry as well. We hope that our partners will call upon us to test and evaluate their new hardware as it is developed and before it is commercialized.

In addition, ISAACC will have an immediate impact on several projects to be carried out by colleagues at CU Boulder. Some of these are outlined in more detail below.
 
 

Development of a Technique for Real-Time Elemental Analysis of Aerosol Particles.

The research groups of John Birks (C.U.) and of Dan Murphy and Anne Middlebrook (NOAA Aeronomy Laboratory) are collaborating on the development of an "exchanging virtual impactor" for transferring particles having aerodynamic diameters greater than 0.1 m m from sampled air into an inert gas such as helium. Air is accelerated through a jet orifice having a diameter of 1 mm and collected in a second orifice of a similar diameter. A counter flow of inert gas prevents air molecules from entering the collection orifice. The greater momentum of aerosol particles, however, allows them to pass penetrate the counter flow be collected. The impactor both concentrates particles and removes them from surrounding air, making it possible to apply a variety of elemental analysis techniques. Current work is focused on determining the N, S and C contents of aerosols in real time. The total reactive nitrogen content, NOy, will be measured by conversion of species to NO within a molybdenum oxide catalyst tube at elevated temperature, followed by either luminol or NO + O3 chemiluminescence. The elemental sulfur content will be measured by use of the Sulfur Chemiluminescence Detector (SCD; Sievers Instruments, Boulder, CO) in which a fraction of the sulfur compounds are converted to SO in a hydrogen-rich flame and SO + O3 chemiluminescence measured. Carbon will be measured by oxidation to CO2 followed by a high-precision conductivity measurement (Kuck and Birks, 1998). The use of an atomic emission detector (Hewlett-Packard) for elemental analysis also will be investigated.
 
 

Development of a Method to Measure the Size Distributions of the Organic, Inorganic and Graphitic Components of Aerosols Using Atomic Force Microscopy (AFM).

The Birks group is developing a technique for determining size distributions and other physical characteristics of atmospheric aerosols using atomic force microscopy (Lehmpuhl et al., 1999). Particles are collected on an atomically flat surface such as mica and the size distribution obtained by AFM. The sample is then exposed to ozone that is found to completely volatilize all components of aerosol particles, forming CO, CO2 and H2O, and a new AFM image obtained. Finally, the sample is exposed to ground-state oxygen atoms, which vaporizes graphitic particles but not inorganic particles, and a final image obtained. From the three images, it is possible to assign each particle to a general class, either organic, graphitic, or inorganic, and to construct the original size distribution for each chemical class.

Veronica Vaida and Barney Ellison have developed an explicit model for organic aerosols [Ellison, et al. 1999]. This is a chemical model for the composition, structure, and atmospheric processing of organic aerosols. Our model is stimulated by recent field measurements showing that organic compounds are a significant component of atmospheric aerosols. The proposed model organic aerosol is an "inverted micelle" consisting of an aqueous core that is encapsulated in an inert, hydrophobic organic monolayer. The organic material that coats the aerosol particles are surfactants of biological origin. We propose a chemical mechanism by which the organic surface layer will be processed by reactions with atmospheric radicals. The net result of an organic aerosol being exposed to an oxidizing atmosphere is the transformation of an inert hydrophobic film to a reactive hydrophilic layer. The chemical model yields certain predictions that are testable by observations. Among them is a curve of the percent organic material as a function of particle diameter that predicts that a high fraction of the mass of the upper tropospheric aerosol will be organic. Atmospheric processing of organic aerosols will lead to the release of small organic fragments into the troposphere that will play a subsequent role in homogeneous chemistry. Organic aerosols are likely to act as a transport vehicle of organics into the atmosphere.

Ellison has just initiated a laboratory program that is funded by the Organic & Macromolecular Chemistry Division of NSF; Hydrocarbon Thin Films as Models for Organic Aerosols. Vaida and Ellison have just won support from the Camille and Henry Dreyfus Foundation to support an environmental postdoctoral associate; Postdoctoral Opportunities in Laboratory and Field Measurements and Modeling in Environmental and Atmospheric Processes. These awards will support extensive laboratory studies of the hydrocarbon radical chemistry that underpins the atmospheric processing of aerosols. As our chemical model for the structure and reactivity of organic aerosols progresses, we plan to use the "Integrated System for Analyses of Aerosol Composition and Chemistry" as an intense source of well-characterized organic aerosols. Experiments at the ISAACC will be used to validate our "inverted micelle" picture of organic aerosols.
 
 

E. Project and Management Plans

This will be a three-year project. Activities in year one will focus on procurement of equipment items that will form the core of several systems, including the mass spectrometers, turbomolecular pumps, lasers, and optics and the design of vacuum chambers, steering optics, and mechanical interfaces. We will also procure and test aerosol generation and test equipment. In year two, all hardware will be fabricated, and the subsystems will be assembled and tested individually. Also in year two, preliminary experiments will begin on aerosol generation and chemistry utilizing subsystems. At the start of year three we will accept delivery of Aerodyne particle mass spectrometer and it will be incorporated into the rest of the apparatus to complete ISAACC. At this time we will initiate more sophisticated studies of the coupled chemistry and composition of particles and the characterization of laser-ablation techniques, including direct tests of the PALMS instrument.

For the duration of this grant, the main management issue will be supervision of the design, procurement, and assembly of the instrument. Some attention will also have to be paid to planning the preliminary scientific studies. Once the development and testing phase of ISAACC is complete in year three, the management focus will shift to procedures for allocating time on the apparatus. Students under the direct supervision of the co-investigators will carry out the development, assembly, and testing of ISAACC. Several of these students will be directly funded by this project, and we anticipate that several additional students will be funded out of other projects of the co-investigators. The students will also work closely with one or more of the collaborators listed in Section B.4. Depending on the nature of the work, it is even possible for the collaborators to directly supervise a student dissertation project in collaboration with one or more of the lead co-investigators. In addition, there will be a part-time (1.5 months per year) senior scientist who will provide additional oversight of this project. This part-time person is important for the continuity of this project because several of the co-investigators spend extended periods of time in the field.

The work will be conducted in Prof. Toohey's lab at CU, which is centrally located on the CU campus. Prof. Toohey will also be primarily responsible for oversight of the project. He will meet with the students on a weekly basis to monitor the progress of individual projects. Once per month there will be meetings of the larger group that will include the co-investigators and some of the collaborators to monitor the progress of the overall project. These meetings will provide frequent opportunities to make important decisions regarding the direction of the design and fabrication efforts and will also serve to promote discussions that will expose the students to the important scientific issues. The project will have its own web site (presently http://paos.colorado.edu/~toohey/mri.html) for rapid communication of regular project status reports, and important developments, meeting announcements, and agendas. Each student will be required to maintain a web page that includes a monthly summary of activities and a biennial progress report. Links will also be provided for events in the Boulder area that will be of interest to all the project participants.