Environ. Sci. Technol. 2001, 35, 2184-2190
Detection of Alkaline Ultrafine Atmospheric Particles at Bakersfield, California ALBERT CHUNG, JORN D. HERNER, AND MICHAEL J. KLEEMAN* Department of Civil and Environmental Engineering, University of California, Davis, California
Two collocated micro-orifice uniform deposit impactors (MOUDIs) and a filter-based sampler were used to measure the size distribution and chemical composition of atmospheric particulate matter at Bakersfield, CA, between January 14 and 23, 1999. The peak number concentration of airborne ultrafine particles measured was 1.45 × 1011 m-3, which is a factor of approximately 3 higher than the peak airborne ultrafine-particle number concentration measured previously in Pasadena, CA. Chemical analysis revealed that airborne ultrafine particles (Dp < 0.1 µm) at Bakersfield contained significant amounts of watersoluble species, including calcium, sodium, ammonium ion, nitrate, and sulfate. Other chemical species detected in the ultrafine size range included potassium, iron, copper, zinc, and strontium. A balance of aqueous ions showed that ultrafine particles were alkaline in nature with calcium acting as the dominant cation. Bulk samples of airborne particles with diameter less than 2.0 µm (PM2.0) were essentially neutral, but particle acidity was found to be a strong function of particle size. The results of this experiment suggest that areas deep in the human lung that preferentially collect particles in the ultrafine size range could be exposed to locally acidic or alkaline conditions even if the integrated airborne particle complex is essentially neutral.
Introduction There are currently over 150 published articles that describe epidemiological studies considering the health effects of particulate air pollution (1, 2). Although the results of several studies are not in complete agreement, the general consensus among the majority of the epidemiological evidence indicates that particulate air pollution is an important risk factor for increased cardio-pulmonary disease and mortality. Several of the theories that seek to identify a causal relationship between health effects and airborne particulate matter focus on the role of ultrafine airborne particles (those with diameter less than 0.1 µm). Ultrafine particles typically exist at very high number concentrations in the atmosphere and yet, because of their small size, they contribute a negligible amount of mass to fine airborne particle samples. There are a number of possible mechanisms by which ultrafine particles may cause adverse health effects. Because of their small size, ultrafine particles penetrate deep into the respiratory system. Oberdorster et al. (3-5) have proposed that the large number of ultrafine particles penetrating deep into the human lung * Corresponding author phone: 530-752-8386; fax: 530-752-7872; e-mail:
[email protected]. 2184
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
may overwhelm the alveolar macrophages that serve as the natural mechanism for the removal of foreign objects from the respiratory system. Oberdorster et al. (3-5) also found that insoluble ultrafine particles impact the alveolar macrophages to a much greater extent than larger particles that deposit themselves into the small passageways of the lung because ultrafine particles have a “better fit”, leading to greater surface area contact between the particle and the linings of the passageway. Another theory explaining the health effects associated with airborne particles proposes that un-neutralized strong acids contained in ultrafine particles penetrate deep into the lung causing damage (6, 7). Studies have shown that the concentration of inhaled acidic aerosol is strongly correlated to increased coughing frequency and intensification of asthmatic symptoms for those with an asthmatic history (8). Finally, it has been proposed that trace metals found on the surface of ultrafine particles catalyze a series of reactions that lead to lung damage (9, 10). The three broad categories of biological and chemical effects described above encompass a large number of possible mechanisms relating airborne particulate matter to health effects. This complexity is aggravated by the fact that atmospheric particles contain a mixture of compounds that may catalyze biological effects in unknown ways. The combination of these factors makes it extremely difficult to design laboratory experiments that will establish a causal relationship between health effects and exposure to airborne particles. To make this task more tractable the exact features of the atmospheric ultrafine particles that are associated with adverse health effects must be identified. The first step of this process is a comprehensive understanding of the abundance and composition of ultrafine particles typically found in the atmosphere. Current knowledge about the characteristics of atmospheric ultrafine particles is limited. Only one published study, conducted in Los Angeles, simultaneously reports the physical size, number concentration, and chemical composition of ambient ultrafine particles. In that study Hughes et al. (11) found that the number concentration of ultrafine particles in Los Angeles averaged over 24 h periods is consistently in the range of 1.3 × 104 particles cm-3 air. A small amount of sulfate was detected on the ultrafine particles, but the concentration of acids and bases was too low to determine whether this sulfate existed in neutralized form. The most common transition metal detected was iron. With this limited database, it is not possible to evaluate the danger posed by ultrafine particles via the health-effects mechanisms described above. In this paper the results of an experiment designed to measure the composition of airborne ultrafine particles at Bakersfield, CA, are reported. Bakersfield is a city with a population of 400,000 located in the heart of the San Joaquin valley. During winter months strong atmospheric temperature inversions and the presence of mountain ranges on three sides of the city lead to a situation where airborne pollutant concentrations build to high levels. Epidemiological evidence suggests that there is a strong increase in excess mortality associated with high concentrations of airborne particulate matter in the San Joaquin valley (12), and so Bakersfield provides an ideal location to characterize the nature of airborne ultrafine particles. In the sections below the experimental design is described and the results of the analysis of ultrafine-particle concentrations and composition are discussed. 10.1021/es001879l CCC: $20.00
2001 American Chemical Society Published on Web 05/05/2001
Experimental Methods Samples of airborne particulate matter were collected on the roof of the California Air Resources Board (CARB) office located at 5558 California Avenue, Bakersfield, CA, on seven different days between January 14 and 23, 1999. Samples were collected between the hours of 10:00 and 18:00 Pacific Standard Time (PST) using a filter-based sampler and two collocated micro-orifice uniform deposit impactors (MOUDI, MSP Corporation, model no. 110) equipped with AIHL-design cyclone separators. Two MOUDIs were used to collect samples during the study to facilitate a broad range of chemical analysis and to allow for consistency checks in airborne particle measurements (13). The first MOUDI was loaded with 47-mm foil substrates (MSP Corporation) and a 37-mm quartz fiber after-filter (Pallflex 2500 QAO). Each substrate and after-filter was prepared prior to use by baking at 550 °C for 48 h in order to eliminate any preexisting carbon on the sample collection media. Airborne particle samples collected on aluminum substrates and quartz fiber filters were analyzed for total carbon, organic carbon, and elemental carbon using the thermal optical reflectance (TOR) method described by Huntzicker et al. (14) and modified for use with impactor substrates by Birch and Cary (15). To convert organic carbon measurements to organic compound measurements, a factor of 1.4 was used (16). The second MOUDI was loaded with 47-mm Teflon substrates (Teflo, R2PJ047) and a 37-mm Teflon after-filter (Zeflour, P5PJ037). The Teflon media used to collect airborne particle samples were divided in half after collection so that two separate chemical analyses could be conducted. The first half of each Teflon substrate and filter was subjected to ion chromatography analysis to measure the concentration of water soluble ions (Na+, Ca2+, NH4+, K+, Cl-, NO3-, SO42-, and PO43-). The second half of each Teflon substrate and filter was analyzed for trace species content using protoninduced X-ray emissions (PIXE) and X-ray fluorescence (XRF). A filter-based sampler was operated alongside the collocated MOUDIs to provide measurement data for a consistency check and to investigate the correlation between ultrafine-particle concentrations and fine-particle mass concentrations. The filter-based sampler was equipped with an AIHL-design cyclone separator that removed particles greater than 1.8 µm in diameter from the sample stream (17). Because the cutoff point of this sampler is very close to the upper size range sampled by the MOUDI-cyclone combinations described above, integrated MOUDI results can be compared to filter-based measurements to identify any inconsistencies that may have arisen during sample handling or chemical analysis. In the current study, results are reported for fine-particle samples collected on a 47-mm Teflon filter (Teflo, R2PJ047) and a 47-mm quartz fiber filter (Pallflex, 2500 QAO). The airborne particle samples collected on each filter were analyzed for chemical composition using the methods described above. All collection media were stored in sterile Petri dishes sealed with Teflon tape before and after sampling events in order to reduce the possibility of sample contamination. Petri dishes used to store aluminum substrates and quartz fiber filters were lined with aluminum foil that was baked at 550 °C for 48 h to eliminate carbon on the surface. All samples were stored in a freezer at -20 °C until chemical analysis was initiated. Impactor substrates were not coated with an antibounce agent to avoid interference with chemical analysis techniques. To prevent particle bounce, an AIHL-design cyclone separator was operated upstream of each MOUDI to remove particles with diameter larger than 2 µm from the sample stream. Many of these large particles are composed of hydrophobic crustal material that may bounce off impactor collection stages. Removal of larger particles from the sample
FIGURE 1. Comparison between measurements made with collocated cascade impactors and filter-based samplers at Bakersfield, CA, January 14-23, 1999. Agreement between the different measurement devices is good but not perfect, because slightly different size ranges were sampled. Integrated impactor samplers collected particles with diameters less than 1.8 µm while filterbased samplers collected particles with diameters less than 2.0 µm. stream did not interfere with the collection of the ultrafine particles that are the focus of the current study. Aluminum and Teflon collection media were weighed 2-3 times using a CAHN-33 microbalance which had a resolution of ( 1 µg. All filters were allowed 1.5 min to equilibrate the scale oscillations before a measurement was taken. The room’s temperature and humidity were constantly monitored VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2185
FIGURE 2. Time series of fine-particle mass and ultrafine-particle number concentration at Bakersfield, CA, January 14-23, 1999. during this procedure and averaged 22° C and 25%, respectively.
Results A comparison between airborne particulate matter collected using MOUDIs and filter-based samplers is shown in Figure 1 to illustrate data consistency. Each panel of Figure 1 shows the sum of the material collected on all stages of the MOUDI as a function of the material collected on the corresponding filter sample. Minimum detection limits (MDLs) for each instrument and analysis method are shown as straight lines perpendicular to the corresponding plot axis. Figure 1a compares the mass of airborne particulate matter (PM2.0) collected with the filter-based sampler to the mass of airborne particulate matter (PM1.8) collected with the MOUDI. The linear regression line drawn through the data shown in Figure 1a has a slope of 0.812 and a correlation coefficient (R2) of 0.930. Figure 1b shows that the amount of sulfate (SO42-) collected by the filter-based sampler and the MOUDI equipped with Teflon collection media are well correlated with a regression slope of 0.813 and a correlation coefficient (R2) value of 0.995. Figure 1c compares the amount of organic carbon (OC) collected by the filter-based sampler and the MOUDI equipped with foil collection media. This regression also indicates a strong correlation for OC between the two instruments with a slope of 0.871 and a correlation coefficient (R2) value of 0.945. The regression derived correlations shown in each panel of Figure 1 are typical for all the chemical species measured in airborne particles in the current study. The temporal variation of fine-particle mass concentrations (Dp < 2.0 µm, µg m-3) and ultrafine-particle number concentrations (Dp < 0.1 µm, particles m-3) measured at Bakersfield, CA, between January 14 and 23, 1999 is illustrated in Figure 2. Ultrafine number concentrations were calculated by inverting mass measurements collected on impactor substrates. Fine-particle mass concentrations decreased during the study period from an initial value of 92 µg m-3 to a final value of 3 µg m-3. In contrast, ultrafine-particle number concentrations increased during the initial portion of the study period from an initial value of 7.14 × 1010 particles m-3 to a peak value of 1.45 × 1011 particles m-3. Ultrafine-particle number concentration then decreased to a final value of 3.5-1.1 × 1010 particles m-3 by the end of the study. This behavior demonstrates that there are cases when the ultrafine-particle number concentration does not follow the same trend as fine airborne particle mass. It is also worth 2186
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
noting that the peak number concentration of airborne ultrafine particles measured at Bakersfield, CA, is approximately 3 times larger than the concentration of ultrafine particles measured in Los Angeles by Hughes et al. (11). The size and chemical compositions of airborne particulate matter collected at Bakersfield, CA, on January 14, 1999 and January 21, 1999, which are heavily and less heavily polluted events, respectively, are shown in Figure 3. On January 14 the fine airborne particles were composed primarily of nitrate (NO3-), ammonium ion (NH4+), and organic carbon (OC) (Figure 3). The “other” fraction shown in Figure 3 consists primarily of trace metals that were above their respective MDLs, but not in sufficient concentrations to be shown on the bar plots individually. Examples of these trace metals include Mg, Al, Si, Ti, Fe, Ni, Cu, and Zn. The “unknown” fraction shown in Figure 3 consists of the difference between the particle-mass concentration determined gravimetrically and the total sum of species concentrations that were determined by chemical analysis. On January 21 the airborne particulate matter at Bakersfield was composed of sulfate (SO42-), nitrate (NO3-), sodium (Na+), and chloride (Cl-) (Figure 3). Carbonaceous species also were present in the ultrafine size mode. The focus of the current study is the characterization of airborne ultrafine particles at Bakersfield, CA. The final impaction stage of the MOUDIs collects airborne particles in the ultrafine particle range (0.056-0.1 µm), but from Figure 3 it is difficult to determine the actual chemical composition of the ultrafine particles because they make a relatively small contribution to the airborne-particle mass concentration. The chemical compositions of the ultrafine particles measured on January 14 and 21 at Bakersfield, CA, are plotted in Figure 4. Chemical analysis shows that ultrafine particles were composed of primarily organic carbon (OC) on both the heavily polluted day (upper panel) and less polluted day (lower panel). Other chemical species detected in measurable amounts included elemental carbon (EC), silicon (Si), calcium (Ca2+), titanium (Ti), iron (Fe), ammonium ion (NH4+), sulfate (SO42-), phosphate (PO43-), nitrate (NO3-), and chloride (Cl-). These chemical species are common components of the PM10 and PM2.5 airborne-particle size fraction in the Southern San Joaquin Valley (18). “Other trace metals” shown in Figure 4 consist of trace metals that were detected in quantities above their MDLs but that were not present in large enough quantities to be legibly seen on a pie chart of this resolution. Examples of these trace metals include Mg, Ni, Cu, and Zn.
FIGURE 3. Size distribution and chemical composition of airborne particulate matter collected at Bakersfield, CA, on (a) January 14, 1999 and (b) January 21, 1999. Also included in Figure 4 is the distribution of watersoluble cations and anions measured in airborne ultrafine particles. The amount of calcium (Ca2+) contained in ultrafine particles on January 14 (0.077 ( 0.057 µg m-3) is roughly equivalent to the concentration of organic carbon (0.094 ( 0.064 µg m-3) found in these particles. On January 21, the composition of the ultrafine particle fraction is more evenly distributed across many different chemical species. A large fraction of the ultrafine particle mass measured on January 21 is composed of organic carbon, but trace metals such as aluminum and the “other trace metals” also contributed strongly to the ultrafine fraction. Calcium (Ca2+) was high in concentration (0.064 ( 0.057 µg m-3) relative to other species and was close to the organic carbon concentration (0.054 ( 0.062 µg m-3). On both January 14 and January 21, ultrafine particles at Bakersfield, CA, contained more sulfur (S) than can be accounted for by sulfate (SO42-) (Figure 4). On both January 14 and 21, the amount of sulfur contained in airborne ultrafine
particles associated with sulfate (measured by IC) accounted for only half of the total ultrafine particle sulfur concentration (measured by PIXE/XRF). These findings may result from the presence of hydroxymethanesulfonate (HMS) produced by the reaction of sulfur dioxide and formaldehyde in aqueous fog droplets. Airborne particles activate to form fog droplets during periods of high relative humidity. When relative humidity decreases, fog droplets evaporate and the airborne particles are returned to their approximate original size. It is likely that ultrafine airborne particles collected in the current study were processed by fog events prior to sampling, as foggy conditions were common throughout the San Joaquin Valley during the evening hours of the study period.The identified chemical species shown in Figure 4 account for 53-100% (January 14, 1999) and 94-100% (January 21, 1999) of the ultrafine particle mass measured at Bakersfield, CA. The majority of the uncertainty in this calculation is due to limitations in the accuracy of the gravimetric analysis. VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2187
FIGURE 4. Chemical composition of ultrafine particles (0.056 < Dp < 0.1 µm) collected at Bakersfield, CA, on (a) January 14, 1999 and (b) January 21, 1999.
Analysis One of the hypotheses that may explain the health effects associated with the inhalation of ultrafine particles is that inhaled acidic ultrafine particles may cause damage to tissues deep in the lung (6, 7). The acidity of the ultrafine particles collected at Bakersfield, CA, was evaluated by converting the mass of each chemical species contained on these particles to a concentration in units of moles m-3 air and then multiplying by the chemical species charge, resulting in units of equivalent charge m-3 air. The results of this acidity balance constructed for both the fine airborne particles (Dp < 2.0 µm) and the ultrafine airborne particles (Dp < 0.1 µm) for each experiment conducted in the present study are shown in Figure 5. Figure 5a shows that the fine fractions of airborne particles collected at Bakersfield, CA, have roughly equal amounts of anions and cations for all seven experiments. Ammonium ion (NH4+) acts as the dominant cation in the fine-particle fraction for the first three measurement periods when fine-particle mass concentration was high (91.6 ( 0.7 to 50.1 ( 0.7 µg m-3). After January 16, fine-particle mass concentrations decreased sharply (10.3 ( 0.7 to 3.1( 0.7 µg m-3), and calcium (Ca2+) became the dominant cation in the fine-particle fraction. Figure 5b shows that the ultrafine particles collected during the study period at Bakersfield, CA, have approximately 25% more cations than anions during six of the seven measurement periods. Calcium (Ca2+) acts as the dominant cation in the ultrafine-particle size range throughout the entire study period. Calcium (Ca2+) and sodium (Na+) were the only ultrafine cation species present in quantities above their minimum detection limits at Bakersfield during the study period. The anions observed in the ultrafine size mode include chloride (Cl-), nitrate (NO3-), and sulfate (SO4-2). 2188
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
Phosphate (PO4-3) was not present in ultrafine-particle samples in quantities above the minimum detection limits. The distribution of anions and cations in the ultrafine-particle size range was relatively constant for all measurement periods in the present study and showed no dependency on overall mass concentration. These results indicate that airborne ultrafine particles measured at Bakersfield, CA, are primarily alkaline (rather than acidic) in nature. Little is currently known about the potential health effects associated with the inhalation of alkaline ultrafine airborne particles. Calcium is commonly detected in relatively large crustal particles found in the atmosphere. In the present study a laboratory test was conducted to determine if large airborne dust particles may have bounced off the upper stages of the cascade impactors and accumulated on the lower stages to produce the results shown in Figure 5. Particle bounce in the MOUDI is a function of ambient relative humidity (19) so this test was conducted under dry conditions as a worst-case scenario. A Lovelace Nebulizer (model 01-100) was used to produce calcium nitrate particles that were then dried by reducing the relative humidity of the carrier gas to less than 25%. The resulting aerosol was combined with 29 L per minute (L/min) of filtered air (RH, 50%) and drawn into an AIHLdesign cyclone separator and then a MOUDI loaded with Teflon substrates (configuration identical to that used in the ambient experiment). A parallel sample of the calcium nitrate aerosol was drawn into a scanning mobility particle sizer (TSI SMPS 3080) to independently verify the size distribution measured with the MOUDI. The results of this experiment indicate that large calcium nitrate particles do not bounce off upper impaction stages and become fixed to lower impaction stages.
FIGURE 5. Overall acidity of fine airborne particulate matter and ultrafine airborne particulate matter collected at Bakersfield, CA, January 14-23, 1999. Analysis indicates that fine-particle samples are approximately neutral but ultrafine-particle samples contain approximately 75% alkaline material (arrow marks transition from acidic to alkaline material). The aerosol acidity as a function of particle size at Bakersfield, CA, on January 14 and 21, 1999 is shown in Figure 6. These specific days were chosen for analysis because they represent extreme examples of the high and the low pollution events encountered during the experiment. Figure 6a illustrates the amount of excess acidity or alkalinity contained in ambient airborne particles on January 14 and 21, 1999. On January 14, 1999 airborne particles between 0.056 and 1.8 µm diameter were alkaline in nature. On January 21, 1999 airborne particles with diameter less than 0.2 µm were alkaline in nature, while particles with diameter greater than 0.2 µm were acidic. The diameter of inhaled particles determines their deposition efficiency in the human cardio-pulmonary system. In the current study, the amount of particulate matter that could be deposited deep in the pulmonary region of the lung was calculated using the deposition efficiency curves determined by the International Commission on Radiological Protection (ICRP) Task Group on Lung Dynamics with a breathing rate of 15 breaths per minute (20). Figure 6b displays the amount of alkaline or acidic material deposited in the deep lung as a function of particle size. On January 14, all of the material deposited in the pulmonary region of the lung was alkaline, with the majority of the alkaline load
delivered by particles with diameter less than 0.1 µm and particles with diameter between 0.2 and 1.8 µm. On January 21, particles deposited in the pulmonary region of the lung with diameter smaller than 0.3 µm were alkaline while particles larger than 0.3 µm diameter were acidic in nature. The deposition curves that are used to generate the results shown in Figure 6 are approximate relationships that predict the general region of the human cardio-pulmonary system where particles of different sizes will be deposited. It is expected that within each general region of the cardiopulmonary system, specific locations will exist that will preferentially collect particles of certain sizes by deposition. The results of the analysis presented in Figure 6 of the current paper demonstrate that particles deposited into sensitive areas of the lung will likely not have the same acidity as that which would be inferred from analysis of a bulk PM2.5 sample.
Acknowledgments This research was supported by the California Air Resources Board under contract number 97-536. The authors also thank the sponsors and participants of the California Regional PM10/ PM2.5 Air Quality Study for their support. Thanks are due to Lynn Salmon and Dr. Glen Cass (California Institute of VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2189
Literature Cited
FIGURE 6. Aerosol acidity as a function of particle size at Bakersfield, CA, on January 14 (heavily polluted) and 21 (less polluted), 1999. The upper panel shows the acidity of particles in the atmosphere, and the lower panel shows the acidity that would be delivered by inhaled particles to the pulmonary region of an average human lung. Technology) for carbon analysis, Dabrina Dutcher and Dr. Tom Cahill for PIXE/XRF analysis, and Dr. Eva Hardison (Research Triangle Institute) for ion chromatography analysis.
2190
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
(1) Dockery, D. W.; Pope, C. A.; Xu, X.; Spengler, J. D.; Ware, J. H.; Fay, M. E.; Ferris, B. G.; Speizer, F. E. N. Engl. J. Med. 1993, 329, 1753-1759. (2) Pope, C. A.; Dockery, D. W.; Schwartz, J. Inhalation Toxicol. 1995, 7, 1-18. (3) Oberdorster, G.; Ferin, J.; Gelein, R.; Soderholdm, S. C.; Finkelstein, J. Environ. Health Perspect. 1992, 193-199. (4) Oberdorster, G.; Ferin, J.; Lehnert, B. E. Environ. Health Perspect. 1994, 102, 173-179. (5) Oberdorster, G.; Gelein, R. M.; Ferin, J.; Weiss, B. Inhalation Toxicol. 1995, 7, 111-124. (6) Chen, L. C.; Peoples, S. M.; Amdur, M. O. Am. Ind. Hyg. Assoc. J. 1991, 52, 187-191. (7) Schlesinger, R. B. Environ. Health Perspect. 1989, 79, 121-126. (8) Chen, L. C.; Miller, P. D.; Lam, H. F.; Guty, J.; Amdur, M. O. J. Toxicol. Environ. Health 1991, 34, 337-352. (9) Ghio, A. J.; Stonehuerner, J.; Pritchard, R. J.; Piantadosi, C. A.; Quigley, D. R.; Dreher, K. L.; Costa, D. L. Inhalation Toxicol. 1996, 8, 479-494. (10) Dreher, K. L.; Jaskot, R. H.; Lehmann, J. R.; Richards, J. H.; McGee, J. K.; Ghio, A. J.; Costa, D. L. J. Toxicol. Environ. Health 1997, 50, 285-305. (11) Hughes, L. S.; Cass, G. R.; Gone, J.; Ames, M.; Olmez, I. Environ. Sci. Technol. 1998, 32, 1153-1160. (12) Cahill, T. A.; Dutcher, D.; Clark, C.; Martin, J.; McCarthy, T.; Lipnick, D. Submitted to The American Lung Association of Sacramento - Emigrant Trails; Sacramento, CA, June, 1998. (13) Marple, V. A.; Rubow, K. L.; Behm, S. M. Aerosol Sci. Technol. 1991, 14, 434-446. (14) Huntzicker, J. J.; Johnson, R. L.; Shah, J. J.; Cary, R. A. In Particulate Carbon, Atmospheric Life Cycle; Wolff, G. T., Klimisch, R. L., Eds.; Plenum: New York, 1982. (15) Birch, M. E.; Cary, R. A. Aerosol Sci. Technol. 1996, 25, 221-241. (16) Gray, H. A.; Cass, G. R.; Huntzicker, J. J.; Heyerdahl, E.; Rau, J. A. Sci. Total Environ. 1984, 36, 17-25. (17) John, W.; Resischl, G. J. Air Pollut.Control Assoc. 1980, 30, 872876. (18) Chow, J. C.; Watson, J. G.; Zhiquiang, L.; Lowenthal, D. H.; Frazer, C. A.; Solomon, P. A.; Thuillier, R. H.; Magliano, K. Atmos. Environ. 1996, 30, 2079-2112. (19) Zhang, X. Q.; McMurry, P. H.; Hering, S. V.; Casuccio, G. S. Atmos. Environ. Part A 1993, 27, 1593-1607. (20) Raabe, O. G.; Howard, R. S.; Cross, C. E. Aerosol considerations in asthma. In Bronchial Asthma; Gershwin, M. E., Ed.; Grune & Stratton, Inc.: Orlando, FL, 1986; Chapter 22, pp 495-514.
Received for review November 14, 2000. Revised manuscript received March 19, 2001. Accepted March 19, 2001. ES001879L
