University of Southern California Dissertations and Theses

Development of ambient particulate concentration technology using condensational growth/virtual impaction

(USC Thesis Other)

Development of ambient particulate concentration technology using condensational growth/virtual impaction

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content DEVELOPMENT OF AMBIENT PARTICULATE CONCENTRATION TECHNOLOGY USING CONDENSATIONAL GROWTH/VIRTUAL IMPACTION by Seongheon Kim A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSTIY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (ENVIRONMENTAL ENGINEERING) August 2000 Copyright 2000 Seongheon Kim Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3093425 UMI UMI Microform 3093425 Copyright 2003 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES. CALIFORNIA 90007 This dissertation, written by Se<?n<|h£0h Kim under the direction of h Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY Date Dean of Graduate Studies 1 - IV %0vO DISSERTATION COMMITTEE s / ) c c t 'n / V Chairperson Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEGEMENTS There are many people, who have contributed not only to my doctoral research but also to my personal growth as a scientist and a human being I want to acknowledge. First of all, I would like to acknowledge my advisor and friend, Dr. Constantinos Sioutas. He was the driving force behind my doctoral research both by formulating the objectives and by offering creative suggestions through the course of the three years in my doctoral work. We shared the most exciting as well as the most disappointing moments of my doctoral research. I would also like to thank the members of my doctoral research committee, Dr. Ronald Henry and Dr. Henry Gong. Dr. Henry always inspired me during his wonderful lectures and read every line of my proposal and thesis. I want to thank also Dr. Gong for his encouragement in every stage of my research. It has been a great pleasure to work with him in various research projects related to my research. I want to acknowledge the United States Environmental Protection Agency, Health and Environment Institute, and California Air Resources Board for providing the necessary funding that was needed for this research. Their tough quality assurance raised the quality of my work to the much higher level. I am sincerely grateful to my family for their enormous support and understanding. My parents and my wife’s parents did not save any effort to take care of my children during my stay in United States. I owe my lovely two sons, Joohwan and Joosung many things in several ways. Finally, I want to thank my wife, Youngsook, for being the light o f my life. She had to sacrifice her valuable three years just to encourage and support me. ii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents Acknowledgements--------------------------------------------------------------------------------------- ii List of Tables----------------------------------------------------------------------------------------------vi List of Figures---------------------------------------------------------------------------------------------ix Abstract--------------------------------------------------------------------------------------------------- xiii Chapter I INTRODUCTION AND THESIS OVERVIEW 1-1 Background-----------------------------------------------------------------------------------1 1-2 Rationale for developing ambient particle concentrators ---------------------------- 8 1-3 Theoretical background and Design considerations---------------------------------- 10 I-4 Thesis overview----------------------------------------------------------------------------15 References---------------------------------------------------------------------------------------18 Chapter II FACTORS AFFECTING THE STABILITY OF THE PERFORMANCE OF AMBIENT FINE PARTICLE CONCENTRATORS II-1 Abstract------------------------------------------------------------------------------------ 22 n-2 Introduction------------------------------------------------------------------------------- 23 II-3 Methods------------------------------------------------------------------------------------25 II-4 Results and Discussion------------------------------------------------------------------ 32 n-5 Summary and Conclusions-------------------------------------------------------------- 43 References-------------------------------------------------------------------------------------- 44 iii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter III DEVELOPMENT AND EVALUATION OF A PROTOTYPE ULTRA-FINE PARTICLE CONCENTRATOR III-l Abstract------------------------------------------------------------------------------------48 III-2 Introduction--------------------------------------------------------------------------------49 III-3 Methods-----------------------------------------------------------------------------------53 II-4 Results and Discussion --------------------------------------------------------------- 62 III-5 Summary and Conclusions------------------------------------------------------------ 75 References---------------------------------------------------------------------------------------76 Chapter IV A NEW GENERATION OF PORTABLE COARSE, FINE AND ULTRA-FINE PARTICLE CONCENTRATORS FOR USE IN INHALATION TOXICOLOGY IV-1 Abstract-----------------------------------------------------------------------------------82 IV-2 Introduction------------------------------------------------------------------------------83 IV-3 Methods----------------------------------------------------------------------------------- 86 IV-4 Results and Discussion---------------------------------------------------------------- 96 IV-5 Conclusions----------------------------------------------------------------------------- 104 References ------------------------------------------------------------------------------------105 Chapter V VERSATILE AEROSOL CONCENTRATION ENRICHMENT SYSTEM (VACES) FOR SIMULTANEOUS IN VIVO AND IN VITRO EVALUATION OF TEXIC EFFECXTS OF ULTRAFINE, FINE AND COARSE AMBIENT PARTICLES V-l Abstract-----------------------------------------------------------------------------------109 V-2 Introduction-------------------------------------------------------------------------------110 iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V-3 Methods ----------------------------------------------------------------------------------113 V-4Results and Discussion ----------------------------------------------------------------125 V-5 Summary and Conclusions------------------------------------------------------------140 References------------------------------------------------------------------------------------- 142 Chapter VI PRELIMINARY CONCLUSIONS AND FUTURE RESEARCH VI-1 Conclusions----------------------------------------------------------------------------- 146 VI-2 Suggestions for future research -----------------------------------------------------150 References------------------------------------------------------------------------------------- 152 BIBLIOGRAPHY________________________________________________________ 153 v Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Tables Table 1-1 United States National Ambient Air Quality Standards for Particulate M atter.--------------------------------------------------------------------------------- 5 Table II-1 Summarized parameters of ambient air during the entire sampling period.-------------------------------------------------------------------------------- 28 Table III-l Comparison of design and operating parameters of this work with a previously published condensational growth system--------------------56 Table III-2 Characteristics of the ultrafine aerosols used in the experiments.--------61 Table III-3 Summary of the different temperature settings tried for the evaluation of the ultrafine particle concentrator. Data for vapor pressures and concentrations at different temperatures are from McQuiston and Parker (1982).---------------------------------------------------------------------- 62 Table III-4 Concentration enrichment achieved by the ultrafine particle concentrator as a function of particle type and minor flow rate. The temperatures of the aerosol upstream and downstream of the condenser are 35 and 25 °C, respectively.-------------------------------------67 Table III-5 Mass concentrations of ultrafine ammonium nitrate particles measured upstream and downstream of the ultrafine particle concentrator. Total flow: 110 LPM ; minor flow rate: LPM. The temperatures of the aerosol upstream and downstream of the condenser are 35 and 25 °C, respectively-------------------------------------69 vi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table IV-1 Table IV-2 Table IV-3 Table IV-4 Table V-l Table V-2 Table V-3 Table V-4 Table V-5 Table V-6 Comparisons Between the Mass Concentrations Determined using the F+UFPC and the HEADS or MOUDI using Indoor Air as the test Aerosol.------------------------------------------------------------------------------- 99 Comparisons between Sulfate Concentrations Determined Using the F+UFPC and the HEADS using Indoor Air as the test Aerosol. 100 Comparisons between Nitrate Concentrations Determined Using the F+UFPC and the HEADS using Indoor Air as the test Aerosol. 101 Comparisons between Elemental and Organic Carbon Concentrations Determined by Means of the F+UFPC and the MOUDI using Indoor Air as the test Aerosol.---------------------------------------------------------- 102 Comparison of ultrafine mass concentration after multi-slit impactor and MOUDI----------------------------------------------------------------------- 128 Coarse Ambient Particle Mass Concentrations Determined with the MOUDI and the VACES.------------------------------------------------------- 132 Coarse Ambient Particle Sulfate Concentrations Determined with the MOUDI and the VACES-------------------------------------------------------- 133 Coarse Ambient Particle Nitrate Concentrations Determined with the MOUDI and the VACES-------------------------------------------------------- 133 PM2.5 Ambient Particle Mass Concentrations (uncorrected and corrected for nitrate losses) Determined with the MOUDI and the VACES------------------------------------------------------------------------------ 135 PM2.5 Ambient Particle Nitrate Concentrations Determined with the MOUDI and the VACES-------------------------------------------------------- 136 vii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table V-7 PM2.5 Ambient Particle Sulfate Concentrations Determined with the MOUDI and the VACES.------------------------------------------------------- 137 Table V-8 Ultrafine PM Number Concentrations Upstream and Downstream of The 0.2 (rm Cutpoint Impactor and Downstream of the Ultrafine Concentrator of the VACES.----------------------------------------------------139 viii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Figures Figure II-1 Schematic of Two-stage Harvard Ambient Fine Particle Concentrator and single-person exposure chamber.-------------------------------------------26 Figure H-2 Typical Increase in Pressure Drop during the operation of Concentrator. Values in parentheses correspond to mass concentrations and dew point temperature.-------------------------------------------------------------------------- 30 Figure II-3a Dependence of aerosol Mass Median Diameter on Relative Humidity.—33 Figure II-3b Dependence of aerosol Mass Median Diameter on Dew Point Temperature.-------------------------------------------------------------------------- 33 Figure II-4 Mass Median Diameter (MMD) as a function of ambient PM2.5 mass Concentration.------------------------------------------------------------------------ 34 Figure LI-5 Concentration Enrichment as a function of Particle Mass Concentration.— 35 Figure H-6 Normalized Hourly Pressure Drop as a Function of Ambient PM2.5 Mass Concentration.-------------------------------------------------------------------------36 Figure D-7 Schematic of a virtual impactor.------------------------------------------------------38 Figure 1 1 - 8 Dependence of hourly Pressure Drop on Dew Point Temperature.-------------40 F ig u r e D - 9 D e p e n d e n c e o f A m b ie n t PM 2.5 M a s s C o n c e n tr a tio n o n D e w P o in t Temperature--------------------------------------------------------------------------- 40 Figure 11- 10 Time-averaged Increase in the Concentrator Pressure Drop as a Function of the product, Cm x (P^ - Ps > 2) .---------------------------------------------------41 IX Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure II-11 Concentration Enrichment as a Function of the Time-Averaged Increased in the Concentrator Pressure Drop.---------------------------------- 43 Figure III-l Schematic of the prototype ultrafine particle Concentrator and the experimental set-up for its characterization.------------------------------------ 55 Figure III-2 Experimental set-up for the characterization of the virtual impactor......... 58 Figure HI-3 Particle collection efficiency and losses of the virtual impactor. Total Row: 110 LPM; minor flow rate: 7 LPM .-------------------------------63 Figure ni-4 Particle collection efficiency and losses of the virtual impactor. Total Row: 106.5 LPM; Minor Row: 3.5 LPM. ------------------------------ 63 Figure HI-5 Concentration enrichment achieved by the virtual impactor at different two minor flow tested (3.5 and 71 min'1 ) . ---------------------------------------65 Figure III-6 Ultrafine particle concentration enrichment at saturation temperature of 35°C (Ts) and cooling temperature of 25 °C (Tc) for different types of ultrafine particles at two different minor flow rates (3.5 and 71 min'1 ). - 68 Figure IH-7 Ultrafine particle concentration enrichment at saturation temperature of 20°C (Ts) and cooling temperature of 11°C (Tc) for different types of ultrafine particles at two different minor flow rates (3.5 and 71 min'1 ). - 70 Figure HI-8 Ultrafine particle concentration enrichment at saturation temperature of 25°C (Ts) and cooling temperature of 15°C (Tc ) for different types of ultrafine particles at two different minor flow rates (3.5 and 71 min'1 ). - 72 x Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure III-9 Ultrafine particle concentration enrichment at saturation temperature of 30°C (Ts) and cooling temperature of 21°C (Tc) for different types of ultrafine particles at two different minor flow rates (3.5 and 71 m in1 ). —74 Figure III-10 Ultrafine particle concentration enrichment at saturation temperature of 40°C (Ts) and cooling temperature of 31°C (Tc) for different types of ultrafine particles at two different minor flow rates (3.5 and 7 1 min'1 ). —74 Figure IV-la Single-Nozzle Coarse Particle Concentrator.----------------------------------- 88 Figure IV-lb Coarse Particle Concentrator; Concentration enrichment as a function of particle size.----------------------------------------------------------------------------88 Figure IV-2 Schematic of the fine plus ultrafine Particle Concentrator (F+UFPC) and the test apparatus used for its characterization.-------------------------------- 90 Figure IV-3a Performance of various Diffusion Drier materials at 6 LPM .--------------- 92 Figure IV-3b Performance of Diffusion Drier with Silica gel at different minor flow rates-------------------------------------------------------------------------------------- 92 Figure IV-4 Characterization of the Fine + Ultrafine PM Concentrator Using Monodisperse Aerosols for 3 Minor Flows.-------------------------------------97 Figure V -la Versatile Aerosol Concentration Enrichment system (VACES) for in vivo studies: VI: virtual impactors.------------------------------------------- 115 Figure V -lb Versatile Aerosol Concentration Enrichment system (VACES) for in vitro studies.--------------------------------------------------- ---------- -------- 116 Figure V-2 Pressure drop across the multi-slit impactor ad a function of flow rate-126 XI Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure V-3 Figure V-4 Figure V-5 Figure V-6 Figure V-7 Removal efficiency of multi-slit impactor as a function of particle diameter.------------------------------------------------------------------------------- 126 Pressure drop across the BioSampler nozzle as a function of flow ra te -129 Particle collection efficiency of BioSampler as a function of particle aerodynamic diameter. Sampling flow rate: 5 LPM .------------------------ 130 Characterization of the versatile Ambient Particle Concentrator for three minor flows. Total intake flow: 220 LPM .-------------------------------------131 Size distribution of ambient aerosols before and after the Versatile Concentration Enrichment System measured by SMPS.-------------------- 140 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT Findings from the epidemiological survey associating ambient particulate pollution with adverse health effects in humans have raised the need for hypothesis- driven studies to investigate regarding physiochemical properties of particles and underlying pathophysiological mechanisms responsible for adverse effects associated with ambient particulate matter. Although recent development of ambient Particle Concentrator has made it possible to perform laboratory exposures with "real-life" ambient aerosols, conventional particle concentration technologies showed significant technical difficulties in operation, in addition to their inability to effectively concentrate PM outside the 0.35-2 pm range. A novel technology capable of concentrating ambient coarse, fine and ultrafine particles (i.e., the entire size range of PMi0 ) was developed and evaluated by means of condensational particle growth followed by separation of grown particles through a virtual impactor. Operational parameters have been experimentally optimized and the feasibility of using such a technology to provide ambient aerosols at high concentrations (up to 30 times the ambient levels) for conducting inhalation exposure health study has been demonstrated. This ambient particle concentrating technology was further extended to develop a versatile aerosol concentration enrichment system for simultaneous in vivo and in vitro evaluation of toxic effects of ultrafine, fine and coarse ambient particles. The ambient particle concentrators developed in this work have been proven to be a promising technology for conducting in vivo and in vitro health studies, which will test specific hypotheses on the relative toxicity of ambient particle size and composition characteristics and will hopefully elucidate the currently unknown mechanisms that link particulate matter to health effects. xiii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter I in t r o d u c t io n a n d t h e s is o v e r v ie w I-l Background l- l-1 Characteristics o f Ambient Particulate Matter Particulate matter (PM) is the general term used for a mixture of solid particles and liquid droplets suspended in the air. Some particles are large or dark enough to be seen as soot or smoke. Others are so small that they can be detected only with an electron microscope. These particles, which exist in a wide range of sizes, originate from many different stationary and mobile sources as well as from natural sources. Some particles are emitted directly from their sources, such as smokestacks and cars. In other cases, gases such as sulfur oxide and S 02 , NOx , and volatile organic carbons (VOC) interact with other compounds in the air to form fine particles. The chemical and physical composition of particles vary depending on location, time of year, and weather. Particle size is the most important parameter for characterizing the behavior of particles. All basic properties and behavior of particulate matter depend on particle size, including formation mechanisms, lifetime in atmosphere and site of deposition in respiratory tract. Ambient particulate matter is classified into three size modes: coarse particle mode, centered around 10-20 pm diameter, fine particle mode, centered around 0.2-0.8 pm in diameter, and ultrafine particle mode, ranging from 0.01-0.1 pm (Hinds, 1999). Each mode has distinctively different sources, chemical composition, and lifetime (Whitby et al., 1972). Particles in coarse mode are generally emitted from mechanical processes, such as vehicles traveling on unpaved roads, materials handling, and crushing and grinding 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. operations, as well as windblown dust. They are relatively large, and settle out of the atmosphere within hours. The fine particle mode is better known as accumulation mode. These particles are generated through gas-to particle conversion mechanisms including homogeneous and heterogeneous nucleation, and by condensation onto pre-existing particles in the accumulation mode. The major chemical constituents of fine particles are sulfate, nitrate, organic and elemental carbons, as well as variety of trace metals formed in combustion process. Particles in this mode usually account for most of the aerosol surface area and substantial portion of the aerosol total mass (Seinfeld and Pandis, 1998). Because they are too small to settle out, particles of this accumulation mode have lifetimes in the atmosphere on the order of days (Hinds, 1999) and can be transported over long distances. Atmospheric particles smaller than approximately 0.1 pm in diameter are known as ultrafine particles, or Aitken nuclei. They arise from gas-to-particle conversion and combustion processes in which hot, supersaturated vapors are formed and subsequently condense at ambient temperatures (Whitby and Svendrup, 1980; Finlayson-Pitts and Pitts, 1986). Particles of this size range contain the majority of the total number of ambient particles, but very little of the total mass. The lifetime of ultrafine particles in the atmosphere is short due to their rapid coagulation to form particles larger than 0.1 pm (accumulation mode). 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1-1-2 Properties o f particulate matter responsible fo r health effects PM has been classified as one of the six principal pollutants (or criteria pollutants) and it is monitored continuously by the U.S. EPA, as well as national, state and local organizations. Air quality standards for PM were first expressed in terms of mass concentration of Total Suspended Particles (TSP). This standard was subsequently replaced by PM 10 (particulate matter with an aerodynamic diameter less than or equal to 10 Jim ) because PM10 is a measurement more relevant to respiratory health, since particles larger than 10 pm are either not inhalable or deposit in the nasopharyngeal area of the respiratory tract. Since that time, many important new studies have been published, showing that the smaller (or fine) particles are largely responsible for the significant health effects including premature death, hospital admissions, and an increase in respiratory and heart diseases. Ozkaynak and Thurston (1987) used 1980 US vital statistics and available ambient pollution data for sulfate and fine, coarse and total suspended particles, and effectively demonstrated the importance of including size, chemical composition, and source information in modeling health effects. Fine particles and sulfate concentrations were consistently and significantly associated with increased mortality rates. On the other hand, particle mass measurements that included coarse particles (inhalable or total suspended mass) were often found to be nonsignificant predictors of total mortality. Scwartz and Dockery (1992) and Pope et al. (1992) analyzed associations between daily pollution levels and mortality in several cities in the US. Increased daily mortality with increased particle concentration was demonstrated. Kinney and Ozkaynak (1991) analyzed data on daily mortality and found a strong association between mortality rates and exposures to fine particles. Similar associations, 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. however, were not found for aerosol acidity, sulfur dioxide, ozone, or nitrogen dioxide concentrations. Accordingly, the U.S. Environmental Agency (EPA) revised the National Ambient Air Quality Standards (NAAQS) for Particulate Matter (EPA, 1997). New annual and 24-hour PM2.5 standards (regulating particulate matter with an aerodynamic diameter less than or equal to 2.5 |im) were established, while retaining the annual and revising the form of the 24-hour PM10 standards (Table 1-1). There are still concerns about the sufficiency of the NAAQS for particulate matter considering the complexity of PM, especially in terms of regulated species and size mode. Mass concentration alone, the current measurement for particulate matter, may not be a sufficient criterion for standards designed to protect public health since particle toxicity should also depend on chemical composition. The specific physicochemical properties of ambient PM that may be responsible for observed health effects are currently not known. There are, however, numerous candidate hypotheses about such characteristics and the mechanisms that result, either directly or indirectly, in adverse effects of particulate matter. Sulfate is one of the major chemical components of particulate matter known to be consistently and significantly associated with increased mortality rates (Ozkaynak and Thurston, 1987). Nitrate, another hygroscopic component, is being considered as potentially toxic constituent of PM (Gong et al, 2000). Ammonium nitrate levels have increased over the past 2 0 years, due to the reduction of sulfur emissions. Ambient PM contains numerous organic compounds. Among the known organic compounds present in the atmosphere, combustion-generated particles containing polycyclic aromatic hydrocarbons (PAHs) are of particular concern due to their capacity to cause adverse health effects (Atkinson and Arey, 1994; Miguel et al., 1990). 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1-1 United States National Ambient Air Quality Standards for Particulate Matter. Pollutant Previous Primary Standard Revised Primary Standard PMio Annual 50 pg/m3 The arithmetic average of the 24-hour samples for a period of 1 year, averaged over 3 consecutive years. Annual 50 pg/m3 Same as previous standard. 24- hour 150 pg/m3 The concentration of samples taken for 24- hour periods at each monitor within an area. 24- hour 150 pg/m3 The 99th percentile* of the distribution of the 24-hour concentrations for a period of 1 year, averaged over 3 years. PM2.5 No previous standard. Annual 15 pg/m3 The 3-year average of the annual arithmetic mean of the 24-hour concentrations from population oriented monitors1 * . No previous standard. 24- hour 65 pg/m3 The 99th percentile of the distribution of the 24- hour concentrations for a period of 1 year, averaged over 3 years. a By using the 99th percentile concentration approach, the revised standard better accounts for the effects on public health and inherently compensates for missing data. b The focus on population-oriented monitors stems from the health information that formed the basis for the annual PM2.5 standard. This information relates area-wide health statistics to area-wide air quality as measured by one or more monitors. Biological responses to certain compounds in particulate matters are sometimes complicated by the co-existence of pollutant gases. An experimental study (Madden et 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. al., 1999) found that toxicity of diesel exhaust particles (DEP) was enhanced by prior exposure of the particles to ozone suggesting interactive effect with gaseous co pollutants. Another recent report (Linn et al., 2000) has shown that both cardiac and pulmonary diseases increase with ambient PM1 0 , but increase more consistently with ambient CO or N 0 2 , both of which might be a surrogate of vehicular emissions. Size is another important characteristics of particulate matter, together with chemical composition. Ambient PM consists of various size modes, and consideration has been given to the role of particle size mode in adverse health impacts. The fine mode (e.g., PM 2. 5) is considered to have more a higher toxic potential than the coarse mode. In addition, there is a growing recent epidemiological database (Heyder et al., 1996; Peters et al., 1997) suggesting that the fraction of ultrafine particles may be of importance. Furthermore, and in part because of concern related to the ultrafine fraction, particle number concentration may be an important PM property in terms of toxicity. Recently laboratory studies showed that, for deposition of a given material in the lung, toxicity tends to increase as particle size decreases (Donaldson et al., 1998; Ferin et al., 1992). Furthermore, Finch et al. (1999) showed evidence for significant and rapid translocation of inhaled ultrafine particles from lung to liver. This is reasonable because finer particles penetrate more readily into cells and through tissue barriers. Furthermore finer particles have greater surface area per unit mass and toxic reactions presumably occur at the surface. Finally, finer particles dissolve more rapidly in the lungs than do larger particles, thus enhancing the bioavailabilty of desolved agents. Although the situation is further complicated by the difference in chemical composition of various size modes, it is important to relate specific particle size fractions to health effects to distinguish the contribution of the physical properties of size from chemical components. 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Elemental carbon is often used as a marker for soot in ambient PM, and mostly exists in the ultrafine mode. However, the major concern is for the health effects of soot, which is comprised of an elemental carbon matrix with adsorbed organic compounds (Hiura et al., 1999). Oxygenated PAHs with high molecular weight (>248) are associated primarily with fine particles (< 2 pm), while compounds with low molecular weight are quite evenly distributed between coarse and fine PM modes, depending on location and season (Allen et al., 1997). Another study conducted in the Caldecott Tunnel (Miguel et al., 1998) showed that diesel emissions constitute the major source of lighter PAHs, compared to light-duty gasoline vehicles that emit higher molecular weight PAHs. Detailed size-resolved measurements showed that a significant fraction of diesel-emitted PAHs is present in both the ultrafine (<0.12pm) and the fine mode (0.12-2pm). In contrast, gasoline-engine particulate PAHs were found almost entirely in the ultrafine mode. Consequently, toxicological studies focusing the adverse health effects of vehicle exhaust should be accompanied by physicochemical characterization of particles in both fine and ultrafine modes. Analysis of the results from the Harvard six cites study (Schwartz et al., 1996) suggested that coarse particles were not a predictor of daily mortality, in contrast to fine particles. Newer studies conducted in Mexico City (Loomis et al., 1999), however, suggested stronger effects of coarse mode than of fine particles on mortality. These findings indicate that coarse mode cannot be discarded as predictor of daily mortality in certain areas, and more work should be devoted to find the noxious species and underlying mechanism responsible for the adverse health effect of those coarse particles. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1-2 Rationale for developing ambient particle concentrators 1-2-1 Previous attempts to concentrate particles fo r the inhalation health study Despite the abundant epidemiological evidences associating ambient particulate pollution with adverse health effects in humans, there are still uncertainties regarding physiochemical properties of particles that affect health risk and underlying pathophysiological mechanisms. In many cases, this apparent lack of agreement between epidemiological and toxicological studies has been attributed to the inability of past controlled laboratory investigations to deliver a sufficient dose of actual ambient aerosols, which might support the epidemiological findings. Ambient particle concentrations are usually too low to introduce measurable acute effects, while artificially generated particles cannot represent all the potentially toxic components present in ambient particles in the form of either particulate or adsorbed gases. The recent development of ambient Particle Concentrators (Sioutas et al. 1995; Sioutas et al., 1997; Gordon et al., 1999) has made it possible to perform laboratory exposures with "real-life" ambient aerosols at increased (but still realistic) concentrations (Gong et al., 2000). Initial results suggest greatly increased toxic responses (as compared to that with artificial particles) and suggest that this type of exposure system may provide a useful method for assessing the health effects of ambient particles and for identifying specific risk factors and the means of controlling those factors (Godleski et al., 1996). Nevertheless, the currently available concentrators focus mainly on concentrating the accumulation mode of ambient PM (e.g., PM2.5 without its ultrafine component). The concentration enrichment of the Harvard Fine Particle Concentrator depends on particle size, with large particles of the accumulation mode being concentrated more effectively 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. than smaller particles (Sioutas et al., 1997). Consequently, this system cannot be used to increase the concentration of particles below 0.15 pm (i.e., the ultrafine particles). Moreover, these concentrators are bulky, hence not easily transportable, and the performance is often not reliable under certain meteorological circumstances. Another type of concentrator using centrifugal forces achieves restricted concentration enrichment due to coarse particle loss by impaction and diffusional loss of ultrafine particles (Gordon et al., 1999). 1-2-2 Objective o f the research The main goal of this thesis was to develop a new and improved series of ambient particle concentration technologies, the major application of which will be to conduct inhalation exposure studies as well as in vitro studies to real-life particles of all size modes. The research presented in this thesis primarily investigates the possibility of particle concentration using condensational particle growth followed by separation through virtual impaction. Coarse, fine plus ultrafine, and ultrafine ambient particles are concentrated to levels that are sufficient to make inhalation exposure studies feasible. In addition, operating these particle concentrator in conjunction with devices collecting particles by impingement on a liquid surface (such as the SKC BioSampler) can provide a highly concentrated liquid suspension/solution that can be readily used for in vitro studies. The goal of this study has been accomplished through experimental approaches to develop the following systems: a) coarse particle concentrator, which increases the concentration of ambient coarse particles in a size range from 2.5-10|im by a factor up to 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 and supplies them at least at 20 LPM to a whole body exposure chamber; b) fine/ultrafine particle concentrator, which increases the concentration of ambient both ultrafine and coarse particles by a factor up to 20 and supplies them at least at 20 LPM to a whole body exposure chamber; c) ultrafine particle concentrator, which increases the concentration of ultrafine PM only; d) a versatile aerosol concentration enrichment system capable of concentrating distinctive sub-mode of ambient aerosols and of providing them either as suspended in air or collected in water. This system was developed to concurrently concentrate ambient particles segregated into coarse, ultrafine, or fine size modes. Finally, the performance and the reliability of these systems will be continuously monitored and improved during their use in ongoing projects that constitute an integral part of the Southern California Particle Center and Supersite (SCPCS). The SCPCS is a 5-year program funded recently by the U.S. EPA to understand the sources, characteristics and health effects of PM in the Los Angeles Basin. Findings from future health studies using this particle concentrators could provide valuable scientific information to understand adverse health effects of particular matter and, will provide a sound basis for revising the standards or setting new standards for PM (U.S. EPA 2002). 1-3 Theoretical background and Design considerations The biggest limitation of conventional particle concentrators used for inhalation health studies is their limited ability to concentrate entire size range of particles, from the ultrafine to coarse size modes. Developing a particle concentrator covering three orders of magnitude in size range (i.e., 0.01 - 10 pm) is a technical challenge since no single 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. physical or chemical principle can govern the behavior of particles in this wide size range. In order to overcome this difficulty, an innovative combination of two different technologies is proposed and realized in this research. Particle concentration enrichment and separation is accomplished by virtual impaction, which can be utilized alone to concentrate coarse particles. Fine and ultrafine particles are enlarged in size by “condensational growth” under super-saturation to a size that can be inertially separated through a conventional virtual impactor. The extra liquid used to grow these particles, mostly ecologically compatible clean water, is removed by a diffusion process inside a drier so that grown particles can be returned to their original size distribution. The following sections discuss the theoretical background and important design considerations to realize the condensational growth/virtual impaction technology. 1-3-1 Stokes number and the cut point o f virtual impactor Virtual impactor is a modified real impactor by replacing the impaction plate with a collection probe. Particles with sufficient inertia are drawn into the collection nozzle carried by the minor airflow. Therefore, particles larger than the cutoff point size and those smaller than the cutoff size are separated into a minor and a major flow, respectively. The virtual impactor has the advantage of maintaining the collected particles airborne and, by adjusting the ratio of the minor to total flow, the concentration of large particles can be increased to a desired value (Barr et al, 1983). 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The principal parameter determining particle inertial separation is the Stokes number, St, defined as the ratio of the particle stopping distance at the velocity, U, versus radius of the jet, D/2 (Marple and Liu, 1974; Hinds, 1999): S t _ TU = PpCcdlU (1) D / 2 9juD where v. particle relaxation time (s) pp: particle density (g/cm3) dp: particle diameter (cm) p.: dynamic viscosity of the air (g/cm s) Cc: Cunningham slip correction factor. The slip correction factor is given by the equation (Hinds, 1999): where P is the absolute pressure in the impaction region in cm Hg and dp is the particle diameter in pm. Equation (2) emphasizes the dependence of the slip correction factor on the product Pdp. Slip correction increases as the pressure decreases, because the mean free path of the air molecules increases. C c = 1 + —- — [6.32 + 2.01 ex p (-0.1095P d P )J P d n (2) 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Equation (1) indicates that the square root of St provides a measure of the dimensionless particle diameter. The Stokes number corresponding to a 50% collection efficiency, St5 0 , depends on the ratio of the minor to the total flow and is 0.48 for a 10% minor flow (Hinds, 1999). Since minor flow is more contaminated by smaller particles at higher minor flow ratio, St50 decreases with increasing minor flow ratio (Barr et al., 1983). Equation (1) therefore can be used to predict the cutoff particle size of a virtual impactor if the velocity across the impactor’s nozzle and the size of this nozzle are known. The collection probe diameter should be 30-50% greater than nozzle diameter and the spacing between nozzle and collection probe should be kept within the range 1.2 - 1.8 times nozzle diameter to minimize particle loss (Chen and Yeh, 1987). Additional requirements are minor flow rate of 5-15% of the total flow and a very careful alignment of the axis of the nozzle with that of the collection probe (Marple and Liu, 1974). 1-3-2 Particle growth via supersaturation The growth of particles by condensation is easily initiated by the presence of small particles (as nuclei) at certain level of supersaturation.. The saturation ratio, S, is the ratio of the partial pressure of vapor, P, to the saturation vapor pressure, Ps, for the temperature of the system. For pure liquids, the smallest particle size (d*) that can be activated by supersaturation is given by the Kelvin or Tomson-Gibbs equation (Hinds, 1999) as follows: „ P . 4 a M . S - — = e x p (---------------- ) (3) Ps p R T dp Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. where M, p, and a are the molecular weight, density and surface tension of the condensing liquid, and R is the ideal gas constant. Supersaturation ratios required to activate a 0.01 pm ultrafine particle are 1.240, 1.231, and 1.223 at 20, 30, and 40 °C, respectively when water vapor is provided as the condensing medium. In a typical supersaturation process, the aerosol is first drawn though a saturator, in which it is mixed with warm water vapor, and subsequently though a condenser, in which it is cooled by few degrees. Theoretically, decrease in temperature by 4 °C is needed to achieve this level of supersaturation, therefore, the actual temperature difference between saturator and condenser should be kept larger than this value. The amount of water available to grow the particles to a large enough size that can be separated by a virtual impactor (preferably up to 2-3 pm) is another parameter that determines the operational temperatures of the saturator and condenser. This factor becomes more important, especially during pollution episodes with ambient number concentration as high as 10s -106 particles/cm3 . Numerical analysis of particle condensational growth must consider several processes such as heat and mass transfer through diffusion and convection, and include various parameters such as latent heat and diffusion coefficient of water vapor, and Kelvin effect (Hinds, 1999; Horton et al., 1991), and it thus becomes very complicated. Predicted values such as final droplet size deviate from actual data at highly humid conditions (Fang et al, 1991) or cannot accommodate wall condensational losses (Mavliev et al., 1999). The experimental approach of changing the saturator temperature while keeping the flow structure and temperature outside of the condenser unchanged is a 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. more feasible alternative that could help optimize the temperatures of saturator and condenser, with the eventual objective of maximizing particle concentration enrichment. 1-4 Thesis overview The basic objective of this thesis is to develop a fundamental understanding of the performance of prototype ambient coarse, fine and ultrafine particle concentrators and to optimize their design and operating parameters. Results form the development of these prototypes will be used to eventually to build up large scale particle concentration systems capable of providing air stream enriched with ambient particles to the whole human exposure chamber. This thesis is divided into six chapters. The first chapter is devoted to the overview of the characteristics of ambient particulate matter, particularly focusing on the their health effects observed in epidemiological studies, and establishes the rationale for developing ambient particle concentration technologies. Chapter II discusses the reliability of currently available concentrator and reveals the technical difficulties and limitations in their operational principle. The size range of existing technologies is restricted to the accumulation mode of fine particles due to the limitations of inertia separation. In addition, most of the coarse mode particles in present systems are lost via impaction inside the concentrator. Furthermore, pressure drop across current concentrators tends to increase during operation under certain meteorological conditions, such as high relative humidity and high particle mass concentration, resulting in a dramatic decrease in concentration enrichment. Pressure drop increase is explained by particle growth during adiabatic expansion and subsequent loss of grown particles by 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. impaction. Phenomenological problems of hourly pressure drop increase and decrease in concentration enrichment are statistically correlated with ambient air parameters, providing guidelines to predict the stability of the concentrator. Finally, dilution of extremely high ambient mass concentration is suggested to avoid pressure drop increase as well as to maintain the appropriate concentration level delivered to the exposure chamber. Development and experimental characterization of a prototype ultrafine particle concentrator are described in chapter HI. The study focuses on optimizing the design and the operating parameters such as the saturator and condenser temperatures as well as minor-to-total flow ratio of the virtual impactor to achieve maximum obtainable concentrations of ultrafine particles. Various types of aerosols are used to investigate the effect of particle size and chemical composition on the performance of the ultrafine concentrator. These particles include monodisperse 0.05 and 0.1 p it fluorescence PSL particles as well as polydisperse ultrafine ammonium sulfate, ammonium nitrate and indoor air particles. The condensational growth/virtual impaction system operates at a flow rate of either 106.5 or 110 LPM, of which 3.5 or 7 LPM are drawn through the minor flow of the virtual impactor. The effect of the saturator and cooler temperatures in activating and growing ultrafine particles is also investigated. Experimental results identified the saturation of ultrafine particles at 35 °C and cooling to 25 °C as the optimum temperatures for operation of the concentrator. Concentration enrichment was measured and evaluated in a variety of other conditions to verify the performance of the system. Chapter IV discusses how the prototype ultrafine particle concentrator can be extended to concentrate particles in the size range of 0.01 -10 |im. The enrichment in 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concentration is determined experimentally as a function of particle size using monodisperse as well as polydisperse aerosols. In addition, detailed chemical characterization of the ambient and concentrated aerosols was performed. Comparisons between the mass, sulfate, nitrate, and elemental and organic carbon concentrations of ambient and concentrated PM2.5 aerosols are presented and discussed in detail. The results of this chapter establish the feasibility of conducting inhalation exposure health study using this new generation of particle concentrators. The ambient particle concentrator developed in this research was further extended to a versatile aerosol concentration enrichment system (VACES) for simultaneous in vivo and in vitro evaluation of toxic effects of distinctive sub-modes of ambient particles. Chapter V describes the development and evaluation of VACES. Individual components such as a 0.2 pm low pressure drop multi slit impactor and BioSampler were first characterized in laboratory and then the performance of the entire system was validated in field tests using outdoor aerosols. PM concentrations of total mass, sulfate, and nitrate in the coarse and fine modes as well as the number concentration of ultrafine particles were enriched by factor of 19 to 23, close to the ideal value of 22. BioSamplers collected virtually identical amount of chemical species compared with those collected on the dry filters. The size distribution of ultrafine particles was not distorted during the process of condensational growth and drying by diffusion. Finally, conclusions and suggestions for future research are discussed in Chapter VI. As future research activities, utilization of the concentrator to collect concentrated ambient particles for in vitro toxicity tests and scale-up versions of the system for human exposure studies are discussed. 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References Allen, J.O., Dooleran, N.M., Taghizadeh, K., Lafleur, A.L., Smith, K.A., and Sarofim, A.F. (1997) Measurement of oxygenated polycyclic aromatic hydrocarbons associated with a size-segregated urban aerosol. Environ. Sci. Technol. 21, 2064-2070. Atkinson, A. and Arey, J. (1994) Atmospheric chemistry of gas-phase polycyclic aromatic hydrocarbons: formation of atmospheric mutagens. Environ. Health Perspec. 2, 117-126. Barr, E.B., Hoover, M.D., Knapilly, G.M., Yeh, H.C. and Rothenberg, S.J. (1983) Aerosol concentrator: design, calibration and use. Aersol Sci. Technol. 2, 437-442. Burton, R.M., Wilson, W.E., Suh, H.H. and Koutrakis, P. (1996) Spatial variation in particulate concentration within the metropolitan Philadelphia. Envir. Sci. Technol. 30, 400-407. Chen, B.T., and Yeh, H.C. (1987) An improved virtual impactor: Design and performance. J. Aerosol Sci. 18, 203-214. Donaldson, K, Li, X.Y., and MacNee, W., (1998) Ultrafine (nanometer) particle mediated lung injury, J. Aerosol Sci. 2 9 ,553-560. EPA (U.S. Environmental Protection Agency), National Ambient Air Quality Standard fo r Particulate Matter: Final Rules, FR1JL97-18, U.S. EPA, Washington DC, July 1997. Fang, C.P., McMurry, P.H., Marple, V.A. and Rubow, K.L. (1991) Effect of flow- induced relative humidity changes on size cuts for sulfuric acid droplets in the microorfice uniform deposit impactor (MOUDI). Aerosol Sci. Technol. 14, 266-277. Ferin, J., Oberdorster, G., Penney, D.P., (1992) Pulmonary retention of ultrafine and fine particles in rats. Am. J. Respir. Cell Mol. Biol. 6,535-542. Godleski, J., Sioutas, C., Katler, M., and Koutrakis, P. (1996) Death from inhalation of Concentrated Ambient Air Particles in animal models of pulmonary disease. Resp. Crit. Care Med. 155(4), A246. Gong, H. Jr., Sioutas, C., Linn, W.S., Clark, K.W., Terrell, S.L., Terrell, L.L., Anderson, K.R., Kim, S., and Chang, M. (2000) Controlled human exposures to concentrated ambient fine particles in metropolitan Los Angeles: Methodology and preliminary health-effect findings. Inhal. Toxicol. 12(S1), 107-119. Gordon T., Gerber H., Fang, C.P., and Chen L.C. (1999) A centrifugal particle concentrator for use in inhalation toxicology. Inhal. Toxicol. 11, 71-87. 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Heyder, J., Brand, P., Heinrich, J., Peters, A., Scheuh, G., Tuch, T. and Wichmann, E. (1996) Size distribution of ambient particles and its relevance to human health. Presented at the 2nd Colloquium on Particulate Air Pollution and Health, Park City, Utah, May 1-3. Hinds, W.C. (1999) Aerosol Technology. 2n d ed. John Wiley & Sons, New York. Hiura, T.S., Kaszubowski, M.P., Li, N. Nel, A.E. (1999) Chemicals in Diesel exhaust particles generate oxygen radicals and induced apoptosis in Macrophages. J. Immunol. 163(10), 5582-5591. Herring, S.V., Eldering, A.M., and Seinfeld, J.H. (1997) Bimodal character of the accumulation mode aerosol mass distributions in southern California. Atmos. Environ. 31,1-11. Hortonm K.D., Miller, R.D. and Mitchell, J.P. (1991) Characterization of a condensation- type monodispeerse aerosol generator (MAGE). J. Aerosol. Sci. 22, 347-363. Kinney, P.L., and Ozkaynak, H. (1991) Associations of daily mortality and air pollution in Los Angeles County. Envir. Res., 54, 99-120. Linn, W.S., Szachcic. Y., Gong, H Jr., Kinney, P.L., and Berhane, K.T. (2000) Air pollution and daily hospital admissions in metropolitan Los Angeles. Environ. Health Perspectives, 108,427-434. Loomis, D., Castillejos, M., Boija-Aburto, V.H., and Dockery, D.W. (1999) Stronger effects of coarse particles in Mexico city. Proceedings o f the 3rd Colloquium on Particulate Air Pollution and Human Health. Durham, North Carolina, 5-13. Mavliev, R., Wang, H., Hopke, P.K. and Lee, D. (1999) A transition from heterogeneous nucleation in the turbulent mixing CNC. 18th Annual AAAR Conference. Tacoma, WA. 331. Madden, M.C., Richards, J.H., Dailey, L.A., Hatch, and G.E., Ghio. (1999) Ozonation of Diesel exhaust particles affects lung responses. Proceedings o f the 3rd Colloquium on Particulate Air Pollution and Human Health. Durham, North Carolina, June 6-8,087. Marple, V.A. and Liu, B.Y.H. (1974) Characteristics of laminar jet impactors. Environ. Sci. Technol. 8 , 648-654. Miguel, A.G., Arey, J.M. and Sousa, J.A. (1990) Comparative study of the mutagenic and genotoxic activity associated with inhalable particulate matter in Rio de Janeiro air. Environ. M olecular Mutag. 15, 36-43. Miguel, A.H., Kirchastetter, T.W., Harley, R.A. and Hering, S.V., (1998) On-road emissions of polycyclic aromatic hydrocarbons form gasoline and diesel vehicles. Environ. Sci. Technol. 32,450. 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ozkaynak, H. and Thurston, G.D. (1987) Associations between 1980 U.S. morality rates and alternative measures of airborne particle concentration. Risk Analysis. 7,4. Peters, A., Dockery, D.W., Heinrich, J., and Wichman, GH.E. (1997) Short-term effects of particulate air pollution on respiratory morbidity in asthmatic children. Eer. Respir. J. 10, 872-879. Pinch, G.L., Nikula, K.J., Barr, E.B., Seagrave, J.C., Snipes, M.B., Hobbs,C.H., and Mauderly, J.L. (1999) Biokinetics of an ultrafine silver aerosol inhaled by rats. Proceedings o f the 3rd Colloquium on Particulate Air Pollution and Human Health. Durham, North Carolina, June 6-8 ,110. Pope, C.A., Schwartz , J. and Ransom, M.R. (1992) Daily mortality and PM10 pollution in Utah Valley. Arch. Env. Health 47, 211-217. Schwartz, J. and Dockery, D.W. (1992) Particulate air pollution and daily mortality in Steubenville, Ohio. Am. J. Epidemiol. 135,12-19. Schwartz , J., Dockery, D.W., and Neas, L.M. (1996) Is daily morality associated specifically with fine particles. J. Air & Waste Manag. Assoc. 46,927-939. Seinfeld, J.H. and Pandis, S.N. (1998) Atmospheric Chemistry and Physics. John Wiley & Sons, New York. Sioutas, C., Koutrakis, P., Ferguson, S.T. and Burton, R.M. (1995) Development and evaluation of a prototype ambient particle Concentrator for inhalation exposure studies. Inhal. Toxicol 7,633-644. Sioutas, C., Koutrakis, P., Godleski, J., Ferguson, S.T., Kim, C.S. and Burton, R.M. (1997) Harvard/EPA ambient fine particle concentrators for human and animal exposures. J Aerosol Sci. 28(6), 1057-1077. Spengler, J.D., and Thurston, G.D. (1983) Mass and elemental composition of fine and coarse particles in six U.S. cities. JAPCA 33,1162-1171. Whitby, K.T., Husar, R.B., and Liu, B.Y.H. (1972) Aerosol size distribution of Los Angeles smog. J. Colloid and Interface Science 39,177-204. Whitby, K.T. and Svendrup, G.M. (1980) California Aerosols: Their Physical and Chemical Characteristics, Adv. Environ. Sci. Technol. 10,477. 2 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter II FACTORS AFFECTING THE STABILITY OF THE PERFORMANCE OF AMBIENT FINE PARTICLE CONCENTRATORS Seongheon Kim, Constantinos Sioutas, and Ming-Chih Chang University of Southern California Department of Civil and Environmental Engineering 3620 South Vermont Avenue Los Angeles CA 90089 Henry Gong Jr, William S. Linn University of Southern California, Keck School of Medicine Department of Preventive Medicine Rancho Los Amigos Medical Center 7601 East Imperial Highway Los Angeles, CA 90242 Accepted in Inhalation Toxicology January, 2000 2 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. II-l Abstract This paper describes a systematic evaluation of factors affecting the stability of the performance of the Harvard Ambient Fine Particle Concentrators, an essential requirement for controlled exposure studies to ambient aerosols that utilize these technologies. Phenomenological problems during the operation of the concentrator, including pressure drop increase and decrease in concentration enrichment, were statistically correlated with ambient air parameters such as temperature, relative humidity, as well as PM2.5 mass concentration and mass median diameter. The normalized hourly pressure drop across the concentrator was strongly associated (R2 =0.81) with the product of ambient PM2.5 mass concentration and the difference between the vapor pressure downstream of the impactor nozzle and the saturation vapor pressure at the adiabatic expansion temperature (i.e., the temperature of the aerosol immediately downstream of the virtual impactors). From multiple regression analysis, the average enrichment factor was predicted reasonably well (R2 = 0.67) by aerosol mass median diameter and the normalized hourly pressure drop. Based on these results, we can anticipate in any given day whether an exposure study can be conducted without a considerable increase in the concentrator pressure drop, which might lead to an abrupt or premature termination of the exposure. As particle mass concentration and ambient dew point are the two main parameters responsible for raising the pressure drop across the Concentrator, efforts should be made to either desiccate the ambient aerosol at days of high dew points, or to dilute the ambient PM at days of high concentrations, prior to passing the aerosol through the virtual impactors. The latter approach is recommended at days of extremely high ambient mass concentrations because of its simplicity. 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. II-2 Introduction Although epidemiological evidence associating ambient particulate pollution with adverse health effects in humans is extensive (American Thoracic Society, 1996; Environmental Protection Agency, 1996; Pope et al., 1995), there are sill uncertainties regarding physiochemical properties of particles that affect health risk and underlying pathophysiological mechanisms (Vedal, 1997). In many cases, this apparent lack of agreement between epidemiological and toxicological studies was possibly due to the inability of past controlled laboratory investigations to deliver a sufficient dose of actual ambient aerosols, which might support the epidemiological findings. Ambient particle concentrations are usually too low to introduce measurable effects, while artificially generated particles cannot represent all the potentially toxic components present in ambient particles in the form of either particulate or adsorbed gases. The recent development of fine particle concentrators (Sioutas et al., 1995; Sioutas et al., 1997; Gordon et al., 1999) has made it possible to perform laboratory exposures with “real-life” ambient aerosols at highly increased, but still environmentally realistic concentrations. Some preliminary results using these technologies have been reported (Clarke et al., 1999; Gavett et al., 1999; Ledbetter et al., 1999; Ghio and Devlin, 1999; Urch et al., 1999), suggesting appreciable toxic responses to concentrated ambient particles exposures in laboratory rodents, and subtle responses in human volunteers. This new line of investigation may eventually lead to coherence between laboratory studies and epidemiological evidence. 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Experimental inhalation studies require a reliable exposure system for the delivery of known concentrations of aerosols to the subjects. Under certain conditions, however, the performance of concentrators becomes unstable during operation. Typical indications of instabilities in the concentrator performance are abrupt increases in pressure drop across the concentrated flow, followed by a sharp decrease in the concentration enrichment factor. These problems have been observed under conditions of high particle concentration (Gong et al., 1999: Sioutas, 1997) and/or when operating these systems in days with high humidity and temperature (personal communication with operators of these systems around the world, including Drs. Cassee at RIVM, Netherlands, Kim at the Health Effects division of the U.S. EPA, and Godleski at Harvard University). Other than these anecdotal observations, no systematic evaluation has been conducted or published on how these factors affect the performance of the concentrators yet. The purpose of this study was to investigate the effects of parameters such as ambient relative humidity (RH), dew point temperature (Tdp), ambient PM2 .s mass concentration (Cm), ambient PM2.5 mass median diameter (MMD), and total pressure drop per unit time across the Concentrator (AP/At) on the overall concentration enrichment achieved by two-stage Fine Particle Concentrator (Sioutas et al., 1995). The two-stage concentrator was installed in the environment of Rancho Los Amigos Medical Center, in south central Los Angeles. The results of this study can be generalized to similar systems operating in other locations, and can be used to predict the performance of the concentrators and to provide guidelines for operating these systems efficiently and reliably. 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. II-3 Methods II-3-1 Experimental Set-Up The ambient fine particle concentrator used in this study essentially duplicates the first and second stages of the 3-stage fine particle concentrator employed in animal studies at Harvard. Sioutas et al. (1995, 1997) have described the operating principles, design, and construction of the concentrator, which is shown schematically in Figure II-1, along with the instrumentation used for its characterization. Ambient aerosol is drawn first through a Model TE-6001 Size Selective Inlet (SSI) (Tisch Environmental Inc, Cleveland, OH), modified to exclude particles above 2.5 pm in aerodynamic diameter. The ambient aerosol is then drawn through a transition piece made of stainless steel into the first stage of the concentrator, which consists of five rectangular geometry (slit) virtual impactors in parallel. Of the 1000 1 min' 1 flow entering each slit, the minor flow of 200 1 min'1 , enriched in particles by approximately a factor of 3, passes through another transition piece to the second stage, while the major flow of 800 1 min-1 is discarded. The second stage of the concentrator consists of a single slit identical to those of the first stage. The particle-enriched minor flow from the second stage is drawn through a dilution stage, and then through a whole-body human exposure chamber, at a rate of 200 to 2501 min' 1 under a slightly negative pressure (0.98 atmosphere), while the major flow is discarded. In the dilution stage, a gate valve in the main passageway allows any portion of the concentrated-PM flow to be diverted through a parallel passage incorporating a HEPA filter, thus reducing the in-chamber PM concentration to any desired fraction of that 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5000 LPM PM-2.5 Inlet Honeycom b Denuder sam pler Quartz Filter MOUDI 4000 LPM 1000 L 800 LPM 200 LPM H oneycom b Denuder sam pler HEPA Filter Gate Valve a quartz filter MOUDI Whole-Body Human Exposure Chamber Figure II-1 Schematic of Two-stage Harvard Ambient Fine Particle Concentrator and single-person exposure chamber. 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. produced by the concentrator. Clean-air control studies are performed with the gate valve entirely closed and all inlet air passing through the HEPA filter. Two sets of Micro-Orifice Uniform Deposit Impactors (MOUDI, MSP Corporation, Minneapolis, MN), were used as depicted in Figure II-1. One set was connected to the transition zone after the size-selective inlet, just upstream of the first- stage impactors, to sample ambient PM25. Another set was located immediately downstream from the second-stage impactor, and upstream from the human exposure chamber. The MOUDIs (Marple et al., 1991) were used to determine the enrichment factor and its dependence on particle aerodynamic diameter and chemical composition. Each MOUDI sampled at 301 min-1 and classified particles in the following aerodynamic diameter intervals: <0.09, 0.09-0.15, 0.15-0.3, 0.3-0.5, 0.5-1.0, 1.0-1.8 , 1.8-3.2, and 3.2- 10.0 fim. 37-mm Teflon filters (2 |im, PTFE, Gelman, Ann Arbor, MI) were placed at the back up filter position to collect less than 0.09 |im particles; 47-mm Teflon filters were used at all other stages of the MOUDIs. Prior to characterization of the concentrator, the two MOUDIs sampled room air in collocation to confirm that they measured the same size distribution when sampling the same aerosols. The Teflon filters were pre-weighed and post-weighed to determine fine particle mass concentration by means of a MT5 Microbalance (Mettler Toledo Inc, Highstown, NJ) in a room with controlled temperature of 21-24 degrees C, and relative humidity of 40 - 50%. 24 hours was allowed for filter equilibration prior to weighing. Sampling flow rates of MOUDI samplers were measured with calibrated flowmeter (Matheson, VWR Brand, Catalogue #60106-208). The overall concentration enrichment factor was defined 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. as the ratio of total particle mass concentration of the MOUDI sampling downstream to that sampling upstream of the concentrator. Pressure drop through concentrator was recorded every 15 minutes using a sensitive Magnehelic pressure gauge (range of 0-20 inches H20; Dwyer Instruments Inc, Mich, IN). Temperature and relative humidity were also measured every 15 minutes by a thermohygrometer HI-9161F (Hanna Instruments, Italy). The sampling time for each characterization study varied from 2 to 7 hours, depending on the time required to collect enough mass at the prevailing ambient concentrations. Table H-l summarizes the meteorological parameters of ambient air during the study period, which corresponds to approximately one full year of experiments (i.e., from November 17, 1998 to December 20, 1999) in our site at south central Los Angeles. The wide range of these parameters, especially relative humidity and ambient PM2.5 levels, makes it possible to generalize the results of this study to predict the stability of similar concentrators in other areas. Table II-l Summarized parameters of ambient air during the entire sampling period a. Parameters Temperature Relative Mass concentration Mass Median Humidity (PM2 .5 ) Diameter (MMD) Range6 15 - 29 °C 12 - 80 % 6.7 -150.0 pg/m3 0.30 - 0.95 pm Average 23.8 °C 47.2 % 47.9 pg/m3 0.50 pm “Data correspond to one year in south central Los Angeles from November 17,1998 to December 20,1999. b Taken from time averaged values for 2-7 hours during daytime. 11-3-2 Theoretical Considerations The central goal of our analysis was to investigate the effects of ambient relative humidity (RH), dew point (Td p ), ambient PM25 mass concentration (Cm), ambient PM2 5 mass median diameter (MMD) and total pressure drop per unit time across the 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Concentrator (AP/At) on the overall concentration enrichment. The dew point temperature was estimated from the temperature and relative humidity data of each experiment (which were recorded every 15 minutes), and was averaged over the sampling time of each test. The MMD was estimated from the MOUDI size distribution data. The total pressure drop across the Concentrator was defined as the pressure measured downstream of the second virtual impaction stage of the Concentrator, relative to ambient pressure. Ideally, this pressure drop should remain constant during the operation of the Concentrator. However, particle deposition and accumulation primarily on the collection slit nozzle of the virtual impactors would contribute to the increase of this pressure drop with time. Figure 13- 2 shows some typical examples of how this increase in pressure drop may occur during the operation of the Concentrator. Variations in parameters such as ambient relative humidity and temperature as well as in ambient PM mass concentrations over the sampling period affect this pressure drop. We therefore avoided the conventional approach of subtracting the final from the original pressure drop values, as it could lead to substantial over- or underestimation of the actual increase in pressure drop. Instead, we calculated the time-averaged increase in pressure drop across the Concentrator as follows. The average increase in pressure drop over the entire sampling time period (At) was expressed by the following equation: where AP(TWA) represents the time-weighted average pressure drop across the \{P ,-P o )dt \ P , dt !Podt AP(TWA ) = — -------------------------- = -2----------------- -2-------------= PiTW A) - Pn ( 1 ) At A t At concentrator, pt and po are the pressure drop at time t and at in the beginning of sampling, respectively. The first term of equation (1) is the time-averaged pressure drop across the 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Concentrator, whereas the second term is the pressure drop in the beginning of the sampling period. The detailed temporal profile of the pressure drop across the concentrator, resulting from the 15-minute recording intervals, made it possible to assess Operation Time (min) accurately the time-averaged pressure drop values, AP(TWA). Figure II-2 Typical increase in pressure drop during the operation of concentrator. Values in parentheses correspond to mass concentrations and dew point temperature. Unusually high ambient concentrations and/or gradual particle accumulation on the slit nozzles of virtual impactors over many runs appear to contribute to rapid clogging. Accordingly, we adopted a policy of frequent inspection and cleaning of all virtual impactors’ acceleration and collection nozzles before each experiment in order to ensure that nozzle clogging could only be attributed to particle accumulation during that particular test. Our main hypothesis was that the two most important parameters in the performance of the concentrator, expressed parametrically by means of the concentration enrichment factor (E.F.), should be the particle mass median diameter (MMD) and the 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. time-averaged increase in the pressure drop across the concentrator (AP/At). We would expect the enrichment factor to increase with MMD and to decrease with increasing pressure drop across the Concentrator, as the latter implies partial clogging of the collection slit nozzles of the virtual impactors. The effect (if any) of the rest of the parameters on E.F. should be a surrogate of the effect of these parameters on the average PM2.5 ambient aerosol size. Some of the actual data from human exposure runs were excluded in the analysis because the concentrated air was diluted with particle-free air in these tests in order to achieve the desirable exposure concentration level of roughly 2 0 0 pg/m3. Because of the dilution, the enrichment factor values corresponding to these data were not representative of actual concentrator performance. Pressure drop in the minor flow of the concentrator increases due to the increase in the velocity across the jet of the collection nozzle that is associated with the reduction of the open surface area of the collection nozzle. The mechanism responsible for increasing the pressure drop is very similar to that observed in filtration, whereby accumulation of particulate matter over prolonged sampling increases the pressure drop across the filter. Particles impact, rather then passing though, the collection slit nozzle of the virtual impactors due to their large size. Our experimental characterization of concentrators has shown that approximately 15-25% of particles in the 1-2.5 pm range deposit on- or inside the collection slit nozzle of the impactors (Sioutas et al., 1997). Previous studies on clogging of fibrous filter (Japuntich et al., 1994) or porous membrane (Spumy et al., 1969) filter have shown that, to a first approximation, the effective reduction in the open (or available) surface of the collection slit nozzle, AA(t), due to particle accumulation over a sampling period, t, can be expressed as follows: 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Q C t f AA(t) = wL-A(t) = w L - m J ( 2 ) P , H where w and L are the width and length of the collection slit nozzle (0.05 and 28 cm, respectively), H is the depth of the collection slit nozzle (H=0.1 cm), Q is the flow rate delivered to the slit virtual impactors (Q=1000 1 min'1 ), Cm is the particle mass concentration, pp is the particle density, t is the sampling period and f is the fraction of the total particles diverted into the minor flow that deposit on the walls of the nozzle. The reduction in the open surface area of the collection nozzle will result in an increase in the average jet velocity (AU) across the nozzle, which is given by: A U = q - --------------------- 2 - ( 3 ) ( A - A A ) A where q is the minor flow rate of the collection nozzle (q=200 1 min"1 ). This increase in the jet velocity leads to a subsequent increase in the pressure drop across the minor flow of the concentrator. II-4 Results and Discussion To initiate our experimental analysis, we first examined the relationship between MMD and RH, Td p , and Cm . Figures H-3a and 3b show the relationship between MMD and RH and between MMD and Td p , respectively (although correlated with RH, Td p is also a strong function of the ambient temperature, T). Figure II-3a shows a weak association (R2 =0.24) between the MMD and RH, but indicates that, in some experiments, the MMD tends to increase from the range of 0.35-0.6 pm to the range of 0.6-0.9 pm, as RH exceeds the value of 50%. A similar association (R2 =0.25) is obtained in Figure D-3b, in 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. which the MMD is plotted against TdP . The dew point temperature is a direct measure of the water vapor content of the ambient air, and the results of Figure D-3b show some increase in MMD due to hygroscopic growth of the soluble components of ambient particles. As the relative fractions of hygroscopic ammonium nitrate and sulfate and the hydrophobic organic carbon may vary considerably in each test, the increase in MMD with Td p is not consistent, as indicated by the scatter of data points of Figure E-3b. Furthermore, hydrophobic components such as elemental and organic carbon accounted for 50-70% of ambient PM2 .s by mass, which explains the lack of a strong correlation between RH and ambient particle MMD. B •o C D s C A c S 5S 0.008x y = 0.337e R = 0.241 0 50 100 Relative Humidity (%) ^ 3 < D 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 y = 0.376e°'024x R2 = 0.246 5 10 15 20 Dew Point Temperature(°C) Figure II-3 (a)Dependence of mass median diameter on relative humidity (b)Dependence of Mass Median Diameter on Dew Point Temperature. Figure II-4 shows a plot of the MMD as a function of the PM2.5 mass concentration. It should be noted that Cm is the dry mass concentration (i.e., that determined from filters equilibrated to weighing-room temperature and humidity, without taking into account the aqueous PM components). The relationship between particle size and mass concentration is mainly dictated by the mechanism by which particles are formed. This relationship is discussed extensively by Heisler and Friedlander (1977) 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. McMurry and Wilson, (1982), and John et al. (1990). Regardless of the actual formation mechanism(s), however, particle mass concentration and average size are inherently related. A best-fit line (resulting in R2=0.55) was applied to the data of Figure II-4, resulting in the following equation associating particle concentration and MMD: MMD = -2x1 O '7 (Cm)3 + 6x1 O '5 (Cm)2 - 0.002 (Cm) + 0.4618 (4) Here MMD is expressed in pm and Cm in pg/m3. Equation (4) suggests that higher concentration enrichment values should be obtained at days when ambient particle concentration is high, since the average particle size would be larger under these conditions. As virtual impaction is based on inertial separation of particles from the carrier gas, higher enrichment factor should be expected at higher MMD. This is consistent with the results of previously published studies on the fine particle concentrators similar to that used in this study (Gong et al., 2000; Sioutas et al., 1997). ■2E-07X3 + 6E-05X2 - 0.002x+ 0.4618 R2 = 0.5488 M ass Concentration (|j,gtai3) Figure II-4 Mass Median Diameter (MMD) as a function of ambient PM2.5 mass concentration. 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The data plotted in Figure II-5, however, show an exactly opposite trend; the overall concentration enrichment decreases sharply as the ambient PM2.5 concentration increases. As it will be discussed in greater detail in the following section, this decrease in the nrichment factor is attributed to a strong association between the increase in pressure drop across the Concentrator and ambient PM2.5 concentrations, which leads to a substantial decrease in the overall enrichment with increasing mass concentration. ▲ 4 -- -- A-- 120 160 200 Ambient PM25 Mass Concentration (fj.g/m 3) Figure E-5 Concentration enrichment as a function of particle mass concentration. The values of the time-averaged increase in pressure drop across the Concentrator (expressed in units of inches of H2 0 per hour) are plotted as a function of the ambient PM2.5 concentration in Figure H-6 . Figure II-6 shows a strong association between the ambient mass concentration and the increase in pressure drop per hour across the concentrator (correlation coefficient R2 =0.61). This is due to increased particle losses on the collection slit nozzle of the virtual impactors. This phenomenon should be expected to be more pronounced at higher particle concentrations, which would lead to a faster clogging of the collection slit nozzles. The majority of particle losses in the concentrator occur inside the collection slit nozzle. This behavior has been first observed in previous studies on concentrators (Sioutas et al., 1997) and was further corroborated in 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. this study by disassembling the impactors and evaluating the sites at which particle deposit were particular noticeable. Furthermore, the increase in the minor flow pressure drop was not followed by a similar increase in the pressure drop across the major flows of the virtual impactors, which would indicate particle deposition in the acceleration slit nozzle of the impactors. 0 50 100 150 Arrbient FM 2.5 Mass Concentration (Pg/mS) Figure IE -6 Normalized hourly pressure drop as a function of ambient PM2.5 mass concentration. Equations 2 and 3 provide some explanation on the relationship between the increase in pressure drop across the concentrator and particle mass concentration. It is assumed, however, that particles remain undistorted as they pass through the acceleration and collection nozzles of the concentrator. Previous studies, however, on low-cutpoint impactors (Biswas et al., 1987; Fang et al., 1990), operating at pressure drops comparable to those across the major flow of the concentrator (e.g., 100-150 inches H20 ) have shown that flow-induced changes in the aerosol relative humidity may lead to a substantial particle growth or shrinkage. Assuming that the expansion of the air flow through the acceleration nozzle of the impactor occurs adiabatically (Hering et al., 1978), the 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. temperature immediately downstream of the acceleration jet (and upstream of the collection nozzle), T2 is given by the equation (White, 1979): p ( r - 1) T2 = Ty ( - M 7 (5) where y=1.4 (for air), T1 and PI are the temperature and pressure upstream at the virtual impaction stage (i.e., at ambient conditions) and P2 is the pressure downstream at the acceleration jet of the virtual impactor as shown in Figure II-7. For the fine particle concentrators described in this study, P2 = 0.75 atmospheres (equivalently, the pressure drop across the major flows of the virtual impactor is equal to 105 inches H20 , or 0.25 atmospheres). Equation (5) indicates that the aerosol undergoes adiabatic cooling as it expands across the virtual impactor’s acceleration jet, to a temperature (expressed in degrees K) roughly 92% of the ambient. The overall changes in the particle surface area due to either growth or shrinkage can therefore be expressed as follows (Hinds, 1982): d S p _ 8 n M W t Pv,2 ~ P S i2 dt R p p T2 where MW is the molecular weight of water (18 g/mole), R is the ideal gas constant (R= 8.3lx 107 dyn cm/K-mole), Pv ,2 and Ps > 2 are the vapor pressure downstream of the impactors nozzle and the saturation vapor pressure corresponding to the temperature downstream of the virtual impactor (T2 ), respectively. 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Assuming no vapor depletion through the virtual impactor, the vapor pressure at the exit of the acceleration nozzle (Pv > 2) is given by: n.2=V/® ./> ...=0'75P*> ^ \ where RHi is the ambient relative humidity, Ps,i is the saturation vapor pressure at ambient temperature and Pdp ,i is the vapor pressure at ambient dew point temperature, respectively. P ,= 1 atm T , i P V i 1 = R H , * P .,, i 1000 L P M Acceleration Nozzle © O © o |P2<P! :t2<t, ..iP,^=(P^0*Pv,, Major Row 0 © © Major Row 800 LPM \ Collection • Nozzle a particles larger • ™ than 50% outpoint q particles smaller ' 2( Minor flow X ) LPM than 50% outpoint Figure 13-7 Schematic of a Virtual Impactor. The above equation applies to particles roughly larger than 0.1 jxm and assumes that the temperature of the growing particles is the same with that downstream of the 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. impactor’s jet. Whether particles will grow or shrink as they become accelerated though the impactor will depend solely on the difference between Pv ,2 and PS t2 . This difference is a function of ambient relative humidity and temperature and the expansion ratio, which determined by the ratio of the pressures downstream and upstream of the impactor. In order to take into account particle growth across the virtual impactor, the right- hand term of equation (6) needs to be multiplied by the increase in particle surface area as follows: M ( ,) = w L - M O . . t - 2 £ 1 L X ‘ ( °-75 ^ ) X A,- (8) P p H R Pp T2 where At* is the aerosol residence time in the area between the acceleration and collection nozzles of the virtual impactor (which is estimated to be on the order of 20-50 ps). As equation (8) indicates, the higher the ambient dew point temperature, the higher the increase in the overall particle surface area and the higher the increase in pressure drop. The importance of the dew point temperature on the overall performance of the concentrator is illustrated best in Figure H-8 . In this figure, the time-averaged increase in pressure drop across the minor flow of the concentrator (expressed in inches of H2 0 per hour of operation) is plotted against the dew point temperature. With the exception of one outlier (a data point that corresponds to a high ambient particle concentration), when the dew point is below roughly 12-13 degrees C, the increase in pressure drop across the concentrator is relatively constant and low (i.e., about 0.3-0.4 inches of H2 0 per hour). However, as the dew point exceeds 13 degrees C, the pressure drop across the minor flows of the concentrator increases sharply by a factor of 5-10, depending on the dew point temperature. Figure II-9 shows that the ambient mass concentration and dew point 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. temperature are poorly associated (R2 =0.13), therefore both of these parameters contribute to the increase in pressure drop across the concentrator independently. te o ^ h S < C L ^ 0 •a 2 C D O 3 -S 5 / 1 d C/3 . 2 1 ~ 1 - □ □ ° Cm=120 pgta3 ■ l D □ □ D □ Dn D ----------------------------------------------- cP □ Qd □ fffib d3 5 10 15 20 Dew Point Temperature( °C) 25 Figure II-8 Dependence of hourly pressure drop on dew point temperature. 200 y = 3.1171x + 12.397 R 2 = 0.1255 e 80 - u 40 - Dew Point Temperature (°C) Figure E-9 Dependence of ambient PM2.5 mass concentration on dew point temperature. Motivated by equation (8), in which the reduction in the open area of the collection slit nozzle (hence the increase in pressure drop) depends on the product of the aerosol mass concentration times the vapor pressure difference upstream of the collection nozzle, (Pv,2 - Ps, 2), we plotted the time-averaged increase in the concentrator pressure drop as a function of the product [ Cm x (PV j2 - Ps, 2)]- Results are shown in Figure 11-10. 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 y =0.0049x + 0.1836 R2 = 0.809 -100 100 300 500 700 C oncentration x (difference in v ap o r pressures), C m x (Pv2-Ps ,2) (g g ta 3 x m m Hg) Figure 11-10 Time-averaged increase in the concentrator pressure drop as a function of the product, Cm x (Pv2-Ps,2). Figure 11-10 shows a very strong correlation (R2 = 0.81) between the product [Cm x (Pv ,2 - Ps, 2)] and the increase in hourly pressure drop per hour across the concentrator. The hourly increase in pressure drop can be predicted by the following equation: ^ = 0.1836 + 0.0049Cm (Pv2- P j2) (9) At where Cm is expressed in |fg/m3 , PV j 2 and Ps > 2 are expressed in mm Hg and AP/At is expressed in inches H2 0/hr. An important implication of equation (9) is that the overall time-averaged increase in pressure drop can be predicted with reasonable accuracy (R2 =0.81) by knowing the ambient PM25 mass concentration, relative humidity and temperature. This makes it possible to anticipate in any given day whether an exposure study can be conducted without a considerable increase in the concentrator pressure drop, which might lead to an abrupt or premature termination of the exposure. The effect of the increase in the hourly pressure drop across the concentrator on the enrichment factor is shown in Figure 11-11. Our previous analysis on the relationship 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. between the enrichment factor and ambient particle MMD has shown that, in general, particles larger than about 0.35 pm are concentrated by a factor in the range of 7-10 (Sioutas et al., 1995). Particles in the size range of 0.25-0.35 pm are concentrated less efficiently (i.e., factor of 3-5) as they are comparable in size to the cutpoint of the virtual impactors of the concentrator (i.e., 0.2 pm). Accordingly, our data were separated in two groups; those in which ambient MMD was smaller than 0.35 pm and those with an ambient MMD larger than 0.35 pm in order to disassociate the effect of particle size from that related to the increase in pressure drop on the concentration enrichment. Figure 13-11 shows clearly a steady decrease in the overall concentration enrichment as the pressure drop across the concentrator increases to values higher than about 0.7-1 inches H2 0 per hour. The decrease in the enrichment factor is primarily due to excessive particle losses due to accumulation on the collection slit nozzle of the virtual impactors after prolonged operation of the concentrator. Inspection of the collection nozzles after prolonged sampling revealed the formation of particle clusters in non-uniform distances along the slit nozzle length. The exact formation mechanism of these clusters remains unknown, but it might be related to non-uniformity in the width of the slit nozzle or in inequalities in surface roughness along the slit. In any event, such non-uniformity will lead to preferential particle deposition on these sites and the formation of small particle clusters. Once formed, these clusters may further act as deposition nuclei, in a mechanism similar to that observed in fiber filters (Brown, 1993; Japuntich, 1994). As stated previously, our main hypothesis was that the two most important parameters in the performance of the concentrator, expressed parametrically by means of the concentration enrichment factor (E.F.), should be the particle mass median diameter (MMD) and the time-averaged increase in the pressure drop across the concentrator. 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Accordingly, multiple regression between the enrichment factor, the time-averaged increase in the pressure drop across the Concentrator (AP/At) and MMD was performed in order to quantitatively assess the effect of both of these parameters on concentration enrichment. Based on this regression, the concentration enrichment factor can be predicted reasonably well (R2 =0.67) by the following equation: E.F. = 2.1 + 8.8 (MMD) - 1.41 (AP/At) (10) In the above equation, MMD is expressed in pm and AP/At is expressed in inches H2 0/hr. ambient MMD<0.35 jjtn 2 - Tim e-averaged increase in concentrator pressure drop (inches H 20 /h r) Figure II-11 Concentration enrichment as a function of the time-averaged increase in the concentrator pressure drop. II-5 Summary and Conclusions In this study we investigated the effect of ambient RH, dew point temperature, ambient PM2.5 mass concentration and MMD, and the total pressure drop per unit time across a two- stage particle concentrator on the overall particle concentration enrichment factor. As we originally hypothesized, the enrichment factor was found to increase with increasing particle MMD and decreases with increasing pressure drop across the 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concentrator. It is therefore essential that the increase in pressure drop across the concentrator be kept to a minimum in order to maximize the concentrator’s output and to guarantee a smooth operation during the exposure studies. As particle mass concentration and ambient dew point are the two main parameters responsible for raising the pressure drop across the Concentrator, efforts should be made to either desiccate the ambient aerosol at days of high dew points, or to dilute the ambient PM at days of high concentrations, prior to passing the aerosol through the virtual impactors. The former approach is non-trivial to accomplish without heating the aerosol (by using, for example, diffusion dryers), due to the high intake flow (5000 1 min'1 ) of the concentrator. The latter approach may be preferable, since at days of high ambient PM concentrations it is not essential that the output of the concentrator becomes maximized, particularly if human exposures are to be conducted (which normally require that PM exposure levels should not exceed 300 pg/m3 ). As our experimental results indicate (Figure II-8 ), it is possible to operate the concentrator even at days with relatively high ambient dew point temperatures (i.e., around 15 degrees C), as long as the ambient PM2 .s concentration does not exceed 20-30 pg/m3. Assuming that the two-stage Concentrator increases particle concentration by a factor of 7-9, the resulting exposure level would be roughly 200 (±60) pg/m3 , and thus be high enough for exposures to be conducted without any disruptions that would be caused due to the pressure drop increase. References American Thoracic Society (ATS), Committee of the Environmental and Occupational Health Assembly; Bascom, R., Chair. (1996) State of the art. Health effects of outdoor air pollution. Am. J. Respir. Crit. Care Med. 153,3-50. 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Biswas, P., Jones, C.L., and Flagan, R.C., (1987) Distortion of size distributions by condensation and evaporation in aerosol instruments. Aerosol Sci. Technol. 7, 231- 246. Brown, R.C. (1993) Aerosol Filtration: An Integrated Approach to the Theory and Applications o f Fibrous Filters, Perganunon Press, Oxford. Clarke, R.W., Catalano, P.J., Koutrakis, P., Krishna Murthy, G.G., Sioutas, C., Paulauskis, J., Coull, B., Ferguson, S., and Godleski, J.J. (1999) Urban air particulate inhalation alters pulmonary function and induces pulmonary in a rodent model of chronic bronchitis. Inhal. Toxicol. 11,101-120. Environmental Protection Agency (1996) Air Quality Criteria for Particulate Matter. EPA- 600/P-95/001af, Office of Research and Development, Washington. Fang, C.P., Marple, V.A., and Rubow, K.L., (1991) Influence of cross-flow on particle collection characteristics of multi-nozzle impactors. J. Aerosol Sci. 22(4), 403-415. Gavett, S.H., Hoyle, G.W., Madison, S.L., Walsh, L.C., Hilliard, H.G., Lappi, E.R., Evansky, P.E., and Costa, D.L. (1999) Pulmonary responses to concentrated air particles in allergen challenged hyperinnervated transgenic mice. Am. J. Respir. Crit. Care Med. 159, A29. Ghio, A.J., and Devlin, R.B. (1999) Healthy volunteers demonstrate no lung inflammation after exposure to fine particles concentrated from Chapel Hill ambient air. Am. J. Respir. Crit. Care Med. 159, A318. Gong, H. Jr., Sioutas, C., Linn, W.S., Clark, K.W., Terrell, S. L., Terrell, L.T., Anderson, K.R., Kim, S. and Chang, M.C. (2000) A pilot study of controlled human exposures to concentrated ambient fine particles in metropolitan Los Angeles. Inhal. Toxicol. 12(S1), 107-119. Gordon T., Gerber H., Fang, C.P., Chen L.C. (1999) A centrifugal particle concentrator for use in inhalation toxicology. Inhal. Toxicol. 11,71-87. Heisler, S.L. and Friedlander, S.K. (1977) Gas-to-particle conversion in photochemical smog; growth laws and mechanisms for organics. Atmos. Envir. 11,158-168. Hering, S.V., Flagan, R.C., and Friedlander, S.K. (1978) Design and evaluation of a new low pressure impactor-1. Environ. Sci. Technol. 12, 667-673. Hinds, W.C. (1999) Aerosol Technology, John Wiley & Sons Inc., New York. Japuntich, D.A., Stenhouse, J.I.T., and Liu, B.Y.H. (1994) Experimental results of solid monodisperse particle clogging of fibrous filters. J. Aerosol Sci. 25, 385-393. John, W., Wall, S.M., Ondo, J.L. and Winklmayr, W. (1990) Modes in the size distributions of atmospheric inorganic aerosol. Atmos. Environ. 22, 1627-1635. 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Koutrakis, P., Sioutas, C., Ferguson, S.T., and Wolfson, J.M. (1993) Development and evaluation of a glass honeycomb denuder/filter pack system to collect atmospheric gases and particles. Environ. Sci. Technol. 27,12:2497-2501. Ledbetter, A., Mebrane, R., Krantz, T., Jackson, M.C., Walsh, L., Hilliard, H., Richards, J., Chen, B., Costa, D.A., and Kodavanti, U.P. (1999) Variable pulmonary responses from exposure to concentrated ambient particles in a rat model of bronchitis. Am. J. Respir. Crit. Care Med. 159, A29. Marple, V.A., Rubow, K.L. and Behm, S. (1991) A Microorifice uniform deposit impactor (MOUDI): description, calibration, and use. Aerosol Sci. Technol. 14,434-446. McMurry, P. and Wilson, J. (1882) Growth laws for the formation of secondary ambient aerosols: Implications for chemical conversion mechanisms. Atmos. Environ. 16, 121- 134. Pope, C. A. Ill, Dockery, D. W., and Schwartz, J. (1995) Review of epidemiologic evidence of health effects of air pollution. Inhal. Toxicol. 7,1-18. Sioutas, C., Koutrakis, P., Ferguson, S. and Burton, R (1995) Development and evaluation of a prototype ambient particle Concentrator for exposure studies. Inhal. Toxicol. 7,633- 644. Sioutas, C., Wolfson, M., Ferguson, S.T., Ozkaynak, H. and Koutrakis, P.K. (1997) Inertial collection of fine particles using a high-volume rectangular geometry conventional impactor. J. Aerosol Sci. 6,1015-1028. Spumy K., Lodge J.P., Frank E.R., and Sheesley, D.C. (1969) Aerosol filtration by means of Nuclepore filters: structural and filtration properties. Environ. Sci. and Technol. 3,453-464. Urch, B., Liu, L., Brook, J., Purdham, J., Tarlo, S., Broder, I., Lukic, Z., Datema, J., Koutrakis, P., Sioutas, C., Ferguson, S., Dales, R., and Silverman, F. (1999) Pulmonary function responses after inhalation of controlled levels of concentrated urban particles in healthy individuals. Am. J. Respir. Crit. Care Med. 159, A318. Vedal, S. (1997) Ambient particles and health: lines that divide. J. Air Waste Manage. Assoc. 47,551-581. White, F.M. (1979) Fluid Mechanics. McGrow-Hill, Inc. New York. 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter III DEVELOPMENT AND EVALUATION OF A PROTOTYPE ULTRAFINE PARTICLE CONCENTRATOR Constantinos Sioutas, Seongheon Kim, and Mingchih Chang Department of Civil and Environmental Engineering University of Southern California 3620 South Vermont Avenue Los Angeles, CA 90898 Journal of Aerosol Science Volume 30(8), 1001-1017,1999 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. III-l Abstract This paper presents the development and experimental characterization of a prototype ultrafine particle concentrator. In this system, ultrafine particles pass over a pool of warm water where they become saturated with vapor, and subsequently drawn through a condenser, kept at a lower temperature, that allows the ultrafine particles to grow to super-micrometer size by vapor condensation on their surface. In order to increase particle concentration, the grown particles are drawn through a virtual impactor with an approximate 50% cutpoint at 1.5 pm. The concentrated particles from the minor flow of the virtual impactor finally pass through a diffusion dryer that removes the excess water on the ultrafine particles and returns them back to their original size and relative humidity. In its optimum configuration, the ultrafine concentrator operates at a sampling flow rate of 106.5 or 1101 min' 1 and concentrates the ultrafine particles to 3.5 or 71 min' 1 by an enrichment factor of approximately 15 and 25.5, respectively. Our experimental results identified saturation of the ultrafine aerosols at 35 °C and cooling to 25 °C as the optimum temperatures for operation of the ultrafine particle concentrator. Lower temperatures either do not concentrate, or concentrate less efficiently the ultrafine particles. Increasing the saturation temperature to 40 °C and cooling to 31 °C does not improve the concentration enrichment achieved by the optimum configuration. Our results also indicated that the concentration enrichment does not depend on the chemical composition of the ultrafine aerosol. Hygroscopic ammonium sulfate, volatile ammonium nitrate, hydrophobic polystyrene latex and actual “real-life” indoor air ultrafine particles were all concentrated by practically the same factor. More importantly, the experimental results show that particle concentration occurs without any coagulation, which would have distorted the size distribution of the original ultrafine aerosols. 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. III-2 Introduction There is abundant epidemiological evidence associating increased air particulate pollution and the incidence of adverse health effects on humans (Dockery et al., 1992; Schwartz and Dockery, 1992). These associations have been primarily demonstrated for fine particles (e.g., particles smaller than 2.5 Jim in size, also referred to as PM2.5) and its components such as sulfate (SO4 2 ) and strong acidity (H+ ) (Bates and Sizto 1989; Thurston et al., 1993; Dockery et al., 1994). Few investigators have recognized the importance of the complex nature of ambient particles (Amdur, 1989; Brain et al., 1976; Anderson et al., 1992). Previous studies have demonstrated acute effects of inhaled components of fine particles (0.1-2.5 |im) using artificial multicomponent mixtures (Kleinman et al., 1995 and 1997). Although these studies have made an attempt to simulate ambient particle exposures, the approach did not fully achieve a “real-life” situation. This discordance between the outcomes of laboratory and epidemiological studies indicated that artificial particles do not replicate the adverse effects of the complex and heterogeneous mixtures that occur in ambient air (Lippmann, 1989). The difficulty in establishing causal relationships between ambient fine and ultrafine particle exposures and health effects suggested that new paradigms in controlled particle exposure studies are needed to elucidate possible toxicity mechanisms of ambient particulate matter. Recently developed ambient fine particle concentrators (Sioutas et al., 1995a; Sioutas et al., 1995b; Sioutas et al., 1997) have been used in order to expose animals and 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. humans to "real-life" aerosols. These technologies focused on separating ambient particles in the fine (or accumulation) mode (0.15-2.5 pm) from the majority of the surrounding air volume in order to investigate health effects resulting from exposures to elevated (but realistic) levels of actual ambient particles. Ambient fine particle concentrations were enriched by as much as a factor of 35 prior to being supplied to animal exposure chambers. Mortality has been observed already in some of our studies exposing animals with pulmonary inflammation and chronic bronchitis to concentrations of about 300 pg/m3 of ambient fine particles in Boston (Godleski et al., 1998). Nevertheless, these systems cannot be used to increase the concentration of particles below 0.15 pm, known as ultrafine particles. Atmospheric ultrafine particles, or Aitken nuclei, arise from gas-to-particle conversion and combustion processes, in which hot, supersaturated vapors undergo condensation upon being cooled to ambient temperatures (Whitby and Svendrup, 1980; Finlayson-Pitts and Pitts, 1986). Particles in the size range 0.01 to about 0.2 pm are formed entirely during these processes. Although the mass fraction of the ultrafine mode is negligible, this size range contains the highest number of ambient particles by counts as well as surface area. Because of their increased number and surface area, ultrafine particles are particularly important in atmospheric chemistry and environmental health. To-date there has been limited epidemiological evidence linking respiratory health effects and exposures to ultrafine particles, primarily due to the lack of adequate methods for ultrafine particle measurement. Recent epidemiological studies (Heyder et al., 1996; Peters et al., 1997), however, demonstrated a stronger association between health effects and exposures to ultrafine particles compared to fine or coarse particles. 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Inhaled ultrafine particles deposit primarily in the lower respiratory tract and alveoli by diffusional mechanisms (ICRP 1994). The pulmonary toxicity of ultrafine particles has been demonstrated in several controlled laboratory exposure studies. Inhalation of fumes consisting mainly of ultrafine particles lead to the well-known effects of metal or polymer fume fever (Drinker et al., 1927; Rosenstock and Cullen, 1986; Gordon et al., 1992; Blanc et al., 1993). Oberdoster et al. (1992) have showed that apparently "inert" dusts consisting of ultrafine particles can be highly toxic to the lung because of the interstitial access across the alveolar epithelium where they interact with macrophages. Separation and concentration of any type of particles from the airstream typically is based on particle inertia. Devices such as virtual impactors have been developed for this purpose (Marple and Chien, 1980; Sioutas et al., 1994). Nevertheless, the development of virtual impactors to concentrate particles as small as 0.01 pm (e.g., “inertia-less” particles) has been a major technical challenge. Recently, hypersonic impactors operating at very low pressures (typically on the order of 500 Pa or less) have been developed to sample ultrafine particles (Fernandez de la Mora et al., 1990; Hering and Stolzenburgh, 1995; Olawoyin et al., 1995). Inertial separation of ultrafine particles from the surrounding air becomes possible due to the combined effect of very high velocities achieved in the impactor due to pressure reduction. Use of hypersonic virtual impactors to concentrate ultrafine particles, however, can be problematic for the following reasons; a) the separated (and concentrated) particles are at a very low pressure, which makes it impossible to conduct inhalation exposures. Any volatile constituents of the ultrafine aerosol will be lost at low pressure. A mechanism that transports the concentrated ultrafine particles from a low pressure to atmospheric 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pressure (i.e., a blower) may result in excessive particle losses, and; b) high sampling flow rates (500 1 min'1 , or higher) are practically impossible because they require operation in parallel of tens of thousands of such impactors. An alternative method to inertially separate ultrafine particles from the majority of the surrounding air without subjecting them to low pressure condition by condensational growth / virtual impaction was recently developed (Sioutas and Koutrakis, 1996). Ultrafine particles undergo condensational growth using supersaturation of water vapor and are separated by the virtual impactor. At low flow rate of 8 1 m in1 , the average particle collecting efficiency is 0.9 and the particle losses through the system were less than 5%. Neither the collection efficiency nor losses were found to depend on the particle chemical composition, hygroscopicity, or size distribution of the original aerosol. Other than inertially separating ultrafine particles grown to super micrometer size and collecting them on a filter, however, that system did not provide airborne concentrated ultrafine aerosols, suitable for use in inhalation exposure studies. Furthermore, the delivery flow rate of the concentrated droplets was only 0.8 1 m in1 , clearly too small to be of any practical use in inhalation exposures. This paper discuses the development and experimental characterization of a prototype ultrafine particle concentrator. This system is an expansion of the work by Sioutas and Koutrakis (1996) to a larger scale, suitable for inhalation studies. Specifically, the ultrafine particle concentrator samples particles at a flow rate of 110 1 m in 1 and concentrates them to a flow of either 7 or 3.5 1 min'1 . The prototype ultrafine particle concentrator is operated at high flow rates and concentrates the ultrafine particles with high collecting efficiency and low losses at atmospheric pressure. Therefore, it 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. provides a powerful tool to conduct human and/or animal exposures to real-time concentrated ultrafine particles. III-3 Methods III-3-1 Description of the Ultrafine Particle Concentrator Figure IE-1 is the schematic of the Ultrafine Particle Concentrator, as well as the experimental setup for its characterization. Air first passes over a pool of warm water to achieve saturation, and subsequently it is drawn through a condenser that allows the ultrafine particles to grow to super-micrometer size. In order to increase particle concentration, the grown particles are then drawn through a virtual impactor. The concentrated particles from the minor flow of the virtual impactor finally pass through a diffusion dryer to remove the excess water on the ultrafine particles and return them back to their original size and relative humidity. A more detailed description of the components of the system is given in the following paragraphs. The saturator consists of a vertical glass cylindrical tube 40 cm long and 15cm in diameter, partially filled with distilled deionized water. The tube is submerged in a water bath (Model 1204, Sheldon, MFG, Inc., Comellius, OR) that heats up the water during the experiments. The aerosol residence time is about 3 s to guarantee the saturation of the sampled air. The descending-ascending motion of particles as they enter and exit the saturator further enhances mixing and saturation. The vapor-saturated aerosol then enters the condenser, which consists of one horizontal aluminum tube, 80 cm long and 2.4 cm in diameter. These dimensions yield a residence time in the condenser of about 0.2 s. The condenser tube is submerged in a mixture of ice and water, plus a few grams of rock salt 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to keep the solution at -9°C. The condenser was designed to decrease the temperature of saturated aerosol by about 10°C at a flow rate in the range of 106.5 to 110 1 min'1 . The grown particles are then drawn through the minor flow of a virtual impactor connected immediately downstream of the condenser. The concentrated droplets are finally drawn through a diffusion dryer that removes excess moisture to become concentrated ultrafine particles. The diffusion dryer consists of a cylindrical screen, 1.8 cm in diameter, placed in the center of a glass tube, 6 cm in diameter. Both glass tube and screen are 20 cm long. The inner space between the two tubes is filled with calcium sulfate (Drierite™), to remove the excess water in the air stream. Relative humidity was measured immediately downstream of the dryer with a temperature/relative humidity probe (Cole-Parmer® Model 37960, Cole-Parmer® Instruments Co., Vernon Hills, IL). In all the experiments that will be described in the following paragraphs, the measured relative humidity downstream of the diffusion dryer ranged from 28% to 34%. The effect of parameters including vapor temperature in the saturator and minor- to- total flow ratio was investigated in order to determine an optimal configuration that concentrates ultrafine particles with high collection efficiency, low losses and high concentration enrichment factor. Table IE-1 shows the comparison of the present design and operating parameters to those by Sioutas and Koutrakis (1996) as well as those of a conventional condensation nucleus counter such as the TSI Condensation Particle Counter (CPC 3022, TSI Inc, St. Paul, MN). 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Ts Tc HEPA Fitter MIXTURE O F ICE MHO ROCK SALT V .V f W W .V .V .V .V .V . - • • A W V A " « JT « 'A • ’ V > A W V • »'.A- I Indoor m m m m m Po-210 Sources CONDENSER Water Trap Sampling Port Nebulizer SATURATOR Major flow 108-117 LPM Pump Pump Diffusion Dryer / Minor Flow (3. To 12 LPM ) VIRTUAL IM PACTOR (CONCENTRATOR) Concentrated Ul rafine particles W ARM DEIONIZED WATER Sampling Port 7 RH/T Probe TS 3022 CPC LA L A Figure 1 1 1 - 1 Schematic of the prototype ultrafine particle concentrtor and the experimental set-up for its cahracterization. Table HI-1 Comparison of design and operating parameters of this work with a previously published condensational growth system (Sioutas and Koutrakis, 1996) and the TSI 3022 CPC. Current Work Sioutas and Koutrakis(1996) TSI CPC 3022 Flow Rate (1 • min'1) 106.5-110 8 0 L O 1 Saturator Liquid Deionized H2 0 Water 1-butanol Temperature ( °C) 2 0 -4 0 50 35 Residence Time (s) 3 10 6 Condenser Temperature ( °C) 1 0 -3 0 8 8 Residence Time (s) 0 .2 0.5 0 .2 III-3-2 Experimental characterization o f the virtual impactor The first series of experiments was done to characterize particle collection efficiency and losses of the virtual impactor used to concentrate the grown ultrafine particles. The virtual impactor has an acceleration nozzle diameter of 0.37cm and collection nozzle diameter of 0.55 cm. The impactor has been designed to have a theoretical 50% collection efficiency cutpoint at about 1.5 pm. The ratio of the collection-to-acceleration nozzles is chosen 1.5 because it has been shown to minimize internal particle losses (Marple and Chen, 1980; Sioutas et al., 1994b). The principal parameter determining particle inertial separation is the Stokes number of a particle having a 50% probability of separation, St5 0 . St50 is defined as the 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ratio of the product of the jet velocity, U, and the particle relaxation time, x, vs. the impactor’ s nozzle diameter (round nozzle) or width (rectangular nozzle), W (Marple and Liu, 1974; Hinds, 1982): p d 502C U St5 Q = (2) 50 9p W where d50 is the geometrical diameter of particle having a 50% probability of impaction, U is the average velocity of the jet (cm/s), pp is the particle density (g/cm3 ), p, is the dynamic viscosity of the air (g/cm.s), and Cc is the Cunningham slip correction factor. The slip correction factor is given by the equation (Hinds, 1999): C c = l + — ^— [6.32 + 2.01 exp(-0.1095P d p ) ] (3) Pdp where P is the absolute pressure upstream of the impaction zone (in cm Hg) and dp is the particle diameter in pm. For the aforementioned values for U, W, and P, the St50 value corresponding to 1.5 pm is approximately 0.6. Two pressure taps were placed in the major and minor flows of the virtual impactor, respectively. The pressure drop across the major flow of the impactor at a flow rate of 106.5 and 110 1 min' 1 was 25 and 28 kPa, respectively. The pressure drop between the airstream in the minor flow and the atmosphere, however, was only 1 kPa (e.g., the absolute pressure of the concentrated particles is about 0.99 atmospheres). This pressure recovery is very important in using this system for inhalation studies, which cannot be conducted under a high vacuum. The experimental setup for the characterization of the virtual impactor is shown in Figure EI-2. Monodisperse aerosols were generated by atomizing dilute aqueous 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure HI-2 Experimental set-up for the characterization of the virtual impactor. Po-210 Neutralizes t HEPA Indoor Air Nebulizer Mixing Bottle Teflon Filter AP Filter 0 ~ f 1 Virtual Impactor 1 Pressure Gauge AP Filter Minor Flow (3.5 or 7 LPM) Major Flow (103 LPM) suspensions of fluorescent polystyrene latex particles using a constant output nebulizer (HEART, VORTRAN Medical Technology, Inc., Sacramento, CA). The range of monodisperse particles varied from 0.5 to 9 |tm. A ll the generated particles were drawn through a 1-liter bottle to remove the excess moisture and subsequently mixed with dry room air. The dry aerosol was drawn through a tube containing ten Po-210 neutralizes that reduces particle charges prior to entering the virtual impactor. A glass fiber filter (2 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pm pore, Gelman Science, Ann Arbor, MI) was connected to each of the major and minor flows, respectively, to collect the test particles. Each filter was connected to a pump with a calibrated flowmeter in line (Cole-Parmer® Instruments Co., Vernon Hills, IL). In addition, a similar glass fiber filter was connected in parallel to the test system to measure the concentration of the generated aerosol. At the end of each run, each glass fiber filter was placed in 5 ml of ethyl acetate to extract the fluorescent dye from the collected particles. The quantities of the fluorescent dye in the extraction solutions were measured by a Fluorescence Detector (FD-500, GTI, Concord, MA) to determine particle concentration. The collection efficiency of the virtual impactor was determined by dividing the amount of fluorescence on the minor flow filter to the sum of the amounts collected on both major and minor flow filters. Furthermore, particle losses were determined by comparing the concentration determined with the reference filter to the total concentration in the major and minor flows of the virtual impactor. 111-3-3 Experimental characterization of the Ultrafine Particle Concentrator The experiment setup for the characterization of the Ultrafine Particle Concentrator is shown in Figure HI-1. Ultrafine particles were generated by atomizing suspensions of ultrafine particles with a constant output HEART nebulizer (VORTRAN Medical Technology, Inc., Sacramento, CA). Different types of suspensions were used for this purpose, including monodisperse 0.05 and 0.1 pm PSL fluorescent latex particles (Polysciences Inc., Warrington, PA) as well as polydisperse aerosols of ammonium sulfate and ammonium nitrate. Finally, ultrafine indoor air particles were used as the test 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. aerosol. The generated ultrafine aerosols were dried and neutralized in a process identical to that described previously in the tests to characterize the virtual impactor. Table IH-2 summarizes the characteristics of the different aerosols tested. Prior to testing the ultrafine particle concentrator, the size distribution based on mass of the ammonium sulfate and ammonium nitrate aerosols were determined using the Microorifice Uniform Deposit Impactor (MOUDI, MSP Corp, Minneapolis, MN). The performance of the MOUDI is described by Marple et al. (1991). Measurement of the size-dependent mass concentration of indoor air was not considered necessary for the following reason: the CPC measures total particle concentration by counts, which, for indoor or ambient air, is dominated by ultrafine particles (Whitby and Svendrup, 1980; Seinfeld, 1986). The particle number concentration for pure deionized water was also measured to provide an estimate of background particles related to impurities in the water. The dilution air in this series of tests was drawn through a HEPA filter to ensure that only particles generated by atomizing deionized water are counted by the CPC. As seen in Table HI-2, the concentration of these particles is very low compared to the concentrations of the test aerosols, thereby eliminating any ambiguities on the nature of the aerosols used in each test. The effect of the saturator temperature in activating and growing ultrafine particles was also investigated. The temperature of the water bath was adjusted so that the ultrafine aerosol leaves the saturator a specific temperature. Five different aerosol temperature settings were tried: 20, 25, 30, 35 and 40°C, respectively. The temperature of the aerosol was recorded immediately upstream and downstream of the condenser as well as downstream of the diffusion dryer with a thermometer (VWR Brand, Cat610lb- 208). Total flow rates were set at 106.5 and 110 1 m in1 , respectively, depending on the 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. minor flow rate (e.g., 3.5 and 7 1 min'1 , respectively). In each experiment, room air was drawn through the system for about 30-45 minutes prior to generating particles to allow for temperature stabilization. Subsequently, the TSI Condensation Particle Counter (CPC 3022, TSI Inc., St. Paul, MN) was connected immediately upstream of the saturator and downstream of the diffusion dryer (as shown in Figure E H ) to measure the number concentrations of the original and concentrated ultrafine aerosols. For each saturator temperature, the five aerosol species, 0.05 and 0.1 pm fluorescent PSL particles, ammonium nitrate and ammonium sulfate, and ambient indoor air, were tested for two different minor-to-total flow ratios of the virtual impactor. Table IE-2 Characteristics of the ultrafine aerosols used in the experiments. Chemical Composition MMD Number Concentration Range (GSD) 1 (pm) (particles/cm3 ) Fluorescent Polystyrene Latex 0.05 (1.11) 1.5-1.95 x 103 Fluorescent Polystyrene Latex 0.10(1.05) 1.1-1.5 x 105 Ammonium Sulfate 0.09(1.8) 1.45-2.0 x 105 Ammonium Nitrate 0.09(1.9) 5.3-6.5 x 104 Indoor Air NM§ 7.5-11 x 103 Deionized Water NM§ 0.3-0.4 x 103 1 1 MMD is the mass median diameter measured with the Microorifice Uniform Deposit Impactor (MOUDI). GSD is the geometric standard deviation. § Not measured. Table IE-3 presents a summary of the saturated and cooled aerosol temperatures (Ts and Tc, respectively), the theoretical supersaturation achieved in each configuration (equal to the ratio of water vapor pressures at temperatures Ts and Tc, respectively), and the theoretical vapor concentration available for condensation on ultrafine particles. This concentration is equal to the difference in vapor concentrations at temperatures Ts and Tc, 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. respectively. It should be noted, however, that the available vapor does not only condense onto the ultrafine particles, but also it condenses on any other available solid surface (such as the walls of the condenser). Therefore the data on Table HI-3 present only qualitative differences between the various configurations tried. Table IH-3 Summary of the different temperature settings tried for the evaluation of the ultrafine particle concentrator. Data for vapor pressures and concentrations at different temperatures are from McQuiston and Parker (1982). Temperature of the saturated aerosol, Ts (°C) Temperature of the cooled aerosol, To (°C) Theoretical Supersaturation Theoretical vapor concentration available for condensation (g/mV 20 11 1.82 7.8 25 15 1.80 10.3 30 21 1.67 12.4 35 25 1.70 16.5 40 31 1.59 20.2 Defined as the ratio of vapor pressures at Ts and Tc, respectively. # Defined as the difference in the saturation vapor concentrations at Ts and Tc, respectively. III-4 Results and discussion III-4-1 Performance of the Virtual Impactor Results from the characterization of the virtual impactor are shown in Figures III- 3~5. Figure El-3 shows the collection efficiency and losses for a total flow of 1101 min'1 and a minor flow of 7 1 min'1 . Particle losses are generally 10% or less and do not seem to depend on particle aerodynamic diameter. Figure ni-4 shows the collection efficiency and losses for a total flow of 106.5 1 min'1 and a minor flow of 3.5 1 min'1 . Similarly, 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. particle losses are 10% or less and generally do not depend strongly on particle size. In both configurations, the aerodynamic diameter corresponding to a 50% collection efficiency is about 1.5 pm. 0.8 — • — Efficiency . . -o- - - L o sses 0.6 - - - " 2 0.4 -- 0.2 - - da- ....... o -j — o -----1 — n Aerodynamic Particle Diameter (jrm) Figure IH-3 Results from the coarse particle concentrator tests. Total flow: 110 LPM, minor flow: 7 LPM. 0.8 Efficiency L o sse s 0.6 0.4 •o 0.2 Aerodynamic Particle Diameter (jxm ) Figure IH-4 Results from the coarse particle concentrator tests. Total flow: 106.5 LPM, minor flow: 3.5 LPM. The sharpness of the collection efficiency curve of an impactor can be defined in terms of the geometric standard deviation (ag ), which is the ratio of the aerodynamic particle diameter corresponding to 84% collection efficiency to the 50% cutpoint (Marple and Willeke, 1976). Based on this definition, the value of ogis approximately 1.3 and 1.4 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. for minor flow rates of 3.5 and 7 1 min'1 , respectively (or equivalently, for minor-to-total flow ratios of 0.033 and 0.065, respectively). In their theoretical analysis on the effect of the minor-to-total flow ratio on the performance of virtual impactors, Marple and Chien (1980) showed that the steepness of the efficiency curve increases as the minor-to-total flow ratio decreases. Our experimental results provide corroboration to the theoretical analysis of Marple and Chien (1980). Figure III-5 shows the enrichment in the concentration of particles in the size range from 0.75-9 pm as a function of particle aerodynamic diameter. The concentration enrichment, CE, is given by the following equation: CE = ^ ( l - W L ) T ] vi (4) ? min where Qto t and q ^ are the intake and minor flows of the impactor, respectively, and t|vi and WL are the collection efficiency and fractional losses of the impactor. Figure 1 3 1 - 5 shows that for a minor flow of 3.5 1 min'1 , the concentration enrichment increases sharply from about 5 to about 28 as particle aerodynamic diameter increases from 1 to 2.2 pm. The enrichment is practically the same for particle in the aerodynamic diameter range of 2.2 to 9 pm. Similarly, for a minor flow of 7 1 min'1 , the concentration enrichment increases sharply from 3 to about 13 as particle aerodynamic diameter increases from 1 to 2.5 pm. For particles having aerodynamic diameters in the range of 3 to 9 pm, the enrichment value is about 15 and practically independent of particle size. The results of Figure DI-5 also suggest that the virtual impactor could be also used by itself (e.g., without the condensational growth component of the system) to concentrate particles having aerodynamic diameters in the range of 2.5 to 10 pm. Ambient particles in this 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. range are defined as coarse particles (Whitby and Svendrup, 1980; Finlayson-Pitts and Pitts, 1986) and technologies to concentrate these particles will be useful in assessing their toxicity, especially when conducted concurrently to inhalation studies using concentrated fine or ultrafine particles. 30 - 25 -- 20 Minor Flow: 7 LPM Minor Flow: 3.5 LPM 15 - Aerodynam ic Particle D iam eter (|j.m) Figure DI-5 Concentration enrichment achieved by the virtual impactor at different two minor flow tests (3.5 and 7 LPM). These values for the concentration enrichment are identical to the maximum obtainable concentration factors when dry, coarse particles (e.g., larger than about 3 pm in aerodynamic diameter) are sampled by the virtual impactor. Figure DI-5 shows that the maximum concentration enrichment obtained for particles in the aerodynamic diameter range of 3 to 9 pm is by a factor of about 15 and 27 at minor flow rates of 7 and 3.5 1 min-1, respectively. A very important implication of this result is that it shows conclusively that no particle coagulation occurs during the concentration enrichment process. If any coagulation had occurred, the measured number concentrations downstream of the diffusion dryer (thus the enrichment factors) would have been substantially smaller than the maximum obtainable values, shown in Figure DI-5. 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. III-4-2 Performance evaluation of the Ultrafine Particle Concentrator To initiate the parametric evaluation of the ultrafine particle concentrator, we set the temperature of the water bath so that the aerosol temperatures upstream and downstream of the condenser (Ts and Tc, respectively) are 35 and 25 °C. For these values of Ts and Tc we evaluated the concentration enrichment achieved by the ultrafine concentrator for five different ultrafine aerosol species and for two different minor-to- total flow ratios of the virtual impactor. Results from this first series of tests are shown in Table HI-4 and summarized in Figure HI-6. Figure HI-6 shows clearly that the concentration enrichment at a minor flow rate of either 3.5 or 7 1 min'1 does not depend on particle species. The average concentration enrichment for indoor air, 0.05 pm PSL, 0.1pm PSL, ammonium nitrate and ammonium sulfate particles is by 15.1, 14.8, 15.3, 15.1 and 15.2, respectively, when the virtual impactor operates with a minor flow of 7 1 min'1 . When the virtual impactor operates with a minor flow of 3.5 1 min'1 , the enrichment for indoor air, 0.05 pm PSL, 0.1pm PSL, ammonium nitrate and ammonium sulfate particles becomes 26.0, 24.7, 27.1, 23.8 and 25.1, respectively. 6 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table IE-4. Concentration enrichment achieved by the ultrafine particle concentrator as a function of particle type and minor flow rate. The temperatures of the aerosol upstream and downstream of the condenser are 35 and 25 °C, respectively. Aerosol Type Concentration Upstream of the Saturator (particles/cm3 ) Concentration Downstream of the Diffusion Dryer (particles/cm3 ) Concentration Enrichment Minor Flow 7 1 min Indoor Air 8.1 (±0.9) 103 1.3 (±0.2) 105 15.1 (±0.8) 0.05 pm PSL 1.7 (±0.2) 105 2.5 (± 0.2) 106 14.8 (±0.9) 0.1 pm PSL 1.3 (±0.2) 105 2.0 (±0.1) 106 15.3 (±0.7) Ammonium Nitrate 5.9 (±0.6) 104 8.9 (± 0.3) 105 15.1 (±0.6) Ammonium Sulfate 1.7 (±0.2) 105 2.6 (±0.1) 106 15.2 (±0.8) Minor Flow 3 .5 1 m in1 Indoor Air 8.1 (±0.9) 103 2.1 (±0.2) 105 26.0 (±1.1) 0.05 pm PSL 1.7 (±0.2) 105 4.2 (± 0.2) 106 24.7 (±1.8) 0.1 pm PSL 1.3 (±0.2) 105 3.5 (±0.2) 105 27.1 (±1.4) Ammonium Nitrate 5.9 (±1) 104 1.4 (±0.2) 106 23.8 (±0.8) Ammonium Sulfate 1.7 (±1) 105 4.3 (±0.2) 106 25.1 (± 1.3) * The particle concentration data correspond to averages of at least three tests. An important consideration regarding the ultrafine particle concentrator’s performance was its ability to increase the concentration of particles without substantial losses of volatile or semi-volatile materials during sampling. Ammonium nitrate is one of the most predominant volatile constituents of ambient particles, thus comparisons in the concentration enrichment based on ammonium nitrate would illustrate the magnitude (if any) of these losses during particle sampling and concentration. As ultrafine particles are heated during saturation, losses due to volatilization of nitrate could be possible. It should be noted that tests with pure ultrafine ammonium nitrate particles would tend to exaggerate such losses, since the vapor pressure of any substance over the surface of a particle increases with decreasing particle size (Finlayson-Pitts and Pitts, 1986). 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0 7 LPM □3.5 LPM Ts=35 deg C Tc=25 dea C I Indoor Air Ammonium Nitrate Ammonium Sulfate 0.05 PSL 0.1 PSL Figure HI-6 Ultrafine particle concentration enrichment at saturation temperature of 35 degrees C (Ts) and cooling temperature of 25 degrees C (Tc) for different types of ultrafine particles at two different minor flow rates (3.5 and 7 LPM). Figure EI-6 shows the average concentration enrichment of ultrafine ammonium nitrate particles is practically identical to that observed for non-volatile ammonium sulfate or polystyrene latex particles. This suggests that no significant volatilization losses occur during the concentration enrichment process. Ammonium nitrate dissociates to ammonia and nitric acid, with its dissociation constant increasing exponentially with temperature. However, the dissociation constant decreases sharply as the relative humidity (RH) exceeds 90-95% (Stelson and Seinfeld, 1982). For example, even at 50°C and at RH=95%, the dissociation constant of ammonium nitrate is approximately 7 ppb2 , which is the value of the dissociation constant at 18°C. Therefore, despite the increase in the aerosol temperature (which would have increased exponentially the value of the dissociation constant), saturation of the aerosol seems to prevent nitrate losses due to volatilization. In addition to the tests described above, the mass concentrations of ammonium nitrate upstream and downstream the ultrafine concentrator were measured in few additional experiments to confirm that no nitrate losses occur through the system. The 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Harvard/EPA Annular Denuder System (HEADS, Koutrakis et al., 1988) was used as the reference. The HEADS consisted of an annular denuder, coated with sodium carbonate to remove nitric acid from the air sample, followed by a sodium carbonate-coated glass fiber filter to collect ammonium nitrate particles. The HEADS sampled immediately upstream of the saturator. An identical HEADS was connected to the minor flow of the virtual impactor, downstream of the diffusion dryer. Both HEADS samplers operated at 7 1 m in1 . At the end of each experiment, the glass fiber filters were extracted with 0.15 mL ethanol in 5 mL ultrapure water. After sonication for about 5 minutes, the extracts were analyzed for nitrate by ion chromatography. Table El-5. Mass concentrations of ultrafine ammonium nitrate particles measured upstream and downstream of the ultrafine particle concentrator. Total flow: 110 1 min'1; minor flow rate: 7 1 min'1 . The temperatures of the aerosol upstream and downstream of the condenser are 35 and 25 °C, respectively._____________________________________ Experiment Concentration upstream Concentration downstream Concentration No. of the saturator of the diffusion dryer Enrichment _______________ (Eg/m3 )_________________(Eg/m3 )______________________ I 22.9 325.2 14.2 H 14.7 225.1 15.3 m 28.4 445.9 15.7 IV 11.3 163.9 14.5 V 19.6 298.0 15.2 Results from these tests are shown in Table HI-5. The concentration enrichment values for ultrafine ammonium nitrate particles based on mass are virtually identical to those based on particle counts (Figure EI-6) without any substantial loss of nitrate. In subsequent experiments, the temperature in the saturator was varied, and the performance of the ultrafine concentrator for each ultrafine particle species and each minor-to-total flow ratio was evaluated. Results from these tests are summarized in Figures EI-7~10. Figure IE-7 shows the concentration enrichment for indoor air, 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ammonium sulfate, ammonium nitrate, and 0.05 and 0.1 Jim PSL particles for two different virtual impactor configurations at aerosol temperatures upstream and downstream of the condenser (Ts and Tc, respectively) of 20 and 11 °C. The results of Figure ID-7 show that none of the ultrafine aerosols becomes concentrated at these temperature settings. Indoor Air Ammonium Nitrate Ammonium Sulfate Figure IH-7 Ultrafine particle concentration enrichment at saturation temperature of 20 degrees C (Ts) and cooling temperature of 11 degrees C (Tc) for different types o f ultrafine particles at two different minor flow rates (3.5 and 7 LPM). The supersaturation achieved in this configuration is 1.82, which should be sufficient to activate particles as small as 0.003 pm. The smallest particle size (d*) that can be activated by supersaturation is given by the Kelvin equation (Hinds, 1999): ri- = 4 g M ln W (5) pR T where S is the supersaturation, M, p, and a are the molecular weight, density, and surface tension of the condensing liquid, and R is the ideal gas constant. It is more likely that ultrafine particles become activated by the supersaturation, but the vapor 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concentration that is actually available for condensation does not grow the ultrafine particles to a size that can be concentrated by the virtual impactor (e.g., particles probable grow to sizes smaller than 1 pm). Figure IH-8 shows the concentration enrichment for indoor air, ammonium sulfate, ammonium nitrate, and 0.05 and 0.1 pm PSL particles for two different virtual impactor configurations at aerosol temperatures upstream and downstream of the condenser (Ts and Tc, respectively) of 25 and 15 °C, respectively. Indoor air ultrafine particles seem to be marginally concentrated, whereas the concentration of ammonium nitrate, ammonium sulfate and 0.05 pm particles increases by a factor of about 3 (± 0.4). The concentration of 0.1pm particles increases by factor of about 5. The higher concentration enrichment obtained for 0.1 pm PSL particles could be due to their somewhat larger initial size, compared to the rest of the ultrafine particles. This may also explain the reason for which the lowest enrichment is obtained for indoor air particles. The size distribution of indoor aerosols, measured by counts, is dominated by particles smaller than 0.05 pm (Thatcher and Layton, 1995), which apparently do not grow to a droplet size that can be concentrated by the virtual impactor. Although studies on technologies using condensational growth (such as the TSI Condensation Particle Counters) have shown that the ultrafine particles grow to the same final size (Ahn and Liu, 1990), it is possible that the combination of parameters in our system such as the temperatures of the saturated and cooled aerosol, as well as the relatively short residence times in the saturator and condenser do not allow all particles to reach their final droplet size. 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Indoor Air Ammonium Nitrate Ammonium Sulfate Figure ID-8 Ultrafine particle concentration enrichment at saturation temperature of 25 degrees C (Ts) and cooling temperature of 15 degrees C (Tc) for different types of ultrafine particles at two different minor flow rates (3.5 and 7 LPM). Figure HI-9 shows the concentration enrichment for indoor air, ammonium sulfate, ammonium nitrate, and 0.05 and 0.1 pm PSL particles for two different virtual impactor configurations at aerosol temperatures upstream and downstream of the condenser (Ts and Tc, respectively) of 30 and 21 °C, respectively. The growth of ultrafine particles in this configuration seems to be more uniform across the different particle species than those at lower temperatures. The average concentration enrichment values for minor flow rates of 7 and 3.5 1 min'1 range from approximately 8 to 13 for a minor flow rate of 7 1 min"1 , and from about 15 to 19 at a minor flow rate of 7 1 min"1 , respectively. By examining the concentration enrichment curves of the virtual impactor (Figure IH-5), we could hypothesize that, for these temperature settings, ultrafine particles grow to about 1.4-1.7 pm on the average. This higher enrichment is evidently due to the higher vapor concentration available for condensation in this configuration (12.4 g/m3 , compared to 7.8 and 10 g/m3 , corresponding to Ts=20 and 25 °C, respectively). The values of the concentration enrichment corresponding to either minor 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. respectively). The values of the concentration enrichment corresponding to either minor flow rate are lower than those obtained at Ts =35 °C and Tc=25 °C (main configuration), for the same reason; there is more vapor available when the aerosol becomes saturated at 35 °C and is cooled down to 25 °C. Finally, Figure IH-10 shows the concentration enrichment for indoor air, ammonium sulfate, ammonium nitrate, and 0.05 and 0.1 pm PSL particles for the two different virtual impactor configurations at aerosol temperatures upstream and downstream of the condenser (Ts and Tc, respectively) of 40 and 31 °C, respectively. The average concentration enrichment is by a factor of 14.4 (±0.3) and 25.5 (±0.7) at a minor flow rate of 7 and 3.5 1 min'1 , respectively. The enrichment values are practically identical to those obtained at Ts=35 °C and Tc=25 °C (main configuration) and similarly do not depend on the type of ultrafine particles used as the test aerosol. The main configuration (Ts=35 °C and Tc=25 °C) was therefore considered our optimum configuration, as it results in the highest concentration enrichment obtainable at the lowest possible saturation temperature. This feature minimizes the energy requirement for particle growth as well as the risk of losses of volatile compounds from the ultrafine aerosols. In order to use this technology for inhalation exposure studies to ambient ultrafine aerosols, a device separating the fine and coarse ambient particles (e.g., those having aerodynamic diameter larger than 0 .1-0 .2 pm) must be used upstream of this system. Technologies such as conventional impactors with small 50% cutpoints, already available in the literature, can be either readily used (Sioutas et al., 1997), or scaled-up (Marple et al., 1991; Hering et al., 1997) to separate particles larger than 0.15-0.2 pm from the air stream at sufficiently high flow rates, and with relatively low pressure drop 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ts=40 deg C Tc=31 deg C ■ 7 LPM □ 3.5 LPM U 20 Indoor Air Ammonium Nitrate Ammonium Sulfate 0.05 PSL 0.1 PSL Figure HI-9 Ultrafine particle concentration enrichment at saturation temperature of 30 degrees C (Ts) and cooling temperature of 21 degrees C (Tc) for different types of ultrafine particles at two different minor flow rates (3.5 and 7 LPM). Ts=30 deg C Tc= 21 deg C H 7 LPM □ 3.5 LPM Indoor Air Ammonium Nitrate Ammonium Sulfate 0.05 PSL 0.1 PSL Figure HI-10 Ultrafine particle concentration enrichment at saturation temperature of 40 degrees C (Ts) and cooling temperature of 31 degrees C (Tc) for different types of ultrafine particles at two different minor flow rates (3.5 and 7 LPM). (e.g., less than 10 kPa). One additional feature of an ultrafine particle concentrator for inhalation exposures should be the inclusion of devices that remove soluble gases (such as ammonia, sulfur dioxide, ozone and nitric acid) from the air sample, so that these gases 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. do not interfere with the ultrafine particles in the presence of water. Technologies such as diffusion denuders (Coutant et al., 1989; Eatough et al., 1993; and Koutrakis et al., 1993) can be used to efficiently remove these gases even at flow rates as high as 110 1 min'1 . Finally, an issue that should be addressed in future toxicology studies is potential redistribution of water soluble compounds from deep in the particle to the surface, which may enhance particle toxicity. III-5 Summary and conclusions We have developed and evaluated experimentally a prototype ultrafine particle concentrator. Ultrafine particles are first grown by means of supersaturation to a size that can be easily concentrated by a virtual impactor. The concentrated droplets are dried in a diffusion dryer to obtain a concentrated ultrafine aerosol. The condensational growth/virtual impaction system has been evaluated using monodisperse 0.05 and 0.1 pm fluorescent PSL particles, as well as polydisperse ultrafine ammonium sulfate, ammonium nitrate and indoor air particles. The concentrator operated at a flow rate of either 106.5 or 1101 m in1 , of which 3.5 or 7 min"1 were drawn through the minor flow of the virtual impactor. The performance of the concentrator was evaluated at different aerosol saturation and cooling temperatures. Our experimental results identified saturation of the ultrafine aerosols at 35 °C and cooling to 25 °C as the optimum temperatures for operation of the ultrafine particle concentrator. Lower temperatures either do not concentrate, or concentrate less efficiently the ultrafine particles. Increasing the saturation temperature to 40 °C and cooling to 31 °C does not improve the concentration enrichment achieved by the system. 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Our results indicated that the concentration enrichment does not depend on the chemical composition of the ultrafine aerosol. Hygroscopic ammonium sulfate, volatile ammonium nitrate, hydrophobic polystyrene latex and actual “real-life” indoor air ultrafine particles were all concentrated by practically the same factor. The average concentration enrichment was by a factor of 15.1 (± 0.4) and 25.5 (±1.9) for minor flow rates of 7 and 3.5 1 min"1 (or, equivalently, for minor-to-total flow ratios of 0.033 and 0.064, respectively). These enrichment factors are very similar to the maximum concentration factors that could be obtained by either configuration of the virtual impactor, thereby suggesting that ultrafine particle concentration is achieved with a very high efficiency and low losses. More importantly, the experimental results show that particle concentration occurs without any coagulation, which would have distorted the size distribution of the original ultrafine aerosols. It should be emphasized that this investigation is a pilot study whose main goal was to demonstrate the feasibility of concentrating efficiently ultrafine particles using a large cutpoint virtual impactor, with minimum losses and at a reasonably high output flow rate. If higher output flow rates are desired, more than one of these systems could be placed in parallel. The simplicity of the design of this technology (e.g., low energy input, small volume and uses water instead of organic fluids as a working liquid) make it ideal for use in a modular configuration. References Agarwal, J.K. and Sem, G.J. (1980) Continuous flow, single-particle-counting Condensation Nucleus Counter, J. Aerosol Sci., 11, 343-357. Ahn K.H. and Liu, B.Y.H. (1990) Particle activation and droplet growth processes in condensation nucleus counter-1. Theoretical background. J. Aerosol Sci., 21,249-261. 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Amdur, M.O., Chen, L.C., Guty, J., Lam, H.F. and Miller, P.D. (1988) Speciation and pulmonary effects of acidic SO2 formed on the surface of ultrafine zinc oxide aerosols, Atmos. Environ., 22:557-560. Anderson, K.R., Avol, E.L., Edwards, S.A., Shamoo, D.A., Peng, R.C., Linn, W.S., and Hackney J.D. (1992) Controlled exposures of volunteers to respirable carbon and sulfuric acid aerosols. J. Air and Waste Manage. Assoc., 42,437-442. Bates, D.V. and Sizto, R. (1989) Hospital admissions and air pollutants in Southern Ontario; the acid summer haze effect. Environ. Health Perspect. 79,69-76. Brain, J.D., Knudson, D.E., Sorokin, S.P., Davis, M.A. (1976) Pulmonary distribution of particles given by intratracheal instillation or by aerosol inhalation. Environ. Res. 11,13-16. Chen, L.C., Miller, P.D., Amdur, M.O., and Gordon, T. (1992) Airway hyperresponsiveness in guinea pigs Exposed to acid-coated ultrafine particles, J. Toxicol, and Environ. Health, 35,165-174. Coutant R.W., Callahan P.J., Kuhlman M.R., and Lewis R.G., (1989) Design and performance of a high-volume compound annular denuder. Atmos. Environ. 23 (10): 2205-2211 Dockery, D.W., Speizer, F.E., Stram, D.O., Ware, J.H., Spengler, J.D., and Ferris, B.J. (1989) Effects of inhalable Particles on respiratory health of children, Am. Rev. Res. Dis. 139,587-594. Drinker, P., Thomson, R.M., and Finn, J.L. (1927) Metal Fume Fever: II. Resistance acquired by inhalation of zinc oxide on two successive days, J. Ind. Hyg. Toxicol. 9(3), 98-105. Eatough, D.J., Wadsworth, A, Eatough, D.A., Crawford, J.W., Hansen, L.D., and Lewis, E.A., (1993) A multiple-system, multi-channel diffusion denuder sampler for the determination of fine-particulate organic material in the atmosphere. Atmos. Environ. 27A: 1213-1219 Ferin, J., Oberdorster, G., and Penny, D.P. (1992) Pulmonary retention of ultrafine and fine particles in rats, Am. J. Respir. Cell Mol. Biol., 6,535-542. Fernandez de la Mora, J., Hering, S.V., Rao, N., and McMurry, P.H. (1990) Hypersonic impaction of ultrafine particles, J. Aerosol Sci. 21,701-711. Finlayson-Pitts, B.J. and Pitts, J.N. (1986) Atmospheric Chemistry: Fundamentals and Experimental Techniques, John Wiley & Sons, New York. Gordon, T., Chen, L.C., Fine, J.M., Schlesinger, R.B., Su, W.Y., Kimmel, T.A., and Amdur, M.O. (1992) Pulmonary effects of zinc oxide in human subjects, guinea pigs, rats, and rabbits, Am. Ind. Hyg. Assoc. J., 53,503-509. 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hering S.V., and Stolzenburg, M.R. (1995) On-line determination of particle size and density in the nanometer size range, Aerosol Sci. and Technol., 23, 155-173. Hering, S.V., Gundel, L. and Daisey, J.M (1997) A microslot impactor for organic aerosol sampling. J. Aerosol Sci., 28,1283-1290. Heyder, J., Brand, P., Heinrich, J., Peters, A., Scheuh, G., Tuch, T. And Wichmann, E. (1996) Size distribution of ambient particles and its relevance to human health. Presented at the 2nd Colloquium on Particulate Air Pollution and Health, Park City, Utah, May 1-3. Hinds, W.C. (1999) Aerosol Technology. John Wiley & Sons Inc., New York. International Commission on Radiological Protection (1994) Human Respiratory Tract Model for Radiological Protection; a Report of Committee 2 of the ICRP. Oxford: Pergamon Press. Kleinman, M.T., Bhalla, D.K., Mautz, W.J and Phalen, R.F. (1995). Cellular and immunologic injury with PM-10 inhalation. Inhal. Toxicol. 7,589-602. Kleinman, M.T., Bhalla, D.K., Ziegler, B., Bucher-Evans, S., and McClure, T. (1997) Effects of inhaled fine particles and ozone on pulmonary macrophages and epithelia. Inhal. Toxicol, 5 ,371-388. Koutrakis, P., Wolfson, J.M., Slater, J.L., Brauer, M., and Spengler, J.D. (1988) Evaluation of an annular denuder/filter pack system to collect acidic aerosols and gases. Environ. Sci. and Technol. 22(12), 1463-1468. Koutrakis P., Sioutas C., Ferguson S., J.M. Wolfson, J.D. Mulik and R.M. Burton (1993) Development and evaluation of a glass honeycomb denuder/filter pack system to collect atmospheric particles and gases. Environ. Sci. & Technol 27:2497-2501. Lippmann M. (1989) Airborne acidity: estimates of exposure and human health effects. Environ. Health Perspect.s, 63,63-70. Marple V.A. and Liu B.Y.H. (1974) Characteristics of laminar jet impactors. Environ. Sci. & Technol., 8 , 648-654. Marple, V.A. and Chien, C.M., (1980) Virtual Impactors: A Theoretical Study, Environ. Sci.& Technol. 8,976-985. Marple, V.A., and Willeke, K. (1976) In Fine Particles: Aerosol Generation, Measurement, Sampling, and Analysis (Edited by B.Y.H. Liu). Academic press, New York. Marple, V.A., Rubow, K.L. and Behm, S. (1991) A Microorifice Uniform Deposit Impactor (MOUDI): Description, Calibration, and Use, Aerosol Sci. and Technol., 14, 434-446. 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. McQuiston, F. and Parker J.D. (1982). Heating, Ventilation and Air Conditioning. John Wiley and sons, New York. Oberdorster, G., Ferin, J., Gelein, R., Soderholm S.C., and Finkelstein, J. (1992) Role of alveolar macrophage in lung injury; studies with ultrafine particles, Environ. Health Perspect. 97, 193-197. Oberdorster, G., Ferin, J., and Lehnert, B.E. (1994) Correlation between particle size, in vivo particle persistence and lung injury, Environ. Health Perspect. 102,173-179. Oberdorster, G., Gelein, R.M., Ferin, J., and Weiss, B. (1995) Association of particulate air pollution and acute mortality: involvement of ultrafine particles?, Inhal. Toxicol. 7,111-124. Olawoyin, O.O., Raunemaa, T.M. and Hopke, P.K. (1995) A system for aerodynamically sizing ultrafine radioactive particles, Aerosol Sci. and Technol. 23,121-130. Peters, A.; Dockery, D.W.; Heinrich, J; Wichman, H.E. (1997). Short-term effects of particulate air pollution on respiratory morbidity in asthmatic children. Eur. Respir. J. 10, 872-879. Schwartz, J. and D.W. Dockery. (1992) Increased mortality in Philadelphia associated with daily air pollution concentrations, Am. J. Epidemiol. 135,12-19. Seinfeld, J.H. (1986) Atmospheric Chemistry and Physics o f Air Pollution. John Wiley and sons, New York, p. 23. Sioutas, C., Koutrakis, P., and Burton, R.M. (1994a) Development of a low cutpoint slit virtual impactor for sampling ambient fine particles, J. Aerosol Sci., 25(7), 1321- 1330. Sioutas, C., Koutrakis P., and Olson, B.A. (1994b) Development and evaluation of a low cutpoint virtual impactor, Aerosol Sci. & Technol., 21(3), 223-236. Sioutas, C., Koutrakis, P., and Burton, R.M. (1995) A technique to expose animals to concentrated fine ambient aerosols, Environ. Health Perspect., 103, 172-177. Sioutas, C., Koutrakis, P., Ferguson, S.T. and Burton, R.M. (1995b) Development and evaluation of a prototype Ambient Particle Concentrator for inhalation exposure studies, Inhal. Toxicol. 7, 633-644. Sioutas, C. and Koutrakis, P. (1996) Inertial separation of ultrafine particles using a condensational growth/virtual impaction system. Aerosol Sci. and Technol., 25, 424- 436. Sioutas, C., Koutrakis, P., Godleski, J., Ferguson, S.T., Kim, C.S. and Burton, R.M. (1997) Harvard/EPA ambient fine particle concentrators for human and animal exposures. J. Aerosol Sci., 28(6), 1057-1071. 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sioutas, C., Wolfson, M., Ferguson, S.T., Ozkaynak, H. and Koutrakis, P.K. (1997) Inertial collection of fine particles using a high-volume rectangular geometry conventional impactor. J. Aerosol Sci., 6,1015-1028. Stelson, A.W. and Seinfeld, J.H. (1982), Atmos. Environ., 16:983-992. Thatcher, T.L. and Layton, D.H (1995) Deposition, resuspension and penetration of particles within a residence. Atmos. Environ., 29(13), 1487-1497. Thurston, Ito, K., Hayes, C.G., Bates, D.V., and G., Lippmann, M. (1994) Respiratory hospital admissions and summertime haze air pollution in Toronto, Ontario: Consideration of the role of acid aerosols, Environ. Res. 65,270-290. Whitby, K.T. and Svendrup, G.M. (1980) California Aerosols: their physical and chemical characteristics, Adv. Environ. Sci. Technol. 10, A ll. 8 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter IV A NEW GENERATION OF PORTABLE COARSE, FINE AND ULTRAFINE PARTICLE CONCENTRATORS FOR USE IN INHALATION TOXICOLOGY Seongheon Kim, Ming-Chih Chang, Daeik Kim and Constantinos Sioutas Department of Civil and Environmental Engineering University of Southern California 3620 South Vermont Avenue Los Angeles, CA 90089-2531 Inhalation Toxicology Volume 12(S1): 121-137, 2000 8 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IV-1 Abstract This study presents the development of prototype portable coarse, fine and ultrafine particle concentrators. A single-round nozzle virtual impactor operating at an intake flow of 120 L/min is used to concentrate coarse particles (e.g., 2.5-10 pm) by a factor up to 40 depending on the minor flow rate. Fine and ultrafine particles are concentrated by first growing to super-micrometer sizes via supersaturation. This is accomplished by first drawing these particles over a pool of warm, deionized, distilled water to achieve saturation and then through a condenser that allows the particles to grow to super-micrometer size, followed by concentration in a virtual impactor. After concentration, particles are returned back to their original size distribution and relative humidity by removing excess moisture in a diffusion drier. The performance of these concentrators was evaluated using generated monodisperse particles as well as ambient air particles. Average concentration enrichment factors were 9.5,20 and 37 for a minor flow of 12,6 , and 3 L/min, respectively. The average concentration enrichment based on particulate sulfate and nitrate was by a factor of 2 0 and 22.6, respectively. The HEADS sampler was used as the reference sampler. The enrichment values based on particulate nitrate indicate that no nitrate loss occurs during particle concentration enrichment. The concentration of particulate elemental (EC) and organic carbon (OC) was also evaluated, using the MOUDI as a reference sampler. The average concentration enrichment factors obtained for EC and OC were 20.4 and 21.6, respectively. Our experimental results indicated that the enrichment in concentration is not dependent on particle size and chemical composition. Because of their compact size and high concentration efficiency, the concentrators described in this study are inexpensive and portable so can be moved easily to several locations over seasons that differ in PM chemical composition and source profiles. 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IV-2 Introduction The National Research Council’s Committee on Research Priorities for Airborne Particulate Matter (1998) has recognized appropriately the need for hypotheses-driven studies to investigate mechanisms responsible for adverse effects associated with ambient particulate matter (PM). Epidemiological evidence associating ambient particulate pollution with adverse health effects in humans is extensive (American Thoracic Society, 1996; Environmental Protection Agency, 1996). Nevertheless, fundamental uncertainty and disagreement persist regarding what physical and chemical properties of particles (or unidentified confounding environmental influences) influence health risks, what pathophysiological mechanisms are operative, and what air quality regulations should be adopted to deal with the health risks (Vedal, 1997). This lack of understanding reflects an inability of controlled laboratory investigations to detect effects of low levels of artificially generated particulates, which might support the epidemiological findings. The recent development of ambient Particle Concentrators (Sioutas et al., 1995; Sioutas et al., 1997; Gordon et al., 1999) has made it possible to perform laboratory exposures with "real-life" ambient aerosols at increased (but still environmentally realistic) concentrations. Initial results suggest greatly increased toxic responses (as compared to these with artificial particles) and suggest that this type of exposure system may provide a useful method for assessing the health effects of ambient particles and for identifying specific risk factors and the means of controlling them (Godleski et al., 1996; Clarke et al., 1999). Current North American toxicity studies involving particle concentrators are being conducted in the Northeast, where the primary constituents of PM2.5 are sulfate and 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. organics (Spengler and Thurston, 1983; Burton et al., 1996). Similar studies in the West Coast of the U.S. have been initiated very recently and are rather limited in examining the effects of particles in the size range of 0.1-2.5 pm without targeting specific constituents of ambient PM. Nevertheless, the currently available concentrators focus mainly on concentrating the accumulation mode of ambient PM (e.g., PM2.5 without its ultrafine component), they are bulky, hence not easily transportable, and the concentration enrichment depends on particle size, with larger particles of the accumulation mode being concentrated more effectively than smaller particles (Sioutas et al., 1997; Gordon et al., 1999). Particle size and composition are two very important parameters in determining particle toxicity. There is a great need for PM toxicity studies that target specific chemical and/or physical PM properties, in their “real-life” state and at realistic levels. Such data are needed to address many of the most important air pollution-related health problems in the U.S., including non-cancer health effects, asthma, respiratory and cardio pulmonary disease, and the role of particulate matter in human mortality. The work presented in this paper discusses the development of a new generation of recently developed portable particle concentrators^ These technologies, known as the California Particle Concentrators, maintain the concentrated particles in an airborne state and supply them to exposure chambers for human or laboratory animal inhalation studies. The Concentrators presented in this paper represent an extension of a prototype Ultrafine Concentrator developed by Sioutas et al. (1999). In that system, ultrafine PM was first grown through condensational supersaturation to super-micrometer droplets, concentrated by means of a 1.5 pm virtual impactor and returned to its original size by passing through a diffusion dryer. The study by Sioutas et al. (1999) focused on optimizing the design 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and operating parameters of that system, such as the saturator and condenser temperatures, to yield maximum obtainable concentrations of ultrafine particles. A successful pilot study utilizing a prototype Ultrafine Particle Concentrator to expose aged rats to selected components of ambient ultrafine particles has been completed (Kleinman et al., 1998). The study demonstrated significant cardiophysiological changes in 24-month old rats after exposure to ammonium nitrate and carbon particles having a 90 nm count median diameter. The observed changes were in the direction of those observed in humans with shock; i.e. depression of heart rate, blood pressure and blood delivery by the heart. In this paper we discuss how this technology was extended to concentrate particles in the diameter range of 0.01-10 |im. The enrichment in concentration is determined experimentally as a function of particle size using monodisperse as well as polydisperse aerosols. In addition, detailed chemical characterization of the ambient and concentrated aerosols was performed. Comparisons between the mass, sulfate, nitrate, elemental and organic carbon concentrations of ambient and concentrated PM-2.5 aerosols are presented and discussed in detail. Along with separating the particles from the majority of the air mass, the California Concentrators are capable of concentrating particles of discrete size groups. These groups could be Ultrafine Particles (<0.1 pm), which are freshly generated particles, such as those generated by combustion, Fine Particles of any size sub-range between 0-2.5 pm and Coarse (>2.5 pm) particles. Due to their compact size and high concentration efficiency, these Concentrators are portable and will be deployed to several locations in California (including the first California Supersite at Fresno), over seasons that differ in chemical composition, source profiles and atmospheric chemistry. Thus, 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. specific size ranges and chemical characteristics of concentrated ambient PM will serve as test aerosol to conduct specific hypotheses-driven animal inhalation toxicity studies. IV-3 Methods IV-3-1 Description o f the Coarse Particle Concentrator (CPC) Particles in the Coarse Particle Concentrator (CPC) are drawn at 120 L/min through a 2-cm diameter inlet tube. A 0.8 cm ring, coated with silica grease, is inserted to the inlet to remove particles larger than approximately 10 pm. The design of this ring has been based on the experimental and numerical work by Muyshondt et al. (1996) and Chen and Pui (1995) on particle deposition in abrupt (i.e., 90 ° angle) pipe contractions. The relationship between the fraction of particles depositing on the walls of the contracted part and particle aerodynamic diameter resembles that of conventional impactors. Particle deposition on the contraction can be predicted by means of the product St(l- A J AO, where St is a modified Stokes number, defined as: p .U td P2C p St = p — p - (1) 9 n d Q where dp, pp, Cp are the particle diameter, density and slip correction, p is the air viscosity, Ui is the velocity at the inlet of the contraction and do is the diameter of the contraction. Ao and A; are the areas of the ring and the inlet tube, respectively. The above 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dimensions of the contraction (e.g., ring) were chosen to yield 50% removal efficiency of 10 pm particles at a flow rate of 120L/min through the ring. The CPC is a single-stage, round-jet nozzle virtual impactor (Figure IV-la) with an acceleration nozzle diameter of 0.37cm and collection nozzle diameter of 0.56 cm. The virtual impactor has been designed to have a theoretical 50% collection efficiency cutpoint at about 1.5 pm when operating at an intake flow rate of 120 L/min. Depending on the desirable enrichment factor, the minor flow could vary from 3-12 L/min, resulting in concentration enrichment by a factor of 40 to 10, respectively. The CPC was characterized in laboratory experiments using monodisperse fluorescent particles in the size range of 0.7-9 pm in aerodynamic diameter. Results from the characterization of the virtual impactor are shown in Figure IV-lb at three different minor flow rates, 3, 6 and 12 L/min, respectively. For particles having aerodynamic diameters in the range of 3 to 9 pm, the enrichment value is about 9, 20 and 36 (e.g., very close to the ideal enrichment values, defined as the ratio of the total-to-minor flow rate) and practically independent of particle size. The CPC is also a component of the Fine and Ultrafine PM Concentrators described below, serving as the concentrator for the grown particles by super-saturation. 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PM -10 ring-insert concentrated coarse particles 0.56 cm 0.37 cm Ambient PM inlet collection nozzle acceleration nozzle fine paticles 0.9 0.8 0.7 5 * S S § 0.6 " 8 0.5 I 1 0.4 I 0.3 u 0.2 0.1 — • — Efficiency — o— L o sse s 2 3 8 9 10 0 1 1 4 5 6 ' Aerodynam ic Particle Diameter (jam) 7 Figure IV-1 (a) Single-nozzle coarse particle concentrator, (b) Coarse particle concentrator, concentration enrichment as a function of particle size. IV-3-2 Description of the Fine Plus Ultrafine Particle Concentrator (F+UFPC) This Concentrator is an extension of the prototype Ultrafine Concentrator described in detail elsewhere (Sioutas et al., 1999), and similar to the CPC, it operates at an intake flow rate of 120 L/min. Briefly, the aerosol is passed over a pool of warm 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. deionized distilled water to achieve saturation. Subsequently, it is drawn through a condenser that allows the particles to grow to super-micrometer size. Particle enrichment occurs by drawing the grown particles through the Coarse Particle Concentrator (described above). The concentrated particles from the minor flow of the virtual impactor pass through a diffusion dryer to remove the excess vapor and return to their original size and relative humidity (Figure IV-2). The effect of parameters including vapor temperature in the saturator and minor-to- total flow ratio was investigated in order to determine an optimal configuration that concentrates ultrafine (e.g., 0 .0 1 -0.1 pm) particles with high collection efficiency, low losses and high concentration enrichment factor. Our experimental results identified saturation of the ultrafine aerosols at 35 °C and cooling to 25 °C as the optimum temperatures for operation of the Ultrafine Particle Concentrator. Lower temperatures either do not concentrate, or concentrate ultrafine particles less efficiently. Increasing the saturation temperature to 40 °C and cooling to 31 °C does not significantly improve the concentration enrichment. All of these experiments are described in detail by Sioutas et al. (1999). IV-3-3 Design and Evaluation of the Diffusion Dryers The concentrated droplets are drawn through a diffusion dryer that removes excess moisture so that the grown particles return to their original size. The diffusion dryer consists of a cylindrical screen, 1.8 cm in diameter, placed in the center of a glass tube, 6 cm in diameter. Both glass tube and screen are 20 cm long. The inner space between the two tubes is filled with a desiccant to remove the excess water in the air 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. HEPA Filter MUCTURE OF ICE AND ROCK SALT Indoor Po-210 Sources CONDENSER Water Trap Sampling Port Nebulizer SATURATOR Major flow 108-117 LPM Pump Pump Diffusion Dryer / Minor Flow (3. To 12 LPM ) VIRTUAL IMPACTOR (CONCENTRATOR) Concentrated Ull rafine particles W ARM DEIONIZED W ATER r - c RH/T Probe Sampling Port TSI 3022 CPC v O O Figure IV-2 Schematic of the fine plus ultrafine particle concentrates' (F+UFPC) and the test apparatus used for its characterization. stream. Three different materials were tested as the desiccant used to dry the grown liquid particles: 1. Drierite: Anhydrous Calcium Sulfate (CaS04 ) with 3% cobalt chloride (CoCl2 ) as indicator, 8 mesh; (W.A. Hammond Drierite Company LTD. Xenia, OH) 2. Desiccant; 99.6% Si0 2 as 100% indicating coat, 6 -8 mesh (EM Industries, Inc. Gibbstown, NY) 3. Silica Gel; 100% plain Si0 2 6-12 mesh (Eagle Chemical CO., INC., Mobile, AL). The purpose of these tests was to investigate whether these desiccants would reduce the aerosol RH at a given flow rate from 100% to less than 40% (e.g., crystallization point of most hygroscopic salts) to ensure that the concentrated fine or ultrafine PM is dry. Relative humidity was measured immediately downstream of the dryer with a temperature/relative humidity probe (Cole-Parmer® Model 37960, Cole-Parmer® Instruments Co., Vernon Hills, IL). Tests were conducted at an intake flow of 120 L/min and at two different minor flow rates (6 and 12 L/min, respectively). Experiments lasted for a period of 6 hours. Results from the diffusion dryer tests are shown in Figures IV-3a and 3b. Figure IV-3a shows that after the first hour of operation at 6 L/min, the Drierite material becomes saturated and cannot further remove any excess vapor. The RH of the concentrated aerosol increases from about 50% to 90% within 2 hours of operation. Both the EM Si0 2 as well as the silica gel desiccants maintain their vapor removal efficiency over a 6 -hour sampling period, with the silica gel reducing RH more effectively (e.g., to less than 40% over the entire 6 -hour period) than the EM desiccant. 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 90 S' 80 ^ 70 | 60 I 50 a c3 t 20 10 0 0 50 100 150 200 250 300 350 400 Elapsed time (min) A- - A - a A A A A * ■ - a A □ D □□ n r- - - - - A Drierite □ Desiccant • Silica gel Figure IV-3a Performance of various Diffusion Drier materials at 6 LPM. ' v i (4-H *3 Is > » ^ § * § 5 a t " a > f* 80 60 A i A A i * ---------A--A--- 40 o o - - o O “ O- ©~ ~ O O^O 20 0 0 100 200 300 400 O at 6LPM A at 12LPM Elapsed time (min) Figure IV-3b Performance of Diffusion Drier with Silica gel at different minor flow rates. Figure IV-3b shows the vapor removal efficiency of the silica gel as a function of flow rate trough the diffusion dryer. As expected, the RH of the dried aerosol is lower at 6L/min (38-40%) than that at 12 L/min ( 55-60%) due to the longer time available for vapor diffusion to the dryer walls. Regardless of the minor flow rate, our tests identified silica gel as the optimum desiccant, and the rest of the F+UFPC evaluation was conducted using this material in the diffusion dryer. 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FV-3-4 Laboratory Evaluation o f the Fine Plus Ultrafine Concentrator (F+UFPC) The experimental setup for the characterization of the Fine Plus Ultrafine Particle Concentrator is shown in Figure IV-2. Monodisperse aerosols were generated by atomizing suspensions of ultrafine and fine particles using a constant output HEART nebulizer (VORTRAN Medical Technology, Inc., Sacramento, CA). Different types of suspensions were used, including monodisperse PSL fluorescent latex particles (size range 0.05-5 pm; Polysciences Inc., Warrington, PA) as well as monodisperse silica beads (particle size range 0.15 to 0.9 pm; Bangs Laboratories, Inc., Carmel, IN). Finally, ultrafine indoor air particles were used as the test aerosol. The generated PSL ultrafine aerosols were dried and neutralized and were drawn though the F+UFPC at 120 L/min. The dilution air in this series of tests was drawn through a HEPA filter to ensure that only particles generated by atomizing deionized water are counted by the CPC. The aerosol was mixed and saturated with water vapor at 35 °C, and subsequently drawn through the condenser. The temperature of the aerosol exiting the condenser was about 24 (± 1)°C. The grown droplets were subsequently drawn through the Coarse Concentrator. Three different minor flow rates were tested, 3, 6 and 12 L/min, respectively (corresponding to theoretical enrichment factors of 40, 20 and 10, respectively). The TSI Condensation Particle Counter (CPC 3022, TSI Inc., St. Paul, MN) was connected immediately upstream of the saturator and downstream of the diffusion dryer (as shown in Figure IV-2) to measure the number concentrations of the original and concentrated aerosols. 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IV-3-5 Evaluation of the Fine Plus Ultrafine Concentrator Using Indoor Aerosols In addition to laboratory experiments, the performance of the F+UFPC was evaluated in a field study, conducted indoors in the Aerosol Laboratory of the University of Southern California. The mass, sulfate and nitrate PM2.5 indoor concentrations of the F+UFPC were compared to those measured by means of a collocated Harvard/EPA Annular Denuder Sampler (HEADS; Koutrakis et al., 1989). The HEADS operated at a flow rate of 10 L/min, and consisted of a conventional impactor inlet with a 50% cutpoint of 2.5 pm in aerodynamic diameter, a sodium carbonate-coated denuder to remove nitric acid from the air sample, followed by a 4.7-cm Teflon membrane to collect particles, and a sodium carbonate-coated glass fiber filter to collect nitric acid that volatilized from the collected PM on the Teflon filter. A 4.7 cm Teflon filter (2 pm pore, Gelman, Science, Ann Arbor, MI) was placed immediately downstream of the diffusion dryer of the F+UFPC, which operated at a total flow of 120 L/min, of which 6 L/min was drawn as the minor flow. The Teflon filters were weighed before and after each field tests in a Mettler 5 Microbalance (MT 5, Mettler-Toledo Inc., Hightstown, NJ) under controlled relative humidity (e.g. 40-45%) and temperature (e.g., 22-24 degrees C) conditions, in order to determine the mass concentrations. Filters were weighed immediately at the end of each experiment as well as after a 24-hour equilibration time period. Laboratory and field blanks were used for quality assurance. Filters and filter blanks were weighed twice in order to increase precision. In case of a difference of more than 2 pg between consecutive weightings, a filter was weighed for a third time. The Teflon filters of the F+UFPC and HEADS as 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. well as the glass fiber HEADS filter were then analyzed by means of ion chromatography to determine the concentrations of particulate sulfate and nitrate. In addition to these tests, the indoor elemental and organic carbon (EC/OC) concentrations of the F+UFPC were compared to those determined using a modified Microorifice Uniform Deposit Impactor (MOUDI, MSP Corporation, Minneapolis, MN), described in more detail by Marple et al. (1991). The MOUDI operates at 30 L/min and classifies particles in the following size intervals: <0.1, 0.1-0.18, 0.18-0.3, 0.3-0.56, 0.56-1.0, 1.0-1.8, 1.8-3.2, 3.2-5.0, and 5.0-10 pm. For the purposes of our experiments, we only used the first two stages of the MOUDI and all particles smaller than 3.2 pm in aerodynamic diameter were collected on a 3.7-cm diameter quartz filter (Pallflex Corp., Putnam, CT). A 4.7-cm diameter quartz filter (Pallflex Corp., Putnam, CT) was connected to the minor flow of the F+UFPC, immediately downstream of the diffusion dryer. The mass concentrations measured by the MOUDI and F+UFPC were determined gravimetrically using the same process described above. The EC/OC concentrations were determined by thermo-analysis. An aliquot of approximately 0.2 cm2 from each filter was placed in a platinum boat containing M n02. The sample was acidified with a dilution of HC1 and heated to 115 degrees C to remove the water and C 0 2 (from sample carbonates). The boat was then advanced into a dual zone furnace where M n02 oxidized OC in the sample at 550 degrees C and EC at 850 degrees C. The C 0 2 formed was converted to CH4 for detection by a Flame Ionization Detector (FID). The analytical method is described in detail by Fung (1990). The MOUDI instead of the HEADS was used to measure EC/OC concentrations because of its higher sampling flow rate, which allowed us to reduce the sampling time to 4-5 hours. The F+UFPC was used without any pre-selective inlet to remove particles 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. above 2.5 or 3.2 pm. As it will be shown in the Results and Discussion section of this paper, particles larger than 3 pm are not concentrated by the F+UFPC because inertial deposition mechanisms remove these particles prior to reaching the virtual impactor. IY-4 Results and discussion IV-4-1 Laboratory Tests Results from the laboratory evaluation of the Fine Plus Ultrafine Particle Concentrator are summarized in Figure IV-4 at three different minor flow rates (e.g., 3, 6 and 12 L/min). In all configurations, the major flow rate is adjusted to yield a total intake flow of 120 L/min. Hence, the maximum obtainable concentration enrichment factors for each configuration are 40, 20, and 10, respectively. The concentration enrichment factors as a function of particle size, shown in Figure IV-4, are based on particle number concentrations measured upstream and downstream of the F+UFPC, and have been obtained using monodisperse aerosols in the size rage of 0.05-5 pm, except of the data corresponding to 0.025 pm particles. The enrichment values corresponding to 0.025 pm were obtained for indoor air particles, measured again by the Condensation Particle Counter (CPC 3022, TSI inc., St. Paul, MN). The count-based size distribution of ambient or indoor aerosols is dominated by particles smaller than 0.05 pm, peaking at around 0.02-0.035 pm (Whitby and Svendrup, 1980). We chose the size of 0.025 pm as an approximate number median diameter representing indoor aerosols, in order to include all of our experimental results in one graph. 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45.0 u § 40.0 P u « 35.0 c < u g 30.0 J S | 25.0 w 20.0 I 15-0 I 10.0 o 0.01 0.10 1.00 10.00 Particle Diameter (jim) .. 1 ' 1 1 1 ------------A -------- A -------------------------------------A-------------------------- A 3LPM O 6LPM 1 1 1 1 ■ 12LPM 1 1 O ! ° O ° cj 0 o ° o f 8 1 1 1 1 ■ ■ I ■ ■ I ■ ■ ■ i ----------1 ----- 1 --- 1 -- 1 --L . - U . . U J - I --------- 1 -----. --- 1 - - 1 --■ — ■ . . . . ■ . . I . 1 ---------. -----L , * . , Figure IV-4 Characterization of the Fine+Ultrafine PM concentrator using monodisperse aerosols for 3 minor flows. Total intake flow: 120 LPM. Figure IV-4 shows clearly that the concentration enrichment corresponding to a minor flow rate of 3, 6 or 12 L/min does not depend on particle size for all particles smaller than 2 pm. The average concentration enrichment for ultrafine indoor air as well as monodisperse 0.05-2 pm PSL particles is by a factor of 9.5, 20 and 37, when the virtual impactor operates at a minor flow of 3, 6 and 12 L/min, respectively. These concentration enrichment values are essentially identical to the maximum obtainable concentration factors. An important implication of these results is that no particle coagulation occurs during the concentration enrichment process. If any coagulation had occurred, the measured number concentrations downstream of the diffusion dryer (hence the enrichment factors) would have been substantially lower than the maximum obtainable values. The concentration enrichment values decrease rapidly to 2 or less for particles larger than 3 pm in diameter. Inertial deposition mechanisms (most likely impingement on the surface of the water in the saturator) remove these particles before they reach the 1.5 pm cutpoint virtual impactor, where they would have been concentrated by the same 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. factor as the rest of the aerosols. This is a desirable (albeit fortuitous) result, as it makes the use of a pre-selective PM-2.5 inlet to remove these particles from the air sample unnecessary. IV-4-2 Indoor tests Results from the comparisons between the PM-2.5 mass, sulfate, nitrate, elemental and organic carbon concentrations determined using the F+UFPC and those using the HEADS or MOUDI are summarized in Tables IV-1~4, respectively. The total flow of the F+UFPC is 120 L/min, of which 6 L/min are drawn through the collection nozzle as a minor flow, ideally containing all of the particles smaller than 2.5 pm, enriched in concentration by a factor of 20. The first two columns in Table IV-1 show the PM-2.5 mass concentrations measured indoors by the HEADS or MOUDI (depending on the type of filter used) and the F+UFPC. The third column shows the values of the collection efficiency of the F+UFPC, defined as the ratio of the minor flow concentration of the F+UFPC to that of the HEADS or MOUDI, divided by 20 (i.e. the ideal enrichment factor). We employed this term to obtain an estimate of the fraction of the total particulate mass that wasactually collected by the minor flow of the F+UFPC and thus account for particle losses (a similar definition of the collection efficiency of a concentrator has been employed by Sioutas et al., 1995; Sioutas et al., 1997 and Gordon et al., 1999). As the results in Table IV-1 suggest, a virtually perfect mass balance was obtained between the HEADS or MOUDI and the F+UFPC. The average collection efficiency of the F+UFPC was 1.01 (± 0.11), and the resulting concentration enrichment factor was 20.04 (± 2.2), 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. both very close to the ideal values. It should be noted that there was no detectable difference (e.g., less than 4 pg) between the weights of the Teflon filter of the minor flow of the F+UFPC immediately at the end of each test and after the 24-hour equilibration period. This is another indication of complete water vapor removal by passing the particles through the diffusion dryer. Table IV-1 Comparisons between the mass concentrations determined using the F+UFPC and the HEADS or MOUDI using indoor air as the test aerosol. Reference (jlg/m3 ) F+UFPC (jxg/m3 ) Collection Efficiency Enrichment factor HEADS Experiments 17.8 351.1 0.99 19.8 14.0 302.1 1.08 21.5 13.7 210.0 0.77 15.3 23.7 484.4 1.02 20.4 17.8 347.8 0.98 19.6 17.8 335.6 0.94 18.9 10.3 264.4 1.28 25.5 19.2 422.2 1.1 21.9 10.8 235.0 1.08 21.7 16.3 278.9 0.86 17.1 9.6 190.0 0.99 19.7 MOUDI experiments 10.93 235.20 1.08 21.5 16.1 334.4 1.04 20.7 18.3 324.4 0.88 17.7 25.9 496.7 0.96 19.1 25.5 420.0 0.82 16.4 33.9 666.7 0.98 19.7 19.4 398.9 1.03 20.5 55.6 1145.6 1.03 20.6 32.4 685.6 1.06 21.15 42.0 922.2 1.1 21.94 Average 1.00 20.04 S.D. 0.11 2.22 Table IV-2 shows the PM-2.5 concentrations obtained using the HEADS and F+UFPC. Similarly to the results based on mass concentrations, excellent agreement was 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. obtained between the two samplers, with the average concentration enrichment factor being 20.04 (±3.6) and the average collection efficiency of the F+UFPC being 1.00 (± 0.18). Table IV-2 Comparisons between sulfate concentrations determined using the F+UFPC and the HEADS using indoor air as the test aerosol.______________________________ HEADS (Mg/m3) F+UFPC (Mg/m3) Collection Efficiency F+UFPC Enrichment factor 2.1 39.0 0.92 18.47 2.1 42.11 1.00 20.00 1.7 33.22 0.97 19.5 0.70 19.56 1.38 27.79 3.9 66.78 0.84 16.85 0.81 17.89 1.09 21.95 0.66 10.75 0.80 16.13 1.4 27.67 0.98 19.66 2.02 37.04 0.91 18.28 Average S.D. 1.002 0.18 20.04 3.62 Table IV-3 shows the PM-2.5 nitrate concentrations measured by the HEADS and F+UFPC. The reported HEADS nitrate concentrations represent the sum of nitrate collected on both Teflon and glass fiber filters. The collection efficiency of the F+UFPC was on the average 1.13 (±0.18), whereas the obtained concentration enrichment based on nitrate was 22.5 (±3.7). The somewhat higher efficiency and enrichment values than the ideal may be due to some uncertainty in the nitrate levels measured by means of the HEADS. This was due to the overall low nitrate levels that were measured indoors (e.g., less than 15% of the total mass concentrations), a rather surprising result, given the high particulate outdoor nitrate levels generally observed in Los Angeles. As the sampling flow rate of the F+UFPC was 120 L/min (e.g., 12 times higher than that of the HEADS), the F+UFPC nitrate concentration data are more robust. The generally low nitrate 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. content of the indoor aerosol also explains the reason for obtaining a perfect mass balance between HEADS and F+UFPC, although only the Teflon filter of the HEADS was weighed. While some volatilization loss of ammonium nitrate from the Teflon filter of the HEADS sampler occurred (ranging from about 10% to as high as 55%), the very low nitrate levels did not contribute significantly to the overall mass concentrations and therefore did not affect the HEADS-to-F+UFPC comparison based on mass. Table IV-3 Comparisons between nitrate concentrations determined using the F+UFPC and the HEADS using indoor air as the test aerosol.______________________________ HEADS (lig/m3 ) F+UFPC (pg/m3 ) Collection Efficiency F+UFPC Enrichment factor 1.1 28.22 1.28 25.65 0.49 14.73 1.5 30.0 0.44 10.88 1.22 24.5 0.55 11.71 1.06 21.2 0.41 7.61 0.95 19.16 0.48 8.78 0.91 18.23 0.62 12.37 0.99 19.80 0.48 11.22 1.16 23.30 0.29 6.33 1.06 21.37 0.45 6.72 0.74 14.93 Average 1.13 22.58 S.D. 0.18 3.71 The results of Table IV-3 show conclusively that concentration enrichment though the F+UFPC occurs without any measurable loss of particulate nitrate, despite heating and saturation of the aerosol to about 35 °C. Ammonium nitrate dissociates to ammonia and nitric acid, with its dissociation constant increasing exponentially with temperature. However, the dissociation constant decreases sharply as the relative humidity (RH) exceeds 90-95% (Stelson and Seinfeld, 1982). For example, even at 50°C 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and at RH=95%, the dissociation constant of ammonium nitrate is approximately 7 ppb2, which is the value of the dissociation constant at 18°C. Therefore, despite the increase in the aerosol temperature (which would have increased exponentially the value of the dissociation constant), saturation of the aerosol seems to prevent nitrate losses due to volatilization. Table IV-4 Comparisons between elemental and organic carbon concentrations determined by means of the F+UFPC and the MOUDI using indoor air as the test aerosol. ______________ Elemental Carbon__________________ Organic Carbon___________ MOUDI F+UFPC Collection Enrichment MOUDI F+UFPC Collection Enrichmen (pg/m3 ) (pg/m3 ) Efficiency factor (pg/m3 ) (pg/m3 ) Efficiency t factor F+UFPC F+UFPC 0.38 7.35 0.95 19.0 5.4 120.1 1.11 22.23 0.77 18.2 1.16 23.4 7.8 228.7 1.42 29.0 0.85 15.9 0.93 18.5 8.3 206.6 1.25 25.02 0.48 12.14 1.25 25.2 6 .2 165.1 1.32 26.56 0.55 7.28 0 .6 6 13.20 10.7 76.86 0.36 7.2 0.89 19.00 1.06 21.3 12.2 287.3 1.17 23.48 0.57 14.26 1.2 24.1 9.1 214.0 1.17 23.44 0.84 18.1 1.08 2 1 .6 9.9 259.7 1.30 26.10 0.72 14.2 0.99 19.7 6 .6 112.3 0.85 17.0 0.61 10.6 0.87 17.4 6.9 116.1 0.84 16.9 Average 1.02 20.4 1.08 21.68 S.D. 0.16 3.28 0.32 6.43 Results from the comparisons between the indoor PM2.5 elemental and organic carbon concentrations determined using the F+UFPC and those by means of the MOUDI are summarized in Table IV-4. Similar to the results based on mass, sulfate and nitrate concentrations, excellent agreement was obtained between the F+UFPC and MOUDI EC concentrations, with the average concentration enrichment factor being 20.4 (+3.3) and the average collection efficiency of the F+UFPC being 1.02 (± 0.16). Good agreement was also obtained between the F+UFPC and MOUDI organic carbon (OC) 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concentrations. The overall concentration enrichment factor was 21.6 (± 6.4) and the average collection efficiency of the F+UFPC was 1.08 (± 0.32). It should be noted that the OC concentrations determined by means of either of the two samplers may be overestimated due to adsorption of gas-phase OC on the quartz filters or underestimated due to evaporation of volatile organic compounds from the quartz filters during sampling (Eatough et al., 1993). Positive sampling artifacts (i.e., adsorption) should be more pronounced in the MOUDI than the F+UFPC because of the higher MOUDI flow rate (i.e., 30 L/min) compared to that of the minor flow of the F+UFPC (i.e., 6 L/min). Moreover, negative sampling artifacts (i.e., volatilization) would also be more pronounced in the MOUDI than the F+UFPC data. This is because of the higher flow rate and smaller size filter of MOUDI (3.7 cm) compared to those of the F+UFPC, both of which result in a higher pressure drop across its filter. The enrichment in concentration would also tend to reduce evaporative losses from the quartz filter of the F+UFPC. Recent studies showed that that nitrate losses from Teflon filter media could be virtually eliminated by placing the sampler downstream of a particle concentrator (Chang et al., 1999). The uncertainties introduced by the aforementioned artifacts may explain the somewhat higher standard deviation value obtained for the OC-based sampler comparison compared to those for the other species. The good overall agreement between the F+UFPC and MOUDI, however, suggests that these artifacts either negate each other or may not be significant under the specific conditions at which the experiments were conducted. 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IV-5 Conclusions This study presented the development of prototype portable Concentrators, capable of enriching the concentration of particles in the range of 0-10 pm by a factor up to 40, depending on the output flow rate. These systems are compact in size, so that they can be easily transported in various locations in order to conduct primarily animal inhalation studies to concentrated PM, as these studies require lower output flow rates. The modular design of these concentrators, however, makes them readily adaptable to accommodate higher output flow rates that are desirable in conducting human exposure studies. This can be easily achieved by placing several single-nozzle virtual impactors in parallel. Coarse PM (2.5-10 pm) are concentrated in a single-stage, round nozzle virtual impactor, operating at an intake flow of 120 L/min. Fine and ultrafine PM (F+UFP, smaller than 2.5 pm) are concentrated by first removing larger particles by impaction and then growing the remaining particles via supersaturation to super-micrometer droplets. The droplets are then concentrated using the same Coarse Particle Concentrator. Concentrated ultrafine and fine particles are returned to their original size by passing through a diffusion dryer using silica gel. The experimental characterization of the F+UFPC showed clearly that the concentration enrichment does not depend on particle size or chemical composition. Volatile species such as ammonium nitrate are preserved through the concentration enrichment process under the laboratory conditions used in this study. Excellent 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. agreement was found between mass, sulfate and nitrate concentrations measured by means of the F+UFPC and a collocated HEADS. Very good agreement was also found between the elemental and organic carbon concentrations determined using the F+UFPC and the MOUDI. Furthermore, the concentration enrichment based on particle counts showed clearly that no particle coagulation occurs during the enrichment process, for any of the three minor-to-total flow configurations tested. The ability of the F+UFPC to enrich the concentrations of all particles in the fine mode (including its ultrafine particle component) enables inhalation toxicologists to conduct exposures to any selected sub-range of PM-2.5. For example, previous studies in California showed the presence of two sub-modes within the accumulation mode of ambient PM (Hering et al., 1997; John et al., 1990); one mode peaks at around 0.2 pm (consisting mainly of gas-to-particle reaction products, such as carbonaceous PM) and the other peaks at about 0.7 pm (mainly associated with hygroscopic PM such as ammonium sulfate and nitrate). Both modes have a geometric standard deviation of about 2 (John et al., 1990). By placing a conventional impactor upstream of the F+UFPC having a 0.3 pm cutpoint, inhalation studies could be conducted to ultrafine PM plus the elemental and organic carbon content of the accumulation mode, but without the majority of its sulfate and nitrate constituents. Similarly, a 0.15 pm conventional impactor would remove all but ultrafine PM from the air-sample, thereby resulting in an Ultrafine Particle Concentrator. References American Thoracic Society (ATS), Committee of the Environmental and Occupational Health Assembly; Bascom, R., Chair. (1996) State of the art. Health effects of outdoor air pollution. Am. J. Respir. Crit. Care Med., 153:3-50. 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Burton, R.M., Wilson, W.E., Suh, H.H. and Koutrakis, P. (1996) Spatial variation in particulate concentration within the metropolitan Philadelphia. Envir. Sci. Technol., 30, 400-407. Chang, M.C., Sioutas, C., Kim, S., Gong, H., and Linn, W. (2000) Reduction of nitrate losses from filter and impactor samplers by means of concentration enrichment. Atmos. Environ., 34,85-98. Chen, D. and Pui, D.Y.H. (1995) J. Aerosol Sci., 26,563-574. Clarke, R.W., Catalano, P., Gazula, G., Sioutas, C., Ferguson, S.T., Koutrakis, P., Godleski, J.J. (1999) Inhalation of concentrated ambient particles (CAPS) induces pulmonary alterations in normal and chronic bronchitic rats. Inhal. Toxicol., 11,101- 120. Eatough D.J., Wadsorth A., Eatough D.A., Crawford J.W., Hansen L.D., and Lewis E.D. (1993) A multiple system, multi-channel diffusion denuder sampler for the determination of fine particulate organic material in the atmosphere. Atmos. Environ. 27, 12123-1219. Environmental Protection Agency (1996) Air Quality Criteria for Particulate Matter. EPA- 600/P-95/001af, Office of Research and Development, Washington. Fung, K. (1990) Particulate Carbon Speciation by M n02 Oxidation. Aerosol Sci. Technol. 12,122-127 Godleski, J., Sioutas, C., Katler, M., and Koutrakis, P. (1996) Death from inhalation of Concentrated Ambient Air Particles in animal models of pulmonary disease. Resp. And Crit. Care Med. 155(4), A246. Gordon T., Gerber H., Fang, C.P., Chen L.C. (1999) A centrifugal particle concentrator for use in inhalation toxicology. Inhal. Toxicol. 11,71-87. Hering, S.V., Eldering, A. and Seinfeld, J.H. (1997) Bimodal characteristics of accumulation mode aerosol mass distributions in Southern California. Atmos. Environ., 31(1), 1-11. John W., Wall S.M., Ondo J.L. and Winklmayr W. (1990) Modes in the size distributions of atmospheric inorganic aerosol. Atmos. Environ. 22, 1627-1635 Kleinman, M.T., Mautz, W.J., Hyde, D.M. and Sioutas, C. (1998) Cardiopulmonary effects of inhaled ultrafine particles in aged rats. Presented at the American Association for Aerosol Research, June 1998 Koutrakis, P., Sioutas, C., Ferguson S.T., and Wolfson, J.M. (1993) Development and evaluation of a glass honeycomb denuder/filter pack system to collect atmospheric gases and particles. Environ. Sci. Technol., 27(12), 2497-2501. 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Marple, V.A., Rubow, K.L. and Behm, S. (1991) A Microorifice uniform deposit impactor (MOUDI): description, calibration, and use. Aerosol Sci. Technol., 14,434-446. Muyshondt, A., McFarland, A.R. and Anand, N.K. (1996) Deposition of aerosol particles in contraction fittings. Aerosol Sci. Technol., 2 4 ,205-216. Sioutas, C., Koutrakis, P., and Burton, R.M. (1995) A technique to expose animals to concentrated fine ambient aerosols, Environ. Health Perspect. 103,172-177. Sioutas, C., Koutrakis, P., Ferguson, S.T. and Burton, R.M. (1995) Development and evaluation of a prototype ambient particle Concentrator for inhalation exposure studies. Inhal. Toxicol,!, 633-644. Sioutas, C., Koutrakis, P., Godleski, J., Ferguson, S.T., Kim, C.S. and Burton, R.M. (1997) Harvard/EPA ambient fine particle concentrators for human and animal exposures. J Aerosol Sci., 28(6), 1057-1077. Sioutas, C., Kim, S., and Chang, M. (1999) Development and evaluation of a prototype ultrafine particle Concentrator. J.Aerosol Sci, 30(8), 1001-1012. Stelson A. W. and Seinfeld J.H. (1982) Relative humidity and temperature dependence of the ammonium nitrate dissociation constant. Atmos. Environ. 16, 983-992. Spengler, J.D., and Thurston, G.D. (1983) Mass and elemental composition of fine and coarse particles in six U.S. cities. JAPCA 33,1162-1171. Whitby, K.T. and Svendrup, G.M. (1980) California Aerosols: Their Physical and Chemical Characteristics, Adv. Environ. Sci. Technol. 10,477. Vedal, S. (1997) Ambient particles and health: lines that divide. J. Air Waste Manage. Assoc., 47,551-581. 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter V Versatile Aerosol Concentration Enrichment System (VACES) for Simultaneous in vivo and in vitro Evaluation of Toxic Effects of Ultrafine, Fine and Coarse Ambient Particles Seongheon Kim*, Peter Jaques*, Mingchih Chang*, and Constantinos Sioutas**. * Department of Civil and Environmental Engineering, University of Southern California, 3620 South Vermont Avenue, Los Angeles, CA 90089 * Department of Environmental Health, UCLA, 650 Charles E. Young Drive South, Los Angeles, California 90095 Manuscript prepared for publication in Journal of Aerosol Science July, 2000 1 0 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V-l Abstract This paper presents the development and evaluation of a versatile aerosol concentration enrichment system. This system is capable of concentrating distinctive sub modes of ambient aerosols including coarse, accumulation plus ultrafine, and ultrafine modes. The concentrated particles can be provided either as suspended in air or collected into water, so that both in vivo and vitro health studies can be conducted concurrently. Individual components of the system were characterized in laboratory experiments. Field tests were conducted to evaluate the performance of the entire system. Size selective multi slit impactors for ultrafine particles were designed and characterized using both SMPS (number counts) and DataRAM (mass concentration). The 50 % cutpoint of multi slit impactor was determined to be at 0.2 pm with a low-pressure drop of 10 inches of water. The collection efficiency of the BioSampler is close to 100% for particles larger than 2 pm at adjusted flow rate of 5 lmin"1 , regardless of liquid volume in the reservoir. The pressure drop across the BioSampler was decreased to 18 inches of water at this reduced flow rate of 5 lmin"1 . The enrichment factors for three minor flows of 7, 10, and 20 lmin'1 and for an intake flow rate of 220 lmin'1 were close to the ideal values. Results from the field tests indicated that a perfect agreement in total mass, nitrate, and sulfate concentrations for the coarse mode was obtained between the versatile aerosol concentration enrichment system and the reference MOUDI. The enrichment factor for the fine PM sulfate was in very good agreement with the ideal value, while the enrichment factor for the total mass of fine plus ultrafine PM became close to the ideal value after correcting the MOUDI concentrations for nitrate losses. The size distribution of ambient ultrafine particles was not distorted during the process of concentration enrichment, with number based enrichment factor of 21(±1.4). 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V-2 Introduction Despite the abundant epidemiological evidence associating ambient particulate pollution with adverse health effects in humans (Bates and Sizto 1989; Pope et al, 1995; Thurston et al., 1994), there are still uncertainties regarding physiochemical properties of particles that affect health risk and underlying pathophysiological mechanisms (Vedal, 1997). In many cases, this apparent lack of agreement between epidemiological and toxicological studies has been attributed to the inability of controlled laboratory investigations to deliver a sufficient dose of actual ambient aerosols, which might support the epidemiological findings. Ambient particle concentrations are usually too low to introduce measurable acute effects, while artificially generated particles cannot represent all the potentially toxic components present in ambient particles in the form of either particulate or adsorbed gases. Ambient particulate matter smaller than 10 pm in aerodynamic diameter (PMi0 ) consists of at least three discrete size modes (i.e., ultrafine, fine and coarse, covering the size ranges of 0-0.1, 0.1-2.5 and 2.5-10 pm, respectively). Consideration has been given to the role of particle size mode in adverse health impacts. The fine PM mode (PM2. 5) is considered to have a higher toxic potential than the coarse PM mode. Recent studies conducted in Mexico City (Loomis et al., 1999) and the Netherlands (Kleinman et al, 2 0 0 0), however, suggested stronger effects of coarse mode than of fine particles on mortality, indicating that coarse mode cannot be discarded as predictor of daily mortality in certain areas. In addition, there is a growing recent epidemiological database (Heyder et al., 1996; Peters et al., 1997) suggesting that the fraction of ultrafine particles may be 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of importance. These findings imply that ambient ultrafine particles, which dominate the particle number concentration, may be an important PM property in terms of toxicity. The recent development of ambient Particle Concentrators (Sioutas et al., 1995 and 1997; Kim et al., 2000a; Gordon et al., 1999) has made it possible to perform laboratory exposures with "real-life" ambient aerosols at increased, but still realistic concentrations. Initial results suggest greatly increased toxic responses (as compared to that with artificial particles) and suggest that this type of exposure system may provide a useful method for assessing the health effects of ambient particles and for identifying specific risk factors and the means of controlling those factors (Godleski et al., 1996; Gong et al., 2000). Nevertheless, currently available particle concentrators focus on concentrating only one or two mode(s) of ambient particulate matter. The concentration enrichment of the Harvard Fine Particle Concentrator (Sioutas et al, 1995) depends on particle size, with the 0.4-1 pm particles of the accumulation mode being concentrated far more effectively than particles smaller than 0.4 pm (Sioutas et al., 1997). Consequently, this system cannot be used to increase the concentration of particles below 0.15 pm (i.e., the ultrafine particles). Furthermore, under certain conditions, the performance of concentrators becomes unstable during operation. Typical indications of instabilities are abrupt increases in pressure drop across the slit nozzles of the virtual impactors, followed by a sharp decrease in the concentration enrichment factor. The effects of parameters such as ambient relative humidity, dew point temperature, ambient PM2.5 mass concentration, ambient PM2.5 mass median diameter (MMD), and total pressure drop per unit time across the Concentrator on the overall concentration enrichment achieved by the Harvard Fine Particle Concentrator are investigated in details by Kim et al. (2000b). I l l Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Another type of concentrator using centrifugal forces achieves restricted particle concentration enrichment (mostly in the 0.5-1.0 |im aerodynamic size range) due to coarse particle loss by impaction and diffusion loss of ultrafine particles (Gordon et al., 1999). Recently developed portable concentrators make use of condensational particle growth via supersaturation, followed by virtual impaction, to concentrate fine and ultrafine PM (Kim et al., 2000a). These technologies are attractive due to their size and ability to concentrate the fine PM mode, including its ultrafine fraction. However, these technologies can only provide concentrated ambient particles in either the coarse or fine plus ultrafine modes, without making provisions for concurrent exposures to coarse, fine and ultrafine PM separately. Furthermore, these technologies operate at relatively moderate flow rates (the intake flow is 110 lmin'1 and the concentrated flow is typically 5 lmin'1 ), thereby allowing inhalation studies using a small number of animals. The purpose of this study is to develop and evaluate a versatile aerosol concentration enrichment system (VACES) capable of conducting simultaneously in-vivo and in-vitro evaluations of toxicity of ambient particles in the coarse, fine and ultrafine size modes. This versatile system is a further extension of previous work by Kim et al. (2000a), strengthened by major new features including: 1) The ability of concentrating ultrafine particles only, and supplying them to an exposure chamber at virtually atmospheric pressure (i.e., 0.99 atmospheres) 2) The ability to allow concurrent animal exposures to coarse, fine and ultrafine particles. 3) If exposures to one PM mode only are desirable, these technologies can operate at three times higher flow rates than the system described by Kim et al. (2000a), thereby concentrating 330 lmin'1 of ambient PM to a flow rate as low as 10 lmin'1 . 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This feature makes it possible to use more animals in an inhalation study, hence increase the confidence level in the observed outcomes. 4) The capability of collecting concurrently very high quantities (i.e., on the order of mg) of coarse, fine and ultrafine PM in a very small liquid volume (i.e., 4-10 ml). The resulting highly concentrated suspensions can be then used for in vitro tests to evaluate the relative toxicity of ambient PM, collected simultaneously in a given location. This paper discusses the development and the laboratory evaluation of the individual components of the VACES. Also, the characterization of the entire system in the field and the preservation of particle mass, number, and chemical species through the concentration enrichment process are presented. V-3 Methods V-3-1 Description o f the Portable Coarse, Fine and Ultrafine PM concentrators Figures V-la, b show a schematic of two different configurations of the new concentrators, which we have tentatively named Versatile Aerosol Concentration Enrichment Systems (VACES). Figure V -la shows the configuration used for inhalation (in vivo) exposures, whereas Figure V -lb shows another version of the same system for in vitro toxicity studies. The VACES consists of three parallel sampling lines. In each line, ambient coarse, fine and ultrafine aerosols are drawn at the flow rate of 110 lmin"1 . Coarse PMs are drawn through a round nozzle, single-stage virtual impactor, having a 50% cutpoint at 1.5 pm. The performance of these virtual impactors is described in greater detail by Kim et al. 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (2000a). Coarse particle in this sampling line can be concentrated by as much as a factor of 35, and supplied to the exposure chamber at a flow rate of 3.3 lm in1 . The other two sampling lines of the VACES consist of identical components, with the only exception of the cutpoints of the impactors through which the samples are drawn prior to passing through the saturator. In the line concentrating fine plus ultrafine PM, air samples are first drawn through a single slit impactor, having a 50% cutpoint at 2.5 |im at a flow rate of 110 lmin'1 . The impactor’s acceleration slit is 0.2 cm wide and 5 cm long. In order to remove all but the ultrafine PM, particles in the third sampling line of the VACES are drawn through a multi-nozzle, high volume conventional impactor with a design 0.15 pm cut-off size at a flow rate of 110 lmin"1 . Separation of these particles is accomplished under a very low-pressure drop (i.e., 10 inches of H2 0). This is a very important feature of these new concentrators, since inhalation studies cannot be conducted under a substantial vacuum. The impactor consists of 6 slit-shaped nozzles in parallel, each 5 cm long and 0.015 cm wide. A 6 x 0.2 cm quartz fiber strip is placed underneath each acceleration nozzle, and at a distance of 0.04 cm. The strips are coated with mineral oil and serve as bounce-free impaction substrates for collecting particles above 0.2 pm in aerodynamic diameter. It should be noted that concentration of ultrafine particles is optional. Without the use of the 0.2 pm impactor, the VACES can also deliver fine and ultrafine PM at 10 lmin'1 , enriched in concentration by a factor of 22. After the impactor pre-separators, the aerosol in both the fine and ultrafine lines is drawn through a stainless steel container used as the aerosol saturator. The container has a capacity of 10 liters and is used to mix the aerosol with warm, distilled deionized water vapor at a temperature of about 32 (± 2) degrees C. The stainless steel container is 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. 0.15 jim cutpoint impactor 2.5 pm cutpoint impactor distilled c (ionized w ater 110 LPM VI-1 5 LPM fine PM diffusion dryer ultrafine PM 110 LPM ---------------» VI-2 —» diffusion dryer Cooler 5 LPM 110 LPM 110 LPM Condensation Particle Counter (CPC) VI-3 coarse PM &lpm _________________ J 15 LPM 2067 Gast pump Saturator (20-L w ater bath; 0.5 kW) Heating Bath animal chamber with separate compartments 315 LPM Figure V-1a Versatile Aerosol Concentration Enrichment System (VACES) for in vivo studies: VI: virtual impactors. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. 0.15 pm cutpoint impactor 2.5 pm cutpoint impactor Cooler 110 LPM distilled d < ionized water 110 LPM 110 LPM 110 LPM Saturator (20-L water bath; 0.5 kW) Heating Bath VI-1 ultrafine PM BioSampler I * 5 LPM VI-2 fine + t j — p BioSampler ► 5 LPM VI-3 coarse PM — ► BioSampler 5 LPM i 2067 Gast pump 315 LPM V I: virtual impactors. Figure V-1b Versatile Aerosol Concentration Enrichment System (VACES) for in vitro studies. Os placed inside a heating bath (VWR Scientific, Model 1024), with a heating power of 0.5 kW. Subsequently, the saturated aerosol is drawn through a cooler, which is a simple icebox with two aluminum tubes (2.2 cm in diameter and 80 cm long) through it. In each cooler, the saturated and warm air is cooled by about 9-10 degrees C. Hence the produced supersaturation causes all particles to grow to about 2.5-2.6 pm droplets. The grown droplets are subsequently drawn through two identical virtual impactors. Each virtual impactor separates particles to two different size ranges, approximately above and below 1.5 pm, respectively. These virtual impactors are also identical in design to that used for concentration of coarse ambient particles. The virtual impactors are made of anodized aluminum. The grown fine and ultrafine particles are drawn into the minor flow of virtual impactor (which can be as small as 3 lmin'1 ), and thereby become concentrated by a factor up to 40. The concentrated droplets are then drawn through a Diffusion Dryer (TSI Model 3062, TSI Inc., St. Paul, MN), placed immediately downstream of the collection nozzle of each virtual impactor. The diffusion dryer is used to remove the excess moisture around the particles and return these grown particles to their original size. Operating at a maximum flow rate of 10 lmin'1 , each diffusion dryer reduces the relative humidity of the incoming aerosol from 100% to 50%, thereby returning the grown particles to their original size. Figure V -lb shows the alternative configuration of the VACES, used for simultaneous coarse, fine and ultrafine PM collection for in vitro toxicity tests. In this configuration, instead of passing through a diffusion dryer, the concentrated coarse, fine and ultrafine particles in each parallel sampling line are drawn through a liquid impinger 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (BioSampler, SKC West Inc., Fullerton, CA). Traditionally, particle collection for in vitro tests has been conducted by using collection substrates such as filters or impactors. The collected particles are subsequently extracted from the substrates and administered into the in vitro culture either directly or after lyophilization of the solvent. This process suffers from several shortcomings, including imperfect particle extraction from the substrate, potential loss of toxicologically important semi-volatile PM constituents as well as loss of biologically active agents of airborne PM. In addition, a recent study by Dick et al. (2000) showed that components of filters used to collect particles could contaminate the preparation and interfere with biological investigation. Particle collection using liquid impingers has been shown to be advantageous over the traditional filtration or impaction methods for collection of airborne particles because impingers are not easily overloaded (Willeke et al., 1998) and because impingement eliminates the need for elaborate extraction procedures (Zucker et al., 2000). Operating at its nominal flow rate of 12.5 lmin'1 , the BioSampler has collection efficiency close to 100% for particles larger than about 1.5 pm. For particles smaller than 1.0 pm in aerodynamic diameter, the collection efficiency decreases sharply to less than 50% (Willeke et al., 1998). Operating in conjunction with the VACES, however, the BioSampler can collect any of the PM size ranges with 100% efficiency and at sampling flow rate that is at least one order of magnitude higher than its nominal operating flow rate. Thus, the condensation growth of even ultrafine PM to super micrometer particles enables effective trapping of these particles by the impinger and allows us to “concentrate” large volumes of ambient PM into a very small solution (i.e., order of 4-10 ml). Furthermore, the ability to collect large volumes of particles directly 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. into a small volume of any solution is a particularly attractive feature when intratracheal instillation is used as the method to conduct particle toxicity tests. All three major flows of the parallel virtual impactors are drawn by a single rotary vane pump (Gast model 2067, Gast Manufacturing, Cerritos, CA). V-3-2 Laboratory Characterization o f the VACES V-3-2-1 Characterization o f the 0.2 pon Low Pressure Drop multi Slit impactor The collection efficiency of the multi-slit nozzle 0.2 pm cutpoint impactor was estimated using an indoor air as the test aerosol. For particles in the size range of 0.015 to 0.5 pm, penetration was determined by measuring the aerosol concentrations upstream and downstream of the impactor by means of the Scanning Mobility Particle Sizer (SMPS Model 3096,TSI Inc., St. Paul, MN). The SMPS sampled 0.2 lm in1 of the total flow rate of 110 lmin'1 through the impactor. Number concentration of ambient aerosols was measured with and without the block of the acceleration slits of the impactor to account for possible diffusion loss of ultrafine particles through the sampling lines connecting to the SMPS. Particle size was selected in the interval of 0.02-0.5 pm by adjusting manually the voltage to the Differential Mobility Analyzer of SMPS and measuring the particle counts upstream and downstream of the 0.2 pm cutpoint impactor. In addition to the SMPS, the DataRAM was used to evaluate the collection efficiency of the multi-slit impactor for particles in the 0.2 to 1.0 pm range, using artificially generated monodisperse PSL particles (Bangs Laboratories Inc.,) in a constant output nebulizer (HEART, VORTRAN Medical Technology, Inc., Sacramento, CA). The generated aerosols passed through Po-210 neutralizes and were mixed with filtered 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. air prior to passing through the slit impactor. The mass concentrations of the monodisperse aerosols upstream and downstream of the impactor were measured by means of a Nephelometer (DataRAM, MIE, Inc., Billerica, MA). For each test, repeated measurements of the concentrations upstream and downstream of the impactor were taken. The concentrations of the generated aerosols were in the range of 100-400 pg/m3 , thus several orders of magnitude higher than the noise level of the DataRAM (i.e., about 1-5 pg/m3 ). As a nephelometer, the response of the DataRAM is dependent on particle size (Sioutas et al., 2000). However, as particle penetration was defined as the ratio of the downstream to the upstream concentrations, it should not be dependent on particle size for the same particle size. Finally, limited field tests were conducted in which the ambient aerosol concentrations measured by the 0.2 pm cutpoint impactor was compared to that measured by means of the Microorifice Uniform Deposition Impactor (MOUDI, MSP Corp., Minneapolis, MN), which was as a reference sampler. A 4.7 cm Teflon filter (2 pm pore, Gelman Science, Ann Arbor, MI) was placed immediately downstream of the multi-slit impactor, which was operated at a flow rate of 110 lmin'1 . The MOUDI was placed at a distance of 1 m from the impactor and sampled at 30 lmin"1 . Ambient particle smaller than 0.18 pm in aerodynamic diameter were collected on a 3.7 cm Teflon filter following the last impaction stage of the MOUDI. Both MOUDI and multi-slit nozzle impactor Teflon filters were weighed before and after each test in a Mettler Microbalance (MT5, Mettler-Toledo, Inc., Highstown, NJ) under the controlled relative humidity (40-45%) and temperature (22-24 °C) conditions in order to determine the mass concentrations. 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V-3-2-2 Characterization o f the BioSampler At the standard operation flow rate of 12.5 lmin'1 , the pressure drop across the BioSampler is close to 0.5 atm, which has been shown to cause excessive evaporation of liquid collection media such as water (Lin et al., 2000). It is also expected that under these sampling conditions, excessive loses of semi-volatile components of ambient particles would occur. In order to reduce the pressure drop across the BioSampler used in conjunction with the VACES virtual impactors, a flow rate of 5 lmin'1 was used instead. The decrease in flow rate was expected to increase the cutpoint of the BioSampler. However, as most of fine and ultrafine PM is grown to 2.5-2.1 pm via supersaturation in the VACES, our primary concern was to ensure that particles in that size range are efficiently collected by the modified BioSampler. Another modification of the BioSamplers used in conjunctions with the VACES was the amount of water used in its reservoir collect the impinging particles. In its nominal configuration, 20 ml of liquid are required in the biosampler reservoir. However, from the standpoint of toxicological studies, it is highly desirable to maximize the concentration of the collected ambient particles in the liquid medium of the BioSampler. We thus investigated the effect of different volumes of water on the collection efficiency of BioSampler at the reduced flow rate of 5 1 min'1 . We specifically tested the BioSampler using water volumes of 2, 4, 10 and 20 ml, respectively. For each liquid volume, the collection efficiency for particles in the range of 0.5-5 pm was determined by measuring the upstream and downstream the BioSampler monodisperse aerosol concentrations using the DataRAM, as described above. At 5 lmin"1 , the pressure drop across the BioSampler was approximately 18 inches of H2 0 . The exhaust of the 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DataRAM pump was returned downstream of the BioSampler in order to avoid sampling biases, which might occur when this instruments samples under a vacuum. V-3-2-3 Laboratory Characterization o f the Fine and Ultrafine Concentrators o f the VACES As the coarse particle concentrator component of the VACES has been already developed and described elsewhere (Kim et al., 2000a), the laboratory tests focused on the experimental characterization of the fine and ultrafine concentrators of the VACES. It should be noted that the use of the 0.2 pm impactor to remove all but ultrafine particles is optional. The VACES can also be used to concentrate fine PM including the ultrafine fraction from 220 lmin'1 to a flow as small as 7 lmin'1 . Thus experiments were conducted at a sampling flow of 220 lmin'1 as a worst case scenario. The experimental characterization of the VACES was conducted using laboratory monodisperse particles as well as real-life ambient particles as the test aerosols. Monodisperse aerosols were generated by atomizing suspensions of ultrafine and fine particles using a constant output HEART nebulizer (VORTRAN Medical, Inc., Sacramento, CA). Different types of suspensions were used, including monodisperse PSL fluorescent latex particles (size range 0.05 - 2 pm; Polysciences, Inc., Warrington, PA) as well as monodisperse silica bead (0.36 pm; Bangs Laboratories, Inc., Carmel, IN). In addition, aqueous solutions of ammonium sulfate and ammonium nitrate were atomized. Finally, indoor aerosol was also used as test aerosol. The size distributions of the polydisperse aerosols were determined using SMPS. 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The generated aerosols were dried and neutralized and were drawn to the saturator at 220 lmin'1 . The aerosol was mixed and saturated with water vapor at about 35 °C, and drawn through two condenser tubes at 110 lmin"1 each. The temperature of the aerosol exiting the condenser was about 24 (± 1)°C. The grown droplets were subsequently drawn through the two virtual impactors. Three different minor flow rates were tested, 7, 10, and 21 lmin'1 , respectively (corresponding to theoretical enrichment factors of 30, 22, and 10.5, respectively). The TSI Condensation Particle Counter (CPC 3022, TSI, Inc., St. Paul, MN) was connected immediately upstream of the saturator and downstream of the diffusion drier (as shown in Figure V-la) to measure the number concentrations of the original and concentrated aerosols. For each particle size, concentration enrichment was defined as the ratio of the concentration measured downstream of the diffusion dryer to that measures upstream of the saturator. V-3-3 Field Evaluation o f the VACES In addition to laboratory experiments, the performance of the VACES was evaluated in a field study, conducted outdoors in the facilities of Rancho Los Amigos National Rehabilitation Center in Downey, CA. The coarse, fine and ultrafine concentrations of the VACES were compared to those measured by means of a collocated MOUDI, operating at 30 lmin'1 , which was used as a reference sampler. The MOUDI was used as a reference sampler because of its higher sampling flow rate, which allowed us to reduce the sampling time to 4-5 hours. It should be noted, however, that the MOUDI is not a reference sampler for semi-volatile species, such as ammonium nitrate, and losses 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of these compounds under conditions of high temperature and low relative humidity (Chang et al, 2000). Instead of using all of the stages of the MOUDI, only those having cutpoints of 2.5 and 0.18 pm were used. Thus, the first MOUDI stages (2.5-10 pm) was used as a reference sampler for coarse ambient particles, the second stage (0.18-2.5 pm) was used as a reference for the ambient PM accumulation mode and the last stage was used to determine ambient ultrafine PM concentrations. Particle size distributions based on the number counts were obtained upstream of the VACES and immediately downstream of the diffusion dryer of the VACES line sampling fine plus ultrafine PM by means of the SMPS. The MOUDI and VACES coarse PM concentrations were compared based on mass, nitrate, sulfate. The fine (accumulation plus ultrafine) MOUDI and VACES concentrations were compared based on mass, number, nitrate, sulfate. Finally, ultrafine concentrations were compared based on elemental and organic carbon (EC/OC), as these are the predominant ultrafine PM constituents in this location (Sioutas et al., 2000). Depending on the type of chemical analysis, 4.7 cm Teflon (Gelman Science, 2 pm pore, for mass, inorganic ions, metals) or quartz (Pallflex Corp., Putnam, CT, for elemental and organic carbon) filters were placed immediately downstream of the diffusion dryer of the fine and ultrafine concentrators of the VACES, and downstream of the minor flow of its coarse concentrator. In order to demonstrate that there is no difference in the particle collection process if the in vitro version of the VACES is used, in selected experiments the BioSamplers instead of the filters (preceded by diffusion dryers) were used to collect concentrated PM. For the samples collected by means of the BioSamplers, only the inorganic ions were determined, because of the logistical 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. difficulties associated with weighing (for mass) the BioSampler or analyzing its aqueous extract for EC and OC. The Teflon filters were weighed before and after each field tests in a Mettler 5 Microbalance (MT 5, Mettler-Toledo Inc., Highstown, NJ) under controlled relative humidity (e.g. 40-45%) and temperature (e.g., 22-24 degrees C) conditions, in order to determine the mass concentrations. Filters were weighed immediately at the end of each experiment as well as after a 24-hour equilibration time period. Laboratory and field blanks were used for quality assurance. Filters and filter blanks were weighed twice in order to increase precision. In case of a difference of more than 2 |lg between consecutive weightings, a filter was weighed for a third time. The Teflon filters were then analyzed by means of ion chromatography to determine the concentrations of particulate sulfate and nitrate. V-4 Results and discussion V-4-1 Characterization o f the 0.2 fjm cutpoint multi-slit impactor Figure V-2 shows the pressure drop across the multi-slit impactor as a function of flow rate. The pressure drop across the multi-slit impactor is about 10 inches of H2 0 at the standard flow rate of 110 lmin'1 . The ability of this impactor to remove all but ultrafine particles with a very low-pressure drop is a very important feature of the VACES, since inhalation health studies cannot be conducted under the substantial vacuum. 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a C /3 I l I £ C f l cl o n .a 0 ^ ft U H 3 2 & e s •a I a s c n 1 o 14 12 8 6 4 2 0 20 40 60 80 100 120 Flow rate through the multi-slit impactor (1/min) 140 Figure V-2 Pressure drop across the multi-slit impactor as a function of flow rate. The collection efficiency of multi-slit impactor, determined from the decrease of both number (SMPS) and mass (DataRAM) concentrations measured downstream of the impactor, is plotted as a function of particle aerodynamic diameter in Figure V-3. Error bars represent the standard deviation of the experimental results. 100% 80% - 60% - 40% -- A SMPS b DataRAM 20% 0% 0.01 0.1 Particle diameter (pm) Figure V-3 Removal efficiency of multi-slit impactor as a function of particle diameter. 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The particle collection efficiency curve obtained from data using the SMPS increases sharply starting at 0.1 pm and reaches the value of about 90 % at particles larger than 0.3 pm in aerodynamic diameter. The collection efficiency values obtained by means of the DataRAM are in a good agreement with those obtained by SMPS for the overlapping particle size range between 0.2 and 0.5 pm. Furthermore, there was no detectable particle removal below 0.1 pm. The comparison between the mass concentrations measured by multi-slit impactor and the reference MOUDI is shown in Table V -l. Despite the small number of data points, the mass concentrations of ultrafine particles obtained with the two samplers are in practically excellent agreement, with the average slit impactor-to-MOUDI ultrafine particle concentration being 1.05 (± 0.15). The agreement between the sampler is remarkable given that, even a small cutpoint difference in the 0.1-0.2 pm range might result in a substantial difference between masses collected by two different impactors for ultrafine particles. This is because this is the size range in which a sharp decreases in mass concentrations of ambient fine aerosols is observed (Whitby and Svendrup, 1980) and a small entrainment of accumulation mode particles, which would be the result of one of the samplers having a slightly higher cutpoint that the other, would result in a significantly higher mass concentration measured by that sampler. The low cut point of the high volume multi-slit impactor with the low pressure drop makes it possible for toxicologists to conduct health study on the ambient particles containing only ultrafine mode, therefore to verify hypothesis of particle size mode responsible for the adverse health effect. 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table V-l Comparison of ultrafine mass concentration after multi-slit impactor and MOUDI. Ambient ultrafine mass Sampling concentration3 (hr) time Ratio of mass concentrations between Multi-slit impactor and MOUDIb 1.89 pg/m3 5 131 % 2.78 pg/m3 5 89% 3.28 pg/m3 10 96% Average 105% Standard Deviation 23% a. Determined by reference MOUDI sampler. b. MOUDI Collected particles in the size below 0.18 pm. V-4-2 Characterization o f the BioSampler Figure V-4 shows the pressure drop across the BioSampler as a function of flow rate. The pressure drop at 5 lmin'1 is 18 inches of H2 0 (0.035 atm), which is substantially lower value than the 145 inches of H2 0 at the standard flow rate of 12.5 lmin"1 . As a consequence of this small pressure drop, less than 1ml of water volatilized after 6 hours of sampling ambient concentrated air at relative humidities ranging from 45 to 65%. By comparison, 80 % of 20 ml of water normally evaporates within 2 hours under reduced pressure at 12.5 lmin'1 (Willeke et al., 1998). Furthermore, the small pressure drop is essential in preserving semi-volatile species such as ammonium nitrate and a host of organic compounds would be more pronounced under the high pressure drop across the sampler (Zhang and McMurry, 1987). The collection efficiency of the BioSampler at 5 lm in1 is shown in Figure V-5 as a function of particle size for various amounts of water in the BioSampler reservoir. 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 150 O 120 U Flow rate through BioSamper (1/min) Figure V-4 Pressure drop across the BioSampler nozzle as a function of flow rate. Error bars represent the standard deviation of repeated tests. Data shown in Figure V-5 indicate that, for any particle size, there is no significant dependence of the collection efficiency of the BioSampler on the amount of water in its reservoir, at least for the range of 2-20 ml. Based on these results, even 4-5 ml in the BioSampler reservoir would ensure high particle collection efficiency, while maximizing the particle concentration in the aqueous suspension to be sued for in vitro tests. Water volume of 5 ml is also sufficient to ensure complete wetting of the bottom of the reservoir. The collection efficiency of the BioSampler is close to 100% for particles larger than 2 pm at a flow rate of 5 lmin'1 , regardless of liquid volume in the reservoir. For particles less than 1 pm in aerodynamic diameter, the collection efficiency decreases sharply to about 50% at 0.5 pm. No significant decrease in the collection efficiency due to particle bounce was observed up to about 5 pm of aerodynamic diameter. Figure V-5 also shows that the BioSampler collects fine and ultrafine particles that were grown to water droplets (depicted as 2.6 pm particles) more efficiently than dry PSL particles of similar size. 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. S — PSL over 2ml of water O — PSL over 10ml of water • water droplet over 5ml of water A PSL over 4ml of water X PSL over 20ml of water 100% 20% 0.1 Particle aerodynamic diameter (pm) Figure V-5 Particle collection efficiency of BioSampler as a function of particle aerodynamic diameter. Sampling flow rate: 5 LPM. Operating in conjunction with our prototype ultrafine, fine or coarse particle concentrator, the BioSampler can collect any of the PM size ranges with virtually 100% efficiency and at sampling flow rates that are 10-30 times higher than its nominal operating flow rate. Thus, the condensational growth of even ultrafine PM to super micrometer particles enables effective trapping of these particles by the impinger, thereby allowing concentration of large quantities of ambient PM into a small liquid volume (i.e., 5 ml). V-4-3 Laboratory Characterization o f the VACES Results from the laboratory evaluation of the VACES at three different minor flow rates are summarized in Figure V-6. In all three minor flow configurations, the major flow rate is adjusted to yield a total intake flow of 220 lmin"1 . Hence, the maximum obtainable concentration enrichment factors for each configuration are 31, 22, 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and 10.5, respectively. The concentration enrichment factors as a function of particle size, shown in Figure V-6 , have been obtained using monodisperse aerosols in the size range of 0.05 - 1.9 pm, except for the data corresponding to 0.025, 0.31, and 0.32 pm particles. The number mean diameter (NMD) of polydisperse aerosols were obtained from the count-based size distributions of ammonium sulfate, ammonium nitrate and indoor aerosols using the SMPS. ■ 2 0 1/min A 7 1/min 35 30 25 -£b~cr-o 20 15 10 5 0 0.1 1 10 0.01 Particle Diameter (pm) Figure V-6 Characterization of the Versatile Ambient Particle Concentrator for three minor flows. Total intake flow: 220 lm in1 . The enrichment factors averaged over the tested size range at minor flow rates of 7 lmin'1 , 10 lmin'1 , and 20 lmin' 1 are 30.1, 20.4, and 9.6, respectively, which are very close to the ideal values. In addition, hygroscopic ammonium sulfate and ammonium nitrate aerosols did not show any observable difference in the enrichment factors compared to the hydrophobic PSL particles. 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V-4-4 Field evaluation of the VACES In each of the sampling lines of the VACES, coarse, fine and ultrafine particles were concentrated from a flow of 110 to a minor flow of 5 lm in1 . Thus the ideal concentration enrichment factor for any chemical PM species should be 22. Results form these field tests are summarized in Tables V-2 ~ 8 and in Figure V-7. The first two columns in each Table show the ambient and concentrated PM concentrations measured by the MOUDI and the corresponding sampling line of the VACES. The third column shows the values of the concentration enrichment, as the ratio the VACES coarse, fine or ultrafine PM concentration to that of the MOUDI. Table V-2 Coarse Ambient Particle Mass Concentrations Determined with the MOUDI and the VACES. Ambient VACES Enrichment Factor (pg/m3 ) (pg/m3 ) 15.5 298.9 19.3 2 0 .6 333.3 16.2 2 2 .2 612.5 27.6 22.4 552.2 24.6 26.7 600.0 22.5 26.8 514.4 19.2 27.4 717.8 26.2 29.2 662.2 22.7 36.5 584.4 16.0 45.1 1097.8 24.3 Average 21.9 Standard Deviation 4.0 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. As the results in Table V-2 suggest, a virtually perfect mass balance was obtained between the MOUDI and the VACES for coarse particles based on mass. The average concentration enrichment of the VACES was 22.0 (± 4.0), thus very close to the ideal values. The data obtained for coarse particle sulfate and nitrate (shown in Tables V-3 and 4) also indicate a very good agreement between the VACES and MOUDI, with the concentration enrichment factors for sulfate and nitrate of 21.0 (± 6 .6) and 19.9 (± 4.3). Table V-3 Coarse Ambient Particle Sulfate Concentrations Determined with the MOUDI and the VACES. Ambient (pg/m3 ) 1.0 1.4 0 .8 1.4 2.1 VACES (pg/m3 ) 2 0 .8 2 2 .0 6 .6 44.0 58.5 Enrichment Factor 2 1 .6 15.2 8.1 31.7 28.5 Average 2 1 .0 Standard Deviation 9.6 Table V-4 Coarse Ambient Particle Nitrate Concentrations Determined with the MOUDI and the VACES. Ambient (pg/m3 ) VACES (pg/m3 ) Enrichment Factor 6 .2 0 83.85 13.51 6.43 135.87 21.13 3.71 57.80 15.58 6.78 155.77 22.98 5.41 118.28 2 1 .8 6 Average 19.01 Standard Deviation 4.2 No nitrate loss was assumed since both VACES and the high stage of MOUDI have negligible pressure drop (less than 1 inch of water) across the sampler, which is the 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. driving force of evaporation loss of semi-volatile species. Nevertheless, the overall VACES and MOUDI agreement for coarse particles should be considered excellent. Table V-5 shows the PM2.5 mass concentrations measured by the MOUDI and VACES. The overall concentration enrichment obtained for the fine PM mode is slightly higher (26.3 ±3.7) than the ideal value of 22. As it will be explained in subsequent paragraphs, this is due to some losses of volatile species, such as ammonium nitrate, from the MOUDI substrates. The experiments were conducted during the months of May and June 2000 in south central Los Angeles, with temperatures averaging to 32 (± 3) degrees C and low relative humidity values (i.e., 40% or less). These conditions have been shown to favor loss of ammonium nitrate from impactor samplers (Chang et al., 2000; Zhang and McMurry, 1987) due to the higher values of the dissociation constant of ammonium nitrate. For this temperature and humidity range, the study by Chang et al. (2000) showed that the MOUDI losses of amount nitrate average to about 40-60%. Furthermore, the study of Kim et al. (2000) showed conclusively that concentration enrichment through the smaller scale portable fine PM concentrators, which are similar in principle and design to those of the VACES, occurs without any measurable loss of particulate nitrate, despite heating and saturation of the aerosol to about 35 °C. Ammonium nitrate dissociates to ammonia and nitric acid, with its dissociation constant increasing exponentially with temperature. However, the dissociation constant decreases sharply as the relative humidity (RH) exceeds 90-95% (Stelson and Seinfeld, 1982). For example, even at 50°C and at RH=95%, the dissociation constant of ammonium nitrate is approximately 7 ppb2 , which is the value of the dissociation constant at 18°C. Therefore, despite the increase in the aerosol temperature (which would have increased 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. exponentially the value of the dissociation constant), saturation of the aerosol seems to prevent nitrate losses due to volatilization. Table V-5 PM 2.5 Ambient Particle Mass Concentrations (uncorrected and corrected for nitrate losses) Determined with the MOUDI and the VACES. Uncorrected Ambient Mass VACES Enrichment Factor Enrichment Ambient Mass concentration Mass Based on Nitrate- Factor Based on Concentration corrected for Concentrations corrected uncorrected mass (pg/m3 ) Nitrate Loss (pg/m3 ) concentrations concentrations 24.2 27.3 704.4 25.8 29.1 25.5 29.5 792.5 26.9 31.1 29.6 37.1 805.6 21.7 27.2 18.3 21.6 518.9 24.1 28.3 36.4 38.7 698.9 18.1 19.2 40.6 49.1 1153.3 23.5 28.5 27.4 34.9 666.7 19.1 24.3 28.2 30.3 560.0 18.5 19.9 29.2 32.0 741.1 23.2 25.4 24.4 23.8 633.3 26.6 25.95 13.3 15.5 370.0 23.9 27.8 17.8 18.6 518.3 27.9 29.1 36.9 759.4 20.6 26.9 653.9 24.3 31.6 944.4 29.9 average 23.6 26.3 SD 3.5 3.7 These conclusions are further supported by the results shown in Table V-6, where the fine PM nitrate concentrations measured by means of the MOUDI are compared to those measured by the VACES. The average concentration enrichment based on nitrate is 40.4 (± 20.3), roughly twice the value of the ideal concentration 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. enrichment. Given that nitrate losses depend significantly on parameters that vary quite randomly, such as temperature, humidity and overall particle concentration, the MOUDI- to-VACES agreement should be highly variable, which is indicated by the somewhat large standard deviation on the concentration enrichment values obtained for nitrate. By comparison, the concentration enrichment obtained for the non-volatile fine particulate sulfate (shown in Table V-7) was 19.6 (± 4.3) and thus practically identical to the ideal value of 22. The above results confirm that the disparity between the ideal and actual concentration enrichment factors based on nitrate is due to sampling artifacts of the MOUDI. Table V-6 PM2 .5 Ambient Particle Nitrate Concentrations Determined with the MOUDI and the VACES. MOUDI (fig/m3 ) VACES (jig/m3 ) Enrichment Factor VACES 1.3 72.4 56.8 Teflon filter 2.2 22.8 10.3 Teflon filter 2.6 108.3 41.7 Teflon filter 2.7 131.7 49.4 Teflon filter 2.7 140.1 51.9 Teflon filter 3.2 54.1 16.9 Teflon filter 5.6 275.9 48.9 Teflon filter 6.2 148.2 24.1 Teflon filter 2.5 118.5 47.1 BioSampler 3.8 311.5 82.7 BioSampler 5.3 159.3 30.2 BioSampler 8.6 483.4 56.4 BioSampler 9.3 373.4 40.2 BioSampler average 39.4 Standard Deviation 20.5 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table V-7 PM2.5 Ambient Particle Sulfate Concentrations Determined with the MOUDI and the VACES. Ambient VACES Enrichment Factor VACES (l-ig/m3 ) (|Xg/m3 ) 3.0 61.4 20.4 Teflon filter 5.5 72.4 13.2 Teflon filter 3.6 56.6 15.7 Teflon filter 9.5 150.9 15.9 Teflon filter 4.1 66.7 16.3 Teflon filter 5.8 102.8 17.7 Teflon filter 3.5 62.2 18.0 Teflon filter 2.2 52.6 23.7 Teflon filter 5.0 81.6 16.5 BioSampler 7.8 143.0 18.4 BioSampler 9.2 209.8 22.7 BioSampler 6.4 160.0 25.1 BioSampler 4.7 134.7 28.7 BioSampler Average 19.3 Standard Deviation 4.6 The MOUDI fine PM mass concentrations were corrected for nitrate losses as follows: PM = PM + 1 29 ( ^ ^ V A C E S — NO 3 I r m co n r lV 1 MOUDI ~ 1 \ 2 2 MOUDI > where N 0 3 vaces a n d N 0 3 M O udi are the nitrate concentrations measured by the V A C E S and MOUDI, respectively, and P M moudi is the total MOUDI fine PM mass concentration determined gravimetrically. The above equation assumes that all nitrates found in the 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fine particulate mode is associated with ammonium nitrate. The corrected values of the MOUDI mass concentrations are also shown in Table V-5, along with the adjusted concentration enrichment factor. The nitrate-adjusted concentration enrichment factor is 23.6 (±3.4), thus very close to the ideal enrichment value of 22. The results of Table V-5 also indicate that the overall impact of nitrate losses from the MOUDI substrates on the mass concentration determined by the MOUDI is rather small. This is because ammonium nitrate accounts on the average for 25-30% of the total PM2.5 mass concentration at Rancho Los Amigos (Sioutas et al., 2000). Thus, even if nitrate losses are as high as 50%, the overall difference between the uncorrected and nitrate-adjusted mass concentrations is not substantial, as indicated by the data shown in Table V-5. The effect of these losses on the concentration enrichment may have been more pronounced in other areas. Table V-8 shows the concentration enrichment achieved by the ultrafine concentrator of the VACES based on ultrafine particle counts. The first column of Table V-8 shows the ambient concentration based on particle counts, the second column shows that number concentration measured immediately downstream of the 0 .2 pm cutpoint impactor and the third column corresponds to the particle number concentrations measured immediately downstream of the diffusion dryer of the ultrafine concentrator of the VACES. The fourth column of this Table shows the ratio of particle counts downstream to that upstream of the 0.2 Jim impactor, indicating that about 84% of ambient particle counts are associated with particles smaller than that size. The final column of Table V-8 shows that concentration enrichment obtained for ultrafine particles, defined as the ratio of the count-based concentration downstream of the VACES to that downstream of the 0.2 pm impactor (i.e., prior to the aerosol entering the saturator). The 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. overall concentration enrichment averaged over 6 experiments was 21 (± 1.4), hence virtually identical to the ideal value of 22. Table V-8 Ultrafine PM number concentrations upstream and downstream of the 0.2 pm cutpoint impactor and downstream of the ultrafine concentrator of the VACES. VACES Number Concentration (particles/cm3 Concentration Downstream of the Impactor (particles/cm3 ) Ambient Particle Number Concentration (particles/cm3 ) Ratio of Concentration Downstream to Upstream Enrichment Factor 551429 26714 32185 83% 20.7 801429 35000 43166 86% 22.9 420000 23000 31750 74% 18.3 600000 29857 33666 85% 20.1 648571 30428 35714 85% 21.3 795714 38285 45142 84% 20.8 574286 26880 31523 86% 21.4 Average 0.83 20.8 Standard Deviation 0.042 1.41 Finally, Figure V-7 shows the size distributions of outdoor aerosols upstream of the VACES and immediately downstream of the diffusion dryer of the VACES line sampling fine plus ultrafine PM by means of the SMPS. The results of Figure V-7 show clearly that there is no distortion in the size distributions between ambient and concentrated aerosols, as the number median diameters (41 nm) and geometric standard deviation (1.7) of the concentrated and ambient aerosols are virtually identical. 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18000 16000 ^ 12000 I 10000 § 8000 <U -2 6000 4000 2000 0 0° oo ° NMD = 41.8 nm O A / GSD= 1.71 v O o O o ° o ° 0 1 o o Concentrated ▲ Ambient NMD = 41.3 nm GSD= 1.79 . 2 2 2 2 2 2 2 - afl' fla a* a a * a - 100 200 300 400 Particle diameter (nm) 500 600 Figure V-7 Size distribution of ambient aerosols before and after the Versatile Aerosol Concentration Enrichment System measured by SMPS. V-5 Summary and Conclusions This study focused on the development and the laboratory/field evaluation of a versatile particle concentration enrichment system, capable of enriching concurrently the concentration of ambient coarse, ultrafine only, and fine plus ultrafine particles by a factor up to 30, either as suspended in air or collected inside water. A high-flow rate, low pressure drop multi-slit nozzle impactor was developed as selective size inlet for ultrafine particles. The 50 % cut point was about 0.2 pm and the collection efficiencies for particles larger than 0.3 pm was more than 95 %. The comparison in ultrafine mass concentrations obtained using the multi-slit impactor and a reference MOUDI using indoor aerosols showed excellent agreement between the two samplers. 140 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The BioSampler was tested at a reduced flow rate and with the as small volume of water as possible. The pressure drop across the BioSampler decreased from 145 in. H2 0 to less than 20 inches. H2 0 as the flow rate was reduced from 12.5 lmin'1 to 5 lmin'1 . This low pressure drop across the sampler becomes beneficial especially in collecting semi-volatile species such as ammonium nitrate. The collection efficiency of BioSampler at the reduced flow rate was still high enough to collect particles larger than 2 pm. In addition, the collection efficiency of BioSampler was not influenced by the amount of water in the sampler’s reservoir, thereby making it possible to achieve highly concentrated suspensions for use in in vitro studies. The ability of the VACES to concentrate particles was first tested in laboratory experiments using different type of particles in the size range of 0.05-1.9 pm and at three minor flow rates of two 7, 10, and 20 lmin'1 with the total intake flow rate of 220 lmin'1 . The enrichment factors based on number concentrations were close to the ideal values. Hygroscopic aerosols, such as ammonium sulfate and ammonium nitrate were concentrated as effectively as hydrophobic PSL particles. The experimental characterization of the VACES demonstrated the concentration enrichment does not depend on particle size or chemical composition. Volatile species such as ammonium nitrate are preserved through the concentration enrichment process under the laboratory conditions used in this study. The performance of the VACES was further evaluated in a field study, conducted outdoors at Rancho Los Amigos National Rehabilitation Center in Downey, CA. The coarse, fine and ultrafine concentrations of the VACES were compared to those measured by means of a collocated MOUDI as a reference sampler. A virtually perfect balance in the total mass, sulfate, and nitrate of coarse mode was achieved between the VACES and 141 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the reference MOUDI. The enrichment factors for sulfate and nitrate-corrected mass concentrations were in very good agreement with the ideal value, while the enrichment factor for nitrate was overestimated due to the evaporation loss of nitrate in the low stages of MOUDI under the negative pressure. Enrichment factor based on the number concentration of ultrafine particles was 21 (± 1.4), virtually identical to the ideal value of 22. Furthermore, the comparison of size distribution between ambient and concentrated particles showed no significant distortion in terms of number median diameter and the geometric standard deviation. In summary, the versatile particle concentrator concentrated ambient aerosols with a very high efficiency and without any substantial distortion in their physico chemical characteristics. One of the most important features of the VACES is the ability to concurrently concentrate all three major PM size modes and to provide them either as suspended in air or collected into water enables toxicologists to simultaneously conduct in-vivo and in-vitro evaluation of toxic effects from ambient particles. References Bates, D.V. and Sizto, R. (1989) Hospital admissions and air pollutants in Southern Ontario; the acid summer haze effect. Environ. Health Perspect. 79, 69-76. Chang, M.C., Sioutas, C., Kim, S., Gong, H. Jr., and Linn, W. S. (2000) Reduction of nitrate losses from filter an impactor samplers by means of concentration enrichment. Atmos. Environ. 34, 85-98. Fung, K. (1990) Particulate Carbon Speciation by M n02 Oxidation. Aerosol Sci. Technol. 12,122-127. Godleski, J., Sioutas, C., Katler, M., and Koutrakis, P. (1996) Death from inhalation of Concentrated Ambient Air Particles in animal models of pulmonary disease. Resp. Crit. Care Med. 155(4), A246. 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Dick, C.A.J., Stone, V., Brown, D.M., Watt, M., Cherrie, JW ., Howarth, S., Seaton, A., and Donaldson, K. (2000) Toxic and inflammatory effects of filters frequently used for the collection of airborne particulate matter. Atmos. Environ. 34, 2587-2592. Gong, H. Jr., Sioutas, C., Linn, W.S., Clark, K.W., Terrell, S.L., Terrell, L.L., Anderson, K.R., Kim, S., and Chang, M. (2000) Controlled human exposures to concentrated ambient fine particles in metropolitan Los Angeles: Methodology and preliminary health-effect findings. Inhal. Toxicol. 12(S1), 107-119. Gordon T., Gerber H., Fang, C.P., and Chen, L.C. (1999) A centrifugal particle concentrator for use in inhalation toxicology. Inhal. Toxicol. 11,71-87. Grinshpun, S.A., Willeke, K., Ulevicius, V.Juozaitis, A., Terzieva, S., Donnelly, J., Stelma, G.N., and Brenna, K.P. (1997) Effect of impaction, bounce and reaerosolization on the collection efficiency of impingers. Aerosol Sci. Tech. 26, 326- 342. Heyder, J., Brand, P., Heinrich, J., Peters, A., Scheuh, G., Tuch, T. And Wichmann, E. (1996) Size distribution of ambient particles and its relevance to human health. Presented at the 2nd Colloquium on Particulate Air Pollution and Health, Park City, Utah, May 1-3. Hinds, W.C. (1999) Aerosol Technology. 2n d Edition John Wiley & Sons, New York. Kim, S., Chang, M.C., Kim, D., and Sioutas, C. (2000) A new generation of portable coarse, fine, and ultrafine particle concentrators for use in inhalation toxicology. Inhal. Toxicol. 12,121-137. Kim, S., Sioutas, C., Chang, M., Gong, G., and Terrell, L L. (2000) Factors affecting the stability of the performance of ambient fine particle concentrators. Accepted in Inhal. Toxicol. Lin, X., Reponen, T., Willeke, K., Wang, Z., Grinshpun, S. A., and Trunov, M. (2000) Survival of airborne microorganisms during swirling aerosol collection. Aerosol Sci. Tech. 32,184-196. Loomis, D., Castillejos, M., Boija-Aburto, V.H., and Dockery, D.W. (1999) Stronger effects of coarse particles in Mexico city. Proceedings o f the 3rd Colloquium on Particulate Air Pollution and Human Health. Durham, North Carolina, 5-13. Peters, A., Dockery, D.W., Heinrich, J., and Wichman, G.H.E. (1997) Short-term effects of particulate air pollution on respiratory morbidity in asthmatic children. Eer. Respir. J. 10, 872-879. Pope, C. A. Ill, Dockery, D. W., and Schwartz, J. (1995) Review of epidemiologic evidence of health effects of air pollution. Inhal. Toxicol. 7,1-18. 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sioutas, C., Wolfson, M., Ferguson, S.T., Ozkaynak, H. and Koutrakis, P.K. (1997) Inertial collection of fine particles using a high-volume rectangular geometry conventional impactor. J. Aerosol Sci. 6,1015-1028. Sioutas, C., Abt, E. Wolfson, J.M., and Koutrakis, P. (1999) Evaluation of the measurement performance of the scanning mobility particle Sizer and aerodynamic particle Sizer. Aerosol Sci. Tech. 30,84-92. Sioutas, C., Kim, S., and Chang, M. C. (1999) Development and evaluation of a prototype ultrafine particle Concentrator. J.Aerosol Sci. 30(8), 1001-1012. Sioutas, C., Kim, S., Chang, M. C., Terrell, L., and Gong, H. (2000) Field evaluation of a modified Data-RAM MIE scattering monitor for real time PM2.5 mass concentration measurements. Manuscript accepted for publication in Atmos. Environ. Sioutas, C., Koutrakis, P., and Burton, R.M. (1995) A technique to expose animals to concentrated fine ambient aerosols, Environ. Health Perspect. 103,172-177. Stelson, A. W. and Seinfeld J. H. (1982) Relative humidity and temperature dependence of the ammonium nitrate dissociation constants. Atmos. Environ. 16, 983-992. Thurston, Ito, K., Hayes, C.G., Bates, D.V., and Lippmann, M. (1994) Respiratory hospital admissions and summertime haze air pollution in Toronto, Ontario: Consideration of the role of acid aerosols, Environ. Res. 65, 270-290. Vedal, S. (1997) Ambient particles and health: Lines that divide. J. Air Waste Manage. Assoc. 47,551-581. Willeke, K., Lin, X., and Grinshpun, S.A. (1998) Improved aerosol collection by combined impaction and centrifugal motion. Aerosol Sci. Tech. 28,439-456. Whitby, K.T. and Svendrup, G.M. (1980) California Aerosols: Their Physical and Chemical Characteristics, Adv. Environ. Sci. Tech. 10, A ll. Zhang, X., McMurry, P.H., (1987) Theoretical analysis of evaporative losses from impactor and filter deposits. Atmos. Environ. 21,1779-1789. Zucker, B.A., Draz, A.M., and Muller, M. (2000) Comparison of filtration and impingement for sampling airborne endotoxins. J. Aerosol Sci. 31(6), 751-755. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter VI CONCLUSIONS Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The central objective of this thesis was to develop and evaluate a technology for concentrating ambient coarse, fine and ultrafine particles (i.e., the entire size range of PM1 0 ). This technology employs condensational particle growth followed by separation of grown particles through a virtual impactor to overcome the limitations of conventional ambient particle concentrators. Operational parameters have been experimentally optimized and the feasibility of using such a technology to provide ambient aerosols at high concentrations (up to 30 times the ambient concentration levels) for conducting inhalation exposure health study has been successfully demonstrated. The ambient particle concentrating technology was further extended to develop a versatile aerosol concentration enrichment system for simultaneous in vivo and in vitro evaluation of toxic effects of ultrafine, fine and coarse ambient particles. The results of this work and suggestions for future research are summarized in the following sections of this chapter. VI-1 Conclusions To initiate this work and establish the rationale for our motivation, the reliability of conventional concentrators was studied using the Harvard Ambient Fine Particle Concentrator. Systematic evaluation of factors affecting the stability of the performance of these systems was performed. Phenomenological problems during the operation of the conventional concentrators, including pressure drop increase and decrease in concentration enrichment, were statistically correlated with ambient air parameters such as temperature, relative humidity, as well as PM2.5 mass concentration and mass median diameter. Physical mechanisms responsible for clogging the nozzles of the virtual impactors of the concentrator are investigated. This analysis led to a better understanding 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of the effects of parameters such as temperature, humidity and ambient particle concentration on the stability of the performance of conventional concentrator. The normalized hourly pressure drop across the concentrator was strongly associated (R2 = 0.81) with the product of ambient PM2.5 mass concentration and the difference between the vapor pressure downstream of the impactor nozzle and the saturation vapor pressure at the adiabatic expansion temperature (i.e., the temperature of the aerosol immediately downstream of the virtual impactors). From multiple regression analysis, the average enrichment factor was predicted reasonably well (R2 = 0.67) by the aerosol mass median diameter and the normalized hourly pressure drop. Dilution of the ambient PM at days of high mass concentrations, prior to passing the aerosol through the virtual impactors, was suggested as a compromising approach to maintain reasonable concentration enrichment without a increase of pressure drop across the concentrator. In summary, conventional particle concentration technologies showed significant technical difficulties in operation, in addition to their inability to effectively concentrate PM outside the 0.35-2 pm range. Consequently, a prototype condensational growth/virtual impaction system was developed and evaluated to overcome problems of already developed concentrators and to extend the particle size range that can be concentrated to the ultrafine and coarse particle modes. Ultrafine particles are first grown by means of supersaturtion to a size that can be easily concentrated by a virtual impactor. The concentrated droplets are dried in a diffusion drier to obtain concentrated ultrafine aerosols. The system was evaluated using monodisperse 0.05 and 0.1 pm fluorescence PSL particles as well as polydisperse ultrafine ammonium sulfate, ammonium nitrate and indoor air particles. The condensational growth/virtual impaction system operated at a flow rate of either 106.5 or 147 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 110 LPM, of which 3.5 or 7 LPM were drawn through the minor flow of the virtual impactor. Experimental results identified saturation of the ultrafine aerosols at 35 °C and cooling to 25 °C as the optimum temperatures for operation of the concentrator. The average concentration enrichment was by a factor of 15.1 (+ 0.4) and 25.5 (± 1.9) for minor flow of 7 an 3.5 LPM, respectively. There was no evaporational loss of volatile ammonium nitrate at the high temperature of the saturator so far as the relative humidity is maintained high. More importantly, the experimental results showed that particle concentration occurs without any coagulation, which would have distorted the size distribution of the original ultrafine aerosols. Overall, this pilot study demonstrated the feasibility of concentrating ultrafine particles using a large cut point virtual impactor, with minimum losses and at a reasonably high output flow rate. This technology was further developed to concentrate particles in the diameter range of 0.01 - 10 pm (entire size range of PMi0 segregated into two coarse and ultrafine/fine mode). Coarse PM (2.5-10 pm) was concentrated in a single-stage, round nozzle virtual impactor, operating at an intake flow of 120 LPM. Fine and ultrafine PM (F+UFP, smaller than 2.5 pm) were concentrated by first removing larger particles by impaction and then growing the remaining particles via supersaturation to super- micrometer droplets. The droplets were then concentrated using the same coarse particle concentrator. Concentrated ultrafine and fine particles were returned to their original size by passing through a diffusion dryer filled with silica gel. The laboratory characterization of the fine and ultrafine concentrator (F+UFPC) demonstrated that the concentration enrichment values were essentially identical to the maximum obtainable concentration factors up to 2 pm particle size. Particles larger than 148 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 pm were not concentrated by the F+UFPC because of the inertial removal of these particles in saturator, which makes the use of pre-selective PM2.5 inlet unnecessary. Field tests of F+UFPC using indoor aerosols and chemical analysis for major constituents of PM2.5 showed that concentration enrichment did not depend on the chemical composition. Excellent agreement was found between mass, sulfate and nitrate concentrations measured by means of the F+UFPC and a collocated HEADS. Very good agreement was also found between the elemental and organic carbon concentrations determined using the F+UFPC and the MOUDI. The ambient particle concentrator developed in the early stages of this research was further extended to a versatile aerosol concentration enrichment system (VACES) for simultaneous in vivo and in vitro evaluation of toxic effects of coarse, fine, and ultrafine ambient particles. A multi-slit impactor with low-pressure drop was developed as selective size inlet for ultrafine particles. The 50 % cut point was 0.2 pm determined by number counts from SMPS and the collection efficiency of this impactor for particles larger than 0.25 pm was more than 95 % in terms of mass concentration measured by DataRAM. Field tests using this multi-slit impactor showed that nearly ideal concentration enrichment was obtained for ultrafine particles based on number counts. Number median diameters and geometric standard deviations of concentrated and ambient aerosols were virtually identical, hence indicating that the size distribution was not distorted in the process of concentrating particles by condensation growth and virtual impaction. Operating in conjunction with the coarse, ultrafine and fine particle concentrator, the BioSampler (a device in which particles are collected by inertial mechanisms directly in a liquid solution) can collect PM from any particle size range with virtually 100% 149 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. efficiency. This particle sampling strategy allows concentration of large quantities of ambient particles into a small liquid volume, which is essential for in vitro studies to obtain dose-response relationships of potentially toxic ambient particles. The ambient coarse and ultrafine/fine particle concentrators developed in this work have been proven to be a promising technology for conducting in vivo and in vitro health studies. These studies will test specific hypotheses on the relative toxicity of ambient particle size and composition characteristics, and will hopefully elucidate the currently unknown mechanisms that link particulate matter to health effects. The primarily application of these technologies and suggestions for future research are discussed in the following session. VI-2 Suggestions for future research The versatile aerosol concentration enrichment system (VACES) developed in this thesis can be utilized to conduct in-vitro studies for obtaining dose-response relationships between concentrations of ambient aerosols and cytotocity. In-vitro tests can quickly generate toxicological data using controlled amount of known chemical species. Without a diffusion drier attached to its end, the VACES can collect highly concentrated fine and ultrafine particles suspended in ultrapure water, which can be either transferred to other culture media or used directly for studying biological effects on cells. Concentrated indoor and outdoor air samples are being currently harvested in different locations of the Los Angels Basin using the VACES. The potential toxicity of ambient particles is currently investigated in a series of in vitro tests conducted in collaboration with the UCLA School o Medicine as one of the research activities of the Southern 150 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. California Particle Center and Supersite (SCPCS). Potential toxicity of ambient particles will be assayed by the exactly same procedure used for diesel exhaust particles (DEP), which already demonstrated substantial cell death ratio (Hiura et al., 1999) by DEP. Highly concentrated ambient fine and ultrafine particles are expected to show measurable cytotoxic effects. Samples collected by the BioSamplers are also supplied to Human Studies division of U.S. EPA in order to investigate the difference in the biological response to the ambient particles between the eastern and western United States. Animal exposures to highly concentrated ambient aerosols are another application of the ambient particle concentrators developed in this work. Animal exposures are being currently conducted also as art of the activities of the SCPCS. In general, animal exposure requires flow rates of 10 - 12 LPM for 10 rodents considering breathing flow rate of 0.2 LPM per rodent. The 10 LPM can be provided from two legs (concentration enriched by 20-fold) or all the three legs (enriched by 30-fold) of the VACES, while the third leg in the former case can be sued to simultaneously collect samples for in vitro test. A challenging task would be to build a large-scale particle concentration system capable of providing concentrated aerosols for human exposure studies. Humans breathe about 12 LPM at rest and more than 35 LPM during exercise, which requires that the concentrated air stream should be supplied at least at 100 LPM for exposure studies. Based on the above considerations, human exposures would require a similar system to the VACES, but scaled up by factor of three. The future version of this system for human exposure should be still compact in size in order to make it transportable to various locations with different size distribution and chemical composition of ambient particles. The modular design of the new 151 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concentrators described in this research opens the possibility to scale up their design without compromising their basic design principles. For example, the scale up can be easily achieved by placing several single-nozzle virtual impactors in parallel. Reference Hiura, T.S., Kaszubowski, M.P., Li, N. Nel, A.E. (1999) Chemicals in Diesel exhaust particles generate oxygen radicals and induced apoptosis in Macrophages. J. Immunol. 163(10), 5582-5591. 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bibliography Agarwal, J.K. and Sem, G.J. (1980) Continuous flow, single-particle-counting Condensation Nucleus Counter, J. Aerosol Sci., 11, 343-357. Ahn K.H. and Liu, B.Y.H. (1990) Particle activation and droplet growth processes in condensation nucleus counter-1. Theoretical background. J. Aerosol Sci., 21, 249-261. Allen, J.O., Dooleran, N.M., Taghizadeh, K., Lafleur, A.L., Smith, K.A., and Sarofim, A.F. (1997) Measurement of oxygenated polycyclic aromatic hydrocarbons associated with a size-segregated urban aerosol. Environ. Sci. Technol. 21, 2064-2070. Amdur, M.O., Chen, L.C., Guty, J., Lam, H.F. and Miller, P.D. (1988) Speciation and pulmonary effects of acidic S02 formed on the surface of ultrafine zinc oxide aerosols. Atmos. Environ., 22,557-560. American Thoracic Society (ATS), Committee of the Environmental and Occupational Health Assembly; Bascom, R., Chair. (1996) State of the art. Health effects of outdoor air pollution. Am. J. Respir. Crit. Care Med., 153,3-50. Anderson, K.R., Avol, E.L., Edwards, S.A., Shamoo, D.A., Peng, R.C., Linn, W.S., and Hackney J.D. (1992) Controlled exposures of volunteers to respirable carbon and sulfuric acid aerosols. J. Air and Waste Manage. Assoc., 42,437-442. Atkinson, A. and Arey, J. (1994) Atmospheric chemistry of gas-phase polycyclic aromatic hydrocarbons: formation of atmospheric mutagens. Environ. Health Perspec. 2 , 117-126. Barr, E.B., Hoover, M.D., Knapilly, G.M., Yeh, H.C. and Rothenberg, S.J. (1983) Aerosol concentrator: design, calibration and use. Aersol Sci. Technol. 2 ,437-442. Bates, D.V. and Sizto, R. (1989) Hospital admissions and air pollutants in Southern Ontario; the acid summer haze effect. Environ. Health Perspect. 79,69-76. Bates, D.V. and Sizto, R. (1989) Hospital admissions and air pollutants in Southern Ontario; the acid summer haze effect. Environ. Health Perspect. 19, 69-76. Biswas, P., Jones, C.L., and Flagan, R.C., (1987) Distortion of size distributions by condensation and evaporation in aerosol instruments. Aerosol Sci. Technol. 7, 231- 246. Brain, J.D., Knudson, D.E., Sorokin, S.P., Davis, M.A. (1976) Pulmonary distribution of particles given by intratracheal instillation or by aerosol inhalation. Environ. Res. 11,13-16. Brown, R.C. (1993) Aerosol Filtration: An Integrated Approach to the Theory and Applications of Fibrous Filters, Pergammon Press, Oxford. 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Burton, R.M., Wilson, W.E., Suh, H.H. and Koutrakis, P. (1996) Spatial variation in particulate concentration within the metropolitan Philadelphia. Envir. Sci. Technol. 30, 400-407. Chang, M.C., Sioutas, C., Kim, S., Gong, H., and Linn, W. (2000) Reduction of nitrate losses from filter and impactor samplers by means of concentration enrichment. Atmos. Environ., 34,85-98. Chen, B.T., and Yeh, H.C. (1987) An improved virtual impactor: Design and performance. J. Aerosol Sci. 18,203-214. Chen, L.C., Miller, P.D., Amdur, M.O., and Gordon, T. (1992) Airway hyperresponsiveness in guinea pigs Exposed to acid-coated ultrafine particles, J. Toxicol, and Environ. Health, 35,165-174. Clarke, R.W., Catalano, P.J., Koutrakis, P., Krishna Murthy, G.G., Sioutas, C., Paulauskis, J., Coull, B., Ferguson, S., and Godleski, J.J. (1999) Urban air particulate inhalation alters pulmonary function and induces pulmonary in a rodent model of chronic bronchitis. Inhal. Toxicol. 11,101-120. Coutant R.W., Callahan P.J., Kuhlman M.R., and Lewis R.G., (1989) Design and performance of a high-volume compound annular denuder. Atmos. Environ. 23 (10), 2205-2211. Dick, C.A.J., Stone, V., Brown, D.M., Watt, M., Cherrie, J.W., Howarth, S., Seaton, A., and Donaldson, K. (2000) Toxic and inflammatory effects of filters frequently used for the collection of airborne particulate matter. Atmos. Environ. 34, 2587-2592. Dockery, D.W., Speizer, F.E., Stram, D.O., Ware, J.H., Spengler, J.D., and Ferris, B.J. (1989) Effects of inhalable Particles on respiratory health of children, Am. Rev. Res. Dis. 139,587-594. Donaldson, K, Li, X.Y., and MacNee, W., (1998) Ultrafine (nanometer) particle mediated lung injury, J. Aerosol Sci. 29,553-560. Drinker, P., Thomson, R.M., and Finn, J.L. (1927) Metal Fume Fever: II. Resistance acquired by inhalation of zinc oxide on two successive days, J. Ind. Hyg. Toxicol. 9(3), 98-105. Eatough, D.J., Wadsworth, A, Eatough, D.A., Crawford, J.W., Hansen, L.D., and Lewis, E.A., (1993) A multiple-system, multi-channel diffusion denuder sampler for the determination of fine-particulate organic material in the atmosphere. Atmos. Environ. 27A: 1213-1219 Environmental Protection Agency (1996) Air Quality Criteria for Particulate Matter. EPA-600/P-95/001af, Office of Research and Development, Washington. 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. EPA (U.S. Environmental Protection Agency), National Ambient Air Quality Standard fo r Particulate Matter: Final Rules, FR1JL97-18, U.S. EPA, Washington DC, July 1997. Fang, C.P., Marple, V.A., and Rubow, K.L., (1991) Influence of cross-flow on particle collection characteristics of multi-nozzle impactors. J. Aerosol Sci. 22(4), 403-415. Fang, C.P., McMurry, P.H., Marple, V.A. and Rubow, K.L. (1991) Effect of flow- induced relative humidity changes on size cuts for sulfuric acid droplets in the microorfice uniform deposit impactor (MOUDI). Aerosol Sci. Technol. 14,266-277. Ferin, J., Oberdorster, G., Penney, D.P., (1992) Pulmonary retention of ultrafine and fine particles in rats. Am. J. Respir. Cell Mol. Biol. 6,535-542. Fernandez de la Mora, J., Hering, S.V., Rao, N., and McMurry, P.H. (1990) Hypersonic impaction of ultrafine particles. J. Aerosol Sci. 21,701-711. Finlayson-Pitts, B.J. and Pitts, J.N. (1986) Atmospheric Chemistry: Fundamentals and Experimental Techniques, John Wiley & Sons, New York. Fung, K. (1990) Particulate Carbon Speciation by M n02 Oxidation. Aerosol Sci. Technol. 12,122-127 Gavett, S.H., Hoyle, G.W., Madison, S.L., Walsh, L.C., Hilliard, H.G., Lappi, E.R., Evansky, P.E., and Costa, D.L. (1999) Pulmonary responses to concentrated air particles in allergen challenged hyperinnervated transgenic m ice.. Am. J. Respir. Crit. Care Med. 159, A29. Ghio, A.J., and Devlin, R.B. (1999) Healthy volunteers demonstrate no lung inflammation after exposure to fine particles concentrated from Chapel Hill ambient air. Am. J. Respir. Crit. Care Med. 159, A318. Godleski, J., Sioutas, C., Katler, M., and Koutrakis, P. (1996) Death from inhalation of Concentrated Ambient Air Particles in animal models of pulmonary disease. Resp. Crit. Care Med. 155(4), A246. Gong, H. Jr., Sioutas, C., Linn, W.S., Clark, K.W., Terrell, S.L., Terrell, L.L., Anderson, K.R., Kim, S., and Chang, M. (2000) Controlled human exposures to concentrated ambient fine particles in metropolitan Los Angeles: Methodology and preliminary health-effect findings. Inhal. Toxicol. 12(S1), 107-119. Gordon T., Gerber H., Fang, C.P., and Chen L.C. (1999) A centrifugal particle concentrator for use in inhalation toxicology. Inhal. Toxicol. 11, 71-87. Gordon, T., Chen, L.C., Fine, J.M., Schlesinger, R.B., Su, W.Y., Kimmel, T.A., and Amdur, M.O. (1992) Pulmonary effects of zinc oxide in human subjects, guinea pigs, rats, and rabbits, Am. Ind. Hyg. Assoc. J., 53, 503-509. 155 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Grinshpun, S.A., Willeke, K., Ulevicius, V.,Juozaitis, A., Terzieva, S., Donnelly, J., Stelma, G.N., and Brenna, K.P. (1997) Effect of impaction, bounce and reaerosolization on the collection efficiency of impingers. Aerosol Sci. Tech. 26, 326- 342. Heisler, S.L. and Friedlander, S.K. (1977) Gas-to-particle conversion in photochemical smog; growth laws and mechanisms for organics. Atmos. Environ. 11,158-168. Hering S.V., and Stolzenburg, M.R. (1995) On-line determination of particle size and density in the nanometer size range, Aerosol Sci. Technol., 23,155-173. Hering, S.V., Eldering, A. and Seinfeld, J.H. (1997) Bimodal characteristics of accumulation mode aerosol mass distributions in Southern California. Atmos. Environ., 31(1), 1-11. Hering, S.V., Hagan, R.C., and Friedlander, S.K. (1978) Design and evaluation of a new low pressure impactor-1. Environ. Sci. Technol. 12, 667-673. Hering, S.V., Gundel, L. and Daisey, J.M (1997) A microslot impactor for organic aerosol sampling. J. Aerosol Sci., 28,1283-1290. Herring, S.V., Eldering, A.M., and Seinfeld, J.H. (1997) Bimodal character of the accumulation mode aerosol mass distributions in southern California. Atmos. Environ. 31,1-11. Heyder, J., Brand, P., Heinrich, J., Peters, A., Scheuh, G., Tuch, T. And Wichmann, E. (1996) Size distribution of ambient particles and its relevance to human health. Presented at the 2nd Colloquium on Particulate Air Pollution and Health, Park City, Utah, May 1-3. Hinds, W.C. (1999) Aerosol Technology. 2n d ed. John Wiley & Sons, New York. Hiura, T.S., Kaszubowski, M.P., Li, N. Nel, A.E. (1999) Chemicals in Diesel exhaust particles generate oxygen radicals and induced apoptosis in Macrophages. J. Immunol. 163(10), 5582-5591. Hortonm K.D., Miller, R.D. and Mitchell, J.P. (1991) Characterization of a condensation- type monodispeerse aerosol generator (MAGE). J. Aerosol. Sci. 22, 347-363. International Commission on Radiological Protection (1994) Human Respiratory Tract Model for Radiological Protection; a Report of Committee 2 of the ICRP. Oxford: Pergamon Press. Japuntich, D.A., Stenhouse, J.I.T., and Liu, B.Y.H. (1994) Experimental results of solid monodisperse particle clogging of fibrous filters. J. Aerosol Sci. 25, 385-393. John, W., Wall, S.M., Ondo, J.L. and Winklmayr, W. (1990) Modes in the size distributions of atmospheric inorganic aerosol. Atmos. Environ. 22, 1627-1635. 156 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kim, S., Chang, M.C., Kim, D., and Sioutas, C. (2000) A new generation of portable coarse, fine, and ultrafine particle concentrators for use in inhalation toxicology. Inhal. Toxicol. 12, 121-137. Kim, S., Sioutas, C., Chang, M., Gong, G., and Terrell, L L. (2000) Factors affecting the stability of the performance of ambient fine particle concentrators. Accepted in Inhal. Toxicol. Kinney, P.L., and Ozkaynak, H. (1991) Associations of daily mortality and air pollution in Los Angeles County. Emir. Res., 54, 99-120. Kleinman, M.T., Bhalla, D.K., Mautz, W.J and Phalen, R.F. (1995). Cellular and immunologic injury with PM-10 inhalation. Inhal. Toxicol. 7, 589-602. Kleinman, M.T., Bhalla, D.K., Ziegler, B., Bucher-Evans, S., and McClure, T. (1997) Effects of inhaled fine particles and ozone on pulmonary macrophages and epithelia. Inhal. Toxicol, 5, 371-388. Kleinman, M.T., Mautz, W.J., Hyde, D.M. and Sioutas, C. (1998) Cardiopulmonary effects of inhaled ultrafine particles in aged rats. Presented at the American Association for Aerosol Research, June 1998. Koutrakis P., Sioutas C., Ferguson S., J.M. Wolfson, J.D. Mulik and R.M. Burton (1993) Development and evaluation of a glass honeycomb denuder/filter pack system to collect atmospheric particles and gases. Environ. Sci. Technol. 27,2497-2501. Koutrakis, P., Wolfson, J.M., Slater, J.L., Brauer, M., and Spengler, J.D. (1988) Evaluation of an annular denuder/filter pack system to collect acidic aerosols and gases. Environ. Sci. Technol. 22(12), 1463-1468. Ledbetter, A., Mebrane, R., Krantz, T., Jackson, M.C., Walsh, L., Hilliard, H., Richards, J., Chen, B., Costa, D.A., and Kodavanti, U.P. (1999) Variable pulmonary responses from exposure to concentrated ambient particles in a rat model of bronchitis. . Am. J. Respir. Crit. Care Med. 159, A29. Lin, X., Reponen, T., Willeke, K., Wang, Z., Grinshpun, S. A., and Trunov, M. (2000) Survival of airborne microorganisms during swirling aerosol collection. Aerosol Sci. Tech. 32,184-196. Linn, W.S., Szachcic. Y., Gong, H Jr., Kinney, P.L., and Berhane, K.T. (2000) Air pollution and daily hospital admissions in metropolitan Los Angeles. Environ. Health Perspectives, 108,427-434. Lippmann M. (1989) Airborne acidity: estimates of exposure and human health effects. Environ. Health Perspect. 63, 63-70. 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Loomis, D., Castillejos, M., Boqa-Aburto, V.H., and Dockery, D.W. (1999) Stronger effects of coarse particles in Mexico city. Proceedings o f the 3rd Colloquium on Particulate Air Pollution and Human Health. Durham, North Carolina, 5-13. Madden, M.C., Richards, J.H., Dailey, L.A., Hatch, and G.E., Ghio. (1999) Ozonation of Diesel exhaust particles affects lung responses. Proceedings o f the 3rd Colloquium on Particulate Air Pollution and Human Health. Durham, North Carolina, June 6-8,087. Marple, V.A. and Chien, C.M., (1980) Virtual Impactors: A Theoretical Study, Environ. Sci. Technol. 8 , 976-985. Marple, V.A. and Liu, B.Y.H. (1974) Characteristics of laminar jet impactors. Environ. Sci. Technol. 8, 648-654. Marple, V.A., and Willeke, K. (1976) In Fine Particles: Aerosol Generation, Measurement, Sampling, and Analysis (Edited by B.Y.H. Liu). Academic press, New York. Marple, V.A., Rubow, K.L. and Behm, S. (1991) A Microorifice uniform deposit impactor (MOUDI): description, calibration, and use. Aerosol Sci. Technol. 14, 434- 446. Mavliev, R., Wang, H., Hopke, P.K. and Lee, D. (1999) A transition from heterogeneous nucleation in the turbulent mixing CNC. 18lh Annual AAAR Conference. Tacoma, WA. 331. McMurry, P. H. and Wilson, J.C. (1882) Growth laws for the formation of secondary ambient aerosols: Implications for chemical conversion mechanisms. Atmos. Environ. 16(1), 121-134. McQuiston, F. and Parker J.D. (1982). Heating, Ventilation and Air Conditioning. John Wiley and sons, New York. Miguel, A.G., Arey, J.M. and Sousa, J.A. (1990) Comparative study of the mutagenic and genotoxic activity associated with inhalable particulate matter in Rio de Janeiro air. Environ. Molecular Mutag. 15, 36-43. Miguel, A.H., Kirchastetter, T.W., Harley, R.A. and Hering, S.V., (1998) On-road emissions of polycyclic aromatic hydrocarbons form gasoline and diesel vehicles. Environ. Sci. Technol. 32,450. Muyshondt, A., McFarland, A.R. and Anand, N.K. (1996) Deposition of aerosol particles in contraction fittings. Aerosol Sci. Technol., 24, 205-216. Oberdorster, G., Ferin, J., and Lehnert, B.E. (1994) Correlation between particle size, in vivo particle persistence and lung injury. Environ. Health Perspect. 102,173-179. 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Oberdorster, G., Ferin, J., Gelein, R., Soderholm S.C., and Finkelstein, J. (1992) Role of alveolar macrophage in lung injury; studies with ultrafine particles, Environ. Health Perspect. 97, 193-197. Oberdorster, G., Gelein, R.M., Ferin, J., and Weiss, B. (1995) Association of particulate air pollution and acute mortality: involvement of ultrafine particles?. Inhal. Toxicol. 7,111-124. Olawoyin, O.O., Raunemaa, T.M. and Hopke, P.K. (1995) A system for aerodynamically sizing ultrafine radioactive particles, Aerosol Sci. Technol. 23,121-130. Ozkaynak, H. and Thurston, G.D. (1987) Associations between 1980 U.S. morality rates and alternative measures of airborne particle concentration. Risk Analysis. 7,4. Peters, A., Dockery, D.W., Heinrich, J., and Wichman, GH.E. (1997) Short-term effects of particulate air pollution on respiratory morbidity in asthmatic children. Eer. Respir. J. 10, 872-879. Pinch, G.L., Nikula, K.J., Barr, E.B., Seagrave, J.C., Snipes, M.B., Hobbs,C.H., and Mauderly, J.L. (1999) Biokinetics of an ultrafine silver aerosol inhaled by rats. Proceedings o f the 3rd Colloquium on Particulate Air Pollution and Human Health. Durham, North Carolina, June 6 -8,110. Pope, C. A. Ill, Dockery, D. W., and Schwartz, J. (1995) Review of epidemiologic evidence of health effects of air pollution. Inhal. Toxicol. 7,1-18. Pope, C.A., Schwartz , J. and Ransom, M.R. (1992) Daily mortality and PM10 pollution in Utah Valley. Arch. Env. Health 47, 211-217. Schwartz , J., Dockery, D.W., and Neas, L.M. (1996) Is daily morality associated specifically with fine particles. J. Air Waste Manag. Assoc. 4 6 ,927-939. Schwartz, J. and Dockery, D.W. (1992) Particulate air pollution and daily mortality in Steubenville, Ohio. Am. J. Epidemiol. 135, 12-19. Seinfeld, J.H. and Pandis, S.N. (1998) Atmospheric Chemistry and Physics. John Wiley & Sons, New York. Sioutas, C. and Koutrakis, P. (1996) Inertial separation of ultrafine particles using a condensational growth/virtual impaction system. Aerosol Sci.Technol., 25,424-436. Sioutas, C., Abt, E. Wolfson, J.M., and Koutrakis, P. (1999) Evaluation of the measurement performance o f the scanning mobility particle Sizer and aerodynamic particle Sizer. Aerosol Sci. Tech. 30,84-92. Sioutas, C., Kim, S., and Chang, M. C. (1999) Development and evaluation of a prototype ultrafine particle Concentrator. J.Aerosol Sci. 30(8), 1001-1012. 159 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sioutas, C., Kim, S., Chang, M. C., Terrell, L., and Gong, H. (2000) Field evaluation of a modified Data-RAM MIE scattering monitor for real time PM2.5 mass concentration measurements. Manuscript accepted for publication in Atmos. Environ. Sioutas, C., Koutrakis P., and Olson, B.A. (1994) Development and evaluation of a low cutpoint virtual impactor, Aerosol Sci. Technol., 21(3), 223-236. Sioutas, C., Koutrakis, P., and Burton, R.M. (1994) Development of a low cutpoint slit virtual impactor for sampling ambient fine particles, J. Aerosol Sci., 25(7), 1321- 1330. Sioutas, C., Koutrakis, P., and Burton, R.M. (1995) A technique to expose animals to concentrated fine ambient aerosols, Environ. Health Perspect. 103, 172-177. Sioutas, C., Koutrakis, P., Ferguson, S.T. and Burton, R.M. (1995) Development and evaluation of a prototype ambient particle Concentrator for inhalation exposure studies. Inhal. Toxicol 7,633-644. Sioutas, C., Koutrakis, P., Godleski, J., Ferguson, S.T., Kim, C.S. and Burton, R.M. (1997) Harvard/EPA ambient fine particle concentrators for human and animal exposures. J Aerosol Sci. 28(6), 1057-1077. Sioutas, C., Wolfson, M., Ferguson, S.T., Ozkaynak, H. and Koutrakis, P.K. (1997) Inertial collection of fine particles using a high-volume rectangular geometry conventional impactor. J. Aerosol Sci. 6 , 1015-1028. Spengler, J.D., and Thurston, G.D. (1983) Mass and elemental composition of fine and coarse particles in six U.S. cities. JAPCA 33,1162-1171. Spumy K., Lodge J.P., Frank E.R., and Sheesley, D.C. (1969) Aerosol filtration by means of Nuclepore filters: structural and filtration properties. Environ. Sci. Technol. 3,453-464. Stelson, A. W. and Seinfeld J. H. (1982) Relative humidity and temperature dependence of the ammonium nitrate dissociation constants. Atmos. Environ. 16, 983-992. Thatcher, T.L. and Layton, D.H (1995) Deposition, resuspension and penetration of particles within a residence. Atmos. Environ., 29(13), 1487-1497. Thurston, Ito, K., Hayes, C.G., Bates, D.V., and G., Lippmann, M. (1994) Respiratory hospital admissions and summertime haze air pollution in Toronto, Ontario: Consideration o f the role of acid aerosols, Environ. Res. 65, 270-290. Urch, B., Liu, L., Brook, J., Purdham, J., Tarlo, S., Broder, I., Lukic, Z., Datema, J., Koutrakis, P., Sioutas, C., Ferguson, S., Dales, R., and Silverman, F. (1999) Pulmonary function responses after inhalation of controlled levels of concentrated urban particles in healthy individuals. Am. J. Respir. Crit. Care Med. 159, A318. 160 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Vedal, S. (1997) Ambient particles and health: lines that divide. J. Air Waste Manage. Assoc. 47,551-581. Whitby, K.T. and Svendrup, G.M. (1980) California Aerosols: Their Physical and Chemical Characteristics, Adv. Environ. Sci. Technol. 10,477. Whitby, K.T., Husar, R.B., and Liu, B.Y.H. (1972) Aerosol size distribution of Los Angeles smog. J. Colloid and Interface Science 39,177-204. White, F.M. (1979) Fluid Mechanics. McGrow-Hill, Inc. New York. Willeke, K., Lin, X., and Grinshpun, S.A. (1998) Improved aerosol collection by combined impaction and centrifugal motion. Aerosol Sci. Tech. 28,439-456. Zhang, X., McMurry, P.H., (1987) Theoretical analysis of evaporative losses from impactor and filter deposits. Atmos. Environ. 21,1779-1789. Zucker, B.A., Draz, A.M., and Muller, M. (2000) Comparison of filtration and impingement for sampling airborne endotoxins. J. Aerosol Sci. 31(6), 751-755. 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Development and evaluation of a personal particle sampler and a mobile high-capacity particle size classifier

PDF

Investigation of the physical and chemical characteristics of ambient coarse particulate matter in indoor and outdoor environments

PDF

Evaluating gas-phase biofilter performance in treating chlorinated VOCs from industrial emissions: Principle, modeling and design

PDF

A percolation biofilm-growth model for biomass clogging in biofilters

PDF

Investigation of chemical characteristics of ambient particulate matter and the associations between particle and gaseous copollutant concentrations in the Los Angeles Basin

PDF

Biological nitrification-denitrification and sulfate reduction of reverse osmosis brine rejects

PDF

Assessment of virus coagulation and filterability using particle density distribution analysis

PDF

Investigation of asphaltic macrostructure and its free radical conversion

PDF

Experimental studies of biopolymers behavior in subsurface porous strata

PDF

Instrumentation and methods for PM(2.5) and ultrafine particulate matter exposure assessment

PDF

Biopolymer plugging effect in porous media: Gelation modeling and lead stabilization

PDF

Health monitoring of structures under ambient vibrations using semiactive devices

PDF

Gas migration from oil /gas fields and gas storage fields: Environmental implications

PDF

Advances in receptor modeling with application to ambient atmospheric concentrations of volatile organic compounds

PDF

Development of high frequency annular array ultrasound transducers

PDF

Design and performance analysis of low complexity encoding algorithm for H.264 /AVC

PDF

Analysis of the effect of varying port spacing on initial dilution for a multiport outfall diffuser

PDF

Geotechnical applications of biopolymer technology

PDF

Algorithms and architectures for robust video transmission

PDF

Biofiltration of ethanol vapor: Head-loss studies of heavily loaded systems

Asset Metadata

Creator Kim, Seongheon (author)

Core Title Development of ambient particulate concentration technology using condensational growth/virtual impaction

School Graduate School

Degree Doctor of Philosophy

Degree Program Environmental Engineering

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag engineering, environmental,geophysics,OAI-PMH Harvest,Physics, Atmospheric Science

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Siuotas, Contantinos (committee chair), Gong, Henry (committee member), Henry, Ronald (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-241849

Unique identifier UC11339269

Identifier 3093425.pdf (filename),usctheses-c16-241849 (legacy record id)

Legacy Identifier 3093425.pdf

Dmrecord 241849

Document Type Dissertation

Rights Kim, Seongheon

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA