University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INVESTIGATIO N OF THE PHYSICAL AND CHEM ICAL CHARACTERISTICS OF A M B IEN T COARSE PARTICULATE M ATTER IN INDOOR AND OUTDOOR ENVIRONMENTS Copyright 2003 by Michael David Geller A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL U NIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (ENVIRO NM ENTAL ENGINEERING) August 2003 Michael David Geller Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3116702 INFORMATION TO USERS The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. ® UMI UMI Microform 3116702 Copyright 2004 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. U N IV E R S IT Y O F SO U TH ER N C A LIFO R N IA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES, CALIFORNIA 90089-1695 This dissertation, written by M ichael D. G e lle r under the direction o f h dissertation committee, and approved by all its members, has been presented to and accepted by the D irector o f Graduate and Professional Programs, in partial fulfillm ent o f the requirements fo r the degree o f DOCTOR OF PHILOSOPHY Director Date C-lo- 1003 Dissertation Committee Chair Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS The past three years have passed so very quickly, yet I have discovered an inestimable amount about my field and myself. I fear that I w ill commit an injustice to all who lent me a hand along the way or simply alleviated the daily strain of graduate student life. I would like to thank the following people for their contributions and support without which this dissertation would not have been possible. First and foremost, I would like to express my deepest gratitude to Constantinos Sioutas, who has become much more than a thesis advisor. His invaluable guidance throughout my career as a student has enabled me to develop as both an individual and a professional. His ears were always the first to hear my ideas, and his eyes were the last to edit my final drafts. The other members of my committee, Henry Gong and Philip Fine, have provided very thoughtful commentary on this document. I am fortunate to have had the opportunity to work with both of them on many projects. I am also indebted to Ming-Chih Chang and Seongheon Kim for the many hours that they spent showing me the ropes. A ll of the people associated with the Southern California Particle Center and Supersite have been essential to my research, for they continually provided support in both field and laboratory experiments and analyses. I would also like to acknowledge the United States Environmental Protection Agency and California Air Resources Board as the primary sponsors of all of this research. ii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Finally, I come to my friends and family, who deserve much more than an honorable mention in this paragraph. The hours of studying, researching, and writing were offset by mere seconds spent with all of you. I appreciate Jennifer Holsten’s support and devotion throughout my college years. Mom, Dad and Kory, thank you from the bottom of my heart. I cannot convey how much your love and encouragement mean to me. Thank you all for helping to shape my life for the past 25 years. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS i Acknowledgements ii List of Tables viii List of Figures ix Abstract xiv 1. Introduction 1 1.1 Background 1 1.1.1 Characteristics of Ambient Particulate Matter 1 1.1.2 Health Effects of Particulate Matter 4 1.2 Rationale for the Proposed Research 5 1.3 Theory 7 1.4 Thesis Overview 8 References 12 2. Indoor/Outdoor Relationship and Chemical Composition o f Fine and Coarse Particles in the Southern California Deserts 14 2.1 Abstract 14 2.2 Introduction 16 2.3 Methods 17 2.3.1 Study Design 17 2.3.2 Sample Analysis 20 2.4 Results and Discussion 22 2.4.1 Indoor and Outdoor Coarse and Fine PM Mass 22 Concentrations iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.4.2 Indoor and Outdoor Elemental and Organic Carbon Concentrations for Fine PM 25 2.4.3 Indoor and Outdoor Trace Element and Metal Concentrations for Fine PM 29 2.4.4 Indoor and Outdoor Trace Element and Metal Concentrations for Coarse PM 33 2.5 Summary and Conclusions 37 References 39 3. Development and Evaluation of a Compact, Highly Efficient Coarse Particle Concentrator for Toxicological Studies 41 3.1 Abstract 41 3.2 Introduction 42 3.3 Experimental Methods 45 3.3.1 Laboratory Characterization of the Virtual Impactors 48 3.3.2 Field Evaluation of the Scaled-up Coarse Particle Concentrator 50 3.4 Results and Discussion 53 3.4.1 Laboratory Characterization of the Virtual Impactors 53 3.4.2 Field Evaluation Tests 57 3.5 Conclusions 63 References 66 4. Development and Evaluation o f a Continuous Coarse Particle Monitor 70 4.1. Abstract 70 v Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.2. Implications 71 4.3. Introduction 72 I 4.4. Methods 77 4.4.1 Description of the Continuous Coarse Particle Monitor 77 4.4.2 Laboratory Evaluation of 2.5 pm Cutpoint Round Nozzle Virtual Impactor 80 4.4.3 Field Study 82 4.5. Results and Discussion 84 4.5.1 Evaluation of the PM j o Inlet 84 4.5.2 Laboratory Evaluation of the 2.5 pm Cutpoint Round Nozzle Virtual Impactor 85 4.5.3 Field Evaluation of the Continuous Coarse Particle Monitor 87 4.6. Summary and Conclusions 97 References 99 5. Development and Evaluation of a P M j0 Impactor-Inlet for a Continuous Coarse Particle Monitor 101 5.1. Abstract 101 5.2. Introduction 102 5.3. Materials and Methods 104 5.3.1 Description of the PMio Impaction Inlet 104 5.3.2 Laboratory Tests for Determination o f Cutpoint 105 5.3.3 Wind Tunnel Tests 106 5.3.4 Field Evaluation o f the P M j o Inlet 112 vi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.4. Results and Discussion 114 5.4.1 Experimental Determination of Cutpoint 114 i 5.4.2 Wind Tunnel Evaluation o f the PMio Inlet 115 5.4.3 Field Evaluation of the Inlet 117 5.5 Summary and Conclusion 122 References 125 6. The Relationship Between Both Real-time and Time-integrated Coarse, Intermodal, and Fine Particulate Matter in the Los Angeles Basin 126 6.1 Abstract 126 6.2 Implications 127 6.3 Introduction 128 6.4 Methods 130 6.4.1 Sampling Location 130 6.4.2 Instrumentation 131 6.5 Results and Discussion 134 6.5.1 Site-by-Site Comparisons 135 6.5.2 Comparisons Between PM Modes Based on Chemical Composition 143 6.6 Summary and Conclusions 155 References 157 7. Conclusion 160 7.1 Summary 160 7.2 Conclusions 162 vii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7.3 Recommendations for Future Research References Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES Table 2.1 Table 2.2 Table 2.3 Table 3.1 Table 3.2 Table 5.1 Characteristics of Coachella Valley Indoor/Outdoor Sampling Locations Descriptive Statistics for PM with Respect to Sampling Location and Size Cut Average Ratio of Indoor-to-Outdoor Particle Concentrations for Selected Trace Elements and Metals Comparisons Between Coarse Particle Concentrator and M O UDI Based on Mass, Sulfate and Nitrate Concentrations Ambient Concentrations and Enrichment Factor for Selected Metal/Elements Based on 13 Sets o f Comparisons Between CPC and M O UDI Comparison Between Coarse PM Concentrations of Various Crustal Metals Measured by the PMio Inlet and Partisol Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES Figure 2.1 Indoor Vs. Outdoor Coarse Particle Mass Concentrations 24 Figure 2.2 Indoor Vs. Outdoor Fine Particle Mass Concentrations 25 Figure 2.3 PM2.5 Elemental Carbon Indoor/Outdoor Comparison By Home 27 Figure 2.4 PM2.5 Organic Carbon Indoor/Outdoor Comparison By Home 28 Figure 2.5 PM2.5 Trace Element and Metal Indoor/Outdoor Comparisons By Home 32 Figure 2.6 Coarse Particle Trace Element and Metal Indoor/Outdoor Comparisons By Home 34 Figure 3.1 Schematic of the Round Jet Impactors of the Multi-Nozzle Coarse Particle Concentrator 46 Figure 3.2 a. Top View of the Coarse Concentrator 47 b. Side View o f the 10-nozzle Coarse Particle Concentrator 48 Figure 3.3 Concentration Enrichment of Each Individual Virtual Impactor at Three Different Minor Flow Rates 54 Figure 3.4 Comparisons Between the Coarse Particle Concentrator and M O UDI Coarse PM Mass Concentrations 56 Figure 3.5 Comparisons o f Coarse PM Nitrate Concentrations Between CPC and M O U D I 59 Figure 3.6 Comparisons o f Coarse PM Sulfate Concentrations Between CPC and M O U D I 59 Figure 3.7 Comparisons of Coarse PM Concentrations Between CPC and M O UDI for Selected Trace Elements and Metals 60 Figure 3.8 Comparison of Multiple Metal/Elemental Concentrations Between CPC and M O U D I Based on Coarse PM Concentrations 64 Figure 4.1 Schematic of the Continuous Coarse Particle Monitor 78 Figure 4.2 Particle Penetration Through the PMio Inlet 85 x Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.3 Concentration Enrichment Factor as a Function of Particle Aerodynamic Diameter 86 Figure 4.4 TEOM™ vs MOUDI™ CM Concentrations; TEOM™ at 50 °C 88 Figure 4.5 TEOM™ vs Partisol™ CM Concentrations; TEOM™ at 50 °C 89 Figure 4.6 Dependence of TEOM™-MOUDI™ and TEOM™-Partisol™ Ratio on FM-to-CM concentration ratio; TEOM™ at 50 °C 90 Figure 4.7 Relationship Between Coarse Particle Mass Median Diameter (M M D ) and FM-to-CM concentration ratio. 90 Figure 4.8 TEOM™ vs Partisol™ CM Concentrations; TEOM™ at 30 °C 93 Figure 4.9 Plot of the Partisol™-to-MOUDI™ CM Concentrations as a Function of Ambient Relative Humidity 94 Figure 4.10 Plot of the Ratio of Partisol™-to-MOUDI™ FM Concentrations as a Function of Relative Humidity 95 Figure 4.11 Dependence of TEOM™-Partisol™ Ratio on FM -to-CM concentration ratio; TEOM™ at 30 °C 96 Figure 4.12 Time-series of TEOM™ and APS™ CM Concentrations 96 Figure 5.1 a. Modified PMio Inlet for CCPM 107 b. Dimensions of Modified Acceleration Jet Nozzle 107 Figure 5.2 Schematic of the Experimental Set-up Used for the Laboratory Characterization o f the PM io Inlet 110 Figure 5.3 PMio Inlet Particle Penetration Curve 115 Figure 5.4 Plot of Penetration Vs. Particle Diameter for Various Wind Speeds 116 xi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 5.5 Coarse PM Concentrations Determined by the 50 LPM PMio Inlet and the R&P Partisol 118 Figure 5.6 Overall Comparison Between Coarse PM Concentration of Five Crustal Metal Measured by Partisol and PMio Inlet 119 Figure 5.7 a. Plot of Coarse PM Nitrate Concentrations Between PMio Inlet and Partisol b. Plot of Coarse PM Sulfate Concentrations Between PMio Inlet and Partisol 121 121 Figure 5.8 Plot of Ratio of PMio/Partisol Coarse Concentrations Vs. Wind Speed 122 Figure 6.1 PM i.2.5 Versus PM2.5 at A ll Sites 134 Figure 6.2 Chemical Composition of Intermodal PM Averaged by Location 135 Figure 6.3 a. Intermodal Versus Coarse PM at USC b. Intermodal Versus PM i at USC 137 137 Figure 6.4 a. Coarse, Intermodal, and Fine 11/25-12/1 2002 b. Coarse, Intermodal, and Fine 12/02-12/09 2002 138 138 Figure 6.5 a. Intermodal Versus Coarse PM at Downey, CA b. Intermodal Versus PMi at Downey, CA 140 140 Figure 6.6 a. Intermodal Versus Coarse PM at Riverside/Rubidoux, CA b. Intermodal Versus PMi at Riverside/Rubidoux, CA 142 142 Figure 6.7 a. Intermodal Versus Coarse PM at Claremont, CA b. Intermodal Versus PM i at Claremont, CA 144 144 Figure 6.8 a. PM2.5-10 Vs. PM j-2.5 Nitrate Mass Concentrations for Claremont, CA b. PM i.2.5 Vs. PM2.5 Nitrate Mass Concentrations for Claremont, CA 146 146 Figure 6.9 Continuous P M I-2.5 Versus P M I Nitrate at Claremont, CA in September 2001 147 Figure 6.10 Daily Wind Speed and PM 1 -2.5 /PM 2.5 Nitrate at Claremont, CA in September 2001 148 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6.11 Monthly Average of the Ratio of Intermodal PM to Total PM2.5 149 Figure 6.12 a. OC and Sulfate Versus Nitrate Concentration in the 1-2.5 p irn Range at Source Sites 150 b. OC and Sulfate Versus Nitrate Concentration in the 1-2.5 Hm Range at Source Sites 150 Figure 6.13 Intermodal Versus Coarse PM Crustal Elements for Claremont, CA 152 Figure 6.14 PM i.2.5 Versus Estimated PM i.2.5 Soil Concentration 153 Figure 6.15 OC, Nitrate and Sulfate Concentrations Versus Soil Concentration in Intermodal PM 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT Atmospheric coarse particles have been studied thoroughly but have recently come back into the spotlight due to a combination of bioterroristic threats, new technologic discoveries, and health effects studies. The longtime standard for sampling and measuring coarse PM, the 24-hour time-integrated filter sample, has been questioned for accuracy due to possible measurement error. This thesis is intended to alleviate the paucity o f information on both the short-term variation o f ambient coarse PM and to what extent humans are exposed to it. An indoor versus outdoor study in a desert location was conducted to demonstrate that even in environments with high outdoor coarse PM concentrations, indoor home exposure to coarse particles is almost entirely influenced by indoor sources and not penetration of outdoor coarse PM. Due to the relatively low ambient levels of coarse PM in many locations, a coarse particle concentrator was developed and evaluated. This device can concentrate ambient levels up to 30 times for in vivo exposure studies or sample collections over much shorter time intervals. Another device designed to measure coarse PM over short time intervals is also presented. The Continuous Coarse Particle Monitor measures real-time coarse particle mass concentrations and thus can be used for simple monitoring or detailed human or animal exposure studies. A PMio size-selective inlet was modified to operate at 50 LPM, the designed flow rate of the Continuous Coarse Particle Monitor. This inlet was tested in a laboratory wind tunnel and an outdoor field location and found to have a very sharp cutpoint near 9 pm. The final chapter o f this work is dedicated to the overlap xiv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. between the ambient coarse and fine PM modes. A study was conducted to determine the effect of the tail of the coarse PM mode on the intermodal size range, i or particles between 1 and 2.5 pm. Results indicate that in Los Angeles, intermodal PM is very similar to PMi and shows nearly no correlation with coarse PM. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 1 INTRODUCTION 1.1 Background 1.1.1 Characteristics o f Ambient Particulate Matter Ambient particulate matter (PM) is a general name given to an atmospheric aerosol, which is a suspension of solid particles and/or liquid droplets in the atmosphere. PM is the keyword and not “aerosol” because the particulate mass and volume in an aerosol is too insignificant to discern from the air in which it is contained. Thus, researchers must focus on the each individual particle’s properties to acquire and understanding of the complete mixture o f particles. The sources o f PM are many, and each source can change on a daily basis. Sources can be divided into anthropogenic and natural, after which they can be divided into mobile and stationary. Natural sources tend to be stationary and include volcanoes, wind-blown dust, and salt from sea-spray. Anthropogenic sources can be either stationary or mobile. Examples of the former are coal-fired power plants, wood-burning stoves, and soil disturbance from man-made construction projects. The largest category of mobile sources of PM is the combined contribution of vehicle exhaust from the millions of automobiles and trucks driven throughout the world. Particles can be formed by a variety of processes, but the two on which most air pollution scientists focus are direct emission and secondary formation. Secondary formation can occur so close to a source that the particles are actually Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. considered to be directly emitted. This is the case in automobile exhaust because high concentrations of gaseous pollutants, such as nitrogen oxides, sulfur dioxide, and a host of organic gases, at very high temperatures mix with ambient air upon exit from the tailpipe. The result of this rapid cooling is condensation o f vapors onto very small nuclei, and particles are formed. Particles are directly emitted from vehicles as well as any combustion source, resuspension of soil and road dust, and from plants in the form o f pollen and spores. Secondary formation occurs by many processes that occur in the atmosphere and is driven by vapor concentration, particle concentration, temperature, relative humidity, and chemical reaction kinetics. The source and formation of a particle determines its size. Mechanically produced particles arise from disintegration o f liquids and solids and are usually larger than 1 pm ( 1C4 cm) in diameter while particles formed from gas phase vapors normally have sub-micron diameters. The distance between the large particles and smaller particles is spanned by growth o f the smaller particles by condensation of vapors onto their surface and coagulation o f two or more particles (Friedlander, 2000). Because of the variety of sources, chemical composition, and sizes of particles, it has become useful to divide them into three size ranges. The largest particle size range considered in modem aerosol science is the coarse mode and consists of particles with aerodynamic diameters between 2.5 pm and 10 pm. Coarse particles are windblown dust, salt particles from sea spray, bioaerosols (e.g. pollen, mold), and mechanically generated man-made particles (Hinds, 1999). The 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. latter originates from agriculture, surface disturbances (e.g. mining, construction development), disintegration of tires and brake linings from cars, etc. The size range of 0.1 pm to 2.5 pm is termed the accumulation mode. Some also call this the fine mode while others define the fine mode as 0-2.5 pm. Accumulation mode particles are the products of combustion, atmospheric chemical reactions (smog), and particles with diameters less than 0.1 pm that have coagulated with these larger particles. Many combustion particles may originally be less than 0.1 mm in diameter, but they reach thermodynamic equilibrium quickly upon release from the combustion source. While equilibrating, these particles serve as condensation nuclei and thus grow as vapors concentrate on their surfaces. Fine particles can exist in the atmosphere on the order of days to weeks, which enables them to be transported long distances (Hinds, 1999). For this reason fine PM is responsible for regional and global air pollution. The smallest PM size range is that o f the ultrafine or nuclei mode, which includes all particles with aerodynamic diameters less than 0.1 pm. These particles form from both homogeneous and heterogeneous nucleation. The former is the process by which a high concentration of vapor is rapidly cooled and condenses into a liquid droplet while the latter is the process by which vapors are cooled and condense on a solid pre-existing particle. Some speculate that particles this small are either sulfate or elemental carbon. Because ultrafine particles are almost as small as gas molecules, they behave similarly. Thus, they are dominated by diffusion and 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. w ill either coagulate with larger particles or grow by condensation. The result is that their lifetimes as ultrafine particles are very short. 1.1.2 Health Effects o f Particulate Matter In 1970 the United States Environmental Protection Agency (USEPA) began regulating total suspended particulates (TSP) in the atmosphere. After health researchers began to study respiratory deposition o f PM, the upper bound was set at 10 pm because particles larger than this are not respirable, which means that they cannot travel into the alveolar region of the lungs. Particles between 10 and 35 pm can still be inhaled, but they are removed by the nose, throat, and bronchi (Yeh and Schum, 1980; Yeh et al., 1996). The USEPA, in turn, modified its regulatory criteria in 1987 and developed a National Ambient A ir Quality Standard (NAAQS) for PMio at 50 pg/m3 annually averaged and 150 pg/m3 daily averaged (Wark et al., 1998). In 1997 a new standard for PM2.5 was set at 15 pg/m and 65 pg/m annually averaged and daily averaged, respectively, due to the huge variation between the coarse and fine modes and recent experimental results that demonstrate greater toxicity of the fine mode. Cohort studies, such as the Harvard Six City Study, that began in the 1970s published results in the early 1990s, which prompted regulators to establish the PM2.5 standard. This study concluded that mortality was strongly associated with fine PM (Dockery et al., 1993). Other studies in the since the Harvard Six City Study have concluded the same results: fine PM is highly correlated with daily mortality (Pope et al., 1995; Schwartz and Neas, 2000; Tsai et 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. al., 2000). A recently published study, however, found that coarse mode particles increased the biological markers associated with alveolar macrophage response in the lungs more than the fine and ultrafine size fractions (Becker et al., 2003). Current studies are now finding that ultrafine PM has the largest toxic response per unit mass in animals and cells. Oberdorster found that ultrafine particles induced an immune response in rats more than larger particles per unit mass (2000, 2001). Ultrafine particles can enter the cell and cause oxidative stress by damaging the mitochondria, which are cells’ energy producers (Li et al., in press). 1.2 Rationale for the proposed research Although of the three size fractions coarse PM has been studied for the longest amount of time, it is still not completely understood. The reasons for problems in characterizing coarse PM are: 1) coarse PM concentrations are very variable and heavily depend on changes in wind speed and direction; 2) sampling coarse PM can be difficult because the particles can settle out and/or are removed by impaction in the inlet before collection; 3) current measurement o f coarse PM is conducted by the difference method, in which PMio and PM 2.5 are collected on a filter and the difference between the two equals the coarse PM concentration. For the same reasons listed above, it has also become very complex to determine the human exposure to coarse PM. Coarse particles are generally thought to originate from local sources and have very short residence times in the atmosphere, thus a centralized monitoring site w ill not accurately reflect the 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concentrations to which someone living miles away w ill be exposed (Wallace et al., 1997). Additionally, the penetration of coarse PM indoors is very low, so indoor sources of coarse particles are more important when assessing human exposure. Many health studies have been conducted to determine the health effects of coarse PM (Becker et al., 2003; Kleinman et al., 2003; Li et al., 2002). While these have involved the exposure of cell cultures and animals to coarse PM, little to no research has been conducted on human subjects. The development o f a coarse particle concentrator with the potential to deliver concentrated coarse PM at flow rates required for humans is a necessity for future particle research. This is much more important for coarse PM exposure to health effect extrapolation because o f the difficulty in determining ambient coarse PM exposure. The objective o f this thesis is to demonstrate the intricacies involved in characterizing coarse PM by presenting technologies that measure and concentrate coarse particles with the intention o f reducing collection time. By doing this coarse PM can be studied on shorter time scales, and errors from variables such as wind speed and direction can be reduced. Further study of the relationship between coarse and fine PM is also investigated in order to determine the composition of the intermediate PM mode and the contribution of coarse PM to this submode of fine PM. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.3 Theory In order to understand the basics of aerosol science, a few equations must be presented. Because the focus o f the research presented herein centers around the coarse size fraction, certain laws that are not relevant (i.e. diffusion) w ill not be discussed. Because an aerosol is a mixture o f particles in air, it can be described by the equations o f fluid motion. The Reynolds number is a dimensionless number that is characteristic of fluid flow around an object or objects. In the case o f aerosols, the fluid is air and the objects are particles. Simply stated, the Reynolds number is the ratio o f inertial forces to viscous (or frictional) forces acting on each small part of a fluid. The Reynolds number is commonly used as a yardstick to determine whether fluid flow is laminar or turbulent. When Reynolds number is expressed in terms of aerosols, it is written Re = pVd where p = density of the fluid V = relative velocity between the fluid and particle d = diameter of the particle r) = viscosity of the fluid A general differential equation that describes fluid dynamics is the Navier- Stokes equation for incompressible flow. By neglecting gravity and buoyancy forces, it is written as (dU ^ { 8 U 1 f 8U 1 + u + V [ d t ) I dx ) I dy) • • ( f ) dp_ \dxj ( d2U d2U d2U^ dx2 dy2 dz2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in the x direction. Stoke’s law is a solution to the above equation under the assumption that the inertial forces acting on a particle are negligible compared with the viscous forces on that particle. This is a valid supposition for particulate matter in air because the particles in question are very small and are subject to low velocities (Hinds, 1999). Stoke’s Law can be solved from the equation and assumption above to yield: Fd = 3 arjVd By setting the above equation equal to the gravitational force, the terminal settling velocity of a particle can be determined. The simplified form after mathematical manipulation and assuming d < 1 pm and Re < lis: r _ P f s 7 5 \Sjj Since the majority o f this research deals with coarse PM, the equations above w ill serve as an introduction to the behavior of these particles in air. 1.4 Thesis Overview The purpose of this thesis is to demonstrate the importance of accurate sampling and measurement of coarse PM by developing technologies to characterize it. Furthermore, the variability of coarse PM depending on location w ill be validated in order to discourage large-scale averaging and generalizing o f coarse PM on a 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. regional basis. Seven chapters compose this thesis with the first being the introduction and brief layout o f the fundamentals of aerosol science. Chapter two w ill present a study of indoor versus outdoor concentrations of PM in the California desert during a time of maximum penetration. The aim of this study is to prove that exposure to coarse PM is not well represented by a centralized monitoring station that measures PMio. Another benefit of this survey is the analysis of the chemical composition o f coarse PM in a desert environment that is also influenced by the plume that is advected from the LA basin into the desert that lies to the east. In contrast to the measurement of ambient and indoor coarse PM discussed in the second chapter, chapter three w ill describe the development and characterization of a technology to concentrate coarse PM in order to deliver a concentrated aerosol to the subject of a controlled exposure. The coarse particle concentrator was developed to determine the toxicity of coarse particles on human and animal subjects. Consideration was give to the ratio of the minor to total flow through each of ten virtual impactors in order to maximize the efficiency o f the coarse particle collection. Another obstacle that must be overcome is the joining o f ten working virtual impactors via air ducts without creating pressure and velocity fluctuations. Any changes in velocity or pressure drop in the collection manifold w ill lead to losses of coarse PM because of their high mass and thus high inertia and settling velocities. 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter four shifts gears once again as a methodology for measuring continuous coarse PM is developed and characterized. The relevance of this device is based on the finding that coarse PM is very localized and influenced by events that occur over time scales on the order of minutes to hours, such as wind speed and direction. The instrument is the combination of a virtual impactor and an already proven mass measurement technology— the Tapered-Element Oscillating Microbalance (TEOM™ 1400A, Rupprecht and Patashnick, Albany, N Y ). Since this device is the first continuous coarse-only PM monitor, it was characterized with time-integrated samples and compared with a time-of-flight particle counter by assuming the density of ambient particles. Because the virtual impactor’s flow rate is set at fifty liters per minute (LPM ) for the continuous coarse monitor described by the previous chapter, a new PMio inlet (a device which removes particles larger than 10 pm in aerodynamic diameter) is discussed in chapter five. A PMio inlet with a sharp fifty percent cutpoint at 10 pm is necessary for accurate measurement of coarse PM because any particles larger than this that escape into the device can severely bias the total mass concentration. The reason for this is that a particles mass increases with the cube o f its diameter. This chapter presents the development and characterization of a 50 LPM PMio inlet in both laboratory and field conditions. Chapter six concerns an emerging topic in the arena of PM regulation and monitoring. Recent studies have shown that a significant fraction o f coarse PM is collected with fine PM because the coarse mass distribution has a tail that does not 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reach a negligible amount until around 1 |jm. This coupled with the large masses of coarse particles compared with those of fine particles has ignited a debate over whether or not to establish a new PMi standard. The sixth chapter of this thesis presents data collected over two years in the Los Angeles basin at various sites. This data includes monitoring o f coarse and fine PM but also incorporates mass concentrations of PM in the size ranges of 0-1 pm and 1-2.5 pm. The latter size range has been termed the intermediate mode, or intermodal. W hile rural and desert community studies have proven that the intermediate mode is highly influenced by coarse PM, the data collected in the Los Angeles basin suggest that the intermediate mode is in fact also heavily influenced by fine particles that grow by condensation into the 1-2.5 pm size range. Chapter seven of this work is the concluding chapter and ties together the ideas presented herein. Further research and new directions are also presented to the reader. The cohesiveness of the topics outlined above will become apparent as the characterization of and technologies to concentrate and monitor coarse particulate matter are presented. 1 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References Becker, S.; Soukup, J.M.; Sioutas, C.; Cassee, F.R. (2003). “Response o f human alveolar macrophages to ultrafine, fine, and coarse urban air pollution particles,” Exp. Lung Research, 29: 29-44. Dockery, D.W .; Pope, C.A.; Xu, X.; Spengler, J.D.; Ware, J.H.; Fay, M .E.; Ferris, B.G.; Speizer, F.E. (1993). “An association between air pollution and mortality in six U.S. cities,” New Eng. Journ. Med., 329(24): 1753-1759. Friedlander, S.K. Smoke, Dust, and Haze: Fundamentals o f Aerosol Dynamics. New York: Oxford University Press, 2000. Hinds, W.C. Aerosol Technology: Properties, Behavior, and Measurement o f Airborne Particles. New York: John W iley & Sons, Inc., 1999. Kleinman, M .T.; Sioutas, C.; Chang, M.C.; Boere, A.J.F.; Cassee, F.R. (2003). “Ambient fine and coarse particle suppression o f alveolar macrophage functions,” Tox. Letters, 137(3): 151-158. Li, N.; Kim, S.; Wang, M .; Froines, J.; Sioutas, C.; Nel, A. (2002). “Use of a stratified oxidative stress model to study the biological effects o f ambient concentrated and diesel exhaust particulate matter,” Inhal. Tox., 14(5): 459- 486. Li, N.; Sioutas, C.; Cho, A.; Schmitz, D.; Misra, C.; Sempf, J.; Wang, M .; Oberley, T.; Froines, J.; Nel, A. (in press). “Ultrafine particulate pollutants induce oxidative stress and mitochondrial damage,” Submitted to Envir. Health Perspect. Oberdorster, G. (2000). “Toxicology of ultrafine particles: in vivo studies,” Phil. Trans. Roy. Soc. London A, 358 (1775): 2719-2739. Oberdorster, G. (2001). “Pulmonary effects o f inhaled ultrafine particles,” Int. Arch. Occup. Envir. Health, 74(1): 1-8. Pope, C.A.; Dockery, D.W .; Schwartz, J. (1995). “Review of epidemiological evidence o f health effects of particulate pollution,” Inhal. Toxicol., 7: 1-18. Schwartz J. and Neas, L.M . (2000). “Fine particles are more strongly associated than coarse particles with acute respiratory health effects in schoolchildren,” Epidemiology, 11(1): 6-10. 1 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tsai, F.C.; Apte, M .G.; Daisey, J.M. (2000). “An exploratory analysis of the relationship between mortality and the chemical composition of airborne particulate matter,” Inhal. Toxicol, 12(Supplement 2): 131-135. I Wallace, L.; Quakenboss, J.; Rhodes, C. (1997). AW M A/EPA Symposium on the Measurement of Toxic and Related A ir Pollutants. Research Triangle Park, NC: 860-871. Wark, K.; Warner, C.F.; Davis, W .T. Air Pollution: Its Origin and Control. Menlo Park, CA: Addison-Wesley, 1998. Yeh, H.C. and Schum, G.M. (1980). “Models for human lung airways and their application to inhaled particle deposition,” Bull. Math. Biology, 42: 461-480. Yeh, H.C.; Cuddihy, R.G.; Phalen, R.F.; Chang, I-Y . (1996). “Comparisons of calculated respiratory tract deposition o f particles based on the proposed NCRP model and the new ICRP model,” Aerosol Sci. Tech., 25: 134-140. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 2 INDOOR /OUTDOOR RELATIONSHIP AND CHEMICAL COMPOSITION OF FINE AND COARSE PARTICLES IN THE SOUTHERN CALIFORNIA DESERTS Atmospheric Environment 36 (2002) 1099-1110 2.1 Abstract The work presented in this paper examines the characteristics, chemical composition and relationship between indoor and outdoor particulate matter (PM ) in the Coachella Valley, a unique desert area in southern California. Fine (0 - 2.5 pm) and coarse (2.5 - 10 pm) PM concentrations were measured concurrently indoors and outdoors in 13 residences during the winter and spring of 2000. Maximum outdoor PM penetration in indoor environments was expected during this period in the California deserts, as the mild climate minimizes the use o f heating and/or air conditioning. Filter and impaction substrates were analyzed for mass, selected trace elements and metals, as well as elemental and organic carbon content (for fine PM only). Fine PM concentrations accounted, on average, for 74.3 (±11.0)% of the total PMio concentrations indoors, whereas fine PM contributed to 61.3 (± 13.1)% of the outdoor PMio concentrations. The indoor-to-outdoor mass concentration ratios were 0.66 (± 0.27) and 1.03 (± 0.29), for coarse and fine PM, respectively. Chemical analysis o f the filters revealed well-correlated indoor-to-outdoor concentrations of trace elements and metals in the fine PM mode, while lower correlations were obtained for the coarse PM mode. Elemental carbon concentrations indoors were Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.84 (± 0.32) of those measured outdoors, whereas organic carbon concentrations t indoors were on the average 77% higher than outdoors, presumably due to the contributions o f indoor sources. Coarse PM concentrations based on mass, trace elements and metals showed similar trends, with the average indoor-to-outdoor concentrations varying from about 50 to 70%. Although the outdoor air o f the specific study area has been traditionally considered to be rich in coarse particles, the results o f this study suggest that indoor PM concentrations are still dominated by the contribution of fine particles. 2.2 Introduction Epidemiological evidence associating ambient particulate pollution with adverse health effects in humans is extensive (ATS, 1996; US EPA, 1996). Nevertheless, fundamental uncertainty and disagreement persist regarding which physical and chemical properties of particles influence health risks, which pathophysiological mechanisms are operative, and what air quality regulations should be adopted to address the health risks (Vedal, 1997). This lack of understanding stems, in part, from the paucity of reliable data linking personal exposure to observed health outcomes in large-scale epidemiological studies. People spend most of their time indoors. Yet, the majority of particle concentration data is based on measurements conducted outdoors, in one or more central monitoring sites. Outdoor particulate concentrations may not be reliable indicators of indoor and personal particulate exposures (Wallace et al., 1997). For 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. example, poor associations between outdoor and indoor or personal levels may result from the presence of indoor particulates sources. The work presented in this paper therefore examines the indoor and outdoor relationships of particulate matter in the Coachella Valley (hereinafter "the Valley"), a unique desert area in southern California. The Valley is located approximately 80 miles southeast of Los Angeles, between the Mojave and Colorado deserts. Bordered by mountains on the north, east, and west and by the Salton Sea on the south, the Valley intermittently accumulates high particle concentrations, sufficient to warrant designation by the EPA as one of five areas in serious non-attainment with the 24-hour ambient air quality standard for PMio. The Valley is also home to many retirees and attracts large numbers of winter tourists because of its warm climate. Recently, Ostro and Lipsett (Ostro et al., 1999) found that ambient concentrations of particulate matter smaller than 10 pm (PMio) in California's Coachella Valley are associated with an increased risk of daily mortality. Given that about 50-60% of the total PMio mass concentration in that area measured in fixed- site monitoring stations was found to be in the coarse fraction (i.e., 2.5 to 10 pm in aerodynamic diameter), the results of that study raised concerns on the mortality and morbidity effects o f coarse particles. The purpose of this field investigation was to determine the chemical characteristics of outdoor coarse and fine particles in the Valley and compare them to 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. those measured in the epidemiological study subjects’ residences (the study is conducted by the California Environmental Protection Agency), lliis is a cross- sectional field study, investigating the indoor-outdoor PM characteristics in several different residences over three 24-hour periods. Data collected from the study presented here examined the following questions: 1. What are the relationships, for both coarse and fine particles, between ambient and indoor concentrations? 2, What are the major chemical constituents of indoor and outdoor particles? 2.3 Methods 2.3.1 Study Design This investigation of the indoor/outdoor relationships of coarse and fine particles was designed for the unique desert climate in the winter and spring of 2000 for two reasons. First, geologic particles comprise a significant percentage of the total particulate mass throughout the year in the Valley, especially during gusty wind episodes (SCAQMD, 1990). Second, during the winter and spring seasons, typically from January to early May, outdoor PM may have marked influence on indoor air quality in residences whose occupants tend to open windows for prolonged time periods and do not use air conditioning. Thirteen homes of volunteers who concurrently were participating in the epidemiological study noted above were included in this study. Seven locations were in Palm Springs (at the northwest end of the populated corridor) and six in 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Indio, about thirty miles southeast of Palm Springs. The characteristics of each sampling site are shown in Table 2.1. These residences were all within five miles of the fixed site air quality monitors at the north (Palm Springs) and south (Indio) ends of the Valley. These locations were targeted since the Indio area, which is in the Valley's wind belt, typically has higher concentrations of PMio and coarse particles (Ostro et al., 1999), while the Palm Springs area historically has had a mix of coarse and fine particles, including materials transported from the Los Angeles basin, but lower overall concentrations of PMio. The principal sampler utilized in residences of this study was the USC Personal Particle Sampler (PPS), designed for coarse and PM2.5 sampling at 5 LP M (Sioutas et al., 1999). A major feature of this sampler is the separation from the air sample and collection of coarse (2.5-10 pm) particles by impaction on an uncoated quartz substrate, without any particle bounce. The substrate is a quartz fiber filter disk, 0.8 cm in diameter. Collection efficiency of coarse particles as large as 10 pm in aerodynamic diameter exceeds 95% with a quartz fiber filter used as an impaction surface (Chang et al., 1997). Using this uncoated substrate for particle collection is a particularly attractive feature of the PPS, as adhesive coating materials typically used to reduce particle bounce in impactors would likely interfere with the chemical analysis of the substrate. Fine (i.e., <2.5 pm in aerodynamic diameter) particles in the PPS are collected on either a 37-mm Teflon filter (PTFE Teflon, 2 pm pore, Gelman Science, Ann Arbor, M I) or a 37-mm quartz filter (Pallflex Corp., Putnam, 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CT), depending on the type of chemical analysis to be performed after sample collection. Teflon filters were used for gravimetric and elemental analysis, whereas quartz filters were used for gravimetric and organic analysis, described below. TABLE 2.1. Characteristics of Coachella Valley Indoor/ Outdoor Sampling Locations. (P.S. stands for Palm Springs). Sampling period; January - May 2000 Residence No. of occupants City Stove Type Heater Type Type of Analysisa A 1 Indio Gas Electric EC/OC, Elements B 1 P.S. Electric Electric EC/OC, Elements C 1 Indio Electric Gas EC/OC, Elements D 1 P.S. Electric Electric EC/OC, Elements E 4 P.S. Gas Gas EC/OC, Elements F 1 Indio Gas Gas EC/OC, Elements G 1 Indio Electric Electric EC/OC, Elements H 2 P.S. Electric Gas EC/OC, Elements I 2 Indio Gas Electric EC/OC, Elements J 1 Indio Electric Electric EC/OC, Elements K 1 P.S. Electric Electric Elements L 1 P.S. Gas Gas EC/OC M 3 P.S. Gas i Gas EC/OC a EC/OC = elemental and organic carbon analysis At each home a PPS was placed immediately outdoors, such as in the front yard, in a courtyard, or on a balcony. The indoor PPS was placed in either the living room or family room where the residents spent the majority of their time at home. 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Efforts were made to secure the PPS inlet at a height o f 1.0-1.5 m above the ground in each residence in order to avoid potential interferences from excessive resuspension of particles due to residential activities and to also sample aerosol concentrations that would reflect those in a “typical” breathing zone. Each set of indoor/outdoor PPS sampling lasted twenty-three hours in order to incorporate time needed for filter collection and replacement. Two to four 23-hour experiments were conducted in each residence to capture the intra-home variability. 2.3.2 Sample Analysis For mass concentration measurements, the Teflon and quartz filters (for fine particles) as well as the quartz substrates (coarse particles) o f the indoor and outdoor PPS were weighed before and after each field test, using a Mettler 5 Microbalance (M T 5, Mettler-Toledo Inc., Highstown, NJ), under controlled relative humidity (e.g., 40-45%) and temperature (e.g., 22-24 °C) conditions in the facilities of the Aerosol Laboratory at the University of Southern California. Filters were weighed after a 24-hour equilibration period. Laboratory and field blanks were used for quality assurance. Subsequent to weighing, the PPS substrates, collecting coarse PM , were analyzed for trace element and metal content, whereas half o f the PPS filters, collecting PM2.5, were analyzed for trace elements and metals (Teflon filters) and half for elemental carbon (EC) and organic carbon (OC) concentrations (quartz filters). 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Teflon filters were analyzed by means o f x-ray fluorescence (XRF) to determine fine particle concentrations o f selected trace elements and metals. Trace element and metal concentrations for coarse particles were determined by analyzing the quartz PPS substrates by means of inductively coupled plasma-mass spectroscopy (ICP/MS). These analyses were conducted by the Monitoring and Laboratory Division of the California A ir Resources Board. The indoor and outdoor elemental and organic carbon (EC/OC) concentrations o f fine particles were determined by thermo-analysis of the quartz PPS filters. This analytical method is described more elaborately by Fung (1990). It should be noted that particle collection using quartz filters for measurements of the OC content of PM may be prone to artifacts, mostly related to adsorption of gas- phase organics onto the quartz filter. Given the relative similarity between the sampling conditions (i.e., type of sampler, flow rate, duration, mass loading) indoor and outdoors, we expect that these artifacts (if they occurred), would not differ substantially between the outdoor and indoor samples collected concurrently, and thus would have little effect on the I/O comparisons for OC. Moreover, because many atmospherically relevant organic compounds, such as polycyclic aromatic hydrocarbons (PAH) partition between gas and particulate phases, and the gas-to- particle partitioning is temperature dependent, total (i.e., gas plus particle) measurements might be more desirable for source apportionment (Carlton et ah, 1999). 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.4 Results and Discussion 2.4.1 Indoor and Outdoor Coarse and Fine PM Mass Concentrations Results from the field tests comparing indoor to outdoor particle concentrations based on mass and chemical composition are shown in Figures 2.1- 2.6 and summarized in Tables 2.2 and 2.3. Similar to previously published studies on the relationship between indoor-outdoor PM (Colome et al., 1992; Clayton et al., 1993), the value of the coefficient of determination (R2 ) between the indoor and outdoor data was used as an indicator of the degree to which a PM species measured indoors is attributed to infiltration from outdoors. In Figures 2.1 and 2.2, indoor mass concentrations for both coarse and fine particles were plotted against their respective outdoor concentrations. Table 2.2 presents a summary of the indoor and outdoor measurements for coarse and fine particle concentrations. Fine PM concentrations accounted on the average for 74.3 (±11.0)% of the total PMio concentrations indoors, whereas fine PM contributed to 61.3 (+ 13.1)% of the outdoor PMio concentrations. Previously published SCAQMD data indicated that geologic (which are primarily coarse) particles in the California deserts constitute about 50-60% of the total PMio (SCAQMD, 1990), thus somewhat higher than those observed in our study. It needs to be emphasized, however, that the outdoor measurements collected in the SCAQMD study were taken in a variety o f nonresidential locations at which intermittent gusty winds caused substantial resuspension of particles. Such resuspension may not occur in the immediate 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. proximity o f residences, which usually have planted courtyards and regular daily i watering of the ground. Figure 2.1 shows that coarse indoor concentrations are markedly lower than those outdoors. Pairwise comparisons between indoor and outdoor coarse PM concentrations indicated that outdoor concentrations were significantly higher (p=0.005) than those indoors, with the average indoor-to-outdoor mass concentration ratio equal to 0.66 (± 0.27). Figure 2.1 also shows a relatively weak correlation between the outdoor and indoor data (R2 = 0.35). The weak indoor-to-outdoor PM association suggests that a substantial fraction of coarse particles are generated by indoor sources and activities, such as dusting, cleaning, washing and resuspension, all of which depend on the residents o f each individual home. Our results are in agreement with the findings of a recent study by Abt et al. (2000), in which indoor activities, such as vacuuming, dusting, washing and carpet cleaning, contributed from 50% to 80% of the indoor concentrations of 2-10 pm particles. The same study also showed that the penetration o f outdoor 2-10 pm particles indoors varies from 10% to 40% and generally decreases with size. These two factors would be responsible for decreasing the R between the indoor and outdoor data. Our results (excluding the one outlier, which increases R2 to 0.53) are also in very good agreement with those in a study by Monn and Becker (1999), in which the indoor-to-outdoor coarse PM ratio was 0.7 with the coefficient o f multiple determination equal to 0.55. 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.2 shows a plot of the indoor versus outdoor concentrations for fine particles. Pairwise comparisons between these concentrations (p = 0.7) indicate that the indoor fine PM levels are very similar to those measured outdoors. The average 30.00 25.00 y = 0.30x + 2.95 R2 = 0.35 =L 20.00 15.00 i » 10.00 5.00 outlier 0.00 0 5 10 15 20 25 30 Outdoor Concentration (ng/m3) Figure 2.1. Indoor versus outdoor coarse particle mass concentrations. indoor-to-outdoor fine particle mass concentration ratio was 1.03 (± 0.29). However, the data plotted in Figure 2.2 indicate that outdoor concentrations can explain only about 37% o f the variation of the indoor concentrations, thereby suggesting that there may be significant contributions by indoor sources to the overall PM2.5 concentrations in the 13 residences monitored in this study. [Excluding the one outlier shown by the arrow in Figure 2.2 increases the indoor-outdoor correlation to R = 0.51, while the resulting regression line changes to y = 0.81 x + 3.4]. 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Additional differences in the home structural characteristics, number of occupants and activity patterns of individuals in each o f the 13 residences also contribute to the overall variability in the indoor PM concentrations. 30.00 outlier 25.00 - o > 20.00 ■ c o 3 5 E 4 - 1 y = 0.74x + 4.26 R2 = 0.37 c 15.00- 0 ) o c o W 1 0 .0 0 - o o ■ o - 5.00 - 0.00 0.00 5.00 30.00 10.00 15.00 20.00 25.00 Outdoor Concentration (ug/m3 ) Figure 2.2. Indoor versus outdoor fine particle mass concentrations. 2.4.2 Indoor and Outdoor Elemental and Organic Carbon Concentrations for Fine PM Figure 2.3 shows a plot of indoor-to-outdoor elemental carbon concentrations. Indoor EC concentrations are generally lower than those outdoors (with one notable exception observed for one o f the 13 residences), with the average indoor-to-outdoor EC concentration being equal to 0.85 (± 0.46). The coefficient of multiple determination between indoor and outdoor EC concentrations, Rz, is 0.45. It should be noted that omission of the outlier increases R2 to 0.71, suggesting that a 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. substantial fraction of the variation in indoor EC concentrations can be attributed to that o f the outdoor EC concentrations, at least in all but one residence. At this moment, we have no explanation for the elevated EC levels observed in the “outlier” residence. For the remaining data, the high R2 value indicates that a substantial fraction o f indoor elemental carbon penetrates from outdoors. This is consistent with results reported by Jones et al. (2000), who found that elemental carbon originates outdoors, mostly from vehicular emissions, except in homes in which smokers are present. (Sampling in our study was conducted exclusively in non-smoking residences.) Elemental carbon (soot) is a byproduct o f incomplete combustion emitted primarily in exhaust from road traffic (QUARG, 1993). These carbon particles can exist singly or as agglomerates with aerodynamic diameters mostly in the ultrafine range, between 0.05-0.2 pm (Seinfeld and Pandis, 1998). According to the recently published data by Abt et al. (2000), this size range has the highest outdoor-to-indoor penetration, varying generally from 75-100%; hence, in the absence of smoking and wood burning, most of the elemental carbon measured indoors would be expected to originate outdoors. Figure 2.4 shows a plot of the indoor to outdoor concentrations o f organic carbon, including the linear regression line. Organic carbon was the most significant component of both outdoor and indoor PM2.5 mass concentrations, accounting for 0.41 (+ 0.14) and 0.61 (± 0.17) of the total fine PM mass concentrations measured outdoors and indoors, respectively. Indoor organic carbon concentrations were 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. found to be substantially higher than those outdoors in all 13 residences of this study, with the average indoor to outdoor PM2.5 organic carbon concentration ratio being 1.77 (± 0.36). The high indoor concentrations are undoubtedly due to the 1.40 1.20 - y = 0.74x + 0.03 R2 = 0.45 n E TO 1.00 - c o - 0.80- 2 • M C 0 o c o o k o o ■ o c 0.60 0.40 - 0.20 - - * 0.00 1.40 1.20 0.80 1.00 0.00 ' 0.20 0.40 0.60 Outdoor Concentration (^g/m3 ) Figure 2.3. PM2.5 elemental carbon (EC) indoor/outdoor comparison by home. contribution of indoor sources, including cooking, waxes, cleaners/polishes, plasticizers, and pesticides. Abt et al. (2000) reported that cooking was the predominant source of indoor particles smaller than 0.5 pm in diameter, contributing on average about 0.27 pm3/cm3/min. The same study showed that cooking generated substantial amounts of super-micrometer particles. However, such particles have also significantly higher decay rates than those smaller than 0.5 pm. 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The data plotted in Figure 2.4 also indicate that a substantial fraction of the variability in the indoor data can be explained by the outdoor concentrations (R2 =0.63), thereby suggesting that a relatively high percentage of outdoor organic carbon also penetrates the home. The majority of particle-bound organic compounds 30.00 outlier _ 25.00 - 20.00 - c 15.00- y = 1.46x + 1.97 R2 = 0.63 10.00 - 0.00 0.00 2.00 4.00 6.00 14.00 18.00 8.00 10.00 12.00 16.00 Outdoor Concentration (ng/m3) Figure 2.4. PM2.5 organic carbon (OC) indoor/outdoor comparison by home. exist typically as submicrometer particles (McMurry and Zhang, 1989), often partitioned in a bimodal distribution with one peak in the 0.4-1.0 pm range and another between 0.1-0.2 pm (Pickle et al., 1990; Mylonas et al., 1991). The latter mode normally dominates the ambient OC size distribution (Hildemann et al., 1991; Hildemann et al., 1993). The fraction of outdoor particles in this size range reported to penetrate indoors varies between 80% and 100% (Abt et al., 2000). Thus, in addition to the considerable contribution of indoor sources to the overall organic 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. carbon in fine PM, a substantial fraction o f the particle-bound organic carbon measured indoors infiltrates from outdoors. 2.4.3 Indoor and Outdoor Trace Element and Metal Concentrations for Fine PM Only these trace elements and metals in fine PM found in concentrations greater than three times the XRF limit of detection were included in this analysis. Thus, results for only the following metals are available: aluminum, silicon, sulfur, potassium, calcium, titanium, iron, and zinc. The indoor concentrations were plotted against the outdoor concentrations and are shown in Figures 2.5a-h. O f these particle-bound metals, sulfur, titanium, zinc, iron and aluminum show high correlations (R2 ranging from 0.59 to 0.85), which suggests that a significant fraction of their indoor concentrations can be attributed to infiltration of outdoor particles containing these metals. Similar conclusions can be drawn for silicon, calcium and potassium, which have lower yet still significant correlations (R2 o f 0.50, 0.48, and 0.46 respectively). Sulfur is a classic marker of atmospheric outdoor aerosols (Jones et al., 2000; Clayton et al., 1993) with no known indoor sources, which explains the very high correlation obtained between its indoor and outdoor concentrations (R2 =0.85). Sulfur exists mostly as sulfate fine particles (<1 pm) that form from the oxidation of sulfur dioxide (Clayton et al., 1S93; US EPA, 1975; Moschandreas et al., 1979). The indoor-to-outdoor (I/O ) concentration ratio of sulfur was calculated to be 0.80 (± 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.18), which is consistent with the findings of Jones et al. (2000) and the PTEAM study (Clayton et al., 1993). The somewhat lower indoor levels are due to some particle removal during penetration indoors. The I/O sulfur ratio probably provides a reasonable estimate of the I/O species ratios for species that infiltrate indoors from the ambient. The near-equality of the indoor and outdoor fine PM concentrations obtained in this study is probably a coincidence occurring because o f the higher OC concentrations measured indoors, which compensate for the particles lost during infiltration. Similarly, the results shown in Figure 2.5b indicate that the indoor concentrations of aluminum are highly correlated with those measured outdoors (R2 = 0.85), but on the average 0.64 (± 0.23) times the outdoors values. The high correlation between the indoor and outdoor data indicates that this metal likely originates outdoors; however, probably because of the larger particle size associated with Al, a smaller fraction infiltrates the house relative to sulfur. Particle-bound metals of crustal origin, such as silicon and calcium, are also expected to display indoor-to-outdoor characteristics similar to those observed for aluminum (Clayton et al., 1993). This probably explains the similar indoor-to-outdoor concentration ratios obtained for silicon and calcium (i.e., 0.69 ± 0. 31 and 0.66 ± 0.24, respectively). Since we expect all of these metals to be found primarily in the coarse mode, the sizes of these particles in the fine PM mode are probably greater than 1 pm, thus representing a “tail-end” of the coarse outdoor mode. Coarse particles settle inside 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and do not become resuspended except transiently, so indoor concentrations o f the associated metals are expected to be lower than those outdoors. The high correlations between indoor and outdoor concentrations obtained for iron, titanium and zinc indicate that these metals are also more likely to originate from outdoor sources with approximate indoor-to-outdoor concentration ratios of 0.69 (± 0.22), 0.63 (± 0.25), and 0.91 (± 0.29), respectively. Moschandreas et al. (1979) reported similar results. That study, however, did not find a strong correlation between PMio indoor and outdoor iron concentrations, probably because of the substantial fraction o f iron in the coarse PM mode, both indoors and outdoors, which would tend to decrease these correlations. The higher indoor-to-outdoor Zn concentration ratio is probably due to the higher infiltration rate of outdoor particle- bound Zn indoors. Jones et al. (2000) and Clayton et al. (1993) found similar indoor-to-outdoor ratios for zinc. Zinc is also primarily an outdoor metal, whose only known indoor source is smoking (Wallace et al., 1997; Jones et al., 2000). Zinc is known to exist in very small particles (mostly smaller than 0.5 pm) that originate from anthropogenic sources such as fossil fuel combustion (Seinfeld and Pandis, 1998). The small size of these particles explains the high penetration values from outdoors. Indoor and outdoor potassium concentrations were moderately correlated (R2 = 0.46), with an average indoor-to-outdoor concentration ratio of 0.79 (± 0.27). These results are inconsistent with those of Moschandreas et al. (1979) who observed variations of potassium to be dependent mostly on indoor activities, such as smoking 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and wood burning (hence no correlations were found between indoor and outdoor K concentrations). A possible explanation for this difference might' be related to differences in the study locations, as potassium may be a soil constituent of the deserts and is also a marker for burning organic matter (i.e. cooking) (Moschandreas et al., 1979). C ° 1.20 a) y«0.627x + 0.03821 R2« 0.8471 I fO 0.90 3 0.60 Outdoor Concentration (^g/m ) b) 0.50 y * 0.4097x + 0.0096 R2 * 0.8477 0.40 ~ 0.30 « 0.20 0.001= 0.00 0.20 0.30 O utdoor Concentration (pg/m3) 0.50 0.40 0.10 C — v II) « * 4 5 1 O o c 02Q 0.40 c) Outdoor Concentration (ng/m ) 0.40 0.30 o « C ■§. 020 O P > 0.00 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Outdoor Concentration (}xg/m ) d) Figure 2.5. PM2.5 trace element and metal indoor/outdoor comparison by homes: (a) sulfur, (b) aluminum, (c) silicon, (d) calcium, (e) iron, (f) titanium, (g) zinc, (h) potassium. 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.5 y * 0.5754x ♦ 0.0972 R2 = 0.7057 O to 1.5 0.5 0.5 0 1.5 2.5 1 2 e) Outdoor Concentration (ug/m3 ) y * 0.4058X + 0.0032 R2* 0.5875 0.03 o o 0.01 c 0.035 0.005 0.01 0.015 0.02 0.025 0.03 f) Outdoor Concentration (ug/m ) 0.015 y = 0 .9 7 5 6 x -0.00021 R2* 0.6186 | 0.012 n "* 0 .< 0.003 g) 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 Outdoor Concentration (ng/m3 ) 0.25 y * 0.4447x + 0.02341 R2« 0.4579 0.20 0 ) * V 0.15 o < 3 0<1° 0.05 0.00 o.oo 0.05 0.15 0.25 0.10 0.20 h) Outdoor Concentration (i*g/m ) Figure 2.5. Continued. 2.4.4 Indoor and Outdoor Trace Element and Metal Concentrations for Coarse PM To ensure accuracy, similar to the analysis o f fine particles, the concentrations o f metals in coarse PM were only included if greater than three times the limit o f detection of ICP-MS. The results for magnesium, aluminum, silicon, sulfur, potassium, calcium, and iron are summarized in Figures 2.6 a-g, with the average ratios of indoor-to-outdoor coarse PM concentrations listed in Table 2.3. 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Indoor to outdoor concentrations for coarse PM metals display considerably lower correlations than those observed for fine particles. Nevertheless, indoor concentrations still moderately depend on their outdoor concentrations. Silicon and magnesium have the highest indoor/outdoor correlations with R2 values of 0.47 and 0.48, respectively, while the rest of the metals, except sulfur, have remarkably similar R2 values, ranging between 0.37 - 0.42. The R2 values for most metals are in agreement with that obtained for mass concentration (R2 = 0.35). The data listed in a) y * 0.2146x + 0.296 RJ » 0.0576 1.2 0 0.8 0 0.2 04 0 . 6 0.8 1 1.2 1.4 Outdoor Concentration (ng/m ) b) 0.5 C O y * 0.2993x + 0.02211 R3 > 0.4820 C ^ 0 ) r > - I o 3 O o t 0 -2 u o o ■ o C 0.3 0.1 0 0.1 0.2 0.3 0.4 0.5 Outdoor Concentration (ng/m ) 0.8 0.7 0.6 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0 . S 0.6 0.7 C ) Outdoor Concentration (jig/m3 ) d) n . o.8 0 0.5 1.6 1 2 Outdoor Concentration (ng/m3 ) Figure 2.6. Coarse particle trace element and metal indoor/outdoor comparison by homes: (a) sulfur, (b) magnesium, (c) aluminum, (d) silicon, (e) calcium, (f) iron, (g) potassium. 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. y=0.5614x +0.1053 R2=0.3778 Q4 0 0 5 1 1.5 2 e) Outdoor Concentration (ng/m 3 ) y=0.2146x.+0.1001 R2 =0.3611 1.6 C n . 0 0 ) C 1 2 o t r * ? .3 * a s 0 4 0 0.4 0 8 1.2 1.4 1.6 0 2 0 6 1 f) Outdoor Concentration (jig/m 3) y * 0.3781x ♦ 0.0245 R* » 0.4207 C o 0.25 | = I ° ? o 3 0.15 O o ■ o c 0.05 0.1 0 0.15 0.2 0.25 0.3 Outdoor Concentration (ng/m3 ) g) Figure 2.6. Continued. Table 2.3 also indicate that the indoor-to-outdoor ratios for aluminum, silicon, sulfur, potassium and calcium are remarkably similar, varying from about 0.57 to 0.69. The indoor-to-outdoor concentration ratios for magnesium and iron are substantially lower (i.e., 0.42 and 0.38, respectively), suggesting that these two metals might be associated with somewhat larger particles than the rest of the particle-bound metals measured indoors. 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE 2.2. Descriptive Statistics for PM with Respect to Sampling Location and Size Cut. Location and PM type No. samples Mean Mass Cone. p.g/m3 Median Mass Cone. M -g/m 3 Range Indoor— Fine PM 39 15.45 13.50 4.21-49.30 Outdoor-Fine PM 39 15.02 14.63 3.66-34.43 Indoor— Coarse PM 39 5.63 4.55 0.57-20.22 Outdoor— Coarse PM 39 8.61 8.42 0.64-24.71 The data plotted in Figure 2.6 show that there is no correlation between coarse particle-bound sulfur concentrations measured indoors and outdoors. Sulfur in ambient air o f the Coachella Valley was mostly found in the fine PM mode (which accounted for approximately 80 ± 5% of total S by mass). The lack of correlation between the indoor and outdoor sulfur data suggests that indoor sources, such as gypsum found in certain wallboards or resuspension of previously deposited sulfate particles may be important contributors to sulfur measured in coarse PM indoors. The considerably lower metal concentrations measured indoors compared to outdoors are undoubtedly due to the lower infiltration rates as well as higher deposition velocities of these particles (Abt et al., 2000). Many processes taking place within a house such as cleaning, dusting, washing, and vacuuming have been shown to affect coarse particle concentrations. As these processes vary randomly between individual homes, correlations between indoor and outdoor coarse particles would be expected to be low. 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE 2.3. Average (± standard deviation) ratio of indoor-to-outdoor particle concentrations for selected trace elements and metals; N = number of paired indoor and outdoor data collected simultaneously PM species N Indoor to Outdoor Fine PM Ratio N Indoor to Outdoor Coarse PM Ratio Mass 35 1.03 (± 0.29). 35 0.66 (±0.27). Elemental 17 0.85 (± 0.46) N /A Carbon Organic 18 1.77 (± 0.36) N /A Carbon Magnesium N /A 33 0.42 (±0.14) Aluminum 18 0.64 (± 0.23) 34 0.58 (± 0.22) Silicon 18 0.69 (+ 0.31) 34 0.69 (± 0.20) Sulfur 18 0.80 (±0.18) 35 0.68 (± 0.21) Potassium 18 0.79 (± 0.27) 33 0.57 (±0 .1 8 ) Calcium 17 0.66 (± 0.24) 35 0.68 (±0.15) Titanium 17 0.63 (± 0.25) N /A Zinc 14 0.91 (±0.29) N /A Iron 18 0.69 (± 0.22) 33 0.38 (±0.15) 2.5 Summary and Conclusions This study identified the concentrations and composition o f both indoor and outdoor air in 13 residencies of the Coachella Valley, CA during the winter and spring of 2000, a period during which maximum outdoor PM penetration in indoor environments was expected. These data w ill be incorporated in an ongoing epidemiological study to examine associations between any o f the PM measurements and health effects o f elderly people who live in this area. Fine PM concentrations 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. indoors, whereas fine PM contributed to 61.3 (± 13.1)% of the outdoor PMio i concentrations. The indoor-to-outdoor mass concentration ratios were 0.66 (± 0.27) and 1.03 (± 0.29), for coarse and fine PM, respectively. Although the outdoor air of the specific study area has been traditionally considered to be rich in coarse particles, the results of this study suggest that indoor PMio concentrations are still dominated by the contribution of fine particles. Acknowledgements This work was supported by the California Public Health Institute through Contract # 849A-8705-S3051 to USC (primary funding source to PHI: EPA STAR Grant # R 826783-01-0). The research described in this article has not been subjected to the Agency’s required peer and policy review and therefore does not necessarily reflect the views of the Agency and no official endorsement should be inferred. The authors would like to express their gratitude to M r. Michael Poore, Chief Chemist, California A ir Resources Board, for providing the elemental composition data obtained by means of ICPMS. 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References Abt, E.; Suh, H.; Catalano, P.; Koutrakis, P. (2000). Relative contribution of outdoor and indoor particle sources to indoor concentrations. Environ. Sci. TechnoL, 34: 3579-3587. American Thoracic Society. (1996). State o f the art: Health effects of outdoor air pollution. Am. J. Respir. Crit. Care Med., 153: 3-50. Carlton, A.G.; Turpin, B.J.; Johnson, W .; Buckley, B.T.; Simcik, M .; Eisenreich, S.J; Porcja, R.J. (1999). Microanalysis methods for characterization of personal aerosol exposures. Aerosol. Sci. Technol., 31: 66-80. Clayton, C.; Perritt, R.; Pellizzari, E.; Thomas, K.; Whitmore, R.; Wallace, L.; Ozkaynak, H.; Spengler, J. (1993). Particle total exposure assessment methodology (PTEAM ) study: distributions o f aerosol and elemental concentrations in personal, indoor, and outdoor air samples in a southern California community. J. Exposure Anal. Environ. Epidem., 3: 227-250. Colome, S.; Kado, N.; Jaques, P.; Kleinman, M . (1992). Indoor-outdoor air pollution relations: particulate matter less than 10 pm in aerodynamic diameter (PM10) in homes of asthmatics. Atmos. Environ., 26A: 2173-2178 Fung, K., (1990). Particulate carbon speciation by M N O 2 oxidation. Aerosol. Sci. Technol., 12: 122-127. Hildemann, L.M .; Cass, G.R.; Mazurek, M .A .; Simoneit, B.R.T. (1993). Mathematical modeling of urban organic aerosols: properties measured by , high-resolution gas chromatography. Environ. Sci. Technol., 27: 2045-2055. Hildemann, L.M .; Markowski, G.R.; Jones, M.C.; Cass, G.R. (1991). Submicrometer aerosol mass distributions o f emissions from boilers, fireplaces, automobiles, diesel trucks, and meat cooking operations. Aerosol. Sci. Technol., 14: 138-152. Jones, N.C.; Thorton, C.A.; Mark, D.; Harrison, R.M . (2000). Indoor/outdoor relationships of particulate matter in domestic homes with roadside, urban and rural locations. Atmos. Environ., 34: 2603-2612. McMurry, P.H. and Zhang, X.Q. (1989). Size distributions of ambient organic and elemental carbon. Aerosol. Sci. Technol., 10: 430-437. 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Monn, C. and Becker, S. (1999). Cytotoxicity and pro-inflammatory cytokines from human monocytes exposed to fine (PM2.5) and coarse particles (PM2.5-10) in indoor and outdoor air. Toxicol. Apply Pharmacol., 155: 24. Moschandreas, D.J.; Winchester, J.W.; Nelson, J.W.; Burton, R .M . (1979). Fine particle residential indoor air pollution. Atmos. Environ., 13: 1413-1418. Mylonas, D.T.; Allen, D.T.; Ehrman, S.H.; Pratsinis, S.E. (1991). The sources and size distributions of organonitrates in the Los Angeles aerosol. Atmos. Environ., 25A: 2855-2861. Ostro, B.D.; Hurley, S.; Lipsett, M.J. (1999). A ir pollution and daily mortality in the Coachella Valley, California: A study o f PMIO dominated by coarse particles. Environ. Res., 81: 231-238. Pickle, T.; Allen, D .T.; Pratsinis, S.E. (1990). The sources and size distributions of aliphatic and carbonyl carbon in Los Angeles aerosol. Atmos. Environ., 24: 2221-2228. QUARG, (1993). Urban A ir Quality in the United Kingdom. Quality of Urban A ir Review Group, Department of Environment, London. Seinfeld, J. and Pandis, S. Atmospheric chemistry and physics. New York: John Wiley & Sons, Inc., 1998. Sioutas, C.; Chang, M .C.; Kim, S.; Ferguson, S.T.; Koutrakis, P. (1998). Design and experimental characterization o f a PMi and a PM2.5 personal sampler. J. Aerosol Sci., 30: 693-707. South Coast Air Quality Management District, (1990). Final state implementation plan for PMio in the Coachella Valley. El Monte, CA. U.S. EPA, (1996). A ir Quality Criteria for Particulate Matter, April (EPA/600/P- 95/001cF). Office of Research and Development, Washington, DC. Vedal, S., (1997). Ambient particles and health: lines that divide. J. Air Waste Manage. Assoc., 47: 551-581. Wallace, L.; Quakenboss, J.; Rhodes, C. (1997). In AW M A/EPA Symposium on the Measurement o f Toxic and Related A ir Pollutants. Research Triangle Park, N.C., 860-871. 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 3 DEVELOPMENT AND EVALUATION OF A COMPACT, HIGHLY EFFICIENT COARSE PARTICLE CONCENTRATOR FOR TOXICOLOGICAL STUDIES Aerosol Science and Technology 36 (2002) 492-501 3.1 Abstract A high-efficiency Coarse-mode Particle Concentrator (CPC) has been developed and evaluated in the laboratory as well as validated by performing field experiments at the University o f Southern California, in Los Angeles, CA, and in Bilthoven, the Netherlands. The CPC operates with a total intake flow o f 1000 LPM. The minor flow rate, containing the concentrated coarse-mode particles (2.5 - 10 pm), can be adjusted from 33 to 120 LPM in order to enrich ambient coarse PM concentrations by a factor of 8 to 30, depending on the desirable exposure level and flow rate needed. The laboratory evaluation o f the virtual impactors at three minor flow rates (3.3, 7 and 10 LPM, respectively) indicated that extremely efficient concentration enrichment was obtained for 2.5 - 10 pm particles. In the field tests, the CPC operated at a minor flow rate of 33 LPM, and the mass obtained was compared to the mass collected by a reference sampler, a (rotating) Micro-Orifice Uniform Deposit Impactor (M O U D I), which sampled at 30 LPM. Concentration enrichment factors in the range of 26-30 were achieved based on particle mass, sulfate, nitrate as well as selected trace element and metal concentrations (A l, Si, Ca, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fe, K, Mn, Cu, Zn, Ti). CPC and M O U D I concentrations were highly correlated for « • all species, with R in the range of 0.74 to 0.89. The use of round (compared to rectangular geometry) nozzle virtual impactors in the CPC results in a high concentration efficiency, which reduces the CPC size as well as the power requirement that is required for its operation. The compact size o f the CPC makes it readily transportable to desired locations for exposures to coarse-mode particles derived from different sources and thus o f a varying chemical composition. 3.2 Introduction Abundant epidemiological literature has indicated a significant relationship between ambient particulate matter (PM) and important clinical end-points, such as respiratory symptoms (including asthma attacks), respiratory-related clinic/emergency room encounters and hospitalization, as well as cardiovascular morbidity and mortality (Dockery et al., 1989; Pope et al., 1991; Koenig et al., 1993;). An average 10 pg/m3 increase in PMio is typically associated with a 1- 10% increase in respiratory symptoms at PMio levels near or even below 150 pg/m3 and with lung function declines o f a s much as 7% during 24-hr PMio concentrations exceeding 150 pg/m3 (Ostro, 1993; Pope et al., 1995). Despite the growing evidence of particulate-related health effects, the paucity of information about specific biological mechanisms remains a critical missing link. 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Although a few studies using artificial multi-component fine particle aerosols (Amdur and Chen, 1989; Anderson et al., 1992; Kleinman et al., 1995 and 2003; Bolarin et al., 1997; Arts et al., 2000) have demonstrated mild effects in animals, they have not consistently provided support for a causal relationship between serious health effects in humans and realistic exposure levels. This discordance between the outcomes of laboratory and epidemiological studies may indicate that such artificial particles do not truly replicate the adverse effects o f the complex and heterogeneous mixtures that occur in ambient air. The recent development of fine particle concentrators based on the principle o f virtual impaction (Sioutas et al., 1995a,b; Sioutas et al., 1997) or centrifugation (Gordon et al., 1999) has made it possible to perform laboratory exposures with “real-life” ambient aerosols at highly increased, yet still environmentally realistic, particle concentrations. Some preliminary results using these technologies have been reported (Godleski et al., 1996; Clarke et al., 1999; Gavett et al., 1999; Ghio and Devlin, 1999; Urch et al., 1999), suggesting physiological toxic responses to concentrated ambient particle exposures in laboratory animals and subtle responses in human volunteers. This new line of investigation w ill hopefully eventually lead to coherence between in vivo studies and epidemiological evidence. The aforementioned particle concentrator technologies primarily concentrate the accumulation mode (i.e., 0.2 - 2.5 pm) of atmospheric aerosols. Coarse PM may 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. also consist of several potentially toxic components, such as resuspended particulate I matter from paved and unpaved roads, industrial materials, brake linings, tire residues, trace metals, and bioaerosols. A considerable fraction o f these particles may deposit in the upper airways and to a lesser extent into the lower airways, and may be responsible for the exacerbation of asthma. Recent data from a small number of epidemiological studies indicate that, apart from— or in addition to— the fine PM fraction, health effects may also be closely associated with the coarse PM fraction and sometimes even to a larger extent (Ostro et al., 1999, Kleinman et al., 2003) than PM2.5. In vitro studies with human monocytes (Monn and Becker, 1998; Becker et al., 1996) show that cellular toxicity and inflammation may also be associated with the coarse fraction (2.5-10 (im) and its biological components. Also, in vitro data from Homberg et al. (1998a,b) on genotoxicity o f ambient fine and coarse mode PM collected from an urban area characterized by a high traffic density suggests that coarse mode PM may have comparable or even higher activity. Collectively, these studies indicated that the coarse mode PM might still contribute to a certain extent to observed health conditions, especially those occurring in the higher airways, like asthma. To investigate real world ambient coarse mode particles in experimental studies, the research presented here describes the development o f a high concentration efficiency coarse particle concentrator (CPC), extended from a previously developed single-nozzle, portable coarse particle concentrator (Kim et al., 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2000). The scaled up CPC maintains the advantage o f portability and compact size i (80cmx75cmx45cm) while increasing coarse particle concentrations by a factor up to 40, and it can be readily used for human and/or animal exposure studies. Detailed comparisons between concentrated and ambient coarse aerosols based on mass, sulfate, nitrate, and selected trace elements and metals were performed. The flexibility o f varying the minor-to-total flow ratio of the CPC allows for alternating the desired exposure concentration level for animal exposures as well as for higher output flow rates needed when conducting human exposures. 3.3 Methods The CPC consists of ten single-nozzle virtual impactors (Figure 3.1) developed by the department o f Civil Engineering o f the University of Southern California. These single-nozzle virtual impactors are placed in a 2 by 5 array (Figure 2a and 2b). A 90° elbow with inside diameter of 1.8 cm is connected to the inlet of each virtual impactor. Each virtual impactor operates at an intake flow rate of 100 LPM and therefore comprised a total intake flow rate o f 1,000 LPM. The dimensions o f the 90° elbow inlet were chosen to yield a theoretical 50% removal efficiency o f 10 pm particles (PMio) at a flow rate o f 100 LPM based on the well- established impaction theory (Marple and Liu, 1974). The 50% cut point can be estimated from the Stokes number, St, defined as (Hinds, 1982): 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. where dp , pp , Cc are the particle diameter, density and slip correction, p is the air viscosity (1.81x1c4 g/cm*sec), Uj is the velocity through the 90° elbow, and d o is the inside diameter o f the elbow (do = 1.8 cm). The St corresponding to 10 pm particles is 0.24, based on the nozzle dimensions and the flow rate through each nozzle, which is close to the value typically corresponding to the 50% cut point of round-nozzle impactors (Marple and Liu, 1974). ^ 6.5 cm 0.56 cm 0.37 cm 30 deg Concentrated coarse particles Ambient PM inlet Collection nozzle Acceleration nozzle ► Fine particles Figure 3.1 Schematic of the round jet impactors of the mulitnozzle CPC. 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Particles smaller than 10 jam in aerodynamic diameter are drawn through the i virtual impactor and become accelerated through a circular nozzle, which was designed to have a theoretical 50% cut point at about 2.0 pm for an intake flow rate o f 100 LPM (Sioutas et al., 1999; Kim, et al., 2000). Coarse-mode particles (2.5 - 10 pm) cross the deflected air streamlines and are drawn through the collection nozzle (minor flow). Particles smaller than the cut point o f the virtual impactor are diverted through the major flow. The ten minor flows are joined at the center o f the CPC and lead to a 5-cm diameter tube, which can be connected to an animal or human exposure chamber (Figures 3.2a and 3.2b). The minor flow rate can vary from 3~20% o f the intake flow rate, depending on desired exposure concentration level and/or exposure flow rate needed. Two major flow ducts, each 2.54 cm in diameter, were each connected to five of the virtual impactor major flows (i.e., 500 LPM) and placed on either side of minor flow, as shown in Figures 3.2a and 3.2b. PM10 elbow Inlet virtual Impactors minor flow Ring support plate minor flow Figure 3.2a. Side view of the 10-nozzle Coarse Particle Concentrator. 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cap sealed ends major flow of each VI s' 2" open to chamber (minor flow) to major flow pump ring support minor flow of each VI virtual impactor holding plate 80 cm Figure 3.2b. Side view of the 10 nozzle CPC. 3.3.1 Laboratory Characterization o f the Virtual Impactors The first series of experiments was conducted in the laboratory to investigate the relationship between the concentration enrichment as a function of particle size and minor-to-total flow ratio. This relationship was investigated for each individual virtual impactor. Briefly, monodisperse aerosols in the size range of 1 to 10 pm were generated by atomizing dilute aqueous suspensions o f fluorescent polystyrene latex particles (Polysciences Inc., Warrington, PA) with a constant output nebulizer 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (HEART, VORTRAN Medical Technology, Inc., Sacramento, CA) at a rate o f 15 LPM. 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 then drawn through a tube containing ten Po-210 neutralizers that reduce particle charges prior to entering the virtual impactor. Monodisperse particles were subsequently drawn through the 90° elbow and entered the virtual impactor. For particles in the range of 1 to 5 pm, a nephelometer (DataRAM, R AM -1, M IE , Inc., Billerica, M A ) was used to first measure the mass concentration of the generated aerosols prior to entering the 90° elbow of virtual impactors. The DataRAM was subsequently connected downstream o f the minor flow o f the virtual impactor to measure the mass concentration of the aerosols after concentration enrichment. The measurements were repeated at least three times, and the average concentration enrichment was determined as a function of particle size. The contributions from background ambient concentrations before and after the enrichment were recorded and subtracted from those o f the input and concentrated aerosols prior to determining the collection efficiencies at the given particle size. It should be noted that indoor air levels were on the order of 7 - 15 pg/m3 , and substantially smaller than those o f the generated aerosols (prior to concentration), which varied from 170 to about 500 pg/m3 . Therefore the contributions of the indoor aerosol to the overall concentrations measured upstream of- and in the minor flows of the virtual impactors were considered insignificant. 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Concentration enrichment for 5 to 9 jam particles was determined by comparing the mass collected on a glass fiber filter (2 pm pore, Gelman Science, Ann Arbor, M I) connected to the minor flow of a virtual impactor, and the mass of a similar glass fiber filter in parallel to the test system to measure the concentration of the generated aerosol. The filter sampling in parallel was connected to a pump operating at 30 LPM. At the end o f 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, M A ) to determine particle concentration. Concentration enrichment for each particle size was defined as the ratio o f the concentration measured in the minor flow o f the virtual impactor to that of the aerosol immediately upstream of the virtual impactor inlet. Each virtual impactor was tested at three different minor-to-total flow ratios. The total flow was kept constant at 100 LPM, whereas the minor flows were adjusted to 3.3, 7 and 10 LPM, resulting in ideal enrichment factors o f 30, 15 and 10, respectively. 3.3.2 Field Evaluation o f the Scaled-up Coarse Particle Concentrator Following laboratory characterization, the CPC was evaluated in collocation with a modified Micro-Orifice Uniform Deposit Impactor (M O U D I, MSP Corporation, Minneapolis, M N ) at two locations: University of Southern Califomiaa, 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in downtown Los Angeles (13 samples in mid August, 2000) and at the National Institute of Public Health and the Environment (RTVM, Bilthoven, The Netherlands, 6 samples in late September, 2000). The ten 90° elbow CPC inlets were arranged at 360° to ensure isokinetic sampling. In the field tests, the CPC operated at a total flow rate of 1,000 LPM and with a minor flow rate adjusted to 33 LPM. The ideal enrichment factor corresponding to this minor-to-total flow ratio is 30. O f the 33 LPM minor flow, 11 LPM were drawn into a 4.7 cm filter (2pm, PTFE, Gelman, Ann Arbor, M I) through an isokinetic sampling probe inserted in the 5 cm tube leading to the exposure chamber while the remaining 22 LPM were drawn by a separate pump. The reason for using only 11 LPM through the filter was to minimize potential coarse particle losses due to the flow contraction through the filter holder inlet. The pressure drops through the minor and major flows were about 2" H 2O and 110" H 2O, respectively. The M O U D I sampled at 30 LPM and was modified (from its original 8-stage configuration) to include only 3 stages, collecting size-segregated particles in aerodynamic diameter ranges of 0-2.5, 2.5-10 and 10-18 pm, respectively. 4.7 cm PTFE filters were used as impaction substrates in coarse PM M O U D I stages. The CPC 90° elbows and M O U D I stage 10-18 pm were coated with a thin layer o f silicone grease to reduce potential particle bounce. The sampling flow rates of the M O U D I and CPC minor flows were measured before and after the sampling with calibrated flow meters (Cole-Parmer, #EW-32458-64 and #EW- 32458-58, Cole-Parmer Instrument Company, Vernon Hills, IL 60061). Additionally 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. during sampling, the CPC minor flow filter sampler and the major flow were monitored by means of inline rotameters. Particle mass, sulfate, nitrate concentrations as well as concentrations of trace elements and metals were determined for both ambient and concentrated aerosols. The sampling periods varied from 3 to 12 hours depending on observed PM levels. To determine particle mass concentrations, the PTFE filters o f the M O U D I and minor flow were pre-weighed and post-weighed using a Microbalance (M T 5, Mettler-Toledo Inc., Highstown, NJ; Sartorius microbalance M C -5, Sartorius AG, Goettingen, Germany) in a room with controlled temperature o f 21-24 °C and relative humidity of 40-50%. Filters were weighed twice in order to increase precision. In case of a difference of more than 3 pg between consecutive weighings, the filter was weighed a third time or reweighed until two consecutive weighings differed by less than 3 pg. Thirteen out of nineteen pairs of PTFE filter samples collected by the CPC and M O U D I were then analyzed by means o f x-ray fluorescence (XRF) to determine concentrations of selected elements and metals. These samples as well as the remaining six pairs (corresponding to samples collected in the Netherlands) were subsequently extracted with 0.15 ml o f ethanol and 5 ml o f ultrapure water. Ethanol was used in order to wet the hydrophobic Teflon filter. The samples were sonicated for 15 minutes and analyzed for sulfate and nitrate ions by means of ion chromatography (IC). Samples that were lower than three times the lower limits of 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. detection (LOD) o f either XRF or IC were excluded. Using XRF prior to IC is a t procedure that is typically not recommended for analysis o f PM 2.5 because it might cause volatilization of ammonium nitrate. However, nitrate in the coarse mode in both the Netherlands (ten Brink et al., 1997) as well in Los Angeles (Solomon et al., 1988; Liu et al., 2000) is mostly associated with the non-volatile sodium nitrate (Solomon et al., 1988; Liu et al., 2000), the concentrations o f which are not expected to be altered by XRF. 3.4 Results and Discussion 3.4.1 Laboratory Characterization o f the Virtual Impactors The results of the evaluation of the virtual impactors are shown in Figure 3.3. The concentration of generated monodisperse particles was in the range o f 170-500 pg/m3 and was enriched to 580-13,000 pg/m3 , and thus several orders of magnitude higher than the lower limit of detection of the DataRAM (1-5 pg/m3 ) while below the instrument’s upper lim it (40 mg/m3 ). It should be noted here that mass concentrations obtained by the DataRAM are quite dependent upon particle size (Sioutas et al., 2000). Therefore, particle concentrations measured by the DataRAM upstream and downstream of the virtual impactor were only used for determining the concentration enrichment of monodisperse particles, which is the ratio o f the 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. downstream to the upstream concentrations and not for representing actual mass concentrations in the air stream. 30 - ■ R 25 - ui 20 - ~ 15- 0 2 8 10 4 6 Aerodynamic Particle Diameter (pm) Figure 3.3. Concentration enrichment of each individual virtual impactor at three different minor flow rates. Total Intake Flow rate/impactor; 1001/min. Figure 3.3 shows the concentration enrichment for each individual virtual impactor at three minor flow rates as a function of aerodynamic particle diameter. The plotted data correspond to the averages o f the ten virtual impactors, whereas the error bars represent the standard deviation in the enrichment values between the ten impactors. For a minor flow o f 3.3 LPM, the concentration enrichment increases sharply from about 2 to about 28 as particle aerodynamic diameter increases from 1 to 2.5 pm. The enrichment is practically the same for particles in the aerodynamic 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. diameter range of 2.5 to 9 pm. Similarly, for a minor flow of 7 LPM, the t concentration enrichment increases sharply from 2.7 to approximately 13.5 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 13 to 14 and practically independent of particle size. The same trends can also be observed for the 10 LPM minor flow configuration. The data shown in Figure 3 also indicate that the 50% cut point of the virtual impactor (defined as the aerodynamic particle size at which the enrichment factor is half o f its ideal value) is approximately 2.0 pm and does not seem to depend significantly on the minor-to-total flow ratio. The slight decrease in concentration enrichment values (still higher than 85% of the ideal value) observed at 9 pm particles is probably due to some internal losses through the collection nozzle. The overall high concentration efficiencies o f 9 pm particles, however, prove that there is no significant loss of these particles due to the 90° elbow. More importantly, these tests imply that the size distribution of enriched coarse particles in the CPC was the same as that o f the ambient air, since the concentration enrichment does not depend on particle size— at least for particles larger than 2.5 pm in aerodynamic diameter. The near-ideal concentration enrichment factors for particles larger than 2.5 pm in aerodynamic diameter clearly illustrate the impressive performance o f round nozzle virtual impactors, which have been proven to be superior to those having 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. acceleration and collection nozzles of rectangular geometry. Experimental flow visualization studies by Masuda et al. (1988) and Gotoh and Masuda (2000) as well as modeling studies by Marple and Chien (1980) showed that end effects associated 1600 1400 | 1200 = L 1000 c o 2 4 -* c 800 © C 600 o o o 400 CL O 200 y = 23.05x + 4.7 R2 = 0.92 0 10 20 30 40 50 60 70 MOUDI Concentration (pg/m3 ) Figure 3.4. Comparisons between the coarse particle concentrator and M O U D I coarse PM mass concentrations. with rectangular geometry virtual impactors result in excessive particle losses and decrease in the sharpness of the particle collection efficiency curve. The previously mentioned studies also indicated that particle losses become particularly high as the minor-to-total flow ratio decreases. Our laboratory experiments indicate that even at a minor- to -total flow ratio o f 3.3 %, the collection efficiency of the virtual impactors o f the CPC are high and internal losses are very low. 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.4.2 Field Evaluation Tests , Results o f concentration comparisons based on mass, sulfate, nitrate, and selected trace elements and metals between Coarse Particle Concentrator and M O U D I are shown in Figures 3.4 to 3.8 and summarized in Tables 3.1 and 3.2. Table 3.1 Comparisons between Coarse Particle Concentrator (CPC) and M O U D I based on mass, sulfate and nitrate concentrations (in pg/m ). Range of Ambient Concentration Range of Minor Flow Concentration Mean E.F.a Coefficient of Determination (R2 ) Mass 3 .4 -5 9 .7 131.9-1,181.2 25.9 ±4 .6 ■ 0.92 Sulfate 0.32 - 2.04 9 .6 0 -6 5 .2 4 30.0 ± 2.9 0.90 Nitrate 1 .1 -1 5 .4 2 9 .4 -32 5 .8 25.9 ± 4.2 0.81 a Ideal Enrichment Factor (EF) determined by the flow ratio o f total flow rate to minor flow rate, 1000 LPM to 33 LPM in this case. Figure 3.4 shows that ambient coarse particle mass concentrations ranged from 3 to 60 pg/m3 and were enriched from roughly 132 to 1,200 pg/m3 . The average concentration enrichment factor obtained for mass was 25.9 with standard deviation of 4.6 (Table 3.1). CPC and M O U D I mass concentrations were also very highly correlated, with the coefficient o f determination (R2 ) based on linear 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. regression being equal to 0.92. Given that the ideal enrichment factor for this minor- to-totql flow rate configuration would be 30, these enrichment factor values indicate that the CPC operates with extraordinary collection efficiency (0.86± 0.15) and very few internal particle losses. Even though lower coarse particle mass concentrations (3.5-10 pg/m3 ) were generally observed in the Netherlands during the field evaluation, there seems to be no significant difference between the actual enrichment factors obtained in Los Angeles and Bilthoven. This overall agreement is important, considering that these two locations have substantially different meteorological conditions (weather, temperature, and relative humidity) as well as aerosol sources; hence, they are expected to have coarse PM with different chemical composition. The results of comparing coarse particulate nitrate and sulfate concentrations collected by the CPC and M O U D I are shown in Figures 3.5-3.6 and Table 3.2. The average amounts of sulfate and nitrate in the coarse mode were 4.6% and 19.6% by mass, respectively. The average enrichment factor obtained for nitrate was 25.9 (± 4.2) with R2 of 0.81. Both enrichment and correlation coefficient values are very close to those based on mass concentrations. A slightly, but not statistically significant (p=0.29), higher enrichment factor of 30.0 (± 2.9) was observed when comparing sulfate concentrations between the CPC and M O U D I. 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 300 250 to E 1 200 c o 3 5 2 y = 25.252X + 6.8 R2 = 0.81 150 c « o o 100 o Q . O 50 0 0 2 4 6 8 10 12 MOUDI ambient concentration (ng/m3 ) Figure 3.5. Field Comparisons of coarse PM nitrate concentrations between CPC and M O UDI. 70 60 - f 50 • 30 ■ 10 ■ 0.0 0.5 2.0 2.5 1.0 1.5 MOUDI ambient concentration (|xg/m3 ) Figure 3.6. Field comparison of coarse PM sulfate concentrations between CPC and M OUDI. 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The CPC and M O U D I comparison based on concentrations o f trace elements and metals is shown in Figures 3.7 a-j and summarized in Table 3.2. For each element a total of thirteen reliable CPC-MOUDI paired data were obtained. The following metals and elements were selected based on their relative high amounts in the coarse mode relative to the fine mode PM: A l, Si, Ca, K, S, Fe, Cu, Mn, Zn and Ti. Results from our field tests indicated that the concentrations o f A l, Si and Ca, which originate from crustal material, contributed on average by 2.0%, 5.4% and 2.4% to the overall coarse PM mass, respectively (Table 3.2). Table 3.2 also shows a) 40 0.0 0.5 1.0 1.5 2.0 MOUDI Ambient Concentration (ng/m ) b) MOUDI Ambient Concentration (jig/m3 ) - *2 O ^ 2 0 y ■ 27.057* ♦ 2.11 R* - 0J0 I 0.0 0.5 1.0 1.5 2.0 MOUDI Ambient Concentration (ng/m3) y - 27.06*+ 1.1 R* - 0.74 I 0.0 0 1 0.2 0.3 0.4 0.5 0.6 0.7 MOUDI Ambient Concentration (iig/m ) c) d) Figure 3.7. Comparisons of coarse PM concentrations between CPC and M O U D I for selected trace elements and metals: (a) Al, (b) Si, (c) Ca, (d) K , (e) S, (f) Fe, (g) Cu, (h) Mn, (i) Zn, (j) Ti. 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 25 8 IS c o > o O 1 0 5 0 0.4 0.0 0.2 0.6 0.6 1.0 1.2 1.4 O w ’ 20 1.6 2.1 1.1 0.1 0.6 MOUDI Ambient Concentration (ng/m ) MOUDI Ambient Concentration (jig/m3 ) 1.0 0.6 55 0-8 2 0.7 C 0 .6 o 2 O < 0.5 C 09 o a 0 4 O 0.3 0.2 0.1 0.0 0.015 0.02 .025 0 0.005 0.01 O .i 0.03 MOUDI Ambient Concentration (fig/m ) g) 4.5 4.0 3.5 C « 1.5 0.5 0.0 0.00 0.05 0.10 0.15 0.20 MOUDI Ambient Concentration (iig/m ) i) Figure 3.7. Continued. h) n F 0.6 o > 0.8 3 * 0.7 c o 0.6 2 0.5 c 04 4) O c o 0? o O 0.1 Q . o MOUDI Ambient Concentration fag/m3 ) 1.6 C o 2 c 0 ) E o -5 c w o O w o £L a 0.2 0.00 0.01 0.02 0.03 0.04 0.05 0.06 MOUDI Ambient Concentration (jig/m3 ) j) that the average concentration enrichment factors of Al, Si, and Ca were 30.5 (± 3.6), 28.8 (± 4.4), and 28.9 (± 5.2), respectively. Based on linear regression between the CPC and ambient metal concentrations, the coefficients o f correlation (R2 ) were 0.83, 0.78, and 0.80, respectively, as displayed in Figures 3.7 a-c. The coarse mode 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. particles also contained K, S, and Fe with mass percentages of 1.0%, 0.8% and 2.8%, respectively. The enrichment factors for K, S, and Fe ranged from 24 to 32 (Table 3.2) with standard deviations between 3.3 and 5.6, and thus similar to the results obtained for A l, Si, and Ca. Figures 3.7d-f show CPC and M O U D I concentrations for K, S and Fe were also highly correlated, with R2 values of 0.74, 0.83, and 0.80, respectively. Slightly higher enrichment factors (30-33) were obtained for trace elements (Cu, Mn, Zn, and Ti), which may be in part due to some uncertainty generated by the very small (but detectable) amounts collected in the M O U D I samples. The total contributions o f Cu, Mn, Zn, and T i were less than 0.5% of the coarse mode mass. Nevertheless, the comparison of concentrations between the CPC and M O U D I o f Cu, Mn, Zn, and Ti (shown in Figures 3.7 g-j) indicate consistency with other species in terms of concentration enrichment. Particle separation and concentration by means of virtual impaction is a technique based on particle inertia; therefore, the enrichment factor should be solely dependent on particle aerodynamic diameter and not chemical constituents. The CPC concentrations for all ten metals/elements are plotted against those measured by the M O U D I in Figure 3.8. Integration of all data in one graph was done to reduce potential analysis errors due to the XRF limits of detection. The linear regression o f this integrated comparison indicated an overall average enrichment factor o f 27 ± 7.6 for the CPC. Figure 3.8 also indicates that a very high correlation was obtained between the ambient and concentrated coarse 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. aerosols (R2=0.93) and that this correlation is independent o f the amount of the chemical constituent in the aerosol. Table 3.2 Ambient concentrations (pg/m3 ) and enrichment factor (E.F.)a for selected metal/elements based on 13 sets of comparisons between CPC and M O U D I. Al Si C a K S Fe Cu Mn Zn Ti Min. Cone. 0.24 0.82 0.41 0.17 0.26 0.4 0.00 0.00 0.00 0.03 (pg/m3 ) 5 7 7 Max. Cone. 1.48 4.4 1.69 0.64 1.19 1.90 0.03 0.02 0.06 0.15 (Ug/m3 ) Average Mass 2.0 5.4 2.4 1.0 0.8 2.8 0.05 0.04 0.06 0.21 Fraction (% ) Mean E.F. 30.5 28.8 28.9 28.5 24.8 32.2 33.4 30.0 31.4 32.4 E.F. Standard Deviation 3.6 4.4 5.2 5.6 4.3 3.3 2.8 3.5 2.7 3.1 Slope of Regression Line 22.9 26.1 27.1 27.1 21.1 30.1 29.2 29.2 24.5 23.8 a Ideal Enrichment Factor (EF) determined by the flow ratio of total flow rate to minor flow rate, 1000 LPM to 33 LPM in this case. 3.5 Conclusions A compact, portable coarse particle concentrator (CPC) with a flow capacity of 1,000 LPM was assembled by connecting ten virtual impactors in parallel. Characterization o f the CPC in the laboratory and field-testing in two locations confirmed that the CPC enriches coarse particles by a factor o f approximately 26 (± 5) when operating with a minor flow of 33 LPM. This enrichment factor is based on 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ♦ Al 100 y = 27.03 x xK R2 = 0.93 □ S 10 - :Xti • Fe Mn oTi ♦ Cu 0.1 ■ Si 0.001 0.01 MOUDI ambient concentration (pg/m3 ) 0.1 Figure 3.8. Field Comparison o f multi metal/elemental concentrations between CPC and M OUDI. the gravimetric analysis of mass, ion chromatography analysis o f sulfate and nitrate, and x-ray fluorescence analysis of trace elements and metals. Ambient coarse PM levels as low as a few micrograms per cubic meter can be concentrated for exposure to humans and animals as well as for rapid collection of samples for chemical analyses. The system itself, while compact and easily transportable, is variable in design and can be configured to fit many existing systems (sampling, exposure, etc.). The flow rates, also highly adjustable, determine the enrichment of the aerosol and thus, give the researcher latitude in exposure 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. studies. Therefore, the system can be adapted for human exposures by simply increasing the minor flow rate (or increasing the minor-to-total ratio) until a desired concentration is attained. The compact size o f the CPC (bulk dimensions 80cmx75cmx45cm) makes it readily transportable to desired locations for exposures to coarse particles of varying chemical composition. Acknowledgements This work has been supported in part by the Ministry of Housing, Planning and the Environment of the Netherlands through contract number 53-4507-4860 to the University of Southern California School of Engineering. Additional support was provided by the Southern California Particle Center and Supersite (SCPCS), funded by the U.S. EPA under the STAR program through Grant # 53-4507-0482 to USC and by the California Air Resources Board under grant # 53-4507-8360 to USC. The research described in this article has not been subjected to the Agency’s required peer and policy review and therefore does not necessarily reflect the views of the Agency and no official endorsement should be inferred. 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References Amdur, M.O. and Chen, L. (1989). Furnace-generated acid aerosols: speciation and pulmonary effects. Environ. Health Perspect., 79: 147-150. Anderson, K.R.; Avol, E.L.; Edwards, S.A.; Shamoo, D.A.; Peng, R.C.; Linn, W.S.; Hackney J.D. (1992). Controlled exposures of volunteers to respirable carbon and sulfuric acid aerosols. J. Air Waste Manage. Assoc., 42: 437-442. Arts, J.H.E.; Spoor, S.M.; Muijser, H.; Kleinman, M .T.; van Bree, L.; Cassee, F.R. (2000) Short-term inhalation exposure of healthy and compromised rats and mice to fine and ultrafine carbon particles. Inhal. Toxicol., 12: 261-266. Becker, S.; Soukup, J.M.; Gilmour, M .I.; Devlin, R.B. (1996). Stimulation of human and rat alveolar macrophages by urban air particulates: effects on oxidant radical generation and cytokine production. Toxicol. Appl. Pharmacol., 141: 637-648. Bolarin, D.M .; Bhalla, D.K.; Kleinman, M .T. (1997) Effects of repeated exposures of geriatric rats to ozone and particle containing atmospheres: an analysis of bronchoalveolar lavage and plasma proteins. Inhalation. Toxicology, 9: 423- 434. Clarke, R.W.; Catalano, P.; Gazula, G.; Sioutas, C.; Ferguson, S.T.; Koutrakis, P.; Godleski, J.J. (1999). Inhalation of concentrated ambient particles (CAPS) induced pulmonary alterations in normal and chronic bronchitic rats. Inhal. Toxicol., 11: 101-120 Dockery, D.W .; Speizer, F.E.; Stram, D.O.; Ware, J.H.; Spengler, J.D.; Ferris, B.J. (1989). Effects of inhalable particles on respiratory health o f children. Am. Rev. Respir. Dis., 139: 587-594. Gavett, S.H.; Hoyle, G.W.; Madison, S.L.; Walsh, L.C.; Hilliard, H.G.; Lappi, E.R.; Evansky, P.E.; 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 H ill ambient air. Am. J. Respir. Crit. Care Med., 159: A318. 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Godleski, J.; Sioutas, C.; Katler, M .; Koutrakis, P. (1996). Death from inhalation of concentrated ambient air particles in animal models o f pulmonary disease. Proceedings of the 2n d Colloquium on Particulate A ir Pollution and Health. Park City, 4/136-4/143. Gordon T.; Gerber, H.; Fang, C.P.; Chen, L.C. (1999). A centrifugal particle concentrator for use in inhalation toxicology. Inhal. Toxicol., 11: 101-117. Gotoh, K. and Masuda, H. (2000). Improvement o f the classification performance of a rectangular jet virtual impactor. Aerosol Sci. and Technol., 32: 221-232. Hinds, W .C. (1982). Aerosol Technology, John W iley and Sons, New York. Homberg C.; Maciuleviciute L.; Seemayer N.H.; Kainka E. (1998a). Induction of sister chromatid exchanges (SCE) in human tracheal epithelial cells by the fractions PM-10 and PM-2.5 of airborne particulates. Toxicology letters, 96,97: 215-220. Homberg C.; Seemayer N.H.; Kainka E. (1998b). Strong genotoxicity on human tracheobronchial epithelial cells (beas 2B) in vitro by the coarse (PM-10) and fine fraction (PM-2.5) of airborne particulates as an indicator o f potential adverse health effects. J. Aerosol Sci., 29 Suppl. 1: S317-S318. Kim, S.; Chang, M .C.; Sioutas, C. (2000). A new generation of portable coarse, fine and ultrafine particle concentrators for use in inhalation toxicology. Inhal. Toxicol., 12 (supplement 1): 121-137. Kleinman, M .T.; Bhalla, D.K.; Mautz, W.J; Phalen, R.F. (1995). Cellular and immunologic injury with PM-10 inhalation. Inhal. Toxicol., 7: 589-602. Kleinman, M .T.; Sioutas, C.; Chang, M .C.; Cassee, F.R. (2003). Ambient fine and coarse particle suppression of alveolar macrophage functions. Tox. Lett., 137(3): 151-158. Koenig, J.; Larson, T.V.; Hanley, Q.S.; Rebolledo, V.; Dumler, K.; Checkoway, H.; Wang, S.Z.; Lin, D.; Pierson, W.E. (1993). Pulmonaiy function changes in children associated with fine particulate matter. Environ. Res., 63: 26-38. Liu D .Y.; Prather, K.A; Hering S.V. (2000). Variations in the size and chemical composition o f nitrate-containing particles in PJverside, CA. Aerosol Sci. and Technol., 33: 71-86. 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Marple V.A. and Liu B.Y.H. (1974). Characteristics o f 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. Masuda H. and Nagasita, S. (1988). Classification performance of a rectangular jet virtual impactor - Effect o f nozzle width ratio of collection nozzle to acceleration jet. J. Aerosol Sci., 19: 243-252 Monn, C. and Becker, S. (1998) Fine and coarse particles: induction o f cytokines in human monocytes. J. Aerosol. Sci., 29: 305-306. Ostro, B.D. (1993). The association of air pollution and mortality: examining the case for interference of organonitrates in the Los Angeles aerosol. Atmos. Environ., 25A: 2855-2861 Ostro, B.D.; Hurley, S.; Lipsett, M J . (1999). A ir pollution and daily mortality in the Coachella Valley, California; a study of PM10 dominated by coarse particles. Environ. Res., 81: 231-238 Pope, C.A., HI; Dockery, D.W.; Spengler, J.D.; Raizenne, M.E. (1991). Respiratory health and PMio pollution. A daily time series analysis. Am. Rev. Respir. Dis., 144: 668-674. Sioutas, C.; Koutrakis, P.; Burton, R.M. (1995a). A technique to expose animals to concentrated fine ambient aerosols, Environ. Health Perspect., 103: 172-177. Sioutas, C.; Koutrakis, P.; Ferguson, S.T.; Burton, R.M. (1995b). Development and evaluation o f a prototype ambient particle concentrator for inhalation exposure studies. Inhal. Toxicol., 1: 633-644. Sioutas, C.; Koutrakis, P.; Godleski, J.; Ferguson, S.T.; Kim, C.S.; Burton, R.M. (1997). Harvard/EPA ambient fine particle concentrators for human and animal exposures. J Aerosol Sci., 28(6): 1057-1071. Sioutas, C.; Kim , S.; Chang, M . (1999b). Development and evaluation of a prototype ultrafine particle concentrator. J. Aerosol Sci., 30(8): 1001-1012. Sioutas, C.; Kim, S.; Chang, M .; Terrell, L.L.; Gong, H. (2000). Field evaluation of a modified DataRAM M IE scattering monitor for real-time PM2.5 mass concentration measurements. Atmos. Environ., 34: 4829-4838. 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Solomon, P.A.; Larson, S.M.; Fall, T.; Cass, G.R. (1988). Basin wide nitric acid and related species concentrations observed during the Claremont Nitrogen Species Comparisons Study. Atmos. Environ., 22:1587-1594. Ten Brink, H .M .; Kruisz, C.; Kos, P.A; Berner, A. (1997). Composition/size of the light-scattering aerosol in the Netherlands. Atmos. Environ., 31: 3955-3962. Urch, B.; Liu, L.; Brook, J.; Purdham, J.; Tarlo, S.; Broder, I.; Lukic, Z.; Datema, J.; Koutrakis, P.; Sioutas, C.; Ferguson, S.; Dales, R.; 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. 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 4 DEVELOPMENT AND EVALUATION OF A CONTINUOUS COARSE (PM 1 0 - PM2.5) PARTICLE MONITOR Journal of the Air & Waste Management Association 51 (2001) 1309- 1317 4.1 Abstract In this paper, we describe the development and laboratory and field evaluation o f a continuous coarse (2.5 - 10 pm) particle mass (PM ) monitor that can provide reliable measurements o f the coarse mass (CM ) concentrations in time intervals as short as 5- 10 minute. The operating principle of the monitor is based on enriching CM concentrations by a factor of about 25 by means o f a 2.5 pm cutpoint round nozzle virtual impactor, while maintaining fine mass, i.e., mass o f PM 2.5 (FM ) at ambient concentrations. The aerosol mixture is subsequently drawn through a standard TEOM™, the response of which is dominated by the contributions of the CM , due to concentration enrichment. Findings from the field study ascertain that a TEO M ™ coupled with a P M 1 0 inlet followed by a 2.5 pm cutpoint round nozzle virtual impactor can be used successfully for continuous CM concentration measurements. The average concentration-enriched CM concentrations measured by the TEOM™ were approximately 26-27 times higher than those measured by the time-integrated PM 1 0 samplers (M OUDI™ and Partisol™ sampler), and highly correlated. CM concentrations measured by the concentration-enriched TEOM™ were independent Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of the ambient FM -to-CM concentration ratio, due to the decrease in ambient coarse particle mass median diameter (M M D ) with an increasing FM -to-CM concentration ratio. Finally, our results illustrate one o f the main problems associated with the use of real impactors to sample particles at relative humidity (RH) values lower than 40%. While PMio concentrations obtained by means o f the M O UDI™ and Partisol were in excellent agreement, CM concentrations measured by the M OUDI™ were low by 20%, while FM concentrations were high by a factor o f 5, together suggesting particle bounce at low RH. 4.2 Implications Several researchers have raised the issue o f the quality of C M concentrations data used in PM exposure assessment and epidemiological studies' Poor CM precision could lead to potential biases in exposure-health effect models that include both FM and CM exposure variables, and make it more difficult to properly assess the spatial correlations o f CM over metropolitan areas. Since these issues may be important in evaluating the health effects o f CM relative to PMio or PM2.5, it is desirable to have CM measurements that are sufficiently precise to resolve the uncertainty surrounding existing PM studies that include CM data. This paper describes the development and performance evaluation of a CM monitor that can provide reliable measurements in time intervals as short as 5 minutes. The simplicity and reliability o f this monitor makes it ideal for use in large scale monitoring networks. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.3 Introduction i Ambient particles in the size range 2.5 to 10 pm are referred to as coarse particles or coarse mode (C M ) aerosols. Coarse particles may consist of several potentially toxic components, such as resuspended particulate matter from paved and unpaved roads, industrial materials, brake linings, tire residues, trace metals, and bioaerosols. Since a considerable fraction of these particles may deposit in i the upper airways and to a lesser extent into the lower airways, they may be responsible for the exacerbation o f asthma. Recent data from a small number of epidemiological studies indicate that, apart from -or in addition to— the fine fraction (FM ) o f particulate matter (also called PM2.5), health effects also may be closely associated with the CM fraction and sometimes even to a larger extent than FM (Ostro, 1993; Mar et al., 1999; Ostro et al., 1999). In vitro studies with human monocytes show that cellular toxicity and inflammation also may be associated with the CM and its biological components (Becker et al., 1996; Homberg et al., 1998; Monn and Becker, 1998). Several researchers have raised the issue of the quality o f CM concentrations data used in PM exposure assessment and epidemiological studies (Lipfert and Wyzga, 1995; Wilson and Suh, 1997; White, 1998). These researchers state that poor CM precision could lead to potential biases in exposure-health effect models that include both FM and CM exposure variables, and make it more difficult to properly assess the spatial correlations o f CM over metropolitan areas. Since these issues may be important in evaluating the health effects of CM relative to PMio or Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PM2.5, it is desirable to have CM measurements that are sufficiently precise to resolve the uncertainty surrounding existing PM studies that include CM data. ' According to the Federal Reference Method (FRM ), current measurements of both the PMio and PM2.5 mass concentrations are based on gravimetric analysis of particles collected on filters over a period of 24 horns (Federal Register, 1997). Gravimetric analysis was selected because most of the particle data used for the epidemiological studies investigating associations between mortality and morbidity outcomes and ambient particle exposures are based on PM concentrations (Dockery et al., 1989; Pope et al., 1995). Typically, a time-integrated sample (e.g., over 24 hours) is collected on the filter, which is later equilibrated at designated temperature and RH conditions, and subsequently weighed to determine the mass of the deposited PM. Dividing by the amount of air sample yields the atmospheric concentration. Since the values of atmospheric parameters influencing ambient particle concentration, hence human exposure, such as the emission strengths o f particle sources, temperature, RH, wind direction and speed and, mixing height, fluctuate in time scales that are substantially shorter than 24 hours, a 24-hour measurement may not reflect an accurate representation o f human exposure. Thus, more accurate, better quality data on the physico-chemical characteristics o f particles are needed to understand their atmospheric properties and health effects. Methods that are capable of providing continuous or near continuous measurements (i.e. 1-hour average or less) are highly desirable because they can provide accurate information on human exposure and atmospheric processes in short 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. timer intervals. Over the past decade, a significant number o f state-of-the-art methods were developed for continuous PMio and PM2.5 mass concentration measurements. These include the Tapered Element Oscillating Microbalance (TEOM™ 1400A; Rupprecht and Patashnick, Albany N Y ), a host o f nephelometers, such as the DataRAM™ (RAM -1, M IE Inc., Billerica, M A ), and the DUSTTRACT™ (Model 8520, TSI Inc., St. Paul, M N ), and the Continuous Ambient Mass Monitor (CAM M ™ , Thermo Andersen, Smyrna, GA) (Babich et al., 1999). The latter method can only provide measurements of FM. Mass concentration measurements using photometers or nephelometers are based on light scattering, and are dependent on particle size and chemical composition showed that variations in particle size and chemical composition may introduce considerable errors in predicting the response o f nephelometers such as the DataRAM (Sloane, 1984; McMurry et al., 1996; Sioutas et al., 2000). The TEOM™ measures either PMio or PM2.5 (but not directly C M ) by recording the decrease in the oscillation frequency o f a particle-collecting element due to the increase in its mass associated with the depositing particles. In its standard configuration, the TEOM™ collects particles at a flow rate of 2-4 liter per minute (lpm) on an oscillating filter heated to 50 °C. The TEOM™ filter is heated to eliminate interferences from changes in RH that can change the amount of particle-bound water associated with the collected PM (Allen et al., 1997). Determining CM concentrations by difference, as currently proposed by EPA introduces significant uncertainties in cases where FM account for a large fraction of the PMio (Wiener and 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bachmann, 2000). Moreover, since much of the semi-volatile particulate matter (which is mostly associated with FM ) is expected to be lost from the TEOM™ filter during1 and after collection at 50 °C, there is the potential for a substantially different measurement of PMio mass between the TEOM™ and FRM. This is most likely to occur in urban areas (or areas affected by urban plumes) where volatile compounds, such as ammonium nitrate and organic compounds can comprise a substantial fraction of the FM. Heating is not likely to affect the mostly non-volatile constituents of coarse particles, thus the accuracy of CM concentrations determined as the difference between PMio and PM2.5 will be compromised by the generally random loss of volatile compounds from FM. In theory, continuous measurements o f CM concentrations also could be conducted by means of optical, electrical, and time-of-flight monitors. These monitors measure size-resolved particle concentrations based on particle numbers, which could be subsequently converted to volume concentrations assuming spherical particles and an assumption about particle density; both assumptions are required td convert particle volume to mass concentrations. As in most air sampling applications, information on particle density is generally not available and assumptions about its value w ill introduce uncertainties in the resulting mass concentrations estimates. A far more important limitation o f the aforementioned particle number-based monitors results from the sharply decreasing number of ambient particles with increasing particle size. The ambient particle size distribution, by number, is dominated by ultrafine particles (i.e., smaller than 0.1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pm). As well, when converting a number to volume distribution, a 1.0 pm particle weighs as much as 103 times a 0.1 pm particle and 106 times a O.pi pm particle. Consequently, counting errors associated with this conversion, which may be substantial for large particles, due to their relatively low numbers combined with electronic noise, may lead to significant uncertainties in volume and consequently mass as a function o f particle size. This was demonstrated in a recent study by Sioutas et al., which showed that the mass concentrations obtained with the Scanning Mobility Particle Sizer/Aerodynamic Particle Sizer system (SMPS, Mode 3936, TSI Inc., St. Paul, M N; APS, Model 3320, TSI Inc., St. Paul, M N ) were higher by 70- 200% than those determined with a reference gravimetric method (1999). In this paper, we describe the development and laboratory and field evaluation of a Continuous Coarse Particle Monitor (CCPM) that can provide reliable measurements of the CM concentrations in time intervals as short as 5-10 minute. The operating principle of the monitor is based on enriching the CM concentrations by a factor o f about 25 while maintaining FM at ambient concentrations. The aerosol mixture is subsequently drawn through a standard TEOM™, the response of which is dominated by the contributions o f the CM due to concentration enrichment. This paper also presents a comparison between the CM and FM concentrations obtained different time-integrated samplers (i.e., filters and impactors), which was conducted during the field evaluation study of the CCPM. 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.4 Methods 4.4.1 Description o f the Continuous Coarse Particle Monitor The CCPM, shown schematically in Figure 4.1, operates at an intake flow of 50 lpm, and consists of three main components: a) a PMio inlet; b) a 2.5 pm cutpoint round nozzle virtual impactor (or, coarse particle concentrator), and; c) TEOM™. Particles are drawn at 50 LPM through a circular nozzle, 1.1 cm inside diameter, attached to a 90° aluminum duct elbow, 3.2 cm in diameter. The nozzle protrudes 3 cm from the rest of the inlet section o f the continuous monitor and extends up to a distance o f 1.5 cm from the inside wall of the 90° elbow, as shown in Figure 4.1. The nozzle has been designed with a cutpoint o f approximately 10 pm aerodynamic diameter (AD). During the field tests, a thin layer (approximately 1 mm) of silicon grease (Chemplex ™ 710, NFO Technologies, Kansas City, KS) was applied periodically to the inside wall of the elbow to prevent particle bounce. The collection efficiency of the PMio inlet was evaluated in field tests by measuring the mass-based concentrations of ambient particles in the 2.5 to 20 pm range by means o f an APS™. For these tests, the TEOM™ was disconnected from the virtual impactor and the minor flow was drawn directly the APS™. The sampling flow o f the APS™ is 5 lpm, thus higher than the minor flow o f the CCPM (2 LPM). Since the cutpoint o f the PMio inlet does not depend on the minor-to-total flow o f the virtual impactor but on the total aerosol flow entering the impactor-inlet, 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the major flow of the virtual impactor was adjusted to 45 LPM in order to maintain the total flow entering the PMio inlet and virtual impactor at 50 lpm. PM-10 inlet Aerosol inlet 50 LPM 2.5-micron cutpoint virtual TEOM Fine (0-2.5 pm) PM 48 LPM Coarse (2.5-10 pm) particles Pump 2 LPM Figure 4.1. Schematic of the Continuous Coarse Particle Monitor The concentration o f particles in the 2.5 to 20 pm (enriched by a factor of approximately 10) was obtained for a sampling period of 3 minutes. Subsequently, the PMio inlet was removed and the mass-based concentration of 2.5 to 20 pm particles was obtained for a period of 3 minutes. The above test sequence was repeated five times. Particle penetration through the PMio inlet was determined for each size by dividing the average concentration (based on five tests) obtained with the PMio inlet connected to the sampler to the concentration without the inlet. The 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. wind speed (a crucial parameter in for the performance evaluation o f the inlet) was recorded during these experiments and varied from 1 to 7 miles per hour (mph), which1 is a typical range for Los Angeles. Particles smaller than 10 pm in AD are drawn through the virtual impactor, which was designed to have a theoretical 50% cut point at about 2.5 pm for an intake flow rate of 50 LPM. This is single-stage, round-jet nozzle virtual impactor with an acceleration nozzle diameter of 0.37 cm and collection nozzle diameter o f 0.56 cm. The distance between the acceleration and collection nozzles is 0.7 cm. The flow field in a virtual impactor is determined by the Reynolds number, which, is defined as: U W p Re = ------- - (1) where U is the average jet velocity through the acceleration nozzle o f the impactor, W is the diameter of that nozzle, and p and p are the dynamic viscosity and density of air, respectively. The value o f Re corresponding to the operating configuration of the virtual impactor is 18,927. Coarse particles follow the minor (concentrated) flow, while particles smaller than the cutpoint of the virtual impactor follow the major flow. The minor flow in these experiments was set at 2 LPM to achieve a nominal enrichment factor of 25. Concentrated CM, including a small fraction of FM (about 4%) are drawn through the TEOM, whose flow was adjusted to 2 LPM. In its 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. most common configuration, the aerosol is heated to 50 °C before collection on the TEOM™ filter, which is attached to the oscillating element. Our experiments were performed at sample temperatures of 50 °C and 30 °C to determine whether differences in these temperatures would result in significant differences in the response o f the CCPM. While the standard configuration o f the TEOM™ is to operate it at 50 °C, due to loss o f semi-volatile species at this temperature, many TEOMs are being operated at 30 °C with a nation dryer to remove water vapor prior to the collection substrate. No nation dryer was used in our configuration. The remaining 48 LPM (major flow) through the virtual impactor is drawn through a separate, lightweight, rotary vane pump (Gast, Model 1023, Gast M fg. Corp., Benton Harbor, M I). The pressure drops across the major and minor flows o f the virtual impactor are 5.8 and 0.25 kPa, respectively. 4.4.2 Laboratory Evaluation o f 2.5 pm Cutpoint Round Nozzle Virtual Impactor The first series of experiments were conducted in the laboratory to investigate the relationship between the concentration enrichment achieved by the 2.5 pm cutpoint round nozzle virtual impactor as a function of particle size. Briefly, monodisperse aerosols in the size range o f 1 to 10 pm were generated by atomizing dilute aqueous suspensions of fluorescent polystyrene latex particles (Polysciences Inc., Warrington, PA) with a constant output nebulizer (HEART™, VORTRAN Medical Technology, Inc., Sacramento, CA). The generated particles were mixed with dry room air in a 1-liter bottle to remove the excess moisture . The dry aerosol 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. was then drawn through a tube containing ten Po-210 neutralizers that reduced particle charges prior to entering the virtual impactor. For each o f the monodisperse particles in the range of 1 to 5 pm, the DataRAM was used to first measure the mass concentration o f the generated aerosols prior to entering the 90° elbow of virtual impactor. The DataRAM was subsequently connected downstream o f the minor flow of the virtual impactor to measure the mass concentration of the aerosols after concentration enrichment. The measurements were repeated at least three times, and the average concentration enrichment was determined as a function of particle size. The contributions from background ambient concentrations before and after the enrichment were recorded and subtracted from those of the input and concentrated aerosols prior to determining the collection efficiencies at the given particle size. It should be noted that indoor air levels were on the order of 7 - 15 pg.m'3 , and substantially smaller than those of the generated aerosols (prior to concentration enrichment), which varied from 170 to about 500 pg.m'3 . Therefore the contributions of the indoor aerosol to the overall concentrations measured upstream of- and in the minor flow o f the virtual impactor were considered negligible. Concentration enrichment for 5 to 10 pm particles was determined by comparing the mass collected on a glass fiber filter (2 pm pore, Gelman Science, Ann Arbor, M I) connected to the minor flow of the virtual impactor, and the mass of a similar glass fiber filter in parallel to the test system to measure the concentration of the monodisperse aerosol. The filter sampling in parallel was connected to a pump operating at 30 LPM. At the end of each run, each glass fiber filter was placed 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 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, M A ) to determine particle concentration. Concentration enrichment for each particle size was defined as the ratio of the concentration measured in the minor flow to that o f the aerosol immediately upstream of the virtual impactor inlet. 4.4.3 Field Study Following the completion of the laboratory experiments, the performance of the CCPM was evaluated in a field study which was part o f the Los Angeles Supersite project at the Rancho Los Amigos National Rehabilitation Center in Downey, CA. Situated near the Los Angeles “Alameda corridor”, Downey has some of the highest inhalable PM ]0 concentrations in the US, very often exceeding the 24- hour National Ambient A ir Quality Standard for PMio of 150 pg.m-3. The field experiments were performed during the period of October to December 2000. Concentrated CM were provided directly to the TEOM™ from the minor flow (2 LPM) of the 2.5 pm cutpoint round nozzle virtual impactor. Measurements o f concentration-enriched CM measured by the TEOM™ were compared to direct measurements with a co-located Microorifice Uniform Deposit Impactor (MOUDr™, MSP Corp. Minneapolis, M N ) and Dichotomous Partisol-Plus™ (Model 2025 Sequential A ir Sampler, Rupprecht and Patashnick Co. Inc., Albany, N Y ). The MOUDI™ sampled at 30 LPM. Instead of using all available MOUDI™ stages , only 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. those having cut-points of 10 pm and 2.5 pm were used. Thus the first' MOUDI™ stage (2.5-10 pm) was used as a reference sampler for CM concentrations and the last stage (i.e., the after-filter) was used to determine the ambient FM concentrations. Teflon filters with diameters of 4.7 and 3.7 cm (2 pm pore size, Gelman Science, Ann Arbor, M I) were used to collect CM and FM in the two MOUDI™ stages, respectively. The Partisol™ uses a PMio inlet operating at 16.7 LPM to remove particles larger than 10 pm in AD. The remaining PMio aerosol is drawn through a virtual impactor, or, “dichotomous splitter”, located after the inlet. Two separate flow controllers maintain the CM at 1.67 LPM and the FM stream at 15 LPM. CM and FM are collected on two 4.7cm Teflon filters, placed in the minor and major flows of the Partisol virtual impactor, which are housed in reusable cassettes. The Teflon filters of both MOUDI™ and Partisol™ samplers were pre- and post-weighed using a Mettler Microbalance (M T5, Mettler-Toledo, Inc, Hightstown, NJ) after 24-hour equilibration under controlled humidity (35-40%) and temperature (22-24 °C). The experiments were performed with simultaneous sampling from the TEOM™ and the MOUDI™ and/or the Partisol™. The sampling time varied from 90- minute to 210 minute depending on the ambient concentrations to allow sufficient mass to be collected on the time-integrated samplers. The majority of the experiments were for sampling periods of 120-minute. The volume concentration of ambient CM also was recorded in 15-minute intervals using an APS™ for a number 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of experiments. In addition, in selected experiments, the time-weighed mass median diameter (M M D ) of the ambient coarse particles was determined by means of the APS™. Temperature and RH data, for each experiment were also measured continuously by the Partisol™ and recorded automatically by the systems software. The mass concentration o f the CCPM was determined both by the 1- or 2-hour time integrated TEOM™ readings and by directly dividing the mass deposited on the TEOM™ filter by the total air volume sampled. In all experiments, these two concentrations differed by less than 5%. CM and FM concentrations of the MOUDI™ were determined by dividing the total PM collected on the MOUDI™ substrates by the total sampled air volume. The CM concentration of Partisol was determined after dividing by the appropriate sample flow and subtracting 10% of FM concentration from it, which corresponded to the ratio minor flow to the total flow of the Partisol™l virtual impactor. 4.5 Results and Discussion 4.5.1 Evaluation o f the PMj o Inlet Particle penetration values through the PMio are plotted as a function o f AD in Figure 4.2. The data plotted in this figure indicate that particle penetration is 90% or higher for particles in the range of 2.5 to 8 pm. Penetration decreases sharply to about 50% at 10 pm and further to less than 10% for particles larger than 12 pm in AD. The sharpness of the particle penetration curve o f an impactor can be defined in 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. P article Penetration terms of the geometric standard deviation (crg ), which is the square root of the ratio of the particle AD corresponding to 16% penetration to that corresponding to 84 % penetration (Marple and Willeke, 1976). Based on this definition, the value o f a g is approximately 1.2 (roughly the ratio of 11 pm / 8 pm) for the PMio inlet, thereby indicating reasonably sharp aerodynamic particle separation characteristics. 0.8 - 0.6 0.4 0.2 - 0 5 10 15 20 25 Aerodynamic Particle Diameter (|im) Figure 4.2. Particle Penetration Through the PMio Inlet. 4.5.2 Laboratory Evaluation o f the 2.5 pm Cutpoint Round Nozzle Virtual Impactor Figure 4.3 presents the concentration enrichment o f the 2.5 pm cutpoint round nozzle virtual impactor as a function of particle AD. The data in Figure 4.3 confirm the rise o f the enrichment factor as a function of particle AD. As seen from 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the figure, the enrichment factor increases sharply up to its ideal value of 25, as predicted based upon the intake and minor flow rates of 50 and 2 LPM, respectively. The plotted data correspond to the averages o f at least three experiments per particle size, whereas the error bars represent the standard deviation in the enrichment values. The concentration enrichment factor increases sharply from about 2 to 23 as particle AD increases from 2 to 3 pm. The enrichment factor is practically the same 25 - £ 2 0 o 1 2 3 4 5 6 7 8 9 10 Particle Aerodynamic Diameter (urn) Figure 4.3. Concentration enrichment as a function of particle aerodynamic diameter, total flow rate = 50 LPM, minor flow rate = 2 LPM. for particles in the AD range of 3 to 9 pm. The data shown in Figure 4.3 also indicate that the 50% cut point of the virtual impactor, defined as the aerodynamic particle size at which the enrichment factor is half of its ideal value (i.e. about 12.5) is approximately 2.4 pm. [The enrichment factor measured at 2.5 pm is about 15]. 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The overall high concentration efficiencies of 9 pm particles, proves that there are no significant losses of these particles in the 90° elbow o f the PM^o inlet. More importantly, these tests imply that the size distribution o f concentrated CM before entering the TEOM is the same as that of the ambient air, since the concentration enrichment factor does not depend on particle size— at least for particles larger than 2.5 pm in AD. 4.5.3 Field Evaluation o f the Continuous Coarse Particle Monitor The results of the field evaluation of the CCPM are shown in Figures 4.4 to 4.7 for experiments performed at a TEOM™ temperature o f 50 °C. Figure 4.4 shows the comparison between the TEOM™ and MOUDI™ CM concentrations at 50 °C. As indicated, the data are highly correlated (R2 =0.88) with a slope of 25 and a near zero intercept. The ratio of concentrations equal to 26.1 (± 3.6) also is close to the expected value.. Figure 4.5 shows the comparison between the TEOM™ and Partisol™ CM concentrations at 50 °C. Again, these datas are highly correlated (R2 =0.88) with a slope of 24 and a near zero intercept. The ratio o f concentrations equal to 25.8 (± 4.1) also is close to the expected value. It is worthwhile noting, that the TEOM™ concentrations are not corrected for the contributions of the FM , which is present in the inlet stream. The purpose of concentrating the CM by a factor o f 25 is to eliminate the need for knowing a priori the FM concentration. Ideally, the mass concentrations measured by the CCPM are related to the actual ambient CM concentrations as follows: 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CCPM = 25 CM + FM (2) Thus a 1:1 FM-to-CM concentration ratio would result in the CCPM being 26 times higher than the actual CM concentration. An important implication of equation (2) is that unusually high (but not impossible) FM-to-CM concentration ratios (i.e., 4-6) would lead to a positive bias 1400 y= 25.1x + 24.7 R2 = 0.88 ^ 1200 • 1000 - 800 - 600 - O o ° 400 - 200 - 0 5 10 15 20 25 30 35 40 45 MOUDI Concentration (ug/m3 ) Figure 4.4. TEOM versus M O U D I coarse PM concentrations; TEOM at 50°C. (or overestimation) of the CM concentration by the CCPM, if the concentrations are not corrected to account for the contribution o f FM. To investigate the effect o f the FM-to-CM concentration ratio on the response of the CCPM, the ratio o f the concentration-enriched TEOM™ -to-MOUDPM and TEOM™-to-FartisolT h ' concentrations were plotted as a function of the FM-to-CM concentration ratio. The 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. results, shown in Figure 4.6, clearly indicate that the ratio of TEOM™ -to-MOUDFM CM concentration and the ratio o f TEOM™-to-Partisol™ CM concentration are, i under the conditions o f this experiment independent of the ratio of ambient FM-to- CM concentrations. (R2 = 0.0064). This independence can be further explained by the data plotted in Figure 4.7, which shows the decrease in the ambient M M D (determined by the APS) as the FM-to-CM concentration ratio increases. There is a marked shift in M M D from 4.8 - 5 pm to 2.8 - 3 pm as the ratio o f FM -to-CM concentration increases from 1 to 5, respectively. 1400 1200 y = 24.0x + 27.8 R2 = 0.88 1000 800 600 400 200 0 0 10 20 30 40 50 Partisol Concentration (jig /m 3) Figure 4.5. TEOM versus Partisol coarse PM concentrations; TEOM at 50°C. 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40 35 30 25 - 20 5 15 + 3 o s s o UJ r — 10 5 O . □ □ d □ ; D ° m □ % r r XT □ □ □ • TEOM-Partisol □ TEOM-MOUDI 0.5 1.5 2 2.5 3 3.5 Fine to Coarse PM ratio 4.5 Figure 4.6. TEO M -M O UD I and TEOM-Partisol concentration ratio as a function of fine/coarse PM ratio; TEOM at 50°C. 5 4 o S 5 i a 3 a e a o S * LL 1 0 0 1 2 8 3 4 5 6 7 Coarse PM Mass Median Diameter (i*m) Figure 4.7. Relationship between coarse particle mass median diameter (M M D ) and fine-to-coarse PM ratio. 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The highest values of FM-to-CM concentrations, ranging from about 3.5 to 4.6, were obtained on October 20 and 21, 2000. During these two days, stagnation conditions occurred in Downey, with the average wind speed during the sampling periods being less than 1 miles per hour (mph). 2-hour averaged FM concentrations measured by either the MOUDI™ or Partisol™ during these two days ranged from 80 to 146 pg.m' . These conditions are expected to result in high FM concentrations in locations such as Downey, which is primarily impacted by vehicular emissions from nearby freeways, while the relatively low CM concentrations may be explained by the lack of sufficient wind velocity to either generate or transport coarse particles. As the virtual impactor-particle concentrator preceding the TEOM™ has a 50% cutpoint at about 2.5 pm, particles in the 2.5 - 3 pm AD range would be concentrated somewhat less efficiently than those larger than 3 pm. For example, the laboratory evaluation of the 2.5 pm cutpoint virtual impactor (Figure 4.3) indicated that 2.5 to 3 pm particles are concentrated by a factor ranging from 16 to 22, compared to particles in the 3 - 10 pm range that are concentrated by a factor of 25. This slightly uneven concentration enrichment, combined with the intrinsic relationship between the coarse particle M M D and the FM -to-CM concentrations ratio, brings the CCPM-to-CM concentration ratio closer to the range of 25-26, and thus, compensates for the increase in the FM-to-CM concentration ratio. As a result, the CCPM can be used efficiently for measuring the ambient CM concentrations even in cases where the ratio o f FM -to-CM concentration is unusually high. 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The results o f the field experiments conducted at a TEOM™ temperature at 30 °C are presented in Figures 4.8 to 4.11. Similar to the 50 °C configuration, highly correlated data (R2 =0.85) are obtained for the comparison of the TEOM™ and Partisol™ CM concentrations as shown in Figure 4.8. The ratio of concentrations is 27.4 (± 3.7), which is slightly higher, but not statistically different (p=0.69) than that at 50 °C. No comparisons between the CCPM and the MOUDI™ concentrations were conducted for the 30 °C TEOM configuration, although M O U D FM data were collected concurrently to the continuous monitor and the Partisol™. This is because the ambient RH was unusually low (even by the standards o f the generally arid climate o f the Los Angeles Basin), often below 20 to 30 %. As a result, while the comparison between TEOM™ and Partisol™ CM concentrations is robust, the CM concentrations measured by the MOUDI™ were low, resulting in unrealistically high ratios between the TEOM™ and MOUDI™ CM concentrations. This is confirmed by plotting the CM concentration ratio o f Partisol™-to-MOUDFM versus RH, as shown in Figure 4.9. From the data plotted in Figure 4.9 there is a well-defined inverse relationship between this ratio and the RH. This ratio achieves an ideal value o f 1 as the RH reaches 45-50 %. For lower RH, this ratio increases sharply and becomes as high as 5 when the RH reaches the 10 to 15% range. To confirm that this phenomenon is related to particle bounce, which would be more pronounced at lower RH, the ratio of FM concentration of Partisol™-to-MOUDFM vs RH was plotted, as shown in Figure 4.10. The reverse trend is observed, with the ratio o f the FM 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1000 900 - y = 27.7x+ 10.5 R2 = 0.85 800 - 700 - ♦ ♦ 500 - H- 200 - ♦ ♦ 100 ■ 0 5 10 20 25 30 35 15 Partisol Coarse PM Concentration (iig/m3 ) Figure 4.8. TEO M versus Partisol coarse PM concentrations; TEO M at 30°C. concentration o f the Partisol™-to-MOUDFM increasing from 0.2 to about 1, as the RH increases from 10 to 50 %. Further, the total PMio Partisol™-to-MOUDI™ ratio was 0.99 (± 0.13) based on 30 field experiments, thereby suggesting that since both samplers agreed well for PMio, the only difference is in the FM and CM concentrations measurements, that is, CM concentration is low and FM concentration is high at low RH, suggesting particle bounce. These field observations illustrate one of the main drawbacks of impactors, and raises serious implications on the appropriateness o f using impactors with uncoated substrates to obtain the size distributions o f aerosols under low (< 30%) RH conditions. 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 ' 5 4 3 2 AA • I 1 n C L 0 0 10 20 30 40 50 60 70 80 Ambient Relative Humidity Figure 4.9. Plot o f the Partisol-to-MOUDI coarse particle concentrations as a function o f ambient relative humidity. Experiments at a TEOM™ temperature setting o f 30 °C also showed independence of the ratio of the TEOM™-to-Partisol™ CM concentrations to the ambient FM -to-CM concentration ratio (Figure 4.11). Data plotted in Figures 4.6 and 4.11 indicate that the mass concentration ratio of the concentration-enriched TEOM™ to either the M O U D I or Partisol™ is independent of the FM -to-CM concentration ratio over a range o f values extending from about 0.2 to 5, thereby covering a broad spectrum of ambient sampling conditions, and thus, strengthening the applicability o f the CCPM to other locations and times of the year. During these experiments, ambient PM data for a few selected runs were recorded using an APS™. Figure 4.12 shows the time series in CM concentrations 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. measured by the TEOM™ and the APS™ during one day of the field experiments. A particle density of 1.6 g/m3 was assumed in the APS™ data. The TEOM™ CM concentrations were converted to ambient CM concentrations by dividing by 26. 1.4 .2 12 < 0 o c . S 1 Q . < u il 0.8 Q Z ) 0 0.6 1 o "5 0.4 0 ) ■ - E m Q- 0.2 0 0 10 20 30 40 50 60 70 80 Ambient Relative Humidity Figure 4.10. Plot of the ratio o f Partisol-to-MOUDI fine PM concentrations as a function of relative humidity. Direct comparison between the actual concentrations measured by the two monitors cannot be made, since knowledge of the real (as opposed to an assumed) density of ambient coarse particles is required in order to convert the APS™ concentrations to actual mass concentrations. However, the data plotted in Figure 4.12 clearly show that very good overall agreement is observed in the time series of the CM concentrations obtained by means of the two samplers. 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35 - O 30 - s 0 25 - M 1 2 0 Q . ■ O 15- V s o U J i- 10 • 0 0.2 1.2 1.4 0.4 0.6 0.8 1 Fine to Coarse PM ratio Figure 4.11. Dependence o f TEOM-Partisol ratio on fine/coarse PM ratio; TEOM at 30°C. 30 E 25 o > 2. » C o c o c o o (0 tL < u O S o U J , v % 20 15 - 10 ■ ‘O . ,Q * r />< C l Jo V d.' ♦' d \ o -♦--APS o TEOM 50 100 150 200 250 Time (min) 300 350 400 Figure 4.12. Time-series o f TEOM and APS coarse particle concentrations. 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.6 Summary and Conclusions i This paper describes the development and laboratory and field evaluation of a CCPM that is based on enriching the CM concentrations by a factor of 25, while maintaining FM concentration at ambient concentrations. The aerosol mixture is subsequently drawn through a standard TEOM™, the response o f which is dominated by the contributions o f the CM due to enrichment of the coarse particles. The laboratory evaluation of the 2.5 pm cutpoint round nozzle virtual impactor confirms the rise in the enrichment factor as a function o f particle AD. The concentration enrichment factor increases sharply from about 2 to about 25 as particle AD increases from 2 to 3 pm. The enrichment is the same, with in the error o f the measurement, for particles in the AD range of 3 to 9 pm. Findings from the field study ascertain that the TEOM™ coupled with a 2.5 pm virtual impactor can be used successfully for continuous CM concentration measurements. The results indicate excellent correlation between the concentration- enriched TEOM™ and time integrated samplers (MOUDI™ and Partisol™), with the average TEOM™ CM concentration being approximately 26-27 times higher than those measured by the time-integrated samplers. No substantial differences in the response of the concentration-enriched TEOM™ are observed between TEOM™ operating temperatures of 30 and 50 °C. Results from the field experiments also show that the CM concentrations measured by the concentration-enriched TEOM™ are independent of the ambient FM -to-CM concentration ratio. This is due to the 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. decrease in ambient coarse particle M M D with increasing FM -to-CM concentration ratio, as might be expected, since FM concentrations tend to increase and coarse particle loadings tend to decrease during stagnation conditions. This also strengthens the applicability of the CCPM in cases where the FM -to-CM concentration ratio is very high. Finally, our results illustrate one of the main problems associated with the use of impactors to sample particles under conditions o f RH values lower than 40%. While PMio concentrations obtained by means of the MOUDI™ and Partisol™ are in excellent agreement, CM concentrations measured by the MOUDI™ are as low as 20% compared to those measured by the Partisol™, while MOUDI™ FM concentrations were high by as much as a factor o f 5, together suggesting particle bounce at low RH. Acknowledgements This work was supported by the Southern California Particle Center and Supersite (SCPCS), funded by the U.S. EPA under the STAR program through Grants # 53- 4507-0482 and 53-4507-7721 to USC. The U.S. Environmental Protection Agency through its Office o f Research and Development collaborated in this research and preparation o f this manuscript. The manuscript has been subjected to Agency review and approved for publication. Mention of trade names or commercial products does not constitute an endorsement or recommendation for use. Finally, a provisional patent application has been filed to the United States Patent Office by the USC Office of Technology and Licensing (USC File No. 3102). 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References , Allen, G.; Sioutas, C.; Koutrakis, P.; Reiss, R.; Lurmann, F.W .; Roberts, P.T. (1997). Evaluation o f the TEOM method for the measurement of ambient particulate mass in urban areas. J. Air Waste Manage. Assoc., 47: 682-689. Babich, P.; Wang, P.Y.; Allen, G.A.; Sioutas, C.; Koutrakis, P. (1999). Development and evaluation o f a continuous PM 2.5 ambient mass monitor. Aerosol Sci. Technol, 32: 309-325. Becker, S.; Soukup, J.M.; Gilmour, M .I.; Devlin, R.B. (1996). Stimulation o f human and rat alveolar macrophages by urban air particulates: effects on oxidant radical generation and cytokine production. Toxicol. Appl. Pharmacol., 141: 637-648. Dockery, D.W.; Speizer, F.E.; Stram, D.O.; Ware, J.H.; Spengler, J.D.; Ferris, BJ. (1989). Effects o f inhalable particles on respiratory health o f children. Am. Rev. Respir. Dis., 139: 587-594. Fed. Regist., July 18,1997; 62 (138) 40 CFR, Part 50. Homberg C.; Maciuleviciute L.; Seemayer N.H.; Kainka E. (1998). Induction of sister chromatid exchanges (SCE) in human tracheal epithelial cells by the fractions PMio and PM2.5 o f airborne particulates. Toxicology Letters, 96,97: 215- 220. Mar T.; Norris G.; Koenig J.; Larson T. (1999). Associations between air pollution and mortality in Phoenix. Environ.. Health Perspect., 108: 347 -353. Lipfert, F.; Wyzga, R. (1995). Uncertainties in identifying “responsible” pollutants in observational epidemiology studies. J. Air & Waste Manage. Assoc., 47: 517- 523. Marple, V.A . and Willeke, K. In Fine Particles: Aerosol Generation, Measurement, Sampling, and Analysis (Edited by B.Y.H. Liu); Academic press: New York, 1976. McMurry, P.H.; Zhang X.; Lee, Q.T. (1996). Issues in aerosol measurement for optical assessments. Journal o f Geophysical Research., 101 (19): 188-197. Monn, C. and Becker, S. (1998). Fine and coarse particles: Induction of cytokines in human monocytes. J. Aerosol. Sci., 29: 305-306. 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ostro, B.D. (1993). The association of air pollution and mortality: examining the case for interference of organonitrates in the Los Angeles aerosol. Atmos. Environ., 25A: 2855-2861. Ostro', B.D.; Hurley, S.; Lipsett, M.J. (1999). A ir pollution and daily mortality in the Coachella Valley, California; a study of PMIO dominated by coarse particles. Environ. Res., 81: 231-238. Pope, C.A., HI; Bates, D.V.; Raizenne, M.E. (1995). Health effects of particulate air pollution: time for reassessment. Environ. Health Perspect., 103: 472-480. Sioutas, C.; Abt., E.; Wolfson, J.M.; Koutrakis, P. (1999). Effect o f particle size on mass concentration measurement by the Scanning Mobility Particle Sizer and the Aerodynamic Particle Sizer. J. Aerosol Sci. Technol., 30: 84-92. Sioutas, C.; Kim , S.; Chang, M .; Terrell, L.L.; Gong, H. (2000). Field evaluation of a modified DataRAM M IE scattering monitor for real-time PM 2.5 mass concentration measurements. Atmos. Environ., 34: 4829-4838. Sloane, C.S. (1984). Optical properties o f aerosols of mixed composition. Atmos. Environ., 18: 871-878. White, W .H. (1998). Statistical considerations in the interpretation of size-resolved particulate mass data. J. Air & Waste Manage. Assoc., 48: 454-458. Wiener, R.; Bachmann, J.D. Coarse Particle Monitoring. Presented at the US Environmental Protection Agency, Science Advisory Board, Clean A ir Scientific Advisory Committee, Technical Subcommittee on Fine Particle Monitoring Meeting, April 18-19, 2000. Wilson, W ., and Suh, H.H. (1997). Fine particles and coarse particles: Concentration relationships relevant to epidemiological studies. J. Air & Waste Manage. Assoc., 47: 1238-1249. 1 0 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 5 DEVELOPMENT AND EVALUATION OF A PM10IMPACTQR- INLET FOR A CONTINUOUS COARSE PARTICLE MONITOR Aerosol Science and Technology 37 (2003) 271-281 5.1 Abstract Conventional PMio inlets available operate at a flow rate o f 16.7 LPM. The purpose of this study was to develop and test a PMio inlet designed to operate at 50 LPM to be used with a recently developed Continuous Coarse Particle Monitor (Misra et al., 2001). Laboratory tests using polystyrene latex particles established the inlet’s 50% cutpoint at 9.5 pm. Further evaluation of PMio inlet was performed in a wind tunnel at wind speeds of 3, 8 and 24 km/h. Tests showed that the 50% efficiency cutpoint as well as the very sharp particle separation characteristics of the inlet were maintained at these wind speeds. Field evaluation of the PMio inlet was performed in Riverside and Rubidoux, CA. A 2.5 pm cutpoint round nozzle virtual impactor was attached downstream o f the developed PMio inlet. The Dichotomous PMio Partisol Sampler, operating at a flow rate of 16.7 LPM was used as a reference sampler. The Dichotomous Partisol uses an FRM PMio inlet operating at 16.7 LPM to remove particles larger than 10 pm in aerodynamic diameter. Commercially available 4.7 cm Teflon filters were used in both the Partisol and the PMio inlet to collect particulate matter (PM ). Results showed good agreement between coarse PM (2.5-10 pm) mass concentrations measured by means o f the PMio inlet and Partisol. Chemical analyses showed excellent agreement between coarse PM concentrations of A l, K, Si, Ca and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fe obtained by the tw o samplers. The agreement also persisted fo r nitrate and sulfate. F inally, the excellent agreement between coarse concentrations o f the PM io in le t and Partisol persisted fo r w ind speeds up to 19 km /h. 5.2 Introduction Since the advent o f the particulate m atter (PM ) standards b y the U nited States Environm ental Protection Agency (US EPA), particle sam pling has been a prim ary goal o f both scientists and lawmakers. The addition o f the PM2.5 (fin e particle) and the soon to be developed PM10-PM2.5 (coarse particle) standards to the PM io standard has created the need fo r reliable continuous coarse and fin e PM measurement devices. One such device, a Continuous Coarse P article M o n ito r' (C C PM ) is described by M isra et al. (2001). A n essential component o f any modem PM m onitoring device is a size pre- selective inlet. This is even m ore im portant when the size range to be removed p rio r to sam pling consists o f large particles. I f the in le t allow s even a sm all fractio n o f the undesirable PM in to the measuring device, the error could be large. The reason fo r this is that large particles have large masses, w hich w ill hea vily influence the measurement o f a mass-based m onitor. Because o f their large mass, hence inertia, coarse particles are d iffic u lt to sample and collect. W hen these heavy particles are accelerated in an im pactor je t, their substantial inertia causes them to h it the im pactor collectio n plate. M any tim es the particle w ill bounce o ff o f this plate and become re-entrained in the a ir stream. 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This causes overestimation of the mass downstream of the inlet. Another problem that occurs because of the inertia of these particles is the underestimation or overestimation of particle mass that results from anisokinetic sampling. Unlike the PM in smaller size ranges, coarse PM is not as uniformly dispersed in the atmosphere. It settles and becomes resuspended due to localized events (i.e. high wind episodes). The original inlet employed to characterize the CCPM (described in more detail by Misra et al., 2001) was subject to this error. That inlet consisted o f a simple 90-degree elbow with a jet fashioned from a pipe with a preset diameter. Particles larger than about 10 pm impacted on the throat o f that elbow. Among its problems were that it sampled from only one direction (thus was prone to substantial anisokinetic sampling errors if the wind direction was not in alignment with the inlet) and also needed to be re-greased periodically in order to prevent bounce o f large particles. Additionally, the inlet’s efficiency curve was not as steep as most commercially available P M i0 inlets. Currently, several manufacturers have developed commercially available PMio inlets, some of which have received designation from the US EPA to be used as federal reference methods. These inlets operate at sampling flow rates ranging from 16.7 to 1133 LPM. A comprehensive review of PMio and other federal reference method inlets is given by Chow (1995). The SA (Thermo-Anderson, Smyrna, GA) or GM W (General Metal Works, Tisch Environmental, Village of Cleves, OH) Models 321A, 321B, 1200, and Wedding IPioPMio operate at the highest flow rate. The SA 254 Medium-Volume PMio Inlet operates at a flow of 113 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LPM, and the SA 246B operates at a design flow of 16.7 LPM (Chow, 1995). None of the devices listed above, however, is compatible with the CCPM ’s flow rate of 50 LPM. To these ends, an efficient and low-maintenance PMio inlet was designed and evaluated for a flow rate of 50 LPM. The inlet’s design flow rate accommodates the CCPM developed by Misra et al. (2001). The goals o f this newly developed inlet are to overcome anisokinetic sampling, sharpness o f cutpoint, and limited capacity of other commercially available PMio inlets. 5.3 Materials and Methods 5.3.1 Description o f the PMj o Impaction Inlet The inlet used in these tests is a modification o f a commercially available PMio inlet, from Rupprecht and Patashnick (Model P/N 57-00596, R&P Inc., Albany, N Y ), which operates at a flow rate o f 16.7 LPM. This inlet is described in detail in Federal Register, 1996. To adapt the inlet so that it could operate at 50 LPM, the nozzle was modified by widening the nozzle diameter and also shortening the nozzle length so as to increase the jet-to-plate distance. The schematic of the inlet is shown in Figure 5.1a. The modifications along with the new dimensions are given in Figure 5.1b. The nozzle design parameters were modified so that the predicted cutpoint o f the impactor is about 10 pm at the flow rate of 50 LPM. The principal parameter determining particle capture is the Stokes number of a particle having a 50% probability of impacting, St, defined as the ratio of the 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. particle stopping distance to the characteristic dimension o f the impactor (Hinds, 1982): St = — t U _ P p C c d j U W 9juW (1 ) where W is the diameter o f the impactor’s nozzle, U is the average velocity of the impactor jet, pp is the particle density, p is the dynamic viscosity of the air and Cc is the Cunningham slip correction factor, given by the following equation (Hinds, 1982): the particle diameter in pm. The acceleration nozzle diameter of the impactor was 1.7 cm and the corresponding jet velocity for a flow o f 50 LPM was 367 cm/s. The gap between the impaction jet and the collection plate was 1.1 cm . The Stokes number corresponding to 10 pm was approximately 0.135. 5.3.2 Laboratory Tests for Determination o f Cutpoint The schematic diagram of the experimental setup for testing the cutpoint of the PMio inlet is shown schematically in Figure 5.2. Monodisperse aerosols in the range of 2.6 to 12 pm were generated by atomizing dilute aqueous suspensions of Cc = 1 + — —[6.32 +2.01 exp P d p (-0.1095P d p) (2 ) where P is the pressure at the location of the particle in the flow (in cm Hg) and dp is 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. polystyrene latex particles (PSL, Bangs Laboratories Inc., Fisher, IN)* using a constant output nebulizer (HOPE, B&B Medical Technologies, Inc., Orangevale, I CA). The generated particles were drawn through a 2-liter glass container in which they were mixed with dry room air in order to remove excess moisture. The dry aerosols were then passed through a series of Po-210 neutralizes (NDR Inc., Grand Island, N Y ) to bring the particle charge distribution to Boltzmann equilibrium. Particle penetration through the impactor was measured as a function of particle size by means of a nephelometer, DataRAM (RAM -1, M IE Inc., Billerica, M A ), which was used to measure the mass concentrations of the monodisperse aerosols upstream and downstream o f the PMio inlet. The upstream and downstream measurements were repeated at least three times. The contributions from background ambient concentrations before and after the PMio inlet were recorded and subtracted from those of the input and concentrated aerosols prior to determining the collection efficiencies at the given particle size. It should be noted that indoor air levels were on the order o f 7-12 pg/m3 , and substantially smaller than those of the generated aerosols (prior to entering the PMio inlet), which varied from 95 to about 300 pg/m3. Therefore the contributions of the indoor aerosol to the overall concentrations measured upstream of- and downstream of- PMio inlet were considered negligible. 5.3.3 Wind Tunnel Tests The performance o f the PMio inlet was evaluated in the wind tunnel facility of School o f Public Health, UCLA. The wind tunnel is described in detail by Hinds 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Insect a) screen Air intake Air intake Deflector cone Impaction plate Modified Nozzle Receiver tube Exiting air PM,0 T 4.7 cm b) 1.7 cm 2.54 cm 1.15 cm Figure 5.1. a) Modified PMio inlet for CCPM. b) Dimensions o f modified acceleration jet nozzle. 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and Kuo, 1995. In its original design, the wind tunnel has a 1.6 x 1.6 m cross-section and was operated at two wind speeds (3 and 8 km/h). The blower downstream of the sampler test area was capable of driving the wind tunnel speed up to 8 km/h (Kennedy at al., 2001). The cross-section of the wind tunnel was modified later to achieve a wind speed of 24 km/h. A plywood baffle was placed about 0.5 m upstream o f the aerosol generation system to promote mixing. The vibrating aerosol orifice (VOAG) (Model 3450, TSI Inc., St. Paul, M N ) was itself mounted on a shaft, which moved both up and down and sideways to promote uniform injection. Three isokinetic samplers were placed around the PMio inlet. The PMio inlet was placed such that it was equidistant from the three isokinetic samplers. Two o f these samplers were lateral to the PMio inlet while the third one was above the PMio inlet. The positioning of the isokinetic samplers corresponded to uniformity in concentration around the inlet. The earlier work by Hinds and Kuo, 1995 describes the positioning of isokinetic samplers in detail. The sampling characteristics of the PMio inlet were determined by comparing the mass concentration obtained by the PMio inlet to that measured by isokinetic samplers. Seven different particle sizes - 5, 7, 9, 10, 12, 15 and 20 pm were selected to evaluate the performance of the PMio inlet. A vibrating orifice aerosol generator (VOAG) (Model 3450, TSI Inc., St. Paul, M N ) was used to generate monodisperse particles. It is known that when a solution containing a non-volatile solute is sprayed through an orifice, the solvent 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. eventually evaporates from the droplets and non-volatile particles o f solute are obtained. The diameters for these non-volatile solute particles are given by: Dp = C1/sDd (3) where C is volumetric concentration of the non-volatile solute in the solution and Dd is the initial droplet diameter. A vibrating orifice produces one droplet per cycle and the Da is given by: Dd = (6Q/701 /3 (4) where Q is the liquid flow rate and f is the disturbance frequency (Berglund and Liu, 1973). The droplet size primarily depends on the orifice size for a given solution feed rate and the frequency. For these experiments, a 20pm orifice was used for generating particle sizes of 5 and 7 pm while a 35 pm orifice was chosen for generating particles in size ranges of 9, 10, 12, 15, 20 pm. Typical VOAG operating parameters were 0.150 mLPM of feed rate at 65-70 kHz utilizing a 20 pm orifice. A feed rate of 0.3 mLPM at an operating frequency of 35 kHz was found to be optimum for a 35 pm orifice. Uranine tagged oleic acid was used as a non-volatile solute for generating particles with acetone as the solvent. Approximately, 2 g o f uranine dye was dissolved in 50 mL o f methanol to prepare the tracer solution and was left overnight to dissolve the dye in the solution with concomitant settling o f the undissolved uranine dye. A 20-40 % of this solution was then added to the oleic acid-acetone solution to generate particles of a desired size using Equations 3 and 4. The 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. monodispersity of the generated aerosols was confirmed by observing the generated particles under a microscope, which also corroborated the size o f the particles. Room A ir Nebulizer PSL Po-210 Neutralizers mm i mu 1 1 nmim 1 1 mr Dilution air exhaust PM in Inlet DataRAM 0523 Gast Pump 50LPM Figure 5.2. Schematic of the experimental set-up used for the laboratory characterization o f the PM j o inlet. The isokinetic samplers and the PMio inlet were positioned at the same distance from the sample injection point (same axial plane). These samplers were constructed from 2.5 cm in-line stainless steel filter holders (P/N 1209, Gelman Sciences Inc., Ann Arbor, M I) fitted with 8.5 mm ID brass probes that extended 32 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. mm from the face of the filter holder and sampled at a flow rate of 10 LPM for wind speeds of 3, 8 and 24 km/h. Millipore membrane filters (SMWP 02500, Millipore, Bedford, M A ) were placed downstream each of the isokinetic samplers to collect the generated monodisperse uranine tagged oleic acid particles. A 4.7 cm Millipore membrane filter (SMWP 04700, Millipore, Bedford, M A ) was placed after the P M i0 inlet in order to collect the oleic acid particles. Each of the experiments was characterized by particle size and wind speed and lasted for about 10-15 min, which was sufficient to obtain detectable mass on the filters. Detection of the deposited uranine tagged oleic acid particles on the Millipore filters was performed using a fluorescence detector (Model FD-500, Programmable Fluorescence Detector, G TI, Concord, M A ). Prior to their fluorescence detection, the extraction of the uranine was done using a buffer solution. The buffer solution was prepared by dissolving 12.4 g of boric acid in 1000 mL o f water (solution A ) and 19.05 g o f sodium borate in 1000 mL of water (solution B). Diluting 50 mL of solution A and 59 mL of solution B to 200 mL using distilled water yielded the buffer solution. The Millipore filters were then extracted in glass vials using the buffer solution. Most o f the extractions were done using 5-10 mL o f solution. Standard uranine dye solutions of 0.005, 0.01, 0.025, 0.05, 0.075 and 0.1 ppm were used to plot the calibration curve. For each wind speed, comparison between the mass concentrations obtained by means of the three isokinetic samplers and the PMio inlet was performed. For 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. each particle size, the averaged value o f the mass concentration for. the three isokinetic samplers was used. Finally, the particle penetration through the PMio inlet i was plotted against the particle diameter for each wind speed. 5.3.4 Field Evaluation o f the PMj o Inlet The performance of the PMio inlet was evaluated in Riverside and Rubidoux, CA in the months o f May and July, 2001. For these tests, a virtual impactor with a cutpoint of 2.5 pm at 50 LPM was attached downstream of the PMio inlet. This is the same virtual impactor used in the continuous coarse PM monitor (CCPM) described by Misra et al., 2001 to separate coarse from fine PM. [The concentrated coarse PM in the minor flow of this impactor is drawn into a Tapered Element Oscillating Microbalance (TEO M Mode 1400A, Rupprecht and Pataschnick Inc., Albany, N Y ) for near-continuous measurement of coarse mass concentration]. A thin film of silicone grease (Chemplex 710, NFO Technologies, Kansas City, KS) was applied to the impaction plate of the PMio inlet to prevent particle bounce. Coarse PM follows the minor flow, while particles smaller than the cutpoint o f the virtual impactor follow the major flow. The minor flow in these experiments was set at 2 LPM to achieve a nominal enrichment factor of 25. This minor flow also corresponds to the inlet flow of the CCPM. Concentrated coarse particles, including a small fraction of fine PM (about 4%), were drawn in the minor flow (2 LPM), which was pulled by an oil-less pump (Model DOA-V191-AA, Gast Manufacturing Inc., Benton Harbor, M I). The remaining 48 LPM (major flow) through the virtual impactor was drawn by a separate, oil less, light-weight, rotary vane pump (Model 0523-101Q-G588DX, 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Gast Manufacturing Inc., Benton Harbor, M I). Concentrated coarse particles in the minor flow stream were collected on a 4.7 cm Teflon filter (2pm pore size, Gelman Science, Ann Arbor, M I). Measurements o f concentration-enriched coarse particle mass were compared to measurements with a Dichotomous Partisol-Plus sampler (Model 2025 Sequential A ir Sampler, Rupprecht and Patashnick Co. Inc., Albany, N Y). The Dichotomous Partisol uses a Federal Reference Method (FRM ) PMio inlet operating at 16.7 LPM to remove particles larger than 10 pm in aerodynamic diameter. The remaining PMio aerosol is drawn through a virtual impactor, or, “dichotomous splitter”, located after the inlet. Two separate flow controllers maintain the coarse particle stream at 1.67 LPM and the fine particle stream at 15 LPM. Coarse and fine PM were collected on two 4.7 cm Teflon filters, placed in the minor and major flows of the Partisol virtual impactor, which are housed in reusable cassettes. The Teflon filters of PMio inlet and Partisol samplers were pre- and post weighed using a Mettler Microbalance (M T5, Mettler-Toledo, Inc, Hightstown, NJ) after 24-h equilibration under controlled humidity (35-40%) and temperature (22-24 °C). The experiments were performed with simultaneous sampling from the P M j o inlet and the Dichotomous Partisol. The sampling time was around 120 min for each experiment to ensure sufficient mass was collected on the filters. The coarse concentration of PMio inlet was determined after dividing by the appropriate sample flow and subtracting 2.5% of fine concentration from it, which corresponded to the ratio of minor flow to the total flow o f the PMio inlet virtual 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. impactor. Similarly, the coarse concentration o f Dichotomous Partisol was determined after dividing by the appropriate sample flow and subtracting 10% of fine concentration from it, which corresponded to the ratio of minor flow to the total flow of the Partisol virtual impactor. In addition to mass concentrations, comparisons were made between coarse PM concentrations of selected trace elements as well as particulate nitrate and sulfate measured by the PMio inlet and Partisol. Ten of twenty-one pairs of PTFE filter samples collected by the PMIO inlet and Partisol were analyzed by means of x- ray fluorescence (XRF) to determine concentrations of selected elements and metals. The remaining eleven pairs were extracted with 0.15 ml of ethanol and 5 ml of ultrapure water. [Ethanol was used in order to wet the hydrophobic Teflon filter]. The samples were sonicated for 15 minutes and analyzed for sulfate and nitrate ions by means o f ion chromatography (IC ). Samples that were lower than three times the lower limits of detection (LOD) of either XRF or IC were excluded. 5.4 Results and Discussion 5.4.1 Experimental Determination o f Cutpoint Particle penetration, the ratio of downstream to upstream mass concentration, through the PMio inlet is plotted as a function of aerodynamic diameter in Figure 5.3. The data shown in this figure indicate that particle penetration is 90% or higher for particles in the range o f 2.5 to 8 pm. Penetration decreases sharply to about 50% at 9.5-9.7 pm and further to less than 10% for particles larger than 11 pm in 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. aerodynamic diameter. An estimate of the “sharpness” o f the particle penetration curve of an impactor can be defined by means of a geometric standard deviation (ag ), which is the square root of the ratio o f the aerodynamic particle diameter corresponding to 16% penetration to that corresponding to 84 % penetration (Marple and Willeke, 1976). Based on this definition, the value o f a g is approximately 1.1 (roughly the square root ratio of 11 pm / 9 pm) for the PMio inlet, thereby indicating very sharp aerodynamic particle separation characteristics. 100% 90% 60% w 70% 81 c 0 ) o 60% E L L I C 5 0 % -- o ! S 2 % c 40% - 4 ) 30% 20% 1 0% -• 0 % 100 P article D iam eter dp (pm ) Figure 5.3. PMio Inlet Particle Penetration Curve 5.4.2 Wind Tunnel Evaluation o f the PMio Inlet The results o f the wind tunnel test are summarized in Figure 5.4. As evident from the figure, the particle penetration characteristics o f the PMio inlet are 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. unaffected by the wind speeds. The penetration for all the wind speeds tested, viz., 3, 8 and 24 km/h show a very close agreement. This is a particularly important result because it demonstrates that the inlet can be used throughout the various ambient conditions found in all normal environments. The wind tunnel tests show that the 50% cut is slightly shifted left at around 9-9.5 pm. The shift may be due to the limited resolution in that data o f Figure 5.4, considering that experiments were conducted for particles with aerodynamic diameters of 9 and 10 pm and data in- between were fitted by the graph. 100% ♦ 3 km/h □ 8 km/h A ; 1 24 km/h >> 75%- c a > 'o E ui C 50%- o s 2 © c © Q . 25% - 0% 1 10 100 Particle Diameter dp (pm) Figure 5.4. Plot o f Penetration vs Particle Diameter for Various Wind Speeds 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.4.3 Field Evaluation o f the Inlet The results o f the field evaluation of the PMio inlet are shown in Figures 5.5 to 5.8.' Figure 5.5 shows comparison between coarse PM mass concentrations obtained from the Partisol and the PMio inlet. The figures display a very good agreement between the two samplers. The geometric mean ratio o f PMio inlet to Partisol coarse PM concentration is around 0.94. Coarse PM concentrations determined by both samplers appear to be also well correlated with R2 =0.91. It should be mentioned that for all the mass calculations, the contributions o f the fine PM was subtracted from the coarse PM, both for Partisol and the PMio inlet. The calculations were performed using equations 5 and 6: Actual Coarse Mass (Partisol) = Coarse Mass (Partisol) - 0.1 x Fine Mass (Partisol) (5) Similarly, Actual Coarse Mass (PM10)= Coarse Mass (PMIO) - 0.025 x 50 x Fine Mass (Partisol)/16.7 (6) The Teflon filters after gravimetric analysis were analyzed for elements using XRF. Five major crustal metals were chosen to see the correlation between their coarse PM concentrations between the two samplers. Figure 5.6 and Table 5.1 show the comparisons and correlations between coarse concentrations o f A l, Si, K , Ca and Fe obtained by the two samplers. 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 70 - 9 y = 0.92x R2 = 0.87 60 - 50 - 40 -- 30 - 20 ■ 10 - 80 0 10 70 20 30 40 50 60 Partisol Coarse PM Concentration (|xg/m3) Figure 5.5. Coarse PM Concentrations Determined by the 50 LPM PMio Inlet and the R&P Partisol The average (± standard deviation) ratio of the PMio inlet to Partisol coarse concentration for a given element and metal are shown in Table 5.1, along with the correlation coefficient between these concentrations obtained for the specific metal. Figure 5.6 depicts an overall comparison o f all the five crustal metals. The results summarized in Table 5.1 as well as the data plotted in Figure 5.6 reveal excellent agreement between the metal concentrations of the two samplers. The mean PMio inlet- to- Partisol coarse PM concentrations vary from 0.98 to 1.19. The concentrations between the two samplers also appear to be highly correlated for each metal and element, with R2 varying from 0.79 to 0.92. As seen in Figure 5.6, the 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. overall PMio inlet-to-Partisol ratio based on all metal concentrations is virtually identical to 1. Table 5.1. Comparison between coarse PM concentrations o f various crustal metals measured by the PMio inlet and Partisol Element Ratio ofPMio/Partisol Coarse Concentration [ Average (± S.D.) ] Coefficient of Determination R 2 Aluminum 1.19 (±0.23) 0.81 Silicon 1.06 (±0.22) 0.79 Potassium 1.09 (±0.19) 0.86 Calcium 1.07 (±0.18) 0.92 Iron 0.98 (±0.16) 0.85 5 E o > 3; 4 C o 2 4 - > c 0 ) o c o O 2 0 y=0.98x + 0.18 R2 = 0.84 2 ♦ A l • SI AK □Ca AFe n o O 1 o 0 0 1 2 3 4 5 Partisol Coarse PM Concentration(|xg/m3 ) Figure 5.6. Overall comparison between coarse PM concentration o f five crustal metal measured by Partisol and PMio inlet 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The comparison of coarse PM nitrate and sulfate concentrations measured by Partisol and the PMio inlet are shown in Figures 5.7a and 5.7b, respectively. Fine PM contributions were subtracted using Equations 5 and 6, as in the case of the concentrations determined for metals. The results plotted in Figures 5.7a and 5.7b show also very good agreement between these two samplers, with the average PMio inlet-to-Partisol concentration ratios being 1.13 (±0 .1 5 ) and 1.08 (± 0.14) for nitrate and sulfate, respectively. Finally, the coarse PM concentration ratios of PMio inlet and Partisol were plotted against the wind speed. The average ratio o f PMio inlet to Partisol coarse PM concentration is around 0.91 (+ 0.11). The slightly smaller PMio inlet concentrations may be due to its somewhat smaller cutpoint (about 9.5 pm estimated from Figure 5.3) compared to that of the Partisol inlet. The results shown in Figure 5.8 clearly indicate that the ratio is independent o f the wind speed, thereby indicating that there is no systematic bias in the coarse particle concentrations measured by the PMio inlet when sampling is conducted at high wind speeds. This result further supports the findings of the wind tunnel tests and establishes the applicability of using the new 50 LPM PMio inlet in conditions of winds as high as 24 km/h. 1 2 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. E 1 o c o o 0 ) 2 + * z 6 -• 0 ) 2 4 •• C O o O © c o 10 6 6 0 2 4 Partisol Coarse Nitrate Cone (|ig/m ) Figure 5.7a. Plot of coarse PM nitrate concentrations between PMio Inlet and Partisol. 2.5 o > o c o o © J 2 3 C O 0 ) & C O o o © c © 0.5 S C L 0 0.5 15 2.5 1 2 Partisol Coarse Sulfate Cone (jig/m3 ) Figure 5.7b. Plot o f coarse PM sulfate concentrations between PMio Inlet and Partisol. 1 2 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.6 1.4 • 1.2 - ♦ ♦ 0.6 - 0.4 0.2 18 20 8 16 0 2 6 10 12 14 4 Wind Speed (km/h) Figure 5.8. Plot o f Ratio of PMIO/Partisol Coarse Concentrations vs Wind Speed. 5.5 Summary and Conclusion A PMio inlet was developed and evaluated for the Continuous Coarse Particle Monitor developed by Misra et al. 2001. Laboratory evaluation was done using polystyrene latex particles and the cutpoint was found to be approximately 9.5 pm. The steepness of the penetration curve, the value o f a g , was calculated to be as 1.1. This indicated reasonably sharp aerodynamic particle separation characteristics. The PM io inlet sampling characteristics were then evaluated in a wind tunnel. Tests were performed at three different wind spends, viz., 3, 8 and 24 km/h. 5, 7,10, 12, 15 and 20 pm particles were generated for these tests. Results showed that the particle 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. penetration characteristics of the PMio inlet were unaffected by the wind speeds. The penetration for all the wind speeds tested, viz., 3, 8 and 24 km/h showed a very close agreement. The 50% cutpoint appeared to have shifted slightly left to around 9 pm. Field evaluation of the PMio inlet was performed in Riverside and Rubidoux, CA in the months of May and July, 2001. For the field evaluation, a 2.5 pm cutpoint round nozzle virtual impactor was attached downstream o f the developed PMio inlet. Dichotomous Partisol was used as a reference sampler. Partisol has a preselective FRM PMio inlet to remove particles larger than 10 pm aerodynamic diameter. Results showed excellent correlation between coarse PM concentrations measured by the PMio inlet and the Partisol with slope of 0.94 with R2 =0.93. XRF analysis of filters was done and coarse PM concentrations of five different crustal metals measured by the PMio inlet and Partisol were determined. Excellent agreement between the two samplers was obtained, with the average PMio inlet to Partisol coarse concentration ratio for various crustal metals being between 0.98 and 1.19. The overall coarse PM metal concentration ratio for these five crustal metals was 0.98. Nitrate and sulfates were analyzed via ion chromatography. Coarse PM nitrate concentration agreed very well (e.g., within 15%) between the PMio inlet and the Partisol. Finally, the coarse PM concentration ratios of PMio inlet and Partisol were plotted against the wind speed and the results clearly showed that this ratio was independent o f the wind speed. This further strengthens the wind tunnel tests and establishes the applicability of using the new 50 LPM PMio inlet in conditions of wind speeds at least as high as 24 km/h. 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Acknowledgements This work was supported in part by the Southern California Particle Center and Supersite (SCPCS), funded by the U.S. EPA under the STAR program, and the California A ir Resources Board. Although the research described in this article has been funded in part by the United States Environmental Protection Agency through grants # 53-4507-0482 and 53-4507-7721 to USC, it has not been subjected to the Agency’s required peer and policy review and therefore does not necessarily reflect the views of the Agency and no official endorsement should be inferred. The authors would like to express their gratitude to Mr. Erich Rupprecht (Rupprecht and Patashnick Inc., Albany, N Y) for availing the to us the PMio inlets that were used in this study, and.Drs. William Hinds and Nola Kennedy for assisting with the wind tunnel experiments at UCLA. 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References Berglund, R.N. and Liu, B.Y.H. (1973). Generation of monodisperse aerosol standards. Environ. Sci. Technol., 7:147-153 Chow, J. (1995). Measurement methods to determine compliance with ambient air quality standards for suspended particles. J. Air & Waste Manage. Assoc., 45: 320-382. Hinds, W.C. (1982). Aerosol Technology. John Wiley, New York. Hinds, W.C. and Kuo, T.L. (1995). A Low-velocity wind tunnel to evaluate inhalability and sampler performance for large dust particles. Applied Occ. Environ. Hyg., 32: 549-556 Kennedy, N.J.; Tatyan, K.; Hinds W.C. (2001). Comparison o f a simplified and full- size mannequin for the evaluation o f inhalable sampler performance. Aerosol Sci. Technol., 35: 564-568. Marple, V.A . and Willeke, K. (1976). In Fine Particles: Aerosol Generation, Measurement, Sampling, and Analysis (Edited by Liu, B .Y.H .) Academic press, New York. Misra, C., Geller, M .D ., Shah, P., Solomon, P. and Sioutas, C. (2001) Development and evaluation of a continuous coarse (PM io - PM2.5) particle monitor. J. Air & Waste Manage. Assoc., 51: 1309-1317 U.S. EPA (1996) Federal Register. 40 CFR Part 50, December 13: 65685-65701. 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 6 THE RELATIONSHIP BETWEEN BOTH REAL-TIME AND TIME- INTEGRATED COARSE (2.5-10 pm), INTERMODAL (1-2.5pm), AND FINE (<2.5pm) PARTICULATE MATTER IN THE LOS ANGELES BASIN Submitted to the Journal of the Air & Waste Management Association June 2003 6.1 Abstract Population exposure to ambient particulate matter (PM ) has received considerable attention due to the association between ambient particulate concentrations and mortality. Current toxicological and epidemiological studies and controlled human and animal exposures suggest that all size fractions o f PM may be responsible for observed health effects. Recently, the governments of European countries and the U.S. have been discussing a new PM i standard. The purpose o f this standard is to preclude invasion o f coarse particles into the fine PM mode. This notion is predicated on evidence that suggests that PM 1 .2.5 is dominated by coarse PM. In this study, coarse (PM10-PM2.5), intermodal (PM1.2.5), and fine (PM2.5) PM mass concentrations in four different sites are measured with both continuous and time-integrated sampling devices. Two source sites, USC and Downey, CA and two receptor sties, Claremont and Riverside, CA, were monitored for at least three months each. The main objective is to document both short-term and diumal variations in ambient fine, intermodal, and coarse particulate mass concentrations Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with respect to each other while considering the effects of sources, weather, wind speed, and wind direction. O f particular interest are the relationships between PMi and PMi-2.5 and coarse PM with PM1.2.5. Results show strong correlations between PMi and intermodal PM in receptor t 2 sites. These two modes in source sites show moderate correlation (R ~0.5). The contribution of PM1.2.5 to PM2.5 shows seasonal variation with the largest contribution in the summer months, most likely due to enhanced long range transport. Coarse PM is poorly correlated with intermodal PM in USC and Riverside. The correlation is dependent upon the mass concentration at Claremont, with smaller mass concentrations being moderately-to-well-correlated. This correlation becomes moderate in Downey, most likely because the local freeway is a source of all particle sizes. Continuous data yield insight into the possibility that PMi is growing into PM1.2.5 via a complex process that involves stagnation of the ambient aerosol during high relative humidity conditions, followed by advection during daytime hours. 6.2 Implications The regulatory community is currently deliberating the establishment o f a PMi standard that would eventually replace the current PM2.5 standard. The new standard is predicated on the assumption that the PMi component of PM2.5 may be more reflective of the health effects associated with fine PM, based on the fact that PM i is mainly composed o f combustion and atmospheric reaction byproducts, while PM 1 .2.5 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. is dominated by crustal materials that are associated with the coarse PM mode. While a few studies have been conducted with results that support the change to a new standard, the study locations were not representative o f all areas because they were surrounded by desert or rural areas and not downwind o f metropolitan areas. This study presents the relationship between PM 1 -2.5 and both fine and coarse PM in source and receptor sites in an urban region. 6.3 Introduction Currently, ambient particulate matter (PM ) is divided into three modes by both the regulatory and scientific communities. These modes are based on the aerodynamic diameter of the particles and were chosen due to convenience, physical properties, source, and to some extent the chemical composition. Coarse mode PM consists of particles with diameters between 2.5 and 10 pm and contains crustal metals (A l, Si, Ca, Fe, Ti) and bioaerosols (pollen, mold spores, etc.) Accumulation mode PM includes particles from 0.1 to 2.5 pm in diameter and is comprised of combustion aerosols and particles that grow from photochemical and physical processes that occur in the atmosphere. The remainder o f particles (with diameters less than 0.1 pm) comprises the ultrafine PM mode, which consists of combustion- formed particles and nucleation of vapors in the atmosphere (Hinds, 1999). Since the focus of PM regulation has switched from PM 10 (particulate matter with aerodynamic diameters less than 10 pm) to PM2.5 (particulate matter with 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. aerodynamic diameters less than 2.5 pm), both scientists and regulators have pondered whether or not another PM standard should be developed for smaller particles. The new standard in question is P M i. The reason given to set the standard at 1-pm particles is due to the sources of these particles. One prevailing theory is that 1-2.5 pm particles primarily originate from the “tail” o f the coarse mode PM mass distribution, thus these particles have the same sources as coarse particles and do not significantly contribute to the accumulation mode and more specifically to the health effects caused by fine (PM2.5) particles. The evidence for this was seen in Spokane, W A and Phoenix, AZ, where coarse and intermodal PM (PM1.2.5) were highly correlated (Haller et al., 1999; Kegler et al., 2001). Proponents of the standard argue that recent research demonstrates that PM i has greater health implications due to its sources, size, chemical composition, and results o f health studies. If the standard is approved and a sizable fraction of PM ] originates from the same sources as PM1.2.5 and/or grows to particles in the 1-2.5 pm size range, PM2.5 would essentially be studied in two halves. Dividing a mode in this fashion only complicates regulations and may ignore an important fraction o f toxic PM . The goal o f this paper is to demonstrate the correlation between intermodal PM and PMi in various areas of the Los Angeles basin. The existence o f this correlation has extreme significance for regulators because intermodal PM cannot be discounted if it is both correlated with PM] and comprises a substantial fraction of PM2.5. Furthermore, a PMi standard, if adopted, should reflect studies in multiple cities that have both urban and rural characteristics. 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.4 Methods 6.4.1 Sampling Location The instruments described below were operated inside a mobile particle laboratory developed by the Southern California Particle Center and Supersite (SCPCS) measurement and monitoring program that is funded by the US EPA. During the period of this study, measurements were conducted at four sites for about 5-12 months each and across separate seasons. From October 2000 - February 2001, sampling was done in Downey, a typical urban site in south central Los Angeles impacted mostly by primary vehicular emissions. From mid-February through August 2001, sampling was conducted in Riverside/Rubidoux, and from September through August 2002 in Claremont. Riverside/Rubidoux and Claremont are both considered receptor areas in the eastern inland valleys of the basin because they lie downwind o f the aerosol plume generated by the millions o f vehicles in the western portion of the Los Angeles Basin. This plume is advected by the predominantly westerly winds after aging for several hours to a day (Pandis et al., 1992). Riverside/Rubidoux (unlike Claremont) also lies downwind o f significant ammonia emissions from nearby farming and livestock, resulting in high concentrations of ammonium nitrate after atmospheric chemical reactions (Christoforou et al., 2000). From October 2002 through February 2003, sampling occurred near the University of Southern California (USC) at an urban site impacted by freeway emissions, local vehicle emissions, and construction site emissions due to a local sewer replacement project. 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.4.2 Instrumentation Both continuous and time-integrated data were collected for coarse, fine, and intermodal PM. The sampling devices employed included the Aerodynamic Particle Sizer (APS™, TSI Model 3320), the Beta-Attenuation Monitor (BAM™, Met-One Instruments), the Micro-Orifice Uniform Deposit Impactor (MOUDI™ , MSP Corp., Minneapolis, M N ), the Dichotomous Partisol-Plus™ (Model 2025 Sequential Air Sampler, Rupprecht and Patashnick Co. Inc., Albany, N Y ), and the Harvard/EPA Annular Denuder System (HEADS, Koutrakis et al., 1993). Additionally, intensive studies were conducted for short time intervals in which the USC Continuous Coarse Monitor (Misra et al., 2001) and Cascaded A D I Continuous Nitrate Monitor (Stolzenburg et al., 2003) were collocated with the above instruments for a short time. A ll samples were drawn through conductive stainless steel pipes with diameters proportional to sample air velocities for each instrument. Both the M O U D I and Partisol sampled approximately once per week and over time periods varying from 4 to 24 hrs, depending on location and observed pollution levels. Particles were classified by the M O U D I in the following aerodynamic particle diameter ranges: <0.10, 0.10-0.32, 0.32-0.56, 0.56-1.0, 1.0-2.5, and 2.5-10 pm. Teflon filters with diameters of 4.7 and 3.7 cm (2 pm pore size, Gelman Science, Ann Arbor, M I) were used to collect particles in the M O U D I stages and after-filter, respectively. The Partisol uses a PMio inlet operating at 16.7 LPM to remove particles larger than 10 pm in aerodynamic diameter. The remaining PMio aerosol is drawn 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. through a virtual impactor, or, “dichotomous splitter”, located after the inlet. Two separate flow controllers maintain coarse aerosol velocity at 1.67 LPM and that of the fine aerosol stream at 15 LPM. Coarse and fine particles are collected on two 4.7 cm Teflon filters, which are housed in reusable cassettes placed in the minor and major flows o f the Partisol virtual impactor. The Teflon filters o f both the MOUDI™ and Partisol™ samplers were pre- and post-weighed using a Mettler Microbalance (M T5, Mettler-Toledo, Inc, Hightstown, NJ) after 24-hour equilibration under controlled humidity (40 ± 5%) and temperature (24 ± 3 °C) to determine particle mass concentrations. The Teflon filters were then used to determine sulfate and nitrate concentrations by ion chromatography. For measurement o f metals and trace elements, a second set of Teflon filters was collected in a second M O U D I configured identically to the first. After weighing, filters were analyzed by X-ray fluorescence for metals and other trace elements. Samples to determine the size-fractionated concentrations of elemental carbon (EC) and organic carbon (OC) were obtained by simultaneous sampling with a third M O UDI. 47 mm aluminum substrates were used for the impaction stages and a 37 mm quartz fiber filter (Pallflex Corp., Putnam, CT) was used as the after-filter (ultrafine stage). EC and OC values were determined using the Thermal Evolution/Optical Transmittance (TOT) analysis of Birch and Cary (1996). Concurrent to the 24-hour M O U D I sampling, fine and coarse mass measurements were performed with 47 mm Teflon filters in a dichotomous sampler (Partisol-Plus™, Model 2025 Sequential A ir Sampler, Rupprecht and Patashnick Co. 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Inc., Albany, N Y ). Mass and elemental concentrations o f coarse and' fine size fractions were determined by the same methods as described previously for M O U D I i sampling. Partisol results were compared to M O U D I data to check for consistency, and, in a few cases, used in the analysis when M O U D I results were not available. In addition to time integrated chemical composition data, the USC Continuous Coarse Monitor (Misra et al., 2001) operated during a winter intensive study at the USC site. The operating principle of the monitor is based on enriching coarse particle concentrations by a factor of about 25 by means of a 2.5-pm cutpoint round nozzle virtual impactor, while maintaining mass o f PM 2.5 at ambient concentrations. The aerosol mixture is subsequently drawn through a Tapered Element Oscillating Microbalance (TEOM™ 1400A, Rupprecht and Patashnick, Albany, N Y ), the response of which is dominated by the contributions o f coarse PM due to concentration enrichment. Another intensive study was conducted during the month o f September 2001 in Claremont, in which the newly developed A D I Continuous Nitrate Monitor was located at the Claremont site for an intensive study. The operation and characterization of this device is described by Stolzenburg et al. (2003) and Fine et al. (2003). 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.5 Results and Discussion PM i-2.5 and PM2.5 mass concentrations are very highly correlated (R2 = 0.76) for all sites (Figure 6.1). Figure 6.1 also indicates that, contrary to the prevailing perception, the 1-2.5 pm range accounts for a substantial fraction o f the total PM2.5 mass, ranging from 20-45%, depending on location and season, as it w ill be discussed in following sections. The chemical characteristics of intermodal PM are displayed by site in Figure 6.2. These results are similar to those discovered by Hughes et al. (2000), in which they determined the chemical composition o f the fine PM mode at four locations in Southern California. In contrast, the coarse size mode does not demonstrate as high o f a positive correlation with intermodal PM. Intermodal PM tracks much better with PM i than the coarse mode in receptor sites, 45 - y = 0.43x - 2.92 R2 = 0.76 i 40 - 30 - O 2 0 - 20 0 40 100 60 80 PM2 .5 Mass Concentration (pg/m3 ) Figure 6.1. PM ].2.5 Versus PM2.5 at A ll Sites 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. □ N 0 3 E S 0 4 B O C □ Metals BOther -m -m : wm + > > > > < # I V . V . V . V . * w . v . v . v v .v .s v .v • v w v s a a * v w w w w w w v w Downey Riverside Claremont USC Location Figure 6.2. Chemical Composition of Intermodal PM Averaged By Location while the correlations between coarse and intermodal PM as well as between PM] and intermodal PM become closer in source sites. The following section will describe this in detail. 6.5.1 Site-by-Site Comparison University of Southern California Figure 6.3a shows the relationship between intermodal PM and the coarse mode at the USC sampling site 1 mile south of downtown Los Angeles. The correlation coefficient (R2 ) of 0.11 reflects the divergence of these two modes in the ambient air near this site. Because the site is urban and crustal particles are not locally emitted, this result is not unusual. The site is also not directly impacted by a 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14 freeway (unlike the Downey site), which would emit coarse particles in the form of road dust. Intermodal PM correlates moderately with PM i, however, with an R2 of 0.53 (Figure 6.3b). At a source site such as USC, this correlation is most likely driven by direct emission of both PM] and intermodal particles from the same automobile sources. The continuous data displayed in Figures 6.4a,b represent two weeks o f intensive sampling at USC. Winter days are normally cool (~10°C) and humid (>70%) in the mornings and dry (<40%) and temperate (~19°C) in the afternoons. Figure 6.4a illustrates a week during early winter in which a Santa Ana wind event occurred froml 1/25 to 11/29 and a rain event occurred from 11/29 to 12/1. From the graph it is apparent that coarse mode concentrations increase and wind becomes erratic at the beginning of the week. Intermodal PM tracks the coarse mode during this time because Santa Ana winds blow crustal particles from the California deserts while creating hot and dry conditions that do not favor particulate growth. The drastic mass concentration drop in all size modes marks the arrival o f the rainstorm. Figure 6.4b presents a more typical winter week in December 2002. Intermodal PM concentration tracks with both PM2.5 and coarse concentrations. The continuous data do not show whether intermodal P M is more closely associated with coarse or fine PM during this week as all modes appear interrelated. For the week o f 12/02/02 to 12/09/02, a correlation was performed between the continuous intermodal mass data and both the coarse and fine PM data. The resulting Pearson correlation coefficients (R-values) were 0.35 and 0.64, respectively. This demonstrates the obvious 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 E 8 D ) a 7 c o k . 6 = 5 0 o o 4 O ( A o 0 ) ^ ( 0 S 2 U ) 2 1 Q . 0 50 40 0 10 20 30 PM2 5 .1 0 Mass Concentration (pg/m3 ) Figure 6.3a. Intermodal Versus Coarse PM at USC 20 18 j± 16 1 14 ‘ iM S 12 c O 10 O 8 < / > W R « b 2 i n 4 E N E ^ 0 0 10 20 30 40 50 60 PM1 Mass Concentration (ng/m3 ) Figure 6.3b. Intermodal Versus PMi at USC 137 y = 0.21x + 0.97 R2 = 0.53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Coarse TEOM Coarse APS Intermodal PM Wind Speed Figure 6.4a. Coarse, Intermodal, and Fine 11/25-12/1 2002 45 ^ 40 E 35 o > A 30 25 c o a > o c o o 1 0 ( 0 r a 20 ! 15 10 - ■ — * — Coarse TEOM — A— Intermodal PM APS — m — Coarse APJ — • — Fine BAM ■ > n i ft n « L_ j k t p Jl S \ f l m J t 11 .J T T i i p i t TT V L\ Iftlm f J V " f l ft 4 _ ra F j& i © f y £ $ < § * & $ < v £ V < § * <V fy s' < v < V » ■ * * £ < V < § > fy s £ £ < v 5s 5 5 5 5 ry o ' < v fy ^ < v < V w K . Date < § > « § < y q > * <v # £ © © cy' o ' t v f y © v § § < § * t^' ^ < v » < v A A # < \ > # V > - v 5 < § S’ < v / £ v Figure 6.4b. 12/09/2002 Coarse, Intermodal, and Fine PM Mass Concentrations 12/02- 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. correlation between intermodal and fine PM while a much weaker one exists between intermodal and coarse PM. i Downey, C a lifo rn ia Figures 6.5a,b demonstrate the correlations between intermodal PM and both coarse and PM] mass concentrations, respectively. The correlations between intermodal PM and the other two modes are similar with R2 =0.49 for intermodal- coarse and R =0.47 for intermodal-PMi. A possible explanation for this finding is that the site is located downwind of a high capacity freeway (1-710) with a large number of trucks and automobiles, and which may be the dominant particle source at that site for particles o f all ranges. Coarse particles in Downey are dominated by resuspended road and tire dust while fine particles are a mixture of about 40% combustion emissions and 30% resuspended road and tire dust (Singh et al., 2002). Moreover, this site is not near any source of crustal particles (i.e. all nearby surfaces are paved), so intermodal PM correlation with coarse PM is most likely driven by road dust, which is known to contain toxic compounds (Chow et al., 2003). Riverside, C a lifo rn ia Intermodal PM is graphed versus coarse PM in Figure 6.6a. The R2 o f 0.14 is very surprising in this location because it is a rural site with local dust emissions. This result indicates that the tail of the coarse PM distribution is not significantly affecting intermodal concentrations and thus fine PM concentrations. The 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. O ) a. c o c « o c o o M 1 0 m 2 2 Q . 14 y = 0.55x + 2.05 R2 = 0.49 12 10 8 6 ♦ ♦ 4 2 0 16 12 14 0 2 6 8 10 4 PM2.5 .io Mass Concentration (ng/m3) Figure 6.5a. Intermodal Versus Coarse PM at Downey, CA I 12 ' o > a. c a > u c o o ( A W n S i n 2 a. 0 10 50 20 30 40 PM1 Mass Concentration (ug/m3 ) Figure 6.5b. Intermodal Versus PMi at Downey, CA Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. relationship between intermodal PM and PM i is strong at this site (Figure 6 .6b). The R2 of 0.74 is significant and may be explained by condensational growth of particles less than 1 pm in diameter to particles between 1 and 2.5 pm in diameter. The sub micron particles are emitted west of Riverside in the Los Angeles area and grow while advection transports them to Riverside. This process often occurs over 6-12 hour time intervals (Allen et al., 2000). While a linear correlation does provide a good fit, the data in Figure 6 .6b appear to have an exponential trend. To explore this further, the data points for which chemical data were known were split into two regions. Group 1 (n = 5) included PM 1 .2.5 concentrations less than 5 pg/m3 and group 2 (n = 4) included those points in which PM 1 .2.5 was greater than 5 pg/m3 . These groups were then statistically tested to determine whether their chemical compositions are significantly different based on the following three parameters: nitrate fraction, organic carbon fraction, and sulfate fraction. The non-parametric Wilcoxon Rank Sum test was employed because o f small sample size and assumed non-normality of the distribution. The results of this test showed that the median nitrate fraction of intermodal concentrations less than 5 pg/m3 in Riverside is significantly lower than the median nitrate fraction of intermodal concentrations greater than 5 pg/m (p < 0.01). This test also confirmed that median organic carbon fraction o f intermodal concentrations less than 5 pg/m3 in Riverside is significantly higher than the median organic carbon fraction of intermodal concentrations greater than 5 pg/m3 (p = 0.026). Median sulfate fraction was nearly significantly lower in group 1 than group 141 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ^ 45 - o > o 35 - y = 0.38x + 3.62 R2 = 0.14 40 - ^ 30 - c 3 25 o O 20 M w 1 C » 15 - 5 i n 10 - ♦ ♦ s a. 40 50 60 20 30 0 10 P M 2.s.io Mass Concentration (ng/m3 ) Figure 6.6a. Intermodal Versus Coarse PM for Riverside/Rubidoux, CA "E 45 - 1 40 - o 35 y= 1.27x-10.96 2 + * c A O c 30 - 25 o O 20 - w <° 1C A 1 5 - S i n 10 - s Q . 0 10 15 5 20 25 30 35 40 PM1 Mass Concentration (fxg/m3 ) Figure 6.6b. Intermodal Versus PMi at Riverside/Rubidoux, CA 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 (p = 0.111). These results reinforce our hypothesis that higher PM 1 .2.5 concentrations are due to increases in hydroscopic compounds such as nitrate and sulfate instead o f the more hydrophobic organic carbon. Clarem ont, C a lifo rn ia At first glance, intermodal PM are not correlated with coarse mode PM with an R2 = 0.10 (Figure 6.7a). However, the correlation increases to R2 = 0.43 when all points with intermodal PM mass concentrations greater than 8 pg/m3 are excluded, suggesting that there is some association between coarse and intermodal PM in that location, but the highest concentrations in the 1 - 2.5 pm range are not associated with days during which coarse PM concentrations were high. This is counterintuitive to the argument that the tail o f the coarse mode contributes to fine PM' during times when high coarse concentrations are measured. Because Claremont is also a rural site, coarse PM would be expected to correlate with intermodal PM, but this is only true for relatively low coarse PM mass concentrations. Intermodal PM correlates well (R2 = 0.65) with PMi without excluding high mass concentrations (Figure 6.7b). Thus, at the receptor site Claremont, advection o f fine PM outweighs the local coarse emissions with respect to contributions to intermodal PM. 6.5.2 Comparisons Between PM Modes Based on Chemical Composition Particulate nitrate is the predominant chemical constituent of the 1-2.5 pm range,1 accounting (with the exception of the Downey site) for about 40 - 65% of the 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C O I 1 6- = l c 14 T o 12 2 g 1 0 - o § 8 O w 6 - ( 0 n 2 4 u > * 9 - ♦♦ s CL 0 10 20 30 40 50 PM2 .5.1 0 Mass Concentration (fig/m3 ) Figure 6.7a. Intermodal Versus Coarse PM in Claremont, CA n E 16 - O ) 3 1 4 - c ~ 12 - y = 0.47x - 0.86 R2 = 0.65 4 3 2 g 10 - o o 8 - o I # (3 W o n * . 4 - C NI r* 2 0. 0 5 10 15 20 25 30 PMi Mass Concentration (fig/m3 ) Figure 6.7b. Intermodal Versus PMi for Claremont, CA 144 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. total mass in that range (Figure 6.2). The relationship between coarse, intermodal and fine PM was also investigated for this species. The correlation between coarse and intermodal nitrate is weak (R2 = 0.13), which demonstrates the divergence of the sources o f these particles (Figure 6.8a). Previous studies in Southern California indicated that coarse mode nitrate is a mixture of sodium and ammonium nitrate, while nitrate in the fine mode is mostly ammonium nitrate (Kleeman et al., 1999). Figure 6.8b shows that intermodal nitrate is well correlated with PM2.5 nitrate with R2 = 0.70, and it also comprises a substantial fraction of PM2.5 nitrate (slope = 0.32). A similarly high degree o f correlation between continuously measured intermodal nitrate and PM i nitrate (R2 = 0.80) is shown in Figure 6.9, using the data generated by the Cascaded A D I Continuous Nitrate Monitor (Stolzenburg et al., 2003). These two figures confirm that nitrate in the PM 1 .2.5 is a significant portion o f total PM2.5 bound nitrate, probably originating from the sub-micrometer range by condensational growth. The exact mechanism through which accumulation mode PM grows to the micrometer range have been debated in several previous publications, including aqueous phase reactions (Hering and Friedlander, 1982) as well as activation of sub-0.5 pm particles to form fog or cloud droplets, followed by aqueous phase chemistry and fog evaporation (Meng et al., 1994). Growth of hygroscopic ambient PM beyond the 1 pm range has been observed in several other studies (Koutrakis et al., 1989; Kelly and Koutrakis, 1998) when relative humidities reach greater than ninety percent. i 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. y = 0.51x + 2.83 R2 = 0.13 0 6 8 10 PM^q-2.5 Nitrate Concentrations (pg/m) Figure 6.8a. PM2.5-10 Vs. PM1.0-2.5 Nitrate Mass Concentrations for Claremont, CA o > c o « 3 S 4 - 1 c ® o c o O Q > 2 a. 14 y = 0.32x + 0.04 R2 = 0.70 12 10 8 6 4 2 0 5 10 15 20 25 PM2 .s Nitrate Concentration (pg/m3 ) 30 Figure 6.8b. PM1.2.5 Versus PM2.5 Nitrate Concentrations at A ll Sites 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 0 2 4 6 8 10 12 14 16 PM! Nitrate (pg/m3 ) Figure 6.9. Continuous PM 1 .2.5 Versus PMi Nitrate at Claremont, CA in September 2001 The period between April and July in the Los Angeles basin is characterized by frequent fog-like conditions with high relative humidities in the overnight and early morning hours. Particles emitted mostly to the west travel by advection eastwards towards the inland valleys o f the basin, such as Claremont. While in transit, these particles experience condensational growth and participate in photochemical reactions. Some particles remain in the sub-micron range after growth, while others become intermodal particles. This process takes place over multi-hour time spans, whereby PMi particles may grow into PM1.2.5 following several hours of advection and stagnation. Growth of sub-micrometer nitrate into the super-micrometer range is also supported by previous studies showing very similar 147 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. chemical composition between PMi and PM 1 -2.5 (Hughes et al., 1999; Kleeman et al., 1999; Hughes et al., 2000). The data plotted in Figure 6.10 further support the argument that intermodal nitrate originates from fine and not coarse PM. While wind speed peaks at 3 PM, the ratio of intermodal to PMj nitrate peaks at 6 PM. The Claremont site is surrounded by many large unpaved areas and gravel pits. If the wind were creating resuspension of local coarse particles that were in turn affecting intermodal PM concentrations, wind speed would peak concurrently with mass concentration. Advection, however, would be represented by the time lag seen here. After wind speed peaks, intermodal PM nitrate particles, that have been undergoing photochemistry and growth, blow toward Claremont and peak shortly thereafter. 0.5 0.45 0.4 0.35 © 0.3 | 0.25 J ~ 3 > 0.2 2 2 0.15 0.1 0.05 0 0 4 8 12 16 20 24 Hour Figure 6.10. Daily Wind Speed and PM 1 -2.5/PM 2.5 Nitrate at Claremont, CA in September 2001 1 148 £ a E T J 0 0 a c o T3 C 5 WS Ratio 3 2 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6.11 also illustrates the photochemical growth and advection o f intermodal PM in the Los Angeles basin. The mass concentration ratio o f intermodal to fine mode PM in all sites (based on time integrated data) increases during the summer months due to increased solar radiation and enhanced advection. 0.55 0.5 0.45 0.4 U ) I 0.35 " " " I h j 0.3 Q . 0.25 0.2 0.15 0.1 X X X • • ♦ DW ■ RV • CL XUSC ■ ■ ------------------ • « ♦ ♦ /K • • X X ♦ • ~W ' ' • • X . X Jan 1 1 Feb ! 1 Mar 1 1 1 Apr May Jun Jul 1 1 1 Aug Sep Oct 1 1 Nov Dec Month Figure 6.11. Monthly Average of the Ratio o f Intermodal PM to Total PM2.5 Figures 6.12a and b show the OC and sulfate concentrations vs. nitrate concentrations for the 1 - 2.5 pm PM range in source and receptors sites, respectively. Nitrate and sulfate are correlated for this size range in source sites (R2 = 0.63) whereas the OC concentrations are poorly correlated with nitrate concentrations (R2 = 0.14). This suggest that even in the source sites, nitrate and sulfate in the intermodal PM range share a common origin,, i.e., secondary formation and growth by condensation into the super-micrometer range, whereas OC in that 149 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. oOC o > y = 0.45x - 0.13 R2 = 0.63 ■ Sulfate o m 0 4 1 2 3 5 6 7 8 Nitrate Concentration (pg/m3 ) Figure 6.12a. OC and Sulfate Versus Nitrate Concentration in the 1 - 2.5 jam Range at Source Sites 5 oOC 4 y = 0.29x - 0.01 R2 = 0.77 ■ Sulfate 3 2 1 J C 0 0 2 4 6 8 10 12 14 Nitrate Concentration (p.g/m3 ) Figure 6.12b. OC and Sulfate Versus Nitrate Concentration in the 1-2.5 pm Range at Receptor Sites 150 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. range originates most likely from traffic road dust (the correlation between OC concentrations in the 1 - 2.5 fjm and 2.5 - 10 pm size ranges yielded a R2 = 0.61. A • • • 2 similar correlation between these two OC modes in receptor sites yielded a R =0.21 suggesting that the presence o f OC in that range is not due to road dust in receptor areas). The sulfate and nitrate concentrations of the 1 - 2.5 pm range in receptor sites are very well correlated (R2 = 0.77) while moderate correlations (R2 = 0.51) were also observed between the nitrate and OC concentrations. These results are consistent with the findings of John et al. (1990) that nitrate and sulfate are uniformly mixed in the so-called “droplet” mode, defined as one containing accumulation mode PM exceeding about 0.5-0.7 pm in diameter. Whether organics are externally or internally mixed is not clear from our data. However, Pandis et al. (Pandis et al., 1993) showed that organics in the larger size range of the accumulation mode can result only if there exist sufficient primary particles in the > 0.5 pm range and/or if the condensable organic species have a strong affinity for that size range. The former condition is consistent with our field data of several years in the Los Angeles Basin, showing that the aerosol size distribution in receptor areas o f this basin contains a much larger number of particles in the larger size range of the accumulation mode compared to source sites (Kim et al., 2002; Fine et al., 2003), as result of aerosol aging in the atmosphere due to advection and long-range transport. The existence o f a pronounced inorganic “droplet” mode, which possibly extends beyond the 1 pm range, w ill likely influence the presence of organics in that range 151 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. through condensation, given that at least some portion of OC is soluble (Meng et al., 1994). Figure 6.13 displays the relationship between coarse and intermodal crustal metal concentrations. The crustal metal concentrations were obtained by analyzing M O U D I Teflon substrates via x-ray fluorescence (XRF) and utilizing the following formula (Malm et al., 1994): PM so il = 2.2 * Al + 2.49 * Si + 1.63 * Ca + 2.42 * Fe +1.94 * Ti y = 0.09x + 0.63 R2 = 0.49 "g 3.5 c o 2 2.5 ♦♦ c •« o c o o o C O i n i 0.5 0 5 10 15 20 25 30 35 PM2 .5 .io Soil Concentration (ng/m3 ) Figure 6.13. Intermodal Versus Coarse PM Crustal Elements for Claremont, CA The moderate correlation (R2 = 0.49) between intermodal and coarse crustal metals is expected because the tail of the coarse mode crustals infiltrates PM2.5 to a limited extent, which is indicated by the rather low concentrations o f crustal metals found in 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. intermodal PM. The relationship between intermodal PM mass to its soil (crustal) component is shown in Figure 6.14. As evident by the slope of 3.42, the soil component o f PM 1 .2.5 is less than 25% of the total mass and has a moderate correlation with total mass (R2 = 0.54). This result differs from the Kegler et al. (2001) study in that they found a similar correlation between intermodal mass and soil concentrations, but with a much higher slope. The crustal component in Spokane, W A is closer to 50% of the total intermodal mass, which is to be expected in a city that is surrounded by rural areas. Although the contribution o f crustal metals to the intermodal PM mass is not negligible, it is far lower than that of nitrate, as illustrated in Figure 6.2. This may also explain the overall low correlation between the intermodal and coarse PM concentrations obtained in our study. y = 3.11x - 0.28 R2 = 0.54 ■ T * 16 - d L 14 - 0 0.5 1 1.5 2 2.5 3 3.5 4 Estimated PM^.s Soil Concentration (fig/m3 ) Figure 6.14. PM1-2.5 Versus Estimated PM1.2.5 Soil Concentration 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6.15 displays the lack of correlation between the most prominent chemical species that comprise intermodal PM and the crustal metals found in that mode. While nitrate, sulfate and (at least in receptor sites) OC correlate well with one another, none correlate with the crustal component. While the crustal component o f intermodal PM is moderately correlated with the intermodal mass (Figure 6.14) the majority of the mass, consisting of nitrate, sulfate, and OC, does not correlate with the crustal component and is chemically more similar to particles in the PMi range. 14 3 12 | 10 n w *» c ® Q o o c o O w 6 _ a > w 4 m o ♦ OC I Sulfate A Nitrate y = 0.31x + 0.43 R2 = 0.16 y = 0.04x + 0.58 R2 = 0.00 y = 0.21x + 1.77 R2 = 0.00 ■ a A A A A J® L 2 3 Soil Concentration (pg/m3 ) Figure 6.15. OC, Nitrate and Sulfate Concentrations Versus Soil Concentration in Intermodal PM 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.6 Summary and Conclusions Although previous research does suggest a relationship between coarse mode and intermodal PM, the sites in which these studies were conducted are not representative o f all locations. Similar data for coarse, intermodal, and fine PM were collected across four sites in the Los Angeles Basin. This study included a large database of time-integrated samples that spans nearly three years and complementary continuous measurements during intensive campaigns. While some similarities exist between these results and those of comparable studies, the main finding established here is that intermodal PM consists of a significant portion of particles that are similar in chemical composition to smaller particles that are thought to cause the greatest health effects. In general, some fraction of intermodal PM originates from the lower-size range “tail” of the coarse PM size distribution. In Los Angeles, however, that correlation is not as strong as the one between PMi and intermodal PM. Even the rural locations in this study demonstrated high correlations between PMi and intermodal PM, which validates the strength o f the PM2.5 standard for locations that have both a crustal source and advected aerosol from an urban area upwind. The receptor sites in this study showed a peak in the ratio between intermodal nitrate and fine nitrate in the early evening, which was three hours after the peak wind speed, indicating advection o f particulate nitrate from upwind sources and growth into the intermodal size range. Overall, intermodal nitrate correlated very well with both PM i and PM2.5 nitrate, signifying its strong relationship to the fine 155 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. mode. Intermodal sulfate and nitrate demonstrated similar correlations and were also correlated with each other and OC in receptor sites. Intermodal crustal material did not correlate with any other chemical constituent. This study was performed to shed light on the origin and chemical composition of intermodal particles between the coarse and fine PM modes in Los Angeles, a unique city where crustal, oceanic, anthropogenic primary, and secondary sources are responsible for the high observed PM levels. Our results indicate that a PMi standard would not constitute an unambiguous separation of coarse and fine mode PM in this urban air shed. Further studies at various locations are warranted, especially sites in areas o f the eastern United States where air parcels are advected across much larger distances than those in Los Angeles, in order to determine the degree to which the promulgation of a PM i standard would be justifiable. Acknowledgements This work was supported by the Southern California Particle Center and Supersite (SCPCS), funded by EPA under the STAR program through Grants #53- 4507-0482 and 53-4507-7721 to the University of Southern California (USC). EPA, through its Office o f Research and Development, collaborated in this research and preparation of this manuscript. The manuscript has been subjected to Agency review and approved for publication. Mention o f trade names or commercial products does not constitute an endorsement or recommendation for use. 156 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References Allen, J.O., Hughes, L.S., Salmon, L.G., Mayo, P.R., Johnson, R.J., Cass, G.R. Characterization and Evolution of Primary and Secondary Aerosols during PM2.5 and PM10 Episodes in the South Coast Air Basin, Report A-22 to the Coordinating Research Council (CRC). 2000. Birch, M.E. and Cary, R.A. (1996). Elemental carbon-based method for monitoring occupational exposures to particulate diesel exhaust. Aerosol Sci. &Technol., 25(3): 221-241. Chow, J.C.; Watson, J.G.; Ashbaugh, L.L.; Magliano, K.L. (2003). Similarities and differences in PM 10 chemical source profiles for geological dust from the San Joaquin Valley, California. Atmos. Environ., 37(9-10): 1317-1340. Christoforou, C.S., Salmon, L.G., Hannigan, M.P., Solomon, P.A., Cass, G.R. (2000). Trends in fine particle concentration and chemical composition in Southern California. J. Air & Waste Manage. Assoc., 50: 43-53. Fine, P.M., Jaques, P.A., Hering, S.V., Sioutas, C. (2003). Performance evaluation and use o f a continuous monitor for measuring size-fractionated PM2.5 nitrate. Aerosol Sci. Technol., 36(4): 342-354. Fine, P. M .; Shen, S.; Sioutas, C. Inferring the sources of fine and ultrafine particulate matter at downwind receptor sites in the Los Angeles Basin using multiple continuous measurements; Accepted by Aerosol Sci. & Technol. February, 2003. Haller, L., Claibom, C., Larson, T., Koenig, J., Norris, G., Edgar, R. (1999). Airborne particulate matter size distributions in an arid urban area. J. Air & Waste Manage. Assoc., 49: 161-168. Harrison, R.M. and Pio, C.A. (1983). Size differentiated composition o f inorganic atmospheric aerosols of both marine and polluted continental origin. Atmos. Environ. 17 (9): 1737-1738. Hering S.V. and Friedlander, S.K. (1982). Origins of aerosol sulfur size distributions in the Los Angeles basin. Atmos. Environ., 16(11): 2647-2656. Hinds, W.C. Aerosol Technology: Properties, Behavior, and Measurement o f Airborne Particles. New York: John Wiley & Sons, Inc., 1999. 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hughes, L.S.; Allen, J.O.; Bhave, P.; Kleeman, M.J.; Cass, G.R.; Liu, D.Y.; Fergenson, D.P.; Morrical, B.D.; Prather, K.A. (2000). Evolution of atmospheric particles along trajectories crossing the Los Angeles Basin. Environ. Sci. & Technol., 34(15): 3058-3068. Hughes, L.S.; Allen, J.O.; Kleeman, M.J.; Johnson, R.J.; Cass, G.R.; Gross, D.S.; Gard, E.E.; Morrical, B.D.; Fergenson, D.P.; Dienes, T.; Noble, C.A.; Liu, D.Y.; Silva, P.J.; Prather, K.A. (1999). Size and composition distribution of atmospheric particles in Southern California. Environ. Sci. & Technol., 33(20): 3506-3515. John, W.; W all, S.M.; Ondo, J.L.; Winklmayr, W. (1990). Modes in the size distributions of atmospheric inorganic aerosol. Atmos. Environ., 24(9): 2349- 2359. Kegler, S.R., Wilson, W .E., Marcus, A.H. (2001). PM i, intermodal (PM 2.5-1 ) mass, and the soil component of PM2.5 in Phoenix, AZ, 1995-1996. Aerosol Sci. & Technol., 35: 914-920. Kelly, B.P. and Koutrakis, P. (1998). Equilibrium size o f atmospheric aerosol sulfates as a function of particle acidity and ambient relative humidity. J. Geophys. Res.-Atmos., D4: 7141-7147. Kim S.; Shen S.; Sioutas C.; Zhu, Y.; Hinds, W.C. (2002). Size distribution and diurnal and seasonal trends of ultrafine particles in source and receptor sites of the Los Angeles basin. J. Air Waste Manage., 52(3): 297-307. Kleeman, M.J.; Hughes, L.S.; Allen, J.O.; Cass, G.R. (1999). Source contribution to the size and composition distribution of atmospheric particles: Southern California in September 1996. Environ. Sci. & Technol., 33(23): 4331-4341. Koutrakis, P., Wolfson, J.M., Spengler, J.D., Stem, B., Franklin, C.A. (1989). Equilibrium size of atmospheric aerosol sulfates as a function of the relative humidity. J. Geophys. Research, 94 (D5): 6442-6448. Koutrakis, P., Sioutas, C., Ferguson, S.T., 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: 2497-2501. Laden, F.; Neas, L.M .; Schwartz, J. (1999). The association of cmstal particles in the fine particulate fraction with daily mortality in six U.S. cities. American Journal o f Respiratory and Critical Care Medicine, 159: A332. 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Malm, W.C., Sisler, J.F., Huffman, D., Eldred, R.A, Cahill, T.A. (1994). Spacial and seasonal trends in particle concentration and optical extinction in the United States. J. Geophys. Res., 99: 1347-1370. Misra, C.; Geller, M .D.; Shah, P.; Sioutas, C. Solomon, P.A. (2001). Development and evaluation of a continuous coarse (PM10-PM2.5) particle monitor. J. Air & Waste Manage. Assoc., 51 (9): 1309-1317. Meng Z.Y. and Seinfeld, J.H. (1994). On the source of the submicrometer droplet mode o f urban and regional aerosols. Aerosol Sci. Tech., 20(3): 253-265. Pandis, S.N., Harley, R.A., Cass, G.R., Seinfeld, J.H. (1992). Secondary organic aerosol formation and transport. Atmos. Environ., 26A: 2269-2282. Pandis, S.N.; Wexler, A.S.; Seinfeld, J.H. (1993). Secondary organic aerosol formation and transport 2: Predicting the ambient secondary organic aerosol size distribution. Atmos. Environ., 27(15): 2403-2416. Singh, M ., Jaques, P. and Sioutas, C. (2002). Particle-bound metals in source and receptor sites of the Los Angeles Basin. Atmos. Environ., 36(10): 1675-1685. Stolzenburg M .R ., Dutcher D.D, Kirby B.W ., Hering S.V. (2003). Automated measurement o f the size and concentration of airborne particulate nitrate. Aerosol Sci. Technol., 37: 537 - 546. 159 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 7 CONCLUSION 7.1 Summary A study was conducted in the Coachella Valley in Southern California to determine the indoor penetration o f coarse particles in a site heavily influenced by crustal sources during a time o f maximum penetration indoors. The findings o f this study were that outdoor coarse particles do not efficiently penetrate indoors, and thus indoor sources of these particles are more important for exposure assessment. Fine PM was found to dominate PM j o in indoor environments. In order to study the relatively low ambient concentrations o f coarse particles, a high volume coarse particle concentrator was developed and characterized. By concentrating the ambient coarse PM fraction, health researchers can conduct dose-response relationships, and aerosol scientists can measure these particles over shorter time intervals. The desired endpoint is reduction of measurement variability o f this PM mode. The CPC can concentrate coarse PM levels in the 1-5 micrograms per cubic meter range, which can then be delivered to animal or human exposure chambers and/or collection devices. While time-integrated samples have the advantage of yielding a sample with enough mass to perform many chemical and biological assays, they may mask the variations that occur over the course of the sample interval. Since this especially true I for coarse PM measurement, the Continuous Coarse Particle Monitor was developed. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This device measures real time coarse PM mass concentration by concentrating the coarse fraction o f PMio prior to delivery to a TEOM continuous mass monitor. This device has many relevant future applications in which the user must know the coarse concentration at a particular time, such as ambient monitoring and in vivo human exposure studies. A PMio inlet was developed in order to accurately separate particles larger than 10 pm in aerodynamic diameter from the ambient aerosol prior to sampling with the CCPM. This was accomplished by modifying a currently available inlet for the flow rate requirement of the CCPM. The addition of this inlet to the CCPM greatly reduces the error incurred if no inlet is employed. The coarse mode’s intrusion into the fine PM mode was explored in Chapter 6 . While coarse and intermodal PM did correlate with each other in some instances, intermodal PM correlated much better with total PM2.5 and even P M i. If PMi is in fact a source of intermodal PM , then intermodal PM may be growing into coarse particles. This may explain some of the correlation that exists between these two size ranges. The main finding established here is that intermodal PM consists of a significant portion of particles that are similar in chemical composition to smaller particles that are thought to cause health effects. In general, some fraction of intermodal PM originates from the lower-size range “tail” o f the coarse PM size distribution. In Los Angeles, however, that correlation is not as strong as the one between PM i and intermediate PM. Even the mral locations in this study demonstrated high correlations between PM i and intermodal PM , which validates the 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. strength o f the PM2.5 standard for locations that have both a crustal source and advected aerosol from an urban area upwind. Results indicate that a PM i standard would not constitute an unambiguous separation o f coarse and fine mode PM in this urban air shed. 7.2 Conclusions This thesis presented the idea that while the behavior o f coarse particles has been studied for a very long time, the chemical composition and health effects of coarse particulate matter as a whole is variable depending upon time and location and must therefore be examined with different tools. The high degree of variability surrounding coarse PM results in increased uncertainty when trying to assess its health effects. Prior to this research, many have concluded that the coarse mode contains mostly benign materials due to its crustal component. The investigation presented here demonstrates that coarse PM found at certain locations in the Los Angeles Basin can contain significant amounts of nitrate, sulfate, and organic carbon, all of which have been attributed to health effects. The aforementioned technologies aim at reducing the inconsistency in coarse PM measurements while increasing the body of knowledge surrounding this size mode o f ambient particles. The coarse particle concentrator and continuous coarse monitor have already been employed in two “in vivo” health studies— one exposing animals and one exposing humans— and many in vitro assays. The preliminary results implicate coarse PM as a cause of adverse health effects in animals and i humans. 162 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The continuous coarse monitor has also been used to measure continuous coarse PM mass concentrations for intensive studies in the Los Angeles Basin and is now being considered for approval as an EPA reference method for monitoring this size range o f particles. 7.3 Recommendations for Future Research The studies that comprise this thesis demonstrate a small blueprint that can be expanded to a large-scale investigation. Obviously, some aspects o f coarse particles are very well understood and need not be studied in further detail, such as physical characteristics and chemical composition at some locations. There is still insufficient evidence that resolves a person’s exposure to coarse PM and describes the mechanism by which coarse particles harm humans. The only method of fixing these inadequacies is by performing many studies that target human exposure and health with respect to coarse PM. Health effects researchers have already begun using particle concentrators in their studies, and the future w ill most likely see the role of concentrators increase in this research. Based on a coarse particle human exposure study of which I was a part, the delivery system and measurement of coarse particles are the two major components o f the coarse particle concentration system that need improvement. Because coarse particles have very high losses due to impaction and settling, new particle delivery systems should consist o f small, well-mixed chambers that can ideally fit over a subject’s breathing zone (i.e. not full body exposures): Real-time 163 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. measurement instruments should sample from this well-mixed breathing zone chamber, which w ill yield the most accurate PM measurement. The characterization o f human exposure to coarse PM has been very difficult because studies usually utilize one of two sampling methods. In some cases an impactor or filter sampler is placed in a common area where the majority of the subject spends his time. While this allows for a larger sampler and thus higher flow rates, the sample is not representative of what the subject actually breathes because coarse particles are certainly not uniformly distributed in a room. The second method is a personal sampler that a subject wears and collects particles in his breathing zone. This sampler can only operate at low flows due to weight constraints of the pump that is worn by the subject. Low flow rates do not always isokinetically sample coarse particles. Thus, in both current exposure assessment methodologies, the accuracy can be questioned. Future research is necessary to correct this problem for it is a very difficult one to solve. The use o f continuous monitors for both exposure assessment and ambient monitoring w ill be the wave o f the future. The change is already taking place, and the result has been an increase in the perception of coarse particle behavior and temporal variability. As these monitors become smaller, human exposure and health effect assessment w ill become simpler. 164 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES Abt, E.; Suh, H.; Catalano, P.; Koutrakis, P. (2000). Relative contribution of outdoor and indoor particle sources to indoor concentrations. Environ. Sci. Technol., 34: 3579-3587. Allen, J.O., Hughes, L.S., Salmon, L.G., Mayo, P.R., Johnson, R.J., Cass, G.R. Characterization and Evolution of Primary and Secondary Aerosols during PM2.5 and PMIO Episodes in the South Coast Air Basin, Report A-22 to the Coordinating Research Council (CRC). 2000. Allen, G.; Sioutas, C.; Koutrakis, P.; Reiss, R.; Lurmann, F.W .; Roberts, P.T. (1997). Evaluation of the TEOM method for the measurement of ambient particulate mass in urban areas. J. Air Waste Manage. Assoc., 47: 682-689. Amdur, M.O. and Chen, L. (1989). Furnace-generated acid aerosols: speciation and pulmonary effects. Environ. Health Perspect., 79: 147-150. American Thoracic Society. (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.; Hackney J.D. (1992). Controlled exposures of volunteers to respirable carbon and sulfuric acid aerosols. J. Air Waste Manage. Assoc., 42: 437-442. Arts, J.H.E.; Spoor, S.M.; Muijser, H.; Kleinman, M .T.; van Bree, L.; Cassee, F.R. (2000) Short-term inhalation exposure of healthy and compromised rats and mice to fine and ultrafine carbon particles. Inhal. Toxicol., 12: 261-266. Babich, P.; Wang, P.Y.; Allen, G.A.; Sioutas, C.; Koutrakis, P. (1999). Development and evaluation of a continuous PM2.5 ambient mass monitor. Aerosol Sci. Technol., 32: 309-325. Becker, S.; Soukup, J.M.; Gilmour, M .I.; Devlin, R.B. (1996). Stimulation o f human and rat alveolar macrophages by urban air particulates: effects on oxidant radical generation and cytokine production. Toxicol. Appl. Pharmacol., 141: 637-648. Becker, S.; Soukup, J.M.; Sioutas, C.; Cassee, F.R. (2003). Response o f human alveolar macrophages to ultrafine, fine, and coarse urban air pollution particles. Exp. Lung Research, 29: 29-44. Berglund, R.N. and Liu, B.Y.H. (1973). Generation o f monodisperse aerosol standards. Environ. Sci. Technol., 7:147-153. 165 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Birch, M .E. and Cary, R.A. (1996). Elemental carbon-based method for monitoring occupational exposures to particulate diesel exhaust. Aerosol Sci. &Technol., 25(3): 221-241. Bolarin, D .M .; Bhalla, D.K.; Kleinman, M .T. (1997) Effects of repeated exposures of geriatric rats to ozone and particle containing atmospheres: an analysis of bronchoalveolar lavage and plasma proteins. Inhalation. Toxicology, 9: 423- 434. Carlton, A.G.; Turpin, B.J.; Johnson, W.; Buckley, B.T.; Simcik, M .; Eisenreich, S.J; Porcja, RJ. (1999). Microanalysis methods for characterization of personal aerosol exposures. Aerosol. Sci. Technol., 31: 66-80. Chow, J. (1995). Measurement methods to determine compliance with ambient air quality standards for suspended particles. J. Air & Waste Manage. Assoc., 45: 320-382. Chow, J.C.; Watson, J.G.; Ashbaugh, L.L.; Magliano, K.L. (2003). Similarities and differences in PMIO chemical source profiles for geological dust from the San Joaquin Valley, California. Atmos. Environ., 37(9-10): 1317-1340. Christoforou, C.S., Salmon, L.G., Hannigan, M.P., Solomon, P.A., Cass, G.R. (2000). Trends in fine particle concentration and chemical composition in Southern California. J. Air & Waste Manage. Assoc., 50: 43-53. Clarke, R.W.; Catalano, P.; Gazula, G.; Sioutas, C.; Ferguson, S.T.; Koutrakis, P.; Godleski, JJ. (1999). Inhalation of concentrated ambient particles (CAPS) induced pulmonary alterations in normal and chronic bronchitic rats. Inhal. Toxicol., 11: 101-120. Clayton, C.; Perritt, R.; Pellizzari, E.; Thomas, K.; Whitmore, R.; Wallace, L.; Ozkaynak, H.; Spengler, J. (1993). Particle total exposure assessment methodology (PTEAM ) study: distributions o f aerosol and elemental concentrations in personal, indoor, and outdoor air samples in a southern California community. J. Exposure Anal. Environ. Epidem., 3: 227-250. Colome, S.; Kado, N.; Jaques, P.; Kleinman, M. (1992). Indoor-outdoor air pollution relations: particulate matter less than 10 pm in aerodynamic diameter (PM10) in homes o f asthmatics. Atmos. Environ., 26A: 2173-2178. Dockery, D.W .; Pope, C.A.; Xu, X.; Spengler, J.D.; Ware, J.H.; Fay, M.E.; Ferris, B.G.; Speizer, F.E. (1993). An association between air pollution and mortality in six U.S. cities. New Eng. Journ. Med., 329(24): 1753-1759. 166 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Dockery, D.W .; Speizer, F.E.; Strain, D.O.; Ware, J.H.; Spengler, J.D.; Ferris, B.J. (1989). Effects of inhalable particles on respiratory health of children. Am. Rev. Respir. Dis., 139: 587-594. Fed. Regist., July 18, 1997; 62 (138) 40 CFR, Part 50. Fine, P.M., Jaques, P.A., Hering, S.V., Sioutas, C. (2003). Performance evaluation and use o f a continuous monitor for measuring size-fractionated PM2.5 nitrate. Aerosol Sci. Technol., 36(4): 342-354. Fine, P. M .; Shen, S.; Sioutas, C. Inferring the sources of fine and ultrafine particulate matter at downwind receptor sites in the Los Angeles Basin using multiple continuous measurements; Accepted by Aerosol Sci. & Technol. February, 2003. Friedlander, S.K. Smoke, Dust, and Haze: Fundamentals o f Aerosol Dynamics. New York: Oxford University Press, 2000. Fung, K., (1990). Particulate carbon speciation by M NO2 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.; 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 H ill ambient air. Am. J. Respir. Crit. Care Med., 159: A318. Godleski, J.; Sioutas, C.; Katler, M .; Koutrakis, P. (1996). Death from inhalation of concentrated ambient air particles in animal models of pulmonary disease. Proceedings of the 2n d Colloquium on Particulate A ir Pollution and Health. Park City, 4/136-4/143. Gordon T.; Gerber, H.; Fang, C.P.; Chen, L.C. (1999). A centrifugal particle concentrator for use in inhalation toxicology. Inhal. Toxicol., 11: 101-117. Gotoh, K. and Masuda, H. (2000). Improvement of the classification performance of a rectangular jet virtual impactor. Aerosol Sci. and Technol., 32: 221-232. 167 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Haller, L., Claibom, C., Larson, T., Koenig, J., Norris, G., Edgar, R. (1999). Airborne particulate matter size distributions in an arid urban area. J. Air & Waste Manage. Assoc., 49: 161-168. Harrison, R.M . and Pio, C.A. (1983). Size differentiated composition of inorganic atmospheric aerosols of both marine and polluted continental origin. Atmos. Environ. 17 (9): 1737-1738. Hering S.V. and Friedlander, S.K. (1982). Origins of aerosol sulfur size distributions in the Los Angeles basin. Atmos. Environ., 16(11): 2647-2656. Hildemann, L.M .; Cass, G.R.; Mazurek, M .A.; Simoneit, B.R.T. (1993). Mathematical modeling o f urban organic aerosols: properties measured by high-resolution gas chromatography. Environ. Sci. Technol., 27: 2045-2055. Hildemann, L.M .; Markowski, G.R.; Jones, M.C.; Cass, G.R. (1991). Submicrometer aerosol mass distributions of emissions from boilers, fireplaces, automobiles, diesel trucks, and meat cooking operations. Aerosol. Sci. Technol., 14: 138-152. Hinds, W.C. Aerosol Technology: Properties, Behavior, and Measurement o f Airborne Particles. New York: John Wiley & Sons, Inc., 1999. Hinds, W .C., and Kuo, T.L. (1995). A Low-velocity wind tunnel to evaluate inhalability and sampler performance for large dust particles. Applied Occ. Environ. H yg, 32: 549-556. Homberg C.; Maciuleviciute L.; Seemayer N.H.; Kainka E. (1998a). Induction of sister chromatid exchanges (SCE) in human tracheal epithelial cells by the fractions PM-10 and PM-2.5 of airborne particulates. Toxicology letters, 96,97: 215-220. Homberg C.; Seemayer N.H.; Kainka E. (1998b). Strong genotoxicity on human tracheobronchial epithelial cells (beas 2B) in vitro by the coarse (PM-10) and fine fraction (PM-2.5) o f airborne particulates as an indicator of potential adverse health effects. J. Aerosol Sci., 29 Suppl. 1 : S317-S318. Hughes, L.S.; Allen, J.O.; Bhave, P.; Kleeman, M.J.; Cass, G.R.; Liu, D.Y.; Fergenson, D.P.; Morrical, B.D.; Prather, K.A. (2000). Evolution of atmospheric particles along trajectories crossing the Los Angeles Basin. Environ. Sci. & Technol., 34(159: 3058-3068. 168 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hughes, L.S.; Allen, J.O.; Kleeman, M.J.; Johnson, R.J.; Cass, G.R.; Gross, D.S.; Gard, E.E.; Morrical, B.D.; Fergenson, D.P.; Dienes, T.; Noble, C.A.; Liu, D .Y.; Silva, P.J.; Prather, K.A. (1999). Size and composition distribution of atmospheric particles in Southern California. Environ. Sci. & Technol., 33(20): 3506-3515. John, W.; W all, S.M.; Ondo, J.L.; Winklmayr, W. (1990). Modes in the size distributions of atmospheric inorganic aerosol. Atmos. Environ., 24(9): 2349- 2359. Jones, N.C.; Thorton, C.A.; Mark, D.; Harrison, R.M. (2000). Indoor/outdoor relationships of particulate matter in domestic homes with roadside, urban and rural locations. Atmos. Environ., 34: 2603-2612. Kegler, S.R., Wilson, W .E., Marcus, A.H. (2001). PM i, intermodal (PM2.5-1 ) mass, and the soil component of PM2.5 in Phoenix, AZ, 1995-1996. Aerosol Sci. & Technol., 35: 914-920. Kelly, B.P. and Koutrakis, P. (1998). Equilibrium size o f atmospheric aerosol sulfates as a function of particle acidity and ambient relative humidity. J. Geophys. Res.-Atmos., D4: 7141-7147. Kennedy, N.J.; Tatyan, K.; Hinds W.C. (2001). Comparison of a simplified and full- size mannequin for the evaluation of inhalable sampler performance. Aerosol • Sci. Technol., 35: 564-568. Kim, S.; Chang, M.C.; Sioutas, C. (2000). A new generation o f portable coarse, fine and ultrafine particle concentrators for use in inhalation toxicology. Inhal. Toxicol., 12 (supplement 1): 121-137. Kim S.; Shen S.; Sioutas C.; Zhu, Y.; Hinds, W.C. (2002). Size distribution and diurnal and seasonal trends of ultrafine particles in source and receptor sites o f the Los Angeles basin. J. Air Waste Manage., 52(3): 297-307. Kleeman, M.J.; Hughes, L.S.; Allen, J.O.; Cass, G.R. (1999). Source contribution to the size and composition distribution of atmospheric particles: Southern California in September 1996. Environ. Sci. & Technol., 33(23): 4331-4341. Kleinman, M .T.; Bhalla, D.K.; Mautz, W.J; Phalen, R.F. (1995). Cellular and immunologic injury with PM-10 inhalation. Inhal. Toxicol., 7: 589-602. 169 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kleinman, M .T.; Sioutas, C.; Chang, M.C.; Cassee, F.R. (2003). Ambient fine and coarse particle suppression of alveolar macrophage functions. Tox. Lett., 137(3): 151-158. Koenig, J.; Larson, T.V.; Hanley, Q.S.; Rebolledo, V.; Dumler, K.; Checkoway, H.; Wang, S.Z.; Lin, D.; Pierson, W.E. (1993). Pulmonary function changes in children associated with fine particulate matter. Environ. Res., 63: 26-38. Koutrakis, P., Wolfson, J.M., Spengler, J.D., Stem, B., Franklin, C.A. (1989). Equilibrium size of atmospheric aerosol sulfates as a function o f the relative humidity. J. Geophys. Research, 94 (D5): 6442-6448. Koutrakis, P., Sioutas, C., Ferguson, S.T., Wolfson, J.M. (1993). Development and evaluation o f a glass honeycomb denuder/filter pack system to collect atmospheric gases and particles. Environ. Sci. & Technol., 27: 2497-2501. Laden, F.; Neas, L.M .; Schwartz, J. (1999). The association o f crustal particles in the fine particulate fraction with daily mortality in six U.S. cities. American Journal o f Respiratory and Critical Care Medicine, 159: A332. Li, N.; Kim, S.; Wang, M .; Froines, J.; Sioutas, C.; Nel, A. (2002). Use of a stratified oxidative stress model to study the biological effects o f ambient concentrated and diesel exhaust particulate matter. Inhal. Tox., 14(5): 459- 486. Li, N.; Sioutas, C.; Cho, A.; Schmitz, D.; Misra, C.; Sempf, J.; Wang, M .; Oberley, T.; Froines, J.; Nel, A. (in press). Ultrafine particulate pollutants induce oxidative stress and mitochondrial damage. Envir. Health Perspect. Lipfert, F.; Wyzga, R. (1995). Uncertainties in identifying “responsible” pollutants in observational epidemiology studies. J. Air & Waste Manage. Assoc., 47: 517- 523. Liu D .Y.; Prather, K.A; Hering S.V. (2000). Variations in the size and chemical composition of nitrate-containing particles in Riverside, CA. Aerosol Sci. and Technol., 33: 71-86. Malm, W .C., Sisler, J.F., Huffman, D., Eldred, R.A, Cahill, T.A. (1994). Spacial and seasonal trends in particle concentration and optical extinction in the United States./. Geophys. Res., 99: 1347-1370. Mar T.; Norris G.; Koenig J.; Larson T. (1999). Associations between air pollution and mortality in Phoenix. Environ.. Health Perspect., 108: 347 -353. 170 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 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. In Fine Particles: Aerosol Generation, Measurement, Sampling, and Analysis (Edited by B.Y.H. Liu); Academic press: New York, 1976. Masuda H. and Nagasita, S. (1988). Classification performance of a rectangular jet virtual impactor - Effect of nozzle width ratio of collection nozzle to acceleration jet. J. Aerosol Sci., 19: 243-252 McMurry, P.H. and Zhang, X.Q. (1989). Size distributions of ambient organic and elemental carbon. Aerosol. Sci. Technol., 10: 430-437. McMurry, P.H.; Zhang X.; Lee, Q.T. (1996). Issues in aerosol measurement for optical assessments. Journal o f Geophysical Research., 101 (19): 188-197. Meng Z.Y . and Seinfeld, J.H. (1994). On the source o f the submicrometer droplet mode o f urban and regional aerosols. Aerosol Sci. Tech., 20(3): 253-265. Misra, C., Geller, M .D ., Shah, P., Solomon, P. and Sioutas, C. (2001). Development ■ and evaluation of a continuous coarse (PMio - PM2.5) particle monitor. J. Air & Waste Manage. Assoc., 51: 1309-1317 Monn, C. and Becker, S. (1998) Fine and coarse particles: induction of cytokines in human monocytes. J. Aerosol. Sci., 29: 305-306. Monn, C. and Becker, S. (1999). Cytotoxicity and pro-inflammatory cytokines from human monocytes exposed to fine (PM2.5) and coarse particles (PM2.5-10) in indoor and outdoor air. Toxicol. Apply Pharmacol., 155: 24. Moschandreas, D.J.; Winchester, J.W.; Nelson, J.W.; Burton, R.M. (1979). Fine particle residential indoor air pollution. Atmos. Environ., 13: 1413-1418. Mylonas, D.T.; Allen, D.T.; Ehrman, S.H.; Pratsinis, S.E. (1991). The sources and size distributions of organonitrates in the Los Angeles aerosol. Atmos. Environ., 25A: 2855-2861. Oberdorster, G. (2000). Toxicology of ultrafine particles: in vivo studies. Phil. , Trans. Roy. Soc. London A, 358 (1775): 2719-2739. 171 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Oberdorster, G. (2001). Pulmonary effects of inhaled ultrafine particles. Int. Arch. Occup. Envir. Health, 74(1): 1-8. Ostro, B.D. (1993). The association of air pollution and mortality: examining the case for interference o f organonitrates in the Los Angeles aerosol. Atmos. Environ., 25 A: 2855-2861. Ostro, B.D.; Hurley, S.; Lipsett, M.J. (1999). A ir pollution and daily mortality in the Coachella Valley, California: A study o f PMIO dominated by coarse particles. Environ. Res., 81: 231-238. Pandis, S.N., Harley, R.A., Cass, G.R., Seinfeld, J.H. (1992). Secondary organic aerosol formation and transport. Atmos. Environ., 26A: 2269-2282. Pandis, S.N.; Wexler, A.S.; Seinfeld, J.H. (1993). Secondary organic aerosol formation and transport 2: Predicting the ambient secondary organic aerosol size distribution. Atmos. Environ., 27(15): 2403-2416. Pickle, T.; Allen, D.T.; Pratsinis, S.E. (1990). The sources and size distributions of aliphatic and carbonyl carbon in Los Angeles aerosol. Atmos. Environ., 24: 2221-2228. Pope, C.A., III; Bates, D.V.; Raizenne, M.E. (1995). Health effects of particulate air pollution: time for reassessment. Environ. Health Perspect., 103: 472-480. Pope, C.A.; Dockery, D.W .; Schwartz, J. (1995). Review o f epidemiological evidence of health effects o f particulate pollution. Inhal. Toxicol., 7: 1-18. Pope, C.A., IE; Dockery, D.W .; Spengler, J.D.; Raizenne, M .E. (1991). Respiratory health and PMio pollution. A daily time series analysis. Am. Rev. Respir. Dis., 144: 668-674. QUARG, (1993). Urban A ir Quality in the United Kingdom. Quality of Urban Air Review Group, Department of Environment, London. Schwartz J. and Neas, L.M . (2000). Fine particles are more strongly associated than coarse particles with acute respiratory health effects in schoolchildren. Epidemiology, 11(1): 6-10. Seinfeld, J. and Pandis, S. Atmospheric chemistry and physics. New York: John W iley & Sons, Inc., 1998. 172 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sioutas, C.; Abt., E.; Wolfson, J.M.; Koutrakis, P. (1999). Effect o f particle size on mass concentration measurement by the Scanning Mobility Particle Sizer and the Aerodynamic Particle Sizer. J. Aerosol Sci. Technol., 30: 84-92. Sioutas, C.; Chang, M .C.; Kim, S.; Ferguson, S.T.; Koutrakis, P. (1998). Design and experimental characterization o f a PMi and a PM2.5 personal sampler. J. Aerosol Sci., 30: 693-707. Sioutas, C.; Kim , S.; Chang, M . (1999b). Development and evaluation of a prototype ultrafine particle concentrator. J. Aerosol Sci., 30(8): 1001-1012. Sioutas, C.; Kim, S.; Chang, M .; Terrell, L.L.; Gong, H. (2000). Field evaluation of a modified DataRAM M IE scattering monitor for real-time PM2.5 mass concentration measurements. Atmos. Environ., 34: 4829-4838. Sioutas, C.; Koutrakis, P.; Burton, R.M. (1995a). A technique to expose animals to concentrated fine ambient aerosols, Environ. Health Perspect., 103: 172-177. Sioutas, C.; Koutrakis, P.; Ferguson, S.T.; Burton, R.M. (1995b). 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.; Burton, R.M. (1997). Harvard/EPA ambient fine particle concentrators for human and animal exposures. J Aerosol Sci., 28(6): 1057-1071. Singh, M ., Jaques, P. and Sioutas, C. (2002). Particle-bound metals in source and receptor sites o f the Los Angeles Basin. Atmos. Environ., 36(10): 1675-1685. Sloane, C.S. (1984). Optical properties of aerosols o f mixed composition. Atmos. Environ., 18: 871-878. Solomon, P.A.; Larson, S.M.; Fall, T.; Cass, G.R. (1988). Basin wide nitric acid and related species concentrations observed during the Claremont Nitrogen Species Comparisons Study. Atmos. Environ., 22: 1587-1594. South Coast A ir Quality Management District, (1990). Final state implementation plan for PMio in the Coachella Valley. El Monte, CA. Stolzenburg M .R ., Dutcher D.D, Kirby B.W., Hering S.V. (2003). Automated measurement of the size and concentration o f airborne particulate nitrate. Aerosol Sci. T ech n o l37: 537 - 546. 173 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ten Brink, H.M .; Kruisz, C.; Kos, P.A; Bemer, A. (1997). Composition/size of the light-scattering aerosol in the Netherlands. Atmos. Environ., 31: 3955-3962. Tsai, F.C.; Apte, M.G.; Daisey, J.M. (2000). An exploratory analysis of the relationship between mortality and the chemical composition of airborne particulate matter. Inhal. Toxicol., 12(Supplement 2): 131-135. Urch, B.; Liu, L.; Brook, J.; Purdham, J.; Tarlo, S.; Broder, I.; Lukic, Z.; Datema, J.; Koutrakis, P.; Sioutas, C.; Ferguson, S.; Dales, R.; Silverman, F. (1999). Pulmonary function responses after inhalation o f controlled levels of concentrated urban particles in healthy individuals. Am. J. Respir. Crit. Care Med., 159: A318. U.S. EPA, (1996). Air Quality Criteria for Particulate Matter, April (EPA/600/P- 95/00lcF). Office of Research and Development, Washington, DC. U.S. EPA (1996) Federal Register. 40 CFR Part 50, December 13: 65685-65701. Vedal, S., (1997). Ambient particles and health: lines that divide. J. Air Waste Manage. Assoc., 47: 551-581. Wallace, L.; Quakenboss, J.; Rhodes, C. (1997). In AW M A/EPA Symposium on the Measurement o f Toxic and Related A ir Pollutants. Research Triangle Park, N.C., 860-871. Wark, K.; Warner, C.F.; Davis, W .T. Air Pollution: Its Origin and Control. Menlo Park, CA: Addison-Wesley, 1998. White, W .H. (1998). Statistical considerations in the interpretation o f size-resolved particulate mass data. J. Air & Waste Manage. Assoc., 48: 454-458. Wiener, R.; Bachmann, J.D. Coarse Particle Monitoring. Presented at the US Environmental Protection Agency, Science Advisory Board, Clean A ir Scientific Advisory Committee, Technical Subcommittee on Fine Particle Monitoring Meeting, April 18-19, 2000. Wilson, W ., and Suh, H.H. (1997). Fine particles and coarse particles: Concentration relationships relevant to epidemiological studies. J. Air & Waste Manage. Assoc., 47: 1238-1249. Yeh, H.C.; Cuddihy, R.G.; Phalen, R.F.; Chang, I-Y . (1996). Comparisons of calculated respiratory tract deposition of particles based on the proposed NCRP model and the new ICRP model. Aerosol Sci: Tech., 25: 134-140. 174 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Yeh, H.C. and Schum, G.M . (1980). Models for human lung airways and their application to inhaled particle deposition. Bull. Math. Biology, 42: 461-480. 175 Reproduced wi,d p e n s io n o f * , c o p * * , owner. Fodder reproduced prod,Med wltdoo. permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

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

PDF

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

PDF

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

PDF

Health monitoring of structures under ambient vibrations using semiactive devices

PDF

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

PDF

Biofiltration of petroleum hydrocarbon vapors and co-treatment of volatile organic compounds in low-pH sulfide biofilters

PDF

A percolation biofilm-growth model for biomass clogging in biofilters

PDF

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

PDF

Assessment of virus coagulation and filterability using particle density distribution analysis

PDF

Experimental studies of biopolymers behavior in subsurface porous strata

PDF

Temporal, spatial and toxicological characteristics of coarse particulate matter in an urban area and relation to sources and regulations

PDF

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

PDF

Technologies for assessment of physical, chemical and toxicological characteristics of ambient ultrafine particles

PDF

Human -like movement in virtual environments

PDF

An integrated environment for modeling, experimental databases and data mining in neuroscience

PDF

Automated segmentation and analysis of human cerebral cortex imagery

PDF

Investigation of physico-chemical characteristics of particulate matter from vehicular sources

PDF

An investigation of compaction grouting for liquefaction mitigation

PDF

Investigation of energy flow in earthquake response of structures

Asset Metadata

Creator Geller, Michael David (author)

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

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, civil,engineering, environmental,environmental sciences,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Sioutas, Constantinos (committee chair), Fine, Philip M. (committee member), Gong, Henry (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-635457

Unique identifier UC11334902

Identifier 3116702.pdf (filename),usctheses-c16-635457 (legacy record id)

Legacy Identifier 3116702.pdf

Dmrecord 635457

Document Type Dissertation

Rights Geller, Michael David

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