Elsevier

Atmospheric Environment

Volume 43, Issue 12, April 2009, Pages 1972-1981
Atmospheric Environment

Source identification of personal exposure to fine particulate matter using organic tracers

https://doi.org/10.1016/j.atmosenv.2009.01.023Get rights and content

Abstract

Personal exposure to fine particulate matter (PM2.5) is due to both indoor and outdoor sources. Contributions of sources to personal exposure can be quite different from those observed at ambient sampling locations. The primary goal of this study was to investigate the effectiveness of using trace organic speciation data to help identify sources influencing PM2.5 exposure concentrations. Sixty-four 24-h PM2.5 samples were obtained on seven different subjects in and around Boulder, CO. The exposure samples were analyzed for PM2.5 mass, elemental and organic carbon, organic tracer compounds, water-soluble metals, ammonia, and nitrate. This study is the first to measure a broad distribution of organic tracer compounds in PM2.5 personal samples. PM2.5 mass exposure concentrations averaged 8.4 μg m−3. Organic carbon was the dominant constituent of the PM2.5 mass. Forty-four organic species and 19 water-soluble metals were quantifiable in more than half of the samples. Fifty-four organic species and 16 water-soluble metals had measurement signal-to-noise ratios larger than two after blank subtraction.

The dataset was analyzed by Principal Component Analysis (PCA) to determine the factors that account for the greatest variance. Eight significant factors were identified; each factor was matched to its likely source based primarily on the marker species that loaded the factor. The results were consistent with the expectation that multiple marker species for the same source loaded the same factor. Meat cooking was an important source of variability. The factor that represents meat cooking was highly correlated with organic carbon concentrations (r = 0.84). The correlation between ambient PM2.5 and PM2.5 exposure was relatively weak (r = 0.15). Time participants spent performing various activities was generally not well correlated with PCA factor scores, likely because activity duration does not measure emissions intensity. The PCA results demonstrate that organic tracers can aid in identifying factors that influence personal exposures to PM2.5.

Introduction

Studies have shown a correlation between mortality and both daily and long-term variations in ambient fine particulate matter (PM2.5) concentrations in the atmosphere (Dockery et al., 1993, Pope et al., 2002). However, personal exposure to PM2.5 can differ significantly from ambient PM2.5 with respect to concentration (Williams et al., 2003), composition (Larson et al., 2004), and contributing sources (Zhao et al., 2006). Americans generally spend much of their day indoors; pollutants encountered at home, at work, and in a vehicle are important components of overall exposure. Improved understanding of personal exposure to PM2.5 could help researchers determine the health effects of various PM2.5 sources more precisely and also determine the most efficient control strategies for total PM2.5 exposure.

Personal sampling can provide a more accurate representation of a person's total exposure to PM2.5 than ambient outdoor measurements. The Harvard Air Pollution Health Study was one of the first large studies to investigate the relationship among indoor, outdoor, and personal exposure to particulate matter (Spengler et al., 1985). This cross-sectional study considered respirable suspended particles, which include particles larger than 2.5 microns, and demonstrated a weak correlation (r2 < 0.01) between outdoor ambient concentrations and personal exposure concentrations across study participants. Subsequent cross-sectional studies also found little correlation between ambient concentrations and personal exposure to PM2.5 (Wallace, 2000). In contrast, longitudinal studies that follow individuals over time show higher correlations, depending on population type and location. These studies have demonstrated average personal-ambient r2 values of 0.74 (children in the Netherlands, Janssen et al., 1999), 0.50 (nonsmoking adults in Amsterdam, Janssen et al., 1998), 0.79 (elderly in Baltimore, MD, Williams et al., 2000), and 0.16 (low socioeconomic status adults with hypertension or cardiovascular disease in North Carolina, Williams et al., 2003). Major outdoor sources of PM2.5 include gasoline and diesel powered motor vehicles, natural gas combustion, and secondary nitrate and sulfate formed through reactions in the atmosphere. Indoor sources and personal activities, such as smoking and meat cooking, contribute significantly to exposure concentrations (Zhao et al., 2006).

Receptor models have been widely used to apportion ambient PM2.5 mass to its sources. Commonly used receptor models include the Chemical Mass Balance (CMB) model (Kim and Henry, 2000) and factor analysis models, such as Positive Matrix Factorization (PMF) (Lee et al., 2003) and Principal Component Analysis (PCA) with Absolute Principal Component Scores (APCS) (Thurston and Spengler, 1985). Factors resulting from factor analysis do not always correspond directly to sources, as meteorological conditions and other factors also contribute variance to the dataset. Receptor models can also be applied to personal exposure data to determine the sources of exposure (Yakovleva et al., 1999, Hopke et al., 2003, Koistinen et al., 2004). However, this has proven difficult for several reasons. Additional exposure sources need to be resolved compared to ambient source apportionment. Within a source category, the individual sources contributing to personal exposure may not have uniform composition; e.g., the composition of PM2.5 from motor vehicles to which one person is exposed might be different from that for another individual. The lower volumetric flow rates used in personal samplers typically result in lower filter mass loadings, so the detection limits of chemical species are degraded. Organic species quantification requires multiple filters to be worn by each participant, increasing the complexity of the sampling setup that participants must wear.

Early personal exposure source apportionment studies (e.g., Yakovleva et al., 1999) used only a Teflon (PTFE) filter to measure particulate matter mass and a limited set of constituents including ions, elements, and light-absorbing carbon. None of the PM2.5 constituents measured in the personal samples of these studies allowed for separation of combustion sources such as meat cooking and motor vehicle exhaust. More recently, Zhao et al. (2006) analyzed mass, elemental and organic carbon (EC/OC), and elements from seven 24-h exposure samples from 38 different individuals in Raleigh and Chapel Hill, NC. The carbon measurements were intended to characterize PM2.5 from combustion sources. However, Zhao et al. (2006) noted that more specific organic speciation would help distinguish among personal exposure sources. Most recently, Sheesley et al. (2008) used organic tracers for different sources originating from motor vehicles to apportion work-related PM2.5 exposure in a freight terminal.

The primary objective of this pilot study was to examine the effectiveness of receptor modeling on an exposure dataset that includes organic tracer species to help resolve additional sources. This approach has previously been evaluated using synthetic data (Brinkman et al., 2006). In this study we collected and analyzed 24-h PM2.5 exposure samples for a non-representative sample of students and faculty at the University of Colorado at Boulder. All subjects worked in the same building (the Engineering Center) and lived within 10 miles of campus. Sixty-four 24-h exposure samples were collected and analyzed for mass, elemental and organic carbon, organic tracer species, water-soluble metals, ammonia, and nitrate. Subjects kept detailed activity logs. PCA and PMF analyses were performed on the exposure concentration data to estimate the sources of PM2.5 exposure. The results were compared to the activity logs to determine if source contributions are correlated with time spent in certain microenvironments (e.g., vehicle or restaurant) or performing certain activities (e.g., cooking or vacuuming).

Section snippets

Sampling

Sixty-four 24-h exposure samples from seven different subjects residing in or near Boulder, CO were collected during March–June of 2006. During each sampling period, the subject chose what time to start and operated the sampling pack for 24 consecutive hours, wearing it during periods of activity and placing it on a nearby table during periods of inactivity, such as sleeping. Each sampling pack consisted of two SKC AirLite pumps attached to two cyclone filter samplers. The samplers were

Exposure concentrations

Summary statistics for selected compounds measured in the study are shown in Table 1. Organic Carbon (OC) was the dominant constituent of the PM2.5 mass. However, it is difficult to assess the exact percentage because there appears to have been a positive OC artifact, as previously noted during indoor sampling by Landis et al. (2001). This artifact was probably caused by semi-volatile compounds adsorbing onto the quartz filter. Mcdow and Huntzicker (1990) demonstrated that OC measurements

Discussion

This study demonstrates the value of including organic tracer species in receptor modeling of personal exposures to PM2.5. Although PM2.5 exposure concentrations in the study were generally low compared to levels observed in prior studies, many organic tracer compounds were significantly above detection and quantification limits in most of the samples. Signal-to-noise ratios were larger than two for most of these compounds. Correlations were seen between species that are typically emitted by

Acknowledgements

Research described in this article was supported in part by Philip Morris USA Inc. and by Philip Morris International. The authors would like to thank Jeff DeMinter for the EC and OC analysis. We acknowledge Steven Dutton at the University of Colorado for helping with the organic analysis and Adam Eisele, Nick Facciola, Brian Cone, and Morgan Meadows at the University of Colorado for their work on the project.

References (39)

  • D.W. Dockery et al.

    An Association between air-pollution and mortality in 6 United-States cities

    New England Journal of Medicine

    (1993)
  • Dutton, S.J., Williams, D.E., Vos, C.A., Garcia, J., Vedal, S., Hannigan, M.P., 2009. PM2.5 characterization for time...
  • M.P. Field et al.

    Magnetic sector ICPMS with desolvating micronebulization: interference-free subpicogram determination of rare earth elements in natural samples

    Analytical Chemistry

    (1998)
  • P.M. Fine et al.

    Chemical characterization of fine particle emissions from the fireplace combustion of wood types grown in the Midwestern and Western United States

    Environmental Engineering Science

    (2004)
  • J.D. Herner et al.

    Measuring the trace elemental composition of size-resolved airborne particles

    Environmental Science & Technology

    (2006)
  • L.M. Hildemann et al.

    Chemical-composition of emissions from urban sources of fine organic aerosol

    Environmental Science & Technology

    (1991)
  • N.A.H. Janssen et al.

    Personal sampling of particles in adults: relation among personal, indoor, and outdoor air concentrations

    American Journal of Epidemiology

    (1998)
  • N.A.H. Janssen et al.

    Personal exposure to fine particles in children correlates closely with ambient fine particles

    Archives of Environmental Health

    (1999)
  • I.T. Jolliffe

    Principal Component Analysis

    (2002)
  • Cited by (0)

    View full text