Environ. Sci. Technol. 2005, 39, 932-944
Size-Fractionated Measurements of Ambient Ultrafine Particle Chemical Composition in Los Angeles Using the NanoMOUDI SATYA B. SARDAR, PHILIP M. FINE, PAUL R. MAYO, AND CONSTANTINOS SIOUTAS* Department of Civil and Environmental Engineering, University of Southern California, 3620 South Vermont Avenue, Los Angeles, California 90089
Ambient ultrafine particles have gained attention with recent evidence showing them to be more toxic than larger ambient particles. Few studies have investigated the distribution of chemical constituents within the ultrafine range. The current study explores the size-fractionated ultrafine (10-180 nm) chemical composition at urban source sites (USC and Long Beach) and inland receptor sites (Riverside and Upland) in the Los Angeles basin over three different seasons. Size-fractionated ultrafine particles were collected by a NanoMOUDI over a period of 2 weeks at each site. Measurements of ultrafine mass concentrations varied from 0.86 to 3.5 µg/m3 with the highest concentrations observed in the fall. The chemical composition of ultrafine particles ranged from 32 to 69% for organic carbon (OC), 1-34% for elemental carbon (EC), 0-24% for sulfate, and 0-4% for nitrate. A distinct OC mode was observed between 18 and 56 nm in the summer, possibly indicating photochemical secondary organic aerosol formation. The EC levels are higher in winter at the source sites due to lower inversion heights and are higher in summer at the receptor sites due to increased long-range transport from upwind source areas. Nitrate and sulfate were measurable only in the larger particle size ranges of ultrafine PM. Collocated continuous measurements of particle size distributions and gaseous pollutants helped to differentiate ultrafine particle sources at each site.
Introduction A number of studies have found positive correlations between increased ambient particulate matter (PM) concentrations and respiratory related mortality and morbidity (1, 2). Atmospheric ultrafine particles (PM with physical diameters less than about 100 nm) have received recent attention as emerging literature continues to show that these smaller particles are comparatively more toxic than larger particles with similar chemical composition and mass (3-7). Although ultrafine particles, because of their small size, contribute relatively little to the total PM mass, they account for the majority of airborne particle number concentrations (3, 8). A study by Penttinen et al. (9) demonstrated a negative association between daily mean number concentration and peak expiratory flow (PEF). However, whether the observed adverse PM health effects are related to particle number, * Corresponding author e-mail:
[email protected]. 932
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005
particle surface area, particle mass, and/or particle chemical composition remains an open question. In any case, ultrafine particles with their high ambient number and surface area concentrations are extremely important in both atmospheric chemistry and environmental health. Vehicular emissions are the primary sources of ultrafine particles in urban atmospheres (10, 11). Recent studies have shown a dramatic decrease of ultrafine number concentrations with increasing distance from busy freeways in Los Angeles, thereby demonstrating that vehicular pollution is a major source of ultrafine particles and that high particle numbers can be a localized phenomenon (on scales of 100500 m) (12, 13). McMurry and Woo (14) measured the size distribution of ambient aerosol in urban Atlanta, GA from August 1998 through August 2000. The study found that for particles between 10 and 100 nm, average concentrations tended to be highest during winter, during rush hour, and on weekdays. On the other hand, concentrations of particles in the 3-10 nm range increased in the summer due to photochemical nucleation. Kim et al. (15) have studied the effects of different sources and formation mechanisms on the size distributions and temporal trends of ultrafine particles at a source and a receptor site in the Los Angeles basin. The results showed that source sites are primarily affected by fresh emissions from nearby freeways, and the inland receptor sites are influenced by photochemical secondary particle formation. Fine et al. (16) have measured the ambient particle characteristics in eastern Los Angeles Basin locations affected by both local emissions as well as advection from the more intense upwind particle sources closer to downtown Los Angeles. Diurnal patterns in this study showed a peak in particle number concentrations due to morning traffic. An increase in particle number was also observed in afternoon periods in the warmer months, possibly resulting from photochemical secondary formation. The study also showed that long-range transport of primary emissions also influences the ultrafine particle concentrations and size distributions at inland sites, resulting in unusually large aerosol number median diameters (i.e., 90 to 120 nm) for these locations during the warmer periods of the year. Few studies have examined how the chemical constituents of ultrafine particles are distributed within the ultrafine size range. Such information will prove important in light of the increased toxicological and epidemiological evidence linking respiratory health effects and exposure to ultrafine particles. A study by Cass et al. (17) reported the 24-h averaged chemical composition of ambient ultrafine particles at several sites in the Los Angeles Basin but only for particles greater than 56 nm. Hughes et al. (18) sampled ultrafine particles at a single site in Los Angeles over the course of 1 winter month, using a MOUDI impactor. Geller et al. (19) examined diurnal variations of size-fractionated ultrafine particle chemistry in three additional ultrafine size bins, using a NanoMOUDI (Model 115, MSP Corp., Minneapolis, MN) preceded by a particle concentrator at a source and a receptor site in the Los Angeles Basin. The objective of the current study is to give a more complete representation of size-fractionated ultrafine chemical characteristics by sampling across four different locations and three seasons in the Los Angeles Basin. Sampling was conducted with a NanoMOUDI for a period of 2 weeks at each site in order to collect sufficient material for mass and chemical analysis. A collocated Scanning Mobility Particle Sizer (SMPS 3936, TSI Inc., Shoreview, MN) was deployed at several locations to measure semicontinuous particle size distributions, and data from collocated gaseous pollutant 10.1021/es049478j CCC: $30.25
2005 American Chemical Society Published on Web 12/31/2004
TABLE 1. Seasonal Meterological Data by Site during the Sampling Periods Riverside Tmax,avg (°C) Tmin,avg (°C) RHmax,avg (%) RHmin,avg (%)
Upland Tmax,avg (°C) Tmin,avg (°C) RHmax,avg (%) RHmin,avg (%)
FIGURE 1. Locations of source sites (Long Beach and USC) and receptor sites (Upland and Riverside) in the Los Angeles Basin. monitors were examined to help differentiate the ultrafine particle sources. The SMPS also complimented the NanoMOUDI by providing an estimate of the diurnal variability of the ultrafine PM size distributions for each location and season, given that the NanoMOUDI measurements represent time-integrated, 2-week averages.
Methods Fourteen-day composite size-fractionated PM2.5 samples (particle diameters less than 2.5 µm), including several ultrafine size ranges below 180 nm in particle, were collected at four sites and three seasons in the Los Angeles Basin (Figure 1). The seasons were defined as the following: fall (September-November 2002); winter (December 2002-March 2003); and summer (June-August 2003). The first site was at the University of Southern California near downtown Los Angeles. This sampling location is about 150 m downwind of a major freeway and represents an urban mix of industrial, vehicular, and construction sources. The Long Beach station is located on one of the busiest streets in Long Beach, about 1 km northeast of a major freeway and is downwind of several industrial facilities such as power plants and oil refineries. The Upland site is located in a residential area at least 3 km away from a major freeway. The sampling location at Riverside is at the Citrus Research Center and Agricultural Experiment Station (CRS-AES), a part of the University of California-Riverside. Riverside is 90 km east of downtown Los Angeles and is primarily a residential and commercial center. The site is also about 25 km downwind of the Chino area dairy farms, a strong ammonia source leading to high concentrations of ammonium nitrate (20). The Riverside site is upwind of surrounding freeways and major roads. The four sites can be characterized based on previously studied meteorological and air pollution patterns in the Los Angeles Basin (20, 27). Because of the generally upwind locations and the surrounding urban environment, Long Beach and USC are considered “source” sites where fresh particles are emitted primarily from vehicular and industrial source, thus representing a well-mixed urban air mass. The other sites, Riverside and Upland, are designated “receptor” sites, with comparatively less vehicular influences and where local emissions mix with advected, aged, and photochemically transformed air masses from the central Los Angeles area (21). The transport time of air masses from source to receptor sites can vary from a few hours to more than a day. Although these distinctions are general and are not always predictive of air pollution conditions at these sites, they are widely accepted as descriptive of Los Angeles Basin sampling sites (15, 19). A NanoMOUDI/MOUDI (Nano-Micro orifice uniform deposit impactor, Model 115, MSP Corporation, Minneapolis,
USC Tmax,avg (°C) Tmin,avg (°C) RHmax,avg (%) RHmin,avg (%)
Long Beach Tmax,avg (°C) Tmin,avg (°C) RHmax,avg (%) RHmin,avg (%)
fall winter summer (Oct 30-Nov 13) (Jan 27-Feb 13) (Jul 14-Jul 28) 28.4 11.9 90.1 34.6
23.0 7.8 70.8 21.4
33.6 16.8 87.8 34.9
fall winter summer (Nov 13-Nov 27) (Feb 13-Feb 27) (Jul 28-Aug 11) 26.7 13.4 70.0 28.3
21.0 8.6 92.1 57.9
32.0 17.5 79.8 29.0
fall winter summer (Oct 2-Oct 16) (Feb 27-Mar 13) (June 30-Jul 14) 24.7 14.1 88.8 47.3
21.8 10.9 90.7 44.6
27.4 17.9 95.0 53.8
fall winter summer (Oct 16-Oct 30) (Mar 13-Mar27) (Aug 11-Aug 25) 24.4 13.8 91.0 61.8
21.0 10.0 95.3 52.5
26.9 18.0 91.2 41.6
MN) impactor collected samples continuously for a period of 2 weeks at each site. The NanoMOUDI was placed in a box at ambient temperature during sampling. Further, the sampling line was straight, made of stainless steel and about 50 cm long. The NanoMOUDI is a cascade impactor operating under low pressure connected to a regular MOUDI (Micro Orifice Uniform Deposit Impactor, Model 110, MSP Corporation, Minneapolis, MN) after the 56 nm cut-point stage. It includes three additional stages with cut points of 32 nm, 18 nm, and 10 nm (22). An Electrical Low-Pressure Impactor (ELPI) pump (Model 393501, TSI, Inc., St. Paul, MN) was used to draw 10 L min-1 of air through the system. Since the NanoMOUDI has a design flow rate of 10 L min-1, and the MOUDI is designed for a flow rate of 30 L min-1, the acceleration nozzle plates of the regular MOUDI were masked to block two-thirds of the nozzles such that a flow rate of 10 L min-1 through the MOUDI does not change the designated size-cuts. An examination of the literature reveals varying definitions of the ultrafine particle size range, alternately taken as particles smaller than 100 nm (18), 150 nm (23), and 180 nm (19). With respect to the NanoMOUDI data, the cutpoints of which are based on aerodynamic and not mobility diameters, we selected 180 nm as the upper cutpoint for the definition of ultrafine particles. Assuming an ambient aerosol density of 1.6 g/cm3, this size would correspond to a spherical particle with a mobility diameter of approximately 130 nm (24). Furthermore, since ultrafine particles are characterized by high particle number concentrations, an upper size cut of 180 nm ensures that the majority of the particle number distribution is included (15, 16). Table 1 displays the average daily maximum and minimum of the temperature and relative humidity (RH) for each site and season when sampling was performed. The consistent temperature data among sites within each season suggest that the sampling periods were representative of the prevailing seasonal conditions. No precipitation occurred VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
933
TABLE 2. Ultrafine (PM0.18) Mass and Chemical Species Concentrations by Sampling Site fall
mass (ng/m3)
sulfate (ng/m3)
nitrate (ng/m3)
EC (ng/m3)
OC (ng/m3)
USC Long Beach Riverside Upland
3490 ( 130 3100 ( 120 2890 ( 110 2870 ( 70
204 ( 8 154 ( 7 61 ( 3 47 ( 4
20 ( 3 35 ( 3 70 ( 4 27 ( 3
1190 ( 160 181 ( 15 704 ( 39 153 ( 13
1980 ( 170 2150 ( 110 1450 ( 70 1880 ( 92
winter
mass (ng/m3)
sulfate (ng/m3)
nitrate (ng/m3)
EC (ng/m3)
OC (ng/m3)
USC Long Beach Riverside Upland
1490 ( 60 1240 ( 50 1310 ( 90 1450 ( 60
