Selective detection and characterization of nanoparticles from motor vehicles
Murray V Johnston,Joseph P Klems,Christopher A Zordan,M Ross Pennington,James N Smith,HEI Health Review Committee
Abstract:Numerous studies have shown that exposure to motor vehicle emissions increases the probability of heart attacks, asthma attacks, and hospital visits among at-risk individuals. However, while many studies have focused on measurements of ambient nanoparticles near highways, they have not focused on specific road-level domains, such as intersections near population centers. At these locations, very intense spikes in particle number concentration have been observed. These spikes have been linked to motor vehicle activity and have the potential to increase exposure dramatically. Characterizing both the contribution and composition of these spikes is critical in developing exposure models and abatement strategies. To determine the contribution of the particle spikes to the ambient number concentration, we implemented wavelet-based algorithms to isolate the particle spikes from measurements taken during the summer and winter of 2009 in Wilmington, Delaware, adjacent to a roadway intersection that approximately 28,000 vehicles pass through daily. These measurements included both number concentration and size distributions recorded once every second by a condensation particle counter (CPC*; TSI, Inc., St. Paul, MN) and a fast mobility particle sizer (FMPS). The high-frequency portion of the signal, consisting of a series of abrupt spikes in number concentration that varied in length from a few seconds to tens of seconds, accounted for 3% to 35% of the daily ambient number concentration, with spike contributions sometimes greater than 50% of hourly number concentrations. When the data were weighted by particle volume, this portion of the signal contributed an average of 10% to 20% to the daily concentration of particulate matter (PM) < or = 0.1 microm in aerodynamic diameter (PM0.1). The preferred locations for observing particle concentration spikes were those surrounding the measurement site at which motor vehicles accelerated after a red traffic light turned green. As the distance or transit time from emission to sampling increased, the size distribution shifted to a larger particle size, which confirmed the source assignments. To determine the distribution of emissions from individual vehicles, we correlated camera images with the spike contribution to particle number concentration at each time point. A small percentage of motor vehicles were found to emit a disproportionally large concentration of nanoparticles, and these high emitters included both spark-ignition (SI) and heavy-duty diesel (HDD) vehicles. In addition to characterizing the contribution of the spikes (local sources) to the ambient number concentration, we developed a method to determine the net contribution of motor vehicles (all sources) to the total mass concentration of ambient nanoparticles. To do this, we correlated the concentration of spikes with measurements of fast changes in the chemical composition of nanoparticles measured with the nano aerosol mass spectrometer (NAMS; built by the Johnston group). The NAMS irradiates individual, size-selected nanoparticles with a high-energy laser pulse to generate a mass spectrum consisting of multiply charged atomic ions. The elemental composition of each particle was determined from the ion signal intensities of each element. However, overlapping mass-to-charge ratios (m/z) at 4 m/z (O(+4) and C(+3)) and at 8 m/z (O(+2) and S(+4)) needed to be separated into their component ions to obtain a representative composition. To do this, we developed a method to deconvolute these ion signals using sucrose and ammonium sulfate [(NH4)2SO4] as calibration standards. With this approach, the differences between the expected and measured elemental mole fractions of carbon (C), oxygen (O), nitrogen (N), and sulfur (S) for a variety of test particles were generally much less than 10%. Ambient nanoparticles were found to consist mostly of C, O, N, and S. Many particles also contained silicon (Si). The elemental compositions were apportioned into molecular species that are commonly found in ambient aerosol: sulfate (SO4(2-)), nitrate (NO3-), ammonium (NH4+), carbonaceous matter, and when present, silicon dioxide (SiO2). Correlating NAMS chemical-composition measurements with spike contributions allowed for the development of a chemical profile representing motor vehicle emissions, which could be used to apportion their total contribution to the ambient nanoparticle mass. Particles originating from motor vehicles had compositions dominated by unoxidized carbonaceous matter, whereas non-motor vehicle particles consisted mostly of SO42-, NO3-, and oxidized carbonaceous matter. Motor vehicles were found to contribute up to 48% and 60% of the nanoparticle mass and number concentrations, respectively, in the winter measurement period, but only 16% and 49% of the nanoparticle mass and number concentrations, respectively, in the summer period. Chemical-composition profiles and contributions of SI versus HDD vehicles to the nanoparticle mass concentration were estimated by correlating still camera images, chemical composition, and spike contributions at each time point. The total mass contributions from SI and HDD vehicles were roughly equal, but the uncertainty in the split was large. The results of this study suggest that nanoparticle concentrations will be higher adjacent to an intersection than along the same roadway but further from an intersection. Possible ways to reduce the motor vehicle contribution to ambient nanoparticulate matter include minimizing stop-and-go activity at an intersection (i.e., vehicles accelerating after a red light turns green) and identifying the small fraction of motor vehicles that emit a disproportionally large number of nanoparticles.