Research Submicrometer and Supermicrometer Particulate Emission from Spark Ignition Vehicles Z. D. RISTOVSKI,* L. MORAWSKA, N. D. BOFINGER, AND J. HITCHINS Center for Health and Medical Physics, Queensland University of Technology, P.O. Box 4233, Brisbane Qld, 4001, Australia Particulate emissions from 11 gasoline-powered and 2 liquefied petroleum (LPG)-powered passenger vehicles were characterized during the Accelerated Simulation Mode driving cycles on a chassis dynamometer. The test fleet consisted of 10 catalyst-equipped vehicles operated with unleaded gasoline (5Ford Falcons and 5Holden Commodores), 2 LPG-powered vehicles (both Ford Falcons), and 1 older type noncatalyst vehicle operated with leaded gasoline. Particulate characterization included determination of total particulate number concentration and size distribution using the scanning mobility particle sizer (SMPS) and the aerodynamic particle sizer (APS). The average particle number concentrations in the SMPS range for all modes was lower for Ford Falcons and somewhat higher for Commodores, with values of 1.5 × 10 4 and 4.1 × 10 4 cm -3 , respectively. This difference is significant and was observed for all modes. The number concentration levels were higher for the LPG-fueled cars (8.4 × 10 4 cm -3 ) and for the leaded gasoline-powered vehicle (7.9 × 10 5 cm -3 ). There was not a significant variation in particle count median diameter in the SMPS and the APS ranges, either for different operating conditions of the vehicles investigated or between different vehicle groups. The observed size distributions were bimodal with average values of CMD ranging from 39.1 to 60.2 nm in the SMPS range and from 0.9 to 1.4 μm in the APS range. The results obtained from this study can be used as a first- order estimation toward emission inventories for vehicle groups included in the investigations. Introduction Exhaust emissions from spark ignition engine consist of gaseousand particulate phases. The main gaseousemissions include hydrocarbons (HC), CO, NOx, CO2, SO2, and water vapor. Gasoline particulate matter is formed as a result of incomplete combustion ofgasoline. The particlesare mostly carbonaceous spherical submicron agglomerates ranging from 10 to 80 nm, consisting of a carbon core with various associated organic compounds. The main components of particulate phase include soot; ash; trace elements such as lead, iron, chlorine, and bromine; organic compounds; and lowto medium boilingfractionsofengine oil(1). Lubricating oil and other fuel hydrocarbons may also contribute. The sulfate particles present in gasoline engine emission are mainly from catalyst-equipped vehicles utilizing unleaded gasoline (2). Common organic compounds are polycyclic aromatic hydrocarbons (PAHs), such as pyrene, chrysene, benzo[a ]pyrene, and BaP. Particle emissions from vehicles are currentlyunder close scrutiny with respect to their contribution to ambient particles relative to other sources. The PM10 fraction of the particles (mass concentration of particles smaller than 10 μm)has been linked to various health effects (3-7). The size distribution of these particles plays an important role as it influences the depth of penetration and deposition in the lung and hence the toxicological effects. The lack of knowledge ofthe size distribution makes it difficult to explain the toxicological mechanism related to the reported health effects. However, there is evidence of association between concentrations of particles and changes in a number of respiratory health indicators ranging from changes in lung function to hospital admissions and death. There is significantly less information available on par- ticulates from gasoline spark ignition emissions than from diesel emissions. Gasoline particulate matter is emitted at levels lower than a milligram per kilometer. Because it is emitted at such low levels,it is difficult to measure accurately. Totalparticulate massemissionsfrom sparkignition vehicles are significantly lower than diesel emissions and are usually well below any emission standards (1, 8-10). It has been noted, however, that while particulate emissions from individual spark ignition vehicles are lower than individual emissionsfrom dieselvehicles,the totalcontribution ofboth types of vehicles to air emissions could be similar due to the exceedingly higher number of spark ignition vehicles than diesel-fueled vehicles (11). The low mass concentration of emissions from spark ignition engines is related to a lower number concentration in the emissions and also to the fact that the emitted particles are smaller than particles from diesel emissions (12). Both air quality standards and current vehicle emissions legislation are based on mass measurements: the former on mass per unit volume (PM10 and PM2.5), and the latter on the total mass of particles emitted per kilometer. In both, there is no reference either to the size of the particles or to the number concentration of emitted particles. The available emissions data are limited and scattered. Most of the available data are on particle mass emission (8-11, 13),and there is only limited data on the size distribution of these particles or on the particle number emissions (9, 10), either in the submicrometer or supermicrometer region. Insuf- ficient data on critical factors related to spark ignition emissions, including particulate emissions, have been iden- tified as the most important issue in developing emission inventories (14, 15). An important factor requiring further investigations and consideration in developing in-service vehicle testing pro- cedures is documented test-to-test variability in emissions (16). The authorshowed that malfunctioningvehicles,which are amongthe highest emitters,often showveryinconsistent emissions and are likely to escape emission testing if these are not frequent enough. Those vehicles are most often modern computer-controlled vehicles that have malfunc- tioning emission control systems that have not been tam- pered with. *Corresponding author fax: 61 7 3864 1521; e-mail: z.ristovski@ qut.edu.au. 10.1021/es980102d CCC: $15.00  1998 American Chemical Society VOL. 32, NO. 24, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 3845 Published on Web 10/29/1998