Enhanced PM2.5 Source Apportionment Using Chemical Mass Balance Receptor Modeling and Scanning Electron Microscopy

One of the weaknesses of using receptor models to apportion the sources of ambient particulate matter is their inability to separate collinear sources such as different types of geological material. In order to develop a methodology to separate the different geological source contributions an ambient monitoring and source apportionment study was carried out for the cities of Reno and Sparks, NV during summer 1998. Chemical Mass Balance (CMB) receptor modeling was performed to estimate the contributions of both anthropogenic and natural sources to the observed ambient concentrations. Scanning electron microscopy was used to examine the geological component of the PM2.5 to determine the sources of that component. Chemical mass balance receptor modeling showed the dominant contribution to summertime PM2.5 mass in Reno and Sparks to be motor vehicle sources (~68%). Geological material was the second most abundant component of the PM2.5 (~14.5%). Sulfate was the predominant secondary species during the measurement period (~11%). The remaining components of significance were vegetative burning (~4%), secondary nitrates (~2%), and salt (NaCl) (0.6%). Scanning electron microscopy of selected ambient samples on a particle-by-particle basis showed the mineral component of the PM2.5 was predominantly aluminum-silicate in nature with a wide range of composition percentages for the major aluminum-silicate minerals (Na, Mg, Al, Si, K, and Ca). Virtually all of the particles examined had P and S in the typical aluminum-silicate spectra, which is attributed to contact with mobile source emissions. In approximately 10% of the examined particles were metallic in nature. Barium was also noted as a minor constituent of some particles, suggesting incorporation of diesel vehicle emissions. This evidence suggests that the source of the majority of the PM2.5 of geological origin in Reno and Sparks during the study period was from the resuspension of paved road dust. Thus, the amount of PM2.5 attributed to mobile source activity was in excess of 80%.