Estimating the contribution of point sources to atmospheric metals using single-particle mass spectrometry

Single-particle mass spectra were collected using an Aerosol Time-of-Flight Mass Spectrometer (ATOFMS) during December of 2003 and February of 2004 at an industrially impacted location in East St. Louis, IL. Hourly integrated peak areas for twenty ions were evaluated for their suitability in representing metals/metalloids, particularly those reported in the US EPA Toxic Release Inventory (TRI). Of the initial twenty ions examined, six (Al, As, Cu, Hg, Ti, and V) were found to be unsuitable due to strong isobaric interferences with commonly observed organic fragments, and one (Be) was found to have no significant signal. The usability of three ions (Co, Cr, and Mn) was limited due to suspected isobaric interferences based on temporal comparisons with commonly observed organic fragments. The identity of the remaining ions (Sb, Ba, Cd, Ca, Fe, Ni, Pb, K, Se, and Zn) was substantiated by comparing their signals with the integrated hourly signals of one or more isotope ions. When compared with one-in-six day integrated elemental data as determined by X-ray fluorescence spectroscopy (XRF), the daily integrated ATOFMS signal for several metal ions revealed a semi-quantitative relationship between ATOFMS peak area and XRF concentrations, although in some cases comparison of these measurements were poor at low elemental concentrations/ion signals due to isobaric interferences. A method of estimating the impact of local point sources was developed using hourly integrated ATOFMS peak areas, and this method attributed as much as 85% of the concentration of individual metals observed at the study site to local point sources. Hourly surface wind data were used in conjunction with TRI facility emissions data to reveal likely point sources impacting metal concentrations at the study site and to illustrate the utility of using single-particle mass spectral data to characterize atmospheric metals and identify point sources.