Comparative elemental characterization of atmospheric fine and coarse PM samples by PIXE and ICP-MS/MS: results from a one-year measurement campaign in a coal-powered thermal power plant city

Daily fine (PM2.5) and coarse (PM2.5-10) particulate matter samples were collected from rural and urban monitoring stations located in an industrial region of mid-western Turkey, Kutahya city for one year. All fine- and coarse-mode samples were analyzed by both triple quadrupole inductively coupled plasma-mass spectrometry (ICP-MS/MS) and particle-induced X-ray emission (PIXE) techniques. The two techniques were comparatively evaluated and analytic performances were compared for 21 identified elements including Al, As, Br, Ca, Cr, Cu, Fe, Mg, Mn, Mo, Na, Ni, Pb, Rb, S, Se, Sr, Ti, V, Zn, and Zr. MDLs for PIXE were relatively higher than those of ICP-MS/MS by analyzing over 800 PM samples. The highest and lowest MDLs estimated for PIXE were in the range of 5.0 x 10(-3) (Na)-9 x 10(-6) mu g m(-3) (Ti) and between 8.5 x 10(-5) (S)-6 x 10(-8) mu g m(-3) (As) for ICP-MS/MS, respectively. Other metrics such as percent below the detection limit (% BDL), precision, and SRM recoveries were also used as variables of comparison. % BDL for PIXE varied between 0% (for Zn) and 83% (for Mo) with important marker elements such as Ca, S, K, Cl and Na being 4%, 8%, 11%, 12% and 21%, respectively. For ICP-MS/MS, % BDL varied between 0% (for Mg) and 98% (for Se) and important marker elements such as Na, Al, Ca, Cr, V and S were between 0% and 1% below detection. Absolute atmospheric concentration ratios (C-PIXE/CICP-MS/MS) of most elements except for Br, Mg, Na, and Zr were approximately 1, suggesting a good relationship between the two analytical techniques for elemental characterization of PM samples. In terms of atmospheric concentrations, the two techniques yielded almost the same magnitudes for the elements As, Br, Cr, Fe, Mn, Mo, Pb, Rb, Se, Sr, Ti, V, and Zn - hence, they were found to be equally competitive for measurement of these elements. Ca, Si, Ti, Cu, Zn, S, and Cl were detected in 90 to 100% of the samples. Likewise, Na, Mg, Al, S, Ca, Ti, V, Cr, Mn, Fe, Ni, and As were detected by ICP-MS/MS in almost every sample (similar to 99-100%).