Trace metal analysis of atmospheric particulate matter: A comparison of personal and ambient samplers

This study compared the size-resolved trace metal concentrations from a standard ambient sampler to those of three commercial personal samplers. Trace metals data from each personal sampler were also inter-compared. Co-located ‘fixed site’ samples were collected in two urban sites in the United States using three types of personal samplers (BGI PM2.5 cyclone, URG PM2.5 cyclone, and Sioutas TM personal cascade impactor sampler [PCIS]) and one standard ambient sampler fitted with a PM2.5 cyclone. This experimental design enabled direct evaluation of the relative performance of the personal samplers. Results of mass measurements showed good agreement between the standard equipment and personal samplers; however, trace metal comparisons (measured by ICP–MS) showed reproducible concentration biases among personal samplers. These biases can be summarized as BGI < URG < PCIS (0.27 ± 0.08, 0.72 ± 0.08, and 1.32 ± 0.34 relative to the ambient sampler). Elements exhibiting larger variation between personal samplers were associated with higher mass fractions in the coarse fraction, suggesting the PM2.5 size-cut a potential cause of this deviation.