Mercury, trace elements and organic constituents in atmospheric fine particulate matter, Shenandoah National Park, Virginia, USA: A combined approach to sampling and analysis

Abstract

Compliance with U.S. air quality regulatory standards for atmospheric fine particulate matter (PM 2.5 ) is based on meeting average 24 hour (35 μ m −3 ) and yearly (15 μg m −3 ) mass‐per‐unit‐volume limits, regardless of PM 2.5 composition. Whereas this presents a workable regulatory framework, information on particle composition is needed to assess the fate and transport of PM 2.5 and determine potential environmental/human health impacts. To address these important non‐regulatory issues an integrated approach is generally used that includes (1) field sampling of atmospheric particulate matter on filter media, using a size‐limiting cyclone, or with no particle‐size limitation; and (2) chemical extraction of exposed filters and analysis of separate particulate‐bound fractions for total mercury, trace elements and organic constituents, utilising different USGS laboratories optimised for quantitative analysis of these substances. This combination of sampling and analysis allowed for a more detailed interpretation of PM 2.5 sources and potential effects, compared to measurements of PM 2.5 abundance alone. Results obtained using this combined approach are presented for a 2006 air sampling campaign in Shenandoah National Park (Virginia, USA) to assess sources of atmospheric contaminants and their potential impact on air quality in the Park. PM 2.5 was collected at two sampling sites (Big Meadows and Pinnacles) separated by 13.6 km. At both sites, element concentrations in PM 25 were low, consistent with remote or rural locations. However, element/Zr crustal abundance enrichment factors greater than 10, indicating anthropogenic input, were found for Hg, Se, S, Sb, Cd, Pb, Mo, Zn and Cu, listed in decreasing order of enrichment. Principal component analysis showed that four element associations accounted for 84% of the PM 2.5 trace element variation; these associations are interpreted to represent: (1) crustal sources (Al, REE); (2) coal combustion (Se, Sb), (3) metal production and/or mobile sources (Mo, Cd, Pb, Cu, Zn) and (4) a transient marine source (Sr, Mg). Concentrations of Hg in PM 2.5 at background levels in the single pg m −3 were shown by collection and analysis of PM 2.5 on filters and by an automated speciation analyser set up at the Big Meadows air quality site. The speciation unit revealed periodic elevation of reactive gaseous mercury (RGM) that co‐occurred with peaks in SO 2 , indicating an anthropogenic source. GC/MS total ion current chromatograms for the two sites were quite similar indicating that organic signatures were regional in extent and/or that the same compounds were present locally at each site. Calculated carbon preference index values for n ‐alkanes indicated that plant waxes rather than anthropogenic sources, were the dominant alkane source. Polycyclic aromatic hydrocarbons (PAHs) were detected, with a predominance of non‐alkylated, and higher molecular weight PAHs in this fraction, suggestive of a combustion source (fossil fuel or forest fires).