Influence of Sampling Artefacts on Measured PM, OC, and EC Levels in Carbonaceous Aerosols in an Urban Area

Sampling of particulate matter (PM) with filter media is subject to the influence of positive and negative artefacts. The complexity of these measurements derives from the use of different instrumentation, face velocities, and filter types. Carbonaceous compounds are one of the main PM chemical components affected by these artefacts. In this study, two main types of artefacts were evaluated at an urban site in Ghent (Belgium) during the summer period: those related to the sampling procedure (high- or low-volume samplers) and those related to the adsorption and/or volatilization of carbonaceous species. For their measurement, two high-volume (PM10 and PM2.5) samplers with single quartz fiber filters and two low-volume PM2.5 samplers (one of them coupled to an annular diffusion denuder) with double quartz fiber filters (front and backup) were used. PM2.5 mass concentrations measured by means of the low-volume sampler (front filter) were, on average, 31% higher than those registered with the high-volume sampler. PM2.5 mass concentrations registered with the low-volume sampler coupled to the diffusion denuder represented 88% of the PM2.5 mass collected with the undenuded low-volume sampler. Similar values for organic carbon (OC) were recorded for the high- and low-volume samplers. Front filter OC levels for the denuded samples represented 66% of the front filter OC mass on the undenuded samples. Using the backup filter OC data to apply a correction to the front filter data as suggested by Mader et al. (2003) only marginally reduced the discrepancy between the undenuded and denuded OC data, indicating that such correction does not result in consistent OC data for an urban area in Western Europe during summer.