Source apportionment of PM10 and its oxidative potential by spatially resolved samplings.

In the last few years, the relevance of high spatial resolution data for PM source apportionment methods has been highlighted. This method allows reducing analytical costs by sampling at various sites for long periods, thus obtaining a PM chemical characterization at high spatial resolution. Furthermore, the variance of concentrations is mainly due to the spatial distribution of the emission sources, with a low contribution of the meteorological conditions and the application of multivariate methods for source apportionment allows obtaining more stable and representative chemical profiles, especially for the local sources. In this study, automatic and very-low volume samplers were employed to evaluate the spatial variability of PM10 concentration and composition at 9 sampling sites distributed over the Sacco River Valley (southern area of Latium Region). The Valley is characterized by frequent and severe atmospheric stability conditions, by high-density population and by the presence of significant number of industrial settlements that had led to a significant pollution phenomena over the years (Battisti et al., 2022; Donateo et al., 2020; Sozzi et al., 2017). PM daily limits exceedances are frequently recorded in the Sacco Valley during the cold season, probably due to the low atmospheric mixing properties (ARPA Lazio, 2022; ISPRA, 2020). At each site, samples have been collected on both quartz and Teflon membranes. Teflon PM filters were subjected to a chemical fractionation procedure (soluble and insoluble fractions), that permits to increase the selectivity of the elements as source tracers (Massimi et al., 2021; Perrino et al., 2010; Canepari et al., 2009). Concentration of elements was determined through a quadrupole inductively coupled plasma mass spectrometer (ICP-MS). The soluble PM10 fraction was also used for the application of three oxidative potential assays (Ascorbic Acid -A A, Dithiothrietol - DTT, Dichlorofluorescein - DCFH) and for the analysis of inorganic ions and levoglucosan (LVG). Quartz membrane filters were instead used for the determination of elemental and organic carbon (EC and OC, respectively). Positive Matrix Factorization (PMF) was applied to identify and assess the main sources of particulate emissions in the sampling areas and to determine the weight of each emission source on the OPAA, OPDTT and OPDCFH. Results showed that, during winter, half of PM mass is due to biomass burning (BB) processes; PMF identified two different profiles of BB, characterized by a different tracers ratio, whose relative contributions strongly depends on the monitored site. The first profile (BB1) contains high concentration of OC, EC, LVGSN and K+; the second one (BB2) contains a higher contribute of other inorganic tracers, such as Cs, Rb and Tl in their soluble fraction. Together with LVGS, these last parameters have been demonstrated to be very selective tracers of biomass combustion for domestic heating (biblio). It is reasonably hypothesized that the two contributions correspond to different heating appliance and/or wood material. The oxidative potential measured by the DCHF method appear to be predominantly associated with the biomass burning (mainly BB1) and the AA assay to the non-exhaust vehicular traffic. OPDTT had significant contributes by both the sources.