Demonstrating that speciation of organic fraction does matter for source apportionment : use of specific primary and secondary organic markers

Organic aerosol (OA) is a major part of atmospheric fine particulate matter (PM), accounting for approximately 20-60% w/w of PM in the continental mid-altitudes. However, ambient OA remains poorly understood due to mixed source origins and processes (anthropogenic and biogenic, primary and secondary). The objective of this study was to refine source apportionment of PM10 OA by Positive Matrix Factorization (PMF) using specific primary and secondary organic molecular markers. PM10 samples were collected over a one year period (2013) at an urban station in Grenoble (France) every third day (24 h-basis sampling) on quartz filters, and an extended chemical characterization was performed (~216 species quantified) including specific primary organic markers i.e. levoglucosan, polyols (arabitol, mannitol), 1-nitropyrene (diesel emission), PAHs, alkanes, hopanes, etc. and secondary markers i.e. nitro- and oxy-PAHs, hydroxyglutaric acid (α-pinene secondary organic aerosol SOA), α-methyl glyceric acid (isoprene SOA), DHOPA (toluene SOA), etc. together with other PM chemical species such as OC/EC, Humic Like Substances (HuLiS), ions/cations (Na+, Mg2+, NH4+, Cl-, SO42-, NO3-), metals (Ba, Cu, Ti, Zn, Sb,...). Results showed that a better source apportionment of PM10 OA fraction was achieved using specific organic markers with additional sources resolved such as biogenic SOA, anthropogenic SOA, primary biogenic (fungi) and plant debris by comparison to more traditional PMF. More than 50% of OC seemed secondary in nature, and a high contribution of anthropogenic SOA was noticed in winter during a specific PM pollution event. Primary and secondary sources of HuLiS were also investigated in this study. Discussion will further underline the details of the chemical and temporal/seasonal profiles of each factor, and their relative contributions