Quantification of Fossil and Non-Fossil SOA from Combined 14C/AMS-PMF Analysis for the SOAS Field Campaign

Abstract

The Southern Oxidant and Aerosol Study (SOAS) was a large field campaign during June-July 2013 in the southeast USA (Hidy et al., 2014; Hu et al., 2015; Carlton et al., 2018). Vast forested areas emitting large amounts of organic compounds and proximity to metropolitan areas present an ideal environment to investigate the influence of anthropogenic emissions on the biogenic secondary organic aerosol (SOA) formation. The main site of this study, located in rural Centreville, AL, was equipped with a wide variety of state-of-the-art analytical instruments. This project focuses on the source apportionment of the organic carbon (OC) fraction of ambient aerosol samples, through combination of radiocarbon (14C) data with positive matrix factorization information from online aerosol mass spectrometry measurements (AMS-PMF). Analysis of the long-lived radioactive isotope 14C is a unique source apportionment tool: it unambiguously separates fossil from non-fossil sources, as 14C has completely decayed in fossil fuels, whereas modern materials have the contemporary radiocarbon level (Szidat, 2009). 14C was measured for total carbon (TC) and elemental carbon (EC) (Zhang et al., 2012) from quartz filters that were collected at Centreville with daily resolution. This allowed the apportionment of fossil vs. non-fossil sources for EC and OC. Although OC mainly originated from non-fossil sources, a certain fraction was attributed to emissions from fossil sources. These results were compared with AMS-PMF data from a high-resolution time-of-flight AMS (Hu et al., 2015), which identified six different factors, i.e. biomass burning organic aerosol (BBOA), SOA formed through direct condensation of low-volatility oxidation products from isoprene (ISOPOOH-SOA), isoprene epoxydiols-derived SOA (IEPOX-SOA), low-oxidized oxygenated organic aerosol I, attributed to mostly biogenic sources (LO OOAI), low-oxidized OOA II, attributed mostly to anthropogenic sources (LO-OOAII) and more-oxidized OOA (MO-OOA). On average, the less well-understood fractions LO-OOAI, LO-OOAII and MO-OOA comprise ~3/4 of the total organic aerosol mass. 14C analysis of EC enables the distinction of sources of this carbonaceous aerosol fraction between fossil-fuel combustion (mainly from traffic) and biomass burning. It indicated a larger contribution from biomass burning compared to other source apportionment techniques or results from bottom-up emission inventories. The combination of 14C and AMS-PMF analysis provides the potential to apportion fossil vs. non-fossil sources for components for which the non-fossil fraction cannot by analyzed directly, such as SOA (Zotter et al., 2014). In this work, we present such results for the SOAS campaign based on Markov chain Monte Carlo calculations to gain more insight into the sources of SOA precursors. LO-OOAI, LO-OOAII and MO-OOA reveal different contributions of fossil and non-fossil sources, which allows a better understanding of these AMS-PMF factors. Carlton, A. M., et al. (2018), Bull. Am. Met. Soc., in press. Hidy, G. M., et al. (2014), Atmos. Chem. Phys., 14, 11893-11914. Hu, W. W., et al. (2015), Atmos. Chem. Phys., 15, 11807-11833. Szidat, S. (2009), Chimia, 63, 157-161. Zhang, Y. L., et al., (2012) Atmos. Chem. Phys., 12 (22), 10841-10856. Zotter, P., et al. (2014), J. Geophys. Res. Atmos., 119, 6818–6835.