Study on the oxidation potential of the water-soluble components of ambient PM2.5 over Xi’an, China: Pollution levels, source apportionment and transport pathways

Abstract

Reactive oxygen species (ROS) are a class of substances that are of general concern in terms of human health and are used to represent the oxidation potential (OP) of the atmosphere. In this study, the ROS levels in 116 daily fine particulate matter (PM2.5) samples taken over Xi'an in 2017 were measured with the dithiothreitol (DTT) method. The sources of DTTv (volume-based DTT consumption) in PM2.5 as well as their contributions were identified by both positive matrix factorization (PMF) and multiple linear regression (MLR) based on the measured chemical species in particulate matter (PM). The results showed that the yearly average DTTv over Xi’an was 0.53 nmol/min/m3 (0.19–1.10 nmol/min/m3). The highest DTTv level occurred in winter, followed by spring, summer and autumn. DTTv was the most strongly correlated with the water-soluble organic carbon (WSOC; r = 0.85), but the effects of WSOC on DTTv were very limited. SO2, NO2, CO, elemental carbon (EC) and K+ (r > 0.64) had moderate correlations with DTTv and were moderately related to environmentally persistent free radicals (EPFRs) (r = 0.56). The linear mixed-effects model showed that pollutants originating from incomplete combustion had greater effects on DTTv than those from complete combustion. Source apportionment results from PMF showed that motor vehicle emissions (27.4%), secondary sulfates (21.6%) and coal combustion sources (18.8%) were more important contributors to the DTTv in PM2.5 than dust sources (8.4%), metal processing (4.9%), industrial emissions (11.3%) and secondary nitrates (7.5%). The PMF results for the DTTv were consistent with the MLR results, which verified that both PMF and MLR are feasible methods for source apportionment of PM2.5 as well as specific species such as ROS and EPFRs. Backward trajectory clusters showed that the dominant cluster groups were local and regional transport, while the OP of the PM2.5 over Xi’an was affected more by long-range transport than by local transport. As stated above, the improvement of atmospheric oxidation potential require not only regional efforts but also large-scale joint cooperation. Furthermore, this study on the OP of PM as well as the specific source information provides important guidance for health effect research. Keywords: PM2.5, Source apportionment, DTT, PMF, MLR