Temporal variations and potential sources of organophosphate esters in PM2.5 in Xinxiang, North China

Abstract

We monitored the concentrations of 10 organophosphate esters (OPEs) in 52 fine particulate matter (PM2.5) samples in Xinxiang, Henan Province, North China, in 2015. During the sampling period, the OPE concentrations in most samples (n = 47) differed minimally and were relatively stable (mean: 2.02 ± 0.93 ng m−3), although several samples (n = 5) had high total OPE (Ʃ10OPE) concentrations (mean: 9.99 ± 5.69 ng m−3), which may have been influenced by high PM2.5 levels. Meanwhile, some samples had high PM2.5 concentrations but low Ʃ10OPE concentrations (i.e. low OPE/PM2.5 ratios) or low PM2.5 concentrations but high Ʃ10OPE concentrations, which might have been influenced by air mass sources. Therefore, we assessed air mass sources using the Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) model and wind direction frequency data, and subsequently analysed PM2.5 and OPE sources using a potential source contribution function (PSCF) model. The results revealed that air mass sources couldn't represent the source of specific pollutants, including PM2.5 and OPEs. Generally, both PM2.5 and OPEs were from Henan and Shandong Provinces; however, the major source areas differed, which may have resulted from diverse pollution characteristics in various source areas. The principal component analysis and PSCF results revealed that the 10 OPEs could be segmented into three groups, which were associated with different source areas. These results suggested that pollution characteristics of contaminants in source areas should be considered in source apportionment.