Multimedia-Based Source Apportionment and Health Implications of Indoor Organophosphate Esters in Various Scenes of Urban Beijing, China

Studies on source apportionment of indoor OPEs rarely involve multiple media and characteristic source markers on emissions. Herein, we present a novel framework for quantitative and integrative source appointment of indoor OPEs by integrating OPE concentrations in multimedia into equivalent indoor concentrations (<i>EICs</i>) and using characteristic source markers on indoor emissions. Utility was demonstrated by applying it to five types of microenvironments where the 13 OPEs were ubiquitous (indoor dust: 993-14000 ng/g; indoor air: 0.549-14.1 ng/m³). The paired dust and air samples were adopted to construct the "Measured-Paired-<i>EICs</i>" data set for source apportionments via positive matrix factorization (PMF). Moreover, an alternative model method for constructing <i>EIC</i> data sets from single-medium measurements was established to improve the applicability of the framework. Accordingly, "Modeled-Paired-<i>EICs</i>", "Dust-Only", and "Air-Only" data sets were constructed for comparison. The extracted factors exhibited consistent contributions for "Measured-Paired-<i>EICs</i>" and "Modeled-Paired-<i>EICs</i>" solutions (RSD: 0.14-5.4%), while the "Air-Only" solution identified incomplete factors and the "Dust-Only" solution showed errors of 40.6-262%. Specifically, PMF analysis resolved seven known sources and other unknown sources. Furthermore, heterogeneities in source identification and source contributions were observed across various exposure scenarios; the probabilistic carcinogenic risk of TCEP (up to 7.19 × 10^-7) was close to the acceptable level (1 × 10^-6) and demands further attention.