Chemical composition of outdoor and indoor PM 2.5 collected during haze events: Transformations and modified source contributions resulting from outdoor-to-indoor transport.

Changes in the chemical constitution and sources of ambient PM2.5 following the infiltration of air into indoor environments were investigated. We collected PM2.5 samples from air inside and outside 31 rooms in Beijing residences during hazy episodes. We calculated the indoor-to-outdoor ratios and the correction (ki ) of each infiltration factor for each chemical component of PM2.5 to determine the effects of infiltrative behavior. The outdoor and indoor mass concentrations of PM2.5 during the sampling period were 70-460 and 10-315 μg/m3 , respectively. Differences in the average indoor-to-outdoor ratios of PM2.5 mass and each component (mean value ± standard deviation: PM2.5 mass = 0.53 ± 0.26, organic matter = 0.75 ± 0.34, elemental carbon = 0.62 ± 0.31, trace elements = 0.62 ± 0.26, SO 4 2 - = 0.67 ± 0.32 , NH 4 + = 0.53 ± 0.54 , NO 3 - = 0.45 ± 0.36 , Cl- = 0.37 ± 0.35, and crustal dust = 0.30 ± 0.19) may be attributed to size distribution, chemical properties, temperature, and humidity. The positive matrix factorization model was applied to calculate the source contributions to equivalent population exposure (Indoor concentration·Indoor time fraction + Outdoor concentration·Outdoor time fraction). The contributions of fossil fuel combustion, secondary source, vehicle exhaust, and mixed dust to the equivalent PM2.5 population source exposure were 37%, 24%, 22%, and 17%, respectively.