The objective of this study is to investigate the pollution sources of PM2.5 in Chiayi-Tainan area by using the Positive Matrix Factorization (PMF) receptor model with C and Pb isotopic fingerprints. In this study, ambient PM2.5 samples were collected from seven sites (Xingang, Puzi, Chiayi, Shanhua, Beimen (the sampling site was moved to Xuejia in summer), Tainan, and Matou (the sampling site was moved to Annan in summer) during spring (March) and summer (August) of 2021, respectively. We also collected PM2.5 samples from 6 specific sources: two petrochemical factories, two electric arc-furnace steel mill factories, and two incinerators, for pollution sources identification. All the PM2.5 samples were characterized with the measurements of water-soluble ions, organic carbon, elemental carbon, crustal elements, heavy metals, and isotopic compositions of carbon (13C and 14C) and lead (206Pb, 207Pb, and 208Pb).
The results showed that condensable particulate matter (CPM) was the predominant component in total PM2.5 emitted from stationary pollution sources. The level of inorganic CPM was higher than that of organic CPM in all stationary pollution sources. The major chemical compositions from six stationary pollution sources were: (1) petrochemical factory (coal-fired boiler): sulfate and magnesium ion were important species; the predominant elementals were different from two petrochemical factories; the ranges of 206Pb/207Pb and 208Pb/207Pb were 1.2094~1.2269 and 2.4314~2.4474, respectively; 2.) electric arc-furnace steel mill factory: the predominant chemical compositions were sulfate and potassium ion; the predominant elementals was Zn; the ranges of 206Pb/207Pb and 208Pb/207Pb were 1.1441~1.1583 and 2.4088~2.4319, respectively; 3.) incinerators: the predominant chemical compositions were organic carbon, sodium ion, potassium ion, calcium ion, chlorine ion, and sulfate; the predominant elementals were different from two incinerators; the ranges of 206Pb/207Pb and 208Pb/207Pb were 1.1408~1.1565 and 2.4138~2.4328, respectively.
In all ambient PM2.5 samples, the average level of PM2.5 was 19.5 ± 17.9 μg/m3 in the study area. The predominant chemical compositions of the ambient PM2.5 were sulfate, nitrate, ammonium, and organic carbon, which accounted for 18%, 25%, 13%, and 16%, respectively. The abundance of modern carbon (13%) was slightly higher than that of fossil carbon (10%). During high PM2.5 episodes, our data showed that the concentrations of sulfate, nitrate, ammonium, and fossil and modern carbonaceous aerosols were elevated. These results implied that photochemical reactions and burning activities were important factors influencing the PM2.5 concentrations in the study areas. The averages of δ13C values for PM2.5 was -25.4‰ (-30.8 to -20.4‰) and -26.5‰ (-31.5 to -22.4‰) in the spring and summer, respectively. The average 206Pb/207Pb and 208Pb/207Pb ratios in PM2.5 were 1.1527 (1.1454~1.1617) and 2.4380 (2.4213~2.4568) in the spring, respectively, and 206Pb/207Pb and 208Pb/207Pb ratios were 1.1453 (1.1124~1.15540) and 2.4211 (2.3832~2.4471) in the summer, respectively. Our data showed that there was heavier carbon and lead isotopic compositions in spring than that in summer, which reflects that the pollution sources differ in summer and spring.
We combined the PMF and isotopic compositions of ambient and pollution sources to investigate the pollution sources for PM2.5 in the study area. Our results indicated that traffic emission, biomass burning, oil combustion were major pollution sources to carbonaceous aerosols in PM2.5. Traffic emission, biomass burning, and steel factories were major pollution sources to lead in PM2.5. However, we need more data to investigate the pollution source of lead for Chiayi, Puzi, and Beimen sites in spring.
Finally, we combined the study data from 2016 to 2021 (central Taiwan, Kaohsiung and Pintung, and Chiayi and Tainan) to compare the difference between chemical compositions, influencing factors, and pollution sources of PM2.5. We found that the predominant chemical compositions and influencing factors in PM2.5 were similar in all study areas. However, the variations of carbon or lead isotopic compositions and pollution sources differ according to the study areas. Consequently, we should make specific control policies for different study areas.