Source apportionment of PM_2.5-bound elemental carbon via stable carbon isotope analysis: A comparison study with...

Isotopes serve as effective tracers for identifying pollutant sources, with carbon isotopes (δ¹³C) enabling the determination of the sources of carbonaceous aerosols in PM_2.5. In this study, PM_2.5 samples were collected in Changzhi, China, during winter and summer to determine the mass concentrations and stable carbon isotope (δ¹³C) of the organic carbon (OC) and elemental carbon (EC). The Bayesian mixing model (MixSIAR) was employed to determine the contributions to EC sources in PM_2.5 based on δ¹³C values, and the results were compared with those from positive matrix factorization (PMF) receptor modeling, which used chemical concentrations. During sampling, the δ¹³C_OC values exhibited clear seasonal variations in their isotope levels of -23.44‰ in the winter and -26.81‰ in the summer, while δ¹³C_EC exhibited no significant variations in the winter (-25.33‰) or summer (-26.37‰). δ¹³C_OC levels in the summer were obviously negative than in the winter, which indicated that δ¹³C_OC levels in the summer might be more influenced by secondary organic aerosols. The MixSIAR model identified five sources of EC in PM_2.5: gasoline vehicle exhaust (summer: 26.5%; winter: 25.0%), diesel vehicle exhaust (43.5%; 28.0%), coal combustion (14.2%; 22.9%), biomass burning (11.6%; 17.2%), and road dust (4.2%; 6.9%), with distinct seasonal variability in contributions. The results obtained by the PMF model were generally consistent with the results mentioned above. Minor discrepancies in source contributions (e.g., biomass burning, road dust) were attributed to the uncertainty of the tests and methods. The consistency of the results between PMF and MixSIAR models provides cross-validation that reduces overall uncertainty in source apportionment. The preliminary application of carbon isotopes shown in this study highlights that isotopes are effective indicators for source apportionment of the atmospheric pollutants, providing a reliable and cross-validating perspective.