Nitrate aerosol formation and source assessment in winter at different regions in Northeast China

To fully identify the formation and source characteristics of aerosol nitrate during winter which accompanied by frequent haze in Northeast China, the nitrate concentrations, δ15N–NO3- and δ18O–NO3- characteristics of PM2.5 samples from four sites of different types including the (a) urban area, (b) suburban area, (c) suburban-rural transition area and (d) rural area in Northeast China were analyzed in this work. δ15N isotope fractionation (εN) during the conversion of atmospheric NOX (δ15N-NOX) to nitrate (δ15N–NO3-) was discussed on the basis of the estimated contributions of the gas phase oxidation pathway of OH radical (·OH) (γOH) and the heterogeneous hydrolysis pathway of N2O5 (γN2O5) to nitrate formation. Besides, the contribution of biological soil, coal combustion, biomass burning and motor vehicles to NOX emissions had also been analyzed using Bayesian isotope mixing model. Results showed that different types of sites had inconsistent source contributions. Obviously, coal combustion plays an absolutely indispensable role in the production of NOX in all the sites, especially in rural areas (d), where coal combustion contributed up to 89.8%. In addition, biomass burning still has a great effect on NOX emissions in the Northeast, showing a contribution of 5.2%–35.8%. It is worth noting that in urban site (a) where the tropospheric O3 concentration (34±9 μg m−3) was significantly lower than the other regions ((b): 54 ± 16 μg m−3, (c): 101 ± 21 μg m−3, (d): 98 ± 17 μg m−3), the contribution of the ·OH pathway to nitrate was as high as 69.4%, which was very different from the other three sites. The different formation mechanisms of nitrate in urban and non-urban may be due to different O3. Haze not only affects the sources of nitrate, but also changes the contribution of nitrate formation. Except for rural site (d), the contribution of the heterogeneous reaction (N2O5 pathway) of the other three sites during the haze period was greater (a: 35.1%, b: 60.0%, c: 59.0%, d: 60.8%) than in the non-haze period (a: 22.5%, b: 44.7%, c: 41.9%, d: 64.2%).