Study Of Formation Mechanisms And Source Apportionment Of Nitrate In Urban Atmospheric PM_2.5 By Using Its Dual Isotopes

In recent years,NO₃^-is gradually overtaking SO₄^2-as the most important secondary inorganic component in the fine particulate matter of urban haze in China.In view of the limitations of using mass concentration equilibrium to trace the secondary pollutants,current studies have started to use Bayesian isotope mixing model to analyze the source and formation process of NO₃^-.However,as the basis and prerequisite of the above studies,theδ¹⁵N characteristic spectra of NO_x（as the gaseous precursors of NO₃^-）sources in China are still lacking.In addition,it is beneficial to further improve the traceability of Bayesian isotope mixing models by clarifying the influencing factors of NO₃^-and its nitrogen and oxygen isotope compositions in urban aerosols at different time scales.On the basis of this,the study firstly constructed theδ¹⁵N characteristic spectra of six NO_x sources in Tianjin city.Subsequently,theδ¹⁵N characteristic spectra of six NO_x sources were applied to a Bayesian isotope mixture model to elucidate the sources of NO₃^-in PM_2.5 in Tianjin urban areas during five different time periods from 2018 to 2019 and the mid-winter heating period from 2017 to 2021,and to calculate the relative contribution of its formation pathways.A new idea of source analysis by PMF-Bayesian isotope mixing model was established to quantify the relative contribution of primary and secondary sources of NO₃^-,and we have achieved the goal of refining the source apportionment of NO₃^-in PM_2.5.The main research findings are as follows:（1）The δ¹⁵N of NO_x emitted from coal-fired power plants in Tianjin area was significantly more positive than other sources,the most depleted ¹⁵NO_x was released from soil,and theδ¹⁵N-NO_xfrom vehicle exhaust was more negative than that from biomass burning.Theδ¹⁵N of NO_x collected at the automobile tailpipe mouth is more negative than that at intersections and tunnels;since the NO_x in the tunnels is mainly contributed by vehicle exhaust,there the correspondingδ¹⁵N was more negative compared to the intersections,and the NO_x over intersections may be influenced by other sources in the urban atmosphere.Theδ¹⁵N of NO_x emissions from steel smelting mainly depend on the different equipment or processes.Compared with the results already reported abroad,theδ¹⁵N of NO_x released from coal-fired and gas-fired power plants in Tianjin area are both negative.（2）The concentrations of NO₃^-in atmospheric PM_2.5 in urban Tianjin were mainly influenced by different NO_x emission intensities,NO₃^-migration pathways and formation processes during five different sampling periods in 2018～2019.During the winter heating period,NO₃^-mainly comes from the local area of Tianjin and the surrounding Hebei area;while in spring and summer,the main potential source area of NO₃^-was located in southwest of Tianjin and partly from the Bohai Sea.Theδ¹⁵N-NO₃^-during early-and mid-winter heating period were consistent with the Rayleigh kinetic fractionation process,and the potential source areas of relatively enriched ¹⁵N and high concentrations of NO₃^-do not overlap at this time.In the transition from the end of heating to spring and summer,δ¹⁵N-NO₃^-was gradually influenced by the equilibrium fractionation of ¹⁵N in the NO_xtransformation process,and the potential source areas of enriched ¹⁵N and high concentration of NO₃^-gradually overlap.（3）By inputting the δ¹⁵N of NO_x sources obtained previously in the literature and in Tianjin into a Bayesian isotope mixing model,the calculated source fractions of NO₃^-in PM_2.5 for five different sampling periods from 2018 to 2019 in Tianjin were significantly different.As a result,this study concluded that the establishment ofδ¹⁵N characteristic spectra of NO_x sources is necessary,which could help to eliminate the uncertainties caused by the calculation of foreign source spectra.The study further entered 4～6δ¹⁵N-NO_x characteristic spectra in a cumulative increasing manner,and the interpretability of the results obtained from the model calculation was further enhanced,indicating that the NO_x source subdivision also helps to reduce the uncertainty in the model source contribution calculation.（4）The variation of δ¹⁵N-NO₃^-in atmospheric PM_2.5 in urban Tianjin during the mid-winter heating period from 2017 to 2021 was mainly influenced by the Rayleigh kinetic nitrogen isotope fractionation.δ¹⁸O-NO₃^-was significant positively correlated with the NOR,which increased with the atmospheric relative humidity.It was also significant negatively correlated with the O₃/O_x ratio,which decreased with the increase of Cl^-concentration,suggesting that NO₃^-in atmospheric PM_2.5in urban Tianjin during mid-winter heating is mainly generated by the heterogeneous hydrolysis reaction of N₂O₅,in addition to being influenced by the Cl^-in particulate matter.The results of potential source analysis and Pearson correlation analysis both showed that NO₃^-in the mid-winter heating period was mainly came from secondary transformation of NO_x emitted from anthropogenic combustion sources（e.g.,steel smelting,coal combustion and biomass burning）in Tianjin and Hebei.（5）By inputting the characteristic spectra of δ¹⁵N of six NO_x sources established in Tianjin,the Bayesian isotope mixing model quantified the relative contribution of NO₃^-sources to atmospheric PM_2.5 in Tianjin urban area in January for the last 5 years.The results showed that the relative contribution of coal-fired NO_x sources decreased from 42.1%in 2017 to 24.6%in 2021 year by year,mainly influenced by the background of large anthropogenic emission reduction in the North China;among which,the largest decrease of about 8%is observed between 2017 and 2018.The relative contribution of biomass burning increased by about 2.6%in 2018 compared to 2017,and it continued to increase by 0.8%in 2019;while it decreases successivelly to 25.4%and 24.5%in the 2020 and 2021,respectively.In addition,the relative contributions of the remaining NO_x sources all increase year by year.（6）Using the PMF receptor model,the results of this study show that not all the NO₃^-in PM_2.5collected in urban Tianjin in winter and summer were come from the secondary transformation of NO_x,but also included primary sources,such as biomass burning,sand and dust or sea salt.Based on this,by referring to previous literatures and formulated the stable nitrogen and oxygen isotopic compositions of the above three sources of NO₃^-,we recalculated the stable nitrogen and oxygen isotopes for the NO₃^-generated from secondary transformations.In addition,the NO_x sources of NO₃^-from secondary sources were quantified using a Bayesian isotope mixing model and finally combined with the analytical results of PMF.The results showed that the NO₃^-in both winter and summer PM_2.5 in urban areas of Tianjin was mainly from biomass combustion.By combining the PMF and Bayesian isotope mixing model,we achieved the division of primary and secondary NO₃^-in PM_2.5,and further enhanced the interpretability of NO₃^-source apportionment results compared to conventional Bayesian isotope mixing model.