大气细颗粒物中的硝酸盐（NO₃^-）与许多大气环境问题密切相关：PM_2.5中高含量硝酸盐能促进PM_2.5吸湿增长而影响大气能见度，增加对太阳光的散射作用并影响云层状态进而影响全球的降水和气候；此外，NO₃^-的光解可能影响大气氧化能力，促进SO₄^2-的形成和大气新粒子的生成。乌兰巴托和北京是东亚地区空气污染比较严重的两个城市。自2019年乌兰巴托控制原煤燃烧的政策颁布后，其大气颗粒物中NO₃^-的质量分数逐渐升高；而在北京大气颗粒物中，NO₃^-的质量分数也已超过SO₄^2-，成为主要的二次无机气溶胶组分。探究乌兰巴托和北京PM_2.5中NO₃^-的形成机制与来源，有助于阐明相关大气化学过程，明确当地PM_2.5中NO₃^-的来源贡献，为控制大气污染提供理论依据和数据支持。

本研究采集了乌兰巴托城区2020 - 2021年春、夏、秋、冬四个季节共83个PM_2.5样品，北京城区2020年11 - 12月共59个PM_2.5样品。对所采集样品分别进行了硝酸盐的¹⁵N、¹⁸O同位素特征分析及水溶性无机离子组分、碳质组分、痕量元素组分分析，利用同位素溯源技术探究了颗粒态硝酸盐的形成机制和来源贡献，同时结合正交矩阵因子分析法（Positive Matrix Factorization, PMF）量化了各污染源对NO₃^-的一次排放和二次转化贡献。主要研究结果如下：

（1）乌兰巴托冬季PM_2.5浓度是北京的3倍以上，但其NO₃^-浓度（2.1 ± 1.7 μg m^-3）比北京冬季（7.1 ± 7.4 μg m^-3）低51.3 - 61.5%，且表现为冬季>秋季>春季>夏季的季节变化趋势，北京冬季NO₃^-的均值浓度则随着污染程度的加重而增加。乌兰巴托城区大气PM_2.5中NO₃^-的¹⁵N、¹⁸O同位素值均呈现显著的季节浮动，寒冷季节高而温暖季节低；北京冬季δ¹⁸O-NO₃^-和δ¹⁵N-NO₃^-的观测值均高于乌兰巴托，且范围更加的宽泛。乌兰巴托的颗粒态NO₃^-生成机制以?OH + NO₂为主导（>57%），N₂O₅ + H₂O过程在冬季有所增强（<40%）；N₂O₅ + H₂O和?OH + NO₂对北京冬季NO₃^-生成的贡献大致相当，分别为46.8%，48.2%。与乌兰巴托相比，N₂O₅ + H₂O过程对北京NO₃^-生成的贡献较高。乌兰巴托城区氮同位素的分馏也具有明显的季节变化趋势，且与δ¹⁵N-NO₃^-值的季节变化趋势相反，表现为温暖季节较高而寒冷季节较低的特点。北京冬季NO₃^-生成过程中发生的氮同位素分馏值比乌兰巴托高了60%以上，说明北京冬季的大气物理化学条件，包括气-粒分配、化学转化、传输以及沉积过程等，更有利于氮同位素的分馏。

（2）贝叶斯同位素混合模型（Bayesian Isotopic Mixing Model, MixSIAR）的结果显示，无论是乌兰巴托全年还是北京冬季，机动车排放都是最重要的NO₃^-贡献源。就具体车载燃料类型而言，在乌兰巴托，汽油车>柴油车；而在北京，柴油车>汽油车>液化石油气（Liquified Petroleum Gas, LPG），这与两个城市机动车燃料利用结构一致。单就冬季而言，煤炭燃烧是乌兰巴托冬季NO₃^-的首要贡献源（~35%），机动车排放次之（~31%）；而在北京，机动车排放对NO₃^-浓度的贡献占绝对主导地位（>50%），生物源性土壤是第二大贡献源（~19%），燃煤和生物质燃烧等燃烧源的贡献相对较低。

（3）PMF模型分析结果表明，二次转化在两个城市都是绝对主导的NO₃^-来源，二次转化在乌兰巴托和北京的贡献比例分别是75.3%和77.4%。燃煤、机动车排放、扬尘和工业排放等污染源对NO₃^-的一次排放贡献在两个城市均不足10%。PMF与MixSIAR模型耦合结果显示，机动车排放和生物质燃烧产生的NO₃^-在两个城市绝大多数是通过二次转化形成的；在工业排放和煤炭燃烧方面，乌兰巴托市的二次形成比例是一次排放的2倍以上，北京一次则是二次的2倍以上，这与两个城市的污染控制技术和治理能力的先进程度有关，乌兰巴托粗放的工业排放导致NOx的大量排放，促进NO₃^-的二次生成，而北京对污染的全过程控制有效削减了NOx，减少了二次转化生成的NO₃^-；扬尘排放的NO₃^-在乌兰巴托NO₃^-以一次排放为主，而北京以二次转化生成为主，这可能与两个城市的风速、风向、温度、湿度等气象条件和土壤质地等有关。乌兰巴托的沙尘暴天气比较频繁，增加了扬尘源对一次排放的NO₃^-的贡献。

Nitrate (NO₃^-) in atmospheric particle matter is closely related to many atmospheric environmental problems. High nitrate content in PM_2.5 can promote the increase of PM_2.5 water absorption and affect the atmospheric visibility. In addition, NO₃^- photolysis may affect atmospheric oxidation capacity and promote the formation of sulfate and new atmospheric particles. The air pollution of Ulaanbaatar and Beijing is among the worst in the world. Ulaanbaatar issued the policy to control raw coal combustion in 2019, and the mass fraction of nitrate in atmospheric particle matter has gradually increased ever since. In Beijing, the mass fraction of NO₃^- has already surpassed that of SO₄^2-, becoming the main component of secondary inorganic aerosol (SIA). To explore the formation mechanism and sources of nitrate in PM_2.5 in Ulaanbaatar and Beijing is helpful to clarify the related atmospheric chemical process and source contributions of local nitrate in PM_2.5, thus providing theoretical basis and data support for air pollution control.

In this study, 83 PM_2.5 samples were collected in the urban area of Ulaanbaatar in 2020 - 2021, covering spring, summer, autumn and winter, and 59 PM_2.5 samples were collected in the urban area of Beijing in November to December 2020. The ¹⁵N and ¹⁸O isotopic characteristics of nitrate were analyzed, the concentration of water-soluble inorganic ion, carbonaceous species and trace element were measured. The formation mechanism and source contribution of particulate nitrate were explored using isotopic tool. At the same time, positive matrix factor analysis (PMF) was used to quantify the contribution from the primary and secondary emission to NO₃^-. The main results are as follow:

(1) The PM_2.5 concentration in Ulaanbaatar in winter was 3 times more than that of Beijing, but the NO₃^- concentration (2.1 ± 1.7 μg m^-3) was lower 51.3 - 61.5% than that in Beijing (7.1 ± 7.4 μg m^-3), with the seasonal variation trend of winter > autumn > spring > summer, while the mean NO₃^- concentration in Beijing increased with the aggravation of air pollution in winter. The ¹⁵N and ¹⁸O isotopic values of NO₃^- in PM_2.5 in Ulaanbaatar showed significant seasonal fluctuations, with higher values in cold season and lower values in warm season. The δ¹⁸O-NO₃^- and δ¹⁵N-NO₃^- values were higher in winter in Beijing than those in Ulaanbaatar. ?OH+NO₂ (>57%) was the dominant formation mechanism, the N₂O₅ + H₂O process was enhanced in winter (<40%) in Ulaanbaatar. While the contribution of N₂O₅ + H₂O and ?OH+NO₂ to NO₃^- was 46.8% and 48.2% in Beijing, respectively. N₂O₅ + H₂O contributed more to NO₃^- formation in Beijing compared with Ulaanbaatar. The fractionation of nitrogen isotope in urban area of Ulaanbaatar also showed an obvious seasonal variation trend and was higher in warmer season and lower in colder season, which was contrary to δ¹⁵N-NO₃^- values. The nitrogen isotope fractionation value during the process of NOx-NO₃^- conversion was significantly higher in Beijing than that in Ulaanbaatar in winter, indicating that the atmospheric physical and chemical conditions, including gas-particle distribution, chemical transformation, transport and deposition processes, were more beneficial to nitrogen isotope fractionation in Beijing.

(2) The results of Bayesian Isotopic Mixing Model (MixSIAR) showed that vehicle emissions were the most important contributor to NO₃^- in both cities. In terms of the contributions from specific fuel types employed by vehicles, gasoline > diesel in Ulaanbaatar, and diesel > gasoline > liquefied petroleum gas (LPG) in Beijing, which was consistent with the structure of fuel utilization in the two cities. In the case of winter alone, coal combustion (~35%) was the primary contributor to NO₃^- in Ulaanbaatar, followed by vehicle emission (~31%). However, in Beijing, the contribution of vehicle emission to NO₃^- concentration was absolutely dominant (>50%). Soil biomass emissions were the second contributor (~19%). The contribution of combustion sources such as coal combustion and biomass burning were relatively low.

(3) The results of PMF showed that secondary formation was dominant for NO₃^-, and the contribution proportion was 75.3% and 77.4% respectively in Ulaanbaatar and Beijing. Primary emissions from coal combustion, vehicle emission, dust and industrial emission contributed less than 10% to NO₃^- in both cities. The result from coupled model between PMF and MixSIAR showed that the majority of NO₃^- from vehicle emission and biomass burning generated by secondary transformation in two cities. In terms of industrial emission and coal combustion, the proportion of secondary formation was more than twice of that of primary emissions in Ulaanbaatar, and the proportion of primary emission more than twice of that of secondary formation in Beijing. This was probably related to the difference of pollution control technology and governance capacity of the two cities. In Ulaanbaatar, extensive industrial emission can lead to significant NOx emission and promote the secondary transformation of NO₃^-, while the whole process of pollution control has effectively reduced NOx resulting in the reduction of secondary formation of NO₃^- in Beijing. In Ulaanbaatar, NO₃^- emitted by dust was mainly primary, while, it was mainly derived from secondary formation in Beijing, which may be related to the wind speed, wind direction, temperature, humidity and other meteorological conditions as well as soil texture of the two cities, the frequent sandstorm weather increased the contribution of dust source to primary NO₃^- emission in Ulaanbaatar.