沈阳大气PM2.5中多环芳烃的污染特征和来源解析研究

Polycyclic aromatic hydrocarbons (PAHs) associated with PM2.5 have been recognized as a class of organic pollutants of the utmost concern in China due to their adverse health impact. A critical step toward risk assessment and management of atmospheric PAHs is to identify their emission sources. Therefore, study on the methodology of source apportionment for PAHs is very important. Receptor models have been widely used for source apportionment of oganic pollutants in various environmental media. Receptor models assess contributions from all the major sources based on the observed sample composition. In this study, gas chromatography with triple quadrupole mass spectrometry (GC-MS/MS) will be used to analyze the PAHs in PM2.5 samples (receptor) in typical urban areas. Typical PAH source profiles will be determined by field measurement, and literature source compositions are to be summarized. An advanced receptor model (factor analysis with non-negative constraints, FA-NNC) is adopted for the source apportionment of atmospheric PAHs. Monte Carlo simulation is used for quantitatively evaluating the uncertainties of the model results, and optimizing the input parameters and enhancing the prediction ability of the model. In addition, a level III fugacity model based on steady state assumption will be adopted to predict PAH profiles in various environmental media. PAH source profiles rectified through the fugacity model are to be used for the receptor modeling.The objective of this research is to establish a novel receptor model for source apportionment of PAHs by the combination of a level III fugacity model and Monte Carlo uncertainty analysis.

PM2.5中携带的多环芳烃（PAHs）对人体健康危害较大，了解大气PM2.5中PAHs的污染水平并定量解析其主要来源对有效控制大气污染具有重要意义，因此有必要在实测数据的基础上开展针对PAHs的来源解析方法研究。本研究拟通过气相色谱三重串联四极杆质谱（GC-MS/MS）测定城市典型区域PM2.5中的PAHs，并且实测和总结PAHs典型污染源指纹谱图。在实测数据的基础上，以目前较先进的一种受体模型--非负约束的因子分析模型（FA-NNC）--为理论基础，通过蒙特卡罗不确定性分析定量评价模型的主要影响参数和模型结果的不确定度，优化模型的输入参数，改进模型评价指标，提高模型结果的准确性。此外，采用三级稳态逸度模型对PAHs典型污染源的指纹谱图进行修正，建立逸度模型结合受体模型的来源解析方法。

受体模型的成功应用很大程度上要依靠对污染物的大量采集和准确分析，而受体模型的结果往往难以进行验证。本研究对沈阳大气PM2.5进行了采样，通过三重串联四级杆质谱仪分析了样品中的多环芳烃，并开发了多种受体模型方法，采用受体模型对沈阳大气PM2.5中PAHs的来源进行了解析。在模型开发方面，使用matlab语言编写开发并改进了一种受体模型，绝对因子得分/多元线性回归模型。同时，本研究采用数据模拟矩阵，假设多氯联苯（PCBs）产品Arolcor1016，Aroclor1248和Aroclor1260为PCBs的三个主要污染源，并设定它们在40个样品中的贡献值，由此组成一个有25种PCBs系列物和40个样品的虚拟数据矩阵，使用非负约束的因子分析模型应用于该虚拟矩阵，很好的复原了污染源的指纹谱图和贡献率，说明非负约束的因子分析模型具有较好的来源解析能力。非负约束的因子分析模型、正矩阵因子分解模型和绝对因子得分/多元线性回归模型基本得到了一致的结果。