我国典型细颗粒物排放源单颗粒质谱特征对比研究

Single particle mass spectra of typical fine particulate matter sources in China

针对现有污染源单颗粒质谱成分谱较少且缺乏对比总结的问题,本论文收集了工艺过程源、扬尘源、机动车尾气源、燃煤源和生物质燃烧源的单颗粒质谱图数据,使用聚类算法分析了各源类单颗粒物的主要化学组成特征的差异性.结果表明工艺过程源和燃煤源颗粒物均含有OC类、矿物质类、EC类、重金属类、富钾/左旋葡聚糖/硫酸盐硝酸盐类、OCEC类,但各颗粒类型的占比有明显差异,燃煤源中EC类占比明显高于工艺过程源而重金属类占比低于工艺过程源.生物质燃烧源主要由OC类、其他有机物类、富钾/左旋葡聚糖/硫酸盐硝酸盐类和OCEC类组成.扬尘颗粒的主要类型为矿物质类.从单颗粒化学组成上看,工艺过程源排放的有机颗粒中OC常与SO_(4)^(2-)内混,元素碳颗粒中EC与NO_(3)^(-)或SO_(4)^(2-)内混.燃煤源排放的碳质颗粒中EC、OC成分往往与NO_(3)^(-)和SO_(4)^(2-)或者单独与SO_(4)^(2-)发生内混,而生物质燃烧排放的有机颗粒物上OC与K、左旋葡聚糖、NO_(3)^(-)、SO_(4)^(2-)等成分内混.扬尘颗粒的常见成分是Si,且Si可能与Fe、Ca、Al或者NO_(3)^(-)等成分内混.进一步使用统计学方法提取了对各污染源具有标识作用的谱图特征,以下简称为示踪组分类,并定量化的对比分析了各污染源质谱特征的差异性.扬尘源、机动车尾气源、生物质燃烧源、燃煤源和工艺过程源分别获得52、16、19、23和53个示踪组分类,它们对各污染源原始颗粒总数分别贡献了82.1%、72.8%、64.7%、48.7%和27.2%.本研究可为今后单颗粒源解析算法不确定性评估提供支撑.

There are only few single particle mass spectral profiles of atmospheric particulate sources available to date.In this study,the mass spectrometry data of single particle emitted from industrial process,crustal dust,vehicle exhaust,coal combustion,and biomass burning were collected and clustered via clustering algorithm to investigate the differences in their chemical composition.The results show that particles emission from both industrial processes and coal combustion were dominated by OC,minerals,EC,heavy metals,potassium-rich/levoglucosan/sulfate-nitrate,and OCEC particle types,but the particle number fraction of each particle type in the two sources was significantly different with each other.For example,the proportion of EC particles from coal combustion was significantly higher than that from industrial processes,as totally opposed to heavy metal particles.Biomass burning particles were mainly composed of OC,other organics,K-rich/levoglucosan/secondary,and OCEC particle types.Crustal dust particles were largely composed of mineral particle types.In terms of chemical composition of single particles,in general,OC was mixed with sulfate in the organic particle emitted by industrial processes,while EC was mixed with nitrate or sulfate in the elemental carbon particles.As to coal combustion,EC and OC were often mixed with nitrate and sulfate or with sulfate alone in the carbonaceous particles.OC was mixed with K,levoglucosan,nitrate and sulfate in the organic particles emitted by biomass burning.Dust particles always contain Si,which may be mixed with Fe,Ca,Al or nitrate.Furthermore,in order to distinguish the various emission sources(tracer chemical type for short),statistical method was applied to the measurement data to extract and compare their spectral characteristics.52,16,19,23,and 53 tracer chemical types were obtained for crustal dust source,vehicle exhaust,biomass burning,coal combustion,and industrial processes,respectively,contributed 82.1%,72.8%,64.7%,48.7%,and27.2%of the total particle number of each emission source.Results of this work are capable of reducing the uncertainty of single particle source apportionment by providing real-world source sampling data.