Abstract
本研究主要以四種標準參考樣品：都市懸浮微粒（NIST SRM 1648），廢棄物飛灰（BCR No.176），土壤（NIST SRM 2709）與燃煤飛灰（NIST SRM 1633b），混合不同比例的模擬樣品，經由ICP-MS實際消化處理，分析其中之化學元素濃度，並經由CMB受體模式計算其可能污染源組成；除了希望藉由實驗驗證分析可信度外，並實際探討在不同分析方法組合下，受體模式的應用性。 經由模式分析所得之結果，我們可將造成污染源解析誤差的因素歸納為三類，即分析方法選取、實驗分析結果和模式本身解析能力；其相對解決方法為增加指標元素、減少分析誤差及合併污染源。假設在不考慮低濃度元素分析上的誤差，共線性問題，及分析元素的數目等因素下，CMB對主要污染源解析誤差應可控制在±10%內，而次要污染源大約控制在±20%內。但在實驗樣品的分析上，由於往往難以控制上述的條件，CMB模式解析主要與次要污染源的誤差或能分別保持在±20%與±50%，應為可接受之合理範圍。 本研究亦嘗試直接利用LA-ICP-MS量測空氣樣品及污染源樣品，所得之質荷比譜圖直接經由指紋比對的方式，應用在CMB受體模式的分析上。模式分析之結果顯示以LA-ICP-MS分析空氣粒塵樣品應用在受體模式是可行的。Four different standard reference materials (SRM), includingurban particulate matter (NIST SRM 1648), city wasteincineration ash (BCR No. 176), San Joaquin soil (NIST SRM 2709)and coal fly ash (NIST SRM 1633b), were used to prepare standardfilter samples for testing the effectiveness of receptor modelin air pollution studies. Mixed standard filters were digestedand analyzed to determine the concentrations of over twentyelements. Based on the obtained chemical data, the CMB receptormodel analysis was performed to identify the compositions ofsimulated pollution sources on the filters.Capabilities of thereceptor model in real sample analysis is therefore examined.It was found that factors to cause the analysis error can beclassified into three categories: (a) the selected analyticalmethods; (b) fluctuations of analytical results; and (c) co-linearity problem of the CMB Model. Several solutions areproposed to reduce such errors, including the identification of□arker elements", improvement of the analytical fluctuationsand conciliation of pollution sources. Experimental resultsshow that only a maximum of 20 % errors can be achieved for themajor pollution sources ( > 25%) and 50% for minor pollutionsources (<10%).The feasibilities of laser ablated inductivelycoupled plasma mass spectrometry (LA-ICP-MS) analyticaltechniques applied for receptor model analysis were alsodeveloped. It demonstrated that the direct mass spectraobtained from air samples as well as pollution sources can beemployed for CMB receptor analysis.