| 摘要: | 為深入瞭解台灣環境中之重金屬來源與其傳輸過程，本研究建立了一套適用於環境樣品之高精準鍶-釹-鉛同位素比值分析技術，並應用於探討台灣地區的環境物質，包含大氣氣溶膠、河川水體樣品以及固體廢棄物中之金屬來源，分別於第二至第五章中以個別應用案例說明。
論文第二章以高精準鉛同位素技術探討台灣中部地區氣膠PM10中鉛的來源。分析結果顯示鉛同位素搭配金屬元素比值以及ECMWF資料庫，可對PM10的來源以及傳輸更清楚掌握。在低風速期間，PM10中的鉛主要來自境內石油燃燒和煉油廠（48-88%），而燃煤的貢獻最低（< 21%）。在高風速期間，自然源的貢獻從13%增加至31%。儘管鉛僅佔PM10的一小部分，在本研究區觀察到PM10質量與鉛、釩和鋁濃度之間存在高相關性（r = 0.89，p < 0.001，多重線性分析），顯示PM10的化學特徵和鉛同位素應可用於追踪台灣中部PM10的來源。此外，分析境外事件期間PM10的鉛同位素比值證實了大陸氣溶膠長程傳輸的影響，且PM10的化學特徵與境內事件期間的PM10的化學特徵顯著不同。
論文第三章利用鉛同位素探討在東南亞生質燃燒活動發生時(泰國清邁地區之氣膠樣品)，其氣膠化學以及傳輸至鹿林山之化學特性變化。分析結果顯示當清邁氣膠中非海鹽鉀(nss-K+)濃度增加時，鐵濃度與鉛同位素比值皆增加，此現象與文獻中觀測到的生質燃燒會將地殼物質帶入大氣當中形成氣膠同時增加微量元素之溶解度(如鐵)一致。除了生質燃燒之訊號外，本研究也發現清邁氣膠中也含有人為活動之訊號，這些氣膠隨著微量元素濃度增加而具有較低之鉛同位素(206Pb/207Pb)比值，顯示可能受到燃油或其他工業活動之影響。透過此案例之探討，分析不同粒徑的氣膠顯示該方式能進一步解析不同來源之訊號(例如：生質燃燒之影響主要集中在0-0.95 µm顆粒，其餘人為源貢獻集中在0.95-3.0 µm顆粒)。
第四章應用已建立之鍶-鉛同位素分析技術探討受金屬污染之台灣南部河川。透過此二同位素系統之分析可以對河川水體中之金屬來源有更進一步之瞭解，鍶同位素之分析可瞭解水體來源；鉛同位素之分析則可追蹤重金屬鉛(或具親顆粒性之重金屬元素)之可能來源。在阿公店溪流域的調查結果發現，上游河水與工業廢水之鍶同位素比值較高，若搭配鉬/鍶元素比能進一步區別兩者。鉛同位素分析結果顯示上游河水206Pb/207Pb比值較高，而本研究調查之工業污水處理單元廢水則具有較低之鉛同位素比值。透過同位素混合模型搭配可能的端源之同位素比值可估算出各端源對河水中鍶、鉛的貢獻量。模擬估計污染熱點之鍶可能主要來源應為自然源(上游河水)與一具有低87Sr/86Sr比值之工業源排放水所混合，分別佔63~76%以及25~36%，而其他已調查之工業排水佔不到1%。河川中鉛的來源變化較大，來自於上游河水和工業源之比例分別為9-85%和15-91%。時間序列之鉛同位素分析結果顯示降雨期間的高濃度鉛最有可能為自然來源。
第五章應用鍶-釹-鉛同位素系統探討環境固體廢棄物是否具可辨別之同位素訊號。本研究針對台灣地區電弧爐煉鋼廠爐碴及其所使用副料進行同位素分析，以瞭解煉鋼廠鍶、釹、鉛之來源，並建立不同工廠之數據資料。結果顯示，鍶-釹-鉛同位素比值能夠區分三間工廠的還原碴或氧化碴。然而爐碴的同位素比值可能會隨時間而有變化，應與所添加的副料來源有關。進一步分析副料之同位素比值，結果顯示鍶和釹同位素比值應主要受所添加的石灰和矽鐵控制，鉛同位素比值則可與矽鐵或錳鐵較相關，因此這些添加的材料可能是同位素鑑別的關鍵。若能建立各工廠廢棄物及副料之同位素資料庫將能以鍶-釹-鉛同位素比來區分爐碴的來源，在爐碴來源示蹤上極具潛力。
總體而言，本研究證實了以多重同位素示蹤法(鍶-釹-鉛同位素)研究大氣及水體環境重金屬污染來源和傳輸的實用性，於環境鑑識或模擬驗證等方面的皆具潛力。對於未來台灣地區重金屬的污染來源及傳輸過程應能提供重要的參考依據。
;In this study, we have established an analytical protocol for high-precision Sr-Nd-Pb isotope ratio measurements on a variety of environmental materials, including airborne particles, riverine materials, and solid waste. We have successfully used this technique to better constrain metal sources and transport in the highly human-impacted environment of Taiwan. Detailed information on the literature review and methodology are given in Chapter 1.
In Chapter 2, we present a comprehensive approach for tracing possible Pb sources in PM10 in central Taiwan by using chemical characteristics, Pb isotope ratios, and reanalysis datasets (ECMWF). The results suggested that Pb in PM10 was predominantly contributed from oil combustion and oil refineries during the local events (48–88%), whereas the lowest contributions were from coal combustion (< 21%). During periods of high wind speed, the contribution from natural sources increased significantly from 13 to 31%. Despite Pb represented only a small portion of PM10, a strong correlation (r = 0.89, p < 0.001, multiple regression analysis) between PM10 mass and the concentrations of Pb, V, and Al was observed in the study area, suggesting that the sources of PM10 in central Taiwan can be possibly tracked by using chemical characteristics and Pb isotopes in PM10. Moreover, the Pb isotopic signatures of PM10 collected during the LRT event confirmed the impact of airborne pollutants from Mainland China, and the chemical characteristics of the PM10 significantly differed from those collected during local events.
In Chapter 3, we further applied Pb isotope ratios to study the characteristics of aerosols and transport from Chiang Mai (Thailand) to Mt. Lulin (Taiwan) during the Southeast Asia biomass burning season. We analyzed the inorganic compositions (water-soluble ions and trace metals) of size-fractionated aerosols collected at both sites. The chemical and Pb isotopic signatures from biomass burning in Chiang Mai aerosols are characterized by high non-sea-salt K+ concentrations, high Fe concentrations, and high 206Pb/207Pb (208Pb/207Pb) isotope ratios, suggesting greater contributions from crustal materials during biomass burning, consistent with those observed in previous studies. Aside from biomass burning signatures, we also found that aerosols with high contents of trace metals had lower 206Pb/207Pb (208Pb/207Pb) isotope ratios, indicating oil combustion and/or other industrial activities are possible sources of the observed enrichments of trace metals. The analysis of Mt. Lulin aerosols showed that biomass burning from Southeast Asia might be an important aerosol source influencing aerosol compositions at Mt. Lulin, especially for the fine particles (<0.95 µm). The chemical analyses of size-fractionated aerosols demonstrated that more information about sources of aerosols could be obtained through this approach.
Applications of the Sr-Pb isotope ratios for studying metal sources and transport in two highly polluted rivers (i.e. Agongdian River and Ji-Shuei River) in southern Taiwan are further discussed in Chapter 4. Of special interest is that in the Agongdian River, upstream waters are characterized by high Sr and Pb isotope ratios; on the other hand, most effluents collected from wastewater treatment units had low Pb isotope ratios. This makes it possible to estimate relative contributions from natural and anthropogenic sources. Time-series studies were carried out in both river catchments. A simple isotope-mixing model estimated that a major portion of Sr could be originated from upstream waters (63~76%), whereas contributions from industrial sources are relatively small (25~36% and 0.3~0.4% for industrial sources with low 87Sr/86Sr and other industrial effluents, respectively) in the Agongdian River. Contributions of Pb vary significantly, with 9-85% and 15-91% originating from natural and industrial sources, respectively. The results of Pb isotope analyses further suggested that high concentrations of Pb during the rain events were most likely natural Pb derived from crustal materials.
In Chapter 5, we further evaluate the potential of using Sr-Nd-Pb isotope ratios to distinguish sources of solid wastes from the same types of industries. Materials including solid waste (slag) and auxiliary materials were analyzed to further constrain sources of Sr-Nd-Pb in slags from three steel smelting plants. The results show that Sr-Nd-Pb isotope ratios in slags and auxiliary materials have great potential for discriminating reductive or oxidative slags from the three plants. However, isotope ratios of slags can be significantly different over time, most likely due to different sources of auxiliary materials. The measured isotope data of the auxiliary materials indicated that Sr and Nd isotope ratios in slags were mainly controlled by lime and ferro-silicon, and Pb isotope ratios were most likely related to ferro-silicon or ferromanganese, implying that these auxiliary materials are the key to discriminate the origins of the slags. Our results demonstrated that sources of slags can be discriminated by the combined Sr-Nd-Pb isotope ratios, and thus serves as a probe for tracing sources of slags if the databases for plants and associated materials are better constrained.
Overall, this study demonstrates the robustness of using Sr-Nd-Pb isotope ratios to trace sources and transport pathways of metal pollution in a variety of environmental materials, and has great potential for studies in environmental forensics and model validations. The multi-tracer approach developed in this study should provide important information for tracing metal sources and transport pathways in the environment increasingly impacted by human activities. |
