| 摘要: | 本研究之主要目的為量測台中市都會區大氣超細微粒(<0.1 ?慆)質量濃度(PM0.1)、粒徑分佈與無機成分，進而初步探討PM0.1之可能排放源。2009年11月至2010年2月的冬季期間，本研究於台中市都會區大新國小，約20 m高之頂樓使用奈米微孔均勻沉降衝擊器(Nano-MOUDI)進行每週2次連續48小時之大氣超細微粒採樣。採獲之23個微粒樣本進行萃取後，以離子層析分析8種離子及感應耦合電將質譜儀分析10種金屬元素，最後依據PM0.1之無機化學組成，利用濃縮因子、因子分析與正矩陣因子法鑑別其可能排放源。
研究結果顯示，PM0.1平均值與標準差為3.53±0.98 μg/m3，且佔PM1、PM2.5、PM10之百分比分別為14.2%、11.5%、8.2%，此濃度值與百分比高於國內外其他研究，顯示台中都會區有相當程度之大氣超細微粒污染問題，且此問題可能源自於鄰近之排放源。此外，PM0.1與PM1、PM2.5與PM10之相關係數分別為0.23、0.22、0.21；反觀，後三者間呈現非常高的正相關(r > 0.9) ，顯示PM0.1傳輸與排放特性不同於其它粒徑微粒，可能是因為超細微粒在時間、空間尺度上皆相較為短小。大氣懸浮微粒平均質量濃度粒徑分佈多呈現雙峰分佈，其波峰位於0.51±0.13 ?慆 (累積性微粒粒徑區間)及6.10±0.76 ?慆 (粗微粒粒徑區)，而超細微粒粒徑區間內無明顯波峰存在，其質量濃度與百分比隨著粒徑減小而減少。PM0.1化學組成中，45.8%為無機鹽類，5.4% 為金屬元素，而剩餘48.8%之未知化學物種可能包括無機碳與有機碳。受體模式正矩陣因子法結果顯示，台中都會區PM0.1可能的排放源及貢獻量依序為道路揚塵(Mg、Sr、Al、K; 51.6%)、衍生性微粒(SO42-; 17.9%)、交通排放源(Sb、V、SO42-、K; 16.6%)及營建工程(Cr、Fe、Zn; 13.8%)。
本研究首度在台中都會區量測大氣超細微粒之質量濃度粒徑分佈與無機成分，其結果將有助台灣本土資料之建立，進而填補相關之知識缺口。雖然樣本數有限，本研究仍為未來後續相關研究之重要參考，包括微粒生成機制、排放源鑑別與控制策略擬定等。最後，本研究建議未來工作可考量擴大採樣期間以涵蓋夏季、採取不分徑之超細微粒採樣策略以縮短採樣時間、及增加微粒化學組成之量測以利排放源之鑑別。
The primary objective of this study is to characterize the mass concentrations, size distributions, and inorganic composition of wintertime ultrafine particles (UFPs; < 0.1 ?慆) in Taichung urban air. The study took place in the Taichung urban area in central Taiwan from Oct 2009 to Feb 2010. The sampling site is located ca. 20-m above ground level on the rooftop of Da-Shin Elementary School. A nano micro-orifice uniform-deposit impactor (Nano-MOUDI) was used to classify and collect ambient particulate matter (PM) into 9 size fractions on Teflon filters over a period of 48-hr, and twice a week. A total of 23 sets of samples were obtained and analyzed for 8 water-soluble ions and 10 acid-extractable metals using IC and ICP-MS, respectively. The resulting composition data were then used in enrichment factor calculation, factor analysis, and positive matrix factorization for preliminary screening for influential sources of UFPs.
The results show that the UFP mass concentrations (PM0.1) averaged 3.53±0.98 μg/m3, making up 14.2%、11.5%、8.2% of PM1, PM2.5, and PM10, respectively. This suggests that the UFP pollution in the Taichung urban area is comparable to or worse than several urban areas in Taiwan and the US. In addition, the high contribution to PM indicates that the UFP sources are of local origin. The correlation coefficients between PM0.1 and PM1, PM2.5, and PM10 are 0.23 0.22, and 0.21, respectively; on the other hand, the latter three have very strong positive correlations (r > 0.9). This indicates that the UFPs may have different transport characteristics and emission sources than PM of larger sizes. More specifically, the UFPs are likely to have shorter spatial and time scales than larger particles. The size distributions show that ambient PM consists of two major modes: one in the accumulation mode (0.51±0.13 ?慆) and the other in the coarse mode (6.10±0.76 ?慆). In addition, there is no obvious mode within the UFP size range, where the mass concentration decreases with decreasing particle size. The chemical composition data show that the determined inorganic ions and metals contribute 45.8% and 5.4% of the UFP mass, respectively. The remaining 48.8% were not chemically resolved, which are likely to consist of inorganic and organic carbon. Receptor models show that the major emission sources of UFPs in Taichung urban air, in descending contribution, are road dust (Mg, Sr, Al, K; 51.6%), secondary formation (SO42-; 17.9%), vehicle emission (Sb, V, SO42-, K; 16.6%), and construction work (Cr, Fe, Zn; 13.8%).
The present study was the first to characterize the mass-size distribution and inorganic composition of UFPs in Taichung urban air. The results are expected to complement existing database and to bridge current knowledge gap with regard to Taiwan ambient UFPs. Although with limited sample size, the present study serves as an important starting point for subsequent related studies, including aerosol formation mechanisms, source apportionment, and control strategy development. Recommended future work includes expanding the study duration to include summer season, shortening the sampling duration by bulk sampling, instead of size-segregated sampling of UFPs, and analyzing an increased number of chemical species in UFPs to improve the accuracy of source apportionment. |
