| 摘要: | 工業區有害空氣污染物（Hazardous Air Pollutants, HAPs）對民眾健康的威脅是近年受關注的議題，行政院環保署2021年2月訂定公布「固定污染源有害空氣污染物排放標準」，優先納管22項有害空氣污染物。為了掌握揮發性有機HAPs較長時間的平均濃度，以評估長時間的暴露健康風險，開發一套類似於美國325A/B之擴散式採樣技術。為了計算樣品濃度，須求取該物質之專屬吸取速率（Uptake Rate）才可定量HAPs，且吸取速率依吸附劑種類與個別目標物質皆有其專屬值，故須個別測定。在樣品分析上採用自動化熱脫附儀搭配氣相層析質譜儀作為分析系統。希望以更簡易的採樣方法長時間監測周界HAPs濃度。
本研究選擇Carbopack X以及Carboxen 569吸附管，為了求取個別物質的吸取速率，建立暴露腔系統以模擬擴散式採樣的環境，首先驗證現行U.S. EPA Method 325B方法中的18項已知HAPs的吸取速率，差異小於11%的結果確認本研究設備與方法之可靠性。再者，利用暴露腔系統求得包含氯乙烯、1,3-丁二烯、丙烯?等HAPs之吸取速率，補足現有標準方法之不足，強化台灣針對優先管制物的監測技術。
本研究除了建立品質管制方法之外，為了瞭解擴散式採樣性質，經由實驗測試採樣時可能的干擾或影響，探討包括溫度與濕度變化、反擴散採樣現象、物種間競爭吸附的現象等，考量吸附劑選擇性，建議兩吸附管各別適合採集的物種。本研究結果顯示Carbopack X與Carboxen 569共可針對65項HAPs進行擴散式採樣之監測，且無須考量物種之間的競爭吸附干擾。Carbopack X可針對包含1,3-丁二烯等共52項有機HAPs進行擴散式採樣，Carboxen 569則可針對包含氯乙烯等共44項有機HAPs進行擴散式採樣。Carbopack X檢量線相關係數介於0.993至1.000之間。Carboxen 569檢量線相關係數介於0.990至1.000之間。兩者除了二氯甲烷之外皆有小於1 ppbv的方法偵測極限，分別介於0.11 ppbv至0.97 ppbv之間以及0.47 ppbv至1.11 ppbv之間。樣品保存回收率分別介於89%至110%之間以及91%至113%之間。目標物濃度變化之反擴散回收率分別介於93%至115%之間以及介於90%至117%之間。適用於溫度23°C至38°C、相對濕度23%至90%的環境條件，其吸取速率變化量RSD%介於5%至17%之間，Carboxen 569之吸取速率變化量則RSD%介於5%至19%之間。
本研究尚探討時間線性之特性，在較長採集時間下，吸附管中含有較大量的物質時是否影響既有的採集行為，導致失去時間線性，這將可能影響定量結果。然而，現今在標準方法中僅有少數物質有相關的參考值。本研究觀察到Carbopack X採集1,3-丁二烯的吸取速率下降14%，這個結果與ASTM D6196描述採集7天後吸取速率下降12%的結過相似。藉由實驗得知擴散式採樣採集量與吸取速率變化之關係，推估更多物質的適用採集質量範圍，提供更多使用資料以建立完善的擴散式採樣技術。
實驗室評估完成後，本研究於北部某工業區進行一個場次的實地監測，以驗證擴散式採樣技術之實用性，時間為2019年11月12日至11月19日。實場的測試結果與線上熱脫附氣相層析質譜分析方法的平均值相當，擴散式採樣採集苯的結果為0.31 ppbv，與線上熱脫附氣相層析質譜分析方法0.27 ppbv的結果相近，同樣在其他物種有相近的結果，甲苯的結果分別為6.13 ppbv與5.37 ppbv，乙苯的結果分別為0.26 ppbv與0.24 ppbv，間/對二甲苯分別為0.40 ppbv與0.36 ppbv。擴散採樣方法不僅可以輕鬆地部署採樣測點，還可以在相對較長時段內提供周界之目標HAPs平均濃度，提升環境監測實力。
;Hazardous Air Pollutants (HAPs) are closely associated with public health issue, which has received much attention in recent years. In 2021, Taiwan EPA announced the Emission Standards for Hazardous Air Pollutants from Stationary Pollution to control 22 HAPs as the first stage. In order to obtain average concentrations over a relative long period of time for these target HAPs, a diffusive sampling technique similar to the U.S. EPA Method 325A/B method was developed. To calculate sample concentrations diffusive uptake rates are required for individual target compounds. For sample analysis a thermal desorption (TD) technique for preconcentration coupled with gas chromatography mass spectrometry (GC-MS) was employed.
In this research, we use both Carbopack X and Carboxen 569 sampling tubes. The uptake rates of the 22 target compounds including 1,3-butadiene and vinyl chloride which are not covered by the U.S. EPA Method 325A/B were determined in this study in a self-built exposure chamber. We found that the uptake rates were in high agreement with those determined by the U.S. EPA Method 325B using Carbopack X based on two repeated results, and the uptake rates were between -4% to 13% for Carbopack X and -13% to 7% for Carboxen 569.
Both types of sorbents were tested for optimized conditions for the 22 HAPs from the perspectives of correlation coefficients (R2) and method detection limits (MDL), as well as other relevant factors in diffusive sampling, such as the effect of humidity, back diffusion, tolerance to high ambient concentration variability, etc. Our results show that Carbopack X can be used to sample 55 HAPs including 1, 3-butadiene, with R2 ranging from 0.993 to 1.000 and MDL ranging from 0.11 to 0.97 ppbv. The relative standard deviations (RSD) in uptake rates are all lower than 20%, between 4% and 17%. The recoveries are between 91% and 116%. On the other hand, Carboxen 569 can detect 44 HAPs including vinyl chloride, with R2 ranging from 0.990 to 1.000 and MDL ranging from 0.47 to 0.95 ppbv. The RSD of uptake rates are between 5% to 19%, and the recoveries are between 87% and 117%.
After completion of laboratory assessment, one-week field measurement was conducted from November 12, 2020 to November 19, 2020. The field measurement of diffusive sampling showed comparable results to TD-GC/MS measurements by averaging hourly data from online TD-GC/MS. For instance, benzene was detected as 0.31 ppbv and 0.27 ppbv for diffusive sampling and on-line method, respectively. Toluene and ethylbenzene were measured as 6.13 ppbv vs. 5.37 ppbv, and 0.26 ppbv vs. 0.24 ppbv, respectively. This diffusive sampling method not only can make sample deployment with ease, but also provide concentration estimates for target HAPs over a relatively long time period of more than 2 weeks. |
