| 摘要: | 有害空氣污染物 (Hazardous Air Pollutants, HAPs) 隨著工業發展被大量的排放至環境空氣中，對於人體的健康造成極大的危害。在國際間現行的HAPs中屬於揮發性有機物 (VOC) 的普遍分析方法為U.S. EPA TO-15，台灣環保署參考此方法訂定相關標準方法NIEA A715.16B，使用採樣罐採集樣品，隨後將樣品帶回實驗室進行分析，稱為離線式方法，其數據無法即時反應環境濃度變化。本研究延續前人研究使用線上熱脫附氣相層析質譜技術 (On-line Thermal Desorption GC-MS，簡稱On-line TD-GC-MS)，在定點直接連續進樣分析，可以偵測空氣中數十種VOC濃度變化，改善離線式採樣的樣品保存、數據代表性不足、耗費人力之缺點。
根據以往經驗，質譜儀應用於線上分析時離子源劣化速度快，分析周界樣品四至五天感度會下降至80%，導致定量誤差，此時需要將離子源進行替換，更換過程需耗費數小時至十幾小時，造成連續數據中斷。為了改善離子源快速劣化使儀器頻繁停機的問題，觀察長期監測後的離子源表面附著髒污，推測空氣中複雜的基質及水氣容易造成離子源表面上之髒污，進而使其快速劣化。為了減緩離子源汙染速度，本研究針對質譜參數進行測試，藉由質譜燈絲電流強度的調整，使降低樣品離子化程度，以有效減緩離子源汙染速度。研究結果顯示將電流從50 µA降低至15 µA，可有效延長離子源偵測壽命長達至15天以上，降低游離源維護頻率。
利用調控後的參數建立相關品保品管測試，檢量線相對標準偏差 (RSD) 介於0.226~7.204%、R2介於0.993~1.000、精密度RSD介於1.20～11.16%、回收率介於75.95～94.11%、方法偵測極限所有物種皆小於0.1 ppb。使用此系統參數於台灣北部某工業區進行六個月的實場監測，利用內標準品觀察儀器穩定性及感度衰退趨勢，監測期間儀器運轉穩定性皆大於80%，離子源平均壽命長達14天，單次離子源壽命最長可到21天，大幅延長儀器運轉時間，並降低維護保養次數，減少因維護保樣造成的數據損失，對於線上連續監測方法產生十分重要的正面影響。
;Hazardous air pollutants (HAPs) are extensively emitted into the ambient air due to industrial development, posing significant risks to human health. The most commonly used analytical method for toxic volatile organic compounds (VOCs) as a class of HAPs is the standard method of U.S. EPA TO-15. Taiwan′s Environmental Protection Administration (EPA) has adopted a similar standard method, i.e., NIEA A715.16B, based on the revision of TO-15. Both methods are offline types which involve sampling VOCs using treated stainless canisters that are brought back from the field to the laboratory for analysis. These offline techniques do not provide real-time data to reveal the instantaneous variation of analytes in ambient air.
In this study, we build upon previous research and utilize online thermal desorption gas chromatography/mass spectrometry (TD-GC-MS) technology. This method enables direct and continuous analysis at fixed locations, allowing direct near real-time analysis of numerous VOCs in the air. The online TD-GC-MS method greatly improves the drawbacks of offline methods in terms of sample preservation, data representativeness, and the.
From the past experience, when MS are used for online analysis, the ion source could degrade rapidly by as much as 80% in sensitivity in 4 days. Maintenance usually takes several hours, causing interruptions in continuous data collection. Observations have revealed that the surface of the ion source becomes contaminated with presumably oxides and carbon residues after a period of usage. Moisture and oxygen in the air are likely the cause for contamination, leading to rapid degradation.
To reduce the rate of contamination of the ion source, this study conducted tests on the optimization of MS parameters. By adjusting the emission current of the ion source, which reduces the intensity of ionization, the contamination rate can be effectively reduced. When reducing the emission current from 50 µA to 15 µA, the effective runtime of the ion source can significantly extend to 15 days or more, thereby greatly reducing the frequency of ion source maintenance.
By using the optimized parameters, the quality assurance/control (QA/QC) results are as follows: the relative standard deviation (RSD) of the calibration curves ranging from 0.226% to 7.204%, linearity (R2) ranging from 0.993 to 1.000, recoveries between 75.95% to 94.11%, the precision (RSD) results between 1.20% to 11.16%, and MDL for all species between below 0.1 ppb. All the QA/QC results complied with the requirements set by the NIEA A715.16B method.
With the optimized parameters, field monitoring was conducted for six months. The instrument′s stability and sensitivity trend were closely monitored using internal standards. Throughout the monitoring period, the instrument′s data completeness exceeded 80%. The average runtime of the ion source was 14 days, with the most extended single continuous runtime reaching up to 21 days. This study significantly prolonged the instrument′s runtime, reduced maintenance frequency, and minimized data loss caused by maintenance. These results have a positive impact on the online continuous monitoring method. |
