懸浮微粒 (Particulate Matter)是重要的空氣汙染指標物，因其容易吸附空氣中其他污染物，並作為光化反應的界面形成二次有機污染物，所以具有種類多樣且含量不一有機成分，若能得知懸浮微粒的化學組成勢必對於了解懸浮微粒的排放來源及形成機制有所幫助，本研究之目的即是透過新穎的分析技術對懸浮微粒的組成進行定性分析，期望藉由定性結果推論出成分的排放來源以及可證明光化反應機制的關鍵化合物。 以往使用一維GC-MS分析懸浮微粒時，容易發生大量層析峰共析的現象，影響質譜定性成效，本研究利用具高解析能力之GC×GC-TOF MS系統有效地分離高複雜度的懸浮微粒組分。其中GC×GC的核心使用了市售的氣流型調制器，有別於常規冷凍型調制器，在無需消耗大量冷劑的情況下，透過氣流壓力的調整與流向的切換達到與冷凍型調制器相仿的調制、聚焦效果，可以應用於野外線上監測。而本研究使用的TOF MS作為定性偵測器除具有高資料擷取速率 (50 Hz以上)以完美呈現GC×GC圖譜外，還可耐受約5 mL/min的載流流速而不致破除真空，提升了進入TOF MS端的分析樣本分流量，增加層析峰感度。 本研究使用HPLC級溶劑，降低溶劑雜質可能產生的干擾，以超音波震盪輔助溶劑萃取收集在石英濾紙上的懸浮微粒，透過吹氮獲得濃縮後的萃取溶液，以液態注射方式注入GC×GC-TOF MS進行定性分析。萃取懸浮微粒中有機成分的方經最佳化測試後，決定使用1/2張面積為8” × 10”之高通量採樣濾紙為萃取樣本，以甲苯/丙酮等體積混合溶劑進行萃取。為確認分析結果的有機成分確實源自懸浮微粒，萃取懸浮微粒採樣濾紙的過程中，予以相同條件同時萃取實驗室空白濾紙，將其分析結果當作背景值排除污染及干擾，作為實驗方法檢驗，避免溶劑雜質及分析系統殘留污染導致誤判。 此外，透過添加直鏈烷烴標準品可見分析物是懸浮微粒中沸點範圍介於十二烷至三十烷之間的半揮發性有機成分(Semi-Volatile Organic Compounds，SVOCs)。無法分析低沸點化合物的主要原因是使用甲苯作為萃取溶劑，因其沸點高，在濃縮過程中為移除甲苯，勢必連同沸點較低的有機物一併移除。 分析高通量採樣之PM10及PM2.5的GC×GC-TOF MS分析結果中，排除烷類化合物後初步定性的成分數量分別有45種和56種，且多為醇、醛、酮和酯類等含氧化合物，又以直接排放、未經光化反應的一級有機氣膠 (primary organic aerosol，POA)為主，其中也包含較為特殊的成分如；塑化劑及有機磷阻燃劑等人造化合物，若能作為具有代表性的排放標記物 (Marker)，就能藉此回溯該污染物的排放源，並進一步實施排放管制以期降低此類人造化合物對環境的影響。 ;The presence of organic compounds on the particular matter (PM) or aerosols can be arisen from the absorption of gaseous organic compounds on the existing aerosols, or from organic precursors to form secondary organic aerosols (SOA) through photochemistry. As a result, knowing the chemical composition of aerosols can shed light to the understanding of particle formation and source characteristics. The objective of this study is to characterize organic constituents on aerosols relevant to their emission sources and the key compounds revealing the evolution of aerosols with the use of a novel analytical technique. The conventional GC-MS technique has been used to analyze the organic composition of aerosols. However, the high organic complexity of aerosol samples often renders GC-MS technique unsatisfactory in terms of compound separation and identification. As a result, time-of-flight mass spectrometry (TOF-MS) coupled with comprehensive two dimensional GC (GC×GC-TOF) was preferred. A flow type of modulator instead of a thermal type was used in GC×GC as a prelude to field applications without the need of cryogen. The high data sampling rate of 50 Hz for the TOF MS is pivotal to produce very detailed and reproducible GC×GC results. The tolerance of high carrier gas flow rates of up to 5 mL/min played the key role to achieve high sample throughput and thus better sensitivity. We used ultrasound assisted extraction to extract PM samples collected on quartz filter papers. The solution extracted was then purged by nitrogen to reduce volume before GC injection. Various organic solvents were tested and the combination of acetone and toluene was found to be most suitable to achieve optimal extraction. HPLC grade solvents were used to minimize background interference. Both solvent and filter blanks were made to determine the background contribution to the sample results. The GC×GC results have been obtained by analyzing both PM10 and PM2.5 samples collected by high-volume samplers. By spiking with a known amount of long-chain alkanes as the markers of molecular size, we found that the majority of the organic analytes were in the range of 12 - 30 carbon numbers falling in the category of semi-volatile organic compounds (SVOCs). If excluding alkanes, 45 and 56 compounds of alcohol, aldehyde, ketone, and ester varieties were able to be tentatively identified for the PM10 and PM2.5 samples, respectively, which are mostly primary organic aerosols (POA). Intriguingly, trace amounts of plasticizers and phosphorus flame retardants were also found. Compounds such as these are unique to the specific sources, called markers, demonstrating the wide spread of these hazardous compounds in the environment.
