現今大氣中VOCs的監測方法大多以固態吸附劑做為捕捉媒介，並藉由熱脫附方式將樣品送至氣相層析系統中進行分離及偵測，然而熱脫附行為往往在第一時間左右了層析結果的層析峰形及解析度，提高了VOCs監測方法在定量上的不確定度。而本實驗利用丁式切換器對熱脫附峰進行切片檢查，以診斷影響層析峰形之變因，瞭解熱脫附行為，並拓廣丁式切換器的應用層面。Heart-cut系統架構以丁式切換器為核心，預管柱選用無滯留能力的毛細管空管，分析管柱則為DB-1，藉由丁式切換器的切換，可控制預管柱沖堤出之樣品流往空管或DB-1，藉以達到樣品切片之目的。 VOCs未經任何分離作用即到達偵測器所測得之峰形，在此稱之為熱脫附峰，此亦代表著物種熱脫附行為的輪廓，實驗中發現VOCs的熱脫附峰呈現拖尾情形，初步推測此乃系統間的無益體積所造成，經由該切片技術對無益體積所增長之樣品帶進行分流後，樣品拖尾現象獲得大幅改善，然而卻亦意外發現被分流丟棄之部分似乎大多為重碳成分，隨後亦進一步針對VOCs的熱脫附峰進行精細切片診斷。診斷結果發現熱脫附峰各區段之物種呈現非均勻分佈趨勢，且其後端拖尾部份主要為重碳物種所貢獻。根據熱脫附峰之診斷報告，吾人認為物種分佈之非均性可能為物種傳送速率之差異所致，為驗證此闡釋，在此亦進一步針對樣品傳輸管之各項參數進行一連串最佳化探討。 最佳化之傳輸管條件為79.5 cm x 0.32 mm i.d.之毛細管空管，且必須另以加熱帶及溫度讀取裝置將溫度控制在200 °C左右。最佳化系統之RSD值可控制在0.66%〜2.00%；線性關係皆在0.999之上；而對稱因子與解析度表現相較於傳統系統，則分別有0.30%〜16.85%及10%左右的提升。本實驗使用切片技術成功達到診斷熱脫附行為之目的，並由實驗結果獲得之資訊進一步對系統進行改良，最終之最佳化系統亦皆有優於傳統系統之層析表現。 Using sorbent traps coupled with GC techniques has become the most common method for ambient VOC measurements. However, the peak shape and resolution is mainly affected by the process of thermal desorption of the sorbent trap. This might induce the uncertainties in quantification. In this study, a Deans switch heart-cut system is used to study the phenomenon. It is equipped with a deactivated capillary column, and a DB-1 column as both the precolumn and the analytical column. By controlling the Deans switch, the analytes can be routed to either the the DB-1 or the deactivated column. In this study, the desorption of VOCs from the trap without being separated by any column is termed a thermal desorption (TD) peak. The TD peak can represent the profile during thermal desorption of a sorbent trap. Pronounced peak-tailing of the TD peak was found, and was possible arising from the dead volumes in the analytical system. The TD peak was sliced into six portions to be separated by the DB-1 column to the diagnose any discrimination with the VOC composition during the TD process. As an important and related issue of the TD process, the directions of thermal desorption of a sorbent trap were also investigated. Surface condition of a transfer line after TD was also found to determine to some degree the transfer rates of different species, particularly the heavier compounds. Hence, different types and conditions of the transfer line are also discussed in our research. The optimum transfer line was found to be the deactivated capillary column (length = 79.5 cm, 0.32 mm ID) heated at approximately 200°C. The RSDs are within 0.66-2.00% with R2 no less than 0.999. A novel application of the Deans switch has been developed by applying the heart-cut technique to diagnose the TD process. This peak-slicing technique not only can exam the TD profile of a sorbent trap, it also further broadens the application aspects of the Deans switch.
