阿達瑪轉換在氣相層析飛行時間質譜儀上的開發與研究

Abstract

本研究嘗試使用脈衝紫外雷射( Nd:YAG, 266 nm) 為光源，以非共振式多光子離子化法( non-resonant multi-photon ionization)，配合阿達瑪轉換-氣相層析/飛行時間質譜儀 ( Hadamard transform-gas chromatography/time-of-flight mass spectrometer, HT-GC/TOFMS)，針對戴奧辛前驅物之一的對-二氯苯進行測量。本實驗首先使用脈衝分子束閥當作氣相層析/飛行時間質譜儀的介面，發現脈衝分子束閥的死腔體積(dead volume)過大，造成管柱層析分離效果不佳，不適合運用在聯結氣相層析/飛行時間質譜儀。之後將氣相層析儀用的分離毛細管(內徑250 m)直接放進飛行時間質譜儀的電極區內，並使雷射保持聚焦在出口端下游 ~ 1 mm之處。為了使真空度能保持在10-5 torr以下，毛細管出口端以高溫火焰燒結封口，再使用2000號的細砂紙，研磨開出直徑50 m的出口，如此可得最佳效果。樣品的進樣方式則使用本研究開發的阿達瑪進樣器，藉由電腦以LabVIEW程式，控制進樣器按照阿達瑪序碼(Hadamard code)，依次將樣品注射到氣相層析儀的氣化室內。阿達瑪進樣器是由脈衝電磁閥所改裝，液態樣品的單一注射量可控制在30 nL ~ 3 mL之間。同日與異日的相對標準偏差值分別為0.56 ~ 1.06%及0.92 ~ 1.67%,%，顯示該進樣器具有絕佳的穩定性。樣品經過雷射激發離子化後，利用箱車波形器設定閘寬（gate width）為200 ns後再連接到類比/數位轉換器，可以單獨擷取並記錄二氯苯的母離子峰訊號。實驗發現將濃度為25.7 g/次的對-二氯苯溶液以一次進樣的方式注射到GC/TOFMS中，僅能得到訊噪比值為1.75的離子峰 ( 1.5 mV)。相對於此，在相同條件之下，樣品以阿達瑪序列的方式，當n = 103與n = 255次時，編碼的層析圖譜經阿達瑪轉換後，訊噪比值分別是7.10與11.58，提高4.05倍與6.62倍。阿達瑪序列n = 103跟n = 255的理論值為5.12跟8.02倍。The non-resonant multi-photon ionization (MPI) method was successfully coupled with a Hadamard transform-gas chromatography/time-of-flight mass spectrometer (HT-GC/TOFMS), for the first time. A well-known dioxin precursor,para-dichlorobenzene, and the fourth harmonic generation (FHG, 266 nm) of a Nd:YAG laser were employed as model sample and an ionization laser, respectively. The findings show that, when a commercially available pulsed supersonic valve (PSV) is utilized as the interface of GC and TOF-MS, the dead volume of the PSV is too large to create a link between the GC with TOF-MS. As a result, the GC chromatography data were not easily identified. A capillary-injector was developed to overcome this. Herein, a piece of a capillary (I.D. = 250 μm; 50 cm in length) was sealed using a high temperature flame, and the sealed part was then polished until the diameter of the outlet reached approximately 50 μm. Following this, the capillary-injector was directly inserted into the electrode region of the TOF-MS, where the laser beam was focused at ~ 1 mm downstream of the capillary-injector. Using the capillary-injector, it was possible to maintain the vacuum of the TOF chamber below 10-5 torr during the GC-TOFMS experiments. On the other hand, a Hadamard-injector, modified from a regular pulse nozzle, was also developed for sample injection. On a portion of the tip of the Hadamard-injector, a piece of a capillary (I.D. = 50 μm; 8 cm in length) was used for the introduction of the sample solution. The injection volume of the pressurized sample solution can be adjusted by changing the background pressure (nitrogen gas), the inner diameter of the capillary, the capillary length and the injection time. The Hadamard-injector can be controlled by a personal computer with the LabVIEW program through a PCI 6221 device, according to a series of Hadamard codes. The RSD (relative standard deviation) values of within-day and between-day were determined to 0.56 ~ 1.06% and 0.92 ~ 1.67%, respectively, indicating that the procedure is stable and reproducible. The sample injection time, volume and split-ratio were investigated in detail during the GC separation experiments. Under optimized conditions, when Hadamard matrices of 103 and 255 were used, the S/N (signal-to-noise) ratios of the signals for para-dichlorobenzene (25.7 μg/single-injection) were substantially improved to 4.1- and 6.6-fold, respectively, and quite matched with those obtained by theory (5.1- and 8.0-fold). Compared to the single injection used in most GC/MS systems, the S/N ratios were substantially improved after inverse Hadamard transformation of the encoded chromatogram.