高粱酒為台灣國產量前三大酒類,市售高粱酒數不勝數,價格更是高低不一,消費者在購買時往往只能依據酒廠的知名度,而無其他依據。因此,建立完整的高粱酒指紋圖譜,將有助於民眾選購。此外,2011年塑化劑案件顯示出針對特定化合物做偵測無法得到樣品的完整資訊,錯失發現違法添加物的機會。 本研究利用氣相層析質譜儀以全掃描方式鑑定高粱酒中成分,藉由其指紋圖譜區別高粱酒廠家,並利用氣相層析火焰離子化偵測器分析酒中甲醇含量。高粱酒的成分鑑定利用氣相層析質譜儀,並將分析物分成非揮發性及揮發性兩部分,第一部分我們利用分散式液液微萃取以及減壓濃縮方式,將高粱酒非揮發物分成鹼性/中性、酸性和水溶性三種,定性成功的分析物共有74種,利用判別分析以1,3,3-三乙氧基丙烷(propane,1,1,3,triethoxy-)、己酸(hexanoic acid)、異丁醛二乙縮醛(propane,1,1-diethoxy-2-methyl-)、辛酸乙酯(octanoic acid,ethyl ester)、異戊醛二乙縮醛(butane,1,1-diethoxy-3-methyl-)五種分析物區分六家廠家的正確率為100%。第二部分採用吹氣捕捉方式求得高粱酒中揮發性物質,共測得50種化合物,以乙縮醛二乙醇(ethane, 1,1-diethoxy-)、乙醛(acetaldehyde)、異丁醇(1-propanol, 2-methyl-)、丁酸乙酯(butanoic acid, ethyl ester)、3-甲基丁醛(butanal, 3-methyl-)進行判別分析正確率為94.12%。實驗結果發現利用非揮發物鑑定高粱酒品牌的結果較為準確。甲醇以氣相層析火焰離子化偵測器偵測,方法偵測極限為0.34 ppm,全數樣品皆小於1000 ppm,回收率落於80%~120%,符合公告標準。本研究成功地分析高粱酒中成分,並證實特定分析物以區別品牌之可行性。
The production of kaoliang wines in Taiwan is on top three among a wine industry. Varieties of kaoliang wines are available in market with a varying price. Hence, it became difficult to choose among them. This motivated us to develop a method to obtain kaoliang wine’s fingerprint profile. Moreover, in 2011, the Taiwan food scandal demonstrated that detecting specific compounds alone from samples can’t give complete information. We may miss the chance to discover illegal additives. In this study, we develop gas chromatograph-mass spectrometer method to identify components in 31 kaoliang wines and differentiate 6 brands by their fingerprint profiles. The methanol content of wine was analyzed by gas chromatograph-flame ionization detector. In GC-MS part, the compounds were separated into nonvolatile and volatile. The nonvolatile analytes were divided into three components viz basic/neutral, acid and water soluble by dispersive liquid-liquid microextraction and rotavapor techniques. Qualitative analysis of 74 analytes was successfully carried out. 1,1,3,triethoxypropane, hexanoic acid, 1,1-diethoxy-2-methyl propane, ethyl octanoate, 1,1-diethoxy-3-methylbutane were able to differentiate wine brands and the classification correctness was found to be 100%. We were able to identify 50 volatile components by using purge and trap technique. The classification correctness was found to be 94.12% using 1,1-diethoxyethane, acetaldehyde, 2-methyl-1-propanol, ethyl butanoate, 3-methylbutanal as variables. The results showed nonvolatile components have better ability to differentiate brands. Finally, GC-FID was used to quantify methanol concentration. The method detection limit was 0.34 ppm and recovery was within 80%~120%. This study has successfully analyzed the components in kaoliang wines and confirms the feasibility to distinguish brands of wine by specific analytes.