Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/51142
標題: 台灣省產水果酒物理化學品質及以近紅外線光譜技術進行快速分析之研究
Studies on the Physicochemical Quality of Taiwan Local Wines and Rapid Analysis by Near Infrared Spectroscopy
作者: 劉士綸
Liu, Shih-Lun
關鍵字: Fruit wines;水果酒;Near Infrared Spectroscopy;Principal omponent analysis;Physicochemical quality;近紅外線光譜技術;主成分分析;物化品質
出版社: 食品科學系
摘要: 
壹、中文摘要
本研究先就台灣全省地區進行民間釀就情形之調查及瞭解,再自水果產量、種類均高之自苗栗、台中、彰化、南投四縣市中共十四個鄉鎮,蒐集以葡萄、楊桃、梅子為原料製作之釀造與蒸餾酒120種,加上本研究室及自公賣局購入共20種各式釀造與蒸餾酒,總計共72件蒸餾酒與68件釀造酒作為試驗材料,進行物理化學分析及近紅外線光譜掃瞄,配合主成分分析探討各物化特性與水果酒間之相關性。另探討應用近紅外線光譜技術於酒類樣品中各物化特性快速分析之可行性,其結果如下:
1.經調查結果顯示,全省中以中部地區之農民釀酒情形較普遍,所釀製之水果酒種類及產量均較其他地區為大。
2.酒類樣品中各項物化特性之量值可溶性固形物含量在7.5至35.78 %間;酸度含量在0.04至4.04%間;pH在2.83至5.5間;甲醇含量在1.16至3769.1 ppm間;乙醇含量在0至56.46 %間;還原糖含量在0.02至340.17 mg/100g間;總酚含量在0.69至2583.16間;L值在4.82至100.72間;a值在-1.04至59.32間;b值在-0.17至38.35間;酒石酸含量在0至3233.31 mg/100g間;蘋果酸含量在0至1208.35 mg/100g間;抗壞血酸含量在0至310.78 mg/100g間;檸檬酸含量在0至3377.29 mg/100g間;琥珀酸含量在0至1014.75 mg/100g間,顯示所有樣品特別是民間所釀水果酒各物化特性差異極大。
3.於酒類之物化特性來看,蒸餾酒中其還原糖、總酚、酸度、各類有機酸含量均極低,pH、甲醇、乙醇含量均較高,且色澤透明具較高之L值與較低之a值、b值,此些特性可作為區分蒸餾酒與釀造酒間差異之依據。
4.利用主成分分析探討水果酒中各物化特性之相關性,以全部酒樣加入生產方式之類別變異後其主成分分析效果最佳,累積至第二種成分可有55 %之解釋能力。在區分酒樣之能力上則以釀造酒結合水果種類之類別變異效果最佳。
5.將酒類之各物化特性進行近紅外線光譜技術之檢量線製作,於所有140種酒樣中以可溶性固形物、酸度、乙醇、還原糖、L值、a值及b值之檢量線預測效果最佳,其R2值分別為0.99、0.98、0.99、0.99、0.98、0.96及0.97,相關係數r也分別高達0.99、0.99、0.99、0.98、0.97、0.83及0.91。68種釀造酒樣中以可溶性固形物、酸度、乙醇、還原糖、L值及檸檬酸之檢量線預測效果最佳,其R2值分別為0.99、0.96、0.98、0.96、0.93及0.91,相關係數r也分別高達0.99、0.97、0.93、0.93、0.93及0.75;72種蒸餾酒樣中以可溶性固形物、乙醇及總酚之檢量線預測效果最佳,其R2值分別為0.99、0.95及0.94,相關係數r也分別高達0.99、0.85及0.96,顯示這些檢量線已可應用於酒類中物化特性之之快速分析上。
6.選取以利用近紅外線光譜技術獲得可用檢量線之物化特性項目,配合水果種類之類別變異進行主成分分析,鑑別出水果酒原料種類效果更佳,顯示近紅外線光譜技術可應用於快速判別水果酒之種類上。

貳、英文摘要
Abstract
This study first surveyed the situation of wine making in Taiwan. Then totally 140 fruit wine and liquor samples (68 and 72, respectively) selected from Miaoli, Taichung, Nantou and Zhanghwa which in production and abundant variations of fruit wines and liquors (120) and supermarkets which the samples mainly produced by Taiwan Tobacco and Wine Board (20) were used as experimental materials to examine their physicochemical characteristics and near infrared spectroscopic scanning. Using principal component analysis, the correlations between their physicochemical characteristics and the differentiating ability for the types of fruit wine and liquor samples were studied. The feasibility of using near infrared spectroscopy to rapidly determine the physicochemical characteristics was evaluated. The results were summarized as follows:
1.From the results of the physicochemical analysis for all wine and liquor samples, the total soluble solids content ranged from 7.5 % to 35.78 %; the titratable acidity ranged from 0.04 % to 4.04 %; the pH content ranged from 2.83 to 5.5; the methanol content ranged from 0 ppm to 3769.1 ppm; the alcohol content ranged from 0 % to 56.46 %; the residual sugar content ranged from 0 ppm to 3401.7 ppm ; the total phenol content ranged from 0.69 ppm to 2583.16 ppm ;.the L value ranged from 4.82 to 100.72; the a value ranged from —1.04 to 59.32; the b value ranged from —0.17 to 38.35; the tartaric acid content ranged from 0 to 3233.31 mg/100 g; the malic acid content ranged from 0 to 1208.35 mg/100 g; the ascorbic acid content ranged from 0 to 310.78 mg/100 g; the citric acid content ranged from 0 to 3377.29 mg/100 g; the succinic acid content ranged from 0 to 1014.75 mg/100 g. The results showed the great variation in the samples, especially those private-make ones
2.Base on the physicochemical analysis results of all wine sample and liquor samples. The content of residual sugar, the total phenol , the titratable acidity and the organic acid in fruit liquor samples were extremely low while the content of methanol and ethanol were very high as compared to wine samples. The fruit liquor samples had higher Hunter L value and lower Hunter a and b valuethan the fruit wine samples. These physicochemical characteristics could be used for distinguish the fruit wine and liquor samples.
3.Based on the results of principal component analysis (PCA), PCA with wine-making-methods category had the best ability for explaining the variation. The first and second component could explain the sample variation up to 46% and 9%, respectively. PCA with kinds-of-fruit category was the best to distinguish the wine and liquor samples.
4.The R2 of NIRS calibration curves of 140 fruit wine and liquor samples for total soluble solids, titratable acidity, ethanol, residual sugar, L value, a value and b value were 0.99, 0.98, 0.99, 0.99, 0.98, 0.96 and 0.97, respectively. The correlation coefficients (r) for prediction of these 7 constituents were 0.99, 0.99, 0.99, 0.98, 0.97, 0.83 and 0.91, respectively. The R2 of NIRS calibration curves of 68 fruit wine samples for total soluble solids, titratable acidity, ethanol, residual sugar, L value and citric acid were 0.99, 0.96, 0.98, 0.96, 0.93 and 0.91, respectively. The correlation coefficients (r) for prediction of these 6 constituents were 0.99, 0.97, 0.93, 0.93, 0.93 and 0.75, respectively. The R2 of NIRS calibration curves of 72 fruit liquor samples for total soluble solids, ethanol and total phenol were 0.99, 0.95 and 0.94, respectively. The correlation coefficients (r) for prediction of these 3 constituents were 0.99, 0.85 and 0.93, respectively. It showed that these calibration curves could be used for rapidly determining these physicochemical characteristics.
5.Using PCA with category variables and those physicochemical characteristic items having good NIRS calibration curves, the different kinds of fruit for making those wine and liquor samples could be identified. It also indicates that NIRS help can help to identify those wine and liquor samples in this study.
URI: http://hdl.handle.net/11455/51142
Appears in Collections:食品暨應用生物科技學系

Show full item record
 
TAIR Related Article

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.