【作者】 张莉；

【导师】 王华； 王少芳；

【作者基本信息】 西北农林科技大学 ， 葡萄与葡萄酒学， 2012， 博士

【摘要】 近几年随着我国进口葡萄酒的增加，澳大利亚已经成为中国第二大葡萄酒进口国。本文旨在研究西澳葡萄酒的质量特征，用于对其风格特色的准确鉴别。本试验以澳大利亚西澳36款干红葡萄酒和6款干白葡萄酒为研究对象，采用高效液相色谱法（highperformance liquid chromatography，HPLC）和固相萃取-气相色谱-质谱联用技术（solidphase extraction gas chromatography-mass spectrometry，SPE-GC-MS）对不同品种、原产地及年份葡萄酒的花色素苷、非花色素苷酚类物质及香气成分进行系统、深入的比较研究，并结合多元统计方法建立葡萄酒类别的判别函数，以期实现葡萄酒的真伪判别。主要研究结果如下：（1）通过分析5个单品种和4个混酿葡萄酒，研究葡萄酒中总酚、花色素苷、黄烷醇、酚酸、黄烷酮、反式-白藜芦醇及香气成分间的品种差异。西澳红葡萄酒总酚浓度范围为0.99-2.56g/L，白葡萄酒是0.14-0.38g/L，前者的平均值是后者的近十倍。红葡萄品种间及白葡萄品种间总酚含量差异不显著（P＞0.05）。不同品种红葡萄酒3种花色素苷含量关系均为二甲基花翠素-3-O-葡萄糖苷＞花翠素-3-O-葡萄糖苷＞花青素-3-O-葡萄糖苷，并且花翠素-3-O-葡萄糖苷和花青素-3-O-葡萄糖苷含量呈线性显著正相关（P＜0.05）。白葡萄酒中均未检测到这3种花色素苷。不同品种葡萄酒中没食子酸、丁香酸、反式-白藜芦醇、槲皮素、杨梅素的含量以及儿茶素/表儿茶素比值存在显著差异（P＜0.05）。西澳干红葡萄酒和干白葡萄酒香气成分浓度差异较大，但同类葡萄酒间差异相对较小。己酸乙酯、辛酸乙酯、乙酸异戊酯、乙酸己酯以及三种脂肪酸（辛酸、癸酸和己酸）可用于干红和干白葡萄酒的区分。戊酸乙酯、顺式-3-己烯-1-醇可用于区分赤霞珠和西拉葡萄酒。乙酸异戊酯、异戊醇、己醇、顺/反式-3-己烯-1-醇、苯甲醇、苯乙醇、丁酸乙酯、戊酸乙酯可用于区分赤霞珠和黑比诺葡萄酒。戊酸乙酯、异戊醇、癸酸、苯甲醛、苯乙醇、月桂酸乙酯、己酸乙酯、乙酸己酯、顺/反式-3-己烯-1-醇、香叶醇可用于霞多丽和长相思/赛美蓉干白葡萄酒的区分。（2）通过分析Frankland River、Geographe及Margaret River三个产地的赤霞珠葡萄酒和Frankland River、Margaret River两个产地的西拉葡萄酒，研究产地对葡萄酒中花色素苷、黄烷醇、酚酸、黄烷酮、白藜芦醇及香气成分的影响。产地对葡萄酒中三种花色素苷无显著影响（P＞0.05），而对非花色素苷单体酚有显著影响（P＜0.05）。酚酸、白藜芦醇、杨梅素、表儿茶素的含量均存在显著性差异（P＜0.05），可以作为鉴别西拉和赤霞珠不同产地葡萄酒的指标。戊酸乙酯、反式-3-己烯-1-醇、顺式-3-己烯-1-醇、辛酸乙酯、癸酸乙酯、月桂酸乙酯、棕榈酸乙酯、香叶醇、异戊醇、辛酸及苯乙醇浓度在3个产地赤霞珠葡萄酒中有显著性差异（P＜0.05）。己酸乙酯、辛酸乙酯、癸酸乙酯、月桂酸乙酯、棕榈酸乙酯、己醇、香叶醇、己酸和苯甲酸在Frankland River西拉葡萄酒中浓度均显著高于Margaret River西拉葡萄酒（P＜0.05）。乙酯类化合物在葡萄酒产地分类中有非常重要的作用，其中癸酸乙酯和棕榈酸乙酯可以区分不同产地赤霞珠和西拉葡萄酒。（3）通过分析7个年份（2002-2008年）的赤霞珠葡萄酒及4个年份（2004-2007年）的西拉葡萄酒，研究年份对葡萄酒中总酚、花色素苷、黄烷醇、酚酸、黄烷酮、反式-白藜芦醇及香气成分含量的影响。不同年份总酚含量存在显著差异（P＜0.05），但随年份变化无明显规律。年份对葡萄酒花色素苷含量有显著影响（P＜0.05），从2008年至2003年，3种花色素苷总含量呈幂函数下降。年份对葡萄酒非花色素苷类单体酚影响不大，特别是2004-2007年份，西拉和赤霞珠各年份酒间单体酚含量基本没有显著性差异（P＞0.05）。二甲花翠素-3-O-葡萄糖苷是葡萄酒中含量最高的花色素，其受年份影响显著并且在陈酿过程中显著降低（P＜0.05），可用于鉴别红葡萄酒的年份。相对于品种和产地而言，年份对西澳葡萄酒中24种香气成分影响较小。在赤霞珠各年份葡萄酒中只有8种化合物有显著性差异（P＜0.05），在西拉各年份葡萄酒中只有7种化合物有显著性差异（P＜0.05）。琥珀酸乙酯、乙酸异戊酯、顺式-3-己烯-1-醇和苯甲醇可用于区分不同年份赤霞珠和西拉葡萄酒。（4）利用葡萄酒酚类化合物和香气成分指标建立判别函数，对不同品种、产地及年份的葡萄酒进行分类鉴别。利用总酚、花翠素-3-O-葡萄糖苷、花青素-3-O-葡萄糖苷、二甲基花翠素-3-O-葡萄糖苷、丁香酸、儿茶素、白藜芦醇、槲皮素、杨梅素、戊酸乙酯、己酸乙酯、乙酸异戊酯、反式-3-己烯-1-醇、顺式-3-己烯-1-醇建立的判别函数可以实现葡萄酒品种的鉴别，判别分析分类与原始个案分类的一致率达到95%。利用总酚、没食子酸、丁香酸、p-香豆酸、表儿茶素、白藜芦醇、杨梅素、癸酸乙酯、棕榈酸乙酯建立的判别函数可以实现葡萄酒原产地的分类鉴别，赤霞珠和西拉原产地判别分析分类与原始个案分类的一致率均为100%。利用总酚、花翠素-3-O-葡萄糖苷、花青素-3-O-葡萄糖苷、二甲基花翠素-3-O-葡萄糖苷、山萘酚、琥珀酸乙酯、乙酸异戊酯、顺式-3-己烯-1-醇建立的判别函数可以实现葡萄酒年份的分类鉴别，赤霞珠和西拉的年份判别分析分类与原始个案分类的一致率分别为93.8%和86.4%。更多还原

【Abstract】 Thirty-six red wines and six white wines from Western Australia were analysed for theiranthocyanins, polyphenols, trans-resveratrol by means of high performance liquidchromatography (HPLC). Aromas from above thirty-six red wines and six white wines wereanalysed using solid phase extraction gas chromatography-mass spectrometry (SPE-GC-MS).The principal aim of this study was to clarify the possibility of correlating the levels of thesecompounds with the variety, origin and vintage of the wine and to use the properties of thesecompounds for wine classification. The discriminant analysis was used to form thediscriminatory function for the prediction of the wine classification. The main results weresummarized as follows:(1) Total phenol, anthocyanins, polyphenols, trans-resveratrol and aroma concentrationsof5varietal wines and4blended wines were examined for the discrimination among grapevarieties. The total phenol levels of red wines from Western Australia were0.99-2.56mg/L,which are almost10folds as white wines (0.14-0.38mg/L). Of the3anthocyanins, thecontent of malvidin-3-glucoside was the highest while cyanidin-3-O-glucoside was thelowest in red grape varieties. Cyanidin-3-O-glucoside was positively correlated withdelphindin-3-O-glucoside. Three anthocyanins were not detected in all white wines. Varietyhad a dramatic effect on the contents of gallic acid, syringic acid, trans-resveratrol, quercetin,myricetin and the ratio of catechin/epicatechin (P＜0.05). Aromas including ethyl hexanoate,ethyl octanoate, isomayl acetate, hexyl acetate and three fatty acids (octanoic acid, decanoicacid and hexanoic acid) can be used to identify red wines and white wines. The levels of ethylvalerate and cis-3-hexen-1-ol had significant difference between Cabernet Sauvignon andShiraz wine (P＜0.05). Pinot Noir wine was differed from Cabernet Sauvignon by isomaylacetate, isoamyl alcohol, hexanol, trans/cis-3-hexenol, benzyl alcohol, phenylethyl, ethylbutyrate and ethyl valerate. Ethyl valerate, isoamyl alcohol, decanoic acid, benzealdehyde,phenylethyl, ethyl laurate, ethyl hexanoate, hexyl acetate, trans/cis-3-hexenol, geraniol can beused to classify Chardonnay from blended white wine.(2) Anthocyanins, polyphenols, trans-resveratrol and aroma concentrations of CabernetSauvignon wines from3geographical origins and Shiraz wines from2geographical originsof Western Australia were assessed for the discrimination among origin wines. Origin hadno obvious effect on the anthocyanins (P＞0.05) but on non-anthocyanin ployphenol, which had significant effect on the levels of trans-resveratrol, myricetin, and epicatechin. The levelsof ethyl valerate, trans/cis-3-hexenol, ethyl octanoate, ethyl decanoate, ethyl laurate, ethylhexadecanoate，geraniol, isoamyl alcohol, octanoic acid and phenylethyl alcohol differedobviously among Cabernet Sauvignon wines from3origins. The concentrations of ethylhexanoate, ethyl octanoate, ethyl decanoate, ethyl laurate, ethyl hexadecanoate，geraniol,hexanoic acid and benzyl alcohol in the Frankland River Shiraz wine were significantlyhigher than that in Margaret River Shiraz wine. Ethyl ester took an important role in theclassification of the wines from different origins. Ethyl decanoate and ethyl hexadecanoatecan be used as an index to classify the Cabernet Sauvignon and Shiraz wine from differentorigins.(3) Total phenol, anthocyanins, polyphenols, trans-resveratrol and aroma concentrationsof7vintage Cabernet Sauvignon wines and4vintage Shiraz wines were examined for thediscrimination among vintage wines. The results showed that the total phenol contents weresignificantly varied in different vintage samples with an irregular pattern (P＜0.05). With theextension of vintage years, the total content of3anthocyanins decreased97.7%in exponentialform, delphindin-3-glucoside decreased from28.26to0mg/L, and malvidin-3-glucoside waslinearly decreased by97.2%. Vintage almost had no effect on the non-anthocyanins phenol,especially during the years of2004-2007. Compared to variety and origin, vintage had lesseffect on the levels of phenols and aroma. There were only8aroma compounds differingamong Cabernet Sauvignon wines from different vintages, and7aroma compounds for Shirazwines. Ethyl succinate, isomayl acetate, trans-3-hexen-1-ol and benzyl alcohol were taken asthe indices to differ the Cabernet Sauvignon and Shiraz wines from different vintages.(4) Discriminant functions were set by the selected phenolic and aroma compounds toclassify wines on the basis of variety, origin and vintage. Discriminatory analysis performedat95%significance level revealed a98.4%categorization of wines in terms of variety,100%categorization of Cabernet Sauvignon and Shiraz wines in terms of origin, and93.8%and86.4%categorization of Cabernet Sauvignon and Shiraz wines in terms of vintage. Our resultsindicated the analytical differentiation of those wines by means of the phenol and aromafunction could support their protection against fraud in the context of variety, origin andvintage.更多还原