【作者】 张妍；

【导师】 潘思轶；

【作者基本信息】 华中农业大学 ， 食品科学， 2008， 硕士

【摘要】 不同果汁的理化品质特征和所含芳香成分的类别和数量千差万别。本文对2006与2007两年度不同品种橙汁理化品质进行了研究,同时考察了市售不同品牌不同类型的橙汁和橙汁饮料的特征理化品质差异。并采用HS-SPME与GC-MS联用的方法,对实验室自制NFC橙汁以及市售100%NFC橙汁与RFC橙汁的芳香成分进行分析判别。还采用FOX 4000电子鼻对同一产地不同品种的橙汁和桔汁进行风味区分,并对掺假不同梯度的橙汁进行了定性和定量分析。此外还利用缓冲能力和电子鼻技术对不同品种橙汁的原橙汁含量进行了研究。研究结论如下:1.橙汁特征理化品质分析2006年度与2007年度的同一品种原橙汁的理化品质没有显著差异,理化品质较稳定。PCA分析结果表明:pH值、a~*值、b~*值、可溶性固形物含量、总酸含量、固/酸比值、蔗糖含量、葡萄糖/总糖比值、富马酸含量、VC含量、水溶性黄酮含量及矿物元素Cu、Fe、Mg、Mn、Zn的含量等在第一主成分上有较高的载荷,所以第一主成分是这16个指标的综合反映,而且这16个指标的系数相当,进而说明这16项指标用于评价不同品种原橙汁的理化品质都是必不可少的。果糖含量、葡萄糖含量、果糖/葡萄糖比值、氨基态氮含量以及矿物元素Cr的含量在第二主成分上有较高载荷。L~*值和Ca含量对第三主成分贡献较大。以上这三个主成分从三个方面刻画分析了不同品种原橙汁的理化品质情况具有98.3%的可靠性。考察了市售不同品牌不同类型的橙汁和橙汁饮料的特征理化品质,并对其进行判别分析后得到:有pH值、L~*值、a~*值、b~*值、可溶性固形物含量、总酸含量、固/酸比值、果糖和葡萄糖含量等9种理化指标值进入判别函数,并且所得判别模型既可以用判别函数对未知样品的类型进行判别,也可以用已知类型的样品对取得的函数进行判别效果检验。对原始数据进行回代检验后显示,取得的判别函数对原始数据的类型判别效果很好,正确率为100%。因此,利用这种理化品质的差异性对橙汁的类型进行科学的判别是可行的。2.利用芳香成分识别橙汁类型的初步研究采用峰面积归一化法对橙汁的芳香成分进行相对定量,参考中外文献报道,借助多元统计分析,以橙汁样品中18种芳香化合物的相对含量为自变量,对NFC和RFC两种类型橙汁进行判别分析后得到:从10个橙汁样品中共检测到烃类、醇类、酯类、醛类、酮类等共127种芳香化合物,相比于其他种类芳香成分,烃类和醇类化合物含量是其中的主要组成成分。以芳香化合物相对含量为自变量建立了2个Fisher线性判别函数,所得判别模型既可以用判别函数对未知样品的类型进行判别,也可以用己知类型的样品对取得的函数进行判别效果检验。并对原始数据进行回代检验,显示取得的判别函数对原始数据的类型判别效果很好,正确率为100%。因此,可以利用这种芳香成分的差异性对橙汁的类型进行识别分析。3.基于电子鼻对桔汁与橙汁的鉴别分析电子鼻可以有效地鉴别不同品种的橙汁与桔汁,并具有良好的灵敏度和重复性。因此,可以通过建立不同品种橙汁的特征指纹库,有效的进行品种橙汁的鉴别和保护。通过对掺假不同比例桔汁的橙汁样品进行检测,PCA分析结果显示,掺假不同比例桔汁的橙汁随着桔汁混合比例的提高,橙汁样品呈现良好的分别趋势,气味指纹呈一定趋势变化。对橙汁比例的PLS拟合分析,拟合效果良好,相关系数达0.9996,并对桔汁含量的拟合也取得了较好的效果。说明电子鼻不仅对定性检测橙汁掺假具备很好的灵敏度,而且对掺假比例的定量结果也有较高的准确性,且预测效果好。4.原橙汁含量的分析利用溶液缓冲能力测原橙汁含量的方法简便、快捷,对检测仪器要求不高,适合工厂企业等非专业检测机构检测使用。但对于不同品种的橙汁,其缓冲系数有一定的差别。该方法可能可以推广到更多的果汁含量的测定上。采用电子鼻技术对不同原橙汁含量的样品进行识别,在用电子鼻定量检测原橙汁的含量时,随着原橙汁含量的增大,电子鼻传感器响应信号逐渐增大,灵敏度较好,区分度较高。借助仪器内置的多元统计软件对不同原橙汁含量的样品进行了PLS拟合分析,拟合效果良好,相关系数达0.9996。说明用电子鼻定量检测原橙汁的含量具有可行性。更多还原

【Abstract】 Different juices have different physical and chemical character and different aroma compositions. In this paper, physical and chemical character of several varieties of orange juice in 2006 and 2007 were researched, also analysed the different commercial brands of different types of orange juice and Orange beverage . Aroma components of not from concentrated orange juice and refrigerated orange juice from concentrated were analysed by integrated using HS-SPME-GC-MS, and then distinguished them by SPSS. FOX 4000 electronic nose was used to distinguish the different orange juice and mandarin orange juice which come from one region, and also used to analyze adulterated orange juice qualitatively and quantitatively. In addition, we calculated the contents of raw orange juice by measuring buffer capacity and by the electronic nose technology. The main results of this paper are as follows.1. Analysis of physical and chemical characterThere was no significant difference in physical and chemical character of same species of orange juice between 2006 and 2007. PCA result showed that: pH, a~*, b~* , solid/acid ratio, glucose/total sugar ratio and content of soluble solids, total acid, sucrose, fumaric acid, VC, flavones, Cu, Fe, Mg, Mn, Zn have a higher load in the first principal component, and the first principal component is the comprehensive reflection of 16 indicators, and it is essential. The content of fructose, glucose, amino nitrogen, Cr and fructose/glucose ratio were remarkable higher in the second principal component. L~*and the content of Ca were remarkable higher in the third principal component. Over the three principal components accounted for physical and chemical character of orange juice for about 98.3% reliability. LDA result showed that: there were 9 kinds of physical and chemical parameters entered the gained discriminant functions, such as pH, L~*, a~* , b~* , solid/acid ratio and content of soluble solids, total acid, fructose, glucose. The gained discriminant functions can not only determine the unknown samples, but also can be determined from the known samples. The discriminant functions with 100% accuracy were constructed. Therefore, it was practical to make use of physical and chemical character to establish a scientific judgment of orange juice identification.2. Preliminary research on the identification of orange juice types by the use of aroma compositionsThe quantitative results showed that there were 127 aroma components in the 10 juice samples, and hydrocarbons, esters, alcohols, ketones, and aldehydes were the major constituents. With Chinese and foreign literature, 18 of which mainly contributed to orange juice aroma were analyzed by LDA, and established two Fisher linear discriminant function according to aroma compounds which quantified relatively. The gained discriminant functions can not only determine the unknown samples, but also can be determined from the known samples. The discriminant functions with 100% accuracy were constructed. Therefore, it was practical to make use of aroma compounds to identify the type of orange juice.3. Distinguish of orange juice and mandarin orange juice by electronic noseElectronic nose can identify different species of orange juice and mandarin orange juice effectively, and has a good sensitivity and repeatability. Therefore, we can establish the characteristics fingerprint of different varieties of orange juice to identify and protect varieties of orange juice effectively. Meanwhile, through detecting the different adulterated orange juice samples, PCA result showed that the trend of aroma fingerprint changed along with the proportion of mandarin orange juice to orange juice. PLS result is good, and the correlation coefficient is 0.9996. Therefore, electronic nose can detect adulterated orange juice qualitatively and quantitatively.4. Analysis of content of raw orange juiceCalculating the contents of raw orange juice by measuring buffer capacity is simple and rapid, and be appropriate for other non-professional testing organizations. But different orange juice maybe have different buffer capacity coefficient. This method may be extended to calculate the contents of other juice.Using electronic nose technology to detect samples which had different content of the raw orange juice, the result showed that the electronic nose sensor response signal increased along with the content of raw orange juice, and the sensitivity is good. With the built-in multivariate statistical analysis, the samples which had different content of the raw orange juice were analyzed by PLS. PLS result is good, and the correlation coefficient is 0.9996. It illuminated that using electronic nose to detect the content of the raw orange juice quantitatively was feasible.更多还原