Differentiation of “Claret”, Rosé, Red and Blend wines based on the content of volatile compounds by headspace solid-phase microextraction and gas chromatography

We studied 46 samples of wine from Denomination of Origin of La Rioja: 8 rosé, 8 “claret”, 10 red, 10 artificial blend and 10 ungrouped wines (3 rosé, 2 “claret”, 2 red and 3 blend). The samples were analysed by gas chromatography using headspace-solid-phase microextraction (HS-SPME) with a Carbowax-divinylbenzene (CW/DVB) fibre. The variables considered were 3-methyl butyl acetate, 3-methyl-1-butanol, ethyl hexanoate, 1-hexanol, ethyl octanoate, diethyl succinate, hexanoic acid, 2-phenylethanol, octanoic acid, decanoic acid and geraniol, all significant for distinguishing the types of wines except 3 metil-1-butanol and 1-hexanol. The different wine samples were classified by multivariate statistical methods. Good differentiation between the different groups of samples was achieved by means of principal component analysis and hierarchical cluster analysis. Linear discriminant analysis was used to differentiate and classify these wines. 3-methyl-butyl acetate, ethyl octanoate, diethyl succinate, hexanoic acid, 2-phenylethanol and decanoic acid were the most discriminant variables and gave 100% recognition ability and 100% prediction ability.     

Characterisation of the volatile fraction of Andalusian sweet wines

The characterization of the volatile composition that could have relevance on the organoleptic profile of Andalusian sweet wines obtained from Muscat and Pedro Ximenez grapes has been carried out. Twenty commercial samples, ten from Pedro Ximenez grapes and ten for Muscat grapes, have been studied. Various different chemometric tools have been applied in the analysis of the results. Using Cluster analysis and principal component analysis the samples have been grouped according to grape variety. Linear discriminate analysis allowed a 100% differentiation according to grape variety.     

Characterization of different annatto extracts based on antioxidant and colour properties

Annatto seeds are known as the only natural source of bixin, a carotenoid widely used in the food industry as colourant. Thus, the aim of this study was to obtain annatto extracts with different solvents and use statistical multivariate analysis to correlate the antioxidant and colourant properties of these extracts. According to the Principal Component Analysis, PC1 was associated to colour and PC2 to antioxidant capacity. Extracts with similar polarities were joined together by Cluster Analysis. The annatto extracts obtained with methanol/water and ethanol/water were characterized by their low bixin levels and a soft yellowish colour, whilst those obtained with methanol, ethanol and ethyl acetate showed red colour and higher bixin contents. The best solvent for bixin extraction was ethyl acetate (4.9 mg bixin/g seeds), whereas the highest total phenolic levels were verified in the most polar extracts, achieving maximum of 1.84 gallic acid equivalent/g seeds in the methanol/water extract. On the other hand, the hexane extract showed both low phenolic and bixin contents, as well as low free radical scavenging activity. Taking into consideration the antioxidant and colourant properties, solvents with medium polarity, especially methanol, should be used to obtain functional annatto extracts.     

Content in metallic ions of wines from the Madeira and Azores archipelagos

This study determines for the first time Na, K, Ca, Mg, Fe, Cu, Zn, Mn, Sr, Li and Rb contents in wines from the archipelagos of Madeira and Azores (Portugal). The greater part of the mean content for the different parameters fell within the ranges described in the literature, except for sodium whose higher content may be due to the effect of marine spray. ANOVA was used to establish the metals with significant differences in mean content between the wines from both archipelagos, between table and liquor wines of Madeira, and between wines of Pico and Terceira Islands from the Azores archipelago. Principal component analysis shows differences in the wines according to the wine-making process and/or the equipment employed. Stepwise linear discriminant analysis achieves a good classification and validation of wines according to the archipelago of origin, and the island in the case of Azores wines.     

Classification of wines according to type and region based on their composition

Zusammenfassung In 25 Riesling- und 20 Müller-Thurgau-Weinen aus derselben Gegend wurden dieflüchtigen Bestandteile, nichtflüchtigen Säuren und Aminosäuren quantitativ bestimmt. Bei Anwendung unserer Erkennungstechnik (pattern recognition techniques) auf diese Daten ergab sich eine einwandfreie Klassifizierung der Proben. In der Gruppe der Rieslingweine waren die sechs Ausleseweine klar zu erkennen. Auf Grund der Spurenelementanalyse wurde die Erkennungstechnik auch zur Klassifikation von 51 in zwei verschiedenen Gegenden in Westdeutschland (Mosel und Rhein-Pfalz) hergestellten Rieslingweinen angewendet. Auch hier wurde eine richtige Klassifizierung erhalten. Die Pfalzweine konnten sogar nach er nördlichen und der südlichen Anbaugegend unterschieden werden.

---

Classification of wines according to type and region based on their composition

Summary In 25 Riesling and 20 Müller-Thurgau wines from the same district volatile compounds, nonvolatile acids and amino acids were quantitatively determined. Application to this data set of pattern recognition techniques resulted in correct classification of the samples. In the group of the Riesling wines the six Auslese wines formed a clearly separated subgroup. Based on trace element analysis, pattern recognition techniques were also used to classify 51 Riesling wines produced in two different regions of West Germany, i.e. the Mosel and Rhein-Pfalz districts.Again a correct classification was obtained; the Rhein-Pfalz wines could even be distinguished into those produced in the northern and in the southern regions.

     

Classification of South African red and white wines according to grape variety based on the non-coloured phenolic content

The phenolic content of South African wines was investigated for the purpose of their classification according to variety or cultivar. Twenty-two phenolic compounds were quantified by liquid chromatography in 55 red wines of five varieties and in 38 white wines of three varieties. Data were compared to literature values reported for wines from across the globe. ANOVA was used to determine significant concentration differences between the different varieties. Further grouping of the data was investigated by principal component analysis and cluster analysis. Finally, linear discriminant analysis was used to derive classification functions for the effective recognition of red and white wine varieties which allowed 100.0% and 97.4% correct recognition of red and white wines, respectively.     

绢纺面料的分类研究

利用KES织物风格仪测量了25种绢纺面料的拉伸、剪切、弯曲等力学性能,运用系统聚类分析法进行了面料分类,基于多元方差分析确定面料最佳分类数,并探讨了分类后各类绢纺面料风格、规格参数的异同性。结果表明：所选面料最佳分类数为4类,4类面料的风格、规格参数具有明显的差异。     

Classification of juices and fermented beverages made from unripe,ripe and senescent apples based on the aromatic profile using chemometrics

The aim of this study was to assess differences between apple juices and fermented apple beverages elaborated with fruits from different varieties and at different ripening stages in the aroma profile by using chemometrics. Ripening influenced the aroma composition of the apple juice and fermented apple. For all varieties, senescent fruits provided more aromatic fermented apple beverages. However, no significant difference was noticed in samples made of senescent or ripe fruits of the Lisgala variety. Regarding the juices, ripe Gala apple had the highest total aroma concentration. Ethanal was the major compound identified in all the samples, with values between 11.83 mg/L (unripe Lisgala juice) and 81.05 mg/L (ripe Gala juice). 3-Methyl-1-butanol was the major compound identified in the fermented juices. Principal component analysis (PCA) and hierarchical cluster analysis (HCA) were applied and classified the juices and fermented juices based on physicochemical and aroma profile, demonstrating their applicability as tools to monitor the quality of apple-based products.     

Changes in terpenoid composition of milk and cheese from commercial sheep flocks associated with seasonal feeding regimens throughout lactation

Changes in the terpenoid content of milk and cheese from commercial sheep flocks monitored throughout lactation in the Cantabrian area of northern Spain were investigated. The flocks followed the same seasonal feeding strategy during lactation: indoor feeding in winter (early lactation) based on concentrate and forage; part-time grazing in the valley in early spring (mid lactation); and from mid spring on (late lactation), flocks were managed under extensive mountain grazing. In the present study design, seasonal feeding and lactation stage were intrinsically linked and could not be considered in isolation, and a holistic approach was necessary to consider the whole production management of the commercial flocks studied. Furthermore, the study focused on the identification of sesquiterpenoid ratios to differentiate milks and cheeses produced under extensive mountain grazing from those produced under other seasonal feeding regimens. Total abundance of mono- and sesquiterpenoids and that of individual compounds such as α-pinene, β-caryophyllene, α-humulene, α-amorphene, and γ-cadinene significantly increased in milk and cheese from indoor feeding to mountain extensive grazing. Sesquiterpenoid ratios such as γ-cadinene/α-muurolene, γ-cadinene/δ-cadinene, β-caryophyllene/α-muurolene, and (β-caryophyllene + γ-cadinene)/α-muurolene were used to differentiate mountain milks and cheeses from those from indoor feeding and part-time grazing in the valley. Multivariate discriminant analysis applied to individual terpenoids and sesquiterpenoid ratios showed milk and cheese samples classified into 2 groups: samples from indoor feeding and part-time grazing in the valley were classified together, and clearly separated from mountain milks and cheeses. The results of the present study showed that the sesquiterpenoid ratios approach could help to differentiate mountain dairy products from others obtained under other specific feeding regimens in a local environment.     

Volatile fraction of DOP “Castelo Branco” cheese: Influence of breed

“Castelo Branco” cheese is a Portuguese DOP cheese made from raw ewe’s milk coagulated with Cynara cardunculus, ripened for at least 40 days. “Merino da Beira Baixa” pure race is frequently used to produce milk for this cheese, however, exotic races such as Assaf and crusade of these two races are also used. The aim of this work has been to compare the volatile profile and sensory characteristics of DOP “Castelo Branco” cheeses manufactured during winter season with milk of breeds from Merino, Assaf and crusade of these two races and identify volatile compounds that can distinguish these cheeses.     

Free and bound volatile compounds as markers of aromatic typicalness of Moravia Dulce, Rojal and Tortosí red wines

Rojal, Moravia Dulce and Tortosí are three red minority grape varieties grown in La Mancha region (Spain). In this work, wines from these grape varieties were elaborated across four harvests (2006-2009). The aroma of wines was studied by instrumental and sensory analysis to determine their typicality and quality. Free and glycosidically-bound aroma compounds were isolated by solid phase extraction (SPE) and analysed using gas chromatography-mass spectrometry (GC-MS). GC-MS analysis of wines identified 75, 78 and 80 free aroma compounds and 59, 62 and 62 bound aroma compounds in Moravia Dulce, Rojal and Tortosí wines, respectively. C₆ and benzenic compounds were the major components of free varietal aroma. Bound aroma fraction was characterised by a larger concentration of benzenic and C₁₃-norisoprenoid compounds, which suggest that these wines possess a great aroma potential.     

Seasonal variation in the odour characteristics of whole milk powder

Seasonal changes in whole milk powder odour characteristics were monitored using SPME-GC and E-nose methodologies. ANOVA showed a significant effect of the season on dimethyl sulphide, n-pentanal, n-hexanal, and butyric acid. Whole milk powder manufactured in summer season had significantly (p < 0.05) higher levels of n-hexanal, n-pentanal and dimethyl sulphide as compared to autumn and winter seasons. On the other hand, butyric acid showed significant differences (p < 0.05) among autumn and spring. Applying linear discriminant analysis, seasonal variations in the odour profiles assessed by electronic nose were observed. A better classification outcome was obtained when volatile compounds and electronic nose data were analysed together, obtaining a success rate of 89.5% and 70.7% of the original cases and after cross-validation respectively. As in other application reported in the literature, E-nose approach represents an alternative technique to traditional methods of odour measurement.     

Classification of Marsala wines according to their polyphenol, carbohydrate and heavy metal levels using canonical discriminant analysis

Marsala is a popular Sicilian fortified aged wine with ancient tradition. Nowadays Marsala is exported all over the world and is considered one of the most important dessert wines. The aim of this study was to determine the concentration of carbohydrates, polyphenols and heavy metals in different types of Marsala wines and to achieve statistical classifications by stepwise forward canonical discriminant analysis (CDA). The obtained results provided evidence that different types of Marsala were correctly classified according to their phenolic and carbohydrate compositions. In particular, the residual sugars allowed a good discrimination among Marsalas having similar total sugar contents. CDA, performed using heavy metals as independent variables, showed that Superiore Ambra Secco and Vergine Marsalas were not discriminated, whereas a good separation among Fine Oro Dolce, Superiore Riserva and Fine Ambra Secco wines was obtained. Finally, an overall statistical model showed that the variables with the highest discriminant power were: tyrosol, caffeic acid, procyanidin B1, catechin, quercetin, kaempferol, lactose, rhamnose, zinc, copper and lead.     

基于指纹图谱和化学计量学的浓香型白酒分类研究

该研究以市售畅销的7个品牌、不同等级的浓香型白酒为研究对象,利用气相色谱-质谱联用（GC-MS）技术建立了浓香型白酒中风味成分的指纹图谱,结合相似度分析、主成分分析（PCA）和判别分析（DA）对不同品牌的浓香型白酒样品进行了有效区分和鉴别。结果表明,不同品牌间样品的相似度存在一定差异;PCA表明前三个主成分累计方差贡献率达到89.71%,能对样品进行聚类和区分;利用判别分析可以将不同产地的酒样区分开,正确率为100%。综上,相同品牌不同等级的白酒具有明显的相关性,不同品牌白酒可以利用指纹图谱结合化学计量学方法进行鉴别和分类,为白酒质量控制及真伪鉴定提供了一种新方法。     

Classification of some heat-treated liver pastes according to container type, using heavy metals content and manufacturer's data, by principal components analysis and potential curves

A total of 115 pork liver pastes were randomly collected in local markets from different brands, countries and containers. The concentrations of nine heavy metals (Cu, Ni, Cd, Fe, Mn, Pb, Cr, Co and Zn), determined by atomic absorption spectrometry, and some qualitative variables described on the labelling constituted the data set. Chemometrics analysis was performed combining principal components analysis (PCA), factor analysis (FA) and typical classification techniques, such as linear discriminant analysis (LDA) and potential curves (PoCu) to classify pork liver pastes. Origin of the sample, manufacturer and effect of manufacturing process were taken into account to verify traceability, which is an important issue in food safety policies.     

Statistical interpretation of the color parameters of red wines in function of their phenolic composition during aging in bottle

The relationships between the color parameters (colorimetric indexes and CIELAB variables) and the phenolic components (anthocyanins, pyranoanthocyanins, hydroxybenzoic and hydroxycinnamic acids, flavanols and flavonols) of young red wines from Vitis vinifera L. cv. Tempranillo, Graciano and Cabernet Sauvignon (vintage 2000, Navarra, Spain), have been investigated during 26 months of wine aging in bottle (a period embracing their commercial life), through the application of different statistical analysis (principal component, correlation and polynomial regression). The results of the principal component analysis first indicated that for each variety a high degree of interrelation existed between the color parameters and the anthocyanin pigments. Moreover, it was found that for each variety the color parameters were correlated with the anthocyanins during aging in bottle. Finally, by the application of polynomial regression analysis, both anthocyanins (simple glucosides and acetyl-glucosides) and pyranoanthocyanins (anthocyanin-pyruvic acid adducts) were selected as the variables that best described the different color parameters during aging in bottle. However, differences were found between varieties in relation to the type of anthocyanin pigment selected for describing each color parameter, finally indicating the importance of the grape variety factor in the definition of the wine chromatic characteristics.     

Chemometrics Classification of Argentine Clover and Eucalyptus Honeys According to Palynological,Physicochemical, and Sensory Properties

Palynological, physicochemical, and sensory methods were applied to ascertain the most discriminant variables for honey characterization. Fifteen physicochemical parameters, six indicators of antioxidant capacity and eight sensory attributes were considered. Clover and eucalyptus honeys were differentiated when the linear discriminant analysis was applied. Electrical conductivity, sugars, ferric ion reducing, and Trolox equivalent antioxidant capacity were the most discriminant variables. Odor and color intensities were the sensory attributes scoring the major differences between honeys. The canonical correlations technique pointed out that differences were related primarily to the content of pollen from Trifolium, Medicago sativa, and Lotus. These standards of quality provide a differentiating tool to classify unifloral honeys.     

Citrus juice classification by SPME-GC-MS and electronic nose measurements

According to suppliers indications, 76 commercial and 120 self-prepared citrus juices were grouped corresponding to fruit type, cultivar and treatment by linear discriminant analysis of the volatile compounds measured by an electronic nose (EN) instrument and solid-phase micro extraction gas chromatography mass spectrometry (SPME-GC-MS) from the headspace. While the number of available variables for SPME-GC-MS with 311 was rather high and always allowed for reasonable group separation, the number of variables for the EN instrument was limited to 12. Nevertheless, wrong and misleading supplier information was uncovered especially by EN data. Analysis revealed a commercial orange juice grouped close to grapefruit and two pummelos declared as grapefruits, while further conspicuous samples were noted but require more detailed investigations.