Rapid headspace solid-phase microextraction-gas chromatographic–time-of-flight mass spectrometric method for qualitative profiling of ice wine volatile fraction: II: Classification of Canadian and Czech ice wines using statistical evaluation of the data

https://doi.org/10.1016/j.chroma.2007.02.052Get rights and content

Abstract

The previously developed and optimized headspace solid-phase microextraction (HS-SPME)-GC–time-of-flight (TOF) MS analytical method for the determination of compounds with a wide range of polarities and volatilities was successfully used in this study to characterize and classify a large set of ice wines according to their origin, grape variety and oak or stainless steel fermentation/ageing conditions, based on a statistical evaluation (principal component analysis (PCA)) of the measured data. More than 130 ice wine samples collected directly from Canadian and Czech wine producers were analyzed in this study. The SPME step was beneficially carried out utilizing the new-generation super elastic divinylbenzene/Carboxen/polydimethylsiloxane (DVB/CAR/PDMS) 50 μm/30 μm fiber assembly. One fiber was used for the whole sequence of ice wine samples, control and blank experiments, which consisted of more than 600 individual extraction/injection cycles. Utilizing the high-speed TOF analyzer, full spectral information within the range of 35–450 u was collected for the entire GC run (as short as 4.5 min) without compromising in the detection sensitivity, as compared to other scanning mass analyzers operated in selected ion monitoring or MSn mode to achieve similar sensitivity. The identification of analytes was performed by a combination of the linear temperature-programmed retention index (LTPRI) approach with the comparison of the obtained spectra with three libraries included in the ChromaTOF software. A total of 201 peaks were tentatively assigned as ice wine aroma components and 58 of those compounds were evaluated in all of the examined samples.

Introduction

The acceptance of food depends mainly on its flavor composition. On the other hand, consumers are also more and more concerned with the food origin and safety. All food products potentially targeted for adulteration are either products of a high commercial value or those produced in large amounts. Ice wine belongs to the first group of these products. Due to the relatively high cost of authentic ice wine, it is very important to characterize and classify this expensive commodity to help to determine fraudulent products that have recently emerged on the market, especially in Asia [1]. In general, wine is typically adulterated by dilution with water, chaptalization (the process of adding sugar to the fermenting wine to raise the final alcohol content) or mislabeling the age or origin of the wine [2]. Isotope ratio mass spectrometric method (IR-MS) can be applied with a great performance to determine, if a wine is faked in terms of water or sugar addition [3]. Fast and clear identification (and quantification) of the patterns and/or markers suitable for food differentiation is becoming an important issue when considering international trade policies and regulations.

Ice wine is a sweet wine belonging to the group of dessert wines. Since grapes are harvested and pressed while still frozen and water in the berries is crystallized, only a few drops of juice are collected. Therefore, the ice wine yields are very low (only 5–10%), which explains the high price of this commodity [4], [5], [6]. Only ice wine approved by the Appellation of Origin system called Vintners Quality Alliance (VQA) is allowed to be produced in Canada [4] and several requirements set by VQA must be met, including the prohibition of artificial freezing and sugar addition during the entire ice wine (“icewine” in the VQA terminology) production process [4]. The high sugar levels in ice wine lead to a slower fermentation process that typically takes several months to be naturally finished. Due to their relatively thick skins, Vidal and Riesling are two typical cultivars that are used for the production of ice wine [7]. Only two previous studies reported in the literature (Cliff et al. [7] and Nurgel et al. [8]) provide any scientific information on the chemical composition, sensory analysis and differentiation of ice wines according to their origin.

The typical wine “bouquet” is a complex combination of hundreds of components belonging to various chemical groups [9], [10], [11]. The use of oak wood in processing of premium table or dessert wine products can be performed primarily to increase the flavor complexity. Wood-related compounds (mainly syringaldehyde, but also vanillin, eugenol, cis- and trans-oak lactone, guaiacol, 4-methylguaiacol, 4-ethylguaiacol and furfural) have commonly been used to differentiate between stainless steel and barrel aged wines [11], [12], [13], [14].

Regarding food and fragrance analysis in general, solid-phase microextraction (SPME) is becoming one of the most prevalent isolation techniques This technique is highly reproducible, offers a good linear dynamic range and, due to pre-concentration, typically provides comparable or lower limits of detection to those of other extraction techniques [15], [16], [17], [18], [19], [20]. The recently developed cold-fiber SPME technique [21] offers further expansion of this approach for highly volatile flavor analysis, including compounds typically found in food and fragrances.

Several classification studies have recently been published in the field of wine analysis, some of them involving SPME sample preparation technique prior to separation and quantification procedures. Spranger et al. reported the successful differentiation of wines according to winemaking procedures [17] and Kotseridis et al. classified wines of different grape varieties [22]. Wine origin, grape variety and ageing procedure were subjects used to differentiate wines in a Martí et al. study [23], and white wines from Galicia region in Spain were characterized by their aromatic index by Falqué et al. [24]. Several markers of oak-aged wines were summarized by Pollnitz et al. [13], [25], Díaz-Maroto et al. [12] and in studies of some other authors. A study by Marengo et al. successfully classified Barbaresco, Barolo, Langhe Nebbiolo and Nebbiolo grape wines from Italy using SPME sample preparation and extraction and several chemometric models. These wines were differentiated from each other, and the age of the wine could be predicted according to ester and alcohol profiles, which could in future help to prevent adulteration attempts [26]. The volatile composition of wines was found to be dependent on the grape variety, as reported by Pozo-Bayón et al. [27]. Principal component analysis (PCA) [7], [13], [17], [23], [24], [26], [28], [29], [30], [31], [32], [33], [34], linear discriminant analysis (LDA) [11], [13], [24], [26], [28], [29], [35], [36], clustering [17], [26], soft independent modeling of class analogy (SIMCA) [23], [26], partial least-squares regression (PLS) [33] or analysis of variance (ANOVA) [7], [11], [22], [28], [32], [36], [37], [38], [39] are typically the most common statistical procedures that are used to differentiate the data in these classification/differentiation studies.

To the authors’ best knowledge, SPME has not previously been used for the analysis of ice wine. Therefore, the objective of this study was to use the previously developed rapid HS-SPME-GC–MS method [40] for the analysis of ice wine aroma components. A new-generation super elastic DVB/CAR/PDMS 50 μm/30 μm fiber assembly was used for the headspace extraction of analytes from ice wine samples. The high-speed time-of-flight (TOF) mass analyzer was used to achieve rapid GC–MS analysis. The profiles of volatile and semi-volatile compounds in various samples of Canadian (Ontario, ON and British Columbia, BC) and Czech ice wine were compared using PCA for the classification of the wines according to origin, grape variety, oak or stainless steel fermentation/ageing procedures used during the wine production or aroma profile differences between ice wines and late harvest wines.

Section snippets

Analytical reagents and supplies

5-Butyldihydro-4-methyl-2(3H)-furanone, also called whiskey lactone (>98%, mixture of cis- and trans-isomers, 1:1, w/w) was obtained from Aldrich (Milwaukee, MI, USA). The super elastic DVB/CAR/PDMS 50 μm/30 μm fibers were obtained from Supelco (Bellefonte, PA, USA). The new fibers were conditioned according to the manufacturer's recommendations prior to the first use and were kept at the desorption temperature for an additional 30 min prior to the actual sample sequence. Purchase of all other

Results and discussion

Table wine analysis is a subject of hundreds of published scientific articles, however, relatively few studies have been published to date on the analysis of desert wines, including ice wines. Ice wines are considered products of a high commercial value and in order to compare the profiles of the particular (ice) wine samples collected for this study (see Table 1), sample ONI42a with a chromatographic profile rich in volatile and semi-volatile compounds was used as a reference sample, to which

Conclusions

The rapid headspace SPME-GC–TOF-MS analytical method for the determination of ice wine aroma profile components was used to study a large set of wine samples for characterization and classification purposes. The qualitative evaluation of the selected volatile components in this study allowed for the characterization of ice wines according to the origin, grape varieties and stainless steel tank/oak barrel processing, as well as for differentiation between ice wines and late harvest wines. Only a

Acknowledgements

The authors would like to acknowledge the assistance of LECO (St. Joseph, MI, USA) that provided us with the SPME-GC–TOF-MS system. The financial assistance of the Leap Technologies (Carrboro, NC, USA) and Natural Sciences and Engineering Research Council of Canada (NSERC), the valuable consultations with Professor Jana Hajslova (Institute of Chemical Technology, Prague, Czech Republic), cooperation with the individual Canadian and Czech wineries and the Proneco s.r.o. company that collected

References (84)

  • R.E. Subden et al.

    Food Res. Int.

    (2003)
  • M. del Alamo Sanza et al.

    Anal. Chim. Acta

    (2004)
  • P. Arapitsas et al.

    Food Chem.

    (2004)
  • M. Isabel Spranger et al.

    Anal. Chim. Acta

    (2004)
  • C. Sala et al.

    J. Chromatogr. A

    (2002)
  • R. López et al.

    J. Chromatogr. A

    (2002)
  • V. Ferreira et al.

    J. Chromatogr. A

    (2003)
  • M.P. Martí et al.

    J. Chromatogr. A

    (2004)
  • E. Falqué et al.

    Talanta

    (2001)
  • A.P. Pollnitz et al.

    J. Chromatogr. A

    (1999)
  • M.A. Pozo-Bayón et al.

    J. Chromatogr. A

    (2001)
  • J.S. Câmara et al.

    Anal. Chim. Acta

    (2004)
  • J.J. Rodríguez-Bencomo et al.

    J. Chromatogr. A

    (2003)
  • V. Ferreira et al.

    Lebensm. Wiss. Technol.

    (1996)
  • M.S. Pérez-Coello et al.

    J. Chromatogr. A

    (1997)
  • R. Castro et al.

    Anal. Chim. Acta

    (2004)
  • L. Setkova et al.

    J. Chromatogr. A

    (2007)
  • A. Calleja et al.

    Food Chem.

    (2005)
  • R. Boulanger et al.

    Food Chem.

    (2001)
  • E.H.A. Andrade et al.

    J. Food Composition Anal.

    (2000)
  • A. Hognadottir et al.

    J. Chromatogr. A

    (2003)
  • E. Tudor

    J. Chromatogr. A

    (1997)
  • J.C.R. Demyttenaere et al.

    J. Chromatogr. A

    (2003)
  • L. Setkova et al.

    Anal. Chim. Acta

    (2007)
  • J.D. Carrillo et al.

    J. Chromatogr. A

    (2006)
  • E. Guchu et al.

    Food Chem.

    (2006)
  • B. Fernández de Simón et al.

    Anal. Chim. Acta

    (2006)
  • M. Esti et al.

    Anal. Chim. Acta

    (2006)
  • J.L. Giraudel et al.

    J. Chromatogr. A

    (2007)
  • R. Castro Mejías et al.

    J. Chromatogr. A

    (2003)
  • P. Romano et al.

    Int. J. Food Microbiol.

    (2003)
  • C. Cordella et al.

    J. Agric. Food Chem.

    (2002)
  • N. Ogrinc et al.

    Anal. Bioanal. Chem.

    (2003)
  • D.J. Erasmus et al.

    Am. J. Enol. Vitic.

    (2004)
  • M. Cliff et al.

    Am. J. Enol. Vitic.

    (2002)
  • C. Nurgel et al.

    J. Sci. Food Agric.

    (2004)
  • D. de la et al.
  • S. Cabredo-Pinillos et al.

    Chromatographia

    (2004)
  • M.C. Díaz-Maroto et al.

    J. Agric. Food Chem.

    (2004)
  • A.P. Pollnitz et al.

    J. Agric. Food Chem.

    (2004)
  • Cited by (0)

    View full text