Classification of Chinese Rice Wine According to Geographic Origin and Wine Age Based on Chemometric Methods and SBSE-TD-GC-MS Analysis of Volatile Compounds

Abstract

The feasibility of the discrimination of Chinese rice wines with respect to their geographic origin and wine age using volatile profiles and chemometric methods was investigated. Stir bar sorptive extraction with thermal desorption and gas chromatography-mass spectrometry (SBSE-TD-GC-MS) was used for the determination of volatile compounds. A total of 63 volatile compounds were identified, including 30 esters, 9 alcohols, 9 aldehydes, 7 acids, 4 ketones, 3 volatile phenols and 1 miscellaneous compound. The major compounds in Chinese rice wine were ethanol, isoamyl formate, ethyl acetate, diethyl succinate, phenylethyl alcohol, benzaldehyde, ethyl lactate, 2-butanol and ethyl butyrate. Principal component analysis (PCA) and cluster analysis (CA) were used to characterize the different Chinese rice wine samples by geographic origin and wine age. The PCA results showed that the characterization of Chinese rice wine by SBSE-TD-GC-MS was highly related to geographic origin and wine age. The CA results indicated that geographic origin had more of an effect than wine age on the character of Chinese rice wine, and that 5-year old wines had a transition of flavor. The results showed that SBSE-TD-GC-MS together with chemometric methods could provide a practical reference for the characterization of Chinese rice wines.

Introduction

Chinese rice wine, in addition to wine and beer, is one of the most ancient wines in the world (Wang and Yi, 2001). It is fermented from rice with wheat Qu (source of microorganisms and crude enzymes) and yeast (Mo et al., 2010). The best-known Chinese rice wine is Shaoxing rice wine. Comparable to the European Denomination of Origin, the National Standard of China GB/T 17946-2008 (Shaoxing rice wine) defines that Shaoxing rice wine is fermented from high quality glutinous rice and wheat as well as water from Mirror Lake in the region near Shaoxing city in the Zhejiang province, China. Before their sale, certain Chinese rice wines are stored in ceramic pottery to age. The aged Chinese rice wines have a smoother, mellower and richer mouth feel and a more harmonious flavor than the non-aged wines (Shen et al., 2012).

To protect producers by guaranteeing their reputation and to protect consumers by ensuring consistent character of the wine, research has been conducted to discriminate wines by their geographic origin or their wine age/vintage. The volatile compounds are centrally important to ensuring the identity of wine (Pereira et al., 2010), as their contents depend on the brewing technology, geographic origin and aging process (Tredoux et al., 2008). Much attention has been paid to the volatile profiles of wine, including research on the classification of liqueur wines (Pereira et al., 2010), the discrimination of white wine (Hernanz et al., 2009), the characterization of Vino (Gil et al., 2006) and the classification of red wine (Zhang et al., 2010; Alejandro Calleja and Falqué, 2005).

Sample preparation is vitally important in gas chromatography-mass spectrometry (GC-MS) analysis, as these levels range from µg/L to mg/L (Tredoux et al., 2008). The solid-phase microextraction (SPME) method has been widely used in flavor analysis of wine (Ryan et al., 2005; Rodrigues et al., 2008; De León-Rodríguez et al., 2006). Luo et al. (2008) identified a total of 97 volatile and semi-volatile compounds using SPME-GC-MS; among them, 39 were volatile. Stir bar sorptive extraction (SBSE) has recently been introduced as another sample preparation method. Because of the increased quantity of stationary phase utilized, SBSE offers increased sensitivity compared with SPME (Tredoux et al., 2008; Maggi et al., 2008; Hayasaka et al., 2003). This new method has been used for identifying the volatile compounds in Madeira wine (Alves et al., 2005) and “Terras Madeirenses” table wines (Perestrelo et al., 2009).

In the research herein, the volatile profiles of Chinese rice wine samples sourced from different geographic origins and different wine ages were analyzed by stir bar sorptive extraction with thermal desorption and gas chromatography-mass spectrometry (SBSE-TD-GC-MS). Using multivariate analysis tools, such as principal component analysis (PCA) and cluster analysis (CA), the Chinese rice wine samples were characterized according to geographic origin and wine age based on the volatile profiles.

Materials and Methods

Chinese rice wine samples    Twenty-seven Chinese rice wine samples were collected from local breweries. Table 1 lists the sample group code, sample code, sample name, total sugar, alcohol content, wine age and geographic origin of the 27 Chinese rice wine samples. The samples from the same group were all selected from different production dates.

Table 1. Details information of the 27 Chinese rice wine samples sourced from different geographic origins and wine ages

Sample group code	Sample code	Sample name	Total sugar (g/L)	Alcohol (% V/V)	Wine age	Geographic origin
JF1	JF1-1	JinFeng rice wine	15.1 – 40.0	15.0	1 year age	Shanghai
	JF1-2
	JF1-3
JF3	JF3-1	JinFeng rice wine	15.1 – 40.0	17.0	3 year age	Shanghai
	JF3-2
	JF3-3
JF5	JF5-1	JinFeng rice wine	15.1 – 40.0	16.0	5 year age	Shanghai
	JF5-2
	JF5-3
TP1	TP1-1	TaPai rice wine	15.1 – 40.1	≥15.5	1 year age	Shaoxing city, Zhejiang province
	TP1-2
	TP1-3
TP3	TP3-1	TaPai rice wine	15.1 – 40.1	≥15.0	3 year age	Shaoxing city, Zhejiang province
	TP3-2
	TP3-3
TP5	TP5-1	TaPai rice wine	15.1 – 40.1	≥14.0	5 year age	Shaoxing city, Zhejiang province
	TP5-2
	TP5-3
XT	XT-1	XiTang rice wine	15.1 – 40.2	15.0	5 year age	Jiashan city, Zhejiang province
	XT-2
	XT-3
FH	FH-1	FenHu rice wine	15.1 – 40.3	12.0	5 year age	Jiaxing city, Zhejiang province
	FH-2
	FH-3
SML	SML-1	ShuiMingLou rice wine	15.1 – 40.4	11.0	5 year age	Nantong city, Jiangsu province
	SML-2
	SML-3

Volatile analysis by SBSE-TD-GC-MS    Eight milliliters of each Chinese rice wine sample were stirred with a polydimethylsiloxane (PDMS)-coated stir bar (10 mm in length and 0.5 mm in film thickness) at 500 rpm for 30 min at room temperature. 2-Octanol (99%, Sigma-Aldrich, St. Louis, MO) at 411 mg/L was used as the internal standard. The stir bars were then removed from the samples, rinsed with distilled water, dried with cellulose tissue and finally transferred into thermal desorption system (TDS, Gerstel GmbH & Co.KG, Germany) for GC-MS analysis. This method was based on a previously optimized method reported by Jaeho et al. 2014.

The volatile compounds were transferred into the Agilent 6890 gas chromatograph (GC) equipped with a mass selective detector (MSD) (model 5973N, Agilent Technologies, New York, USA). The capillary column used was an HP-Innowax (30 m × 0.25 mm i.d. × 0.25 µm film thickness) from Agilent Technologies. The GC operation conditions were as follows: the carrier gas, helium (99.999% purity), was at a constant flow rate of 1 mL/min; the injection was conducted in splitless mode with the injector temperature at 250°C; the chromatographic program was set at 40°C (held for 0.5 min), raised to 230°C at 3°C/min and held for 15 min. For mass spectrometry analysis, an electron impact mode (EI) of 70 eV was used. The mass range varied from 10 to 450 m/z and the detector temperature was 150°C.

Compound identification and quantification    The volatile compounds were identified by comparison to Kovats retention indices (RI) (Cates and Meloan, 1963) and by comparison with the MS fragmentation patterns of the reference compounds; and mass spectra in the Wiley7n.l and NIST05 (Agilent Technologies Inc., New York, USA) databases and previously reported Kovats RI values were also used. The Kovats RI of the unknown compounds were determined via sample injection with a homologous series of alkanes (C6-C30) (Sigma-Aldrich, St. Louis, MO). The GC-MS conditions were the same as described above. To quantify the volatiles, the integrated areas, based on the total ion chromatogram, were divided by internal standard peak, assuming a response factor of one. The relative volatile concentrations in the 27 samples were determined by comparison with the concentration of the internal standard (2-octanol).

Chemometric analysis    The volatile compounds of the Chinese rice wine samples sourced from different geographic origins and different wine ages were analyzed by PCA and CA to classify these wines by the natural groupings that exist between the samples. In this research, PCA was performed on the mean values of the relative volatile concentrations using a correlation matrix with no rotation. A Euclidean metric was used for the CA method. All statistical analyses were performed using XLSTAT ver. 7.5 (Addinsoft, New York, NY, USA).

Results and Discussion

Volatile compounds in Chinese rice wines based on geographic origin    A total of 63 volatile compounds were identified in the 27 samples, including 30 esters, 9 alcohols, 9 aldehydes, 7 acids, 4 ketones, 3 volatile phenols and 1 miscellaneous compound. Among the identified volatiles, the esters, aldehydes, alcohols and acids were the principal groups. The most abundant compounds in the Chinese rice wine samples were isoamyl formate, ethyl acetate and phenylethyl alcohol. This result was in accordance with that obtained by Cao et al. (2010).

Esters    Esters were formed by either enzymatic or chemical esterification of organic acids and alcohols. The most common in wines are ethyl esters (Tredoux et al., 2008). A total of 30 esters were detected in the 27 samples. Most of these compounds were previously found in Yakju and Chinese rice wine (Kwon et al., 2010; Lee et al., 2007; Luo et al., 2008). Ethyl acetate, isoamyl formate, diethyl succinate and ethyl lactate were the major esters in the 27 samples. The fatty acid esters primarily contributed fruity aromas (Escudero et al., 2004), diethyl succinate contributed a fruity and sweet aroma (Fan and Qian, 2006), and isoamyl formate produced a mulberry aroma. These compounds may be important contributors to the aroma of the Chinese rice wine samples. In addition, two lactones (γ-butyrolactone and γ-nonalactone) were detected in all samples. These compounds were most likely produced by the bacteria (Pinto et al., 2006) from the wheat Qu. The TP5 sample group sourced from Shaoxing city in the Zhejiang province had the highest ethyl acetate contents among the wine samples from different geographic origins. The brewing technologies of the different regions may affect the rice lipid utilization and the formation of high molecular weight esters (Chen and Xu, 2010; Mo et al., 2010).

Alcohols    The alcohols were another large class, containing 9 compounds. The most abundant alcohol in the Chinese rice wine samples was phenylethyl alcohol and the second was 2-butanol. Phenylethyl alcohol had a rose-honey-like flavor, and it not only served as a perfume formulation ingredient in the flavor industry but also acted as an aroma enhancer in the fermentation products (Chen et al., 2009). Phenylethyl alcohol was found at higher levels in the XT, FH and TP5 sample groups. These three sample groups were all sourced from the Zhejiang province. The results of this comparison agreed with that obtained by Chen et al. (Chen and Xu, 2010). The concentrations of 1-hexanol and 2-pentanol were higher in the JF sample group from Shanghai. The results of this comparison were in accordance with previous studies (Chen and Xu, 2010; Jian et al., 2009).

Aldehydes    Nine aldehydes were identified in the samples, and these compounds were previously found in Chinese rice wine (Cao et al., 2010; Mo et al., 2010). The most abundant aldehyde was benzaldehyde, followed by furfural. The TP5 sample group had the highest content of benzaldehyde, which was described as having sweet, fruity, nutty and caramel-like odors (Xiao et al., 2011). 2-Phenyl-2-butenal, which had green, powder and cocoa aromas, was also a principal constituent of Chinese rice wine.

Acids    A total of seven acids were identified in the samples, including short carbon chain saturated acids (C2–C10). Acetic acid, isovaleric acid, nonanoic acid and benzoic acid were the principal acids of the Chinese rice wine samples. Fatty acids are important for the flavor and taste of Chinese rice wine (Cao et al., 2010). Several fatty acids in rice wine originated from the raw materials (Gu, 1996). Most were released or produced by yeast during the fermentation process (Gallart et al., 1997; Ng, 2002). The XT sample group lacked acids such as hexanoic acid. However, high contents of corresponding esters (ethyl hexanoate) were found in these samples. These ethyl esters were derived from the esterification of the free fatty acids with ethanol. Therefore, their corresponding fatty acids were expected to be present during the fermentation process.

Phenols    Only three phenols (phenol, 4-ethylguaiacol and 4-ethylphenol) were identified in the 27 Chinese rice wine samples. The FH sample group had the highest total phenol content. Phenols are typically produced by thermal degradation through depolymerization or the oxidation of lignin, which was composed of repeating phenol units with three carbon side chains (Natera et al., 2003). Phenolic compounds, such as guaiacol, 4-vinylguaiacol and 4-ethylguaiacol, were responsible for spicy, smoky and phenolic odors in wine (Mo et al., 2010). The threshold values for phenol, 4-ethylguaiacol, and 4-ethylphenol were 0.031 mg/kg, 1.10 mg/kg, and 0.042 mg/kg, respectively. Because of their low threshold values, the flavor of the phenols could be regarded as a characteristic aroma of Chinese rice wine, even though the phenol content was not the highest.

Miscellaneous compound    2-Ethoxythiazole, the additional miscellaneous compound, was identified. It could be formed through both enzymatic and non-enzymatic pathways (Owens et al., 1997; Rizz, 1972).

Volatile compounds in Chinese rice wines based on wine age    Chinese rice wine is traditionally stored in ceramic pottery with mud seals to age. During the aging process, air penetrates through the mud seals of the pottery and slowly advances until contact catalysis oxidation reactions occur on the inner surfaces of ceramic pottery. Table 3 shows the volatile compounds and their contents in the samples aged for 1, 3 and 5 years and sourced from two different geographic origins (Shanghai and Shaoxing city in the Zhejiang province). From Table 3, it can be observed that the levels of some esters in the Chinese rice wine samples increased with wine age, increasing as esterification occurred. These esters included ethyl acetate , isobutyl acetate, isoamyl formate, ethyl lactate, phenethyl acetate, γ-decalactone and ethyl phenylacetate. However, for some ethyl esters, such as ethyl hexanoate, ethyl octanoate, ethyl myristate, ethyl oleate and ethyl linoleate, there was a decrease with wine age, increasing as the excess fatty acid esters formed by yeast were hydrolyzed during the aging process (Escudero et al., 2004). The reduction of furfural to furfuryl alcohol (and further products) occurred during the aging process. Furthermore, the total aldehyde contents increased as wine age increased, which reflected the oxidation reaction of alcohols that occurs during aging (Chen and Xu, 2010), and the results obtained in this study followed this trend. Additionally, Table 3 indicates that the samples aged for 5 years has the highest contents of free fatty acids, followed by the wines aged for 3 years and then those aged for 1 year. These free fatty acids may be formed by the oxidization of primary alcohols during the aging process.

Table 2. Volatile compounds in the Chinese rice wine samples sourced from different geographic origins (mg/kg)

RIa	Code	Volatile compoundsb	Sample group code
			JF5	TP5	FH	XT	SML
		Aldehydes
683	AD1	Acetaldehyde	1.15 ± 0.13	1.06 ± 0.27	0.89 ± 0.06	1.12 ± 0.03	1.37 ± 0.93
735	AD2	Butanal	3.64 ± 2.17	nd	nd	9.21 ± 2.72	0.40 ± 0.07
1035	AD3	2-Methyl-2-butenal	0.22 ± 0.14	0.21 ± 0.05	0.44 ± 0.24	nd	0.18 ± 0.07
1482	AD4	Furfural	1.81 ± 0.14	3.18 ± 0.14	1.68 ± 0.13	1.55 ± 0.33	0.96 ± 0.35
1542	AD5	Benzaldehyde	4.93 ± 0.95	12.02 ± 1.84	6.20 ± 0.17	5.05 ± 0.05	1.58 ± 0.56
1089	AD6	Hexanal	nd	nd	1.45 ± 2.16	nd	0.20 ± 0.04
1891	AD7	2-Phenyl-2-butenal	1.44 ± 0.15	2.51 ± 0.32	2.56 ± 1.68	1.54 ± 0.19	0.55 ± 0.30
1625	AD8	2-Pentylfuran	0.06 ± 0.10	nd	0.30 ± 0.08	0.14 ± 0.04	0.18 ± 0.05
1734	AD9	1,1-Diethoxy-butanal	nd	nd	0.28 ± 0.17	0.15 ± 0.01	0.08 ± 0.07
		Total	13.25 ± 3.77	18.97 ± 2.62	13.81 ± 4.66	18.76 ± 3.36	5.49 ± 2.44
		Esters
865	ES1	Ethyl formate	1.69 ± 0.14	1.67 ± 0.43	1.76 ± 0.35	0.20 ± 0.06	0.45 ± 0.40
894	ES2	Ethyl acetate	46.29 ± 4.07	54.89 ± 12.40	30.91 ± 3.21	37.02 ± 4.94	36.01 ± 14.45
964	ES3	Ethyl propanoate	1.34 ± 0.20	1.43 ± 0.26	1.00 ± 0.07	1.61 ± 0.31	0.72 ± 0.32
978	ES4	Ethyl isobutyrate	2.80 ± 0.85	1.61 ± 0.40	1.57 ± 0.14	2.19 ± 0.27	1.10 ± 0.37
1020	ES5	Isobutyl acetate	0.68 ± 0.23	0.51 ± 0.11	0.38 ± 0.17	0.53 ± 0.05	0.55 ± 0.18
1018	ES6	Ethyl butyrate	0.80 ± 0.27	4.24 ± 0.29	1.16 ± 0.05	1.89 ± 0.28	0.42 ± 0.10
1060	ES7	Ethyl 2-methylbutyrate	0.32 ± 0.13	0.46 ± 0.09	0.41 ± 0.27	0.41 ± 0.11	0.14 ± 0.11
1076	ES8	Ethyl isovalerate	0.73 ± 0.14	1.53 ± 0.39	0.78 ± 0.14	0.64 ± 0.25	0.17 ± 0.11
1129	ES9	Isopentyl acetate	2.42 ± 0.39	1.50 ± 0.25	1.37 ± 0.37	1.79 ± 0.11	1.78 ± 0.92
1118	ES10	Ethyl valerate	0.38 ± 0.19	0.18 ± 0.03	0.88 ± 0.26	0.17 ± 0.12	0.16 ± 0.14
1080	ES11	Isoamyl formate	49.69 ± 2.47	32.87 ± 1.69	38.11 ± 4.38	49.14 ± 2.70	30.43 ± 5.75
1241	ES12	Ethyl hexanoate	1.82 ± 0.26	1.64 ± 0.14	1.24 ± 0.12	1.65 ± 0.44	0.63 ± 0.25
1356	ES13	Ethyl lactate	9.58 ± 1.28	8.10 ± 0.98	4.62 ± 0.83	5.53 ± 2.27	2.16 ± 0.37
1442	ES14	Ethyl octanoate	0.71 ± 0.39	0.43 ± 0.11	0.44 ± 0.13	0.51 ± 0.40	0.27 ± 0.13
1540	ES15	Ethyl nonanoate	0.33 ± 0.06	nd	0.37 ± 0.10	0.05 ± 0.01	0.09 ± 0.04
1414	ES16	Butyl crotonate	1.95 ± 0.48	0.37 ± 0.32	1.26 ± 0.41	1.09 ± 0.48	0.34 ± 0.15
1280	ES17	Octyl formate	0.30 ± 0.19	0.13 ± 0.03	0.33 ± 0.10	0.13 ± 0.05	0.20 ± 0.10
1660	ES18	γ-Butyrolactone	0.13 ± 0.04	0.20 ± 0.02	0.75 ± 0.26	0.19 ± 0.06	0.31 ± 0.07
1688	ES19	Diethyl succinate	25.54 ± 3.76	2.04 ± 0.45	15.16 ± 10.15	18.65 ± 14.39	5.75 ± 2.04
1804	ES20	Ethyl phenylacetate	1.04 ± 0.08	2.22 ± 0.22	0.68 ± 0.43	1.27 ± 0.14	0.52 ± 0.18
1835	ES21	Phenethyl acetate	1.72 ± 0.09	1.15 ± 0.16	1.43 ± 0.26	1.11 ± 0.30	2.60 ± 0.25
1908	ES22	Butyl butyryllactate	0.39 ± 0.03	0.60 ± 0.61	0.50 ± 0.18	0.76 ± 0.57	11.78 ± 5.44
1914	ES23	Ethyl-3-Phenyl Propionate	0.24 ± 0.06	0.16 ± 0.07	0.72 ± 0.44	0.17 ± 0.03	0.20 ± 0.04
2158	ES24	Ethyl myristate	0.48 ± 0.32	0.55 ± 0.07	0.51 ± 0.21	nd	0.35 ± 0.20
1340	ES25	Ethyl heptanoate	nd	nd	nd	0.14 ± 0.04	nd
2180	ES26	γ-Decalactone	0.83 ± 0.30	0.96 ± 0.27	0.55 ± 0.29	0.71 ± 0.16	0.47 ± 0.07
1646	ES27	Ethyl caprate	3.49 ± 1.72	0.87 ± 0.82	0.62 ± 0.33	1.91 ± 0.29	0.78 ± 0.12
2276	ES28	Phenyl ethyl hexanoate	0.34 ± 0.24	nd	nd	0.24 ± 0.17	nd
1986	ES29	Ethyl oleate	nd	0.76 ± 0.17	nd	nd	nd
1977	ES30	Ethyl linoleate	nd	1.01 ± 0.20	nd	nd	nd
		Total	156.03 ± 18.38	122. 08 ± 20.98	107. 51 ± 23.65	129. 70 ± 28.99	98.38 ± 32.33
		Ketones
1296	KE1	3-Hydroxy-2-butanone	nd	nd	nd	nd	0.49 ± 0.21
1290	KE2	2-Octanone	0.09 ± 0.03	0.59 ± 0.17	nd	0.03 ± 0.01	nd
1625	KE3	Acetophenone	nd	nd	0.40 ± 0.16	nd	0.32 ± 0.06
1828	KE4	3-Methyl-2(5H)-furanone	0.13 ± 0.02	0.10 ± 0.03	0.35 ± 0.08	0.10 ± 0.01	0.53 ± 0.21
		Total	0.22 ± 0.05	0.69 ± 0.20	0.75 ± 0.23	0.13 ± 0.01	1.34 ± 0.48
		Alcohols
1034	AL1	1-Propanol	0.91 ± 0.21	0.61 ± 0.09	0.86 ± 0.06	1.08 ± 0.09	0.97 ± 0.34
1027	AL2	2-Butanol	8.85 ± 1.31	3.51 ± 0.31	8.75 ± 2.70	9.94 ± 1.03	7.46 ± 1.56
1148	AL3	1-Butanol	0.45 ± 0.03	0.33 ± 0.09	0.44 ± 0.13	nd	0.30 ± 0.11
1469	AL4	Heptanol	nd	nd	nd	0.12 ± 0.06	nd
1362	AL5	1-Hexanol	0.51 ± 0.14	0.22 ± 0.03	0.28 ± 0.11	0.29 ± 0.12	0.16 ± 0.03
1129	AL6	2-Pentanol	0.58 ± 0.18	0.32 ± 0.06	0.47 ± 0.16	0.47 ± 0.16	0.51 ± 0.11
1678	AL7	Furfuryl alcohol	nd	nd	nd	0.05 ± 0.01	nd
1898	AL8	Benzyl alcohol	0.15 ± 0.08	nd	0.65 ± 0.16	0.05 ± 0.01	0.21 ± 0.11
1935	AL9	Phenylethyl alcohol	26.42 ± 1.29	30.68 ± 0.91	28.40 ± 2.81	31.05 ± 2.30	19.77 ± 1.16
		Total	37.87 ± 1.36	35.66 ± 1.12	39.85 ± 6.13	43.05 ± 3.77	29.38 ± 3.41
		Acids
1481	AC1	Acetic acid	0.62 ± 0.18	0.50 ± 0.22	1.49 ± 1.59	0.21 ± 0.23	0.83 ± 0.84
1721	AC2	Isovaleric acid	0.27 ± 0.28	0.24 ± 0.09	0.45 ± 0.13	0.12 ± 0.02	0.40 ± 0.07
1871	AC3	Hexanoic acid	0.28 ± 0.07	0.29 ± 0.06	0.33 ± 0.05	nd	0.26 ± 0.08
2077	AC4	Octanoic acid	0.44 ± 0.07	0.29 ± 0.04	1.02 ± 0.52	0.29 ± 0.24	0.45 ± 0.24
2185	AC5	Nonanoic acid	0.46 ± 0.10	0.28 ± 0.09	0.48 ± 0.51	1.96 ± 0.29	0.48 ± 0.09
2296	AC6	Decanoic acid	0.39 ± 0.06	0.26 ± 0.02	1.00 ± 0.13	2.03 ± 0.63	0.65 ± 0.11
2401	AC7	Benzoic acid	0.49 ± 0.22	0.65 ± 0.14	2.58 ± 1.02	0.39 ± 0.13	0.42 ± 0.12
		Total	2.95 ± 0.98	2.51 ± 0.65	7.35 ± 3.39	5.00 ± 1.54	3.49 ± 1.55
		Phenols
2038	PH1	Phenol	0.11 ± 0.01	0.19 ± 0.05	0.72 ± 0.07	0.14 ± 0.03	0.36 ± 0.08
2227	PH2	4-Ethylguaiacol	nd	0.14 ± 0.01	0.32 ± 0.17	0.28 ± 0.22	0.26 ± 0.33
2206	PH3	4-Ethylphenol	0.22 ± 0.10	0.31 ± 0.03	nd	0.41 ± 0.10	0.22 ± 0.08
		Total	0.33 ± 0.11	0.64 ± 0.09	1.04 ± 0.23	0.83 ± 0.35	0.84 ± 0.49
		Miscellaneeous
1035	MI1	2-Ethoxythiazole	nd	nd	0.25 ± 0.12	0.71 ± 0.04	0.48 ± 0.09

a  Retention index of volatile compounds calculated on HP-INNOWAX (30 m × 0.25 µm i.d. × 0.25 µm, Agilent) column;

b  Volatile Compounds identified by the SBSE-GC-MS; nd: not detected by SBSE-TD-GC-MS.

Table 3. Volatile compounds and their content in the samples of 1-year, 3-year, 5-year age sourced from two different geographic origins (mg/kg)

Code	Sample group code
	JF1	JF3	JF5	TP1	TP3	TP5
Aldehydes
AD1	0.45 ± 0.03	0.60 ± 0.10	1.15 ± 0.13	0.74 ± 019	0.63 ± 0.06	1.06 ± 0.27
AD2	nd	nd	3.64 ± 2.17	0.05 ± 0.01	0.08 ± 0.01	nd
AD3	nd	nd	0.22 ± 0.14	nd	0.44 ± 0.43	0.21 ± 0.05
AD4	1.26 ± 0.14	1.53 ± 0.06	1.81 ± 0.14	2.50 ± 0.68	3.25 ± 0.66	3.18 ± 0.14
AD5	4.02 ± 0.26	5.99 ± 0.25	4.93 ± 0.95	8.66 ± 1.30	10.21 ± 0.49	12.02 ± 1.84
AD6	2.17 ± 0.05	3.31 ± 0.09	nd	nd	nd	nd
AD7	0.77 ± 0.17	0.94 ± 0.06	1.44 ± 0.15	1.44 ± 0.30	2.80 ± 0.73	2.51 ± 0.32
AD8	0.16 ± 0.06	0.19 ± 0.02	0.06 ± 0.10	0.15 ± 0.01	0.19 ± 0.03	nd
AD9	nd	nd	nd	0.20 ± 0.01	0.32 ± 0.10	nd
Total	8.83 ± 0.19	12.56 ± 0.23	13.25 ± 3.77	13.74 ± 2.50	17.92 ± 2.11	18.98 ± 2.62
Esters
ES1	nd	nd	1.69 ± 0.14	1.19 ± 0.03	1.05 ± 0.16	1.67 ± 0.43
ES2	29.02 ± 2.13	38.13 ± 9.69	46.29 ± 4.07	46.86 ± 4.67	49.96 ± 1.81	54.89 ± 12.4
ES3	0.97 ± 0.17	1.37 ± 0.41	1.34 ± 0.20	1.49 ± 0.27	1.40 ± 0.29	1.43 ± 0.26
ES4	1.60 ± 1.37	1.40 ± 0.14	2.80 ± 0.85	0.64 ± 0.18	0.61 ± 0.27	1.61 ± 0.40
ES5	0.50 ± 0.07	0.53 ± 0.03	0.68 ± 0.23	nd	nd	0.51 ± 0.11
ES6	0.73 ± 0.01	1.10 ± 0.38	0.80 ± 0.27	5.87 ± 0.76	5.47 ± 0.73	4.24 ± 0.29
ES7	0.31 ± 0.07	0.32 ± 0.03	0.32 ± 0.13	0.39 ± 0.19	0.50 ± 0.29	0.46 ± 0.09
ES8	1.32 ± 0.57	1.44 ± 0.22	0.73 ± 0.14	1.56 ± 0.17	2.17 ± 0.64	1.53 ± 0.39
ES9	2.32 ± 0.41	3.09 ± 0.62	2.42 ± 0.39	1.78 ± 0.15	1.55 ± 0.94	1.50 ± 0.25
ES10	0.39 ± 0.05	1.11 ± 0.98	0.38 ± 0.19	0.68 ± 0.65	1.82 ± 1.07	1.63 ± 0.30
ES11	42.76 ± 0.55	45.85 ± 3.04	49.69 ± 2.47	36.13 ± 2.45	32.26 ± 1.88	32.87 ± 1.69
ES12	1.71 ± 0.17	4.40 ± 2.50	1.82 ± 0.26	2.47 ± 0.14	2.41 ± 0.073	1.64 ± 0.14
ES13	8.20 ± 1.09	8.57 ± 0.60	9.58 ± 1.28	8.63 ± 1.38	9.27 ± 1.42	8.10 ± 0.98
ES14	0.46 ± 0.42	1.87 ± 1.84	0.71 ± 0.39	1.28 ± 0.46	0.95 ± 0.13	0.43 ± 0.11
ES15	2.18 ± 0.06	1.68 ± 0.42	0.33 ± 0.06	nd	nd	nd
ES16	nd	nd	1.95 ± 0.48	0.28 ± 0.09	0.24 ± 0.09	0.37 ± 0.32
ES17	nd	nd	0.30 ± 0.19	0.12 ± 0.01	0.21 ± 0.06	0.13 ± 0.03
ES18	0.15 ± 0.01	0.28 ± 0.20	0.13 ± 0.04	0.16 ± 0.04	0.18 ± 0.03	0.20 ± 0.02
ES19	25.03 ± 6.16	32.17 ± 7.79	25.54 ± 3.76	27.64 ± 2.97	31.48 ± 40.14	23.22 ± 2.66
ES20	0.78 ± 0.05	1.08 ± 0.06	1.04 ± 0.08	1.29 ± 0.13	2.12 ± 0.63	2.22 ± 0.22
ES21	1.10 ± 0.23	1.47 ± 0.07	1.72 ± 0.09	0.97 ± 0.02	2.00 ± 1.49	1.15 ± 0.16
ES22	nd	nd	0.39 ± 0.03	0.62 ± 0.32	1.47 ± 0.06	0.60 ± 0.61
ES23	0.28 ± 0.10	0.48 ± 0.17	0.24 ± 0.06	0.12 ± 0.01	0.83 ± 1.16	0.16 ± 0.07
ES24	nd	nd	0.48 ± 0.32	1.29 ± 0.27	0.78 ± 0.08	0.55 ± 0.07
ES25	0.12 ± 0.01	0.38 ± 0.21	nd	nd	0.18 ± 0.05	nd
ES26	0.65 ± 0.16	0.73 ± 0.18	0.83 ± 0.30	0.78 ± 0.12	0.74 ± 0.15	0.96 ± 0.27
ES27	0.37 ± 0.03	0.54 ± 0.22	0.23 ± 0.13	0.40 ± 0.35	0.15 ± 0.03	0.87 ± 0.82
ES28	0.38 ± 0.08	0.60 ± 0.072	0.34 ± 0.24	nd	nd	nd
ES29	0.31 ± 0.17	0.52 ± 0.19	nd	1.21 ± 0.17	0.99 ± 0.11	0.76 ± 0.17
ES30	0.77 ± 0.11	1.43 ± 0.65	nd	1.89 ± 0.31	1.38 ± 0.31	1.01 ± 0.20
Total	122.41 ± 13.25	150.54 ± 30.64	152.77 ± 20.46	145.74 ± 16.38	152.17 ± 17.98	144.71 ± 20.96
Ketones
KE1	0.16 ± 0.013	0.15 ± 0.02	nd	0.08 ± 0.01	0.17 ± 0.03	nd
KE2	0.56 ± 0.062	0.57 ± 0.18	0.09 ± 0.03	0.54 ± 0.06	0.56 ± 0.03	0.59 ± 0.17
KE3	nd	nd	nd	0.29 ± 0.01	0.32 ± 0.01	nd
KE4	nd	nd	0.13 ± 0.02	0.12 ± 0.01	0.14 ± 0.02	010 ± 0.03
Total	0.72 ± 0.070	0.72 ± 0.17	0.22 ± 0.05	1.03 ± 0.04	1.19 ± 0.09	0.69 ± 0.05
Alcohols
AL1	0.56 ± 0.02	0.44 ± 0.02	0.91 ± 0.21	nd	nd	0.61 ± 0.09
AL2	6.75 ± 0.87	6.79 ± 0.07	8.85 ± 1.31	3.74 ± 0.55	3.49 ± 1.10	3.51 ± 0.31
AL3	2.36 ± 3.50	1.37 ± 0.76	0.45 ± 0.03	0.44 ± 0.17	0.20 ± 0.03	0.33 ± 0.09
AL4	0.23 ± 0.01	0.26 ± 0.02	nd	nd	nd	nd
AL5	0.17 ± 0.05	0.23 ± 0.02	0.51 ± 0.14	0.23 ± 0.04	0.25 ± 0.02	0.22 ± 0.03
AL6	0.30 ± 0.14	0.47 ± 0.20	0.58 ± 0.18	0.18 ± 0.01	0.35 ± 0.07	0.32 ± 0.06
AL7	0.34 ± 0.02	0.32 ± 0.13	nd	0.33 ± 0.05	0.19 ± 0.01	nd
AL8	0.17 ± 0.02	0.15 ± 0.02	0.15 ± 0.08	0.16 ± 0.01	0.20 ± 0.01	nd
AL9	25.48 ± 3.48	30.41 ± 4.61	26.42 ± 1.29	31.93 ± 1.47	32.51 ± 4.37	30.68 ± 0.91
Total	36.36 ± 8.10	40.44 ± 5.85	37.87 ± 1.36	37.01 ± 2.30	37.19 ± 4.61	35.67 ± 1.12
Acids
AC1	0.59 ± 0.05	0.67 ± 0.12	0.62 ± 0.18	0.70 ± 0.09	0.60 ± 0.18	0.50 ± 0.22
AC2	nd	nd	0.27 ± 0.28	0.18 ± 0.02	0.12 ± 0.01	0.24 ± 0.09
AC3	nd	nd	0.28 ± 0.07	0.19 ± 0.05	1.09 ± 0.64	0.29 ± 0.06
AC4	0.49 ± 0.08	0.36 ± 0.07	0.44 ± 0.07	0.64 ± 0.24	0.38 ± 0.03	0.29 ± 0.04
AC5	nd	nd	0.46 ± 0.10	0.28 ± 0.10	0.21 ± 0.05	0.28 ± 0.09
AC6	0.45 ± 0.13	0.22 ± 0.02	0.39 ± 0.06	0.90 ± 0.27	0.27 ± 0.07	0.26 ± 0.02
AC7	4.33 ± 0.08	7.33 ± 0.50	0.49 ± 0.22	nd	nd	0.65 ± 0.14
Total	5.86 ± 1.06	8.58 ± 0.71	2.95 ± 0.98	2.89 ± 0.76	2.67 ± 0.99	2.51 ± 0.65
Phenols
PH1	0.30 ± 0.07	0.36 ± 0.27	0.11 ± 0.01	0.26 ± 0.02	0.21 ± 0.02	0.19 ± 0.05
PH2	0.40 ± 0.30	0.20 ± 0.08	nd	nd	nd	0.14 ± 0.01
PH3	0.14 ± 0.01	0.19 ± 0.06	0.22 ± 0.10	0.29 ± 0.02	0.41 ± 0.11	0.31 ± 0.030
Total	0.84 ± 0.31	0.75 ± 0.40	0.33 ± 0.11	0.55 ± 0.01	0.62 ± 0.13	0.64 ± 0.08
Miscellaneeous
MI1	0.37 ± 0.24	0.27 ± 0.11	nd	nd	nd	nd

Characterization of the aroma profiles of Chinese rice wines by PCA and CA    PCA provided partial visualization of the data in a reduced-dimension plot. Figure 1 shows the PCA results for the 15 Chinese rice wine samples sourced from 5 geographic origins. Moreover, Figure 1 indicates that the first two principal components (PCs) accounts for 56.63% of the total variation across the samples. When examining the volatile distributions (Figure 1a), the major compounds positively contributing to PC1 (factor loadings >0.7) were aldehydes such as 2-methyl-2-butenal (AD3), hexanal (AD6) and 2-pentylfuran (AD8); esters such as octyl formate (ES17); ketones such as acetophenone (KE3) and 3-methyl-2(5H)-furanone (KE4); alcohols such as benzyl alcohol (AL8); acids such as isovaleric acid (AC2) and octanoic acid (AC4); and phenols such as phenol (PH1). Ethyl propanoate (ES3), ethyl phenylacetate (ES20) and 4-ethylphenol (PH3) were the major compounds on the negative side of PC1 (factor loadings < −0.7). The key compounds for PC2 were furfural (AD4), benzaldehyde (AD5), ethyl butyrate (ES6), ethyl isovalerate (ES8), ethyl oleate (ES29), ethyl linoleate (ES30) and 2-octanone (KE2). On the negative side of PC2, 1-propanol (AL1), 2-butanol (AL2) and 2-ethoxythiazole (MI1) were the major compounds (factor loadings < −0.7).

Fig. 1.

PCA results of the 15 Chinese rice wine samples sourced from 5 geographic origins

Figure 1 illustrates that the Chinese rice wine samples can be clearly divided into five groups according to their geographic origins. Therefore, it can be concluded that the geographic origin is highly related to Chinese rice wine, which is in accordance with previous studies (Hu et al., 2009).

Dendrograms results by CA    The similarities between the 15 samples of 5 geographic origins were calculated on the basis of the Euclidean distance, and the CA dendrogram is shown in Figure 2. Two different groups can be observed in this dendrogram. The first group included the JF, FH and XT samples, which corresponded with Shanghai, Jiaxing city in the Zhejiang province and Jiashan city in the Zhejiang province, respectively. The SML and TP samples formed the second group, which corresponded to Nantong city in the Jiangsu province and Shaoxing city in the Zhejiang province, respectively. Furthermore, the Euclidean distance between FH and XT is much closer. The Shaoxing rice wine represented traditional rice wine, which has a full-bodied aroma and mellow flavor, while the Shanghai rice wine was representative of the refreshing type Chinese rice wine that has an elegant aroma.

The CA dendrogram for the similarities between the 18 samples from the 2 brands (TP and JF) and 3 aging times (1, 3 and 5 years) is shown in Figure 3. There were two different groups. The first group was the 3 aging times (1, 3 and 5 years) for the JF rice wines, and the other group was the 3 aging times (1, 3 and 5 years) for the TP rice wines. Furthermore, it could be concluded that geographic origin was more effective than wine age on the character of Chinese rice wine. From Figure 3, it can also be observed that the 3 aging time (1, 3 and 5 years) from the 2 brands were clear discriminated. One reason for this clear discrimination is that numerous physical and chemical reactions occurred in the ceramic pottery during the aging process. From Figure 3, it can also be noted that the 1- and 3-year old samples remain close, which indicates these two Chinese rice wines had a high similarity of flavor, while the 5-year old Chinese rice wine had a flavor transition.

Fig. 2.

CA dendrogram for the 15 Chinese rice wine samples sourced from 5 different geographic origins.

Fig. 3.

CA dendrogram for the 18 Chinese rice wine samples of 1-year, 3-year, 5-year from 2 geographic origin. (Shanghai, Shaoxing city in the Zhejiang province).

Conclusions

This study represented the approaches taken to characterize Chinese rice wine using the volatile profiles and chemometric methods. SBSE-TD-GC-MS analysis indicated that the volatile profile of Chinese rice wine were esters, alcohols, aldehydes, acids, ketones, volatile phenols and miscellaneous compounds. The excellent performance of the discrimination models indicated that the volatile compounds are of central importance in ensuring the identity of Chinese rice wines.

Acknowledgments    This work was financially supported by the National Natural Science Foundation of China (No. 21105065) and A Foundation for the Author of National Excellent Doctoral Dissertation of PR China (No. 201059).

References

•  Alves,  R.F.,  Nascimento,  A.M.D., and  Nogueira,  J.M.F. (2005). Characterization of the aroma profile of Madeira wine by sorptive extraction techniques. Anal. Chim. Acta, 546, 11-21.
•  Calleja,  A., and  Falqué,  E. (2005). Volatile composition of Mecía wines. Food Chem., 90, 357-363.
•  Cao,  Y.,  Xie,  G.F.,  Wu,  C., and  Lu,  J. (2010). A study on characteristic flavor compounds in traditional chinese rice wine-guyue longshan rice wine. J. Inst. Brew., 116, 182-189.
•  Cates,  V.E. and  Meloan,  C. E. (1963). Separation of sulfones by gas chromatography. J. Chromatogr. A, 11, 472-478.
•  Chen,  S. and  Xu,  Y. (2010). The influence of yeast strain on the volatile flavor compounds of chinese rice wine. J. Inst. Brew., 116, 190-196.
•  Chen,  S.,  Luo,  T.,  Xu,  Y.,  Fan,  W.L., and  Zhao,  G. (2009). Effects of yeast strains and raw materials on β-phenylethanol production in Chinese rice wines. China Brewing, 4, 23-28 (in Chinese).
•  Escudero,  A.,  Gogorza,  B.,  Melus,  M. A.,  Ortin,  N.,  Cacho,  J., and  Ferreira,  V. (2004). Characterization of the aroma of a wine from maccabeo. key role played by compounds with low odor activity values. J. Agric. Food Chem., 52, 3516-3524.
•  Fan,  W.L. and  Qian,  M.C. (2006). Identification of aroma compounds in Chinese ‘Yanghe Daqu’ liquor by normal phase chromatography fractionation followed by gas chromatography/olfactometry. Flavour Fragr. J., 21, 333-342.
•  Gallart,  M.,  Francioli,  S.,  Viu-Marco,  A.,  López-Tamames,  E., and  Buxaderas,  S. (1997). Determination of free fatty acids and their ethyl esters in musts and wine. J. Chromatogr. A, 776, 283-291.
•  Gil,  M.,  Cabellos,  J.M.,  Arroyo,  T., and  Prodanov,  M. (2006). Characterization of the volatile fraction of young wines from the Denomination of Origin “Vinos de Madrid” (Spain). Anal. Chim. Acta, 563, 145-153.
•  Gu,  G. X. (1996). Technology of Brewing Wine. 2nd Ed., Light Industrial Press of China: Beijing, pp. 484-489.
•  Hayasaka,  Y.,  MacNamara,  K.,  Baldock,  G.A.,  Taylor,  R.L., and  Pollnitz,  A.P. (2003). Application of stir bar sorptive extraction for wine analysis. Anal. Bioanal. Chem., 375, 948-955.
•  Hernanz,  D.,  Gallo,  V.,  Recamales,  A.F.,  Meléndez-Martínez,  A.J.,  González-Miret,  M.L., and  Heredia,  F.J. (2009). Effect of storage on the phenolic content, volatile composition and colour of white wines from the varieties Zalema and Colombard. Food Chem., 113, 530-537.
•  Hu,  J.,  Chi,  G.H.,  Wu,  M. Y. W., and  Wang,  X.H. (2009). Identification of the producing areas of Chinese rice wine by use of flavoring compositions. Liquor-Making Science & Technology, 180, 17-19 (in Chinese).
•  Kwon,  Y.H.,  Jo,  S.J.,  Kim,  J.H., and  Ahn,  B.H. (2010). Fermentation characteristics and volatile compounds in Yakju made with various brewing conditions; glutinous rice and pre-treatment. Korean Journal of Microbiology and Biotechnology, 38, 46-52 (in Korean).
•  Lee,  H.S.,  Lee,  T.S., and  Noh,  B.S. (2007). Volatile flavor components in the mashes of Takju prepared using different yeasts. Korean Journal of Food Science and Technology, 39, 593-599.
•  León-Rodríguez,  A. de,  González-Hernández,  L.,  Barba de la Rosa,  A.P.,  Escalante-Minakata,  P., and  López,  M.G. (2006). Characterization of volatile compounds of mezcal, an ethnic alcoholic beverage obtained from Agave salmiana. J. Agric. Food Chem., 54, 1337-1341.
•  Luo,  T.,  Fan,  W.L., and  Xu,  Y. (2008). Characterization of volatile and semi-volatile compounds in Chinese rice wines by headspace solid phase micro-extraction followed by gas chromatography-mass spectrometry. J. Inst. Brew., 114, 172-179.
•  Maggi,  L.,  Zalacain,  A.,  Mazzoleni,  V.,  Alonso,  G.L., and  Salinas,  M. R. (2008). Comparison of stir bar sorptive extraction and solid-phase microextraction to determine halophenols and haloanisoles by gas chromatography-ion trap tandem mass spectrometry. Talanta, 75, 753-759.
•  Mo,  X.,  Xu,  Y., and  Fan,  W. (2010). Characterization of aroma compounds in Chinese rice wine qu by solvent-assisted flavor evaporation and headspace solid-phase microextraction. J. Agric. Food Chem., 58, 2462-2469.
•  Natera,  R.,  Castro,  R.,  García-Moreno,  M. V.,  Hernández,  M. J., and  García-Barroso,  C. (2003). Chemometric studies of vinegars from different raw materials and processes of production. J. Agric. Food Chem., 51, 3345-3351.
•  Ng,  L. K. (2002). Analysis by gas chromatography/mass spectrometry of fatty acids and esters in alcoholic beverages and tobaccos. Anal. Chim. Acta, 465, 309-318.
•  Owens,  J. D.,  Allagheny,  N.,  Kipping,  G., and  Ames,  J. M. (1997). Formation of volatile compounds during Bacillus subtilis fermentation of soya beans. J. Sci. Food Agric., 74, 132-140.
•  Pereira,  A.C.,  Reis,  M.S.,  Saraiva,  P.M., and  Marques,  J.C. (2010). Aroma ageing trends in GC/MS profiles of liqueur wines. Anal. Chim. Acta, 659, 93-101.
•  Perestrelo,  R.,  Nogueira,  J.M.F., and  Câmara,  J.S. (2009). Potentialities of two solventless extraction approaches-Stir bar sorptive extraction and headspace solid-phase microextraction for determination of higher alcohol acetates, isoamyl esters and ethyl esters in wines. Talanta, 80, 622-630.
•  Pinto,  U. M.,  Viana,  E. d. S.,  Martins,  M. L., and  Vanetti,  M. C. D. (2006). Detection of acylated homoserine lactones in gram-negative proteolytic psychrotrophic bacteria isolated from cooled raw milk. Food Control, 18, 1322-1327.
•  Rizz,  G. P. (1972). Mechanistic study of alkylpyrazine formation in model systems. J. Agric. Food Chem., 20, 1081-1085.
•  Rodrigues,  F.,  Caldeira,  M., and  Câmara,  J.S. (2008). Development of a dynamic headspace solid-phase microextraction procedure coupled to GC-qMSD for evaluation the chemical profile in alcoholic beverages. Anal. Chim. Acta, 609, 82-104.
•  Ryan,  D.,  Watkins,  P.,  Smith,  J.,  Allen,  M., and  Marriott,  P. (2005). Analysis of methoxypyrazines in wine using headspace solid phase microextraction with isotope dilution and comprehensive two-dimensional gas chromatography. J. Sep. Sci., 28, 1075-1082.
•  Shen,  F.,  Li,  F.Z.,  Liu,  D.L.,  Xu,  H.R.,  Ying,  Y.B., and  Li,  B.B.(2012). Ageing status characterization of Chinese rice wines using chemical descriptors combined with multivariate data analysis. Food Control, 25, 458-463.
•  Tredoux,  A.,  De Villiers,  A.,  Májek,  P.,  Lynen,  F.,  Crouch,  A., and  Sandre,  P. (2008). Stir Bar Sorptive Extraction combined with GC-MS analysis and chemometric methods for the classification of south african wines according to the volatile composition. J. Agric. Food Chem., 56, 4286-4296.
•  Wang,  J.G. and  Yi  S.M. (2001). Understand and develop yellow wine. China Brewing, 11, 8-10 (in Chinese).
•  Xiao,  Z.B.,  Dai,  S.P.,  Niu,  Y.W.,  Yu,  H.Y.,  Zhu,  J.C.,  Tian,  H.X., and  Gu,  Y.B. (2011). Discrimination of Chinese vinegars based on headspace solid-phase microextraction-gas chromatography mass spectrometry of volatile compounds and multivariate analysis. J. Food Sci., 76, 1125-1135.
•  Xu,  R. N.,  Bao,  Z. D.,  Pan,  Z. Z., and  Hu,  P. X. (2003). Maturization of Chinese rice wine. Liquor Making, 30, 50-52 (in Chinese).
•  Zhang,  J.,  Li,  L.,  Gao,  N.F.,  Wang,  D.P.,  Gao,  Q., and  Jiang,  S.P. (2010). Feature extraction and selection from volatile compounds for analytical classification of Chinese red wines from different varieties. Anal. Chim. Acta, 662, 137-142.