Volatiles and antioxidant activity of fermented Goji (Lycium Chinese) wine: Effect of different oak matrix (barrel, shavings and chips)

ABSTRACT

The effect of oak matrix (medium toast barrel, medium toast shavings, non-toast chips, light toast chips, medium toast chips, and heavy toast chips) on the volatiles and antioxidant activity in Goji wine, a novel functional but aroma lackluster wine, were investigated. Results showed that the oak matrix used for Goji wine during aging improved its volatile profile and enhanced its antioxidant activity, depending on the toast degree and particularly the type of oak matrix. With regard to the volatiles of Goji wine, the oak shavings sample was the most efficient for the esters formation and the release of oak-related compounds. The analysis of odour activity values of volatiles in Goji wine also showed that the flavour of woody, vanilla, and clove was the highest in the oak shavings sample, which were related to the compounds such as cis- and trans-whisky lactone, vanillin, eugenol, isoeugenol, and 4-vinylguaiacol. According to the sensory analysis, the oak shavings samples had the highest score in olfactory and gustative. In addition, results of the effect of oak matrix on antioxidant activity of Goji wine indicated that Goji wine treated with oak shavings exhibited the highest antioxidant activity thanks to the high content of phenols (5.47 g/L GAE) and flavonoids (0.36 g/L).

KEYWORDS:

• Goji wine
• oak matrix
• volatiles
• antioxidant activity
• shavings

Introduction

Goji berry (Lycium barbarum fruit), having functional effect such as anti-aging, antioxidant, enhancing vision, nourishing both kidney and liver, has been widely used as a traditional Chinese herb medicine and functional food.^[1,2] With the active ingredients extracted, Goji liquor steeped with xiaoqu liquor, has been produced for a long history in China.^[3] As a key functional component in Goji liquor, Lycium barbarum polysaccharide (LBP) has been found to play several roles in wine, such as controlling wine stability and improving organoleptic properties.^[4,5] With the effective extract of LBP, the production process of Goji wine was developed in our previous research.^[6] In addition, it’s well known that the antioxidant activity of wine is also closely related with the presence of phenolics.^[7] Wine polyphenols mainly include phenolic acids, flavonols, tannins and anthocyanins, and they play an important role during the oxidation of wine when aging in barrels, thus affecting their composition, organoleptic properties and stability.^[8,9]

Aging is a common winemaking practice to improve wine quality, and the aging process is traditionally carried out in oak barrel. Aging in oak barrel promotes colour, structure, and particularly aroma of wine; it is interesting to observe different reactions among phenolics and the extraction of several compounds from wood to wine. It is reported to be able to increase the complexity and stability of wine.^[10] It is also shown in literature that only a few wood volatiles have an obvious impact on the sensory characteristics of wine, such as syringol, furfural, β-methyl-γ-octalactone (particularly the cis isomer), eugenol, vanillin, syringaldehyde, guaiacol, and its derivatives (e.g., 4-vinylguaiacol).^[11] In addition to the characteristics of used wood (geographic origin, seasoning, toast, etc.), the quality of a wood-aged wine also depends on the factors that control the mass transfer, such as the concentration of alcoholic, type of process (static or dynamic), duration time, and contact surface between wine and wood.^[12,13] In addition, the use of oak fragment is often preferred because the aging in barrel is very expensive (long permanence of wine inside, high cost of oak casks).^[14] Wine treated with oak chips or shavings matures even more quickly than that wine aged in barrel.^[15] Recent studies on wine treated with oak chips also indicated that the volatiles and phenolic profiles were affected by botanical origins and toasting degree.^[14,16,17]

However, little research about the oak matrix-based aging of Chinese traditional fermented Goji wine was reported. Thus, the objective of the paper was to study the influence of various oak matrix (barrel, chips, and shaving) on the volatile profiles and antioxidant activity of Goji wine. To the best of our knowledge, this study may provide useful information to the Goji wine industry to evaluate quality and potential consumer acceptability.

Materials and methods

Materials and chemicals

Dried Goji berries were purchased from the local market (Chengdu City, Sichuan province, China). In addition, 10 L of new oak barrels (Quercus alba) in this study were purchased from huanzemuye (Tianjin City, China), with medium toast level (toasting at 200°C, for 2 h). With the same material of barrel, four types of oak chips (sized 1 cm × 0.5 cm × 0.1 cm) were employed: one non-toast and the others toasted for 2 h at three levels (light (L) toast chips with the surface temperature of = 150 ± 10°C; medium (M) = 200 ± 10^oC; heavy (H) = 230 ± 10^oC). Oak shavings were medium toasted and shaving size was 3–5 mm, which is the ideal size for optimum extraction of phenolics from wood using hydroalcohol mixtures according to Martinez et al.^[18]

Methyl caprylate (internal standard) was purchased from Sigma-Aldrich (St. Louis, MO, USA). Glucose, rutin and gallic acid; 1,1-diphenyl-2-picrylhydrazyl (DPPH), 2,2ʹ-azinobis(3-ethylbenzonthiazoline-6-sulfonic acid) (ABTS), trolox were obtained from Aladdin (Shanghai City, China). All reagents were of analytical grade.

Goji wine samples and wood maturations

A Goji wine fermentation trial was carried out in our laboratory according to the process described by Xia et al.^[6], and the procedure is summarized in Figure 1. Briefly, 20 kg of dried Goji berries were crushed with a coffee mill and mixed well with water by 1:10 (w/v). The initial contraction of sugar was adjusted to 15% (v/v). Saccharomyces cerevisiae 2.178 culture suspension (10⁵ cell/mL of the finial concentration) was inoculated into the mixture. The fermentation was carried out in a stainless steel container in an incubator at 12–15^oC for 15 d. Once finished, the obtained 200 L of fresh Goji wine was blended with edible alcohol to adjust the ethanol concentration into 30% (v/v). The major physicochemical properties of Goji wine were evaluated before aging: titrable acidity (express as citric acid): 0.92 g/100 mL; residual sugar (express as glucose): 0.49 g/100 mL. The fresh Goji wine sample was labelled as AG0.

Figure 1. Manufacture procedure of Goji wine samples.

After calculating the relation surface/volume of oak barrel, 4.0 g/L of oak chips for were added for each brown glass bottle of 10 L. As for the shavings, the quantity was 0.4% (w/v) proportion in each bottle. The barrels and bottles were filled up by Goji wine and then aged for 3 months with or without oak matrix (barrel, chips and shavings) under the same conditions (relative humidity of air and temperature were at approx. 75–85% and 15–25°C, respectively). The wine bottles were shaken for 3–5 min every month during the aging period and the aging experiment were carried out in triplicate. The Goji wine samples successively labelled as AG1 (without oak matrix as control), AG2 (non-toast chips), AG3 (light toast chips), AG4 (medium toast chips), AG5 (heavy toast chips), AG6 (medium toast shavings) and AG7 (medium toast barrel) (Figure 1).

Extraction of volatiles

A 50/30 μm DVB/CAR/PDMS fibre (Supelo, USA) was used for volatiles extraction. It validated that this fibre has been validated to be appropriate for trapping a range of volatiles with different polarities and efficient in covering the wide range of physical-chemical properties of flavour volatiles because of their intermediary polarity.^[19] To begin with, sample was diluted with deionized water to a final concentration of ethanol in 15% (v/v), and then it was saturated with sodium chloride. Head space-solid phase micro extraction (HS-SPME) conditions were based on reported method with slight modifications.^[20] 1 mL of Goji wine sample was added into a 15 mL vial and saturated with sodium chloride. 10 μL of methyl octanoate (877 mg/L in absolute methanol) was added into the wine samples as internal standard. The vial was tightly capped with a silicon septum. Then, it was equilibrated at 60°C in a magnetic stirring plate for 15 min and extracted for 30 min. Finally, the fibre was immediately introduced into the injection port of the gas chromatography (GC) to thermally desorb the analytes at 250°C for 5 min.

Gas chromatography-mass spectrometry (GC-MS) analysis

Samples were analysed on a Trace GC Ultra gas chromatography-DSQ II mass spectrometer (Thermo Electron Corporation, Waltham, MA, USA) equipped with a HP-INNOWAX capillary column (30.0 mm×0.25 mm i.d.,0.25-μm film thickness; Agilent, Santa Clara, CA, USA). The GC operation condition was achieved according to protocols reported previously.^[21] Briefly, an inlet temperature of 250°C, split ratio of 10:1, and helium (purity: 99.999%) carrier gas flow of 1 mL/min. The over temperature was held at 40°C for 5 min, followed by an increase of 5°C/min to 100°C and then programmed at 6°C/min until 220^oC, and held for 10 min. The temperatures of the transfer line, quadruple, and ionization source were 250, 150, and 230°C, respectively. The mass spectrum was generated in the electron impact (EI) mode at 70 eV. The molecular weight scanned was 40–400 amu. The constituents were tentatively identified by matching mass spectrum with NIST05 spectrum database and verified by comparison of their Kováts retention indices (RI) with the RI reported in literatures, which was calculated by using C8–C20 n-alkanes.^[22] Relative amounts (μg/L) of certain volatiles were calculated by the peak area ratio to the internal standard on GC total ion chromatograms.

Sensory analysis

Sensory analysis was carried out by a tasting panel consisting of 25 students, aged between 20 and 30 years old. These members were students of our college, and were trained repeatedly by professional wine connoisseur according to the process described by Frangipane et al.^[23] Different aspects of the samples were considered: visual, olfactory and gustative. Standard of environment (room, sample, taster) was as follows: odour – free, quiet, well-lit; temperature – 21°C, humidity – 45% to 55%; samples – anonymous and uniformly poured and covered, tasting at the same time. A volume of 30 mL of each sample was evaluated in tulip-shaped ISO (International Standardization Organization) tasting glasses at 20°C. With the form, samples were scored on a 10-point scale from 0 (poor) to 10 (good).

Determination of LBP, total flavonoid and total phenolics contents

Using glucose as standard, the LBP content in Goji wine was determined by the phenol–sulfuric acid method.^[24] The total flavonoid (TF) content in Goji wine was measured according to the method of Jia et al.^[25] The TF was calculated from a calibration curve with rutin as standard. The total phenolics (TP) content in Goji wine was determined by Atanacković et al.^[26] The TP content was expressed as g of gallic-acid equivalents (g/L GAE).

DPPH and ABTS free radical scavenging capacity

All wine samples were filtered through a membrane filter (0.45 μm) and diluted with methanol. The free radical scavenging activity of wine sample on DPPH was carried out as previous document.^[27] The method used for ABTS •scavenging assay was reported by Re et al.^[28] All the results for the DPPH (ABTS) scavenging capacity of wine samples were expressed as trolox equivalent antioxidant capacity (TEAC). The standard solutions of trolox ranged from 0 to 400 mg/L. The trolox standard curve was built by plotting inhibition percentage versus trolox concentration at 517 nm (734 nm) for DPPH (ABTS) assays.

Statistical analysis

Every sample was analysed in triplicate and results were means ± SD of replicates. The content of volatile compound was expressed as μg/L. Analysis of variance (ANOAV) with Duncan’s test was performed to evaluate significant differences in volatiles from different samples. Significant difference was defied at p < 0.05 (n = 3). One-way ANOAV was conducted using SPSS software (version 17.0). Odour activity values (OAVs) were calculated by dividing concentrations with their respective odour thresholds. An aroma-like profile was obtained by taking logarithm of OAVs. Origin 7.5 was used to create the figures in this study.

Results and discussion

Effect of oak matrix treatment on volatiles profile of Goji wine

The effect of oak matrix on the volatiles profiles of Goji wine is shown in Table 1. A total of 39 kinds of volatiles were identified in various Goji wine samples, sorted into seven families, including alcohols (4), esters (18), lactones (2), acids (6), aldehydes (5), ketones (1), and phenols (3). These volatiles were also reported in some grape wine.^[29–34] As for Goji wine before maturation (AG0), only benzyl alcohol, 2-phenylethyl acetate, and decanoic acid were present above its thresholds. There were only 3 compounds that played a major aromatic role in Goji wine before maturation, suggesting that fresh Goji wine was an aroma lackluster wine. In addition, the content of volatiles in Goji wine with natural maturation (AG1) had no distinct difference in comparison with AG0 (p > 0.05). By contrast, the content of volatiles in Goji wine treated with oak matrix changed significantly (p < 0.05). Compared with AG0, the total amount decreased by only 1.6% in AG1 but decreased by 11.7%-46.6% in oak matrix samples. Compared with AG1, the quantity of alcohols and acids in oak matrix samples decreased by 16.3%–56.5% and 17.7%–54.8%, respectively. The quantity of esters increased by 29.2%–65.8%. Moreover, nine kinds of new volatiles were only identified in oak matrix samples, resulting in the increase of lactones, aldehydes and phenols. These oak-related volatiles were cis- and trans-whisky lactone, furfural, 5-methyl-furfural, syringaldehyde, vanillin, eugenol, isoeugenol, and 4-vinylguaiacol.^[35] All of them were also reported in the literature and endowed Goji wine with special sensory.^[29] Therefore, the effect of aging naturally on the volatiles profiles in Goji wine could be ignored, and Goji wine needed aging in oak matrix (wooden barrel or other container with oak wood addition) to acquire wine organoleptic properties.^[36]

Table 1. Contents of volatiles in Goji wine before and after maturation.

		Concentration (μg/L)^α
RI*	Compounds	AG0	AG1	AG2	AG3	AG4	AG5	AG6	AG7	OTS (μg/L)^η	Odour descriptor	OAV^&
	Alcohols
1260	1-Pentanol	16128.76±1307.25^a	15571.63±1247.17^a	8836.68±163.01^c	6672.64±515.35^d	7048.54±47.80^d	10106.07±383.29^b	9790.42±82.32^bc	6482.71±112.26^d	64000	Bitter, almond, balsamic	<1
1299	3-Methyl-1-butanol	352.34±38.66^b	330.70±13.43^bc	359.01±5.98^b	313.04±2.28^c	262.67±26.74^d	231.84±5.73^d	436.06±1.33^a	358.77±44.79^b	30000	Burnt, alcohol	<1
1307	3-Methyl-1-pentanol	115.23±4.05^a	112.93±1.08^a	65.78±6.45^c	50.84±0.22^d	44.46±1.00^e	37.01±4.04^f	80.54±6.69^b	47.62±4.42^de	50000	Vinous, herbaceous, cocoa	<1
1846	Benzyl alcohol	87997.16±3922.04^a	86478.81±4114.88^a	58479.15±592.64^e	70932.95±383.82^c	37227.23±313.39^g	52106.30±1755.58^f	62787.52±301.44^d	78926.99±2379.06^b	1000	Floral, rose	>1
	Total	104593.49	102494.06	67740.63	77969.47	44582.90	62481.22	73094.54	85816.09
	Esters
985	Butyrolactone	4.03±0.36^f	4.21±0.25^f	37.43±6.50^d	73.90±3.80^b	32.58±1.09^d	23.28±0.58^e	47.97±0.18^c	103.60±8.47^a	500	NF
1228	Ethyl caproate	13.11±5.00^d	14.61±6.04^d	8.26±0.06^d	200.09±13.49^b	4.26±0.01^d	92.43±3.48^c	19.50±1.86^d	218.11±31.41^a	14	Green apple	0.59–15.58
1364	Ethyl lactate	32.17±23.80^abcd	45.88±2.03^a	17.73±0.65^de	16.00±0.04^e	22.97±1.28^cde	40.53±0.74^ab	36.84±0.14^abc	25.53±0.44^bcde	154636	acid, medicinal	<1
1434	Ethyl caprylate	349.11±16.63^de	357.43±11.76^d	846.11±59.66^a	629.63±25.77^b	318.83±15.96^e	497.21±3.36^c	599.87±1.64^b	622.28±2.78^b	580	Sweet, fruity, pear	0.55–1.46
1486	Methyl nonanoate	25.51±2.75^d	26.89±1.94^d	44.37±3.56^b	52.93±5.64^a	36.06±0.66^c	36.63±3.08^c	37.73±0.97^c	49.11±2.12^a	NF	NF
1624	Methyl ethyl butanedioate	55.27±1.23^d	55.91±0.79^d	151.74±3.88^b	134.25±6.24^c	134.59±2.72^c	128.90±9.79^c	165.97±0.54^a	132.59±4.29^c	NF	NF
1668	Ethyl phenylacetate	174.20±2.44^a	175.42±1.72^a	84.25±6.72^c	95.08±2.78^b	48.31±2.02^d	82.28±4.22^c	84.42±5.33^c	90.33±4.09^b	1800	Flowery	<1
1690	Diethyl succinate	736.96±22.79^cd	750.12±0.42^bc	661.73±48.85^e	780.02±12.48^b	428.26±46.68^g	523.12±13.36^f	711.25±1.12^d	914.55±6.01^a	200000	Wine	<1
1770	Diethyl glutarate	289.93±43.66^a	313.98±18.52^a	66.83±7.12^e	192.18±3.59^c	19.33±1.06^f	141.84±19.48^d	49.40±0.40^e	237.53±10.52^b	760000	Over-ripe, peach, prune	<1
1792	2-Phenylethyl acetate	4085.23±15.20^g	4092.83±10.75^g	7120.28±229.78^c	5872.20±12.79^e	9610.98±14.14^a	6241.79±58.98^d	7478.41±260.63^b	4995.38±15.15^f	250	Floral	>1
1800	Methyl laurate	272.28±43.14^cd	295.90±19.40^c	247.27±1.89^de	345.71±4.78^b	210.72±4.83^e	225.48±6.39^e	221.92±2.07^e	429.58±45.85^a	NF	NF
1995	Ethyl palmitate	180.67±43.21^f	204.84±15.15^f	724.17±36.09^c	895.74±55.43^b	610.18±21.17^d	555.56±6.08^e	1039.39±80.26^a	900.78±7.71^b	1500	Fatty, fruity, sweet	<1
>2000	Isopropyl myristate	404.92±3.50^a	406.68±2.48^a	149.55±0.78^f	207.06±36.61^d	64.92±3.17^g	272.31±1.36^b	239.83±5.57^c	179.95±8.57^e	800	NF	<1
>2000	Ethyl myristate	6.12±0.93^d	6.63±0.42^d	6.77±0.31^d	31.45±4.49^b	2.87±0.28^e	7.30±0.48^d	61.52±0.18^a	25.53±4.82^c	4000	NF	<1
>2000	Ethyl cinnamate	254.58±22.23^d	265.21±17.62^d	339.70±13.77^b	301.42±8.35^c	172.96±4.55^f	217.77±2.57^e	362.33±0.12^a	189.82±0.89^f	NF	NF
>2000	Ethyl monosuccinate	73.93±2.35^a	75.11±1.66^a	45.47±0.59^d	51.32±1.95^c	28.83±2.37^e	49.82±0.40^c	61.17±2.28^b	76.43±3.90^a	1000000	caramel，coffee	<1
>2000	Ethyl oleate	14.19±0.78^g	14.58±0.55^g	81.85±1.03^d	151.59±0.11^b	72.00±1.26^e	59.60±1.77^f	107.23±1.78^c	165.20±0.37^a	NF	NF
>2000	Ethyl (2E)-3-(4-hydroxy-3-methoxyphenyl)-2-propenoate	53.79±0.41^bc	53.99±0.29^bcd	46.28±4.94^e	58.83±0.52^a	53.71±3.44^bc	57.38±4.01^bc	52.64±7.25^bc	50.33±3.07^cd	NF	NF
	Total	7026.02	7160.21	10679.80	10089.40	11872.37	9253.25	11377.39	9406.63
	Lactones
1861	trans-Whisky lactone	–	–	223.60±6.42^b	186.31±1.41^c	122.52±6.87^d	84.55±2.02^f	342.28±6.52^a	112.03±6.91^e	32	Woody, coconut notes, vanilla	>1
1928	cis-Whisky lactone	–	–	100.93±6.07^b	89.96±3.11^c	63.81±1.57^e	68.42±3.48^d	120.69±0.28^a	70.26±3.65^d	74	Woody, coconut notes, vanilla	0.86–1.63
	Total	–	–	324.53	276.26	186.33	152.98	462.97	182.29
	Acids
1407	Acetic acid	143.72±2.76^e	142.34±1.95^e	265.47±6.66^c	252.83±0.96^cd	656.50±6.78^b	228.85±5.89^d	161.55±4.46^e	906.25±56.86^a	200000	Sour, pungent, vinegar	<1
1655	Isobutyric acid	36.13±1.46^a	36.86±1.03^a	–	15.02±0.02^c	12.05±0.19^d	9.13±0.74^e	–	19.47±1.39^b	200000	Cheese	<1
1876	Hexanoic acid	411.05±8.59^b	414.23±9.30^b	188.53±0.62^e	352.00±2.58^c	207.01±5.27^d	184.21±1.53^e	177.76±4.07^e	670.94±23.43^a	420	Sweat	0.42–1.60
>2000	Nonanoic acid	1064.31±45.21^a	1086.91±31.97^a	121.96±0.51^de	168.59±2.92^c	100.99±5.95^e	176.24±1.72^bc	137.36±0.84^d	203.01±1.41^b	3000	Fatty	<1
>2000	Decanoic acid	2472.49±70.60^a	2494.59±83.89^a	1273.31±19.85^d	1673.13±15.96^b	1329.09±10.43^cd	1308.31±40.15^cd	1326.10±5.27^cd	1372.96±44.84^c	6	Fatty, rancid	>1
>2000	Dodecanoic acid	161.86±3.44^de	163.34±3.28^cde	152.52±3.45^e	321.02±5.75^b	176.87±0.57^cd	182.49±7.41^c	158.17±3.78^de	396.36±41.29^a	1000	Dry, metallic	<1
	Total	4289.56	4338.27	2001.79	2782.60	2482.51	2089.22	1960.93	3568.99
	Aldehydes
1260	2,5-Dimethoxy-benzaldehyde	497.24±11.93^c	503.75±5.52^c	287.25±15.79^e	523.62±3.76^b	259.63±1.58^f	330.99±3.89^d	157.06±9.11^g	602.00±15.76^a	NF	NF
1448	Furfural	–	–	271.55±4.05^e	287.91±1.41^e	626.97±4.26^d	1172.84±86.72^b	1514.34±46.06^a	969.86±32.83^c	20000	Burned almonds, incense, floral	<1
1551	5-Methyl-furfural	–	–	8.30±0.26^f	27.33±0.16^e	82.04±5.69^d	726.70±16.02^b	811.77±48.28^a	390.72±15.40^c	16000	Bitter almond, spice	<1
>2000	Vanillin	–	–	39.35±2.10^d	41.80±2.85^d	176.87±5.90^b	87.48±3.16^c	606.46±52.72^a	164.91±5.50^b	60	Vanillin	0.66–10.11
>2000	Syringaldehyde	–	–	627.43±9.83^f	913.64±26.58^e	1508.02±11.77^b	1453.92±14.09^c	1647.21±25.61^a	1188.62±27.68^d	25000	Vanilla	<1
	Total	497.24	503.75	1233.88	1794.30	2653.53	3771.94	4736.83	3316.12
	Ketones
1585	2-Methyl-cyclopentanone	2.76±0.13^g	2.83±0.09^g	7.66±0.62^e	4.42±0.03^f	11.61±0.54^c	9.95±0.31^d	26.23±0.14^a	14.19±2.07^b	NF	NF
	Total	2.76	2.83	7.66	4.42	11.61	9.95	26.23	14.19
	Phenols
>2000	Eugenol	–	–	117.47±3.03^e	125.35±1.63^d	159.85±3.91^a	152.22±1.41^b	161.94±5.28^a	133.09±3.71^c	6	Spices, clove, honey	>1
>2000	Isoeugenol	–	–	12.94±0.15^d	14.28±0.34^d	42.76±1.41^b	41.99±0.94^b	53.95±3.02^a	23.80±5.17^c	6	Clove	>1
>2000	4-Vinylguaiacol	–	–	26.63±0.31^f	38.56±2.47^e	219.98±0.51^b	76.52±0.14^c	258.37±25.17^a	53.77±1.89^d	40	Spicy, curry	0.67–6.46
	Total	–	–	157.03	178.19	422.59	270.73	474.26	472.17

Note: Different superscripted letters indicate significant differences (p < 0.05).

*: Retention indices (RI) are determined by using n-alkanes C₈-C₂₀.

α: Values are means of triplicate replicates ± SD (n = 3).

η: Odour threshold obtained from literature:^{[29,30,48,49]}

&: Odour activity value.

NF: Not Found, odour threshold or odour descriptors are not available in the literature.

–: No data.

As mentioned before, the effect of oak matrix on the volatiles in Goji wine depended on the toast degree and the type of oak matrix. A detailed analysis was therefore performed to demonstrate the effect of these two factors. Among oak chips samples, the total content of alcohol and esters exhibited the greatest change in AG4 and the total content of ketones in AG2 increased the most. Among the four compounds in alcohols, the content of 3-methyl-1-pentanol decreased with the toast intensity. As for oak-related volatiles, content of cis- and trans-whisky lactone decreased during toast process because of lipid pyrolysis,^[37] and their contents were the lowest in AG5. However, furfural and 5-methyl-furfural were the major furan derivatives detected, and their contents increased with toast degree,^[11] which were the lowest in AG2. Besides, the content of phenolic aldehydes (syringaldehyde, vanillin) and lignin-derived volatile phenols (eugenol, isoeugenol and 4-vinylguaiacol) increased with toast level except for heavy toast degree, it is therefore the highest in AG4 (medium toast degree).

According to samples at medium toast degree, the quantity of alcohols and acids decreased by 28.7% and 54.8% in AG6, and decreased by 17.7% and 16.3% in AG7. While the quantity of esters increased by 58.9% in AG6, and increased by 31.4% in AG7. The quantity of esters in AG6 and AG4 were higher than that in AG7. Compared with AG1, the content of 1-pentanol decreased the most in AG6 and AG7 and the content of 3-methyl-1-pentanol decreased by 60.6% in AG4. Samples at medium degree also changed significantly in acids and esters. Among the six compounds in acids, the contents of all the compounds in AG7 were higher than that in AG4 and AG6. For the esters, the contents of 2-phenylethyl acetate and ethyl (2E)-3-(4-hydroxy-3-methoxyphenyl)-2-propenoate were higher in AG4. The contents of ethyl lactate, ethyl myristate, ethyl cinnamate, ethyl palmitate, isopropyl myristate, and methyl ethyl butanedioate were higher in AG6. The contents of other compounds in esters were higher in AG7. In addition, all of the oak-related volatiles had the highest content in AG6 and the content of those compounds in AG7 had no significantly different from others (p > 0.05). In general, the treatment with medium toast shavings was beneficial for the esters formation and the release of oak aroma compounds into Goji wine during aging.

Effect of oak matrix treatment on the flavour profile of Goji wine

OAV can reflect the contribution of each aroma compound to the wine sample characteristic flavour, and compounds with OAV >1 contribute to the overall aroma significantly.^[38,39] In order to study the effect of oak matrix on the flavour profile of Goji wine, the OAVs of volatiles are presented in Table 1. Among 39 kinds of identified volatiles, there were 12 kinds of compounds with their OAV changed significantly. For the alcohol, only benzyl alcohol demonstrated OAV >1. Benzyl alcohol is an important compound in some China liquors^[40] and grape wine,^[41] which has a floral and rose aroma. The OAVs of benzyl alcohol ranging from 37.23 to 86.48, decreased in oak matrix samples especially in AG4. Among the 18 identified esters, the OAVs of 2-phenylethyl acetate increased in oak matrix samples particularly in AG4, which increased by almost two times. Nevertheless, the OAVs of ethyl caproate and ethyl caprylate were the lowest in AG4 with OAV<1 but they were the highest in AG2 (1.46) and AG7 (15.58), respectively. 2-Phenylethyl acetate, ethyl caproate, and ethyl caprylate were rich in the flavour of floral, green apple and sweet, fruity, which were of great importance to the sensory of Goji wine.^[30] In addition to acid, decanoic acid gave fatty and rancid odours to Goji wine and the OAVs of decanoic acid had little difference among oak matrix samples. Hexanoic acid contributed a sweat note to Goji wine, whose OAV>1 only in AG7. As for the oak-related volatiles, cis- and trans-whisky lactone, vanillin, eugenol, isoeugenol and 4-vinylguaiacol were considered as the dominant compounds. Cis- and trans-whisky lactone contribute resin fragrance into products, which was one of the characteristic flavour in wine.^[10] Vanillin, eugenol, isoeugenol, and 4-vinylguaiacol endowed the wine with clove and curry flavour.^[29] The OAVs of all the six compounds with the flavour of woody, vanilla and clove, were the highest in AG6. As suggested by the results, Goji wine with medium toast shavings treatment had higher OAVs of cis- and trans-whisky lactone, vanillin, eugenol, isoeugenol, and 4-vinylguaiacol.

In order to further reveal the distinction among the samples of oak treated Goji wine, an aroma-like profile with a five-point scale from −1 (little contribution) to 3 (high contribution) was constructed to compare the flavour feature on Goji wine sample with and without oak matrix (Figure 2). Compared with the control sample, flavours of sweet and fruity were increased in samples with non-toast chip and no significant change in flavour was induced by the samples with different toast chips. Intense flavours of woody, vanilla, and clove notes were enhanced by samples with medium toast shavings and that of green apple was rich in samples of oak barrel. In one word, oak shavings sample was benefited to the release of woody, vanilla and clove flavour, which were related to the high OAVs of compounds such as cis- and trans-whisky lactone, vanillin, eugenol, isoeugenol, and 4-vinylguaiacol.

Figure 2. Analysis of OAV profile of Goji wine treated with and without oak matrix.

According to sensory analysis, the average score of Goji wine samples for four attributes of visual, olfactory, gustative, and texture is investigated in Figure 3. Compared with the control sample, samples with oak matrix had higher score in visual and olfactory. Among oak chips samples, samples at medium toast degree were the highest in all the four attributes. Among the samples at medium toast degree, oak chips and shavings samples were higher than oak barrel sample for all the four attributes, especially oak shavings sample had the highest score in olfactory and gustative.

Figure 3. Sensory scores of Goji wine treated with and without oak matrix.

Different letters indicate statistically significant differences for p < 0.05.

Analysis of oak matrix treatment on antioxidant activity of Goji wine

The changes of LBP, phenolics, flavonoids in Goji wine treated with and without oak matrix were carried out (Table 2). LBP had a major impact on the antioxidant properties of Goji wine^[42] and the changes of LBP during aging via different oak matrix were of no obvious difference (p > 0.05). Phenolics play an important role in flavour stability and colloidal stability of wine, which are also considered to be important antioxidant sources in wine.^[43] The Goji wine samples exhibited significant differences in TP content, varying from 4.25 g/L GAE for AG1 to 5.47 g/L GAE for AG6, which were higher than those of red wine (1.40–3.13 g/L GAE), white wine (0.19–0.50 g/L GAE), and rosé wine (0.74–1.09 g/L GAE) reported by Li et al.^[44] The contents of TP were increased in oak matrix samples and decreased with toast degree for the oak chips samples. It was in agreement with the findings of Alañón et al.^[8] Regarding to samples at medium degree, the contents of TP were the highest in AG6 (5.47 g/L GAE). In addition, flavonoids are an important group of phenolic compounds in wine.^[45] The results of TF in Goji wine samples were consistent with that of TP and the highest content of TF was 0.36 g/L in AG6.

Table 2. Content of LBP, TF, TP, and free radicals scavenging capacities of Goji wine treated with and without oak matrix.

Samples	LBP (g/L) ^β	TF (g/L) ^γ	TP (g/L GAE) *	DPPH (TE) ^£	ABTS (TE) ^£
AG1	0.88 ± 0.02^a	0.27 ± 0.01^f	4.25 ± 0.05^g	1.27 ± 0.01^g	1.61 ± 0.02^f
AG2	0.90 ± 0.01^a	0.34 ± 0.01^bc	5.43 ± 0.01^b	1.78 ± 0.03^c	1.99 ± 0.03^b
AG3	0.91 ± 0.02^a	0.32 ± 0.03^c	4.43 ± 0.02^e	1.43 ± 0.02^d	1.89 ± 0.02^c
AG4	0.90 ± 0.01^a	0.30 ± 0.02^d	4.31 ± 0.03^f	1.40 ± 0.01^e	1.81 ± 0.04^d
AG5	0.89 ± 0.02^a	0.29 ± 0.01^ef	4.72 ± 0.02^d	1.37 ± 0.02^f	1.74 ± 0.01^e
AG6	0.91 ± 0.01^a	0.36 ± 0.01^a	5.47 ± 0.02^a	1.87 ± 0.02^a	2.13 ± 0.03^a
AG7	0.91 ± 0.02^a	0.35 ± 0.01^ab	5.30 ± 0.04^c	1.58 ± 0.01^b	1.97 ± 0.01^b

Note: Values are means of triplicate replicates ± SD, and different superscripted letters indicate significant differences (p < 0.05).

β: LBP expressed as glucose equivalents.

γ: TF expressed as rutin equivalents.

*: TP expressed as gallic-acid equivalents.

£: DPPH and ABTS expressed as g/L trolox equivalents.

In order to determine the antioxidant capacity of Goji wine samples, the DPPH and ABTS assays, the two most widely used methods, were also performed, as shown in Table 2. The capacities of free radicalscavenging of Goji wine determined by DPPH and ABTS trials were 1.27–1.87 g/L and 1.61–2.13 g/L, respectively, which were much higher than those of Chinese wine.^[44] Although the capacities of free radicalscavenging decreased with toast degree among oak chips samples, they had the highest content in oak shavings sample. No obvious differences were observed in oak barrel sample (p > 0.05), which were in accordance with the results of TP and TF. Therefore, the oak shavings sample had the highest antioxidant activity, related to the high contents of phenolics and flavonoids.^[46,47]

Conclusion

In the present study, results suggested that fresh Goji wine was an aroma lackluster wine and the oak matrix used were effective for the alteration of alcohols, esters and acids in Goji wine and the spread of oak-related volatiles from oak to wine. Moreover, the effect of oak matrix on the volatile profiles of Goji wine depended on the toast degree and particularly the type of oak matrix. The oak shavings sample was the most beneficial for the esters formation and the release of oak aroma compounds. With the high OAVs of compounds such as cis- and trans-whisky lactone, vanillin, eugenol, isoeugenol, and 4-vinylguaiacol, the oak shavings sample exhibited a pronounced flavour of woody, vanilla, and clove. According to sensory analysis, oak shavings samples had the highest score in olfactory and gustative. Additionally, it also revealed that Goji wine with oak shavings treatment possessed the highest antioxidant activity due to the high content of phenols and flavonoids. Thus, the medium toast oak shavings seems to be an ideal choose for the maturation of Goji wine.

Additional information

Funding

This work was financially supported by the Sichuan University – Luzhou City Cooperation Fund [grant number 2013CDLZ-S03].