Characterization of volatile aroma compounds in litchi (Heiye) wine and distilled spirit

Abstract This study used litchi (Heiye) wine and distilled spirit as raw experimental materials to analyze the volatile aroma compounds. Qualitative and quantitative determination of aromatic components was studied using stir bar sportive extraction (SBSE) and gas chromatography coupled to mass spectrometry (GC/MS). Results indicated that a total of 128 different types of aroma compounds were observed, which belonged to six chemical groups, including 39 esters, 16 alcohols, 16 acids, 22 terpenes, 17 aldehydes and ketones, and 18 other compounds. In particular, esters were the highest among all six categories and represented approximately 52% of the total flavor component content in litchi distilled spirit. The odor activity values (OAVs) revealed 22 types of aroma compounds with OAVs >1 in this test. It is possible that the produced litchi distilled spirit had a stronger varietal character due to the increased concentrations and OAVs of β‐damascenone, linalool, ethyl butyrate, ethyl isovalerate, ethyl caproate, trans‐rose oxide, and cis‐rose oxide. Taking the OAVs into account, we evaluated the characteristic aromas for litchi wine and litchi distilled spirit.


| INTRODUC TI ON
Litchi (Litchi chinensis Sonn.) is an important fruit crop originated in China and has earned its popularity worldwide due mainly to its charming aroma, unique taste, and possible health benefits. In recent years, litchi cultivation has been distributed primarily for countries in Southeast Asia, such as China, India, Thailand, and Vietnam which are the leading litchi-producing countries in the world, among which China is the largest producing country (Pareek, 2016). Several cultivars in the west of Guangdong region have a long history of cultivation and account for approximately one-third of all fruit exports in China, while others are relatively new (Emanuele et al., 2017;Xiong et al., 2018). Litchi is suitable for producing fruit wine, with its high sugar content (up to 19.2%) and rose-floral and citrus-like aroma (Chen et al., 2014). Various physiological functions have also been reported for litchi, such as cancer preventive, antioxidant activity, antimicrobial, anti-inflammatory activities, and so on (Emanuele et al., 2017;Kilari & Putta, 2016;Varzakas et al., 2016). These days, the fermented fruit alcohol industry shows increasing interest in functional food, which displays an additional function related to disease prevention or health promotion. The naturally occurring litchi is a prime candidate. A wide range of litchi cultivars and growing climates in addition to orchard management practices allow for countless variables in starting fruit material. Likewise, prefermentation treatments of fruit and juice, fermentation management, and postfermentation treatments of litchi wine will also have impacts on final product quality. Li et al. (2012) found a 3.2-fold difference in phenolic content between the highest and lowest litchi varieties, Heiye and Chanchutou, respectively. However, litchi pericarp browning and pulp decay lead to its short shelf life; it is liable to lose its attractive feature and rapidly goes unpleasant once harvested from trees (Zhao et al., 2020). Therefore, much attention has been paid to postharvest quality preservation via various approaches and further processing, including the production of litchi wine and distilled spirit, aiming to keep its excellent characteristics for a more extended period.
Today, the consumption of distilled spirit has become increasingly and wildly popular in the world. Globally, famous distilled spirits such as whiskey, brandy, and rum occupy most of the domestic alcohol market because of recognition as high-quality products (Lee et al., 2018). And there have been numerous studies regarding the volatile aroma compounds in fermented wine and distilled spirit. The purpose of this study was to explain the difference in aroma compounds between litchi (cv. Heiye) wine and distilled spirit by SBSE-GC/MS technology and to objectively evaluate the contribution of aroma compounds to litchi wine and distilled spirit through the calculation of OAVs. Based on these results, the study may improve our understanding of the essence of the aroma difference between litchi wine and litchi distilled spirit. It can provide a substantial basis for further research into the control of flavor during the distilled process for litchi distilled spirit.

| Litchi materials
Two varieties of litchi (cvs. Heiye an Guiwei) were from Maoming, Guangdong Province, China, 2013 vintage, at their optimum point of maturity and sound quality without the disease. Fresh litchi fruits were kept at −20°C until processing and analysis.

| Winemaking processes
The method of Lee et al. (2018) was used with some modifications. 70 kg of raw materials was divided, peeled, destemmed, and crushed, and then all placed in 20-L stainless steel tanks, and 60 mg/L sulfur dioxide was added for preservation. Physicochemical characteristics of litchi juice were as follows: total sugar, 125 g/L; titratable acidity (expressed as tartaric acid), 2.5 g/L. Afterward, the modified litchi juice was added with 20 mg/L pectinase and then kept at 5°C for 24 hr for clarification. As a starter, commercial yeast (Saccharomyces cerevisiae EC-1118, Chr.-Han, Horsholm, Denmark) 200 mg/L activated in advance (hydration in 5% of the glucose solution at 37°C for 30 min). The starter was added into the clear juice to start fermentation. And alcoholic fermentation was maintained at 25°C. Cap punching was performed three times a day during fermentation.
When the residual sugar was lower than 2 g/L, the wines were centrifuged at 1,500 × g for 20 min, and the supernatants were transferred to the clean jars, treated with 60 mg/L sulfur dioxide, and stored at 4°C for six months. During the storage, a cold treatment at −4°C for 15 days was used to stabilize, and three times general racking was performed to clarify.
The distilled spirit was gained by alcoholic distillation with some modifications (Paolo et al., 1997). 200 ml of litchi wine, at 12% alcohol (v/v), was distilled in a copper distiller (Hoga Co., Salvaterra de Mino, Spain). A fraction of 50 ml at 43.9% alcohol (v/v) was collected.
Four of these fractions were redistilled together, and a second 50 ml fraction at 83% alcohol was collected. The body of distillation was selected for testing; samples were recorded as distilled spirit.

Highlights
• The naturally occurring litchi (Heiye) has potential applications in wine products, particularly in distilled spirit.
• 128 different types of aroma compounds were identified in this study.
• Alcoholic fermentation and distilled spirit process lead to a series of by-products.
• It is possible that the produced litchi distilled spirit had a stronger varietal character.
Samples were collected at three different phases: raw juice, litchi wine, and litchi distilled spirit. Each sample was made in triplicate.

| Volatile compounds analysis
The volatile compounds of the samples were isolated by stir bar sportive extraction (SBSE) according to the modified method proposed by Aguirre et al. (2014), and the volatile compounds of the samples were analyzed by GC-MS as previously described by Lin et al. (2019).

| Extraction of volatile compounds
About 10 ml of each sample was placed in a 15-ml airtight vial containing a magnetic stirrer bar (SBSE, 20 mm × 0.5 mm, Gerstel Co. Ltd, Germany), including 50 µl of 2-octanol (0.234 mg/ml water, internal standard) and 2 g NaCl. Then, the airtight vial was placed on a magnetic stirrer (PC-420D, Corning Co. Ltd, USA) and extracted for 90 min at 40°C with stirring (1,100 rpm). After the extraction, the stir bar was taken out with tweezers, washed it until there is no residual sample with distilled water, and dried with filter paper. Finally, it was inserted into the GC injector for 3 min to desorb analytes.
The desorption temperature is 250°C. Each sample was extracted in triplicate.

| GC-MS analysis
The separation, detection, and quantification of volatile com- applied as previously reported with some modifications . Ultrapure helium was used as the carrier gas at 1.0 ml/ min. The initial column temperature was set at 40°C for 3 min.
Afterward, the temperature was raised to 160°C at a rate of 5°C/ min, then 7°C/min to 230°C, and held at 230°C for 8 min. Mass detector conditions were as follows: voltage of electron impact was 70 eV, scan range was m/z 30-350, and scanning frequency in full scan mode was 5.27 times/s.

| Determination of odor activity values
Odor activity value (OAV) is a parameter frequently used to assess each volatile compound's contribution to wine's aroma. Compounds existing in greater concentrations than their odor threshold (OAVs >1) are defined as aroma impact compounds and contribute individually to the wine aroma (Zhu et al., 2014). The influence of each aromatic compound on the scent of litchi wine and distilled spirit was determined by the odor activity values (OAVs), which were calculated as the ratio between the concentration and the odor threshold of the individual aroma compounds (Velázquez et al., 2015).

| Statistical analysis
The data were reported as means ± standard deviation of three trip-

| Esters
Esters are of primary industrial interest because they have very low thresholds and can directly affect wine flavor and via complex synergistic interactions (Dzialo et al., 2017;Lytra et al., 2016). In this study, esters were the largest group in terms of the composition and concentration of aroma compounds. The esters' concentration ranged  Table 2).
The most inadequate variety of esters was observed in litchi juice; there were only seven esters, but 15 and 28 kinds of esters in litchi wine and distilled spirit (Figure 1), respectively. All esters those detected in litchi juices showed the increasing trend in the fermented ones except diisopropyl adipate, diethyl phthalate, and dibutyl phthalate ( Table 2). In general, these compounds had a similar trend with a sharp increase during the winemaking and distilling process.
Since esters are one of the most important by-products of alcoholic fermentation and mainly produced by yeast as secondary products of sugar metabolism (Zhang et al., 2019 ethyl. Some of these esters have also been found to be dominant in strawberry wine (Kafkas et al., 2006) and orange wine (Selli, 2007   Note: Data are expressed as mean ± standard deviation (n = 3). a-c means with different lowercase letters in the same row indicate significant differences (p <.05); "-": Not detected.  (Feng et al., 2018;Guth, 1997;Lukić et al., 2016;Wu et al., 2011).

| Acids
Some acids in litchi distilled spirit are naturally present in litchi juice, while others are by-products of fermentation (Venkatachalam et al., 2018). This group of volatiles could actively contribute to wine flavor at low levels. The widest variety of acids was observed in litchi wine, and there were 11 kinds of acids, but 7 and 6 kinds of acids in litchi juice and litchi distilled spirit, respectively. The concentration of total acids ranged from 446.59 ( (Table 3), namely decanoic acid and octanoic acid. The OAVs of octanoic acid and decanoic acid increased sharply during the distilled process. They were 3.5 times higher in distilled spirit than in litchi wine, consistent with the trends of their concentrations. However, unpopular sweat and rancid aromas might not be generated in distilled spirit because the OAVs of octanoic acid (OAV of 6.7) and decanoic acid (OAV of 3.5) were much lower than its odor thresholds.

| Terpenes
Terpenes, synthesized from glucose by the isoprenoid pathway, provide a powerful floral and fruity aroma to berries (Venkatachalam et al., 2018;Wu et al., 2016). In the present study, 22 terpenes were detected in three treatments ( Table 2). The kinds of total aroma compounds did not differ significantly among the three treatments ( Figure 1). In contrast, the total contents of aroma compounds significantly differed among these treatments. Thirteen terpenes were found in litchi juice, namely cis-rose oxide (169.34 ± 18.42 µg/L),  Table 2). Except for 4-terpineol, cis-carveol, and elemicin, which were typically present in litchi juice but faded away in litchi wine F I G U R E 2 Comparison of the concentration (μg/L) of aroma compounds in litchi juice, wine, and distilled spirit. Values in the same column with different superscripted alphabet are significantly different at p < .05 and distilled spirit, these terpenes were sharply increased during the whole winemaking and distilling process. Geraniol, linalool, and cisrose oxide were higher than other terpenes in litchi juice; this trend was similar to the results reported by Chen and Liu (2016) in litchi juice. Geraniol could be metabolized into cis-rose oxide by some yeast strains (Steyer et al., 2012). Simultaneously, cis-rose oxide had a positive effect on the rose aroma of wine because of its high odor activity (Guth, 1997). Linalool is produced by many linalool, geraniol, citronellol, and carveol, were previously reported in litchi wine (Chen & Liu, 2016;Tang et al., 2019;Wu et al., 2011). A total of 7 and 6 terpenes were presented with OAVs higher than 1 in litchi juice and distilled spirit (Table 3)

| Aldehydes and ketones
Aldehydes and ketones could confer a more abundant, more elegant, and unique aroma to wine (Nyanga et al., 2013). The widest variety of aldehydes and ketones was observed in litchi juice. There were eight kinds of aldehydes and ketones, but 5 and 5 kinds of aldehydes ketones in litchi wine and litchi distilled spirit (Table 2 and  Therefore, apple and plum odor could be expected in the distilled spirit, as suggested by Genovese et al. (2007) and Lukić et al. (2016), who observed a similar trend in wines made from overripe grapes.

| Others
The variety and content of others in litchi wine were actually the lowest, while the total contents of others in litchi distilled spirit represented approximately 94%. Nine aroma compounds of others were identified in litchi juice. Among these aroma compounds, 2,6-methyl-2-octene, 2,4,6-triisopropylphenol, dureno, 1,4-benzenediol, dimethyl-, pentylcyclopropane, trimethylhydrazine, mesylazide were just found in litchi juice and fade away in litchi wine. In particular, 2-pentylfuran was the only one and presented in the whole winemaking and distilling process. 2-Pentylfuran was also determined in cv. Dimrit grape seed oil. These author also reported that presence of 2-pentylfuran may be associated with the applied high temperature during the oil extraction procedure (Sevindik et al., 2020). Six aroma compounds were newly generated during the distilling process. In the finished litchi distilled spirit,

| Analysis of the characteristic volatiles by principal component analysis (PCA) with OAVs (>1)
PCA was conducted to understand the correlation and segregation among those volatile compounds, which are significantly different among the three samples. Twenty-two essential volatile compounds were selected for PCA to determine the contribution with OAVs of >1 in Table 3. The richest variety of OAVs >1 was observed in litchi distilled spirit; there were 19 kinds of volatile aroma compounds, but 6 and 16 kinds of aroma compounds in litchi juice and litchi wine, respectively. As shown in Figure 3, the three samples were differentiated by their main aroma profiles. The compound cis-rose oxide (T1) was more related to litchi juice. Four kinds of volatile compounds may be related to litchi wine, including E5 (ethyl butyrate), T13 (pmenth-1-en-8-ol), T15 (D-citronellol), and A7 (1-octen-3-ol). And six kinds of volatile compounds were more related to litchi distilled spirit, including E12 (ethyl caproate), E6 (ethyl isovalerate), T2 (transrose oxide), T11 (linalool), and AK14 (β-damascenone).
The kinds of total volatile aroma compounds appeared only in distilled spirit were highest ( Figure 4 and Table 2), and there were 41 kinds of compounds in distilled spirit, but 25 and 23 in litchi juice and litchi wine, respectively. This illustrated fermentation and distilled spirit process lead to a series of by-products. They include esters, alcohols, acids, terpenes, and so on, all of which influence the final wine's quality. The concentration of the by-products can vary widely from a few g/L to hundreds of mg/L. Conversely, five kinds of volatile aroma compounds were disappeared in the alcohol fermentation process, and 33 disappeared in the distilled spirit process.
Fourteen kinds of volatile aroma compounds detected in litchi juice showed increasing trends during alcoholic fermentation and distilled spirit process. Notably, some volatile aroma compounds detected in distilled spirit were ten times higher than litchi wine, such as ethyl caprylate (11.02), ethyl caprate (32.57), ethyl palmitate (643.76), isoamyl alcohol (11.11), 7-methyl-3-methylene-6-octen-1-ol ( trans-2-hexenal, linalool, and geraniol had the highest OAVs in litchi (Heiye and Guiwei), much higher than its odor thresholds (Table S1) (Table 2 and Table 3) since these compounds are known to play a decisive fruit and flower aroma in the distilled spirit. Similar to previous results reported by Wu et al. (2011), who also observed higher OAVs of ethyl butyrate F I G U R E 4 Changes of the numbers of aroma compounds in winemaking process F I G U R E 3 Principal components analysis for the main volatiles (OAVs >1) in litchi juice, wine, and distilled spirit (OAV of 69.1 ± 4.0), cis-rose oxide (OAV of 62.7 ± 0.1), and transrose oxide (OAV of 20.6 ± 0.4) in litchi wine.

| CON CLUS ION
This study provides the first comprehensive characterization of the volatile aroma compounds contributing to the aroma profile of litchi distilled spirit (Heiye). One twenty-eight different aroma compounds were identified in this study, which belonged to 6 chemical groups, including 39 esters, 16 alcohols, 16 acids, 22 terpenes, 17 aldehydes and ketones, and 18 other compounds. According to the determined concentrations of total aroma components, they were ranked in the following order (highest to lowest concentration): esters; terpenes; acids; alcohols; others, and aldehydes and ketones.
The concentrations and kinds of total volatile aroma compounds were higher in distilled spirit than litchi wine. The richest variety of OAVs >1 was also observed in litchi distilled spirit; there were 19 kinds of volatile aroma compounds, but only 16 kinds of them in litchi wine. β-damascenone, linalool, ethyl butyrate, ethyl isovalerate, ethyl caproate, trans-rose oxide, and cis-rose oxide were the essential aroma-active compounds and played a decisive fruit and flower aroma in litchi distilled spirit. These findings provided comprehensive knowledge on the aroma character of litchi wine and distilled spirit.

| INFORMED CON S ENT
Written informed consent was obtained from all study participants.

ACK N OWLED G M ENTS
This study was supported by the National Key R&D Program of Ch ina (Nos. 2019YFD1002503) and the Natural Science Foundation of Guangdong province (Nos. 2011A01003). We thank the staff at the Fermentation Engineering Laboratory, College of Enology, Northwest A & F University, for their technical assistance.

CO N FLI C T S O F I NTE R E S T
The authors declare no competing financial interests.

E TH I C A L A PPROVA L
This study does not involve any human or animal testing.