Flavor profile variations of Huangjiu brewed in different traditional Chinese solar terms

Abstract

Background

Through long-term research on Huangjiu fermentation, it has been found that the quality of Huangjiu is closely associated with Chinese 24-solar term. Therefore, this study was conducted to explore the characteristic and property index differences of Huangjiu, such as aroma and flavor profile, and physicochemical properties—brewed in different solar terms by choosing five representative fermentation times of Shaoxing Huangjiu.

Results

Huangjiu samples in the current study all met the national standards of traditional semi-dry Huangjiu. There are significant differences in physicochemical properties such as acidity and amino nitrogen among different solar-term groups. Forty three detected volatiles were applied by principal component analysis and partial least squares discriminant analysis analyses to differentiate the main factors. Volatiles mainly loaded to four principal components, which accounted for 86.5%. Nineteen volatiles were discriminated to significantly differentiate solar-term groups. Odor activity values (OAV) analysis found 14 compounds with OAV>1, while correlation analysis between volatiles and the outcomes of sensory evaluation displayed the various properties of Huangjiu on aroma and flavor due to the various combinations of volatiles, reducing sugar, acidity and amino nitrogen. After a national wine inspector evaluated and scored Huangjiu fermented in different solar terms, Huangjiu brewed in Winter Solstice exhibited the highest performance, with a score of 91.0, and praised as a kind of gold medal product.

Conclusion

The methodology of this study can help to produce more types of appealing flavor and aroma of alcoholic beverages to consumers, build varietal Huangjiu or other alcoholic beverages through fermenting guidance by solar term, and even expand the applications of the traditional Chinese 24-solar term.

Huangjiu, Chinese traditional 24-solar term, volatile, gas chromatography–mass spectrometry, flavor profile

Introduction

Huangjiu is a traditional alcoholic beverage with alcohol content over 6% (in volume). Because of its special flavor, it has been adored by many consumers since ancient times. The fermentation materials of Huangjiu are grains (glutinous rice, japonica rice, millet rice, etc.) and mash fermentation including wheat Qu, yeast, and other saccharification starters (Jiao et al., 2017; Yu et al., 2019). Huangjiu—a traditional Chinese alcohol—has over 2500 years history (Sun et al., 2020), and with beer and wine, is regarded as the world’s three largest ancient alcoholic beverages (Wang et al., 2020). Shaoxing Huangjiu wins the highest reputation among varietal Huangjiu due to its favorable palate.

Twenty four solar terms originated in the Yellow River Basin of China by ancient Chinese farmers to guide agricultural affairs and farming activities (Shen, 2019). Because the 24 solar term can precisely reflect changes in climate, it plays an essential role in agricultural affairs, food processes, clothing, housing, and transportation (Han et al., 2020). The 24 solar term is even considered to be ‘The Chinese Fifth Great Invention’, and it is officially listed in the United Nations Educational, Scientific, and Cultural Organization (UNESCO) of Representative Works of Intangible Cultural Heritage of Humanity on November 30, 2016 (Shen, 2019). Oenological parameters like fermentation temperature and humidity would be strongly impacted by the selection of the 24 solar term. Therefore, oenological practices with various solar terms can strongly influence Huangjiu’s properties, especially flavors. The flavor profile of Huangjiu can directly influence consumers’ choices, and certain flavors draw purchasing intentions. Normally, Shaoxing Huangjiu are fermented from early winter to early spring, hence the fermentation is called ‘Winter Brewing’. By exploring the aroma profile variations of Huangjiu, Yu et al. (2021) found that Huangjiu brewed in Winter Solstice had a higher content of flavor volatiles and a better quality of aroma than that brewed at other times. Furthermore, Du et al. (2018) found that certain solar terms have better flavor of Baijiu.

In this study of flavor profile variations of Shaoxing Huangjiu, investigations proceeded to: (1) explore the physicochemical properties of Shaoxing Huangjiu brewed in different solar terms; (2) understand the correlation between the flavor and volatile compounds of Huangjiu; and (3) investigate the impacts of solar terms on Huangjiu properties. Then, this concept of the traditional Chinese 24 solar term transfers to properties of Huangjiu, even building different varieties of Huangjiu through fermentation in different solar terms.

To deeply understand the impacts of solar term selection on a molecular and a flavor basis, analytical techniques (headspace solid-phase microextraction/gas chromatography–mass spectrometry (HS-SPME/GC-MS)) and sensory evaluation were applied. The outcomes clarify the characteristics of varietal Shaoxing Huangjiu through analyses of physicochemical properties, volatile organic compounds (VOCs), odor activity values (OAVs) and principal component analysis (PCA). The method of current study can not only help to understand the role of the 24 solar term in Huangjiu fermentation and makes the flavor more attractive to consumers, but also guide the oenological strategy of alcoholic beverages also.

Materials and Methods

Sample collection

Huangjiu samples were provided by Zhejiang Pagoda Brand Shaoxing Rice Wine Co., Ltd. (Shaoxing, China). A total of 15 samples (five batches of Huangjiu, three duplicates for each) were brewed from the same raw materials, using the same ratios and brewing technique, while the brewing duration started at different solar terms. These samples were transferred to the laboratory. The fermentation process of Huangjiu is as follows. First, glutinous rice was selected, soaked, steamed, cooled, and mixed (wheat Qu, yeast, and the water of Jianhu). Second, after 4-d primary fermentation and 90-d post fermentation, the fermented mixture was filtered, clarified, sterilized, and then stored for 12 months in cellars.

Meteorological data study

Five solar terms were selected to ferment Huangjiu, including Beginning of Winter (T1), the intermediate moment between Lesser Snow and Greater Snow (T2), Winter Solstice (T3), the intermediate moment between Lesser Cold and Greater Cold (T4), and Beginning of Spring (T5). Different solar terms represent variation in temperature, hence the meteorological data were accessed to explore how solar terms influence the fermentation alcohols (Huangjiu in the current study). Two main websites were accessed: http://data.cma.cn/data/detail/dataCode/A.0012.0001.html and http://waptianqi.2345.com/wea_history/.

Reagents

Copper sulfate (≥98.0%), methine blue (99%), potassium sodium tartrate (99.99%), sodium hydroxide (≥97.0%), potassium ferrocyanide (98%), hydrochloric acid (37%), and formaldehyde solution (36%–38%) were analytical reagents obtained from Sinopharm Chemical Reagent Co., Ltd. (Hangzhou, China). 2-Octanol (purity greater than 99.8%) standard product was the gas chromatography (GC) reagents and was obtained from Sigma-Aldrich Co. (St Louis, MO, USA).

Quantitative analysis of physicochemical properties of Huangjiu

Reducing sugar, pH, acidity, amino nitrogen, ethanol content, and VOCs of Huangjiu brewed in different solar terms were determined and analyzed. Physicochemical properties of Huangjiu except ethanol content were measured by GB/T13662-2018 (State Administration for Market Regulation and Standardization Administration, 2018), while ethanol content was evaluated by densitometry at 20 °C after distillation. The specific procedures were as follows. The potassium ferrocyanide method was conducted to determine reducing sugar. The pH values were measured by a FiveEasy Plus FE20 pH Benchtop (Mettler Toledo, Columbus, OH, USA). Acidity and amino nitrogen were measured by titration that Huangjiu and water with 1:5 (in volume) was placed in a beaker. After the pH value was titrated to 8.2 with 0.1 mol/L NaOH, 10 mL of formaldehyde solution was added, and then titration was conducted with 0.1 mol/L NaOH until pH 9.2 was achieved. The volume of NaOH consumed was recorded and then converted to acidity and amino nitrogen.

Extraction and quantification of VOCs of Huangjiu

VOCs of Huangjiu were analyzed by solid-phase microextraction (SPME, Supelco, Bellefonte, PA, USA) combined with GC-MS (QP-2010, equipped with AOC-20I autosampler, Shimadzu, Japan). We followed the similar process of Chen et al. (2020). Briefly, the Huangjiu samples of 5 mL were put into 20 mL headspace class vials, and then 5 mL of deionized water and 20 μL of 2-octanol internal standard (4.120 µg/mL in absolute ethanol solution) were added. Then, the fiber (100 μm polydimethylsiloxane, Supelco, Bellefonte, PA, USA) was introduced to collect volatiles for 30 min after preheating at 50 °C for 5 min. Volatiles were subsequently desorbed at 250 °C for 15 min, and then were identified by GC-MS equipped with RTX-WAX elastic quartz capillary column (30 m×0.25 mm×0.25 µm; Restek Corporation, Bellefonte, PA, USA). The sample was injected in splitless mode. Helium was delivered as carrier gas at a flow rate of 1.0 mL/min. The initial temperature was 40 °C, held for 5 min; then it was increased to 120 °C at a rate of 6 °C/min, and held for 5 min; then the temperature was increased from 120 to 190 °C at a rate of 3 °C/min, and held for 5 min. The MS conditions were set as follows: EI source, electron energy 70 eV, full scan mode to collect data, ion source temperature 230 °C, interface temperature 250 °C, scanning range 35–500 m/z (mass-to-charge ratio), and solvent removal time of 3 min. The extracted volatile compounds were identified by comparing the spectra with mass spectrum libraries (NIST08 and NIST08s). The quantity of volatile compounds was determined using an internal standard method (Qin et al., 2013). Quantitative data of the flavor compounds were acquired using the following formula:

$${\displaystyle \begin{array}{l}{\displaystyle Ci(\mathrm{m}\mathrm{g}/\mathrm{L})=\frac{Ai}{Aj}\times C\mathrm{i}\mathrm{s}}\end{array}}$$

where Ci is the relative concentration of VOCs; Ai is the peak area of the flavor compounds, while Aj is the peak area of the internal standard; Cis is the internal concentration of standards.

Determination of odor activity values

The OAVs of volatile flavor compounds were calculated according to the following equation:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{O}\mathrm{A}\mathrm{V}i=\frac{Ci}{\mathrm{O}\mathrm{T}i}}\end{array}}$$

where OAVi is the odor active value of VOCs; Ci is the relative concentration of VOCs; and OTi is the minimum concentration that causes the sense of smell (Gemert, 2011).

Sensory evaluation

Sensory evaluation was conducted in a sensory laboratory, where a trained panel was recruited. The panel comprised 20 people (10 males and 10 females, aged 22–30 years), who had experience in sensory evaluation of Huangjiu. A total of six sensory attributes (sweetness, acidity, bitterness, pungency, astringency, and umami) were chosen to characterize the sensory properties of the Huangjiu samples (Chen et al., 2020). During training, the judges were instructed to read the definitions (Table S1; dos Santos Navarro et al., 2012; Jung et al., 2014). After 2 weeks of intermittent training (30–40 min per day), they precisely described the characteristics of Huangjiu taste. The samples of Huangjiu were randomized and the whole evaluation was a blind test. The 20 sensory assessors were randomly divided into five groups, each with four assessors. Twenty judges rated the test samples on a 9-point scale from 0 (very weak) to 9 (very strong) under standardized conditions. Water was provided to the judges for rinsing their palate between samples. All evaluations were conducted in a room with a uniform source of lightening, the absence of noise and distracting stimuli at 20 °C. Besides, to further understand the difference in taste and quality of Huangjiu brewed in different solar terms, we invited three professional National Sommelier Wine Tasters to evaluate the quality and taste, and the scoring criteria are shown in Table S2.

Statistical analysis

Each experiment was conducted at least three replications, and the results are presented as means±standard deviations (SD). One-way analysis of variance (ANOVA) followed by the Duncan test was employed to determine the significant differences using R (version 4.1.1; https://www.r-project.org/) with the ‘Agricola’ package by RStudio. P<0.05 was considered statistically significant. A heatmap was produced by R (version 4.1.1) with the package ‘pheatmap’. Principal component analysis (PCA) was conducted by Simca (version 14.1; Umetrics, Umea, Sweden) and R (version 4.1.1) with package ‘FactoMineR’, while partial least squares discriminant analysis (PLS-DA) was created with Simca 14.1. Simca 14.1 and R with package ‘ggplot 2’ were used for data visualization. The scale() function in R (version 4.1) was used to normalize data. This is also known as standardizing data, which simply converts each original value into a z-score. This function uses the following formula to calculate scaled values:

$${\displaystyle \begin{array}{l}{\displaystyle X_{\mathrm{s}\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{e}\mathrm{d}}=\frac{(X_{\mathrm{o}\mathrm{r}\mathrm{i}\mathrm{g}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{l}}-{\displaystyle \overline{X}})}{S}}\end{array}}$$

where X_original is the original x-value; $X$ is the sample mean; and S is the sample standard deviation.

Results and Discussion

Temperature variations of solar terms accessed through meteorological data

The fermentation process of Huangjiu is divided into two stages: primary and post fermentation, which include 4 d and 90 d, respectively. Primary fermentation has taken place indoors, whereas post fermentation has been processed outdoors. Temperature is an important factor that can affect the fermentation processes by influencing growth and metabolism of microorganism. A higher temperature during primary fermentation can increase the proliferation of mainly yeast (Saccharomyces cerevisiae) and other microbes. Hence, primary fermentation is conducted outdoors at high temperature. The exact date of brewing and cumulative temperature of post fermentation are shown in Table 1. The cumulative temperature of post fermentation of T5 (1285.5 °C) is the highest; T1 and T4 are 914.5 °C and 977 °C, respectively; T2 (880 °C) and T3 (846 °C) share a similar cumulative temperature.

Table 1.

The definite brewed date of the samples of Huangjiu

Sample	Solar term	Date	Primary fermentation	Post fermentation	Active cumulative temperatures of post fermentation (°C)
T1	Beginning of Winter	6 Nov	6–9 Nov	10 Nov–8 Feb	914.5
T2	The intermediate moment between Lesser Snow and Greater Snow	27 Nov	27–30 Nov	1 Dec–28 Feb	880.0
T3	Winter Solstice	20 Dec	20–23 Dec	24 Dec–23 Mar	846.0
T4	The intermediate moment between Lesser Cold and Greater Cold	10 Jan	10–13 Jan	14 Jan–13 Apr	977.0
T5	Beginning of Spring	2 Feb	2–5 Feb	6 Feb–6 May	1285.5

Active cumulative temperatures: The accumulation of daily average temperatures or daily average temperatures that meet specific requirements within a certain period of time.

Physicochemical properties of Huangjiu fermented in different solar terms

As Figure 1 shows, the physicochemical properties meet the national standard of traditional semi-dry Huangjiu (GB/T 13662-2018), whose criteria contain minimum ethanol (8%, in vol), reducing sugar (15.1–40.0 g/L), acidity range (3.0–7.0 g/L). There was no significant difference among different groups in ethanol contents, which was approximately 16%. There was similar reducing sugar in Huangjiu brewed in different solar terms of T3, T4, and T5. Huangjiu brewed in T1 had the highest reducing sugar, while T1 had the relatively lowest sugar. Wheat Qu and yeast can influence sugar contents by affecting polysaccharide hydrolysis (Khattak et al., 2013) and monosaccharide glycolysis. Important fermentation conditions like temperature could influence activity of wheat Qu and yeast to affect the reducing sugar of Huangjiu, and relative higher temperature is more suitable for brewing microorganisms to accumulate sugar in Huangjiu. Therefore, higher environmental temperature of T1 may contribute to higher reducing sugar. Acidity varied significantly in different groups. The acidity of Huangjiu brewed in T2 and T4 was lower than others, and T3 had relatively higher acidity, while T3 had the highest acidity and the acidity of T5 was parallel to T3. The amino nitrogen of Huangjiu in this study was in accord with the superior criterion (0.50 g/L). The amino nitrogen of Huangjiu is mainly produced by the amino acid metabolism from hydrolysis of rice protein and yeast autolysis (Martinez-Rodriguez and Polo, 2000). The contents and varieties of amino nitrogen can affect Huangjiu’s quality and its flavor profile (Caliari et al., 2014). Amino nitrogen of Huangjiu brewed in T1, T2, and T4 was similar, while T3 had higher and T5 had the highest content of amino nitrogen. The pH could influence the flavor of Huangjiu and its quality, and proper pH contributes to the flavor profile of Huangjiu. (Zhao et al., 2020). The pH of Huangjiu samples (3.65–4.13) are close to wine and sour (Esteban-Fernandez et al., 2018; Sanchez-Palomo et al., 2019). Huangjiu samples of the current study shared similar pH except T1, whose pH is lower (3.65±0.31).

Figure 1.

Analysis of physicochemical properties (ethanol, sugar, amino nitrogen, and acid) of Huangjiu brewed in different solar terms. Different letters indicated significance among multi-comparisons conducted by Duncan test. The data shown are mean±standard deviation (n=3).

Volatile organic compounds of Huangjiu fermented in different solar terms

Table S3 shows the quantitative data of VOCs in Huangjiu brewed in different solar terms, as well as the aroma thresholds and the corresponding OAVs of each compound. A total of 43 VOCs were detected by GC-MS, including 19 esters, six alcohols, six acids, six aldehydes, three ketones and three others, which are also shown in a heatmap (Figure 2). The highest contents are isoamylol (B2), phenylethanol (B4), ethyl caprylate (A8), ethyl lactate (A7), ethyl acetate (A1), and diethyl succinate (A11) with concentrations of 4773.14, 2705.20, 1363.48, 1250.77, 1161.70, and 863.04 µg/L, respectively.

Figure 2.

Heatmap of volatiles in Huangjiu brewed in different solar terms. The different lowercase letters (a, b, c) represent three repetitions of the same process.

Esters in the current study are more diverse than other kinds of volatiles, and Huangjiu brewed in T3 had the highest ester diversity, while Huangjiu brewed in T1 and T2 had the highest ester contents at a concentration of 6165.09 µg/L. The diversities and contents of ester varied in different groups due to fermentation in long-term low temperature, which was conducive to producing esters and alcohols by yeast (Deed et al., 2017). Ethyl esters comprised 72% of total esters, probably because there was a large amount of ethanol produced by fermentation, which were esterified to ethyl esters by alcohol acetyltransferase (Liu et al., 2019). Various ethyl esters contributed to the special aromas of Shaoxing Huangjiu: banana aroma and creamy—ethyl caprylate (A8); pineapple aroma—ethyl caprate (A9); creamy aroma—ethyl lactate (A7); pineapple aroma—ethyl acetate (A1) (Jung et al., 2014; Jin et al., 2021). T3 has the lowest ethyl acetate (A1) at (177.81±3.79) µg/L, whereas T5 has the lowest concentration at (316.43±19.62) µg/L in ethyl lactate (A7).

Higher alcohols constitute the main aroma skeleton in Huangjiu and the content of higher alcohols in Huangjiu is more than that in other wines (de-la-Fuente-Blanco et al., 2016). Higher alcohols in Huangjiu are mainly produced from amino acid deamination, where the amino acids are mainly from hydrolysis of rice protein and saccharometabolism of fermented starch (Liang et al., 2020). Normally, higher alcohols are formed under anaerobic condition, and amino acids are converted to higher alcohols via the Ehrlich pathway under adequate amino acid condition, while certain aldehydes can also be reduced to higher alcohols (Fan and Qian, 2006). The concentrations of higher alcohols of Huangjiu in the current study were higher than those of other types of compounds, particularly isoamylol (B2) and phenylethanol (B4), which reached (4773.14±253.49) µg/L and (2705.20±218.60) µg/L, respectively. Typical banana and mellow flavors are contributed by isoamylol (B2) (Hazelwood et al., 2008), while phenylethanol (B4) provides floral and rose aroma (Pires et al., 2014). These adorable aromas can be fused into a harmonious and delicate aroma and give people a pleasant feeling (Sharma et al., 2018).

Other volatiles contributing to flavor profiles in Huangjiu are acids and aldehydes, which are important volatiles during fermentation. However, a lack of acid, particularly acetic acid (C1), which is an indispensable feature of Huangjiu, will lead to unbalanced flavor. However, an excess of acid in Huangjiu would cause unpleasant flavors with a pungent smell (Yang et al., 2018). Acetic acid (C1) had the highest concentration, ranging from (126.92±16.81) to (532.62±63.42) µg/L, and acetic acid (C1) in T2 was higher than that in the other groups, with a concentration of (532.62±63.42) µg/L due to more esters generated in T2. Butyric acid (C3) had the highest content in T1 and T4, which could reach (61.91±4.06) µg/L and (66.6±6.89) µg/L, respectively, because their post fermentation of cumulative temperatures were similar. Undecanoic acid (C6) had the highest concentration in T1, with (241.46±121.02) µg/L; propionic acid (C2) was detected in T2 with a higher content of (31.78±1.26) µg/L. Butyric acid, undecanoic acid and propionic acid can bring unpleasant flavors and tastes to Huangjiu, while Huangjiu brewed in T1 had the highest content in T1 ((60.37±18.98) μg/L), which could bring the aroma of cheese and sweet cream. Aldehydes are another important type of volatiles in Huangjiu, and the appropriate content can expand the aroma of the wine body (Wang et al., 2016). Benzaldehyde (D4) had the highest content in T1 and T3, reaching (535.80±5.60) μg/L and (502.51±116.95) μg/L, respectively, and provides a nutty aroma in Huangjiu. Furfural (D2), which was higher in T3 and T4, could achieve (325.99±4.02) μg/L and (338.29±5.12) μg/L, respectively, and would give a bitter almond smell to Huangjiu. Besides, the reactions of aldehydes and higher alcohols can enrich the wine’s flavor (Sharma et al., 2018).

Ketones are not flavor substances, but a small amount of carbonyl compounds can provide medicinal and caramel aromas (Kang et al., 2016). 2-Nonanone (E1), 2-ctanone (E2), and 2-undecanone (E3) were detected in all samples and the total content of ketones was higher in T2 ((212.42±34.65) µg/L) and T3 ((290.10±34.98) µg/L). These three ketones could bring herbal, milky, and cheese aroma to Huangjiu, and had positive effects on the flavor of Huangjiu, especially the effect on the warmth and softness of the wine body.

Selecting important volatiles for Huangjiu fermented in different solar terms

Differential volatile compounds discriminated by PCA

As shown in the scree plot of Figure 3, principal components (PCs) contributed variously, and the decline reduced dramatically after PC4, as shown in Figure 3A. Contributions of variables were hardly explained by the remaining PCs. Hence, four major components were extracted from 43 volatiles, which together accounted for 86.5%. PC1 loaded isoamyl acetate (A3), ethyl lactate (A7), decanoic acid (C5), ethyl acetate (A1), acetoin (F3), 1-hexadecanol (B5), ethyl phenylacetate (A12), nonanoic acid (C4), 3-methylbutyl heptanoate (A16) and isoamylol (B2) as mainly positive, whereas substances as 2-phenyl-2-butenal (D6) and 2-hydroxy-ethyl ester (A19) were negatively loaded most; PC2 loaded phenylacetaldehyde (D5), decanal (D3), 2,3-butanediol (B3), ethyl laurate (A14), ethyl hexanoate (A4) and phenylethanol (B4) as mostly positive, whereas volatiles such as ethyl benzoate (A10), ethyl caprate (A9) and benzaldehyde (D4) were negatively loaded most (Table S4, Figure 3B). Similarly, PC3 loaded furfural (D2), hydroxyacetone (B6), 2-undecanone (E3) mainly positively, while isopropyl myristate (A17) and ethyl heptanoate (A6) were negatively loaded most; PC4 loaded 2-nonanone (E1) and maltol (F1) as the most positive variables, while 2,4-di-tert-butylphenol (F2) was negatively loaded most (Table S4, Figures 3C and 3D). Squared cosines are used to refine the factor loadings of volatiles and solar terms. Thus, cosine squares were used to confirm the factor loadings of variables and samples in current study (Figures S1A–S1B). The solar term T1 could be explained by PC1, while PC2 could explain T2. Squared cosine values displayed that only T3 blend (samples as T3a, T3b, T3c) explained relatively evenly all PC1, PC2, PC3 and PC4 relatively evenly. PC4 mainly loaded T4, while T5 blend explained both PC1 and PC3. Similarly, T2, T5, and T1 were clearly separated in Figure 3B (PC1 vs. PC2) and Figure 3C (PC1 vs. PC3), whereas T4 and T3 were only clearly distinguished in Figure 3D (PC1 vs. PC4).

Figure 3.

Principal component analysis (PCA) biplot of loading plot and score plot. (A) Percentage of explained variance of each principal component (PC); (B, C, and D) Comparisons of PC1 vs. PC2, PC1 vs. PC3, and PC1 vs. PC4, respectively.

Variable importance in the projection of volatiles discriminated by PLS-DA

PLS-DA can be used for discriminative variable selection. Therefore, PLS-DA was used to discriminate volatiles of Huangjiu brewed in different solar terms. The variable importance in the projection (VIP) of VOCs was used to qualify the contribution. The larger VIP indicates greater importance for the classification, generally, variables with VIP>1 are significantly different among categories. There were 19 volatiles with VIP>1 and P-value<0.05 (Table 2). It can be observed that they mostly contributed to distinguishing different solar terms of Huangjiu.

Table 2.

P-values of compounds with the variable importance in the projection (VIP)>1 in the partial least squares discriminant analysis (PLS-DA) model

No.	Compounds	VIP	P value	No.	Compounds	VIP	P value
A2	Ethyl butanoate	1.020	0.011	C2	Propionic acid	1.211	0.008
A4	Ethyl hexanoate	1.050	0.008	C3	Butyric acid	1.053	0.037
A5	Isoamyl N-butyrate	1.142	0.011	D1	Nonanal	1.078	0.013
A6	Ethyl heptanoate	1.138	0.029	D2	Furfural	1.206	0.014
A11	Diethyl succinate	1.019	0.025	D3	Decanal	1.064	0.009
A17	Isopropyl myristate	1.158	0.012	E1	2-Nonanone	1.188	0.024
A18	Apricolin	1.137	0.008	E3	2-Undecanone	1.278	0.012
B1	Isobutanol	1.007	0.012	F1	Maltol	1.324	0.028
B3	2,3-Butanediol	1.077	0.014	F2	2,4-Di-tert-butylphenol	1.282	0.021
C1	Acetic acid	1.151	0.016	–	–	–	–

Aroma and flavor profiles of Huangjiu fermented in different solar terms

The OAV discrimination of Huangjiu in different solar terms

There was no direct correlation between aroma and the VOCs of Huangjiu in different solar terms. It is necessary to discriminate VOCs contributing to the aroma of Huangjiu by aroma threshold analysis. Aroma threshold is the minimum concentration that causes the smell. Only concentrations higher than this threshold can be detected by human nose (Capone et al., 2013). Therefore, aroma threshold can quantify the aroma as well as judge the aroma of the wine. OAVs can describe the contribution of VOCs in the aroma system of Huangjiu (Han et al., 2020). When OAVs>1 of the compounds, they were considered as the key volatile components (Wang et al., 2020). As Table S3 displays, there were 14 volatiles with OAVs>1 that mainly contributed to the flavors of Huangjiu, and most of them bringing floral and fruity aromas. The OAVs of phenylethanol (B4), ethyl caprylate (A8), phenylacetaldehyde (D5), ethyl caprate (A9), ethyl hexanoate (A4), and isoamylol (B2) were over 10, all of which contributed to Huangjiu’s stronger aroma profile, indicating that they played significant roles in the aroma of Huangjiu.

Most esters can affect aromas of Huangjiu, due to its lower aroma thresholds and relatively large OAVs. Even though the ester content was low, they could strongly affect Huangjiu’s aroma. The OAVs of certain esters (ethyl acetate (A1), ethyl phenylacetate (A12) and phenylethyl acetate (A13)) are lower than many compounds (0.035<OAV<1.494), but they contribute to the fruity aroma of wines (Zhang et al., 2021). Higher alcohols can enrich the flavor profile of Huangjiu and promote the coordination of the wine. Phenylethanol (B4) had the highest OAV among all substances, particularly Huangjiu brewed in T2. Due to its strong sweet and rosy aroma, it has become the most fragrant substance that contributes to the aroma of Huangjiu (Chen et al., 2021). The second highest OAV is ethyl caprylate (A8), which has a sweet banana, pineapple, and creamy aroma. The highest OAV of ethyl caprylate was from T1 (852.177), while the lowest was from T4 (631.861). Ethyl caprylate is believed to be the most important flavor substance in wine (Yu et al., 2019), and plays a vital role in the aroma profile of Huangjiu. Ethyl caprylate, isoamylol (B2), ethyl butanoate (A2), ethyl caprate (A9), and isoamyl acetate (A3) increased the fruit aroma in T1. The different special aroma profiles in various solar terms are due to different content of phenylethanol (B4), ethyl hexanoate (A4) and nonanal (D1). These combined VOCs increased the floral and fruity aroma in T2, the pineapple aroma in T3, the coconut, honey aroma in T4 and the cocoa aroma in T5. Seven VOCs with OAVs>1 were discriminated by VIP compounds from PLS-DA. These VOCs included pleasant odors such as ethyl butanoate (A2) with pineapple odor, ethyl hexanoate (A4) with fruity odor, apricolin (A18) with coconut odor, nonanal (D1) with rose and fresh orange peel odor, acetoin (F2) with fatty frankincense odor; while unpleasant odors were produced by VOCs such as decanal (D3) with a sour smell and 2-undecanone (E3) with a bitter odor.

Flavor profiles differentiation of Huangjiu fermented in different solar terms

Organoleptic assessments were separated into six descriptors as sweetness, acidity, bitterness, pungency, astringency, and umami to reduce the possibilities of objective bias of sensory evaluation (ANOVA was used to compare differences among groups). There were significant differences of flavors except acidity among Huangjiu brewed in different solar terms (Figure 4A). The umami flavor among different groups was similar, and the flavor of Huangjiu brewed in T3 and T2 was relatively higher than that in the other groups. The sweetness of Huangjiu brewed in T1 was significantly stronger than that in the others, while T4 had the mildest sweet flavor. Unpleasant flavors like bitterness, pungency, and astringency of Huangjiu brewed in T2 are relatively lower than others especially for flavors of pungency and astringency, whereas those unpleasant flavors of T3 are relatively higher than other groups, especially for bitterness and pungency. The astringency score was highest in T5, followed by T4 and T3. To further uncover the relationship between sensory profiles and other factors, the correlation analysis of sensory score, physicochemical properties and volatile components was conducted (Qin et al., 2013; Figure 4B). The outcomes suggested that a strong correlation of sweet flavor was with factors such as reducing sugar, ethyl acetate (A2), isoamylol (B2), nonanoic acid (C4), decanoic acid (C5), undecanoic acid (C6), ethyl 3-hydroxybutyrate (A15), apricolin (A18), and 2-hydroxy-ethyl ester (A19). Acidity was found to be strongly correlated with the contents of acetic acid (C1) and propionic acid (C2) at concentrations of 32.62 and 31.78 µg/L, respectively. Similarly, 2-nonanone (E1), 2-undecanone (E3), maltol (F1) and 2,4-di-tert-butylphenol (F2) were significantly correlated with bitterness. The pungency was mainly related to total acid, ethyl hexanoate (A4), phenylethyl alcohol (B4), 1-hexadecanol (B5), propionic acid (C2), nonanal (D1), maltol (F1) and 2,4-di-tert-butylphenol (F2). Astringency was primarily related to isoamyl N-butyrate (A5), ethyl lactate (A7), benzoic acid, 2-hydroxy-ethyl ester (A19), and 2-phenyl-2-butenal (D6). Umami was significantly correlated with 2-undecanone (E3), maltol (F1), and 2,4-di-tert-butylphenol (F2). In conclusion, Huangjiu brewed in different solar terms presented different taste profiles, and the variation of Huangjiu taste was strongly correlated with physicochemical and the combinations of VOCs. The 15 VOCs that were strongly associated with flavor were discriminated by VIP compounds from PLS-DA (Table S5), whereas 13 VOCs with OAVs>1 were significantly associated with flavor (Table S6).

Figure 4.

(A) Sensory evaluation of Huangjiu brewed in different solar terms, * indicates significant differences in sensory scores between samples (*P<0.05; **P<0.01). (B) Pearson correlation network plot between sensory scores and physicochemical results and volatiles, showing only the results with significant correlations of |r|>0.5. The red, black, thin, and thick lines represent positive, negative, significant and very significant correlations, respectively.

The panelists of National Sommelier Wine Tasters evaluated Huangjiu samples (Table 3) and found that Huangjiu brewed in Winter Solstice grains had the highest scores, which indicated its pleasant aromas; mellow, smooth and harmonious flavors; and a long aftertaste. The relatively lower cumulative temperature may have reduced the rapidity of microorganism proliferation and metabolism in Huangjiu, and moderate growth and metabolic speed can optimize the profile of aroma and flavor, which benefits the whole taste of Huangjiu (Mao, 2013; Jin et al., 2021).

Table 3.

The evaluation of Huangjiu by three experts and their average scores

Sample	Sensory assessment	Score
		Color (10)	Smell (25)	Taste (50)	Style (15)	General comment (100)
T1	The aroma is relatively pure, the wine is fresh and smooth, the taste is a little weak	9.3	19.3	40.0	12.7	81.3
T2	The aroma is pure, with mellow aroma, but not strong, the wine is fresh and smooth, the taste is slightly weak	9.0	20.7	41.3	12.7	83.7
T3	Mellow and rich, gentle and full-bodied, with a harmonious taste and a lasting aftertaste, with the typical style of Shaoxing Huangjiu	9.3	24.0	43.7	14.0	91.0
T4	Full-bodied, distinctly mellow, flavorful and smooth, with a harmonious component	9.3	21.7	41.0	13.7	85.7
T5	The aroma is richer and fuller-bodied, with some pungency and with a slightly bitter aftertaste	9.0	19.0	39.0	12.7	79.7

Conclusions

The fermentation of Huangjiu is affected by factors such as microbes, temperature, etc. Ancient Chinese people used traditional Chinese 24 solar terms to separate periods by temperature. The proliferation and metabolism of yeast and other microorganism, which contribute to the development of flavor and aroma of Huangjiu, are mainly affected by temperature. Hence, five representative solar terms were selected to investigate the influence of different solar terms on characteristic and property indexes of Huangjiu. The results showed significant differences in acidity and amino nitrogen. Forty three VOCs were detected by HSPM–GC-MS analysis, and they were composed of esters, alcohols, acids, aldehydes, and ketones. PCA and PLS-DA analysis of the volatiles found that Huangjiu brewed in different solar terms could be distinguished significantly, and 19 essential compounds were identified that could be used to distinguish Huangjiu brewed in different solar terms. There were 14 compounds with OAVs>1, while phenylethyl alcohol, ethyl caprylate, phenylacetaldehyde, ethyl hexanoate, ethyl butanoate, and nonanal had OAVs>10. The sensory evaluation further revealed significant differences in tastes of different Huangjiu samples, besides the significant correlation between tastes and volatiles found. Three National Sommelier Wine Tasters evaluated Huangjiu samples, and Huangjiu brewed in Winter Solstice gained the highest scores. This study extends the application of the traditional Chinese 24 solar terms and even allows the fermentation of different types of Huangjiu by selected solar terms. However, in this study, the fermentation of Huangjiu was according to traditional procedures, which was not as precise as in controlled laboratory conditions. During the open fermentation process, factors like humidity and temperature difference between day and night were not considered in this study. Further study will precisely control solar term-imitating factors under laboratory conditions to explore the Huangjiu profile of aroma and flavor.

Author Contributions

Lina Lu: Conceptualization, methodology, software, formal analysis, visualization, writing original draft, review and editing. Jiaojiao Zhang: Formal analysis, software, data curation, visualization, investigation, writing, review and editing. Fenghua Wu: Resources, supervision, project administration. Guangfa Xie: Resources, supervision. Zhichu Shan: Resources, project administration. Xingquan Liu: Writing original draft, review and editing, project administration, funding acquisition.

Acknowledgements

We thank Zhejiang Pagoda Brand Shaoxing Rice Wine Co., Ltd. for their generous supply of materials for the present study, and technical support provided by the panelists of three National Sommelier Wine Tasters.

Funding

This research was supported by the Science Technology Department of Zhejiang Province Project (L20200079), China.

Conflict of Interest

The authors declare no conflict of interest.

Author notes

Lina Lu and Jiaojiao Zhang are co-first authors.