Black tea, coffee, fresh and mature citrus, white granulated sugar: commercially available; Saccharomyces cerevisiae: Angel yeast Co., Ltd.; Pectinase (enzyme activity 10⁵ U/mL): Shanghai jietu industry and Trade Co., Ltd.; Potassium metabisulfite (food grade), vitamin C (purity ≥ 99.0%): Guilin baozuan Food Co., Ltd.; Anhydrous ethanol (food grade): Shandong Luying Chemical Co., Ltd.; Other reagents are domestic analytical pure.

The black tea (coffee) was crushed after removing impurities, sieved through 40 mesh sieve, added with 15 times the weight of ethanol solution of tea (coffee) powder (volume fraction was 60%), soaked for 48 h, filtered, the filtrate was concentrated to 1/3 of the original volume under reduced pressure, and then spray dried at 160 °C to obtain instant powder. Fresh and mature oranges were selected, peeled and seeded, and then crushed with the same quality of purified water. Pectinase (20 U/g) was added to the pulp juice and treated at room temperature for 3 h, and then filtered. According to the experimental design, the instant powder was added into the juice, and then the initial sugar content was adjusted by white granulated sugar, and then the total acid content was adjusted to 0.9% (w/w) by adding potassium metabisulfite, and 0.08% (w/w) vitamin C was added, and then pasteurized (65 °C, 30 min); After cooling to room temperature, 0.05% (v/v) expanded and cultured Saccharomyces cerevisiae suspension (1.3 × 10⁹ CFU/mL) was inoculated, followed by static fermentation for 5.5 days under the experimentally set conditions; After the completion of fermentation, the finished product is obtained after four layers of gauze filtration and pasteurization (65 °C, 30 min).

The previous study found that when the addition amount of instant coffee, instant black tea, initial sugar content and fermentation temperature were 1.5‰ (w/w), 1.5% (w/w), 22% (w/w) and 24 °C respectively, the fermented tea wine with excellent sensory quality could be prepared. Therefore, the effects of the amount of instant coffee (0.5‰, 1.0‰, 1.5‰, 2.0‰, 2.5‰), the amount of instant black tea (0.5%, 1.0%, 1.5%, 2.0%, 2.5%), the initial sugar content (18%, 20%, 22%, 24%, 26%) and fermentation temperature (20 °C, 22 °C, 24 °C, 26 °C, 28 °C) on the sensory quality of BTCCFW were investigated, and each experiment was repeated three times.

On the basis of single factor experiment, the sensory score (Y) of BTCCFW was taken as the index, and the addition amount of instant coffee (A), instant black tea (B), initial sugar content (C) and fermentation temperature (D) were taken as the experimental factors to design a 4-factor 3-level response surface test to optimize the brewing process of BTCCFW (Grygorcewicz et al., 2023; Humberg and Grund, 2022). See Table 1 for response surface test factors and levels.

Referring to the sensory evaluation method of gb/t 15038–2006 “general methods for analysis of wine and fruit wine”, 11 professionals (6 males and 5 females) aged between 28 and 38 with professional qualification certificate of senior wine taster were invited to form an evaluation team to evaluate the sensory quality of BTCCFW from four aspects: appearance and color, aroma, taste and typicality. Take the average score of 11 people, the full score is 100 points (de Santis, 2024). See Supplementary Table S2 for sensory scoring criteria of BTCCFW.

A 5.0 mL aliquot of BTCCFW sample was mixed with 10 μL of internal standard (100 μg/g 2-octanol) in a headspace vial and equilibrated at 60 °C with 250 rpm agitation for 15 min. Then, 0.5 mL of headspace gas was sampled using a preheated (80 °C) 2.5 mL airtight syringe. VFCs were extracted with a DVB/CAR/PDMS (Divinylbenzene/Carboxen™/Polydimethylsiloxane) fiber for 45 min at 60 °C, then desorbed in the GC-MS injection port at 250 °C for 3 min (Yang et al., 2022). Each sample was analyzed in triplicate. VFCs were identified by comparing mass spectra with NIST 1.6 and Wiley 6.0 databases, and relative concentrations were calculated using the internal standard method (Rahayu et al., 2017). Additionally, the Equation 1 was used to calculate the odor activity values (OAV) of the VFCs (Bragagnolo et al., 2018). O A V i = c i T i (1) where c _i represents the relative VFCs concentration and T _i denotes the VFCs odor threshold in water.

SPSS statistics 20.0 was used for data statistical analysis, originpro 12.1 was used to draw the single factor test result chart, and design expert 12.0 was used to analyze the response surface test design results.

It can be seen from Figure 1A that when the amount of instant coffee is between 0.5‰ and 2.5‰, the sensory score of BTCCFW increases first and then decreases with the increase of the amount of instant coffee. When the amount of instant coffee is 2.0‰, the sensory score reaches the highest. When the addition amount of instant coffee is 0.5‰–2.0‰, the scorched aroma substances in instant coffee and tea polyphenols form a composite aroma, which enhances the layering of the wine. Coffee polysaccharides and wine esters form a stable colloidal structure, which improves the smoothness of the taste (Ballesteros et al., 2015; Ballesteros et al., 2017). When the addition amount of instant coffee is greater than 2.0‰, excessive instant coffee will cause “burnt bitterness” and sour feeling. Polyphenols in coffee and proteins in compound fruit wine are easy to form insoluble complexes, resulting in wine turbidity. The diffusion of aroma substances can also be inhibited by excessive coffee polysaccharides (Reichembach et al., 2024). Therefore, the addition amount of instant coffee was selected as 1.5‰∼2.5‰ for subsequent response surface test.

As shown in Figure 1B, the sensory score of BTCCFW increased first and then decreased with the increase of the amount of instant black tea. The tannin provided by an appropriate amount of instant black tea can support the metabolic activity of yeast. At the same time, the concentration of tea polyphenols within this range can not only inhibit the growth of miscellaneous bacteria, but also will not interfere with the sugar metabolism of S. cerevisiae, which optimizes the adaptability of fermentation to a certain extent. Tea polyphenols can also neutralize the bitter taste of coffee, and improve the freshness of the wine through convergence. Theaflavins form a compound aroma with the burnt aroma of coffee and the esters of the wine, enhancing the flavor level (Meilhoc and Teissie, 2020; Parapouli et al., 2020). When the addition amount of instant black tea was 1.5%, the characteristic aroma of black tea, coffee aroma and wine aroma reached the best ratio. When the addition amount of instant black tea exceeds 2.0%, the activity of S. cerevisiae is inhibited, the proportion of fermentation products is unbalanced, and pungent odor is generated. At the same time, excessive tea polyphenols will form insoluble complex with caffeine, resulting in wine turbidity and affecting sensory score (Kim et al., 2021). Therefore, the addition amount of instant black tea was selected as 1.0%∼2.0% for the follow-up response surface test.

It can be seen from Figure 1C that the sensory score of BTCCFW increased first and then decreased with the increase of initial sugar content. When the initial sugar content increased from 18% to 24%, the sensory score increased significantly; When the initial sugar content exceeded 24%, the score gradually decreased. From the perspective of fermentation mechanism, sugar is the main carbon source and energy source for the growth and metabolism of S. cerevisiae. If the sugar content is insufficient, it will limit the proliferation and metabolic activity of S. cerevisiae, resulting in incomplete fermentation, which will affect the flavor and taste of the wine; However, under the condition of sufficient sugar, yeast will rapidly proliferate and accelerate the consumption of nutrients in the fermentation system, causing metabolic imbalance (Korovesi et al., 2022; Schulze et al., 2024). The experimental data showed that when the initial sugar content increased from 18% to 24%, the sensory score of BTCCFW increased by 19.2% compared with the initial value; When the initial sugar content increased to 26%, the high osmotic pressure produced by high sugar environment would inhibit the activity of yeast, reduce the fermentation rate, and promote the metabolism of yeast to produce some bad flavor substances; In addition, high sugar concentration will cover up the diversified flavor characteristics of black tea and coffee to a certain extent, resulting in the simplification of liquor flavor, which is easy to cause taste fatigue and discomfort in sensory evaluation, and ultimately lead to a decline in score (Amberg and Burke, 2016; Schmidt et al., 2024). Therefore, the initial sugar content of 22%∼26% was selected for the subsequent response surface test.

As shown in Figure 1D, when the fermentation temperature changes between 20 °C and 28 °C, the sensory score of BTCCFW shows a trend of first increasing and then decreasing with the increase of fermentation temperature. When the fermentation temperature is 24 °C, the sensory score reaches the highest. When the fermentation temperature was lower than 24 °C, the metabolic activity of S. cerevisiae was limited, the synthesis of esters, higher alcohols and other flavor substances decreased, the fruit aroma characteristics of the wine were not prominent, and the overall taste was not full; In particular, when the temperature is as low as 20 °C, the yeast proliferation rate decreases significantly, which not only prolongs the fermentation cycle, but also makes the wine less flavored and thin due to the lack of metabolites (Tous Mohedano et al., 2023; Wu et al., 2023). When the fermentation temperature was 24 °C, the inhibitory effect of tea polyphenols and caffeine on yeast activity was the least, the utilization rate of sugar was high, the fermentation was stable, the residual sugar was low, the accumulation of aroma substances was sufficient and the persistence was significantly enhanced. When the fermentation temperature is higher than 24 °C, the metabolic rate of S. cerevisiae is accelerated, and the BTCCFW begins to have a bitter taste. At the same time, the high temperature accelerates the escape of volatile aroma substances, which gradually reduces its sensory score (Maneira et al., 2025). Therefore, the fermentation temperature of 22∼26 °C was selected for the follow-up response surface test.

According to the single factor test results and product characteristics, the box Behnken central composite design was used to carry out the response surface optimization test. The results are shown in Table 1. Taking sensory score (Y) as the response value, after regression fitting, the corresponding regression equation is obtained as follows. Y = 94.5 + 0.4 A + 0.16 B − 0.33 C + 1.69 D + 1.03 A B + 0.1 A C + 2.43 A D − 0.37 B C + 0.68 B D − 2.33 C D − 3.86 A 2 − 2.93 B 2 − 3.44 C 2 − 3.3 D 2

The results of variance analysis of the test data are shown in Supplementary Table S1. According to the analysis of variance, the influence order of the four factors on the sensory score of BTCCFW was D > A > C > B, that is, the fermentation temperature had the most significant effect on the sensory score of BTCCFW, followed by the amount of instant coffee, and then the initial sugar content. As the core factor regulating yeast metabolic activity, fermentation temperature directly affects the growth, reproduction and enzyme synthesis of S. cerevisiae, appropriate temperature promotes yeast to efficiently decompose sugars into ethanol and intermediate metabolites (e.g., esters, aldehydes), while extreme temperatures inhibit metabolic balance, leading to insufficient flavor substance generation or accumulation of off-flavors. Moreover, temperature significantly influences the volatility of aroma components (e.g., terpenoids, phenols), determining the intensity and coordination of tea, coffee and fruit aromas in the final product, thus becoming the most critical factor affecting sensory quality. The quadratic terms A ² , B ² , C ² and D ² had significant effects on the sensory score of BTCCFW (P < 0.01); One time item D, interactive items AD and CD had significant effects on the sensory score of BTCCFW (P < 0.05). It can be seen from Figure 2 that the slope of the response surface formed by factor A and factor D is the steepest, which verifies that the interactive term ad has a significant effect on the sensory score of BTCCFW (P < 0.05).

According to the regression equation of sensory score of BTCCFW, the highest sensory score of BTCCFW can be obtained under the addition amount of 2.1‰ of instant coffee, 1.6% of instant black tea, 23.6% of initial sugar content and 24.8 °C of fermentation temperature (test 30). Compared with traditional fruit wines (typically fermented at 18 °C–22 °C) and tea wines (usually with 10%–15% initial sugar content), the optimal fermentation temperature of 24.8 °C in this study is moderately higher, which facilitates the coordinated release of tea polyphenols and coffee volatile compounds while maintaining yeast activity; the 23.6% initial sugar content balances alcohol production and flavor formation, avoiding excessive sweetness or bitterness. Moreover, the combined addition of instant coffee and black tea is rarely reported in similar compound fermented wines, highlighting the novelty of this brewing process. Therefore, three groups of validation tests were carried out on the above conditions, and the sensory score of BTCCFW in test group 30 was (95.6 ± 1.3). Therefore, the optimal brewing conditions of BTCCFW were determined as follows: the addition amount of instant coffee was 2.1‰, the addition amount of instant black tea was 1.6%, the initial sugar content was 23.6% and the fermentation temperature was 24.8 °C.

Through mass spectrometry analysis, a total of 454 aroma compounds (7 more than those identified before the process optimization) were detected in the BTCCFW after process optimization, which were classified into 13 categories, namely esters (106 aroma compounds), terpenoids (84 aroma compounds), heterocyclic compounds (64 aroma compounds), aldehydes (43 aroma compounds), ketones (41 aroma compounds), alcohols (37 aroma compounds), aromatics (21 aroma compounds), acids (16 aroma compounds), phenols (14 aroma compounds), hydrocarbons (11 aroma compounds), ethers (8 aroma compounds), sulfur compounds (6 aroma compounds) and nitrogen compounds (3 aroma compounds), with the specific results shown in Figure 3.

After the process optimization, there were 63 aroma compounds with significantly increased relative contents compared with those before optimization, including 16 terpenoids (e.g., Linalool, Cyclohexene, 1-methyl-4-(1-methylethylidene)-), 11 heterocyclic compounds (e.g., Pyrazine, 2-methoxy-3-(1-methylethyl)-, 2-Thiophenemethanethiol, Ethanone, 1-(2-thienyl)-), 8 esters (e.g., Benzoic acid, methyl ester, Methyl salicylate), 7 aldehydes (e.g., Nonanal, 2-Hexenal), 4 alcohols (e.g., 1-Octen-3-ol, 1-Butanol, 3-methyl-), 4 ketones (e.g., 3,5-Octadien-2-one, (E,E)-), 4 sulfur compounds (e.g., Benzenemethanethiol, Methyl ethyl disulfide), 3 phenols (e.g., 2-methoxy-Phenol, Phenol, 2-methyl-, Ethyl maltol), 3 acids, 2 aromatics, and 1 hydrocarbon. In contrast, 37 aroma compounds showed significantly decreased relative contents, which were composed of 12 heterocyclic compounds (e.g., 2-Ethoxy-3-methylpyrazine, 2-Methyl-1,3-dithiacyclopentane), 10 esters (e.g., Benzene, (2-isothiocyanatoethyl)-), 5 aromatics (e.g., Benzene, 1,2,4,5-tetramethyl-, Benzene, 1,3-dimethyl-), 3 aldehydes (e.g., Undecanal), 2 ketones (e.g., Ethanone, 1-(4-methylphenyl)-), 2 phenols (e.g., 2,4-Di-tert-butylphenol, Phenol, 3,5-dimethyl-), 1 alcohol (Benzyl alcohol), 1 ether (1,3-Benzodioxole, 4-methoxy-6-(2-propenyl)-) and 1 hydrocarbon (Dodecane).

Volatile aroma compounds with significantly altered odor activity values (OAV) > 1 in BTCCFW before and after technological optimization are presented in Table 2. Beneficial aroma compounds with coffee (Cai et al., 2022) (Benzenemethanethiol, 2-Thiophenemethanethiol), nutty (Whetstine et al., 2006) (Ethanone, 1-(2-thienyl)-, Pyrazine, 2-methoxy-3-(1-methylethyl)-, 2-methoxy-Phenol), and fruity (Lu D. et al., 2022) (2-Hexenal, 1-Octen-3-ol, 3,5-Octadien-2-one, (E,E)-) were significantly increased, with the increase rates of these 8 beneficial aroma compounds being 1888.4%, 238.3%, 7121.6%, 371.0%, 133.2%, 775.0%, 647.4% and 240.3%, respectively. In contrast, the harmful aroma compound with horseradish (Lu T.et al., 2022) (Benzene, (2-isothiocyanatoethyl)-) was significantly decreased by 47.9%.