Quality characterization of fermented rice wines: A multivariate study on physicochemical indices, biogenic amines, volatile profiles, and sensory evaluation Authors: HuaWei Yuan, XinYu Zhao, Hao Yuan, ZiYing Tian, Zhou Xu, TingNa Xie, Kai Lou Published: 2026 ABSTRACT This study comprehensively evaluated seven Chinese fermented rice wines by integrating physicochemical indices, free amino acids, organic acids, biogenic amines (BAs), volatile compounds, and sensory scores to identify key quality determinants. Results revealed substantial regional variations. Sample MJ5 (Guangxi) achieved the highest sensory acceptance (93.6 points), attributed to a harmonious balance of high alcohol content, umami-enhancing amino acids, and abundant floral/fruity esters. Conversely, MJ6 (Hubei) displayed comparable flavor complexity but carried the highest BA load, raising safety concerns, while MJ2 (Chongqing) exhibited the safest profile yet was sensorially plain. Notably, no single parameter dictated overall quality; instead, optimal rice wine requires synergistic integration of volatile and taste-active components alongside rigorous BA control. These findings provide a scientific foundation for establishing multi-parameter quality standards for fermented rice wines Introduction Fermented rice wine, known as one of the oldest traditional alcoholic beverages in China, has been an integral part of Asian dietary culture for thousands of years. This distinctive beverage is produced through the fermentation of various cereals, primarily glutinous rice, using complex starter cultures such as Qu that contain diverse assemblages of microorganisms including molds, yeasts, and bacteria (Liu et al., 2022; Peng et al., 2025). The resulting product is characterized by its smooth mouthfeel, low alcohol content, and rich nutritional profile, which includes oligosaccharides, polypeptides, amino acids, vitamins, and minerals. These nutritional attributes, combined with its unique flavor profile, have contributed to the widespread popularity of fermented rice wine among consumers not only in China but also in other parts of the world (Koay et al., 2024; Wong et al., 2023). The sensory appeal of rice wine derives from the intricate balance of volatile and non-volatile compounds generated during the fermentation process, including esters, alcohols, organic acids, amino acids, and phenolic compounds, which collectively contribute to its characteristic aroma and taste (Chen et al., 2018; Yu et al., 2019). Despite its long history and cultural significance, the Chinese rice wine industry faces considerable challenges related to product standardization and quality consistency. The market encompasses a diverse array of rice wine products manufactured using different raw materials, fermentation technologies, and production scales, ranging from traditional household brewing to modern industrial manufacturing (Lin et al., 2021; Yang, Xia, Wang, Hu, et al., 2017). This diversity, while contributing to the richness of rice wine varieties, has resulted in considerable variability in product quality and compositional characteristics. Studies have demonstrated that the concentrations of flavor compounds such as esters, amino acids, phenolic acids, and organic acids vary significantly depending on the raw materials and fermentation conditions employed (Shen et al., 2024; Zhao et al., 2020). For instance, Shen et al. (2024) reported that buckwheat-fermented rice wine exhibited the highest ester content among five different cereal-based wines, while glutinous rice wine showed the highest organic acid concentration. Such variability, although inherent to traditional fermented products, poses significant obstacles to the establishment of comprehensive quality standards and the consistent delivery of products that meet consumer expectations. The quality of fermented rice wine is determined by a complex interplay of physicochemical parameters and flavor-active compounds. Key indicators such as total acid, total sugar, alcohol content, reducing sugar, and total soluble solids serve as fundamental quality markers that influence both the stability and sensory characteristics of the final product (Qian et al., 2023; Yang et al., 2021). During the fermentation process, dynamic changes occur in these parameters; for example, total sugar content typically decreases while alcohol and total acid contents increase as fermentation progresses (Liu et al., 2022; Zhao et al., 2023). Beyond these basic indices, amino acids and organic acids play particularly important roles in shaping the taste profile of rice wine. Amino acids contribute not only to nutritional value but also to taste characteristics, with bitter amino acids often comprising more than 50% of the total amino acid content in some rice wine varieties (Deng et al., 2025; Shen et al., 2024). Organic acids, derived from both raw materials and microbial metabolism during fermentation, influence the acidity, buffering capacity, and overall flavor balance of the final product (Yang, Li, et al., 2024; Huang et al., 2023). The complex interactions among these components ultimately determine the sensory quality and consumer acceptability of rice wine. Recent advances in analytical methodologies have enabled more comprehensive characterization of rice wine composition and its relationship to sensory quality. Headspace solid-phase microextraction coupled with gas chromatography–mass spectrometry (HS-SPME-GC–MS) has been widely employed to identify volatile flavor compounds, while high-performance liquid chromatography (HPLC) facilitates the quantification of non-volatile components such as organic acids and amino acids (Liu et al., 2025; Yang, Xia, Wang, Hu, et al., 2017). Studies utilizing these techniques have revealed the presence of numerous flavor-active compounds in rice wine, including esters, alcohols, aldehydes, acids, and phenolic compounds (Cui et al., 2024; Yu et al., 2019). The application of odor activity values (OAV) has further enabled the identification of key aroma compounds that contribute most significantly to the overall flavor profile (Chen et al., 2013; Jiao et al., 2025). Deng et al. (2025) demonstrated that different fermentation microorganisms can substantially influence the volatile organic compound profile, with certain non-Saccharomyces yeasts enhancing the production of pleasant aroma compounds while reducing the content of bitter amino acids. These methodological advances provide powerful tools for establishing correlations between chemical composition and sensory attributes, thereby supporting quality improvement and standardization efforts. The relationship between physicochemical parameters and sensory quality has been the subject of increasing research attention, yet substantial knowledge gaps remain. While individual studies have examined specific aspects of rice wine composition, integrated investigations that simultaneously analyze multiple physicochemical indices, flavor components, and sensory attributes are relatively scarce (Yu et al., 2022; Zhang et al., 2024). Furthermore, the lack of comprehensive quality standards for fermented rice wine in China has resulted in products with inconsistent quality characteristics, potentially undermining consumer confidence and limiting market development (Peng et al., 2025; Yuan et al., 2025). Sensory evaluation, which integrates olfactory, gustatory, and tactile perceptions, provides a holistic assessment of rice wine quality that complements instrumental analysis (Koay et al., 2024; Wong et al., 2023). However, systematic studies correlating comprehensive physicochemical profiles with sensory evaluation results across different commercial rice wine products are needed to establish meaningful quality benchmarks. Although numerous studies have separately investigated the physicochemical properties, volatile profiles, or sensory characteristics of fermented rice wines, a comprehensive analysis that simultaneously examines all these parameters—together with free amino acids, organic acids, and biogenic amines—within a single set of commercial samples remains scarce. This study addresses this gap by adopting a holistic analytical framework that integrates multiple quality indicators and directly correlates them with sensory evaluation data. Our work offers several novel contributions: (i) it provides a systematic quality comparison of seven geographically distinct commercial rice wine products, reflecting real-world variability that laboratory-scale fermentations cannot capture; (ii) it quantifies the critical trade-off between flavor complexity (driven by esters and higher alcohols) and food safety (reflected by biogenic amine accumulation), revealing that high sensory scores do not necessarily accompany low safety risks; (iii) it employs multivariate statistical approaches to identify the specific compounds that most strongly influence consumer acceptance, thereby offering actionable targets for quality control; and (iv) it establishes a scientific rationale for moving beyond single-parameter quality assessment toward a multi-indicator standardization system. Accordingly, the specific objectives of this study were to comprehensively characterize the physicochemical, amino acid, organic acid, biogenic amine, and volatile profiles of seven commercial rice wines, correlate these chemical parameters with sensory evaluation scores to identify key quality drivers, and provide a scientific basis for establishing multi-parameter quality standards that balance flavor complexity with safety considerations, ultimately supporting product consistency and sustainable development of the rice wine industry. Materials and methods Collection of fermented rice wine samples Fermented rice wine samples were collected from seven rice wine manufacturers located in Anyuan County, Jiangxi Province; Chongqing City; Dazhou City, Sichuan Province; Jinping County, Guizhou Province; Laibin City, Guangxi Zhuang Autonomous Region; Xiaogan City, Hubei Province; and Yibin City, Sichuan Province, China. The samples were sequentially labeled MJ1, MJ2, MJ3, MJ4, MJ5, MJ6, and MJ7. Despite the general consistency in the basic production workflow—including steaming, cooling, starter mixing, saccharification and fermentation, filtration, and clarification, without any distillation or blending—across all manufacturers, notable differences in raw materials, processing parameters, and storage conditions cannot be excluded. For instance, variations in glutinous rice cultivars, the microbial and enzymatic composition of Qu starters, fermentation temperature and duration, as well as post-fermentation aging and storage practices, may substantially influence the final physicochemical and sensory profiles. However, these specific parameters were not disclosed by the manufacturers owing to proprietary protocols, which constitutes a limitation of this study and underscores the need for future research that systematically monitors these variables to better elucidate their individual contributions to product quality and safety. All samples were collected in March 2025 and stored at −18 °C for up to 10 days prior to analysis. Pretreatment of fermented rice wine After opening, samples were centrifuged (Centrifuge: TGL-20 M, Luxiangyi, Shanghai, China; 5000 ×g, 10 min, 4 °C). The supernatant was transferred into pre-sterilized 50-mL amber glass bottles, sealed with PTFE/silicone septa, and stored at 4 °C in darkness. All analyses were completed within 7 days post-opening to prevent oxidation or microbial spoilage. For assays requiring particle-free solutions, samples were filtered through 0.22 μm nylon membrane filters (Jinteng, Tianjin, China) or diluted 1:5 to 1:20 (v/v) with ultrapure water (18.2 MΩ·cm, Milli-Q®). Measurement of physicochemical indices The ethanol content was measured using a fully automatic density-refractometer (RM40, Mettler Toledo) after distillation. Total acid (titratable acidity) was determined by titration with 0.1 M NaOH to pH 8.2 (pH meter: FiveEasy™ FE28, Mettler Toledo), expressed as lactic acid equivalents (g/L), following the method of Chen et al. (2020). Volatile acids were quantified by steam distillation followed by titration (GB/T 15038—2006, General Administration of Quality Supervision, Inspection and Quarantine of China, & Standardization Administration of China, 2006). Reducing sugar was analyzed by the DNS colorimetric method (He et al., 2022) using a UV–Vis spectrophotometer (Meixi UV-1500, Shanghai, China) at 540 nm against a glucose standard curve (0–1.0 mg/mL, R2 = 0.999). Amino acid nitrogen (AAN) was determined by formol titration (GB 5009.235—2016; National Health Commission of China and Standardization Administration of China, 2016a) with 0.05 M NaOH. Determination of amino acids in fermented rice wine Free amino acids were quantified using an automatic amino acid analyzer (Sykam S433, Munich, Germany) following the method of Yan et al. (2021) with minor modifications. Samples were deproteinized by mixing 1 mL of supernatant with 1 mL of 5% sulfosalicylic acid, vortexed for 30 s, incubated at 4 °C for 1 h, and then centrifuged at 10,000 ×g for 15 min at 4 °C. The clear supernatant was filtered through 0.22-μm filter membraneChromatographic separation was performed on LCA K07/Li cation-exchange resin column (150 × 4.6 mm, 7 μm) using gradient elution with lithium citrate buffers (pH 2.2 and 4.3) at a flow rate of 0.4 mL/min, with the column oven at 57 °C and the reaction coil at 135 °C. Post-column ninhydrin derivatization enabled detection at 570 nm (primary amines) and 440 nm (proline). Quantification was carried out using external standard curves (0.1–50 μmol/L) prepared from 17 amino acid standards (Sigma-Aldrich, purity ≥ 99%), with correlation coefficients (R2) exceeding 0.995 for all analytes. Determination of organic acids in fermented rice The organic acids—including oxalic, acetic, lactic, malic, succinic, citric, and tartaric acids—were qualitatively and quantitatively determined following the method of Yang et al. (2018) with appropriate modifications, using Shimadzu LC-20 A HPLC equipped with Inertsil ODS-SP C18 column (250 × 4.6 mm, 5 μm). The mobile phase was 5 mM H₂SO₄ at an isocratic flow rate of 0.6 mL/min, with the column temperature maintained at 60 °C and UV detection at 210 nm. Prior to injection, samples were filtered through 0.22 μm membranes and diluted 10-fold. Quantification was performed using external standards (Sigma-Aldrich) over a linear range of 1–200 mg/L, with all calibration curves yielding R2 > 0.99. Determination of biogenic amines in fermented rice wine The determination of biogenic amines was carried out according to GB 5009.208—2016 (National Health Commission of China and Standardization Administration of China, 2016b), using dansyl chloride derivatization. The derivatized samples were analyzed by HPLC (Shimadzu LC-20 A) equipped with a ZORBAX Eclipse Plus C18 column (250 × 4.6 mm, 5 μm) and a gradient elution program with mobile phase A (0.1 M ammonium acetate) and B (acetonitrile). Fluorescence detection (FLD) was performed at excitation and emission wavelengths of 254 nm. The method exhibited satisfactory recovery rates of 92–105% in spiked samples. Analysis of volatile flavor compounds Volatile compounds in the rice wine samples were analyzed by headspace solid-phase microextraction coupled with gas chromatography–mass spectrometry (HS-SPME-GC–MS). Briefly, each sample (3.0 mL) was mixed with 1.2 g NaCl in a 15-mL headspace bottle, and 10 μL of 2-octanol (0.0716 mg/mL in methanol) was added as the internal standard. The vial was pre-equilibrated in a water bath at 60 °C for 15 min, and volatiles were then extracted at the same temperature for 45 min using a 50/30 μm DVB/CAR/PDMS fiber (Supelco, Bellefonte, PA, USA). The extracted compounds were desorbed and analyzed on a Shimadzu QP2020 GC–MS system equipped with an HP-INNOWax column (60 m × 0.25 mm i.d., 0.25 μm film thickness, J&W Scientific). Ultra-high-purity helium was used as the carrier gas at a constant flow of 1.0 mL/min. The oven temperature was programmed as follows: held at 40 °C for 5 min, raised to 100 °C at 5 °C/min, then increased to 230 °C at 6 °C/min and held for 10 min. The transfer line and ion source temperatures were set to 250 °C and 230 °C, respectively. Compound identification was performed by matching mass spectra against the NIST-14.0 library (match >85%), and semi-quantification was achieved using the internal standard method based on 2-octanol, following the procedure described by Xiao et al. (2021). Sensory properties analysis Sensory evaluation of the rice wine samples was conducted by a trained panel consisting of twelve judges (six females and six males, aged 20 to 50 years), including rice wine critics and professional rice wine technicians, following the guidelines of ISO 11035 (International Organization for Standardization, 1994) and similar to the approach described by Wang et al. (2013). Fourteen sensory descriptors were selected to characterize the samples, covering aroma (alcoholic, fruity, honey, floral, cereal), taste (sweetness, acidity, astringency, umami, bitterness), mouthfeel (continuation, fullness), and appearance (yellowness, turbidity). Each descriptor was rated on a nine-point intensity scale (0 = none, 1–2 = very weak, 3–4 = ordinary, 5–6 = moderate, 7–8 = intense; 9 = very intense). A 100-point scoring system was also applied, allocating 60 points to taste, 20 points to aroma, 10 points to color, and 10 points to full-body. A total score of 100 indicated outstanding quality. In accordance with earlier reports (Yuan et al., 2022), the sensory tests were performed in a dedicated testing room maintained at 20 °C, equipped with uniform lighting, and free from noise or other distracting stimuli. Blind assessments were performed using randomly coded samples. The panelists vigorously swirled each cup and inhaled the headspace vapor to evaluate the aroma. Rice wine samples (30 mL) were poured into identical cups, served at 20 °C, and presented to the judges with only the sample codes visible. Between samples, panelists rinsed their mouths with water. A reference card listing all sensory attributes along with their definitions and the nine-point intensity scale (0–9) was provided to assist the tasting. The judges were asked to rate each descriptor for every sample. Statistical analysis All analyses were performed in triplicate. Results were examined by one-way analysis of variance (ANOVA) employing SPSS (version 22.0, IBM Corp., Armonk, NY, USA). Using the Duncan test at p < 0.05, significant variations across the data were found. The findings were presented as mean ± standard deviation (SD). Results and discussion Physicochemical properties and sensory quality of fermented rice wine Before evaluating the individual samples, it is important to define what constitutes ‘good quality’ in the context of this study. Based on our integrated analytical framework and sensory validation, we considered a rice wine to be of good quality when it exhibited: (1) a high overall sensory score (> 90 points), reflecting consumer acceptance; (2) a chemically complex and balanced profile—characterized by abundant esters and higher alcohols, moderate-to-high umami amino acids, and a well-structured organic acid composition; (3) biogenic amine levels within acceptable safety limits (provisionally <100 mg/L total BAs), and (4) harmonized integration of these attributes rather than excellence in any single parameter. This multi-dimensional definition guided our comparative assessment of the seven commercial samples. The fundamental physicochemical parameters and sensory scores of the seven rice wine samples (MJ1–MJ7) are presented in Table 1(Supplementary Table S1 and S2). These indices directly reflect the metabolic activities of the microbial community and significantly influence the final product's quality and consumer acceptance (Peng et al., 2025; Yang, Xia, et al., 2024).Table 1The physicochemical parameters and sensory scores of the seven rice wine samples.MJ1MJ2MJ3MJ4MJ5MJ6MJ7Total Acidity (g/L)3.23 ± 0.16c1.90 ± 0.09e1.99 ± 0.01e2.74 ± 0.09d5.53 ± 0.15a5.02 ± 0.23b2.51 ± 0.08dVolatile acid (g/L)0.18 ± 0.01f0.23 ± 0.01e0.24 ± 0.01e0.50 ± 0.01b0.67 ± 0.02a0.43 ± 0.01c0.32 ± 0.02dTotal sugar (g/L)96.84 ± 0.13g171.07 ± 0.91c159.19 ± 0.32d188.89 ± 0.29b137.29 ± 0.19e209.75 ± 0.47a127.80 ± 0.83fReducing Sugar(g/L)85.28 ± 0.18e160.78 ± 0.61a112.85 ± 0.62b108.62 ± 0.56c71.29 ± 0.64f109.04 ± 1.01c87.19 ± 0.89dAmino acid nitrogen (mg/100 mL)12.13 ± 0.20e20.80 ± 0.59c24.53 ± 0.74b20.59 ± 0.43c27.07 ± 0.40a14.72 ± 0.02d11.49 ± 0.39eAlcohol Content (% vol)6.81 ± 0.02b4.45 ± 0.01e4.43 ± 0.02e5.42 ± 0.03d8.35 ± 0.03a5.43 ± 0.01d6.14 ± 0.03cSensory score (points)83.4 ± 2.2d88.8 ± 2.4c81.4 ± 2.6e80.3 ± 2.4e93.6 ± 1.8a91.4 ± 2.5b88.9 ± 1.4cNote: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). Table 1The physicochemical parameters and sensory scores of the seven rice wine samples.MJ1MJ2MJ3MJ4MJ5MJ6MJ7Total Acidity (g/L)3.23 ± 0.16c1.90 ± 0.09e1.99 ± 0.01e2.74 ± 0.09d5.53 ± 0.15a5.02 ± 0.23b2.51 ± 0.08dVolatile acid (g/L)0.18 ± 0.01f0.23 ± 0.01e0.24 ± 0.01e0.50 ± 0.01b0.67 ± 0.02a0.43 ± 0.01c0.32 ± 0.02dTotal sugar (g/L)96.84 ± 0.13g171.07 ± 0.91c159.19 ± 0.32d188.89 ± 0.29b137.29 ± 0.19e209.75 ± 0.47a127.80 ± 0.83fReducing Sugar(g/L)85.28 ± 0.18e160.78 ± 0.61a112.85 ± 0.62b108.62 ± 0.56c71.29 ± 0.64f109.04 ± 1.01c87.19 ± 0.89dAmino acid nitrogen (mg/100 mL)12.13 ± 0.20e20.80 ± 0.59c24.53 ± 0.74b20.59 ± 0.43c27.07 ± 0.40a14.72 ± 0.02d11.49 ± 0.39eAlcohol Content (% vol)6.81 ± 0.02b4.45 ± 0.01e4.43 ± 0.02e5.42 ± 0.03d8.35 ± 0.03a5.43 ± 0.01d6.14 ± 0.03cSensory score (points)83.4 ± 2.2d88.8 ± 2.4c81.4 ± 2.6e80.3 ± 2.4e93.6 ± 1.8a91.4 ± 2.5b88.9 ± 1.4cNote: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). The physicochemical parameters and sensory scores of the seven rice wine samples. Note: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). The alcohol content, a primary indicator of fermentation efficiency, varied significantly (p < 0.05) among the samples, ranging from 4.43%vol (MJ3) to 8.35%vol (MJ5). MJ5 exhibited the highest alcohol content, which was consistent with its highest total acidity (5.53 g/L) and volatile acid (0.67 g/L) levels. This suggests a more vigorous fermentation process, potentially driven by a more robust yeast population, as Saccharomyces cerevisiae is known to be a primary producer of ethanol and organic acids during rice wine fermentation (Tang & Peng, 2024; Yang, Xia, Wang, Yu, & Ai, 2017). The high sensory score of MJ5 (93.6 points) corroborates that a balanced increase in these parameters contributes positively to the overall flavor profile. Conversely, MJ2 and MJ3 showed the lowest alcohol content (∼4.4%vol) and total acidity (∼1.9 g/L), which correlated with their lower sensory scores, indicating that insufficient fermentation leads to a flat and less desirable taste (Koay et al., 2024). Total sugar and reducing sugar contents displayed a complex pattern. MJ6 had the highest total sugar (209.75 g/L), while MJ1 had the lowest (96.84 g/L). Interestingly, MJ1 also had the highest reducing sugar (85.28 g/L) as a proportion of its total sugar, indicating a more complete saccharification. The high sugar content in MJ6 did not translate into the highest alcohol, suggesting that its saccharification power may have exceeded the yeast's fermentation capacity or that the microbial community favored sugar accumulation over alcohol production. This observation aligns with studies showing that the balance between Rhizopus and Saccharomyces species dictates the final sugar-to-alcohol ratio (He et al., 2025; Zhao et al., 2023). The amino acid nitrogen (AAN) content, crucial for flavor and yeast nutrition, was highest in MJ5 (27.07 mg/100 mL) and lowest in MJ7 (11.49 mg/100 mL). High AAN contributes to the umami taste and serves as a precursor for higher alcohols, which explains the robust and complex flavor of MJ5 (Deng et al., 2025; Shi et al., 2022). The strong positive correlation between AAN, alcohol content, and sensory score across the samples underscores the importance of nitrogen metabolism for overall rice wine quality (Qian et al., 2023). The sensory evaluation, depicted by the radar chart, revealed a distinct shared flavor framework across all seven samples (MJ1–MJ7), characterized by high intensities in Alcoholic, Fruity, Floral, and Honey attributes (predominantly scoring above 6 on a 9-point scale) and uniformly low scores for Bitterness, Umami, Astringency, and Cereal (consistently hovering around 3), indicating a clean and aromatically driven profile (Fig. 1); however, the polygonal areas exhibited substantial inter-sample variation, with MJ5 demonstrating the most robust overall intensity—approaching the maximum score of 9 across multiple dimensions including Alcoholic, Fruity, Honey, Fullness, and Continuation—closely followed by MJ3, whereas MJ7 presented the weakest performance with the smallest radar envelope, particularly in the core aromatic attributes (Fig. 1).Fig. 1Radar chart of sensory evaluation scores for samples across 14 attributes.Fig. 1 Radar chart of sensory evaluation scores for samples across 14 attributes. Furthermore, MJ1 displayed a polarized profile characterized by high aromatic notes but notably lower Fullness and Yellowness, while MJ6 was distinctly differentiated by the highest Turbidity score and above-average Fullness, with MJ2 and MJ4 maintaining intermediate profiles (Fig. 1). Collectively, these findings establish MJ5 and MJ3 as the samples with the most comprehensive and intense sensory qualities, whereas MJ7 exhibits the lowest overall sensory strength, and the distinct features of MJ6 and MJ1 serve as key distinguishing factors within the analyzed group. Free amino acid and organic acid profiles Free amino acids (FAAs) are critical for both the nutritional value and the taste of rice wine, contributing to sweetness, bitterness, and umami (Shen et al., 2024). The total FAA content varied dramatically, with MJ7 showing the highest concentration (95.23 mg/100 mL), followed by MJ5 (88.53 mg/100 mL), while MJ1 had the lowest (39.44 mg/100 mL) (Table 2, Supplementary Table S3).Table 2Amino acid contents of different fermented rice wines (mg/100 mL).MJ1MJ2MJ3MJ4MJ5MJ6MJ7Asp3.81 ± 0.02e6.41 ± 0.02d6.45 ± 0.16d6.35 ± 0.07d8.98 ± 0.09b6.82 ± 0.03c9.15 ± 0.14aThr1.62 ± 0.03e2.40 ± 0.05d2.35 ± 0.01d2.43 ± 0.03c3.48 ± 0.01a2.70 ± 0.02b3.54 ± 0.06aSer2.18 ± 0.03f3.37 ± 0.02d3.24 ± 0.03e3.36 ± 0.02d4.87 ± 0.02b3.81 ± 0.03c5.09 ± 0.05aGlu8.25 ± 0.04e12.48 ± 0.05d12.49 ± 0.03d12.40 ± 0.09d19.08 ± 0.05b14.06 ± 0.14c19.40 ± 0.04aGly2.43 ± 0.03f3.16 ± 0.01e3.16 ± 0.01e3.28 ± 0.03d5.05 ± 0.01b3.64 ± 0.03c7.32 ± 0.11aAla2.39 ± 0.09e3.64 ± 0.02cd3.60 ± 0.02d3.83 ± 0.02c5.84 ± 0.03b3.75 ± 0.04cd6.56 ± 0.26aCys0.57 ± 0.02d0.56 ± 0.01d0.51 ± 0.01f0.53 ± 0.01e1.15 ± 0.02b0.76 ± 0.02c1.22 ± 0.03aVal1.95 ± 0.04e3.53 ± 0.05d3.73 ± 0.03c3.60 ± 0.02d4.40 ± 0.06b3.48 ± 0.16d4.76 ± 0.05aMet0.15 ± 0.01e0.23 ± 0.01d0.23 ± 0.01d0.23 ± 0.01d0.40 ± 0.02b0.30 ± 0.01c0.75 ± 0.04aIle1.40 ± 0.03e2.56 ± 0.03c2.72 ± 0.06b2.65 ± 0.03b3.04 ± 0.02a2.36 ± 0.11d3.11 ± 0.07aLeu2.33 ± 0.03f4.44 ± 0.03e4.51 ± 0.02d4.65 ± 0.04c5.05 ± 0.04b4.41 ± 0.02e5.13 ± 0.04aTyr1.14 ± 0.02f1.55 ± 0.01c1.55 ± 0.02c1.55 ± 0.03c1.55 ± 0.04c1.55 ± 0.05c1.55 ± 0.06cPhe1.69 ± 0.02e3.28 ± 0.02d3.25 ± 0.04d3.27 ± 0.03d3.77 ± 0.03b3.50 ± 0.08c4.03 ± 0.07aLys1.83 ± 0.03f2.74 ± 0.05e2.83 ± 0.08d3.00 ± 0.02c3.69 ± 0.05b2.96 ± 0.06c3.89 ± 0.02aHis1.25 ± 0.02e1.63 ± 0.02d1.74 ± 0.06c1.73 ± 0.04c3.38 ± 0.01a2.07 ± 0.03b3.42 ± 0.05aArg3.92 ± 0.08f6.60 ± 0.03e6.77 ± 0.04d6.65 ± 0.06e9.07 ± 0.02b7.30 ± 0.07c9.71 ± 0.03aPro2.53 ± 0.09e3.08 ± 0.02d3.02 ± 0.07d3.07 ± 0.02d5.05 ± 0.06b3.62 ± 0.02c5.83 ± 0.02aTotal39.44 ± 0.14e61.64 ± 0.33d62.15 ± 0.44d62.42 ± 0.51d88.53 ± 0.41b66.81 ± 0.73c95.23 ± 0.53aNote: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). Table 2Amino acid contents of different fermented rice wines (mg/100 mL).MJ1MJ2MJ3MJ4MJ5MJ6MJ7Asp3.81 ± 0.02e6.41 ± 0.02d6.45 ± 0.16d6.35 ± 0.07d8.98 ± 0.09b6.82 ± 0.03c9.15 ± 0.14aThr1.62 ± 0.03e2.40 ± 0.05d2.35 ± 0.01d2.43 ± 0.03c3.48 ± 0.01a2.70 ± 0.02b3.54 ± 0.06aSer2.18 ± 0.03f3.37 ± 0.02d3.24 ± 0.03e3.36 ± 0.02d4.87 ± 0.02b3.81 ± 0.03c5.09 ± 0.05aGlu8.25 ± 0.04e12.48 ± 0.05d12.49 ± 0.03d12.40 ± 0.09d19.08 ± 0.05b14.06 ± 0.14c19.40 ± 0.04aGly2.43 ± 0.03f3.16 ± 0.01e3.16 ± 0.01e3.28 ± 0.03d5.05 ± 0.01b3.64 ± 0.03c7.32 ± 0.11aAla2.39 ± 0.09e3.64 ± 0.02cd3.60 ± 0.02d3.83 ± 0.02c5.84 ± 0.03b3.75 ± 0.04cd6.56 ± 0.26aCys0.57 ± 0.02d0.56 ± 0.01d0.51 ± 0.01f0.53 ± 0.01e1.15 ± 0.02b0.76 ± 0.02c1.22 ± 0.03aVal1.95 ± 0.04e3.53 ± 0.05d3.73 ± 0.03c3.60 ± 0.02d4.40 ± 0.06b3.48 ± 0.16d4.76 ± 0.05aMet0.15 ± 0.01e0.23 ± 0.01d0.23 ± 0.01d0.23 ± 0.01d0.40 ± 0.02b0.30 ± 0.01c0.75 ± 0.04aIle1.40 ± 0.03e2.56 ± 0.03c2.72 ± 0.06b2.65 ± 0.03b3.04 ± 0.02a2.36 ± 0.11d3.11 ± 0.07aLeu2.33 ± 0.03f4.44 ± 0.03e4.51 ± 0.02d4.65 ± 0.04c5.05 ± 0.04b4.41 ± 0.02e5.13 ± 0.04aTyr1.14 ± 0.02f1.55 ± 0.01c1.55 ± 0.02c1.55 ± 0.03c1.55 ± 0.04c1.55 ± 0.05c1.55 ± 0.06cPhe1.69 ± 0.02e3.28 ± 0.02d3.25 ± 0.04d3.27 ± 0.03d3.77 ± 0.03b3.50 ± 0.08c4.03 ± 0.07aLys1.83 ± 0.03f2.74 ± 0.05e2.83 ± 0.08d3.00 ± 0.02c3.69 ± 0.05b2.96 ± 0.06c3.89 ± 0.02aHis1.25 ± 0.02e1.63 ± 0.02d1.74 ± 0.06c1.73 ± 0.04c3.38 ± 0.01a2.07 ± 0.03b3.42 ± 0.05aArg3.92 ± 0.08f6.60 ± 0.03e6.77 ± 0.04d6.65 ± 0.06e9.07 ± 0.02b7.30 ± 0.07c9.71 ± 0.03aPro2.53 ± 0.09e3.08 ± 0.02d3.02 ± 0.07d3.07 ± 0.02d5.05 ± 0.06b3.62 ± 0.02c5.83 ± 0.02aTotal39.44 ± 0.14e61.64 ± 0.33d62.15 ± 0.44d62.42 ± 0.51d88.53 ± 0.41b66.81 ± 0.73c95.23 ± 0.53aNote: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). Amino acid contents of different fermented rice wines (mg/100 mL). Note: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). The high total FAA in MJ7 and MJ5 was primarily driven by elevated levels of umami-contributing amino acids like glutamic acid (Glu, 19.40 and 19.08 mg/100 mL, respectively) and aspartic acid (Asp, 9.15 and 8.98 mg/100 mL, respectively). These findings are significant because Glu and Asp are known to enhance the pleasant, full-bodied mouthfeel of fermented products (Cen et al., 2023; Shi et al., 2022). The presence of bitter amino acids, such as Val, Ile, and Leu, was also higher in MJ5 and MJ7. However, their high sensory scores suggest that these bitter notes were well-balanced by the high sugar and umami components, creating a complex and desirable flavor profile, a phenomenon also observed in other traditional rice wines (He et al., 2025; Qian et al., 2023). The lowest FAA content in MJ1 correlates with its lower sensory score, confirming that adequate proteolysis is essential for flavor development (Liu et al., 2022). The organic acid composition, which influences the taste and microbial stability of the wine, also showed significant regional variation (Table 3, Supplementary Table S4). MJ5 and MJ6 were characterized by very high levels of citric acid (CA, 245.60 and 222.14 mg/100 mL, respectively) and acetic acid (AA, 33.51 and 18.09 mg/100 mL). While high acetic acid can be detrimental, its presence in MJ5, combined with high levels of malic acid (MA, 233.43 mg/L) and succinic acid (SA, 6.71 mg/L), likely contributed to a sharp yet complex acidity that is often appreciated in certain rice wine styles (Liu et al., 2022). In contrast, MJ2 and MJ3 had much lower levels of most organic acids, which aligns with their lower total acidity and flat sensory profile. The exceptionally high tartaric acid (TA) in MJ7 (111.36 mg/L) and MJ6 (100.69 mg/L) is noteworthy, as TA is not a common microbial product but may be derived from raw materials or specific fermentation conditions (Yang et al., 2021). The data suggest that the interplay between different organic acids, rather than any single one, defines the sensory perception of sourness and overall taste balance (Cao et al., 2025).Table 3Organic acid contents of different fermented rice wines (mg/100 mL).MJ1MJ2MJ3MJ4MJ5MJ6MJ7Oxalic acid (OA)0.22 ± 0.01e0.53 ± 0.01c0.57 ± 0.01b0.21 ± 0.01e1.00 ± 0.02a0.42 ± 0.01d1.02 ± 0.04aTartaric acid (TA)75.06 ± 0.43e47.90 ± 2.42g64.54 ± 0.87f97.60 ± 0.48c83.26 ± 0.10d100.69 ± 0.42b111.36 ± 0.11aMalic acid (MA)231.55 ± 0.94b135.34 ± 1.04f218.51 ± 0.85c191.44 ± 1.83d233.43 ± 1.87b282.32 ± 0.71a153.32 ± 0.69eLactic acid(LA)50.71 ± 0.12d63.62 ± 0.46c66.67 ± 0.09b74.27 ± 0.36a32.45 ± 0.87f40.98 ± 0.34e67.10 ± 0.22bAcetic acid (AA)8.42 ± 0.05c4.82 ± 0.12f3.96 ± 0.06g5.73 ± 0.12e33.51 ± 0.82a18.09 ± 0.60b7.25 ± 0.05dCitric acid (CA)84.30 ± 0.15f78.49 ± 1.02g156.80 ± 1.50e187.18 ± 1.89d245.60 ± 2.22a222.14 ± 1.50b197.27 ± 0.41cSuccinic acid (SA)4.78 ± 0.13d1.72 ± 0.07e1.91 ± 0.08e7.61 ± 0.19b6.71 ± 0.09c7.85 ± 0.14b11.82 ± 0.31aNote: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). Table 3Organic acid contents of different fermented rice wines (mg/100 mL).MJ1MJ2MJ3MJ4MJ5MJ6MJ7Oxalic acid (OA)0.22 ± 0.01e0.53 ± 0.01c0.57 ± 0.01b0.21 ± 0.01e1.00 ± 0.02a0.42 ± 0.01d1.02 ± 0.04aTartaric acid (TA)75.06 ± 0.43e47.90 ± 2.42g64.54 ± 0.87f97.60 ± 0.48c83.26 ± 0.10d100.69 ± 0.42b111.36 ± 0.11aMalic acid (MA)231.55 ± 0.94b135.34 ± 1.04f218.51 ± 0.85c191.44 ± 1.83d233.43 ± 1.87b282.32 ± 0.71a153.32 ± 0.69eLactic acid(LA)50.71 ± 0.12d63.62 ± 0.46c66.67 ± 0.09b74.27 ± 0.36a32.45 ± 0.87f40.98 ± 0.34e67.10 ± 0.22bAcetic acid (AA)8.42 ± 0.05c4.82 ± 0.12f3.96 ± 0.06g5.73 ± 0.12e33.51 ± 0.82a18.09 ± 0.60b7.25 ± 0.05dCitric acid (CA)84.30 ± 0.15f78.49 ± 1.02g156.80 ± 1.50e187.18 ± 1.89d245.60 ± 2.22a222.14 ± 1.50b197.27 ± 0.41cSuccinic acid (SA)4.78 ± 0.13d1.72 ± 0.07e1.91 ± 0.08e7.61 ± 0.19b6.71 ± 0.09c7.85 ± 0.14b11.82 ± 0.31aNote: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). Organic acid contents of different fermented rice wines (mg/100 mL). Note: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). Biogenic amines and safety considerations Biogenic amines (BAs) are nitrogenous compounds formed by the decarboxylation of amino acids by microorganisms. High levels of BAs, particularly histamine and tyramine, are of food safety concern.The total BA content varied widely, from 20.47 mg/L in MJ2 to a high of 132.54 mg/L in MJ6 (Table 4, Supplementary Table S5). Putrescine was the most abundant BA in most samples, followed by tyramine and cadaverine. MJ6 showed the highest levels of tyramine (57.38 mg/L) and tryptamine (6.15 mg/L). The high tyramine content in MJ6 is a potential safety concern, as it can cause hypertensive effects in sensitive individuals. This high BA level likely results from the activity of specific lactic acid bacteria (LAB) or other decarboxylase-positive bacteria present during fermentation (Luo et al., 2024; Yuan et al., 2025). In contrast, MJ2 had remarkably low levels of almost all BAs, particularly putrescine (7.59 mg/L) and tyramine (6.73 mg/L), indicating superior microbial control or a fermentation environment less conducive to BA formation.Table 4Biogenic amines contents of different fermented rice wines (mg/L).MJ1MJ2MJ3MJ4MJ5MJ6MJ7Histamine4.32 ± 0.05f3.04 ± 0.07g5.49 ± 0.09e6.12 ± 0.06d14.39 ± 0.12a9.26 ± 0.21b7.11 ± 0.14cPutrescine27.19 ± 0.39d7.59 ± 0.05f47.23 ± 1.02a12.58 ± 0.07e39.15 ± 1.20c42.59 ± 0.78b46.39 ± 0.86aCadaverine1.32 ± 0.04f0.61 ± 0.02g4.53 ± 0.06a2.05 ± 0.07e4.17 ± 0.13b3.79 ± 0.08c2.76 ± 0.06dTryptamine0.42 ± 0.02g0.82 ± 0.01e0.54 ± 0.01f0.91 ± 0.03d2.35 ± 0.05b6.15 ± 0.10a1.49 ± 0.04cTyramine10.64 ± 0.31e6.73 ± 0.03f16.75 ± 0.07b25.52 ± 0.11c27.42 ± 0.16b57.38 ± 0.74a17.28 ± 0.12dSpermidine0.73 ± 0.01e0.62 ± 0.01f1.29 ± 0.02b0.98 ± 0.03c2.61 ± 0.02a2.58 ± 0.04a0.91 ± 0.02dSpermine2.38 ± 0.07a0.34 ± 0.01f0.82 ± 0.02e1.65 ± 0.03b1.27 ± 0.03c2.43 ± 0.05a0.94 ± 0.04dPhenethylamine1.14 ± 0.02f0.72 ± 0.02j3.78 ± 0.03e5.34 ± 0.04d6.45 ± 0.05c8.36 ± 0.06a7.49 ± 0.05bTotal48.14 ± 0.72f20.47 ± 0.13j80.42 ± 1.14d55.15 ± 0.08e97.82 ± 1.32b132.54 ± 1.66a84.38 ± 0.69cNote: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). Table 4Biogenic amines contents of different fermented rice wines (mg/L).MJ1MJ2MJ3MJ4MJ5MJ6MJ7Histamine4.32 ± 0.05f3.04 ± 0.07g5.49 ± 0.09e6.12 ± 0.06d14.39 ± 0.12a9.26 ± 0.21b7.11 ± 0.14cPutrescine27.19 ± 0.39d7.59 ± 0.05f47.23 ± 1.02a12.58 ± 0.07e39.15 ± 1.20c42.59 ± 0.78b46.39 ± 0.86aCadaverine1.32 ± 0.04f0.61 ± 0.02g4.53 ± 0.06a2.05 ± 0.07e4.17 ± 0.13b3.79 ± 0.08c2.76 ± 0.06dTryptamine0.42 ± 0.02g0.82 ± 0.01e0.54 ± 0.01f0.91 ± 0.03d2.35 ± 0.05b6.15 ± 0.10a1.49 ± 0.04cTyramine10.64 ± 0.31e6.73 ± 0.03f16.75 ± 0.07b25.52 ± 0.11c27.42 ± 0.16b57.38 ± 0.74a17.28 ± 0.12dSpermidine0.73 ± 0.01e0.62 ± 0.01f1.29 ± 0.02b0.98 ± 0.03c2.61 ± 0.02a2.58 ± 0.04a0.91 ± 0.02dSpermine2.38 ± 0.07a0.34 ± 0.01f0.82 ± 0.02e1.65 ± 0.03b1.27 ± 0.03c2.43 ± 0.05a0.94 ± 0.04dPhenethylamine1.14 ± 0.02f0.72 ± 0.02j3.78 ± 0.03e5.34 ± 0.04d6.45 ± 0.05c8.36 ± 0.06a7.49 ± 0.05bTotal48.14 ± 0.72f20.47 ± 0.13j80.42 ± 1.14d55.15 ± 0.08e97.82 ± 1.32b132.54 ± 1.66a84.38 ± 0.69cNote: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). Biogenic amines contents of different fermented rice wines (mg/L). Note: values with distinct superscript letters within the same row were statistically significant (p < 0.05).All findings are shown as the mean ± standard deviation (n = 3). Interestingly, MJ5, despite having high total FAA and high sensory scores, had moderate BA levels (97.82 mg/L), suggesting that its dominant microbial community was efficient in converting amino acids to alcohols and esters rather than to undesirable amines. This is consistent with findings that a well-managed fermentation with a balanced microbial ecosystem can reduce BA accumulation (Jiao et al., 2025; Yuan et al., 2025). The data highlight that while MJ6 and MJ7 have complex flavor profiles, their higher BA content necessitates careful process control to ensure safety. The low BAs in MJ2, while indicating safety, unfortunately coincide with a less flavorful product, demonstrating the challenge in balancing safety and sensory quality. Volatile flavor compounds in different fermented rice wines In total, 118 aroma compounds were identified in the seven rice wine samples: 29 alcohols, 47 esters, 19 ketones and aldehydes, 13 acids, 4 phenols, 2 alkenes, and 4 alkanes (Table 5, Supplementary Table S6). The types and numbers of these aroma components indicate that the major flavor contributors in fermented rice wine are alcohols, esters, ketones/aldehydes, and acids, with esters playing a particularly crucial role. The clustered heatmap revealed that the seven rice wine samples segregated into two major groups, with MJ5 and MJ6 forming a distinct cluster characterized by elevated levels of key esters (e.g., ethyl acetate, phenethyl acetate, ethyl lactate, and ethyl caprylate) and higher alcohols (e.g., β-phenylethanol and isoamyl alcohol), alongside unique floral compounds such as linalool, terpineol, and β-ionone in MJ5 (Fig. 2).Table 5Volatile compound composition and content in fermented rice wines (μg/L).Flavor componentsMJ1MJ2MJ3MJ4MJ5MJ6MJ7Alcohols1-Propanol22.03 ± 1.33e9.78 ± 0.16f38.1 ± 1.15c38.87 ± 0.46c35.65 ± 0.48d64.65 ± 2.36a45.03 ± 2.07bIsobutanol654.39 ± 11.01a364.24 ± 27.15e403.17 ± 31.2d299.14 ± 19.91f524.7 ± 19.6b460.92 ± 19.55c557.97 ± 9.51b1-Butanol11.4 ± 0.27c6.82 ± 0.5d7.68 ± 0.41d7.41 ± 0.34d25.6 ± 1.69a12.49 ± 0.85c14.07 ± 1.1bIsoamyl alcohol4731.32 ± 361.83a2139.91 ± 69.64d2323.63 ± 146.19d2513.79 ± 103.79d4993.27 ± 244.58a4287.02 ± 354.1b3833.57 ± 254.75cHexanolND123.62 ± 6.62b119.61 ± 11.34b148.17 ± 2.06aNDNDNDHeptanol13.82 ± 0.67c20.98 ± 1.63b19.7 ± 0.87b21.94 ± 0.55bND14.75 ± 1.02c32.37 ± 1.99a2-Ethylhexanol50.06 ± 1.77b10.84 ± 0.51d16.63 ± 0.97c14.45 ± 0.81c55.62 ± 1.05a55.41 ± 3.01a11.33 ± 0.5d2-Nonanol4.58 ± 0.15c6.04 ± 0.57b8.87 ± 0.45aNDNDND9.1 ± 0.68a2,3-Butanediol220.99 ± 2.74bND51.77 ± 0.58f145.71 ± 3.66c373.69 ± 4.85a98.86 ± 9.64e109.11 ± 7.53dLinalool5.3 ± 0.34cND15.77 ± 0.89b4.33 ± 0.14c227.96 ± 19.32a17.14 ± 1.61b10.65 ± 0.86bOctanol61.02 ± 4.11b44.31 ± 1.84d52.78 ± 3.26c95.75 ± 2.28a49.96 ± 1.85c23.09 ± 1.89f33 ± 2.04e2-Octen-1-ol4.52 ± 0.26aNDND2.48 ± 0.09aNDNDNDNonanol16.55 ± 0.51c15.08 ± 1.14d13.31 ± 0.71e24.93 ± 0.45a15.82 ± 0.37 cd22.71 ± 1.12b16.17 ± 0.21 cdTerpineol7.51 ± 0.3cNDND5.55 ± 0.36c350.53 ± 6.04aND16.76 ± 0.84bMethylthiopropanol11.44 ± 1.12cNDND8.26 ± 0.46c41.22 ± 3.34a14.45 ± 1.04b11.22 ± 0.73cDecanol3.3 ± 0.27c14.94 ± 0.5b95.08 ± 7.89a14.61 ± 0.39bND12.41 ± 0.98b16.52 ± 1.29bUndecanolND7.79 ± 0.36b64.24 ± 4.1a6.46 ± 0.21bNDNDNDCitronellolNDNDNDND9.47 ± 0.1343.68 ± 2.22NDBenzyl alcohol16.68 ± 1.54b13.2 ± 1.2c19.47 ± 0.59a6.46 ± 0.53d12.4 ± 0.5c5.62 ± 0.49d19.68 ± 1.21aβ-Phenylethanol3592.63 ± 150.36d2153.98 ± 146.21f2495.68 ± 217.11f2629.66 ± 63.29e8456.95 ± 307.84a5758.81 ± 233.54b4354.49 ± 166.35cLauryl alcohol84.6 ± 1.66c185.9 ± 12.12b1147.81 ± 104.76a72.5 ± 6.49c1142.32 ± 18.8a35.87 ± 2.12d15.24 ± 0.35dIonol3.59 ± 0.18aNDNDND129.57 ± 5.53NDNDHexadecanolND33.11 ± 2.39a18.11 ± 1.54bND7.16 ± 0.16cNDNDNerolidolND7.8 ± 0.49a5.86 ± 0.43bNDNDNDNDα-CadinolNDND1.38 ± 0.06c5.39 ± 0.34b34.79 ± 1.79aNDNDSpathulenol27.59 ± 2.03c34.43 ± 3.04b32.65 ± 1.89b54.66 ± 4.56aNDND5.41 ± 0.14dα-SantalolND1.18 ± 0.06c11.66 ± 0.63b13.33 ± 0.53aNDNDNDα-Bisabolol8.82 ± 0.12NDNDNDNDNDNDEstersNDNDNDNDNDNDNDEthyl acetate204.95 ± 3.77c46.77 ± 2.03e66.46 ± 4.4e117.82 ± 0.78e1619.27 ± 148.33a459.97 ± 35.58b252.82 ± 6.73cEthyl isobutyrate1.76 ± 0.04NDNDNDNDNDNDIsoamyl acetate139.03 ± 6.81b55.48 ± 4.94c34.18 ± 2.82d62.67 ± 5.28c139.02 ± 4.73b160.54 ± 8.65b295.48 ± 28.57aEthyl caproate26.25 ± 1.05e48.45 ± 3.65dND160.92 ± 7.36b275.64 ± 8.79a135.17 ± 9.26c127.41 ± 8.48cEthyl heptanoateNDNDND4.44 ± 0.43NDNDNDEthyl pyruvateNDNDNDND18.4 ± 1.08a17.57 ± 0.46aNDEthyl lactate216.04 ± 0.77cNDNDND750.25 ± 11.94a317.84 ± 11.02b142.29 ± 1.07dEthyl caprylate6.23 ± 0.19e3.51 ± 0.33eND82.02 ± 5.52d95.55 ± 8.82c477.35 ± 1.42a257.71 ± 4.22bEthyl leucate57.2 ± 2.72b15.15 ± 1.11c12.08 ± 1.08 cd3.88 ± 0.08e197.74 ± 6.66a8.92 ± 0.71de6.21 ± 0.15eHeptyl formateNDNDNDND1.54 ± 0.05a12.78 ± 0.8bNDIsobutyl lactateND2.83 ± 0.27c3.21 ± 0.23c5.35 ± 0.22b3.08 ± 0.19c12.03 ± 1.07a5.07 ± 0.14bAmyl hexanoateNDNDND13.85 ± 0.79a23.1 ± 0.61bNDNDEthyl 3-hydroxybutyrate5.84 ± 0.3NDNDNDNDNDNDEthyl nonanoate6.73 ± 0.61aNDNDNDNDND2.29 ± 0.17bγ-Butyrolactone9.47 ± 0.34e28.48 ± 2.46c32.87 ± 2.58b37.58 ± 3.67aND13.88 ± 0.69dNDEthyl decanoate22.18 ± 1.88cNDND78.76 ± 5.96b82.62 ± 6.07b293.91 ± 12.52a5.49 ± 0.25dDiethyl succinate757.02 ± 32.27b128.71 ± 6.01f228.75 ± 0.79de193.97 ± 2.89e1407.65 ± 50.61a426.16 ± 19.55c244.44 ± 15.44dEthyl benzoateND9.07 ± 0.83dND12.47 ± 0.74c26.01 ± 2.6a15.15 ± 0.74b1.42 ± 0.09eEthyl glutarateNDNDND1.56 ± 0.14a17.68 ± 1.01bNDNDEthyl phenylacetate22.01 ± 1.23c4.42 ± 0.39e5.94 ± 0.23e15.2 ± 0.46d168.37 ± 3.16a33.62 ± 2.05b24.41 ± 1.39cPhenethyl acetate458.03 ± 3.99d176.13 ± 10.96f123.25 ± 4.98 g326.94 ± 11.11e1066.68 ± 28.51a723.84 ± 27.31b628.9 ± 20.51cEthyl laurateNDNDNDNDND62.75 ± 5.9NDEthyl phenylpropionateND0.94 ± 0.04aNDND50.5 ± 1.46bNDNDPhenethyl isobutyrateNDNDNDND15.19 ± 0.19a3.99 ± 0.39b2.55 ± 0.15cDiethyl pimelate6.02 ± 0.52b1.82 ± 0.1 cd0.96 ± 0.04d2.32 ± 0.07c20.62 ± 1.11aNDNDEthyl myristate134.05 ± 5.77b11.75 ± 1.05d8.44 ± 0.72d163.38 ± 9.73aND76.96 ± 7.58c3.91 ± 0.35dγ-Octanolactone3.29 ± 0.31cNDNDND10.26 ± 0.91aND4.83 ± 0.46bDiethyl suberateNDNDND5.73 ± 0.51NDNDNDEthyl pentadecanoateNDNDND2.66 ± 0.12NDNDNDγ-Decanolactone9.37 ± 0.46aNDNDND154.94 ± 3.6bNDNDDiethyl azelateNDNDND15.52 ± 1.28b28.85 ± 1.24aND2.78 ± 0.16cEthyl palmitate12.52 ± 0.43d18.34 ± 1.22d35.81 ± 2.95 cd621.93 ± 45.63a53.99 ± 5.4c399.18 ± 8.86b15.98 ± 0.5dEthyl palmitoleateNDNDND8.67 ± 0.72NDNDNDHexyl acetateNDND1.91 ± 0.05a7.59 ± 0.38bNDNDNDMethyl stearate7.45 ± 0.39b4.95 ± 0.29bc2.58 ± 0.24cND52.94 ± 3.8aND5.83 ± 0.43bcEthyl stearate64.1 ± 3.08c6.41 ± 0.41e5.61 ± 0.46e12.87 ± 0.21d282.08 ± 3.48a113.64 ± 6.35b65.16 ± 5.07cEthyl oleateND3.41 ± 0.1c4.92 ± 0.37c167.7 ± 8.75a25.54 ± 1.36d171.75 ± 9.23a10.53 ± 0.8cEthyl linoleateND5.49 ± 0.27c3.75 ± 0.29c221.74 ± 2.46aND158.36 ± 10.13b5.74 ± 0.45cEthyl linolenateNDNDND7.64 ± 0.61NDNDND Ketones2-Octanone92.31 ± 6.49b86.67 ± 4.76b89.77 ± 6.87b108.52 ± 4.25a75.7 ± 6.74c69.36 ± 3.9c97.54 ± 6.53bAcetoinNDNDNDND117.8 ± 10.59a583.47 ± 16.37bND2-NonanoneND9.44 ± 0.74NDNDNDNDND3-Nonen-2-oneND5.4 ± 0.5a5.09 ± 0.35a2.97 ± 0.11bNDNDNDAcetophenoneNDNDND5.94 ± 0.21aND2.11 ± 0.2c4.9 ± 0.24b2,3-Butanedione14.85 ± 0.17a3.95 ± 0.26cND9.82 ± 0.53bNDNDNDPhenylacetoneNDNDNDND15.57 ± 0.55b5.19 ± 0.21NDβ-Ionone73.62 ± 3.68aNDNDND973.66 ± 19.92bNDNDGeranyl acetone31.94 ± 0.16b9.96 ± 0.9eND14.25 ± 0.84d50.28 ± 3.71a20.16 ± 1.94c32.93 ± 3.07bα-Ionone69.11 ± 5.25aNDNDND1542.16 ± 45.7bNDND AldehydesNonanal13.6 ± 0.42d21.07 ± 1.49c15.86 ± 1.26d32.74 ± 1.79a33.75 ± 2.49a33.17 ± 1.55a26.63 ± 0.87b2-OctenalND13.41 ± 0.74a2.77 ± 0.25b2.97 ± 0.11bNDNDNDFurfural5.67 ± 0.1c5.12 ± 0.44c4.52 ± 0.03c5.97 ± 0.49c182.96 ± 6.74a31.37 ± 1.68bNDBenzaldehydeND74.99 ± 7.41b55.29 ± 3.16c68.94 ± 5.87cND268.03 ± 12.66aND2,4-NonadienalND10.18 ± 0.5a7.98 ± 0.79b5.81 ± 0.21cNDNDND2,4-DecadienalND6.17 ± 0.5a2.21 ± 0.16b6.9 ± 0.68cNDNDNDVeratraldehydeND736.26 ± 63.06NDNDNDNDNDCuminaldehyde7.29 ± 0.37bc6.47 ± 0.54c5 ± 0.14d8.61 ± 0.15a3.51 ± 0.13e8.06 ± 0.79ab8.4 ± 0.75a5-Hydroxymethylfurfural7.92 ± 0.12aNDNDND51.79 ± 5.08bNDND AcidsAcetic acid64.22 ± 2.12c48.52 ± 1.32d32.76 ± 2.88e53.82 ± 3.49d317.33 ± 22.91a156.96 ± 11.02b68.24 ± 4.12cIsobutyric acid26.87 ± 1.66c22 ± 2d24.66 ± 0.94c12.82 ± 0.86e78.65 ± 7.09a20.31 ± 1.4d33.09 ± 1.59bn-Butyric acid4.59 ± 0.4c3.9 ± 0.34d3.55 ± 0.22e5.12 ± 0.4c11.75 ± 1b13.57 ± 0.58a4.76 ± 0.39c2,4-Dimethylvaleric acid36.74 ± 1.26a8.19 ± 0.07d7.27 ± 0.26d25.83 ± 1.54bNDND23.28 ± 1.9cValeric acidNDNDNDNDND135.79 ± 3.03NDCaproic acid98.27 ± 5.26d215.51 ± 6.27bND188.71 ± 11.13c471.57 ± 3.34a85.96 ± 4.75d174.59 ± 17.01cHeptanoic acidND1.53 ± 0.13d3.13 ± 0.14b2.5 ± 0.22cNDND5.58 ± 0.3aCaprylic acid349.19 ± 20.32c101.65 ± 0.95d100.55 ± 4.41d58.61 ± 3.17d732.06 ± 14.71b913.19 ± 24.4a897.52 ± 37.57aNonanoic acid31.35 ± 0.74b23.04 ± 2.26c16.54 ± 1.33d22.87 ± 2.27c58.57 ± 2.53a23.5 ± 1.92c31.96 ± 1.15bDecanoic acid73.54 ± 0.91d23.74 ± 2.17e31.50 ± 1.96e33.73 ± 0.98e741.08 ± 31.98a522.05 ± 32.56b190.15 ± 8.25cBenzoic acid12.72 ± 1.1b8.47 ± 0.54c5.56 ± 0.29dND21.58 ± 1.62aND6.41 ± 0.42dLauric acid8 ± 0.33NDNDNDNDNDNDTetradecanoic acidND3.07 ± 0.15cND22.47 ± 1.78b26.98 ± 2.55aNDND4-Ethylphenol16.28 ± 0.8d16.03 ± 1.49d32.91 ± 1.73c181.85 ± 6.8a109.23 ± 2.98bNDND Olefins3-Decene11.08 ± 0.94NDNDNDNDNDNDStyrene16.39 ± 0.58b5.82 ± 0.56cND26.97 ± 1.39aNDNDND Alkanes4-DecaneNDNDND23.57 ± 1.74NDNDNDUndecaneNDNDND5.57 ± 0.31b17.28 ± 0.77a2.81 ± 0.16c0.76 ± 0.04dHexadecaneNDND7.32 ± 0.4a1.93 ± 0.04bNDND1.69 ± 0.07bNote: values with distinct lowercase characters within the same row were statistically significant (p < 0.05). All findings are shown as the mean ± standard deviation (n = 3). ND, not detected.Fig. 2Clustered heatmap of volatile compounds across seven rice wine samples.Fig. 2 Table 5Volatile compound composition and content in fermented rice wines (μg/L).Flavor componentsMJ1MJ2MJ3MJ4MJ5MJ6MJ7Alcohols1-Propanol22.03 ± 1.33e9.78 ± 0.16f38.1 ± 1.15c38.87 ± 0.46c35.65 ± 0.48d64.65 ± 2.36a45.03 ± 2.07bIsobutanol654.39 ± 11.01a364.24 ± 27.15e403.17 ± 31.2d299.14 ± 19.91f524.7 ± 19.6b460.92 ± 19.55c557.97 ± 9.51b1-Butanol11.4 ± 0.27c6.82 ± 0.5d7.68 ± 0.41d7.41 ± 0.34d25.6 ± 1.69a12.49 ± 0.85c14.07 ± 1.1bIsoamyl alcohol4731.32 ± 361.83a2139.91 ± 69.64d2323.63 ± 146.19d2513.79 ± 103.79d4993.27 ± 244.58a4287.02 ± 354.1b3833.57 ± 254.75cHexanolND123.62 ± 6.62b119.61 ± 11.34b148.17 ± 2.06aNDNDNDHeptanol13.82 ± 0.67c20.98 ± 1.63b19.7 ± 0.87b21.94 ± 0.55bND14.75 ± 1.02c32.37 ± 1.99a2-Ethylhexanol50.06 ± 1.77b10.84 ± 0.51d16.63 ± 0.97c14.45 ± 0.81c55.62 ± 1.05a55.41 ± 3.01a11.33 ± 0.5d2-Nonanol4.58 ± 0.15c6.04 ± 0.57b8.87 ± 0.45aNDNDND9.1 ± 0.68a2,3-Butanediol220.99 ± 2.74bND51.77 ± 0.58f145.71 ± 3.66c373.69 ± 4.85a98.86 ± 9.64e109.11 ± 7.53dLinalool5.3 ± 0.34cND15.77 ± 0.89b4.33 ± 0.14c227.96 ± 19.32a17.14 ± 1.61b10.65 ± 0.86bOctanol61.02 ± 4.11b44.31 ± 1.84d52.78 ± 3.26c95.75 ± 2.28a49.96 ± 1.85c23.09 ± 1.89f33 ± 2.04e2-Octen-1-ol4.52 ± 0.26aNDND2.48 ± 0.09aNDNDNDNonanol16.55 ± 0.51c15.08 ± 1.14d13.31 ± 0.71e24.93 ± 0.45a15.82 ± 0.37 cd22.71 ± 1.12b16.17 ± 0.21 cdTerpineol7.51 ± 0.3cNDND5.55 ± 0.36c350.53 ± 6.04aND16.76 ± 0.84bMethylthiopropanol11.44 ± 1.12cNDND8.26 ± 0.46c41.22 ± 3.34a14.45 ± 1.04b11.22 ± 0.73cDecanol3.3 ± 0.27c14.94 ± 0.5b95.08 ± 7.89a14.61 ± 0.39bND12.41 ± 0.98b16.52 ± 1.29bUndecanolND7.79 ± 0.36b64.24 ± 4.1a6.46 ± 0.21bNDNDNDCitronellolNDNDNDND9.47 ± 0.1343.68 ± 2.22NDBenzyl alcohol16.68 ± 1.54b13.2 ± 1.2c19.47 ± 0.59a6.46 ± 0.53d12.4 ± 0.5c5.62 ± 0.49d19.68 ± 1.21aβ-Phenylethanol3592.63 ± 150.36d2153.98 ± 146.21f2495.68 ± 217.11f2629.66 ± 63.29e8456.95 ± 307.84a5758.81 ± 233.54b4354.49 ± 166.35cLauryl alcohol84.6 ± 1.66c185.9 ± 12.12b1147.81 ± 104.76a72.5 ± 6.49c1142.32 ± 18.8a35.87 ± 2.12d15.24 ± 0.35dIonol3.59 ± 0.18aNDNDND129.57 ± 5.53NDNDHexadecanolND33.11 ± 2.39a18.11 ± 1.54bND7.16 ± 0.16cNDNDNerolidolND7.8 ± 0.49a5.86 ± 0.43bNDNDNDNDα-CadinolNDND1.38 ± 0.06c5.39 ± 0.34b34.79 ± 1.79aNDNDSpathulenol27.59 ± 2.03c34.43 ± 3.04b32.65 ± 1.89b54.66 ± 4.56aNDND5.41 ± 0.14dα-SantalolND1.18 ± 0.06c11.66 ± 0.63b13.33 ± 0.53aNDNDNDα-Bisabolol8.82 ± 0.12NDNDNDNDNDNDEstersNDNDNDNDNDNDNDEthyl acetate204.95 ± 3.77c46.77 ± 2.03e66.46 ± 4.4e117.82 ± 0.78e1619.27 ± 148.33a459.97 ± 35.58b252.82 ± 6.73cEthyl isobutyrate1.76 ± 0.04NDNDNDNDNDNDIsoamyl acetate139.03 ± 6.81b55.48 ± 4.94c34.18 ± 2.82d62.67 ± 5.28c139.02 ± 4.73b160.54 ± 8.65b295.48 ± 28.57aEthyl caproate26.25 ± 1.05e48.45 ± 3.65dND160.92 ± 7.36b275.64 ± 8.79a135.17 ± 9.26c127.41 ± 8.48cEthyl heptanoateNDNDND4.44 ± 0.43NDNDNDEthyl pyruvateNDNDNDND18.4 ± 1.08a17.57 ± 0.46aNDEthyl lactate216.04 ± 0.77cNDNDND750.25 ± 11.94a317.84 ± 11.02b142.29 ± 1.07dEthyl caprylate6.23 ± 0.19e3.51 ± 0.33eND82.02 ± 5.52d95.55 ± 8.82c477.35 ± 1.42a257.71 ± 4.22bEthyl leucate57.2 ± 2.72b15.15 ± 1.11c12.08 ± 1.08 cd3.88 ± 0.08e197.74 ± 6.66a8.92 ± 0.71de6.21 ± 0.15eHeptyl formateNDNDNDND1.54 ± 0.05a12.78 ± 0.8bNDIsobutyl lactateND2.83 ± 0.27c3.21 ± 0.23c5.35 ± 0.22b3.08 ± 0.19c12.03 ± 1.07a5.07 ± 0.14bAmyl hexanoateNDNDND13.85 ± 0.79a23.1 ± 0.61bNDNDEthyl 3-hydroxybutyrate5.84 ± 0.3NDNDNDNDNDNDEthyl nonanoate6.73 ± 0.61aNDNDNDNDND2.29 ± 0.17bγ-Butyrolactone9.47 ± 0.34e28.48 ± 2.46c32.87 ± 2.58b37.58 ± 3.67aND13.88 ± 0.69dNDEthyl decanoate22.18 ± 1.88cNDND78.76 ± 5.96b82.62 ± 6.07b293.91 ± 12.52a5.49 ± 0.25dDiethyl succinate757.02 ± 32.27b128.71 ± 6.01f228.75 ± 0.79de193.97 ± 2.89e1407.65 ± 50.61a426.16 ± 19.55c244.44 ± 15.44dEthyl benzoateND9.07 ± 0.83dND12.47 ± 0.74c26.01 ± 2.6a15.15 ± 0.74b1.42 ± 0.09eEthyl glutarateNDNDND1.56 ± 0.14a17.68 ± 1.01bNDNDEthyl phenylacetate22.01 ± 1.23c4.42 ± 0.39e5.94 ± 0.23e15.2 ± 0.46d168.37 ± 3.16a33.62 ± 2.05b24.41 ± 1.39cPhenethyl acetate458.03 ± 3.99d176.13 ± 10.96f123.25 ± 4.98 g326.94 ± 11.11e1066.68 ± 28.51a723.84 ± 27.31b628.9 ± 20.51cEthyl laurateNDNDNDNDND62.75 ± 5.9NDEthyl phenylpropionateND0.94 ± 0.04aNDND50.5 ± 1.46bNDNDPhenethyl isobutyrateNDNDNDND15.19 ± 0.19a3.99 ± 0.39b2.55 ± 0.15cDiethyl pimelate6.02 ± 0.52b1.82 ± 0.1 cd0.96 ± 0.04d2.32 ± 0.07c20.62 ± 1.11aNDNDEthyl myristate134.05 ± 5.77b11.75 ± 1.05d8.44 ± 0.72d163.38 ± 9.73aND76.96 ± 7.58c3.91 ± 0.35dγ-Octanolactone3.29 ± 0.31cNDNDND10.26 ± 0.91aND4.83 ± 0.46bDiethyl suberateNDNDND5.73 ± 0.51NDNDNDEthyl pentadecanoateNDNDND2.66 ± 0.12NDNDNDγ-Decanolactone9.37 ± 0.46aNDNDND154.94 ± 3.6bNDNDDiethyl azelateNDNDND15.52 ± 1.28b28.85 ± 1.24aND2.78 ± 0.16cEthyl palmitate12.52 ± 0.43d18.34 ± 1.22d35.81 ± 2.95 cd621.93 ± 45.63a53.99 ± 5.4c399.18 ± 8.86b15.98 ± 0.5dEthyl palmitoleateNDNDND8.67 ± 0.72NDNDNDHexyl acetateNDND1.91 ± 0.05a7.59 ± 0.38bNDNDNDMethyl stearate7.45 ± 0.39b4.95 ± 0.29bc2.58 ± 0.24cND52.94 ± 3.8aND5.83 ± 0.43bcEthyl stearate64.1 ± 3.08c6.41 ± 0.41e5.61 ± 0.46e12.87 ± 0.21d282.08 ± 3.48a113.64 ± 6.35b65.16 ± 5.07cEthyl oleateND3.41 ± 0.1c4.92 ± 0.37c167.7 ± 8.75a25.54 ± 1.36d171.75 ± 9.23a10.53 ± 0.8cEthyl linoleateND5.49 ± 0.27c3.75 ± 0.29c221.74 ± 2.46aND158.36 ± 10.13b5.74 ± 0.45cEthyl linolenateNDNDND7.64 ± 0.61NDNDND Ketones2-Octanone92.31 ± 6.49b86.67 ± 4.76b89.77 ± 6.87b108.52 ± 4.25a75.7 ± 6.74c69.36 ± 3.9c97.54 ± 6.53bAcetoinNDNDNDND117.8 ± 10.59a583.47 ± 16.37bND2-NonanoneND9.44 ± 0.74NDNDNDNDND3-Nonen-2-oneND5.4 ± 0.5a5.09 ± 0.35a2.97 ± 0.11bNDNDNDAcetophenoneNDNDND5.94 ± 0.21aND2.11 ± 0.2c4.9 ± 0.24b2,3-Butanedione14.85 ± 0.17a3.95 ± 0.26cND9.82 ± 0.53bNDNDNDPhenylacetoneNDNDNDND15.57 ± 0.55b5.19 ± 0.21NDβ-Ionone73.62 ± 3.68aNDNDND973.66 ± 19.92bNDNDGeranyl acetone31.94 ± 0.16b9.96 ± 0.9eND14.25 ± 0.84d50.28 ± 3.71a20.16 ± 1.94c32.93 ± 3.07bα-Ionone69.11 ± 5.25aNDNDND1542.16 ± 45.7bNDND AldehydesNonanal13.6 ± 0.42d21.07 ± 1.49c15.86 ± 1.26d32.74 ± 1.79a33.75 ± 2.49a33.17 ± 1.55a26.63 ± 0.87b2-OctenalND13.41 ± 0.74a2.77 ± 0.25b2.97 ± 0.11bNDNDNDFurfural5.67 ± 0.1c5.12 ± 0.44c4.52 ± 0.03c5.97 ± 0.49c182.96 ± 6.74a31.37 ± 1.68bNDBenzaldehydeND74.99 ± 7.41b55.29 ± 3.16c68.94 ± 5.87cND268.03 ± 12.66aND2,4-NonadienalND10.18 ± 0.5a7.98 ± 0.79b5.81 ± 0.21cNDNDND2,4-DecadienalND6.17 ± 0.5a2.21 ± 0.16b6.9 ± 0.68cNDNDNDVeratraldehydeND736.26 ± 63.06NDNDNDNDNDCuminaldehyde7.29 ± 0.37bc6.47 ± 0.54c5 ± 0.14d8.61 ± 0.15a3.51 ± 0.13e8.06 ± 0.79ab8.4 ± 0.75a5-Hydroxymethylfurfural7.92 ± 0.12aNDNDND51.79 ± 5.08bNDND AcidsAcetic acid64.22 ± 2.12c48.52 ± 1.32d32.76 ± 2.88e53.82 ± 3.49d317.33 ± 22.91a156.96 ± 11.02b68.24 ± 4.12cIsobutyric acid26.87 ± 1.66c22 ± 2d24.66 ± 0.94c12.82 ± 0.86e78.65 ± 7.09a20.31 ± 1.4d33.09 ± 1.59bn-Butyric acid4.59 ± 0.4c3.9 ± 0.34d3.55 ± 0.22e5.12 ± 0.4c11.75 ± 1b13.57 ± 0.58a4.76 ± 0.39c2,4-Dimethylvaleric acid36.74 ± 1.26a8.19 ± 0.07d7.27 ± 0.26d25.83 ± 1.54bNDND23.28 ± 1.9cValeric acidNDNDNDNDND135.79 ± 3.03NDCaproic acid98.27 ± 5.26d215.51 ± 6.27bND188.71 ± 11.13c471.57 ± 3.34a85.96 ± 4.75d174.59 ± 17.01cHeptanoic acidND1.53 ± 0.13d3.13 ± 0.14b2.5 ± 0.22cNDND5.58 ± 0.3aCaprylic acid349.19 ± 20.32c101.65 ± 0.95d100.55 ± 4.41d58.61 ± 3.17d732.06 ± 14.71b913.19 ± 24.4a897.52 ± 37.57aNonanoic acid31.35 ± 0.74b23.04 ± 2.26c16.54 ± 1.33d22.87 ± 2.27c58.57 ± 2.53a23.5 ± 1.92c31.96 ± 1.15bDecanoic acid73.54 ± 0.91d23.74 ± 2.17e31.50 ± 1.96e33.73 ± 0.98e741.08 ± 31.98a522.05 ± 32.56b190.15 ± 8.25cBenzoic acid12.72 ± 1.1b8.47 ± 0.54c5.56 ± 0.29dND21.58 ± 1.62aND6.41 ± 0.42dLauric acid8 ± 0.33NDNDNDNDNDNDTetradecanoic acidND3.07 ± 0.15cND22.47 ± 1.78b26.98 ± 2.55aNDND4-Ethylphenol16.28 ± 0.8d16.03 ± 1.49d32.91 ± 1.73c181.85 ± 6.8a109.23 ± 2.98bNDND Olefins3-Decene11.08 ± 0.94NDNDNDNDNDNDStyrene16.39 ± 0.58b5.82 ± 0.56cND26.97 ± 1.39aNDNDND Alkanes4-DecaneNDNDND23.57 ± 1.74NDNDNDUndecaneNDNDND5.57 ± 0.31b17.28 ± 0.77a2.81 ± 0.16c0.76 ± 0.04dHexadecaneNDND7.32 ± 0.4a1.93 ± 0.04bNDND1.69 ± 0.07bNote: values with distinct lowercase characters within the same row were statistically significant (p < 0.05). All findings are shown as the mean ± standard deviation (n = 3). ND, not detected. Volatile compound composition and content in fermented rice wines (μg/L). Note: values with distinct lowercase characters within the same row were statistically significant (p < 0.05). All findings are shown as the mean ± standard deviation (n = 3). ND, not detected. Clustered heatmap of volatile compounds across seven rice wine samples. In contrast, MJ2 and MJ3 clustered closely together, showing higher relative abundances of certain aldehydes (e.g., benzaldehyde, veratraldehyde, and 2,4-nonadienal) and medium-chain fatty acids, while MJ1, MJ4, and MJ7 occupied intermediate positions with mixed profiles (Fig. 2). These distribution patterns corroborate the sensory evaluation results, confirming that ester- and alcohol-rich volatile profiles are the primary drivers of superior sensory quality, whereas samples with lower volatile diversity and abundance exhibited plainer sensory characteristics. Higher alcohols contribute to both pleasant (fruity, floral) and harsh (alcoholic) notes. β-Phenylethanol, which imparts a rose-like fragrance, was exceptionally high in MJ5 (8456.95 μg/L) and MJ6 (5758.81 μg/L), which correlates strongly with their high sensory scores (93.6 and 91.4, respectively). This compound is a key contributor to the desirable floral aroma in rice wine (Cui et al., 2024; Yu et al., 2019). The presence of linalool (227.96 μg/L) and terpineol (350.53 μg/L) only in MJ5 provides a unique floral and citrus dimension, likely contributing to its top sensory rating. In contrast, MJ2 and MJ3 showed much lower levels of these beneficial alcohols. Isoamyl alcohol, which can be malty or harsh depending on concentration, was highest in MJ5 (4993.27 μg/L) and MJ1 (4731.32 μg/L). While high in MJ5, its effect was likely balanced by the high ester content, whereas in MJ1, it may have contributed to a more pungent aroma, explaining its moderate sensory score (83.4) despite relatively high alcohol (Chen et al., 2013). Esters are the primary contributors to fruity aromas and are highly desired in rice wine (Yang, Li, et al., 2024). MJ5 and MJ6 again stood out. Ethyl acetate, which provides a pineapple-like note, was dramatically higher in MJ5 (1619.27 μg/L) compared to other samples (e.g., 204.95 μg/L in MJ1). Furthermore, phenethyl acetate (honey, rose) was also highest in MJ5 (1066.68 μg/L), which, together with β-phenylethanol, explains the intense floral and fruity character of this sample. Diethyl succinate, which imparts a fruity and wine-like aroma, was most abundant in MJ5 (1407.65 μg/L) and MJ1 (757.02 μg/L). These results are consistent with studies showing that co-fermentation or specific starter cultures can significantly boost ester production (Cui et al., 2024; He et al., 2025). The presence of unique esters like γ-decanolactone in MJ5 (154.94 μg/L) and MJ1 (9.37 μg/L) adds a peachy, coconut-like nuance, further enhancing flavor complexity. While volatile acids can contribute to rancid notes, they also serve as precursors for esters. Acetic acid was highest in MJ5 (317.33 μg/L), which, if present in excess, could be a defect. However, its high concentration in MJ5 was apparently balanced by the high ester and alcohol content, resulting in a well-integrated aroma. Caprylic acid, which has a slightly oily and rancid note, was exceptionally high in MJ6 (913.19 μg/L) and MJ7 (897.52 μg/L). This may explain why, despite having high FAAs and some beneficial alcohols, MJ7's sensory score (88.9) did not surpass MJ5's, as high levels of certain fatty acids can be perceived as off-flavors (Shen et al., 2024). The presence of β-ionone (973.66 μg/L) and α-ionone (1542.16 μg/L), which have violet and woody aromas, exclusively in MJ5 provides a unique and highly desirable floral complexity, solidifying its position as the highest-quality sample in this study (Chen et al., 2018).The volatile profiles clearly distinguished the seven samples in terms of chemical diversity and abundance, with the highest ester and higher alcohol contents observed in MJ5 and MJ6, aligning with their superior sensory ratings, while MJ2 and MJ3 exhibited notably lower volatile complexity (Table 5). Implications for quality standardization The integrated analysis presented in this study offers a practical roadmap for establishing comprehensive quality standards for fermented rice wines. First, we propose that a core set of mandatory indicators should include alcohol content (reflecting fermentation completeness), total acidity (influencing microbial stability and taste), amino acid nitrogen (contributing to flavor and nutrition), key esters (e.g., ethyl acetate and phenethyl acetate, which drive consumer preference), and total biogenic amines (ensuring safety). Second, our data suggest a provisional safety threshold of <100 mg/L for total biogenic amines, as samples exceeding this level (e.g., MJ6 at 132.54 mg/L) raised legitimate safety concerns, while the highest-quality sample (MJ5) remained below this limit. Third, sensory-chemical correlations enable quality grading: premium wines (> 90 points) are characterized by high ester contents (> 2000 μg/L) and elevated umami amino acids, whereas plain wines (< 85 points) exhibit low volatile diversity and limited amino acid complexity. Fourth, regional benchmarking is essential, as local raw materials and starter cultures produce distinct profiles; standards should therefore be adaptable while maintaining core safety and sensory criteria. Finally, our findings guide process optimization—promoting ester formation through selected starter cultures and fermentation conditions, while controlling biogenic amine accumulation via microbial management. These recommendations provide an empirical foundation that can support future regulatory efforts and industrial quality improvement. Limitations and future perspectives It should be noted that although a uniform storage protocol was applied to all samples (−18 °C before analysis and 4 °C after opening, with measurements completed within 7 days), the stability of individual quality indicators under these conditions was not experimentally verified. Freezing does not fully inactivate enzymes (e.g., lipases, esterases, and proteases) or suppress psychrotolerant microbial activity, which may theoretically cause subtle changes over time, including ester hydrolysis, organic acid interconversion, or decarboxylation of amino acids to biogenic amines (Luo et al., 2024; Yuan et al., 2025). Moreover, volatile compounds with low boiling points or high hydrophobicity (e.g., ethyl acetate and isoamyl alcohol) may partition into the headspace and be lost during thawing or repeated pipetting (Wang et al., 2013). Consequently, the absolute concentrations in Table 5 may not fully reflect the original fresh-state profiles. Nevertheless, since all samples were subjected to identical storage duration, temperature, and handling, the relative comparisons and quality rankings remain robust; any storage-induced deviations would have affected all groups uniformly. Future studies should include systematic stability tests under various storage conditions to validate absolute quantitation. Additionally, this study analyzed only a single batch per manufacturer and did not assess batch-to-batch reproducibility. The findings thus represent a snapshot of product quality at the time of collection (March 2025), rather than consistency across production cycles. Traditional rice wine fermentation, conducted in open or semi-open systems, is inherently variable due to differences in raw materials, starter cultures, fermentation conditions, and post-fermentation storage. Moreover, as noted in Section 2.1, specific production parameters were not disclosed for proprietary reasons, precluding a systematic evaluation of their influence on reproducibility. Future work should incorporate multiple batches from the same manufacturer to identify critical control points affecting consistency. The multi-parameter analytical framework established here—covering physicochemical indices, amino acids, organic acids, biogenic amines, volatiles, and sensory evaluation—provides a standardized platform for such investigations. Conclusion This study confirms that the quality of fermented rice wine is not governed by any single physicochemical or flavor parameter, but rather by the intricate synergy among volatile compounds (particularly esters and higher alcohols), non-volatile taste components (amino and organic acids), and the critical safety index of biogenic amines. Among the seven commercial samples evaluated, MJ5 (Guangxi) emerged as the benchmark for optimal flavor balance, whereas MJ6 (Hubei) exemplified the challenge of managing flavor complexity alongside BA safety risks, and MJ2 (Chongqing) illustrated the trade-off between safety and sensory richness. These findings underscore the necessity for a comprehensive, multi-target quality assurance framework for the Chinese rice wine industry. Future production improvements should prioritize the selection of starter cultures and fermentation conditions that enhance desirable ester synthesis and controlled proteolysis while actively suppressing the decarboxylase activity of spoilage microorganisms to limit BA accumulation. This study provides a scientific reference for standardizing regional rice wine quality and fostering consumer confidence in traditional fermented beverages. Studies in human This study was performed in compliance with relevant laws, regulatory frameworks and guidelines where the research took place. The Science and Technology Ethics Committee of Yibin University granted an exemption. The authors provided the following explanation: It does not involve any clinical intervention, collection of personal health information, or experimentation on human subjects. CRediT authorship contribution statement HuaWei Yuan: Writing – original draft, Software, Methodology, Investigation, Funding acquisition, Data curation, Conceptualization. XinYu Zhao: Validation, Investigation. Hao Yuan: Resources, Methodology. ZiYing Tian: Methodology, Investigation. Zhou Xu: Investigation, Funding acquisition. TingNa Xie: Validation, Funding acquisition, Formal analysis. Kai Lou: Writing – review & editing, Supervision, Project administration, Conceptualization. Informed consent and patient details Written informed consent to take part in the study and to publish the article has been obtained from all participants or their legal representatives. The privacy rights of participants have been observed. Ethics statement Prior to the sensory tests, panelists from Yibin University (Yibin, China) were fully informed of the test details, requirements, and risks, and signed informed consent forms. The study was authorized by the Science and Technology Ethics Committee of Yibin University for sensory evaluation. Participants' rights and privacy were protected through voluntary participation, full disclosure, informed consent, data confidentiality, and the right to withdraw at any time. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.