Pottery jars are key for Baijiu aging, with their microstructure and macroscopic properties regulating liquor quality. This study investigated three origin-specific pottery jars (P1, P2, P3) for soy sauce-flavored Baijiu, analyzing their corrosion, mechanical and adsorption properties via multiple characterization and test methods. A novel "diffusion-complexation-crystallization-growth" cycle was revealed: iron ions leached from jars form iron oxide/salt crystals in Baijiu, modulating flavor evolution. The jars showed distinct superiorities: P3 had the optimal corrosion resistance (Fe leaching: 0.307 mg/L, P1 the poorest with 6.9 mg maximum mass loss); P2 possessed the largest specific surface area (0.323 m(2)/g), presenting the highest adsorption capacity (25.7 mg/g) and best mechanical stability (8.22% strength loss after corrosion); P1 selectively adsorbed high-boiling-point alcohols and ketones (2-undecanone residue: 10.71 mg/L). All jars followed the pseudo-second-order adsorption model (R-2 > 0.999),adsorption involves a rapid surface diffusion phase and a slow intraparticle diffusion phase; the overall adsorption rate is primarily governed by surface diffusion (liquid film diffusion) and had unique metal ion leaching fingerprints. This study establishes the structure-property-function relationship of pottery jars, providing a theoretical basis for their scientific selection in Baijiu aging.

ABSTRACT Solid‐state fermentation (SSF) is central to Luzhou‐flavor Baijiu production, yet the temperature evolution inside pits and its coupling with substrate conversion remain poorly understood. We combined in situ temperature sensing, offline physicochemical sampling, and mechanism‐based modeling to analyze production‐scale pits. A multi‐point system (36 sensors) was deployed in one winter pit and two summer pits. Samples from the center, middle, and bottom layers were collected every 5 days and analyzed for starch, reducing sugar, moisture, acidity, and ethanol. Cross‐season results showed a reproducible rise‐then‐fall temperature trajectory and a parabolic spatial profile (center high, edges low; R 2 > 0.90, reaching a maximum of 0.996). In winter, the core temperature rose from 20°C to 30°C, with a 5°C difference between the bottom and middle layers. In summer, the two pits reached maximum central temperatures of 37°C and 34°C, with maximum spatial temperature differences of 15°C and 12°C. Physicochemical indices revealed rapid starch consumption, transient accumulation then depletion of reducing sugar, a moisture increase of about 6%, acidity rising from 2.2 to 5.0 mmol/10 g, and ethanol accumulation to about 35 µg/g. Mechanistic analysis indicated that starch hydrolysis provides sugars, which drive microbial heat generation; temperature promotes growth before the thermal maximum, while acidity accumulation causes irreversible inhibition later. Governing equations were formulated: The heat production rate is linearly driven by available reducing sugar, modulated by an inverted parabolic temperature function, and subjected to cumulative acidity inhibition. This work offers a quantitative basis for online monitoring, early warning, and bioreactor modeling of Baijiu SSF.

This study employed GC–MS, sensory evaluation, and odor activity value (OAV) analysis combined with machine learning to differentiate the volatile flavor profiles of fermented green plum wines. Most green plum wines contained high levels of esters, with ethyl benzoate (up to 4820.53 μg/L), ethyl caprylate (up to 2640.83 μg/L), and benzaldehyde (up to 3432.96 μg/L) being the major contributors. Among six compared algorithms, fuzzy c-means clustering performed best, distinguishing three distinct flavor categories, while a decision tree model achieved 95.13% classification accuracy. SHAP analysis identified ethyl caprylate, benzyl acetate, and phenethyl acetate as the most influential volatile compounds. These results demonstrate that machine learning provides an effective approach for flavor classification and interpretation in fruit wines.

This study investigates the release behavior of ferric oxide (Fe₂O₃) in Baijiu and its regulation of flavor evolution. Kinetics analysis indicated the leaching process follows the shrinking core model (Ea = 44.52 kJ·mol-1), controlled by surface chemical reactions. Fe₂O₃ particles undergo surface erosion and amorphization, eventually recrystallizing into a FeOOH shell to form a core-shell composite structure. Electrochemical tests confirmed that this surface reconstruction alters interfacial electron transfer properties. The leaching of Fe₂O₃ significantly catalyzed pyrazine formation, resulting in a twofold increase in tetramethylpyrazine at 25 °C and a 107% rise in 2,3-dimethyl-5-isopentylpyrazine at 65 °C. The total ester content exhibited a 42.0% decrease under accelerated aging at 65 °C, compared to a 28.4% reduction at 25 °C. The migration and surface reconstruction of Fe₂O₃ facilitate the catalytic evolution of Baijiu compounds, driving the evolution and reshaping of its theoretical aroma profile based on OAV assessment.

To address production challenges in tea processing, freeze-withering was introduced for yellow tea. This study conducted a comprehensive assessment of the sensory quality, volatile compounds and metabolomics characteristics of freeze-withering yellow tea (FYT) and unfrozen withered yellow tea (UYT). The results indicated that the FYT samples contained higher levels of volatile compounds, including (E,E)-3,5-octadien-2-one, 1-octanol, linalool, beta-myrcene, 2,6-dimethyl-3,7-octadiene-2,6-diol, 2-pentylfuran, and phenylethyl alcohol, compared to UYT. Consequently, FYT exhibited more pronounced sweet, woody, and baked aromas. A total of 15 aroma compounds from the two tea samples were identified as key components and integrated into the flavor wheel model. The enhancement of umami, sweetness, and richness in yellow tea was attributed to increased levels of metabolites, including p-Coumaroyl quinic acid, D-Gluconic Acid, Theaflavin 3,3 '-digallate, Theaflavin-3-gallate, Theaflavin, and 9S,12R,13S-Trihydroxy-10E,15Z-octadecadienoic acid. Conversely, the reduction in bitterness was linked to decreased levels of compounds such as Epigallocatechin 3,4 '-di-O-gallate, (-)-gallocatechin gallate, (-)-catechin gallate, and Procyanidin B3. In addition, after freeze-withering, the content of theaflavins in yellow tea increased significantly. In conclusion, the freeze-withering process is demonstrated to enhance the quality of yellow tea, and it provides a one-month buffer period for tea processing, thereby providing a theoretical basis for alleviating the processing pressure of tea in spring.

The physicochemical properties and volatile composition of fruits are critical determinants of fruit quality and processing performance. This study evaluated major green plum cultivars from Sichuan and Yunnan Provinces by analyzing fruit morphology, nutritional composition, bioactive compounds, and volatile profiles. Multivariate statistical analyses, including orthogonal partial least squares discriminant analysis (OPLS-DA), principal component analysis (PCA), and cluster analysis (CA), were applied to comprehensively assess cultivar-dependent quality differences. EH exhibited the highest total acid and glucose contents, whereas MD showed superior soluble solids, total sugars, solid-acid ratio, and several organic acids and sugars. Yunnan cultivars generally showed higher flavonoid contents and stronger antioxidant activities than Sichuan cultivars. Citric acid was the predominant organic acid. A total of 97 volatile compounds were identified. Ten volatile compounds were detected in all eight varieties, including butyl acetate, hexyl acetate, and butyl butyrate. EH and MD released the higher volatile, and PCA-based comprehensive evaluation ranked the cultivars as follows: EH, MD, EY, YZ, PX, EZ, DY and DN. Therefore, EH and MD exhibited superior overall quality in physicochemical properties and volatile composition. These findings provide a theoretical basis for evaluating green plum quality and their rational utilization in production and processing.

The study examined the effects of oxygen,light,temperature,non-covalent bonds,and metal ions on the precipitation amount,as well as the protein,starch,pectin,and tannin content in the supernatant of Sichuan bran vinegar,to determine the key factors influencing turbidity reversion.Using transmittance as the response value,the clarification conditions were optimized via single-factor experiments combined with response surface methodology(RSM).The results indicated that oxygen and temperature were critical factors promoting turbidity reversion,with significantly shorter turbidity reversion time under aerobic conditions compared to anaerobic conditions,and higher temperature further accelerating the process.Investigations into non-covalent interactions revealed that the addition of NaCl and guanidine hydrochloride disrupted electrostatic forces and hydrogen bonds,significantly reducing precipitation while increasing the protein,starch,pectin,and tannin content in the supernatant.This suggested that electrostatic interactions and hydrogen bonds facilitated macromolecular complexation and subsequent precipitation.Metal ions(Fe3+,Ca2+)had no significant effect on turbidity reversion.The optimal clarification conditions determined by single-factor and RSM optimization were:acidic protease dosage of 2.10 g/L,treatment time of 2.20 h,treatment temperature of 30 ℃,and standing time of 2 h,achieving a transmittance of 85.75％±0.24％.This study provides a data foundation for understanding the mechanism of turbidity reversion and offers practical solutions for industrial production of Sichuan bran vinegar.

This study systematically investigated differences in volatile flavor profiles among fermented green plum wines by integrating gas chromatography–mass spectrometry (GC–MS), sensory evaluation, and odor activity value (OAV) analysis with machine learning and SHapley Additive exPlanations (SHAP) based feature interpretation. The primary objective was to evaluate the applicability of machine learning algorithms for flavor profiling of green plum wine. The results indicated that floral and fruity aromas were predominant in samples NG9, YM7, and YM9. Most green plum wines contained high levels of esters, with ethyl benzoate (up to 4820.53 μg/L), ethyl octanoate (up to 2640.83 μg/L), and benzenecarbaldehyde (up to 3432.96 μg/L) being the major contributors. Among the six classification algorithms compared, fuzzy c-means clustering provided the most distinct clustering structure, identifying three distinct flavor categories. Six machine learning models were subsequently established, of which the decision tree (DT) model exhibited the highest performance, with an accuracy of 95.13%. SHAP analysis further revealed that ethyl octanoate, benzyl ethanoate, and 2-phenylethyl ethanoate exerted the greatest influence on model predictions. Overall, these findings highlight the effectiveness of machine learning as a robust tool for the classification and interpretation of flavor characteristics in fermented fruit wines, with broad applicability in flavor science.

Daqu, the essential fermentation starter for Baijiu production, harbors stratified microbial communities that orchestrate flavor formation. This study integrated headspace solid-phase microextraction gas chromatography-mass spectrometry (HS-SPME-GC-MS), metagenomics, and PacBio full-length 16S rRNA and ITS sequencing to characterize the spatiotemporal heterogeneity of volatile flavor compounds (VFCs) and functional microbial communities across three layers—the surface (QP), middle (HQ), and core (QX) of strong-flavor Daqu over a 90-day fermentation period. A total of 47 VFCs were identified, with alcohols, aldehydes, and esters as the predominant components. VFCs peaked at 2-4 days, then declined and stabilized, showing a progressive accumulation gradient from QP to QX. Aldehydes and pyrazines were enriched in HQ and QX, while esters and acids were highest in QX. Metagenomic analysis demonstrated that the abundance of flavor-related enzyme genes followed spatiotemporal trends of VFCs. Although filamentous fungi and Lactobacillaceae constituted the primary microbial sources of flavor-related enzyme genes across all layers, their community composition and successional dynamics exhibited pronounced spatiotemporal specificity. Positive microbial interactions increased from QP (71.70%) to HQ (76.73%) to QX (89.95%), indicating stronger cooperation in QX. These findings provide mechanistic insights into the ecological underpinnings of flavor stratification in Daqu and offer a scientific framework for future targeted optimization of strong-flavor Baijiu fermentation.

Pit mud (PM) hosts diverse microbial communities, which serve as a medium to impart flavor and quality to Baijiu and exhibit long-term tolerance to ethanol and acids, resulting in a unique ecosystem. However, the ecology and metabolic functions of PM remain poorly understood, as many taxa in PM represent largely novel lineages. In this study, we used a combination of metagenomic analysis and chemical derivatization LC–MS analysis to provide a comprehensive overview of microbial community structure, metabolic function, phylogeny, horizontal gene transfer, and the relationship with carboxyl compounds in spatiotemporal PM samples. Our findings revealed three distinct stages in the spatiotemporal changes of prokaryotic communities in PM: an initial phase dominated by Lactobacillus, a transitional phase, and a final state of equilibrium. Significant variations in α- and β-diversity were observed across different spatial and temporal PM samples. We identified 178 medium- and high-quality non-redundant metagenome-assembled genomes (MAGs), and constructed their phylogenetic tree, depicting their roles in the carbon, nitrogen, and sulfur cycles. The Wood-Ljungdahl pathway and reverse TCA cycle were identified as the main carbon fixation mechanisms, with both hydrogenotrophic and aceticlastic methanogens playing a major role in methane production, and methylotrophic pathway observed in older PM. Furthermore, we identified relationships between prokaryotes and 29 carboxyl metabolites, including medium- and long-chain fatty acids. Horizontal gene transfer (HGT) was widespread in PM, particularly among clostridia, Bacteroidota, Bacilli, and Euryarchaeota, and was shown to play critical roles in fermentation dynamics, carbon fixation, methane production, and nitrogen and sulfur metabolism. Our study provides new insights into the evolution and function of spatiotemporal PM, as well as its interactions with carboxyl metabolites. Lactobacillus dominated in new PM, while methanogens and clostridia were predominant in older or deeper PM layers. The three distinct stages of prokaryotic community development in PM and HGT played critical roles in metabolic function of spatiotemporal PM. Furthermore, this study highlights the importance of α-diversity, β-diversity, methanogens, and Clostridium as useful indicators for assessing PM quality in the production of high-quality Baijiu.

In order to determine the best culture conditions of caproic acid-producing compound bacteria liquid, and explore a brewing technology of baijiu with both qingxiangxing and nongxiangxing flavor. In this study, caproic acid-producing compound bacteria solution was used as the experimental object, and the optimum process of caproic acid-producing compound bacteria solution was optimized by single factor, Plackett-Burman and response surface experiments. Then the caproic acid-producing compound bacteria solution was applied to the brewing process of xiaoqu qingxiangxing baijiu. The optimum culture conditions for caproic acid production were as follows: ethanol 2%, culture temperature 37 °C, dipotassium hydrogen phosphate 0.8g /L, anhydrous sodium acetate 5.0g /L, corn steep liquor dry powder 2.0g /L, initial pH value 8, liquid volume 70%. Under these conditions, the caproic acid yield could reach 5.07 g/L. Compared with the alcoholic fermentative material of traditional xiaoqu qingxiangxing baijiu, the alcoholic fermentative material inoculated with caproic acid compound bacteria solution had no significant difference in acidity, significantly increased moisture, and significantly decreased starch and reducing sugar. At the same time, compared with the traditional xiaoqu qingxiangxing alcoholic fermentative material, the content of caproic acid was significantly increased to 3.382 ug/g, the content of acetic acid was significantly decreased to 0.126 ug/g, the content of caproic acid was significantly decreased to 0.126 ug/g, the content of ethyl caproate was significantly increased to 1.593 ug/g, the content of ethyl caproate was significantly increased to 1.593 ug/g, and the content of ethyl acetate was significantly decreased to 0.945 ug/g. This study revealed the effect of caproic acid-producing compound bacteria on the physical and chemical indexes and flavor substances of traditional xiaoqu qingxiangxing alcoholic fermentative material, and significantly increased the content of ethyl caproate, which had guiding significance for the production of qingxiangxing and nongxiangxing baijiu.

Ultrasonic-assisted maceration offers a promising strategy to optimize flavor quality in fermented fruit wines, yet its application on kiwi wine (KW) remains unexplored. This study systematically investigated the impact of ultrasonic-assisted maceration on KW's metabolomics, flavoromics, sensory evaluation and antioxidant properties. 1H NMR identified 58 molecules in KW. Ultrasonic-assisted maceration significantly increased the concentrations of phenylalanine, acetoacetate, N,N-dimethylglycine, creatinine, and levulinate, while decreasing the levels of epicatechin, acetoin, cytidine, acetone, and hydroxyacetone. GC-IMS and GC-MS characterized 29 and 46 volatile molecules, respectively. The levels of esters and organic acids showed a direct proportionality to the intensity of the ultrasonic treatment, while an opposite trend was found for alcohols. Sensory scores of KW improved under the ultrasonic treatment. In detail, ultrasonic treatment for 40 min at 240 W led to the highest scores for taste-lasting and mouthfeel. A treatment of equal duration at 480 W led to the highest aroma score. Etongue effectively discriminated taste features brought to KW by the treatments. Moreover, ultrasonic treatments significantly enhanced KW's scavenging activity of DPPH, hydroxyl and superoxide anion radicals. This study provides theoretical references and basis for the application of ultrasonic-assisted maceration as a technique to potentially enhance the flavor quality of KW.

The study explores the effect of pre-fermentation heat treatment (PFHT) on the flavor and metabolomic profiles of kiwi wine (KW) derived from three kiwifruit cultivars. Six KW groups were involved, namely with/without PFHT for green (GWH/GW), yellow (YWH/YW), and red (RWH/RW) kiwifruit. E-tongue analysis effectively distinguished the taste profiles across these KW groups, identifying significant variations. A total of 97 volatile components were characterized using GC-MS and GC-IMS, 12 of them were identified as key volatile compounds based on a combination of t-tests (p < 0.05) and variable importance in projection (VIP) scores. GC-MS and GC-IMS results demonstrated that PFHT significantly altered volatile profiles, specifically decreasing ester content while increasing aldehyde levels in comparison to untreated samples. Furthermore, 71 non-volatile compounds were identified by 1H-NMR, with 10 key metabolites (p < 0.05, VIP > 1) contributing to the observed differences. PFHT notably influenced metabolomic profiles, particularly in carbohydrate and organic acid levels, displaying cultivar-specific differences. Green kiwifruit-derived KW showed the most pronounced sensitivity to PFHT, as reflected in both flavor and metabolic profiles. These findings offer valuable insights for optimizing KW production processes and scaling up industrial production.

The chicken feces, landfill leachate, landfill compost products, and effective microbiological (EM) composite bacteria were used to screen high-efficiency deodorizing strains. The preparation of composite microbial deodorizing bacteria was provided for the harmless and resource treatment of livestock and poultry feces. Firstly, the domestication enrichment method, plate separation method, primary screening, and re-screening method were used to screen the deodorizing strains. Then, all the screened strains were identified by the combination of conventional morphological identification and physiological and biochemical identification. At last, the antagonism experiment was used to study the antagonistic characteristics of all the strains, and the growth curve was used to study all the strains logarithmic growth period. Results indicated that four strains have dominant deodorization ability, including two strains of bacteria (BA13, BA11), one fungus strain (FU10), and one yeast strain (YE8). Strains BA13 and BA11 of hydrogen sulfide and ammonia removal efficiency were 88 +/- 3.45% and 88 +/- 3.98%, respectively. The four strains were prepared as the compound microbial deodorizer, and it obtained a dominant odor removal ability; the hydrogen sulfide and ammonia removal efficiencies were 70 +/- 5.43% and 68 +/- 4.56%, respectively. The compound microbial deodorizer can reduce the concentrations of hydrogen sulfide and ammonia and obtain excellent potential in deodorizing livestock and poultry feces composting.

Daqu, a representative solid-state fermentation product, produces saccharifying enzymes to degrade sorghum starch into fermentable sugars for ethanol synthesis. Spatial heterogeneity in Daqu drives community assembly. However, its regulatory role in enzyme-driven saccharification remains unclear. By integrating metagenomics and PacBio full-length sequencing, this study investigated how microenvironmental gradients across distinct Daqu layers (QP (surface layer), HQ (middle layer), QX (center layer)) shape saccharifying microbiota and activity. Saccharifying activity exhibited a declining surface-to-center gradient (e.g., QP: 870.9 ± 21.2 U/mL > HQ: 631.2 ± 16.4 U/mL > QX: 296.5 ± 16.1 U/mL on day 30, p < 0.05), paralleled by divergence in microenvironments. Metagenomics identified α-amylase and α-glucosidase as key saccharifying enzymes, primarily encoded by fungi; their abundance was inhibited by heat and humidity, yet promoted by acidity. Enzymatic validation confirmed higher saccharifying activity in QP and HQ core microbes (e.g., Lichtheimia ramosa: 43.16 ± 1.97 U/mL) than in QX (e.g., Paecilomyces variotii: 14.27 ± 1.25 U/mL). Network analysis revealed Lactobacillaceae are closely linked with saccharifying communities. This study establishes microenvironmental gradients as critical regulators of spatial saccharification in Daqu, informing strategies to optimize microbial consortia for baijiu production.

Transforming industrial by-products into reusable high-value resources, this study investigates the untapped potential of Luzhou-flavor baijiu lees as a substrate for cultivating pink Auricularia cornea. Forty-one baijiu lees- based substrate formulations were tested, identifying two optimal formulations that increased yields by 33.56 % and 28.54 % compared to the control group, while also improved surface area. Transcriptomic and metabolomic of bodies from the optimal formulations and the control group were analyzed across four growth stages. A total of 50-616 unique differentially expressed genes and 6-120 differentially expressed metabolites were identified, with KEGG analysis indicating enrichment in pathways related to amino acid and nitrogen metabolism, oxidation-reduction processes, oxidoreductase and catalytic activity, tetrapyrrole binding, and alcohol oxidase activity. Furthermore, 6-50 biomarkers were identified across the four growth stages. This groundbreaking study sheds light on the adaptive transcriptional and metabolic dynamics of pink A. cornea on baijiu lees substrates.

In the brewing of Luzhou-lavor Baijiu (LFB), starch content in fermented grains is critical for product quality and fermentation efficiency. Conventional Fehling's reagent assays, however, is labor-intensive and time-consuming, limiting real-time on-site quality control. To address this, a portable agar-based colorimetric hydrogel sensor integrated with Au@MnO2 nanoparticles (AH) was developed, combined with smartphone image analysis for label-free visual detection. The sensor operated via sequential reactions: starch in fermented grains was hydrolyzed to glucose by acid treatment, and the glucose was oxidized by glucose oxidase (GOx) to generate H2O2. The H2O2 triggered etching of Au@MnO2 nanoparticles, inducing a blue-to-pink color transition correlated with glucose concentration. With smartphone analysis, the sensor detected glucose in the range of 0-160 mu M (detection limit: 3.38 mu M) and converted results to starch content according to DB 34/T 2264-2014 for real-time assessment. Notably, the detection time for 100 samples was reduced from 2 h (Fehling's method) to 25 min (an 87.5 % time saving), eliminating the need for complex laboratory equipment. This simple and efficient sensor advanced on-site quality control in industrial brewing and held promise for scalable applications in food fermentation.

This study aimed to explore the influence of glutathione-enriched inactive dry yeast (g-IDY) addition on the changes in aroma and taste compounds in kiwi wine (KW), produced from green-, red-, and yellow-flesh kiwifruits. In total, 42 aroma compounds and 67 taste compounds were characterized using GC-IMS and 1H-NMR, respectively. Among them, six aroma compounds and thirty-one taste compounds were determined as key compounds based on a t-test, variable importance in projection scores, and relative odor activity value. Results indicated that g-IDY addition significantly decreased the concentration of hexyl acetate and increased the concentrations of 1-hexanol-M and pentanal in KW produced from green- and yellow-flesh kiwifruits. Pearson correlation analysis revealed strong associations between key aroma and taste compounds, particularly highlighting significant negative correlations between amino acids and aroma compounds. The findings could shed light on KW processing optimization and provide theoretical support for integrating g-IDY into KW industrial production.

An acidic polysaccharose (YL-D2N2) was isolated from crude polysaccharides of pink Auricularia cornea and characterized for its structural and antioxidant properties. YL-D2N2 consists of fucose, galactose, glucose, xylose, mannose and glucuronic acid in a molar ratio of 0.85: 1.50: 4.44: 27.52: 46.56: 19.13. It has a number-average molecular weight of about 52.811 kDa and a weight-average molecular weight of about 135.457 kDa. Structural characterization showed that YL-D2N2 consists of nine residues (Xylp-(1 ->, GlcpA-(1 ->, -> 2)-Xylp-(1 ->, -> 3)-Galp-(1 ->, -> 3)-Manp-(1 ->, -> 4)-GlcpA-(1 ->, -> 2,3)-Manp-(1 ->, -> 3,4)-Glcp-(1 ->, -> 3,6)-Manp-(1 ->), with a backbone of -> 3)-beta-D-Manp-(1 ->, -> 2,3)-alpha-D-Manp-(1 ->, -> 3,6)-alpha-D-Manp-(1 -> and side chains containing beta-D-Xylp-(1 -> and alpha-D-GlcpA-(1 ->. Notably, YL-D2N2 exhibits significant radical scavenging activity for superoxide anions, reaching 50.82 +/- 0.64 % at a concentration of 3.2 mg/mL. Overall, YL-D2N2 exhibits a unique chemical structure and specialized applications for targeting superoxide anion radicals, providing valuable insights for further exploration of its structure-activity relationship.

Fruit wines, produced through the fermentation of various fruits, are well-documented for their distinct flavor profiles. Intelligent sensory analysis, GC-TOF/MS and GC-IMS were used for the analysis of the volatile profile of eight types of fruit wines including 5 grape wine (SJ, LS, HY, TJ, FT), 1 fermented plum wine (FZ), 1 blueberry wine (HZ), as well as 1 configured plum wine (LM). A total of 281 compounds were identified through GC-TOF/MS, with esters and acids constituting over 80% of all samples. GC-IMS identified 60 compounds, predominantly including 16 esters, 11 alcohols, and 6 ketones, and 7 sulfur-containing compounds. This observation leads to the assumption that the IMS and MS data contain different information about the composition of the volatile profile. 37 and 18 differential compounds for TOF/MS data and IMS data were obtained, respectively. Three ranking algorithms combined with five machine learning models Neural Networks (NN), Random Forests (RF), Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Logistic Regression (LR) applied and identified both 58 key features from volatiles. LR and KNN achieved an overall classification of 0.95 and an F1 score greater than 0.9. For the IMS data, NN, LR, and KNN models exhibited accuracies and F1 scores greater than 0.9. This study advances fruit wine classification, benefiting the beverage industry and food chemistry research.