Optimizing the Maillard reaction for sweet flavor enhancement in instant tea: Integrating sensory evaluation and metabolomics Authors: Jie Wang, Shuyi Zhou, Qian Wang, Xueming Ma, Yijing Wang, Songyan Han, Yuan Feng, Fengfeng Qu, Guofeng Yu, Xinfu Zhang Published: 2026 ABSTRACT AbstractThis study aimed to optimize the Maillard reaction (MR) conditions for instant green tea (GT), black tea (BT), and oolong tea (OT) and to characterize the aroma profiles of the resulting MR products (MRPs). Supercritical CO₂ pretreatment reduced tea polyphenols and caffeine in instant teas. Key parameters (time, pH, temperature, exogenous substrate ratio) were optimized for each instant tea. The MRPs developed desirable sweet, roasted, and caramel-like aroma attributes. Volatile analysis revealed that the MR increased the levels of ketones, aldehydes, and esters, while decreasing hydrocarbons in the MRPs. Based on odor activity value and chemometric analysis, key aroma-active compounds such as indole, β-damascone, β-ionone, and dihydroactinidiolide were identified as crucial odorants responsible for the characteristic aroma of the MRPs. This work provides a practical strategy for producing potential aroma-modifying materials with sweet and roasted flavor, and offers insights into the broader utilization of instant tea by-products.Keywords: Instant tea, Maillard reaction, Sweet aroma, Aroma-active compounds Introduction Tea is one of the most popular beverages worldwide, due to its pleasant flavor and health benefits such as antioxidant and anticancer activities (Shang et al., 2021). Instant tea is a soluble processed product made from tea through hot water extraction, centrifugation, concentration, and drying (Min et al., 2025). Recently, an increasing number of instant tea products have been produced to meet consumers' demand for convenient and efficient tea beverages. In addition to being used as a beverage, instant tea can also be added to other foods to enhance flavor and nutritional value (Fu et al., 2021; Liu, Cai, et al., 2024). Maillard reaction (MR) is a nonenzymatic reaction occurring at high temperatures between amino and carbonyl groups. It imparts unique flavor and color to foods and is widely applied in baking, oil processing, and milk processing (Han et al., 2025; Lee et al., 2024; Pucci et al., 2024). In particular, with increasing demand for diverse food flavors, MR is also widely applied in the preparation of flavorings and aroma enhancers for food. MRPs of camellia seed meal, Lentinula edodes, and red pepper have been found to improve sensory qualities and possess potential as flavoring agents (Ni et al., 2021; Qiu et al., 2024; Song et al., 2021). Similarly, sugars and amino acids in tea can serve as MR substrates, and the impact of MRPs on aroma generation has attracted growing attention. Some studies have focused on the influence of the MR during tea processing, with volatiles such as 1-ethyl-1H-pyrrole, 2-methylpyrazine, and 2-ethylfuran identified as characteristic MRPs that enhance roasted aromas. (Gao et al., 2025; Li et al., 2025). Specifically, instant tea is not only rich in MR precursors such as amino acids (e.g., L-theanine, glutamic acid) and sugars (fructose, glucose) (Cesarettin et al., 2013; Min et al., 2025), but also exhibits high solubility and uniformity (Liu et al., 2026), offering favorable conditions for the MR. Currently, instant tea is mainly used for preparing tea beverages and bakery products (Zhang et al., 2019). However, its abundant aroma precursors offer an untapped potential for novel flavor creation. The development of flavorings that combine Maillard-type flavors with inherent tea notes could expand the applications of instant tea. MR is a complex process influenced by multiple factors. High contents of tea polyphenols and caffeine are considered as inhibitors that impede the MR (Liang et al., 2024; Passos et al., 2017). Therefore, properly removing tea polyphenols and caffeine from instant tea may help to reduce potential inhibition of the MR. In addition, temperature, time, pH, and exogenous substrate ratio are crucial parameters influencing the characteristics of the MR. For example, higher temperatures or longer times promote the formation of more MRPs, but may also generate undesirable off-flavor compounds (Cai et al., 2016; Yin et al., 2023). pH has also been reported to affect the volatile profile of the MR (Feng et al., 2024). Nevertheless, the optimization of parameters to achieve desirable flavor in instant tea MRPs warrants further investigation. Utilizing the MR to enhance instant tea flavor represents a viable strategy for adding value to instant tea products. Although MR flavoring is widely applied in food processing, its use in instant tea production remains rare. To date, the suitability of instant teas with different degrees of fermentation (green, black, and oolong tea) as MR substrates has not been systematically investigated; their optimal reaction parameters have not been established. More importantly, the aroma transformation patterns and key aroma compound formation mechanisms during MR processing remain unclear. Therefore, this study aims to optimize the MR of instant tea to produce sweet and roasted flavorings, and to elucidate the mechanism of characteristic aroma changes induced by the MR. The sensory qualities and chemical compositions of GT, BT, and OT were initially characterized. The efficacy of supercritical CO2 (SC-CO2) extraction for selective removal of tea polyphenols and caffeine was evaluated under varying times and pressures. The effects of temperature, time, pH, and exogenous substrate ratio on the MR were investigated, with optimal model parameters determined based on aroma assessments of MRPs. In particular, GC–MS coupled with chemometric analysis was used to identify key aroma compounds in the MRPs. This study provides insights into the flavor profile of instant tea products, offering valuable reference for developing sweet-roasted flavoring agents. Materials and methods Experimental materials Instant green tea (GT) and oolong tea (OT) powders were purchased from Fujian Da Ming Co., Ltd. (Zhangzhou, China), and black tea (BT) powder was obtained from Zhejiang Minghuang Natural Food Development Co., Ltd. (Zhejiang, China). Sensory evaluation Sensory evaluation was performed by 6 trained panelists (3 males and 3 females) according to a previous method (Chen et al., 2024) and GB/T 31740.1–2015. Briefly, 0.5 g of instant tea sample was infused with 150 mL of water (85 ± 5 °C), and aroma and taste were scored on a 100-point scale. The protocol was approved by the Institutional Review Board of Qingdao Agricultural University (20260306001). All participants provided written informed consent. Determination of biochemical components The content of tea polyphenols was determined using the Folin-Ciocalteu method (Qiao et al., 2023). The amino acid content was quantified according to National Standard of China (GB/T8314–2013). Catechins and caffeine were isolated and identified by ultra-high performance liquid chromatography (UPLC) (ACQUITY UPLC H-Class, Waters, Denmark), and the analytical conditions were based on a previous report with slight modifications (Wang et al., 2024). Soluble sugar content was analyzed by the anthrone‑sulfuric acid colorimetric method. Dephenolization and decaffeination with supercritical CO2 The tea samples were dephenolized and decaffeinated using a supercritical fluid extraction system model SFT-110 (Supercritical Fluid Technologies Inc., Newark, DE, USA). 5.0 g of sample was placed in the extraction chamber and the vessel was subsequently sealed. To determine the influence of time on the efficiency of caffeine and tea polyphenols elimination, the extraction times of 1, 2, and 3 h were tasted. The temperature and initial pressure were set at 50 °C and 20 MPa, respectively. After optimizing the time condition, the experiments were conducted at 20, 25, and 30 MPa. The optimal conditions were established by the removal ratio of caffeine and polyphenols. All experiments were repeated three times. To verify that polyphenol and caffeine removal enhances MR progression, samples both before and after SC-CO₂ extraction were subjected to the MR under the following conditions: 110 °C, 120 min, pH 8, and an exogenous substrate ratio of 1:3. Optimization of the Maillard reaction An L₁₆(44) orthogonal array (four factors at four levels, for a total of 16 runs) was used to evaluate the effects of temperature, time, pH, and exogenous substrate ratio on the MR. The specific parameters are listed in Table 1. In brief, 3.0 g of instant tea sample was placed in a crucible and dissolved in 80 °C water at a tea-to-water ratio of 1:10. The baking soda was prepared as a 1:10 (w/v) working solution for pH adjustment of the tea soup. Honey (containing 60.7% fructose and 21.4% glucose) and L-theanine (98% purity) were added as exogenous substrates to the MR system at honey-to-L-theanine weight ratios of 1:1, 1:2, 1:3, and 2:1, with unit ratio “1” set to 18 g/L (actual concentration = coefficient × 18 g/L). The tea soup was then placed in an oven at 100, 110, 120, and 130 °C, and reacted for 30, 60, 90, and 120 min, respectively. After reaction completion, the tea infusion was cooled to room temperature for subsequent analysis. Table 1The orthogonal test schemes of Maillard reaction.ModelFactorTemperature (°C)Time(min)pHExogenoussubstrate ratio110012091:1211012081:3312012071:2413012062:151109092:161009081:271309071:381209061:191206091:3101306081:1111006072:1121106061:2131303091:2141203082:1151103071:1161003061:3 The orthogonal test schemes of Maillard reaction. The absorbances of MRPs at 294 and 420 nm serve as indicators of MR progression. These values were measured using a microplate reader equipped with a high-sensitivity multifunctional detector (Biotek, USA). Range analysis was employed to evaluate the influence of each factor on the MR progression using the following formulas: where Y is the experimental data for each factor at each level, K is the total experimental data for each factor at each level, X represents the number of levels for each factor, Kavg represents the mean experimental data for each factor at each level, Kmax and Kmin are the maximum and minimum mean experimental data for each factor at each level, respectively. The aroma profile of MRPs was evaluated and scored by a systematically trained panel of seven experts using 5 aroma descriptors: sweet, roasted, floral, fruity, and caramel-like. Odor intensity was evaluated with a 5-point scale: “0” represented no perception, “1” represented very weak, “2” represented weak, “3” represented moderate, “4” represented strong, and “5” represented extremely strong. All experiments were performed in triplicate at room temperature (22 ± 1 °C), with distilled water as the blank control (assigned an intensity of 0). Analysis of volatile compounds in Maillard reaction products The volatile compounds in instant tea MRPs were extracted by solid-phase microextraction (SPME) and determined by gas chromatography–mass spectrometry (GC–MS). 10 mL of the MRPs and 5 μL of internal standard (ethyl decanoate, 10 μg/mL) were added to a headspace bottle. The sample was statically incubated at 60 °C for 15 min, followed by a 45-min extraction of the volatile compounds by an SPME fiber (50/30 μm DVB/CAR/PDMS) at the same temperature. After the extraction, the SPME fiber was inserted into the GC injector at 250 °C for 5 min to desorb the analytes. GC–MS analytical was performed based on a previous study (Song et al., 2016) using a 8890-7000E GC–MS system (Agilent, USA) equipped with a HP-INNOWAX capillary column (30 m × 0.25 mm × 0.25 μm). The GC injector temperature was 250 °C. Helium (purity >99.999%) served as the carrier gas at a constant flow rate of 1.85 mL/min. The oven temperature was programmed from 40 °C hold for 2 min, then increased to 160 °C at 6 °C /min, and finally increased to 250 °C at 8 °C /min and hold for 2 min. The MS spectra were acquired electron impact mode at 70-eV, with the ion source temperature set at 230 °C and the mass scan range set from 30 to 500 amu. The test was repeated three times for each sample. The volatile compounds were identified by matching mass spectra with the NIST (match score ≥ 75) and comparing retention index (RI) values with those reported in the literature. RI values were calculated relative to a series of n-alkanes series (C6-C25) under identical conditions. The internal standard method was conducted to semi-quantify the volatiles: where C is the concentration of any compound (μg/kg, semi-quantitative); Av is the chromatographic peak area of any compound; As is the chromatographic peak area of the internal standard; ms is the mass of the internal standard (μg); and mv is the mass of the MRPs sample (g). Due to differences in HS-SPME extraction efficiency and mass spectrometric response among compounds, the reported concentrations are semi-quantitative and intended for intergroup comparisons rather than absolute quantification. Calculation of odor activity values Odor activity values (OAVs) are considered a key indicator for assessing the contribution of volatile compounds to aroma (Liao et al., 2020). OAVs were determined from semi-quantitative values according to the equation: where C is the concentration of volatile compounds (μg/kg), and OT is its odor threshold (μg/kg). Cytotoxicity analysis Human gastric epithelial GES-1 cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum in a humidified incubator (5% CO2) at 37 °C. Exponentially growing cells were seeded into 96-well plates at a density of 5000 cells per well in a final volume of 100 μL of medium and allowed to adhere overnight. The cells were then treated with instant tea MRPs (3 mg/mL, based on the mass of the instant tea raw material), and control cells were treated with medium alone for 24 h. After treatment, 10 μL of CCK-8 solution was added to each well, followed by incubation at 37 °C for 4 h. Cytotoxicity was measured using the CCK-8 kit (Dojindo, CK04) according to the manufacturer's instructions. Cell viability was calculated as follows: Cell viability (%) = (A_sample − A_blank) / (A_control − A_blank) × 100. Statistical analysis Duncan's test in one-way ANOVA was used to analyze the significance (p < 0.05) by SPSS software version 25.0 (IBM Corporation, USA). Principal component analysis (PCA) and orthogonal partial least squares-discriminant analysis (OPLS-DA) were performed using SIMCA-P software (Umetrics, Umea, Sweden). The key components of MRPs in instant teas were identified by variable importance in projection (VIP) > 1 and odor activity value (OAV >1) (Hu et al., 2023). Hierarchical cluster analysis (HCA) was conducted using MultiExperiment Viewer (MEV) software (Oracle Corporation, Redwood Shores, CA, USA). Results and discussion Sensory evaluation The sensory score and description of the instant teas are presented in Table 2. Aroma evaluation showed that GT had the highest score with chestnut-like aroma followed by BT (floral aroma) and OT (pure and light floral aroma). A similar trend was observed for taste scores, which followed the order GT > BT > OT. These results indicated that the instant teas exhibited excellent aroma and taste qualities with distinct characteristics, confirming their suitability for subsequent experiments. Table 2The sensory evaluation of the instant teas.SamplesAromaTasteRemarksScoreRemarksScoreGTChestnut-like94.00 ± 2.65aAstringent, mellow93.33 ± 0.58aBTFloral92.67 ± 2.52bSweet93.00 ± 1.00aOTPure with light floral92.67 ± 1.53bMellow, slightly astringent91.67 ± 1.53bNote:The means with different superscript letters (a, b, c) in the same column are significantly different (p < 0 0.05). The sensory evaluation of the instant teas. Note:The means with different superscript letters (a, b, c) in the same column are significantly different (p < 0 0.05). Analysis of the major bioactive compounds As basic substrates for the MR, amino acids are directly involved in the formation of MRPs (Yang et al., 2023). In this study, amino acids were classified into four taste-related categories: umami, sweet, bitter, and odorless. Notably, GT exhibited significantly higher contents across all four categories compared to BT and OT, whereas no significant difference was observed between BT and OT (Fig. 1A-D). Soluble sugars contribute to the sweetness of tea liquor and also serve as MR substrates (Kitts, 2021). As shown in Fig. 1E, BT exhibited the highest soluble sugar content (55.23 mg/g), with GT and OT showing lower levels of 24.52 and 16.06 mg/g, respectively. The high soluble sugar content in BT may be attributable to the fermentation process (Jia et al., 2022). The content of tea polyphenols in GT (388.26 mg/g) was markedly higher than that in OT (298.48 mg/g) and BT (176.89 mg/g) (Fig. 1F). Similarly, the catechin content was highest in GT (20.51 mg/g), followed by OT (19.29 mg/g) and BT (5.06 mg/g) (Fig. 1G). A similar descending order (GT > OT > BT) was observed for most individual catechins (EC, CG, ECG, EGC, GCG, and EGCG), with the exception of C and GC (Fig. 1H-O). Consistent with a previous report (Salman et al., 2022), GT, as an unfermented tea, retains higher levels of phenolic compounds than BT and OT. Regarding caffeine, the content in GT (39.93 mg/g) was the highest, followed by OT (37.37 mg/g) and BT (28.99 mg/g). The biochemical traits of the three instant teas. (A) Bitter amino acids. (B) Sweet amino acids. (C) Umami amino acids. (D) Odorless amino acids. (E) Soluble sugars. (F) Tea polyphenols. (G) Catechins. (H) C. (I) EC. (J) CG. (K) GC. (L) ECG. (M) EGC. (N) GCG. (O)EGCG. (P) Caffeine. Therefore, these three instant teas contain multiple bioactive substances, including sugars and amino acids, which contribute to their attractive flavor profile by serving as potential substrates for the MR. However, given that the high levels of tea polyphenols and caffeine in the samples may inhibit the MR, appropriate removal of these compounds is necessary in subsequent experiments. Dephenolization and decaffeination of the instant teas In order to better facilitate subsequent MR progression, SC-CO2 was performed to remove tea polyphenols and caffeine. The selection of SC-CO₂ pretreatment conditions followed a two-step sequential procedure: optimal time was determined first, followed by optimal pressure. The selection criterion was to maximize polyphenol removal efficiency, with caffeine removal also monitored but not used as a primary decision factor. Specifically, removal efficiencies at 1, 2, and 3 h were compared at a fixed pressure of 20 MPa. GT and BT achieved the highest polyphenol removal at 1 h (40.72% and 15.83%, respectively), while OT reached its maximum at 3 h (60.40%); thus, the optimal times were 1 h for GT and BT, and 3 h for OT (Fig. 2A-C). The dephenolization rates of the three instant teas did not increase continuously with time, possibly owing to approaching saturation. For decaffeination, all three teas achieved their highest removal rates (16.36%–18.14%) at 3 h (Fig. 2A-C), consistent with a previous study that decaffeination efficiency increases with extraction time (Sun et al., 2021). Effects of SC-CO₂ treatment on polyphenols and caffeine removal and MR progression in instant teas. (A–C) Time-dependent changes in polyphenols and caffeine contents in GT, BT, and OT; (D—F) pressure-dependent changes in polyphenols and caffeine contents in GT, BT, and OT; (G–I) effects of SC-CO₂ pretreatment on MR progression in GT, BT, and OT. To assess the effect of pressure on removal efficiency, the time was fixed at 1 h, and pressures of 20, 25, and 30 MPa were tested. For GT and OT, the highest dephenolization efficiency was achieved at 20 MPa (40.72% and 60.40%), which was selected as optimal (Fig. 2D, F). In contrast, BT exhibited increasing dephenolization efficiency with increasing pressure, reaching its highest at 30 MPa, which was therefore selected as optimal (Fig. 2E). In addition, the highest decaffeination rates for GT (25.06%) and OT (25.62%) were observed at 30 MPa and 25 MPa, respectively (Fig. 2D, F); however, the decaffeination rate for BT did not vary significantly with pressure (Fig. 2E). In summary, the optimal SC-CO2 pretreatment conditions were 1 h and 20 MPa for GT, 1 h and 30 MPa for BT, and 3 h and 20 MPa for OT. The role of SC-CO₂ pretreatment in this study was to reduce tea polyphenols and caffeine interference, thereby creating more favorable substrate conditions for the subsequent MR, rather than to achieve complete removal of tea polyphenols and caffeine. To further verify that the removal of tea polyphenols and caffeine indeed promotes MR progression, both SC-CO₂-treated and untreated instant teas were subjected to the MR, and the reaction extent was monitored using A₂₉₄ and A₄₂₀ values. As shown in Fig. 2G-I, the A₂₉₄ and A₄₂₀ values were significantly higher in the SC-CO₂-treated groups than in the untreated control groups. This study demonstrates that partial removal of tea polyphenols and caffeine by SC-CO₂ pretreatment promotes MR progression. Notably, SC-CO₂ is applicable not only to polyphenol and caffeine extraction, but also to the recovery of carotenoids, tocopherols, amino acids, and other compounds (Le, 2025). Indeed, previous studies have demonstrated that SC-CO₂ can co-extract amino acids and total chlorophyll from tea during decaffeination (Park et al., 2012; Wang et al., 2022). In the present study, while the pretreatment mainly targeted tea polyphenols and caffeine, it may have also partially removed fat-soluble compounds and amino acids. In future research, the effects of SC-CO₂ pretreatment on other substances in instant tea could be analyzed in detail. Optimization of the Maillard reaction in instant tea Effects of time, temperature, pH, exogenous substrate ratio of MR Temperature, time, pH, and exogenous substrate ratio are key factors influencing the MR; therefore, an orthogonal scheme was designed to investigate their effects (Table 1). The absorbance values at 294 nm (A294) and 420 nm (A420) are indicative of the intermediate and final stages of the MR, respectively (Zhang, Tan, et al., 2025). In this study, the A294 and A420 were measured for all 16 MR model systems. Fig. 3A-B showed that the trends of A294 and A420 varied among the three instant tea samples across the different model numbers, suggesting that the effects of MR factors differ depending on tea type. Effect of different models of the Maillard reaction. (A) A294 value of different MR models. (B) A420 value of different MR models. (C) Aroma evaluation of screened MR models in GT. (D) Aroma evaluation of screened MR models in BT. (E) Aroma evaluation of screened MR models in OT. To further evaluate the influence of each factor on MR progression, range analysis was performed based on the combined A294 and A420 values, as presented in Table 3. The four experimental levels are specified in Table S1, and the effects of each factor at the four levels were calculated and denoted as K1、K2、K3、and K4. The importance of each factor was evaluated by the range value (R value), with a greater R value indicating a stronger effect of the corresponding factor (Mao et al., 2021). As shown in Table 3, the R values of the four factors for the three instant teas were ranked differently. Based on R values, the order of influence of the four factors on MR progression in GT was time > temperature > pH > exogenous substrate ratio, and the parameters that most favored MR progression for GT were 120 °C, 60 min, pH 8, and an exogenous substrate ratio of 1:3. For BT, the R values followed the order pH > temperature > time > exogenous substrate ratio, with the conditions favoring MR progression being 130 °C, 120 min, pH 7, and a ratio of 1:1. In terms of OT, the order was temperature > time > pH > exogenous substrate ratio, with the conditions favoring MR progression being 110 °C, 120 min, pH 8, a ratio of 1:3. Table 3Range analysis of the evaluation indexes.SamplesTemperature (°C)Time (min)pHExogenous substrate ratioGTK16.576.386.776.27K26.646.026.936.63K36.707.406.236.91K46.606.726.596.70R0.141.380.700.65BTK16.717.016.626.99K26.876.736.966.80K36.926.957.196.93K47.026.836.746.79R0.210.180.220.06OTK16.116.555.735.92K26.526.276.435.81K36.255.886.386.46K45.475.655.816.16R1.060.900.690.65 Range analysis of the evaluation indexes. Overall, temperature, time, pH, and exogenous substrate ratio are key factors influencing the MR of instant tea. Temperature is a primary factor affecting the rate of MR and the characteristics of MRPs (Hosry et al., 2025). The results of this study indicate that instant tea produced greater production of MRPs at temperatures between 110 and 130 °C, whereas at 100 °C, insufficient heat might have limited MRP generation. This is consistent with a previous report suggesting that 110 °C serves as the critical threshold for the MR (Lan et al., 2010). Heating time is another critical variable, determining the accumulation of intermediate and the conversion of end products (Jiang et al., 2018). In general, prolonged heating leads to the accumulation of more MRPs but may also result in the loss of flavor compounds. Consistent with a previous study (Xiang et al., 2022), the MR proceeded more readily at higher pH values. Compared with temperature, time, and pH, which typically exert more direct and stronger driving effects on the MR (Hosry et al., 2025), the exogenous substrate ratio had the least influence in this study. This may be attributed to the endogenous sugars and amino acids inherently present in instant tea attenuating the effect of exogenous substrates, as well as to the difference from model systems that rely solely on exogenously added sugars and amino acids (Hu et al., 2025). It is noteworthy that excessive MR readily occurs under specific conditions (such as high temperature, prolonged heating, and alkaline environments), resulting in adverse effects on the aroma profile of MRPs (Yang, Li, et al., 2024). A higher MR degree does not necessarily imply better aroma; therefore, in addition to investigating the effect of reaction conditions on MRPs yield, assessing the aromatic characteristics of MRPs is of greater significance. Accordingly, the A₂₉₄ and A₄₂₀ values were used to monitor the progress of the MR (intermediate products and browning degree) and to evaluate the effects of various factors (temperature, time, pH, and exogenous substrate ratio) on the reaction extent, rather than to directly predict aroma quality. Based on these results, six models per tea type with high A₂₉₄ and A₄₂₀ values were selected for the subsequent sensory evaluation to identify the model with preferred aroma profiles. Specifically, the selected models are as follows: for GT, models 1, 2, 9, 10, 12, and 14; for OT, models 3, 4, 5, 6, 7, and 14; and for BT, models 2, 3, 7, 10, 11, and 15. Selection of optimal MR model based on aroma evaluation To identify the optimal aroma MR model, the selected models were subjected to aroma evaluation, with results shown in Fig. 3C-E. The MRPs of the selected MR models in GT, BT, and OT exhibited superior aromatic characteristics compared with the CK. As shown in Fig. 3C, Model 1 in GT produced a more prominent roasted aroma, along with enhanced sweet, caramel-like, and fruity notes, resulting in the most favorable overall aroma profile. This suggests that the optimal flavor outcome for GT was achieved under the MR conditions of 100 °C, 120 min, pH 9, and an exogenous substrate ratio of 1:1. Fig. 3D presents the aroma profiles of the six MR models in BT. Model 10 was distinguished by its high-intensity roasted and sweet notes, which were generated under the following MR conditions: 130 °C, 60 min, pH 8, and an exogenous substrate ratio of 1:1. Under the conditions of 120 °C, 30 min, pH 8, and a 2:1 exogenous substrate ratio, Model 14 in OT (Fig. 3E) yielded the most favorable aroma profile, marked by a prominent sweet note and a harmonious blend of roasted, caramel-like, fruity, and floral aromas. Notably, models 2, 3, and 5 in OT not only failed to enhance the roasted and sweet aromas but actually diminished these flavors compared to the CK. This is presumed to be related to the high temperature and prolonged heating conditions. This result suggests that the degree of MR and the flavor enhancement are not always positively correlated, and some MRPs may negatively effect on the aroma (Shakoor et al., 2022). In summary, the aroma of instant tea was significantly enhanced after MR processing, particularly in terms of sweet, roasted, and caramel-like flavors. The MR exerts a significant influence on the aroma of instant teas, yet its effects vary across different tea types. For instance, OT tended to lose its characteristic aroma under prolonged reaction conditions, whereas BT achieved optimal aroma at 130 °C for 60 min. The aroma profiles of MRPs from different instant teas exhibit varying responses to MR parameters, which can be attributed to differences in their intrinsic characteristics and bioactive compound contents. In general, temperature and time play dominant roles in shaping the aroma profiles of MRPs. Low temperature and short time limit the MR to its initial stage (Cao et al., 2022; Chen et al., 2019), thereby limiting the development of desirable flavor (Wang et al., 2026). For instance, reaction at 120 °C for 120 min was more conducive to the formation of meaty flavor, whereas fewer aromatic compounds were produced at 100 °C (Zhan et al., 2025). Conversely, excessive conditions, such as prolonged heating at high temperature (≥170 °C), may generate undesirable byproducts, including advanced glycation end products (AGEs), melanoidins, and even bitter flavors (Liu, Liu, et al., 2024; Lund & Ray, 2017). Hence, striking a balance between MR conditions and flavor is of paramount importance (Huang et al., 2023). In addition, similar to a previous study (Zhang, Dong, et al., 2025), the MRPs of instant teas in this study exhibited pronounced caramelized, sweet, and roasted aromas. This discovery offers a novel strategy for developing promising flavor-modifying ingredients, thereby overcoming the limitations of traditional tea beverages. It should also be noted that although the laboratory-scale parameters optimized in this study fall within the conventional operating ranges of the food industry, several engineering challenges remain to be addressed before industrial implementation, such as uniform heat transfer and the impact of drying methods on aroma retention. Future pilot-scale studies will systematically evaluate these issues, and cost-benefit analyses will be conducted to further validate economic feasibility. GC–MS of Maillard reaction products in instant teas The optimal models (GT-model 1, BT-model 10, and OT-model 14) were selected for aroma enhancement. HS-SPME/GC–MS was employed to detect differences in the volatile profiles of the three instant teas before and after the MR. As shown in Table S2, a total of 141 volatile compounds were identified across all samples, comprising 76 in GT-CK, 75 in GT-MR, 61 in BT-CK, 64 in BT-MR, 86 in OT-CK, and 85 in OT-MR. The Venn diagram was constructed based on the data in Table S2 to visualize common and unique elements between CK and MR group. As depicted in Fig. 4A, 69 volatiles were common to GT-CK and GT-MR, 58 to BT-CK and BT-MR, and 77 to OT-CK and OT-MR. Notably, 7, 6, 3, 6, 9, and 8 compounds were only detected in GT-CK, GT-MR, BT-CK, BT-MR, OT-CK, and OT-MR, respectively. These unique volatiles might act as indicators of MR-induced changes. Interestingly, dimethyl ether and β-damascone were detected only in all samples following the MR, suggesting the new compounds formed through MR may play a key role in shaping the aroma of MRPs of instant teas. The effect of the Maillard reaction on the composition and content of aroma compounds in instant teas. (A) Venn diagram of volatile compounds in instant teas (GT-CK vs GT-MR, BT-CK vs BT-MR, and OT-CK vs OT-MR). (B) Types and relative contents of volatile compounds in instant teas. The volatile components included 21 alcohols, 43 hydrocarbons, 28 ketones, 17 esters, 6 aldehydes, 9 oxygen compounds, 12 nitrogen compounds, and 5 others. Fig. 4B depicted the changes in the major categories of volatile compounds in instant teas induced by MR processing. The alcohol content in GT and BT showed no obvious changes after MR, whereas OT displayed an increasing trend, which is in agreement with results from the previous study (Chen et al., 2025). Ketones contribute fruity, floral, and woody scent, playing a pivotal role in aroma due to their low odor threshold (Zhang, Lu, et al., 2025). Aligning with prior research (Hu et al., 2025), MR treatment markedly elevated the contents of ketones in GT and OT, but resulted in only a minor increase in BT. Notably, the marked increase in aldehydes in GT and BT may be attributed to Strecker degradation and the oxidative degradation of hydrocarbons (Liu, Sun, et al., 2022). Following MR treatment, a decrease in hydrocarbon content was observed in all three instant teas, with OT and GT showing a significant decrease. As previously reported, hydrocarbons can be converted into ketones and aldehydes under specific conditions (120 °C, 45 min, pH 7), which accounts for the MR-induced reduction in hydrocarbons (Fu et al., 2023). As for esters and oxygen compounds, the contents were markedly higher in the MR group than in the CK group. Additionally, the nitrogenous compound content showed a significant increase in OT but a decrease in BT. Overall, the MR exerted a remarkable influence on the volatile compound composition, especially on the levels of hydrocarbons, ketones, esters, and aldehydes. In order to visualize overall volatiles differences among the samples, PCA was performed on the data from Table S2. In this PCA model, principal component 1 (PC1) and principal component 2 (PC2) contributed 39.8% and 31.4% of total variation (71.4%). As observed in Fig. 5A, the three types of instant tea were clearly separated into distinct quadrants in the PCA score plot, with a clear separation between pre- and post-MR samples. This pattern demonstrates that the aroma profiles differ not only among tea types but are also significantly altered by the MR. OPLS-DA models were subsequently employed to further delineate the MR-induced changes in volatiles. In the score plots (Fig. 5B-C), samples from GT, BT, and OT were distinctly separated along the Y-axis, with pre- and post-MR groups located on opposite sides, highlighting the pronounced impact of the MR. As shown in Table S3 and Fig. S1, the negative Q2 intercept in the permutation test demonstrated that these models were stable and reliable. Variable importance in the projection (VIP) with values greater than 1 are typically regarded as an important contribution to classification (Zheng et al., 2024). A total of 57, 49, and 63 differential compounds were screened with a VIP > 1 from the OPLS-DA models comparing CK and MR groups in GT, BT, and OT, respectively (Fig. S2). PCA and OPLS-DA of different groups. (A) PCA score plot. (B) OPLS-DA plot of GT-CK vs GT-MR. (C) OPLS-DA plot of BT-CK vs BT-MR, (D) OPLS-DA plot of OT-CK vs OT-MR. The perceived aroma of a compound is governed by both its concentration and its odor threshold. The odor activity value (OAV), defined as the ratio of a volatile compound's concentration to its odor detection threshold, is a critical indicator for assessing its contribution to food aroma (Xi et al., 2024). Accordingly, the OAVs of the differential compounds screened above were subsequently calculated. Key aroma compounds were then identified by applying the dual criteria of VIP > 1 and OAV > 1 (Tables S4-S6, Fig. 6). It should be noted that OAVs were calculated using water-based odor thresholds, whereas actual perception thresholds in tea infusions may be influenced by matrix components such as caffeine and polyphenols (Ding et al., 2025; Du, Ma, et al., 2025). Nevertheless, OAV remains the most widely accepted approach for screening key aroma compounds in food flavor research. In this study, OAV was used to semi-quantitatively compare the relative changes of aroma compounds before and after MR treatment, rather than to predict their absolute aroma intensity. Clustering patterns of key compounds in instant teas based on HCA, and cytotoxicity assessment of the MRPs. (A-C) HCA of GT-CK vs GT-MR, BT-CK vs BT-MR, and OT-CK vs OT-MR, respectively. (D) Cytotoxicity assessment of GT-MR, BT-MR, and OT-MR. As shown in Fig. 6A, 34 key compounds were clearly divided into two groups. The contents of 23 key compounds in GT increased after MR, including dimethyl ether, β-damascone, β-ionone, β-cyclocitral, safranal, indole, 1-octen-3-ol, dihydroactinidiolide, etc. On the contrary, 11 key compounds such as hexanoic acid, dimethyl sulfide, and d-limonene were decreased after MR. As shown in Fig. 6B, a total of 18 key aroma compounds were identified as differential substances between the BT-CK and BT-MR groups. Among these, 9 compounds—including cis-3-hexenyl hexanoate, β-damascone, α-terpineol, indole, and β-ionone epoxide—were present at higher levels in BT-MR. In contrast, the remaining 9 compounds, such as ethylbenzene, β-ocimene, d-limonene, and mesitylene, showed lower contents following MR treatment. Fig. 6C illustrated that in comparison with OT-CK group, the levels of 9 compounds including β-damascone, eucalyptol, ethyl caprylate, indole, 6-methyl-5-hepten-2-one, α-terpineol, dimethyl ether, 2,4-di-tert-butylphenol, and 4-methyl-3-penten-2-one were elevated, and 17 compounds—mostly hydrocarbons, including camphene, β-phellandrene, dimethyl sulfide, and d-limonene were reduced in OT-MR group. The three instant teas exhibited divergent responses to MR, resulting from the specific MR conditions applied in their respective optimal models, combined with their inherent compositional differences. Interestingly, some compounds such as β-damascone (honey, floral) (Du, Wang, et al., 2025), indole (floral) (Yin, Xiao, et al., 2023), dihydroactinidiolide (roasted, fruity) (Gong et al., 2017; Zhao et al., 2024), and safranal (herb, sweet) (Liu, Zheng, et al., 2022) repeatedly appeared in the MRPs of three or two instant teas. These compounds may be key contributors to the sweet and caramel-like aromas of MRPs in instant teas, corroborating the sensory evaluation results. Indole was identified as a key pyrolysis product of MR contributing to baked aromas (Yang, Wang, et al., 2024), which is consistent with the classification of indole as a key compound in instant tea MRPs in this study. Additionally, other key compounds including β-damascone, dihydroactinidiolide, β-ionone, β-cyclocitral, and safranal are derived from carotenoid degradation, also increased in concentration during thermal processing via the MR. These reactions are accelerated by high-temperature MR conditions but are not exclusive to the MR. Therefore, for clarity, all heat-induced volatiles (MR-derived and non-MR-derived) are collectively termed “key differentiating compounds”, as defined by their elevated post-MR levels. In summary, MR significantly changes the volatile profile to promote aroma transformation in instant tea. Unlike previous studies (Gong et al., 2025; Yao et al., 2024), this research did not identify numerous furans, pyrazines, and pyrroles as characteristic MRPs. This discrepancy may stem from several factors. One contributing factor is the inhibitory effects of tea polyphenols and caffeine on the MR, together with differences in substrates and reaction conditions (Han et al., 2022; Starowicz & Zieliński, 2019), which may have limited the generation of heterocyclic compounds. Another factor is that the MR conditions in this study were optimized based on sweet aroma characteristics, leading to limited formation of heterocyclic compounds associated with roasted and burnt notes. Additionally, detection conditions may also have affected compound identification. Although the extraction fiber and chromatographic column used in this study are suitable for the enrichment or separation of polar compounds (Guo et al., 2022; Dekić et al., 2025), future studies could optimize settings such as extraction parameters and temperature programming to improve the detection sensitivity of polar heterocyclic compounds. Furthermore, there is no direct positive correlation between MRPs yield and aroma quality, as excessive reaction may produce burnt off-notes. MRPs are complex mixtures generated from sugars and amino acids under thermal reaction conditions, including taste and aroma compounds, pigments, and other functional components (Hosry et al., 2025). Among these, only certain volatiles, such as 2,3-dihydro-3,5-dihydroxy-6-methyl-4H-pyran-4-one, furfural, and 2-methylpyrazine, possess odor activity and contribute to the sweet and roasted aroma (Shen et al., 2026; Wang et al., 2025). Therefore, this study employed OAV to identify key aroma contributors from MRPs, rather than simply equating the total amount of MRPs with aroma quality. Cytotoxicity assessment To evaluate the potential cytotoxic effect of the tested samples on cell viability, a CCK-8 assay was performed. As shown in Fig. 6D, the cell viability values ranged from 96.26% to 99.14%, and no statistically significant decrease in cell viability was observed compared with the control. The results indicated that the MRPs from the three instant teas exhibited no significant cytotoxicity under the tested conditions. The absence of cytotoxicity may be reasonably attributed to the tea polyphenols present in the tea matrix. Previous studies have demonstrated that tea polyphenols can effectively inhibit the formation of AGEs and 5-hydroxymethylfurfural through mechanisms including trapping α-dicarbonyl compounds, scavenging free radicals, and chelating metal ions (Li et al., 2024; Ye et al., 2024; Yin et al., 2014). These protective effects likely contributed to the low toxicity profile of the MRPs observed in this study. Nevertheless, the formation of hazardous compounds during the MR is highly dependent on processing conditions. For instance, the content of AGEs increases rapidly when the temperature rises to 170 °C (Liu et al., 2024b). Consequently, the relatively mild heating conditions employed in the present study may have been insufficient to induce substantial formation of highly toxic MRPs. Conclusion This study investigated the optimal MR conditions for instant teas and the consequent impact on their flavor profiles. Characteristic analysis confirmed that instant teas possess favorable aroma and a variety of bioactive components, making them suitable substrates for the MR. SC-CO₂ treatment partially removed tea polyphenols and caffeine. By monitoring A294 and A420, the effects of temperature, time, pH, and exogenous substrate ratio on MR progression were evaluated. Aroma assessment indicated that the MRPs of instant teas exhibited pleasant sweet and roasted flavors. Furthermore, GC–MS analysis revealed that the MR significantly increased the contents of ketones, aldehydes, and esters, while reducing hydrocarbons levels. A total of 23, 15, and 9 volatiles were identified as key aroma compounds in GT-MR, BT-MR, and OT-MR, respectively. Among these, key volatiles such as indole, β-damascone, β-ionone, and dihydroactinidiolide played crucial roles in contributing sweet aroma notes to the MRPs of instant teas. Additionally, the instant tea MRPs did not exhibit cytotoxicity under the tested conditions. In conclusion, MRPs derived from instant teas represent a novel ingredient with unique sweet, caramel-like, roasted, and tea-specific nuances, offering an innovative strategy for developing potential aroma-modifying materials. CRediT authorship contribution statement Jie Wang: Writing – original draft, Methodology, Investigation. Shuyi Zhou: Software, Investigation. Qian Wang: Methodology, Conceptualization. Xueming Ma: Validation. Yijing Wang: Methodology. Songyan Han: Investigation. Yuan Feng: Formal analysis. Fengfeng Qu: Writing – review & editing, Supervision. Guofeng Yu: Validation, Resources, Conceptualization. Xinfu Zhang: Project administration, Funding acquisition, Conceptualization. Ethical statement The sensory evaluation study involved in this research has been reviewed and formally approved by the Institutional Review Board of Qingdao Agricultural University (approval number: 20260306001). Prior to the evaluation, all participants were informed about the purpose and procedures of the study and provided written informed consent. Participation was entirely voluntary, and participants were allowed to withdraw at any time. The privacy and confidentiality of all participants were strictly protected, and all data were anonymized prior to analysis. Declaration of competing interest All authors declared that they had no known competitive financial interests or personal relationships that could have appeared to influence the study reported in this paper. Acknowledgments This study was supported by the National Natural Science Foundation of China (32372764), the Technology System of Modern Agricultural Industry in Shandong Province (SDARS-19-01), the cooperation project of China Tobacco Shandong Industrial Co., Ltd. (2022370000340592), and the Project of Shandong Provincial Natural Science Foundation (ZR2022MC088). Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.fochx.2026.104375. Contributor Information Guofeng Yu, Email: jingyu825158@163.com. Xinfu Zhang, Email: zxftea@163.com. Supplementary data Supplementary material Data availability Data will be made available on request. References Associated Data Supplementary Materials Supplementary material Data Availability Statement Data will be made available on request.