Next Article in Journal
The Mechanism of GABA in Attenuating Neuroinflammation in Alzheimer’s Disease: CP/CEBPα/miR-34a-Mediated Suppression of HDAC2/3 in Astrocytes
Previous Article in Journal
Comparative Evaluation of Different Fruit Parts of Phyllanthus emblica L. for Functional Food Applications: Nutritional Composition, Safety, and Gut Microbiota-Associated Aging Benefits
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Taste Modulation of White Tea by Red/Blue-LED-Assisted Withering Revealed via Non-Volatile Metabolomics

1
College of Tea Science and Tea Culture, Zhejiang A&F University, Hangzhou 311300, China
2
College of Horticulture, South China Agricultural University, Guangzhou 510640, China
3
School of Food Science and Technology, Nanchang University, Nanchang 330031, China
*
Authors to whom correspondence should be addressed.
These authors contributed equally to this work.
Foods 2026, 15(5), 836; https://doi.org/10.3390/foods15050836
Submission received: 26 January 2026 / Revised: 13 February 2026 / Accepted: 20 February 2026 / Published: 3 March 2026
(This article belongs to the Topic Advances in Analysis of Food and Beverages, 2nd Edition)

Abstract

Background: Red/blue- light-emitting diode (LED)-assisted withering provides a controllable spectral input to steer tea quality, yet metabolite-level evidence linking spectrum composition to quantitative taste phenotypes in white tea remains insufficient. Methods: Fresh leaves were withered under supplemental red/blue LEDs—S0, S1, S2, S3, S4, and S5—and the resulting white teas were evaluated by quantitative descriptive analysis (QDA), untargeted metabolomics, weighted gene co-expression network analysis (WGCNA), and high-performance liquid chromatography (HPLC) quantification of caffeine, gallic acid, and eight catechin monomers. Results: Red/blue-mixed spectrum enhanced the overall sensory quality relative to the incandescent lamp; S3 maximized sweetness and freshness, whereas S4 minimized bitterness and astringency and achieved the highest overall score. Untargeted metabolomics showed the largest deviation for S0 vs. S4. A total of 18 common metabolites were identified between the S0 and light-supplemented withering treatments, dominated by saccharides and related derivatives. WGCNA linked a saccharide-centered module to higher sweetness/freshness/smoothness and a lipid-oxylipin-centered module to stronger bitterness/astringency. HPLC independently confirmed that S4 contained the lowest catechins and caffeine, supporting its reduced bitter/astringent attributes. Conclusions: Overall, the application of mixed red-blue spectra offered a promising approach to enhancing the palatability of white tea by coordinately intensifying saccharide metabolism while simultaneously suppressing key bitter and astringent components. Our study provided a scientific basis for standardizing white tea processing to enhance sensory quality.

1. Introduction

White tea is characterized by its minimal processing and distinctively fresh, delicate flavor, and its sensory quality is highly contingent upon moisture loss and endogenous metabolic restructuring during the withering phase [1]. Research has indicated that withering induces dynamic fluctuations in non-volatile constituents—including soluble sugars, amino acids, phenolic acids, flavonoids, and certain lipid derivatives—which collectively define critical taste attributes such as sweetness, umami (freshness), bitterness, and astringency [2]. Natural environmental fluctuations such as light, temperature, humidity, and airflow have a substantial impact on the metabolic trajectory of the withering process [3]. Consequently, reshaping non-volatile metabolites by manipulating the withering microenvironment, while maintaining uniform raw material grades and processing protocols, has become a pivotal strategy in improving the stability of white tea quality through standardized and controllable processing.
Light quality represents one of the most versatile parameters in controlled environments. Beyond the mere presence or absence of light, spectral quality acts as a regulatory trigger. Within a controlled environment, specific red-blue ratios are converted by photoreceptors into downstream molecular cues, thereby orchestrating the rechanneling of metabolic pathways [4]. Red light perception, mediated by phytochromes, links light signaling to carbon assimilation, allocation, and saccharide signaling, thereby fundamentally reshaping sugar metabolism and the supply of carbon skeletons at the source [5,6]. By contrast, blue light is primarily perceived by cryptochromes and phototropins. This cascade broadly activates the light-responsive transcriptional network and upregulates secondary metabolic pathways, such as phenylpropanoid-flavonoid and polyphenol biosynthesis, which are closely associated with the synthesis and modification of compounds contributing to bitterness and astringency [5,7]. Furthermore, the interplay between red and blue light signaling is integrated with stress and redox pathways, thereby reconfiguring defense-related metabolic profiles and protein turnover [4]. This coordination drives a fundamental shift in carbon and nitrogen allocation, which redirects saccharide availability and dictates amino acid synthesis and transamination, while altered phenylpropanoid-flavonoid flux reshapes the chemical drivers of bitterness and astringency [8]. Therefore, the light environment serves as a precision flux switch, offering the potential for directional and predictable optimization of tea flavor by balancing key taste attributes.
In tea plants, light quality has been demonstrated to modulate metabolites associated with both aroma and taste [9]. For instance, monochromatic light treatments significantly alter the formation of volatile organic compounds (VOCs) in fresh leaves, suggesting a direct regulatory potential over fatty acid- and phenylpropanoid-derived aroma pathways [10]. During the processing stage, withering and illumination conditions similarly reshape the metabolic profile and quality of the final product; in black tea, specific light-induced changes in key quality attributes have been documented during the withering process [11]. Furthermore, adopting LED technology in controlled-environment withering has shown promise, particularly in oolong tea, where it reconfigures non-volatile profiles (such as flavonoids, amino acids, alkaloids) to favor a sweeter, more umami, and less bitter sensory profile [12]. Similarly, comparative metabolomic analyses in white tea underscore the role of withering environments (e.g., sunlight versus indoor) in restructuring the balance of key flavor precursors, including sugars and polyphenols [3]. While existing evidence supports the correlation between light quality, metabolism, and flavor, several knowledge gaps persist. The research predominantly addresses the presence or absence of light, whereas the systemic influence of specific red/blue light spectral compositions on the white tea metabolome remains under-explored. Building upon previous research, we proposed the hypothesis that an optimized red-to-blue light ratio during the withering can effectively regulate the metabolite composition of white tea. This is achieved through the modulation of carbon-nitrogen partitioning and phenylpropanoid/flavonoid metabolism, ultimately leading to predictable metabolic shifts in the sensory flavor profiles of white tea. Therefore, in this study, we focused on a gradient of red/blue light ratios within a controlled facility-based environment to comprehensively assess their role in shaping the taste attributes and non-volatile profiles of white tea.

2. Materials and Methods

2.1. Chemicals and Reagents

Tea components including C ((+)-Catechin, B21722, high-performance liquid chromatography (HPLC) purity ≥ 98%), GC ((-)-Gallocatechin, B20849, HPLC ≥ 98%), EC (Epicatechin, B20102, HPLC ≥ 98%), EGC ((-)-Epigallocatechin, B20105, HPLC ≥ 98%), CG (Catechin gallate, B20350, HPLC ≥ 98%), ECG ((-)-Epicatechin gallate, B20103, HPLC ≥ 98%), GCG ((-)-Gallocatechin gallate, B20850, HPLC ≥ 98%), EGCG (Epigallocatechin gallate, B20106, HPLC ≥ 98%), and gallic acid (B20851, HPLC ≥ 98%) were purchased from Yuanye Biological technology (Shanghai, China). All reagents were stored at 2–8 °C. Epicatechin gallate (Cat#B20103, HPLC ≥ 98%) was purchased from Yuanye Biological Technology (Shanghai, China) and stored at −20 °C. Caffeine (P/N:71559, HPLC = 99.8%, TMstandard, Beijing, China) was also stored at −20 °C. Chromatographic-grade methanol methanol was purchased from Merck KGaA (Darmstadt, Germany), Chromatographic-grade acetonitrile was purchased from CNW Technologies Co., Ltd. (Shanghai, China), and formic acid was purchased from Aladdin (Shanghai, China). All three were stored at room temperature.

2.2. Preparation of Samples

Fresh leaves of the tea cultivar Fuyun No. 6 were harvested from the Tea Garden of the Teaching and Research Base, South China Agricultural University (Guangzhou, Guangdong, China; 113°37′34″ E, 23°14′11″ N), on 17 March 2022. Leaves were plucked at the standard of one bud with two leaves and processed into white tea in accordance with the Chinese national standard GB/T 32743–2016. The freshly harvested leaves were thoroughly mixed and randomly divided into six groups: control (S0), blue light (S1, S1), red/blue = 1:3 (S2, R25B75), red/blue = 1:1 (S3, R50B50), red/blue = 3:1 (S4, R75B25), and red light (S5, R100). Each group comprised 1.0 kg of fresh leaves, which were evenly spread in a withering tank (30 × 50 cm; leaf-bed thickness 3 cm). An LED plant-lighting system was positioned above the withering tank and equipped with red (610–630 nm) and blue (430–440 nm) LEDs (n = 4 tubes per group), and the overall photosynthetic photon flux density was maintained at 100 μmol·m−2·s−1. The S0 group was withered with an incandescent lamp at the same photosynthetic photon flux density. All samples were withered for 36 h under well-ventilated conditions at 25 ± 2 °C and 75 ± 5% relative humidity. Then, the six groups of samples in a tea roasting machine were dried at 75 °C for 1 h and then at 105 °C for 10 min (Figure 1A). All samples were sealed and stored at −20 °C, and all analyses were completed within 3 months.

2.3. QDA

Each tea was brewed with boiling water at 100 °C at a leaf/water ratio of 1:50 (w/w) for 5 min [13]. Tea infusions, obtained after filtering to remove the leaves, were immediately placed in a cooling tank for cooling to room temperature (RT, 24 ± 1 °C). The cooled tea infusions were directly used for QDA [14]. QDA was performed in an odorless room to accurately describe and analyze the tea taste characteristics. The tea infusions were scored from 1 to 6 by nine trained panelists (six females and three males, 22–32 years old) from the College of Horticulture at South China Agricultural University at 24 ± 1 °C. Sensory attributes were evaluated in terms of sweetness, bitterness, astringency, freshness, strength and smoothness. All the panelists had obtained the certificate of China Professional Tea Evaluator (Level 3) and systematically learned about typical flavor types and descriptive vocabulary of white tea. Prior to the start of the experiment, all the experimenters received a total of more than 40 h of sensory recognition training over four weeks. A 5-point scale, adapted from Xu et al. [13,15], was used for scoring each of the above characteristics, in which 5 was “extremely strong”, 4 “strong”, 3 “neutral”, 2 “weak”, 1 “extremely weak” and 0 “none”, and the overall flavor/acceptability was also scored 0–5. Panelists rinsed their mouths with purified water between successive samples. The sniffing was repeated three times for each person in each experiment.

2.4. Metabolite Extraction and UPLC–MS/MS Analysis

2.4.1. Sample Preparation and Extraction

Samples were freeze-dried (Scientz-100F) and milled (MM 400, Retsch, Haan, Germany) with a zirconia bead for 1.5 min at 30 Hz. An amount of 50 mg of powder was extracted with 1.2 mL of 70% methanol, vortexing for 30 s every 30 min for six cycles. After centrifugation (12,000 rpm, 3 min), the supernatant was filtered (0.22 μm, ANPEL, Shanghai, China) for UPLC–MS/MS.

2.4.2. UPLC Conditions

Chromatographic separation was performed on a SHIMADZU Nexera X2 using an Agilent SB-C18 column (1.8 μm, 2.1 × 100 mm) at 40 °C, with a flow rate of 0.35 mL·min−1 and an injection volume of 4 μL. Mobile phase A was water with 0.1% formic acid, and B was acetonitrile with 0.1% formic acid. Gradient: 5–95% B (0–9 min), 95% B (9–10 min), 95–5% B (10–11.1 min), and then 5% B (11.1–14 min).

2.4.3. ESI-QTRAP-MS/MS

MS detection was carried out on an AB Sciex 4500 QTRAP equipped with an ESI source (550 °C; 5500 V positive/−4500 V negative; gas I 50 psi, gas II 60 psi, curtain gas 25 psi; CAD high). Instrument tuning and calibration were performed using 10 and 100 μmol·L−1 of polypropylene glycol (QQQ and LIT modes, respectively). Data were acquired in scheduled MRM mode with nitrogen as the collision gas (medium), and DP/CE were optimized for each transition. L-2-chlorophenylalanine (1 mg/L) was utilized as an internal standard to monitor the reproducibility of the extraction and the stability of the LC-MS system.

2.5. WGCNA

To investigate the correlation between sample taste attributes and non-volatile metabolites, the WGCNA R package on the Metware Cloud platform (Metware, Wuhan, China) was used to construct a co-expression network between metabolites and traits, clustering metabolites with similar expression patterns into modules. Stable metabolite modules were identified using hierarchical clustering and dynamic tree cutting algorithms, and module eigengenes were calculated. Furthermore, by integrating trait/group data, correlations between modules and traits/groups were computed to screen for key modules significantly associated with the traits of interest. Meanwhile, hub metabolites within each module were identified based on intra-modular connectivity and module membership, with the top 10 metabolites selected by default. All analytical procedures were performed using the R language and the WGCNA package, with soft-threshold selection following the scale-free topology criterion (R2 > 0.85), a module merging cut height set to 0.25, and a minimum module size of 50 metabolites.

2.6. HPLC Analysis

Caffeine, gallic acid, and catechin levels were measured by HPLC [16].

2.7. Statistical Analysis

Peak picking and integration were performed using MultiQuant™ 3.0.3 (SCIEX, Foster City, CA, USA). Global metabolomic variations among treatments were evaluated via unsupervised principal component analysis (PCA) using the prcomp function in R, after unit variance scaling. Orthogonal proportional latent structures–discriminant analysis (OPLS-DA) was subsequently performed using the MetaboAnalyst R package 1.0.1 (McGill University, Canada) and model robustness was validated through 200 permutation tests. Hierarchical cluster analysis (HCA) and Pearson correlation coefficients (PCCs) were generated using the ComplexHeatmap 2.14.0 (Zuguang Gu, https://github.com/jokergoo/ComplexHeatmap (accessed on 25 January 2026), European Molecular Biology Laboratory, Heidelberg, Germany) to assess metabolite accumulation patterns and sample relationships. Differential metabolites (DMs) were identified based on a combined criterion of VIP ≥ 1, p < 0.05, and fold change (FC) ≥ 2 or FC ≤ 0.5. Functional annotation and pathway enrichment were conducted by mapping DMs to the KEGG database and performing metabolite set enrichment analysis (MSEA) with hypergeometric tests.
Each independent experiment was performed in triplicate and the results were expressed as mean ± standard deviation (SD). Significant differences (p < 0.05) between means were determined by one-way analysis of variance (ANOVA), followed by Tukey’s HSD post hoc test indicated by different letters, using GraphPad Prism 8.0. (Boston, Massachusetts 02110) Data visualization, including radar plots and bar charts, was executed using OriginPro 2025 (Northampton, MA, USA) and GraphPad Prism 8.0.

3. Results

3.1. QDA of Taste Attributes

Sensory evaluation indicated that varying the red/blue light ratios during withering exerted negligible influence on the appearance and color of the finished white tea, both before and after brewing (Figure 1B). By contrast, QDA revealed that light supplement reshaped the taste by altering the relative intensities of six taste descriptors (Figure 1C). As can be seen from Figure 1D, the sweetness and freshness QDA scores of the S3 were the highest, and they were significantly higher than those of other processing groups. The sweetness and freshness QDA scores of the S4 were also significantly higher than those of the S0, S1, S2 and S5, and there was a significant difference from that of the S3. There were significant differences in the bitterness QDA scores of different treatment groups, and the order (from high to low) was S0, S5, S2, S1, S3 and S4. For the astringency, the S0 has the highest astringency QDA score. In the light supplement treatment groups, the S4 group has the lowest astringency score, followed by the S3, both of which were significantly lower than the S1 and S2. The smoothness QDA scores of the S3 and S4 were significantly higher than those of the S0 group, and also higher than those of the S2 and S5 groups. In addition, the strength of tea infusions in the light supplementation treatment groups was significantly higher than that of the S0. In general, light supplementation withering treatment can improve the taste of tea soup, and the overall taste is better under the condition that the proportion of red light is higher.

3.2. Data Quality Assessment

To ensure data reliability, analytical stability and repeatability were evaluated using quality control (QC) injections. The overlapped total ion current (TIC) of QC samples showed high consistency, indicating stable instrument performance throughout the batch (Figure S1A). In addition, more than 85% of detected features in QCs exhibited a coefficient of variation (CV) < 0.3 (Figure S1B), supporting acceptable signal variation for downstream multivariate analyses and differential metabolite screening.

3.3. Metabolite Annotation and Global Metabolomic Patterns

A total of 1118 metabolites were annotated based on UPLC–MS/MS profiling and the Metware database, dominated by flavonoids (20.30%), phenolic acids (18.43%), lipids (14.04%), amino acids and derivatives (11.45%), organic acids (8.41%), alkaloids (8.05%), and other classes (8.05%) (Figure 2A; Table S1). PCA based on all annotated metabolites explained 22.20% (PC1) and 14.42% (PC2) of the total variance. Samples clustered well within each treatment and were clearly separated across treatments in the PCA space, suggesting that withering under different red/blue ratios induces systematic shifts in the non-volatile metabolome of white tea. Notably, S4 displayed a clearer overall displacement and separation from the S0 and the other LED treatments in the PC1–PC2 score plot, indicating a more distinct global metabolomic signature under this condition (Figure 2B). HCA corroborated the PCA pattern (Figure 2C), further indicating clear discriminability among treatments at the global metabolome level with good within-group consistency.

3.4. Differential Metabolite Responses Caused by Withering of Red/Blue LED

To systematically evaluate the effects of red/blue-LED-light-supplemented withering on the non-volatile metabolic profiles of white tea, we compared the differences between the S0 group and each LED treatment group. The results showed that S0 vs. S4 exhibited the highest number of DMs (52 total: 36 up- and 16 down-regulated), followed by S0 vs. S5 (44 total:34 up- and 10 down-regulated). By contrast, fewer DEMs were identified in other treatment groups relative to the S0 (S0 vs. S1: 37; S0 vs. S2: 32; S0 vs. S3: 30) (Figure S2; Table S2). These findings indicated that under identical withering durations and environmental conditions, different red/blue light combinations exert distinct degrees of influence on metabolic remodeling, with the S4 treatment inducing the most pronounced metabolic shift compared to the S0.
Having established that different light regimes vary in the magnitude of metabolic deviation, we further investigated whether LED-assisted withering elicits a conserved core response regardless of the red/blue ratio. The intersection of DMs between the S0 group and each LED treatment group identified 18 common metabolites (17 up- and 1 down-regulated). This core set exhibited higher peak intensities in the LED-treated groups compared to the S0, primarily comprising carbohydrate- and oligosaccharide-related metabolites and their derivatives, notably D-sucrose, D-maltose, D-trehalose, isomaltulose, galactinol, raffinose/rutinose-related features, and 2′-deoxyuridine (Figure S3; Table S2).
To discern the distinct impacts of red and blue light withering on the white tea metabolome, a comparative analysis was performed between S1 and S5. The results revealed that S5 was characterized by significantly higher peak intensities of lipid-derived oxylipins and oxidized fatty acid derivatives, including 9(10)-EpOME and 12,13-epoxy-9-octadecenoic acid. This was accompanied by a marked elevation in conjugated phenolics and related metabolites, such as p-coumaroylputrescine and jasmonic acid. Conversely, various carbohydrates (such as D-sucrose and D-maltose) and specific flavonoid glycosides, notably vitexin-7-O-(6″-p-coumaroyl)glucoside, were significantly lower in S5 compared to the S1 group (Table S3).
Owing to the superior sensory attributes and significant metabolic divergence of S4, we further compared its discriminatory metabolites with S2, S3, and S5 to identify specific metabolic signatures induced by this optimal ratio. First, regarding the global magnitude of metabolic shifts, both S3 and S4 treatments exhibited a significant accumulation of sugar- and oligosaccharide-related metabolites relative to the S0, including sucrose, maltose, trehalose, isomaltulose, and galactinol (Table S2). Notably, S4 yielded a higher number of DMs compared to S3 (Figure S4). Building on this, comparative analyses of S4 against S2, S3, and S5 revealed that the peak intensities of catechin-5-O-glucoside, p-coumaroylputrescine, and N-glucosyl-p-coumaroylputrescine were significantly higher in S3 than in S4. Conversely, L-serine, benzoic acid, reduced glutathione, and tricin-7-O-guaiacylglycerol were significantly enriched in S4 compared to S3 (Table S4). Furthermore, we identified p-coumaroylputrescine, N-glucosyl-p-coumaroylputrescine, and veratric acid (3,4-dimethoxybenzoic acid) as consistent discriminatory metabolites across the three comparison pairs. These three compounds followed a uniform trend, with their peak intensities in S2, S3, and S5 being significantly higher than those in S4. Meanwhile, in the S4 vs. S0 comparison, several galloylated polyphenol-related metabolites—including strictinin, 1,3,6-tri-O-galloyl-β-D-glucose, and 3,4,5-tri-O-galloylshikimic acid—were significantly decreased in S4. Taken together, these findings distinguish the metabolic profile of S4 from those of its adjacent ratios and the S5 treatment at the metabolite level.

3.5. WGCNA Links Metabolite Modules to Taste Attributes and Treatment Groups

WGCNA was performed based on the fully annotated metabolite matrix to integrate the relationships among non-volatile metabolic profiles, sensory traits, and LED (red/blue) light treatments in white tea. Using hierarchical clustering and dynamic tree cleavage, seven co-abundance modules, including red, turquoise, yellow, green, blue, brown, and gray, were identified (Figure 3A). Correlation analysis between these modules and taste attributes revealed that the turquoise and yellow modules positively correlated with sweetness, freshness, and smoothness, but negatively with bitterness and astringency; by contrast, the blue module displayed the opposite trend (Figure 3B). This distribution aligned with the taste differentiation observed between treatments in the QDA. Correlation analysis was further performed between module eigen-metabolites and samples to investigate their associations. The results showed that the turquoise module exhibited the strongest positive correlation with S4, while showing weak or negative correlations with S0 and other treatment groups. Similarly, the yellow module was positively correlated with S4. By contrast, the blue module showed a positive correlation with the S0 group but a negative correlation with S4 (Figure 3C).
Module–trait correlation analysis indicated that the turquoise, yellow, and blue modules were closely associated with the variations induced by red/blue-LED-supplemented withering treatments in white tea. To identify potential hub metabolites within these co-expression patterns, we further screened candidates based on intra-modular connectivity (kWithin > 5). The yellow module showed clear characteristics of saccharides, nucleoside/related derivatives, including D-maltose, D-sucrose, trehalose-6-phosphate, isomaltulose, D-trehalose, galactose, 2′-deoxyuridine, melibiose, etc. Further association analysis between the yellow module and DMs showed that D-maltose, D-sucrose, D-trehalose, isomaltulose, galactose, and 2′-deoxyuridine were present in both sets. This overlap suggested that red/blue-LED-supplemented withering was characterized by a reproducible saccharide/nucleoside metabolic pattern, which exhibited a consistent phenotypic association with the enhancement of sweet and fresh sensory attributes under mixed-spectrum treatments. The turquoise module comprised a diverse array of metabolites that were highly correlated with metabolic traits, primarily concentrated in flavonoids/phenols, lysophospholipids, and nucleotides. Key hub metabolites identified through connectivity analysis included 10,16-dihydroxypalmitic acid, LysoPC 15:0, LysoPE 20:2, uridine 5′-diphosphate, and multiple flavonoid-related features such as diosmetin, apigenin, and vicenin-type C-glycosides. The enrichment of this module suggested that, alongside saccharide fluctuations, the synergistic variations in lipid-related and phenolic metabolites were closely linked to the divergent taste profiles observed under different red/blue light ratios. In contrast, the blue module was dominated by oxylipin/oxidized fatty acid derivatives, including 9-oxo-10E, 12Z-octadecadienoic acid, 17-hydroxylinolenic acid, 13 (S) -HODE, 9S-hydroxy-10E, 12Z-octadecadienoic acid, and a series of C18 polyhydroxy derivatives. The high degree of overlap between this module’s composition and the DMs indicates that different red/blue spectral treatments were associated not only with overall metabolic shifts but also with a distinct synergistic change mode in lipid oxidation-related metabolism (Table 1, Table S5).

3.6. HPLC Quantification of Caffeine, Gallic Acid, and Catechins

To complement the non-targeted metabolomics, caffeine, gallic acid, and eight major catechin monomers were quantified by HPLC. The results showed that different red/blue light ratios significantly affected the composition of catechin components. For non-ester catechins, S4 had the lowest content, followed by S3, with no significant difference between them; both were significantly lower than that of S5, while S2 had the highest content. For ester catechins, S4 had the lowest content; those of S1 and S3 were not significantly different from it, and all three were significantly lower than those of S0 and S5, with S2 having the highest content. At the monomer level, S4 had significantly lower levels of caffeine and epigallocatechin than S3 did (Figure 4).

4. Discussion

Facility spectroscopy, as a controllable environmental factor, has been applied to modulate metabolite composition in postharvest plant tissues that remain metabolically responsive [23], and light exposure during tea processing steps such as withering has been reported to reshape flavor-related metabolite profiles in final products [24]. Based on this, our research showed that red/blue LED supplemental lighting during withering was associated with a stable improvement in the sensory quality of white tea compared to the S0. The chemical basis for this was not a general increase in the intensity of a single flavor, but rather a structural reorganization of the taste-related metabolism.
Initially, the common metabolites of the red/blue LED treatment compared to the S0 were mainly saccharides and their derivatives (such as sucrose, maltose, trehalose-related signals, isomaltose, galactitol, and rutin), accompanied by small amounts of nitrogenous compounds and phenolic conjugates. Importantly, tea sensory and metabolomics evidence from water extracts has shown a direct association between increased saccharides and enhanced sweetness and palatability [25,26]. Meanwhile, research on oral interaction mechanisms clearly indicated that palatability enhancement often occurred when sweetness increased and bitterness was driven in a controlled manner [27]. Therefore, the increase in saccharide-related metabolites provided a linkable metabolic basis to explain the superior sensory performance of LED-treated white tea compared to the S0 group.
For DMs, the sensory quality of white tea processed under mixed spectra (S3, S4) was generally superior to that under monochromatic blue (B100) or red (S5) LEDs. This phenomenon was aligned with broader plant-light research indicating that combined red–blue illumination was more frequently correlated with coordinated allocation of coordinates, carbon allocation, and secondary metabolism than single wavelengths, resulting in non-additive metabolic outcomes [28,29,30]. Specifically, compared to monochromatic red light, the mixed spectrum was associated with a metabolic configuration that not only maintained higher saccharide levels but also exhibited a synergistic downward shift in pathways typically linked to bitterness and astringency. Consequently, it allowed the enhancement in sweetness and umami and the inhibition of bitterness and astringency to occur simultaneously. It is noteworthy that monochromatic red light (S5) did not form a distinctly different flavor metabolite configuration from the S0 in this study, consistent with its sensory phenotype being close to the S0, suggesting that a single light quality may have a comparatively limited capacity to facilitate the synergistic metabolic rearrangement required for optimal palatability.
Furthermore, based on sensory evaluation and WGCNA results, the superior sensory quality of S4 over S3 exhibits a strong association with established chemical drivers of bitterness and astringency. The bitterness and astringency of tea were closely linked to the composition of polyphenols, particularly galloyl catechins, which are known to interact with salivary proteins [30,31]. Our HPLC data provided independent support for these sensory differences. Specifically, S4 showed the lowest levels of non-esterified catechins, esterified catechins, total catechins, as well as caffeine and gallic acid, followed by S3, with significantly lower levels of caffeine and EGC compared to S4. Considering the understanding of the taste interaction between catechins, caffeine, and sugar in tea infusion (for example, EGCG could enhance caffeine bitterness and inhibit sucrose sweetness [32]), the simultaneous downward shift in polyphenol reactive load and caffeine bitterness load of S4 makes it easier to release sweet and umami perception at the oral level and reduce the significance of bitterness, thus forming a more stable overall advantage than S3. In addition to the quantitative analysis results of catechins and caffeine, we captured consistent directional differences in p-coumaroylputrescine, N-glucosyl-p-coumaroylputrescine, and veratric acid in key comparisons distinguishing S4 from adjacent ratios/monochromatic red light (S2/S3/S5 were significantly higher than S4). Phenolic amides were clearly defined as inducible conjugates of phenylpropanoid metabolites and polyamines, and were a class of metabolic markers highly sensitive to environmental signals and stress responses [33]. Simultaneously, the interaction between phenolic acids/flavonoids and caffeine in tea infusion could jointly shape the bitterness/astringency intensity and sensory weight [32]. Therefore, the relatively lower abundance of metabolites in the phenylpropanoid conjugation/phenolic acid discriminant branch characterized the S4 metabolic profile. This lower abundance paralleled the reduced bitterness perception and aligned with the QDA bitterness ranking.

5. Conclusions

Under the experimental conditions of this study, we demonstrated that red-blue LED-assisted withering could serve as an effective and controllable strategy for improving the sensory quality of white tea by remodeling taste-related metabolic networks rather than simply increasing overall metabolite abundance. Compared with the control and single-wavelength treatments, mixed red–blue light was associated with a coordinated metabolic redistribution characterized by enriched saccharides and concurrent suppression of key bitterness- and astringency-driving compounds. Among all tested light regimes, S4 exhibited the most balanced sensory profile. This advantage was not attributable to any single metabolite but was supported by evidence from QDA, non-targeted metabolomics, WGCNA, and targeted HPLC quantification. Specifically, S4 was linked to a saccharides-centered network module positively associated with sweetness and umami, along with a significant reduction in catechins and caffeine, which were established chemical drivers of bitterness and astringency. In subsequent research, we will further investigate the effects of LED light at different wavelengths on tea quality across various cultivars and seasons. This work will provide a scientific basis for standardizing white tea processing to enhance sensory quality.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/foods15050836/s1, Figure S1: Data quality assessment of UPLC-MS/MS; Figure S2: Differential metabolite analysis of red/blue-LED-supplemented withering treatments vs. control; Figure S3: A total of 18 common DMs across red/blue LED supplemented withering treatments vs. control; Figure S4: Comparison of differential metabolites between key treatment groups; Table S1: All metabolites detected by UPLC-MSMS; Table S2: Differential metabolites between all treatment groups and the control; Table S3: Differential metabolites between S1 and S5; Table S4: Differential metabolites between S3 and S4; Table S5: Association between WGCNA modules and metabolites.

Author Contributions

Conceptualization, visualization, writing—original draft preparation, funding acquisition, methodology, data curation, D.W.; formal analysis, validation, Y.D.; investigation, J.X.; investigation, L.W.; investigation, J.M.; project administration, supervision, D.D.; resources, conceptualization, funding acquisition, supervision, F.C. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Zhejiang Provincial Department of Education (grant no. Y202455649) and the Research Development Foundation of Zhejiang A&F University (grant no. 2024FR019). The funders did not have any role in the study design, data collection, or data analysis.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy reasons.

Conflicts of Interest

The authors declare that they have no competing interests.

Abbreviations

The following abbreviations were used in this manuscript:
ANOVAOne-way analysis of variance
Ccatechin
CADCollision-activated dissociation
CECollision energy
CGCatechin gallate
CVCoefficient of variation
DMsDifferential metabolites
DPDeclustering potential
ECEpicatechin
ECGEpicatechin gallate
EGCEpigallocatechin
EGCGEpigallocatechin gallate
ESIElectrospray ionization
FCFold change
GCGallocatechin
GCGGallocatechin gallate
HCAHierarchical cluster analysis
HPLCHigh-performance liquid chromatography
LEDLight-emitting diode
MRMMultiple reaction monitoring
MSEAMetabolite set enrichment analysis
OPLS-DAOrthogonal proportional latent structures–discriminant analysis
PC 1Principal component 1
PCAPrincipal component analysis
PCCsPearson correlation coefficients
PIFPhytochrome-interacting factor
QCQuality control
QDAQuantitative descriptive analysis
QTRAPQuadrupole-linear ion trap
RTRoom temperature
TICTotal ion current
UPLC–MS/MSUltra-performance liquid chromatography–tandem mass spectrometry
VIPVariable importance in projection
VOCsVolatile organic compounds

References

  1. Wang, J.; Lin, D.; Feng, J.; Ye, S.; Tian, W.; Kong, W.; Zhang, W.; Chu, Q.; Yu, B.; Lin, H.; et al. Airflow: A double-edged sword in shaping white tea quality. Food Chem. X 2025, 28, 102558. [Google Scholar] [CrossRef] [Scilit]
  2. Zhang, L.; Cao, Q.-Q.; Granato, D.; Xu, Y.-Q.; Ho, C.-T. Association between chemistry and taste of tea: A review. Trends Food Sci. Technol. 2020, 101, 139–149. [Google Scholar] [CrossRef] [Scilit]
  3. Zou, L.; Shen, S.; Wei, Y.; Jia, H.; Li, T.; Yin, X.; Lu, C.; Cui, Q.; He, F.; Deng, W.; et al. Evaluation of the effects of solar withering on nonvolatile compounds in white tea through metabolomics and transcriptomics. Food Res. Int. 2022, 162, 112088. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  4. Liu, Y.; Singh, S.K.; Pattanaik, S.; Wang, H.; Yuan, L. Light regulation of the biosynthesis of phenolics, terpenoids, and alkaloids in plants. Commun. Biol. 2023, 6, 1055. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  5. Jones, M.A. Using light to improve commercial value. Hortic. Res. 2018, 5, 47. [Google Scholar] [CrossRef] [Scilit]
  6. Xu, D. COP1 and BBXs-HY5-mediated light signal transduction in plants. New Phytol. 2020, 228, 1748–1753. [Google Scholar] [CrossRef] [Scilit]
  7. Gangappa, S.N.; Botto, J.F. The Multifaceted Roles of HY5 in Plant Growth and Development. Mol. Plant 2016, 9, 1353–1365. [Google Scholar] [CrossRef] [Scilit]
  8. Wu, W.; Wu, H.; Liang, R.; Huang, S.; Meng, L.; Zhang, M.; Xie, F.; Zhu, H. Light regulates the synthesis and accumulation of plant secondary metabolites. Front. Plant Sci. 2025, 16, 1644472. [Google Scholar] [CrossRef] [Scilit]
  9. Li, Y.; He, C.; Yu, X.; Zhou, J.; Ntezimana, B.; Yu, Z.; Chen, Y.; Ni, D. Study on improving aroma quality of summer-autumn black tea by red-light irradiation during withering. Lwt 2022, 154, 112597. [Google Scholar] [CrossRef] [Scilit]
  10. Xie, J.; Wang, Q.; Hu, J.; Wang, L.; Yu, X.; Yuan, H.; Jiang, Y.; Yang, Y. Uncovering the effects of spreading under different light irradiation on the volatile and non-volatile metabolites of green tea by intelligent sensory technologies integrated with targeted and non-targeted metabolomics analyses. Food Chem. 2025, 463, 141482. [Google Scholar] [CrossRef] [Scilit]
  11. Ai, Z.; Zhang, B.; Chen, Y.; Yu, Z.; Chen, H.; Ni, D. Impact of light irradiation on black tea quality during withering. J. Food Sci. Technol. 2017, 54, 1212–1227. [Google Scholar] [CrossRef] [Scilit]
  12. Li, Q.; He, J.; Yu, X.; Wu, Y.; Li, K.; Wu, Z.; Zheng, Y.; Sun, Y.; Lin, H. Exploring the taste code of light-withered tea: The mechanism of nonvolatile metabolite changes in LED light-withered oolong tea (Camellia sinensis). Food Chem. X 2025, 30, 102873. [Google Scholar] [CrossRef] [Scilit]
  13. Deng, S.; Cao, Q.-Q.; Zhu, Y.; Wang, F.; Chen, J.-X.; Zhang, H.; Granato, D.; Liu, X.; Yin, J.-F.; Xu, Y.-Q. Effects of natural spring water on the sensory attributes and physicochemical properties of tea infusions. Food Chem. 2023, 419, 136079. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  14. Wu, H.; Chen, Y.; Feng, W.; Shen, S.; Wei, Y.; Jia, H.; Wang, Y.; Deng, W.; Ning, J. Effects of Three Different Withering Treatments on the Aroma of White Tea. Foods 2022, 11, 2502. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  15. Cao, Q.-Q.; Wang, F.; Wang, J.-Q.; Chen, J.-X.; Yin, J.-F.; Li, L.; Meng, F.-K.; Cheng, Y.; Xu, Y.-Q. Effects of brewing water on the sensory attributes and physicochemical properties of tea infusions. Food Chem. 2021, 364, 130235. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  16. Wu, D.; Chen, R.; Zhang, W.; Lai, X.; Sun, L.; Li, Q.; Zhang, Z.; Cao, J.; Wen, S.; Lai, Z.; et al. Tea and its components reduce the production of uric acid by inhibiting xanthine oxidase. Food Nutr. Res. 2022, 66, 10–29219. [Google Scholar] [CrossRef] [Scilit]
  17. Yin, Q.-C.; Ji, J.-B.; Zhang, R.-H.; Duan, Z.-W.; Xie, H.; Chen, Z.; Hu, F.-C.; Deng, H. Identification and verification of key taste components in wampee using widely targeted metabolomics. Food Chem. X 2022, 13, 100261. [Google Scholar] [CrossRef] [Scilit]
  18. Umino, M.; Onozato, M.; Sakamoto, T.; Koishi, M.; Fukushima, T. Analyzing Citramalic Acid Enantiomers in Apples and Commercial Fruit Juice by Liquid Chromatography-Tandem Mass Spectrometry with Pre-Column Derivatization. Molecules 2023, 28, 1556. [Google Scholar] [CrossRef] [Scilit]
  19. Singh, N.; Kathuria, D.; Barthwal, R.; Joshi, R. Metabolomics of chemical constituents as a tool for understanding the quality of fruits during development and processing operations. Int. J. Food Sci. Technol. 2024, 59, 4169–4184. [Google Scholar] [CrossRef] [Scilit]
  20. Sun, Y.; Zhang, S.; Bao, T.; Jiang, Z.; Huang, W.; Xu, X.; Qiu, Y.; Lei, P.; Wang, R.; Xu, H.; et al. Comprehensive New Insights into Sweet Taste Transmission Mechanisms and Detection Methods. Foods 2025, 14, 2397. [Google Scholar] [CrossRef] [Scilit]
  21. Figueroa, C.M.; Lunn, J.E. A Tale of Two Sugars: Trehalose 6-Phosphate and Sucrose. Plant Physiol. 2016, 172, 7–27. [Google Scholar] [CrossRef] [Scilit]
  22. Sajjad, M.; Xue, S.; Zhou, M.; Li, G.; Xu, Y.; Liu, L.; Zhu, J.; Meng, Q.; Jin, Q.; Du, H.; et al. Decoding comparative taste and nutrition regulation in Chinese cabbage via integrated metabolome and transcriptome analysis. Food Res. Int. 2024, 195, 114943. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  23. Liu, X.; Shi, R.; Gao, M.; He, R.; Li, Y.; Liu, H. Growth of tomato and cucumber seedlings under different light environments and their development after transplanting. Front. Plant Sci. 2023, 14, 1164768. [Google Scholar] [CrossRef] [Scilit]
  24. Zheng, C.; Ma, J.-Q.; Ma, C.-L.; Shen, S.-Y.; Liu, Y.-F.; Chen, L. Regulation of Growth and Flavonoid Formation of Tea Plants (Camellia sinensis) by Blue and Green Light. J. Agric. Food Chem. 2019, 67, 2408–2419. [Google Scholar] [CrossRef] [Scilit]
  25. Li, W.; Zhang, Z.; Chen, R.; Sun, L.; Lai, X.; Li, Q.; Hao, M.; Zhang, S.; Li, Q.; Sun, S.; et al. Metabolomics-based analysis of the effects of differences in soluble sugars on the sweetness quality of six major tea types in China. Food Funct. 2025, 16, 3707–3720. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  26. Wu, H.; Huang, W.; Chen, Z.; Chen, Z.; Shi, J.; Kong, Q.; Sun, S.; Jiang, X.; Chen, D.; Yan, S. GC-MS-based metabolomic study reveals dynamic changes of chemical compositions during black tea processing. Food Res. Int. 2019, 120, 330–338. [Google Scholar] [CrossRef] [Scilit]
  27. Silva, M.S.; García-Estévez, I.; Brandão, E.; Mateus, N.; de Freitas, V.; Soares, S. Molecular Interaction Between Salivary Proteins and Food Tannins. J. Agric. Food Chem. 2017, 65, 6415–6424. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  28. Hwang, H.; Lim, Y.; Oh, M.-M.; Choi, H.; Shim, D.; Song, Y.H.; Cho, H. Spatiotemporal bifurcation of HY5-mediated blue-light signaling regulates wood development during secondary growth. Proc. Natl. Acad. Sci. USA 2024, 121, e2407524121. [Google Scholar] [CrossRef] [Scilit]
  29. Hu, G.; Li, X.; Yang, J.; Yuan, Q.; Yang, S.; Fu, W.; Zhang, X.; Li, Y.; Shen, Z.; Jiang, J. Effects of Photoperiod and Light Quality on Germination and Growth of Camellia sinensis ‘HuangKui’. Plants 2024, 13, 1782. [Google Scholar] [CrossRef] [Scilit]
  30. Herrera, M.A.; Caldeira da Silva, C.C.; Von Dentz, M.; Baptista, M.S.; Kowaltowski, A.J. Mitochondrial fatty acid oxidation is stimulated by red light irradiation. FEBS Lett. 2025, 600, 20–38. [Google Scholar] [CrossRef] [Scilit]
  31. Yue, C.; Wang, Z.; Peng, H.; Jiang, L.; Yang, P.; Li, W. Analysis of Taste Quality Differences Between High and Low Grades of Ninghong Tea: From the Perspective of Sensory, Metabolite, and Taste Activity Values. Foods 2024, 13, 3957. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  32. Chen, Y.-H.; Zhang, Y.-H.; Chen, G.-S.; Yin, J.-F.; Chen, J.-X.; Wang, F.; Xu, Y.-Q. Effects of phenolic acids and quercetin-3-O-rutinoside on the bitterness and astringency of green tea infusion. npj Sci. Food 2022, 6, 8. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  33. King, B.M.; Arents, P.; Bouter, N.; Duineveld, C.A.A.; Meyners, M.; Schroff, S.I.; Soekhai, S.T. Sweetener/sweetness-induced changes in flavor perception and flavor release of fruity and green character in beverages. J. Agric. Food Chem. 2006, 54, 2671–2677. [Google Scholar] [CrossRef] [Scilit] [PubMed]
Figure 1. QDA sensory evaluation results under varying red and blue light ratios. (A) White tea processed under varying red and blue light proportions. S0, the incandescent lamp control group; S1, the blue LED light group; S2, the mixed light (red:blue = 1:3) withering treatment group; S3, the mixed light (red:blue = 1:1) withering treatment group; S4, the mixed light (red:blue = 3:1) withering treatment group; S5, the red LED light group. All the groups were established under a uniform total light intensity of 100 μmol·m−2·s−1. (B) QDA sensory evaluation sample morphology. (C) Taste attribute radar chart. (D) Taste attribute histogram. The bars were arranged in descending order of value, with the highest value labeled ‘a’. Groups sharing a common letter were not significantly different, whereas groups with no common letters were significantly different (p < 0.05).
Figure 1. QDA sensory evaluation results under varying red and blue light ratios. (A) White tea processed under varying red and blue light proportions. S0, the incandescent lamp control group; S1, the blue LED light group; S2, the mixed light (red:blue = 1:3) withering treatment group; S3, the mixed light (red:blue = 1:1) withering treatment group; S4, the mixed light (red:blue = 3:1) withering treatment group; S5, the red LED light group. All the groups were established under a uniform total light intensity of 100 μmol·m−2·s−1. (B) QDA sensory evaluation sample morphology. (C) Taste attribute radar chart. (D) Taste attribute histogram. The bars were arranged in descending order of value, with the highest value labeled ‘a’. Groups sharing a common letter were not significantly different, whereas groups with no common letters were significantly different (p < 0.05).
Foods 15 00836 g001
Figure 2. UPLC-MS/MS metabolite annotation and classification. (A) Donut chart; (B) PCA; (C) HCA.
Figure 2. UPLC-MS/MS metabolite annotation and classification. (A) Donut chart; (B) PCA; (C) HCA.
Foods 15 00836 g002
Figure 3. Modules feature analysis of WGCNA between all annotated metabolites and taste attributes/groups. (A) Cluster dendrogram; (B) module–taste trait relationships between module and taste attributes; (C) module–group trait relationships between module and groups.
Figure 3. Modules feature analysis of WGCNA between all annotated metabolites and taste attributes/groups. (A) Cluster dendrogram; (B) module–taste trait relationships between module and taste attributes; (C) module–group trait relationships between module and groups.
Foods 15 00836 g003
Figure 4. HPLC results of catechin monomers, gallic acid, and caffeine. (A) Gallic acid content; (B) caffeine content; (CJ) GC, EGC, C, EC, ECG, GCG, EGCG, and CG contents, respectively; (K) non-ester catechin content; (L) ester catechin content. Different lowercase letters represent different significance levels.
Figure 4. HPLC results of catechin monomers, gallic acid, and caffeine. (A) Gallic acid content; (B) caffeine content; (CJ) GC, EGC, C, EC, ECG, GCG, EGCG, and CG contents, respectively; (K) non-ester catechin content; (L) ester catechin content. Different lowercase letters represent different significance levels.
Foods 15 00836 g004
Table 1. Top 10 hub metabolites of turquoise, yellow, and blue modules.
Table 1. Top 10 hub metabolites of turquoise, yellow, and blue modules.
ModuleCompoundkWithinClassCAS. No.Taste Descriptor
YellowD-Glucose 6-phosphate *42.08 Saccharides56-73-5Sweet [17]
L-Citramalic acid36.89 Organic acids6236-09-5Sour [17,18]
Glucose-1-phosphate *34.89 Saccharides59-56-3Sweet [19]
L-Alanyl-L-Phenylalanine34.28 Amino acids and derivatives3061-90-3-
Uridine 5′-diphospho-D-glucose33.22 Nucleotides and derivatives133-89-1-
D-Maltose *31.00 Saccharides69-79-4Sweet [20]
Galactinol30.14 Saccharides3687-64-7-
Trehalose 6-phosphate29.41 Saccharides4484-88-2Regulating sucrose accumulation [21]
2′-Deoxyuridine28.57 Nucleotides and derivatives951-78-0-
D-Sucrose *28.54 Saccharides57-50-1Sweet [20]
Turquoise10,16-Dihydroxypalmitic acid28.34 Lipids3233-90-7-
LysoPC 15:027.52 Lipids108273-89-8-
Diosmetin27.49 Flavonoids520-34-3-
6,7,8-Tetrahydroxy-5-methoxyflavone *27.33 Flavonoids--
9,16-Dihydroxypalmitic acid26.44 Lipids38076-49-2-
Apigenin; 4′,5,7-Trihydroxyflavone26.38 Flavonoids520-36-5-
LysoPE 20:2(2n isomer)25.32 Lipids--
LysoPE 20:224.53 Lipids--
Uridine 5′-diphosphate23.63 Nucleotides and derivatives27821-45-0Related to taste quality [22]
2-Hydroxycinnamic acid *22.96 Phenolic acids583-17-5-
Blue9-Oxo-10E,12Z-octadecadienoic acid29.45 Lipids54232-59-6-
17-Hydroxylinolenic acid29.33 Lipids--
9,10,13-Trihydroxy-11-Octadecenoic Acid28.71 Lipids29907-57-1-
13(S)-HODE28.70 Lipids10219-69-9-
9S-Hydroxy-10E,12Z-octadecadienoic acid *28.64 Lipids15514-85-9-
9,10,11-Trihydroxy-12-octadecenoic acid28.26 Lipids61911-67-9-
9,12,13-Trihydroxy-10,15-octadecadienoic acid28.19 Lipids--
9,12,13-TriHOME26.79 Lipids97134-11-7-
Dodecanedioic acid26.49 Lipids693-23-2-
1-Stearidonoyl-Glycerol26.16 Lipids--
Footnote: Asterisk (*) indicated that there were isomers of this metabolite, and the mass spectrometry cannot completely distinguish these isomers under the current detection conditions. Therefore, multiple isomers with co-elution or highly similar secondary mass spectrometry characteristics were reported together and marked with *.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Wu, D.; Deng, Y.; Xing, J.; Wen, L.; Ma, J.; Dong, D.; Cao, F. Taste Modulation of White Tea by Red/Blue-LED-Assisted Withering Revealed via Non-Volatile Metabolomics. Foods 2026, 15, 836. https://doi.org/10.3390/foods15050836

AMA Style

Wu D, Deng Y, Xing J, Wen L, Ma J, Dong D, Cao F. Taste Modulation of White Tea by Red/Blue-LED-Assisted Withering Revealed via Non-Volatile Metabolomics. Foods. 2026; 15(5):836. https://doi.org/10.3390/foods15050836

Chicago/Turabian Style

Wu, Dan, Yongyi Deng, Jiabao Xing, Lianghua Wen, Jiawei Ma, Dubin Dong, and Fanrong Cao. 2026. "Taste Modulation of White Tea by Red/Blue-LED-Assisted Withering Revealed via Non-Volatile Metabolomics" Foods 15, no. 5: 836. https://doi.org/10.3390/foods15050836

APA Style

Wu, D., Deng, Y., Xing, J., Wen, L., Ma, J., Dong, D., & Cao, F. (2026). Taste Modulation of White Tea by Red/Blue-LED-Assisted Withering Revealed via Non-Volatile Metabolomics. Foods, 15(5), 836. https://doi.org/10.3390/foods15050836

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop