Skip to main content
NCBI home page
Food Chemistry: X logoLink to Food Chemistry: X
. 2025 Apr 30;27:102507. doi: 10.1016/j.fochx.2025.102507

Characterization of the changes of aroma profiles in large-leaf yellow tea during processing using GC–MS and electronic nose analysis

Xiangyang Guo a,b,c,d,g, Wilfried Schwab b,c,e, Chi-Tang Ho b,c,f,⁎⁎, Chuankui Song b,c, Xiaochun Wan b,c,⁎
PMCID: PMC12131246  PMID: 40463654

Abstract

Tea processing is vital for aroma formation of large-leaf yellow tea (LYT), yet the volatile compounds during its processing remain poorly understood. Volatile profiles of LYT throughout the processing were characterized by gas chromatography–mass spectrometry and electronic nose (E-nose). Distinct differences in volatile compositions were observed among stages, with 178 volatiles identified, dominated by alcohols during processes. Notably, heterocyclic compounds became prominent after full fire processing, during which 30 N-containing compounds were newly generated, making a significant shift in the volatile profile. Furthermore, tea samples can be distinguished based on the identified volatiles or the responses of E-nose, implying that E-nose is suitable for the detection of volatiles in tea production. Additionally, the aroma-active compounds of LYT were widely different from those of other types of yellow teas, attributed to the unique critical processes involved. These results provide valuable theoretical guidance for the processing and quality control of LYT.

Keywords: Large-leaf yellow tea (LYT), Electronic nose (E-nose), Gas chromatography-mass spectrometry (GC–MS), Tea processing, Volatile compounds, Aroma profiles

Highlights

  • •

    Volatile compositions of LYT during the whole processing were analyzed.

  • •

    Aroma profiles of LYT during processing were revealed by GC–MS and E-nose.

  • •

    Thirty N-containing volatiles were newly formed throughout processing.

  • •

    Full fire processing was crucial for the final aroma formation of LYT.

  • •

    Key odorants were widely different among teas with different critical procedures.

1. Introduction

Tea (Camellia sinensis) is popularly consumed and highly appreciated for its potential health benefits and pleasant flavor (Ho et al., 2015). Among them, large-leaf yellow tea (LYT) has attracted more and more attention and behaved great commercial value in tea market recently. This is mainly attributed to its outstanding health-promoting benefits (Chen et al., 2024; Ge et al., 2024; Wang et al., 2021; Zhao et al., 2023) and a characteristic and unique fried rice-like odor (Guo et al., 2019; Guo et al., 2021a).

Tea processing is crucial to the formation of the characteristic flavor of LYT, which involves the steps of fixing, rolling, yellowing, roasted and full fire processing, commonly known as ‘La Laohuo’ in Chinese with a high intensity of roasting (Li et al., 2024). Yellowing is essential for the presentation of typical yellow color of tea leaves and infusions in the production of yellow tea, which has a major impact on the variations of volatiles, thereby affecting the aroma properties and flavor quality (Wei et al., 2020; Wei et al., 2021). Nevertheless, the stable flavor of LYT is formed within full fire processing, especially through the procedure involving old fire roasting (deep firing) treatment, during which a large number of heterocyclic compounds, exhibiting roasted or nutty flavor with pyrazine, pyrrole, furan and pyran structures are produced (Guo et al., 2021a; Yin et al., 2023). Previously published studies focused on the effects of roasting processing on flavor quality of LYT (Guo et al., 2021a; Li et al., 2023; Sheng et al., 2024) or the analysis of volatile compositions of LYT (Guo et al., 2019; Zhai et al., 2023). However, to the best of our knowledge, hardly any research reported the effects of tea processing for tea aroma of LYT, either in volatile compounds or in aroma properties, and their variations throughout the processes.

Tea aroma is the critical determinant of tea quality (Yang et al., 2013). Generally, tea aroma can significantly affect the desire and provide an important means for consumers to drink tea, and influence the potential possibility of repeated consumption. The source of the tea aroma is refered to numerous volatile compounds, and their compositions in a certain proportions, as present in tea. The volatile compounds are generated during tea production and can be easily varied during tea processes, affecting the tea aroma profiles. Moreover, the influence of individual steps on tea aroma is different. The characteristic aroma of the final product tea is the result and a comprehensive reflection of biochemical reactions and/or thermal reactions during processing (Cho et al., 2007). Therefore, it is of great importance to study the effects of overall processing on tea aroma of LYT, and to reveal the variations of volatile compounds during processing.

During tea production, the rapid aroma evaluation and off-flavor monitoring are important for the accurate judgement of time interval and intensity of each process, which closed relates to the quality control and improvement of final product tea. Traditionally, the evaluation of tea quality and tea aroma in particular is based on human sensory assessment (Zhu et al., 2017), which mainly relays on the experienced tea evaluators to identify or perceive tea samples and then give the relevant evaluation scores of aroma attributes. Nevertheless, such sensory evaluations are inevitably subjective due to the personal preference, ability, physical and psychological factors of evaluators (Zhi et al., 2017). Meanwhile, it is a long-term and continuous process to train the professional assessors, who are key to obtain accurate evaluation results. Thus, it is not easy to determine the quality of tea by sensory assessment. Moreover, it is impossible to score the aroma attributes of tea sample from each process in practical industrial applications, only the final product teas will be assessed. Therefore, it is necessary to seek a new method to cover the insufficient assessment of tea, and can quickly response the aroma changes of tea samples during processing, as well as to provide information of the variations on volatile compounds and aroma profiles.

Currently, gas chromatography–mass spectrometry/olfactometry (GC–MS/O) is the most commonly used instrumental analytical technology for detecting of tea aroma compounds and for evaluating of tea quality (Ma et al., 2017). The results obtained by GC–MS/O can provide sufficient information of volatiles and aroma profiles in tea samples during processing. However, it is not appropriate for the rapid characterization of aroma compounds in tea because of the required complex sample pretreatment, long time of detection and data processing (Song & Liu, 2018). In comparison, electronic nose (E-nose) is the preferred method for aroma compounds analysis, which is characterized by simplicity, high efficiency, low-cost, real-time detection and fast assessment (Peris & Escuder-Gilabert, 2009). The E-nose instrument consists of a sampling system, sensor array and signal processing system, among which the gas sensor array is the core part, determining the rapid response ability, accuracy and reproducibility of system. Of all the sensors available recently, metal oxide sensors are the most popular and commonly used in E-nose system for its high sensitivity and low cost (Srivastava & Sadistap, 2018). It can non-destructively mimic the human olfactory organ for online recognition, detection, and analysis of volatile components in samples (Rao et al., 2020), and is widely applied in tea variety and grade discrimination, as well as quality assessment (Xu et al., 2019; Yuan et al., 2019).

Up till now, E-nose technology has not been employed to detect tea samples during processing, nor has it been used to monitor and analyze the changes of aroma properties of LYT during processing. The current work mainly aims to reveal the aroma properties of LYT during processing by GC–MS in conjunction with E-nose technology, the volatile compositions and aroma profiles of LYT from fresh tea leaves to the final product tea were characterized. Meanwhile, we want to evaluate the potential of using E-nose technology to analyze and monitor the variations of aroma profiles of LYT at different processing stages.

2. Material and methods

2.1. Tea material

Fresh tea leaves (cultivar ‘population’) were plucked at the tea garden of Huibinyi Tea Co., Ltd. (Lu’an City, Anhui, China) in May of 2017. The fresh tea leaves (FTL-LYT) were subjected to a series of conventional LYT processing, including fixing (Fx-LYT), rolling (Ro-LYT), primary roasted (PR-LYT), primary yellowing (PY-LYT), re-roasted (RR-LYT), re-yellowing (RY-LYT) and full fire processing (FF-LYT), which is illustrated in Fig. 1A. The tea materials used in this work are the same batch as those used in the previous publication (Guo et al., 2019), and the detail parameters of LYT processing are listed in Table S1. Tea samples from FTL-LYT to the final tea products after full fire processing treatment were freeze-dried and ground into powder using an ultra-mill (Shanghai Dianjiu Chinese Machinery Manufacturing Co., Ltd., Shanghai, China) to pass through a 0.35 mm sieve.

Fig. 1.

Fig. 1

Open in a new tab

Aroma rador plot of LYT during processing by E-nose.

The E-nose detects volatiles using 10 different metal oxide gas sensors (labeled S1 to S10). (A) FTL-LYT & FF-LYT. (B) Fx-LYT & Ro-LYT. (C) PR-LYT & PY-LYT. (D) RR-LYT & RY-LYT. Sensors labeled with ‘*’ had significant difference (p< 0.05) in response values between two samples.

2.2. Chemicals

The authentic standards of volatile compounds and internal standard compound (ethyl decanoate, 545 mg/kg in diethyl ether) were purchased from Sigma-Aldrich (Shanghai, China) unless specified otherwise. The authentic aroma compounds and their suppliers are list in Table S2. Sodium sulfate (analytical grade) was obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Diethyl ether (Sinopharm Chemical Reagent Co., Ltd.) was distilled prior to use.

2.3. Tea aroma extraction by steam distillation (SD)

The SD method of tea aroma extraction was followed to Guo et al. (2019) and Clevenger apparatus was applied for extracting tea aroma. For SD assay, previously homogenized tea powder (15.0 g) was placed into a 500 mL flask. After the addition of distilled water (300 mL) and ethyl decanoate (10 μL, 545 ppm) as internal standard, the distillation was conducted for 1.5 h. Then, the distillate was transferred to a separatory funnel after cooling and the upper diethyl ether layer (10 mL) was dried (4 °C, 12 h) using anhydrous sodium sulfate, subsequently concentrated under a nitrogen stream to 1 mL before GC–MS analysis.

2.4. GC–MS analysis of volatile composition in tea

The volatile analysis was conducted using an Agilent 7890 A GC coupled to a 5975C MS (Agilent, Santa Clara, CA, USA), and the GC–MS analytical procedure was performed according to Guo et al. (2019). The separation was executed on a fused-silica capillary column (30 m × 0.25 mm × 0.25 μm, HP-5 MS, J&W, Folsom, CA, USA). Helium gas (37 cm/s) was used as a mobile phase. The temperature programming was initially set at 50 °C for 5 min, increased to 210 °C at 3 °C/min, and held for 3 min, then raised to 230 °C at 10 °C/min and subsequent maintained at 230 °C for 2 min. Injection volume was 1 μL with a split ratio of 5:1 for SD samples. The parameters of the mass spectrometer were an electron impact ion source (EI ion source) ionized at an electron energy of 70 eV, with an ion source temperature of 250 °C and a transfer-line temperature of 150 °C. The mass spectra were acquired in full scan mode from 30 to 500 amu. Retention indices (RI) were calculated from the retention time of n-alkanes (C5-C28) by linear interpolation.

2.5. Electronic nose (E-nose) analysis

A PEN3 portable E-nose (Airsense Analytics, Germany) was applied in the research to acquire aroma properties of LYT samples. The E-nose is composed of a sampling apparatus, an array of sensors and data analysis software (Win Muster version 3.0), of which the sensor array contains 10 different metal oxide gas sensors (labeled S1 to S10). The main responded volatile chemical components of the array of sensors used in PEN3 are shown in Table S3.

The preparation of tea samples was conducted according to the method from Xu et al. (2019) with minor modification. Briefly, 3.0 g of tea samples were placed into a 150 mL beaker, and was sealed with parafilm for 30 min to give sufficient time for the release of volatile compounds in the headspace to reach a dynamic balance. After that, the parafilm was punctured using a 2.5 mL syringe and injected the gas 4–5 times to disperse the volatile compounds evenly. Subsequently, 2 mL of the sample gas was pumped into the sensor chamber at a flow rate of 150 mL/min. The detection time and cleaning time was 100 s, 120 s, respectively. The LYT samples during the whole processing were detected by E-nose at a temperature of 25 ± 1 °C. The typical response signals of sensors in E-nose for LYT sample was shown in Fig. S1. Each tea sample was tested three times to obtain the average of the sensor responses for plotting the radar fingerprint of LYT during processing.

2.6. Statistical analysis

One-way analysis of variance (ANOVA) and Duncan's multiple-range tests were used to find the significant difference (p-value <0.05) between samples. The principal component analysis (PCA) was performed using SPSS Statistics 22 (SPSS Inc., Chicago, IL, USA). All the determination were performed in triplicate (n = 3).

3. Results and discussion

3.1. Volatile profiling of LYT during processing determined by E-nose

Significant difference was uncovered in the fingerprint of the eight tea samples, especially between the samples of FTL-LYT and FF-LYT (Fig. 1B), as well as PR-LYT and PY-LYT (Fig. 1D). The largest response value was found in S7, followed by S2 and S9 in all the tea samples. In addition, it can be seen from each radar fingerprint that the greater response values of S7, S2, S9 and S6 were recorded in tea samples with heat treatment (FF-LYT, Fx-LYT, PR-LYT and RR-LYT) in comparison to the non-thermal treatment samples (FTL-LYT, Ro-LYT, PY-LYT and RY-LYT).

To more clearly present and further distinguish the differences of LYT samples during processing, the principal component analysis (PCA) was conducted on the responses of the sensors of the E-nose. PCA plots were presented as a two-dimensional scatter plot using the first two principal component score vectors (Fig. 2A). The cumulative contribution rate of the first two principal components (PC1 of 78.4 %, PC2 of 17.9 %) reached 96.3 %, which represented the major characteristics of tea samples. The LYT samples appeared to be well separated according to the tea processing except for the partial overlap of Ro-LYT and PY-LYT, RR-LYT and RY-LYT samples, consistent with the results presented in the fingerprint of the E-nose, implying a rapid distinction for tea samples during processing. Moreover, the distance between FTL-LYT and FF-LYT was very large, indicating great differences in aroma properties of the two samples. As shown in Fig. 2B, the bi-plot on the PC1 and PC2 was plotted to combine the score plots with the loadings plot. It was found that the correlation of FF-LYT sample with sensors of S2, S8, S7 and S9, which are primarily sensitive to nitrogen oxides and aromatic compounds, helped to localize the final product teas apart from the other processing samples. This is consistent with the results from Guo et al. (2019), showing that the N-containing heterocyclic compounds and aromatics are the major volatiles in LYT.

Fig. 2.

Fig. 2

Open in a new tab

PCA analysis of LYT during processing and sensor contribution rate analysis based on responses of E-nose.

The score plot (A) and biplot (B) of PCA analysis based on responses of E-nose. Compound nos. Correspond to sensor number. (C) The sensor contribution rate analysis.

The sensor contribution rate was analyzed using the software supplied by the manufacturer (Fig. 2C). All the 10 sensors made different contributions, among which S7 contributed the most to main axis 1, and S6 had the greatest contribution to main axis 2. Aside from this, S2, S9, or S8 also made high contribution to main axis1 or 2, whereas, the lowest contributions of the sensors were observed from S4 and S10. Definitely, all these sensors with high contribution rates are primarily sensitive to nitrogen oxides, aromatics, or chemical compounds including amines. Therefore, the greatest aroma difference in LYT samples during processing mainly was attributed to nitrogen oxides and aromatic compounds.

3.2. Volatile profiling of LYT during processing determined by GC–MS

Totally 178 volatiles were identified in tea samples of LYT during processing by the database search in the NIST 2017 library, coupled to the retention time, retention index, in-house database, and literature data (Guo et al., 2019; Guo et al., 2021a). Whenever possible, volatiles were positively identified using authentic compounds, as well as using advanced peak deconvolution and data processing software to reveal the overlapping compounds to analyze and achieve the volatile compounds accurately. The relative concentrations of volatiles were quantified based on the peak areas of the internal standard. Detailed information of volatile compounds identified in LYT during processing and odor qualities are listed in Table 1, and the concentrations are shown in Table S4 (the data of volatile compositions in FTL-LYT and FF-LYT were sourced from the previous publications (Guo, Ho, Schwab, & Wan., 2021b; Guo et al., 2019)).

Table 1.

Odor quality of the identified volatiles in large-leaf yellow tea (LYT) during manufacturing processing.

No. Volatile compounds RT (min) RI-1 RI-2 ID ζ Odor quality ψ
1 1-Methyl-1H-pyrrole 3.22 729 717 MS,RI Burnt, roasted, nutty
2 2-Methyl-1-butanol 3.23 729 723 MS,RI Roasted, cocoa-like, floral, malty
3 Pyrrole 3.41 739 733 MS,RI,S Nutty, sweet
4 3-Methyl-1-butanol 3.73 757 741 MS,RI Apple brandy flavor, spicy
5 Toluene 3.76 759 756 MS,RI,S Sweet, aromatic
6 (Z)-2-Penten-1-ol 3.77 759 765 MS,RI Green, fruity
7 Piperidine 3.84 763 764 MS,RI Animal-like
8 4-Methyl-3-penten-2-one 4.41 795 798 MS,RI Honey-like, card board-like, nutty, woody
9 Hexanal 4.51 800 801 MS,RI,S Grassy, green, fresh, fatty
10 2-Methylthiazole 4.68 805 815 MS,RI Green, nutty, vegetable-like
11 Dihydro-2-methyl-3(2H)furanone 4.71 805 809 MS,RI Roasted, coffee-like
12 1-Ethyl-1H-pyrrole 4.82 808 815 MS,RI Burnt, roasted
13 2,4-Dimethylheptane 5.11 816 821 MS,RI Alkane-like
14 Methylpyrazine 5.22 819 831 MS,RI,S Nutty, coffee, cocoa-like
15 Furfural 5.47 826 833 MS,RI,S Sweet, bready, caramel-like
16 2-Methyl-1H-pyrrole 5.62 830 814 MS,RI Nutty, roasted
17 2,4,5-Trimethyloxazole 5.98 840 829 MS,RI Burnt, nutty, hazelnut-like
18 (E)-3-Hexen-1-ol 6.15 844 852 MS,RI,S Green, grassy, fresh
19 2-Furanmethanol 6.21 846 859 MS,RI,S Burnt, sweet, bready, caramel-like
20 2-Hexenal 6.27 848 851 MS,RI,S Grassy, herbal
21 3-Hexen-1-ol 6.35 850 852 MS,RI,S Green, leafy, grassy
22 Ethylbenzene 6.52 855 852 MS,RI,S Aromatic
23 (E)-2-Hexen-1-ol 6.72 860 856 MS,RI Herbaceous, green
24 p-Xylene 6.87 864 862 MS,RI,S Plastic, green, pungent
25 1-Hexanol 6.88 864 874 MS,RI,S Green, cut grass
26 4-Cyclopentene-1,3-dione 7.29 875 881 MS,RI –
27 2(1H)-Pyrazinone 7.29 875 889 MS,RI Roasted, nutty
28 2,6-Dimethyl-1,5-heptadiene 7.43 879 882 MS,RI –
29 1,3,5,7-Cyclooctatetraene 7.76 888 880 MS,RI,S Solvent-like
30 Heptanal 8.28 902 907 MS,RI,S Green, oily, grassy
31 2-Acetylfuran 8.51 906 914 MS,RI Almond-like, nutty, cocoa-like
32 (E,E)-2,4-Hexadienal 8.64 909 909 MS,RI,S Green, sweet, fruity
33 2,5-Dimethylpyrazine 8.67 909 917 MS,RI,S Peanut, coffee, cocoa-like, nutty
34 Ethylpyrazine 8.83 913 921 MS,RI,S Nutty, coffee, cocoa-like
35 2,3-Dimethylpyrazine 8.91 914 926 MS,RI,S Cocoa-like, powdery, roasted, nutty
36 2-Ethyl-1H-pyrrole 9.25 921 930 MS,RI Burnt, roasted
37 1-Butyl-1H-pyrrole 10.27 942 946 MS,RI Roasted, nutty
38 2-Methyl-4-nonene 10.66 949 970 MS,RI –
39 5-Methyl-2-furancarboxaldehyde 11.02 957 961 MS,RI Caramel-like, bready, coffee-like
40 Benzaldehyde 11.07 958 962 MS,RI,S Almond-like, fruity, cherry-like, powdery
41 1-Heptanol 11.62 969 973 MS,RI Leafy green, vegetative-like, fruity
42 Heptyl formate 11.62 969 982 MS,RI Green, waxy, herbal, apple, cucumber-like
43 Methyl 2-furoate 11.71 970 980 MS,RI Sweet, caramel-like, brown sugar
44 Phenol 11.99 976 980 MS,RI,S Sweet, medicinal odor
45 1-Octen-3-ol 12.12 979 981 MS,RI,S Earthy, green, oily, vegetative-like, fungal
46 6-Methyl-5-hepten-2-one 12.35 983 986 MS,RI Fruity, apple-like, citrus
47 2,3-Octanedione 12.43 985 984 MS,RI Warmed-over flavor
48 2,2,4,6,6-Pentamethylheptane 12.54 987 991 MS,RI Alkane-like
49 β-Myrcene 12.59 988 988 MS,RI,S Woody, resinous, musty, balsamic, ethereal
50 2-Pentylfuran 12.60 988 989 MS,RI Fruity, green, earthy, beany, vegetable-like (faint)
51 6-Methyl-5-hepten-2-ol 12.79 992 998 MS,RI Green, coriander-like, oily
52 2-Ethyl-5-methylpyrazine 12.92 995 1005 MS,RI,S Nutty, caramel-like, coffee-like
53 Trimethylpyrazine 13.09 998 1005 MS,RI,S Cocoa-like, musty, nutty, peanut, potato-like
54 2-Ethyl-6-methylpyrazine 13.16 1000 1003 MS,RI,S Caramel-like, nutty, roasted potato
55 Decane 13.17 1000 1000 MS,RI,S Alkane-like
56 1-Methyl-1H-pyrrole-2-carboxaldehyde 13.18 1000 1016 MS,RI Bitter, almond, nutty
57 Octanal 13.34 1003 1004 MS,RI Fatty, green, citrus, waxy
58 (Z)-3-Hexen-1-ol acetate 13.44 1005 1006 MS,RI Fresh, green, grassy, fruity, sweet
59 3-Ethyl-2,4-dimethyl-1H-pyrrole 13.53 1006 998 MS,RI Nutty, almond-like
60 1H-Pyrrole-2-carboxaldehyde 13.61 1008 1015 MS,RI Roasted, burnt, smoky
61 (E,E)-2,4-Heptadienal 13.70 1010 1008 MS,RI Fatty, nutty, hay, green, oily
62 Hexyl acetate 13.86 1013 1011 MS,RI Fruity, sweet, fatty, fresh, apple, pear-like
63 1-Methyl-4-(1-methylethyl)-1,3-cyclohexadiene 13.95 1014 1010 MS,RI Citrus, woody, spicy, lemon-like
64 1,2,3-Trimethylbenzene 14.06 1016 1015 MS,RI Plastic
65 o-Cymene 14.36 1022 1022 MS,RI Aromatic
66 7-Propylidene-bicyclo[4.1.0]heptane 14.59 1026 1025 MS,RI Cooling, minty
67 Limonene 14.60 1027 1026 MS,RI,S Citrus, lemon, orange-like, green
68 2-Acetyl-5-methylfuran 14.76 1030 1039 MS,RI Nutty, caramel-like, sweet
69 Benzyl alcohol 14.83 1031 1036 MS,RI,S Fruity, rose-like
70 2,2,6-Trimethylcyclohexanone 14.90 1032 1036 MS,RI Alkane-like
71 trans-β-Ocimene 15.06 1035 1048 MS,RI,S Warm, floral, herbal, sweet
72 Benzeneacetaldehyde 15.36 1041 1038 MS,RI,S Floral, rose, cherry-like
73 1-Ethyl-1H-pyrrole-2-carboxaldehyde 15.53 1044 1039 MS,RI Burnt, roasted, smoky
74 β-Ocimene 15.59 1045 1041 MS,RI,S Citrus, herbaceous, sweet
75 1-Pentyl-1H-pyrrole 15.78 1049 1054 MS,RI Burnt
76 γ-Terpinene 16.16 1056 1050 MS,RI,S Citrus, lemon-like, woody, spicy, juicy
77 2-Acetylpyrrole 16.32 1059 1059 MS,RI Nutty, musty
78 Acetophenone 16.47 1062 1061 MS,RI Sweet, cherry pit, vanilla-like
79 Benzylamine 16.69 1066 1069 MS,RI Amine-like
80 Linalool oxide I 16.85 1069 1074 MS,RI,S Sweet, floral, creamy
81 1-Octanol 16.94 1070 1061 MS,RI,S Green, citrus, fatty, coconut-like
82 3-Ethyl-2,5-dimethylpyrazine 17.07 1073 1079 MS,RI,S Cocoa-like, roasted, nutty
83 2-Ethyl-3,5-dimethylpyrazine 17.40 1079 1081 MS,RI,S Peanut, nutty, caramel-like, coffee-like, roasted
84 1-Methyl-4-(1-methylethylidene)cyclohexene 17.58 1082 1088 MS,RI Fresh, woody, sweet, piney, citrus
85 Linalool oxide II 17.74 1085 1081 MS,RI,S Sweet, floral, creamy
86 2,5-Diethylpyrazine 17.87 1088 1085 MS,RI,S Nutty, hazelnut-like
87 Undecane 18.50 1100 1100 MS,RI,S Alkane-like
88 Linalool 18.51 1100 1103 MS,RI,S Floral, sweet, grape-like, woody
89 Hotrienol 18.62 1102 1106 MS,RI,S Fresh, floral, fruity
90 Nonanal 18.74 1104 1112 MS,RI,S Floral, green, lemon-like
91 2,6-Dimethyl-2,4,6-octatriene 19.90 1127 1130 MS,RI Sweet, floral, nutty, herbal, peppery
92 Phenylethyl alcohol 18.98 1109 1120 MS,RI,S Floral, rose-like
93 2-Isobutyl-3-methylpyrazine 20.19 1132 1134 MS,RI Nutty, roasted
94 5H-5-Methyl-6,7-dihydrocyclopentapyrazine 20.33 1135 1149 MS,RI Nutty, roasted, toasted, grainy, coffee-like
95 (E,Z)-2,6-Dimethyl-2,4,6-octatriene 20.51 1138 1131 MS,RI Sweet, floral, nutty, herbal, peppery
96 2,6,6-Trimethyl-2-cyclohexene-1,4-dione 20.71 1142 1138 MS,RI Musty, woody, sweet, tea-like, citrus
97 1,2,3,4-Tetramethylbenzene 20.88 1145 1145 MS,RI –
98 2,3-Diethyl-5-methylpyrazine 21.03 1148 1145 MS,RI,S Roasted (beef), nutty, meaty, hazelnut-like
99 3,5-Diethyl-2-methylpyrazine 21.18 1151 1162 MS,RI Nutty, meaty
100 (Z)-3-Nonen-1-ol 21.22 1152 1150 MS,RI Fresh, waxy, green, mushroom-like
101 (E,Z)-3,6-Nonadien-1-ol 21.33 1154 1156 MS,RI Sweet, fresh, green, waxy, melon, fruity
102 Linalool oxide III 22.03 1168 1175 MS,RI,S Floral, honey-like
103 1-Nonanol 22.22 1171 1172 MS,RI Fresh, clean, fatty, floral, rose-like
104 Linalool oxide IV 22.31 1173 1183 MS,RI,S Floral, honey-like
105 1-Furfurylpyrrole 22.41 1175 1187 MS,RI Cocoa-like, roasted
106 Levomenthol 22.46 1176 1172 MS,RI,S Peppermint, cooling, minty
107 (E,E)-2,6-Dimethyl-1,3,5,7-octatetraene 22.74 1181 1195 MS,RI –
108 3-Methylacetophenone 22.74 1181 1182 MS,RI Sweet, fruity, nutty, vanilla-like
109 (Z)-3-Hexenyl butanoate 22.93 1185 1184 MS,RI Fresh, green
110 Methyl salicylate 23.13 1189 1181 MS,RI,S Peppermint, wintergreen-like
111 α-Terpineol 23.33 1192 1183 MS,RI,S Pleasant, floral
112 2,5-Dimethyl-3-(2-methylpropyl)pyrazine 23.48 1195 1185 MS,RI Nutty, roasted
113 Safranal 23.48 1195 1189 MS,RI Medicinal, spicy, woody, herbaceous, phenolic
114 Dodecane 23.70 1200 1200 MS,RI,S Alkane-like
115 2,6,6-Trimethyl-1-cyclohexene-1-carboxaldehyde 24.48 1215 1209 MS,RI Herbal, clean, rose-like, fruity
116 6,7-Dihydro-2,5-dimethyl-5H-cyclopentapyrazine 24.72 1220 1238 MS,RI Burnt, earthy, nutty, coffee-like, roasted
117 6,7-Dimethyl-3,5,8,8a-tetrahydro-1H-2-benzopyran 24.88 1223 1242 MS,RI Nutty, roasted
118 (Z)-3,7-Dimethyl-2,6-octadien-1-ol 24.83 1222 1231 MS,RI Fresh, citrus, floral, green, lemon-like
119 (Z)-3,7-Dimethyl-3,6-octadien-1-ol 25.08 1228 1232 MS,RI Fresh, citrus, floral, green, sweet, lime-like
120 cis-3-Hexenyl valerate 25.19 1230 1236 MS,RI Fruity, apple, pear-like, green
121 Geraniol 26.21 1250 1263 MS,RI,S Rose-like, sweet, honey-like
122 2,6,6-Trimethyl-1-cyclohexene-1-acetaldehyde 26.30 1252 1257 MS,RI Oily, fruity, cooling, camphoraceous, woody
123 (E)-4-(2-Butenyl)-1,2-dimethylbenzene 26.67 1260 1271 MS,RI –
124 (E)-3,7-Dimethyl-2,6-octadienal 26.98 1266 1270 MS,RI,S Citrus, lemon-like
125 Indole 27.99 1286 1289 MS,RI,S Floral, animal-like
126 Tridecane 28.65 1300 1300 MS,RI,S Alkane-like
127 2,5-Dimethyl-3-(3-methylbutyl)pyrazine 29.09 1309 1314 MS,RI Nutty, roasted
128 2,6,10,10-Tetramethyl-1-oxaspiro[4.5]dec-6-ene 29.18 1311 1310 MS,RI Woody, cooling, minty, herbal
129 1,2-Dihydro-1,1,6-trimethylnaphthalene 30.96 1349 1344 MS,RI Licorice-like
130 1,2,3,4-Tetrahydro-1,4,6-trimethylnaphthalene 31.06 1351 1364 MS,RI –
131 β-Damascenone 32.18 1376 1384 MS,RI,S Apple-like, herbaceous, woody, smoky, citrus, nutty, rose-like, wine-like
132 (Z)-3-Hexenyl hexanoate 32.35 1379 1380 MS,RI Green, waxy, winey, grassy
133 Hexyl hexanoate 32.61 1385 1371 MS,RI,S Fruity, sweet
134 1,7-Dimethylnaphthalene 33.28 1399 1396 MS,RI –
135 Tetradecane 33.31 1400 1400 MS,RI,S Alkane-like
136 β-Damascone 33.52 1404 1416 MS,RI Floral, rose-like, leathery, woody, fruity, minty, sweet
137 1,2,3,4-Tetramethyl-4-(1-methylethenyl)benzene 33.89 1413 1412 MS,RI –
138 α-Ionone 34.13 1418 1419 MS,RI,S Floral, violet, powdery, berry-like
139 Eugenol 35.21 1443 1413 MS,RI,S Sweet, spicy, clove-like, woody
140 Geranylacetone 35.33 1446 1451 MS,RI Fresh, rose-like, floral, green, fruity
141 cis-β-Farnesene 35.58 1451 1447 MS,RI,S Floral, citrus
142 trans-β-Ionone 36.62 1475 1478 MS,RI,S Violet, raspberry, floral
143 Germacrene D 36.73 1477 1482 MS,RI Woody, spicy
144 (Z,E)-α-Farnesene 37.24 1489 1481 MS,RI Floral
145 Bicyclogermacrene 37.36 1492 1492 MS,RI Spicy
146 α-Bulnesene 37.41 1493 1502 MS,RI Herbal, spicy, green
147 α-Farnesene 37.83 1502 1508 MS,RI,S Woody, green, floral, herbal
148 (1S-cis)-1,2,3,5,6,8a-Hexahydro-4,7-dimethyl-1-(1-methylethyl)naphthalene 38.32 1514 1516 MS,RI Herbal, woody
149 α-Calacorene 39.28 1537 1542 MS,RI Woody
150 Cadala-1(10),3,8-triene 39.29 1538 1548 MS,RI Fresh, woody
151 Nerolidol 40.18 1559 1556 MS,RI,S Floral, green, citrus, woody, waxy
152 3-Hexen-1-ol benzoate 40.57 1568 1553 MS,RI Fresh, green, leafy
153 1-Isobutyl 4-isopropyl 3-isopropyl-2,2-dimethylsuccinate 41.25 1585 1581 MS,RI –
154 Hexadecane 41.87 1600 1600 MS,RI,S Alkane-like
155 Methyl 8-(2-furyl)octanoate 42.64 1619 1624 MS,RI Waxy, green, herbal
156 Benzophenone 42.79 1623 1628 MS,RI Sweet, floral, rose-like
157 Dimethyl decanedioate 43.66 1645 1648 MS,RI Plastic-like
158 Diethyl-α-naphthylamine 44.52 1667 1692 MS,RI Amine-like
159 Heptadecane 45.82 1700 1700 MS,RI,S Alkane-like
160 Benzyl benzoate 48.19 1763 1762 MS,RI Floral, sweet, balsamic
161 Anthracene 48.51 1771 1775 MS,RI Aromatic
162 Phenanthrene 48.53 1772 1775 MS,RI Aromatic
163 Octadecane 49.57 1800 1800 MS,RI,S Alkane-like
164 Benzyl salicylate 51.90 1865 1870 MS,RI,S Balsam, clean, herbal, oily, sweet
165 Nonadecane 53.14 1900 1900 MS,RI,S Alkane-like
166 Methyl hexadecanoate 53.97 1924 1926 MS,RI Oily, waxy, fatty
167 n-Hexadecanoic acid 55.20 1960 1968 MS,RI Waxy, creamy, candle-like
168 Ethyl hexadecanoate 56.30 1992 1994 MS,RI Oily, waxy, fatty
169 (E,E)-3,7,11,15-Tetramethyl-1,6,10,14-hexadecatetraen-3-ol 57.21 2019 2020 MS,RI Floral, rose-like, balsam
170 Methyl (Z,Z)-9,12-octadecadienoate 59.54 2088 2081 MS,RI Oily, fatty, woody
171 Methyl (Z,Z,Z)-9,12,15-octadecatrienoate 59.73 2094 2098 MS,RI Oily, fatty, woody
172 Phytol 60.18 2107 2114 MS,RI,S Floral, balsam, powdery, waxy
173 Ethyl linoleate 62.05 2162 2163 MS,RI Oily, fatty, woody
174 Octadecanoic acid 62.08 2163 2173 MS,RI Oily, waxy
175 Ethyl (Z,Z,Z)-9,12,15-octadecatrienoate 62.23 2168 2165 MS,RI Oily, fatty, woody
176 Tricosane 66.11 2300 2300 MS,RI,S Alkane-like
177 Pentacosane 69.82 2500 2500 MS,RI,S Alkane-like
178 Squalene 74.47 2811 2814 MS,RI Floral, pleasant flavor
Open in a new tab

Significant differences were found in the number and amount of identified volatiles among the tea samples, which can be attributed to the multiple biosynthetic pathways such as carotenoid degradation, lipid degradation, glycoside hydrolysis, and the Maillard reaction that occur during tea processing (Ho et al., 2015). Each of these pathways contributes uniquely to the generation of volatiles, ultimately influencing the aroma profile of LYT. The number of volatiles presented in eight samples varied from 58 to 104, being the greatest in FF-LYT and the lowest in RR-FYT. Whereas the highest total amount of volatiles was found in FTL-LYT (228.30 μg/kg), followed by Ro-LYT (45.74 μg/kg) and FF-LYT (43.91 μg/kg), with the lowest amount in RY-LYT (22.32 μg/kg) (Fig. 3A). In addition, the total amount of volatiles from samples with thermal treatment, which were of Fx-LYT, PR-LYT, and RR-LYT were lower than those of the samples of the corresponding previous processing step. The concentrations increased significantly from RY-LYT to FF-LYT, the sample which underwent full fire processing, indicating that the full fire processing was beneficial to the liberation of volatiles in LYT.

Fig. 3.

Fig. 3

Open in a new tab

Aroma profiles and PCA analysis of LYT during processing based on identified volatiles by GC–MS.

(A) The total amount of volatiles during LYT processing. (B) The changes in the number of volatile categories during LYT processing. (C) The changes in the proportion of volatile categories during LYT processing. (D) The number and proportion of N-containing heterocyclic compounds of LYT during processing. The score plot (E) and biplot (F) of PCA analysis based on identified volatiles. Compound nos. Correspond to Table 1. Columns labeled with ‘a’, ‘b’, ‘c’, ‘d’, and ‘e’ had significant difference (p< 0.05) with each other.

In detail, there were 95, 67, 66, 59, 58, 55, 68, and 100 volatiles identified in FTL-LYT, Fx-LYT, Ro-LYT, PR-LYT, PY-LYT, RR-LYT, RY-LYT, and FF-LYT samples, respectively. These volatile compounds comprised alcohols, alkenes, esters, aldehydes, heterocyclics, aromatics, ketones, and hydrocarbons. Among them, the alcohols were the most abundant, in number and proportion of LYT samples before the full fire processing. In contrast, the largest number and proportion of volatiles were heterocyclic compounds in FF-LYT after full fire treatment, which accounted for 42 and 56.57 %, respectively (Fig. 3B & C). Heterocyclic compounds and nitrogen-containing heterocyclic compounds in particular, which are derived from Maillard reaction or Strecker degradation during tea processing, are vital for the formation of roasted or nutty odors of LYT flavor (Guo et al., 2019). And indeed, a large number of volatiles containing nitrogen (totally 34) were identified in FF-LYT, of which 30 were newly generated after full fire processing. These compounds are only present in low concentrations or not at all in other samples with lower intensity of heat treatment in comparison to full fire processing or samples without thermal treatment (Fig. 3D). Notably, the presence of indole with floral odor was not detected in tea samples except for FTL-LYT and FF-LYT, which might be related to thermal reaction under pyrolysis conditions during full fire processing which promotes the formation of indole in final product teas (0.35 μg/kg) (Ho et al., 2015). 5-Methyl-2-furancarboxaldehyde exhibits a caramel-like flavor and is known as a thermal product from sugars and amino acids (Kato & Shibamoto, 2001). It was produced after full fire processing. Additionally, 1-ethyl-1H-pyrrole-2-carboxaldehyde having roasted or smoky odors (Yang et al., 2021) was detected in RY-LYT and clearly increased in concentration after full fire treatment, consistent with the results in coffee (Kıvançlı & Elmacı, 2016) where the amount raised with the enhancement of roasting degree.

After alcohols, aldehydes were the most abundant volatiles in tea samples from Fx-LYT to RY-LYT, and 9 or more aldehydes had been identified in these samples, while the proportion of aldehydes of the total amount of volatiles was relatively low, ranged from 2.88 % (Ro-LYT) to 6.06 % (RR-LYT). The lowest number and proportion of aldehydes was found in FF-LYT, which was of 3 and 1.44 %, respectively. Among all identified aldehydes, six volatiles including hexanal, 2-hexenal, heptanal, octanal, nonanal and safranal, which impart green, fatty or medicine-like odors were identified in the previous samples of FF-LYT, but were undetectable in FF-LYT. Benzaldehyde and benzeneacetaldehyde, which convey almond-like and floral notes, respectively (Ho et al., 2015) were identified in all tea samples. The formation of benzeneacetaldehyde was promoted by a thermal reaction (Ho et al., 2015), and the concentration in FF-LYT (0.48 μg/Kg) did increase significantly after full fire processing, which was twice of RY-LYT.

In contrast to aldehydes, few aromatics were identified in a high amount, second only to alcohols in tea samples between the processing of FTL and full fire. The highest number of aromatics was found in FF-LYT (13), and the highest proportion was found in RR-LYT, reaching 21.18 %, which was 8.54 times of the lowest proportion of FTL-LYT and almost two times of FF-LYT. As presented in Fig. 3B & C, alcohols dominated the volatiles in RY-LYT and previous samples, and no less than 17 volatiles were identified in each sample during this period, accounting for more than 50 % of the total amount of volatiles in each corresponding sample, ranging from 50.36 % (RY-LYT) to 69.08 % (FTL-LYT). However, the proportion of alcohols dropped sharply to 6.27 % in FF-LYT, and only 8 volatiles were identified and also were the common alcohol compounds over the whole processing. Among them, the four major components were linalool, linalool oxide I & II, and phytol, and their contents exceeded 0.40 μg/kg in FF-LYT. The latter one having green odor was found in green tea with a high content (Guo et al., 2021b). The former three volatiles exhibiting floral, sweet or earthy odors (Guo et al., 2025; Ho et al., 2015) were selected for further analysis because they were confirmed as the key aroma compounds in large-leaf yellow tea (Guo et al., 2021a) or yellow tea (Shi et al., 2021).

The number of alkenes identified in FTL-/FF-LYT was the highest, but their proportion was the least. And the highest proportion was found in PR-LYT (10.49 %). Four common volatiles, including limonene and α-farnesene were identified in all tea samples. α-Farnesene with woody odor was derived from carotenoids in oolong tea and black tea (Kawakami & Kobyashi, 2000). The amount of α-farnesene decreased gradually during processing, reaching the lowest content in FF-LYT (0.13 μg/kg), whereas limonene imparting citrus odor performed an opposite trend. β-Ocimene and (Z,E)-α-farnesene were only identified in the previous samples of RR-LYT. The highest amount of (Z,E)-α-farnesene, which has a floral odor was noted in FTL-LYT, then declined significantly after fixing and subsequently remained relatively stable in the rest of the samples. β-Ocimene exhibits herbaceous or green flavor in tea (Lin et al., 2013) and is a volatile component involved in the response of the tea plant to herbivore or mechanical damage (Jian et al., 2021). Its content raised with the progress of tea processing, and reached a peak value of 0.09 μg/kg in PR-LYT. In addition, β-myrcene, trans-β-ocimene, γ-terpinene, α-bulnesene and α-calacorene were only identified in FF-LYT. The proportion of ketones was relatively low in all tea samples, ranging from 0.67 % (FTL-LYT) to 3.63 % (RY-LYT), though no less than 7 volatiles were identified in tea samples. In the post-processing stages, the proportion of ketones increased from the sample of PY-LYT, which might be attributed to the high abundance of trans-β-ionone (violet-like odor) and geranylacetone (rose-like odor) in these samples. Both volatiles together with acetophenone imparting sweet odor were commonly identified in all samples. Different from 4-methyl-3-penten-2-one with honey-like flavor, which was measured in Fx-LYT and subsequent samples, 6-methyl-5-hepten-2-one (fruity odor) and β-damascenone (fruity odor) were identified in the samples prior to FF-LYT. Similarity, 2,3-octanedione with warmed-over flavor, which is reported as a diet tracer in lamb fats (Sivadier et al., 2009), occurred in relatively low amount in relevant tea samples and was not detected in final teas. Additionally, β-damascenone and β-damascone, which are part of the volatile fraction of rose essential oils and formed by the biodegradation of β-carotene (Uenojo & Pastore, 2010), were identified simultaneously in tea aroma for the first time.

Although the number and proportion of esters were high in FTL-LYT, which were next to that of alcohols, both decreased to varying degrees thereafter, and reached the lowest value of 2 and 2.60 %, respectively in RR-LYT, and then significantly increased to 3.03 times from RY-LYT to FF-LYT. Twelve volatiles with green, fruity or fatty odors were present in FTL-LYT but not in subsequent samples. Three of the 12 volatiles, identified as ethyl hexadecanoate, ethyl linoleate and ethyl (Z,Z,Z)-9,12,15-octadecatrienoate with fatty odors, had high contents greater than 1.36 μg/kg. The nine remaining FTL-specific compounds had relatively low amounts below 0.38 μg/kg. The one common ester that was found in all samples was methyl salicylate, which possesses minty or green odors and is formed by glycoside hydrolysis in tea (Wang et al., 2011; Wang & Ruan, 2009). Another ester compound with green odor was (Z)-3-hexenyl hexanoate. Its content gradually decreased throughout the LYT production but was undetectable in FF-LYT. Unlike other teas, LYT samples contained much more linear alkanes, either in number or proportion, especially in FTL-LYT, where 10 compounds were identified. Dodecane and tridecane were commonly identified with high abundance in all samples (Fig. S2). Moreover, the 23- and 25‑carbon alkanes, identified as tricosane or pentacosane were present in tea samples prior to FF-LYT but not in FF-LYT. This might be due to the degradation of the carbon chain during full fire processing.

The variations in the number or percentage of volatiles were mainly observed during processing, which is the basis for E-nose technology to detect and distinguish tea samples with different processes.

3.3. Principal component analysis

The PCA analysis was also applied to further discriminate the differences of LYT samples with different processes, which was carried out with the identified volatiles. The 8 tea samples appeared to be well separated according to the LYT processes except for the partial overlap of PR-LYT and PY-LYT (Fig. 3E), consistent with the volatiles in a similar number and proportion from the two samples. The distance between FTL-LYT and FF-LYT was large, which was consistent with their different characteristic aroma compounds. Moreover, fresh tea leaves had green flavor, while they were replaced by crispy rice-like and nutty odors with high intensities in the actual evaluation of LYT (Guo et al., 2021a), indicating that tea processing had a significant impact on the final flavor formation of LYT. As shown in Fig. 3F, the FTL-LYT showed high scores in positive PC1 and PC2 which contained high loadings of (E,E)-2,4-heptadienal, benzeneacetaldehyde, α-terpineol, 1-octanol, heptanal, α-farnesene, (Z)-3,7-dimethyl-2,6-octadien-1-ol, (E,E)-2,4-hexadienal, etc., the characteristic volatiles with green, fatty, woody or floral odors. The FF-LYT exhibited high scores on negative PC1 and positive PC2 where the loadings of characteristic aroma compounds including p-xylene, β-myrcene, 1-methyl-1H-pyrrole, furfural, pyrrole, 1,2-dihydro-1,1,6-trimethylnaphthalene, 1,2,3,4-tetrahydro-1,4,6-trimethylnaphthalene, and 2(1H)-pyrazinone, belonging to heterocyclic compounds or aromatics were high. In addition, the levels of nonanal, 6-methyl-5-hepten-2-one, tetradecane, tridecane, dodecane, octanal, and α-terpineol in the remaining samples were similar and they were regarded as the co-characteristic volatiles in these samples.

3.4. Aroma profiling of key aroma compounds

The previous study reported that 13 volatiles with odor activity value (OAV) above one were the key aroma compounds in LYT (Table S5) (Guo et al., 2021a), of which 11 volatile compounds (including 6 N-containing heterocyclic compounds) were newly produced in FF-LYT (Fig. 4A). The remaining two volatiles, identified as trans-β-ionone and linalool, which were the principal contributor to the violet-like or floral odors were measured in all tea samples (Fig. 4B & C). The amount of linalool decreased except after rolling as compared with Fx-LYT during the LYT production. However, the content of trans-β-ionone showed a U-shaped curve during processing, decreased first to the valley value of 0.07 μg/kg in Ro-LYT and then gradually rose to the peak value of 0.47 μg/kg in FF-LYT. Significantly, the concentration of trans-β-ionone was enhanced by 1.36 times from RY-LYT to FF-LYT, and the high intensity roasting of full fire processing had the potential to promote the formation of trans-β-ionone (Kanasawud & Crouzet, 1990). Besides these volatiles mentioned above, four other volatiles, namely 1-octanol, β-damascenone, geranylacetone and nerolidol with OAV higher than one were perceived as the key aroma compounds in summer green tea (Table S6) made with the same batch of fresh tea leaves of LYT (Guo et al., 2021b). As can be seen in Fig. 4D, E & F, their amounts were significantly decreased with different variation degrees from FTL-LYT to FF-LYT. The highest amounts of these volatiles were found in FTL-LYT, and their contents in RY-LYT were higher than that of RR-LTY. The former two were undetectable in FF-LYT, and the latter two giving floral flavor to tea had a low level in FF-LTY. In addition, the contents of geranylacetone in the samples of two times of yellowing (PY-LYT & RY-LYT) were much greater than those of the corresponding previous samples (Fig. 4E), and the enhancement might be related to the non-enzymatic degradation of carotenoids (Ho et al., 2015). Even if the teas are made with the same materials, different processing methods produce different types of tea, which had different volatiles, consistent with the actual results of the sensory evaluation from Guo et al. (2021b) that LYT and green tea possessed distinctive aroma characteristics.

Fig. 4.

Fig. 4

Open in a new tab

The variations of selected volatile compounds of LYT during processing.

(A) The formation of aroma-active compounds in LYT. Compound nos. Correspond to Table 1. (B) trans-β-Ionone. (C) Linalool. (D) 1-Octanol. (E) β-Damascenone & Geranylacetone. (F) Nerolidol.

Apart the large-leaf yellow tea, bud or little yellow tea are the other two kinds of yellow tea because yellowing is a key process during manufacturing. Twenty-five volatiles were considered as the key odorants (Table S7) with aroma intensities greater than 1.5 in bud and little yellow teas (BL-YT) (Shi et al., 2021), among which 14 were also identified in both LYT and BL-YT, including linalool and trans-β-ionone. These volatiles were roughly classified into two groups with either decreasing or increasing levels from RY-LYT to FF-LYT (Fig. 5A & B). The group of decreasing content mainly consisted of 1-octen-3-ol, phenylethyl alcohol, geraniol, benzaldehyde, linalool oxide II, and geranylacetone. The former three volatiles, which are responsible for mushroom-like, floral and rose-like odors, respectively, were not found in FF-LYT. 1-Octen-3-ol was significantly increased after the first yellowing process. The five remaining volatiles were slightly elevated in Ro-LYT, and the aroma compounds might be liberated along the loss of tea juice within rolling. In addition, the concentrations of (E,E)-2,4-heptadienal, 1-ethyl-1H-pyrrole-2-carboxaldehyde, linalool oxide I, methyl salicylate, indole and α-ionone increased. Similar to the cases of linalool oxide II, the contents of linalool oxide I and methyl salicylate were significantly increased in Ro-LYT, and finally rose to 1.19 and 4.23 times, respectively in FF-LYT as compared with RY-LYT, which might be due to the hydrolysis of the related glycosidic bonds by thermal treatment (Sasaki et al., 2017). Six abundant volatiles in yellowing were selected for further discussion (Fig. 5C) because yellowing is crucial for yellow tea manufacturing and ultimately affects the flavor quality (Wei et al., 2020; Wei et al., 2021). Of the six compounds, five were not present in FF-LYT but were abundant in the samples of either primary yellowing or re-yellowing. In addition, hotrienol imparting floral odor was not found in FF-LYT, inconsistent with the result in green tea that the content was enhanced after drying (Guo et al., 2021b), which might be related to the different roasting degrees. Although 3-methyl-1-butanol having apple brandy or spicy odors was identified in FF-LYT, its amount was significantly decreased from RY-LYT to FF-LYT and therefore of low level. To a certain extent, these results indicated that the full fire processing beyond yellowing was pivotal for the final formation of LYT flavor.

Fig. 5.

Fig. 5

Open in a new tab

The changes of selected volatile compounds during LYT processing.

The content variations of key aroma compounds of yellow tea during LYT processing (A) & (B). The changes in concentration of volatiles with high abundance in yellowing during LYT processing (C).

4. Conclusions

The present study examined the variations in volatile compounds and aroma profiles of LYT during processing using GC–MS in conjunction with E-nose technology. A total of 178 volatiles were found during the overall processing, with 25 being commonly present and 62 newly generated, notably including 30 N-containing compounds. The number and concentration of these volatile compounds changed significantly during processing. Alcohols, aromatics, alkenes and aldehydes were the main volatiles during processes while the heterocyclics and N-containing compounds in particular dominated the volatiles in the final product teas. Accordingly, the aroma properties shifted from green, fruity and fatty during processes to roasted or nutty odors in the final teas. The processing samples were distinguished by PCA analysis based on the identified volatiles or E-nose responses, which were significantly different among tea samples. In addition, the key aroma compounds of LYT were different from those of green tea made with the same batch of fresh tea leaves, and bud or little yellow teas, suggesting that aroma profiles differ largely among different tea types or teas within the same category but processed differently. The current studies also demonstrated the potential of E-nose as a simple and effective tool for detecting tea volatiles, in combination with GC–MS, for discrimination aroma profiles and quality control during tea processing.

CRediT authorship contribution statement

Xiangyang Guo: Writing – review & editing, Writing – original draft, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Wilfried Schwab: Writing – review & editing, Methodology, Formal analysis. Chi-Tang Ho: Writing – review & editing, Supervision, Methodology, Formal analysis, Data curation, Conceptualization. Chuankui Song: Writing – review & editing. Xiaochun Wan: Writing – review & editing, Supervision, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was financially supported by the National Natural Science Foundation of China (32072634), The Open Fund of National Key Laboratory for Tea Plant Germplasm Innovation and Resource Utilization (NKLTOF20240116), Key Scientific Research Project of Higher Education Institutions in Henan Province (24B210015), and Science and Technology Key Project of Henan Province (252102110093). We sincerely thank engineer Xuchun Li (Yaozhun Pharmaceutical Technology Co., Ltd., Shanghai, China) for kind assistance of detection and data processing of E-nose.

The following chemical compounds were identified in this study:

1-Octanol (PubChem CID: 3301); Benzeneacetaldehyde (PubChem CID: 7103); Linalool (PubChem CID: 2795); Furfural (PubChem CID: 980); Pyrrole (PubChem CID: 7962); 1-Methyl-1H-pyrrole (PubChem CID: 7471); trans-β-Ionone (PubChem CID: 638115); Geranylacetone (PubChem CID: 13826); Nerolidol (PubChem CID: 5280450); Indole (PubChem CID: 7372).

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.fochx.2025.102507.

Contributor Information

Chi-Tang Ho, Email: ctho@sebs.rutgers.edu.

Xiaochun Wan, Email: xcwan@ahau.edu.cn.

Appendix A. Supplementary data

Supplementary material

mmc1.docx (400.4KB, docx)

Data availability

Data will be made available on request.

References

  1. Chen H., Wang Z., Gong L., Chen J., Huang Y., Guo W., Zhang Q., Li Y., Bao G., Li D., Chen Y. Attenuation effect of a polysaccharide from large-leaf yellow tea by activating autophagy. Internatinal Journal of Biological Macromolecules. 2024;265 doi: 10.1016/j.ijbiomac.2024.130697. [DOI] [PubMed] [Google Scholar]
  2. Cho J.Y., Mizutani M., Shimizu B., Kinoshita T., Ogura M., Tokoro K.…Sakata S. Chemical profiling and gene expression profiling during the manufacturing process of Taiwan oolong tea “oriental beauty”. Bioscience, Biotechnology, and Biochemistry. 2007;71(6):1476–1486. doi: 10.1271/bbb.60708. [DOI] [PubMed] [Google Scholar]
  3. Ge H., Qi F., Shen Z., Wang H., Zhu S., Zhou S., Xie Z., Li D. Large-leaf yellow tea protein derived-peptides alleviated dextran sodium sulfate-induced acute colitis and restored intestinal microbiota balance in C57BL/6 J mice. Food Chemistry. 2024;456 doi: 10.1016/j.foodchem.2024.139936. [DOI] [PubMed] [Google Scholar]
  4. Guo X., Ho C.T., Schwab W., Song C., Wan X. Aroma compositions of large-leaf yellow tea and potential effect of theanine on volatile formation in tea. Food Chemistry. 2019;280:73–82. doi: 10.1016/j.foodchem.2018.12.066. [DOI] [PubMed] [Google Scholar]
  5. Guo X., Ho C.T., Schwab W., Wan X. Effect of the roasting degree on flavor quality of large-leaf yellow tea. Food Chemistry. 2021;347 doi: 10.1016/j.foodchem.2021.129016. [DOI] [PubMed] [Google Scholar]
  6. Guo X., Ho C.T., Schwab W., Wan X. Aroma profiles of green tea made with fresh tea leaves plucked in summer. Food Chemistry. 2021;363 doi: 10.1016/j.foodchem.2021.130328. [DOI] [PubMed] [Google Scholar]
  7. Guo X., Wang L., Huang X., Zhou Q. Characterization of the volatile compounds in tea (Camellia sinensis L.) flowers during blooming. Frontiers in Nutrition. 2025;11 doi: 10.3389/fnut.2024.1531185. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Ho C.T., Zheng X., Li S. Tea aroma formation. Food Science and Human Wellness. 2015;4(1):9–27. doi: 10.1016/j.fshw.2015.04.001. [DOI] [Google Scholar]
  9. Jian G., Jia Y., Li J., Zhou X., Liao Y., Dai G., Zhou Y., Tang J., Zeng L. Elucidation of the regular emission mechanism of volatile β-ocimene with anti-insect function from tea plants (Camellia sinensis) exposed to herbivore attack. Journal of Agricultural and Food Chemistry. 2021;69(38):11204–11215. doi: 10.1021/acs.jafc.1c03534. [DOI] [PubMed] [Google Scholar]
  10. Kanasawud P., Crouzet J.C. Mechanism of formation of volatile compounds by thermal degradation of carotenoids in aqueous medium. 1. Β-carotene degradation. Journal of Agricultural and Food Chemistry. 1990;38(1):237–243. doi: 10.1021/jf00091a052. [DOI] [Google Scholar]
  11. Kato M., Shibamoto T. Variation of major volatile constituents in various green teas from Southeast Asia. Journal of Agricultural and Food Chemistry. 2001;49:1394–1396. doi: 10.1117/12.2062259. [DOI] [PubMed] [Google Scholar]
  12. Kawakami M., Kobyashi A. Carotenoid-derived aroma compounds in tea. Abstracts of Papers of the American Chemical Society. 2000;219:U32. doi: 10.1021/bk-2002-0802.ch011. [DOI] [Google Scholar]
  13. Kıvançlı J., Elmacı Y. Characterization of Turkish-style boiled coffee aroma by gas chromatography and mass spectrometry and descriptive analysis techniques. International Journal of Food Properties. 2016;19:1671–1686. doi: 10.1080/10942912.2015.1080726. [DOI] [Google Scholar]
  14. Li Y., Luo Q., Qin M., Xu W., Wang X., Zhou J., He C., Chen Y., Yu Z., Ni D. Study on color, aroma, and taste formation mechanism of large-leaf yellow tea during an innovative manufacturing process. Food Chemistry. 2024;438 doi: 10.1016/j.foodchem.2023.138062. [DOI] [PubMed] [Google Scholar]
  15. Li Y., Zhang J., Jia H., Pan Y., Xu Y.-Q., Wang Y., Deng W.-W. Metabolite analysis and sensory evaluation reveal the effect of roasting on the characteristic flavor of large-leaf yellow tea. Food Chemistry. 2023;427 doi: 10.1016/j.foodchem.2023.136711. [DOI] [PubMed] [Google Scholar]
  16. Lin J., Zhang P., Pan Z., Xu H., Luo Y., Wang X. Discrimination of oolong tea (Camellia sinensis) varieties based on feature extraction and selection from aromatic profiles analysed by HS-SPME/GC-MS. Food Chemistry. 2013;141(1):259–265. doi: 10.1016/j.foodchem.2013.02.128. [DOI] [PubMed] [Google Scholar]
  17. Ma L., Qiao Y., Du L., Li Y., Huang S., Liu F., Xiao D. Evaluation and optimization of a superior extraction method for the characterization of the volatile profile of black tea by HS-SPME/GC-MS. Food Analyticl Methods. 2017;10(7):2481–2489. doi: 10.1007/s12161-016-0785-y. [DOI] [Google Scholar]
  18. Peris M., Escuder-Gilabert L. A 21st century technique for food control: Electronic noses. Analytica Chimica Acta. 2009;638:1–15. doi: 10.1016/j.aca.2009.02.009. [DOI] [PubMed] [Google Scholar]
  19. Rao J., Zhang Y., Yang Z., Li S., Wu D., Sun C., Chen K. Application of electronic nose and GC-MS for detection of strawberries with vibrational damage. Food Quality and Safety. 2020;4:181–192. doi: 10.1093/fqsafe/fyaa025. [DOI] [Google Scholar]
  20. Sasaki T., Koshi E., Take H., Michihata T., Masachika M., Enomoto T. Characterisation of odorants in roasted stem tea using gas chromatography-mass spectrometry and gas chromatography-olfactometry analysis. Food Chemistry. 2017;220:177–183. doi: 10.1016/j.foodchem.2016.09.208. [DOI] [PubMed] [Google Scholar]
  21. Sheng C., Lu M., Liu Q., Zhou H., Xiong Z., Li T., Wei Y., Zhang J., Ke H., Wu Y., Wang Y., Ning J. Differences in the aroma quality of large-leaf yellow tea subjected to different roasting methods. LWT – Food Science and Technology. 2024;204 doi: 10.1016/j.lwt.2024.116475. [DOI] [Google Scholar]
  22. Shi Y., Wang M., Dong Z., Zhu Y., Shi J., Ma W., Lin Z., Lv H. Volatile components and key odorants of Chinese yellow tea (Camellia sinensis) LWT- Food Science and Technology. 2021;146(33) doi: 10.1016/j.lwt.2021.111512. [DOI] [Google Scholar]
  23. Sivadier G., Ratel J., Engel E. Latency and persistence of diet volatile biomarkers in lamb fats. Journal of Agricultural and Food Chemistry. 2009;57(2):645–652. doi: 10.1021/jf802467q. [DOI] [PubMed] [Google Scholar]
  24. Song H., Liu J. GC-O-MS technique and its applications in food flavor analysis. Food Research Internatinal. 2018;114:187–198. doi: 10.1016/j.foodres.2018.07.037. [DOI] [PubMed] [Google Scholar]
  25. Srivastava S., Sadistap S. Data procesing approaches and strategies for non-destructive fruits quality inspection and authentication: A review. Journal of Food Measurement and Characterization. 2018;12(4):2758–2794. doi: 10.1007/s11694-018-9893-2. [DOI] [Google Scholar]
  26. Uenojo M., Pastore G.M. Β-Carotene biotransformationt o obtain aroma compounds. Food Science and Technology. 2010;30(3):822–827. doi: 10.1590/S0101-20612010000300039. [DOI] [Google Scholar]
  27. Wang H., Chen J., Ren P., Zhang Y., Onyango S.O. Ultrasound irradiation alters the spatial structure and improves the antioxidant activity of the yellow tea polysaccharide. Ultrasonics Sonochemistry. 2021;70 doi: 10.1016/j.ultsonch.2020.105355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Wang K., Liu F., Liu Z., Huang J., Xu Z., Li Y., Chen J., Gong Y., Yang X. Comparison of catechins and volatile compounds among different types of tea using high performance liquid chromatograph and gas chromatograph mass spectrometer. International Journal of Food Science & Technology. 2011;46(7):1406–1412. doi: 10.1111/j.1365-2621.2011.02629.x. [DOI] [Google Scholar]
  29. Wang K., Ruan J. Analysis of chemical components in green tea in relation with perceived quality. International Journal of Food Science & Technology. 2009;44(12):2476–2484. doi: 10.1111/j.1365-2621.2010.02224.x. [DOI] [Google Scholar]
  30. Wei Y., Fang S., Jin G., Ni T., Hou Z., Li T., Deng W., Ning J. Effects of two yellowing process on colour, taste and nonvolatile compounds of bud yellow tea. International Journal of Food Science and Technology. 2020;55(8):2931–2941. doi: 10.1111/ijfs.14554. [DOI] [Google Scholar]
  31. Wei Y., Li T., Xu S., Ni T., Deng W.W., Ning J. The profile of dynamic changes in yellow tea quality and chemical composition during yellowing process. LWT – Food Science and Technology. 2021;139 doi: 10.1016/j.lwt.2020.110792. [DOI] [Google Scholar]
  32. Xu M., Wang J., Gu S. Rapid identification of tea quality by E-nose and computer vision combining with a synergetic data fusion strategy. Journal of Food Engineering. 2019;241:10–17. doi: 10.1016/j.jfoodeng.2018.07.020. [DOI] [Google Scholar]
  33. Yang P., Song H., Lin Y., Guo T., Wang L., Granvogl M., Xu Y. Differences of characterisitic aroma compounds in Rougui tea leaves with different roasing temperatures analyzed by switchable GC-O-MS and GC×GC-O-MS and sensory evaluation. Food & Function. 2021;12:4797. doi: 10.1039/d1fo00165e. [DOI] [PubMed] [Google Scholar]
  34. Yang Z., Baldermann S., Watanabe N. Recent studies of the volatile compounds in tea. Food Research International. 2013;53(2):585–599. doi: 10.1016/j.foodres.2013.02.011. [DOI] [Google Scholar]
  35. Yin X., Wei Y., Li T., Zhang J., Zou L., Cui Q., Lu C., Ning J. Heterocyclic compounds formation in large-leaf yellow tea induced by the Maillard reaction at different roasting temperatures. LWT – Food Science and Technology. 2023;182 doi: 10.1016/j.lwt.2023.114856. [DOI] [Google Scholar]
  36. Yuan H., Chen X., Shao Y., Cheng Y., Yang Y., Zhang M., Hua J., Li J., Deng Y., Wang J., Dong C., Jiang Y., Xie Z., Wu Z. Quality evaluation of green and dark tea grade using electronic nose and multivariate statistical analysis. Journal of Food Science. 2019;84(12):3411–3417. doi: 10.1111/1750-3841.14917. [DOI] [PubMed] [Google Scholar]
  37. Zhai X., Hu Y., Pei Z., Yu J., Li M., Zhang L.…Wan X. Insights into the key odorants in large-leaf yellow tea (Camellia sinensis) by application of the sensomics approach. Journal of Agricultural and Food Chemistry. 2023;71:690–699. doi: 10.1021/acs.jafc.2c05881. [DOI] [PubMed] [Google Scholar]
  38. Zhao G., Teng J., Dong R., Ban Q., Yang L., Du K.…Ren Z. Alleviating effects and mechanisms of action of large-leaf yellow tea drinking on diabetes and diabetic nephropathy in mice. Food Science and Human Wellness. 2023;12:1660–1673. doi: 10.1016/j.fshw.2023.02.023. [DOI] [Google Scholar]
  39. Zhi R., Zhao L., Zhang D. A framework for the multi-level fusion of electronic nose and electronic tongue for the quality assessment. Sensors. 2017;17(5):1007–1022. doi: 10.3390/s17051007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Zhu H., Ye Y., He H., Dong C. Evaluation of green tea sensory quality via process characteristics and image information. Food and Bioproducts Processing. 2017;102:116–122. doi: 10.1016/j.fbp.2016.12.004. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary material

mmc1.docx (400.4KB, docx)

Data Availability Statement

Data will be made available on request.

Articles from Food Chemistry: X are provided here courtesy of Elsevier

RESOURCES