Manufacturing process differences give Keemun black teas their distinctive aromas

Highlights

• •Four Keemun black tea types with different processing possess distinctive aromas.
• •Volatiles with flowery or sweet notes had high aroma intensity in Keemun black tea.
• •Congou Keemun black teas with a refining stage had high volatiles concentration.
• •Phenylethyl alcohol oxidation resulted in phenylacetaldehyde formation.
• •Processing changes the proportion of flowery to green volatile compounds.

Abstract

Traditional Keemun black tea is also known as Congou black tea (CF). Over the last three decades, three other types of Keemun black tea (Jinzhen, JZ; Maofeng, MF; Xiangluo, XL) made by different processing have been introduced into the tea market. Total CF volatile concentrations ranged from 1666.3 to 2185.7 μg/L, followed by XL (1193.5–1916.1 μg/L), JZ (1058.9–1811.0 μg/L), and MF (987.5–1518.0 μg/L) tea infusions. A total of 79 volatiles in tea infusions was identified by two methods, among which fourteen with OAVs > 1 were identified and OAVs proportion of volatiles with flowery, fruity, or sweet notes to those with other notes differed in four Keemun black teas (CF = 6.58:1, MF = 5.16:1, JZ = 4.04:1, XL = 5.11:1). Phenylethyl alcohol oxidation resulted in phenylacetaldehyde formation which is the characteristic odorant in Keemun black tea. We clearly show that changes in tea processing gives the distinctive aroma to different Keemun black teas.

1. Introduction

Black tea as a fully fermented tea is discriminated against other tea products, such as non-fermented green tea and semi-fermented oolong tea. Total volatile compound concentrations in black tea products are higher than in other tea products that do not undergo fermentation (Feng et al., 2019). Moreover, volatile compounds in black tea are abundant and complex on account of enzymatic reactions (Tan et al., 2016, Qu et al., 2020), contributing to its charming and pleasant aroma. Keemun black tea, produced in the Qi-men area, with a high honey-like, flowery and fruity aroma (also described as ‘Keemun aroma’), is different from black tea manufactured in other regions of China and from other countries (Wang and You, 1996, Kang et al., 2019). For example, Dianhong black teas manufactured in China’s Yunnan province have a caramel-like, roasted and flowery odor (Wang et al., 2017); Assam black teas from India exhibit malty and rose-like scents (Kang et al., 2019); Ceylon black tea produced in Sri Lanka has a typical floral and juicy-sweet notes (Kumazawa, Wada, & Masuda, 2006).

Aroma formation in tea products can be attributed to tea varieties, field management, local climate, and processing (Yener et al., 2016, Zheng et al., 2018, Zhao et al., 2017, Ohno et al., 2011). Recent studies using phylogenetic analysis separated ancestral Camellia sinensis var. assamica, C. sinensis var. assamica and var. sinensis tea varieties into three subgroups (Zhang et al., 2021, Xia et al., 2020). Among them, the majority of processed Chinese tea varieties belong to the C. sinensis var. sinensis group, except for Yunan and Guangzhou where tea varieties mostly belong to the C. sinensis var. assamica group. Assam and Darjeeling black teas from India were made by fresh leaves of C. sinensis var. assamica varieties (Mahanta, Baruah, Owuor & Murai, 1988). It is generally recognized that the characteristic aroma of Keemun black tea depends on the premium tea variety ‘Zhuyezhong’ belonging to the subgroup of C. sinensis var. sinensis (Yang et al., 2022, Wang and You, 1996).

Aroma precursors, such as fatty acids, glycosides, amino acids, and carbohydrates contribute the most to black tea aroma (∼92%) (Feng et al., 2019). Phenylacetaldehyde (with honey-like, floral trait), a derivant of l-phenylalanine was identified as the definitive odorants in Keemun black tea (Kang et al., 2019). Geraniol (with rose-like scent) is the most abundant glycosidically bound volatile in Keemun black tea (Wang & You, 1996). Twenty-one volatile compounds, such as 2-hexenal, benzaldehyde, linalool, methyl salicylate, geraniol, hexanal, (Z)-3-hexen-1-ol, and phenylacetaldehyde had a high content in Keemun, Assam, Ceylon, and Darjeeling black teas (Kang et al., 2019), but their concentrations in the four black teas were different. Different grades of Keemun black tea showed that 11 volatile compounds (rOAV > 1) were the key volatile compounds, particularly geraniol, linalool, phenylacetaldehyde, and methyl salicylate (Su, He, Zhou, Li & Zhou, 2022).

Nowadays, consumers are beginning to pursue the premium quality and security of black tea products, such as Gold Jun Mei (Jin Jun Mei) (DeBernardi & Ma, 2022), which is made from tender buds using innovative processing methods. A new method originating from oolong tea processing, namely, additional turning over stages before fermentation gives a peach-like aroma to black tea products (Wu, Zhou, Zhang, Huang, Ou & Sun, 2023). Different withering methods were conducted to determine what affects Keemun black tea aroma quality (Huang et al., 2022). Over the last three decades, apart from Congou Keemun black tea (CF), the Keemun black tea industry has developed three new black teas, namely, Jinzhen Keemun (JZ), Xiangluo Keemun (XL) and Maofeng Keemun (MF). These four Keemun black teas have different processing stages (Fig. 1). Based on sensory evaluation, although CF, JZ, MF and XL black teas had the characteristic Keemun black tea fragrance, distinctive flavor differences were perceived among them. CF black tea generally had a high flowery and fruity odor, while JZ tea had strong sweet and slight flowery characteristics. Whether the mechanism underlying this slight difference between the four Keemun black tea aromas was caused by tea processing still needs to be elucidated.

Fig. 1.

Processing stages for four Keemun black tea types.

We used gas chromatography–mass spectrometry (GC–MS) coupled with headspace micro extraction (HS-SPME), and gas chromatography olfactometry-mass spectrometry (GC-O-MS) coupled with the stir bar sorptive extraction (SBSE) method to identify aroma profiles. Our aims were to (i) profile volatile metabolites of the four Keemun black tea types and (ii) explore the differential volatiles and mechanisms underlying the distinctive odor differences between them.

2. Materials and methods

2.1. Black tea samples

Forty black tea samples from the Qi-men area (latitude/longitude 29°51′/117°43′) were used in this study: thirteen MF, twelve JZ, nine XL, and six CF black tea samples. After sensory evaluation (Table S1), six products for each black tea (24 tea products) were selected for aroma profile analysis.

2.2. Manufacturing processes of Keemun black tea

The Keemun black tea manufacturing process is as follows. For the four Keemun black tea types, the withering, rolling, and fermentation stages were the same. Fresh shoots (one bud and one leaf) of tea variety ‘Zhuyezhong’ were withered with hot blower for 5–6 h until their moisture content was about 58–60%. A barrel-type rolling machine (6CR-55, diameter 55 cm, Sunyoung Machinery Co., Ltd, Zhejiang, Quzhou, China) was then used for rolling the withering leaves at 35–45 rpm for 1.5–2 h. Lastly, the rolled leaves were fermented in a fermentation room (23–25 °C, moisture content ≥ 95%) for 3–3.5 h.

Congou Keemun black tea is a traditional black tea product in the Qi-men area and well-known around the world. The refining stage for this tea is different from the other three types of black tea, and is a complicated and long-duration process. After the drying stage (first-drying, 110 °C, 20–30 min; final-drying, 80–90 °C, 30–40 min), a large, combined machinery group consisting of screening, cutting, winnowing modules (Sunyoung Machinery Co., Ltd, Zhejiang, Quzhou, China) separates the primary black teas into different sizes (based on length and thickness). A re-drying stage (70–80 °C, 30 min) is conducted before the blending stage. During the blending stage, premium parts are set as the raw materials for high grade of Keemun black tea and other parts of different sizes are used to blend other grades. As mentioned above, black tea samples after the drying stage were the primary materials for Maofeng Keemun black tea. A winnowing machine (EF-40, Sunyoung Machinery Co., Ltd, Zhejiang, Quzhou, China) is used to remove impurities such as pieces of old leaves and stems and then these black tea products are prepared to flow into the tea market. Xiangluo Keemun black tea is a kind of curly black tea product. After the first-drying stage (110 °C, 10–20 min), a pan-firing machine (6CCQ-50, Chunjiang Tea Machinery Co., Ltd, Hangzhou, Zhejiang, China) is used for modeling tea products at 100–110 °C for 30 min and the final-drying stage is conducted at 80–90 °C for 30 min. Jinzhen Keemun black tea is a kind of black tea product with a straight appearance. A carding machine (6CLZ-60/8(11), Sunyoung Machinery Co., Ltd, Quzhou, Zhejiang, China) is used for modeling of tea products at 120–130 °C for 30–60 min after the first-drying stage.

2.3. Sensory evaluation

Six tea experts were recruited to conduct the sensory evaluation. Each black tea product (3 g) was infused with boiling distilled water (150 mL, 98 °C), and the tea infusion was poured after brewing for 5 mins. Experts evaluated and described the quality of each black tea product according to the evaluation criteria (GB/T 23776-2018).

2.4. Chemicals

Analytical standards of volatile compounds used in this study were purchased from Aladdin Company (Shanghai, China), including (E)-2-hexenal (≥98.0%), heptanal (≥99.5%), benzaldehyde (≥99.5%), myrcene (≥95.0%), limonene (≥95.0%), benzyl alcohol (≥99.5%), ocimene (≥90.0%), phenylacetaldehyde (≥95.0%), linalool (≥98.0%), nonanal (≥95.0%), neral (≥97.0%), geraniol (≥99.0%), methyl geranate (≥94.0%), (Z)-3-hexenyl hexanoate (≥98.0%), β-ionone (≥97.0%), cedrol (≥98.0%), ethyl decanoate (≥99.5%), and n-alkane (≥99.0%). Analytical standards of hexanal (≥98.0%), β-cyclocitral (≥95.0%), α-ionone (≥95.0%), decanal (≥95.0%), jasmone (≥92.0%), nerolidol (≥97.0%), citral (≥98.0%), methyl salicylate (≥99.0%), 2-phenylethyl alcohol (≥98.0%), and linalool oxide mixture (≥97.0%) were purchased from Tokyo Chemical Industry Co., Ltd (Tokyo, Japan).

2.5. GC–MS analysis

Quantification of volatile compounds in tea infusions were made according to Flaig et al. (2020) with a slight modification. Tea infusions (3 g tea samples; 150 mL 98 °C boiling water; 300 mg NaCl; 5 min) were prepared. After cooling to room temperature, 10 mL aliquots containing 10 μL 10 μg/mL ethyl caprate (for repeatability) were used for volatiles extraction using a SPME (solid phase microextraction) fiber (30/50 μm DVB/CAR/PDMS, 57348-U, Supelco, Merck KGaA, Darmstadt, Germany) at 40 °C for 40 min, with an agitation rate of 1000 rpm.

After an absorption stage, the SPME fiber was immediately desorbed at 250 °C in a split mode (10:1). A SHIMADZU GC–MS system (QP2020 NX, Japan) equipped with a DB-5 MS column (30 × 0.25 mm × 0.25 μm, Agilent, USA) was used to conduct the volatile analysis. The GC gradient program was as follows: 40 °C for 5 min; from 40 °C to 110 °C at a rate of 3 °C/min and held for 3 min; from 110 °C to 200 °C at a rate of 5 °C/ min and held for 1 min; from 200 °C to 260 °C at a rate of 15 °C/min and held for 3 mins. The flow rate was 1 mL/min constant helium. MS operating conditions were: 230 °C ion source, 250 °C interface, 0.1 kV voltage, and 30 to 600 m/z scan range at 5 scans/sec. Triplicate or six injections were obtained.

2.6. Volatile compounds identification and quantification

Volatile compound identification was performed according to Zhou et al., (2019). n-Alkane was used to calculate the retention indices of volatile compounds and then compared with the retention time and mass spectrum found in the NIST database and their standards.

Quantification of volatiles was conducted according to Han et al., (2016) with a slight modification. Deodorized black tea samples were obtained using a distilling apparatus (RE-2000A, Yarong Company, Shanghai, China) (Wang et al., 2020). Distilling stages were repeated until the overall abundance of volatile compounds in the tea infusion had a very low response value when detected by GC–MS (Fig. S1). Tea infusion prepared by deodorized black teas was set as the substance to perform the volatile compound calibration curves.

Five points of twenty-eight compounds were used to conduct the calibration curves. These standards were extracted and analyzed in the same way as in Section 2.6. Finally, twenty-eight volatile compound calibration curves were established (Table S2). Twenty-seven volatiles were quantified against their authentic standard calibration curves and the others without calibration curves were quantified based on ethyl caprate calibration curves. Volatile concentrations were expressed as μg/L in tea infusions (W/V, 1:50).

2.7. GC-O analysis

The GC-O analysis was conducted according to Wang, Ma, Shi, Zhu, Lin, & Lv (2020). Tea sample for GC-O analysis was a mixture of ten premium Keemun black tea samples. Black tea volatile compounds were extracted by the SBSE (stir bar sorptive extraction) method using a PDMS (polydimethyl siloxane) twister (10 × 1 mm, Gerstel, Germany). Aqueous infusions were prepared for volatile extraction (1000 mg tea powder, 500 mg NaCl, and10 mL 98 °C hot water). The PDMS twister was used to absorb volatiles on a stirrer (SP200-2T, Miu Instruments Co. Ltd, Hangzhou, China) at 70 °C for 30 min, with an agitation rate of 1000 rpm. Before GC-O analysis, the twister was washed with distilled water and dried with a clean, lint-free tissue and then transferred to a thermal desorption tube.

A GC–MS system (6890A-5975C, Agilent, USA) equipped with a sniffing port (ODP2; Gerstel GmbH & Co. KG, Germany) was used to analyze the aroma profiles. A split mode of 1:1 ratio was conducted between the olfactory detection port and the MS. The GC column, temperature program and flow rate were HP-5 MS column (30 × 0.25 mm × 0.25 μm, Agilent, USA), 50 °C (held for 3 min) and then ramped to 265 °C (held 5 min) at a rate of 4 °C/min, 1.6 mL/min constant nitrogen (99.99%), respectively. The temperature of the GC-O injector and transfer line were 150 °C and 220 °C, respectively.

Volatile aroma descriptions were performed by an experienced panel, consisting of six healthy, non-smoking judges. The assessors were trained for chemical standard odor characteristics for up to 90 h, as well as the intensity scales with different aroma standard concentrations (four different concentrations for each compound were made to distinguish ‘weak’, ‘moderate’, ‘strong’, and ‘extremely strong’). There are five odor attributes applied by aroma standards, including hexanal and methyl salicylate for ‘green’ or ‘grassy’ attribute, linalool, geraniol, and β-ionone for ‘floral’ attribute, benzaldehyde for ‘fruity’ note, phenylacetaldehyde for ‘honey-like’ attribute, and heptanal and (E)-2-octenal for ‘nutty’ attribute (Wang et al., 2022, Feng et al., 2020). The assessors recorded aroma characteristics and scored odor intensities according to a 4-point scale (1, weak; 2, moderate; 3, strong; 4, extremely strong) (Wang et al., 2020). Odorant descriptions were identified by comparing their aroma characteristics, retention indices (RIs), and mass spectra with those of the corresponding standards. The average score of the six members were set as the aroma intensity of volatile compounds. The experiment was performed in triplicate.

2.8. Statistical analysis

Statistical analysis was conducted to differentiate significant or non-significant differences (p < 0.05) in volatile concentrations using ANOVA (SPSS statistics 19.0 software). To obtain the differential components between black tea samples, the software package SIMCA 13.0 (Umetrics, Sweden) was used to perform principal component analysis (PCA).

3. Results and discussion

3.1. Volatile compounds identified in tea infusion by HS-SPME-GC-MS

Aroma extraction by HS-SPME preserves the original properties of volatile compounds (Yang et al., 2013). In this study, fifty volatile compounds with high confidence were identified and quantified in tea infusions by HS-SPME-GC-MS (Table 1). It is well-known that response of compounds in mass spectrometry is influenced by their structures, for instance benzyl alcohol and β-ionone (Fig. S2). Hence, twenty-seven volatiles were quantified by their authentic standard curves and the remaining twenty-three compounds were quantified by standard ethyl caprate curve.

Table 1.

Concentrations of 50 odorants in four Keemun black tea infusions.

NO.	Compoundsa	RIb	RIc	Concentrations (μg/L)d
				MF	JZ	CF	XL
1	Dimethyl sulfide	577	565	0.21–1.68	0.83–1.18	0.50–1.13	0.56–1.31
2	3-Methyl butanal	682	686	1.75–3.89	3.62–4.16	2.60–4.50	3.08–5.97
3	2-Methyl butanal	686	688	1.61–4.01	2.65–4.02	2.03–4.87	2.47–5.63
4	Pentanal	702	705	0.46–0.79	0.74–1.10	0.55–0.94	0.52–0.92
5	2-ethyl-Furan	713	713	0.49–1.32	1.04–1.42	0.52–1.20	0.68–1.46
6	1-Pentanol	777	770	0.25–0.49	0.39–0.92	0.43–0.79	0.30–0.59
7	Hexanal*	812	812	5.10–14.09	11.13–16.38	7.29–14.54	7.93–15.12
8	(E)-2-Hexenal*	855	855	9.18–26.99	11.48–19.03	17.65–28.20	9.70–27.32
9	(Z)-3-Hexen-1-ol	858	856	0.89–2.35	0.90–1.99	1.47–2.54	0.90–2.43
10	(E)-2-Hexen-1-ol	869	868	0.18–0.26	0.15–0.23	0.16–0.30	0.14–0.21
11	1-Hexanol	872	874	0.35–0.71	0.39–0.98	0.49–0.75	0.35–0.59
12	2-Heptanone	890	889	0.22–0.35	0.20–0.50	0.26–0.51	0.17–0.34
13	Heptanal*	903	901	0.28–0.57	0.41–0.59	0.43–0.67	0.39–0.57
14	Benzaldehyde*	958	960	5.11–11.31	3.96–12.91	8.83–19.12	4.31–16.38
15	1-Octen-3-ol	980	983	0.17–0.30	0.19–0.44	0.21–0.39	0.19–0.33
16	6-methyl-5-Hepten-2-one	984	988	0.35–0.60	0.43–0.71	0.53–0.84	0.51–0.77
17	β-Myrcene*	989	991	0.38–0.77	0.28–0.62	0.00–1.06	0.37–0.62
18	Octanal	1010	1005	0.19–0.33	0.23–0.37	0.26–0.40	0.21–0.37
19	d-Limonene*	1033	1032	0.13–0.21	0.14–0.24	0.21–0.31	0.19–0.23
20	Benzyl alcohol*	1037	1033	259.45–590.37	388.03–783.99	500.11–677.53	361.97–874.50
21	3,5-Octadien-2-ol	1041	1039	0.00–0.30	0.16–0.30	0.10–0.30	0.15–0.26
22	Phenylacetaldehyde*	1044	1043	41.33–91.41	33.01–126.55	44.71–164.31	7.83–111.04
23	1-Ethyl-2-pyrrolecarbaldehyde	1047	1046	0.14–0.31	0.22–0.33	0.17–0.36	0.20–0.65
24	β-Ocimene*	1050	1046	1.90–12.15	0.00–5.49	6.50–15.32	4.21–9.03
25	(E)-2-Octenal	1059	1058	0.12–0.27	0.18–0.31	0.16–0.30	0.18–0.27
26	(Z)-linalool oxide (furanoid)*	1070	1070	33.16–71.04	27.68–81.23	78.56–103.81	43.33–76.31
27	(E)-Linalool oxide (furanoid)*	1084	1082	150.70–244.99	84.33–243.58	200.61–338.30	154.11–191.37
28	Linalool*	1098	1097	63.31–87.18	40.65–75.00	65.65–86.49	38.14–77.66
29	Hotrienol	1090	1104	0.30–0.52	0.25–0.54	0.38–0.66	0.46–1.68
30	Nonanal*	1092	1098	0.08–0.50	0.09–0.53	0.21–0.91	0.27–0.78
31	Phenylethyl alcohol*	1097	1103	203.71–284.66	186.63–346.54	229.67–362.99	217.15–328.77
32	5-Ethyl-6-methyl-3(E)-hepten-2-one	1140	1143	0.15–0.28	0.24–0.36	0.17–0.31	0.14–0.24
33	Linalool oxide (pyranoid)	1170	1173	0.25–0.31	0.24–0.38	0.28–0.41	0.28–0.40
34	Methyl salicylate*	1186	1185	27.16–39.84	19.14–34.12	29.17–40.98	23.15–32.04
35	Decanal*	1176	1185	0.35–0.61	0.37–0.72	0.43–0.98	0.69–1.12
36	β-Cyclocitral*	1189	1200	0.24–0.53	0.32–0.66	0.41–0.65	0.31–0.67
37	Neral*	1216	1231	3.77–8.64	5.84–8.40	5.99–9.84	4.61–8.53
38	Geraniol*	1234	1234	138.63–225.33	144.05–233.83	219.58–260.06	173.79–241.05
39	(E)-Citral*	1254	1267	3.29–7.65	5.21–7.65	5.44–8.68	4.39–7.20
40	(E, E)-2,4-Decadienal	1293	1295	0.00–0.20	0.11–0.13	0.00–0.14	0.07–0.15
41	(E)-Geranic acid methyl ester*	1299	1301	0.12–0.16	0.12–0.24	0.13–0.19	0.12–0.15
42	(Z)-Hexanoic acid, 3-hexenyl ester*	1375	1376	0.17–0.37	0.15–0.41	0.22–0.47	0.19–0.47
43	(Z)-Jasmone*	1386	1389	5.37–14.22	5.99–14.28	7.80–21.07	6.90–17.41
44	α-Ionone*	1406	1401	0.00–0.13	0.00–0.13	0.10–0.12	0.10–0.12
45	Geranyl acetone	1438	1435	0.13–0.28	0.20–0.39	0.29–0.47	0.36–0.39
46	β-Ionone*	1473	1493	0.00–0.63	0.07–0.70	0.53–1.16	0.27–0.99
47	2,4-Di-tert-butylphenol	1493	1502	0.18–0.24	0.19–0.22	0.20–0.23	0.19–0.24
48	δ-Cadinene	1508	1509	0.06–0.12	0.11–0.12	0.11–0.12	0.10–0.12
49	Nerolidol*	1555	1558	0.01–0.13	0.02–0.10	0.04–0.12	0.06–0.15
50	Cedrol*	1600	1597	0.48–0.91	0.48–1.60	0.44–0.95	0.55–0.76

^aVolatile compound labeled with * was quantified using its authentic standard calibration curve.

^bRetention index calculated in this study.

^cRetention index from the NIST library.

^dConcentration of volatile compounds in tea infusions (tea sample: water, W/V, 1:50).

Volatile total concentrations in four types of Keemun black tea was quite different. CF tea infusions had the highest total concentration (1666.3–2185.7 μg/L), followed by XL (1193.5–1916.1 μg/L), JZ (1058.9–1811.0 μg/L) and MF (987.5–1518.0 μg/L) black tea infusions. Clearly, glycosidically bound volatiles (GBVs) were the most concentrated compounds in Keemun black tea infusions, especially benzyl alcohol, phenyl alcohol, geraniol, linalool, linalool oxides, and methyl salicylate. In tea plants, the aglycone part of GBVs mainly comprised three classes: monoterpene alcohols (e.g., linalool, geraniol, and linalool oxide), aromatic alcohols (e.g., benzyl alcohol and phenyl alcohol) and phenolic compounds (e.g., methyl salicylate) (Cui et al., 2016).The ratio of aromatic alcohols in glucoside (45%) and primeveroside (55%) forms is almost 1:1, whereas the primeveroside-type of monoterpenes make up almost 98% compared to the glucosides-type. After the fermentation stage, primeverosides had almost disappeared (about 96%), but 70% glucosides were maintained (Wang, Kurasawa, Yamaguchi, Kubota & Kobayashi, 2001). Our study showed that GBVs were highly concentrated in Keemun black tea, suggesting that they are abundant in fresh Keemun black tea materials and that processing also benefited the release of these aglycone parts with the participation of glucosidases.

Phenylacetaldehyde and benzaldehyde also had a high concentration in Keemun black tea. In plants, there are three pathways for phenylacetaldehyde biosynthesis from l-phenylalanine: (1) catalyzed by AADC (aromatic amino acid decarboxylase); (2) catalyzed by AAAT (aromatic amino acid aminotransferase) and PPDC (phenylpyruvic acid decarboxylase); and (3) catalyzed by CYP79D73 and an unknown enzyme (Zeng et al., 2023). Phenylacetaldehyde was highly concentrated in Keemun black tea, probably contributing to the vigorous biosynthesis. Furthermore, we found that phenylacetaldehyde abundance increased with black tea processing, especially after the rolling stage (Fig S3). Supplemental experiments showed that benzyl alcohol and phenyl alcohol oxidation could produce benzaldehyde and phenylacetaldehyde with or without crude enzymes extracted from tea shoots (Fig S3), suggesting that high phenylacetaldehyde and benzaldehyde concentration may be attributed to the abundant GBVs in Keemun black tea. After the rolling stage, the primeveroside forms of GBVs were hydrolyzed and released aglycones, which were oxidized to form aldehydes during the fermentation stage. Oddly, oxidation of 1-octen-3-ol produced 1-octen-3-one was dramatically enhanced after crude enzymes extracted from tea shoots were added (Fig. S4), suggesting that 1-octen-3-ol oxidation is not only a spontaneous reaction, but an enzymatic reaction during tea processing.

Principle component analysis showed that CF black teas were separated from JZ, MF, and XL (Fig. 2A) and the model validation proved it (Fig. 2B). MF Keemun black tea has often been regarded as the raw materials of Congou Keemun black tea (Fig. 1). Herein, PCA and HCA (hierarchical clustering analysis) results indicated that CF tea volatile profiles were differentiated from those of MF teas (Fig. 2C and D). Between the two teas, differential metabolites were mainly GBVs, and their concentrations were higher in CF tea infusions than in MF (Fig. 2E). Geranyl β-primeveroside was highly concentrated in young leaves (Zhou et al., 2022), while linalyl β-primeveroside and 2-phenylethyl β-primeveroside were more abundant in tender stems and young buds (Ohgami et al., 2015). It was notable that removing pieces of old leaves and stems in the refining stage could benefit aroma quality.

Fig. 2.

Principal component analysis and identification of differential volatile compounds. A) Principal component analysis of volatiles in four Keemun black tea infusions. B) The validate model for PCA. C) PCA and (D) HCA (hierarchical clustering analysis) between CF and MF Keemun teas. E) Differential volatile compounds.

3.2. Volatile compounds identified by GC-O-MS

GC-O coupled with mass spectrometry is a powerful tool in identifying volatile compound attributes and contributions to overall aroma. The SBSE method has good repeatability and universality for volatiles analysis (Wang et al.,2020). Fifty-four compounds were identified by SBSE-GC-O-MS (Table 2). Finally, a total of 79 volatile compounds were identified in Keemun black tea infusions, among which 25 common volatiles were identified by HS-SPME-GC–MS and SBSE-GC-O-MS. There were twenty-eight compounds with high aroma intensity (AI values ≥ 3), including eight with flowery character, eight with green, grassy, or fresh odors, ten with sweet or honey-like aroma and one each of woody and fatty odor. Clearly, volatile compounds with flowery or sweet characteristics were abundant in Keemun black tea.

Table 2.

Volatile compounds identified by SBSE-GC-O-MS.

No.	Compounds	CAS	RIa	AIb	Odor qualityc	IMd
1	3-Methyl butanal	590-86-3	709	2.33	Pungent, unpleasant	MS/RI/Std
2	2-Methyl butanal	96-17-3	715	2.33	Pungent, unpleasant	MS/RI/Std
3	Hexanal	66-25-1	804	2.17	Green, grassy, fresh	MS/RI/Std
4	Heptanal	111-71-7	905	3.25	Sweet, balsamic, creamy	MS/RI/Std
5	2,5-Dimethyl pyrazine	123-32-0	927	3.00	Sweet, balsamic, creamy	MS/RI
6	Hexanoic acid, methyl ester	106-70-7	925	2.25	Sweet, balsamic, creamy	MS/RI
7	Dimethyl trisulfide	3658-80-8	974	3.17	Green, grassy, fresh	MS/RI
8	1-Octen-3-ol	3391-86-4	982	3.40	Green, grassy, fresh	MS/RI
9	6-Methyl-5-heptene-2-one	110-93-0	987	3.33	Green, grassy, fresh	MS/RI
10	β-Myrcene	123-35-3	991	2.83	Green, grassy, fresh	MS/RI/Std
11	2-Pentyl furan	3777-69-3	992	2.83	Green, grassy, fresh	MS/RI
12	(E, Z)-2,4-Heptadienal	4313-02-4	999	2.33	Green, grassy, fresh	MS/RI
13	Octanal	124-13-0	1003	3.00	Green, grassy, fresh	MS/RI/Std
14	(Z)-β-Ocimene	3338-55-4	1036	3.00	Floral	MS/RI/Std
15	Phenylacetaldehyde	122-78-1	1047	3.50	Floral, honey-like	MS/RI/Std
16	1-Ethyl-2-pyrrolecarbaldehyde	2167-14-8	1053	2.83	Green, grassy, fresh	MS/RI
17	(E)-2-Octenal	2548-87-0	1058	3.00	Fatty, oil	MS/RI
18	(Z)-linalool oxide (furanoid)	5989-33-3	1091	2.33	Floral	MS/RI/Std
19	Nonanal	124-19-6	1100	3.33	Green, grassy, fresh	MS/RI/Std
20	Linalool	78-70-6	1110	3.67	Floral	MS/RI/Std
21	Hotrienol	29957-43-5	1112	3.40	Floral	MS/RI/Std
22	5-Ethyl-6-methyl-3(E)-hepten-2-one	57283-79-1	1149	2.83	Green, grassy, fresh	MS/RI
23	(E, E)-2,6-Nonadienal	17587-33-6	1156	2.83	Green, grassy, fresh	MS/RI
24	(Z)-3-Nonen-1-ol	10340-23-5	1158	2.83	Green, grassy, fresh	MS/RI
25	(E)-2-Nonenal	18829-56-6	1162	3.17	Green, grassy, fresh	MS/RI
26	Linalool oxide (pyranoid)	39028-58-5	1173	2.83	Woody, herbal	MS/RI/Std
27	(E)-Butanoic acid, 3-hexenyl ester	53398-84-8	1186	2.83	Woody, herbal	MS/RI
28	Methyl salicylate	119-36-8	1207	3.00	Green, grassy, fresh	MS/RI/Std
29	(E, E)-2,4-Nonadienal	5910-87-2	1217	2.80	Fatty, oil	MS/RI
30	Neral	106-25-2	1232	2.67	Green, grassy, fresh	MS/RI/Std
31	Geraniol	106-24-1	1238	3.00	Floral	MS/RI/Std
32	(E)-Citral	106-26-3	1253	2.67	Green, grassy, fresh	MS/RI/Std
33	2-Undecanone	112-12-9	1296	3.00	Floral	MS/RI
34	(E)-Geranic acid methyl ester	105-86-2	1303	3.17	Sweet, balsamic, creamy	MS/RI/Std
35	(E, E)-2,4-Decadienal	25152-84-5	1321	2.83	Woody, herbal	MS/RI
36	Eugenol	97-53-0	1357	3.33	Sweet, balsamic, creamy	MS/RI
37	γ-Nonanolactone	104-61-0	1369	3.00	Sweet, balsamic, creamy	MS/RI
38	(E)-β-Damascenone	23726-93-4	1389	3.00	Sweet, balsamic, creamy	MS/RI/Std
39	(Z)-Jasmone	488-10-8	1406	2.67	Floral	MS/RI/Std
40	Dimethyl anthranilate	85-91-6	1417	3.17	Sweet, balsamic, creamy	MS/RI
41	α-Ionone	127-41-3	1432	3.00	Floral	MS/RI/Std
42	Phenethyl butyrate	103-52-6	1444	2.25	Sweet, balsamic, creamy	MS/RI
43	Coumarin	91-64-5	1455	2.67	Sweet, balsamic, creamy	MS/RI
44	γ-Decanolactone	706-14-9	1473	2.80	Sweet, balsamic, creamy& Fruity	MS/RI
45	1-Dodecanol	112-53-8	1474	3.00	Sweet, balsamic, creamy	MS/RI
46	(E)-β-Ionone	79-77-6	1491	3.17	Floral	MS/RI/Std
47	(Z)-Jasmin lactone	25524-95-2	1501	3.00	Sweet, balsamic, creamy	MS/RI/Std
48	δ-Dodecalactone	705-86-2	1502	3.00	Sweet, balsamic, creamy	MS/RI
49	(Z)-3-Hexenyl benzoate	25152-85-6	1575	3.00	Green, grassy, fresh	MS/RI/Std
50	Hexyl benzoate	6789-88-4	1581	2.60	Floral	MS/RI
51	Methyl jasmonate	1211-29-6	1653	2.00	Floral	MS/RI/Std
52	α-Cadinol (isomer)	481-34-5	1665	3.00	Woody, herbal	MS/RI
53	Methyl epi-jasmonate	42536-97-0	1683	2.50	Floral	MS/RI
54	Benzyl benzoate	120-51-4	1774	2.33	Floral	MS/RI/Std

^aRetention index from calculated on HP-5 MS capillary column.

^bAroma intensity.

^cOdor description from assessors.

^dIdentification method: (MS) compared with mass spectrum from the NIST library; (RI) Volatile identification by retention index; (Std) Volatile identification by authentic standards.

Linalool had the highest score (AI = 3.67), followed by phenylacetaldehyde (AI = 3.50), 1-octen-3-ol (AI = 3.40), hotrienol (AI = 3.40), 6-Methyl-5-heptene-2-one (AI = 3.33), nonanal (AI = 3.33), eugenol (AI = 3.33), heptanal (AI = 3.25), (E)-β-Ionone (AI = 3.17), dimethyl trisulfide (AI = 3.17), and geraniol (AI = 3.00). Linalool has a flowery odor, which is an important contributor to Darjeeling black tea aroma (Kang et al., 2019). It was reported that geraniol (not linalool) was the most abundant monoterpenoid in Keemun black tea (Wang & You, 1996). In this study, geraniol aroma intensity was lower than linalool, possibly due to the differentiated odor thresholds in air (R-linalool, 0.006 mg/m³; S-linalool, 0.035 mg/m³; geraniol, 0.15 mg/m³) (Van Germet, 2011). Linalool as a type of chiral compound, its enantioform R-linalool has a strong flowery odor and is sensitive to human olfaction. The ratio of R-linalool and S-linalool in tea products is significantly influenced by tea varieties (Mu, Zhu, Lv, Yan, & Lin, 2018). Phenylacetaldehyde (with honey-like, flowery odor) was recognized as the definitive odorant in Keemun black tea compared to Darjeeling, Assam, and Ceylon black tea volatile profiles (Kang et al., 2019). The high aroma intensity of phenylacetaldehyde suggests that it has an important contribution to Keemun black tea aroma. 1-Octen-3-ol possesses green, grassy, and fresh notes, while 1-octen-3-one with a strong mushroom-like odor has a high odor activity in Hanzhong black tea (Chen et al., 2019). According to our supplemental experiments, 1-octen-3-ol high aroma intensity suggests that oxidation degree of Keemun black tea is different from Hanzhong black tea.

3.3. Odor activity values of volatile compounds in four black tea infusions

Volatile compound thresholds in water have been well reported (Van Gemert, 2011, Flaig et al., 2020, Han et al., 2016). Based on recent studies (Flaig et al., 2020, Wang et al., 2020, Zhu et al., 2015), we calculated volatile odor activity values (the ratio of concentrations in tea infusion to their threshold in water). There were fourteen volatile compounds with OAVs greater than one (Table 3). Phenylacetaldehyde, linalool, geraniol, β-ionone, 3-methyl butanal had the highest OAVs. Compared to the SBSE-GC-O-MS results, eugenol, and dimethyl trisulfide odorants (AI > 3) were not identified according to HS-SPME-GC–MS and OAVs. Additionally, 6-Methyl-5-heptene-2-one (AI = 3.33) had a relatively high threshold in water (50 μg/L) resulting in an OAV of c. 0.01–0.02 among tea infusions. Other compounds with OAVs >1 mostly had a relatively high aroma intensity. However, benzyl alcohol and phenyl alcohol were not identified using the SBSE method, whose OAVs were about 2.59–5.90 and 4.53–6.33 in Keemun black tea infusions. Only fourteen common volatile compounds among 184 volatiles were identified by HS-SPME-GC–MS and HS-GC-IMS (Xie et al., 2023), suggesting that the aroma profile of tea samples varies with extraction methods and analytical techniques. For example, geraniol as a characteristic compound in green tea products (Wang et al., 2020), was not identified by the HS-GC-IMS method (Liu et al., 2023). Dimethyl sulfide had high OAVs in all four black tea infusions but was not found in the GC-O-MS result. It has a corn-like odor and is the important contributor to ‘Zhonghuang 1′ green tea aroma (Liao, Yan, Wang, Meng, Zhang & Tong, 2020). A recent study showed that dimethyl sulfide was formed from S-methyl methionine decomposition during tea brewing and the tea processing drying step (Zhai et al., 2022).

Table 3.

Odor activity values (OAV) of key odorants in four Keemun black tea infusions.

No.	Compounds	Odor typea	Threshold(μg/L) b	OAVs c				OAVs > 1
				MF	JZ	CF	XL
1	Dimethyl sulfide	Corn-like	0.3 (Flaig et al., 2020)	0.70–5.59	2.77–3.95	1.68–3.76	1.87–4.38	√
2	3-Methyl butanal	Malty	0.5 (Flaig et al., 2020)	3.50–7.79	7.25–8.32	5.20–8.99	6.15–11.93	√
3	2-Methyl butanal	Malty	1.5 (Flaig et al., 2020)	1.07–2.67	1.77–2.68	1.35–3.24	1.65–3.75	√
4	Hexanal	Green, grassy	2.4 (Flaig et al., 2020)	2.12–5.87	4.64–6.83	3.04–6.06	3.30–6.30	√
5	1-Hexanol	Green, grassy	5.6 (Van Gemert, 2011)	0.06–0.13	0.07–0.18	0.09–0.13	0.06–0.11
6	Heptanal	Green, grassy	2.8 (Van Gemert, 2011)	0.10–0.20	0.15–0.21	0.15–0.24	0.14–0.20
7	Benzaldehyde	Fruity	3 (Wang et al., 2020)	1.70–3.77	1.32–4.30	2.94–6.37	1.44–5.46	√
8	1-Octen-3-ol	Green	1 (Wang et al., 2020)	0.17–0.30	0.19–0.44	0.21–0.39	0.19–0.33
9	Octanal	Green	0.7 (Wang et al., 2020)	0.28–0.47	0.32–0.53	0.38–0.56	0.29–0.52
10	Benzyl alcohol	Flowery	100 (Zhu, 2015)	2.59–5.90	3.88–7.84	5.00–6.78	3.62–8.74	√
11	Phenylacetaldehyde	Sweet (honey-like)	4 (Wang et al., 2020)	10.33–22.85	8.25–31.64	11.18–41.08	1.96–27.76	√
12	β-Ocimene	Green	34 (Van Gemert, 2011)	0.06–0.36	0.00–0.16	0.19–0.45	0.12–0.27
13	(Z)-Linalool oxide (furanoid)	Flowery	320 (Wang et al., 2020)	0.10–0.22	0.09–0.25	0.25–0.32	0.14–0.24
14	(E)-Linalool oxide (furanoid)	Fruity	320 (Wang et al., 2020)	0.47–0.77	0.26–0.76	0.63–1.06	0.48–0.60
15	Linalool	Flowery	6 (Wang et al., 2020)	10.55–14.53	6.77–12.50	10.94–14.41	6.36–12.94	√
16	Nonanal	Flowery	1 (Wang et al., 2020)	0.08–0.50	0.09–0.53	0.21–0.91	0.27–0.78
17	Phenylethyl alcohol	Flowery (honey-like)	45 (Wang et al., 2020)	4.53–6.33	4.15–7.70	5.10–8.07	4.83–7.31	√
18	Methyl salicylate	Wintergreen like	40 (Wang et al., 2020)	0.68–1.00	0.48–0.85	0.73–1.02	0.58–0.80
19	Decanal	Fruity (citrus-like)	0.1 (Wang et al., 2020)	3.54–6.07	3.72–7.22	4.30–9.83	6.95–11.22	√
20	β-Cyclocitral	–	5 (Wang et al., 2020)	0.05–0.11	0.06–0.13	0.08–0.13	0.06–0.13
21	Neral	Flowery	53 (Van Gemert, 2011)	0.07–0.16	0.11–0.16	0.11–0.19	0.09–0.16
22	Geraniol	Flowery (rose-like)	6.6 (Van Gemert, 2011)	21.00–34.14	21.83–35.43	33.27–39.40	26.33–36.52	√
23	(E)-Citral	Fruity (citrus-like)	32 (Van Gemert, 2011)	0.10–0.24	0.16–0.24	0.17–0.27	0.14–0.22
24	(E, E)-2,4-Decadienal	Green (herbal)	0.077 (Van Gemert, 2011)	0.00–2.61	1.45–1.73	0.00–1.88	0.90–1.98	√
25	α-Ionone	Flowery	0.4 (Wang et al., 2020)	0.00–0.32	0.00–0.34	0.25–0.30	0.25–0.30
26	β-Ionone	Flowery (violet-like)	0.021 (Flaig et al., 2020)	0.00–30.00	3.53–33.47	25.44–55.46	12.73–47.37	√
27	Cedrol	Woody	0.5 (Wang et al., 2020)	0.97–1.82	0.96–3.20	0.88–1.90	1.10–1.51	√

^aOdor description from previous studies.

^bVolatile threshold in water from previous studies.

^cOAVs calculation of volatile compounds (the ratio of concentration in tea infusions (W/V, 1:50) to threshold in water).

According to our previous studies (Zhou et al., 2019), odor shape based on volatile compounds with the highest OAVs improves our understanding of whether a volatile is important for overall odor. Therefore, we used the top twenty with the highest OAVs to determine odor shape. Odor shape of each tea sample was differentiated from others (Fig. 3). CF black teas had relatively higher volatile OAVs than other black teas. Overall Keemun black tea odor was controlled mainly by geraniol, β-ionone, phenylacetaldehyde, linalool, and phenylethyl alcohol which had sweet or flowery note, as well as 3-methyl butanal and hexanal with a green odor. The proportion of volatiles with a flowery, fruity, or honey-like odor (these characteristics were recognized as the characteristic fragrance of Keemun black tea) to volatiles with other odors (e.g., green, woody, corn-like) was calculated and differentiated among the four black teas (MF = 5.16:1, CF = 6.58:1, JZ = 4.04:1, XL = 5.11:1). According to sensory evaluation, CF black tea had a higher of flowery and sweet odor intensity than the others, most likely attributed to the high proportion of flowery or sweet volatile compounds.

Fig. 3.

Aroma shape of four Keemun black teas using the volatile OAVs.

Geraniol had little impact on shape, whereas β-ionone dramatically changed among the four Keemun black teas and significantly influenced aroma shape. In the tea plant, carotenoid cleavage dioxygenases CsCCD1 and CsCCD4 can catalyze β-carotene to produce β-ionone. β-Carotene and β-ionone concentration was dramatically influenced by tea cultivars (Wang et al., 2022). CF black tea had the highest OAVs of β-ionone, followed by XL black tea, while the OAVs of β-ionone in JZ and MF were not significantly different. Previous research showed that β-ionone content in black tea samples decreased after a prolonged and thermal drought (Yang et al., 2020). The JZ and XL shaping stage endured a long heating pan which probably led to β-ionone emission and transformation. The phenylacetaldehyde OAV was the lowest in MF but the highest in CF black tea, while it did not differ in JZ and XL black teas. Phenylacetaldehyde was oxidized from phenethyl alcohol during tea processing in our study (Fig. S3). As described above, CF black tea after the refining stage could be recognized as premium materials of MF black tea, while the prolonged shaping stage of JZ and XL possibly enhanced the formation of phenylacetaldehyde and benzaldehyde.

Volatile compounds with high OAVs control Keemun black tea overall aroma, especially geraniol, β-ionone, phenylacetaldehyde, and linalool, and they determine the proportion of flowery and honey-like odors to green and grassy odors. It was well-known that overall tea aroma results from the combined contribution of multiple volatile compounds (Wang et al., 2020, Flaig et al., 2020). Therefore, differentiated processing changed the proportion of volatile compounds and influenced the Keemun black tea aroma profile.

4. Conclusion

The volatile profiles of four Keemun black tea types were comprehensively analyzed by HS-SPME-GC–MS and SBSE-GC-O-MS. A total of 79 volatiles compounds in Keemun tea infusions were identified by two methods. Congou Keemun black tea (with a refining stage) had the highest total volatile concentrations. GBVs and their oxidative products such as phenylacetaldehyde and benzaldehyde, were highly concentrated in Keemun black tea infusions. The GC-O-MS results shows that volatile compounds with sweet, honey-like, flowery, or fruity characters (considered as ‘Keemun aroma’) had a high aroma intensity in Keemun black tea, consistent with the SPME-GC–MS and OAVs results. Geraniol, β-ionone, phenylacetaldehyde, linalool, and phenylethyl alcohol are the key compounds influencing Keemun black tea overall aroma. The OAVs ratio shows that proportion of volatiles (namely, ‘Keemun notes’ /other notes) was different among four Keemun black teas. We showed that differentiated processing with consistent raw materials changes the proportion of volatiles and gives a distinctive aroma to black tea products.

CRediT authorship contribution statement

Yujie Xu: Data curation, Formal analysis, Investigation, Software, Visualization, Writing – original draft. Yaqin Liu: Data curation, Formal analysis, Software, Validation. Jihong Yang: Formal analysis, Validation. Hui Wang: Resources, Validation. Hanchen Zhou: Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Software, Validation, Visualization, Writing – review & editing. Pandeng Lei: Conceptualization, Funding acquisition, Methodology, Validation, Supervision, Writing – review & editing.

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.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.