Effect of hydrogen carbonate in brewing water on the aroma of tea infusions

Abstract

Tea aroma is significantly influenced by hydrogen carbonate (HCO₃⁻) in brewing water. This study investigated the impact of HCO₃⁻ in water on the aroma of brewed tea infusions using sensory evaluation, headspace solid-phase microextraction, and gas chromatography–mass spectrometry techniques. As the concentration of HCO₃⁻ in water increased, the purity of tea aroma diminished, while attributes related to cooked flavors and sweetness intensified; this effect is particularly pronounced at high temperatures and prolonged brewing times. HCO₃⁻ changed the concentration of volatiles, resulting in a significant decrease in concentrations of dimethyl sulfide, β-ionone, and other compounds. Furthermore, the presence of HCO₃⁻ markedly decreased EGCG content while increasing GCG content in tea infusion; these variations in catechin concentrations were correlated with changes in dimethyl sulfide and β-ionone concentrations. These findings enhance the understanding of flavor chemistry concerning tea and water, provide valuable insights for the scientific selection of tea brewing water.

Graphical abstract

Highlights

• •Molecular mechanism of HCO₃⁻ affecting the aroma of tea infusion was first explored.
• •HCO₃⁻ in brewing water reduced the purity and increased the sweetness of tea aroma.
• •HCO₃⁻ in water strongly reduced EGCG content and increased GCG content in tea soup.
• •Catechin contents were correlated with the changes of dimethyl sulfide and β-ionone.

1. Introduction

The unique flavor and quality of tea can only be presented after being infused with water. Various types of water commonly used for daily drinking, including tap water, purified drinking water, natural drinking water, natural spring water, and natural mineral water. The sensory qualities of tea brewed with different types of water can vary significantly (Ma et al., 2023). The quality of tea infusion prepared with tap water typically exhibits diminished aroma, reduced purity, and insufficient mellowness; conversely, tea infusion brewed with purified water is better poised to manifest the inherent quality traits of the tea leaves. Natural water sources and spring water often yield favorable flavor enhancements that elevate both the intensity of aroma and mellow taste in tea infusion. On the other hand, tea brewed with natural mineral water tends to display a less distinct aroma profile and weaker tea taste.

To unravel the intricate mechanisms underlying how water quality influences the taste and quality of tea infusion, researchers have delved deep into exploring the effects of water's physicochemical characteristics. Their findings revealed that pH level, mineralization, and hardness emerge as pivotal factors contributing to the varying qualities of tea infusion (Liu et al., 2020; Xu et al., 2017). These properties, in turn, are intimately tied to the composition and concentration of ions present in water. Hence, the fundamental difference in the flavor profile of tea infused with various types of drinking water ultimately stems from the disparities in their ionic compositions (Zhang et al., 2017). Some studies focusing on the ions in water have proved that the content and composition of ions in water will affect the quality of tea infusion. The total ion content exhibits a negative correlation with the quality of tea infusion (Cao et al., 2021). When water contains higher concentrations of cations or anions, it generally leads to a decrease in the purity of aroma, an increase in a dull, overcooked flavor, a notable rise in astringency, and a decline in freshness and mellowness of the brewed tea. The specific concentration thresholds that produce these effects vary depending on the type of ion involved (Zhang & Yin, 2023). Current research on the impact of ions in water on the quality of tea infusion has primarily focused on cations. It was found that cations in water could promote the self-association of major flavor components in tea infusion, such as caffeine, polyphenols, and theaflavins, resulting in the formation of tea cream (Yin et al., 2009). Cations in water were also found to accelerate the oxidative degradation of catechins during tea brewing, which further affected the flavor quality of the tea infusion (Yin et al., 2014). Furthermore, they were found to influence the aroma quality by altering the content and proportion of aroma components (Bai et al., 2023; Ren et al., 2023). However, there is a lack of research on anions, which also significantly influence the flavor and quality of tea infusion.

The hydrogen carbonate (HCO₃⁻) is one of the ubiquitous anions in natural water. It has strong acidity buffering capacity and can significantly affect the flavor quality of tea infusion. When HCO₃⁻ is used to prepare simulated water in combination with different cations (eg. Ca²⁺, Mg²⁺, Na⁺) at the same concentration, the taste of tea brewed with such water is significantly inferior to that brewed with water containing Cl⁻ as the anion (Yin et al., 2018). The flavor of the tea infusion is weakened, and an overcooked or dull sensation may even arise. Previous research has indicated that HCO₃⁻ plays a crucial role in maintaining the tea infusion within a stable weakly alkaline environment. This ultimately leads to the oxidative degradation of phenolic substances, such as catechins, thereby impacting the overall taste quality of the tea (Chen et al., 2012). Numerous studies have demonstrated interaction effects between taste and aroma components. Elevated polyphenol levels may adversely affect the formation of desirable aromatic characteristics, such as fruity and floral notes (Goldner et al., 2011). The hydrogen peroxide (H₂O₂) generated from EGCG oxidation significantly reduces linalool concentration, thereby compromising aroma quality (Wang et al., 2022). Existing studies have demonstrated that aqueous ions can indirectly influence the release of aromatic compounds by modulating taste-related components. Specifically, the presence of Fe²⁺ may catalyze catechin degradation and subsequently react with the resulting H₂O₂ to generate highly reactive free radicals. These radicals can oxidize volatile compounds, ultimately leading to aroma deterioration in green tea infusion (Gao et al., 2021). The observed influence of HCO₃⁻ on polyphenol content may potentially correlate with modifications in aroma quality characteristics, though the underlying mechanism remains unexplored. Our previous research has clearly established a negative correlation between the content of HCO₃⁻ in water and the aroma quality of tea infusion, significantly diminishing the fresh aroma of green tea, floral aroma of oolong tea, and sweet aroma of black tea (Li et al., 2023; Peng et al., 2022). However, the molecular mechanism of how HCO₃⁻ affects the aroma of tea infusion remains unclear. The present study aims to investigate the effect of HCO₃⁻ on the aroma quality of green tea through a combined approach of sensory evaluation, aroma component analysis, and physicochemical analysis. The research findings will provide a basis for selecting ion indicators in water used for brewing tea, and also lay a theoretical foundation for further in-depth studies on the mechanism of how brewing water affects the release of aroma in tea infusion.

2. Materials and methods

2.1. Materials

Tea samples. Longjing tea was provided by Hangzhou Longguan Co., Ltd. (Hangzhou, China). Instant green tea, obtained by water extraction, concentration and freeze-drying using Longjing tea, was provided by SAN ‘an Biotechnology Co., Ltd. (Anhui, China).

Water Samples. The ions of each water quality collected before the experiment were analyzed by the necessary tests, and 11 kinds of commercially available water with large differences in HCO₃⁻ content were selected as the experimental water (The specific content information was shown in Table S1). Deionised water (Hangzhou Wahaha Group Co., Ltd.) was used as the experimental control. Different concentrations of anionic water were configured by dissolving corresponding masses of sodium chloride (NaCl) and sodium bicarbonate (NaHCO₃) with pure water. That is, 50 mg, 100 mg, 200 mg, 300 mg of NaCl or NaHCO₃ reagents were weighed and dissolved with 1 L pure water. The concentration range of HCO₃⁻ in the collected commercial water was 0–356 mg/L. Accordingly, the concentration of NaHCO₃-simulated water was set at several appropriate concentration points within the range of 0–300 mg/L. NaCl-simulated water of the same concentration was prepared for comparison.

2.2. Chemicals

NaCl, NaHCO₃ (Superior pure, purit ≥99.80 %) and p‑xylene‑d₁₀ (purity = 98 %) was purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). Acetic acid, acetonitrile, methanol (chromatographic purity, purity ≥99.99 %) was purchased from Hangzhou Qichuang Scientific Equipment Co., Ltd. (Hangzhou, China). Decanoic acid ethyl ester (analytically acceptability reagent, purity ≥99.50 %) was purchased from TCI Development Co., Ltd. (Shanghai, China). A standard mixture of n-alkanes C8–C30 was purchased from Shanghai Anpu Resplendent Standard Technology Co., Ltd. (Shanghai, China). Dimethyl sulfide, hexanal were purchased from SIGMA-ALDRICH (Beijing, China). 3-Methyl-butanal, 2-methyl-butanal were purchased from Merye (Shanghai, China). Benzeneacetaldehyde, nonanal, decanal, methyl salicylate, β-cyclocitral, α-terpinene were purchased from Shanghai Yuanye Biotechnology Co., Ltd. (Shanghai, China). Linalool was purchased from TCI (Shanghai, China). Benzaldehyde, indole, 1-octanol, β-pinene, 6- methylhept-5-en-2-one, α-ionone, nerol were purchased from Aladdin (Shanghai, China). Heptanal, β-ionone, (E)-2-octenal were purchased from Macklin (Shanghai, China). Pentanol, styrene, 1-octen-3-ol, 2-ethylfuran, (E)-2-heptenal, 2,4-di-t-butylphenol were purchased from CNW (Shanghai, China). 98 % HPLC grade EGCG (epigallocatechin gallate) and GCG (gallocatechin gallate) were purchased from Shanghai Yuanye Biotechnology Co., Ltd. (Shanghai, China).

2.3. Sensory evaluation

2.3.1. Brewing method

For tea leaves. 3.00 g tea leaves were poured into a 150 mL evaluation cup, filled with boiling water and brewing for 4 min. In this case, commercially available water with different HCO₃⁻ contents and simulated water with different concentrations of NaCl or NaHCO₃ were used for tea brewing.

For instant tea. 0.60 g tea powder were poured into a 240 mL evaluation bowl and filled with 150 mL of boiling water (B, 100 °C) or room temperature water (R, 25 °C), and brewing for 10 s (1) or 3 min (2) and whisked with the tea whisk. In particular, 300 mg/L NaHCO₃ simulated water (N) and pure water (W) were used for brewing water. A total of 8 different brewing treatment were designed: BW1- brewed with boiling deionised water for 10 s; BW2- brewed with boiling deionised water for 3 min; BN1- brewed with boiling NaHCO₃ simulated water for 10 s; BN2- brewed with boiling NaHCO₃ simulated water for 3 min; RW1- brewing with room temperature deionised water for 10 s; RW2- brewing with room temperature deionised water for 3 min; RN1- brewed with room temperature NaHCO₃ simulated water for 10 s; RN2- brewed with room temperature NaHCO₃ simulated water for 3 min.

2.3.2. Evaluation method

An 8-member sensory review team was established, consisting of 4 men and 4 women who obtained the national tea assessor qualification certificate and had been trained in reviewing, and they reviewed experimental samples according to the experimental requirement without understanding the conditions of the samples and the meaning of the sample password (GB T 23776–2018). The aroma of tea infusion were described the quality first and scored with the percentage system based on intensity, persistence, and pleasantness, in order to compare the differences in aroma with the overall perception. The aroma of tea brewed with different concentrations NaCl and NaHCO₃ were evaluated from the purity of aroma, fresh flavor (trans-3-hexenol, new leaves-like aroma), roasted bean (soybeans stir-fried in an iron pan), chestnut (boiled chestnuts-like aroma), sweet (roasted sweet potatoes-like aroma), ripeness (smells like dried straw), and scored according to the 10-point scale method (0 means no such aroma, 10 means the strongest such aroma), in order to carry out multi-dimensional comparisons and discover the nuances of the aroma (Li et al., 2023).

The aroma evaluation results were calculated as the mean value of remaining data after removing the highest and lowest scores. Correlation analysis between the overall aroma scores and HCO₃⁻ ion concentrations was performed using analytical software.

2.4. Physicochemical composition determination

The contents of HCO₃⁻ was tested according to Chinese National Standard (GB 8538–2016). The pH values of water samples and tea infusions were measured using a pH meter (FiveGo F2 pH Meter; Mettler Toledo Instruments, Shanghai, China). The content of catechins in tea infusions was determined by high performance liquid chromatography with UV-detector (Shimadzu, Tokyo, Japan). The samples were filtered by a 0.45 μm aqueous membrane before injection, and separated as follows: Diamonsil™ C18 column (4.6 mm × 250 mm, 5 μm; Dikma Technologies Inc., Lake Forest, CA); column temperature 35 °C; injection volume 10 μL; flow rate 1.0 mL/min; detection wavelength 280 nm. The mobile phases were A: 2 % acetic acid; B: 100 % acetonitrile. The elution solvent was initially 6.50 % B, then ramped linearly to 15 % B at 16 min, held at 15 % B until 25 min, then ramped back to 6.50 % B at 30 min, then continue for 5 min.

2.5. Headspace solid-phase microextraction (SPME)

Headspace solid-phase microextraction (HS-SPME) was based on the previously published method (Li et al., 2023). The tea infusion obtained from 2.3.1 was filtered through a nylon netting and cooled down quickly to room temperature in an ice bath. Then 50 mL of tea infusion was mixed with internal standards (10 μL, 2 mg/L p‑xylene‑d₁₀ and 10 μL, 2 mg/L decanoic acid ethyl ester) and sealed in a 100 mL glass bottle, placed in a 30 °C water bath, mixed and balanced for 5 min. An SPME fiber [50/30 μm Divinylbenzene/ Carboxen/Polydimethylsiloxane (DVB / CAR / PDMS), Stable flex (2 cm)] (Supelco, Inc., Bellefonte, PA, USA) was used for the headspace sampling of the tea infusion sample for 30 min.

2.6. Gas chromatography-mass spectrometry (GC–MS)

The SPME fiber was then desorbed in the injection port of a gas chromatography–mass spectrometry (GC–MS) at 250 °C for 5 min. GC–MS analysis was conducted with an Agilent 6890 GC equipped with 5975B mass selective detector using fused silica capillary columns DB-5MS, 30 m × 0.25 mm, 0.25 μm film thickness. Gas chromatography conditions were as follows: inlet temperature 250 °C; high purity helium (99.999 %) as carrier gas, flow rate of 1.0 mL/min; splitless injection. Temperature program: initial temperature 40 °C for 2 min; increased to 120 °C at a rate of 4 °C/min; rising to 260 °C at a rate of 30 °C/min; hold for 5 min. Mass spectra was recorded in electron impact (EI) ionisation mode at 70 eV. The quadrupole mass detector, ion source and transfer line temperatures were set, respectively, at 150, 230 and 280 °C. The MS scan range was set at 30–350 amu.

2.7. Aroma analysis

The data obtained by GC–MS were processed with the following software: the ion fragments of the detected compounds were analyzed by qualitative analysis software (version B.07.00), automated mass spectral deconvolution and identification system (AMDIS version 2.72 National Institute of Standards and Technology, Gaithersburg, MD), and matched with the ion fragments of NIST (version 2.2. National Institute of Standards and Technology) mass spectrometry search program to match the volatile compounds. The RI of this peak-time volatile was calculated using n-alkanes and compared to the matched compounds in the NIST library. Finally, the mass concentration of the selected compounds was calculated according to the internal standard method, and the contribution degree of the aroma compounds was calculated according to the relative odor activity value (ROAV), the ratio of the relative concentration of the compounds in water to its threshold, was calculated by the formula (Xiao et al., 2022):

$${\text{ROAV}}_{\mathrm{i}}=\frac{C_{\mathit{i}}}{O{T}_{\mathit{i}}}$$

where $C_{\mathit{i}}$is the relative content of the compound, and OT_i is the odor threshold of the compound. A high ROAV signifies that a particular component makes a substantial contribution to the overall flavor profile of the sample.

2.8. Data processing

Each independent experiment was performed in triplicate, and the results were expressed as mean ± standard deviation (SD) were calculated by Microsoft Excel 2019 software. The statistical significant analysis were evaluated based on analysis of variance (ANOVA) using IBM SPSS Statistics 27. The graphs were plotted with Origin 2021 software and GraphPad Prism 9 software. Partial least square discriminate analysis (PLS-DA) were conducted with SIMCA 13.0 software. Relevance heatmaps were analyzed by the mapping tool website https://www.chiplot.online/.

3. Results analysis

3.1. Differences in aroma of tea infusion brewed with natural water containing different HCO₃⁻ levels

To investigate the influence of HCO₃⁻ concentration in water on the aroma quality of brewed tea infusion, various commercial drinking waters with different HCO₃⁻ concentration distributions were selected for tea brewing. Sensory evaluations were performed on the aroma quality of the tea infusions. Measurements revealed that the HCO₃⁻ concentration in the 11 selected commercial drinking waters ranged from 0 to 356 mg/L (Fig. 1). The overall aroma scores of the tea infusions ranged from 87.34 to 93.00, exhibiting distinct aroma profiles including tender chestnut, fresh aroma with tender chestnut notes, pure fresh aroma, chestnut, and cooked chestnut. As the HCO₃⁻ concentration increased, the aroma profile of the brewed tea infusion became more overcooked and sweet, with a corresponding decline in the overall aroma score. A correlation analysis was performed examining the relationship between HCO₃⁻ concentrations and the overall aroma scores of tea infusions. The analysis revealed a negative correlation between HCO₃⁻ concentrations and the aroma quality of tea infusions ($r^2$ = − 0.92). It has been confirmed that HCO₃⁻ can have complex effects on aroma substances by adjusting pH value, changing solubility, participating in chemical reactions, affecting volatility and sensory experience, which may be the cause of aroma changes of tea infusion (Zhang et al., 2017).

Fig. 1.

Influence of natural water with different HCO₃⁻ content on aroma quality of brewed tea. “**” indicates the level of significance, P < 0.01.

3.2. Analysis of aroma and volatiles of tea infusions brewed with simulated water containing different anions

3.2.1. Sensory evaluation

Given the complexity of natural water composition, in order to eliminate interference from other factors present in natural water, simulated water was designed with sodium ions (Na⁺) as the cation, paired with different anions (Cl⁻, HCO₃⁻). Tea leaves were brewed using this simulated water, and the aroma profile of the tea infusion was quantitatively described and analyzed (Fig. 2). The results showed that the aroma profile of the tea infusion brewed with the simulated water formulated with NaCl was very close to that of the control deionised water (WA), with score differences in attributes such as purity, fresh, roasted bean, chestnut, sweet being less than 1. Sodium chloride is often used as a flavor enhancer in food and beverage, and as an electrolyte in volatile substance detection to improve the response of aroma substances (Zhang et al., 2017). The results verified that NaCl could enhance the perception of tea aroma without changing the aroma profile. Brewing tea with simulated water formulated with NaHCO₃ changed the aroma profile of the tea infusion compared to brewing with WA. Furthermore, this change clearly demonstrated a concentration effect. As the NaHCO₃ concentration increased (0 to 300 mg/L), the scores for attributes such as purity (7.80 to 4.80), freshness (4.56 to 2.00), roasted bean (3.60 to 2.10), and chestnut (5.63 to 3.25) in the brewed tea infusion gradually decreased. Conversely, the scores for sweet (3.42 to 4.50) and ripeness aroma (0 to 3.00) showed an upward trend. It can be confirmed that HCO₃⁻ can significantly change the aroma profile of tea infusion by comparing the results of tea brewing using two kinds of simulated water. The simulated water containing sodium chloride did not alter the aroma profile of tea soup, indicating that neither Na⁺ nor Cl⁻ exert a significant influence on the aroma characteristics. Consequently, the observed impact of NaHCO₃ simulated water on the aroma profile of tea soup can be primarily attributed to the presence of HCO₃⁻. This finding is consistent with the results of previous studies (Li et al., 2022).

Fig. 2.

Sensory evaluation of the aroma of tea infusion brewed with different concentrations of simulated water.

3.2.2. Volatile compounds variation analysis

To characterize the aroma differences in tea infusions brewed with different simulated waters, the volatile components were analyzed using the HS-SPME coupled with GC–MS method, resulting in the detection of 86 volatile components in each sample (Table 1). The volatile concentration of tea infusions, which had been brewed with deionised water, ranged from 13.11 to 13.44 μg/L. Meanwhile, the volatile concentration of tea infusions brewed with the simulated water of NaHCO₃ (0 to 300 mg/L), ranged from 10.52 to 11.75 μg/L. As for the tea infusions brewed with the simulated water of NaCl (0 to 300 mg/L), the volatile concentration was found to fall within the range of 11.68 to 12.68 μg/L.

Table 1.

The aroma compounds and its ROAV of tea infusion brewed with different simulated water.

No.	Name	RI Calculation	threshold	WA		NaCI-1		NaCI-2		NaCI-3		NaCI-4		NaHCO3–1		NaHCO3–2		NaHCO3–3		NaHCO3–4
				Contents	ROAV	Contents	ROAV	Contents	ROAV	Contents	ROAV	Contents	ROAV	Contents	ROAV	Contents	ROAV	Contents	ROAV	Contents	ROAV
1	Dimethyl sulfide	633	0.12	1.181 ± 0.009	9.836	1.107 ± 0.089	9.222	1.046 ± 0.01	8.711	1.216 ± 0.051	10.133	1.242 ± 0.055	10.342	0.957 ± 0.036	7.974	0.955 ± 0.01	7.952	0.903 ± 0.016	7.521	0.853 ± 0.065	7.106
2	3-Methyl-butanal	672	0.5	0.439 ± 0.017	0.877	0.435 ± 0.028	0.870	0.381 ± 0.009	0.760	0.446 ± 0.017	0.891	0.496 ± 0.111	0.992	0.44 ± 0.022	0.879	0.462 ± 0.104	0.923	0.491 ± 0.025	0.981	0.488 ± 0.019	0.976
3	2-Methyl-butanal	677	1	0.512 ± 0.022	0.512	0.451 ± 0.039	0.450	0.4 ± 0.007	0.399	0.578 ± 0.089	0.577	0.55 ± 0.045	0.549	0.538 ± 0.037	0.537	0.649 ± 0.034	0.649	0.585 ± 0.036	0.585	0.522 ± 0.052	0.521
4	1-Penten-3-ol	690	23	0.106 ± 0.014	0.005	0.05 ± 0.054	0.002	0.116 ± 0.002	0.005	0.074 ± 0.014	0.003	0.066 ± 0.004	0.003	0.092 ± 0.012	0.004	0.081 ± 0.015	0.003	0.054 ± 0.005	0.002	0.081 ± 0.008	0.004
5	Pentanal	696	12	0.157 ± 0.041	0.013	0.084 ± 0.004	0.007	0.098 ± 0.001	0.008	0.115 ± 0.005	0.010	0.134 ± 0.007	0.011	0.073 ± 0.056	0.006	0.09 ± 0.008	0.007	0.079 ± 0.007	0.007	0.06 ± 0.003	0.005
6	2-Ethyl-furan	699	8000	0.53 ± 0.002	0.000	0.495 ± 0.01	0.000	0.466 ± 0.014	0.000	0.518 ± 0.01	0.000	0.504 ± 0.009	0.000	0.527 ± 0.008	0.000	0.542 ± 0.025	0.000	0.499 ± 0.023	0.000	0.456 ± 0.022	0.000
7	2,5-Dimethyl-furan	709	/	0.055 ± 0.021	/	0.026 ± 0.016	/	0.032 ± 0.001	/	0.047 ± 0.007	/	0.034 ± 0.007	/	0.032 ± 0.001	/	0.05 ± 0.014	/	0.025 ± 0.005	/	0.028 ± 0.002	/
8	(E)-2-Pentenal	742	310	0.033 ± 0.013	0.000	0.035 ± 0.017	0.000	0.026 ± 0.004	0.000	0.023 ± 0.003	0.000	0.03 ± 0.001	0.000	0.024 ± 0.004	0.000	0.033 ± 0.009	0.000	0.023 ± 0.002	0.000	0.026 ± 0.004	0.000
9	2-Methyl-pentanal	754	0.7	0.004 ± 0.002	0.005	0.005 ± 0.004	0.006	0.004 ± 0.004	0.004	0.007 ± 0.003	0.010	0.008 ± 0.002	0.011	0.009 ± 0.001	0.013	0.007 ± 0.003	0.009	0.001 ± 0.001	0.001	0.007 ± 0.001	0.009
10	1-Pentanol	759	150.2	0.036 ± 0.001	0.000	0.03 ± 0.004	0.000	0.035 ± 0.006	0.000	0.038 ± 0.002	0.000	0.039 ± 0.005	0.000	0.036 ± 0.007	0.000	0.04 ± 0.003	0.000	0.032 ± 0.002	0.000	0.03 ± 0.003	0.000
11	Toluene	758	527	0.241 ± 0.019	0.000	0.215 ± 0.001	0.000	0.257 ± 0.033	0.000	0.299 ± 0.009	0.001	0.287 ± 0.023	0.001	0.34 ± 0.014	0.001	0.375 ± 0.027	0.001	0.381 ± 0.038	0.001	0.335 ± 0.021	0.001
12	4-Methyl-3-penten-2-one	794	300	0.199 ± 0.014	0.001	0.187 ± 0.013	0.001	0.192 ± 0.03	0.001	0.218 ± 0.016	0.001	0.223 ± 0.017	0.001	0.227 ± 0.007	0.001	0.206 ± 0.014	0.001	0.211 ± 0.047	0.001	0.188 ± 0.028	0.001
13	Hexanal	798	1.1	1.421 ± 0.097	1.291	1.304 ± 0.038	1.185	1.275 ± 0.152	1.158	1.401 ± 0.009	1.273	1.453 ± 0.117	1.320	1.461 ± 0.064	1.328	1.523 ± 0.083	1.384	1.467 ± 0.065	1.333	1.334 ± 0.09	1.213
14	1-Ethyl-1H-pyrrole	806	/	0.17 ± 0.017	/	0.171 ± 0.005	/	0.173 ± 0.021	/	0.246 ± 0.007	/	0.201 ± 0.021	/	0.171 ± 0.024	/	0.194 ± 0.032	/	0.235 ± 0.024	/	0.215 ± 0.014	/
15	2-Methyl-2-pentenal	825	290	0.102 ± 0.006	0.000	0.106 ± 0.016	0.000	0.117 ± 0.023	0.000	0.11 ± 0.003	0.000	0.112 ± 0.013	0.000	0.104 ± 0.015	0.000	0.103 ± 0.021	0.000	0.115 ± 0.008	0.000	0.094 ± 0.006	0.000
16	3-Methyl-2-hexanone	832	4.1	0.022 ± 0.005	0.005	0.024 ± 0.001	0.006	0.029 ± 0.002	0.007	0.034 ± 0.001	0.008	0.028 ± 0.004	0.007	0.031 ± 0.003	0.007	0.035 ± 0.006	0.008	0.036 ± 0.003	0.009	0.027 ± 0.005	0.006
17	1,2,5,5-Tetramethyl-1,3-cyclopentadiene	837	/	0.017 ± 0.003	/	0.012 ± 0.001	/	0.014 ± 0.001	/	0.015 ± 0.001	/	0.015 ± 0.001	/	0.014 ± 0.002	/	0.015 ± 0.003	/	0.011 ± 0.002	/	0.012 ± 0.002	/
18	(E)-2-Hexenal	848	17	0.032 ± 0.004	0.002	0.02 ± 0.021	0.001	0.028 ± 0.002	0.002	0.023 ± 0.008	0.001	0.021 ± 0.003	0.001	0.026 ± 0.001	0.001	0.026 ± 0.002	0.001	0.026 ± 0.003	0.002	0.024 ± 0.002	0.001
19	(E)-3-Hexen-1-ol	851	110	0.075 ± 0.002	0.001	0.069 ± 0.065	0.001	0.097 ± 0.011	0.001	0.058 ± 0.005	0.001	0.063 ± 0.004	0.001	0.077 ± 0.011	0.001	0.075 ± 0.006	0.001	0.065 ± 0.006	0.001	0.058 ± 0.005	0.001
20	Ethylbenzene	854	2205.25	0.063 ± 0.001	0.000	0.092 ± 0.003	0.000	0.084 ± 0.002	0.000	0.063 ± 0.002	0.000	0.062 ± 0.002	0.000	0.109 ± 0.006	0.000	0.106 ± 0.02	0.000	0.108 ± 0.007	0.000	0.088 ± 0.007	0.000
21	p-Xylene	863	1000	0.196 ± 0.015	0.000	0.195 ± 0.006	0.000	0.171 ± 0.003	0.000	0.185 ± 0.013	0.000	0.167 ± 0.009	0.000	0.196 ± 0.008	0.000	0.211 ± 0.012	0.000	0.262 ± 0.01	0.000	0.21 ± 0.003	0.000
22	1-Hexanol	864	5.6	0.027 ± 0.006	0.005	0.039 ± 0.006	0.007	0.057 ± 0.01	0.010	0.041 ± 0.016	0.007	0.039 ± 0.001	0.007	0.065 ± 0.012	0.012	0.093 ± 0.051	0.017	0.038 ± 0.011	0.007	0.055 ± 0.02	0.010
23	2-Heptanone	885	1	0.053 ± 0.006	0.053	0.043 ± 0.003	0.042	0.052 ± 0.004	0.051	0.046 ± 0.004	0.046	0.056 ± 0.005	0.056	0.057 ± 0.009	0.057	0.066 ± 0.012	0.065	0.057 ± 0.01	0.057	0.056 ± 0.022	0.056
24	Styrene	886	65	0.115 ± 0.005	0.002	0.122 ± 0.008	0.002	0.136 ± 0.017	0.002	0.143 ± 0.003	0.002	0.149 ± 0.011	0.002	0.177 ± 0.006	0.003	0.203 ± 0.082	0.003	0.19 ± 0.011	0.003	0.162 ± 0.005	0.002
25	(Z)-4-Heptenal	897	0.025	0.009 ± 0.001	0.320	0.005 ± 0.001	0.187	0.007 ± 0.001	0.269	0.006 ± 0.001	0.216	0.007 ± 0.001	0.275	0.007 ± 0.004	0.264	0.005 ± 0.001	0.164	0.004 ± 0.001	0.150	0.004 ± 0.001	0.148
26	Heptanal	899	2.8	0.653 ± 0.027	0.233	0.575 ± 0.046	0.205	0.531 ± 0.036	0.189	0.617 ± 0.012	0.220	0.593 ± 0.054	0.212	0.622 ± 0.071	0.222	0.588 ± 0.023	0.210	0.536 ± 0.025	0.191	0.532 ± 0.019	0.190
27	Hexanoic acid, methyl ester	922	70	0.033 ± 0.003	0.000	0.038 ± 0.009	0.001	0.03 ± 0.003	0.000	0.033 ± 0.002	0.000	0.031 ± 0.002	0.000	0.034 ± 0.003	0.000	0.029 ± 0.001	0.000	0.026 ± 0.004	0.000	0.025 ± 0.002	0.000
28	(E)-2-Heptenal	953	40	0.04 ± 0.005	0.001	0.007 ± 0.001	0.000	0.006 ± 0.002	0.000	0.004 ± 0.001	0.000	0.007 ± 0.003	0.000	0.031 ± 0.006	0.001	0.036 ± 0.001	0.001	0.027 ± 0.008	0.001	0.028 ± 0.002	0.001
29	Benzaldehyde	957	350	0.184 ± 0.002	0.001	0.143 ± 0.007	0.000	0.116 ± 0.006	0.000	0.142 ± 0.002	0.000	0.142 ± 0.01	0.000	0.184 ± 0.027	0.001	0.182 ± 0.014	0.001	0.185 ± 0.014	0.001	0.186 ± 0.014	0.001
30	1,3,5-Trimethylbenzene	964	700	0.061 ± 0.001	0.000	0.012 ± 0.001	0.000	0.005 ± 0.001	0.000	0.06 ± 0.001	0.000	0.006 ± 0.001	0.000	0.017 ± 0.002	0.000	0.012 ± 0.001	0.000	0.029 ± 0.001	0.000	0.021 ± 0.003	0.000
31	1-Heptanol	966	5.4	0.002 ± 0.001	0.000	0.021 ± 0.003	0.004	0.016 ± 0.001	0.003	0.021 ± 0.001	0.004	0.019 ± 0.002	0.003	0.03 ± 0.003	0.006	0.028 ± 0.004	0.005	0.018 ± 0.004	0.003	0.011 ± 0.002	0.002
32	4-Ethyltoluene	973	600	0.018 ± 0.001	0.000	0.004 ± 0.001	0.000	0.015 ± 0.001	0.000	0.021 ± 0.001	0.000	0.022 ± 0.001	0.000	0.01 ± 0.002	0.000	0.014 ± 0.001	0.000	0.032 ± 0.008	0.000	0.024 ± 0.003	0.000
33	1-Octen-3-ol	978	1	0.047 ± 0.003	0.046	0.034 ± 0.005	0.034	0.036 ± 0.009	0.035	0.032 ± 0.003	0.031	0.032 ± 0.006	0.032	0.032 ± 0.007	0.032	0.037 ± 0.002	0.037	0.031 ± 0.003	0.030	0.033 ± 0.004	0.032
34	6-Methyl-5-hepten-2-one	981	50	0.058 ± 0.006	0.001	0.059 ± 0.008	0.001	0.062 ± 0.017	0.001	0.075 ± 0.003	0.001	0.056 ± 0.01	0.001	0.034 ± 0.008	0.001	0.046 ± 0.005	0.001	0.047 ± 0.005	0.001	0.064 ± 0.015	0.001
35	2-Pentylfuran	987	5.8	0.139 ± 0.008	0.024	0.18 ± 0.004	0.031	0.19 ± 0.022	0.033	0.278 ± 0.006	0.048	0.278 ± 0.038	0.048	0.097 ± 0.013	0.017	0.074 ± 0.009	0.013	0.037 ± 0.009	0.006	0.033 ± 0.005	0.006
36	(E,Z)-2,4-Heptadienal	994	94.8	0.004 ± 0.001	0.000	0.003 ± 0.001	0.000	0.003 ± 0.001	0.000	0.003 ± 0.001	0.000	0.011 ± 0.002	0.000	0.004 ± 0.001	0.000	0.005 ± 0.001	0.000	0.005 ± 0.001	0.000	0.005 ± 0.001	0.000
37	Octanal	1001	0.52	0.248 ± 0.023	0.476	0.21 ± 0.026	0.403	0.21 ± 0.028	0.404	0.219 ± 0.003	0.421	0.193 ± 0.002	0.371	0.207 ± 0.027	0.397	0.174 ± 0.013	0.334	0.174 ± 0.007	0.333	0.148 ± 0.011	0.285
38	(E)-3-Hexen-1-ol, acetate,	1003	870	0.268 ± 0.024	0.000	0.216 ± 0.07	0.000	0.181 ± 0.001	0.000	0.175 ± 0.045	0.000	0.178 ± 0.011	0.000	0.194 ± 0.014	0.000	0.189 ± 0.01	0.000	0.127 ± 0.09	0.000	0.199 ± 0.035	0.000
39	(E,E)-2,4-Heptadienal	1009	15.4	0.067 ± 0.001	0.004	0.042 ± 0.018	0.003	0.026 ± 0.001	0.002	0.025 ± 0.004	0.002	0.026 ± 0.003	0.002	0.052 ± 0.004	0.003	0.026 ± 0.005	0.002	0.024 ± 0.018	0.002	0.037 ± 0.001	0.002
40	α-Terpinene	1013	85	0.016 ± 0.001	0.000	0.01 ± 0.004	0.000	0.009 ± 0.004	0.000	0.007 ± 0.001	0.000	0.008 ± 0.001	0.000	0.004 ± 0.001	0.000	0.005 ± 0.001	0.000	0.008 ± 0.006	0.000	0.006 ± 0.004	0.000
41	1,2,3-Trimethyl-benzene	1016	260	0.018 ± 0.001	0.000	0.02 ± 0.003	0.000	0.012 ± 0.001	0.000	0.015 ± 0.003	0.000	0.015 ± 0.002	0.000	0.068 ± 0.009	0.000	0.073 ± 0.009	0.000	0.043 ± 0.008	0.000	0.023 ± 0.008	0.000
42	p-Cymene	1021	5.01	0.038 ± 0.003	0.007	0.039 ± 0.007	0.008	0.036 ± 0.003	0.007	0.038 ± 0.008	0.008	0.031 ± 0.002	0.006	0.035 ± 0.004	0.007	0.047 ± 0.008	0.009	0.033 ± 0.002	0.007	0.031 ± 0.002	0.006
43	d-Limonene	1026	34	0.198 ± 0.063	0.006	0.277 ± 0.015	0.008	0.224 ± 0.029	0.007	0.276 ± 0.031	0.008	0.263 ± 0.005	0.008	0.261 ± 0.018	0.008	0.261 ± 0.01	0.008	0.249 ± 0.017	0.007	0.229 ± 0.011	0.007
44	2,2,6-Trimethyl-cyclohexanone	1031	0.1	0.084 ± 0.009	0.840	0.095 ± 0.018	0.946	0.137 ± 0.032	1.366	0.075 ± 0.006	0.747	0.084 ± 0.01	0.833	0.086 ± 0.017	0.852	0.069 ± 0.005	0.684	0.069 ± 0.008	0.686	0.057 ± 0.005	0.568
45	E-3-Octen-2-one	1035	/	0.014 ± 0.003	/	0.018 ± 0.004	/	0.016 ± 0.005	/	0.013 ± 0.001	/	0.015 ± 0.002	/	0.017 ± 0.004	/	0.014 ± 0.003	/	0.013 ± 0.002	/	0.011 ± 0.001	/
46	Benzeneacetaldehyde	1039	4	0.057 ± 0.001	0.014	0.053 ± 0.016	0.013	0.05 ± 0.009	0.012	0.056 ± 0.002	0.014	0.056 ± 0.001	0.014	0.058 ± 0.007	0.014	0.046 ± 0.004	0.011	0.051 ± 0.011	0.013	0.039 ± 0.005	0.010
47	1-Ethyl-2-formyl-1H-pyrrole	1043	65,000	0.114 ± 0.017	0.000	0.09 ± 0.015	0.000	0.082 ± 0.012	0.000	0.094 ± 0.002	0.000	0.083 ± 0.005	0.000	0.103 ± 0.016	0.000	0.09 ± 0.014	0.000	0.098 ± 0.018	0.000	0.088 ± 0.002	0.000
48	trans-β-Ocimene	1044	10	0.033 ± 0.005	0.003	0.013 ± 0.004	0.001	0.014 ± 0.001	0.001	0.018 ± 0.002	0.002	0.014 ± 0.001	0.001	0.016 ± 0.001	0.002	0.014 ± 0.001	0.001	0.012 ± 0.002	0.001	0.01 ± 0.002	0.001
49	(E)-2-Octenal	1055	0.34	0.08 ± 0.002	0.234	0.06 ± 0.006	0.176	0.054 ± 0.007	0.157	0.059 ± 0.007	0.171	0.051 ± 0.001	0.150	0.057 ± 0.011	0.167	0.041 ± 0.004	0.118	0.044 ± 0.005	0.126	0.038 ± 0.007	0.111
50	1-Octanol	1068	0.022	0.089 ± 0.002	4.023	0.061 ± 0.012	2.756	0.069 ± 0.02	3.125	0.082 ± 0.005	3.695	0.077 ± 0.001	3.469	0.089 ± 0.02	4.003	0.085 ± 0.005	3.836	0.075 ± 0.005	3.391	0.068 ± 0.002	3.081
51	(E)-Linalool oxide (furanoid)	1083	190	0.027 ± 0.002	0.000	0.025 ± 0.003	0.000	0.022 ± 0.004	0.000	0.025 ± 0.004	0.000	0.013 ± 0.004	0.000	0.024 ± 0.003	0.000	0.026 ± 0.005	0.000	0.029 ± 0.002	0.000	0.023 ± 0.003	0.000
52	1-Ethyl-4-(1-methylethyl)-benzene	1086	n.f.	0.047 ± 0.002	/	0.024 ± 0.002	/	0.027 ± 0.006	/	0.039 ± 0.001	/	0.035 ± 0.007	/	0.027 ± 0.002	/	0.033 ± 0.009	/	0.04 ± 0.01	/	0.045 ± 0.005	/
53	Linalool	1097	0.22	0.336 ± 0.003	1.524	0.302 ± 0.04	1.368	0.261 ± 0.017	1.185	0.287 ± 0.023	1.302	0.147 ± 0.16	0.666	0.292 ± 0.033	1.325	0.297 ± 0.035	1.347	0.291 ± 0.043	1.322	0.271 ± 0.006	1.230
54	Nonanal	1102	1.1	1.492 ± 0.001	1.356	1.281 ± 0.046	1.164	1.217 ± 0.198	1.106	1.28 ± 0.034	1.163	1.16 ± 0.121	1.054	1.129 ± 0.149	1.026	1.059 ± 0.066	0.962	1.122 ± 0.066	1.020	1.015 ± 0.112	0.922
55	Octanoic acid, methyl ester	1121	200	0.008 ± 0.001	0.000	0.006 ± 0.002	0.000	0.006 ± 0.001	0.000	0.006 ± 0.002	0.000	0.005 ± 0.001	0.000	0.006 ± 0.002	0.000	0.005 ± 0.001	0.000	0.005 ± 0.001	0.000	0.005 ± 0.002	0.000
56	(E)- 2-Nonenal	1145	0.19	0.01 ± 0.001	0.052	0.02 ± 0.002	0.104	0.005 ± 0.001	0.026	0.013 ± 0.001	0.064	0.012 ± 0.001	0.058	0.017 ± 0.003	0.088	0.017 ± 0.001	0.088	0.009 ± 0.003	0.046	0.005 ± 0.001	0.022
57	Naphthalene	1157	6	0.034 ± 0.003	0.006	0.031 ± 0.003	0.005	0.021 ± 0.001	0.003	0.027 ± 0.001	0.004	0.026 ± 0.001	0.004	0.027 ± 0.002	0.004	0.026 ± 0.005	0.004	0.029 ± 0.005	0.005	0.026 ± 0.002	0.004
58	(Z)-3-Hexenyl butyrate	1178	320	0.5 ± 0.001	0.002	0.388 ± 0.037	0.001	0.325 ± 0.048	0.001	0.354 ± 0.022	0.001	0.326 ± 0.019	0.001	0.395 ± 0.062	0.001	0.341 ± 0.076	0.001	0.313 ± 0.026	0.001	0.333 ± 0.086	0.001
59	Methyl salicylate	1183	40	0.081 ± 0.016	0.002	0.069 ± 0.006	0.002	0.061 ± 0.009	0.002	0.057 ± 0.008	0.001	0.053 ± 0.001	0.001	0.047 ± 0.015	0.001	0.047 ± 0.012	0.001	0.04 ± 0.01	0.001	0.038 ± 0.007	0.001
60	2,6,6-Trimethyl-1,3-cyclohexadiene-1-carboxaldehyde	1187	3	0.039 ± 0.001	0.013	0.025 ± 0.001	0.008	0.027 ± 0.001	0.009	0.024 ± 0.001	0.008	0.027 ± 0.002	0.009	0.025 ± 0.007	0.008	0.028 ± 0.014	0.009	0.023 ± 0.012	0.007	0.023 ± 0.014	0.007
61	Dodecane	1194	13,000	0.006 ± 0.001	0.000	0.009 ± 0.002	0.000	0.01 ± 0.003	0.000	0.008 ± 0.001	0.000	0.008 ± 0.001	0.000	0.012 ± 0.003	0.000	0.01 ± 0.001	0.000	0.008 ± 0.002	0.000	0.008 ± 0.001	0.000
62	Decanal	1198	0.19	0.342 ± 0.003	1.797	0.391 ± 0.005	2.055	0.321 ± 0.072	1.686	0.442 ± 0.027	2.321	0.334 ± 0.025	1.757	0.22 ± 0.025	1.158	0.133 ± 0.013	0.695	0.219 ± 0.067	1.149	0.23 ± 0.077	1.209
63	2,6,6-Trimethyl-1-cyclohexene-1-carboxaldehyde	1203	3	0.126 ± 0.007	0.042	0.104 ± 0.003	0.034	0.087 ± 0.01	0.029	0.089 ± 0.004	0.030	0.083 ± 0.006	0.027	0.098 ± 0.018	0.033	0.092 ± 0.008	0.030	0.082 ± 0.011	0.027	0.07 ± 0.001	0.023
64	cis-3-Hexenyl-α-methylbutyrate	1215	/	0.456 ± 0.018	/	0.41 ± 0.004	/	0.347 ± 0.061	/	0.32 ± 0.024	/	0.316 ± 0.014	/	0.351 ± 0.022	/	0.241 ± 0.033	/	0.299 ± 0.022	/	0.284 ± 0.117	/
65	cis-3-Hexenyl isovalerate	1229	20	0.055 ± 0.003	0.003	0.06 ± 0.005	0.003	0.048 ± 0.002	0.002	0.044 ± 0.002	0.002	0.044 ± 0.009	0.002	0.062 ± 0.005	0.003	0.053 ± 0.005	0.003	0.052 ± 0.003	0.003	0.052 ± 0.013	0.003
66	Nerol	1235	7.5	0.074 ± 0.019	0.010	0.079 ± 0.003	0.011	0.061 ± 0.013	0.008	0.084 ± 0.016	0.011	0.06 ± 0.016	0.008	0.071 ± 0.018	0.009	0.064 ± 0.007	0.008	0.062 ± 0.009	0.008	0.06 ± 0.007	0.008
67	2,6,6-Trimethyl-1-cyclohexene-1-acetaldehyde	1249		0.043 ± 0.004	/	0.041 ± 0.001	/	0.035 ± 0.007	/	0.036 ± 0.001	/	0.034 ± 0.001	/	0.039 ± 0.008	/	0.037 ± 0.004	/	0.028 ± 0.01	/	0.031 ± 0.002	/
68	E-Citral (geranial)	1253	32	0.002 ± 0.001	0.000	0.004 ± 0.003	0.000	0.002 ± 0.001	0.000	0.002 ± 0.001	0.000	0.002 ± 0.001	0.000	0.002 ± 0.001	0.000	0.002 ± 0.001	0.000	0.002 ± 0.001	0.000	0.002 ± 0.001	0.000
69	Indole	1266	11	0.007 ± 0.002	0.001	0.009 ± 0.001	0.001	0.007 ± 0.001	0.001	0.008 ± 0.001	0.001	0.007 ± 0.001	0.001	0.011 ± 0.001	0.001	0.008 ± 0.002	0.001	0.008 ± 0.001	0.001	0.006 ± 0.001	0.000
70	Undecanal	1288	14	0.041 ± 0.01	0.003	0.041 ± 0.004	0.003	0.03 ± 0.003	0.002	0.04 ± 0.003	0.003	0.037 ± 0.002	0.003	0.004 ± 0.001	0.000	0.007 ± 0.001	0.000	0.007 ± 0.001	0.000	0.007 ± 0.001	0.000
71	2,6,10,10-Tetramethyl-1-oxaspiro[4.5]dec-6-ene	1306	/	0.008 ± 0.002	/	0.009 ± 0.002	/	0.007 ± 0.001	/	0.006 ± 0.001	/	0.005 ± 0.001	/	0.005 ± 0.002	/	0.006 ± 0.001	/	0.006 ± 0.001	/	0.004 ± 0.001	/
72	Cubebene	1315	/	0.012 ± 0.003	/	0.009 ± 0.01	/	0.01 ± 0.001	/	0.008 ± 0.004	/	0.008 ± 0.003	/	0.007 ± 0.002	/	0.005 ± 0.001	/	0.005 ± 0.001	/	0.005 ± 0.001	/
73	Dehydro-ar-ionene	1349	2.5	0.016 ± 0.002	0.006	0.017 ± 0.003	0.007	0.014 ± 0.002	0.005	0.014 ± 0.001	0.005	0.012 ± 0.001	0.005	0.014 ± 0.003	0.005	0.012 ± 0.001	0.005	0.011 ± 0.002	0.004	0.012 ± 0.002	0.004
74	(Z)-Hex-3-enyl hexanoate	1359	781	0.549 ± 0.057	0.001	0.391 ± 0.032	0.000	0.409 ± 0.039	0.001	0.399 ± 0.051	0.001	0.378 ± 0.074	0.000	0.415 ± 0.027	0.001	0.442 ± 0.011	0.001	0.39 ± 0.023	0.000	0.446 ± 0.08	0.001
75	cis-3-Hexenyl cis-3-hexenoate	1382	/	0.015 ± 0.004	/	0.054 ± 0.005	/	0.05 ± 0.003	/	0.053 ± 0.009	/	0.053 ± 0.009	/	0.054 ± 0.006	/	0.04 ± 0.005	/	0.051 ± 0.006	/	0.046 ± 0.01	/
76	cis-Jasmone	1385	7	0.015 ± 0.001	0.002	0.012 ± 0.004	0.002	0.014 ± 0.001	0.002	0.015 ± 0	0.002	0.016 ± 0.006	0.002	0.015 ± 0.002	0.002	0.015 ± 0.003	0.002	0.011 ± 0.002	0.002	0.009 ± 0.002	0.001
77	Dodecanal	1387	1.07	0.042 ± 0.014	0.039	0.031 ± 0.004	0.028	0.044 ± 0.016	0.041	0.028 ± 0.002	0.026	0.026 ± 0.006	0.024	0.028 ± 0.004	0.026	0.012 ± 0.001	0.011	0.013 ± 0.003	0.012	0.007 ± 0.001	0.006
78	α-Cedrene	1408	/	0.012 ± 0.001	/	0.012 ± 0.001	/	0.02 ± 0.002	/	0.014 ± 0.003	/	0.015 ± 0.004	/	0.017 ± 0.001	/	0.015 ± 0.001	/	0.015 ± 0.004	/	0.009 ± 0.003	/
79	α-Ionone	1419	0.4	0.008 ± 0.002	0.018	0.005 ± 0.001	0.012	0.009 ± 0.002	0.021	0.005 ± 0.001	0.011	0.009 ± 0.002	0.022	0.011 ± 0.001	0.026	0.009 ± 0.001	0.022	0.007 ± 0.002	0.017	0.005 ± 0.002	0.011
80	trans-Geranylacetone	1422	60	0.021 ± 0.002	0.000	0.044 ± 0.001	0.001	0.046 ± 0.028	0.001	0.098 ± 0.002	0.002	0.048 ± 0.01	0.001	0.021 ± 0.005	0.000	0.016 ± 0.003	0.000	0.011 ± 0.002	0.000	0.007 ± 0.004	0.000
81	1-Dodecanol	1449	16	0.216 ± 0.073	0.013	0.301 ± 0.045	0.019	0.35 ± 0	0.022	0.275 ± 0.012	0.017	0.221 ± 0.029	0.014	0.036 ± 0.004	0.002	0.024 ± 0.001	0.001	0.015 ± 0.004	0.001	0.01 ± 0.003	0.001
82	β-ionone	1475	0.007	0.063 ± 0.018	8.982	0.066 ± 0.005	9.399	0.075 ± 0.003	10.574	0.062 ± 0.001	8.733	0.051 ± 0.007	7.153	0.065 ± 0.003	9.213	0.052 ± 0.002	7.393	0.049 ± 0.004	6.873	0.039 ± 0.014	5.537
83	Butylated hydroxytoluene	1473	1000	0.016 ± 0.004	0.000	0.012 ± 0.002	0.000	0.011 ± 0.005	0.000	0.015 ± 0.002	0.000	0.008 ± 0.003	0.000	0.01 ± 0.003	0.000	0.01 ± 0.002	0.000	0.008 ± 0.003	0.000	0.006 ± 0.004	0.000
84	Δ-Cadinene	1501	/	0.066 ± 0.003	/	0.075 ± 0.015	/	0.084 ± 0.016	/	0.089 ± 0.006	/	0.089 ± 0.024	/	0.061 ± 0.016	/	0.036 ± 0.003	/	0.04 ± 0.004	/	0.031 ± 0.013	/
85	cis-Calamenene	1522	/	0.035 ± 0.01	/	0.051 ± 0.018	/	0.041 ± 0.01	/	0.052 ± 0.004	/	0.048 ± 0.013	/	0.041 ± 0.008	/	0.035 ± 0.003	/	0.034 ± 0.004	/	0.029 ± 0.009	/
86	α-Calacorene	1527	/	0.007 ± 0.001	/	0.005 ± 0.002	/	0.006 ± 0.002	/	0.008 ± 0.002	/	0.004 ± 0.003	/	0.002 ± 0.001	/	0.005 ± 0.001	/	0.003 ± 0.002	/	0.003 ± 0.003	/

Note: The unit of the compound content is ug/L. /: the threshold is not found and ROAV cannot be calculated. Odor thresholds were obtained from: Odor & Flavor Detection Thresholds in Water (In Parts per Billion, μg/L); doi:https://doi.org/10.1016/j.foodchem.2018.04.117; doi:https://doi.org/10.1016/j.foodchem.2021.131933; doi:https://doi.org/10.1007/s00217-022-03967-3; Compilations of odor threshold values in air, water and other media (second enlarged and revised edition).

The concentrations of volatile compounds in tea infusions were examined through the application of Partial Least Squares Discriminant Analysis (PLS-DA), to investigate the influence of three brewing waters on the volatile compounds in the tea infusions (Fig. 3A). The model exhibited a coefficient of determination for the independent variable (R2X) of 0.954, a coefficient of determination for the dependent variable (R2Y) of 0.947, and a predictive index (Q2) of 0.928, which indicated a good fit for the model. The tea infusions were categorized into two distinct groups: one group comprised NaCl-simulated water and deionised water, while the NaHCO₃-simulated water was clearly separated from them. This indicated that there was no significant difference in the volatile compounds of tea infusions brewed with NaCl-simulated water compared to deionised water; however, a significant difference was observed in infusions brewed with NaHCO₃-simulated water. Combining the aforementioned results, it can be confirmed that HCO₃⁻ in water plays a significant role in altering the aroma quality and composition of volatile compounds in brewed tea infusion.

Fig. 3.

PLS-DA of tea infusion aroma brewed with simulated water and the heat maps with the components of VIP > 1. A: PLS-DA of tea infusion aroma brewed with deionised water (WA), NaCl simulated water (NaCl) and NaHCO₃ simulated water (NaHCO₃); B: PLS-DA of tea infusion aroma brewed with deionised water (WA) and NaHCO₃ simulated water; C: heat maps with VIP > 1 compound in the two models (WA and NaHCO₃).

To further investigate the impact of HCO₃⁻ on tea infusion aroma, PLS-DA analysis was conducted on the aroma of tea infusions brewed with NaHCO₃-simulated water and deionised water (Fig. 3B and C). The independent variable (R2X, cum), dependent variable (R2Y, cum), and the model prediction index (Q2, cum) of the model were 0.659, 0.996 and 0.968. Variable influence on projection (VIP) values are often used to screen key contributing compounds. A compound with VIP > 1 can be considered a relevant contributor. The VIP values for important compounds in the model were analyzed. The heat map and cluster analysis were carried out on the 38 volatile compounds with a VIP value >1. The results of the cluster analysis revealed that deionised water was initially segregated from NaHCO₃-simulated water, forming a distinct category. Furthermore, the clustering outcomes confirmed that the presence of HCO₃⁻ in water exerted a notable influence on the aroma compounds of tea infusion. The differences in aroma compounds between tea infusions brewed with deionised water and NaHCO₃-simulated water were clearly highlighted by the heatmap. The concentrations of toluene, 3-methyl-2-hexanone, ethylbenzene, 1-heptanol, (E,Z)-2,4-hexadienal, 1,2,3-trimethyl-benzene, dodecane, and cis-3-hexenyl cis-3-hexenoate were higher in the tea infusions brewed with NaHCO₃-simulated water compared to those brewed with deionised water. Conversely, the levels of dimethyl sulfide, pentanal, 1,2,5,5-tetramethyl-1,3-cyclopentadiene, (E)-2-heptenal, 1,3,5-trimethylbenzene, 1-octen-3-ol, 2-pentylfuran, octanal, (E,E)-2,4-heptadienal, α-terpinene, trans-β-ocimene, (E)-2-octenal, nonanal, methyl octanoate, naphthalene, (Z)-3-hexenyl butyrate, methyl salicylate, 2,6,6-trimethyl-1-cyclohexene-1-carboxaldehyde, cis-3-hexenyl-α-methylbutyrate, trans-citral, undecanal, theaspirane, cubebene, dehydro-ar-ionene, cis-3-hexenyl hexanoate, dodecanal, 1-dodecanol, butylated hydroxytoluene, δ-cadinene, and α-calacorene were lower in the simulated water tea infusions than in the WA tea infusions. Notably, the concentrations of dimethyl sulfide, 2-pentylfuran, trans-β-ocimene, and 2,6,6-trimethyl-1-cyclohexene-1-carboxaldehyde exhibited a decreasing trend with varying concentrations of NaHCO₃ in water.

The ROAV was commonly utilized to assess the contributions of individual aroma compounds to the overall aroma profile (Xiao et al., 2022). Components with ROAV ≥1 are generally considered as the key aroma compounds of the analyzed samples, whereas components with 0.1 ≤ ROAV <1 have important modifying effects on the overall aroma of the samples. The ROAVs of aroma compounds in tea infusions brewed with deionised water or NaHCO₃ -simulated water were calculated (Table 1). By comparing the results of VIP and ROAV, five key aroma compounds were identified (Fig. 4). These compounds were dimethyl sulfide (fresh aroma), β-ionone (violet-like fragrance), linalool (sweet floral scent), 1-octanol (fatty and citrus-like aroma), and hexanal (grassy). Notably, both dimethyl sulfide and β-ionone were found to exhibit lower concentrations in tea infusions brewed with NaHCO₃-simulated water, and their concentrations demonstrated a decreasing trend as the HCO₃⁻ concentration increased. Some previous studies also focused on the influence of HCO₃⁻ on sensory aroma quality (Chen et al., 2023), and this result further explained the influence of HCO₃⁻ on aroma substances.

Fig. 4.

Changes of key aroma compounds in tea infusion brewed with different concentrations of NaHCO3 simulated water.

3.3. Comprehensive analysis of instant tea aroma brewed with HCO₃⁻ simulated water

The aforementioned studies have confirmed that the concentration of HCO₃⁻ in water can affect the aroma quality and volatiles of tea infusions. The brewing process of tea includes the leaching and release of aroma substances in tea leaves, as well as the chemical transformation process. Ions present in water may impact the aroma quality of tea infusion through various pathways. To eliminate the influence of leaching rate of aroma substances, instant tea was chosen as the research subject to specifically investigate the impact of HCO₃⁻ on the aroma in tea infusion system and its possible pathways.

3.3.1. Sensory evaluation of instant tea aroma brewed with simulated water

With deionised water (W) as the control, the effect of NaHCO₃ simulated water (N, 300 mg/L NaHCO₃ in deionised water) on the aroma quality of instant tea at different brewing temperatures and brewing times was studied (Fig. 5). Under short-time (10s) brewing conditions at room temperature, both the simulated water and deionised water-brewed instant tea exhibited a chestnut aroma, with their overall aroma scores being close (less than 1 point apart). When the brewing time was extended to 3 min, the instant tea brewed with deionised water maintained its chestnut aroma, while the instant tea brewed with simulated water began to develop a sweet potato aroma, resulting in a 6.50-point decrease in its overall aroma score. Under high-temperature short-time brewing conditions, the tea brewed with deionised water retained its chestnut aroma, whereas the chestnut aroma in the instant tea brewed with simulated water weakened and was accompanied by a hint of sweet potato aroma, leading to a 6-point reduction in its overall aroma score. Upon reaching a brewing time of 3 min, the tea's aroma shifted completely to that of sweet potato, along with an ripeness odor, causing an 8.50-point drop in its overall aroma score. Compared with the aroma of instant tea that had been brewed with deionised water, the aroma of instant tea brewed with simulated water gradually shifted from a chestnut-like fragrance to a sweet potato-like fragrance, resulting in an overall reduction in the aroma score. The results indicated that the aroma of brewed tea infusion had been tended to be made ripeness and sweet by HCO₃⁻ in water, which was found to be consistent with the findings from the sensory experiments where tea leaves were brewed with simulated water. Upon comparing the sensory results obtained under varying brewing times and temperatures, it was observed that the degree to which HCO₃⁻ in water contributed to the transformation of tea infusion aroma towards ripeness and sweetness was accelerated by temperature and demonstrated a cumulative effect over time. Focusing on instant tea powder as the research object, complex reactions such as aroma leaching and release during tea brewing were excluded. Therefore, through sensory experiments, it was determined that HCO₃⁻ could directly affect the aroma profile of tea soup headspace and lead to changes in sensory aroma. Unlike prior investigations, this methodology offers a more controlled and precise elucidation of the specific mechanistic role of HCO₃⁻ in modulating the aromatic properties of tea (Liu, 2014).

Fig. 5.

Sensory evaluation of instant tea under different brewing conditions. RW1- brewed with room temperature (25 °C) deionised water for 10 s; RN1- brewed with room temperature NaHCO₃ simulated water for 10 s; RW2- brewing with room temperature deionised water for 3 min; RN2- brewed with room temperature NaHCO₃ simulated water for 3 min. BW1- brewed with boiling (100 °C) deionised water for 10 s; BN1- brewed with boiling NaHCO₃ simulated water for 10 s; BW2- brewed with boiling deionised water for 3 min; BN2- brewed with boiling NaHCO₃ simulated water for 3 min.

3.3.2. Effects of HCO₃⁻ in water on key aroma components

The effects of HCO₃⁻ on the volatile composition of instant tea infusion at different brewing temperatures and brewing times was studied, and 100 volatile compounds were identified in instant tea infusions brewed with W and N (Table S2). The instant tea infusions brewed with the W had the volatile concentrations of 23.81–27.46 μg/L, whereas the volatile concentrations of the instant tea infusions brewed with the N were 20.81–25.61 μg/L.

For several key aroma compounds (Fig. 6), under high-temperature (100 °C) brewing conditions, the concentration of dimethyl sulfide in the instant tea infusion prepared with N (0.45 μg/L for 3 min, 0.52 μg/L for 10 s) was significantly lower than that in the instant tea infusion prepared with W (0.65 μg/L for 3 min, 0.75 μg/L for 10 s), while the changes in concentration was not significant under room-temperature (25 °C) brewing conditions. This indicated that HCO₃⁻ in water could resulted in a decrease in the concentration of dimethyl sulfide volatilized in the headspace of tea infusions at high temperatures. Under different brewing conditions, the content of β-ionone in N-brewed instant tea was lower than that in W-brewed instant tea, and the difference reached a very significant level under high temperature brewing for 3 min (0.21$\to 0.35\mathrm{\mu g}/\mathrm{L}$). The findings revealed that HCO₃⁻ in water could reduced the concentration of β-ionone in the headspace of tea infusions, both at high- and room- temperatures, with this reduction demonstrated a noticeable time accumulation phenomenon. The changes in the contents of linalool and hexanal were not significant during different brewings. Under both high-temperature and room-temperature brewing conditions, after a brewing duration of 10 s, the concentration of 1-octanol in tea infusions prepared using N (0.07 μg/L for room temperature, 0.10 μg/L for high temperature) was markedly greater than that in tea infusions prepared using W (0.05 for room temperature, 0.08 for high temperature μg/L). However, as brewing duration increased to 3 min, compared to W, the concentration of 1-octanol in tea infusions prepared with N increased under room-temperature conditions (0.07 → 0.08 μg/L) but decreased under high-temperature conditions (0.09 → 0.07 μg/L). The correlation analysis was conducted to investigate the correlation between the overall aroma quality and the concentration of aroma substances in instant tea, taking into account various brewing conditions. The findings revealed a positive correlation between the overall aroma quality score and both dimethyl sulfide ($r^2$= 0.67) and β-ionone ($r^2$= 0.71). It was further speculated that the reduction of dimethyl sulfide and β-ionone in the tea infusion, attributed to the presence of HCO₃⁻ in the brewing water, might be the underlying factor influencing the aroma quality of the tea infusion. Recent studies have demonstrated that HCO₃⁻, as a pivotal ionic species, may directly contribute to the reduction in the concentrations of dimethyl sulfide and β-ionone (Li et al., 2023).

Fig. 6.

Analysis of key aroma compounds in tea infusion brewed under different brewing conditions. RW1- brewed with room temperature (25 °C) deionised water for 10 s; RN1- brewed with room temperature NaHCO₃ simulated water for 10 s; RW2- brewing with room temperature deionised water for 3 min; RN2- brewed with room temperature NaHCO₃ simulated water for 3 min. BW1- brewed with boiling (100 °C) deionised water for 10 s; BN1- brewed with boiling NaHCO₃ simulated water for 10 s; BW2- brewed with boiling deionised water for 3 min; BN2- brewed with boiling NaHCO₃ simulated water for 3 min.

3.3.3. Exploration of the influence of HCO₃⁻ on the catechins in tea infusion and its correlation with aroma changes

It was proved that HCO₃⁻ maintained the tea infusion in a stable weakly alkaline environment, resulting in the oxidation and degradation of phenolic substances such as catechins, thereby affecting the taste of the tea infusion (Chen et al., 2012). To explore the effect of HCO₃⁻ on the interaction between taste and aroma substances, the catechin content of instant tea infusion brewed with NaHCO₃ simulated water under different brewing conditions was measured (Fig. 7A). The results showed that, under the same brewing temperature and time, the EGCG content in tea infusions brewed with N (242.56 mg/L) was significantly lower than that in tea infusions brewed with W (406.85 mg/L), while the GCG content was significantly higher in N brewed tea infusions (16.52 mg/L) than W brewed tea infusions (7.71 mg/L). This suggested that there may be a certain correlation between the changes in catechins and the changes in aroma.

Fig. 7.

The influence of simulated water under different brewing conditions on catechins in tea infusion and its relationship with aroma changes. A:Analysis of catechin content in tea infusion brewed under different brewing conditions. RW1- brewed with room temperature (25 °C) deionised water for 10 s; RN1- brewed with room temperature NaHCO₃ simulated water for 10 s; RW2- brewing with room temperature deionised water for 3 min; RN2- brewed with room temperature NaHCO₃ simulated water for 3 min. BW1- brewed with boiling (100 °C) deionised water for 10 s; BN1- brewed with boiling NaHCO₃ simulated water for 10 s; BW2- brewed with boiling deionised water for 3 min; BN2- brewed with boiling NaHCO₃ simulated water for 3 min. B: Correlation analysis between important catechins content and key aroma substances changes. The size of the circle indicates the degree of relevance.

In order to explore the relationship between catechin content and aroma under different brewing conditions, a correlation analysis was conducted between the two (Fig. 7B). The results showed that the EGCG content exhibited a significant positive correlation with aroma quality ($r^2$= 0.87), while the GCG content showed a significant negative correlation with aroma quality ($r^2$= −0.84), confirming the hypothesis that HCO₃⁻ could altered aroma quality through the taste substances-aroma substance interaction. The correlation between the content of key aroma compounds and these two catechins revealed that GCG exhibited a significant negative correlation with dimethyl sulfide ($r^2$= −0.70) and β-ionone ($r^2$= −0.86), while EGCG exhibited a significant positive correlation with β-ionone content ($r^2$= 0.72). Studies have demonstrated that EGCG plays a pivotal role in modulating the levels of key volatile compounds in Longjing tea infusion. The thermal-induced degradation of EGCG is hypothesized to drive compositional changes in the aroma profile, ultimately promoting the evolution of the tea's aromatic properties (Weng, 2019). The interaction is principally pi-pi stacking between the galloyl ring and the aromatic ring of the aroma compounds, but secondary hydrogen-bonding effects help to stabilize the complex and enhance the specificity (Jung et al., 2000). This dual-interaction mechanism potentially mediates EGCG-induced alterations in β-ionone concentrations. Thus, the variation in EGCG and GCG content that was induced by HCO₃⁻ in water may have been one of the pathways through which HCO₃⁻ affected the concentration of key aroma compounds in tea infusions, thereby altering the aroma quality.

4. Conclusion and discussion

In summary, HCO₃⁻ in water exhibited a dose-dependent influence on the aroma quality of tea infusion, transforming the chestnut-like aroma into a more sweet and cooked off-flavor. The experimental results confirmed that water characterized by high HCO₃⁻ content was found to be unsuitable for green tea brewing. This effect was particularly evident under high-temperature conditions compared to room temperature, and it intensified over time. The presence of HCO₃⁻ in water changed the concentration of aroma components in tea infusion, resulting in a significant reduction in key compounds like β-ionone and dimethyl sulfide, which subsequently decreased the overall aroma quality. HCO₃⁻ also decreased the content of EGCG in tea infusion while increasing the content of GCG, and these changes were closely correlated with alterations in key aroma substances.

Amino acids, fatty acids, and related compounds significantly influence tea aroma quality (Xiao et al., 2024). Furthermore, prior research has confirmed molecular interactions between catechins and volatile aroma compounds (Dufour & Bayonove, 1999). And it was known that catechins could modify the threshold values and release characteristics of aroma compounds (Lorrain et al., 2013). Thus, HCO₃⁻ was probability to have altered the aroma quality of tea infusion by influencing the interaction between taste and aroma. Specifically, HCO₃⁻ may affect key aroma compounds such as dimethyl sulfide and β-ionone through its influence on EGCG. Dimethyl sulfide has been found to be produced either through thermal degradation of sulfur-containing amino acids (e.g., methionine) or via β-lyase enzymatic activity in tea leaves. Notably, volatile aromatic compounds in wine, including esters, benzaldehyde, and limonene, have been shown to be significantly reduced by polyphenols, a phenomenon analogous the catechin-associated aroma modifications observed in the present study (Dufour & Bayonove, 1999). Specifically, EGCG can form adducts with glucose and methional, thereby decreasing the reactivity of dimethyl sulfide precursors in tea and significantly inhibiting dimethyl sulfide formation (Yao et al., 2025). β-Ionone is derived from the oxidative degradation or enzymatic cleavage of carotenoids such as β-carotene and lutein, or through the hydrolysis of β-ionone glucosides by β-glucosidase. During dark tea fermentation, a simultaneous decrease in catechins and β-ionone was observed (Zhang et al., 2025). Current research indicates that EGCG may bind with β-ionone via π-π stacking or hydrogen bonding, reducing its volatility (Lyu et al., 2021).

The regulatory mechanism of HCO₃⁻ mediated by EGCG on aroma, as well as the influence of the interaction between HCO₃⁻ and other ions on aroma characteristics, remains to be further investigated. Additionally, HCO₃⁻ generally imparted a more sweet and cooked aroma to infused tea soup. However, this study did not pinpoint the specific substances responsible for this cooked off-flavor characteristic, which warrants further attention in future research. This research provides actionable insights for tea industry applications, enabling both smarter water source selection for differentiated tea varieties and quality enhancement in processed tea products like concentrates and extracts. Furthermore, as widespread buffering agents in aqueous solutions, HCO₃⁻ significantly influence the flavor profiles of many beverages. Elevated levels of HCO₃⁻ can reduce the perception of fruity, acidic, and sweet notes in coffee, affecting key flavor compounds such as 1-furfurylpyrrole, 5-methylfurfural, and furfuryl methyl sulfide (Chen et al., 2024). Moreover, increasing HCO₃⁻ concentrations in beer have been associated with enhanced bitterness and a more pronounced astringent mouthfeel, which collectively contribute to an undesirable sensory experience (Gao, 2022). Therefore, in brewing practices, the content of HCO₃⁻ in brewing water must be carefully controlled, and the direct use of high-HCO₃⁻, high-pH water sources should be avoided (Cao et al., 2021), given its essential role in ensuring consistent flavor quality.

Ethical approval and consent

Ethical permission was not required. The rights and privacy of the participants were fully protected, participation was not forced, and the research requirements and risks were fully disclosed before sensory evaluation, with informed consent being obtained from all participants.

CRediT authorship contribution statement

Mingming Zhang: Writing – original draft, Investigation, Formal analysis, Data curation. Meiqin Li: Investigation, Formal analysis, Data curation. Fang Wang: Investigation. Jianxin Chen: Investigation. Yifan Li: Investigation. Gensheng Chen: Investigation. Yuwan Wang: Investigation. Zhihui Feng: Writing – review & editing, Methodology, Funding acquisition, Formal analysis. Junfeng Yin: Writing – review & editing, Supervision, Funding acquisition, Conceptualization.

Funding

This work was financially supported by the National Natural Science Foundation of China (32272771, 32102444) and the Agricultural Sciences and Technology Innovation Program (ASTIP-1610212024001, Y2024CG04).

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.

Data availability

The data that has been used is confidential.