Accepted Manuscript

Title: Enantiomeric and Quantitative Analysis of Volatile

Terpenoids in Different Teas (Camellia Sinensis)

Authors: Yin Zhu, Chen-Yang Shao, Hai-Peng Lv, Yue Zhang,

Wei-Dong Dai, Li Guo, Jun-Feng Tan, Qun-Hua Peng, Zhi Lin

PII: S0021-9673(17)30219-4
DOI: http://dx.doi.org/doi:10.1016/j.chroma.2017.02.013
Reference: CHROMA 358283

To appear in: Journal of Chromatography A

Received date: 7-10-2016

Revised date: 6-2-2017
Accepted date: 8-2-2017

Please cite this article as: Yin Zhu, Chen-Yang Shao, Hai-Peng Lv, Yue Zhang,
Wei-Dong Dai, Li Guo, Jun-Feng Tan, Qun-Hua Peng, Zhi Lin, Enantiomeric and
Quantitative Analysis of Volatile Terpenoids in Different Teas (Camellia Sinensis),
Journal of Chromatography A http://dx.doi.org/10.1016/j.chroma.2017.02.013

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
 Enantiomeric and Quantitative Analysis of Volatile Terpenoids in

Different Teas (Camellia Sinensis)

Yin Zhua,1, Chen-Yang Shaoa, b, 1, Hai-Peng Lva*, Yue Zhanga, Wei-Dong Daia, Li Guoa, Jun-Feng

Tana, Qun-Hua Penga, Zhi Lina**

1
These authors contributed equally to this work.

(a. Key Laboratory of Tea Biology and Resource Utilization of Ministry of Agriculture, Tea Research Institute,

Chinese Academy of Agricultural Sciences, Hangzhou 310008, China)

(b. Graduate School of Chinese Academy of Agricultural Sciences, Beijing 100081, China)

*Corresponding Author

Prof. Dr. Zhi Lin, Dr. Hai-Peng Lv

Key Laboratory of Tea Biology and Resource Utilization of Ministry of Agriculture,

Tea Research Institute, Chinese Academy of Agricultural Sciences, Hangzhou 310008,

China

E-mail address: teachem108@hotmail.com (Hai-Peng Lv), linzhi@caas.cn (Zhi Lin).

Tel./fax: +86 571 86650617

1
Highlights

• A new method for chiral separation of 12 volatile terpenoids in tea was established
• The 12 terpenoids in different teas presented diverse enantiomeric distributions
• Unique distributions were observed in Indonesia white and some black teas
• Significant concentration differences of terpenoids were shown among different
teas
• The results provide theoretical basis for quality or authenticity assessment of tea

Abstract:

Volatile terpenoids play important roles in the formation of tea aroma quality due to

their pleasant scents and low odor thresholds. Most volatile terpenoids contain

stereogenic centers, which results in various stereo distributions of their enantiomers

and diastereoisomers in different types of tea. However, the distribution

characteristics of terpenoid enantiomers in teas were still unclear, which poses an

obstacle to the scientific understanding of tea aroma. In this work, a new and efficient

analysis approach based on headspace solid phase microextraction (HS-SPME)-chiral

gas chromatography-mass spectrometry (GC-MS) was established to analyze 12 pairs

of familiar terpenoid enantiomers in different teas. The extraction efficiency of the

HS-SPME method to extract volatile terpenoids in teas was the greatest when using

CAR-DVB-PDMS (50/30 μm) fibers and 1:10 proportions between tea and boiling

water at a 50 °C extraction temperature for 40 min, and the stability observation of

enantiomeric ratios of the terpenoids well proved the feasibility of the extraction

method. The favorable limits of detection, limits of quantitation, repeatability,

linearity, and concentration ranges of each terpenoid enantiomer demonstrated the

2
repeatability and reliability of the analytical approach. The enantiomeric and

quantitative analyses indicated that S-limonene, S-linalool, (2S, 5S)-linalool oxide A,

(2S, 5R)-linalool oxide B, R-4-terpineol, (2S, 5R)-linalool oxide C, (2S, 5S)-linalool

oxide D, S-α-terpineol, R-α-ionone, peak 1 of theaspirane A and peak 2 of theaspirane

B were the major terpenoid components in most Chinese teas; instead, higher

proportions of the opposite enantiomers of the above terpenoids were frequently

detected in black teas with large leaf origin and Indonesia white teas. Besides, great

diversities of enantiomeric ratios and concentrations among different teas were

observed. Furthermore, partial least-squares discriminant analyses were performed to

distinguish the concentration differences of the terpenoid enantiomers among different

teas; the analysis results indicated that highly significant concentration differences

existed between large and small leaf origins of black teas, and significant differences

of the concentrations of linalool oxides A, B and C were observed between green,

white and dark teas. The successful application of this chiral analysis technique of tea

aroma will lay a scientific foundation for further quality assessment, botanical origin

determination and authenticity assessment of teas.

Abbreviations

CAR-DVB-PDMS, Carboxen/divinylbenzene/polydimethylsiloxane; DVB-PDMS,

divinylbenzene/polydimethylsiloxane; PA, polyacrylate; ER, enantiomeric ratio; GC,

gas chromatography; HCA, hierarchical clustering analysis; HS-SPME, headspace

solid‐phase microextraction; LOD, limits of detection; LOQ, limits of quantitation;

3
PLS-DA, partial least-squares discriminant analysis; R2, correlation coefficients; RSD,

relative standard deviation; TIC, total ion chromatogram; VIP, variable importance in

the projection.

Keywords: Chiral GC-MS; Enantiomeric ratio; HS-SPME; Tea; Volatile

terpenoids

1. Introduction

Essential oils of tea (Camellia Sinensis) are composed of diverse volatile

compounds at various concentrations, and the volatiles can be classified into

aliphatics, aromatics and volatile terpenoids based on their basic carbon chains and

parents of compounds [1]. Terpenoids have a significant influence on the formation of

tea aroma quality due to their low odor thresholds and pleasant fragrances, and

hundreds of them have already been separated and identified by gas

chromatography-mass spectrometry (GC-MS) [2]. Most identified terpenoids contain

one or more stereogenic center, which results in various stereo distributions of their

enantiomers and diastereoisomers in different teas. Among them, linalool, linalool

oxides (A, B, C and D), citronellol, nerolidol (Z, E), citronellal, α-ionone, theaspirane

(A, B), α-terpineol, α-pinene, limonene, phellandrene and 4-terpineol are considered

to be the familiar aroma components with high contents in most teas [3] (Figure 1).

It has been verified that different enantiomers of the same molecule may possess

4
different properties including odor characteristics, odor thresholds and even biological

activities [4]. For instance, R-linalool (odor threshold of 0.8 μg/L) has a woody and

lavender-like smell [5] and shows sedative effects on autonomic nerve activity and

mood states at a very low intensity [6], whereas S-linalool (odor threshold of 7.4 μg/L)

has a sweet, flowery and orange-like aroma [7] and shows no sedative effects [6].

S-Citronellol presents a gentle rose-like smell, but R-citronellol presents a sweet and

rose-like smell [8]. S-α-Ionone (odor threshold of 20-40 μg/L) has a woody odor,

whereas R-α-ionone (odor threshold of 0.5-5 μg/L) has a violet-like aroma, and

racemic α-ionone has a strong sweet-floral odor, reminiscent of violets [9].

S-Limonene exhibits a lemon-like aroma, whereas R-limonene exhibits a turpentine

smell [10]. In view of the different aroma characteristics of enantiomers, it is

speculated that different enantiomeric distributions of terpenoids could trigger various

fragrance characteristics, and would significantly influence the aroma quality of tea

due to permutations and combinations of terpenoids with different enantiomeric

excess values [4] and concentrations. Therefore, determining the enantiomeric

distributions of volatile terpenoids in different types of tea is a significant and

valuable research topic.

Enantiomeric analysis of terpenoids is currently widely studied in the field of food

chemistry. It has been found that R-linalool exists in basil oil, hops and Japanese

pepper, whereas S-linalool exists in strawberries, orange oil, and cocoa products

[11-12]. Racemic linalool oxides have four chemical constructions named linalool

oxide A, B, C, and D; the chiral separation of the four pairs of enantiomers in linalool

5
oxides has been studied in grape [13], honey [14] and food essential oil [15].

R-α-Ionone has been detected in raspberry [16], violet [17] and carrot [18].

Additionally, enantiomers of α-terpineol, α-pinene, limonene, phellandrene and

4-terpineol have been successfully separated in essential oils of lamiaceae [19],

cardamom [20], citrus [21], lemon [22] and passion fruit [4]. In light of the

enantiomeric distribution characteristics of terpenoids in certain foods, chiral analysis

technology has been successfully applied in the quality assessment, determination of

botanical origin and authenticity assessment of honey, Melaleuca alternifolia oils and

fruit beverages [14, 23-24].

Although great progress has been made, research on the enantiomeric analysis of

terpenoids in teas is still at an initial stage. The enantiomers of linalool and linalool

oxides in black and oolong teas were originally separated in 1994 [25], and it has

proven that the enantiomeric ratios and even the major configurations of linalool and

linalool oxides presented different distributions in the different teas; very recently, the

enantiomeric distribution of linalool in various types of teas, tea cultivars and tea

samples during processing has been reported by our group [26], and a similar

distinction was observed. Beyond those, however, few chiral separations of other

terpenoids that likely play important roles in the formation of tea aroma quality have

been reported.

Enantiomeric analysis of terpenoids has generally been performed using chiral gas

chromatography. However, multifarious structures, similar boiling points and

different adhesive abilities in chiral stationary phases of terpenoids cause great

6
difficulty separating them in a single chiral chromatographic system. The multi-peak

overlap phenomenon among different terpenoids and the unsuccessful separation of

terpenoid enantiomers frequently occurred by using unsuitable chiral GC columns.

Therefore, the optimization of chiral separation efficiency of terpenoids is the most

critical step of the chiral analysis and presents many challenges. On the other hand, in

prior work, identification of terpenoid enantiomers was frequently performed through

maintaining consistency between the retention time of peaks of both standard and

samples to be tested under the same experimental conditions by using a GC-flame

ionization detector. Nevertheless, erroneous judgment of peaks might occur from

similar retention times of other compounds with similar chemical properties,

especially for some terpenoids with relatively low contents in teas (such as α-pinene,

limonene etc.). Therefore, a MS technique is quite necessary in the enantiomeric and

quantitative analysis of terpenoids in teas.

In this work, a simple and effective analytical approach for enantiomeric and

quantitative analysis of various terpenoids (up to 12 pairs) in teas by using HS-SPME

combined with chiral GC-MS technology has been successfully established.

Subsequently, the enantiomeric distribution of 12 pairs of separated terpenoid

enantiomers and their corresponding contents in various representative tea samples

(up to 45 samples) have been examined. Finally, the distributing characteristics of

terpenoid enantiomers in different types of teas were illuminated by multivariate

statistical analysis.

7
 2. Materials and methods

2.1 Reagents and materials

Standards: Linalool: (±)-linalool, R-linalool; linalool oxides: (±)-linalool oxide

(mixture of A and B), (±)-linalool oxide (mixture of C and D); 4-terpineol:

(±)-4-terpineol, R-4-terpineol, S-4-terpineol; α-terpineol: (±)-α-terpineol,

R-α-terpineol; citronellol: (±)-citronellol, S-citronellol; citronellal: (±)-citronellal,

S-citronellal; nerolidol: Z-(±)-nerolidol, E-(±)-nerolidol; α-ionone: (±)-α-ionone;

α-pinene: (±)-α-pinene, R-α-pinene, S-α-pinene; limonene: (±)-limonene, R-limonene,

S-limonene; α-phellandrene: (±)-α-phellandrene, S-α-phellandrene; theaspirane: (Z,

E)-Theaspirane. The above standards were purchased from J&K Scientific Ltd.

(Beijing, China). The standard solutions were stored at 4 ºC. Ether (GC grade) was

purchased from Tedia (Fairfield, OH, USA). Distilled water was purchased from

Wahaha Group Company (Hangzhou, China).

Carboxen/divinylbenzene/polydimethylsiloxane (CAR-DVB-PDMS; 50/30 μm),

divinylbenzene/polydimethylsiloxane (DVB-PDMS; 65 μm), and polyacrylate (PA;

85 μm) microextraction fibers were purchased from Supelco (Bellefonte, PA, USA).

2.2 Tea Samples

Samples of green, black, white, oolong, and dark tea were purchased from local

markets in China and other countries in 2015-2016. Seven green teas, nineteen black

teas, six oolong teas, eight white teas and five dark teas were used in this study; their

detailed names and sources are listed in Table 1. The tea samples were stored in a

refrigerator at a temperature below −20 °C for further analysis. Among these samples,

8
Maoshan Bai tea, Yingde-1 black tea, Tieguanyin oolong tea, Baimudan-1 white tea

and Pu-erh-1 dark tea as representative samples of each type of tea were mixed in

equal amounts and the mixed samples were used for optimization of HS-SPME

parameters, checking the repeatability and reliability of the analytical method and

preparation of the“volatile-free” tea sample.

“Volatile-free” tea sample was considered as a blank matrix which identical to the

commercial tea samples without volatile components [26-27]. “Volatile-free” tea was

prepared by repeatedly heating the tea infusion (10.0 g grounded tea dissolved in 30

mL boiling water) at 50 °C under reduced pressure in a rotary evaporator for five

times to remove all volatile terpenoids, and then drying the tea powder at 80 °C for 5

days. The standard “volatile-free” tea was examined by optimized HS-SPME-chiral

GC-MS method, and the signal-to-noise ratios (S/N) of terpenoids in the GC

chromatogram were all lower than 3.0.

2.3 Instrument and equipment

A solid-phase microextraction (SPME) fiber holder-manual sampler was purchased

from Supelco (Bellefonte, PA, USA), and the extraction bottle for extracting volatile

terpenoids of teas was fabricated in-house; analyses of terpenoid enantiomers in the

tea samples were performed using an Agilent GC-7890A system (Agilent, USA)

coupled to a 5975C quadrupole MS (MSD; Agilent Technologies). A series of chiral

capillary columns with different chiral stationary phases were applied to separate the

terpenoid enantiomers, including an Astec CHIRAL DEX B-PM (30 m × 0.25 mm ×

0.12 μm), Astec CHIRAL DEX B-DA (30 m × 0.25 mm × 0.12 μm), Astec CHIRAL

9
DEX G-TA (30 m × 0.25 mm × 0.12 μm), β-DEX 110 (30 m × 0.25 mm × 0.25 μm),

β-DEX 120 (30 m × 0.25 mm × 0.25 μm), Astec CHIRAL DEX G-DM (30 m × 0.25

mm × 0.12 μm) (all from Supelco, Bellefonte, PA, USA) and a

CP-Cyclodextrin-B-2,3,6-M-19 (50 m × 0.25 mm × 0.25 μm), CHIRASIL-DEX CB

(25 m × 0.25 mm × 0.25 μm) and HP-Chiral 20β (30 m × 0.25 mm × 0.25 μm) (all

from Agilent, Santa Clara, CA, USA).

2.4 Optimization of HS-SPME parameters

To establish an optimal method to extract the contents of volatile terpenoids at a

satisfactory level, several experimental HS-SPME parameters including the type of

extraction fiber, volume of extraction bottle, tea/water brewing proportions, extraction

temperature, and extraction time were screened in sequence.

The optimization of HS-SPME parameters was carried out according to our

previous work, and the original experiment conditions were the same as the extraction

of linalool enantiomers [26]. Firstly, three types of microextraction fibers

(CAR-DVB-PDMS (50/30 μm), DVB-PDMS (65 μm), and PA (85 μm)) were

examined under equivalent extraction conditions. Each fiber was preconditioned at

250 ºC for 1.0 h in the injection port of the GC-MS to fully activate their extraction

ability. For each tea sample, 1.0 g of tea samples was accurately weighed into a 250

mL extraction bottle, and 6.0 mL of boiling water were added. The bottle was

immediately placed in a water bath to equilibrate for 3.0 min at 60 °C, and then the

SPME fiber was exposed to the headspace of the bottle for 1.0 h at 60 °C. The SPME

fiber was then introduced into the GC injector and kept there for 3.0 min to allow

10
thermal desorption of the analytes. Subsequently, the tea/water brewing proportions,

extraction temperature, and extraction time were tested at 30, 40, 50, 60, and 70 °C

for 30, 40, 50, 60, and 70 min using different brewing proportions of tea and water

(1:4, 1:6, 1:8, 1:10, and 1:12; w(g)/v(mL)). All samples were extracted in triplicate to

ensure the repeatability of experiments.

2.5 Chiral GC-MS analysis

GC conditions: the temperatures of the GC injector and the transfer line were set to

220 °C and 250 °C, respectively; helium (99.999%) was used as the carrier gas at a

constant flow of 1.2 mL/min; ether solutions of standards were injected using an

Agilent 7683B autosampler (Agilent, USA) with an injection volume of 1.0 μL and

split ratio of 10:1; the extracted terpenoids of teas were analyzed by manual injection

through the front GC injector; the column temperature program was initial

temperature at 50 °C (held for 2 min), increased to 150 °C at 2 °C/min (held for 10

min), and increased at 4 °C/min to 180 °C (held for 5 min).

MS conditions: the electron ionization, ion source temperature and the electron

multiplier were set at −70 eV, 220 °C, 1780 V, respectively. The mass range was set as

m/z 30-600 u.

2.6 Enantiomeric analysis of volatile terpenoids in tea samples

The terpenoid enantiomers in tea samples were identified by comparing their

retention times and mass spectrums with those of terpenoid standards analyzed under

the same experimental conditions.

Enantiomeric analysis of separated terpenoids in tea samples was performed under

11
the optimized conditions. Enantiomer ratio (ER), which may widely be applied to the

research area of flavors and fragrances in the future, were applied to describe the

enantiomeric distributions of volatile terpenoids in teas [28]. ER values were

calculated from the following equation (E1: the percent composition of one of the

enantiomers in a certain terpenoid; E2: the percent composition of another enantiomer

in a certain terpenoid; and E1 + E2 =100%):

ER = E1:E2

2.7 Quantitative analysis of volatile terpenoids in tea samples

The volatile terpenoid concentrations in tea samples were calculated according to

the standard addition method. 10 μL mixture of standard solutions at increasing

concentrations (more than 8 gradients and diluted in ether; including S-α-pinene,

R-limonene, R-linalool, S-4-terpineol, R-α-terpineol, α-ionone, mixture of linalool

oxides A-B, mixture of linalool oxides C-D and mixture of theaspiranes A-B) were

sequentially added to the “volatile-free” tea sample under the optimized HS-SPME

conditions. The calibration curves of the terpenoid enantiomers were established

according to the peak areas of the selected quantitative ions and the corresponding

concentrations of standards. The contents of a target terpenoid enantiomer in tea

samples were calculated by plugging peak areas of the terpenoid enantiomers in the

tested samples into the corresponding linear equations of standards under the same

conditions. Note that the identical calibration curve was applied in the quantification

of the two enantiomers of the same terpenoid or the four isomers of the terpenoid with

12
two stereogenic centers (such as linalool oxides A-B, linalool oxides C-D and

theaspiranes A-B) because they present the same characteristics ions in the mass

spectrums.

2.8 Validation of the optimized method

The optimized approach for enantiomeric and quantitative analysis of terpenoids in

tea samples was validated according to linearity, detection and quantification limits,

and repeatability by using the mixed tea sample as representative. A series of

concentration gradients of different terpenoid standards was prepared by diluting the

stock solutions in ether, and then the linearity of each terpenoid was examined and

expressed through correlation coefficients. The detection limits of terpenoids were the

corresponding concentrations when the S/N of the chromatographic peak was

approximately 3, and the quantification limits were the corresponding concentrations

when the S/N of the chromatographic peak was approximately 10. The repeatability of

the method for each terpenoid was represented by using the relative standard

deviation (RSD) of ER values based on five repetitions of the mixed tea sample.

2.9. Statistical analysis

Bar graphs and the total ion chromatograms (TICs) of standards and tea samples

were drawn using Excel 2010 (Microsoft, USA); the mass spectrums of standards

were drawn using OriginPro 9.0 (OriginLab, Hampton, USA); partial least squares

discriminant analysis (PLS-DA) was performed using SIMCA-P 11.5 software

(Umetrics, Umea, Sweden); and hierarchical clustering analysis (HCA) was

performed using MultiExperiment Viewer 4.8.1 (Oracle Corporation, Redwood

13
Shores, USA).

3. Results and discussion

3.1. Optimization of chiral capillary columns

As mentioned, the selection of chiral capillary columns is the key step of this study.

A series of gas chromatographic columns filled by different chiral stationary phases

(modified β-cyclodextrins and γ-cyclodextrins) were applied to enantiomerically

separate the racemic terpenoid standards under the same experimental conditions. The

optimal chiral capillary column should separate the maximum numbers of terpenoid

enantiomers (standards) with satisfactory resolutions while avoiding the multi-peak

overlap phenomenon of mixed standards. A total of 17 racemic terpenoids (2.1

Reagents and materials) were examined in this study, and a comparison of

separating capacity of different chiral capillary columns is presented. As shown in

Table 2, the gas chromatographic columns filled by modified γ-cyclodextrins (Astec

CHIRAL DEX G-TA and Astec CHIRAL DEX G-DM) presented unsatisfied

separation efficiencies, and only racemic α-ionone could be separated. The modified

β-cyclodextrin stationary phases showed better separation efficiencies except

2,6-di-pentyl-O-methyl derivative of β-cyclodextrin (Astec CHIRAL DEX B-DA). A

superior separation effect was observed when the same modified β-cyclodextrin

derivative were diluted in polysiloxanes [29-30], for instance, β-DEX 120 (20% full

methylation of β-cyclodextrin) and HP-Chiral 20β (similar to β-DEX 120) could

separate at most four pairs of racemic terpenoids, while β-DEX 110 (10% full

14
methylation of β-cyclodextrin) showed stronger separating capacity and 9 pairs of

terpenoid enantiomers were separated. Moreover, the separating capacity was further

enhanced by using the column filled by another modified β-cyclodextrin stationary

phase named 2,3,6-tri-O-methyl derivative of β-cyclodextrin (Astec CHIRAL DEX

B-PM), and 11 pairs of terpenoid enantiomers (linalool, linalool oxide A, C, D,

4-terpineol, α-terpineol, α-ionone, α-pinene, limonene and theaspiranes A-B) were

successfully separated. Although citronellol, citronellal, Z-nerolidol, E-nerolidol and

α-phellandrene were still inseparable after all attempts (such as changing the chiral

columns, regulating the GC conditions and changing concentrations etc.), the present

chiral separation efficiency was satisfying in the research area of tea aroma, and the

maximum numbers of volatile terpenoid enantiomers separated in a single chiral

chromatographic system have been obtained. Therefore, the Astec CHIRAL DEX

B-PM capillary column was considered the optimal chiral gas chromatographic

column, and β-DEX 110 was also applied to detect the ER values and contents of

linalool oxides A-B in tea samples. Subsequently, the retention times and

characteristic ions of separated terpenoid enantiomer standards were recorded for their

identification in the tea samples, and the most notable ions were considered the

quantitative ions of the corresponding terpenoids for further quantitative analysis

(Figure 2, Table 3).

In addition, for lacking of the corresponding standards, the absolute configurations

of linalool oxides A-D were determined according to the reported reference. Linalool

oxides A-D in the mixed tea samples were analyzed by using the same chiral capillary

15
column (CP-Cyclodextrin-B-236-M-19) and temperature program as those reported in

the reference [25], and then the ERs of linalool oxides A-D in the mixed tea samples

were calculated. Subsequently, the peak order of linalool oxides A-D enantiomers by

using Astec CHIRAL DEX B-PM or DEX 110 columns was determined through

comparing their ERs which obtained from the above step, and the detailed retention

times and absolute configurations of them were listed in Figure 2 and Table 3.

Furthermore, aroma components of the mixed tea sample were analyzed to measure

the feasibility and necessity of the experimental investigation by using HS-SPME

with chiral GC-MS technology (using the Astec CHIRAL DEX B-PM as the chiral

column), and the TIC is shown in Figure 3A. The qualitative analysis of enantiomers

was performed by verification of standards and mass spectrum identification and

indicated that S-limonene, R-limonene, R-linalool, S-linalool, (2R, 5R)-linalool oxides

A, (2S, 5S)-linalool oxide A, (2R, 5S)-linalool oxide B, (2S, 5R)-linalool oxide B (the

enantiomers of linalool oxide B could not be completely separated), S-4-terpineol,

R-4-terpineol, (2S, 5R)-linalool oxide C, (2S, 5S)-linalool oxide D, (2R, 5R)-linalool

oxide D, S-α-terpineol, R-α-terpineol and a single enantiomer of α-ionone could be

detected with various abundance levels by extracting their characteristic ions (Figure

3B-H). Nevertheless, the peaks of α-pinene and theaspirane enantiomers failed to be

detected in the mixed tea samples, and their extremely low concentration might be

responsible for the above results. Overall, the approach was practicable and requires

optimization by other experimental parameters.

3.2 Optimization of extraction of terpenoids in teas

16
 HS-SPME was widely considered a convenient and solvent-free analytical

technology [2], and thus was used for the extraction of terpenoids in teas. Several

experimental HS-SPME parameters that may possibly influence extraction efficiency

were screened to obtain the greatest amount of volatile terpenoids in tea. Summation

of GC peak areas of the separated terpenoid enantiomers in tea was used to compare

the investigated parameters. Firstly, three common silica fiber materials

(CAR-DVB-PDMS (50/30 μm), DVB-PDMS (65 μm), and PA (85 μm)) were tested

under the same conditions to optimize the extraction fiber. As shown in Figure 4A,

the extracted terpenoid peak areas demonstrated wide variation among the three fibers,

with CAR-DVB-PDMS (50/30 μm) showing the strongest extraction ability, followed

by the DVB-PDMS (65 μm) and PA (85 μm) fibers. Thus, CAR-DVB-PDMS (50/30

μm) fiber was chosen as the optimal fiber for HS-SPME.

Next, the proportion between tea and boiling water was screened, aimed at

extracting the maximum amount of target terpenoids with as little boiling water as

possible. The largest terpenoid peak area occurred with a 1:10 proportion between tea

and boiling water, and the remaining proportions (1:4, 1:6, 1:8 and 1:12) exhibited

lower or nearly equal amounts (Figure 4B). Hence, the 1:10 proportion between tea

and boiling water was selected.

Subsequently, a series of extraction temperatures were investigated under fixed

other experimental conditions. As shown in Figure 4C, the total peak areas of

extracted terpenoids increased as the extraction temperature increased from 30 ºC to

50 ºC, however, the total area sharply decreased when the temperature continued to

17
rise to 60 ºC, and the least quantity of terpenoid was obtained when the extraction

temperature was 70 ºC. As a result, 50 °C was considered to be the best extraction

temperature for the subsequent optimization.

Finally, the extraction time of terpenoids in the mixed tea samples was studied

(Figure 4D). The results proved that the extraction time was not proportional to the

extraction efficiency, and the summation of peak areas of terpenoids was the greatest

after 40 min; the terpenoid extraction efficiency gradually reduced during further

extraction. Consequently, the extraction efficiency of the HS-SPME method to extract

terpenoids in teas was the best when using CAR-DVB-PDMS (50/30 μm) fibers, a

1:10 proportion between tea and boiling water, and a 50 °C extraction temperature for

40 min.

3.3 The stability observation of terpenoid enantiomers in the tea sample

It was well known that the stabilities of chiral compounds might be influenced by

many environmental factors especially temperature, time, humidity and pH et al. In

our study, the sampling conditions were screened in wide ranges, and it was necessary

to investigate the stability of the terpenoid enantiomers in the mixed tea samples while

changing the HS-SPME parameters. It was found that the types of fibers and the

proportions between tea and boiling water almost did not affect the stability of ERs of

terpenoids, and slight effects were observed when the extraction temperature and time

were changed. As shown in Figure S1A (see details in the supporting information),

ERs of most detected terpenoids were constant or fluctuated in acceptable ratio ranges

(±5%) when the temperature was increased from 30 °C to 70 °C, such as linalool,

18
linalool oxide A, linalool oxide C, α-terpineol and α-ionone; while ERs of 4-terpineol

and linalool oxide D tended to be stable when the temperature ranged from 50-70 °C,

and ER of limonene was sharply changed when the extraction temperature was raised

to 70 °C. Similarly, the ERs of the most terpenoids were stable (with differences less

than 5%) when changing the extraction time, whereas the ratios of R-limonene was

sharply decreased when the extraction time was extended to 70 min (Figure S1B in

the supporting information). Therefore, suitable stabilities of ERs of the detected

terpenoids in the mixed tea samples were presented under the optimized HS-SPME

conditions (3.2 Optimization of extraction of terpenoids in teas).

3.4 The preparation of quantitative analysis for separated terpenoid

enantiomers and validation of the optimized HS-SPME-chiral GC approach

Quantitative analysis of aroma components in tea has usually been performed using

a stable and volatile compound (such as ethyl decanoate) as an internal standard, with

the respective concentrations of aroma components calculated according to the

proportion of the peak area between the tested compounds and internal standard.

However, different abundances of analytes and internal standards have generally led

to large errors of quantitative results of aroma components in tea. In this study, the

quantitative analysis of terpenoid enantiomers proceeded using standard addition

method, and the standard curves of related terpenoid standards in wide concentration

ranges were established under the same matrix as the tested tea samples, and then the

linear equations between the concentrations of terpenoid standards and peak areas of

the quantitative ions of terpenoid standards were established.

19
 As shown in Table 3, the concentration ranges, correlation coefficients (R2), limits

of detection (LOD), limits of quantitation (LOQ), and relative standard deviations

(RSDs, n=5) of the terpenoid enantiomers were investigated. The LOD and LOQ of

each terpenoid enantiomer were satisfactory, with ranges of 0.11-1.59 ng/g and

0.53-7.94 ng/g, respectively. All of the terpenoid enantiomers showed good linearity

in wide concentration ranges (up to 20410 ng/g) with excellent correlation coefficients

(R2, up to 0.9999). Moreover, the low RSD values (1.04-3.98%) of the terpenoid

enantiomers clearly demonstrated the good repeatability of the optimized

HS-SPME-chiral GC-MS approach.

3.5 The enantiomeric and quantitative analysis of volatile terpenoid

enantiomers in different types of teas

Subsequently, a total of forty-five representative teas that could be classified into

five types (green, black, oolong, white and dark teas) were examined by the optimized

HS-SPME-chiral GC-MS approach. The ER values and concentrations of the 12 pairs

of terpenoid enantiomers in tea samples are summarized. As shown in Table S1 (see

details in the supporting information), the 12 terpenoids existed in different teas

with various concentration levels. Among them, α-pinene could be detected only in

Xinyang Maojian, however, it failed to be quantitated because of its extremely low

concentration. Limonene was found in all the black teas, most of the oolong and white

teas, and some green teas with generally low concentrations, whereas R-limonene

presented extremely high concentrations in Earl Grey teas (up to 10833 ng/g); linalool

and linalool oxides A and B could be detected in all forty-five tea samples, and they

20
presented high concentrations in most teas, especially in black and white teas

(R-linalool: up to 17417 ng/g in Silver Needle-2 which is from Indonesia; linalool

oxide A: up to 1440 ng/g in Keemun black tea; linalool oxide B: up to 2896 ng/g in

Keemun black tea). 4-Terpineol was detected in most black and white teas and two of

the oolong teas with a similar and relatively low concentration level, and

R-4-terpineol in Earl Grey-2 presented highest concentrations (up to 11.58 ng/g).

Linalool oxide C could be detected in 22 tea samples, including Matcha samples,

some oolong and dark teas and most black and white teas, and it presented the highest

concentration in Wuye Dancong (up to 683.6 ng/g); linalool oxide D was widely

distributed in most teas with various concentrations except Matcha, Earl Grey teas and

some oolong teas, and it presented a higher concentration level in black teas and

Chinese white teas, especially in Darjeeling-1 black tea ((2S, 5S)-Linalool oxide D, up

to 875.4 ng/g). α-Terpineol could be detected in all black, white and dark teas and

some oolong teas with concentration ranges from trace (in some black teas, oolong tea

and Pu-erh teas) to 1114 ng/g (R-α-Terpineol, in Earl Grey-1 black tea); α-ionone

could be detected in all white and dark teas, most black teas and some green and

oolong teas, with the highest concentration detected in Baihao Yinzhen-1 (peak 1, up

to 11.93 ng/g) and Rizhao green tea (peak 2, up to 8.80 ng/g). Both theaspirane A and

B presented low concentration levels in teas, and they could be detected in almost half

of the tested tea samples, with their highest concentrations in Fenghuang Shuixian

oolong tea (theaspirane A, 9.28 ng/g) and Lingtou Dancong oolong tea (theaspirane B,

9.72 ng/g), respectively.

21
 On the other hand, the 12 terpenoids displayed various enantiomeric distributions

and even distinct major stereo configurations in different types of teas. The varieties

of detected terpenoid enantiomers were the fewest in green teas, and the two

enantiomers of 4-terpineol, α-terpineol and theaspirane A were not found in any green

teas. Single and same stereo configurations of (2S, 5S)-linalool oxide A, (2S,

5R)-linalool oxide B and (2S, 5S)-linalool oxide D were observed in green teas

(Figure 5C, D, G). Notably, both the two enantiomers of linalool with a wide range of

ER values (5:95-44:56 R/S) could be detected, and the proportions of S-linalool were

higher than those of R-linalool in all of the green tea samples (Figure 5B); it was

interesting to note that the proportions of S-linalool in teas from the north of China

(81-89%; Xinyang Maojian and Rizhao green teas) were generally lower than those of

teas from the south of China (92-95%; Lu’an Guapian, Maoshan Bai and Biluochun),

and the climatic differences between southern and northern China might be

responsible for the phenomenon. The highest proportions of R-linalool existed in

Matcha samples (up to 44%), and its special processing technology probably led to

the large gaps between Matcha samples and other green teas. Linalool oxide C was

only detected in Matcha samples, and the single stereo configuration of (2R,

5S)-linalool oxide C was observed (Figure 5F). In addition, a single stereo

configuration of theaspirane B (the absolute configuration of peak 2) and a unique

configuration of α-ionone (45:55 ER; the absolute configuration of peak 1) could be

found only in Rizhao green tea.

A series of Chinese and other worldwide famous black teas were examined in our

22
study, and the most enriched terpenoid enantiomers have been detected. As shown in

Figure 5, linalool oxides A-D, α-ionone and theaspiranes A and B presented

consistent major configurations among most black teas; notably (2S, 5R)-Linalool

oxide C and α-ionone (the absolute configuration of peak 2) existed in all black teas

with single configurations. It was noteworthy that Yingde and Dianhong black teas

exhibited similar enantiomeric tendencies for most detected terpenoids, for instance,

the relatively high proportions of R-limonene (61-69% for Yingde; 63-70% for

Dianhong), R-linalool (64-82% for Yingde; 86-92% for Dianhong) and R-α-terpineol

(61-68% for Yingde; 63-70% for Dianhong) in them were contrast to those of other

Chinese black teas; the ratios of (2S, 5S)-linalool oxide A (63-71% for Yingde; 50-58%

for Dianhong), (2S, 5R)-linalool oxide B (74-84% for Yingde; 62-80% for Dianhong)

and (2S, 5S)-linalool oxide D (63-78% for Yingde; 59-72% for Dianhong) in the two

teas were similar and much lower than those in other Chinese black teas (linalool

oxide A: 77-96%; linalool oxide B: 94-99%; linalool oxide D: >99%). These

experimental phenomena clearly demonstrate the specialization and high similarity of

Yingde and Dianhong black teas, which was worthy of in-depth investigation. It was

initially speculated that the common large leaf origins of Yingde and Dianhong black

teas might be responsible for the large difference between them and other Chinese

black teas [31]. As for other worldwide famous black teas (Darjeeling, Ceylon and

Ntri Parc black teas), various enantiomeric distributions were observed probably due

to the special processing from traditional processing technology, climatic and cultivar

differences, especially nearly equal proportions of R- and S-linalool were presented

23
and different from those of other teas. Additionally, equal proportions of the two

enantiomers of linalool were detected in an authentic Earl Grey tea which were

fortified with bergamot oil, while extremely high ratios of R-linalool (90%) were

detected in the Earl Grey tea which contained excess food flavors (Figure 5B); and it

was different from other black teas that linalool oxides A-D were not detected or with

trace contents in the two Earl Grey teas, whereas the highest proportions of

R-limonene (88-94%) in them were presented.

Six typical oolong teas with different aroma characteristics were investigate, and

(2S, 5R)-linalool oxide C, α-ionone (the absolute configuration of peak 2), theaspirane

A (the absolute configuration of peak 2) and B (the absolute configuration of peak 2)

existed in oolong teas with single configurations (Figure 5F, Table S1). Approximate

proportions of S-limonene (54-63%), S-linalool (73-86%), R-4-terpineol (54-56%)

and (2S, 5S)-linalool oxide D (87->99%) among the oolong teas were detected;

beyond that, diverse enantiomeric distributions of other tested terpenoids in oolong

teas were observed, notably the major configuration of linalool oxide A ((2R, 5R),

34:66 ER) and α-terpineol (racemic, 50:50 ER) in Lingtou Dancong were contrary to

those of other oolong teas. In addition, the major configuration of linalool oxide B

among different oolong teas was consistent and with 67:33 to >99:1 ER values.

The enantiomeric distributions of terpenoids showed a high similarity among

Chinese white teas (Shoumei, Baimudan and Baihaoyinzhen), whereas great contrast

between Chinese and Indonesia white teas was observed. As shown in Figure 5, (2S,

5S)-linalool oxide A, (2S, 5R)-linalool oxide C and (2S, 5S)-linalool oxide D,

24
α-ionone (the absolute configuration of peak 2) and theaspirane B (the absolute

configuration of peak 2) were distributed in Chinese white tea samples with single and

consistent configurations. Notably, it was found that the proportions of S-limonene

(55-74%), S-linalool (97-99%), (2S, 5R)-linalool oxide B (98->99%), R-4-terpineol

(56-67%) and S-α-terpineol (78-83%) were generally higher than those of other types

of teas, which provides a preliminary theoretical basis for the discrimination of white

tea species. In addition, slight discrepancies were also observed in the enantiomeric

distribution of some terpenoids among Chinese white teas: the two enantiomers of

4-terpineol were unable to be detected in Shoumei, whereas the two enantiomers of

theaspirane A (40:60 ER) was only observed in Shoumei; and theaspirane B could be

detected in most Chinese white teas except Baihao Yinzhen samples. The standards of

plucking from tea plants (Baihao Yinzhen: single bud; Baimudan: two leaves and a

bud; Shoumei: two or three leaves without buds) might give rise to the above

difference. Notably, the enantiomeric distribution of terpenoids in Silver Needle was

unique; R-limonene (61-72%), R-linalool (95-98%), (2R, 5R)-Linalool oxide A

(54-69%) and R-α-terpineol (70-77%) presented high proportions in the

corresponding terpenoid enantiomers, which were reversed from the results of

Chinese white teas and most other tested teas. Moreover, in light of the lower odor

threshold (0.8 μg/L) and different odor characteristics (a woody and lavender-like

smell) of R-linalool [5] compared with S-linalool, the remarkable concentrations of

R-linalool (8976-17417 ng/g) probably played an important role in the formation of

the unique and strong fragrances of Indonesia white teas. Additionally, the ERs of

25
linalool oxides B and D showed great differences between Indonesia white teas

(47:53-75:25, 45:55-65:35) and Chinese white teas (98:2->99:1, >99:1). And the

above specific distributions could possibly provide a new idea for the authenticity

assessment of Indonesia white teas.

Compared with the abundant species of terpenoids in other fermented teas, some

common terpenoids were not detected in the dark teas, including limonene,

4-terpineol, theaspiranes A and B. R-Linalool was the major configurations of linalool

in dark teas, however, a wide range of proportions (43-70%) were observed although

the same type of dark tea was applied in the study. It was speculated that linalool

probably decomposed in different extents by the microbial metabolisms during

post-fermentation step of processing and even gradually volatilized after different

storage times [32], and the much higher concentrations of linalool oxide A and B

which could be generated by the oxidation of linalool in dark teas might support this

speculation. In addition, other terpenoids presented similar enantiomeric distribution

with 62-67% (2S, 5S)-linalool oxides A, 70-75% (2S, 5R)-linalool oxide B and

58-66% R-α-terpineol, and the extremely high proportions of (2R, 5S)-linalool oxides

C (>99:1 ER), (2S, 5S)-linalool oxide D (>99:1 ER) and α-ionone (peak 2, >99:1 ER)

in all dark tea samples.

3.6 The comparison of enantiomeric distributions and concentrations of

terpenoids among the five types of teas

Both similarities and differences of enantiomeric distributions of terpenoids were

observed among the five types of teas. Linalool oxide B (2S, 5R), linalool oxide D (2S,

26
5S) and α-ionone (peak 2) presented consistent major configurations in all tea types.

Notably, although the stereo configurations of α-ionone were unable to be identified

by commercially available standards, it might provide a reliable reference for our

identification from past work indicating that R-α-ionone was a natural constituent of

black teas through correlation with manool [33]. Therefore, the single existence of

peak 2 of α-ionone in all of black teas was considered as R-α-ionone. S-Limonene, (2S,

5S)-linalool oxide A and the peak 2 of theaspirane A were the major configurations in

most teas; while high proportions of their reverse configurations also existed in a

small number of teas, such as Yingde, Dianhong and Earl Grey black teas, Lingtou

Dancong oolong tea, and Indonesia white teas et al. Large differences were observed

in the enantiomeric distributions of linalool, α-terpineol and linalool oxide C;

S-linalool was the major configuration of linalool in all the green, oolong and Chinese

white teas and most black teas with different ERs, and Chinese white teas showed the

highest proportions. S-α-Terpineol was the major configuration in most oolong and

white teas; while conversely, R-α-terpineol was the major configuration in Indonesia

white teas and most black and dark teas; >99% proportions of (2S, 5R)-Linalool oxide

C was detected in black, oolong and white teas, while another enantiomers of linalool

oxide C existed only in Matcha and dark teas.

Subsequently, the concentration differences of terpenoid enantiomers among

different teas were studied. In view of the great disparities of ERs and concentrations

displayed between large and small leaf origins of black teas, PLS-DA [34] was

applied to find out the detailed differences. As shown in Figure 6A, a significant

27
discrimination between large and small leaf origins of black teas was observed by

using a PLS-DA model, and the well-explained variance (R2Y = 0.984) and

cross-validated predictive capability (Q2 = 0.916) indicated the model’s good

feasibility. Next, the cross-validation experiment (200 repeats of calculations) used to

measure the over-fitting phenomenon was performed, and the sufficiently low

intercepts (R2= 0.383, Q2= −0.389) demonstrated in the validation model strongly

proved the high dependability of the PLS-DA model (Figure 6B). Furthermore, six

compounds with variable importance in the projection (VIP) [35] values larger than

1.0 were identified as the components with significant differences between the two

types of black teas, including (2S, 5R)-linalool oxide C (1.77), (2S, 5R)-linalool oxide

B (1.71), (2S, 5S)-linalool oxide D (1.68), (2S, 5S)-linalool oxide A (1.61), S-linalool

(1.13) and R-linalool (1.09). To clarify the specific contribution of the compounds,

HCA [36] based on the concentrations of the six compounds in the black teas was

performed. As shown in Figure 6E, the concentration distribution of the compounds

presented an obvious tendency: the concentrations of the four linalool oxides A-D

enantiomers and S-linalool have significantly higher levels in the black teas with

small leaf origins, and conversely, the concentration of R-linalool in the Chinese black

teas with large leaf origins (Yingde and Dianhong black teas) were significantly

higher than those in other teas. Besides, the above compounds showed the least

amounts in the Ntri Parc and Ceylon teas, which indicating the obvious differences

exist among the black teas with different producing areas. Similarly, the concentration

differences of terpenoid enantiomers among different types of teas were distinguished

28
by PLS-DA. However, the six oolong tea samples were unable to be clustered

probably due to the large differences of cultivars and processing steps, and thus they

needed in-depth study in the future based on a large amount of related samples. A well

PLS-DA model (Figure 6C, R2Y = 0.874, Q2 = 0.813; Figure 6D, R2 = 0.362, Q2 =

-0.442) was established to differentiate the concentration differences of terpenoid

enantiomers among the Chinese green, white and dark teas. The HCA result (Figure

6F) indicated that (2S, 5S)-linalool oxide A and (2S, 5R)-linalool oxide B presented

the highest concentration levels in white teas, and the high concentrations of the

opposite configurations of linalool oxides A-B and low concentration of (2R,

5S)-Linalool oxide C were only detected in dark teas. Besides, all the above terpenoid

enantiomers in the green teas showed the lowest concentration levels.

4. Conclusions

In conclusion, an efficient HS-SPME/chiral GC-MS method was established to

analyze up to 12 pairs of volatile terpenoid enantiomers in teas for the first time, and

the enantiomeric and quantitative analysis of terpenoids in different types of teas were

carried out. The sufficiently low LOD, LOQ and RSD values, great linearity and wide

concentration ranges of each terpenoid enantiomer demonstrated the ideal

repeatability and reliability of the analysis approach. The analysis revealed that the 12

terpenoids were detected in teas with different enantiomeric distributions and

concentrations. α-Pinene was only detected in Rizhao green tea with an equal

proportions of enantiomers and trace concentration; S-limonene, S-linalool, (2S,

5S)-linalool oxide A (2S, 5R)-linalool oxide B, R-4-terpineol, (2S, 5R)-linalool oxide

29
C, (2S, 5S)-linalool oxide D, S-α-terpineol, R-α-ionone, peak 1 of theaspirane A and

peak 2 of theaspirane B were the major existences of the corresponding terpenoids in

most teas. Moreover, the enantiomeric distributions of volatile terpenoids showed

high similarities among Chinese white teas, and their ER values were frequently

higher than those in other types of teas. In the contrast, Yingde and Dianhong black

teas and the teas from outside of China, especially Indonesia white teas, tended to

exhibit the reverse enantiomeric distributions of terpenoids. Furthermore, large and

small leaf origins of black teas were clustered respectively based on the

concentrations of terpenoid enantiomers, and (2S, 5R)-linalool oxide C, (2S,

5R)-linalool oxide B, (2S, 5S)-linalool oxide D, (2S, 5S)-linalool oxide A, S-linalool

and R-linalool showed significant differences between the two origins of black teas.

Moreover, the PLS and HCA results indicated that both enantiomers of linalool oxides

A-B and (2R, 5S)-linalool oxide C showed significant differences among Chinese

green, white and dark teas. The breakthrough progress of the enantiomeric and

quantitative analysis in this work improves the understanding and scientific evaluation

of tea aroma and provides a preliminary theoretical basis for the discriminations of tea

origins.

Conflict of interest

The authors have declared no conflict of interest.

Acknowledgments

This work was supported by the Natural Science Foundation of Zhejiang Province

30
(LY14C160002, LQ15C160007), the National Natural Science Foundation of China

(31470694), the Special Funds for the Construction of Modern Agricultural

Technology Systems (CARS-23), and the Science and Technology Innovation Project

of the Chinese Academy of Agricultural Sciences (CAAS-ASTIP-2014-TRICAAS).

References

[1] X. C. Wan, Tea Biochemistry, Beijing: China Agriculture Press, (2003) 39. (in Chinese)

[2] Z. Yang, S. Baldermann, N. Watanabe, Recent studies of the volatile compound in tea, Food

Res. Int. 53 (2013) 585–599.

[3] C.T. Ho, X. Zheng, S.M. Li, Tea aroma formation, Food Sci. Hum. Wellness 4 (2015) 9–27.

[4] C. Bicchi, A. D'Amato, P. Rubiolo, Cyclodextrin derivatives as chiral selectors for direct gas

chromatographic separation of enantiomers in the essential oil, aroma and flavour fields, J.

Chromatogr. A 843 (1999) 99–121.

[5] A. Padrayuttawat, T. Yoshizawa, H. Tamura, T. Tokunaga, Optical isomers and odor thresholds

of volatile constituents in Citrus sudachi, Food Sci. Technol. Int. 3 (1997) 402–408.

[6] K. Kuroda, N. Inoue, Y. Ito, K. Kubota, A. Sugimoto, T. Kakuda, T. Fushiki, Sedative effects

of the jasmine tea odor and (R)-(−)-linalool, one of its majorodor components, on autonomic

nerve activity and mood states, Eur. J. Appl. Physiol. 95 (2005) 107–114.

[7] G. Ohloff, E. Klein, Die absolute konfiguration des linalools durch verknüpfung mit dem

pinansystem, Tetrahedron 18 (1962) 37–42.

[8] K. Takoi, K. Koie, Y. Itoga, Biotransformation of hopderived monoterpene alcohols by lager

yeast and their contribution to the flavor of hopped beer, J. Agric. Food Chem. 58 (2010)

31
 5050–5058.

[9] F. Quattrini, G. Biressi, M. Juza, M. Mazzotti, C. Fuganti, M. Morbidelli, Enantiomer

separation of α-ionone using gas chromatography with cyclodextrin derivatives as chiral

stationary phases, J. Chromatogr. A 865 (1999) 201–210.

[10] V. Schurig, M. Schleimer, M. Jung, S. Mayer, A. Glausch, P. Schreier, Ed, P. Winterhalter,

Enantiomer separation by GLC, SFC and CZE on high-resolution capillary columns coated

with cyclodextrin derivatives, Progress in Flavour Precursor Studies, Allured Publ. Corp.

(1993) 63–75.

[11] M. Steinhaus, H. T. Frisch, P. Schieberle, Quantitation of (R)-and (S)-linaloolin beer using

solid phase microextraction (SPME) in combination with a stableisotope dilution assay

(SIDA), J. Agric. Food Chem. 51 (2003) 7100–7105.

[12] B. Bonnländer, R. Cappuccio, F. S. Liverani, P. Winterhalter, Analysis of enantiomeric

linalool ratio in green and roasted coffee, Flavour Frag. J. 21 (2006) 637–641.

[13] F. Luan, A. Mosandl, M. Gubesch, M. Wüst, Enantioselective analysis of monoterpenes in

different grape varieties during berry ripening using stir bar sorptive extraction- and solid

phase extraction-enantioselective-multidimensional gas chromatography-mass spectrometry,

J. Chromatogr. A. 1112 (2006) 369–374.

[14] I. Špánik, A. Pažitná, B. Šiška, P. Szolcsányi, The determination of botanical origin of honeys

based on enantiomer distribution of chiral volatile organic compounds, Food Chem. 158

(2014) 497–503.

[15] A. Mosandl, Authenticity Assessment: A permanent challenge in food flavor and essential oil

analysis, J. Chromatogr. Sci. 42 (2004) 440–449.

32
[16] M. Winter, E. Sundt, Recherches sur les arǒmes. 5e communication. Analyse de I'arǒme des

framboises. I. Les constituants carbonylés volatils, Helv. Chim. Acta 45 (1962) 2195‒2211.

[17] G. Uhde, G. Ohloff, Parmon, eine phantomverbindung im veilchenblütenöl, Helv. Chim. Acta

55 (1972) 2621‒2625.

[18] R. R. Linko, H. Kallio, T. Pyysalo, K. Rainio, K. Volatile monocarbonyl compounds of carrot

roots at various stages of maturity, Z. Lebensm. Unters. Forsch. 166 (1978) 208‒211.

[19] H. G. Shutava, N. A. Kavalenka, H. N. Supichenka, V. N. Leontiev, T. G. Shutava, Essential

oils of lamiaceae with high content of α-,β-pinene and limonene enantiomers, TEOP 17 (2014)

18–25.

[20] S. Chin, Y. Nolvachai, P. J. Marriott, Enantiomeric separation in comprehensive

two-dimensional gas chromatography with accurate mass analysis, Chirality 26 (2014) 747–

753.

[21] I. Bonaccorsi, D. Sciarrone, A. Cotroneo, L. Mondello, P. Dugo, G. Dugo, Enantiomeric

distribution of key volatile components in Citrus essential oils, Rev. Bras. Farmacogn. 21

(2011) 841–849.

[22] M. Junge, S. Bieri, H. Huegel, P. J. Marriott, Fast comprehensive two-dimensional gas

chromatography with cryogenic modulation, Anal. Chem. 78 (2007) 4448–4454.

[23] N. W. Davies, T. Larkman, P. J. Marriott, I. A. Khan, Determination of enantiomeric

distribution of terpenes for quality assessment of Australian tea tree oil, J. Agric. Food Chem.

64 (2016) 4817−4819.

[24] C. Cagliero, C. Bicchi, C. Cordero, P. Rubiolo, B. Sgorbini, E. Liberto, Fast

headspace-enantioselective GC-mass spectrometric-multivariate statistical method for routine

33
 authentication of flavoured fruit foods, Food Chem. 132 (2012) 1071−1079.

[25] D. Wang, K. Ando, K. Morita, K. Kubota, A. Kobayashi, Optical isomers of linalool and

linalool oxides in tea aroma, Biosci. Biotech. Biochem. 58 (1994) 2050–2053.

[26] T. Yang, Y. Zhu, C. Y. Shao, Y. Zhang, J. Shi, H. P. Lv, Z. Lin, Enantiomeric analysis of

linalool in teas using headspace solid-phase microextraction with chiral gas chromatography,

Ind. Crop. Prod. 83 (2016) 17–23.

[27] L. P. Du, C. Wang, J. X. Li, D. G. Xiao, C. W. Li, Y. Q. Xu, Optimization of headspace

solid-phase microextraction coupled with gas chromatography–massspectrometry for

detecting methoxyphenolic compounds in Pu-erh tea, J. Agric. Food Chem. 61 (2013) 561–

568.

[28] V. Schurig, Terms for the quantitation of a mixture of stereoisomers, Top. Curr. Chem. 340

(2013) 21-40.

[29] V. Schurig, H. -P. Nowotny, Separation of enantiomers on diluted permethylated

ß-cyclodextrin by high-resolution gas chromatography, J. Chromatogr. 441 (1988) 155–163.

[30] V. Schurig, M. Schleimer, H. -P. Nowotny, Enantiomer analysis of α-pinene and limonene,

Naturwissenschaften 77 (1990) 133–134.

[31] Y. J. Yang, Y. R. Yang, Clonal tea varieties of China, Shanghai: Shanghai Scientific and

Technical Publishers (2014) 157, 270. (in Chinese)

[32] H. P. Lv, Y. J. Zhang, Z. Lin, Y. R. Liang. Processing and chemical constituents of Pu-erh tea:

A review, Food Res. Int. 53 (2013) 608–618.

[33] J. Bricout, R. Viani, F. Müggler-Chavan, J. P. Marion, D. Reymond, R. H. Egli, Sur la

composition de I'arôme de thé noir II. Helv. Chim. Acta 50 (1967) 1517‒1522.

34
[34] P. S. Gromski, H. Muhamadali, D. I. Howbeer, A tutorial review: Metabolomics and partial

least squares-discriminant analysis-a marriage of convenience or a shotgun wedding, Anal.

Chim. Acta 879 (2015) 10–23.

[35] S. Wold, M. Sjӧstrӧm, L. Eriksson, PLS-regression: a basic tool of chemometrics, Chemometr.

Intell. Lab. Syst. 58 (2001) 109–130.

[36] J. P. Peeters, J. A. Martinelli, Hierarchical cluster analysis as a tool to manage variation in

germplasm collections, Theor. Appl. Genet. 78 (1989) 42–48.

35
Figure captions

Figure 1. The familiar volatile terpenoids with high contents in teas

Figure 2. The characteristic ions and GC-MS peaks of the separated terpenoid

enantiomer standards (the absolute configurations of linalool oxide (A-D), α-ionone

and theaspirane (A and B) are not marked due to a lack of the corresponding

standards)

Figure 3. A) The total ion chromatograms (TICs) of aroma components in the mixed

tea samples; B-H) the identified terpenoid enantiomers by extracting the characteristic

ions (S-limonene, R-linalool, S-linalool, linalool oxides A-D, S-4-terpineol,

R-4-terpineol, S-α-terpineol, R-α-terpineol and α-ionone)

Figure 4. The comparisons of summation of peak areas of separated terpenoid

enantiomers by using A) different extraction fibers; B) different proportions between

tea and boiling water; C) different extraction temperatures; D) different extraction

times

Figure 5. The enantiomeric ratios of terpenoids in different teas A) Limonene; B)

Linalool; C) Linalool oxide A; D) Linalool oxide B; E); 4-Teroineol; F) Linalool

oxide C; G) Linalool oxide D; H) α-Terpineol

Figure 6. A) The PLS-DA model used to distinguish the concentration differences of

terpenoid enantiomers between large (Class 1) and small (Class 2) leaf origins of

black teas; B) the validation model used to examine the over-fitting of the PLS-DA

model of black teas; C) The PLS-DA model used to distinguish the concentration

differences of terpenoid enantiomers among Chinese green (Class 1), white (Class 2)

36
and dark teas (Class 3); D) the validation model used to examine the over-fitting of

the PLS-DA model of green, white and dark teas; E) the heat map of the key terpenoid

enantiomers in the two origins of black teas; F) the heat map of the key terpenoid

enantiomers among Chinese green, white and dark teas

37
 HO HO HO  O  HO  O  O O
  
OH  OH
R-Linalool S-Linalool Linalool oxide (A, B, C, D)


HO HO
H H
S-Citronellol R-Citronellol OH
E-Nerolidol

OH
O S--Pinene R--Pinene 
O
S-Citronellal R-Citronellal Z-Nerolidol

O O
 O  O
 

S--Ionone R--Ionone S-Limonene R-Limonene Theaspirane A Theaspirane B

OH OH

OH OH
R-Terpinen-4-ol S-Terpinen-4-ol R-Phellandrene S-Phellandrene S--Terpineol R--Terpineol

Figure 1.

Figure 2.

38
Figure 3.

39
Figure 4.

40
41
Figure 5.

42
Figure 6.

43
Table 1. The tested tea samples used in the study
No Name Source Classification Comments
1 Matcha-1 Japan Green tea -
2 Matcha-2 Japan Green tea -
3 Lu’an Guapian Anhui province of China Green tea -
4 Xinyang Maojian Henan province of China Green tea -
5 Rizhao Shandong province of China Green tea -
6 Maoshan Bai Jiangsu province of China Green tea -
7 Biluochun Jiangsu province of China Green tea -
8 Yingde-1 Guangdong province of China Black tea -
9 Yingde-2 Guangdong province of China Black tea -
10 Yingde-3 Guangdong province of China Black tea -
11 Yingde-4 Guangdong province of China Black tea -
12 Keemun Anhui province of China Black tea
13 Dianhong -1 Yunnan province of China Black tea -
14 Dianhong -2 Yunnan province of China Black tea -
15 Dianhong -3 Yunnan province of China Black tea -
16 Dianhong -4 Yunnan province of China Black tea -
17 Jinjunmei Fujian province of China Black tea -
18 Jiuqu Hongmei Zhejiang province of China Black tea -
19 Ntri Parc Pakistan Black tea Broken tea
20 Ceylon-1 Sri lanka Black tea Broken tea
21 Ceylon-2 Sri lanka Black tea Broken tea
22 Ceylon-uva Uva, Sri lanka Black tea -
23 Darjeeling-1 Darjeeling, India Black tea -
24 Darjeeling-2 Darjeeling, India Black tea -
25 Earl Grey-1 Sri lanka Black tea Authentic
26 Earl Grey-2 India Black tea Excess food flavors have been
added
27 Wuye Dancong Guangdong province of China Oolong tea -
28 Lingtou Dancong Guangdong province of China Oolong tea -
29 Wenshan Baozhong Taiwan province of China Oolong tea -
30 Dongding Taiwan province of China Oolong tea -
31 Fenghuang Shuixian Guangdong province of China Oolong tea -
32 Tieguanyin Fujian province of China Oolong tea -
33 Shoumei Fujian province of China White tea -
34 Baimudan-1 Fujian province of China White tea -
35 Baimudan-2 Fujian province of China White tea -
36 Baihao Yinzhen-1 Fujian province of China White tea -
37 Baihao Yinzhen-2 Fujian province of China White tea -
38 Silver Needle-1 Indonesia White tea Processed by tea buds
39 Silver Needle-2 Indonesia White tea Processed by tea buds
40 Gray Dragon Indonesia White tea Processed by tea buds and leaves
41 Pu-erh-1 Yunnan province of China Dark tea Loose tea

44
42 Pu-erh-2 Yunnan province of China Dark tea Loose tea
43 Pu-erh-3 Yunnan province of China Dark tea Loose tea
44 Pu-erh-4 Yunnan province of China Dark tea Loose tea
45 Pu-erh-5 Yunnan province of China Dark tea Tea cake

45
Table 2. The comparison of separating capacity of different chiral capillary columns
Name B-PM[1] B-DA[2] G-TA[3] 110[4] 120[5] CP[6] Dex-CB[7] 20β[8] G-DM[9]
The separation of terpenoid enantiomers
Linalool √ × × √ × √ × Partly[10] ×
Linalool oxide A √ × × √ × √ × × ×
Linalool oxide B Partly × × √ × √ × × ×
Linalool oxide C √ × × √ × × × × ×
Linalool oxide D √ × × √ × × × × ×
4-Terpineol √ × × √ × √ √ √ ×
α-Terpineol √ × × √ × √ × √ ×
Citronellol × × × × × × × × ×
Citronellal × × × × × × × × ×
Z-nerolidol × × × × × × × × ×
E-nerolidol × × × × × × × × ×
α-Ionone √ √ √ √ √ √ √ √ √
α-Pinene √ × × √ √ √ √ √ ×
Limonene √ × × × × × √ × ×
α-Phellandrene × × × × × × × × ×
Theaspiroane A √ × × × × × √ × ×
Theaspiroane B √ × × × × × √ × ×
[1] Astec CHIRAL DEX B-PM, using 2,3,6-tri-O-methyl derivative of β-cyclodextrin as the chiral stationary phase;
[2] Astec CHIRAL DEX B-DA, using 2,6-di-pentyl-O-methyl derivative of β-cyclodextrin as the chiral stationary
phase; [3] Astec CHIRAL DEX G-TA, using 2,6-di-O-pentyl-3-trifluoroacetyl derivative of γ-cyclodextrin phase; [4]
β-DEX 110, using 10% full methylation of β-cyclodextrin as the chiral stationary phase; [5] β-DEX 120, using 20%
full methylation of β-cyclodextrin as the chiral stationary phase; [6] CP-Cyclodextrin-B-236-M-19, using 30%
7-(2,3-di-O-methyl-6-O-t-butyldimethylsilyl)-β-cyclodextrin as the chiral stationary phase; [7] CHIRASIL-DEX
CB, using dimethyl siloxane-β-cyclodextrin as the chiral stationary phase; [8] HP-Chiral 20β, using 35% phenyl
methylpolysilicone-β-cyclodextrin (20%) as the chiral stationary phase; [9] Astec CHIRAL DEX G-DM, using
2,3-di-O-methyl-6-t-butyl silyl derivative of γ-cyclodextrin phase; [10] The enantiomers of the terpenoid were partly
separated.

46
Table 3. Retention times, concentration ranges, correlation coefficients (R2), limits of detection
(LOD), limits of quantitation (LOQ), relative standard deviations (RSDs, n=5) and the
characteristic ions of the terpenoid enantiomers[1]
Peak Characteristic ions[2] Compounds RT[3] (min) Linear range LOD[4] LOQ[5] R2 RSD

(ng/g) ( ng/g) (ng/g) (%)

S: 10.27
1-2 53, 77, (93), 105, 136 α-Pinene 1.59-9016 1.59 7.94 0.9962 1.08
R: 10.49

S: 14.16
3-4 53, (68), 79, 93, 136 Limonene 0.80-16340 0.40 1.05 0.9930 1.49
R: 14.39

R: 24.64
5-6 55, (71), 93, 121 Linalool 0.86-20410 0.86 1.77 0.9979 2.45
S: 24.85

(2R, 5R): 25.45

Linalool oxide A[6]
(2S, 5S): 26.35
7-10 (59), 68, 94, 81, 137 0.98-4827 0.98 4.90 0.9996 3.25
(2R, 5S): 26.68
Linalool oxide B[6]
(2S, 5R): 27.12

(71), 93, 111, 136, S: 29.90

11-12 4-Terpineol 0.63-1289 0.32 1.21 0.9985 2.72
154 R: 30.17

(2S, 5R): 30.69

Linalool oxide C
59, (68), 94, 109, 137, (2R, 5S): 31.06
13-16 1.21-1214 1.21 3.00 0.9978 1.04
155 (2S, 5S): 31.49
Linalool oxide D
(2R, 5R): 32.03

S: 32.60
17-18 (59), 93, 121, 81, 136 α-Terpineol 1.04-2003 1.04 3.65 0.9989 3.85
R: 32.87

Peak 1: 40.92
19-20 93, (121), 136, 192 α-Ionone 0.90-1834 0.90 2.60 0.9999 2.72
Peak 2: 41.50

Peak 1: 29.45
Theaspirane A
Peak 2: 30.53
21-24 55, 82, 96, 109, (138) 1.06-10260 0.11 0.53 0.9975 3.98
Peak 1: 31.47
Theaspirane B
Peak 2: 32.11
[1] Unless otherwise noted, the chiral separations of terpenoids were carried out by using Astec CHIRAL DEX B-PM as the chiral

capillary column; [2] the two enantiomers of a certain terpenoids or the four isomers of the terpenoids with two stereogenic centers (such

as linalool oxides A and B, linalool oxides C and D and theaspiranes A and B) were identified by the same characteristic ions, and the

characteristic ions in the parentheses were used as the quantitative ions of the corresponding terpenoids; [3] RT: retention time; [4] limits of

detection; [5] limits of quantitation; [6] the chiral separations of terpenoid enantiomers were carried out by using β-DEX 110 as the chiral

capillary column.

47