Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Dec 24.
Published in final edited form as: Food Chem. 2011 Jun 1;126(3):1269–1277. doi: 10.1016/j.foodchem.2010.11.055

Tentative identification, quantitation, and principal component analysis of green pu-erh, green, and white teas using UPLC/DAD/MS

Yang Zhao a,b,c, Pei Chen a,*, Longze Lin a, JM Harnly a, Liangli (Lucy) Yu c, Zhangwan Li b
PMCID: PMC4276396  NIHMSID: NIHMS504183  PMID: 25544798

Abstract

Tea (Camellia sinensis L.), an important drink and a natural medicine for thousands of years, contains many health beneficial compounds. Growing season, geographical region, and fermentation methods create many variations in tea compositions, which contribute to each tea's uniqueness. In this study, a simple, rapid, and efficient ultra-performance liquid chromatography (UPLC) method combined with diode array detector (DAD) and mass spectroscopic (MS) detection and chemometrics analysis was used to analyse three different types of teas (green pu-erh, green tea, white tea). Using the developed method, 68 compounds were identified and 54 were quantified based on retention times, UV spectra, and MS spectra by referencing to available standards and data in the literatures. The results showed the chemical differences between the tested teas. Principal component analysis (PCA) was applied to classify and distinguish between tea samples.

Keywords: Camellia sinensis, Tentative identification, Quantification, UPLC/DAD/MS, PCA

1. Introduction

Tea (Camellia sinensis L.) has been used as an important drink for over 1000 years because of its health benefits for the human body. It is one of the most widely consumed beverages in the world, next to water (Alan Crozier, Jaganath, & Clifford, 2009; Dufresne & Farnworth, 2001; Editorial committee of encyclopedia of Chinese medicinal plant-sEds, 1999). Tea is cultivated in more than 30 countries. The growing seasons, geographical regions, processing, and fermentation methods create many varieties that contribute to each tea's uniqueness. Teas are commercially distinguished into three major categories according to the degree of fermentation: non-fermented (green, white, and green pu-erh teas), partially fermented (oolong and paochong teas), and fully fermented (black and pu-erh teas), although other classifications have been proposed (Alcazar et al., 2007; Fernandez, Martin, Gonzalez, & Pablos, 2000; Lin, Lin, Liang, Lin-Shiau, & Juan, 1998). In the fermentation process, the enzymatic oxidation of polyphenols takes place, leading to the formation of theaflavins and thearubigins, which are responsible for their characteristic aroma and colour (Palmer, 1984). In general, green and oolong teas are consumed mainly in Asia and Northern Africa, white tea in Asia and Europe, pu-erh tea in Asia, and black tea worldwide (Wu et al., 2007).

Taxonomically,C. sinensis (L.) O. Kuntze has two subspecies, C. sinensis var. sinensis (China tea) and C. sinensis var. assamica (Masters) Kitamura (Assam tea or pu-erh tea in China). There are two definitions for white teas. The traditional Chinese definition of white tea belongs only to “khenghe bai hao” and “fudin bai hao”, found only in Fujian province of China. The definition in other countries of white tea is less sophisticated, it could be any tea made from newly grown buds and young leaves of the tea plants. The buds may be shielded from sunlight during growth to reduce the formation of chlorophyll, giving the young leaves a white appearance (Hilal & Engelhardt, 2007).

Many studies have been reported on the determination of chemical components of teas, especially on tea phenolics. It has been shown that teas contain purine (xanthine) alkaloids, phenolic compounds (catechins, O-glycosylated flavonols, C-glycosylated flavones, proanthocyanidins, and phenolic acids and their derivatives), terpenoids, fatty acids, essential oils, and amino acids. Oolong and black teas also contain the fermented oxidation products of catechins, theaflavins, and polymeric thearubigins (Alan et al., 2009; Chiu & Lin, 2005; Clifford, Stoupi, & Kuhnert, 2007; Del Rio et al., 2004; Dufresne et al., 2001; Editorial committee of encyclopedia of Chinese medicinal plant-sEds., 1999; Ito, Sugimoto, Kakuda, & Kubota, 2002; Kato & Shibamoto, 2001; Lakenbrink, Engelhardt, & Wray, 1999; Miketova et al., 2000; Naef, Jaquier, Velluz, & Maurer, 2006; Zeeb, Nelson, Albert, & Dalluge, 2000; Zhou, Zhang, Xu, & Yang, 2005; Zhu et al. 2004). Biological studies have demonstrated the health benefits from tea phenolics, especially epigallocatechin gallate (EGCG), the major catechin constituting over 10% of the dry weight of green teas (Alan et al., 2009; Dufresne et al., 2001; Editorial committee of encyclopedia of Chinese medicinal plant-sEds., 1999; Friedman, 2007; Friedman et al., 2007; Jie, Lin, Zhang, Lv., He, & Zhao, 2006; Lambert & Yang, 2003).

Modern analytical methods are also used for tea classification and quality control based on the chemical components of teas. Reported methods include gas chromatography mass spectrometry (GC/MS, Pongsuwan, Fukusaki, Bamba, Yonetani, Yamahara, & Kobayashi, 2007) and liquid chromatography mass spectrometry (LC/MS) for characterisation of teas (Pongsuwan, Bamba, Harada, Yonetani, Kobayashi, & Fukusaki, 2008; Xie et al., 2009), capillary electrophoresis (CE) analysis of major amino acids in tea leaves and beverages (Hsieh & Chen, 2007), near infrared spectroscopy (NIR) identification of green, black, and oolong teas (Zhao, Chen, Huang, & Fang, 2006), NMR measurements (Fujiwara, Ando, & Arifuku, 2006), and HPLC with UV detection (Fernandez et al., 2000; Zheng, Watson, Johnston, Clark, Edrada-Ebel, & Elseheri, 2009). Some of these methods were coupled with chemometrics for studies of the chemical compounds and chromatograms of tea extracts.

In a previous study, we used a standardised profiling method based on HPLC/DAD/MS to examine 41 green teas and 25 fermented teas (Lin, Chen, & Harnly, 2008). However, the differences between green pu-erh teas, green teas, and white teas were not investigated. As part of our continuous effort in botanical research, a newly developed UPLC/DAD/MS method was used for tentative identification and quantification of the major constituents in green pu-erh teas, green teas, and white teas. The method proved to be an effective alternative for the standardised profiling method used previously. Principal component analysis (PCA) was performed to categorise the tested teas.

2. Material and methods

2.1. Standard compounds and other chemicals

(+)-catechin (C), (−)-epicatechin (EC), (−)-epigallocatechin (EGC), (−)-epigallocatechin 3-O-gallate (EGCG), gallic acid (GA), caffeine (CAF), theobromine (TB), theophylline (TP), apigenin, p-coumaric acid, caffeic acid, chlorogenic acid, kaempferol, quercetin, and kaempferol 3-O-p-coumaroylglucoside were obtained from Sigma Chemicals Co. (St. Louis, MO). Quercetin 3-O-glucoside, kaempferol 3-O-glucoside, and myricetin were purchased from Extrasynthese (Genay, Cedex, France). HPLC grade methanol, acetonitrile and formic acid were purchased from VWR International, Inc. (Clarksburg, MD). HPLC grade water was prepared from distiled water using a Milli-Q system (Millipore Laboratory, Bedford, MA).

2.2. Tea samples

A total of 15 tea samples were obtained from various sources. The 5 green pu-erh teas and 5 white teas were purchased through internet. The 5 green teas were purchase at local stores in China. The detailed information for the tea samples are shown in Table 1. All the samples were finely powdered and passed through a 20-mesh sieve and then kept in desiccators at room temperature.

Table 1.

The detailed information of tested teas.

Label Category Information
GP1 Green pu-erh tea Holy mountain trading company, green Tibetan mushroom pu-erh
GP2 Green pu-erh tea Anjing, Sheng Tai Qing Raw Pu'er
GP3 Green pu-erh tea Ba Jiao Ting
GP4 Green pu-erh tea Bulang Shan
GP5 Green pu-erh tea Guang Yao Hao
GT1 Green tea Jiming Gong Tea, Sixth Grade, Chengkou, Chongqing City
GT2 Green tea Mingqian green tea, Sichuan Province
GT3 Green tea China Zhu Tea
GT4 Green tea Japanese green tea
GT5 Green tea Japanese Sencha
WT1 White tea Adagio tea, white leaf tea, silver needle
WT2 White tea Adagio tea, white leaf tea, snowbud
WT3 White tea Adagio tea, white leaf tea, white symphony
WT4 White tea Adagio teas, white leaf tea, white peony
WT5 White tea Ji'an white tea
Open in a new tab

2.3. Standard solutions and sample preparation

Approximately 3.0–5.0 mg of each standard was accurately weighed using a Mettler M3 Micro Balance and was then brought to a concentration of 1.0 mg/mL with methanol–water (60:40, v/v) to provide stock solutions. The stock solutions were mixed and diluted to provide a series of standard solutions at appropriate concentrations for the calibration curves. All the solutions were stored at 4 °C in refrigerator.

100 mg of each tea sample was sonicated (FS30 Ultrasonic sonicator, Fisher Scientific, Pittsburgh, PA) with 5.00 mL of methanol–water (60:40, v/v) for 60 min at room temperature. The slurry mixture was centrifuged at 5000 g for 15 min (IEC Clinical Centrifuge, Damon/IEC Division, Needham, MA). The supernatant was filtered through a 0.45 μm PVDF syringe filter (VWR Scientific, Seattle, WA).

2.4. UPLC/MS conditions

A Waters ACQUITY UPLC system (Waters, Milford, MA) equipped with a binary solvent delivery manager, a column manager, a sample manager, and a diode array detector (DAD) was used. The system was coupled with a Waters Micromass Quattro Micro™ mass spectrometer (Waters, Milford, MA). The instrument was fitted with an Acquity UPLC® HSS C18 column, 1.8 μm, 2.1 × 100 nm (Waters, Milford, MA). Chromatographic separation was carried out with mobile phase A (H2O containing 0.1% formic acid) and mobile phase B (acetonitrile containing 0.1% formic acid). The linear gradient program was as follows: 0–12 min, 5–26% B; 12–14 min, 26–65% B. The flow rate was 0.5 mL/min, the column temperature was 40 °C, the autosampler temperature was 10 °C, and the injection volume was 5 μL. The DAD was set at 270 and 348 nm for quantitation and UV spectra were recorded from 190–400 nm for tentative identification. Both positive and negative ion modes were used with the ESI-MS. Positive ion mode parameters were as follows: capillary voltage, 3 kV; cone voltage, 20 V; extractor voltage, 3 V; RF lens voltage, 0.3 V; the desolvation gas flow, 650 L/h; the desolvation temperature, 275 °C; and source temperature, 100 °C. Negative ion mode parameters were as follows: capillary voltage, 3.5 kV; cone voltage, 30 V; extractor voltage, 2 V; RF lens voltage, 0.2 V; desolvation gas flow, 700 L/h; desolvation temperature, 300 °C; and source temperature, 100 °C. Data were collected in centroid mode over a mass range of m/z 100–1100 with acquisition time of 400 ms and inter-scan delay of 100 ms.

2.5. Quantification of the chemical compounds

Since only 18 standards were available, GA was used as the molar equivalent to quantify its derivatives; EC for epicatechin 3-O-benzoate; C for all the tea catechins; EGCG for catechin gallates; caffeic acid for its derivatives; quercetin 3-O-glucoside for all the quercetin 3-O-glycosides; kaempferol 3-O-glucoside and kaempferol 3-O-p-coumaroylglucoside for all of its kaempferol 3-O-glycosides without and with p-coumaroyl, respectively. All samples were run in triplicate for both quantitation and multivariate statistical analysis.

2.6. Multivariate statistical analysis

In this study, 3 approaches were used for PCA analysis. The 1st one was to use all the aligned data points of a LC/DAD chromatogram, the 2nd one was to use the quantitative results, and the 3rd one was to use characteristic fingerprints (the relative peak areas of significant chromatographic peaks).

  1. All the data points from the LC/DAD chromatogram of each sample was exported from Masslynx and saved in *.csv format. Then all the data were combined into 1 matrix in Microsoft Excel. The matrix was imported to SpecAlign software (Version 2.4 Cartwright Group, PTCL, University of Oxford) for chromatogram alignment, after which, several thousands of data points were converted into decades of peak intensities.

  2. The contents of the 54 determined chemical compounds were used as input data, where there the first column consist of different tea samples and the remaining columns to the right consist of the contents of the compounds.

  3. This approach selected the chromatogram of an authentic sample of a certain botanical as a reference fingerprint (RF, in this study, WT1). The most obvious peak of the RF was selected as the reference peak (RP, peak at retention time 3.71 min in this case). The areas of all other peaks in the chromatograms were normalised against the area of the RP and the ratios of the peaks were entered into a peak table and used for PCA analysis. The detailed process has been discussed previously (Chen, Ozcan, & Harnly, 2007; Chen, Song, & Lin, 2009).

The different kinds of data were introduced into SIMCA-P 11.5 software package (Umetrics) AB, Umea°, Sweden) for the PCA analysis. Unit variance scaling (autoscaling) was performed before PCA scatter plots were obtained based on two first components.

3. Results and discussion

3.1. Optimisation of the UPLC method

Good separation was achieved after screening a series of mobile phases, UPLC columns, and gradient profiles (Fig. 1). Among the standards used, GA, TB, TP, EGC, C, CAF, EC, EGCG and p-coumaric acid were quantified at 270 nm, while chlorogenic acid, caffeic acid, quercetin 3-O-glucoside, kaempferol 3-O-glucoside, myricetin, quercetin, kaempferol 3-O-p-coumaroylglucoside, apigenin, and kaempferol were quantified at 348 nm. In the previous study (Lin et al., 2008), the HPLC method needed a long analysis time of 70 min, whereas the UPLC/DAD/MS method required only 14min. The combination of the shortened running time with a lower flow rate (0.5 mL/min vs 1.0 mL/min) reduced solvent consumption from 70 to 7 mL per analysis.

Fig. 1.

Fig. 1

Open in a new tab

UPLC/DAD chromatograms of mixed standards (total scan 200–400 nm).

3.2. Tentative identification of the chemical compounds

Representative UPLC chromatograms detected at 270 and 348 nm are shown in Fig. 2. The retention times (tR), wavelengths of maximum absorbance (λmax), deprotonated/protonated molecules ([M−H]−/[M + H]+), and major fragment ions (NI/PI) are listed in Table 2. Peak identification was based on analysis of the data in the table, comparison with standards, and comparison with data in previously published literatures. The identification process has been described in details previously (Lin et al., 2008). All the results are listed in Table 2.

Fig. 2.

Fig. 2

Open in a new tab

Representative UPLC chromatograms of tea samples: (a) green tea registered at 348 nm; (b) green pu-erh tea registered at 348 nm; (c) white tea registered at 348 nm; and (d) white tea registered at 270 nm.

Table 2.

Identified components in tested tea samples.

Peak No. tR (min) UVλmax (nm) [M−H]−/[M−H]+ (m/z) NI/PI aglycone, other ions (m/z) Identification
1 0.49 ND 173/175 −/− Theanine
2 0.71 254 173/175 −/− Unknown
3 0.84 273 331/333 169/− Galloylglucose
4A 1.04 273 343/345 169,191/151,173,187 3-Galloylquinic acid
4B 1.07 274 343/345 169,191/151,173,187 5-Galloylquinic acid
4C 1.08 274 169/171 −/− Gallic acid
5 1.24 273 343/345 169,191/173,187 4-Galloylquinic acid
6 1.37 264 164/166 113,147/120,149 Unknown
7A 1.69 273 305/307 −/− Gallocatechin
7B 1.74 273 179/181 −/− Theobromine
8 2.02 275 609/611 −/− Gallocatechin dimer
9 2.23 240,326 353/355 191,291/275 3-Caffeoylquinic acid
10 2.31 240,326 353/355 179,191/275 5-Caffeoylquinic acid
11 2.40 271 179/181 −/− Theophylline
12A 2.44 278 456/458 −/− Unknown
12B 2.49 272 456/458 −/− Unknown
13 2.66 273 287/289 −/195 Unknown
14 3.02 210,270 305/307 −/− (−)-epigallocatechin
15A 3.12 − 337/339 163,305/163,231 3-p-Coumaroylquinic acid
15B 3.16 310 337/339 163,305/163,209,231 5-p-Coumaroylquinic acid
16 3.26 274 183/185 169/− Gallic acid methyl ester
17 3.32 276 483/485 −/− Digalloylglucose
18 3.43 276 289/291 −/− (+)-Catechin
19A 3.49 235sh,300sh,326 353/− 191/291 Chlorogenic acid
19B 3.50 282 295/297 113,191/− Caffeoylmalic acid
20A 3.65 271 633/− Strictinin
20B 3.71 270 −/195 −/157 Caffeine
21 3.80 235sh,300sh,326 353/355 179,191/195 4-Caffeoylquinic acid
22 3.97 240sh,292sh,323 179/181 135/157 Caffeic acid
23 3.97 278 577/579 −/− Procyanidin dimer
24A 4.31 279 745/747 −/− Gallocatechin catechingallate
24B 4.37 279 745/747 −/− Gallocatechin catechingallate
25 4.78 210,279 289/291 −/− (−)-epicatechin
26 4.91 213,274 457/459 −/− (−)-Epigallocatechin 3-0-gallate
27 5.15 278 457/459 −/− (−)-Gallocatechin 3-0-gallate
28 5.32 277 383/385 −/− Epicatechin 3-O-benzoate
29 5.52 278 457/459 169/195 Gallocatechin gallate
30 5.58 226,310 163/165 119/120,142,157 p-Coumaric acid
31 5.70 276 897/899 730/195,437,731 Digallocatechin-catechin
32 5.74 276 897/899 730/195,437,731 Digallocatechin-catechin
33 5.82 277 729/731 635/195,446,451 Catechin-4α-epicatechin 3-O-gallate
34 5.91 278 635/− /− Trigalloylglucose
35 6.16 262,356 479/481 −/319 Myricetin 3-O-galactoside
36 6.31 262,355 479/481 −/319 Myricetin 3-O-glucoside
37 6.42 276 471/473 287/195,289 (−)-Epigallocatechin 3-O-(4″-O-methyl) gallate
38 6.62 276 471/473 287/195,289 (−)-Epigallocatechin 3-O-(3″-O-methyl) gallate
39 6.67 256,354 771/773 −/303,465,611 Quercetin 3-O-galactosylrutinoside
40 6.98 256,354 771/773 −/303,465,611 Quercetin 3-O-glucosylrutinoside
41 7.16 277 441/443 −/− (−)-Epicatechin gallate
42 7.22 278 441/443 −/− (+)-Catechin gallate
43 7.42 256,354 609/611 −/303,465 Quercetin 3-O-rhamnosylgalactoside
44 7.59 348 609/611 −/303,465 Rutin
45 7.68 210,256,354 463/465 −/120,157,303 Quercetin 3-O-glucoside
46 7.97 256,354 755/757 −/303,465 Quercetin 3-O-dirhamnosylglucoside
47 8.02 266,348 755/757 −/287,449,595 Kaempferol 3-O-glucosylrutinoside
48A 8.39 265,347 447/449 −/195,287 Kaempferol 3-O-galactoside
48B 8.43 265,347 739/741 −/− Kaempferol 3-O-rhamnosylrutinoside
49 8.56 265,347 593/595 −/287,449 Kaempferol-3-O-rutinoside
50 8.74 283,330sh 603/605 −/− Unknown
51 8.86 210,265,347 447/449 −/303 Kaempferol 3-O-glucoside
52 9.37 266,368 317/319 −/− Myricetin
53 9.75 266,348 781/783 −/287 Kaempferol 3-O-acetyldirhamnosylhexoside
54 9.88 220,275 705/707 −/− Quercetin dihexoside sulphate
55 10.03 221,282 499/501 −/− Unknown
56 11.52 266,310 1063/1065 −/303 Quercetin 3-O-acylglycoside
57 11.92 266,368 301/303 −/− Quercetin
58 12.73 266,310 593/595 −/287 Kaempferol 3-O-p-coumaroylglucoside
59 12.85 266,310 901/903 −/287 Kaempferol 3-O-p-coumaroylglucosylrhamnosylglactoside
60 12.87 266,310 901/903 −/287 kaempferol 3-O-p-coumaroyldirhamnosylhexoside
61 12.96 268 447/449 −/287 Quercetin 3-O-rhamnoside
62 13.01 260,310 885/887 −/287 Kaempferol 3-O-p-coumaroyldirhamnosylhexoside
63 13.08 260,310 593/595 −/287 Kaempferol 3-O-6″-p-coumaroylglucoside
64 13.25 218,267,338 269/271 −/− Unknown
65 13.33 266,368 285/287 −/− Kaempferol
66 13.56 226,285 467/469 −/− Unknown
67 13.65 266,316 739/741 285/287 Kaempferol 3-O-di-p-coumaroylhexoside
68 13.72 266,316 739/741 285/287 Kaempferol 3-O-2″,6″-di-p-coumaroylglucoside
Open in a new tab

3.3. Quantification of the chemical compounds

The described UPLC method was subsequently applied to simultaneously determine 54 compounds in the tea samples. The total contents (mg/g tea samples, mean ± SD) of phenolic acid relatives, catechins, purine alkaloids, proanthocyanidins, O-glycosylated flavonols, flavonols and flavones, and acylated glycosylated flavonols for the 3 different groups of teas are listed in Table 3. The results showed that catechins existed at the highest concentrations in all the tested samples. The total contents of catechins in the white teas and green teas were similar and greater than those in green pu-erh teas. Between the tea groups, the contents of phenolic acid derivatives were the highest in white teas, followed by that in green pu-erh teas, and much lower in green teas. Total flavonoid contents were highest in white teas, much lower in pu-erh teas and green teas. Other notables: TP was not detected in green teas, quercetin 3-O-dirhamnosylglucoside and kaempferol 3-O-p-cou-maroyldirhamnosylhexoside were not detected in white teas, apigenin was not detected in green pu-erh teas.

Table 3.

Quantitation results (mg/g) of some important chemical compounds in tested tea samples.

Peak No Compound Contents (mg/g, mean ± SD)
Green pu-erh Green tea White tea
Phenolic acid relatives
3 Galloylglucose 0.4116 ± 0.2175 0.2853 ± 0.1813 0.3547 ± 0.1158
4A + 4B + 4C 3-Galloylquinic acid + 5-galloylquinic acid + gallic acid 30.6355 ± 12.2719 16.6109 ± 5.8619 40.6541 ± 12.3919
9 3-Caffeoylquinic acid 0.1387 ± 0.0266 0.0425 ± 0.0282 0.0443 ± 0.0098
10 5-Caffeoylquinic acid 0.4065 ± 0.1020 0.0903 ± 0.0582 0.0356 ± 0.0112
16 Gallic acid methyl ester 0.1950 ± 0.1625 0.1106 ± 0.0342 6.7001 ± 6.3004
17 1,6-Digalloylglucose 0.2357 ± 0.1861 0.2114 ± 0.1476 0.7519 ± 0.4278
19A + 19B Chlorogenic acid + caffeoylmalic acid 1.1004 ± 0.8073 0.3330 ± 0.1757 0.6435 ± 0.3064
21 4-Caffeoylquinic acid 5.1895 ± 1.7253 0.8125 ± 0.7868 0.0924 ±0.0835
22 Caffeic acid 0.0191 ± 0.0427 2.5669 ± 1.3071 3.0477 ± 0.8248
30 p-Coumaric acid 3.8126 ± 1.2359 3.9910 ± 1.2861 2.9695 ± 1.6811
34 1,2,6-Trigalloylglucose 4.4141 ± 2.6049 2.9223 ± 1.6583 19.4409 ± 16.0670
Total 46.5586 ± 14.5440 27.9767 ± 9.0805 74.7346 ± 31.8080
Catechins
7A Gallocatechin 0.4287 ± 0.0985 0.3800 ± 0.1826 0.4415 ± 0.3101
14 (−)-Epigallocatechin 15.5085 ± 18.2513 47.0838 ± 29.4833 17.2873 ± 14.3654
18 (+)-Catechin 3.8890 ± 1.6334 1.2264 ± 0.5175 0.6592 ± 0.4646
20A (−)-Methylepigallocatechin gallate 0.6670 ± 0.1264 0.8923 ± 0.1131 0.8671 ± 0.2337
25 (&minus)-Epicatechin 18.4774 ± 4.4958 15.4012 ± 3.6785 9.3480 ± 3.7320
26 (−)-Epigallocatechin 3-O-gallate 88.3355 ± 37.5361 144.2102 ± 29.1159 169.7685 ± 4.7624
27 (−)-Gallocatechin 3-O-gallate 0.2405 ± 0.0681 0.3326 ± 0.2538 0.3849 ± 0.1964
28 Epicatechin 3-O-benzoate 1.0737 ± 0.4470 1.9573 ± 1.0008 1.2857 ± 0.9623
33 Catechin-4α-epicatechin 3-O-gallate 0.7991 ± 0.4140 0.3315 ± 0.2937 0.6431 ± 0.1915
38 (−)-Epigallocatechin 3-O-(3″-O-methyl) gallate 1.6149 ± 2.2873 1.6435 ± 1.1365 0.2546 ± 0.1547
Total 131.0343 ± 58.9121 213.7414 ± 52.9722 200.9399 ± 15.6054
Purine alkaloids
7B Theobromine 2.1380 ± 0.7187 1.6380 ± 0.7871 1.9818 ± 1.2951
11 Theophylline 0.2277 ± 0.2538 ND 0.3489 ± 0.6978
20B Caffeine 24.1144 ± 6.2303 29.8456 ± 4.3371 43.0819 ± 4.9826
Total 26.4800 ± 6.5912 31.4836 ± 5.0835 45.4126 ± 4.0074
Proanthocyanidins
24A + 24B Gallocatechin catechingallate 0.4174 ± 0.1259 0.9587 ± 0.2905 0.7988 ± 0.4608
31 Digallocatechin-catechin 0.1082 ± 0.1162 0.6727 ± 0.7336 0.7224 ± 1.1990
32 Digallocatechin-catechin 0.3403 ± 0.2164 0.1761 ± 0.1385 3.1458 ± 1.6156
Total 0.8660 ± 0.3216 1.8075 ± 0.9498 4.6670 ± 2.7229
O-Glycosylated flavonols
40 Quercetin 3-O-glucosylrutinoside 0.5014 ± 0.3690 4.4178 ± 1.7759 6.4777 ± 3.2404
43 Quercetin 3-O-rhamnosylgalactoside 3.5999 ± 1.3702 1.2165 ± 1.3354 0.8465 ± 0.4835
44 Rutin 0.8781 ± 0.8544 0.0933 ± 0.0260 0.0063 ± 0.0142
45 Quercetin 3-O-glucoside 2.5978 ± 0.7897 0.7904 ± 0.4351 0.7641 ± 0.5335
46 Quercetin 3-O-dirhamnosylglucoside 0.0054 ± 0.0121 0.0733 ± 0.1000 ND
47 Kaempferol 3-O-glucosylrutinoside 0.5145 ± 0.2190 5.3761 ± 2.0145 5.2526 ± 1.5471
48A + 48B Kaempferol 3-O-galactoside + kaempferol 3-O-rhamnosylrutinoside 0.3127 ± 0.2498 1.1999 ± 1.3701 0.5445 ± 0.2332
49 Kaempferol-3-O-rutinoside 1.9691 ± 0.5651 1.0491 ± 0.6987 0.7587 ± 0.3128
51 Kaempferol 3-O-glucoside 1.3033 ± 0.3286 0.9552 ± 0.6125 0.9692 ± 0.3647
61 Quercetin 3-O-rhamnoside 0.0725 ± 0.0325 0.0840 ± 0.0607 0.1867 ± 0.0955
Total 11.7548 ± 3.0255 15.2556 ± 3.9826 15.8064 ± 3.6666
Flavonols and flavone
52 Myricetin 0.0372 ± 0.0365 0.0723 ± 0.0130 0.1788 ± 0.1449
57 Quercetin 0.1964 ± 0.1985 0.0176 ± 0.0363 0.0283 ± 0.0387
64 Apigenin ND 0.0642 ± 0.0254 0.0533 ± 0.0143
65 Kaempferol 0.0529 ± 0.0219 0.0216 ± 0.0182 0.1042 ± 0.0998
Total 0.2865 ± 0.2451 0.1756 ± 0.0336 0.3646 ± 0.1034
Acylated glycosylated flavonols
53 Kaempferol 3-O-acetyl-dirhamnosylhexoside 0.0917 ± 0.0668 0.0442 ± 0.0645 0.0386 ± 0.0272
58 Kaempferol 3-O-p-coumaroylglucoside 0.0807 ± 0.0348 0.0692 ± 0.0269 0.2217 ± 0.1213
59 + 60 kaempferol 3-O-p-coumaroylglucosylrhamnosylglactoside + kaempferol 3-O-p-coumaroyldirhamnosylhexoside 0.0690 ± 0.0294 0.0253 ± 0.0290 0.0138 ± 0.0192
62 Kaempferol 3-O-p-coumaroyldirhamnosylhexoside 0.0176 ± 0.0110 0.0477 ± 0.0362 ND
63 Kaempferol 3-O-6″-p-coumaroylglucoside 0.0145 ± 0.0153 0.0141 ± 0.0055 0.0322 ± 0.0314
67 Kaempferol 3-O-di-p-coumaroylhexoside 0.0695 ± 0.0218 0.0433 ± 0.0234 0.4431 ± 0.5119
68 Kaempferol 3-O-2″,6″-di-p-coumaroylhexoside 0.0203 ± 0.0118 0.0190 ± 0.0126 0.0903 ± 0.0775
Total 0.3633 ±0.1392 0.2628 ± 0.1050 0.8398 ± 0.7088
Open in a new tab

ND: not detected.

3.4. Principal component analysis (PCA)

PCA is an unsupervised mathematical procedure that transforms a number of possibly correlated variables into a smaller number of uncorrelated variables called principal components. Its operation can be thought of as revealing the internal structure of the data in a way which best explains the variance in the data.

When chromatographic fingerprints are used for PCA, there are basically two sources of variation contained in chromatographic fingerprints. One comes from the experimental process, and the other from the samples themselves. The main variation arising from the experimental process are retention time shifts that make the comparison of fingerprints acquired from different systems or at different times difficult for PCA analysis. A retention time align step is usually needed for the PCA to be successful. Despite numerous studies, computerised peak alignment programs have had only limited success, as discussed in our previous studies (Chen et al., 2007; Lin et al., 2008).

Fig. 3a shows the PCA score plot from the 1st approach. Since all the chromatograms was collected from one system within a week, the alignment procedure finished flawlessly. However, as shown in the PCA score plot, this approach did not work very well even under those optimal conditions. The primary reasons are (1) tea is a very complex botanical and (2) the data used in PCA included all the information acquired and a lot of them were either redundant or no-critical.

Fig. 3.

Fig. 3

Open in a new tab

(a) PCA plot based on the original profile matrix exported from Masslynx after spectra alignment by SpecAlign software; (b) PCA plot based on the contents of the 54 determined chemical compounds; (c) PCA plot based on the relative peak areas of main peaks.

Fig. 3b shows the PCA score plot from the 2nd approach. All the quantitative data were summarised into 1 matrix in Excel. The matrix was then used for PCA. The PCA score plot produced clear separation of the three tea groups (Green pu-erh tea, Green tea and White tea). Although the sample WT1 was separated from the white tea cluster, it is in the direction of white tea group. From the results, we found that the contents of gallic acid, 1,2,6-trigalloylglucose and caffeine in WT1 were much higher compared to other white teas. The method, although effective, is not practical since quantitation of all the major components of a botanical is a very expensive and time-consuming process.

Fig. 3c shows the PCA score plot from the 3rd approach. The biggest advantage of the approach is that it is nearly fool-proof, as the retention time shifts introduced by the instruments and other experimental conditions (fluctuation of the temperatures, the reagents used, and the columns, etc.) could be minimised. The tea samples could be classified into three groups, which were marked as “Green pu-erh”, “Green tea”, and “White tea”. The data showed that the teas of the same category were clustered into the same group. “Green tea” group and “White tea” group were separated from “Green pu-erh” group in PC1, also, the three groups were separated well in PC2. Although GP2 was a little farther from other GP samples, it was separated distinctly from green teas and white teas. It could be asserted that the samples classified into one group were associated with similar chemical properties/components.

Among the 3 approaches, the 1st approach is the easiest, but unfortunately not very effective when used in very complex samples like teas even under the best possible scenarios. The second approach works, but is too labour intensive for PCA since it used the quantitation results of 58 compound. The 3rd approach gives the best separation among the 3 approaches and can be used to compare fingerprints obtained from different instruments, operated by different people, and under different conditions.

4. Conclusion

The results of this study show that the optimised UPLC fingerprint techniques combined with the chemometrics method are able to classify tea samples objectively and successfully in accordance with the different categories. The results also indicate that the UPLC method has the practical advantages of shorter analysis time and reduced solvent consumption, which makes it an attractive alternative to conventional HPLC technique in routine fingerprint analysis, especially in situations where high sample throughput and fast analytical speed are needed. The method developed in the present study can provide an important reference to the quality control of teas.

Acknowledgments

This research is supported by the Agricultural Research Service of the US Department of Agriculture and an Interagency Agreement with the Office of Dietary Supplements of the National Institutes of Health.

References

  1. Alan Crozier A, Jaganath IB, Clifford MN. Dietary phenolics: Chemistry, bioavailability and effects on health. Natural Product Reports. 2009;26(8):1001–1043. doi: 10.1039/b802662a. [DOI] [PubMed] [Google Scholar]
  2. Alcazar A, Ballesteros O, Jurado JM, Pablos F, Martin MJ, Vilches JL, et al. Differentiation of green, white, black, oolong, and pu-erh teas according to their free amino acids content. Journal of Agricultural and Food Chemistry. 2007;55(15):5960–5965. doi: 10.1021/jf070601a. [DOI] [PubMed] [Google Scholar]
  3. Chen P, Ozcan M, Harnly J. Chromatographic fingerprint analysis for evaluation of Ginkgo biloba products. Anaytical and Bioanalytical Chemistry. 2007;389(1):251–261. doi: 10.1007/s00216-007-1386-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chen P, Song F, Lin LZ. Chromatographic fingerprint analysis of pycnogenol® dietary supplements. Journal of AOAC International. 2009;92(2):624–632. [PubMed] [Google Scholar]
  5. Chiu FL, Lin JK. HPLC analysis of naturally occurring methylated catechins, 3″- and 4″-methyl-epigallocatechin gallate, in various fresh tea leaves and commercial teas and their potent inhibitory effects on inducible nitric oxide synthase in macrophages. Journal of Agricultural and Food Chemistry. 2005;53(18):7035–7042. doi: 10.1021/jf0507442. [DOI] [PubMed] [Google Scholar]
  6. Clifford MN, Stoupi S, Kuhnert N. Profiling and characterization by LC–MSn of the galloylquinic acids of green tea, tara tannin, and tannic acid. Journal of Agricultural and Food Chemistry. 2007;55(8):2797–2807. doi: 10.1021/jf063533l. [DOI] [PubMed] [Google Scholar]
  7. Del Rio D, Stewart AJ, Mullen W, Burns J, Lean MEJ, Brighenti F, et al. HPLC–MSn analysis of phenolic compounds and purine alkaloids in green and black tea. Journal of Agricultural and Food Chemistry. 2004;52(10):2807–2815. doi: 10.1021/jf0354848. [DOI] [PubMed] [Google Scholar]
  8. Encyclopedia of chinese medicinal plants. Vol. 3. Shanghai: Shanghai Press of Science and Technology; 1999. Tea leaves (#2159) Editorial committee of encyclopedia of Chinese medicinal plant-sEds; pp. 565–574. [Google Scholar]
  9. Dufresne CJ, Farnworth ER. A review of latest research finding on the health promotion properties of tea. The Journal of Nutritional Biochemistry. 2001;12(7):404–421. doi: 10.1016/s0955-2863(01)00155-3. [DOI] [PubMed] [Google Scholar]
  10. Fernandez PL, Martin MJ, Gonzalez AG, Pablos F. HPLC determination of catechins and caffeine in tea. Differentiation of green, black and instant teas. Analyst. 2000;125(3):421–425. doi: 10.1039/a909219f. [DOI] [PubMed] [Google Scholar]
  11. Friedman M. Overview of antibacterial, antitoxin, antiviral and antifungal activities of tea flavonoids and teas. Molecular Nutrition & Food Research. 2007;51(1):116–134. doi: 10.1002/mnfr.200600173. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Friedman M, Mackey BE, Kim HJ, Lee IS, Lee KP, Lee SU, et al. Structure-activity relationships of tea compounds against human cancer cells. Journal of Agricultural and Food Chemistry. 2007;55(2):243–253. doi: 10.1021/jf062276h. [DOI] [PubMed] [Google Scholar]
  13. Fujiwara M, Ando I, Arifuku K. Multivariate analysis of 1H–NMR spectra of two hundred kinds of tea in the world. Analytical Sciences. 2006;22(10):1307–1314. doi: 10.2116/analsci.22.1307. [DOI] [PubMed] [Google Scholar]
  14. Hilal Y, Engelhardt U. Characterisation of white tea - Comparison to green and black tea. Journal für Verbraucherschutz und Lebensmittelsicherheit. 2007;2(4):414–421. [Google Scholar]
  15. Hsieh MM, Chen SM. Determination of amino acids in tea leaves and beverages using capillary electrophoresis with light-emitting diode-induced fluorescence detection. Talanta. 2007;73(2):326–331. doi: 10.1016/j.talanta.2007.03.049. [DOI] [PubMed] [Google Scholar]
  16. Ito Y, Sugimoto A, Kakuda T, Kubota K. Identification of potent odorants in Chinese jasmine green tea scented with flowers of Jasminum sambac. Journal of Agricultural and Food Chemistry. 2002;50(17):4878–4884. doi: 10.1021/jf020282h. [DOI] [PubMed] [Google Scholar]
  17. Jie G, Lin Z, Zhang L, Lv H, He P, Zhao B. Free radical scavenging effect of pu-erh tea extracts and their protective effect on oxidative damage in human fibroblast cells. Journal of Agricultural and Food Chemistry. 2006;54(21):8058–8064. doi: 10.1021/jf061663o. [DOI] [PubMed] [Google Scholar]
  18. Kato M, Shibamoto T. Variation of major volatile constituents in various green teas from Southeast Asia. Journal of Agricultural and Food Chemistry. 2001;49(3):1394–1396. doi: 10.1021/jf001321x. [DOI] [PubMed] [Google Scholar]
  19. Lakenbrink C, Engelhardt UH, Wray V. Identification of two novel proanthocyanidins in green tea. Journal of Agricultural and Food Chemistry. 1999;47(11):4621–4624. doi: 10.1021/jf9813081. [DOI] [PubMed] [Google Scholar]
  20. Lambert JD, Yang CS. Cancer chemopreventive activity and bioavailability of tea and tea polyphenolis. Mutation Research. 2003;523–524:201–208. doi: 10.1016/s0027-5107(02)00336-6. [DOI] [PubMed] [Google Scholar]
  21. Lin LZ, Chen P, Harnly JM. New phenolic components and chromatographic profiles of green and fermented teas. Journal of Agricultural and Food Chemistry. 2008;56(17):8130–8140. doi: 10.1021/jf800986s. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lin JK, Lin CL, Liang YC, Lin-Shiau SY, Juan IM. Survey of catechins, GA, and methylxanthines in green, oolong, pu-erh and black teas. Journal of Agricultural and Food Chemistry. 1998;46(9):3635–3642. [Google Scholar]
  23. Miketova P, Schram KH, Whitney J, Li M, Huang R, Kerns E, et al. Tandem mass spectrometry studies of green tea catechins. Identification of three minor components in the polyphenolic extract of green tea. Journal of Mass Spectrometry. 2000;35(7):860–869. doi: 10.1002/1096-9888(200007)35:7<860::AID-JMS10>3.0.CO;2-J. [DOI] [PubMed] [Google Scholar]
  24. Naef R, Jaquier A, Velluz A, Maurer B. New constituent related to 3- methyl-2, 4-nonanedieone identified in green tea. Journal of Agricultural and Food Chemistry. 2006;54(24):9201–9205. doi: 10.1021/jf061738o. [DOI] [PubMed] [Google Scholar]
  25. Palmer JK. Enzyme reactions and acceptability of plant foods. Journal of Chemical Education. 1984;61(4):284–288. [Google Scholar]
  26. Pongsuwan W, Bamba T, Harada K, Yonetani T, Kobayashi A, Fukusaki E. High throughput technique for comprehensive analysis of Japanese green tea quality assessment using ultra performance liquid chromatography with time of flight mass spectrometry (UPLC/TOF MS) Journal of Agricultural and Food Chemistry. 2008;56(22):10705–10708. doi: 10.1021/jf8018003. [DOI] [PubMed] [Google Scholar]
  27. Pongsuwan W, Fukusaki E, Bamba T, Yonetani T, Yamahara T, Kobayashi A. Prediction of Japanese green tea ranking by gas chromatography/mass spectrometry-based hydrophilic metabolite fingerprinting. Journal of Agricultural and Food Chemistry. 2007;55(2):231–236. doi: 10.1021/jf062330u. [DOI] [PubMed] [Google Scholar]
  28. Wu SC, Yen GC, Wang BS, Chiu CK, Yen WJ, Chang LW, Duh PD. Antimutagenic and antimicrobial activities of pu-erh tea. LWT–Food Science and Technology. 2007;40(3):506–512. [Google Scholar]
  29. Xie GX, Ye M, Wang YG, Ni Y, Su MM, Huang H, et al. Characterization of pu-erh tea using chemical and metabolic profiling approaches. Journal of Agricultural and Food Chemistry. 2009;57(8):3046–3054. doi: 10.1021/jf804000y. [DOI] [PubMed] [Google Scholar]
  30. Zeeb DJ, Nelson BC, Albert K, Dalluge JJ. Separation and identification of twelve catechins in tea using liquid chromatography/atmospheric pressure chemical ionization-mass spectrometry. Analytical Chemistry. 2000;72(20):5020–5026. doi: 10.1021/ac000418f. [DOI] [PubMed] [Google Scholar]
  31. Zhao J, Chen Q, Huang X, Fang CH. Qualitative identification of tea categories by near infrared spectroscopy and support vector machine. Journal of Pharmaceutical and Biomedical Analysis. 2006;41(4):1198–1204. doi: 10.1016/j.jpba.2006.02.053. [DOI] [PubMed] [Google Scholar]
  32. Zheng L, Watson DG, Johnston BF, Clark RL, Edrada-Ebel R, Elseheri W. A chemometric study of chromatograms of tea extracts by correlation optimization warping in conjunction with PCA, support vector machines and random forest data modeling. Analytica Chimica Acta. 2009;642(1–2):257–265. doi: 10.1016/j.aca.2008.12.015. [DOI] [PubMed] [Google Scholar]
  33. Zhou ZH, Zhang YJ, Xu M, Yang CR. Puerins A and B, two new 8-C substituted flavan-3-ols from Pu-er tea. Journal of Agricultural and Food Chemistry. 2005;53(22):8614–8617. doi: 10.1021/jf051390h. [DOI] [PubMed] [Google Scholar]
  34. Zhu XL, Chen B, Ma M, Luo XB, Zhang F, Yao SZ, et al. Simultaneous analysis of theanine, chlorogenic acid, purine alkaloids and catechine in tea samples with the help of multi-dimension information of on-line high performance liquid chromatography/electrospray–mass spectrometry. Journal of Pharmaceutical and Biomedical Analysis. 2004;34(3):695–704. doi: 10.1016/s0731-7085(03)00605-8. [DOI] [PubMed] [Google Scholar]

RESOURCES