Properties of leaf and infusion col

Properties of leaf and infusion colours, chemical components and volatile flavour compounds of oolong teas and their correlation with perceived quality score given by tea-tasting panel were analysed. The scores for appearance and infused leaf correlated strongly with concentrations of chlorophyll a (chl a), chlorophyll (chl b) and chlorophyll (chl) (r = 0.355–0.433, P < 0.05) and the total quality score (TQS) positively correlated with concentrations of chl a, chl b and chl (r = 0.517–0.533, P < 0.01). The perceived taste score and TQS positively correlated with the concentration of total free amino acid (r = 0.514, 0.694, P < 0.01) and theanine (r = 0.500, 0.684, P < 0.01). The volatile composition and their quantities varied widely among oolong tea samples. Nerolidol, indole, benzeneacetaldehyde, linalool, linalool oxide I, hexanal, benzyl nitrile, geraniol and 1-penten-3-ol were prevailing volatile compounds detected in most of oolong tea samples. Principal component analysis screened ten principal components with the first three (glutamic acid, total catechins and benzeneacetaldehyde) explaining 27.86%, 20.00% and 14.46% of the total variance, respectively. Regression analysis upon the ten principal components formulated a prediction model on the total quality score with 78.5% probability. The result showed that oolong teas could be partially classified by cluster analysis based on principal components.
Tea (Camellia sinensis) is one of the most popular beverages consumed worldwide. There are three types of tea: green, oolong and black. Oolong tea is partially fermented during processing, whereas green tea is not fermented and black tea is fully fermented. Oolong tea is manufactured predominantly in Fujian, Guangdong and Taiwan provinces of China. It is well established that the flavour of tea is principally determined by chemical components it contains, such as volatile compounds contributing to the property of aroma and nonvolatile compounds to the taste (Hara et al., 1995; Scharbert & Hofmann, 2005). Considerable attempts have been made to link the sensory assessment of tea quality index to chemical compounds for black tea (Liang et al., 2003; Scharbert
& Hofmann, 2005), green tea (Wang & Ruan, 2009), pu-erh tea (Liang et al., 2005b) and jasmine tea (Liang et al., 2007). Although oolong tea is getting more and more popular in the world, especially in China and Japan, there is much less investigation on the quality of different oolong tea in comparison with the vigorous studies on the quality of green and black teas. Oolong tea is a semi-fermented tea as partially chlorophylls (chl), catechins and other polyphenols (PPs) are preserved after processing owing to inactivation of enzyme by dry heating. The perceived quality of oolong teas is assessed according to their appearance of leaf tea and the colour, taste and aroma of the brew and features of infused young shoots. Only a few studies have been performed on oolong teas with emphasis on taste properties (Huang et al., 2003; Chen et al., 2010), colour difference (Liang et al., 2005a) and aroma properties (Wang et al., 2008) while comprehensive study on chemical compositions associated with colour, taste, aroma and other essential features of this type of tea is relatively limited. The purpose of the present study was to explore the relationship of chemical components with the perceived quality index of oolong tea assessed by sensory evaluation.
Materials and methods
Materials A total of thirty-one oolong tea samples were collected from tea factories and tea companies (Fujian, Guangdong & Taiwan, from April to June 2008) in China (Table S1). All analyses including sensory evaluation were conducted within 3 months after sample collection. All analyses including sensory evaluation were conducted within 24 h of preparing tea infusion. Catechins and volatile compound standards for HPLC and gas chromatograph⁄mass spectrometer (GC⁄MS), chemicals (analytical reagent or above) were purchased from Changsha Chemical Company (Changsha, China), Sigma Co. (St Louis, MO, USA) or (Fluka AG, Chemische Fabrik, Switzerland).
Sensory evaluation Perceived quality score was blindly assessed according to a standardised procedure by a tea-tasting panel consisted of six professional panelists (The panelists were divided into two groups with each consisted of three panelists) (Liang et al., 2007; Wang & Ruan, 2009). The grading was performed to each of five attributes, the appearance (including colour, shape, regularity and uniformity of leaf tea), infusion colour, taste and aroma of the infusion and features of infused leaves (postinfusion, based mainly on bud and leaves proportions). The assessment of appearance was performed on leaf tea samples and other features after infusion process. Procedure of sensory evaluation is given as follows (Hui et al., 2004; Wang & Ruan, 2009). The total score of a sample was calculated from summing of each feature by weighting factors, of which 20% was awarded to the appearance of dry tea, 30% for the tea aroma, 10% for the infusion colour, 30% for the taste and 10% for the infused leaves (Gong, 2001).
CIELAB colour scale parameters DL, Da and Db of infusion Leaf tea sample of 3 g was infused similarly as those for sensory evaluation described above. After tea infusion was filtered using filter paper and cooled to room temperature, the infusion was immediately measured with colourimeter SMY-2000 (Shenmingyang Science Co., Ltd, Beijing, China) against distilled water for parameters DL, Da and Db based on the CIELAB colour scale.
Chlorophylls (chl) and carotenoids Grounded tea sample of 0.5 g was extracted with 100 mL 80% acetone in water (v⁄v) until colourless by grinding with sand and a small amount of CaCO3 in a mortar using a pestle (Wang & Ruan, 2009). The contents of chl and carotenoids were calculated with following equations (1–3): chlaðmg g1Þ¼ð12:7D6632:69D645Þ=ðð1SÞWÞ ð1Þ chlbðmg g1Þ¼ð22:9D6454:68D663Þ=ðð1SÞWÞ ð2Þ Carotenoids(mgg1Þ¼ð4:7D440ð1:38D663 þ5:48D645ÞÞ=ðð1SÞWÞ ð3Þ where S (%) is the moisture content and W is the weight of sample. D663, D645 and D440 are the corresponding (absorbance) readings from the spectrophotometer of the above solutions. Tea moisture was measured using a vacuum oven based on an international standard method (ISO 1573, 1980); chl a is chlorophyll a and chl b is chlorophyll b.
Polyphenols (PPs), catechins, free amino acids, caffeine and water extract Concentrations of PPs and total free amino acid (AA) were determined with routine methods after extracted with deionised water in a boiling bath for 45 min with occasional hand shaking (Liang et al., 2003; Wang & Ruan, 2009). The extract was immediately filtered and cooled to room temperature and measured for the concentrations of PPs as previously described (Erdemoglu et al., 2000). The absorbance was measured at 760 nm on UV⁄Vis spectrophotometer (Shimadzu UV 2550, Kyoto, Japan) and the results were expressed as gallic acid equivalents. Contents of total AA in the tea infusions were determined by a spectraphotometric method except that theanine was used as amino acid standard to make calibration graph (Liang et al., 2007). Water extract was determined by an international standard method (ISO 9768, 1994). Briefly a volume of 50 mL tea extract prepared as described above was evaporated in dish on boiling water bath to rough dry and further dried in an oven at 103 C to complete dryness and then weighted after cooling down to room temperature in a silicagel desiccator. The composition of free AAs was measured by an L-8800 amino acid analyser (Hitachi High-Technologies,Tokyo, Japan). The composition of caffeine and catechins in the extract was determined with a HPLC system (LC-2010AHT; Shimadzu Corp.) equipped with a Shim-pack VP-ODS column (5 lm, 4.6mm · 150 mm, 35 C) at 278 nm as previously described with some modifications (Wang et al., 2004). Solvents A (water) and B (N,N-dimethylformamide:methanol:acetic acid, 20:1:0.5, v⁄v) were run in linear gradients with B increasing from 14% to 23% within 13 min, from 23% to 36% within next 12 min and maintained for 3 min thereafter at a rate of 1.0 mL min)1. Concentrations of caffeine and catechins were quantified by their peak areas against those of standards prepared from authentic compounds.
Volatile compounds The preparation of oil extract and analysis of aromatic compounds were essentially the same as previously described (Wang & You, 1996). Briefly, 50 g of ground tea sample was extracted with 1000 mL hot water in a Liens-Nickerson simultaneous steam distillation continuous extraction (SDE) with ether as the solvent. Before SDE extraction 1.0 mL ethyl caproate (5 lL ethyl caproate of in 100 mL of ether) was added to the tea as an internal standard. The ethyl ether phase was then dehydrated with 5 g of anhydrous sodium sulphate overnight. The dehydrated ethyl ether phase was concentrated to 400 lL under a purified nitrogen stream. The volatile compounds in the extract were analysed by GCMS-QP2010 (Shimadzu). The column was RTX-5ms fused silica capillary column (30 m · 0.25 mm · 0.25 lm). The temperature was programmed from 50 C (held for 4 min) to 150 C (held for 1 min) at 2 C min)1, then programmed from 150 to 180 C (held for 5 min) at 5 C min)1, and then programmed from 180 to 280 C (held for 30 min) at 10 C min)1. MS ion source temperature was 210 C, and electron energy was 70 eV. Peak identification was achieved by interpretation of mass spectra and by coincidence of retention times with authentic standards. Concentrations of aromatic compounds were expressed as ratios of peak area to that of internal standard (ethyl caproate).
Data analysis and statistics All above analyses except sensory evaluation were duplicated for each sample and means were presented. Analyses of Spearman’s linear correlation, principal component (PCA) and linear regression based on means of duplicat