Abstract
A reverse phase gradient elution HPLC method using 3 um ODS has been developed to separate over 40 pigmented polyphenols from a black tea liquor with detection at 450 nm. Four of these peaks were identified as the well characterised compounds, theaflavin, theaflavin-3-monogallate, theaflavin-3'-monogallate and theaflavin-3,3'-digallate. Diode array spectral data indicated that the remaining peaks belong to the previously poorly characterised thearubigin (TR) class of compounds. The HPLC method has been applied to both green and black tea liquors and further demonstrated that these non-theaflavin peaks were not untransformed green tea polyphenols such as flavonol glycosides, but were products of fermentation. In addition, the HPLC profiles obtained for black tea liquors showed a noticeable rising baseline and suggested the presence of TR material beyond the resolution capabilities of the HPLC system. By combining the HPLC method with an in vitro model fermentation system, containing different combinations of the six major green tea catechins (flavan-3-ols) and tea PPO, the formation and transformation of TF and TR have been studied. The effects of varying fermentation time, oxygen tension, pH and temperature were investigated. It was observed that in vitro oxidations yielded products very similar to those separated from a black tea liquor and thus acted as a good model for in vivo black tea fermentation. Moreover, the rising baseline observed from analysis of black tea liquors could also be produced from the model system studies. In vitro oxidation of single catechins showed that many of the resolvable TR compounds arose from the gallocatechins (epigallocatechin gallate and epigallocatechin), whereas unresolvable TR material, observed as a rising baseline on the HPLC profile, were obtained from the simple catechins (epicatechin and catechin). However, it could not be explained why the simple catechin gallate, epicatechin gallate, did not produce an equivalent baseline rise. Varying pH and temperature had little effect on the oxidation of the simple catechins, but a much greater effect on that of the gallocatechins. Paired catechin oxidations led to the formation of tf, tfmg, tf'mg and tfdg as well as TR. The greatest number of resolvable TR compounds arose from (EGCG + EC) and (EGCG + ECG), whereas the greatest yields of unresolvable TR were obtained from (EGC + EC) and (EGCG + EC). These observations highlighted the importance of the relevant catechins used for paired oxidations in determining the extent to which resolvable and unresolvable TR is produced. The formation of unresolvable TR through coupled oxidation of TF with simple catechins was considered to relate to the molar ratio of gallocatechin to simple catechin. (EGC + ECG) was the only model system where this ratio was greater than 1, indicating a deficit of the simple catechin and providing one explanation for this model system resulting in negligible production of unresolvable TR. In vitro oxidation of catechin mixes, containing the six major catechins, showed that resolvable TR peaks, particularly those eluting in the first half of the HPLC profile, could be selectively increased or decreased by altering pH and temperature. These latter results further demonstrated the significant effect of pH and temperature on the rate at which simple and gallocatechins are differentially oxidised. In vitro studies on the oxidation of individual TF compounds in the presence of EC showed a dramatic loss of TF and rise in baseline 5-10 min after initiating the reaction. These observations strongly suggested the role played by this simple catechin or rather it's quinone in coupled oxidative breakdown of TF to produce non-resolvable TR. The results from the in vitro oxidation studies highlight several important points. First, resolvable TR compounds are likely to arise from the catechins directly via their quinones or via intermediates prior to TF formation. Secondly, unresolvable TR compounds, probably of higher molecular weight, arise mainly from either the simple catechin quinones directly or via coupled oxidative breakdown of TF in which the simple catechin quinones act as electron carriers. Thirdly and most significantly, there is much greater potential to produce relatively pure TR compounds which can be isolated by preparative HPLC more easily from the soluble products of the model system than from a black tea liquor. Some preliminary structural studies were carried out on material isolated from both a black tea liquor and a model system fermentation. MS and NMR data obtained for Peak 13 from in vitro oxidation of EGCG are discussed, but remain inconclusive regarding it's precise structure. Suggestions for further work have been made. These include extending the model system studies to incorporate other substrates and enzymes and the further development of analytical methods better suited to investigating the nature of the materials responsible for the rising baseline.