Comparison of three second-order multivariate calibration methods for the rapid identification and quantitative analysis of tea polyphenols in Chinese teas using high-performance liquid chromatography

Retention time shifts in second-order calibration-assisted chromatographic analysis seriously impact the modeling and quantitative accuracies in complex systems. In this work, three second-order methods, i.e. alternating trilinear decomposition (ATLD) algorithm, multivariate curve resolution-alternating least squares (MCR-ALS), alternating trilinear decomposition-assisted multivariate curve resolution (ATLD-MCR), were compared their performance to process liquid chromatographic data in the presence of retention time shifts and overlapped peaks. Firstly, the validation samples contain five tea polyphenols at three concentrate levels within the calibration ranges, helped to understand, visualize and interpret these features of three second-order multivariate methods. Secondly, experimental data were studied concerning the determination of polyphenols in Chinese tea samples by HPLC-DAD. The results showed that all three second-order multivariate methods realized satisfactory quantification for five targeted analytes in Pu-Er ripe tea samples and Green tea samples even with the interference of slight retention time shifts, average recoveries were 91.23% –113.16% for ATLD, 89.96%–115.96% for ATLD-MCR, 90.64%–117.60% for MCR-ALS, respectively. However, ATLD was disappointing in the case of larger time shifts (approx. 4.00 s and 6.40 s) occurring for the quantitative analysis of Black tea and Clinacanthus nutans tea, the average recoveries were just 67.33–84.05%. Relatively, MCR-ALS and ATLD-MCR were more significantly excellent, satisfactory results still can be obtained, the average recoveries for MCR-ALS and ATLD-MCR were in the range of 86.04–117.60% and 89.96–115.96%, respectively.