Applications of chemometrics techniques on chromatographic and spectroscopic methods to advance chemical analysis of Radix Ligustici Chuanxiong, Radix Angelicae Sinensis, Cortex Phellodendri and other Chinese herbal medicines

Abstract

Chromatography and spectroscopy are the most commonly used analytical methods in various fields of chemistry and their instrumental technologies have been well developed. Yet, Chinese Herbal Medicine (CHM) is not easy to be analyzed because of their complexity of chemical composition. Data processing techniques provides good opportunity for mining more useful chemical information from original information-rich data. Chemometrics is a chemical discipline that uses mathematics, statistics to extract more useful chemical information from chemical data. Owing to its rapid development, several effective and precise techniques are invented to process data from advanced hyphenated instruments so as to characterize the chemical composition of CHM in more detail. In our works, conventional approach, 'marker' and recently introduced approaches including 'multi-component' and 'pattern' were devised and applied into chromatographic and spectroscopic data sets together with chemometrics techniques for developing the chromatographic fingerprints of three CHMs, Radix Ligustici Chuanxiong (CX), Radix Angelicae Sinensis (DG), Cortex Phellodendri (HB) for both qualitative and quantitative analyses. Chemometrics resolution method and spectral correlative chromatography were applied for finding out the common constituents of these CHMs while local least square method was used for chromatographic alignment during constructing the chromatographic fingerprint. The overall results indicated that pattern approach, with the use of the whole chromatogram, not only distinguishes geographical locations including Szechuan, Yunnan and Guizhou provinces of CX as well as Phellodendron chinense Schneid and Phellodendron amurense Rupr of HB through multivariate analysis, but is also feasible on classifying the two CHMs, CX and DG, with 100% correctness in which five-folds cross validations were involved via several pattern recognition methods. Finally, the rapid and non-destructive near-infrared reflectance spectroscopy (NIRRS) was utilized in analyzing chemical compositions of CHMs. With the proper use of chemometric pre-treatment and processing techniques on the full NIR spectra, the feasibility of applying NIRRS to CHM for identification of different parts of Herba menthae, of species of HB and of five different CHMs and quantification of major ingredients were achieved. In conclusion, we believe that using most chemical information in data analysis can advance the standard of quality control of CHM as demonstrated in this work.