Image analysis approaches to improve the thin layer chromatography – Chemometric-based investigations of natural extracts

Abstract

The combination of high-performance thin-layer chromatography (HPTLC) with image analysis (IA) and chemometrics becomes an attractive tool for natural extracts investigations. The large variability of these samples requires powerful image acquisition devices, multivariate image processing techniques and advanced chemometric methods to facilitate the interpretation of the chromatographic data. In the current study, two image acquisition devices and different image processing procedures were investigated using the HPTLC chromatograms of hydroalcoholic extracts of Gallium verum. Different sets of chromatographic data were generated for both UV chromatograms (obtained at 254 nm and at 366 nm) using images acquired with a digital camera and an UV-Vis TLC scanner. In all cases the Principal Component Analysis (PCA) technique was used in order to extract the information from chromatographic profiles. Variables of gray and pure colour red, green and blue intensities of pixels from start to front were used as input data in all cases. The results obtained by PCA investigations of HPTLC data from UV chromatograms at 254 nm and 366 nm respectively, provided complementary information related to the characteristics of the investigated extracts. Moreover, important steps as appropriate color scale selection and image processing/analysis procedures were pointed out based on the obtained results.