IMPLEMENTATION OF IMAGE ANALYSIS FROM THIN LAYER CHROMATOGRAM OF PURPLE SWEET POTATO LEAF EXTRACT IN CHEMICAL FINGERPRINTING

Abstract

Background: Purple sweet potato (Ipomoea batatas L.) is a crop that contains rich amounts of anthocyanins and flavonoids. For maintaining its quality in preparations, fingerprint analysis using Thin Layer Chromatography (TLC) is commonly used. Objective: In this study, image analysis was implemented on the chromatogram to develop fingerprint profiles of the leaves of various types of purple sweet potatoes.  Methods: Fingerprint analysis and acquisition of data were carried out by combining modern TLC equipment with software for image analysis. The in-situ fingerprint analysis using WinCats was compared with the image analysis results obtained in this study. The chemometric technique, cluster analysis, was performed to measure their difference in terms of sample classification. Cluster analysis was carried out to confirm that the clusters formed from image analysis were the same as those from WinCats. Results: The results obtained were two dendrograms, both showing the formation of two clusters but with a different order of samples. This study concluded that the data obtained from image analysis using ImageJ software resulted in clustering with a similarity of 38.15%, while the data obtained from WinCats software resulted in classification with a similarity of 28.31%. Both methods had issues in determining the fingerprint profiles, as evidenced by the low level of similarity. Conclusion: These results indicate that image analysis has the potential to be implemented for fingerprint profile development. However, further analysis is needed to align the Rf values and confirm the compounds in the peaks.