¹Sekolah Tinggi Ilmu Farmasi Makassar, Makassar 90242, Indonesia
²Faculty of Health Sciences, Almarisah Madani University, Makassar 90245, Indonesia
³Departement of Pharmacy, Faculty of Mathematics and Natural Sciences, Universitas Tadulako, Palu 94118, Indonesia
⁴Departement of Biology Pharmacy, Sekolah Tinggi Ilmu Farmasi Makassar, Makassar 90242, Indonesia
⁵Department of Pharmacognosy-Phytochemistry Laboratory, Faculty of Pharmacy, Hasanuddin University, Makassar 90245, Indonesia
⁶Faculty of Pharmacy, Universitas Halu Oleo, Kendari 93232, Indonesia
⁷Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Universitas Gadjah Mada, Yogyakarta, 55281, Indonesia DOI : 10.29228/jrp.2022.00 Applying chemometrics in analyzing the chemical data from natural materials is very important for understanding the complex data of herbal raw materials. This study aims to apply chemometric principal component analysis (PCA) and cluster analysis (CA) techniques to know the fingerprint profiling of 20 types of ethanolic extract of Mangosteen obtained from different area in Sulawesi Island, Indonesia. The extraction uses variations for solvent kinds (70% and 96% ethanol) with maceration techniques and fruit parts (pericarp and seed) representing plant parts from Sulawesi Island. All extracts were analyzed using FTIR spectroscopy at wavenumbers of 4000-400 cm-1, and the obtained FTIR spectra were analyzed using chemometric or multivariate data analysis. The results showed that some peaks were identified as markers for mangosteen extracts, namely at wavenumbers of 3414 cm-1 (O-H stretching), 2961 cm-1 (CH3 asymmetrical stretching), 2922 cm-1 (CH2 asymmetrical stretching), 2859 cm-1 (CH2 symmetrical stretching), 1724 cm-1 (C=O stretching), 1643 (hydrogen bonded), 1611 cm-1 (carboxyl groups), 1582 cm-1 (C=C stretching), 1458 cm-1 (CH2 bending), and 1283 cm-1 (O-H). PCA using the first principle component (PC1) and second principle components accounting of 85.8% variations could differentiate Mangosteen extracts with different area. Furthermore, CA could classify the samples of Mangosteen extracts into 4 clusters with a similarity percentages of 25.65%-99.99%. It can be concluded that chemometric-assisted fingerprinting profiles using variable of FTIR spectral absorbance values can distinguish variations in mangosteen extract according to the type of solvent and the fruit part according to Sulawesi Island origin. Keywords : Chemometrics; Fingerprinting profiling; Mangosteen; Multivariate data analysis; Spectroscopy