Application Chemometrics-Assisted Fingerprinting Profiling of Extract Variation from Mangosteen (Garcinia mangostana L.) Using FTIR Method

Abstract

Garcinia mangostana L. (mangosteen) is a plant that comes from the tropics and is very loved by the public because of its pleasant taste. It has anti-diabetic effects, wound treatment, diarrhea, skin anti-infection, antioxidant, anticancer, anti-inflammatory, antimicrobial, antioxidant, and anti-tumorigenic effects. This study aims to determine the fingerprint profile of mangosteen seed extract and peel using variations of ethanol using FTIR and combined chemometrics with multivariate analysis methods. Mangosteen seeds and peels were extracted with a variety of solvents (ethanol 70%, and ethanol 96%) by maceration obtained 8 types of extracts. All extracts were analyzed through FTIR spectroscopy with KBr pellets and solvent methanol PA at wavelengths 400-4000 cm-1. Mangosteen peel and seed extracts have functional groups with fingerprint markers at wavelength positions 3414 cm-1 (O-H), 2964.59 cm-1 (C-H₃), 2920.23 cm-1 (C-H₂), and 2856.58 cm-1 (C-H₂). Multivariate analysis showed loading plot of extract had normally distributed and 96% seed Bulukumba extract had a major effect on PC1 (0.424) and 96% ethanol peel Bulukumba extract (0.439). The plot score affected PC1 (20.1%) and PC2 (57.4%) with a similarity between 31.57-99.89% in the extracts. The correlation of each extract showed a very positive correlation with the equation (y = 0.9805x + 65.177; R² = 0.9926). The use of FTIR spectra in observing the fingerprint profile of mangosteen peel and seed extracts with ethanol variations and chemometric combinations is highly applicable and can be applied in the validation and study of the chemical components of plant extracts.