Near-infrared spectroscopy and HPLC combined with chemometrics for comprehensive evaluation of six organic acids in Ginkgo biloba leaf extract

Abstract

Objectives: To investigate the feasibility of using near-infrared spectroscopy for rapid determination of main organic acids in Ginkgo biloba leaf extract (EGBL).

Methods: Main organic acids in EGBL were assayed using the HPLC method. Critical factors of the chromatographic separation were optimized by a novel analytical quality by design approach. Partial least squares-discriminant analysis (PLS-DA) was performed to screen the marker components, and principal component analysis (PCA) was utilized to distinguish the different samples. Then, spectral quantification potential was investigated using PLS and support vector machine (SVM) approaches. For modelling, different spectral preprocessing and wavelength selection methods were systematically compared.

Key findings: It was found that quinic acid, protocatechuic acid and 6-hydroxykynurenic acid were identified as possible index components. PLS-DA based on contents and PCA based on near-infrared spectra can both effectively distinguish the different EGBL samples. The calibration models with wonderful prediction performance can be both developed by the PLS and SVM algorithms.

Conclusions: NIR spectroscopy combined with chemometrics can realize the rapid and non-destructive qualitative and quantitative analysis of EGBL. The proposed method may be applied to quality control of EGBL and other natural products in commercial use.

Keywords: Ginkgo biloba leaf extract; near-infrared spectroscopy; partial least squares; quality evaluation; support vector machine.