Non-destructive quality assessment of herbal tea blends using hyperspectral imaging

作者：Sandasi Maxleene; Chen Weiyang; Vermaak Ilze; Viljoen Alvaro^*

来源：Phytochemistry Letters, 2018, 24: 94-101.

DOI：10.1016/j.phytol.2018.01.016

• Cyclopia genistoides
• Sceletium tortuosum
• Herbal tea
• Hyperspectral imaging
• Chemometrics
• Quality control

The consumption of herbal teas is increasing as consumers become more appreciative of the health benefits. Herbal tea blends comprise of two or more plant species blended to improve taste and multiply the health benefits. Quality control (QC) of herbal teas like other nutraceuticals, is important to ensure safety and efficacy. Current QC methods are chromatography-based and require destructive sample preparation using solvents. In this study, hyperspectral imaging is applied as a fast and non-destructive method for the quality control of herbal tea blends. The technique combines conventional spectroscopy and digital imaging to gather chemical information and visualise spatial distribution of chemical constituents within a matrix. Certified raw materials (Sceletium tortuosum and Cyclopia genistoides) and herbal tea blends were acquired from Parceval Pty (Ltd). Hyperspectral images of the raw material and tea blends were captured on a SisuChema" SWIR (short wave infrared) hyperspectral pushbroom imaging system using ChemaDAQ (R) software. The images were analysed using Evince multivariate analysis software 2.4.0. Principal component analysis (PCA) revealed 54.2% chemical variation between S. tortuosum and C. genistoides raw materials. A partial least squares-discriminant analysis (PLS-DA) model with predictive ability of 95.8% was developed. Based on pixel classification, it was possible to visualise the tea blend constituents as S. tortuosum and C. genistoides and quantitatively predict C. genistoides as the major constituent (> 97%) while S. tortuosum was present in lower amounts (< 3%). The predictions confirm that HSI is a potentially favourable visual tool for the quality assessment of herbal tea blends. However, due to low instrument sensitivity quantitative determinations showed some deviation from the company formulation.

• 出版日期2018-4