Vibrational spectroscopy and chemometric data analysis: the principle components of rapid quality control of herbal medicines

Researchers are constantly exploring alternative methods to ensure the quality of raw plant materials and herbal products. Conventional quality control techniques may fall short on speed of analysis. Vibrational spectroscopy provides rapid results, uses no solvents and is non-destructive. The application of vibrational spectroscopy and hyperspectral imaging (HSI) in combination with chemometric data analysis as a quality control method was demonstrated using several examples: 1) species differentiation; 2) biomarker quantification and 3) percentage composition. Spectral data was acquired in the mid-infrared (4000-500 cm^-1) and near infrared (10 000-4000 cm^-1) wave regions and hyperspectral images were acquired in the shortwave infrared region (920-2514 nm). The spectral data obtained was processed using chemometric data analysis techniques.1) Differentiation between closely related species was demonstrated for mid-infrared (MIR) spectral data in the case of powdered Agathosma betulina vs. Agathosma crenulata leaves as well as for whole fruits of Illicium anisatum vs. Illicium verum using hyperspectral imaging; 2) Calibration models (R2≥0.75) based on spectral data were developed for seven major compounds identified in Melaleuca alternifolia essential oil; The biomarker Sutherlandioside B (SU1) was quantified in external samples of powdered Sutherlandia frutescens leaves using the calibration model (R2>0.95) developed; 3) The hyperspectral imaging model developed to determine percentage tea blend composition (Aspalathus linearis/Agathosma betulina) in intact teabags had an R2X_cum of 0.767 and Q2_cum of 0.932 showing good prediction ability. The results showed that these techniques have great potential to be implemented as non-destructive quality control methods depending on the application and the desired accuracy.