Development of a portable Raman and near infrared spectrometer for quality assessment of medicinal plants

Abstract

Abstract Medicinal plants are the integral components of alternative medical care in the world and the herbal medicines manufactured from these plants serve the primary healthcare needs of the majority of the world’s population. There are varied problems in this sector primarily due to absence of knowledgebase on the chemotypic diversity of medicinal plants. One of the most important attributes of these plant species is the chemotypic variation with respect to the major bioactive therapeutic compounds and the most important synergistic compounds that enhance the potential of the primary compounds. The conventional methods of measurement using High-performance liquid chromatography (HPLC), Thin-layer chromatography (TLC), Liquid chromatography–mass spectrometry (LCMS), and Gas chromatography–mass spectrometry (GCMS) are very accurate and precise, but are complex, expensive and requires skilled manpower as most of the measurements are based on chemical methods and lab based. It cannot be used in the field or in the industry for day-to-day use. In this respect, the portable measurement systems based on Raman and Near Infrared (NIR) Spectroscopy are promising and, if tuned properly, can be used for many such applications. The basic goal of the thesis is to develop sensor systems for detection and quantification of the major active groups of compounds in a few medicinal plants such as Andrographis paniculate (Kalmegh), Piper nigrum (black pepper) plants. The design and development of a portable Raman and NIR spectrometer with customized GUI is presented in this research work. The integrated Raman spectrometer consists of laser power supply, optical head and detector and a customized GUI was developed to estimate the piperine content in black pepper for assessing its quality. The portable NIR has a light source and optical components and detector, and a GUI for the quantitative estimation of andrographolides in Andrographis paniculata for its gradation and quality assessment In this work, a methodology is proposed for determining the content of piperine in black pepper berries using Raman spectroscopy. In this study, we present a simple, rapid and green analytical method based on Raman spectroscopy for the quantitative assessment of piperine. To assess the potential of the technique, we report the complete vibrational characterisation of the piperine with density functional theory (DFT) calculations. The theoretical peaks were obtained at 1097 cm-1, 1388 cm-1, 1528 cm-1,1578 cm-1, and at 1627 cm-1, and this result was verified in a Raman spectrometer followed by a preliminary experiment. Twenty black pepper samples were analysed using high-performance liquid chromatography (HPLC) and used as reference data for Raman analysis. The Raman shift spectra were analysed using partial least squares (PLS) and good prediction accuracy with correlation coefficient of prediction (Rp2)= 0.93, root mean square error of prediction (RMSEP) = 0.13 and residual prediction deviation (RPD) = 3.9 obtained. The results demonstrate the efficacy of the Raman technique for the estimation of piperine in the dry fruit of Piper nigrum. Thus, this method can be suitably applied for in situ quality detection of inward medicinal plant leaves in medicinal plant industry. In this study, a methodology has been proposed to grade Andrographis paniculata samples using NIR spectroscopy and SVM classifier based on the content of marker molecules, andrographolides. Relative accuracies of estimating the andrographolides by NIR spectra of methanol extracts of the samples, and powdered leaf samples were compared taking the estimates obtained from HPLC analysis as the standard. The accuracy of estimation based on extracts was a little higher than the powder leaf samples. But it did not change the grading pattern of the samples. Support vector machine was used to grade the samples into three classes–Class I (best quality), Class II (intermediate quality) and Class III (poor quality). The average classification accuracy of tenfold cross validation of SVM was obtained as 83%. Thus, NIR based estimation of powdered leaf samples combined with SVM classifier can be a low-cost solution to grade the samples rapidly. A small range portable NIR instrument would serve as a field-lab deployable instrument for gradation of Andrographis paniculata samples by the industry.