Rapid determination of rebaudioside A content in the macroporous resin elution process using a portable near-infrared spectrometer

Rebaudioside A (RA) is an ideal sweet component extracted from leaves of Stevia rebaudiana Bertoni and has been widely used in the food and pharmaceutical fields. Macroporous adsorbent resin is often used to enrich RA due to its low cost, easy regeneration, and high stability. However, the traditional method for the determination of the elution process end point often relies on personal experience, which has hysteresis and ambiguity and seriously restricts the broad development prospects of stevia. To develop a new method for rapidly and accurately determining the RA content in the macroporous resin elution process, portable near-infrared (NIR) spectroscopy combined with chemometrics was used in this study. The linear partial least squares regression (PLSR) model and the non-linear support vector machine regression (SVMR) model were used to establish the relationships between the NIR spectrum and the RA content. The results of model fitting showed that the PLSR model displayed higher predictive accuracy of the RA content than the SVMR model with a determination coefficient external test set (Rp2) value of 0.9796, a root mean square error external test set (RMSEP) value of 2.7555 and a residual prediction deviation (RPD) value of 6.5566. This study demonstrates that NIR spectroscopy techniques are potential tools for rapidly determining the RA content in the macroporous resin elution process, which will save time, effort, and cost in the analysis of large quantities of unknown samples. Portable NIR spectroscopy is used to monitor the macroporous resin elution process. The PLSR and SVMR models displayed good performance for determining the RA content. The PLSR model displayed higher predictive accuracy than the SVMR model.