Development of calibration model for determination of sweeteners additives in Indonesia rice flour-based food by FT-NIR spectroscopy

Abstract

Cane sugar and other artificial sweeteners are usually used as a food additive to provide sweetness in food. Sweetener additives can be consumed safely by considering the acceptable daily intake (ADI). This research was aimed to determine the type of sweeteners and the level of sweeteners added in food with regard to the ADI. The food sample used in this study was geplak, traditional Indonesian food, made of rice flour, coconut, and sugar, or other sweeteners. The reflectance of geplak powder was measured using the NIRFlex N500 Fiber Optic Solids Cell at 4000 - 10,000 cm-1. The reflectance spectra obtained were pre-treated and analyzed using The Matlab version R2018a. Principal component analysis (PCA) and partial least square discriminant analysis (PLS-DA) were used for data exploration and qualitative classification. PCA model was able to classify food added with sugar, saccharin, and cyclamate. PLS-DA calibration model using the 2nd derivative Savitzky-Golay as the spectral pretreatment achieved 100% accuracy in predicting high sucrose and high saccharin, as well as low cyclamate and low saccharin, while achieved a slightly low accuracy of 90.0 and 85.7% in predicting low sucrose and high cyclamate, respectively. PLS-DA prediction model, which also applies the 2nd derivative Savitzky-Golay pretreatment spectra, achieved 100% accuracy in predicting high and low sucrose, low cyclamate, and high saccharin.