Prediction of Copra Chemical Content with FT-NIR Spectroscopy Using PLS

Abstract

Copra is a coconut derivative product used as a raw material for making coconut oil and its derivatives. The chemical content of copra is determined using chemical method that damage the material, take a long time, and use chemicals. Therefore, it is necessary to have another alternative in determining  the chemical content of copra, which is non-destructive, fast and does not contain chemicals. The purpose of this study was to examine the FT-NIRS method as a rapid method for predicting the water  and free fatty acids (FFA) content of copra. The reflectance of 45 whole copra samples was measured using the NIRFlex N-500 spectrometer at the wavelength of 1000-2500 nm. After that, the water and free fatty acid content of copra were determined using chemical method. Several spectral data processing (SNV, MSC, Normalization, and OSC) were carried out then the processed spectra were calibrated with chemical data using PLS. The best prediction for copra water content is using SNV pretreatment with 12 PLS factors (r = 0.97, SEC = 0.70%, SEP = 0.73, CV = 9.08%, RPD = 3.41 and consistency 95.75%). Meanwhile, the prediction of free fatty acids (FFA) content is using OSC pretreatment with 13 PLS factors ( r = 0.90, SEC = 0.04%, SEP = 0.05%, CV = 24.45%, RPD = 2.05 and 71.78% consistency). These results indicate that the water  and FFA content of copra can be predicted  by FT-NIR spectroscopy.