Coconut sugar is a local sugar from the blossoms of a coconut tree. It has been considered
a healthy sugar due to its low glycemic index. There is an attempt to add other sugar to it
to lower the cost. Thus, this research aimed to identify Thai coconut sugar and to establish
models for predicting the moisture content of coconut sugar by using FT-NIR
spectroscopy. Thai coconut sugar samples were purchased from local grocery stores in
four provinces, online, and the community market. Their moisture contents were varied
and equilibrated for 24 hrs prior to the measurements of moisture and FT-NIR spectra.
The results showed that FT-NIR spectra of Thai coconut sugar differ from sucrose,
glucose and fructose at the absorbance spectrum of 5379-5011 cm-1
. FT-NIR spectroscopy
of 54 known moisture samples of Thai coconut sugar was used to obtain a model to
predict moisture content. The predicted equation, using the PLS technique with the
Spectrum Quant program, was found to give a standard error of prediction (SEP) 0.077%
(less than 0.10%), indicating a non-destructive method of accurately and precisely
predicting moisture levels in the coconut sugar. The results obtained suggested that FTNIR spectroscopy has the potential to be used as a tool to identify Thai coconut sugar
accurately. It can rapidly predict the moisture content in the sample which will be useful
in quality control standards.