Abstract
Results indicate that sufficient information is present in near-infrared (near-IR) spectra covering the range from 5000 to 4000 cm−1 to monitor the acid-catalyzed hydrolysis (thinning) of starch. Spectra taken from the aqueous fraction of samples collected during actual thinning reactions display information corresponding to soluble carbohydrates. The observed absorbance bands increase in magnitude according to the extent of the thinning reaction. A digital Fourier filter is applied to isolate the carbohydrate band centered at 4400 cm−1. This filtering step effectively removes choppy noise from the raw absorbance spectra and improves the measurement reproducibility by 3.6-fold. A third-order polynomial is used to empirically relate the integrated area under this 4400-cm−1 absorption band to the extent of starch thinning as measured as fluidity of the resulting starch paste. This model is able to accurately predict fluidities with an overall standard error of prediction of 1.94 mL for fluidity values ranging from 22 to 67 mL.
Get full access to this article
View all access options for this article.
