Prediction of Rheological and Chemical Properties of Different Starches Used in the Paper Industry by Near Infrared Spectroscopy (NIRS)

Starch is one of the most abundant biopolymers, consisting of the polysaccharides amylose and amylopectin and other minor but characteristic constituents depending on the origin. Potatoes, maize, wheat, rice and tapioca are the sources of commercial interest with a large variety of applications both in food and non food areas. The measurements of over hundred starch samples were accomplished over a range of 4000–10000 cm^?1 in diffuse reflectance and transmission mode. Quantitative prediction models, based on a partial least squares (PLS) algorithm for the determination of fat content, the main fatty acids, water content, tendency to retrogradation and viscosity were developed. All models are characterized by high correlation coefficients (R² > 0,98). Principal components analysis (PCA) was applied successfully to distinguish starches regarding their source, provenience and their degree of modification. Both quantitative and qualitative models are cross validated.