Desenvolvimento e validação de modelos quimiométricos baseados em espectroscopia no infravermelho próximo para a predição de parâmetros do caldo e da biomassa do sorgo sacarino visando à produção de etanol

Abstract

The sweet sorghum juice is gaining importance as a raw material for the first generation ethanol production in the period between harvests of sugar cane. Furthermore, there is a great expectation on the use of sorghum bagasse for the lignocellulosic ethanol production, since this culture has a high yield of biomass per hectare and low production costs. Consequently, breeding programs are seeking to improve this raw material quality in order to increase productivity. This practice has generated an excessive number of samples to be analyzed. Thus, the aim of this study was to develop rapid and low cost methods based on multivariate calibration and near infrared spectroscopy for the determination of quality parameters of sweet sorghum biomass (cellulose, hemicellulose and lignin) and juice (brix, saccharimetric reading, reducing sugars, pol and apparent purity), besides the estimate of the theoretical ethanol yield (TEY) based on cell wall components (cellulose and hemicellulose). The biomass models were built using 957 samples, obtained from over 100 different genotypes, in the ranges of 21.4 to 49.1% w/w, 18.4 to 34.8% w/w, 221 to 412 Lt-1 and 1.8 to 11.5% w/w for cellulose, hemicellulose, lignin and TEY, respectively. The juice models were built with 500 samples, part of them from 275 contrasting strains for quality and for amount of sugars in the juice, in the ranges of 5.5 to 18.1 °B for brix, 1.1 to 53.2 °Z for saccharimetric reading, 1.2 to 5.2 % w/v for reducing sugars, 0.3 to 13.0 % w/v for pol and 9.8 to 83.0% w/v for apparent purity. Finally, a complete multivariate analytical validation was carried out in accordance with the Brazilian and international guidelines, and the methods were considered accurate, linear, sensitive and unbiased.