Acidogenic Fermentation of Sugarcane Vinasse and Glucose Using Two Different Anaerobic Digestion Inocula

Abstract

Vinasse is an agroindustrial wastewater obtained at large amounts in countries such as Brazil, which is one of the world greatest sugar and ethanol producers from sugarcane. Despite its most common use is as nutrient for fertirrigation in sugarcane crops nowadays, alternative processing has been intensely sought in last decades for vinasse aiming primarily the production of energy carriers and high-added value chemicals. One of these alternatives is the anaerobic digestion promoted by bacteria, which results in several products of economic importance. In view of these information, this work presents an investigation of acidogenic fermentation by C. acetobutylicum ATCC 824 and C. beijerinckii ATCC 25752 was conducted in batch reactors by using synthetic vinasse and aqueous glucose solution as substrates. Evaluation of effects of microorganism and carbon source was focused mainly on the production of short chain organic acids (SCOAs), although other metabolites, such as ethanol, were also detected. Carbohydrates were effectively consumed by bacteria, so remaining concentrations of sucrose and glucose (Suc and Glu) in both substrates were considerably low at the end of tests. Lactic acid (HLa) was the prevailing metabolite in all experiments, determining the trends observed for total SCOAs content as a function of incubation time, while other acids were generated in much lesser amounts. Moreover, ethanol (EtOH) attained notably higher concentrations in batch reactors containing glucose as substrate, which showed more relevant production of this chemical. Maximum of 14.1 g HLa L-1 was quantified for lactic acid at incubation time 96 h in reactor using synthetic vinasse and C. beijerinckii (VB), linked to a productivity P=146.46 mg HLa L-1 h-1 and a yield Y=1108.03 mg HLa g Suc-1. For ethanol, maximum was 1.6 g EtOH L-1, occurring in reactor containing glucose and C. acetobutylicum (GA) also at incubation time 90 h, resulting in P=13.60 mg EtOH L-1 h-1 and Y=73.04 mg EtOH g Glu-1. Finally, kinetic models were adjusted to experimental data of carbohydrate content, lactate concentration and optical density (indicative of microorganism growth), showing good prediction capability as suggested by values of fit quality parameters such as Residual Sum of Squares (RSS) and R2.