Batch tests of biological hydrogen production from food industry wastes by four thermotoga thermophilic strains in 0.12-L microcosms and in a 19-L fermentor

This work is aimed at evaluating the feasibility of a cost-effective process of biological H2 production from food industry wastes under thermophilic conditions. The H2-producing performances of 4 Thermotoga strains (T. neapolitana, T. petrophila, T. naphtophila, T. maritima) were compared at 77 °C by means of tests conducted in 120-mL batch bioreactors containing a nutrient-rich growth medium (ATCC 1977; Van Ooteghem et al., Biotechnol. Lett., 2004, 26:1223) additioned with glucose, molasses or cheese whey as carbon source. For all the substrates tested, T. neapolitana resulted the best-performing strain under suspended-cell conditions, with a 0.9-1.9 mmol / L medium / h H2 production rate at an 8-10 g/L initial substrate concentration, and a 1.6-2.6 mmol H2 / mmol monosaccharide consumed yield. To compare the 4 strains also under attached-growth conditions, preliminary tests were conducted with glucose-growing T. neapolitana, with the goal to select the best biomass carrier among 4 porous materials utilized in biofiltration applications. The best results were obtained with the carrier characterized by the highest interfacial area, equal to 2.1 m2 g-1. Also under attached-growth conditions, T. neapolitana resulted the best strain for all the 3 substrates tested, with a 1.2-1.8 mmol / L medium / h H2 production rate. Further tests, aimed at simplifying the growth medium composition, led to encouraging results. For example, in the case of molasses a minimum medium composed only by NH4Cl, K2HPO4, NaCl, buffer and cysteine resulted – in comparison with the ATCC 1977 complete medium – in a 73% reduction of medium cost and in a 12% increase of the H2/substrate yield. Finally, the process scale-up to a 19-L reactor is in progress. The preliminary results indicate that, in the scale-up of batch H2 production from glucose by T. neapolitana, the values of the three monitored yields (H2/substrate, organic acids/substrate, biomass/substrate) did not show significant variations, whereas the H2 volumetric production and the initial H2 production rate resulted to be significantly affected by the concentration of the initial inoculum