In situ lignocellulolytic enzyme production by solid state fermentation

Résumé

Experimental plans were designed in order to develop solid state fermentation (SSF) for in situ lignocellulolytic enzyme production. The enzyme extract will be used to increase biodegradation and methane production in downstream anaerobic digestion (AD). In that perspective, four lignocellulolytic enzymes were identified of interest: total cellulase (FPase), carboxymethylcellulase (CMCase), ß-glucosidase and xylanase. Their corresponding concentrations or activities were monitored all along the conducted research. Following an extensive literature review, substrate type, autoclaving, inoculum type, pH, moisture and nutrient addition were identified as the factors that impact enzyme production under SSF conditions at the laboratory scale. A scale-down of the enzyme activities analytical methods to 96-well plate test format was however first proven necessary. Analysis had to adapt to large factorial experiments. Two consecutive sets of experiments were run focusing on the development of the SSF process using three substrates: municipal solid waste (MW), paper/cardboard (PC) and brewer¿s spent grain (SG). First the factors impacting the SSF process and their corresponding ranges were narrowed down using a D-optimal design experimental plan over a 5-day fermentation period. A full design experimental plan later optimized the experimental conditions. Humidity level and pH were the most significant impacting factors. Their effect was non-linear over the tested range. Since the optimal enzyme mixture could not be defined, all three substrates were kept in the different experiments. Using the identified optimal conditions, a follow-up over two weeks of triplicates of the different substrates identified the production profile of the four enzymes. The above defined fermentation time falls well within the optimal range. These assays also put forward the importance of the indigenous microflora which could substitute external inoculum. Finally, brewer's spent grain was used as a model substrate to study the impact of two enzyme samples, one commercial as a reference and the enzyme extract obtained from SSF of the three substrates, on methane production in three lab-scale AD systems: single-stage, two-stage and a percolation system coupled with AD. Hydrolysis results were encouraging but the impact on methane production using the experimental setup was shown on the third tested system. At least a 40% increase of methane production can be achieved using enzymes.