An analysis of published industrial enzymatic and microbial biotransformation processes showed that in about 10% of the cases in-situ product recovery has been implemented. However, for de-novo fermentations it is hard to find industrial examples. Several explanations can be given. In de-novo fermentations (as opposed to biotransformations) removing product from a reaction equilibrium is not an issue. Also, any product degradation might be minimized by knocking out undesired enzyme activities rather than by in-situ product removal.

We will show some of our recent work on recovering inhibiting fermentation products (L-phenylalanine, ethanol, 1-butanol, fumaric acid, and succinic acid). To facilitate fermentation, we considered capturing these inhibiting products in a loop around the fermenter by methods like adsorption, crystallization and pervaporation. Experiments and techno-economic evaluations were performed. In general the preferred process option involves: 1) feeding such that not much nutrient leaves the fermenter; 2) cell removal or retention; 3) complete product recovery. Then, the remaining aqueous stream contains impurities only and should not be recycled to the fermenter. The main integration issue is to find fermentation conditions such that the recovery is not compromised too much. Metabolic engineering for minimizing impurities can be very helpful for this.