In‐Silico Optimization of a Bi‐Enzymatic Reactor for Mannitol Production Using Pareto‐Optimal Fronts

Abstract

For multi‐enzymatic cases, the determination of the batch reactor (BR) optimal operating policy often translates into a difficult multi‐objective problem. Exemplification is made here for the enzymatic reduction of D‐fructose to mannitol by using the mannitol dehydrogenase (MDH) enzyme and nicotinamide adenine dinucleotide (NADH) cofactor, with in situ regeneration of NADH at the expense of formate degradation by using the FDH enzyme. This paper presents an original rule to in silico generate the problem solution, by using the Pareto optimal‐front approach with accounting for pairs of competing economic goals and constraints. The optimal BR is then compared to an optimal fed‐BR (FBR), or a series of equal BRs (SeqBR). As proved, the Pareto‐optimal front alternative is an advantageous option, compared to the classical nonlinear programming technique, being simple to apply, by considering pairs of opposite objective functions. In the present case study, the Pareto‐optimal BR operating mode predicts an M‐productivity 1.5x better than those of an optimized FBR, with comparable enzymes consumption. The MDH consumption of this Pareto‐optimal BR is 10x smaller than an optimal SeqBR, and 130x smaller vs. heuristic (sub)optimal BR.