Effects of solvent, water activity and temperature on lipase and hydroxynitrile lyase enantioselectivity

The influence of the reaction conditions on the enantioselectivity of reactions catalysed by lipases or hydroxynitrile lyases (HNLs) in organic solvents was investigated. The lipases catalysed kinetic resolution of chiral secondary alcohols or chiral carboxylic acids and the HNLs catalysed asymmetric addition of hydrogen cyanide to aldehydes. The temperature effects on enantioselectivity were studied in detail. From measurements of the enantiomeric ratio (E) at different temperatures the activation parameters ΔΔH# and ΔΔS# were determined. In the lipase-catalysed reactions the enthalpic and entropic effects on E always counteracted, while in a few of the HNL-catalysed reactions, ΔΔH# and ΔΔS# had opposite signs and therefore the effects cooperated to give high E values (−RTlnE = ΔΔG# = ΔΔH# − TΔΔS#). In all the HNL-catalysed reactions and most of the lipase-catalysed ones, the enantioselectivity increased with decreasing reaction temperature. However, in one of the lipase-catalysed reactions, the enantioselectivity decreased with decreasing temperature. The theoretical background of these observations was discussed. In the HNL-catalysed reactions, the enantioselectivity increased with increasing water content up to water saturation, while in the lipase-catalysed reactions the opposite trend was found in one case and in the others no significant effect was observed. Solvent mixtures of diisopropylether and hexane were used to obtain solvents with different log P values. The log P value of the solvent did not influence the enantioselectivity in the HNL-catalysed reactions, while the enantioselectivity increased with increasing log P value in two of the lipase-catalysed reactions. The reaction temperature was shown to be a very useful way to influence enzyme selectivity and the effects obtained could be rationalised. The influence of the reaction medium (solvent and water activity) is much more difficult to rationalise and predict.