[Optical resolution of 2-alkanol by lipase-catalyzed acetylation with vinyl acetate in packed-bed reactor with recycling system].

The lipase-catalyzed acetylation of 2-alkanol with vinyl acetate was studied using Burkholderia cepacia lipase (BCL), three alcohol and three organic solvents in a packed-bet reactor with a recycling system (flow method). The optical resolution data were found in agreement with those of the batch method in which BCL was suspended in the substrate solution. Repeated reaction results clearly indicated BCL in the packed-bed to be quite stable and to be usable for at least 50 reaction runs or to remain effective for as long as two months in the water-insoluble solvents such as hexane and 1,2-dichloroethane. In the reaction using a water-soluble solvent such as acetonitrile, the catalytic power of BCL showed only a 1% decrease of conversion per run or solvent recycling possibly owing to compression of BCL in the bed although enantioselectivity was independent of the number of reaction repetitions. The present method showed thus be applicable to kinetic resolution by enzyme-catalyzed acylation in hydrophobic organic solvents with no waste of enzyme.

Kinetic studies on lipase-catalyzed acetylation of 2-alkanol with vinyl acetate in organic solvent.

Lipase-catalyzed acetylation of 2-alkanol with vinyl acetate has been studied kinetically using Burkholderia cepacia lipase (BCL), enantiomerically pure (R)- and (S)-2-alkanols and different organic solvents. The rate equation was derived by the steady state method for the simplified mechanism. The second order rate constants (k(R) and k(S)) for (R)- and (S)-2-alkanols were evaluated from the slopes of the double reciprocal plots, v(-1) vs. [2-alkanol](-1), where v is the initial rate of the reaction. The log k(R) value increased with the solvent hydrophobicity log P, where P is a partition coefficient of a given solvent between octanol and water. The log k(S) value also increased with log P except the bulky solvents such as 1,4-dioxane and cyclohexane, in which the rates were faster than those expected from the log k(S) vs. log P plot. The slope of log k(S) vs. log P plot was larger than that for (R)-2-alkanol. Thus, log E (E=k(R)/k(S): enantioselectivity) decreased with log P except the bulky solvents. The rate constants and the enantioselectivity were different depending on the structure (carbon number CN) of 2-alkanol. The log E vs. CN plot was minimized at CN=8 and 10 and the log k(S) vs. CN plot maximized at CN=8 and 10. In contrast the log k(R) vs. CN plot showed a different feature from the log E vs. CN plot. These facts suggest that dependence of E on CN is more strongly affected by the reactivity of (S)-2-alkanol than that of (R) isomer in this acetylation.