The diastereoselectivity of zirconium alkoxide catalysed Meerwein‐Ponndorf‐Verley reductions

Abstract

The new variation of the Meerwein-Ponndorf-Verley reduction using 1 -[ 4 -(dimethy1amino)phenyll ethanol (DMAPE) (6) or 1-tetralol (12) (3 equiv.) as the reducing alcohols and Zr(0-tBu)., as the catalyst (0.2 equiv.) is kinetically controlled and highly stereoselective. Preferential axial attack is achieved with the sterically less bulky alcohol 6 in the case of 4-ferfbutyl-and 3-methylcyclohexanones ( l a --$ 96 % axial and 3a --f 93 % axial attack). The combination tetralol (12)/Zr(OtBu)4 behaves as a very bulky reducing agent, and the thermodynamically less stable alcohols 2c (80%) and 4c (92-96%) are formed preferentially in the reduction of 2a and 4b. The fused bicyclic systems 1-methyl-2-tetralone (14a) and flavanone (15a) and the steroids 16a-18a are reduced with high stereoselectivity to the corresponding cis alcohols 14b and 15b and the P-alcohols 16b-18b. The stereochemical outcome is in agreement with the Felkin-Ahn model.In a preceding communication we described a new operationally simple variation of the Meerwein-Ponndorf-Verley (MPV) reduction [']. The main features were the application of a substoichiometric amount (in most cases 20%) of zirconium catalyst z r ( O -t B~)~, the use of alcohols of high reduction potential [e.g. 1-[4-(dimethylamino)phenyl]ethanol (DMAPE) (6); 1-tetra101 (1211 in moderate excess (2-4fold), and reaction in unpolar solvents (eg. CC14, cyclohexane, or toluene) at ambient temperature and with short reaction times. The mild conditions and high yields of the new variant offer considerable advantages in the reduction of sensitive substrates. Furthermore, the deliberate choice of a favorable redox potential decreases the rate of the reverse reaction (Oppenauer oxidation) making this process kinetically controlled (see below) in contrast to the thermodynamically controlled conventional MPV reductions (for reviews see ResultsReduction to stereoisomers can give experimental evidence for kinetic reaction control if the ratio of the isomers is remote from equilibrium, and if this ratio is maintained unchanged over the course of the reaction. In our initial experiments we have thus followed the stereochemical course of a number of reductions using different substrates and reducing alcohls. 4-tert-Butyl-(la), 2-methyl-(2a), 3-methylcyclohexanone (3a), and 3,3,5-trimethylcyclohexanone (4a) were selected as substrates, DMAPE (6) and 1-tetralol (12) as reducing alcohols, and Z r ( 0 -t B~)~ (20%) as the catalyst in these experiments. The results are summarized in Figure 1 and show the small changes in the isomeric ratio of the cis-and trans-4-tert-butylcyclohexanols l c and l b (ratio ca. 4:96; equilibrium ca. 20 : 80U5l), the cis-and trans-2-met hylcyclohexanols 2c and 2b (ratio ca. 80:20; equilibrium 1 :99r61, and the cis-and trans-3-methylcyclohexanols 3b and 3c (ratio ca. 4:6; equilibrium 77:23 [']). Small deviations (in some cases with the opposite direction) are only seen in the initial phase of the reaction. This effect can be explained by exchange of tert-butyl alcohol with t...