Kinetic Resolution of Hydroperoxides with Enantiopure Phosphines: Preparation of Enantioenriched Tertiary Hydroperoxides

Link: http://pubs.acs.org/cgi-bin/abstract.cgi/jacsat/asap/abs/ja070482f.html

From Prof. Keith A. Woerpel's group at the University of California, Irvine

A new method for reductive kinetic resolution of tertiary hydroperoxide employing cyclophane diphosphine as a selective reducing agent was recently published in JACS ASAP.

Several phosphines were investigated as shown.
The test substrate was chosen as tertiary benzyl hydroperoxide 11. The results of the initial screening for the appropriate phosphine are shown in Table 1. As seen in the table, cyclophane-derived phosphine 10 showed the most promising result (entry 9). This commercially available phosphine was used to investigate the scope of the reaction with several benxyl tertiary hydroperoxides as shown in Table 2.
This phosphine worked well with all substrates, including secondary hydroperoxides (entries 7-10). Functionalized hydroperoxide (entry 6) could be resolved well, although selectivity was lower. In all cases, (R)-10 reduced the (-)-(S)-hydroperoxide preferentially, and the enantiomer, (S)-10, had the opposite selectivity.

Selectivities diminished with increasing length of the alkyl linker in the resolutions of non-benzylic hydroperoxides 23 and 24 with (R)-10 (Scheme 1). Presumably, as the tether length increases, the steric differentiation decreases at the reactive center.

Preparative-scale resolution was also possible. For instance, one gram of hydroperoxide (+/-)-11 was subjected to resolution conditions with commercially available phosphine (R)-10 (71% conversion, Scheme 2). The resulting enantiopure hydroperoxide (+)-(R)-11 and enriched alcohol (-)-(S)-12 could not be separated by physical means, but a strategy was developed to facilitate purification. When the mixture of hydroperoxide (+)-(R)-11 and alcohol (-)-(S)-12 was treated with Et3SiCl, the hydroperoxide was protected selectively, and the resulting silylperoxy ether could be separated from the alcohol by column chromatography. Subsequent desilylation provided enantiopure (>99% ee) hydroperoxide (+)-(R)-11 in 24% overall yield. This route also allows access to enantiopure tertiary alcohol (+)-(R)-12 by reduction with triphenyl phosphine (Scheme 2).

Preliminary mechanistic studies reveal that the two phosphines of xylyl-PHANEPHOS (10) operate independently. The supposed intermediate, mono(phosphine oxide) (R)-25, was isolated from the reaction of phosphine (R)-10 and 1 equiv of hydroperoxide 17. Utilizing this compound in the resolution of hydroperoxide 11 afforded starting material with 84% ee at 51% conversion (krel = 25, Scheme 3). This experiment demonstrates that the monophosphine intermediate (R)-25 reduces hydroperoxides with a similar selectivity to that of xylyl-PHANEPHOS. It also suggests that less complex monophosphines may also be useful for this type of resolution.

In conclusion, the authors have described a method for the stoichiometric kinetic resolution of hydroperoxides employing commercially available phosphines. The reaction provides access to enantiopure hydroperoxides and, therefore, the corresponding alcohols as well. In addition, the resulting bis(phosphine oxide) can be converted back to the phosphine in high yield (the bis(phosphine oxide) isolated from the resolution reaction can be reduced with HSiCl3 in >90% yield.)