The Sharpless asymmetric epoxidation of allylic alcohols

Original Web page: http://www.cmbi.kun.nl/~borkent/projects/sharpless.html

What makes it asymmetric?

This discussion is based on the paper by Y.D. Wu and D. Lai, J.Amer.Chem.Soc. 117 11327 (1995), a DFT study on the selectivity of this reaction. The structure below is not the outcome of their calculations; it is a hand-made reconstruction of the complex they suggest as a transition state that explains the experimental findings.

The reagents in this reaction are: the catalysts titanium tetraisopropoxide, and enantiomerically pure diethyl tartarate (DET, either RR or SS), tert-butyl hydroperoxide (TBHP) as the oxygen source, and of course the allylic alcohol to be epoxidated.
It is proposed that these reagents form an asymmetric complex, in which the allylic alcohol exposes one side of its double bond towards the oxygen to be transferred.
Hydrogens on/off

Where are the components to be found?
The heart of the complex is formed by two titanium atoms, surrounded by six oxygen atoms each (two shared in a Ti-O-Ti-O 'square'). (Such an octahedral dimer is often found in crystal structures, e.g. COPPIG.)
These oxygens originate from two remaining isopropoxides, the other ones being replaced by
the oxygens of the hydroxyl groups of two diethyl tartarate molecules , of which the green one not only donates its hydroxyl oxygens, but
one carbonyl oxygen as well.
Question: is this the RR (= L-(+) ) or the SS ( D-(-) ) enantiomer of DET?
The tert.butylperoxide (brown) occupies two oxygen positions at one Ti-atom.
In this asymmetric complex the allylic alcohol takes the last position, prefering a conformation in which one side of the double bond (greenblue) is facing the peroxide oygen.

Now a possible explanation of the enantioselectivity of this reaction:
The green coloured tartarate has the two ester groups in a kind of axial position: one takes part in the complexation of a titanium atom, and it is the other one that interacts with the tert.butyl group of the peroxide. (With methylhydroperoxide the reaction is not selective).
It pushes the tert.butyl in the direction of the blue tartarate.
This in turn has its effect on the position of the allylic system: the calculations show that its O-C bond prefers to be gauche with respect to (in between) the peroxide oxygen and the Ti-O-Ti bridging oxygen.
Because of this one face of the double bond forms a spiro-type of transition state with the oxygen to be transferred. This yields an enantiomeric epoxide:

All color functions are toggle buttons, and you can reset all atom colors and positions here.

With the other enantiomer of DET (imagine the ester groups and hydrogens interchanged) one should imagine that now the blue one participates in Ti-complexation. The positions of the t-butylperoxide and allyl system are mirrored, resulting in the opposite epoxide enantiomer.

Hens Borkent, CMBI, University of Nijmegen, The Netherlands