Published on Web 02/04/2006
N-Sulfonyloxy Carbamates as Reoxidants for the Tethered
Aminohydroxylation Reaction
Timothy J. Donohoe,*,† Majid J. Chughtai,† David J. Klauber,† David Griffin,‡ and
Andrew D. Campbell§
Department of Chemistry, Chemistry Research Laboratory, UniVersity of Oxford, Mansfield Road,
Oxford, OX1 3TA, U.K., Syngenta, Jealott’s Hill International Research Centre, Bracknell, Berks, RG42 6EY, U.K.,
and AstraZeneca Pharmaceuticals, Mereside, Alderley Park, Macclesfield, Cheshire, SK10 4TG, U.K.
Received October 29, 2005; E-mail: timothy.donohoe@chem.ox.ac.uk
Scheme 1
The Sharpless aminohydroxylation reaction has proven itself to
be an extremely useful method of oxidizing alkenes to form vicinal
amino alcohols. The scope of this chemistry is greatly enhanced
by the fact that the reaction is stereospecific (strictly syn addition)
and also capable of forming single enantiomers of amino alcohol
products when performed in the presence of cinchona alkaloid lig-
ands.1 Our contribution to the development of this reaction was to
address the regioselectivity problems that arose when unsymmetrical
alkenes were oxidized. We recognized that tethering the nitrogen
souce for an aminohydroxylation (such as a carbamate) to an allylic
Scheme 2
alcohol would allow an intramolecular aminohydroxylation to ensue
with complete control of regioselectivity (and high levels of
stereoselectivity when chiral allylic carbamates were subjected to
the tethered aminohydroxylation, TA, reaction; Scheme 1).2
The reoxidant for an aminohydroxylation reaction is typically
an N-halocarbamate salt prepared in situ by the action of NaOH
and tBuOCl on a primary carbamate (Scheme 1).3 In our hands,
we found that the N-chlorocarbamates thus produced have a limited
lifetime in the reaction. In an AA reaction, one can compensate
for this by using 3-3.5 equiv of carbamate (and NaOH and
tBuOCl). Clearly, we do not have this luxury in the TA reaction
and thus we sometimes have difficulty in driving the reaction to
gave a yield of 43% of 2 in the TA reaction. Of course, there is no
need to add a chlorinating agent to this TA reaction, and
consideration of the likely pKa of the NH proton7 within 1 led us
to run the hydroxyamination not just without tBuOCl but also
without any hydroxide base.8 Pleasingly, the reaction worked as
planned and gave 70% yield of oxazolidone 2 (Scheme 3).
completion.
Furthermore, in detailed studies on the TA reaction we discovered
that, occasionally, chlorination of the alkene unit was a competing
Next, a range of allylic alcohols was converted to the corre-
side reaction (promoted by tBuOCl) that was responsible for
sponding N-sulfonyloxy derivatives and then subjected to the
lowering the yields. This problem is especially pronounced with
chlorine-free TA reaction conditions (Scheme 3). In each case, the
homoallylic carbamates, and this substrate class is barely compatible
product of the reaction was identical to that obtained by conven-
tional TA oxidation of the corresponding carbamate, but the yields
with the TA conditions.
We have just completed an investigation into alternative reoxi-
were always higher. All of the main classes of allylic alcohol were
dants for the TA reaction with our efforts centered around
examined, showing that the reaction works well for derivatives of
achiral and chiral (both cyclic and acyclic) allylic alcohols.
The new regimen is much easier to perform and has none of the
replacement of the chloride leaving group on the N-halocarbamate
salt.4 It soon became apparent that N-sulfonyloxy derivatives were
superior reoxidants for this process. Recently, Lebel et al. have
complications that arise when excess chlorinating agent is present
shown that NHOTs derivatives of carbamates are effective nitrene
in situ. As a testament to the extra simplicity of the reaction, it
could be run at much lower catalyst loadings (1 mol % Os) and
still gave acceptable yields of products; in our experience, this low
precursors for C-H insertion and alkene aziridination.5
The requisite N-sulfonyloxy carbamates were readily prepared,
in generally good yields, in a two-pot procedure by sequential
catalyst loading is well out of range of a typical TA reaction.9
reaction of an alcohol with CDI and hydroxylamine, followed by
Extra information on the mechanism of reaction was gleaned
by conducting a reaction on homoallylic alcohol derivative 11 using
both catalytic and then stoichiometric potassium osmate (Scheme
4).
As expected, the catalytic reaction gave six-membered carbamate
12, which was identical to that formed from the corresponding
ROCONH2 derivative, but with a yield greatly increased from 16
sulfonylation (Scheme 2). While we discovered that a variety of
O-sulfonyl groups (e.g. OTs) were generally efficient as leaving
groups in the TA reaction, we chose to use mesitylsulfonyl as a
common standard.
Our studies into the aminohydroxylation reaction commenced
with 1 (Scheme 3). Note that the corresponding carbamate of 1
to 63%. Performing a reaction with 1 equiv of osmium, in the
† University of Oxford.
presence of TMEDA,2b allowed us to isolate osmium azaglycolate
‡ Syngenta, Jealott’s Hill International Research Centre.
§ AstraZeneca Pharmaceuticals.
13 in 36% yield (syn addition across the alkene was proven by
9
2514
J. AM. CHEM. SOC. 2006, 128, 2514-2515
10.1021/ja057389g CCC: $33.50 © 2006 American Chemical Society