Communications
Asymmetric Rearrangement
DOI: 10.1002/anie.200601731
Practical, Highly Active, and Enantioselective
Ferrocenyl–Imidazoline Palladacycle Catalysts
(FIPs) for the Aza-Claisen Rearrangement of
N-para-Methoxyphenyl Trifluoroacetimidates**
their preparation requires two ortho lithiations. Moreover, in
the case of N-para-methoxyphenyl trifluoroacetimidates, the
catalyst loading had to be as high as 5–7.5 mol% (that is, 10–
15 mol% PdII) for both 1 and 2, while long reaction times
were still required in order to obtain preparatively useful
conversions.
Matthias E. Weiss, Daniel F. Fischer, Zhuo-qun Xin,
Sascha Jautze, W. Bernd Schweizer, and RenØ Peters*
Dedicated to Professor Dieter Enders
on the occasion of his 60th birthday
Our goal was to develop a practical, highly active catalyst
for the asymmetric rearrangement of N-para-methoxyphenyl
trifluoroacetimidates as a means to produce chiral primary
allylic amines. Ideally the catalyst should be easily accessible
without the necessity of low-temperature lithiations.
Recently, we described the first preparation of optically
pure 2-ferrocenyl-1-alkyl imidazolines and their pentamethyl-
as well as pentaphenylferrocenyl derivatives.[6,7] The direct
cyclopalladations of these systems are diastereoselective only
in the case of the pentaphenylferrocenyl derivative (R = Ph in
3).[7] This Pd complex gave up to 88% ee for the aza-Claisen
rearrangement of trifluoroacetimidates with 5 mol% catalyst
[3,3]-Sigmatropic rearrangements rank among the most
fundamental reactions in organic chemistry.[1] Particularly
attractive is the PdII-catalyzed aza-Claisen rearrangement,
also known as the Overman rearrangement, which allows for
the formation of chiral enantioenriched protected allylic
amines starting from achiral allylic imidates, which are easily
synthesized from allylic alcohols in a single high-yielding
step.[2] The resulting allylic amine derivatives are valuable
building blocks for the synthesis of important compound
classes such as unnatural amino acids.[3] However, most of the
catalytic asymmetric aza-Claisen rearrangements that have
been investigated have been limited to N-aryl benzimidates,
which are of little practical value since cleavage of the amide
protecting group is typically very low-yielding. Notable
exceptions are the use of allylic trichloroacetimidates[4] and
N-para-methoxyphenyl trifluoroacetimidates[5] since the pro-
tecting groups can generally be removed in preparatively
useful yields. Only a few studies have been devoted to these
practical, but less reactive substrates, and the planar chiral
oxazoline-based palladacycle complexes 1 (COP-X) and 2
(FOP-X) have emerged as the most versatile catalyst
systems.[4,5] Their design is based upon the premise that
steric bulk has to be projected above and below the Pd square
plane in order to allow for a face-selective coordination of the
substrate olefin to the PdII complex.[2] FOP catalysts 2 are not
accessible by direct cyclopalladation of the ferrocene moiety;
I [8]
loading at 408Cafter activation with Ag . Based on these
results, we have designed a less electron-rich catalyst system
in which a bulky N-sulfonyl residue is the key constituent to
permit a direct diastereoselective cyclopalladation of 4. Steric
repulsion in 4 between the residue R1 at the 5-position of the
imidazoline and the sulfonyl group effects a transfer of
chirality to the sulfonylated nitrogen atom, thus resulting in a
preferred equilibrium conformation in which the sulfonyl
group is oriented away from the ferrocenyl moiety, thereby
allowing for a diastereoselective cyclopalladation.
The modular design allowed us to create catalysts 5 in
which the steric demand and the electronic properties could
be adjusted by each single module (Scheme 1), the modules
[*] M. E. Weiss, D. F. Fischer, Z.-q. Xin, S. Jautze, Dr. W. B. Schweizer,
Prof. Dr. R. Peters
Laboratory of Organic Chemistry
ETH Zürich
Wolfgang-Pauli-Strasse 10, Hönggerberg HCI E111
8093 Zürich (Switzerland)
Fax: (+41)44-633-1226
E-mail: peters@org.chem.ethz.ch
[**] This work was financially supported by ETH Research Grant TH-30/
04-2F, F. Hoffmann-La Roche, and Novartis (masters fellowship to
M.E.W. and Ph.D. fellowship to Z.-q.X.). We thank Prof. B. Jaun and
co-workers for performing nOesy experiments, Prof. E. M. Carreira
for sharing laboratory equipment, and Dr. Martin Karpf and Dr. Paul
Spurr (both F. Hoffmann-La Roche, Basel) for critically reading this
manuscript.
Supporting information for this article is available on the WWW
Scheme 1. Modular design of catalysts 5.
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2006, 45, 5694 –5698
5694