COMMUNICATION
www.rsc.org/chemcomm | ChemComm
0
The first syntheses of enantiopure 2,2 -biindolinew
a
b
a
Steven M. Wales, Anthony C. Willis and Paul A. Keller*
Received 24th September 2010, Accepted 14th October 2010
DOI: 10.1039/c0cc04045b
0
The first two syntheses of chiral 2,2 -biindoline are reported
0
competing mono-addition of the chloride and iodide anions.
The desired diol (R,R)-4 was isolated (31%) by selective
crystallisation from the crude mixture. This was converted to
dimesylate (R,R)-5 (84%) which in turn was subjected to
standard azidation conditions, affording – after an extremely
sluggish reaction – a crude mixture of the desired diazide
either in five steps from 2,2 -bioxirane, or three steps from
0
2
,2 -biaziridine, both with exceptional enantiopurity.
Chiral biamine compounds are well established as ligands in
stereoselective metal-catalysed reactions, including allylic
1
additions, reductions and asymmetric dihydroxylations. Our
(S,S)-6 (B60% conversion) and a chromatographically
inseparable azidoalkene arising from mono-elimination
interest in chiral ligand design focusses on a new principle that
uses helix sense discrimination to achieve stereoselective
outcomes. In particular, ligand types we are interested in
include biphosphines, biarsines, biamines and ligands that
possess a mixture of heteroatoms. The molecular helix itself
is defined by two stereogenic atoms, flanked by metal
co-ordinating heteroatoms, manifesting an arc of helicity once
bound to a metal e.g. palladium. The helical groove depth and
degree of twist could be modulated by a range of substituents,
including the presence of fused rings. Importantly, an appropriate
synthetic strategy towards such structures should not only be
highly stereoselective, but should allow articulation to efficiently
produce a range of substituted ligands including those with
different ring sizes, and/or containing fused ring structures.
One example of a biamine that would meet our helical
(B25% conversion, not illustrated). The diazide was isolated
after oxidation of the by-product with KMnO and subsequent
4
silica gel column chromatography (47% yield overall). Staudinger
reduction provided complete conversion to the diamine
(S,S)-7 which was isolated by successive crystallisation as its
dihydrochloride salt and liberation with sodium hydroxide
6
(71%). Finally, after several unsuccessful attempts to form
the biindoline using standard palladium and copper amination
protocols, the molecule was found to undergo the desired
regioselective cyclisation under microwave irradiation, providing
7
the target ligand (S,S)-1 in 57% yield. The product was found
to be enantiomerically pure (>99% ee) after preparing the
8
diBoc derivative and performing HPLC with analysis
relative to a sample enriched in the R,R-enantiomer that
0
design criteria when metal-bound is 2,2 -biindoline 1. Despite
was previously synthesised in our laboratory using the
3
metathesis – AD strategy.
the illustrious history of the indoline molecule, including
its place in natural product chemistry, drug design and
development and its use in industry as a key element in
catalysts, there has been surprisingly little investigation into
Although the synthesis of biindoline (S,S)-1 from bioxirane
(R,R)-3 provided the target compound in enantiomerically
pure form, the strategy needed to be modified to avoid the
low yielding ring opening with Grignard 2 and the inefficient
azidation reaction.
2
the dimeric structure 1 – even more surprising is the lack of
any reporting of the stereoselective synthesis of 1 or related
compounds that contain 1 as a substructure. In our first
Therefore, an alternative synthesis of 1 was proposed using
the analogous biaziridine as the chiral building block, such
that ring opening would directly install the required nitrogen
atoms. This would provide a shorter entry to the penultimate
diamine (S,S)-7, which could be cyclised under the established
palladium catalysed microwave conditions.
0
attempt at a stereoselective synthesis of 2,2 -biindoline 1 using
a metathesis–asymmetric dihydroxylation (AD) strategy, poor
yields and stereoselective outcomes resulted from steric
3
hinderence in the AD step. Therefore, we devised a new
strategy to the key intermediate chiral diols (e.g. 4) and in this
communication report the first synthesis of the parent
0
Efficient ring opening of the biaziridine with Grignard 2
(Scheme 2) would require N-protection with an electron
withdrawing group and two such derivatives of the molecule
2
,2 -biindoline 1 in enantiopure form.
Our chiral pool starting material was the bioxirane (R,R)-3
9
10
(
Scheme 1), prepared from D-tartaric acid as previously
are known – the Boc (8, R = Boc) and the tosyl (R = Ts).
4
described. This was ring opened with Grignard 2 under
5
Given that strong reducing conditions are required to remove
11,12
4
CuI catalysis, giving a mixture of diols arising from the
the tosyl group in good yield,
we chose to begin our
investigations with the acid labile Boc carbamate as the
protecting group.
a
School of Chemistry, University of Wollongong, Wollongong,
Australia. E-mail: keller@uow.edu.au; Fax: +61 2 4221 4287;
Tel: +61 2 4221 4692
Research School of Chemistry, The Australian National University,
Canberra, Australia. E-mail: willis@rsc.anu.edu.au;
Tel: +61 2 6125 4109
The biaziridine (S,S)-8 has been prepared previously from
D-tartaric acid using two cyclisation strategies – a Staudinger
9
2 substitution. In our hands,
b
reduction and a traditional S
only the S
N
N
2 cyclisation sequence provided the biaziridine in
w Electronic supplementary information (ESI) available: Synthetic
procedures, X-ray data, NMR spectra and HPLC traces of
high yield – initial attempts to prepare the product using the
alternative Staudinger approach resulted in poor yields.
0
2,2 -biindoline structures. CCDC [794927-794928]. For ESI and
1
3
The CuBrꢀSMe catalysed ring opening of the biaziridine
crystallographic data in CIF or other electronic format see DOI:
0.1039/c0cc04045b
2
1
(S,S)-8 with Grignard 2 gave the expected dicarbamate
This journal is c The Royal Society of Chemistry 2010
9
226 Chem. Commun., 2010, 46, 9226–9228