One-pot formation of nitrogen-containing heterocyclic ring systems using a deprotection-cyclisation-asymmetric reduction sequence

Article abstract of DOI:10.1039/b509231k

A one-pot sequence of amine deprotection, intramolecular C=N bond formation and subsequent asymmetric reduction may be promoted by a ruthenium catalyst. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b509231k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
One-pot formation of nitrogen-containing heterocyclic ring systems
using a deprotection–cyclisation–asymmetric reduction sequence{{
Glynn D. Williams,a Charles E. Wadeb and Martin Wills*a
Received (in Cambridge, UK) 1st July 2005, Accepted 2nd August 2005
First published as an Advance Article on the web 26th August 2005
DOI: 10.1039/b509231k
A one-pot sequence of amine deprotection, intramolecular
CLN bond formation and subsequent asymmetric reduction
may be promoted by a ruthenium catalyst.
The catalytic asymmetric reduction of CLN bonds represents a
highly practical and efficient method for the synthesis of
enantiomerically pure amines.1 Previously reported methods for
this process include hydrogenation (i.e. with hydrogen gas),2
hydrosilylation3a,3b and reduction with borane.3c In the mid-1990s,
Noyori reported the efficient asymmetric reduction of cyclic imines
using ruthenium–TsDPEN complex 1.4 In 1999, Baker reported
that structurally-related Rh(III) complexes 2 could also be
successfully employed in this application.5 Since these early
Scheme 1 Reagents and conditions: (i) CH2LCH(CH2)3MgBr 5, THF.
reports, many other groups have reported synthetic applications
(ii) OsO4, NaIO4, dioxane–H2O. (iii) HCO2H, then Et3N and 2 mol%
of asymmetric transfer hydrogenation (ATH) of CLN bonds.6
catalyst R,R-1 (formed in situ), MeCN, 3 h. (iv) HCO2H, then Et3N and
Whilst the majority of the interest in this area is focused on Ru(II)
2 mol% catalyst R,R-8 (formed in situ from 1 mol% dimer 9), MeCN, 2 h.
complexes, the Rh(III) derivatives have been developed commer-
cially by Avecia.7
Grignard reagent 5 with the known Boc lactam 6 resulted in
formation of the ketone 7 in 71% isolated yield. Oxidative cleavage
of the terminal double bond in 7, using the osmium-catalysed
periodate conditions reported by Johnson,9 then gave the required
dicarbonyl compound 3 in 85% yield. When subjected to our
conditions for asymmetric one-pot reductive amination, 3 was
converted to 4 in 64% isolated yield and 71% enantiomeric excess
[determined by chiral HPLC against a racemic standard prepared
by TFA-catalysed cyclisation of
3 and reduction with
NaBH(OAc)3]. In view of the relative complexity of the process,
this was considered a pleasing result. We also took the opportunity
to test our tethered catalyst 8, which we have previously reported
and is formed in situ by exposure of dimer 9 to the reaction
conditions.10 In the event, 8 gave a very similar result to that
obtained with 1. At present, the configuration of 4 has been
assigned by analogy with that of known cyclic substrates, and has
not been confirmed.
As a part of our own ongoing studies into ATH reactions, we
have reported the one-pot conversion of tBoc-protected ketoa-
mines into cyclic amines in high enantioselectivity. In our studies
we utilised the Ru(II) catalyst 1.8 In this paper, we report the
extension of our studies of this enantioselective reductive
amination process to more complex systems, and demonstrate
that two rings can be formed in a single enantioselective process.
In order to establish the scope of the reductive amination, we
first wished to investigate the transformation of ketoaldehyde 3
into the tricyclic product 4. Towards this end, substrate 3 was
prepared using the sequence in Scheme 1. The reaction of
aDepartment of Chemistry, The University of Warwick, Coventry, UK
CV4 7AL. E-mail: m.wills@warwick.ac.uk.; Fax: +44 24 7652 4112;
Tel: +44 24 7652 3260
bGlaxoSmithKline, Medicines Research Centre, Gunnels Wood Road,
Stevenage, Hertfordshire, UK SG1 2NY
{ Electronic supplementary information (ESI) available: Experimental
procedures and characterisation data for intermediates and cyclisation
products. See http://dx.doi.org/10.1039/b509231k
Having obtained a promising result, we wished to establish
whether we could extend our one-pot procedure to the conversion
of 10 into 11. Although this is similar to the conversion of 3 to 4 in
terms of bond formation, it provides a potential route to a number
of berberine natural products with this tetracyclic ring structure.11
{ Dedicated to Professor Steven V. Ley on the occasion of his 60th
Birthday.
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 4735–4737 | 4735

View Article Online
Simple berberine derivatives have been used in medicinal chemistry
applications. One example is the synthesis of curare-like agents by
GSK, which has been achieved through the use of an ATH
step.6a,b The synthesis of 10 proved to be challenging, however the
successful route to this is shown in Scheme 2. In this sequence, the
Bischler–Napieralski cyclisation of amide 12 into cyclic imine 13
was followed by treatment with Boc anhydride and acid-catalysed
ring-opening of the resultant acyliminium cation to give 14. The
use of p-toluenesulfonic acid was found to be essential; in its
absence the formation of water-stable N-tBoc enamines was
observed. A key step was the palladium-catalysed coupling12 of 14
with potassium vinyltrifluoroborate 15 to furnish 16 in an
acceptable 80% yield. Finally, oxidative cleavage of the alkene in
16 resulted in formation of the required ketoaldehyde 10. The
conversion of 10 into the cyclic amine 11 proved to be successful.
Using non-tethered catalyst 1, the product was isolated in an
adequate 70% yield, but proved to be of only 10% ee. A better
result was obtained using tethered complex 8; 73% isolated yield
and 50% ee. The reaction was also faster; 4 hours with 8 compared
to 48 hours for full reaction using the untethered 1. Enantiomeric
excesses were calculated by chiral HPLC against a racemic sample
prepared from 10. The lower ee for the formation of 11 compared
to 4 may reflect the similarity of the functional groups flanking the
intermediate imine.
Scheme 3 Possible cyclisation modes.
ring 18 is formed in the first step, then this engages in a
transannular cyclisation and reduction to give the product.
In the cyclisation of 10, NMR inspection of the reaction mixture
at 5, 10 and 24 hours revealed that there was a slow step in the
cyclisation process. The first cyclisation–reduction appeared to
take place slowly, but then the second reduction was rapid. The
crude 1H NMR spectrum of the intermediate suggested that it was
the aldehyde which condensed first with the amine to give the
imine which was slowly reduced. This was surprising because
normally six-membered ring formation would be expected to
outpace that of a ten-membered ring. On the other hand,
aldehydes are more electrophilic than ketones and it may be that
this reactivity is dominating the selectivity of the reaction.
There is some precedent to support the proposed ten-membered
ring formation in the recent literature however. For example the
natural products 19 and 20 have both been isolated and speculated
to have been formed via an imine condensation,13 even though
there is an alternative possibility of formation of a smaller ring
with the ketone in the molecule.
The mechanism of the cyclisation–reduction process has not
been fully determined. However, two synthetic courses can be
envisaged (Scheme 3). In the first of these, a six-membered ring is
formed and reduced asymmetrically to give 17, then a second ring
is built upon this. In the second possible route, a ten-membered
If it is indeed the case that 18 is the key intermediate on the way
to 11, then the key asymmetric step will be the reduction of the
iminium cation 21. By analogy, the iminium cation 22 would
occupy a corresponding position on the pathway to 4.
The participation of 21 and 22 in the reduction pathway may
account, in part, for the decreasing enantioselectivity upon going
from the smaller to the larger substrate, since the steric and
electronic difference between each side of the iminium cation in 21
is significantly less than in 22. What is less clear, however, is the
reason for the improved performance of catalyst 8 over 1. Unlike
Scheme 2 Reagents and conditions: (i) POCl3, PhCH3, 110 uC. (ii)
Boc2O, DMF, 90 uC, then p-TsOH, DMF, H2O, 130 uC. (iii)
CH2LCHBF3?K 15, Pd(PPh3)4, PhCH3, EtOH, K2CO3, 85 uC. (iv)
OsO4, NaIO4, dioxane–H2O. (v) HCO2H, then Et3N and 2 mol% catalyst
R,R-1 (formed in situ), MeCN, 48 h. (vi) HCO2H, then Et3N and 2 mol%
catalyst R,R-8 (formed in situ from 1 mol% dimer 9), MeCN, 4 h.
4736 | Chem. Commun., 2005, 4735–4737
This journal is ß The Royal Society of Chemistry 2005

View Article Online
acknowledged.15 Mr J. C. Bickerton of the Warwick MS service
and Dr B. Stein of the EPSRC MS service are thanked for their
support of this project. We also acknowledge the generous loan of
ruthenium salts by Johnson-Matthey Ltd.
Notes and references
Scheme 4 Reagents and conditions: (i) HCO2H, then Et3N and 2 mol%
catalyst R,R-1 (formed in situ) or TFA.
1 (a) Comprehensive Asymmetric Catalysis, ed. E. N. Jacobsen, A. Pfaltz
and H. Yamamoto, Springer, Berlin, 1999; (b) J. M. J. Williams,
Catalysis in Asymmetric Synthesis, Blackwell Science, Sheffield, 1999; (c)
Catalytic Asymmetric Synthesis, ed. I. Ojima, VCH Press, Berlin, 2nd
edn, 2000.
2 (a) M. J. Burk and J. E. Feaster, J. Am. Chem. Soc., 1992, 114, 6266; (b)
C. A. Willoughby and S. L. Buchwald, J. Am. Chem. Soc., 1994, 116,
8952–11703; (c) P. Schnider, G. Koch, R. Preˆtoˆt, G. Wang,
F. M. Bohnen, C. Kru¨ger and A. Pfaltz, Chem. Eur. J., 1997, 3,
887.
3 (a) X. Verdaguer, U. E. W. Lange, M. T. Reding and S. L. Buchwald,
J. Am. Chem. Soc., 1996, 118, 6784; (b) X. Verdaguer, U. E. W. Lange
and S. L. Buchwald, Angew. Chem. Int. Ed., 1998, 37, 1103; (c) borane
CLN reduction; S. Itsuno, Y. Sakurai, K. Ito, A. Hirao and
S. Nakahama, Bull. Chem. Soc. Jpn., 1987, 395.
the reduction of ketones, the mechanism of imine reduction by
transfer hydrogenation is not well understood. The proposal of a
stereochemical model would be premature at this stage, before
further data has been acquired and molecular modelling studies
completed. However we believe that the improvement is probably
not the result of a steric effect, because the tether is some way from
the reaction sphere. It is more likely that a significant stereoelec-
tronic effect may be operating through the ligand, for example
involving the SO2 dipole, as a result of the stereochemically-locked
nature of the system.
In order to extend the methodology further, we attempted to
prepare the 6,6,5 analogue of 4 through a one-pot cyclisation.
However, under the reaction conditions, the pyrrole 24 was formed
from the precursor 23 (Scheme 4), presumably through a Pall–
Knorr process. Ketoaldehyde 23 was prepared in an analogous
manner to 3. The cyclisation, which may also be effected simply by
treating 23 with TFA, may be due to the strong thermodynamic
effect of forming a stable heteroaromatic structure. Adjustment of
the pH to 12 with aq. NaOH promoted the precipitation of the
product in essentially quantitative yield. Although 24 is a known
compound,14 we are not aware of this approach having been taken
to its synthesis previously, and we are currently studying the scope
of this process.
4 N. Uematsu, A. Fujii, S. Hashiguchi, T. Ikariya and R. Noyori, J. Am.
Chem. Soc., 1996, 118, 4916.
5 J. Mao and D. C. Baker, Org. Lett., 1999, 1, 841.
6 (a) I. Kaldor, P. L. Feldman, R. A. Mook, Jr., J. A. Ray, V. Samano,
A. M. Sefler, J. B. Thompson, B. R. Travis and E. E. Boros, J. Org.
Chem., 2001, 66, 3495; (b) V. Samano, A. J. Ray, J. B. Thomson,
R. A. Mook, Jr., D. K. Jung, C. S. Koble, M. T. Martin, E. C. Bigham,
C. S. Regitz, P. L. Feldman and E. E. Boros, Org. Lett., 1999, 1, 1993;
(c) L. F. Tietze, N. Rackelmann and I. Muller, Chem. Eur. J., 2004, 10,
2722; (d) L. S. Santos, R. A. Pilli and V. H. Rawal, J. Org. Chem., 2004,
69, 1283; (e) E. Vedejs, P. Trapencieris and E. Suna, J. Org. Chem.,
1999, 64, 6724; (f) G. J. Meuzelaar, M. C. A. van Vliet, L. Maat and
R. A. Sheldon, Eur. J. Org. Chem., 1999, 2315; (g) K. H. Ahn, C. Ham,
S.-K. Kim and C.-W. Cho, J. Org. Chem., 1997, 62, 7047; (h) K. H. Ahn,
C. Ham and S.-K. Kim, Tetrahedron Lett., 1998, 39, 6321; (i)
L. F. Tietze, Y. Zhou and E. Topken, Eur. J. Org. Chem., 2000, 2247.
7 S. Richards, M. Ropic, D. Blackmond and A. Walmsley, Anal. Chim.
Acta, 2004, 519, 1.
In conclusion, we have demonstrated that it is possible to form
tri- or tetracyclic heterocyclic products in good yields by using a
8 Y. Xu, N. W. Alcock, G. J. Clarkson, G. Docherty, G. Woodward and
M. Wills, Org. Lett., 2004, 6, 4105.
9 R. Pappo, D. S. Allen, Jr. and R. U. Johnson, Synth. Commun., 2001,
31, 401.
10 J. Hannedouche, G. Clarkson and M. Wills, J. Am. Chem. Soc., 2004,
126, 986.
11 P. M. Dewick, Medicinal Natural Products, Wiley, New York, 2001,
p. 339.
12 G. A. Molander and M. R. Rivero, Org. Lett., 2002, 4, 107.
13 L. Rastrelli, A. Cpasso, C. Pizza and N. De Tommasi, J. Nat. Prod.,
1997, 60, 1065.
14 I. Moreno, I. Tellitu, E. Dominguez and R. SanMartin, Eur. J. Org.
Chem., 2002, 2126.
one-pot deprotection–reductive amination process. Using
a
transfer hydrogenation catalyst to reduce the intermediate iminium
cations provides products in ees of 50–70%. Since both
enantiomers of the C2-symmetric diamine precursor of catalysts
1 and 8 are readily available, either enantiomer of imine-reduction
product may be prepared using this method. We are currently
studying the further applications of this methodology using
alternative substrates and catalysts, including some based on Rh.
The results of this study will be reported in due course.
EPSRC and GSK are thanked for financial support of this
project (to GDW) through an industrial CASE award. The use
of the EPSRC Chemical Database Service at Daresbury is
15 D. A. Fletcher, R. F. McMeeking and D. Parkin, J. Chem. Inf. Comput.
Sci., 1996, 36, 746.
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 4735–4737 | 4737