A One-Pot Process for the Enantioselective Synthesis of Amines via Reductive Amination under Transfer Hydrogenation Conditions

Article abstract of DOI:10.1021/ol035746r

(Equation presented) Cyclic amines may be prepared via a sequence of deprotection followed by intramolecular reductive amination of t-Boc-protected amino ketones under asymmetric transfer hydrogenation conditions. In cases where the corresponding imine reaction proceeds with high enantioselectivity, this is reflected in the one-step process.

Full text of DOI:10.1021/ol035746r

ORGANIC
LETTERS
2003
Vol. 5, No. 22
4227-4230
A One-Pot Process for the
Enantioselective Synthesis of Amines
via Reductive Amination under Transfer
Hydrogenation Conditions
,†
Glynn D. Williams,† Richard A. Pike,† Charles E. Wade,‡ and Martin Wills*
Department of Chemistry, UniVersity of Warwick, CoVentry CV4 7AL, U.K., and
Process Chemistry Department, GlaxoSmithKline Pharmaceuticals Ltd.,
Gunnels Wood Road, SteVenage, Herts SG1 2NY, U.K.
m.wills@warwick.ac.uk
Received September 11, 2003
ABSTRACT
Cyclic amines may be prepared via a sequence of deprotection followed by intramolecular reductive amination of t-Boc-protected amino
ketones under asymmetric transfer hydrogenation conditions. In cases where the corresponding imine reaction proceeds with high
enantioselectivity, this is reflected in the one-step process.
Asymmetric transfer hydrogenation has recently emerged as
an excellent method for the enantioselective reduction of
ketones to alcohols.1 Although a long-established method
for ketone reduction, the significant recent increase in interest
in this method has been largely due to the introduction by
Noyori of several new ruthenium(II)-based catalysts.2 In
particular, the monotosylated diamine TsDPEN 1 (Figure 1)
The Ru(II)/TsDPEN system also works well for the
reduction of CdN double bonds. However, this variation has
been significantly less developed. Noyori has reported that
several amines, most particularly tetrahydroisoquinolines, can
be prepared by this method in high yield and enantioselec-
tivity.4 Vedejs et al. have since extended their work.5 Baker
and Mao have also described the analogous use of Rh(III)/
(1) (a) Palmer, M. J.; Wills, M. Tetrahedron: Asymmetry 1999, 10,
2045. (b) Noyori, R.; Hashiguchi, S. Acc. Chem. Res. 1997, 30, 97. (c)
Zassinovich, G.; Mestroni, G.; Gladiali, S. Chem. ReV. 1992, 92, 1051.
(2) (a) Takehara, J.; Hashiguchi, S.; Fujii, A.; Inoue, S.-I.; Ikariya, T.;
Noyori, R. Chem. Commun. 1996, 233. (b) Fujii, A.; Hashiguchi, S.;
Uematsu, N.; Ikariya, T.; Noyori, R. J. Am. Chem. Soc. 1996, 118, 2521.
(c) Hashiguchi, S.; Fujii, A.; Takehara, J.; Ikariya, T.; Noyori, R. J. Am.
Chem. Soc. 1995, 117, 7562. (d) Haack, K. J.; Hashiguchi, S.; Fujii, A.;
Ikariya, T.; Noyori, R. Angew. Chem., Int. Ed. Engl. 1997, 36, 285. (e)
Murata, K.; Ikariya, T.; Noyori, R. J. Org. Chem. 1999, 64, 2186.
(3) (a) Palmer, M.; Walsgrove, T.; Wills, M. J. Org. Chem. 1997, 62,
5226. (b) Nordin, S. J. M.; Roth, P.; Tarnai, T.; Alonso, D. A.; Brandt, P.;
Andersson, P. G. Chem. Eur. J. 2001, 7, 1431. (c) Everaere, K.; Mortreux,
A.; Bulliard, M.; Brussee, J.; van der Gen, A.; Nowogrocki, G.; Carpentier,
J. F. Eur. J. Org. Chem. 2001, 275. (d) Hennig, M.; Puntener, K.; Scalone,
M. Tetrahedron: Asymmetry 2000, 11, 1849. (e) Blacker, A. J.; Mellor, B.
J. (Avecia/Zeneca), WO 98/42643 A1, 1998.
Figure 1. Structures of ligands used for asymmetric transfer
hydrogenation.
has been adopted as one of the optimal ligands for this
process, particularly when used in formic acid/triethylamine,
which acts both as the solvent and the hydrogen source.2,3
(4) Uematsu, N.; Fujii, A.; Hashiguchi, S.; Ikariya, T.; Noyori, R. J.
Am. Chem. Soc. 1996, 118, 4916.
(5) Vedejs, E.; Trapencieris, P.; Suna, E. J. Org. Chem. 1999, 64, 6724.
† University of Warwick.
‡ GlaxoSmithKline Pharmaceuticals.
10.1021/ol035746r CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/08/2003

Cp* catalysts.6 It is noteworthy that CdN reduction requires
TsDPEN/Ru reagents in formic acid/triethylamine with an
organic cosolvent such as acetonitrile, dichloromethane, or
DMF. In contrast the popular combination of Ru(II) with
amino alcohol ligands such as 2 and 3 (Figure 1) is
incompatible with formic acid and therefore not suitable for
imine reduction.
substrate for the reaction under study, as the yield was high.
Our first attempt at effecting a one-pot deprotection/
cyclization/reduction sequence under transfer hydrogenation
conditions did, however, result in failure. Analysis of the
reaction mixture revealed that t-Boc deprotection had been
unsuccessful, presumably because the pH of the solution was
too high for this process. In view of this we believed that
we might have more success if we first performed the
deprotection/cyclization sequence under more strongly acidic
conditions and then adjusted the reaction mixture later for
the transfer hydrogenation process. After a short series of
tests we found that the use of 9 volumes of formic acid was
sufficient to promote full t-Boc removal after 16 h, resulting
in formation of imine 6. The reaction was then repeated,
but after the time required to remove the t-Boc group,
sufficient triethylamine was added to generate a 5:2 formic
acid/triethylamine mixture to which the Ru(II)/TsDPEN
catalyst was added. Anhydrous acetonitrile was also added
to reproduce the conditions reported as being essential for
imine reduction.4 The deprotection and reduction events
could all be conveniently monitored by NMR analysis, which
revealed essentially full reduction to 7 after 24 h at 28 °C
(Scheme 2).9 To our knowledge this represents the first report
In the applications described above, the imine was
prepared and isolated prior to the key reduction step. We
reasoned that the synthetic potential of the method might be
further increased if it was possible to perform the imine
formation and reduction in one synthetic step, i.e., through
a reductive amination process. This would be particularly
useful for the synthesis of cyclic amines through an intra-
molecular cyclization process. Furthermore, given that the
formic acid/triethylamine (5:2 molar ratio) reductions are
typically carried out at ca. pH 5, we reasoned that this would
be compatible with the reductive amination process.7
Toward this end, we identified the t-Boc-protected amino
ketone 4 as a convenient system for our initial studies. The
use of t-Boc is significant because, as well as being a
commonly used protecting group, it may be removed under
acidic conditions not dissimilar to those employed for the
formic acid/triethylamine transfer hydrogenation reaction.
Ketone 4 is a known compound and was prepared in one
step from the t-Boc-protected δ-lactam 5 following a
literature precedent (Scheme 1).8 To assess whether the imine
Scheme 2
Scheme 1
of a reductive alkylation of amines under Ru(II)/TsDPEN
conditions. Although the yield was good, the product was
again essentially racemic.
Determination of the ee of the product 7 proved difficult,
no success being achieved with either chiral HPLC (OD
column) or europium shift reagents. We were successful,
however, when we employed the pyrrolidine reagent 8 to
derivatize the amine (Figure 2).10 Displacement of the
chlorine atom from the chloromethyl position results in
formation of two diastereoisomers in which the methyl
doublet conveniently indicated the ratio of enantiomers.
Despite the low enantioselectivity observed in the reduc-
tion, we were encouraged by the high conversion. To confirm
derived from 4 was convenient for transfer hydrogenation,
we first removed the t-Boc group with TFA to give an amine,
which cyclized readily to 6 under the reaction conditions
(92% isolated yield). The reduction of 6 to amine 7 was
achieved through the use of reported Ru(II)/TsDPEN condi-
tions (with acetonitrile cosolvent) in 94% isolated yield,
although the product was essentially racemic (Scheme 1).
The method for the determination of ee is described later in
this paper.
(9) Typical Procedure. t-Boc-amino-ketone (0.2 g) was stirred in freshly
distilled formic acid (1.8 mL) for 16 h. The flask was then sealed and cooled
to 0°C; triethylamine (3 mL) was added cautiously with vigorous shaking
until all gas had been readsorbed. In a separate flask a mixture of (p-cymene)
ruthenium(II) chloride dimer (0.25 mol %) and (1R,2R)-TsDPEN (0.5 mol
%), triethylamine (1 drop), and anhydrous acetonitrile (1 mL) were stirred
at 28 °C for 40 min. The catalyst solution was transferred to the formic
acid/triethylamine solution, and the mixture was stirred at 28°C until
complete by NMR. The mixture was made basic (pH 9-10) with saturated
Na2CO3 solution and extracted with DCM (3 × 25 mL). The combined
organics were dried (MgSO4) and filtered, and the solvent removed under
reduced pressure. The residue was purified by flash column chromatography
(10-15% v/v ethyl acetate/hexane on silica pretreated with Et3N) to afford
the amine.
Although the reduction had proceeded without any selec-
tivity, we reasoned that 4 still represented a good test
(6) (a) Mao, J. Baker, D. C. Org. Lett. 1999, 1, 841. (b) Meuzelaar, G.
J.; van Vliet, M. C. A.; Maat, L.; Sheldon, R. A. Eur. J. Org. Chem. 1999,
2315. (c) Ahn, K. H.; Ham, C.; Kim, S.-K.; Cho, C.-W. J. Org. Chem.
1997, 62, 7047.
(7) Borch, R. F.; Bernstein, M. D.; Durst, H. D. J. Am. Chem. Soc. 1971,
93, 2897.
(8) Giovanninni, A.; Savaoia, D.; Umnani-Ronchi, A. J. Org. Chem.
1989, 54, 228.
(10) Smith, M. B.; Bembofsky, B. T.; Son, Y. C. J. Org. Chem. 1994,
59, 1719.
4228
Org. Lett., Vol. 5, No. 22, 2003

Scheme 4
Figure 2. NMR derivatizing agent for ee determination.
that the low selectivity was not the result of our modified
conditions, we repeated the reduction on 9, a substrate that
has been reported to give a product of high enantioselectivity
under normal asymmetric transfer hydrogenation conditions.
Compound 9 was prepared following a literature method and
was successfully reduced to amine 10 in 95% yield and 88%
ee using TsDPEN in formic acid/triethylamine with aceto-
nitrile as cosolvent (Scheme 3).4 For the comparative one-
introduction of an aromatic ring on the side of the imine
bearing the lone pair is detrimental to enantioselectivity,
which suggests some sort of steric clash between this group
Scheme 3
Table 1. One-Pot Deprotection/Cyclization/Asymmetric
Transfer Hydrogenation of δ-t-Boc Amino Ketones
R
yield, %
o-(MeO)C6H4
n-(MeO)C6H4
p-(MeO)C6H4
p(CF3)C6H4
m-(CF3)C6H4
2-thiophene
c-C6H11
78
96
94
99
98
20
98
pot cyclization/reduction the t-Boc-protected material 11 was
used as the substrate. Using the protocol described previ-
ously, the product 10 was formed in 85% yield and 88% ee,
comparable to the result from the direct imine reduction
(Scheme 3). This method represents a viable practical
alternative to reduction of the preformed imine 9; it was
pleasing to see that the results were not significantly
deteriorated using the alternative method. We therefore
conclude that 6 is a poor substrate with regard to enantio-
selectivity in this particular application.
and a group in the catalyst. An aromatic ring on the side of
the imine opposite the nitrogen lone pair, by contrast, appears
to be essential for high selectivity (Figure 3).
To test the generality of the one-pot method we also
investigated the preparation of the cyclic amines from a series
of compounds (Scheme 4), the results of which are illustrated
in Table 1. In the majority of cases examined the conversions
were excellent, although the products were essentially
racemic. It was also possible to form the seven-membered
ring product 12 from 13 in only 12% yield (racemic);
however, attempts to form the five-membered heterocyclic
product from 14 failed.
The reason for the loss of stereocontrol in the cyclization
of 4, under conditions where 10 is formed in high ee, is
unclear. Unlike the ketone reduction process, the asymmetric
transfer hydrogenation of imines is less well understood. One
pattern that does appear to emerge, however, is that the
Figure 3. Relationship of aromatic ring position to extent of
enantioselectivity in reduction.
By analogy with the ketone reduction, it may be the case
that this ring is able to engage in a π-stabilizing interaction
with the η6-ring on the catalyst.11
This analysis is supported by the original observation by
Noyori that the reduction of the derivative of 9 bearing a
(11) Yamakawa, M.; Yamada, I.; Noyori, R. Angew. Chem., Int. Ed. 2001,
40, 2818.
Org. Lett., Vol. 5, No. 22, 2003
4229

phenyl ring in place of the exocyclic methyl group gives a
product of some 10% lower ee.4
Swansea EPSRC MS service for MS analysis of certain
compounds. We wish to acknowledge the use of the EPSRC's
Chemical Database Service at Daresbury.12
In conclusion, we have demonstrated that the synthesis
of cyclic amines may be achieved under modified conditions
of asymmetric transfer hydrogenation and that in cases where
the intermediate imines are known to give products of high
ee, this is reproduced in our process. We are currently
examining the extension of this methodology to further
substrates, including acyclic imines.
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL035746R
Acknowledgment. We thank the EPSRC and Glaxo-
SmithKline for the support of an industrial CASE award (to
G.W.) and Dr. B. Stein and Professor D. Games of the
(12) Fletcher, D. A.; McMeeking; R. F.; Parkin, D. J. Chem. Inf. Comput.
Sci. 1996, 36, 746.
4230
Org. Lett., Vol. 5, No. 22, 2003