1
00
Chemistry Letters Vol.38, No.1 (2009)
Synthesis of Melatonin Receptor Agonist Ramelteon via Rh-catalyzed Asymmetric
Hydrogenation of an Allylamine
Ã
Masayuki Yamashita and Toru Yamano
Medicinal Chemistry Research Laboratories, Pharmaceutical Research Division,
Takeda Pharmaceutical Company, Ltd., Osaka 532-8686
(Received November 7, 2008; CL-081050; E-mail: yamashita masayuki@takeda.co.jp)
a
In the course of developing a practical synthetic method for
Table 1. Asymmetric hydrogenation of compound 1
the selective melatonin MT1/MT2 receptor agonist Ramelteon,
a rhodium Josiphos complex was found to be an excellent cata-
lyst for asymmetric hydrogenation of the key precursor, allyl-
amine 1.
O
NH2
2 mol% [Rh(cod)Cl]2 / ligand 3a-3p
O
NH2
H2 (0.7 MPa), MeOH, rt, 5 h
1
2
2
(
R )2P
2
1
P(R )2
(R )2P
Ramelteon1,2f [(S)-N-[2-(1,6,7,8-tetrahydro-2H-indeno[5,4-
b]furan-8-yl)ethyl]propionamide] (Figure 1), effective in the
treatment of circadian rhythm sleep disorders, consists of a
unique three-fused-ring system with an asymmetric center at
the benzylic position (C8). We have previously reported a prac-
tical synthesis of Ramelteon based on a ruthenium-catalyzed
asymmetric hydrogenation, in which the acylamino moiety of
the substrate 1 might play an important role as an anchoring
2
1
P(R )2
P(R )2
1
(
R )2P
Fe
Fe
Fe
NMe2
ligand 3a-3n
ligand 3o
ligand 3p
1
b
2
c
c
Entry Ligand
R
R
ee/%
yield/%
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
3l
3m
3n
3o
3p
C6H5
4-MeO-Xyl
c-Hex
c-Hex
c-Hex
c-Hex
t-Bu
74
73
35
78
14
43
49
42
34
10
92
27
23
50
80
77
>99
95
57
93
91
87
89
96
92
93
95
32
86
79
92
93
1
c
3,5-(CF ) C H
3
3
2
6
group to realize the excellent enantioselectivity.
A great deal of effort has been made to develop efficient and
4-CF3C6H4
C6H5
4-MeO-Xyl
2
potent chiral ligands in the field of asymmetric hydrogenation.
Among those chiral ligands, ferrocene-based diphosphines, es-
pecially Josiphos ligands enjoy a broad range of applications
t-Bu
3,5-(CF3)2C6H3 t-Bu
3
4-CF3C6H4
4-F-C6H4
4-MeO-C6H4
2-Furyl
t-Bu
t-Bu
t-Bu
t-Bu
Xyl
C6H5
Xyl
c-Hex
C H
due to their high reactivity and diversity that offers differing
steric and electronic properties. The fact that the Josiphos type
ligand has been utilized for the economical manufacture of
1
1
1
1
1
1
(S)-metolachlor, also encouraged us to investigate its perfor-
2-Furyl
c-Hex
mance. In the course of exploring novel catalysts for the synthe-
sis of Ramelteon, we investigated the use of the Josiphos family
of ligands with the allylamine 1, which is a straightforward
4-MeO-Xyl
C6H5
C H
4
substrate for asymmetric hydrogenation. Precatalysts of Rh-
Josiphos were prepared in situ from 2 mol % of [Rh(cod)Cl]2
16
6
5
6
5
a
The reaction was conducted on a 0.1 mmol scale at room temper-
b
5
and each ligand in MeOH at room temperature. We started by
ature under 0.7 MPa of H2. Abbreviations: 4-MeO-Xyl =
c
using the most widely utilized ligand, Josiphos 3a, and surpris-
ingly found that the reaction proceeded very smoothly even
under 0.7 MPa of H2 at room temperature, to afford compound
4
-CH3O-3,5-(CH3)2C6H2, Xyl = 3,5-(CH3)2C6H3. Determined
by HPLC analysis (CHIRALCEL OD-RH column).
1
2
in quantitative yield with 74% ee (Table 1, Entry 1). To the
best of our knowledge, this is the first asymmetric hydrogenation
The effect of the electronic properties of the R groups on
the enantioselectivity was found to be limited. With c-Hex as
4
2
of an allylamine using Rh-Josiphos catalysts.
Josiphos ligands have two sets of tunable substituents on the
the R group, C6H5, 4-MeO-Xyl, and 4-CF3C6H4 gave almost
identical enantioselectivities (Table 1, Entry 1, 2, and 4). With
2
1
phosphorous atoms. Among the ligands tested (3b–3n), ligand
3k gave the best result, with satisfactory enantioselectivity
t-Bu as the R group, all substitutions on the R groups also gave
low enantioselectivities (Table 1, Entry 5, 6, 7, 8, 9, and 10). The
1
(
Table 1, Entry 11). Consideration of the differences between
R groups seem to have little effect on enantioselectivity, except
for the furyl group, where drastic improvement was observed
and the best ee was obtained.
The furyl group is unique and outstanding as a substituent. It
has been generally recognized that 2-Furyl phosphines are poor
the ligands 3a–3n and the outcome of the reaction, can give
some clues to further understanding of the Joshiphos ligands.
All ligands listed preferably gave the identical stereoisomer.
ꢀ
-donor ligands and useful ligands for asymmetric reactions and
O
6
transition metal-catalyzed reactions such as the Stille reaction.
Therefore, we anticipated that the ligand having a 2-furyl sub-
stituent might lead to substantial change in the complex.
O
8
N
H
2
The bulkiness of the R group may have a weak correlation
1
Figure 1. Structure of Ramelteon.
to the enantioselectivities. With 4-MeO-Xyl as the R group,
Copyright Ó 2009 The Chemical Society of Japan