A novel and practical synthesis of ramelteon

Article abstract of DOI:10.1021/op500386g

An efficient and practical process for the synthesis of ramelteon 1, a sedative-hypnotic, is described. Highlights in this synthesis are the usage of acetonitrile as nucleophilic reagent to add to 4,5-dibromo-1,2,6,7-tetrahydro-8H-indeno[5,4-b]furan-8-one 2 and the subsequent hydrogenation which successfully implement four processes (debromination, dehydration, olefin reduction, and cyano reduction) into one step to produce the ethylamine compound 13 where dibenzoyl-l-tartaric acid is selected both as an acid to form the salt in the end of hydrogenation and as the resolution agent. Then, target compound 1 is easily obtained from 13 via propionylation. The overall yield in this novel and concise process is almost twice as much as those in the known routes, calculated on compound 2.

Full text of DOI:10.1021/op500386g

Technical Note
pubs.acs.org/OPRD
A Novel and Practical Synthesis of Ramelteon
Sa Xiao, Chao Chen, Hongyan Li, Kuaile Lin, and Weicheng Zhou*
State Key Lab of New Drug & Pharmaceutical Process, Shanghai Institute of Pharmaceutical Industry, State Institute of
Pharmaceutical Industry, Shanghai 200437, China
S
* Supporting Information
ABSTRACT: An efficient and practical process for the synthesis of ramelteon 1, a sedative-hypnotic, is described. Highlights in
this synthesis are the usage of acetonitrile as nucleophilic reagent to add to 4,5-dibromo-1,2,6,7-tetrahydro-8H-indeno[5,4-
b]furan-8-one 2 and the subsequent hydrogenation which successfully implement four processes (debromination, dehydration,
olefin reduction, and cyano reduction) into one step to produce the ethylamine compound 13 where dibenzoyl-L-tartaric acid is
selected both as an acid to form the salt in the end of hydrogenation and as the resolution agent. Then, target compound 1 is
easily obtained from 13 via propionylation. The overall yield in this novel and concise process is almost twice as much as those in
the known routes, calculated on compound 2.
INTRODUCTION
RESULTS AND DISCUSSION
■
■
Ramelteon 1 or (S)-N-[2-(1,6,7,8- tetrahydro-2H-indeno[5,4-
b]furan-8-yl)ethyl]propionamide is well-known as a selective
melatonin MT1/MT2 receptor agonist which primarily has
been used as a treatment of insomnia.1,2 It was developed by
the Japanese company Takeda and approved by the US Food
and Drug Administration (FDA) in 2005. With negligible
affinity for other neuronal receptors related with gamma-
aminobutyric acid and benzodiazepine, ramelteon seldom has
the adverse effects, occurring for the traditional sedative
hypnotics, such as learning and memory impairment and
drug dependence.3,4 This advantage makes it the first and only
prescription sleep medication that is not designated as a
controlled substance by the US Drug Enforcement Admin-
istration (DEA).5
Our studies commenced with the nucleophilic addition of
compound 2 (prepared according to the literature8) and
acetonitrile in tetrahydrofuran with n-butyllithium as a base. An
exciting result was obtained in the concept-proof test where 2
equiv n-butyllithium was employed and the reaction temper-
ature was controlled at −15 °C (Table 1, entry 1). Then, the
following experiments showed that the reaction went better
with the lower temperature and higher ratio of base and
acetonitrile/the substrate. Notably, an excellent yield (77%,
Table 1, entry 5) was achieved at −75 °C with the presence of
4 equiv of n-butyllithium and acetonitrile. To explore the
viability of other base, the reactions were carried out with
different alkalis, but the outcome was disappointing with
reduced yield or complicated products. For example, when
lithium diisopropylamide was used, the yield was declined to
67% (Table 1, entry 7). Even lower yield (33%) was obtained
when the reaction was performed at −45 °C (entry 8).
Furthermore, in the tests with the nonorganolithium base (e.g.,
t-BuOK, NaH or NaOEt), either the reaction did not go (the
starting material 2 was recovered), or it gave too complicated
products to purify (Table 1, entries 9−12).
With the nucleophilic addition product 11 in hand, the next
step was the reduction. In order to convert compound 11 into
the ethylamine 14, a series of reactions including debromina-
tion, dehydration, olefin reduction, and cyano reduction was
involved (Scheme 3). The catalytic hydrogenation was expected
to complete all reactions in one-pot procedure. The reduction
of the cyano group should be critical and might require high
pressure of hydrogen. In order to avoid the additive 17 from
the primary amine and the imine in the process, acidic (HOAc/
H2O or HOAc/alcohols) or basic solvent (NH3/alcohols)
should be of the first choice.14,15
Recently, Zhang et al.6 and Fu et al.7 reported new
approaches to synthesize ramelteon 1 by asymmetric Michael
addition, and both of them required long reaction steps and
adopted expensive chiral auxiliaries. However, as the practical
synthesis, the key part is the installation of the side chain into
the three-fused-ring system at the benzylic position. The
Wittig−Horner reaction was the only way to complete the
installation so far found in the literature.8−12 As outlined in
Scheme 1, it requires different kinds of phosphorous ylide,
which is usually prepared from a trialkyloxyphosphine and
chloroacetate or chloracetonitrile.13 In Addition, more than one
step of reduction were employed in both existing routes, and
they required either long preparation steps or complicated
procedures. Most of all, they all suffer from very low yields.
Our group was interested in exploiting a novel method to
install the side chain without involving any Wittig reaction.
Since the side chain of key intermediate 8 was composed of two
carbons and one nitrogen atom, we attempted to construct the
skeleton with acetonitrile as the building block. After the
cyanomethyl was added to the ketone 2 with acetonitrile as a
nucleophilic reagent, amino-compound 8 could be obtained via
a series of reduction by catalytic hydrogenation. Herein, a novel
and optimized synthesis of ramelteon was established with
short preparation steps and cheap reagents (Scheme 2).
If the hydrogenation in ambient atmosphere was carried out
in the presence of Pd/C as catalyst in CH3COOH/H2O (v/v =
Received: December 12, 2014
© XXXX American Chemical Society
A
DOI: 10.1021/op500386g
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Technical Note
Scheme 1. Existing process I and II8,10
1:1) at 60 °C for 18 h, the debromination product 15 was
obtained in a yield of 77% (Table 2, entry 1). When hydrogen
pressure was elevated to 0.5 MPa, the dehydration product 9
was isolated along with 15 (Table 2, entry 2). Fortunately, the
desired product 14 was obtained when the pressure was
increased to 2.0 MPa, and the best yield (89%) was
encountered at 3.0 MPa (Table 2, entries 3 and 4).
Surprisingly, if the solvent (HOAc/H2O) was replaced by
HOAc/CH3OH, the reaction, carried out in the same time as in
entry 4, did not go to the end, giving a mixture of compound 9
and 12, terribly, together with the additive 17. Under the same
condition, the basic solvent resulted in a lower yield of 12
(Table 2, entries 8 and 9). To assess the alternative catalyst,
hydrogenation in CH3COOH/H2O or NH3/CH3OH at 3.0
MPa with Raney nickel were carried out, and the dehydration
compound 9 was the main product (Table 2, entries 10 and
11).
As a base, compound 14 was a slightly yellow oil and
unstable in ambient conditions. In general, salt formation is an
easy way to get the stable solid. We did obtain the
B
DOI: 10.1021/op500386g
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Technical Note
Scheme 2. New route
a
Table 1. Optimization of nucleophilic addition
hydrochloride or hydrobromide of 14 as a stable solid.
However, since the next step was the resolution where the
racemic amine 14 would react with an acidic resolution agent to
deliver the desired (S)-isomer product, dibenzoyl-L-tartaric acid
was selected both as an acid to form the salt 12 in the end of
hydrogenation and as the resolution agent. After recrystalliza-
tion of 12, the (S)-amino salt 13 was obtained. Eventually,
target compound 1 was converted from 13 via propionylation
with propionyl chloride.
CH3CN
(equiv)
temperature
b
entry base (equiv)
(°C)
yield (%)
1
n-BuLi (2)
n-BuLi (2)
n-BuLi (2)
n-BuLi (3)
n-BuLi (4)
n-BuLi (6)
2
2
2
3
4
6
4
4
4
4
4
−15
−45
−75
−75
−75
−75
−75
−45
12
28
43
59
77
56
67
33
CONCLUSION
■
2
In summary, the novel and practical synthesis of ramelteon
offers distinctive advantages over the reported routes.
Beginning with the same material, our new synthetic routes
only included three reactions, such as nucleophilic addition,
hydrogenation, and propionylation with an overall yield of 22%,
which was almost twice as much as those existing processes.
The conventional Wittig−Horner reaction was abolished, with
no phosphorous ylide, but acetonitrile was used as the
nucleophilic reagent. Besides, four reduction processes were
successfully implemented into one step of hydrogenation.
3
4
5
6
c
7
LDA (4)
c
8
LDA (4)
d
9
t-BuOK (4)
NaH (4)
25
complex mixture
complex mixture
d
10
11
25
NaOEt (4)
25
recovery of starting
material 2
12
NaOEt (4)
4
60
complex mixture
EXPERIMENTAL SECTION
a
b
■
The reaction was carried out in tetrahydrofuran for 2 h. Isolated
yield. LDA = lithium diisopropylamide. Recovery of starting material
2 below 0 °C.
c
d
General. Unless otherwise noted, stated reagents were
commercially available and used without purification. NMR
data were obtained using either Bruker V-400 or Varian
Scheme 3. Processes in hydrogenation
C
DOI: 10.1021/op500386g
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Technical Note
a
solution of 11 (12g 0.03 mol) in acetic acid (60 mL) and
H2O (60 mL) was subjected to hydrogenation in the presence
of palladium 10% on carbon (1.8 g) in a hydrogen pressure (3.0
MPa) at 60 °C for 18 h. After hydrogen absorption ceased, the
catalyst was removed by filtration. The filtrate was concentrated
to dryness, and the residue was neutralized with 10% NaOH
(aq). The mixture was extracted with 4 × 30 mL ethyl acetate,
dried over MgSO4, and filtered. Then, dibenzoyl-L-tartaric acid
(11.517 g, 0.03 mol) was added to the filtrate and stirred at
room temperature for 1 h and filtered to afford a white solid of
2-(1,6,7,8-tetrahydro-2H-indeno[5,4-b]furan-8-yl)ethylamine
dibenzoyl-L-tartaric acid (12) (16.063 g, 89% yield).
Table 2. Optimization of catalytic hydrogenation
pressure
product
c
b
entry
solvent (v/v)
catalyst
Pd/C
(MPa)
(yield, %)
1
CH3COOH/H2O
(1:1)
atm
15 (77)
2
3
4
5
CH3COOH/H2O
(1:1)
Pd/C
Pd/C
Pd/C
Pd/C
0.5
2.0
3.0
3.0
15, 9
The above 12 (11.220 g, 0.02 mol) was recrystallized with
acetonitrile: ethyl alcohol = 5:4 to afford a white solid of (S)-2-
(1,6,7,8-tetrahydro-2H-indeno[5,4-b]furan-8-yl)-ethylamine di-
benzoyl-L-tartaric acid (13) (4.151 g, 37% yield, 98.6% ee). Mp:
CH3COOH/H2O
(1:1)
12 (49)
12 (89)
12 (87)
CH3COOH/H2O
(1:1)
1
1.69−171 °C. H NMR(400 MHz, DMSO-d6): δ 1.60 (m,
CH3COOH/H2O
(2:1)
2H); 2.05 (m, 2H); 2.71 (m, 4H); 3.07 (m, 3H); 4.45 (m, 2H);
5.66 (s, 2H,); 6.54 (d, J = 8, 1H); 6.89 (d, J = 8, 1H); 7.50 (m,
4H); 7.63 (m, 2H); 7.93 (d, J = 4, 4H); 8.00 (br s,2H). MS (ES
6
7
CH3COOH/H2O(1:2) Pd/C
3.0
3.0
12 (73)
CH3COOH/CH3OH
(1:1)
Pd/C
9, 17, 12
+
+): m/z 204 (M + H) . [α]D −115.2 (c = 1.0, MeOH).
Preparation of (S)-N-[2-(1,6,7,8-Tetrahydro-2H-indeno-
[5,4-b]furan-8-yl)ethyl]propionamide (1). To a solution of
13 (4 g, 7 mmol) in tetrahydrofuran (40 mL) was added 20%
NaOH (aq) (1.120 g, 28 mmol) and stirred at −10 °C. Then
propionyl chloride (0.777 g, 8.4 mmol) was added dropwise at
−10 °C. After that, the reaction mixture was stirred at room
temperature for 2 h, and ethyl acetate (40 mL) and water (40
mL) were added. After separation, the aqueous layer was
extracted with 2 × 30 mL ethyl acetate. The organic layer was
washed with brine and dried over MgSO4. The filtrate was
concentrated to give a solid, which was recrystallized from
ethanol and water to afford a white solid of 1 (1.570 g, 85%
yield, 99.8% ee). Purity by HPLC 99.6%. Mp: 115−116 °C
8
9
NH3(13%)/CH3OH
Pd/C
Pd/C
3.0
3.0
12 (52)
12 (54)
NH3(13%)/
CH3CH2OH
10 NH3(13%)/CH3OH
Raney-Ni
Raney-Ni
3.0
3.0
9, 12
9, 12
11 CH3COOH/H2O
(1:1)
a
b
Reaction run at 60 °C for 18 h. 15 wt % Pd/C or 20 wt % Raney-Ni.
c
Isolated yields.
INOVA-400 instrument at 400 mHz for 1H and 100 Hz for 13
C
in either CDCl3 or DMSO-d6. The chemical shift were reported
in δ ppm relative to TMS. The melting points were determined
by Buchi M-565 apparatus. Specific optical rotations were
performed using Rudolph Automatic Polarimeter IV, at
measurement wavelength: 589 nm. HPLC was used to establish
the purity of compound 1 using Dionex U3000; HPLC
column: Symmetry Shield RP C18 4.6 mm × 250 mm, particle
size 5 μm, 1 mL/min flow, detection at 289 nm. The
enantiomeric purity of compound 13 and compound 1 was also
monitored by HPLC (HPLC column: Diacel Chiralcel OD-RH
4.6 mm × 150 mm, 5 μm). Further information is also shown
in the Supporting Information.
Preparation of 8-Hydroxy-4,5-dibromo-(1,2,6,7-tetrahy-
dro-8H-indeno[5,4-b]furan-8-yl)acetonitrile (11). A 500 mL
flask was charged with anhydrous tetrahydrofuran (200 mL)
and n-butyllithium in hexane 2.5 M (96.4 mL, 0.24 mol) was
added at −75 °C under a nitrogen atmosphere, followed by a
solution of acetonitrile (9.881g, 0.24 mol) in tetrahydrofuran
(50 mL). After stirring for 30 min, compound 2 (20 g, 0.06
mol) was added to the resulting white suspension and stirred
for 2 h at −75 °C. The yellow reaction mixture was poured into
500 mL ice−water. The aqueous mixture was extracted with 2
× 200 mL ethyl acetate, dried over MgSO4, and filtered. The
organic layer was evaporated under vacuum to give the crude
product of 11 as a light yellow solid. It was recrystallized with
toluene to give the desired product as a white solid (17.298 g,
77% yield). Mp: 174−175 °C. 1H NMR(400 MHz, DMSO-d6):
δ 2.16 (m, 1H); 2.32 (m, 1H); 2.84 (m, 2H); 3.08 (dd, J = 1.6,
2H); 3.40 (m, 2H); 4.66 (m, 2H,); 5.94 (br s, 1H). MS (ES+):
m/z 396 (M + Na)+.
8
1
(113−115 °C in literature ). H NMR(400 MHz, CDCl3): δ
1.39 (t, 3H); 1.63 (m, 1H); 1.83 (m, 1H); 2.02 (m, 1H); 2.16
(dd, J = 8, 2H); 2.28 (m, 1H); 2.78 (m, 1H); 2.83 (m, 1H);
3.14 (m, 1H); 3.22 (m, 2H); 3.33 (m, 2H); 4.54 (m, 2H); 5.38
(br s, 1H); 6.61 (d, J = 8, 1H); 6.97 (d, J = 8, 1H). 13C
NMR(100 MHz, CDCl3): δ 173.85, 159.56, 143.26, 135.92,
123.52, 122.28, 107.56, 71.26, 42.37, 38.17, 33.66, 31.88, 30.82,
29.86, 28.73, 10.01. MS (ES+):m/z 282 (M + Na)+. [α]D −57.3
(c = 1.0, CHCl3, −57.8 in literature8). Anal. (C16H21NO2) Calc:
C, 74.10; H, 8.16; N, 5.40; found: 74.09; H, 8.17; N, 5.47.
Preparation of 8-Hydroxy -(1,2,6,7-tetrahydro-8H-indeno-
[5,4-b]furan-8-yl)acetonitrile (15). A solution of 11 (2 g, 0.005
mol) in acetic acid (10 mL) and H2O (10 mL) was subjected to
hydrogenation in the presence of palladium 10% on carbon (1.8
g), while hydrogen was bubbled into the reaction mixture at 60
°C for 18 h. Then, the catalyst was removed by filtration. The
filtrate was concentrated into dryness. Water was added into
the residue and then extracted with 3 × 10 mL ethyl acetate.
The organic layer was washed with brine, dried over MgSO4,
and filtered. After concentration, the residue was separated by
column chromatography with ethyl acetate−hexanes 1:9, and
compound 15 was obtained as a white solid (0.813 g, 77%
yield). Mp: 122−123 °C. 1H NMR (400 MHz, CDCl3): δ 2.21
(m, 1H); 2.52 (m, 1H); 2.80 (dd, J = 16, 2H); 2.91 (m, 2H),
3.32 (m, 2H); 4.57 (m, 2H); 6.73 (d, J = 8, 1H); 6.97 (d, J = 8,
1H). MS (ES+):m/z 238 (M + Na)+.
Preparation of 2-(1,2,6,7-Tetrahydro-8H-indeno[5,4-b]-
furan-8-ylidene)acetonitrile (9). A solution of 11 (2 g, 0.005
mol) in acetic acid (10 mL) and H2O (10 mL) was subjected to
Preparation of (S)-2-(1,6,7,8-Tetrahydro-2H-indeno[5,4-
b]furan-8-yl)ethylamine Dibenzoyl-L-tartaric Acid(13). A
D
DOI: 10.1021/op500386g
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Technical Note
(9) Kansal, V. K.; Mistry, D. N.; et al. Process for the Synthesis of
Ramelteon and Its Intermediates. PCT Int. Appl. WO 2010045565,
April 22, 2010.
(10) Yamano, T.; Yamashita, M.; Adachi, M.; Tanaka, M.;
Matsumoto, K.; Kawada, M.; Uchikawa, O.; Fukatsu, K.; Ohkawa, S.
Tetrahedron: Asymmetry 2006, 17, 184−190.
(11) Padi, P. R.; Polavarapu, S.; et al. Process of Making Ramelteon
and Related Substances. PCT Int. Appl. WO 2008150953, December
11, 2008.
(12) Rao, D. R.; Kankan R. N.; et al. A Novel Process for Synthesis of
Ramelteon, and Key Intermediates for the Synthesis of Ramelteon.
PCT Int. Appl. WO 2012035303, March 22, 2012.
(13) Lan, L.; Zhan, X.; Qin, W.; Liu, Z.; Mao, Z. Heterocycles 2014,
89, 375−397.
(14) Schreifels, J. A.; M, P. C.; Swartz, W. E. J. Org. Chem. 1981, 46,
1263−1269.
(15) Shares, J.; Yehl, J.; Kowalsick, A.; Byers, P.; Haaf, M. P.
Tetrahedron Lett. 2012, 53, 4426−4428.
hydrogenation in the presence of palladium 10% on carbon (1.8
g) in a hydrogen pressure of 0.5 MPa at 60 °C for 18 h. Then,
the catalyst was removed by filtration. The filtrate was
concentrated to dryness. Water was added into the residue
and then extracted with 3 × 10 mL ethyl acetate. The organic
layer was washed with brine, dried over MgSO4, and filtered.
After column chromatography with 5% ethyl acetate in hexanes,
compound 9 was obtained as a yellow solid (0.354 g, 34%
yield). Mp: 149−150 °C (146−151 °C in literature ). H
NMR (400 MHz, CDCl3): δ 3.03 (t, 2H); 3.11 (m, 2H); 3.30
(t, 2H); 4.66 (t, 2H); 5.44 (t, 1H); 6.85 (d, J = 8, 1H); 7.10 (d,
J = 8, 1H). MS (ES+):m/z 198 (M + H)+.
10
1
Preparation of 2,2′-Bis(1,2,6,7-tetrahydro-8H-indeno[5,4-
b]furan-8-yl)-diethylamine (17). A solution of 11 (2 g, 0.005
mol) in acetic acid (10 mL) and CH3OH (10 mL) was
subjected to hydrogenation in the presence of palladium 10%
on carbon (1.8 g, 15 wt %) in a hydrogen atmosphere (3.0
MPa) at 60 °C for 18 h. Then, the catalyst was removed by
filtration. The filtrate was concentrated to dryness. Water was
added into the residue. The mixture was neutralized with 10 wt
% NaOH (aq), extracted with 4 × 30 mL ethyl acetate, dried
over MgSO4, and filtered. After that, compound 17 was
separated by column chromatography (5% methanol in
dichloromethane) as a white solid (0.259 g). Mp: 225−227
1
°C. H NMR (400 MHz, CDCl3): δ 1.69 (m, 2H); 1.99 (m,
2H); 2.24 (m, 2H); 2.40 (m, 2H); 2.82 (m, 8H); 3.09 (m, 4H);
3.28 (m, 2H); 4.50 (m, 4H); 6.58 (d, J = 8, 2H); 6.90 (d, J = 8,
2H); 9.68 (br s, 1H). MS (ES+):m/z 390 (M + H)+.
ASSOCIATED CONTENT
■
S
* Supporting Information
Detailed experimental procedures and spectroscopic character-
ization for all new compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Telephone: +86 21 55514600. E-mail: zhouweicheng58@163.
com.
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) Koike, T.; Hoashi, Y.; Takai, T.; Nakayama, M.; Yukuhiro, N.;
Ishikawa, T.; Hirai, K.; Uchikawa, O. J. Med. Chem. 2011, 54, 3436−
3444.
(2) Zlotos, D. P.; Jocker, R.; Cecon, E.; Rivara, S.; Witt-Enderby, P.
A. J. Med. Chem. 2014, 57, 3161−3185.
(3) Kato, K.; Hirai, K.; Nishiyama, K.; Uchikawa, O.; Fukatsu, K.;
Ohkawa, S.; Kawamata, Y.; Hinuma, S.; Miyamoto, M. Neuro-
pharmacology 2005, 48, 301−310.
(4) Miyamoto, M. CNS Neurosci. Ther. 2009, 15, 32−51.
(5) Srinivasan, V.; Pandi-Perumal, S. R.; Trahkt, I.; Spence, D. W.;
Poeggeler, B.; Hardeland, R.; Cardinali, D. P. Int. J. Neurosci. 2009, 6,
821−846.
(6) Zhang, X.; Yuan, W.; Luo, Y.; Huang, Q.; Lu, W. Heterocycles
2012, 85, 73−84.
(7) Fu, X.; Guo, X.; Li, X.; He, L.; Yang, Y.; Chen, Y. Tetrahedron:
Asymmetry 2013, 24, 827−832.
(8) Uchikawa, O.; Fukatsu, K.; Tokunoh, R.; Kawada, M.;
Matsumoto, K.; Imai, Y.; Hinuma, S.; Kato, K.; Nishikawa, H.; Hirai,
K.; Miyamoto, M.; Ohkawa, S. J. Med. Chem. 2002, 45, 4222−4239.
E
DOI: 10.1021/op500386g
Org. Process Res. Dev. XXXX, XXX, XXX−XXX