A new industrial process for 10-methoxyiminostilbene: Key intermediate for the synthesis of oxcarbazepine

Article abstract of DOI:10.1021/op900127v

A new industrial process, involving only two isolation and drying steps, for 10-methoxyiminostilbene (MISB), an advance intermediate of widely prescribed antiepileptic drug, oxcarbazepine, has been developed. A salient feature of this process is the novel use of 1,3-dibromo-5,5-dimethylhydantoin (DBDMH) to afford bro-mohydrin methyl ether from N-acetyliminostilbene. The byproducts of this process namely acetic acid, 5,5-dimethylhydantoin and Et₃N HBr are recyclable as well as nontoxic This process is amenable for the large-scale production of MISB.

Full text of DOI:10.1021/op900127v

Organic Process Research & Development 2009, 13, 870–874
A New Industrial Process for 10-Methoxyiminostilbene: Key Intermediate for the
Synthesis of Oxcarbazepine
Harnam Singh,† Nitin Gupta,† Pramod Kumar,† Sushil K. Dubey,† and Pawan K. Sharma*,‡
Jubilant Organosys Limited, Chemical and Pharma Research Department, C-26, Sector-59,
NOIDA, Uttar Pradesh - 201301, India, and Department of Chemistry, Kurukshetra UniVersity,
Kurukshetra-1360119, Haryana, India
Abstract:
key starting material, and it involves only a single chemical
step employing the reaction of 3 with phosgene,7a or urea7b or
sodium cyanate in acetic acid.7c There are several efficient
synthetic routes8-10 available for the synthesis of 3 which
enabled the supremacy of 3 as the most logical, economically
viable starting material for 1. Due to the structural similarity of
1 and 2, iminostilbene 3 is considered to be also the logical
starting material for the production of oxcarbazepine 2. A strong
demand for OCBZ 2 led to the development of several synthetic
schemes for its synthesis11,12 including those involving the
transformation, 3 f 2,12 albeit involving multistep procedures
using reagents such as phosgene, cyanogen halides, halogens,
etc. which are known to be potentially hazardous at the
commercial scales of production besides being environmentally
degrading. Search for a practically simpler method led Novartis
Pharma to design two protocols without involving iminostilbene
3 as the key starting material, employing either remote meta-
lation13 or Friedel-Crafts acylations14 as key steps in the
construction of the tricyclic skeleton. The scale-up process for
A new industrial process, involving only two isolation and drying
steps, for 10-methoxyiminostilbene (MISB), an advance intermedi-
ate of widely prescribed antiepileptic drug, oxcarbazepine, has
been developed. A salient feature of this process is the novel use
of 1,3-dibromo-5,5-dimethylhydantoin (DBDMH) to afford bro-
mohydrin methyl ether from N-acetyliminostilbene. The byprod-
ucts of this process namely acetic acid, 5,5-dimethylhydantoin and
Et3N·HBr are recyclable as well as nontoxic. This process
is amenable for the large-scale production of MISB.
Introduction
Carbamazepine (CBZ, 1) is a well-known anticonvulsant and
mood-stabilizing drug which is also used for the treatment of
trigeminal neuralgia and acute mania and for prophylactic
treatment in bipolar disorder.1 CBZ possesses some structural
similarity to the tricyclic antidepressant drug imipramine, and
in fact, CBZ was first synthesized as a potential antidepressant.
Oxcarbazepine (OCBZ, 2) is a relatively novel second-genera-
tion antiepileptic drug primarily used for the treatment of
psychometric disturbance,2 epilepsy, and trigeminal neuralgia,3
and for Parkinson’s disease.4 In fact 2 was initially isolated as
an active metabolite of 1,5 and it is structurally a derivative of
carbamazepine, adding an extra oxygen atom on the dibenza-
zepine ring. This small difference helps reduce the impact on
the liver of metabolizing the drug. OCBZ 2 has assumed the
status of the most widely prescribed drug for the treatment of
epilepsy due to its improved tolerability profile as compared to
that of CBZ 1, but without loss in its therapeutic potency.6
(3) Gomez-Arguelles, J. M.; Dorado, R.; Sepulveda, J. M.; Huet, R.;
Arrojo, F. G.; Aragon, E.; Herrera, A.; Trron, C.; Anciones, B. J. Clin.
Neurosci. 2008, 15, 516.
(4) Boireau, A.; Bordier, F.; Doble, A.; Dubedat, P.; Louvel, E.; Meunier,
M.; Miquet, J.-M.; Tutzmann, J.-M. U.S. Patent 5,658,900, 1997.
(5) Lertratanangkoon, K.; Honing, M. G. Drug Metab. Dispos. 1982, 10,
1.
(6) Clemens, B.; Menes, A.; Nagy, Z. Acta Neurol. Scand. 2004, 109,
324.
(7) (a) Schindler, W. U.S. Patent 2,948,718, 1960. (b) Vyas, D. K.; Jafri,
S. W.; Kulkarni, K. A. U.S. Patent 6,245,908, 2001. (c) Eckardt, R.;
Janch, J. H. U.S. Patent 7,015,322, 2006.
(8) (a) Schramek, H.; Riethmann, J.; Kllen, J. Swiss Patent 442319, 1968.
(b) Craig, P. N.; Lester, B. M.; Saggiomo, A. J.; Kaiser, C.; Zirker,
C. L. J. Org. Chem. 1961, 26, 135.
(9) Knell, A.; Monti, D.; Maciejewski, M.; Baiker, A. Appl. Catal. 1995,
121, 139.
(10) (a) Lapworth, A. J. Chem. Soc. 1901, 79, 126. (b) Craig, P. N. U.S.
Patent 3,074,931, 1963; Chem. Abstr. 1963, 59, P3900f.
(11) (a) Fuenfschilling, P. C.; Zaugg, W.; Beutler, U.; Kaufmann, D.; Lohse,
O.; Mutz, J. P.; Onken, U.; Reber, J. L.; Shenton, D. Org. Process
Res. DeV. 2005, 9, 272. (b) Carril, M.; San, M. R.; Churruca, F.; Tellitu,
I.; Dominguez, E. Org. Lett. 2005, 7, 4787. (c) Lohse, O.; Beutler,
U.; Funfeschilling, P.; Furel, P.; Julien, F.; Kaufmann, D.; Penn, G.;
Zaugg, W. Tetrahedron Lett. 2001, 42, 385. (d) Ernst, A. European
Patent 0028028, 1981. (e) Gosteli, J. Swiss Patent 642950, 1984. (f)
Miller, L.; Ellinwood, A. E. U.S. Patent 3,642,775, 1972. (g)
Heckendorn, R.; Zergenyi, J.; Menard, F. Canadian Patent 1112241,
1981.
(12) (a) Manadakini, M.; Muthukumaran, N.; Thennati, R. WO 2005/
096709, 2005. (b) Aufderhaar, K. E.; Sprecher, B. K.; Zergenyi, S. J.
U.S. Patent 4,436,660, 1984. (c) Aufderhaar, K. E. U.S. Patent
4,452,738, 1984. (d) Aufderhaar, K. E. U.S. Patent 4,540,514, 1985.
(e) Aufderhaar, K. E. U.S. Patent 4,559,174, 1985. (f) Aufderhaar,
K. E. U.S. Patent 4,579,683, 1986. (g) Banti, M. A.; Bollini, R. D.;
Lombardia, R.; Serra, C. M. U.S. Patent 2007/0149507, 2007.
(13) Lohse, O.; Beutler, U.; Funfschilling, P.; Furet, P.; France, J.;
Kauffmann, D.; Penn, G.; Zaugg, W. Tetrahedron Lett. 2001, 42, 385.
Historically, CBZ 1 is synthesized primarily following
synthetic schemes utilizing the tricyclic iminostilbene 3 as the
* Author for correspondence: E-mail: talk2pawan@gmail.com. Telephone:
+91 94164 57355. Fax: +91 1744 238277.
† Jubilant Organosys Limited.
‡ Kurukshetra University.
(1) Ambrosio, A. F.; Soares-da-Silva, P.; Carvalho, C. M.; Carvalho, A. P.
Neurochem. Res. 2002, 27, 121–30.
(2) Vajda, F. J. E. J. Clin. Neurosci. 2008, 7, 88.
870
•
Vol. 13, No. 5, 2009 / Organic Process Research & Development
10.1021/op900127v CCC: $40.75  2009 American Chemical Society
Published on Web 08/20/2009

Scheme 1. Novartis Pharma’s synthetic process11a for the preparation of oxcarbazepinea
a Reagents and conditions: (a) PPA; (b) H2O, CH3OH; (c) PEG-200, NaOH; (d) NaOCN, CH3COOH; (e) HCl, HCOOH, H2O.
Scheme 2. Sun Pharma’s synthetic process12a for the preparation of oxcarbazepinea
a Reagents and conditions: (a) COCl2/toluene; (b) CH3COOH/Br2; (c) CH3ONa/CH3OH; (d) CH3ONa/DMF; (e) NaOCN/CH3COOH; (f) H2SO4.
the latter protocol has been successfully achieved11a,15 employing
1,3-dihydro-1-phenyl-2H-indol-2-one (4) as a starting material
which was converted into 2-[(methoxycarbonyl)phenylamino]-
benzeneacetic acid (5) in multistep procedures involving the
use of reagents such as phosgene or butyllithium (Scheme 1).
5 was made to participate in a Friedel-Crafts cyclization
strategy providing OCBZ 2 in five further steps. Although this
protocol provides ready access to 2 via intermediate 10-
methoxyiminostilbene (MISB, 8), it is too long in addition to
having some environmental and toxicological drawbacks.
Another synthetic procedure of industrial importance12a was
developed by Sun Pharma utilizing a multistep strategy involv-
ing halogenation-dehalogenation on iminostilbene 3 to generate
intermediate 12 (Scheme 2). The enol ether 10-methoxyimi-
nostilbene (MISB, 8) is prepared by nucleophilic substitution
of the vinylic bromide 12. MISB 8 could readily be converted
into OCBZ 2 by treatment with sodium cyanate in acetic acid.
However, the major disadvantages of this route are the use of
potentially lethal reagents in addition to generation of large
amounts of effluents.
A perusal of these synthetic processes revealed that 10-
methoxyiminostilbene (MISB, 8) is a key intermediate in
both these industrial processes11a,12a as well as several other
synthetic strategies,12g,16 and the transformation 8 f 2 is
rather simple, often employing a reaction with urea or
sodium cyanate which introduces the N-carboxamide
(14) Kauffmann, D.; Funfschilling, P.; Beutler, U.; Hoehn, P.; Lohse, O.;
Zaugg, W. Tetrahedron Lett. 2004, 45, 5275.
(16) (a) Parenky, C.; Chaturvedi, R. U.S. Patent, 10,032,647, 2007. (b)
Gutman, D.; Baidossi, W. U.S. Patent 2006/0241293, 2006. (c)
Sivakumar, B. V.; Bhirud, S. B.; Batchu, C.; Kale, S. A. U.S. Patent
7,459,553, 2008. (d) Milanese, A. WO 118550, 2005.
(15) Funfschilling, P.; Kauffmann, D.; Lohse, O.; Beutler, U.; Zaugg, W.
Chem. Abstr. 2001, 135, 166785. WO 01/56992, 2001.
Vol. 13, No. 5, 2009 / Organic Process Research & Development
•
871

Scheme 3. New process for the preparation of MISB (8) starting from iminostilbene (3)a
a Reagents and condition: (a) Ac2O, 120 °C, 3-4 h; (b) DBDMH, CH3OH, 6-8 h 5-15 °C, 86-92%; (c) Et3N, toluene, reflux, 24 h; (d) KOH/toluene, PEG-200,
reflux, 4-6 h, 85-91%.
group followed by acid-mediated ether cleavage. We
envisioned that an efficient route to intermediate 8 without
involving halogenated solvents or hazardous reagents
should make economic sense besides giving a new
practical industrial method within the acceptable envi-
ronmental and toxicological concerns. We further reasoned
that the new method would be more appealing if it can
employ commercially available conventional starting
material for CBZ and OCBZ; iminostilbene 3. In this
contribution, we report the design and technical aspects
of our new process,17 which enables the commercial
production of 8 from 3.
whereas at a temperature range of 120-125 °C the reaction is
complete in about 3-4 h.
After the completion of the acetylation, the volatiles
are removed under vacuum, the resulting viscous material
is dissolved in methanol at 40-45 °C and water is added
to the clear solution resulting in separation of solid
N-acetyliminostilbene (10) in 85-90% yield. Considering
our strategy of simultaneous functionalization of the
carbon-carbon double bond of the central ring of imi-
nostilbene with a leaving group and an ether functionality,
we attempted a reaction of 10 with liquid bromine in
methanol expecting to get a bromohydrin methyl ether
13. However, a mixture of 13 and 10,11-dibromo-N-
acetyliminostilbene was obtained. Reaction of 10 with
NBS in methanol was the next to be tried. and it resulted
in isolation of the expected 13 in 80-85% yield. Although
the yield of 13 in this case was satisfactory, however, the
reproducibility of yield was not up to the mark. We shifted
our attention to a cheap commercially available bromi-
nating agent, 1,3-dibromo-5,5-dimethylhydantoin (DB-
DMH) which is well-known to effect brominations of
aromatic rings.18,19 Thus, 10 is taken up in methanol (3-4
volumes) and treated with DBDMH at a temperature of
0-10 °C. After stirring for 1 h the product bromohydrin
methyl ether 13 starts crystallizing out; however, the
reaction takes 6-8 h for completion. The resulting slurry
is filtered to give 13 in about 88-92% yield.20 After a
fewtrials,wewereabletoperformthesetwostepssacetylation
and the formation of bromohydrin methyl ethersin tandem
Results and Discussion
The first task of our project was to remove or minimize the
use of halogenated solvents and reagents. However, it was
apparent that, in order to achieve an efficient synthesis, the best
option would be to find a suitable method that can functionalize
the carbon-carbon double bond of the central ring of iminos-
tilbene in such a way so as to install an ether functionality (such
as a OCH3 group) as well as a suitable leaving group in a single
step on the adjacent carbons at positions 10 and 11. Being a
symmetrical molecule, no regioselectivity issues are involved
here. The leaving group then can be made to participate in a
ꢀ-elimination process leading to the 10-methoxyiminostilbene
(8). As described in Scheme 3, our synthesis of 8 begins with
the protection of the imino nitrogen by heating iminostilbene
with slight excess (4.35 mol equiv) of acetic anhydride for 3 h
at 120-125 °C. The solubility of iminostilbene is poor in acetic
anhydride, thus warranting the use of over 4 equiv of acetic
anhydride which is used as a reagent as well as a solvent. It
was found that, at this concentration, a clear solution results as
the reaction temperature approaches 70 °C. TLC examination
of this clear solution indicates a clean reaction with partial
formation of 10. A temperature study of this reaction was carried
out that shows that the reaction is very slow below 100 °C,
(18) Alam, A.; Takaguchi, Y.; Tsuboi, S. J. Fac. EnViron. Sci.Technol.
2005, 10, 105; Chem. Abstr. 2005, 142, 409887.
(19) Chassaing, C.; Haudrechy, A.; Langlois, Y. Tetrahedron Lett. 1997,
38, 4415.
(20) The only byproduct at this stage is 5,5-dimethylhydantoin. Efforts were
made to recover 5,5-dimethylhydantoin from the resulting filtrate. The
filtrate is distilled to remove methanol, the resulting concentrate is
dissolved in water (5 volumes) at 30-35 °C and extracted with
dichloromethane (2 × 1.5 volumes). The aqueous phase is partially
distilled (2 volumes), and the residual aqueous layer is cooled to 10-
15 °C when a solid starts separating out. Filtration of the slurry affords
5,5-dimethylhydantoin in 50-60% yield.
(17) Gupta, N.; Singh, H.; Kumar, P.; Dubey, S. K. WO 141798, 2007.
872
•
Vol. 13, No. 5, 2009 / Organic Process Research & Development

without isolating the intermediate 10. Thus, volatiles are
removed under vacuum from the reaction mixture after
the completion of the acetylation reaction, and the residue
is taken up in methanol. The methanolic solution is treated
with 1,3-dibromo-5,5-dimethylhydantoin (DBDMH) for
6-8 h at 5-10 °C to afford, after filtration, 13 in 88-92%
overall yield from iminostilbene. Combining these two
steps offers significant advantages, as both steps neces-
sitate the use of a single reactor at the plant level, thus
lowering the reaction time and utility cost besides saving
the costs involved in isolation of 10. Moreover, the overall
yield from 3 to 13 is about 10-13% higher by a one-pot
procedure. Dehydrobromination of 13 to 14 was achieved
by treating it with triethylamine21 in toluene and heating
the resulting mixture under reflux for 24 h. As the reaction
proceeds, the byproduct Et3N · HBr keeps on precipitating
and is filtered at the end of the reaction at room
temperature. The remaining organic phase is washed with
water, excess toluene is evaporated, and the resulting thick
mass is dissolved in methanol at 50-55 °C. 14 is obtained
in 83-85% yield as it is precipitated out from the
methanolic solution after addition of water at room
temperature. Out of many other reaction conditions
attempted for dehydrobromination, the use of DBU is
worth mentioning, as the reaction takes only 30-40 min
to complete at room temperature. However, economic
considerations led us to adopt the route involving triethy-
lamine in spite of long reaction times. No further attempts
were made to optimize the dehydrobromination by other
processes involving inorganic bases, as they were liable
to generate more effluent. Deacetylation of 14 is achieved
by refluxing in n-butanol and KOH for 6-8 h. After
completion of the reaction, n-butanol is distilled off and
the residue is stirred with water for 1-2 h and filtered to
afford MISB 8 in 85-90% yield. We were apprehensive
about the feasibility of this step at the plant level as the
nature of the aqueous slurry at this stage is very thick
and thus difficult to handle, resulting in unloading
problems. We could achieve a clean deacetylation reaction
without any concomitant unloading problems by replacing
n-butanol with toluene in the presence of a catalytic
amount of PEG-200. This gave us the idea of combining
the two stepssdehydrobromination and deacetylation.
Thus, the reaction mixture obtained after Et3N/toluene
reflux in the dehydrobromination step is filtered to remove
Et3N · HBr. KOH and a catalytic amount of PEG-200 are
added to the filtrate, and the reaction mixture is refluxed
for another 4-6 h. The reaction mixture is washed with
water at room temperature, and the toluene is distilled
off completely to afford a yellow solid mass which is
dissolved in methanol at reflux. Water is added to the
cooled reaction mixture under stirring to precipitate the
pure 8, resulting in 88-90% overall yield from 13. By
combining the last two steps, the synthesis is again
economically rewarding in terms of increased yield,
reduced timings as well as workup hassles. The newly
developed protocol for 8 has been found to be a robust
process as it has achieved reproducible results on com-
mercial scale.
Conclusion
Starting from the conventional tricyclic starting material,
iminostilbene 3, the newly developed production process for
MISB 8 has an overall yield of around 80% and involves only
two isolation and drying steps, namely the intermediate 13 and
the target product 8. The new process offers distinctive
advantages over earlier reported procedures in terms of less
effluent generation, avoiding the use of liquid bromine as well
as significant cost reduction on commercial scale. To the best
of our knowledge, we have used the industrially important 1,3-
dibromo-5,5-dimethylhydantoin (DBDMH) for halohydrin ether
formation for the first time. The byproducts of this process,
namely acetic acid, 5,5-dimethylhydantoin, and Et3N·HBr, are
recyclable as well as nontoxic. This process is amenable for
the large-scale production of 8.
Experimental Section
Starting materials, reagents, and solvents were obtained from
commercial suppliers and were used without further purification.
Water used at any stage was deionized. All melting points were
uncorrected and determined on Buchi B-545 melting point
1
apparatus. NMR spectra (400 MHz for H and 100 MHz for
13C) were recorded on a Bruker Avance-400 instrument. Mass
spectra were recorded on LQC-Avatage Max-Ion Trap (Source:
APCI, ESI) instrument. Water content of all reagents, solvents,
and intermediates was done on Metrohm-Modetro-787 KF-
Titrino instrument. TLC was done on Merck TLC silica gel 60
F254 plates using mobile phase hexane/EtOAc (8:2).
Preparation of N-Acetyl-10-bromo-11-methoxy-10,11-
dihydro-5H-dibenzo[b,f]azepine (13). A mixture of acetic
anhydride (110 mL, 1.16 mol) and iminostilbene 3 (100 g, 0.518
mol) was heated to 120-125 °C for a period of 3-4 h. After
completion of reaction (TLC), volatiles were removed under
vacuum. Methanol (250 mL) was added to the residue and
cooled to 5-10 °C. Solid 1,3-dibromo-5,5-dimethylhydantoin
(DBDMH; 81.5 g, 0.284 mol) was added at 5-10 °C, and the
reaction mixture stirred for 6-8 h at 5-10 °C. The resulting
slurry was filtered, washed with chilled methanol, and dried
under vacuum to give the intermediate 13 (162 g, 90.6%): Rf
) 0.46; mp 135-138 °C; 1H NMR (400 MHz, CDCl3) δ 2.30
(s, 3H, -COCH3), 3.60 (s, 3H, OCH3), 5.08 (d, 1H, CHOMe),
5.21 (d, 1H, -CHBr), 7.26-7.59 (m, 8H, arom-H); 13C NMR
(100 MHz, CDCl3) δ 22.7, 54.8, 59.3, 83.8, 125.7, 126.0, 127.1,
127.4, 128.3, 128.6, 128.7, 128.9, 133.7, 133.8, 169.6; IR (KBr,
cm-1) 3066, 2948, 2932, 1671, 1598, 1438, 1372, 1125, 772,
602, 588 cm-1; MS (ES+) m/z: 346 (MH+) and 348 (MH+ +
2). Anal. Calcd for C17H16BrNO2: C, 58.97; H, 4.66; N, 4.05;
Found: C, 58.62; H, 4.76; N, 3.81.
Preparation of 10-Oxo-10,11-dihydro-5H-dibenzo[b,f]-
azepine-5-carboxamide (10-Methoxyiminostilbene; MISB;
8). A mixture of compound 13 (160 g; 0.463 mol), toluene (800
mL), and Et3N (95 g, 0.938 mol) was heated to reflux under
stirring for 24 h. After completion of reaction (TLC), the
reaction mixture was cooled to room temperature, and the
precipitated Et3N·HBr was filtered. A mixture of toluene layer,
(21) Charles, C. P.; Joseph, M. Org. Synth. 1965, 45, 22.
Vol. 13, No. 5, 2009 / Organic Process Research & Development
•
873

KOH (70 g, 1.24 mol), and PEG-200 (4.8 g, 3% w/w) was
refluxed for 4-6 h. After completion of reaction (TLC), the
reaction mixture was cooled to room temperature and washed
with water (2 × 100 mL). The organic layer was concentrated
to dryness. Methanol (300 mL) was added to the residue and
heated under reflux for 1 h. The resulting mixture was cooled
to room temperature, and water (100 mL) was added slowly
and stirred for 2 h. The resulting slurry was filtered, and the
solid was washed with chilled methanol (100 mL). The resulting
yellow solid was dried under vacuum to give 8 (91.1 g, 88%):
6.67-7.6 (m, 8H, arom-H); IR (KBr, cm-1) 3360, 1640, 1480,
1220, 1105, 779, 740; MS (ES+) m/z: 224 (MH+).
Acknowledgment
We thank the management of Jubilant Organosys Limited
for supporting and sponsoring us to do this work. The Analytical
Department of Jubilant is gratefully acknowledged for providing
support for analytical and spectral data.
Received for review May 16, 2009.
OP900127V
1
Rf ) 0.35; mp 120-124 °C; H NMR (400 MHz, CDCl3) δ
3.90 (s, 3H, -OCH3), 5.16 (s, 1H, NH), 5.89 (s, 1H, CH),
874
•
Vol. 13, No. 5, 2009 / Organic Process Research & Development