Efficient catalytic system for Ru-catalyzed C-H arylation and application to a practical synthesis of a pharmaceutical

Article abstract of DOI:10.1021/cs501328x

A series of K salts of sulfonic acids have been tested as a cocatalyst for Ru-catalyzed C-H arylation. Among them, K 2, 4, 6-trimethylbenzenesulfonate (TMBSK) was found to be most active, and generality of the reaction was confirmed for a variety of nitrogen-containing heterocycles to give corresponding functionalized biaryls in high yields. The present methodology was applied to a practical synthesis of Candesartan Cilexetil.

Full text of DOI:10.1021/cs501328x

Letter
pubs.acs.org/acscatalysis
Efficient Catalytic System for Ru-Catalyzed C−H Arylation and
Application to a Practical Synthesis of a Pharmaceutical
Masahiko Seki*
Process R&D Department, Healthcare Business Division II, API Corporation, The Kaiteki Building, 1-13-4, Uchikanda, Chiyoda-ku,
Tokyo 101-0047, Japan
S
* Supporting Information
ABSTRACT: A series of K salts of sulfonic acids have been
tested as a cocatalyst for Ru-catalyzed C−H arylation. Among
them, K 2, 4, 6-trimethylbenzenesulfonate (TMBSK) was
found to be most active, and generality of the reaction was
confirmed for a variety of nitrogen-containing heterocycles to
give corresponding functionalized biaryls in high yields. The
present methodology was applied to a practical synthesis of
Candesartan Cilexetil.
KEYWORDS: C−H activation, direct arylation, ruthenium, sulfonic acids, biaryls, heterocycles
In our initial study, the C−H arylation employing K
sulfonate as the cocatalyst was tested for the reaction of 1-
benzyl-2-phenyl-1H-tetrazole (4) with 4-bromobenzyl acetate
(5a) whose product 6a is a key common intermediate for
angiotensin II receptor blockers (ARBs) (Table 1).13,16−19,23
The reaction was conducted in the presence of [RuCl2(p-
cymene)]2 (0.5 mol %), PPh3 (2.0 equiv to Ru), cocatalyst (2.0
equiv to Ru), and K2CO3 in NMP at 138 °C for 6 h. The data
employing BEHPK is listed for comparison (Table 1, Entry
1).13 For water content in the reaction mixture, even adding a
small amount of water (>2% to NMP) resulted in no
conversion. This might be due to generation of 4-bromo-
benzylalcohol, which considerably retards the reaction.17 When
K methanesulfonate was employed as the cocatalyst, a high
monoarylation selectivity similar to BEHPK was obtained,
though the yield was declined (6a/7a = 94/6, 60% yield, Table
1, Entry 2, versus 6a/7a = 95/5, 82% yield, Table 1, Entry 1).
The use of aromatic K benzenesulfonate, K p-toluenesulfonate,
or K p-dodecylbenzene- sulfonate did not improve the yield
while the monoarylation selectivity was retained (6a/7a = 94/
6−97/3, 28−46% yield, Table 1, Entries 3, 4, and 5). Further
screen of the cocatalyst revealed that K 2, 4, 6-trimethylbenze-
nesulfonate (TMBSK) was found to provide a higher yield
which is similar to BEHPK (6/7 = 92/8, 77% yield, Table 1,
Entry 6). The amount of PPh3 in the reaction is significant.
Actually the conversion was considerably declined when the
amount of PPh3 was reduced to 1.0 equiv to Ru (18%, Table 1,
Entry 7, versus 86%, Table 1, Entry 6). Of particular interest is
the result when bulky benzoyl derivative 5b was employed as
the substrate where TMBSK provided a higher yield than
BEHPK (80%, Table 1, Entry 9 versus 68%, Table 1, Entry 8).
he Ru-catalyzed C−H arylation reaction has received keen
Tinterest as a highly atom-economical approach to prepare
biaryl compounds (1 to 2, Scheme 1).1−7 The new approach is
able to construct various biaryls of pharmaceutical and material
importance without use of any activating group or stoichio-
metric amount of hazardous organometallic compounds which
are required for the conventional cross coupling reactions. The
C−H arylation is important not only from a theoretical
viewpoint but also from a commercial application, although the
use of the methodology is still early in its development in
various industries. In order to promote the catalytic system
efficiently, the proper choice of a cocatalyst is significant, and
extensive research has been conducted to discover the correct
catalyst. Among them, K salts of Brønstead acids such as acetic
acid,8,9 pivalic acid,10,11 adamantanecarboxylic acid,12 and
mesitylenecarboxylic acid12 have been employed. Very recently,
we have disclosed that K bis(2-ethylhexyl)phosphate (BEHPK)
exerted high activity in the C−H arylation.13 However, BEHPK
is not completely satisfactory, because the performance is not
good for less reactive 2-phenyloxazoline derivatives13 and bulky
substrates such as 1-benzyl-5-phenyl-1H-tetrazole and 4-
bromobenzyl benzoate (vide infra). Reactivity of the cocatalyst
was attributed to a proper ligand excahnge by means of
hydrogen bonding.14 This type of activation might be enhanced
by using a stronger acid. Hence, in a search for a better
cocatalyst which has higher activity and versatility to
compensate for the lack of BEHPK, the author came up with
an idea to employ K sulfonate as the cocatalyst whose conjugate
acid has a lower pKa value than the corresponding phosphoric
and carboxylic acids (RSO3H > RP(O)(OH)3 > RCO2H).15
Disclosed herein is a novel and highly efficient catalytic system
of the C−H arylation which involves 2, 4, 6-trimethylbenzene-
sulfonate (TMBSK) as a cocatalyst and application of the
protocol to a practical of Candesartan Cilexetil (3).
Received: September 4, 2014
Revised: October 8, 2014
Published: October 9, 2014
© 2014 American Chemical Society
4047
dx.doi.org/10.1021/cs501328x | ACS Catal. 2014, 4, 4047−4050

ACS Catalysis
Letter
Scheme 1. C−H Arylation of Arenes and Candesartan Cilexetil
a
Table 1. Screen of a Cocatalyst for C−H Arylation of 1-Benzyl-2-phenyl-1H-tetrazole (4)
b
b
c
entry
R
cocatalyst
6/7
conv (%)
yield of 6 (%)
15
Me
Me
Me
Me
Me
Me
Me
Ph
Bis(2-Et-HexO)2P(O)OK (BEHPK)
95/5
94/6
97/3
94/6
94/6
92/8
98/2
92/8
89/11
87
75
35
60
59
86
18
77
83
82
60
28
50
46
77
14
68
80
2
MeSO3K
3
PhSO3K
4
4-MePhSO3K
4-n-C12H25SO3K
2, 4, 6-Me3PhSO3K (TMBSK)
TMBSK
5
6
d
7
8
9
BEHPK
Ph
TMBSK
a
Reactions were conducted by employing 4 (2.0 g, 8.46 mmol), 5 (2.13 g, 9.31 mmol, 1.1 equiv to 4), [RuCl2(p-cymene)]2 (26 mg, 0.0423 mmol,
0.5 mol %), PPh3 (44 mg, 0.168 mmol, 2.0 equiv to Ru), cocatalyst (0.168 mmol, 2.0 equiv to Ru), K2CO3 (1.17 g, 8.46 mmol, 1.0 equiv to 4) in
b
c
d
NMP (10 mL) at 138 °C for 6 h. Determined by HPLC. Assay yield. PPh3 (22 mg, 0.084 mmol, 1.0 equiv to Ru).
a
Table 2. C−H Arylation of Arenes 8 in the Presence of TMBSK
a
Reactions were conducted by employing 8 (6.44 mmol), 9 (14.2 mmol, 2.2 equiv), [RuCl2(p-cymene)]2 (20 mg, 0.0322 mmol, 0.5 mol %), PPh3
(34 mg, 0.129 mmol, 2.0 equiv to Ru), cocatalyst (27 mg, 0.129 mmol, 2.0 equiv to Ru), K2CO3 (0.89 g, 6.44 mmol, 1.0 equiv) in NMP (5 mL) at
b
c
138 °C for 6 h. For characterization data of 10 and 11, see ref 13. Determined by HPLC. Assay yield.
4048
dx.doi.org/10.1021/cs501328x | ACS Catal. 2014, 4, 4047−4050

ACS Catalysis
Letter
Scheme 2. Synthesis of Candesartan Cilexetil (3)
The benzoyl derivative 6b has a practical advantage over acetyl
derivative 6a in terms of better crystallinity. It remarkably
facilitates the isolation without silica gel column chromatog-
raphy (vide infra).
conducted in DMF, a poor selectivity was obtained (N1/N3 =
2:1). However, gratifyingly, the selectivity was considerably
improved when the reaction was conducted in 2-propanol
(IPA) (N1/N3 = 9.6:1). Practically, the regioisomer needs not
be separated at this stage, and the crude product 14 was
subjected to hydrolysis and purified by forming crystalline N,
N-dicyclohexylamine (DCHA) salt. By such a simple
procedure, the undesired N3-regioisomer was removed
completely to provide 15 in high yield and high purity (67%
based on 12). Furthermore, the DCHA salt underwent
esterification directly with cilexetil chloride to give an ester
16 in high yield (94%). The final intermediate 16 was quite
smoothly deprotected by our previously developed debenzyla-
tion protocol employing Rosenmund catalyst to furnish
Candesartan Cilexetil (3) in a high yield (80%).23
The reaction with TMBSK was further tested employing
various nitrogen-containing heterocycles, as shown in Table 2.
The use of pyridine and pyrazole derivatives provided arylated
products in high yields (Table 2, Entries 1−6). It should be
noted TMBSK provided oxazoline derivatives in much higher
conversions than BEHPK (100%, Table 2, Entries 7 and 9,
versus 45% and 50%, Table 2, Entries 8 and 10). Practically,
TMBSK, whose corresponding free sulfonic acid is commer-
cially available as a stable dihydrate, is nonhygroscopic, and its
handling is much easier than very hygroscopic BEHPK.
The present method was applied to a synthesis of
Candesartan Cilexetil (3), one of the most potent anti-
hypertension drugs.20,21 In our previous synthesis of 3,18 acetyl
derivative 6a was employed as the intermediate. However, in an
improved process, a bezoyl derivative 6b was chosen as the
intermediate, because it has a practical advantage over 6a on
ease of crystallization. Considering the entire route of synthesis
of 3, the previous methods consist of a long linear-type pathway
where benzimidazole moiety was formed consecutively in the
main reaction sequence.20,21 To avoid the impracticality,22
preformed benzimidazole 13 was coupled with a chloride 12,
which was readily prepared from 6b via debenzoylation and
chlorination (Scheme 2). The regioselectivity of the alkylation
of 13 with 12 is significant because undesired N3-alkyl
derivative might be formed and carried over to the final active
pharmaceutical ingredient (API). When the reaction was
The quality of 3 obtained was extremely high because the
deprotection is carried out under neutral conditions in marked
contrast to the previous method (removal of trityl group),
which needs strong acid.20,21
The present synthesis of 3 is innovative and has the
advantages shown below:
1. The biphenyl core is produced by means of C−H
arylation and hence has high atom economy22 and
sustainability: atom efficiency24 of the present process is
much higher than previous synthesis using cross coupling
reaction (61% versus 34%):
2. The number of steps in the present synthesis is 7,
whereas that of the previous one is 10.
4049
dx.doi.org/10.1021/cs501328x | ACS Catal. 2014, 4, 4047−4050

ACS Catalysis
Letter
(6) Ackermann, L. Chem. Rev. 2011, 111, 1315−1345.
(7) Arockiam, P. B.; Bruneau, C.; Dixneuf, P. H. Chem. Rev. 2012,
112, 5879−5918.
(8) Pozgan, F.; Dixneuf, P. H. Adv. Synth. Catal. 2009, 351, 1737−
1743.
(9) Li, B.; Bheeter, C. B.; Darcel, C.; Dixneuf, P. H. ACS Catal. 2011,
1, 1221−1224.
(10) Arockiam, P. B.; Poirier, V.; Fischmeister, C.; Bruneau, C.;
Dixneuf, P. H. Green Chem. 2009, 11, 1871−1875.
(11) Arockiam, P. B.; Fischmeister, C.; Bruneau, C.; Dixneuf, P. H.
Angew. Chem., Int. Ed. 2010, 49, 6629−6632.
(12) Ackermann, L.; Lygin, A. V. Org. Lett. 2011, 13, 3332−3335.
(13) Seki, M. RSC Adv. 2014, 4, 29131−29133.
3. The present synthesis is highly convergent using a key
common intermediate 12, from which every ARB is
synthesized.
4. It involves easy operation and no need of cryogenic
conditions and is able to be conducted under quite mild
conditions using very common multipurpose facilities
available all over the world.
5. It does not need expensive and/or hazardous reagents.
6. The quality of 3 is extremely high to prove complete
equivalence with previous process.
7. Ru and Pd content in the final API produced by the
present method were less than the detection limit, which
was assayed by IPC analysis. Recovery of Ru catalyst was
not tested because of the low price of the catalyst.
8. The process is remarkably practical and thus readily
applicable to multihundred kilograms scale commercial
production.
(14) Ferrer-Flegeau, E.; Bruneau, C.; Dixneuf, P. H.; Jutand, A. J. Am.
Chem. Soc. 2011, 133, 10161−10170.
(15) Hatano, M.; Maki, T.; Moriyama, K.; Arinobe, M.; Ishihara, K. J.
Am. Chem. Soc. 2008, 130, 16858−16860.
(16) Seki, M. ACS Catal. 2011, 1, 607−610.
(17) Seki, M.; Nagahama, M. J. Org. Chem. 2011, 76, 10198−10206.
(18) Seki, M. Synthesis 2012, 44, 3231−3237.
(19) Seki, M. J. Synth. Org. Chem., Jpn. 2012, 70, 1295−1304.
(20) Kubo, K.; Kohara; Yoshimura, Y.; Inada, Y.; Shibouta, Y.;
Furukawa, Y.; Kato, T.; Nishikawa, K.; Naka, T. J. Med. Chem. 1993,
36, 2343.
(21) Naka, T.; Nishikawa, K.; Kubo, T. Such as 1-
(cyclohexyloxycarbonyloxy)ethyl-2-ethoxy-1-((2′-(1H-tetrazol-5-yl)-
biphenyl-4-yl)methyl)benzimidazol-7-carboxylate; exhibits antihyper-
tensive activity and strong angiotensin II antagonistic activity. U.S.
patent 20020151723, January 16, 2002.
(22) Trost, B. M. Science 1991, 254, 1471−1477.
(23) Seki, M. Synthesis, in press (DOI: 10.1055/s-0034-1378585).
(24) Sheldon, R. A. Green Chem. 2007, 9, 1273−1283.
In conclusion, sterically demanding TMBSK was found to be
extremely efficient as a cocatalyst for the Ru-catalyzed C−H
arylation. The catalytic system is economically as well as
environmentally sustainable. Versatility and commercial
importance of the present process was apparently demonstrated
by an application to a novel synthesis of Candesartan Cilexetil.
Ready availability and need of very small amount of TMBSK
and Ru catalyst as well as ease of operation would form a basic
tool in accessing functionalized biaryls of pharmaceutical
importance. The use of TMBSK as an additive for C−H
activation has never been reported and the insights obtained in
the present study would inspire new ideas for C−H activation.
ASSOCIATED CONTENT
* Supporting Information
Experimental information, including protocols for compound
synthesis and NMR data. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: seki.masahiko@mm.api-corp.co.jp. Fax: +81-3-5217-
7173.
■
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) Oi, S.; Fukita, S.; Hirata, N.; Watanuki, N.; Miyano, S.; Inoue, Y.
Org. Lett. 2001, 3, 2579−2581.
(2) Oi, S.; Ogino, Y.; Fukita, S.; Inoue, Y. Org. Lett. 2002, 4, 1783−
1785.
(3) Ackermann, L.; Vicente, R.; Kapdi, A. R. Angew. Chem., Int. Ed.
2009, 48, 9792−9826.
(4) Ackermann, L. Chem. Commun. 2010, 46, 4866−4877.
(5) Ackermann, L.; Vicente, R. Top. Curr. Chem. 2010, 292, 211−
229.
4050
dx.doi.org/10.1021/cs501328x | ACS Catal. 2014, 4, 4047−4050