Synthesis of a sodium-hydrogen exchange type 1 inhibitor: An efficient Cu-catalyzed conjugated addition of a Grignard reagent to an acetyl pyridinium salt

Article abstract of DOI:10.1021/op300331b

A facile and economical five-step process for the synthesis of a sodium-hydrogen exchange type I inhibitor (NHE-1) was developed from readily available starting materials in 43% overall yield. Key transformations included a highly efficient copper-catalyzed conjugate addition of 2- trifluoromethylphenyl Grignard reagents to acetyl pyridinium salts, a facile hydrogenation of 4-aryl dihydropyridines, a regioselective aromatic bromination, an efficient palladium-catalyzed carbonylation of aryl bromides, and a high-yielding acyl guanidine formation. A safe and scalable protocol for preparation of 2-trifluoromethyl phenyl Grignard reagent was developed, and a facile method for controlling the palladium content with N-acetyl-L-cysteine as the scavenger was demonstrated. Process issues in controlling the formation of a key diacylation side product during acyl guanidine formation are also addressed.

Full text of DOI:10.1021/op300331b

Article
pubs.acs.org/OPRD
Synthesis of a Sodium−Hydrogen Exchange Type 1 Inhibitor: An
Efficient Cu-Catalyzed Conjugated Addition of a Grignard Reagent to
an Acetyl Pyridinium Salt
Wenjun Tang,† Nitinchandra D. Patel, Xudong Wei,* Denis Byrne, Ashish Chitroda,
Bikshandarkoil Narayanan, Alexander Sienkiewicz, Laurence J. Nummy, Max Sarvestani, Shengli Ma,
Nelu Grinberg, Heewon Lee, Soojin Kim, Zhibin Li, Earl Spinelli, Bing-Shiou Yang, Nathan Yee,
and Chris H. Senanayake
Chemical Development, Boehringer Ingelheim Pharmaceuticals Inc., Ridgefield, Connecticut 06877, United States
S
* Supporting Information
ABSTRACT: A facile and economical five-step process for the synthesis of a sodium−hydrogen exchange type I inhibitor
(NHE-1) was developed from readily available starting materials in 43% overall yield. Key transformations included a highly
efficient copper-catalyzed conjugate addition of 2-trifluoromethylphenyl Grignard reagents to acetyl pyridinium salts, a facile
hydrogenation of 4-aryl dihydropyridines, a regioselective aromatic bromination, an efficient palladium-catalyzed carbonylation of
aryl bromides, and a high-yielding acyl guanidine formation. A safe and scalable protocol for preparation of 2-trifluoromethyl
phenyl Grignard reagent was developed, and a facile method for controlling the palladium content with N-acetyl-L-cysteine as the
scavenger was demonstrated. Process issues in controlling the formation of a key diacylation side product during acyl guanidine
formation are also addressed.
overall yield (∼20%). In addition to the requirement of several
chromatographic purifications, some costly reagents such as
triflimide 3, octamethyl-2,2′-bi(1,3,2-dioxaborolane), and Boc-
protected guanidine were also employed. We herein report an
efficient and highly convergent synthesis of compound 1 from
the readily available starting material 2-bromobenzotrifluoride
(7) in only five synthetic steps and in 42% overall yield. Key
transformations including an efficient copper-catalyzed con-
jugated addition of 2-trifluoromethylphenyl Grignard reagents
to acetyl pyridinium salts, a facile hydrogenation of 4-aryl
dihydropyridines, a regioselective aromatic bromination, an
efficient carbonylation of aryl bromides, and a high-yielding acyl
guanidine formation are described at multikilogram scales, and
the key process issues are addressed.
INTRODUCTION
■
Sodium−hydrogen exchangers (NHEs) are ion transporters
expressed in a variety of cells that maintain intracellular pH
homeostasis by the electroneutral exchange of intracellular
hydrogen for extracellular sodium.1 Among the nine identified
isoforms of NHEs, NHE type 1 (NHE-1) as the major subtype
in myocardial cells is known to be deeply involved in ischemic
and reperfusion injury. The NHE-1 inhibitors are proven to
improve myocardial contractility and metabolic status as well as
to reduce arrhythmia, apoptosis, necrosis, and intracellular
overload of sodium and calcium ions. They can be effectively
used for prevention and treatment of ischemic heart diseases
such as acute myocardial infarction, arrhythmia, angina pectoris,
etc., and they are also promising candidates for heart-protecting
agents applied to reperfusion therapy or cardiac surgery
including coronary artery bypass graft and percutaneous
transluminal coronary angioplasty.2 Our research department
has recently discovered a potent NHE-1 inhibitor 1 for
preclinical development to fully define safety and pharmaco-
logical properties. An efficient and economical process to acyl
guanidine 1 was thus required in order to supply ample drug
substances for preclinical studies.
RESULTS AND DISCUSSION
■
Choice of Synthetic Strategy. The original synthesis of
NHE-1 inhibitor 1 from Medicinal Chemistry adopted a key
Suzuki coupling reaction4 between vinyl boronic ester 6 and
trisubstituted aryl bromide 10 followed by a transfer hydro-
genation to construct its 4-aryl piperidine framework. Although
the Suzuki coupling was facile and proceeded in high yield
(85%), the preparation of both vinyl boronic ester 6 and aryl
bromide 10 were tedious and required several synthetic steps
from readily available starting materials. Thus, this synthetic
strategy was not ideal for further scale-up activity of compound
1. Several alternative methods for the synthesis of 4-aryl
piperidines were studied on the basis of the availability of
starting materials and practicality (Scheme 2). These included
The original synthetic route to compound 1 from Medicinal
Chemistry3 (Scheme 1) consisted of 14 synthetic steps and,
while suitable for medicinal chemistry’s requirements, had a low
Received: November 14, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/op300331b | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. Original synthetic route of compound 1 from Medicinal Chemistry
Scheme 2. Route scouting for the synthesis of 4-aryl piperidine framework
Scheme 3. Proposed new synthetic route to NHE-1 inhibitor 1
Suzuki coupling of an aryl halide with 4-pyridine boronic acid
(Route A), nucleophilic attack by an aryl Grignard species on a
protected piperidinone (Route B), and reaction of an aryl
Grignard reagent with an acyl pyridium salt (Route C).
Comin’s nucleophilic addition of an aryl Grignard reagent to an
acyl pyridium salt5 to form a 4-aryl dihydropyridine is
B
dx.doi.org/10.1021/op300331b | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
advantageous over the other two methods in terms of both
convergence and cost-effectiveness. The product 4-aryl
dihydropyridine is also easily reduced to form the correspond-
ing 4-aryl piperidine derivative by catalytic hydrogenation. This
method was thus adopted for the synthesis of compound 1.
Since the synthesis of trisubstituted phenyl bromide 10 was
tedious and required five steps from the readily available
starting material (Scheme 1), we proposed to synthesize the
dihydropyridine 15 via Comin’s reaction from 2-trifluoromethyl
phenyl Grignard reagent, which can be prepared conveniently
from 2-bromo benzotrifluoride (7) via Knochel’s metal−
halogen exchange protocol.6 We expected that an aromatic
bromination of the corresponding 4-aryl piperidine derivative
16 should furnish the bromination product 17 regioselectively
(Scheme 3). A subsequent palladium-catalyzed carbonylation
would provide its corresponding methyl ester 18, which could
further react with guanidine to form the acyl guanidine final
product 1 in an overall five-step sequence.
event can be effectively moderated by decreasing its
concentration. An important concentration effect on process
safety was thus discovered, and the detailed results were
reported by us separately.11 ARSST studies showed much
milder exothermic activity with a higher onset temperature
(∼110 °C) when the concentration was decreased to 0.5−0.6
M. Thus, the solution of 2-trifluoromethylphenyl magnesium
chloride was recommended to be best prepared at a low
concentration (<0.6 M). In our scale-up program, we
successfully produced 83 L (50 mols) of a solution of 2-
trifluoromethylphenyl magnesium bromide in THF at 0.6 M
concentration.
With a scalable protocol for preparing 2-trifluoromethyl-
phenyl Grignard reagent available, the Cu-catalyzed addition of
2-trifluoromethylphenyl Grignard reagent to acetyl pyridinium
salts was explored (Scheme 5). The acetyl pyridinium salts were
Scheme 5. Synthesis of dihydropyridine 15
Synthesis of 4-Aryl Dihydropyridine 15 via Comin’s
Reaction. In contrast to many reports on addition of
organometallic reagents to 1-alkoxycarbonylpyridinium salts,7,5a
the corresponding reaction on 1-acetyl pyridinium salt has
received significantly less attention. An early report by Lyle et
al8 showed the addition of PhMgBr to 1-acetyl pyridinium salt
occurred primarily at the 2-position of the pyridinium ring
without using a copper catalyst. In the presence of a copper
catalyst, the addition at the 4-position of the pyridinium ring
was generally preferred.5a,7a,b We envisioned that the bulkiness
of 2-trifluoromethyl phenyl Grignard reagent in conjunction
with employment of a copper catalyst should facilitate the
addition preferentially at the 4-position of the 1-acetyl
pyridinium ring.
In order to study this reaction, it was required to scale up the
formation of 2-trifluoromethyl phenyl Grignard reagent. It
should be noted that the preparation of trifluoromethyl-
substituted phenyl Grignard reagent from its corresponding aryl
halide and magnesium metal is dangerous, and in fact several
severe explosions have been reported.9 Knochel’s method of
generating Grignard solutions via metal−halogen exchange has
proven to be a better method with mild conditions. Applying
this method, the formation of 3,5-bis(trifluoromethyl)phenyl
Grignard reagent in kilogram scale was reported.10 Thus, by
formed instantaneously and precipitated out while acetyl
chloride was added to a light-yellow solution of pyridine and
10 mol % CuI in THF. 2-Trifluoromethylphenyl Grignard
reagent was then added slowly into the slurry at −5−0 °C. The
addition was highly exothermic, indicating the rapid reaction
between 2-trifluoromethylphenyl Grignard reagent and the
acetyl pyridinium salt. The reaction provided an excellent
conversion (>90%), assay yield (∼90%), and regioselectivity.
The mole ratio of 1,4-addition adduct 15 to 1,2-addition adduct
19 was ∼97:3. Further experiments showed that the excellent
regioselectivity can even be maintained in the presence of only
0.1 mol % CuI. In order to understand the role of CuI in the
regioselectivity of this reaction, a parallel study without CuI was
studied. The regioselectivity decreased to 85:15 in the absence
of CuI, demonstrating the importance of the copper catalyst for
the regioselectivity. It is noteworthy that the CuI loading (0.1
mol %) for this reaction is among the lowest reported to date
for addition of an organometallic reactant to an acyl pyridinium
salt. With this protocol, we successfully produced over 10 kg of
the dihydropyridine product 15 in 76% isolated yield and >99%
purity.
i
using PrMgCl (2.0 M in THF, 1.1 equiv) as the reagent and
THF as the solvent, the magnesium−bromide exchange
reaction of 2-bromobenzotrifluoride was run at 15−25 °C for
2 h and >95% conversion was observed. The resulting 2-
trifluoromethyl phenyl magnesium chloride solution (∼1.1 M)
was stable over 24 h at ambient temperature (Scheme 4).
However, an adiabatic calorimetry study of this solution in an
Advanced Reactive Systems Screening Tool (ARSST, http://
www.fauske.com) showed a highly exothermic event with an
onset temperature at ∼85 °C, leading to a safety concern for
scale-up activities. Further studies showed that this exothermic
Hydrogenation of Dihydropyridine 15 and Bromina-
tion. The hydrogenation of dihydropyridine 15 with 10% Pd/
C as the catalyst proceeded smoothly under 200 psi H2 at 50−
55 °C. The hydrogenation of the first double bond proceeded
much faster than that of the second double bond, possibly due
to the conformational and electronic change of the ring system
during hydrogenation. With 1 mol % of 10% Pd/C as the
catalyst, the hydrogenation of the second double bond reached
>99% conversion after 10 h (Scheme 6). After filtration to
Scheme 4. Formation of 2-trifluoromethyl phenyl Grignard
reagent
C
dx.doi.org/10.1021/op300331b | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
remove Pd/C catalyst, the filtrate was directly used for the
ensuing bromination step.
Employment of a high reaction temperature (110 °C) greatly
accelerated the reaction rate, and >99% conversion was
observed after 20 h. However, the formation of the carboxylic
acid 22 increased accordingly (27%). Further screening of bases
revealed that an organic base such as triethylamine provided a
decrease in the formation of 22 (5%). Fine tuning of the
reaction temperature to 90 °C ensured the sufficient reactivity
for this carbonylation and also controlled the formation of acid
22 to ∼3%. A small amount of the reduction product 16 (<2%)
was also observed during the reaction. A simple workup
procedure was developed, in which the reaction mixture was
filtered over Celite and the filtrate then solvent-switched to an
NMP-water system. The product was crystallized directly from
this solution in 90−93% yield. In our scale-up program, we
were able to employ only 0.2 mol % Pd(OAc)2 and 0.22 mol %
DPPP for this carbonylation, which not only allowed us to
reduce costs but also facilitated the control of the Pd content
within 18 (Scheme 8).
Scheme 6. Hydrogenation of dihydropyridine 15
Several reagents were initially screened for the bromination
of 16, and 1,3-dibromo-5,5-dimethylhydantoin (DBH) proved
to be the best for its higher reactivity than Br2 and N-bromo-
succinimide (NBS) (Scheme 7). It was also found that an acidic
Several methods14 were applied to control the Pd content
including Cuno ZetaCarbon filters15 and N-acetyl-L-cystei-
ne16,14a (Table 3). While Cuno ZetaCarbon filters were
ineffective with MeOH as the solvent, N-acetyl-L-cysteine was
a fine scavenger for this process. Since the palladium complex
of N-acetyl-L-cysteine is water-soluble, we simply added 3 mol
% N-acetylcysteine to the NMP/water solution of the reaction
mixture prior to crystallization and the Pd content of 18 was
controlled to ∼10 ppm. No additional operations such as phase
cuts or filtrations were then required for robust palladium
content control. In our scale-up program, we successfully
prepared 7.5 kg of methyl ester 18 in 90−93% isolated yield
and 97% purity with <5 ppm residual Pd content.
Scheme 7. Bromination of compound 16
Synthesis of Acyl Guanidine 1 and Its HCl Salt
Formation. Next we studied the synthesis of acyl guanidine
1 from ester 18 and guanidine (Scheme 9).17 Since anhydrous
guanidine is not commercially available, it was generated in situ
from guanidine·HCl salt by reaction with NaOtAm as base in
NMP solution. A more common base KOtBu can also be used
for generating guanidine on laboratory scale. However, its NMP
solution proved to be too viscous for operations at larger scales.
In addition to the desired product 1, the two common side
products during the reaction were acid 22 and diacylation
product 23, which were generally formed in about 3%.
Abnormally high levels of the diacylation side product 23
(3−30%) were observed in some small batches. Investigation
revealed that the heterogeneous reaction between guanidi-
ne·HCl salt and NaOtAm was incomplete in those batches,
leading to a strongly basic media at the time of addition of the
ester 18. The formed product 1 can be further deprotonated
under the strong basic conditions used to promote the
formation of the diacylation product 23 (Scheme 10). To
avoid the incomplete reaction between guanidine·HCl salt and
NaOtAm, two strategies were enforced: (1) additional NMP
solvent was employed to make the reaction more homoge-
neous, and (2) excess guanidine·HCl salt (2.2 equiv) was
employed to ensure the complete consumption of NaOtAm. In
this way, the formation of diacylation side product 23 was
consistently controlled to ∼3% and the acyl guanidine 1 was
isolated by crystallization from NMP/water solution in 88%
yield and 98.9 A% purity.
media could enhance the reactivity, therefore various
combinations of solvents and acids were tested, and the results
are listed in Table 1. A combination of sulfuric acid and
dichloromethane was found to be a good medium to promote
the bromination (entry 6). Other than the desired product 17,
a major regioisomer 20 and a dibromo compound 21 were also
identified, together with two other uncharacterized minor
regioisomers. All experiments were run at ambient temper-
atures (20−25 °C); higher temperature was found to generate
more impurities. It was found that addition of acetic acid
provided a better regioselectivity between 17 and 20.12 The
selectivity increased from ∼12:1 to 18:1 when 3 equiv of acetic
acid was employed. Thus, with H2SO4/HOAc/DCM as the
media and DBH (0.6 equiv) as the reagent, compound 17 was
formed in ∼90−92% assay yield after 12 h at 20−22 °C (entry
10), along with the side products 20 (∼5%) and 21 (∼2%).
These impurities could be easily controlled to <1.5% by a single
crystallization of 17 from methylcyclohexane. We thus
successfully produced 9.5 kg of 17 in 78−80% isolated yield
and in >98.5% purity by HPLC.
Carbonylation of Bromide 17. The palladium-catalyzed
carbonylation of bromide 17 to form methyl ester 18 was then
studied.13 With MeOH as the solvent and K2CO3 as the base at
70 °C under 100 psi CO, three common bidentate ligands
DPPP, BINAP, and DPPF were screened and DPPP provided
the best reactivity under otherwise identical conditions (entry
1, 3, Table 2). Thus, DPPP was chosen for further screening.
The wet cake of acyl guanidine 1 was directly used for the
formation of its HCl salt with isopropanol as the solvent and
concentrated HCl as the reagent (Scheme 11). In order to
consistently form the right polymorph and morphology of
D
dx.doi.org/10.1021/op300331b | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Table 1. Screening of reaction conditions for bromination of 16
scale
time product (17) SM (16) regioisomers (20 + other
dibromo
entry
1
(g)
conditions
(h)
(%)
(%)
two) (%)
(21) (%)
0.1 DCM (10 vol), DBH (0.5 equiv), MeSO3H (5 equiv)
12
32
61
2.0
1.5
3.2
0
0
2
3
0.1 DCM (10 vol), NBS (1.05 equiv), MeSO3H (5 equiv)
12
12
5
95
8
0
0
0.1 DCM (4 vol), DBH (0.75 equiv), MeSO3H (10−15 equiv)
80
7.4
1.3
2.2
4.6
2.1
4.7
0.9
1.2
7.2
0.7
4.1
2.3
2.9
1.0
5.9
0.8
5.5
0.6
6.2
0.6
1.5
0.2
−
4
5
0.1 TFA(2.5 vol), DBH (0.55 equiv), H2SO4 (1.0 equiv)
12
12
91
90
0
1.5
0
2
TFA(2.5 vol), DBH (0.55 equiv), H2SO4 (1.0 equiv)
2.8
6
0.1 DCM, DBH (0.62 equiv), H2SO4 (4 equiv)
8
8
82
3.2
14.8
40.6
4.0
4.6
0.6
0
7
0.1 DCM, DBH (0.62 equiv), H2SO4 (4 equiv)
75.3
53
8
0.1 DCM, DBH (0.62 equiv), AcOH (6 equiv), H2SO4 (2 equiv)
0.1 DCM, DBH (0.62 equiv), AcOH (4 equiv), H2SO4 (3.4 equiv)
8
9
3
88
0.5
1.5
0.6
0.8
−
10
11
12
20
DCM, DBH (0.62 equiv), AcOH (4 equiv), H2SO4
(3.4 equiv)
8
92.2
89.3
40
0.0
3.1
5.0 MeCN, DBH (0.65 equiv), AcOH (2.5 equiv), H2SO4
(5.0 equiv)
12
15
0.1 MCH, DBH (0.65 equiv), AcOH (5 equiv), H2SO4 (2 equiv)
56
13
14
0.1 iPrOH, DBH (0.62 equiv), AcOH (4 equiv), H2SO4 (3.4 equiv)
0.1 THF, DBH (0.57 equiv), H2SO4 (2 equiv)
15
15
−
1
100
99
−
Table 2. Screening of reaction conditions for carbonylation
of 17
Scheme 8. Carbonylation of aryl bromide 17
Table 3. Palladium content control in ester 18
Pd(OAc)2 loading (mol
%)
Pd content in 18
treatment
(ppm)
HPLC yield (A%)
0.5
0.5
none
1557
1399
CUNO ZetaCarbon R53
filter
entry
L
base
T (°C)
18
22
1
2
3
4
5
6
7
DPPF
BINAP
DPPP
DPPP
DPPP
DPPP
DPPP
K2CO3
K2CO3
K2CO3
K2CO3
TMEDA
Et3N
70
70
20
10
44
72
87
83
92
−
−
−
27
8
0.5
0.5
0.5
0.2
CUNO ZetaCarbon R55
filter
1255
136
83
70
1.5 mol % N-acetyl-L-
cysteine
110
110
110
90
3 mol % N-acetyl-L-
cysteine
5
3 mol % N-acetyl-L-
cysteine
4
Et3N
3
CONCLUSIONS
1·HCl salt, the water content of the isopropanol solution of 1
was controlled to below 1% prior to the addition of HCl. Thus,
the HCl salt of 1 was isolated in ∼90% isolated yield and in
99.6% purity.
■
In summary, we have successfully developed an economical and
facile process for the synthesis of a NHE type I inhibitor, 1, in
five steps from readily available starting material and in 43%
E
dx.doi.org/10.1021/op300331b | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 9. Formation of acyl guanidine 1
Scheme 11. Synthesis of acyl guanidine 1 and its HCl salt
°C for 20 min. To this mixture at 0−5 °C was added acetyl
chloride (1.94 kg, 24.2 mol, 1.1 equiv) over 10 min, while the
temperature was controlled <25 °C. To the slurry at −5−5 °C
was added the aforementioned 2-trifluoromethyl Grignard
solution over 1.5 h, while the temperature was controlled at
−5−5 °C. The mixture was kept at 0−5 °C for 0.5 h, warmed
to 20−25 °C over 20 min, and further stirred at 20−25 °C for 1
h. To the cold reaction mixture at 10−15 °C was charged 2 N
HCl solution (10 L) over 15 min while controlling the reaction
temperature <35 °C. The two phases were separated, and the
aqueous layer was discarded. To the organic phase was added
methylcyclohexane (10 L) and 10% NaCl solution (10 L), and
the mixture was stirred at 20−25 °C for 20 min. The organic
phase was separated, and methylcyclohexane (5 L) was added.
The mixture was distilled to ∼10 L to form a slurry, while the
internal temperature was controlled <70 °C. Additional
methylcyclohexane (7.5 L) was added, and the mixture was
heated to ∼90 °C to form a clear solution, cooled to 40 °C over
1 h, held at 40 °C for 0.5 h, further cooled to 20−22 °C over 1
h, and finally stirred at 20−22 °C for 1 h to form a slurry. The
slurry was filtered, and the cake was washed further with
methylcyclohexane (2 L × 2) and then dried under vacuum at
20−25 °C for 12 h to give the desired product 15 (4.45 kg, 76%
yield, >99% purity by HPLC) as a white crystalline solid. 15:
mp (DSC) 91 °C; 1HNMR (400 MHz, DMSO-d6): δ 7.68 (m,
2H), 7.58 (m, 1H), 7.43 (m, 1H), 7.21 (d, J = 6.5 Hz, 1H), 6.96
(d, J = 6.0 Hz, 1H), 4.99 (br s, 1H), 4.87 (br s,1H), 4.46 (s,
1H), 2.23 (s, 3H); 13CNMR (100 MHz, DMSO-d6): δ166.4,
143.8, 133.1, 131.8, 127.1, 125.4 (m), 125.3 (m), 124.6, 124.4
(q, J = 272 Hz), 121.1, 109.5, 107.8, 34.5, 21.2; ESI-MS: m/z
268 [M + H]+.
overall yield. This process allowed us to produce compound 1
in multikilogram scales in 3 months. Key transformations
included a highly efficient copper-catalyzed conjugated addition
of 2-trifluoromethylphenyl Grignard reagents to acetyl
pyridinium salts at 0.1 mol % catalyst loading, a facile
hydrogenation of 4-aryl dihydropyridines 15, a regioselective
aromatic bromination of 16, an efficient palladium-catalyzed
carbonylation of aryl bromide 17 at 0.2 mol % Pd loading, and
finally a high-yielding acyl guanidine formation of 1. In addition
to the development of these key transformations, a safe and
scalable protocol for the preparation and use of 2-
trifluoromethyl phenyl Grignard reagent was developed, and a
facile method for controlling the palladium content with N-
acetyl-L-cysteine was defined.
EXPERIMENTAL SECTION
■
1-(4-(2-(Trifluoromethyl)phenyl)pyridin-1(4H)-yl)-
ethanone (15). To a solution of 2-bromobenzotrifluoride (7,
5.0 kg, 22.0 mol) in THF (5 L) at 20−25 °C was added a
solution of iPrMgBr in THF (26.2 kg, 13.6 wt %, 24.2 mol, 1.1
equiv) over 15 min while controlling the reaction temperature
<30 °C (CAUTION: For safety reasons, it is critical to add the
iPrMgBr solution slowly and keep the temperature under 30 °C!).
After the addition, the mixture was stirred at 25−30 °C for 3−4
h. In a separate reactor was charged CuI (42 g, 0.022 mol, 0.1
mol %), THF (10 L) and pyridine (1.64 kg, 33.0 mol. 1.5
equiv). The resulting light yellow solution was stirred at 20−25
1-(4-(4-Bromo-2-(trifluoromethyl)phenyl)piperidin-1-
yl)ethanone (17). To a 20 L autoclave was charged 1-(4-(2-
Scheme 10. Formation of unwanted diacylation side product 23
F
dx.doi.org/10.1021/op300331b | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
(trifluoromethyl)phenyl)pyridin-1(4H)-yl)ethanone (15, 2.5
kg, 9.31 mol) and 10 wt % Pd/C (Degussa type E101 NE/
W, wet, 50%, 0.2 kg). The system was purged with nitrogen
three times and then MeOH (12.5 L) was charged. The
mixture was further purged with hydrogen three times and then
pressurized with 200 psi H2. The resulting mixture was stirred
at 50−55 °C for 12 h and then filtered over a Celite pad to give
a crude MeOH solution of 1-(4-(2-(trifluoromethyl)phenyl)-
piperidin-1-yl)ethanone (16) in 99% HPLC purity. A crude
MeOH solution of 1-(4-(2-(trifluoromethyl)phenyl)piperidin-
1-yl)ethanone (16) from hydrogenation (containing ∼5 kg 16,
18.4 mol) was charged to a 50 L reactor. The mixture was
distilled to ∼7 L, while the internal temperature was controlled
at 50−55 °C. To the mixture was charged dichloromethane (16
L), and the resulting mixture was further distilled to ∼16 L to
control the MeOH content <1 wt %. To the mixture at 20−25
°C was charged 1,3-dibromo-5,5-dimethylhydantoin (3.12 kg,
10.9 mol, 0.6 equiv) and acetic acid (3.28 kg, 54.6 mol, ∼3
equiv). After the mixture was cooled to ∼10 °C, conc. sulfuric
acid (5.35 kg, 54.6 mol, ∼3 equiv) was added over 0.5 h, while
the temperature was controlled at 15−20 °C. The resulting
mixture was stirred at 20−23 °C for 10 h. Methylcyclohexane
(25 L) and water (10 L) were added over 30 min, while the
temperature was kept at <22 °C. The two phases were
separated, and the aqueous layer was removed. To the organic
layer at ∼10 °C was charged 10% NaOH solution (7.3 kg). The
pH of the aqueous layer was adjusted to 8−10. To the mixture
was further added 10% Na2S2O3 solution (9.2 L). The two
phases were separated, and the bottom aqueous layer was
removed. The organic layer was further washed with 5% NaCl
solution (12 L) and distilled under vacuum to ∼10 L.
Methylcyclohexane (6.5 L) was added, and the mixture was
warmed to ∼65 °C to form a clear solution, cooled to 20−22
°C over 8 h, and stirred further at this temperature for 1 h. The
slurry was filtered, and the cake was washed with methyl-
cyclohexane/heptane (1/1, 3 L × 2) and dried at 50 °C under
vacuum for 12 h to provide the desired product 17 (4.93 kg,
78%, 99.0% purity by HPLC) as white solid. 17: mp (DSC)
mixture at ∼60 °C was charged N-acetyl-L-cysteine (35.5 g,
0.22 mol, 3 mol %), and the mixture was further stirred at ∼60
°C for 30 min before water (20 L) was charged. To the mixture
at ∼40 °C was charged seed crystals (12 g, ∼0.5% of theoretical
yield), and the resulting solution was cooled to 20−22 °C over
3 h and held at this temperature for 6−8 h. Water (5 L) was
added, and the resulting slurry was further stirred at 20−22 °C
for 1 h. The slurry was filtered, and the wet cake was washed
with water (2 L × 3) and dried at 50 °C under house vacuum
to provide the desired product 18 (2.12 kg, 93.6% 96.5 A%
purity by HPLC, Pd content: 4.1 ppm) as white solid. 18: mp
(DSC) 105 °C; 1H NMR (400 MHz, DMSO-d6): δ 8.14 (d, J =
8.2 Hz, 1H), 8.13 (s, 1H), 7.80 (d, J = 8.1 Hz, 1H), 4.56 (d, J =
13.2 Hz, 1H), 3.94 (d, J = 13.7 Hz, 1H), 3.87 (s, 3H), 3.13 (m,
2H), 2.57 (td, J = 12.9, 2.6 Hz, 1H), 2.04 (s, 3H), 1.50−1.85
(m, 4H); 13C NMR (100 MHz, DMSO-d6): δ 168.0, 164.9,
149.6, 133.1, 129.6, 128.1, 126.9 (q, J = 29.0 Hz), 126.0 (q, J =
5.9 Hz), 124.0 (q, J = 272.5 Hz), 52.4, 46.1, 41.2, 38.4, 32.9,
32.4, 21.2; ESI-MS: m/z 330 [M + H]+.
4-(1-Acetylpiperidin-4-yl)-N-(diaminomethylene)-3-
(trifluoromethyl)benzamide (1). To a 50 L reactor was
charged guanidine·HCl (2.39 kg, 24.5 mol, 2.2 equiv) followed
by NMP (11.7 kg). The mixture was cooled to 10−15 °C. A
solution of sodium tert-pentoxide (2.58 kg, 22.3 mol, 2.0 equiv)
in NMP (7.8 kg) was charged over ∼20 min, while the reaction
temperature was controlled at <25 °C. The resulting mixture
was stirred at 20−25 °C for 1 h. To the mixture at 20−25 °C
was charged a solution of methyl 4-(1-acetylpiperidin-4-yl)-3-
(trifluoromethyl) benzoate (18) over 1 h, while the temper-
ature was controlled at <25 °C. The resulting mixture was
stirred at 20−25 °C for 2 h. Water (39.9 kg) was added over 20
min, while the temperature was controlled <40 °C. The
solution was cooled to 30 °C, and seed crystals (30 g, ∼0.6% of
theoretical yield) were charged. The resulting mixture was
cooled to 20−22 °C over 3 h and stirred at this temperature for
12 h. The slurry was filtered, and the cake was rinsed with water
(2 L × 3), air-dried for 2 h, and further dried under house
vacuum at 20−25 °C for 12 h to provide the wet desired
product 1 (4.18 kg, 88.4% yield, 98.9 A% purity by HPLC,
containing 15.5% water and 0.9% NMP) as a white crystalline
solid. 1: mp (DSC) 219 °C; 1H NMR (400 MHz, DMSO-d6):
δ 8.35 (s, 1H), 8.23 (d, J = 8.0 Hz, 1H), 7.62 (d, J = 8.2 Hz,
1H), 4.56 (d, J = 13.0 Hz, 1H), 3.92 (d, J = 13.3 Hz, 1H), 3.09
(m, 2H), 2.56 (t, J = 12.7 Hz, 1H), 3.10 (s, 3H), 1.45−1.80 (m,
4H); 13C NMR (100 MHz, DMSO-d6): δ 174.0, 168.1, 163.1,
145.5, 137.5, 132.4, 128.3, 126.1, 125.6 (q, J = 6.0 Hz), 124.6
(q, J = 243.7 Hz), 46.3, 41.4, 38.2, 33.2, 32.7, 21.3; ESI-MS: m/
z 357 [M + H]+.
1
107 °C; H NMR (400 MHz, CDCl3): δ 7.76 (d, J = 2.1 Hz,
1H), 7.64 (dd, J = 8.4, 2.0 Hz, 1H), 7.26 (d, J = 8.4 Hz, 1H),
4.80 (d, J = 13.5 Hz, 1H), 3.93 (d, J = 13.4 Hz, 1H), 3.16 (m,
2H), 2.61 (t, J = 11.4 Hz, 1H), 2.13 (s, 3H), 1.83 (m, 2H), 1.62
(m, 2H); 13CNMR (100 MHz, CDCl3): δ 168.7, 143.1, 135.1,
129.6, 129.6 (q, J = 119.6 Hz), 128.9 (q, J = 6.2 Hz), 123.5 (q, J
= 272.8 Hz), 119.9, 46.9, 42.0, 38.1 (q, J = 1.9 Hz), 33.8, 32.7,
21.4; ESI-MS: m/z 350 [M + H]+.
Methyl 4-(1-Acetylpiperidin-4-yl)-3-(trifluoromethyl)-
benzoate (18). To a 20 L autoclave was charged 1-(4-(4-
bromo-2-(trifluoromethyl)phenyl)piperidin-1-yl)ethanone (17,
2.5 kg, 7.1 mol) followed by MeOH (10 L) and triethylamine
(1.44 kg, 14.1 mol, 2 equiv). The mixture was purged with
nitrogen (5 psi) three times before a solution of Pd(OAc)2 (3.2
g, 14 mmol, 0.2 mol %) and DPPP (6.6 g, 16 mmol, 0.22 mol
%) in degassed MeOH (2 L) was charged. The resulting
solution was purged with CO three times, pressurized with 100
psi CO, and then heated to 90−92 °C for 20 h. After cooling to
rt, the mixture was filtered through a Celite pad to provide a
crude solution of methyl 4-(1-acetylpiperidin-4-yl)-3-
(trifluoromethyl)benzoate (18) in MeOH in 93.6% assay
yield. The solution was distilled to ∼2 L, while the internal
temperature was controlled at <65 °C. NMP (6.5 kg) was
added, and the resulting mixture was further distilled at ∼55 °C
under reduced pressure to remove most of MeOH. To the
4-(1-Acetylpiperidin-4-yl)-N-(diaminomethylene)-3-
(trifluoromethyl)benzamide HCl Salt (1·HCl). To a 50 L
reactor was charged wet 4-(1-acetylpiperidin-4-yl)-N-(diamino-
methylene)-3-(trifluoromethyl)benzamide (1, 4.1 kg, 84.0%,
9.66 mol) followed by isopropanol (41.3 L). The mixture was
stirred at 20−25 °C for 15 min to form a clear solution and
then polish filtered. The filtrate was distilled under vacuum to
∼25 L, while the temperature was controlled at 50−65 °C. A
sample was taken to confirm the water content of the solution
was <1.5% (actual: 1.38%). Additional isopopanol (20 L) was
added, and the mixture was cooled to ∼40 °C. Conc. HCl
solution (1.05 kg, 37%, 10.63 mol, 1.1 equiv) was charged to
the mixture over 15 min followed by seeding (19 g, ∼0.5% of
theoretical yield). The mixture was held at ∼40 °C for 2 h,
cooled to 20−22 °C over 3 h, and further stirred at this
G
dx.doi.org/10.1021/op300331b | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
(9) (a) Waymouth, R.; Moore, E. J. Chem. Eng. News 1997, 75, 6.
(b) Ashby, E. C.; Al-Fekri, D. M. J. Organomet. Chem. 1990, 390, 275.
(c) Appleby, I. C. Chem. Ind. 1971, 120.
(10) Leazer, J. L., Jr.; Cvetovich, R.; Tsay, F.-R.; Dolling, U.; Vickery,
T.; Bachert, D. J. Org. Chem. 2003, 68, 3695.
(11) Tang, W.; Sarvestani, M.; Wei, X.; Patel, N.; Narayanan, B.;
Nummy, L. J.; Byrne, D.; Lee, H.; Yee, N.; Senanayake, C. H. Org.
Process Res. Dev. 2009, 13, 1426.
(12) A similar effect of HOAc on regioselectivity in bromination was
also reported, see ref 9.
temperature for 10 h. The slurry was filtered and the wet cake
dried under house vacuum at 80 °C for 24 h to provide the
desired product 1·HCl (3.40 kg, 90% yield, 99.6 A% purity by
HPLC) as a white crystalline solid. 1·HCl: mp (DSC) 238 °C;
1H NMR (400 MHz, DMSO-d6): δ 12.4 (s, 1H), 8.81 (br s,
2H), 8.69 (br s, 2H), 8.42 (d, J = 8.36 Hz, 1H), 8.38 (s, 1H),
7.88 (d, J = 8.3 Hz, 1H), 4.56 (d, J = 12.7 Hz, 1H), 3.95 (d, J =
13.3 Hz, 1H), 3.13 (m, 2H), 2.57 (t, J = 11.0 Hz, 1H), 2.03 (s,
3H), 1.50−1.90 (m, 4H); 13C NMR (100 MHz, DMSO-d6): δ
168.0, 166.1, 155.6, 150.2, 132.5, 129.7, 129.6, 127.0 (q, J =
29.5 Hz), 125.9 (q, J = 5.8 Hz), 124.0 (q, J = 273 Hz), 46.1,
41.2, 38.5, 32.8, 32.3, 21.3.
(13) For general reviews, see: (a) Mori, M. In Handbook of
Organopalladium Chemistry for Organic Synthesis; Negishi, E.-i., Ed.;
John Wiley & Sons, Inc.: New York, 2002; p 2313. (b) Skoda-Foldes,
̈
́
R.; Kollar, L. Curr. Org. Chem. 2002, 6, 1097. (c) Beller, M.; Indolese,
A. F. Chimia 2001, 55, 684.
ASSOCIATED CONTENT
(14) For general reviews on palladium removal methods, see:
(a) Garrett, C. E.; Prasad, K. Adv. Synth. Catal. 2004, 346, 889.
(b) Welch, C. J.; Albaneze-Walker, J.; Leonard, W. R.; Biba, M.;
DaSilva, J.; Henderson, D.; Laing, B.; Mathre, D. J.; Spencer, S.; Bu, X.;
Wang, T. Org. Process Res. Dev. 2005, 9, 198. (c) Bien, J. T.; Lane, G.
C.; Oberholzer, M. R. Top. Organomet. Chem. 2004, 6, 263.
(15) CUNO ZetaCarbon filters are products of CUNO Inc.
■
S
* Supporting Information
1H NMR and 13C NMR of compounds 15, 17, 18, 1, and
1·HCl. This material is available free of charge via the Internet
at http://pubs.acs.org.
(16) Konigsberger, K.; Chen, G.-P.; Wu, R.; Girgis, M. J.; Prasad, K.;
Repic, O.; Blacklock, T. J. Org. Process Res. Dev. 2003, 7, 731.
̌
(17) For similar transformations, see: (a) Lee, S.; Yi, K. Y.; Hwang, S.
K.; Lee, B. H.; Yoo, S.-e.; Lee, K. J. Med. Chem. 2005, 48, 2882.
(b) McAllister, L. A.; Hixon, M. S.; Kennedy, J. P.; Dickerson, T. J.;
Janda, K. D. J. Am. Chem. Soc. 2006, 128, 4176.
̈
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: xudong.wei@boehringer-ingelheim.com
Present Address
†Shanghai Institute of Organic Chemistry, 345 Ling Ling Road,
Shanghai 200032, China.
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) (a) Karmazyn, M.; Gan, X. T.; Humphreys, R. A.; Yoshida, H.;
Kusumoto, K. Circ. Res. 1999, 85, 777. (b) Karmazyn, M. Drug Dev.
Res. 2002, 55, 22. (b) Karmazyn, M.; Kilic,
Cardiol. 2008, 44, 647.
(2) (a) Spitznagel, H.; Chung, O.; Xia, Q.-G.; Rossius, B.; Illner, S.;
́
A.; Javadov, S. J. Mol. Cell.
Jahnichen, G.; Sandmann, S.; Reinecke, A.; Daemen, H. J. A. P.;
̈
Unger, T. Cardiovasc. Res. 2000, 46, 102. (b) Fliegel, L. J. Mol. Cell.
Cardiol. 2008, 44, 228.
(3) Huber, J. D.; Bentzien, J.; Boyer, S. J.; Burke, J.; De Lombaert, S.;
Eickmeier, C.; Guo, X.; Haist, J. V.; Hickey, E. R.; Kaplita, P.;
Karmazyn, M.; Kemper, R.; Kennedy, C. A.; Kirrane, T.; Madwed, J.
B.; Mainolfi, E.; Nagaraja, N.; Soleymanzadeh, F.; Swinamer, A.;
Eldrup, A. B. J. Med. Chem. 2012, 55 (16), 7114−7140.
(4) For a review of palladium-catalyzed cross-coupling reactions of
organoboranes, see: (a) Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95,
2457. (b) Suzuki, A. In Metal-Catalyzed Cross-Coupling Reactions;
Diederich, F., Stang, P. J., Eds.; Wiley-VCH: Weinheim, Germany,
1998; p 49.
(5) (a) Comins, D. L.; Abdullah, A. H. J. Org. Chem. 1982, 47, 4315.
(b) Comins, D. L.; O’Connor, S. Adv. Heterocycl. Chem. 1988, 44, 199.
(6) (a) Abarbri, M.; Dehmel, F.; Knochel, P. Tetrahedron Lett. 1999,
40, 7449. (b) Boymond, L.; Rottlander, M.; Cahiez, G.; Knochel, P.
̈
Angew. Chem., Int. Ed. 1998, 37, 1701. (c) Jensen, A. E.; Dohle, W.;
Sapountzis, I.; Lindsay, D. M.; Cu, V. A.; Knochel, P. Synthesis 2002, 4,
565. (d) Knochel, P.; Dohle, W.; Gommermann, N.; Kneisel, F. F.;
Kopp, F.; Korn, T.; Sapountzis, I.; Vu, V. A. Angew. Chem., Int. Ed.
2003, 42, 4302.
(7) (a) Boice, G. N.; Savarin, C. G.; Murry, J. A.; Conrad, K.; Matty,
L.; Corley, E. G.; Smitrovich, J. H.; Hughes, D. Tetrahedron 2004, 60,
11367. (b) Wang, X.; Kauppi, A. M.; Olsson, R.; Almqvist, F. Eur. J.
Org. Chem. 2003, 4586. (c) Comins, D. L.; Chung, G.; Foley, M. A.
Heterocycles 1994, 37, 1121.
(8) Lyle, R. E.; Marshall, J. L.; Comins, D. L. Tetrahedron Lett. 1977,
12, 1015.
H
dx.doi.org/10.1021/op300331b | Org. Process Res. Dev. XXXX, XXX, XXX−XXX