Convergent kilo-scale synthesis of a potent renin inhibitor for the treatment of hypertension

Article abstract of DOI:10.1021/op2001063

Process research and development of a synthetic route towards a novel renin inhibitor (1) is described. The highly convergent synthetic route provided 1 in 15% yield on multikilogram scale with a longest linear sequence of 11 steps. The use of catalytic h

Full text of DOI:10.1021/op2001063

ARTICLE
pubs.acs.org/OPRD
Convergent Kilo-Scale Synthesis of a Potent Renin Inhibitor for the
Treatment of Hypertension
Louis-Charles Campeau,*,† Sarah J. Dolman,† Danny Gauvreau,† Ed Corley,‡ Jinchu Liu,‡
Erin N. Guidry,‡ Stꢀephane G. Ouellet,† Dietrich Steinhuebel,‡ Mark Weisel,‡ and Paul D. O’Shea†
†Global Process Chemistry, Merck-Frosst, 16711 Trans Canada Highway, Kirkland, Quꢀebec H9H 3L1, Canada
‡Global Process Chemistry, Merck, P.O. Box 2000, Rahway, New Jersey 07065, United States
S Supporting Information
b
ABSTRACT: Process research and development of a synthetic route towards a novel renin inhibitor (1) is described. The highly
convergent synthetic route provided 1 in 15% yield on multikilogram scale with a longest linear sequence of 11 steps. The use of
catalytic hydrogenation features prominently in our design. The proper choice of N-methylpyridone surrogate was also important,
and we describe a method for the easy conversion of 2-methoxypyridines to N-methylpyridones using cheap and readily available
reagents.
’ INTRODUCTION
Essential hypertension is known to affect over a billion people
worldwide.1 If left untreated, hypertension can lead to failure of a
host of biological functions in the brain, kidney, and heart.2 The
reninꢀangiotensinꢀaldosterone system3 is a tightly regulated
biological cascade known to play a role in the control of blood
pressure. Consequently, its various steps have been targets of
novel agents for relief of hypertension as well as end-organ
protection.4,5 As part of a drug discovery program in our
laboratories, 1-HCl was identified as a potent and selective
inhibitor of renin with a suitable profile for further clinical
development.6 In order to do so, we developed a synthesis which
would allow for its production on kilogram scale. Herein, we
report our efforts to develop an enantioselective, practical, and
convergent approach to 1 which was demonstrated on multi-
kilogram scale for the first GMP delivery.
’ RESULTS AND DISCUSSION
Our retrosynthetic analysis is shown in Figure 1. We envi-
sioned that 1 could be assembled in a convergent manner with
final amidation step between the key acid (2) and amine
fragments (3). The chirality in 2 would be installed via asym-
metric hydrogenation of the tetrasubstituted olefin which could
be readily obtained from commercially available piperidine (4).
The choice of pyridone surrogate was crucial to the success of
this strategy. We envisaged the amine fragment (3) as accessible
from amidation and reduction of the corresponding 3-bromo-4-
toluic acid (5), while the alkyne moiety could be installed on an
appropriately substituted halogenated arene.
Amine Fragment (3) Synthesis. Previously, gram-scale
synthesis of amine (3) was achieved via a nonselective Suzuki
coupling between a 3,5-dibromo aryl and the appropriate alke-
Special Issue: Asymmetric Synthesis on Large Scale 2011
nyl-boronic acid (eq 1).6a The resulting mixture of mono- and
bis-coupled products required column chromatography in order
to isolate the desired intermediate and afforded low yield (46%).
Received: April 18, 2011
Published: June 24, 2011
r
2011 American Chemical Society
1138
dx.doi.org/10.1021/op2001063 Org. Process Res. Dev. 2011, 15, 1138–1148
|

Organic Process Research & Development
ARTICLE
Figure 1. Retrosynthetic approach to 1.
Table 1. Hydrogenation of alkyne 7
entry
catalysta
solvent
additiveb
conversionc (%)
8:8ac
8bc (%)
1
2
3
4
5
6
7
RhCl(PPh3)3
Pd/C
EtOAc
EtOH
EtOH
EtOH
PhMe
PhMe
PhMe
none
0
0
ꢀ
ꢀ
ꢀ
ꢀ
none
Pd/C
MgBr2
none
100
100
100
100
100
9:1
3.3
ꢀ
PtO2
1:1
PtO2
none
13:1
19:1
>50:1
<3
ꢀ
PtO2
AcOH
Cs2CO3
PtO2
1.8
a 5 mol % catalyst loading. b 20 mol % additive. c Determined by HPLC.
To circumvent this problem, we employed arene (6) which
could be subjected to monoselective cross-coupling due to the
differentially activated halides.7 Additionally, by moving from a
Suzuki to a Sonagashira coupling, we were able to replace the
expensive vinyl-boronic ester with a propargyl ether (eq 2).
Alkyne 7 was obtained from successful cross-coupling, and then
studied under a variety of hydrogenation conditions.
We found that the major issue associated with this hydro-
genation step was the formation of two side products, 8a and 8b,
resulting respectively from reduction of C-OMe and CꢀBr
bonds. Formation of 8b could be minimized by switching from
Pd- to Pt-based catalysts. For example, Adam’s catalyst afforded
<3% 8b; however, concomitantly high amounts of 8a were
observed (Table 1, entry 4). Screening revealed that 8a levels
could be reduced by switching the solvent from EtOH to PhMe
(Table 1, entries 4ꢀ5) and that various additives could improve
that ratio further (Table 1, entries 6ꢀ7).8 It was also noted that
higher levels of residual palladium, from the previous Sonoga-
shira coupling, resulted in greater levels of impurity 8b. In order
to avoid this byproduct, we required palladium levels to be <100
ppm. We therefore chose to reorganize the synthesis around
opportunities for purification via crystallization. Since amides 11
and 12 are highly crystalline,9 we opted to run the amidation step
first, followed by crystallization and finally reduction (Scheme 1).
Moreover, this sequence allowed for elimination of one step from
our synthesis as esterification would not be required.
Synthesis of the amine fragment is summarized in Scheme 1.
Iodination of commercially available 3-bromo-4-methylbenzoic
acid with N-iodosuccinimide (NIS) proved to be highly selective
for the 5-position.10 Excess NIS (1.15 equiv) allowed complete
conversion and did not form any bis-iodo impurities after 2 h at
20 °C. Upon completion of the reaction, the mixture was slowly
poured into cold 1 N HCl to precipitate product 9. The solid was
then triturated in ACN to recover benzoic acid 9 in high purity.
Sonogashira coupling with methyl propynolate was performed
directly on benzoic acid 9. Gratifyingly, this reaction was highly
selective for coupling of the iodide over the bromide with no bis-
coupling observed.7 By performing the coupling at 60 °C and
with a large excess (15 equiv) of triethylamine, the palladium
loading could be reduced below 1 mol %. Use of excess base
allowed for complete conversion after 7 h, whereas longer
reaction times were needed with less Et3N. At the end of the
reaction, a simple extractive workup11 permitted rejection of
impurities and isolation of acid-alkyne 10 in 77% yield.
Tripropylphosphonic acid anhydride (T3P)-mediated12 ami-
dation between acid 10 and cyclopropyl amine afforded amide 11
1139
dx.doi.org/10.1021/op2001063 |Org. Process Res. Dev. 2011, 15, 1138–1148

Organic Process Research & Development
ARTICLE
Scheme 1. Synthesis of amine fragment 3-MsOH
in 93% yield. Crystallization of this intermediate from toluene/
heptane proceeded with high recovery (91%) to provide a purity
upgrade with low metal content.13 Hydrogenation with Adam’s
catalyst in PhMe with basic additives which had previously shown
to be beneficial14 showed high yield, with Et3N affording the best
purity profile for production of amide 12.
With our strategy firmly established we undertook the synth-
esis of acid (2) (Figure 2). The piperidine core could be
conveniently constructed from commercially available benzyl-
protected β-keto ester (4). The benzyl protecting group was
switched for a Boc followed by activation of the ketone as a
triflate as previously described.16,19 With the triflate in hand, the
stage was set for the installation of the required methylpyridone
in 1. The choice of coupling partner for this Suzuki reaction was
largely guided by the availability of the desired starting materials
in the appropriate timelines. While 4-bromo-2-methoxy pyridine
was shown to be a viable coupling partner in a one-pot
borylationꢀSuzuki sequence to provide 15 on gram-scale, it
was not available on kilogram scale within reasonable timelines.
An alternative preparation of this building block using 4-bromo-
2-chloropyridine was developed. Unfortunately, we were still
faced with unacceptable lead times for the raw material, and
another route had to be designed. Nonetheless, this sequence
could be used in future deliveries of 1 (eq 3).
Amide reduction was performed using borane prepared in situ
from NaBH4 and BF3 THF.15 Slow addition of BF3 THF (4.5
3
3
equiv) to a mixture of amide and sodium borohydride (4 equiv)
was needed to control the heat and gas evolution associated with
this transformation. The resulting amine was an oil. To facilitate
handling and storage, as well as guarantee high purity, its MsOH
salt was prepared. 3-MsOH was prepared via slow addition of
MsOH to a solution of the amine in MTBE/i-PrOAc in the
presence of 2 wt % seeds. Using this procedure, 3-MsOH was
obtained in 85% yield over the two steps with a purity
profile suitable for the final amidation step and end-game
(HPLC 97.8 A%). The MsOH salt is air stable, and no degrada-
tion was observed over >2 months at room temperature.
Acid Fragment (2) Synthesis. Synthesis of the acid fragment
(2) centered around asymmetric hydrogenation of a highly
substituted tetrahydropyridine.16,17 Success of this strategy
would be highly dependent on the nature of the protecting
group as well as the nature of the aryl substituent. Early screening
revealed liabilities associated with over-reduction of the pyri-
done-ring during high-pressure hydrogenation of substrates
derived from the requisite methylpyridone. Various pyridine
derivates as well as pyridine N-oxides were investigated as po-
tential alternative substrates for hydrogenation. Any surrogate
would have to be amenable to facile, high-yielding and scalable
pyridine-to-pyridone rearrangement. Substrates derived from
pyridine N-oxide also suffered from various side reactions
associated with NꢀO bond reduction which ultimately led to
catalyst deactivation. Installation of a 2-methoxy pyridine ap-
pendage was selected due to a superior performance in the
asymmetric hydrogenation (vide infra), as well as precedence for
conversion to the requisite N-methylpyridone.18
For the first GMP delivery of 1 we opted for the use of readily
available 2,4-dichloropyridine. SNAr reaction with sodium meth-
oxide afforded the 4-chloro-2-methoxypyridine as the major
1140
dx.doi.org/10.1021/op2001063 |Org. Process Res. Dev. 2011, 15, 1138–1148

Organic Process Research & Development
ARTICLE
Figure 2. Different strategies for synthesis of acid fragment 2.
product with solvent showing the greatest effect on product
ratios.20 Reaction run in toluene afforded a good ratio of 13:13a,
albeit with low conversion. Inspired by a recent report that
copper salts catalyze this type of displacement whilst improving
the isomeric ratio,21 we screened various copper catalysts
and identified CuI and N,N0-dimethylethylenediamine as an
effective catalyst/ligand combination to improve the rate of
reaction. We found that complete conversion could be achieved
within 20 h with reduced catalyst loading (2 mol %) if the
reaction was run more concentrated (1.0 M) and with a larger
excess of sodium methoxide (1.7 equiv). These optimized
conditions maintained a good ratio (100:1) of 13:13a and
afforded good isolated yield (70%) of the desired regioisomer
after distillation (eq 4).21
HBF4 OEt2. Tetrafluoroboric acid was required to protonate
3
the pyridyl-nitrogen, which otherwise acts as a catalyst
poison. High enantioselectivity (98% ee) and complete con-
version were obtained using the above reaction conditions.
Further pressure and temperature optimization showed the
reaction can proceed within 16 h at 350 psi of H2 at 50 °C with
reduced catalyst loadings (1.5 mol %), with no erosion of
enantioselectivity. Using these optimized conditions, the
catalyst loading could be further reduced to 1 mol % without
effect on conversion or rate.25 Following hydrogenation,
charcoal treatment was performed to reduce transition metal
levels, and chiral ester 17 was afforded in 85% yield and
>98.5% ee.
The crude stream from hydrogenation was then used directly
in the pyridine/pyridone conversion. Early studies showed
it was possible to effect this transformation with a large excess
of iodomethane at high temperature. This approach was viable
and was successfully performed on small scale in sealed
vessels. However, the low boiling point and high toxicity
associated with MeI prevented scale-up of this process. There-
fore, a superior alternative would have to be devised. Drawing
from previous experience with selective N-alkylation of
heterocycles,26 we were pleased to find that treatment of
crude 17 with trimethylsulfoxonium iodide and magnesium
hydroxide in DMF at 98 °C afforded clean conversion to the
N-methylpyridone within 2 h (Scheme 3). The use of wet DMF
was crucial to achieve high conversion; therefore, one
equivalent of water was added to the reaction.27 The crystalline
pyridone was isolated as a solid after a trituration in MTBE/i-
PrOAc in 79% yield. The final four-step sequence for the
preparation of acid 2 was carried out in one-pot (eq 5).16
Treatment of 18 with sodium ethoxide afforded a ratio of
trans:cis (18a:18) of approximately 6:1. Despite this modest
ratio, after TFA removal by addition of water and reprotection
with tert-butyl carbamate (18c), the ratio continued to upgrade
to ∼22:1; after hydrolysis with 5 N NaOH the ratio had
improved to >100:1 for crude acid 2. Trituration of the acid in
4-Chloro-2-methoxypyridine was then used in the Suzuki
reaction with vinyl boronꢀpinacol ester 14. Optimization of
the Suzuki-coupling conditions revealed that a combination of 1
mol % Pd(OAc)2, 2 mol % X-Phos,22 and aqueous K3PO4 in
2-methyltetrahydrofuran (2-MeTHF) at 50 °C afforded com-
plete conversion of the chloropyridine starting material to the
desired product 15 in a 75% assay yield. All attempts at
immediately performing hydrogenation on Boc protected 15
resulted in partial deprotection and catalyst poisoning.23 Thus, it
was necessary to perform a protecting group switch from Boc to
TFA. Fortunately, this protecting group switch, while inelegant,
could be performed as a through process without isolation of the
free piperidine (Scheme 2).
Olefin 16 was subjected to chiral hydrogenation previously
developed in our laboratories.16,24 Initial conditions were Ru-
(cod)(methallyl)2/SL-J212-1 (10 mol % preformed catalyst) at
room temperature, 500 psig, in 2-MeTHF, with 1.2 equiv of
1141
dx.doi.org/10.1021/op2001063 |Org. Process Res. Dev. 2011, 15, 1138–1148

Organic Process Research & Development
ARTICLE
Scheme 2. Synthesis of hydrogenation precursor 16
Scheme 3. Chiral hydrogenation and pyridone rearragement in the synthesis of acid 2
MTBE/2-MeTHF afforded complete rejection of the undesired
epi-2 affording 2 in 91% yield and >98.5% ee.
acid 2.6 However, previous experience led us to consider the use
of a p-toluenesulfonyl mixed anhydride as a more economic
mode of activation.28 In this procedure, the acid is first activated
as an electrophile using p-toluenesulfonyl chloride and N-
methylimidazole (NMI) in 2-MeTHF/MeCN. After addition
of the requisite amine, the desired amide was obtained. We were
pleased to find that we could directly use 3-MsOH salt in this
coupling reaction by simple modification: an additional equiva-
lent of NMI was used to neutralize the salt (Scheme 4). To make
the process more easily scalable, the salt was first dissolved in
DCM and added as a solution which allowed complete control of
the rate of addition. Careful optimization of this process with
respect to temperature was performed in order to avoid any
formation of the diastereomeric cis-19. Stress tests on the
reaction revealed that performing the reaction at higher tem-
perature increased the relative proportion of cis-diastereomer,
with room temperature reaction yielding as much as 5%.
Gratifyingly, by keeping the temperature below ꢀ5 °C during
the acid activation and subsequent addition of the amine, the
levels of cis-19 were kept below 0.5%. The crude solution of
amide 19 was directly used in the Boc deprotection step.
Development on this step revealed that TFA, H3PO4, formic
acid, and anhydrous HCl could potentially be used to perform
Final Amidation and End-Game. Work from the discovery
group had shown that HATU effectively coupled amine 3 with
1142
dx.doi.org/10.1021/op2001063 |Org. Process Res. Dev. 2011, 15, 1138–1148

Organic Process Research & Development
ARTICLE
Scheme 4. Final amidation and end-game
the reaction. The two acids which afforded the highest yields and
cleanest conversion were H3PO4 and HCl. Mainly for its ease of
manipulation, phosphoric acid was used for this step. Reactions
could be run in DCM as solvent with 9 equiv of H3PO4. It was
observed that the 1 phosphoric acid salt would gum out during
the course of the reaction. To obviate this issue, 0.5 mL/g of
MeOH was added to the reaction which provided a homoge-
neous solution. Extractive workup allowed for rejection of major
impurities. The solution of the free base was then carried forward
to the final salt formation and isolation.
1-HCl salt was identified as a stable crystalline and bioavailable
form. The HCl salt formation was achieved by dissolving the free
base in IPA (5 mL/g) with 0.1 mL/g of water and 2 wt % of seeds.
Seeds were critical in order to obtain the desired crystalline
polymorph. To this cooled solution is added HCl in diethyl ether
(1.3 equiv). After ageing for 18 h at room temperature, MTBE is
added, and the suspension is then cooled for maximum recovery
(95%) after filtration. 1-HCl was obtained as a white solid, HPLC
98.4 A%, >99.9% ee.
60 min at 20 °C, the mixture was filtered. The cake was rinsed
with cold (2 °C) acetonitrile (2 ꢁ 2 mL/g, 2 ꢁ 15 L). The
product was dried under nitrogen atmosphere for 15 h. Crude
mass (dried): 10.05 kg, 85%, as a pale-pink solid (93.7 A% by
HPLC). 1H NMR (400 MHz, DMSO-d6): δ 8.27 (s, 1 H); 8.02
(s, 1 H); 2.60 (s, 3 H). 13C NMR (101 MHz, DMSO-d6): δ
164.5, 144.4, 138.9, 132.9, 131.5, 122.9, 101.8, 29.1. IR (neat):
3068, 2683, 2533, 1683, 1534, 1370, 1284, 1141, 902, 766, 718.
HRMS: (ESI) m/z calcd for C8H6BrIO2 (M + H) 340.8669,
found 340.8666. Melting point: 152.5ꢀ154.2 °C.
3-Bromo-5-(3-methoxyprop-1-yn-1-yl)-4-methylbenzoic
Acid (9). 3-Bromo-5-iodo-4-methylbenzoic acid (8.00 kg, 1.0
equiv) is dissolved in THF (16 L, 2 mL/g) and Et3N (49 L, 15
equiv). The mixture was degassed bubbling nitrogen in the
solution for a period of 10 min. PdCl2(PPh3)2 (82 g, 0.005
equiv) was added, followed by the CuI (45 g, 0.01 equiv) and
methylpropargyl ether (2.97 L, 1.5 equiv). The mixture was
warmed to 60 °C and aged for a period of 7 h. After 5 h, 0.1%
PdCl2(PPh3)2 (16 g) and 0.2% CuI (9 g) were added. After 7 h,
HPLC showed >99% conversion. The volatiles were removed
under reduced pressure. The residue was dissolved in 1N NaOH
(8 mL/g, 64 L) and MTBE (4 mL/g, 32 L). The layers were
separated, the aqueous layer was washed 1 ꢁ MTBE (1 ꢁ 4 mL/
g, 32 L). The aqueous layer was acidified to pH 1 using 6 N HCl
(9 L), then extracted with i-PrOAc (1 ꢁ 7 mL/g, 1 ꢁ 55 L, 1 ꢁ
5 mL/g, 1 ꢁ 40 L, 1 ꢁ 2.5 mL/g, 1 ꢁ 20 L). Combined organic
layers were washed with brine (4 mL/g, 32 L). To the dark
i-PrOAc solution was added Darco KB-G (25 wt %, 2.0 kg). The
suspension was stirred for a period of 2 h. The suspension was
filtered over Solka Floc, rinsed 2 ꢁ 2.5 mL/g i-PrOAc (2 ꢁ 20 L).
The filtrate was treated with Na2SO4 (0.5 kg/20 L) for 16 h, then
concentrated to dryness under reduced pressure. Assay yield:
5.11 kg, 77%, as an orange solid (91.3 A% by HPLC). 1H NMR
(400 MHz, acetone-d6): δ 8.13 (s, 1 H); 8.00 (s, 1 H); 4.40 (s, 2
H); 3.41 (s, 3 H); 2.59 (s, 3 H). 13C NMR (101 MHz, acetone-
d6): δ 165.6, 145.1, 134.1, 133.0, 130.8, 125.6, 125.5, 92.3, 84.0,
60.4, 57.7, 21.8. IR (neat): 1692, 1548, 1421, 1304, 1099, 767.
HRMS: (ESI) m/z calcd for C12H11BrO3 (M + H) 280.9818,
found 280.9821. Melting point: 134.2ꢀ135.2 °C.
’ CONCLUSION
In summary, a highly convergent and scalable route to a novel
renin inhibitor was developed. The amine fragment (3) required
six steps and the chiral acid (2) nine steps. The longest linear
sequence is 11 steps starting from commercially available piper-
idine 4. The key chiral center was installed via an enantioselective
hydrogenation of a tetrasubstituted olefin and the amine frag-
ment synthesis features a highly selective Sonogashira coupling
and alkyne hydrogenation. This process was performed on
multikilogram scale to provide >2.5 kg of 1-HCl salt.
’ EXPERIMENTAL SECTION
3-Bromo-5-iodo-4-methylbenzoic Acid (5). 3-Bromo-4-
methyl benzoic acid (7.50 kg, 1.0 equiv) was dissolved in
H2SO4 (22.5 L, 3 mL/g). The mixture was cooled to 5 °C.
The N-iodosuccinimide (9.00 kg, 1.15 equiv) was added in six
portions (1.5 kg each) over a period of 1 h to control the
exotherm (internal temperature between 5 and 20 °C). Ten liters
of H2SO4 was added to facilitate stirring. The mixture was
warmed to 20 °C and aged for a period of 1 h. HPLC showed
>98% conversion to the desired iodo-aryl. The mixture was
poured onto a cold (2 °C) 1 N HCl solution (60 L, 8 mL/g) over
a period of 1 h at <35 °C. The iodo-aryl precipitated out of
solution. The mixture was stirred 20 min at 20 °C, then filtered,
rinsed 2 ꢁ 2.5 mL/g 1 N HCl (2 ꢁ 20 L). The pink precipitate
was triturated in acetonitrile (8 mL/g, 60 L). After stirring for
3-Bromo-N-cyclopropyl-5-(3-methoxyprop-1-yn-1-yl)-4-
methylbenzamide (10). 3-Bromo-5-(3-methoxyprop-1-yn-1-
yl)-4-methylbenzoic acid (5.05 kg, 1.0 equiv) is dissolved in
THF (30 L, 6 mL/g). DIPEA (6.84 L, 2.20 equiv) and the
cyclopropyl amine (1.70 L, 1.35 equiv) are added. The mixture
was cooled to 9 °C. Propylphosphonic anhydride (T3P) (50 wt %
in EtOAc, 14.4 L, 1.35 equiv) was added to the solution over a
period of 90 min to control the exotherm (9 to 19 °C). The
mixture was warmed to 19 °C and aged for a period of 15 min.
1143
dx.doi.org/10.1021/op2001063 |Org. Process Res. Dev. 2011, 15, 1138–1148

Organic Process Research & Development
ARTICLE
HPLC showed >98% conversion. The reaction was quenched by
pouring the solution on a cold (5 °C) saturated solution of
NH4Cl (8 mL/g, 40 L). The mixture was allowed to warm to
20 °C and stirred 60 min. The mixture was diluted with i-PrOAc
(8 mL/g, 40 L). The layers were separated, the aqueous layer was
back extracted 1 ꢁ i-PrOAc (6 mL/g, 30 L). Combined organic
layers were washed 1 ꢁ 1 N HCl (4 mL/g, 20 L), 1 ꢁ saturated
NaHCO3 (4 mL/g, 20 L), 1 ꢁ brine (4 mL/g, 20 L). The organic
layer was dried with Na2SO4 (500 g/20 L) over 18 h, and
concentrated under reduced pressure, then solvent switched to
toluene. Assay yield: 5.34 kg, 93%. The crude 3-bromo-N-
cyclopropyl-5-(3-methoxyprop-1-yn-1-yl)-4-methylbenzamide
(5.33kg, 1.0equiv) was then suspended in toluene(36L, 7 mL/g).
The mixture was warmed to 50 °C to dissolve the solids. The
solution was cooled to 25 °C. Seeds (0.5%, 25 g) were added, and
the mixture was aged at 25 °C for 45 min. Heptane (7 mL/g,
36 L) was added over a period of 5 min. The mixture was aged at
20 °C for 1 h, then cooled to 5 °C, aged 10 min, then filtered. The
cake was rinsed 2 ꢁ 3 mL/g cold (2 °C) toluene/heptane 1:2
(2 ꢁ 15 L), 1 ꢁ 3 mL/g heptane (15 L). The material was dried
on frit under a nitrogen atmosphere for a period of 40 h. Yield:
4.84 kg, 91% as a beige solid (97.5 A% by HPLC). 1H NMR (500
MHz, Acetone-d6): δ 8.01 (s, 1 H); 7.92 (s, 1 H); 7.86 (s, 1 H);
4.37 (s, 2 H); 3.39 (s, 3 H); 2.95ꢀ2.89 (m, 1 H); 2.53 (s, 3 H);
0.76ꢀ0.67 (m, 2 H); 0.68ꢀ0.60 (m, 2 H). 13C NMR (126 MHz,
Acetone-d6): δ 166.1, 142.9, 134.8, 132.1, 130.7, 125.5, 125.1,
91.8, 84.3, 60.4, 57.7, 23.9, 21.5, 6.3. IR (neat): 3268, 2821, 1635,
1525, 1313, 1098, 1028. HRMS: (ESI) m/z calcd for
C15H16BrNO2 (M + H) 322.0437, found 322.0433 Melting
point: 109.5ꢀ111.4 °C.
cake was rinsed with MTBE/heptane 1:10 (3 mL/g, 12 L), then
1 ꢁ 2 mL/g heptane (1 ꢁ 8 L). The material was dried on the frit
for 40 h under a nitrogen atmosphere. Yield: 3.23 kg = 77% as a
beige solid (97.3 A% by HPLC). 1H NMR (500 MHz, acetone-
d6): δ 7.87 (m, 2 H); 7.65 (s, 1 H); 3.35 (t, J = 6.12 Hz, 2 H); 3.27
(s, 3 H); 2.93ꢀ2.86 (m, 1 H); 2.75 (t, J = 7.91 Hz, 2 H); 2.39 (s, 3
H); 1.80ꢀ1.72 (m, 2 H); 0.75ꢀ0.67 (m, 2 H); 0.61ꢀ0.56 (m, 2
H). 13C NMR (126 MHz, acetone-d6): δ 167.0, 143.6, 139.4,
134.7, 129.7, 128.0, 126.2, 72.2, 58.5, 31.7, 31.0, 23.8, 19.2, 6.4.
IR (neat): 3281, 2924, 2856, 1636, 1520, 1458, 1308, 1110,
1021. HRMS: (ESI) m/z calcd for C15H20BrNO2 (M + H)
326.0750, found 326.0745. Melting point: 84.6ꢀ85.4 °C.
N-[3-Bromo-5-(3-methoxypropyl)-4-methylbenzyl]cyclo-
propanamine Mesylate (3-MsOH). 3-Bromo-N-cyclopropyl-
5-(3-methoxypropyl)-4-methyl benzamide (3.23 kg, 1.0 equiv)
is dissolved in THF (26 L, 8 mL/g). Sodium borohydride (1.69
kg, 4.5 equiv) was added to the solution in three portions (200 g,
500 g, 990 g). BF3 THF (5.5 L, 5.0 equiv) was added to the
3
mixture over a period of 2 h to control the exotherm and keep
the internal temperature below 30 °C (temperature range from
19.8 to 28.7 °C). After complete addition of the BF3.THF, the
mixture was warmed to 35 °C over 1.5 h and aged for a period
of 18 h. The mixture was slowly poured on 3 N HCl (10 mL/g,
33 L) over a period of 2 h, while under a nitrogen atmosphere.
The mixture was warmed to 50 °C and aged for a period of 2 h.
The mixture was cooled to 25 °C and diluted with MTBE/
heptane 1:1 (10 mL/g, 33 L). The layers were separated, and the
organic layer was back extracted 1 ꢁ 2 mL/g 2 N HCl (2 ꢁ 7 L).
Combined aqueous layers were basified to pH 14 using 10 N
NaOH (12 L), then extracted with MTBE (2 ꢁ 7 mL/g, 2 ꢁ
23 L). Combined organic layers were washed with half-brine
(5 mL/g, 16 L), brine (5 mL/g, 16 L). The organic layer was
dried with Na2SO4 for 15 h, filtered via in-line filter, and concen-
trated under reduced pressure. MTBE (31 L, 10 mL/g) and
i-PrOAc (19 L, 6 mL/g) are added to the crude N-[3-bromo-
5-(3-methoxypropyl)-4-methylbenzyl]cyclopropanamine. MsOH
(879 g, 0.91 equiv) was added as a THF (1 mL/g, 3.2 L) solution
over a period of 90 min. After the addition of ∼5% of the MsOH,
the solution was seeded (2% seeds, 62 g). The mixture was aged
2.5 h (final temperature = 24.8 °C), then filtered and rinsed 1 ꢁ
5 mL/g MTBE/i-PrOAc 5/1 (15 L), 1 ꢁ 5 mL/g MTBE (15 L).
The salt was dried on frit under a N2 atmosphere for 40 h. Yield:
3.67 kg, 89% as a snow-white powder (97.8 A% by HPLC).
3-Bromo-N-cyclopropyl-5-(3-methoxypropyl)-4-methyl
Benzamide (11). A visually clean and dry 10- gal Hastelloy
autoclave, pretested for leaks, was charged with the 3-bromo-N-
cyclopropyl-5-(3-methoxyprop-1-yn-1-yl)-4-methylbenzamide
(1.60 kg, 1.0 equiv) and toluene (15 L, 10 mL/g). To the mixture
was added the PtO2 (64 g, 0.04 equiv) as a toluene (0.5 L)
suspension, followed by the addition of the triethylamine
(140 mL, 0.20 equiv). Toluene (30 L) was added to dilute the
mixture. The vessel was purged three times with nitrogen, then
three times with hydrogen. The final pressure of hydrogen was 25
psig. The mixture was initially agitated at 350 rpm at 20 °C.
During the first 30 min, a 15 °C exotherm was observed (20 to
35 °C). After this initial exotherm, the agitation was increased to
650 rpm, and the mixture was stirred for another 4 h. The
solution was transferred to 5-gal plastic containers. The reaction
was performed three times on similar scale, then combined in a
140-L extractor. The assay yield on the combined organic layers
was 4.40 kg, 91% as a dark-brown oil. To the crude toluene
solution of 3-bromo-N-cyclopropyl-5-(3-methoxypropyl)-4-
methyl benzamide (4.20 kg, 1 equiv) was added Darco KB-G
(2.00 kg, 50 wt %), and the mixture was stirred at 20 °C for a
period of 1.5 h. The mixture was filtered on Solka Floc (∼2 kg),
rinsed 1 x toluene (3 mL/g, 12 L), 1 ꢁ toluene/MTBE 1:1
(3 mL/g, 12 L), 1 ꢁ toluene/MTBE 1:1 (2 mL/g, 8 L). Assay
yield: 3.87 kg = 92%. The filtrate was concentrated to dryness
(residual toluene: 1.2 equiv). The liquid residue was dissolved in
MTBE (8 mL/g, 31 L). The temperature of the mixture was
adjusted to 27 °C, then the solution was seeded with crystallized
amide (0.4%, 15.7 g). The mixture was aged at 27 °C for 30 min,
during which period the amide crystallized out of solution.
Heptane (8 mL/g, 31 L) was then added over 1.5 h to further
crystallize the amide. The mixture was aged 1 h, then filtered. The
1
Charcterized as free base: H NMR (500 MHz, CHCl3-d): δ
7.34 (s, 1 H); 7.01 (s, 1 H); 3.71 (s, 2 H); 3.36 (t, J = 6.2 Hz,
2 H); 3.32 (s, 3 H); 2.69 (t, J = 7.8 Hz, 2 H); 2.34 (s, 3 H);
2.13ꢀ2.07 (m, 1 H); 1.97 (s, 1 H); 1.83ꢀ1.75 (m, 2 H);
0.43ꢀ0.37 (m, 2 H); 0.37ꢀ0.33 (m, 2 H). 13C NMR (126 MHz,
acetone-d6): δ 142.0, 139.3, 134.0, 130.0, 128.2, 126.0, 71.8,
58.5, 52.7, 31.1, 30.2, 29.9, 18.6, 6.4. IR (neat): 3085, 2921,
2868, 2824, 1554, 1431, 1333, 1200, 1113, 1010. HRMS: (ESI)
m/z calcd for C15H22BrNO (M + H) 312.0957, found 312.0955.
2-Chloro-4-methoxypyridine (13). 2,4-Dichloropyridine
(2.09 kg) was dissolved in 14 L of toluene at room temperature.
CuI (53.7 g), DMEDA (150 mL) and NaOMe (1.14 kg) were
added in order under N2. The mixture was stirred at 105ꢀ110 °C
for 21 h. HPLC showed <0.4% of starting material, and the ratio
of 13:13a was 12.4 (270 nm). The mixture was filtered through
6 kg of silica with DCM (∼30 L). The desired product 13 elutes
first. Fractions were concentrated to a minimum volume at
40 °C, 74 mmHg, followed by atmospheric distillation through
1144
dx.doi.org/10.1021/op2001063 |Org. Process Res. Dev. 2011, 15, 1138–1148

Organic Process Research & Development
ARTICLE
a glass helix-packed column to remove toluene and provide the
title compound in 74% yield of 98 A% used as-is in the next step.
1-tert-Butyl 3-ethyl 20-methoxy-5,6-dihydro-4,40-bipyri-
dine-1,3(2H)-dicarboxylate (15). The vinyl triflate (prepared
according to literature precedent)19b in 55 L of dioxane was
treated with 7.36 kg of bis-pinacol diboron, 600 g of Cl2Pd-
(dppf), 8.13 kg of KOAc. The mixture was degassed with N2 and
heated to 90 °C. After aging for 8 h, the reaction was at 92.5%
conversion. The reaction was allowed to cool slowly to room
temperature overnight. The following morning the reaction had
reached complete conversion. Isopropyl acetate (45 L) and water
(80 L) were added and stirred for 10 min, and the layers were
separated. The organic layer was washed with 40 L of water and
then concentrated using a batch concentrator and flushed with 4
L of heptane to give an oil which was further diluted with 10 L of
heptane. The oil was purified by silica gel eluting with heptane
then a gradient of increasing amounts of ethyl acetate (up to
20%). The product-rich fractions were combined and stripped to
an oil using the batch concentrator to give 7.9 kg (75%). A
separate 22-L round-bottom flask (RBF) was charged with 7.75 L
of 2-MeTHF which was degassed with multiple vacuum/nitro-
gen purges. To this solution was added X-Phos and palladium
acetate which were stirred at room temperature for 5 min
(a burgundy-colored solution formed). The solution was aged
for one hour. This solution was then transferred to a 75-L RBF
charged with 2.37 kg of the chloromethoxypyridine, the vinyl
boronate, 10.36 L 2 M K3PO4 solution, and 15 L of 2-MeTHF.
The solution was degassed with multiple vacuum/nitrogen
purges. This solution was then heated to 58 °C at which point
an exotherm further heated the mixture to 69 °C over the next
20 min. The reaction mixture then maintained a temperature of
67 °C for 1 h. The reaction was assayed and was complete by
HPLC. The reaction mixture was then cooled to 25 °C, 600 g of
EcoSorb C-941 was added, and the mixture was aged for 30 min.
The mixture was then filtered through Solka Floc. The flask was
rinsed with 2 L of 2-MeTHF. The cake was then washed with an
additional 5 L of 2-MeTHF. The combined filtrate and wash
were transferred into a 100-L extractor, and the aqueous layer
was drawn off. The organic layer was washed with 14 L of dilute
brine. The organic layer was then collected and transferred to a
50-L RBF. The solvent was then evaporated off and flushed with
16 L of heptane to remove residual 2-MeTHF. The mixture was
stored on ice overnight and then chilled to ꢀ7 °C to give a thick
slurry. The solids were filtered and rinsed with 2 ꢁ 2 L of cold
heptane. The cake was then washed with 3 ꢁ 4 L of cold heptane
and sucked dry to give 4.34 kg of a tan-gray solid. 1H NMR (400
MHz, acetone-d6): δ 8.09 (d, J = 5.24 Hz, 1 H); 6.74 (d, J = 5.23
Hz, 1 H); 6.55 (s, 1 H); 4.20 (s, 2 H); 3.93 (q, J = 7.11 Hz, 2 H);
3.87 (s, 3 H); 3.60 (s, 2 H); 2.49ꢀ2.45 (m, 2 H); 1.46 (s, 9 H);
0.97ꢀ0.89 (m, 3 H). 13C NMR (101 MHz, acetone-d6): δ
166.4, 165.0, 154.8, 153.9, 147.4, 144.5, 126.8, 116.5, 109.3, 80.1,
72.7, 60.9, 53.5, 49.3, 28.5, 27.2, 13.8. IR(νmax /cmꢀ1): 2981,
1696, 1387, 1235, 760; HRMS calculated for C19H26N2O5 (M+):
362.1842; Found: 362.1823. Melting point: (EtOAc/Hex)
85.5ꢀ86.6 °C.
reaction was cooled. The solvent was then switched for IPAc
(12 L), and 16 L of 10% sodium carbonate was added along with
900 g of solid sodium carbonate (pH ≈ 11). The layers were
separated, and the aqueous layer was washed with 15 L of IPAc
twice. The combined organic layers were washed with 10 L of
water. This organic layer was concentrated to minimum volume,
and then flushed with 5 L of 2-MeTHF. The volume was ajusted
to 28 L of 2-MeTHF (Kf = 200). Triethylamine (12.7 L ) was
added to the mixture followed by TFAA (1.9 L) via addition
funnel. The addition causes a temperature rise to 52 °C. The
reaction was complete 5 min after the addition. The solution was
cooled to 30 °C, 30 L of water was added. The layers were
separated, and the organic layer was washed with 10 L of water
and 12 L of 5 N HCl, and then the organic layer was washed with
15 L of water. The pH of the aqueous layer was then adjusted to 8
using 5 N NaOH, and then this aqueous layer was extracted with
2-MeTHF. The combined organic layers were dried over sodium
sulfate. The suspension was then filtered and concentrated to ∼8
L (Kf = 500). This solution was assayed to contain 3.6ꢀ3.7 kg
1
(84ꢀ86% yield). H NMR (600 MHz, DMSO-d6, 363 K): δ
8.10 (d, J = 5.20 Hz, 1 H); 6.77 (d, J = 5.17 Hz, 1 H); 6.60 (s, 1
H); 4.40 (s, 2 H); 3.97ꢀ3.91 (m, 2 H); 3.86 (s, 3 H); 3.78 (s, 2
H); 2.57 (s, 2 H); 0.92 (s, 3 H). 13C NMR: Peaks were not
assigned because of rotamers at room temperature. Spectra
shown in Supporting Information. IR(νmax /cmꢀ1): 2984,
1692, 1389, 1131, 1042. HRMS calculated for C16H17F3N2O4
(M+): 358.1140; Found: 358.1130.
Ethyl (3S, 4S)-4-(1-Methyl-2-oxo-1,2-dihydropyridin-4-yl)-
1-(trifluoroacetyl)piperidine-3-Carboxylate (18). In a nitro-
gen-filled glovebox, (O2 < 10 ppm) (R)-S-SL-J212-1 (46.9 g)
was combined with (COD)Ru(Me-allyl)2 (23.0 g) in a 1.0-L
screw-cap bottle. CH2Cl2 (500 mL, N2 degassed, anhydrous)
was added. and the bottle was sealed with a cap. A red-brown
catalyst solution was observed. The catalyst solution was trans-
ferred to a 2.0-L stainless steel vessel (see below) followed by a
CH2Cl2 rinse (1.2 L). Additional CH2Cl2 (1.7 L) was added to a
2.0-L stainless steel vessel. The two stainless steel vessels were
removed from the glovebox and connected via a t-joint. The
t-joint was purged with a nitrogen stream prior to connection to
the autoclave.
Ethyl 20-methoxy-1-(trifluoroacetyl)-1,2,5,6-tetrahydro-4,40-
bipyridine-3-carboxylate (5.95 kg, 28.8 wt % in 2-MeTHF) was
drawn into the autoclave under partial vacuum followed by
2-MeTHF rinse (4.0 L). HBF4 OEt2 (781 mL) was drawn into
3
the autoclave under partial vacuum followed by 2-MeTHF rinse
(1.0 L). An exotherm was observed upon addition of HBF4
OEt2 to 33 °C. An additional volume of 2-MeTHF (2.7 L) was
3
added to the autoclave. The autoclave solution was cooled to
25 °C before catalyst addition. The autoclave was sealed and
purged with N2 (3 ꢁ 40 psig). The catalyst assembly described
above was connected to the autoclave via flex tubing and under
partial vaccum. The catalyst solution was drawn into the autoclave
followed by the CH2Cl2 (1.8 L) rinse. The reactor was sealed and
the headspace exchanged with H2 (3 ꢁ 350 psig). The H2
pressure was set at 300 psig, agitation was begun, and the reaction
was heated to 50 °C. Once the reaction had reached a tempera-
ture of 45 °C, the H2 pressure was adjusted to reach 350 psig. The
reaction was sampled after 5 h. The product conversion was
observed to be 99.4%. The product ee was observed to be above
98.5%. The batch (17 L) was removed through the eductor tube
prior to cooling and drummed. CH2Cl2 (16.0 L) was added
to the autoclave, and agitation was performed for 30 min. The
Ethyl 20-Methoxy-1-(trifluoroacetyl)-1,2,5,6-tetrahydro-
4,40-bipyridine-3-carboxylate (16). 1-tert-Butyl 3-ethyl 20-
methoxy-5,6-dihydro-4,40-bipyridine-1,3(2H)-dicarboxylate was
dissolved (4.34 kg, 11.98 mol) in ethanol (28.2 L). To this
mixture was added methanesulfonic acid (1.944 L, 29.9 mol).
The mixture was heated to 55 °C, then slowly up to 65 °C. After
heating for 1.5 h, 99.65% conversion was obtained, and the
1145
dx.doi.org/10.1021/op2001063 |Org. Process Res. Dev. 2011, 15, 1138–1148

Organic Process Research & Development
ARTICLE
solution was then filtered on Solka Floc, and the organics were
then washed with brine.
bath for 1 h. HPLC showed complete hydrolysis to the acid, with
a ratio of ∼trans:cis = 150:1. The mixture was cooled to 30 °C,
and the solution was concentrated to ∼20 L volume, such that
most of the ethanol was removed. The material was transferred to
a 50 L extractor, and washed with MTBE (6 L). The aqueous
layer (pH ≈ 14) was recharged to the extractor, then acidified
to pH ≈ 1 by addition of concentrated HCl (2.7 L) and 6 N HCl
(2 L) to afford a slurry. This slurry was extracted with 2-methyl-
tetrahydrofuran (2-MeTHF, 15 kg). The aqueous layer was then
back-extracted with 2-MeTHF (5 kg). The combined organic
fractions were then washed with half-brine (10 L). The organics
were then concentrated in vacuo to afford a slushy pale-yellow
solid, ∼4.3 kg. The solid was transferred to a 50-L RBF equipped
with mechanical stirrer, nitrogen inlet, and temperature probe
and suspended in MTBE (12 L). The slurry was stirred at room
temperature for 18 h. The slurry was filtered, washing the flask
with 2 L of the residual mother liquors. The filter cake was then
washed with MTBE (1 L) and dried under vacuum under an N2-
tent. After 4 h, the white solid was transferred into five Pyrex trays
and stored in the vacuum oven at room temperature. After 3 days,
2.10 kg (91% yield) of white solid was collected. ICP-Metals
The crude 2-methoxypyridyl ester was transferred to a 50-L
RBF flask, and solvent was switched to DMF (3.4 L/kg, 10.7 L).
The flask was charged with trimethylsulfoxonium iodide (3.85
kg) and magnesium hydroxide (1.02 kg). The mixture was heated
to 98 °C and aged for a period of 2 h. HPLC showed >98%
conversion to the desired N-methylpyrrolidone. The mixture was
transferred via vacuum into a 50-L extractor charged with
biphasic solution of 4 N HCl (9 L) and dichloromethane (8 L)
over 20 min with cooling to keep the internal temperature below
30 °C; the flask was rinsed with 2 L dichloromethane and 3 L 2 N
HCl and this solution was charged to the extractor as well. The
organic layer was removed and the remaining aqueous layer back-
extracted with dichloromethane (6 L). The organic layers were
then subjected to successive washes: water (6 L ꢁ 2), 10% LiClꢀ
water (6 L ꢁ 2) then 1% LiClꢀwater (6 L ꢁ 1). The organics
were then diluted with MTBE (40 L) then washed with 10%
LiClꢀwater (20 L ꢁ 3). Organics were collected, then all three
10% LiClꢀwater washes were back-extracted with 9:1 i-PrOAc/
DCM (20 L ꢁ 2). All organic fractions were transferred via in-
line filter (10 μm) into a 50-L RBF connected to batch
concentrator to remove solvents in vacuo. After concentration
to a thick, yellow slurry, the material was flushed with MTBE (20 L)
and concentrated to ∼8 L volume. i-PrOAc (1 L) was added,
rinsing the flask edges, and additional MTBE was added until the
volume was ∼12 L. The slurry was stirred at room temperature
for 24 h. The slurry was filtered into a Teflon filter-pot, washing
the flask with 10% i-PrOAc/MTBE (2 L). The filter cake was
then washed with 10% i-PrOAc/MTBE (2 L ꢁ 2) and dried
under vacuum under an N2-tent. After 3 h, the yellow solid was
transferred into three Pyrex trays and stored in the vacuum oven.
After 5 days, 2.50 kg (79% yield) of mustard-yellow solid was
collected. ICP-Metals analysis indicated Ru 95 ppm, Fe 65 ppm,
1
analysis indicated Ru, Fe, and Pd were all <10 ppm. H NMR
(400 MHz, DMSO-d6): δ 12.42 (s, 1 H); 7.57 (d, J = 6.83 Hz, 1
H); 6.20ꢀ6.14 (m, 2 H); 4.20 (s, 1 H); 4.00 (s, 1 H); 3.34 (s, 3
H); 2.81ꢀ2.43 (m, 4 H); 1.66 (d, J = 12.98 Hz, 1 H); 1.40 (m, 10
H). 13C NMR (101 MHz, DMSO-d6): δ 173.0, 161.9, 156.0,
153.6, 139.2, 116.3, 105.1, 79.1, 46.2, 43.7, 36.3, 30.9, 28.0.
(overlapping peaks, no other peaks detected even after pro-
longed scans.) IR(νmax /cmꢀ1): 2919, 1682, 1533, 1426, 1167,
785; [r]D: +19.9° (c = 3.0, CH2Cl2). HRMS calculated for
C17H24N2O5 (M+): 336.1685; Found: 336.1680; Melting point
(water): 208.0ꢀ209.1 °C.
tert-Butyl (3R,4S)-3-{[[3-Bromo-5-(3-methoxypropyl)-4-
methylbenzyl](cyclopropyl)amino]carbonyl}-4-(1-methyl-
2-oxo-1,2-dihydropyridin-4-yl)piperidine-1-carboxylate
(19). (3R,4S)-1-(tert-Butoxycarbonyl)-4-(1-methyl-2-oxo-1,2-
dihydropyridin-4-yl)piperidine-3-carboxylic acid (1.99 kg, 5.92
mol) was suspended in 2-MeTHF (24 L). TsCl (1.466 kg, 7.69
mol) was then added followed by MeCN (4 L). The temperature
was cooled to ꢀ20 °C (IPA/dry ice). To this stirring suspension
was added NMI (2.122 L, 26.6 mol) via addition funnel over
25 min (initial temp ꢀ19.9 °C, final temperature = ꢀ3.7 °C, max
T = ꢀ2.9 °C). Another 5 L of MeCN was added to dissolve the
solids forming which caused a colour change to yellow and
dissolved the suspension. The reaction was stirred at ꢀ8 °C for
55 min. (max T = ꢀ5 °C, min T = ꢀ9 °C). An aliquot was
removed and quenched with R-methylbenzylamine. HPLC
analysis revealed near complete (>95%) consumption of the
starting acid. The amine-MsOH salt (2.66 kg, 6.51 mol) was
dissolved in DCM (7.96 L) and added to the reaction which
caused the reaction to become immediately clear. The solution
was added over 25 min, and the temperature was at 3 °C by the
end of the addition. The reaction was left stirring overnight at
room temperature. In the morning an aliquot was analyzed by
HPLC showing near complete conversion. Acetic anhydride
(0.084 L, 0.887 mol) was added to the mixture and stirred at
room temperature for another 30 min. The reaction was then
concentrated by removing 20 L of solvent. The resulting
suspension was transferred to the 100-L extractor. Six liters of
water and 2 L of i-PrOAc were added to rinse the reactor. Four
liters of water, 10 L of brine, and 18 L of i-PrOAc were then
added, and the layers were cut. The aqueous layer was then back
1
and Pd 22 ppm. Product was 92.7 A% by HPLC analysis. H
NMR (600 MHz, DMSO-d6, 323 K) Major rotamer: δ 7.57
(dd, J = 7.00, 4.61 Hz, 1 H); 6.15ꢀ6.11 (m, 2 H); 4.54 (d, J =
13.57 Hz, 1 H); 3.98 (d, J = 14.07 Hz, 1 H); 3.92ꢀ3.80 (m, 2 H);
3.39 (d, J = 13.33 Hz, 1 H); 3.34 (s, 3 H); 3.25 (dd, J = 13.59, 3.55
Hz, 1 H); 3.14 (s, 1 H); 3.07ꢀ3.02 (m, 1 H); 2.30ꢀ2.15 (m, 1
H); 1.78 (t, J = 13.70 Hz, 1 H); 0.98 (td, J = 7.09, 3.95 Hz, 3 H).
13C NMR: Peaks were not assigned because of rotamers at
room temperature. Spectra shown in Supporting Information.
IR(νmax /cmꢀ1): 1681, 1657, 1594, 1139, 867. [r]D: ꢀ18.8°
(c = 3.0, CH2Cl2). HRMS calculated for C16H19F3N2O4 (M+):
360.1297; Found: 360.1306; Melting point: (i-PrOAc/MTBE):
140.1ꢀ141.3 °C.
(3R,4S)-1-(tert-Butoxycarbonyl)-4-(1-methyl-2-oxo-1,2-
dihydropyridin-4-yl)piperidine-3-carboxylic Acid (2).
(3R,4S)-1-(tert-butoxycarbonyl)-4-(1-methyl-2-oxo-1,2-dihy-
dropyridin-4-yl)piperidine-3-carboxylic acid (2.46 kg) was dis-
solved in ethanol (4.1 L/kg, 10 L). Solid sodium ethoxide (0.56
kg) was added to the flask, (the temperature climbed from 18 to
28 °C) and the yellow solution was stirred for 1 h. Water
(150 mL) was added to the reaction mixture, over 5 min. (the
temperature climbed from 24 to 29 °C) and the solution was
stirred for 1 h. Boc-anhydride (1.90 L) was added to the reaction
mixture, over 15 min. (the temperature climbed from 22 to
29 °C) and the solution was stirred for 1 h. HPLC showed
complete conversion to the Boc-piperidine, with a ratio of
∼trans:cis =23:1. Sodium hydroxide (17.1 L, 2M) was added
to the flask, and the solution heated to 70 °C via external steam
1146
dx.doi.org/10.1021/op2001063 |Org. Process Res. Dev. 2011, 15, 1138–1148

Organic Process Research & Development
ARTICLE
extracted with 15 L of i-PrOAc. The organic layers were then
combined, cooled to 14 °C, and washed with 10 mL/g of 0.5 N
HCl. This was followed by a 2/3 sat. NaHCO3 (10 mL/g) and
brine wash (10 mL/g). The crude organic solution was carried
through to the next step. 1H NMR (400 MHz, CHCl3-d): δ 7.07
(d, J = 6.82 Hz, 1 H); 7.01 (s, 1 H); 6.81 (s, 1 H); 6.32 (s, 1 H);
5.99 (d, J = 6.64 Hz, 1 H); 4.45ꢀ4.81 (m, 1 H); 4.43ꢀ3.70 (m, 3
H); 3.40 (s, 3 H); 3.36ꢀ3.28 (m, 6 H); 2.98 (t, J = 11.68 Hz, 1
H); 2.80 (s, 2 H); 2.62 (t, J = 7.74 Hz, 2 H); 2.39 (s, 1 H); 2.29 (s,
3 H); 1.77ꢀ1.68 (m, 3 H); 1.44 (s, 10 H); 0.84 (s, 3 H); 0.64 (s, 1
H). 13C NMR (101 MHz, CHCl3-d): δ 174.0, 162.9, 156.5,
154.2, 142.0, 137.8, 136.6, 134.3, 129.6, 128.3, 125.9, 116.8,
107.5, 80.0, 77.2, 71.7, 58.5, 48.9, 46.7, 45.5, 44.4, 43.4, 37.3, 31.1,
30.2, 29.7, 28.4, 18.7, 9.4. IR(νmax /cm-1): 2921, 1695, 1646,
1421, 1116, 877. [r]D: +29.0° (c = 3.0, CH2Cl2). HRMS
calculated for C32H44BrN3O5 (M+): 629.2464; Found:
629.2483; Melting point (2-MeTHF): 172.4ꢀ173.8 °C
acetone/ice bath to ꢀ3.5 °C. Added to the mixture was 61.5 g of
1-HCl seeds along with water (0.254 L). HCl in ether (3.1 L, 6.20
mol) was then added over 60 min (start T = ꢀ3.5 °C, end =
11 °C). The solution was then warmed to room temperature and
aged overnight. The next day, MTBE (25.9 L) was added over 30
min, and the solution was cooled to 5 °C. After aging for 1 h, the
suspension was then filtered. The flask was rinsed with combined
4 L of MTBE and 4 L of mother liquors, then another 4 L of
MTBE. The pad was rinsed with 8 L of 1:1 IPA/MTBE. The solid
was then dried under a N2 tent overnight. In the morning (drying
for 20 h) a small sample was taken, and residual solvent analysis
was performed which showed it to be dry. After drying for an
additional 5 h, 1 HCl (2.57 kg, 4.53 mol, 95% recovery) was
collected as a white solid (98.4 LCAP, > 99.9% ee, residual
solvents: 0.07% Et2O and 0.42% IPA, Moisture 0.3%). 1H NMR
(400 MHz, H2O-d2): δ 7.16 (d, J = 6.85 Hz, 1 H); 6.74ꢀ6.56 (m,
1 H); 6.57 (s, 1 H); 6.20 (s, 1 H); 6.10 (d, J = 6.94 Hz, 1 H); 4.51
(d, J = 14.48 Hz, 1 H); 3.81 (t, J = 11.54 Hz, 1 H); 3.50ꢀ3.30 (m,
3 H); 3.22ꢀ3.06 (m, 8 H); 3.06ꢀ2.78 (m, 3 H); 2.31 (d, J = 9.94
Hz, 3 H); 1.95 (s, 3 H); 1.88 (d, J = 14.15 Hz, 1 H); 1.77 (q, J =
13.60 Hz, 1 H); 1.45 (t, J = 7.48 Hz, 2 H); 0.94 (d, J = 7.68 Hz, 1
H); 0.75ꢀ0.62 (m, 2 H); 0.50 (s, 1 H). 13C NMR (101 MHz,
H2O-d2): δ 173.7, 164.0, 156.2, 142.3, 139.9, 136.0, 134.7, 129.6,
128.8, 125.6, 116.3, 108.6, 71.8, 58.1, 45.2, 43.4, 41.7, 38.2, 30.8,
30.6, 29.6, 28.3, 18.5, 10.6, 8.2. IR(νmax /cmꢀ1): 1657, 1597,
1398. 1118, 962; HRMS calculated for free base C27H36BrN3O3
(M+): 529.1940; Found: 529.1923. Melting point (IPA/
MTBE): 222.1ꢀ222.9 °C.
(3R,4S)-N-[3-Bromo-5-(3-methoxypropyl)-4-methylbenzyl]-
N-cyclopropyl-4-(1-methyl-2-oxo-1,2-dihydropyridin-4-yl)-
piperidine-3-carboxamide (1). The 2-MeTHF solution of
N-Boc-1 (3.568 kg, 5.66 mol) was placed via in-line filter into
a 100-L RBF. The solvent was evaporated, which caused solids to
crash out. After 30 L were evaporated and all of organics phases
had been added, MeOH was added slowly while continuing
evaporation which caused everything to dissolve. Once the
solution was completely clear, addition of MeOH was ceased
and evaporation was continued. After evaporation of another
25 L, DCM was added to the mixture while continuing evapora-
tion. After 35 L of DCM had been added the evaporation was
ceased. NMR analysis suggests that DCM/MeOH ratio is >15:1.
The final volume was adjusted by adding 1 L of MeOH. HPLC
analysis revealed the solution to be 25 wt%. To this solution was
added slowly via addition funnel phosphoric acid (3.5 L, 51.8
mol) at 15 °C. This caused a controlled exotherm which rose to
∼26 °C. The reaction was heated to 28 °C which caused a slow,
delayed exotherm to 34.3 °C over 40 min. After 1.25 h the
reaction was complete by HPLC analysis. The reaction was
cooled to 10 °C and diluted with water (12 L, 4 mL/g). The
contents of the flask were transferred to the extractor which is
precharged with MTBE (∼5.5 mL/g ∼20 L). The flask was
rinsed with 4 L of MTBE and 4 L of water which were transferred
to the extractor, and the layers were cut. The combined aqueous
layers were washed with 20 L of MTBE/DCM (4:1). The acidic
aqueous layer was then added to a cooled (10 °C) solution of
KOH 8 N (12.9 L, 103 mol) and 20 L of 2-MeTHF. After the
addition was complete (causes solution to become milky) pH of
the aqueous (wet pH paper) was 13. Eight liters of 2-MeTHF and
28 L of i-PrOAc were then added, and the phases were cut.
The aqueous was then reloaded and extracted with 30 L of 1:1
(2-MeTHF/i-PrOAc). The organics were then combined and
washed with 20 L of 2/3 brine (3.4 kg of NaCl) and then stored
over the weekend at room temperature protected from light
(with foil). Assay of the organics revealed 1 (2.615 kg, 4.93 mol,
83% over two steps) with LCAP of ∼97%.
’ ASSOCIATED CONTENT
S
Supporting Information. Complete characterization
b
data and spectra for all new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org/
’ AUTHOR INFORMATION
Corresponding Author
*Louis-Charles.Campeau@merck.com.
’ ACKNOWLEDGMENT
We thank Ravi Sharma and Wayne Mullet for analytical
support. Austin Chen is thanked for helpful discussions. We thank
Dan Sorensen for helpful assistance with NMR experiments.
’ REFERENCES
(1) Kearney, P. M.; Whelton, M.; Reynolds, K.; Muntner, P.;
Whelton, P. K.; He, H. Lancet 2005, 365, 217.
(2) Ong, K. L.; Cheung, B. M. Y.; Man, Y. B.; Lau, C. P.; Lam, K. S. L.
Hypertension 2007, 49, 69.
(3) (a) MacGregor, G. A.; Markandu, N. D.; Roulston, J. E.; Jones,
J. C.; Morton, J. J. Nature 1981, 291, 329. (b) Weber, M. A. Am. J.
Hypertens. 1999, 12, 189S. (c) Weir, M. R.; Dzau, V. J. Am. J. Hypertens.
1999, 12, 205S. (d) Brewster, U. C.; Perazella, M. A. Am. J. Med. 2004,
116, 263.
(4) (a) Zaman, M. A.; Oparil, S.; Calhoun, D. A. Nat. Rev. Drug
Discovery 2002, 1, 621. (b) Kasani, A.; Subedi, R.; Stier, M.; Holsworth,
D. D.; Maiti, S. N. Heterocycles 2007, 73, 47. (c) Fisher, N. D. L.;
Hollenberg, N. K. J. Am. Soc. Nephrol. 2005, 16, 592. (d) Rahuel, J.;
Rasetti, V.; Maibaum, J.; Rueger, H.; Goschke, R.; Cohen, N.-C.; Stutz,
S.; Cumin, F.; Fuhrer, W.; Wood, J. M.; Gruetter, M. G. Chem. Biol.
2000, 7, 493. (e) Wood, J. M.; Stanton, J. L.; Hofbauer, K. G. J. Enzyme
(3R,4S)-N-[3-Bromo-5-(3-methoxypropyl)-4-methylbenzyl]-
N-cyclopropyl-4-(1-methyl-2-oxo-1,2-dihydropyridin-4-yl)-
piperidine-3-carboxamide (1-HCl). The solution of 1 (2.54 kg,
4.79 mol) was charged in 100-L flask, and the solvent was
switched to IPA. The free base was flushed with 20 L of IPA,
and the volume was adjusted to yield a 33 wt % solution in
IPA (∼2 mL/g). Another 7.5 L of IPA was added, bringing the
final concentration at ∼5 mL/g. The solution was cooled in an
1147
dx.doi.org/10.1021/op2001063 |Org. Process Res. Dev. 2011, 15, 1138–1148

Organic Process Research & Development
ARTICLE
Inhib. 1987, 1, 169. (f) Rahuel, J.; Priestle, J. P.; Gruetter, M. G. J. Struct.
Biol. 1991, 107, 227. (e) Webb, R. L.; Schiering, N.; Sedrani, N.;
Maibaum, J. J. Med. Chem. 2010, 53, 7490.
D. E.; Miller, R. A.; Reamer, R. A.; Weissman, S. A. Org. Process Res. Dev.
2011, 15, 73.
(27) Reactions under anhydrous conditions were very slow and led
to incomplete conversion of the pyridine to the pyridone.
(28) Wakasugi, K.; Iida, A.; Misaki, T.; Nishii, Y.; Tanabe, Y. Adv.
Synth. Catal. 2003, 345, 1209.
(5) Norris, K.; Vaughn, C. Expert Rev. Cardiovasc. Ther. 2003, 1, 51.
(6) (a) Chen, A.; Campeau, L.-C.; Cauchon, E.; Chefson, A.;
Ducharme, Y.; Dubꢀe, D.; Fournier, P.-A.; Gagnꢀe, S.; Grimm, E.; Han,
Y.; Houle, R.; Huang, J.-Q.; Lacombe, P.; Lalibertꢀe, S.; Lꢀevesque, J.-F.;
Liu, S.; MacDonald, D.; Mackay, B.; McKay, D.; Regan, C.; Regan, H.;
Stump, G.; Toulmond, S. Bioorg. Med. Chem. Lett. 2011, 21, 3970.
(b) Chen, A.; Bayly, C.; Bezenc-on, O.; Richard-Bildstein, S.; Dubꢀe, D.;
Dubꢀe, L.; Gagnꢀe, S.; Gallant, M.; Gaudreault, M.; Grimm, E.; Houle, R.;
Lacombe, L.; Lalibertꢀe, S.; Lꢀevesque, J.-F.; Liu, S.; MacDonald, D.;
Mackay, B.; Martin, D.; McKay, D.; Powell, D.; Remen, L.; Soisson, S.;
Toulmond, S. Bioorg. Med. Chem. Lett. 2010, 20, 2204.
(7) Sonogashira coupling on the ester 6 afforded only ∼5% of bis-
coupled product.
(8) For effects of metal ion salts on hydrogenations, see:Yang, X.;
Liu, H. Appl. Catal. 1997, 164, 197.
(9) In contrast esters 6 and 7 are oils.
(10) Elliott, E. L.; Ray, C. R.; Kraft, S.; Atkins, J. R.; Moore, J. S. J.
Org. Chem. 2006, 71, 5282.
(11) The acid dissolved in MTBE (4 mL/g) could be extracted into
1 N NaOH (8 mL/g), and additional MTBE washes removed impu-
rities. Reacidification of the aqueous layer using 6 N HCl followed by
extraction with i-PrOAc resulted in high purity of the desired compound.
(12) Kleiner, H.-J.; Wissmann, H. Angew. Chem., Int. Ed. Engl. 1980,
19, 133.
(13) ICP analysis showed residual Pd = 31 ppm and Cu = 2 ppm.
(14) Both Cs2CO3 and Et3N were screened.
(15) For the large-scale in situ preparation of BH3ꢀTHF see: Kanth,
J. V. B.; Brown, H. C. Inorg. Chem. 2000, 39, 1795 and references therein.
(16) Molinaro, C.; Roy, A.; Schultz, S.; Lau, S.; Trinh, T.; Angelaud,
R.; O’Shea, P. D.; Cameron, M.; Corley, E.; Steinhuebel, D.; Weisel, M.;
Krska, S. J. Org. Chem. 2011, 76, 1062.
(17) For a review on the use of asymmetric hydrogenation at
Merck, see: Shultz, S. C.; Krska, S. K. Acc. Chem. Res. 2007, 12, 1320.
(18) (a) Ikehara, M.; Tanaka, S. Tetrahedron Lett. 1974, 15, 497.
(b) Lanni, E. L.; Bosscher, M. A.; Ooms, B. D.; Shandro, C. A.;
Ellsworth, B. A.; Anderson, C. E. J. Org. Chem. 2008, 73, 6425.
(c) Bowman, W. R.; Bridge, C. F. Synth. Commun. 1999, 29, 4051.
(d) Itami, K.; Yamazaki, D.; Yoshida, J.-I. Org. Lett. 2003, 5, 2161.
(19) (a) Elitzin, V. I.; Harvey, K. A.; Kim, H.; Salmons, M.; Sharp,
M. J.; Tabet, E. A.; Toczko, M. A. Org. Process Res. Dev. 2010, 14, 912.
(20) Solvents affording the best ratios (>10:1) were dioxane (100%
conversion), toluene (<50% conversion), and CPME (90% conversion).
A reaction in dioxane on 100-g scale afforded a ratio, after fractional
distillation, of ∼100:1 in 70% isolated yield.
(21) Ruggeri, S. G.; Vanderplas, B. C.; Anderson, B. G.; Breitenbach,
R.; Urban, F. J.; Stewart, A. M., III; Young, G. R. Org. Process Res. Dev.
2008, 12, 411.
(22) Billingsley, K. L.; Anderson, K. W.; Buchwald, S. L. Angew.
Chem., Int. Ed. 2006, 45, 3484.
(23) More than one equivalent of HBF4 is required for protonation
of the pyridine which otherwise acts as a poison for the reaction. This
resulted in partial deprotection of the piperidine, which is also a catalyst
poison.
(24) For reviews on asymmetric hydrogenations, see: (a) Takaya, T.;
Ohta, T.; Noyori, R. In Catalytic Asymmetric Synthesis; Ojima, I., Ed.;
VCH Publishers: New York, 1993, p 1. (b) Cui, X.; Burgess, K. Chem.
Rev. 2005, 105, 3272. (c) Jakel, C.; Paciello, R. Chem. Rev. 2006, 106,
2912.
(25) <1 mol% catalyst loadings led to variable results: some reactions
afforded 100% conversion while others afforded <5% conversion. We
opted to use 1 mol% catalyst in order to guarantee a robust process.
(26) (a) Humphrey, G. R.; Miller, R. A.; Maligres, P. E.; Weissman,
S. PCT Int. Appl. 68; WO/2009/088729 A1 20090716 , 2009; CAN
151:14835; AN 2009:859685 (b) Humphrey, G. R.; Pye, P. J.; Zhong,
Y.-L.; Angelaud, R.; Askin, D.; Belyk, K. M.; Maligres, P. E.; Mancheno,
1148
dx.doi.org/10.1021/op2001063 |Org. Process Res. Dev. 2011, 15, 1138–1148