Development and Scale-Up of an Asymmetric Synthesis Process for Alogliptin

Article abstract of DOI:10.1021/acs.oprd.0c00544

Alogliptin (1) benzoate is a potent, highly selective inhibitor of serine protease dipeptidyl-peptidase IV, approved by US FDA for the treatment of type 2 diabetes. Herein, we report a more cost-effective process that includes ruthenium-catalyzed asymmetric hydrogenation followed by Hofmann rearrangement of 2-((6-chloro-3-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-yl)methyl)benzonitrile (10) to introduce a chiral amino moiety at a late stage. Use of an inexpensive and readily available nicotinamide (6) for a chiral aminopiperidine core and iodobenzene diacetate (PIDA) under mild and specific conditions allowed us to access 1 with excellent total yield and comparable quality to that manufactured by the original process.

Full text of DOI:10.1021/acs.oprd.0c00544

Made available through a Creative Commons CC-BY-NC-ND License
pubs.acs.org/OPRD
Article
Development and Scale-Up of an Asymmetric Synthesis Process for
Alogliptin
Masatoshi Yamada,* Sayuri Hirano, Ryoji Tsuruoka, Masahiro Takasuga, Kenichi Uno,
Kotaro Yamaguchi, and Mitsuhisa Yamano
Cite This: Org. Process Res. Dev. 2021, 25, 327−336
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Alogliptin (1) benzoate is a potent, highly selective inhibitor of serine protease dipeptidyl-peptidase IV, approved by
US FDA for the treatment of type 2 diabetes. Herein, we report a more cost-effective process that includes ruthenium-catalyzed
asymmetric hydrogenation followed by Hofmann rearrangement of 2-((6-chloro-3-methyl-2,4-dioxo-3,4-dihydropyrimidin-1(2H)-
yl)methyl)benzonitrile (10) to introduce a chiral amino moiety at a late stage. Use of an inexpensive and readily available
nicotinamide (6) for a chiral aminopiperidine core and iodobenzene diacetate (PIDA) under mild and specific conditions allowed us
to access 1 with excellent total yield and comparable quality to that manufactured by the original process.
KEYWORDS: asymmetric hydrogenation, enamide, Hofmann rearrangement, PIDA, alogliptin, trelagliptin
luoylnitrile (3) to give the key intermediate 4. Then, (R)-3-
aminopiperidine (5) is introduced to 4 by nucleophilic
substitution, followed by salt formation with benzoic acid to
afford alogliptin (Scheme 2).2
INTRODUCTION
■
Takeda Pharmaceutical Company subsidiary Takeda San Diego
(formerly Syrrx) disclosed 2-{6-[3(R)-amino-piperidin-1-yl]-3-
methyl-2,4-dioxo-3,4-dihydro-2H-pyrimidin-1-ylmethyl}-ben-
zonitrile (alogliptin, 1) as a dipeptidyl peptidase IV (DPP-4)
inhibitor and an agent for the treatment of type 2 diabetes
mellitus (Scheme 1).1−3 Alogliptin inhibits DPP-4, which
With growing demand around the world, it was considered
that a more efficient and cost-effective process for alogliptin
would be beneficial. Among the starting materials of the original
process, the optically active amine 5, a relatively expensive
material, has mainly been synthesized via inefficient routes
employing optical resolution, with yields of 50% at most.6 We
envisioned that an alternative process, where the chiral
stereocenter is constructed via an asymmetric method, would
improve the total yield and streamline the process. The atom
economical enantioselective approach would also enable
reduction of the raw material cost of alogliptin. Herein, we
report a new efficient asymmetric process for the synthesis of
alogliptin. A key transformation is asymmetric hydrogenation of
1,4,5,6-tetrahydropyridine-3-carboxamide (7), which is synthe-
sized in a single step from the readily available and low-cost
nicotinamide (6). We found that ruthenium catalysis in the
presence of strong acids was highly effective for the trans-
formation. After condensation with a pyrimidine-2,4-dione core,
an amide moiety was transformed into an amine without erosion
of enantioselectivity via Hofmann rearrangement, using
iodobenzene diacetate (PIDA) as an oxidizing agent.7
Scheme 1. Chemical Structure of Alogliptin (1)
normally degrades the incretins glucose-dependent insulino-
tropic polypeptide (GIP) and glucagon-like peptide 1 (GLP-1).
The inhibition of DPP-4 increases the amount of active plasma
incretins, which helps with glycemic control. GIP and GLP-1
stimulate glucose-dependent secretion of insulin in pancreatic β
cells. GLP-1 has the additional effects of suppressing glucose-
dependent glucagon secretion, inducing satiety, reducing food
intake, and reducing gastric emptying.4 Alogliptin or a fixed dose
combination with the antidiabetic agent metformin hydro-
chloride was approved and launched first in Japan in 2010 and is
now commercially available in many countries, including
Australia, China, Europe, Japan, Mexico, South Korea, and the
United States, under varying brand names.5
Received: December 19, 2020
Published: February 10, 2021
The original process for the synthesis of alogliptin starts with
condensation of 6-chloro-3-methyluracil (2) and α-bromoto-
© 2021 The Authors. Published by
American Chemical Society
https://dx.doi.org/10.1021/acs.oprd.0c00544
Org. Process Res. Dev. 2021, 25, 327−336
327

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Scheme 2. Original Synthetic Process of Alogliptin (1) Benzoate
Scheme 3. Partial Reduction of 6
a
Table 1. Asymmetric Hydrogenation of 7
b
c
d
e
entry
metal precursor
ligand
s/c
additive
conversion (%)
% ee
1
2
3
[RhCl(cod)]2
[IrCl(cod)]2
Pd(OAc)2
(R)-BINAP
(R)-BINAP
(R)-BINAP
20
20
20
74
67
0
racemic
racemic
4
5
6
7
8
9
10
11
12
13
RuCl2{(R)-binap}
RuCl2{(R)-binap}
RuCl2{(R)-binap}
RuCl2{(R)-binap}
RuCl2{(R)-binap}
RuCl2{(R)-binap}(dmf)n
RuCl2{(S)-binap}(dmf)n
RuCl2{(R,S)-mandyphos}(dmf)n
RuCl2{(R,S)-xylyl-mandyphos}(dmf)n
RuCl2{(R)-phanephos}(dmf)n
20
20
20
20
39
84
61
100
100
100
100
100
100
100
−59
−15
−12
−70
−70
−68
69
−69
−81
−86
b
CH3CO2H
PhCO2H
PhSO3H
p-TsOH·H2O
p-TsOH·H2O
p-TsOH·H2O
p-TsOH·H2O
p-TsOH·H2O
p-TsOH·H2O
20
100
100
100
100
100
a
Unless otherwise stated, reactions were conducted under 1 MPa of H2 for 20 h at 50 °C using 7 in MeOH containing a metal complex. Substrate-
c
d
to-catalyst molar ratio. 1 equivalent of 7. Determined by chiral high-performance liquid chromatography (HPLC) analysis, with the conditions:
SHISEIDO CD-Ph column (5 μm, 250 × 4.6 mm), 0.1 M aq KPF6/MeCN 95/5, flow rate: 0.5 mL/min, oven temperature: 25 °C, detection: 200
e
nm (UV), tR: 13.7 min (7), tR: 15.2 min (ent-8), and tR: 17.0 min (8). Assayed by chiral HPLC analysis, with the conditions: CHIRALPAK IC
column (5 μm, 250 × 4.6 mm), MeCN/0.02 M aq H3PO4 3/7, flow rate: 0.5 mL/min, oven temperature: 25 °C, detection: 200 nm (UV), The
sample was derived to benzoylated form, tR: 13.1 min (N-benzoylated ent-8), and tR: 17.2 min (N-benzoylated 8).
in the reaction mixture were found to be important to obtain 7 in
acceptable yield with constant reproducibility.
RESULTS AND DISCUSSION
■
Partial Reduction of Nicotinamide (6). The classical
Transformation of 6 to 7 was carried out with Pd/C type-K,
purchased from N.E. CHEMCAT Corporation, in MeOH under
0.1 MPa of hydrogen pressure at 40 °C for 21 h. Although the
reaction gave a mixture in the ratio of 6/7/8 = 1/77/22, 7 was
successfully isolated in 70% yield after crystallization from
MeOH. Enamide 7 was then converted to its mono p-
method for the partial reduction of 6 into 7 with Pd/C as a
heterogeneous hydrogenation catalyst has been reported to
proceed in 70−92% yield.8 Optimization of the partial reduction
was investigated, to suppress over-reduction to racemic
nipecotamide (8). The stirring speed and hydrogen pressure
328
https://dx.doi.org/10.1021/acs.oprd.0c00544
Org. Process Res. Dev. 2021, 25, 327−336

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Scheme 4. Asymmetric Hydrogenation of 7·TsOH with Ru(CF3CO2)2{(S)-Phanephos} at s/c 1000
Scheme 5. Large-Scale Asymmetric Hydrogenation of 7·TsOH with Ru(CF3CO2)2{(S)-Binap} at s/c 1000
toluenesulfonate as a 0.85 ethanol solvate, with 96% yield from
EtOH, for the next asymmetric hydrogenation step (Scheme 3).
On a 70 kg scale of 6, 7·TsOH was obtained in 60% yield in two
steps and used directly in the next step as a substrate for
asymmetric hydrogenation.
Asymmetric Hydrogenation of 1,4,5,6-Tetrahydropyr-
idine-3-carboxamide (7). We first designed a small screen to
assess the feasibility of asymmetric hydrogenation of enamide 7
to hydrogenation product 8. To our knowledge, transition
In 1998, Pye et al. reported a practical and reproducible
procedure using the readily prepared [2.2]-PHANEPHOS-
Ru(II)bistrifluoroacetate salt, Ru(CF3CO2)2{(S)-phanephos},
in the presence of tetrabutylammonium iodide (TBAI), that
allowed the asymmetric hydrogenation of β-ketoesters up to
96% ee.11 Moreover, they mentioned that Ru(CF3CO2)2{(S)-
phanephos}could be stored up to several weeks in an argon
atmosphere without noticeable deterioration from a practical
standpoint. We prepared Ru(CF3CO2)2{(S)-phanephos} ac-
cording to the literature11 and conducted an optimization using
the catalyst for a practical manufacturing method. We found that
not only TBAI but also other inexpensive inorganic salts
dramatically increased the catalytic activity. After optimization,
the process for asymmetric hydrogenation of 7 was finalized as
shown in Scheme 4, using 7·TsOH as the substrate for the
hydrogenation reaction. Addition of KBr (10 equiv to Ru) as a
halide source was effective to enhance the reaction activity even
at s/c 1000. Fortunately, we also found that crystallization from
ethanol, after concentration of the asymmetric hydrogenation
reaction mixture, gave 8 as a salt with p-toluenesulfonic acid in
almost perfect enantioselectivity (99.7% ee) and 76% yield.
Thus, we have developed an asymmetric process that is cost-
effective, facile, and robust for scale-up manufacturing.
The asymmetric hydrogenation of 7·TsOH has been scaled
up without any issues. After further optimization, the reaction
solvent was switched from methanol to 2-propanol for substrate
stability, and the chiral catalyst was also changed from
Ru(CF3CO2)2{(S)-phanephos} to Ru(CF3CO2)2{(S)-binap},
bearing the inexpensive and widely used (S)-binap ligand, in
thorough consideration of the impact on manufacturing cost.
The designated asymmetric hydrogenation of 7·TsOH (50.0
kg), under the conditions shown in Scheme 5, gave crude 8·
TsOH in full conversion with 70.1% ee. Subsequent
recrystallization from ethanol finally provided 25.5 kg of 8·
metal-catalyzed asymmetric hydrogenation of 7-like non-
protected tetrahydropyridines has not been reported, although
some catalytic examples for 3-carboxy-protected 1,4,5,6-
tetrahydropyridines were recorded with good enantioselectiv-
ity.9 The screen commenced with a relatively small set of
commercially available chiral transition metal catalysts (Table
1). The combination with (R)-BINAP and [RhCl(cod)]2 or
[IrCl(cod)]2 resulted in moderate conversion with no
enantioselectivity (Table 1, entries 1 and 2). The combination
with (R)-BINAP and Pd(OAc)2 provided a negligible amount of
8 (entry 3). In the case of RuCl2{(R)-binap}, 8 was generated in
39% conversion with moderate enantioselectivity (−59% ee,
entry 4). In the asymmetric hydrogenation of 7, addition of
Bronsted acid was found to be very effective to improve the
reaction conversion (entries 5 and 6). Among the Bronsted acids
tested, sulfonic acids such as benzenesulfonic acid and p-
toluenesulfonic acid monohydrate achieved full conversion of 7
to provide 8 with −70% ee (entries 7 and 8). We selected p-
toluenesulfonic acid monohydrate as an acid additive and
subsequently evaluated the performance of various chiral
diphosphine ligands, as depicted in entries 9−13. The
corresponding chiral ruthenium catalyst was prepared as a
dimethylformamide (DMF) complex according to the classical
method.10 RuCl2{(R)-binap}(dmf)n gave the same result as that
in entry 8 at s/c (substrate-to-catalyst molar ratio) 100 (entry 9).
When RuCl2{(S)-binap}(dmf)n was employed, the enantiose-
lectivity was completely reversed, and the desired configuration
of 8 was obtained with 69% ee (entry 10). In the screening of 44
types of chiral diphosphine ligands, RuCl2{(R,S)-mandyphos}-
(dmf)n showed the same performance as (R)-BINAP (entry 11).
The use of (R,S)-Xylyl-MANDYPHOS and (R)-PHANEPHOS
showed the best results, giving 8 in 81% ee and 86% ee,
respectively (entries 12 and 13).
329
https://dx.doi.org/10.1021/acs.oprd.0c00544
Org. Process Res. Dev. 2021, 25, 327−336

Organic Process Research & Development
pubs.acs.org/OPRD
Article
TsOH in 68% yield with 99.6% ee. The residual metal contents
of Pd and Ru were <0.081 ppm and 3.6 ppm, respectively.
2-((6-Chloro-3-methyl-2,4-dioxo-3,4-dihydropyrimi-
din-1(2H)-yl)methyl)benzonitrile (10) Synthesis. Dis-
placement of the chlorine on the 1,3-disubstituted chlorouracil
9 with 8·TsOH was accomplished in the presence of potassium
carbonate in aqueous 2-propanol at 65 °C for 16 h. To promote
crystallization of 2-((6-chloro-3-methyl-2,4-dioxo-3,4-dihydro-
pyrimidin-1(2H)-yl)methyl)benzonitrile (10), additional water
was added to the reaction mixture. On a 12.0 kg scale of 9, the
manufacturing operation was implemented to give 15.3 kg of 10
without any scale-up issues (Scheme 6).
yield. We first studied the effect of the PIDA equivalence to 10 in
MeCN/H2O (1/1) at room temperature (Table 2). Increasing
the amount of PIDA was effective to improve the reaction yield
and to prevent the formation of 11 at room temperature. In
order to reduce the amount of PIDA used as much as possible,
for cost reduction, optimization of the equivalence of PIDA was
required. Judging from the HPLC area % in Table 2, at least 1.3
equiv of PIDA was needed to complete the reaction (entry 4).
Dimer formation was clearly suppressed by the excess amount of
PIDA (entry 6).
The decomposition (hydration) of PIDA was considered as
one of the major reasons for the necessity of excess PIDA, so the
effect of additives was examined (Table 3). As a side note,
controlling of urea (11) transformation under a small amount of
PIDA was difficult, although we tried the various adding method
of PIDA.13a The addition of inorganic bases showed that K2CO3
enhanced consumption of 10, but a substantial amount of 11
was formed. Cooling the reaction to 0 °C had no effect to reduce
11 (entry 3). Although the addition of organic amines like Et3N
increased 11 formation, the addition of pyridine was effective for
reaction completion, and the formation of 11 also slightly
decreased (entries 6 and 7). It was interesting that PIDA was
solubilized in the presence of pyridine, whereas it was hardly
soluble in all the other solvent systems. Although the
solubilization mechanism is still unclear, we considered that
increasing the initial concentration of PIDA could facilitate the
reaction to proceed. We suspected that the solubilizing effect
would be obtained by a catalytic amount of pyridine, and indeed,
no significant difference between the pyridine amounts was
observed (entries 8−10). The reaction with a catalytic amount
of pyridine also enabled us to change the solvent from MeCN to
2-propanol, a much more inexpensive and environmentally
benign solvent (entry 11), while the reaction without pyridine
did not go to completion, with 22% of 10 remaining, due to
faster PIDA decomposition. Less 11 formation was observed in
2-propanol/H2O than in MeCN/H2O, with a little longer
reaction time.
Scheme 6. Synthesis of 10
Hofmann Rearrangement of 10 and Salt Formation to
1. A variety of oxidizing agents have been known to promote
Hofmann rearrangement, including NaOCl, NaOBr, and N-
bromosuccinimide as principal examples.6b,12 Recently, PIDA or
iodobenzene bistrifluoroacetate as effective oxidizing agents
have provided good reaction selectivity and high practicality.13
We initially attempted using aq NaOCl, an inexpensive reagent,
for Hofmann rearrangement of 10. However, the resulting
reaction mixture was complex, with side products such as the
nonhydrolyzable symmetrical urea (11) being generated in the
presence of diisopropylethylamine in MeCN/H2O. The isolated
yield of 1·BzOH was 50% based on 10 with 97% assay, and these
side products were difficult to remove by phase separation or
crystallization. Next, we investigated the Hofmann rearrange-
ment reaction condition using PIDA, to improve the reaction
a
Table 2. Effect of the PIDA Amount in Hofmann Rearrangement of 10
bc
,
HPLC (area %)
1
entry
PIDA (equiv)
10
11
1
2
3
4
5
6
1.0
1.1
1.2
1.3
1.5
2.0
5.5
2.0
0.5
0.1
0.2
0.1
84.8
89.4
92.3
94.5
95.1
96.1
6.6
5.7
5.2
3.6
2.4
1.1
a
Unless otherwise stated, to a four-necked round bottomed flask, MeCN/H2O (1/1) and 10 were charged. The solution was warmed to 50 °C for
b
completely dissolving 10 and then cooled to 25 °C. PIDA was added to the solution, and then, the suspension was stirred for 1 h at 25 °C. Any
area contribution of iodosobenzene was ignored. Determined by HPLC analysis, with condition: SHISEIDO MG II column (5 μm, 250 × 4.6
c
mm), 0.02 M aq H3PO4/MeCN = 6/4, flow rate: 1.0 mL/min, oven temperature: 25 °C, detection: 278 nm (UV), tR: 2.1 min (1), tR: 2.5 min
(10), and tR: 6.0 min (11).
330
https://dx.doi.org/10.1021/acs.oprd.0c00544
Org. Process Res. Dev. 2021, 25, 327−336

Organic Process Research & Development
pubs.acs.org/OPRD
Article
a
Table 3. Effect of Bases for Hofmann Rearrangement of 10 by PIDA
b c
,
HPLC (area %)
entry
base
temp. (°C)
time (h)
10
1
11
1
2
3
4
5
6
7
8
9
10
Rt
Rt
0
Rt
0 to rt
Rt
Rt
Rt
Rt
1
1
24
1
4
1
1
1
1
1
4.8
0.4
0.6
7.4
9.5
2.9
1.0
0.2
0.1
0.1
0.3
88.5
65.4
53.7
70.7
30.3
52.7
92.9
95.1
95.6
95.5
94.7
4.5
26.9
38.6
17.2
48.7
36.1
3.6
3.0
3.1
3.2
3.5
K2CO3 (3.0)
K2CO3 (3.0)
NaHCO3 (2.0)
NaOH (2.0)
Et3N (2.0)
pyridine (2.0)
pyridine (1.0)
pyridine (0.05)
pyridine (0.025)
pyridine (0.025)
Rt
Rt
d
11
2
a
Unless otherwise stated, to a four-necked round bottomed flask, MeCN/H2O (1/1), base, and 10 were charged. PIDA was added to the solution,
b
c
and then, the suspension was stirred. Any area contribution of iodosobenzene was ignored. Determined by HPLC analysis, with condition:
SHISEIDO MG II column (5 μm, 250 × 4.6 mm), 0.02 M aq H3PO4/MeCN = 6/4, flow rate: 1.0 mL/min, oven temperature: 25 °C, detection:
d
278 nm (UV), tR: 2.1 min (1), tR: 2.5 min (10), and tR: 6.0 min (11). 2-propanol/H2O = 1/1 was used as a reaction solvent.
a
Table 4. Hofmann Rearrangement and Benzoic Acid Salt Formation (50 g Scale)
b c
,
isolated yield (%)
1·BzOH
HPLC (area %) of 1·BzOH
residual metals (ppm)
Pd
run
1
1
11
Ru
1
2
3
89
87
91
93
92
90
99.79
99.88
99.80
0.07
0.03
0.04
0.026
<0.025
<0.026
3.1
2.0
2.2
a
b
c
The operation procedure is described in the experimental section. Any area contribution from benzoic acid was ignored. Determined by HPLC
analysis, with condition: SHISEIDO MG II column (5 μm, 250 × 4.6 mm), 0.02 M aq H3PO4/MeCN = 6/4, flow rate: 1.0 mL/min, oven
temperature: 25 °C, detection: 278 nm (UV), tR: 2.1 min (1), tR: 2.5 min (10), and tR: 6.0 min (11).The optical yield of 1·BzOH was 99.5% by
chiral HPLC analysis, with condition: SUMICHIRAL OA-4600R column (5 μm, 250 × 4.6 mm), hexane/2-propano; /methanol/trifluoroacetic
acid = 430/45/25/1, flow rate: 1.0 mL/min, oven temperature: 35 °C, detection: 275 nm (UV), tR: 42.7 min (enantiomer), and tR: 45.4 min (1).
Isolation of 1 and Salt Formation to 1·BzOH. Although 1
had been isolated as a hydrochloride in the previous method, the
salt exchange process from hydrochloride to benzoate before the
final step was not operationally friendly. Here, 1 obtained via
Hofmann rearrangement showed an improved HPLC profile
and was able to be isolated as a crystalline-free base in around
90% yield from toluene/heptane. While the urea 11 was difficult
to be removed by crystallization, due to its low solubility, 11 was
found to be removable by back extraction of 1 in an acidic
aqueous phase during work-up (for further details, see
Experimental Section). After the work-up operation, 1 could
be isolated in around 90% yield from toluene/heptane (10/15),
which was adopted as the best crystallization solvent among the
solvents screened. Additionally, we investigated an alternative
solvent instead of iPrOAc, which was used as the extraction and
crystallization solvent in the original route because of its
significant effect on the manufacturing cost. An examination of
1·BzOH salt formation using 2-propanol/EtOAc (1/1) showed
that it gave high quality 1·BzOH in good yield, which satisfied
the established specification.
Next, the optimized reaction conditions from 10 to 1·BzOH
were applied to a 50 g scale synthesis, which was performed
three times (Table 4).
The three trials showed good reproducibility, and 1 and 1·
BzOH were obtained in 89 2 and 92 2%, respectively. As
contamination by 11 in the first run was close to 0.1 area %, the
amount of EtOAc for back extraction to remove 11 was
increased from 5 to 10 vol. Hence, in the second and third runs,
contamination by 11 could be reduced to less than 0.04 area %.
331
https://dx.doi.org/10.1021/acs.oprd.0c00544
Org. Process Res. Dev. 2021, 25, 327−336

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Scheme 7. Synthesis of Trelagliptin (12) Succinate
Scheme 8. Overall Scheme of the Developed Synthetic Route to 1·BzOH
Thus, we could confirm the reproducibility of the economical
reaction system on a 50 g scale, without any quality issues.
Application to Trelagliptin Succinate. Trelagliptin (12)
succinate, a novel once-weekly oral DPP-4 inhibitor, was
approved for the Japanese market in 2015 from Takeda
Pharmaceutical Company.1,14 As 12 has a similar molecular
structure to 1 with only a single aromatic hydrogen atom being
replaced by a fluorine, we investigated whether the developed
method could be applied to the synthesis of 12 (Scheme 7).
The same conditions as for the synthesis of 1 were applied to
the synthesis of 12, and for the final succinate formation, the
current conditions giving 12 were traced as well. Each reaction
proceeded in good yield and gave the desired product in 70%
yield in three steps from 13.
CONCLUSIONS
■
In summary, the application of catalytic asymmetric hydro-
genation to a key intermediate in the synthesis of alogliptin (1)
resulted in the discovery of an enantioselective synthesis
332
https://dx.doi.org/10.1021/acs.oprd.0c00544
Org. Process Res. Dev. 2021, 25, 327−336

Organic Process Research & Development
pubs.acs.org/OPRD
Article
pathway, as the key step (Scheme 8). The asymmetric
hydrogenation of 1,4,5,6-tetrahydropyridine-3-carboxamide
(7) and crystallization as the p-toluenesulfonate from ethanol
gave (R)-nipecotamide (8) in >99% ee. With Ru-
(CF3CO2)2{(S)-phanephos}, it was possible to achieve an
enantioselectivity of 88% ee in the reduction of 7·TsOH to 8 in
the presence of KBr at a catalyst loading of 1000:1 on a 66.7 g
laboratory scale of 7·TsOH, and then, 99.7% ee of 8·TsOH was
obtained by crystallization from ethanol, with 76% yield. The
new procedure could be applied to a 50.0 kg scale 7·TsOH
manufactured using Ru(CF3CO2)2{(S)-binap}, and 25.5 kg of
8·TsOH was prepared in 68% yield.
Hofmann rearrangement of 2-((6-chloro-3-methyl-2,4-dioxo-
3,4-dihydropyrimidin-1(2H)-yl)methyl)benzonitrile (10), pre-
pared by the coupling of 8·TsOH and the 1,3-disubstituted
chlorouracil 9, was successful using PIDA in the presence of a
catalytic amount of pyridine. These new synthetic routes
afforded 1 as crystals of the free base, and finally, 1 was
converted to the benzoate salt in a 2-propanol/EtOAc system.
Moreover, the synthetic strategy was applicable for the
preparation of trelagliptin (12), which has a similar molecular
structure to 1.
96% yield, 100 area % HPLC purity). IR (KBr): 3379, 3296,
2978, 2949, 1665, 1609, 1437, 1410, 1177, 1123, and 1101 cm−1.
The product was detected as two isomers in the ratio of 87
1
(major):13 (minor) by NMR measurements. H NMR (500
MHz, D2O): δ 7.62 (d, J = 8.5 Hz, 2H), 7.29 (d, J = 8.5 Hz, 2H),
5.09 (s, 0.13H), 4.79 (s, 0.87H), 3.57 (q, J = 6.5 Hz, 0.85 × 2H),
3.24−3.40 (m, 1H), 2.91−3.11 (m, 1H), 2.31 (s, 3H), 1.76−
2.08 (m, 2H), 1.53−1.71 (m, 2H), 1.10 (t, J = 6.5 Hz, 0.85 ×
3H). 13C NMR (126 MHz, D2O): δ 176.1 (0.13C), 175.9
(0.87C), 142.5 (1C), 139.9 (1C), 129.6 (1C), 125.6 (1C), 80.0
(0.87C), 78.1 (0.13C), 57.6 (0.87C), 47.5 (0.13C), 47.4
(0.13C), 47.2 (0.13C), 43.2 (1C), 40.7 (0.13C), 25.6
(0.87C), 22.0 (0.13C), 20.7 (1C), 20.5 (0.87C), 19.1
(0.13C), 17.0 (0.87C). HRMS (m/z, ESI): calcd 127.0866 for
C6H10N2O (M + H)+; found, 127.0873; calcd 171.0121 for
C7H8O3S (M − H)−; found, 171.0124. HPLC: SHISEIDO CD-
Ph column (5 μm, 250 × 4.6 mm), MeCN/aq KPF6 (0.10 M) 5/
95, flow rate: 0.5 mL/min, oven temperature: 25 °C, detection:
200 nm (UV), tR: 8.5 min (p-toluenesulfonic acid), tR: 12.0 min
(7), tR: 19.0 min (6).
Lab-Scale Synthesis of (R)-Nipecotamide (8) p-
Toluenesulfonate. To a 1 L autoclave were added Ru-
(CF3CO2)2{(S)-phanephos} (0.1785 g, 0.198 mmol, 0.001
equiv = s/c 1,000), potassium bromide (0.2350 g, 1.98 mmol,
0.01 equiv), and 7·TsOH 0.85 ethanol solvate (66.67 g, 197.5
mmol, 1.00 equiv). The atmosphere was evacuated and filled
with argon gas seven times. Dehydrated methanol (500 mL) was
added to the autoclave by the argon feed cannula. The mixture
was stirred at room temperature for 5 min, and hydrogen was
initially introduced into the autoclave at a pressure of 0.1 MPa,
before being reduced to atmospheric pressure by carefully
releasing the stop valve. After this procedure was repeated ten
times, the hydrogen pressure was introduced at 1.0 MPa, and the
solution was stirred at 50 °C for 15 h. The solution was cooled to
room temperature, and hydrogen gas was then carefully vented.
After evaporation of the solvent, the residue was evaporated to
afford a pale-green solid. The residue was dissolved with ethanol
(201 mL) at 75 °C, and the insolubles were removed by
filtration. The filtrates were evaporated to dryness. The residue
was diluted with ethanol (201 mL) and stirred at 80 °C for 1 h.
The seed crystal of 8·TsOH was inoculated to the solution. The
resulting suspension was cooled to room temperature and aged
for 3 h. The solid was collected by filtration, and the cake was
washed with ethanol (142 mL) before being dried at 60 °C
under vacuum to yield 8·TsOH as a colorless solid (44.99 g, 76%
yield, 99.7% ee as benzylated 8). [α]2D50.3° (c = 0.996,
methanol); IR (KBr): 3400−3300, 3159, 2950−2800, 1670,
1435, and 1171 cm−1. 1H NMR (500 MHz, D2O): δ 7.73 (d, J =
8.5 Hz, 2H), 7.41 (d, J = 8.5 Hz, 2H), 3.37−3.45 (m, 1H), 3.28−
3.36 (m, 1H), 3.16−3.24 (m, 1H), 3.05−3.14 (m, 1H), 2.82−
2.92 (m, 1H), 2.43 (s, 3H), 2.82−2.92 (m, 1H), 2.43 (s, 3H),
2.03−2.14 (m, 1H), 1.92−2.03 (m, 1H), 1.74−1.89 (m, 2H).
13C NMR (126 MHz, D2O): δ 177.4, 142.6, 139.9, 129.6, 125.6,
44.9, 44.1, 38.5, 25.6, 20.7, 20.6. Anal. Calcd for C13H20N2O4S:
C, 51.98; H, 6.71; N, 9.33; S, 10.68. Found: C, 52.19; H, 6.79; N,
9.31; S, 10.57. HRMS (m/z, ESI): calcd 129.1022 for
C6H12N2O (M + H)+; found, 129.1046; calcd 171.0121 for
C7H8O3S (M − H)−; found, 171.0130. HPLC: CHIRALPAK
IC column (5 μm, 250 × 4.6 mm), MeCN/aq H3PO4 (0.020 M)
3/7, flow rate: 0.5 mL/min, oven temperature: 25 °C, detection:
200 nm (UV). The sample was converted to the benzylated
form; tR: 13.1 min (N-benzylated ent-8), tR: 17.2 min (N-
benzylated 8).
EXPERIMENTAL SECTION
■
General Remarks. All reagents and solvents were purchased
commercially and used without further purification. HPLC
analysis was performed on HITACHI 7400 instruments or
SHIMADZU LC-10ADvp instruments. 1H NMR and 13C NMR
spectra were measured in CDCl3, CD3OD, or D2O solutions at
500 and 126 MHz, respectively, using a Bruker Avance-III
spectrometer for analysis of 7·TsOH, 8·TsOH, 10, 1·BzOH, 14,
and 12·HOOC(CH2)2COOH. The IR spectrum was measured
using a SHIMADZU IR Prestige-21. High-resolution mass
spectrometry (HRMS) spectra were measured using a
SHIMADZU LCMS-IT-TOF, and melting points were
measured using a SIBATA B-545 apparatus. Specific rotations
were measured using a JASCO P-1030.
Lab-Scale Synthesis of 1,4,5,6-Tetrahydropyridine-3-
carboxamide (7) p-Toluenesulfonate. To a 1 L autoclave
were added nicotinamide (50.00 g, 0.41 mol, 1.0 equiv) Pd/C
type-K (5.00 g, 50%-wet) and methanol (500 mL). The
atmosphere was purged with nitrogen gas seven times.
Hydrogen was initially introduced into the autoclave at a
pressure of 0.02 MPa, before being reduced to atmospheric
pressure by carefully releasing the stop valve. After this
procedure was repeated ten times, the hydrogen pressure was
introduced at 0.1 MPa, and the solution was stirred at 40 °C for
21 h under 0.1 MPa of hydrogen pressure. The solution was
cooled to room temperature, and hydrogen gas was then
carefully vented. After removing Pd/C by filtration, the filtrates
were evaporated to produce a white crystalline powder. The
residue was diluted with methanol (100 mL), and the resulting
suspension was aged at room temperature for 1 h. The solid was
collected by filtration, and the cake was washed with methanol
(50 mL) before being dried at 50 °C under vacuum to yield 7 as
a colorless solid (36.20 g, 70% yield). To a four-necked round
bottomed flask were charged the obtained powder (36.20 g), p-
toluenesulfonic acid monohydrate (54.63 g, 0.29 mol, 1.00
equiv), and ethanol (290 mL, 8 vol). The suspension was aged at
room temperature overnight and collected by filtration, and
then, the cake was washed with a 1/4 v/v ethyl acetate-
diisopropylether mixture and dried at 60 °C under vacuum to
yield 7·TsOH 0.85 ethanol solvate as a colorless solid (92.84 g,
333
https://dx.doi.org/10.1021/acs.oprd.0c00544
Org. Process Res. Dev. 2021, 25, 327−336

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Lab-Scale Synthesis of 2-((6-Chloro-3-methyl-2,4-
dioxo-3,4-dihydropyrimidin-1(2H)-yl)methyl)-
benzonitrile (10). To a 1 L four-necked round bottomed flask
were charged 2-((6-chloro-3-methyl-2,4-dioxo-3,4-dihydropyr-
imidin-1(2H)-yl)methyl)benzonitrile (9) (27.54 g, 99.90 mmol,
1.0 equiv), 8·TsOH (30.00 g, 99.98 mmol, 1.0 equiv), potassium
carbonate (27.61 g, 199.8 mmol, 2.00 equiv), 2-PrOH (210
mL), and water (3 mL). This suspension was heated to 70 °C
and stirred for 18 h. The reaction mixture was cooled to 25 °C
and evaporated under reduced pressure. The resulting
suspension was stirred at room temperature for 2 h. The solid
was collected by filtration, and the cake was washed with water
(90 mL) and dried at 60 °C under vacuum to yield 10 as a
colorless solid (34.13 g, 93% yield, 95.9 area % HPLC purity).
[α]2D5 + 21.4° (c = 0.977, methanol); IR (KBr): 3387, 3319,
3202, 2941, 2853, 2226, 1690, 1676, and 1628 cm−1. 1H NMR
(500 MHz, CDCl3): δ 7.61 (dd, J = 7.5, 1.5 Hz, 1H), 7.51 (td, J =
8.0, 1.5 Hz, 1H), 7.32 (t, J = 7.5 Hz, 1H), 7.11 (d, J = 8.0 Hz,
1H), 5.73 (bs, 1H), 5.67 (bs, 1H), 5.34 (s, 1H), 5.32 (d, J = 16.0
Hz, 1H), 5.13 (d, J = 16.0 Hz, 1H), 3.25 (s, 3H), 3.12−3.20 (dm,
1H), 2.88 (d, J = 11.5 Hz, 1H), 2.72 (t, J = 11.5 Hz, 1H), 2.55 (t,
J = 11.5 Hz, 1H), 2.47 (tt, J = 11.5, 3.5 Hz, 1H), 1.91 (dd, J =
13.0, 3.5 Hz, 1H), 1.70 (dt, J = 13.5, 3.5 Hz, 1H), 1.57 (qd, J =
13.0, 3.5 Hz, 1H), 1.46 (qt, J = 13.0, 4.0 Hz, 1H). 13C NMR (126
MHz, CDCl3): δ 174.8, 163.2, 159.8, 153.0, 141.0, 133.6, 133.3,
128.2, 127.2, 117.7, 110.7, 91.3, 54.0, 51.9, 46.4, 42.5, 28.2, 27.3,
24.3. HRMS (m/z, ESI): calcd 368.1717 for C19H21N5O3 (M +
H)+; found, 368.01731. HPLC: YMC-Pack ODS-A302 column
(5 μm, 150 × 4.6 mm), MeCN/aq H3PO4 (0.020 M) 3/7, flow
rate: 0.5 mL/min, oven temperature: 25 °C, detection: 225 nm
(UV), tR: 8.7 min (10), tR: 25.1 min (9).
Lab-Scale Synthesis of Alogliptin (1). To a 2 L four-
necked round bottomed flask, H2O/2-PrOH (1/1, 1.5 L),
pyridine (550 μL, 6.9 mmol, 0.05 equiv), and 10 (50.0 g, 136
mmol, 1.00 equiv) were charged. PIDA (48.2 g, 150 mmol, 1.1
equiv) was added to the solution, and then, the suspension was
stirred for 3 h at 20 °C. The solvents (750 50 mL) were
evaporated in vacuo at 40 °C, and then, EtOAc (500 mL) was
added, and the aqueous layer was separated. The aqueous layer
was washed by EtOAc (500 mL) and treated with K2CO3 (400
g) at 0−15 °C. The organic products were extracted with
toluene (100 mL) and 2-PrOH (150 mL), followed by rinsing
with saturated aq NaCl (50 mL). After evaporation of the
organic solvents, toluene (150 mL) was added and evaporated
for solvent substitution. 1 was crystallized from toluene (100
mL) and heptane (150 mL, dropwise for 1 h) as a yellowish
white crystalline solid (40.3 g, 87.2% yield, 99.31 area % HPLC
purity). HPLC: YMC-Pack ODS-A302 column (5 μm, 150 ×
4.6 mm), MeCN/aq H3PO4 (0.020 M) 3/7, flow rate: 0.5 mL/
min, oven temperature: 25 °C, detection: 225 nm (UV), tR: 3.2
min (1), tR: 8.7 min (10).
7.28−7.42 (m, 4H), 7.22 (d, J = 8.0 Hz, 1H), 5.43 (s, 1H), 5.13
(d, J = 16.0 Hz, 1H), 5.05 (d, J = 16.0 Hz, 1H), 3.25−3.45 (m,
2H), 3.03 (s, 3H), 2.77−2.96 (m, 2H), 2.57−2.77 (m, 1H),
1.95−2.08 (m, 1H), 1.68−1.80 (m, 1H), 1.47−1.63 (m, 2H).
13C NMR (126 MHz, D2O) Peaks of ethanol were detected: δ
175.3, 165.4, 160.4, 152.9, 140.1, 47.1, 136.3, 133.3, 133.7,
131.1, 128.9, 128.3, 128.2, 128.0, 118.0, 109.4, 90.2, 57.5, 52.2,
51.7, 46.9, 27.9, 27.2, 21.2, 16.9. HRMS (m/z, ESI): calcd
340.1768 for C19H21N5O3 (M + H)+; found, 340.1769. HPLC:
SHISEIDO MG II column (5 μm, 150 × 4.6 mm), 0.02 M aq
H3PO4/MeCN = 6/4, flow rate: 1.0 mL/min, oven temperature:
25 °C, detection: 278 nm (UV), tR: 2.1 min (1), tR: 2.5 min (10),
tR: 6.0 min (11). Chiral HPLC: SUMICHIRAL OA-4600R
column (5 μm, 250 × 4.6 mm), hexane/2-propano; /methanol/
trifluoroacetic acid = 430/45/25/1, flow rate: 1.0 mL/min, oven
temperature: 35 °C, detection: 275 nm (UV), tR: 42.7 min
(enantiomer), tR: 45.4 min (1).
Lab-Scale Synthesis of (R)-1-(3-(2-Cyano-5-fluoroben-
zyl)-1-methyl-2,6-dioxo-1,2,3,6-tetrahydropyrimidin-4-
yl)piperidine-3-carboxamide (14). To a 50 mL round
bottomed flask equipped with a reflux condenser were added
H2O/2-PrOH (8/3, 27.5 mL), 8·TsOH (5.0 g, 16.6 mmol, 0.99
equiv), and 2-((6-chloro-3-methyl-2,4-dioxo-3,4-dihydropyri-
midin-1(2H)-yl)methyl)-4-fluorobenzonitrile (13) (4.9 g, 16.8
mmol, 1.00 equiv). K2CO3 (4.6 g, 33.2 mmol, 1.98 equiv) was
added, and then, the reaction was heated to 65 °C and stirred for
24 h. During the reaction, the resulting amide started to
crystallize. After the addition of H2O (30 mL), the reaction
mixture was cooled to 0 °C, stirred for 1 h, and then filtered. The
crystals obtained were washed with H2O (10 mL) and dried in
vacuo at 45 °C to give a reddish white crystalline solid (5.6 g,
1
87% yield, 97.6 area % HPLC purity). H NMR (500 MHz,
CDCl3): δ 7.70 (dd, J = 8.7 Hz, 5.2 Hz, 1H), 7.10 (td, J = 8.0 Hz,
2.5 Hz, 1H), 6.90 (dd, J = 9.1 Hz, 2.5 Hz, 1H), 5.66 (brs, 1H),
5.48 (brs, 1H), 5.42 (s, 1H), 5.38 (d, J = 16.08 Hz, 1H), 5.17 (d,
J = 16.4 Hz,1H), 3.34 (s, 3H), 3.14−3.28 (m, 1H), 2.94 (d, J =
12.0 Hz, 1H), 2.82 (t, J = 10.6 Hz, 1H), 2.64 (t, J = 10.9 Hz,1H),
2.49−2.59 (m, 1H), 1.92−2.07 (m, 1H), 1.80 (m, 1H), 1.62−
1.72 (m, 1H), 1.45−1.60 (m, 1H). HPLC: SHISEIDO MG II
column (5 μm, 150 × 4.6 mm), 0.02 M aq H3PO4/MeCN = 6/4,
flow rate: 1.0 mL/min, oven temperature: 25 °C, detection: 278
nm (UV), tR: 2.7 min (14).
Lab-Scale Synthesis of Trelagliptin (12). To a 100 mL
four-necked round bottomed flask, H2O/2-PrOH (1/1, 60 mL),
pyridine (21.4 μL, 0.26 mmol, 0.05 equiv), and 14 (2.0.0 g, 5.2
mmol) were charged. PIDA (1.84 g, 5.7 mmol, 1.10 equiv) was
added to the solution, and then, the suspension was stirred for 3
h at 20 °C. The solvents (ca. 40 mL) were evaporated in vacuo at
30 °C, and then, EtOAc (20 mL) was added, and the aqueous
layer was separated. The aqueous layer was washed with EtOAc
(20 mL) and treated with K2CO3 (16 g) at 0−15 °C. The
organic products were extracted with toluene (6 mL) and 2-
PrOH (6 mL), followed by rinsing with saturated aq NaCl (10
mL). After evaporation of the organic solvents, toluene (6 mL)
was added and evaporated for solvent substitution. Product 12
was crystallized from toluene (6 mL) and heptane (6 mL,
dropwise for 1 h) at 0 °C as a white crystalline solid (1.6 g, 86%
yield, 99.2 area % HPLC purity). 1H NMR (500 MHz, CDCl3):
δ 7.69 (dd, J = 8.5 Hz, 5.4 Hz, 1H), 7.09 (td, J = 8.0 Hz, 2.5 Hz,
1H), 6.86 (dd, J = 9.0 Hz, 2.4 Hz, 1H), 5.39 (s, 1H), 5.23−5.32
(m, 2H), 3.32 (s, 3H), 2.99−3.05 (m, 1H), 2.87−2.98 (m, 2H),
2.61 (m, 1H), 2.41 (m,1H), 1.95 (dd, J = 12.8 Hz, 3.9 Hz, 1H),
1.72−1.83 (m, 1H), 1.56−1.67 (m, 1H), 1.30 (brs, 2H), 1.23 (d,
Lab-Scale Synthesis of Alogliptin (1) Benzoate. 1 (35 g,
103 mmol, 1.00 equiv) was dissolved in hot IPA (140 mL, 60
°C) then filtered through a membrane filter to a 1 L four-necked
round bottomed flask. Benzoic acid (13.8 g, 113 mmol, 1.1
equiv) in EtOAc (140 mL) was added dropwise to the solution
for 1 h at 60 °C. The crystals were aged for 19 h at room
temperature. The resulting solid material was filtered and dried
in vacuo at 50 °C to give a slightly yellowish crystalline solid
(43.9 g, 92% yield, 99.79 area % HPLC purity). mp 184.0−186.1
°C; IR (KBr): 3082, 2976, 2961, 2860, 2230, 1695, 1609, 1591,
1445 and 1362 cm−1. 1H NMR (500 MHz, D2O): δ 7.76 (d, J =
7.0 Hz, 2H), 7.62 (d, J = 7.5 Hz, 1H), 7.54 (t, J = 7.5 Hz, 1H),
334
https://dx.doi.org/10.1021/acs.oprd.0c00544
Org. Process Res. Dev. 2021, 25, 327−336

Organic Process Research & Development
pubs.acs.org/OPRD
Article
J = 11.0 Hz, 1H). 13C NMR (126 MHz, CDCl3): δ 162.9, 159.5,
152.7, 144.6, 135.5, 135.4, 116.4, 115.8, 115.6, 114.7, 114.6,
90.8, 59.7, 51.9, 46.1, 33.4, 28.0. HPLC: SHISEIDO MG II
column (5 μm, 150 × 4.6 mm), 0.02 M aq H3PO4/MeCN = 6/4,
flow rate: 1.0 mL/min, oven temperature: 25 °C, detection: 278
nm (UV), tR: 1.4 min (12), tR: 2.7 min (14).
potassium bromide (176.20 g, 1.481 mol, 0.01 equiv), and
Ru(CF3CO2)2{(S)-binap} (140.60 g, 1.480 mol, 0.01 equiv, s/c
1,000). The reactor was evacuated to −0.09 MPa and then filled
with nitrogen to 0.10 MPa. This operation was repeated seven
times. To the reactor was added anhydrous 2-propanol (370 L,
7.4 vol) by nitrogen gas feed. The reactor was stirred at 15−21
°C for 2 h. Then, the reactor was pressurized with nitrogen (0.60
MPa), vented three times, pressurized with hydrogen (0.10
MPa), and vented nine times. The hydrogen pressure was set to
0.70 MPa, and heating was started. After a reaction temperature
of 45 °C was reached, the pressure was adjusted to 0.80 MPa.
The reaction mixture was stirred with the hydrogen pressure
maintained within the range of 0.79−0.81 MPa at 46−50 °C (set
point 50 °C) for 20 h. The reactor was cooled below 10 °C, and
the hydrogen was carefully vented. The reactor was pressurized
with nitrogen (0.05 MPa) and vented four times. Then, 2-
propanol (25 L, 0.5 vol) was added, and the mixture was stirred
at 8−9 °C for 3 h. The solid was collected by filtration, and the
cake was washed with 2-propanol (75 L, 1.5 vol) before being
dried at 60 °C under vacuum to yield crude 8·TsOH (39.7 g,
89% yield, 70.1% ee). To a reactor were added crude 8·TsOH
and ethanol (302 L, 8 vol). The suspension was heated to
around 75 °C to dissolve the solid and then stirred at 73 °C for 1
h. The resulting solution was cooled to 60 °C before adding a
seed crystal of 8·TsOH (30 g). The resulting suspension was
cooled to 25 °C over 2 h and aged at 25 °C for 3 h. The solid was
collected by filtration, and then, the cake was washed with 2-
propanol (57 L, 1.5 vol) and dried at 50 °C under vacuum to
yield 8·TsOH as a colorless solid (25.5 kg, 68% yield, 99.7% ee).
HPLC analysis of a sample matched that of one from the
laboratory-scale synthesis.
Kilogram-Scale Synthesis of 2-((6-Chloro-3-methyl-2,4-
dioxo-3,4-dihydropyrimidin-1(2H)-yl)methyl)benzonitrile
(10). To a reactor were charged 2-((6-chloro-3-methyl-2,4-
dioxo-3,4-dihydropyrimidin-1(2H)-yl)methyl)benzonitrile (9)
(12.0 kg, 43.53 mol, 1.00 equiv), 8·TsOH (13.2 kg, 43.95 mol,
1.01 equiv), potassium carbonate (12.1 kg, 87.55 mol, 2.01
equiv), 2-PrOH (14.4 L, 1.2 vol), and water (58.6 L, 4.8 vol).
This suspension was heated to 65 °C and stirred for 18 h. To the
reaction mixture was added water (79.2 L, 6.6 vol), and then, it
was gradually cooled to 5 °C. The resulting suspension was
stirred at 5 °C for 3 h. The solid was collected by filtration, and
the cake was washed with water (48.0 L, 4.0 vol) and dried at 60
°C under vacuum to yield 10 as a pale-yellow solid (15.3 kg, 96%
yield, 99.4 area % HPLC purity). HPLC analysis of a sample
matched that of one from the laboratory-scale synthesis.
Lab-Scale Synthesis of Trelagliptin (12) Succinate.
Crystalline 12 (1.0 g, 2.8 mmol, 1.00 equiv) was dissolved in hot
THF (4.5 mL) with 2 drops of water, and the solution was
filtered through a membrane filter. Then, the solution was slowly
added to an already filtered succinic acid (331 mg, 2.8 mmol)
solution in THF (4 mL) and 2-PrOH (2.5 mL) at 65 °C.
Crystals formed and were aged at 65 °C for 30 min; then, the
suspension was cooled to room temperature. After stirring for 16
h, the suspension was cooled to 0−5 °C and stirred for a further
2 h. The crystals were filtered and dried in vacuo at 45 °C to give
a white crystalline solid (1.2 g, 93% yield, 99.8 area % HPLC
purity). The 1H NMR spectrum was identical with the reported
data.14 1H NMR (500 MHz, DMSO-d6): δ 7.95 (dd, J = 8.7 Hz,
5.5 Hz, 1H), 7.35 (td, J = 8.5 Hz, 2.5 Hz, 1H), 7.17 (dd, J = 9.6
Hz, 2.4 Hz, 1H), 5.38 (s, 1H), 5.20 (d, J = 16.4 Hz, 1H), 5.12 (d,
J = 16.1 Hz, 1H), 3.14 (m, 1H), 3.09 (s, 3H), 3.08 (m, 1H),
3.00−3.07 (m, 1H), 2.91 (d, J = 11.4 Hz, 1H), 2.54−2.77 (m,
2H), 1.66−1.97 (m, 2H), 1.42−1.57 (m, 1H), 1.35 (d, J = 8.8
Hz, 1H). 13C NMR (126 MHz, DMSO-d6): δ 175.2, 166.1,
164.1, 162.7, 159.7, 152.3, 145.8, 136.53, 136.45, 117.1, 115.7,
106.9, 90.3, 55.8, 51.7, 47.0, 46.3, 31.6, 27.9. HPLC: SHISEIDO
MG II column (5 μm, 150 × 4.6 mm), 0.02 M aq H3PO4/MeCN
= 6/4, flow rate: 1.0 mL/min, oven temperature: 25 °C,
detection: 278 nm (UV), tR: 1.4 min (12), tR: 2.7 min (14).
Manufacturing for Scale-Up Study. Kilogram-Scale
Synthesis of 1,4,5,6-Tetrahydropyridine-3-carboxamide (7)
p-Toluenesulfonate. A nitrogen-substituted 1,500 L pressure
reactor was charged with nicotinamide (6) (70.0 kg, 0.573 kmol,
1.00 equiv), Pd/C type-K (7.0 kg, 50% wet), and methanol (700
L, 10 vol). The reactor was pressurized with hydrogen (0.07
MPa) and vented nine times. The hydrogen pressure was set to
0.08 MPa, and heating was started. After a reaction temperature
of 37 °C was reached, the pressure was adjusted to 0.10 MPa.
The reaction mixture was stirred with the hydrogen pressure
maintained within the range of 0.09−0.10 MPa at 37−43 °C for
33 h. The reactor was cooled below 40 °C, and the hydrogen was
carefully vented. The reactor was pressurized with nitrogen
(0.02 MPa) and vented four times. The reaction mixture was
heated to 55 °C and stirred at 55 °C for 1 h. Pd/C was removed
by nitrogen-pressurized filtration at 50−57 °C with methanol
rinsing (70 L). The filtrates were concentrated under reduced
pressure to about 210 L (3 vol). The resulting slurry was cooled
to 5 °C and aged at 5 °C for 2 h. The solid was collected by
filtration, and the cake was washed with cool methanol (42 L, 0.6
vol). A 1,200 L reactor was charged with the wet-collected
powder and ethanol (420 L, 6 vol). p-Toluenesulfonic acid (87.2
kg, 0.458 kmol, 0.80 equiv) was added to the reactor at 26 °C,
and then, a seed crystal of 7·TsOH (70 g) was inoculated. The
resulting suspension was aged at 25 °C for 3 h before being
filtered, and then, the cake was washed with ethanol (140 L, 2
vol) and dried at 60 °C under vacuum to yield 7·TsOH 0.85
ethanol solvate as a colorless solid (116.3 kg, 60% yield). HPLC
analysis of a sample matched that of one from the laboratory-
scale synthesis.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.oprd.0c00544.
■
sı
General analytical methods and available copies of HPLC
chromatograms and 1H NMR and 13C NMR for
compounds 7·TsOH, 8·TsOH, 10, 1·BzOH, 14, and
12·HOOC(CH2)2COOH (PDF)
AUTHOR INFORMATION
Corresponding Author
■
Masatoshi Yamada − Process Chemistry, Pharmaceutical
Sciences, Takeda Pharmaceutical Company Limited, Osaka,
Japan; orcid.org/0000-0002-8390-6021;
Email: masatoshi.yamada@spera-pharma.co.jp
Kilogram-Scale Synthesis of (R)-Nipecotamide (8) p-
Toluenesulfonate. A 500 L pressure reactor was charged with
7·TsOH 0.85 ethanol solvate (50.0 kg, 0.148 kmol, 1.00 equiv),
335
https://dx.doi.org/10.1021/acs.oprd.0c00544
Org. Process Res. Dev. 2021, 25, 327−336

Organic Process Research & Development
Authors
Sayuri Hirano − Process Chemistry, Pharmaceutical Sciences,
Takeda Pharmaceutical Company Limited, Osaka, Japan
Ryoji Tsuruoka − Process Chemistry, Pharmaceutical Sciences,
Takeda Pharmaceutical Company Limited, Osaka, Japan;
orcid.org/0000-0003-3748-9172
Masahiro Takasuga − Process Chemistry, Pharmaceutical
Sciences, Takeda Pharmaceutical Company Limited, Osaka,
Japan
Kenichi Uno − Process Chemistry, Pharmaceutical Sciences,
Takeda Pharmaceutical Company Limited, Osaka, Japan
Kotaro Yamaguchi − Process Chemistry, Pharmaceutical
Sciences, Takeda Pharmaceutical Company Limited, Osaka,
Japan
pubs.acs.org/OPRD
Article
(7) Yamada, M.; Hirano, S.; Tsuruoka, R.; Yamano, M. Process for
Producing Heterocyclic Compound. WO 2015137496 A1, 17 Sep
2015.
(8) (a) Quan, P. M.; Quin, L. D. 1,4,5,6-Tetrahydropyridines from
Catalytic Reduction of Nicotinoyl Derivatives and Their Ring Opening
with Hydrazine. J. Org. Chem. 1966, 31, 2487−2490. (b) Wenkert, E.;
Dave, K. G.; Haglid, F.; Lewis, R. G.; Oishi, T.; Stevens, R. V.;
Terashima, M. Tetrahydropyridines. J. Org. Chem. 1968, 33, 747−753.
(9) (a) Fuchs, R.; Roduit, J.-P. Process for the Preparation of Optically
Active Piperidine 2-Carboxylic Acid Derivatives. U.S.Patent 5,886,181
A, 3 March 1999. (b) Lei, A.; Chen, M.; He, M.; Zhang, X. Asymmetric
Hydrogenation of Pyridines: Enantioselective Synthesis of Nipecotic
Acid Derivatives. Eur. J. Org. Chem. 2006, 2006, 4343−4347.
(10) Kitamura, M.; Tokunaga, M.; Ohkuma, T.; Noyori, R.
Convenient Preparation of Binap-Ruthenium (II) Complexes Catalyz-
ing Asymmetric Hydrogenation of Functionalized Ketones. Tetrahe-
dron Lett. 1991, 32, 4163−4166.
Mitsuhisa Yamano − Process Chemistry, Pharmaceutical
Sciences, Takeda Pharmaceutical Company Limited, Osaka,
Japan
(11) Pye, P. J.; Rossen, K.; Reamer, R. A.; Volante, R. P.; Reider, P. J.
[2.2]Phanephos-Ruthenium(II) Complexes: Highly Active Asymmet-
ric Catalysts for the Hydrogenation of β-Ketoesters. Tetrahedron Lett.
1998, 39, 4441−4444.
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.oprd.0c00544
(12) (a) Hofmann, A. W. Ueber die Einwirkung des Broms in
̈
alkalischer Losuug auf Amide. Ber. Dtsch. Chem. Ges. 1881, 14, 2725−
Notes
2736. (b) Chai, Y.; Wan, Z.-L.; Wang, B.; Guo, H.-Y.; Liu, M.-L.
Synthesis and in vitro antibacterial activity of 7-(4-alkoxyimino-3-
amino-3-methylpiperidin-1-yl)fluoroquinolone derivatives. Eur. J. Med.
Chem. 2009, 44, 4063−4069. (c) Ewing, W. R.; Zhu, Y.; Sun, C.;
Huang, Y.; Karatholuvhu, M. S. Diaminocyclohexane Compounds and
Uses Thereof. WO 2013012829 A1, 24 Jan 2013.
The authors declare the following competing financial
interest(s): The research described was totally funded by
TAKEDA Pharmaceutical Company Limited.
ACKNOWLEDGMENTS
(13) (a) Abrecht, S.; Adam, J.-M.; Bromberger, U.; Diodone, R.;
Fettes, A.; Fischer, R.; Goeckel, V.; Hildbrand, S.; Moine, G.; Weber, M.
An Efficient Process for the Manufacture of Carmegliptin. Org. Process
Res. Dev. 2011, 15, 503−514. (b) Taylor, G. M.; Baker, S. J.; Gedney, A.;
Pearson, D. J.; Sibley, G. E. M. On the Ritter Reaction of Cyclic
Hydroxyamines: Synthesis of Conformationally-Restricted Reduced
Amide Dipeptide Isosteres. Tetrahedron Lett. 1996, 37, 1297−1300.
(c) Mercado-Martin, E. V.; Garcia, R. P.; Romminger, S.; Pimenta, E.
F.; Romney, D. K.; Lodewyk, M. W.; Williams, D. E.; Andersen, R. J.;
Miller, S. J.; Tantillo, D. J.; Berlinck, R. G. S.; Sarpong, R. Nature 2014,
509, 318−324.
(14) Zhang, Z.; Wallace, M. B.; Feng, J.; Stafford, J. A.; Skene, R. J.;
Shi, L.; Lee, B.; Aertgeerts, K.; Jennings, A.; Xu, R.; Kassel, D. B.;
Kaldor, S. W.; Navre, M.; Webb, D. R.; Gwaltney, S. L. Design and
Synthesis of Pyrimidinone and Pyrimidinedione Inhibitors of
Dipeptidyl Peptidase IV. J. Med. Chem. 2011, 54, 510−524.
■
We thank our industrial partners Hamari Chemicals, Ltd for the
manufacturing operation of the asymmetric hydrogenation and
FUJIFILM Wako Pure Chemical Corporation for catalyst
preparation. Also, we thank our TAKEDA colleague Kenji
Kuroda and SPERA colleagues Hideya Mizufune, Toshihiko
Fujitani, David Cork, Koichiro Fukuoka, and Koji Mukai for
their helpful discussions regarding this project.
REFERENCES
■
(1) Feng, J.; Gwaltny, S. L.; Stafford, J. A.; Zhang, Z. Dipeptidyl
Peptidase Inhibitors. WO 2005095381 A1, 13 Oct 2005.
(2) Feng, J.; Gwaltny, S. L.; Stafford, J. A.; Zhang, Z.; Elder, B. J.;
Isbester, P. K.; Palmer, G. J.; Salsbury, J. S.; Ulysse, L. Dipeptidyl
Peptidase Inhibitors. WO 2007035629 A2, 29 March 2007.
(3) Feng, J.; Zhang, Z.; Wallace, M. B.; Stafford, J. A.; Kaldor, S. W.;
Kassel, D. B.; Navre, M.; Shi, L.; Skene, R. J.; Asakawa, T.; Takeuchi, K.;
Xu, R.; Webb, D. R.; Gwaltney, S. L. Discover of Alogliptin: A Potent,
Selective, Bioavailable, and Efficacious Inhibitor of Dipeptidyl
Peptidase IV. J. Med. Chem. 2007, 50, 2297−2300.
(4) Fujita, H.; Taniai, H.; Murayama, H.; Ohshiro, H.; Hayashi, H.;
Sato, S.; Kikuchi, N.; Komatsu, T.; Komatsu, K.; Komatsu, K.; Narita,
T.; Yamada, Y. DPP-4 inhibition with alogliptin on top of angiotensin II
type 1 receptor blockade ameliorates albuminuria via up-regulation of
SDF-1α in type 2 diabetic patients with incipient nephropathy. Endocr.
J. 2014, 61, 159−166.
(5) https://www.takeda.com/en-ca/newsroom/news-releases/
2015/new-takeda-type-2-diabetes-treatments-nesina-alogliptin-and-
kazano-alogliptin-and-metformin-hydrochloride-now-available-for-
quebec-patients/ (accessed before September 2020).
(6) (a) Samuel, H. J.; Meek, G. A. Method for Producing 3-
Aminopiperidine Diastereomer. WO 2007075630 A1, 5 July 2007.
(b) Mori, K.; Nishiyama, A.; Nojiri, M.; Taoka, N. Method for
Producing Optically Active 3-Aminopiperidine or salt thereof. WO
2008102720 A1, 28 Aug 2008. (c) Wallace, M. B.; Cody, J.; Fornicola,
R.; Garcia-Rubio, S.; Kisanga, P. B.; Reeve, M. M.; Wilson, C.
Preparation of (R)-3-aminopiperidine dihydrochloride. WO
2007112368 A1, 4 Oct 2007.
336
https://dx.doi.org/10.1021/acs.oprd.0c00544
Org. Process Res. Dev. 2021, 25, 327−336