A Second Generation Synthesis of Benzyl Piperidine Derivatives: A Key Intermediate for the Preparation of SERT/5-HT1A Dual Inhibitor

Article abstract of DOI:10.1021/acs.oprd.5b00207

A second generation process for benzyl piperidine 10 is described. By the use of a Horner-Wadsworth-Emmons reaction and selective hydrogenation with Pt/C in ethyl acetate, 2-bromo-5-(hydroxymethyl) phenol 14 was efficiently converted to the compound 10 on a 5 kg scale. A small amount of water was found to be critical to complete the selective hydrogenation with low levels of debrominated byproduct 15. Impurities of the compound 10 were controlled by limiting the quality of 2-bromo-5-(hydroxymethyl) phenol 14 and 1-bromo-2-methoxyethane 15.

Full text of DOI:10.1021/acs.oprd.5b00207

Article
pubs.acs.org/OPRD
A Second Generation Synthesis of Benzyl Piperidine Derivatives: A
Key Intermediate for the Preparation of SERT/5-HT1A Dual Inhibitor
Atsushi Ueno, Nobuyuki Ae, Hideo Terauchi, Koji Fujimoto, and Yuji Fujiwara*
Process Chemistry Research and Development Laboratories, Sumitomo Dainippon Pharma Co., Ltd., 3-1-98 Kasugade-naka,
Konohana-ku, Osaka 554-0022, Japan
S
* Supporting Information
ABSTRACT: A second generation process for benzyl piperidine 10 is described. By the use of a Horner−Wadsworth−Emmons
reaction and selective hydrogenation with Pt/C in ethyl acetate, 2-bromo-5-(hydroxymethyl) phenol 14 was efficiently converted
to the compound 10 on a 5 kg scale. A small amount of water was found to be critical to complete the selective hydrogenation
with low levels of debrominated byproduct 15. Impurities of the compound 10 were controlled by limiting the quality of 2-
bromo-5-(hydroxymethyl) phenol 14 and 1-bromo-2-methoxyethane 15.
iii. At the hydrogenation step, 40 wt % (vs starting material
9) of expensive 10% Rh/C (50% wet) was needed for full
conversion.
INTRODUCTION
■
Increasing numbers of patients are affected by neuropsychiatric
disorders such as schizophrenia, bipolar disorder, and
depression. Selective serotonin reuptake inhibitors (SSRIs)
such as escitalopram and paroxetine are commonly used for the
treatment of depressions. However, the slow onset of the action
is an issue. It has been reported that the combination of SSRIs
with 5-HT1A antagonists can accelerate the onset action of
SSRIs,1 so we have started a program to explore such new drug
candidates. We were therefore pleased to find that the novel
benzyl piperidine compound connected to a chromone core (1,
Scheme 1) was a promising candidate as it is both a dual
serotonin transport inhibitor and a 5-HT1A inhibitor.2
In the early stages of the clinical development, the
intermediate 2 was manufactured by the route shown in
Scheme 2. Nucleophilic addition of 2-methoxyethanol to 4-
bromo-3-fluorobenzoic acid 4 gave 5 in 87% yield. A diborane
reduction of the carboxylic acid 5 gave the benzyl alcohol 6,
which was sufficiently pure for the next step. Bromination with
47% aqueous HBr/toluene system, followed by treatment with
triphenyl phosphine, precipitated the phosphonium salt 8 in
78% yield over 3 steps. A Wittig reaction of 8 with N-Boc
piperidone in 2-propanol in the presence of K2CO3 gave a clean
conversion to 9, which was hydrogenated with Rh/C in ethyl
acetate to give 10, together with the debrominated 18.
Refluxing 10 in 2-propanol with concentrated hydrochloric
acid removed the Boc group, and after recrystallization, the
hydrochloride of 2 was obtained in 87% yield over three steps.
The overall yield of the hydrochloride of 2 from 4-bromo-3-
fluorobenzoic acid 4 was 58%.
RESULTS AND DISCUSSION
■
To address these problems, new routes were investigated. A
potential second generation synthesis was found as shown in
Scheme 3.
We were delighted to find that the selective decarbonylative
Heck reaction3 of the acid chloride 11 and the methylene
piperidine 12 gave 13, which was then hydrogenated with Pt/C
in toluene to give 10. The overall yield from commercially
available 4 was 24% in 4 steps. Although this route was shorter
than the first generation procedure, the yield from the coupling
reaction was moderate (∼50%). The purity of 10 was also not
satisfactory, and so a second recrystallization at the final step
was necessary to improve its purity. Although we felt that this
approach is attractive, due to the issues described, we decided
to move to another process for further scale-up.
The synthetic route we then established for the second
generation synthesis of compound 10 is shown in Scheme 4.4,5
Alkylation of 2-bromo-5-(hydroxymethyl)phenol 14 with 15
proceeded efficiently in N-methyl pyrrolidone with powdered
potassium carbonate as a base. When 1-chloro-2-methoxy-
ethane was used instead of the bromide, the reaction did not
complete. With KI as an additive, the reaction was faster but
impurities increased, leading to a diminished yield and quality.
After extraction, 6 was obtained as a toluene solution which was
directly treated with 48% aqueous HBr at 70 °C for 2 h to give
the benzyl bromide 7. Although compound 7 is a highly
crystalline material, we were aware that this is strong irritant to
eyes and skin even in a laboratory-scale experiment. So we
decided to use it as a toluene solution in the next step without
further purification. An Arbuzov reaction of the benzyl bromide
7 with triisopropyl phosphite gave the corresponding
phosphonate 16. As neat conditions gave increased impurities,
The route was successfully scaled up to supply 40 kg of API
for early clinical trials. However, the route needed to be
improved for future material supply due to the following three
problems:
i. In the first step, hydrogen fluoride is formed at acidic
workup.
ii. After the Wittig reaction, silica gel column chromatog-
Received: June 25, 2015
raphy was needed to remove phosphine oxide.
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.5b00207
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. Process To Prepare 1
Scheme 2. Synthetic Route of Hydrochloride of 2 for Early Stage Development
Scheme 3. Route to 10 via Decarbonylative Heck Reaction
Scheme 4. Second Generation Synthesis of Compound 10
B
DOI: 10.1021/acs.oprd.5b00207
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
a
Table 1. Condition Optimization for the Horner−Wadsworth−Emmons Reaction
entry
reagent
solvent
NMP
temp. (°C)
time (h)
16 area %
9 area %
17/(9 + 17) %
1
(i-Pr)2NEt/LiCl
DBU/LiCl
Cs2CO3
80
6.5
6.5
99.6
93.3
97.2
99.6
96.6
96.7
6.9
< 0.1
< 0.1
< 0.1
< 0.1
< 0.1
2.9
< 0.1
< 0.1
< 0.1
< 0.1
< 0.1
< 0.1
0.2
2
MeCN
reflux
reflux
reflux
reflux
23
3
THF
6.5
4
K2CO3
THF
6.5
5
K3PO4
THF
6.5
6
NaOEt
THF/PhMe 2:1
THF/PhMe 2:1
THF/PhMe 2:1
THF/PhMe 2:1
THF/PhMe 2:1
THF/PhMe 2:1
4.0
7
t-BuONa
LiHMDS
t-BuOK
0
14.0
14.0
14.0
14.0
14.0
87.8
7.5
8
0
57.3
0.0
22.2
3.4
9
0
89.2
89.3
80.0
10
11
KHMDS
NaHMDS
0
2.8
0.3
0
12.0
0.2
a
All reactions were performed in 0.5 g scale under nitrogen atmosphere and were analyzed by HPLC (condition A, see Experimental Section).
a
Table 2. Catalyst and Solvent Screening for the Hydrogenation Reaction
entry
catalyst
solvent
AcOEt
9
10
18
1
2
3
4
5
6
7
10%Rh/C (40 wt %)
SGS-10DR (10 wt %)
SGS-10DR (10 wt %)
PdBN (5 wt %)
0.7%
0.5%
98.1%
77.3%
32.2%
13.3%
< 0.1
1.3%
22.2%
0.5%
MeOH
AcOEt
67.2%
84.9%
> 99%
0.4%
MeOH
AcOEt
13.5%
< 0.1
6.6%
PdBN (2.5 wt %)
5%Pt/C (10 wt %)
5% Pt/C (10 wt %)
2-propanol
AcOEt
92.9%
98.1%
0.8%
1.1%
a
All reactions were performed in 0.2 or 0.5 g scale under hydrogen atmosphere (baloon) at room temperature (20−23 °C). The reaction mixture
was diluted with acetonitrile, filtered, and analyzed by HPLC (condition B, see Experimental Section).
dilution with toluene (same weight as the crude benzyl
bromide) was critical for a clean conversion. When trimethyl
phosphite was used, the reaction did not complete because the
boiling point of the reagent (bp 110 °C) was too low. The
removal of isopropyl bromide (bp 59−60 °C) was also
important for full conversion, as the inner temperature did not
reach 120 °C when the reaction was performed with refluxing.
As the phosphonate 16 is an oil, it was used in the next step
without further purification.
The results of the base and solvent screening of the Horner−
Wadsworth−Emmons (HWE) reaction are shown in Table 1.
Weak bases such as i-Pr2NEt/LiCl, DBU/LiCl, Cs2CO3,
K2CO3, K3PO4, or NaOEt did not give 9 even at elevated
temperature (entries 1−6). Use of sodium t-butoxide at 0 °C
gave the desired product but with incomplete conversion (entry
7). LiHMDS gave more than 20% HPLC area of 17 (entry 8).
Although the byproduct 17 could be converted to the desired
10 under hydrogenation conditions, we noticed that the
formation of 17 needed to be suppressed for the following
reasons. Recrystallization of compound 9 was necessary
because it was confirmed that excess triisopropyl phosphite,
which was used in the previous step, is a catalyst poison.6 It was
revealed that triisopropyl phosphite could be completely
removed when the product 9 was recrystallized from aqueous
2-propanol. However, 17 was also removed to the mother
liquor, leading to a diminished yield. To maximize the product
yield, conditions that do not form byproduct 17 were
investigated. With higher loading of t-BuONa, the reaction
still did not complete even at elevated temperature and gave
irreproducible results. Gratifyingly, about 3% of 17 was formed
with t-BuOK when the reaction was performed below 5 °C,
while the starting material was consumed in 6 h.7 As KHMDS
and NaHMDS are more expensive, t-BuOK was chosen as the
base in this step. After this coupling reaction, 9 was extracted
with toluene, and the solvent was switched to 2-propanol.
Recrystallization from aqueous 2-propanol gave triisopropyl
phosphite-free 9 in 83% yield from 2-bromo-5-(hydroxymeth-
yl) phenol 14 with >99.0% HPLC area.
Then the stage was set for the reduction of the double bond
of 9. In our previous procedure, 40 wt % of 10%Rh/C (50%
wet) was used for full conversion (Table 2, entry 1). Poisoned
Pd catalysts such as SGS-10DR8 (entries 2, 3), Pd-BN catalyst9
(entries 4,5) did not give fruitful results, but 5% Pt/C was
found to be promising with only 10 wt % of catalyst (50% wet)
loading (entry 6). Switching the solvent from 2-propanol to
ethyl acetate was effective but gave only 1.1% of debrominated
18 when reaction completed (entry 7). The catalyst loading
was then optimized to 13 wt %, as decreased catalyst loading
gave incomplete conversion.
C
DOI: 10.1021/acs.oprd.5b00207
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Solvent screening with 5% Pt/C revealed that toluene and
methyl tert-butyl ether (MTBE) were also promising. As
debrominated 18 was found to be efficiently removed by
recrystallization from aqueous 2-propanol, it was desirable to
use a solvent with a low boiling point. Toluene was omitted for
this reason. MTBE (bp 55−56 °C) was better than ethyl
acetate (bp 76−77 °C), but a slight difference of catalyst
loading led to an increase in the formation of debrominated 18
(Figure 1).
Table 3. HPLC Area of Unreacted Starting Material 9 vs
Desired Product 10 over Time
entry
1
2
3
4
weight ratio of water
added vs dried Pt/C
catalyst
0.0
(dried)
1.2
(54% wet)
2.9
4.5
reaction time
unreacted starting material 9 vs product 10
3 h
6 h
9 h
note
52.7%
26.9%
14.4%
6.9%
3.1%
0.1%
0.1%
0.1%
0.1%
1.5%
standard conditions
Figure 1. Catalyst loading of 5% Pt/C and HPLC area % of
debrominated 18.
Figure 2. Weight ratio of water to dry 5% Pt/C catalyst and HPLC
area % of debrominated 18 vs desired product 10. The catalyst used in
this experiment contained 54% of water, which corresponded to 1.2
weight ratio of water vs dried catalyst.
For stable yield and reproducibility, ethyl acetate was chosen
as the solvent for this step.
It was also found that the hydrobromide of 2 is formed
(about 2% by HPLC area) during hydrogenation, presumably
due to the hydrogen bromide formed through the debromina-
tion reaction. With the preliminary study, the formation of this
compound was suppressed by addition of powdered K2CO3
during the hydrogenation without affecting the reaction rate.
Regarding the catalyst, 13 wt % of 50% wet 5% Pt/C Nx type10
was optimal, leading to reproducible reactions which completed
in about 10 h under atmospheric pressure at 25 °C.
The use of Pt/C slurry in water is normally desirable for
handling purposes when charging a catalyst to a reactor upon
scale up. However, it was revealed that the addition of water
significantly increased the formation of debrominated 18. For
example, addition of two weight ratios of water vs 50% wet
catalyst, which provided a good slurry of the catalyst, led to the
formation of 15.4% of debrominated 18 by HPLC area %.
However, this is reasonable considering that polar, protic
solvents such as methanol, 2-propanol, and 1-butanol have a
tendency to facilitate undesired debromination reactions (Table
2, entries 6, 7). In laboratory scale, when 50% wet 5% Pt/C was
dried under reduced pressure and used in this reaction, less
debrominated 18 was formed. However, this condition is not
acceptable, not only because of the safely issue, but also because
the reaction rate was slower than the original conditions (i.e.,
50% wet catalyst with no additional water) as shown in Table 3.
The water content of the catalyst positively correlates to the
amount of 18 formed as shown in Figure 2. Although less water
is favorable in terms of selectivity, the reaction rate was slower.
We decided to use 50% wet 5% Pt/C suspended in ethyl
acetate under nitrogen and charged to the reactor.
this purification. These impurities were classified into two
groups as follows.
The first group consisted of process-derived impurities which
were present in the first generation procedure (Figure 3).
Compound 17 (less than 0.10% by HPLC area %) was formed
by the HWE reaction as described before, 18 (0.3−0.4%) was
formed at the hydrogenation step and 19 (0.1−0.2%) was
formed at the bromination step. These compounds could be
easily removed by recrystallization during the following steps.
One exception was 9, which was an unreacted starting material
from the hydrogenation step. We decided to control 9 by
HPLC in process control analysis.
The second group consisted of material derived impurities
(Figure 4). Compounds 20 to 24 were derived from impurities
in 1-bromo-2-methoxyethane 15. Analysis of the reagent grade
15 revealed that it contained corresponding bromides that
would lead to these impurities (i.e., methyl bromide, 1-bromo-
2-ethoxyethane, 1-bromo-2-chloroethane, 1-bromo-3-methox-
ypropane, and 1-bromo-2-isopropoxyethane), which also
indicates that the formation of these compounds could be
avoided by limiting its impurity levels. The precursor of
compound 25 was considered to be formed through the
reaction of 7 with the unreacted starting material 14. It was
found that 14 could be easily removed to 0.03% (HPLC area
%) by basic washing at workup of 6. However, the impurity 25
was still observed, indicating the presence of another pathway
which gives 25. LC-MS analysis and confirmation by synthesis
revealed that 0.3 to 0.9% (HPLC area %) of 26 is included in
14 (Figure 5).11 However, as long as the level of 26 in 14 did
not exceed 0.9% and basic washing at workup to prepare 6 was
performed, the level of 25 in 10 did not exceed 0.05% (HPLC
area%).
After filtration of the catalyst, the solvent was switched to 2-
propanol, and recrystallization from aqueous 2-propanol gave
90% yield of 10 with an HPLC area of >99.0%. Although 2% of
debrominated 18 was formed, it was rejected to 0.3−0.4% after
D
DOI: 10.1021/acs.oprd.5b00207
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 3. First group; process derived impurities in 10.
Figure 4. Second group; material derived impurities in 10.
Condition A; SUMIPAX ODS C-212, 5 μm (6 mmφ × 150
mm). Solvent A: 0.05% aqueous trifluoroacetic acid, solvent B:
methanol. Gradient program; solvent A/solvent B 50:50 to
10:90 over 20 min, hold 15 min, flow rate = 1.0 mL/min, 40
°C, UV detection at 220 nm.
Condition B; SUMIPAX ODS C-212, 5 μm (6 mmφ × 150
mm). Solvent A: 0.02% aqueous trifluoroacetic acid, solvent B:
0.02% trifluoroacetic acid in methanol. Gradient program;·
solvent A/solvent B, hold 25:75 for 30 min, 25:75 to 15:85 over
30 min, hold 10 min, flow rate = 0.7 mL/min, 40 °C, UV
detection at 220 nm.
Condition C; Kinetex (2.6 μm, 3 mmφ × 100 mm). Solvent
A: 0.05% aqueous trifluoroacetic acid, solvent B: acetonitrile.
Gradient program; solvent A/solvent B 90:10 to 10:90 over 24
min, hold 4 min, flow rate = 0.8 mL/min, 40 °C, UV detection
at 220 nm.
Condition D; YMC-Pack Pro C18 RS (3.0 μm, 4.6 mmφ ×
150 mm). Solvent A: 0.02% aqueous trifluoroacetic acid,
solvent B: 0.02% trifluoroacetic acid-methanol. Gradient
program; solvent A/solvent B 25:75 for 30 min, then 15:85
over 30 min, hold 10 min, flow rate = 0.7 mL/min, 40 °C, UV
detection at 220 nm.
Figure 5. Impurity in 14.
Starting from 2-bromo-5-(hydroxymethyl)phenol 14 and 1-
bromo-2-methoxyethane 1512 with controlled impurities, we
were able to synthesize 10 that is as pure as that obtained with
the first generation procedure.
To confirm the manufacturing process thus developed, a
scale-up of compound 10 was performed. Starting from 10 kg
of 14, the toluene solution of 7 was prepared in one batch. The
solution was divided into two batches, and the subsequent steps
were conducted. Thus, 6.16 and 6.32 kg of compound 10 were
obtained without issue. The overall yields from compound 14
were 72 and 75% respectively, which were consistent with the
average yields of laboratory experiments (71−77%).
CONCLUSION
■
In conclusion, the second generation process to manufacture
benzyl piperidine compound 10 was demonstrated. The HWE
reaction with t-BuOK proceeded with minimum isomerization
of the double bond and also avoided the use of silica gel column
purification. Recrystallization of the HWE product removed
triisopropyl phosphite efficiently, leading to reproducible
hydrogenation in the next step. Selective hydrogenation was
accomplished by employing 50% wet 5% Pt/C in ethyl acetate
with an adequate amounts of water, furnishing the Boc
protected benzyl piperidine 10 in 90% yield. The catalyst
loading was reduced from 40 wt % (10% Rh/C) to 13 wt %
(5% Pt/C). It was confirmed that, by limiting the level of
impurities in 2-bromo-5-(hydroxymethyl) phenol 14 and 1-
bromo-2-methoxyethane 15, the quality of 10 was sufficient for
further clinical trials. This process was successfully scaled up to
5 kg (2 batches) in 72 and 75% yield without issue. The
decarbonylative Heck reaction was found to be potential future
manufacturing method.
The purities listed were determined by area %. NMR spectra
were recorded on a Bruker Avance 400 spectrometer. Chemical
shifts (δ) are given in parts per million, and residual internal
CHCl3 (δ 7.26) and DMSO (δ 2.50). All reactions were carried
out under a nitrogen atmosphere unless otherwise mentioned.
Reagents and solvents were used as obtained from commercial
suppliers without further purification.
Conditions for original route to (10) have already been
reported.2
( 4 - B r o m o - 3 - ( 2 - m e t h o x y e t h o x y ) b e n z y l ) -
triphenylphosphonium Bromide (8). Mp 200−203 °C
1
(decomp.). H NMR (d6-DMSO, 400 MHz): 7.95−7.88 (m,
3H), 7.79−7.68 (m, 12H), 7.47 (d, J = 8.0, 0.8 Hz, 1H), 6.68
(dd, J = 2.2, 2.2 Hz, 1H), 6.58 (ddd, J = 8.0, 2.2, 2.2 Hz, 1H),
5.21 (d, J = 16.0 Hz, 2H), 3.71−3.67 (m, 2H), 3.55−3.51 (m,
2H), 3.29 (s, 3H).
13C NMR (d6-DMSO, 100 MHz): 154.5 (d, J = 3.4 Hz),
135.1 (d, J = 2.7 Hz), 134.1 (d, J = 10.0 Hz), 133.2 (d, J = 3.2
Hz), 130.1 (d, J = 12.4 Hz), 129.0 (d, J = 8.6 Hz), 124.3 (d, J =
5.7 Hz), 117.6 (d, J = 85.2 Hz), 116.1 (d, J = 5.2 Hz), 111.1 (d,
J = 5.0 Hz), 69.8, 68.1, 58.4, 28.0 (d, J = 46.3 Hz).
EXPERIMENTAL SECTION
General Information. HPLC was performed on Agilent
1100 system.
■
E
DOI: 10.1021/acs.oprd.5b00207
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
4-Bromo-3-(2-methoxyethoxy)benzoic Acid (5). A flask
was charged with potassium tert-butoxide (92.2 g, 822 mmol,
3.00 equiv), N-methylpyrrolidone (300 g), and a stir bar. To
this solution, 2-methoxyethanol (52.1 g, 685 mmol, 2.50 equiv)
was added dropwise, and 4-bromo-3-fluorobenzoic acid (4)
(60.0 g, 274 mmol) was added. The resulting mixture was
heated to 90 °C for 3 h. After the reaction completed, it was
cooled to 40−50 °C, and water (1200 g) was added dropwise
over 30 min, followed by concentrated HCl (27.0 g, 1096
mmol, 4.00 equiv) over 30 min, during which time HF gas
evolved. The resulting slurry was cooled to room temperature,
filtered, and dried under reduced pressure to give (5) (65.5 g,
238 mmol) in 86.8% yield.
(m, 1H), 1.88−1.74 (br, 1H), 1.53−1.43 (m, 1H), 1.48 (s,
9H);
13C NMR (CDCl3, 100 MHz): 155.2, 140.9, 133.0, 125.5,
122.9, 114.7, 109.9, 108.5, 80.7, 70.9, 69.0, 59.5, 42.1, 41.0,
40.0, 33.8, 28.3, 27.7;
LRMS (ESI) calcd for C15H21BrNO2 (M-Boc + H + H)
326.08, found 326.0.
tert-Butyl 4-(4-bromo-3-(2-methoxyethoxy)benzyl)-
piperidine-1-carboxylate (10). Crude tert-butyl 4-(4-
bromo-3-(2-methoxyethoxy)benzyl)-3,6-dihydropyridine-
1(2H)-carboxylate (13) in xylene (2.56 g) was diluted with
toluene (12.8 g), and the flask was evacuated and backfilled
with nitrogen. The flask was then charged with 5% Pt/C (50%
wet, 0.512 g), and the flask was filled with nitrogen. The flask
was filled with hydrogen and stirred at room temperature for 4
h. The catalyst was filtered, and the filtrate was washed with
toluene under a stream of nitrogen. 5% Pt/C (50% wet, 0.640
g) was charged to this filtrate, and hydrogenation continued
another 21 h. Then the catalyst was filtered, and the filtrate was
washed with toluene under nitrogen. The solvent was removed,
and the residue was recrystallized from 67% aqueous 2-
propanol at −10 °C to give (10) (961 mg, 2.24 mmol, 97.34%
with condition C) in 37.3% yield as a yellow powder. The loss
in the filtrate was determined to be 7.3% by HPLC (118 mg,
0.438 mmol).
Mp 176−178 °C. 1H NMR (d6-DMSO, 400 MHz): 13.2 (br,
1H), 7.70 (d, J = 8.4 Hz, 1H), 7.55 (d, J = 1.6 Hz, 1H), 7.45
(dd, J = 8.4, 1.6 Hz, 1H), 4.27−4.21 (m, 2H), 3.73−3.69 (m,
2H), 3.34 (s, 3H).
13C NMR (d6-DMSO, 100 MHz): 166.6, 154.8, 133.2, 131.7,
122.9, 116.4, 113.7, 70.2, 68.6, 58.4.
4-Bromo-3-(2-methoxyethoxy)benzoyl Chloride (11).
4-Bromo-3-(2-methoxyethoxy)-benzoic acid (5) (5.00 g, 18.2
mmol) and a stir bar were added to a flask, which was then
evacuated and backfilled with nitrogen three times. Toluene
(20 mL) and DMF (35 μL, 0.91 mmol, 0.05 equiv) were added
via syringe, and the resulting mixture was stirred at room
temperature. Oxalyl chloride (3.46 g, 27.3 mmol, 1.50 equiv)
was added dropwise over 5 min via syringe, and the resulting
solution was stirred for 3 h. Solvent was evaporated, and the
residue was distilled under reduced pressure (1.0−0.2 mmHg,
120−140 °C) to give (11) as colorless liquid, which solidified
upon standing.
1
Mp 89−90 °C. H NMR (CDCl3, 400 MHz): 7.41 (d, J =
8.0 Hz, 1H), 6.70 (d, J = 2.0 Hz, 1H), 6.62 (dd, J = 8.0, 2.0 Hz,
1H), 4.18−4.14 (m, 2H), 4.15−3.95 (br, 2H), 3.82−3.78 (m,
2H), 3.49 (s, 3H), 2.62 (dd, J = 12.2, 12.2 Hz, 2H), 2.47 (d, J =
6.8 Hz, 2H), 1.68−1.55 (m, 3H), 1.45 (s, 9H), 1.19−1.05 (m,
2H).
1
Mp 39−41 °C. H NMR (CDCl3, 400 MHz): 7.69 (d, J =
13C NMR (CDCl3, 100 MHz): 155.1, 154.8, 141.1, 132.9,
123.0, 114.8, 109.7, 79.3, 70.9, 69.0, 59.5, 44.0, 42.9, 38.1, 31.9,
28.4.
8.4 Hz, 1H), 7.62 (dd, J = 8.4, 2.0 Hz, 1H), 7.58 (d, J = 8.4 Hz,
1H), 4.27−4.23 (m, 2H), 3.85−3.82 (m, 2H), 3.49 (s, 3H).
13C NMR (CDCl3, 100 MHz): 167.6, 155.7, 133.8, 133.3,
125.1, 121.5, 114.5, 70.6, 69.3, 59.5.
LRMS (APCI) calcd for C15H23BrNO2 (M-Boc + H + H)
328.09, found 328.2.
tert-Butyl 4-(4-bromo-3-(2-methoxyethoxy)benzyl)-
3,6-dihydropyridine-1(2H)-carboxylate (13). 4-Bromo-3-
(2-methoxyethoxy)benzoyl chloride (11) (1.76 g, 6.00 mmol),
palladium trifluoroacetate (90.2 mg, 0.30 mmol, 0.05 equiv),
and a stir bar were added to a flask, which was then evacuated
and backfilled with nitrogen three times. o-Xylene (30.0 mL, 0.2
mol/L), 1-Boc-4-methylenepiperidine (12) (1.18 g, 6.0 mmol,
1.00 equiv), and N,N-diisopropylethylamine (0.97 g, 7.50
mmol, 1.25 equiv) were added via syringe, and the resulting
mixture was heated at 120 °C for 7 h. The mixture was cooled
to room temperature, and 1 mol/L aq. HCl (10 mL) and water
(20 mL) were added. The phases were separated, and the
aqueous phase was extracted with toluene (20 mL). The
combined organic phases were washed with 2.5% aq. NaOH
(30 mL) and water (30 mL) and concentrated under reduced
pressure below 40 °C to give an amber residue, which was used
in the next step without further purification.
(4-Bromo-3-(2-methoxyethoxy)phenyl)methanol (6).
2-Bromo-5-(hydroxymethyl)phenol (14) (10.0 kg, 49.3 mol),
N-methyl piperidone (20.0 kg), and potassium carbonate (10.2
kg, 73.8 mol, 1.50 equiv) were charged and warmed to 70 °C.
Then 1-bromo-2-methoxyethane (7.5 kg, 54.0 mol, 1.10 equiv)
was added dropwise over 45 min, and the mixture was stirred at
70 °C for 5 h. Additional powdered (D80 < 45 μm) potassium
carbonate (5.0 kg, 36.1 mol, 0.73 equiv) was added and stirred
for another 2 h (HPLC condition A). Toluene (60 kg) and
water (60 kg) were added, and the organic phase was separated.
The aqueous phase was extracted with toluene (30 kg), and the
combined organic phase was successively washed with 5%
aqueous potassium carbonate solution (30 kg) and water (30
kg). The organic phase was concentrated under reduced
pressure below 40 °C to a weight of 77.2 kg, which was used in
the next step without further purification.
1H NMR (CDCl3, 400 MHz): 7.47 (d, J = 8.0 Hz, 1H), 6.92
(d, J = 1.6 Hz, 1H), 6.80 (m, 1H), 4.62 (s, 2H), 4.17−4.13 (m,
2H), 3.81−3.78 (m, 2H), 3.48 (s, 3H), 2.10 (br, 1H).
13C NMR (CDCl3, 100 MHz): 155.4, 141.8, 133.2, 120.3,
112.0, 111.2, 70.8, 68.9, 64.6, 59.5.
Analytical sample was purified by silica gel column
chromatography (hexane/ethyl acetate). The compound was
a mixture of rotamers of the Boc group (3:2 mixture).
1H NMR (CDCl3, 400 MHz, 3:2 mixture of rotational
isomer): 7.42 (d, J = 8.0 Hz, 1H), 6.85 (d, J = 7.4 Hz, 0.4H),
6.73 (d, J = 2.0 Hz, 1H), 6.71 (d, J = 7.2 Hz, 0.6H), 6.66 (dd, J
= 8.0, 2.0 Hz, 1H), 4.76 (d, J = 7.4 Hz, 0.4H), 4.64 (d, J = 7.2
Hz, 0.6H), 4.18−4.15 (m, 2H), 3.82−3.78 (m, 2H), 3.78−3.61
(m, 1H), 3.49 (s, 3H), 3.35 (ddd, J = 13.0, 9.6, 3.2 Hz, 1H),
2.59 (dd, J = 13.0, 7.6 Hz, 1H), 2.55−2.47 (m, 1H), 2.46−2.38
LRMS (ESI) calcd for C10H11BrO2 (M − H2O + H) 243.10,
found 243.1.
1-Bromo-4-(bromomethyl)-2-(2-methoxyethoxy)-
benzene (7). (4-Bromo-3-(2-methoxyethoxy)phenyl)-
methanol in toluene (77.2 kg) and aqueous 47% HBr (42.4
kg, 246.3 mol) were charged to a reactor. This was heated at 70
F
DOI: 10.1021/acs.oprd.5b00207
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
°C for 2 h with vigorous stirring and then cooled to 25 °C
(HPLC condition A). The aqueous phase was separated, and
the organic phase was washed twice with water (38.6 kg x 2).
Then the toluene solution of the title compound was
concentrated under reduced pressure below 40 °C to the
weight of 32.0 kg and separated into two batches.
3.52−3.47 (m, 2H), 3.48 (s, 3H), 3.39 (dd, J = 5.8, 5.8 Hz,
2H), 2.43 (dd, J = 5.6, 5.6 Hz, 2H), 2.31 (dd, J = 5.6, 5.2 Hz,
2H), 1.47 (s, 9H).
13C NMR (CDCl3, 100 MHz): 155.0, 154.7, 139.4, 138.0,
132.9, 123.7, 122.8, 114.4, 110.3, 79.6, 70.9, 69.0, 59.5, 44.5,
36.1, 29.3, 28.4; (17C observed for expected 18C).
LRMS (APCI) calcd for C15H21BrNO2 (M − Boc + H + H)
326.08, found 326.1.
1
Mp 60−61 °C. H NMR (CDCl3, 400 MHz): 7.48 (d, J =
8.0 Hz, 1H), 6.94 (d, J = 2.0 Hz, 1H), 6.86 (dd, J = 8.0, 2.0 Hz,
1H), 4.43 (s, 2H), 4.21−4.17 (m, 2H), 3.83−3.79 (m, 2H),
3.49 (s, 3H);
tert-Butyl-4-(4-bromo-3-(2-methoxyethoxy)benzyl)-
3,6-dihydropyridine-1(2H)-carboxylate (17). This com-
pound was isolated by silica gel column chromatography.
1H NMR (CDCl3, 400 MHz): 7.42 (d, J = 8.0 Hz, 1H), 6.72
(d, J = 1.8 Hz, 1H), 6.65 (dd, J = 8.0, 1.6 Hz, 1H), 5.38 (br,
1H), 4.17−4.13 (m, 2H), 3.87 (br, 2H), 3.82−3.78 (m, 2H),
3.48 (s, 3H), 3.44 (t, J = 5.8 Hz, 2H), 3.24 (br, 2H), 1.97 (br,
2H), 1.45 (s, 9H).
13C NMR (CDCl3, 100 MHz): 155.5, 138.3, 133.5, 122.6,
114.2, 112.6, 70.8, 69.0, 59.5, 32.7;
Diisopropyl (4-bromo-3-(2-methoxyethoxy)benzyl)-
phosphonate (16). 1-Bromo-4-(bromomethyl)-2-(2-
methoxyethoxy)benzene (7) (16.0 kg as toluene solution)
and triispropyl phosphite (5.60 kg, 26.89 mol) were charged
and the solution was warmed to 120 °C over 1.5 h, during
which time isopropyl bromide and toluene were distilled off
(8.3 kg, HPLC condition A). The mixture was cooled to room
temperature. The solution was transferred to the next reactor
with toluene to give 16.65 kg of (16) as a toluene solution.
1H NMR (CDCl3, 400 MHz): 7.43 (dd, J = 8.0, 1.2 Hz, 1H),
6.91 (dd, J = 2.0, 1.6 Hz, 1H), 6.75 (ddd, J = 8.0, 2.0, 1.6 Hz,
1H), 4.60 (m, 2H), 4.18−4.14 (m, 2H), 3.81−3.78 (m, 2H),
3.47 (s, 3H), 3.03 (d, J = 21.6 Hz, 2H), 1.27 (d, J = 6.4 Hz,
6H), 1.17 (d, J = 6.4 Hz, 6H).
13C NMR (CDCl3, 100 MHz): 155.2, 154.9, 140.0, 135.6,
133.0, 122.9, 120.3, 114.6, 110.1, 79.5, 70.9, 69.0, 59.5, 43.5,
41.0, 39.8, 28.5, 28.0.
LRMS (ESI) calcd for C15H21BrNO2 (M − Boc + H + H)
326.08, found 326.0.
tert-Butyl-4-(4-bromo-3-(2-methoxyethoxy)benzyl)-
piperidine-1-carboxylate (10). tert-Butyl 4-(4-bromo-3-(2-
methoxyethoxy)benzylidene) piperidine-1-carboxylate (9)
(7.00 kg, 16.42 mol), ethyl acetate (25.0 kg), and slurry of
5% Pt/C (50% wet, 0.91 kg) in ethyl acetate (10.0 kg) were
charged to a reactor. Then the reactor was charged with
hydrogen, and the pressure was maintained between 1.0 and 2.0
atm at 25 °C. Stirring was performed for 19 h, and then the
atmosphere of the reactor was replaced with nitrogen. After
completion of the reaction was confirmed by HPLC analysis
(condition B), and the amount of H2 used, the catalyst was
filtered off and washed with 2-propanol (7.0 kg). The combined
filtrate was concentrated under reduced pressure below 40 °C
(33.5 kg distilled off) to give a solid; then the additional 2-
propanol (40.0 kg) was added, and again removal of solvent
(35.0 kg) gave a solid. 2-Propanol was added to the residue to
adjust the weight to 24.5 kg and heated to 57 °C to dissolve.
Water (8.75 kg) was added and cooled to 35 °C before seeding
(40 g). The resulting slurry was allowed to stir for 1 h and
cooled to 0 °C over 2 h. Stirring was performed for another 1 h
before filtration. The collected crystals were washed twice with
5.3 kg of 67% aqueous 2-propanol and dried under vacuum at
50 °C for 18 h to give (10) (6.32 kg, 14.75 mol, HPLC area
99.22% with condition D) in 89.9% yield.
13C NMR (CDCl3, 100 MHz): 155.1 (d, J = 3.2 Hz), 133.0
(d, J = 3.0 Hz), 132.8 (d, J = 9.0 Hz), 123.6 (d, J = 7.2 Hz),
115.2 (d, J = 6.3 Hz), 110.7 (d, J = 4.6 Hz), 70.7 (d, J = 5.6
Hz), 70.6, 68.9, 59.4, 34.6 (d, J = 139.2 Hz), 24.0 (d, J = 3.8
Hz), 23.8 (d, J = 5.0 Hz).
LRMS (APCI) calcd for C16H27BrO5P (M + H) 409.08,
found 409.1.
tert-Butyl 4-(4-bromo-3-(2-methoxyethoxy)-
benzylidene)piperidine-1-carboxylate (9). Diisopropyl
(4-bromo-3-(2-methoxyethoxy)benzyl)phosphonate (16) in
toluene (16.65 kg), tert-butyl 4-oxopiperidine-1-carboxylate
(5.40 kg, 27.10 mol), and toluene (13.53 kg) were charged to a
reactor. After dissolution, the mixture was cooled to −5 °C. A
solution of potassium tert-butoxide (3.87 kg, 34.49 mol) in
tetrahydrofuran (40.32 kg) was added to the mixture dropwise
over 65 min, during which time the temperature was kept
below 5 °C. After 7 h at −7 to −1 °C (HPLC area of the
starting material was less than 1.0% vs the product by condition
A), water (60.5 kg) was added,and phase separated. The
aqueous phase was removed and the organic phase was washed
with 2% sodium chloride solution (61.7 kg). To the organic
layer 2-propanol (40.8 kg) was added and concentrated under
reduced pressure below 40 °C to distill off 42.0 kg of solvent.
Residue was adjusted to 30.28 kg with additional 2-propanol
and heated to 50 °C. Water (10.1 kg) was added, and the
solution was cooled to 35 °C before seed crystal (50 g) was
added. Stirring was performed for 1 h, and then additional
water (10.1 kg) was added dropwise over 1 h. The white slurry
was cooled to 5 to 1 °C, and the crystals was filtered. The wet
cake was washed three times with 67% aqueous 2-propanol (6.7
kg × 3) and dried under a stream of nitrogen at 40 °C for 17 h
to give the title compound (8.70 kg, 20.41 mol, HPLC area
99.31% with condition D) in 82.9% yield over four steps from
(14).
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.oprd.5b00207.
NMR spectra of important intermediates (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: yuji-fujiwara@ds-pharma.co.jp.
Notes
The authors declare no competing financial interest.
1
ACKNOWLEDGMENTS
Mp 80−81 °C. H NMR (CDCl3, 400 MHz): 7.45 (d, J =
■
8.0 Hz, 1H), 6.74 (d, J = 1.6 Hz, 1H), 6.67 (dd, J = 8.0, 1.6 Hz,
1H), 6.28 (br, 1H), 4.17−4.14 (m, 2H), 3.82−3.78 (m, 2H),
We thank Ms. Norie Tsuboya for analytical support and Dr.
Kiichi Kuroda for proofreading of the manuscript.
G
DOI: 10.1021/acs.oprd.5b00207
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
REFERENCES
■
(1) (a) Artigas, F.; Perez, V.; Alvarez, E. Arch. Gen. Psychiatry 1994,
51, 248−251. (b) Zhou, D.; Stack, G. P.; Lo, J.; Failli, A. A.; Evrard, D.
A.; Harrison, B. L.; Hatzenbuhler, N. T.; Tran, M.; Croce, S.; Yi, S.;
Golembieski, J.; Hornby, G. A.; Lai, M.; Lin, Q.; Schechter, L. E.;
Smith, D. L.; Shilling, A. D.; Huselton, C.; Mitchell, P.; Beyer, C. E.;
Andree, T. H. J. Med. Chem. 2009, 52, 4955−4959.
(2) (a) Toyoda, T.; Yoshinaga, H. PCT Int. Appl. WO2009/099087,
2009. (b) Toyoda, T.; Nishida, T.; Yoshinaga, H. PCT Int. Appl.
WO2011/016468, 2011. (c) Nishida, T.; Yoshinaga, H.; Toyoda, T.
Org. Process Res. Dev. 2012, 16, 625−634.
(3) (a) Decarbonylative Heck reaction with Pd: Blaser, H.-U.;
Spencer, A. J. Organomet. Chem. 1982, 233, 267−274. (b) Spencer, A.
J. Organomet. Chem. 1982, 240, 209−216. (c) Spencer, A. J. Organomet.
Chem. 1983, 247, 117−122. (d) Spencer, A. J. Organomet. Chem. 1984,
265, 323−331. (e) Sugihara, T.; Satoh, T.; Miura, M. Tetrahedron Lett.
2005, 46, 8269−8271. Decarbonylative Heck reaction with Rh:
(aa) Sugihara, T.; Satoh, T.; Miura, M.; Nomura, M. Angew. Chem., Int.
Ed. 2003, 42, 4672−4674.
(4) Reviews on the preparation of piperidines; (a) Weintraub, M. P.;
Sabol, S. J.; Kane, M. J.; Borcherding, R. D. Tetrahedron 2003, 59,
2953−2989. (b) Buffat, G. P. M. Tetrahedron 2004, 60, 1701−1729.
(5) Examples of the synthesis of benzyl piperidine compounds. (a)
Method with Horner−Wadsworth−Emmons reaction and reduction
of double bond : Miki, S.; Nakamoto, K.; Takeda, M. PCT Int. Appl.
WO2002/057235. (b)
Method via Wolff−Kishner reduction of
carbonyl group after Friedel Crafts reaction : Wick, A.; Frost, J.;
Gaudilliere, B.; Bertin, J.; Dupont, R.; Rousseau, J. U.S. Patent
4,690,931, 1987.
(6) Addition of triisopropyl phosphite to pure 9 completely shuts
down the reduction, thus proved to be a catalyst poison.
(7) The order of addition was critical to the outcome of the reaction.
When the phosphonate 16 was treated with t-BuOK and then N-Boc
piperidone was added, more byproduct 17 was observed. The best
result was obtained when mixture of 16 and N-Boc piperidone was
added t-BuOK in THF below 5 °C.
(8) Mori, A.; Mizusaki, T.; Kawase, M.; Maegawa, T.; Monguchi, Y.;
Takao, S.; Takagi, Y.; Sajiki, H. Adv. Synth. Catal. 2008, 350, 406−410.
SGS-10DR is commercially available from N. E. Chemcat Corporation.
(9) Yabe, Y.; Yamada, T.; Nagata, S.; Sawama, Y.; Monguchi, Y.;
Sajiki, H. Adv. Synth. Catal. 2012, 354, 1264−1268.
(10) Purchased from N. E. Chemcat Corporation.
(11) Although we are not aware of the procedure to manufacture 14,
the reduction of the benzoic acid to form benzyl alcohol would be
reasonable based on the literature precedents. Burkitt, A. S.; Cardozo,
G. M.; Cushing, D. T.; DeGraffenreid, R. M.; Farthing, N. C.; Hao, X.;
Jaen, C. J.; Jiao, Y. X.; Kopecky, J. D.; Labelle, M.; Lively, E. S.;
McMinn, L. D.; Rasmussen, P. S.; Shin, Y.; Smith, A.; Smith, M. U.S.
Patent 0,097,485, 2004.
(12) Purchased from Manac, Inc..
H
DOI: 10.1021/acs.oprd.5b00207
Org. Process Res. Dev. XXXX, XXX, XXX−XXX