Exploratory Process Development of Pexidartinib through the Tandem Tsuji-Trost Reaction and Heck Coupling

Article abstract of DOI:10.1055/s-0037-1612421

A novel synthetic route to CSF1R inhibitor pexidartinib was designed and demonstrated. Crucial to the successful synthesis is a tandem Tsuji-Trost reaction and Heck coupling in combination with palladium and silver catalysis. The final product was obtained via five steps in 49percent yield with purity as high as 99.2percent. The cheap and available materials and reagents and easy operations for workup and purification make this route more practical.

Full text of DOI:10.1055/s-0037-1612421

SYNTHESIS
0
0
3
9
-
7
8
8
1
1
4
3
7
-
2
1
0
X
©
Georg Thieme Verlag Stuttgart · New York
2019, 51, A–H
paper
A
en
Synthesis
D. Chen et al.
Paper
Exploratory Process Development of Pexidartinib through the
Tandem Tsuji–Trost Reaction and Heck Coupling
Dongsheng Chena,b
Yi Zhanga,b
H
N
Jianqi Li*a,b
Yu Liu*a,b
Tsuji–Trost/Heck
tandem reaction
NH2
N
Cl
OBoc
Cl
Br
N
0
0
0
0-0
0
0
3-2
6
0
8-6
9
4
1
Cl
+
N
N
N
N
NHTs
PMB
N
H
F3C
a
N
N
N
Novel Technology Center of Pharmaceutical
Chemistry, Shanghai Institute of
Ts
Boc
5 steps, 49% overall yield
Pharmaceutical Industry, China State Institute
of Pharmaceutical Industry, 285 Gebaini Road,
Shanghai 201203, P. R. China
Shanghai Engineering Research Center of
Pharmaceutical Process, Shanghai Institute of
Pharmaceutical Industry, China State Institute
of Pharmaceutical Industry, 285 Gebaini Road,
Shanghai 201203, P. R. China
b
lijianqb@126.com
liuyu_tianjin@126.com
Received: 31.01.2019
Accepted after revision: 26.02.2019
Published online: 01.04.2019
Prior to this work, there were only two reported syn-
thetic routes, which were both developed by Daiichi Sankyo
Company for the preparation of pexidartinib (1). In the
3
DOI: 10.1055/s-0037-1612421; Art ID: ss-2019-f0063-op
lead optimization phase, a few hundred milligrams had
been prepared using the first-generation synthesis depicted
in Scheme 1.3a Specific problems with this synthetic route
include the following: (1) The very expensive material 3
was used in the first step, which significantly increased
costs; (2) The introduction of the aldehyde group required
ultra-low temperature and used the quite dangerous re-
agent tert-butyllithium; (3) The yield of the reaction of
fragment 6 and relatively expensive material 7 was quite
low; and (4) Compounds 6, 8, and 1 were purified by col-
umn chromatography. For these reasons, it is not suitable
for a large-scale preparation of 1.
Abstract A novel synthetic route to CSF1R inhibitor pexidartinib was
designed and demonstrated. Crucial to the successful synthesis is a tan-
dem Tsuji–Trost reaction and Heck coupling in combination with palla-
dium and silver catalysis. The final product was obtained via five steps
in 49% yield with purity as high as 99.2%. The cheap and available mate-
rials and reagents and easy operations for workup and purification
make this route more practical.
Key words pexidartinib, palladium-silver-catalyzed, tandem, Tsuji–
Trost reaction, Heck coupling
Tenosynovial giant cell tumor (TGCT) is usually a benign
1
tumor that appears in joints or soft tissues. Because of the
special location of TGCT, the high risk of surgical therapy
makes the damage to the patients greater than the benefit.
Therefore, drug therapy has become a preferred choice.
Pexidartinib (1, Figure 1) is a CSF1R inhibitor developed by
The second-generation route (Scheme 2) provided a
much more concise method for the preparation of 1 with a
3b
high overall yield on a scale of a few hundred grams. How-
ever, the materials 7, 9, and 3 are expensive. Especially,
compound 9 is not easy to prepare and difficult to obtain
commercially. Additionally, corrosive trifluoroacetic acid
was used in large quantities in the last two steps, which in-
creased the requirement for equipment and made the pro-
cess of workup and purification quite complicated.
2
Daiichi Sankyo Company, which effectively inhibits the
binding of CSF1 (colony stimulating factor-1) to CSF1R (col-
ony stimulating factor-1 receptor), a primary growth driver
of abnormal cells in the synovium that cause TGCT. There-
fore, pexidartinib can control the occurrence of CGCT from
the root. At present, the drug has achieved multiple end-
points in the phase III clinical trial of TGCT treatment.
Therefore, a more practical and low-cost synthetic
method is needed to enable the large-scale preparation of
peridartinib. Herein we report our efforts to develop a novel
peridartinib synthesis using a palladium- and silver-cata-
lyzed cross-coupling via Tsuji–Trost/Heck tandem reaction.
In our work, we first performed a retrosynthetic analy-
sis on the target product 1 (Scheme 3). Retrosynthetically, 1
can be accessed from 11 and 3 by reductive amination. The
fragment 11, which has the basic skeleton similar to indole-
H
N
N
Cl
N
H
N
N
F3C
1
3-acetic acid could be synthesized theoretically from moi-
Figure 1 Chemical structure of pexidartinib (1)
eties 12 and 13 through Tsuji–Trost/Heck tandem reaction
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2019, 51, A–H

B
Synthesis
D. Chen et al.
Paper
Br
CHO
n-BuLi (1 equiv)
then t-BuLi (2 equiv)
THF, –78 °C
Br
CF3
N
OHC
CF3
N
CF3CO2H, Et3SiH
toluene, reflux
+
N
N
N
then DMF, –78 °C to rt
N
HN
HN
NH2
CF3
2
3
4
5
Boc
HO
N
OHC
CF3
Cl
N
(
Boc)2O, DMAP
Et3N, DCM, rt
KOH
Cl
N
+
N
N
MeOH, rt
N
N
N
N
N
H
F3C
Boc
H
6
7
8
H
N
N
Cl
CF3CO2H, Et3SiH
CH3CN, reflux
N
N
N
H
F3C
1
Scheme 1 First-generation synthesis of 1
4
developed previously by our group. Furthermore, the in-
termediate 13 could be obtained easily from materials 14
and 15. The advantage of this retrosynthetic approach is
that the core fragment 11 could be prepared from quite
simple starting materials. So, the key to our new route is
whether the construction of 9-azaindole ring of 11 could be
completed in one-pot process via Tsuji–Trost/Heck tandem
1). The transformation of 16 to Grignard reagent was also
failed when using isopropylmagnesium bromide (entry 2).
7
Ultimately, some modifications to the reaction conditions
8
were successfully made as follows: one equivalent of iso-
propyl Grignard reagent was first added to the suspension
of 16 to form a soluble magnesium salt, then two equiva-
lents of n-butyllithium was added at 0 ° for metalization of
16, and finally DMF was added to introduce the aldehyde
group (entry 3). Those modifications avoided the ultra-low
temperature required for the use of n-butyllithium alone.
Subsequently, the intermediate 13 was obtained
smoothly in a one-pot process under mild conditions. How-
ever, the purification of 13 was somewhat difficult due to
its small polarity and low melting point (Scheme 4).
5,6
reaction.
Our next work was to prepare the intermediate 13.
However, the material 15 is a new compound and was diffi-
cult to access commercially, so we had to synthesize it. Ini-
tially, we tried to lithiate the material 16 with n-butyllithi-
um, but this could not be performed smoothly at low tem-
perature because of its poor solubility in THF (Table 1, entry
3a
Et3SiH (5 equiv)
CF3CO2H (9 equiv)
HO
N(Boc)2
Potassium tert-pentylate
Cl
OHC
n-Bu HSO
N
3
CH CN
4
4
Cl
+
N
N
N(Boc)2
i-PrOH, 20–24 h
1%
60 °C, 20–48 h
78%
N
9
N
N
H
H
7
9
10
H
NH2
N
Et3SiH (7.4 equiv)
CF3CO2H (7.6 equiv)
CH3CN
N
N
Cl
OHC
Cl
+
N
80 °C, 20–48 h
N
N
N
N
N
CF3
89%
H
H
F3C
11
3
1
Scheme 2 Second-generation synthesis of 1
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2019, 51, A–H

C
Synthesis
D. Chen et al.
Paper
H
N
NH2
reductive
amination
N
N
OHC
Cl
Cl
+
N
CF3
N
N
N
N
N
H
H
F3C
1
11
3
Tsuji–Trost/Heck
tandem reaction
1
2
Cl
Br
NR R
+
N
N
NHTs
A
PdII
12
0
Pd Ln
OBoc
OHC
MgCl
+
N
NHPMB
PMB
14
15
N
N
Boc
13
Scheme 3 Novel retrosyntheses for the preparation of 1
With intermediates 12 and 13 in hand, we put our focus
on the key step for the construction of 9-azaindole ring of
intermediate 11 through a novel tandem transition-metal-
catalyzed coupling (Table 2). At first, the reaction was per-
formed under the conditions developed previously by our
group as follows: Pd(OAc) (5 mol%), P(o-tol) (10 mol%),
2
3
and DIPEA (2.0 equiv) in DMA at 100 ° for 12 hours. Excit-
ingly, the reaction could smoothly generate a large amount
of Tsuji–Trost product 17 and a small amount of target
product 18, which was the Ts-protected compound 11. Al-
though 17 was not the desired product, it could be further
converted to 18 via Heck coupling. Next, we tried to in-
crease the conversion rate for 18 by prolonging the reaction
time and increasing the reaction temperature, but the re-
sults were not satisfactory (Table 2, entries 1, 2).
Table 1 Optimization for the Preparation of Compound 15
O
Br
Reaction conditions
H
PMB
PMB
N
N
H
N
N
H
Observing the structure of 17 and analyzing the bond
formation process of Heck coupling (Scheme 5), we envi-
sioned that the formation of 17c is the key to the conver-
16
15
9
Entry
1
Reaction conditions
Results
sion of 17 to 18. In terms of mechanism, the formation of
17c needs to undergo the stages of oxidative addition,
n-BuLi (1 equiv), THF, 0 °C;
n-BuLi (1 equiv), –78 °C;
DMF (2.8 equiv), –78 °C → rt
No reaction
alkene coordination, and migratory insertion. Therefore, in
order to perform the Heck coupling smoothly, it is neces-
sary to increase the activity of palladium(0) in the oxidative
addition stage or the activity of palladium(II) in the alkene
coordination and migratory insertion stages. In theory, the
activities of palladium(0) and palladium(II) are affected by
2
3
i-PrMgBr (2.2 equiv), THF, rt → 50 °C;
DMF (2.8 equiv), 0 °C → rt
No reaction
i-PrMgCl (1 equiv), THF, 0 °C → rt, 0.5 h;
n-BuLi (2 equiv), 0 °C, 0.5 h;
Yield: 73%
HPLC: 99.8%
DMF (2.8 equiv), 0 °C → rt, 1 h
O
O
OBoc
14 (2 equiv), THF
then Boc2O (2.1 equiv)
H
PMB
PMB
PMB
N
N
15 °C to rt, 0.5 h
N
N
0 °C, 1 h
N
N
H
Boc
15
13
Scheme 4 One-pot synthesis of intermediate 13
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2019, 51, A–H

D
Synthesis
D. Chen et al.
Paper
factors such as catalysts, ligands, solvents, and additives. To
make minimum changes to the reaction conditions, we en-
visaged optimizing the reaction conditions by screening
When the solvent DMA was replaced by NMP, the reac-
tion did not improve at all (Table 2, entry 3). After replacing
DMA and NMP with DMSO as solvent,10 the reaction was
obviously accelerated, but the yield of 18 was still low (en-
try 4).
solvents and additives while maintaining the Pd(OAc) /P(o-
2
tol) /DIPEA catalytic system.
3
Table 2 Summary of Tsuji–Trost/Heck Tandem Reaction Conditionsa
PMB
Boc
N
N
PMB
OBoc
Pd(OAc)2 (0.05 equiv)
N
Cl
Br
P(o-tol)3 (0.1 equiv)
Boc
N
+
Cl
Br
+
Cl
PMB
N
NHTs
N
N
DIPEA (2.0 equiv)
additive, solvent
N
Boc
N
N
N
Ts
Ts
12
13
17
18
b
b
Entry
Additive
Solvent
Temp (°C)
Time (h)
Yield (%) 17
Yield (%) 18
1
2
3
4
5
6
7
8
9
–
–
–
–
–
DMA
100
24
24
24
24
24
24
12
12
12
12
12
12
24
24
36
65
60
59
63
64
57
–
13
17
15
14
16
21
70
26
58
67
64
68
69
56
57
DMA
120
NMP
120
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
120
70–120
70–120
70–100
70–100
70–100
70–100
70–100
70–100
70–100
70–100
70–100
n-Bu
Ag CO
AgOAc (2.0 equiv)c
4
NBr (1.0 equiv)
(1.0 equiv)c
2
3
49
–
Ag
Ag
Ag
Ag
Ag
Ag
Ag
3
2
2
2
2
2
2
PO
4
(0.67 equiv)c
(0.8 equiv)c
(0.6 equiv)c
(0.8 equiv)
(1.0 equiv)
(1.0 equiv)
(0.8 equiv)
1
1
1
1
1
0
CO
CO
CO
CO
CO
CO
3
3
3
3
3
3
–
1
–
2
–
3d
4e
–
–
1 f
5
–
a
b
c
d
e
f
2
Reaction conditions: 12 (1.0 mmol), 13 (1.2 mmol), c = 0.2 M, N atmosphere.
Isolated yields.
Silver salts were added after the completion of Tsuji–Trost reaction.
Reaction scale: 12 (25 g, 99.1% HPLC, 68 mmol), 13 (40 g, 96.3% HPLC, 82 mmol).
Reaction scale: 12 (22 g, 99.1% HPLC, 60 mmol), 13 (40 g, 86.9% HPLC, 74 mmol).
Reaction scale: 12 (112 g, 99.1% HPLC, 307 mmol), 13 (218 g, 86.9% HPLC, 403 mmol).
PMB
Boc
N
PMB
Boc
N
PMB
Boc
N
N
N
N
PMB
PdII
N
Br
PdII
PdII
Boc
Pd0
N
Cl
Br
Cl
Cl
Cl
oxidative
addition
alkene
cordination
migratory
insertion
N
N
N
N
N
N
N
N
Ts
Ts
Ts
Ts
17
17a
17b
17c
Scheme 5 The bond formation process of Heck coupling
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2019, 51, A–H

E
Synthesis
D. Chen et al.
Paper
During the experiment we found that a lower tempera-
ture was favorable for the Tsuji–Trost reaction. So the trans-
formation was carried out at 70 °C for 12 hours and then
with 86.9% purity. The difference between those two reac-
tions was the purity of intermediate 13. The higher the pu-
rity of 13, the better the yield of 18. It seems that impurities
inhibited the catalytic process. The reaction could give a
similar result as entry 14 by increasing the scale of 13 (en-
try 15). It is worth mentioning that compound 18 has poor
solubility, which makes 18 easy to precipitate from the re-
action system, and it could be furnished with high purity
after slurring sequentially with dimethyl sulfoxide and eth-
yl acetate.
120 °C for another 12 hours (Table 2, entry 5). As a result,
there was hardly any improvement in the overall yield of 17
and 18, as well as for the transformation to 18. Based on en-
try 5, we attempted to increase the stability of palladium(0)
by adding phase transfer catalyst such as n-Bu NBr to in-
4
11
crease its catalytic activity, but the effect was not signifi-
cant (entry 6).
According to the literature,12 the use of silver salt could
convert the neutral palladium(II) generated in the stage of
oxidative addition into the more active cationic palladi-
um(II), which is beneficial to the alkene coordination and
migratory insertion of nucleophilic olefin bonds, by precip-
itating halogen ions. Excitingly, the yield of the desired
product 18 was significantly improved as expected when
silver salts such as AgOAc, Ag CO , and Ag PO were added
Deprotection of 18 was completed with a one-pot oper-
ation (Scheme 6).13 The protecting group Ts was first re-
moved under the condition of KOH/MeOH at 70 °C for 4
hours, PMB and Boc were then removed with concentrated
hydrochloric acid at the same temperature for another 8
hours after distillation of methanol.14 Thus, after crystalli-
zation from toluene, compound 19 was smoothly obtained
in 91% yield and 98.9% HPLC purity.
2
3
3
4
to the reaction system after the completion of the Tsuji–
Trost reaction at 70 °C (Table 2, entries 7, 8, and 9). In par-
ticular, the presence of Ag CO made the transformation of
Ultimately, the target product 1 was achieved in 88%
yield and 99.2% HPLC purity under the conditions reported
by Daiichi Sankyo Company (Scheme 6).
2
3
1
7 more complete and the yield of 18 was up to 70%. Fur-
In conclusion, we have successfully developed a novel
efficient synthetic route for pexidartinib (1) via Tsuji–
Trost/Heck tandem reaction in combination with palladium
and silver catalysis. The new synthesis avoids the need for
ultra-low temperature and greatly reduces the amount of
corrosive trifluoroacetic acid. The cheap and easily available
materials and reagents and easy operations for workup and
purification make this route more practical. Currently, we
are focusing on establishing a scalable manufacturing pro-
cess. Further modifications are carried out to improve and
stabilize the yield of the tandem reaction through improv-
ing the purity of intermediate 13 and reducing the amount
of catalysts. These efforts will be discussed in a future pa-
per.
thermore, the investigation of the amount of Ag CO used
2
3
showed that 0.8–1.0 equivalent was preferable (entries 10,
1). And the same result as entry 10 could be achieved by
adding Ag CO before the Tsuji–Trost reaction (entry 12).
1
2
3
Finally, the optimal reaction conditions were determined as
follows: 12 (1.0 equiv), 13 (1.2 equiv), Pd(OAc) (5 mol%),
P(o-tol) (10 mol%), Ag CO (0.8–1.0 equiv), and DIPEA (2.0
equiv) in DMSO.
Based on the optimized reaction conditions, the scale
application of this tandem reaction was then examined.
The reaction gave a 69% isolated yield under the scale of 40
g of 13 (Table 2, entry 13) with 96.3% purity, while it gave a
2
3
2
3
56% isolated yield under the scale of 40 g of 13 (entry 14)
PMB
1
2 (0.83 equiv)
O
OBoc
N
1
4 (2 equiv)
Pd(OAc)2 (0.05 equiv)
P(o-tol)3 (0.1 equiv)
Boc
THF, 150 °C to rt, 5 h
N
H
Cl
PMB
PMB
then Boc2O (2.1 equiv)
Ag CO (0.8–1.0 equiv)
DIPEA (2.0 equiv)
DMSO, 70–100 °C, 24 h
N
N
N
N
2
3
N
0
°C, 1 h
N
H
Boc
Ts
15
13
18
89%
69%
H
NH2
N
3
(1.8 equiv), Et3SiH (6.0 equiv)
CF CO H (6.0 equiv)
N
N
KOH/MeOH, 70 °C, 4 h
Cl
3
2
Cl
then HCl (con.), 70 °C, 8 h
3
CH CN, 65 °C, 36 h
N
N
N
N
N
H
H
F3C
91%
88%
19
1
4
9% overall yield from 15
Scheme 6 The novel synthetic route to pexidartinib (1)
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2019, 51, A–H

F
Synthesis
D. Chen et al.
Paper
tert-Butyl (5-{1-[(tert-Butoxycarbonyl)oxy]allyl}pyridin-2-yl)(4-
methoxybenzyl)carbamate (13)
All commercially available chemicals and solvents were used directly
1
as received. H NMR spectra were recorded on a Bruker Avance 400
To a four-necked flask charged with compound 15 (29 g, 0.12 mol)
and anhyd THF (0.12 L) was added slowly vinylmagnesium chloride
(14; 1.0 M, 0.24 L, 0.24 mol) under N atmosphere at 10–20 °C. After
completion of the addition of compound 15, the mixture was cooled
13
and 600 MHz NMR spectrometer. C NMR spectra were recorded on a
Bruker Avance 100 and 150 MHz NMR spectrometer. Chemical shifts
2
1
13
are reported in ppm () (CDCl  = 7.26 for H NMR and 77.1 for
C
3
1
13
NMR; DMSO-d  = 2.50 for H NMR and 39.5 for C NMR). Standard
to 0 °C, and a solution of Boc O (55 g, 0.252 mol) in anhyd THF (0.05 L)
6
2
abbreviations were used to denote proton coupling patterns. HRMS
spectra were measured on a Waters Xevo G2-XS QTof. Reactions were
monitored by TLC or reversed-phase HPLC (Dionex U3000) under the
following conditions: Column: Ultimate XB-C18 column (5 m, 250 ×
was added slowly below 10 °C, after which the reaction mixture was
stirred for 2 h below 10 °C and monitored by TLC until the reaction
was complete. The mixture was poured into sat. aq NH Cl (0.6 L) at
4
0 °C. After stirring for 20 min, the mixture was added to EtOAc (0.2 L),
and filtered through a pad of kieselgur. The layers were separated and
the organic layer was washed with brine (2 × 0.1 L), concentrated in
vacuo, and purified by silica gel column chromatography (200–300
mesh, EtOAc/hexane 1:20) to give a colorless oil with 96.3% HPLC pu-
rity, which turned to a hard white solid on keeping for 24 h; yield: 50
4
.6 mm). Eluent: (A) 1% H PO /99% H O; (B) 100% CH CN. The flow
3 4 2 3
rate was 1.0 mL/min with a column temperature of 35 °C. The detec-
tion wavelength was 220 nm. The HPLC gradient method is shown in
Table 3.
Table 3 HPLC Gradient Method
g (89%); mp 88.8–89.2 °C; HPLC: t = 19.7 min.
R
1
H NMR (400 MHz, DMSO-d ):  = 8.37 (d, J = 2.0 Hz, 1 H), 7.75 (dd, J =
6
Time (min)
0
Eluent A (%)
Eluent B (%)
8
.4, 2.4 Hz, 1 H), 7.63 (d, J = 8.8 Hz, 1 H), 7.14–7.16 (m, 2 H), 6.83–8.85
(m, 2 H), 6.03–6.12 (m, 2 H), 5.26–5.34 (m, 2 H), 5.03 (s, 2 H), 3.70 (s,
90
50
5
10
50
95
10
3
H), 1.40 (s, 9 H), 1.39 (s, 9 H).
1
2
3
0
13
C NMR (100 MHz, DMSO-d ):  = 158.6, 154.3, 153.8, 152.3, 146.6,
6
0–30
5–40
136.5, 136.2, 131.3 (2 C), 130.4, 128.7, 119.6, 117.9 (2 C), 114.0, 82.5,
81.4, 76.6, 55.4, 49.1, 28.2 (3 C), 27.8 (3 C).
90
+
HRMS (ESITOF): m/z [M + H] calcd for C26H35N O : 471.2490; found:
2
6
471.2503.
6-[(4-Methoxybenzyl)amino]nicotinaldehyde (15)
To a four-necked flask charged with compound 1615 (500 g, 1.71 mol)
and anhyd THF (2 L) was added slowly isopropylmagnesium chloride
Larger scale: To a four-necked flask charged with compound 15 (290
g, 1.2 mol) and anhyd THF (1.2 L) was added slowly vinylmagnesium
chloride (1.0 M, 2.4 L, 2.4 mol) under N atmosphere at 10–20 °C. Af-
(2.0 M, 0.94 L, 1.88 mol) under N2 atmosphere below 20 °C, after
2
ter completion of the addition of compound 15, the mixture was
which the mixture was stirred for 0.5 h, and then n-BuLi (2.5 M solu-
tion in hexanes, 1.37 L, 2.42 mol) was subsequently added below
cooled to 0 °C, and a solution of Boc O (550 g, 2.52 mol) in anhyd THF
2
(0.5 L) was added slowly below 10 °C, after which the reaction mix-
20 °C. After the disappearance of compound 16, the mixture was
ture was stirred for 2 h below 10 °C and monitored by TLC until the
cooled to 0 °C, and anhyd DMF (350 g, 4.8 mol) was added slowly be-
low 20 °C, after which the reaction mixture was stirred for 2 h below
reaction was complete. The mixture was poured into sat. aq NH Cl (6
4
L) at 0 °C. After stirring for 20 min, the mixture was added to EtOAc (2
L) and filtered through a pad of kieselgur. The layers were separated
and the organic layer was concentrated in vacuo to a volume of about
20 °C and monitored by TLC until the reaction was complete. The
mixture was poured into sat. aq NH Cl (4 L) at 0 °C. After stirring for
4
20 min, the mixture was added to EtOAc (4 L) and filtered through a
1.2 L. To the concentrate was added silica gel (200–300 mesh, 350 g)
pad of kieselgur. The layers were separated and the organic layer was
concentrated in vacuo to a volume of about 4 L. The concentrate was
and stirred for 10 min at rt, filtered through a pad of kieselgur, which
was washed with a mixed solution of EtOAc and PE (1 L/1 L). The fil-
trate was concentrated in vacuo to a volume of about 1.2 L. To the
concentrate was added again silica gel (200–300 mesh, 350 g) and
stirred for 10 min at rt, filtered through a pad of kieselgur, which was
washed with a mixed solution of EtOAc and PE (1 L/1 L). The filtered
solution was concentrated in vacuo until no solvent was left. Cyclo-
hexane (2 L) was then added to the residue over 10 min while stirring
at rt, at which time a grayish yellow solid was precipitated. After stir-
ring for another 20 min, the solution was filtered through a pad of
added to DMF (0.1 L) and H O (2 L) and stirred at 50 °C until complete
2
dissolution, after which NaHSO (355 g, 3.42 mol) was added. The re-
3
action mixture was stirred for another 5 h at 50 °C, cooled to rt, and
filtered. The filter cake was washed subsequently with EtOAc (1 L)
and H O (1 L), and transferred to a mixture of EtOAc (4 L) and H O (2
2
2
L). To this suspension was added NH OAc (660 g, 8.5 mol), heated to
4
70 °C and stirred for 12 h, then cooed to rt. The layers were separated
and the organic layer was washed with brine (2 × 2 L), concentrated in
vacuo, and then dried to give a light yellow solid with 99.7% HPLC pu-
kieselgur and washed with brine (2 × 1 L), dried (MgSO ), filtered, and
rity; yield: 302 g (73%); mp 133.6–134.5 °C; HPLC: t = 10.2 min.
4
R
concentrated in vacuo until dryness to provide a yellow oil with 86.9%
HPLC purity, which was used directly in the next step without further
purification; yield: 477 g (85%).
1
H NMR (400 MHz, DMSO-d ):  = 9.67 (s, 1 H), 8.51 (d, J = 2.0 Hz, 1
6
H), 8.15 (t, J = 5.8 Hz, 1 H), 7.75 (dd, J = 8.8, 2.4 Hz, 1 H), 7.26 (d, J = 8.4
Hz, 2 H), 6.89 (d, J = 8.8 Hz, 2 H), 6.60 (d, J = 9.2 Hz, 1 H), 4.52 (d, J = 4.0
Hz, 2 H), 3.73 (s, 3 H).
tert-Butyl {5-[(5-Chloro-1-tosyl-1H-pyrrolo[2,3-b]pyridin-3-
yl)methyl]pyridin-2-yl}(4-methoxybenzyl)carbamate (18) (Table
13
C NMR (150 MHz, DMSO-d ):  = 189.6, 161.9, 158.8, 155.7, 135.0,
6
131.6, 129.2, 122.6, 114.0, 110.0, 55.5, 44.0.
2, entry 13)
+
HRMS (ESITOF): m/z [M + H] calcd for C 4H N O : 243.1128; found:
1
15
2
2
To a three-necked flask charged with DMSO (170 mL) was added com-
243.1126.
16
pound 12 (25 g, 99.1% HPLC purity, 68 mmol), compound 13 (40 g,
96.3% HPLC purity, 82 mmol), Pd(OAc)2 (0.76 g, 3.4 mmol), P(o-tol)3
(
2.1 g, 6.8 mmol), Ag CO (18.8 g, 68 mmol), and DIPEA (17.6 g, 136
2
3
mmol). The mixture was stirred under N2 atmosphere at 70 °C for
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2019, 51, A–H

G
Synthesis
D. Chen et al.
Paper
about 10 h until compound 12 had disappeared, and then heated to
00 °C. After stirring for another 14 h, silica gel (200–300 mesh, 25 g)
5-[(5-Chloro-1H-pyrrolo[2,3-b]pyridin-3-yl)methyl]pyridin-2-
amine (19)
1
was added. The reaction mixture was stirred for 20 min and filtered
immediately through a pad of kieselgur, which was washed with hot
DMSO (100 °C, 30 mL). The filtered solution was cooled at ambient
conditions to about 20–24 °C while stirring, at which time a yellow
solid precipitated. The precipitate was filtered and DCM (200 mL) was
A two-necked flask fitted with a condenser was charged with com-
pound 18 (95 g, 0.15 mol) and MeOH (0.24 L). To this mixture was
added KOH (27 g, 0.48 mol) and the flask was heated in an oil bath at
70 °C and the reaction mixture was stirred for 4 h until compound 18
had disappeared. The solvent was evaporated and to the residue was
added aq 12 M HCl (0.24 L). The mixture was stirred for 8 h at 70 °C
until the reaction was completed, cooled to about 20–24 °C, and add-
added to the filtrate. The DCM layer was washed with H O ( 3 × 100
2
mL), and concentrated in vacuo to give a gray solid, which was com-
bined with the above precipitate and treated with DMSO (50 mL). The
slurry was heated to 80 °C and stirred for about 1 h. The mixture was
cooled at ambient conditions to about 20–24 °C while stirring. The
precipitate was filtered, washed with a mixture of DMSO and EtOAc
ed to H O (0.24 L). The mixture was extracted with EtOAc (2 × 0.24 L)
2
and the aqueous layer was adjusted pH to 8–9 by adding K CO under
2
3
cooling in an ice bath. The precipitate formed was filtered, washed
with H O (0.12 L), and then transferred to a mixture of MeOH and tol-
2
(25 mL/25 mL) at 0 °C. The filter cake was slurried in EtOAc (50 mL) at
uene (0.05 L/0.1 L). The suspension was heated to 70 °C until complete
dissolution and then MeOH was evaporated off to leave about 45–48
mL. The residual solution was cooled under ambient conditions to
80 °C for about 1 h. Then the mixture was cooled to about 0–5 °C
while stirring. The precipitate was filtered, washed with EtOAc (50
mL) at 0 °C, and dried to provide a bright yellow powder with 99.1%
HPLC purity; yield: 29.7 g (69%); mp 204.3–205.8 °C; HPLC: tR = 12.9
min.
20–24 °C while stirring, at which time a light yellow solid was precip-
itated. The precipitate was filtered, washed with toluene and hexane
in sequence, and dried to provide a light yellow powder with 98.9%
1
H NMR (400 MHz, DMSO-d ):  = 8.44 (d, J = 1.6 Hz, 1 H), 8.18 (dd, J =
HPLC purity; yield: 35 g (91%); mp 212.4–213.5 °C; HPLC: t = 8.317
R
6
16.0, 2.4 Hz, 2 H), 7.98 (d, J = 8.0 Hz, 2 H), 7.80–7.73 (m, 2 H), 7.38 (d,
min.
J = 8.0 Hz, 2 H), 7.26 (t, J = 3.0 Hz, 1 H), 7.19 (d, J = 8.4 Hz, 2 H), 6.87 (d,
J = 8.8 Hz, 2 H), 5.21 (d, J = 2.8 Hz, 2 H), 5.09 (s, 2 H), 3.71 (s, 3 H), 2.36
1
H NMR (600 MHz, DMSO-d ):  = 11.64 (s, 1 H), 8.16 (d, J = 1.8 Hz, 1
6
H), 7.91 (dd, J = 18.6, 1.8 Hz, 2 H), 7.36 (s, 1 H), 7.26 (dd, J = 8.4, 1.8 Hz,
1
(s, 3 H), 1.42 (s, 9 H).
H), 6.37 (d, J = 8.4 Hz, 1 H), 5.69 (s, 2 H), 3.82 (s, 2 H).
13
C NMR (150 MHz, CDCl ):  = 158.6, 154.7, 154.1, 153.6, 147.7,
13
3
C NMR (150 MHz, DMSO-d ):  = 158.6, 147.5, 147.4, 147.8, 130.8,
6
1
1
46.9, 144.7, 135.9, 135.1, 131.2, 129.7, 128.6, 127.9, 127.8, 126.7,
26.2, 125.4, 118.9, 118.7, 113.6, 81.9, 55.2, 53.9, 49.1, 28.2, 21.6.
1
26.3, 126.0, 124.3, 122.2, 120.5, 113.8, 108.3, 27.6.
+
HRMS (ESITOF): m/z [M + H] calcd for C13H12ClN : 259.0745; found:
259.0743.
4
HRMS (ESITOF): m/z [M + H]+ calcd for C33H34ClN O S: 633.1933;
4
5
found: 633.1938.
5
-[(5-Chloro-1H-pyrrolo[2,3-b]pyridin-3-yl)methyl]-N-{[6-(triflu-
tert-Butyl {5-[(5-Chloro-1-tosyl-1H-pyrrolo[2,3-b]pyridin-3-
yl)methyl]pyridin-2-yl}(4-methoxybenzyl)carbamate (18) (Table
oromethyl)pyridin-3-yl]methyl}pyridin-2-amine (Pexidartinib, 1)
The process of preparation for 1 starting from 19 (26 g) was the same
as given in the literature. The reaction provided a light yellow pow-
der with 99.2% HPLC purity; yield: 37 g (88%); mp 197.1–198.8 °C;
HPLC: t = 11.3 min.
2, entry 15)
3b
To a three-necked flask charged with DMSO (0.8 L) was added com-
pound 12 (112 g, 99.1% HPLC purity, 0.31 mol), compound 13 (218 g,
R
86.9% HPLC purity, 0.4 mol), Pd(OAc)2 (2.8 g, 12.4 mmol), P(o-tol)3
1
H NMR (400 MHz, DMSO-d ):  = 11.65 (s, 1 H), 8.70 (s, 1 H), 8.15 (d,
6
(
7.5 g, 24.8 mmol), Ag CO (68 g, 0.25 mol), and DIPEA (80 g, 0.62
2
3
J = 2.0 Hz, 1 H), 7.95–7.94 (m, 3 H), 7.83 (d, J = 8.0 Hz, 1 H), 7.36 (d, J =
mol). The mixture was stirred under N atmosphere at 70 °C for about
2
1.6 Hz, 1 H), 7.31 (dd, J = 8.4, 2.4 Hz, 1 H), 7.10 (t, J = 6.0 Hz, 1 H), 6.48
12 h until compound 12 had disappeared, and then heated to 100 °C.
(d, J = 8.8 Hz, 1 H), 4.56 (d, J = 6.0 Hz, 2 H), 3.82 (s, 2 H).
After stirring for another 24 h, silica gel (200–300 mesh, 110 g) was
added. The reaction mixture was stirred for 20 min and filtered im-
mediately through a pad of kieselgur, which was washed with hot
DMSO (100 °C, 0.3 L). The filtered solution was cooled under ambient
conditions to about 20–24 °C while stirring, at which time a yellow
solid was precipitated. The precipitate was filtered and the filtrate
13
C NMR (150 MHz, DMSO-d ):  = 157.2, 149.8, 147.4, 147.1, 145.2,
6
1
1
41.2, 140.9, 137.8, 137.1, 126.3, 126.0, 124.9, 123.2, 122.3, 120.8,
20.4, 113.7, 108.8, 42.1, 27.5.
HRMS (ESITOF): m/z [M + H]+ calcd for C20H16ClF N : 418.1041;
found: 418.1053.
3
5
was added to DCM (1 L). The DCM layer was washed with H O (3 × 0.6
2
L), concentrated in vacuo to give a gray solid, which was combined
with the above precipitate and was added to DMSO (0.3 L). The slurry
was heated to 80 °C and stirred for about 1 h. The mixture was cooled
under ambient conditions to about 20–24 °C while stirring. The pre-
cipitate was filtered, washed with a mixed solution of DMSO and
EtOAc (0.15 L/0.15 L) at 0 °C. The filter cake was slurried in EtOAc (0.3
L) at 80 °C for about 1 h. Then the mixture was cooled to about 0–5 °C
while stirring. The precipitate was filtered, washed with EtOAc (0.3 L)
at 0 °C, and dried to provide a grayish yellow powder with 96.7% HPLC
purity; yield: 112 g (57%).
Funding Information
National Natural Science Foundation of China (81502930).
N
ati
o
n
a
l Natural S
c
i
e
n
c
e
F
o
u
n
d
ati
o
n
of
C
h
i
n
a
(8
1
5
0
2
9
3
0)
Acknowledgment
We thank Mr. Kai Yao for the HPLC support.
Supporting Information
Supporting information for this article is available online at
https://doi.org/10.1055/s-0037-1612421.
S
u
p
p
orit
n
g Inform ati
o
n
S
u
p
p
orit
n
g Inform ati
o
n
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2019, 51, A–H

H
Synthesis
D. Chen et al.
Paper
References
Hirao, H.; Zhou, J. S. Angew. Chem. Int. Ed. 2015, 54, 6531.
(e) Pires, M. J. D.; Poeira, D. L.; Purificação, S. I.; Marques, M. M.
(
(
(
1) (a) Bertoni, F.; Unni, K. K.; Beabout, J. W.; Sim, F. H. Am. J. Surg.
B. Org. Lett. 2016, 18, 3250.
Pathol. 1997, 21, 153. (b) West, R. B.; Rubin, B. P.; Miller, M. A.;
Subramanian, S.; Kaygusuz, G.; Montgomery, K.; Zhu, S.;
Marinelli, R. J.; De Luca, A.; Downs-Kelly, E.; Goldblum, J. R.;
Corless, C. L.; Brown, P. O.; Gilks, C. B.; Nielsen, T. O.; Huntsman,
D.; van de Rijn, M. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 690.
2) (a) Giustini, N.; Bernthal, N. M.; Bukata, S. V.; Singh, A. S. Clin.
Sarcoma. Res. 2018, 8, 14. (b) Tap, W. D.; Gelderblom, H.;
Stacchiotti, S.; Palmerini, E.; Ferrari, S.; Desai, J.; Bauer, S.; Blay,
J.; Alcindor, T.; Ganjoo, K. N.; Broto, J. M.; Ryan, C. W.; Shuster,
D. E.; Zhang, L.; Wang, Q.; Hsu, H.; Lin, P. S.; Tong, S.; Wagner, A.
J. J. Clin. Oncol. 2018, 11502.
(
(
(
7) (a) D’Errico, S.; Oliviero, G.; Borbone, N.; Nici, F.; Piccialli, V.;
Pinto, B.; D’Alonzo, D.; Mayol, L.; Piccialli, G. Eur. J. Org. Chem.
2015, 2244. (b) Wang, Y.; Przyuski, K.; Roemmele, R. C.;
Hudkins, R. L.; Bakale, R. P. Org. Process Res. Dev. 2013, 17, 846.
8) (a) Xu, J.; Jain, N.; Sui, Z. Tetrahedron Lett. 2004, 45, 6399.
(
b) Struk, L.; Sosnicki, J. G. Synthesis 2012, 44, 735. (c) Tian, Q.;
Shang, S.; Wang, H.; Shi, G.; Li, Z.; Yuan, J. Molecules 2017, 22,
952.
1
9) (a) Crisp, G. T. Chem. Soc. Rev. 1998, 27, 427. (b) Amatore, C.;
Jutand, A. Acc. Chem. Res. 2000, 33, 314.
(
10) Tanaka, D.; Romeril, S. P.; Myers, A. G. J. Am. Chem. Soc. 2005,
27, 10323.
3) (a) Zhang, C.; Zhang, J.; Ibrahim, P. N.; Artis, D. R.; Bremer, R.;
Wu, G.; Zhu, H.; Nespi, M. PCT Int. Appl WO 2008064255 A2
1
(
11) (a) Beller, M.; Fischer, H.; Kühlein, K.; Reisinger, C.-P.;
Herrmann, W. J. J. Organomet. Chem. 1996, 520, 257. (b) Reetz,
M. T.; Maase, M. Adv. Mater. 1999, 11, 773. (c) Reetz, M. T.;
Westermann, E. Angew. Chem. Int. Ed. 2000, 39, 165.
20080529, 2008. (b) Ibrahim, P. N.; Jin, M.; Matsuura, S. PCT Int.
Appl WO 2016179412 A1 20161110, 2016.
(4) Chen, D.; Chen, Y.; Ma, Z.; Zou, L.; Li, J.; Liu, Y. J. Org. Chem. 2018,
83, 6805.
(
12) (a) Beletskaya, I. P.; Cheprakov, A. V. Chem. Rev. 2000, 100, 3009.
(5) (a) Tsuji, J.; Takahashi, H.; Morikawa, M. Tetrahedron Lett. 1965,
(b) Jeong, S.; Chen, X.; Harran, P. G. J. Org. Chem. 1998, 63, 8640.
4
2
387. (b) Trost, B. M.; Fullerton, T. J. J. Am. Chem. Soc. 1973, 95,
92. (c) Frost, C. G.; Howarth, J.; Williams, J. M. Tetrahedron:
(
13) Zhou, Q.; Phoa, A. F.; Abbassi, R. H.; Hoque, M.; Reekie, T. A.;
Font, J. S.; Ryan, R. M.; Stringer, B. W.; Day, B. W.; Johns, T. G.;
Munoz, L.; Kassiou, M. J. Med. Chem. 2017, 60, 2052.
Asymmetry 1992, 3, 1089. (d) Kljajic, M.; Puschnig, J. G.; Weber,
H.; Breinbauer, R. Org. Lett. 2017, 19, 126. (e) Braun, J.; Ariens,
M. I.; Matsuo, B. T.; de Vries, S.; van Wordragen, E. D. H.;
Ellenbroek, B. D.; Van de Velde, C. M. L.; Orru, R. V. A.; Ruijter, E.
Org. Lett. 2018, 20, 6611.
(
(
(
14) Xing, L.; Petitjean, A.; Schmidt, R.; Cuccia, L. Synth. Commun.
2007, 37, 2349.
15) The preparation of compound 16 is described in the Supporting
Information.
(6) (a) Mizoroki, T.; Mori, K.; Ozaki, A. Bull. Chem. Soc. Jpn. 1971, 44,
16) The preparation of compound 12 is described in the Supporting
5
(
81. (b) Heck, R. F.; Nolley, J. P. J. Org. Chem. 1972, 37, 2320.
c) Nallapati, S. B.; Adepu, R.; Ashfaq, M. A.; Sreenivas, B. Y.;
Mukkanti, K.; Pal, M. RSC Adv. 2015, 5, 88686. (d) Yue, G.; Lei, K.;
Information.
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2019, 51, A–H