Improved Manufacturing Route and Polymorphic Control of a Potent and Selective Anaplastic Lymphoma Kinase (ALK) Inhibitor ASP3026

Article abstract of DOI:10.1021/acs.oprd.8b00427

Our effort toward the process improvement of anaplastic lymphoma kinase (ALK) inhibitor ASP3026 (1) is described. A cost-effective and practical synthesis of 1 was accomplished as a result of the change of starting material from 2,4-dichloro-1,3,5-triazine (6) to cyanuric chloride (9) and late-stage introduction of a highly reactive N-methyl piperazine moiety by reductive amination of intermediate ketone 13. The modified process avoided the challenges with the original synthesis and furnished the several hundred kilograms of high-quality API with high economic efficiency, operability, and reproducibility. Furthermore, a sequence of investigation of polymorphic control in the second-generation synthetic route to obtain the thermodynamically desired, most stable polymorph Form A04 is also discussed.

Full text of DOI:10.1021/acs.oprd.8b00427

Article
pubs.acs.org/OPRD
Cite This: Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Improved Manufacturing Route and Polymorphic Control of a
Potent and Selective Anaplastic Lymphoma Kinase (ALK) Inhibitor
ASP3026
Yuji Takahama,* Kazuyoshi Obitsu, Kazuhiro Takeguchi, Shun Hirasawa, Koji Kobayashi,
Takahiro Akiba, Norihiro Ueda, Ryoki Orii, Atsushi Ohigashi, Takumi Takahashi, Minoru Okada,
and Shigeru Ieda
Process Chemistry Laboratories, Astellas Pharma Inc., 160-2 Akahama, Takahagi-shi, Ibaraki 318-0001, Japan
S
* Supporting Information
ABSTRACT: Our effort toward the process improvement of anaplastic lymphoma kinase (ALK) inhibitor ASP3026 (1) is
described. A cost-effective and practical synthesis of 1 was accomplished as a result of the change of starting material from 2,4-
dichloro-1,3,5-triazine (6) to cyanuric chloride (9) and late-stage introduction of a highly reactive N-methyl piperazine moiety
by reductive amination of intermediate ketone 13. The modified process avoided the challenges with the original synthesis and
furnished the several hundred kilograms of high-quality API with high economic efficiency, operability, and reproducibility.
Furthermore, a sequence of investigation of polymorphic control in the second-generation synthetic route to obtain the
thermodynamically desired, most stable polymorph Form A04 is also discussed.
KEYWORDS: ALK inhibitor, SNAr reaction, 1,3,5-triazine, polymorph
the presence of DBU and by following hydrogenation of the
nitro group utilizing Pd/C catalyst and salt formation with
trifluoroacetic acid, which gave the best result in the
subsequent SNAr reaction of 5 and intermediate 8, furnishing
aniline 5·TFA. Aside from the synthetic stream of intermediate
5·TFA, 1,3,5-triazine derivative 8 was obtained by SNAr
reaction between 6 and aniline 7.
INTRODUCTION
■
Anaplastic lymphoma kinase (ALK) is a validated therapeutic
target for treating echinoderm microtubule-associated protein-
like 4 (EML4)-ALK positive non-small cell lung cancer
(NSCLC).1−5 ASP3026 (1) (Figure 1) is a potent and
Various conditions of consecutive substitution reactions of 8
and 5 were examined (Table 1). First, a series of protic acids
were screened (entries 1−4). While MsOH and TFA gave 1 in
a high HPLC ratio (entries 1,2), the use of formic acid and
acetic acid led to poor conversion of 5 (entries 3,4).
Eventually, TFA was selected as a suitable acid considering a
possible generation of genotoxic alkyl mesylates in the MsOH
condition. Also, the conditions of Lewis acids gave 1 in
moderate to low yield (entries 5−9), and screening of solvent
with TFA revealed that various solvents are applicable (entries
10−13). With respect to the stoichiometry of TFA, it was
found that 3 equiv. is the best condition (entries 14−18). Also,
we examined the effects of temperature (entry 19) and H2O
(entry 20), though both did not provide significant change.
Although we learned that 2-PrOH is the best solvent, we finally
chose 2-butanone as the reaction solvent (entry 13). A reaction
with 8 and 5·TFA had to be conducted continuously from a
reaction of 6 and 7 without any isolation of 8, since 8 was
unstable under air. 8 was obtained as 2-butanone solution after
workup of the previous reaction, and a subsequent solvent
switch to 2-PrOH generated gummy solid, which seems to
Figure 1. ASP3026, a potent ALK inhibitor.
selective ALK inhibitor that Astellas designed and synthesized
through detailed structure−activity relationship (SAR) stud-
ies.6,7 We describe herein our effort toward the development of
an improved synthetic route for a commercial stable supply of
1 and method to stably obtain desired polymorph Form A04.
RESULTS AND DISCUSSION
■
First-Generation Manufacturing Route. During the
preclinical and early clinical phases of the development, 1
was supplied in kilogram quantities through the first-
generation synthetic route, which was established by
introducing minor changes into the discovery route using
2,4-dichloro-1,3,5-triazine (6) and 5-fluoro-2-nitroanisole (2)
as starting material (Scheme 1). Anisole 2 was converted to
intermediate 4 by a substitution reaction with piperidine 3 in
Special Issue: Japanese Society for Process Chemistry
Received: December 6, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. First-Generation Synthetic Route
synthetic scheme by obtaining 5 as a TFA salt after a
hydrogenation of 4 and using such salt directly in a
substitution reaction with 8.
Table 1. Screening of Reaction Conditions: SNAr between 8
and 5
a
Since crude 1 was obtained as a crystal of a metastable
polymorph (Form A02), solvent-mediated polymorphic trans-
formation with aqueous acetone was employed to convert 1 to
thermodynamically desired most stable polymorph, Form
A04.8,9 By this synthetic protocol, multiple kilograms of 1
were delivered for GLP/GMP studies.
Necessity of Process Improvement. While the first-
generation synthetic route was able to be applied to the API
supply in discovery and the early clinical stage, the process was
not fully satisfactory as the manufacturing method for future
commercial production because of the following points: (1)
starting 1,3-dichlorotriazine (6) was considerably expensive
and had low commercial availability, (2) aniline 5·TFA was
unstable under air and had difficulties in quality control,
storage, and transportation, and (3) the efficiency of an SNAr
reaction between 8 and 5 was low despite of the thorough
investigation shown above because of significant side-reaction.
In order to establish a cost-effective, easy-to-operate,
commercially sustainable manufacturing process, we decided
to develop an another synthetic route of 1.
Unsuccessful Route Using Cyanuric Chloride 9. At
first, we tried a synthetic method utilizing cyanuric chloride
(9) as the starting material instead of 6 (Scheme 2). Although
an additional hydrogenation to remove the extra chloro group
would be required in a later stage of synthetic stream, the
higher cost efficiency and commercial availability of 9 were
considered to be vital to develop an economically efficient
manufacturing process. 9 was smoothly reacted with 7 to give
intermediate 10. Next, we investigated various solvents in the
SNAr reaction of 10 and 5·TFA (Table 2, entries 1−9).10
When THF was used, the reaction proceeded with the highest
conversion, and the generation of byproduct was comparatively
inhibited (entry 3). As a result of examination of the lower
reaction temperature (entries 10, 11), it was found that 40 °C
is enough to ensure higher conversion (entry 11). However,
despite the sequence of investigation above, the efficiency of
the SNAr reaction was not improved sufficiently. One of the
reasons for this was thought to be the extreme instability of
resulting 11, and this nature not only led to poor yield but also
made it difficult to isolate 11 in high purity. Furthermore, it
was found that 11 easily decomposed under a condition of
HPLC area%
entry
acid
equiv
solvent
EtOH
EtOH
EtOH
EtOH
EtOH
EtOH
EtOH
EtOH
1
5
1
2
3
4
5
6
7
8
MsOH
TFA
3
3
3
3
3
3
3
3
3
3
3
3
3
0
1
2
3
4
3
3
73.4
77.6
58.2
39.3
56.5
10.1
1.0
17.7
13.5
79.9
33.9
73.7
72
4.6
5.7
14.4
18.8
8.3
36.6
2.2
0.9
6.0
1.1
37.2
3.1
3.8
3.0
7.8
2.7
3.4
3.8
6.0
3.0
formic acid
acetic acid
BF3·Et2O
AlCl3
FeCl3
ZnCl2
LiCl
TFA
TFA
TFA
TFA
TFA
TFA
TFA
TFA
TFA
TFA
TFA
9
EtOH
MeCN
DMF
10
11
12
13
14
15
16
17
18
EtOAc
2-butanone
2-PrOH
2-PrOH
2-PrOH
2-PrOH
2-PrOH
2-PrOH
2-PrOH
9.9
39.8
73.9
d
81.1
78.6
77.1
81.3
b
19
c
20
a
b
d
8 (0.30 mmol), 5 (0.30 mmol), and solvent (4.5 mL) were used.
Reaction was conducted at 40 °C. H2O (1.0 equiv) was added.
c
Isolated yield was 49%.
indicate that 2-PrOH cannot be applied to manufacturing.
From this context, we selected 2-butanone considering
manufacturability, although the yield with it was slightly
inferior than that with 2-PrOH. TFA was embedded into the
B
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 2. Unsuccessful Route from Cyanuric Chloride (9)
and 15. We expected that this new protocol could resolve all of
the issues in first synthetic route mentioned above and
embarked on the investigation based on this hypothesis.
Second-Generation Synthetic Route. A synthetic route
designed on the basis of the retrosynthetic study is shown
below (Scheme 4). A substitution reaction of 2 and piperizine
16·HCl using DBU proceeded well and furnished intermediate
17 in good yield, and following hydrogenation with Pd/C,
aniline 15 was yielded. In keeping with our strategy, we reacted
15 with aniline 10. As we expected, a reaction proceeded in
excellent yield to give intermediate 14 owing to the absence of
a reactive N-methyl piperazine moiety. A residual chloro group
on 14 was easily removed by hydrogenation with Pd/C under
a basic condition and further converted to ketone 13 by
subsequent deprotection with aqueous HCl in good yield with
high reproducibility. A reductive amination of 13 to introduce
N-methylpiperazine under a condition of NaBH(OAc)3 and
AcOH gave crude 1 in good yield. However, a crystal was
obtained as a mixed polymorph of the desirable Form A04 and
metastable Form A02 and A06; furthermore, the ratio of
polymorph in obtained crystal was unstable. Hence, recrystal-
lization with 2-butanone was incorporated in a process to
consolidate polymorphic mixture into Form A04. This second-
generation synthetic route was applied to plural manufacturing
to provide over 200 kg of API of 1. The overall yield was
significantly improved in comparison with the first-generation
synthetic route, and the manufacturing cost has been reduced
to approximately half (Table 3). Furthermore, although the
number of overall steps and reactions were increased due to
additional dechlorination and late-stage introduction of
piperazine, the production lead time was not affected by
these changes as a result of the optimization of workup and
isolation protocol.
Table 2. Screening of Reaction Conditions: SNAr between
10 and 5·TFA
a
HPLC area%
entry
solvent
temp (°C)
time (h)
11
5
1
2
3
4
5
6
7
8
2-butanone
2-PrOH
THF
MeCN
toluene
MeOH
DMF
EtOAc
CPME
THF
60
60
60
60
60
60
60
60
60
20
40
2
2
3
34.9
30.7
56.8
28
10.6
18
12.6
26.2
7.4
59.5
69.9
18.4
21.3
11.6
31.5
17.5
29.2
30.9
25.2
17.2
8.7
14
14
14
14
14
14
18
18
9
10
11
THF
11.1
a
10 (0.30 mmol), 5 (0.30 mmol), and solvent (3 mL) were used.
following hydrogenation, and the subsequent yield of these two
steps fell 17% eventually. From these results, we relinquished
this synthetic route.
Retrosynthetic Analysis. We assumed that the low yield
of the SNAr reaction between 10 and 5·TFA and the instability
of intermediate 11 comes from the highly nucleophilic tertiary
amine group on the piperazine moiety, based on the analysis of
a byproduct generated in a substitution reaction.11 From this
reasoning, we turned our attention to late-stage introduction of
an N-methylpiperazine moiety. As a result of a retrosynthetic
analysis (Scheme 3), we expected that the N-methylpiperazine
moiety could be introduced by reductive amination of ketone
13, which forms by a hydro-dechlorination and acidic
deprotection of 14. Furthermore, 14 was thought to be
obtained through a subsequent substitution reaction of 7, 9,
Polymorph Control of ASP3026 (1). Polymorphic
control is also a crucial issue to a stable API supply of 1,8,9
since (1) 1 has multiple anhydrous polymorphs (Forms A01−
A06) as shown in the following XRD and DSC charts (Figures
2 and 3);12 (2) a thermodynamic stability between the most
stable Form A04 (desired polymorph) and metastable Form
A03 is considerably close (Figure 4);13 (3) Moreover, it was
impossibly difficult to convert Form A03 to Form A04 by
solvent-mediated polymorphic transformation within a realistic
operation time, once a concomitant polymorph of Form A04/
A03 is obtained.9a
C
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 3. Retrosynthetic Analysis
Scheme 4. Second-Generation Synthetic Route
Table 3. Comparison of the Synthetic Routes
first generation
second generation
starting material
low availability
expensive
31%
high availability
inexpensive
45%
a
overall yield
manufacturing cost
overall steps
production lead time
-
approximately half
eight (seven reactions)
comparable
six (four reactions)
-
a
From 6 or 9.
Polymorph Control in the Second-Generation Syn-
thetic Route. In the first-generation synthetic route, as we
previously reported, desired Form A04 was reproducibly
obtained by solvent-mediated polymorphic transformation of
crude 1 (Form A02) in aqueous acetone.9b,c In the second
Figure 2. XRD chart of polymorphs of 1 (Forms A01−A04 and A06).
D
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 5. Recrystallization of Crude 1
unexpected change in quality of crude 1 during manufacturing
might lead to this uncontrollability; in other words, some
impurity could prominently inhibit a formation of Form A04 in
a recrystallization process.14 In regard to manufacturing, an
activated carbon treatment was introduced in a recrystallization
process to improve purity (Scheme 6), and this enabled
consolidation of polymorph into Form A04.
Figure 3. DSC chart of polymorphs of 1 (Forms A01−A04 and A06)
While we met our obligations to the API supply, we ended
up facing the following challenges for future commercial
production: (1) the yield of recrystallization step was
decreased from 90 to 75% because of adsorption loss on
activated carbon; (2) purified 1 was colored yellow through
the activated carbon process; and (3) another new impurity
was generated possibly because of oxidation by the activated
carbon−molecular oxygen system.15
Through careful investigation, we found that the contents of
a specific impurity (18, Scheme 7) in crude 1 show correlation
with the polymorph ratio of purified 1 (Figure 6). This is an
impurity that was generated by “over-reaction” of an extra C−
Cl bond on the triazine ring of 10 with 15, i.e., this impurity is
derived from the change of starting material from 6 to 9. The
analytical result indicating the possibility that 18 would exert a
negative impact to polymorphic control prompted us to
implement further investigation.
To prove this hypothesis, we conducted two types of
verification experiments: (1) spike test with impurity 18 to
recrystallization process of crude 1 and (2) complete
elimination of 18 from a synthetic stream by modification of
the reaction condition of SNAr between 10 and 15 and
subsequent confirmation of the polymorph in the final crystal.
Spike Test with 18. For preparation of impurity 18, we
first reacted an intermediate 14 with 1 equiv of aniline 15
under basic conditions to give 19, and resulting 19 was
converted to precursor 20 by an acidic deprotection with
aqueous HCl. Finally, 18 was obtained by reductive amination
of 20 with N-methylpiperazine. To confirm a polymorphic
inhibitory effect of 18, we added 1.0 wt % (0.75 area% in
HPLC) of synthesized 18 into a solution of 1 after a hot
filtration Scheme 5 and then checked the polymorphic ratio of
resulted crystal. As we expected, the crystal to which 18 was
added in the process clearly contained 45% (ratio of amount of
heat in DSC) of Form A03, while pure Form A04 was obtained
in the control experiment in which no 18 was added (Figure
7). This result strongly supports that 18 was a root cause of
contamination of Form A03.
Figure 4. Solubility of polymorphs of 1 in 2-butanone (Forms A01−
A04)
route, on the contrary, a crude 1 was obtained as a mixed
polymorph comprising Forms A02, A04, and A06 without
reproducibility possibly because of a difference in quality.
Given an instability in solvent-mediated polymorphic trans-
formation by using a mixed polymorph, we decided to turn our
attention to recrystallization. 2-Butanone was selected from
these following reasons: (1) solvent-mediated polymorphic
transformation from Form A02 to A04 was best accelerated in
2-butanone9b and (2) 1 was soluble enough in 2-butanone to
conduct recrystallization within a practical solvent volume
(Table 4). On the basis of the fact that Form A04 is preferably
Table 4. Solubility of 1 (Form A04) to Various Solvents
Form A04
solubility
2-
ethyl
acetate
2-
butanone acetone
MIBK EtOH PrOH
25 °C (g/L)
reflux (g/L)
5.8
53.8
4
13
3.8
-
3.2
-
3.1
-
0.8
-
nucleated at a higher temperature,9 we established a procedure
shown as Scheme 5 and obtained several hundred grams of
purified 1 as desirable Form A04 in a premanufacturing study.
However, the in-laboratory recrystallization of the manu-
facturing lot of crude 1 gave a concomitant polymorph of
Forms A04 and A03 with a ratio of 85:15 (ratio of amount of
heat in DSC), although the same recrystallization protocol as
that used in the premanufacturing study was applied (Figure 5,
SEM; VE-8800, Keyence Corp., Japan). We assumed that an
Elimination of 18 by Reaction Optimization. To verify
our hypothesis and to ensure a solid polymorphic control to
Form A04, next we attempted to completely eliminate 18 from
a synthetic stream (Table 5). As shown in Scheme 7, impurity
19, a source of 18, pertains to the SNAr reaction between 10
E
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 5. SEM image (500×) of Form A04 (left) and mixture of Forms A04 and A03 (right).
solution of this unstable polymorph. As we expected, a
decrease of 15 to 0.95 equiv. and change of the reaction
temperature to 0 °C effectively inhibited 19 (Table 5, entries
2,3). When −10 °C was adopted, 19 was best controlled to the
level of ND (Table 5, entry 4). As per the second-generation
synthetic route, resulting 14, in which 19 was completely
eliminated, was led to crude 1 which does not contain impurity
18. Finally, it was confirmed that such crude 1 can be
reproducibly converted to pure Form A04 by same
recrystallization condition as manufacturing without an
activated carbon process.
Scheme 6. Recrystallization of Crude 1 with Activated
Carbon
From the result of these two types of experiments, we
concluded that impurity 18 was the root cause of
contamination of Form A03 in manufacturing. This finding
and modification of a reaction condition made it possible to
conduct robust polymorphic control of 1 even in the second-
generation synthetic route and eliminate problematic activated
carbon process.
and 15. In an original protocol, which was adopted in the
manufacturing, the substitution reaction was performed with a
slight excess of 1.05 equiv of 15 (against 10) at 25 °C and
furnished intermediate 14 containing 0.21 area% of 19 (Table
5, entry 1). Importantly, in the actual manufacturing, 19 was
further increased to 0.40% during the overnight storage of the
workup solution. We therefore expected that the use of a
smaller amount of 15 and lowering the reaction temperature
would suppress generation of 19 and eventually lead to a
CONCLUSIONS
■
We have developed an improved practical manufacturing
process for ALK inhibitor ASP3026 (1). A change of starting
material from 2,4-dichloro-1,3,5-triazine (6) to cyanuric
chloride (9) significantly improved the economic efficiency
Scheme 7. Generation Path of 18
F
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
EXPERIMENTAL SECTION
■
General. Starting materials (2, 3, 6, 7, 9, 12, and 16·HCl),
reagents, and solvents were purchased from suppliers and used
without further purification. Shimadzu LC-2010 AHT and
Hitachi Lachrom Elite L-2000 were used for HPLC analysis.
HPLC Analysis. For 1, 4, 5, 11, 18, 19, 20: L-column2
ODS (Chemical Evaluation and Research Institute, Japan) 150
× 4.6 mm at a flow rate of 1.0 mL/min; peaks were monitored
at 225 nm. For 8, 14: Inertsil ODS-4 (GL Sciences, Japan) 150
× 4.6 mm at a flow rate of 1.0 mL/min; peaks were monitored
at 225 nm. For 10, 13: YMC-Pack ODS-A (YMC, Japan) 150
× 4.6 mm at a flow rate of 1.0 mL/min; peaks were monitored
at 225 nm. For 15, 17: YMC-Pack Pro C18 RS (YMC, Japan)
150 × 4.6 mm at a flow rate of 1.0 mL/min; peaks were
monitored at 225 nm. Improved condition to detect 18 in 1:
CAPCELL PAK C18 AQ (Shiseido, Japan) 150 × 4.6 mm at a
flow rate of 1.2 mL/min; peaks were monitored at 238 nm.
Purities are reported by relative HPLC area.
NMR Measurement. 1H and 13C NMR spectra were
acquired with a JEOL ECA600 NMR spectrometer at 600 and
150 MHz, respectively. Chemical shifts (δ) of 1H NMR spectra
are reported in parts per million (ppm) referenced to TMS at
0.00 or acetone-d at 2.09. Chemical shifts of 13C NMR spectra
are reported in parts per million (ppm) referenced to solvent
peak (chloroform-d at 77.0; DMSO-d at 39.5; acetone-d at
30.6). Coupling constants (J) are reported in hertz (Hz).
Mass Measurement. Mass spectra data were acquired on a
Waters Xevo TQ MS, triple quadrupole mass spectrometer, in
ESI positive mode. Column chromatography was performed
with silica gel (63-210 mesh).
Figure 6. HPLC charts of crude 1: relationship between 18 and
polymorph ratio of purified 1.
PXRD Measurement. ASP3026 polymorphs were eval-
uated with Miniflex (Rigaku Corp., Japan) under the following
conditions: X-ray tube, copper; tube current, 15 mA; tube
voltage, 30 kV; sampling width, 0.010°; scanning speed, 2°/
min; wavelength, 11.49 mm; and range of measurement
diffraction angles (2θ), 3−35°.
DSC Measurement. DSC analysis was conducted with a
TA Instruments Q-2000 calorimeter (TA Instruments, U.S.A.)
under the following conditions: temperature range of measure-
ment, ambient temperature to 493 K or higher; rate of
temperature increase, 30 °C/min; nitrogen flow rate, 50 mL/
min; and sample pan material, aluminum.
Solubility Measurement. An excess quantity of each
polymorph was added to the solvent. The temperature of the
slurry was controlled with a thermostat bath, and the slurry was
agitated over 2 h and then filtrated with a 0.45 μm membrane
filter. The concentration of a supernatant was measured by the
HPLC condition above.
First-Generation Synthetic Route. 1-[1-(3-Methoxy-4-
nitrophenyl)piperidin-4-yl]-4-methylpiperazine (4). A 1000 L
reactor was charged with 2 (23.0 kg, 134 mol), 3 (29.6 kg, 161
mol), and THF (230 L). 1,8-Diazabicyclo[5.4.0]undec-7-ene
(24.6 kg, 162 mol) was added, and the mixture was stirred at
50 °C for 8 h and then cooled to room temperature. Ethyl
acetate (460 L) and NaCl aqueous solution (46 kg in 230 L of
water) were added to the reaction mixture, and the organic
layer was washed twice with NaCl aqueous solution (46 kg in
230 L of water) and concentrated to 46 L under vacuum. The
resulting residue was added to EtOH (115 L) and
concentrated to 46 L under vacuum, and this protocol was
repeated again. n-Heptane (138 L) was added to the residue,
and the mixture was stirred at 40 °C for 1 h. To the resulting
Figure 7. DSC charts of resulting crystal (control experiment vs spike
test of 18).
of the process owing to its high commercial availability.
Furthermore, late-stage introduction of N-methyl piperazine by
reductive amination was effective in avoiding undesirable side-
reactions in the SNAr reaction. In regard to polymorphic
control in the second-generation synthetic route, we found that
a small amount of impurity 18 behaves in a way that inhibits
formation of desired polymorph Form A04 in a recrystalliza-
tion process. An improved process, in which 18 is completely
eliminated, enabled stable control to Form A04 with high
practicality and reproducibility.
G
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
a
Table 5. Screening of Reaction Condition of SNAr of 10 and 15
a
b
10 (1.00 mmol), 15, DIPEA (1.10 mmol), and THF (3.5 mL) were used. In actual manufacturing, 19 was further increased to 0.40% during
overnight storage after the reaction.
slurry was added n-heptane (276 L) at 40 °C, and this mixture
was stirred at 40 °C for 1 h and then cooled to 0 °C over 2 h.
The crystals were collected by centrifugation, washed with the
precooled mixture of EtOH/n-heptane (6.9 L/62.1 L), and
dried at 50 °C under vacuum to give 4 (40.5 kg, 121 mol, 90%
yield) as a dark purple solid. HPLC purity: 99.2 A%. 1H NMR
(600 MHz, chloroform-d) δ 8.00 (d, J = 8.9 Hz, 1 H), 6.42
(dd, J = 8.9, 2.1 Hz, 1 H), 6.31 (d, J = 2.1 Hz, 1 H), 3.97−3.92
(m, 2 H), 3.95 (s, 3 H), 3.01−2.93 (m, 2 H), 2.75−2.55 (m, 4
H), 2.55−2.34 (m, 5 H), 2.29 (s, 3 H), 2.00−1.94 (m, 2 H),
1.66−1.57 (m, 2 H). 13C NMR (150 MHz, chloroform-d) δ
156.4, 155.2, 128.9, 128.8, 105.4, 96.8, 61.1, 56.1, 55.3, 49.0,
46.8, 45.9, 27.7. LC−MS (ESI): calcd for C17H27N4O3 ([M +
H]+) 335.2; found 335.1.
2-Methoxy-4-[4-(4-methylpiperazin-1-yl)piperidin-1-yl]-
aniline Tris(trifluoroacetate) (5·TFA). A 1000 L autoclave was
charged with 4 (40.0 kg, 120 mol) and THF (400 L), and the
mixture was stirred at 25 °C until the solid was dissolved. 10%
Pd/C (AD-type, 50% wet, 6.0 kg dry-based, Kawaken Fine
Chemicals, Japan) was added, and the mixture was stirred
under H2 atmosphere (0.10 MPa) at 25 °C for 5 h. The
reaction mixture was warmed to 35 °C and transferred to the
another 1000 L reactor through a 1 μm filter, and the autoclave
was washed with THF (80 L). To the mixture was added TFA
(42.3 kg, 371 mol) at 35 °C. Seed crystals (4 g) were added at
35 °C and stirred at the same temperature for 1 h and then
stirred at 20 °C for 3 h. n-Heptane (320 L) was added to the
reactor, and the mixture was stirred at 20 °C for 3 h. The
crystals were collected by centrifugation, washed with the
mixture of THF/n-heptane (72 L/48 L), and dried at 50 °C
under vacuum to give 5·TFA(69.6 kg, 108 mol, 90% yield) as a
dark purple solid. HPLC purity: 99.9A%. 1H NMR (600 MHz,
MeOH-d) δ 7.11 (d, J = 8.9 Hz, 1 H), 6.81 (d, J = 2.8 Hz, 1
H), 6.70 (dd, J = 8.9, 2.8 Hz, 1 H), 3.94 (s, 3 H), 3.87 (d, J =
13.1 Hz, 2 H), 3.52−3.35 (m, 4 H), 3.35−3.18 (m, 4 H),
3.12−3.05 (m, 1 H), 3.01 (t, J = 11.7 Hz, 2 H), 2.89 (s, 3 H),
2.14 (d, J = 12.4 Hz, 2 H), 1.89−1.79 (m, 2 H). 13C NMR
(150 MHz, DMSO-d) δ 159.2, 159.0, 158.8, 158.5, 119.6,
117.6, 115.7, 113.7, 108.0, 108.0, 100.9, 60.9, 55.9, 50.9, 48.2,
45.5, 42.1, 25.9. LC−MS (ESI): calcd for C17H30N4O ([M +
H]+) 305.2; found 305.1.
4-Chloro-N-[2-(propane-2-sulfonyl)phenyl]-1,3,5-triazin-
2-amine (8) and N2-{2-Methoxy-4-[4-(4-methylpiperazin-1-
yl)piperidin-1-yl]phenyl}-N4-[2-(propane-2-sulfonyl)phenyl]-
1,3,5-triazine-2,4-diamine (Crude 1, Form A02). A 1000 L
reactor was charged with 6 (18.0 kg, 134 mol), 7 (26.3 kg, 132
mol), diisopropylethylamine (17.1 kg, 132 mol), and toluene
(180 L). After 3 h of stirring at 70 °C, the mixture was cooled
to 45 °C. To the resulting mixture 2-butanone (180 L) was
added and was concentrated under vacuum. 2-Butanone (90
L) was added to the residue and concentrated under vacuum.
To the residue was added 2-butanone (360 L) and K2CO3
aqueous solution (18 kg in 180 L of water), and the residue
was separated. The reactor was washed twice with 2-butanone
(18 L). 8 was able to be isolated by concentration of this
combined organic phase and further purified by column
chromatography to obtain analytical sample. HPLC purity:
97.0A%. 1H NMR (600 MHz, chloroform-d) δ 9.88 (bs, 1 H),
8.61 (s, 1 H), 8.50 (d, J = 8.3, 1 H), 7.93 (dd, J = 8.3, 2.1 Hz, 1
H), 7.74−7.69 (m, 1 H), 7.36−7.31 (m, 1 H), 3.27−3.18 (m, 1
H), 1.32 (d, J = 6.9, 6 H). 13C NMR (150 MHz, chloroform-d)
δ 170.9, 167.5, 163.9, 136.8, 135.0, 131.4, 124.9, 124.4, 123.0,
56.1, 15.2. LC−MS (ESI): calcd for C12H14ClN4O2S ([M +
H]+) 313.1; found 313.0.
A 1000 L reactor was charged with the combined 2-
butanone solution and 5·TFA (57.9 kg, 89.5 mol), and the
mixture was stirred at 65 °C for 3 h and then cooled to 20 °C.
NaOH aqueous solution (17.2 kg in 364 L of water) premixed
with NaCl (36.4 kg) was added to the reaction mixture, and
the organic layer was washed twice with NaCl aqueous
solution (72.8 kg in 364 L of water). The mixture was
transferred to a 1000 L reactor through a 0.6 μm filter, and the
reactor was washed with 2-butanone (42 L). The solution was
concentrated to 168 L under vacuum, seed crystals (28 g,
H
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Form A02) were added, and the mixture was stirred for 1 h.
The mixture was concentrated to 56 L under vacuum. Acetone
(98 L) was added, and the mixture was concentrated to 56 L
under vacuum, and this procedure was repeated two more
times. To the slurry acetone (36 L) was added, and the mixture
was stirred at 45 °C for 1 h. Isopropyl acetate (91 L) was
added to the slurry, which was cooled to 0 °C over 4 h and
stirred at 0 °C for 5 h. The crystal was obtained by filtration,
washed with a mixture of acetone and isopropyl acetate (18 L/
18 L) and isopropyl acetate (56 L), and dried at 50 °C under
vacuum to give crude 1 (25.0 kg, 43.0 mol, 48% yield) as a pale
mixture of Acetone and Isopropyl acetate (0.1 mL/0.1 mL)
and Isopropyl acetate (0.2 mL), and dried at 50 °C under
vacuum to give crude 1 (29.6 mg, 0.051 mmol, 17% yield) as
pale yellow solid. HPLC purity: 97.1A%.
Second-Generation Synthetic Route. 4,4-Dimethoxy-1-
(3-methoxy-4-nitrophenyl)piperidine (17). A 1000 L reactor
was charged with 16·HCl (70.1 kg, 386 mol), DMF (150 L),
and 1,8-diazabicyclo[5.4.0]undec-7-ene (117.4 kg, 771 mol).
To the mixture was added a solution of 2 (60.0 kg, 351 mol) in
DMF (90 L), and this was stirred at 25 °C for 5 h. Seed crystal
(6 g) was added at 25 °C, and the resulting slurry was stirred at
25 °C for 3 h. Water (240 L) was added to the mixture, with
continued stirring at 25 °C for 3 h. The crystals were collected
by centrifugation, washed twice with a mixture of DMF/water
(60 L/60 L), and dried at 50 °C under vacuum to give 17
(98.7 kg, 333 mol, 95% yield) as dark purple solid. HPLC
purity: 99.9A%. 1H NMR (600 MHz, chloroform-d) δ 8.01 (d,
J = 8.9 Hz, 1 H), 6.43 (dd, J = 8.9, 2.8 Hz, 1 H), 6.32 (d, J =
2.8 Hz, 1 H), 3.95 (s, 3 H), 3.48−3.43 (m, 4 H), 3.24 (s, 6 H),
1.90−1.85 (m, 4 H). 13C NMR (150 MHz, chloroform-d) δ
156.4, 155.2, 129.1, 128.8, 105.6, 98.1, 97.1, 56.1, 47.6, 44.6,
31.9. LC−MS (ESI): calcd for C14H21N2O5 ([M + H]+) 297.2;
found 297.1.
1
yellow solid in Form A02. HPLC purity: 97.9A%. H NMR
(600 MHz, chloroform-d) δ 9.28 (s, 1H), 8.47−8.61 (bs, 1H),
8.30−8.48 (m, 1H), 8.10 (bs, 1H), 7.88 (d, J = 8.2 Hz, 1H),
7.49−7.72 (m, 2H), 7.23 (t, J = 8.2 Hz, 1H), 6.45−6.60 (m,
2H), 3.88 (s, 3H), 3.63−3.75 (m, 2H), 3.18−3.32 (m, 1H),
2.31−2.78 (m, 11H), 2.30 (s, 3H), 1.92−2.01 (m, 2H), 1.65−
1.79 (m, 2H), 1.31 (d, J = 6.9 Hz, 6H). 13C NMR (150 MHz,
chloroform-d) δ (ppm) 166.3, 163.6, 149.6, 148.6, 138.4,
134.5, 131.1, 124.4, 123.6, 123.0, 122.1, 121.4, 119.6, 108.0,
100.4, 61.6, 55.6, 55.5, 55.4, 50.0, 49.0, 46.0, 28.1, 15.3. LC−
MS (ESI): calcd for C29H41N8O3S ([M + H]+) 581.3; found
581.2.
N2-{2-Methoxy-4-[4-(4-methylpiperazin-1-yl)piperidin-1-
yl]phenyl}-N4-[2-(propane-2-sulfonyl)phenyl]-1,3,5-triazine-
2,4-diamine (1, Form A04). A 1000 L reactor was charged
with crude 1 (26.0 kg, 44.7 mol), and hot acetone/water (78
L/78 L) and seed crystals (26 g of 1, Form A04) were added.
The mixture was stirred at 62 °C for 6 h, and hot acetone/
water (52 L/52 L) was added. After 18 h of stirring at 62 °C,
the mixture was cooled to 25 °C over 8 h and was stirred at 25
°C for 8 h. The crystals were filtered, washed with a mixture of
acetone/water (78 L/78 L), and dried at 50 °C under vacuum
to give 1 (22.1 kg, 38.0 mol, 85% yield) as a pale yellow solid
in Form A04. HPLC purity: 99.9A%.
Unsuccessful Route from Cyanuric Chloride (9). 6-
Chloro-N2-{2-methoxy-4-[4-(4-methylpiperazin-1-yl)-
piperidin-1-yl]phenyl}-N4-[2-(propane-2-sulfonyl)phenyl]-
1,3,5-triazine-2,4-diamine (11) and N2-{2-Methoxy-4-[4-(4-
methylpiperazin-1-yl)piperidin-1-yl]phenyl}-N4-[2-(propane-
2-sulfonyl)phenyl]-1,3,5-triazine-2,4-diamine (1). Procedure
for Table 2, entry 11: A test tube was charged with 10 (104
mg, 0.30 mmol) and THF (3 mL). To the mixture was added
5·TFA (194 mg, 0.30 mmol) at room temperature. After
stirring at 40 °C for 18 h, the precipitated solid was collected
by filtration to give 11, which was used for the next step
without characterization by NMR for stability issues. LC−MS
(ESI): calcd for C29H40ClN8O3S ([M + H]+) 615.3; found
615.2.
An autoclave was charged with above solid, ammonium
formate (95 mg, 1.5 mmol), MeOH (3 mL) and 10% Pd/C
(NX-type, 21 mg, N.E. CHEMCAT, Japan), and the mixture
was stirred under H2 atmosphere (0.10 MPa) at 50 °C for 3 h.
The reaction mixture was transferred to the another flask
through a 1 μm filter, and the autoclave was washed with
MeOH (5 mL). The filtrate was concentrated under vacuum,
and then diluted with 2-butanone (10 mL), NaOH aqueous
solution (47 mg in 10 mL of water) premixed with NaCl (1 g)
was added, then organic layer was separated. The organic layer
was washed twice with NaCl aqueous solution (2 g in 10 mL of
water). The 2-butanone solution was concentrated to dryness,
and Acetone (0.2 mL) and Isopropyl acetate (0.2 mL) were
added. The crystal was collected by filtration, washed with the
4-(4,4-Dimethoxypiperidin-1-yl)-2-methoxyaniline (15). A
1000 L autoclave was charged with 17 (90.0 kg, 304 mol),
THF (450 L) and 10% Pd/C (M-type, 50% wet, 4.5 kg dry-
based, Kawaken Fine Chemicals, Japan). The mixture was
stirred under H2 atmosphere (0.30 MPa) at 25 °C for 3 h. The
reaction mixture was transferred to another 1500 L reactor
through a 1 μm filter, and the autoclave was washed with THF
(180 L). The filtrate was concentrated to 180 L under vacuum,
and seed crystal (9 g) was added. The slurry was stirred at 40
°C, and then, n-heptane (270 L) was added. After 1 h of
stirring at 40 °C, the mixture was cooled to 0 °C over 4 h and
stirred at 0 °C for 2 h. To the mixture was added n-heptane
(810 L) at 0 °C, and the mixture was stirred at 0 °C for 2 h.
The crystals were collected by centrifugation, washed with a
mixed solution of THF/n-heptane (18 L/108 L), and dried at
50 °C under vacuum to give 15 (72.8 kg, 273 mol, 90% yield)
1
as dark purple solid. HPLC purity: 99.9A%. H NMR (600
MHz, chloroform-d) δ 6.64 (d, J = 8.3 Hz, 1 H), 6.54 (d, J =
2.8 Hz, 1 H), 6.44 (dd, J = 8.3, 2.8 Hz, 1 H), 3.84 (s, 3 H),
3.53 (bs, 2 H), 3.23 (s, 6 H), 3.08−3.04 (m, 4 H), 1.94−1.88
(m, 4 H). 13C NMR (150 MHz, chloroform-d) δ 147.9, 145.0,
130.0, 115.4, 110.0, 102.9, 98.3, 55.4, 48.7, 47.4, 32.5. LC−MS
(ESI): calcd for C14H23N2O3 ([M + H]+) 267.2; found 267.1.
4,6-Dichloro-N-[2-(propane-2-sulfonyl)phenyl]-1,3,5-tria-
zin-2-amine (10). A 1000 L reactor was charged with 9 (40.0
kg, 217 mol), NaHCO3 (21.9 kg, 261 mol) and Acetone (320
L). The mixture was added 7 (47.5 kg, 238 mol), and stirred at
25 °C for 21 h. Water (320 L) was added, and the mixture was
stirred at 25 °C for 21 h. The crystals were collected by
centrifugation, washed with a mixture of Acetone/water (80 L/
80 L), and dried at 50 °C under vacuum to give 10 (71.5 kg,
206 mol, 95% yield) as white solid. HPLC purity: 99.9A%. 1H
NMR (600 MHz, ACETONE-d) δ 8.41 (d, J = 7.6 Hz, 1 H),
7.97 (dd, J = 7.6, 1.4 Hz, 1 H), 7.89−7.85 (m, 1 H), 7.55−7.50
(m, 1 H), 3.52−3.45 (m, 1 H), 1.26 (d, J = 6.2 Hz, 6 H). − 1
H (NH) 13C NMR (150 MHz, chloroform-d) δ 165.6, 137.1,
136.1, 132.4, 127.7, 126.6, 125.2, 100.9, 56.6, 15.5. LC−MS
(ESI): calcd for C12H13Cl2N4O2S ([M + H]+) 347.0; found
346.9.
I
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
6-Chloro-N2-[4-(4,4-dimethoxypiperidin-1-yl)-2-methoxy-
phenyl]-N4-[2-(propane-2-sulfonyl)phenyl]-1,3,5-triazine-
2,4-diamine (14). A 1000 L reactor was charged with 10 (60.0
kg, 173 mol) and THF (600 L). To the mixture was added 15
(48.3 kg, 181 mol), followed by diisopropylethylamine (24.6
kg, 190 mol) at 25 °C. After stirring at 25 °C for 3 h, isopropyl
acetate (60 L) and K2CO3 aqueous solution (3.0 kg in 60 L of
water) were added to the mixture. After the phase separation,
the organic layer was concentrated under vacuum. Seed crystal
(6 g) was added at 25 °C, and the mixture was stirred at 25 °C
for 1 h. To the slurry was added 2-propanol (120 L) and n-
heptane (300 L), and the mixture was stirred at 25 °C for 2 h.
The mixture was cooled to 0 °C over 2 h and was stirred at 0
°C for 5 h. The crystal was collected by centrifugation, washed
with a mixture of THF/2-propanol/n-heptane (72 L/48 L/120
L), and dried at 50 °C under vacuum to give 14 (93.7 kg, 162
N2-{2-Methoxy-4-[4-(4-methylpiperazin-1-yl)piperidin-1-
yl]phenyl}-N4-[2-(propane-2-sulfonyl)phenyl]-1,3,5-triazine-
2,4-diamine (Crude 1, Polymorphic Mixture). A 1000 L
reactor was charged with 13 (50.0 kg, 101 mol), 12 (20.2 kg,
202 mol), and toluene (500 L). To the mixture was added
acetic acid (22.5 L). The mixture was stirred at 20 °C for 1 h,
and NaBH(OAc)3(42.7 kg, 201 mol) was added. After 16 h of
stirring at 20 °C, MeOH (50 L) and water (150 L) were added
and separated. The organic layer was extracted with water (50
L), and the combined aqueous layer was washed twice with
isopropyl acetate (500 L). The aqueous layer was charged into
a 2000 L reactor using water (25 L). To the aqueous solution,
MeOH (550 L), NaOH aqueous solution (24.2 kg in 120 L of
water), and water (25 L) were added. Seed crystals (5 g, Form
A04) were added at 25 °C, and the mixture was stirred at 25
°C for 1 h. To the mixture, water (550 L) was added and
stirred at 25 °C for 2 h. The crystal was collected by
centrifugation, washed with a mixture of MeOH/water (100
L/100 L), and dried at 50 °C under vacuum to give crude 1 as
pale yellow solid comprising polymorphs of Form A02/A04/
A06 (42.7 kg, 73.5 mol, 73% yield). HPLC purity: 99.6A%.
N2-{2-Methoxy-4-[4-(4-methylpiperazin-1-yl)piperidin-1-
yl]phenyl}-N4-[2-(propane-2-sulfonyl)phenyl]-1,3,5-triazine-
2,4-diamine (1, Form A04). Procedure using activated carbon:
A 2000 L reactor was charged with crude 1 (50.0 kg, 64.6
mol), activated carbon (5.0 kg), and 2-butanone (1200 L).
After 1 h of stirring at 70 °C, the mixture was transferred to a
2000 L reactor through a filter paper and 1 μm cartridge filter,
and the reactor was washed with 2-butanone (100 L). To the
filtrate was added activated carbon (5.0 kg), which was stirred
at 70 °C for 1 h. The mixture was transferred to a 2000 L
reactor through a filter paper and 1 μm cartridge filter, and the
reactor was washed with 2-butanone (100 L). Again, to the
filtrate was added activated carbon (5.0 kg), which was stirred
at 70 °C for 1 h. The mixture was transferred to a 2000 L
reactor through a filter paper and 1 μm cartridge filter, and the
reactor was washed with 2-butanone (100 L). The filtrate was
concentrated to 400 L. Seed crystal (5 g, Form A04) was
added at 70 °C and stirred for 3 h at same temperature. The
slurry was cooled to 0 °C over 4 h, the crystal was collected by
centrifugation, washed with 2-butanone (200 L), and dried at
50 °C under vacuum to give 1 (37.5 kg, 64.6 mol, 75% yield)
as a pale yellow solid in Form A04. HPLC purity: 99.7A%.
Synthesis of Impurity 18. N2,N4-Bis[4-(4,4-dimethox-
ypiperidin-1-yl)-2-methoxyphenyl]-N6-[2-(propane-2-
sulfonyl)phenyl]-1,3,5-triazine-2,4,6-triamine (19) and 1,1′-
({6-[2-(Propane-2-sulfonyl)anilino]-1,3,5-triazine-2,4-diyl}-
bis[azanediyl(3-methoxy-4,1-phenylene)])di(piperidin-4-
one) (20). A 100 mL three-neck round-bottom flask was
charged with 14 (5.0 g, 8.66 mmol), 15 (2.3 g, 8.66 mmol),
and THF (35 mL). To the mixture was added 1 M NaOH
aqueous solution (17.3 mL), and the mixture was stirred at 63
°C for 32 h. The mixture was transferred to a separating funnel
with THF (5.0 mL) and separated. The organic layer was
washed three times with K2CO3 aqueous solution (2.0 g in 38
mL water) and charged into a 100 mL three-neck round-
bottom flask. 19 was able to be isolated as a white solid by
concentrating this organic layer in 99.9A% of HPLC purity. 1H
NMR (600 MHz, chloroform-d) δ 9.03 (s, 1 H), 8.60 (bs, 1
H), 8.20 (bs, 2 H), 7.85 (dd, J = 7.6, 1.4 Hz, 1 H), 7.63−7.56
(m, 1 H), 7.27 (bs, 2 H), 7.16 (t, J = 7.6 Hz, 1 H), 6.61−6.52
(m, 4 H), 3.88 (s, 6 H), 3.33−3.27 (m, 1 H), 3.24 (s, 12 H),
3.22−3.18 (m, 8 H), 1.95−1.91 (m, 8 H), 1.32 (d, J = 6.9 Hz,
1
mol, 94% yield) as white solid. HPLC purity: 99.9A%. H
NMR (600 MHz, chloroform-d) δ 9.48−9.31 (m, 1 H), 8.46
(d, J = 8.3 Hz, 1 H), 8.16−8.00 (m, 1 H), 7.90 (d, J = 7.6 Hz, 1
H), 7.72 (s, 1 H), 7.68−7.54 (m, 1 H), 7.30−7.23 (m, 1 H),
6.61−6.48 (m, 1 H), 6.54 (d, J = 2.8 Hz, 1 H), 3.87 (s, 3 H),
3.31−3.24 (m, 1 H), 3.24 (s, 6 H), 3.23−3.18 (m, 4 H), 1.94−
1.89 (m, 4 H), 1.31 (d, J = 6.9 Hz, 6 H). 13C NMR (150 MHz,
chloroform-d) δ 169.3, 164.3, 163.5, 149.8, 148.7, 137.7, 134.5,
131.1, 124.9, 123.6, 121.6, 119.0, 108.1, 100.2, 98.3, 55.8, 55.6,
47.5, 47.1, 32.1, 15.3. LC−MS (ESI): calcd for
C26H34ClN6O5S ([M + H]+) 577.2; found 577.0.
1-[3-Methoxy-4-({4-[2-(propane-2-sulfonyl)anilino]-1,3,5-
triazin-2-yl}amino)phenyl]piperidin-4-one (13). A 1500 L
autoclave was charged with 14 (90.0 kg, 156 mol), THF (810
L), and 2-propanol (90 L), and the mixture was stirred at 25
°C until the solid was dissolved. 10% Pd/C (AD-type, 50%
wet, 9.0 kg dry-based, Kawaken Fine Chemicals, Japan) and
diisopropylethylamine (24.2 kg, 187 mol) were added, and the
mixture was stirred under H2 atmosphere (0.25 MPa) at 40 °C
for 7 h. The reaction mixture was transferred to another 1000
L reactor through a 1 μm filter, and the autoclave was washed
with THF (180 L). HCl aqueous solution (32.5 kg of 36 wt %
hydrochloric acid in 180 L of water) was added, and the
mixture was stirred at 25 °C for 10 h. To the reaction mixture
was added K2CO3 aqueous solution (53.9 kg in 180 L of
water), and this was separated. To the organic layer was added
activated carbon (9.0 kg), and the mixture was stirred at 25 °C
for 12 h. The reaction mixture was transferred to the another
1000 L reactor through 1 μm filter, and the used reactor was
washed with THF (90 L). The mixture was concentrated to
315 L under vacuum, and acetone (270 L) was added. Seed
crystals (9 g) were added at 25 °C and stirred at 25 °C for 1 h.
To the residue was added water (720 L), and the mixture was
stirred at 25 °C for 1 h. The crystal was collected by
centrifugation, washed with a mixture of acetone/water (54 L/
144 L), and dried at 50 °C under vacuum to give 13 (71.2 kg,
143 mol, 92% yield) as white solid. HPLC purity: 99.9A%. 1H
NMR (600 MHz, chloroform-d) δ 9.31 (s, 1 H), 8.53 (bs, 1
H), 8.50−8.31 (m, 1 H), 8.18 (bs, 1 H), 7.89 (dd, J = 8.3, 1.4
Hz, 1 H), 7.63 (bs, 1 H), 7.23 (t, J = 7.6 Hz, 1 H), 6.67−6.56
(m, 2 H), 3.91 (s, 3 H), 3.58 (t, J = 5.6 Hz, 4 H), 3.31−3.22
(m, 1 H), 2.60 (t, J = 5.6 Hz, 4 H), 1.31 (d, J = 6.2 Hz, 6 H).
13C NMR (150 MHz, chloroform-d) δ 207.8, 166.4, 163.6,
163.5, 150.0, 146.5, 138.5, 134.5, 131.1, 124.4, 123.4, 122.9,
121.7, 120.2, 107.9, 100.1, 55.7, 55.6, 49.6, 40.8, 15.3. LC−MS
(ESI): calcd for C24H29N6O4S ([M + H]+) 497.2; found 497.0.
J
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
6 H). 13C NMR (150 MHz, chloroform -d) δ 164.3, 164.2,
149.5, 147.7, 139.5, 134.3, 130.9, 123.8, 122.0, 121.6, 121.5,
121.1, 108.6, 100.8, 98.4, 55.6, 55.4, 47.6, 47.5, 32.3, 15.4.
LC−MS (ESI): calcd for C40H55N8O8S ([M + H]+) 807.4;
found 807.1.
ACKNOWLEDGMENTS
■
We thank Hodogaya JRF Contract Laboratories for their
support on documentation of the manufacturing procedure as
well as NMR and MS measurement. We thank Dr. Yasuhiro
Kondo for the support on NMR measurement.
6 M HCl aqueous solution (7.2 mL) was added to the
mixture and stirred at 25 °C for 21 h. To the mixture was
added 5 M NaOH aqueous solution (7.2 mL), and the
aqueous layer was extracted with THF (35 mL). The
combined organic layer was concentrated under vacuum, and
the residue was purified by silica gel column chromatography.
The resulting crystal was further purified by washing with
mixed solvent of acetone/water (30 mL/20 mL) and dried at
50 °C under vacuum to give 20(4.1 g, 5.74 mmol, 66% yield)
REFERENCES
■
(1) Morris, S. W.; Kirstein, M. N.; Valentine, M. B.; Dittmer, K. G.;
Shapiro, D. N.; Saltman, D. L.; Look, A. T. Fusion of a kinase gene,
ALK, to a nucleolar protein gene, NPM, in non-Hodgkin’s lymphoma.
Science 1994, 263, 1281−1284.
(2) Soda, M.; Choi, Y. L.; Enomoto, M.; Takada, S.; Yamashita, Y.;
Ishikawa, S.; Fujiwara, S.; Watanabe, H.; Kurashina, K.; Hatanaka, H.;
Bando, M.; Ohno, S.; Ishikawa, Y.; Aburatani, H.; Niki, T.; Sohara, Y.;
Sugiyama, Y.; Mano, H. Identification of the transforming EML4-ALK
fusion gene in non-small-cell lung cancer. Nature 2007, 448, 561−
566.
(3) Mano, H. Non-solid oncogenes in solid tumors: EML4-ALK
fusion genes in lung cancer. Cancer Sci. 2008, 99, 2349−2355.
(4) Soda, M.; Takada, S.; Takeuchi, K.; Choi, Y. L.; Enomoto, M.;
Ueno, T.; Haruta, H.; Hamada, T.; Yamashita, Y.; Ishikawa, Y.;
Sugiyama, Y.; Mano, H. A mouse model for EML4-ALK-positive lung
cancer. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 19893−19897.
(5) Shaw, A. T.; Solomon, B. Targeting anaplastic lymphoma kinase
in lung cancer. Clin. Cancer Res. 2011, 17, 2081−2086.
1
as a white solid. HPLC purity: 99.9A%. H NMR (600 MHz,
chloroform-d) δ 9.09 (s, 1 H), 8.60 (bs, 1 H), 8.27 (bs, 2 H),
7.88 (dd, J = 7.6, 1.4 Hz, 1 H), 7.65−7.57 (m, 1 H), 7.32 (bs, 2
H), 7.19 (t, J = 7.6 Hz, 1 H), 6.64−6.55 (m, 4 H), 3.91 (s, 6
H), 3.57 (t, J = 5.5 Hz, 8 H), 3.34−3.26 (m, 1 H), 2.60 (t, J =
5.5 Hz, 8 H), 1.33 (d, J = 6.9 Hz, 6 H). 13C NMR (150 MHz,
DMSO-d) δ 207.7, 164.6, 163.5, 152.3, 146.8, 139.3, 134.7,
130.6, 125.2, 122.2, 122.1, 121.7, 119.1, 106.5, 99.8, 55.5, 54.9,
48.0, 40.0, 14.8. LC−MS (ESI): calcd for C36H43N8O6S ([M +
H]+) 715.3; found 715.1.
N2,N4-Bis{2-methoxy-4-[4-(4-methylpiperazin-1-yl)-
piperidin-1-yl]phenyl}-N6-[2-(propane-2-sulfonyl)phenyl]-
1,3,5-triazine-2,4,6-triamine (18). A 300 mL autoclave was
charged with 20 (2.0 g, 2.8 mmol), 12 (1.12 g, 11.2 mmol),
and THF (150 mL), and the mixture was stirred at 25 °C until
the solid was dissolved. 10% Pd/C (EB-type, 50% wet, 0.2 g
dry-based, Kawaken Fine Chemicals, Japan) was added, and
the mixture was stirred under H2 atmosphere (0.20 MPa) at 50
°C for 18 h. The mixture was transferred to another flask
through a Celite pad. The filtrate was concentrated under
vacuum, and the residue was purified by column chromatog-
raphy to give 18 (1.8 g, 2.04 mmol, 73% yield) as white solid.
(6) Kondoh, Y.; Iikubo, K.; Kuromitsu, S.; Shindo, N.; Soga, T.;
Furutani, T.; Shimada, I.; Matsuya, T.; Kurosawa, K.; Kamikawa, A.;
Mano, H. Di(Arylamino)Aryl Compound. PCT Int. Appl. WO
2009008371 Al, 2009.
(7) Iikubo, K.; Kondoh, Y.; Shimada, I.; Matsuya, T.; Mori, K.;
Ueno, Y.; Okada, M. Discovery of N-{2-Methoxy-4-[4-(4-methyl-
piperazin-1-yl)piperidin-1-yl]phenyl}-N’-[2-(propane-2-sulfonyl)-
phenyl]-1,3,5-triazine-2,4-diamine (ASP3026), a Potent and Selective
Anaplastic Lymphoma Kinase (ALK) Inhibitor. Chem. Pharm. Bull.
2018, 66, 251−262.
(8) Shimada, I.; Hirakura, Y.; Yamazaki, K.; Takeguchi, K.; Ueda, N..
Crystals of Di(Arylamino)Aryl Compound. PCT Int. Appl.
WO2011145548 A1, 2011.
1
(9) (a) Takeguchi, K.; Hirakura, Y.; Yamazaki, K.; Shimada, I.; Ieda,
S.; Okada, M.; Takiyama, H. Characterization and Thermodynamic
Stability of Polymorphs of Di(arylamino) Aryl Compound ASP3026.
Chem. Pharm. Bull. 2015, 63, 418−422. (b) Takeguchi, K.; Obitsu, K.;
Hirasawa, S.; Orii, R.; Ieda, S.; Okada, M.; Takiyama, H. Strategy for
Controlling Polymorphism of Di(Arylamino) Aryl Compound
ASP3026 and Monitoring Solution Structures via Raman Spectros-
copy. Org. Process Res. Dev. 2015, 19, 1966−1972. (c) Takeguchi, K.;
Obitsu, K.; Hirasawa, S.; Orii, R.; Ieda, S.; Okada, M.; Takiyama, H.
Effect of Temperature and Solvent of Solvent-Mediated Polymorph
Transformation on ASP3026 Polymorphs and Scale-up. Org. Process
Res. Dev. 2016, 20, 970−976.
HPLC purity: 97.1A%. H NMR (600 MHz, chloroform-d) δ
9.03 (s, 1 H), 8.60 (bs, 1 H), 8.20 (bs, 2 H), 7.86 (dd, J = 8.3,
1.4 Hz, 1 H), 7.64−7.56 (m, 1 H), 7.27 (bs, 2 H), 7.17 (t, J =
7.6 Hz, 1 H), 6.58−6.49 (m, 4 H), 3.88 (s, 6 H), 3.68 (d, J =
12.4 Hz, 4 H), 3.34−3.26 (m, 1 H), 2.71 (t, J = 12.4 Hz, 4 H),
2.68−2.42 (m, 12 H), 2.42−2.35 (m, 2 H), 2.30 (s, 6 H), 1.96
(d, J = 11.7 Hz, 4 H), 1.77−1.66 (m, 8 H), 1.31 (d, J = 6.9 Hz,
6 H). 13C NMR (150 MHz, DMSO-d) δ 164.6, 163.5, 151.9,
148.9, 139.3, 134.6, 130.5, 124.9, 122.1, 122.0, 121.6, 119.0,
106.9, 100.1, 60.8, 55.4, 55.1, 54.9, 48.6, 48.5, 45.6, 27.7, 14.8.
LC−MS (ESI): calcd for C46H67N12O4S ([M + H]+) 883.5;
found 883.2.
(10) When an organic base such as DMEDA was used, the reaction
became a complicated mixture.
(11) The N-methyl piperazine adduct was proposed as the major
byproduct of this reaction from LC−MS analysis (ESI calcd for
C17H24ClN6O2S [M+H]+ 411.1, found 411.0).
(12) Form A05 was excluded from XRD/DSC charts, since this is a
polymorph that can be obtained only in water-containing medium and
is outside the scope of this study. In regard to the XRD and DSC
chart of Form A05, see ref 9a.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.oprd.8b00427.
1H and 13C NMR spectra (PDF)
(13) It was deemed difficult to determine the solubility of Form A06,
since solvent-mediated polymorphic transformation of Form A06 to
the other stable/metastable polymorph immediately occurs during a
solubility measurement.
AUTHOR INFORMATION
Corresponding Author
*E-mail: yuji.takahama@astellas.com.
■
(14) Examples of impurity inhibiting polymorphic control of a
pharmaceutical compound, see: (a) Mukuta, T.; Lee, A. Y.;
Kawakami, T.; Myerson, A. S. Influence of Impurities on the
Solution-Mediated Phase Transformation of an Active Pharmaceutical
Ingredient. Cryst. Growth Des. 2005, 5, 1429−1436. (b) Okamoto,
ORCID
Yuji Takahama: 0000-0002-3863-1997
Notes
The authors declare no competing financial interest.
K
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
M.; Hamano, M.; Igarashi, K.; Ooshima, H. The Effects of Impurities
on Crystallization of Polymorphs of a Drug Substance AE1−923. J.
Chem. Eng. Jpn. 2004, 37, 1224−1231. (c) Gong, Y.; Collman, B. M.;
Mehrens, S. M.; Lu, E.; Miller, J. M.; Blackburn, A.; Grant, D. J. W.
Stable-Form Screening: Overcoming Trace Impurities That Inhibit
Solution-Mediated Phase Transformation to the Stable Polymorph of
Sulfamerazine. J. Pharm. Sci. 2008, 97, 2130−2144. (d) Machiya, K.;
Ieda, S.; Hirano, M.; Ooshima, H. Effects of Impurities on Crystal
Polymorphism of an Imidazopyridine Derivative Developed as a Drug
Substance for Osteoporosis. J. Chem. Eng. Jpn. 2009, 42, 147−152.
(15) When crude 1 was treated with an activated carbon in a
recrystallization step, less than 0.2% of new impurity was observed in
purified 1. LC−MS analysis indicated that this impurity is derived
from an oxidation of the methylene group on 1 (ESI calcd for
C29H41N8O3S (1) [M+H]+ 581.3, found 595.2). For examples of
oxidation by an activated carbon−molecular oxygen system, please
see: (a) Nakamichi, N.; Kawabata, H.; Hayashi, M. Oxidative
Aromatization of 9,10-Dihydroanthracenes Using Molecular Oxygen
Promoted by Activated Carbon. J. Org. Chem. 2003, 68, 8272−8273.
(b) Hayashi, M. Oxidation Using Activated Carbon and Molecular
Oxygen System. Chem. Rec. 2008, 8, 252−267.
L
DOI: 10.1021/acs.oprd.8b00427
Org. Process Res. Dev. XXXX, XXX, XXX−XXX