A Practical and Scalable Method for Manufacturing JAK Inhibitor ASP3627

Article abstract of DOI:10.1021/acs.oprd.9b00269

ASP3627 (1) is a potent Janus kinase inhibitor discovered and developed by Astellas Pharma Inc. Here, we report the development of a practical and scalable method for manufacturing ASP3627 featuring a six-reaction telescoping process consisting of DIBAL reduction, hydrolysis, Wittig reaction, cyclization, and a two-step sequence to remove the SEM group. In addition, a simple palladium removal procedure using l-cysteine is described. This method was used to successfully perform the first multikilogram synthesis of ASP3627, producing 30 kg with high purity in excellent overall yield.

Full text of DOI:10.1021/acs.oprd.9b00269

Subscriber access provided by LUNDS UNIV
Full Paper
A Practical and Scalable Method for Manufacturing JAK Inhibitor ASP3627
Shun Hirasawa, Takashi Kikuchi, and Souichirou Kawazoe
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.9b00269 • Publication Date (Web): 09 Sep 2019
Downloaded from pubs.acs.org on September 9, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
A Practical and Scalable Method for Manufacturing
JAK Inhibitor ASP3627
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Shun Hirasawa,†,* Takashi Kikuchi,† and Souichirou Kawazoe†
Process Chemistry Labs., Pharmaceutical Technology, Astellas Pharma Inc., 160-2 Akahama,
Takahagi-shi, Ibaraki 318-0001, Japan
1
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 37
1
2
3
4
5
6
7
TOC GRAPHIC
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
· first scale-up
· six isolations
· 54.6% overall yield
· telescoping process
· Pd scavenging by L-cysteine
JAK inhibitor
ASP3627
30 kg
2
ACS Paragon Plus Environment

Page 3 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
ABSTRACT
8
9
ASP3627 (1) is a potent Janus kinase inhibitor discovered and developed by Astellas
Pharma Inc. Here, we report the development of a practical and scalable method for
manufacturing ASP3627 featuring a six-reaction telescoping process consisting of DIBAL
reduction, hydrolysis, Wittig reaction, cyclization, and a two-step sequence to remove the SEM
group. In addition, a simple palladium removal procedure using ʟ-cysteine is described. This
method was used to successfully perform the first multi-kilogram synthesis of ASP3627,
producing 30 kg with high purity in excellent overall yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
KEYWORDS
JAK inhibitor, telescoping process, palladium, ʟ-cysteine, SNAr reaction, SEM protection.
3
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 37
1
2
3
4
5
6
7
INTRODUCTION
8
9
Janus kinase (JAK) inhibitors are widely known to be useful in the treatment of immune
and inflammatory diseases such as rheumatoid arthritis, inflammatory bowel disease, psoriasis,
and transplant rejection.1 Several JAK inhibitors (e.g., tofacitinib2 and ruxolitinib3) have been
approved by the U.S. Food and Drug Administration (FDA) and a number of others are under
development. ASP3627 (1),4 a tricyclic dipyrrolopyridine derivative, was discovered by Astellas
Pharma Inc. as a highly potent JAK inhibitor with good oral bioavailability,5 and is a promising
candidate for the treatment of acute and chronic rejection in cardiac transplantation.6 While a
discovery synthetic approach appears to be suitable for large-scale synthesis of ASP3627, there
are a number of problems associated with manufacturing 30 kg of an active pharmaceutical
ingredient (API). This article describes the challenges we encountered in the process of
developing a practical, scalable and reproducible manufacturing process for the first scale-up
synthesis of ASP3627.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. ASP3627 (1), a highly potent JAK inhibitor.
ASP3627 (1)
4
ACS Paragon Plus Environment

Page 5 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
RESULTS AND DISCUSSION
8
9
Medicinal chemistry synthetic route
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of ASP3627 via the discovery route is shown in Scheme 1.5 Protection of 4-
chloro-1H-pyrrolo[2,3-b]pyridine-5-carbonitrile (2)7 with SEMCl gave 3. SNAr reaction of 3
with racemic 1-benzyl-4-methylpiperidin-3-amine (4)8 yielded rac-5 under a high reaction
temperature (180 °C). The cyano moiety of rac-5 was transformed to a pyrrole ring using the
following sequence: DIBAL reduction, hydrolysis, Wittig reaction to give an E/Z mixture of rac-
7, and two-step cyclization with acid via a dimethyl acetal intermediate to give rac-8.
Subsequently, the SEM group of rac-8 was removed using a two-step sequence, followed by
enantiomeric separation using chiral column chromatography to yield (3R,4R)-9. Debenzylation
of (3R,4R)-9 using Pd(OH)2/C and ammonium formate gave (3R,4R)-10. Condensation of
(3R,4R)-10 and cyanoacetic acid (11) was performed using EDC·HCl and HOBt to give
ASP3627 (1) in good overall yield (33% yield/11 reactions).
5
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 37
1
2
3
4
5
6
7
Scheme 1. Synthesis of ASP3627a via the discovery route (from ref 5).
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
b
a
c
d
2
3
rac-5
rac-6
rac-4
e
f,g
h
i
ASP3627 (1)
11
rac-7
(3R,4R)-10
rac-8
(3R,4R)-9
a
Reagents and conditions. (a) SEMCl, NaH, THF, DMF, rt, 3 h, column chromatography,
88%; (b) 4, DIPEA, NMP, 180 °C, 2 h, column chromatography, 91%; (c) DIBAL in n-hexane,
THF, 0 °C, 1 h; MeOH, −78 °C; then 1 M aq HCl, 0 °C, 20 min, column chromatography, 82%;
(d) MeOCH2P+Ph3Cl−, NaHMDS, THF, 0 °C, 30 min; then 6, rt, 13 h, column chromatography;
(e) AcCl, MeOH, 80 °C, 0.5 h; then H2O, 80 °C, 1 h, column chromatography, 83% from rac-6;
(f) TFA, CH2Cl2, rt, overnight; ethylenediamine, 1 M aq NaOH, CH2Cl2, rt, 2 h, 74%; (g)
enantiomeric separation using chiral column chromatography; (h) Pd(OH)2/C, ammonium
formate, MeOH, 80 °C, 0.5 h, column chromatography, 95%; (i) 11, EDC·HCl, HOBt, Et3N,
DMF, 50 °C, 2 h, column chromatography, 87%.
SEMCl = 2-(trimethylsilyl)ethoxymethyl chloride; DIPEA = N,N-diisopropylethylamine; NMP =
1-methyl-2-pyrrolidinone; DIBAL = diisobutylaluminium hydride; NaHMDS = sodium
bis(trimethylsilyl)amide; TFA = trifluoroacetic acid; EDC·HCl = N-(3-dimethylaminopropyl)-
N’-ethylcarbodiimide hydrochloride; HOBt = 1-hydroxybenzotriazole.
6
ACS Paragon Plus Environment

Page 7 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
Large-scale synthesis
8
9
We determined that the discovery synthetic approach was suitable for the first scale-up
synthesis of ASP3627 for two reasons: (1) starting material 2 was readily available in high purity
and, because chiral amine (3R,4R)-4 was commercially available, the enantiomeric separation
process could be avoided; and (2) all reactions in the discovery route had good yields (11
reactions in 33% yield). However, a number of difficulties were also associated with
manufacturing multi-kilograms of ASP3627, namely: (1) a high reaction temperature (3 → 5);
(2) inefficient workup procedures and complicated processes for DIBAL reduction, hydrolysis,
Wittig reaction, cyclization and deprotection (5 → 9; see Table 2, entry1 and Table 3, entry 1);
(3) multiple column chromatography purifications; (4) removal of residual palladium from 10
using column chromatography; (5) low solubility of ASP3627 in various solvents; and (6) the
need to avoid a number of reagents and solvents, namely NaH with DMF, TFA, and CH2Cl2.
Given that Wittig product 7 with SEM protection dissolved easily in n-heptane, we expected that
triphenylphosphine oxide, the byproduct of the Wittig reaction, would be removed by
crystallization with n-heptane. Therefore, SEM protection was necessary in our scale-up
synthetic strategy. The challenges we encountered in solving these problems for manufacturing
30 kg of ASP3627 are described in detail below.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1) Synthesis of (3R,4R)-5
In SEM protection of starting material 2, NaHMDS was used as a substitute for NaH,
which was adopted in the discovery conditions to give 3. NaH should be avoided for the scale-up
because it generates hydrogen as a byproduct. Additionally, NaH is known to cause a thermal
runaway reaction with DMF.9 In the SNAr reaction of 3 with commercially available chiral amine
7
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 37
1
2
3
4
5
6
7
8
9
(3R,4R)-4, the discovery conditions required a high reaction temperature (180 °C). In contrast, a
steam heating system, which is commonly used in scale-up manufacturing, requires a lower
reaction temperature (<120 °C). After optimizing the base, solvent, and temperature (Table 1),
we concluded that Na2CO3 in sulfolane at 120 °C was the best reaction condition (entry 12). This
step may be a good candidate for future flow reactions given that the reaction time at 120 °C is
still long (>18 h) for a batch process. Subsequently, development of an efficient crystallization
purification method was required to control the quality of (3R,4R)-5 due to the production of a
number of byproducts. Because most of the impurities appeared to be more polar than (3R,4R)-5,
as shown in Figure S1, we expected that crystallization with a highly polar solvent such as
alcohol and water would be sufficient for their removal. After solubility and crystallization
studies with various solvents, we found that crystallization in aqueous EtOH efficiently produced
(3R,4R)-5 with high quality (99.3 HPLC area percent purity (A%)) as expected (Figure S2).
Quality control for (3R,4R)-5 production was important for the subsequent telescoping process.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
8
ACS Paragon Plus Environment

Page 9 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
Table 1. Optimization of the SNAr reaction of 3 with 4
8
9
SEM
N
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
SEM
NH2
N
N
CN
+
NH
CN
N
Cl
Bn
N
Bn
3
4
5
HPLC A%a
entry
base
solvent
temp
isolated yield
5
3
1b
2c
3c
4c
5c
6c
7c
8c
9c
DIPEA
DIPEA
pyridine
DBU
NMP
NMP
NMP
NMP
NMP
NMP
NMP
NMP
NMP
NMP
180 °C N/A N/A
130 °C 73.4 9.4
130 °C 41.0 11.6
130 °C 37.0 0.2
130 °C 60.2 18.8
130 °C 43.1 20.5
130 °C 38.0 0.3
130 °C 80.7 0.3
130 °C 87.5 0.2
120 °C 85.9 0.9
91%
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
85%e
78%e
88%e
N/A
NMM
Et3N
Cs2CO3
K2CO3
Na2CO3
10d Na2CO3
11d Na2CO3 DMSO 120 °C 80.9 0.8
12d Na2CO3 sulfolane 120 °C 84.9 0.4
13d Na2CO3
DMAc 120 °C 70.6 1.4
a
b
Except for the solvent peak. Discovery conditions (ref 5): rac-4 (1.1 equiv), DIPEA (3.0
equiv), NMP, 180 °C, 2 h, column chromatography. (3R,4R)-4 (1.2 equiv), base (3.0 equiv),
130 °C, overnight. (3R,4R)-4 (1.2 equiv), base (1.2 equiv), 120 °C, 7−20 h. Crystallization
with EtOH/H2O (7/3). DBU 1,8-diazabicyclo[5.4.0]undec-7-ene; NMM 4-
methylmorpholine; DMSO = dimethyl sulfoxide; DMAc = N,N-dimethylacetamide.
c
d
e
=
=
9
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 37
1
2
3
4
5
6
7
8
2) Transformation from (3R,4R)-5 to (3R,4R)-9 using the telescoping process
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the transformation from (3R,4R)-5 to (3R,4R)-9, we determined that the telescoping
process would be effective for avoiding an inefficient workup and complicated process.
In general, the telescoping process has many advantages for scale-up synthesis.10
However, it is typically very challenging to control the quality of the product and to ensure
reproducibility of the process without isolating the intermediates. To develop a reproducible and
robust telescoping process, we focused on the following: (1) preparation of (3R,4R)-5 with high
quality (see purification of (3R,4R)-5 above); (2) optimization of each reaction, the workup
conditions, and whole process; and (3) efficient crystallization of (3R,4R)-9 to remove impurities
that accumulate across the reactions.
The procedure described in the previous section yielded high quality (3R,4R)-5. We
subsequently attempted to optimize the process. In the transformation from (3R,4R)-5 to
(3R,4R)-7, the workup conditions needed to be optimized. Specifically, Al-related impurities
generated from DIBAL and triphenylphosphine oxide generated from the Wittig reagent needed
to be efficiently removed without column chromatography. The discovery conditions required
filtration of the Al-related impurities and subsequent column chromatography (Table 2, entry 1).
Table 2 shows the findings from our attempts to remove Al-related impurities. After DIBAL
reduction and quenching with MeOH, Al-related impurities were efficiently removed by washing
with aqueous Rochelle salt. Under these extraction conditions, separation of the two layers was
sufficient for scale-up synthesis. Subsequently, hydrolysis with aqueous HCl, followed by
neutralization with aqueous NaOH gave (3R,4R)-6 (Table 2, entry 2). That washing with aqueous
Rochelle salt after hydrolysis resulted in poor separation of layers suggested that the order in
which these processes were performed was very important for ease of operation and reducing the
10
ACS Paragon Plus Environment

Page 11 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
reaction time (Table 2, entry 3). In the Wittig reaction (6 → 7), removal of the byproduct
triphenylphosphine oxide is typically a serious problem.11 We developed an efficient purification
method that employed crystallization with n-heptane. The residual amount of triphenylphosphine
oxide was reduced from 3.9 A% to 0.2 A% using this crystallization process. This removal
method can be applied to any products that dissolve in n-heptane.12 In the case of intermediate 7,
the SEM group probably enhanced its solubility in n-heptane. Therefore, SEM protection was
necessary in our synthetic strategy to purify the Wittig product by crystallization of the
byproduct.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the cyclization and deprotection sequence for transformation from (3R,4R)-7 to
(3R,4R)-9, we attempted to develop a process that solved the following problems: (1) inefficient
two-reaction sequence via dimethyl acetal intermediate 13 for cyclization; (2) a complicated
workup requiring column chromatography; and (3) the use of 18 equiv of TFA, a corrosive acid,
and CH2Cl2, which has a negative environmental impact, for deprotection (Table 3, entry 1).
Table 3 shows the optimization of the cyclization and deprotection processes. We found that
aqueous HCl/MeOH or DME conditions yielded 8 in one reaction (Table 3, entries 2 and 3).
Moreover, in the subsequent reaction, the first step of deprotection was conducted in “one pot”
by adding another 12 equivalents of aqueous HCl to the reaction mixture to yield the alcohol
intermediate (3R,4R)-14. The resulting product contained triphenylphosphine derived from the
Witting reagent, which needed to be removed because residual triphenylphosphine would poison
the palladium catalyst used in the next step.13 To remove triphenylphosphine, phase split was
attempted using n-heptane. As expected, most of the triphenylphosphine was purged into the n-
heptane layer and (3R,4R)-14 was obtained in the aqueous layer. After the subsequent
crystallization of (3R,4R)-9, the amount of triphenylphosphine was successfully reduced (0.6
11
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 37
1
2
3
4
5
6
7
8
9
A%) such that it would not affect the next debenzylation step. Addition of aqueous NaOH to the
aqueous layer containing (3R,4R)-14 progressed the second step of deprotection to give crude
(3R,4R)-9 (Table 3, entry 3).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
12
ACS Paragon Plus Environment

Page 13 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
Table 2. Optimization of workup process of DIBAL reduction and hydrolysis reaction.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
hydrolysis
DIBAL reduction
R =
6
5
12
DIBAL reduction
reaction workup
hydrolysis
yield of
entry
6
reaction
workup
1. neutralize with 1 M aq
NaOH, then filter
1 M aq
HCl, 0 °C,
1 M DIBAL/n-
hexane (2.5 equiv),
THF, 0 °C, 1 h
2. extract with EtOAc
3. wash with water
1a
add MeOH
82%
99%
20 min
4. dry over MgSO4
5. column chromatography
1. neutralize with 5 M aq
NaOH
1. add MeOH
1 M
3 M aq
HCl, rt,
DIBAL/toluene
(2.5 equiv), THF,
−50 °C, then 0 °C,
1 h
2. wash with aq
Rochelle salt
(good
2
3
2. separate
45 min
3. wash with aq NaCl and
water
separation)
1. neutralize with 3 M aq
NaOH
1 M
3 M aq
HCl, rt,
2. add aq Rochelle salt
(bad separation)
DIBAL/toluene
(2.5 equiv), THF,
−60 °C, then 0 °C,
1 h
add MeOH
90%
3.5 h
3. extract with CPME
4. wash with aq NaCl and
water
a Discovery conditions (ref 5). CPME = cyclopentyl methyl ether.
13
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 37
1
2
3
4
5
6
7
Table 3. Optimization of cyclization and removal of SEM.
8
9
OH
Si
H
N
Si
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
O
N
N
cyclization
deprotection
1st step
N
N
N
N
2nd step
N
R
N
OMe
N
R
NHR
R
7
8
14
9
Si
O
N
N
R =
N
OMe
OMe
NHR
13
cyclization
deprotection
entry
reaction
workup
first step
second step
1. concentrate
1. ethylenediamine (3.0
equiv), 1 M aq NaOH,
CH2Cl2, rt, 2 h
1. TFA (18 equiv),
CH2Cl2, rt, overnight
2. add aq NaHCO3
2. concentrate
3. extract with
EtOAc
1. AcCl (3.0
equiv), MeOH,
80 °C, 0.5 h
2. add water
3. add aq NaHCO3
3. extract with IPA/CHCl3
4. wash with aq NaCl
5. dry over MgSO4
1a
4. wash with aq NaCl
5. dry over MgSO4
4. extract with
IPA/CHCl3
2. H2O, 80 °C, 1
h
5. wash with aq NaOH
(×2) and aq NaCl
6. column
chromatography
6. slurry wash with IPA/IPE
74% yield
6. dry over MgSO4
83% yield from 6
1. add aq NaHCO3
concd aq HCl
(3.4 equiv),
MeOH (5 vol),
50 °C, 6 h, IPCb:
8/7 = 82.4/0.8
6 M aq HCl (15
equiv), DME (5 vol),
50 °C, 18 h, IPCb: 14/8
= 78.0/0.2
2. extract with
EtOAc
add 5 M aq NaOH (18
equiv), 75 °C, 0.5 h, IPCb:
9/14 = 76.1/0.5
2
3. wash with aq NaCl
4. concentrate
6 M aq HCl (3.0
equiv), DME (5
vol), 50 °C, 4.5
h, IPCb: 8/7 =
84.5/0.4
add 6 M aq HCl (12
equiv), 50 °C, 21 h,
IPCb: 14/8 = 81.9/0.2,
then wash with n-
heptane
add 5 M aq NaOH (18
equiv), 75 °C, 0.5h, IPCb:
9/14 = 80.5/0.6
3
none
a
b
Discovery conditions (ref 5). HPLC A%. DME = 1,2-dimethoxyethane; IPA = 2-propanol;
IPE = diisopropyl ether.
14
ACS Paragon Plus Environment

Page 15 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Despite developing a practical synthetic method for the production of crude (3R,4R)-9, a
large number of impurities were observed. Therefore, we developed the workup process below to
obtain pure (3R,4R)-9. First, to remove highly polar water-soluble impurities, the reaction
mixture was diluted with EtOAc and washed with aqueous NH4Cl followed by water at 50 °C.
This improved the quality from 80.5 A% to 84.2 A%. Second, the crystallization conditions were
developed after completing the solubility and crystallization experiments using various solvents.
Finally, pure (3R,4R)-9 (94.9 A%) was obtained by performing the crystallization process under
aqueous EtOH conditions in 81.9% yield with 9.1% loss to the mother liquor. As described
above, the six-reaction telescoping process—DIBAL reduction, hydrolysis, Wittig reaction,
cyclization, and two-step removal of the SEM group—was successfully completed without the
need for column chromatography purification. The whole process of this six-reaction sequence
was optimized for the following: (1) selection of the appropriate order of hydrolysis and Al
removal; (2) removal of triphenylphosphine in the back-extractable intermediate; and (3) one pot
reaction in the same solvent as that used for the cyclization and deprotection. In addition, we set
appropriate in-process controls (IPCs) to control the quality of all intermediates. The criteria for
IPCs are described in the experimental section.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3) Synthesis of 1 and removal of Pd
Debenzylation of (3R,4R)-9 was conducted by employing Pearlman's catalyst
(Pd(OH)2/C), H2, and HCl. After optimizing the amount of HCl, we found that the reaction was
highly reproducible using two equivalents of HCl. This result indicates that the protonation of
two nitrogens of pyrrolopyridine and piperidine with two equivalents of acid suppressed the
15
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 37
1
2
3
4
5
6
7
8
9
poisoning of palladium and improved reproducibility. Meanwhile, residual palladium up to 560
ppm remained present in (3R,4R)-10 when left untreated. According to the ICH Q3D guideline,
the Permitted Daily Exposure (PDE) for oral exposure to palladium is 100 μg/day.14 The amount
of residual palladium in the API must therefore be strictly controlled to ensure that it is below the
acceptable level. Although approximately 85% of residual palladium could be removed in
subsequent steps, this was not sufficient to meet the criterion. Our efforts to remove palladium
are shown in Table 4. We found that palladium could be efficiently removed using a combination
of activated carbon and 2,4,6-trimercapto-s-triazine (TMT)15 (Table 4, entry 2), a combination of
activated carbon and ʟ-cysteine15a,16 (Table 4, entry 3), or ʟ-cysteine alone without activated
carbon (Table 4, entry 4). When activated carbon was used, the additive had to be filtered out
before crystallization of (3R,4R)-10 (Table 4, entries 1−3). Further, addition of carbon into the
reactor is undesirable for cleaning of the reaction vessel under Good Manufacturing Practice
(GMP) and there is a risk of carbon contamination. In contrast, use of only ʟ-cysteine (Table 4,
entry 4) did not require additional operations like filtration to remove the additive because ʟ-
cysteine was dissolved in the crystallization solvent and purged into the filtrate. We therefore
adopted the use of only ʟ-cysteine. The resulting simple procedure developed to remove residual
palladium is shown in Figure 2. Although N-acetylcysteine is known to work as a Pd scavenger
in the crystallization,15a,17 to our knowledge, the method of removing Pd to the mother liquor
using ʟ-cysteine has never been reported.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Condensation of (3R,4R)-10 and cyanoacetic acid (11) was performed using EDC·HCl
and HOBt·H2O to yield crude 1.18 However, an impurity that was thought to be the overreacted
compound was observed in the reaction mixture. Addition of aqueous HCl to the reaction
mixture successfully hydrolyzed this impurity to give the desired compound 1 as expected.
16
ACS Paragon Plus Environment

Page 17 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
For the final purification, we investigated the solubility of 1 and found that it was very
low in a number of solvents except for DMF and DMSO. However, we determined that DMF
and DMSO should be avoided for the following reasons: (1) DMF is categorized as a Class 2
solvent in the ICH Q3C guideline;19 and (2) recrystallization of 1 using aqueous DMSO
deteriorated the solution’s filterability, and the level of residual DMSO in the API was
unacceptably high. Although free form 1 had low solubility, the HCl salt of 1 dissolved easily in
aqueous HCl/EtOH. After the polish filtration, aqueous NaOH was added to produce a slurry of
free form 1. Following filtration and drying, ASP3627 (1) was obtained with high quality (≥99.0
A%, residual EtOH: 0.2%).20
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ultimately, the target compound was prepared using six isolations in an overall yield of
52.9% at the lab-scale.
17
ACS Paragon Plus Environment

Organic Process Research & Development
Page 18 of 37
1
2
3
4
5
6
7
Table 4. Removal of Pd from 10.a
8
9
entry
1
Pd initial
560 ppm
additive
removal of additive
filtration
Pd residue
160 ppm
adsorption
71.4%
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
activated carbon (0.1 wt)
activated carbon (0.1 wt),
TMT (0.1 wt)
2
560 ppm
filtration
30 ppm
94.6%
activated carbon (0.1 wt),
ʟ-cysteine (0.1 wt)
3
4
560 ppm
560 ppm
filtration
none
8.8 ppm
17 ppm
98.4%
97.0%
ʟ-cysteine (0.1 wt)
a Reagents and conditions. 10 (0.3 g), additive, 6 M aq HCl (2.0 equiv), EtOH/H2O (3/5), rt, 1 h,
followed by filtration of additive (entries 1−3), neutralization, then crystallization with aq EtOH.
Figure 2. Scale-up procedure for (3R,4R)-10.
20% Pd(OH)2/C
(0.05 wt, dry basis)
EtOH/H2O (1/2)
(3 vol) wash
9,
stirring under
H2 (0.2 MPa),
30−40 °C
filtration
6 M aq HCl (2.0 equiv),
EtOH (2 vol), H2O (3 vol)
ʟ-cysteine (0.1 wt)
5 M aq NaOH
stirring 1 h,
rt
aging 1 h,
20−30 °C
filtration
pH 9.5−10.0
EtOH/H2O (3/7)
(4 vol) wash
drying at 50 °C
10
18
ACS Paragon Plus Environment

Page 19 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
Scale-up synthesis
8
9
We used the optimized methods described above to demonstrate the first multi-kilogram
synthesis of ASP3627 (1). The scaled-up manufacturing process effectively reproduced the lab
experiments to prepare 30 kg of 1 with high purity (99.2 A%) in excellent overall yield (54.6%)
(Table 5).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 5. Scale-up synthesis of 1.
step 1
step 2
step 3
step 4
step 5,6
2
3
(3R,4R)-5
(3R,4R)-9
(3R,4R)-10
ASP3627 (1)
lab experiments
first scale-up synthesis
step
reaction
yield
(%)
purity (HPLC
A%)
yield
(%)
purity (HPLC
A%)
start
(kg)
target
(kg)
1
2
SEM protection
SNAr reaction
84.1
90.7
97.9
99.3
86.9
92.3
98.0
99.5
36.0
52.0
54.2
74.2
six-reaction
telescoping
3
4
81.9
96.4
94.9
97.1
80.8
95.1
96.8
98.0
69.0
39.7
40.4
removal of Bn
condensation
27.9a
91.0;
90.9
98.7;
98.8
13.6;
13.6
15.7;
15.7
5b
6
90.3
97.2
98.8
99.3
recrystallization
97.4
54.6
99.2
31.1
30.3
overall
52.9
a
b
Residual palladium: 5 ppm. Step 5 for the first scale-up procedure was performed in two
batches at the same scale.
19
ACS Paragon Plus Environment

Organic Process Research & Development
Page 20 of 37
1
2
3
4
5
6
7
CONCLUSION
8
9
We developed a practical, scalable, and highly reproducible method for the synthesis of
ASP3627 that features the following: (1) elimination of the need for a high reaction temperature
to synthesize (3R,4R)-5 by setting appropriate reaction conditions, (2) a challenging and efficient
six-reaction telescoping process, (3) no column chromatography purification, (4) a simple
palladium removal method using ʟ-cysteine, (5) a robust debenzylation reaction condition using
two equivalents of HCl, and (6) recrystallizaion of a low solubility API by neutralization. This
method was used to successfully perform the first scale-up synthesis of ASP3627, producing 30
kg with high quality in 54.6% overall yield over six steps.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
General methods
Unless otherwise noted, all reactions were performed under a nitrogen atmosphere. All
reagents and solvents purchased from suppliers were used as received unless otherwise noted.
(3R,4R)-4 was obtained commercially from a custom synthesis supplier. NMR spectra were
recorded on a Bruker ADVANCE III HD500. Chemical shifts (δ) are reported in ppm in
reference to the residual solvent signal (δ 2.50 for 1H NMR in DMSO-d6, δ 39.52 for 13C{1H}
NMR in DMSO-d6, and δ 77.16 for 13C{1H} NMR in CDCl3). High-resolution mass spectra were
obtained on a Thermo Scientific EXACTIVE Plus. Elemental analyses were performed on a
Elementar Vario MICRO cube and DIONEX ICS-3000-I. IR spectra were recorded on a
SHIMADZU IRAffinity-1S.
20
ACS Paragon Plus Environment

Page 21 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
HPLC conditions
8
9
Method A. YMC-Pack ODS-A (5 μm, 12 nm), 150 × 4.6 mm I.D., eluent: 0.01 M aqueous
KH2PO4/MeCN = 25/75, flow rate: 1.0 mL/min, temperature: 40 °C, detection: UV at 254 nm.
Retention time: 2.1 min for 2, 6.7 min for 3, 20.6 min for 5, 16.9 min for 12, 18.3 min for 6, 5.3
min for 9.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Method B. YMC-Pack ODS-A (5 μm, 12 nm), 150 × 4.6 mm I.D., eluent: 0.02 M aqueous
K2HPO4 (pH 7.0)/MeCN = 25/75, flow rate: 1.0 mL/min, temperature: 40 °C, detection: UV at
254 nm. Retention time: 18.3 min for 6, 2.3 min for triphenylphosphine oxide, 20.3 min for (Z)-7,
22.4 min for (E)-7, 24.8 min for 8, 4.2 min for 14, 9.4 min for triphenylphosphine, 5.3 min for 9.
Method C. YMC-Pack Pro C18 (5 μm, 12 nm), 150 × 4.6 mm I.D., eluent A: 0.02 M aqueous
K2HPO4 (pH 7.0), eluent B: MeCN, flow rate: 1.0 mL/min, temperature: 40 °C, detection: UV at
254 nm. Gradient: 0 → 5 min, A/B = 75/25; 5 → 15 min, A/B = 75/25 → 30/70; 15 → 40 min,
A/B = 30/70. Retention time: 19.8 min for 9, 4.0 min for 10.
Method D. YMC-Pack Pro C18 (5 μm, 12 nm), 150 × 4.6 mm I.D., eluent A: 0.02 M aqueous
K2HPO4 (pH 7.0), eluent B: MeCN, flow rate: 1.0 mL/min, temperature: 40 °C, detection: UV at
254 nm. Gradient: 0 → 12 min, A/B = 75/25; 12 → 20 min, A/B = 75/25 → 50/50; 20 → 40 min,
A/B = 50/50. Retention time: 4.0 min for 10, 21.9 min for overreacted compound, 10.0 min for 1.
Method E. CHIRALPAK IC (5 μm), 250 × 4.6 mm I.D., eluent: 0.02 M aqueous NH4HCO3 (pH
9.0)/MeCN = 67/33, flow rate: 1.0 mL/min, temperature: 25 °C, detection: UV at 222 nm.
Retention time: 15.2 min for 1, 17.6 min for ent-1.
21
ACS Paragon Plus Environment

Organic Process Research & Development
Page 22 of 37
1
2
3
4
5
6
7
8
9
4-Chloro-1-{[2-(trimethylsilyl)ethoxy]methyl}-1H-pyrrolo[2,3-b]pyridine-5-carbonitrile (3). A
500 L reactor-1 was charged with 2 (36.0 kg, 203 mol) and DMF (170.1 kg, 5 vol). NaHMDS
(38.1% in THF, 107.3 kg, 1.1 equiv, 223 mol) was added to the resulting slurry at ≤10 °C, and
the mixture was stirred for ≥30 min at 0−10 °C. Following the addition of SEMCl (37.2 kg, 1.1
equiv, 223 mol) over ≥30 min at 0−10 °C, the reaction mixture was stirred for ≥1 h at 0−10 °C
(IPC: 2 ≤ 1.0 A% by HPLC Method A). The reaction mixture was transferred to a 2000 L
reactor-2 and quenched with aqueous NH4Cl (prepared from 72.0 kg (2 wt) of NH4Cl and 360 L
(10 vol) of water), and EtOAc (97.4 kg, 3 vol) and Radiolite (18.0 kg, 0.5 wt) were added at
≤30 °C. The mixture was stirred for ≥1 h at 20−30 °C, filtered, and the wet cake was washed
with EtOAc (64.9 kg, 2 vol). The filtrate was transferred to a 1500 L reactor-3. The biphasic
mixture was separated. The organic layer was washed with aqueous NH4Cl (prepared from 36.0
kg (1 wt) of NH4Cl and 360 L (10 vol) of water) (×2) at 10−30 °C, transferred to a 500 L reactor-
4 and reactor-3 was rinsed with EtOAc (20.0 kg, 0.6 vol). The mixture was concentrated under
vacuum at 50 °C to ≤108 L (3 vol). EtOH (113.6 kg, 4 vol) was added to the residue and the
mixture was concentrated under vacuum at 50 °C to ≤108 L (3 vol). More EtOH (113.6 kg, 4
vol) was added to the residue and the mixture was concentrated under vacuum at 50 °C to
97−119 L (3 vol). EtOH (113.6 kg, 4 vol) was again added, the mixture was warmed to
45−55 °C, and stirred for ≥1 h. The solution was cooled to 20−30 °C (approximately 10 °C/ h),
and stirred for ≥1 h. Water (108 L, 3 vol) was added over ≥30 min at 20−30 °C. The slurry was
aged for ≥2 h, cooled to −5 to 5 °C (approximately 10 °C/h), aged for ≥5 h (IPC: 3 ≤ 4 g/L by
HPLC Method A), and then filtered. The wet cake was washed with cooled aqueous EtOH
(prepared from 73.9 kg (2.6 vol) of EtOH and 47 L (1.3 vol) of water), dried at 50 °C (internal
temperature) for ≥1 h under reduced pressure (IPC: loss on drying (LOD) ≤ 1.0%) to give 35
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
22
ACS Paragon Plus Environment

Page 23 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
(54.2 kg, 176 mol, 86.9% yield) as a pale yellow solid. HPLC purity: 98.0 A% (specification:
≥95.0 A% by HPLC Method A). 1H NMR (DMSO-d6, 500 MHz): δ –0.13 (s, 9H), 0.80 (t, 2H, J
= 8.0 Hz), 3.51 (t, 2H, J = 8.0 Hz), 5.66 (s, 2H), 6.74 (d, 1H, J = 3.7 Hz), 7.97 (d, 1H, J = 3.7
Hz), 8.70 (s, 1H). 13C{1H} NMR (DMSO-d6, 125 MHz): δ –1.5, 17.1, 65.8, 73.0, 99.7, 101.5,
115.7, 118.8, 133.0, 137.3, 146.6, 148.5. HRMS–ESI (m/z): [M + H]+ calcd for C14H19N3OClSi,
308.0980; found, 308.0981. Anal. Calcd for C14H18N3OClSi: C, 54.62; H, 5.89; N, 13.65; Cl,
11.52. Found: C, 55.00; H, 5.99; N, 13.66; Cl, 11.11. IR (ATR, cm–1): 3088, 2955, 2926, 2220,
1589, 1346, 1248, 1233, 1070, 858, 833, 737, 698. DSC (onset): 66.4 °C.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4-{[(3R,4R)-1-Benzyl-4-methylpiperidin-3-yl]amino}-1-{[2-(trimethylsilyl)ethoxy]methyl}-1H-
pyrrolo[2,3-b]pyridine-5-carbonitrile (5). A 500 L reactor-1 was charged with 3 (52.0 kg, 169
mol), (3R,4R)-4 (41.4 kg, 1.2 equiv, 203 mol), and sulfolane (524.2 kg, 8 vol). Na2CO3 (21.5 kg,
1.2 equiv, 203 mol) was added to the resulting solution, and the reaction mixture was warmed to
115−125 °C and stirred for ≥18 h (IPC: 3 ≤ 1.0 A% by HPLC Method A). The reaction mixture
was cooled to 10−30 °C, transferred to a 2000 L reactor-2, diluted with EtOAc (469.0 kg, 10 vol),
and quenched with aqueous NaCl (prepared from 15.6 kg (0.3 wt) of NaCl and 780 L (15 vol) of
water). The biphasic mixture was separated. The organic layer was washed with aqueous NaCl
(prepared from 10.4 kg (0.2 wt) of NaCl and 520 L (10 vol) of water) and aqueous NaCl
(prepared from 104.0 kg (2 wt) of NaCl and 520 L (10 vol) of water) at 10−30 °C. The resulting
solution was transferred to a 1500 L reactor-3 and reactor-2 was rinsed with EtOAc (20.0 kg, 0.4
vol). The solution was concentrated under vacuum at 50 °C to ≤156 L. EtOH (164.1 kg, 4 vol)
was added to the residue and the mixture was concentrated under vacuum at 50 °C to ≤156 L (3
vol). More EtOH (164.1 kg, 4 vol) was added the residue and the mixture was concentrated
23
ACS Paragon Plus Environment

Organic Process Research & Development
Page 24 of 37
1
2
3
4
5
6
7
8
9
under vacuum at 50 °C to 146−166 L (3 vol). EtOH (205.1 kg, 5 vol) was again added to the
residue, the mixture was cooled to −5 to 5 °C, seeded with (3R,4R)-5 (1.0 g), and aged for ≥1 h.
Water (156 L, 3 vol) was added over ≥30 min, the slurry was aged for ≥1 h at −5 to 5 °C,
warmed to 40−50 °C, and aged for ≥1 h. The slurry was cooled to −5 to 5 °C (approximately
10 °C/h), aged for ≥5 h (IPC: 5 ≤ 5 g/L by HPLC Method A), and then filtered. The wet cake
was washed with cooled aqueous EtOH (prepared from 114.9 kg (2.8 vol) of EtOH and 62 L (1.2
vol) of water), dried at 50 °C (internal temperature) for ≥1 h under reduced pressure (IPC: LOD
≤ 1.0%) to give (3R,4R)-5 (74.2 kg, 156 mol, 92.3% yield) as a pale yellow solid. HPLC purity:
99.5 A% (specification: ≥97.0 A% by HPLC Method A). 1H NMR (DMSO-d6, 500 MHz): δ –
0.13 (s, 9H), 0.77 (t, 2H, J = 8.0 Hz), 0.83 (d, 3H, J = 6.7 Hz), 1.42–1.56 (m, 2H), 1.81–1.93 (m,
1H), 2.05–2.16 (m, 1H), 2.27 (d, 1H, J = 10.6 Hz), 2.73–2.87 (m, 2H), 3.43 (d, 1H, J = 13.4 Hz),
3.47 (t, 2H, J = 8.0 Hz), 3.56 (d, 1H, J = 13.4 Hz), 4.39–4.47 (m, 1H), 5.53 (s, 2H), 6.31 (d, 1H,
J = 9.9 Hz), 6.84 (d, 1H, J = 3.8 Hz), 7.17 (t, 1H, J = 7.3 Hz), 7.24 (t, 2H, J = 7.5 Hz), 7.35 (d,
2H, J = 7.3 Hz), 7.43 (d, 1H, J = 3.8 Hz), 8.19 (s, 1H). 13C{1H} NMR (DMSO-d6, 125 MHz): δ –
1.5, 17.1, 17.4, 28.8, 33.7, 51.8, 52.6, 57.4, 61.6, 65.4, 72.5, 84.2, 101.3, 104.3, 118.5, 125.9,
126.9, 128.1, 128.5, 138.3, 147.5, 149.4, 149.8. HRMS–ESI (m/z): [M + H]+ calcd for
C27H38N5OSi, 476.2840; found, 476.2842. Anal. Calcd for C27H37N5OSi: C, 68.17; H, 7.84; N,
14.72. Found: C, 68.74; H, 7.98; N, 14.73. [α]D23 – 27.7 (c 1.00, EtOH). IR (ATR, cm–1): 3339,
2951, 2922, 2818, 2199, 1591, 1557, 1524, 1512, 1225, 1080, 858, 833, 712, 698. DSC (onset):
75.2 °C.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1-[(3R,4R)-1-Benzyl-4-methylpiperidin-3-yl]-1,6-dihydrodipyrrolo[2,3-b:2',3'-d]pyridine (9). A
1500 L reactor-1 was charged with (3R,4R)-5 (69.0 kg, 145 mol) and THF (368.0 kg, 6 vol). The
24
ACS Paragon Plus Environment

Page 25 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
resulting solution was cooled to −65 to −45 °C, and DIBAL (17% in toluene, 300.4 kg, 2.5 equiv,
363 mol) was added at −65 to −45 °C. The reaction mixture was warmed to −5 to 5 °C, and
stirred for ≥1 h (IPC: 5 ≤ 0.3 A% by HPLC Method A). The reaction was cooled to −65 to
−45 °C, quenched with MeOH (27.3 kg, 0.5 vol), and transferred to a 5000 L reactor-2. Aqueous
Rochelle salt (prepared from 207.0 kg (3 wt) of Rochelle salt and 690 L (10 vol) of water) was
added to the solution at ≤10 °C. The biphasic mixture was stirred vigorously for ≥1 h at
15−25 °C and separated to give (3R,4R)-12 in toluene solution (the organic layer).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Aqueous HCl (prepared from 162.8 kg (1599 mol) of concd aqueous HCl and 414 L (6 vol) of
water) was added to the toluene solution of (3R,4R)-12, and the reaction mixture was stirred for
≥3 h at 15−25 °C. The pH of the aqueous layer was adjusted to 8−9 by adding 5 M aqueous
NaOH (359.8 kg, 1497 mol) at 15−25 °C. The biphasic mixture was separated. The organic layer
was washed with aqueous NaCl (prepared from 69.0 kg (1 wt) of NaCl and 345 (5 vol) L of
water) (×2) at 15−25 °C, transferred to a 1500 L reactor-3, and reactor-2 was rinsed with THF
(20.0 kg, 0.3 vol). The solution was concentrated under vacuum at 50 °C to ≤138 L (2 vol).
Toluene (298.8 kg, 5 vol) was added to the residue and the mixture was concentrated under
vacuum at 50 °C to ≤138 L (2 vol). More toluene (298.8 kg, 5 vol) was added to the residue and
the mixture was concentrated under vacuum at 50 °C to 128−148 L (2 vol, IPC: residual water ≤
0.10%). THF (184.0 kg, 3 vol) was added to the residue to give (3R,4R)-6 in THF solution.
A 1500 L reactor-4 was charged with (methoxymethyl)triphenylphosphonium chloride (74.6 kg,
1.5 equiv, 218 mol) and THF (184.0 kg, 3 vol). The resulting slurry was stirred, cooled to −20 to
−10 °C, and NaHMDS (38.7% in THF, 137.5 kg, 2.0 equiv, 290 mol) was added over ≥30 min at
−20 to −10 °C. The mixture was stirred for ≥1 h, the THF solution of (3R,4R)-6 was added over
≥30 min at −20 to −10 °C, and reactor-3 was rinsed with THF (122.7 kg, 2 vol). The reaction
25
ACS Paragon Plus Environment

Organic Process Research & Development
Page 26 of 37
1
2
3
4
5
6
7
8
9
mixture was warmed to 15−25 °C and stirred (IPC: 6 ≤ 1.0 A% by HPLC Method B). The
reaction was quenched with aqueous NH4Cl (prepared from 69.0 kg (1 wt) of NH4Cl and 345 L
(5 vol) of water), transferred to a 5000 L reactor-5, and reactor-4 was rinsed with THF. The
biphasic mixture was separated at 0−30 °C. The organic layer was washed with aqueous NaCl
(prepared from 69.0 kg (1 wt) of NaCl and 345 L (5 vol) of water) at 0−30 °C, transferred to a
1500 L reactor-6, and reactor-5 was rinsed with n-heptane (13.7 kg, 0.3 vol). The mixture was
concentrated under vacuum at 50 °C to ≤207 L (3 vol). n-Heptane (236.0 kg, 5 vol) was added to
the residue and the mixture was concentrated under vacuum at 50 °C to ≤207 L (3 vol). More n-
heptane (236.0 kg, 5 vol) was added to the residue and the mixture was concentrated under
vacuum at 50 °C to 197−217 L (3 vol). n-Heptane (424.8 kg, 9 vol) was again added to the
residue at 45−55 °C. The slurry was cooled to −5 to 5 °C (approximately 10 °C/h), aged for ≥10
h, warmed to approximately 50 °C, stirred for 3 h, cooled to approximately 30 °C (IPC:
triphenylphosphine oxide ≤ 0.7 A%, by HPLC Method B), and then filtered. The wet cake was
washed with n-heptane (188.8 kg, 4 vol). The filtrate was transferred to a 2000 L reactor-7 and
concentrated under vacuum at 50 °C to ≤138 L (2 vol). DME (297.7 kg, 5 vol) was added to the
residue and the mixture was concentrated under vacuum at 50 °C to 128−148 L (2 vol). More
DME (238.2 kg, 4 vol) was added to the residue to give an E/Z mixture of (3R,4R)-7 in DME
solution.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Aqueous HCl (3.0 equiv, 436 mol, prepared from 44.5 kg of concd aqueous HCl and 37 L of
water) was added to the E/Z mixture of (3R,4R)-7 in DME solution, the reaction mixture was
warmed to 45−55 °C, and stirred for ≥3 h (IPC: 7 ≤ 1.0 A% by HPLC Method B) to give
(3R,4R)-8 in aqueous DME solution.
26
ACS Paragon Plus Environment

Page 27 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Aqueous HCl (12.0 equiv, 1742 mol, prepared from 177.3 kg of concd aqueous HCl and 145 L
of water) was added to the aqueous DME solution of (3R,4R)-8 at 45−55 °C, and the reaction
mixture was stirred for ≥19 h (IPC: 8 ≤ 1.0 A% by HPLC Method B). n-Heptane (94.4 kg, 2 vol)
was added to the reaction mixture and the biphasic mixture was separated to give (3R,4R)-14 in
aqueous DME solution (the aqueous layer).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A 6000 L reactor-8 was charged with NaOH (104.4 kg, 18.0 equiv, 2610 mol), water (524 L, 7.6
vol), and DME (59.5 kg, 1 vol). The aqueous DME solution of (3R,4R)-14 was added to the
reaction mixture and the biphasic reaction mixture was stirred for ≥0.5 h at 70−80 °C (IPC: 14 ≤
1.5 A% in organic layer by HPLC Method B). A 5000 L reactor-9 was charged with NH4Cl (69.0
kg, 1 wt) and water (345 L, 5 vol). EtOAc (373.4 kg, 6 vol) was added to the solution and the
mixture was stirred at 45−65 °C. The reaction mixture in reactor-8 was added to the biphasic
mixture in reactor-9 at 45−65 °C and reactor-8 was rinsed with DME (29.8 kg, 0.5 vol). The
biphasic mixture was separated. The organic layer was washed with water (345 L, 5 vol) at
45−65 °C, diluted with EtOAc (124.5 kg, 2 vol), and then washed with water (345 L, 5 vol) at
45−65 °C. The organic layer was transferred to a 1500 L reactor-10 and reactor-9 was rinsed
with EtOAc (20.0 kg, 0.3 vol). The solution was concentrated under vacuum at 50 °C to ≤207 L
(3 vol). EtOH (272.2 kg, 5 vol) was added to the residue and the mixture was concentrated under
vacuum at 50 °C to ≤207 L (3 vol). More EtOH (272.2 kg, 5 vol) was added to the residue and
the mixture was concentrated under vacuum at 50 °C to 197−217 L (3 vol). EtOH (435.5 kg, 8
vol) was again added to the residue, the mixture was warmed to 55−65 °C, and stirred for ≥1 h.
The solution was cooled to 20−30 °C (approximately 10 °C/h) and stirred for ≥2 h. Water (345 L,
5 vol) was added to the slurry over ≥30 min at 20−30 °C, and the slurry was cooled to −5 to 5 °C
(approximately 10 °C/h), aged for ≥3 h (IPC: 9 ≤ 5 g/L by HPLC Method B), and then filtered.
27
ACS Paragon Plus Environment

Organic Process Research & Development
Page 28 of 37
1
2
3
4
5
6
7
8
9
The wet cake was washed with cooled aqueous EtOH (prepared from 141.5 kg (2.6 vol) of EtOH
and 90 L (1.3 vol) of water), dried at 50 °C (internal temperature) for ≥1 h under reduced
pressure (IPC: LOD ≤ 1.0%) to give (3R,4R)-95 (40.4 kg, 117 mol, 80.8% yield) as a white solid.
HPLC purity: 96.8 A% (specification: ≥94.0 A% by HPLC Method A). 1H NMR (DMSO-d6,
500 MHz): δ 0.51 (d, 3H, J = 6.8 Hz), 1.58–1.70 (m, 2 H), 2.13–2.31 (m, 2H), 2.67 (dd, 1H, J =
11.8, 3.2 Hz), 2.84–2.94 (m, 1H), 3.04 (dd, 1H, J = 11.8, 3.8 Hz), 3.53 (d, 1H, J = 13.2 Hz), 3.59
(d, 1H, J = 13.2 Hz), 5.01 (q, 1H, J = 4.0 Hz), 6.61 (d, 1H, J = 3.3 Hz), 6.74–6.77 (m, 1H), 7.22
(tt, 1H, J = 7.2, 1.4 Hz), 7.28–7.37 (m, 5H), 7.99 (br s, 1H), 8.43 (s, 1H), 11.6 (br s, 1H).
13C{1H} NMR (CDCl3, 125 MHz): δ 16.0, 30.2, 34.1, 52.2, 56.1, 56.8, 63.5, 97.6, 101.5, 105.1,
118.3, 121.3, 126.3, 127.3, 128.4, 129.1, 135.8, 137.7, 138.3, 145.2. HRMS–ESI (m/z): [M + H]+
calcd for C22H25N4, 345.2074; found, 345.2071. Anal. Calcd for C22H24N4: C, 76.71; H, 7.02; N,
16.27. Found: C, 76.70; H, 7.08; N, 16.19. [α]D23 – 28.5 (c 1.00, 0.1 M aq HCl). IR (ATR, cm–1):
3111, 3028, 2926, 2808, 1612, 1582, 1452, 1350, 1327, 885, 837, 820, 731, 719, 698. DSC
(onset): 185.0 °C.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1-[(3R,4R)-4-Methylpiperidin-3-yl]-1,6-dihydrodipyrrolo[2,3-b:2',3'-d]pyridine (10). A 500 L
reactor-1 was charged with (3R,4R)-9 (39.7 kg, 115 mol), EtOH (62.6 kg, 2 vol), and water (119
L, 3 vol). Aqueous HCl (2.0 equiv, 230 mol, prepared from 23.4 kg of concd aqueous HCl and
19 L of water) and 20% Pd(OH)2/C (wetted with water, 2.0 kg, 0.05 wt, dry basis) were added to
the resulting mixture. The reaction mixture was stirred for 6 h at 30−40 °C under a H2
atmosphere (0.15−0.25 MPa) (IPC: 9 ≤ 1.0 A% by HPLC Method C). The reaction mixture was
cooled to 20−30 °C, filtered, and the wet cake was washed with aqueous EtOH (prepared from
31.3 kg (1 vol) of EtOH and 79 L (2 vol) of water). The filtrate was transferred to a 500 L
28
ACS Paragon Plus Environment

Page 29 of 37
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
reactor-2 and ʟ-cysteine (4.0 kg, 0.1 wt) was added. The mixture was stirred for ≥1 h at
20−30 °C. The pH of the solution was adjusted to 9.5−10.0 by adding 5 M aqueous NaOH (63.1
kg, 263 mol) at 20−30 °C. The slurry was aged for ≥1 h at 20−30 °C (IPC: 10 ≤ 2 g/L by HPLC
Method C), and then filtered. The wet cake was washed with aqueous EtOH (prepared from 37.6
kg (1.2 vol) of EtOH and 111 L (2.8 vol) of water), dried at 50 °C (internal temperature) for ≥1 h
under reduced pressure (IPC: LOD ≤ 1.0%) to give (3R,4R)-105 (27.9 kg, 110 mol, 95.1% yield)
as a pale yellow solid. HPLC purity: 98.0 A% (specification: ≥94 A% by HPLC Method C). 1H
NMR (DMSO-d6, 500 MHz): δ 0.59 (d, 3H, J = 7.0 Hz), 1.51–1.60 (m, 1H), 1.67–1.75 (m, 1H),
2.27–2.54 (m, 2H), 2.70 (ddd, 1H, J = 12.2, 7.4, 3.7 Hz), 2.96 (ddd, 1H, J = 12.1, 7.3, 3.3 Hz),
3.08 (dd, 1H, J = 12.2, 3.7 Hz), 3.29 (dd, 1H, J = 12.3, 6.6 Hz), 4.89 (dt, 1H, J = 6.5, 4.1 Hz),
6.58 (d, 1H, J = 3.3 Hz), 6.71 (dd, 1H, J = 3.3, 1.6 Hz), 7.32 (t, 1H, J = 2.9 Hz), 7.76 (br s, 1H),
8.44 (s, 1H), 11.58 (br s, 1H). 13C{1H} NMR (DMSO-d6, 125 MHz): δ 14.7, 30.9, 33.2, 43.2,
47.8, 55.7, 96.9, 100.7, 104.3, 117.7, 121.7, 125.1, 134.3, 137.2, 144.5. HRMS–ESI (m/z): [M +
H]+ calcd for C15H19N4, 255.1604; found, 255.1603. Anal. Calcd for C15H18N4: C, 70.84; H,
7.13; N, 22.03. Found: C, 70.77; H, 7.24; N, 21.68. [α]D23 − 159.0 (c 1.00, 0.1 M aq HCl). IR
(ATR, cm–1): 3310, 3109, 2899, 1614, 1352, 1339, 822, 716. DSC (onset): 84.3, 260.1 °C.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Crude 1. A 500 L reactor-1 was charged with (3R,4R)-10 (13.63 kg, 53.6 mol) and DMF (128.8
kg, 10 vol). HOBt·H2O (0.82 kg, 0.1 equiv, 5.4 mol) and cyanoacetic acid (11) (5.01 kg, 1.1
equiv, 58.9 mol) were added to the resulting slurry, followed by EDC·HCl (16.44 kg, 1.6 equiv,
85.8 mol) over ≥0.5 h, and the reaction mixture was stirred for ≥1 h at 20−30 °C (IPC: 10 ≤ 1.0
A% by HPLC Method D). Next, 1 M aqueous HCl (54.4 kg, 1.0 equiv, 53.3 mol) was added to
the reaction mixture and the mixture was stirred for ≥1 h at 45−55 °C (IPC: overreacted
29
ACS Paragon Plus Environment