Development of a Quality Controllable and Scalable Process for the Preparation of 7,8-Difluoro-6,11-dihydrodibenzo[ b, e]thiepin-11-ol: A Key Intermediate for Baloxavir Marboxil

Article abstract of DOI:10.1021/acs.oprd.9b00389

A novel six-step synthesis of 7,8-difluoro-6,11-dihydrodibenzo[b,e]thiepin-11-ol (1) is described. Starting with 3,4-difloro-2-methylbenzoic acid and using diphenyl disulfide as an ideal sulfur source effectively solve the problems such as harsh reaction conditions, usage of smelly thiophenol, which might restrict the known process from pilot plant application. Large-scale applicability of this new route has been successfully demonstrated on kilogram-scale production to afford 1 with 98.04% purity in 75% overall yield. Meanwhile, the corresponding impurity profile was thus studied in detail and well documented.

Full text of DOI:10.1021/acs.oprd.9b00389

Subscriber access provided by University of Winnipeg Library
Full Paper
Development of A Quality Controllable and Scalable Process
for the Preparation of 7, 8-Difluoro-6,11-Dihydrodibenzo[b,e]
Thiepin-11-ol: A Key Intermediate for Baloxavir Marboxil
Zihong Zhou, Zhongqing Wang, jingping Kou, shuming wu, Junrong
Zhang, Xi Yuan, Xiwei Wu, Chuan-Hua Li, and Gaohong Liao
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.9b00389 • Publication Date (Web): 28 Nov 2019
Downloaded from pubs.acs.org on November 28, 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 21
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Development of A Quality Controllable and Scalable Process for the Preparation
of 7, 8-Difluoro-6,11-Dihydrodibenzo[b,e] Thiepin-11-ol: A Key Intermediate for
Baloxavir Marboxil
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
‡,
Zihong Zhou‡,§,‖, Zhongqing Wang*,†, §,‖, Jingping Kou‡, Shuming Wu‡, Junrong Zhang‡, Xi Yuan‡,
Xiwei Wu‡, Chuan-hua Li†, Gaohong Liao‡
†Hunan Provincial Key Laboratory of Xiangnan Rare-Precious Metals Compounds Research and
Application, Xiangnan University, Chenzhou 423000, China
‡HEC Research and Development Center, HEC Pharm Group, Dongguan 523871, P. R. China
§State Key Laboratory of Anti-Infective Drug Development, Sunshine Lake Pharma Co.,Ltd,
Dongguan 523871, P. R. China
*E-mail:Wangzhongqing@hec.cn
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 21
1
2
3
4
TOC Graphic:
5
6
7
8
9
S
S
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
0.50 eq
6 steps
O
F
S
OH
F
F
OH
F
16
1
>98% purity by HPLC
up to 75% overall yield
ACS Paragon Plus Environment

Page 3 of 21
Organic Process Research & Development
1
2
3
4
ABSTRACT:
5
6
7
8
9
A novel 6-step synthesis of 7,8-difluoro-6,11-dihydrodibenzo[b,e]thiepin-11-ol (1) is described.
Starting with 3,4-difloro-2-methylbenzoic acid and using diphenyl disulfide as an ideal sulfur source
effectively solve the problems such as harsh reaction conditions, usage of smelly thiophenol, which
might restrict the known process from pilot plant application. Large-scale applicability of this new
route has been successfully demonstrated on kilogram-scale production to afford 1 with 98.04% purity
in 75% overall yield. Meanwhile, the corresponding impurity profile was thus studied in detail and well
documented.
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
Key words: Baloxavir marboxil; Diphenyl disulfide; Thiophenol; impurity profile; process
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 21
1
2
3
4
INTRODUCTION
5
6
7
8
9
Baloxavir marboxil (Xofluza; baloxavir), which had been developed by Roche and Shionogi, was a
small molecule cap-dependent endonuclease inhibitor and exhibited broad and potent antiviral activity
against clinically isolated influenza A and B virus, including viruses harboring NAI-resistance
mutation.1 Baloxavir marboxil was first approved for the treatment of influenza A or B virus infections
in Japan in February 2018,1b and then approved by the U.S. Food and Drug Administration (FDA) on
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
October 24, 2018.2
Scheme 1. Three Synthetic Routes towards Baloxavir Marboxil (Xofluza)
OBn
O
CO2Et
NH
OBn
N
O
OBn
N
O
OH
S
1st generation
O
O
O
N
O
N
N
O
O
O
O
O
O
O
F
N
H
N
H
NAlloc
N
F
2
3
4
N
O
1
N
(S)
OBn
O
O
O
OBn
N
O
O
2nd generation
F
O
COOH
O
O
O
N
S
N
O
F
H
N
O
N
H
Baloxavir Marboxil
NHBoc
9
5
6
O
O
COOMe
O
O
O
3rd generation
O
O
COOMe
N
O
H
N
O
NHBoc
7
8
Since the discovery of Baloxavir marboxil, three generations of synthetic route have been developed
by Shionogi & CO., LTD to satisfy different needs during the drug development process.3 As outlined
in Scheme 1, these three approaches shared a similar retrosynthetic strategy of combining two major
parts of the molecule, key intermediate 1 and the other motif 2 or 5, by an SN2 substitution reaction.
Notably, an asymmetric directing group was introduced in the third generation route disclosed by
Shionogi recently to improve overall yield and avoid cumbersome operations. Fragment 2 or 5 can both
be prepared from compound 9 by intermolecular aminolysis or acetalization according to known
methodologies. However, when refers to the production of 1 there has been relatively few reports by
far.3b,4
Scheme 2. Previous Route to intermediate 1
ACS Paragon Plus Environment

Page 5 of 21
Organic Process Research & Development
1
2
3
Previous route 3b
4
5
6
7
O
O
O
COOH
OH
S
O
AlCl3, TMDSO
toluene, 0-25 o
81.3% yield, 3 steps
n-BuLi, iPr2NH
THF, -40oC
O
D-CSA, PhSH
toluene, 60 o
F
C
C
F
F
F
S
F
OH
F
F
F
8
13
12
10
11
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
O
HO
PPA
NaBH4
120 oC, 91%
S
S
F
iPrOH/H2O, 40C
97% yield
F
F
F
14
1
In 2017, Shionogi3b developed an intramolecular Friedel-Crafts acylation process to construct
intermediate 1 from 3,4-difluorbenzoic acid 10 (Scheme 2). The methylene phenyl sulfide part in the
derivative was introduced from N,N-Dimethylformamide (DMF) and thiophenol in 3 steps. As an
original route to acquire sufficient amount of 1 for follow-up research, the drawbacks were obvious.
Large-scale synthesis of 1 was limited by ultra-low reaction temperature and usage of
moisture-sensitive base lithium diisopropylamide (LDA). Besides, thiophenol was so smelly and
air-sensitive that it was not suitable for scale-up procedure. Furthermore, when we repeated this
reported procedure a defluorinated impurity 15 was detected and proved difficult to remove. Moreover，
the corresponding impurity originated from 15 was found in the final active pharmaceutical ingredient
(API) and therefore affect the product quality (Figure 1). Thus, it is still highly desirable to develop an
efficient and readily scalable process for the synthesis of this key intermediate 1 with high purity which
could meet the quality remand for Baloxavir marboxil .
O
O
O
O
O
O
N
OH
N
N
O
S
F
15
S
F
the corresponding impurity in the API
substance
about 0.23% HPLC content
Figure 1. Key impurity 15 and the corresponding impurity in the final API
To overcome all the above disadvantages, we wished to develop a convenient and readily scalable
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 21
1
2
3
4
5
6
7
8
9
approach for the synthesis of 1 based on the retrosynthetic strategy depicted in Scheme 3. It was
envisioned that with introduction of a methyl group in ortho-position to the carboxylic acid beforehand,
alkylation under ultra-low temperature could be avoided. Then after bromination of the benzylic
position, substitution by a thiol nucleophile would be implemented to afford a cyclization precursor.
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
Another 2 steps were needed to complete the route after the intramolecular cyclization.
Scheme 3. Retrosynthetic Analysis of Baloxavir marboxil Key Intermediate 1
O
F
O
F
S
S
OH
S
F
F
OH
F
F
OH
F
O
F
1
Following this line of retrosynthetic reasoning, we set out to develop this synthetic strategy. During
our production for this procedure on kilogram-scale, some Chinese patents using similar strategies
were disclosed earlier this year (Scheme 4).4 However, whether these processes could be scaled up in
batch mode was still suspicious due to commercial unavailability of starting materials 6-Cyano-2,3-
difluorobenzyl bromide4a or (6-bromo-2,3-difluorophenyl)mechanol.4b Besides, carboxyl group
introduction by reaction between Grignard reagent and CO2 might not be easily manipulated and the
yield was less than moderate.4b Moreover, no quality researches concerning impurities profiles were
presented in the patents. Therefore, we still continued our ongoing work to seek an industrial solution
which could meet the corresponding quality demand for the API substance.
Scheme 4. Relative Chinese patents for intermediate 1 and intermediate 14
ACS Paragon Plus Environment

Page 7 of 21
Organic Process Research & Development
1
2
3
CN109134428 4a
4
5
6
7
S-Na+
O
HO
N
S
N
P2O5
NaBH4
Br
MeOH, -10C
92.7%
F
Et3N, acetone/H2O, rt
96.9%
Xylenes, reflux
82.3%
F
S
S
F
F
F
F
F
F
8
1
9
CN109721585 4b
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
Br
Br
Br
iPrMgBr in THF
CO2, hexane, -35C, 70%
PhSSPh, NaBH4
NaOH, H2O, 70C, 95%
40% HBr(aq.)
S
OH
Br
F
F
F
MeCN, 89.5%
F
F
F
O
COOH
S
(COCl)2, AlCl3
DCM, 60%
F
S
F
F
F
14
RESULTS AND DISCUSSION
Herein we report our readily scalable process for the synthesis of intermediate 1 (Scheme 5). Initial
esterification with 3,4-difloro-2-methylbenzoic acid (16) and dimethyl sulfate (DMS) in
dimethylacetamide (DMAc) gave compound 17 in nearly quantitative yield. Slow crystallization
proved to be critical to remove the defluorinated impurity (<0.08% detected by HPLC). MeOH/SOCl2
was also an alternative for methylation but led to an oily product, which had brought difficulties to the
subsequent purification procedure.
Scheme 5. New Synthetic Route to 1
This work:
PhSSPh (0.48 equiv),
Zn (2.0 equiv),
FeCl3 (1.0 equiv)
O
O
O
O
DMS (1.5 equiv),
NaHCO3 (1.8 equiv)
NBS (1.2 equiv),
BPO (0.03 equiv)
O
S
OH
O
O
DMAc, 40 o
98%
C
Br
MeCN/H2O = 4:1(v/v)
-5 oC~rt
DCM, reflux
97%
F
F
F
F
F
F
F
F
16
17
18
19
O
F
F
S
S
KOH (1.6 equiv)
PPA
100 o
94%
NaBH4 (0.3 equiv)
F
OH
S
F
C
EtOH, 50 o
98%
C
i-PrOH/H2O = 1:1 (v/v)
F
80 oC, 86% for two steps
O
OH
F
13
14
1
5
CCl4 was the best solvent for the following Wohl-Ziegler Reaction to furnish 18 due to its
immiscibility with N-bromosuccinimide (NBS) and resulting succinimide, reaching a 97% yield in only
18 h (Table 1, entry 1). The poor solubility of NBS in CCl4 successfully kept the concentration of the
reagent low and the insolubility of byproduct succinimide also promoted the reaction. However, high
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 21
1
2
3
4
5
6
7
8
9
toxicity of CCl4 made it unsuitable for industrial scale-up. In addition, other solvents such as
acetonitrile (MeCN) and ethyl acetate, which could lead to an undesired dibromo-substituted
byproduct, failed to provide product 18 in good yield or satisfying quality. Thus, it prompted us to
choose dichloromethane (DCM) as the reaction solvent, with which 18 was obtained in 97% yield after
36 h (Table 1, entries 2-4). Other bromination reagent such as 1,3-dibromo-5,5-dimethylhydantoin
(DBDMH) led to an inferior result, same as an reduced reaction temperature to 40 oC (Table 1, entries
5-6). Therefore, benzylic bromination of 17 with NBS using dibenzoyl peroxide (BPO) as free-radical
initiator in DCM at reflux was preferred as the optimal conditions. Together with esterification with
DMS, both steps could be scaled up to multi-kilogram batch size.
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 1. Effect of solvent, reagent, and temperature on bromination reaction a
O
O
O
O
O
O
bromination reagent
solvent
Br
Br
F
F
F
F
Br
F
17
F
dibromo-substituted
impurity
18
Yield of
18c(%)
entry
solvent
bromination reagent
T (oC) b
1
2
3
4
5
6
CCl4
DCM
MeCN
NBS
NBS
80
50
80
60
50
40
97d
97
58(24e)
NBS
ethyl acetate
DCM
NBS
63
DBDMH
NBS
85
DCM
81
a
reaction conditions: 17 (0.11 mol), reagent (1.2 equiv), and solvent (8.0 mL/g) , for 36 h; b reactor
jacket temperature; c isolated yield; d for 24 h; e the content of dibromo-substituted impurity.
Additionally, during the investigation of this step the defluorinated impurity in 17 was considered
material-related rather than process-related as 0.3% of 3-fluoro-2-methylbenzoic acid had been found
in 3,4-difluoro-2-methylbenzoic acid 16. After the optimization of the first step, the corresponding
defluorinated impurity was easy to purge and remained only 0.08% determined by HPLC. Moreover,
ACS Paragon Plus Environment

Page 9 of 21
Organic Process Research & Development
1
2
3
4
no 15 could be detected in 1 followed by successive steps in the later studies (Scheme 6). Therefore, by
5
6
7
controlling
the
amount
of
3-fluoro-2-methylbenzoic
acid
in
starting
material
8
9
3,4-difluoro-2-methylbenzoic acid 16, defluorinated impurity 15 can be well avoided in key
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
intermediate 1.
Scheme 6. The origin of impurity 15
O
O
F
S
O
OH
OH
F
F
3-fluoro-2-methylbenzoic acid
15
impurity
corresponding deflorinated
17
impurity in
For the subsequent step, it is known that most of the sulfur nucleophiles are not suitable for pilot
plant production due to horrible smells. To solve this problem, we tried to choose a solid and
non-volatile sulfur source as a nucleophile. We initially examined sodium thiophenolate for its strong
nucleophility and alkalinity. However, it proved more challenging than anticipated as sodium
thiophenolate turned out to have a poor solubility in organic solvents and was instable under reaction
conditions with a possible mechanism shown in Scheme 7.6 After 17 hours of reaction, only 39% of 18
was converted into product 19, with a content of about 15% of hydrolysis product 13 and 10% of
diphenyl disulfide. Significant efforts including usage of other solvents such as dimethylformamide
(DMF), tetrahydrofuran (THF), acetonitrile (MeCN), DCM and toluene were not able to provide
acceptable results, nor were elevated temperature (50 oC) or increased dosage of the sodium
thiophenolate (3.0 equiv). Although the reaction proceeded under nitrogen, the content of diphenyl
disulfide still increased considerably while the reaction temperature was raised. Another impurity,
methyl 2-(bromomethyl)-3-fluoro-4-(phenylthio)benzoate 20, was also detected in reaction mixture
(Scheme 7). In addition, excessive sodium thiophenolate would result in a large amount of smelly
thiophenol after acidic work-up which might pass on to the crystallization of 13. Thus, sodium
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 21
1
2
3
4
thiophenolate was not a good sulfur nucleophile for this reaction and had to be replaced.
5
6
7
8
Scheme 7. Impurities and possible formation mechanism of diphenyl disulfide in the thiol
etherification reaction
9
O
O
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
O
S
PhSNa
O
O
S
S
Br
solvents
Br
F
F
S
F
F
F
20
18
19
diphenyl disulfide
Possible mechanism of diphenyl disulfide formation from sodium thiophenolate
2PhS-Na+ +
+
2Na+
+
2-
O2
2PhS
O2
2PhS
2Na+
PhSSPh
2NaOH
+
0.5O2
2-
H2O + O2
+
All the mentioned difficulties led us to pursue an alternative method of preparing thiol ether 19. In
the original procedure, diphenyl disulfide was an undesired but intrinsic impurity possibly attributed to
air oxidation. From another perspective, diphenyl disulfide is also an odorless solid sulfur-containing
compound. Thus, we were interested in the possibility to utilize diphenyl disulfide as a sulfur source in
the synthesis of intermediate 1 as several benefits could be envisioned. And to our delight, a research
by S. Roy et. al7 was later found to support this idea, which clearly described a new radical assisted
synthesis to sulfides from disulfides via in-situ generated organocobalt(III) intermediates. Moreover, 18
might still be suitable for this reaction system.8 Based on that, we set out to verify our proposal. To
prove our concept (Table 2), 18 was treated with diphenyl disulfide in the presence of Lewis acid AlCl3
together with zinc powder in DMF/H2O. Gratifyingly, the reaction was clean and 19 was achieved
o
successfully in 83% yield at 65 C. Although the yield of 19 was not affected by lowering the loading
of diphenyl disulfide to 0.5 equiv (Table 2, entries 1-3), different kinds of Lewis acid had a great
influence on the conversion of 18, among which FeCl3 gave the best result (85% yield, Table 2, entries
4-8).
On the other hand, a drastic drop of yield to only 14% of 19 occurred when the Zn loading was
ACS Paragon Plus Environment

Page 11 of 21
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
reduced to 1.5 equiv and no product was observed when lowered to 1.0 equiv (Table 2, entries 9-11).
An abundant of zinc powder demonstrated to be necessary for good yields implying a potential
different mechanism from Roy’s methodology. Meanwhile, the yield decreased dramatically when the
Lewis acid loading is reduced to less than 1 equivalent (Table 2, entries 12-13). A subsequent solvent
screening indicated MeCN/H2O as the optimal solvent system for this conversion, with a product yield
of 91% (Table 2, entries 14-18). Besides, water played an important role in this process by showing
only 46% yield of 19 and a content of 17 up to 38% when absent (Table 2, entry 18). We were also
glad to find that by-product generation could be suppressed by decline of reaction temperature, with
only 0.25% of 17 was detected while 19 was obtain in 91% yield under room temperature (Table 2,
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
entry 19).
Table 2. Investigation of the optimized conditions for synthesizing 19a
O
O
O
O
O
S
Zn/ lewis acid
solvent
O
S
+
S
Br
F
F
F
F
F
F
18
19
17
Zn/ lewis acid
(mol ratio)
AlCl3 (2:1)
AlCl3 (2:1)
AlCl3 (2:1)
FeCl3 (2:1)
CuCl2 (2:1)
CoCl2 (2:1)
MgCl2 (2:1)
entry PhSSPh (equiv)
solvent b
T (oC) Yield of 19 c:17d (%)
1
2
1.0
0.8
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
DMF/H2O
DMF/H2O
DMF/H2O
DMF/H2O
DMF/H2O
DMF/H2O
DMF/H2O
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
83:2.0
84:2.0
83:2.0
85:3.0
31:36
82:4.0
78:2.0
80:3.0
84:2.0
14:3.0
N/A
3
4
5
6
7
8
DMF/H2O BF3·Et2O (2:1)
9
DMF/H2O
DMF/H2O
DMF/H2O
FeCl3 (2.5:1)
FeCl3 (1.5:1)
FeCl3 (1:1)
10
11
12
13
14
15
DMF/H2O FeCl3 (1.6:0.8)
DMF/H2O FeCl3 (1.0:0.5)
62:2.0
45:3.0
85:3.0
91:2.0
THF/H2O
MeCN/H2O
FeCl3 (2:1)
FeCl3 (2:1)
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 21
1
2
3
4
5
6
7
8
9
16
17
18
19
20
21
22
0.5
0.5
0.5
0.5
0.5
0.5
0.5
toluene/H2O
MeOH/H2O
MeCN
FeCl3 (2:1)
FeCl3 (2:1)
FeCl3 (2:1)
FeCl3 (2:1)
FeCl3 (2:1)
FeCl3 (2:1)
FeCl3 (2:1)
65
65
65
rt
30:2.0
88:2.0
46:38
MeCN/H2O
MeCN/H2O
MeCN/H2O
MeCN/H2O
91:0.25
90:1.0
87:3.0
81:4.0
40
50
80
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
b
reaction conditions: 18 (3.8 mmol), lewis acid (1.0 equiv), and solvent (10 mL/g) , for 3 h; unless
c
otherwise stated, all reactions were carried out using organic solvent : H2O = 4 : 1 (v/v) as solvent;
isolated yield; d detected by HPLC.
Subsequently, an investigation on solvents of the saponification for crude product 19 to provide 13
was carried out (Step 4, Scheme 5). According to the experiment results, isopropyl alcohol was
preferred over MeOH or EtOH. This might probably be attributed to Le Chatelier's principle as
methanol is also produced during this hydrolysis process. Furthermore, higher temperature accelerated
the reaction while iPrOH has a higher boiling point over MeOH. Although KOH was considered as the
optimal base for this reaction, other bases such as K2CO3 and NaOH were acceptable while a reaction
time of at least 18 hours was demanded to guarantee a complete conversion. Ultimately, acid 13 was
achieved smoothly as a white solid, with an overall yield up to 85% over two steps from compound 18.
Considering the disavantage of the high reaction temperature, different cyclization conditions from
the reported ones by Shiongi 3b were evaluated initially. Standard conditions using SOCl2/AlCl3/DCM
as a Lewis acid system was supposed to be suitable for this following Friedel-Crafts acylation
cyclization to afford 14. However, contrary to our initial expectation, it failed with only a 35%
conversion. Tremendous efforts had been made with the attempt to improve the yield of 14 under
standard conditions by changing solvents into ClCH2CH2Cl, PhNO2, CHCl3 or CCl4 at different
temperatures including 50 oC, 60 oC and 70 oC. To our disappointment, only moderate yield (73%) was
achieved in CCl4 at 60 oC. The difficulty in this cyclization urged us to pursue an alternative sequence.
Finally, when we came to a Brønsted acid system adopting neat polyphosphoric acid (PPA) as reaction
ACS Paragon Plus Environment

Page 13 of 21
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
reagent which was the same as that proposed by Shiongi,3b a successful cyclization of compound 13
was furnished and the desired 14 was given in 94% yield. Lower reaction temperature, a simplified
work-up procedure and a slightly higher yield were achieved through the improvements of this step
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
compared to the original conditions.
Table 3. Optimization of reduction conditionsa
F
F
S
S
NaBH4
solvent
F
F
O
14
OH
1
entry
NaBH4 (equiv)
solvent
MeOH
i-PrOH
EtOH
Yield of 1 (%)b
1
2
3
4
5
6
1.0
1.0
89
97
98
97
98
76
1.0
0.50
0.30
0.25
EtOH
EtOH
EtOH
a reaction conditions: 14 (46 mmol), solvent (3 mL/g) , 50 oC , for 3 h; b isolated yield;
For the last step, compound 14 was reduced to form key intermediate 1 as presented in Table 3.
EtOH was more favored over MeOH or iPrOH in terms of yields while MeOH was the least beneficial,
partly due to a better solubility of product in methanol (Table 3, entries 1-3). When above 0.30 equiv,
loading of NaBH4 had no big influence on the reaction yield (Table 3, entries 3-5), whereas the yield
slumped to 76% as the loading lowered to 0.25 equiv (Table 3, entry 6). This indicated that at least 0.25
equivalent of sodium borohydride was required for a clean and complete reaction. After an easy
crystallization by simply adding water to the reaction mixture at 50 oC, product 1 was collected through
a smooth filtration with high quality (>98% pure by HPLC). No sodium hydroxide was needed for the
reaction compared to the original conditions.3b
For the optimization of the last step, ethyl ether impurity 21 was one of the major impurities that had
been observed during acidic work-up of 1, which was owe to an ether formation between hydroxyl
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 21
1
2
3
4
5
6
7
8
9
group of 1 and the solvent ethanol under treatment of aqueous hydrochloride. By quenching the
reaction with water instead of aqueous hydrochloride, 21 was finally avoided. Even if trace 21 was
found in intermediate 1, it still had no effect on the quality of final API substance as the mentioned
impurity could also react with 5 to deliver the same final product (Scheme 8).
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
Scheme 8. The formation and control strategy of impurity 21
OH
OEt
O
H
O
O
O
O
O
N
F
F
EtOH
S
N
S
F
F
N
O
impurity 21
1
F
O
O
S
F
O
N
Baloxavir Marboxil
N
O
N
H
5
Impurity control is an important part of drug development and regulatory assessment for efficiency
and safety reasons. The impurity 15 and 21 were initially found critical for the quality of 1, thus the
formation and control strategies for these two were carefully studied and well documented during this
process development. Based on that, 1 was well obtained from 16 via six steps with 98.04% purity in
75% overall yield. Moreover, the process has been successfully applied for multikilo-gram scale
production showing reproducible and consistent results all along the development.
ACS Paragon Plus Environment

Page 15 of 21
Organic Process Research & Development
1
2
3
4
CONCLUSIONS
5
6
7
8
9
We have developed an effective and scalable process to synthesize key motif 1, that avoids harsh
reaction conditions by starting from 3,4-difloro-2-methylbenzoic acid and addresses smelly issue by
using diphenyl disulfide as a stable sulfur source compared to the known process. This six-step
procedure has been demonstrated in multikilo-gram scale production to afford 1 with 98.04% purity in
75% overall yield while the corresponding impurity file is thus studied in detail. It is expected to be
designated as the manufacturing route in large quantities and therefore supports the kilogram-scale
preparation of the baloxavir marboxil that meets drug substance standards for future research and
commercialization.
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
If no specially indicated, all reagents and solvents were used as commercially available without
further purification. NMR spectra were measured on a BrukerAvance 400 spectrometer in the solvents
indicated; chemical shifts are reported in units (ppm) by assigning TMS resonance in the 1H spectrum
as 0.00ppm. Coupling constants are reported in Hz with multiplicities denoted as singlet (s), doublet
(d), triplet (t), quartet (q), dd (doublet of doublets); m (multiplets), and etc. HRMS were performed on
Fourier Transform Ion Cyclotron Resonance Mass Spectrometer. Analytical HPLC for liquid phase
was carried out on an Agilent HPLC workstation, equipped with a Water Xbridge-C18 (4.6x 75mm,
2.5μm) column. Gradient elution from 65:35 to 20:80 (0.1% H3PO4 in water)/(acetonitrile) over 19 min,
from 20:80 to 10:90 (0.1% H3PO4 in water)/(acetonitrile) over 1 min, hold for 4 min, then gradient
from 10:90 to 65:35 over 4 min; flow rate 1.0 mL/min; temperature, 50 oC; wave length 216 nm.
Preparation of methyl 3,4-difluoro-2-methylbenzoate (17). To
a
stirred solution of
3,4-difluoro-2-methylbenzoic acid (2 kg, 11.6 mol) in N,N-dimethylacetamide (DMAc, 6L) was added
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 21
1
2
3
4
5
6
7
8
9
dimethyl sulfate (DMS, 2.2 kg, 17.4 mol) and NaHCO3 (1.8 kg, 20.9 mol). The reaction mixture was
stirred at 40 oC for 4 h. After fully conversion of the starting material, the reaction mixture was cooled
to room temperature and H2O (18L) was added dropwise over at least 3 hours. The white solid product
was collected by filtration and dry under vacuum at room temperature to yield the titled compound
(2.12 kg, 98% yield). HPLC purity: 99.86%, 1H NMR (600 MHz, CDCl3) δ 7.72 (ddd, J = 8.6, 5.2, 1.7
Hz, 1H), 7.04 (dd, J = 16.9, 8.8 Hz, 1H), 3.91 (s, 3H), 2.55 (d, J = 2.7 Hz, 3H). 13C NMR (101 MHz,
CDCl3) δ 166.57 (d), 154.23 (d), 151.70 (d), 150.60 (d), 148.16 (d), 130.58 (d), 126.89 (dt), 113.99 (d),
52.22, 12.25 (dd). HRMS (ESI) calcd for C9H9F2O2 [M+H]+ 187.0560, found 187.0565.
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
Preparation of methyl 2-(bromomethyl)-3,4-difluorobenzoate (18). To a solution of methyl
3,4-difluoro-2-methylbenzoate 17 (2 kg, 10.7 mol) in dichloromethane (DCM, 16 L) was added
N-bromosuccinimide (NBS, 2.3 kg, 12.9 mol) and dibenzoyl peroxide (BPO, 78.1 g, 0.3 mol), the
resulting mixture was stirred for about 36 h at 50 oC. After fully conversion of the starting material, the
insoluble matters were removed by filtration and the organic layer was washed with 10% of Na2SO3
aqueous solution (1 L), then with water (1L). The organic layer was concentrated under vacuum (40
1
oC) to give 2.7 kg product as yellow oil, yield 97%. HPLC purity: 92.29%, H NMR (600 MHz,
CDCl3) δ 7.81 (ddd, J = 8.8, 5.0, 1.9 Hz, 1H), 7.19 (ddd, J = 13.5, 8.4, 6.3 Hz, 1H), 5.01 (d, J = 2.2 Hz,
2H), 3.95 (s, 3H). 13C NMR (101 MHz, CDCl3) δ 165.41 (d), 154.18 (d), 151.63 (d), 150.56 (d), 148.05
(d), 130.88 (m), 128.96, 127.90 (dd), 116.88 (d), 52.67, 21.06 (dd). HRMS (ESI) calcd for C9H8BrF2O2
[M+H]+ 264.9615, found 264.9670.
Preparation of methyl 3,4-difluoro-2-((phenylthio)methyl)benzoate (19). In a 30 L reaction kettle,
diphenyl disulfide (450 g, 2.06 mol), zinc powder (564 g, 8.63 mol) and MeCN-H2O (12L, 4:1, v/v)
were placed. The solution was cooled to -5 °C, then FeCl3 (696 g, 4.3 mol) was added by portions over
ACS Paragon Plus Environment

Page 17 of 21
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1 h. After that, the reaction was allowed to reach room temperature and stirred for 3 h, and then methyl
2-(bromomethyl)-3,4-difluorobenzoate 18 (1.2 kg, 4.3 mol) was added. The reaction was stirred at
room temperature until conversion was completed (about 8 h). The residue was removed by filtration,
the filter cake was washed with ethyl acetate (500 mL×3), and the filtrate was concentrated under
vacuum to give a crude product. The crude product was dissolved in the mixture of ethyl acetate-H2O
(12 L, 1:1, v/v). After stirring for 30 min, the aqueous phase was removed and the organic layer was
washed with water (2L) before being concentrated. The product was used in the next step directly
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
without further purification. HPLC purity: 84.35%, H NMR (600 MHz, CDCl3) δ 7.68 (ddd, J = 8.8,
5.1, 1.9 Hz, 1H), 7.40-7.29 (m, 2H), 7.26-7.16 (m, 3H), 7.05 (td, J = 9.0, 7.8 Hz, 1H), 4.55 (d, J = 1.9
Hz, 2H), 3.82 (s, 3H). 13C NMR (101 MHz, CDCl3) δ 166.00 (d), 153.87 (d), 151.33 (d), 150.18 (d),
147.71 (d), 134.84, 132.55, 131.00 (d), 128.87, 127.61, 127.36 (dd), 126.31 (dd), 115.30 (d), 52.38,
29.89 (dt). HRMS (ESI) calcd for C15H13F2O2S [M+H]+ 295.0568, found 295.0599.
Preparation of 3,4-difluoro-2-((phenylthio)methyl)benzoic acid (13). Crude product of methyl
i
3,4-difluoro-2-((phenylthio)methyl)benzoate 19 was dissolved in a mixed solution of PrOH-H2O
(10.05 L, 2:1, v/v), then KOH (383 g, 6.84 mol, dissolved in 3.35 L of H2O) was added. The resulting
o
o
solution was heated to 80 C and stood for 8 h, then cooled to 50 C. The solvent was removed under
reduced pressure to give crude product as a gray solid. The crude product was dissolved in water (6.7
L), then the pH was adjusted to 1 with HCl (2 M) and stirred for another hour at 15 oC. After filtration
a solid was observed and then suspended in n-hexane (6.7 L), stirred for 1 h, then filtered and dried to
give pure product as a white solid (1.1 kg, 86% yield over two steps). HPLC purity: 97.37%, 1H NMR
(400 MHz, CDCl3) δ 7.97-7.80 (m, 1H), 7.35 (dd, J = 6.4, 2.8 Hz, 2H), 7.30-7.20 (m, 3H), 7.12 (dd, J =
16.7, 8.7 Hz, 1H), 4.61 (s, 2H). 13C NMR (101 MHz, CDCl3) δ 171.16, 154.72 (d), 152.16 (d), 150.38
ACS Paragon Plus Environment

Organic Process Research & Development
Page 18 of 21
1
2
3
4
5
6
7
8
9
(d), 147.91 (d), 134.60, 133.02, 132.22 (d), 129.05, 128.58 (dd), 127.96, 125.20-124.14 (m), 115.57
(d), 30.11 (d). NMR were in accordance with literature.3b HRMS (ESI) calcd for C14H11F2O2S [M+H]+
281.0399, found 281.0442.
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
Preparation
of
7,8-difluorodibenzo[b,e]thiepin-11(6H)-one
(14).
A
mixture
of
3,4-difluoro-2-((phenylthio)methyl)benzoic acid 13 (1.0 kg, 3.57 mol) and polyphosphoric acid (PPA,
o
4.8 kg) was heated to 100 C. After 10 h, the reaction mixture was cooled to room temperature, and
ice-water (10 kg) was added. The solid product was filtrated and was suspended in saturated NaHCO3
solution (3 L), stirred for 8 h, then filtered and dried in vacuum to yield the titled compound as a gray
1
solid (880 g, 94% yield). HPLC purity: 98.43%, H NMR (400 MHz, CDCl3) δ 8.19 (dd, J = 8.1, 1.3
Hz, 1H), 7.38 (dddd, J = 18.1, 9.1, 6.4, 1.5 Hz, 3H), 7.28 (ddd, J = 12.8, 7.1, 3.6 Hz, 1H), 7.12 (dt, J =
16.6, 8.3 Hz, 1H), 4.13 (s, 2H). 13C NMR (101 MHz, CDCl3) δ 193.42 (d), 153.86 (d), 151.33 (d),
146.60 (d), 144.13 (d), 140.51 (s), 137.59 (d), 135.14, 133.10, 132.68, 129.41, 128.04 (d), 126.02,
125.78 (dd), 116.34 (d), 27.30 (dd). NMR were in accordance with literature.3b HRMS (ESI) calcd for
C14H9F2OS [M+H]+ 263.0350, found 263.0337.
Preparation
of
7,8-difluoro-6,11-dihydrodibenzo[b,e]thiepin-11-ol
(1).
Compound
7,8-difluorodibenzo[b,e]thiepin-11(6H)-one 14 (2.0 kg, 7.63 mol) was dissolved in EtOH (6 L) and
NaBH4 (87 g, 2.29 mol) was added by portions over 1 h at room temperature. The resulting solution
was stirred for 4 h at 50 oC, then water (18 L) was added dropwise under this temperature. The reaction
o
mixture was then cooled to 0 C and stirred for a further hour, then filtered and dried to give pure
1
product as a white solid (1.98 kg, 98% yield). HPLC purity: 98.04%, H NMR (400 MHz, CDCl3) δ
7.50-7.35 (m, 1H), 7.21-7.06 (m, 4H), 6.99 (dd, J = 17.8, 8.4 Hz, 1H), 6.01 (s, 1H), 4.61 (d, J = 14.5
Hz, 1H), 4.14 (dd, J = 14.5, 0.7 Hz, 1H), 2.97 (s, 1H). 13C NMR (101 MHz, CDCl3) δ 151.61 (d),
ACS Paragon Plus Environment

Page 19 of 21
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
149.13 (d), 148.61 (d), 146.15 (d), 138.73, 137.74 (d), 133.40, 129.57, 128.53, 127.44, 126.16, 123.15
(d), 121.20 (dd), 115.69 (d), 74.60, 25.16 (dd). NMR were in accordance with literature.3b HRMS (ESI)
calcd for C14H11F2OS [M-OH]+ 247.0388, found 247.0372.
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
SUPPORTING INFORMATION
The Supporting Information is available free of charge via the internet at http://pubs.acs.org
1H and 13C NMR spectra for these compounds are given.
CORRESPONDING AUTHORS
E-mail: wangzhongqing@hecpharm.com.
AUTHOR CONTRIBUTIONS
‖ Z.Z. and Z.W. contributed equally to this work.
ACKNOWLEDGMENT
This work was financially supported by the Anti-infective Drug National Key Laboratory Open Grant
(No. SKLAID201805). We also gratefully acknowledge financial support from the Key Science &
Technology Brainstorm Project of Dongguan (No. 2019622102009) and The Hunan Provincial Natural
Science Foundation of China (No. 2017JJ2240). We gratefully thank Guozhu Liu (HEC Pharm Co.,
Inc.) for NMR assistance, and Meiyan Qiu(HEC Pharm Co., Inc.) for HPLC assistance.
REFERENCES
(1) (a) Noshi, T.; Yamamoto, A.; Kawai, M.; Yoshida, R.; Sato, A.; Shishido, T.; Naito, A.
S-033447/S-033188, a Novel Small Molecule Inhibitor of Cap-Dependent Endonuclease of Influenza
A and B Virus: In Vitro Antiviral Activity Against Clinical Strains. Open Forum Infectious Diseases
2016, 3(1), 645. (b) Heo, Y.-A. Baloxavir: First Global Approval. Drugs 2018, 78(6), 693-697.
(2) According
to
U.S.
Food
and
Drug
Administration
(FDA),
ACS Paragon Plus Environment

Organic Process Research & Development
Page 20 of 21
1
2
3
4
5
6
7
https://www.fda.gov/news-events/press-announcements/fda-approves-new-drug-treat-influenza,
acess date Sep.17, 2019.
last
8
9
(3) (a) Makato, K.; Kenji, T.; Toshiyuki, A.; Azusa, O.; Masayoshi, M. Substituted polycyclic
pyridone derivative and prodrug thereof. Patent Appl. WO 2016/175224 A1, 2016. (b) Setsuya, S.;
Nobuaki, F.; Toshikatsu, M. Method for producing substituted polycyclic pyridone derivative and
crystal of same. Patent Appl. WO 2017/221869 A1, 2017. (c) Kazuya, O.; Tatsuhiko, U.; Yoshio, H.;
Toshikazu, H.; Shohei, M. Method for stereoselectively producing substituted polycyclic pyridone
derivative. Patent Appl. WO 2019/070059 A1, 2019. (d) Hughes, D. L. Review of the Patent
Literature: Synthesis and Final Forms of Antiviral Drugs Tecovirimat and Baloxavir Marboxil. Org.
Process Res. Dev. 2019, 23(7), 1298-1307.
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) (a) Song, J.; He, X.; Leng, B.; Wu, R.; Zhang, X.; Zhao, M. Preparation method of baloxavir
marboxil intermediate. Patent Appl. CN 109134428 (A), 2019. (b) Yu, C.; Zheng, X.; Wei, H.; Jiang,
X.; Shao, H.; Jin, Y. Preparation method of key intermediate of baloxavir marboxil. Patent Appl. CN
109721585 (A), 2019.
(5) (a) Akiyoshi, M.; Masamichi, K.; Hideyuki, K. Novel diamide derivative. Patent Appl. WO
2008/111300 A1, 2008. (b) Lee, H. J.; Lim, S. J.; Oh, S. J.; Moon, D. H.; Kim, D. J.; Tae, J.; Yoo, K.
H. Isoindol-1,3-dione and isoindol-1-one derivatives with high binding affinity to b-amyloid fibrils.
Bioorg. Med. Chem. Lett. 2008, 18, 1628-1631. (c) Joseph, B. S.; Francis, K. R.; Lucja, O.
Isoindolinone derivatives. Patent Appl. 2012/150202 A1, 2012.
(6) Joshi, V. A.; Bhusare,S.; Baidossi, M.; Qafisheh, N.; Sasson, Y. Oxidative coupling of thiols to
disulfides using a solid anhydrous potassium phosphate catalyst. Tetrahedron Lett. 2005, 46,
3583-3585.
ACS Paragon Plus Environment

Page 21 of 21
Organic Process Research & Development
1
2
3
4
5
6
7
(7) Chowdhury, S.; Roy, S. Cobalt Assisted Cleavage of S-S Bonds and a Base-free Synthesis of
Mercapturic Acids. Tetrahedron Lett. 1997, 38, 2149-2152.
8
9
(8) (a) Lu, G.; Zhang, Y. A Novel Synthesis of Benzyl Sulfides and Selenides via Sm/BiCl3 System
in Aqueous Media. Synth. Commun. 1998, 28(23), 4479-4484. (b) Movassagh, B.; Lakouraj, M. M.;
Fadaei, Z. Synthesis of organic sulfides from disulfides using a Zn/AlCl3 system in aqueous media. J.
Chem. Research 2000, 350-351. (c) Lakouraj, M. M.; Movassagh, B.; Fadaei, Z. Syhthesis of organic
sulfides via Zn/AlCl3 system in aqueous media. Synth. Commun. 2002, 32(8), 1237-1242. (d) Tian, F.;
Zhu, Y. M.; Zhang, S.; Wang, Y. A novel method for the synthesis of unsymmetrical sulfides,
thioesters and β-thioesters. J. Chem. Research 2002, 582-583. (e) Du, J.; Zheng, R.; Li, X. Synthesis of
allylic sulfides via cleavage of S-S bonds promoted by zinc metal in DMF. J. Chem. Research 2005,
180-181.
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
ACS Paragon Plus Environment