Synthetic process development of BMS-599793 including azaindole Negishi coupling on kilogram scale

Article abstract of DOI:10.1021/op400012p

A new approach to the synthesis of 1 (DS003, BMS-599793), a small-molecule HIV entry inhibitor, is described. The initial medical chemistry route has been modified by rearranging the sequence of synthetic steps followed by replacement of the Suzuki coupling step by the Negishi conditions. Acylation of the resulting azaindole 7 under the Friedel-Crafts conditions is studied using monoesters of chlorooxalic acid in the presence of aluminum chloride. Polymorphism of 1 is also investigated to develop conditions suitable for preparation of the desired Form 1 of the target compound. The new route is further optimized and scaled up to establish a new process that is applied to the synthesis of kilogram quantites of the target active pharmaceutical ingredient.

Full text of DOI:10.1021/op400012p

Subscriber access provided by University of South Dakota
Full Paper
Synthetic Process Development of BMS-599793
including Azaindole Negishi Coupling on Kilogram Scale
Stanislaw Pikul, Hua Cheng, Allen Cheng, CD Huang, Alexander
Ke, LungHuang Kuo, Andrew S. Thompson, and Susan Wilder
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/op400012p • Publication Date (Web): 24 May 2013
Downloaded from http://pubs.acs.org on June 11, 2013
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 free 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 accessible to all
readers and 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.
Organic Process Research & Development 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 20
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Synthetic Process Development of BMS-599793
including Azaindole Negishi Coupling on Kilogram
Scale
Stanislaw Pikul,1* Hua Cheng,2 Allen Cheng,3 C. D. Huang,3 Alex Ke,3 LungHuang Kuo,3 Andrew Thompson,2
and Susan Wilder4
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 Princeton API Services, LLC, 42 Chase Road, Lumberton, NJ 08048, USA
2 J-Star Reserach, Inc., 3001 Hadley Road, Suites 1-4, South Plainfield, NJ 07080, USA
3 ScinoPharm Taiwan, Ltd., No. 1, Nan-Ke 8th Road, Tainan Science-Based Industrial Park, Shan-Hua, Tainan
County, 74144 Taiwan, R.O.C.
4 International Partnership for Microbicides, 8401 Colesville Road, Suite 200, Silver Spring MD 20910 USA
Abstract
A new approach to the synthesis of 1 (DS003, BMS-599793), a small molecule HIV entry inhibitor, is
described. The initial med-chem route has been modified by rearranging the sequence of synthetic
steps followed by replacement of the Suzuki coupling step by the Negishi conditions. Acylation of
the resulting azaindole 7 under the Friedel-Crafts conditions is studied using mono-esters of
chlorooxalic acid in the presence of aluminum chloride. Polymorphism of 1 is also investigated to
develop conditions suitable for preparation of the desired Form 1 of the target compound. The new
route is further optimized and scaled-up to establish a new process that is applied to the synthesis
of kg quantites of the target active pharmaceutical ingredient.
* Author to whom correspondence may be sent. E-mail: spikul@princetonapi.com
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 20
1
2
3
4
5
6
7
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
Introduction
1 (DS003, BMS-599793) is a small molecule entry inhibitor that interferes with HIV infection by binding
to the gp120 protein.1 The International Partnership for Microbicides (IPM) licensed 1 from Bristol-Myers
Squibb (BMS) with the goal to develop it as a topical microbicide for use in resource-poor countries.
Microbicides are vaginal dosage forms of potent inhibitors of HIV that women can use to prevent sexual
transmission of HIV from male partners. At the time of licensing, a synthesis developed by the BMS
medicinal chemistry group was provided (Scheme 1). Milligram quantities of 1 were prepared at BMS
during the course of the discovery program. This discovery route utilized a Stille coupling reaction in the
last step of the synthesis, yielding 1 containing excessive levels of residual tin that proved difficult to
remove. In addition, the Stille reaction2 required a pyrazine tributyltin reagent (6) that is toxic, expensive,
unpleasant to use, and not readily available in larger quantities. Given these synthetic liabilities, the
medicinal chemistry process was not suitable for scale-up and preparation of pre-clinical material. A new
1 (a) Maddon, P. J.; Dalgleish, A. G.; McDougal, J. S.; Clapham, P. R.; Weiss, R. A.; Axel. R. Cell 1986, 47,
333-348. (b) McDougal, J. S.; Kennedy, M. S.; Sligh, J. M.; Cort, S. P.; Mawle, A.; Nicholson, J. K. Science
1986, 231, 382-385. (c) Moore, J. P.; Jameson, B. A.; Weiss, R. A.; Sattentau, Q. J. in Viral fusion mechanisms.
ed. J. Bentz, CRC Press, Boca Raton, Fla. 1993, p. 233-289.
2 (a) Kosugi, M.; Sasazawa, K.; Shimizu, Y.; Migita, T. Chem. Letters 1977, 6, 301-302. (b) Milstein, D.;
Stille, J. K. J. Am. Chem. Soc. 1978, 100, 3636-3638. (c) Scott, W. J.; Crisp, G. T.; Stille, J. K. Organic Syntheses,
1993, 68, 116-129; 1990, Coll. Vol. 8, 97-103. (d) Stille, J. K.; Echavarren, A. M.; Williams, R. M.; Hendrix, J.
A. Organic Syntheses, 1993, 71, 97-106; 1998, Coll. Vol. 9, 553-558.
,
,
ACS Paragon Plus Environment
2

Page 3 of 20
Organic Process Research & Development
synthesis process was required to address these deficiencies. Our efforts to develop such a process are
described herein.
1
2
3
4
5
6
Scheme 1. Med-chem synthesis of 1
7
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
Results and Discussion
Step sequence change. In the intial approach to address the high residual metal level in 1 an attempt
was made to rearrange the synthetic steps and to perform the Stille coupling as the first, rather than the
last, step of the sequence as shown on Scheme 2. This approach worked well synthetically and provided the
desired product with an overall yield comparable to that obtained by following the original med-chem
route. However, despite the fact that three synthetic steps followed the Stille coupling step to reach 1, the
final material still contained residual levels of tin in the 40 ppm range. This result demonstrated the need
for further research to identify a better alternative to the Stille reaction.
Scheme 2. Modified med-chem route with Stille coupling as the first step.
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 20
1
2
3
4
5
6
7
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
Alternative coupling conditions. At the initial stage the coupling reaction of azaindole 2 was explored
using a pyrazine Grignard reagent derived from iodopyrazine 83 (Scheme 3). Transmetallation of the
Grignard to form a zinc intermediate and its reaction with 2 was performed under standard conditions for
Negishi coupling4 using a palladium catalyst. These conditions afforded the desired coupled product 7 in
yields of over 60%.
The Suzuki reaction5 was also briefly investigated using the pyrazine Grignard to form the corresponding
pyrazine boronic acid derivatives. However, a number of reaction conditions that were tried produced
impure mixtures. Application of pinacol diborane afforded a relatively clean pinacol boronate derivative, but its
Suzuki reaction with 2 proved to be rather slow. No coupled product was observed when the reaction was run at
100 0C, and at 140 0C only a small amount of 7 was observed. In view of these results and the early positive
result with the Negishi reaction, further research focused on the optimization of the Negishi chemistry.
Negishi reaction. The pyrazine Grignard reagent was prepared via an exchange reaction of iodopyrazine
83 using commercially available n-butylmagnesium chloride in THF at -20 0C to -10 0C. Only isopropyl
magnesium chloride was examined as an alternative, but the outcome was less clean. Temperatures of -20
3
Plé, N.; Turck, A.; Heyndericks, A.; Quéguiner, G. Tetrahedron 1998, 54, 9701-9710.
4 Klein, M.; Boche, G. Synthesis 1999, 1246-1250.
5 (a) Miyaura, N.; Yamada, K.; Suzuki, A. Tetrahedron Lett. 1979, 3437-3440. (b) Miyaura, N.; Suzuki, A.
Chem. Commun. 1979, 866-867. (c) Suzuki, A. Pure Appl. Chem. 1991, 63, 419-422. (d) Miyaura, N.; Suzuki,
A. Chem. Rev. 1995, 95, 2457-2483.
ACS Paragon Plus Environment
4

Page 5 of 20
Organic Process Research & Development
0C to -10 0C were found optimal for the exchange reaction. At higher temperatures, the addition of a butyl
group to pyrazine was observed, probably via the reaction of n-butyl halide formed in the Grignard
exchange reaction. The stoichiometry was critical in that excess n-butyl magnesium chloride would
compete with pyrazine in the subsequent coupling reaction and consume expensive azaindole. In addition,
this scenario produced a process impurity that was relatively difficult to remove. The optimized conditions
used exactly one equivalent of the Grignard reagent.
1
2
3
4
5
6
7
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
Scheme 3. Negishi coupling
The next step involved transmetallation of the Grignard reagent with zinc chloride. The zinc chloride
stoichiometry has a large effect on the species produced at this point, and the reactivity of those species.6
When a 1:1 ratio of zinc to magnesium was used, zinc chloride was added to the Grignard reagent at -10 0C
and the mixture was allowed to warm to room temperature and stirred for 1 h under a nitrogen
atmosphere. The zinc species with one aryl ligand and one halide ligand that were formed this way were
effective in the Negishi reaction. Even better results were observed with diaryl zinc species formed using a
0.5 equiv of zinc chloride versus pyrazine Grignard. Under such conditions a good reaction rate and full
consumption of the azaindole were observed within about 6 hours at 60 0C. The reactive zinc species exist
as a suspension and should be used within several hours of preparation. Attempts to filter this mixture to
remove insoluble materials resulted in an unreactive solution.
6 Gauthier, D.R.; Szumigala, R.H. Jr.; Dormer, P.G.; Armstrong, J.D. III.; Volante, R.P.; Reider, P.J. Org. Lett.
2002, 4, 375-378.
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 20
The amount of iodopyrazine, and thus how much diaryl zinc reagent is produced, has a significant effect
on the conversion of this Negishi reaction. Typically, three equiv of iodopyrazine versus 2 were used,
resulting in 1.5 equiv of the diaryl zinc species after Grignard formation and transmetallation with zinc
chloride. Experiments were conducted with less iodopyrazine and it was found that with a ratio of 2:1
(iodopyrazine to azaindole), the Negishi reaction would still go to completion. Those reactions were
conducted through a portionwise addition of the suspension of zinc species to the reaction mixture
containing azaindole 2 and palladium catalyst. With ratios of less than two equiv of starting iodopyrazine
the azaindole was not completely consumed. With ratios in excess of two equiv, the azaindole was
completely consumed, but this was accompanied by the formation of dimer 9 at low levels (usually
<0.6%).7 When completed, the Negishi reaction was subjected to aqueous work-up and the crude product 7
was initially isolated as the hydrochloride salt. The poor aqueous and organic solubility of this salt
facilitated easy removal of all reaction byproducts. The hydrochloride salt was neutralized with aqueous
sodium carbonate and extracted into DCM. DCM was removed and the material precipitated from methanol
to afford 7 in approximately 60% yield and over 99% purity by HPLC.
1
2
3
4
5
6
7
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
The Negishi reation was successfully optimized and scaled up to provide over 2 kg of 7 in several batches
(Table 1).
Table 1. Analysis of kilo-lab scale batches of 7.
7 obtained
Batch 2 used (kg)
Yield (%)
Purity by HPLC (%)
(kg)
1
2
3
4
5
0.4
0.4
0.4
0.4
0.4
0.205
0.210
0.291
0.282
0.294
41
42
59
57
59
98.9
99.0
99.4
99.4
99.4
7 If needed, dimer 9 could be removed by crystallization of the crude product from hexane-ethyl acetate.
ACS Paragon Plus Environment
6

Page 7 of 20
Organic Process Research & Development
6
1.5
0.995
54
99.6
1
2
3
4
5
6
Acylation of Azaindole 7. Acylation of azaindole 7 was achieved by adding the substrate to a mixture of
aluminum chloride in the mixed solvent system of DCM and nitromethane, ~4:1 v/v (Scheme 4).8 Methyl
chlorooxalate was then added and the reaction was allowed to proceed to completion. Once complete, the
reaction was quenched with ammonium acetate, and the product was isolated by an extractive work-up.
Although optimization efforts focused on reaction temperatures and aluminum chloride and methyl
chlorooxalate stoichiometry, the main issue with this step in the synthesis was the formation of unwanted
acid 11 produced from the hydrolysis of the desired intermediate 10a. While 11 is the product of the
subsequent step, it translates to a yield loss in this step, thus the need to reduce its levels. Compound 11
could not be recovered from the aqueous layer and prolonged reaction times used to complete the
conversion of 7 into 10, also led to the formation of 11. Therefore, it was critical to ensure that the
hydrolysis reaction could be suppressed to a low level, allowing large scale manufacturing with the
associated longer processing time cycles. Attempts to suppress the hydrolysis reaction by varying the
aluminum chloride and methyl chorooxalate amounts were explored but, ultimately, it was found that the
reaction temperature was a critical parameter effecting hydrolysis. When run at 0 0C, the acylation step
would procede at a reasonable rate, consuming almost all of 7 in about 15 h with, typically, less than 3% of
the hydrolysis product 11 formed. The hydrolysis reaction would level off at 1-2% and stay at that level for
many hours. When run at ambient temperature, the acylation would again proceed to completion, but the
extent of hydrolysis would increase over time. The results of these laboratory scale experiments are
presented on Figure 1, compare a) versus b) and c).
7
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
8 Although we are aware of a potential hazard associated with using aluminum chloride and
nitromethane, no observations of hazardous results were seen over hundreds of experiments under a
variety of conditions. This data fulfilled our safety requirements at this scale for a fit-for-purpose process.
Future experiments will explore ways to eliminate nitromethane and/or to establish reaction condition
limits for larger scale preparations.
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 20
1
2
3
4
Scheme 4. Acylation of azaindole 6.
5
6
7
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
Lowering the aluminum chloride charge to six equiv resulted in the acylation reaction stalling after 85-
90% conversion (see Figure 1d). Increases in the aluminum chloride charge would increase the rate
without affecting the level of hydrolysis (when run at 0 0C), but the ultimate percent conversion was
unaffected by an increase in aluminum chloride.
Figure 1. Acylation of 7 with methyl chlorooxalate (2 equiv) in the presence of aluminum chloride.
a) 25 0C, 6.2 equiv of AlCl3
100%
b) 0 0C, 8.1 equiv of AlCl3
100%
80%
60%
40%
20%
0%
80%
60%
40%
20%
0%
7
7
10a
11
10a
11
0
10
20
30
0
10
20
30
time (h)
time (h)
c) 25 0C, 7.9 equiv of AlCl3
d) 0 0C, 15 equiv of AlCl3
ACS Paragon Plus Environment
8

Page 9 of 20
Organic Process Research & Development
100%
80%
60%
40%
20%
0%
100%
80%
1
2
3
4
5
6
7
8
60%
7
7
40%
10a
10a
11
20%
0%
11
9
10
11
12
13
14
15
16
17
18
0
10
20
30
0
10
20
30
time (h)
time (h)
Variations in the acylating agent included the use of oxalyl chloride, ethyl chlorooxalate and isopropyl
chlorooxalate. Only a few percent of product formation was observed with oxalyl chloride. The acylation
proceeded slower with ethyl or isopropyl chlorooxalate when compared to methyl chlorooxalate. Finally,
the optimal conditions for the synthesis of acylated intermediate 10 utilized four equiv of methyl
chloroxalate, eight equiv of aluminum chloride and a reaction temperature of 0 0C.
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
Although the acylation reaction using methyl chlorooxalate worked well in the lab, it was still challenging
to control the undesired hydrolysis on a larger scale. In addition, the considerably lower cost of ethyl
chlorooxalate as compared to its methyl analog encouraged further exploration of the reaction. As it turned
out, good yields of the ethyl ester 10b could be obtained by simply extending reaction times to over 8 h
while increasing temperature to 5 0C. The results of multiple kilo-lab batches run with ethyl chlorooxalate
are summarized in Table 2.
Table 2. Analysis of kilo-lab scale batches of acylated azaindole 10b.
10b obtained
Batch 7 used (kg)
Yield (%)
Purity by HPLC (%)
(kg)
1
2
3
4
5
0.11
0.40
0.38
0.37
0.45
0.116
0.398
0.334
0.412
0.474
73
69
61
77
73
99.53
99.70
99.89
99.34
99.60
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 20
1
2
3
4
5
6
7
8
9
Hydrolysis of azaindole ester 10. The penultimate step in the synthesis includes hydrolysis of acylated
azaindole 10 into carboxylic acid 11 (Scheme 5). This step was performed on laboratory scale using 10a
and aqueous potassium carbonate in methanol. Once the reaction was complete methanol was removed
and the mixture diluted with water and extracted with MTBE to remove neutral impurities. The precipitate
obtained after the mixture was acidified to pH ~1 was isolated by filtration to provide the target product
11 in 95% yield and >99% purity by HPLC.
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 5. Hydrolysis of azaindole ester 10.
The laboratory scale hydrolysis conditions were applied in the kilo-lab using ethyl ester 10b as a
substrate. However the work-up was further simplified by eliminating the methanol evaporation and
MTBE extraction steps. Instead, when the hydrolysis was complete the reaction mixture was diluted with
water, acidified to pH ~1 with hydrochloric acid, and then the precipitate was collected by filtration. This
procedure was used to produce 1.2 kg of 11 in 96% yield in a single batch. The obtained product had
99.4% purity by HPLC and was used in the final step of the synthesis without further purification.
Coupling with piperidine 4. The final step of the synthesis of 1 involved HATU mediated coupling of
carboxylic acid 11 with piperidine 4 in DMF to afford 1 (Scheme 6). Ethanol was added to the reaction
mixture and the precipitated product isolated by filtration. In the laboratory-scale experiments 1 was
obtained in approximately 70% yield and >99% purity. However, 16 ppm of Pd, 400 ppm of iron and 100
ACS Paragon Plus Environment
10

Page 11 of 20
Organic Process Research & Development
ppm of zinc impurities were also found requiring that this liability to be addressed before larger-scale
production be attempted.
1
2
3
4
5
6
Scheme 6. Final coupling of carboxylic acid 11 with piperidine 4.
7
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
Alternative coupling conditions were explored for the final synthetic step. HOBT was substituted for
HATU, but this reagent led to a complex mixture of products. Conversion of 11 to the corresponding acid
chloride using oxalyl chloride in DCM9 and subsequent reaction with 4 in the presence of iPr2NEt afforded
1 in only 20% yield. Some effort was made to react methyl ester 10a directly with 4 to afford 1, without
going through the intermediate acid 11. Thus, 4 was treated with NaHMDS, KHMDS, or NaOtBu in THF
followed by addition of compound 10a. Although reported10 to be successful on similar systems, this
transformation did not work in our case.
In addition to a potential excessive level of heavy metals in 1 carried over from the Negishi reaction the
kilo-lab runs were faced with some residual yellow-to-brown discoloration carried over from the Friedel-
Crafts acylation step. To address the color issue the work-up procedure after the final coupling reaction
was modified to include treatment with active carbon. Thus, the solid material obtained after precipitation
9 This acid chloride formation step was not a clean reaction and caused partial decomposition of 10.
10 Private communication from BMS.
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 20
of the product with ethanol was redissolved in DCM11 and treated with Darco KB at 35 0C. After filtration
through a Celite bed, the DCM solution was washed with 5% aqueous ammonium chloride to remove
1
2
3
4
5
6
7
8
9
residual DMF. Following the solvent swap to n-heptane the precipitated product was collected by filtration.
Unfortunately, the crystalline form of 1 (Form 2) obtained this way was different from the desired Form 1
targeted for manufacturing of the drug product and for the pre-clinical toxicology studies.
Polymorphism of 1. To better understand polymorphism of 1 and to develop a procedure to convert
Form 2 to Form 1 a brief polymorphism study was conducted. After full characterization of both forms it
was determined that Form 2 has a monotropic relationship to Form 1. Both forms were shown to be stable at
elevated temperatures and humidities when present as a single form. However, a mixture of Forms 1 and 2 was
seen to gradually convert to Form 1 over time when exposed to elevated temperatures and humidities.
A polymorph screen was carried out to try to identify other forms of 1, as well as conditions under which they
would be formed. Despite many attempts it was not possible to make suitable amorphous material, so, maturation
of amorphous solid in a range of solvents could not be carried out. Therefore the polymorph screen was carried
out using several sovent evaporation or antisolvent addition methods. As a result three polymorphic forms and
two solvated forms (a mono DMSO solvate and a hemi dioxane solvate) were identified and characterized.
Relationships between these forms are summarized in Figure 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
Figure 2. Polymorphic relationships of 1.
11 The use of DCM at this stage of the process is necessary because of very low solubility of 1 in many
organic solvents.
ACS Paragon Plus Environment
12

Page 13 of 20
Organic Process Research & Development
Form 3
(mp 199 0C)
monotropic relation
to Form 1
1
2
3
4
DMF/MEK
5
6
7
8
9
Two
solvated forms
Form 2
(mp 170 0C)
monotropic relation
to Form 1
melt/recryst.
on heating
10
11
12
13
14
15
16
17
18
19
20
21
22
23
melt/recryst.
on heating
heat
Form 1
(mp 223 0C)
most stable form
DCM
To develop a reliable procedure to convert Form 2 to Form 1 a series of slurry experiments was performed
using a range of Class III solvents at ambient, 30 0C and 60 0C temperatures (Table 3). Thirteen solvents were
successful in producing the Form 1 material, but three of these caused a significant reduction in crystallinity. Of
the remaining ten solvents, four required prolonged slurrying for the conversion to occur, which was not suitable
for large-scale use. Of the six solvents that were successful in producing Form 1 material after 1 day of slurrying,
four needed temperatures of 60 0C. From the two solvents: n-propanol and 3-methyl-1-butanol, that efficiently
converted Form 2 to Form 1 in less than 1 day at 30 0C, the former was selected for use on a preparative scale.
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 3. Conversion of Form 2 to Form 1 by slurrying in selected solvents at ambient temperature, 30 0C
and 60 0C for 1 or 7 days.
Solvents used
Complete conversion after
7 days
1 day
Acetone
Anisole
D
C
-
C
-
n-butanol
Aa
A
sec-butanol
n-butyl acetate
C
-
C
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 20
sec-butyl acetate
C
A
-
C
-
1
2
3
4
Cumene
Ethanol
C
5
6
7
8
9
3-methyl-1-butanol
methyl iso-butyl ketone
n-pentanol
B
B
C
Ca
B
C
Ba
Ca
A
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-propanol
iso-propanol
B
A – all temperatures; B – 30 0C and 60 0C; C – 60 0C only; D – 30 0C only
a partial loss of crystallinity
To speed up the polymorphic conversion in the kilo-lab a slurry of Form 2 in n-propanol was heated to 90 0C,
and 1 h at this temperature was sufficient to achieve complete conversion of Form 2 to the desired Form 1. After
cooling, filtration and drying of the solid, 1.856 kg of the desired Form 1 of 1 was obtained in two batches (Table
4). Based on the largest batch size performed, the overall yield of the four-step process was 32% with purity of
the final material exceeding 99.5%. In addition, the process executed in the kilo-lab provided 1 containing less
than 1 ppm each of palladium, iron and zinc, and less than 3 ppm of each magnesium and aluminum. This
material was used in formulation studies and in preclinical toxicology evaluation of 1.
Table 4. Analysis of kilo-lab scale batches of the taget 1.
11 used
1 obtained
Purity by HPLC
(Area %)
Batch
Yield (%)
(kg)
(kg)
1
2
0.700
0.705
0.903
0.953
80
84
99.58
99.60
Conclusions
ACS Paragon Plus Environment
14

Page 15 of 20
Organic Process Research & Development
A robust synthesis of 1, which does not involve tin reagents, has been developed. This improved process
was scaled up to yield kilogram quantities of. The Stille coupling step was moved from the last step of the
med-chem synthesis route to the first step of the new process, and has been further replaced by the
Negishi conditions to eliminate the tributyltin-based reagent that was used previously. The protocol for the
acylation step to produce ester 10 was optimized and inexpensive ethylchlorooxalate was introduced to
allow for kilogram scale runs without concern over ester hydrolysis. Conversion of ester 10 into 1 was
accomplished in two steps including ester hydrolysis followed by HATU mediated coupling with piperidine
4. Based on a polymorphism study of 1, an efficient procedure for the conversion of the undesired
polymorphic Form 2 of 1 produced in the process, to the most stable Form 1 of 1 was established. The new
process allows for the synthesis of kilogram quantities of 1 in 30-35% overall yield starting from azaindole
1. The final product is obtained with high purity of over 99.5% and heavy metal impurities meeting
standard ACH guidelines.
1
2
3
4
5
6
7
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
Experimental Section
General: Chloroazaindole 2, piperidine 4, iodopyrazine 8 used in the work described here were custom
made in kg quantities by Z.D. Chemipan, Warsaw, Poland.
HPLC method used for monitoring laboratory scale experiments:
Column: Phenomenex Jupiter C18 300A, 4.6x250 mm, 5 µm of particle size
Flow rate: 1.5 mL/min
Injection volume: 10 µL
Detection: UV, 220 nm
Column temp.: 30 0C
Run time: 56 minutes
Mobile phase A: 0.025%TFA in water
Mobile phase B: 0.025%TFA in ACN
Gradient Elution
Time (min)
0
A (%)
90
B (%)
10
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 20
15
17
18
22
10
10
90
90
90
90
10
10
1
2
3
4
5
6
7
8
9
Retention times: compound 1 - 10.8 min, compound 2 - 6.9 min, compound 4 - 6.7 min, compound 7 - 5.5 min,
compound 9 - 4.9 min, compound 10a - 6.3 min, compound 11 - 4.1min.
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-Methoxy-7-(pyrazin-2-yl)-1H-pyrrolo[2,3-c]pyridine (7, kilo-lab scale process). To a solution of
iodopyrazine 8 (721 g, 3.5 mol) in THF (8.4 L) at -18 0C, n-BuMgCl (2M in THF, 1.75 L, 3.5 mol) was added
over 30-60 min. While maintaining the temperature at -18 to -20 0C ZnCl2 (0.5M in THF, 3.5 L, 1.75 mol)
was added over 30-60 minutes and the mixture was warmed up to 25 0C in 2 h. Chloroazaindole 1 (210 g,
0
1.15 mol) and PdCl2(dppf)*CH2Cl2 (100 g, 0.122 mol) were added and mixture was heated at 58 C for 8 h
when HPLC analysis showed >20:1 ratio of the product 7 to the azaindole 2. The reaction mixture was
cooled to 0 0C and quenched with NH4Cl (6 N aqueous, 2.5 L). The inorganic salts were filtered off and
washed with THF (420 mL). The combined organic solutions were treated with aqueous HCl (1.2 L of conc
HCl plus 7 L of water). The mixture was stirred at 25 0C for 1 h and the precipitated hydrochloride of
azaindole 7 was filtered and washed with ethyl acetate (2.1 L). The solids were suspended in DCM (8 L),
aqueous sodium carbonate (10%, 4 L) was added and the mixture was stirred for 2 h at 25 0C. The
precipitated salts were removed by filtration and the organic phase was separated and concentrated to ~
0.4 L volume. Methanol (1.5 L) was added and the solvents were distilled off until a volume of ~ 0.5 L was
reached. The process of methanol addition and solvent distillation was repeated two more times and the
mixture was stirred at 25 0C for 1 h. The precipitate was filtered, washed with methanol (0.3 L) and dried
to give 158 g (60.2% yield) of the azaindole 7 (99.5% pure by HPLC).12 1H NMR (400 MHz, DMSO-d6) 4.08
(s, 3H), 6.64 (d, J = 2.8 Hz, 1H), 7.56 (t, J = 2.4 Hz, 1H), 8.03-8.05 (m, 1H), 8.64 (d, J = 2.4 Hz, 1H), 8.74-8.75
(m, 1H), 9.62 (d, J = 1.6 Hz, 1H), 11.78 (br, s, 1H); 1H NMR (400 MHz, CDCl3) 4.10 (s, 3H), 6.70 (dd, J = 2.6
12 The lab trial material prepared this way contained 1669 ppm Fe and 83 ppm Zn when tested by ICP-
MS.
ACS Paragon Plus Environment
16

Page 17 of 20
Organic Process Research & Development
Hz, 3.0 Hz, 1H), 7.36 (dd, J = 2.6 Hz, 3.0 Hz, 1H), 8.48-8.51 (m, 2H), 9.80 (d, J = 1.6 Hz, 1H), 10.88 (br s, 1H);
13C NMR (100 MHz, CDCl3) 56.1, 99.5, 120.5, 124.5, 126.9, 131.5, 133.2, 142.2, 142.5, 143.5, 150.9, 152.4;
LCMS: m/e 227 (M+H)+.
1
2
3
4
5
6
Methyl 2-(4-methoxy-7-(pyrazin-2-yl)-1H-pyrrolo[2,3-c]pyridin-3-yl)-2-oxoethanoate (10a,
laboratory scale process). To a solution of dichloromethane and nitromethane (4:1, 200 mL) cooled with
ice-water bath, was added portionwise AlCl3 (22.3 g, 168 mmol) followed by azaindole 7 (4.75 g, 21.0
mmol). CAUTION: mixtures of CH3NO2 and AlCl3 are potentially hazardous. Conduct a proper safety
evaluation before running this chemistry. The internal temperature was raised from 1 0C to 6 0C then
lowered back to 1 0C. Methyl chlorooxalate (3.9 mL, 41.1 mmol) was added into the solution dropwise in
ca. 5 min. The resulting homogeneous solution was stirred at 0 0C for 10 min and then put in a cold room
(ca. 0 0C) for 15 h without stirring. Analysis by HPLC after 15 h showed the ratio of 7:10a:11 to be 0:92:3.
The reaction solution was poured into cold 25% aqueous NH4OAc solution (500 mL) and the organic layer
was separated. The aqueous layer was extracted with DCM (300 mL, then 2x150 mL) and the combined
organic layers were washed with brine (2 x 300 mL), and dried (Na2SO4). Removal of solvents in vacuo
gave ester 10a as a solid (4.85 g, 74% yield), which was used in the next step without further purification.
1H NMR (300 MHz, CDCl3) 3.95 (s, 3H), 4.10 (s, 3H), 8.15 (s, 1H), 8.35 (d, J = 3.3 Hz, 1H), 8.57-8.60 (m, 2H),
9.81(d, J = 1.2 Hz, 1H), 11.7 (br s, 1H). Analysis by ICP-MS showed 1535 ppm Fe, 103 ppm Zn.
Ethyl 2-(4-methoxy-7-(pyrazin-2-yl)-1H-pyrrolo[2,3-c]pyridin-3-yl)-2-oxoethanoate (10b, kilo-lab
scale process). To a solution of dichloromethane (7 L) and nitromethane (1.9 L) at 5 0C was added in
portions AlCl3 (930 g, 7.0 mol), followed by ethyl chlorooxalate (193 mL, 1.73 mol) and azaindole 7 (194 g,
0.86 mol). CAUTION: mixtures of CH3NO2 and AlCl3 are potentially hazardous. Conduct a proper safety
evaluation before running this chemistry. The reaction mixture was stirred at 5 0C for 8 h and progress
monitored by HPLC. When the analysis indicated complete disappearance of the starting material the
mixture was slowly added to a solution of ammonium acetate (4.75 kg) in water (15 L) cooled to 0 0C. The
organic layer was separated and the aqueous layer was extracted with DCM (2 x 3 L). The combined
organic layers were concentrated in vacuo to ~ 0.4 L volume. Methanol (1.0 L) was added and the solvents
were distilled off until volume of ~ 0.5 L was reached. The process of methanol addition and solvent
7
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
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 20
distillation was repeated one more time and the mixture was stirred at 25 0C for 1 h. The precipitate was
filtered off, washed with methanol (0.2 L) and dried to give 170 g (60.75% yield) of the azaindole 10b
(99.6% pure by HPLC). 1H NMR (400 MHz, CDCl3) 1.41 (t, J = 7.2 Hz, 1H), 4.07 (s, 3H), 4.42 (q, J = 7.2 Hz,
1H), 8.08 (s, 1H), 8.31 (d, J = 3.2 Hz, 1H), 8.54-8.57 (m, 2H), 9.76 (d, J = 1.2 Hz, 1H), 11.69 (br s, 1H); 13C
NMR (100 MHz, CDCl3) 14.1, 56.5, 62.1, 114.3, 121.3, 124.1, 131.8, 133.6, 136.2, 142.2, 143.3, 143.5, 151.1,
151.5, 164.2, 180.8
1
2
3
4
5
6
7
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
2-(4-methoxy-7-(pyrazin-2-yl)-1H-pyrrolo[2,3-c]pyridin-3-yl)-2-oxoethanoic acid (11) by
hydrolysis of 10a. To a suspension of ester 10a (10.00 g, 32.1 mmol) in methanol (150 mL) at room
temperature was added aqueous K2CO3 (1M, 150 mL, 150 mmol). The reaction mixture was stirred for 1 h
when complete hydrolysis of 10a was indicated by HPLC analysis. Methanol was removed under vacuum
and the remaining mixture was diluted with water to 1.2 L and washed with MTBE (2 x 400 mL). Aqueous
phase was acidified with HCl (2M, 185 mL, 370 mmol) to pH=1 and the gray precipitate was filtered off and
dried to give 9.29 g of acid 11 (97% yield). Analysis by ICP-MS showed 143 ppm Fe, 96 ppm Zn.
2-(4-methoxy-7-(pyrazin-2-yl)-1H-pyrrolo[2,3-c]pyridin-3-yl)-2-oxoethanoic acid (11) by
hydrolysis of 10b (kilo-lab scale process). A reactor was charged with ester 10b (1.61 kg, 4.93 mol),
methanol (1.5 L) and aqueous K2CO3 (1 M, 6.5 L). The reaction mixture was heated to 40 0C and stirred for
1 h when complete hydrolysis of 10b was indicated by HPLC analysis. Water (15 L) was added the mixture
was acidified to pH ~1 with 2 N hydrochloric acid. The precipitate was filtered off, washed with water (12
L) and dried to give 1.416 kg of acid 11 (96% yield). 1H NMR (400 MHz, DMSO-d6) 4.04 (s, 3H), 8.21 (s, 1H),
8.32 (d, J = 1.6 Hz, 1H), 8.68 (d, J = 2.4 Hz, 1H), 8.74 (dd, J = 1.6, 2.4 Hz, 1H), 9.59 (d, J = 1.6 Hz, 1H), 12.64
(br s, 1H); 13C NMR (100 MHz, DMSO-d6) 56.5, 113.2, 121.5, 124.3, 132.6, 132.9, 138.6, 142.9, 143.5, 144.0,
151.0, 151.3, 166.9, 183.1.
2-(1-(2-(4-methoxy-7-(pyrazin-2-yl)-1H-pyrrolo[2,3-c]pyridin-3-yl)-2-oxoethanoyl)piperidin-4-
ylidene)-2-phenylethanenitrile (1, laboratory scale process). A flask was charged with acid 11 (9.29 g,
31.2 mmol), DIPEA (12.9 mL, 78 mmol), 4 (7.18 g, 36.3 mmol) and DMF (95 mL) subsequently. HATU
(13.66 g, 35.9 mmol) was added to the reaction mixture over 10 minutes, which was accompanied by
increase of internal temperature from 19 0C to 27 0C. After the reaction mixture was stirred at 25 0C for 3.5
ACS Paragon Plus Environment
18

Page 19 of 20
Organic Process Research & Development
h the HPLC analysis showed complete disappearance of acid 11. Ethanol (950 mL) was added and the
resulting suspension was heated at reflux for 1 h. The mixture was then cooled to 25 0C and 1 was isolated
by filtration and washed with ethanol (50 mL). The material was dried under vaccum at 50 0C to afford
10.58 g (71% yield) of 1 as a colorless solid. 1H NMR (300 MHz, CDCl3) 2.58-2.65 (m, 2H), 2.91-2.99 (m,
2H), 3.48-3.51(m, 1H), 3.68-3.78 (m, 2H), 3.95-3.99 (m, 1H), 4.11 (s, 3H), 7.27-7.46 (m, 5H), 8.16 (d, J = 5.1
Hz, 1H), 8.21-8.25 (m, 1H), 8.60 (s, 2H), 9.82 (d, J = 3.9 Hz, 1H), 11.75(br s, 1H). LCMS: m/e 479.3 (M+H)+.
Analysis by ICP-MS showed 16 ppm Pd, 79 ppm Fe, 102 ppm Zn. This material was found to be a mixture of
two polymorphs: Form 1 and Form 2.
1
2
3
4
5
6
7
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
2-(1-(2-(4-methoxy-7-(pyrazin-2-yl)-1H-pyrrolo[2,3-c]pyridin-3-yl)-2-oxoethanoyl)piperidin-4-
ylidene)-2-phenylethanenitrile (1, kilo-lab scale process including polymorph conversion). A reactor was
sequentially charged at 25 0C with DMF (4 L), acid 11 (705 g, 2.36 mol), piperidine 4 (668 g, 2.85 mol, 1.2
eq) and DIPEA (1.8 L, 4.5 eq). To the clear solution was added a solution of HATU (1.080 kg, 2.85 mol, 1.2
eq) and the mixture was stirred at 25 0C until HPLC analysis indicated complete disappearance of acid 11.
Ethanol (48 L) was added and the mixture was stirred for 1 hour. The precipitate was collected by
filtration, washed with ethanol (10 L) and dried by purging nitrogen for 4 h. The crude product was
dissolved in DCM (80 L), Darco KB (2.3 kg) was added and the mixture was stirred at 30 0C for 1 hour.
Darco KB was removed by filtration through a bed of Celite and the clear solution was washed with 5%
aqueous solution of ammonium chloride (41 L). The DCM solution was concentrated to a volume of ~ 3 L
by vaccum distillation, heptane (21 L) was added and the mixture was concentrated again to a volume of ~
3 L. This step was repeated one more time using 10 L of heptane. The resulting suspension was filtered; the
precipitate was washed with heptane (3.5 L) and dried under vaccum to give 1 as a pure Form 2
polymorph. This material was suspended in n-propanol (15 L) and the mixture was stirred for 1 h at 90 0C.
After cooling to ambient temperature the precipitate was filtered, washed with n-propanol (8 L) and dried
under vaccum at 65 to give 0.953 kg of 1 (84% yield) and 99.6% purity by HPLC. This material was a pure
Form 1 polymorph. 1H NMR (400 MHz, CDCl3) 2.61 (t, J = 5.8 Hz, 1H), 2.65 (t, J = 5.9 Hz, 1H), 2.94 (t, J = 5.8
Hz, 1H), 2.99 (t, J = 5.9 Hz, 1H), 3.51 (t, J = 5.8 Hz, 1H), 3.72 (t, J = 5.8 Hz, 1H), 3.78 (t, J = 6.0 Hz, 1H), 3.98 (t,
J = 6.0 Hz, 1H), 4.12, 4.12 (two s, 3H), 7.29-7.47 (m, 5H), 8.16, 8.18 (two s, 1H), 8.23, 8.25 (two d, J = 3.1 Hz,
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 20
1H), 8.60-8.63 (m, 2H), 9.83, 9.84 (two d, J = 1.4 Hz, 1H), 11.76, 11.78 (two br s, 1H); 13H NMR (100 MHz,
1
2
3
4
5
6
7
8
9
CDCl3) 30.3, 31.0, 33.7, 34.3, 41.7, 42.0, 46.0, 46.3, 56.8, 111.4, 115.0 (2C), 117.7 (2C), 120.9, 124.2, 128.9
(2C), 129.0, 129.1 (2C), 131.9, 132.6 (2C), 133.9 (2C), 136.6, 142.2, 143.4, 143.7 (2C), 151.1, 151.3, 154.8,
166.4, 166.5, 185.6 Anal. Calcd for C27H22N6O3: C, 67.77; H, 4.63; N, 17.56. Found: C, 67.84; H, 4.64; N,
17.56.
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
Acknowledgement
We thank Jerzy Raczko of Chemipan, Warsaw, Poland for providing kg quantities of reagents 2, 4, and 8,
Emma Sharp and Adrian St. Clair Brown of Pharmorphix Ltd., Cambridge, UK, for performing
polymorphism studies of 1.
Supporting Information Available
1H and 13C NMR of 1, DSC and XRPD traces of Form 1 and Form 2 of 1. This material is available free of
charge via Internet at http:/pubs.acs.org.
ACS Paragon Plus Environment
20