A Biocatalytic Route to the Novel Antiepileptic Drug Brivaracetam

Article abstract of DOI:10.1021/acs.oprd.6b00094

An asymmetric synthesis of the novel antiepileptic drug Brivaracetam 1 is described. The stereochemistry of the 4-n-propyl substituent is introduced by a biocatalytic resolution of (rac)-methyl 2-propylsuccinate 4-tert-butyl ester 4. The selection of the resolution substrate and the screening of enzymes were carried out from our in-house screening platform. The development and scale-up of the best conditions, including solvent media, pH control, workup, and enzyme supply, led up to a successful demonstration conducted at 1 kg scale in a 10 L vessel. The chiral intermediate (R)-2-propylsuccinic acid 4-tert-butyl ester 6 was reproducibly obtained in 42% yield and 97% ee all along the development. The control of the stereochemistry via the biocatalytic resolution allowed the production of Brivaracetam 1 within the required commercial quality specifications.

Full text of DOI:10.1021/acs.oprd.6b00094

Subscriber access provided by Northern Illinois University
Full Paper
A bio-catalytic route to the novel anti-epileptic drug Brivaracetam
Arnaud Schule, Alain Merschaert, Christophe Szczepaniak,
Christophe Marechal, Nicolas Carly, John O'Rourke, and Celal Ates
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.6b00094 • Publication Date (Web): 24 Aug 2016
Downloaded from http://pubs.acs.org on August 25, 2016
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 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
A bioꢀcatalytic route to the novel antiꢀepileptic drug
Brivaracetam
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
Arnaud Schülé*, Alain Merschaert, Christophe Szczepaniak, Christophe Maréchal, Nicolas
Carly, John O’Rourke and Célal Ates.
UCB Pharma S.A., Chemical Process Research and Development, Pharma Sciences, Chemin du
Foriest, 1420 Braine l’Alleud, Belgium.
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 31
1
2
3
4
Table of Contents
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
Enzymatic
resolution
42% yield
97.1% ee
4
6
Brivaracetam 1
ACS Paragon Plus Environment
2

Page 3 of 31
Organic Process Research & Development
1
2
3
4
Abstract
5
6
7
8
9
An asymmetric synthesis of the novel antiepileptic drug Brivaracetam 1 is described. The
stereochemistry of the 4ꢀnꢀpropyl substituent is introduced by a bioꢀcatalytic resolution of (rac)ꢀ
methyl 2ꢀpropylsuccinate 4ꢀtertꢀbutyl ester 4. The selection of the resolution substrate and the
screening of enzymes were carried out from our inꢀhouse screening platform. The development
and scaleꢀup of the best conditions including solvent media, pH control, workꢀup and enzyme
supply led up to a successful demonstration conducted at 1 kg scale in a 10 L vessel. The chiral
intermediate (R)ꢀ2ꢀpropylsuccinic acid 4ꢀtertꢀbutyl ester 6 was reproducibly obtained in 42%
yield and 97% ee all along the development. The control of the stereochemistry via the bioꢀ
catalytic resolution allowed the production of Brivaracetam 1 within the required commercial
quality specifications.
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: enzymatic resolution, scaleꢀup, process chemistry, Brivaracetam
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 31
1
2
3
4
Introduction to Brivaracetam and control of the stereochemistry
5
6
7
8
9
Brivaracetam 1, (2S)ꢀ2ꢀ[(4R)ꢀ2ꢀoxoꢀ4ꢀpropyltetrahydroꢀ1Hꢀpyrrolꢀ1ꢀyl]butanamide also referred
to as ucb 34714, is a novel antiepileptic drug (AED) with a highꢀaffinity for synaptic vesicle
protein 2A ligand1 (SV2A), that successfully went through clinical Phase III2. Brivaracetam 1
was recently approved in Europe and United States as adjunctive therapy in the treatment of
partial onset seizures in patients of 16 years age and older with epilepsy under the trade name of
Briviact®.
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 control of the stereochemistry of the 4ꢀnꢀpropyl moiety is the major challenge from a
synthetic perspective. The stereochemistry of the (2S) center might be accessed and controlled by
the widely available (S)ꢀ2ꢀaminobutanamide 2 building block. In our continuous efforts to find
new and potentially useful synthetic approaches that could be applied on large scale, the
transformations on the succinate derivatives 3ꢀ6 were selected for further investigation (scheme
1).
Scheme 1. Succinate intermediates towards Brivaracetam
2
, R=Et,
3
, R=Me
4
, R=H
5
6
Brivaracetam ,
ucb 34714
ucb 34713
1
The transformation from the racemic series 3ꢀ5 into the chiral intermediate 6 was envisaged by
both an enzymatic resolution3 (3ꢀ4→6) and a salt resolution (5→6). The synthesis of the three
resolution substrates 3, 4 and 5 at laboratory scale is described in the Supporting Information.
ACS Paragon Plus Environment
4

Page 5 of 31
Organic Process Research & Development
1
2
3
4
Enzymatic resolution approach
5
6
7
8
9
The evaluation of the bioꢀcatalytic resolution of substrates 3 and 4 was performed with the
screening of four families of enzymes4: lipases (30 enzymes), acylase (1 enzyme), protease (15
enzymes) and esterase (30 enzymes). The objective was to be able to perform readily a
demonstration at scale if some hits were identified in the perspective of a potential
industrialization.
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 2. Bioꢀcatalytic resolution of (rac)ꢀethyl ester 3 and (rac)ꢀmethyl ester 4
Enzymatic
resolution
7, R=Et
8, R=Me
3, R=Et
4, R=Me
6
The screening of enzymes was performed in the following conditions: 20 mg of substrate, ca.
50% w/w of enzyme load, a phosphate buffer (KH2PO4, K2HPO4, 0.1 M, pH 7.5, 90 mL/g),
methyl tertꢀbutyl ether (MTBE) as coꢀsolvent (12.5 mL/g), 27.5 °C during ca. 16 hours on a
shaking plate at 250 rpm5.
The full screening of the enzymes was performed on the ethyl ester substrate 3. No conversion
was observed with the acylase and esterase families. The best results were obtained with lipase
and protease families, especially alkaline protease B and protease C from Bacillus subtilis with
97ꢀ98% ee. The results are summarized in the table 1.
Table 1. Hydrolysis results on (rac)ꢀethyl 2ꢀpropylsuccinate 4ꢀtertꢀbutyl ester 3
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 31
1
2
3
4
5
6
7
8
9
Subꢀclass
of the
hydrolase
Lipase
Lipase
ee of (S)ꢀ
ethyl ester succinic
7 (%)
11
69
64
15
ee of (R)ꢀ
Conversion
(%)
Strain of the hydrolase
E value6
acid 6 (%)
A from Rhizomucor miehei 13
75
63
97
98
8
9
C from Rhizomucor miehei
Alkaline B
C from Bacillus subtilis
52
40
14
Protease
Protease
148
96
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 influence of alternative coꢀsolvents and the absence of a coꢀsolvent were also evaluated on
the four identified hits by testing cyclohexane, dimethyl sulfoxide, acetone and toluene, all other
parameters were kept unchanged. The results are reported in the table 2.
Table 2. Influence of alternative coꢀsolvents and absence of a coꢀsolvent
Subꢀclass
of the
hydrolase
Protease Alkaline B
Protease C from Bacillus subtilis
Lipase
Lipase
Protease Alkaline B
ee of (S)ꢀ
ee of (R)ꢀ
ethyl ester succinic
Coꢀ
solvent
Conv.
(%)
Strain of the hydrolase
E value
7 (%)
99
64
82
ꢀ
acid 6 (%)
54
40
58
100
33
0
85
98
59
ꢀ
98
ꢀ
65
40
81
94
65
ꢀ
77
95
70
30
ꢀ
65
190
9
No
A from Rhizomucor miehei
C from Rhizomucor miehei
ꢀ
47
ꢀ
176
ꢀ
Protease C from Bacillus subtilis
Cycloꢀ
Lipase
Lipase
Protease Alkaline B
Protease C from Bacillus subtilis
Lipase
Lipase
A from Rhizomucor miehei hexane
C from Rhizomucor miehei
60
71
55
51
99
99
99
99
72
ꢀ
23
10
48
167
10
ꢀ
38
210
9
8
ꢀ
ꢀ
ꢀ
Dimethyl
A from Rhizomucor miehei sulfoxide 53
C from Rhizomucor miehei 100
Protease Alkaline B
Protease C from Bacillus subtilis
56
51
41
77
≤ 15
0
≤ 15
≤ 15
99
99
48
99
ꢀ
ꢀ
ꢀ
ꢀ
Acetone
Toluene
Lipase
Lipase
A from Rhizomucor miehei
C from Rhizomucor miehei
Protease Alkaline B
Protease C from Bacillus subtilis
ꢀ
ꢀ
ꢀ
Lipase
Lipase
A from Rhizomucor miehei
C from Rhizomucor miehei
ꢀ
ACS Paragon Plus Environment
6

Page 7 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
In general, the results using the protease hits were better than the ones using the lipase hits: a
conversion close to 50%, a high ee for the desired product 6 and also a higher E value. Dimethyl
sulfoxide and acetone emerged as alternative coꢀsolvents to methyl tertꢀbutyl ether used in the
screening. The absence of a coꢀsolvent also led to similar good results. The lipase hits appeared
to be less stereoselective with higher conversions, sometimes nearly complete.
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
Based on the resolution results obtained with the ethyl substrate 3, the methyl derivative 4 was
tested with the four enzymes hits only in order to optimize the screening duration. The screening
conditions were the same than the ones used for the ethyl derivative 3, expending the screening
to the best coꢀsolvents identified with the ethyl substrate 3. The results are reported in the table 3.
Table 3. Hydrolysis results on (rac)ꢀmethyl 2ꢀpropylsuccinate 4ꢀtertꢀbutyl ester 4
Subꢀclass Strain of the hydrolase
of the
Coꢀ
solvent
Conv.
(%)
ee of (S)ꢀ
methyl ester succinic
ee of (R)ꢀ E value
hydrolase
8 (%)
acid 6 (%)
Protease
Protease
Lipase
Alkaline protease B
57
52
61
100
50
≤ 10
62
~75
99
99
95
ꢀ
96
ꢀ
76
91
60
ꢀ
96
ꢀ
37
115
14
ꢀ
180
ꢀ
C from Bacillus subtilis
A from Rhizomucor miehei
C from Rhizomucor miehei
Alkaline protease B
C from Bacillus subtilis
A from Rhizomucor miehei
C from Rhizomucor miehei
No
Lipase
Protease
Protease
Lipase
Cycloꢀ
hexane
99
ꢀ
62
ꢀ
21
ꢀ
Lipase
Protease
Protease
Lipase
Alkaline protease B
C from Bacillus subtilis
A from Rhizomucor miehei sulfoxide 52
C from Rhizomucor miehei
~74
55
ꢀ
ꢀ
ꢀ
99
77
ꢀ
83
72
ꢀ
54
14
ꢀ
Dimethyl
Lipase
~98
Protease
Protease
Lipase
Alkaline protease B
64
53
~24
55
99
99
ꢀ
82
99
ꢀ
57
89
ꢀ
68
94
ꢀ
17
93
ꢀ
13
162
ꢀ
C from Bacillus subtilis
A from Rhizomucor miehei
C from Rhizomucor miehei
Alkaline protease B
C from Bacillus subtilis
A from Rhizomucor miehei
C from Rhizomucor miehei
Acetone
MTBE
Lipase
Protease
Protease
Lipase
51
~33
≤ 10
~28
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
Lipase
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 31
1
2
3
4
5
6
7
8
9
The trends obtained previously with the ethyl substrate 3 were also observed with the methyl
derivative 4. Proteases were again more efficient than the lipases that appeared significantly less
stereoselective. At this stage of the screening, with the results in hands and insight on enzymes
prices at larger scale7, the two proteases hits were selected to progress further in the development
of the enzymatic resolution.
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
Both ethyl and methyl substrates 3 and 4 were similarly attractive at this stage of the evaluation.
In order to differentiate the two substrates, parameters of the screening conditions were
challenged to reflect more realistic industrial resolution conditions: 10%ꢀw of enzyme load (vs.
50% w/w), 25 mL/g of phosphate buffer/pH 7.5 (vs. 90 vol), 2 mL/g of coꢀsolvent (vs. 12.5
mL/g) and 30 °C for the temperature. The tests were performed on 100 mg of substrates 3 and 4.
The results are reported in the table 4.
Table 4. Hydrolysis results on 3 and 4 with more realistic operating conditions
ee of (S)ꢀ
derivative 7 succinic
or 8 (%)
75
39
ee of (R)ꢀ
Conv.
(%)
Substrate Strain of the hydrolase
Coꢀsolvent
No
E value
acid 6 (%)
97
98
98
99
4, R=Me
4, R=Me
3, R=Et
3, R=Et
4, R=Me
4, R=Me
4, R=Me
3, R=Et
3, R=Et
4, R=Me
4, R=Me
3, R=Et
3, R=Et
4, R=Me
3, R=Et
Alkaline B
C from Bacillus subtilis
Alkaline B
C from Bacillus subtilis
Alkaline B
Alkaline B
C from Bacillus subtilis
Alkaline B
C from Bacillus subtilis
Alkaline B
C from Bacillus subtilis
Alkaline B
43
28
31
6
149
126
154
211
135
162
170
182
156
168
139
182
169
143
175
44
6
Cyclohexane 33
45
Dimethyl
sulfoxide
48
80
57
59
27
82
50
55
98
97
98
98
98
97
98
98
36
38
22
46
34
36
20
43
25
Acetone
MTBE
C from Bacillus subtilis
Alkaline B
Alkaline B
25
72
32
99
97
98
ACS Paragon Plus Environment
8

Page 9 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
In these resolution conditions, the methyl ester derivative 4 emerged as the best resolution
substrate with both proteases. Additionally, the methyl ester derivative 4 was also more attractive
from a synthesis perspective thanks to the efficient oneꢀpot dealkoxycarbonylation
transformation (scheme 3). The alkaline protease B gave also better results than the protease C
from Bacillus subtilis allowing to nominate a lead enzyme and a backꢀup one if necessary. It was
worthwhile to note that the absence of a coꢀsolvent gave similar results than the resolution using
a coꢀsolvent. This option without a coꢀsolvent could be attractive to simplify the workꢀup of the
reaction.
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 temperature influence was also evaluated (35 °C vs. 30 °C) without significant impact on the
outcome. In a similar way, the quality of the substrate – a polishing distillation vs. a
chromatography – did not reveal any resolution difference.
Work-up of the enzymatic resolution
The opportunity to conduct the resolution without any coꢀsolvent was prioritized to simplify the
workꢀup. Once the resolution completed (pH 7.5), the unreacted (S)ꢀmethyl ester derivative 8
was extracted with cyclohexane. The analysis of the discarded organic layer showed that ca. 12%
of the succinic acid derivative 6 was also eliminated. To circumvent that issue the pH was
adjusted at 9 before the extraction. In these conditions, the product 6 was totally ionized into the
succinate form and remained in the aqueous phase without any racemization identified. The pH
was then adjusted to pH 1 in order to extract the product 6 with isopropyl acetate. During both
liquidꢀliquid separations at pH 9 and pH 1, some foaming was observed probably due to protein
material but also due to amphiphilic properties of the product 6. Moreover, at pH 1, the
precipitation of the denatured enzyme was observed. After some tests, the most appropriate
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 31
1
2
3
4
5
6
solution found was to preꢀfilter the biphasic mixture prior to separation with addition of
diatomaceous earth (Celite®) which enhances the filtration by adsorbing the protein material.
7
8
9
The final workꢀup procedure was conducted in the following conditions: the unreacted ester 8 is
extracted in cyclohexane (1 x 2 mL/g) at pH 9 (adjustment with NaOH 0.1 M solution, ca. 0.2
mL/g). The pH is then adjusted to pH=1 with HCl 3 M (1 mL/g). Isopropyl acetate (3 mL/g) and
Celite (17% w/w) are added then the slurry is filtered. The cake is washed with isopropyl acetate
(1 mL/g) and the liquidꢀliquid separation is performed. The organic layer is finally evaporated to
give 6 as an oil8.
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
Supply of the enzyme at scale
At the time of performing the initial scaleꢀup, the most promising enzymes – protease alkaline B
and protease C from Bacillus subtilis – could unfortunately not be provided immediately by the
supplier at 2 kg scale. In agreement with the supplier, alternative enzymes from the same strains
available from the supplier’s stock were therefore quickly evaluated.
The backꢀup protease C from Bacillus subtilis was replaced by two similar strains referenced as
type 1 and type 2 respectively. The lead enzyme protease alkaline B was replaced by alkaline B
n°2 whose formulation was described as different. The enzymatic resolution was tested on the
methyl ester substrate 4 in the second screening conditions: 100 mg of substrate, 10%w/w of
enzyme load, 25 mL/g of phosphate buffer (pH 7.5), 2 mL/g of coꢀsolvent (or not) and 30 °C for
the temperature. The results are reported in the table 5.
Table 5. Hydrolysis results on 4 with new operating conditions
ACS Paragon Plus Environment
10

Page 11 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
ee of (S)ꢀ
methyl
ester 8 (%)
20
19
48
11
6
ee of (R)ꢀ
succinic
acid 6 (%)
97
99
97
95
99
97
96
98
97
97
98
97
96
Conv.
(%)
Coꢀsolvent
No
Strain of the protease
Alkaline B n°2
E value
17
117
240
105
43
211
83
Protease C from Bacillus subtilis (type 1) 16
Protease C from Bacillus subtilis (type 2) 33
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
Alkaline B n°2
Protease C from Bacillus subtilis (type 1)
11
6
Cyclohexane
DMSO
Protease C from Bacillus subtilis (type 2) 20
Alkaline B n°2
25
9
9
56
Protease C from Bacillus subtilis (type 1) 15
Protease C from Bacillus subtilis (type 2) 36
18
54
34
17
56
23
2
121
116
90
144
138
69
Alkaline B n°2
26
Protease C from Bacillus subtilis (type 1) 15
Protease C from Bacillus subtilis (type 2) 36
Acetone
MTBE
Alkaline B n°2
Protease C from Bacillus subtilis (type 1)
20
2
99
98
203
114
Protease C from Bacillus subtilis (type 2) 27
35
The backꢀup from the lead enzyme alkaline B n°2 and the protease C from Bacillus subtilis (type
1) gave poor results in the conditions tested. Protease C from Bacillus subtilis (type 2) gave
acceptable results whereas poorer than the best results obtained during the screening phases
especially regarding the conversion (33ꢀ36% vs. 43ꢀ45% with the former lead enzyme).
The influence of the pH and the enzyme loading was tested in order to evaluate the scope of
potential improvements for the resolution. A pH 8.0 was tested and the enzyme loading was
fixed between 10% w/w and 20% w/w. All other parameters being unchanged compared to the
second screening conditions. Two enzymes were tested: the best from each backꢀup strain
provided. The results are reported in the table 6.
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 31
1
2
3
4
Table 6. Influence of pH and enzyme load on the hydrolysis of 4
5
6
7
8
9
ee of (S)ꢀ
Conv. methyl
Enzyme
load
(%ꢀw)
ee of (R)ꢀ
succinic
acid 6 (%)
Coꢀ
solvent
E
value
pH Strain of the protease
(%)
ester 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
10
15
20
15
8.0 C from Bacillus subtilis (type 2)
8.0 C from Bacillus subtilis (type 2)
8.0 C from Bacillus subtilis (type 2)
8.0 Alkaline B n°2
36.6
39.5
41.2
33.8
55.7
63.0
67.8
49.6
96.6
96.4
96.9
97.0
101
104
130
107
No
10
15
20
15
8.0 C from Bacillus subtilis (type 2)
8.0 C from Bacillus subtilis (type 2)
8.0 C from Bacillus subtilis (type 2)
8.0 Alkaline B n°2
22.0
26.9
33.8
21.2
27.4
35.7
49.6
25.7
97.3
97.1
97.2
95.6
95
96
115
58
Cycloꢀ
hexane
10
15
20
15
8.0 C from Bacillus subtilis (type 2)
8.0 C from Bacillus subtilis (type 2)
8.0 C from Bacillus subtilis (type 2)
8.0 Alkaline B n°2
41.7
42.4
44.2
39.3
68.7
70.8
75.9
62.2
96.2
96.3
95.8
96.1
107
111
106
96
Dimethyl
sulfoxide
10
15
20
15
8.0 C from Bacillus subtilis (type 2)
8.0 C from Bacillus subtilis (type 2)
8.0 C from Bacillus subtilis (type 2)
8.0 Alkaline B n°2
38.2
42.0
43.2
38.0
60.1
69.8
73.3
59.0
97.0
96.4
96.2
96.3
123
114
114
95
Acetone
MTBE
10
15
20
15
8.0 C from Bacillus subtilis (type 2)
8.0 C from Bacillus subtilis (type 2)
8.0 C from Bacillus subtilis (type 2)
8.0 Alkaline B n°2
28.1
32.0
34.0
26.9
38.1
45.9
50.0
35.4
97.8
97.6
97.3
96.3
131
128
119
74
A positive effect of the pH on the conversion was observed for both enzymes, maintaining
similar ee (96ꢀ97%) for the product 6. This was particularly true for the alkaline B n°2 with a
conversion improvement from 9ꢀ26% at pH 7.5 (table 5) to 21ꢀ38% at pH 8.0 (Table 6).
Nevertheless, the best enzyme remained the protease C from Bacillus subtilis (type 2) either
without coꢀsolvent or with dimethyl sulfoxide and acetone as coꢀsolvent. A 10ꢀ15% w/w of
enzyme loading might appeared as an acceptable compromise between conversion and enzyme
price at this stage of the study. The new lead enzyme was supplied at kg scale without any issue
during the development of the project9.
ACS Paragon Plus Environment
12

Page 13 of 31
Organic Process Research & Development
1
2
3
4
Scale-up preparation for the enzymatic resolution
5
6
7
8
9
The switch from the phosphate buffer, used in the screenings, to aqueous media with pH control
in order to limit the overall volume of reaction for industrial perspectives was identified as a key
factor to ensure the successful enzymatic scaleꢀup from the screening conditions to standard
agitated vessels.
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 phosphate buffer solution (pH 7.5 at 0.1 M) was replaced by water with a pH stat Metrohm®
Titrando 902 using a NaOH solution in order to control the pH and to limit the overall dilution.
Controlling the pH of the reaction mixture during the resolution by adding a NaOH solution
might lead to the ester hydrolysis as a sideꢀreaction in case of pH overshoot. This was confirmed
by a stress test with the substrate 4 in NaOH 1M at 30 °C overnight. As expected, a significant
amount of chemical hydrolysis was observed (ca. 12%). Hence, it was crucial to perform the
reaction and the scaleꢀup in agitated vessels with the appropriate agitation design and the
appropriate positioning of the pH probe and the dosing pipe (NaOH).
The setꢀup of the resolution at laboratory scale was done in a 250 mL doubleꢀjacketed reactor
equipped with a multistage mechanical stirrer and the pH stat Metrohm® Titrando 902. Some
preliminary blank tests were run by simulating the bioꢀresolution to determine the appropriate i)
dilution of the NaOH solution, ii) positioning of probes/pipes and iii) agitation conditions in
order to achieve an smooth pH control10.
A first bioꢀcatalytic resolution in the 250 mL reactor set was run in the following conditions:
30 g of the methyl ester substrate 4, 10% w/w of enzyme (protease C from Bacillus subtilis, type
2), 6 mL/g of water, no coꢀsolvent, pH 8.0, 30 °C and a 2 M NaOH solution to control the pH
along the resolution. An inꢀprocess control after 18.5 hours showed 43% conversion and 97% ee.
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 31
1
2
3
4
5
6
7
8
9
The volume of 2 M NaOH required to control the pH during the resolution was about 1 mL/g.
The workꢀup developed was applied and the (R)ꢀsuccinic acid derivative 6 was isolated in 33%
yield, 95% purity and 97% ee. The modest isolated yield was explained by the number of
manipulations in order to recover the final product in this first attempt.
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 reaction was repeated – same scale and conditions – to evaluate the robustness of the
resolution and to assess a more representative yield. The inꢀprocess control after 20 hours
showed again 43% conversion and 97% ee. After workꢀup, the (R)ꢀsuccinic acid derivative 6 was
isolated in 42% yield, 95% purity and 97% ee. These results confirmed the bioꢀresolution route
as a valuable stereoꢀcontrolling step.
When the resolution was conducted at pH 8.5, the yield (42%) and enantiomeric excess (97%)
were reproducible in a shorter time without any chemical hydrolysis observed. The usual
conversion (ca. 43%) was obtained after ca. 17 hours at pH 8.5 compared to ca. 20 hours at pH
8.0. This result was useful in the perspective of the definition of an acceptable pH range to
ensure robustness of the process.
The resolution was also tested with 15% w/w of enzyme loading. As expected, the initial rate
was higher compared to 10% w/w. But since the kinetic trend is asymptotic, the increase of the
average rate was more modest. These observations were confirmed, at the same reaction time, by
a yield slightly higher (<1%). Furthermore, an increase of enzyme loading would likely impact
the workꢀup with an expected filtration time increase due to the enzyme denaturation observed at
pH 1. Hence, the enzyme charge was kept at 10% w/w.
ACS Paragon Plus Environment
14

Page 15 of 31
Organic Process Research & Development
1
2
3
4
Salt resolution approach
5
6
7
8
9
The evaluation of the salt resolution of (rac)ꢀ2ꢀpropylsuccinic acid 4ꢀtertꢀbutyl ester 5 was
carried out at 100 mg scale in 5 mL/g of solvent11 with 1 molar equivalent of chiral amine. 15
chiral amines have been screened12. The mixtures were agitated in 2.5 mL vials on a shaking
plate at room temperature for ca. 24 hours. The salt screening allowed to identify 3 potential hits:
(S)ꢀphenylethylamine, Lꢀphenylalaninol and Lꢀalaninol (Table 7).
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 7. Screening results for the hits identified
EtOH
EtOAc
d.e. (%)
27.8
THF
d.e. (%)
23.0
MIBK
d.e. (%)
0.6
d.e. (%)
Solid
Mother liquor
Solid
No
(S)-phenylethylamine
L-alaninol
solid
ꢀ49.6
0.2
14.2
ꢀ52.2
ꢀ1.2
ꢀ0.2
No
solid
Mother liquor
Solid
33.4
2.4
34.6
ꢀ34.2
56.4
ꢀ24.8
67.8
No
solid
No
solid
L-phenylalaninol
Mother liquor
(S)ꢀphenylethylamine was rapidly confirmed as the most promising resolution agent in ethyl
acetate. Additionally, (S)ꢀphenylethylamine is widely available and cheap13. LꢀPhenylalaninol
was discarded because the production of the desired (R)ꢀsuccinic acid salt would have required
the unnatural Dꢀphenylalaninol isomer that was anticipated to be less available and more
expensive14. Lꢀalaninol was also eliminated for the same reason and a poor isolated yield (14%).
After preliminary development, a salt resolution test at 5 g scale gave the desired (R)ꢀ2ꢀ
propylsuccinic acid 4ꢀtertꢀbutyl ester.(S)ꢀPEA salt 9 in 43% yield and 64% de. The
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 31
1
2
3
4
5
6
recrystallization in ethyl acetate afforded the salt 9 in 61% yield and 95.2% de. The overall
resolution and recrystallization sequence gave the salt 9 in 26% yield and 95.2% de. (scheme 3).
7
8
9
Scheme 3. Salt resolution of (rac)ꢀ2ꢀpropylsuccinic acid 4ꢀtertꢀbutyl ester 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
EtOAc
EtOAc
43%
64% d.e.
61%
95.2%
9
5
9
Choice of the stereo-control approach
After preliminary parallel development, the enzymatic resolution appeared to us significantly
more attractive than the salt resolution. Indeed, the enzymatic approach offered the advantage of
using directly the methyl ester 4 without any additional hydrolysis step to access the salt
resolution substrate 5. Moreover, the potential of the enzymatic resolution both in term of chiral
purity (97% ee vs. 95.2% de) and isolated yield (42% vs. 26%) was better at this stage of
development. With these results in hands, it was decided to further progress with the enzymatic
resolution of the methyl ester 4 as the key stereoꢀcontrolling step.
Nevertheless, as (R)ꢀ2ꢀpropylsuccinic acid 4ꢀtertꢀbutyl ester 6 produced via bioꢀcatalysis was
isolated as an oil with the potential carryꢀover of all organic impurities originating from the
previous chemical steps, we decided to keep this salt resolution as a backꢀup plan for chemical
purity control15. The salt resolution might also be considered as a reworking option of 6 for
ACS Paragon Plus Environment
16

Page 17 of 31
Organic Process Research & Development
1
2
3
4
5
6
chiral upgrade in case of issue with the enzymatic resolution and also as a stable solid salt
intermediate for the control of the impurity profile if required later in the development.
7
8
9
Downstream chemistry towards Brivaracetam
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 transformations towards Brivaracetam are described in the scheme 4. The (R)ꢀsuccinic acid
derivative 6 was selectively reduced into the corresponding alcohol 10 then telescoped with the
addition of TFA to catalyze the cyclization into (R)ꢀ3ꢀnꢀpropyllactone 11.
Scheme 4. Transformation of 6 into Brivaracetam 1
1- ClCO
Et N (1.2 eq),
toluene (10 mL/g), 0°C
2
Et (1.1 eq),
O
O
3
TFA (0.3 eq)
25 °C
HO
HO
MeO
O
O
O
+
O
O
O
O
100%
91%
95.9% ee
2- NaBH (2 eq),
4
O
MeOH (2 mL/g), -20 °C
3- MeOH
11
10
6
12
HBr 33% w/w
in AcOH (4 eq)
80 °C, 2 h
1- TBAI (0.3 eq), 2 (1.5 eq)
Na CO (2 eq),
IPAC (5 mL/g), reflux, 20 h
HCl (0.25 eq),
EtOH (4 mL/g),
40 °C, o/n
2
3
O
Br
O
Br
O
EtO
N
HO
2- MTBE (10 mL/g)
3- IPAC (1 mL/g)
77%
92%
93.2%
NH2
98% 1H NMR
14
13
30-40%
O
1
The control of the reduction was key to ensure a proper quality of the intermediate 11 in a safety
environment. Indeed, a sideꢀreaction occurred above ꢀ20 °C with the addition of methanol onto
the activated chloroformate derivative to generate the methyl ester impurity 12 – typically 2ꢀ4%.
From a safety perspective, the use of a slight excess of triethylamine (1.2 eq vs. 1.1 eq) avoided
the presence of acidic methanol in contact with sodium borohydride16. Furthermore, the use of
powdered sodium borohydride versus a granular form ensured that the reduction started at the
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 31
1
2
3
4
5
6
7
8
9
beginning of the methanol addition without any accumulation of reducing agent. The evolution
of hydrogen gas during the reduction was accompanied with foaming, both controlled with the
rate of addition of methanol.
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 transformation of 11 into ethyl (R)ꢀγꢀbromoꢀ3ꢀpropylbutyrate 14 was initially envisaged
through a oneꢀpot reaction using HBr gas in ethanol17. Nevertheless, due to a supply shortage of
HBr gas lecture bottle at the time of the study, the transformation was performed in two steps: a
ring opening with HBr in acetic acid to give (R)ꢀγꢀbromoꢀ3ꢀpropylbutyric acid 13 followed by a
standard esterification. The major impurity from the ring opening reaction was the unreacted
lactone 11, typically 2%18. The purity of the bromoester 14 by HPLC revealed that the amount of
residual lactone 11 was finally around 5ꢀ6%.
The condensation of 14 with (S)ꢀ2ꢀaminobutanamide 2 was catalyzed with tetrabutylammonium
iodide in refluxing isopropyl acetate. The reaction proceeded in 74ꢀ78% w/w in situ yield despite
of a nearly quantitative consumption of the substrate 14 (ca. 96%). Some substrate or
intermediate degradation was highly suspected but no root cause was clearly identified. The
reaction workꢀup was also troublesome to get rid of all the salts engaged due to the high water
solubility of Brivaracetam19. In the laboratory, a preꢀtreatment of the crude product by
suspension in methyl tertꢀbutyl ether was helpful to proceed with the final crystallization in
isopropyl acetate to isolate the API 1 within the specifications in 30ꢀ40% overall step yield.
Kilo-scale demonstration
The demonstration was performed starting from 15 kg of dimethyl nꢀpropylmalonate 15 (scheme
5). The two steps towards the racemic substrate 4 were scaled up without any issue to supply 17
kg of 4 in 4 and 6 batches respectively and 94% overall yield.
ACS Paragon Plus Environment
18

Page 19 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
The key enzymatic resolution of 4 was scaled up in a 10 L vessel without technical nor chemical
issue affording the same results than the ones obtained in the laboratory (42% yield and 97% ee),
Figure 1. Overall, 6.2 kg of the required enantiomer 6 were produced in 19 batches20. This
transformation exhibited high robustness in terms of pH (8.0, 8.5), concentration (4.5 vol,
6.0 vol) and agitation type (twoꢀstage stirrer, anchor).
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. Final conditions towards Brivaracetam 1
NaHMDS 40% w/w
in THF (1.05 eq),
THF (2 L/kg)
O
2
LiBr (1 eq), H O ( 2eq)
DMF (5 L/kg)
140 °C, 5 h
O
tert
-butyl-2-bromoacetate
MeO
O
O
(1.05 eq)
O
MeO
MeO
MeO
MeO
O
O
94%, 89.8% HPLC,
95.9% GC
-5 °C to RT
100%, 91.8% HPLC,
95.5% GC
O
O
4
16
15
O
2
1- ClCO Et (1.1 eq),
3
Enzyme (10% w/w)
Et N (1.2 eq),
toluene (10 L/kg), -5 °
C
TFA (0.3 eq)
25 °C
2
H O (4.5 L/kg), 30 °C
HO
HO
NaOH 2 M (1 L/kg)
O
O
O
O
O
O
100%
93.2% GC
95.7% ee
4
2- NaBH (2 eq),
42%, 97.1% ee
92.4% HPLC,
95.1% GC
MeOH (2 L/kg), -20 °C
3- HCl 4 M (2 L/kg)
11
10
6
1- TBAI (0.3 eq), (1.5 eq)
2
HBr 33% w/w
in AcOH (4 eq)
80 °C, 2 h 30
2
3
HCl (0.25 eq),
EtOH (4 L/kg),
40 °C, o/n
Na CO (2 eq),
IPAC (5 L/kg), reflux, 28 h
2- MTBE (10 L/kg)
O
O
Br
Br
O
N
HO
EtO
3- IPAC (1 L/kg)
89%
93.3% HPLC
93.0% GC
81%
NH2
>95% 1H NMR
32%, 97.1% w/w
95.9% ee
13
14
O
1
Figure 1. (a) bioꢀresolution in a 10 L vessel with pHꢀstat control equipment. (b) pH monitoring
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 31
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 reductionꢀdeprotection transformation of 6 into the propyllactone 11 (5 batches, 2.9 kg,
nearly quantitative yield, 91% purity, 95.9% ee21) and the twoꢀstep sequence into the bromoester
14 (2 x 3 batches, 3.04 kg, 70% overall yield, 92.5% purity) were achieved without any technical
issue at the scale studied.
As expected, the final condensation remained the less optimal step from the demonstration with a
32% overall step yield. The condensation was done in 3 batches (2.47 kg, 99% yield, 74% w/w,
85.4%, 1.94% ucb 34713), the preꢀtreatment in methyl tertꢀbutyl ether was performed in 2
batches but finally appeared as useless with respect to the purity (2.06 kg, 82% yield, 64% w/w,
86.9%, 2.52% ucb 34713). A single batch crystallization afforded the API 1 (0.5 kg, 39% yield)
with a purity of 97.1% w/w and 2.03% ucb 34713.
The assessment of the potential genotoxic compounds and impurities (GTIs) associated to the
synthetic route unsurprisingly showed alerting structures for the tertꢀbutyl 2ꢀbromoacetate and
the intermediates 13 and 14 by inꢀsilico Derek analysis22. The GLP Ames testing showed that the
compounds tertꢀbutyl 2ꢀbromoacetate and 13 were not mutagenic23.
ACS Paragon Plus Environment
20

Page 21 of 31
Organic Process Research & Development
1
2
3
4
Conclusion
5
6
7
8
9
The control of the stereochemistry of the 4ꢀnꢀpropyl substituent borne by the novel AED
Brivaracetam 1 was efficiently achieved by the bioꢀcatalytic resolution of (rac)ꢀmethyl 2ꢀ
propylsuccinate 4ꢀtertꢀbutyl ester 4. The selection of the best substrate and the protease C from
Bacillus subtilis (type 2) was carried out from our inꢀhouse screening platform. The
transformation was then optimized in 250 mL laboratory scale reactors with a final
demonstration at 10 L scale showing reproducible and consistent results all along the
development. The chiral intermediate (R)ꢀ2ꢀpropylsuccinic acid 4ꢀtertꢀbutyl ester 6 was obtained
in 42% yield and 97% ee. A fitꢀtoꢀpurpose development of upstream and downstream chemistry
towards Brivaracetam 1 allowed to produce the API within the required specifications.
Furthermore, the racemization24 of the unwanted enantiomer 8 was demonstrated offering a
recycling opportunity to maximize the overall yield of the resolution. The selection of the bioꢀ
catalytic route over the salt resolution one also gave the chance to identify an important impurity
control point with the isolation of a (S)ꢀphenylethylamine salt of 6. The process mass intensity
(PMI) calculated with the kiloꢀscale demonstration data was 1428 kg/kg API hampered by the
weak final step. Indeed, the workꢀup, the methyl tertꢀbutyl ether treatment and the isolation of
the final step represented 68% of the overall PMI. A recalculated metric taken account of the
racemization and recycling of 8 and without the useless methyl tertꢀbutyl ether treatment gave a
PMI estimation of 785 kg/kg API showing there was still room for improvement. Finally, an
optimization of the final step minimizing the suspected substrate degradation and improving the
isolation of the API would be necessary to consider this bioꢀcatalytic route to Brivaracetam 1 as
a viable commercial 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
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 31
1
2
3
4
Experimental section
5
6
7
8
9
Analytical methods for the determination of chemical purity by HPLC and GC: HPLC
analysis were run on a Water Alliance 2695 apparatus, column Xbridge C18 (3.5 µm, 4.6 x
50mm), 35 °C, H2O with H3PO4 1%/acetonitrile from 90/10 to 10/90 v/v in 5 min, 3 mL/min,
UV detection at 210 nm. GC analysis were run on an Agilent 6890, column VFꢀ5ms (15 m x
0.25 mm x 1µm), 25 min run (50 °C/5 min, 50 °C to 280 °C/10 min, 280 °C/10 min), helium,
FID detection at 300 °C. The chemical purity of 1 was monitored on HPLC Waters Alliance
2695, column Ascentis Express C18 (2.7 ꢁm, 100 x 4.6 mm), 40 °C, (A) water/phosphoric acid
1000/1 (v/v) and (B) acetonitrile gradient (t = 0 min 95/5 A/B, t = 25 min 64/36 A/B, t = 26 min
95/5 A/B, t = 28 min 95/5 A/B), 1.5 mL/min, UV detection at 205 nm.
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
Analytical methods for the determination of chiral purity: the chiral purity of 6 (enzymatic
and salt resolutions) was monitored on an HPLC Water Alliance 2695 apparatus, column ASꢀH
(5 µm, 4.6 x 250mm), 25 °C, IPA 3%/nꢀheptane 97%/TFA 0.01% v/v, 1 mL/min, UV detection
at 210 nm. The chiral purity of 20 was monitored on a GC Agilent 6890, column Macherey –
Nagel Lipodex E 25 m x 250 µm x 1 µm, isotherm 130 °C, helium (1 mL/min), FID detection at
220 °C. The chiral purity of 1 was monitored on HPLC, column Chiralpak AD (10 ꢁm, 4.6 x 250
mm), 20 °C, 80/20 (v/v) nꢀhexane/ethanol, 1 mL/min, UV detection at 205 nm.
NMR spectra were recorded on a Bruker Avance 400 MHz instrument. Elemental analyses
were performed on an Elementar Vario EL3 apparatus.
ACS Paragon Plus Environment
22

Page 23 of 31
Organic Process Research & Development
1
2
3
4
(R)ꢀ2ꢀpropylsuccinic acid 4ꢀtertꢀbutyl ester.(S)ꢀphenylethylamine salt 9:
5
6
7
8
9
A 100 mL bottom flask equipped with refrigerant and mechanical agitation is charged with (rac)ꢀ
2ꢀpropylsuccinic acid 4ꢀtertꢀbutyl ester 5 (5 g, 23.12 mmol, 1.0 eq), ethyl acetate (30 mL, 6 vol)
and (S)ꢀphenylethylamine (2.80 g, 23.12 mmol, 1.0 eq) at room temperature. The salt precipitates
within 10 min. The suspension is heated to gentle reflux (80ꢀ85°C, oil bath) until the salt has
entirely dissolved. The reflux is maintained for 20 min then cooled to room temperature in ca.
1 h 30 min. The salt precipitates at room temperature and is agitated an extra 1 h 30 min at room
temperature. The suspension is then cooled to 0ꢀ5 °C and filtered. The cake is washed with cold
ethyl acetate (10 mL) and dried under vacuum oven at 40 °C to give 9 (3.35 g, 9.94 mmol, 43%)
with a chiral purity of 64% de.
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 100 mL bottom flask equipped with refrigerant and mechanical agitation is charged with 9
(3.28 g, 9.7 mmol) and ethyl acetate (33 mL) at room temperature. The suspension is heated to
gentle reflux (85ꢀ90 °C, oil bath) until the salt has entirely dissolved. The reflux is maintained
for 20 min then slowly cooled to room temperature in ca. 2 h 30 min. The salt precipitates at ca.
30 °C. The suspension is agitated at room temperature for an extra 1 h 30 min then cooled to 0ꢀ5
°C and filtered. The salt is dried under vacuum oven at 40 °C to give 9 (2.01 g, 61%) with a
chiral purity of 95.2% de.
O1ꢀTertꢀbutyl O2,O2ꢀdimethyl pentaneꢀ1,2,2ꢀtricarboxylate 16:
A 60 L double jacketed reactor equipped with a mechanical stirrer and a condenser is charged
with dimethyl nꢀpropylmalonate 15 (3000 g, 17.22 mol, 1.0 eq) and THF (6 L, 2 L/kg) under a
nitrogen atmosphere. The solution is cooled to ꢀ5.0 °C. NaHMDS 40.9% w/w solution in THF
(8140 g, 18.08 mol, 1.05 eq) is added maintaining the temperature between ꢀ2.8 °C and ꢀ0.3 °C
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 31
1
2
3
4
5
6
7
8
9
over a period of 1 h 10 min. The addition funnel is washed with THF (0.3 L, 0.1 L/kg). The
solution is agitated at ꢀ5 °C for 1 hour. Tertꢀbutyl 2ꢀbromoacetate (3529 g, 18.096 mol, 1.05 eq)
is added maintaining the temperature between ꢀ8.0 °C and ꢀ0.3 °C over a period of 1 hour. The
addition funnel is washed with THF (0.3 L, 0.1 L/kg). The reaction mixture is warmed to 17 °C
in 1 h 15 min and agitated at 20 °C for an additional 1 hour. Water (3 L, 1 L/kg), saturated
ammonium chloride solution (3 L, 1 L/kg) and ethyl acetate (9 L, 3 L/kg) are successively added
at 20 °C. After decantation the aqueous layer is separated and discarded. The organic layer is
washed with 50% saturated ammonium chloride solution (2 x 6 L, 2 x 2 L/kg). After decantation
and phase separation, the organic layer is evaporated to give 13 as an oil (5154 g, 17.87 mol,
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
104%) with a purity of 95.9% GC and 91.8% HPLC. H NMR (CDCl3, 400 MHz) 3.67 (s, 6 H)
2.82 (s, 2 H) 1.90ꢀ1.86 (m, 2 H) 1.36 (s, 9 H) 1.23ꢀ1.12 (m, 2 H) 0.86 (t, J = 7.3 Hz, 3 H) ppm;
13C NMR (CDCl3, 100 MHz) 171.0, 169.4, 81.1, 55.7, 52.5, 38.9, 35.5, 27.9, 17.7, 14.2 ppm.
Anal. Calcd for C14H24O6: C, 58.32; H, 8.39. Found: C, 57.75; H, 8.49.
(rac)ꢀMethyl 2ꢀpropylsuccinate 4ꢀtertꢀbutyl ester 4:
A 60 L double jacketed reactor equipped with a mechanical stirrer and a condenser is charged
with lithium bromide (985 g, 11.34 mol, 1.0 eq), intermediate 16 (3272 g, 11.35 mol, 1.0 eq),
water (0.409 L, 22.72 mol, 2.0 eq) and DMF (16.4 L, 5 L/kg) under a nitrogen atmosphere. The
reaction mixture is heated to 139 °C for 5 hours. The reaction mixture is cooled to 20 °C in
1 h 05 min. Diisopropyl ether (13.2 L, 4 L/kg), water (6.6 L, 2 L/kg) and saturated ammonium
chloride solution (6.6 L, 2 L/kg) are added. After decantation, an additional portion of saturated
ammonium chloride solution (2.5 L, 0.76 L/kg) is added in order to have a clear phase
separation. After phase separation, the organic layer is washed with water (2 x 10 L, 2 x
3.05 L/kg) then evaporated to give 4 as a crude oil (2467 g, 10.71 mol, 94%) with a purity of
ACS Paragon Plus Environment
24

Page 25 of 31
Organic Process Research & Development
1
2
95.9% GC and 89.8% HPLC. 1H NMR (CDCl3, 400 MHz) 3.54 (s, 3 H) 2.70ꢀ2.60 (m, 1 H) 2.46
3
4
5
6
(dd, J = 16.3, 9.4 Hz, 1 H) 2.20 (dd, J = 16.3, 5.2 Hz, 1 H) 1.52ꢀ1.40 (m, 1 H) 1.35ꢀ1.25 (m, 1 H)
7
8
9
13
1.28 (s, 9 H) 1.22ꢀ1.10 (m, 2 H) 0.76 (t, J = 7.3 Hz, 3 H) ppm; C NMR (CDCl3, 100 MHz)
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
175.6, 171.1, 80.6, 51.5, 41.2, 37.4, 34.0, 28.0, 20.2, 13.8 ppm. Anal. Calcd for C12H22O4: C,
62.58; H, 9.63. Found: C, 61.64; H, 9.53.
(R)ꢀ2ꢀpropylsuccinic acid 4ꢀtertꢀbutyl ester 6:
A 10 L double jacketed reactor equipped with a mechanical stirrer and a condenser is charged
with protease C from Bacillus subtilis type 2 (100 g, 10% w/w) and water (4.5 L, 4.5 L/kg).
NaOH 0.1 M (30 mL, 0.03 L/kg) is added so as to reach an initial pH of 8.55. Then the racemic
substrate 4 (1000 g, 4.34 g, 1.0 eq) is added one pot. The reaction mixture is stirred (200 rpm) at
30 °C maintaining the internal pH at 8.55 with a 2 M NaOH solution (pHꢀstat) for 18.7 hours
(NB: 930 mL, 0.93 L/kg of 2 M NaOH is added overall). The unreacted ester is extracted with
cyclohexane (2 L, 2 L/kg) at pH 9 (0.1 M NaOH, 60 mL, 0.06 L/kg). The pH is then adjusted to
2 with 3 M HCl (640 mL, 0.64 L/kg) and isopropyl acetate (3 L, 3 L/kg) is added to extract the
product. Celite (167 g, 16.7% w/w) is added into the reactor to facilitate the filtration of the
slurry (NB: removal of a part of the denatured enzyme). The cake is washed with isopropyl
acetate (0.5 L, 0.5 L/kg). The layers are separated and the organic phase (emulsion) is filtered
through the pad of Celite previously obtained. The cake is washed with isopropyl acetate (5 L,
0.5 L/kg). The final organic layer is evaporated (45 °C, 3 mbar) to give 6 (394.4 g, 1.82 mol,
1
42.0%) with a purity of 95.1% GC, 92.4% HPLC and 97.1% ee. H NMR (CDCl3, 400 MHz)
2.88ꢀ2.78 (m, 1 H) 2.63 (dd, J = 16.4, 9.3 Hz, 1 H) 2.39 (dd, J = 16.4, 5.2 Hz, 1 H) 1.72ꢀ1.60 (m,
13
1 H), 1.55ꢀ1.48 (m, 1 H) 1.45 (s, 9 H) 1.45ꢀ1.35 (m, 2 H) 0.94 (t, J = 7.3 Hz, 3 H) ppm ; C
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 31
1
2
3
4
5
6
NMR (CDCl3, 100 MHz) 181.5, 171.1, 80.9, 41.2, 37.1, 33.8, 28.0, 20.1, 13.8 ppm; HRMS
calculated for [MꢀH] C11H19O4 215.1283, found 215.1328.
7
8
9
(R)ꢀ3ꢀnꢀPropyllactone 11:
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 20 L double jacketed reactor equipped with a mechanical stirrer and a condenser is charged
with 6 (1000.6 g, 4.63 mol, 1.0 eq) and toluene (8.5 L, 8.5 L/kg). The solution is cooled to ꢀ5.1
°C under nitrogen flow. Triethylamine (561.5 g, 5.55 mol, 1.2 eq) is added maintaining the
temperature between ꢀ6.0 °C and ꢀ5.1 °C over a period of 35 min. Ethyl chloroformate (551.9 g,
5.09 mol, 1.1 eq) is added maintaining the temperature between ꢀ6.6 °C and ꢀ3.4 °C over a
period of 38 min. The heterogeneous reaction mixture is stirred for 1 hour at ꢀ5 °C. Then the
reaction mixture is warmed to 20 °C and filtered to remove the triethylammonium chloride salt.
The cake is washed with toluene (1.5 L, 1.5 L/kg) at 20 °C. The filtrate is transferred into the
reactor and the solution is cooled to ꢀ18.7 °C. Under nitrogen flow, powdered sodium
borohydride (349.8 g, 9.25 mol, 2.0 eq) is added one pot at ꢀ18.5 °C. Then methanol (2 L, 2
L/kg) is transferred dropwise in order to maintain the temperature below ꢀ18 °C and to limit the
formation of foaming (08 h 40 min addition time between ꢀ21.4 °C and ꢀ19.1 °C). The reaction
mixture is stirred for minimum 1 hour at ꢀ20 °C (15 h 15 min, overnight). At ꢀ21.4 °C, under
nitrogen flow, HCl 4 M (2 L, 2 L/kg) is added dropwise in order to maintain the temperature
below ꢀ18 °C and to limit the formation of foaming (05 h 20 min addition time between ꢀ21.9 °C
and ꢀ21.5 °C). Then the reaction mixture is warmed to 20 °C. Water (5 L, 5 L/kg) is added until
dissolution of salts and the aqueous layer is discarded. An additional water wash (4 L, 4 L/kg) is
done. At 25 °C, trifluoroacetic acid 95% v/v (0.1 L, 0.1 L/kg, ~0.27 eq) is added one pot and the
reaction mixture is stirred until the completion of the reaction (~1 h). The organic layer is
washed with water (2 x 3 L, 2 x 3 L/kg) then evaporated to give 11 (608.9 g, 4.75 mol, 103%)
ACS Paragon Plus Environment
26

Page 27 of 31
Organic Process Research & Development
1
2
1
3
4
with a purity of 93.2% GC and 95.7% ee GC. H NMR (CDCl3, 400 MHz) 4.42 (tapp, J = 8.0
5
6
Hz, 1 H) 3.93 (tapp, J = 8.0 Hz, 1 H) 2.65ꢀ2.54 (m, 2 H) 2.19 (dd, J = 16.3, 7.3 Hz, 1 H) 1.48ꢀ
7
8
9
13
1.44 (m, 2 H) 1.40ꢀ1.30 (m, 2 H) 0.95 (t, J = 7.1 Hz, 3 H) ppm ; C NMR (CDCl3, 100 MHz)
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
177.3, 73.4, 35.4, 35.2, 34.5, 20.5, 13.9 ppm; HRMS calculated for [M+H] C7H13O2 129.0916,
found 129.0916.
1
Non isolated intermediate 10 (tertꢀbutyl 3ꢀhydroxymethyl hexanoate): H NMR (CDCl3, 400
MHz) 4.42 (dd, J = 9.0, 7.4 Hz, 1 H) 3.92 (dd, J = 9.0, 7.1 Hz, 1 H) 2.65ꢀ2.50 (m, 2 H) 2.18 (dd,
13
J = 16.4, 7.4 Hz, 1 H) 1.50ꢀ1.44 (m, 2 H) 1.40ꢀ1.30 (m, 2 H) 0.94 (t, J = 7.2, 3 H) ppm; C
NMR (CDCl3, 100 MHz) 177.4, 73.4, 69.1, 35.4, 35.2, 34.5, 31.1, 20.5, 13.9 ppm. Anal. Calcd
for C11H22O3: C, 65.31; H, 10.96. Found: C, 62.65; H, 10.44.
(R)ꢀγꢀbromoꢀ3ꢀpropylbutyric acid 13:
A 10 L double jacketed reactor equipped with a mechanical stirrer, a condenser and a sodium
hydroxide scrubber is charged under a nitrogen flow with HBr 33% w/w in acetic acid (4.4 L,
25.49 mol, 4 eq). The solution is cooled to 5.9 °C and a solution of 11 (816.7 g, 6.37 mol, 1.0 eq)
in acetic acid (0.4 L, 0.5 L/kg) is added slowly maintaining the temperature at ca. 5 °C over a
period of 40 min. The solution is warmed to room temperature and then heated to 80 °C for
2 h 30 min. The reaction mixture is cooled to 20 °C, diluted with water (4.1 L, 5 L/kg) and
extracted with dichloromethane (4.1 L, 5 L/kg then 2 L, 2.5 L/kg). The combined organic phases
are washed with water (3 x 2.4 L, 3 x 3 L/kg), dried over anhydrous sodium sulfate and
evaporated to 13 as a pale yellow oil (1075.9 g, 5.15 mol, 81%) with a purity ≥ 95% by 1H
NMR. 1H NMR (CDCl3, 400 MHz) 11.0ꢀ10.0 (s, 1 H) 3.60 (dd, J = 10.3, 4.1 Hz, 1 H) 3.51 (dd, J
= 10.3, 5.2 Hz, 1 H) 2.59 (dd, J = 16.4, 7.3 Hz, 1 H) 2.44 (dd, J = 16.4, 5.9 Hz, 1 H) 2.28ꢀ2.18
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 31
1
2
13
3
4
(m, 1 H) 1.50ꢀ1.30 (m, 4 H) 0.96 (t, J = 7.0) ppm ; C NMR (CDCl3, 100 MHz) 179.2, 38.2,
5
6
7
8
37.4, 36.0, 34.7, 19.7, 13.9 ppm; HRMS calculated for [M+H] C7H14O2Br 209.0177, found
209.0182.
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
Ethyl (R)ꢀγꢀbromoꢀ3ꢀpropylbutyrate 14:
A 10 L double jacketed reactor equipped with a mechanical stirrer and a condenser is charged
under a nitrogen atmosphere with 13 (1014.7 g, 4.85 mol, 1.0 eq) and ethanol (4.04 L, 4 L/kg) at
20 °C. HCl 37% (120 g, 100 mL, 1.21 mol, 0.25 eq) is added and the reaction mixture is heated
to 40.6 °C for 19 h 05 min. Ethanol is removed by evaporation and the residue is diluted in ethyl
acetate (4.04 L, 4 L/kg). The solution is successively washed with aqueous sodium hydroxide
(97 g, 2.42 mol, 0.5 eq in 1.52 L, 1.5 L/kg of water) and water (1.52 L, 1.5 L/kg). The organic
phase is then evaporated to give 14 as a pale yellow oil (1025 g, 4.32 mol, 89%) with a purity of
1
93.0% GC and 93.3% HPLC. H NMR (CDCl3, 400 MHz) 4.16 (q, J = 7.2, 2 H) 3.57 (dd, J =
10.2, 4.2 Hz, 1 H) 3.51 (dd, J = 10.2, 5.1 Hz, 1 H) 2.50 (dd, J = 16.0, 7.3 Hz, 1 H) 2.35 (dd, J =
16.0, 6.0 Hz, 1 H) 2.25ꢀ2.15 (m, 1 H) 1.50ꢀ1.30 (m, 4 H) 1.28 (t, J = 7.2, 3 H) 0.94 (t, J = 7.1, 3
H) ppm ; 13C NMR (CDCl3, 100 MHz) 172.3, 60.4, 38.5, 37.6, 36.2, 34.7, 19.7, 14.2, 14.0 ppm;
HRMS calculated for [M+H] C9H18O2Br 237.0490, found 237.0493.
(2S)ꢀ2ꢀ[(4R)ꢀ2ꢀoxoꢀ4ꢀpropyltetrahydroꢀ1Hꢀpyrrolꢀ1ꢀyl]butanamide 1:
A 10 L double jacketed reactor equipped with a mechanical stirrer and a condenser is charged
under a nitrogen atmosphere with 14 (1041.8 g, 4.39 mol, 1.0 eq), (S)ꢀ2ꢀaminobutanamide 2
(673.3 g, 6.60 mol, 1.5 eq), tetrabutylammonium iodide (487 g, 1.32 mol, 0.3 eq), sodium
carbonate (932 g, 8.79 mol, 2.0 eq) and isopropyl acetate (5.2 L, 5 L/kg). The suspension is
heated to reflux (Tmass = 88.8 °C) for 28 hours then cooled to 15.7 °C. The suspension is filtered
ACS Paragon Plus Environment
28

Page 29 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
at 15 °C and the cake is washed with isopropyl acetate (1 L, 1 L/kg). The filtrate is charged into
the reactor, diluted to ca. 9ꢀ10 vol overall with isopropyl acetate (+4 L, +4 L/kg) and then heated
to 60.3 °C. Acetic acid (197.9 g, 3.30 mol, 0.75 eq) is added slowly over a period of 40 min then
the reaction mixture is agitated at 60 °C for 1 h 30 min. The suspension is cooled to 25.1 °C and
filtered. The cake is washed with isopropyl acetate (1 L, 1 L/kg). Water (1 L, 1 L/kg) is added
and sodium bicarbonate (185 g) is added until pH 7 is reached in the aqueous layer then the
phases are separated. An additional wash with water (1 L, 1 L/kg) is done. The organic phase is
evaporated to give crude product (892 g, 95.7%) as a yellow oily solid.
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 blended crude product (1560 g) is partially dissolved in methyl tertꢀbutylether (15.6 L) into a
20 L reactor and agitated at 20 °C for 22 h 30 under a nitrogen atmosphere. The insoluble solid is
eliminated by filtration and the filtrate is evaporated to a crude 1 (1283 g, 82%) as a yellow oily
solid. After complete dissolution in isopropyl acetate (500 mL), the mass temperature is set at
36 °C and seeded with 1 (8 g). After 30 min aging time at the seeding temperature, the
suspension is cooled down to 23 °C in 2 hours then to 15 °C in 1 hour with a post stirring time of
30 min at 15 °C. The product is isolated by filtration. The cake is washed with isopropyl acetate
(350 mL and 150 mL). The product is dried 40 °C under vacuum to afford Brivaracetam 1
1
(500 g, 32% overall step yield) with a purity 97.1% w/w assay and 2.03% of ucb 34713. H
NMR (CDCl3, 400 MHz) 6.51 (s, 1 H) 5.93 (s, 1 H) 4.46 (dd, J = 8.9, 7.9, 1 H) 3.47 (dd, J = 9.8,
7.8 Hz, 1 H) 3.05 (dd, J = 9.8, 7.1 Hz, 1 H) 2.54 (dd, J = 16.7, 8.6, 1 H) 2.39ꢀ2.23 (m, 1 H) 2.06
(dd, J = 16.7, 8.1, 1 H) 1.99ꢀ1.85 (m, 1 H) 1.70ꢀ1.62 (m, 1 H) 1.45ꢀ1.37 (m, 2 H) 1.37ꢀ1.25 (m, 2
13
H) 0.94ꢀ0.84 (m, 6 H) ppm ; C NMR (CDCl3, 100 MHz) 175.4, 172.6, 55.7, 49.4, 37.8, 36.4,
31.9, 21.2, 20.5, 13.9, 10.4 ppm; HRMS calculated for [M+H] C11H21O2N2 213.1603, found
213.1607.
ACS Paragon Plus Environment
29