Process development and optimization for production of a potassium ion channel blocker, ICA-17043

Article abstract of DOI:10.1021/op3000916

A scalable process for the manufacture of a potassium ion channel blocker was developed and optimized. Key features of the process include an optimized Grignard reaction, a direct cyanation of the intermediate trityl alcohol derivative, and an improved nitrile hydrolysis protocol, relative to the original acidic hydrolysis conditions, to generate the crude active pharmaceutical ingredient (API) with >95% HPLC purity. The Grignard and the cyanation reactions could be telescoped, resulting in an improved throughput compared to the original four-step process. An effective recrystallization of the API was also developed and the process scaled up to manufacture multiple batches at the pilot scale.

Full text of DOI:10.1021/op3000916

Subscriber access provided by FONDREN LIBRARY, RICE UNIVERSITY
Full Paper
Process Development and Optimization for Production
of the Potassium Ion Channel Blocker ICA-17043
Sripathy Venkatraman
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/op3000916 • Publication Date (Web): 19 Jul 2012
Downloaded from http://pubs.acs.org on July 20, 2012
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 28
Organic Process Research & Development
1
2
3
4
5
6
Process Development and Optimization for Production of a Potassium
Ion Channel Blocker ICA-17043
Bingidimi I. Mobele,1¥ Sripathy Venkatraman,1* Grant McNaughton-Smith,2 Cameron
Gibb,1 Luckner G. Ulysse,1 Carl A. Lindmark,1 Stephen Shaw,1 Brian Marron,2 Kerry
Spear,2 and Mark J. Suto2†
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
1AMRI, 26 Corporate Circle, P. O. Box 15098, Albany, NY 12212-5098 USA
2Icagen, Inc., Post Office Box 14487, Research Triangle Park, NC 27709, USA
Abstract
A scalable process for the manufacture of a potassium ion channel blocker was developed
and optimized. Key features of the process include an optimized Grignard reaction, a
direct cyanation of the intermediate trityl alcohol derivative and an improved nitrile
hydrolysis protocol relative to the original acidic hydrolysis conditions, to generate the
crude API with >95% HPLC purity. The Grignard and the cyanation reactions could be
telescoped, resulting in an improved throughput compared to the original four-step
process. An effective recrystallization of the API was also developed and the process
scaled up to manufacture multiple batches at the pilot scale.
TOC Graphic
O
i. PhMgBr/THF
O
KOH
NC
Toluene
H2N
t-AmOH
ii.
F
F
TMSCN, InCl3 (20 mol%)
F
F
F
F
RT - 50 oC
90 °C
1
3
ICA-17043
* Author for correspondence. E-mail: sripathy.venkatraman@amriglobal.com
¥ jpmobele@okapi-chemtech.com; OKAPI Chemtech, Inc., P. O. Box 3715, Albany, NY
12203
† Southern Research, Birmingham, Alabama
1
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 28
1
2
3
4
1. Introduction
5
6
7
8
Ion channels have emerged as attractive targets for the development of new
therapies as they are implicated in the pathogenesis of several disease states, including
vascular dysfunction, gastrointestinal dysmotility, memory disorders, neurological
disorders, epilepsy, autoimmune and inflammatory diseases, to name a few.1-4 In
particular, the modulation of the intermediate conductance calcium-activated potassium
ion channel, KCa3.1, a membrane protein found in lymphocytes, erythrocytes,
fibroblasts, intestinal and airway epithelia, proliferating vascular smooth muscle and
vascular endothelium3,5 offers avenues for potential pharmacological intervention, which
can lead to the development of much needed therapies for several diseases. ICA-17043
is a selective and potent inhibitor of KCa3.1 for which a practical and scalable synthesis
was required.
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 medicinal chemistry procedures used for producing the initial batches of the
API for toxicological and formulation evaluation (Scheme 1) suffered from several
drawbacks that precluded scale-up and required further development and optimization to
establish an efficient and robust process.
2
ACS Paragon Plus Environment

Page 3 of 28
Organic Process Research & Development
1
2
3
4
Scheme 1. Original Synthesis of ICA-17043
5
6
O
Cl
7
8
9
HO
PhMgBr
CH3COCl
DCM
F
F
Et2O / MTBE
F
F
F
F
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
2
2'
O
Cu2O
CuCN
NC
H2N
H SO
2
4
Xylenes
135 - 145 oC
HOAc
F
F
F
F
125–130 °C
3
ICA-17043
The process development and optimization efforts needed to address several issues
summarized in Table 1.
Table 1. Issues in Original Synthesis
Process Step
Issues
Grignard reaction
Cyanation reaction
High volatility and flammability of diethyl ether
• Moisture lability of trityl chloride intermediate 2’
• Requirement for an extensive azeotropic drying of copper
reagents to minimize hydrolysis of 2’ and yield losses
• High reaction temperatures
• Impractical workup and purification (decantation,
triturations and filtration of crude extracts through silica
gel to remove highly colored impurities)
Nitrile Hydrolysis
• Harsh reaction conditions
• Loss of API through further hydrolysis to the
corresponding acid and decarboxylation
• Volume inefficient workup; emulsions during extraction;
filtration of extracts through silica gel prior to isolation of
crude API
3
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 28
1
2
3
4
5
6
7
8
Accordingly, the main development objectives were the following: (1) to optimize
the Grignard reaction; (2) to evaluate a direct conversion of alcohol 2 to the penultimate
nitrile 3, thereby avoiding the moisture labile chloride 2’; (3) to evaluate and optimize
alternative hydrolysis conditions for the conversion of nitrile 3 to the API and optimize
the yield, while improving its quality. Additionally, the process could be further
optimized and streamlined by removing column chromatography, reducing the number of
solvents used and telescoping steps as appropriate.
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. Results and Discussions
2.1 Optimization of the Grignard Reaction
Tetrahydrofuran (THF), methyl tert-butyl ether (MTBE) and toluene were
evaluated as potential alternative solvents for the Grignard reaction. In all cases, the
reaction was complete within 2 h (reversed phase HPLC, % area/area, UV detection at
220 nm), affording a quantitative yield of the crude alcohol. Toluene was chosen as the
reaction solvent to facilitate azeotropic drying of the crude alcohol 2 and because it made
possible to telescope crude 2 into the next step. Phenylmagnesium bromide was chosen
over phenylmagnesium chloride as the former generally gave a cleaner reaction.6 The
reaction volume in the range of 6 to 10 volumes of toluene relative to the
difluorobenzophenone showed no appreciable difference in the reaction outcome, and so
the lowest volume was chosen for scale-up. Likewise, the order of addition of the
4
ACS Paragon Plus Environment

Page 5 of 28
Organic Process Research & Development
1
2
3
4
5
6
Grignard made no difference in the reaction time or product purity within the ranges
studied.
7
8
9
Prior to scale-up, a calorimetric screening (RC1) of both the Grignard addition
reaction and the quench was conducted. The screening indicated a moderately exothermic
Grignard reaction, with an enthalpy of reaction of 151.5 kJ/mol, corresponding to a
theoretical adiabatic temperature rise (ꢀTad) of 66.9 K. The quench of the Grignard
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
reaction was also exothermic (ꢀTad of 47.9 K). Accordingly on scale, the
difluorobenzophenone was dissolved in 12 volumes of toluene and approximately half of
the solvent was stripped to azeotropically dry the reactor contents (KF <0.1%). On
cooling to 30 °C, the Grignard reagent was added slowly, thereby allowing control of the
resulting exotherm. Once complete (HPLC), the reaction was quenched by adding
aqueous HCl over 1 – 2 h, while holding the reactor contents at 5 – 20 °C. The crude
toluene solution obtained after the phase split was washed (aqueous sodium bicarbonate,
then de-ionized water), and then azeotropically distilled to dry the crude product (3 – 4
volumes of residual solution, KF <0.1%), which was used in the cyanation step without
further purification.
2.2 Conversion of Alcohol 2 to Nitrile 3 – Development of a Direct Catalytic
Cyanation Reaction
2.2.1 Cyanation Screening Experiments
The original process for producing nitrile 3 relied on a classical procedure typically
used for the preparation of triaryl halomethanes7a-d and proceeded via the moisture labile
chloride 2’ (Scheme 1). Subsequent nucleophilic displacement of the chloride using
5
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 28
1
2
3
4
5
6
7
8
9
CuCN in the presence of cuprous oxide in refluxing xylenes afforded the nitrile in
moderate isolated yields. Inconsistent conversion profiles were observed due to
hydrolysis of chloride 2’ back to alcohol 2. Due to the anticipated difficulties in scaling
up such a process, we sought to develop an alternative process that would obviate all the
aforementioned issues. We briefly considered a Lewis acid mediated cyanation of
chloride 2’ using trimethylsilyl cyanide as a soluble cyanide source based on literature
precedence for a related structure,8 but we reasoned that a direct cyanation of the alcohol
under similar reaction conditions would be simpler and might be achievable, as it would
likely proceed through a stabilized carbenium species, affording the nitrile via SN1
nucleophilic substitution.9,10 We were encouraged by a report from Caron et al.11 in
which an aryl cyclohexanol intermediate was converted to the corresponding nitrile in
near quantitative yields using a catalytic amount (25 mol%) of tin chloride and
trimethylsilyl cyanide as the cyanide source.
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
Several Lewis acids were screened for their ability to mediate a direct cyanation of
alcohol 2 using trimethylsilyl cyanide (TMSCN) as a soluble cyanide source and
dichloromethane (DCM) or toluene as solvents (Scheme 2). The evaluated Lewis acids
were TiCl4, BF3·Et2O, InCl3, ZnCl2, Zn(CN)2 and Zn dust.
Scheme 2. Screening of Lewis Acid Mediated Direct Cyanation of Alcohol 2
TMSCN
HO
NC
Lewis Acid
DCM or Toluene
F
F
F
F
2
3
6
ACS Paragon Plus Environment

Page 7 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Gratifyingly, a clean conversion of alcohol 2 to nitrile 3 was observed in all cases, except
when zinc reagents (ZnCl2, Zn(CN)2 and Zn dust) were used as Lewis acids. In these
cases, the reaction proceeded to approximately 50% conversion using roughly one molar
equivalent of TMSCN. A single byproduct, determined to be the trimethylsilyl ether 4
resulting from a competing silylation of alcohol 2 (Scheme 3) was generated in equal
amounts (~ 50%).
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. Silylation Side Reaction using Zinc Reagents
TMSCN
NC
TMSO
HO
+
Lewis Acid
Solvent
F
F
F
F
F
F
2
3
4
Lewis Acid: ZnCl2, Zn(CN)2, Zn dust
Solvent: DCM, toluene
Nonetheless, the cyanation reaction could be pushed to complete conversion in these
cases by adding another equivalent of TMSCN. This indicates that the trimethylsilyloxy
moiety is also a suitable leaving group under these reaction conditions.
In light of these screening experiments, indium trichloride proved to be the Lewis acid
of choice for the desired cyanation reaction. Indeed, it afforded a clean reaction profile
and it could be used catalytically. Furthermore, unlike the other Lewis acids (TiCl4,
BF3·Et2O, and ZnCl2), InCl3 did not require stringently anhydrous conditions. Also,
using toluene as a reaction solvent for the cyanation reaction, the first two steps in the
synthesis could be telescoped to improve the overall process throughput.
7
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 28
1
2
3
4
5
6
7
8
9
Concurrently with the optimization of the cyanation of alcohol 2, the generality of
this reaction toward the conversion of other benzylic alcohols to the corresponding
nitriles was briefly examined (Scheme 4). A few representative examples are summarized
in Table 2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 4. InCl3-Catalyzed Cyanation of Benzylic Alcohols
TMSCN (1.2 equiv)
R1
R2
R1
R2
R1
Ar
OH
R2
R1
Ar
CN
R2
R1
Ar
OTMS
R2
InCl3 (10-20 mol%)
Toluene, RT
+
Ar
O
Ar
+
Ar = aryl
Ar = aryl
R1, R2 = H, aryl, alkyl
R1, R2 = H, aryl, alkyl
Triaryl methanols (Entries 6 – 8) reacted readily to afford the corresponding nitriles in
good isolated yields. Furthermore, substrates bearing electron donating groups (Entries 2
and 4) reacted well to afford the expected nitriles in good yields (unoptimized). By
contrast, benzyl alcohol and diaryl methanols bearing electron withdrawing groups or no
substituents at all (Entries 1, 3 and 5) afforded a mixture of products in which the desired
nitrile was generated in low to fair yields (in the case of benzyl alcohol, no cyanation was
observed). The main byproducts obtained in these reactions were the trimethylsilyl
ethers derived by silylation of the alcohols and the symmetrical ethers resulting from a
dehydrative dimerization of the alcohols, presumably via a nucleophilic attack of the
alcohols on the less stabilized putative benzylic carbocations. That triaryl carbinols, as
well as benzhydryl and benzylic alcohols bearing electron donating substituents on the
aromatic rings afforded the best reaction profiles strongly suggest the intermediacy of a
carbenium ion that is highly resonance stabilized. These observations are consistent with
8
ACS Paragon Plus Environment

Page 9 of 28
Organic Process Research & Development
1
2
3
4
Table 2. InCl3-Catalyzed Cyanation of Benzylic Alcohols
5
6
7
8
Byproducts (% AUC)2
Equiv
InCl3
Nitrile Product
(% Yield)1
Entry
Substrate
OTMS
Dimer
9
1
2
3
4
5
0.2
0.2
0.2
0.2
0.1
0
99
0
OH
OH
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
76
23
74
44
0
30
0
0
26
0
MeO
OH
OH
MeO
F
OMe
OH
0
56
F
HO
6
7
0.2
0.2
89
80
0
0
0
0
HO
F
HO
8
0.2
75
0
0
F
F
1Isolated yields by column chromatography on silica gel (unoptimized). 2Rough estimate by HPLC analysis (%AUC);
clean reactions showed none of the OTMS or dimer byproducts and modest isolated yields of nitrile product in these
cases reflect physical losses (1 mmol scale) or incomplete conversions in toluene at RT.
those recently published by Ding et al.10 on the cyanation of electron rich benzylic
alcohols using catalytic InBr3 and TMSCN. No isonitrile products that could be
potentially generated due to the ambident nature of the cyanide ion were observed,12
consistent with mechanistic studies by Ziegler et al., who demonstrated that trityl
isonitriles readily isomerize to the corresponding nitriles in the presence of trityl cations.8
The nitrile product 3 obtained from alcohol 2 using InCl3 and TMSCN was identical in all
respects to the product derived from the cyanation of trityl chloride 2’ with CuCN/CuO.
9
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 28
1
2
3
4
5
6
This cyanation reaction further demonstrates the synthetic usefulness of indium in
organic synthesis.13
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
2.2.2 Optimization of the Cyanation Reaction
The optimization of the cyanation of alcohol 2 focused on defining the mode of
addition of TMSCN and the optimal reaction temperature, adjusting the TMSCN
stoichiometry, and developing a suitably crystallization procedure to isolate the nitrile
with high purity. While small scale reactions conducted at ambient temperature in
toluene typically afforded >60% conversion within the first hour, a complete conversion
required 16 to 18 h. On the other hand, reactions run at higher temperatures (45 – 75 °C)
afforded >85% conversion within the first hour, but required an additional 4 to 5 h to
proceed to completion. It was suspected that due to its noticeable volatility (bp 118 – 119
°C) a substantial amount of the TMSCN was likely in the vapor phase and consequently
only marginally available for reaction over time. A solution to circumvent this issue was
to use slightly more TMSCN (1.35 molar equivalents rather than 1.2 equivalents) and to
add it to the reaction mixture slowly over time (~1 h), while holding the reaction < 40 °C
(typically 25 – 35 °C). The reaction mixture was then agitated at ambient temperature for
1 – 2 h and sampled to assess the conversion of alcohol 2 to ketone 3. The temperature
was then raised to 40 – 50 °C to complete the conversion, usually within 1 h.
A calorimetric screening of the cyanation reaction was conducted prior to scale up.
The study indicated a very exothermic reaction (enthalpy of reaction of 221.4 kJ/mol;
ꢀTad of 90.7 K). The thermal conversion profile at 40 °C showed the reaction to be
essentially complete at the end of the trimethylsilyl cyanide addition (2 h), with no
10
ACS Paragon Plus Environment

Page 11 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
observed accumulation. Based on this calorimetric screening and consistent with the
need to minimize losses of TMSCN to the vapor phase, a dose-controlled addition of the
TMSCN was established and implemented on scale-up.
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 brief study to develop a suitable recrystallization of nitrile 3 was carried out using
crude nitrile of 93.5% purity ( Table 3).
Table 3. Nitrile Crystallization Screening
Entry Solvent
Water
(vol)
Mass
Recovery
HPLC Purity
(%AUC)
Remarks
(%)
-
60
1
2
EtOH
EtOH
-
0.5
-
-
No crystallization
Crystallization on
standing at RT for 10 h
Good crystallization
Good crystallization
No crystallization
3
4
5
6
EtOH
EtOH
IPA
1
2
-
0.5
67
69
-
99.9
99.7
-
IPA
61
99.1
Crystallization on
standing at RT for 10 h
Good crystallization
Good crystallization
Good crystallization
7
8
9
IPA
IPA
IPA
1
2
3
55
70
72
99.2
99.9
97.6
Ethanol (EtOH) and isopropyl alcohol (IPA) were chosen as the primary solvents in an
effort to limit the number of solvents used in the process, since these solvents were also
concurrently being evaluated in the final crystallization of the API. Water was chosen as
the preferred co-solvent. In a typical procedure, the crude nitrile was dissolved in 5
volumes of solvent at reflux (78 – 82 °C), water was added, then the reaction mixture was
cooled to ambient temperature over 2 h. When crystallization was achieved the slurries
were further cooled to 0 °C and stirred for 30 min then filtered, washed with chilled
11
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 28
1
2
3
4
5
6
7
8
9
solvent (1 vol) and dried to afford the purified nitrile 3. The nitrile did not crystallize
from ethanol or IPA as single solvents (Entries 1 and 5). The addition of 0.5 volumes of
water in both cases induced a crystallization of the nitrile upon further standing at
ambient temperature for 10 h. The best results in these screening experiments (1 g scale)
were obtained using 5 volumes of IPA and 2 volumes of water, affording 70% mass
recovery, with high purity (Entry 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
This crystallization procedure was successfully demonstrated at lab scale (starting
from 300 g of 4,4’-difluorobenzophenone 1) to afford pure nitrile 3 (> 99% AUC) in >
85% overall yield from the difluorobenzophenone 1. Analysis of the nitrile 3 by ICP
showed the material to contain <1 ppm of residual indium.
During the production of early batches at scale, severe emulsions were observed in
the workup of the cyanation reaction. This issue was remedied by incorporating
Rochelle’s Salt (sodium potassium tartrate) in the basic aqueous quench and filtering the
batch through Celite® impregnated filter pads. This modification suppressed the
emulsions and streamlined the workup. The two-step telescoped process was scaled-up
at the pilot scale in multiple batches, with the largest batch run at 110 kg input of 4,4’-
difluorobenzophenone 1 (1,000 gallon reactor), affording 76 – 80% overall yields of the
nitrile 3 (see Table 4 for representative batches).
Table 4. Representative Pilot Batches of Nitrile 3
Entry
Ketone 1 Input
Nitrile 3 Output
Yield
(%)
80.0
76.4
77.9
80.0
(kg)
10.0
20.0
20.0
110.0
(kg)
11.2
21.4
21.8
123.4
1
2
3
4
12
ACS Paragon Plus Environment

Page 13 of 28
Organic Process Research & Development
1
2
3
4
5
6
2.3 Development of an Alternative Hydrolysis Process
7
8
9
The hydrolysis process used to produce early batches of the API (concentrated
sulfuric acid, refluxing glacial acetic acid) was evaluated as a baseline experiment and
to generate samples of the API for optimization of the recrystallization. The nitrile 3
(10 g) was heated in refluxing H2SO4/HOAc (0.8 vol each) at 120 – 125 °C for 16
hours to afford the amide with 87% purity (AUC), 5% residual unreacted nitrile 3 and
4.7% of a polar impurity determined to be carboxylic acid 5 (Scheme 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
Scheme 5. Original Acidic Hydrolysis Conditions
O
O
H SO
2 4
HO
NC
H2N
+
+
HOAc
F
F
F
F
F
F
F
F
125–130 °C
3
ICA-17043
(major)
5
6
acid
4,4'-difluoro-
triphenylmethane
The reaction quench was very exothermic, which required a slow, inverse addition to
chilled water while maintaining the temperature at 10 – 20 ºC. To extract the crude
product, the pH of the mixture was adjusted to 6.5–7.5 by slowly adding concentrated
ammonium hydroxide. Dichloromethane was then added and the layers separated. The
aqueous phase, which contained a substantial rag layer, was back extracted with DCM,
resulting in significant emulsions. The combined extracts (including some undissolved
solids) were filtered through a bed of Celite®/silica gel. The filtrate was then
concentrated under reduced pressure to afford the crude amide as a dark yellow powder
(84% crude yield, 95% AUC). Further analysis of the isolated product and impurities
showed that the main impurities observed in the hydrolysis reaction stemmed from
13
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 28
1
2
3
4
5
6
7
8
9
further hydrolysis of the amide to the corresponding carboxylic acid 5 and subsequent
decarboxylation of the acid to the corresponding 4,4’-difluorotriphenylmethane
derivative 6 (Scheme 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
2.3.1 Basic Hydrolysis Screening Experiments
The harsh acidic hydrolysis conditions and more importantly the impractical isolation
and purification process were unattractive to scale without substantial development time.
For this reason, we sought to develop a cleaner reaction with a simpler isolation protocol.
A quick assessment of basic hydrolysis conditions14, 15 using powdered potassium
hydroxide (5.5 molar equiv) in refluxing t-butanol indicated a cleaner conversion profile
that warranted a thorough evaluation (Scheme 6). t-Amyl alcohol (t-AmOH) was then
evaluated as an alternative tertiary alcohol, both because it was easier to handle at
ambient temperature compared to t-BuOH (mp 24 – 25 °C) and because it made possible
to run the hydrolysis reaction at higher temperatures (bp 103 °C).
Scheme 6. Alternative Basic Hydrolysis Conditions
O
KOH
NC
H2N
+
ROH, heat
F
F
F
F
F
F
3
6
ICA-17043
(major)
R = t-Bu, t-Am
4,4'-difluoro-
triphenylmethane
14
ACS Paragon Plus Environment

Page 15 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
A faster conversion was achieved in this solvent at 100 °C (67% conversion in 2 h in t-
AmOH vs. 65% conversion in 6 h in refluxing t-BuOH), with a complete conversion
within 8 h.
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
Encouraged by these initial results we screened potentially suitable common
alcoholic solvents. The results from this screen are summarized in Table 5.
Table 5. Basic Hydrolysis Solvent Screen
Entry
Solvent
Temp
(°C)
82
Time
% Amide1 % Unreacted
Nitrile1
% Total
Impurities1
1.1
(h)
6
1
2
3
4
5
6
7
8
t-BuOH
t-AmOH
IPA
n-BuOH
i-BuOH
i-AmOH
n-Pentanol
2-BuOH
69.7
90.2
51.7
47.9
61.3
61.6
44.6
73.9
29.2
5.5
100
82
5
10
7
7
7
4.3
6.9
36.7
24.1
20.7
39.6
15.3
41.4
15.4
14.6
17.7
15.8
10.8
100
100
100
100
98
6
6
1% AUC by HPLC analysis
.
The cleanest and fastest reaction profile was achieved in t-AmOH. Therefore, further
process optimization was carried out using this solvent.
To determine the most important parameters in the hydrolysis reaction we relied on a
design of experiments (DoE) methodology.16a-c The goals of the study were: (1) to
identify reaction conditions that afforded the highest conversion of the nitrile 3 to the API
and, (2) to identify key parameters that minimized the formation of impurities. Two
factors, the amount of KOH and the solvent volume, were evaluated in the DoE study.
All experiments were run at 100 °C. The experimental details and the complete set of
DoE data are provided in the Supporting Information.
15
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 28
1
2
3
4
5
6
7
8
9
The DoE study helped gain the following insights into the performance of this
hydrolysis reaction: (1) both the amount of KOH and the solvent volume, as well as the
interaction between these two factors have a significant impact on the reaction profile and
the amount of impurities generated; (2) the fastest conversion is achieved at high KOH
loading and low solvent volumes (i.e. high concentration); (3) impurities are minimized
under dilute reaction conditions; (4) the amide is gradually depleted over time when the
reaction is allowed to proceed over an extended time. This loss is most significant at high
reaction concentration and high KOH loading. These conclusions agreed well with the
results achieved under typical conditions in most preliminary preparative scale
experiments in t-AmOH (5.5 molar equiv KOH, 10 vol solvent, 100 °C). Indeed, these
conditions consistently provided the highest conversion of nitrile 3 to the API within 6 –
8 h with minimal impurities, but showed increasing API degradation through hydrolysis
to the acid 5 and decarboxylation to the triphenylmethane derivative 6 when the reaction
was allowed to proceed at 100 °C for 12 h.
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
2.3.2 Hydrolysis Process Optimization
Additional experiments were conducted to determine the best reaction temperature,
concentration and time that would maximize the conversion while minimizing the
amount of impurities. The experiments were conducted at 10 g scale, using 5.4 molar
equivalents of KOH and the solvent volumes and temperatures shown in Table 6.
16
ACS Paragon Plus Environment

Page 17 of 28
Organic Process Research & Development
1
2
3
4
Table 6. Effect of Temperature and Concentration on Reaction Profile
5
6
7
8
API
(%AUC)
4,4’-Difluoro-
triphenylmethane 6
(%AUC)
Total Impurities
(%AUC)
Solvent
Volumes
Temp
(°C)
Entry
9
5 h
90
9 h
---
12 h
86.1
5 h
3.3
0.6
9 h
---
12 h
8.9
5 h
4.4
9 h 12 h
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
2
3
10
10
15
100
90
100
---
6.1
---
12.3
7.3
9.4
74
86.61
87.8 90.2
---
1.4
---
2.0
5.4
4.9
7.31
89.8 2.61
1HPLC analysis after 5.5 hours
As expected and consistent with the DoE study, a slower conversion was observed at
lower concentration (compare Entries 1 and 3) and at lower temperature (compare Entries
1 and 2). Overall, the hydrolysis reaction conducted at 90 °C using 10 volumes of solvent
(Entry 2) afforded the best balance for the highest conversion and the lowest amount of
impurities, particularly residual unreacted nitrile 3, which was the most difficult process
impurity to purge out during the final crystallization of the API. These reaction
conditions were selected for further scale-up.
In contrast to the original sulfuric acid-mediated hydrolysis in glacial acetic acid, the
isolation of the crude API from the basic hydrolysis process was simpler and
straightforward. A comparison between the workup and isolation procedures is provided
in Table 7.
17
ACS Paragon Plus Environment

Organic Process Research & Development
Page 18 of 28
1
2
3
4
Table 7. Comparison of Workup and Isolation Procedures of Crude API
5
Processing
Step/Appearance
Quench
Hydrolysis in
H2SO4/AcOH
Hydrolysis in
KOH/t-AmOH
6
7
8
9
Cool reaction mixture to RT;
inversely add it to cold water (5
– 10 °C) over 60 to 90 min
Cool reaction mixture to 5 – 10
°C; add 3 volumes of cold (5 °C)
6N HCl over 30 to 60 min
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
Isolation of crude - Add conc. NH4OH and adjust
API
- filter and wash cake with water
- dry under reduced pressure
pH to 6.5 – 7.5
- Extract with DCM (severe
emulsions)
- Filter DCM extracts through
Celite®/silica gel
- Strip to a solid residue
Dark yellow powder
Appearance
Off-white to white powder
Further process optimization focused on developing a robust crystallization
procedure to purify the API. Both single solvents and binary solvent systems were
evaluated in screening experiments (Table 8).
Table 8. Recrystallization Screening
Entry
Solvent
(mL/g)
Antisolvent
(mL/g)
---
Total
Volume
16.4
24
Recovery
HPLC
(% AUC)
95.3
(%)
74
1
2
3
4
EtOH (16.4)
IPA (24)
EtOH (18)
MEK (10)
---
74
80
94.3
81.2
H2O (4)
n-Heptane
(12)
22
22
58
93.1
5
6
EtOAc (8)
n-Heptane
(10)
n-Heptane (8)
18
22
60
86
97.4
96.2
Toluene (14)
Saturated solutions of the crude API (93.5% AUC; material from acidic hydrolysis) were
prepared near the boiling point of the chosen primary solvent. In experiments using a
single solvent, crystallization was induced by cooling the solution, whereas in the case of
18
ACS Paragon Plus Environment

Page 19 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
binary solvent systems, it was achieved by adding the chosen antisolvent to the hot
solution. The crystallizing mixtures were then cooled to 0 °C over 2 h and filtered.
Conditions that afforded materials of lower purity than that of the starting crude API
(Entries 3 and 4) and those yielding low recoveries (Entries 4 and 5) were eliminated
from further consideration. The best solvent systems (Entries 1, 2, and 6) were
reevaluated at higher concentrations using crude API from the basic hydrolysis process
(96% AUC), with the goal of improving the mass recoveries. Results are summarized in
Table 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
Table 9. Further Recrystallization Screening Experiments
Entry
Solvent
(mL/g)
Antisolvent
Total
Volume
11
Recovery
(%)
HPLC
(% AUC)
99.6
(mL/g)
---
---
1
2
3
EtOH (11)
IPA (13)
Toluene (8) n-Heptane (4)
82.0
88.2
86.0
13
12
99.7
99.6
We chose IPA as the preferred solvent for purifying the API since it afforded the best
combination of mass recovery and product purity and since it was used upstream in the
purification of nitrile 3. This was an important consideration in order to minimize the
number of organic volatile impurities in the API. Finally, a polish filtration of the hot
solution of the API through a 1.2 µm Teflon® filter was incorporated into the
crystallization process to remove inorganics (KCl) trapped in the crude API at the quench
of the hydrolysis reaction.
19
ACS Paragon Plus Environment

Organic Process Research & Development
Page 20 of 28
1
2
3
4
5
6
2.3.3 Process Demonstration and Pilot Scale Production
7
8
9
The optimized, alternative chemistry is shown in Scheme 7. Prior to technology
transfer for pilot scale production, incremental scale-up experiments were conducted at
25 g, 80 g and 300 g scale. Results from the 300 g batch are provided below.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 7. Alternative Chemistry to ICA-17043.
O
TMSCN
NC
HO
PhMgBr/THF
F
Toluene
InCl3 (20 mol%)
Toluene
RT - 50 oC
F
F
F
F
F
3
1
2
O
KOH
H2N
t-AmOH
90 °C
F
F
ICA-17043
The Grignard reaction followed by the telescoped indium trichloride mediated
cyanation of crude alcohol 2 performed well, consistent with smaller scale reactions,
affording 345 g of nitrile 3 as an off-white, crystalline solid (89% overall yield, >99%
AUC by HPLC analysis). Subsequent base mediated hydrolysis of the nitrile (345 g
scale) afforded 307 g of crude API (94% AUC). The crude (300 g) was recrystallized
from 13 volumes of IPA to afford the purified API as a white, crystalline solid (280 g,
77.1% overall yield from nitrile 3, 99.2 wt. % by validated HPLC assay). The overall
yield from 4,4’-difluorobenzophenone 1 was ~ 68.5%. Other analytical data were
consistent with smaller scale batches (<1 ppm residual indium, 876 ppm residual IPA).
20
ACS Paragon Plus Environment

Page 21 of 28
Organic Process Research & Development
1
2
3
4
5
6
The process was further scaled up at pilot scale to produce several batches of the API.
Representative results from pilot scale production are shown in Table 10.
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
Table 10. Representative Pilot Scale Production Results
Entry
Input
Output
Yield
(%)
Reactor Size
(Gal)
1
2
3
4
5
Ketone 1
110.0 kg
Nitrile 3
60.0 kg
Nitrile 3
63.4 kg
Crude API
54.0 kg1
Crude API
54.0 kg1
Nitrile 3
123.4 kg
Crude API
60.5 kg1
Crude API
62.7 kg2
Pure API
42.3 kg
80.1
1,000
95.21
93.32
78.3
80.3
300
300
300
300
Pure API
43.4 kg
1KF = 6.5% and ROI = 7.2%; 2ROI = 10.0%
3. Conclusions
In conclusion, an improved process suitable for the pilot scale production of a
potassium ion channel blocker from 4,4’-difluorobenzophenone 1 was developed and
optimized to produce multiple batches of the API. The Grignard reaction was
conveniently conducted in toluene rather than diethyl ether to afford crude alcohol 2
suitable for processing without further purification. The original two-stage cyanation
reaction employing a moisture labile tertiary chloride intermediate 2’ was replaced with a
direct cyanation employing a catalytic amount of indium trichloride as the Lewis acid and
trimethylsilyl cyanide as the cyanide source. The process throughput was further
enhanced by telescoping the alcohol into the cyanation reaction. A basic hydrolysis of the
intermediate nitrile 3 was developed and optimized using potassium hydroxide in t-amyl
alcohol and employing a direct drop isolation process instead of an extractive workup. An
21
ACS Paragon Plus Environment

Organic Process Research & Development
Page 22 of 28
1
2
3
4
5
6
improved crystallization process using isopropyl alcohol as the solvent was implemented
to afford the API in good yield and high purity.
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
4. Experimental
General. HPLC data during process development (non-validated method) was collected
using a LUNA C8 (3µm) column using water (modified with 0.1% TFA) and acetonitrile
(modified with 0.1% TFA) as mobile phases. The following gradient was used: 30%
acetonitrile isocratic for 1 min, 30% to 90% acetonitrile over 7 min, 90% acetonitrile
isocratic for 5 min, 90% to 30% acetonitrile over 2 min and 30% acetonitrile isocratic for
2 min, detector at 220 nm; RT for API is 9.7 min, alcohol 2 is 10.7 min and nitrile 3 is
11.8 min. HPLC data during pilot scale production (validated method) was collected
using a Waters Symmetry Shield RP18 (5µm) column using water and acetonitrile as
mobile phases. The following gradient was used: 50% acetonitrile isocratic for 25 min,
50% to 100% acetonitrile over 15 min, 100% acetonitrile isocratic for 10 min, 100% to
50% acetonitrile over 1 min and 50% acetonitrile isocratic for 15 min, detector at 210nm;
RT for API is 16.9 min, alcohol 2 is 32.7 min and nitrile 3 is 36.0 min.
Preparation of Bis-(4-Fluorophenyl) phenylmethanol 2. A glass-lined stainless
steel reactor was charged with 4,4’-difluorobenzophenone 1 (110.0 kg, 504.1 mols),
followed by toluene (1,320 L). Approximately half the amount of toluene was distilled
under reduced pressure (50 °C, 25 inches vacuum), leaving approximately 770 L of
residual solution, which was then cooled to 30 °C. A solution of phenylmagnesium
bromide in tetrahydrofuran (1 M, 609.4 kg) was added to the batch over 1 h and 30 min,
22
ACS Paragon Plus Environment

Page 23 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
while maintaining a temperature of 20 – 40 °C. The reaction mixture was further agitated
at 20 – 30 °C for 1 h, whereupon in-process analysis by HPLC indicated <3% (AUC)
ketone remaining (reaction considered complete when no more than 3% ketone remains).
The batch was cooled to 5 °C, and the reaction quenched with 5 °C aqueous HCl
(prepared by mixing 128.7 kg of conc. HCl with 160.3 kg of deionized water) over 1 h 15
min, while maintaining the batch temperature between 5 and 20 °C. The batch was
allowed to warm to 25 – 30 °C and further agitated for 1 h. The bottom layer was
separated and the organic layer was washed with aqueous sodium bicarbonate (prepared
by dissolving 33 kg of sodium bicarbonate in 78.6 gal of deionized water) then deionized
water (2 x 87.3 gal). The crude alcohol 2 solution was then concentrated by distillation
under reduced pressure (50 °C, 25 inches vacuum) to approximately 440 L of residual
solution. Toluene (1,100 L) was added and the batch was further distilled to
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
approximately 440 L of residual solution. Analysis of an aliquot from the residual
solution by Karl Fisher titration indicated 0.0% (w/w) water. The residual toluene
solution of alcohol 2 was used in the cyanation step without further processing.
Preparation of Bis-(4-Fluorophenyl) phenylacetonitrile 3. A slurry of indium
trichloride (22.44 kg) in toluene (308 L) was added to the residual solution of alcohol 2
from above, which was agitated at ambient temperature for 30 min. Trimethylsilyl
cyanide (68.2 kg) was then added over 1 h, while maintaining the temperature at 20 – 40
°C. The reaction mixture was further agitated for 1.5 h at ambient temperature and
analyzed by HPLC for residual alcohol 2 (~60% AUC). The batch was then heated and
agitated at 40 – 50 °C for 1 h and analyzed by HPLC (3% AUC for residual alcohol 2;
reaction considered complete when no more than 3% alcohol 2 remains). After further
23
ACS Paragon Plus Environment

Organic Process Research & Development
Page 24 of 28
1
2
3
4
5
6
7
8
9
reaction at 40 – 50 °C for 1 h, the reaction was quenched with aqueous KOH containing
Rochelle’s salt (74.8 kg KOH and 74.8 kg of sodium potassium tartrate dissolved in
176.5 gal of deionized water), maintaining the temperature at 20 – 30 °C. The quenched
reaction mixture was further agitated overnight (16 hours) and the layers separated. The
organic layer was filtered through Celite® impregnated filter pads into a glass-lined
reactor, and further washed with deionized water (3 x 145.5 gal). The crude nitrile 3
solution was distilled under reduced pressure (50 °C, 25 inches vacuum) to the minimum
stir volume (until no further toluene distilled). Isopropyl alcohol (157.1 gal) was added,
followed by deionized water (78.6 gal), and the resulting suspension was heated to 65 –
75 °C and held at that temperature for an additional 30 min. The hot solution was then
cooled to 5 °C at a rate of 10 °C/h (9.5 h) to crystallize the product. The resulting slurry
was aged at 5 °C for 1 h, then transferred to a stainless steel centrifuge and filtered in
sections. The reactor was rinsed with isopropyl alcohol (26.5 gal) and the rinse
transferred to the centrifuge to wash the wet cake. The wet solids were spun to dryness,
transferred to a tray dryer and dried at 40 – 45 °C for 17 h to afford nitrile 3 as a white
crystalline powder (123.4 kg, 80.2% isolated yield over two steps). Analytical data:
HPLC assay: 97.4 % AUC; residue on ignition (sulfated ash): 0.04% (w/w); residual
indium: none detected (ICP; limit of detection: 1 ppm); 1H NMR (300 MHz, CDCl3):
7.01-7.08 (m, 4H); 7.15-7.34 (m, 6H); 7.34-7.39 (m, 3H). 13C NMR (75 MHz, CDCl3):
56.6; 116.0; 116.2; 123.5; 128.8; 128.9; 129.2; 130.8; 131.0; 136.3; 140.2; 161.1; 164.4.
19F NMR (282 MHz, CDCl3): -113.71.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Preparation of 2,2-Bis-(4-Fluorophenyl) 2-phenylacetamide, ICA-17043. A
Hastelloy reactor was charged with nitrile intermediate 3 (60.0 kg, 196.5 mols), followed
24
ACS Paragon Plus Environment

Page 25 of 28
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
by potassium hydroxide flakes (60.0 kg) and t-amyl alcohol (480 kg). The resulting
mixture was then heated and agitated at 85 – 95 °C for 10.5 h, whereupon in-process
analysis by HPLC indicated 1.1% (AUC) nitrile 3 remained. The batch was cooled to 5
°C and the reaction quenched with 5 °C aqueous HCl (prepared by diluting 104.7 kg of
conc. HCl with 93 L of deionized water) over 1 h and 30 min, maintaining the batch
temperature at 5 – 15 °C. The resulting slurry was further agitated at 15 – 25 °C for 45
min, then transferred to a pressure filter dryer and filtered. The wet cake was washed with
deionized water (3 x 300 L), conditioned for 3 h then further dried at 45 – 50 °C for 12 h
to constant weight, affording crude ICA-17043 as a white crystalline powder (60.5 kg,
95.2% crude yield). HPLC assay: 99.3% AUC; KF: 6.5% (w/w); residue on ignition
(sulfated ash): 7.2% (w/w); residual indium: none detected (ICP; limit of detection: 1
ppm).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Crude ICA-17043 (54.0 kg) was charged into a glass-lined reactor, followed by
isopropyl alcohol (847.8 kg), and the batch was heated and held at 75 – 85 °C for 1 h to
dissolve the solids. The resulting solution was polish-filtered through a 1.2 µm in-line
Teflon® filter (with the filter housing held at 80 °C) into a Hastelloy reactor, then distilled
at atmospheric pressure to approximately 756 L of residual solution. The batch was
cooled to 5 °C over 8 h to crystallize the API. The resulting slurry was transferred to a
pressure filter dryer and filtered. The wet cake was washed with 5 °C isopropyl alcohol
(52.8 kg), conditioned at ambient temperature for 3 h, then further dried at 45 – 50 °C to
constant weight for 12 h to afford ICA-17043 as a white crystalline powder (43.4 kg,
80.3% recovery; 61.3% overall yield from the difluorobenzophenone 1). Analytical data:
DSC melting endotherm: 187.2 °C; HPLC assay: 100.5% w/w (specifications: 98.0 –
25
ACS Paragon Plus Environment

Organic Process Research & Development
Page 26 of 28
1
2
3
4
5
6
7
8
9
102.0% w/w); total related impurities: 0.25% (AUC); residue on ignition (sulfated ash):
0.09% w/w; residual IPA: 461 ppm; residual t-amyl alcohol: none detected; residual
indium: none detected (ICP; limit of detection: 1 ppm); particle size distribution by laser
diffraction: D10 = 8.2 micron, D50 = 26.9 micron, and D90 = 53.6 micron; KF = 0.0%
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
w/w; H NMR (300 MHz, CDCl3): 5.71(bs, 1H); 6.73(bs, 1H); 6.95-7.00 (m, 4H); 7.19-
7.32 (m, 9H). 13C NMR (75 MHz, CDCl3): 66.8; 115.0; 115.3; 127.8; 128.6; 130.5;
132.4; 132.5; 139.2; 143.3; 160.4; 163.7; 176.2. 19F NMR (282 MHz, CDCl3): δ-115.66.
References
1. Ledoux, J.; Werner, M. E.; Brayden, J. E.; Nelson, M. Physiology 2006, 21, 69.
2. Conejo-Garcia, A.; Campo, J. M. Curr. Med. Chem. 2008, 15(13), 1305.
3. Wulff, H.; Kolski-Andreaco, A.; Sankaranarayanan, A.; Sabatier, J-M.; Shakkottai,
V. Curr. Med. Chem. 2007, 14(13), 1437.
4. Baell, J. B. Expert Opin. Therap. Patents, 2005, 15(9), 1209.
5. Wulff, H.; Zhorov, B. S. Chem. Rev. 2008, 108, 1744.
6. Impurities observed using phenylmagnesium chloride were likely present in the
commercial supply of this reagent. However, since the available supplies of
phenylmagnesium bromide afforded clean reactions, the use of phenylmagnesium
chloride was discontinued.
7. For general references on the synthesis of nitriles, see (a) North, M. Comprehensive
Organic Functional Group Transformations, Vol 3, Part III; Elsevier, 1995, pp 611-
640; (b) Mowry, D. T. Chem. Rev. 1948, 42, 198; (c) Rappoport, Z. The Chemistry of
the Cyano Group; Wiley-Interscience, 1970; (d) McClelland, R. A.;
26
ACS Paragon Plus Environment

Page 27 of 28
Organic Process Research & Development
1
2
3
4
5
6
Kanagasabapathy, V. M.; Banait, S. N.; Steenken, S. J. Am. Chem. Soc. 1989,
111(11), 3966.
7
8
9
8. Ziegler, H. E.; Wo, S. J. Org. Chem. 1994, 59, 3838.
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
9. At the time of this development work (September to December 2003), and during
subsequent pilot scale production of the API (2004 to 2006), few examples of such a
direct cyanation of benzylic alcohols were reported in the literature. However Ding et
al.10 have recently (2008) published an excellent account of such a direct cyanation of
electron rich benzylic alcohols, catalyzed by indium tribromide, which confirms the
general reactivity trends observed during this development work and further
highlights the synthetic usefulness of this methodology.
10. Chen, G.; Wang, Z.; Wu, J.; Ding, K. Org. Lett. 2008, 10, 4573.
11. Caron, S.; Vazquez, E. Org. Process Res. Dev. 2001, 5(6), 587.
12. See for example Kitano, Y.; Chiba, K.; Tada, M. Synthesis, 2001, 437.
13. See for example the following review articles, (a) Podlech, J.; Maier, T. C. Synthesis,
2003, 633; (b) Cintas, P. Synlett, 1995, 1087; Nair, V.; Ros, S.; C. N. Jayan; Pillai, B.
S. Tetrahedron, 2004, 60, 1959.
14. Hall, J. H.; Gisler, M. J. Org. Chem., 1976, 41, 3769.
15. Kim et al. recently developed a highly efficient InCl3-catalyzed hydration of nitriles
using acetaldoxime as a water surrogate. This new methodology is potentially a
superior alternative to the basic hydrolysis of α-aryl nitriles with KOH using tertiary
alcohols as such as t-BuOH or t-AmOH as solvents: Kim, E. S.; Lee, H. S.; Kim, S.
H.; Kim, J. N. Tetrahedron Lett. 2010, 51, 1589.
27
ACS Paragon Plus Environment

Organic Process Research & Development
Page 28 of 28
1
2
3
4
5
6
7
8
9
16. For representative overviews of DoE, see for example (a) Carlson, R. Design and
Optimization in Organic Synthesis; Elsevier: Amsterdam, 1992; (b) Hicks, C. R.;
Turner, K. V., Jr. Fundamental Concepts in the Design of Experiments; Oxford
University Press: New York, 1999; (c) Owen, M. R.; Luscombe, C.; Lai, L-W.;
Godbert, S.; Crookes, D. L.; Emiabata-Smith, D. Org.Process Res. Dev. 2001, 5(3),
308.
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
Acknowledgments
We thank Dr. Sridhar Vaddeboina and his team (AMRI) for analytical methods
development. We are grateful to the manuscript reviewers for helpful comments and
feedback.
Supporting Information. (1) DoE study data, including 2 Tables and 12 Figures; (2)
NMR and HPLC data for laboratory scale (300 g) process demonstration batch (alcohol
2, nitrile 3 and ICA-17043). This information is available free of charge via the Internet
at http://pubs.acs.org/.
28
ACS Paragon Plus Environment