Multkilogram scale-up of a reductive alkylation route to a novel PARP inhibitor

Article abstract of DOI:10.1021/op200238p

Novel PARP inhibitor 1 is a promising new candidate for treatment of breast and ovarian cancer. A modified synthetic route to 1 has been developed and demonstrated on 7 kg scale. In order to scale up the synthesis to multikilogram scale, several synthetic challenges needed to be overcome. The key issues included significant thermal hazards present in a Leimgruber-Batcho indole synthesis, a low-yielding side-chain installation, a nonrobust Suzuki coupling and hydrogen cyanide generation during a reductive amination. In addition to these issues, changing from intravenous to oral delivery required a new salt form and therefore a new crystallization procedure. This contribution describes development work to solve these issues and scaling up of the new process in the pilot plant.

Full text of DOI:10.1021/op200238p

Subscriber access provided by FORDHAM UNIVERSITY
Full Paper
Multi-Kilogram Scale-Up of a Reductive
Alkylation Route to a Novel PARP Inhibitor.
Adam Thomas Gillmore, Matthew Badland, Clare L Crook, Nieves Castro, Doug J Critcher,
S FUSSELL, Katherine J Jones, Matthew Jones, Eleftherios kougoulos, Jinu S. Mathew,
Lynne McMillan, John Pearce, Fiona Rawlinson, Alexandra Sherlock, and Robert Walton
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/op200238p • Publication Date (Web): 14 Nov 2012
Downloaded from http://pubs.acs.org on November 28, 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 23
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
MultiꢀKilogram ScaleꢀUp of a Reductive Alkylation
Route to a Novel PARP Inhibitor.
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
Adam T. Gillmore,* Matthew Badland, Clare L. Crook, Nieves M. Castro, Douglas J. Critcher, Steven
J. Fussell, Katherine J. Jones, Matthew C. Jones, Eleftherios Kougoulos, Jinu S. Mathew, Lynne
*
McMillan, John E. Pearce, Fiona L. Rawlinson, Alexandra E. Sherlock and Robert Walton.
Chemical Research and Development, Pfizer Global Research and Development, Sandwich
Laboratories, Ramsgate Road, Sandwich, Kent CT13 9NJ, United Kingdom.
*
To whom correspondence should be addressed. Eꢀmail: Adam.Gillmore@Pfizer.com and
Robert.Walton@Pfizer.com.
RECEIVED DATE (to be automatically inserted after your manuscript is accepted if required
according to the journal that you are submitting your paper to)
TOC graphic
H
N
CO H
2
O
12 steps
.
(S)-CSA
F
N
N
F
H
H
Abstract Novel PARP inhibitor 1 is a promising new candidate for treatment of breast and ovarian
cancer. A modified synthetic route to 1 has been developed and demonstrated on 7 kg scale. In order to
scale up the synthesis to multiꢀkilogram scale, several synthetic challenges needed to be overcome. The
1
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 23
key issues included significant thermal hazards present in a LeimgruberꢀBatcho indole synthesis, a lowꢀ
yielding sideꢀchain installation, a nonꢀrobust Suzuki coupling and hydrogen cyanide generation during a
reductive amination. In addition to these issues, changing from IV to oral delivery required a new salt
form and therefore a new crystallization procedure. This paper describes development work to solve
these issues and scaling up of the new process in the pilot plant.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Introduction
Inhibition of Poly(ADP ribose) polymerase, or PARP, is an exciting new mechanism for the treatment
1
of cancer. The PARP enzyme is responsible for repair of damaged DNA in both normal and tumor
cells, and inhibition of this repair mechanism is expected to make the cell more likely to undergo
apoptosis. Preꢀclinical work has shown that PARP inhibitors coꢀadministered with a standard
2
chemotherapuetic agent are more effective than the standard treatment alone. Several PARP inhibitors,
3
including PFꢀ01367338 (1) which is currently in development (Chart 1), have shown promising results
4
in recent clinical trials for the treatment of advanced solid tumors. This paper details the development
work carried out to design and scale up a safe process for the manufacture of 1 on multiꢀkilogram scale.
H
N
O
N
N
F
H
H
1
Chart 1. PFꢀ01367338, a PARP inhibitor.
The early synthetic route to 1 is shown in Schemes 1 and 2. The key starting material for the inꢀhouse
synthesis of 1 is indolazepine 8. The synthesis of indolazepine 8 was outsourced for the first GMP
campaign and started with nitration of 2ꢀmethylꢀ5ꢀfluorobenzoic acid (2) followed by Leimgruberꢀ
Batcho indole synthesis. The resulting indole 5 was then reacted with 1ꢀdimethylaminoꢀ2ꢀnitroethylene
(DMANE) to append the side chain. Subsequent 2ꢀstage reduction then gave the desired lactam 8.
2
ACS Paragon Plus Environment

Page 3 of 23
Organic Process Research & Development
OMe
O
O
OMe
1
2
3
4
5
6
7
8
9
DMFDMA,
CO H
i) HNO , H SO
2
3 2 4
DMF, Et N
120 C
3
N
ii) MeOH, H SO
2
4
o
(
40%)
F
NO2
F
NO2
F
4
3
2
NO2
i) TFA, DCM
O
OMe
O
OMe
i) H , Pd/C,
2
1-dimethylamino
-2-nitroethylene
ii) DCM/MeOH
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
MeOH, AcONa
ii) DCM/hexane
chromatography
iii) NaHCO , reslurry
3
N
H
N
F
F
H
(
45% from 3)
(98%)
6
5
NO2
H
O
OMe
N
O
i) NaBH , THF,
4
i) H , Pd/C, MeOH
2
MeOH
ii) Column, DCM
ii) DCM/MeOH
reslurry
N
H
F
N
F
H
(
33%)
(88%)
7
8
Scheme 1. Outsourced route to key starting material indolazepine 8.
Indolazepine 8 was selectively brominated with pyridinium tribromide to give 9 which was coupled
with boronic acid 10 under standard Suzuki conditions. Reductive amination with methylamine then
gave 1 which was converted to the phosphate salt 13 which is the final active pharmaceutical ingredient
(API).
3
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 23
H
N
H
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
O
N
HO
HO
H
O
N
B
CHO O
Pyr.HBr , THF
DCM
3
Br
10
N
F
CHO
H
75%
N
F
N
H
F
70%
H
8
11
9
Pd(PPh ) , DMAc
Na CO
3
4
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
H
i) MeNH /MeOH,
HCl (g)
NaCNBH3
ii) Aq. HCl
H
2
N
O
N
O
NaOH, MeOH
0%
N
N
F
N
N
H
80%
H
F
H
6
H
12
.HCl
1
H
N
i) H PO4
ii) Water reslurry
3
O
.
H PO
3 4
N
N
F
H
H
1
3
Scheme 2. Inꢀhouse synthesis of 13 for first GMP campaign.
Whilst this route was successfully used to prepare 2 kg of phosphate salt for use in preꢀclinical and
early clinical studies, some challenges still remained for the preparation of larger multiꢀkilogram
quantities, as outlined below:
1
. Although preparation of indole 5 from 2 had previously been successfully outsourced, whilst
running this chemistry inꢀhouse we discovered that the onset of an exothermic event during the
LeimgruberꢀBatcho enamine formation (conversion of 3 to 4) occurred at the current reaction
temperature.
2
. The reduction of nitroalkene 6 to 7 had not performed well on scale leading to a low yield and the
requirement for large scale chromatography.
3. The Suzuki coupling between aryl bromide 9 and boronic acid 10 failed to go to completion on
scaleꢀup and required an additional catalyst charge.
4
. The generation of HCN during the reductive amination of aldehyde 11 was not desirable and
4
ACS Paragon Plus Environment

Page 5 of 23
Organic Process Research & Development
alternative conditions were sought.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
. The phosphate salt 13 was initially delivered as an IV formulation during Phase I clinical trials,
however we were required to develop an oral product which required a new salt. After screening and
evaluation work, the (S)ꢀcamphorsulfonate salt ((S)ꢀCSA) was selected for tablet formulation and meant
that a new API crystallization procedure was required.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Results and Discussion
During the course of our lab development work stocks of indole 5 were required to allow us to
address the scaleꢀup issues observed during the reduction of nitroalkene 6. Routine process safety
testing of enamine 4 as a concentrate showed the onset of a significant exothermic event from 75 °C.
Preliminary testing of the reaction mixture showed the onset of an exothermic event from 120 °C, which
was the same as the intended reaction temperature. Our general purpose lab and plant scaleꢀup facilities
are not configured to routinely handle this type of hazard so inherently safer conditions were sought,
either through avoiding this intermediate or through use of alternative conditions. The analogous
pyrrolidine enamine was prepared in the hope that the safety profile of the reaction would be improved,
however this was shown not to be the case with the exothermic onset from 90 °C. Replacement of
5
dimethylformamide dimethylacetal (DMFDMA) with Bredereck’s reagent did afford complete reaction
within 4 h at 50 °C, however the impurity profile was poor and the crude mixture did not work in the
6
subsequent reduction step. Tris(dimethylamino)methane (TDMAM) gave complete reaction to the
dimethylenamine 4 within 4.5 h at 50 °C and the reaction profile was better. However, the exothermic
onset was also lower with this reagent (75 °C) and unfortunately TDMAM was not readily
commercially available on the scale that would have been required for the next campaign (~100 kg) so
we discontinued this line of investigation.
In parallel to this work, simply reducing the reaction temperature was also investigated. The reaction
rate rapidly dropped off below 90 °C resulting in extended reaction time and product degradation. A
5
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 23
suitable compromise was achieved by running the reaction at 95 °C since isothermal Accelerating Rate
Calorimetry (ARC) testing detected no significant thermal events over 3 days at 100 °C. The yield
obtained under these conditions was lower (32% from 3) but the safety risks were now better understood
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
7
and manageable in general purpose equipment. The preparation of indole 5 was then outsourced to
provide 17.5 kg in 17% yield over four steps from 2.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
With indole 5 now in hand we turned our attention to the sequence of steps to convert it through to 8.
The sodium borohydride reduction of 6 had not performed well on scale due to poor chemoselectivity
and the generation of a significant impurity thought to be conjugate addition of the product onto the
starting material. In addition to this, it was known that the reaction of 5 with DMANE had some process
8
safety concerns which needed careful management and control. We therefore sought an inherently safer
route which would convert indole 5 to 8 avoiding DMANE and the subsequent reduction steps.
9
Highꢀthroughput screening of a variety of twoꢀcarbon electrophiles under both Lewis and Brønsted
acid catalysis revealed that the 3ꢀposition of this indole was surprisingly unreactive. Even preꢀformation
of the indolyl magnesium salt did not improve the reactivity. The only promising hit came with
10
reductive alkylation of phthalimidoacetaldehyde diethyl acetal (14) using triethylsilane (TES) and
TFA to give 16, however this reaction only gave a meager 26% yield after 4 days with the remainder of
the mass balance comprising various adducts arising from reaction of the starting material or product
with the intermediate benzylic cation (up to 15%). In order to increase the reaction rate we investigated
1
1
the use of the deprotected aldehyde 15. This did indeed speed up the reaction to give complete
conversion in 4 h, however a significant level of alcohol 17 (14% by HPLC) was observed due to direct
reduction of the aldehyde, as well as multiple other byꢀproducts. While the desired product 16 only
constituted 40% of the mixture fortunately, 16 crystallized from the reaction mixture with excellent
purity purging all of the byꢀproducts. Although the yield was still relatively low (typically 35ꢀ38%
isolated yield), given the high quality of the material obtained and the lack of a suitable alternative, this
process was selected for scale up. Our main concern was whether the product distribution (and hence
6
ACS Paragon Plus Environment

Page 7 of 23
Organic Process Research & Development
the yield) would remain the same at larger scale.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
O
O
i) 1M HCl
ii) DCM, Toluene
EtO
EtO
TFA, TES
DCM
N
N
O
O
O
O
N
O
OMe
O
14
15
45%
O
OMe
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
F
H
N
H
F
1
6
5
H
N
O
Aq MeNH2
O
85%
N
H
N
F
HO
O
8
Impurity 17
Scheme 3. Alternative reductive alkylation route to azepine 8.
We conducted a series of experiments investigating the order of addition of the various reagents to
give an insight into factors affecting impurity formation. Initially we thought that a controlled addition
of the triethylsilane (TES) might limit reduction of aldehyde 15 since it would not be present in excess,
however, new impurities were formed when the indole 5 was stirred with TFA. Dosing 5 into the
reaction mixture resulted in very high levels of aldehyde reduction. The best process involved adding
TFA to a solution of the indole 5, TES and aldehyde 15. Increasing the quantity of TFA resulted in an
increased level of one of the byꢀproducts, so we decided to add the TFA as a DCM solution to avoid
high local concentration upon dosing. It was also noted that cooling the suspension prior to filtration
resulted in poor quality product due to incorporation of impurities. This process was successfully run in
the pilot plant using 15.7 kg of 5 and delivered 12.9 kg of product 16 in 43% yield as expected. The
subsequent deprotection of the phthalimide group was carried out in aqueous methylamine and was
scaled up uneventfully to give 8 in 82% yield. The bromination of 8 proceeded with excellent chemical
7
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 23
yield, so we turned our attention to the subsequent Suzuki coupling reaction.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
Pd(PPh ) is often a capricious catalyst due to its airꢀsensitivity and the Suzuki coupling of 9 and 10
3
4
had often failed to reach completion using this catalyst. We desired a more robust and active catalyst,
and Pd(dppf)Cl .DCM was identified from a highꢀthroughput screen. This catalyst consistently gave
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
complete reaction in 2 h at 90 °C compared to variable completion after 18 h at 90 °C for Pd(PPh ) .
3
4
However, when we ran a useꢀtest of the reagents intended for use in the pilot plant campaign no
reaction was observed after heating overnight. Repeating the reaction with other batches of substrate
9
/base/boronic acid 10/catalyst that were known to work from lab trials showed that both the catalyst
and the boronic acid 10 intended for use in the pilot plant did not perform well in the reaction; the
boronic acid gave only 28% conversion after heating overnight, and the catalyst proved unreactive.
Since the catalyst is readily available, we sourced an alternative batch which did work in the useꢀtest.
The boronic acid 10 was more problematic as we were limited by supply of this material. A thorough
1
1
suite of analytical tests (including B NMR, residual pH, and sulfated ash) were carried out on the
boronic acid 10 in an effort to identify why the reaction was proceeding poorly, however these tests did
not show any differences between this batch of boronic acid and other batches which worked well in the
coupling (note that all batches were from a single supplier). Some key experiments suggested that the
boronic acid 10 contained a low level of a stiochiometric poison, and that once this had been consumed,
a standard reaction profile could be achieved. For example, charging 10 mol% of catalyst instead of the
standard 2.5 mol% pushed the reaction to completion in 3 h. Similarly, using a 50:50 mix of ‘good’ and
‘bad’ boronic acid gave complete reaction after heating overnight with the same batch of catalyst at 2.5
mol% loading. We therefore decided to see if we could devise a process which would be able to cope
with any batch of boronic acid used.
1
3
No debromination and no protodeboronation was observed so we hypothesized that deactivation of
the catalyst was occurring prior to oxidative insertion. This is probably not surprising given that all of
8
ACS Paragon Plus Environment

Page 9 of 23
Organic Process Research & Development
the reagents are added together and then the reaction is heated to 90 °C. Our aim was to ensure that the
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
II
0
catalyst was activated (i.e. converted from Pd to Pd ) and oxidative insertion had been initiated by the
time boronic acid 10 is added. Additionally, dosing of 10 should limit the amount of poison that is
present in the initial stages of the reaction, allowing the boronic acid to react as it is added thereby
ensuring completion of the reaction prior to catalyst deactivation. The addition mode was therefore
changed so that the catalyst and aryl bromide 9 were heated to 90 °C in dimethyl acetamide (DMAc)
and held at this temperature for 1 h. In a separate vessel the boronic acid 10 and base were dissolved in
water and DMAc and this solution was stirred for 30 min under nitrogen. This was then added to the
reaction. Running the reaction in this way gave complete conversion 2 h after the addition of the boronic
acid 10 was complete.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We also carried out an experiment to evaluate the stability of the boronic acid and base solution and
actually found that the reaction worked better if this solution was stirred at room temperature for 2 h
prior to addition. Conversely, using the boronic acid solution immediately after preparation resulted in
incomplete reaction. This may be deactivation of the catalyst poison during this time, or hydrolysis of
14
15
boroxines/anhydrides to the more active monomeric or boronate species, however, we did not
examine this any further. To summarise, the Suzuki coupling now consistently goes to completion if the
substrate 9 and catalyst are preꢀheated to 90 °C and the boronic acid 10 is dosed in after preꢀstirring for
2
h with aqueous base. This reaction was then scaled up in the pilot plant on 7.3 kg scale and was
complete in 2 h giving an excellent 92 % yield of aldehyde 11.
The initial development work on the reductive amination of 11 with sodium cyanoborohydride had
shown that a low reaction pH was required to stop addition of the product 12 into the intermediate imine
18 resulting in impurity 19. However, at low pH, HCN was produced as a sideꢀproduct. On smaller
scale this was manageable, but we were required to change the conditions for this campaign since
scrubbing of HCN from hydrogenꢀventing vessels is difficult on pilot plant scale.
9
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 23
H
i) MeNH in EtOH,
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
N
H
O
NaCNBH , HCl (g)
3
MeOH/THF
ii) Aq. HCl
N
O
CHO
N
F
H
N
F
N
H
11
1
8
H
N
O
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
.
HCl
N
N
F
H
H
12
H
H
H
N
N
N
O
O
O
N
N
F
OH
N
H
N
F
F
H
H
Impurity 20
Impurity 19
Scheme 4. Reductive amination of 11 and associated impurities.
Gratifyingly, we found that sodium borohydride could be used as a direct replacement for sodium
1
6
cyanoborohydride if the HCl is omitted. Although this was a promising hit, we did observe higher
levels of alcohol 20 (from reduction of the aldehyde 11) at larger lab scales. Since this presented a
robustness risk, and 20 would constitute a new unqualified impurity if still present in the API, this was
unacceptable. Further work showed that the intermediate imine 18 is insoluble in methanol and can be
isolated in very good yield after removal of THF, thus purging any residual aldehyde 11. The clean
imine 18 can then be redissolved in methanol and THF and reduced with sodium borohydride as before
to give the hydrochloride salt 12. In the previous campaign the aqueous solution of this HCl salt was
treated with activated carbon to reduce the palladium to acceptable levels (<20 ppm), so this was
incorporated into later experiments when we were assessing the quality of 12 produced by this new
route.
Due to a lack of material, development work for the new (S)ꢀcamphorsulfonate salt 21 was carried out
10
ACS Paragon Plus Environment

Page 11 of 23
Organic Process Research & Development
on 1 that had been obtained from neutralization of the phosphate salt 13 prepared in the first GMP
campaign. It was acknowledged that this material would have a different impurity profile compared to 1
prepared via the new route, but this was unavoidable at this stage of development. From a combination
of highꢀthroughput screening and small scale lab experiments, a process was developed wherein an
isopropanol (IPA) solution of (S)ꢀcamphorsulfonic acid was added to a solution of free base 1 in IPA.
Cooling and addition of water then resulted in crystallization of 21 with the desired purity and properties
for tablet formulation. When we switched our source of 1 to free base derived from the new route the
solution turned brown upon addition of (S)ꢀcamphorsulfonic acid, the salt 21 was very reluctant to
crystallize and was obtained in very low yield. An extensive set of troubleshooting experiments
identified the cause; samples of 1 that had not been carbon treated during the reductive amination
performed very poorly in the crystallization, whereas samples which had been carbon treated gave (S)ꢀ
CSA salt 21 of the desired purity. Fortunately, this was discovered shortly before a planned large labꢀ
scale run to demonstrate the conditions for transfer to the pilot plant. We therefore ensured that the
workꢀup incorporated the activated carbon treatment, although a suite of analytical tests did not reveal
the nature of the impurity or impurities that were causing the inhibition of the crystallisation in this step.
When used on pilot plant scale this afforded 6.4 kg of product 12 (76% yield; 12 ppm Pd) which
performed well in the salt formation. The remaining steps were carried out uneventfully in the pilot
plant to deliver 7.1 kg of 21 in 75% yield from Suzuki product 11.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The final modified synthetic route is shown in Scheme 5.
11
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 23
O
O
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
i) 1M HCl
ii) DCM, Toluene
EtO
EtO
N
N
TFA, TES
DCM
O
O
64%
O
O
N
O
OMe
O
15
1
4
43%
CO Me
2
N
F
H
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
H
F
1
6
5
HO
HO
B
CHO
10
H
N
H
H
N
N
i) Pd(dppf)Cl .DCM,
O
O
Pyr.HBr3,
THF, DCM
2
O
DMAc, aq. Na CO
3
2
Aq MeNH2
89%
ii) Methanol reslurry
CHO
Br
N
H
83%
F
N
F
N
92%
F
H
H
8
11
9
H
N
O
H
N
MeNH2
in EtOH
MeOH,
THF
i) NaBH , MeOH/THF
ii) Aq. HCl
iii) Carbon
4
O
.
HCl
N
N
F
H
H
N
H
N
F
76% over 2 steps
1
2
18
H
H
N
N
O
i) Methanol,
Aq. NaOH
ii) THF reslurry
O
(S)-CSA, IPA, water
5%
.CSA
N
N
9
F
79%
N
H
N
F
H
H
H
21
1
Scheme 5. Modified reductive alkylation route to 21.
In conclusion, we have developed an improved, robust process for the synthesis of 21, a novel PARP
inhibitor, and demonstrated the scalability of the route through the manufacture of 7.1 kg of API. Key
highlights include; thorough understanding of the thermal hazards of the LeimgruberꢀBatcho indole
synthesis, installation of the sideꢀchain through a reductive alkylation procedure, improvements in the
12
ACS Paragon Plus Environment

Page 13 of 23
Organic Process Research & Development
robustness of the Suzuki coupling, removal of hydrogen cyanide generation during the reductive
amination and reliable generation of good quality free base for the final salt formation step. The overall
yield for the route has been improved from 1.3% to 2.9%.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Experimental section
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
General. H NMR spectra were recorded on a Varian 400 MHz spectrometer or Jeol ECS 400 MHz
13
spectrometer. C NMR spectra were recorded on a Jeol ECS spectrometer at 100.5 MHz. Melting
points were measured on a Büchi B540 melting point apparatus. Combustion analyses were performed
by Warwick Analytical Service, University of Warwick Science Park, The Venture Centre, Sir William
Lyons Road, Coventry CV4 7EZ, U.K.
5
-Fluoro-2-methyl-3-nitrobenzoic acid methyl ester (3)
To conc. sulfuric acid (349.6 L, 7.6 L/kg) was added 5ꢀfluoroꢀ2ꢀmethylbenzoic acid (46.0 kg, 298.4
mol) and the solution was cooled to 0ꢀ5 °C. Conc. nitric acid (70% w/w aqueous solution, 23.18 L, 0.50
L/kg, 1.26 eq) was added maintaining the reaction temperature below 40 °C (CAUTION; HIGHLY
EXOTHERMIC). The reaction was stirred at 0ꢀ5 °C for 30 min before cautiously adding into water
(920 L, 20 L/kg) that had been cooled to 0ꢀ5 °C, maintaining the temperature below 30 °C. The reaction
mixture was then extracted with methyl tertꢀbutylether (MTBE) (2 × 460 L, 10 L/kg) at 30 °C and the
combined organic layers were washed with water (3 × 460 L, 10 L/kg). The organic layer was then dried
over anhydrous sodium sulfate (11.5 kg, 0.25 kg/kg) and filtered, washing the filter cake with MTBE
(2.5 L/kg, 115 L). The combined filtrate and washings (555 L) were concentrated by distillation under
vacuum at 35ꢀ40 °C to remove 310 L of solvent. This concentrate was taken directly into the next stage.
To methanol (230 L, 5 L/kg) was added the concentrated MTBE solution from the previous step (note:
quantities based on input of 5ꢀfluoroꢀ2ꢀmethylbenzoic acid). Conc. sulfuric acid (14.72 L, 0.32 L/kg)
was added slowly maintaining the temperature below 40 °C. The reaction mixture was then heated to 65
13
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 23
°
C for 24 h. The reaction was cooled to 40ꢀ45 °C and methanol (110 L, 2.5 L/kg) was removed by
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
distillation under vacuum. The reaction was cooled to 30ꢀ35 °C and water (126.5 L, 2.75 L/kg) was
added. The slurry was stirred at 30 °C for 30 mins and then the product was isolated by centrifuge. The
filter cake was washed with water (2 × 96.6 L, 2.1 L/kg). The product was unloaded from the centrifuge
and reslurried in methanol (105.8 L, 2.3 L/kg) at 50 °C for 20 min. Water (52.9 L, 1.19 L/kg) was
added, the slurry was stirred for 1 h at 50 °C and then the product was isolated by centrifuge. The
reaction vessel was rinsed with methanol (23.9 L, 0.52 L/kg) at 0ꢀ5 °C, and this was used to wash the
filter cake. The product was then dried under vacuum at 40 °C to give 3 as a brown solid (32.9 kg, 154.3
mol, 52% yield). M.p. (methanol) 69 °C; δ (400MHz, DMSOꢀd ) 2.39 (s, 3H), 3.84 (s, 3H), 7.85 (dd,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
H
6
1
=
=
H, J = 2.9, 8.6 Hz), 8.06 (dd, 1H, J = 2.9, 8.6 Hz); δ (100MHz, DMSOꢀd ) 14.88, 52.86, 114.37 (d, J
C 6
26.8 Hz), 120.37 (d, J = 23.0 Hz), 127.31, 134.76 (d, J = 7.4 Hz), 151.67 (d, J = 8.3 Hz), 158.93 (d, J
247.9 Hz), 165.33 (d, J = 2.7 Hz).
6
-Fluoro-1H-indole-4-carboxylic acid methyl ester (5)
To DMF (87.5 L, 3.5 L/kg) was added 3 (25.0 kg, 117.3 mol) and dimethylformamide dimethylacetal
(80 L, 3.2 L/kg). Triethylamine (8.25 L, 0.33 L/kg) was then added and the reaction was heated to 95ꢀ
00 °C for 4 h (CAUTION; STRICT TEMPERATURE CONTROL REQUIRED). The reaction
mixture was cooled to 60 °C and 125 L of solvent was removed by distillation under vacuum
CAUTION; STRICT TEMPERATURE AND VOLUME CONTROL REQUIRED). The reaction
1
(
mixture was then cooled to 25ꢀ30 °C and used directly in the next step without further isolation.
Reagent quantities are based on 25 kg (117.3 mol) of 3. To methanol (160 L, 6.3 L/kg) was added the
solution of the enamine 4 in DMF (57.4 kg total weight), sodium acetate (9.65 kg, 117.3 mol, 1.0 eq)
and palladium on carbon (10% Pd on carbon, 50% wet Degussa E101 NE/W, 9.5 kg, 0.38 kg/kg).
Hydrogen was then applied to the vessel and the reaction was stirred at 30 °C for 4 h (CAUTIONꢀ
HIGHLY EXOTHERMIC). The reaction mixture was filtered through Hyflo filter aid to remove the
14
ACS Paragon Plus Environment

Page 15 of 23
Organic Process Research & Development
catalyst and the catalyst bed was washed with methanol (2 × 50 L, 2 L/kg). The solution was stripped to
dryness under vacuum at 45 °C and dissolved in DCM (200 L, 8 L/kg) The DCM solution was then
washed with water (5 × 125 L, 5 L/kg) and dried over sodium sulfate (17.5 kg, 0.7 kg/kg). The solid was
removed by filtration and washed with DCM (2 × 50 L, 2 L/kg) and the washings were combined with
the filtrate. The filtrate was then stripped to dryness under vacuum at 30ꢀ35 °C to give the crude product
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
. The crude product was then purified as follows. Crude 5 was redissolved in DCM (50 L, 2 L/kg) and
purified by column chromatography using 50 kg of silica gel eluting with DCM (337.5 L, 13.5 L/kg).
The product was collected in 4ꢀ5 fractions of 62.5ꢀ70 L each. The fractions were combined and silica
gel (3.75 kg, 0.15 kg/kg) was added. After stirring for 30 min the silica gel was removed by filtration
and the filter was washed with DCM (2 × 25 L, 1 L/kg). The solution was concentrated to 25 L (1 L/kg)
and hexane (37.5 L, 1.5 L/kg) was added at 40 °C over 25 min. The solution was cooled to 30 °C and
stirred for 3 h before collecting the solid by centrifuge and washing with hexane (2 × 25 L, 1 L/kg). The
solid was then dried under vacuum at 45 °C to give 5 as a white crystalline solid (7.30 kg, 37.79 mol,
32% yield). M.p. (hexane) 119 °C; δ (400MHz, DMSOꢀd ) , 3.87 (s, 3H), 6.9 (dd, 1H, J = 0.8, 3.1 Hz),
H
6
7
.43ꢀ7.49 (m, 2H), 7.51 (d, 1H, J = 3.1 Hz), 11.47 (s, br, 1H); δ (100MHz, DMSOꢀd ) 51.85, 102.05,
C 6
1
02.78 (d, J = 25.8 Hz), 109.49 (d, J = 25.8 Hz), 120.79 (d, J = 9.2 Hz), 123.96, 128.49 (d, J = 3.3 Hz),
136.73 (d, J = 12.4 Hz), 157.33 (J = 234.4 Hz), 166.21 (J = 3.2 Hz); Anal. Calcd for C H FNO
2
1
0
8
requires: C 62.18; H 4.17; F 9.83; N 7.25. Found: C 62.08; H 4.21; F 9.85; N 7.27.
Phthalimidoacetaldehyde (15)
1
4 (45.0 kg, 170.9 mol) was added to aqueous hydrochloric acid (1M, 200 L, 4.45 L/kg) and heated at
7
0ꢀ75 °C for 1 h. Water (225 L, 5 L/kg) was then added and the reaction was cooled to 20 °C and stirred
for 12 h. Dichloromethane (450 L, 10 L/kg) was added to the reaction and the phases were separated.
The aqueous layer was extracted with a further portion of dichloromethane (112.5 L, 2.5 L/kg) and the
layers were separated. The dichloromethane layers were then combined. The solvent was removed by
15
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 23
distillation and replaced with toluene to give a final solvent level of 5 L/kg and the resulting slurry was
stirred at 20 °C for 8 h. The product was filtered under pressure, washed with toluene (45 L, 1 L/kg) and
dried under vacuum at 50 °C to give 15 as a white solid (20.2 kg, 107 mol, 62% yield). M.p. (toluene)
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
16 °C; δ (400MHz, DMSOꢀd ) 4.59 (s, 2H), 7.83ꢀ7.90 (m, 4H), 9.56 (s, 1H); δ (100MHz, DMSOꢀd )
H 6 C 6
4
7.29, 123.27, 131.42, 134.67, 167.23, 196.68.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
-Ethyl-(2-phthalimido)-6-Fluoro-1H-indole-4-carboxylic acid methyl ester (16)
To a solution of 15 (20 kg, 106 mol) and 5 (15.7 kg, 81.3 mol) in dichloromethane (157 L, 10 L/kg),
was added triethylsilane (37.8 kg, 325.2 mol, 4 eq). A solution of trifluoroacetic acid (18.5 kg, 12.2 L, 2
eq) in dichloromethane (31.4 L, 2 L/kg) was then added over 30 min maintaining the temperature below
30 °C. The reaction was then stirred for 28 h at 20 °C and the resultant slurry was filtered under
pressure. The filter cake was washed with dichloromethane (7.9 L, 0.5 L/kg) and then dried under
vacuum at 40 °C to give 16 as a pale yellow solid (12.9 kg, 3.51 mol, 43% yield). M.p.
(
dichloromethane) 216 °C; δ (400MHz, DMSOꢀd ) 3.16 (t, 2 H, J=6.96 Hz) 3.76 (t, 2 H, J=6.96 Hz)
H 6
3
.92 (s, 3 H) 7.27 (d, br, 1 H, J=2.69 Hz) 7.29 (dd, 1 H, J=10.26, 2.44 Hz,) 7.38 (dd, 1 H, J=9.28, 2.44
Hz) 7.78 ꢀ 7.82 (m, 4 H) 11.27 (s, br, 1 H); δ (100MHz, DMSOꢀd ) 26.14, 52.72, 102.31 (d, J = 24.9
C
6
Hz), 109.93 (d, J = 25.8 Hz), 112.05, 121.58, 123.44, 124.24 (d, J = 8.8 Hz), 127.96 (d, J = 3.2 Hz),
32.05, 134.81, 138.41 (d, J = 12.0 Hz), 157.5 (d, J = 235 Hz), 167.6 (d, J = 2.8 Hz), 168.19.
1
8
-Fluoro-1,3,4,5-tetrahydro-azepino[5,4,3-cd]indol-6-one (8)
1
6 (12.9 kg, 35.1 mol) was added to methylamine (40% solution in water, 90 L, 7 L/kg) and the slurry
was stirred at 20 °C for 16 h. Water (90 L, 7 L/kg) was charged to the reaction and the mixture was
stirred for 1 h. The solid was then filtered under pressure and washed with water (26 L, 2 L/kg). The
solid was then reslurried in water (129 L, 10 L/kg), filtered, washed with water and dried under vacuum
at 70 °C to give the product (8) as a tan brown solid (6.42 kg, 31.4 mol, 89% yield); M.p. (water) 187
16
ACS Paragon Plus Environment

Page 17 of 23
Organic Process Research & Development
°
C; δ (400MHz, DMSOꢀd ) 2.86ꢀ2.88 (m, 2H), 3.37 (q, 2H, J = 5.6 Hz), 7.22 (s, br, 1H), 7.30 (dd, 1H,
H 6
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
J = 2.4, 9.3 Hz), 7.37 (dd, 1H, J = 2.4, 9.3 Hz), 8.11 (t, br, J = 5.6 Hz), 11.13 (s, br, 1H); δ (100MHz,
C
DMSOꢀd ) 27.91, 42.06, 100.82 (d, J = 25.8 Hz), 109.12 (d, J = 25.8 Hz), 114.40, 121.87, 123.72 (d, J
6
=
3.3 Hz), 125.48 (d, J = 8.7 Hz), 136.71 (d, J = 12.4 Hz), 158.21 (d, J = 233.6 Hz), 168.39.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
-Bromo-8-fluoro-1,3,4,5-tetrahydro-azepino[5,4,3-cd]indol-6-one (9)
Dichloromethane (96.3 L, 15 L/kg) and THF (96.3 L, 15 L/kg) were mixed together and 8 (6.42 kg,
31.4 mol) was added. After stirring for 20 min, the slurry was cooled to 0ꢀ5 °C. Pyridinium tribromide
11.0 kg, 34.5 mol, 1.1 eq) was then added and the solution was stirred at 0ꢀ5 °C for 1 h. Water (64.2 L,
0 L/kg) was then charged and the organic solvents were removed by distillation to give a final volume
(
1
of 15 L/kg. THF was added (64.2 L, 10 L/kg) and the reaction mixture was added slowly to a solution of
saturated aqueous sodium carbonate (481.5 L, 75 L/kg). The resultant slurry was then stirred overnight
at 20 °C. The product was filtered under pressure, washed with water (64.2 L, 10 L/kg) and dried under
vacuum at 55 °C to give 9 as a beige solid (7.42 kg, 26.2 mol, 83% yield); M.p. (water) 215 °C
(decomposition); δ (400MHz, DMSOꢀd ) 2.73ꢀ2.76 (m, 2H), 3.38 (q, 2H, J = 5.6 Hz), 7.25 (dd, 1H, J
H 6
= 2.4, 9.0 Hz), 7.40 (dd, 1H, J = 2.4, 9.0 Hz), 8.18 (t, br, J = 5.6 Hz), 12.05 (s, br, 1H); δ (100MHz,
C
DMSOꢀd ) 27.70, 41.22, 100.49 (d, J = 26.2 Hz), 109.08 (d, J = 2.8 Hz), 109.85 (d, J = 25.3 Hz),
6
1
14.04, 121.78, 124.88 (d, J = 8.8 Hz), 136.53 (d, J = 12.5 Hz), 158.30 (d, J = 235.0 Hz), 167.87; Anal.
Calcd for C H BrFN O: C 46.67; H 2.85; Br 28.22; F 6.71; N 9.90. Found: C 46.81; H 2.82; Br 28.00;
1
1
8
2
F 6.64; N 9.82.
4-(8-Fluoro-6-oxo-3,4,5,6-tetrahydro-1H-azepino[5,4,3-cd]indol-2-yl)-benzaldehyde (11)
,1′ꢀBis(diphenylphosphino)ferrocene palladium (II) dichloride dichloromethane (0.539 kg, 0.66 mol,
.5 mol%) was added to a mixture of degassed dimethylacetamide (119 L, 16 L/kg) and 9 (7.41 kg, 26.2
mol) and stirred for 1 h at 20 °C. The mixture was then heated to 95 °C and stirred for 1 h. In a separate
1
2
17
ACS Paragon Plus Environment

Organic Process Research & Development
Page 18 of 23
vessel, 4ꢀformylbenzene boronic acid (4.71 kg, 31.4 mol, 1.2 eq) was dissolved in dimethylacetamide
30 L, 4 L/kg) before adding to an aqueous solution of sodium carbonate (5.6 kg in 89 L of water, 2 eq)
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
and was then stirred for 3 h at 20 °C. The boronic acid solution was then added to the catalyst and
substrate solution whilst maintaining the temperature above 90 °C. After stirring for 2 h the reaction was
cooled to room temperature and stirred for 4 h. Water (296 L, 40 L/kg) was added and the slurry was
stirred for 1 h before filtering the solid under pressure and washing with water (74 L, 10 L/kg) to give
crude 11. The product 11 was further purified by resuspending in methanol (59L, 8 L/kg) and stirring at
60 °C for 1 h. The slurry was cooled to 20 °C and stirred for 1 h. The product was then filtered, washed
with methanol (15 L, 2 L/kg) and dried under vacuum to give 11 as an olive green solid (7.44 kg, 24.1
mol, 92% yield); M.p. (methanol) 219 °C (decomposition); δ (400MHz, DMSOꢀd ) 3.10 (s, br, 2 H),
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
H
6
3.41 (s, br, 2 H), 7.37 (d, 1 H, J = 7.82 Hz) 7.46 (d, 1 H, J = 10.01 Hz), 7.86 (d, 2 H, J = 6.84 Hz), 8.05
(
d, 2 H, J = 7.08 Hz), 8.28 (s, br, 1 H), 10.06 (s, br, 1 H) 11.88 (s, br, 1 H); δ (100MHz, DMSOꢀd )
C 6
2
8.89, 41.71, 100.75, (J = 25.8 Hz), 110.13 (J = 25.8 Hz), 114.23, 122.95, 126.55 (J = 8.3 Hz), 128.14,
29.83, 133.90, 134.75, 137.20, 158.88 (J = 235.9 Hz), 168.12, 192.49.
1
8-Fluoro-2-(4-methylaminomethyl-phenyl)-1,3,4,5-tetrahydro-azepino[5,4,3-cd]indol-6-one
hydrochloride salt (12)
To a mixture of methanol (313 L, 43 L/kg) and THF (153 L, 21 L/kg) was added 11 (7.30 kg, 23.7
mol). A solution of methylamine was then added (8 M in ethanol, 2.0 eq, 4.44 kg) and the mixture was
stirred for 2 h. The reaction mixture was then concentrated by distillation under atmospheric pressure to
approximately 73 L and the solvent was replaced with methanol to give a final solvent volume of 73 L
(10 L/kg). The resultant slurry was cooled to 20 °C and stirred for 4 h. The solid was filtered under
nitrogen pressure, washed with methanol (37 L, 5 L/kg) and dried under vacuum at 45 °C to give 18 as a
solid (6.92 kg, 21.5 mol, 91% yield) which was used directly in the next stage. To a mixture of methanol
(298 L, 43 L/kg) and THF (145 L, 21 L/kg) was added 18 (6.92 kg, 21.5 mol) and the slurry was cooled
18
ACS Paragon Plus Environment

Page 19 of 23
Organic Process Research & Development
to 0ꢀ5 °C. Sodium borohydride (3.0 eq, 2.44 kg, 64.59 mol) was then added and the reaction mixture
was stirred at 0ꢀ5 °C for 2 h. The reaction was warmed to 20 °C over 1 h and then added to a mixture of
methanol (69 L, 10 L/kg), water (111 L, 16 L/kg) and conc. hydrochloric acid (27.7 L, 4 L/kg) over 1 h.
Methanol (35 L, 5 L/kg) was then added as a line wash. Activated carbon (Norit SX plus, 3.46 kg, 0.5
kg/kg) was added to the reaction mixture as a slurry in methanol (45 L, 6.5 L/kg) and the slurry was
stirred for 18 h. The mixture was filtered and the filter cake was washed with water (28 L, 4 L/kg), THF
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(43 L, 6L/kg) and methanol (69 L, 10 L/kg). The filtrate was concentrated by vacuum distillation to a
level of 130 L and the resultant slurry was filtered. The filter cake was washed with water (28 L, 4 L/kg)
and dried under vacuum at 50 °C to give 12 as an offꢀwhite solid (6.44 kg, 17.90 mol, 76% yield). M.p.
(
water) 293 °C; δ (400MHz, DMSOꢀd ) 2.52 (t, br, 3H, J = 4.6 Hz), 3.0ꢀ3.02 (m, 2H), 3.34ꢀ3.36 (m,
H 6
2H), 4.12 (t, br, 2H, J = 5.0 Hz), 7.32 (dd, 1H, J = 2.4, 9.3 Hz), 7.40 (dd, 1H, J = 2.4, 11.0 Hz), 7.65 (s,
4
4
1
H), 8.22 (t, br, J = 5.7 Hz), 9.37 (s, br, 1H), 11.83 (s, 1H); δ (100MHz, DMSOꢀd ) 28.74, 31.94,
C 6
1.77, 50.80, 100.63 (d, J = 25.8 Hz), 109.67 (d, J = 25.8 Hz), 112.33, 123.03, 126.00 (d, J = 8.8 Hz),
27.90, 130.28, 131.32, 132.10, 134.59, 136.83 (d, J = 12.4 Hz), 158.48 (d, J = 234 Hz), 168.28.
8-Fluoro-2-(4-methylaminomethyl-phenyl)-1,3,4,5-tetrahydro-azepino[5,4,3-cd]indol-6-one (1)
To a solution of aqueous sodium hydroxide (40% w/w, 3.6 kg, 2.0 eq) in water (88 L, 14 L/kg) and
methanol (35 L, 5.5 L/kg) was added 12 (6.44 kg, 17.90 mol). Water (90 L, 14 L/kg) was then added
and the slurry was stirred for 18 h at 20 °C. The solid was filtered, washed with water (2 × 32 L, 5 L/kg)
and dried under vacuum at 50 °C to give 1 as an offꢀwhite solid (5.40 kg, 16.70 mol, 93% yield). The
product (5.40 kg, 16.70 mol ) was further purified by reslurrying in THF (54 L, 10 L/kg) and passing
through a high shear wet mill for 1.5 h. The slurry was then heated to 60 °C for 12 h. After cooling to 20
°
C the product was filtered under pressure, washed with THF (16.2 L, 3 L/kg) and dried at 45 °C under
vacuum to give 1 as a 1:1 THF solvate (5.57 kg, 14.08 mol, 84% yield). M.p. (THF) decomposed at 220
C; δ (400MHz, DMSOꢀd ) 2.25 (s, 3H), 2.99ꢀ3.01 (m 2H), 3.65 (s, 2H), 7.27 (dd, 1H, J = 2.4, 9.3
°
H
6
19
ACS Paragon Plus Environment

Organic Process Research & Development
Page 20 of 23
Hz), 7.39 (dd, 1H, J = 2.4, 9.3 Hz), 7.42 (d, 2H, J = 8.5 Hz), 7.53 (d, 2H, J = 8.3 Hz), 8.18 (t, br, 1H, J =
.7 Hz), 11.60 (s, 1H); δ (100MHz, DMSOꢀd ) 28.74, 35.58, 41.84, 54.74, 100.47 (d, J = 25.8 Hz),
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
C
6
109.44 (d, J = 25.8 Hz), 111.47, 123.19, 125.72 (d, J = 8.8 Hz), 127.55, 128.20, 129.86, 135.38 (d, J =
.7 Hz), 136.67 (d, J = 12.4 Hz), 140.52, 158.31 (d, J = 233), 168.39.
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
-Fluoro-2-(4-methylaminomethyl-phenyl)-1,3,4,5-tetrahydro-azepino[5,4,3-cd]indol-6-one (S)-
camphorsulfonate salt (21)
To a slurry of 1 (5.32 kg, 13.48 mol) in isopropanol (30 L, 5.5 L/kg) and water (39 L, 7.3 L/kg) was
added a solution of (S)ꢀcamphorsulfonic acid (3.75 kg, 16.18 mol, 1.2 eq) in water (10.6 L, 2 L/kg). The
resultant slurry was then heated to 70 °C and held for 1 h to ensure dissolution. The solution was filtered
to remove particulates and then a line wash of water (3.7 L, 0.7 L/kg) and isopropanol (2.1 L, 0.4 L/kg)
was added. The solution was cooled to 40 °C, seeded with 21 (0.266 kg, 0.05 eq) and then held at 40 °C
for 8 h. The slurry was then cooled to 20 °C, water was added (4 × 13.3 L, 10 L/kg total) and the slurry
was cooled further to 0ꢀ5 °C and stirred for 12 h. The product was filtered under pressure, washed with
water (10.6 L, 2 L/kg) and dried under vacuum at 50 °C to give 21 as a white crystalline solid (7.09 kg,
12.76 mol, 95% yield). M.p. (IPA/water) 303 °C; δ (400MHz, DMSOꢀd ) 0.74 (s, 3H), 1.05 (s, 3H),
H
6
1
.28 (m, 1H), 1.80 (d, 1H, J = 18.0 Hz), 1.81ꢀ1.88 (m, 1H), 1.93 (app t, 1H, J = 4.5 Hz), 2.24 (m, 1H),
.41 (d, 1H, J = 14.6 Hz), 2.62 (s, 3H), 2.66ꢀ2.72 (m, 1H), 2.91 (d, 1H, J = 14.7 Hz), 3.04ꢀ3.07 (m, br,
2
2H), 3.36ꢀ3.45 (m, br, 2H), 4.20 (s, 2H), 7.37 (dd, 1H, J = 2.4, 9.3 Hz), 7.44 (dd, 1H, J = 2.4, 11.0 Hz),
.63 (d, 2H, J = 8.3 Hz), 7.71 (d, 2H, J = 8.3 Hz), 8.26 (t, br, 1H, J = 5.5 Hz), 11.76 (s, 1H); δC
100MHz, DMSOꢀd ) 19.51, 20.02, 24.14, 26.37, 28.74, 32.28, 41.77, 42.13, 42.22, 46.71, 47.00, 51.06,
7
(
6
5
1
8.21, 100.65 (d, J = 25.8 Hz), 109.72 (d, J = 25.8 Hz), 112.41, 123.03, 126.04 (d, J = 8.7 Hz), 127.98,
30.19, 131.22, 132.22, 134.50, 136.83 (d, J = 12.0 Hz), 158.52 (d, J = 235 Hz), 168.27, 216.24.
Acknowledgements The authors thank Julian Meekings, Onyechi Obonyo, Kevin Meldrum, Dean
20
ACS Paragon Plus Environment

Page 21 of 23
Organic Process Research & Development
Brick and Patricia Basford for valuable contributions to the project.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
References
1
CananꢀKoch, S. S.; Thoresen, L. H.; Tikhe, J. G.; Maegley, K. A.; Almassy, R. J.; Li, J.; Yu, X.ꢀH.;
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Zook, S. E.; Kumpf, R. A.; Zhang, C.; Boritzki, T. J.; Mansour, R. N.; Zhang, K. E.; Ekker, A.;
Calabrese, C. R.; Curtin, N. J.; Kyle, S.; Thomas, H. D.; Wang, L.ꢀZ.; Calvert, A. H.; Golding, B. T.;
Griffin, R. J.; Newell, D. R.; Webber, S. E.; Hostomsky, Z. J. Med. Chem. 2002, 45(23), 4961ꢀ4974.
2
Thomas, H. D.; Calabrese, C. R.; Batey, M. A.; Canan, S.; Hostomsky, Z.; Kyle, S.; Maegley, K. A.;
Newell, D. R.; Skalitzky, D.; Wang, L.ꢀZ.; Webber, S. E.; Curtin, N. J. Molecular Cancer Therapeutics,
2007, 6(3), 945ꢀ956; Skalitzky, D. J.; Marakovits, J. T.; Maegley, K. A.; Ekker, A.; Yu, X.ꢀH.;
Hostomsky, Z.; Webber, S. E.; Eastman, B. W.; Almassy, R.; Li, J.; Curtin, N. J.; Newell, D. R.;
Calvert, A. H.; Griffin, R. J.; Golding, B. T. J. Med. Chem. 2003, 46(2), 210ꢀ213.
3
CananꢀKoch, S. S.; Chu, J.ꢀJ.; Liu, J.; Matthews, J. J. PCT Int. Appl. WO 2004087713 A1; Liu, J.;
Nayyar, N.; Guo, M.; Wu, Z.ꢀP.; Borer, B. C.; Srirangam, A. N.; Mitchell, M. B.; Li, Y.; Chu, J.ꢀJ.
PCT Int. Appl. WO 2006033007 A2; Webber, S. E.; CananꢀKoch, S. S.; Tikhe, J.; Thoresen, L. H.
PCT Int. Appl. WO 2000042040 A1.
4
Plummer, R.; Jones, C.; Middleton, M.; Wilson, R.; Evans, J.; Olsen, A.; Curtin, N.; Boddy, A.;
McHugh, P.; Newell, D.; Harris, A.; Johnson, P.; Steinfeldt, H.; Dewji, R.; Wang, D.; Robson, L.;
Calvert, H. Clinical Cancer Research, 2008, 14(23), 7917ꢀ7923; Wang, Y.; Frederick, M. D.; Drugs of
the Future 2009, 34(3), 177ꢀ182; Mukhopadhyay, A.; Elattar, A.; Cerbinskaite, A.; Wilkinson, S. J.;
Drew, Y.; Kyle, S.; Los, G.; Hostomsky, Z.; Edmondson, R. J.; Curtin, N. J. Clinical Cancer Research
2010, 16(8), 2344ꢀ2351.
5
Bredereck, H.; Simchen, G.; Rebsdat, S.; Kantlehner, W.; Horn, P.; Wahl, R.; Hoffmann, H.;
Grieshaber, P. Chem.Ber. 1968, 101, 41; Rosso, G. B. Synlett 2006, 5, 809ꢀ810.
21
ACS Paragon Plus Environment

Organic Process Research & Development
Page 22 of 23
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
Kantlehner, Willi, e-EROS Encyclopedia of Reagents for Organic Synthesis 2001, no page number
given.
7
8
Further safety evaluation of this process is advised prior to further scaleꢀup.
Differential Scanning Calorimety (DSC) carried out on DMANE detected an exotherm from 240°C
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
with ꢁH = ꢀ2.8kJ/g. Thermal Stability Unit (TSU) testing showed that a mixture of trifluoroacetic acid
and DMANE in DCM undergoes an exothermic runaway from 50 °C with the temperature rising to 220
°
C within 2 min. The pressure also increased from 1 bar to 60 bar in under 30 seconds. DSC tests are
carried out in 40ꢂL gold HP test cells and heated at 5K/min from 25°C to 400°C. TSU tests are carried
out in glass test cells and heated at 2K/min from 25°C to 250°C.
9
Electrophiles included 2ꢀaminoacetaldehyde dimethylacetal, 2ꢀaminoethylchloride hydrochloride, 2ꢀ
aminoethylhydrogensulfate, NꢀBocꢀ2ꢀaminoethyl bromide, 2ꢀchloroethyl tosylate, and 2ꢀ
phthalimidoethyl bromide
1
0 Li, W.; Li, J.; DeVincentis, D.; Mansour, T. S.; Tetrahedron, 2008, 64, 7871.
1
1 The aldehyde could be easily prepared by deprotection of the commercially available diethyl acetal
with aqueous hydrochloric acid. Some care was required in isolating the aldehyde however, as it
appeared to undergo facile dimerisation/trimerisation to give a very insoluble solid. This mixture could
be converted back to the monomer by heating in toluene at 80 °C for 3 hours.
12 Ennis, D. S.; McManus, J.; WoodꢀKaczmar, W.; Richardson, J.; Smith, G. E.; Carstairs, A. Org.
Process Res. Dev. 1999, 3, 248.
1
3 Butters, M.; Harvey, J. N.; Jover, J.; Lennox, A. J. J.; LloydꢀJones, G. C.; Murray, P. M. Angew.
Chem. Int. Ed. 2010, 49, 5156 –5160.
14 AntoftꢀFinch, A.; Blackburn, T.; Snieckus, V. J. Am. Chem. Soc. 2009, 131(49), 17750ꢀ17752.
22
ACS Paragon Plus Environment

Page 23 of 23
Organic Process Research & Development
1
2
3
4
1
5 Cammidge, A. N.; Goddard, V. H. M.; Gopee, H.; Harrison, N. L.; Hughes, D. L.; Schubert C. J.;
Sutton, B. M.; Watts G. L.; Whitehead, A. J. Org. Lett., 2006, 8, 4071.
16 Lane, Clinton F. Synthesis 1975, 3, 135ꢀ46.
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
23
ACS Paragon Plus Environment