An Alternative Indazole Synthesis for Merestinib

Article abstract of DOI:10.1021/acs.oprd.8b00016

A new synthesis of a key indazole-containing building block for the MET kinase inhibitor merestinib was designed and demonstrated. Crucial to the successful construction of the challenging indazole is an S_NAr reaction, which forges the heterocy

Full text of DOI:10.1021/acs.oprd.8b00016

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Communication
An Alternative Indazole Synthesis for Merestinib
Yu Lu, Kevin Paul Cole, Jared Fennell, Todd Maloney,
David Mitchell, Ramesh Subbiah, and Balakumar Ramadas
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.8b00016 • Publication Date (Web): 21 Feb 2018
Downloaded from http://pubs.acs.org on February 21, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
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 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
An Alternative Indazole Synthesis for Merestinib
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
†
§
Yu Lu,* Kevin P. Cole, Jared W. Fennell, Todd D. Maloney, David Mitchell, Ramesh Subbiah,
§
and Balakumar Ramadas
Small Molecule Design and Development, Lilly Research Laboratories, Eli Lilly and Company,
Indianapolis, Indiana 46285
§
Syngene International Ltd., Biocon Park, Plot No. 2 & 3, Bommasandra IV Phase, Jigani Link
Road, Bangalore 560 099, India
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 33
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
Table of Contents Graphic
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
ACS Paragon Plus Environment
2

Page 3 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
ABSTRACT. A new synthesis of a key indazoleꢀcontaining building block for the MET kinase
inhibitor merestinib was designed and demonstrated. Crucial to the successful construction of
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 challenging indazole is an S Ar reaction, which forges the heterocyclic ring. Continuous
N
processing was applied to two of the five steps: nitration of a benzaldehyde and high temperature
hydrolysis of an aniline to phenol. When compared to a highlyꢀdeveloped historical route, the
new route shows clear benefit in terms of product quality and potentially manufacturability and
robustness.
Keywords: merestinib, indazole, nitration, S Ar, continuous flow, thermohydrolysis
N
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 33
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
INTRODUCTION
1
Merestinib (Figure 1, 1) is a MET inhibitor currently in clinical trials as therapy for a
2
3
number of oncolytic indications. Marketed MET inhibitors include crizotinib (Xalkori®) and
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
4
cabozantinib (Carbometyx®). They were the first MET inhibitors approved by the U.S. Food
and Drug Administration (FDA). Crizotinib was approved in 2011 to treat lateꢀstage nonꢀsmall
cell lung (NSCL) cancers. Also, in 2011, carbozantinib was approved by the FDA as an orphan
drug to treat medullary thyroid cancer. Both MET inhibitors are currently being evaluated for
5
other types of cancers.
The current synthesis approach for 1 involved a convergent approach in which three
fragments (2, 3, and 4) are united. The pieces can be assembled using Suzuki crossꢀcoupling,
nitro group reduction, amide bond formation, and unmasking of the pyrazole moiety; thus far in
development two different step orderings have been employed for API manufacturing. Indazole
6
fragment 2 has been a constant intermediate, and has served as a key building block. The
complexity of 2 presents several daunting synthetic challenges: the 5,6ꢀsubstitution pattern of the
methylindazole ring (which can also be viewed as a tetrasubstituted benzene derivative),
electronic activation requirements for indazole formation, and the Nꢀmethyl substitution of the
heterocycle which introduces the possibility of N1 vs. N2 regioisomerism in any methylation
approach.
Figure 1. Retrosynthesis of Merestinib
ACS Paragon Plus Environment
4

Page 5 of 33
Organic Process Research & Development
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
A number of synthesis routes have been evaluated for the preparation of 2. Prior to this
work, the most recently developed approach involved a four step transformation utilizing Cuꢀ
7
catalyzed cyclization of a methyl hydrazone to form the indazole ring (Scheme 1). The
substitution pattern is first set by selective dibromination of 3ꢀhydroxybenzaldehyde (5). Phenol
6
was reacted with 1,2ꢀdifluoroꢀ4ꢀnitrobenzene 7 in an S Ar reaction to form diaryl ether 8.
N
Hydrazone 9 was then obtained by reacting aldehyde 8 with aqueous methylhydrazine. Indazole
formation finally provides 2 by means of an intramolecular Cuꢀcatalyzed Ullmannꢀlike ring
8
closure between the hydrazone and aryl bromide.
Scheme 1. Previous Synthesis of Indazole 2
On multiꢀkilogram scale, the first three steps have proven robust and typically provide
5
0–55% yield across the sequence. However, even after several rounds of development, the Cuꢀ
catalyzed step continues to present challenges at plant scale. Specific problems with this
synthetic route in the pilot plant were: 1) Use of dichloromethane and bromine for the
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 33
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
bromination step proved optimal for reactivity and regioselectivity. For environmental reasons,
dichloromethane is considered highly undesirable at Eli Lilly for long term manufacturing use.
2) For the Ullmann indazole ring closure, a rigorously oxygenꢀfree atmosphere was necessary to
avoid low product recovery. Despite our best efforts this step still proved to exhibit substantial
variability in terms of performance, impurity profile, product quality, and yield (37–57%). 3)
For the Ullmann indazole ring closure, long cycle time during the extensive purification process
limited throughput. The purification consisted of a technical grade crystallization, carbon
treatment, and recrystallization in order to ensure acceptable quality of indazole 2 in terms of
residual copper content, product color, and overall impurity load. Notably, as shown in Figure 2,
at least 50 discrete impurities are formed during the Cuꢀcatalyzed indazole ring formation! In
parallel with further attempts at development of the existing route, these findings caused us to
revisit the route selection strategy for a possible alternative approach to 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
Figure 2. HPLC chromatogram (240 nm) of solid 2 after initial isolation (red trace, top)
compared to blank injection (blue trace, bottom). In this chromatogram, 2 elutes at ~8 min, but
ACS Paragon Plus Environment
6

Page 7 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
at least 50 peaks can be integrated, and the main peak has an approximate area percentage of
8
8.3%.
RESULTS AND DISCUSSION
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
1. Route Design and Selection. Since we chose to revisit the route selection approach
for 2, the data from previously investigated routes was reꢀevaluated. A summary of the route
design is presented in Scheme 2. In addition to the copperꢀenabled synthesis route (Scheme 2,
Equation 1), there were three main additional approaches considered. The key transformation in
each approach are as follows: Equation 2: selective bromination of a preformed indazole ring;
Equation 3: An S Ar approach to the indazole using an electronically unactivated substrate; and
N
Equation 4: An intentionally activated substrate for S Ar formation of the indazole.
N
Scheme 2. Route Design Summary
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 33
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
For the selective bromination investigation, methylindazole 14 was utilized since it is
commercially available to some extent. Selective monoꢀbromination at C6 would provide a
penultimate intermediate 13. Although the most electrophilic site of 14 for halogenation appears
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
2
3
4
5
6
7
8
9
0
to be C4 based on a literature example, our plan was to incorporate dibromination followed by
selective debromination, should monobromination prove unselective. Reaction of 14 with NBS
in THF at 25 °C for 16 h provided 15 instead of the desired regioisomer 13. Attempts at
dibromination were also unsuccessful: when 15 was further reacted with NBS, 3,4ꢀ
dibromoindazole 16 was obtained. A similar outcome was noted when NIS was substituted for
NBS in the reaction with 14. Since no useful C6 halogenation conditions were identified, all
investigations utilizing 14 as a raw material were terminated.
2
. Alternate Indazole Ring Formation through S Ar Reaction. Based on our
N
learnings surrounding the Cuꢀcatalyzed indazole formation step, we focused on alternative means
to close the indazole ring. One approach was to investigate an S Ar reaction for the indazole
N
ring forming step. To this point, such an approach had gone underꢀinvestigated because previous
starting materials had an electronically deactivating ether or phenol functionality in the 5ꢀ
position, such as compounds 9, 10, and 11. In order to enable S Ar reactivity in this system, it
N
was determined that a strongly electronꢀwithdrawing group in the C5ꢀposition was crucial.
Furthermore, these intermediates present a bromine atom as the leaving group, which was also
unfavorable for S Ar reaction. A brief investigation of the S Ar approach with the unactivated
N
N
substrate 17 was undertaken. After formation of hydrazone 18, all attempts at intramolecular
SNAr cyclization to the desired indazole 12 failed (Scheme 2, Equation 3).
In order to perturb the electronics of the aromatic ring, we decided to radically alter the
synthetic approach. Readily available 4ꢀbromoꢀ2ꢀfluorobenzaldehyde (19) was identified as a
ACS Paragon Plus Environment
8

Page 9 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
logical starting point for the activated S Ar investigation. With the bromide, aldehyde and
N
fluoride in place, we focused on nitration at C5 in order to obtain the correct substitution pattern
as well as to introduce S Ar reactivity. Then, the nitro group in the product would serve as an
N
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
activator for S Ar reactions while acting as a masked phenol. It was envisioned that the
N
chemistry needed to enable this synthetic approach would consist of: 1) Aromatic ring nitration;
2
) S Ar reaction to form indazole 20; 3) Conversion of a nitroarene to phenol; and 4) Diaryl
N
ether formation, which had already been demonstrated in a previous synthesis.
A. Aromatic Ring Nitration. Nitration of benzaldehyde 19 was initially performed as a
conventional batch process with KNO in concentrated H SO to provide the nitrobenzene 21
3
2
4
10
according to a known literature method. Although high yield and excellent regioselectivity
were achieved for the isolated product (90% yield, >99.5% desired regioisomer), the portionꢀ
wise addition of KNO to the solution of 19 in H SO was noted to be strongly exothermic.
3
2
4
Even with active cooling, at 100 g scale, an exotherm from 0 to 30 °C occurred in a span of <5
min, which gave us great concern for runaway potential. The exotherm was reduced by
switching from KNO to concentrated HNO (65%, w/w). Upon controlled addition of HNO to
3
3
3
a solution of 19 in concentrated H SO at 0 °C, the reaction demonstrated a subdued exotherm,
2
4
and the internal temperature did not exceed 10 °C. Although differential scanning calorimetry
(DSC) demonstrated the stability of isolated 21 was well maintained at temperatures up to 200
°
C, further mitigation of the potential safety hazard from rapid heat evolution was deemed
necessary, especially when considering scaleꢀup.
Aromatic nitration in continuous flow has been demonstrated on both lab and production
11
scale. Here, continuous processing offers a significant safety benefit compared to batch mode.
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 33
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
Since this nitration was determined to be a rapid reaction, the necessary volume of a continuous
flow reactor would be very small relative to a batch reactor. The small reactor dimensions would
impart superior heat and mass transfer to the process, limiting the potential for thermal runaway,
as well as minimizing the overall amount of material at risk, should a runaway occur. Therefore,
nitration of 19 was investigated in continuous mode. For the continuous reaction, a mixture of
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
HNO /H SO was combined using a Teeꢀmixer with a premixed solution of 19 in H SO and
3
2
4
2
4
introduced into a plug flow reactor (PFR). The PFR (ID = 6 mm, length = 24 m) was
constructed of Polytetrafluoroethylene (PTFE) and immersed in a cooling bath at 3 °C, but the
1
2
temperature inside the reactor was not measured. A calculated residence time (reactor volume /
volumetric flow rate) of about 23 min in the PFR was sufficient to provide >98% reaction
conversion (HPLC area%) with <2 HPLC area% of the undesired 3ꢀnitro regioisomer. Although
the level of the 3ꢀnitro regioisomer was higher than the batch process (typically 0.3–0.8%), it
was well rejected during the crystallization step and posed little impact in the downstream
chemistry. The reactor outflow was collected and quenched by adding it to ice. Extraction with
dichloromethane and solvent exchange into heptane afforded 21 as a crystalline solid in 88%
yield. The majority of the product loss occurred during the extraction and crystallization, which
are not yet optimized. It should be noted that the use of dichloromethane as extraction solvent
was to enable rapid proof of concept and a study to replace it with a more sustainable solvent is
ongoing. The continuous nitration process was demonstrated at 540 g scale, whereas the batch
process was not developed beyond 100 g batch size. Overall, the nitration of 19 to afford 21
proved robust and versatile in that a batch or continuous process may be utilized for
manufacturing. Given the equipment simplicity and inherent safety benefits of continuous flow
nitration, the continuous process is currently in favor for further scale up.
ACS Paragon Plus Environment
10

Page 11 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
B. Indazole Ring Formation via Hydrazone Formation and S Ar Reactions. Following a
N
literature procedure for preparation of 6ꢀbromoꢀ1ꢀisopropylꢀ5ꢀnitroꢀ1Hꢀindazole from 21 and
1
3
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
isopropylhydrazine in N,Nꢀdimethylformamide (DMF) at 150 °C for 5 h, 21 was subjected to
analogous reaction conditions (Table 1). Such a single pot transformation was desired, since it
would avoid isolation of the hydrazone intermediate. Using methylhydrazine instead of
isopropylhydrazine with a reaction time of 2 h in a pressurized autoclave, indazole 20 was
obtained in good isolated yield (64%; Table 1, Entry 1). Optimization of the reaction parameters
included screening of solvent and temperature in an effort to avoid the high temperature
conditions. At lower temperature, the reaction was found to be complete after heating at 50 °C
for 2 h (Table 1, Entry 2) to provide 20 in 70% yield. This result may indicate that the
methylhydrazone intermediate 22 is more activated for the S Ar than was the reported
N
1
3
isopropylhydrazone. Changing the reaction solvent from DMF to a mixture of EtOH/H O (5:1)
2
caused slower cyclization, as a 40:52 ratio of the desired product 20 and hydrazone intermediate
2
2 was observed by HPLC at the end of reaction (Table 1, Entry 3). We postulated that a base
additive would better facilitate the indazole formation by neutralizing the HF produced from
cyclization and maintaining a high pH throughout the process. Additional rate acceleration may
also be possible through deprotonation of the transient hydrazone intermediate, which may serve
as a stronger nucleophile in the intramolecular nucleophilic substitution. Accordingly, addition
of 1.5 equiv Et N with higher reaction temperature resulted in higher conversion (Table 1, Entry
3
4
). Given that potassium carbonate (K CO ) proved to be the optimal base for similar
2 3
1
4
transformations, it was applied to the formation of 20. Indeed, addition of 1.5 equiv K CO
2
3
was found to promote the S Ar reaction, providing 20 as the dominant product by HPLC (Table
N
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 33
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
6
, Entry 5). Further optimization was realized by screening additional solvents (Table 1, Entries
–9). An aqueous mixture of isopropanol (IPA) proved optimal as it gave the highest conversion
rate and HPLC purity of the isolated product (Table 1, Entry 7). After the reaction was
complete, addition of water and isolation of the resulting solid afforded 20 in 76% yield at 500 g
scale (Table 1, Entry 10). It is noteworthy that no evidence of the isomeric 2Hꢀindazole was
observed using this synthetic approach.
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
Table 1. Solvent screen and additive effect on the synthesis of 20 from 21 and MeNHNH2
Scale
g)
Additive
(1.5 eq)
HPLC area% Ratio
(20:22:others)
Entry
Solvent
T (°C)
(
a
1
300
5
DMF
DMF
150
50
none
none
ND
ND
b
2
EtOH:H O
2
3
4
5
6
7
5
2
4
5
5
50
80
80
80
80
none
40:52:8
78:10:12
85:3:12
14:29:57
95:2:3
(
5:1)
EtOH:H O
2
Et N
3
(2:1)
EtOH:H O
2
K CO3
2
(
2:1)
MeOH:H O
2
K CO3
2
(
2:1)
IPA:H O
2
K CO3
2
(2:1)
8
9
5
5
THF
80
80
K CO3
95:0:5
2
2ꢀMeTHF
2
K CO3
80:3:17
c
IPA:H
O
2
10
500
80
K CO3
2
80:1:19
(
2:1)
a
b
reaction was conducted in autoclave, 64% isolated yield, 99.4% HPLC purity; reaction was
c
conducted in a round bottom flask, 70% isolated yield, 97.8% HPLC purity; reaction time was
6
h, 76% isolated yield, 95.8% HPLC purity. ND = not determined.
ACS Paragon Plus Environment
12

Page 13 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
C. Transformation of Nitroindazole to Aminoindazole. The nucleophilic displacement of a
nitro group with an oxygen nucleophile has been achieved on nitroarenes bearing electronꢀ
1
5
withdrawing groups or halogens, which we intended to adapt for nitroindazole 20. We began
this investigation by screening hydroxide, phenoxide, and methoxide with 20 in the hope of
direct conversion of the nitro group to the corresponding phenol or ether. Unfortunately, direct
functionalization of indazole 20 to the desired phenol or ether was unsuccessful; in all cases
unreacted 20 was recovered.
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
Next, we turned our attention to a twoꢀstep process consisting of nitro reduction to
aminoindazole 23 (Table 2) followed by hydrolysis of the amino functionality to
hydroxylindazole 13 (Table 3). Three different nitro reduction methods were investigated for the
16
conversion of 20 to 23: 1) Bechamp type reaction with Fe/NH Cl; 2) hydrazine or
4
1
7
18
methylhydrazine with Raney nickel (RaꢀNi); and 3) catalytic hydrogenolysis.
In the
Bechamp method, Fe powder (4 equiv) was added to an EtOH/H O (3:2) solution containing
2
NH Cl (10 equiv) at 90 °C for 2 h. After extractive work up with EtOAc, 23 was isolated in 81%
4
yield as a crystalline solid. This method was not viewed favorably within current process
development, as the heavy iron powder and excess NH Cl solid raised mixing and suspension
4
concerns, as well as extra operational burden in the workꢀup and the handling of metal waste.
19
Hydrazines are known hydrogen donors in various hydrogenation reactions.
In the
Raney nickel approach, hydrazine or methylhydrazine serves as the stoichiometric reductant in
the presence of the nickel catalyst. This reaction could be performed either under hydrogen
pressure or with an inert environment (Table 2, Entries 1–3). In all cases, the debrominated
byproduct 24 was observed as an impurity. Although the initial results proved promising from a
reactivity perspective, scale up of RaꢀNi reductions can prove challenging due to the difficulty of
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 33
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
suspending the dense metal catalyst. Additionally, a process safety assessment indicated a high
level of concern with regards to the use of pyrophoric Raney nickel in combination with a
hydrazine. Therefore, the decision was made to explore other reduction options for the
conversion of 20 to 23.
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
In the metal catalyzed hydrogenolysis approach (Table 2, Entries 4–8), our goal was to
develop a batch reaction which could then potentially be converted to a continuous packedꢀbed
transformation. In the batch process, while the reaction employing Raney nickel catalyst at 25
°
C resulted in incomplete conversion, reaction of 20 with Pt/C catalyst (5% w/w metal loading
on carbon, dry basis) in 5:1 EtOH/water at 25 °C provided a 6:1 ratio of desired product 23 and
debrominated byproduct 24 (Table 2, Entries 4 and 5). When the catalyst was changed to
sulfided Pt/C, reduced production of 24 was observed, but the reaction did not go to completion
after 20 h at 60 °C (Table 2, Entry 6). Finally, by replacing EtOH/water with THF as the
reaction solvent, the catalytic hydrogenation achieved completion with only a small amount of
2
4 generated. After filtration to remove the catalyst and concentration of the solvent, addition of
EtOH/H O precipitated the product 23, which was isolated in 83% yield with 98.6% HPLC
2
purity on 283 g scale. With respect to these conditions, there were clear performance differences
between the 5 g front run and the larger batch (Table 2, Entries 7 and 8). Further work must be
done to understand the scale up parameters that will enable robust production.
In all three nitro reduction methods, debrominated byꢀproduct 24 was observed.
Fortunately, 24 is largely purged during subsequent synthetic steps. The debrominated analog
does not participate in the downstream Suzuki cross coupling and is simply purged during
isolation of the crossꢀcoupled product (>90% rejection at 1% w/w loading with respect to 2).
Nitroindazole 20 may be reduced to aniline 23 utilizing at least three different methods, any of
ACS Paragon Plus Environment
14

Page 15 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
which could potentially be further developed into a commercial manufacturing process. We
investigated all three methods in an attempt to provide flexibility for future manufactures. The
ultimate selection of methodology will be based on a combination of process performance and
the available equipment and technical expertise at the commercial manufacturing site.
Table 2. Preparation of aminoindazole 23
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
Scale
(g)
Catalyst
(0.1 g/g)
HPLC area% ratio
(23:20:24)
Entry
Reductant
Solvent
a
NH
NH •H O (2.5 eq)
2 2 2
1
20
RaꢀNi
EtOH
93:0:6
with H (1 atm)
2
b
methylhydrazine (2.5 eq)
2
80
5
RaꢀNi
RaꢀNi
RaꢀNi
EtOH
EtOH
97:0:2
85:6:5
30:18:0
with H (1 atm)
2
3
methylhydrazine (2.5 eq)
c
EtOH:H
5:1)
2
O
O
4
1.5
H (1 atm)
2
(
c
EtOH:H
(5:1)
2
5
2
5
H (1 atm)
2
Pt/C
78:0:12
EtOH:H O
2
6
7
H (1 atm)
2
Pt(S)/C
54:43:0.5
98:0:0.5
(
5:1)
5
H (1 atm)
2
Pt(S)/C
Pt(S)/C
THF
THF
d
8
283
H (1 atm)
2
90:3:3
c
a
b
79% isolated yield, 97% HPLC purity; 76% isolated yield, 97% HPLC purity; reaction at 25
d
°
C; Reaction was conducted for 10 h, 83% isolated yield, 98.6% HPLC purity.
Since methylhydrazine was used for the indazole formation, it could potentially serve as a
hydrogen source in the subsequent nitro reduction. This approach would allow for a oneꢀpot
preparation of 23 from 21 using methylhydrazine and Raney nickel catalyst. In practice, after
indazole formation using the conditions in Table 1, additional methyhydrazine was added to the
reaction mixture, followed by Raney nickel catalyst. The first stage of the optimized process
used K CO (1.5 equiv) in IPA/H O with methylhydrazine (1.5 equiv) at 60 °C for 2 h. After
2
3
2
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 33
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
cooling the reaction mixture, the second stage was then initiated by addition of methylhydrazine
and Raney nickel. After refluxing the reaction for 15 h, removal of the catalyst by filtration and
cooling crystallization of the product from IPA/H O provided 23 in 60% yield (96% HPLC
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
purity) from 21.
D. Hydroxyindazole Formation from Aminoindazole Hydrolysis. The penultimate step to 2
2
0
involved hydrolysis of 23 to 5ꢀhydroxyindazole 13. Diazotization conditions with NaNO and
2
H SO or HCl resulted in formation of the corresponding diazonium intermediate as evidenced
2
4
+
by LCMS, which showed complete conversion to the diazonium salt of 23 ([M] m/z = 237/239).
Despite extensive efforts, the diazonium intermediate did not undergo the desired hydrolysis
step. When higher temperatures were applied to the hydrolysis attempts, multiple unidentified
byproducts were observed, likely due to several decomposition pathways of the diazonium
intermediate.
Since aqueous H PO at high temperature has been reported for the hydrolysis of aniline
3
4
21
to phenol, it was investigated for the hydrolysis of 23 (Table 3). Initially, a solution of 23 in
5:1 v/v of 85% w/w aq. H PO /H O was refluxed at 135 °C for 40 h. We were pleased when
3
4
2
these conditions afforded 71% conversion of 23 to 13 (Table 3, Entry 1). A 144 h (6 days)
reaction time delivered 98% conversion (Table 3, Entry 2). However, the long hold at high
temperature required to achieve high reaction conversion was deemed impractical. In order to
achieve higher temperatures under pressure, screening reactions were conducted in sealed glass
tubes. While it required 11 h at 160 °C to reach 80% of 13 with 18% unreacted aminoindazole
23 by HPLC area%, reaction at 200 °C gave 92% of 13 with only 2% of 23 remained after 1 h
reaction time (Table 3, Entries 3 and 4). Guided by process safety considerations under extreme
ACS Paragon Plus Environment
16

Page 17 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
conditions (acidic conditions at high temperature and pressure), continuous processing in a PFR
®
was investigated. Proof of concept for the continuous reaction was achieved using a Hastelloy
type Bꢀ2 PFR with pressurization to 83 psig (poundꢀperꢀsquareꢀinch, gauge): a 1 h nominal
residence time (reactor volume / volumetric flow rate) at 200 °C produced 130 g of 13 in 71%
isolated yield (Table 3, Entry 5). However, after completion of a flow process lasting about 16
h, contamination of the collected reaction mixture with nickel (Ni is a major component of
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
®
Hastelloy type Bꢀ2) was noted by Inductively Coupled Plasma Mass Spectrometry (ICPꢀMS)
®
analysis, indicating corrosion of the Hastelloy coil. Although it did not ultimately affect the
quality of isolated 13 or the indazole 2 in the subsequent step, extreme caution is merited in this
instance, as reactor rupture would occur upon extended exposure and it is not recommended that
®
this process be run in Hastelloy . Inherently unsafe operations such as this are not conducted as
a general practice at Eli Lilly. A corrosion study of different materials to hopefully identify a
®
more suitable PFR material of construction revealed a 0.35% w/w loss for a Hastelloy coupon
in the mixture of H PO /H O after 4 days at 160 °C, while a tantalum coupon showed 0.015%
3
4
2
w/w loss under identical conditions. No weight loss or visible physical change (e.g., softening,
coloring, cracking) was observed for PFA (perfluoroalkoxy) coupon under the same conditions.
Therefore, hydrolysis of 23 was next accomplished using a stainless steel encompassed PFA PFR
on lab scale (5 g) at 160 °C with a residence time of 10 h, affording 13 in 77% isolated yield with
9
7% HPLC purity. In this case, 160 °C was selected due to concern over softening of the PFA
tubing at higher temperatures. Backpressure was applied (70 psig), as Aspen prediction indicates
33 psig of pressure is required to prevent vaporization of the reactor contents. Overall, the
conversion from 23 to 13 in H PO /H O under thermal conditions is a relatively clean reaction,
3
4
2
as only the starting material and desired product are observed as the major components
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 33
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
(
generally >93% HPLC area combined, as shown in Table 3) when monitoring by HPLC.
Neutralization of the acidic reaction mixture with base and extraction into EtOAc provided good
recovery of isolated product (Table 3, Entry 1). The more operationally straightforward
precipitation method from diluted reaction mixture gave less efficient recovery of product due to
material loss in aqueous solution (Table 3, Entries 5 and 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
Table 3. Reaction conditions for conversion of aminoindazole 23 to hydroxyindazole 13
H N
HO
Br
2
H3PO4:H2O (3:1 v/v)
N
N
N
N
Br
Me
Me
23
13
Scale
Temp
(°C)
Time
(h)
HPLC area% ratio Isolated Yield
Entry
Reactor
(
g)
(13:23:others)
(HPLC area%)
a
1
1
135
135
160
200
200
40
144
11
1
RBF, condenser
RBF, condenser
Sealed glass tube
Sealed glass tube
71:27:2
95:2:3
70% (99%)
ND (ND)
2
3
4
5
20
0.5
2
80:18:2
92:2:6
ND (ND)
60% (92%)
b
®
c
182
ca. 1
Hastelloy PFR
81:5:14
71% (94%)
b
c
6
5
160
b
10
PFA PFR
89:4:7
c
77% (97%)
a
H PO :H O (5:1, v/v); nominal residence time of flow reaction; ~13% product loss in aq.
3
4
2
solution during workꢀup. RBF = Round Bottom Flask, ND = Not Determined.
Additional screening of the H PO : H O ratio (from 6:1 to 1:6, v/v) indicated an optimal
3
4
2
operational range of 2–3:1 H PO :H O (or 65–71% w/w solvent H PO strength) in order to
3
4
2
3
4
achieve the highest reaction conversion at 160 °C within 5 h (Figure 3). Further investigation
and scaleꢀup of this reaction conditions is currently ongoing and will be reported in due course.
a
Figure 3. Reaction screening for optimal solvent H PO strength
3
4
ACS Paragon Plus Environment
18

Page 19 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
00
9
8
0
0
70
6
0
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
50
Solids Present
Homogeneous
4
0
30
2
0
10
15%
25%
35%
45%
55%
65%
75%
Solvent H PO Strength (w/w)
3
4
a
Reactions were conducted in a mixture of H PO and H O (12 mL per gram of 23) and stopped
after stirring for 5 h at 160 °C.
3
4
2
E. Preparation of 2. We have previously disclosed a method for similar diaryl ether
6a
formation. The key challenge to the success of this approach is in formation of an appropriate
phenolic coupling partner. The newlyꢀdeveloped route benefits in that 13 has the isomerically
pure methylated indazole ring fully formed and no further manipulations are necessary after the
SNAr ether formation with fluoride 7. Therefore, the aryl ether formation conditions required
little effort for optimization in terms of a base, solvent, temperature and reaction rate. The
optimized conditions used K CO as base, with THF as solvent at 55 °C for 14 h to give >90%
2
3
assay yield of 2. After aqueous workꢀup, crystallization from toluene afforded 72% yield of
purified 2 (Scheme 3).
Scheme 3. Developed alternative synthesis of indazole aryl ether 2
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 33
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
Figure 4 shows an HPLC chromatogram of indazole 2 derived from the new route (“nitro
route”) compared with crude and recrystallized material from the Ullmann route. The overall
impurity load from the new method is significantly reduced versus the Ullmann method both in
terms of overall number of impurities as well as area% of individual species. Although the
impurities from the Ullmann route are ultimately purged to acceptable levels in the downstream
steps, the new synthesis is viewed as advantageous given the greatly improved purity profile for
this proposed regulatory starting material (2). Given that 2 contains a genotoxic alerting
functionality (aromatic nitro), it is not surprising that impurities related to 2 would likely also be
of similar concern. We have in fact identified several such impurities derived from the Ullmann
route and have characterized their downstream behavior in order to understand their fate as well
as demonstrate purge to acceptable limits. Since the new method has fewer overall such
impurities, this task is greatly simplified should long term manufacturing utilize the new route.
Finally, shown in Figure 5 are photographs of typical material derived from the Ullmann route
compared to that from the new route. While the color from 2 has been shown to normally purge
in downstream operations, there have been individual developmental batches of 2 that have
resulted in colored intermediates and API.
ACS Paragon Plus Environment
20

Page 21 of 33
Organic Process Research & Development
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
Figure 4. Comparison of compound 2 productions by HPLC analysis (Trace identification from
top to bottom: Green = Ullmann route crude solids, Blue = Ullmann route recrystallized solids,
Purple = nitro route solids, Red = blank injection).
A
B
Figure 5. Comparison of compound 2 productions by color (A. Sample of 2 from Ullmann
route; B. Sample of 2 from nitro route).
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 33
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
CONCLUSION
An alternative indazole synthesis for merestinib starting material 2 has been developed.
Indazole 2 is common to multiple existing merestinib syntheses, and the new method of
preparation offers multiple advantages over previous methods. The key transformation in the
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
new preparation of 2 is an intramolecular S Ar reaction to form the indazole heterocycle; this
N
reaction ensures that the methyl group is installed in a selective manner. The intramolecular
SNAr precursor is derived from a readily available aldehyde building block and the cyclization
itself is enabled by use of an excellent S Ar leaving group (fluoride) and the strongly electronꢀ
N
withdrawing nitro group para to the fluorine. These two factors appear to be necessary for the
success of this transformation, as less activated systems didn’t successfully yield product. We
chose to develop the nitration using continuous flow methodology and showed that the reaction
could be run using a small PFR, which has several safety benefits. The new synthesis avoids the
need for any tedious cleanꢀup procedures such as metal scavenging during manufacturing. The
most significant challenge during the development was identification of suitable conditions to
convert the nitro group into the phenolic functionality. Ultimately, nitro reduction to the aniline
followed by a high temperature acidic hydrolysis enabled by continuous processing allowed for a
successful transformation. The synthesis of 2 was completed by diaryl ether formation, which
relied on conditions modified from a previously reported approach. Although the new synthesis
route is one step longer than the Ullmann route, it is believed that the new route will prove to be
more robust in terms of process performance and yield than the problematic Ullmann cyclization.
Additionally, indazole 2 derived from the new route is superior to that from the Ullmann route in
terms of impurity profile and color.
ACS Paragon Plus Environment
22

Page 23 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
EXPERIMENTAL SECTION
General Information
All commercially available chemicals and solvents were used directly as received.
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
1
Melting points were recorded (uncorrected) on a Buchi Mꢀ565 melting point monitor. H NMR
1
3
spectra were recorded on Bruker AV 300 and 400 MHz spectrometer. C NMR spectrum were
recorded using the same spectrometers at 75 and 100 MHz, respectively. Chemical shifts are
reported in parts per million (δ), coupling constants (Jꢀvalues) are reported in hertz (Hz), and
spin multiplicities are indicated by the following symbols: s (singlet), d (doublet), t (triplet), q
(quartet), m (multiplet), bs (broad singlet). Mass spectra were obtained from Agilent 1290
Infinity II with 6130 Chemstation and Agilent 1200 Infinity II with 6130 Chemstation. GCMS
were recorded using Agilent technologies; 7890B GC with 5977B MS. Infrared spectra were
obtained on a PerkinElmer Spectrometry.
Reactions were monitored by TLC (Thin Layer Chromatography), or reverseꢀphase
HPLC using an Atlantis dꢀC18 4.6 mm × 250mm column with 5 µm particle size. For HPLC:
Solvent A was a 0.1% solution of TFA in water; Solvent B was MeOH. The flow rate was 1.0
mL/min with a column temperature of 40 °C. Detection wavelength: 210 nm. The HPLC
gradient method is shown below:
Time/min
0
Solvent A %
Solvent B%
70
5
30
95
95
30
30
1
5
18
8.1
5
1
70
70
2
0
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 33
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
4
-Bromo-2-fluoro-5-nitrobenzaldehyde (21) (batch reaction): To a stirred solution of 4ꢀ
bromoꢀ2ꢀfluorobenzaldehyde 19 (100 g, 493 mmol) in conc. H SO (500 mL) at 0 °C (iceꢀwater
2
4
bath) was added KNO (54.3 g, 547 mmol) portionꢀwise [CAUTION: exothermic; internal
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
temperature increased to 30 °C]. The mixture was stirred at 0 °C for 2 h, at which time TLC
(heptane:acetone = 2:1, UV) indicated completion of the reaction. The reaction mixture was
slowly poured into 5 kg of crushed ice with stirring. The solid thus precipitated was filtered,
washed with water (3 × 500 mL), and dried to afford 21 (110 g, 90% yield) as an offꢀwhite solid.
ꢀ1
m.p.: 46.4 – 48.5 °C; IR (thin film, cm ) 3102, 3050, 2882, 1700, 1604, 1571, 1532, 1459, 657,
1
605; H NMR (400 MHz, DMSOꢀd ): δ 10.14 (s, 1H), 8.50 (d, J = 6.4 Hz, 1H), 8.22 (d, JHꢀF
=
6
HꢀF
1
3
1
0.0 Hz, 1H); C NMR (DMSOꢀd , 75 MHz): δ 186.4, 163.39 (d, J = 266.7 Hz), 146.9, 127.1,
6 CF
1
24.3, 124.1, 121.6; HPLC t = 10.7 min (210 nm); GCMS: calculated for C H BrFNO 246.9,
R 7 3 3
found 246.9.
4
-Bromo-2-fluoro-5-nitrobenzaldehyde (21) (flow reaction): PFR: PTFE tubing, ID = 6 mm,
length = 24 m, maintained at 3 °C in ice bath. Feed solution A: 4ꢀbromoꢀ2ꢀfluorobenzaldehyde
(0.50 kg, 2.46 mol) in H SO (1.84 kg). Feed solution B: HNO (65% w/w, 0.36 kg, 3.69 mol) in
2
4
3
H SO (2.76 kg). At first, CH Cl was pumped into the PFR with Pump A and Pump B (both are
2
4
2
2
piston pumps made of PTFE) to achieve a steady state of the corresponding flow rates (Pump A
rate = 9.2 g/min; Pump B rate = 12.2 g/min), which is adjusted by weight loss of two CH Cl2
2
feed solutions and reflects a theoretical 1.5 equiv. of HNO in relative to the aldehyde starting
3
material. After the PFR was filled with CH Cl , and confirming normal operation for 30 min,
2
2
the pumps were stopped. Feed solutions A and B were then pumped into the PFR by Pumps A
and B respectively with the temperature of the reactor maintained at 3 °C. The mass flow rates
ACS Paragon Plus Environment
24

Page 25 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
were monitored such that 9.2 g/min was delivered from Pump A and 12.2 g/min from Pump B.
These conditions resulted in a nominal residence time of 23 min. The reactor outflow was
collected in a receiving reactor maintained at 3 °C. After the feed solutions were consumed,
conc. H SO (0.92 kg) was pumped into the PFR to push out the reaction mixture, followed by
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
2
4
CH Cl until colorless liquid was observed at the reactor outlet. The collected mixture was then
2
2
poured onto cracked ice (7.5 kg) slowly. The mixture was extracted with CH Cl (3 × 2.5 L). The
2
2
combined organic phase was washed with 10% aqueous K CO solution (2 × 2.5 L) to adjust the
2
3
pH to ≥7, washed with H O (2.5 L) and concentrated under reduced pressure. Heptane (1.5 L)
2
was added and the resulting mixture was stirred for 1 h and then cooled to 2 °C, held for 50 min.
A solid precipitated, which was collected by filtration and rinsed with heptane (0.5 L). The solid
was dried under nitrogen at 23 °C to afford 21 (0.54 kg, 88% yield) as a yellow solid.
6
-bromo-1-methyl-5-nitro-1H-indazole (20): To a 10 L jacketed reactor charged with aldehyde
1 (0.51 kg, 2.05 mol), isopropanol (1.53 L) and H O (0.765 L) at 24 °C was added K CO
3
2
2
2
(0.425 kg, 3.08 mol) in portions. After the suspension was warmed to 41 °C, methylhydrazine
0.142 kg, 3.08 mol) was added dropwise over 1 h. The reaction mixture was stirred for 1 h at
1 °C and then heated to 77 °C for 6 h and monitored by HPLC until the reaction was complete.
(
4
The reaction mixture was cooled to 30 °C, water (5.1 L) was added into reaction mixture over 1
h, and the mixture was stirred at 15 °C. After stirring for 1.5 hours, the solid was collected by
filtration, the wet cake was rinsed with water (1.02 L) and the solids were dried under nitrogen at
23 °C to provide 20 as a yellow solid (0.40 kg, 76% yield). m.p.: 161.9 – 165.9 °C; IR (thin film,
ꢀ1
1
cm ): 3104, 3067, 3027, 2952, 1686, 1608, 1566, 1515, 1487, 655, 623; H NMR (400 MHz,
13
DMSOꢀd ): δ 8.61 (s, 1H), 8.33 (s, 2H), 4.11 (s, 3H); C NMR (DMSOꢀd , 75 MHz): δ 143.6,
6
6
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 33
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
40.5, 135.6, 121.7, 120.4, 116.2, 110.7, 36.4; HPLC t = 12.4 min (210 nm); GCMS: calculated
R
for C H BrN O 255.0, found 255.0.
8
6
3
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
6
-bromo-1-methyl-1H-indazol-5-amine (23): To a 10 L 4ꢀnecked round bottom flask charged
with a solution of 20 (0.283 kg, 1.11 mol) in THF (3.0 L) at 23 °C was added Pt(S)/C (5% w/w
metal loading on carbon, dry basis; 0.03 kg, 7.7 mmol) in one portion. The reaction mixture was
purged with nitrogen (5×), and then purged with hydrogen (3×) at atmospheric pressure. The
mixture was heated to 60 °C and stirred under hydrogen for 10 h, at which time HPLC analysis
indicated completion of the reaction. The reaction mixture was cooled to 25 °C, and THF (3.0 L)
®
was added. After stirring for 1 hour, the mixture was filtered through Celite (128 g) to remove
the catalyst, and the filtrate was concentrated to 0.4 L. To the residue was added a mixture of
EtOH (0.32 L) and H O (3.0 L) slowly with stirring. A solid formed and was stirred for 1 h and
2
then collected by filtration. The solid was dried under nitrogen at 24 °C to afford 23 (0.21 kg,
ꢀ1
8
3% yield) as a brown solid. m.p.: 153.6 – 159.2 °C; IR (thin film, cm ) 3415, 3304, 3206,
1
3
090, 2940, 1677, 1624, 1563, 975, 868, 841, 621; H NMR (400 MHz, DMSOꢀd ): δ 7.85 (s,
6
1
3
1H), 7.77 (s, 1H), 7.04 (s, 1H), 4.95 (bs, 2H), 3.93 (s, 3H); C NMR (DMSOꢀd , 100 MHz): δ
6
1
39.8, 135.2, 130.9, 124.5, 113.4, 112.7, 102.8, 35.9; HPLC t = 5.7 min (210 nm); LCMS:
R
+
[
M+H] : calculated for C H BrN 226.0, found 225.8.
8
9
3
6
-bromo-1-methyl-1H-indazol-5-amine (23) (one-pot procedure from aldehyde 21 with
MeNHNH /Raney-Ni): A 4ꢀnecked round bottom flask equipped with a mechanical stirrer and
2
thermocouple was charged with aldehyde 21 (10.0 g, 40.3 mmol), K CO (8.35 g, 60.5 mmol),
2
3
isopropanol (30 mL) and H O (15 mL) at 25 °C. 1ꢀMethylhydazine (2.76 g, 60.5 mmol) was
2
ACS Paragon Plus Environment
26

Page 27 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
added dropwise at 25 °C. The reaction mixture was heated to 80 °C and stirred for 2 h, at which
time LCMS showed completed conversion to indazole 20. After the reaction was cooled to
25 °C, RaneyꢀNi (2.0 g, 50% w/w in aq. solution, 17.8 mmol) was added along with additional 1ꢀ
methylhydazine (1.84 g, 40.3 mmol) and the mixture was heated to reflux for 20 h. After cooling
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
®
and addition of THF (100 mL), the mixture was filtered through Celite (20 g), and the residue
was concentrated to remove the organic solvents. The solid was dissolved in MTBE (50 mL) and
then acidified to pH = 2 with 1 N HCl. The separated aqueous product solution was washed with
MTBE and then basified with 30% aq. NaOH solution to pH = 8–9. The resulting suspension
was stirred and solid was collected by filtration. The solid was washed with water (2 × 50 mL)
and dried under vacuum to provide 23 (5.5 g, 60% yield) as a brown solid.
6-bromo-1-methyl-1H-indazol-5-ol (13) (batch reaction): To a stirred suspension of 23 (2.00
g, 8.9 mmol) in water (6 mL) in a glass tube, phosphoric acid (85% in water, 18 mL) was added
at 25 °C. The tube was sealed and heated to 200 °C for 1 h. After completion of the reaction
(
monitored by TLC & LCMS), the reaction mixture was cooled to 25 °C and poured into
crushed ice (300 g) with stirring. The solid thus precipitated was filtered, washed with water (3 ×
0 mL) and dried to afford 13 (1.22 g, 60% yield) as a yellow solid. m.p.: 184.3 – 185.7 °C; IR
3
ꢀ1
1
(
thin film, cm ) 3110, 2946, 1671, 1624, 1563, 1495, 627, 587 ; H NMR (400 MHz, DMSOꢀd )
6
1
3
δ 9.93 (bs, 1H), 7.93 (s, 1H), 7.86 (s, 1H), 7.17 (s, 1H), 3.97 (s, 3H); C NMR (DMSOꢀd , 75
6
MHz): δ 148.2, 135.7, 131.5, 123.9, 114.0, 112.9, 103.9, 35.9; HPLC t = 9.2 min (210 nm);
R
+
LCMS: [M+H] : calculated for C H BrN O 227.0/229.0, found 227.0/228.9.
8
8
2
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 33
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
-bromo-1-methyl-1H-indazol-5-ol (13) (flow reactor run): Two identical PFRs were
®
constructed from Hastelloy (Bꢀ2) tubing, ID = 6 mm, length = 32 m. One of the PFR was
heated to 200 °C using high temperature oil bath. The second PFR was maintained at 7 °C for
the reactor outflow. Feed solution A: a mixture of 23 (0.182 kg, 0.805 mol), water (0.60 L) and
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
5% w/w aq. H PO (1.82 L) was heated to 51 °C and stirred until the solid dissolved
3 4
completely. The mixture was cooled to 20 °C and was ready for use. Feed solution B: a solution
of 85% w/w aq. H PO /H O (3:1, v/v). The outflow receiving flask consisted of a 20 L fourꢀ
3
4
2
necked round bottom flask containing H O (9.1 L) cooled to 4 °C. Solution B was pumped into
2
the PFR at 4.0 g/min for 30 min. Then, the flow of Solution B was stopped and the flow of
Solution A was initiated into the PFR at 4.0 g/min. The flow rate was monitored by matching
the actual weight loss of the feed solution with the calculated theoretical weight loss. The
reactor outflow passed through the cooling coil and flowed into the receiving flask after 60 min
residence time (not corrected for thermal expansion). After Solution A was fully consumed,
Solution B was pumped into the system to push out the reaction mixture until colorless liquid
was observed at the outlet. The collected mixture was stirred in the receiving flask at 5 °C for 1.5
h, and the precipitated solid was filtered and rinsed with H O (0.91 L). The solid was dried under
2
nitrogen at 25 °C to afford 13 (130 g, 71% yield) as a dark brown solid.
Preparation of 6-bromo-5-(2-fluoro-4-nitrophenoxy)-1-methyl-1H-indazole (2): To a 2 L
fourꢀnecked round bottom flask charged with 13 (0.115 kg, 0.51 mol) and THF (0.55 L) at 25 °C
was added K CO (0.77 kg, 0.55 mol) and 3,4ꢀdifluoronitrobenzene (0.83 kg, 0.52 mol). The
2
3
resulting mixture was stirred at 55 °C for 14 h and then concentrated to 2–3 vol (<345 mL).
Water (1.15 L) was slowly added, and the resulting solid was collected through filtration, rinsed
ACS Paragon Plus Environment
28

Page 29 of 33
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
with H O (0.22 L), followed by EtOH (1.1 L). The yellow solid was dissolved in toluene (1.1 L)
2
at 80 °C, to which activated carbon (57 g) was added. The resulting suspension was stirred at 80
°C for 5 h and the activated carbon was removed by filtration. After concentration to 2–3 vol, the
mixture was stirred at 85 °C for 3 h and was cooled to 5 °C slowly. The suspension was stirred at
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
°C for 2 h. The solid was collected by filtration, rinsed with EtOH (0.22 L), and dried under
nitrogen at 25 °C to afford 2 as a white solid (0.134 kg, 72% yield). m.p.: 182.2 – 187.5 °C; IR
ꢀ1
1
(
thin film, cm ) 3044, 1675, 1601, 1521, 1484, 1431, 1405, 668, 623; H NMR (400 MHz,
DMSOꢀd ) δ 8.36 (dd, J = 11.1, 2.8 Hz, 1H), 8.30 (s, 1H), 8.10 (s, 1H), 8.04–8.01 (m, 1H), 7.85
6
13
(s, 1H), 6.89 (t, J = 8.8 Hz, 1H), 4.09 (s, 3H); C NMR (DMSOꢀd , 75 MHz): δ 152.6, 151.7,
6
1
1
51.6, 149.3, 144.0, 142.5, 142.4, 138.6, 133.3, 123.5. 121.8 (2C), 117.1, 115.6, 114.6, 114.1,
13.7, 113.4, 36.3; HPLC t = 15.5 min (210 nm); GCMS: calculated for C H BrFN O
3
R
14
9
3
1
13
365.0/367.0, found 365.0/366.9. H and C NMR spectra are in agreement with those reported in
6
a
literature.
ASSOCIATED CONTENT
Supporting Information
1
13
Schematic drawings of equipment set for Step 1 and Step 4 in PFR, as well as H and C NMR
spectra for the compounds 2, 13, 20, 21, 23. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Email: ylu@lilly.com
ACS Paragon Plus Environment
29