Process Safety Considerations for the Supply of a High-Energy Oxadiazole IDO1-Selective Inhibitor

Article abstract of DOI:10.1021/acs.oprd.9b00105

The development of a stereospecific synthesis of a IDO1-selective inhibitor is described. The synthetic strategy toward enabling early discovery efforts along with additional findings pertaining to process safety that limited scalability are outlined. A convergent approach that supported the synthesis of material suitable for early preclinical and/or good laboratory practice toxicology studies and avoided the formation of key high-energy intermediates is summarized.

Full text of DOI:10.1021/acs.oprd.9b00105

Subscriber access provided by ALBRIGHT COLLEGE
Full Paper
Process Safety Considerations for the Supply of a
High Energy Oxadiazole IDO1-Selective Inhibitor
Theodore A Martinot, Michael J. Ardolino, Lu Chen, Yu-hong Lam, Chaomin Li, Matthew L Maddess, Daniel
J. Muzzio, Ji Qi, Josep Saurí, Zhiguo Jake Song, Lushi Tan, Thomas P. Vickery, Jingjun Yin, and Ralph Zhao
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.9b00105 • Publication Date (Web): 28 May 2019
Downloaded from http://pubs.acs.org on June 2, 2019
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.
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 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Process Safety Considerations for the Supply of a
High Energy Oxadiazole IDO1-Selective Inhibitor
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
Theodore A. Martinot,*,†,※ Michael Ardolino, Lu Chen, Yu-hong Lam, Chaomin Li,
†
┴
‡
†,⌂
†
§
§
†
§
Matthew L. Maddess, Daniel Muzzio, Ji Qi, Josep Saurí, Zhiguo J. Song, Lushi
§
§
§
§
Tan, Thomas Vickery, Jingjun Yin, and Ralph Zhao
AUTHOR ADDRESS
†Process Research & Development, MRL, Merck & Co., Inc., Boston, Massachusetts,
02115, USA
§Process Research & Development, MRL, Merck & Co., Inc., Rahway, New Jersey
07065, USA
‡Modeling and Informatics, MRL, Merck & Co., Inc., Rahway, New Jersey 07065, USA
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 60
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
┴WuXi AppTec, 288 Fute Zhong Road, Waigaoqiao Free Trade Zone, Shanghai
200131, China
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 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
TOC GRAPHIC
O
N
N
O
N
X = R'O, R'NH, R'S
N
N
O
N
O
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
O
O
NH2
N
N
NO2
N O
O
N
Route Comparison
N
Cl
N
Process Safety Assessment
Strategic considerations
HN
HO
N
NH2
N
X
HO
N O
N O
H
N
N
[RAR]
N
H N
2
X = R'NH
N
HN
N
NH2
HO
HO
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 60
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
ABSTRACT: The development of a stereospecific synthesis of a IDO1-selective inhibitor
is described. The synthetic strategy towards enabling early discovery efforts along with
additional findings pertaining to process safety which limited scalability are outlined. A
convergent approach that supported the synthesis of material suitable for early pre-
clinical and/or GLP toxicology studies, and which avoided the formation of key high
energy intermediates is summarized.
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
KEYWORDS: Process Safety, Oxadiazole, Benzocyclobutylamine, IDO1
ACS Paragon Plus Environment
4

Page 5 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
INTRODUCTION
Continued interest in immuno-oncology has been spurred by the demonstrated
durable responses observed in patients treated with anti-PD1 mAb (e.g.,
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
pembrolizumab), among other therapies. Indoleamine-2,3-dioxygenase-1 (IDO1) has
2
been shown to play an important role in immunomodulation. Moreover, IDO1 is over-
expressed in multiple tumor types (including melanoma, colon, and ovarian), and this
mechanism is believed to contribute to immunosuppression, which in turn may promote
tolerance to the cancer.3,4 The combination of IDO1 inhibitors and anti-PD1 mAbs has
5
been shown to be synergistic in efficacy models, and multiple clinical trials are currently
evaluating the efficacy of this combination as a means of improving response rates
while maintaining durability.
6
Notably, epacadostat (INCB24360; 1, Figure 1) has been developed to this end, and
is currently being evaluated in multiple clinical trials in combination with anti-PD1 mAb
therapy (pembrolizumab, nivolumab, etc.). Additional IDO1 inhibitors have been
developed and reported upon in the literature.7
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 60
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
F
F
Br
O
N
O
N
N
N
HN
HN
HO
O
O
HN
N
HN
S
N
NH2
HO
NH
OH
N
H
O
1
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 1. Structure of epacadostat (1) and 2
Recently, we sought a synthetic approach to structures such as 2 that enabled the
discovery effort around this target (IDO1) and facilitated the construction of Structure-
7
Activity Relationship (SAR) around both sides of the oxadiazole. The following
discussion will feature strategic elements surrounding the supply of 2, including route
planning with regards to availability of the building blocks and safety, as well as impact
of the selection of these building blocks or intermediates on the delivery timeline.
INITIAL ROUTE TO 1 AND PRELIMINARY EVALUATION
Initial access to 2 (Scheme 1) started from benzocyclobutyl carboxylic acid 3, which
was initially available in small quantities, and converted to the amine using a Curtius
7
rearrangement. Initial efforts to support SAR for this target (IDO1) carried the racemic
carbamate (rac-4) forward, with a resolution of the final inhibitor (e.g., 2), typically using
supercritical fluid chromatography with a chiral stationary phase (“chiral SFC”).
ACS Paragon Plus Environment
6

Page 7 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
However, once the configuration of the optimal benzocyclobutylamine building block
was determined to be (S), this intermediate was amenable to early resolution, which
facilitated downstream operations by avoiding resolution of each new compound. The
racemic N-Boc amine was resolved using chiral SFC to afford 4. After deprotection to
afford the benzocyclobutylamine hydrochloride salt (5•HCl), hydroxyamidine 7 was
formed by condensation with commercially-available imidoyl chloride 6.8 The
hydroxyamidine was protected as the cyclic carbamate by reaction with
carbonyldiimidazole (CDI), and the amino-oxadiazole (8) was oxidized to the nitro-
oxadiazole (9) using hydrogen peroxide in trifluoroacetic acid (TFA) or sulfuric acid.9
The final product was obtained by condensation of the ethanolamine side chain (10)
through displacement of the nitrooxadiazole and subsequent deprotection of the
carbamate.
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
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 60
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
HCl / dioxane
then conc.
DPPA / tBuOH
then SFC
•HCl
F
F
NH2
5•HCl
F
NHBoc
COOH
3
4
O
N
N
Cl
F
F
NH2
N
O
N
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
N
O
CDI
OH
N
N
HN
N
O
NaHCO3
EtOH / H2O
O
N
NH2
N
NH2
HO
7
8
F
H
N
H2N
OH
O
F
N
N
O
O
10
H2O2 / TFA
N
HN
HO
N
N
1. Et3N / THF
2. NaOH (aq) / MeOH / THF
HN
N
O
NH
OH
N
NO2
O
O
9
2
Scheme 1. Initial route to 2
A significant advantage of the route outlined in Scheme 1 is that intermediate 9 was a
versatile building block for rapidly enabling SAR evaluation of the tailpieces, with the
nitro group being a suitable leaving group for a variety of nucleophiles (amine, alcohol,
1
0
thiol, etc.). However, the route was also limiting in that the stereogenic and arguably
more precious portion of the molecule was installed first. This produced a potential
bottleneck, as availability of benzocyclobutylamine 5 was limited early on in the
program. Moreover, the safety element associated with handling compounds 8 and
particularly 9 were a concern because of the low carbon-count relative to the number of
ACS Paragon Plus Environment
8

Page 9 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
heteroatoms (e.g., in 8) or relative to the number of energetic functional groups in the
molecule (oxadiazole, nitro, etc. in 8 and 9).11,12
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
Differential Scanning Calorimetry (DSC) was used as a primary means of evaluating
and triaging intermediates of concern. Specifically, amino-oxadiazole 8 was first
analyzed by DSC (Figure 2) and showed a clean melt with an onset at 213.9 °C,
followed by an exothermic decomposition, which is attributed primarily to decomposition
of the oxadiazole and/or the benzocyclobutylamine. The integrated heat for this
decomposition, which has an onset of 254 °C, is 1567 J/g, which is considered
13
significant. Drop weight tests were performed on 8 at up to 30 J, and were negative (6
of 6 drop tests), indicating that the compound was not shock-sensitive. The high onset
of decomposition (both absolute and relative to a reaction temperature of 25 °C),
coupled with the apparent high crystalline nature of the intermediate (indicated by the
sharp melt) led us to conclude that 8 was an intermediate that could be synthesized and
handled safely at intermediate scale (< 1 kg), as both the melt and the decomposition
occur more than 100 °C above the highest achievable reaction temperature.14
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 60
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 2. DSC thermogram of 8 (sample size: 5.70 mg)
A similar evaluation of 9 was conducted, but the DSC of the nitro-oxadiazole was
shown to be significantly different (Figure 3). In this case, the intermediate showed a
lower onset temperature for its melt (of approx. 87 °C) relative to its precursor (8), and
that transition was also poorly defined. Following its melt, the compound showed a
decomposition exotherm with an onset of 109 °C, and an integrated heat of
decomposition of >3000 J/g; it is noted that the onset of each these transitions is a best
ACS Paragon Plus Environment
10

Page 11 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
case scenario, as solutions or impure phases with no melting obstacles. Drop weight
tests of 9 gave 3 of 6 positive results at 30 J impact (decomposition (charring); no
smoke, flame or audible report), but no positive results at 20 J intensity. Combined,
these results constituted a significant safety risk for scale-up. Therefore, a better
understanding of the synthetic process to the formation of 9 was deemed necessary in
order to fully evaluate whether the approach to 2 outlined in Scheme 1 could be safely
conducted.
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
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 60
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
Figure 3. DSC thermogram of 9 (sample size: 3.40 mg)
FORMATION OF 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
The conversion of 8 to 9 can be effected by a number of different reaction conditions.
16
9
The “Piranha” (sulfuric acid / hydrogen peroxide ) conditions were evaluated, but
resulted in a complex mixture, and alternative options were therefore considered. After
some additional evaluation, it was determined that either trifluoroacetic acid (TFA)17
/
urea hydrogen peroxide (UHP)18 or TFA / hydrogen peroxide afforded the cleanest
reaction profile. Both sets of conditions are believed to proceed through the same
reactive species (pertrifluoroacetic acid, PTFA19,20,21), therefore selection of conditions
was based on more practical considerations. Further development led to the conclusion
that the latter set of conditions using hydrogen peroxide was likely easier to operate
because hydrogen peroxide can be dosed continuously more easily than UHP (a solid),
which in turn means that the individual reaction steps could be better controlled, thus
avoiding potential runaway exotherms.
ACS Paragon Plus Environment
12

Page 13 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
The main impurity of the reaction is nitrile oxime 11 (Scheme 2), which is believed to
originate from the condensation of the intermediate nitroso-oxadiazole with the starting
material (8) followed by degradation of the corresponding diazo-oxadiazole dimer.15
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
F
F
F
OH
CN
O
N
N
N
O
N
N
PTFA
N
O
N
N
O
+
O
O
N
NH2
O
N
N
NO2
O
8
9
11
Approx. 60% LCAP
Approx. 30% LCAP
Scheme 2. Oxidation of 8 to 9
Empirically, only 8, 9, and 11 have been observed (by HPLC-MS analysis) over the
course of the reaction; therefore, it is believed that the relative rate for the initial
oxidation to the hydroxylamine must therefore be a rate-determining step (i.e., no
appreciable accumulation of the hydroxylamine or nitroso intermediates) and the
formation of 11 as a byproduct of the reaction may be inevitable under these reaction
conditions unless the mechanism for the formation of 9 itself were to be changed.
Indeed, varying the reaction temperature, concentration, hydrogen peroxide addition
rate, or stoichiometry did not have a statistically-significant effect on the ratio of 11 to 9
at the end of the reaction. While reaction temperature was shown to have an effect on
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 60
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
the reaction conversion, there was no statistically-significant variation in the amount of
impurity relative to the product under any of the conditions tested. This observation
indicates that formation of the impurity under the reaction conditions used may not be
easily suppressed. Fortunately, upon completion of the reaction, the reaction mixture is
diluted with water, which causes the desired product (9) to nucleate, and most of the
impurities (including 14) can be purged through the mother liquor.
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 addition, reaction calorimetry for the oxidation of 8 to 9 was performed (both in a
SuperCRC™ calorimeter and in a Mettler Toledo EasyMax™ equipped with the
HFCal™) to determine safe operating conditions. First, the addition of hydrogen
peroxide to TFA was evaluated. From the data collected in the EasyMax™, it was
determined that the enthalpy associated with the addition of 0.10 mol of hydrogen
peroxide to 0.78 mol of TFA and the subsequent formation of PTFA is approximately
1
kJ, which corresponds to an adiabatic temperature rise (ΔT ) of < 10 K. Moreover,
ad
because this exotherm is addition-controlled with no apparent thermal accumulation or
delay, the risk posed to the operator in terms of a runaway exotherm can be minimized.
ACS Paragon Plus Environment
14

Page 15 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Separately, the enthalpy measured in the EasyMax™ for the oxidation reaction
conducted on 3 g (0.010 mol) of 8 with 0.10 mol of hydrogen peroxide was determined
to be approx. 6 kJ when the reaction was run at ≤ 40 °C, but increased to 16 kJ when
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 same reaction was run at 60 °C. These reaction enthalpies correspond to a ΔT of
ad
40 K and 120 K, respectively. In the CRC, the same reactions performed at 25 °C and
4
0 °C produced ΔT of 9 K and 90 K, respectively. Though the absolute magnitude of
ad
these temperature rises do not match absolutely, the trends are compelling, and this
remarkable increase in reaction enthalpy for the higher temperature experiments
appears to be caused by the acceleration of the degradation of PTFA as a function of
temperature. Finally, Accelerating Rate Calorimetry (ARC) performed on this reaction
following a rapid addition of hydrogen peroxide showed a significant temperature and
pressure increase at approx. 50 °C, which was also indicative of a potential runaway
exotherm due to PTFA degradation (Figure 4).
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 60
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
a.
b.
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. ARC data showing Self Heat Rate (a) and Pressure Rise Rate (b) as a
function of temperature for the oxidation of 8 to 9. Experiments were performed by
dissolving 8 in TFA and heating to 50 °C, then injecting H O at once (see annotation)
2
2
with the ARC in adiabatic tracking mode.
These data as a whole indicate that care should be taken when performing this
oxidation in order to avoid the thermal decomposition of the PTFA formed. Based on the
results collected, as well as the lack of benefit of performing this oxidation at
significantly higher temperature (Figure 5), the recommended batch temperature is ≤
45 °C at intermediate scale (< 200 g) with a slow dosing of hydrogen peroxide: Although
this temperature is close to the potential degradation of the PTFA, the slow dosing of
ACS Paragon Plus Environment
16

Page 17 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
hydrogen peroxide allows for minimal accumulation of this reagent, and accelerates the
rate of oxidation, thereby minimizing risk to the operator. It is recognized that the edge
of failure for this runaway reaction has not been systematically confirmed (i.e., the
temperature at which the PTFA degradation will cause the reaction to self-heat), and for
this reason, alternative strategies for the assembly of 2 were strongly considered in
order to avoid going through this high energy intermediate and reaction.
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
100
8
0
0
6
40
0
2
4
6
°C
0°C
0°C
0°C
20
0
0
5
10
15
Reaction Time (h)
Figure 5. Effect of temperature on the reaction conversion of 8 to 9
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 60
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
One final concern over the approach to 2 outlined in Scheme 1 is in the condensation
of benzocyclobutylamine 5 with imidoyl chloride 6. As noted elsewhere,22,23 the nitrile
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
oxide formed from treatment of imidoyl chloride 6 with base (12) can homodimerize to
form the symmetrical dioxadiazene (13) and/or its isomeric oxadiazole oxide (15),
depending on the reaction conditions used (Scheme 3). As expected, these species are
highly energetic, and their corresponding oxidation products (14 and 16) are known
explosives.22,23 While the outcome of the condensation of 5 and 6 seems to primarily
result in the formation of 7, we have not attempted to prove that neither 13 nor 15 are
present in the isolated product from this reaction because synthesis of the
corresponding markers and their handling is considered high risk. Therefore, 7 as well
as 8 could technically contain these corresponding impurities, depending on the
reaction parameters used. As a result of this known but uncharacterized risk, and the
risk associated with the formation of nitro-oxadiazole 9, an alternative approach to 2
was carefully evaluated.
ACS Paragon Plus Environment
18

Page 19 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
N
O
N
N
O
N
O
O
O
O
NH2
N
Oxidation
NO2
N
N
NH2
N
NO2
N
N
O
N
O
N
O
O
13
BADD"
14
"BNDD"
N
N
N
N
"
Base
Cl
HO
NH2
NH2
N
N
O
NH2
N
O
NO2
N
O
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
6
12
N
N
O
Oxidation
O
N
O
N
N
NH2
O
N
NO2
N
N
O
N
1
5
O
N
16
"BNFF"
"BAFF"
Scheme 3. Undesired dimerization of imidoyl chloride 622,23
IMPROVED ACCESS TO THE BENZOCYCLOBUTYLAMINE
Early on in development, the synthesis of benzocyclobutylamine 5 was optimized to
allow access to kilogram quantity of the desired product using the existing route. Most
notable in the improved synthesis is the classical resolution of the
benzocyclobutylamine,24 which obviated the need for costly SFC of this intermediate.
Indeed, the Process Mass Intensity (PMI) for the Curtius rearrangement step and the
SFC resolution alone in the original approach was estimated at 3839 due to the high
material demand (solvent and CO ) of the chromatography. In contrast, the process
2
outlined in Scheme 4 decreased the PMI for the Curtius rearrangement and classical
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 60
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
resolution to 98.7. While the resulting PMI is still relatively high for a single step, the
approach was deemed fit-for-purpose at this stage in development.
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
O
H
EtO
P
CN
NaBH , H PO
THF / MeOH
10-20 °C
CN
4
3
4
LDA
O
OEt CN
THF, -60 to -70 °C
0% AY, 58% LCAP
7
F
Br
K PO , THF, 10-20 °C
F
Br
F
Br
3
4
95% LCAP, 93% AY
18
81% IY (2 steps), 87% LCAP
19
17
telescoped
KOH
DPPA
EtOH / H2O, 80 °C
50% IY (2 steps)
Et N
3
F
F
N3
F
CN
F
15-25 °C
CO2H
NCO
2
0
O
21
22
3
telescoped
HO
O
O
0.55 equiv. L-DBTA
15-25 °C
O
Ph
1
. HCl, 85-95 °C
Ph
NH2 •1/2
O
F
2. K2CO3, 10-20 °C
F
32% IY (2 steps)
O
NH2
O
OH
96% ee
74% (2 steps)
rac 5
5•DBTA
>
12 kg made
Scheme 4. Synthesis of benzocyclobutylamine 5 on kilogram scale
APPROACH TO 2 VIA OXADIAZOLE REARRANGEMENT
To enable a safer supply of 2, an approach that takes advantage of the previously-
6
demonstrated Boulton-Katritzky rearrangement (RAR) was attempted (Scheme 5). This
approach provides a number of benefits over the initial approach shown in Scheme 1,
namely: [1] the route is more convergent; [2] the hydroxyamidine functionality does not
need to be protected; [3] with the exception of imidoyl chloride 6, no exceedingly high
ACS Paragon Plus Environment
20

Page 21 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
energy intermediates are used; [4] the benzocyclobutylamine 5 is introduced late, which
results in a higher overall throughput when this material was of limited availability.
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
O
O
N
O
O
O
N
N
N
H
N
N
H2N
HO
N
NH2
Cl
R
R
RAR
HN
NH2
F
N
NH2
N
R
N
HO
OH
O
23
24
6
O
O
N
N
N
N
Cl
HN
HN
HN
N
N
HO
R
HO
NH
OH
25
O
O
2
Scheme 5. Revised strategy to 2
However, it was determined that under the conditions needed to effect the oxadiazole
6
rearrangement on our substrate (aq. KOH, 100 °C), the ethanolamine amide in the
fully-elaborated tailpiece is hydrolyzed. A screen aimed at finding non-aqueous
conditions that effect this rearrangement failed to afford conditions that were
competitive. As a result, a modified strategy was implemented that still installed the
hydroxyamidine first, but with a sacrificial ester that would get saponified during the
rearrangement conditions (Scheme 6). To better understand the thermodynamics of the
transformation, the starting acid (28) and product (29) were modeled. From the analysis
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 60
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
performed, it is determined that the rearrangement is thermodynamically-favored
(
ΔG = -3.9 kcal/mol), and that the geometry of the hydroxyamidine is very important to
rxn
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 conformation of the starting material and product, with the (Z)-isomer being favored
in both cases. Therefore, because the mechanism of the Boulton-Katritzky
rearrangement is believed to be concerted, these data suggest that the starting oxime
((Z)-28, presumably as the dianion due to the high pH of the reaction mixture) must first
isomerize prior to rearrangement to the less favored (E)-isomer which then has the
proper geometry to enable the rearrangement. The product of the rearrangement then
will undergo isomerization to the final most stable conformer, (Z)-29 (Scheme 6), and it
is the release of energy from the combined rearrangement and isomerization that
makes the rearrangement favorable thermodynamically.
ACS Paragon Plus Environment
22

Page 23 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
O
OtBu
N
O
O
O
H2N
O
N
O
N
N
N
KOH (aq)
100 °C
H
N
N
H
N
N
N
O
Cl
26
O
Grxn = -3.9 kcal/mol
H N
2
O
O
OH
NH2
H3O+
N
Et3N
NH2
NH2
HN
(Z)-29
N
N
N
OH
HO
O
HO
6
27
(Z)-
28
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
(
Z)-28
(E)-28
less stable than (Z)-28 by 2.1 kcal/mol
(E)-29
less stable than (Z)-29 by 6.6 kcal/mol
(Z)-29
most stable conformer
most stable conformer
O
O
N
O
N
N
O
O
N
N
O
N
O
N
H
N
N
N
N
N
O
O
NH2
Z)-28
N
NH NH2
(E)-28
NH NH2
(E)-29
NH NH2
O
(
(Z)-29
O
O
O
O
O
O
Scheme 6. Installation of the tailpiece and rearrangement as well as computed energies
for the rearrangement (M06-2X-D3/6-311G(d,p)–SMD(water) geometries and
frequencies. M06-2X-D3/def2-QZVPP–SMD(water) single-point energies.)
The optimized synthesis of 2 is provided in Scheme 7. The formation of
hydroxyamidine 27 proceeded smoothly, with the product being isolated by direct
crystallization from ethyl acetate. The hydrolysis and rearrangement of 27 affords 29
cleanly, although it does require a 48 h reaction time. Acid 29 is also isolated directly
from the reaction mixture after neutralization, thus facilitating the process steps.
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 60
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
However, the condensation of 29 with 3025 to complete the tailpiece assembly was
problematic at first, with varying yields of product being obtained, and unreliable
outcomes requiring complex chromatography.
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
O
O
O
O
N
N
N
O
Et3N (1.5 equiv.)
EtOAc
H
N
N
KOH (3 equiv.)
N
O
N
H2N
HO
Cl
NH2
H O
2
O
O
OH
NH2
reflux, ~ 3 days
then HCl
HN
NH2
0-10 °C then RT / 16 h
N
N
N
HO
OH
26
2
7
29
56 - 66% yield
(1.1 equiv.)
90% yield
6
Direct Crystallization
Direct Crystallization
O
T3P (1.25 equiv.) in EtOAc
DIPEA (6 equiv.)
N
O
O
N
N
O
N
O
O
O
H2N
2
H N
THF, RT / 2h
H2N
HO
O
•HCl
OH
HN
N
N
80% yield
HN
N
H
HO
29
30
1.2 equiv.)
31
(
F
F
NH2
O
O
N
1.
N
5
O
NaNO2 (1.05 equiv.)
HCl / NaCl (3 equiv.)
O
HCl or DBTA
O
N
N
O
N
O
H2N
Cl
N
O
O
DIPEA or NaHCO3
HN
HO
N
N
HN
H
HN
N
0-5 °C / 1 h
N
H
THF
HN
HO
HO
N
reverse addition
90% yield
NH
OH
31
32
2. K2CO3 / MeOH
O
77% yield (2 steps)
2
Scheme 7. Final delivery approach
Initially, EDC and HOBt were used as coupling reagents, but product 31 proved
difficult to separate from HOBt and related reagents, thereby limiting the use of many
common amide-coupling reagents.26 Small-scale runs had shown propylphosphonic
anhydride (T3P) to be a favorable amide coupling reagent for the transformation.
ACS Paragon Plus Environment
24

Page 25 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
However, upon repeated reactions and increase in scale, multiple byproducts including
the pseudo-dimer 33 (Figure 6) were observed, resulting in low yield and purity of 31.
In order to quickly identify the key parameters driving reaction efficiency and
cleanliness, a Design of Experiments (DoE) approach was taken. For this, a 24-
experiment design was initially created using the JMP software.27 The design allowed
use of a common 24-well plate array (1 mL shot vials) to facilitate reaction setup and
analysis while minimizing starting material consumption, and evaluated 4 parameters:
solvent (THF, DCM and EtOAc), DIPEA charge (2 – 4 equiv.), T3P charge (0.5 – 1.5
equiv.), and amine hydrochloride 30 charge (0.9 – 1.2 equiv.). Following completion of
these experiments, analysis indicated that the DIPEA charge was insufficient, even at 4
equiv. Therefore, the DoE was augmented with 5 additional reactions to probe the
effects of higher DIPEA (up to 8 equiv.). The model generated from the data collected
showed DIPEA and T3P equivalents to be the main drivers of purity (panel I, Figure 6),
with THF showing the best balance of conversion and purity among the solvents.
Setting response levels at >80% conversion, >95% purity of converted product and <5%
LCAP 33, the contour profile (panel II, Figure 6) predicted a T3P loading of 1.25
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
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 60
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
equivalents and 6 equivalents of DIPEA with THF as the solvent to be ideal.
Mechanistically, this makes sense, but was not obvious to us a priori: Each equivalent
of T3P can generate, over the course of the reaction, up to 3 equiv. of phosphonic acid
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
(therefore, the total base needed becomes: 3 x 1.25 (T3P) + 1 (amine HCl 30) + 1 (acid
29) = 5.75 ~ 6). With insufficient base, the amine becomes inaccessible (protonated) as
the reaction proceeds and therefore the hydroxyamidine oxygen becomes the next
possible nucleophile and results in the formation of 33.
O
N
NH2
•HCl
O
N
O
AcO
O
H2N
O
O
N
N
N
O
HO
O
30
O
N
N
O
H2N
HO
HN
H
H2N
HO
O
N
N
O
H2N
OH
N
HN
H
H
HN
T3P
N
N
N
33
Pseudo-dimer byproduct
29
DIPEA
DCM
3
1
N
O
N
I
II
(%)
Product
d
e
rt
Conve
Purity
T3P
v.
equi
equi
v.
DIPEA
(solvent = THF)
T3P equiv.
Figure 6. Optimization of the formation of 31
ACS Paragon Plus Environment
26

Page 27 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
These conditions were applied to the overall delivery of 2 (approx. 200 g scale;
Scheme 7), allowing for a cleaner reaction, which after crystallization afforded 31 with
high purity and yield. The optimized conditions demonstrated a robust design space
across a wide scale range (~ mg in DoE optimization to > 200 g).
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
From 31, the synthesis is completed by diazotization of the hydroxyamidine and in situ
formation of imidoyl chloride 32, followed by installation of 5 and deprotection of the
tailpiece. As was observed for the formation of hydroxyamidine 8, some dimerization
may occur if the order of addition is reversed (Scheme 8): Namely, if imidoyl chloride 32
is pre-mixed with the salt of 5 (either HCl or DBTA), then base is added, the
deprotonation of the imidoyl chloride appears to occur faster than the neutralization of
the salt of 5 (i.e., the neutralization reaction may be mass-transfer limited due to
solubility limitations), and therefore results in formation of the corresponding nitrile
oxide, which was found to effectively homodimerize (Scheme 8). The solution to afford
clean coupling is to first neutralize 5 by pre-mixing the salt with the base in the reaction
solvent, and then adding in the imidoyl chloride solution from the previous step. After
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 60
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
final deprotection, the desired product (2) is obtained in approx. 70% yield over three
steps from 31.
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
O
O
O
N
N
H
N
O
O
DIPEA (slow addition)
EtOAc
O
N
O
N
O
O
Cl
O
NH
N
O
O
N
N
HN
H
HN
H
F
O
N
N
NH2
N
HO
HCl or DBTA
N
34
O
3
2
5
(
plus traces of desired hydroxyamidine adduct)
Scheme 8. Homodimerization of 32
CONCLUSION
A fit-for-purpose synthesis of high energy oxadiazole-bearing molecules has been
evaluated. Taking advantage of the Boulton-Katritzky rearrangement the new approach
enabled not only a more convergent route that allowed for the late-stage installation of
the chiral building block 5, but also avoided the generation of high energy intermediates
such as nitrooxadiazole 9. The optimized approach culminates in a longest linear
sequence of nine (9) steps – several of which are telescoped – from commercially-
available aldehyde 17. This work ultimately enabled the safe synthesis of 2 in quantities
sufficient to support toxicology studies.
EXPERIMENTAL SECTION
ACS Paragon Plus Environment
28

Page 29 of 60
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
All reactions were carried out under a nitrogen atmosphere. All solvents and reagents
1
were purchased from commercial sources and were used without further purification. H
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
NMR chemical shifts are reported relative to residual proton solvent peaks or TMS.
(S)-3-(4-amino-1,2,5-oxadiazol-3-yl)-4-(4-fluorobicyclo[4.2.0]octa-1(6),2,4-trien-7-yl)-
1
,2,4-oxadiazol-5(4H)-one (8). To a 5 L 4-neck round-bottomed flask, were charged (S)-
4-fluorobicyclo[4.2.0]octa-1,3,5-trien-7-amine hemi L-dibenzoyl tartaric acid salt
(
5•DBTA) (435 g, 1.38 mol, 1.00 equiv) and hydrogen chloride in 1,4-dioxane (4500 mL
of 4 M solution; 18 mol, 13.0 equiv.). The resulting mixture was stirred overnight at room
temperature. The solids were collected by filtration. This resulted in 200 g (crude) of (S)-
4-fluorobicyclo[4.2.0]octa-1,3,5-trien-7-amine hydrochloride as a white solid. Into a 10-L
4
-neck round-bottomed flask, was charged sequentially a solution of (S)-4-
fluorobicyclo[4.2.0]octa-1,3,5-trien-7-amine hydrochloride (200 g, 1.15 mol, 1.00 equiv)
in acetonitrile (2000 mL, 10 vol.), a solution of sodium hydroxide (61 g, 1.53 mol, 1.30
equiv) in water (2000 mL, 10 vol.), a solution of (E)-4-amino-N-hydroxy-1,2,5-
oxadiazole-3-carbonimidoyl chloride (6) (191 g, 1.18 mol, 1.00 equiv) in acetonitrile
ACS Paragon Plus Environment
29