Development of a Scalable Synthetic Route to BMS-986251. Part 1: Synthesis of the Cyclohexane Dicarboxylate Fragment

Article abstract of DOI:10.1021/acs.oprd.1c00124

The cyclohexane dicarboxylate unit of BMS-986251 (1), a potent and efficacious RORγt inverse agonist, was synthesized starting from Hagemann's ester in seven chemical transformations with five isolated intermediates. The synthesis involved an enzymatic ki

Full text of DOI:10.1021/acs.oprd.1c00124

pubs.acs.org/OPRD
Article
Development of a Scalable Synthetic Route to BMS-986251. Part 1:
Synthesis of the Cyclohexane Dicarboxylate Fragment
Sankar Kuppusamy, Aravind S. Gangu, Srinivas Kalidindi, Muthukrishnan Ponnusamy,
Shankar Tendulkar, Alla Venu, Senthil Palani, Vedhachalam Nagappan, Arun Vinodini,
Boguslaw Mudryk, Sanjeewa Rupasinghe, Candice L. Joe, John R. Coombs, William P. Gallagher,
Nathaniel Kopp, Francisco Gonzalez-Bobes,* Martin D. Eastgate, and Rajappa Vaidyanathan*
Cite This: Org. Process Res. Dev. 2021, 25, 1547−1555
Read Online
Metrics & More
Article Recommendations
ACCESS
ABSTRACT: The cyclohexane dicarboxylate unit of BMS-986251 (1), a potent and efficacious RORγt inverse agonist, was
synthesized starting from Hagemann’s ester in seven chemical transformations with five isolated intermediates. The synthesis
involved an enzymatic kinetic resolution, a two-step telescoped enol tosylation followed by carboxylation using a benign CO
surrogate for the installation of the second carboxylate functionality, and a Crabtree catalyst-mediated diastereoselective olefin
hydrogenation. This process was successfully demonstrated to produce 3.6 kg of compound 3.
KEYWORDS: enzymatic ester hydrolysis, carboxylation using a CO surrogate, Crabtree hydrogenation, nuclear hormone receptors, RORγt
13, which were subjected to a Pd-catalyzed carbonylation
reaction using CO gas4 to yield a regioisomeric mixture of
diesters 14. A diastereoselective reduction of this mixture using
Crabtree’s catalyst5,6 followed by TFA-mediated chemo-
selective cleavage of the t-butyl ester furnished the desired
methyl ester 4. The cyclohexane dicarboxylate unit 4 prepared
via this sequence was coupled with the tricyclic core 2 to
provide API 1 after hydrolysis of the methyl ester.
This sequence to synthesize 4 was demonstrated successfully
on a multigram scale. As the molecule progressed into
development, significantly higher quantities of materials were
required, which necessitated a refinement of this synthetic
route. This work needed to be accomplished expediently to
enable the supply of API for preclinical and clinical studies. All
of the intermediates in the synthetic route depicted in Scheme
2 (except 4) were oils and required chromatographic
purification. In addition, the absence of a chromophore in
this synthetic sequence posed analytical challenges. We
envisioned that the use of a phenyl ester at C4 would alleviate
these concerns by imparting crystallinity to several of the
intermediates, facilitating the use of HPLC analysis due to the
presence of a chromophore, and promoting faster hydrolysis
under milder conditions (after coupling with 2) compared to
the methyl ester analog. Therefore, dicarboxylate 3 containing
a phenyl ester moiety at C4 was chosen as our target. We
INTRODUCTION
■
BMS-986251 (1) was discovered and developed within Bristol-
Myers Squibb as a potent small molecule RORγt inverse
agonist.1 In order to supply the active pharmaceutical
ingredient (API) for preclinical and early clinical studies, we
needed a rapid, safe, and robust synthetic approach that would
allow for the large-scale production of the structurally complex
target 1. Retrosynthetically, 1 can be derived by coupling
advanced intermediate 2 with an appropriately substituted
enantiomerically pure cyclohexane dicarboxylate unit (Scheme
1). While the discovery approach to 1 involved the coupling of
the methyl ester 4 with the tricyclic fragment 2, we chose to
employ the phenyl ester 3 (vide infra). This manuscript details
the synthesis of the cyclohexane dicarboxylate unit 3, while the
synthesis of the tricyclic fragment 2 and its subsequent
coupling with the dicarboxylate fragment to ultimately produce
BMS-986251 (1) will be described in the next paper.
We envisioned that the absolute configuration at C1 and C2
of 3 could be established via an enzyme-mediated kinetic
resolution of an ester derived from readily available
Hagemann’s ester 6.2 A carbonylation reaction on intermediate
5 would install the phenyl ester functionality at C4, and a
directed diastereoselective hydrogenation would provide access
to the desired configuration at this position.
The discovery approach to the dicarboxylate fragment 43
commenced with the synthesis of t-butyl Hagemann’s ester
( )-10 in two steps from t-butyl acetoacetate 7 and methyl
vinyl ketone 8 (Scheme 2). This racemic compound was
separated by supercritical fluid chromatography (SFC), and
the desired enantiomer 11 was hydrogenated over Pd/C to
afford the cis keto-ester 12. Treatment of ketone 12 with a base
and PhNTf2 provided a mixture of regioisomeric enol triflates
Received: April 11, 2021
Published: June 15, 2021
https://doi.org/10.1021/acs.oprd.1c00124
Org. Process Res. Dev. 2021, 25, 1547−1555
© 2021 American Chemical Society
1547

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Scheme 1. Retrosynthetic Analysis for 1
Scheme 2. First-Generation Synthesis of 1 via Methyl Ester 4
sought to develop a process for the large-scale synthesis of 3
that would obviate the need for SFC separation as well as avoid
the use of CO gas to install the carboxylate functionality at C4.
Our efforts are described in this manuscript.
utilized pig liver esterase for this transformation;1 however, its
animal origin precluded its use in our synthesis. An array of
readily available enzymes was screened in an attempt to
selectively hydrolyze one of the enantiomers of 16. Gratify-
ingly, several enzymes were effective for this kinetic resolution
with esterase AR “Amano”8 (10 wt %) in 0.1 M aqueous tris
buffer (pH 8) providing the best result (desired acid
enantiomer 17 in ∼92% ee at ∼40% conversion). After
completion of the reaction, the enzyme was removed from the
reaction mixture by filtration after a charcoal treatment, and
the unreacted ester 18 was separated from the desired
carboxylic acid 17 by extraction with 2-MeTHF at pH 8.
Acidification of the aqueous layer followed by extraction with
2-MeTHF gave the required acid 17 in 92−94% ee. The
enantiomeric purity was upgraded to 99% ee by recrystalliza-
tion of the crude from ethyl acetate/n-heptane. This step was
RESULTS AND DISCUSSION
■
We envisioned that an enzymatic approach could be leveraged
to establish the absolute configuration at C1. Initial attempts at
an enzymatic resolution by hydrolysis of the t-butyl cyclo-
hexenone ester ( )-10 and its cyclohexanone analog were not
successful, thus prompting us to examine the corresponding
ethyl esters. Commercially available cyclohexenone ester 6 was
subjected to a transfer hydrogenation using Pd/C, formic acid
(2.0 equiv), and i-Pr2NEt (2.2 equiv)7 in 2-methyltetrahy-
drofuran (2-MeTHF) at 25−35 °C to furnish cis-cyclo-
hexanone ester ( )-16 in an 86% yield (along with ∼2.5%
trans-isomer ( )-19), as an oil. Our discovery colleagues had
1548
https://doi.org/10.1021/acs.oprd.1c00124
Org. Process Res. Dev. 2021, 25, 1547−1555

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Scheme 3. Olefin Hydrogenation and Enzymatic Ester Hydrolysis
successfully demonstrated on multiple 25 kg batches in the
pilot plant (Scheme 3).
of 8 °C. This transformation was successfully carried out on a
17 kg scale.
It was necessary to mask the carboxylic acid in 17 as an ester
in order to effect subsequent transformations. We chose to
protect it as the t-butyl ester since that would impart hydrolytic
stability to the molecule and provide the orthogonality needed
for selective deprotection in downstream processing. Several
conditions were attempted to effect the esterification; most of
them either led to low conversions (e.g., H2SO4/MgSO4/t-
BuOH provided ∼30% conversion) or led to epimerization of
the carboxyl center (e.g., DMAP/Boc2O/t-BuOH furnished a
2.8:1 ratio of 12:20 at 63% conversion). Eventually,
enantiomerically enriched carboxylic acid 17 was converted
to the corresponding t-butyl ester 12 in an 85% yield by
treatment with POCl3 (2.3 equiv), pyridine (12 equiv), and t-
BuOH9 in acetonitrile at 20−35 °C (Scheme 4). The trans-
isomer impurity 20 was also formed in <0.5% by HPLC area
due to partial epimerization and was carried through to the
subsequent steps.
In the discovery approach, the carboxylate unit at C4 was
introduced through conversion of ketone 12 to the
corresponding mixture of enol triflates 13 followed by a Pd-
catalyzed carboxylation with CO gas. From a scalability
standpoint, this approach presented two major challenges:
(1) the enol triflate preparation was capricious and led to
several impurities, and (2) the use of CO gas on scale was not
preferred. We hypothesized that the sequence could be carried
out via enol tosylates 5 (which would be more stable than the
corresponding enol triflates 13 and would also be amenable to
monitoring by HPLC due to the presence of a chromophore).
Further, carboxylation using a CO surrogate such as HCO2Ph/
Et3N10−12 would obviate the need to use a large excess of CO
gas on scale (Scheme 5).
Enol tosylates 5 were initially synthesized via the addition of
NaHMDS (2 M in THF, 1.4 equiv) to a solution of 12 in THF
at below −50 °C to generate an isomeric mixture of the
enolate intermediates. A solution of Ts2O (1.75 equiv) in THF
(10 vol) was added to the enolate, maintaining the
temperature below −35 °C, and the mixture was allowed to
warm to ambient temperature. This approach worked well in
the laboratory, but the reaction provided inconsistent yields at
a 1 kg scale. This problem was solved by simply modifying the
addition sequence: addition of the enolate solution to a
solution of Ts2O in THF at below −5 °C (reverse addition)
led to clean conversion to enol tosylates 5. The reaction was
quenched with water, the mixture was extracted with toluene,
and the product-rich toluene layer was azeotropically dried and
carried through to the carboxylation reaction.
Scheme 4. t-Butyl Ester Synthesis
The addition of POCl3 to a mixture of carboxylic acid 17,
pyridine, and t-BuOH in acetonitrile at 10 °C was exothermic
with an adiabatic temperature rise of 68 °C and was found to
be dosing-controlled. The addition was controlled such that
the temperature of the reaction mixture did not exceed 25 °C.
The mixture after product formation was thermally stable in
the operating range and exhibited only a very minor self-
heating event (1 °C/min) at 110 °C by ARSST. The reaction
was quenched by the addition of water, which was also found
to be only mildly exothermic with an adiabatic temperature rise
To the solution of 5 in toluene, HCO2Ph (2.5 equiv), Et3N
(2.5 equiv), Pd(OAc)2 (0.03 equiv), and Xantphos (0.06
equiv) were added and heated at 85−95 °C for 20 h to provide
phenyl esters 21 as a 1:1 mixture of inconsequential
regioisomers after workup, treatment with SiliaMetS Thiol
(in order to remove residual Pd that could interfere with the
subsequent hydrogenation), and crystallization from aqueous
methanol. It is important to note that CO gas is generated in
situ during the reaction, and hence, the reactor outlet was
Scheme 5. Enol Tosylate Formation Followed by Carboxylation
1549
https://doi.org/10.1021/acs.oprd.1c00124
Org. Process Res. Dev. 2021, 25, 1547−1555

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Table 1. Effect of Applied Hydrogen Pressure and Crabtree Catalyst Loading on the Diastereoselective Hydrogenation
in process results (HPLC area %)
entry
catalyst loading (mol %)
applied H2 pressure (barg)
21
22
23
1
2
3
4
5
10
10
10
5
1
2.5
5
2.5
2.5
44.9
0.6
2.2
5.9
0.4
53.9
92.9
86.1
84.1
95.9
0.5
5.7
11.0
9.2
14
2.8
connected to an exhaust, and the processing area was equipped
with CO sensors. This telescoped process was scaled up to 9 kg
in the pilot plant to furnish the mixture of regioisomeric esters
21 as a crystalline solid in a 46% yield over two steps.
The discovery approach utilized a directed diastereoselective
hydrogenation mediated by Crabtree’s catalyst5,6 (7 mol %
Crabtree’s catalyst in CH2Cl2 at 1 barg H2 pressure) of a
regioisomeric mixture of methyl esters 14 to establish the
stereocenter at C4 (Scheme 2). In the process chemistry
approach (this work), the reduction of phenyl esters 21 reliably
went to completion on <10 g scales in a 200 mL hydrogenator
under these conditions. However, as the reaction scale was
increased beyond 100 g in a 10 L hydrogenator, the reaction
took much longer (up to 72 h) and led to higher levels of
unreacted 21 (up to 3% by HPLC area).13 Simultaneous
exploration of downstream processing conditions revealed that
compound 21 was difficult to purge in subsequent steps, and
consequently, we focused our efforts on identifying conditions
that would reproducibly lead to <1.0 area % 21 remaining at
the end of the reaction.
The lower selectivity (higher all-syn formation) observed at
higher pressures suggested a background reaction perhaps due
to catalyst decomposition. We postulated that this selectivity
issue could be addressed by increasing the catalyst loading.
Accordingly, the next two reactions depicted in Table 1 were
carried out under 2.5 barg pressure with varying catalyst
loadings. At a 5 mol % loading, both the starting material 21
and the all-syn isomer 23 were observed above their
acceptance criteria at 5.9 and 9.2 area %, respectively (entry
4). At a 14 mol % loading, the hydrogenation proceeded
smoothly to >99% conversion (∼0.4 area % 21 remaining),
with excellent selectivity (2.8 area % 23, entry 5); therefore,
these conditions were chosen for the scale-up runs.
The gas−liquid mass transfer coefficient (kLa) for the 2 L
lab-scale hydrogenators was measured to be <0.01 s−1. The kLa
for the pilot plant hydrogenator, as estimated using Dynochem
correlations, was >0.3 s−1; the higher kLa indicated better mass
transfer on scale. The hydrogenation was successfully carried
out in multiple batches up to an 8 kg scale in a 300 L
hydrogenator (15 mol % catalyst, 20 volumes of CH2Cl2 at
2.5−3 barg H2 pressure), and the product was isolated by
treatment with SiliaMetS Thiol (to reduce the Ir content from
∼11,000 to <400 ppm in isolated 22) and crystallization from
aqueous methanol. Under these conditions, phenyl esters 21
and the all-syn diastereomer 23 were controlled to <0.5 and <5
area %, respectively. The next step (hydrolysis of the t-butyl
ester, vide infra) effectively purged the all-syn diastereomer 23
completely and also reduced the residual Ir content to <10
ppm.
We examined the effect of hydrogen pressure (1, 2.5, and 5
barg) and catalyst loading (5, 10, and 14 mol %) on the rate
and selectivity in the Crabtree hydrogenation, and our results
are summarized in Table 1. These reactions were carried out in
14
10 volumes of CH2Cl2 wrt the phenyl esters 21, in a 2 L
hydrogenator with 30% occupancy. This occupancy was
chosen considering the average batch sizes in the plant; higher
occupancies (up to 60%) and lower concentrations (20
volumes of CH2Cl2) in laboratory-scale experiments exhibited
similar reaction profiles. In general, the reactions were clean
and produced only the desired compound 22 and the all-syn
diastereomer 23. In the first three reactions (entries 1−3,
Table 1), the catalyst loading was held constant at 10 mol %,
while the hydrogen pressure was varied. At 1 barg, the reaction
remained incomplete after 7 h (∼45 area % starting material
21, entry 1). As the pressure was increased to 2.5 barg, the
reaction proceeded to completion (0.55 area % 21 remaining
after 7 h, entry 2). However, up to 5.7 area % of the all-syn
isomer 23 was also observed under these conditions.
Interestingly, increasing the pressure to 5 barg not only led
to a slightly lower conversion (2.2 area % 21 remaining, entry
3) but also furnished ∼11 area % of the all-syn isomer 23.
While up to 10 area % 23 at this stage could be effectively
purged in downstream processing, its formation was
detrimental to the overall yield.
Selective hydrolysis of the t-butyl ester in 22 was
accomplished by treatment with trifluoroacetic acid15,16 in
CH2Cl2 (Scheme 6). Compound 3 was isolated in a 92% yield
(>99.9% purity; >99.8% ee) after recrystallization from
CH2Cl2/n-heptane; the all-syn diastereomer 23 carried from
the Crabtree hydrogenation was completely purged during this
step. The overall synthesis of 3 is depicted in Scheme 7.
Scheme 6. t-Butyl Ester Hydrolysis
1550
https://doi.org/10.1021/acs.oprd.1c00124
Org. Process Res. Dev. 2021, 25, 1547−1555

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Scheme 7. Synthesis of Compound 3
Melting points were obtained using a Buchi B-545 instrument.
HRMS was recorded on a Q Exactive Plus Orbitrap Thermo
mass spectrometer.
SUMMARY
■
We rapidly developed a scalable approach to the stereochemi-
cally complex cyclohexyl fragment of BMS-986251. This
compound was synthesized in seven steps from ethyl 2-
methyl-4-oxocyclohex-2-ene-1-carboxylate 6. The stereocen-
ters in this molecule were established as follows: a transfer
hydrogenation to set the relative stereochemistry at C1 and
C2, an enzyme-mediated kinetic resolution to generate the
absolute stereochemistry at C1 and C2, and a directed
diastereoselective hydrogenation to install the stereocenter at
C4. The key carboxylation reaction was effected using
HCO2Ph-Et3N as a CO surrogate, thereby circumventing the
need to use excess CO gas under pressure. The choice of the
phenyl ester was significant: it provided a chromophore that
rendered it amenable to HPLC analysis, which allowed for
excellent purity control and impurity tracking, imparted
crystallinity to several downstream intermediates, and most
importantly, could be cleaved under mild conditions after
coupling with the tricyclic core 2 (this is described in the next
manuscript). The process described in this manuscript was
successfully scaled up to produce 3.6 kg of dicarboxylate 3.
Ethyl (cis)-2-Methyl-4-oxocyclohexane-1-carboxylate
( )-16. To a 600 L Hastelloy hydrogenation reactor, 2-
MeTHF (330 L), ethyl 2-methyl-4-oxocyclohex-2-ene-1-
carboxylate 6 (33.0 kg, 181.1 mol), N,N-diisopropylethylamine
(51.5 kg, 398.5 mol, and 2.2 equiv), and 10% Pd/C 50% wet
(6.6 kg, 20% wt/wt) were charged sequentially under an N2
atmosphere. The reaction mixture was cooled to 10−15 °C,
and formic acid (17.0 kg, 369.3 mol, and 2.0 equiv) was added
to the reactor under a N2 atmosphere. The reaction mass was
warmed to 25−35 °C and stirred for 6 h. After completion of
the reaction, the catalyst was removed by filtration through a
bed of Celite. The reactor was rinsed with 2-MeTHF (230 L),
and the rinse was filtered through the Celite bed and combined
with the product-containing filtrate in a separate 600 L
Hastelloy reactor. The solution was washed sequentially with
1.5 N aqueous hydrochloric acid (330 L), water (330 L), 10%
aqueous NaHCO3 (2 × 165 L), and 10% aqueous NaCl (2 ×
165 L). The organic layer was concentrated in vacuo (at less
than 55 °C) until the 2-MeTHF content was less than 5% by
GC (actual 2-MeTHF content, 1.58%) to afford 30.4 kg of
ethyl (cis)-2-methyl-4-oxocyclohexane-1-carboxylate ( )-16 as
a colorless liquid (86% assay-corrected yield, 97.3% purity by
GC area %, 2.65% trans-isomer ( )-19, and 94.1% assay).
1H NMR (400 MHz, CDCl3): δ = 4.09 (m, 2H), 2.77 (dt, J
= 8.4 Hz, 4.0 Hz, 1H), 2.48−2.39 (m, 2H), 2.35 (m, 2H),
2.27−2.18 (m, 1H), 2.11−1.90 (m, 2H), 1.19 (t, J = 7.2 Hz,
3H), 0.88 (d, J = 6.8 Hz, 3H).
EXPERIMENTAL SECTION
■
General Information. All reactions were performed under
a nitrogen atmosphere. Reaction conversion was measured by
GC and/or HPLC. Chemical shifts for protons are reported in
parts per million downfield from tetramethylsilane and are
referenced to residual protons in the NMR solvent (CDCl3 = δ
7.29). Chemical shifts for carbons are reported in parts per
million downfield from tetramethylsilane and are referenced to
the carbon resonances of the solvent (CDCl3 = δ 77.01).
1551
https://doi.org/10.1021/acs.oprd.1c00124
Org. Process Res. Dev. 2021, 25, 1547−1555

Organic Process Research & Development
pubs.acs.org/OPRD
Article
13C NMR (100 MHz, CDCl3): δ = 210.4, 173.2, 60.2, 46.7,
43.9, 38.6, 33.6, 24.5, 16.3, 14.0.
added portion-wise in 2 h intervals as follows: 32.8 kg of
pyridine (415 mol, 3.8 equiv) and 8.6 kg of POCl3 (56 mol, 0.5
equiv); 18.6 kg of pyridine (235 mol, 2.1 equiv) and 5.0 kg of
POCl3 (33 mol, 0.3 equiv); 14.9 kg of pyridine (188 mol, 1.7
equiv) and 4.0 kg of POCl3 (26 mol, 0.24 equiv). After
completion of the reaction, the mixture was cooled to <15 °C,
and 180 L of purified water was added maintaining the
temperature below 25 °C. The mixture was extracted with
MTBE (3 × 86 L), and the combined organic layer was
washed with 1 N aqueous HCl (3 × 86 L), 5% aqueous citric
acid (2 × 690 L), 10% aqueous sodium bicarbonate (2 × 86
L), purified water (2 × 86 L), and saturated aqueous NaCl (86
L). The organic layer was concentrated in vacuo until the
MTBE content was less than 5% by GC to afford 18.7 kg of
tert-butyl (1R,2S)-2-methyl-4-oxocyclohexane-1-carboxylate 12
(84.5% yield, 98.4% purity by chiral column GC) as a viscous
liquid.
HRMS (ESI) (m/z): [M + H]+ calcd for C10H17O3,
185.1172; found, 185.1170.
(1R,2S)-2-Methyl-4-oxocyclohexane-1-carboxylic
Acid 17. Purified water (270 L) and tris(hydroxymethyl)-
aminomethane (3.3 kg, 27.0 mol) were added to a 1500 L
glass-lined reactor at 20−30 °C and stirred for 5 min. The pH
of the mass was adjusted to less than 8.1 by adding
hydrochloric acid (11 M, 1.5 L). Forty kilograms of this
solution (tris buffer) was unloaded and kept aside, and the
remaining solution was heated to 32−37 °C in the reactor.
Ethyl 2-methyl-4-oxocyclohexane-1-carboxylate ( )-16 (13.5
kg, 73.3 mol) was charged to the reactor followed by a solution
of esterase AR “Amano” (1.35 kg) in 30.5 L of tris buffer, and
the resulting mixture was stirred at 32−37 °C for 10 h. During
this time, the pH of the reaction was maintained between 7.5
and 8.5 by the addition of 5 N aqueous sodium hydroxide
every 30 min. At the end of the 10 h stirring period, the
reaction mass was cooled to 25−30 °C, and 3.4 kg of activated
charcoal was charged. The slurry was stirred for 1 h, filtered
through a Celite bed, and washed with 68 L of 50% aqueous 2-
MeTHF. The combined filtrate was extracted with 2-MeTHF
(2 × 135 L) to remove the unreacted starting material. The
aqueous layer was acidified with conc. HCl (4.5 L) to pH 2−3
and extracted with 2-MeTHF (4 × 68 L). The combined
organic layer was filtered through a Celite bed, and the bed
washed with 68 L of 2-MeTHF. The filtrate was washed with
saturated brine solution (2 × 34 L) and concentrated in vacuo
to a final volume of 13 L. n-Heptane (34 L) was added, and the
reaction mass was concentrated in vacuo to a final volume of
13 L. n-Heptane (34 L) was charged to the concentrated mass
and stirred at 20−35 °C for 1 h, and resulting slurry was
filtered to give 4.6 kg of crude (1R,2S)-2-methyl-4-
oxocyclohexane-1-carboxylic acid 17 as a white crystalline
solid.
1H NMR (400 MHz, CDCl3): δ = 2.77 (dt, J = 8.8 Hz, 4.4
Hz, 1H), 2.55−2.39 (m, 4H), 2.30 (ddd, J = 6.4 Hz, 9.2 Hz,
14.8 Hz, 1H), 2.16−1.96 (m, 2H), 1.48 (s, 9H), 0.98 (d, J =
6.8 Hz, 3H).
13C NMR (100 MHz, CDCl3): δ = 210.8, 172.7, 80.6, 47.0,
44.9, 38.8, 33.9, 28.0, 24.6, 16.3.
HRMS (ESI) (m/z): [M−H]+ calcd for C12H19O3,
211.1340; found, 211.1336.
Mixture of 4-(tert-Butyl) 1-Phenyl (4R,5S)-5-Methyl-
cyclohex-1-ene-1,4-dicarboxylate and 4-(tert-Butyl) 1-
Phenyl (3S,4R)-3-Methylcyclohex-1-ene-1,4-dicarboxy-
late 21. Tosylation. To a 600 L Hastelloy reactor, THF (66
L, water content < 0.05%) and tert-butyl (1R,2S)-2-methyl-4-
oxocyclohexane-1-carboxylate 12 (6.0 kg, 28.3 mol) were
charged and stirred at 20−30 °C. After 5 min, the solution was
cooled to −60 °C, and a 2 M solution of NaHMDS in THF
(22.2 kg, 48.3 mol, and 1.7 equiv) was added slowly while
maintaining the temperature below −45 °C. At the end of the
addition, the mixture was cooled to −60 °C and transferred
slowly to a precooled (−10 to −5 °C) solution of p-
toluenesulfonic anhydride (16.1 kg, 49.5 mol, and 1.75 equiv)
in THF (60 L, water content < 0.05%) in a separate 600 L
Hastelloy reactor while maintaining the temperature below 5
°C. The resulting thick slurry was warmed to 20−30 °C over 2
h and was stirred for an additional 2 h. The reaction was
quenched by the addition of purified water (36 L) and
extracted twice with toluene (60 and 36 L). The combined
organic layer was washed with aqueous NaCl solution (4 × 30
L) and concentrated in vacuo to a final volume of 20 L.
Toluene (60 L) was added, and the mixture was concentrated
in vacuo to 20 L. This operation was repeated two more times
to provide a solution of tosylates 5 in toluene (<0.5% v/v
THF; <1.0% water).
Carboxylation. The solution of tosylates 5 in toluene in a
600 L Hastelloy reactor was subjected to five vacuum/N2
purge cycles to remove headspace oxygen. Phenyl formate (8.7
kg, 71.2 mol, and 2.5 equiv wrt ketone 12) and triethylamine
(7.2 kg, 71.2 mol, and 2.5 equiv) were charged to the vessel,
and the resulting solution was sparged with nitrogen via a dip
tube for 10 min. Pd(OAc)2 (192.0 g, 0.8 mol, and 0.03 equiv)
was added followed by Xantphos (1.0 kg, 1.7 mol, and 0.06
equiv) and then sparged with nitrogen via a dip tube for 20
min. The reaction mixture was maintained at 85−95 °C for 20
h (Caution: CO gas is evolved in the reaction. The reactor
outlet was connected to an exhaust, and the processing area
was equipped with CO sensors). The reaction mixture was
This solid was added to a Hastelloy reactor containing ethyl
acetate (10 L), and the solution was maintained at 45−55 °C
for 30 min. n-Heptane (50 L) was slowly added into the mass
at 45−55 °C, and the slurry was cooled to 20−30 °C and
allowed to granulate for 3 h. The resulting solid was filtered,
washed with 67 L of n-heptane, and deliquored for 4 h to give
4.2 kg of (1R,2S)-2-methyl-4-oxocyclohexane-1-carboxylic acid
17 as a white crystalline solid (33% assay-corrected yield,
99.68% purity by chiral column GC, and 90.80% assay).
1H NMR (400 MHz, CDCl3): δ = 10.10 (br s, 1H), 2.92 (m,
1H), 2.62−2.40 (m, 4H), 2.39−2.26 (m, 1H), 2.24−2.03 (m,
2H), 1.02 (d, J = 7.2 Hz, 3H).
13C NMR (100 MHz, CDCl3): δ = 210.7, 179.2, 46.8, 43.9,
38.6, 33.5, 24.3, 16.5.
HRMS (ESI) (m/z): [M−H]+ calcd for C8H11O3, 155.0714;
found, 155.0704.
tert-Butyl (1R,2S)-2-Methyl-4-oxocyclohexane-1-car-
boxylate 12. To a 1000 L glass-lined reactor, dry acetonitrile
(180 L), (1R,2S)-2-methyl-4-oxocyclohexane-1-carboxylic acid
17 (17.2 kg, 94.5% assay, and 110.0 mol), pyridine (89.5 kg,
1131 mol, and 10.3 equiv), and dry tert-butyl alcohol (73.4 kg,
990 mol, and 9.0 equiv) were charged sequentially under a
nitrogen atmosphere. The reaction mass was cooled to 0−10
°C, and POCl3 (23.9 kg, 156.1 mol, and 1.42 equiv) was added
maintaining the temperature below 25 °C. The reaction
mixture was stirred at 20−35 °C, and the progress of the
reaction was monitored by GC. Pyridine and POCl3 were
1552
https://doi.org/10.1021/acs.oprd.1c00124
Org. Process Res. Dev. 2021, 25, 1547−1555

Organic Process Research & Development
pubs.acs.org/OPRD
Article
cooled to ambient temperature, and a kicker charge of phenyl
formate (1.3 kg, 11.0 mol), triethylamine (1.1 kg, 11.0 mol),
Pd(OAc)2 (29.7 g, 0.13 mol), and Xantphos (154.5 g, 0.26
mol) was added. The reaction mixture was heated at 85−95 °C
for an additional 14 h, cooled to 20−30 °C, and washed with
purified water (31 L), 0.1 N aqueous HCl, (33 L), 10%
aqueous N-acetylcysteine (33 L), 10% aqueous LiOH twice
(23 and 13 L),17 and purified water (3 × 30 L). The organic
layer was stirred with 3 kg of silica gel (230−400 mesh) for 1 h
at 20−30 °C and filtered through a Nutsche filter. The filter
cake was washed with toluene (24 L), and the combined
filtrate was concentrated in vacuo to 19 L under reduced
pressure, diluted with n-heptane (72 L) and filtered through
Celite to remove fine particulate residues, and concentrated in
vacuo to 10 L. n-Heptane (30 L) and ethyl acetate (30 L) were
added followed by SiliaMetS Thiol (1.5 kg) and activated
charcoal (1.5 kg), and the resulting slurry was stirred for 2 h
and filtered through Celite. The filtrate was concentrated in
vacuo and swapped with methanol (3 × 50 L) to the final
volume of 30 L (<0.5% v/v each of toluene, n-heptane, and
ethyl acetate by Raman analysis). Purified water (11 L) was
added to this solution over 20 min, and the resulting slurry was
allowed to granulate at 0−5 °C for 2 h and filtered. The cake
was washed with water (16 L), and the solid was dried at 55 °C
under vacuum for 16 h to give 4.5 kg of a mixture of 4-(tert-
butyl) 1-phenyl (4R,5S)-5-methylcyclohex-1-ene-1,4-dicarbox-
ylate and 4-(tert-butyl) 1-phenyl (3S,4R)-3-methylcyclohex-1-
ene-1,4-dicarboxylate (phenyl esters 21, 46.2% yield from 12,
93.41% purity by HPLC area) as a pale yellow solid.
the cake was washed with a mixture of dichloromethane (8 L)
and n-heptane (79 L). The combined filtrate was concentrated
in vacuo to 24 L and swapped with methanol twice (79 and 64
L) to a final volume of 24 L. An additional 64 L of methanol
was added, the contents were cooled to 5−15 °C, 79 L of
purified water was added, and the resulting slurry was allowed
to age for 2 h. The precipitated solids were filtered through a
Nutsche filter, and the cake was rinsed with 79 L of water. The
material was dried under vacuum at 50−55 °C for 18 h to give
7.8 kg of 1-(tert-butyl) 4-phenyl (1R,2S,4R)-2-methylcyclohex-
ane-1,4-dicarboxylate 22 (90.0% yield, 93.3% purity by HPLC
area, 5.94% all-syn isomer 23, and 92.2% assay) as an off-white
solid.
1H NMR (400 MHz, CDCl3): δ = 7.36 (t, J = 7.2 Hz, 2H),
7.21 (t, J = 0.2 Hz, 1H), 7.04 (d, J = 7.6 Hz, 2H), 2.74 (tt, J =
3.6 Hz, 12.0 Hz, 1H), 2.52−2.38 (m, 2H), 2.16 (m, 1H), 2.00
(m, 1H), 1.88−1.78 (m, 2H), 1.70 (dq, J = 3.6 Hz, 12.0 Hz,
1H), 1.62−1.48 (m, 1H), 1.45 (s, 9H), 0.97 (d, J = 7.2 Hz,
3H).
13C NMR (100 MHz, CDCl3): δ = 174.6, 173.9, 150.8,
129.4, 125.7, 121.5, 80.1, 46.1, 37.4, 34.9, 29.8, 28.1, 27.8, 21.3,
13.9.
HRMS (ESI) (m/z): [M + H]+ calcd for C19H27O4,
319.1904; found, 319.1904.
(1R,2S,4R)-2-Methyl-4-(phenoxycarbonyl) Cyclohex-
ane-1-carboxylic Acid 3. Dichloromethane (17 L), 1-(tert-
butyl) 4-phenyl (1R,2S,4R)-2-methylcyclohexane-1,4-dicarbox-
ylate 22 (3.2 kg, 10.0 mol corrected for assay) and
trifluoroacetic acid (6.74 L) were charged into a 60 L glass-
lined reactor, and the mixture was stirred at 20−35 °C for 3 h.
Upon reaction completion the solution was concentrated in
vacuo to 9 L. n-Heptane (30 L) was added, and the solution
was concentrated to 9 L. This operation was repeated once
more. Finally n-heptane (30 L) was added to the residue, and
the resulting slurry was allowed to age at 20−35 °C for 1 h.
The solids were filtered, washed with n-heptane (19 L) and
deliquored for 6 h to afford 2.7 kg of crude 3.
1H NMR (300 MHz, CDCl3): δ = 7.36 (t, J = 7.8 Hz, 2H),
7.20 (m, 1.5H), 7.11 (m, 2.5H), 2.83 (m, 0.5H), 2.65−2.20
(m, 4.5H), 1.95 (m, 0.5H), 1.75 (m, 0.5H), 1.47 (s, 9H), 1.05
(d, J = 7.2 Hz, 1.5H), 0.96 (d, J = 6.9 Hz, 1.5H).
13C NMR (75 MHz, CDCl3): δ = 173.3, 173.2, 165.7, 165.6,
151.0, 145.1, 139.3, 129.4, 128.8, 128.1, 125.6, 121.7, 80.5,
80.4, 43.5, 42.6, 32.0, 31.2, 28.7, 28.1, 25.0, 24.1, 19.4, 15.4,
15.3.
HRMS (ESI) (m/z): [M + H]+ calcd for C19H25O4,
317.1747; found, 317.1744.
Crude 3 was dissolved in dichloromethane (4 L) in a 60 L
glass-lined reactor. n-Heptane (26 L) was added to the reactor
and the slurry was allowed to age at 20−35 °C for 1 h. The
resulting solids were isolated by filtration through a Nutsche
filter, and dried under vacuum at 50−55 °C for 18 h to afford
2.4 kg of (1R,2S,4R)-2-methyl-4-(phenoxycarbonyl) cyclo-
hexane-1-carboxylic acid 3 (91.6% yield, 99.98% purity by
HPLC area, 99.80% ee by chiral column HPLC, and 99.8%
assay) as a white crystalline solid.
1-(tert-Butyl) 4-Phenyl (1R,2S,4R)-2-Methylcyclohex-
ane-1,4-dicarboxylate 22. To a 300 L stainless steel
hydrogenation reactor equipped with an induction stirrer,
dichloromethane (162 L), a mixture of 4-(tert-butyl) 1-phenyl
(4R,5S)-5-methylcyclohex-1-ene-1,4-dicarboxylate and 4-(tert-
butyl) 1-phenyl (3S,4R)-3-methylcyclohex-1-ene-1,4-dicarbox-
ylate (phenyl esters 21, 7.9 kg, and 25.1 mol assay-corrected
charge), and Crabtree’s catalyst (3.2 kg, 4.0 mol) were charged
sequentially. The reactor was purged with nitrogen (applied
1.0 barg pressure and evacuated for not less than 5 min; two
cycles) and then hydrogen (applied 1.0 barg pressure and
evacuated for not less than 5 min; two cycles). The reactor was
then charged with hydrogen (2−3 barg pressure) and stirred
for 15 h at 25−35 °C. After completion of the reaction, the
mixture was transferred to a 600 L reactor, and 159 L of n-
heptane was added. The solution was concentrated in vacuo to
∼80 L. n-Heptane (159 L) was charged, and the mixture was
stirred for 1 h at 20−35 °C. This solution was filtered through
a Celite bed to remove the precipitated catalyst-related
residues. The reactor was rinsed with 79 L of n-heptane, and
the rinse was passed through the Celite bed. The combined
filtrate was transferred to a clean reactor. Dichloromethane (16
L) and SiliaMetS Thiol (8.1 kg) were added, and the mixture
was stirred for 2 h. The slurry was filtered through Celite, and
1H NMR (400 MHz, CDCl3): δ = 11.66 (br s, 1H), 7.41 (t,
J = 8.0 Hz, 2H), 7.26 (t, J = 7.6 Hz, 1H), 7.10 (d, J = 8.0 Hz,
2H), 2.81 (t, J = 11.6 Hz, 1H), 2.63 (m, 2H), 2.22 (d, J = 11.2
Hz, 1H), 2.09 (d, J = 13.6 Hz, 1H), 2.03−1.72 (m, 3H), 1.62
(q, J = 13.6 Hz, 1H), 1.07 (d, J = 6.8 Hz, 3H).
13C NMR (100 MHz, CDCl3): δ = 180.9, 174.3, 150.7,
129.3, 125.7, 121.4, 45.2, 37.2, 34.6, 29.5, 27.4, 21.1, 14.0.
HRMS (ESI) (m/z): [M−H]+ calcd for C15H17O4,
261.1132; found, 261.1133.
AUTHOR INFORMATION
■
Corresponding Authors
Francisco Gonzalez-Bobes − Chemical Process Development,
Bristol-Myers Squibb, New Brunswick, New Jersey 08903,
United States; orcid.org/0000-0001-7876-7154;
Email: francisco.gonzalezbobes@bms.com
1553
https://doi.org/10.1021/acs.oprd.1c00124
Org. Process Res. Dev. 2021, 25, 1547−1555

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Rajappa Vaidyanathan − Chemical Development and API
Supply, Biocon Bristol-Myers Squibb Research and
Development Center, Bangalore 560099, India; orcid.org/
0000-0002-2236-5719; Email: rvaidyanathan@
seagen.com
ACKNOWLEDGMENTS
■
We would like to thank Jayaprakash Karamil, Sabuj Mukherjee,
Raghukumar Vellingiri, and Prantik Maity for their insightful
suggestions during the course of this work. Process safety
testing was carried out by Thirumalai Lakshminarasimhan. We
thank Abhishek Shrikant for process engineering support for
the enzymatic hydrolysis step. Analytical support from
Haranath Voguri, Mallikarjun Narayanam, Saravanan Natar-
ajan, Manjunatha Reddy, and Laurent Lehmann is gratefully
acknowledged. Our sincere thanks to David Kronenthal and
Robert Waltermire for their support of this work.
Authors
Sankar Kuppusamy − Chemical Development and API
Supply, Biocon Bristol-Myers Squibb Research and
Development Center, Bangalore 560099, India
Aravind S. Gangu − Chemical Development and API Supply,
Biocon Bristol-Myers Squibb Research and Development
Center, Bangalore 560099, India
Srinivas Kalidindi − Chemical Development and API Supply,
Biocon Bristol-Myers Squibb Research and Development
Center, Bangalore 560099, India
Muthukrishnan Ponnusamy − Chemical Development and
API Supply, Biocon Bristol-Myers Squibb Research and
Development Center, Bangalore 560099, India
Shankar Tendulkar − Chemical Development and API
Supply, Biocon Bristol-Myers Squibb Research and
Development Center, Bangalore 560099, India
Alla Venu − Chemical Development and API Supply, Biocon
Bristol-Myers Squibb Research and Development Center,
Bangalore 560099, India
Senthil Palani − Chemical Development and API Supply,
Biocon Bristol-Myers Squibb Research and Development
Center, Bangalore 560099, India
Vedhachalam Nagappan − Chemical Development and API
Supply, Biocon Bristol-Myers Squibb Research and
Development Center, Bangalore 560099, India
Arun Vinodini − Chemical Development and API Supply,
Biocon Bristol-Myers Squibb Research and Development
Center, Bangalore 560099, India
REFERENCES
■
(1) Cherney, R. J.; Cornelius, L. A. M.; Srivastava, A.; Weigelt, C. A.;
Marcoux, D.; Duan, J. J.-W.; Shi, Q.; Batt, D. G.; Liu, Q.; Yip, S.; Wu,
D.-R.; Ruzanov, M.; Sack, J.; Khan, J.; Wang, J.; Yarde, M.; Cvijic, M.
E.; Mathur, A.; Li, S.; Shuster, D.; Khandelwal, P.; Borowski, V.; Xie,
J.; Obermeier, M.; Fura, A.; Stefanski, K.; Cornelius, G.; Tino, J. A.;
Macor, J. E.; Salter-Cid, L.; Denton, R.; Zhao, Q.; Carter, P. H.; Dhar,
T. G. M. Discovery of BMS-986251: a clinically viable, potent and
selective RORγt inverse agonist. ACS Med. Chem. Lett. 2020, 11,
1221−1227.
(2) Pollini, G. P.; Benetti, S.; Risi, C. D.; Zanirato, V. Hagemann’s
ester: a timeless building block for natural product synthesis.
Tetrahedron 2010, 66, 2775−2802.
(3) Cornelius, L. A. M.; Li, J.; Smith, D.; Krishnananthan, S.; Yip, S.;
Wu, D.-R.; Pawluczyk, J.; Aulakh, D.; Sarjeant, A. A.; Kempson, J.;
Tino, J. A.; Mathur, A.; Dhar, T. G. M.; Cherney, R. J. Synthesis of 1-
(tert-butyl) 4-methyl (1R,2S,4R)-2-methylcyclohexane-1,4-dicarbox-
ylate from Hagemann’s tert-butyl ester for an improved synthesis of
BMS-986251. J. Org. Chem. 2020, 85, 10988−10993.
(4) Cacchi, S.; Morera, E.; Ortar, G. Palladium-catalyzed carbon-
ylation of enol triflates. A novel method for one-carbon homologation
of ketones to α, β-unsaturated carboxylic acid derivatives. Tetrahedron
Lett. 1985, 26, 1109−1112.
(5) Crabtree, R. H.; Davis, M. W. Directing effects in homogeneous
hydrogenation with [Ir(cod)(PCy3)(py)]PF6. J. Org. Chem. 1986, 51,
2655−2661.
Boguslaw Mudryk − Chemical Process Development, Bristol-
Myers Squibb, New Brunswick, New Jersey 08903, United
States
̀
(6) (6.) Verendel, J. J.; Pamies, O.; Diéguez, M.; Andersson, P. G.
Asymmetric hydrogenation of olefins using chiral Crabtree-type
catalysts: scope and limitations. Chem. Rev. 2014, 114, 2130−2169.
(7) Cortese, N. A.; Heck, R. F. Palladium-Catalyzed Reductions of α,
β-unsaturated carbonyl compounds, conjugated dienes, and acety-
lenes with trialkylammonium formates. J. Org. Chem. 1978, 43, 3985−
3987.
(8) Esterase AR “Amano” was purchased from Amano Enzyme, Inc.
(9) Banville, J.; Remillard, R.; Ruediger, E. H.; Deon, D. H.; Gagnon,
M.; Dube, L.; Guy, J.; Priestley, E. S.; Posy, S. L.; Maxwell, B. D.;
Wong, P. C. Imidazothiadiazole and imidazopyrazine derivatives as
protease activated receptor 4 (part 4) inhibitors for treating platelet
aggregation. WO 2,013,163,279 A1, 2013.
(10) Ueda, T.; Konishi, H.; Manabe, K. Palladium-catalyzed
carbonylation of aryl, alkenyl and allyl halides with phenyl formate.
Org. Lett. 2012, 14, 3100−3103.
(11) Fujihara, T.; Hosoki, T.; Katafuchi, Y.; Iwai, T.; Terao, J.; Tsuji,
Y. Palladium-catalyzed esterification of aryl halides using aryl formates
without the use of external carbon monoxide. Chem. Commun. 2012,
48, 8012−8014.
(12) Ueda, T.; Konishi, H.; Manabe, K. Preparation of α, β-
unsaturated esters and amides via external-CO-free palladium-
catalyzed carbonylation of alkenyl tosylates. Tetrahedron Lett. 2012,
53, 5171−5175.
(13) Xu, Y.; Mingos, D. M. P.; Brown, J. M. Crabtree’s catalyst
revisited: Ligand effects on stability and durability. Chem. Commun.
2008, 199−201.
(14) Attempts to substitute CH2Cl2 with more sustainable options
were met with limited success. Crabtree’s catalyst is known to form
Sanjeewa Rupasinghe − Chemical Process Development,
Bristol-Myers Squibb, New Brunswick, New Jersey 08903,
United States
Candice L. Joe − Chemical Process Development, Bristol-
Myers Squibb, New Brunswick, New Jersey 08903, United
States; orcid.org/0000-0002-7167-2416
John R. Coombs − Chemical Process Development, Bristol-
Myers Squibb, New Brunswick, New Jersey 08903, United
States; orcid.org/0000-0003-1473-2651
William P. Gallagher − Chemical Process Development,
Bristol-Myers Squibb, New Brunswick, New Jersey 08903,
United States; orcid.org/0000-0001-7472-0463
Nathaniel Kopp − Chemical Process Development, Bristol-
Myers Squibb, New Brunswick, New Jersey 08903, United
States
Martin D. Eastgate − Chemical Process Development, Bristol-
Myers Squibb, New Brunswick, New Jersey 08903, United
States; orcid.org/0000-0002-6487-3121
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.oprd.1c00124
Notes
The authors declare no competing financial interest.
1554
https://doi.org/10.1021/acs.oprd.1c00124
Org. Process Res. Dev. 2021, 25, 1547−1555

Organic Process Research & Development
pubs.acs.org/OPRD
Article
catalytically inactive species in nonpolar solvents. Similarly, the
Crabtree catalyst-mediated reaction is known to proceed sluggishly in
coordinating solvents. Please see (a)) Crabtree, R. Iridium
compounds in catalysis. Acc. Chem. Res. 1979, 12, 331−337.
(b)) Crabtree, R. H.; Demou, P. C.; Eden, D.; Mihelcic, J. M.;
Parnell, C. A.; Quirk, J. M.; Morris, G. E. Dihydrido olefin and
solvento complexes of iridium and the mechanisms of olefin
hydrogenation and alkane dehydrogenation. J. Am. Chem. Soc. 1982,
104, 6994−7001. (c)) Crabtree, R. H.; Davis, M. W. Occurrence and
origin of a pronounced directing effect of a hydroxyl group in
hydrogenation with [Ir(cod)P(C6H11)3(py)]PF6. Organometallics
1983, 2, 681−682. (d)) Nell, P. G. Crabtree’s Catalyst: [Ir(cod)pyr-
(PCy3)]PF6. Synlett 2001, 160.
(15) Marcoux, D.; Beaudoin Bertrand, M.; Dhar, T. G. M.; Yang, M.
G.; Xiao, Z.; Xiao, H.-Y.; Zhu, Y.; Weigelt, C. A.; Batt, D. G.
Preparation of tricyclic sulfones as RORγ modulators. US
20,180,127,368, 2018.
(16) Horiguchi, M.; Hayashi, M.; Saito, Y.; Teng, Y.; Tong, C. J.;
Fang, Y. Carboxylic acid compound, production method thereof, and
liquid crystal composition using the compound. JP 2,016,216,433 A,
2016.
(17) The base wash was introduced in order to remove the phenol
formed in the reaction (from excess phenyl formate). The use of 10%
aq NaOH led to partial hydrolysis of phenyl ester 21; however, with
10% aq LiOH, product hydrolysis was not observed.
1555
https://doi.org/10.1021/acs.oprd.1c00124
Org. Process Res. Dev. 2021, 25, 1547−1555