You are on page 1of 12

pubs.acs.

org/OPRD Article

Toward the Development of a Manufacturing Process for Milvexian:


Scale-Up Synthesis of the Side Chain
Simon Wagschal,* Diego Broggini,* Trung D.C. Cao, Pascal Schleiss, Kristian Paun, Jessica Steiner,
Anna-Lena Merk, Joachim Harsdorf, Winfried Fiedler, Stefan Schirling, Sven Hock, Tobias Strittmatter,
Jan Dijkmans, Ivan Vervest, Tim Van Hoegaerden, Brecht Egle, Matthew P. Mower, Zhi Liu,
Zhiyong Cao, Xiaoning He, Lei Chen, Lei Qin, Hongyu Tan, Jun Yan, Nicolas Lucien Cunier̀ e,
Carolyn S. Wei, Venkata Vuyyuru, Rajaram Ayothiraman, Sundaramurthy Rangaswamy, Mohamed Jaleel,
Rajappa Vaidyanathan, Martin D. Eastgate, Richard Klep, Cyril Benhaïm, Ilse Vogels, Koen Peeters,
and Sébastien Lemaire
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Downloaded via UNIVERSIDADE DO PORTO 01100 on November 23, 2023 at 17:17:29 (UTC).

Cite This: Org. Process Res. Dev. 2023, 27, 680−691 Read Online

ACCESS Metrics & More Article Recommendations *


sı Supporting Information

ABSTRACT: Anticoagulants play a critical role in the prevention and treatment of thrombotic-driven cardiovascular diseases.
Factor XIa (FXIa) inhibitors have the potential to improve the benefit/risk profile of existing anticoagulants through a safer bleeding
profile in a variety of conditions where patients are predisposed to a high risk of thrombotic or bleeding events. To support the
clinical development program of milvexian (BMS-986177/JNJ-70033093), a FXIa inhibitor that recently completed phase II clinical
trials, we improved the discovery route to deliver the suitable quantity of key intermediate 1 for clinical supply. This paper describes
our optimization of the Suzuki cross-coupling and how we simplified and improved the isolation of 4-trimethylsilyl-1,2,3-triazole 6
after the azidation−click sequence. On top of streamlining the processes for the chlorination and demethylation steps, we
demonstrated that the recrystallization of the penultimate intermediate 7 was key to control the purity and the color of the desired 4-
chloro-1,2,3-triazole 1, which could be obtained in a 70% yield over five steps.
KEYWORDS: milvexian, Suzuki, click, azide, chlorination, factor XIa inhibitor

■ INTRODUCTION
Cardiovascular diseases (CVDs) are among the leading causes
supply, prompting the team to further optimize the synthesis of
target 1 with cost and sustainability in mind.
of death worldwide.1 Despite significant advances in the Several aspects of the discovery route shown in Scheme 1
prevention and management of thrombosis, the residual risk of needed reinvestigation for the synthesis of compound 1 at the
ischemic events remains high. Genetic, pre-clinical, and clinical plant scale. The Pd-catalyzed Suzuki−Miyaura coupling of step
evidence suggest that FXIa inhibition has the potential to 1 required a 10 mol % Pd catalyst and afforded aniline 4 in a
reduce the burden of vascular and thromboembolic diseases moderate 56% yield. The click reaction required 3 equiv of
while preserving hemostasis.2 Milvexian (BMS-986177/JNJ-
expensive trimethylsilylacetylene (TMSA). The chlorination of
70033093) is a potentially first-in-class oral small-molecule
FXIa inhibitor, which recently completed phase II studies to compound 6 was performed with silica gel, and 10 equiv of
prevent thrombotic events in multiple patient populations.3 corrosive HBr in acetic acid were required in the final
Milvexian is a complex molecule that can be accessed from demethylation step. Overall, all intermediates including
4-chloro-1,2,3-triazole 1, a key intermediate first described in compound 1 were isolated using chromatography, making
the discovery route used to prepare milvexian (Scheme 1). this process not sustainable to deliver the suitable quantity for
Boronic ester 2 reacted with 4-chloro-6-methoxypyrimidine 3 clinical supply. Herein, we report our research in addressing
via a Suzuki−Miyaura coupling4 to deliver aniline 4. This
these shortfalls to enable successful deliveries and support
intermediate was further functionalized to the 4-TMS-
substituted triazole 6, employing a sequence of diazotization/ clinical studies.
azidation5 and regioselective click chemistry.6 A chlorination
reaction on compound 6 to exchange the trimethyl silyl group Received: December 23, 2022
with a chlorine atom followed by an O-demethylation provided Published: April 3, 2023
the desired compound 1. This enabling process showed
acceptable yields, quality, and good reproducibility at a lab
scale. However, the current route proved to be insufficient at
the plant scale to deliver the quantity suitable for clinical

© 2023 American Chemical Society https://doi.org/10.1021/acs.oprd.2c00399


680 Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

Scheme 1. Discovery Chemistry Route for 1

Table 1. Base Screening for the Suzuki-Miyaura Cross-Coupling between 2 and 3

entry base compound 4 (area %)a compound 11 (area %)a


1 K3PO4 52.5 18.2
2 AcOK 46.9 14.4
3 NaHCO3 50.9 6.9
4 DIPEA 53.7 3.9
5 Et3N 65.7 1.7
6 TMG 85.9 NDb
a
Area % on the crude reaction mixture after 5 h reaction time. bNot detected.

■ RESULTS AND DISCUSSION


Initial reaction optimization focused on the reaction selectivity
Table 2. Solvent Screening for the Suzuki−Miyaura Cross-
Coupling between 2 and 3
between the formation of the desired aniline 4 and undesired
Buchwald−Hartwig impurity 11, resulting from a C−N
coupling of 3 with 4 (Table 1). The purging for this impurity
required up to three HBr salt formation/free basing/re-slurry
cycles to reach the required >99.5 area % purity for compound
4. Screening the reaction conditions with different bases in 2-
MeTHF/water (3.6/1 v/v ratio) quickly identified 1,1,3,3-
tetramethylguanidine (TMG) for minimizing impurity 11.7
While NaHCO3 performed better than K3PO4 and AcOK as an entry solvent reaction purity (area %) in situ yield (%)
a
inorganic base (entries 1−3), the best outcome was obtained 1 2-MeTHF 93.3 98.1
when using organic bases such as DIPEA, Et3N, or TMG 2 MeCN 94.9 100
(entries 4−6).8 3 THFa 93.5 97.4
After further optimization, aniline 4 could be obtained in a 4 2-propanol 92.5 98.1
a
>98% in situ yield using 1 mol % Pd(dppf)Cl2·DCM in the Biphasic reaction mixture.
presence of 2 equiv TMG and 10 equiv water in 2-MeTHF
within 4 h at 65 °C (Table 2). However, the conversion
profiles obtained with this biphasic system were not performed similarly under the same reaction conditions,
reproducible, which led us to investigate water-miscible affording the desired product in a more than 97% in situ
solvents to avoid mixing effects on a larger scale and allow yield in all cases. MeCN was finally selected for further
direct anti-solvent crystallization after the addition of water. development due to the slightly higher purity and yield
Acetonitrile (MeCN), 2-propanol, and tetrahydrofuran (THF) obtained.
681 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

Figure 1. Same excess experiment showing the overlay between standard and same excess reactions, indicating no catalyst deactivation under these
conditions. Left: raw reaction curves before the time shift. Right: Same excess curve time-shifted to place the initial point on the curve of the
standard reaction.

We then performed a kinetic analysis of the reaction to when a Pd(II) salt is mixed with TMS-azide in MeCN/2-
assess the robustness of the optimized catalytic conditions. The MeTHF. It precipitates as a brown solid that explodes on
reaction progress was monitored with both ReactIR and UPLC touch if allowed to dry. We first screened several metal
sampling (Figure 1). Kinetic interpretation of the data was scavengers (e.g., N-acetyl cysteine, EDTA, citric acid, and
accomplished using the technique of variable time normal- tributylphosphine) and found that 1,3,5-triazine-2,4,6-trithiol
ization analysis (VTNA).9 Catalyst robustness was first trisodium salt (TriNaTMT) was the best performer.12 After
investigated using a series of “same excess” experiments.10 reaction completion, an aqueous solution of TriNaTMT (10
The standard and same-excess curves displayed a good overlay, equiv relative to the Pd catalyst) was dosed at the reaction
indicating that no catalyst deactivation occurred during the temperature to trigger crystallization followed by a cooling
first 50% of the turnovers performed by the standard reaction. ramp from 65 to 10 °C to maximize the yield. The addition of
A series of “different excess” experiments revealed a first order TriNaTMT at the end of the reaction minimized the formation
in 3 and zeroth order in 4 (see the Supporting Information).
of the Buchwald−Hartwig amination impurity 11 from 4 by
The main impurities observed under those conditions were
quenching the active catalyst. We also observed a beneficial
the boronic ester protodeboronation 8 and homocoupling
side-product 9 along with the phosphine oxide 10 and effect of TriNaTMT on the product desaturation kinetics
previously mentioned Buchwald−Hartwig impurity 11 (Figure during the crystallization (2 h instead of 20 h without
2). Apart from 11, usually observed at 1−2 area %, all TriNaTMT). On a 100 g scale, these conditions afforded
impurities were formed below 1 area % and could be depleted compound 4 in a 90−95% isolated yield and >98 area % purity
below 0.1 area % after crystallizing 4 in MeCN/water. with 50−100 ppm residual palladium.
During our first scale-up, the conversion and crystallization
of the desired product matched the 100 g scale reaction, but
the Pd levels in the isolated product were approximately 1000
ppm. The isolation from MeCN/water also resulted in a low
weight assay (90−95% w/w) and long drying times due to the
high water content at the end of filtration.13 These issues were
resolved by dissolving wet solid 4 in 2-MeTHF (the next step’s
solvent) and performing two washes with an aqueous solution
of N-acetyl cysteine followed by a sodium bicarbonate and a
Figure 2. Main impurities identified in the Suzuki−Miyaura cross- brine wash. This procedure reliably decreased the residual
coupling. palladium levels to ≤100 ppm. The water content could be
reduced below 0.1% after an azeotropic distillation, thus
With the next step involving azide chemistry, we needed to reducing the cycle time by avoiding the need to dry compound
control the residual palladium to very low levels to avoid the 4. On multi-kilogram scale, the yield for this telescoped process
formation of highly shock-sensitive palladium azide11 in the was 93% with 97.7 area % purity, containing 0.1% w/w water
next step. We found that palladium azide is rapidly formed and 55 ppm residual palladium. The resulting intermediate 4

Scheme 2. Outcome of Azidation when Dosing tert-Butyl Nitrite Followed by TMS Azide

682 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

Figure 3. DSC thermogram of the isolated azide 5.

Figure 4. Influence of the copper source and base on the reduction of 5 to 4.

solution in 2-MeTHF was directly used in the subsequent with significant formation of an unknown solid later identified
azide formation step. as triazene 13 (Scheme 2).
Due to the poor aqueous solubility of aniline 4, the aqueous We hypothesized that the diazene intermediate 12 would
azidation conditions using sodium nitrite and sodium azide did react more selectively to form desired product 5 if the TMS
not work. 2-MeTHF and THF exhibited the highest solubility azide were present in excess. When adding slowly tert-butyl
for compounds 4 and 5, and 2-MeTHF was selected to allow nitrite to a solution containing 4 and 1.2 equiv TMS azide, the
better phase separation after the aqueous NaOH washes. reaction proceeded smoothly with full conversion and no
formation of 13 was observed. Since TMS azide reacts directly
MeCN as a co-solvent increased the azidation kinetics and
with tert-butyl nitrite, as evidenced by a gas release upon
improved the phase separation after the basic washes. A sharp
mixing the two reagents, a slight excess of both reagents was
contrast in the reaction outcome was observed when used in the process.
comparing two possible addition orders for the azidation From a safety perspective, all lab and plant scale reactions
step. Upon the addition of tert-butyl nitrite to a mixture of were run under a constant nitrogen flow to avoid accumulation
compound 4 in MeCN/2-MeTHF, intermediate 12 formed of the highly explosive and toxic hydrazoic acid (HN3) if it
rapidly, and a thick slurry was obtained. A subsequent slow were to form in the reactor head space.14 Furthermore, the
addition of TMS azide produced azide 5 in low conversion nitrogen flow was passed through a double aq. NaOH trap.
683 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

Table 3. Optimization of the Click Reaction of Azide 5 with TMS Acetylene

entry Cu source reductant base solvent 6 (area %) 4 (area %)


1 CuSO4−5H2O Na ascorbate none THF/H2O 82 17
2 CuSO4−5H2O Na ascorbate K2CO3 THF/H2O 93 3.6
3 CuSO4−5H2O Na ascorbate Et3N THF/H2O 95 2.1
4 CuI K2CO3 THF/H2O 98 0.7
5 CuI Et3N THF/H2O 99 0.4
6 CuI Et3N THF 99 0.4

Scheme 3. Optimized Process for the Azidation−Click Sequence to 4-TMS-1,2,3-triazole 6

Hydrazoic acid was indirectly measured as sodium azide in the h in the presence of CuSO4 and sodium ascorbate. These
NaOH trap. Hydrazoic acid content in solution was controlled control studies indicated that the reduction of 5 by sodium
prior to the click step by four consecutive aqueous NaOH ascorbate is catalyzed by CuSO4, which is consistent with
washes and one brine wash to avoid the formation of the literature reports of ortho-directed azide reduction occurring in
explosive copper azide Cu(N3)2. Each organic and aqueous the presence of copper.16 Since the role of sodium ascorbate is
layer was tested for azide content using ion chromatography, to reduce the Cu(II) source to the catalytically reactive Cu(I)
and the results were typically below 3 ppm after two washes. species, we reasoned that the replacement of CuSO4/sodium
The solubility of compound 5 turned out to be quite low in ascorbate with CuI could obviate the need for a reductant.
most organic solvents (toluene, EtOAc, IPAc, MEK, MIBK, Indeed, the reduction of 5 to 4 was reduced to ≤5 area % when
and MeCN), and the isolation of azide 5 was considered to CuI was the copper source (Figure 4). The role of base in the
reduce the solvent consumption. Thermal stability testing click reaction was also examined, and it was found that,
using differential scanning calorimetry (DSC) showed that although the CuSO4/sodium ascorbate conditions proceeded
compound 5 has a very high decomposition energy (1001 J/g) to full conversion in the absence of K2CO3, the amount of 4
with autocatalytic behavior (Figure 3). The maximum was nearly 17 area % after 3 h (Table 3, entry 1). In the
allowable temperature was set at 50 °C based on isothermal presence of K2CO3, the click reaction gave only ≤4 area % 4,
DSC experiments. While this intermediate was not shock- suggesting that the presence of a base inhibits azide reduction
sensitive according to the BAM fallhammer test, degradation (entry 2). Since K2CO3 has poor solubility in the reaction
under impact was observed (color change and partial triggering solvent THF, we switched to organic base Et3N and found a
of the decomposition exotherm). The potential of this further decrease in the amount of 4 (entry 3). Using a
compound to degrade by impact or friction combined with combination of CuI and Et3N, we were able to lower the
its unfavorable thermal stability profile prompted us to discard aniline impurity to ≤0.4 area % (entries 4 and 6). The final
the option of isolating 5 on the scale. The resulting 2-MeTHF conditions using CuI/Et3N in THF delivered up to 12 kg of
solution containing 5 was then used as such in the next step. triazole 7 with good control of the aniline impurity 4.
For the click chemistry development, we first evaluated the While the impurity profile and scalability were acceptable at
CuSO4/sodium ascorbate system to perform the coupling of the time, this sequence required further development as the
azide 5 with trimethylsilylacetylene and found that the initial clinical demand increased. We optimized the click step using
reaction conditions were satisfactory to support a rapid scale- the 2-MeTHF azide solution previously described, and the
up.15 However, the formation of aniline 4 from azide 5 was amount of TMS acetylene could be reduced from 4 to 1.2
observed between 3 and 20 area % during the plant-scale equiv in the presence of 10 mol % CuI and 2.2 equiv Et3N
synthesis, indicating a lack of reaction robustness. This without an impact on the reaction purity. After full conversion
prompted us to investigate the role of copper on the reduction to the click product 7, the reaction mixture was quenched with
of azide 5 to aniline 4 (Figure 4). In the absence of CuSO4, aqueous ammonia. After a celite bed filtration, the Pd and Cu
only a trace amount of 4 was observed in a mixture of 5 and levels (100 and 2000 ppm, respectively) remained the same,
sodium ascorbate. In contrast, 66 area % 4 was formed after 3 even after treatment with silica thiol. Additional ammonia
684 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

Figure 5. Major impurities observed during the formation of 6.

washes did not further decrease the metal content. We later


demonstrated that adding solid TriNaTMT at the end of the
reaction followed by a celite bed filtration, an aqueous
ammonia wash, and an aqueous sodium sulfate wash could
reliably reduce the copper level below 100 ppm. Finally, put-
and-take solvent switch from 2-MeTHF to heptane17 afforded
compound 6 in an acceptable purity and yield with a low Figure 6. Alternative chlorinating agents.
mother liquor loss (Scheme 3).
With this optimized process, we next turned our attention to
the other impurities present in 7 (Figure 5). On the lab scale, methanol and TCCA results in a mixture that is thermally
the des-amino impurity 14 formed during the azidation step unstable in the presence of metals.
was observed in low amounts (0.1−0.2 area %). However, in 1,3-Dichloro-5,5-dimethylhydantoin (DCDMH) provided
the first 1 kg scale up, the azidation produced up to 3.5 area % full conversion to compound 7 in 2 h in DMF with 0.3−0.5
area % des-TMS impurity 15. In other solvents such as NBP,
14. After investigation, we found that the nitrogen flow
MeCN, 2-MeTHF, and DCM, the reaction gave poor yields
through the reactor headspace removed part of the TMS azide
and selectivity. When looking more closely at the reaction
that was dosed on top of the reaction mixture. This
conditions, we found that 0.5 equiv water and 0.6 equiv
phenomenon was confirmed in the lab by undercharging
chlorinating agent were necessary to reach full conversion and
TMS azide. In the following batch, impurity 14 was controlled
a catalytic amount of sulfuric acid was not required. The
by performing sub-surface addition of TMS azide. We also
reaction temperature turned out to be key to control the
found that it was critical to control the pH of the solution amount of protodesilylation: 5 °C was found to be optimal,
containing azide 5 to below 9 to minimize the cleavage of the leading to 0.2 area % des-TMS impurity after 23 h (1 and 5
carbon-silicon bond of 6 leading to impurity 15. Additional area % were obtained at 15 and 25 °C, respectively). When
washes with aqueous saturated Na2SO4 solutions following the varying the amount of water between 0.4 and 2.0 equiv, we
aqueous NaOH wash effectively controlled the pH of the found that 0.5 equiv water was required for the reaction to
telescoped stream of 5. The oxidative dimer 16 was first proceed and additional water had minimal impact on the
observed at 1.6 area % after an oxygen ingression in one of the reaction. In order to ensure reaction robustness, the final
manufacturing batch. This impurity could not be purged conditions used 1.0 equiv water (Table 4, entries 4−7). We
during the isolation of compound 6 and was reacted in the next also examined higher amounts of DCDMH and found no
step to give the corresponding chlorinated impurity in significant impact on the purity (entries 4 and 5).
compound 7 (vide infra). Impurity 11 carried over from the From a safety perspective, the use of DCDMH is safer than
Suzuki coupling step was unreactive under the azidation and NCS (decomposition energy of 30 vs 104 J/g in 5 V of DMF).
click reaction conditions and had poor purge efficiency during The thermal stability of the reaction mixtures was assessed
the isolation of 6. On the scale, compound 6 was isolated as a using DSC since it has been reported that N-haloimides may
brown solid in a 90% yield, 99.4 area % purity, and 98.6% assay be incompatible with DMF.19 Using DCDMH as the
and contained 46 ppm Pd and 94 ppm Cu. halogenating agent instead of NCS resulted in a thermally
Optimization of the N-chlorosuccinimide (NCS) chlorina- more stable reaction mixture. An RC1 experiment with single-
tion of TMS-triazole 6 to compound 7 indicated that DMF portion addition of DCDMH was performed on a 30 g scale to
was by far the best solvent when compared to DMAc, EtOAc, obtain calorimetric data: The reaction is classified as a Stoessel
AcOH, DCM, MeCN, 2-MeTHF, and MTBE (Figure S4). N- class 1 process, meaning that it has a very low thermal risk.20
Butyl pyrrolidinone (NBP) provided slower conversion, a The addition of DCDMH in one portion was however
lower purity, and higher levels of des-TMS impurity 15 (up to discounted from a quality perspective since the potential
90 area % in 2-MeTHF/DMF). In order to use a more adiabatic temperature rise (24 °C) could result in a reaction
environmentally benign solvent, the reaction in MeCN was temperature that may generate significant amounts of the des-
optimized and found to proceed in the presence of water (0.5 TMS impurity in case of inadequate cooling. The option of
equiv) and sulfuric acid (0.1 equiv), but the purity was lower slow-dosing DCDMH in solution was evaluated in DMF, but
(89−91%) than in DMF (>98%). the concentration tested (2 L of DMF per kg of DCDMH) was
Other chlorinating agents were also evaluated to avoid the not stable as gas evolution was observed at elevated
use of DMF (Figure 6). Palau’chlor18 did not provide any temperatures. Portion-wise addition of a neat chlorinating
promising results, and trichloro isocyanuric acid (TCCA) in agent to the reaction mixture was then tested from −5 to 5 °C
methanol, although initially promising, was abandoned after and gave satisfactory results with the addition of three portions
compatibility experiments showed that the combination of of 0.3 equiv DCDMH each (Table 4, entries 6 and 7). The
685 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

Table 4. Impact of DCDMH and Water on the Chlorination of 6

entry scale (g) DCDMH (equiv) water (equiv) 6 (area %) 7 (area %) 15 (area %) isolated yield (%) purity (area %)
1 10 0.6 0.4 20.6 77.0 1.3 88.5 97.4
2 10 0.6 0.5 0 97.5 0.4 89.5 99.0
3 20 0.6 2.0 0 97.4 0.1 87.5 98.9
4 20 0.8 1.0 0 97.8 0.2 89.8 99.4
5 20 1.0 1.0 0 97.6 0.2 91.8 98.8
6 30 3 × 0.2 1.0 0 98.1 0.2 91.3 99.3
7 30 3 × 0.3 1.0 0 97.6 0.3 93.2 98.8

current conditions feature first the addition of water followed Table 5. Influence of the HCl Amount and Temperature on
by the addition of three portions of DCDMH (0.25 equiv the Conversion and Purity of the Demethylation of 7
each) at −5 °C to ensure reaction robustness. The reaction is
then warmed to 5 °C to reach full conversion within 2−3 h.
We improved the crystallization of compound 7 by seeding
after adding a first portion of water (0.5 L/kg). One hour after
seeding, the desaturation reached a plateau; adding the second
water portion (4.5 L/kg) over 4 h allowed the crystallization of
99% of compound 7. On the scale, this step performed well
and the impurity 15 was not observed during the reaction
conversion. As mentioned earlier, the oxidative dimer impurity
16 formed in the click step was chlorinated and the resulting entry HCl (equiv) temp (°C) time (h) conversion (%) 1 (area %)
impurity was not purged in the isolation of 7 from DMF/ 1 10 35 21 99.9 99.5
water. This prompted us to evaluate a recrystallization of 2 10 45 4 99.6 98.2
compound 7 in non-polar systems such as DCM/heptane.21 3 10 55 4 99.0 98.3
Dosing heptane slowly to a DCM solution of 7 at low 4 5 45 22 99.8 99.0
temperature efficiently removed all non-polar impurities such 5 15 45 5.5 100 99.5
as oxidative dimer derivatives and delivered 7 in a 99.9 area %
purity, starting from crude 7 with a 96.0 area % purity. In the due to the material sticking to the reactor wall. Addition of
long term, we implemented a charcoal treatment of the DCM water at 45 °C at the end of reaction followed by slow cooling
solution of 7 to remove the color coming from 6 and obtained to 5 °C turned out to be optimal with respect to the particle
7 as a white solid. On the scale, crude 7 was isolated in a 92% size, filterability, yield, and total volume (entry 3). Despite a
yield with a 99.3 area % purity, which was improved to 99.9 highly acidic pH, a low chlorine content measurement on the
area % and a 100% assay after the DCM/heptane isolated product suggested no retention of HCl. While direct
recrystallization (94% recovery). filtration of 1 in acidic aqueous media would be very attractive
For the last step, HCl was investigated as an alternative to from a productivity perspective, it would require specific plant
HBr for the O-demethylation of compound 7 to target 1 as it is equipment to avoid corrosion. Neutralizing the reaction
cheaper and less corrosive. When testing different temper- mixture prior to isolation would avoid these technical
atures with 10 equiv HCl, we found that full conversion could limitations but would also require a good pH control. Indeed,
be achieved at 35−55 °C overnight, giving 1 in excellent purity the phenol function of 1 is quite acidic (pKa of ∼7) and 1 is
in all cases (Table 5, entries 1−3). A reaction temperature of fully dissolved in water at pH 10. Attempts to crystallize the
45 °C was selected for further examination, and various product by acidification from basic pH were not satisfactory.
amounts of HCl were tested. Full conversion to 1 can be We tested several aqueous bases and found that potassium
reached overnight with 5 or 15 equiv of concentrated aqueous phosphate or ammonia offered a better pH control over NaOH
HCl (entries 4 and 5). However, the reaction volume was not (entries 4−6). In both cases, excellent yields were obtained,
sufficient for proper stirring with 5 equiv HCl (1.5 L/kg) and and ammonia was selected due the lower peak volume.
45 °C with 10 equiv HCl was chosen for process develop- On a 750 g scale starting from 98.5 area % pure 7, full
ment.22 conversion and 98.8 area % purity were achieved within 16 h
Those concentrated conditions (3 L/kg) offered the followed by isolation of compound 1 in a 96% assay-corrected
advantage of generating highly pure compound 1 in acidic yield and 99.5 area % purity. This straight-forward process was
aqueous media (pH = −1.5 at the end of the reaction).23 however not purging some of the impurities present in 7. As a
Target compound 1 was insoluble in water below pH 7 and control strategy, we developed an acetone/heptane recrystal-
crystallized at room temperature upon the addition of water or lization of 1 to increase the purity of 1 up to >99.9 area % with
an aqueous sodium chloride solution (Table 6, entries 1 and a 91% recovery on the 500 g scale. During a second run on the
2). The lower yield observed when quenching with water was 450 g scale, the use of recrystallized 7 (99.9 area % purity) was
686 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

Table 6. Different Quenching Methods Tested for the Isolation of 1


entry quenching method temperature (°C) filtration temperature (°C) resulting pH at rt yield 1 (%) purity 1 (area %) ML loss (%)
1 water (10 V) 25 20 0.5−1 85 99.6 3
2 aq. 10% NaCl (10 V) 25 20 0.5−1 94 98.4 5
3 water (10 V) 45 5 0.5−1 95 99.2 3
4 aq. 2 M NaOH (12 V) 45 20 7−8 92 99.8 0.3
5 aq. 32% K3PO4 (12 V) 45 20 6−7 96 98.2 0.5
6 aq. 10% NH3 (5 V) 45 5 6−7 96 98.3 1.7

Scheme 4. Optimized Medicinal Chemistry Route Demonstrated on the Scale for Compound 1

able to suppress the unknown impurity formation and deliver 1


in a 93% yield and 100% area % purity with no need for
■ EXPERIMENTAL SECTION
All reactions were performed under a protective nitrogen
acetone/heptane recrystallization. On the scale, compound 1 atmosphere using standard laboratory techniques. All reagents
was isolated in a 96.4% assay-corrected yield with no impurities and solvents were purchased from local suppliers and were
and within specifications for residual metals and solvents. It used as received. NMR spectra were recorded on a Bruker
should be noted that compound 1 has a high decomposition spectrometer operating at 300 MHz for 1H, 75 MHz for
13
energy (610 J/g) that is accompanied by strong gas evolution, C{1H}. Chemical shifts are expressed in parts per million (δ,
owing to the triazole ring. The maximum allowable temper- ppm) and referenced to residual solvent peaks. All coupling
ature is set at 130 °C based on isothermal DSC experiments. constants (J) are absolute values (Hz), and descriptions of
signals are s (singlet), d (doublet), t (triplet), and m
■ CONCLUSIONS
The synthesis of pyrimidinol 1, a key intermediate for the
(multiplet). HPLC analyses were performed on a Waters
ACQUITY H-Class UPLC system equipped with a quaternary
pump, refrigerated sample chamber, flow-through needle
synthesis of milvexian, was further optimized, bearing the autosampler/injector operating in the partial loop needle
scalability, robustness, and control strategy in mind. The Pd overfill mode, column heater, and photodiode array detector.
loading for the Suzuki cross-coupling was reduced from 10 to 1 The amounts of the isolated product are corrected by assay
mol %, and the resulting aniline 4 was isolated with a direct (weight percent purity). Assays were determined by HPLC
anti-solvent crystallization, increasing the isolated yield from against a purified external reference standard. Mettler Toledo
56 to >90%. The azidation−click sequence could be performed ReactIR 15 was used for inline FTIR monitoring. Wavelengths
on the scale after a thorough safety analysis, and 4-TMS of interest were selected manually and verified against HPLC
triazole 6 was isolated by crystallization of 2-MeTHF/heptane samples. The latter were also used to calibrate the IR trends for
in a >90% yield over the two steps. Performing the use in the VTNA analysis. The BAM fallhammer test was
chlorination in DMF in the presence of DCDMH alleviated performed as described in the “UN Recommendation on the
Transport of Dangerous Goods, Manual of Tests and Criteria:
the need for silica gel, and 4-chloro-1,2,3-triazole 7 was isolated
Test 3-a-ii” using a 5 kg weight at a height of 80 cm. DSC
by the addition of water to the reaction mixture. A
experiments were performed on a Mettler-Toledo DSC 822C
recrystallization of penultimate intermediate 7 in DCM/ using M20 gold-plated, pressure-resistant crucibles (TÜ V
heptane together with a charcoal treatment was implemented SÜ D). All analytical data for isolated compounds 4, 5, 6, 7,
to control color and impurity profiles. Corrosive HBr was and 1 were aligned with reported literature.3
replaced by HCl in the last demethylation step, and final target 4-Chloro-2-(6-methoxypyrimidin-4-yl)aniline (4). A 2
1 crystallized upon the addition of aqueous ammonia in a L reactor was charged with acetonitrile (500 mL, 5 L/kg),
>90% yield. Overall, compound 1 was obtained in a 70% compound 3 (100 g, 1.0 equiv), compound 4 (60 g, 1.05
overall yield starting from boronic ester 3 (Scheme 4). equiv), TMG (93.2 g, 2.05 equiv), and water (70.8 g, 10
687 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

equiv). The reactor headspace was purged using a nitrogen Copper iodide (8.0 g, 10 mol %) was charged to the reactor,
flow. Pd(dppf)Cl2·DCM (3.2 g, 1 mol %) was added to the which was then purged again with nitrogen until the oxygen
reaction mixture, and the reaction mixture was heated to 65 °C level was below 0.1%. The reaction mixture was stirred at 5−15
in 1 h and stirred at that temperature for 16 h. After reaction °C for 16 h. After reaction completion, the reaction mixture
completion, 1,3,5-triazine-2,4,6(1H,3H,5H)-trithione sodium was warmed up to 25 °C and 1,3,5-triazine-2,4,6(1H,3H,5H)-
salt (TriNaTMT) (10 g, 10% w/w) was dissolved in water (50 trithione sodium salt (TriNaTMT) (10 g, 10% w/w ) was
mL, 0.5 L/kg) and added to the reaction mixture. Warm water added. After 1 h of stirring at 25 °C, the mixture was filtered
(280 mL, 2.8 L/kg) was then slowly dosed over 4 h to the through Celite (30 g, 0.3 kg/kg) and the cake was rinsed with
reaction mixture while keeping the internal temperature at 60− 2-MeTHF (250 mL, 2.5 L/kg). The resulting solution
65 °C. After aging for 4 h, warm water (350 mL, 3.5 L/kg) was containing crude compound 6 was then washed with aqueous
then slowly dosed over 6 h to the reaction mixture while ammonia solution (100 g in 1 L of water) then with aqueous
keeping the internal temperature at 60−65 °C. The reaction sodium sulfate solution (100 g in 1 L of water). After phase
mixture was then cooled down to 10−15 °C over 4 h and separation, the upper layer was filtered through Celite (30 g,
further aged at 10−15 °C for 3 h. The slurry was then filtered, 0.3 kg/kg). After washing the Celite cake with 2-MeTHF (0.5
and the cake was washed with a cooled MeCN/water solution L, 5 L/kg), the mixture was concentrated at 45 °C under a
(2:1 v/v ratio, 5 L/kg). Wet compound 5 was obtained in a reduced pressure to 500−700 mL (5−7 L/kg). n-Heptane (1
90−95% assay yield and 98 area % purity and was used as such L, 10 L/kg) was added dropwise to the reactor, and the
in the next step. A 5 L reactor was charged with wet compound resulting mixture was concentrated at 45 °C under a reduced
4 (100 g, 1.0 equiv), 2-MeTHF (1.5 L, 15 L/kg), and an pressure to 500−700 mL (5−7 L/kg). The reaction mixture
aqueous N-acetyl L-cysteine solution (32 g in 1.5 L of water). was then warmed up to 55−60 °C and kept stirring at that
The resulting mixture was stirred for 2 h at 25 °C and filtered temperature for 4 h. After cooling the reactor to 5−15 °C over
through Celite (30 g, 0.3 kg/kg). After phase separation, the 3 h, the slurry was aged at 10 °C and filtered. Compound 6 was
organic layer was then successively washed with aqueous N- washed with n-heptane (1 L, 10 L/kg) and dried under a
acetyl L-cysteine solution (32 g in 1.5 L of water), aqueous reduced pressure at 40−45 °C for 12 h. Compound 6 was
sodium bicarbonate (70 g in 1 L water), and aqueous sodium isolated as a brown solid (137−147 g, >99.5 area % purity,
sulfate (100 g in 1 L of water). The upper layer was kept in the >99.0% w/w assay, and 90−96% corrected yield over the two
reactor and diluted with 2-MeTHF (1 L, 10 L/kg), and the steps).
resulting solution was concentrated under vacuum to 500−600 4-(5-Chloro-2-(4-chloro-1H-1,2,3-triazol-1-yl)phenyl)-
mL (5−6 L/kg). The solution was then diluted with 2-MeTHF 6-methoxypyrimidine (7). A reactor was charged with
(1 L, 10 L/kg), and the resulting solution was concentrated compound 6 (100 g, 1.0 equiv), DMF (500 mL, 5 L/kg),
under vacuum to 500−600 mL (5−6 L/kg). The solution cooling to −10 °C, and then purified water (5 g, 1.0 equiv).
containing compound 4 was then diluted with 2-MeTHF (1 L, 1,3-Dichloro-5,5-dimethylhydantoin (DCDMH, 13.7 g, 0.75
10 L/kg), and the water content was measured (KF below equiv) was charged in portions while keeping the internal
0.1%). temperature below 5 °C. The internal temperature was then
4-(5-Chloro-2-(4-(trimethylsilyl)-1H-1,2,3-triazol-1-yl)- adjusted to 0−10 °C, and the mixture was stirred at that
phenyl)-6-methoxypyrimidine (6). A reactor was charged temperature for 12 h. After reaction completion, water (70 mL,
with the 2-MeTHF solution containing compound 4 (100 g in 0.7 L/kg) was dosed to the reaction mixture over 1 h while
1.5 L of 2-MeTHF), 2-MeTHF (1.5 L, 15 L/kg), and MeCN keeping the internal temperature below 15 °C. Compound 7
(1 L, 10 L/kg), and the resulting mixture was cooled to 5−10 seeds (0.1 g, 0.001 kg/kg) were added to the reaction mixture,
°C. TMS azide (59.0 g, 1.2 equiv) was dosed slowly to the which was then aged for 2 h at 0−10 °C. Water (530 mL, 5.3
reactor. tBuONO (53.0 g, 1.2 equiv) was dosed slowly to the L/kg) was dosed over 3 h at 0−10 °C, and the slurry was aged
reactor at 5−10 °C, and the mixture was stirred at 10 °C for 4 for 4 h at 0−10 °C. The mixture was filtered, and the cake was
h. An aqueous NaOH solution (100 g in 1 L of water) was washed with cold water (0.5 L). Crude compound 7 was dried
slowly added to the mixture, which was then warmed up to 25 under reduced pressure at 40−50 °C for 15 h. A reactor was
°C, stirred for 30 min, and left standing for 30 min at 25 °C. charged with crude compound 7 (100 g, 1.0 equiv) and DCM
After phase separation, the upper layer was left in the reactor (0.5 L, 5 L/kg), and the resulting solution was stirred for 0.5−
and was washed two more times with an aqueous NaOH 2 h at 20−30 °C. The resulting mixture was filtered through
solution (100 g in 1 L of water). After the third wash, the charcoal and circulated for 6 h and then concentrated under a
upper layer was left in the reactor and the residual azide was reduced pressure to 300−360 mL (3.0−3.6 L/kg). The
measured (residual N3 < 3 ppm). An aqueous sodium sulfate mixture was then warmed up to 35−45 °C and refluxed for
solution (100 g in 1 L of water) was then added to the mixture, 30 min before being cooled down to 0−10 °C over 4 h then
which was then stirred for 30 min and left standing for 30 min further aged at 0−10 °C for 0.5−2 h. n-Heptane (1.6 L, 16 L/
at 25 °C. After phase separation, the upper layer was left in the kg) was then charged to the reactor at 0−10 °C over 3 h, and
reactor and was washed two more times with an aqueous the slurry was aged at 0−10 °C for 3 h. After filtration of the
sodium sulfate solution (100 g in 1 L of water). After phase slurry, purified compound 7 was washed with n-heptane (500
separation, the pH was measured (pH < 9) and the solution mL, 5 L/kg) and dried under reduced pressure at 40−50 °C
containing azide 5 was cooled to 10 °C. After bubbling for 6−12 h. Compound 7 was isolated as an off-white solid
nitrogen for 30 min, triethylamine (95.0 g, 2.2 equiv) was (78−82 g, >99.5 area % purity, >99.0% w/w assay, and 88−
charged slowly to the reaction mixture while keeping the 92% corrected yield over the two steps).
temperature at 5−15 °C. Trimethylsilylacetylene (50.0 g, 1.2 6-(5-Chloro-2-(4-chloro-1H-1,2,3-triazol-1-yl)phenyl)-
equiv) was charged slowly to the reaction mixture, while pyrimidin-4-ol (1). A reactor was charged with purified
keeping the temperature at 5−15 °C, and the reactor was compound 7 (100 g, 1.0 equiv) and aqueous HCl (35% w/w,
purged with nitrogen until the oxygen level was below 0.1%. 320 g, 10 equiv), and the reactor was warmed up to 40−50 °C
688 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

in 3 h, and then the reaction mixture was stirred at that Winfried Fiedler − Chemical Process Research and
temperature for 18 h. After reaction completion, an aqueous Development, Janssen Pharmaceuticals, Schaffhausen 8200,
ammonia solution (50 g in 500 mL) was added dropwise to Switzerland
the reactor over 4 h at 40−50 °C to reach pH 5−7. The Stefan Schirling − Chemical Process Research and
reaction mixture was then cooled down to 0−10 °C in 3 h and Development, Janssen Pharmaceuticals, Schaffhausen 8200,
aged at that temperature for 3 h. After filtration of the slurry, a Switzerland
crude compound 1 cake was washed with cold water (1 L at Sven Hock − Chemical Process Research and Development,
0−10 °C, 10 L/kg) and dried under a reduced pressure at 40− Janssen Pharmaceuticals, Schaffhausen 8200, Switzerland
50 °C for 24 h. Crude compound 1 was isolated as an off-white Tobias Strittmatter − Chemical Process Research and
solid (89−93 g, >99.5 area % purity, >99.5% w/w assay, and Development, Janssen Pharmaceuticals, Schaffhausen 8200,
93−97% corrected yield). A reactor was charged with crude Switzerland
compound 1 (100 g, 1.0 equiv) and acetone (1.4 L, 14 L/kg), Jan Dijkmans − Chemical Process Research and Development,
and the reactor was warmed up to 50−60 °C and stirred at that Janssen Pharmaceuticals, Beerse 2340, Belgium
temperature for 1−3 h. After seeding with compound 1 (0.5 g, Ivan Vervest − Chemical Process Research and Development,
0.005 kg/kg), n-heptane (1.7 L, 17 L/kg) was dosed at 50−60 Janssen Pharmaceuticals, Beerse 2340, Belgium
°C over 6 h, and the mixture was stirred at that temperature for Tim Van Hoegaerden − Chemical Process Research and
2 h before being cooled down to 5−15 °C over 4 h. After Development, Janssen Pharmaceuticals, Beerse 2340, Belgium
filtration, pure compound 1 was washed with cold n-heptane Brecht Egle − Chemical Process Research and Development,
(0.5 L, 5 L/kg) and dried under reduced pressure at 80−90 °C Janssen Pharmaceuticals, Beerse 2340, Belgium
for 16 h. Compound 1 was isolated as an off-white solid (90− Matthew P. Mower − Chemical Process Research and
95 g, >99.9 area % purity, >99.5% w/w assay, and 90−95% Development, Janssen Pharmaceuticals, Beerse 2340, Belgium
corrected yield). Zhi Liu − Janssen (China) R&D Center, Shanghai 200030,
China


*
ASSOCIATED CONTENT
sı Supporting Information
Zhiyong Cao − Janssen (China) R&D Center, Shanghai
200030, China
Xiaoning He − Janssen (China) R&D Center, Shanghai
The Supporting Information is available free of charge at 200030, China
https://pubs.acs.org/doi/10.1021/acs.oprd.2c00399. Lei Chen − Changzhou SynTheAll Pharmaceutical Co., Ltd.,
Changzhou, Jiangsu 213127, China
Kinetic study of the Suzuki cross-coupling using a Lei Qin − Changzhou SynTheAll Pharmaceutical Co., Ltd.,
VTNA approach, solvent screen for chlorination step Changzhou, Jiangsu 213127, China
using NCS, calorimetry data for azidation, click and Hongyu Tan − Changzhou SynTheAll Pharmaceutical Co.,
chlorination steps, and characterization and DSC data Ltd., Changzhou, Jiangsu 213127, China
for isolated compounds 4, 5, 6, 7, and 1 (PDF) Jun Yan − Changzhou SynTheAll Pharmaceutical Co., Ltd.,
Changzhou, Jiangsu 213127, China

■ AUTHOR INFORMATION
Corresponding Authors
Nicolas Lucien Cunière − Chemical Process Development,
Bristol-Myers Squibb, New Brunswick, New Jersey 08903,
United States
Simon Wagschal − Chemical Process Research and Carolyn S. Wei − Chemical Process Development, Bristol-
Development, Janssen Pharmaceuticals, Schaffhausen 8200, Myers Squibb, New Brunswick, New Jersey 08903, United
Switzerland; orcid.org/0000-0003-1979-5838; States; Present Address: Process Development, Amgen
Email: swagscha@its.jnj.com Inc., Thousand Oaks, California 91320, United States
Diego Broggini − Chemical Process Research and Venkata Vuyyuru − Chemical Development and API Supply,
Development, Janssen Pharmaceuticals, Schaffhausen 8200, Biocon Bristol-Myers Squibb Research and Development
Switzerland; Email: dbroggin@its.jnj.com Center, Bengaluru 560099, India
Rajaram Ayothiraman − Chemical Development and API
Authors Supply, Biocon Bristol-Myers Squibb Research and
Trung D.C. Cao − Chemical Process Research and Development Center, Bengaluru 560099, India
Development, Janssen Pharmaceuticals, Schaffhausen 8200, Sundaramurthy Rangaswamy − Chemical Development and
Switzerland API Supply, Biocon Bristol-Myers Squibb Research and
Pascal Schleiss − Chemical Process Research and Development Center, Bengaluru 560099, India
Development, Janssen Pharmaceuticals, Schaffhausen 8200, Mohamed Jaleel − Chemical Development and API Supply,
Switzerland Biocon Bristol-Myers Squibb Research and Development
Kristian Paun − Chemical Process Research and Development, Center, Bengaluru 560099, India
Janssen Pharmaceuticals, Schaffhausen 8200, Switzerland Rajappa Vaidyanathan − Chemical Development and API
Jessica Steiner − Chemical Process Research and Development, Supply, Biocon Bristol-Myers Squibb Research and
Janssen Pharmaceuticals, Schaffhausen 8200, Switzerland Development Center, Bengaluru 560099, India; Present
Anna-Lena Merk − Chemical Process Research and Address: SEAGEN INC, 22515 29th Dr. SE, Bothell,
Development, Janssen Pharmaceuticals, Schaffhausen 8200, Washington 98021, United States; orcid.org/0000-
Switzerland 0002-2236-5719
Joachim Harsdorf − Chemical Process Research and Martin D. Eastgate − Chemical Process Development, Bristol-
Development, Janssen Pharmaceuticals, Schaffhausen 8200, Myers Squibb, New Brunswick, New Jersey 08903, United
Switzerland States; orcid.org/0000-0002-6487-3121

689 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

Richard Klep − Chemical Process Research and Development, Memish, Z. A. Global and Regional Mortality from 235 Causes of
Janssen Pharmaceuticals, Beerse 2340, Belgium Death for 20 age groups in 1990 and 2010: A Systematic Analysis for
Cyril Benhaïm − Chemical Process Research and the Global Burden of Disease Study 2010. Lancet 2012, 380, 2095−
Development, Janssen Pharmaceuticals, Beerse 2340, Belgium 2128. (b) Raskob, G. E.; Angchaisuksiri, P.; Blanco, A. N.; Buller, H.;
Ilse Vogels − Chemical Process Research and Development, Gallus, A.; Hunt, B. J.; Hylek, E. M.; Kakkar, A.; Konstantinides, S. V.;
Janssen Pharmaceuticals, Beerse 2340, Belgium McCumber, M.; Ozaki, Y. Thrombosis: A major contributor to
disease burden. Arterioscler., Thromb., Vasc. Biol. 2014, 12, 1580−
Koen Peeters − Chemical Process Research and Development,
1590. (c) Virani, S. S.; Alonso, A.; Benjamin, E. J.; Bittencourt, M. S.;
Janssen Pharmaceuticals, Beerse 2340, Belgium
Callaway, C. W.; Carson, A. P.; Chamberlain, A. M.; Chang, A. R.;
Sébastien Lemaire − Chemical Process Research and Cheng, S.; Delling, F. N.; Djousse, L.; Elkind, M. S. V.; Ferguson, J.
Development, Janssen Pharmaceuticals, Beerse 2340, F.; Fornage, M.; Khan, S. S.; Kissela, B. M.; Knutson, K. L.; Kwan, T.
Belgium; orcid.org/0000-0002-9343-0493 W.; Lackland, D. T.; Lewis, T. T.; Lichtman, J. H.; Longenecker, C.
Complete contact information is available at: T.; Loop, M. S.; Lutsey, P. L.; Martin, S. S.; Matsushita, K.; Moran, A.
https://pubs.acs.org/10.1021/acs.oprd.2c00399 E.; Mussolino, M. E.; Perak, A. M.; Rosamond, W. D.; Roth, G. A.;
Sampson, U. K. A.; Satou, G. M.; Schroeder, E. B.; Shah, S. H.; Shay,
Author Contributions C. M.; Spartano, N. L.; Stokes, A.; Tirschwell, D. L.; VanWagner, L.
The manuscript was written through contributions of all B.; Tsao, C. W.; American Heart Association Council on
authors. All authors have given approval to the final version of Epidemiology and Prevention Statistics Committee and Stroke
the manuscript. Statistics Subcommittee. Heart disease and stroke statistics-2020
update: A report from the American Heart Association. Circulation
Notes 2020, 141, e139−e596.
The authors declare no competing financial interest. (2) (a) Weitz, J. I.; Chan, N. C. Advances in Antithrombotic

■ ACKNOWLEDGMENTS
This study was sponsored by Bristol-Myers Squibb and Janssen
Therapy. Arterioscler., Thromb., Vasc. Biol. 2019, 39, 7−12. (b) Quan,
M. L.; Pinto, D. J. P.; Smallheer, J. M.; Ewing, W. R.; Rossi, K. A.;
Luettgen, J. M.; Seiffert, D. A.; Wexler, R. R. Factor XIa Inhibitors as
Research & Development, LLC. New Anticoagulants. J. Med. Chem. 2018, 61, 7425−7447. (c) Al-
Horani, R. A.; Afosah, D. K. Recent Advances in the Discovery and
■ REFERENCES
(1) (a) Lozano, R.; Naghavi, M.; Foreman, K.; Lim, S.; Shibuya, K.;
Development of Factor XI/XIa Inhibitors. Med. Res. Rev. 2018, 38,
1974−2023. (d) Al-Horani, R. A. Factor XI(a) Inhibitors for
Thrombosis: An Updated Patent Review (2016-present). Expert
Aboyans, V. R.; Abraham, J.; Adair, T.; Aggarwal, R.; Ahn, S. Y.;
Alvarado, M.; Anderson, H. R.; Anderson, L. M.; Andrews, K. G.; Opin. Ther. Pat. 2020, 30, 39−55. (e) Hussain, Z.; Cooke, A. J.;
Atkinson, C.; Baddour, L. M.; Barker-Collo, S.; Bartels, D. H.; Bell, M. Neelamkavil, S.; Brown, L.; Carswell, E.; Geissler, W. M.; Guo, Z.;
L.; Benjamin, E. J.; Bennett, D.; Bhalla, K.; Bikbov, B.; Bin, A. A.; Hawes, B.; Kelly, T. M.; Kiyoi, Y.; Lai, K. D.; Lesburg, C.; Pow, E.;
Birbeck, G.; Blyth, F.; Bolliger, I.; Boufous, S.; Bucello, C.; Burch, M.; Zang, Y.; Wood, H. B.; Edmondson, S. D.; Liu, W. Design and
Burney, P.; Carapetis, J.; Chen, H.; Chou, D.; Chugh, S. S.; Coffeng, Synthesis of Novel Proline-based Factor XIa Selective Inhibitors as
L. E.; Colan, S. D.; Colquhoun, S.; Colson, K. E.; Condon, J.; Connor, Leads for Potential New Anticoagulants. Bioorg. Med. Chem. Lett.
M. D.; Cooper, L. T.; Corriere, M.; Cortinovis, M.; de Vaccaro, K. C.; 2020, 30, 127072. (f) Lorthiois, E.; Roache, J.; Barnes-Seeman, D.;
Couser, W.; Cowie, B. C.; Criqui, M. H.; Cross, M.; Dabhadkar, K. Altmann, E.; Hassiepen, U.; Turner, G.; Duvadie, R.; Hornak, V.;
C.; Dahodwala, N.; De Leo, D.; Degenhardt, L.; Delossantos, A.; Karki, R. G.; Schiering, N.; Weihofen, W. A.; Perruccio, F.; Calhoun,
Denenberg, J.; Des Jarlais, D. C.; Dharmaratne, S. D.; Dorsey, E. R.; A.; Fazal, T.; Dedic, D.; Durand, C.; Dussauge, S.; Fettis, K.; Tritsch,
Driscoll, T.; Duber, H.; Ebel, B.; Erwin, P. J.; Espindola, P.; Ezzati, F.; Dentel, C.; Druet, A.; Liu, D.; Kirman, L.; Lachal, J.; Namoto, K.;
M.; Feigin, V.; Flaxman, A. D.; Forouzanfar, M. H.; Fowkes, F. G. R.; Bevan, D.; Mo, R.; Monnet, G.; Muller, L.; Zessis, R.; Huang, X.;
Franklin, R.; Fransen, M.; Freeman, M. K.; Gabriel, S. E.; Gakidou, E.; Lindsley, L.; Currie, T.; Chiu, Y.-H.; Fridrich, C.; Delgado, P.; Wang,
Gaspari, F.; Gillum, R. F.; Gonzalez-Medina, D.; Halasa, Y. A.; S.; Hollis-Symynkywicz, M.; Berghausen, J.; Williams, E.; Liu, H.;
Haring, D.; Harrison, J. E.; Havmoeller, R.; Hay, R. J.; Hoen, B.; Liang, G.; Kim, H.; Hoffmann, P.; Hein, A.; Ramage, P.; D’Arcy, A.;
Hotez, P. J.; Hoy, D.; Jacobsen, K. H.; James, S. L.; Jasrasaria, R.; Harlfinger, S.; Renatus, M.; Ruedisser, S.; Feldman, D.; Elliott, J.;
Jayaraman, S.; Johns, N.; Karthikeyan, G.; Kassebaum, N.; Keren, A.; Sedrani, R.; Maibaum, J.; Adams, C. M. Structure-Based Design and
Khoo, J.-P.; Knowlton, L. M.; Kobusingye, O.; Koranteng, A.; Preclinical Characterization of Selective and Orally Bioavailable
Krishnamurth, R.; Lipnick, M.; Lipshultz, S. E.; Ohno, S. L.; Factor Xia Inhibitors: Demonstrating the Power of an Integrated S1
Mabweijano, J.; MacIntyre, M. F.; Mallinger, L.; March, L.; Marks, Protease Family Approach. J. Med. Chem. 2020, 63, 8088−8113.
G. B.; Marks, R.; Matsumori, A.; Matzopoulos, R.; Mayosi, B. M.; (g) Lei, Y.; Zhang, B.; Zhang, Y.; Dai, X.; Duan, Y.; Mao, Q.; Gao, J.;
McAnulty, J. H.; McDermott, M. M.; McGrath, J.; Mensah, G. A.; Yang, Y.; Bao, Z.; Fu, X.; Ping, K.; Yan, C.; Mou, Y.; Wang, S. Design,
Merriman, T. R.; Michaud, C.; Miller, M.; Miller, T. R.; Mock, C.;
Synthesis and Biological Evaluation of Novel FXIa Inhibitors with 2-
Mocumbi, A. O.; Mokdad, A. A.; Moran, A.; Mulholland, K.; Nair, M.
phenyl-1H-imidazole-5-carboxamide Moiety as P1 Fragment. Eur. J.
N.; Naldi, L.; Narayan, K. M. V.; Nasseri, K.; Norman, P.; O’Donnell,
M.; Omer, S. B.; Ortblad, K.; Osborne, R.; Ozgediz, D.; Pahari, B.; Med. Chem. 2021, 220, 113437.
Pandian, J. D.; Rivero, A. P.; Padilla, R. P.; Perez-Ruiz, F.; Perico, N.; (3) (a) Dilger, A. K.; Pabbisetty, K. B.; Corte, J. R.; De Lucca, I.;
Phillips, D.; Pierce, K.; Pope, C. A., 3rd; Porrini, E.; Pourmalek, F.; Fang, T.; Yang, W.; Pinto, D. J. P.; Wang, Y.; Zhu, Y.; Mathur, A.; Li,
Raju, M.; Ranganathan, D.; Rehm, J. T.; Rein, D. B.; Remuzzi, G.; J.; Hou, X.; Smith, D.; Sun, D.; Zhang, H.; Krishnananthan, S.; Wu, D.
Rivara, F. P.; Roberts, T.; De Leon, F. R.; Rosenfeld, L. C.; Rushton, R.; Myers, J. E., Jr.; Sheriff, S.; Rossi, K. A.; Chacko, S.; Zheng, J. J.;
L.; Sacco, R. L.; Salomon, J. A.; Sampson, U.; Sanman, E.; Schwebel, Galella, M. A.; Ziemba, T.; Dierks, E. A.; Bozarth, J. M.; Wu, Y.;
D. C.; Segui-Gomez, M.; Shepard, D. S.; Singh, D.; Singleton, J.; Crain, E.; Wong, P. C.; Luettgen, J. M.; Wexler, R. R.; Ewing, W. R.
Sliwa, K.; Smith, E.; Steer, A.; Taylor, J. A.; Thomas, B.; Tleyjeh, I. Discovery of Milvexian, a High-Affinity, Orally Bioavailable Inhibitor
M.; Towbin, J. A.; Truelsen, T.; Undurraga, E. A.; of Factor XIa in Clinical Studies for Antithrombotic Therapy. J. Med.
Venketasubramanian, N.; Vijayakumar, L.; Vos, T.; Wagner, G. R.; Chem. 2022, 65, 1770−1785. (b) Goodwin, N. C. Persistence Pays
Wang, M.; Wang, W.; Watt, K.; Weinstock, M. A.; Weintraub, R.; Off: Milvexian Emerges from the Industry’s Longstanding Search for
Wilkinson, J. D.; Woolf, A. D.; Wulf, S.; Yeh, P.-H.; Yip, P.; Zabetian, Orally Bioavailable Factor XIa Inhibitors. J. Med. Chem. 2022, 65,
A.; Zheng, Z.-J.; Lopez, A. D.; Murray, C. J. L.; AlMazroa, M. A.; 1767−1769.

690 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691
Organic Process Research & Development pubs.acs.org/OPRD Article

(4) Hutton, C. A.; Skaff, O. A Convenient Preparation of Dityrosine (20) F., Stoessel (2020), Thermal Safety of Chemical Processes: Risk
via Miyaura Borylation−Suzuki Coupling of Iodotyrosine Derivatives. Assessment and Process Design, 2nd edition, Wiley-VCH.
Tetrahedron Lett. 2003, 44, 4895−4898. (21) Due to the low solubility of compound 7, the choice of suitable
(5) For a related example on scale, see: Lehmann, H.; Ruppen, T.; solvents was limited. Compound 7 was highly soluble in DCM which
Knoepfel, T. Scale-Up of Diazonium Salts and Azides in a Three-Step allowed the filtration of the DCM stream of 7 through a charcoal
Continuous Flow Sequence. Org. Process Res. Dev. 2022, 26, 1308− cartridge. Efforts to replace DCM for commercial production
1317. activities are ongoing.
(6) (a) Panja, C.; Puttaramu, J. V.; Chandran, T. K.; Nimje, R. Y.; (22) Under those conditions, the initial slurry of compound 7 turned
Kumar, H.; Gupta, A.; Arunachalam, P. N.; Corte, J. R.; Mathur, A. into a slightly cloudy suspension after heating for a few hours, and a
Methyl-2,2-difluoro-2-(fluorosulfonyl) Acetate (MDFA)/copper (I) slurry formed again later as product 1 formed and precipitated for the
Iodide Mediated and Tetrabutylammonium Iodide Promoted reaction mixture.
Trifluoromethylation of 1-aryl-4-iodo-1,2,3-triazoles. J. Fluorine (23) The chloromethane off-gas was collected using a basic scrubber.
Chem. 2020, 236, 109516. (b) Barral, K.; Moorhouse, A. D.;
Moses, J. E. Efficient conversion of aromatic amines into azides: A
one-pot synthesis of triazole linkages. Org. Lett. 2007, 9, 1809−1811.
(7) Murray, J. I.; Zhang, L.; Simon, A.; Silva Elipe, M. V.; Wei, C. S.;
Caille, S.; Parsons, A. T. Kinetic and Mechanistic Investigations to
Enable a Key Suzuki Coupling for Sotorasib Manufacture - What a
Difference a Base Makes. Org. Process Res. Dev. 2023, 27, 198−205.
(8) Sunesson, Y.; Limeí, E.; Nilsson Lill, S. O.; Meadows, R. E.;
Norrby, P.-O. Role of the Base in Buchwald-Hartwig Amination. J.
Org. Chem. 2014, 79, 11961−11969.
(9) (a) Bureís, J. Variable Time Normalization Analysis: General
Graphical Elucidation of Reaction Orders from Concentration
Profiles. Angew. Chem. 2016, 55, 16084−16087. (b) Blackmond, D.
G. Reaction Progress Kinetic Analysis: A Powerful Methodology for
Mechanistic Studies of Complex Catalytic Reactions. Angew. Chem.,
Int. Ed. 2005, 44, 4302−4320. (c) Blackmond, D. G. Kinetic Profiling
of Catalytic Organic Reactions as a Mechanistic Tool. J. Am. Chem.
Soc. 2015, 137, 10852−10866.
(10) ”Same-excess” experiments probe for changes in active catalyst
concentration during a reaction. A standard reaction is compared to
one initiated with the same absolute amount of catalyst but half the
concentration of substrate. If the reaction profiles of these two
experiments give good visual overlay after a superficial x-axis shift
(timeshift), this indicates that no changes in catalyst concentration Recommended by ACS
occurred during the first 50% of turnovers. For further details see
reference 6.
(11) Clem, R. G.; Huffman, E. H. Azide Complexes of Palladium Process Development for the Synthesis of BI 1702135: A
(II). J. Inorg. Nucl. Chem. 1965, 27, 365−369. Concise Design Enabled by Selective Acylation of a 2-
(12) Rosso, V. W.; Lust, D. A.; Bernot, P. J.; Grosso, J. A.; Modi, S. Aminobenzimidazole Intermediate
P.; Rusowicz, A.; Sedergran, T. C.; Simpson, J. H.; Srivastava, S. K.; Jiang-Ping Wu, Géraldine Garavel, et al.
Humora, M. J.; Anderson, N. G. Removal of Palladium from Organic APRIL 05, 2023
Reaction Mixtures by Trimercaptotriazine. Org. Process Res. Dev. ORGANIC PROCESS RESEARCH & DEVELOPMENT READ
1997, 1, 311−314.
(13) The presence of residual water may lead to the hydrolysis of Leveraging Synergistic Solubility in the Development of a
TMS azide and the formation of explosive hydrazoic acid HN3. In the Direct Isolation Process for Nemtabrutinib
presence of residual palladium, hydrazoic acid may also lead to the
formation of highly energetic palladium azide Pd(N3)2. Yonggang Chen, Eric Sirota, et al.
MARCH 09, 2023
(14) González-Bobes, F.; Kopp, N.; Li, L.; Deerberg, J.; Sharma, P.;
ORGANIC PROCESS RESEARCH & DEVELOPMENT READ
Leung, S.; Davies, M.; Bush, J.; Hamm, J.; Hrytsak, M. Scale-up of
Azide Chemistry: A Case Study. Org. Process Res. Dev. 2012, 16, 2051.
(15) Fletcher, J. T.; Walz, S. E.; Keeney, M. E. Monosubstituted Development of a Scalable Asymmetric Process for the
1,2,3-Triazoles from Two-step One-pot Deprotection/Click Addi- Synthesis of GLYT1 Inhibitor BI 425809 (Iclepertin)
tions of Trimethylsilylacetylene. Tetrahedron Lett. 2008, 49, 7030− Rogelio Frutos, Chris H. Senanayake, et al.
7032. MARCH 03, 2023
(16) Zhao, H.; Fu, H.; Qiao, R. Copper-Catalyzed Direct Amination ORGANIC PROCESS RESEARCH & DEVELOPMENT READ
of Ortho-Functionalized Haloarenes with Sodium Azide as the Amino
Source. J. Org. Chem. 2010, 75, 3311−3316. Control of Hydrogen Fluoride Formation and Release during
(17) Papadakis, E.; Tula, A. K.; Gani, R. Solvent Selection a Large-Scale Nitroarene Reduction in the Manufacture of
Methodology for Pharmaceutical Processes: Solvent swap. Chem.
AZD9977
Eng. Res. Des. 2016, 115, 443−461.
(18) Rodriguez, R. A.; Pan, C.-M.; Yabe, Y.; Kawamata, Y.; Eastgate, Carl J. Mallia, Paul M. Gillespie, et al.
M. D.; Baran, P. S. Palau’chlor: A practical and reactive chlorinating DECEMBER 28, 2022

reagent. J. Am. Chem. Soc. 2014, 136, 6908−6911. ORGANIC PROCESS RESEARCH & DEVELOPMENT READ
(19) Shimizu, S.; Imamura, Y.; Ueki, T. Incompatibilities Between
Get More Suggestions >
N-Bromosuccinimide and Solvents. Org. Process Res. Dev. 2014, 18,
354−358.

691 https://doi.org/10.1021/acs.oprd.2c00399
Org. Process Res. Dev. 2023, 27, 680−691

You might also like