Synthesis of BACE1 Inhibitors E2609 / E2071 via Oxime-Olefin Cycloaddition following a Process Risk Mitigation Strategy

Article abstract of DOI:10.1021/acs.oprd.1c00223

Process development of E2609 from the preclinical stage to the clinical stage following a process risk mitigation strategy is described here. Key features include a turbo Grignard reaction monitored by in-situ IR, [3 + 2] cycloaddition in water, chemoselective amide coupling via in-situ protection, and a Reformatsky/decarboxylation approach to install a difluoromethyl group. Toward safe and scalable manufacture of E2071, an analog of E2609, a flow-reaction process for trifluoromethylation of aldehydes is presented here.

Full text of DOI:10.1021/acs.oprd.1c00223

pubs.acs.org/OPRD
Article
Synthesis of BACE1 Inhibitors E2609/E2071 via Oxime−Olefin
Cycloaddition Following a Process Risk Mitigation Strategy
Dae-Shik Kim,* Kazuhiro Yoshizawa, Branko Mitasev, Matthew Schnaderbeck, Huiming Zhang,
Masayuki Omori, Akio Kayano, Mitsuo Nagai, Kazunori Wakasugi, Yuzo Watanabe, Farid Benayoud,
Yuichi Suzuki, Takafumi Motoki, Toshihiko Kaneko, Mamoru Takaishi, Tasuku Ishida,
Kunitoshi Takeda, Yoichi Kita, Noboru Yamamoto, Afzal Khan, Paschalis Dimopoulos,
Christopher N. Farthing, Adrian Hall, Arani Chanda, Lily Lu, Markus Bracke, and Francis G. Fang
Cite This: https://doi.org/10.1021/acs.oprd.1c00223
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Process development of E2609 from the preclinical stage to the clinical stage following a process risk mitigation
strategy is described here. Key features include a turbo Grignard reaction monitored by in-situ IR, [3 + 2] cycloaddition in water,
chemoselective amide coupling via in-situ protection, and a Reformatsky/decarboxylation approach to install a difluoromethyl group.
Toward safe and scalable manufacture of E2071, an analog of E2609, a flow-reaction process for trifluoromethylation of aldehydes is
presented here.
KEYWORDS: BACE1 inhibitor, turbo Grignard reaction, [3 + 2] cycloaddition, chemoselective amide coupling, Reformatsky reaction,
flow chemistry
resolution and aryl substitution protocols could be imple-
mented. E2609 and E2071 differ in the distribution and total
number of fluoro substituents on the two methyl groups
attached to the THF and pyrazine rings. In particular,
incorporation of a difluoromethylpyrazine on E2609 and a
trifluoromethyl group on the THF ring of E2071 presented
practical and safety-related concerns that needed to be
addressed prior to scale-up and manufacturing. Herein, are
described (1) the development and scale-up of an oxime−
olefin cycloaddition approach to directly and diastereoselec-
tively assemble 3 contiguous stereogenic centers on a THF
template, (2) development of a practical diastereomeric salt
resolution of the amino−alcohols derived from oxime−olefin
cycloaddition, (3) a practical decarboxylative Reformatsky
approach to efficiently incorporate a difluoromethyl group
onto a pyrazine ring, and (4) a safe and practical flow
chemistry process for incorporating a trifluoromethyl group
into the oxime−olefin cycloaddition substrate required for
manufacture of E2071.
INTRODUCTION
■
The amyloid hypothesis of Alzheimer’s disease (AD) suggests
that interfering with the production and accumulation of
amyloid beta (Aß) peptides, especially Aß42, could provide a
disease-modifying therapy.1 β-Amyloid cleaving enzyme 1
(BACE1) was found to be a key enzyme responsible for the
production of Aß peptides and therefore has been a prime
therapeutic target.2 E2609 and E2071, novel potent orally
administered BACE1 inhibitors, were developed as Aß peptide-
lowering small molecule agents that could serve as disease-
modifying therapeutics for AD patients (Figure 1).3 E2609 and
RESULTS AND DISCUSSION
■
The initial synthetic route used during the discovery phase was
adapted and optimized to generate early batches of E2609 for
preclinical studies. The resulting first-generation route outlined
Figure 1. Structures of BACE1 inhibitors E2609 and E2071.
E2071 share a common cis-fused 5,6-tetrahydrofuran (THF)−
isothiourea scaffold substituted with a 2-fluoro-5-aminophenyl
aromatic acylated on the amino moiety with a 2-fluoromethy-
lated-pyrazine-5-carboxylic acid. The oxime−olefin cyclo-
addition approach initially utilized to discover the cis-fused
5,6-THF-isothiourea scaffold was viewed as a solid strategy
that could be developed further provided that efficient chiral
Special Issue: Excellence in Industrial Organic Syn-
thesis 2021
Received: June 12, 2021
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX
© XXXX American Chemical Society
A

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Scheme 1. First-Generation Manufacturing Route of E2609
Scheme 2. Generation of Dimer Impurity 22 from 15 and Its Plausible Mechanism
in Scheme 1 consists of 19 synthetic stages and proceeds in
4.8% overall yield. Process highlights of the route include
multiple crystalline intermediates in the late stages, a
diastereoselective [3 + 2] cycloaddition (dr 13:1)4 to establish
the relative stereochemistry of all three stereogenic centers
(Stage 5), and a classical diastereomeric salt resolution utilizing
dibenzoyl-D-tartaric acid (D-DBTA) to isolate the desired
enantiomer whose configuration was confirmed by X-ray
B
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Scheme 3. Crossed Claisen Condensation/Decarboxylation Approach to Ketone 6
Scheme 4. Generation of Ketone 6 via Turbo Grignard Reaction
(Stage 7). Pyrazinoic acid 16 was prepared from acid 18 in 4
steps employing deoxofluorination as the key step. While the
majority of the route developed for E2609 was applicable to
E2071 and is therefore not discussed separately here, the
presence of the trifluoromethyl−THF ring in E2071
necessitated developing a process for manufacture of the
corresponding allylic alcohol cycloaddition substrate [1,1,1-
trifluoro-2-hydroxy-3-butene (2)].
fluorobenzoate (23) with lithium hexamethyldisilazide
(LiHMDS) provided β-ketoester 24 in 84% yield. Decarbox-
ylation in formic acid then generated ketone 6 in 71% yield.
Although the reaction can proceed at noncryogenic temper-
atures (>−20 °C), considerable amounts of impurities were
generated when the reaction temperature was raised above
−40 °C: alcohol 25 via 2,3-Wittig rearrangement8 and methyl
ester 26 via transesterification. Therefore, retaining cryogenic
conditions was required to minimize these side reactions.
A shorter, noncryogenic route to ketone 6 was developed
upon evaluation of commercial chloroacetamide compounds as
an alternative starting material. Coupling reaction of N-
(chloroacetyl)morpholine 27 with alcohol 1 in the presence of
aqueous sodium hydroxide and phase transfer catalyst
tetrabutylammonium hydrogensulfate (n-Bu4NHSO4)9 pro-
ceeded well to give butenyl ether 28 in good yield (66%,
Scheme 4). Although the resulting amide (28) would undergo
the cryogenic aryl lithium addition, a search for noncryogenic
conditions led to the evaluation of aryl magnesium complexes.5
Treatment of 1-bromo-2-fluorobenzene with isopropylmagne-
sium chloride lithium chloride complex (turbo Grignard
reagent)10 provided aryl magnesium complex 29, which then
reacted with amide 28 to give tetrahedral adduct 30 at
noncryogenic temperatures (0 °C). Upon treatment with
aqueous hydrochloric acid, ketone 6 was obtained in >90%
yield.
Formation of aryl Grignard reagent 29 and tetrahedral
adduct 30 could not be directly monitored using conventional
in-process control (IPC) methods (e.g., HPLC, GC). There-
fore, in-situ IR spectroscopy was employed to monitor these
species directly.11 First, based on IR spectra collected at
various reaction time points, the most intense and well-
resolved IR absorption bands for each species were selected for
reaction monitoring: the peak at 1161 cm−1 for aryl Grignard
reagent 29, the peak at 1200 cm−1 for adduct 30, and the peak
at 1635 cm−1 for amide 28.12 Next, the changes in the
absorption intensities at these three wavelengths were plotted
as a function of time, highlighting how the concentration of
Through our initial research and early manufacturing
experiences described above, five main risks associated with
this route were identified as follows.
• In Stage 3, the lithiation of 1-bromo-2-fluorobenzene
required cryogenic conditions (below −60 °C) to
prevent benzyne formation,5 which was throughput
limiting and increased manufacturing cost.
• Two late-stage intermediates, Boc-protected isothioureas
14 and 15, posed handling challenges due to instability.
For example, a significant dimeric urea impurity 22 was
generated from 15 in either a solid or solution even at
ambient temperature (Scheme 2).
• The Stage 3′ deoxofluorination has a limitation for scale-
up due to the high cost of Deoxo-Fluor and the
requirement of special facilities for the handling of
hazardous reagents.6
• The route contains 15 linear synthetic stages with 4.8%
overall yield, leading to a long production time.
• Trifluoromethylation of aldehydes with Ruppert−
Prakash reagent [trifluoromethyltrimethylsilane
(CF3TMS)], a process for the generation of 1,1,1-
trifluoro-2-hydroxy-3-butene (2), is an autocatalytic
reaction7 and therefore raised a safety concern of
controlling the reaction heat on scale.
Development of an Alternative Approach to the
Cryogenic Aryl Lithiation. An approach which would
preclude the need for an aryl lithiation was envisioned by
utilizing a crossed Claisen condensation as shown in Scheme 3.
Treatment of a mixture of tert-butyl ester 4 and methyl 2-
C
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
each component changed with reaction progress (Figure 2).
Such data showed that aryl magnesium bromide formation
Figure 3. HPLC overlay chromatograms of the oxime formation/[3 +
2] cycloaddition reaction.
Figure 2. In-line monitoring of the turbo Grignard reaction using in-
situ IR spectroscopy: Normalized time-dependent relative concen-
tration trends for amide 28, aryl Grignard reagent 29, and tetrahedral
adduct 30 based on univariate (one wavelength) analysis of IR.
oximes underwent the [3 + 2] cycloaddition to give 8. After 42
h, most of the oximes were consumed (93% conversion), and a
14:1 mixture of the desired product 8 and the methyl epimer
(epi-8) was obtained after workup in 81% yield. In scale-up,
without isolation of the resulting product, treatment of the
reaction mixture with zinc and acetic acid led to the N−O
bond cleavage to give amino alcohol 9 (Scheme 5). This one-
pot, three-step process reduced two synthetic stages and
increased the overall yield from ca. 50% to ca. 80%.16
seems to be complete in 1.5 h after starting the addition of 1-
bromo-2-fluorobenzene.13 In addition, toward the end of the
reaction, the conversion rate from 28 to 6 was measured with
an HPLC IPC method. At time points marked as IPC1 and
IPC2, the conversion rates were found to be 97.1% and 98.9%,
respectively. These data correlated well with the calculated
transformation obtained from IR data.
[3 + 2]-Cycloaddition in Water. One of the reasons for
the moderate yield of the [3 + 2] cycloaddition was attributed
to be the modest Z/E ratio (2.7/1) of oxime 7. Only the Z
isomer undergoes the desired [3 + 2] cycloaddition via NH
nitrone intermediate 31 formed by tautomerization of the Z
isomer, while the E isomer would give byproducts such as 32
via an azaprotiocyclo transfer reaction (APT) (Scheme 5).14
The first attempt for the optimization was to increase the Z/E
ratio by changing the reaction solvents and temperature, but
the improvement was only marginal. Next, based on literature
reports that the E/Z isomerization of oximes is promoted in
water, we decided to explore water as a reaction solvent for the
[3 + 2] addition.15 Given that the oxime formation proceeded
well in water, the oxime formation and cycloaddition steps
were combined into a single step; the desired product (8) was
generated via intermediate 7 upon treatment of ketone 6 with
hydroxylamine−hydrochloride in water at 100 °C.
Development of Chemoselective Coupling of Dia-
mine 34 via In-Situ Protection. To address the stability
issue of intermediates 14 and 15, alternative protecting groups
were searched. Serendipitously, a hint for the best protecting
group was given during the process impurity fate studies.
When compound 15 spiked with 10.75% of compound 13 was
subjected to the following stage, it was noticed that the
expected byproduct 33 was not detected at all while 13 still
remained at a level of 2.66% (Scheme 6) after isolation. These
results indicated that protection of the isothiourea moiety of
compound 15 might not be needed during the amide coupling.
To our delight, diamine 34, which was prepared from 13 via
nitro reduction with iron powder, underwent a chemoselective
coupling with acid chloride 35 to give E2609 in a quantitative
yield along with <0.2% byproduct 36 in the crude product
(Scheme 7).17 We reasoned that 1 equiv of hydrogen chloride
(HCl) in-situ generated from the acid chloride formation
would selectively protonate the more basic isothiourea amine
moiety of 34 to give an in-situ-protected form 34-HCl. In
contrast, the undesired amide (37) was selectively generated
under the base-mediated 2-(1H-benzotriazole-1-yl)-1,1,3,3-
The reaction progress was monitored by HPLC (Figure 3).
The Z/E ratio of oxime 7 was 1.9/1 at 0.5 h after 40%
conversion of ketone 6 and became 1.2/1 at 4 h, at which
point all of ketone 6 was consumed and some of the resulting
Scheme 5. One-Pot, Three-Step Process to Amino Alcohol 9 and Plausible Mechanism
D
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Scheme 6. Fate Study of Process Impurity 13
Scheme 7. Chemoselective Amide Coupling of Diamine 34 with Acid 16
Scheme 8. Second-Generation Manufacturing Route for Difluoromethyl Pyrazinoic Acid 16
tetramethyluronium hexafluorophosphate (HBTU) coupling
conditions, underscoring the importance of utilizing the base-
free coupling conditions. This chemoselective amide coupling
eliminated two unstable intermediates (14 and 15) and
shortened the overall synthesis by two stages: installation and
removal of the isothiourea protecting group.
Development of the Reformatsky/Decarboxylation
Approach for Difluoromethyl Pyrazinoic Acid 16. To
avoid the hazardous deoxofluorination step (Scheme 1),
introduction of the difluoromethyl group by Reformatsky
reaction of halo-pyrazinoic ester with 2-bromo-2,2-difluor-
oethyl acetate followed by decarboxylation reaction was
investigated (Scheme 8). Due to the low reactivity of chloride
38 for the Reformatsky coupling reaction, chloride 38 was
converted into more reactive bromide 39 by treating with
trimethylsilyl bromide (TMSBr) while continuously distilling
off byproduct trimethylsilyl chloride (TMSCl).18 The coupling
reaction of 39 with the Reformatsky reagent (BrZnCF2CO2Et)
prepared from BrCF2CO2Et with activated Zn19 proceeded
well in the presence of CuBr to give 40.20 Hydrolysis followed
by salt formation with benzylamine afforded bis(benzylamine)
salt 41 in 58% overall yield from 39. The decarboxylation
E
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Scheme 9. Synthetic Route to 45, and Schematic Representation of the Flow Reactor
Scheme 10. Second-Generation Manufacturing Route of E2609
reaction21 of 41 was performed under acidic and thermal
conditions, leading to difluoromethylpyrazinoic acid 16 in 68%
yield after crystallization. According to this method, any special
facilities and use of hazardous fluorinating reagents were not
required to make further scale-up manufacturing feasible.
Development of a Flow Chemistry Process for 1,1,1-
Trifluorobut-3-en-2-ol (2). In the initial experiment, treat-
ment of a solution of acrolein and Ruppert−Prakash reagent
(CF3TMS)7 in THF with a catalytic amount of tetrabuty-
lammonium fluoride (TBAF) rapidly generated TMS ether 44
(Scheme 9). Moreover, by performing the reaction in toluene,
the crude product could be used directly in a phase-transfer
alkylation of 3 to afford desired intermediate 45. However, the
temperature of the reaction mixture changed from ca. 0 to 80
°C almost instantaneously (within seconds) even on a
relatively small scale (e.g., 1 g) upon addition of a sufficient
amount of TBAF to initiate this autocatalytic process: the
fluoride ion from TBAF initiated the reaction which was
subsequently self-catalyzed by the resulting alkoxide ion of the
product.7 The autocatalytic trifluoromethylation reaction
precluded conduction of a batch process on scale and thus
provided a clear and compelling reason to utilize flow
F
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
chemistry to directly address the safety concerns. Moreover,
the kinetics and homogeneity of the process seemed ideal for
the adaptation to a flow system.22 A two-streamflow system
with a solution of CF3TMS and acrolein in toluene as stream 1
and a solution of TBAF (catalyst) in toluene/THF as stream 2
(Scheme 9) was designed. The two streams were mixed by
simultaneous flow through a static mixer, and the released heat
was cooled by a subsequent temperature-controlled heat-
exchange coil. In order to assess the effects of flow rate on the
reaction and enable rapid optimization via close temperature
control of the process, three temperature probes (T1, streams
meeting point before the static mixer; T2, immediately after
the static mixer; T3, after the heat exchange cooling loop) were
used. A React IR probe was also inserted along with T3, and
specific high-intensity IR peaks for key starting materials (1700
cm−1 for acrolein, 1050 cm−1 for CF3TMS) and product (44,
1100−1200 cm−1) were monitored for real-time assessment of
the reaction.23 The resulting reactor setup provided the
possibility of high production rates of up to 6 kg of 44 per hour
and was utilized to produce 45 on a 24 kg scale in a single run.
cooled to 0−10 °C. While maintaining the internal temper-
ature below 20 °C, 50 wt % aqueous sodium hydroxide (146
kg, 1834 mol, 3.0 equiv) was added. After the addition, the
mixture was stirred between 10 and 20 °C for 1 h and the
reaction was monitored for complete consumption of 27
(target >97.0% conversion). After being cooled down to below
5−10 °C, the reaction mixture was treated with deionized
water (153 kg) while maintaining an internal temperature
below 20 °C. After being stirred for 15 min at 15−23 °C, the
layers were allowed to partition. The lower aqueous layer was
separated out and extracted with methyl-tert-butyl ether
(MTBE) (220 kg). The combined organic layers were washed
with 1.0 M aqueous hydrochloric acid (50 L, 50 mol, 0.08
equiv) and concentrated under reduced pressure (jacket T <
45 °C) to give 28 (80.9 kg, 96.3 GC area % purity, 406 mol,
66.4% yield) as a pale yellow oil. 1H NMR (400 MHz, CDCl3)
δ 5.71 (ddd, J = 7.6, 10.1, 17.4 Hz, 1H), 5.33−5.10 (m, 2H),
4.19−4.12 (m, 1H), 4.07−3.99 (m, 1H), 3.97−3.83 (m, 1H),
3.68 (br s, 4H), 3.64−3.48 (m, 4H), 1.29 (d, J = 6.2 Hz, 3H);
13C NMR (125 MHz, CDCl3) δ 168.19, 138.94, 117.24, 77.65,
67.62, 66.87, 66.78, 45.78, 42.14, 21.13.
2-(But-3-en-2-yloxy)-1-(2-fluorophenyl)ethanone (6).
A reactor was charged with a solution of isopropylmagnesium
chloride−lithium chloride in THF (14%, 326 kg, 450 mol, 1.1
equiv) and cooled to 0−5 °C. 1-Bromo-2-fluorobenzene (81
kg, 449 mol, 1.1 equiv) was added while maintaining the
internal temperature between 0 and 10 °C. Upon complete
addition, the mixture was stirred between 0 and 10 °C for 1 h.
A solution of amide 28 (80.9 kg, 406 mol) in THF (46 kg) was
added while maintaining the internal temperature between 0
and 10 °C. A 1 kg amount of THF was used for rinsing. After
stirring the reaction for 3 h, the reaction was monitored for
complete consumption of amide 28 (>97.0% conversion). The
reaction mixture was transferred into a reactor containing a
mixture of MTBE (195 kg) and 2.0 M HCl (470 L, 2.3 equiv)
while maintaining the internal temperature below 20 °C. A 5
kg amount of THF was used for rinsing. The mixture was
stirred for 15 min, and then the phases were separated. The
aqueous layer was extracted with 125 kg of MTBE. The
combined organic layers were washed with 5.0 wt % aqueous
sodium bicarbonate (169 kg), deionized water (163 kg), and
17 wt % aqueous sodium chloride solution (196 kg). The
organic phase was concentrated under reduced pressure (T <
40 °C) to give ketone 6 (80.0 kg, 86.6 GC area % purity, 384
CONCLUSION
■
In summary, the second-generation manufacturing route of
E2609 in Scheme 10 was developed to address process issues
identified in the first manufacturing route for the preclinical
stage. A turbo Grignard reaction with in-situ monitoring by IR
spectroscopy was employed in Stage 2 to realize noncryogenic
arylation. In Stage 3, oxime formation followed by [3 + 2]
oxime−olefin cycloaddition and reduction was performed in
water to improve the yield and shorten the reaction steps. In
Stage 9, the chemoselective amide coupling of diamine 34 was
realized via in-situ protection, eliminating two unstable
intermediates and shortening two reaction stages. For
pyrazinoic acid 16, a much safer and scalable Reformatsky/
decarboxylation protocol to install the difluoromethyl group
was established and implemented on scale. These process
optimization steps resulted in shortening the linear synthetic
stage count from 15 to 9 and improving the overall yield to
10% from 4.8% in the first-generation route. Using this route,
19 kg of E2609 with clinical trial quality was manufactured
successfully. For E2071, the flow process for trifluoromethyl
ether 44 was developed to control the exothermic autocatalytic
process and was successfully implemented on industrial scale.
1
mol, 94.6% yield) as a pale yellow oil. H NMR (500 MHz,
EXPERIMENTAL SECTION
CDCl3) δ 7.93 (dt, J = 7.6, 1.8 Hz, 1H), 7.57−7.49 (m, 1H),
7.24 (dt, J = 14.6, 3.8 Hz, 1H), 7.12 (ddd, J = 11.0, 8.3, 0.8 Hz,
1H), 5.77 (ddd, J = 17.3, 10.2, 7.6 Hz, 1H), 5.21 (d, J = 17.6
Hz, 1H), 5.17 (dd, J = 10.3, 0.8 Hz, 1H), 4.69 (dd, J = 18.1,
3.2 Hz, 1H), 4.60 (dd, J = 18.1, 3.4 Hz, 1H), 4.00 (dq, J =
12.9, 6.4 Hz, 1H), 1.36 (d, J = 6.4 Hz, 3H); 13C NMR (125
■
NMR spectra were recorded on a Varian Inova 500 or 400
MHz or Bruker 600 or 400 MHz spectrometer. Chemical shifts
are reported in parts per million (ppm). For H NMR spectra
1
[CDCl3, CD3OD, or (CD3)2SO], the residual solvent peak was
used as the internal reference (7.26 ppm in CDCl3; 3.31 ppm
in CD3OD; 2.50 ppm in (CD3)2SO), while the central solvent
peak was the reference (77.0 ppm in CDCl3; 49.0 ppm in
CD3OD; 39.5 ppm in (CD3)2SO) for 13C NMR spectra. Water
contents were measured by Karl Fischer titration (Metrohm,
831 KF coulometer). All of the positive/negative-ion mode
electrospray ionization mass spectra were recorded on an LTQ
Orbitrap XL mass spectrometer (Thermo Fisher Scientific).
2-(But-3-en-2-yloxy)-1-morpholinoethanone (28). A
reactor was charged with toluene (260 kg), 4-(chloroacetyl)-
morpholine (27) (100 kg, 611 mol), tetrabutylammonium
hydrogen sulfate (9.4 kg, 28 mol, 0.05 equiv), and 3-buten-2-ol
(1) (53 kg, 735 mol, 1.20 equiv). The resulting mixture was
MHz, CDCl3) δ 195.24 (d, JCF = 5.3 Hz), 162.03 (d, JCF
253.9 Hz), 139.42, 134.86 (d, JCF = 8.9 Hz), 130.64 (d, JCF
=
=
3.4 Hz), 124.67 (d, JCF = 3.2 Hz), 123.57 (d, JCF = 15.3 Hz),
116.89, 116.45 (d, JCF = 23.7 Hz), 77.93, 74.11 (d, JCF = 11.2
Hz), 21.19.
4-Amino-4-(2-fluorophenyl)-2-methyltetrahydrofur-
an-3-yl)methanol (9). A reactor was charged with ketone 6
(80.0 kg, 384 mol), and deionized water (400 kg), hydroxyl-
amine hydrochloride (28.0 kg, 403 mol, 1.05 equiv), and
sodium acetate (34.7 kg, 423 mol, 1.10 equiv) were added. The
resulting mixture was diluted with 320 kg of deionized water,
heated to 101−102 °C for at least 40 h, and monitored for
G
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
complete conversion to tetrahydrooxazole 8 (>90.0% con-
version). The mixture was cooled to below 25 °C, and THF
(140 kg) and acetic acid (69.3 kg, 1155 mol, 3.0 equiv) were
added. Zinc dust (75.4 kg, 1155 mol, 3.0 equiv) was added
while maintaining the internal temperature below 45 °C. The
resulting mixture was stirred between 35 and 45 °C for 1 h,
and the reaction was monitored for complete conversion to
amino alcohol 9 (>98.0% conversion). The mixture was cooled
to 15−20 °C, and dichloromethane (205 kg) was added.
Aqueous ammonium hydroxide (25%, 107 kg) was added until
the pH reached 10−12. The resulting mixture was stirred for
15 min, and 40 kg of Celite was added. After 30 min of stirring,
the suspension was filtered in two equal portions by means of a
centrifuge. Each portion was washed with 20 kg of dichloro-
methane. The combined filtrates were transferred to a reactor,
and the layers were separated. The upper aqueous layer was
extracted twice with dichloromethane (230 kg each). The
combined organic layers were washed with 65 kg of deionized
water, treated with 7 kg of sodium sulfate, and filtered through
a bag filter. The resulting organic layer was concentrated under
reduced pressure (jacket T < 35 °C); meanwhile, solvent
exchange was performed with 235 kg of absolute ethanol. After
solvent exchange, 236.5 kg of concentrate containing amino
alcohol 9 (75.2 kg, 80.3 area % purity, 334 mol, 86.9% yield)
was obtained. 1H NMR (500 MHz, DMSO) δ 7.63 (td, J = 8.6,
1.7 Hz, 1H), 7.32−7.24 (m, 1H), 7.20−7.08 (m, 2H), 4.43 (s,
1H), 4.07 (dd, J = 8.7, 1.3 Hz, 1H), 4.02 (dq, J = 8.5, 6.1 Hz,
1H), 3.68−3.57 (m, 2H), 3.55−3.46 (m, 1H), 2.23 (dd, J =
14.4, 6.7 Hz, 1H), 1.24 (d, J = 6.1 Hz, 3H); 13C NMR (125
Chiral HPLC Parameters. Column: Chiralcel OD-H, 4.6 ×
250 mm, 5 μm, mobile phase (isocratic):n-heptane/2-
propanol/triethylamine = 98/2/0.05, flow rate 1.0 mL/min,
injection volume 20 μL, temperature 40 °C, detection 262 nm,
total running time 80 min, dilution solvent 2-propanol, relative
retention time (RRT) of (ent)-9 = 0.90, RRT of 9 = 1.00
N-((3S,4R,5R)-3-(2-Fluorophenyl)-4-(hydroxymethyl)-
5-methyltetrahydrofuran-3-ylcarbamothioyl)-
benzamide (10). A reactor was charged with 9-DBT (30.0
kg, 48.5 mol) and ethyl acetate (81 kg). The mixture was
cooled to 0−10 °C. A 1.00 M aqueous NaOH (120 L, 120
mol, 2.47 equiv) solution was added. Benzoyl isothiocyanate
(9.42 kg, 56.5 mol, 1.16 equiv) was added with vigorous
stirring while maintaining the internal temperature between 0
and 10 °C. After complete addition, stirring was continued at
0−5 °C over 2 h and the reaction was monitored for complete
consumption of 9-DBT (target >98.0% conversion). The
reaction mixture was filtered, and the filter cake was rinsed
with (1) deionized water (60 kg) and (2) a mixture of heptane
(27 kg) and ethyl acetate (36 kg). The solid was dried under
reduced pressure (T < 40 °C) to give 10 (17.0 kg, 93.5 wt %
purity, 40.9 mol, 84.3% yield). 1H NMR (500 MHz, CDCl3) δ
11.78 (s, 1H), 8.88 (s, 1H), 7.88−7.82 (m, 2H), 7.67 (td, J =
8.1, 1.6 Hz, 1H), 7.65−7.59 (m, 1H), 7.50 (dd, J = 10.8, 4.8
Hz, 2H), 7.30−7.23 (m, 1H), 7.15 (td, J = 7.8, 1.2 Hz, 1H),
7.01 (ddd, J = 12.3, 8.2, 1.1 Hz, 1H), 4.73 (d, J = 10.1 Hz,
1H), 4.42 (dd, J = 10.1, 1.7 Hz, 1H), 4.10−4.03 (m, 1H),
4.03−3.91 (m, 2H), 2.80 (t, J = 5.3 Hz, 1H), 2.63−2.52 (m,
1H), 1.35 (d, J = 6.1 Hz, 3H); 13C NMR (125 MHz, CDCl3) δ
179.42, 166.60, 160.35 (d, JCF = 247.4 Hz), 133.60, 131.81,
129.27 (d, JCF = 8.9 Hz), 129.20 (d, JCF = 3.7 Hz), 129.15,
127.55, 127.39 (d, JCF = 10.0 Hz), 123.96 (d, JCF = 3.3 Hz),
116.18 (d, JCF = 23.0 Hz), 76.90 (d, JCF = 2.8 Hz), 76.27,
69.26, 59.49, 58.03, 20.04.
MHz, DMSO) δ 160.69 (d, JCF = 244.7 Hz), 133.57 (d, JCF
=
11.3 Hz), 128.95 (d, JCF = 8.8 Hz), 128.81 (d, JCF = 4.8 Hz),
124.33 (d, JCF = 3.2 Hz), 116.32 (d, JCF = 23.5 Hz), 79.58 (d,
JCF = 4.9 Hz), 78.52, 63.99 (d, JCF = 3.5 Hz), 59.39, 56.91 (d,
JCF = 2.2 Hz), 21.37.
((2R,3R,4S)-4-Amino-4-(2-fluorophenyl)-2-methylte-
trahydrofuran-3-yl)methanol (2S,3S)-2,3-Bis-
(benzoyloxy)succinate (9-DBT). A reactor was charged
with 236 kg of an ethanol solution containing amino alcohol 9
(75.2 kg, 334 mol) and ethanol (253 kg). The mixture was
heated to an internal temperature of 65−70 °C. To the mixture
was added a solution of dibenzoyl-D-tartaric acid (125 kg, 332
mol, 1.0 equiv) in ethanol (148 kg) and deionized water (12.8
kg) in 5−10 min while maintaining the internal temperature
above 60 °C. After being stirred for 60 min at 60−70 °C, the
mixture was cooled to 50 °C at a rate 3 °C/h, while crystal
seeds were added in every hour. After being stirred for 1 h at
50 °C, the mixture was cooled to 10 °C (at 10 °C/h) and
stirred for at least 2 h at 5−10 °C. The crystals were then
filtered, rinsed with prechilled ethanol (15 kg), and dried
under vacuum at 50 °C to a constant weight. A 74.0 kg amount
of 9-DBT (98.7 area % purity, 95.5 ee %, water 5.6%, 120 mol,
(4aS,5R,7aS)-7a-(2-Fluorophenyl)-5-methyl-4a,5,7,7a-
tetrahydro-4H-furo[3,4-d][1,3]thiazin-2-amine (12). A
reactor was charged with isothiourea 10 (17.0 kg, 40.8 mol),
pyridine (49 kg) and toluene (44 kg). The resulting mixture
was cooled to below −15 °C. Trifluoromethanesulfonic
anhydride (12.7 kg, 45.0 mol, 1.10 equiv) was added while
maintaining the internal temperature below 0 °C. Upon
complete addition, stirring between −5−0 °C was continued
for 1.5 h and the reaction was monitored for complete
conversion of 10 (≥97.0%). Toluene (88 kg) and 20 wt %
aqueous ammonium chloride solution (71 kg) were added.
The resulting biphasic mixture was warmed to 15−25 °C,
stirred for at least 15 min and then allowed to partition. The
lower aqueous layer was separated, and the upper organic
phase was collected. The aqueous layer was extracted with
toluene (44 kg). The combined organic layers were washed
with deionized water (34 kg) and concentrated under reduced
pressure to give crude 11 as a thick brown oil. 2-propanol (41
kg) and 3.0 M aqueous sodium hydroxide (60 kg, 51 L) were
added, and the resulting reaction mixture was heated to 80 °C
for 6 h and monitored for complete conversion of 11
(>99.0%). n-Heptane (24 kg) and deionized water (17 kg)
were added while keeping the internal temperature above 60
°C. The resulting mixture was cooled to 0−5 °C, maintained at
this temperature for at least 2 h, and then filtered. The reactor
and filter cake were rinsed with a mixture of deionized water
(31 kg) and 2-propanol (2.7 kg), and then a mixture of
heptane (20 kg) and ethyl acetate (3.1 kg). The solids were
dried to give isothiourea 12 (9.45 kg, 96.7 wt % purity, 34.3
1
35.9% yield) was obtained as a white powder. H NMR (500
MHz, CD3OD) δ 8.14−8.10 (m, 4H), 7.60 (t, J = 7.4 Hz, 2H),
7.55 (td, J = 8.1, 1.3 Hz, 1H), 7.47 (t, J = 7.8 Hz, 4H), 7.46−
7.42 (m, 1H), 7.30−7.19 (m, 2H), 5.90 (s, 2H), 4.39−4.26
(m, 2H), 4.15 (dd, J = 10.4, 2.4 Hz, 1H), 3.93 (dd, J = 11.7,
4.4 Hz, 1H), 3.87 (dd, J = 11.7, 5.4 Hz, 1H), 2.52 (dt, J = 8.4,
5.0 Hz, 1H), 1.31 (d, J = 6.1 Hz, 3H); 13C NMR (125 MHz,
CD3OD) δ 171.37, 167.32, 161.76 (d, J = 245.5 Hz), 134.36,
132.46 (d, JCF = 9.8 Hz), 131.16, 130.98, 129.47, 128.98 (d, JCF
= 2.9 Hz), 126.23 (d, JCF = 3.9 Hz), 125.65 (d, JCF = 10.8 Hz),
117.84 (d, JCF = 23.5 Hz), 77.75, 75.73 (d, JCF = 4.9 Hz),
74.91, 67.01 (d, JCF = 2.0 Hz), 58.84, 56.28, 20.40.
H
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
mol, 84.1% yield). 1H NMR (500 MHz, CDCl3) δ 7.41 (td, J =
8.2, 1.6 Hz, 1H), 7.28−7.20 (m, 1H), 7.15−7.09 (m, 1H), 7.03
(ddd, J = 12.4, 8.1, 0.7 Hz, 1H), 4.81−4.42 (s, 2H), 4.61 (dd, J
= 8.7, 0.5 Hz, 1H), 4.40−4.27 (m, 1H), 3.80 (dd, J = 8.7, 2.2
Hz, 1H), 3.07 (dd, J = 13.3, 3.9 Hz, 1H), 2.70 (dd, J = 13.3,
rinsed with ethanol (4.1 kg), and the ethanol wash was
transferred into the first reactor. The reaction mixture was
stirred between 55 and 65 °C until complete consumption of
starting material 13 (target ≥99.0% conversion). The reaction
mixture was diluted with ethanol (51 kg) and cooled below 20
°C. After being stirred for 1−2 h, the mixture was filtrated
through a Celite pad (11 kg), rinsing with ethanol (106 kg).
Concentration of the filtrate in vacuo gave an orange solid,
which was dissolved in deionized water (84 kg) and stirred at
18−25 °C for at least 15 min. The resulting solution was
filtered through a Celite pad (2 kg) to remove the remaining
particles, rinsing with deionized water (21 kg). A 3.0 M sodium
hydroxide in deionized water solution (35 L) was added over 1
h. The resulting slurry was stirred for 1 h at 18−25 °C while
monitoring the pH. If necessary, additional 3 M NaOH was
added until pH ≥ 10. After 1−2 h of stirring at 18−25 °C, the
mixture was filtered, and the filter cake was rinsed with
deionized water three times (21 kg each). Drying in an oven
vacuum at 45 °C provided 34 (8.73 kg, 96.4 wt % purity, 29.9
3.9 Hz, 1H), 2.56−2.47 (m, 1H), 1.34 (d, J = 6.1 Hz, 3H); 13
C
NMR (125 MHz, CDCl3) δ 160.03 (d, JCF = 247.0 Hz),
149.41, 130.73 (d, JCF = 10.0 Hz), 129.80 (d, JCF = 4.1 Hz),
128.83 (d, JCF = 8.7 Hz), 124.02 (d, JCF = 3.4 Hz), 116.38 (d,
JCF = 23.5 Hz), 78.54 (d, JCF = 4.8 Hz), 76.43, 66.97 (d, JCF
=
4.6 Hz), 44.50 (d, JCF = 3.7 Hz), 23.70, 19.83.
(4aS,5R,7aS)-7a-(2-Fluoro-5-nitrophenyl)-5-methyl-
4a,5,7,7a-tetrahydro-4H-furo[3,4-d][1,3]thiazin-2-amine
(13). A reactor (reactor 1) was charged with trifluoroacetic
acid (35 kg) and cooled down to 0−5 °C. 12 (9.45 kg, 34.3
mol) was added portionwise with stirring, while maintaining
the internal temperature below 20 °C. The holding vessel and
reactor were rinsed with trifluoroacetic acid (14.2 kg). The
mixture was cooled below 5 °C and then sulfuric acid (12.3 kg)
was added while maintaining the internal temperature below
15 °C. The resulting solution was cooled down to 0−5 °C and
then fuming nitric acid was added (purity >90%, 3.0 kg, 41
mol, 1.2 equiv) while maintaining the internal temperature
below 25 °C. Upon complete addition, stirring was continued
with cooling (bath temperature: 0−5 °C) over 0.5 h, and the
reaction was monitored for complete consumption of 12
(target >98.0% conversion). Another reactor (reactor 2) was
charged with 3.0 M NaOH (246 kg, 217 L) and cooled to 3−7
°C. The reaction mixture was transferred into the second
reactor while maintaining the internal temperature below 30
°C. The first reactor was rinsed with deionized water (19 kg).
Additional 3 M NaOH to make the pH > 12 was added if
necessary. After stirring for 0.5−1 h at 23−27 °C, the mixture
was filtered and the reactor and filter cake were rinsed with
deionized water (38 kg). The filter cake was dried under
reduced pressure (T < 30 °C) overnight to give 13 (12.89 kg,
81.9 wt % purity, 33.9 mol, 98.9% yield). 1H NMR (500 MHz,
CDCl3) δ 8.39 (dd, J = 6.9, 2.9 Hz, 1H), 8.20−8.14 (m, 1H),
7.21 (dd, J = 10.8, 9.0 Hz, 1H), 4.55 (dd, J = 9.1, 0.9 Hz, 1H),
4.41−4.31 (m, 1H), 3.80 (dd, J = 9.1, 1.8 Hz, 1H), 3.05 (dd, J
= 13.5, 3.8 Hz, 1H), 2.74 (dd, J = 13.5, 4.1 Hz, 1H), 2.52−2.42
(m, 1H), 1.37 (d, J = 6.1 Hz, 3H); 13C NMR (125 MHz,
CDCl3) δ 163.56 (d, JCF = 258.9 Hz), 150.91, 144.23, 133.73
1
mol, 88.2% yield) as an off-white solid. H NMR (500 MHz,
DMSO) δ 6.76 (dd, J = 12.3, 8.6 Hz, 1H), 6.59 (dd, J = 7.1,
2.8 Hz, 1H), 6.45−6.36 (m, 1H), 5.91 (s, 2H), 4.87 (s, 2H),
4.29 (d, J = 8.2 Hz, 1H), 4.20−4.11 (m, 1H), 3.61 (dd, J = 8.0,
2.7 Hz, 1H), 2.93 (dd, J = 13.3, 3.7 Hz, 1H), 2.82 (dd, J =
13.3, 3.7 Hz, 1H), 2.37−2.28 (m, 1H), 1.23 (d, J = 6.1 Hz,
3H); 13C NMR (126 MHz, DMSO) δ 151.44 (d, JCF = 231.8
Hz), 147.68, 144.51, 130.86 (d, JCF = 11.7 Hz), 116.04 (d, JCF
= 25.4 Hz), 114.91 (d, JCF = 3.9 Hz), 112.87 (d, JCF = 7.8 Hz),
78.05 (d, JCF = 4.9 Hz), 75.48, 66.03 (d, JCF = 4.9 Hz), 43.22
(d, JCF = 3.9 Hz), 22.44, 19.81.
N-(3-((4aS,5R,7aS)-2-Amino-5-methyl-4a,5,7,7a-tetra-
hydro-4H-furo[3,4-d][1,3]thiazin-7a-yl)-4-fluorophenyl)-
5-(difluoromethyl)pyrazine-2-carboxamide (E2609). 5-
(Difluoromethyl)pyrazine-2-carboxylic acid 16 (9.5 kg, 54.7
mol, 1.1 equiv) was charged into a reactor. DMI (29 kg) was
added, and the mixture was cooled down to 0−5 °C. Thionyl
chloride (7.39 kg, 62.1 mol, 1.30 equiv) was added while
keeping the internal temperature below 10 °C. The resulting
solution was stirred at 4−10 °C until the conversion was
>95.0% by HPLC. The mixture was cooled down to 0 °C, and
a solution of 34 (14.0 kg, 47.8 mol) in DMI (36 kg) was
charged while keeping the internal temperature below 5 °C.
The vessel was rinsed with DMI (7.3 kg), and the reaction
mixture was stirred until the conversion was >99.0%.
Deionized water (168 kg) was charged, and the resulting
mixture was stirred at 15−20 °C for 0.5 h. EtOAc (168 kg) and
then 50% aq. NaOH (21 kg, 253 mol, 5.3 equiv) were added
while keeping the internal temperature below 30 °C. The pH
was monitored to make sure that it was above 10. The aqueous
phase was separated and extracted with EtOAc (63 kg). The
organic layers were combined and washed with 30 wt %
aqueous sodium chloride (94 kg) and deionized water twice
(70 kg). The mixture was filtered through a Celite pad (7.0
kg), rinsing with EtOAc (38 kg). The filtrate was concentrated
under reduced pressure at 40−50 °C. 1-Propanol (169 kg) was
charged, and the resulting mixture was warmed up to 90−100
°C. Stirring continued until a clear solution was obtained. The
mixture was cooled down to 0−5 °C over 2 h. After 1 h of
stirring, the mixture was filtered, rinsing with cold 1-propanol
twice (28 kg). Drying in vacuo at 45 °C provided the title
compound E2609 (18.5 kg, 97.4 wt % purity, 41.2 mol, 86.2%
(d, JCF = 12.3 Hz), 126.05 (d, JCF = 6.2 Hz), 124.76 (d, JCF
=
10.8 Hz), 117.53 (d, JCF = 26.3 Hz), 78.83 (d, JCF = 4.0 Hz),
76.96, 66.91 (d, JCF = 4.8 Hz), 45.26 (d, JCF = 3.1 Hz), 23.88,
19.63.
(4aS,5R,7aS)-7a-(5-Amino-2-fluorophenyl)-5-methyl-
4a,5,7,7a-tetrahydro-4H-furo[3,4-d][1,3]thiazin-2-amine
(34). A reactor (reactor 1) was charged with powder iron (11
kg, 170 mol, 5.0 equiv) and ethanol (67 kg) at 18−25 °C. To
the resulting slurry was added concentrated hydrochloric acid
(37 wt %, 1.7 kg, 0.50 equiv). The resulting mixture was
warmed up to 60−70 °C, stirred for 2 h, and cooled down to
50−55 °C. A saturated aqueous ammonium chloride solution
(27 wt %, 42 kg, 6.3 equiv) was added while keeping the
internal temperature between 45 and 55 °C. Another reactor
(reactor 2) was charged with 13 (12.89 kg, 33.9 mol) and
ethanol (38 kg). The resulting mixture was cooled to 10−15
°C and treated with concentrated hydrochloric acid (37 wt %,
3.4 kg, 1.0 equiv) while keeping the internal temperature below
25 °C. The resulting solution was added into the first reactor
while keeping the internal temperature below 65 °C; it was
1
yield) as an off-white solid. H NMR (500 MHz, CD3OD) δ
9.38 (s, 1H), 9.00 (s, 1H), 7.89−7.77 (m, 2H), 7.14 (dd, J =
I
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
11.9, 8.8 Hz, 1H), 6.94 (t, J = 54.3 Hz, 1H), 4.80 (s, 2H), 4.58
(d, J = 8.9 Hz, 1H), 4.37−4.28 (m, 1H), 3.81 (dd, J = 8.8, 2.4
Hz, 1H), 3.14 (dd, J = 13.5, 4.0 Hz, 1H), 2.88 (dd, J = 13.5,
twice (5.6 wt % solution, 200.4 kg; 2.9 wt % solution, 199.5
kg) to give a solution of 40. The remaining half of the solution
of 39 was processed in the same manner as described above.
The product solutions were combined and used in the next
4.1 Hz, 1H), 2.61−2.54 (m, 1H), 1.33 (d, J = 6.1 Hz, 3H); 13
C
1
NMR (125 MHz, CD3OD) δ 162.59, 158.36 (d, JCF = 244.5
Hz), 156.34, 151.63 (t, JCF = 25.9 Hz), 147.96, 144.72, 141.45
step without further purification (assumed 100% yield). H
NMR (600 MHz, DMSO-d6) δ 9.30 (brs, 1H), 9.28 (brs, 1H),
4.35 (q, J = 7.1 Hz, 2H), 3.95 (s, 3H), 1.22 (t, J = 7.1 Hz, 3H);
13C NMR (150 MHz, DMSO-d6) δ 163.3, 161.7 (t, JCF = 32
Hz), 148.2 (t, JCF = 28 Hz), 145.5, 145.1, 141.8 (t, JCF = 3 Hz),
111.5 (t, JCF = 252 Hz), 64.1, 53.2, 13.8; HRMS (ESI+) calcd
(t, JCF = 3.9 Hz), 135.15 (d, JCF = 2.0 Hz), 132.13 (d, JCF
=
11.7 Hz), 123.47 (d, JCF = 3.9 Hz), 122.90 (d, JCF = 8.8 Hz),
117.81 (d, JCF = 24.5 Hz), 114.38 (t, JCF = 239.6 Hz), 78.98 (d,
JCF = 4.9 Hz), 78.21, 67.91 (d, JCF = 4.9 Hz), 46.57 (d, JCF
=
+
for C10H11F2N2O4 ([M + H]+) 261.0681, found 261.0684.
2.9 Hz), 24.02, 20.00.
Methyl 5-Bromopyrazine-2-carboxylic Acid (39). To a
mixture of methyl 5-chloropyrazine-2-carboxylic acid 38 (35.4
kg, 205.1 mol) and acetonitrile (139.2 kg) was added TMSBr
(62.7 kg, 409.6 mol, 2.00 equiv), and 35 L of volatile was
distilled off at 80 °C under reduced pressure (atmosphere to
−25 kPa). After acetonitrile (28.1 kg) was added, 35 L of
volatile was distilled off. After cooling to room temperature,
EtOAc (159.4 kg) and 6.2 wt % aqueous NaHCO3 (283.5 kg)
were added. The aqueous layer was separated and re-extracted
with EtOAc (161.7 kg). The combined organic layers were
washed with 4.9 wt % aqueous NaCl (93.4 kg). After addition
of N,N-dimethylacetamide (DMA, 166.9 kg), the resulting
solution was concentrated to 177 L under reduced pressure at
60 °C. After the solution was azeotroped three times with
EtOAc (160 kg × 3) and then two times with an EtOAc/DMA
mixture (160 kg/22 kg × 2) under reduced pressure at 60 °C,
DMA (199.8 kg) was added to give 39 (397.1 kg, content 40.7
kg, 91% yield, 39:38 = 91:9) as a solution. The 39:38 ratio was
confirmed by HPLC as follows: Sunniest RP-AQUA (3 μm,
4.6 mm × 150 mm), 1.0 mL/min, oven temperature = 45 °C,
UV detection at 270 nm, mobile phase A = 990:10:1 v/v/v
H2O/MeCN/CF3CO2H, mobile phase B = 100:900:1 v/v/v
H2O/MeCN/CF3CO2H, gradient (time (min)/B conc (%)) =
0/0 → 30/30 → 35/100 → 45/100 → 45.01/0 → 55/stop,
relative retention time (RRT) of 38 = 0.90, RRT of 39 = 1.00.
1H NMR (600 MHz, DMSO-d6) δ 9.01 (s, 2H), 3.91 (s, 3H);
13C NMR (150 MHz, DMSO-d6) δ 163.7, 147.5, 146.3, 144.2,
Benzylamine Salt of 5-Carboxy(difluoro)-
methylpyrazine-2-carboxylic Acid (41). To a solution of
40 (theoretical content 48.9 kg, 187.9 mol) in toluene (117.6
kg) was added 48% aqueous NaOH (62.52 kg, 750.4 mol, 4.0
equiv). The resulting mixture was stirred at rt for 20 h. After
cooling to below 15 °C, the reaction mixture was treated with
concentrated HCl (20.3 kg) to adjust the pH to 3−4 (actual
4.2) and warmed to rt. The organic layer was separated and re-
extracted with water (24.6 kg). To the combined aqueous
extracts were added i-PrOAc (257.8 kg) and concentrated HCl
(68.2 kg). The aqueous layer was separated and re-extracted
with i-PrOAc (116.1 kg). To the combined organic extracts
were added i-PrOAc (20.1 kg) and 2-propanol (352.2 kg). The
resulting mixture was cooled to 10 °C. Benzylamine (60.3 kg,
562.8 mol, 3.00 equiv) was added while keeping the internal
temperature under 25 °C. The precipitated solids were
collected and washed with 2-propanol (77.8 kg). The solids
were dried under reduced pressure at 50 °C to give 41 (46.8
kg, 108 mol, 58% yield from 39) as an orange-brown powder.
1H NMR (600 MHz, DMSO-d6) δ 8.99 (d, J = 0.6 Hz, 1H),
8.73 (d, J = 0.7 Hz, 1H), 8.58 (brs, 6H), 7.46−7.33 (m, 10H),
4.02 (s, 4H); 13C NMR (150 MHz, DMSO-d6) δ 166.7, 163.8
(t, JCF = 30 Hz), 151.3, 150.4 (t, JCF = 30 Hz), 143.7, 140.7 (t,
JCF = 8.0 Hz), 134.9, 128.9, 128.7, 128.4, 113.5 (t, JCF = 255
Hz), 42.5; HRMS (ESI+) calcd for C7H3F2N2O4− ([M - H]−)
217.0066, found 217.0070.
5-(Difluoromethyl)pyrazine-2-carboxylic Acid (16).
To a mixture of 41 (44.8 kg, 103.6 mol), water (209.3 kg),
and toluene (194.1 kg) was added 48% aqueous NaOH (21.5
kg, 258.1 mol, 2.49 equiv), and the aqueous layer was
separated. To the aqueous extract was added 85% aqueous
H3PO4 (83.6 kg, 725.1 mol, 7.00 equiv), and then it was stirred
at 110 °C under a nitrogen atmosphere for 21 h. To the
reaction mixture was added 48% aqueous NaOH (77.6 kg,
931.4 mol, 9.0 equiv) to adjust the pH to 6−7 (actual 6.4), and
then it was stirred at 110 °C for 17 h. After cooling to rt, i-
PrOAc (157.9 kg) and concentrated HCl (91.7 kg) were added
and polish filtration of the resulting solution was performed,
rinsing with i-PrOAc (39.4 kg). The aqueous layer was
separated and re-extracted with i-PrOAc (197.0 kg). After the
extracts were combined with i-PrOAc (10.5 kg), the mixture
was washed with water five times (67.0 kg ×5). The five
aqueous layers were combined and then re-extracted with i-
PrOAc (197.0 kg). After the extracts were combined with i-
PrOAc (10.8 kg), the resulting solution was concentrated to
300 L under reduced pressure at 50 °C, and then polish
filtration was performed, rinsing with i-PrOAc (14.7 kg). The
filtrate was reconcentrated to 45 L under reduced pressure at
50 °C. The resulting mixture was heated at 60 °C to
completely dissolve the solid. The resulting solution was then
cooled to rt to induce crystallization. n-Heptane (42.9 kg) was
added dropwise over 1 h. After aging at 0 °C, the crystals were
+
141.8, 53.0; HRMS (ESI+) calcd for C6H5BrN2NaO2 ([M +
Na]+) 238.9427, found 238.9417.
Methyl 5-(2-Ethoxy-1,1-difluoro-2-oxoethyl)pyrazine-
2-carboxylate (40). To a mixture of zinc (13.0 kg, 198.8 mol,
2.11 equiv) and triglyme (97.6 kg) was added TMSBr (3.1 kg,
20.2 mol, 0.21 equiv). The resulting mixture was stirred at 70
°C for 1.5 h and cooled to rt. BrCF2CO2Et (48.0 kg, 236.5
mol, 2.52 equiv) was added for 2 h while keeping the internal
temperature around 25 °C. Formation of the Zn reagent was
monitored by 1H NMR. After the reaction mixture was
quenched with H2O/CD3CN to generate the corresponding
ethyl difluoroacetate, the proton of the resulted difluoromethyl
group (HF2C−) was quantitated by comparison with those of
triglyme, which is one of the reaction solvents.
After cooling to below 10 °C, the zinc reagent mixture was
added to a mixture of 39 in DMA (198.6 kg, content 20.4 kg,
94.0 mol) and DMA (101.6 kg) at 0 °C. The resulting mixture
was warmed to rt. CuBr (28.4 kg, 198.0 mol. 2.00 equiv) was
then added in five portions over 1 h, and the resulting reaction
mixture was stirred for 30 min at rt. After cooling to below 5
°C, the reaction mixture was added to a precooled (5 °C)
mixture of 6.2 wt % aqueous NaCl (317.1 kg), concentrated
HCl (97.2 kg), and toluene (302.1 kg), rinsing with toluene
(43.5 kg). The resulting mixture was warmed to rt. The
organic layer was separated and washed with aqueous NaCl
J
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
1H and 13C{1H} NMR spectra of compounds 6, 9-DBT,
16, 34, and E2609 (PDF)
collected and washed with n-heptane (15.4 kg). The crystals
were dried under reduced pressure at 50 °C to give 16 (12.4
kg, 70.4 mol, 68% yield) as an orange-yellow powder. 1H NMR
(600 MHz, CD3OD) δ 9.32 (s, 1H), 9.00 (s, 1H), 6.92 (t, J =
54.3 Hz, 1H); 13C NMR (150 MHz, CD3OD) δ 165.9, 151.8
(t, JCF = 26 Hz), 146.7, 146.5, 142.5 (t, JCF = 4 Hz), 114.3 (t,
AUTHOR INFORMATION
■
Corresponding Author
J
CF = 240 Hz); HRMS (ESI+) calcd for C6H3F2N2O2− ([M −
Dae-Shik Kim − Eisai Incorporated, Cambridge,
Massachusetts 02140, United States; orcid.org/0000-
0002-9519-6354; Email: dae-shik_kim@eisai.com
H]−) 173.0168, found 173.0164.
Procedure for the Synthesis of tert-Butyl 2-(1,1,1-
Trifluorobut-3-en-2-yloxy)acetate (45) by Lab-Scale
Flow Reaction. In a glass bottle were added acrolein (133
mL, 1.99 mol), CF3TMS (353 mL, 2.39 mol, 1.20 equiv), and
toluene (335 mL), resulting in stream 1 solution. In another
glass bottle were added TBAF (34.8 mL, 34.8 mmol of a 1.0 M
solution in THF, 0.0175 equiv), THF (56 mL), and toluene
(725 mL), resulting in stream 2 solution. Using the lab-scale
flow reactor, the two solutions were pumped through the
reactor setup with equal flow rates starting with 20 mL/min
and gradually increasing to ca. 200 mL/min until the reaction
solutions were consumed (the temperature profile of the
reaction run in this manner is shown in SI Table 1). Several
other options for the streams had been considered. For
example, placing the initiator (TBAF) with acrolein was found
to promote polymerization, leading to insoluble material which
would be detrimental to the flow reactor. Alternatively, a flow
system with three separate inputs (acrolein, CF3TMS, and
TBAF) was considered but not explored due to the increased
complexity. Therefore, premixing acrolein and CF3TMS in
desired stoichiometry and pumping as a single stream were
chosen to ensure initiation of the reaction in the desired
stoichiometry upon contact with the TBAF solution.
Authors
Kazuhiro Yoshizawa − API Research Japan, Pharmaceutical
Science & Technology Core Function Units, Medicine
Creation Center, Eisai Company, Limited, Kamisu-shi,
Ibaraki 314-0255, Japan
Branko Mitasev − Eisai Incorporated, Cambridge,
Massachusetts 02140, United States
Matthew Schnaderbeck − Eisai Incorporated, Cambridge,
Massachusetts 02140, United States
Huiming Zhang − Eisai Incorporated, Cambridge,
Massachusetts 02140, United States
Masayuki Omori − API Research Japan, Pharmaceutical
Science & Technology Core Function Units, Medicine
Creation Center, Eisai Company, Limited, Kamisu-shi,
Ibaraki 314-0255, Japan
Akio Kayano − API Research Japan, Pharmaceutical Science
& Technology Core Function Units, Medicine Creation
Center, Eisai Company, Limited, Kamisu-shi, Ibaraki 314-
0255, Japan
Mitsuo Nagai − API Research Japan, Pharmaceutical Science
& Technology Core Function Units, Medicine Creation
Center, Eisai Company, Limited, Kamisu-shi, Ibaraki 314-
0255, Japan
Kazunori Wakasugi − API Research Japan, Pharmaceutical
Science & Technology Core Function Units, Medicine
Creation Center, Eisai Company, Limited, Kamisu-shi,
Ibaraki 314-0255, Japan
The static mixer and heat exchange coil were submerged in
an ice bath during the reaction. The reaction flow was collected
in a 4 L glass container and analyzed by NMR, providing ca. 25
wt % of 44 in toluene, which was used in next step as is. To 44
(244 g total weight, 307 mmol, 1.20 equiv) from the flow-
reaction output were added tert-butyl bromoacetate (3) (50 g,
256 mmol) and nBu4NHSO4 (8.7 g, 26 mmol, 0.10 equiv).
The reaction mixture was cooled to 0 °C, and NaOH (205 g of
a 50 wt % solution in water, 2.56 mol, and 10 equiv) was added
in a controlled manner while keeping the internal temperature
below 10 °C. The reaction mixture was warmed to 15−20 °C
and analyzed by GC. Upon completion, the reaction mixture
was diluted with MTBE (300 mL) and then deionized water
(150 mL) while keeping the internal temperature below 30 °C.
The organic layer was separated and washed with deionized
water (150 mL) and aqueous saturated NaCl (75 mL). The
resulting organic solution was concentrated under vacuum to
afford 45 as a pale yellow oil (90 g, quantitative yield, contains
Yuzo Watanabe − API Research Japan, Pharmaceutical
Science & Technology Core Function Units, Medicine
Creation Center, Eisai Company, Limited, Kamisu-shi,
Ibaraki 314-0255, Japan
Farid Benayoud − Eisai Incorporated, Cambridge,
Massachusetts 02140, United States
Yuichi Suzuki − Neurology Business Group, Eisai Tsukuba
Research Laboratories, Tokodai, Ibaraki 300-2635, Japan
Takafumi Motoki − Neurology Business Group, Eisai Tsukuba
Research Laboratories, Tokodai, Ibaraki 300-2635, Japan;
orcid.org/0000-0002-5638-1629
Toshihiko Kaneko − Neurology Business Group, Eisai
Tsukuba Research Laboratories, Tokodai, Ibaraki 300-2635,
Japan
Mamoru Takaishi − Neurology Business Group, Eisai
Tsukuba Research Laboratories, Tokodai, Ibaraki 300-2635,
Japan
Tasuku Ishida − Neurology Business Group, Eisai Tsukuba
Research Laboratories, Tokodai, Ibaraki 300-2635, Japan
Kunitoshi Takeda − Neurology Business Group, Eisai
Tsukuba Research Laboratories, Tokodai, Ibaraki 300-2635,
Japan
1
ca. 30% residual solvents). H NMR (500 MHz, CDCl3) δ
5.86−5.74 (m, 1H), 5.59 (d, J = 17.5 Hz, 1H), 5.56 (d, J =
10.9 Hz, 1H), 4.37−4.30 (m, 1H), 4.11 (d, J = 16.5 Hz, 1H),
4.06 (d, J = 16.4 Hz, 1H), 1.40 (s, 9H); 13C NMR (125 MHz,
CDCl3) δ 168.51, 128.49 (d, JCF = 1.7 Hz), 123.86, 123.71 (q,
JCF = 281.8 Hz), 82.22, 78.67 (q, JCF = 31.5 Hz), 66.60, 28.02.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.oprd.1c00223.
Yoichi Kita − Neurology Business Group, Eisai Tsukuba
Research Laboratories, Tokodai, Ibaraki 300-2635, Japan
Noboru Yamamoto − Eisai Europe Limited, Hatfield
Hertfordshire AL10 9SN, United Kingdom
Additional experimental details/spectroscopic data for
Schemes 1, 2, 3, 7, and 9 and Figures 2 and 3; copies of
K
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
78, 2217−2222. (b) Heaney, H. The Benzyne and Related
Intermediates. Chem. Rev. 1962, 62, 81−97.
(6) L’Heureux, A.; Beaulieu, F.; Bennett, C.; Bill, D. R.; Clayton, S.;
LaFlamme, F.; Mirmehrabi, M.; Tadayon, S.; Tovell, D.; Couturier,
M. Aminodifluorosulfinium Salts: Selective Fluorination Reagents
with Enhanced Thermal Stability and Ease of Handling. J. Org. Chem.
2010, 75, 3401−3411.
(7) (a) Ruppert, I.; Schlich, K.; Volbach, W. Die ersten CF3-
substituierten organyl(chlor)silane. Tetrahedron Lett. 1984, 25, 2195−
2198. (b) Prakash, G. K. S.; Yudin, A. K. Perfluoroalkylation with
Organosilicon Reagents. Chem. Rev. 1997, 97, 757−786. (c) Prakash,
G. K. S.; Mandal, M. Nucleophilic trifluoromethylation tamed. J.
Fluorine Chem. 2001, 112, 123−131. (d) Liu, X.; Xu, C.; Wang, M.;
Liu, Q. Trifluoromethyltrimethylsilane: Nucleophilic Trifluoromethy-
lation and Beyond. Chem. Rev. 2015, 115, 683−730. and references
therein.
Afzal Khan − Eisai Europe Limited, Hatfield Hertfordshire
AL10 9SN, United Kingdom
Paschalis Dimopoulos − Eisai Europe Limited, Hatfield
Hertfordshire AL10 9SN, United Kingdom
Christopher N. Farthing − Eisai Europe Limited, Hatfield
Hertfordshire AL10 9SN, United Kingdom
Adrian Hall − Eisai Europe Limited, Hatfield Hertfordshire
AL10 9SN, United Kingdom; orcid.org/0000-0001-
7869-6835
Arani Chanda − Eisai Incorporated, Cambridge,
Massachusetts 02140, United States
Lily Lu − Eisai Incorporated, Cambridge, Massachusetts
02140, United States
Markus Bracke − Eisai Incorporated, Cambridge,
Massachusetts 02140, United States
(8) (a) Takahashi, O.; Saka, T.; Mikami, K.; Nakai, T. [2,3]Wittig
Rearrangements of “Lithium Enolates” Generated from α-Allyloxy
Esters, Ketones, and Oximes. Some Implications of the Enolate
Structures. Chem. Lett. 1986, 15, 1599−1602. For data of compound
25, see: (b) Kiefer, A.; Gawas, D.; Kazmaier, U. Palladium-Catalyzed
Allylic Alkylation Reactions of Glycolates as a Tool for the Synthesis
of Functionalized α-Hydroxy Esters. Eur. J. Org. Chem. 2015, 2015,
5810−5816.
Francis G. Fang − Eisai Incorporated, Cambridge,
Massachusetts 02140, United States
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.oprd.1c00223
Author Contributions
The manuscript was written through contributions of all
authors.
(9) Stojanowa-Antoszczyszyn, M.; Brenk, D.; Janus, E. Application
of Phase Transfer Catalysts in the 0-Alkylation of Sorbitol with Allyl
Chloride. C. R. Acad. Bulg. Sci. 2001, 54, 51−54.
Notes
The authors declare no competing financial interest.
(10) Krasovskiy, A.; Knochel, A. A LiCl-Mediated Br/Mg Exchange
Reaction for the Preparation of Functionalized Aryl- and Hetero-
arylmagnesium Compounds from Organic Bromides. Angew. Chem.,
Int. Ed. 2004, 43, 3333−3336.
ACKNOWLEDGMENTS
■
We thank our colleagues N. Shibuguchi (Eisai Co., Ltd), I.
Kushida (Eisai Co., Ltd), K. Kono (Eisai Co., Ltd), E. Troy
(Eisai Inc.), and Dr. M. Alaimo (Eisai Inc.) for data analysis
support, M. Vachon (Eisai Inc.) and M. Takahashi (Eisai Co.,
Ltd) for process validations support, and Dr. H.-w. Choi for
valuable feedback and proofreading. We also thank Dr. R.
Staples (Crystallographic Resources Inc.) for X-ray crystallo-
graphic data of compound 9-DBT.
(11) Am Ende, D. J.; Clifford, P. J.; DeAntonis, D. M.; SantaMaria,
C.; Brenek, S. J. Preparation of Grignard Reagents: FTIR and
Calorimetric Investigation for Safe Scale-Up. Org. Process Res. Dev.
1999, 3, 319−329.
(12) See Supporting Information for details.
(13) In early scale-up runs, the formation of Grignard reagent 29
step was held over 4 h before addition of morpholine amide 28.
However, based on accumulated ReactIR data, the time has been
shortened to 1.5 h.
(14) Doyle, L.; Heaney, F. An investigation of structure-reactivity
relationships of δ-alkenyl oximes; Competitive thermal reactions
leading to cyclic nitrones and/or N-unsubstituted bicyclic isoxazoli-
dines. Tetrahedron 2010, 66, 7041−7049.
(15) (a) Heaney, F.; O’Mahony, C. The influence of oxime
stereochemistry in the generation of nitrones from ω-alkenyloximes
by cyclization or 1,2-prototropy. J. Chem. Soc., Perkin Trans. 1 1998,
341−350. (b) Nsikabaka, S.; Harb, W.; Ruiz-López, M. F. The role of
water on the acid-promoted E/Z isomerization of oximes in aqueous
solution. J. Mol. Struct.: THEOCHEM 2006, 764, 161−166.
(c) Another explanation for our results is that the rate of the
unwanted side reaction (APT) might slow down in aqueous media
(suggested by one of the reviewers).
REFERENCES
■
(1) Selkoe, D. J.; Hardy, J. The amyloid hypothesis of Alzheimer’s
disease at 25 years. EMBO Mol. Med. 2016, 8, 595−608.
(2) Moussa-Pacha, N. M.; Abdin, S. M.; Omar, H. A.; Alniss, H.; Al-
Tel, T. H. BACE1 inhibitors: Current status and future directions in
treating Alzheimer’s disease. Med. Res. Rev. 2020, 40, 339−384.
(3) (a) Suzuki, Y.; Motoki, T.; Kaneko, T.; Takaishi, M.; Ishida, T.;
Kita, Y.; Takeda, K.; Yamamoto, N.; Khan, A.; Dimopoulos, P. Fused
aminodihydrothiazine derivatives. Patent WO200991016A1, 2009.
(b) Albala, B.; Kaplow, J. M.; Lai, R.; Matijevic, M.; Aluri, J.; Satlin, A.
S4−04−01: CSF amyloid lowering in human volunteers after 14 days’
oral administration of the novel BACE1 inhibitor E2609. Alzheimer's
Dementia 2012, 8 (4S_Part_20), S743−S743. (c) Das, B.; Yan, R. A.
Close Look at BACE1 Inhibitors for Alzheimer’s Disease Treatment.
CNS Drugs 2019, 33, 251−263.
(16) On the basis of overall yields from ketone 6 to 9-DBTA in
scale-up synthesis with a 40% assumed yield of the DBTA resolution
step. See Supporting Information for detailed HPLC conditions for
Figure 2.
(4) (a) Hassner, A.; Maurya, R.; Mesko, E. Intramolecular oxime
olefin cycloadditions. Stereospecific formation of functionalized
pyrrolidines. Tetrahedron Lett. 1988, 29, 5313−5316. (b) Garcia-
Losada, P.; DeBaillie, A. C.; Eugenio de Diego, J.; Green, S. J.;
Hansen, M. M.; Jaramillo, C.; Johnson, M.; Kaoudi, T.; Li, J.; Lindsay-
Scott, P. J.; Mateos, C.; Mergott, D. J.; Rincon, J. A.; Rothhaar, R. R.;
Seibert, K. D.; Watson, B. M.; Winneroski, L. L.; Gangula, S.; Jing, D.;
Sun, H.; Zhang, L.; Frederick, M. O. Synthesis, Optimization, and
Large-Scale Preparation of the Low-Dose Central Nervous System-
Penetrant BACE1 Inhibitor LY3202626 via a [3 + 2] Nitrone
Cycloaddition. Org. Process Res. Dev. 2020, 24, 306−314.
(17) (a) Mitasev, B.; Kim, D.-S.; Zhang, H.; Schnaderbeck, M. J.;
Farthing, C. Methods and Compounds Useful in the Synthesis of
Fused Aminodihydrothiazine Derivatives. US Patent US
20130197244A1, 2013. After we developed this chemoselective
coupling, similar selective couplings for diamines were reported, see:
(b) Hansen, M. M.; Jarmer, D. J.; Arslantas, E.; DeBaillie, A. C.;
Frederick, A. L.; Harding, M.; Hoard, D. W.; Hollister, A.; Huber, D.;
Kolis, S. P.; Kuehne-Willmore, J. E.; Kull, T.; Laurila, M. E.; Linder, R.
J.; Martin, T. J.; Martinelli, J. R.; McCulley, M. J.; Richey, R. N.;
Starkey, D. R.; Ward, J. A.; Zaborenko, N.; Zweifel, T. Synthesis of
BACE Inhibitor LY2886721. Part II. Isoxazolidines as Precursors to
(5) (a) Gilman, H.; Gorsich, R. D. Some Reactions of o-
Halobromobenzenes with n-Butyllithium. J. Am. Chem. Soc. 1956,
L
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
pubs.acs.org/OPRD
Article
Chiral Aminothiazines, Selective Peptide Coupling, and a Controlled
Reactive Crystallization. Org. Process Res. Dev. 2015, 19, 1214−1230.
(18) Schlosser, M.; Cottet, F. Silyl-Mediated Halogen/Halogen
Displacement in Pyridines and Other Heterocycles. Eur. J. Org. Chem.
2002, 2002, 4181−4184.
(19) Zn activation by TMSBr instead of TMSCl did not generate
aggregation. see: (a) Osato, H.; Osaki, M.; Oyama, T.; Takagi, Y.;
Hida, T.; Nishi, K.; Kabaki, M. Pilot Plant Preparation of tert-Butyl-4-
(2-hydroxyethyl)-4-(pyrrolidin-1-yl)-piperidine-1-carboxylate, An In-
termediate of Novel Antiarteriosclerotics, Via a Safe, Scalable
Reformatsky-Type Reaction. Org. Process Res. Dev. 2011, 15, 1433−
1437. BrZnCF2COO2R can be stabilized in triglyme, see: (b) Burton,
D. J.; Easdon, J. C. The α,α-Difluoro Reformatsky Reagent:
Pregeneration and Structural Determination. J. Fluorine Chem. 1988,
38, 125−129.
(20) Pan, Y.; Holmes, C. P.; Tumelty, D. Novel α,α-Difluor-
ohomophthalimides via Copper-Catalyzed Tandem Cross-Coupling−
Cyclization of 2-Halobenzamides with α,α-Difluoro Reformatsky
Reagent. J. Org. Chem. 2005, 70, 4897−4900.
(21) Fujikawa, K.; Fujioka, Y.; Kobayashi, A.; Amii, H. A New
Method for Aromatic Difluoromethylation: Copper-Catalyzed Cross-
Coupling and Decarboxylation Sequence from Aryl Iodides. Org. Lett.
2011, 13, 5560−5563.
(22) (a) Okusu, S.; Hirano, K.; Yasuda, Y.; Tokunaga, E.; Shibata, N.
Flow trifluoromethylation of carbonyl compounds by Ruppert-
Prakash reagent and its application for pharmaceuticals, efavirenz
and HSD-016. RSC Adv. 2016, 6, 82716−82720. (b) Hirano, K.;
Gondo, S.; Punna, N.; Tokunaga, E.; Shibata, N. Gas/Liquid-Phase
Micro-Flow Trifluoromethylation using Fluoroform: Trifluoromethy-
lation of Aldehydes, Ketones, Chalcones, and N-Sulfinylimines.
ChemistryOpen 2019, 8, 406−410.
(23) See the Supporting Information for additional details regarding
reactor design, IR graphs, and temperature profile.
M
https://doi.org/10.1021/acs.oprd.1c00223
Org. Process Res. Dev. XXXX, XXX, XXX−XXX