Development of an Enantioselective [3 + 2] Cycloaddition to Synthesize the Pyrrolidine Core of ABBV-3221 on Multikilogram Scale

Article abstract of DOI:10.1021/acs.oprd.9b00292

The tetrasubstituted pyrrolidine core of ABBV-3221 was synthesized by catalytic, enantioselective cycloaddition. A Cu(I) catalyst system was identified as ideal for further development, which gave a 78% yield of 99% purity product after optimization. The

Full text of DOI:10.1021/acs.oprd.9b00292

Communication
pubs.acs.org/OPRD
Cite This: Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Development of an Enantioselective [3 + 2] Cycloaddition To
Synthesize the Pyrrolidine Core of ABBV-3221 on Multikilogram
Scale
John Hartung,*,† Stephen N. Greszler,‡ Russell C. Klix,† and Jeffrey M. Kallemeyn†
†Process Research and Development, AbbVie, Inc., 1 North Waukegan Road, North Chicago, Illinois 60064, United States
‡Discovery Chemistry and Technology, AbbVie, Inc., 1 North Waukegan Road, North Chicago, Illinois 60064, United States
S
* Supporting Information
steric interactions from the tert-butyl group and enabling
ABSTRACT: The tetrasubstituted pyrrolidine core of
ABBV-3221 was synthesized by catalytic, enantioselective
cycloaddition. A Cu(I) catalyst system was identified as
ideal for further development, which gave a 78% yield of
99% purity product after optimization. The processes of
catalyst selection, optimization, and crystallization of the
cycloaddition product are described herein.
further elaboration with the ether and amide substituents.
Early in our efforts it was established that an enantioselective
[3 + 2] cycloaddition was an ideal method to forge the
stereogenic pyrrolidine core in a convergent manner using a
mature synthetic methodology.5−8 To achieve acceptable
reactivity in the cycloaddition, an activating nitro group was
appended on the olefin component that could later be
converted to the requisite ether linkage by functional group
interconversion.8,9 (See Scheme 1.)
KEYWORDS: enantioselective catalysis, copper,
[3 + 2] cycloaddition, pyrrolidine
Scheme 1. Abbreviated Synthetic Strategy toward ABBV-
3221
INTRODUCTION
■
Cystic fibrosis (CF) is a debilitating and ultimately fatal genetic
disease impacting more than 70 000 people worldwide.1 CF is
caused by a large variety of mutations leading to dysfunction of
the cystic fibrosis transmembrane receptor (CFTR) protein,2
which is responsible for anion transport in tissues throughout
the body.3 Malfunction of the CFTR protein results in
recurring lung infections and pancreatic insufficiency due to
buildup of mucus. ABBV-3221 is a modulator of the CFTR
protein in development as a potential therapy for cystic
fibrosis.4 (See Figure 1.)
In the project’s transition from Discovery to Process
chemistry, work was carried out to develop a robust and
scalable cycloaddition to fuel multikilogram synthesis of
ABBV-3221. In particular, identification of a catalyst system,
optimization of reaction parameters impacting stereoselectiv-
ity, and controls for process-related impurities impacting
catalytic activity were explored to achieve an optimized
process. Ultimately, a streamlined process utilizing crude
imine in the enantioselective [3 + 2] cycloaddition and
subsequent crystallization of high-purity pyrrolidine product
directly from the reaction mixture was successfully executed on
a multikilogram scale.
Figure 1. Structure of ABBV-3221.
The structure of 1 is dominated by its penta-substituted
pyrrolidine core containing four contiguous stereocenters.
Given the challenges in rejecting stereoisomeric impurities, a
highly stereoselective transformation was required to limit the
formation of the 15 undesired stereoisomers arising from the
four stereocenters. The stereogenic tetrahydropyranyl amide,
epimerizable carboxylate functionality, and sterically bulky tert-
butyl group appended on the pyrrolidine core further
contributed to the candidate’s structural complexity. A strategy
to synthesize 1 would require a robust transformation to set
the ring stereochemistry while accommodating significant
RESULTS AND DISCUSSION
■
Synthesis of Cycloaddition Starting Materials. Nitro
olefin 2 was synthesized by condensation of nitromethane and
pivaldehyde, and subsequent elimination mediated by
trifluoroacetic anhydride (TFAA) and triethylamine (Scheme
2).10 The condensation produced varying amounts of 7 and 2,
which were funneled to 2 upon treatment with TFAA and
Received: June 27, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.9b00292
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
a
Scheme 2. Synthesis of Nitro Olefin 2
Scheme 4. Design of Catalyst Screen for Cu-Catalyzed
Enantioselective [3 + 2] Cycloaddition
a
a
Reagents and conditions: (a) 5 (1.0 equiv), 6 (1.1 equiv), sodium
hydroxide (1.3 equiv), ethanol, 0−10 °C, 1 h; (b) TFAA (1.15 equiv),
triethylamine (2.3 equiv), methylene chloride, −15 to −10 °C, 1 h,
65% yield.
a
Reagents and conditions: (a) 2 (2 μmol, 1.0 equiv), 3 (2.2 μmol, 1.1
equiv), [Cu(OTf)]2*toluene (0.1 μmol, 5 mol %), ligand (0.22 or
0.44 μmol, 10 or 20 mol %), KOtBu (0.16 μmol, 8 mol %), THF, 0
°C, 18 h.
triethylamine. Product 2 was purified by column chromatog-
raphy to give an isolated yield of 65% at 99% purity.
Imine 3 was synthesized by condensation of glycine ethyl
ester hydrochloride with ortho-tolualdehyde in the presence of
triethylamine and drying agent in methylene chloride (Scheme
3).11 Higher yields were obtained by sequestration of water
Table 1. Selected Results from Catalyst Screen
a
Scheme 3. Synthesis of Imine Reactant 3
a
Reagents and Conditions: (a) 8 (1.1 equiv), 9 (1.0 equiv),
triethylamine (1.1 equiv), sodium sulfate (anhydrous, 2.0 equiv),
methylene chloride, 20 °C, 2 h, 93% yield.
with drying agent (magnesium or sodium sulfate) than with
azeotropic distillation (Dean−Stark apparatus). Azeotropic
removal of water resulted in decomposition of the imine due to
its thermal instability.12 Residual solids after workup could be
minimized by using sodium sulfate over magnesium sulfate.
Rejection of reaction byproduct triethylamine HCl, excess 8,
and drying agent could be achieved by filtration and washing of
the organic layer with water and brine. The resultant solution
of 3 in methylene chloride was solvent swapped to cyclopentyl
methyl ether (CPME; to a final concentration of 45% w/w) to
give a potency adjusted yield of 93%. The CPME solution was
then used directly in the cycloaddition reaction. The two
starting materials are the only impurities observed in isolated 3
in this process (8 0.7% w/w, 9 0.5% w/w), which did not have
an impact on downstream chemistry. The resulting solution of
3 in CPME (45% w/w) was stable for up to 7 days at 0 °C as
determined by qNMR.
Cu-Catalyzed Enantioselective Cycloaddition. Devel-
opment of the [3 + 2] cycloaddition reaction was focused on
achieving high reaction stereoselectivity, yield, and reproduci-
bility at what was deemed an acceptable catalyst loading of
approximately 1 mol %. Discovery efforts toward related
pyrrolidines were aided by an enantioselective and high-
yielding [3 + 2] cycloaddition catalyzed by Cu(I) and ligand
10.13 A high-throughput screen of Cu(I)/ligand complexes was
carried out to exclude potentially more efficient or lower-cost
alternatives to the initially identified system (Scheme 4 and
Table 1; see the SI for full details). An array of ligands was
combined with the precatalyst [CuOTf]2*toluene and catalytic
base14 to form the requisite azomethine ylide from imine 3. A
high endo diastereomeric ratio (dr) was observed with diverse
ligand scaffolds and electronics. Ligands achieving >80% ee
were limited to Segphos and push−pull bidentate ferrocenyl
derivatives, with 10 affording significantly higher enantiose-
lectivity compared to ferrocenyl bisphosphines (Table 1). The
results of the screen clearly indicated that ligand 10 gave the
highest enantioselectivity (99.5:0.5 er endo product; er,
enantiomeric ratio), complete conversion, and acceptable
diastereoselectivity (83:17 dr endo/exo) and was the lead for
further development.8a
The impact of reaction parameters on reaction outcome
with CuOTf/10 were investigated to reduce catalyst loading
from the initial screening level of 10 mol % and further
optimize stereoselectivity and yield. Studies aimed at reducing
catalyst loading determined that 0.25 mol % [Cu-
(OTf)]2*C6H6 and 0.55 mol % 10 were minimally required
B
DOI: 10.1021/acs.oprd.9b00292
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
for full conversion (Table 2). Below this quantity, conversion
and reaction er dropped significantly, and dr of the product
Scheme 5. Cu-Catalyzed Enantioselective [3 + 2]
Cycloaddition
a
Table 2. Optimization of Precatalyst, Ligand, and Base
a
Charge
[CuOTf]2C6H6
mol %
10 mol % KOtBu mol % yield
dr
er endo
0.5
0.25
0.125
1.1
0.55
0.275
0.8
0.4
0.2
92
7:1
8.6:1
13:1
98.5:1.5
98.2:1.8
70:30
87
b
18
a
Reagents and conditions: (a) 2 (1.0 equiv), 3 (1.1 equiv),
[Cu(OTf)]2*C6H6 (0.25 mol %), 10 (0.55 mol %), KOtBu (0.40
mol %), CPME, −15 °C, 85% assay yield, 8.1:1 dr (endo/exo),
98.5:1.5 er.
a
Conditions: 2 (1.0 equiv), 3 (1.1 equiv), CPME (1 M in 3), 0 °C,
b
0.5 h. Yield after 1.5 h.
increased, indicative of a highly diastereoselective but slower
background reaction. Ligand to Cu ratio had no observable
impact on reaction outcome in the range studied (0.9−1.2
equiv ligand to Cu).
It was found that reaction solvent had the largest impact on
diastereoselectivity (endo/exo), enantioselectivity, and con-
version as shown in Table 3.15 Ethereal solvents tended to give
to 16 000 ppm, which corresponded to 0.005−1.8 mol % with
respect to limiting reagent 2. These levels were sufficient to
partially or fully neutralize the 0.4 mol % KOtBu charge,
resulting in abrogation of reactivity. TFA levels in nitro olefin
lots supplying plant campaigns were controlled to not more
than 300 ppm because neutralization of acid by an increase of
the base charge did not fully restore the reaction rate, although
stereoselectivity was not impacted (Table 4). Reactions stalled
Table 3. Impact of Reaction Solvent on Stereoselectivity
and Conversion
a
Table 4. Effect of Increasing KOtBu Charge in
Cycloaddition of 2 Containing 1.39 mol % TFA
solvent
dr (endo/exo)
er endo product
conversion (%)
MTBE
CPME
toluene
DME
2-MeTHF
THF
9.1:1
8.1:1
7.5:1
5.5:1
5.3:1
3.9:1
90:10
98:2
97:2
92:8
99:1
97:2
34
>99
98
63
98
KOtBu (mol %)
assay yield (%)
dr (endo/exo)
er endo product
1.79
2.79
51
45
8:1
8:1
98.8:1.2
98.7:1.3
80
at incomplete conversion without the presence of additional
byproducts, suggesting premature catalyst decomposition. This
result suggested that modification of the CuOTf/10 complex
by trifluoroacetate was occurring in situ to give a less stable or
active catalytic species. Acceptable levels of TFA in 2 could be
achieved by implementing additional sodium bicarbonate
washes during workup.
Crystallization of Cycloaddition Product 2. Because the
reaction mixture contained catalytic Cu, 10, and KOtBu, it was
envisioned that a direct crystallization of 4 from the reaction
mixture would be possible, obviating the need for a workup.
To explore this, a solubility study was carried out on pure 4
(Figure 2). n-Heptane was found to be a suitable antisolvent
a
Conditions: 2 (3.9 mmol, 1.0 equiv), 3 (4.3 mmol, 1.1 equiv),
[Cu(OTf)]2*C6H6 (9.7 μmol, 0.25 mol %), 10 (21 μmol, 0.55 mol
%), KOtBu (15 μmol, 0.40 mol %), 0 °C.
high conversion with the exception of MTBE, which was likely
due to low catalyst solubility in this solvent as made evident by
heterogeneity during precomplexation. While conversion and
diastereoselectivity could be significantly improved on moving
from the screening solvent tetrahydrofuran (THF) to 2-
methyltetrahydrofuran (2-MeTHF), cyclopentyl methyl ether
(CPME) was chosen as the reaction solvent due to the higher
stereoselectivity and conversion observed in lab-scale experi-
ments in this solvent and its low environmental impact.
Optimization studies culminated in the conditions chosen
for scale-up of 0.25 mol % [Cu(OTf)]2*C6H6, 0.55 mol % 10,
and 0.4 mol % KOtBu in CPME (1 M in 2) at 0 °C for 30 min
giving an assay yield of 85%, dr of 8.1:1, and er of 98:2
(Scheme 5), but variability in the reaction outcome led us to
investigate possible sources of irreproducibility before scale-up.
Two contributing factors were identified: adventitious oxygen
and purity of nitro olefin. To prevent catalyst decomposition
from adventitious oxygen, plant runs employed an oxygen
sensor to maintain O2 levels below 10 ppm in reactors.16
Observed lot to lot variability in the nitro olefin led to a
deeper investigation into impurities present in the starting
material. As shown in Scheme 2, the final step in the synthesis
of 2 is a TFAA mediated elimination to form the olefin.
Consequently, nitro olefin lots had varying levels of the
byproduct trifluoroacetic acid (TFA), which neutralizes the
catalytic base charge in the [3 + 2] cycloaddition. TFA levels
were determined by ion chromatography and ranged from 45
Figure 2. Solubility of 4 as a function of n-heptane fraction in CPME/
n-heptane mixtures.
C
DOI: 10.1021/acs.oprd.9b00292
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
(solubility: 2.6 mg/mL) relative to the CPME reaction solvent
(solubility: 101 mg/mL). At full conversion, the reaction
reached a supersaturation ratio of approximately 2.9, resulting
in spontaneous seed formation. The reaction mixture was
subsequently concentrated by distillation and diluted with n-
heptane to produce a readily filtered product slurry with
acceptable product purity and loss to mother liquors.
The desired end-point for the crystallization was determined
by exploring product purity and recovery as a function of
solvent composition in lab-scale experiments. Major process-
related impurities observed in the reaction crude were the
imine 3 (present in 10% excess), diastereomer 11, and
enantiomer 12 (Figure 3).
Regardless of the reaction enantioselectivity, crystallization
of the product from the reaction mixture took place readily on
scale. After reaction completion, the mixture was concentrated
to 0.5 L/kg 4; n-heptane (3 L/kg 4) was charged over 1 h, and
then, the slurry was cooled to 0 °C over 4 h and filtered. The
crystallization conditions resulted in rejection of the exo
diastereomer 11 from a crude level of 8 a% to 0.3 a% and imine
from 20 a% to 0.4 a% in the isolated product, while the
enantiomer 12 level remained unchanged from the crude
(1.5%). Crystallization studies of 4 at different enantiopurities
suggested that there did not exist a significant difference in
solubility between racemic and enantiopure crystal forms,
resulting in an unavoidably poor rejection of enantiomer 12.
Nonetheless in the first pilot plant campaign, a very fine
racemate solid was observed in the crystallization, which could
be partially excluded by filtering the product into multiple
portions to give a batch er of 97:3. Formation of the racemic
solid, which had not been observed in lab-scale experiments,
was attributed to lower reaction enantioselectivity (96:4 er)
and uncontrolled nucleation during concentration of the
reaction mixture rather than at reaction completion. In the
second campaign, in which a reaction er of 98.5:1.5 was
achieved, a total of 19.21 kg of 4 (78% yield) at 98.8a% purity
(1.5% enantiomer) and 97.5% potency could be isolated
directly from the reaction mixture through the crystallization
process described above. The level of enantiomer 12 was
acceptable because isolations downstream of this step resulted
in rejection to <0.05a%.17
Figure 3. Reaction byproducts observed in [3 + 2] cycloaddition.
While a 5 L/kg process at 0.75 fraction n-heptane gave high-
purity product (99.2 a%), product loss (9%) was deemed to be
too high (Table 5). It was determined that lower (13% v/v)
CPME fraction reduced the loss of product to the mother
liquor to an acceptable level and did not significantly reduce
product purity.
CONCLUSION
■
A highly enantioselective cycloaddition was developed to form
the pyrrolidine core of ABBV-3221 from readily available
starting materials. High-throughput screening validated an
originally identified catalyst system, which was optimized by
straightforward exploration of reaction parameters. Solvent
choice was critical to improving selectivity and yield and also
benefitted the isolation of cycloaddition product by enabling
direct crystallization from the reaction mixture. Strict control
of TFA in the nitro olefin starting material 2 was required to
achieve reproducible reactivity, while a crude solution of imine
3 could be used directly after workup. The cycloaddition was
carried out on a multikilogram scale to provide material for
early deliveries of ABBV-3221 to supply preclinical and clinical
trial demand.
Table 5. Recovery and Purity of 4 by Crystallization from
CPME/n-Heptane at 0 °C
11 in
solid (a
%)
fraction
CPME
fraction total volume
loss of 4 to
liquors (%)
purity of
4 (a%)
heptane
(L/kg 4)
0.25
0.13
0.75
0.87
5.1
10.3
9
4
0.3
0.3
99.2
98.9
Further process intensification was achieved on scale-up by
reducing the final volume from 10.3 to 3.5 L/kg with respect to
4 with no negative impact on isolated product purity.
Pilot Plant Campaign Results. The cycloaddition was
executed in two pilot plant campaigns aimed at producing
multikilogram quantities of ABBV-3221 for preclinical
toxicology and clinical trial supply. Optimized reaction
conditions were scaled to produce two 20 kg batches of
product 4 in high enantiopurity and yield. Observations and
modifications on scale are summarized in the following
paragraphs.
On scale, control of the exotherm generated in the reaction
was crucial to achieving high enantioselectivity. While lab runs
on the gram scale gave a reaction er of 98.5:1.5 at 0 °C reactor
temperature, the first campaign gave a reaction enantioselec-
tivity of 96:4 er at 0 °C internal temperature. This result
suggested that the rapid and exothermic cycloaddition reaction
produced a feed zone of relatively higher temperature due to
poorer mixing, which eroded the enantioselectivity. Control of
the exotherm upon addition of nitro olefin by diluting the
reactant in CPME to 1.6 L/kg, increasing the addition time
from 0.5 to 3.5 h, and maintaining an internal temperature of
not more than −10 °C was shown to improve enantiose-
lectivity to 98.5:1.5 er in the second campaign.
EXPERIMENTAL DETAILS
■
General Remarks. Reagents and solvents were used as
received without further purification. Reactions were carried
out under nitrogen atmosphere. Large-scale reactions were
carried out in pilot plant equipment (glass-lined reactors with
agitator and jacket, stainless steel canisters) after a safety
evaluation was completed. NMR spectra were obtained using a
Varian 400 or 700 MHz NMR spectrometer. 1H and 13C NMR
chemical shifts are expressed in parts per million (δ) relative to
CDCl3 solvent. Analytical HPLC was performed on an Agilent
1100/1200 HPLC system equipped with either an Ascentis
Express C18 (4.6 mm × 15 cm, 2.7 μm) or Chirapak IC (4.7
mm × 15 cm, 5 μm) column.
Preparation of Nitro Olefin (2). To the solution of
pivalaldehyde (3.51 kg, 40.8 mol, 1 equiv) and MeNO2
(2.74 kg, 44.8 mol, 1.1 equiv) in ethanol (7 L) was added
NaOH (10 M, 5.3 L, 1.3 equiv) dropwise at 0 °C. The reaction
D
DOI: 10.1021/acs.oprd.9b00292
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
mixture was stirred at 0−10 °C for 0.5 h. HPLC showed that
the reaction was complete. The reaction mixture was poured
into water (14 L) and acidified with 6 M HCl until pH = 1−3
and extracted with MTBE (7 L × 3). The combined organic
layers were washed with water (7 L) and brine (7 L), and dried
over Na2SO4. The organic layers from three runs were
concentrated to give 7 (16.5 kg, crude) as dark oil. Note:
the crude product containing 7 and 2 was used directly in the
next step without further purification.
1.1 equiv), which had been previously sparged with N2, was
added at a rate such that the internal temperature did not
exceed 0 °C. A 1.0 M potassium tert-butoxide solution in THF
(260 g, 0.291 mol, 0.4 mol %) was added to the reactor and
stirred for 15 min.
To a separate canister, nitro olefin 2 (9.33 kg, 72.2 mol, 1.0
equiv) and CPME (14.6 L) were charged and sparged by N2.
The nitrobutene solution was charged to the reactor at an O2
level of less than ∼10 ppm over 3 h 20 min (internal temp
NMT −10 °C, targeted −15 °C). At reaction completion, the
reactor was warmed to 23 °C.
Spectroscopic data for 7 matched previously measured
values.10
The reaction mixture was distilled to a volume of
approximately 30 L. After cooling to 23 °C, n-heptane (59
L) was added over 1 h. The reactor was stirred for 1 h at 23 °C
after completed n-heptane addition and then cooled to 0 °C
over 4 h. The slurry was then filtered and the filter cake washed
with 19 L of n-heptane. The isolated material was dried at 50
°C to give 4 as a light brown crystalline solid (19.21 kg, 97.5%
w/w, 56.0 mol, 77.6% yield, 98.8pa% purity, 1.5% enantiomer).
To the crude solution of 7 and 2 (3.30 kg, 22.4 mol, 1.0
equiv) in CH2Cl2 (10.0 L) was added TFAA (5.42 kg, 25.8
mol, 3.59 L, 1.15 equiv) at −15 °C, and then, triethylamine
(5.22 kg, 51.6 mol, 7.15 L, 2.30 equiv) was added dropwise,
maintaining a reactor temperature below 10 °C. The mixture
was stirred at −15 to 10 °C for 1 h. At reaction completion, the
mixture was washed with saturated NH4Cl aqueous solution (3
× 10 L), NaHCO3 aqueous solution (5 L× 2), and brine (10
L), and dried over Na2SO4.
1
Data for 4 are as follows. H NMR (700 MHz, CHCl3): δ
7.27 (dd, J = 7.3, 1.8 Hz, 1H), 7.24−7.15 (m, 3H), 5.15 (dd, J
= 6.1, 2.6 Hz, 1H), 4.53 (dd, J = 11.9, 6.1 Hz, 1H), 4.33 (qd, J
= 7.1, 2.1 Hz, 2H), 3.79 (dd, J = 9.4, 7.3 Hz, 1H), 3.34−3.24
(m, 1H), 3.04 (dd, J = 7.3, 2.6 Hz, 1H), 2.39 (s, 3H), 1.35 (t, J
= 7.2 Hz, 3H), 1.06 (s, 9H). 13C NMR (101 MHz, CHCl3): δ
172.4, 135.1, 132.3, 130.4, 128.4, 126.4, 124.9, 91.6, 65.3, 61.8,
61.8, 61.0, 32.6, 27.7, 19.5, 14.1. MS (APCI+): [M + H] calcd
for C18H27N2O4, 335.4; found, 335.1.
The organic layer was concentrated and purified by column
chromatography on silica gel with petroleum ether to give 2
(9.35 kg, 72.4 mol, 64.6% yield, 99a% purity) as a yellow oil.
Spectroscopic data for 2 matched previously measured
values.10
Preparation of Imine (3). To a suitable reactor was added
glycine ethyl ester hydrochloride (15.3 kg, 110 mol, 1.1 equiv),
sodium sulfate (28.4 kg, 200 mol, 2.0 equiv), methylene
chloride (83 L, 1.20 M in o-tolualdehyde), and triethylamine
(11.1 kg, 110 mol, 1.1 equiv). The thick slurry was stirred for
30 min at approximately 20 °C after which time o-
tolualdehyde (12.0 kg, 100 mol, 1.0 equiv) was added. The
reactor was stirred at approximately 20 °C for 2 h, at which
point conversion reached >90% consumption of o-tolualde-
1
Data for exo isomer 11 are as follows. H NMR (400 MHz,
CHCl3): δ 7.77 (dd, J = 7.8, 1.4 Hz, 1H), 7.32−7.25 (m, 1H),
7.21 (td, J = 7.4, 1.5 Hz, 1H), 7.17−7.12 (m, 1H), 5.16 (t, J =
8.5 Hz, 1H), 4.95 (d, J = 8.4 Hz, 1H), 4.27 (m, 1H), 4.20 (m,
2H), 3.13 (t, J = 8.4 Hz, 1H), 2.29 (s, 3H), 2.19 (br s, 1H),
1.34 (t, J = 7.2 Hz, 2H), 1.02 (s, 9H). MS (APCI+): [M + H]
calcd for C18H27N2O4, 335.4; found, 335.1.
1
hyde as determined by H NMR. The reaction mixture was
then filtered and the filter washed with 34 L of methylene
chloride. The product solution in methylene chloride was
washed with water (2 × 60 L) and 23% sodium chloride
solution (71 kg).
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.oprd.9b00292.
■
S
The organic layer was distilled to a volume of approximately
38 L. Then, a constant volume distillation with addition of
CPME (55 L) was carried out until methylene chloride was
present at less than 0.15 wt % to afford the product solution.
The product solution was filtered through an in-line Teflon
filter to a canister and the reactor rinsed with 4.7 L CPME to
give 3 (18.98 kg, 98% purity, 45.2 wt %, 9.25 mol, 93% yield).
Details of HTE catalyst screen, 1H and 13C NMR of new
compounds, HPLC trace analysis of cycloaddition
products, and PXRD of 4 (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: john.hartung@abbvie.com.
ORCID
John Hartung: 0000-0001-5379-0307
Stephen N. Greszler: 0000-0003-1993-2417
Notes
The authors declare the following competing financial
interest(s): All authors are employees of AbbVie. The design,
study conduct, and financial support for this research were
provided by AbbVie. AbbVie participated in the interpretation
of data, review, and approval of the publication.
■
1
Data for 3 are as follows. H NMR (700 MHz, CHCl3): δ
8.60 (s, 1H), 7.92 (dd, J = 7.8, 1.5 Hz, 1H), 7.32 (td, J = 7.5,
1.5 Hz, 2H), 7.27−7.22 (m, 1H), 7.20−7.17 (m, 1H), 4.42 (d,
J = 1.3 Hz, 2H), 4.24 (q, J = 7.1 Hz, 2H), 2.52 (s, 3H), 1.31 (t,
J = 7.2 Hz, 3H). 13C NMR (176 MHz, CHCl3): δ 170.3, 164.1,
138.0, 133.77, 130.9, 130.8, 127.9, 126.3, 62.7, 61.1, 19.4, 14.3.
Preparation of Pyrrolidine (4). Note: the reaction is
sensitive to oxygen, and care must be taken to exclude oxygen
until a passing reaction completion sample is obtained. To a
reactor purged with N2 to an O2 level of approximately 10 ppm
were added [Cu(OTf)]2 benzene complex (106 g, 0.18 mol,
0.25 mol %), ligand 10 (329 g, 0.40 mol, 0.55 mol %), and
CPME (35 L) that had been previously sparged with N2 for 1
h in a separate canister.
ACKNOWLEDGMENTS
■
The resulting darkly colored suspension was stirred for 1 h
at approximately 23 °C, after which time the reactor jacket was
cooled to −25 °C. Imine 3 solution (36 kg, 45.2 wt %, 79 mol,
The authors thank Robert W. Miller, Jerry G. Gaertner, Zhe
Wang, and Brian Kotecki (all of AbbVie) for experimental
assistance.
E
DOI: 10.1021/acs.oprd.9b00292
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
(9) Ballini, R.; Petrini, B. The Nitro to Carbonyl Conversion (Nef
Reaction): New Perspectives for a Classical Transformation. Adv.
Synth. Catal. 2015, 357, 2371−2402.
ABBREVIATIONS
■
MTBE, tert-butyl methyl ether; DME, dimethoxyethane
(10) Denmark, S. E.; Marcin, L. R. Asymmetric Nitroalkene [4 + 2]
Cycloadditions: Enantioselective Synthesis of 3-Substituted and 3,4-
Disubstituted Pyrrolidines. J. Org. Chem. 1995, 60, 3221−3235.
(11) CPME was investigated as the reaction solvent because it would
enable more efficient workup and telescoping. However, this met with
no success because conversion to the desired product was not
observed.
(12) Onset of an exothermic event was observed at 62 °C when
heating isolated 3 during a routine safety evaluation.
(13) The development and application of this methodology will be
disclosed in a future publication.
(14) Strong base was used because it has been shown to suppress
formation of the Michael adduct from the stepwise reaction of the
azomethine ylides with nitro olefins. See ref 8a for details.
(15) The solvent effect was also observed in the initial disclosure of
the catalyst system (ref 8a). The higher dr observed in less polar
solvents is consistent with the authors’ proposal that the endo
transition state is stabilized by an electrostatic interaction.
(16) A 13% drop in yield was observed in reactions run under an
atmosphere of air.
REFERENCES
■
(1) Cutting, G. R. Cystic Fibrosis Genetics: From Molecular
Understanding to Clinical Application. Nat. Rev. Genet. 2015, 16, 45−
56.
(2) Kerem, B.-S.; Rommens, J. M.; Buchanan, J. A.; Markiewicz, D.;
Cox, T. K.; Chakravarti, A.; Buchwald, M.; Tsui, L.-C. Identification
of the Cystic Fibrosis Gene: Genetic Analysis. Science 1989, 245,
1073−1080.
(3) (a) Kartner, N.; Hanrahan, J. W.; Jensen, T. J.; Naismith, A. L.;
Sun, S.; Ackerley, C. A.; Reyes, E. F.; Tsui, L.-C.; Rommens, J. M.;
Bear, C. E.; Riordan, J. R. Expression of the Cystic Fibrosis Gene in
Non-epithelial Invertebrate Cells Produces a Regulated Anion
Conductance. Cell 1991, 64, 681−691. (b) Welsh, M. J.; Smith, A.
E. Molecular Mechanisms of CFTR Chloride Channel Dysfunction in
Cystic Fibrosis. Cell 1993, 73, 1251−1254.
(4) Altenbach, R. J.; Bogdan, A.; Desroy, N.; Gfesser, G. A.;
Greszler, S. N.; Koenig, J. R.; Kym, P. R.; Liu, B.; Scanio, M. J.; Searle,
X.; Wang, X.; Yeung, M. C.; Zhao, G. Substituted Pyrrolidines as
CFTR Modulators. Intl. Pat. Appl. 2018065962, 2018.
(17) Analysis for Cu content was carried out downstream of this step
and was found to be at an acceptable level per ICH Q3D.
(5) Wilson, R. M.; Danishefsky, S. J. Pattern Recognition in
Retrosynthetic Analysis: Snapshots in Total Synthesis. J. Org. Chem.
2007, 72, 4293−4305.
(6) For leading reviews, see: (a) Husinec, S.; Savic, V. Chiral
Catalysts in the Stereoselective Synthesis of Pyrrolidine Derivatives
via Metalloazomethine Ylides. Tetrahedron: Asymmetry 2005, 16,
2047−2061. (b) Pandey, G.; Banerjee, P.; Gadre, S. R. Construction
of Enantiopure Pyrrolidine Ring System via Asymmetric [3 + 2]-
Cycloaddition of Azomethine Ylides. Chem. Rev. 2006, 106, 4484−
4517. (c) Stanley, L. M.; Sibi, M. P. Enantioselective Copper-
Catalyzed 1,3-Dipolar Cycloadditions. Chem. Rev. 2008, 108, 2887−
2902. (d) Hashimoto, T.; Maruoka, K. Recent Advances of Catalytic
Asymmetric 1,3-Dipolar Cycloadditions. Chem. Rev. 2015, 115,
5366−5412.
(7) For previous applications of enantioselective [3 + 2] cyclo-
additions on scale, see: (a) Shu, L.; Gu, C.; Fishlock, D.; Li, Z.
Practical Synthesis of MDM2 Antagonist RG7388. Part 1: A Cu(II)-
Catalyzed Asymmetric [3 + 2] Cycloaddition. Org. Process Res. Dev.
2016, 20, 2050−2056. (b) Rimmler, G.; Alker, A.; Bosco, M.;
Diodone, R.; Fishlock, D.; Hildbrand, S.; Kuhn, B.; Moessner, C.;
Peters, C.; Rege, P. D.; Schantz, M. Practical Synthesis of MDM2
Antagonist RG7388. Part 2: Development of the Cu(I) Catalyzed [3
+ 2] Asymmetric Cycloaddition Process for the Manufacture of
Idasanutlin. Org. Process Res. Dev. 2016, 20, 2057−2066. (c) Shu, L.;
Li, Z.; Gu, C.; Fishlock, D. Synthesis of a Spiroindolinone
Pyrrolidinecarboxamide MDM2 Antagonist. Org. Process Res. Dev.
2013, 17, 247−256. (d) Agbodjan, A. A.; Cooley, B. E.; Copley, R. C.
B.; Corfield, J. A.; Flanagan, R. C.; Glover, B. N.; Guidetti, R.; Haigh,
D.; Howes, P. D.; Jackson, M. M.; Matsuoka, R. T.; Medhurst, K. J.;
Millar, A.; Sharp, M. J.; Slater, M. J.; Toczko, J. F.; Xie, S. Asymmetric
Synthesis of an N-Acylpyrrolidine for Inhibition of HCV Polymerase.
J. Org. Chem. 2008, 73, 3094−3102.
(8) (a) Yan, X.-X.; Peng, Q.; Zhang, Y.; Zhang, K.; Hong, W.; Hou,
X.-L.; Wu, Y.-D. A Highly Enantio- and Diastereoselective Cu-
Catalyzed 1,3-Dipolar Cycloaddition of Azomethine Ylides with
Nitroalkenes. Angew. Chem., Int. Ed. 2006, 45, 1979−1983. (b) Arai,
T.; Mishiro, A.; Yokoyama, N.; Suzuki, K.; Sato, H. Chiral
Bis(imidazolidine)pyridine-Cu(OTf)2: Catalytic Asymmetric Endo-
Selective [3 + 2] Cycloaddition of Imino Esters with Nitroalkenes. J.
Am. Chem. Soc. 2010, 132, 5338−5339. (c) Li, J.-Y.; Kim, H. Y.; Oh,
K. Brucine Diol-Copper-Catalyzed Asymmetric Synthesis of endo-
Pyrrolidines: The Mechanistic Dichotomy of Imino Esters. Org. Lett.
2015, 17, 1288−1291. (d) Conde, E.; Rivilla, I.; Larumbe, A.; Cossío,
F. P. Enantiodivergent Synthesis of Bis-Spiropyrrolidines via
Sequential Interrupted and Completed (3 + 2) Cycloadditions. J.
Org. Chem. 2015, 80, 11755−11767.
F
DOI: 10.1021/acs.oprd.9b00292
Org. Process Res. Dev. XXXX, XXX, XXX−XXX