Convergent asymmetric synthesis of two complex TRPV1 antagonists

Article abstract of DOI:10.1021/op200177b

The convergent scale-up synthesis of two complex TRPV1 antagonists to support exploratory toxicology studies is described. Both compounds contain three chiral centers introduced by asymmetric synthesis with chiral control being critical for the success of the project. Preparation of the key cyclopropyl intermediate utilised an asymmetric cyclopropanation using thermally unstable ethyl diazoacetate. Ellman's auxiliary was used to synthesize the chiral α-methyl benzylamine fragments. This paper highlights some of the key synthetic challenges, processing issues, and safety aspects from the scale-up of this chemistry.

Full text of DOI:10.1021/op200177b

ARTICLE
pubs.acs.org/OPRD
Convergent Asymmetric Synthesis of Two Complex TRPV1
Antagonists
Kenneth J. Butcher, Steven M. Denton, Stuart E. Field, Adam T. Gillmore, Gareth W. Harbottle,
Roger M. Howard, Daniel A. Laity,* Christian J. Ngono, and Benjamin A. Pibworth*
Pfizer Worldwide Research and Development, Sandwich Laboratories, Ramsgate Road, Sandwich CT13 9NJ, United Kingdom
S Supporting Information
b
ABSTRACT: The convergent scale-up synthesis of two complex TRPV1 antagonists to support exploratory toxicology studies is
described. Both compounds contain three chiral centers introduced by asymmetric synthesis with chiral control being critical for the
success of the project. Preparation of the key cyclopropyl intermediate utilised an asymmetric cyclopropanation using thermally
unstable ethyl diazoacetate. Ellman’s auxiliary was used to synthesize the chiral R-methyl benzylamine fragments. This paper
highlights some of the key synthetic challenges, processing issues, and safety aspects from the scale-up of this chemistry.
’ INTRODUCTION
Two routes were investigated for the preparation of the
acetophenone intermediates (10 and 11) needed for the Ellman’s
sulfinamide chemistry (Scheme 2).
The transient vanilloid receptor TRPV1 has been identified as
a potential target for the treatment of chronic pain.1 TRPV1
antagonists inhibit the transmission of inflammatory pain signals
from the periphery to the CNS and so provide the possibility of
developing novel analgesic and anti-inflammatory agents.2 Cap-
sazepine was published as the first competitive TRPV1 antago-
nist by Novartis in 1994,3 and subsequently this has led to further
development in this area.
Previous work has shown that for a series of R-substituted
N-(4-tert-butylbenzyl)-N-[4-(methylsulfonylamino)benzyl]-
thiourea analogues the incorporation of a benzylic R-methyl
group with specific (R)-configuration enhanced the specific bind-
ing to the TRPV1 receptor when compared with capsaicin, a
TRPV1 antagonist.4
Preparation of 10 and 11 was initially trialled using literature
precedent,4,5 involving the installation of the ketone via a Heck
coupling with butyl vinyl ether. This approach, however, was
found to be low yielding without the addition of toxic thallium
salts.6 The SNAr approach, implemented by medicinal chemistry,
proved more efficient. Formation of byproduct 16, formed from
the degradation of DMF,7 could be eliminated by switching
solvent from DMF to DMSO which gave the products in high
yield (75ꢀ82%) with only a 2% impurity, believed to be 17 by
mass spectrometry.
ꢀ
Two potential TRPV1 antagonists (1 and 2) have been
discovered which incorporate both a benzylic R-methyl group
with an (R)-configuration along with a chiral cyclopropyl group.5
The synthesis of these two highly complex compounds was
rapidly enabled within our process development group in order
to support early-phase toxicology studies. Retrosynthetic analysis
of both targets highlighted a suitable amide disconnection
strategy which provided a convergent synthesis allowing rapid
parallel enabling which we describe here (Scheme 1).
The resulting potassium salts of 10 and 11 were taken into the
aqueous layer, and the pH was adjusted with aqueous acid;
however, we were concerned about the potential generation of
hydrogen fluoride (HF) during this acidic work up. We reasoned
that a pH window between pH 5ꢀ8 would protonate the
product, allowing extraction without the risk of HF generation.
Acetic acid was therefore selected as the acid of choice, as
an accidental overcharge would not exceed the designated
pH range.
’ RESULTS AND DISCUSSION
Preparation of r-Methyl Benzylamine Fragments (3 and
4). The routes chosen for the synthesis of the R-methyl benzyl-
amine fragments (3 and 4) were based on literature precedent
(Scheme 2).4,5 Several issues were identified which needed
addressing prior to scale-up; these included improving the
synthesis of the acetophenone fragments, 10 and 11, due to
poor regioselectivity of the Heck chemistry and improving the
diastereoselectivity along with overcoming the safety concerns of
the sodium borohydride (NaBH4) quench for the Ellman’s
sulfinamide chemistry.
The aqueous soluble potassium salts of 10 and 11 enabled us
to remove DMSO and any impurities by extraction with ethyl
Special Issue: Asymmetric Synthesis on Large Scale 2011
Received: June 30, 2011
Published: August 01, 2011
r
2011 American Chemical Society
1192
dx.doi.org/10.1021/op200177b Org. Process Res. Dev. 2011, 15, 1192–1200
|

Organic Process Research & Development
ARTICLE
Scheme 1. Retrosynthetic approach to desired compounds 1 and 2
Scheme 2. Synthesis of r-methyl benzylamine fragments 3 and 4a
a Reagents and conditions: (a) CH3SO2Cl, pyridine, 0 °C, 95%; (b) butyl vinyl ether, Pd(OAc)2, DPPP, TlOAc, DMF, 95 °C, 78%; (c) CH3SO2NH2,
K2CO3, DMSO, 130 °C, 75ꢀ82%; (d) Ti(OEt)4, THF, 66 °C; (e) NaBH4, ꢀ20 °C, 42ꢀ53%; (f) 1.25 M HCl/MeOH, rt, 86%.
acetate (EtOAc) prior to pH adjustment with acetic acid and
isolation. This resulted in the products being isolated in high
purity.8
any exotherm. The equivalents of NaBH4 relative to ketone were
also reduced from 3 equiv to 2 equiv without any detrimental
effect to the reaction.
Installation of the chiral R-methyl benzylamine was achieved
using Ellman’s protocol.9,4 Sulfinyl ketimines 12 and 13 were
formed using (R)-butanesulfinamide and then were diastereose-
lectively reduced in situ with NaBH4 to give the chiral tert-
butanesulfinyl-protected amines (14 and 15). The final R-methyl
benzylamines (3 and 4) were produced by cleaving the sulfinyl
group with 1.25 M hydrochloric acid in methanol (Scheme 2).
The diastereoselectivity of the reduction of 12 and 13 was
found to be solvent dependent, with THF providing the best
selectivity.10 The use of titanium ethoxide (Ti(OEt)4) in the
reaction is also important as it serves not only as a Lewis acid and
water scavenger for the imine condensation but also to enhance
the rates and distereoselectivity of the reduction.11 Initially
NaBH4 was charged at room temperature, resulting in a low
60% de with column purification being required; however, this
ratio was significantly improved to 96% de12 by lowering the
internal temperature to ꢀ20 °C. Further upgrade to >98% de12
was achieved by slurrying the isolated product in methyl tert-
butyl ether (MTBE).
The quench of the Ti(OEt)4 with water resulted in significant
deposits of titanium oxide on the vessel walls and a slow filtration
of the titanium residues. These issues were overcome, enabling
the reaction to be carried out on scale, by reducing the water
quench to only 3 L/kg ketone (10 or 11) and adding Arbocel
filter aid to the vessel prior to filtration to bind the titanium salts.
Following asymmetric reduction, the N-tert-butanesulfinyl
group was hydrolysed using hydrochloric acid in methanol to
give the R-methyl benzylamine (3 and 4). The hydrochloride salt
was isolated in high yield (86ꢀ87%) following a solvent swap
into ethyl acetate.
Preparation of the Cyclopropyl Acid Core (5). The initial
synthesis of cyclopropyl acid core 5 was developed by medicinal
chemistry (Scheme 3). Several issues were highlighted with this
chemistry which included a nonrobust synthesis of styrene 20, a
complex chiral cyclopropanation, and the need for chiral pre-
parative purification.
20 required for the cyclopropanation was synthesised in two
steps from a commercial starting material 18, involving Grignard
initiation using Knochel’s14 protocol followed by an acetone
quench and an acidic dehydration.
The initial Grignard reaction was capricious and produced
yields varying from 40 to 80%. Additionally, process groups in
Merck15 and Pfizer16 highlighted the use of Grignard reagents in
the presence of fluoro-alkyl groups as potentially hazardous;
A major hindrance to the scalability of this chemistry was the
quench of the NaBH4 and Ti(OEt)4. The initial quench of
13
NaBH4 following reduction of the ketimine posed safety
concerns due to vigorous evolution of hydrogen gas. To elim-
inate the formation of hydrogen the reaction was first quenched
with acetone, resulting in a scalable quench and also controlling
1193
dx.doi.org/10.1021/op200177b |Org. Process Res. Dev. 2011, 15, 1192–1200

Organic Process Research & Development
ARTICLE
Scheme 3. Initial discovery route to targets 1 and 2a
a Reagents and conditions: (a) iPrMgCl LiCl, THF, acetone, ꢀ5 °C, 47%; (b) TsOH H2O, toluene, reflux, 94%; (c) ethyldiazoacetate (EDA), toluene,
3
3
80 °C, 80%; (d) EtOH, LiOH, H2O, rt, 81% ; (e) chiral HPLC purification; (f) T3P in EtOAc, MeCN, DIPEA, 50 °C, 63ꢀ77%.
Scheme 4. Synthesis of 5-isopropenyl-2-trifluoromethyl-pyridine 20a
a Reagents and conditions: (a) Pd(dppf)Cl2, CO(g), MeOH, 100 °C, 50Psi, 70%; (b) 1 M MeMgBr in THF, ꢀ10 °C, 100%; c) TsOH H2O, toluene,
3
DeanꢀStark, 94%.
consequently, all chemistry was assessed to ensure these safety
concerns were alleviated. Safety implications over the use of
ethyldiazoacetate (EDA), a thermally unstable reagent,17 in the
cyclopropanation of 20 at elevated temperature required closer
investigation. Finally, alternative purification of the 5 was needed
as the chiral chromatography previously used would not be
amenable for scale-up.
Scheme 5. Outsourced synthesis of 20a
Initial investigations centered on an alternative synthesis of 20
from the same commercially available starting material 18
(Scheme 4).
a Reagents and conditions: (a) nBuLi, toluene, ꢀ60 °C, acetone, 99%;
(b) TsOH H2O, toluene, DeanꢀStark, 75%.
3
Knochel’s chemistry to form tertiary alcohol 19 from 18 using
iPrMgCl LiCl in THF followed by acetone quench produced
Knochel’s reagent (Scheme 5). Ultimately, these alternative
nBuLi conditions were scaled up as accessing 22 involved
carbonylation of 18 which adds an extra processing step to
the route.
3
variable yields and purity with aldol byproduct being the main
concern. Methyl Grignard addition to 22, however, proved more
robust and scalable. Careful consideration of the Grignard
reagent used was required as significant differences in reaction
profile was seen with 1 M MeMgBr in THF being preferable to 3
M MeMgCl in THF. Formation of 20 from 19 was then achieved
by dehydration using toluene sulfonic acid monohydrate
Cyclopropanation of 20 proved the most challenging step due
to poor enantioselectivity (ee ∼60ꢀ65%) and trans/cis selec-
tivity (70/30) (Scheme 6). Nishiyama and co-workers18 high-
light the use of chiral Ru-pybox ligand systems which afford good
enantioselectivity for a range of substrates. However, most
literature precedent for cyclopropanations of this nature are
performed on substrates without a pyridine functionality present
and no other substitution on the styrene. Attempts to prepare
trisubstituted cyclopropanes generally result in either poor ee or
poor trans/cis ratio,19 and there is only one report of an
alkenylpyridine substrate in a noncatalysed racemic system.20
We wondered if the inferior trans/cis selectivity and enantios-
electivity of 21 could be explained by competitive co-ordination
(TsOH H2O) in toluene under DeanꢀStark conditions, with
3
variable yields obtained on the basis of its volatility. A salt screen
was performed by our technology group as an alternative method
of isolation of 20 but was unsuccessful. 20 could be purified by
column chromatography; however, purification was also
achieved by vacuum distillation with an excellent yield obtained.
Preparation of 20 was outsourced to an external vendor who
demonstrated that the original organometallic addition of 18 to
n
acetone could be reliably carried out with BuLi in place of
1194
dx.doi.org/10.1021/op200177b |Org. Process Res. Dev. 2011, 15, 1192–1200

Organic Process Research & Development
ARTICLE
or ligation of the pyridine to the chiral ruthenium catalyst and
sought an improved catalyst and ligand system. A high-through-
put reaction screen21 including copper-, cobalt-, and palladium-
based catalysts with several oxazoline-substituted ligands failed to
identify any improvement from the original conditions (Table 1).
On the basis of literature precedent of thermal instability of
ethyl diazoacetate (EDA),17 investigations of the cyclopropana-
tion reaction were considered necessary as it was performed at
80 °C. Slow addition of EDA to the reaction was key in providing
a controlled evolution of nitrogen gas, suitable reaction conver-
sion, and the minimising of byproduct associated with EDA
dimerization. tert-Butyl diazoacetate was attempted as an alter-
native diazo source but offered no advantage in either enantios-
electivity or trans/cis selectivity. Achiral cobalt conditions22 were
trialled with the aim of improving the trans/cis selectivity, but
these proved unsuccessful. The original catalyst was prepared by
mixing dichloro(p-cymene)ruthenium(II) dimer with 2,6-bis[-
(4R)-(+)-isopropyl-2-oxazolin-2-yl]pyridine ligand in dichloro-
methane and then bubbling ethylene through the solution before
isolation. It was found that the intermediate cymene catalyst was
just as effective in the reaction and could be telescoped directly
into the cyclopropanation, negating any stability issues seen with
catalyst isolation23 with similar chiral results being obtained
(Scheme 6).
Hydrolysis of 21 to give 5 was achieved using lithium hydroxide
(LiOH) in ethanol (EtOH). Careful examination of the reaction
indicated a subtle rate difference which upgraded the trans/cis
selectivity further to 85:15.
The complete removal of chiral chromatography was desirable
for the purification of 5, consequently, a series of chiral bases
were screened21 to upgrade the trans/cis selectivity and enantios-
electivity. Unfortunately, no salt provided the desired upgrade in
a single step, quinine was identified as a method to upgrade the
trans/cis selectivity to >95:5 with no upgrade of the enantios-
electivity. Following a salt-break process using 2 M HCl(aq)
/
MTBE, chiral upgrade was achieved using (S)-1,2,3,4-tetrahy-
dro-1-naphthylamine to provide >93% ee24 on a first crystal-
lization. This could be further recrystallised to upgrade the ee or
taken into the amide coupling directly with the unwanted
diastereoisomer being purged via traditional crystallisation meth-
ods (Scheme 7).
A biocatalytic alternative for purifying the crude cyclopropyl
acid 5, avoiding the use of chiral chromatography, was investigated.
Scheme 7. Chiral purification of cyclopropyl acid 5a
Purification of Crude Cyclopropyl Intermediate (5). Fol-
lowing the cyclopropanation it was considered necessary to
upgrade the trans/cis ratio of 21 from 70/30. This was achieved
via silica plug purification to give a trans/cis ratio of 80/20.
Scheme 6. Modified cyclopropanation of 20a
a Reagents and conditions: (a) quinine, MTBE, 55 °C, 75%; (b) 2 M
HCl(aq), MTBE 25 °C, 100%; (c) (S)-1,2,3,4-tetrahydro-1-naphthyla-
mine, 2-propanol, 50 °C, 77%).
a Reagents and conditions: (a) toluene, EDA, 80 °C, 80%.
Table 1. Selected results from cyclopropanation screena,21
HPLC analysis (area %)
metal/precatalyst
ligand
20
cis-21
trans-21
Cl2(p-cymene)2ruthenium(II) dimer
ruthenium(II) catalyst b
2,6-bis[(4S)-(ꢀ)-isopropyl-2-oxazolin-2yl]pyridine
6
11
5
10
1
30
20
16
13
10
1
N/A
carbonylhydridotris(triphenylphosphine)rhodium(I)
(+)-N-methylephedrine
(R,S)-tert-ButylꢀJosiphos
(R,S)-tert-butyl-Josiphos
BINAP
9
Cu(OTf)2
CoCl2
8
7
1
1
Pd2(dba)3
6
1
a NOTE: All the reactions were carried out in 2.5 mL of HPLC crimp-cap vials on 10-mg scale. The reaction mixtures were analysed by HPLC using an
SB-C18 column, a gradient of 95:5 to 5:95%, 0.1% TFA in water/MeCN, over 3.75 min with a flow rate of 3 mL/min and using 0.04 μL injections.
b Chloro[(1S)-(ꢀ)-5,50-dichloro-6,60-dimethoxy-2,20-bis(diphenylphosphino)-1,10-biphenyl](p-cymene)ruthenium(II)chloride DCM adduct.
1195
dx.doi.org/10.1021/op200177b |Org. Process Res. Dev. 2011, 15, 1192–1200

Organic Process Research & Development
ARTICLE
Scheme 8. Biocatalytic hydrolysis of crude cyclopropane mixturea
a Reagents and conditions: (a) alcalase (Bacillus licheniformis protease), pH 9.0 carbonate buffer, 30 °C, 14 days.
Table 2. Screen of commercial enzymes for hydrolysis of 21a
enzyme description
vendor
conversion to 5 (chiral HPLC)/%
chiral purity of 5 (% ee.)
control (no enzyme)
n/a
0
n/a
>99 (2S)
>99 (2S)
>99 (2S)
unknown
>99 (2S)
>99 (2S)
54 (2R)
Bacillus lentus protease Esperase 8.0 L
acid protease II
Novozymes
Amano
1.1
2.1
4.3
4.2
5.8
6.5
5.1
Bacillus licheniformis protease Alcalase 2.5 L
PROTIN SD-AY 10
Novozymes
Amano
Protex 6 L (A01424)
Geneco
Protease P “Amano” 6
Amano
Rhizomucor miehei lipase
a NOTE: All reactions were carried out on 10-mg scale.
Novozymes
Scheme 9. Convergent coupling to synthesize final product 1
fragments (3 and 4) and cyclopropyl acid core (5) the two
fragments were coupled to complete the convergent synthesis
(Scheme 9). It was found that high chiral purity of the two
fragments prior to coupling was critical, as all attempts to purge
high levels of the diastereoisomers in the product were
unsuccessful.26,27
and 2a
Alternative conditions to the initial discovery conditions
involving 1-propylphosphonic acid cyclic anhydride (T3P) were
evaluated; these included traditional 1,10-carbonyldiimidazole
(CDI) activation, thionyl chloride activation via acid chloride
formation, and boronic acid coupling methodologies.28 None of
these conditions provided a superior reaction profile, and in
some cases there were indications of epimerization; thus, T3P in
THF was used. Following coupling both products (1 and 2)
required a reslurry in MTBE, and 2 was recrystallised from
acetonitrile to reduce the level of diastereoisomers to 1%,
resulting in an HPLC achiral purity >95%.
a Reagents and conditions: (a) 50% T3P in EtOAc, DIPEA, 50 °C,
63ꢀ77%.
This investigation centred around the use of enzymes to selec-
tively give the required diastereoisomer via enzymatic hydrolysis
of 21 (65% ee trans/cis 70:30) (Scheme 8).
A set of commercially available enzymes were screened for the
selective hydrolysis of the desired trans-2S-stereoisomer using
the protocol described by Tao and co-workers25 and Novozymes’
protease ‘Alcalase’ from Bacillus licheniformis was selected for
further development based on cost and previous experience
(Table 2).
This enzyme displayed >99% selectivity for the desired
trans-(2S) isomer. Unfortunately, despite efforts to optimize
the reaction conditions, the biocatalyst displayed very low
activity, and therefore long reaction times of multiple days and
multiple 1000 wt % charges of the crude liquid enzyme prepara-
tion were required to reach high conversion (Scheme 8). A yield
comparable to the chemical hydrolysis and stereoselective crys-
tallisation route could be obtained via this method (59% of the
available stereoisomer), although for scale-up we decided to scale
up the previous conditions due to the confidence we had in the
chemistry.
’ SUMMARY
In summary, we have developed a safe and scalable process to
convergently produce two complex chiral compounds (1 and 2)
in high purity in support of toxicology studies. This has involved
delivering two chirally pure R-methyl benzylamine fragments
(3 and 4) by utilizing Ellman’s sulfinamide chemistry which was
successfully scaled. The safe scale-up of a complex cyclopropana-
tion step involving thermally unstable ethyl diazoacetate was also
achieved. This was then followed by a stereoselective purification
of 5 via traditional crystallisation techniques (without the need
for chiral preparative chromatography) and coupling of the
convergent fragments.
’ EXPERIMENTAL SECTION
Convergent Coupling to Give Final Products (1 and 2).
Following parallel synthesis of the R-methyl benzylamine
General. All starting materials are available commercially or
described in the literature. Chromatography was carried out
1196
dx.doi.org/10.1021/op200177b |Org. Process Res. Dev. 2011, 15, 1192–1200

Organic Process Research & Development
ARTICLE
using Merck silica gel 60 (9385) using standard glassware. Thin
layer chromatography (TLC) was carried out on Biotage kp-Sil
flash TLC plates. H NMR spectra were recorded on a Varian
charged to the vessel. The reaction mixture was stirred at ꢀ20 °C
(internal temperature) for approximately 16 h and then warmed
to ꢀ10 °C. Acetone (76.3 mL, 1.04 mol) was added slowly
followed by methanol (26.8 mL, 662 mmol) and the reaction
allowed to warm to room temperature and stirred for 2 h. To the
reaction mixture was added water (100 mL) followed by Arbocel
filter aid (32 g, 0.80 g/g) and stirred for 1 h before filtering the
titanium salts. The cake was washed with 2MeTHF (3 ꢁ
100 mL) and the filtrate concentrated to remove THF and
methanol. 2MeTHF (300 mL) was added followed by water
(300 mL) and aqueous hydrochloric acid (1 M, 60 mL), and the
organic layer was retained. The aqueous was extracted with
2MeTHF (300 mL), and then the organic layers were combined.
The organic layer was washed with water (300 mL) and brine
(300 mL), dried over MgSO4, and evaporated in vacuo. The
residue was purified by slurrying in MTBE (120 mL) at room
temperature for 2 h to give the title compound 14 as an off-white
solid (30.9 g, 91.8 mmol, 53%, 98.2% de12). 1H NMR (400 MHz,
DMSO-d6) δ 1.19 (s, 9H), 1.45ꢀ1.48 (d, 3H), 2.97 (s, 3H),
4.29ꢀ4.37 (m, 1H), 5.62ꢀ5.66 (d, 1H), 7.14ꢀ7.18 (dd, 1H),
7.25ꢀ7.31 (m, 2H), 9.45 (s, 1H); 13C NMR (100 MHz, DMSO-
d6) δ 1.57.4, 155.0, 154.9, 154.8, 127.3, 123.8, 123.7, 123.3,
123.3, 114.9, 114.7, 55.8, 54.6, 40.9, 24.5, 23.1; LC/MS m/z
(ES+) 337 [MH]+.
N-(4-{(1R)-1-[(tert-Butylsulfinyl)amino]ethyl}-2-chlorophe-
nyl)methanesulfonamide (15). To a solution of 11 (105 g, 424
mmol) in THF (840 mL, 8 mL/g), was added Ti(OEt)4
(146 mL, 699 mmol) followed by (R)-tert-butylsulfinamide
(54.4 g, 445 mmol) and the reaction mixture heated to reflux
with stirring. After heating at reflux for 16 h the reaction was
cooled to ꢀ30 °C, and NaBH4 (45.6 g, 1.27 mol) was slowly
charged to the vessel. The reaction mixture was stirred at ꢀ20 °C
(internal temperature) for approximately 5 h and then warmed to
ꢀ10 °C for 12 h. Acetone (280 mL, 3.82 mol) was added slowly
followed by methanol (105 mL, 2.59 mol) and the reaction
allowed to warm to room temperature and stirred for 2 h. To the
reaction mixture was slowly added aqueous hydrochloric acid
(1 M, 840 mL) followed by Celite filter aid (150 g, 1.0 g/g) and
stirred for 1 h before filtering the titanium salts. The cake was
washed with THF and the filtrate concentrated to remove THF
and methanol. 2MeTHF (500 mL) was added, and the organic
layer was retained. The aqueous layer was extracted with
2MeTHF (500 mL), and then the organic layers were combined.
The organic layer was washed with water (300 mL) and brine
(300 mL), dried over MgSO4, and evaporated in vacuo. The
residue was purified by slurrying in MTBE (480 mL) at room
temperature for 3 h to give the title compound 15 as an off-white
solid. (62.8 g, 178 mmol, 42%, 99.6% de12). 1H NMR (400 MHz,
DMSO-d6) δ 1.08 (s, 9H), 1.32ꢀ1.36 (d, 3H), 2.99 (s, 3H),
4.29ꢀ4.37 (m, 1H), 5.66ꢀ5.70 (d, 1H), 7.29ꢀ7.36 (m, 2H),
7.52ꢀ7.53 (m, 1H) 9.34 (s, 1H); 13C NMR (100 MHz, DMSO-
d6) δ 145.7, 133.1, 129.3, 128.5, 128.1, 126.6, 55.8, 54.6, 41.5,
24.5, 23.1; LC/MS m/z (ES+) 353 [MH]+
1
300 MHz spectrometer or Jeol ECS 400 MHz spectrometer. 13C
NMR spectra were recorded on a Jeol ECS spectrometer at 100.5
MHz. HPLC analyses were performed using a reverse-phase
technique. LC/MS analysis was performed using the following
system: Hewlett-Packard 1100 with SB C18 3.0 mm ꢁ 50 mm,
1.8 μm particles; mobile phase consisting of solvent A, 0.05%
TFA in water, solvent B, 0.05% TFA in acetonitrile; 0 min = 5%
solvent B; 3.5 min = 100% solvent B; 4.5 min = 100% solvent B;
4.6 min = 5% solvent B; run time 5 min; column temperature
50 °C; λ = 225 nm; with Waters Micromass ZQ 2000/4000 mass
detector. Combustion analyses were performed by Warwick
Analytical Service, University of Warwick Science Park, The
Venture Centre, Sir William Lyons Road, Coventry CV4 7EZ, U.K.
N-(4-Acetyl-2-fluorophenyl)methanesulfonamide (10). To a
solution of 1-(3,4-difluorophenyl)ethanone (125.0 g, 0.80 mol)
in DMSO (625 mL, 5 mL/g) was added potassium carbonate
(221.3 g, 1.60 mol), followed by methane sulfonamide (152.3 g,
1.60 mol). The reaction mixture was heated to 120 °C with
overhead stirring for 16 h and then cooled to room temperature.
The cooled mixture was poured into water (1.6 L) and washed
with ethyl acetate (4 ꢁ 625 mL). The pH of the aqueous layer
was adjusted to pH 6 by controlled addition of 1:1 acetic acid/
water and the resulting solid collected by filtration and dried to
give the title compound 10 as a brown solid (139.7 g, 0.60 mol,
1
75%). H NMR (400 MHz, DMSO-d6) δ 2.51 (s, 3H), 3.10
(s, 3H), 7.50ꢀ7.55 (dd, 1H), 7.72ꢀ7.77 (m, 2H), 10.06 (s, 1H);
13C NMR (100 MHz, DMSO-d6) δ 196.5, 155.2, 152.7, 134.4,
134.4, 131.0, 130.9, 125.7, 125.7, 123.4, 116.2, 115.9, 41.22,
27.11; Anal. Calcd for C9H10FNO3S: C, 46.75; H, 4.36; N, 6.06;
F, 8.22; S, 13.87. Found: C, 46.60; H, 4.31; N, 5.94; F, 8.29;
S, 14.02.
N-(4-Acetyl-2-chlorophenyl)methanesulfonamide (11). To
a solution of 1-(3-chloro-4-fluorophenyl)ethanone (93.2 g, 0.54
mol) in DMSO (466 mL, 5 mL/g) was added potassium
carbonate (149.2 g, 1.1 mol), followed by methane sulfonamide
(102.7 g, 1.1 mol). The reaction mixture was heated to 120 °C
with overhead stirring for 16 h and then cooled to room
temperature. Water (1.2 L) and ethyl acetate (475 mL) were
added to the vessel, and the aqueous layer was retained and
washed with ethyl acetate (3 ꢁ 400 mL). To the aqueous layer
was added 2MeTHF (750 mL) followed by controlled addition
of 1:1 acetic acid/2MeTHF (200 mL). The organic layer was
retained and the aqueous layer extracted with 2MeTHF
(800 mL). The organic layers were combined, dried over
MgSO4, and evaporated in vacuo to give the title compound 11
as a light-brown solid (110.0 g, 0.44 mol, 82%). 1H NMR (400
MHz, DMSO-d6) δ 2.53 (s, 3H), 3.12 (s, 3H), 7.57ꢀ7.59 (d,
1H), 7.85ꢀ7.88 (dd, 1H), 7.98ꢀ7.99 (d, 1H), 9.66 (s, 1H); 13C
NMR (100 MHz, DMSO-d6) δ 196.5, 139.2, 134.9, 130.3, 128.4,
127.1, 124.7, 41.7, 27.2; Anal. Calcd for C9H10ClNO3S: C, 43.64;
H, 4.07; N, 5.65; Cl, 14.31; S, 12.94. Found: C, 44.44; H, 3.96; N,
5.31; Cl, 14.56; S, 12.25.
N-{4-[(1R)-1-Aminoethyl]-2-fluorophenyl}methanesulfona-
mide Hydrochloride (3). To a vessel containing 14 (100 g, 297
mmol) was added hydrochloric acid (1.25 M)/ methanol
(713 mL, 7 mL, g) and the reaction mixture was stirred at room
temperature for 16 h. The solution was evaporated in vacuo to give
a yellow powder. The solid was purified by slurrying in EtOAc
(400 mL) for 2 h at room temperature and the resulting
solid collected by filtration and dried to give the title compound
N-(4-{(1R)-1-[(tert-Butylsulfinyl)amino]ethyl}-2-fluorophe-
nyl)methanesulfonamide (14). To a solution of 10 (40.0 g, 173
mmol) in THF (320 mL, 8 mL/g), was added Ti(OEt)4
(59.7 mL, 285 mmol) followed by (R)-tert-butylsulfinamide
(22.2 g, 182 mmol) and the reaction mixture heated to reflux
with stirring. After heating at reflux for 16 h the reaction was
cooled to ꢀ30 °C, and NaBH4 (12.4 g, 346 mmol) was slowly
1
3 as an off-white solid (68.7 g, 256 mmol, 86%). H NMR
1197
dx.doi.org/10.1021/op200177b |Org. Process Res. Dev. 2011, 15, 1192–1200

Organic Process Research & Development
ARTICLE
(400 MHz, MeOD) δ 1.60ꢀ1.62 (d, 3H), 3.00 (s, 3H), 4.43ꢀ4.49
(q, 1H), 7.25ꢀ7.28 (m, 1H), 7.31ꢀ7.36 (dd, 1H), 7.54ꢀ7.59
(t, 1H); 13C NMR (100 MHz, MeOD) δ 158.3, 153.8, 136.8,
136.7, 126.3, 126.1, 125.4, 123.0, 122.9, 114.4, 114.2, 50.1,
39.0, 19.1;
ꢀ25 °C over 45 min, and aqueous HCl (2 M, 863 mL) was then
added. The organic layer was retained, and the aqueous layer was
extracted with diethyl ether (1.9 L). The organic layers were
combined and washed with brine (2ꢁ 1.9 L), dried over MgSO4,
and evaporated in vacuo to yield the title compound 19 quantita-
1
N-{4-[(1R)-1-Aminoethyl]-2-chlorophenyl}methanesulfona-
mide Hydrochloride (4). To a vessel containing 15 (60 g, 170
mmol) was added hydrochloric acid (1.25 M)/ methanol
(600 mL, 10 mL/g) and the reaction mixture was stirred at
room temperature for 3 h. The solution was evaporated in vacuo
to give an off-white powder. The solid was slurried in EtOAc
(240 mL) for 0.5 h at room temperature and then cooled to
0ꢀ5 °C for 0.5 h before being collected by filtration. The
compound was dried to give the title compound 4 as an off-
tively as an oil (51 g, 248 mmol, 99%). H NMR (400 MHz,
CDCl3) 1.60 (s, 6H), 7.60 (dd, 1H), 7.95ꢀ8.00 (dd, 1H) and
8.75ꢀ8.85 (s, 1H); 19F NMR (400 MHz, CDCl3) ꢀ68 (s, 3F);
13C NMR (400 MHz, DMSO-d6) δ 147.5, 146.5, 146.2, 133.6,
123.5, 119.9, 71.8, 32.0; LC/MS m/z (ES+) 206 [MH]+
5-Isopropenyl-2-trifluoromethyl-pyridine (20). To a solution
of 19 (270 g, 1.31 mol) in toluene (2.3 L, 9 mL/g) was added
TsOH H2O (95.0 g, 499 mmol), and the resulting mixture was
3
heated to reflux under DeanꢀStark conditions for 6 h. The
reaction mixture was cooled to room temperature, and then sat.
sodium carbonate (2.5 L) was charged to the vessel and the
resulting mixture stirred for 20 min. The phases were separated,
and the resulting organic layer was washed with brine (2 ꢁ 2.5 L),
dried over MgSO4, and evaporated in vacuo with temperature
<45 °C (to minimise volatile product losses). Purification was by
high vacuum distillation at 65 °C (bp 40 °C @ 0.17 mmHg) to
yield the title compound 20 as an oil (131 g, 987 mmol, 75%). 1H
NMR (400 MHz, CDCl3) 2.18 (s, 3H), 5.29 (s, 1H), 5.50
(s, 1H), 7.62ꢀ7.70 (dd, 1H), 7.85ꢀ7.90 (dd, 1H) and 8.81 (s,
1H); 13C NMR (400 MHz, CDCl3) 147.5, 139.9, 134.0, 129.0,
128.1, 125.2, 120.2, 116.1, 21.8; LC/MS m/z (ES+) 188 [MH]+
2-Methyl-2-(6-trifluoromethyl-pyridin-3-yl)cyclopropanecar-
boxylic Acid Ethyl Ester (21). Preforming the active catalyst: to a
100-mL round-bottomed flask was charged DCM (40 mL),
followed by Ru(cymene)2Cl2 (4.91 g, 8.0 mmol) and 2,6-bis[-
(4R)-(+)-isopropyl-2-oxazolin-2-yl] pyridine (4.83 g, 16 mmol),
and the mixture was stirred at room temperature for 30 min,
resulting in a deep-purple reaction mixture.
1
white solid (42.2 g, 148 mmol, 86%). H NMR (400 MHz,
MeOD) δ 1.60ꢀ1.63 (d, 3H), 3.01 (s, 3H), 4.43ꢀ4.49 (q, 1H),
7.40ꢀ7.43 (m, 1H), 7.60ꢀ7.63 (m, 2H); 13C NMR (100 MHz,
MeOD) δ 137.0, 135.1, 128.3, 127.9, 126.2, 126.0, 50.0, 39.6,
19.1; Anal. Calcd for C9H13ClN2O2S HCl: C, 37.90; H, 4.95; N,
3
9.82; Cl, 24.86; S, 11.24. Found: C, 37.73; H, 4.92; N, 9.71; Cl,
24.86; S, 11.23.
Methyl 6-(Trifluoromethyl)pyridine-3-carboxylate (22). To a
solution of 5-bromo-2-(trifluoromethyl) pyridine (0.5 g, 2.21
mmol) in methanol (15 mL, 30 mL/g) was added Pd(dppf)Cl2
(81.2 mg, 0.11 mmol) followed by DIPEA (1.16 mL, 6.64
mmol). The resulting mixture was carbonylated under CO(g)
atmosphere at 100 °C and 50 Psi and left for 12 h. Upon reaction
completion the mixture was filtered through Celite filter aid, and
the solvent was removed via distillation in vacuo. The solid was
redissolved in EtOAc (20 mL). The organic layer was washed
with aqueous sodium carbonate (1 M, 2 ꢁ 20 mL) followed by
saturated aqueous brine (2 ꢁ 20 mL), dried over Na2SO4, and
evaporated in vacuo. The residue was redissolved in minimal
DCM and purified via silica plug using heptane/EtOAc (75:25).
The product fractions were combined and evaporated in vacuo to
give the title compound 22 as a solid (320 mg, 1.55 mmol, 70%).
1H NMR (300 MHz, CDCl3) 4.00 (s, 3H), 7.78ꢀ7.81 (dd, 1H),
8.47ꢀ8.50 (dd, 1H), 9.31 (s, 1H); LC/MS m/z (ES+) 206 [MH]+
2-(6-Trifluoromethyl-pyridin-3-yl)-propan-2-ol (19): from
Grignard Addition to 22. MeMgBr (1 M in THF, 731.2 mL,
731 mmol) was charged to a vessel and cooled to ꢀ10 °C. A
preformed solution of 22 (50 g, 243 mmol) in anhydrous THF
(500 mL, 10 mL/g) was added over 2 h while maintaining the
internal temperature at ꢀ10 °C. Following addition, the reaction
mixture was warmed to 10 °C and analysed by HPLC (>97%
conversion to desired product). The mixture was recooled to
0 °C, and aqueous NH4Cl (500 mL) was added. The product was
extracted with 2MeTHF (500 mL), and the lower aqueous layer
was re-extracted with 2MeTHF (250 mL). The organic layers
were combined and washed with brine (250 mL), dried over
MgSO4, and evaporated in vacuo to yield the title compound 19
quantitatively as an oil (51 g, 248 mmol, >100%).
Toasolutionof20 (100g, 534mmol) intoluene(1L, 10 mL/g)
was added the preformed catalyst solution, and the resulting
mixture was heated to 80 °C over 30 min. A preformed solution
of ethyldiazoacetate (EDA) (90.7 mL, 854 mmol) in toluene
(400 mL) was added to the reaction mixture over a period of 5 h
using a peristaltic pump to provide a constant feed of the EDA
solution. After 5 h HPLC analysis indicated the reaction was
complete with 77.5% trans product and 21.7% cis product. The
reaction was cooled to room temperature and then evaporated in
vacuo to give a purple residue. Purification by silica plug using
heptane/EtOAc (80:20) gave the title compound 21 as an oil
1
(132.25 g, 484 mmol, 81%, trans/cis ratio 80:20, 65% ee). H
NMR (400 MHz, CDCl3) 1.25ꢀ1.30 (t, 3H), 1.41ꢀ1.46
(m, 1H), 1.52ꢀ1.56 (m, 4H), 1.95ꢀ2.0 (m, 1H), 4.15ꢀ4.25
(m, 2H), 7.55ꢀ7.61 (dd, 1H), 7.71ꢀ7.75 (dd, 1H), 8.65 (s, 1H);
13C NMR (400 MHz, CDCl3) 171.0, 149.5, 146.0, 144.0, 136.0,
123.0, 120.0, 61.0, 28.0, 27.5, 20.5, 19.5 and 14.0; LC/MS m/z
(ES+) 274 [MH]+
2-Methyl-2-(6-trifluoromethyl-pyridin-3-yl)cyclopropanecar-
boxylic Acid (5). To a solution of LiOH (17.24 g, 410 mmol) in
water (1125 mL, 10 mL/g) was added a preformed solution of 21
(112.25 g, 410 mmol) in EtOH (560 mL, 5 mL/g). The addition
was controlled to maintain the temperature <30 °C. HPLC
analysis after 2 h showed full conversion of the desired trans-ester
21 to the corresponding acid 5. To the reaction mixture was
added aqueous K2CO3 (1 M, 140 mL) followed by addition of
MTBE (800 mL), and the organic layer was retained. The lower
aqueous layer was re-extracted with MTBE (600 mL). The
organic layers were combined and washed with saturated brine
2-(6-Trifluoromethyl-pyridin-3-yl)-propan-2-ol (19): Direct
Addition of 18 to Acetone. To a solution of toluene (1.75 L,
6 mL/g) at ꢀ75 °C was added nBuLi in hexane (1.6 M, 912 mL,
1.46 mol) over 5 min. A preformed solution of 18 (300 g, 1.33
mol) in anhydrous toluene (750 mL, 2.5 mL/g) was added over
15 min while maintaining the internal temperature below
ꢀ60 °C. Following addition the reaction mixture was left for
30 min and then a solution of acetone (195 mL, 2.65 mol) in
THF (600 mL) was added over 30 min while maintaining the
internal temperature below ꢀ55 °C. The reaction was warmed to
1198
dx.doi.org/10.1021/op200177b |Org. Process Res. Dev. 2011, 15, 1192–1200

Organic Process Research & Development
ARTICLE
(300 mL), dried over MgSO4, and evaporated in vacuo to yield
the title compound 5 as an orange oil (81.40 g, 332 mmol, 81%).
1H NMR (400 MHz, CDCl3) 1.54ꢀ1.59 (m, 2H), 1.60ꢀ1.65
(m, 3H), 1.99ꢀ2.05 (q, 1H), 7.60ꢀ7.65 (dd, 1H), 7.76ꢀ7.80
(dd, 1H), 8.70 (s, 1H), 11.35ꢀ11.80 (broad singlet, 1H); 13C
NMR (400 MHz, CDCl3) 176.5, 149.5, 146.0, 144.2, 136.0,
123.5, 120.5, 29.0, 27.0, 21.0, 19.5; LC/MS m/z (ES+) 246
[MH]+ Anal. Calcd for C11H10F3NO2: C, 53.88; H, 4.11; F,
23.24; N, 5.71; O, 13.05. Found: C, 53.85; H, 4.07; N, 5.71
2-Methyl-2-(6-trifluoromethyl-pyridin-3-yl)cyclopropane-
carboxylic Acid Quinine Salt (5). To a solution of 5 (102.47 g,
468 mmol) in MTBE (614 mL, 6 mL/g) was added quinine
(135.58 g, 417 mmol). To the slurry was added further
MTBE (500 mL, 5 mL/g), and the slurry was heated to 55 °C
for a period of 2 h. The slurry was cooled to room tempera-
ture and filtered with the cake being washed with MTBE and
pulled dry under vacuum. Isolated solid was dried in vacuum
oven at 50 °C to yield title compound 5 as the quinine salt
(202.38 g, 355 mmol, 76%). 1H NMR (400 MHz, CDCl3) 1.15
(s, 3H), 1.15ꢀ1.30 (m, 2H), 1.45ꢀ1.49 (t, 1H), 1.59 (s, 3H),
1.65ꢀ1.72 (m, 1H), 1.92ꢀ2.00 (m, 3H), 2.54 (broad singlet,
1H) 2.91ꢀ3.03 (m, 1H), 3.19 (s, 1H), 3.22ꢀ3.30 (t, 1H),
3.35ꢀ3.43 (t, 1H), 3.73 (s, 3H), 4.15ꢀ4.26 (m, 1H),
4.93ꢀ5.0 (m, 2H), 5.48ꢀ5.58 (m, 1H), 6.20 (s, 1H), 6.92 (s,
1H), 7.13ꢀ7.16 (dd, 1H), 7.54ꢀ7.60 (m, 2H), 7.73ꢀ7.79 (m,
2H), 8.62 (s, 1H), 8.71 (s, 1H).
2-Methyl-2-(6-trifluoromethyl-pyridin-3-yl)cyclopropanecar-
boxylic Acid (S)-1,2,3,4-tetrahydronaphthylamine salt (5). To a
solution of 5 (83.92 g, 342 mmol) in 2-propanol (840 mL,
10 mL/g) was added a preformed solution of (S)-1,2,3,4-tetra-
hydronaphthylamine (50.38 g, 342 mmol) in 2-propanol
(200 mL, 4 mL/g). The reaction was heated to 50 °C for 3 h
and was cooled to room temperature. The slurry was cooled
using an ice bath, isolated by filtration, washed with 3 mL/g
2-propanol, and dried in a vacuum oven at 50 °C to yield title
compound 5 as the (S)-1,2,3,4-tetrahydro-1-naphthylamine salt
(104.54 g, 266 mmol, 77%). 1H NMR (400 MHz, CDCl3) 1.19ꢀ
1.26 (m, 1H), 1.28ꢀ1.32 (m, 1H), 1.48 (s, 3H), 1.71ꢀ1.80 (m,
2H), 1.81ꢀ1.98 (m, 2H), 2.04ꢀ2.13 (m, 1H), 2.66ꢀ2.80 (m,
2H), 4.16ꢀ4.21 (t, 1H), 7.00ꢀ7.15 (m, 6H), 7.40ꢀ7.45 (d, 1H),
7.54ꢀ7.57 (d, 1H), 7.62ꢀ7.67 (dd, 1H), 8.57 (s, 1H).
(dd, 1H), 8.60ꢀ8.62 (m, 1H); 13C NMR (100 MHz, DMSO-d6)
δ 168.5, 157.6, 155.2, 149.3, 146.0, 145.3, 145.3, 145.1, 144.8,
144.5, 144.1, 136.6, 127.7, 126.4, 123.7, 123.7, 123.6, 122.8,
122.8, 121.0, 120.7, 118.3, 114.3, 114.1, 48.2, 30.3, 27.4, 25.8,
22.9, 19.6, 18.1; LC/MS m/z (ES+) 460 [MH]+; Anal. Calcd for
C20H21F4N3O3S: C, 52.28; H, 4.61; N, 9.15; F, 16.54; S, 6.98.
Found: C, 52.28; H, 4.76; N, 8.86; F, 16.86; S, 6.95.
(1S,2S)-N-[(1R)-1-{3-Chloro-4-[(methylsulfonyl)amino]phenyl}-
ethyl]-2-methyl-2-[6-(trifluoromethyl)pyridin-3-yl]cyclopropa-
necarboxamide (2). To a solution of 5 (63.1 g, 257 mmol) in
THF (734 mL, 10 mL/g) was added 4 (73.4 g, 257 mmol)
followed by 1-propylphosphonic acid cyclic anhydride (T3P)
(232 mL, 386 mmol). Diisopropylethylamine (224 mL, 1.29
mol) was added and the reaction mixture heated to 60 °C for 2 h.
The reaction was cooled to room temperature and to the vessel
was added water (750 mL) followed by ethyl acetate (750 mL).
The organic layer was retained and washed with water (750 mL),
sat. sodium carbonate (750 mL), sat. citric acid (750 mL), and
water (750 mL), dried over MgSO4, and evaporated in vacuo to
give a foam. The residue was purified by slurrying in MTBE
(325 mL) at room temperature and recrystallised with acetoni-
trile (390 mL) to give the title compound 2 as an off-white solid.
1
(65.0 g, 137 mmol, 63%, 98.5% de27). H NMR (400 MHz,
CDCl3) δ 1.40ꢀ1.45 (dd, 1H), 1.48ꢀ1.51 (d, 3H), 1.58 (s, 3H),
1.60ꢀ1.63 (dd, 1H), 1.70ꢀ1.75 (dd, 1H), 2.99 (s, 3H), 5.07ꢀ
5.15 (m, 1H), 6.02ꢀ6.07 (d, 1H), 6.72 (brs, 1H), 7.24ꢀ7.27 (dd,
1H), 7.38ꢀ7.40 (d, 1H), 7.58ꢀ7.61 (dd, 1H), 7.66ꢀ7.70 (dd,
1H), 8.61ꢀ8.63 (m, 1H); 13C NMR (100 MHz, DMSO-d6) δ
168.5, 149.3, 146.0, 145.2, 144.8, 144.4, 144.1, 136.6, 133.0,
129.6, 128.5, 127.8, 126.4, 126.1, 123.7, 121.0, 120.7, 118.3, 48.2,
41.5, 30.3, 25.8, 22.9, 19.6, 18.1; LC/MS m/z (ES+) 476 [MH]+;
Anal. Calcd for C20H21F4N3O3S: C, 50.48; H, 4.45; N, 8.83; Cl,
7.45; S, 6.74. Found: C, 50.38; H, 4.42; N, 8.79; Cl, 7.52; S, 6.80.
’ ASSOCIATED CONTENT
S
Supporting Information. Further information summar-
b
izing the chiral base screen of cyclopropyl acid 5 and the
cyclopropanation screen of 20. This material is available free of
charge via the Internet at http://pubs.acs.org.
(1S,2S)-N-[(1R)-1-{3-Fluoro-4-[(methylsulfonyl)amino]phenyl}-
ethyl]-2-methyl-2-[6-(trifluoromethyl)pyridin-3-yl]cyclopropane-
carboxamide (1). To a solution of 5 (49.0 g, 200 mmol) in THF
(250 mL, 5 mL/g) was added 3 (53.7 g, 200 mmol) followed by
THF (125 mL, 2.5 mL/g). 1-propylphosphonic acid cyclic
anhydride (T3P) (178 mL, 300 mmol) was added followed by
diisopropylethylamine (209 mL, 1.20 mol) and THF (125 mL,
2.5 mL/g) and the reaction mixture heated to 55 °C for 2 h. The
reaction was cooled to room temperature, and to the vessel was
added water (500 mL) followed by ethyl acetate (500 mL). The
organic layer was retained and washed with water (750 mL). sat.
sodium carbonate (500 mL), sat. citric acid (500 mL), and water
(500 mL), dried over MgSO4, and evaporated in vacuo to give a
foam. The residue was purified by slurrying in MTBE (450 mL)
at 50 °C for 1 h, cooled to room temperature, and filtered to give
the title compound 1 as an off-white solid. (71.1 g, 154 mmol,
’ AUTHOR INFORMATION
Corresponding Author
*daniel.laity@pfizer.com; benjamin.pibworth@pfizer.com
’ ACKNOWLEDGMENT
We thank our colleagues in Analytical Chemistry for analytical
support along with colleagues in Process Safety, Medicinal
Chemistry, and Chemical Research and Development for their
support and ideas.
’ REFERENCES
(1) Hutchins, C. W.; Kort, M. E.; Kym, P. R. Biochem. Pharmacol.
2009, 78, 211–216.
(2) Appendino, G. J. Med. Chem. 2004, 47, 2717–2723.
(3) Bevan, S.; Boelsteri, J.; Bovermann, G.; Breckenridge, R.; Davies,
J. W.; Hughes, G. A.; James, I.; Oberer, L.; Walpole, C. S. J.; Winter, J.;
Wrigglesworth, R. J. Med. Chem. 1994, 37, 1942–1954.
(4) Lee, J. Bioorg. Med. Chem. 2007, 15, 6043.
1
77%, 98.9% de27). H NMR (400 MHz, CDCl3) δ 1.38ꢀ1.43
(m, 1H), 1.47ꢀ1.50 (d, 3H), 1.58 (s, 3H), 1.59ꢀ1.62 (dd, 1H),
1.72ꢀ1.76 (dd, 1H), 3.00 (s, 3H), 5.07ꢀ5.16 (m, 1H), 6.20ꢀ
6.25 (d, 1H), 6.60 (brs, 1H), 7.09ꢀ7.11 (m, 1H), 7.12ꢀ7.14
(m, 1H), 7.46ꢀ7.51 (m, 1H), 7.57ꢀ7.60 (d, 1H), 7.65ꢀ7.69
1199
dx.doi.org/10.1021/op200177b |Org. Process Res. Dev. 2011, 15, 1192–1200

Organic Process Research & Development
ARTICLE
(5) (a) Hester, J B. U.S. Patent 5,155,268 A, 1992. (b) Hanazawa, T.;
Hirano, M.; Inoue, T.; Nagayama, S.; Nakao, K.; Shishido, Y.; Tanaka, H.
International Patent WO 2006/097817 A1, 2006. (c) Hanazawa, T.;
Hirano, M.; Inoue, T.; Nagayama, S.; Nakao, K.; Shishido, Y.; Tanaka, H.
U.S. Patent 7,622,589 B2, 2009.
(6) Yields were consistently below 50% unless toxic thallium salts
were added to improve the regioselectivity of the coupling. Product also
required column purification.
(7) Up to 20% of dimethyl amine impurity 16 formed when reaction
was carried out in DMF.
(8) Purity as determined by HPLC was >98.5%.
(9) For reviews on the chemistry of sulfinimines see: (a) Ellman, J. A.
J. Org. Chem. 1999, 64, 1278. (b) Davis, F. A. Tetrahedron 2004,
60, 8003. (c) Ellman, J. A.; Herbage, M. A.; Robak, M. T. Chem. Rev.
2010, 110, 3600.
(10) Other solvents examined for the imine reduction (ꢀ20 °C)
included EtOAc (82% de) and 2-methyl tetrahydrofuran (32% de).
(11) Ellman, J. A. Tetrahedron Lett. 1999, 40, 6709.
(12) The diastereoselectivities of 14 and 15 were calculated
by HPLC.
(13) Original conditions involved quench of NaBH4 by first adding
methanol dropwise over an extended period of time followed by water.
(14) Knochel, P. Angew. Chem., Int. Ed. 2004, 43, 3333–3336.
(15) Leazer, J. L. J. Org. Chem. 2003, 68, 3695–3698.
(16) Appleby, I. C. Chem. Ind. 1971, 120.
(17) Clark, J. D. Thermochim. Acta 2002, 386, 65–72 for Monsanto.
(18) Nishiyama, H. Bull. Chem. Soc. Jpn. 1995, 68, 1247–1262.
(19) Gallo, E.; Rose, E. Organometallics 2008, 27, 6143–6151.
(20) Hanazawa, T.; Hirano, M.; Inoue, T.; Nagayama, S.; Nakao, K.;
Shishido, Y.; Tanaka, H. Patent WO 2006/09787
(21) Further information is available in the Supporting Information.
(22) Hanazawa, T.; Nagayama, S.; Nakao, K.; Shishido, Y.; Tanaka,
H. Patent WO 2007/129188, 2007
(23) Cornejo, A. Tetrahedron 2005, 12107–12110.
(24) Ee measured by chiral HPLC (OJ_H column) heptane (90):
IPA (10): TFA (0.1), 20 °C
(25) Hu, S.; Kline, B. J.; Martinez, C.; Tao, J.; Yazbeck, D. R. Adv.
Synth. Catal. 2003, 345, 524–532.
(26) 5 with a chiral purity of 60% ee was trialled in the coupling, but a
screen of column and crystallisation conditions failed to purge the
diastereoisomer to acceptable levels. As a result, the salt resolution using
(S)-1,2,3,4-tetrahydro-1-naphthylamine had to be implemented to in-
crease the chiral purity of 5 to 93% ee.
(27) The diastereoisomers could be resolved by HPLC, providing
confidence that the impurities could be accurately measured.
(28) Montalbetti, C. A. G. N. Tetrahedron 2005, 61, 10827.
1200
dx.doi.org/10.1021/op200177b |Org. Process Res. Dev. 2011, 15, 1192–1200