Kilogram-lab-scale oxindole synthesis via palladium-catalyzed c-h functionalization

Article abstract of DOI:10.1021/op200332p

A scalable method for the preparation of oxindole 8, a key intermediate en route to a serine palmitoyl transferase inhibitor, compound 1, is presented. A three-step, chromatography-free route has been designed that takes advantage of Buchwald's palladium-catalyzed C-H functionalization to cyclize an α-chloroacetanilide to form the five-membered ring. This process has been successfully carried out in our kilogram laboratory facility on 10-kg scale in 76% yield.

Full text of DOI:10.1021/op200332p

Technical Note
pubs.acs.org/OPRD
Kilogram-Lab-Scale Oxindole Synthesis via Palladium-Catalyzed C−H
Functionalization
E. Jason Kiser,* Javier Magano,* Russell J. Shine, and Michael H. Chen
,§
,†
†
⊥
†Chemical Research and Development, Pfizer Inc., Eastern Point Road, Groton, Connecticut 06340, United States
*
S Supporting Information
ABSTRACT: A scalable method for the preparation of oxindole 8, a key intermediate en route to a serine palmitoyl transferase
inhibitor, compound 1, is presented. A three-step, chromatography-free route has been designed that takes advantage of
Buchwald’s palladium-catalyzed C−H functionalization to cyclize an α-chloroacetanilide to form the five-membered ring. This
process has been successfully carried out in our kilogram laboratory facility on 10-kg scale in 76% yield.
INTRODUCTION
Compound 1 (Figure 1) is a molecule designed to inhibit the
serine palmitoyl transferase (SPT) enzyme. It has been
the use of NaH and two chromatographic purifications. In
addition, the high cost of 2 made this starting material
undesirable in the long run for large-scale manufacturing. On
the basis of a report from Buchwald’s group on the preparation
of substituted oxindoles from α-chloroacetanilides via palla-
dium-catalyzed C−H functionalization, we envisioned that this
method could provide rapid access to oxindole 8 and streamline
the synthesis of compound 1. The original conditions reported
■
1
proposed that inhibition of SPT may elevate levels of HDL
2
cholesterol. This effect is associated with decreased
6
cardiovascular health risk, and therefore, compound 1
represents a potential treatment for heart disease.
3
by this group employed Pd(OAc) as catalyst and 2-(di-tert-
2
butylphosphino)biphenyl as ligand in the presence of Et N and
3
toluene at 80 °C (Scheme 2). This method represents an
innovative and direct approach for the preparation of oxindoles
that avoids the high temperatures and strongly acidic
conditions usually required when a Friedel−Crafts-type of
approach is employed with α-halo- or α-hydroxy acetanilides
to generate similar substrates. In addition, the α-chloroaceta-
nilide precursor can be readily accessed from the corresponding
aniline and chloroacetyl chloride.
7
8
Figure 1. Structure of SPT inhibitor 1.
MEDICINAL CHEMISTRY ROUTE
The original medicinal chemistry route for synthesizing the
target 1 involved a key oxindole intermediate 8 which was
■
To the best of our knowledge, only one report on the
application of this method for the large-scale preparation of an
1a
9
accessed through a five-step synthetic process (Scheme 1).
The synthesis started with the esterification of 3-fluoro-4-
oxindole has been found in the literature. Conditions similar to
6
those described by Buchwald et al. were employed with the
nitrobenzoic acid (2) in MeOH as solvent and catalytic H SO4
to provide methyl ester 3 in 89% yield. This intermediate then
underwent reaction with di-tert-butyl malonate and NaH as
only major change being trifluorotoluene in place of toluene as
solvent. Encouraged by this precedent on multikilogram scale,
we decided to investigate the application of this approach to
our substrate.
2
base to generate intermediate 4 after fluoride displacement via
4
S Ar in 79% yield. The nitro group was hydrogenated with
N
5
RESULTS AND DISCUSSION
Before we could test Buchwald’s C−H activation method for
Ra−Ni as catalyst to afford aniline 5 in 88% yield, which
underwent reductive amination with 1-benzyloxycarbonyl-4-
■
6
piperidone (6) to give secondary amine 7 in 82% yield after
the preparation of oxindole 8, we had to access the appropriate
α-chloroacetanilide 11. This was expediently accomplished in
two steps from commercially available methyl 4-aminobenzoate
(9), a considerably cheaper starting material than 3-fluoro-4-
nitrobenzoic acid (2), and 1-benzyloxycarbonyl-4-piperidone
(6) (Scheme 3).
4
a,5
chromatography. Indole 8 was produced in 81% yield after
treating 7 with p-TsOH·H O in toluene at reflux followed by
2
4
a
chromatography. Four additional steps completed the syn-
thesis of 1 from 8.
Even though this route was satisfactory for the preparation of
small batches of 1, the need for kilo quantities of material to
support clinical studies required the development of either an
optimized or a new route. We focused our attention on the
preparation of oxindole 8 as a key advanced intermediate en
route to 1. The Medicinal Chemistry route provided this
material in five steps and 41% overall yield from 2 but required
Several conditions were tested for the reductive amination
between 6 and 9. Initially, mixtures of 6, 9 (1 equiv), and
HOAc (1 equiv) were added in a number of solvents (THF,
Received: November 21, 2011
Published: December 23, 2011
©
2011 American Chemical Society
255
dx.doi.org/10.1021/op200332p | Org. ProcessRes. Dev. 2012, 16, 255−259

Organic Process Research & Development
Technical Note
a
Scheme 1. Medicinal Chemistry route to oxindole 8
a
Reagents and conditions: (a) H SO , MeOH, reflux, 5.5 h, 89%. (b) CH (CO t-Bu) , NaH, THF, 0 °C to rt, 79%. (c) H , Ra−Ni, rt, 88%. (d) (i)
2
4
2
2
2
2
NaBH(OAc) , HOAc, DCE, rt, 72 h; (ii) chromatography, 82%. (e) (i) p-TsOH·H O, toluene, reflux, 2 h; (ii) chromatography, 81%. (f) 2 N
3
2
NaOH, MeOH, 80 °C, 2 h, 97%. (g) (i) MeNH (2 M in THF), EDC·HCl, HOBt, CH Cl , rt, 72 h; (ii) chromatography, 56%. (h) H , 10% Pd/C,
2
2
2
2
THF, 10 h, 66%. (i) (3aR,6aS)-2-(chloroacetyl)octahydrocyclopenta[c]pyrrole, KHCO , MeCN, 72 °C, 1.5 h, 65%.
3
Scheme 2. Oxindole formation via C−H functionalization by
(1:1). Further improvements were the discovery that adding
,
6 a
Buchwald’s group
NaBH(OAc) in several portions and the use of 1.2 equiv of
3
piperidone 6 gave complete consumption of 9, an acceptable
level (10−12%) of piperidol byproduct 12, and a substantial
increase in yield in either halogenated solvent (86−87%). DCE
(
class 1 solvent) was initially used as the solvent in this
reaction; however, CH Cl (class 2 solvent) was preferred
2
2
11
during large-scale operations due to its lower toxicity. The
optimal conditions involved dissolving the two substrates in
a
1
2
R = Me, OMe, Cl, CF , OTBS, Me Si, NO ; R = Me, Et, Ph, Bn, p-
3
3
2
CH Cl with 1 equiv of acetic acid and adding NaBH(OAc) in
methoxybenzyl, diphenylmethyl.
2
2
3
several portions over a few hours. Upon reaction completion,
the mixture was quenched into 2 N NaOH, the layers were
separated, and the organic layer was washed with water to
ensure boron byproducts removal. Concentration of the
organic phase and crystallization from a mixture of heptane
and MTBE afforded the product in 85−90% yield and >98%
purity (HPLC area %) on laboratory-scale (25 g). Two runs
were carried out in our kilo lab facility (4.2 and 6.0 kg) in 76%
and 86% yield, respectively (HPLC purities >98 area %).
The original method for carrying out the subsequent
acylation step to produce α-chloroacetanilide 11 employed
CH Cl as solvent and solid K CO as base. These conditions
a
Scheme 3. Preparation of α-chloroacetanilide 11
2
2
2
3
a
Reagents and conditions: (a) NaBH(OAc) , HOAc, CH Cl , 20 °C,
3
2
2
tended to give variable reaction rates, often resulting in
incomplete conversion most likely due to the biphasic nature of
the mixture. Schotten−Baumann conditions (CH Cl or EtOAc
5
.5 h, 86%. (b) ClC(O)CH Cl, py, EtOAc, 20 °C, 1.5 h, 89%.
2
2
2
1
,2-dichloroethane (DCE), CH Cl ) to a suspension of
2 2
NaBH(OAc) in the same solvent, but incomplete reactions
and considerable amounts of alcohol 12 (up to 17%) resulting
from the reduction of 6 were observed (Figure 2). Never-
and aqueous K CO ) were also investigated, but low
2 3
3
conversion was observed due to hydrolysis of the acid chloride
before it had a chance to react with poorly nucleophilic amine
10. Better results were obtained when the reaction was
performed by treating an EtOAc solution of 10 with
chloroacetyl chloride in the presence of pyridine. Excesses of
both reagents (1.5 equiv) were required to drive the reaction to
completion. Et N as a less toxic alternative than pyridine was
3
found to afford intermediate 11 with a yellow color that could
not be removed via crystallization from MTBE (vide infra). The
workup involved quenching the reaction with water, separating
the layers, and washing the organic layer with half-saturated
brine to ensure removal of chloroacetic acid and pyridine
hydrochloride byproducts and prevent emulsion formation. α-
Chloroacetanilide 11 was isolated by concentration of the
organic layer and crystallization of the residue from MTBE.
This optimized protocol using pyridine in EtOAc was
implemented in the laboratory on 100-g scale in >85% yield
and >98% HPLC purity (area %). Yields and purities were
Figure 2. Alcohol byproduct from the reduction of piperidone 6.
theless, higher conversions were obtained in DCE and CH Cl2
2
(
61−65%) compared to that in THF (35%). When the reverse
10
order of addition was implemented (addition of solid
NaBH(OAc) in one portion to a solution of 6, 9, and
3
HOAc in either CH Cl or DCE), respective yields of 62% and
2
2
73% of desired product 10 were obtained. The level of
piperidol 12 was 8−12%, and this byproduct could be purged
in the filtrates after a final crystallization from heptane/MTBE
2
56
dx.doi.org/10.1021/op200332p | Org. ProcessRes. Dev. 2012, 16, 255−259

Organic Process Research & Development
Technical Note
reproducible during two campaigns in our kilogram laboratory
facility on 7.8- and 12.5-kg scale.
With 11 in hand, we focused our attention on the oxindole6-
forming step. Initial efforts at applying Buchwald’s method
toward cyclizing compound 11 (Scheme 4) gave incomplete
by adding heptane or MTBE as antisolvent resulted in the
coprecipitation of TEA·HCl, which necessitated multiple
aqueous washes of the filter cake to be removed and longer
product drying times. The 2-propanol purification gave product
of high quality (>98 area % by HPLC) in 70−80% overall yield.
Residual palladium levels varied between 100−800 ppm. The
optimized process was carried out multiple times on 25−50-g
scale, and it was subsequently scaled up in our kilo laboratory
facility twice on 5-kg and once on 10-kg scale with yields in the
15
a
Scheme 4. Preparation of 8 via C−H functionalization
7
6−84% range. The application of this chemistry to our
substrate (bearing a Cbz-protecting group) also represents an
extension of the type of protecting group that can be employed
on the nitrogen of 11 with respect to the original article.
6
Even though Buchwald stated in his original article that, after
a
Reagents and conditions: (a) Pd(OAc)2 (10 mol %), 2-di-(tert-
testing several ligands, only 2-(di-tert-butylphosphino)biphenyl
butylphosphino)biphenyl (20 mol %), TEA, MeTHF−IPA (4:1 v/v),
7
6
afforded the desired oxindole, we decided to investigate the
0−75 °C, 2.5 h, 76%.
use of 5-(di-tert-butylphosphino)-1′,3′,5′-triphenyl-1′H-[1,4′]-
bipyrazole (Bippyphos, 13, Figure 3) in this type of coupling.
reaction and significant side products. It was soon recognized
that the problem was the use of toluene as solvent, since the
reaction mixture was heterogeneous and tended to produce
sticky solids that prevented good agitation as the reaction
proceeded. As a result, several solvent and base combinations
were tested to find a system capable of providing a more
homogeneous mixture. Thus, 2-MeTHF with TEA gave
complete conversion at 80 °C but still produced tarry and
somewhat sticky solids, whereas that with (n-Bu) N gave low
3
conversion to the desired product and higher level of impurities
due to the prolonged reaction time (overnight at 80 °C).
MeCN/TEA showed very low conversion and some decom-
position after 4 h at 80 °C, whereas complete consumption of
Figure 3. Structure of Bippyphos ligand.
16
This nonproprietary ligand was originally developed at Pfizer,
11 was observed at the same temperature in DMF/TEA to
became available to us shortly after the work described in this
article had been implemented, and is currently commercially
available from several sources. When α-chloroacetanilide 11
was subjected to the same reaction conditions described in
afford a complex mixture with a small amount of oxindole 8.
The small improvement observed with 2-MeTHF led us to
investigate this system further. A major breakthrough came
when the reaction was performed in a 2-MeTHF/2-propanol
mixture. Initially, the reaction was carried out with a 1:1 (v/v)
ratio to provide oxindole 8 in 49% yield after 1.5 h at 80 °C.
However, when a 4:1 (v/v) mixture of 2-MeTHF and 2-
propanol was tested, complete cyclization occurred in less than
Scheme 4 in the presence of Pd(OAc) (10 mol %) and
2
Bippyphos (20 mol %), complete conversion was observed to
oxindole 8 after 2 h and this material was isolated in 84% yield
on 25 g-scale. Following the successful application of this
catalytic system, we have further expanded the use of this
technology for the preparation of other oxindoles in our
laboratories.
1
h at 70−80 °C to afford 8 in 70% yield on gram scale. Tarry
solids were still present in the reaction, but the addition of 2-
propanol resulted in a more homogeneous and stirrable
mixture. The ideal amounts of palladium catalyst and ligand
were found to be 10 mol % and 20 mol %, respectively. The use
of half of these loadings of catalyst and ligand resulted in the
reaction stalling out at 80−85% conversion as well as the
In summary, we have developed a high-yielding, chromatog-
raphy-free, scalable process for producing kilogram quantities of
6
oxindole 8 utilizing Buchwald’s C−H functionalization
chemistry. The application of this methodology allowed us to
remove two processing steps from the original route to
compound 8 and eliminated the need for NaH and two
chromatographic purifications. The original stirring issues
associated with the use of toluene have been overcome by
replacing this solvent with a 4:1 2-MeTHF/2-propanol mixture
to provide a more homogeneous reaction mixture. We have
also demonstrated the applicability of Bippyphos as an
alternative ligand to 2-(di-tert-butylphosphino)biphenyl in this
type of coupling. Even though this technology was suitable for
early development campaigns, further optimization would be
required to decrease the amount of catalyst and ligand to turn it
into a more cost-effective process. On the basis of some
additional work carried out in our laboratories for the
generation of other oxindoles on gram-scale, we believe that
this methodology can become a general and scalable approach
for the manufacture of this type of materials.
12
formation of higher levels of impurities. The 1:2 ratio of metal
to ligand was also found to be important; an attempt with the
13
reverse ratio (2:1) gave little or none of oxindole 8. Due to
the relatively low solubility of 8 at ambient temperature,
attempts at a conventional aqueous workup were abandoned, as
the compound had a tendency to precipitate unexpectedly from
solution. In addition, the postreaction mixture contained
insoluble black, tarry solids (TEA·HCl and metal/ligand
1
4
residues) which caused intractable emulsions if partitioning
with water was attempted. These workup issues were eradicated
by filtering the hot reaction mixture through diatomaceous
earth. The filtrate was then concentrated and the crude
oxindole 8 recrystallized from 2-propanol. The product could
also be precipitated directly from the reaction filtrate by
cooling; however, yields were lower due to the modest
solubility of 8 in 2-MeTHF. Attempts to increase this yield
2
57
dx.doi.org/10.1021/op200332p | Org. ProcessRes. Dev. 2012, 16, 255−259

Organic Process Research & Development
Technical Note
EXPERIMENTAL SECTION
and filtered. The solids were washed with MTBE (37.6 L)
and dried under vacuum at 50 °C for 25 h to afford 13.53 kg
■
Reaction completion and product purity were evaluated by
HPLC using the following conditions: column: YMC PackPro
C18, 150 mm × 4.6 mm, 5 μm; flow rate: 1.0 mL/min;
wavelength: 215 nm; temperature: 30 °C; injection volume:
(
89%) of acetanilide 11 as a white solid. Mp: 107 °C (DSC).
HPLC retention time: 8.1 min. HPLC purity: 99.3% (area %).
1
H NMR (400 MHz, CDCl ) δ 1.15−1.35 (m, 2 H), 1.85 (d, J =
3
1
4
1.67 Hz, 2 H), 2.89 (m, 2 H), 3.70 (s, 2 H), 3.98 (s, 3 H),
.24 (br s, 2 H), 4.72−4.85 (m, 1 H), 5.06 (br s, 2 H), 7.24 (m,
5
μL after diluting the sample with 9:1 CH CN/water to a
3
concentration of approximately 0.5 mg/mL of analyte; eluent
A: water (0.2% HClO ); B: CH CN; gradient: 30% B to 95% B
J = 8.28 Hz, 2 H), 7.27−7.38 (m, 5 H), 8.14 (d, J = 8.66 Hz, 2
H). C NMR (100 MHz, CDCl ) δ 30.22, 42.12, 43.32, 52.55,
4
3
13
3
over 10 min; hold 95% B for 5 min; return to 30% B over 1
min. HPLC purities are given in area %.
5
1
3.47, 67.20, 127.92, 128.04, 128.47, 130.24, 131.10, 131.22,
36.54, 141.23, 154.92, 165.52, 165.80. HRMS (EI) Calcd for
Synthesis of Benzyl 4-(4-(methoxycarbonyl)-
phenylamino)piperidine-1-carboxylate (10). A solution
of methyl 4-aminobenzoate (9, 6.00 kg, 39.7 mol) and 1-
benzyloxycarbonyl-4-piperidone (6, 11.14 kg, 47.8 mol) in
CH Cl (60 L) and glacial HOAc (2.38 kg, 39.6 mol) was
C H ClN O (M + H) 445.15248, found 445.15246.
23
26
2
5
Synthesis of Methyl 1-(1-(Benzyloxycarbonyl)-
piperidin-4-yl)-2-oxoindoline-5-carboxylate (8). A solu-
tion of α-chloroacetanilide 11 (10.95 kg, 24.6 mol), Pd(OAc)2
2
2
(
(
(
0.553 kg, 2.46 mol), and 2-(di-tert-butylphosphino)biphenyl
1.47 kg, 4.93 mol) in 2-MeTHF (71.2 L) and 2-propanol
17.7 L) was sparged subsurface with nitrogen for 30 min.
treated with NaBH(OAc) (12.60 kg, 59.4 mol) in 5 portions
3
over 4 h at 15− 25 °C. The reaction mixture was stirred until
HPLC analysis indicated complete reaction (5.5 h). The
reaction was transferred into a solution of sodium hydroxide
Triethylamine (3.74 kg, 37.0 mol) was added and the reaction
was heated to 71 °C and stirred until HPLC analysis indicated
complete consumption of 11 (2.5 h). Without cooling, the
reaction mixture was filtered through Celite followed by a
second in-line polishing filter (0.2 μm). The reactor and filters
were rinsed with hot (70 °C) 2-MeTHF (11 L), and the
combined filtrates were cooled to 51 °C and concentrated at
reduced pressure to 20 L. The slurry was diluted with 2-
propanol (153 L) and heated to reflux until all the solids
dissolved (30 min). The mixture was cooled to 25 °C over 3 h,
and the resulting slurry was stirred for 5 h. The solids were
filtered, washed with 2-propanol (32.9 L), and dried under
vacuum at 35 °C for 72 h to give 7.70 kg (76%) of oxindole 8
(
50 wt %, 4.76 kg, 59.5 mol) in water (30 L) followed by a
CH Cl rinse (6 L). The reaction mixture was stirred for 1 h at
2
2
2
0 °C, and the layers were separated. The organic layer was
washed with water (2 × 30 L), concentrated at atmospheric
pressure to a volume of 30 L, and diluted with methyl tert-butyl
ether (MTBE, 30 L). The mixture was concentrated to a
volume of 30 L, and this solvent displacement operation was
repeated twice more. The residue was diluted with MTBE
(
60 L), and to the resulting slurry was added heptane (60 L).
The suspension was stirred for 3 h at 20 °C, and the solids were
filtered and washed with 1:1 MTBE/heptane (24 L). The solid
was dried under vacuum at 50 °C for 24 h to give 12.54 kg
86%) of amine 10 as a white solid. Mp: 94 °C (DSC). HPLC
retention time: 8.4 min. HPLC purity: 98.7% (area %). H
as a white solid. Mp: 154 °C (DSC). HPLC retention time:
(
1
7
.90 min. HPLC purity: 98.4% (area %). H NMR (400 MHz,
1
CDCl ) δ 1.66 (d, J = 11.80 Hz, 2 H), 2.26 (dd, J = 12.61, 3.83
3
NMR (400 MHz, CDCl ) δ 1.41 (d, J = 10.04 Hz, 2 H), 2.07
3
Hz, 2 H), 2.84 (br s, 2 H), 3.48 (s, 2 H), 3.83 (s, 3 H), 4.21−
(
(
6
8
4
1
d, J = 11.67 Hz, 2 H), 3.04 (t, J = 11.73 Hz, 2 H), 3.47−3.60
4
7
.46 (m, 3 H), 5.10 (s, 2 H), 6.89 (d, J = 8.41 Hz, 1 H), 7.23−
13
m, 1 H), 3.87 (s, 3 H), 4.17 (br s, 3 H), 5.12−5.20 (m, 2 H),
.36 (m, 5 H), 7.84. (s, 1 H), 7.88 (d, J = 8.28 Hz, 1 H).
C
.57 (d, J = 8.78 Hz, 2 H), 7.30−7.46 (m, 5 H), 7.88 (d, J =
NMR (100 MHz, CDCl ) δ 28.08, 35.49, 43.72, 50.18, 52.09,
3
.78 Hz, 2 H). 1 C NMR (100 MHz, CDCl ) δ 32.01, 42.76,
3
3
67.40, 109.13, 124.03, 124.59, 125.88, 128.06, 128.17, 128.57,
130.30, 136.62, 147.60, 155.14, 166.70, 174.96. HRMS (EI)
Calcd for C H N O (M + H) 409.17580, found 409.17582.
9.62, 51.58, 67.23, 111.92, 118.61, 127.91, 128.08, 128.54,
31.67, 136.70, 150.40, 155.21, 167.20. HRMS (EI) Calcd for
23
25
2
5
C H N O (M + H) 369.18088, found 369.18100.
21
25
2
4
S y n t h e s i s o f B e n z y l 4 - ( 2 - C h l o r o - N - ( 4 -
methoxycarbonyl)phenyl)acetamido)piperidine-1-car-
boxylate (11). A solution of amine 10 (12.46 kg, 33.8 mol)
and pyridine (4.04 kg, 51.1 mol) in EtOAc (119 L) was treated
with chloroacetyl chloride (5.77 kg, 51.1 mol) over 40 min at
ASSOCIATED CONTENT
Supporting Information
Copies of H and C NMR spectra for compounds 8, 10, and
1. This material is available free of charge via the Internet at
■
(
*
S
1
13
1
http://pubs.acs.org.
15−25 °C. The reaction was stirred until HPLC analysis
indicated complete consumption of 10 (1.5 h), and water (62.7 L)
was added. The resulting mixture was stirred for 1 h at 20 °C,
and the layers were separated. The organic layer was
washed with a solution of NaCl (8.15 kg) in water (62.7 L).
The layers were separated, and the organic phase was
concentrated at reduced pressure to 20 L. EtOAc (37.6 L)
was charged to the reactor, and the mixture was concentrated
again to a volume of 20 L. The product was then fully dissolved
by the addition of EtOAc (157 L), and the resulting solution
was filtered through an inline polishing filter (0.2 μm) to
remove insoluble material. The reactor and filter were rinsed
with EtOAc (12.5 L), and the solution was concentrated at
reduced pressure to a volume of 20 L. MTBE (75.2 L) was
added, and the mixture was heated at reflux for 30 min and
cooled to 25 °C over 3 h. The resulting slurry was stirred for 6 h
AUTHOR INFORMATION
■
Corresponding Author
Javier.Magano@Pfizer.com
Present Addresses
Ash Stevens Inc., 18655 Krause Street, Riverview, MI 48193,
United States
Wuhan Masteam Bio-Technology Research Institute, Wuhan,
30074, P.R. China
§
⊥
4
ACKNOWLEDGMENTS
■
We thank Dr. Robert Singer for providing ligand Bippyphos
and Groton’s kilo laboratory and Research Analytical staff for
their support during the implementation of this project.
2
58
dx.doi.org/10.1021/op200332p | Org. ProcessRes. Dev. 2012, 16, 255−259

Organic Process Research & Development
Technical Note
REFERENCES
■
(
1) (a) Bolton, G. L.; Hutchings, R. H.; Kohrt, J. T; Park, W. K. C.;
Van Huis, C. A. WO 2008084300 A1 20080717, 2008; Chem. Abst.
1
(
49, 153082. (b) Hanada, K. Biochim. Biophys. Acta 2003, 1632, 16.
2) HDL is commonly referred to as “good” cholesterol. See:
deGoma, E. M.; Rader, D. J. Nat. Rev. Cardiol. 2011, 8, 266.
3) (a) Castelli, W. P.; Garrison, R. J.; Wilson, P. W.; Abbott, R. D.;
Kalousdian, S.; Kannel, W. B. JAMA, J. Am. Med. Assoc. 1986, 256,
835. (b) Grundy, S. M. Am. J. Cardiol. 2002, 90, 11i.
4) (a) Johnson, D. J.; Forbes, I. T.; Watson, S. P.; Garzya, V.;
(
2
(
Stevenson, G. I.; Walker, G. R.; Mudhar, H. S.; Flynn, S. T.; Wyman,
P. A.; Smith, P. W.; Murkitt, G. S.; Lucas, A. J.; Mookherjee, C. R.;
Watson, J. M.; Gartlon, J. E.; Bradford, A. M.; Brown, F. Bioorg. Med.
Chem. Lett. 2010, 20, 5434. (b) Zimmermann, K.; Wittman, M. D.;
Saulnier, M. G.; Velaparthi, U.; Langley, D. R.; Sang, X.; Frennesson,
D.; Carboni, J.; Li, A.; Greer, A.; Gottardis, M.; Attar, R. M.; Yang, Z.;
Balimane, P.; Discenza, L. N.; Vyas, D. Bioorg. Med. Chem. Lett. 2008,
1
(
8, 4075.
5) (a) Zaveri, N. T.; Jiang, F.; Olsen, C. M.; Deschamps, J. R.;
Parrish, D.; Polgar, W.; Toll, L. J. Med. Chem. 2004, 47, 2973.
b) Forbes, I. T. Tetrahedron Lett. 2001, 42, 6943.
6) Hennessy, E. J.; Buchwald, S. L. J. Am. Chem. Soc. 2003, 125,
2084.
7) Beckett, A. H.; Daisley, R. W.; Walker, J. Tetrahedron 1968, 24,
093.
(
(
1
(
6
(
8) Ben-Ishai, D.; Peled, N.; Sataty, I. Tetrahedron Lett. 1980, 21, 569.
(
9) For a recent application of this method to the synthesis of an
oxazolidinone antibacterial, see: Choy, A.; Colbry, N.; Huber, C.;
Pamment, M.; Van Duine, J. Org. Process Res. Dev. 2008, 12, 884.
(
10) The addition of a solution of 6 and 9 to a suspension of
NaBH(OAc) in the corresponding solvent would have been preferred
3
for scale-up convenience.
(
11) Dwivedi, A. M. Pharm. Technol. 2002, 26, 42.
(
12) LC/MS data suggested that the major impurity was the des-
chlorinated starting material.
13) The Supporting Information in reference 6 erroneously tells the
reader to use this reverse ratio.
14) These solids were actually the impetus behind using 2-propanol
(
(
as a cosolvent in the reaction. Without the added alcohol, the black,
tarry solids would build up into a sticky material that would prevent
efficient stirring and mixing. We envisioned that this problem would
worsen on scale. The addition of 20% alcohol to the solvent avoided
this issue. Interestingly, 2-propanol also had a tendency to kick-start
the reaction at room temperature; however, without additional heat, it
would stop at only fractional conversion. The exact role of the alcohol
as a reaction mediator was not determined.
(
15) No attempts were made to remove residual palladium from 8
since a subsequent step also employed this metal (Cbz-protecting
group removal via catalytic hydrogenation in the presence of Pd/C).
Pd was purged downstream in the synthesis.
(
16) (a) Withbroe, G. J.; Singer, R. A.; Sieser, J. E. Org. Process Res.
Dev. 2008, 12, 480. (b) Singer, R. A.; Dore, M.; Sieser, J. E.; Berliner,
́
M. A. Tetrahedron Lett. 2006, 47, 3727. (c) Singer, R. A.; Caron, S.;
McDermott, R. E.; Arpin, P.; Do, N. M. Synthesis 2003, 1727.
2
59
dx.doi.org/10.1021/op200332p | Org. ProcessRes. Dev. 2012, 16, 255−259