Scalable synthesis of a nonracemic α-arylpropionic acid via ketene desymmetrization for a glucokinase activator

Article abstract of DOI:10.1021/op400354g

Process research and development for a synthesis of the chiral carboxylic acid (R)-2 as a key intermediate of the glucokinase activator (R)-1 is described. The construction of the stereocenter at the α-carbon is a key point for the synthesis of (R)-2. The proposed process utilizes desymmetrization of a ketene in situ generated from the corresponding racemic carboxylic acid Rac-2 with (R)-pantolactone as a chiral auxiliary followed by hydrolysis of the resulting ester. This key step has been successfully scaled up to 20 kg, which demonstrates that this synthetic approach is comparable with a previously reported approach via enantioselective hydrogenation.

Full text of DOI:10.1021/op400354g

Article
pubs.acs.org/OPRD
Scalable Synthesis of a Nonracemic α‑Arylpropionic Acid via Ketene
Desymmetrization for a Glucokinase Activator
Takafumi Yamagami,* Noriaki Moriyama, Masahiro Kyuhara, Atsushi Moroda, Takeshi Uemura,
Hiroaki Matsumae, Yasunori Moritani, and Isao Inoue
Process Chemistry Research Laboratories, CMC Division, Mitsubishi Tanabe Pharma Corporation, 3-16-89, Kashima, Yodogawa-ku,
Osaka 532-8505, Japan
S
* Supporting Information
ABSTRACT: Process research and development for a synthesis of the chiral carboxylic acid (R)-2 as a key intermediate of the
glucokinase activator (R)-1 is described. The construction of the stereocenter at the α-carbon is a key point for the synthesis of
(R)-2. The proposed process utilizes desymmetrization of a ketene in situ generated from the corresponding racemic carboxylic
acid Rac-2 with (R)-pantolactone as a chiral auxiliary followed by hydrolysis of the resulting ester. This key step has been
successfully scaled up to 20 kg, which demonstrates that this synthetic approach is comparable with a previously reported
approach via enantioselective hydrogenation.
gram preparation of (R)-1.5 A Friedel−Crafts reaction of 4 with
INTRODUCTION
■
ethyl chloroglyoxylate 12 followed by reduction of 5 produced
The world is now facing a huge increase in the number of
people with diabetes, in particular, type 2 diabetes, because of
secondary benzylic alcohol Rac-6 in 75% yield after purification
by crystallization. Subsequently, reduction of Rac-6 with
aging population and rapidly rising numbers of overweight and
obese people.1 Many pharmaceutical companies are focusing
TMSCl and NaI in CH3CN led to the corresponding α-aryl
acetate 7,6 which was treated with oxone to give the crude
their attention on diabetes, and have developed various small-
molecule series as potential therapeutic agents for the treatment
of type 2 diabetes. Among them, small-molecule glucokinase
sulfone. 8 was purified by crystallization (80% yield from Rac-
6), alkylated with 4-THPCH2I (9) in THF and 1,3-dimethyl-
activators (GKAs) represent a new strategy.2 (R)-2-[4-
3,4,5,6-tetrahydro-2(1H)-pyrimidinone (DMPU) using
LHMDS as a base, and saponified with aq NaOH to give the
(Cyclopropylsulfonyl)phenyl]-N-pyrazin-2-yl-3-(tetrahydro-
sodium salt of Rac-2. Rac-2 was isolated as a crystal in 57%
yield through aqueous acid workup followed by repeated
crystallization from toluene, a mixture of i-BuOAc and n-
heptane, and i-BuOAc alone. Subsequently, (R)-2 was obtained
as a crystal in 80% yield by employing the diastereoselective
esterification of a ketene generated from the corresponding acyl
chloride with (R)-pantolactone followed by hydrolysis and
crystallization from i-BuOAc. Finally, the amidation of (R)-2
with 2-aminopyrazine gave the desired active pharmaceutical
ingredient (R)-1. Our detailed investigations of this synthetic
strategy are described below.
Friedel−Crafts Reaction for Benzyl Alcohol. Benzylic
alcohol Rac-6 was synthesized by a direct alkylation of 4 with
ethyl glyoxalate 11 (route e) or by a Friedel−Crafts reaction
with ethyl chloroglyoxylate 12 followed by reduction (route f),
shown in Table 1. The reaction with 11 provided a mixture of
p-adduct Rac-6 and o-adduct Rac-13 in the presence of AlCl3.
Crystallization from the mixture of toluene and n-heptane after
aqueous workup gave purified Rac-6 in 73% yield with a small
amount of o-adduct and other impurities. This direct route
using 11 as a reactant is favorable for the production of Rac-6;
however, bulk availability and physical properties of 11 were
limited.7 Therefore, we ultimately chose a two-step sequence
with 12 as the reactant for the synthesis of Rac-6 in the first
2H-pyran-4-yl) propanamide (R)-1 has been regarded as a
potent GKA.3 For use in nonclinical studies and clinical trials,
Prosidion developed a scalable methodology using the
enantioselective hydrogenation of acrylic acid (E)-3 as a key
reaction for the construction of stereochemistry at the α-carbon
center to give the key intermediate (R)-2, followed by
amidation with 2-aminopyrazine, giving (R)-1 (Scheme 1,
route a).3a,b According to their reaction conditions, a high
pressure (>700 psig) of hydrogen was required to promote the
reaction. Lilly successfully reduced the hydrogen pressure to 70
psig using Et3N as an additive, leading to improved reduction
efficiency.3d We have reported another approach to (R)-2 that
employs the esterification of racemic carboxylic acid Rac-2 via
ketene formation followed by stereoselective protonation
(route b).3c,f The intermediate Rac-2 was easily synthesized
by hydrogenation of (E)-3 (route c); however, this route was
not suitable for production at an industrial scale because it
required more reaction steps than route a. Therefore, we
reconstructed another synthetic route for Rac-2 based on
Merck’s synthetic strategy (route d).4 Herein, we describe our
process research and development for a scalable synthesis of
(R)-2 via ketene desymmetrization.
RESULTS AND DISCUSSIONS
■
Scheme 2 illustrates the manufacturing results of our first
campaign. At first, we selected readily available cyclo-
propylphenylsulfide 4 as a starting material for the multikilo-
Received: December 11, 2013
Published: February 13, 2014
© 2014 American Chemical Society
437
dx.doi.org/10.1021/op400354g | Org. Process Res. Dev. 2014, 18, 437−445

Organic Process Research & Development
Article
Scheme 1. Synthetic strategies for (R)-1 via (R)-2 as a key intermediate
Scheme 2. First campaign for (R)-1 on a multikilogram scale
campaign. Treatment of 4 with 12 in the presence of AlCl3 gave
the corresponding ketone 5 with almost exclusive para
selectivity. This was followed by reduction with NaBH(OAc)3
to give Rac-6 in quantitative yield. PhCl as a solvent was also
reacted with acyl chloride 12 to give the corresponding
chlorobenzene adducts (5%−15% relative HPLC area %);
however, the amount of these adducts was easily reduced below
detectable limits by crystallization in the next step. Unfortu-
nately, use of less expensive NaBH4 as a reductant led to poor
yield because of side reactions. The two-step sequence was
reproducible in laboratory scale (up to 1.2 kg) to give the
desired Rac-6 in 85% crystallized yield; however, the yield
decreased to 75% on a 27-kg scale. This unexpected decrease of
yield could probably be attributed to the instability of 12 in the
presence of AlCl3 and PhCl. The prolonged stirring time of 12
and AlCl3 in PhCl led to a poor HPLC profile of Rac-6 in 1 h
after adding 4 to the reaction mixture [84.1% relative HPLC
area % (stirring time: 1 h) → 46.5% relative HPLC area %
(stirring time: 3 h)], because the degradation and side reactions
would be promoted. This scale-up problem was resolved by
changing the addition order of reagents, i.e., 12 was added last
to the reaction mixture of 4 and AlCl3 in PhCl to give the
reaction mixture containing Rac-6 with the best HPLC profile
in 1 h (92.5% relative HPLC area %).
Deoxygenation of α-Arylacetate. For the multikilogram
preparation of 7, a screen of acid-induced conditions in the
presence of Et3SiH for high conversion of Rac-6 incorporating
a variety of solvents and acids was undertaken (Table 2). Use of
BF3·OEt2 as an acid could not complete the reaction in any
solvent even over a long duration (entries 1−3).8,9 Unfortu-
nately, an attempt with Brønsted acid TFA failed during the
reduction (entry 4).10 Only utilization of TiCl4 as an acid gave
438
dx.doi.org/10.1021/op400354g | Org. Process Res. Dev. 2014, 18, 437−445

Organic Process Research & Development
Article
investigated; however, these led to low yields because of the
formation of multiple byproducts arising from competing
pathways (4-THPCH2OMs: 57% assay yield, 4-THPCH2OTs:
33% assay yield, and 4-THPCH2OTf: 59% assay yield).15 9 was
prepared from the corresponding alcohol in two reactions
without special purification, such as distillation or column
chromatography. DMPU as an additive enhanced the reactivity
of lithium mixed aggregates (ratio of alkylation in 1 h after
adding 9: 15:81 → 4:87) (Scheme 3)16 and accelerated the
Table 1. Selective synthesis of 4
Scheme 3. Alkylation of 8 with 9 via lithiation followed by
hydrolysis
a
reactant
(X)
temp.
(°C)
selectivity
yield
a
,
b
entry
solvent
Rac-6:Rac-13
(%)
1
2
3
4
5
11 (H)
11 (H)
11 (H)
12 (Cl)
12 (Cl)
CH2Cl2
PhCl
30
30
10
10
10
25:1
11:1
26:1
73:1
78
c
80 (73)
78
PhCl
c
c
CH2Cl2
90 (87)
95 (86)
PhCl
199:1
a
b
c
Determined by HPLC. Assay yield of Rac-6. Crystallized yield
based on 4.
Table 2. Acid-induced deoxygenation of Rac-6 with Et3SiH
hydrolysis of 14 (conversion in 1 h: 42% → 90%). Dialkylation,
as a result of additional lithiation and alkylation of 14, was
observed; however, the corresponding dialkylated ester 15 was
effectively removed from the product-rich basic aqueous stream
after hydrolysis by washing with toluene because it was not
hydrolyzed under this basic condition.
a
b
entry
solvent
acid
time (h)
conv. (%)
yield (%)
1
2
3
4
5
CH2Cl2
PhCl
BF3·OEt2
BF3·OEt2
BF3·OEt2
TFA
22
22
22
19
6
87
88
76
68
11
<1
93
CH3CN
CH2Cl2
CH2Cl2
89
Under the optimized conditions on a laboratory scale, the
alkylated acid with desired purity was isolated as a crystal from
the mixture of i-BuOAc and n-heptane (standard isolated yield:
75%−80%). However, the reaction mixture on the 29-kg scale
failed to crystallize and purify to the desired quality. Through
detailed investigations of the problematic factor, we discovered
that contamination with a small amount of sulfur influenced the
processing and profiles of this reaction.17 8 was obtained by
oxidation of 7 using oxone as an oxidant. Oxone (2KHSO5·
KHSO4·K2SO4) is a weakly acidic triple salt. We chose Na2S2O3
to quench the excess peroxides remaining in the resulting
solution after removing the insoluble components of oxone
from the reaction mixture by filtration. Under acidic conditions,
Na2S2O3 decomposed to elemental sulfur, i.e., on pilot scale,
the prolongation of stirring and settling time would produce a
large amount of sulfur as a byproduct. The participation of
sulfur would affect the alkylation step in the presence of
LHMDS to give various unexpected impurities. One impurity
was isolated from the reaction mixture and identified as sulfide
adduct 16 by spectroscopic analysis such as NMR, IR, MS, and
elemental analysis.18 The formation of 16 was detected in the
presence of sulfur (0.2 equiv) as an additive to the usual
reaction mixture (1.6% relative HPLC area %). Unfortunately,
the contamination of Rac-2 with a small amount of impurity 16
proved problematic in the downstream pharmaceutical
ingredient; therefore, in the first campaign, Rac-2 was
rigorously purified by repeated crystallization from the mixture
<5
TiCl4
b
>95
a
Determined by HPLC. Isolated yield of 7 based on Rac-6.
complete and clean deoxygenation of Rac-6 in a shorter time
(entry 5).11 However, abrupt exothermal behavior was
observed with the addition of TiCl4. Therefore, Lewis acid-
induced deoxygenation with Et3SiH was discontinued.
Ultimately, for the effective reduction of Rac-6 on a
multikilogram scale, we employed TMSCl−NaI−CH3CN
conditions,12 even though these conditions were accompanied
by slight hydrolysis of ester (∼10%).13 Through a detailed
optimization of this condition, high conversion (>99%) was
achieved by using >3 equiv of both TMSCl and NaI with Rac-6
in CH3CN at 35−50 °C. 7 was easily converted to 8 using
oxone as an oxidant in acetone and THF.14 These reactions
succeeded in 34-kg scale to give 8 in 80% yield (standard
isolated yield on gram scale: 76%−87%). However, the use of
Na2S2O3 as a quenching agent was not acceptable for the next
alkylation on a multikilogram scale, as shown below.
Alkylation for Racemic Carboxylic Acid. The alkylated
acid Rac-2 was obtained by treatment of 8 with LHMDS,
followed by alkylation with 4-THPCH2I (9) and subsequent
hydrolysis of ester 14 under basic conditions (85% assay
yield).15 Other alkylating analogues of 9 were briefly
439
dx.doi.org/10.1021/op400354g | Org. Process Res. Dev. 2014, 18, 437−445

Organic Process Research & Development
Article
a
of i-BuOAc and n-heptane to reduce the residue level in Rac-2
crystal to <0.05%. We investigated sodium bisulfite (NaHSO3)
as an alternative reductant and found that it did not lead to the
formation of sulfur even in the presence of additional KHSO4
over a long duration, and gave the desired 8 free from sulfur
(Table 3, entry 4). Consequently, we will change the quenching
agent for the previous oxidation step in the next large-scale
campaign.
Table 4. Effect of amines on diastereoselectivity of (R,R)-10
Table 3. Formation of impurity 16
b
c
entry
amine
yield (%)
dr
1
Et3N
81
89
86
75
75
73
91:9
85:15
90:10
91:9
92:8
89:11
−
2
Me2NEt
3
N-methylmorpholine
(n-Pr)3N
(c-Hex)2NMe
(n-Bu)3N
(i-Pr)2NEt
PMP
4
5
6
d
7
<1
d
8
<1
−
9
TMP
82
93:7
e
10
none
0
−
a
An amine was added to a solution of Rac-17 at 25 °C. After stirring
Rac-2
for 0.5 h, a solution of (R)-pantolactone was added to the reaction
mixture dropwise. After stirring for 2−3 h, the resulting mixture was
operation
yield
b
purity
c
16
a
c
b
entry
scale
reductant
time
(%)
(%)
(%)
purified by extraction and column chromatography. Isolated yield
c
d
d
based on Rac-2. Determined by HPLC. A slight amount of (R,R)-10
1
2
>25 kg Na2S2O3
2 h
65
97.0
98.6
95.4
98.6
0.51
<0.05
0.80
e
was detected in HPLC. Formation of (R,R)-10 was not detected in
5 g
5 g
5 g
Na2S2O3
Na2S2O3
NaHSO3
<15 min
24 h
79
74
78
HPLC. c-Hex = cyclohexyl, PMP = 1,2,2,6,6-pentamethylpiperidine,
TMP = 2,2,6,6-tetramethylpiperidine.
e
3
e
4
24 h
<0.05
a
Total time for stirring and settling in the presence of a reductant.
b
c
d
Crystallized yield based on 8. Relative HPLC area %. Assay yield
e
determined by HPLC. KHSO4 (0.87 w/w) was added to the filtrate
before adding the reductant.
Ketene Desymmetrization Followed by Hydrolysis for
Non-Racemic Carboxylic Acid. Rac-2 was effectively
converted into (R)-2 via diastereoselective esterification of
ketene in situ generated from the corresponding acyl chloride
Rac-17 with (R)-pantolactone followed by hydrolysis of the
ester.3c We have previously reported that amines strongly affect
the conversion of Rac-17 to the activated species, such as a
ketene, and an acyl ammonium salt;3f thereby, the diastereo-
meric ratio (dr) of the resulting ester (R,R)-10 was highly
attributable to the size of the alkylamines (Table 4). Bulkier
tertiary trialkylamines tend to prefer the path via the ketene
rather than the path via the acyl ammonium salt (entries 1−6);
however, bulkiness was likely limited. The bulkiest amines, such
as (i-Pr)2NEt and 1,2,2,6,6-pentamethylpiperidine (PMP), gave
only a small amount of esters because their bulky chains
disturbed the approach of the amine to a hydrogen atom on the
α-carbon of acyl chloride (entries 7, 8). The less bulky
secondary amine 2,2,6,6-tetramethylpiperidine (TMP) gave the
ester with the highest selectivity because a nonselective path via
an acyl ammonium salt led to the formation of an amide (∼4%,
entry 9). Finally, we selected the most inexpensive Et3N as a
base because its enantiomeric ratio (er) after crystallization was
almost the same compared to N,N-dicyclohexylmethylamine
((c-Hex)2NMe), which was the most effective amine in terms
of diastereoselectivity.
Figure 1. Monitoring the ketene intermediate by React IR.
ketene complexes, thereby resulting in high yield (>90%).
Lower temperature in the addition step improved the
diastereoselectivity of ester (R,R)-10 (up to 97:3 dr); however,
(R)-pantolactone was sparingly soluble in toluene even at room
temperature, and the reaction rate was determined by solution
rate unless an (R)-pantolactone crystal was added without
dissolving in a solvent. Therefore, we used THF as a cosolvent
to dissolve (R)-pantolactone and performed the dropwise
addition of this solution (room temperature) to the solution of
18 in toluene at <−70 °C. (R,R)-10 was hydrolyzed at 0−10
°C with less epimerization/racemization by aq H2O2 and LiOH
(96:4 er);20 however, this single treatment for hydrolysis led to
a low crystallized yield (up to 65%). For (R,R)-10, this
The ketene-forming reaction was monitored by React IR
(Figure 1),19 which revealed that the treatment with Et3N as a
base below 0 °C successfully suppressed the degradation of
440
dx.doi.org/10.1021/op400354g | Org. Process Res. Dev. 2014, 18, 437−445

Organic Process Research & Development
Article
Scheme 4. Diastereoselective esterification of in situ ketene followed by hydrolysis
peroxide condition produces side esters such as methyl esters
and ring-opening esters, which were removed by extraction
because this condition alone cannot hydrolyze them. Therefore,
the additional treatment of the resulting mixture with LiOH at
40 °C is a key requirement for high conversion of ester (R,R)-
10 to the corresponding carboxylate,21 although the er of the
crude ester was slightly decreased (94:6 er). Finally, acid-
ification by aq HCl followed by crystallization from i-BuOAc
gave the desired enantio-enriched carboxylic acid (R)-2 as a
colorless crystal with >99% purity and >199:1 er in 78%−80%
yield (Scheme 4). These optimized conditions on laboratory
scale were successfully applied to manufacture on a 20-kg scale,
giving (R)-2 with the desired quality in 80% yield.
mol) in PhCl (60.35 kg) was added to it dropwise over a period
of 2.5 h at −15−0 °C. [The procedure on a pilot scale was not
optimized. Our optimized procedure is as follows: A reactor
was charged with PhCl and AlCl3, and cooled to −15−0 °C.
The temperature was maintained between −15 and 0 °C, and 4
was added to the mixture dropwise. After the addition, the
mixture was stirred for 1 h. A solution of 12 in PhCl was added
to it dropwise at −15−0 °C.] After addition, the mixture was
warmed to 10−20 °C and stirred for 2 h. The resulting mixture
was transferred to a second reactor that contained the purified
water (326.00 kg) over a period of 2.5 h at 5−30 °C. The first
reactor and the line were washed with PhCl (30.20 kg). After
quenching the reaction mixture, the aqueous layer was
separated and extracted with EtOAc (122.90 kg), while the
first organic layer was concentrated under vacuum at 50−60 °C
until no further distillate could be collected, and dissolved in
EtOAc (172.05 kg). The organic layers were collected, washed
with 24.6% aq NaCl (159.20 kg), 5.1% aq NaHCO3 (139.10
kg), and 24.6% aq NaCl (159.20 kg) and concentrated under
vacuum at 35−45 °C until no further distillate could be
collected. The residue was dissolved in THF (72.45 kg) and
concentrated under vacuum at 35−45 °C until no further
distillate could be collected. The residue 5 was dissolved in
THF (120.90 kg), and a suspension of NaBH(OAc)3 in THF,
which was prepared in another reactor with NaBH4 (13.70 kg;
362.1 mol) and AcOH (87.00 kg; 1449 mol) in THF (133.05
kg), was added to it dropwise over a period of 1.5 h at 25−35
°C. Another reactor and the line were washed with THF (18.15
kg). The mixture was heated to 45−50 °C and stirred for 2 h.
After cooling below 25 °C, the resulting mixture was washed
with a 15% aq NaCl (136.40 kg) and concentrated under
vacuum at 35−45 °C until no further distillate could be
collected. The residue was dissolved in EtOAc (245.30 kg),
washed with a mixture of NaCl (6.80 kg) and NaHCO3 (13.60
kg) in purified water (116.00 kg) twice and a 15% aq NaCl
(136.40 kg), and concentrated under vacuum at 45−55 °C until
no further distillate could be collected. The residue was
CONCLUSION
■
We developed a scalable route for multikilogram preparation of
nonracemic carboxylic acid (R)-2 with the desired quality. Our
process utilizes desymmetrization of a ketene in situ generated
from the corresponding racemic carboxylic acid Rac-2 with (R)-
pantolactone. This key step has been successfully scaled up to
20 kg, which demonstrates that this synthetic approach is
comparable to the previously reported approach via enantio-
selective hydrogenation as a key step.22
EXPERIMENTAL SECTION
■
General. All air- and moisture-sensitive manipulations were
performed under nitrogen atmosphere. All substrates, reagents,
and solvents were used as received from suppliers without
further purification.
Ethyl [4-(cyclopropylthio)phenyl](hydroxy)acetate
(Rac-6) (CAS No.1196118-13-4). A reactor was charged
with PhCl (241.05 kg) and AlCl3 (36.20 kg; 271.5 mol), and
cooled to −15−0 °C. The temperature was maintained
between −15 and 0 °C, and ethyl chloroglyoxylate 12 (37.10
kg; 271.7 mol) was added to the mixture dropwise over a
period of 1.5 h. After the addition, the mixture was stirred for 1
h. A solution of cyclopropylphenylsulfide 4 (27.20 kg; 181.0
441
dx.doi.org/10.1021/op400354g | Org. Process Res. Dev. 2014, 18, 437−445

Organic Process Research & Development
Article
dissolved in toluene (47.10 kg) and concentrated under
vacuum at 45−55 °C until no further distillate could be
collected. The residue was diluted in toluene (71.50 kg) and
heated to 50−60 °C. After dissolution, n-heptane (93.05 kg)
was added to the solution and the resulting solution was cooled
to 25−35 °C slowly over a period of 2 h. After crystallization,
the resulting slurry was aged for 1 h at 25−35 °C, and cooled to
5−15 °C slowly over a period of 2 h. n-Heptane (93.05 kg) was
added to the slurry dropwise over a period of 1 h. After aging
for 5 h at 5−15 °C, the resulting slurry was filtered by
centrifuge. The filter cake was washed with a cold mixture of
toluene (12.00 kg) and n-heptane (28.00 kg). The cake was
dried using conical dryer under vacuum at 45−55 °C for 13.5 h
to give Rac-6 as a pale-yellow crystal (34.15 kg; 75% yield;
mixture of EtOAc (21.60 kg) and n-heptane (32.55 kg). The
cake was dried using conical dryer under vacuum at 45−55 °C
for 6.5 h to give 8 as a pale-yellow crystal (28.90 kg; 80% yield;
1
3
99.89% HPLC area %). H NMR (CDCl3): δ 7.86 (d, JHH
=
=
8.4 Hz, 2H), 7.48 (d, 3JHH = 8.7 Hz, 2H), 4.18 (quartet, 3JHH
7.2 Hz, 2H), 3.71 (s, 2H), 2.49−2.42 (m, 1H), 1.38−1.33 (m,
3
2H), 1.27 (t, JHH = 7.2 Hz, 3H), 1.06−1.00 (m, 2H). 13C
NMR (DMSO-d6): δ 170.4, 140.4, 139.1, 130.4, 127.2, 60.5,
39.8, 32.0, 14.0, 5.3. Anal. Calcd for C13H16O4S: C, 58.19; H,
6.01; S, 11.95. Found: C, 58.11; H, 5.92; S, 11.89 (laboratory-
scale sample), C, 57.78; H, 5.92; S, 13.61 (pilot-scale sample).
4-(Iodomethyl)tetrahydro-2H-pyran (9) (CAS
No.101691−94−5). A reactor was charged with tetrahydro-
2H-pyran-4-ylmethanol (22.75 kg; 195.9 mol), toluene (137.00
kg) and Et3N (22.70 kg; 224.3 mol), and it was cooled to 0−10
°C. A solution of methanesulfonyl chloride (24.60 kg; 214.8
mol) in toluene (59.00 kg) was added to the mixture dropwise
over a period of 0.5 h. After the addition, the mixture was
warmed to 20−30 °C. After stirring for 1.5 h, purified water
(114.00 kg) was added to the resulting mixture. The aqueous
layer was separated, and extracted with toluene (58.80 kg). The
organic layers were collected and concentrated under vacuum
at 45−55 °C until no further distillate could be collected. The
residue was dissolved in acetone (53.75 kg), NaI (73.20 kg;
488.4 mol) was added to it portionwise over a period of 45 min,
and warmed to 55−65 °C. After stirring for 5 h, purified water
(182.00 kg) was added to the resulting mixture. The resulting
solution was concentrated under vacuum at 45−55 °C until no
further distillate could be collected. The residue was dissolved
in EtOAc (163.80 kg), washed with 17.9% aq Na2S2O3 (101.25
kg), dried over MgSO4 (2.30 kg), and filtered by centrifuge.
The filter cake was washed with EtOAc (27.30 kg). The filtrate
was concentrated under vacuum at 25−35 °C until no further
distillate could be collected. The residue 9 was taken directly
into the next step without further purification (93% assay
yield). 1H NMR (CDCl3): δ 4.00−3.94 (m, 2H), 3.37 (td, 3JHH
= 11.8 Hz, 2.0 Hz, 2H), 3.10 (d, 3JHH = 6.7 Hz, 2H), 1.82−1.75
(m, 2H), 1.75−1.66 (m, 1H), 1.36−1.25 (m, 2H). 13C NMR
(CDCl3): δ 67.6, 37.7, 33.6, 13.9. Anal. Calcd for C6H11IO: C,
31.88; H, 4.90. Found: C, 31.76; H, 4.72.
1
3
97.20% HPLC area %). H NMR (CDCl3): δ 7.36 (d, JHH
=
3
3
8.7 Hz, 2H), 7.33 (d, JHH = 8.7 Hz, 2H), 5.11 (d, JHH = 4.6
3
Hz, 1H), 4.32−4.13 (m, 2H), 3.43 (t, JHH = 5.1 Hz, 1H),
2.21−2.14 (m, 1H), 1.24 (t, 3JHH = 7.2 Hz, 3H), 1.11−1.04 (m,
2H), 0.72−0.66 (m, 2H). 13C NMR (CDCl3): δ 172.4, 137.9,
136.4, 127.1, 125.8, 72.0, 60.3, 13.9, 11.3, 8.2, 8.1. Anal. Calcd
for C13H16O3S: C, 61.88; H, 6.39; S, 12.71. Found: C, 61.87; H,
6.29; S, 12.53.
Ethyl [4-(cyclopropylsulfonyl)phenyl]acetate (8) (CAS
No. 1058167−40−0). A reactor was charged with CH3CN
(79.85 kg), Rac-6 (33.95 kg; 134.5 mol) and NaI (64.50 kg;
430.3 mol), and it was warmed to 30−35 °C. TMSCl (58.40
kg; 537.6 mol) was added to the mixture dropwise over a
period of 0.5 h. After the addition, the mixture was stirred for
11 h. The resulting mixture was quenched with 20% aq
NaHSO3 (81.30 kg). The resulting solution was extracted with
EtOAc (122.00 kg, 122.10 kg) twice. The organic layers were
collected, washed with a 10% aq NaCl (169.00 kg), 10% aq
Na2CO3 (169.00 kg), 10% aq Na2CO3 (169.00 kg), and 10% aq
NaCl (169.00 kg), and concentrated under vacuum at 45−55
°C until no further distillate could be collected. The residue 7
was dissolved in acetone (53.75 kg) and THF (151.10 kg), and
a 23.4% aq oxone (531.00 kg; 404.2 mol) was added to it
dropwise over a period of 3.5 h at 15−25 °C. After stirring for 2
h, EtOAc (245.00 kg) was added to the resulting mixture. The
resulting suspension was filtered by centrifuge, and the filter
cake was washed with EtOAc (61.80 kg). The filtrate was
transferred to a reactor, and 9.5% aq Na2S2O3 (143.30 kg) was
added to it over a period of 45 min at 15−30 °C, and the
resulting mixture was stirred for 40 min. [The procedure on a
pilot scale was not optimized. Our optimized procedure is as
follows: The filtrate was transferred to a reactor, and 9.5% aq
NaHSO3 was added to it at 15−30 °C, and the resulting
mixture was stirred for 0.5−1 h.] After settling for 35 min, the
aqueous layer was separated, and extracted with EtOAc (122.10
kg). The organic layers were collected, washed with a 10% aq
Na2CO3 (135.50 kg) and 10% aq NaCl (135.50 kg), and
concentrated under vacuum at 35−45 °C until no further
distillate could be collected. The residue was dissolved in
toluene (47.10 kg), and concentrated under vacuum at 45−55
°C until no further distillate could be collected. The residue was
diluted in EtOAc (76.55 kg), and heated to 50−70 °C. After
dissolution, the resulting solution was cooled to 40−45 °C
slowly over a period of 1 h. After crystallization, the resulting
slurry was aged for 1.5 h at 40−50 °C, cooled to 5−15 °C
slowly over a period of 3.5 h, and stirred for 3 h. n-Heptane
(116.05 kg) was added to the slurry dropwise over a period of 1
h. After aging for 1 h at 5−15 °C, the resulting slurry was
filtered by centrifuge, and the filter cake was washed with a cold
2-[4-(Cyclopropylsulfonyl)phenyl]-3-(tetrahydro-2H-
pyran-4-yl)propanoic acid (Rac-2) (CAS No. 745052−
93−1). A reactor was charged with 8 (28.70 kg; 107.0 mol) and
THF (128.05 kg) and cooled to −10−0 °C. The temperature
was maintained between −10 and 0 °C, and a 20.5% solution of
LHMDS in THF (93.30 kg; 114.3 mol) was added to the
mixture dropwise over a period of 1.5 h. After the addition, the
mixture was stirred for 1 h. A solution of 9 (33.90 kg; 139.5
mol) in THF (51.10 kg) and DMPU (20.55 kg; 160.3 mol) was
added to the mixture dropwise over a period of 1 h at −10−0
°C. After addition, the mixture was warmed to 15−25 °C and
stirred for 2 h. An 8% aqueous solution of NaOH (100.15 kg)
was added to the resulting mixture, and the resulting solution
was warmed to 45−55 °C. After stirring for 4 h, the organic
layer was separated and extracted with 10% aq NaCl (143.40
kg). The aqueous layers were collected, acidified by 35% aq
HCl (34.20 kg), and extracted with EtOAc (131.25 kg, 130.05
kg) twice. The organic layers were collected, washed with a
10% aq NaHSO3 (143.40 kg), dried over MgSO4 (2.85 kg), and
filtered on activated charcoal (1.45 kg). The filter cake was
washed with EtOAc (80.80 kg). The filtrate was transferred to a
reactor and concentrated under vacuum at 45−55 °C until no
further distillate could be collected. The residue was dissolved
442
dx.doi.org/10.1021/op400354g | Org. Process Res. Dev. 2014, 18, 437−445

Organic Process Research & Development
Article
in i-BuOAc (99.40 kg) and concentrated under vacuum at 45−
55 °C until no further distillate could be collected. The residue
was diluted in i-BuOAc (124.30 kg) and heated to 90−110 °C.
After dissolution, the resulting solution was cooled to 35−45
°C slowly over a period of 1 h. After crystallization, the slurry
was aged for 1 h at 35−45 °C, warmed to 73−83 °C, and
cooled to 5−15 °C slowly over a period of 4.5 h. n-Heptane
(98.25 kg) was added to the slurry dropwise over a period of
3.5 h. (After the addition, the slurry was emulsified
immediately.) The emulsion was concentrated under vacuum
at 45−55 °C until no further distillate could be collected.
Toluene (124.10 kg) was added to the residue in the reactor,
and the mixture was warmed to 85−95 °C. After dissolution,
the resulting solution was cooled to 45 °C over a period of 4 h
and stirred for 0.5 h at 40−45 °C. After cooling to 15 °C, the
resulting slurry was aged at 10−15 °C for 2 h, filtered by
centrifuge. The filter cake was washed with cold toluene (49.70
kg). The cake was dried using a conical dryer under vacuum at
45−55 °C for 4 h to give crude Rac-2 as a pale-yellow crystal
containing toluene with a weight of 32.25 kg (65% assay yield,
96.97% HPLC area %). A reactor was charged with i-BuOAc
(49.85 kg) and the dry crude Rac-2 (32.25 kg) and heated to
110 °C. After dissolution, the solution was cooled to 45 °C
slowly over a period of 4.5 h. After the resulting slurry was aged
at 40−45 °C for 0.5 h, n-heptane (39.30 kg) was added to it
dropwise over 1 h, and the resulting slurry was cooled to 15 °C
slowly over a period of 2 h. After stirring at 10−15 °C for 1 h,
the resulting slurry was filtered by centrifuge, and the filter cake
was washed with a cold mixture of i-BuOAc (24.90 kg) and n-
heptane (19.60 kg). The wet cake (28.70 kg) was transferred to
a reactor that contained i-BuOAc (50.05 kg), and the mixture
was heated to 110 °C. After dissolution, the solution was
cooled to 45 °C slowly over a period of 3 h. The resulting slurry
was aged at 40−45 °C for 0.5 h, n-heptane (39.30 kg) was
added to it dropwise over 1 h, and the resulting slurry was then
cooled to 15 °C slowly over a period of 1 h. After stirring at
40−45 °C for 1 h, the resulting slurry was filtered by centrifuge,
and the filter cake was washed with a cold mixture of i-BuOAc
(24.90 kg) and n-heptane (19.60 kg). The wet cake (27.25 kg)
was transferred to a reactor that contained i-BuOAc (49.80 kg),
and the mixture was heated to 110 °C. After dissolution, the
solution was cooled to 45 °C slowly over a period of 4.5 h.
After the resulting slurry was aged at 40−45 °C for 0.5 h, n-
heptane (39.35 kg) was added to it dropwise over 1 h, and the
resulting slurry was cooled to 15 °C slowly over a period of 2 h.
After stirring at 10−15 °C for 1.5 h, the resulting slurry was
filtered by centrifuge, and the filter cake was washed with a cold
mixture of i-BuOAc (24.95 kg) and n-heptane (19.65 kg). The
wet cake (26.50 kg) was transferred to a reactor that contained
i-BuOAc (49.85 kg), and the mixture was heated to 110 °C.
After dissolution, the solution was cooled to 45 °C slowly over
a period of 3.5 h. The resulting slurry was warmed to 78 °C
slowly over a period of 1 h. After stirring at 78 °C for 0.5 h, the
slurry was cooled to 15 °C slowly over a period of 5.5 h. After
stirring at 10−15 °C for 1.5 h, the resulting slurry was filtered
by centrifuge, and the filter cake was washed with cold i-BuOAc
(24.90 kg). The wet cake (23.95 kg) was transferred to a
reactor that contained i-BuOAc (49.75 kg), and the mixture
was heated to 110 °C. After dissolution, the solution was
cooled to 45 °C slowly over a period of 4 h. The resulting slurry
was warmed to 73 °C slowly over a period of 0.5 h. After
stirring at 73−76 °C for 0.5 h, the slurry was cooled to 15 °C
slowly over a period of 5.5 h. After stirring at 10−15 °C for 1.5
h, the resulting slurry was filtered by centrifuge, and the filter
cake was washed with cold i-BuOAc (24.85 kg). The cake was
dried using a conical dryer under vacuum at 45−55 °C for 4.5 h
to give Rac-2 as a colorless crystal (20.50 kg; 57% yield; 99.94%
1
3
HPLC Area %). Rac-2: H NMR (CDCl3): δ 7.87 (d, JHH
=
3
8.4 Hz, 2H), 7.51 (d, JHH = 8.4 Hz, 2H), 3.96−3.90 (m, 2H),
3
3.80 (t, 3JHH = 7.8 Hz, 1H), 3.30 (tdd, JHH = 11.8 Hz, 5.1 Hz,
3
2.0 Hz, 2H), 2.46 (tt, JHH = 7.9 Hz, 4.9 Hz, 1H), 2.12−2.04
(m, 1H), 1.81−1.72 (m, 1H), 1.64−1.55 (m, 2H), 1.46−1.25
(m, 5H), 1.08−1.01 (m, 2H). 13C NMR (DMSO-d6): δ 174.1,
145.5, 139.1, 128.9, 127.4, 66.8, 47.5, 39.7, 32.5, 32.3, 32.1,
31.9, 5.3. Anal. Calcd for C17H22O5S: C, 60.33; H, 6.55; S, 9.47.
1
Found: C, 60.26; H, 6.52; S, 9.45. 14: H NMR (CDCl3): δ
7.86 (d, 3JHH = 8.4 Hz, 2H), 7.50 (d, 3JHH = 8.0 Hz, 2H), 4.22−
4.06 (m, 2H), 3.97−3.90 (m, 2H), 3.77 (t, 3JHH = 7.8 Hz, 1H),
3.36−3.26 (m, 2H), 2.51−2.43 (m, 1H), 2.12−2.03 (m, 1H),
1.77−1.69 (m, 1H), 1.64−1.54 (m, 3H), 1.45−1.25 (m, 5H),
3
1.23 (t, JHH = 7.2 Hz, 3H), 1.08−1.01 (m, 2H). 13C NMR
(CDCl3): δ 172.8, 144.8, 139.5, 128.7, 127.8, 67.6, 67.5, 61.1,
48.3, 40.3, 32.7, 32.7, 32.6, 32.5, 14.0, 5.8. Anal. Calcd for
C19H26O5S: C, 62.27; H, 7.15; S, 8.75. Found: C, 62.10; H,
7.13; S, 8.84. 15: 1H NMR (CDCl3): δ 7.85 (d, 3JHH = 8.8 Hz,
2H), 7.53 (d, 3JHH = 8.8 Hz, 2H), 4.16 (d, 3JHH = 7.2 Hz, 1H),
4.12 (d, 3JHH = 7.2 Hz, 1H), 3.86−3.78 (m, 4H), 3.30−3.18 (m,
4H), 2.47 (tt, 3JHH = 8.0 Hz, 4.8 Hz, 1H), 2.11 (dd, 2JHH = 14.4
3
2
3
Hz, JHH = 5.6 Hz, 2H), 2.02 (dd, JHH = 14.4 Hz, JHH = 5.6
Hz, 2H), 1.46−1.32 (m, 4H), 1.32−1.17 (m, 11H), 1.08−1.00
(m, 2H). 13C NMR (CDCl3): δ 175.0, 148.8, 138.9, 127.5,
127.3, 67.7, 61.0, 52.9, 43.4, 34.3, 33.9, 32.8, 31.5, 13.8, 5.8.
Anal. Calcd for C25H36O6S: C, 64.63; H, 7.81; S, 6.90. Found:
1
C, 64.62; H, 7.80; S, 6.78. 16: H NMR (CDCl3): δ 7.91 (d,
3JHH = 8.4 Hz, 2H), 7.53 (d, 3JHH = 8.4 Hz, 2H), 5.27 (bs, 1H),
3.98−3.90 (m, 4H), 3.81 (t, 3JHH = 7.8 Hz, 1H), 3.39−3.26 (m,
3
3
4H), 2.63 (dd, JHH = 11.6 Hz, 6.6 Hz, 1H), 2.58 (dd, JHH
=
11.6 Hz, 7.0 Hz, 1H), 2.15−2.06 (m, 1H), 1.98−1.92 (m, 2H),
1.81−1.72 (m, 1H), 1.72−1.56 (m, 5H), 1.50−1.38 (m, 1H),
1.38−1.13 (m, 6H). 13C NMR (CDCl3): δ 176.7, 144.7, 137.1,
129.9, 128.5, 67.6, 67.6, 67.6, 48.2, 45.1, 40.3, 40.0, 34.4, 32.7,
32.6, 32.5, 32.3, 32.3, 30.9, 17.1, 17.0. Anal. Calcd for
C23H32O6S2: C, 58.95; H, 6.88; S, 13.68. Found: C, 58.95; H,
6.83; S, 13.52.
(R)-2-[4-(Cyclopropylsulfonyl)phenyl]-3-(tetrahydro-
2H-pyran-4-yl)propanoic acid ((R)-2) (CAS No. 745053-
49-0). A reactor was charged with toluene (87.85 kg), Rac-2
(20.25 kg; 59.84 mol) and Et3N (182.0 g; 1.798 mol), and it
was warmed to 35−40 °C. SOCl2 (9.25 kg; 77.8 mol) was
added to the reaction mixture under a scrubber system. After
stirring at 40−45 °C for 3 h, the reaction mixture was
concentrated under vacuum at 45−46 °C until no further
distillate could be collected. The residue was dissolved in
toluene (35.35 kg), and the solution was concentrated under
vacuum at 46 °C until no further distillate could be collected.
The residue was dissolved in toluene (141.15 kg) and cooled to
−6 °C. Et3N (12.14 kg; 120.0 mol) was added to the solution
and stirred at −10−0 °C for 12 h. After cooling to −77 to −74
°C, a solution of (R)-pantolactone (9.40 kg; 72.2 mol) in THF
(73.25 kg) was added dropwise to the yellow solution over a
period of 2.5 h, and the resulting mixture was stirred for 2 h.
After warming to −5 °C, citric acid monohydrate (22.20 kg)
and purified water (40.60 kg) were added to the reaction
solution. After warming to 20 °C, the resulting solution was
separated. The organic layer was washed with purified water
(40.60 kg) and 10% aq NaCl (81.60 kg), and concentrated
443
dx.doi.org/10.1021/op400354g | Org. Process Res. Dev. 2014, 18, 437−445

Organic Process Research & Development
Article
under vacuum at 48−51 °C until no further distillate could be
collected. The residue was dissolved in MeOH (48.25 kg) and
concentrated under vacuum at 50−51 °C until no further
distillate could be collected, to give the residue (R,R)-10.
Another reactor was charged with 35.4% H2O2 (8.05 kg; 83.8
mol) and purified water (40.65 kg), and was cooled to 5 °C. A
solution of lithium hydroxide monohydrate (3.00 kg, 71.4 mol)
in purified water (20.30 kg) was added dropwise. The resulting
lithium peroxide solution was transferred to a second reactor
containing a solution of (R,R)-10 in MeOH (96.50 kg)
dropwise over 1 h at 0−3 °C. After stirring at −1−3 °C for 1.5
h, a solution of Na2SO3 (18.30 kg; 145.2 mol) in purified water
(163.50 kg) was added dropwise to the resulting solution over a
period of 1.5 h at 8−9 °C. After warming to 35 °C, lithium
hydroxide monohydrate (3.80 kg; 90.6 mol) was added to the
solution. After stirring at 38−40 °C for 2.5 h, the resulting
solution was washed with toluene (87.90 kg), and acidified by
35% aq HCl (27.50 kg). The aqueous layer was extracted with
EtOAc (91.35 kg, 91.35 kg, 91.25 kg) three times, and the
collected organic layers were washed with 10% aq NaCl (81.75
kg), dried over MgSO4 (4.05 kg), and filtered. After the filtrate
was concentrated at 50 °C under vacuum, the residue was
dissolved in i-BuOAc (70.30 kg), and the solution was
concentrated at 50−51 °C under vacuum. The residue was
dissolved in i-BuOAc (87.95 kg) at 110 °C. The solution was
cooled slowly to 85 °C for crystallization. After crystallization,
the suspension was stirred at 85−81 °C for 1.5 h, and cooled
slowly to 15 °C over a period of 8 h. After stirring at 15−11 °C
for 3 h, a crystal was filtered, washed with i-BuOAc (35.20 kg),
and dried at 46−51 °C for 6 h under vacuum to give (R)-2 as a
colorless crystal (16.20 kg; 80.0% yield from Rac-2, 99.84:0.16
REFERENCES
■
(1) Danaei, G.; Finucane, M. M.; Lu, Y.; Singh, G. M.; Cowan, M. J.;
Paciorek, C. J.; Lin, J. K.; Farzadfar, F.; Khang, Y.-H.; Stevens, G. A;
Rao, M.; Ali, M. K; Riley, L. M; Robinson, C. A; Ezzati, M. Lancet
2011, 378, 31.
(2) (a) Moller, D. E. Nature 2001, 414, 821. (b) Grimsby, J.; Sarabu,
R.; Corbett, W. L.; Haynes, N.-E.; Bizzarro, F. T.; Coffey, J. W.;
Guertin, K. R.; Hilliard, D. W.; Kester, R. F.; Mahaney, P. E.; Marcus,
L.; Qi, L.; Spence, C. L.; Tengi, J.; Magnuson, M. A.; Chu, C. A.;
Dvorozniak, M. T.; Matschinsky, F. M.; Grippo, J. F. Science 2003, 301,
370. (c) Sarabu, R.; Grimsby, J. Curr. Opin. Drug Discovery Dev. 2005,
8, 631. (d) Bertram, L. S.; Black, D.; Briner, P. H.; Chatfield, R.;
Cooke, A.; Fyfe, M. C. T.; Murray, P. J.; Naud, F.; Nawano, M.;
Procter, M.; Rakipovski, G.; Rasamison, C. M.; Reynet, C.; Schofield,
K. L.; Shah, V. K.; Spindler, F.; Taylor, A.; Turton, R.; Williams, G. M.;
Wong-Kai-In, P.; Yasuda, K. J. Med. Chem. 2008, 51, 4340. (e) Sarabu,
R.; Berthel, S. J; Kester, R. F; Tilly, J. W Expert Opin. Ther. Pat. 2008,
7, 759. (f) Coghlan, M.; Leighton, B. Expert. Opin. Investig. Drugs
2008, 2, 145. (g) Matschinsky, F. M. Nat. Rev. Drug Discovery 2009, 8,
399. (h) Li, F.; Zhu, Q.; Zhang, Y.; Feng, Y.; Leng, Y.; Zhang, A.
Bioorg. Med. Chem. 2010, 18, 3875.
(3) (a) Fyfe, M. C. T.; Gardner, L. S.; Nawano, M.; Procter, M. J.;
Rasamison, C. M.; Schofield, K. L.; Shah, V. K.; Yasuda, K. PCT Int.
Appl. WO/2004/072031 A2 20040826, 2004. (b) Briner, P. H.; Fyfe,
M. C. T.; Madeley, J. P.; Murray, P. J.; Procter, M. J.; Spindler, F. PCT
Int. Appl. WO/2006/016178 A1 20060216, 2006. (c) Sugawara, K.;
Toshikawa, S. PCT Int. Appl. WO/2009/139438 A1 20091119, 2009.
(d) Magnus, N. A.; Braden, T. M.; Buser, J. Y.; DeBaillie, A. C.; Heath,
P. C.; Ley, C. P.; Remacle, J. R.; Varie, D. L.; Wilson, T. M. Org.
Process Res. Dev. 2012, 16, 830. (e) DeBaillie, A. C.; Magnus, N. A.;
Laurila, M. E.; Wepsiec, J. P.; Ruble, J. C.; Petkus, J. J.; Vaid, R. K.;
Niemeier, J. K.; Mick, J. F.; Gunter, T. Z. Org. Process Res. Dev. 2012,
16, 1538. (f) Yamagami, T.; Hatsuda, M.; Utsugi, M.; Kobayashi, R.;
Moritani, Y. Eur. J. Org. Chem. 2013, 33, 7467.
(4) (a) Larsen, R. D.; Corley, E. G.; Davis, P.; Reider, P. J.;
Grabowski, E. J. J. J. Am. Chem. Soc. 1989, 111, 7650. (b) Chen, C.;
Dagneau, P.; Grabowski, E. J. J.; Oballa, R.; O’Shea, P.; Prasit, P.;
Robichaud, J.; Tillyer, R.; Wang, X. J. Org. Chem. 2003, 68, 2633.
(5) (a) Truce, W. E.; Hollister, K. R.; Lindy, L. B.; Parr, J. E. J. Org.
Chem. 1968, 33, 43. (b) Masson, E.; Leroux, F. Helv. Chim. Acta 2005,
88, 1375.
(6) (a) Olah, G. A.; Narang, S. C. Tetrahedron 1982, 38, 2225.
(b) Sakai, T.; Miyata, K.; Utaka, M.; Takeda, A. Bull. Chem. Soc. Jpn.
1987, 60, 1063. (c) Sakai, T.; Miyata, K.; Tsuboi, S.; Takeda, A.;
Utaka, M.; Torii, S. Bull. Chem. Soc. Jpn. 1989, 62, 3537. (d) Stoner, E.
J.; Cothron, D. A.; Balmer, M. K.; Roden, B. A. Tetrahedron 1995, 51,
11043.
1
3
er). (R)-2: H NMR (CDCl3): δ 7.87 (d, JHH = 8.4 Hz, 2H),
7.51 (d, 3JHH = 8.2 Hz, 2H), 3.97−3.90 (m, 2H), 3.80 (t, 3JHH
=
2
3
7.8 Hz, 1H), 3.30 (tdd, JHH = 11.8 Hz, JHH = 5.1 Hz, 2.0 Hz,
2H), 2.46 (tt, 3JHH = 7.9 Hz, 4.9 Hz, 1H), 2.12−2.04 (m, 1H),
1.81−1.72 (m, 1H), 1.64−1.55 (m, 2H), 1.46−1.25 (m, 5H),
1.08−1.01 (m, 2H). 13C NMR (DMSO-d6): δ 174.1, 145.5,
139.1, 128.9, 127.4, 66.8, 47.5, 39.7, 32.5, 32.3, 32.1, 31.9, 5.3.
Anal. Calcd for C17H22O5S: C, 60.33; H, 6.55; S, 9.47. Found:
C, 60.08; H, 6.50; S, 9.49. [α]20 −53.4 (c 1.06, MeOH).
D
ASSOCIATED CONTENT
■
S
(7) Ethyl glyoxalate is commercially available in its polymeric form in
toluene solution, and a monomer is generated by pyrolysis of a
polymer. The monomer form is so reactive that it polymerizes easily
and reacts readily with water to generate the hydrated form. Ethyl
glyoxalate is therefore distilled just prior to use, after pyrolysis, and is
used under nonaqueous conditions.
* Supporting Information
Spectra of all new compounds, such as intermediate 14,
impurities 15 and 16. This material is available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
(8) Hanaoka, M.; Yoshida, S.; Annen, M.; Mukai, C. Chem. Lett.
1986, 15, 739. (b) Yoda, H.; Kimura, K.; Takabe, K. Synlett 2001, 3,
400.
(9) Other Lewis acids such as AlCl3, FeCl3, and ZnCl2 produced a
sticky substance in the middle of conversion, prohibiting mixture
stirring before the reaction completed. For an example using AlCl3,
see: Jaxa-Chamiec, A.; Shah, V. P.; Kruse, L. I. J. Chem. Soc. Perkin
Trans. I 1989, 1705.
■
Corresponding Author
*yamagami.takafumi@mk.mt-pharma.co.jp
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
(10) Luzzio, F. A.; Wlodarczyk, M. T.; Duveau, D. Y.; Chen, J.
Tetrahedron Lett. 2007, 48, 6704.
■
We are grateful to the API Corporation for manufacturing (R)-
1 on pilot scale. In addition, we thank Mr. Shigeru Kimura, Mr.
Koki Harigaya, Mr. Masanobu Ota, Dr. Hajime Hiramatsu, Dr.
Masanori Hatsuda, Dr. Masayuki Utsugi, and Mr. Ryo
Kobayashi for analytical supports and helpful discussions.
(11) Joshi, R. R.; Narasimhan, N. S. Synthesis 1987, 943.
(12) (a) Sakai, T.; Miyata, K.; Tsuboi, S.; Takeda, A.; Utaka, M.;
Torii, S. Bull. Chem. Soc. Jpn. 1989, 62, 3537. (b) Stoner, E. J.;
Cothron, D. A.; Balmer, M. K.; Roden, B. A. Tetrahedron 1995, 51,
11043.
444
dx.doi.org/10.1021/op400354g | Org. Process Res. Dev. 2014, 18, 437−445

Organic Process Research & Development
Article
(13) (a) Olah, G. A.; Narang, S. C.; Gupta, B. G. B.; Malhotra, R. J.
Org. Chem. 1979, 44, 1247. (b) Olah, G. A.; Narang, S. C. Tetrahedron
1982, 38, 2225.
(14) Oxone (potassium peroxymonosulfate) is a DuPont product
commercially available from Aldrich, and easily oxidizes acetone to
dimethyldioxirane. For oxidation of sulfide, see: Webb, K. S.
Tetrahedron Lett. 1994, 35, 3457.
(15) Assay yields were calculated by isolated yields and relative
HPLC area %.
(16) (a) Sun, X.; Collum, D. B. J. Am. Chem. Soc. 2000, 122, 2452.
(b) Sun, X.; Collum, D. B. J. Am. Chem. Soc. 2000, 122, 2459.
(17) Xiang, Y.; Caron, P.-Y.; Lillie, B. M.; Vadyanathan, R. Org.
Process Res. Dev. 2008, 12, 116.
(18) See the SI for details.
(19) Use of (i-Pr)2NEt as a base gave no absorption of ketene at
2102 cm−1, which demonstrates that the formation of small amounts
of esters was owing to no formation of the ketene.
(20) Evans, D. A.; Britton, T. C.; Ellman, J. A. Tetrahedron Lett. 1987,
28, 6141.
(21) The direct treatment of (R,R)-10 with aq LiOH alone resulted
in 10−20% erosion of the stereocenter at the α-carbon.
(22) Unfortunately, lower temperature (<−70 °C) was not avoidable
at this first campaign. We expect that use of N-benzyl-(R)-pantolactam
as a chiral auxiliary instead of (R)-pantolactone will overcome this
drawback (see, ref 3f); however, our new approach was not
demonstrated on a multikilogram scale due to the termination of
the project.
445
dx.doi.org/10.1021/op400354g | Org. Process Res. Dev. 2014, 18, 437−445