An improved process for pioglitazone and its pharmaceutically acceptable salt

Article abstract of DOI:10.1021/op900131m

An improved process for pioglitazone (1) is described. The process features high-yielding transformations employing inexpensive reagents and recoverable solvents.

Full text of DOI:10.1021/op900131m

Organic Process Research & Development 2009, 13, 1190–1194
Communications to the Editor
An Improved Process for Pioglitazone and Its Pharmaceutically Acceptable Salt†
Lokeswara Rao Madivada,‡ Raghupathi Reddy Anumala,‡ Goverdhan Gilla,‡ Sampath Alla,‡ Kavitha Charagondla,‡
Mukkanti Kagga,§ Apurba Bhattacharya,‡ and Rakeshwar Bandichhor*,‡
InnoVation Plaza, IPD, R&D, Dr. Reddy’s Laboratories Ltd., SurVey Nos. 42, 45, 46, and 54, Bachupally, Qutubullapur, R.R.
District - 500 073, Andhra Pradesh, India, and Institute of Science and Technology, Center for EnVironmental Science, JNT
UniVersity, Kukatpally, Hyderabad - 500 072, Andhra Pradesh, India
Abstract:
An improved process for pioglitazone (1) is described. The process
features high-yielding transformations employing inexpensive
reagents and recoverable solvents.
Introduction
Pioglitazone 1 (Figure 1) is a benzylthiazolidinedione de-
Figure 1. Structure of pioglitazone 1.
rivative approved as a drug for the management of diabetes.
Pioglitazone 1 is found to stimulate peroxisome proliferator-
activated receptor gamma (PPARγ) in order to modulate the
transcription of the insulin sensitive genes that are involved in
glucose and lipid metabolism.1
The synthetic routes for pioglitazone 1, as described in a
previous patent,2 are shown in Scheme 1 where reaction
involves protection of 5-ethyl-2-pyridyl ethanol, 2, with a
p-toluene sulfonyl group to obtain intermediate 3. Subsequently,
this intermediate was subjected to nucleophilic substitution. In
particular, the reaction between intermediate 3 and p-hydroxy-
benzaldehyde 4 in the presence of sodium hydroxide afforded
intermediate 5. The reaction between intermediate 5 and
thiazolidinedione 6, employing Knoevenagel conditions, af-
forded penultimate intermediate 7.
starting material 2 was condensed with p-substituted fluoroben-
zene to obtain intermediates 8 and 9. Raney Ni/HCO2H-
mediated reductive hydrolysis of cyano derivative 9 afforded
aldehyde intermediate 5 which can be converted to title
compound 1 as per the procedure described in Scheme 1a.
Intermediate 8 was subjected to a cascade of reactions to obtain
bromo derivative 10. In this particular strategy, catalytic
hydrogenation, diazotization, bromination, Cu metal insertion
across the aromatic carbon and bromo functionality, followed
by reaction with methyl acrylate was performed to afford 10.
Subsequently, intermediate 10 was further utilized in the
condensation with thiourea and acid-catalyzed hydrolysis to
render desired species 1. Despite the proven potential of these
procedures in Scheme 1 (a and b), there are certain disadvan-
tages: (a) transformation from 2 to 3 was high yielding, but the
enrichment of E2 elimination impurity during the reaction
leading to the cumbersome isolation and purification of
intermediate 5, (b) expensive Pd metal was used in the reduction
of 7 to 1, (c) a plant-unfriendly, pyrophoric non-nucleophilic
base, NaH, is employed in the transformation of 8 or 9 from 2,
(d) a cascade of reactions, involving expensive metal (Pd), toxic
acid (HBr, and diazonium chloride intermediate, was performed
to yield intermediate 10, and (e) use of expensive partially
recoverable solvents, e.g., dioxane, THF and CH2Cl2. A
convergent approach involving condensation of 4 and 6 to
obtain 1 is also precedented.5
Moreover, the reduction of intermediate 7 in the presence
of Pd/C/H2 at 25 °C yielded the desired compound 1. In another
disclosure,3 CoCl2 ·6H2O/NaBH4/dimethyl glyoxime system
was also used for such a reduction to obtain 1. Additionally,
there is a completely different approach (Scheme 1b)4 where
† DRL-IPD Communication number: IPDO-IPM-00169.
* Corresponding author. E-mail: rakeshwarb@drreddys.com. Telephone: +91
4044346000. Fax: +91 4044346285.
‡ Innovation Plaza, IPD, R&D, Dr. Reddy’s Laboratories Ltd.
§ Institute of Science and Technology, Center for Environmental Science,
JNT University.
(1) (a) Le, A.; Pucko, W.; Szelejewski, W. Org. Process Res. DeV. 2004,
8, 157–162. (b) Colca, J. R.; McDonald, W. G.; Waldon, D. J.; Leone,
J. W.; Lull, J. M.; Bannow, C. A.; Lund, E. T.; Mathews, W. R. Am. J.
Physiol. Endocrinol. Metab. 2004, 286, E252-60. (c) Paddock, M. L.;
Wiley, S. E.; Axelrod, H. L.; Cohen, A. E.; Roy, M.; Abresch, E. C.;
Capraro, D.; Murphy, A. N.; Nechushtai, R.; Dixon, J. E.; Jennings,
P. A. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 14342–14347.
(2) (a) Meguro, K.; Fujita, T.; Hatanaka, C.; Ooi, S. U.S. Patent 4,812,570,
1989. (b) Huber, J. E. U.S. Patent 5,585,495, 1996.
Herein, we present our efforts to avoid all of the disadvanta-
geous factors involved in Scheme 1 (a and b) and achieve a
cost-effective, high-yielding, and moderately greener process
for pioglitazone 1 and its pharmaceutically acceptable HCl salt.
(3) Momose, Y.; Meguro, K.; Ikeda, H.; Hatanaka, C.; Oi, S.; Sohda, T.
Chem. Pharm. Bull. 1991, 39, 1440–1447.
(4) Sohda, T.; Momose, Y.; Meguro, K.; Kawamatsu, Y.; Sugiyama, Y.;
Ikeda, H. Arzneim.-Forsch. 1990, 40, 37–42.
(5) Yoshioka, T.; Nishi T.; Kanai, T.; Aizawa, Y.; Wada, K.; Fujita, T.;
Horikoshi, H. European Patent EP549365, xxxx.
1190
•
Vol. 13, No. 6, 2009 / Organic Process Research & Development
10.1021/op900131m CCC: $40.75  2009 American Chemical Society
Published on Web 09/29/2009

Scheme 1. Precedented synthetic approach
Scheme 2. Improved synthetic approach: route A
(feasibility) and B (after optimization)
Table 1. Optimization of solvent quantity used in the
synthesis of intermediate 11
toluene quantity
yield of 11
entry quantity of 2 (g)
(mL)
vol
(g)
(%)
1
2
3
25
25
25
75
100
125
3.0
4.0
5.0
33.7
34.9
33.8
88.8
92.0
89.1
Results and Discussion
In our endeavor to improve the process, the synthesis of
pioglitazone 1, as shown in Scheme 2, starts with the protection
of 5-ethyl-2-pyridyl ethanol 2 with the methyl sulfonyl group
to obtain intermediate 11. Subsequently, this intermediate was
subjected in situ to nucleophilic substitution. In particular, the
reaction between intermediate 11 and p-hydroxybenzaldehyde,
4, in presence of potassium carbonate afforded intermediate 5.
The reaction between intermediate 5 (not isolated) and thiazo-
lidinedione 6, employing Knoevenagel conditions, afforded
penultimate intermediate 7 in 50% yield and 99% purity for
the three steps. Moreover, the reduction of intermediate 7 in
the presence of Co (II) salts as a catalyst at 30 °C yielded the
desired compound 1 in 90-98% yield and 98% purity.
In one of the reactions, following the literature procedure,3
15 mol % CoCl2 ·6H2O/NaBH4/dimethyl glyoxime system
afforded 1 in 98% yield and 98% purity. This reduction was
also performed, in the presence of 15 mol % Co(NO3)2 ·6H2O/
NaBH4/dimethyl glyoxime as a catalytic system, that gave rise
to the product 1 in same yield and purity. However, cobalt
ammonium sulfate as a catalyst, instead of cobalt chloride or
nitrate, did not offer better yields although the purity was found
to be same. Pharmaceutically acceptable salt of 1 was prepared
by employing 1.5 equiv of aq HCl and isopropanol in 90%
yield and >99.7% purity (ICH grade).
The process that we developed at our end was optimized to
find out the best possible conditions for the synthesis of 1. The
optimization of every step is described in detail as given below:
Optimization of Solvent Quantity Used in the Synthesis
of Intermediate 11. Preparation of 11 involves mesylation of
2-(5-ethyl-2-pyridyl) ethanol 2, employing methanesulfonyl
chloride and triethylamine in toluene. During feasibility studies,
7 volumes of toluene were used as shown in Scheme 2 (route
2A). We anticipated that the volume of the solvent can further
be optimized to a lesser quantity. With 3 volumes of solvent, a
thick reaction mass was observed; however, with 4 or 5 volumes
of solvent the stirring of the reaction mass was found to be
efficient [Table 1 (entry 2)]. In this study, except for the solvent,
all other parameters were the same as in Scheme 2 (route 2A).
Eventually, we could perform the reaction with 4 volumes
of toluene and the detailed investigation is summarized in Table
1.
Optimization of Triethylamine Equivalence in the Syn-
thesis of 11. As mentioned before, preparation of 11 involves
triethylamine in toluene medium. During feasibility studies 1.0
equiv of triethylamine was used as shown in Scheme 2 (route
2A). In order to improve the yield, we reasoned that the quantity
Vol. 13, No. 6, 2009 / Organic Process Research & Development
•
1191

Table 2. Optimization of triethylamine equivalence in the
synthesis of 11
Table 5. Optimization of solvent quantity used in the
synthesis of intermediate 5
triethylamine quantity yield of 11
quantity of organic toluene quantity yield of 5
quantity of
purity by
layer containing
purity by
vol (g) (%) HPLC (%)
entry
2 (g)
(mL)
equiv
(g) (%) HPLC (%)
entry
11 (mL)
(mL)
1
2
3
4
5
25
25
25
25
25
23.0
25.8
28.8
31.2
34.6
1.0
32.9
33.1
34.6
34.9
34.4
87
87
91
92
91
61
70
80
81
82
1
2
3
150
150
150
185
155
100
7.4
6.2
4.0
31.3 74.1
32.4 76.7
72.8
80.8
-
1.12
1.25
1.35
1.5
-
-
Table 6. Optimization of equivalents of
4-hydroxybenzaldehyde 4 in the preparation of 5
Table 3. Optimization of methanesulfonyl chloride
equivalents in the synthesis of 11
quantity of organic quantity of 4 yield of 5
layer containing
purity by
MsCl quantity yield of 11
quantity of
purity by
entry
11 (mL)
(g) equiv (g) (%) HPLC (%)
entry
2 (g)
(mL)
equiv
(g)
(%) HPLC (%)
1
2
3
4
150
150
150
150
18.2 0.9
20.2 1.0
21.6 1.07 33.8 80.0
24.2 1.2 32.6 77.2
33.5 79.3
33.1 78.4
76.1
80.9
85.4
84.2
1
2
3
4
25
25
25
25
19.0
21.2
23.7
25.6
1.0
30.2
32.9
33.1
32.4
80
87
87
85
78
84
55
75
1.12
1.25
1.35
In this way we optimized the reaction conditions involved
in the preparation of 11 as shown in Scheme 2, route 2B.
Optimization of Solvent Quantity in the Preparation of
5. Preparation of 5 involves condensation of 4-hydroxybenzal-
dehyde 4 and 2-(5-ethyl-2-pyridyl)ethyl methanesulfonate 11
using potassium carbonate in toluene. During feasibility studies
7.4 volumes of toluene were used as shown in Scheme 2, route
2A. As per our optimization strategy, we realized that the
volume of the solvent can further be optimized to a lesser
quantity. With 4 volumes of solvent, a thick reaction mass was
observed; however, with 6.2 volumes as shown in Table 5 (entry
2) of solvent the stirring of the reaction mass was found to be
efficient. In this study, except for the solvent, all other
parameters remained the same as in Scheme 2 (route 2A).
Optimization of Equivalence of 4-Hydroxybenzaldehyde
4 in the Preparation of 5. Condensation of 4-hydroxybenzal-
dehyde 4 with 2-(5-ethyl-2-pyridyl)ethyl methanesulfonate 11
leads to the synthesis of 5. During feasibility studies 1 equiv of
4 was used as shown in Scheme 2 (route 2A). After optimiza-
tion, the yield and purity in the preparation of 5 were found to
be optimal with 1.07 equiv of 4 (Table 6, entry 3). In different
reactions, less or more than 1.07 equiv of 4 was found to be
inefficient (entries 1, 2, and 4). In this study, except for the
volume of the solvent and quantity of 4, all other parameters
remained the same as in Scheme 2 (route 2A).
Table 4. Optimization of temperature in the synthesis of 11
yield of 11
quantity of
purity by
entry
2 (g)
temperature (°C) (g)
(%) HPLC (%)
1
2
3
4
25
25
25
25
0–5
25–35
-5–0
45–50
32.9
33.3
33.0
24.2
87
89
87
64
85
88
85
83
of base could be detrimental and can be optimized to adequate
quantity. With 1.12 equiv of triethylamine, a lesser purity of
the product was observed. Additionally, different quantities of
base were investigated but the best combination of purity and
yield, considering the amount of base, was observed with 1.25
equiv of triethylamine as shown in Table 2 (entry 3). In this
study, except for toluene and triethyl amine, all other parameters
were the same as in Scheme 2 (route 2A).
Optimisation of Methanesulfonyl Chloride Equivalence
in the Synthesis of 11. As a part of our strategy, the number
of equivalents of methanesulfonyl chloride in the preparation
of 11 is also optimized. During feasibility studies 1.0 equiv of
methanesulfonyl chloride was used as shown in Scheme 2 (route
2A). In order to improve the yield, we anticipated that the
quantity of mesylating reagent could also be detrimental.
Employing 1.12 equiv of methanesulfonyl chloride as shown
in Table 3 (entry 2), we obtained the product with the best yield
and purity. In the different reactions with the increased amount
of methylsulfonyl chloride, we isolated the product with the
lower yield and purity as shown in Table 3. In this study, except
for toluene, triethylamine, and methanesulfonyl chloride, other
reaction conditions remained the same as in Scheme 2 (route
2A).
Optimization of Temperature in the Synthesis of 11.
Temperature always plays a pivotal role in the synthesis;
therefore, as a part of our strategy preparation of 11 at different
temperatures is also studied. During feasibility studies, the
reaction was conducted at 0-5 °C as shown in Scheme 2 (route
2A). In order to improve the yield, during optimization, we
performed the reaction at 25-35 °C and isolated the product
with best purity and yield as shown in Table 4 (entry 2). In the
different reactions at lower or higher temperatures, we were
not able to isolate the product with better yield and purity as
shown in Table 4.
Optimization of Potassium Carbonate Quantity in the
Preparation of 5. Condensation of 4-hydroxybenzaldehyde 4
with 2-(5-ethyl-2-pyridyl) ethyl methane sulfonate 11 using
potassium carbonate in toluene leads to the synthesis of 5.
During feasibility studies 1.0 equiv of potassium carbonate was
used as shown in Scheme 2 (route 2A). After optimization, 1.75
equiv of potassium carbonate was found to be sufficient to
obtain best yield and purity of 5 as shown in Table 7 (entry 2).
In different reactions, less or more than 1.75 equiv of base did
not offer better results.
The optimized reaction conditions involved in the preparation
of 5 are shown in Scheme 2 (route 2B).
Optimization of Methanol Quantity in the Preparation
of 7. Preparation of 7 involves the condensation of 2,4-
thiozolidinedione 6 and 5 using piperidine base in methanol.
As shown in Scheme 2 (route 2A), 13 volumes of methanol
were used. After optimization, 7 volumes of the methanol were
1192
•
Vol. 13, No. 6, 2009 / Organic Process Research & Development

Table 7. Optimization of equivalents of potassium carbonate
in the preparation of 5
Table 11. Optimization of quantity of sodium borohydride
in the preparation of 1
quantity of organic
sodium borohydride
yield
(g) (%) HPLC (%)
quantity of K2CO3 yield of 5
purity by
layer containing
purity by
entry
11 (mL)a
(g)
equiv
(g) (%) HPLC (%)
entry 7 (g)
(g)
(equiv)
1
2
3
4
5
175
175
175
175
175
34.3
40.0
57.1
22.8
17.1
1. 5
1.75
2.5
1.0
0.75
32.3 76.5
33.8 80.0
31.8 75.3
30.2 71.5
28.4 67.2
82.2
85.4
86.0
79.3
76.6
1
2
3
4
5
6
7
20
20
20
20
20
20
20
2.1
2.6
3.0
3.2
4.0
3.6
4.9
1.0
1.2
1.4
1.5
1.9
1.7
2.3
18.2 90.4
18.5 91.9
18.5 91.9
19.0 94.4
18.2 90.4
19.1 94.9
18.0 89.4
88.77
90.73
94.16
97.21
98.69
98.14
98.08
a 175 mL of organic layer is equal to 33 g of 11.
Table 8. Optimization of methanol quantity in the
The optimized reaction conditions involved in the preparation
of 7 are shown in Scheme 2 (route 2B).
preparation of 7
yield
Optimization of Quantity of Sodium Borohydride in the
Preparation of 1. Preparation of 1 involves reduction of 7 using
sodium borohydride in aq methanol and DMF. Dimethyl
glyoxime and cobalt chloride hexahydrate was used as a catalyst.
In feasibility studies, 1.2 equiv of sodium borohydride was used
for the preparation of 1 as shown in Scheme 2 (route 2A). After
optimization, 1.7 equiv of sodium borohydride was found to
be sufficient to effect the transformation with high yield and
purity [Table 11 (entry 6)].
entry
5 (g)
methanol (vol)
(g)
(%)
purity (%)
1
2
3
25
25
25
7.0
10.0
12.0
23.3
23.5
23.3
67.1
67.7
67.1
96.2
95.9
98.2
Table 9. Optimization of equivalents of 2,4-thiozolidinedione
in the preparation of 7
yield
entry
5 (g)
6 (equiv)
(g)
(%)
purity (%)
An improved process presented in Scheme 2 (route 2B),
appears to be advantageous over the existing synthesis, e.g. (a)
transformation from 2 to 11 was high yielding, and the
enrichment of E2 elimination impurity during the reaction was
avoided up to greater extent [the route described in the patent
(Scheme 1a) yields 30-40% E2 elimination impurity; however,
in our optimized process (Scheme 2, route 2B) it was found to
be only 5-8%]; (b) expensive Pd metal was replaced with Co
(II) salts as a catalyst in the reduction of 7 to 1; (c) plant-
unfriendly, pyrophoric, non-nucleophilic base, NaH, was com-
pletely avoided as we developed a process analogous to Scheme
1a; (d) similarly, a cascade of reactions involving expensive
metal (Pd), toxic acid (HBr), and diazonium chloride intermedi-
ate was avoided to yield intermediate 10, as the process
analogous to Scheme 1b was not developed; and (e) expensive,
partially recoverable solvents, e.g., dioxane, THF, and CH2Cl2,
were replaced with toluene, methanol, and isopropanol.
1
2
3
4
5
25
25
25
25
25
0.68
0.85
0.75
0.94
0.61
22.6
23.2
23.2
22.9
21.7
65.1
66.8
66.8
65.9
62.5
95.7
96.1
97.6
96.8
95.19
Table 10. Optimization of piperidine quantity in the
preparation of 7
yield
entry 5 (g) piperidine (equiv)
(g)
(%)
purity (%)
1
2
3
4
25
25
25
25
0.64
0.72
0.83
0.95
22.8 65.6
22.8 65.6
25.5 73.4
23.6 68.0
97.4
98.3
94.52
96.7
found to be sufficient for effective stirring and solubility of the
substrates, Table 8 (entry 1). In this study, except for the volume
of solvent, all other parameters were the same as in Scheme 2
(route 2A).
Optimization of Equivalents of 2,4-Thiozolidinedione in
the Preparation of 7. In feasibility studies, 0.85 equivalent of
2,4-thiozolidinedione 6 was used to prepare 7 as shown in Table
9 (entry 2). After optimization, 0.75 equivalent of 6 was found
to be sufficient to afford 7 in better yield and purity [Table 9
(entry 3)]. In this study, except for the volume of solvent and
equivalents of 6, all other parameters were the same as in
Scheme 2 (route 2A).
Optimization of Piperidine Quantity in the Preparation
of 7. As mentioned in Scheme 2, preparation of 7 involves the
condensation of 6 and 5 using piperidine base in methanol. In
feasibility studies, 0.72 equiv of piperidine was used for the
preparation of 7 [Scheme 2 (route 2A)]. After optimization, 0.83
equiv of piperidine base was found to be sufficient to effect
the transformation [(Table 10 (entry 3)]. In this study, except
for the volume of solvent, equivalents of 6 and equivalents of
piperidine base, all other parameters remained the same as in
Scheme 2 (route 2A).
Conclusions
We have developed an improved process for pioglitazone 1
which appears to be more compatible with industrial scale and
has some advantages over the existing synthesis.
Experimental Section
Solvents and reagents were obtained from commercial
sources and used without further purification. The 1H and 13
C
spectra were measured in DMSO-d6 using 200 or 400 MHz on
a Varian Gemini and Varian Mercury plus 2000 FT NMR
spectrometer; the chemical shifts were reported in δ ppm. IR
spectra were recorded in the solid state as KBr dispersion using
Perkin-Elmer 1650 FT IR spectrometer. The mass spectrum
(70 eV) was recorded on an HP 5989 A LC-MS spectrometer.
The melting points were determined by using the capillary
method on Polmon (model MP-96) melting point apparatus.
The solvents and reagents were used without further purification.
Vol. 13, No. 6, 2009 / Organic Process Research & Development
•
1193

Preparation of 5-[4-[2-(5-Ethyl-2-pyridyl)ethoxy]benzi-
lidene]-2,4-thiazolidinedione (7). To a stirring solution of
triethylamine (57.6 L, 413 mol), 2-(5-ethyl-2-pyridyl)ethanol
(50 kg, 331 mol), and toluene (200 L) at about 25 °C was added
slowly (200 g/min) methanesulfonyl chloride (42.5 kg, 371 mol)
for about 25 min. After stirring for 3 h at 25 °C, the separated
solid was filtered and washed with toluene (60 L). The obtained
filtrate was washed with 4% sodium bicarbonate solution (60
L). The layers were separated, the aqueous phase was extracted
with toluene (50 L), and the combined organic layers were
washed with water (2 × 60 L). To the organic layer containing
intermediate 11 were added 4-hydroxybenzaldehyde (43.2 kg,
354 mol) and potassium carbonate (80 kg, 579 mol) in one
portion at 25 °C. After stirring for 13 h at 90 °C, the reaction
mass was cooled to 50 °C in 1 h and quenched with water (250
L) under stirring. The layers were separated, the aqueous phase
was extracted with toluene (2 × 150 L), and the combined
organic layers were washed with 5% sodium hydroxide solution
(250 L). The clear organic layer was distilled completely under
vacuum below 65 °C to afford the crude compound 5 (68.6
kg), in 81.2% yield and >83.0% purity (unsuccessful attempts
were made to distill the product at below <250 °C). To a
solution of 5 in methanol (450 L) were added piperidine (23.3
kg, 274 mol) and 2,4-thiozolidinedione (29.1 kg, 248 mol) at
25 °C. After stirring for 13 h at 65 °C, the reaction mass was
cooled in 1 h to 25 °C and diluted with methanol (195 L) under
stirring. The pH of reaction mass was adjusted to about 6-6.5
by using acetic acid (35 L) followed by addition of methanol
(130 L) under stirring for about 15 min. After stirring for 1.5 h
at 65 °C, the reaction mass was cooled to 25 °C in 1 h, and the
separated solid was filtered and washed with methanol (65 L)
and dried under vacuum at 65 °C for 4 h to afford compound
7 (58.6 kg) in 50% yield and >98.9% purity. Spectroscopic data
were found to be in agreement with the data collected from an
authentic sample of intermediate 7.
at 35 °C. Subsequently, a mixture of sodium borohydride (11.1
kg dissolved in water 120 L) and 4% sodium hydroxide solution
(27 L) was added slowly (150 mL/min) at 20 °C over a period
of 3 h. After additional stirring for 3 h, the charcoal (3 kg) was
added to the reaction mass and the stirring continued for about
30 min at 20 °C. The reaction mixture was passed through the
Celite bed, and the pH of the filtrate was adjusted within the
range of 6.5-7.0 using acetic acid (24 L). After stirring for 45
min at 25 °C, the separated solid was filtered, washed with water
(60 L) and methanol (2 × 75 L), and dried under vacuum at
70 °C for 8 h to afford compound 1 (59.2 kg) in 98% yield
and >98% purity. Spectroscopic data were found to be in
agreement with the data collected from an authentic sample of
pioglitazone 1.1
Preparation of Pioglitazone Hydrochloride (1 ·HCl) salt.
To a stirring solution of compound 1 (50 kg, 141 mol) in
isopropyl alcohol (250 L) was added hydrochloric acid (75 L)
over the period of 10 min at 25 °C and heated to 80 °C to
achieve clear solution. After additional stirring for 45 min, the
clear solution was in 1 h cooled to about 25 °C, and the
separated solid was filtered and washed with isopropyl alcohol
(50 L) and dried under vacuum at 70 °C for 4 h to afford 1·HCl
(49.6 kg) in 90% yield and >99.7% purity. Spectroscopic data
were found to be in agreement with the data collected from
authentic sample of pioglitazone 1.1
1H NMR (400 MHz, DMSO-d6) 12.08 (s, 1H), 8.73 (d, 1H,
J ) 1.6 Hz), 8.43 (dd, 1H, J ) 2.0, 8.0 Hz), 8.01 (d, 1H, J )
8.0 Hz), 7.16 (d, 2H, J ) 8.8 Hz), 6.89 (d, 2H, J ) 8.8 Hz),
4.88 (dd, 1H, J ) 4.4, 8.8 Hz), 4.34 (t, 2H, J ) 6.2 Hz), 3.55
(t, 2H, J ) 6.2 Hz), 3.30 (dd, 1H, J ) 4.4, 14.0 Hz), 3.06 (dd,
1H, J ) 8.8, 14.0 Hz), 2.80 (q, 2H, J ) 7.6 Hz), 1.24 (t, 3H,
J ) 7.6 Hz); 13C NMR (400 MHz, DMSO-d6) 175.6, 171.5,
156.9, 151.0, 145.2, 141.3, 139.8, 130.3, 129.0, 127.1, 114.4,
65.4, 52.8, 40.1, 39.9, 39.7, 39.5, 39.2, 39.0, 38.8, 36.2, 32.1,
24.5, 14.5; IR (KBr) 2928, 2742, 1743, 1694, 1616, 1510, 1461,
1313, 1243, 1038, 850, 712 cm-1; HRMS (Cl) calcd For
C19H20N3O3S (M+) 356.44; found (MH+) 357.5.
Preparation of 5-[4-[2-(5-Ethyl-2-pyridyl)ethoxy]benzyl]-
2,4-thiazolinedione (1). To a stirring solution of compound 7
(60 kg, 169 mol) in methanol (120 L), water (420 mL), and
4% sodium hydroxide solution (78 L) was added a solution of
cobalt nitrate hexahydrate (0.75 kg, 2.5 mol) in dimethyl
glyoxime (7.8 kg, 67.2 mol) and dimethyl formamide (90 L)
Received for review May 26, 2009.
OP900131M
1194
•
Vol. 13, No. 6, 2009 / Organic Process Research & Development