Vol. 26, No. 12 (2014)
Synthesis of Main Impurity of Vildagliptin 3491
for 4.5 h at room temperature. The reaction was treated with
water (100mL) and extracted with ethyl acetate (100 mL).The
organic layer was separated and the aqueous layer was re-
extracted with ethyl acetate (50 mL × 2), The combined
organic layers was washed with 5 % sodium bicarbonate and
some low cost reagent, such as cyanuric chloride, as dehy-
23
dration agent of amide . The compound 6 in dry DMF was
treated with cyanuric chloride at room temperature, and after
the accomplishment of the reaction, the side product was
quenched with water and the objective product 6 was isolated
from the ethyl acetate extract with saturated aqueous sodium
bicarbonate solution. Therefore, the compound 6 was prepared
in 87 % yield.
2 4
saturated brine solution, dried over anhydrous Na SO . Evapo-
ration of the solvent gave a residue, which was crystallized by
the isopropyl ether to afford compound 5 (3.33 g, 87 %). HPLC
2
0
Purity 98 %. m.p. 62-63 °C (lit. 53-57 °C). IR (KBr, νmax
,
To prepare the ultimate target compound 1, we started
from compound 5 with 3-hydroxy-1-aminoadamantane as the
route 1, and we changed the solvent was 2-butanone refluxing
under potassium carbonate and potassium iodide condition.
According to this method, we avoided the use of expensive
vildagliptin as a reaction material and the column chromato-
graphy was also avoided been used to isolate the product,
meanwhile, the reaction time has been shortened to only 4h
and the purity of the product also was improved. Finally, the
crude product was purified by recrystallization from ethyl
acetate and methanol (1:1) to obtain colorless crystals in 77 %
yield.
-1
1
cm ) : 2952, 2887, 2241, 1655. H NMR (300 MHz, CDCl
3
):
2
.15-2.40 (m, 4H), 3.55-3.65 (m, 1H), 3.70-3.80 (m, 1H), 4.07-
1
3
4
3
.12 (s, 2H), 4.72-4.87 (m, 1H); C NMR (75 MHz, CDCl ):
δ 22.5, 24.5, 25.7, 30.1, 32.5, 41.9, 46.7, 46.7, 46.9, 47.5,
1
+
17.4, 164.9, 165.3; MS (ESI ) 173.1(M +1).
+
Preparation of (2S)-1-{[{2[(2S)-2-cyanopyrrolidin-1-
yl]-2-oxoethyl}[(3-hydroxytricyclo[3.3.1.1(3,7)]dec-1-yl)-
amino]]acetyl}-pyrrolidine-2-carbonitrile (1):The compound
5
(2.2 g, 0.0127 mol) and 3-hydroxy-1-aminoadamantane (1 g,
mmol) were dissolved in 2-butanone (15 mL), then K CO
3
6
2
(
3.3 g, 24 mmol) and KI (50 mg, 0.3 mmol) were added. The
resulting reaction mixture was stirred in refluxing for 4 h.After
completing consumption of the starting material, filtered.
Evaporation of the solvent gave a residue, which was recrysta-
llized by ethyl acetate and methanol (1:1) to afford compound
Thus the method becomes very convenient and gives the
desired products in excellent overall yields of 50.64 % and
the purity of 98.17 %.
17
Conclusion
1
1
2
(2 g, 77 %). HPLC Purity 98.17 %. m.p. 182-184 °C (lit
-1
81-183 °C). IR (KBr, νmax, cm ): 3419, 2920, 2906, 2880,
1
851, 2239, 1649, 1450, 1424, 1404, 1311, 1003. H NMR
In conclusion, a new practical, convenient and high-
yielding method of improving the compound 1 has been
developed, which provides material with high quality. This
progress almost avoids all the drawbacks, as it does not involve
the use of expensive L-prolinamide, TFFA and other non-
commercially available reagents. It also delivers the products
(
300 MHz, DMSO-d
6
) : 1.30-1.68 (m,12H), 1.75-2.38 (m,
1
1
2
6
0H), 3.10-3.30(m,1H), 3.35-3.87 (m, 6.7H), 4.40-4.50 (m,
13
H), 4.57-4.73, (m, 1H); C NMR (75 MHz, DMSO-d
4.4, 28.9, 29.8, 34.4, 37.7, 43.9, 45.7, 45.9, 46.5, 49.9, 57.8,
6
): δ
+
+
7.5, 118.5, 170.0; MS (ESI ) 440.3 (M +1).
5 into a higher yield than previously reported syntheses. This
known method will be used in large-scale preparations.
RESULTS AND DISCUSSION
In this study (Scheme-I), we planned to use L-proline in
ACKNOWLEDGEMENTS
place of L-prolinamide on account of its easy availability.
Furthermore, the chloroacetyl group was used to play the role
of a protecting group and it could also avoid being removed.
This work was supported by Natural Science Foundation
of Chongqing (CSTC, 2006BB5286). Thanks are also due to
the Chongqing Medical University for partial financial support
of this work.
19
As previously reported in the literature , L-proline (2) was
N-acylated with chloroacetyl chloride in refluxing THF to
afford compound 3 and this process should avoid any aqueous
media. Finally, we found that an appropriate extension of time
to 2.5 h can improve its yield up to about 90 %.
REFERENCES
1
2
3
.
.
.
D.K. Arulmozhi and B. Portha, Eur. J. Pharm. Sci., 28, 96 (2006).
J.J. Holst, Exp. Opin. Investig. Drugs, 8, 1409 (1999).
C.F. Deacon, M.A. Nauck, M. Toft-Nielsen, L. Pridal, B. Willms and
J.J. Holst, Diabetes, 44, 1126 (1995).
We attempted to convert the carboxylic acid moiety of
compound 3 to the amide, but unfortunately, most of these
methods had restrictions on the utility and applicability due to
the extreme toxicity, cumbersome operation and high costs
4
.
B.D. Green, P.R. Flatt and C.J. Bailey, Diab. Vasc. Dis. Res., 3, 159
(2006).
2
0-21
in the availability of the reagents . However, after several
attempts, a method was successfully applied to carboxylic acid
just simply by using the di-tert-butyl dicarbonate-pyridine
5
6
.
.
E.J. Verspohl, Pharmacol. Rev., 64, 188 (2012).
S. Dhillon, Drugs, 70, 489 (2010).
7. N.A. Thornberry and A.E. Weber, Curr. Topics Med. Chem., 7,
557 (2007).
22
system . While optimizing the reaction condition, we noted
that the solvent and the temperature have significant effects
on this process. The compound 3 was treated with di-tert-butyl
dicarbonate and pyridine (13:6) at room temperature in MeCN
followed by ammonium bicarbonate. Thus, the compound 4
was prepared in 84 % yield.
8
9
1
.
.
B. Ahren, A. Schweizer, S. Dejager, E.B. Villhauer, B.E. Dunning an
J.E. Foley, Diabetes Obes. Metab., 13, 775 ( 2011).
G. Stamataros and S.H. Schneider, Expert Opin. Pharmacother., 12,
1967 (2011).
0. M. Banerjee, N. Younis and H. Soran, Expert Opin. Pharmacother.,
0, 2745 (2009).
1
1
1
1. D.S. Kania, J.D. Gonzalvo and Z.A. Weber, Clin. Ther., 33, 1005 (2011).
2. J.J. Neumiller and R.K. Campbell, Am J. Health Syst. Pharm., 67, 1515
(2010).
Taking into account our interest for the nitrile synthesis,we
investigated the possibility of synthesizing compound 5 with