Discovery of the decarboxylative biaise reaction and its application to the efficient synthesis of ethyl 2,6-dichloro-5-fluoronicotinoylacetate

Article abstract of DOI:10.1021/op7001694

An efficient synthesis of 2,6-dichloro-5-fluoronicotinoylacetate (1) has been accomplished in a single step using the unprecedented decarboxylative Biaise reaction of 3-cyano-2,6-dichloro-5-fluoropyridine (4) with potassium ethyl malonate in the presence of zinc chloride.

Full text of DOI:10.1021/op7001694

Organic Process Research & Development 2007, 11, 1062–1064
Discovery of the Decarboxylative Blaise Reaction and Its Application to the Efficient
Synthesis of Ethyl 2,6-Dichloro-5-fluoronicotinoylacetate1
Jae Hoon Lee, Bo Seung Choi, Jay Hyok Chang, Sam Sik Kim, and Hyunik Shin*
Chemical DeVelopment DiVision, LG Life Sciences, Ltd./R&D, 104-1 Moonji-dong, Yusong-gu, Daejeon 305-380, Korea
Scheme 1
Abstract:
An efficient synthesis of 2,6-dichloro-5-fluoronicotinoylacetate (1)
has been accomplished in a single step using the unprecedented
decarboxylative Blaise reaction of 3-cyano-2,6-dichloro-5-fluoro-
pyridine (4) with potassium ethyl malonate in the presence of zinc
chloride.
A naphthyridine ring is embedded as a key structural unit
of many potent quinolone antibiotics such as enoxacin,2
tosufloxacin,3 trovafloxacin,4 and gemifloxacin (Scheme 1).5 To
construct this structural segment, most of the reported syntheses
employed as a key starting material ethyl 2,6-dichloro-5-
fluoronicotinoylacetate (1), which was prepared from a common
synthetic route, the reaction of an acetate enolate equivalent
with 2,6-dichloro-5-fluoronicotinoyl chloride (2) (Scheme 2).
The reaction of magnesium enolate of diethyl malonate with
the nicotinoyl chloride 2 proceeded well to give the diester
intermediate 3 (R ) OEt), which was partially hydrolyzed and
decarboxylated to give 1.6 However, this process is complicated
by the selective partial hydrolysis of the diester intermediate
3: the formation of the methyl ketone impurity7 via double
decarboxylation was observed as a side product. To circumvent
this drawback, malonate monoester8 and ethyl acetoacetate9
Scheme 2
were used instead of diethyl malonate as an acetate enolate
equivalent. As an alternative route, we have devised the one-
pot Blaise reaction transformation10 of nitrile 4 to 1 (Scheme
3). Although this process is more efficient than the current
commercial process, its highly exothermic nature, use of excess
materials, and use of lachrymatory ethyl bromoacetate leave
room for further improvement for a large-scale operation.
* To whom correspondence should be addressed. Tel: 82-42-8662471. Fax:
82-42-8665754. E-mail: hisin@lgls.co.kr.
(1) The content of the paper has been published as a part of patent. See:
Shin, H.; Choi, B. S.; Lee, J. H. WO 2007064077, 2007.
(2) (a) Miyamoto, T.; Egawa, H.; Shibamori, K.; Matsumoto, J.-I.
J. Heterocycl. Chem. 1987, 24, 1333–1339. (b) Matsumoto, J.-I.;
Miyamoto, T.; Minamida, A.; Nishimura, Y.; Egawa, H.; Nishimura,
H. J. Heterocycl. Chem. 1984, 21, 673.
(3) Chu, D. T. W.; Fernandes, P. B.; Claiborne, A. K.; Gracey, E. H.;
Pernet, A. G. J. Med. Chem. 1986, 29, 2363.
Scheme 3
(4) (a) Brighty, K. E.; Castaldi, M. J. Synlett 1996, 1097. (b) Braish, T. F.;
Castaldi, M. J.; Chan, S.; Fox, D. E.; Keltonic, T.; McGarry, J.;
Hawkins, J. M.; Norris, T.; Rose, P. R.; Sieser, J. E.; Sitter, B. J.;
Watson, H., Jr. Synlett 1996, 1100.
(5) Hong, C. Y.; Kim, Y. K.; Chang, J. H.; Kim, S. H.; Choi, H.; Nam,
D. H.; Kim, Y. Z.; Kwak, J. H. J. Med. Chem. 1997, 40, 3584.
(6) (a) The reported yield of 1 from the acid chloride 2 is 74%. See:
Bouzard, D.; Di Cesare, P.; Essiz, M.; Jacquet, J. P.; Ledoussal, B.;
Remuzon, P.; Kessler, R. E.; Fung-Tomc, J. J. Med. Chem. 1992, 35,
518. (b) Miyamoto, T.; Matsumoto, J.-I. Chem. Pharm. Bull. 1990,
38, 3211.
On the basis that the copper(I) salt facilitated the decar-
boxylation of malonic acid,11 we envisaged the decarboxylative
Blaise reaction on the premise that zinc ethyl malonate species
would undergo decarboxylation to form the Reformatsky
(7) In an acid-catalyzed decarboxylation of 3 (R ) OEt), ca. 10% (area
% by HPLC) of the methyl ketone impurity was usually formed in
the reaction mixture. The analysis of the isolated 1 showed ca. 2%
contamination of the methyl ketone impurity. 1H NMR data of the
methyl ketone impurity: (300 MHz) δ 7.57 (d, J ) 7.3Hz, 1H), 2.71
(s, 3H).
(10) Shin, H.; Choi, B. S.; Lee, K. K.; Choi, H.-w.; Chang, J. H.; Lee,
K. W.; Nam, D. H.; Kim, N.-S. Synthesis 2004, 2629. (b) Choi, B. S.;
Chang, J. H.; Choi, H.-w.; Kim, Y. K.; Lee, K. K.; Lee, K. W.; Lee,
J. H.; Heo, T.; Nam, D. H.; Shin, H. Org. Process Res. DeV. 2005, 9,
311.
(8) (a) The reported yield of 1 from the acid chloride 2 is 96%. Refer to
ref 2. (b) O’Neill, B. T.; Busch, F. R.; Lehner, R. S. EP 449445, 1991.
(c) Clay, R. J.; Collom, T. A.; Karrick, G. L.; Wemple, J. Synthesis
1993, 290.
(9) The reported yield of 1 from the acid chloride 2 is 70%. See: Urban,
F. J.; Moore, B. S.; Spargo, P. L. Org. Prep. Proced. Int. 1997, 29,
231.
(11) Toussaint, O.; Capdevielle, P.; Maumy, M. Synthesis 1986, 1029. (b)
Darensbourg, D. J.; Holtcamp, M. W.; Khandelwal, B.; Reibenspies,
J. H. Inorg. Chem. 1994, 33, 531.
1062
•
Vol. 11, No. 6, 2007 / Organic Process Research & Development
10.1021/op7001694 CCC: $37.00
2007 American Chemical Society
Published on Web 10/13/2007

Scheme 4
Table 2. Metal salt effect on the decarboxylative Blaise
reaction of 4 to 6a
entry
Lewis acid
time (h)
conversion (%)b
1
2
3
4
5
ZnCl2
3
3
15
3
15
100
99
5
95
0
ZnBr2
ZnI2
Zn(OTf)2
MgX2 (X ) Cl, Br, I)
a All reactions were performed with 1.5 equiv of potassium ethyl malonate in
the presence of 1.0 equiv of Lewis acid in DCE. b The analysis was performed as
in the footnote b of Table 1.
Table 1. Solvent effect on the decarboxylative Blaise
reaction of 4a
entry
solvent
time (h)
conversion (%)b
1
2
3
4
5
6
7
8
9
NMP
2
2
100
100
100
100
9
Figure 1. Measure of CO2 evolution. The reaction was started
with 0.2 equiv of 4 in the presence of 1.5 equiv of potassium
ethyl malonate and 0.5 equiv of zinc chloride. After the initial
evolution of CO2 was complete (at ca. 7000 s), the remaining
0.8 equiv of 4 was added.
DMF
DCE
3
CHCl3
DME
dioxane
THF
3
15
15
15
15
15
2
trace
5
22
EtOH
toluene
of zinc chloride. We found that reduction of the equivalency
below 0.5 led to prolonged reaction time for the completion of
the reaction: 0.5 equiv completed the reaction in 3 h, whereas
0.25 and 0.1 equiv required 2 d to reach 96% and 88%
conversion, respectively. Although the turnover number is
marginal, it clearly provided the feasibility to develop a more
efficient catalytic reaction system.
a All reactions were run at 90 °C with 1.5 equiv of potassium ethyl malonate in
the presence of 1.0 equiv of zinc chloride. b Sample was obtained by quenching the
mixture with aqueous ammonium chloride solution, and the organic layer was checked
by GC: HP-5MS column; injector 280 °C; initial time and temp 2 min and 70 °C;
gradient 10 °C/min; final temp 300 °C; flow rate 1 mL/min. Compounds 4 and 6 were
detected at 8.9 and 17.0 min, respectively.
reagent, which should react with nitrile 4 to form 5
(Scheme 4).
In contrast to the classical Blaise reaction, the decarboxy-
lative version is endothermic (∆H ) -6.5 KJ/mol of 4),
rendering the reaction safer than the classical Blaise reaction.
Moreover, the evolution of carbon dioxide was also readily
controllable by portionwise addition of 4 to the mixture of
potassium ethyl malonate and zinc chloride. Figure 1 shows
how the evolution of carbon dioxide may be controlled in a
dose-dependent manner.12 With optimized conditions in hand,
we successfully scaled up the reaction to the preparation of ca.
20 kg of 1 as described in Experimental Section.
In conclusion, we have discovered the decarboxylative Blaise
reaction of 4 with potassium ethyl malonate in the presence of
zinc chloride. This protocol is more environmentally benign in
terms of reduced effluent discharge and avoids the use of a
lachrymatory reagent. Moreover, successful scale-up to 1 in a
one-pot manner using 20 kg of 4 clearly demonstrated its
reproducibility. The general scope of the decarboxylative Blaise
reaction will be published in due course.
A clue for the possible success for the proposed decarboxy-
lative Blaise reaction was secured by an early outcome: heating
a mixture of potassium ethyl malonate, nitrile 4, and zinc
chloride in toluene to provide 22% conversion of the reaction
(entry 9 in Table 1). Encouraged by this result, further screening
of reaction solvents was performed to disclose a dramatic
solvent effect. Polar aprotic solvents such as DMF and NMP
(entries 1 and 2) showed complete conversion of the reaction.
So did the chlorinated solvents such as 1,2-dichloroethane
(DCE) and chloroform (entries 3 and 4). On the other hand,
ethereal solvents such as DME, dioxane, and THF (entries 5–7)
were poor media for this transformation, and ethanol (entry 8)
showed minimal conversion.
Using DCE as the optimal medium for the reaction, the effect
of metal salts on the reaction was investigated briefly, and the
results are gathered in Table 2. Among zinc salts, zinc chloride
(entry 1 in Table 2), bromide (entry 2), and triflate (entry 4)
led to essentially complete conversion, whereas zinc iodide
(entry 3) showed very low reactivity. Magnesium salts (entry
5) did not show any of the desired reactivity.
(12) A simulated calculation using the experimentally measured CO2
evolution data showed that a 2-in. diameter of ventilation tubing is
sufficient to release CO2 without any pressure build-up when a mixture
of 100 kg of 4, zinc chloride, and potassium ethyl malonate is heated
at reflux.
To check the catalytic turnover of the reaction towards zinc
chloride, we tested the reaction with 0.5, 0.25, and 0.1 equiv
Vol. 11, No. 6, 2007 / Organic Process Research & Development
•
1063

Experimental Section
was dissolved completely. The solution was cooled down
slowly, and the pale yellowish solid started to precipitate at
about 40 °C. The solution was further cooled down to 0 °C
and stirred for another 1 h. The precipitate was filtered off and
washed with EtOH/H2O (5/1, 100 L) at ambient temperature
to give 1 as a white solid (22.0 kg, 75.0%). Physical and spectral
data of 1: mp 68–70 °C; 1H NMR (400 MHz, CDCl3) δ (enol
form, 86%) 12.57 (s, 1H), 7.83 (d, J ) 7.6 Hz, 1H), 5.83 (s,
1H), 4.30 (q, J ) 7.2 Hz, 2H), 1.34 (t, J ) 7.2 Hz, 3H); (keto
form, 14%) 7.83 (d, J ) 7.6 Hz, 1H), 4.20 (q, J ) 7.2 Hz,
2H), 4.09 (s, 2H), 1.26 (t, J ) 7.2 Hz, 3H); 13C NMR (100
MHz, CDCl3) δ (enol form) 172.4, 165.1, 154.0 (d, J ) 270
Hz), 141.6, 138.7 (d, J ) 20 Hz), 130.3, 126.9 (d, J ) 20 Hz),
95.0, 61.2, 14.1; MS (APCl, m/z) 284 (M + 4), 282 (M + 2),
280 (M+).
Solvents and reagents were obtained from commercial
sources and used without further purification. NMR spectra were
obtained on a Bruker 400 MHz and a Jeol 500 MHz spectro-
meter. HPLC analyses were carried out on a Hewlett–Packard
1100 system and a Waters 490E detector and 616 pump system.
Mass spectra were collected using a Finnigan LCQ mass
spectrometer system and a Jeol JMX-700 mass spectrometer.
Ethyl 2,6-Dichloro-5-fluoronicotinoylacetate (1). To a
250-L reactor containing 1,2-dichloroethane (100 L) were added
3-cyano-2,6-dichloro-5-fluoropyridine (4, 20.0 kg, 105 mol),
zinc chloride (7.14 kg, 52.4 mol), and potassium ethyl malonate
(21.4 kg, 126 mol) in sequence. The mixture was heated at
reflux, leading to a gradual color change to red. After 3 h, 6 N
HCl (150 L) was added slowly, and the mixture was heated at
reflux for 30 min. The mixture was cooled down to ambient
temperature, and the organic phase was separated and concen-
trated. To the residue was added EtOH/H2O (5/1, 100 L), and
the mixture was heated to 70 °C until all of the solid material
Received for review July 26, 2007.
OP7001694
1064
•
Vol. 11, No. 6, 2007 / Organic Process Research & Development