Synthesis and Purification of 6-Ethoxy-4-oxo-1,4-dihydro-[1,5] naphthyridine-3-carboxylic Acid Benzylamide

Article abstract of DOI:10.1021/op0341061

The synthesis of 6-ethoxy-4-oxo-1,4-dihydro-[1,5]naphthyridine-3- carboxylic acid benzylamide (1) on multikilogram scale is described. The major challenge for the synthesis of this quinolone GABA partial agonist was in the isolation of product of acceptable purity for clinical studies due to the insolubility of this compound. Also described are efforts to circumvent a high-temperature cyclization required for the synthesis of the quinolone ring system.

Full text of DOI:10.1021/op0341061

Organic Process Research & Development 2003, 7, 873−878
Synthesis and Purification of
6-Ethoxy-4-oxo-1,4-dihydro-[1,5]naphthyridine-3-carboxylic Acid Benzylamide
Justin Beaudin, Dennis E. Bourassa, Paul Bowles, Michael J. Castaldi, Ronald Clay, Michel A. Couturier,
Gregory Karrick, Teresa W. Makowski, Ruth E. McDermott, Clifford N. Meltz, Morgan Meltz, James E. Phillips,
John A. Ragan, David H. Brown Ripin,* Robert A. Singer, John L. Tucker, and Lulin Wei
Chemical Research and DeVelopment, Pfizer Global Research DiVision, Pfizer Inc., Eastern Point Road,
Groton, Connecticut 06340, U.S.A.
Abstract:
The synthesis of ester 7 has been previously reported and
is depicted in Scheme 1. The published synthesis, with minor
improvements, was used to prepare material for the regula-
tory synthesis. 2-Chloro-5-nitropyridine 2 was treated with
KOH in EtOH to provide ethyl ether 3. After an aqueous
workup, the organic (CH2Cl2) layer is concentrated to give
a yellow solid in 66% yield. Pyridine 3 is then taken back
in EtOH and reduced using 8 equiv of iron and 0.5 equiv of
CaCl2 for completion. The resulting product (4) is an oil that
slowly decomposes on standing.
The synthesis of 6-ethoxy-4-oxo-1,4-dihydro-[1,5]naphthyridine-
3-carboxylic acid benzylamide (1) on multikilogram scale is
described. The major challenge for the synthesis of this
quinolone GABA partial agonist was in the isolation of product
of acceptable purity for clinical studies due to the insolubility
of this compound. Also described are efforts to circumvent a
high-temperature cyclization required for the synthesis of the
quinolone ring system.
To streamline the process, a one-pot single-solvent
procedure was devised to transform 2 into 4 which is iso-
lated as a stable phosphate salt (Scheme 2). NaOEt was used
in place of KOH in EtOH. Upon reaction completion, the
nitro group was reduced with 1.2 equiv of BH3-NMe3
complex using 3% loading of Pearlman’s catalyst.4 The spent
catalyst and NaCl were removed by filtration, and H3PO4
was added to deliver the corresponding phosphate salt of 4
in 93% yield.
The coupling of aminopyridine 4 with malonate derivative
5 was shown to proceed as either the phosphate salt or free
base in refluxing CH2Cl2 (83 °C), toluene (100 °C), or
trifluorotoluene (102 °C). The high-temperature cyclization
of diester 6 was successfully scaled, but the reaction was
difficult to run. Rapid heating and cooling as well as fairly
precise control over reaction time were required to minimize
the formation of impurities in this step.
Introduction
6-Ethoxy-4-oxo-1,4-dihydro-[1,5]naphthyridine-3-carbox-
ylic acid benzylamide (1) is a subtype-selective GABA-A
receptor inverse agonist1 and was recently of interest for
clinical evaluation as an agent for the treatment of cognition
disorders such as Alzheimer’s disease. The pharmacological
selectivity of subtype-selective GABA-A receptor modulators
such as 1 is anticipated to provide an efficacious agent with
limited side effects.
A number of alternative methods for the production of
this material without the high temperatures needed for the
cyclization were investigated (Scheme 3) so that the process
could be run in a general-purpose manufacturing plant if
desired. Efforts to facilitate cyclization using Lewis acids
or bases to lower the temperature at which acylketene
formation would occur resulted in no reaction or decomposi-
tion. Protection of the NH with a benzyl group inhibited the
cyclization. Synthesis of a 2-bromopyridine and attempted
metal-halogen exchange followed by cyclization also failed
to provide desired product (e.g., 10 f 7). Heck coupling of
bromide 9 with methyl â-methoxyacrylate did provide the
cyclization product 11 in modest yield (28%). However, the
desired C-C bond formation is almost certainly occurring
in an intramolecular fashion, as all attempts to effect an
intermolecular Heck coupling with an N-blocked substrate
The synthesis of 12 is depicted in Scheme 1.3 Although
the chemistry to make the drug candidate is high yielding
and relatively simple, the insolubility of the quinolones
makes them quite difficult to purify. As a result, it was
necessary for the synthesis to produce material of high
enough purity that a rework would not be necessary or
alternatively needed to identify methods of purification for
the intermediates and product.
* Author for correspondence. E-mail: David_B_Ripin@groton.pfizer.com.
(1) (a) Davies, M. F. The Pharmacology of the Gamma-Aminobutyric Acid
System. In Brain Mechanisms and Psychotropic Drugs; Baskys, A.,
Remington, G., Eds.; CRC: Boca Raton, FL, 1996; pp 101-116. (b)
Krogsgaard-Larsen, P.; Froelund, B.; Joergensen, F. S.; Schousboe, A. J.
Med. Chem. 1994, 37, 2489-2505.
(2) (a) Albaugh, P. A.; Desimone, R. W.; Liu, G. U.S. Patent 6,143,760A, 2000.
(b) Albaugh, P. A.; Desimone, R. W.; Liu, G. World Patent 9910437A1,
1999.
(4) Couturier, M.; Tucker, J. L.; Andresen, B. M.; Dube´, P.; Breneck, S. J.;
(3) Hiroshi, T.; Toshio, A. Japanese Patent 59093080A2, 1984.
Negri, J. T. Tetrahedron Lett. 2001, 42, 2285.
10.1021/op0341061 CCC: $25.00 © 2003 American Chemical Society
Published on Web 09/30/2003
Vol. 7, No. 6, 2003 / Organic Process Research & Development
•
873

Scheme 1
Scheme 2
reflux provided that the pH of the solution was above 1.0.
The pKa of the carboxylic acid in the reaction mixture
was determined by titration to be between 4.5 and 6. A
pH target range of 2-2.5 was selected for the bulk run. As
the pH range proved difficult to hit while acidifying the
reaction mixture, it was determined that using 50% NaOH
to back-adjust the pH provided equivalent results. Under
these conditions, pilot experiments involving heating at reflux
for 48 h showed minimal formation of the previously
observed impurity. These solids filtered much more quickly
and required only 1 day to filter 155 kg in four centrifuge
loads on a 40-in. diameter centrifuge basket. The final bulk
saponification afforded a 94% yield of acceptable carboxylic
acid 8 with 4% of fine solids passing through the centrifuge
bag.
The formation of 1 is best accomplished using CDI in
DMF, followed by the addition of BnNH2. Depending on
the relative amount of CDI used, one of two impurities is
observed, either unreacted acid 8 or dibenzylurea. As
dibenzylurea proved easier to purge (vide infra), an excess
of CDI is preferable in this step. In practice, 1.1 equiv of
CDI and 1.1 equiv of benzylamine are employed in the
transformation. The initial stage of the reaction, wherein the
acylimidazole is formed, is a very thick slurry and at least
4 L/kg of DMF is required to maintain stirring in a pilot
plant reactor. This stage of the reaction is assayed by
derivatization of a sample to the API with BnNH2. After
formation of the acylimidazole, BnNH2 is added. Upon
reaction completion, H2O is added, and the product is isolated
by filtration. The intermediate acylimidazole has sufficient
stability for an alternate procedure wherein the excess CDI
can be quenched by the addition of H2O prior to addition of
BnNH2. This procedure (1.5 equiv of CDI, followed by H2O,
then BnNH2), yielded no dibenzylurea, and free-acid 8 is
observed at a level of <1.0%. Because of the relative ease
of purging the dibenzylurea, this procedure was not adopted
on scale.
(e.g., the formamide or sulfonamide derived from 9 or
formamidine 12) were unsuccessful. Since an intramolecular
Heck cyclization delivers a product with C-2 in the wrong
oxidation state (e.g., 11), this approach was not pursued
further. Attempted Baylis-Hillman cyclization of nitrile or
the corresponding ester (14-7) also failed to deliver the
desired product.
As alternate methods for the preparation of ester 7 were
not identified and using the high-temperature route could
obtain large quantities of material (several hundred kilograms
were prepared), the decision was made to stay with the
original synthesis and investigate the advantages of a high-
temperature flow reactor as an alternative should the program
progress further.
Ethyl ester 7 has a very low solubility in most solvents.
While it could be recrystallized from DMF, high dilutions
and temperatures were required for the purification. Because
of the difficulty in controlling the impurity profile at this
intermediate, carboxylic acid 8 was defined as the regulatory
starting material for the synthesis. The ethyl ester (7) could
be hydrolyzed to the carboxylic acid using NaOH in EtOH
and H2O. The resulting sodium salt solution was acidified
using concentrated HCl to precipitate the free acid (8). In
the first few campaigns, the filtration of this acid intermediate
proved to be the most troublesome operation in the synthesis,
with the filtration of 34 kg on a 33-in. diameter Nutsche
filter routinely taking 3-5 days. Aging the solids at elevated
temperature (85 °C for 4 h) resulted in a faster-filtering
slurry; however, during this process an impurity (8%) was
formed. This impurity was identified as the ethyl ether
cleavage product by HPLC/MS. To exploit the faster fil-
tration without the impurity issue, the stability of acid 8 was
determined at a range of pH under the reaction conditions.
It was found that the solids could be successfully aged at
As a result of the low solubility of 1, low exposure of
the drug candidate was observed in the toxicology studies.
874
•
Vol. 7, No. 6, 2003 / Organic Process Research & Development

Scheme 3
Therefore, several kilograms of the more soluble potassium
salt were required for the toxicology program in which higher
drug exposure was desired. The potassium salt could be
generated in THF using KOt-Bu with a small amount of
H2O.5 In the course of developing this salt formation prior
to scaling, it was noted that the potassium salt was soluble
enough to be recrystallized. This recrystallization proved to
be a very effective purification, especially with regard to
any dibenzylurea formed in the previous step of the synthesis.
The ratio of H2O to THF in this crystallization was critical
to the recovery. In addition, the solution of the potassium
salt provided an ideal point for a 1 micron filtration prior to
isolation of the API.6 As a result, this salt formation was
adopted as a purification step in the synthesis of 1 as the
free acid as well. In practice, the potassium salt was generated
(KOt-Bu, THF, 9 equiv H2O), filtered through a 1 micron
filter at 45 °C, then cooled to 15 °C, granulated for 12 h,
and filtered to provide the potassium salt.7 When the free
acid was the final target, the potassium salt wet cake was
resuspended in H2O (15 L/kg) and acidified with HCl (1
equiv). The solids were slurried and isolated by filtration in
68% yield for the salt formation and break. Although the
product was routinely isolated in very high purity (>99.5%),
this procedure suffered from the resultant slurries being
foamy and giving slow-filtering, clay-like solids, which
required 2 H2O slurries or washes to reduce the residue on
ignition (ROI) content to acceptable levels. Alternatively, a
cheaper potassium salt purification was developed in the lab,
which consisted of adding 1.4 equiv of 50% aqueous KOH
solution to 1 in 7 L/kg THF with 1 vol of i-PrOH and heating
to 50 °C. The resultant solution was filtered, and the salt
crystallized upon cooling to 5 °C. The isolated salt was
dissolved in 12 L/kg of H2O and converted to final product
1 using HCl in an 80% overall yield.
A second alternative for production of acceptably pure
API is to isolate the acylimidazole intermediate during the
amide formation. The acylimidazole precipitates from tri-
ethylamine and can be isolated by filtration. Conversion of
this to the final product by addition of BnNH2 produced high-
purity product in pilot experiments.
The simplest way to produce acceptably pure API is to
add acetone and H2O to the slurry at the conclusion of the
amide formation reaction. The addition of 7 L/kg of acetone
and H2O results in a faster filtration and also proved effective
for the purge of the dibenzylurea byproduct in the lab, giving
clinical-grade API in good yields (>87%) and purities
(>98.5%) for the amide formation.
Compound 1 is highly insoluble in most solvents but can
be recrystallized from either aqueous DMF or HOAc. DMF
was not a preferred final crystallization solvent. Dissolution
in 10 L/kg of HOAc and subsequent dilution with 2 L/kg of
H2O proved very effective at purifying the API, giving a
fast-filtering solid, which was easily washed with H2O. The
purification typically afforded an 86% recovery of 1 (>98.5%)
with residual HOAc levels and residue on ignition analyses
lower than 0.1%.
Ester 7 could be converted directly to amide 1 by treat-
ment with BnNH2 in DMF and DMSO at high temperature.
The DMSO reactions afforded superior yields over the DMF
reactions, with the reaction rate greatly enhanced by the
quantity of BnNH2 used. Reaction of 3 equiv of BnNH2 with
7 in 5 L/kg DMSO at 110 °C afforded a 96% yield of crude
1 (88%). Recrystallization with aqueous HOAc afforded a
69% recovery of pure 1 (>98.0%).
(5) Gassman, P. G.; Schenk, W. N. J. Org. Chem. 1977, 42, 918-920.
(6) All materials going into the final drug substance crystallization are filtered
in solution or neat through a 1 micron filter prior to crystallization per an
internal operating procedure.
(7) The isolated yield of the potassium salt was 77% in a campaign where the
solid salt was dried.
Vol. 7, No. 6, 2003 / Organic Process Research & Development
•
875

Experimental Section
purged 500-gal, glass-lined reactor was charged 29 kg of
6-ethoxy-4-oxo-1,4-dihydro-[1,5]naphthyridine-3-car-
boxylic acid (8), 31 gal of DMF, and 22 kg of CDI. The
mixture was heated to 90 °C and held for 2 h with gas
evolution. The crude mixture was cooled to 35 °C and
sampled.9 Once the reaction was observed to be complete,
14.6 kg of BnNH2 was added, and the mixture was stirred
at 35 °C for 2 h. The reaction was sampled for reaction
completion and then diluted with 192 gal of process H2O.
The slurry was allowed to granulate for 4 h, and then the
product was isolated via filtration. The product was allowed
to dry for 24 h at 45 °C until the KF was 0.5%, yielding
38.9 kg, 97%.
All materials were purchased from commercial suppliers
and used without further purification. All reactions were
conducted under an atmosphere of nitrogen unless noted
otherwise. All reactors were glass-lined steel vessels. Reac-
tions were monitored for completion by removing a small
sample from the reaction mixture and analyzing the sample
by HPLC. HPLC analyses were performed using a YMC
basic column (15 cm × 4.6 mm, 3 µm) and a mobile phase
of 72/28 (v/v) (0.1% H3PO4, 10 mM SDS in H2O)/
acetonitrile. 1H-spectroscopy was performed at 300, 400, and
500 MHz on Bruker-Spectrospin Avance instruments. The
LOD (loss on drying) was determined by concentrating a
weighed sample from the bulk material.
5-Amino-2-ethoxy-pyridine (4).3 To a strirred slurry of
2-chloro-5-nitro-pyridine (76.1 g, 480 mmol) in EtOH (500
mL) was added a solution of sodium ethoxide (21% in EtOH,
196 mL, 576 mmol) over 20 min. After 6 h of additional
stirring at room temperature, EtOH (150 mL), borane
trimethylamine (43.3 g, 576 mmol), and palladium hydroxide
(2.16 g, 50% wet) were successively added, and the resulting
suspension was heated to reflux for 2 h. Upon cooling to
room temperature, the mixture was filtered over Celite, and
the filtrate was concentrated under vacuum to a final volume
of 300 mL. The concentrate was diluted with additional
EtOH (1.50 L), and phosphoric acid (56.5 g, 576 mmol) was
added dropwise. The resulting slurry was filtered, and the
solid material was washed with EtOH (225 mL) and dried
to yield the title compound as a pale yellow crystalline solid
(106 g, 93%).
To a clean and dry 500-gal, glass-lined reactor was
charged 38 kg of 6-ethoxy-4-oxo-1,4-dihydro-[1,5]naphthy-
ridine-3-carboxylic acid benzylamide (1), 80 gal of THF,
13.5 kg of KOt-Bu, and 19 L of USP H2O. The slurry was
heated to reflux and held for 1 h. The solution was then
filtered via a jacketed sparkler filter at a temperature just
below10 reflux to a clean 300-gal, glass-lined reactor. The
filter and lines were rinsed with 10 gal of 55 °C THF. To
the batch was added an additional 80 gal of THF, and the
batch was concentrated atmospherically to 55 gal. The
reaction mixture was allowed to cool and granulate for 16
h. The desired salt (27.4 kg, wet cake, first crop) was isolated
and used in the next step without drying. The mother liquor
from the first crop was diluted with 55 gal of THF and
concentrated atmospherically to a final volume of 15 gal.
The reaction mixture was allowed to cool and granulate for
16 h. A second crop of the desired salt was isolated by
1
filtration (17.4 kg, wet cake). mp 208 °C. H NMR (400
MHz, DMSO): δ 8.72 (s, 1H), 8.03 (d, J ) 9.2, 1H), 7.34-
7.23 (m, 5H), 7.21 (d, J ) 9.2, 1H), 4.54 (d, J ) 5.2, 2H),
4.39 (q, J ) 7.2, 2H), 1.33 (t, J ) 7.2, 3H). 13C NMR (100
MHz, DMSO): δ 172.8, 168.8, 159.2, 150.9, 143.2, 141.2,
141.1, 140.3, 129.0, 127.9, 127.2, 114.8, 112.8, 61.5, 42.4,
15.2.
6-Ethoxy-4-oxo-1,4-dihydro-[1,5]naphthyridine-3-car-
boxylic acid (8). To a clean 1000-gal, glass-lined reactor
was charged 185 kg of ethyl 6-ethoxy-4-oxo-1,4-dihydro-
[1,5]-naphthyridine-3-carboxylic acid, ethyl ester (7), 1480
L of H2O, and 437 kg of 2-B EtOH.8 To the slurry was added
150 kg of 50% NaOH solution and 185 L of H2O at
temperatures below 40 °C. The reaction was heated to reflux
and held for 2 h. The solution was cooled to 45 °C and
transferred through a preheated polish filter to a second clean
1000-gal, glass-lined reactor. The empty reactor and polish
filter were rinsed with 110 L of H2O. At 25 °C, the reaction
filtrate was adjusted to a pH of 2-4 with filtered concen-
trated HCl and 50% NaOH. The slurry was heated to reflux
and held for 1 h. The slurry was stirred for 1 h at 22 °C and
then filtered on a 240-L centrifuge in four loads. Each load
was washed with 95 L of filtered H2O and dried together in
a vacuum tray dryer at 66 °C until the H2O content (KF)
was 0.2%. This procedure afforded a 94% yield (155 kg) of
spec-free 8. mp 260-268 °C. Anal. Calcd for C11H10N2O4:
C, 56.41; H, 4.30; N, 11.96. Found: C, 56.30; H, 3.96; N,
To a 300-gal reactor was charged 120 gal of USP H2O
and 27.4 kg of 6-ethoxy-4-oxo-1,4-dihydro-[1,5]naphthyri-
dine-3-carboxylic acid benzylamide potassium salt (1) (THF-
wet). The slurry was stirred at 19 °C, and 6 L of 37% HCl
was added. The pH was measured to be 2.4, and the slurry
was allowed to granulate for 4 h at 20 °C. The product was
isolated via filtration and was then washed with 2 × 10 gal
of 40 °C H2O.11 The product was dried under vacuum at
45-50 °C for 48 h under a nitrogen bleed. The desired
product was isolated in 42% yield (16.2 kg). mp 199 °C.
Anal. Calcd for C18H17N3O3: C, 66.86; H, 5.30; N, 13.00.
1
Found: C, 66.69; H, 5.18; N, 12.91. H NMR (500 MHz,
DMSO): δ 12.73 (br s, 1H), 10.61 (br t, J ) 5.6, 1H), 8.73
(s, 1H), 8.05 (d, J ) 9.0, 1H), 7.35 (d, J ) 4.4, 2H), 7.27
(dd, J ) 4.4, 8.6, 2H), 7.22 (d, J ) 9.0, 1H), 4.56 (d, J )
5.8, 2H), 4.42 (q, J ) 7.0, 2H), 1.35 (t, J ) 7.1, 3H). 13C
1
11.93. H NMR (400 MHz, DMSO): δ 8.85 (s, 1H), 8.12
(d, J ) 9.2, 1H), 7.33 (d, J ) 9.2, 1H), 4.44 (q, J ) 7.2,
2H), 1.38 (t, J ) 7.2, 3H). 13C NMR (125 MHz, DMSO):
δ 176.6, 166.5, 160.8, 142.8, 137.2, 133.1, 132.0, 118.9,
110.7, 62.1, 14.2.
6-Ethoxy-4-oxo-1,4-dihydro-[1,5]naphthyridine-3-car-
boxylic Acid Benzylamide (1). To a clean and dry nitrogen-
(8) Toluene-denatured ethanol.
(9) Assay was accomplished by derivatization of the acylimidazole to the desired
product with BnNH2 and analyzing by HPLC.
(10) All transfer equipment must be preheated to ensure that the potassium salt
remains in solution during the sparkle step.
(11) This filtration took under 1 h including the wash, which was necessary to
reduce ash levels.
876
•
Vol. 7, No. 6, 2003 / Organic Process Research & Development

NMR (125 MHz, DMSO): δ 174.5, 164.5, 160.1, 141.6,
139.3, 138.6, 132.1, 131.5, 128.4, 127.3, 126.9, 117.5, 113.6,
61.7, 42.1, 14.3.
hexanes-EtOAc-Et3N provided Heck product 11 (Rf ) 0.16
in 2:1 hexanes-EtOAc) as a dark-green oil which solidified
upon storage at 0 °C (434 mg, 1.29 mmol, 28% yield). MS
(CI) m/z: 337 (M + H, 100). 1H NMR (CDCl3): δ 7.97 (s,
1H), 7.80 (d, J ) 9, 1H), 6.68 (d, J ) 9, 1H), 5.70 (t, J )
6, 1H), 4.51 (q, J ) 7, 2H), 3.91 (s, 3H), 3.63 (s, 3H), 3.20
(s, 3H), 3.15 (dd, J ) 16, 7, 1H), 2.93 (dd, J ) 16, 6, 1H),
1.42 (t, J ) 7, 3H).
3-Amino-2-bromo-6-ethoxypyridine (9).12 A solution of
2-ethoxy-5-aminopyridine 4 (3.00 g, 21.7 mmol) was dis-
solved in AcOH (18 mL) at room temperature. NaOAc (1.73
g, 21.1 mmol) was added in a single portion, followed by
bromine (1.04 mL, 1.36 g, 20.2 mmol), added dropwise over
10 min. The thick reaction mixture was stirred for 60 min
and then quenched by portionwise addition to an ice-cold
10% aq NaOH solution (120 mL). This mixture was ex-
tracted with two 150 mL portions of EtOAc. The combined
organic extracts were washed with brine, dried over
Na2SO4, filtered, and concentrated to provide 4.9 g of a
N′-(2-Bromo-6-ethoxypyridin-3-yl)-N,N-dimethylfor-
mamidine (12). Bromide 9 (300 mg, 1.38 mmol) and N,N-
dimethylformamide-dimethylacetal (0.22 mL, 0.20 g, 1.66
mmol) were combined in 8 mL of toluene and placed in a
50 °C oil bath for 3 days. GC/MS indicated incom-
plete conversion; therefore, another portion of dimethyl-
acetal (0.055 mL, 0.049 g, 0.41 mmol) was added. After
warming to 50 °C for another 20 h, the reaction mixture
was cooled to room temperature and concentrated to pro-
vide a dark red oil. The crude product was purified by
chromatography on 15 g of silica gel, eluting with 40:10:1
hexanes-EtOAc-Et3N. The product-containing fractions
were combined and concentrated to provide amidine 12 as
a red solid (295 mg, 1.08 mmol, 78% yield). MS (CI)
1
reddish-orange oil. H NMR analysis showed the desired
product, contaminated with AcOH and some unreacted
starting material. The material was redissolved in EtOAc (150
mL) and washed with H2O, 10% aq NaOH, H2O, and brine.
After drying over Na2SO4, filtering, and evaporation, the
desired product was obtained as an orange oil (4.16 g, 19.2
mmol, 88% yield), which solidified upon storage at 0 °C.
MS (EI) m/z: 218, 216 (M, 30). 1H NMR (CDCl3): δ 7.03
(d, J ) 8, 1H), 6.55 (d, J ) 8, 1H), 4.25 (q, J ) 7, 2H),
1.35 (t, J ) 7, 3H).
1
m/z: 274, 272 (M + H, 100).; H NMR (CDCl3): δ 7.44
(s, 1H), 7.22 (d, J ) 8, 1H), 6.60 (d, J ) 8, 1H), 4.30
(q, J ) 7, 2H), 3.14 (br s, 3H), 3.07 (br s, 3H), 1.35 (t, J )
7, 3H).
2-[(2-Bromo-6-ethoxy-pyridin-3-ylamino)-methylene]-
malonic Acid Diethyl Ester (10). A neat solution of
2-bromo-6-ethoxy-pyridin-3-ylamine (2.20 g, 10.1 mmol)
and 2-ethoxymethylene-malonic acid diethyl ester (2.05
mL, 2.18 g, 10.1 mmol) were heated to 50 °C for 8 h to
form a moist, blackened solid. The solid was recrystal-
lized from a minimum of isopropyl ether. The solid was
filtered and washed with 0 °C ethyl ether to yield opaque,
light pink, needlelike crystals (3.45 g, 88%). mp 166 °C. Rf
) 0.23 (10% EtOAc-hexane). Anal. Calcd for C15H19-
BrN2O5: C, 46.53; H, 4.95; N, 7.23. Found: C, 46.45; H,
3-Amino-6-ethoxy-pyridine-2-carbonitrile (13). To a
stirring solution of 2-bromo-6-ethoxy-pyridin-3-ylamine
(23.0 g, 106 mmol) in DMF (500 mL) was added Zn(CN)2
(18.7 g, 159 mmol), followed by Pd(PPh3)4 (12.20 g, 11
mmol). The mixture was heated to 80 °C for 3 h and then
cooled to 25 °C. The solids were filtered, and the filtrate
was diluted with EtOAc (1.0 L) and washed with 1.0 N HCl
(1.0 L) and H2O (3 × 1.0 L). The organic phase was
collected and dried over MgSO4. The solvent was re-
moved via rotary evaporator, and the resulting solid re-
crystallized from 20:1 hexane-EtOAc (800 mL) to yield
small, rust-colored crystals (12.1 g, 70.0%). mp 104-105
°C. Rf ) 0.65 (10% CH3OH/CH2Cl2). Anal. Calcd for
C8H9N3O: C, 58.88; H, 5.56; N, 25.75. Found: C, 58.72;
1
4.88; N, 7.09. H NMR (400 MHz, CD3OD): δ 11.17 (bd,
J ) 11.4, 1H), 8.33 (d, J ) 13.3, 1H), 7.50 (d, J ) 8.7, 1H),
6.75 (d, J ) 8.7, 1H), 4.35 (q, J ) 6.2, 4H), 4.24 (q, J )
7.3, 2H), 1.38 (t, J ) 7.1, 6H), 1.32 (t, J ) 7.3, 3H). 13C
NMR (300 MHz, CD3OD): δ 168.7, 165.9, 159.9, 151.1,
150.9, 129.8, 129.2, 127.6, 112.7, 111.1, 95.4, 63.3, 60.9,
60.6, 14.6.
1
H, 5.25; N, 25.44. H NMR (400 MHz, CD3OD): δ 7.08
(d, J ) 9.2, 1H), 6.77 (d, J ) 9.2, 1H), 4.25 (q, J ) 7.1,
2H), 3.96 (bs, 2H), 1.33 (t, J ) 7.1, 3H). 13C NMR (300
MHz, CD3OD): δ 156.9, 142.7, 128.3, 118.2, 116.8, 111.5,
62.3, 14.7.
3-(6-Ethoxy-4-methoxy-2-oxo-2H-[1,5]naphthyridin-1-
yl)-3-methoxy-propionic Acid Methyl Ester (11). Bromide
9 (1.0 g, 4.6 mmol), methyl â-methoxyacrylate (1.0 mL, 1.08
g, 9.2 mmol), Pd(OAc)2 (53 mg, 0.23 mmol, 5 mol %), and
tri-(o-tolyl)phosphine (578 mg, 1.84 mmol, 40 mol %) were
combined in a dry, N2-purged flask equipped with a reflux
condenser. Triethylamine (2.3 mL, 1.67 g, 16.5 mmol) was
added, and the flask was placed in a 100 °C oil bath for 5
days. The reaction mixture was cooled to room temperature
and diluted with ca. 50 mL of 1:1 hexanes-EtOAc. The
solution was filtered through Celite and then washed with
two 20-mL portions each of H2O, aq NaHCO3, and brine.
The organic phase was dried over Na2SO4, filtered, and
concentrated to provide 700 mg of a dark-brown solid.
Chromatography on 50 g of silica gel, eluting with 40:10:1
3-(2-Cyano-6-ethoxy-pyridin-3-ylamino)-acrylic Acid
Ethyl Ester (14). To a stirring solution of 3-amino-6-ethoxy-
pyridine-2-carbonitrile (2.0 g, 12.3 mmol) and 3-ethoxy-
acrylic acid ethyl ester (1.95 mL, 1.95 g, 13.5 mmol) in THF
(25 mL) at -78 °C was added NaHMDS (1.0 M in THF,
14.8 mL, 14.8 mmol) dropwise over 20 min. The solution
was allowed to warm to 25 °C and was stirred for 1 h. The
crude reaction mixture was concentrated via rotary evapora-
tor, and the crude residue was purified on a silica gel column,
4:1 hexane-EtOAc. Two major UV active products were
visible. The less-polar UV active product was collected and
concentrated to yield a light-yellow, flaky solid (1.68 g,
52.2%). mp 140-141 °C. Rf ) 0.15 (10% EtOAc-hexane).
(12) den Hertog, J. Recl. TraV. Chim. Pays-Bas 1953, 72, 125-131.
Vol. 7, No. 6, 2003 / Organic Process Research & Development
•
877

Anal. Calcd for C13H15N3O3: C, 59.76; H, 5.79; N, 16.08.
Found: C, 59.69; H, 5.64; N, 15.94. H NMR (400 MHz,
Acknowledgment
1
We thank Dr. Randall DeJong for helpful conversations
and Mr. Matthew L. Jorgensen, Dr. Ricardo Borjas, and Mr.
Douglas W. Heller for analytical support.
CD3OD): δ 10.44 (bd, J ) 11.6, 1H), 7.42 (d, J ) 9.1, 1H),
7.10 (dd, J ) 11.8, 8.3, 1H), 5.02 (d, J ) 8.5, 1H), 4.32 (q,
J ) 7.1, 2H), 4.24 (q, J ) 7.1, 2H), 1.37 (t, J ) 7.3, 3H),
1.31 (t, J ) 7.1, 3H). 13C NMR (300 MHz, CD3OD): δ
169.9, 159.1, 141.0, 137.0, 126.0, 117.9, 116.2, 115.6, 112.5,
92.0, 62.9, 60.2, 14.6.
Received for review July 28, 2003.
OP0341061
878
•
Vol. 7, No. 6, 2003 / Organic Process Research & Development