An efficient synthesis of dibenzo[c,f]-2,7-naphthyridine ring system through design of experiments

Article abstract of DOI:10.1021/op700009t

The dibenzo[c,f]-2,7-naphthyridine ring system was found to be of biological interest, but had limited synthetic accessibility. The initial compound of interest, 10,11-dimethoxy-4-methyldibenzo[c,f]-2,7-naphthyridine-3, 6-diamine, was obtained in <25% yie

Full text of DOI:10.1021/op700009t

Organic Process Research & Development 2007, 11, 450−454
Communications to the Editor
An Efficient Synthesis of Dibenzo[c,f]-2,7-naphthyridine Ring System through
Design of Experiments
Ariamala Gopalsamy,*,† Mengxiao Shi,† and Ramaswamy Nilakantan‡
Chemical and Screening Sciences, Wyeth Research, Pearl RiVer, New York 10965, U.S.A
Abstract:
The dibenzo[c,f]-2,7-naphthyridine ring system was found to
be of biological interest, but had limited synthetic accessibility.
The initial compound of interest, 10,11-dimethoxy-4-meth-
yldibenzo[c,f]-2,7-naphthyridine-3,6-diamine, was obtained in
<25% yield by reacting 4-chloro-6,7-dimethoxy-quinoline-3-
carbonitrile with 2-methyl-benzene-1,3-diamine. A reliable high-
yielding procedure was identified through the use of design of
experiments (DOE). The effects of stoichiometry of reagents,
Figure 1. HTS hit - 10,11-dimethoxy-4-methyldibenzo[c,f]-2,7-
naphthyridine-3,6-diamine.
catalyst, and temperature were explored in this study. The DOE
optimization suggested temperatures higher than the boiling
Scheme 1. Original synthetic route for 1
point of our solvent. Hence, microwave heating was used,
resulting in 80% yield of the desired product.
Introduction
During the course of a high-throughput campaign for a
high-priority discovery project, we identified 10,11-dimethoxy-
4-methyldibenzo[c,f]-2,7-naphthyridine-3,6-diamine, 1, (Fig-
ure 1) as a potent compound with favorable biological profile.
Further database mining was unable to provide additional
analogues with the dibenzo[c,f]-2,7-naphthyridine scaffold.
The high-throughput screening hit 1 had been isolated in a
previous project as a by-product in <25% yield. To
understand the structure activity relationships of this single-
ton, and to explore the potential of this series, we needed a
reliable high-yielding synthetic procedure.
The original synthetic route that produced 1 as a side
product was one that we commonly employ to generate
4-anilino-3-quinolinecarbonitriles1 such as 8 (Scheme 1). The
method involves heating an equimolar mixture of 4-chloro-
3-quinolinecarbonitrile, aniline, and pyridine hydrochloride
in ethoxyethanol at 130 °C for 8 h. This method has found
utility in several instances to generate various other anilino
hetrocyclic carbonitriles such as 7-phenylamino-thieno[3,2-
b]pyridine-6-carbonitriles,2a 4-anilino-benzo[b][1,5]-naph-
thyridine-3-carbonitrile,2b 4-anilino-benzo[b][1,8]-naphthy-
ridine-3-carbonitrile,2b 4-anilino-benzothieno[3,2-b]pyridine-
3-carbonitrile,2c and 4-anilino-benzofuro[3,2-b]pyridine-3-
carbonitrile.2c However, in this instance, the electron-rich
nature of the aniline employed led to the isolation of 1 along
with the expected product 8. While the mechanism for forma-
tion of 1 can be postulated by invoking a carbon nucleophile
of the electron-rich aniline, factors that would favor the
isolation of 1 as the major product remained a challenge.
Instead of attempting extensive ad hoc optimization, we
chose to use design of experiments (DOE) to evaluate the
effects of various synthetic conditions. DOE is used to
explore the response surface using a small number of
carefully chosen experiments. The experiments are carried
out, and a regression equation is fit to the resulting data.
The regression model can then be used to find the optimum
* To whom correspondence should be addressed. Telephone: +845-602-2841.
Fax: +845-602-5045. E-mail: gopalsa@wyeth.com.
† Medicinal Chemistry.
‡ Cheminformatics.
(1) Wissner, A.; Berger, D. M.; Boschelli, D. H.; Floyd, M. B.; Greenberger,
L. M.; Gruber, B. C.; Johnson, B. D.; Mamuya, N.; Nilakantan, R.; Reich,
M. F.; Shen, R.; Tsou, H.; Upeslacis, E.; Wang, Y.; Wu, B.; Ye, F.; Zhang,
N. J. Med. Chem. 2000, 43, 3244-3256.
(2) (a) Boschelli, D. H.; Wu, B.; Sosa, A. B.; Durutlic, H.; Chen, J. J.; Wang,
Y.; Golas, J. M.; Lucas, J.; Boschelli, F. J. Med. Chem. 2005, 48, 3891-
3902. (b) Wang, Y. D.; Boschelli, D. H.; Johnson, S.; Honores, E.
Tetrahedron 2004, 60, 2937-2942. (c) Boschelli, D. H.; Ye, F. J.
Heterocycl. Chem. 2002, 39, 783-788.
450
•
Vol. 11, No. 3, 2007 / Organic Process Research & Development
10.1021/op700009t CCC: $37.00 © 2007 American Chemical Society
Published on Web 03/27/2007

Figure 2. Plot of the coefficients.
Table 1. Factors and ranges
In order to understand the response surface better, we plot-
ted the predicted yields as a function of two variables in a
series of contour plots as seen in Figure 3. In the four plots,
yield is plotted as a function of aniline and catalyst stoichi-
ometry. Four combinations of temperature and time are em-
ployed in the four different plots. An analysis of the plots
shows that the strongest dependency of yield is on tempera-
ture, with higher temperatures giving higher yields. Similarly,
higher stoichiometries of both aniline and catalyst give higher
yields. As per the model, in order to improve our yield be-
yond 65% (the highest yield we obtained from the initial set
of experiments), we needed to increase the stoichiometry of
aniline, catalyst, temperature, or various combinations thereof.
Figure 4 shows a contour plot of aniline versus catalyst at
130 °C with an incubation time of 24 h. It can be seen that
the model predicts yields in the 90% range with high equiva-
lents of both aniline and catalyst. This prompted us to modify
the initial ranges that we set for the factors aniline (1-12
equiv) and catalyst (0-9 equiv) while we maintained the
temperature at the boiling point of the solvent (130 °C). The
results of these experiments are shown in Table 3. As seen
from Table 3, at 9 equiv each of aniline and catalyst we
obtained ∼70% yield as predicted by the model. This was
an acceptable yield for our purposes, but such high equiva-
lents of aniline and catalyst were not too appealing since it
would involve additional work to remove excess reagents at
the end of the reaction, particularly during scale-up. We were
then left with temperature and incubation time as the only
remaining factors. Reaction times of >24 h are inconvenient,
thus leaving temperature as the only remaining factor to vary.
The maximum temperature used in the design is indeed
the boiling point of the solvent, 2-ethoxyethanol. To achieve
higher temperatures under conventional heating, we would
need to use a different, higher-boiling solvent. Switching the
solvent at this stage of optimization might cause unforesee-
able variations and was deemed undesirable. Taking these
factors into consideration, we decided to try the reaction
under microwave irradiation conditions. Use of microwave
irradiation in organic synthesis has gained popularity in
recent years, particularly for parallel synthesis. The higher
reaction temperatures that can be attained using microwave
irradiationswell above the boiling point of the solventswas
especially attractive in our situation. Table 4 shows results
factor
conc. units
range
aniline
equiv
equiv
°C
1-4
0-5
50-130
8-24
catalyst
temperature
time
h
conditions to maximize the yield. This approach has been
extensively used in several disciplines3 and is now commonly
being adopted by synthetic chemists as well.4
Results and Discussion
Experimental design, model building, analysis, and op-
timization of reaction conditions were all performed using
the MODDE5 software application. Four different factors6
were considered in our studysstoichiometry of aniline 7,
stoichiometry of pyridine hydrochloride (henceforth referred
to as catalyst), temperature, and reaction time. The initial
ranges of these factors are shown in Table 1. A two-level
full factorial design was chosen to screen the factor space.
A set of 16 experiments plus three replicates of the center
point was carried out to develop a model of the response
surface. The design and the yields from these experiments
are shown in Table 2. The three center point experiments
had a mean value of 26.1 with a standard deviation of 3.5.
The data was used to build a predictive model using multiple
linear regression. We obtained a good fit, with R2 ) 0.95
and Q2 ) 0.79. Figure 2 shows a plot of the coefficients of
the regression model. It can be seen that temperature,
stoichiometry of aniline, stoichiometry of catalyst, and two
cross terms are significant.
(3) (a) Morris, N. D.; Mallouk, T. E. J. Am. Chem. Soc. 2002, 124, 11114-
111121. (b) Le Mapihan, K.; Vial, J.; Jardy, A. J. Chromatogr., A 2004,
1061, 149-158. (c) Arauzo-Bravo, M. J.; Shimizu, K. J. Biotechnol. 2003,
105, 117-133. (d) Halliwell, C. M.; Cass, A. E. Anal. Chem. 2001, 73,
2476-2483.
(4) (a) Stazi, F.; Palmisano, G.; Turconi, M.; Clini, S.; Marco, S. J. Org. Chem.
2004, 69, 1097-1103. (b) Chen, J. J.; Nugent, T. C.; Lu, C. V.; Kondapally,
S.; Giannousis, P.; Wang, Y.; Wilmot, J. T. Org. Process Res. DeV. 2003,
7, 313-317. (c) Gooding, O. W. Curr. Opin. Chem. Biol. 2004, 8, 297-
304.
(5) MODDE 6.0; Umetrics: Kennelon, NJ. http://www.umetrics.com.
(6) Preliminary experiments carried out in parallel using the conditions detailed
in method A, but varying the concentration (0.025 M, 0.05 M, 0.1 M and
0.2 M) did not show any change. As a result, a detailed study was not
carried out using concentration as a factor.
Vol. 11, No. 3, 2007 / Organic Process Research & Development
•
451

Table 2. Full factorial design: experimental matrix and measured response
temp time yield
temp time yield
exp no exp name aniline catalyst (°C)
(h)
(%)
exp no exp name aniline catalyst (°C)
(h)
(%)
1
2
3
4
5
6
7
8
9
N1
N2
N3
N4
N5
N6
N7
N8
N9
N10
1
4
1
4
1
4
1
4
1
4
0
0
5
5
0
0
5
5
0
0
50
50
50
8
8
8
8
8
8
8
8
24
24
0
0
0
0
0
23
33
65
0
11
12
13
14
15
16
17
18
19
N11
N12
N13
N14
N15
N16
N17
N18
N19
1
4
1
4
1
4
2.5
2.5
2.5
5
5
0
0
5
5
2.5
2.5
2.5
50
50
130
130
130
130
90
24
24
24
24
24
24
16
16
16
0
0
32
47.3
31.4
64
22
27.8
28.4
50
130
130
130
130
50
90
90
10
50
0
Table 3. Iterative design: experimental matrix and measured response
temp time yield
temp time yield
exp no exp name aniline catalyst (°C)
(h)
(%)
exp no exp name aniline catalyst (°C)
(h)
(%)
1
2
3
4
5
6
7
8
N20
N21
N22
N23
N24
N25
N26
N27
1
4
1
0
0
5
5
0
3
0
9
130
130
130
130
130
130
130
130
8
8
8
8
8
8
8
8
0
23
33
65
27
55
10
71
9
10
11
12
13
14
15
16
N28
N29
N30
N31
N32
N33
N34
N35
1
4
1
0
0
5
5
0
3
0
9
130
130
130
130
130
130
130
130
22
22
22
22
22
22
22
22
32
47
31
64
38
73
61
69
4
4
12
12
2
12
12
2
9
9
Table 4. Experiments carried out under microwave irradiation conditions
temp time yield
temp time yield
exp no exp name aniline catalyst (°C)
(h)
(%)
exp no exp name aniline catalyst (°C)
(h)
(%)
1
2
N36
N37
1
2
1
1
210
210
1
1
65.5
80
3
4
N38
N39
4
2
1
1
210
130
1
1
79
20
of our microwave experiments. As can be seen from the table,
we were able to achieve 80% yield by increasing the reaction
temperature to 210 °C in a microwave, without the need to
use high aniline or catalyst concentrations. Experiment N39
in Table 4 was carried out as a control experiment to judge
the microwave effect in the absence of high temperature. It
clearly indicates that the high yield obtained can be attributed
to the effect of higher temperature.
the residue was recrystallized from EtOAc to give 40.0 g
(72%) of 2-cyano-3-(3,4-dimethoxyphenylamino)acrylic acid
ethyl ester 4 as a tan solid. Mp 166-170 °C; MS (ES+):
m/z 277.2 [M + H].
6,7-Dimethoxy-4-oxo-1,4-dihydroquinoline-3-carboni-
trile (5).1 A stirred mixture of 2-cyano-3-(3,4-dimethoxy-
phenylamino)acrylic acid ethyl ester 4 (40 g, 145 mmol) and
1.2 L of Dowtherm A was heated at reflux under N2 for10
h. After cooling to 50 °C the mixture was diluted with
hexane. The product was filtered off, washed with hexane
followed by methylene chloride, and dried to give 21.1 g
(63%) of 6,7-dimethoxy-4-oxo-1,4-dihydroquinoline-3-car-
Conclusion
In conclusion, we have identified a set of practical high-
yielding reaction conditions for the synthesis of the diben-
zo[c,f]-2,7-naphthyridine ring system. The effects of stoi-
chiometry of reagents, catalyst, and temperature were
explored using DOE techniques. The DOE optimization
suggested higher temperatures beyond the boiling point of
our solvent. Hence, microwave heating was used, resulting
in an 80% yield of the desired product. Although the use of
microwave limits the scale-up at present, the method is useful
for rapid synthesis of analogues.
1
bonitrile 5 as a brown solid. Mp 330-350 °C; H NMR
(DMSO-d6): δ 12.57 (s,1H), 8.59 (s, 1H), 7.44 (s, 1H), 7.03
(s, 1H), 3.89 (s, 3H), 3.87 (s, 3H); MS (ES+): m/z 231.0
[M + H].
4-Chloro-6,7-dimethoxyquinoline-3-carbonitrile (6).1 A
stirred mixture of 6,7-dimethoxy-4-oxo-1,4-dihydroquinoline-
3-carbonitrile 5 (20 g, 87 mmol) and 87 mL of POCl3 was
heated at reflux for 2 h. Volatile materials were removed
under vacuum at about 70 °C. The residue was stirred at
0 °C with methylene chloride and H2O; solid K2CO3 was
carefully added until the pH was 8-9. After stirring for 30
min at 25 °C the organic layer was separated, washed with
H2O, dried, filtered through Celite, and concentrated to give
19.8 g (92%) of an off-white solid. A sample recrystallized
from CH2Cl2 gave 4-chloro-6,7-dimethoxyquinoline-3-car-
Experimental Section
2-Cyano-3-(3,4-dimethoxyphenylamino)acrylic Acid Eth-
yl Ester (4).1 A mixture of 3,4-dimethoxyaniline 2 (30.6 g,
200 mmol), ethyl (ethoxymethylene)cyano acetate 3 (33.8
g, 200 mmol), and 80 mL of toluene was stirred at 100 °C
for 1 h at 125 °C for 15 min. After evaporation of the toluene,
452
•
Vol. 11, No. 3, 2007 / Organic Process Research & Development

Figure 4. Contour plot showing dependency on aniline and
catalyst at 130 °C for 24 h.
0.10 mmol), 2,6-diaminotoluene 7 (49 mg, 0.40 mmol), and
pyridine hydrochloride (58 mg, 0.50 mmol) (catalyst) in 2
mL of 2-ethoxyethanol was heated at 130 °C for 8 h. 10,11-
Dimethoxy-4-methyldibenzo[c,f]-2,7-naphthyridine-3,6-di-
amine 1 was formed in 65% yield (based on LC/MS
analysis).
Experiments listed in Tables 2, 3, and 4 were carried out
in parallel. All the experiments listed in a given table were
performed on the same day, while the experiments in
different tables were performed on different days.
Method B. A stirred mixture of 4-chloro-6,7-dimethoxy-
quinoline-3-carbonitrile 6 (25 mg, 0.10 mmol), 2,6-diami-
notoluene 7 (24 mg, 0.20 mmol), and pyridine hydrochloride
(12 mg, 0.10 mmol) in 2 mL of 2-ethoxyethanol was heated
in a sealed tube at 210 °C in a microwave oven (Biotage,
Initiator) for 1 h using a power limit of 300 W and a pressure
inside the reaction vessel of less than 20 psi. 10,11-
Dimethoxy-4-methyldibenzo[c,f]-2,7-naphthyridine-3,6-di-
amine 1 was formed in 80% yield (based on LC/MS
analysis). The reaction mixture was concentrated under
reduced pressure, diluted with ethyl acetate, and washed with
brine. The organic layer was collected, dried over sodium
sulfate, and concentrated. The crude product was purified
by flash column chromatography using silica gel. Isolated
yield: 72%. HPLC: Rt ) 2.13 min; MS (ES+): m/z 335.1
[M + H]. HRMS: 335.15122 [M + H]; 335.15026 [calcu-
lated]; 1H NMR (DMSO-d6): δ 9.4 (1H, s), 8.4 (1H, d), 8.2
(1H, s), 7.5 (1H, s), 6.9 (1H, br), 6.8 (1H, d), 5.4 (2H, br),
Figure 3. Contour plots showing dependency on temperature.
bonitrile 5 as an off-white solid. Mp 220-223 °C; 1H NMR
(DMSO-d6): δ 8.98 (s, 1H), 7.54 (s, 1H), 7.42 (s, 1H), 4.02
(s, 3H), 4.01 (s, 3H).
10,11-Dimethoxy-4-methyldibenzo[c,f]-2,7-naphthyri-
dine-3,6-diamine (1). Method A. A stirred mixture of
4-chloro-6,7-dimethoxyquinoline-3-carbonitrile 6 (25 mg,
Vol. 11, No. 3, 2007 / Organic Process Research & Development
•
453

4.0 (3H, s), 3.9 (1H, s), 2.4 (3H, s); IR (cm-1). Cyano group’s
absorption around 2200 cm-1 is not observed.
data generation, Dr. John Ellingboe and Dr. Diane Boschelli
for helpful suggestions, and the discovery analytical chem-
istry group for the characterization of the synthetic samples.
4-[(3-Amino-2-methylphenyl)amino]-6,7-dimethoxyquin-
oline-3-carbonitrile (8): 1H NMR (DMSO-d6): δ 9.3 (1H,
s), 8.3 (1H, d), 7.8 (1H, s), 7.3 (1H, s), 6.9 (1H, m), 6.6
(1H, d), 6.5 (1H, d), 5.0 (2H, br), 4.0 (3H, s), 3.9 (1H, s),
1.9 (3H, s), HRMS: 335.14954 [M + H]; 335.15026
[calculated]; IR (cm-1). Cyano group’s absorption around
2200 cm-1
Supporting Information Available
1H NMR, 13C NMR, and IR spectroscopic data for
compounds 1 and 8. This material is available free of charge
via the Internet at http://pubs.acs.org.
Received for review January 8, 2007.
OP700009T
Acknowledgment
We thank Mr. Mike Kelly for help with the analytical
454
•
Vol. 11, No. 3, 2007 / Organic Process Research & Development