L. Saikia et al. / Tetrahedron Letters 55 (2014) 1796–1801
1797
O
N
O
N
Ar
O
N
O
N
Ph
NO2
NO2
NO2
NO2
DMSO, reflux
N
N
DMF, reflux, 3 hrs
-NHMe2
N
N
+
+
O
N
N
O
N
Ph
3a
O
N
N
O
N
Ar
3a-h
82%
5a
5a-h
2
2
DMF, rt
overnight stirring
Scheme 2. Synthesis of pyrido[2,3-d]pyrimidines (5a–h) in one step.
-H2
30% conversion
O
N
Ph
O
Ph
N
the final product. The intermediate compound responsible for this
new spot was isolated using column chromatography and its struc-
ture was established as 4a, a Michael type product by single crystal
X-ray analysis. A similar reaction condition was imposed (i.e.
refluxed in DMF) to 4a and to our delight, it offered the same prod-
uct 5a within shorter period of time and thus supported the
involvement of 4a as an intermediate.
NO2
NO2
N
DMF, reflux, 45 min
-NHMe2
N
O
N
N
O
N
92%
5a
4a
Scheme 1. Synthesis of pyrido[2,3-d]pyrimidine 5a.
To study the effect of solvents in the formation of 4a as well as
5a, the reaction was carried out in different solvents (both polar
and non-polar) by stirring the reaction mixture at room tempera-
ture for 12 h followed by refluxing it for 3 h. The results are sum-
marized in Table 1. Table 1 reveals that high boiling aprotic polar
solvents like DMF and DMSO (Table 1, entries 1 & 2) are suitable
for the formation of the final product 5a, but the rate of conversion
towards the intermediate 4a is quite low. In contrast, water is
found to be the best solvent for the formation of the intermediate
4a, but not so satisfactory for furnishing the final product 5a. To
examine the effect of H2O, the reaction was then carried out in a
1:1 mixture of 4a, but not good enough to give 5a in satisfactory
yield. On the basis of all these observations, it can be concluded
that H2O favours the formation of 4a whereas it disfavours the con-
version of 4a to the product (5a). Other solvents like chloroform,
acetonitrile, benzene and methanol were not found to be suitable.
Among all these solvents checked, DMSO was found to be the best
solvent for the synthesis of pyrido[2,3-d]pyrimidine (5a) in a single
step process. Generalization of the reaction was carried out using
various b-nitrostyrenes (3a–h) as shown in Scheme 2 and summa-
rized in Table 2. The reaction did not proceed with aliphatic nitro-
alkenes. The synthesis of 5 was also attempted via a two-step route
(Scheme 3) considering the observation that H2O was the best
solvent for the intermediate (4) formation step whereas DMF and
DMSO were the excellent solvents for the formation of intermedi-
ate (4) to final product (5). It was found that this route was slightly
more beneficial than the single step route when the overall iso-
lated yields of 5 were considered. The 2nd step was carried out
in refluxing DMF although both DMF & DMSO were found to give
almost comparable yield. However, DMF is preferred over DMSO,
as DMSO has an extremely unpleasant smell when refluxed.
Chloro, nitro, hydroxyl, methoxy and heterocyclic moieties
In our initial attempt to investigate the interesting diene behav-
iour of the methylated amidine (2) in [4+2] cycloaddition reaction,
we treated 2 with b-nitrostyrene (3a) (Scheme 1), an electron defi-
cient dienophile. During the process, we observed lower reactivity
of 2 towards 3a than 1. The percentage conversion of 2 in chloro-
form was found to be 35% only (stirring at room temperature for
12 h, (entry 5, Table 1) whereas 1 underwent complete conversion
within 5 minutes (stiring at 0 °C) as demonstrated in our earlier
report.7e This lower reactivity of 2 towards b-nitrostyrene, in spite
of having one extra electron releasing methyl group is contradic-
tory to the general rule of Diels–Alder reaction which suggests en-
hanced reactivity of 2 than 1. This might be due to the reason that
the presence of this extra methyl group hinders the extent of delo-
calization starting from the dimethylamino group and hinders the
planarity. However, the reaction occurs smoothly in refluxing DMF
with complete conversion of 2 within 3 h furnishing 1,3,7-tri-
methyl-6-nitro-5-phenylpyrido[2,3-d]pyrimidine-2,4(1H,3H)-
dione (5a) in 82% yield. The structure of 5a has been established
using FTIR spectroscopy, MS (ESI), 1H & 13C NMR spectroscopy
and elemental analyses. To gain some insight into the mechanism
of the reaction, the reaction was monitored at different time inter-
vals using TLC. As the reaction progressed, we noticed an extra spot
in TLC with distinctly lower Rf value than that of 5a, whose inten-
sity decreased gradually over time. The same spot was observed in
TLC as the single spot other than the starting compounds when we
carried out the reaction at room temperature for 12 h (Scheme 1),
although only 30% conversion of the starting materials was ob-
served here. These observations suggest that this extra spot in
TLC might be due to the formation of an intermediate which
underwent further reaction under the reaction condition to furnish
Table 1
Screening of solvents for Scheme 1
Entry
Solvent
Reaction condition
Refluxed for 3 ha
Stirred at rt for overnight (12 h)
Conversion of 2 (%)
Yield of 4a (%)
Yield of 5a (%)
Conversion of 2 (%)
Yield of 4a (%)
Yield of 5a (%)
1
2
3
4
5
7
8
9
DMF
DMSO
H2O
DMF–H2O (1:1)
CHCl3
CH3CN
C6H6
30
32
100
100
35
38
15
75
100
100
100
100
100
100
100
100
No product formation
No product formation
No product formation
No product formation
No product formation
No product formation
No product formation
No product formation
100
100
100
100
50
68
60
92
<10
<10
8
82
85
22b
28b
NF
12
100
>90
45
<10
18b
<10b
CH3OH
65
a
Reaction mixture was stirred at room temperature for 12 h followed by reflux for 3 h.
Some other unidentified products were formed.
b