2
K. K. Gnanasekaran et al. / Tetrahedron Letters xxx (2015) xxx–xxx
Most reactions proceeded
at
75 °C,
while 2,3,4,5,6-
pentafluorobenzoyl chloride (3e) gave a near quantitative yield of
5e after 1 h at 23 °C (Table 2). The reaction workup involved
removal of DMF under vacuum, dilution with distilled water, and
filtration. The resulting solid was then stirred with ethanol at
reflux, cooled, and filtered. While the benzimidazoquinazolinone
product did not dissolve significantly in hot ethanol, soluble
impurities were removed by this treatment. The low solubility of
the products in common organic solvents rendered
Figure 1. Benzo[4,5]imidazo[2,1-b]quinazolin-12(5H)-one (1) and benzo[4,5]imi-
dazo[1,2-a]quinazolin-5(7H)-one (2) antitumor compounds.
chromatographic purification ineffective as
a
means of
purification. Poor mass balances were obtained when this was
attempted.
exploited the 1,3-disposed nucleophilic centers in 2-aminobenzim-
idazole with the similarly positioned electrophilic sites in a
2-haloaroyl chloride to generate the central six-membered ring.
Interestingly, two regioisomeric products were possible from the
current reaction, one derived from initial acylation of the amino
substituent on the benzimidazole to give structures related to 2
and the other from acylation at the saturated benzimidazole
nitrogen8a,12 to give regioisomers similar to 1. We anticipated
that treatment of 2-fluoro-5-nitrobenzoyl chloride (3a) with
2-aminobenzimidazole (4) in the presence of base would proceed
by a sequence involving acylation of the amino function of 4
followed by an SNAr ring closure to give substituted derivatives
of 2. In practice, however, product 5a, with the ring system
found in 1, was produced. Using N,N-dimethylformamide (DMF)
as the solvent,11b 3a and 4 were mixed in the presence of
different bases at ꢀ10 °C, and then heated. When complete, the
reaction was worked up and purified to give the yields of 5a
shown in Table 1. Organic bases (TEA, pyridine, and DIPEA) in
DMF at 140 °C failed to give acceptable conversions, but sodium
and potassium carbonate as well as the corresponding
bicarbonates were found to promote complete conversion at
140 °C within 1 h. Since sodium bicarbonate is mild as well as
inexpensive, we sought to optimize the temperature using this
base. Variation of this parameter revealed that the reaction
proceeded at temperatures as low as 75 °C, but slowed
dramatically at 45 °C and 60 °C. Thus, our optimized procedure
involved treating 2-aminobenzimidazole (1.2 equiv) with the acid
chloride (1.0 equiv) in the presence of NaHCO3 (2 equiv) in DMF
at ꢀ10 °C for 30 min, followed by heating to 75 °C for 0.5–1 h.
Application of these conditions to substrates 3a–h furnished
high yields of the target heterocycles 5a–h, allowing slightly
broader reaction scope with respect to substitution on the acid
chloride than our previous synthesis of pyrazoloquinazolinones.11b
In order to broaden the scope of the reaction, we prepared
2-amino-5,6-dimethylbenzimidazole (6) from 4,5-dimethylben-
zene-1,2-diamine and cyanogen bromide according to a literature
procedure.13 Reaction of 3a–g with
6 using our standard
protocol generated the corresponding 8.9-dimethylbenzimidazo-
quinazolinones 7a–g in high yields. Similarly, nicotinoyl chloride
(3h) afforded 8,9-dimethylbenzimidazopyridopyrimidin-5-one 7h
(Table 3). Again, the perfluorinated acid chloride afforded excellent
conversion to the acylation–SNAr product at 23 °C.
Previous syntheses of N5- or N6-alkylated benzo[4.5]imidazo-
[2,1-b]quinazolin-12-ones were reported with unambiguous char-
acterization data,10 though most of these papers did not report 13
C
NMR spectra. As the current compounds are not substituted at
either of these sites, spectral interpretation proved considerably
more difficult. While the 1H NMR spectra were reasonably sharp,
it was observed that many of the absorptions in the 13C NMR spec-
tra were broadened to a point where they could not be observed. It
is interesting to note that benzimidazoles normally exist as tau-
tomeric structures having partial double bond character between
C2–N1 and C2–N3,14 with H exchanged between the two nitrogens.
We have observed broadened 13C NMR signals resulting from this
phenomenon in a previous synthesis of these compounds.15 In
the current systems, tautomerization involved a proton shift
between N6 in the benzimidazole moiety and N5 exocyclic to the
benzimidazole; due to its tertiary structure and electron-deficient
substitution, N11 was not involved in this tautomerization. The NH
proton was often not observed in the 1H NMR, presumably due to
H-bonding with the solvent, which would shift the signal well
beyond the normal chemical shift range.16 The broadened signals
in the 13C NMR spectra for the current products clearly indicated
that the bonding situation in these materials was not straightfor-
ward. Earlier studies failed to report 13C NMR spectra, which is
understandable since our attempts to acquire these data were
hampered by the tautomeric nature and low solubility of the
Table 1
Reaction optmization
Table 2
Cyclizations to form products 5a–h
Entry
Base (2 equiv)
Temp (°C)
Time (h)
Yield (%)a
1
2
3
4
5
6
TEA
Pyridine
DIPEA
140
140
140
140
140
140
140
75
12.0
12.0
12.0
0.5
1.0
1.0
1.0
1.0
5.0
5.0
15
18
26
88
88
88
88
88
53
38
Substrate
X
Y
Z
Pdt
Z
Yield (%)a
3a
3b
3c
3d
3e
3f
F
F
F
F
F
F
F
Cl
CH
CH
CH
CH
CF
CH
CH
N
5-NO2
H
5-F
5a
5b
5c
5d
5e
5f
2-NO2
H
2-F
88
87
86
93
96b
76
78
90
K2CO3
Na2CO3
KHCO3
NaHCO3
NaHCO3
NaHCO3
NaHCO3
6-F
1-F
7
4,5,6-F
4-Me
4-Br
H
1,2,3-F
3-Me
3-Br
H
8b
9
60
45
3g
3h
5g
5h
10
a
a
Isolated yields.
Optimized conditions.
Isolated yields.
Reaction was complete in 1 h at 23 °C.
b
b
Gnanasekaran, Krishna Kumar
