The reactivity of the diaminomaleonitrile-based imines containing hydroxyphenyl substituents with diverse aromatic aldehydes has been explored for the synthesis of novel highly substituted nitrogen heterocycles, which are considered privileged scaffolds in drug discovery. We report here a simple and efficient method for the regiocontrolled synthesis of a variety of 2-aryl-5-cyano-1-(2-hydroxyaryl)-1H-imidazole-4-carboxamides from 2-hydroxybenzylidene imines and aromatic aldehydes. Computational studies on the reaction path revealed that the regioselectivity of the reaction toward the formation of imidazole derivatives instead of 1,2-dihydropyrazines, most likely via a diaza-Cope rearrangement, is driven by the 2-hydroxyaryl group in the scaffold. The latter group promotes the intramolecular abstraction and protonation process in the cycloadduct intermediate, triggering the evolution of the reaction toward the formation of imidazole derivatives.
The reactivity of the diaminomaleonitrile-based imines containing hydroxyphenyl substituents with diverse aromatic aldehydes has been explored for the synthesis of novel highly substituted nitrogen heterocycles, which are considered privileged scaffolds in drug discovery. We report here a simple and efficient method for the regiocontrolled synthesis of a variety of 2-aryl-5-cyano-1-(2-hydroxyaryl)-1H-imidazole-4-carboxamides from 2-hydroxybenzylidene imines and aromatic aldehydes. Computational studies on the reaction path revealed that the regioselectivity of the reaction toward the formation of imidazole derivatives instead of 1,2-dihydropyrazines, most likely via a diaza-Cope rearrangement, is driven by the 2-hydroxyaryl group in the scaffold. The latter group promotes the intramolecular abstraction and protonation process in the cycloadduct intermediate, triggering the evolution of the reaction toward the formation of imidazole derivatives.
Diaminomaleonitrile
(DAMN) is a versatile building block in organic
synthesis, extensively used as a precursor in the preparation of a
large variety of nitrogen heterocycles,[1] including pyrimidines,[2] pyrazines,[3,4] purines,[5−7] adenines,[8] imidazoles,[9,10] and pyrroles,[11,12] among others. The synthetic scheme
to achieve these compounds usually involves the modification of one
of the amino groups in DAMN by the reaction with aldehydes, orthoformate,
acid chlorides, acid anhydrides, or isocyanates, followed by an intramolecular
condensation reaction.[1] In other cases,
before the condensation, the modification of the remaining free amino
group is performed to expand the chemical space, as well as to get
access to more complex heterocycles.Among the DAMN-based intermediates
employed in those approaches,
benzylidene imines 1 proved to be versatile and readily
accessible intermediates for the synthesis of highly substituted imidazoles
and pyrazines (Figure ).[13] These nitrogen heterocycles are considered
privileged scaffolds in drug discovery and are the pharmacophore in
numerous successful drugs in clinical use, such as eprosartan (antihypertensive
agent), metronidazole (antibiotic and antiparasitic agent), and gilteritinib
(anticancer agent), among others.[14−23]
Figure 1
Examples
of drugs in clinical use containing imidazole and pyrazine
scaffolds and chemical structures of DAMN and DAMN-based benzylidene
imines 1.
Examples
of drugs in clinical use containing imidazole and pyrazine
scaffolds and chemical structures of DAMN and DAMN-based benzylidene
imines 1.Examples of synthetic
approaches employing benzylidene imines 1 are disclosed
in Scheme . The treatment
of 1 with either MnO2,[9] nicotinamide/N-chlorosuccinimide (NCS),[24] or copper-modified
manganese oxide-based octahedral molecular sieves (CuOx/OMS-2)[10] afforded 2-aryl-4,5-dicyanoimidazoles 2 in yields ranging from 48 to 93%. The reaction is suitable
for a large variety of substituents (electron-donating and electron-withdrawing
groups), as well as positions (ortho, meta, and para) in the ring. On the other hand, substituted
pyrazine carboxamides 5 could be synthesized in two steps
from dibenzylidene imines 3, which are obtained from
the treatment of 1 with benzaldehydes in the presence
of triethylamine in 68–73% yields and with concomitant hydrolysis
of one of the nitrile groups.[25] Thermal
cyclization of 3 (40–90% yields) and subsequent
oxidation of the resulting 1,2-dihydropyrazine derivatives 4 (80–90% yields) gave access to compounds 5 containing
distinct para-substituted phenyl groups at positions
5 and 6 in the pyrazine ring. No examples of meta- and ortho-substituted derivatives were reported.
Scheme 1
Synthesis of Imidazoles 2 and Pyrazine Carboxamides 5 from DAMN-Based Benzylidene Imines 1
Reagents, conditions, and yields:
(a) MnO2, EtOAc/MeCN, Δ, 50–70%; (b) nicotinamide,
NCS, DMF, 50–60%; (c) CuOx/OMS-2, DMA, 120 °C,
48–93%; (d) ArCHO, Et3N, MeOH, EtOH, MeCN or PhH,
−5 °C to RT; (e) DMSO, 80 °C; and (f) MnO2, DMF, 60 °C or H2O2, MeOH, 50 °C.
Synthesis of Imidazoles 2 and Pyrazine Carboxamides 5 from DAMN-Based Benzylidene Imines 1
Reagents, conditions, and yields:
(a) MnO2, EtOAc/MeCN, Δ, 50–70%; (b) nicotinamide,
NCS, DMF, 50–60%; (c) CuOx/OMS-2, DMA, 120 °C,
48–93%; (d) ArCHO, Et3N, MeOH, EtOH, MeCN or PhH,
−5 °C to RT; (e) DMSO, 80 °C; and (f) MnO2, DMF, 60 °C or H2O2, MeOH, 50 °C.Exploring the reactivity of the DAMN-based benzylidene
imines 1 containing hydroxyl substituents in the phenyl
ring with
aromatic aldehydes, we found that while meta and para isomers give 1,2-dihydropyrazines 4, the
same does not happen with the corresponding ortho-substituted derivatives,
compounds 6, which afforded compounds 7 (Scheme ). We report here
a simple and efficient method for the regiocontrolled synthesis of
a variety of 2-aryl-5-cyano-1-(2-hydroxyaryl)-1H-imidazole-4-carboxamides 7 from 2-hydroxybenzylidene imines 6. The role
of the 2-hydroxyaryl group in the regioselectivity of the reaction
was also studied by computational studies on the reaction path.
Scheme 2
Synthetic Approach Explored in This Work
Results
and Discussion
Synthesis of Compounds 7
First, the effect
of the position of the hydroxyl substituent in the phenyl ring of
the benzylidene imines was explored. To this end, the reactivity of
the three possible isomers, compounds 6a, 8, and 9, which were prepared as previously described,[9] with 2-chlorobenzaldehyde was studied (Figure ). The reaction was
performed using 2 equiv of the aldehyde, in the presence of 2 equiv
of triethylamine using ethanol as the solvent and at room temperature
for 96 h. The results showed that the reaction performed with 2-hydroxybenzylidene
imine 6a gave mainly compound 7b, while
the formation of 1,2-dihydropyrazines was not experimentally observed.
By contrast, for benzylidene imine 8, a mixture of the
corresponding diamine derivative 3 and 1,2-dihydropyrazines 4 was obtained, while the formation of the corresponding imidazole
derivative was not detected. In addition, heating a deuterated dimethyl
sulfoxide (DMSO-d6) solution of the latter
mixture at 80 °C for 30 min led to compound 10 as
the only reaction product (Figure ). Finally, the reaction carried out with imine 9 gave a complex mixture. The formation of compounds 7b and 10 was corroborated by a combination of
1D (1H, 13C, and DEPT) and bidimensional (HMQC
and HMBC) NMR experiments. For compound 7b, the most
remarkable NMR data are (i) the signals corresponding to the methylene
group at 5.09 ppm (singlet, integration by two hydrogen atoms) and
46.2 ppm in the 1H and 13C NMR spectra, respectively,
and (ii) the strong correlation signals of the methylene group with
the CN group (110.7 ppm) and the quaternary carbon atom of the phenyl
group (147.6 ppm) in the HMBC spectrum (see the Supporting Information for further details). The NMR spectra
of compound 10 did not reveal the presence of a methylene
group. Instead, a characteristic CH group is identified at 6.16 and
48.6 ppm in the 1H and 13C NMR spectra, respectively.
Taken together, these outcomes revealed that the 2-phenol moiety in
the DAMN-based imine 6a seems to be essential for obtaining
the imidazole carboxamides 7.
Figure 2
Hydroxybenzylidene imines
explored in this work and the chemical
structure of compound 10, obtained using imine 8 and 2-chlorobenzaldehyde.
Hydroxybenzylidene imines
explored in this work and the chemical
structure of compound 10, obtained using imine 8 and 2-chlorobenzaldehyde.The scope of the reaction was then explored using various aromatic
aldehydes 11, specifically benzaldehydes substituted
with electron-withdrawing (Cl, Br, F, NO2, and CN) and
electron-donating (OMe and Me) groups, furan-2-carbaldehyde, 6-methoxy-2-naphthaldehyde,
and piperonal (Scheme and Table ). The
optimal experimental conditions were obtained using 1.0–1.5
equiv of aldehydes 11, 2 equiv of triethylamine at room
temperature, and ethanol as the solvent. Reaction times range from
7 to 68 h. In general, among the benzaldehydes employed, meta-substituted derivatives provided the best results with yields ranging
from 73 to 89%. Ortho- and para-substituted
benzaldehydes led to moderate yields (31–70%). The reaction
was also shown to be useful for other aromatic aldehydes, and in this
case, the yields varied according to the aldehyde that was used.
Scheme 3
Synthesis of Imidazoles 7 from DAMN-Based Benzylidene
Imines 6
Table 1
Reaction Conditions and Yields for
the Synthesis of 7a
comp.
Ar1
Ar2
time (h)
yield (%)
7a
2-OH-phenyl
phenyl
7
93
7b
2-OH-phenyl
2-Cl-phenyl
16
42
7c
2-OH-phenyl
3-Cl-phenyl
26
74
7d
2-OH-phenyl
4-Cl-phenyl
25
31
7e
2-OH-phenyl
2,5-(Cl)phenyl
17
37
7f
2-OH-phenyl
4-CN-phenyl
24
69
7g
2-OH-phenyl
3-Br-phenyl
19
73
7h
2-OH-phenyl
4-Br-phenyl
15
54
7i
2-OH-phenyl
3-F-phenyl
19
65
7j
2-OH-phenyl
4-tolyl
24
70
7k
2-OH-phenyl
2-OMe-phenyl
18
40
7l
2-OH-phenyl
3-NO2-phenyl
15
89
7m
2-OH-phenyl
benzo[d](1,3]dioxol-5-yl
17
28
7n
2-OH-phenyl
6-OMe-naphthalen-2-yl
16
54
7o
2-OH-phenyl
furan-2-yl
24
53
7p
2-OH-3-OMe-phenyl
3-NO2-phenyl
66
49
7q
2-OH-3-OMe-phenyl
2-Cl-phenyl
68
51
7r
2-OH-naphthalenyl
2-Br-phenyl
14
73
1.2 equiv
of 11 were
used except for compounds 7a (1.0 equiv), 7i (1.5 equiv), and 7j (1.5 equiv).
1.2 equiv
of 11 were
used except for compounds 7a (1.0 equiv), 7i (1.5 equiv), and 7j (1.5 equiv).The reported method also proceeds
satisfactorily using other substituted
2-hydroxybenzylidene imines such as 6b or 2-hydroxyarylidene
imines such as 6c (Figure and Table ).Altogether, the synthetic method reported here gives
access, in
a regiocontrolled manner, to a variety of highly substituted imidazole-4-carboxamides
containing a 2-hydroxyarylidene group and an aryl group at positions
1 and 2, respectively.
Computational Studies
Having demonstrated
that the
presence of the 2-hydroxyphenyl moiety in the starting material is
decisive in the regioselective synthesis of imidazoles 7, the reaction mechanism was studied in silico to
further explore its role in product formation. The calculations were
performed assuming that the formation of the five-membered ring (that
further leads to the final product 7, path a) is started
by the nucleophilic attack of the nitrogen atom (N1) of the 2-hydroxybenzylidene
moiety to the carbon atom (Cb) of the in situ formed
diimine, and the formation of the six-membered ring (that further
leads to 1,2-dihydropyrazines, path b) occurs through an aza-Claisen-like
rearrangement by the reaction between the two carbon atoms of the
imine moieties (Ca and Cb) (Scheme ). For our initial studies, the aryl group was not
modulated and was substituted by a hydrogen atom. Once the reaction
path was determined, the effect of the phenyl group was then explored,
which will be discussed later.
Scheme 4
Schematic Representation of (A) the
Intramolecular Cyclization Step
for the Formation of Five- (Path a) and Six-Membered (Path a) Structures
B and C, Respectively, and (B) the Aromatization Step Assisted by
the 2-Phenol Group
The two plausible protonation
states of the phenol group were considered for step 1 and are indicated
with “.OH” (neutral form) and with “.O” (phenoxide form).
Schematic Representation of (A) the
Intramolecular Cyclization Step
for the Formation of Five- (Path a) and Six-Membered (Path a) Structures
B and C, Respectively, and (B) the Aromatization Step Assisted by
the 2-Phenol Group
The two plausible protonation
states of the phenol group were considered for step 1 and are indicated
with “.OH” (neutral form) and with “.O” (phenoxide form).Two main
steps were considered: (1) intramolecular cyclization
of the 2-hydroxybenzylidene structure A into the five-membered
ring structure B (path a) and six-membered ring structure C (path b) (Scheme A) and (2) aromatization of structure B into
the final imidazole form D, intramolecularly assisted
by the 2-phenol moiety in the process (Scheme B).Considering that the reaction conditions
for the formation of either
imidazoles or 1,2-dihydropyrazines involve the use of triethylamine,
the two plausible protonation states of the phenol group were considered
and are indicated with “.OH” (neutral form)
and with “.O” (phenoxide form). Intrinsic
reaction coordinate (IRC) calculations were performed to identify
the transition states of each process.[25] All calculations were performed in the implicit solvent [CPMC (conductor-like
polarizable continuum model,[26] ethanol]
using program Gaussian 09[27] at the B3LYP
level of theory with Grimme’s correction[28] and the basis set TZVP.[29](1) Step 1—intramolecular cyclization: this was explored
with the three possible isomers in the hydroxybenzylidene moiety (ortho, meta, and para)
(see the Supporting Information). Both
paths led to the respective transition states TS1 and TS2 in both protonation states studied (Figure A). When the neutral form was considered,
an activation free energy barrier of 9.2 kcal mol–1 was obtained for the formation of the five-membered compound B, while a value of 20.7 kcal mol–1 was
obtained for the formation of the six-membered derivative C (Figure B). The
energy differences between both paths were when the phenoxide form
was employed (∼8 kcal mol–1). For the three
isomers, the formation of five-membered derivatives is predicted to
be more favorable than the formation of six-membered ones (Figure S1). While no remarkable differences were
obtained among the three isomers when the neutral form was used, a
different scenario was obtained for the anionic form. Thus, while
for the para isomer, the differences in activation energy values between
both paths were similar to that of the ortho isomer (∼6–7
kcal mol–1), for the meta isomer,
the differences between the paths were higher (∼13 kcal mol–1). These findings highlighted a strong contribution
of the electronic effects on the intramolecular cyclization step,
and based on the low energy barriers differences obtained, both paths
would be possible and reversible. Therefore, the explanation for the
different regioselectivities obtained may lie in the subsequent evolution
of the intermediates generated in this step.
Figure 3
(A) Geometries of the
transition states TS1.OH, TS1.O, TS2.OH, and TS2.O obtained
by computational studies. (B) Free energy profiles for the cyclization
steps (paths a and b) to afford five- and six-membered derivatives
substituted with a 2-phenol group. The two plausible protonation states
of the phenol group were considered and are indicated with “.OH” (neutral form) and with “.O” (phenoxide form). For better comparison of the energy barriers
of the cyclization processes, the energy of the most suitable conformer
of A.OH and A.O that leads to the corresponding
transition state is given.
(A) Geometries of the
transition states TS1.OH, TS1.O, TS2.OH, and TS2.O obtained
by computational studies. (B) Free energy profiles for the cyclization
steps (paths a and b) to afford five- and six-membered derivatives
substituted with a 2-phenol group. The two plausible protonation states
of the phenol group were considered and are indicated with “.OH” (neutral form) and with “.O” (phenoxide form). For better comparison of the energy barriers
of the cyclization processes, the energy of the most suitable conformer
of A.OH and A.O that leads to the corresponding
transition state is given.On the other hand, the analysis of the transition states TS1.OH/TS1.O and TS2.OH/TS2.O revealed that for
all the transition states, the NH2 group of the amide moiety
would be pointing toward the N4 nitrogen atom to establish an intramolecular
hydrogen bonding interaction (Figures A and S2). This contact
seems to help in achieving the best arrangement for cyclization. In
addition, some differences among the TS1 structures, for the three
isomers explored, were identified. Thus, the relative arrangement
of the phenolic ring in relation to the five-membered structures,
formed in the cyclization process, was shown to be distinct for the
ortho isomer when compared with that for the meta and para isomers,
which followed a similar architecture. For the ortho isomer, the phenolic ring showed to be almost coplanar to the new
five-membered ring generated.(2) Step 2—aromatization:
an inspection of intermediates B.OH and B.O, which were obtained in step 1,
showed remarkable differences in their spatial disposition depending
on the position of the substituent in the aromatic ring (Figure A). Thus, for the meta and para isomers, an overall planar
arrangement was mainly obtained with the newly formed five-membered
ring and the phenol moiety in the same plane. However, for the ortho isomers, in which the substituent is located on the
same site of the methylene group, the phenol/phenoxide ring is mainly
arranged about 45 and 30°, respectively, from the coplanar disposition.
The optimal conformation of B.O (ortho) to reach the transition state is shown. As a result, the oxygen
atom in the phenol/phenoxide moiety would be close to one of the hydrogen
atoms in the methylene group of the five-membered ring, with distances
of 2.1 and 1.9 Å for the phenol and phenolate derivatives, respectively.
Although there is free rotation around the C(sp2)–C(sp2, aryl)
bond, this suitable arrangement for the ortho intermediates B.OH and B.O suggests that the phenol/phenoxide
moiety could be involved in the final aromatization of the five-membered
ring generated to afford the experimentally observed imidazole derivatives.
These findings led us to explore this intramolecular abstraction reaction,
which would be geometrically difficult for the meta and para isomers
(Figure S3). The fact that the velocity
of the reaction was not increased with the concentration of the base
also supported an intramolecular reaction mechanism.
Figure 4
(A) Superposition of
the geometries of the B.O intermediates
in the three isomeric forms obtained by computational studies. Detailed
view of the B.O intermediate of the ortho isomer highlighting
the dihedral angle between the two rings. The atoms involved in the
dihedral angle are highlighted. Note how for the meta and para isomers,
both rings are coplanar. The non-coplanar arrangement of the ortho
isomer as well as the proximity of the phenoxide group would facilitate
the intramolecular removal of the hydrogen atom in the five-membered
ring. (B) Free energy profiles for the formation of imidazole derivatives
substituted with a 2-phenol group. Only the phenoxide forms of the
starting material, intermediates, and transition states are considered.
(C) Geometries of the transition states TS3 and TS4 and intermediate INT obtained by computational
studies. Note how the intermediate INT undergoes a ∼45°
turn to transfer the hydrogen atom.
(A) Superposition of
the geometries of the B.O intermediates
in the three isomeric forms obtained by computational studies. Detailed
view of the B.O intermediate of the ortho isomer highlighting
the dihedral angle between the two rings. The atoms involved in the
dihedral angle are highlighted. Note how for the meta and para isomers,
both rings are coplanar. The non-coplanar arrangement of the ortho
isomer as well as the proximity of the phenoxide group would facilitate
the intramolecular removal of the hydrogen atom in the five-membered
ring. (B) Free energy profiles for the formation of imidazole derivatives
substituted with a 2-phenol group. Only the phenoxide forms of the
starting material, intermediates, and transition states are considered.
(C) Geometries of the transition states TS3 and TS4 and intermediate INT obtained by computational
studies. Note how the intermediate INT undergoes a ∼45°
turn to transfer the hydrogen atom.As intermediate B.OH could also be converted into B.O by the reaction with the Et3N employed in the
reaction, our next calculations were focused on intermediate B.O. The full energy landscape is presented in Figure B. The calculated activation
free energy barrier for the intramolecular hydrogen atom abstraction
that led to transition state TS3 and intermediate INT was 12.9 kcal mol–1 (Figure C). As a result of this reaction,
the five-membered ring became flat, and a 45° turn in the resulting
2-phenol moiety was observed to locate the OH group close to the carbon
atom that joins both rings, achieving a suitable arrangement for proton
transfer. In the transition state of the latter process, TS4, the transferred proton was located at 1.52 Å to the benzylic
carbon atom and at 1.17 Å to the phenolic oxygen atom. This step
is expected to be very fast as the calculated activation free energy
barrier was 1.4 kcal mol–1. In addition, this step
should be irreversible as the reverse process would have a high energy
barrier of 26.8 kcal mol–1. In contrast with that
for TS1.O and TS3, for TS4,
the carbonyl group of the amide group was pointing toward the heterocyclic
nitrogen atom.From the full energy landscape obtained for diimine A.O, we can conclude that the cyclization reactions for the
formation
of both the five- and six-membered ring derivatives show the following
characteristics: (i) they are energetically favorable and (ii) as
both reactions have relatively close energy barriers (of 10.0 and
18.5 kcal mol–1, respectively), they are competitive
processes and reversible. In addition, for the 3- and 4-hydroxybenzylidene
derivatives 8 and 9, since the generated
five-membered intermediates B.OH/B.O cannot evolve, the
process reverses, favoring the formation of the six-membered systems.
However, for the ortho derivative 6, as this type of
substitution promotes the intramolecular abstraction of one of the
methylene hydrogen atoms of the five-membered system B.O, the evolution of the reaction toward the formation of imidazole
derivatives is triggered. The low energy barriers of the abstraction
and protonation steps, and an ΔG‡ calculated for the global process leading to the imidazole derivative D.O to be −27.4 kcal mol–1, would
make this reaction highly favorable in comparison to the formation
of the six-membered derivative.Finally, following a similar
approach, the effect of the presence
of a phenyl group in the in situ generated diamine
derivative (such as 3) was evaluated by studying the
conversion of diimine A.O_Ph into imidazole D.O_Ph (Figure ). The results
showed that the overall process would also be energetically favorable
with an ΔG‡ calculated to
be −20.5 kcal mol–1 and energy barriers of
13.6 and 12.8 kcal mol–1 for the first and second
steps, respectively. According to the calculations performed here,
while no significant differences were identified for the second step
(intramolecular H abstraction), the energy barrier involving the intramolecular
cyclization leading to the formation of the five-membered ring would
be 3.6 kcal mol–1 higher, probably because of the
stereo hindrance introduced by the substitution of a hydrogen atom
by a phenyl group. No relevant differences were identified in the
geometries and arrangements of the transition states and intermediates.
In addition, the energy activation barriers differences for the formation
of five- and six-membered rings were revealed to be smaller, which
reinforces the idea of competitive processes (Figure S4).
Figure 5
(A) Schematic representation of the intramolecular cyclization
and aromatization steps assisted by the 2-phenol group to afford 1-(2-hydroxyphenyl)-2-phenyl
imidazole derivatives; (B) geometries of the transition states TS1.O_Ph, TS3_Ph, and TS4_Ph and
the intermediate INT_Ph obtained by computational studies;
and (C) free energy profiles for the formation of 1-(2-hydroxyphenyl)-2-phenyl
imidazole derivatives. Only the phenoxide forms of the starting material,
intermediates, and transition states are considered. Note the different
arrangement of the CONH2 group in TS1.O_Ph, in which the NH2 group is pointing toward the imine
moiety (yellow shadow), from that in transition states TS3_Ph and TS4_Ph and intermediate INT_Ph.
(A) Schematic representation of the intramolecular cyclization
and aromatization steps assisted by the 2-phenol group to afford 1-(2-hydroxyphenyl)-2-phenyl
imidazole derivatives; (B) geometries of the transition states TS1.O_Ph, TS3_Ph, and TS4_Ph and
the intermediate INT_Ph obtained by computational studies;
and (C) free energy profiles for the formation of 1-(2-hydroxyphenyl)-2-phenyl
imidazole derivatives. Only the phenoxide forms of the starting material,
intermediates, and transition states are considered. Note the different
arrangement of the CONH2 group in TS1.O_Ph, in which the NH2 group is pointing toward the imine
moiety (yellow shadow), from that in transition states TS3_Ph and TS4_Ph and intermediate INT_Ph.
Conclusions
A series of 2-aryl-5-cyano-1-(2-hydroxybenzyl)-1H-imidazole-4-carboxamides 7 were regioselectively
prepared
from DAMN-based benzylidene imines 6 and diverse aromatic
aldehydes 11. Exploring the reaction mechanism that led
to their formation showed that the presence of the 2-hydroxyaryl moiety
controls the mechanism of the reaction, guiding it mainly through
a self-catalyzed hydrogen atom shift. These conclusions were based
on the reactions performed with hydroxybenzylidene imines 6, 8, and 9, where it was observed that
(i) only the imines possessing a 2-hydroxyl moiety in the aromatic
ring, compounds 6, evolved to imidazoles 7, while imines 8 (3-OH) and 9 (4-OH) led
to the formation of pyrazines 4, much likely via a diaza-Cope rearrangement. The mechanism of the regioselective
formation of imidazoles 7 was supported by computational
studies, where the most energetically favorable process involved a
phenol-assisted hydrogen atom shift, which is not geometrically suitable
for the corresponding meta and para derivatives.
Experimental Section
General
All chemicals
and solvents for synthesis were
of the analytical grade, purchased from commercial sources and used
without further purification, unless otherwise specified. The reactions
were monitored by thin layer chromatography, using glass plates precoated
with silica gel 60 and a fluorescent indicator (Macherey Nagel, DC-Fertigplatten
Duresil 25 UV254) or aluminum plates precoated with silica gel 60
and a fluorescent indicator (Macherey Nagel, DC-Fertigfolien ALUGRAM
Xtra SIL G/UV254). The spots were observed under ultraviolet (UV)
light irradiation in a UV chamber (CN-6 Vilber Lourmat) with a 254
nm lamp and in an iodine chamber. For flash chromatography, silica
gel MN Kieselgel 60 (230 ASTM, particle size <0.063 mm) was used.
Distilled water was used in the reactions performed in aqueous solution.
The NMR spectra were recorded at room temperature on a Varian Unity
Plus spectrometer (1H: 300 MHz and 13C: 75 MHz)
or a Brucker AVANCE III spectrometer (1H: 400 MHz, 500
MHz and 13C: 100 MHz, 125 MHz), using DMSO-d6 as the solvent. The chemical shifts (δ) are expressed
in parts per million (ppm), and the coupling constants (J) are given in hertz (Hz). NMR assignments were obtained by a combination
of 1D and bidimensional (HMQC and HMBC) experiments. IR spectra were
recorded on a Fourier transform infrared (FTIR) Bomem MB apparatus
with Nujol mulls or NaCl plates or on a PerkinElmer Two FTIR spectrometer
with attenuated total reference (ATR). Melting points were determined
using a Stuart SMP3 melting point apparatus or a Büchi M-560
apparatus. Electrospray ionization (ESI) mass spectra were recorded
on a Bruker BIOTOF II mass spectrometer. Elemental analysis was determined
on a Thermo Flash EA 1112 (NC Soil Analyzer) apparatus [carrier gas
flux (He): 130 mL/min; reference gas flux (He): 100 mL/min; oxygen
flux: 250 mL/min; oven temperature: 950 °C; column: multiple
analysis 6 × 5 mm, 2.0 m de Cromlab; column temperature: 75 °C.;
and standard: sulfanilamide (Thermo Scientific)]. Hydroxybenzylidene
imines 6a, 8, and 9 were prepared
as previously described.[9]
DAMN (0.54 g, 5.0 mmol) and 3-methoxy salicylaldehyde
(0.76 g, 5.0 mmol) were combined in the minimum amount of DMSO and
treated with a catalytic amount of H2SO4 (c),
and the resulting mixture was vigorously stirred for 3 min. A solid
precipitate was generated almost instantaneously. The reaction mixture
was diluted with water (15 mL), and the dense suspension was cooled
in an ice bath for 2 h. The solid was filtered, and the filtrate was
successively washed with water, ethanol, and diethyl ether and dried.
Dark-yellow solid (1.19 g, 98%). mp 221 °C (dec). 1H NMR (400 MHz, DMSO-d6): δ/ppm
9.86 (s, 1H, OH), 8.59 (s, 1H, NCHAr), 7.84 (s, 2H, NH2), 7.61 (dd, J1 = 7.6, J2 = 1.2 Hz, 1H, H6′), 7.06 (dd, J1 = 7.6, J2 = 1.2 Hz, 1H,
H4′), 6.83 (t, J = 8.0 Hz, 1H, H5′),
and 3.83 (s, 3H, OMe). IR (nujol) ν̅/cm–1: 3408 (s), 3324 (s), 3208 (w), 2245 (m), 2205 (m), 1645 (s), 1602
(s), 1589 (m), and 1526 (w). MS (ESI) m/z: 243 (MH+). HRMS calcd for C12H10N4O2 (MH+), 243.0877; found, 243.0872.
This compound was prepared following a
similar procedure to that used for 6b. Yellow solid (92%).
mp 223–225 °C. 1H NMR (500 MHz, DMSO-d6): δ/ppm 11.96 (s, 1H, OH), 9.24 (s,
1H, NCHAr), 8.60 (d, J = 8.6 Hz, 1H, H8′),
7.98 (d, J = 8.9 Hz, 1H, H4′), 7.87 (d, J = 8.5 Hz, 1H, H5′), 7.86 (s, 2H, NH2), 7.61 (t, J = 8.5 Hz, 1H, H7′), 7.41 (t, J = 8.5 Hz, 1H, H6′), and 7.22 (d, J = 9.0 Hz, 1H, H3′). 13C NMR (125 MHz, DMSO-d6): δ/ppm 159.8 (C2′), 154.9 (NCHAr),
135.2 (C4′), 131.7 (C8a′), 128.9 (C5′), 128.5
(C7′), 127.8 (C4a′), 125.1 (C1), 123.9 (C6′),
121.9 (C8′), 118.5 (C3′), 114.6 (CN), 114.1 (CN), 110.2
(C1′), and 103.9 (C2). IR (ATR) ν̅/cm–1: 3466 (m), 3340 (m), 3194 (m), 2242 (m), 2197 (m), 1618 (s), 1603
(m), 1557 (m), and 1356 (m). MS (ESI) m/z: 263 (MH+). HRMS calcd for C15H11N4O (MH+), 263.0927; found, 263.0925.
General Procedure for the Synthesis of Imidazoles 7
A suspension of the 2-hydroxybenzylidene imine 6a(9) in ethanol (2 mL, 0.45 M) was treated
with aromatic aldehydes 11 (1.0–1.5 equiv) and Et3N (2 equiv), and the reaction mixture was stirred at room temperature
for 7–24 h. The resulting suspension was then cooled in an
ice bath for 2 h, and the obtained precipitate was filtered, successively
washed with cold methanol or ethanol and diethyl ether, and dried.
A solution of imine 8 (0.2
g, 0.94 mmol) in ethanol (2 mL) at room temperature was treated with
2-chlorobenzaldehyde (127 μL, 1.13 mmol) and Et3N
(1.88 mmol, 262 μL), and the reaction mixture was stirred for
96 h. The resulting yellow solid was filtered and successively washed
with cold ethanol, methanol, and diethyl ether. A sample (9 mg) of
the solid was solubilized in DMSO-d6 (250
μL), and the mixture was heated at 80 °C for 30 min. After
this time, the mixture was studied by 1H and 13C NMR spectroscopy, with dihydropyrazine 10 being identified. 1H NMR (400 MHz, DMSO-d6): δ/ppm
9.62 (s, 1H, OH), 8.66 (d, J = 2.8 Hz, 1H, NH), 8.06
(s, 1H, CONHH), 7.94 (s, 1H, CONHH), 7.55 (dd, J1 = 7.6 Hz, J2 = 1.2 Hz, 1H, H3″), 7.37 (dt, J1 = 7.6 Hz, J2 = 1.6 Hz, 1H,
H4″), 7.31 (dt, J1 = 7.6, J2 = 1.6 Hz, 1H, H5″), 7.22 (t, J = 2.0 Hz, 1H, H2′), 7.18 (t, J = 8.0 Hz, 1H, H5′), 7.07 (dd, J1 = 7.6, J2 = 1.6 Hz, 2H, H6′ +
H6″), 6.78 (dd, J1 = 8.0, J2 = 2.5 Hz, 1H, H4′), and 6.16 (d, J = 2.8 Hz, 1H, CH). 13C NMR (100 MHz, DMSO-d6): δ/ppm 162.3 (CO), 157.6 (C3′),
145.2 (C5), 140.6 (C2), 136.1 (C1′), 135.4 (C1″), 131.4
(C2″), 130.6 (C4″), 130.3 (C3″), 129.8 (C5′),
129.4 (C6″), 119.1 (CN), 117.8 (C4′), 117.1 (C6′),
112.9 (C2′), 96.2 (C3), and 48.6 (C6). IR (nujol) ν̅/cm–1: 3475 (w), 3270 (m), 2205 (s, CN), 1672 (s, CO),
1558 (m), and 1533 (m).All calculations were initially
performed in vacuum using the program Gaussian 09[27] at the B3LYP level of theory and the basis set TZVP.[29] Transition states and intermediates were further
refined considering Grimme’s correction[28] and the solvent effect (CPCM,[26] ethanol). The imaginary frequencies calculated for all the transition
states identified in the reaction path are summarized in Table S1.Step 1—intramolecular
cyclization: to find a starting point for determining the transition
states of the two possible cyclization reactions [formation of the
five-membered compounds (path a) and six-membered heterocycles (path
b)], a relaxed potential energy surface (PES) was calculated for the
three possible isomers (ortho, meta, and para) of each path, and the two possible protonation
forms of the phenol group were explored (neutral and phenoxide). No
substituent (aryl) was considered in the whole calculation. The two
possible relative arrangements of the substituents −CN and
−CONH2 were considered, that is, with either the
NH2 group or the carbonyl group in the CONH2 moiety pointing toward the CN substituent. To identify the transition
state, an elongation of the bond formed in the reaction from 1.6 to
2.5 Å was applied by always taking the last step of the previous
distance as the starting point of the next one. For the formation
of five-membered structures, the distance between the nitrogen atom
connected to the benzyl group (N1) and the methylenic carbon atom
(Cb) was considered. For the formation of six-membered structures,
the distance between the benzylenic carbon atom (Ca) and Cb was used.
The highest energy points of these PES calculations were then used
as the starting point for the identification of the transition states.
Finally, to obtain both the starting and end points of the reaction,
an IRC approach was performed in both the forward and reverse directions.
In all cases, the final step of the IRC calculation was further minimized
to obtain a local minimum for each structure.Step 2—aromatization:
a similar protocol to that described
for step 1 (phenoxide form) was employed here but considering a two-stage
process: (i) abstraction of the hydrogen atom in the previously obtained
nitrogen heterocycle by the phenoxide group and (ii) transfer of the
hydrogen atom from the resulting phenol group to the benzylidene carbon
atom (Ca). For both stages, the elongation was performed from 1.3
to 3.0 Å, starting from the previously obtained intermediate B.O in the case of stage i and from the resulting intermediate
of the latter (intermediate INT) in the case of stage
ii. All the identified transition states were characterized by having
an imaginary frequency that corresponds to the formation of the O–H
bond (TS3 and TS3_Ph) or of the Cα–H
bond (TS4 and TS4_Ph) (Table S1). Cartesian coordinates (PDB1–24), total electronic
energies, and thermal corrections (Table S2) are included in the Supporting Information.
Authors: Pankaj B Miniyar; Prashant R Murumkar; Pallavi S Patil; Mahesh A Barmade; Kailash G Bothara Journal: Mini Rev Med Chem Date: 2013-10 Impact factor: 3.862
Authors: Andrew Potter; Victoria Oldfield; Claire Nunns; Christophe Fromont; Stuart Ray; Christopher J Northfield; Christopher J Bryant; Simon F Scrace; David Robinson; Natalia Matossova; Lisa Baker; Pawel Dokurno; Allan E Surgenor; Ben Davis; Christine M Richardson; James B Murray; Jonathan D Moore Journal: Bioorg Med Chem Lett Date: 2010-09-17 Impact factor: 2.823
Authors: Ana I Ribeiro; Carla Gabriel; Fátima Cerqueira; Marta Maia; Eugénia Pinto; João Carlos Sousa; Rui Medeiros; M Fernanda Proença; Alice M Dias Journal: Bioorg Med Chem Lett Date: 2014-08-17 Impact factor: 2.823
Authors: Ana Bettencourt; Marián Castro; João Silva; Francisco Fernandes; Olga Coutinho; M João Sousa; M Fernanda Proença; Filipe Areias Journal: Molecules Date: 2018-10-03 Impact factor: 4.411