Structural revision in the reactions of 3-cyanochromones with primary aromatic amines. Improved synthesis of 2-amino-3-(aryliminomethyl)chromones

Article abstract of DOI:10.1016/j.tetlet.2009.09.028

3-Cyanochromones react with primary aromatic amines to give 2-amino-3-(aryliminomethyl)chromones as the sole products or as their mixtures with Z- and E-3-anilino-2-salicyloylacrylonitriles, depending on the reaction conditions. With aliphatic amines, 2-a

Full text of DOI:10.1016/j.tetlet.2009.09.028

Tetrahedron Letters 50 (2009) 6515–6518
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Structural revision in the reactions of 3-cyanochromones with primary
aromatic amines. Improved synthesis of 2-amino-3-
(aryliminomethyl)chromones
a
b
Vyacheslav Ya. Sosnovskikh a, , Vladimir S. Moshkin , Mikhail I. Kodess
*
a Department of Chemistry, Ural State University, pr. Lenina 51, 620083 Ekaterinburg, Russia
b Institute of Organic Synthesis, Ural Branch of the Russian Academy of Sciences, 620041 Ekaterinburg, Russia
a r t i c l e i n f o
a b s t r a c t
Article history:
3-Cyanochromones react with primary aromatic amines to give 2-amino-3-(aryliminomethyl)chromones
as the sole products or as their mixtures with Z- and E-3-anilino-2-salicyloylacrylonitriles, depending on
the reaction conditions. With aliphatic amines, 2-amino-3-(alkyliminomethyl)chromones are obtained in
good yields. The reaction of 3-cyanochromone with o-phenylenediamine is reinvestigated and proof for
the product structure and a possible reaction pathway are presented.
Received 15 June 2009
Revised 24 August 2009
Accepted 4 September 2009
Available online 9 September 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
3-Cyanochromones
Primary amines
2-Amino-3-(aryl/
alkylimino)methylchromones
3-Anilino-2-salicyloylacrylonitriles
Introduction of the electron-withdrawing CN group at the 3-po-
sition of a chromone system significantly changes the reactivity of
the pyrone ring with respect to nucleophiles, and provides broad
synthetic potential for 3-cyanochromone 1.1 Among the diverse
transformations of 1, one of the most important is its conversion,
on heating with morpholine in an aqueous DMF,2 with n-propyl-
amine in an aqueous ethanol3 or in the presence of NaOH in water,4
into 2-amino-3-formylchromone. On heating at reflux with ethy-
lenediamine in ethanol, chromone 1 gives 2-amino-3-formylchro-
mone together with 2-(salicyloylmethylene)tetrahydroimidazole
as a minor product.3 However, to the best of our knowledge, very
little is known about the reaction of 1 with primary aromatic
amines. There is only one report on the reaction of 3-cyano- and
3-cyano-6-methylchromones 1 with p-toluidine in benzene at re-
flux.5 It was claimed that benzene induces the initial 1,4-addition
of p-toluidine and the product of this reaction, 2-amino-3-(p-toly-
liminomethyl)chromone 4, depending on the conditions, can exist
solely or partially as its two other tautomeric forms 2 and 3
(Scheme 1). For example, according to the 1H NMR spectrum in
CDCl3, the Schiff base 4 (R = Me) has been described as its enamine
form 3 (3JH,NH = 13.0 Hz), while in the solid state it exists as the
especially since we noted the unusually downfield shifted signal
due to the H-5 aromatic proton (d 8.20 ppm), which is typically
manifested in the 6-methylchroman-4-one series at ca.
d
7.7 ppm. It seemed more logical to us that the reaction of 1 with
p-toluidine would produce a mixture of 4 and open-chain form 2
where the H-6 aromatic proton (H-5 in 3) is shifted downfield
due to the deshielding effect of the cyano group in a planar confor-
H
H
Ar
Ar
O
O
O
R
O
N
R
R
CN
ArNH2
ref. 5
CN
2
1
4
H
O
O
O
O
N
Ar
R
N
acrylonitrile 2 (m
CN = 2210 cmꢀ1 in KBr).5 However, structure 3
NH2
NH
was not established with certainty and remains open to question,
3
R = H (a), Me (b); Ar = p-MeC6H4
* Corresponding author. Fax: +7 343 261 59 78.
E-mail address: vyacheslav.sosnovskikh@usu.ru (V.Ya. Sosnovskikh).
Scheme 1.
0040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2009.09.028

6516
V. Ya. Sosnovskikh et al. / Tetrahedron Letters 50 (2009) 6515–6518
mation. The latter should be stabilized in CDCl3 solution by two
intramolecular hydrogen bonds involving the NH and OH protons
and the carbonyl group.
H
Ar
O
O
OH
O
N
R
R
R
CN
ArNH2
C6H6
H
We therefore repeated the reactions of chromones 1 with ani-
line and p-methyl-, p-methoxy-, and p-bromoanilines following
the literature method (reflux, benzene, 4 h).5 We also used benzyl-
and isopropylamines in order to determine whether aliphatic sub-
stituents influence the structure of the reaction products. In the
case of 3-cyano-6-methylchromone 1 (R = Me) and aniline, we ob-
tained a 77:23 mixture (CDCl3) of isomers Z-2a and E-2a,6 having s-
E conformation at the C–N bond, and no traces of 4a. p-Methyl- and
p-methoxyanilines gave similar product mixtures along with a
small amount of 4b, c (6–17%). It is a well-established phenome-
non with push–pull alkenes, that the Z/E equilibrium is dependent
on the solvent polarity.7 Indeed, in DMSO-d6, the equilibrium was
shifted toward the E-2 isomer (Table 1). This result can be ex-
plained by considering the fact that the formation of Z-2 is pre-
ferred in solvents of low polarity, such as chloroform, due to the
presence of the intramolecular hydrogen bond between the NH
and carbonyl group, whereas in high polarity solvents, such as
DMSO, isomer E-2 may be more favorable due to the strength of
the intermolecular hydrogen bond. The structural assignment of
compounds 2 as the Z- and E-isomer of 3-anilino-2-salicyloylacry-
lonitriles was based on 1H, 13C, and 15N NMR spectroscopy. In addi-
tion, all the signals in the 1H and 13C NMR spectra of 2a were
assigned on the basis of 2D 1H–13C HSQC and HMBC experiments.
Consequently, structure 3 should be revised to 2 (Scheme 2).
The different isomeric forms of acrylonitriles 2 have character-
istic 1H NMR spectra. Thus, the vicinal coupling constant,
3JH,NH = 13.5 Hz, found for Z-2 is in agreement with the fixed s-E
conformation around the C–N bond. The signal corresponding to
the NH proton appears at ca. 12.5 ppm and is only slightly depen-
dent on the solvent polarity; this provides evidence that the proton
is involved in an intramolecular hydrogen bond and is in agree-
ment with the Z configuration. The doublet due to the NH proton
of E-2 exists at much higher field than that of Z-2 (ca. d 8.1 ppm
in CDCl3) and is shifted downfield with increasing solvent polarity
(ca. d 11.0 ppm in DMSO-d6). Therefore, it can be concluded that
this proton is involved in an intermolecular hydrogen bond. In
CN
1
Z-(s-E)-2a-d
R
ArNH2
C6H6/Et3N
O
O
OH
H
Ar
Ar
N
N
H
Hb
CN
O
4a-h
N
Ha
R
E-(s-E)-2a-d
X
H
Ar
O
N
OH
R
O
H
H
N
CN Ar
O
NH
E-(s-Z )-2a-d
5
Scheme 2.
The reaction of 3-cyanochromones with amines turned out to
be very sensitive to the nature of the substituent at the 6-position.
When 1 (R = H) was allowed to react with p-substituted anilines,
and benzyl- and isopropylamines in benzene at reflux for 4 h, the
initially formed intermediates 2 could not be isolated and under-
went rapid cyclization via addition of the phenolic hydroxy to
the CN group to give chromone derivatives 4d–h as the sole prod-
ucts in 54–80% yields (except for 4d, which was isolated as a 1:3
mixture of 2d and 4d). This remarkable change from product 2 to
product 4 on changing the C-6 substituent of 1 from methyl to
hydrogen possibly results from the electron-donating effect of
the 5-Me group of 2, which decreases the acidity of the OH group.
In accordance with this, chromones 4a–d were obtained in pure
form (68–98% yields) under the same reaction conditions in the
presence of two drops of triethylamine, which accelerates the
cyclization step via deprotonation of the phenolic hydroxy group
of intermediate 2.
3
CDCl3, the large JH,NH = 15.2 Hz value allowed us to assign the s-
E conformation to the C–N bond of E-2, however, the correspond-
ing coupling was not observed in DMSO-d6 solution where the
NH and @CH protons appeared as slightly broadened singlets. This
may be attributed to the s-Z conformation in which the coupling
constant due to the NH proton (usually 3JH,NH = 7–8 Hz) can be de-
creased or eliminated, depending on the amount of water present
in the solvent and the solute concentration.7 These spectral data
are in line with the structures proposed for isomers 2a–d.
The 1H NMR spectra of compounds 4 in DMSO-d6 consisted of a
characteristic singlet due to the @CH proton at ca. d 9.0 and two
singlets due to the resonances of the non-equivalent NH2 protons
at d 9.3–9.4 (Ha) and 10.5–10.7 (Hb). Addition of CD3CO2D resulted
in the disappearance of the two latter signals. All the signals in the
1H and 13C NMR spectra of compound 4f8 were assigned on the ba-
sis of 2D 1H–13C HSQC and HMBC experiments. The choice between
the tautomeric forms 4 and 5 was made in favor of the former on
the basis of 2D 1H–15N HMQC and HMBC experiments. Since the
15N chemical shifts relative to ammonia are d 98.9 (NH2) and
287.4 (N@C) (for 4f), the structure 5 with the two imine nitrogen
atoms is excluded. It is reasonable to assume that the non-equiva-
lence of the NH2 protons is connected with an intramolecular
hydrogen bond involving Hb and the imine nitrogen atom. This ex-
plains the formation of E-anils 4, exclusively.
In the light of these findings, we envisaged that the reaction of
3-cyanochromone with o-phenylenediamine would produce the
corresponding anil 4i. However, the reported structures 6 and 8
of the reaction product of 1 with o-phenylenediamine, differed
from our results. It was claimed that the reaction occurred via
1,2-addition of o-phenylenediamine to the cyano group to form
amidine 6, which on further reflux in AcOH gave benzodiazepine
Table 1
1H NMR analysis of the reaction products in CDCl3 and DMSO-d6
Product
R
Ar
Z-2/E-2/4a (%)
Yield (%)
2a
2a
Me
Me
Me
Me
Me
Me
H
H
H
H
H
Ph
Ph
77:23:0b
35:65:0c
84:10:6b,d
35:60:5c
76:7:17b
30:50:20c
25:0:75b,d
0:0:100b,c
0:0:100c
0:0:100b,c
0:0:100c
36
—
51
—
81
–
54
64
80
57
65
2b + 4b
2b + 4b
2c + 4c
2c + 4c
2d + 4d
4e
4f
4g
4h
4-MeC6H4
4-MeC6H4
4-MeOC6H4
4-MeOC6H4
4-MeC6H4
4-MeOC6H4
4-BrC6H4
e
PhCH2
i-Pre
a
b
c
The ratio of products was determined from the 1H NMR spectra.
In CDCl3.
In DMSO-d6.
d
e
Previously5 this product was described as an equilibrium mixture of 2, 3, and 4.
Ar = aliphatic.

V. Ya. Sosnovskikh et al. / Tetrahedron Letters 50 (2009) 6515–6518
6517
NH2
O
O
NH
O
O
[O]
AcOH
X
N
1,2-AN
N
X
H
ref. 3
NH2
N
H
O
O
6
7
8
H2N
H2N
CN
+
H
NH2
OH
O
N
OH
O
1
NH2
NH
1,4-AN
ref. 9
X
CN
N
H2O,
base
2i
O
O
N
O
O
O
CHO
NH2
H2N
X
N
H
N
NH2
AcOH
X = NH2
+
X
O
9
X = NH2 (4i), OH (4j)
AcOH
X = OH
X = NH2
-NH3
O
O
NH
O
N
O
N
N
O
N
O
H
NH2
O
O
H2
N
H
10
B
A
N
HN
HN
O
O
O
NH
NH
OH
N
N
H
Scheme 3.
7,3 or via initial attack at the electrophilic C-2 atom (1,4-addition)
followed by intramolecular cyclization involving the amino and cy-
ano groups to form benzodiazepine 8, which was transformed into
benzimidazole 9 by treatment with AcOH9 (Scheme 3).
amino-3-formylchromone and o-aminophenol, was refluxed in
AcOH, the expected 1,3-benzoxazole derivative was not isolated.
Instead, the known self-condensation product of 3-cyanochro-
mone,
2-(4-oxo-4H-chromen-3-yl)-5H-chromeno[2,3-d]pyrimi-
In order to clarify this discrepancy, we repeated the reaction of
1 with o-phenylenediamine at reflux in ethanol (benzene can also
be used) and found that the product was, in fact, the expected 2-
amino-3-(2-aminophenyliminomethyl)chromone 4i,10 instead of
the reported compounds 6 and 8. Close scrutiny of the 1H and
13C NMR spectral data provided in Ref. 9 (unfortunately, in Ref. 3,
no spectral data of 6 are given) revealed that the benzodiazepine
structure 8 was assigned erroneously to the reaction product. Ris-
itano et al.9 assigned the signal at d 179 to the C@O group of the
salicyloyl moiety, whereas this carbon usually resonates at about
d 190 in similarly constituted molecules. Moreover, in the 1H
NMR spectrum, all the protons of the benzene ring were shifted
to a lower field in comparison with those of the salicyloyl frag-
ment. When anil 4i (this compound was also obtained from 2-ami-
no-3-formylchromone and o-phenylenediamine) was refluxed in
acetic acid for 3 h, the known benzimidazole 911 was obtained,
and not the earlier claimed,3 on the basis of the incorrect structure
6, benzodiazepine 7. Consequently, structures 6 and 8 should be
revised to 4i and structure 7 changed into 9. A possible route for
the transformation of 4i into 9 occurs via rearrangement of inter-
mediates A to B through addition–elimination sequences as out-
lined in Scheme 3. Interestingly, when anil 4j, prepared from 2-
din-5-one 10,3 was obtained in 67% yield.
In conclusion, 3-cyanochromones are very reactive systems and
are useful synthetic building blocks in medicinal chemistry. Since
the identity of some of their reaction products with primary aro-
matic amines was in doubt, we have reinvestigated these reactions
and found that the mixtures of 2-amino-3-(aryliminomethyl)chro-
mones and their open forms, Z- and E-3-anilino-2-salicyloylacrylo-
nitriles, were obtained in variable proportions depending on the
nature of the substituents and the conditions used. When the reac-
tion was carried out at reflux in benzene in the presence of trieth-
ylamine, 2-amino-3-(aryliminomethyl)chromones were formed as
the sole products.
Acknowledgment
This work was financially supported by the RFBR (Grant 06-03-
32388).
References and notes
1. Ghosh, C. K.; Karak, S. K. J. Heterocycl. Chem. 2005, 42, 1035–1042.
2. Nohara, A.; Ishiguro, T.; Ukawa, K.; Sugihara, H.; Maki, Y.; Sanno, Y. J. Med.
Chem. 1985, 28, 559–568.

6518
V. Ya. Sosnovskikh et al. / Tetrahedron Letters 50 (2009) 6515–6518
3. Ghosh, C. K.; Tewari, N. J. Org. Chem. 1980, 45, 1964–1968.
4. Petersen, U.; Heitzer, H. Liebigs Ann. Chem. 1976, 1659–1662.
Calcd for C17H14N2O2: C, 73.37; H, 5.07; N, 10.07. Found: C, 73.56; H, 5.33; N,
10.28.
5. Ghosh, C. K.; Ghosh, C.; Patra, A. Indian J. Chem., Sect B 1998, 37, 387–390.
6. 3-Anilino-2-(2-hydroxy-5-methylbenzoyl)acrylonitrile 2a. This compound was
prepared from chromone 1 (R = Me) and aniline according to the procedure
described previously for 3.5 Yield 36%, mp 226–228 °C; 1H NMR (400 MHz,
CDCl3) (Z-2a, 77%) d 2.34 (s, 3H, Me), 6.91 (d, 1H, H-3, J = 8.4 Hz), 7.22–7.32 (m,
4H, H-20, H-60, H-40, H-4), 7.46 (t, 2H, H-30, H-50, J = 7.9 Hz), 8.05 (d, 1H, @CH,
J = 13.4 Hz), 8.19 (br d, 1H, H-6, J = 1.5 Hz), 11.17 (s, 1H, OH), 12.45 (br d, 1H,
NH, J = 13.4 Hz); (E-2a, 23%) d 2.35 (s, 3H, Me), 6.92 (d, 1H, H-3, J = 8.4 Hz),
7.21–7.50 (m, 6H, H-4, Ar), 8.10 (br d, 1H, H-6, J = 1.7 Hz), 8.18 (br d, 1H, NH,
J = 15.2 Hz), 8.80 (d, 1H, @CH, J = 15.2 Hz), 11.68 (s, 1H, OH); 13C NMR
(100 MHz, CDCl3) (Z-2a) d 20.60 (Me), 82.38 (C2), 118.05 (C200, C600), 118.15
(C30), 119.35 (C10), 120.34 (CN), 126.92 (C400), 128.17 (C50), 129.08 (C60), 130.22
(C300, C500), 136.78 (C40), 137.84 (C100), 155.03 (C3), 159.26 (C20), 193.14 (C1); (E-
2a) d 84.02 (C2), 129.44 (C60), 137.98 (C100), 154.45 (C3), 160.06 (C20), 188.90
(C1) (the remaining carbon atoms were not observed due to their intensities
being too low); 1H NMR (400 MHz, DMSO-d6) (Z-2a, 35%) d 2.23 (s, 3H, Me),
6.81 (d, 1H, H-3, J = 8.2 Hz), 7.10–7.16 (m, 2H, H-4, H-6), 7.24 (tt, 1H, H-40,
J = 7.4, 1.0 Hz), 7.41–7.46 (m, 2H, H-30, H-50), 7.55–7.59 (m, 2H, H-20, H-60), 8.54
(d, 1H,@CH, J = 13.5 Hz), 9.93 (s, 1H, OH), 12.31 (br d, 1H, NH, J = 13.5 Hz); (E-
2a, 65%) d 2.23 (s, 3H, Me), 6.82 (d, 1H, H-3, J = 8.2 Hz), 7.10–7.16 (m, 2H, H-4,
H-6), 7.18 (tt, 1H, H-40, J = 7.4, 1.0 Hz), 7.28–7.32 (m, 2H, H-20, H-60), 7.36–7.41
(m, 2H, H-30, H-50), 8.13 (br s, 1H, @CH), 9.95 (br s, 1H, OH), 10.99 (br s, 1H,
NH); 13C NMR (100 MHz, DMSO-d6) (Z-2a) d 19.91 (Me), 85.48 (C2), 116.27
(C30), 118.19 (C200, C600), 119.92 (CN), 125.77 (C400), 126.59 (C10), 127.24 (C50),
128.24 (C60), 129.62 (C300, C500), 132.45 (C40), 138.46 (C100), 152.67 (C20), 153.59
(C3), 193.05 (C1); (E-2a) d 19.91 (Me), 86.75 (C2), 116.17 (C30), 116.73 (CN),
118.33 (C200, C600), 125.20 (C400), 125.63 (C10), 127.70 (C50), 129.19 (C60), 129.62
(C300, C500), 132.56 (C400), 139.78 (C100), 152.83 (C20), 154.23 (C3), 189.69 (C1);
15N NMR (40 MHz, DMSO-d6) (Z-2a) d 137.6 (NH), (E-2a) 132.1 (NH). Anal.
7. (a) Kozerski, L.; Krajewski, P.; Bednarek, E. Magn. Res. Chem. 1995, 33,
745–748; (b) Wójcik, J.; Domalewski, W.; Kamien´ ska-Trela, K.; Stefaniak,
L.; Vdovienko, S. I.; Gerus, I. I.; Gorbunova, M. G. Magn. Res. Chem. 1993,
31, 808–814.
8. 2-Amino-3-(p-bromophenyliminomethyl)chromone 4f. This compound was
prepared according to the procedure described previously.5 Yield 80%, mp
253–254 °C; IR (KBr) 3211, 3075, 1655, 1605, 1564, 1519 cmꢀ1 1H NMR
;
(400 MHz, DMSO-d6) d 7.22 (d, 2H, H-20, H-60, J = 8.7 Hz), 7.42–7.46 (m, 2H, H-
6, H-8), 7.56 (d, 2H, H-30, H-50, J = 8.7 Hz), 7.72 (ddd, 1H, H-7, J = 8.5, 7.2,
1.7 Hz), 8.05 (dd, 1H, H-5, J = 7.8, 1.7 Hz), 8.97 (s, 1H, @CH), 9.47 (br s, 1H, NHa),
10.52 (br s, 1H, NHb); 13C NMR (100 MHz, DMSO-d6) d 95.05 (C3), 116.73 (C8),
117.64 (C40), 121.83 (C4a), 123.15 (C20, C60), 125.09 (C6), 125.22 (C5), 132.03
(C30, C50), 133.50 (C7), 150.27 (C10), 152.86 (C8a), 156.46 (C@N), 164.25 (C2),
173.89 (C4); 15N NMR (40 MHz, DMSO-d6) d 98.9 (NH2), 287.4 (N@C). Anal.
Calcd for C16H11BrN2O2: C, 56.00; H, 3.23; N, 8.16. Found: C, 55.86; H, 3.35; N,
7.98.
9. Risitano, F.; Grassi, G.; Foti, F. J. Heterocycl. Chem. 2001, 38, 1083–1085.
10. 2-Amino-3-(2-aminophenyliminomethyl)chromone 4i. This compound was
prepared from 3-cyanochromone and o-phenylenediamine according to the
procedure described previously for 6.3 Yield 46%, mp 217–218 °C (lit.3 mp
210 °C, lit.7 mp 217 °C); IR (KBr) 3330, 3180, 1664, 1601, 1559, 1521, 1498,
1462 cmꢀ1 1H NMR (400 MHz, DMSO-d6) d 5.02 (s, 2H, NH2), 6.59 (ddd, 1H, H-
;
50, J = 7.8, 7.3, 1.0 Hz), 6.73 (dd, 1H, H-30, J = 7.8, 0.8 Hz), 6.87 (dd, 1H, H-60,
J = 7.8, 0.8 Hz), 6.93 (ddd, 1H, H-40, J = 7.8, 7.3, 1.0 Hz), 7.41–7.47 (m, 2H, H-6,
H-8), 7.71 (ddd, 1H, H-7, J = 8.2, 7.4, 1.6 Hz), 8.06 (dd, 1H, H-5, J = 7.7, 1.6 Hz),
8.91 (s, 1H, @CH), 9.30 (s, 1H, NHa), 10.39 (s, 1H, NHb). Anal. Calcd for
C16H13N3O2: C, 68.81; H, 4.69; N, 15.05. Found: C, 69.04; H, 4.67; N, 15.09.
11. (a) Sigg, I.; Haas, G.; Winkler, T. Helv. Chim. Acta 1982, 65, 275–279; (b) Rihs, G.;
Sigg, I.; Haas, G.; Winkler, T. Helv. Chim. Acta 1985, 68, 1933–1935.