Theoretical study of the mechanism of formation of 4-aminoisoxazoles. 2*. An unusual rearrangement of oximinonitriles

Article abstract of DOI:10.1007/s10593-013-1342-x

The base-catalyzed cyclization of oximinonitriles in the presence of lithium ions leads to 4-aminoisoxazoles and, sometimes, to 5-aminooxazoles. Modeling using the density functional theory and the HF and MP2 methods indicated that both 4-aminoisoxazoles

Full text of DOI:10.1007/s10593-013-1342-x

Chemistry of Heterocyclic Compounds, Vol. 49, No. 7, October, 2013 (Russian Original Vol. 49, No. 7, July, 2013)
THEORETICAL STUDY OF THE MECHANISM OF FORMATION
OF 4-AMINOISOXAZOLES. 2*. AN UNUSUAL REARRANGEMENT
OF OXIMINONITRILES
V. P. Kislyi1**, E. B. Danilova1, and V. N. Solkan1
The base-catalyzed cyclization of oximinonitriles in the presence of lithium ions leads to
4-aminoisoxazoles and, sometimes, to 5-aminooxazoles. Modeling using the density functional theory
and the HF and MP2 methods indicated that both 4-aminoisoxazoles and 5-aminooxazoles are formed
from the anion of one of the oximinonitrile lithium complexes. The formation of 5-aminooxazoles
proceeds through an unusual anion rearrangement. The thermodynamic parameters of the transition
states were calculated to explain the predominant formation of 4-aminoisoxazoles.
Keywords: aminoisoxazoles, aminooxazoles, lithium complexes, anion rearrangement, density functional
method.
Functionally substituted 4-aminoisoxazoles (AI) 2 hold interest in light of the possibility of using these
compounds to obtain various condensed heterocyclic systems [2], in particular, isoxazolopyridines [3],
isoxazolopyrimidines, as well as analogs of N-alkyl-5-aminoisoxazoles, which display activity relative to
VEGFR-1 and VEGFR-2 receptors [5]. The base-catalyzed cyclization of oximinonitriles 1a-h was used to
obtain AI 2a-h [1, 3, 6]. The cyclization was carried out at room temperature in aqueous ethanol solutions and
the products were isolated simply by filtration.
N
X
NH2
X
X
N
B:
R
R
+
N
R
N
NH2
O
O
O
O
O
O
1a–h
2a–h
3a–h
a R = Me, X = CONH2; b R = Me, X = CONHPh; c R = Ph, X = CONHMe;
d R = Ph, X = CONHPh; e R = 4-BrC6H4, X = CONHPh; f R = 4-MeOC6H4, X = CONHPh;
g R = Ph, X = CONH(C6H3Cl2-3,5); h R = Me, X = Ph
_______
*For communication 1, see [1].
**To whom correspondence should be addressed, e-mail: vkislyi@yandex.ru.
1N. D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, 47 Leninsky Ave., Moscow
119991, Russia.
________________________________________________________________________________________
Translated from Khimiya Geterotsiklicheskikh Soedinenii, No. 7, pp. 1118-1131, July, 2013. Original
article submitted October 31, 2012; revision submitted March 3, 2013.
1042
0009-3122/13/4907-1042©2013 Springer Science+Business Media New York

TABLE 1. Conversion of Oximinonitriles 1a-h, Yields and Cyclization
Product Ratios Depending on the Reaction Conditions
Yield*2, %
Oxime or oxime
Reaction conditions*
20°C, CH3CN, Et3N, 4 h
20°C, 50% MeOH, LiOH*3, 4 h
45°C, 50% MeOH, LiOH*3, 1 h
20°C, 50% MeOH, KOH, 2 h
Conversion, %
complex
2
3
100
100
100
100
2
0
0
0
1
1a
1a·LiClO4·3H2O
1a·LiClO4·3H2O
1a·KBr
38
24
25
1b·LiClO4·3H2O
1b·KBr
20°C, 60% EtOH, LiOH*3, 3.5 h
20°C, 60% EtOH, KOH, 2 h
100
100
48
38
2
6
1c·LiClO4·3H2O
1c·KBr
20°C, 60% EtOH, LiOH*3, 4 h
20°C, 60% EtOH, KOH, 2 h
100
100
57
45
0
0
20°C, EtOH, KOCN, 48 h
20°C, EtOH, NaHCO3, 48 h
20°C, EtOH, K2CO3, 6 h
25
40
6
9
9
12
7
1d
95
25
4
20°C, MeOH, Et3N, 4 h
95
2
20°C, CH3CN, DBU, 2.5 h
20°C, 60% EtOH, LiOH*3, 4.5 h
45°C,60% EtOH, LiOH*3,1.5 h
60°C, 60% EtOH, LiOH*3, 1 h
20°C, 60% EtOH, KOH, 2 h
20°C, 60% MeOH, LiOH*3, 4 h
20°C, 60% EtOH, KOH, 2 h
97
3
2
99
32
13
2
4
100
100
100
99
1
0
32
36
33
7
1d·LiClO4·3H2O
1d·KBr
2
100
13
1e·LiClO4·3H2O
1e·KBr
20°C, 60% EtOH, LiOH*3, 4.5 h
20°C, 60% EtOH, KOH, 2 h
97
95
38
29
8
22
1f·LiClO4·3H2O
1f·KBr
20°C, 50% MeOH, LiOH*3, 3 h
20°C, 50% MeOH, KOH, 2 h
99
23
11
24
34
100
1g·LiClO4·3H2O
1g·KBr
20°C, 60% EtOH, LiOH*3, 4 h
20°C, 60% EtOH, KOH, 2 h
98
99
34
29
7
24
1h·LiClO4·3H2O
1h·KBr
20°C, 50% MeOH, LiOH*3, 4.5 h
20°C, 50% MeOH, KOH, 2 h
100
100
36
28
0
0
_______
*The reaction was carried out in a suspension of the oxime or oxime
complex in the indicated solvent.
*2The yields are indicated relative to the starting oxime by integrating
1
signals in the H NMR spectra of the reaction product isolated by
filtration, washed with water, and dried (without other purification
operations).
*3Fresh LiOH·H2O was used.
However, in some cases, 5-aminooxazoles (AO) 3 were also formed in the reaction [6]. The ratio of AI
2 to AO 3 depended on the structure of oximinonitrile 1 and the reaction conditions, in particular, the presence
of salts. AI 2 and AO 3 were formed in low yields when the reaction was carried out using triethylamine as the
base (Table 1). The complex reaction mixture suggested side reactions involving tar formation from the starting
oximinonitrile 1. Carrying out the reaction at elevated temperatures had an unfavorable effect on the yields of
AI 2, while there was hardly any formation of the desired products in the reaction mixture up to 60°C. The use
of strong inorganic bases such as LiOH and KOH substantially reduced the amount of the tar formation
products. The maximum yields of AI 2 were obtained using LiOH as the base and prior formation of the
complex of the oximinonitrile with LiClO4·3H2O. Under the optimum conditions found, AI 2 was sometimes
obtained without a trace of AO 3 but in the case of oximinonitriles 1b,d-g containing aryl substituents,
conditions could not be found for the complete suppression of the formation of AO 3.
1043

The structures of the compounds synthesized were confirmed by 1H and 13C NMR spectroscopy as well
as elemental analysis or high-resolution mass spectrometry data. When AI 2 and AO 3 had the same
substituents, these compounds had identical empirical formulas and molecular ions, but chemical shifts of the amino
groups in the 1H NMR spectra differed significantly: the amino group signal was found at 6.4-7.1 ppm for AI 2
and 8.1-8.3 ppm for AO 3. Upon TLC analysis of these compound mixtures, AO 3 were observed as a spot with
yellow phosphorescence below the spot of the corresponding AI 2. As a rule, phosphorescence was not found
for AI 2. In previous work [6], we established the structure of AO 3 by X-ray structural analysis.
In the present work, we studied the mechanisms of formation of AI 2 and AO 3 in an attempt to explain
the experimental results, specifically, the impossibility of carrying out the synthesis of AI 2 by the action of
organic bases and the requirement of lithium salts in the reaction mixture, as well as find reaction conditions,
under which AI 2 and AO 3 may be obtained regiospecifically.
In previous work [1], we established that oximinonitriles 1a,b,d existed in the crystal and solution in a
single preferred conformation with E-configuration relative to the double bond in the oxime fragment (Fig. 1).
The formation of AI 2 proceeded through deprotonation of the methylene group, which was close to the nitrile
group in the E-configuration of the oximinonitrile. Triethylamine was a rather efficient base for deprotonation,
but anions 4, formed by the action of triethylamine in the absence of lithium salts, underwent intermolecular
condensations instead of intramolecular cyclization to give AI 2.
In 2006, we proposed a mechanism to explain the formation of AO 3 [6], involving nucleophilic attack
at the nitrogen atom of the C=N double bond in enolates 5, leading to formation of the oxaziridine ring in
intermediate 6 and then to the anion of nitrile 7. The formation of nitrile 7 was followed by usual steps
involving closure of the ring at the O(1)–C(5) bond, protonation, and a proton shift in iminooxazoline 8 [7, 8].
Opening of oxaziridines at the N–O bond has been described by various workers [9-11]. However, this
mechanism appears unlikely in the final step due to high strain energy of the oxaziridine ring [12] and the lack
of examples in the literature of oxaziridine formation through closure of the C–N bond.
X
NH
N
–
X
B:
N
O
1
2
3
R
N
O
H
O
O
R
B:
4
8
N
X
N
N
X
X
–
N
–
N
O
N
O
O
O
R
HC
R
R
O
O
6
7
5
We should note that AO 3 could not be formed from AI 2 under the reaction conditions. Calculations
using the density functional method (B3LYP) and second-order Møller-Plesset perturbation theory (MP2) with
6-31+G* and 6-311++G** basis sets showed that both AI 2 and AO 3 formations proceed exothermally. The
enthalpies are from -25.6 to -34.8 kcal/mol for AI 2 and from -50.5 to -62.9 kcal/mol for AO 3 (Table 2). Thus,
AO 3 is 22-28 kcal/mol more stable than AI 2. The Hartree-Fock (HF) method leads to reduced values of the
formation heat of AI 2 and, thus, these calculations indicate that AO 3 are more stable than AI 2 by
30-37 kcal/mol. However, since the recyclization of isoxazoles to oxazoles requires UV irradiation or
high-temperature pyrolysis in vacuum [7, 8], such a transformation under our reaction conditions featuring no
UV irradiation and room temperature would seem impossible.
1044

TABLE 2. Calculated Values for the Changes in Enthalpy (H, kcal/mol)
and Gibbs Free Energy (G, kcal/mol) in the Cyclization of
Oximinonitriles 1 to AI 2 and AO 3 at 298.15 K and 1 atm
Method
2a
2d
2h
3a
3d
3h

B3LYP/6-31+G(d)
-34.76
-32.24
-33.18
-30.71
-27.69
-24.51
-61.99
-59.65
-60.13
-57.69
-50.59
-47.75
H
G
B3LYP/6-
H
G
-34.80
-32.82
-33.26
-30.96
-27.88
-24.86
-62.92
-60.88
-61.11
-58.52
-50.49
-47.75
311++G(d,p)
HF/6-31+G(d)
H
G
-21.71
-18.68
-19.42
-16.75
-16.22
-13.13
-58.78
-56.03
-56.47
-53.76
-46.19
-43.58
MP2/6-31+G(d)
H
-32.73
-32.46
—
-25.66
-25.95
-59.56
-57.80
—
-48.69
-48.63
G
In the present work, we used the density functional method to study the structure of oximinonitrile
complexes with lithium ions, the potential energy surfaces (PES) of the oximinonitrile free anions in the absence
of salts and deprotonated lithium complexes in vacuum and in polar solvents, and the structure of the transition
states in the formation of AI 2 and AO 3.
The reaction of a lithium ion with oximinonitrile 1 may lead to complexes 9-11, in which three different
types of coordination may be realized with +1 total charge of the complex (Figs. 1, 2).
Fig. 1. General view (a) of oximinonitrile 1a and (b) lithium complex 9a (B3LYP/6-31+G(d)).
Fig. 2. General view of complexes 10a (a) and 11a (b) (B3LYP/6-31+G(d)).
The loss of a proton from the methylene group in complexes 9-11 is followed by formation of the
corresponding deprotonated lithium complexes 12-14 (DLC) with zero total charge, which were found to be
essentially planar in B3LYP/6-31+G(d) and B3LYP/6-311++G(d,p) calculations.
1045

N
Li+
N
H2N
O
N
Li+
H2N
O
N
N O
Li+
O
O
O
O
N
O
H2N
O
Me
Me
Me
9a
B:
10a
11a
B:
B:
N
H2N
O
Li+
N
N
H2N
O
Li
N O
O
N
O
Me
O
O
Li
O
H2N
N
O
Me
13a
12a
Me
14a
+
2a
3a
A study of the PES for DLC 12a showed that rotation about the N–O bond gives conformation A
(Fig. 3), in which a C–C bond is formed with closure of the isoxazole ring of AI 2. At the same time, incomplete
rotation about the N–O bond and an additional rotation about the O–CH2 bond give conformation B, in which a
rearrangement occurs to form lithium nitrile complex 7 and subsequent formation of AO 3.
Fig. 3. Transition states in the formation of (a) AI 2 (conformation A) and (b) AO 3 (conformation B) for
compounds 2a and 3a.
The course of the rearrangement leading to nitrile 7 was studied in greater detail. Our
B3LYP/6-31+G(d) calculation showed that bond lengths either increase or decrease with the corresponding
changes in valence angles during the course of the rearrangement in conformation B (Table 3) with the exception of
the length of the C(2)–N(2) bond, which initially increases but then decreases to virtually the previous value. The
C(2)=N(2) bond length is 1.368 Å in the oxime fragment in planar enolate 12a, later increases to 1.444 Å during
the rearrangement, and then decreases to 1.301 Å in the final lithium nitrile complex 7a. At the point
corresponding to the maximum C(2)–N(2) bond length, the distances between the N(2), C(4), and O(2) atoms
and the valence angles are typical for oxaziridines. The corresponding parameters of oxaziridines were found by
X-ray structural analysis: N–O, 1.49-1.54; O–C, 1.42-1.43; C–N, 1.45-1.47 Å; N–O–C, 59°, O–C–N, 64°
[13-16]. On the other hand, there is no energy minimum corresponding to the oxaziridine ring on the PES,
which would appear to obviate formation of intermediate 6 during the rearrangement.
1046

TABLE 3. Changes in Geometrical Parameters in the Rearrangement of
DLC 12a to Lithium Nitrile Complex 7a
Bond lengths and valence Bond lengths (Å) and valence angles (deg) during the rearrangement*
angles in the transition state
1
2
3
4
С(2)–N(2)
1.294
1.338
2.30
1.368
1.322
2.301
1.339
1.428
1.244
101
1.444
1.501
1.466
1.414
1.533
1.2125
60.3
1.301
2.077
1.486
1.335
1.549
1.220
45.5
N(2)–O(2)
N(2)–C(4)
O(2)–C(4)
1.437
1.541
1.211
111.6
32.8
C(4)–C(5)
C(5)–O(3)
N(2)–O(2)–C(4)
N(2)–C(4)–O(2)
39.2
62.8
94.6
_______
*1 - lithium complex 9a, 2 - enolate 12a, 3 - geometry corresponding to
the maximum C(2)–N(2) bond length upon shift along the reaction
coordinate from the transition state to the products, 4 - lithium nitrile
complex 7a.
A study of the PES for DLC 13a and 14a did not reveal saddle points in the transition to AI 2 and AO 3.
This result may be attributed in classical terms of nucleophilicity and electrophilicity to the positioning of the
lithium ion next to oxygen atom of the enolate, that reduces the negative charge on the enolate carbon atom and
leads to a sharp drop in its nucleophilicity. At the same time, the lithium ion in complex 13 is found between the
nitrile group and enolate fragment, which hinders nucleophilic attack at the nitrile group carbon atom, while
there is no activating effect of the lithium ion on the nitrile group in complex 14. DLC 13 and 14 prove entirely
inert in cyclization reactions when there no activating effect on the nitrile group and sharply reduced
nucleophilicity in the enolate fragment.
AI 2 is formed in the case of anions 4 lacking lithium from conformation A although the activation energies
are much greater than in the case of the lithium complexes (see below). A study of the PES for anions 4 did not
reveal saddle points corresponding to the transition to AO 3, that indicates the impossibility of the first step in the
formation of AO 3 in the absence of lithium ions. This result is in agreement with our experimental data (Table 1).
We should note that generalization of the calculated data for the reaction studied is impossible without
taking account of the solvation of the various types of complexes and the transition states since the solvation
energies for singly-charged cations in polar solvents such as water and methanol fall in the range 60-120
kcal/mol [17]. Taking account of solvation effects may be accomplished by various methods including the
Tomasi polarized continuum model and various discrete models, in particular, models explicitly considering the
structure of the first solvation shell [18, 19].
The structure of the lithium complexes has been studied repeatedly by X-ray structural analysis [20] and
theoretical calculations [21-25]. In individual cases such as in the [Li6(H2O)16]6+ ion, Xie and Ma [26] found
octahedral coordination at the lithium ion involving water molecules. However, the formation of tetrahedral
[Li(H2O)4]1+ ions is observed in the overwhelming majority of cases [20]. The lithium ion forms tetrahedral
Li(OR1R2)4]1+ complexes with bulkier alcohol and ester molecules. In aqueous methanol solutions, equilibrium
is established between [Li(H2O)n(MeOH)4-n]1+ ions, where n = 0-4. The content of the various complexes
depends on the amounts of water and methanol, temperature, and pressure, but was not studied in this work. It
would be natural to propose that oximinonitrile 1 replaces two water molecules in the [Li(H2O)4]1+ ion to give
complex 15a (Fig. 4), in which the lithium ion has tetrahedral coordination (the matrix for structure 15a was
obtained by adding two water molecules to the matrix of complex 9a). However, upon optimization of the
geometry, we found that complexes 15 rearrange to complexes 16 without any barrier. The lithium ion in
complex 16 has coordination number 3 and, thus, is coordinatively unsaturated.
1047

The existence of complexes 16 is possible in the crystalline phase or in nonpolar solvents, but in
aqueous solution complex 16a adds a water molecule to give trihydrate complex 17a, in which lithium has
tetrahedral coordination and coordination number 4 (Fig. 5). Trihydrate complexes 18 and 19 are analogously
formed upon the hydration of complexes 10 and 11.
The formation of lithium trihydrate complexes is partially confirmed by the calculated values of the
formation heats of the anhydrous and trihydrate oximinonitrile complexes. The B32LYP/6-31+G(d) calculation
showed that the reaction of lithium ions with oximinonitriles 1 in vacuum proceeds with a significant
exothermal effect (from -60 to -72 kcal/mol, Table 4).
On the other hand, we obtained a value of from -8 to -13 kcal/mol for the formation of trihydrate
complexes in the reaction
[Li(H2O)4]1+ + 1 → [Li(1)(H2O)3]1+ + H2O
also from a B3LYP/6-31-G(d) calculation (Table 4).
Fig. 4. General view of a molecule of (a) lithium complex 15a (prior to geometry optimization) and (b)
dehydrated complex 16a (after geometry optimization) from B3LYP/6-31+G(d) calculations.
This reaction is slightly exothermal but the equilibrium shifted toward formation of the complexes of the ligand
with lithium. On the whole, the calculated results are in agreement with the experimental findings: strong
heating and considerable tar formation were observed in an attempt to obtain complexes by the reaction of
anhydrous lithium salts with oximinonitriles in benzene. When the reaction is carried out in methanolic solution,
weak heating was observed and the orange-yellow color of the oximinonitrile solutions became less intense.
Fig. 5. General view of lithium trihydrated complexes (a) 17a and (b) 18a from B3LYP/6-31+G(d) calculations.
1048

Using the known equation relating the relation of the free energy difference to the equilibrium constant
K = exp(-G0/RT),
where R is the gas constant and T is the absolute temperature in kelvins, we calculated the ratios of anhydrous
complexes 9a, 10a, 11a to trihydrated complexes 17a, 18a, 19a using the relative Gibbs free energy (Table 5).
B3LYP calculations with the 6-31+G(d) and 6-311++G(d,p) basis sets showed that the relative stabilities of
complexes 9a, 10a, 11a and the corresponding hydrated complexes 17a, 18a, 19a differ significantly. The
reaction of the unsolvated lithium ion with oximinonitrile 1a leads to the formation of complex 10a. The content
of the "reactive" complex 9a is on the level of the content of the enol form in acetone. In the case of the
trihydrated complexes, the "reactive" complex 17a is the most stable, and its content amounts to 97-99%. The
high content of complexes 17 is favorable for achieving cyclization.
TABLE 4. Changes in Enthalpy (H, kcal/mol) and Gibbs Free Energy (G,
kcal/mol) in the Formation of Anhydrous Lithium Complexes 9a, 10a, 11a
and Hydrated Lithium Complexes 17a, 18a, 19a at 298.15 K and 1 atm
Method
9a
10a
11a
17a
18a
19a

B3LYP/6-31+G(d)
-60.88
-51.64
-71.73
-61.63
-62.05
-53.37
-13.36
-7.22
-11.2
-5.26
-8.52
-2.52
H
G
B3LYP/6-311++G(d,p) H
G
-61.17
-51.87
-71.68
-61.49
-62.32
-53.58
-13.50
-9.67
-11.62
-7.63
-8.92
-5.12
Deprotonation of lithium complexes 17 leads to DLC 20 and then to the corresponding transition states
(conformations A and B), which cyclize further to AI 2a or undergo rearrangement to give AO 3a. In the case of
trihydrated DLC 21a and 22a, as in the case of anhydrous DLC 10a and 11a, a study of the PES did not reveal
saddle points for the transition to AI 2a or AO 3a.
H
N
H2N
O
H2N
N
H2O
H
O
H
Li+
H
N
O
Li+
H2O
O
OH2
OH2
N
O
N
O
O
H
O
O
O
N
O
Li+
H
H2N
O
Me
Me
18a
B:
Me
H2O
17a
B:
H2O
19a
B:
H
H2O
H2O
O
H
H2N
N
Li+
N
H
N
H2N
H
O
O
O
OH2
OH2
O
N
Li
O
O
N
O
N
O
H2N
O
H
O
O
20a
Me
H
H2O
Li
H2O
22a
Me
21a
Me
+
2a
3a
Table 6 shows the thermodynamic parameters of the transition states in the formation of AI 2a and
lithium nitrile complex 7a (initial step in the formation of AO 3a) relative to the corresponding planar enolates.
1049

When the reaction is carried out without lithium ions, the activation energy for the formation of AI 2a
from anions 4a is too high (45.9 kcal/mol), and other reactions (intermolecular condensation, saponification
involving the nitrile and/or amide groups, etc.) are realized in the system. This result is in accord with an
experiment, in which tar formation occurs in an attempt to carry out the reaction by the action of organic bases
(Table 1). A very significant decrease in the energy of the transition state in the formation of AI 2 is observed in
the presence of the nonhydrated lithium ion. In the case of lithium trihydrated complexes, the decrease in
activation energy is not as marked. This finding may be attributed to weak bonding of the hydrated lithium ion to
oximinonitrile 1a (Table 4). In the case of oximinonitrile 1a, the transition states to formation of AO 3a both with the
nonhydrated and hydrated lithium ion have higher energies than the transition states to formation of AI 2a.
We should note that, in addition to trihydrated complexes, solvated complexes [Li(1)(H2O)3-n(MeOH)n]1+,
where n = 1-3, with partial or complete replacement of the water molecules by methanol molecules may also
exist in aqueous methanol solutions. Such complexes may also be deprotonated by base and then cyclize to give
AI 2 or AO 3. We propose that in the case of the methanol complexes, the transition state energies are similar to
the corresponding values for the trihydrated complexes and very different from the energies of transition states
without a lithium ion. In this work, we did not calculate the relative contents of the methanol-containing lithium
complexes and thermodynamic parameters of the corresponding transition states in light of the large number of
geometrical and optical isomers of such complexes.
TABLE 5. Relative Gibbs Free Energies (G) and Ratio of the Lithium
Oximinonitrile Complexes at 298.15 K and 1 atm
G, kcal/mol
Ratio, %
Method
9a
10a
11a
9a : 10a : 11a
B3LYP/6-31+G(d)
9.99
9.61
0.0
0.0
8.26
7.91
4.7·10–6 : 99.9999 : 8.7·10–5
8.9·10–6 : 99.9998 : 1.6·10–5
B3LYP/6-311++G(d,p)
17a
0.0
0.0
18a
1.96
2.04
19a
4.70
4.56
17a : 18a : 19a
99.44 : 3.52 : 0.036
96.9 : 3.09 : 0.045
B3LYP/6-31+G(d)
B3LYP/6-311++G(d,p)
TABLE 6. Thermodynamic Parameters of the Transition States in the
Formation of AI 2 and Nitrile 7a at 298.15 K and 1 atm from a
B3LYP/6-31+G(d) Calculation
Starting enolate
Parameter*
Transition state to AI 2a Transition state to nitrile 7a

H
G
S
IF
45.86
51.60
—
4a
-19.25
-378.39
23.53
H
G
S
IF
27.28
30.59
12a
20a
28.45
-16.5
-11.1
-271.19
29.83
-163.52
31.26
H
G
S
IF
36.48
36.92
-22.303
-642.05
-19.00
-183.79
_______
*H is the enthalpy change (kcal/mol), G is the change in Gibbs free
energy (kcal/mol), and S is the entropy change (cal/(mol·K)) in going
from the planar enolate to the transition state. IF is the imaginary
frequency in the transition state (cm-1)
1050

Our experiments showed that the formation of AO 3 is much more facile in the case of oximinonitriles
containing aryl groups in the acetonyl or amide fragment (Table 1). The introduction of bulky aryl groups into
the oximinonitrile molecule probably has a greater effect on the stability of conformation A than on the stability
of conformation B, which leads to change in the relative stabilities of these conformations and, as a
consequence, change in the ratio between the reaction products.
Hence, we may conclude that solvents and/or ligands with weak electron-donating properties should be
used for regioselective preparation of 4-aminoisoxazoles. On the other hand, 5-aminooxazoles probably cannot
be prepared by this method, and other synthetic approaches should be employed to obtain these compounds
[7, 8].
EXPERIMENTAL
1
13
The H and C NMR spectra were recorded on Bruker DRX-500 (500 MHz) and Bruker AM-300 (75
MHz) spectrometers, respectively, in DMSO-d6 with TMS as internal standard. The high-resolution electrospray
ionization positive ion mode mass spectra were recorded on a Bruker micrOTOF II instrument. The melting
points were determined on a Boetius PHMK05 apparatus with heating at 4°C/min. The syntheses of compounds
1a,b,d, 2d, and 3d were carried out according to our previous methods [1]. The syntheses of compounds 1h and
2h were carried out according to the procedures of Gewald et al. [3] with slight changes.
2-Cyano-N-methyl-2-[(2-oxo-2-phenylethoxy)imino]acetamide (1c) was obtained according to our
1
previous procedure [1]. Yield 87%. Mp 162-164°С (2-PrOH). Н NMR spectrum, , ppm (J, Hz): 2.70 (3Н, d,
3J = 4.7, NСН3); 5.95 (2H, s, CH2); 7.61 (2Н, t, 3J = 7.7, Н-3,5 Ph); 7.74 (1Н, t, 3J = 7.3, Н-4 Ph); 8.00 (2Н, d,
3J = 7.1, Н-2,6 Ph); 8.61 (1H, br. s, NH). Found, m/z: 246.0871 [M+H]+. C12H12N3O3. Calculated, m/z:
246.0873.
2-{[2-(4-Bromophenyl)-2-oxoethoxy]imino}-2-cyano-N-phenylacetamide
(1e)
was
obtained
according to our previous procedure [1]. Yield 93%. Mp 195-197°C (PrOH) (mp 194-197°С (EtOH) [27]).
1Н NMR spectrum, , ppm (J, Hz): 5.95 (2H, s, CH2); 7.18 (1Н, t, 3J = 7.5, Н-4 Ph); 7.35 (2Н, t, 3J = 7.5, Н-3,5
3
3
3
Ph); 7.65 (2Н, d, J = 7.4, Н-2,6 Ph); 7.80 (2Н, d, J = 8.3, Н-3,5 Ar); 7.92 (2Н, d, J = 8.2, Н-2,6 Ar); 10.41
(1H, br. s, NH).
2-Cyano-2-{[2-(4-Methoxyphenyl)-2-oxoethoxy]imino}-N-phenylacetamide (1f) was obtained
according to our previous procedure [1]. Yield 91%. Mp 182-184°С (EtOH) (mp 182-185°С (EtOH) [6]). 1Н NMR
spectrum, , ppm (J, Hz) 3.88 (3Н, s, OCH3); 5.94 (2Н, s, CH2); 7.09-7.19 (3Н, m, Н-3,5 Ar, H-4 Ph); 7.36 (2Н, t,
3J = 8.1, Н-3,5 Ph); 7.67 (2Н, d, 3J = 8.1, Н-2,6 Ph); 7.96 (2Н, d, 3J = 8.0, Н-2,6 Ar); 10.42 (1H, br. s, NH).
N-(3,5-Dichlorophenyl)-2-cyano-2-[(2-oxo-2-phenylethoxy)imino]acetamide (1g) was obtained
according to our previous procedure [1]. Yield 94%. Mp 188-191°C (EtOH) (mp 186-190°С (EtOH) [6]).
1Н NMR spectrum, , ppm (J, Hz): 6.04 (2Н, s, CH2); 7.43 (1Н, s, Н-4 Ar); 7.61 (2Н, t, J = 7.3, H-3,5 Ph),
3
7.75 (1Н, t, 3J = 7.3, Н-4 Ph); 7.79 (2Н, s, H-2,6 Ar); 8.02 (2Н, d, 3J = 7.3, Н-2,6 Ph); 10.70 (1H, br. s, NH).
Preparation of 4-Aminoisoxazoles 2a-c,e-g (General Method). Corresponding oximinonitrile 1
(5 mmol) was mixed with an equimolar amount of LiClO4·3H2O and dissolved in a minimal amount of warm
DMF. The solution obtained was added dropwise over 20 min to 2.5% LiOH solution in water or 50% aqueous
methanol (10 ml) and then stirred until the complete disappearance of the starting O-alkylated nitrile as monitored
by thin-layer chromatography. These reactions required 1-4 h. The precipitated product was filtered off, washed on
the filter with 2% hydrochloric acid and then water, and recrystallized (compounds 2c,e) or separated by
chromatography (2f,g).
5-Acetyl-4-aminoisoxazole-3-carboxamide (2a). Yield 32%. Mp 186-189°С (EtOH). 1Н NMR
spectrum, , ppm (J, Hz): 2.45 (3H, s, СН3); 5.95 (2H, br. s, NH2); 7.94 (1H, br. s) and 8.21 (1H, br. s, NH2CO).
Found, %: C 42.28; H 4.31; N 24.35. C6H7N3O3. Calculated, %: C 42.61; H 4.17; N 24.84.
1051

5-Acetyl-4-amino-N-phenylisoxazole-3-carboxamide (2b). Yield 43%. Mp 181-183°С (EtOH) (mp
179-182°С (EtOH) [27]). 1Н NMR spectrum, , ppm (J, Hz): 2.47 (3H, s, СН3); 6.09 (2H, br. s, NH2); 7.15 (1Н,
t, 3J = 7.8, Н-4 Ph); 7.38 (2Н, t, 3J = 7.8, Н-3,5 Ph); 7.80 (2Н, d, 3J = 8.0, Н-2,6 Ph); 10.87 (1H, br. s, NHCO).
4-Amino-5-benzoyl-N-methylisoxazole-3-carboxamide (2c). Yield 52%. Mp 175-178°С (EtOH).
1Н NMR spectrum, , ppm (J, Hz): 2.80 (3H, d, 3J = 4.6, СН3); 6.4 (2H, br. s, NH2); 7.60 (2Н, t, 3J = 7.8, Н-3,5
3
3
Ph); 7.70 (1Н, t, J = 7.8, Н-4 Ph); 8.07 (2Н, d, J = 7.8, Н-2,6 Ph); 8.94 (1H, br. s, NHCO). 13C NMR
spectrum, , ppm: 25.6 (СН3); 128.7 (С Ph); 133.0 (С Ph); 136.0 (С-4); 138.0 (С-1 Ph); 146.7 (С-5); 147.7
(С-3); 159.6 (CONH); 180.2 (COAr). Found, m/z: 246.0874 [M+H]+. C12H12N3O3. Calculated, m/z: 246.0873.
4-Amino-5-(4-bromobenzoyl)-N-phenylisoxazole-3-carboxamide (2e). Yield 32%. Mp 236-239°С
1
(PrOH) (mp 235-238°С (EtOH) [27]). Н NMR spectrum, , ppm (J, Hz): 6.52 (2Н, br. s, NH2); 7.20 (1Н, t,
3J = 7.6, Н-4 Ph); 7.40 (2Н, t, 3J = 7.6, Н-3,5 Ph); 7.80 (2Н, d, 3J = 7.4, Н-2,6 Ph); 7.87 (2Н, d, 3J = 8.4, Н-3,5
Ar); 8.05 (2Н, d, 3J = 8.4, Н-2,6 Ar); 10.94 (1H, br. s, NHCO). 13C NMR spectrum, , ppm: 120.8 (С Ph); 124.6
(С Ph); 127.2 (С Ar); 128.7 (PhNH); 130.7, 131.9 (С-2,3,5,6 Ar); 135.7 (C-4); 137.6 (С-1 Ph); 138.4 (С Ar);
146.7 (C-5); 148.2 (C-3); 157.7 (CONH); 179.0 (COAr).
4-Amino-5-(4-methoxybenzoyl)-N-phenylisoxazole-3-carboxamide (2f). Yield 18%. Mp 235-237°С
1
(EtOH) (mp 236-238°С (EtOH) [6]). Н NMR spectrum, , ppm (J, Hz): 3.90 (3H, s, OCH3); 6.37 (2Н, br. s,
3
3
NH2); 7.15-7.20 (3Н, m, H-3,5 Ar, H-4 Ph); 7.40 (2Н, t, J = 7.6, Н-3,5 Ph); 7.82 (2Н, d, J = 7.4, Н-2,6 Ph);
8.18 (2Н, d, 3J = 8.4, Н-2,6 Ar); 10.90 (1H, br. s, NHCO)
4-Amino-5-benzoyl-N-(3,5-dichlorophenyl)isoxazole-3-carboxamide (2g). Yield 28%. Mp 232-234°С
1
(EtOH) (mp 232-234°С (EtOH–Н2О) [6]). Н NMR spectrum, , ppm (J, Hz): 6.65 (2H, br. s, NH2); 7.19 (1Н,
s, H-4' Ar); 7.59-7.72 (3Н, m, Н-3,4,5 Ph); 7.86 (2Н, s, Н-2',6' Ar); 8.10 (2Н, d, 3J = 8.1, Н-2,6 Ph); 11.06 (1H,
br. s, NHCO).
Preparation of 5-Aminooxazoles 3 (General Method). A mixture of oximinonitrile 1 (5.00 mmol) and
potassium bromide (2.0 g, 16.80 mmol) in 60% aqueous ethanol (4 ml) was left overnight in an open beaker.
The moist precipitate obtained was suspended in 60% aqueous ethanol (15-25 ml), and 10% aqueous potassium
hydroxide (0.6 g, 1.07 mmol) was added with stirring to the suspension. After 2 h, the precipitate was filtered
off, washed on the filter with 2% hydrochloric acid and then water, and dried. The resultant mixture was
dissolved in acetonitrile (25 ml). An equal amount (by weight) of silica gel was added, and the mixture was
evaporated on a rotary evaporator. The resultant powder was placed on a chromatographic column. AI 2 and AO
3 were eluted consecutively with benzene, 10:1 benzene–ethyl acetate, and 2:1 benzene–ethyl acetate. The
composition of the fractions was monitored by thin-layer chromatography. AO 3e,g were isolated as pure
compounds.
5-Amino-2-(4-bromobenzoyl)-N-phenyloxazole-4-carboxamide (3e). Yield 19%. Mp 232-234°C
1
3
(EtOH) (mp 230-232°С (EtOH) [6]). Н NMR spectrum, , ppm (J, Hz): 7.15 (1Н, t, J = 7.9, Н-4 Ph); 7.35
3
3
(2Н, t, J = 8.1, Н-3,5 Ph); 7.80 (4Н, d, J = 8.1, Н-2,6 Ar, Н-2,6 Ph); 8.12 (2Н, br. s, NH2); 8.40 (2Н, d,
3J = 8.5, Н-3,5 Ar); 9.60 (1H, br. s, NHCO). 13C NMR spectrum, , ppm: 108.4 (C-4); 120.3 (C-2,6 Ph); 123.3
(C-4 Ph); 127.4 (C-4 Ar); 128.5 (C-3,5 Ar); 131.5 (C-3,5 Ph); 132.3 (C-2,6 Ar); 134.2 (C-1 Ar); 138.6 (C-1 Ph);
144.9 (C-2); 160.3 (CONH); 160.8 (C-5); 174.6 (COAr).
5-Amino-2-benzoyl-N-(3,5-dichlorophenyl)oxazole-4-carboxamide (3g). Yield 22%. Mp 234-236°C
1
4
(EtOH) (mp 234-236°С (EtOH) [6]). Н NMR spectrum, , ppm (J, Hz): 7.26 (1Н, t, J = 1.8, Н-4 Ar);
4
3
7.57-7.70 (3Н, m, H-3,4,5 Ph); 7.98 (2Н, d, J = 1.8, Н-2,6 Ar); 8.20 (2H, br. s, NH2); 8.43 (2Н, d, J = 8.0,
Н-2,6 Ph); 10.00 (1H, br. s, NHCO). 13C NMR spectrum, , ppm: 107.7 (C-4); 118.1 (C-2,6 Ar); 122.1 (C-4
Ar); 128.5 (C-3,5 Ph); 130.3 (C-2,6 Ph); 133.3 (C-4 Ph); 133.8 (C-3,5 Ar); 135.1 (C-1 Ph); 141.3 (C-1 Ar);
145.1 (C-2); 160.7 (CONH); 161.3 (C-5); 176.0 (COPh).
Calculation Procedures. The calculations were carried out using the Gaussian03 program package [28]
using the B3LYP hybrid method [29] and nonempirical HF [28] and MP2 methods [30] (Tables 1-6).
Calculation of the frequency was carried out for all the stationary points in order to confirm accordance of the
optimized geometry to the energy minimum. In the case of transition states, the imaginary frequency
1052

corresponding to the reaction coordinate was found. Calculation of the reaction pathway by the internal reaction
coordinate method was also carried out for transition states. The frequencies were calculated using the PCM
solvation model in order to calculate the thermodynamic parameters in solution [18]. Numerical frequency
calculations [28] were used for the MP2 method. The JMol 12.2.27 [31, 32] or ChemCraft packages [33] were
used for visualization of the results.
REFERENCES
1.
2.
V. P. Kislyi and E. B. Danilova, Khim. Geterotsikl. Soedin., 1108 (2013). [Chem. Heterocycl. Compd.,
49, 1033 (2013)].
V. P. Kislyi, E. B. Danilova, and V. V. Semenov, in: A. R. Katritzky (editor), Advances in Heterocyclic
Chemistry, Vol. 94, Elsevier, New York (2007), p. 173.
3.
4.
K. Gewald, P. Bellmann, and H.-J. Jänsch, Liebigs Ann. Chem., 1623 (1980).
H. Haning, U. Niewöhner, T. Schenke, T. Lampe, A. Hillisch, and E. Bischoff, Bioorg. Med. Chem.
Lett., 15, 3900 (2005).
5.
6.
7.
8.
A. S. Kiselyov, E. L. Piatnitski, A. V. Samet, V. P. Kisliy, and V. V. Semenov, Bioorg. Med. Chem. Lett.,
17, 1369 (2007).
V. P. Kislyi, E. B. Danilova, V. V. Semenov, A. A. Yakovenko, and F. M. Dolgushin, Izv. Akad. Nauk,
Ser. Khim., 1773 (2006).
G. V. Boyd, in: A. R. Katritzky and C. W. Rees (editors), Comprehensive Heterocyclic Chemistry, Vol. 6,
Elsevier, New York (1984), p. 177.
V. Yeh and R. Iyengar, in: A. Katritzky, C. Ramsden, E. Scriven, and R. Taylor (editors), Heterocyclic
Chemistry III, Vol. 4, Elsevier, New York (2008), p. 487.
9.
10.
11.
12.
K. Suda, T. Umehara, and F. Hino, Chem. Pharm. Bull., 38, 839 (1990).
C. Yijima, F. Hino, and K. Suda, Tetrahedron Lett., 21, 4725 (1980).
P. Duhamel, D. Bénard, and J.-C. Plaquevent, Tetrahedron Lett., 26, 6065 (1985).
F. A. Davis, B.-C. Chen, and P. Zhou, in: A. Katritzky, C. Ramsden, E. Scriven, and R. Taylor (editors),
Comprehensive Heterocyclic Chemistry III, Vol. 1, Elsevier, New York (2008), p. 559.
S. Schoumacker, O. Hamelin, S. Téti, J. Pécaut, and M. Fontecave, J. Org. Chem., 70, 301 (2005).
E. Bourguet, J.-L. Baneres, J.-P. Girard, J. Parello, J.-P. Vidal, X. Lusinchi, and J.-P. Declercq, Org.
Lett., 3, 3067 (2001).
13.
14.
15.
16.
J. Vidal, S. Damestoy, L. Guy, J.-C. Hannachi, A. Aubry, and A. Collet, Chem.-Eur. J., 3, 1691 (1997).
A. Lattes, E. Oliveros, M. Rivière, C. Belzecki, D. Mostowicz, W. A. Abramskj, C. Piccini-Leopardi,
G. Germain, and M. Van Meerssche, J. Am. Chem. Soc., 104, 3929 (1982).
I. D. Zaitsev and G. G. Ageev, Physicochemical Properties of Binary and Multicomponent Solutions of
Inorganic Compounds [in Russian], Khimiya, Moscow (1988), p. 372.
J. Tomasi, B. Mennucci, and R. Cammi, Chem. Rev., 105, 2999 (2005).
E. V. Butyrskaya, Computer Chemistry: Theoretical Fundamentals and Working With Gaussian and
GaussView Programs [in Russian], Solon-Press, Moscow (2011), p. 84.
U. Olsher, R. M. Izatt, J. S. Bradshaw, and N. K. Dalley, Chem. Rev., 91, 137 (1991).
A. C. Hoepker and D. B. Collum, J. Org. Chem., 76, 7985 (2011).
17.
18.
19.
20.
21.
22.
B. Lecachey, H. Oulyadi, P. Lameiras, A. Harrison-Marchand, H. Gerard, and J. Maddaluno, J. Org.
Chem., 75, 5976 (2010).
23.
24.
25.
H. K. Khartabil, P. C. Gros, Y. Fort, and M. F. Ruiz-López, J. Org. Chem., 73, 9393 (2008).
L. M. Pratt, S. C. Nguyen, and B. T. Thanh, J. Org. Chem., 73, 6086 (2008).
A. B. Sannigrahi, T. Kar, B. G. Niyogi, P. Hobza, and P. von Rague Schleyer, Chem. Rev., 90, 1061
(1990).
1053

26.
27.
28.
A.-L. Xie and C.-A. Ma, Acta Crystallogr., C: Cryst. Struct. Commun., 61, i67 (2005).
V. P. Kislyi, E. B. Danilova, and V. V. Semenov, Izv. Akad. Nauk, Ser. Khim., 1159 (2005).
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A.
Montgomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi,
V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada,
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,
M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, B. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratman, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,
P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich,
A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman,
J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and
J. A. Pople, Gaussian 03, Revision C.1, Gaussian Inc., Pittsburgh (2003).
29.
30.
31.
32.
33.
C. Lee, W. Yang, and R. G. Parr, Phys. Rev., B37, 785 (1988).
C. Møller and M. S. Plesset, Phys. Rev., 46, 618 (1934).
Jmol: An open-source Java viewer for chemical structures in 3D, http://www.jmol.org/.
R. M. Hanson, J. Appl. Crystallogr., 43, 1250 (2010).
G. A. Zhurko and D. A. Zhurko, http://www.chemcraftprog.com.
1054