Studies on azole-to-azole interconversion - An interesting case of absence of a "primary steric effect" in the ring-degenerate equilibration between ortho-substituted 3-aroylamino-5-methyl-1,2,4-oxadiazoles and 3-acetylamino-5-aryl-1,2,4-oxadiazoles in me

Article abstract of DOI:10.1002/1099-0690(200204)2002:8<1417::AID-EJOC1417>3.0.CO;2-G

The title reaction has been studied by ¹H NMR, both in CD₃OD and in tBuOK/CD₃OD. The components of the electronic effect of the substituent on the benzene ring show different contributions depending on whether the equilibr

Full text of DOI:10.1002/1099-0690(200204)2002:8<1417::AID-EJOC1417>3.0.CO;2-G

FULL PAPER
Studies on Azole-to-Azole Interconversion ؊ An Interesting Case of Absence of
a ‘‘Primary Steric Effect’’ in the Ring-Degenerate Equilibration between ortho-
Substituted 3-Aroylamino-5-methyl-1,2,4-oxadiazoles and 3-Acetylamino-5-
aryl-1,2,4-oxadiazoles in Methanol
[a]
[a]
[a]
[a]
`
Silvestre Buscemi, Vincenzo Frenna, Andrea Pace, Nicolo Vivona,
Barbara Cosimelli,[b] and Domenico Spinelli*[b]
Keywords: Free energy relationships / NMR spectroscopy / Oxygen heterocycles / Rearrangements / Ring-ring
interconversions
The title reaction has been studied by 1H NMR, both in
CD3OD and in tBuOK/CD3OD. The components of the elec- because of the importance of the internal amidic conjugation.
effect’’ does not significantly affect the reactivity, probably
tronic effect of the substituent on the benzene ring show dif-
ferent contributions depending on whether the equilibrium
between neutral (strong prevalence of the ‘‘proximity polar
effect’’) or anionic species (balance between ‘‘ordinary and
For the ortho-methoxy derivative, interesting ‘‘special’’ prox-
imity effects have been observed.
(
Wiley-VCH Verlag GmbH, 69451 Weinheim, Germany,
proximity polar effects’’) is considered. The ‘‘primary steric 2002)
Introduction
ment is that the starting and final products have similar
stabilities.
Examples of potentially reversible ring interconversions
are collected in Scheme 2; it has been observed, for ex-
ample, that:
• as expected, the (Z)-oximes of 3-acyl-1,2,4-oxadiazoles
(3) rearrange irreversibly[4] into the corresponding 3-acyl-
amino-1,2,5-oxadiazoles (4), the higher stability of which
essentially depends both on the greater aromaticity of 1,2,5-
oxadiazole (Bird’s I5 index:[5] 43) with respect to 1,2,4-oxa-
diazole (Bird’s I5 index: 39) and on the effective resonance
within the amido group;
The Boulton, Katritzky, and Majid-Hamid scheme,[1a]
a
monocyclic rearrangement of heterocycles in which an un-
saturated side chain is involved (see Scheme 1), represents
an interesting pathway for the synthesis of five-membered
nitrogen aromatic heterocycles.[1] Its scope has been en-
larged by Korbonits and co-workers,[2] who pointed out
that in 1,2,4-oxadiazoles a ringϪring interconversion into
five-membered nitrogen dihydroheterocycles (e.g., pyrazol-
ines) could also occur in the presence of saturated side
chains.
• in contrast, an equilibrium, also dependent on the
nature of R, occurs[6] between 3-(o-hydroxyphenyl)-1,2,4-
oxadiazoles (5) and 3-acylaminobenzisoxazoles (6): this is
because 5 is favoured by the diaryloid effect[3a] and by the
strong hydroxyϪaryl interaction (especially important in
the relevant anion), while the resonance within the amido
group effectively contributes to the stability of 6, in which
on the other hand, the resonance of isoxazole (Bird’s I5 in-
dex: 47) is strongly reduced because of the benzocondens-
ation;
• 3-acetylamino-5-methylisoxazole (7) could not be con-
verted[7] into 3-acetonyl-5-methyl-1,2,4-oxadiazole (8), as
was to be expected on the grounds of the higher aromaticity
of 7, the stability of which is furthermore increased by the
amide resonance.
Scheme 1
In some instances the classic monocyclic rearrangement
of heterocycles is reversible,[1,3] as may occur when, for ex-
ample, an OϪN bond is both cleaved and formed (i.e., in 1
and 2: Z ϭ D ϭ O). Of course, the thermodynamic require-
[a]
`
Dipartimento di Chimica Organica ‘‘E. Paterno’’
Viale delle Scienze Parco D’Orleans 2, 90128 Palermo, Italy
Dipartimento di Chimica Organica ‘‘A. Mangini’’
Via S. Donato 15, 40127 Bologna, Italy
Fax: (internat.) ϩ 39-051/244064
[b]
Of special interest are cases in which starting and final
ring are identical: a ring-degenerate interconversion then
E-mail: spinelli@alma.unibo.it
Eur. J. Org. Chem. 2002, 1417Ϫ1423
WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002 1434Ϫ193X/02/0408Ϫ1417 $ 20.00ϩ.50/0
1417

S. Buscemi, V. Frenna, A. Pace, N. Vivona, B. Cosimelli, D. Spinelli
FULL PAPER
Scheme 3
Results and Discussion
Equilibration Reaction
Methanolic solutions of 10 or 11, either as such or con-
verted into their conjugate bases by addition of tBuOK,
were kept at 313 K until analytical (1H NMR) determina-
tions gave constant mixture compositions (Table 1). Signi-
ficantly enough, while 11 had been found to prevail com-
Scheme 2
pletely in the neutral equilibration of both meta- (KN
ϭ
3.6Ϫ10.1) and para-substituted (KN ϭ 5.3Ϫ13.3) derivat-
ives,[3d] the electron-withdrawing ortho substituents signific-
antly shifted the equilibrium composition towards 10,
which became the main component (KN Ͻ 1) for X ϭ o-
CF3 and o-NO2 (Table 1).
occurs and can in turn be reversible, as we pointed out
when reporting some examples in the 1,2,4-oxadiazole
series.[3aϪ3e]
We studied the fully degenerate rearrangement of the an-
ion of 3-acetylamino-5-methyl-1,2,4-oxadiazole (9) in
DMSO[3b] and the quasi-fully degenerate rearrangement of
9 with a trideuterated acetyl group in the same solvent.[3c]
Semiempirical[3c] and ab initio[3e,8] calculations have been
also reported. The ring-degenerate equilibrium between
some meta- or para-substituted 3-aroylamino-5-methyl-
1,2,4-oxadiazoles (10) and the corresponding 3-acetylam-
ino-5-aryl-1,2,4-oxadiazoles (11) both in CD3OD and in
tBuOK/CD3OD (Scheme 3) has also been examined.[3d]
The measured equilibrium constants for the 10/11 couple
were found to be quite differently favoured by electron-re-
pelling substituents depending on whether neutral or an-
Table 1. Equilibrium constants and relevant compositions (%) for
the rearrangement 10aϪr v 11aϪr at 313.15 K
In CD3OD
In CD3OD/tBuOK
[10Ϫ]/[11Ϫ]
X
[10]/[11][a]
KN
KA
a: p-MeO
b: p-Me
c: H
7:93
8:92
9:91
13.3
11.5
10.1
22:78
30:70
38:62
67:33
3.55
2.33
1.63
0.493
d: p-Cl
11:89
14:86
15:85
16:84
11:89
9:91
16:84
17:83
22:78
42:58
7:93
8.09
e: p-F3C
f: p-NC
g: p-O2N
h: m-OMe
6.14
5.67
5.25
8.09
90:10
92:8
41:59
32:68
78:22
0.111
0.087
1.44
2.13
0.282
ionic forms were involved;[3d] the smaller substituent effect i: m-Me
10.1
j: m-Cl
5.25
4.88
3.55
1.38
in the first case (ρm ϭ Ϫ0.62 and ρp ϭ Ϫ0.37) in compar-
k: m-F3C
ison with that in the second one (ρm ϭ Ϫ1.95 and ρp ϭ
l: m-O2N
93:7
24:76
26:74
90:10
93:7
95:5
100:0
0.0753
3.17
2.85
0.11
0.075
0.053
Ϫ1.52) was convincingly interpreted on the grounds of the
m: o-MeO
electronic effects at play (e.g. internal carboxamide reson-
ance, arylϪcarbonyl resonance,[9] amidic nitrogen atom/
1,2,4-oxadiazol-3-yl interaction, diaryloid effect[3a]).
In this paper we extend the study of the equilibration of
the 10/11 couples to the case of ortho-substituted derivatives
(X ϭ H, MeO, Me, Cl, Br, CF3, NO2).
n: o-Me
o: o-Cl
p: o-Br
q: o-F3C
r: o-O2N
13.3
44:56
43:57
58:42
78:22
1.27
1.32
0.72
0.28
[a]
Data concerning 10aϪl/11aϪl couples from ref.[3d]
1418
Eur. J. Org. Chem. 2002, 1417Ϫ1423

Studies on Azole-to-Azole Interconversion
FULL PAPER
As far as the conjugate bases are concerned, effective
negative charge stabilisation is responsible for the great pre-
valence of 10؊ (KA Ͻ 1) observed[3d] in the presence of elec-
logKo,p/KH ϭ ρσo,p ϩ δEs ϩ fFo ϩ i
(2)
Of course, this treatment does not take account of hydro-
tron-withdrawing meta and para substituents (Table 1). gen bonding or of other intramolecular interactions in
Similarly to what has been observed under neutral condi- which an ortho substituent might engage, and which could
tions, electron-withdrawing ortho substituents here dictate therefore significantly affect the reactivity of ortho-substi-
a further shift of the equilibrium composition towards the tuted derivatives.
aroylamino system. Thus, practically, only 10؊ can be ob-
The basis of the use of multiparameter treatments is the
served at equilibrium for X ϭ NO2 and, accordingly, the statistical observation that the various parameters of inter-
relevant KA value has not been included in the free energy est show a low degree of interrelationship. Actually, a pur-
correlations below.
posely carried out check revealed that a significant correla-
tion (r2 ϭ 0.687) existed only between σ and the steric para-
meter Es; in the other instances lower r2 values (0.58 and
0.15) were calculated. As the steric parameter has been ex-
cluded from the correlation [see discussion and Equa-
tion (3) below], the multiparameter treatment can confid-
ently be used.
Free Energy Treatment of Equilibrium Data
A) A Hammett Equation Approach
We first attempted a simple Hammett treatment of our
data. Thus, in Ϫ for example Ϫ equilibration between neu-
tral species, we observed only a rough correlation (n ϭ 7,
r ϭ 0.7, ρo ϭ Ϫ1.2Ϯ0.5, i ϭ Ϫ0.5Ϯ0.2). A glance at the
plot as well as, on a more proper statistical basis, a calcula-
tion of residuals, in particular clearly showed the failure of
the ortho-methoxy group to line up with the other substitu-
ents; on exclusion of the relevant point the correlation im-
proved significantly (n ϭ 6, r ϭ 0.937, ρo ϭ Ϫ1.98Ϯ0.37,
i ϭ Ϫ0.15Ϯ0.15), exhibiting a susceptibility constant more
than one order of magnitude higher than that observed
with para substituents (compare ρp ϭ Ϫ0.37[3d] with ρo ϭ
Ϫ1.98). Such an extraordinary enhancement of the effect
of ortho substituents in comparison with para substituents
induced us to make use of a more specific approach that
would be able to account more properly for the very special
effects (proximity polar as well as steric) played by ortho
substituents on kinetic and thermodynamic data.
Thus, by application of Equations (1) and (2) to the in-
terconversions studied both in neutral and anionic form,
the results listed in Entries 2, 3, 8, and 9 of Table 2 were
obtained, displaying, at first glance, the existence of a small
and a large contribution from the ‘‘primary steric’’ and the
‘‘proximity polar’’ parameters, respectively.
Anyway, the correlation coefficients (R) were no more
than acceptable, and a calculation of residuals again showed
that the ortho-methoxy group did not line up with the other
substituents; thus, with the relevant point excluded, the cor-
relation coefficients improved from 0.967Ϫ0.986 to
0.995Ϫ0.997 (compare Entries 2, 3, 8, and 9 with Entries
4, 5, 10, and 11 in Table 2), clearly indicating the occurrence
of some special effect for this substituent (see below).
A close correspondence was observed between the ρ
values calculated for the para-substituted amides (10aϪg/
11aϪg couples; Entries 1 and 7 of Table 2) and for all ortho-
and para-substituted amides (10aϪg,nϪr/11aϪg,nϪr
couples: compare Entries 1 and 7 with Enries 5 and 11 of
Table 2), indicating that data for ortho and para derivatives
can confidently be treated in terms of a unique multipara-
meter free energy relationship. The values of the susceptibil-
ity constants indicated a large contribution from the ‘‘prox-
imity polar effect’’, especially in neutral solution [compare
B) An Approach by Use of the Fujita؊Nishioka Equation
Several approaches can be used to discuss the effect of
an ortho substituent on kinetic or thermodynamic data. In
an elaboration of Charton’s[10] primary hypothesis of dis-
secting the total ortho substituent effect into its inductive,
mesomeric and steric components, Fujita and Nishioka
suggested[11] the treatment of data as a sum of three contri-
butions [Equation (1)]: the ‘‘ordinary polar effect’’ (as-
sumed to be equivalent to that of a para substituent and
expressed by, for example, the σp constant), the ‘‘proximity
polar effect’’ (represented by the SwainϪLupton constant,
F)[12] and the ‘‘primary steric effect’’ (symbolised by the
Taft constant Es).[13]
(ρo,p)N ϭ Ϫ0.39 and (ρo,p)A ϭ Ϫ1.50 with (f)N ϭ Ϫ1.70 and
(f)A ϭ Ϫ2.08]. Thus, while electron-withdrawing groups
caused a shift towards 10 (or 10؊) from both the ortho and
the para positions, the effect in ortho-substituted derivatives
was evidently larger because of the occurrence of a strong
‘‘proximity polar effect’’ that cooperated with an ‘‘ordinary
polar effect’’ common to both sets of substituents.
As mentioned above, the ‘‘primary steric effect’’ does not
significantly affect the thermodynamic values either of the
neutral or of the anionic equilibration, as the calculated δ
values show uncertainties comparable to their absolute
values (0.07Ϯ0.04 and 0.06Ϯ0.04). Accordingly, fitting of
logKo/KH ϭ ρσo ϩ δEs ϩ fFo ϩ i
(1)
Interestingly, this treatment allows data to be combined data to Equation (3) similarly provides excellent correla-
for ortho and para substituents; meta substituents may also tions (Entries 6 and 12 of Table 2: R Ն 0.993), with only
be included only when meta and para derivatives fit the minor, if any, variations in the susceptibility constants for
same relationship. In the present instance, this condition is the polar effects [(ρo,p A
)
ϭ Ϫ0.38 and (ρo,p A
)
ϭ Ϫ1.50;
not met,[3d] and Equation (2) should be used.
(f)N ϭ Ϫ1.96 and (f)A ϭ Ϫ1.90].
Eur. J. Org. Chem. 2002, 1417Ϫ1423
1419

S. Buscemi, V. Frenna, A. Pace, N. Vivona, B. Cosimelli, D. Spinelli
Table 2. Free energy relationships for the equilibrium between 10aϪg,mϪr/11aϪg,mϪr couples at 313.15 K in methanol
FULL PAPER
Entry
Equation used[a]
ρ Ϯ sρ
δ Ϯ sδ
f Ϯ sf
i Ϯ si
R
n
Substituents
1
2
3
4
5
6
7
8
9
log (KN)X/(KN)H ϭ ρσp
(1)
(2)
(1)
(2)
(3)
Ϫ0.37Ϯ0.02
0.32Ϯ0.46
0.00Ϯ0.01
0.996
0.972
0.967
0.993
0.995
0.993
0.996
0.986
0.984
0.999
0.997
0.996
7
7
13
6
12
12
6
6
12
5
11
11
aϪg
0.13Ϯ0.16
0.03Ϯ0.09
0.06Ϯ0.10
0.07Ϯ0.04
Ϫ2.37Ϯ0.52
Ϫ2.10Ϯ0.40
Ϫ1.60Ϯ0.50
Ϫ1.70Ϯ0.10
Ϫ1.96Ϯ0.10
Ϫ0.07Ϯ0.20
Ϫ0.05Ϯ0.06
Ϫ0.01Ϯ0.10
0.01Ϯ0.03
c, mϪr
Ϫ0.19Ϯ0.13
Ϫ0.54Ϯ0.45
Ϫ0.39Ϯ0.05
Ϫ0.38Ϯ0.06
Ϫ1.52Ϯ0.07
Ϫ2.09Ϯ0.50
Ϫ1.64Ϯ0.13
Ϫ1.05Ϯ0.28
Ϫ1.50Ϯ0.07
Ϫ1.50Ϯ0.07
aϪg, mϪr
c, nϪr
aϪg, nϪr
aϪg, nϪr
aϪg
c, mϪq
aϪg, mϪq
c, nϪq
aϪg, nϪq
aϪg, nϪq
0.00Ϯ0.02
log (KA)X/(KA)H ϭ ρσp
Ϫ0.08Ϯ0.03
Ϫ0.07Ϯ0.21
Ϫ0.03Ϯ0.06
0.01Ϯ0.07
Ϫ0.09Ϯ0.03
Ϫ0.07Ϯ0.03
(1)
(2)
(1)
(2)
(3)
Ϫ0.07Ϯ0.20
Ϫ0.01Ϯ0.09
0.01Ϯ0.05
Ϫ1.62Ϯ0.59
Ϫ1.76Ϯ0.36
Ϫ2.40Ϯ0.20
Ϫ2.08Ϯ0.17
Ϫ1.90Ϯ0.10
10
11
12
Ϫ0.06Ϯ0.04
[a] sρ, sδ, sf, and si represent standard errors of ρ, δ, f, and i, respectively; i: intercept; R: correlation coefficient; n: number of data points.
The parameter correlations were calculated by using classical σp substituent constants (O. Exner, Correlation Analysis of Chemical Data,
Plenum Press, New York, London, 1988, pp. 61Ϫ62 and 143Ϫ144).
‘‘internal’’ conjugation of the amido group itself (structure
log Ko,p/KH ϭ ρσo,p ϩ fFo,p ϩ i
(3)
B), which is not significantly affected by electronic interac-
tions with the aryl moiety (see below).[9] In alkaline media,
on the other hand, stabilisation of 10Ϫ by the adjacent
phenyl ring is most likely to be mainly responsible for the
experimental results.
As already pointed out for meta- and para-substituted
derivatives,[3d] the overall effect of X on the studied equilib-
rations most probably results from a balance between coun-
teracting factors on 10 (or 10؊) and/or on 11 (or 11؊).
Thus, when rationalising the polar effect of, for example, an
electron-withdrawing substituent, one must essentially take
into account: i) a decrease in the electron density of the
nucleophilic oxygen atom of 10, ii) a lower weight of reson-
ance structures (Scheme 4) C and D of 10, and iii) a disfa-
vouring effect on the conjugation between the aryl and the
heteroaryl rings of 11, thus lowering the significance of the
stabilising ‘‘diaryloid’’ effect.[3a]
At first sight, the absence of steric effects may appear
surprising; as a matter of fact we have recently observed
(Scheme 5) that the kinetics of the rearrangement of ortho-
substituted (Z)-phenylhydrazones of 3-benzoyl-5-phenyl-
1,2,4-oxadiazole (12) into 2-aryl-4-benzoylamino-5-phenyl-
1,2,3-triazoles (13) (like the 10/11 equilibrium, an example
of the general class of monocyclic rearrangement of
heterocycles[1Ϫ3]) show[14] a significant contribution of the
‘‘primary steric effect’’ to the overall reactivity observed[15]
in dioxane/water (at pSϩ ϭ 3.80: ρo,m,p ϭ Ϫ1.30, f ϭ Ϫ0.90,
and δ ϭ 0.50; at pSϩ ϭ 11.50: ρo,m,p ϭ 2.23, f ϭ 0.31, and
δ ϭ 1.51 in the presence of electron-withdrawing substitu-
ents).
Scheme 4
Scheme 5
The qualitative similarity of the effects played by ortho
and para substituents, attested to by the excellent correla-
To understand why the reactivities of 12 and of the 10/
tions above, suggests that, as previously proposed for para- 11 couple should be so differently affected by the steric
substituted derivatives,[3d] the ‘‘diaryloid’’ effect governs the component of proximity effects, comparison between the
overall outcome in neutral media, justifying the observed side chains involved in the rearrangement in the two sys-
shift of the equilibrium towards 10 for electron-withdrawing tems appears relevant. The electronic distribution in the
substituents; this would essentially originate from com- ϾCϭNϪNαHϪAr side chain of 12 is essentially charac-
pensation between factors i and ii and would furthermore terised by the possibility of conjugative interactions be-
be strongly in agreement with the observation that struc- tween Nα and the aryl group, electron-withdrawing and -
tures such as C or D play only a minor role in the resonance repelling substituents in the aryl group decreasing or in-
system of 10, the main contribution being provided by the creasing, respectively, the electron density on Nα and hence
1420
Eur. J. Org. Chem. 2002, 1417Ϫ1423

Studies on Azole-to-Azole Interconversion
FULL PAPER
on the proton bound to Nα (i.e., the two factors that influ- hydrogen-bonded structures were not possible (10b and 10n,
ence the reactivity in the pSϩ-independent and -dependent and also 10e and 10q) strengthened this interpretation: the
regions). Of course, the occurrence of through-conjugation NϪH signals were always observed in the δ ϭ 7.8Ϫ8.7
is subject to precise steric requirements that can hardly be range in CDCl3 and in the δ ϭ 11.4Ϫ11.8 range in DMSO
satisfied if an ortho substituent is present on the aromatic for both ortho and para isomers, respectively.
ring.
In contrast, the ϪNHϪCαOϪAr [or ϪNϭC(OH)Ar]
side chain in 10 is essentially characterized, as already men-
1
Table 3. H NMR NϪH chemical shifts of compounds
tioned, by the strong resonance within the amido function,
Compound
δ (CDCl3)
δ ([D6]DMSO)
which is rather insensitive to the electronic effects of the
aryl group. Consequently, as 13C NMR spectroscopic data
of para-substituted benzamides and 2,6-dimethylbenzam-
ides have shown,[9] a low contribution of through-conjuga-
tion between Cα and the adjacent aryl moiety is to be ex-
pected. Their interaction is, rather, dominated by a ‘‘re-
verse’’ polar effect[16] that has been shown not to be signific-
antly affected by 2,6-dimethyl substitution (i.e., by steric
effects).
10a
10b
10e
10m
10n
10q
8.57
8.63
8.48
10.26
7.81
8.71
11.28
11.35
11.77
10.94
11.41
11.77
The situation observed is not new, and is reminiscent of
the findings of Wooldridge et al.,[19] who reported on the
occurrence of strong intramolecular hydrogen bonding be-
tween an ortho-alkoxy substituent and the N(1)ϪH proton
of a series of 2-aryl-8-azopurin-6-ones. As a matter of fact,
the proton of a pyrimidone ring amide group gives an intra-
molecular hydrogen bond with the oxygen atom of the al-
koxy group. The occurrence of the intramolecular hydrogen
bonding was evaluated by IR spectroscopy. The observed
∆ν [that is, the difference in cmϪ1 between the N(1)ϪH
stretching frequency observed for ortho-alkoxy derivatives
and that of the unsubstituted compound] gives a measure
of the strength of the bond formed. Interestingly, the occur-
rence of intramolecular hydrogen bonding strongly in-
creased the pharmacological (antiallergic) activity of the
studied compounds.
The Origin of the Peculiar Behaviour of o-Methoxy
Derivatives 10m and 11m
For the 10m/11m couple, the experimentally determined
equilibrium constants (1.38 and 3.17 under neutral and ba-
sic conditions, respectively) differed significantly from those
(KN ϭ 4.0, KA ϭ 1.2) calculated from the two-parameter
Equation (3). Thus, with respect to expectations, the equi-
librium was shifted more towards 10m under neutral condi-
tions, but towards the acetylamino anion 11m؊ under ba-
sic conditions.
In neutral solution, overstabilisation of 10m could arise
from an intramolecular hydrogen-bonded structure as in-
dicated in 14; actually, a calculation of δ∆G° (∆G°exp
Ϫ
∆G°calc) gives a value (ca. 0.6 kcal·molϪ1) perfectly compat-
ible with such a structure.[17] This hypothesis found further
support in the 1H NMR spectra determined in different
solvents, as the NϪH signal of 10m was found at low fields
both in DMSO (δ ϭ 10.94) and in CDCl3 (δ ϭ 10.26). By
contrast, the NϪH signal in the para isomer 10a was found
at δ ϭ 11.28 in DMSO, but at δ ϭ 8.57 in CDCl3,[18] the
downfield shift in the H-bond-accepting DMSO indicating
the occurrence of hydrogen-bonding with solvent molecules.
In the ortho isomer, in contrast, only a relatively small vari-
ation in the NϪH signal could be observed depending on
the spectroscopic solvent used, and so we can conclude that
an (intramolecular) hydrogen-bonded structure is present in
both solvents (Figure 1).
Finally, as far as the situation in basic solution is con-
cerned, a possible explanation for the observed destabilis-
ation of 10mϪ with respect to 11m؊ can be found in the
electrostatic repulsion between the lone pairs on the oxygen
of the methoxy group and the charged heteroatoms (oxygen
as well as nitrogen) of the anionic amide system.
Conclusion
Study of the 10/11 ring-degenerate equilibration of some
ortho-substituted (X ϭ H, OMe, Me, Cl, Br, CF3, NO2)
derivatives has demonstrated a large contribution from
purely electronic factors in determining the position of
equilibrium; the contribution of the ‘‘proximity polar ef-
fect’’ appears much higher (Ͼ 90%) than that of the ‘‘ordin-
ary polar effect’’ in neutral solution, whilst in basic solution
the two effects balance almost exactly (45% and 55%, re-
spectively). The absence of the ‘‘primary steric effect’’ com-
ponent in this kind of ringϪring interconversion has been
interpreted on the basis of the ‘‘strong’’ internal conjuga-
tion of the amido group, which leaves little role for the con-
jugation between the aryl group and Nα. Interesting prox-
Figure 1. Intramolecular hydrogen bond for compound 14
Comparison with 1H NMR spectra (Table 3) of other imity effects have been observed in the case of the ortho-
pairs of ortho and para isomers for which intramolecular methoxy derivative.
Eur. J. Org. Chem. 2002, 1417Ϫ1423
1421

S. Buscemi, V. Frenna, A. Pace, N. Vivona, B. Cosimelli, D. Spinelli
Table 4. Physical data for the acylamino-1,2,4-oxadiazoles 10mϪr and 11mϪr
FULL PAPER
Compd. M.p. [°C][a][b] IR [cmϪ1
]
1H NMR, δ ([D6]DMSO)
HRMS
Elemental analyses
10m
10n
10o
10p
116Ϫ118 (A) 3250, 3200, 3100, 1665 2.60 (s, 3 H), 3.95 (s, 3 H), 7.10Ϫ7.72 (m, 4 H), found 233.08029
10.94 (s, 1 H) calcd. 233.08004
156Ϫ158 (A) 3240, 3210, 3080, 1690 2.39 (s, 3 H), 2.57 (s, 3 H), 7.27Ϫ8.02 (m, 4 H), found 217.08531
found C 56.72, H 4.65, N 17.98
calcd. C 56.65, H 4.74, N 18.02
found C 60.58, H 5.12, N 19.20
calcd. C 60.82, H 5.10, N 19.34
found C 50.39, H 3.26, N 17.55
calcd. C 50.54, H 3.39, N 17.68
found C 42.39, H 2.71, N 14.81
calcd. C 42.58, H 2.86, N 14.90
11.41 (s, 1 H)
calcd. 217.08513
152Ϫ153 (A) 3240, 3190, 3100, 1665 2.62 (s, 3 H), 7.47-8.13 (m, 4 H), 11.69 (s, 1 H)
160Ϫ162 (A) 3230, 3190, 3100, 1700 2.62 (s, 3 H), 7.45Ϫ8.06 (m, 4 H), 11.70 (s, 1 H)
10q
10r
138 (A)[c]
145 (A)[c]
11m
124 (B)
3240, 3150, 1690
2.17 (s, 3 H), 3.97 (s, 3 H), 7.15Ϫ8.00 (m, 4 H), found 233.08039
11.30 (s, 1 H) calcd. 233.08004
3230, 3200, 3100, 1690 2.19 (s, 3 H), 2.69 (s, 3 H), 7.46Ϫ8.06 (m, 4 H), found 217.08542
11.27 (s, 1 H) calcd. 217.08513
found C 56.81, H 4.67, N 17.88
calcd. C 56.65, H 4.74, N 18.02
found C 60.63, H 5.16, N 19.27
calcd. C 60.82, H 5.10, N 19.34
11n
11o
11p
165 (B)
130Ϫ132 (B) 3210, 3160, 3100, 1690 2.20 (s, 3 H), 7.26Ϫ8.14 (m, 4 H), 11.38 (s, 1 H) found 237.03091[d] found C 50.42, H 3.20, N 17.52
calcd. 237.03050 calcd. C 50.54, H 3.39, N 17.68
143Ϫ145 (B) 3230, 3180, 3100, 1670 2.20 (s, 3 H), 7.45Ϫ8.06 (m, 4 H), 11.38 (s, 1 H) found 280.98032[e] found C 42.64, H 2.69, N 14.76
calcd. 280.97999
calcd. C 42.58, H 2.86, N 14.90
11q
11r
124 (A)[c]
163 (C)[c]
[a]
[b]
Melting points can be affected by a thermally induced rearrangement. All new compounds are colourless. Crystallisation solvents:
A: benzene; B: ethanol; C: ethyl acetate. [c] Ref.[21]: 10q: m.p. 138 °C; 10r: m.p. 145 °C; 11q: m.p. 124 °C; 11r: m.p. 163 °C. [d] 35Cl isotope.
[e] 79Br isotope.
In CD3OD in the Presence of tBuOK: Samples of 3-aroylamino-
Experimental Section
5-methyl-1,2,4-oxadiazoles 10 (1.8·10Ϫ2 mmol) in CD3OD (1 mL)
containing tBuOK (3.6·10Ϫ2 mmol)[22] reached equilibrium after
General Remarks: Melting points were determined with a Kofler
hot-stage apparatus and are uncorrected. IR spectra (Nujol) were
standing for 30 min and were then analysed by NMR. Representat-
ively, the equilibration reaction was also carried out starting from
determined with a PerkinϪElmer 257 instrument, 1H NMR spectra
3-acetylamino-5-aryl-1,2,4-oxadiazoles (i.e., 11mϪn), the same
were recorded with a Bruker 250 E spectrometer (tetramethylsilane
equilibrium compositions being observed.
as internal standard). Analytical determinations were carried out
by 1H NMR, integrating the methyl singlets in the ranges δ ϭ
Acknowledgments
2.39Ϫ2.62 (characteristic of 5-methyl of 10) and 2.17Ϫ2.20 (char-
acteristic of the 3-acetylamino group of 11), respectively. Composi-
We thank the CNR, the MURST and the University of Bologna
tions at equilibrium (expressed in % of the two isomers) represent
for financial support.
average values (with an uncertainty lower than Ϯ2%) of at least
three independent determinations. Compounds 10 were prepared
by treatment of 3-amino-5-methyl-1,2,4-oxadiazole[20,21] with the
appropriate aroyl chloride by the procedure previously descri-
bed.[3a,3d,7] Compounds 11 were prepared by isoheterocyclic re-
arrangement of the corresponding 10 by refluxing for several hours
in ethanol. After removal of the solvent, the mixtures of 10 and 11
were separated by chromatography. Compounds 11q and 11r were
prepared by acetylation of 3-amino-5-aryl-1,2,4-oxadiazoles with
acetyl chloride in pyridine.[21] All new compounds gave satisfactory
analytical data (C, H, N). Significant physical data are collected
in Table 4.
A. J. Boulton, A. R. Katritzky, A. Majid-Hamid, J. Chem.
[1] [1a]
[1b]
Soc. C 1967, 2005Ϫ2007.
A. S. Afridi, A. R. Katritzky, C.
A. Ramsden, J. Chem. Soc., Perkin Trans. 1 1976, 315Ϫ320.
[1c]
A. J. Boulton, Lectures in Heterocyclic Chemistry, Hetero
Corporation, Provo, 1973. [1d]M. Ruccia, N. Vivona, D.
Spinelli, Adv. Heterocycl. Chem. 1981, 29, 141Ϫ169.
vona, S. Buscemi, V. Frenna, G. Cusmano, Adv. Heterocycl.
[1e]
N. Vi-
[1f]
Chem. 1993, 56, 49Ϫ154.
B. Cosimelli, S. Guernelli, D.
Spinelli, S. Buscemi, V. Frenna, G. Macaluso, J. Org. Chem.
2001, 66, 6124Ϫ6129.
[2] [2a]
´
D. Korbonits, I. Kanzel-Szvoboda, K. Horvath, J. Chem.
[2b]
´
K. Horvath, D. Kor-
Soc., Perkin Trans. 1 1982, 759Ϫ766.
Equilibration between 3-Aroylamino-5-methyl-1,2,4-oxadiazoles 10
and 3-Acetylamino-5-aryl-1,2,4-oxadiazoles 11
´
`
bonits, G. Naray-Szabo, K. Simon, J. Mol. Struct. (Theochem)
1986, 136, 215Ϫ227.
[3] [3a]
N. Vivona, G. Cusmano, M. Ruccia, D. Spinelli, J. Hetero-
In CD3OD: Samples of 3-aroylamino-5-methyl-1,2,4-oxadiazoles
10 (1.8·10Ϫ2 mmol) in CD3OD (1 mL) were maintained in NMR
tubes at 313.15 K until constant mixture composition. Equilibrium
was reached in 3Ϫ5 weeks, except for the 10m/11m couple, which
required 8 weeks. Representatively, the equilibration reaction was
also carried out starting from 3-acetylamino-5-aryl-1,2,4-oxadia-
zoles (i.e., 11mϪn), with the same equilibrium compositions being
observed.
[3b]
cycl. Chem. 1975, 12, 985Ϫ988.
N. Vivona, M. Ruccia, G.
Cusmano, M. L. Marino, D. Spinelli, J. Heterocycl. Chem.
1975, 12, 1327Ϫ1328. [3c] G. La Manna, S. Buscemi, V. Frenna,
[3d]
N. Vivona, D. Spinelli, Heterocycles 1991, 32, 1547Ϫ1557.
S. Buscemi, V. Frenna, N. Vivona, G. Petrillo, D. Spinelli, Tet-
rahedron 1995, 51, 5133Ϫ5142.
N. Vivona, J. Mol. Struct. (Theochem) 1998, 452, 67Ϫ74.
H. C. van der Plas, Adv. Heterocycl. Chem. 1999, 74, 153Ϫ240.
[3e]
G. La Manna, S. Buscemi,
[3f]
1422
Eur. J. Org. Chem. 2002, 1417Ϫ1423

Studies on Azole-to-Azole Interconversion
FULL PAPER
[4]
[4a] N. Vivona, V. Frenna, S. Buscemi, M. Ruccia, J. Heterocycl.
ous studies on proximity effects in ortho-substituted anilines
[anilinodebromination of 2-bromo-3,5-dinitrothiophene in
methanol (δ ϭ 1.31) and dissociation of anilinium ions in water
(δ ϭ 0.56)] D. Spinelli, G. Consiglio, R. Noto, C. Arnone, J.
Chem. Soc., Perkin Trans. 2 1979, 219Ϫ221.
It may be remarked that the importance of polar effects ob-
served in NMR spectroscopic data of benzamides agrees with
the observation that in the 10 v 11 equilibrations, both in neut-
ral and anionic forms, meta substituents affect the reaction
more than para substituents and that ortho substituents give
rise to a high ‘‘proximity polar effect’’ contribution.
[4b]
Chem. 1985, 22, 97Ϫ99.
N. Vivona, S. Buscemi, V. Frenna,
`
M. Ruccia, M. Condo, J. Chem. Research (M) 1985,
2184Ϫ2197; (S) 1985, 190. [4c] V. G. Andrianov, A. V. Eremeev,
Chem. Heterocycl. Compd. (Engl. Transl.) 1990, 26,
1199Ϫ1213.
[16]
[5]
[6]
[7]
C. W. Bird, Tetrahedron 1985, 41, 1409Ϫ1414.
K. Harsani, J. Heterocycl. Chem. 1973, 10, 957Ϫ961.
S. Buscemi, V. Frenna, N. Vivona, Heterocycles 1991, 32,
1765Ϫ1772.
[8]
[9]
V. G. Andrianov, S. V. Makushenkov, A. V. Eremeev, Mendeleev
Commun. 1992, 129Ϫ130.
[17]
[18]
G. A. Jeffrey, An Introduction to Hydrogen Bond, Oxford Uni-
versity Press, New York, Oxford, 1997.
As 13C NMR spectroscopic data have also shown, the carba-
moyl carbon chemical shift in para-substituted benzamides is
not significantly affected by through-resonance with the sub-
stituent, π-polarization appearing to be the prevalent resonance
effect; see, for example: C. Dell’Erba, A. Mele, M. Novi, G.
Petrillo, F. Sancassan, D. Spinelli, J. Chem. Soc., Perkin Trans.
2 1990, 2055Ϫ2058 and references therein.
The influence of the nature of the solvent as well as of the
occurrence of intramolecular or intermolecular hydrogen
bonding (i.e., with the solvent molecules) on the proton NMR
chemical shifts has already been investigated in depth (e.g., see:
G. P. Dado, S. H. Gelleman, J. Am. Chem. Soc. 1993, 115,
4228Ϫ4245, and references therein).
[10]
[11]
[19] [19a]
M. Charton, Progr. Phys. Org. Chem. 1971, 8, 235Ϫ317.
T. Fujita, T. Nishioka, Progr. Phys. Org. Chem. 1976, 12,
49Ϫ89.
C. G. Swain, E. C. Lupton, J. Am. Chem. Soc. 1968, 90,
4328Ϫ4337.
R. W. Taft, Jr., Steric Effects in Organic Chemistry (Ed.: M.
S. Newman), Wiley, New York, 1956, p. 556. The reference
substituent has been changed to hydrogen according to Fujita
and Nishioka.[11]
B. J. Broughton, P. Chaplen, P. Knowles, E. Lunt, S. M.
Marshall, D. L. Pain, K. R. H. Wooldridge, J. Med. Chem.
1975, 18, 1117Ϫ1122. [19b] A. Holland, D. Jackson, P. Chaplen,
E. Lunt, S. Marshall, D. Pain, K. Woolfridge, Eur. J. Med.
Chem. 1975, 10, 447Ϫ449.
G. W. Anderson, H. E. Faith, H. W. Marson, P. S. Winneck,
R. O. Roblin, Jr., J. Am. Chem. Soc. 1942, 64, 2902Ϫ2905.
S. Buscemi, A. Pace, V. Frenna, N. Vivona, work in progress.
Experiments previously carried out have shown that complete
salt formation was reached at around 1.5 tBuOK/10.
Received November 14, 2001
[12]
[13]
[20]
[21]
[22]
[14]
[15]
V. Frenna, G. Macaluso, G. Consiglio, B. Cosimelli, D. Spinelli,
Tetrahedron 1999, 55, 12885Ϫ12896.
We have always found significant steric contributions in previ-
[O01549]
Eur. J. Org. Chem. 2002, 1417Ϫ1423
1423