Experimental and DFT studies on competitive heterocyclic rearrangements. 3. A cascade isoxazole-1,2,4-oxadiazole-oxazole rearrangement

Article abstract of DOI:10.1021/jo802081k

(Figure Presented) The thermal rearrangements of 3-acylamino-5- methylisoxazoles 1 have been investigated under basic and neutral conditions and interpreted with the support of computational data. The density functional theory (DFT) study on the competitive routes available for the base-catalyzed thermal rearrangement of isoxazoles 1 showed that the Boulton-Katritzky (BK) rearrangement, producing the less stable 3-acetonyl-1,2,4-oxadiazoles 5, is a much more favored process than either the migration-nucleophilic attack-cyclization (MNAC) or the ring contraction-ring expansion (RCRE). In turn, an increase in reaction temperature will promote the MNAC of oxadiazoles 5, producing the more stable 2-acylaminooxazoles 8. The thermal rearrangement of 3-acylaminoisoxazoles 1 into oxazoles 8 can therefore be interpreted in terms of a cascade BK-MNAC rearrangement involving 3-acetonyl-1,2,4-oxadiazoles 5 as ancillary intermediates.

Full text of DOI:10.1021/jo802081k

Experimental and DFT Studies on Competitive Heterocyclic
Rearrangements. 3.† A Cascade
Isoxazole-1,2,4-Oxadiazole-Oxazole Rearrangement‡
Andrea Pace,*,§ Paola Pierro,§ Silvestre Buscemi,§ Nicolo` Vivona,§ and
Giampaolo Barone*,|
Dipartimento di Chimica Organica “E. Paterno`” and Dipartimento di Chimica Inorganica e Analitica
“S. Cannizzaro”, UniVersita` degli Studi di Palermo, Viale delle Scienze, Parco d’Orleans II, Edificio 17,
I-90128 Palermo, Italy
pace@unipa.it; gbarone@unipa.it
ReceiVed September 19, 2008
The thermal rearrangements of 3-acylamino-5-methylisoxazoles 1 have been investigated under basic
and neutral conditions and interpreted with the support of computational data. The density functional
theory (DFT) study on the competitive routes available for the base-catalyzed thermal rearrangement of
isoxazoles 1 showed that the Boulton-Katritzky (BK) rearrangement, producing the less stable 3-acetonyl-
1,2,4-oxadiazoles 5, is a much more favored process than either the migration-nucleophilic attack-cyclization
(MNAC) or the ring contraction-ring expansion (RCRE). In turn, an increase in reaction temperature
will promote the MNAC of oxadiazoles 5, producing the more stable 2-acylaminooxazoles 8. The thermal
rearrangement of 3-acylaminoisoxazoles 1 into oxazoles 8 can therefore be interpreted in terms of a
cascade BK-MNAC rearrangement involving 3-acetonyl-1,2,4-oxadiazoles 5 as ancillary intermediates.
Introduction
isolation of intermediates was difficult, mechanisms have been
proposed on the basis of the experimental conditions used and
of the structure of the final products or by comparison with
proven mechanisms for similar reactions.
Heterocycles are an important class of organic compounds
largely represented in nature and in daily life applications.1 The
knowledge of their properties and chemical behavior is crucial
for understanding their functions in biological systems and
predicting their features in new materials. Often, heterocyclic
compounds are able to undergo chemical transformation into
other, more stable, heterocycles. These ring-rearrangements can
be either thermally or photochemically induced, and their
mechanistic interpretation often represents a challenging research
area. Indeed, heterocyclic ring rearrangements are widely
documented in the literature2 and could represent useful
synthetic routes to target heterocycles. In the past, when the
Nowadays, with the development of continuously more
affordable and reliable quantum chemical methods,3 heterocyclic
rearrangements can be described on the basis of theoretical
calculations.4 For example, we recently used combined experi-
mental and computational data to rationalize the mechanisms
(2) See, for example: (a) Van der Plas, H. C. Ring Transformation of
Heterocycles, Academic Press: New York, 1973; Vol. 1. (b) Van der Plas, H. C.
Ring Transformation of Heterocycles; Academic Press: New York, 1973; Vol.
2. (c) L’abbe´, G. J. Heterocycl. Chem. 1984, 21, 627–638. (d) Van der Plas,
H. C. AdV. Heterocycl. Chem. 1999, 74, 1–253. See also specific classes of ring
transformations reviewed in: (e) ComprehensiVe Heterocyclic Chemistry;
Katritzky, A. R., Rees, C. W., Eds.; Pergamon Press: Oxford, 1984; Vols. 1-8.
(f) ComprehensiVe Heterocyclic Chemistry II; Katritzky, A. R., Rees, C. W.,
Scriven, E. F. V., Eds.; Elsevier: Amsterdam, 1996; Vols. 1-9.
† For parts 1 and 2, see refs5 and,6 respectively.
‡ In memory of Professor Giuseppe Cusmano.
§ Dipartimento di Chimica Organica “E. Paterno`”.
| Dipartimento di Chimica Inorganica e Analitica “S. Cannizzaro”.
(1) Pozharskii, A. F.; Soldatenkov, A. T.; Katritzky, A. R. Heterocycles in
Life and Society; Wiley: Chichester, UK, 1997.
(3) Cramer, C. J. Essentials of Computational Chemistry, 2nd ed.; Wiley:
New York, 2004.
10.1021/jo802081k CCC: $40.75
Published on Web 12/02/2008
2009 American Chemical Society
J. Org. Chem. 2009, 74, 351–358 351

Pace et al.
SCHEME 1. Reaction Routes Available to the Highly
Energetic Anion of
SCHEME 2. BK, MNAC, and RCRE Routes of the Anionic
Forms of 3-Acylamino-1,2,4-oxadiazoles
3-Amino(alkylamino)-5-perfluoroalkyl-1,2,4-oxadiazole7
of some photochemical5 and thermal6 rearrangements involving
highly energetic or highly reactive nonisolable intermediates.
In particular, we pointed out that in the photoinduced
rearrangements of some 3-amino- and 3-N-alkylamino-5-per-
fluoroalkyl-1,2,4-oxadiazoles, different routes were accessible.
In that case, the deprotonation of the neutral excited state of
the substrate (about 30 orders of magnitude more acidic than
the ground state)5 could lead to highly energetic excited-state
or “hot” ground-state anionic forms. The postphotoexcitation
thermochemical development of such intermediates followed a
downhill pathway where the internal cyclization-isomerization
(ICI), the ring contraction-ring expansion (RCRE), and the
migration-nucleophilic attack-cyclization (MNAC) routes
were involved in product formation (Scheme 1).5,7
Later on, the latter two routes were considered, together with
the well-known Boulton-Katritzky (BK) rearrangement,8 in a
combined experimental and theoretical study on the thermal
reactivity of the anionic forms of 3-acylamino-1,2,4-oxadiazoles.
It was found that the RCRE, the MNAC, and the BK pathways
had well-differentiated energetic profiles which allowed thermal
control of the observed heterocyclic rearrangement. At temper-
atures lower than 80 °C, the reversible BK equilibrium was the
only active process, while at higher temperatures, the irreversible
MNAC rearrangement into the more stable 1,3,4-oxadiazoles
was observed. Both these rearrangements involved the N(2) ring-
nitrogen as a pivotal electrophilic center,8 a typical feature of
O-N bond-containing azoles.9 The internal nucleophilic center
was represented by either the oxygen atom of the side chain,
for the BK rearrangement, or the exocyclic nitrogen atom linked
at the C(3) of the oxadiazole ring, for the migration step of the
MNAC route. The last two processes of the MNAC rearrange-
ment are then completed in the same step through an intramo-
CHART 1
lecular nucleophilic attack-cyclization involving the oxygen
atom of the original side chain (Scheme 2).6,7
The higher energy demand of the MNAC with respect to the
BK can be mainly ascribed to the formation of a ring-strained
three-membered cyclic intermediate (diazirine-exo in Scheme
2). In this framework, the RCRE route, which also requires the
involvement of a three-membered cyclic intermediate (diazirine-
endo in Scheme 2), is the least favorite one since the cleavage
of the O-N(2) ring bond is not assisted by a simultaneous
nucleophilic attack on the N(2), such as in the case of BK and
MNAC.
Considering the importance of understanding heterocyclic
reactivity and of our historical interest in O-N bond-containing
azoles, we decided to verify the competition between BK,
RCRE, and MNAC rearrangements by studying the thermal
reactivity of 3-acylaminoisoxazoles 1a-d (Chart 1). The
substitution of a ring nitrogen atom with a ring C-H group,
with respect to previously studied 3-acylamino-1,2,4-oxadiaz-
oles, might in fact cause a significant change of the thermody-
namics and kinetics involved along the reaction coordinate.
We proceeded by preliminarily performing an extensive DFT
study on the three possible reaction pathways and then by
opportunely planning and performing the experiments required
to verify the proposed reaction mechanism.
(4) For DFT studies on monocyclic BK rearrangements, see: (a) Bottoni,
A.; Frenna, V.; Lanza, C. Z.; Macaluso, G.; Spinelli, D. J. Phys. Chem. A 2004,
108, 1731–1740. Moreover, for DFT studies on bicyclic BK rearrangements,
see, for example: (b) Eckert, F.; Rauhut, G. J. Am. Chem. Soc. 1998, 120, 13478–
13484. (c) Rauhut, G. J. Org. Chem. 2001, 66, 5444–5448. (d) Pen˜a-Gallego,
A.; Rodríguez-Otero, J.; Cabaleiro-Lago, E. M. J. Org. Chem. 2004, 69, 7013–
7017.
(5) Pace, A.; Buscemi, S.; Vivona, N.; Silvestri, A.; Barone, G. J. Org. Chem.
2006, 71, 2740–2749.
(6) Pace, A.; Pibiri, I.; Palumbo Piccionello, A.; Buscemi, S.; Vivona, N.;
Barone, G. J. Org. Chem. 2007, 72, 7656–7666.
(7) Since in the RCRE route the endocyclic N(4) ring nitrogen is involved
in the formation of the diazirine, we named such intermediate as diazirine-endo.
For the same reason, we named the first MNAC intermediate as diazirine-exo,
since the exocyclic side-chain nitrogen is involved in its formation.
(8) (a) Boulton, A. J.; Katritzky, A. R.; Hamid, A. M. J. Chem. Soc. C 1967,
2005–2007. (b) Afridi, A. S.; Katritzky, A. R.; Ramsden, C. A. J. Chem. Soc.,
Perkin Trans. 1 1976, 315–320.
Results and Discussion
3-Acylaminoisoxazoles 1a-d can, in theory, undergo the
same kind of rearrangements observed for 3-acylamino-1,2,4-
(9) (a) Pace, A.; Pibiri, I.; Buscemi, S.; Vivona, N. Heterocycles 2004, 63
(11), 2627–2648. (b) Ruccia, M.; Vivona, N.; Spinelli, D. AdV. Heterocycl. Chem.
1981, 29, 141–169. (c) Vivona, N.; Buscemi, S.; Frenna, V.; Cusmano, G. AdV.
Heterocycl. Chem. 1993, 56, 49–154. (d) Korbonits, D.; Kanzel-Szvoboda, I.;
Horva´th, K. J. Chem. Soc., Perkin Trans. 1 1982, 759–766. (e) Horva´th, K.;
Korbonits, D.; Nara´y-Szabo`, G.; Simon, K. THEOCHEM 1986, 136, 215–227.
(f) Boulton, A. J.; Frank, F.; Huckstep, M. R. Gazz. Chim. Ital. 1982, 112, 181–
183. (g) Sheremetev, A. B.; Makhova, N. N.; Friedrichsen, W. AdV. Heterocycl.
Chem. 2001, 78, 66–188. (h) Makhova, N. N.; Ovchinnikov, I. V.; Kulikov,
A. S.; Molotov, S. I.; Baryshnikova, E. L. Pure Appl. Chem. 2004, 76, 1691–
1703.
352 J. Org. Chem. Vol. 74, No. 1, 2009

1,2,4-Oxadiazole-Mediated Isoxazole-to-Oxazole Rearrangement
SCHEME 3. BK, MNAC, and RCRE Routes Available to
Representative 3-Benzoylamino-5-methylisoxazole 1a in Its
Deprotonated Anionic Form 1a-
to the stabilizing hydrogen bond between the amide moiety and
the C(4)-H of the heterocycle. Nevertheless, the BK route
requires an initial rotation of the C(3)-N bond with formation
of the less stable conformer 1a′- which is an intermediate toward
the formation of 1,2,4-oxadiazole 5a-. Similarly, calculations
showed how the RCRE route of 1a- would lead to the formation
of the less stable azirine conformer 6′-, accessible also from
azirine 6-, precursor of carbodiimide 7-. Furthermore, the
development of 7-produces the oxazole 8a′- which will
eventually give 8a-, a more stable conformer in methanol.
Computational results show undoubtedly that the previously
claimed RCRE route,10 due to its high activation barrier (∆Gq
> 180 kJ/mol in all three media), is not a plausible mechanism
for the thermal transformation of isoxazole 1a into oxazole 8a;
therefore, this rearrangement should be interpreted in terms of
a different mechanistic pathway. Another alternative route
toward oxazole 8a (see Scheme 3) consists of a MNAC of
isoxazole 1a- followed by a one-atom side-chain rearrangement
of diazirine 2- leading to 6- and eventually to 8a-. However,
if diazirine 2- was formed, its development into 3- and then
4-, with an activation barrier of only ∆Gq ) 38.3∼48.3 kJ/
mol, would have been favored over the transformation into
azirine 6- with a ∆Gq ) 64.9∼75.7 kJ/mol.
Moreover, an accurate analysis of data reported in Table 1
reveals a significant change of the activation barrier for the
MNAC rearrangement of deprotonated isoxazole 1a- with
respect to previously studied 3-acylamino-1,2,4-oxadiazoles.6,13
For the general system illustrated in Chart 2, a change in the
ring-atom A from N to CH (with B ) N) causes a stabilization
of the substrate with a consequent increase of the activation
barrier of MNAC from 111.76 to 142.8 (in vacuo) and from
132.56 to 152.7 (in DMSO). This renders the MNAC route less
accessible to 3-acylaminoisoxazole 1a than it is for 3-acylamino-
1,2,4-oxadiazoles.
On the other hand, a change in the side-chain atom B from
N to CH (with A ) N) increases the reactivity of the
nucleophilic site and lowers the activation barrier, with respect
to 3-acylamino-1,2,4-oxadiazoles, from 103.66 to 72.3 (in vacuo)
and from 132.16 to 95.1 (in DMSO). Therefore, the MNAC route
of oxadiazole 5a- is much more favored than either its RCRE
rearrangement or the MNAC of the isoxazole 1a-.
Overall, theoretical data suggest the intriguing hypothesis that
the formation of oxazole 8a originates from a cascade
BK-MNAC rearrangement of the substrate 1a. In other words,
deprotonated isoxazole 1a- will initially rearrange (via BK) into
1,2,4-oxadiazole 5a- which, in turn, rearranges (via MNAC)
into azirine 6- which will eventually produce the final oxazole
8a-.
To obtain experimental support, the thermal rearrangements
of 1a was performed by heating the substrate in DMF and in
the presence of an equimolar amount of t-BuOK, and the
reaction mixtures were monitored by TLC using authentic
synthetic samples of 4,14 5a (see the Experimental Section),
and 8a10 as reference compounds. After 1 h at 110 °C, substrate
1a was partly converted into oxazole 8a, and a repeatedly eluted
TLC15 revealed the presence of 1,2,4-oxadiazole 5a, never
observed in previous studies.10,16 Obviously, a similar product
oxadiazoles since they contain an electrophilic N(2) ring nitrogen
and two potentially nucleophilic centers in the side chain: the
exocyclic nitrogen and the carbonylic oxygen.
A previous report from our laboratories showed that some
3-aroylaminoisoxazoles, heated under basic conditions, rear-
ranged into the corresponding 2-aroylamino-oxazoles with no
evidence of the occurrence of other rearrangements.10 Such
reactivity, involving the anionic forms, was interpreted in terms
of a RCRE pathway, in analogy to the well-proven mechanism
proposed for the photochemical isoxazole-azirine-oxazole
transformation.11 However, based on our recent study showing
that the RCRE of 3-acylamino-1,2,4-oxadiazoles is the highest
energy demanding among the three possible routes,6 we were
prompted to reinvestigate the base-catalyzed rearrangement of
3-acylaminoisoxazoles 1. Therefore, we performed a DFT study
on the anionic forms of substrates, products, intermediates, and
transition states of all of the possible BK, MNAC, and RCRE
routes available to the representative compound 1a (Scheme
3). Moreover, to complete the mechanistic picture, the possibility
of a one-atom side-chain rearrangement between the three-
membered ring intermediates 2- and 6- has been also considered.
Structures and free energy values have been calculated in
vacuo in DMSO and MeOH to simulate the presence of polar
aprotic or protic solvents.12 Standard free energy values (∆G°),
relative to 1a-, of reagents, products and transition states, and
activation barriers (∆Gq) are reported in Table 1. Representative
data calculated in MeOH are also graphically illustrated in
Figure 1.
All possible conformers and tautomers have been considered
for each reagent, intermediate, or product; however, only the
most stable or particularly significant key structures will be
discussed in detail. For instance, the structure 1a- represents
the most stable conformer of the anionic isoxazole reagent, due
(10) Buscemi, S.; Frenna, V.; Vivona, N. Heterocycles 1991, 32, 1765–1772.
(11) (a) Ullman, E. F.; Singh, B. J. Org. Chem. 1966, 31, 1844–1845. (b)
Ullman, E. F.; Singh, B. J. Org. Chem. 1967, 32, 6911–6916. (c) Singh, B.;
Zweig, A.; Gallivan, J. B. J. Am. Chem. Soc. 1972, 94, 1199–1206.
(12) This choice also allows an easier comparison with previously calculated
data for similar reactions such as photoinduced MNAC and RCRE in MeOH
(see ref 5) and thermal BK, MNAC, and RCRE in DMSO (see ref 6).
(13) A complete comparison table between activation barriers of BK, MNAC,
and RCRE rearrangements of deprotonated isoxazole 1a-, oxadiazole 5a-, and
previously studied 3-acylamino-1,2,4-oxadiazoles is given in the Supporting
Information.
(14) Kato, T.; Chiba, T.; Daneshtalab, M. Chem. Pharm. Bull. 1976, 24,
2549–2552.
J. Org. Chem. Vol. 74, No. 1, 2009 353

Pace et al.
TABLE 1. Standard Free Energy Values (∆G°, kJ/mol) Relative to Compound 1a-,a for Anionic Species or Transition States, and Activation
Barriers (∆Gq, kJ/mol), Relative to the Given Starting Compound, Along the Reaction Coordinate, Calculated in Vacuo, in DMSO and in
MeOH at 298.15 K
deprotonated
species
∆G°
(in vacuo)
∆G°
(DMSO)
∆G°
(MeOH)
transition
states
∆G°
(in vacuo)
∆G°
(DMSO)
∆G°
(MeOH)
1a-
1a′-
2-
0.0
20.9
116.7
22.2
-21.8
31.0
0.0
10.0
127.2
48.4
-20.1
15.6
3.0
3.7
-23.3
-90.6
-89.8
0.0
3.2
115.2
39.6
[1a-f1a′-]q
28.2
142.8
181.6
96.4
155.5
181.6
48.8
142.8
103.3
14.2
18.7
152.7
202.9
100.0
173.2
202.9
59.6
152.7
110.7
24.7
11.6
140.4
190.6
89.3
163.5
190.6
53.4
140.4
99.5
12.9
b
[1a-f2-]q
c
[1a-f6′-]q
3-
[1a′-f5a-]q
[2-f3-]q
4-
-27.5
c
5a-
6-
7.1
[2-f6-]q
1.2
-14.5
-10.1
-33.0
-99.5
-104.4
[3-f4-]q
b
6′-
7-
-9.7
-48.3
-94.5
-91.1
[5a-f2-]q
[5a-f6-]q
[6-f6′-]q
[6′-f7-]q
[7-f8a′-]q
[8a′-f8a-]q
8a′-
8a-
51.4
-20.4
-80.7
62.8
-7.6
-78.8
50.9
-16.5
-90.0
reactions
∆G‡ (in vacuo)
∆G‡ (DMSO)
∆G‡ (MeOH)
1a-f1a′-
1a-f2-
1a-f6′-
1a′-f5a-
2-f3-
28.2
142.8
181.6
75.5
38.3
64.9
26.6
111.8
72.3
13.0
61.1
27.9
10.4
18.7
152.7
202.9
90.0
46.0
75.7
11.2
137.1
95.1
21.7
59.1
15.7
11.0
11.6
140.4
190.6
86.1
48.3
75.4
13.8
133.3
92.4
27.4
61.0
16.5
14.4
2-f6-
3-f4-
5a-f2-
5a-f6-
6-f6′-
6′-f7-
7-f8a′-
8a′-f8a-
a The standard free energy value calculated at B3LYP/6-31++G(d,p) level is -684.47037 au for 1a-; the solvation free energy values for 1a-are
-50.62 kcal/mol (in DMSO) and -59.85 kcal/mol (in MeOH) (see Computational Details in the Experimental Section). b The transformation of either
c
1a- or 5a- into 2- occurs through the same transition state. The transformations of 1a- into 6′- and of 2- into 6- occur through the same transition
state.
isoxazoles.17 No trace of 1,3,4-oxadiazole 4 was present in either
reaction mixture, even after prolonged heating (4 h in refluxing
DMF/t-BuOK) of the substrate, thus excluding both the MNAC
of isoxazole 1a and the RCRE of oxadiazole 5a as possible
routes.
Also compounds 1b-d yielded the corresponding oxazoles
8b-d in good yield,16 while traces of 1,2,4-oxadiazoles 5b and
5c were detected. On the other hand, during the rearrangement
of 1d into 8d, no traces of 5d, which is known to undergo a
complete BK transformation into isoxazole 1d,18 were detected
(Scheme 4).
Interestingly, when the reaction of 1a was performed at lower
temperature in refluxing methanol containing only 0.1 equiv of
t-BuOK, the formation of oxadiazole 5a was spotted after 30
min, while oxazole 8a became detectable after 1 h. This
FIGURE 1. Standard free energy profiles along the reaction coordinate
qualitative kinetic observation seemed to support the hypothesis
of anionic routes of isoxazole 1a in MeOH solution, at 298.15 K.
that the formation of oxazole 8a may occur through the
oxadiazole 5a intermediate. Unfortunately, a quantitative NMR
CHART 2
(15) Any attempt to perform either quantitative or analytical HPLC separation
of isoxazole 1a and oxadiazole 5a from the reaction mixture was unsuccessful.
Moreover, dynamic NMR measurements were precluded by the low solubility
of 1a- and 5a- in DMF-d6 at the temperature reachable with the instrument.
Analytical and preparative TLC were repeatedly eluted in a light petroleum ether/
ethyl acetate 5/1 (vol/vol) mixture to achieve satisfactory separation of the two
compounds.
(16) The previously reported yields of oxazoles 8a and 8b (see also ref 10)
were reproduced, within experimental error, after compounds 1a and 1b were
heated for 4 h at 110 °C in DMF/t-BuOK.
distribution was observed during the reaction of oxadiazole 5a
since, under these reaction conditions, it should readily rearrange
into isoxazole 1a. Although the Boulton-Katritzky rearrange-
ment between 1a and 5a is potentially a reversible reaction,
this result represents the first example of the isolation of the
less stable 1,2,4-oxadiazole from a BK rearrangement of
(17) In the benzo-condensed series, the formation of 1,2,4-oxadiazoles was
reported from a base-catalyzed BK rearrangement of 3-acylaminobenzisoxazoles.
Harsa´nyi, K. J. Heterocycl. Chem. 1973, 10, 957–961.
(18) Ku¨bel, B. Monatsh. Chem. 1983, 114, 373–376.
354 J. Org. Chem. Vol. 74, No. 1, 2009

1,2,4-Oxadiazole-Mediated Isoxazole-to-Oxazole Rearrangement
TABLE 3. Standard Free Energy Values (∆G°, kJ/mol) Relative
to Compound 1a,a for Neutral Species or Transition States, and
Activation Barriers (∆Gq, kJ/mol), Relative to the Given Starting
Compound, Along the Reaction Coordinate, Calculated in Vacuo, in
DMSO and in MeOH at 298.15 K
SCHEME 4
compd
∆G° (in vacuo)
∆G° (DMSO)
∆G° (MeOH)
1a
1a′
2
0.0
27.6
178.9
-15.8
158.2
10.9
8.0
60.1
49.5
0.0
10.0
179.7
-21.4
110.1
19.4
7.6
44.6
48.5
0.0
7.2
177.0
-23.5
95.3
28.7
6.5
29.0
40.4
TABLE 2. Mixture Composition (%), Determined by NMR,a as a
Function of Time (h) for the Reaction of Isoxazole 1a and
Oxadiazole 5a in Refluxing Ethanol/t-BuOK
4
5azw
5aenol
5aketo
6
reaction of isoxazole 1a
1a 5a 8a
97.2 ( 0.3 2.8 ( 0.3 ,0.1a,b
95.3 ( 0.5 3.9 ( 0.3 0.8 ( 0.2 93.5 ( 0.5 4.0 ( 0.2 2.5 ( 0.3
92.1 ( 0.4 3.0 ( 0.3 4.9 ( 0.2 90.0 ( 0.4 4.0 ( 0.3 6.0 ( 0.2
88.3 ( 0.4 3.0 ( 0.4 8.7 ( 0.3 86.0 ( 0.2 4.0 ( 0.3 10.0 ( 0.4
reaction of oxadiazole 5a
1a 5a 8a
93.2 ( 0.5 5.0 ( 0.3 1.8 ( 0.2
reaction
time (h)
7zw
8a
-71.2
-88.7
-79.2
-87.5
-82.5
-82.3
0.5
1.0
2.5
4.0
8aimino
transition state
∆G° (in vacuo)
∆G° (DMSO)
∆G° (MeOH)
‡
[1a′f5azw
]
176.8
136.0
121.7
a Under the conditions used (50 mg of substrate), NMR deter-
minations were able to reproduce the composition of a known prepared
mixture of 1a (94%), 5a (4%), and 8a (2%) within a 0.2% experimental
error and a threshold level of 0.1%. b Traces of compound 8a were
detected only by TLC performed on a concentrated sample.
reactions
1a′f5azw
∆Gq (in vacuo)
∆Gq (DMSO)
∆Gq (MeOH)
149.2
126.0
114.5
a The standard free energy value calculated at B3LYP/6-31++G(d,p)
level is -685.00864 au for 1a; the solvation free energy values for 1a
are -6.01 kcal/mol (in DMSO) and -15.15 kcal/mol (in MeOH) (see
Computational Details in the Experimental Section).
analysis of the reaction mixture was not achievable due to the
very low conversion of the substrate at the early stages of the
reaction.15 However, by performing the reaction in refluxing
ethanol/t-BuOK, substrate conversion was increased19 enough
to allow NMR determination of reaction mixture composition
as a function of time (see Table 2).
which can be possibly involved, kinetic studies for the search
of neutral transition states have been reported only for the BK
step. In this frame, we did not consider explicit solvent
molecules which could be also involved along the reaction path.
As a matter of fact, although 1,3,4-oxadiazole 4 is more stable
than either isoxazole 1a or oxadiazole 5a, it is by far less stable
than oxazole 8a. Moreover, its diazirine precursor 2 has a free
energy (relative to 1a) of ∆G° )179.7 kJ/mol. Therefore, in
analogy with the anion route, the formation of 4 should be both
thermodynamically and kinetically unfavored with respect to
the formation of oxazole 8a.
Similar to the anionic BK route, and starting from the more
stable isoxazole 1a, an initial rotation of the C(3)-Nexocyclic bond,
yielding 1a′, is required to achieve the correct geometry to
undergo the neutral BK rearrangement. At this point, if one
hypothesizes that no proton exchange is involved in the reaction,
the BK step will produce the zwitterionic product 5azw which
could intramolecularly stabilize into its enol form 5aenol and
eventually produce the more stable 5aketo (Scheme 5).
A comparison between data in Tables 1 and 3 shows that the
activation barrier of the BK rearrangement is higher for the
neutral than for the anionic route, suggesting that higher
temperatures are needed just to observe the BK process under
neutral conditions. Moreover, while the activation barrier for
the anionic BK route increased by increasing the polarity of
the reaction media, according to the Hughes and Ingold rules,20
for the neutral route a stronger and opposite effect is observed
suggesting the use of polar solvents. As a matter of fact, no
reaction of 1a was observed in refluxing methanol (bp 65 °C),
Data from Table 2 represent the strongest evidence that
anionic oxadiazole 5a- is a reaction intermediate for the
formation of oxazole 8a from a thermal rearrangement of
isoxazole 1a. In fact, yields in oxazole 8a are consistently higher
from the reaction of 5a (an advanced starting point along the
isoxazole-to-oxazole reaction coordinate) than they are in the
reaction of 1a. In other words, when starting from 100% of 1a,
only the BK rearrangement occurs (∆Gq ) 86.1 kJ/mol),
forming a few percent of the less stable oxadiazole 5a (∆G° )
7.1 kJ/mol); the latter undergoes the “retro”-BK rearrangement
(∆Gq ) 82.2 kJ/mol) faster than the MNAC (∆Gq ) 92.4 kJ/
mol); therefore, oxazole 8a becomes significantly detectable only
after 1 h of total reaction time. When starting from 100% of
oxadiazole 5a, instead, the MNAC route was able to produce
almost 2% of oxazole 8a after just 30 min, while the faster BK
route yielded 93% of 1a. Since the BK equilibrium is the fastest
among the competitive pathways, the use of either isoxazole
1a or oxadiazole 5a as starting compounds makes a significant
difference only during the earlier stages of the rearrangements.
In fact, after 1 h, both reactions contained about 4% of
oxadiazole 5a and both increased the yield of oxazole 8a of
7.5% in the next 3 h.
As far as neutral forms are concerned, DFT calculations have
been performed on reagent, products, and intermediates along
the reaction coordinate of the isoxazole-to-oxazole rearrange-
ment of compound 1a (see Table 3). Moreover, since an
exhaustive computational kinetic study on the competitive
mechanisms should consider all coordinates connecting all
accessible tautomers and conformers (even the less stable ones)
(20) The intramolecular reaction between the isoxazole nucleus (neutral
substrate) and the side-chain oxygen (anionic nucleophile) occurs through a
transition state with charge dispersion. Therefore, the solvent stabilization effect
is more pronounced for the reagent than for the transition state and the activation
barrier increases with solvent polarity. See: (a) Hughes, E. D.; Ingold, C. K.
J. Chem. Soc. 1935, 244–255. (b) Ingold, C. K. Structure and Mechanism in
Organic Chemistry, 2nd ed.; Cornell University Press: Ithaca and London, 1969;
pp 457-463. (c) Reichardt, C. SolVents and SolVent Effects in Organic Chemistry,
3rd ed.; Wiley-VCH: Weinheim, 2003; pp 163-173.
(19) The higher temperature reached in refluxing ethanol allowed an easier
overtake of the activation barriers of both BK (of 1a- and 5a-) and MNAC (of
5a-) rearrangements, with respect to the reaction carried out in methanol.
Moreover, in a previous report (see ref 10), no reaction was observed when 1a
was refluxed in ethanol/KOH medium.
J. Org. Chem. Vol. 74, No. 1, 2009 355

Pace et al.
SCHEME 5
SCHEME 6
isopropyl-1,2,4-oxadiazole 5d was not observed, most likely due
to its extremely low concentration. In fact, the free energy
difference between 1d and 5d is almost double (∆G°5d-1d) 13.5
kJ/mol in MeOH)23 with respect to that of the corresponding
phenyl derivatives 5a and 1a (∆G°5a-1a) 6.5 kJ/mol in MeOH)
due to the lack of the stabilizing diaryloid effect24 exerted by
the phenyl group on the oxadiazole.
ethanol (bp 78 °C), or n-butanol (bp 118 °C), while only traces
of 1,2,4-oxadiazoles 5a and 8a were detected from the reaction
of 1a by prolonged (8 h) refluxing in DMF (bp 153 °C).21
Reaction of 1a performed in a sealed tube and in the absence
of solvent at 170 °C for 4 h yielded 5% of 1,2,4-oxadiazole 5a
and 90% of recovered starting material, while oxazole 8a was
only detected in traces.21 Even lower amounts of 5a were
obtained from the reaction at 170 °C in Ph2O. On the other
hand, when a high boiling protic solvent such as ethylene glycol
was used, reaction of 1a yielded 8% of 5a and 10% of 8a
together with recovered starting material (61%) and a few
amount of ethylene glycol monobenzoate (6%) after just 30 min
at 170 °C.
These observations suggest that, besides temperature, the
protic medium plays a key role in promoting the reaction, most
likely favoring proton exchanges. The formation of oxadiazole
5a under neutral conditions could be interpreted in terms of a
BK rearrangement of the isoxazole 1a into 5azw followed by a
solvent mediated, rather than intramolecular, prototropic shift
from the amide to the acetonyl moiety (Scheme 5). An
alternative hypothesis still considers the anionic form of the
starting substrate as the key intermediate which can be formed
at high temperatures more easily in protic than in aprotic
solvents. In other words, the BK and the MNAC rearrangements
observed under neutral conditions could be the results of
preliminary acid-base equilibria, where the solvent acts as a
proton acceptor, followed by rearrangements of the deprotonated
species and final neutralization (Scheme 6).
Conclusions
The DFT study on the competitive routes available for the
base catalyzed thermal rearrangement of 3-acylaminoisoxazoles
1 showed that the BK reaction is a much more favored process
than either the MNAC or the RCRE. This rearrangement
produces an equilibrium mixture of 3-acylaminoisoxazole 1 and
the corresponding 3-acetonyl-1,2,4-oxadiazole 5, where the latter
is, by far, the minor component. Unlike the previously studied
3-acylamino-1,2,4-oxadiazoles, where the MNACs of both BK
equilibrium components had similar activation barriers, here an
increase in reaction temperature promotes only the MNAC of
oxadiazole 5, irreversibly producing the more stable oxazole 8.
Therefore, while the photoinduced rearrangement of isoxazoles
into oxazoles is well-proven to follow a RCRE route, the thermal
rearrangement of 3-acylaminoisoxazoles 1 into 2-acylaminoox-
azoles 8 should be better interpreted in terms of a cascade BK-
MNAC rearrangement involving 3-acetonyl-1,2,4-oxadiazoles
5 as ancillary intermediates. This mechanistic hypothesis is
supported by the isolation of oxadiazoles 5 from the reaction
mixture as well as by the higher yields in oxazole 8 obtained
when oxadiazole 5 is used as starting material. Although the
Boulton-Katritzky rearrangement between 1 and 5 could be
considered as a potentially reversible equilibrium reaction, this
study represents an interesting example of the isolation of the
less stable 1,2,4-oxadiazole heterocycle in a BK rearrangement
of isoxazoles.
This hypothesis is also supported by the fact that yields in
oxadiazole 5a-c (% yields of 5c > 5a > 5b; see the
Experimental Section) increase with the presumed acidity of
their isoxazole precursors 1a-c. In fact, while isoxazoles 1a-c
remain the major component of the BK equilibrium due to the
higher aromaticity of the heterocyclic ring,22 their acid-base
equilibria will tend to be shifted toward the less acidic species
3-acetonyl-1,2,4-oxadiazoles 5a-c. In the case of the isobu-
tyrroylamino derivative 1d, the formation of 3-acetonyl-5-
Experimental Section
Starting Materials. All chromatographic purification/separations
have been performed using silica gel (0.040-0.063 mesh) and
mixtures of light petroleum/EtOAc in various ratios. Compounds
(23) Standard free energy values of compound 5d, calculated at 298.15 K
and relative to compound 1d, are ∆G° ) 15.0 kJ/mol (in vacuo), ∆G° ) 15.4
kJ/mol (in DMSO), and ∆G° ) 13.5 kJ/mol (in MeOH). The standard free energy
value calculated at the B3LYP/6-31++G(d,p) level is-572.04948 au for 1d;
the solvation free energy values for 1d are -4.65 kcal/mol (in DMSO) and
-11.92 kcal/mol (in MeOH).
(21) In a previous report (ref 10), no reaction was observed by either refluxing
1a in DMF for 4 h or keeping it in the melted phase at 170 °C for 30 min.
(22) (a) Bird, C. V. Tetrahedron 1985, 41, 1409–1414. (b) Bird, C. V.
Tetrahedron 1992, 48, 335–340.
356 J. Org. Chem. Vol. 74, No. 1, 2009

1,2,4-Oxadiazole-Mediated Isoxazole-to-Oxazole Rearrangement
SCHEME 7
acetic acid, and extracted with ethyl acetate. The unified organic
layers were dried over Na2SO4, filtered, and evaporated and the
residue chromatographed.
Reaction of Compound 1a in DMF/t-BuOK. Chromatography
of the reaction mixture yielded starting material 1a (0.05 g; 10%)
and oxazole 8a (0.40 g; 80%). Traces of oxadiazole 5a were
detected, by comparison with an authentic sample, on a repeatedly
eluted TLC of concentrated chromatographic fractions collected
during the elution of isoxazole 1a. Compound 8a had mp 133-135
°C (from benzene; lit.10 mp 134 °C).
Reaction of Compound 1b in DMF/t-BuOK. Chromatography
of the reaction mixture yielded starting material 1b (0.08 g; 16%)
and oxazole 8b (0.38 g; 76%). Traces of oxadiazole 5b (isolated
and fully characterized from the reactions under neutral conditions;
see later) were detected on a repeatedly eluted TLC of concentrated
chromatographic fractions collected during the elution of isoxazole
1b. Compound 8b had mp 147-149 °C (from benzene; lit.10 mp
148-150 °C).
1a,b were synthesized as previously reported.10 By following a
similar procedure, compounds 1c,d were prepared by acylation of
3-amino-5-methylisoxazole (2.0 g; 20.41 mmol) with a slight excess
of the appropriate acyl chloride (22 mmol) in anhydrous toluene
containing an equimolar amount of pyridine (22 mmol) at room
temperature. After the reaction was completed, the solvent was
removed and the residue treated with water, neutralized, and filtered.
3-(4′-Chlorobenzoylamino)-5-methylisoxazole 1c (4.30 g; 89%): mp
215-216 °C (from toluene); IR (Nujol) 3216, 3151, 1695, 1635
1
cm-1; H NMR (300 MHz, CDCl3) δ 2.46 (s, 3H), 6.90 (s, 1H),
Reaction of Compound 1c in DMF/t-BuOK. Chromatography
of the reaction mixture yielded starting material 1c (0.06 g; 12%)
and 2-(4′-chlorobenzoylamino)-5-methyloxazole 8c (0.40 g; 80%).
Traces of oxadiazole 5c (isolated and fully characterized from the
reactions under neutral conditions; see later) were detected on a
repeatedly eluted TLC of concentrated chromatographic fractions
collected during the elution of isoxazole 1c. Compound 8c: mp
165-167 °C (from light petroleum/ethyl acetate); IR (Nujol) 3136,
7.48-7.51 (m, 2H), 7.95-7.98 (m, 2H), 9.98 (s, 1H, exchangeable
with D2O); HRMS calcd for C11H9ClN2O2 236.0353, found
236.0351. 3-Isobutyrroylamino-5-methylisoxazole 1d (3.55 g; 75%):
mp 119-120 °C (from light petroleum); IR (Nujol) 3224, 3154,
1704, 1634 cm-1; 1H NMR (300 MHz, CDCl3) δ 1.26 (d, 6H, J )
6.9 Hz), 2.41 (s, 3H), 2.67 (hept, 1H, J ) 6.9 Hz), 6.78 (s, 1H),
9.80 (bs, 1H, exchangeable with D2O); HRMS calcd for C8H12N2O2
168.0899, found 168.0902.
1
1714 cm-1; H NMR (300 MHz, CDCl3) δ 2.15 (s, 3H), 6.29 (s,
Authentic samples of oxadiazoles 5a,d have been prepared
accordingly to a procedure reported for oxadiazole 5d25 from the
reaction of amidoxime 925 with either benzoyl chloride for (5a) or
isobutyryl chloride (for 5d) (Scheme 7).
1H), 7.23 (d, 2H, J ) 8.4 Hz), 7.86 (d, 2H, J ) 8.4 Hz), 10.74
(Very broad s, 1H exchangeable with D2O); HRMS calcd for
C11H9ClN2O2 236.0353, found 236.0350.
Reaction of Compound 1d in DMF/t-BuOK. Chromatography
of the reaction mixture yielded starting material 1d (0.11 g; 22%)
and oxazole 8d (0.37 g; 74%). No traces of oxadiazole 5d, which
was synthesized as reference compound (see above) were detected
either during the reaction or in the collected chromatographic
fractions. 2-Isobutyrroylamino-5-methyloxazole 8d: mp 107-109
°C (from light petroleum/ethyl acetate); IR (Nujol) 3169, 3134,
1705, 1674 cm-1; 1H NMR (300 MHz, CDCl3) δ 1.25 (d, 6H, J )
6.9 Hz), 2.33 (s, 3H), 2.72 (broad, 1H), 6.53 (s, 1H), 11.20 (very
broad s, 1H, exchangeable with D2O); HRMS calcd for C8H12N2O2
168.0899, found 168.0897.
Analytical-Scale Reactions of Compounds 1a and 5a in
EtOH/t-BuOK. Four identical solutions of isoxazole 1a and of
oxadiazole 5a (0.05 g; 0.247 mmol) in ethanol (20 mL) were
prepared. A 1 mL portion of a 0.025 M solution of t-BuOK in
ethanol was added to each of the above solutions. The eight
solutions were then refluxed under identical conditions and the
reactions stopped at different times (0.5, 1.0, 2.5, and 4.0 h) by
removing the heat source and cooling in a water/ice bath. The
solvent was removed at room temperature under reduced pressure
and the residue treated with water (50 mL), neutralized, and
extracted with diethyl ether (4 × 75 mL). The combined organic
layers were dried on Na2SO4 and filtered, and the solvent was
removed under reduced pressure. In all eight cases, quantitative
mass recovery was obtained. The obtained residue was then
dissolved in CDCl3 and analyzed by NMR. Amounts of compounds
1a, 5a, and 8a were determined by integration of the methyl signals
at 2.43, 2.33, and 2.30 ppm, respectively, and the results are reported
in Table 2.
A mixture of amidoxime 9 (2.0 g; 12.5 mmol), pyridine (20 mL),
and either benzoyl or isobutyryl chloride (14.0 mmol) was stirred
at room temperature overnight. The reaction was then diluted with
water and extracted with EtOAc (3 × 150 mL); the combined
organic layers were dried on Na2SO4 and filtered. After solvent
removal, the residue was melted at 130-140 °C for 1 h to thermally
promote cyclization into oxadiazoles 11, which, in turn, were
deprotected by stirring at room temperature with 3.3 M HCl (11
mL) for 2 h. The reaction was then diluted with water, neutralized
with NaHCO3, and extracted with Et2O (3 × 150 mL); the combined
organic layers were dried on Na2SO4 and filtered. After solvent
removal, the residue was chromatographed to yield final oxadiazoles
5a,d. 3-Acetonyl-5-phenyl-1,2,4-oxadiazole 5a (1.10 g; 45%): mp
103-105 °C (from light petroleum/ethyl acetate); IR (Nujol) 1711
1
cm-1; H NMR (300 MHz, CDCl3) δ 2.33 (s, 3H), 3.95 (s, 2H),
7.50-7.60 (m, 3H), 8.11-8.13 (m, 2H); HRMS calcd for
C11H10N2O2 202.0742, found 202.0739. 3-Acetonyl-5-isopropyl-
1,2,4-oxadiazole 5d25 (1.05 g; 50%): colorless oil; IR (Nujol) 1733
1
cm-1; H NMR (300 MHz, CDCl3) δ 1.37 (d, 6H, J ) 6.9 Hz),
2.24 (s, 3H), 3.19 (hept, 1H, J ) 6.9 Hz), 3.82 (s, 2H); HRMS
calcd for C8H12N2O2 168.0899, found 168.0896.
Besides using 5a as a substrate for analytical scale reactions (see
below), samples of 5a and 5d were used as reference compounds
to facilitate their detection from reaction mixtures containing them
in very low concentration.
General Procedure for the Reaction of Compounds 1a-d
in DMF/t-BuOK. The procedure to perform the base-catalyzed
rearrangements of isoxazoles 1a-d was similar to the one previ-
ously reported.10 A 0.5 g portion of isoxazole was dissolved in 15
mL of DMF and added with 1 equiv of t-BuOK. The reaction
mixture was then heated at 110 °C and monitored by TLC. After
4 h, the mixture was diluted with water (100 mL), neutralized with
Reaction of Compound 1a in the Absence of Solvent. Isoxazole
1a (0.6 g; 2.97 mmol) was heated in a sealed tube at 170 °C (at
which temperature the reaction occurred in a melted liquid phase)
for 4 h. The tube was then allowed to cool at room temperature
and the residue chromatographed to yield starting isoxazole 1a (0.54
g; 90%) and oxadiazole 5a (0.03 g; 5%).
General Procedure for the Reaction of Compounds 1a-d
in Ethylene Glycol. Isoxazoles 1a-d (0.50 g) were dissolved in
ethylene glycol (20 mL), and the mixture was allowed to stir, in a
closed tube, at 170 °C for 30 min. The reaction mixture was then
(24) (a) Vivona, N.; Cusmano, G.; Ruccia, M.; Spinelli, D. J. Heterocycl.
Chem. 1975, 12, 985–988. (b) Buscemi, S.; Frenna, V.; Vivona, N.; Petrillo, G.;
Spinelli, D. Tetrahedron 1995, 51, 5133–5142. (c) Buscemi, S.; Frenna, V.; Pace,
A.; Vivona, N.; Cosimelli, B.; Spinelli, D. Eur. J. Org. Chem. 2002, 1417–
1423.
(25) Ku¨bel, B. Monatsh. Chem. 1982, 113, 793–803.
J. Org. Chem. Vol. 74, No. 1, 2009 357

Pace et al.
cooled to room temperature, diluted with water (100 mL), and
extracted with diethyl ether (4 × 100 mL). The combined organic
layers were further washed with water, dried on Na2SO4, and
filtered, and the residue was chromatographed after solvent removal.
Reaction of Compound 1a in Ethylene Glycol. Chromatogra-
phy of the residue yielded starting isoxazole 1a (0.31 g; 62%),
3-acetonyl-5-phenyl-1,2,4-oxadiazole 5a, (0.04 g; 8%), 2-benzoy-
lamino-5-methyloxazole 8a10 (0.05 g; 10%), and ethylene glycol
monobenzoate (0.03 g; 7%) likely formed by solvolysis of
acylamino derivatives 1a or 8a.
Information. Transition-state structures were found by the synchro-
nous transit guided quasi-Newton method.28 Vibration frequency
calculations, within the harmonic approximation, were performed
to confirm whether each obtained geometry represented a transition
state or a minimum in the potential energy surface. All of the
minimum energy structures presented only real vibrational frequen-
cies, whereas all of the transition state structures found were first-
order saddle points. To take into account the relative stabilization
of intermediates, the standard free energies, at 298.15 K, of the
considered species have been evaluated by vibrational frequency
calculations of the in vacuo species. Their free energy in DMSO
and in MeOH solutions was evaluated by adding the solvation free
energy to the free energy of the species in vacuo.3 The solvation
free energy in DMSO and in MeOH was calculated by the
conductor-like polarized continuum model,29 without further op-
timization of the geometry (i.e., single point calculations) within
the solvent medium.
Reaction of Compound 1b in Ethylene Glycol. Chromatog-
raphy of the residue yielded starting isoxazole 1b (0.28 g; 56%),
3-acetonyl-5-(4′-methoxyphenyl)-1,2,4-oxadiazole 5b (0.03 g; 6%),
2-(4′-methoxybenzoylamino)-5-methyloxazole 8b10 (0.03 g; 6%),
and ethylene glycol mono(4′-methoxy)benzoate (0.03 g; 7%) likely
formed by solvolysis of 1b or 8b. Compound 5b: mp 93-94 °C
1
(from light petroleum/ethyl acetate): IR (Nujol) 1719 cm-1; H
NMR (300 MHz, CDCl3) δ 2.33 (s, 3H), 3.90 (s, 3H), 3.93 (s,
2H), 7.02 (d, 2H, J ) 9.0 Hz), 8.07 (m, 2H, J ) 9.0 Hz); HRMS
calcd for C12H12N2O3 232.0848, found 232.0846.
All calculations were performed using the Gaussian98 program
package.30
Acknowledgment. Financial support from the University of
Palermo is gratefully acknowledged.
Reaction of Compound 1c in Ethylene Glycol. Chromatogra-
phy of the residue yielded starting isoxazole 1c (0.30 g; 60%),
3-acetonyl-5-(4′-chlorophenyl)-1,2,4-oxadiazole 5c, (0.07 g; 14%),
2-(4′-chorobenzoylamino)-5-methyloxazole 8c (0.05 g; 10%), and
ethylene glycol mono(4′-chloro)benzoate (0.05 g; 12%) likely
formed by solvolysis of 1c or 8c. Compound 5c: mp 120-121 °C
(from light petroleum/ethyl acetate); IR (nujol) 1718 cm-1; 1H NMR
(300 MHz, CDCl3) δ 2.34 (s, 3H), 3.96 (s, 2H), 7.52 (d, 2H, J )
8.4 Hz), 8.08 (d, 2H, J ) 8.4 Hz); HRMS calcd for C11H9ClN2O2
236.0353, found 236.0356.
Reaction of Compound 1d in Ethylene Glycol. Chromatog-
raphy of the residue yielded starting isoxazole 1d (0.25 g; 50%),
2-isobutyrylamino-5-methyloxazole 8d (0.05 g; 10%), and ethylene
glycol monoisobutyrate (0.05 g; 13%) likely formed by solvolysis
of 1d or 8d.
Supporting Information Available: Comparison table of
calculated activation barriers of BK, MNAC, and RCRE
rearrangements for some O-N bond containing azole systems.
Cartesian coordinates, B3LYP energy (in au), and standard
thermochemical data (in au) at 298.15 K of compounds reported
in Tables 1 and 3 as well as of compounds 1d and 5d. 1H NMR
scpectra of compounds 1c,d, 5a-d, and 8c,d. This material is
available free of charge via the Internet at http://pubs.acs.org.
JO802081K
J. A.; Frisch, M. J. Chem. Phys. Lett. 1992, 197, 499–505.
(28) Peng, C.; Schlegel, H. B. Isr. J. Chem. 1994, 33, 449–454.
Computational Details. The molecular and anionic species
considered, shown in Schemes 3 and 5, were the lowest energy
tautomers and/or conformers found for each isomer in the MeOH
solvent (see below). Their geometry was fully optimized by using
the hybrid DFT B3LYP method26 and the 6-31++G(d,p) basis
set.27 Their Cartesian coordinates are available in the Supporting
(29) Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995–2001.
(30) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.;
Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.;
Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson,
G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A. G.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts,
R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.;
Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.;
Pople, J. A. Gaussian 98, reVision A.8; Gaussian, Inc.: Pittsburgh, PA, 1998.
(26) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(27) (a) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56,
2257–2261. (b) Clark, T.; Chandrasekhar, J.; Schleyer, P. V. R. J. Comput. Chem.
1983, 4, 294–301. (c) Krishnam, R.; Binkley, J. S.; Seeger, R.; Pople, J. A.
J. Chem. Phys. 1980, 72, 650–654. (d) Gill, P. M. W.; Johnson, B. G.; Pople,
358 J. Org. Chem. Vol. 74, No. 1, 2009