Pyridinium N-ylides and (arylmethylene)azol-5-ones. Reaction cascade leading to an unusual spiroisoxazolinone ring

Article abstract of DOI:10.1016/S0040-4020(00)00921-2

The reaction of a series of (arylmethylene)azol-5-ones with phenacylpyridinium salt 3 in glacial acetic acid/ammonium acetate mixture gives different results depending on the starting azolone. The isoxazol-5-ones 1 give the unusual spirans 6 in a reaction cascade involving Michael- and retro-Michael reactions, C-alkylation, aldol addition, and diastereospecific cyclization. The reaction performed with oxazol-5-ones 11 has shown that a literature report has to be corrected since no oxazolopyridines 12 but rather arylideneimidazol-5-ones 13 are produced. In the case of pyrazolin-5-ones 14, the unavoidable formation of bis-adducts 15 always prevents any other type of reaction. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00921-2

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 9669±9674
Pyridinium N-Ylides and (Arylmethylene)azol-5-ones. Reaction
Cascade Leading to an Unusual Spiroisoxazolinone Ring
*
Francesco Risitano, Giovanni Grassi, Francesco Foti and Cristina Bilardo
Á
Istituto di Chimica dei Composti eterociclici, Universita, Vill. S. Agata, I-98166 Messina, Italy
Received 7 July 2000; revised 7 September 2000; accepted 28 September 2000
AbstractÐThe reaction of a series of (arylmethylene)azol-5-ones with phenacylpyridinium salt 3 in glacial acetic acid/ammonium acetate
mixture gives different results depending on the starting azolone. The isoxazol-5-ones 1 give the unusual spirans 6 in a reaction cascade
involving Michael- and retro-Michael reactions, C-alkylation, aldol addition, and diastereospeci®c cyclization. The reaction performed with
oxazol-5-ones 11 has shown that a literature report has to be corrected since no oxazolopyridines 12 but rather arylideneimidazol-5-ones 13
are produced. In the case of pyrazolin-5-ones 14, the unavoidable formation of bis-adducts 15 always prevents any other type of reaction.
q 2000 Elsevier Science Ltd. All rights reserved.
Introduction
and provide the nitrogen of the ®nal heterocycle. This result
has suggested that similar treatment with AcOH/AcONH4 of
arylidene derivatives 1 and ylides 2 or of betaines 4 would
be an attractive and versatile route to known isoxazolo-
[5,4-b]pyridine systems 5.9
4-(Arylmethylene)isoxazol-5-ones
1 remarkably show
heterodienic behaviour towards Diels±Alder reactions
with a variety of electron-rich dienophiles and can be
considered as a useful means for the preparation of fused
heterocycles.1 On the other hand the exocyclic double bond
of these arylmethylene derivatives 1 can also work perfectly
well as an electrophilic or dipolarophilic unit. Nevertheless,
while it is recognised that it can easily undergo addition
reactions of primary and secondary amines2 and with
C-nucleophiles3 to give open-chain adducts, its dipolaro-
philic reactivity with 1,3-dipoles has successfully been
tested in only a few cases and is limited to nitrile oxides,4
munchnones5 and benzonitrilium N-phenylimides.6
In practice, this reaction has led to a variety of products,
depending on the conditions used, which have, however,
never included the desired isoxazolopyridines 5. Here we
report the results obtained and the action of the ylides 2 is
extended both to arylmethylene-pyrazolin-5-ones 14 and to
arylmethylene-oxazol-5-ones 11. With the latter, the correct
course of a previously studied reaction is clari®ed.10
Results and Discussion
Some of our recent work on the reactivity of the above-
mentioned arylidene derivatives 1 towards the easily acces-
sible dipole-like 2 has demonstrated that these intermediates
readily react with 1, but only as C-nucleophiles, producing
high yields of stable betaines 4.7 Although these zwitterions
have shown no tendency to rearrange into fused 5.5 or 5.6
heterocycles, similar non-isolated intermediates have never-
theless been thought to have been successfully converted
into carbocyclic aromatic systems or substituted and anne-
lated pyridines.8 Generally the latter transformation is
achieved by getting a,b-unsaturated carbonylic compounds,
sometimes incorporated in rings, to react with pyridinium
salts 3 in a mixture of glacial acetic acid and ammonium
acetate which may liberate in situ the appropriate ylide 2
Our initial attempt to obtain fused heterocycles 5 involved
reacting equimolecular quantities of pyridinium salt 3 and 1
in MeOH in the presence of glacial acetic acid-ammonium
acetate at room temperature or under re¯ux. Under these
conditions, the reactions stopped with high yields of 4, as
already described using triethylamine.7 Further experiments
carried out using a mixture of pyridinium salt 3 and 1, or
the monocyclic compounds 4, using AcOH/AcONH4 in
re¯uxing benzene or toluene,8 proved unsuccessful in the
formation of the desired compounds 5. Surprisingly, spirans
6 were isolated from the reaction in moderate yields (35%)
as single diastereomers (Scheme 1). All attempts to obtain
these spirans 6 without the arylidene group, for example by
conducting the reaction under strictly anhydrous conditions,
were unsuccessful, giving always the same compounds 6.
The structure of the spirans indicates that in the presence of
AcOH/AcONH4 one molecule of arylaldehyde is further
required to react with stochiometric quantities of 4 or 1
Keywords: pyridinium N-ylides; (arylmethylene)azol-5-ones; spirans;
cascade reactions; cleavage reactions.
* Corresponding author. Tel.: 1390-90-393-897; fax: 1390-90-398-892;
e-mail: frisitan@isengard.unime.it
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00921-2

9670
F. Risitano et al. / Tetrahedron 56 (2000) 9669±9674
Scheme 1.
with 2. Therefore, when a 1:1 molar ratio of betaine/
aldehyde (or 1:1:1 of arylidene/aldehyde/ylide) was used,
yields of spirans 6 were increased (Table 1).
disubstituted isoxazol-5-one ring,12 while the absorption
band at about 1670 cm21 of the benzoyl group originally
present in the betaines 4 disappeared. The 1H NMR spectra
are also consistent with the spiro-structure. A singlet always
appears at d 5.88±6.35 and this can be assigned to the H6
protons; the presence of the two aryls, both derived from the
arylmethylene group of the starting product 1, was,
however, evidenced by the fact that in cases b, c, e, f, g
The identi®cation of 6 was based on their spectral and
analytical data, and con®rmed by X-ray crystallographic
analyses carried out on 6b and 6f.11 The IR spectra showed
a strong absorption at n 1780±1799 cm21, indicating a 4,4-
Table 1. Analytical and spectroscopic data of the spirans 6 and 10
Compound
(Formula)
Mp (8C)
Yield (%)
Found (Calcd)
IRa
C
H
N
1H NMRb
5.95c; 6.89±7.99d
6a
(C31H22N2O2)
6b
(C33H26N2O2)
6c
(C33H26N2O4)
6d
(C31H20N2O2Cl2)
6e
(C33H26N2O2)
6f
(C33H26N2O4)
6g
(C33H26N2O2)
6h
(C33H26N2O4)
10i
180
230
210
231
194
196
176
178
207
202
65
70
71
57
65
66
50
51
18
18
81.93
(81.92)
82.01
(82.13)
77.14
(77.02)
71.09
(71.14)
82.03
(82.13)
77.10
(77.02)
82.05
(82.13)
77.11
(77.02)
79.39
4.90
(4.88)
5.35
(5.43)
5.15
(5.09)
3.80
(3.82)
5.39
(5.43)
5.02
(5.09)
5.38
(5.43)
5.13
(5.09)
5.21
6.07
(6.16)
5.98
(5.81)
5.55
(5.44)
5.46
(5.35)
5.79
(5.81)
5.24
(5.44)
5.96
(5.81)
5.60
(5.44)
5.76
1795
1792
17.83
1791
1799
1796
1789
1786
1780
1784
5.88c; 2.25e; 2.33e; 6.80±7.93f
5.87c; 3.75e; 3.80e; 6.69±7.92f
5.90c; 6.89±7.99f
6.34c; 1.68e; 2.10e; 6.87±7.96f
6.30c; 3.37e; 3.49e; 6.57±7.97f
6.35c; 2.10e; 2.18e; 6.87±7.97f
6.30c; 3.38e; 3.49e; 6.50±7.97f
5.87c; 2.25e; 3.79e; 6.79±7.92f
5.80c; 2.33e; 3.74e; 6.82±7.92f
(C33H26N2O3)
10l
(C33H26N2O3)
(79.50)
79.41
(79.50)
(5.30)
5.28
(5.30)
(5.70)
5.84
(5.70)
a Nujol (nCvO/cm21).
b CDCl3 (d/ppm).
c (s, 1H, C±H).
d (20H, m, ArH).
e (s, 3H, Me).
f (18H, m, ArH).

F. Risitano et al. / Tetrahedron 56 (2000) 9669±9674
9671
Scheme 2.
and h there are two separate methyl resonance singlets, one
at lower ®eld assigned to the aryl methyl bonded at C-sp2
and the other, therefore, the aryl methyl on C-sp3.
gives the spiran 6a and, in addition, 4-(2,4,6-trimethyl-
phenyl)methylene-3-phenylisoxazol-5-one 1 (Arˆmesityl).14
The attempted unsuccessful synthesis of fused heterocycles
5 from all these reactions can therefore be considered as
con®rmation that betaines 4 cannot rearrange to fused
heterocycles. Moreover, in agreement with previous obser-
vations,15 such a heteroaromatic system with both peri-
positions occupied by bulky groups would be unlikely to
form.
We believe that betaines 4 are the starting products of these
reactions, and the formation of spirans 6 depends on the
facility to cleave the C4±CAr bond under the reaction con-
ditions.
The released chalcone 7 can then hydrolyze to salt 3 and the
corresponding aldehyde; these are immediately intercepted
by the anion 8 to form the disubstituted isoxazolone 9 which
®nally evolves stereospeci®cally to spiro compound 6
according to a 5-exo-trig cyclization (Scheme 2).13
Nevertheless, in contrast to our results, arylmethylene-
oxazol-5-ones 15 (Scheme 4) have, under the same con-
ditions, been reported to give fused 2,4,6-triaryl-
oxazolo[5,4-b]pyridine 1210 and, although the afore-
mentioned structures were subsequently disproved,16 no
alternative was advanced.
As proof of this behaviour, besides the observation that the
addition of arylaldehyde always increased yields, it can be
shown (Scheme 3) that, starting both from 4b and 4c and
from the reaction between 1b and p-anisaldehyde or
between 1c with p-tolylaldehyde and N-phenacyl pyridi-
nium 3, under the usual conditions, we have obtained all
four possible spirans 6b and 6c, 10i and 10l, rather than a
single mixed spiran. As further con®rmation, the reaction of
4a with the hindered mesitaldehyde and AcOH/AcONH4
We repeated the reaction of 4-benzylidene-2-phenyloxazol-
5-one 11 (ArˆPh) with phenacylpyridinium salt 3 by re¯ux-
ing both in AcOH/AcONH4, according to Bansal and Jain,10
and also in chloroform containing triethylamine under the
conditions used for isoxazolones 17 or under re¯ux.
Whereas with triethylamine no reaction was observed,
Scheme 3.

9672
F. Risitano et al. / Tetrahedron 56 (2000) 9669±9674
Scheme 4.
Scheme 5.
Table 2. Analytical and spectroscopic data of the bis-derivatives 15
Compound
(Formula)
Mp (8C)
Yield (%)
Found (Calcd)
N
C
H
IRa
1H NMRb
a
(C37H28N4O2)
b
(C38H30N2O2)
c
(C38H30N4O3)
d
(C37H27N4O2Cl)
e
(C38H30N4O2)
f
(C38H30N4O3)
220
166
156
174
216
215
170
160
55
46
47
45
49
37
43
43
79.15
(79.26)
79.52
(79.42)
77.44
(77.27)
79.69
(79.55)
79.39
(79.42)
77.45
(77.27)
79.61
(79.42)
77.35
(77.27)
4.99
(5.03)
5.32
(5.26)
5.26
(5.12)
5.10
(4.96)
5.29
(5.26)
5.22
(5.12)
5.39
(5.26)
5.09
(5.12)
10.11
(9.99)
9.89
(9.75)
9.61
(9.49)
10.16
(10.03)
9.87
(9.75)
9.31
(9.49)
9.90
(9.75)
9.61
(9.49)
1596
1600
1602
1600
1602
1604
1598
1605
5.44c 7.17±7.83d
5.05c; 2.30e; 6.81±7.58f
5.13c; 3.72e; 6.75±7.65f
5.03c; 6.86±7.99f
5.24c; 2.30e; 7.40±7.99f
5.38c; 3.50e; 6.75±7.65f
5.12c; 2.25e; 6.80±7.64f
5.32c; 3.72e; 7.11±7.79f
g
(C38H30N4O2)
h
(C38H30N4O3)
a Nujol (nmax/cm21).
b CDCl3 (d/ppm).
c (s, 1H, C±H).
d (25H, m, ArH).
e (s, 3H, Me).
f (24H, m, ArH).

F. Risitano et al. / Tetrahedron 56 (2000) 9669±9674
9673
under Bansal's conditions we obtained 4-benzylidene-2-
phenylimidazolin-5-one 13 (ArˆPh).17 This compound
showed a melting point and spectroscopic characteristics
very similar to those of the earlier claimed 12 (ArˆPh),
so it seems probable that, in general, the oxazolopyridines
of Bansal and Jain are in fact arylideneimidazolin-5-ones
13.
(5 mmol), ammonium acetate (3.8 g, 50 mmol) and glacial
acetic acid (20 mL) and the appropriate aromatic aldehyde
(5 mmol) was stirred in re¯uxing toluene (20 mL). The
solution was at ®rst orange but later became green-brown.
After 2 h, the reaction mixture was ®ltered and the solvent
evaporated under reduced pressure to give an oily residue
which was crystallized from methanol to yield the spirans 6.
According to this procedure, identical results were obtained
from betaines 4 as starting material. Yields, analytical and
selected spectroscopic data of the new compounds are given
in Table 1.
In con®rmation, the 4-arylmethylene-2-phenyloxazol-5-
ones 11 used by Bansal and Jain10 were treated with
AcOH/AcONH4 in the presence and in the absence of the
salt 3; in both cases only the corresponding imidazolin-5-
ones 13 were obtained, indicating that in our hands these
reactions achieved exclusively the transformation of the
oxazolone ring into the imidazolinone, consistent with
previous results.18
General procedure for the reactions of the 4-(aryl-
methylene)oxazol-5-ones (11) with phenacyl-
pyridinium salt (3)
4-(Arylmethylene)oxazol-5-ones 11 (2.6 mmol) and salt 3
(2.6 mmol) were re¯uxed for 3 h in glacial acetic acid
(20 mL) containing ammonium acetate (26 mmol).The
precipitate which formed on cooling was collected by ®l-
tration and recrystallized from methanol to give products
13. Analytical and spectroscopic data are identical to
those reported in Ref. 10.
Arylmethylenepyrazolin-5-ones 14, however, showed beha-
viour different from both isoxazolone 1 and oxazolone 11
derivatives. Indeed, after treatment under the same con-
ditions with ylid 2 generated in situ from salt 3 in the
presence of triethylamine or AcOH/AcONH4 and with a
signi®cant excess of the arylaldehyde corresponding to the
arylmethylene group, these did not give the expected
products 16 or 18, but only the corresponding bis-deriva-
tives 15 (Scheme 5).
Reactions of the 4-(arylmethylene)pyrazol-5-ones (14)
with phenacylpyridinium salt (3)
The structure of these can easily be deduced from analytical
and spectroscopic data (Table 2) and con®rmed by the
comparison of their IR spectra with that of compound
15a, already known and prepared by other means.19
Obviously, in these cases is possible that hydrolytic scission
of the arylmethylene bond and subsequent unavoidable
formation of bis-adducts 15 take place, preventing any
other type of reaction. Otherwise an initial formation of
the betaine 16 can also occur, which under the reaction
conditions splits into 7 and 17 with the latter immediately
forming 15 in the presence of the starting products 14.
The procedure described above (routes a and b) for isoxa-
zol-5-ones 1 was used for pyrazolones 14. The residues
were chomatographed on silica gel (chloroform) to afford
15. Yields analytical and selected spectroscopic data are
given in Table 2.
References
1. De Simoni, G.; Gamba, A.; Righetti, P. P.; Tacconi, G. Gazz.
Chim. Ital. 1971, 899±922; De Simoni, G.; Tacconi, G. Gazz.
Chim. Ital. 1968, 1329±1342.
2. Knowles, A. M.; Lawson, A. J. Chem. Soc., Perkin Trans. 1
1972, 1240±1243.
Experimental
Á
3. Pastour, P. Comp. Rend. 1957, 244, 2243±2248; Mustafa, A.;
Melting points: Reichert±Ko¯er hot-stage microscope. Micro-
analyses: Carlo Erba EA 1102. IR: Nicolet FT-IR Impact
400D spectrometer. H NMR spectra with a Bruker ARX
300, tetramethylsilane as internal reference, hexadeutero-
dimethyl sulfoxide as solvent. 4-(arylmethylene)azol-5-ones
1,20 1121 and 1422 and phenacylpyridinium salt 323 were
prepared by standard procedures.
Asker, W.; Harhash, A. H.; Kassab, N. A. L. Tetrahedron 1963, 52,
1577±1585; Lo Vecchio, G.; Caruso, F.; Foti, F.; Grassi, G.;
Risitano, F. Atti Accademia Peloritana dei Pericolanti 1974,
303±310; Cook, D. C.; Lawson, A. J. Chem. Soc., Perkin Trans.
1 1974, 1112±1116.
1
4. Lo Vecchio, G.; Grassi, G.; Risitano, F.; Foti, F. Tetrahedron
Lett. 1973, 39, 3777±3780.
5. Clerici, F.; Erba, E.; Mornatti, P.; Trimarco, P. Chem. Ber.
1989, 122, 295±300; Grassi, G.; Risitano, F.; Caruso, F.; Foti, F.
presented in part at the Euchem Conference, Vulcano (Italy), June,
1995.
General procedure for the reactions of the 4-(aryl-
methylene)isoxazol-5-ones (1) with phenacyl-
pyridinium salt (3)
6. Abbass, J. M.; Mosselhi, M. A. N.; Abdallah, M. A.; Shawali,
A. S. J. Chem. Research (Synopsis) 1995, 190±191.
Á
7. Risitano, F.; Grassi, G.; Bruno, G.; Nicolo, F. Liebigs Ann.
1997, 2, 441±445.
È
8. Krohnke, F.; Zecher, W.; Curtze, J.; Drechsler, D.; P¯eghar, K.;
Schnalke, K. E.; Eis, W. W. Angew. Chem. 1962, 74, 811±817;
È
Krohnke, F.; Zecher, W. Angew. Chem., Int. Ed. Engl. 1962, 12,
626±639.
Route (a): A mixture of 4-(arylmethylene)isoxazol-5-ones 1
(5 mmol), salt 3 (5 mmol), ammonium acetate (3.8 g,
50 mmol) and glacial acetic acid (20 mL) in methanol was
stirred for 2 h at room temperature or at re¯ux. The solvent
was removed under reduced pressure, and then the residue
was allowed to cool and quenched with water to yield the
betaines 4.7
9. Adembri, G.; Campanini, A.; Ponticelli, F.; Tedeschi, P.
J. Chem. Soc., Perkin Trans. 1 1975, 2190±2194.
Route (b): A mixture of isoxazol-5-ones 1 (5 mmol), salt 3

9674
F. Risitano et al. / Tetrahedron 56 (2000) 9669±9674
10. Bansal, R. K.; Jain, J. K. Synthesis 1986, 840±842.
Á
11. Bruno, G.; Nicolo, F.; Risitano, F.; Grassi, G.; Foti, F. Acta
Crystallogr., Sect. C, submitted for publication.
18. Williams, D. L.; Ronzio, A. R. J. Am. Chem. Soc. 1946, 647±649.
19. Knorr, L.; Klozt, C. Ber. 1887, 20, 2545±2548; Ehsan, A.; Ali,
S.; Ahmed, I.; Karimullah, G. Pakistan J. Sci. Ind. Res. 1967, 10,
228±229.
12. Boulton, A. J.; Katritzky, A. R. Tetrahedron 1961, 41±50.
13. Baldwin, J. E. J. Chem. Soc., Chem. Commun. 1976, 734±741.
14. Lo Vecchio, G.; Risitano, F.; Grassi, G.; Foti, F.; Caruso, F.
Atti Accademia Peloritana dei Pericolanti 1974, 297±302.
15. Skotsch, S.; Kohl Meyer, I.; Breitmaier, E. Synthesis 1979,
449±452.
20. Quilico, A. The Chemistry of Heterocyclic Compounds;
Wiley-Interscience: New York, 1962; pp 117±124.
21. Rao, Y. S. J. Org. Chem. 1976, 41, 722±725; Carter, H. E.
Org. React. 1974, 3, 199±201.
22. Risitano, F.; Grassi, G.; Caruso, F.; Foti, F. Tetrahedron 1996,
1443±1450.
16. Gelmi, M. L.; Pocar, D.; Viziano, M.; Destro, R.; Merati, F.
J. Chem. Soc., Perkin Trans. 1 1992, 701±705.
17. Erlenmeyer, E. Ber. 1900, 33, 2036±2041.
23. Henrick, C. A.; Ritchie, E.; Taylor, W. C. Aust. J. Chem. 1967,
20, 2441±2453.