Synthesis and properties of azoles and their derivatives. 57. Unexpected results of 3-nitropropene-1 [2+3] cycloaddition to C,C,N-triphenylnitrone

Article abstract of DOI:10.1007/s10593-006-0244-6

[2+3] Cycloaddition of 3-nitropropene-1 with C,C,N-triphenylnitrone leads to a mixture of 5-nitromethyl-2,3,3-triphenylisoxazolidine and 5-methyl-4-nitro-2,3,3-triphenylisoxazolidine. The results obtained can be explained assuming that 3-nitropropene-1 isomerizes to 1-nitropropene-1 under reaction conditions.

Full text of DOI:10.1007/s10593-006-0244-6

Chemistry of Heterocyclic Compounds, Vol. 42, No. 10, 2006
SYNTHESIS AND PROPERTIES OF AZOLES
AND THEIR DERIVATIVES. 57*. UNEXPECTED RESULTS
OF 3-NITROPROPENE-1 [2+3] CYCLOADDITION
TO C,C,N-TRIPHENYLNITRONE
R. Jasinski, M. Kwiatkowska, and A. Baranski
[2+3] Cycloaddition of 3-nitropropene-1 with C,C,N-triphenylnitrone leads to a mixture of
5-nitromethyl-2,3,3-triphenylisoxazolidine and 5-methyl-4-nitro-2,3,3-triphenylisoxazolidine. The results
obtained can be explained assuming that 3-nitropropene-1 isomerizes to 1-nitropropene-1 under
reaction conditions.
Keywords: [2+3] cycloaddition, nitrone, nitropropene, isomerization.
In previous papers of this series [1-4] we described the [2+3] cycloaddition of trans-R-substituted
nitroethylenes (conjugated systems) with aryl nitrones. It was found that the reaction is entirely regioselective
and leads to 4-nitro-5-R-isoxazolidines as the only reaction products. The present contribution is intended as an
extension of our studies. In particular, it initiates investigations on the reactivity of unconjugated nitroalkenes in
the reaction with aryl nitrones. For this purpose 3-nitropropene-1 (1) and C,C,N-triphenylnitrone (2) were
chosen as model compounds. Theoretically [5,6], [2+3] cycloaddition of such reactants should lead to
5-nitromethyl-2,3,3-triphenylisoxazolidine (3).
The reaction was carried out in solvent-free conditions at room temperature using a tenfold molar excess
of alkene. When the reaction was completed, the excess alkene was removed in vacuum and the semiliquid
residue was tested by means of HPLC. Unexpectedly, it was found that the reaction yielded two products rather
than one product as well as a small quantity of benzophenone. Their RT were different from those of the starting
compounds. Separation of the mixture by means of semipreparative HPLC produced compound 3 and 5-methyl-
4-nitro-2,3,3-triphenylisoxazolidine (4) in molar ratio ca. 1:5 (Scheme 1).
It is interesting to note at this point that [2+3] cycloaddition of the same nitroalkene to aromatic nitrile
N-oxides gave only 3-aryl-5-nitromethyl- isoxazolidines [7].
Despite this finding, which at first sight appears to be an unexplainable, it can be understood by
assuming that under the reaction conditions nitropropene 1 isomerizes partially to trans-1-nitropropene-1 (5) by
a [1,3]-sigmatropic shift. The possibility of such isomerization was reported some years ago [8, 9].
_______
* For Part 56, see [1].
__________________________________________________________________________________________
Institute of Organic Chemistry and Technology, Cracow Technical University, 31-155 Krakow, Poland;
e-mail: pcbarans@chemia.pk.edu.pl. Published in Khimiya Geterotsiklicheskikh Soedinenii, No. 10,
pp. 1545-1548, October, 2006. Original article submitted May 11, 2006.
1334
0009-3122/06/4210-1334©2006 Springer Science+Business Media, Inc.

Scheme 1
Ph
Ph
H
H
H
Ph
Ph
Ph
C
C
C
N+
+
N
Ph
CH2NO2
O
CH2NO2
O
3
2
1
Ph
Ph
NO2
Me
H
H
2
C
N
Ph
Me
O
C
NO2
4
5
Probably, the conjugated nitroalkene 5 reacts with nitrone 2 much faster than the unconjugated isomer 1.
This hypothesis is in agreement with an interaction diagram elaborated by us on the basis of FMO energies
estimated for reactants 1, 2 and 5 (Fig. 1). From the diagram it is evident that both cycloadditions are controlled
by HOMOnitrone–LUMOalkene interactions (∆E2–∆E1>1 eV). According to Sustmann's terminology [10] they are
the normal electron demand reactions. However, the energy gap (∆E1) between LUMO of nitroalkene 5 and
HOMO of nitrone 2 is smaller than that between LUMO of nitroalkene 1 and HOMO of nitrone 2. Therefore,
with regard to FMO theory [10, 11], the reaction leading to compound 4 should be preferred. Actually, by
independent experiment it was proved that the [2+3] cycloaddition of nitropropene 5 with nitrone 2 in solvent-
free conditions resulted in compound 4 with almost quantitative yield. Nevertheless, we are aware more rigorous
investigations along similar lines are called for.
∆E1 = ⎢EHOMO nitrone – ELUMO alkene
∆E2 = ⎢EHOMO alkene – ELUMO
⎢
8.39
7.57
⎢
10.53
10.84
nitrone
Fig. 1. FMO interaction diagram for [2+3] cycloaddition of C,C,N-triphenylnitrone 2 to nitropropenes 1,5
(data taken from AM1 calculations).
1335

EXPERIMENTAL
1
Melting points were determined on the Boetius apparatus. H NMR spectra were recorded on a Tesla
BS-567C (80 MHz) spectrometer in CDCl3 using TMS as an internal standard. IR spectra were recorded on a
Bio-Rad-175C spectrometer in KBr discs. Mass spectra (70eV) were measured with a Varian MAT 112
spectrometer. HPLC analyses were carried out with a Knauer apparatus equipped with UV-VIS detector, using a
Lichrospher 100RP column (4 × 240) and methanol–water (3:1 v/v) as an eluent at a flow rate of 1.3 cm3/min.
For separation of cycloadducts semipreparative HPLC was applied with a Lichrospher 100RP column (16 × 250)
and methanol–water (3:1 v/v) as an eluent at a flow rate of 10 cm3/min. 3-Nitropropene-1, trans-1-nitropropene-
1 and C,C,N-triphenylnitrone were prepared according to the methods described in the literature [12-14].
[2+3] Cycloaddition of Triphenylnitrone with 3-Nitropropene-1. The reaction was carried out in the
dark. A mixture of 3-nitropropene-1 (3.48 g, 0.04 mol) and C,C,N-triphenylnitrone (1.09 g, 0.004 mol) was
stirred at room temperature for one week. Then the excess of nitroalkene was evaporated in vacuo, and the
residue was separated by semipreparative HPLC. Evaporation of the eluent from the obtained fractions gave
5-nitromethyl-2,3,3-triphenylisoxazolidine (3) and 5-methyl-4-nitro-2,3,3-triphenylisoxazolidine (4) with RT 9.8
and 16.3 min, respectively. The products were recrystallized from cyclohexane. Thus, 0.22 g (15%) of 3
(mp 107-109°C) and 1.08 g (75%) of 4 [mp 99-100°C (decomposition)] were obtained.
5-Nitromethyl-2,3,3-triphenylisoxazolidine (3). IR spectrum, ν, cm-1: 1557, 1366 (NO2), 1180, 958
(azolidine ring), 760, 695 (C6H5). 1H NMR, δ, ppm (J, Hz): 2.46 (1H, dd, J = 8.8, J = 13.2, H-4); 3.79 (1H, dd,
J = 12.8, J = 13.2, H-4); 4.10 (1H, dd, J = 5.0, J = 12.8, CH2NO2); 4.81 (1H, dd, J = 7.7, J = 12.8, CH2NO2);
5.25 (1H, m, J = 8.8, J = 12.8, J = 5.0, J = 7.7, H-5). Mass spectrum, m/z (Irel, %): 360 [M]+ (39), 257
[(C6H5)2C=NC6H5]+ (11), 253 [M–C6H5NO]+ (50), 180 [(C6H5)C≡NC6H5]+ (58), 91 [C6H5N]+ (100). Found, %: C
73.46; H 5.52; N 7.81. C22H20N2O3. Calculated, %: C 73.32: H 5.59; N 7.77.
5-Methyl-4-nitro-2,3,3-triphenylisoxazolidine (4). IR, ν, cm-1: 1559, 1369 (NO2), 1181, 956, 925
(azolidine ring), 780, 693 (C6H5). 1H NMR, δ, ppm (J, Hz): 1.66 (3H, d, J = 6.4, CH3); 5.18 (1H, dq, J = 8.1, J =
6.4, H-5); 5.75 (1H, d, J = 8.1, H-4). Mass spectrum, m/z (Irel, %): 360 [M]+ (40), 273 [M–O2NCHCHCH3]+ (36),
257 [(C6H5)2C=NC6H5]+ (22), 180 [(C6H5)C≡NC6H5]+, 91 [C6H5N]+ (100). Found, %: C 73.41; H 5.58; N 7.78.
C22H20N2O3. Calculated, %: C 73.32; H 5.59, N 7.77.
[2+3] Cycloaddition of Triphenylnitrone with trans-1-Nitropropene-1. A mixture of 1-nitropropene-
1 (3.48 g, 0.04 mol) and C,C,N-triphenylnitrone (1.09 g, 0.004 mol) was shaken at room temperature for 15 min.
When the reaction was completed, the excess of nitroalkene was evaporated to dryness in vacuo. The residue
was tested by HPLC and then recrystallized from cyclohexane. In this manner 1.38 g (96%) of 5-methyl-4-nitro-
2,3,3-triphenylisoxazolidine; mp 99-100°C (decomp.) was obtained.
The authors acknowledge the financial support of these studies by the Polish State Committee (grant
C-2/230/DS/2005).
REFERENCES
1.
2.
R. Jasinski and A. Baranski, Polish J. Chem., 80 (No.9) (2006).
A. Baranski and R. Jasinski, Khim. Geterotsikl. Soed., 1670 (2001). [Chem. Heterocycl. Comp., 37, 1518
(2001)].
3.
4.
A. Baranski, Khim. Geterotsikl. Soed., 840 (2000). [Chem. Heterocycl. Comp., 36, 738 (2000)].
R. Jasinski, A. Ciezkowska, A. Lyubimtsev, and A. Baranski, Khim. Geterotsikl. Soed., 243 (2004).
[Chem. Heterocycl. Comp., 40, 206 (2004)].
1336

5.
I. A. Motorina and L. A. Sviridova, Khim. Geterotsikl. Soed., 829 (1992). [Chem. Heterocycl. Comp.,
28, 689 (1992)].
6.
7.
8.
P. N. Confalone and E. M. Huie, Org. React., 36, 46 (1988).
A. Baranski, G. A. Shvekhgeimer, and N. I. Kirillova, Polish J. Chem., 54, 23 (1980).
V. Yu. Baskov, T. Urbanski, M. Witanowski, and L. Stefaniak, Tetrahedron., 20, 1519 (1964).
G. Hesse and V. Jäger, Liebigs Ann. Chem., 740, 85 (1970).
9.
10.
11.
12.
13.
14.
R. Sustmann R., Tetrahedron Lett., 717 (1971).
I. Fleming, Frontier Orbitals and Organic Chemical Reactions, Wiley, Chichester, 1976.
V. Meyer and P. Askenazy, Ber., 25, 1701 (1892).
M. Stadinger, M. Miescher, Helv. Chim. Acta., 2, 568 (1919).
G. D. Buckley, C. W. Scaife, J. Chem. Soc., 1417 (1947).
1337