A chemical and thermochemical study of non-observed symmetry allowed reactions

Abstract

A variety of reactions, whilst predicted to be symmetry allowed, are not observed experimentally. Several such non-observed yet allowed reactions have been studied in order to understand them and to possibly bring about, with suitably modified substrates, their observation. For example, although ozone adds to π-bonds, the isoelectronic nitro group does not, due to a high activation energy barrier. This reaction is of the type A⇆B→C. Molecular orbital calculations of a series of 1,3-dipolar species reveal a correlation of the π bond order with their effectiveness in [4 + 2]-additions. Based on this, the nitro group as well as ozone should be considered as poor dipolar agents. The exceptionally high ΔH_f°of ozone (+34.00 kcal mol^-) as compared with that of nitrobenzene (+3.8 kcal mol^-1) could possibly explain the reactivity of the former. This is reflected in an estimated ΔH° value of -45.5 kcal mol^-1 for the ethylene-ozone addition as compared with + 7.5 kcal mol^-1 for the non-occurring ethylene-nitrobenzene addition. Thermochemical calculations demonstrate that the nitrogroup is more prone to undergo addition as a 2π partner than as a 1,3-dipolar reagent. The thermal, nitro group-olefin addition has been examined using a variety of electrophilic nitrobenzenes with tetracyclone as the common substrate. The extent of cyclo-addition was assessed on the basis of the isolation of 2-benzoyl-3,4,5-triphenylfuran. 4-Nitro-4'-methyl diphenyl sulfone and 4,4'-dinitro diphenyl sulfone have been identified as efficient reagents for these reactions. The former compound has been found to add to even non-conjugated olefins. The clean thermal transformation of crystalline 7,7-dimethyl-bicyclo(2.2.1]heptane-1-nitrile oxide to the corresponding isocyanate is an example of the observable A⇆B→C change. In this case, the oxazirine corresponding to B, arising from conrotatory cyclisation, undergoes irreversible rupture (B→C). Such transformations are not observed with the stereochemically more favoured nitrones since these belong to the type A⇆B. The Π⁴s + Π²s addition of carboxylates to olefins would lead to a carbon base at the expense of an oxygen one. This is clearly unfavourable and is not observed. On the other hand, the reverse of this reaction, namely, the fragmentation of ketal conjugate bases to carboxylates would be expected to be facile. This is indeed so and the process has been shown to be concerted. Ketals from ρ-nitrobenzaldehyde are shown to undergo fragmentation with 1,5-diazabicyclo[3.4.0]nonene-5 (DBN) to olefins. The carboxylate-olefin addition belongs to the general category Π + A^-⇆B^-, that encompasses cycloadditions, electrocyclic reactions and sigmatropic shifts. An equation ΔG°=[1.4(ρ K_a^BH - ρK_a^AH) + X] kcal mol^-1, wherein X is dependent on the reaction type has been derived for predicting the ΔG°for these Π + A^-⇆^-B processes. Of practical interest is the development of perturbed Π+ A₁^-B₁^- systems that are predicted to occur in the forward direction, where the parent type Π+A^-⇆B^- is not favoured. Although no worthwhile synthon may be devised that might simulate either the carboxylate-olefin addition, or the enolate-olefin addition, the perturbation of the allyl anion is possible, principally by substitution with electron-withdrawing substituents at 2 position. Thus, the anions derived from propenes that carry -CN, -NO₂, -SO₂Ph, -SO₂CF₃, -COO^tBu, -CO^tBu and -N⁺₂ as 2-substituents are predicted to add to olefins thereby providing novel routes to functionalised cyclopentanes. In preliminary investigations, the 2-nitropropene anion unit, generated via the collapse of nitrocyclopropane-9, 9'-fluorene conjugate base, underwent rapid nitrite elimination leading to allene, thus precluding cycloaddition to added styrene. A preference for [2,3] shift over [3,3] is exhibited by oxime-O-allyl ethers as predicted by the above equation. Several observed and nonobserved symmetry allowed reactions have been rationalised using the ρ K_a concept. Highly unfavourable Π + A^-⇆^-changes can be. in principle, realised via the strategy Π+ A^-⇆B^- →C^-. This is possible by appropriate incorporation of weak bonds such as N-O, S-S and O-O with the B^- unit. Endeavours to illustrate this with an unusual oxadiazolinone carrying a particularly fragile N-O bond and with 3-lithiomethyl benzisoxazole are presented.