The oretical and experimental investigations of structure, reactivity and bonding in some organic systems

Abstract

A theoretical study has been made of some aspects of prototype potential energy surfaces for some simple organic reactions. Addition of prototype electrophiles to simple alkenes has been investigated by means of non-empirical and semi-empirical calculations, within the Hartree-Fock formalism, and the resulting carbonium ions studied. The systems under investigation may be formally considered as being derived from electrophilic addition of H(^+) to ethylene, fluoroethylene and chloroethylene, or of X(^+) (X=F,Cl) to ethylene, and may thus be represented as (C(_2)H(_4)X)(^+), X=H,F and Cl. For the simplest system, C(_2)H(_5)(^+), two basic structures have been considered, the classical ethylcation and the bridge-protonated ethylene. The energies of these species have been minimised with respect to the C-C bond lengths and also, in the case of the latter ion, with respect to the distance of the bridging H from the CC bond centre. Examination of conformational processes in the classical ion has shown a virtual absence of any barrier to rigid rotation about the cc bond. The calculated relative energies of the species has indicated, subject to limitations imposed by the basis set size and partial geometry optimisation, that in the gas phase the classical ion should be ~5.2k cal mole(^-1) more stable than the bridge protonated ethylene. Furthermore calculations along an idealised reaction coordinate representing trans-formation between the two species have indicated the absence of an activation barrier thus suggesting the bridged ion to be the transition state for the scrambling of the hydrogen atoms of the ethyl cation. These results have been compared with mass spectrometric data. The approach of a prototype nucleophile (H(^-)) to ethyl cation has been examined, results suggesting a preferential cis attack. Conformational processes in the 1- and 2- fluoroethyl chlorethyl cations have been examined. The rotational barrier in the 2- fluoroethyl cation has been shown to be very large (10.5k cal mole) and, with the exception of the 2- chloroethyl cation, all the barriers for the substituted ethyl cations have been shown to be dominated by attractive terms. In both the fluoro and chloroethyl systems, predicted ordering of stabilities of cations has been 1- haloethyl > bridge-protonated haloethylene > 2- haloethyl, and idealised reaction coordinates have been constructed relating the ions in the fluoro case, the results predict the total absence of any activation barrier in trans-forming 2- to 1- fluoroethyl cation, whilst, in the analogous chloro case, a small barrier (4.3 k cal mole(^-1)) is predicted. Relative thermochemical stabilities of the ions have been computed, and the stabilising/destabilising effects of halogen substitution in these carbonium ions investigated and compared with experimental data. The halogen bridged 'halonium' ions have been studied, and their total energies minimised with respect to the distance of the halogen atom from the CC bond centre. The calculations have indicated that the fluoronium ion should be of marginally greater stability than the 2- fluoroethyl cation (3.6k cal mole(^-1)) and this has been discussed in the light of published nmr studies of the ionisation of 2-halo-3fluoro 2,3-dimethyl butanes in SO(_2)/SbF(_5). Results for the chloronium ion have indicated that this ion should be considerably more stable (15.8k cal mole(^-1)) than the corresponding 2- chloroethyl cation. Electron Spectroscopy for Chemical Applications (ESCA) has been employed for the measurement of core binding energies in three series of closely related molecules (i) a series of acetyl compounds of general formula CH(_3)COX, X=H, CH(_3), OH, OCH(_3), NH(_2), NHCH(_3), COCH(_3), CO(_2)H, CN and OCOCH(_3). (ii) a series of five membered ring heterocycles. (iii) a series of pyrimidine bases and related compounds. Assignment of core levels has been accomplished in two ways, (i) Direct correlation of measured binding energies with orbital energies derived from SCF calculations, i.e. assuming Koopmans' theorem, (ii) Correlation of shifts in core binding energies with computed electron distributions within the molecule using the charge potential model. In general, assignments based upon the different methods have been found to be in agreement. Furthermore in the case of some members of the pyrimidine series comparison has been possible between charge potential assignments using both ab initio and CNDO/II populations. Agreement between the two sets has been complete.