Semi-empirical MO calculations, EHT, CNDO/2, MINDO/3, and MNDO met hods, were performed on various geometries of n-butane, n-alkyl radical and tetramethylene diracal (triplet) in order to compare eigenvalue and eigenvector properties with those obtain...
Semi-empirical MO calculations, EHT, CNDO/2, MINDO/3, and MNDO met hods, were performed on various geometries of n-butane, n-alkyl radical and tetramethylene diracal (triplet) in order to compare eigenvalue and eigenvector properties with those obtained by STO-3G method. All methods predicted the same relative order of stabilities of various geometries for n-butane; geometrical preferences were found to be dominated by one-electron factor, ${\pi}$-orbital energy changes being more impotant in the semi-empirical methods. The hyperconjugative energy changes accompanying structural changes from $(n-{\sigma}{\ast})_{trans}$ to (n-{\sigma}{\ast})cis were underestimated in the EHT, CNDO/2 and MINDO/3, whereas those were overestimated in the MNDO. The net destabilizing effect of $(n-{\sigma}{\ast})_{trans}$ structure was mainly due to the large internuclear energy involved in the structure. Through-space interaction between $n_1$ and $n_2$ orbitals of diradical caused energy gap narrowing of ${\Delta}E_{sp}$ and ${\Delta}{\varepsilon}={\varepsilon}_0$-${\varepsilon}_{av}$; through-space interaction had opposing effect to that of through-bond interaction. Due to the less severe neglect of differential overlaps in the MNDO, this energy gap narrowing effect appeared amplified in the MNDO. In general orbital properties were found to be reproduced satisfactorily, but eigenvalue properties were not, in all the semi-empirical methods especially when ${\sigma}-{\sigma}{\ast}$ and n-$n-{\sigma}{\ast}$interactions were involved.