AbstractThis thesis presents a computational quantum chemical study on a class of hydrides, the group 13 hydrides. The hydrides of boron, aluminium and gallium are investigated using ab initio quantum mechanical methods. In particular the molecular structures and potential energy surfaces of tetraborane(10), tetralane(10) and tetragallane(10) and their mixed forms have been examined. This work mainly addresses two major issues. The first issue is to investigate appropriate theoretical methods to be used for the system in question. The second issue is to determine the structure and energy of these systems; in particular, to understand theoretically, the non-existence of the bis(diboranyl)- structures which are very close in energy to the arachno- structures experimentally found in the case of tetraborane(10) and gallatetraborane(10). The non-existence of the bis- isomer of the tetraborane structure had been the subject of several previous investigations and the question was not answered.
To enable an objective evaluation of theoretical methods for the study of molecular energies and structures of isomers, different high level ab initio methods in the form of model chemistries are considered. The CBS models, which were available for only first and second row molecules, have been extended to molecules containing third row elements Ga-K, in this work. These are found to produce comparable results to the well-known G2 methods. The accuracy is less than that of the new G3 method, but the CBS methods scale better and are more economical for large molecules. The mean absolute deviation from experiments for the highly accurate methods CBS-Q and CBS-QB3 are 1.11 kcal/mol and 1.08 kcal/ mol, respectively, both lower than for the G2 method. The mean absolute deviation from experiment for the more affordable methods, CBS-4 and CBS-q, are 2.23 kcal/mol and 1.81 kcal/mol respectively. The methods, CBS-4(d), CBS-q(d), CBS-Q(d) and CBS-QB3(d), with the d orbitals included in the correlation space are also investigated, giving results in poorer agreement with experiment. Problems with the CBS extrapolation step when the d orbitals are included in the correlation space are discussed.
Two different approaches have been taken to address the second problem. Firstly, the influence on the structure and energy of the tetraborane isomers were examined, by replacing boron with heavier atoms (Al and Ga). Secondly, the potential energy surfaces of the systems were investigated to find the pathways for the interconversion of the different isomers to the experimentally found isomers.
For the substitution of heavy atoms, aluminium and gallium separately into tetraborane (10) and then gallium into tetraallane, the quantum chemical calculations were performed at G2, G2(MP2), CBS-Q and CBS-QB3 levels of theory. It was found that the incremental replacement of boron atoms by heavy atoms (Al and Ga), considerably lowers the energy of the corresponding bis- isomers compared to the arachno isomers. In case of gallium substitution into tetraborane, the stabilization increases gradually and in tetragallane(l0) the bis- isomer is ∼2 kcal/mol lower in energy that the arachno- isomer. The same trend is seen during aluminium substitution also, but the stabilization magnitude is slightly smaller, and bis-tetralane(10) is ∼0.5 kcal/mol below arachno-tetralane(10). Regarding the replacement of aluminium with gallium in tetralane(10), there was no marked decrease in energy difference between the bis- and arachno forms, but the bis-structure stayed lower in energy compared to the arachno- structure.
In the study of reaction paths for the interconversion of bis-tetraborane(10) to arachnotetraborane( 10), two pathways, a concerted and a step-wise path, were found. The concerted pathway seemed more suitable compared to the stepwise pathway. Clear pathways have been found for the interconversion of isomers of gallatetraborane(10). The potential energy surface of tetraborane(10) and gallaboranes(10) have numerous transition structures due to the exchange of terminal and bridged hydrogens. New penta-coordinated isomers which appear to be promising synthetic targets have also been predicted.
|Date of Award||Aug 2004|
|Supervisor||Brian Salter-Duke (Supervisor)|