Computational Chemistry Driven Understanding for Group-13 Catalysts in CO2 Utilisation

Rising levels of CO2 is a global challenge as a dominant contributor to rising temperatures. CO2 utilisation can reduce atmospheric CO2 levels by converting it into useful products. Of industrial importance, is the cycloaddition of CO2 and epoxides to form cyclic carbonates, which are used as lubricants, in synthesis and as fuel additives. The stability of CO2 means the reaction is economically unviable at an industrial scale, therefore, a catalyst needs to be employed. Lewis acid catalysts are excellent for cyclic carbonate synthesis. Among group 13, aluminium is a popular choice, as it has previously shown high activity for CO2 and epoxide cycloadditions. This activity has always been attributed to aluminium’s stronger Lewis acidity, resulting in better CO2 activation. As a result, little attention is given to heavier group 13 metal catalysts. This work gives a thorough understanding of the mechanistic action of aluminium, gallium and indium catalysts utilising density functional theory (DFT), quantum theory of atoms in molecules (QTAIM). A review of density functional theory (DFT) methods was conducted in Chapter 2, along with a theoretical introduction to all the methods used. Bimetallic aluminium salphen is an excellent catalyst for this the conversion of CO2 and epoxides to cyclic carbonates, whilst the monometallic aluminium salphen was reported as inactive due to the formation of an insoluble precipitate. However, experimentally the Whiteoak group showed In[salphen(Br)] to be highly active. A mechanistic study was conducted to understand why the indium catalyst was active over the aluminium. Furthermore, a ligand abstraction step was needed for the CO2 insertion to occur and associated with the poor activity of the [Al(salphen)Cl] catalyst. A mechanistic study on the cycloaddition of 1,2-diphenylaziridine and 1-isopropyl-2-phenyl aziridine with CO2 catalysed by Ga aminotrisphenolate was undertaken. A temperature selective formation of an oxazolidinone or piperazine product was explored using DFT. Mechanisms were elucidated, and the selectivity attributed to the difference in entropic contributions of the reactants. Importantly, a ligand dissociative ring opening step was seen for the first time between an aminotrisphenolate catalyst and aziridine substrate. Finally, a comparative study on how the electronic and steric properties affect catalyst activity was undertaken on 12 ML3 catalysts [M = Al, Ga and In; L = Cl, Br, I, Me*, tBu*, Ph* (*Only used with Ga)]. Electronic and steric catalyst properties were elucidated and analysed through quantitative structure-activity relationship models. Results were analysed to see which descriptors could predict a highly active tris-ligated Lewis acidic catalyst.