Abstract
D.Phil.
Homogeneous catalysts can give very high activities and selectivities towards desired products; however, they suffer from the difficulties associated with separating the product from catalyst and reaction solvent. Only one two-phase catalyst recovery process has been commercialised in the form of the Rh-catalysed hydroformylation reaction where trisodium triphenylphosphinetrisulfonate is used as ligand. The present study intends to develop ligands that are mobile between two phases and not stationary in one phase, with the view to enhancing catalyst activity. To this end, substituted ligands that may readily protonate and deprotonate by the addition or removal of CO2 are sought, wherein the protonated form of the ligand is designed to be water soluble and the free-base organic soluble. The amidine functional group has been identified as possessing the appropriate pKa value for such proton cycling. With a vision to introduce structural diversity into the amidine functionality and study the effects of variation in the amidine structure on the effectiveness of catalyst switching, arylphosphines bearing acyclic and cyclic amidines were sought. Accordingly, triarylphosphine derivatives were synthesised from P-Cl derivatives, the intermediates of which were converted into their amidine analogues.
Two areas were explored in this project, on the ligand synthesis side:
1. The synthesis of acyclic amidine functionalised phosphines from amino arylphosphines was achieved by condensation with excess dimethyl-acetamide dimethylacetal. The same modification was also successfully applied to Xantphos-type ligands with the aim to achieve high selectivity towards linear aldehyde in hydroformylation reaction while maintaining catalyst recycling.
2. The synthesis of cyclic amidine functionalised phosphines. This approach was found to be successful when making use of nitrile-containing phosphine derivatives. These were converted into their respective amidine analogues by making use of appropriate diamines and AlEt3. In so doing the synthesis has been abbreviated and the approach was shown to be applicable to mono-, di- and tri-substituted triarylphosphines. This method was also applied to Xantphos derivatives.Both of the methodologies described above provided a range of amidine functionalised ligands. Rhodium complexes of these newly synthesised ligands were found to be active catalysts in hydroformylation of 1-octene. Evaluation of catalyst recycling and re-use of such catalytic systems was carried out using aqueous biphasic separation with carbon dioxide induced phase switching. After phase separation and addition of fresh toluene, the complex can be transferred back into the organic phase by bubbling N2.
These systems provide for the hydroformylation reaction of medium to long chain alkenes to be performed in purely organic solvents thereby retaining good reaction rates and selectivities, and only form two phase systems during recycling. Consequently, overcoming mass transport limitations experienced with aqueous biphasic systems.