Search

Computational modelling was conducted in conjunction with experimental collaborators in the Davies group in Leicester1 into the mechanism of C-H functionalisation at Rh, Ir and Ru complexes. In Chapter 3 C-H activation by [Cp*IrCl2]2 and [Cp*RhCl2]2 dimers was investigated, and an intramolecular acetate-assisted mechanism was characterised. Experimental trends in reactivity were rationalised as reactions with Rh operating under thermodynamic control, while Ir complexes operated under kinetic control due to both less stable C-H activation transition states and more stable cyclometallated products. In order to obtain agreement between experiment and computational results, solvation and dispersion effects had to be accounted for and diffuse basis functions were required. Chapter 4 describes computational studies on functionalisation of rhodacycle complexes formed following C-H activation with alkynes to yield 6-membered ring pyridinium and isoquinolinium products. These proceed via a rate-determining migratory insertion followed by reductive coupling. Experimental results also revealed the possibility of 5-membered ring indenylamine formation and this was computed to occur via acetate-association, imine insertion and protonation by acetic acid (Chapter 5). Selectivity between 5- and 6-membered ring formation was found to be determined by the relative barriers of C-C reductive coupling and C-N bond forming imine insertion steps. The related functionalisation of iridacycles formed via C-H activation was computed to have a selectivity determining migratory insertion and rate determining reductive coupling, while the form of the final heterocyclic Ir complex obtained was under thermodynamic control (Chapter 6). In Chapter 7, Ru-catalysed direct C-H alkylation was investigated computationally. To model these reactions in non-polar aromatic solvents, optimisation was required to be performed in solvent. The most stable reaction profile characterised involved initial C-H activation, followed by an SN2 C-I bond cleavage at an arene s-(C-H) complex with a k2 acetate ligand. C-C reductive coupling then completes the cycle.

In the UK a national black smoke (BS) monitoring network has existed since the 1960's and this historic data could be used for long-term retrospective health studies. However, as sources of pollution changed with the introduction of legislation to control domestic coal burning there was a move in the mid 1990's to the use of monitors designed to measure PMlO. The critical evaluation of data from eo-located BS and PM10 monitors would enable historic BS data to be expressed in terms of PMlO which would be useful in assessing exposure profiles in health studies. The analysis of filters used in the different monitors would also give an indication of the composition of airborne particulate matter sampled by historic and modern techniques. In the study reported here BS and PMlO monitors were co-located at 5 locations throughout the United Kingdom. Haydon Bridge in Northumberland and Strabane in Northern Ireland where the main source of airborne particles is domestic coal burning; Belfast also in Northern Ireland measuring particles from traffic and domestic coal burning; a kerbside site in Sunderland, where traffic emissions are the dominant source of particles and finally a background site, also in Sunderland which will have few sources of combustion related particles. Statistical analysis of the monitoring data showed positive correlations between BS and PMlO at all sites, but that the association between the two was far greater at sites where domestic coal was being burnt. The regression equations derived mean that conversions can be made from historic BS data to PMlO across a range of pollution climates. At sites where data was available for other air pollutants BS showed a much stronger positive correlation with NO, N02 and NOx than did PMlO. This was especially true at the kerbside site indicating that BS may be a better measure of traffic related combustion particles. Analysis of the different filters using ion chromatography, atomic absorption spectroscopy, SEM and x-ray microanalysis showed harmful heavy metals were associated with BS whilst secondary particles, such as nitrates and sulphates, were associated with PMlO. Crustal material and sea salt were also found to be important constituents of PM10.