Photocatalytic water splitting using metal complexes as either photosensitisers (PS) or catalysts has been developed to generate H2 by water reduction and O2 by water oxidation. However, stability and efficacy of the metal complex PS and catalysts are current two major issues for the technology; specifically, stability can be related to dissociation of ligands from metal centre, reduction or oxidation of ligands in the redox environment of water splitting, while efficacy can be determined by factors determining charge transfer process, such as energy levels and interactions between different components. Aiming at stable and effective photocatalytic water splitting, this thesis focuses on the development of cyclometallated Ir(III) complexes as the PS for H2 generation and catalysts for O2 evolution. While cyclometallating ligands can form robust Ir-C covalent bonds, their contribution to the highest occupied molecular orbital (HOMO) enables the energy levels to be tuned through ligand modification and increase the interactions between Ir(III) molecules and electron donors.
The first part of the thesis focuses on photocatalytic H2 generation using Ir(III) complex PS. A new family of Ir(III) complexes ([Ir(C^N)2(dmebpy)]+, C^N=cyclometallating ligands, dmebpy=4,4‛-dimethyl-2,2‛-bipyridyl co-ligand) (PS2.2-PS2.6) bearing a range of aryl quinoline C^N ligands were synthesised and employed to investigate the effect of cyclometallating ligands on the photophysical and electrochemical properties and H2-generation activities. It was found that varying the C^N ligands significantly effected the excited-state reduction potential [E(PS*/PS-)], and hence the reduction of the excited-state PS (i.e. PS*) by electron donor (NEt3 in this study), i.e. PS* + NEt3→ ‛ PS- + NEt3+. Importantly, a potential difference between E(PS*/PS-) and E(NEt-3+/NEt3) of > 0.2 V was found necessary for effective reduction of PS* by NEt3, which was first established in this study. Controlled photocatalytic reactions also suggested that while the cyclometallating ligands were relatively stable, the dissociation of the dmebpy co-ligand from Ir metal centre was responsible for the PS degradation. The effect of N^N co-ligands on the H2-generation activity was then further examined. The inclusion of carboxylate/amide co-ligands in Ir(III) complexes (PS3.1-PS3.4) was found to dramatically decrease E(PS*/PS-), resulting in insufficient PS* reduction by NEt3. In addition, the employment of an electron-deficient phenazine N^N co-ligand in Ir(III) complex (PS3.5) was found to lead to insufficient E(PS/PS-), which is unable to provide the driving force for H2 generation, PS- + H+ PS + 1/2H2. However, the incorporation of cyanofluorenyl in the N^N diimine co-ligand for Ir(III) complex PS3.7 was found to have suitable energy levels for both E(PS*/PS-) and E(PS/PS-), and most importantly, it led to much improved H2-generation stability [i.e. with turnover numbers (TONs) of 2,000 – 7,000, depending on the PS concentrations]. The high TONs can be attributed to the possible self organisation of the Ir(III) complex PS on the Pt catalyst via the cyano anchor group for possible fast charge transfer, rendering the radical anion localised at the N^N diimine rings. In addition, attachment of electron-donating carbazole moiety on the C^N ligand rings of Ir(III) complex (PS3.22) with cyano anchors was further investigated to improve electron separation and transfer. Even though the H2-generation turnover frequencies (TOFs) of the new Ir(III) complex PS (PS3.22) were lower than its parent complex PS3.7 due to the lower E(PS*/PS-), the extended photochemical durability made the final TONs (4,000 – 7,000 depending on the PS concentration) rather high.
The second part of the thesis focuses on catalytic water oxidation to generate O2 using cyclometallated aqua Ir(III) complex catalysts. The effect of the cyclometallating ligands on the catalytic activities were investigated using two families of aqua Ir(III) complexes (4.1-4.3 and 4.4-4.6) based on electron-deficient aryl triazole and electron-rich aryl quinoline ligands, respectively. The results of water oxidation using CAN (ceric ammonium nitrate) as sacrificial oxidant showed that the Ir(III) catalysts with lower oxidation potentials (i.e. aryl quinoline Ir(III) complexes) tended to have higher catalytic activities, i.e. higher reaction rate constants in CAN consumption analysis and turnover frequencies (TOFs) in dynamic oxygen evolution analysis. Further study using a new family of aqua Ir(III) catalysts (5.2-5.4) bearing with hydrophilic functional groups (-OH and -PO3H2) on phenyl quinoline ligands showed that water-solubility was found to play an important role in catalytic water oxidation. Namely, TOFs increased doubly (for -OH) or by an order of magnitude (for -PO3H2). However, the Ir(III) complex catalysts were found unstable in water-oxidation reactions due to the CAN oxidant, and degradation even occurred at the very early stage of catalytic water oxidation (i.e. at 5 – 50 TONs). While the isolation and characterisation of the secondary materials degraded under the CAN driven water oxidation are technically difficult, they were assumed to be some clusters containing Ir, O and oxidised ligands and having the same phase with amorphous IrOx, which were the active species for catalytic water oxidation. Therefore, the higher catalytic activities found in the Ir(III) complexes with lower oxidation potentials and good water-solubility can be ascribed to the relatively easier formation of the secondary active species through the oxidation by CAN. In order to investigate whether the Ir(III) complexes could be stable on photoelectrochemical (PEC) conditions, Ir(III) complexes 5.3 and 5.4 with -PO3H2 anchor groups were adsorbed onto TiO2 electrodes. Under applied bias, these electrodes showed significantly decreased onset potentials and much higher current density than bare TiO2 electrodes toward water oxidation. Under light illumination, photocurrents were observed. Future work should focus on O2 detection and stability of Ir(III) complexes in PECs.