Compositionally Complex Titanium Niobium Oxynitride Materials for Solar-Driven Photochemistry

Abstract

The focus of this thesis is to examine co-incorporation of cationic and anionic dopants simultaneously as a strategy for increasing visible light absorption in TiO2. Co-incorporation has been well-studied theoretically as a viable mechanism for introducing low-energy transitions into the stable TiO2 host lattice; the history and fundamental motivation for co-incorporation is examined here in detail. However, experimental preparations of co-doped and co-incorporated materials remain limited, and a general synthetic method that establishes rigorous control over both cationic and anionic dopant stoichiometry has yet to emerge. In this work, TiO2 co-incorporated with the charge-compensating pair Nb5+/N3- to form titanium niobium oxynitride (TiNbON) is prepared by three synthetic routes and its photochemical properties investigated. First, TiNbON with 25% Nb is prepared by a traditional hydrolytic sol-gel method followed by high-temperature ammonolysis. The resultant material is modified with 1 wt % of RuO2 and evaluated as a photochemical water oxidation catalyst in a solution of NaIO3 sacrificial oxidant. Under 6 times the intensity of solar illumination (6 suns), TiNbON-25 produces oxygen via water oxidation at the rate of ~100 μmol h-1 g-1. Water oxidation occurs regardless of excitation wavelength, though diminished proportionally to the material absorptivity at each wavelength. The rate constant exhibits a zero-order dependence on iodate. Finally, oxygen evolution experiments in 18O-labeled water produce primarily 36O2, suggesting that the dominant pathway for oxygen evolution is the coupling of two water molecules. Secondly, we undertake a new synthetic preparation for TiNbON materials by adapting the urea-glass synthesis for metal nitrides. The reaction produces micron-sized particles of mixed-metal titanium niobium nitride for a variety of niobium contents. These materials are then oxidized to form photoactive anatase/rutile TiNbON that absorbs visible light of λ ≤ 550 nm. Contrary to previous results, the optimized material contains 8% Nb of total metals and degrades methylene blue with a first-order Langmuir-Hinshelwood rate constant of 0.704 h-1 under 5 suns illumination (0.595 h-1 when restricted to λ ≥ 400 nm). Full compositional analysis of TiNbON-5 reveals an empirical formula of Ti0.92Nb0.08O1.97N0.03. A further refinement of the urea-glass synthesis for TiNbON is presented in which the alkaline-earth cation Ca2+ is added to the synthesis to slow the rate of ammonia release during the initial heating step. TiNbON-5 prepared by this method exhibits superior visible light absorption to 600 nm. Tauc analyses of optical spectra suggest a direct band gap. TiNbON-5 produced by this route exhibits a superior first-order Langmuir-Hinshelwood rate constant of 1.785 h-1 under 5 suns solar irradiation. It is hypothesized that the material’s improved properties are due to increased nitrogen content. Finally, a preparation for layered transition metal tungstate/tungsten oxide (MWO4/WO3) photoanodes for photoelectrochemical water oxidation is presented. This strategy seeks to improve on the fundamental shortcomings of the well-studied photoanode material WO3, namely its instability and propensity to participate in side reactions, by adding a more stable and chemoselective interface layer. The interface layers chosen for this study are CuWO4 and Cu0.95Ni0.05WO4. The photoelectrochemical reactivity of MWO4/WO3 electrodes does not change appreciably with respect to bare WO3. Furthermore, we introduce α-thujone, a water-soluble organic molecule with the potential to report on radical chemistry near an electrode surface. Experiments including α-thujone suggest decreased radical prevalence near the surface of MWO4/WO3 compared to WO3.