Irani, Rowshanak: Model Systems for Understanding Charge Transfer at Semiconductor/Catalyst Interfaces = Modellsysteme zum Verständnis des Ladungstransfers an Halbleiter-Katalysator-Grenzflächen. , TU Berlin, 2020
Open Accesn Version
In the first part, we investigated the role of an atomic layer deposited (ALD) MnOx co-catalyst and its thickness on the photoelectrochemical performance of BiVO4 photoanodes. The modified MnOx/BiVO4 sample with a thickness of ~4 nm shows the highest photocurrent. X-ray photoelectron spectroscopy (XPS) studies reveal that addition of MnOx results in a modification of the surface band bending of BiVO4, therefore, reducing surface recombination. At the same time, increasing the thickness of MnOx beyond the optimal 4 nm introduces shunting pathways, as shown by energy dispersive x-ray scanning transmission electron microscopy (EDX-STEM) and redox electrochemistry. This cancels out the band bending effect, which explains the observed photocurrent trend. In the second part, we attempted to shift the valence band position of the BiVO4 by the incorporation of nitrogen. Two different approaches were investigated: modification of the precursors for the spray pyrolysis recipe and post-deposition nitrogen ion implantation. The incorporation was successful, but nitrogen was present in the molecular form (i.e., N2). Density functional theory (DFT) calculations confirm the thermodynamic stability of the incorporation and suggest that N2 coordinates to two vanadium atoms in a bridging configuration. This forms an additional intra-band state under the conduction band minimum, which results in a shift in bandgap but unfortunately without any improvement in the photoelectrochemical performance. Since the substitution process of oxygen atoms by nitrogen atoms is not thermodynamically favorable for BiVO4 (vide supra), an alternative (oxy)nitride material system with adjustable band position was desired. Thus, we shifted our focus for the photoanode material to the Ta-O-N system with tunable valence band positions. Introducing nitrogen indeed results in a bandgap reduction, which is a direct result of the valence band shift from 3.8 eV below the Fermi level for Ta2O5 to 2.3 eV for TaON and 2.1 eV for Ta3N5. As the next step, the role of a co-catalyst with adjustable valence band positions (Ni-doped MnOx) and its influence at the interface were investigated. The tunability of the valence band positions of both the semiconductor and the co-catalyst allows us to unravel the energetic effects at the interface. Photocurrents for each Ta-O-N photoanode were improved upon the deposition of the co-catalysts. We found that the catalytic activity of the co-catalysts does not play a major role; instead the properties of the interface between the semiconductor and the co-catalyst does. We also found two interesting correlations. First, the photocurrent of the photoanodes scales with the additional band bending as a result of depositing different co-catalysts. Second, a correlation is established between the relative valence band maximum positions of the semiconductor to those of the co-catalysts and the interfacial charge transfer properties between the semiconductor and co-catalyst. Overall, in this dissertation, a better understanding of the charge transfer processes at the semiconductor/co-catalyst/electrolyte interface has been achieved. Our findings show that the interface energetics play a more important role as compared with the catalytic activity of the catalyst in improving water oxidation for semiconductor-catalyst systems. This consideration is highly beneficial in the development of photoelectrodes for water oxidation, which is one of the main bottlenecks in achieving highly efficient solar water splitting.