Radio frequency sputter deposition of the transparent conducting wide band gap oxide Ga2O3

Abstract

Transparent conducting wide band gap oxides can be used in a variety of promising applications, e. g., thin film solar cells or transparent optoelectronic devices. Exemplary, a Cu2O-based solar cell will represent a sustainable and non-toxic technology with the potential for affordable costs, if the chosen window layer materials also are sustainable and abundant. One suitable window layer material is Ga2O3 due to the favorable conduction band position relative to that of Cu2O even though the growth of high quality material still is considered as rather difficult. However, by variation of deposition parameters, e. g. gas flows, sputtering power or substrate temperature, it is possible to specifically tailor the quality and properties of a deposited layer. Of major interest for commercialization of any technology is the capability for mass production. Thus, the conventional as well as pulsed-mode rf sputter deposition are favorable methods as they allow for industrial upscaling due to high deposition rates, excellent layer uniformity on large-area substrates and ease of automation. However, the latter is a new methodology whose parameter correlations and impacts have to be further investigated and explored. Preliminary studies have already shown the potential of this technology for reducing the allocated thermal energy during deposition while maintaining the same layer quality. In two consecutive publications, the feasibility of rf sputter deposition has been investigated, in conventional design as well as in pulsed-mode operation, in order to obtain Ga2O3 thin films of high quality.In the first publication, the results of conventional rf sputter deposition of Ga2O3 by variation of the rf and heating power were discussed. A set of parameters yielding an optimized layer were found, though the quality was not sufficient for the utilization as a window layer material. Still, the thorough analysis of the results revealed a correlation which is very suitable for a fast determination whether the layer is grown stoichiometrically or not. The discrepancy between the refractive index dispersion of a layer and the respective bulk material can give a first impression about the compound s composition.The second publication highlights the utilization of pulsed-mode rf sputter deposition for growing Ga2O3 layers. Slight improvements of layer quality were achieved but, unfortunately, not yet good enough for targeting an application as functional layer in optoelectronic devices. Only a first impression of the complex relationship between conventional and new deposition parameters and their impact on the layer quality was obtained in these experiments. With increasing rf power and pulse duty cycle value, particle energies were accomplished which cannot be reached in conventional sputtering processes. Besides minor improvements in crystalline structure, another feature was identified which is tentatively assigned to growth of metastable gamma-Ga2O3 at the interface between sapphire and Ga2O3 layer. An additional effect is suggested. The presence of negatively charged oxygen ions in the process leads to etching of the surface or penetrating into the layer inducing an increase of the optical band gap. In conclusion, the results have revealed a new relationship for investigating layer stoichiometry only requiring optical measurement data. Furthermore, the Experiment performed underlined the potential of the new growth methodology. The prospective exploration and understanding of the intertwined parameter spaces of conventional and pulsed rf sputtering bears great potential to overcome current obstacles in material research.