Development of gallium-oxide (Ga2O3) coatings by non-aqueous sol-gel routes for electronic applications

Abstract

The overall aim of this thesis was to develop a sol-gel gallium oxide (Ga2O3) coating and to study the phase evolution in coating, and to evaluate theoretically the crystal structure and electrical properties of coating by modelling the intrinsic- and doped-Ga2O3 structure using Materials Studio software, and to solid theoretical result, experimental work was performed and compared with the simulation results.

The most stable phase of gallium oxide (β-Ga2O3) is a wide band gap (4.9eV) metal oxide having a wide range of important applications. β-Ga2O3 is thermally and chemically stable at high temperatures and so exhibits very stable operating characteristics over large temperature ranges. The major limitations of this coating arise from it having high resistivity. Monoclinic gallium oxide (β-Ga2O3) is one of the most promising materials for the device applications because of its wide band gap which gives high transparency from the visible into the UV wavelength regions (~260 nm). However, the high electrical resistivity of Ga2O3 coatings is going to be a significant issue making limitation for this coating to be used in optoelectronic devices. Therefore, incorporation of dopants to reduce Ga2O3 resistivity is desirable.

To develop Ga2O3 coating, two types of sols were prepared in this work using gallium isopropoxide as the starting precursor. The first sol (Type-I sol) was prepared via an aqueous route based on the method developed by Yoldas for alumina sol-gel coatings. The other sol (Type-II sol) was prepared via a non-aqueous route involving 2-methoxyethanol (MOE) as the solvent. The principle of coating process was the deposition of these sols onto substrate (glass and quartz) by spin coating followed by drying and then heat treating at elevated temperature. The Type-I sol did not gelate during the deposition process as evidenced by the lack of any visible coating on the substrates. In contrast, the Type-II sol produced obvious coatings, albeit with varying extent of cracking depending on the deposition and heat-treatment conditions.

In another stage of project, the effects of the coating thickness, heat treatment conditions, and substrate type on coating structural evolution were investigated. Phase composition in the Ga2O3 sol-gel coating was studied as function of heat-treatment conditions. The initial deposited phase of gallium oxide transformed to α-Ga2O3 and then to β-Ga2O3 with increasing temperature. At 500°C, α-Ga2O3 phase started to form and upon heating at 900°C, β-Ga2O3 was only stable phase of gallium oxide. Subsequent heat treatment at different heating temperature for 2 h affected the coating behaviour in terms of the amount of cracking. The amount of cracking tended to increase with increasing heating rate and with increasing coating thickness. The choice of substrates for Ga2O3 was studied since it is critical and substantial for making high-quality coatings.

In this work, the effect of adding different dopants on the electrical properties of gallium oxide coatings were investigated theoretically using software called Materials Studio. The Modeling of Ga2O3 showed that the introduction of the Sn and Zn caused the impurity energy level at the bottom of the conduction band. Therefore, the conductivity of the Zn-doped and Sn-doped β-Ga2O3 was improved compared to the intrinsic β-Ga2O3. In order to assess the simulation result and obtain how much the results are close to the practical results, experimental work was carried out by measuring the bandgap of pure β-Ga2O3, Zn3%-doped Ga2O3, and Zn6%-doped Ga2O3. The study indicated that in spite of a deviation in values between the experiment and simulation, the values were considered fit well, and there was a consistency between simulation and experiment results. It was found that the experimental value of the bandgap energy for pure β-Ga2O3 agreed reasonably well with values reported in the literature and the experimental values for the pure and doped coatings were consistently ~2 times higher than the simulated values which suggested that the structural model used to calculate the bandgap energies systematically underestimated the values. This was attributed to limitations in the structural model. Regardless, the structural model was considered reliable for predicting the effects of dopants on selected structural and electronic properties of Ga2O3.