Group III sesquioxides oxides are distinguished from other wide band gap semiconductors by the fact that they can be efficiently n-doped despite a wide band gap that ranges from 2.7 eV for In2O3 over 4.8 eV for Ga2O3 to 8.9 eV for Al2O3. Possible fields of applications include unipolar devices like field effect devices, switching memories, intersubband devices, and UV photodetectors. Full exploitation of these properties requires band gap engineering of binary compounds through formation of solid solutions of these oxides which allows the tuning of physical properties over a wide range. The expected high band offsets between Ga2O3 and Al2O3 for example hold the promise for quantum cascade structures at telecommunication wavelengths between 1.3 µm and 1.5 µm. The formation of solid solutions in group III sesquioxides is, however, challenging since at thermodynamic equilibrium the binary compounds crystallize in different structures (cubic, rhombohedral, and monoclinic) and exhibit significant lattice mismatch ranging between 0.03 (between Al2O3 and Ga2O3) and 0.28 (between Al2O3 and In2O3). The oxygen coordination of the metal atoms in the binary alloys is either octahedral, or mixed tetrahedral and octahedral. Moreover, the stability of the phases essentially depends on the strain state and temperature. Recent work on heteroepitaxial growth of Ga2O3, for example, reports on coherent growth of alpha Ga2O3 on sapphire. Even though these results are promising, a thorough experimental study and a predictive model that accounts for phase stability as dependent on composition and strain in binary and mixed sesquioxides does not exist yet.
State of the art
In conventional semiconductors and semiconductor alloys that crystallize in the zincblende und wurtzite lattice, systematic theoretical and experimetal work on phase stability and order/disorder transitions has been performed in the 1980s and 1990s. Predictive approaches were based on non-classical structural coordinate scales by Philips, i.e. homopolar and heteropolar dielectric bandgaps, and on orbital radii. Orbital radii were obtained as linear combinations from screened nonlocal atomic pseudopotentials. Zunger and his colleagues determined orbital radii from ab-initio local density calculations and applied these successfully to determine the favourable structure of AB compounds. Later they accounted for an effective orbital iconicity to rationalize the subtle differences that stabilize AB compounds either in the zincblende or wurtzite phase. By doing this, they were able to unravel simple chemical trends within homological series in a Pauling-esque manner. In sesquioxides simple pseudopotential approaches are not sufficient for this kind of reasoning. Only very recently, Umemoto pointed out that d electron occupancy plays a fundamental role in phase relations in transition-metal sesquioxides. Sabino et al. explained the stabilization of the monoclinic Gallia structure of Gallium Oxide, unexpected from a simplistic point of view of atomic radii by hybridization of Ga d states with O 2s states, which is maximized by the presence of fourfold cation sites. Starting from simple phenomenological concepts such as weighted average bond lengths and effective coordination numbers to get a clue on the phase stability they finally found that it is the Hartree exchange and correlation energies of filled Ga 3d and O 2s states that lead to a gain in energy.
While preliminary work on binary compounds exists, solid solutions of Ga2O3, AlGa2O3 and In2O3 have been studied only very scarcely. This holds for both theory and experiment, for solubilities, thermodynamic stability of the phases as well as for physical properties. The studied materials are mainly polycrystalline or sintered powders and studies are based on X-ray diffraction, dispersive x-ray spectroscopy and Raman spectroscopy but lack detailed structural investigations. Even rarer is theoretical work on ternary solid solutions of group III sesquioxides. Theoretical work by Maccioni et al. indicates low solubility of In in Ga2O3 and preferred incorporation into octahedral sites. They show, based on density functional theory, that band offsets are highly dependent on strain.
With respect to the applications, a fundamental understanding of phase formation is highly desirable. Within this project we focus on fundamentals and applications of ternary solid solutions of Ga2O3, AlGa2O3 and In2O3.
Goals of the project
The project phase formation in group III sesquioxides performed by the groups of M. Albrecht at Leibniz Institut für Kristallzüchtung and C. Koch at Humboldt University performs structural characterization of strained and relaxed alloys. The tasks comprise
Methods applied comprise
Collaboration with partners in the project:
Charlotte Wouters is originally from Belgium, where she studied physics at KU Leuven. During her master’s, she specialized in nuclear physics, but after her studies decided she wanted to continue in the direction of solid state physics. To her, taking part in the GraFOx project is a great opportunity to work in an exciting, ever-growing, and application-oriented research field.
Leibniz-Insitut im Forschungsverbund Berlin e.V.
10117 Berlin, Germany
The Leibniz ScienceCampus GraFOx is a network of two Leibniz institutes, two universities and one institute of the Max Planck Society. The Network is based in Berlin, Germany.