GOX 2022 Session EG-TuA: Bulk & Epitaxy II

In this presentation, we present recent work on the development of β-Ga₂O₃ materials and their properties. The talk will include the following topics:

*Coherently strainedβ-(Al_xGa_1-x)₂O₃ thin films on β-Ga₂O₃ Part I: Growth of (001) β-(Al_xGa_1-x)₂O₃ thin films via metal oxide catalyzed epitaxy. In this work, we report on the growth of (001) β-(Al_xGa_1-x)₂O₃ films in molecular beam epitaxy via metal oxide catalyzed epitaxy. Films with Al contents up to 15% were grown and Al content was measured with atom probe tomography. A relationship between Al content and out of plane lattice parameter was determined. Transmission electron microscopy showed no evidence of extended defects in the (001) β-(Al_xGa_1-x)₂O₃ and reciprocal space maps confirmed that the β-(Al_xGa_1-x)₂O₃ films were coherently strained to the (001) β-Ga₂O₃. Sn was also demonstrated to act as a surfactant for (001) β-(Al_xGa_1-x)₂O₃ growth, allowing for high quality, uniform films with smooth morphologies.

*Coherently strainedβ-(Al_xGa_1-x)₂O₃ thin films on β-Ga₂O₃ Part II - composition determination. We derive the relationships between lattice parameters for β-(Al_xGa_1-x)₂O₃ and Al content x assuming the β-(Al_xGa_1-x)₂O₃ is coherently strained to β-Ga₂O₃. The fundamental stiffness tensor of β-Ga₂O₃and stress-strain relationships are used to determine out of plane lattice parameters for (010) and (001) β-(Al_xGa_1-x)₂O₃. Additionally, transformation of the stiffness tensor allows for derivation of similar relationships for (100) β-(Al_xGa_1-x)₂O₃. For all three orientations, the relationships between peak spacing for β-(Al_xGa_1-x)₂O₃ and β-Ga₂O₃ peaks in HRXRD and Al content x are calculated.

*We describe two recent ultrafast optical pump probe experiments that have determined the electron-phonon scattering time - 4.5 fs for electron-polar optical phonon scattering. These experiments also determined the energy separation of the CBM to the first side valley: 2.6 eV [Marcinkevicius et al., Appl. Phys. Lett. 118, 242107 (2021)]. In a separate ultra-fast pump probe spectroscopy study, the time scale for the formation of polarons (from optically generated free holes) was determined: 0.5 to 1.1 ps [[Marcinkevicius et al., Appl. Phys. Lett. 116, 132101 (2020)].

Epitaxial growth of β-Ga₂O₃ with superior crystal quality has been achieved on different crystal orientations such as (100), (010) and (-201) via plasma-assisted molecular beam epitaxy (PAMBE)^[1]. So far, most of the research has been performed on (010) substrates. However, investigation on (010) substrates has shown that (110) facets are revealed the chevron consistent features in RHEED studies, which indicates (110) is a natural plane in β-Ga₂O₃^[2]. Figure 1 shows atomic models of (110) and (010) planes projected along [001] direction.

Unintentionally doped (UID) β-Ga₂O₃ epitaxial films were grown on (110) substrates by PAMBE while (010) substrates were co-loaded as growth reference. The temperatures of the substrates were kept at 600 ℃ and 700 ℃. To optimize the growth condition, the Ga fluxes were changed from 3.0×10^-8 Torr to 2.5×10^-7 Torr which were measured by beam equivalent pressure (BEP). Prior to the growth, oxygen polishing and Ga polishing were performed to remove the residual impurities from the surfaces. The film thickness was determined by measuring high-resolution X-ray diffraction (HRXRD). The surface morphology of the epitaxial films was measured by atomic force microscopy (AFM). Figure 2 shows the RHEED pattern of (110) and (010) substrates after Ga polishing. Streaky patterns were observed from the surface of (110) substrates, which indicates atomically flat surface. Conversely, crossed lines (red guideline) corresponding to (110) facets were observed from [001] azimuth on (010) substrate. Figure 3 shows the HRXRD result of the (110) β-Ga₂O₃ epitaxial film. Clear thickness fringes indicate abrupt interface between β-Ga₂O₃ and β-(Al_1-xGa_x)₂O₃ spacer layers. Figure 4 shows the growth rate dependence on Ga flux of (010) and (110) substrates at 600 ℃ and 700 ℃. This result suggests that the growth rate is not reduced on the (110) plane compared to (010)^[3]. In the oxygen rich regime, the growth rate increases linearly with Ga flux. In the plateau regime, there was still too low excess Ga flux to have a reduced growth rate. We expect higher Ga flux to yield reduced growth rates. Figure 5 shows the surface morphology of β-Ga₂O₃ films grown at 700 ℃ on (110) and (010) substrates. The RMS values indicate smooth surface morphology was obtained by growing on (110) substrates. Despite the appearance of (110) facets in the growth of (010) β-Ga₂O₃, the (110) plane does not have the tendency to show a well-defined step-terrace structure.

[1] A. Mauze et al., APL Mater. 8, 021104 (2020). [2] P. Mazzolini et al., APL Mater. 7, 022511 (2019). [3] T. Itoh et al., Appl. Phys. Lett. 117, 152105 (2020).

Ultrawide bandgap (UWBG) gallium oxide (Ga₂O₃) represents a promising semiconductor material with excellent chemical and thermal stability. Its wide energy bandgap (4.5-4.9 eV) predicts a breakdown field of 6-8 MV/cm, which is much larger than that of the 4H-SiC or GaN. The key advantages from this material system arise from the availability of high quality scalable bulk substrate and the capability of a wide range of doping.

Metalorganic chemical vapor deposition (MOCVD) growth technique has been demonstrated to produce high quality β-Ga₂O₃ thin films and its ternary (Al_xGa_1-x)₂O₃ alloys. Record charge carrier mobilities approaching theoretical limit were reported from MOCVD grown materials. In this talk, I will discuss the control of background and n-type doping in MOCVD β-Ga₂O₃, and the impact of metalorganic precursor on Ga₂O₃ growth rate and material quality.

Growth and fundamental understanding of (Al_xGa_1-x)₂O₃ with different phases are still limited. The limit of Al incorporation in beta-phase Ga₂O₃ has not been well understood or experimentally verified, although it was predicted up to 60% of Al composition could be incorporated into β-Ga₂O₃. In this talk, MOCVD growth of β-AlGaO with targeted Al composition of > 40%, n-type doping capability as a function of Al composition in (Al_xGa_1-x)₂O₃, and MOCVD growth of different phase AlGaO will be discussed.

Acknowledgement: The authors acknowledge the funding support from the Air Force Office of Scientific Research FA9550-18-1-0479 (AFOSR, Dr. Ali Sayir) and the National Science Foundation (1810041, 2019753, 1755479).