Speaker
Description
Gallium oxide (Ga$_2$O$_3$) is a highly versatile material with applications in power electronics, optoelectronics, and battery technologies. Among its polymorphs, monoclinic $\beta$-Ga$_2$O$_3$ is the most chemically and thermally stable phase. However, controlling the metastable polymorph phases remains challenging, and fabrication technologies for nanoscale structures are still under development. This study aims to enhance the understanding of polymorph conversion mechanisms and to establish novel fabrication techniques for single-phase polymorph films, buried layers, multilayers, and various nanostructures of Ga$_2$O$_3$.
We investigate $\beta$-Ga$_2$O$_3$ samples irradiated with different ions and fluences, as well as $\alpha$- and $\kappa$-Ga$_2$O$_3$ thin films. Broad beam (BB) ion irradiation was employed to induce phase transformations in the near-surface region. The irradiated samples were characterized using transmission electron microscopy (TEM) and X-ray diffraction (XRD) to analyze structural changes. Complementary experiments using Positron Annihilation Lifetime Spectroscopy (PALS) and Doppler Broadening Variable Energy Positron Annihilation Spectroscopy (DB-VEPAS) provided insights into defect types and concentrations.
Our results reveal the evolution of defect types and densities based on DB-VEPAS and positron lifetime measurements. During the phase transition from $\beta$- to $\gamma$-Ga$_2$O$_3$, a significant reduction in positron trapping sites is observed, indicating a decrease in defect density in the newly formed $\gamma$-Ga$_2$O$_3$ layer, consistent with the high radiation hardness of Ga$_2$O$_3$. Additionally, we employed Neon-based helium ion microscopy to investigate the minimal achievable polymorph feature size, producing $\gamma$-Ga$_2$O$_3$ lines as narrow as \qty{20}{nm}. The lateral dimensions and shape of the observed $\gamma$ regions are in excellent agreement with SRIM and TRIDYN collision cascade simulations, directly confirming the defect densities and ion fluences employed in both the FIB patterning and the PAS characterisation.
This work is supported by the m-era.net project GoFIB and funded by the Saxonian government. Additional support from the COST Action CA19140 FIT4NANO is gratefully acknowledged.