Table of Contents
Fetching ...

Defect analysis of the $β$- to $γ$-Ga$_{2}$O$_{3}$ phase transition

Umutcan Bektas, Maciej O. Liedke, Huan Liu, Fabian Ganss, Maik Butterling, Nico Klingner, René Hübner, Ilja Makkonen, Andreas Wagner, Gregor Hlawacek

TL;DR

This study addresses the defect-driven $β$-to-$γ$ Ga$_2$O$_3$ phase transition induced by ion irradiation, mapping defect formation, phase change, and crystal quality with a multi-method approach. The methodology combines XRD, TEM, RBS/c, DB-VEPAS, VEPALS, and DFT to connect open-volume defects and Ga-vacancy chemistry to the phase transformation. Key findings include local $β$-to-$γ$ transformation evidenced by γ reflections at 222 and 444 with suppression of 111/333 due to apb, a γ-layer thickness of ≈260 nm, and a defect-evolution pattern in which Ga vacancies migrate to tetrahedral sites in the γ phase, followed by neon bubble formation at the highest fluence. DFT-based $3$-site γ occupancy modelling reproduces the observed positron lifetimes and supports a vacancy-driven mechanism that underpins the radiation tolerance of the $β$/$γ$ dual-phase Ga$_2$O$_3$ system.

Abstract

In this study, we investigate the ion-irradiation-induced phase transition in gallium oxide (Ga2O3) from the $β$ to the $γ$ phase, the role of defects during the transformation, and the quality of the resulting crystal structure. Using a multi-method analysis approach including X-ray diffraction (XRD), transmission electron microscopy (TEM), Rutherford backscattering spectrometry in channeling mode (RBS/c), Doppler broadening variable energy positron annihilation spectroscopy (DB-VEPAS) and variable energy positron annihilation lifetime spectroscopy (VEPALS) supported by density functional theory (DFT) calculations, we have characterized defects at all the relevant stages before, during, and after the phase transition. Reduction in backscattering yield was observed in RBS/c spectra after the transition to the $γ$ phase. This is corroborated by a significant decrease in the positron trapping center density due to generation of embedded vacancies intrinsic for the $γ$-Ga2O3 but too shallow in order to trap positrons. A comparison of the observed positron lifetime of $γ$-Ga2O3 with different theoretical models shows good agreement with the three-site $γ$ phase approach. A characteristic increase in the effective positron diffusion length and the positron lifetime at the transition point from $β$-Ga2O3 to $γ$-Ga2O3 enables visualization of the phase transition with positrons for the first time. Moreover, a subsequent reduction of these quantities with increasing irradiation fluence was observed, which we attribute to further evolution of the $γ$-Ga2O3 and changes in the gallium vacancy density as well as relative occupation in the crystal lattice.

Defect analysis of the $β$- to $γ$-Ga$_{2}$O$_{3}$ phase transition

TL;DR

This study addresses the defect-driven -to- GaO phase transition induced by ion irradiation, mapping defect formation, phase change, and crystal quality with a multi-method approach. The methodology combines XRD, TEM, RBS/c, DB-VEPAS, VEPALS, and DFT to connect open-volume defects and Ga-vacancy chemistry to the phase transformation. Key findings include local -to- transformation evidenced by γ reflections at 222 and 444 with suppression of 111/333 due to apb, a γ-layer thickness of ≈260 nm, and a defect-evolution pattern in which Ga vacancies migrate to tetrahedral sites in the γ phase, followed by neon bubble formation at the highest fluence. DFT-based -site γ occupancy modelling reproduces the observed positron lifetimes and supports a vacancy-driven mechanism that underpins the radiation tolerance of the / dual-phase GaO system.

Abstract

In this study, we investigate the ion-irradiation-induced phase transition in gallium oxide (Ga2O3) from the to the phase, the role of defects during the transformation, and the quality of the resulting crystal structure. Using a multi-method analysis approach including X-ray diffraction (XRD), transmission electron microscopy (TEM), Rutherford backscattering spectrometry in channeling mode (RBS/c), Doppler broadening variable energy positron annihilation spectroscopy (DB-VEPAS) and variable energy positron annihilation lifetime spectroscopy (VEPALS) supported by density functional theory (DFT) calculations, we have characterized defects at all the relevant stages before, during, and after the phase transition. Reduction in backscattering yield was observed in RBS/c spectra after the transition to the phase. This is corroborated by a significant decrease in the positron trapping center density due to generation of embedded vacancies intrinsic for the -Ga2O3 but too shallow in order to trap positrons. A comparison of the observed positron lifetime of -Ga2O3 with different theoretical models shows good agreement with the three-site phase approach. A characteristic increase in the effective positron diffusion length and the positron lifetime at the transition point from -Ga2O3 to -Ga2O3 enables visualization of the phase transition with positrons for the first time. Moreover, a subsequent reduction of these quantities with increasing irradiation fluence was observed, which we attribute to further evolution of the -Ga2O3 and changes in the gallium vacancy density as well as relative occupation in the crystal lattice.
Paper Structure (7 sections, 5 equations, 21 figures, 4 tables)

This paper contains 7 sections, 5 equations, 21 figures, 4 tables.

Figures (21)

  • Figure 1: Structure analysis of $\beta$--Ga2O3 irradiated with different fluences of 140keV Ne$^+$. (a) xrd patterns of the virgin and irradiated samples with low fluences, (b) xrd patterns of the irradiated samples with high fluences. (c) bfstem image of the converted layer on top of pristine $\beta$--Ga2O3. (d) hrtem image (right) and fft (left) of the $\gamma$ layer close to the interface along [101] zone axis. The diffraction spots highlighted by the blue circles correspond to the $\gamma$--Ga2O3 111 reflections. (e) hrtem image and fft pattern of the $\beta$ region close to the interface along [010] zone axis.
  • Figure 2: rbsc spectra for $\beta$--Ga2O3 irradiated with different fluences. The irradiated area is converted from $\beta$--Ga2O3 to $\gamma$--Ga2O3 for 3.5e16 and higher fluences. (a) rbsc spectra obtained after aligning to the [-201] $\beta$--Ga2O3 channeling direction. (b) rbsc spectra obtained after the phase transformation and realignment to the [111] $\gamma$--Ga2O3 channeling direction. The dashed lines in the sample configuration inset are only a sketch of the lattice planes and do not represent the actual misalignment. Of note, the depth scale is calculated for gallium atoms.
  • Figure 3: dbvepas and vepals results for Ne$^+$-irradiated samples with different fluences. (a) and (c) show the S-parameter as a function of the positron implantation energy, (b) and (d) show the positron lifetime as a function of the positron implantation energy. The colored transparent lines in (a) and (c) represent the fitted S-parameter curves generated from vepfit for each sample. The dashed lines in (b) represent theoretically calculated lifetimes for various defect configurations in $\beta$--Ga2O3 Karjalainen2020, while the violet and pink bar correspond to lifetimes of the different defect configurations as interpreted by Tuomisto for Fe-doped $\beta$--Ga2O3 Tuomisto2023. The dashed lines in (d) represent theoretically calculated lifetimes for various defect configurations in $\gamma$--Ga2O3 (see table \ref{['tab:Lifetime']}).
  • Figure 4: vepals lifetimes and relative intensities as a function of irradiation fluence obtained at a positron energy of 6keV or in depth of 110nm. (a) Positron lifetimes of $\tau_1$ (fast annihilation pathway) and $\tau_2$ (slow annihilation pathway) as a function of irradiation fluence. (b) Relative intensities of the short positron lifetime $\tau_1$ (I$_1$) and long positron lifetime $\tau_2$ (I$_2$) as a function of irradiation fluence.
  • Figure 5: xrd data for the ($\overline{2}01$)-oriented $\beta$--Ga2O3 virgin sample (red) and sample irradiated with 140keV Ne$^+$ using a fluence of 3.5e16 (blue). While the $\gamma$ 222 and $\gamma$ 444 reflections are observed, the expected $\gamma$ 111 and $\gamma$ 333 reflections are missing.
  • ...and 16 more figures