Session Index

β-Ga₂O₃ thin films were grown on Al₂O₃(0006) substrates via MOCVD. Increasing growth temperature improved crystallinity, surface smoothness, and vertical grain alignment, as confirmed by XRD, SEM, AFM, and TEM. These findings highlight temperature’s key role in optimizing film quality for optoelectronic applications.

  Preview abstract

Epitaxial β-Ga₂O₃ films were grown on c-plane Al₂O₃ by MOCVD to study the effect of TEGa flow rate on structural properties. XRD and TEM confirmed monoclinic single-crystalline growth along the <-201> orientation. Lower TEGa flow yielded smoother surfaces and growth rates of 6.14–13.5 nm/min.

  Preview abstract

Several diamond-based-devices require producing, on a diamond substrate, a doped layer with a gradient of doping. The distortion of this layer can be important for the performance of the device.
We investigated, by using Rocking Curve Imaging (RCI), the distortion of a (100) platelet shaped, 250 µm thick, diamond crystal with two deposited boron-doped layers. This doping varies over 3.7 µm, is associated with a higher lattice parameter for the layer. This structure allows helping accommodating the variation of lattice parameter. Similar features were observed for other samples with 2 or 4 layers.




  Preview abstract

Heteroepitaxial β-Ga2O3 films were grown by MOCVD on exact c-plane, 6° A-off-cut, and 6° M-off-cut sapphire. Gate-recessed enhancement-mode MOSFETs fabricated on the off-cut substrates show superior performance: the specific on-resistance drops from 1.85 kΩ·mm (c-plane) to 0.45 kΩ·mm (A-6°) and 0.70 kΩ·mm (M-6°); breakdown voltages reach 434 V, 584 V, and 442 V, respectively. The improvement is attributed to step-flow growth that guiding an ordered crystal arrangement during growth. These results demonstrate off-cut-driven substrate engineering as an effective route toward high-efficiency β-Ga2O3 power devices.

  Preview abstract

We report a non-recess etch fabrication method for β-Ga2O3 MOSFETs that preserves the as-grown surface to eliminate plasma-induced damage. By increasing growth temperature of the N+/N- layers, silicon donor activation was enhanced, yielding higher free-carrier concentration. Devices exhibit increased on-state current, reduced on-resistance, improved field-effect mobility, also simplified the process. This fabrication approach enables high-performance β-Ga2O3 transistors for advanced wide-bandgap power electronics.

  Preview abstract

Gallium oxide (Ga₂O₃), a promising ultra-wide bandgap (UWBG) semiconductor with a bandgap of ~4.9eV. In this study, we investigate the effect of different Schottky metal stacks on the electrical characteristics of Ga₂O₃-based Schottky barrier diodes (SBDs). By comparing pure Pt, Ni/Pt, and Ni/Au contacts, we examine the impact of adhesion, and potential interfacial reactions on turn-on voltage, leakage current, and on/off ratio. Devices with Pt contacts exhibit the best leakage suppression with an on/off ratio exceeding 10⁷, while Ni/Au structures suffer from severe leakage due to possible interfacial oxidation.

  Preview abstract

In this study, a CO gas sensor based on β-Ga2O3 thin films was fabricated. The sensing performance was evaluated under photoactivation, using a xenon lamp as the excitation light source. Key characterization metrics included current–time response, resistance–concentration profiles, gas response percentages, linearity plots at various concentrations, response and recovery times. The results demonstrate that the β-Ga2O3 thin film exhibits excellent sensitivity and a strong linear correlation with CO concentration. The sensor achieved a rapid response time of 3.5 s; however, the recovery time was relatively long (11.1 s), primarily due to the influence of persistent photoconductivity induced by light illumination.

  Preview abstract

A novel growth method for c-In2O3 was proposed, wherein In2O gas is selectively generated by supplying H2O vapor over In metal in the source zone. The c-In2O3 layer is formed via a reaction between the generated In2O gas and additional H2O gas in the growth zone. Thermodynamic analysis revealed that c-In2O3 growth using In2O and H2O gases is feasible at temperatures above 1000 °C. Based on the thermodynamic results, experimental growth was carried out, and c-In2O3 with (111)- and (100)-oriented domains were successfully obtained on sapphire (0001) substrates at 1000 °C.

  Preview abstract

The Indium-Tungsten-Gallium Oxide thin films are applied to thin-film transistors using radio frequency magnetron sputtering. Silicon dioxide is used as the gate dielectric layer for the IWGO TFTs. T The IWGO TFTs are then subjected to light exposure measurements, showing a light-to-dark current ratio of 5.96 and rejection ratio of 9.98*10^2, confirming that IWGO TFTs can also be used in optoelectronic transistors.

  Preview abstract

Chirality, like the distinction between left and right hands, prevents an object from overlapping with its mirror image. In optics, chiral structures enable selective polarization absorption, producing circular dichroism. Here, we present a high-performance dielectric metasurface of amorphous silicon on fused silica, combining chiral geometry with PB phase modulation to break symmetry between left- and right-handed circular polarization. This optically asymmetric, efficient design shows strong potential for multifunctional metalenses and next-generation devices like smartphone cameras and facial recognition systems.

  Preview abstract

We propose a complex electric field optimization strategy to jointly optimize both amplitude and phase in the design of visible-light amorphous-silicon (a-Si) metalenses, significantly enhancing their efficiency. Numerical simulations and experimental validation on a polarization-independent NA = 0.2 lens demonstrate an approximate 30% improvement in focusing efficiency, highlighting the potential of this method to enable scalable production of high-performance metalenses for advanced optoelectronic, nanophotonic, and AR/VR applications.

  Preview abstract

This work presents a successful implementation of the artificial intelligence (AI) model to inverse predict the input structural parameters of the silicon carbide (SiC) Trench MOSFET. Device structural parameters, such as trench depth, width, and angle, for 2.5kV breakdown voltage (BV) device were predicted successfully. For dataset, parameters of baseline 1.2kV BV device were varied to simulate more than 1600 instances to provide a sufficient dataset for the ML training. Technology Computer-aided Design (TCAD) has been used to simulate the device. The reinforcement learning (RL) based deep learning (DL) method has been implemented to inverse predict the parameters.

  Preview abstract