Conference Agenda

Nowadays, Raman spectroscopy is crucial in studying semiconductors by

providing insights into their structural, vibrational, and electronic

properties and stability under light exposition. This presentation focuses on

studies we conducted on III-Nitride heterostructures for solid-state light

emitters and chalcogenide thin films for photovoltaic applications.

For III-Nitrides, Raman spectroscopy helps analyze crystalline phase, strain,

composition, and defect density, particularly in ion-implanted materials. It

detects disorder-induced vibrational changes and enables the study of

phonon lifetimes, anharmonic effects, and size-dependent phonon

behavior, crucial for optimizing optoelectronic and high-power devices.

In chalcogenide thin films, particularly Sb₂(S, Se)₃, Raman is widely used

but can be affected by in situ surface reactions. These reactions may lead to

phase transformations, resulting in misinterpretation of Raman spectra,

highlighting the need for careful experimental conditions.

Silicon carbide (SiC) diodes are crucial in high-power applications due to their excellent electrical and thermal properties. However, the nonlinear nature of SiC diodes is rarely represented by conventional calibration methods. This paper proposes a neural network-based calibration methodology using simulation data for extracting the ideality factor (n), saturation current (Is), series resistance (Rs), and leakage resistance (Rl), of a diode directly from V–I curves. Implemented as a feedforward network with two hidden layers and optimized hyperparameters, the network takes advantage of the logarithmically scaled training data to enhance the performance in the low-current region. The method shows low mean squared error and mean percentage deviation below 5% for a large operating range and hence can be considered a computationally efficient approach compared to the traditional curve-fitting techniques in SiC diode modelling.

In this work, a novel technique used to create β-Ga2O3 microtubes and nanomembranes, ion-beam-assisted exfoliation, was employed in the fabrication of simple metal-semiconductor-metal (MSM) devices to be tested as field-effect transistors (FET) and UV photodetectors, as MSM photodetectors and as phototransistors. Electrical characterisation showed that these devices can reach very high responsivities and detectivities, up to 4.7×10^4 A/W and 2.7×10^17 Jones, respectively. A rejection ratio above 10^6 between 5.08 eV and 3.4 eV light was also obtained. However, some of these devices exhibit high persistent photoconductivity. In order to better weigh the different mechanisms that are responsible for the high and persistent photoresponses, ion beam induced current (IBIC) measurements were performed on the two metal junctions and on the semiconductor channel using a 2 MeV proton microbeam.

This work describes a stacked diode CMOS sensor using a 0.18 µm process for energy-resolved electron imaging. Inspired by optical sensors that use these devices to detect color based on light absorption depth, this sensor applies the analogous principle to electrons, whose energy deposition also varies with depth in silicon. Optical tests confirmed distinct spectral responses separable via a correction algorithm. Electron beam tests using a 10 keV SEM showed sensitivity dependent on beam current and diode depth. This approach shows promise for integrated, lower-cost energy discrimination in SEM.

Electro-optical modulators, operating via linear and non-linear electro-optical effects are well-established; however, typically, this modulation is performed using high-power grid-supply electricity. In this research, a green, environmentally friendly alternative is investigated, combining the field of triboelectric nanogenerators (TENG) with photonics to produce a precise tribophotonic device, capable of stimulating standard electro-optical modulators (EOMs) at the lower frequency range. Varying levels of TENG hand-tapping stimulation (corresponding to output voltage) evidenced a successful non-linear EOM response resulting from electric field exposure. EOM transmission amplitude reduction at 993 nm by 17.5 % at 154 V applied via the TENG and by 64.7 % at 348 V was observed. Furthermore, this investigation characterizes the developed zinc oxide (ZnO) - polyethylene terephthalate (PET) based TENG, further optimizing base conditions for integration with photonics.

In this work, the electrical characteristics of Mo Schottky contact on n-GaN were investigated post-metal annealing at 850 °C. Barrier Height ΦB and ideality factor η were extracted from the measured forward current-voltage linear fit using the thermionic emission (TE) theory and C-V measurements. The leakage current is directly given by the reverse bias I-V measurements. Results indicate dependency of the electrical characteristics upon annealing. The Mo/n-GaN junctions showed ideal electrical characteristics after annealing at 850 °C with an ideality factor of 1.07, a hard breakdown voltage of 118 V/µm and a leakage current of 6.57mA.cm-2 at -10 V. The specific on resistance RON is very low with a value of 0.1 mΩ.cm². SEM images showed no post-annealing physical degradation of the device.