STMicroelectronics (Italy)

Silicon carbide (SiC) is the leading wide bandgap semiconductor for high-power and high-temperature electronics, but the high defect density still limits device performance. This study investigates how inclusions, Basal Plane Dislocations (BPDs), and Threading Screw Dislocations (TSDs) in 4H-SiC substrates affect epitaxial defect formation. Twenty 200 mm SiC wafers were analyzed after epitaxial growth in two industrial Chemical Vapor Deposition (CVD) reactors, one using Trichlorosilane/Ethylene (Reactor A) and the other Silane/Propane (Reactor B). Defects were characterized using Candela (KLA), Altair (KLA), XRTmicron LAB (Rigaku), SICA (Lasertec), and Crossbeam (ZEISS) dual-beam SEM system. Statistical correlation showed that the conversion rate of embedded particles decreases with particle depth and increases with particle size. Reactor A exhibited lower propagation rates, indicating better suppression of substrate-related defects. SEM/FIB-EDX analyses suggested that carbon inclusions generate pits while metallic inclusions induce triangular defects. Dislocation analysis confirmed a strong correlation between TSDs and BPDs with carrots and triangular defects. BPD conversion rates were estimated at about 98.3% (Reactor A) and 99.8% (Reactor B). These results emphasize the importance of substrate quality and buffer layer optimization to minimize defect propagation.

This work reports the successful integration and processing of hundreds of GaN on Silicon wafer lots devoted to 100V Monolithic GaN power device within a standard 8-inch silicon fab primarily dedicated to BCD/CMOS technology production. By addressing key challenges related to gallium cross-contamination and equipment compatibility with thicker GaN on Si substrates, a comprehensive contamination management strategy was developed and implemented. This strategy includes dedicated equipment classification, backside wafer protection, optimized cleaning procedures for GaN etching tools, and rigorous monitoring using TXRF measurements. The approach enabled reliable, high-mechanical yield of GaN wafer device fabrication without impacting existing BCD/CMOS production lines, demonstrating the feasibility of coexisting GaN and silicon technologies in a shared manufacturing environment. This achievement paves the way for cost-effective scaling of GaN power device production within mainstream semiconductor fabs.

The research discusses progresses and challenges towards the performance improvement of amplitude-modulated (AM) mode-split gyroscopes fabricated in commercial <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$30-\mu \mathrm{m}$</tex>-thick processes. From an existing 2.6 <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{mm}^{2}$</tex> sensor with <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$0.1^{\circ} / \sqrt{ } \text{hr}$</tex> angle random walk (ARW) and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$5-10^{\circ} / \text{hr}$</tex> stability, the new architecture theoretically targets noise levels of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$0.02^{\circ} / \sqrt{ } \text{hr}$</tex> and stability below <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$0.1^{\circ} / \text{hr}$</tex>, by increasing the area to <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$4.35\ \text{mm}^{2}$</tex>, doubling the rotor voltage, and triplicating the drive displacement. Experiments, however, while matching noise targets, reveal unexpected modal couplings, triggered by softening of high-order modes. This complex mechanism is modelled to explain the observed stability degradation.

Abstract This paper presents the design and the experimental characterization of an 800 V 11 kVA three-level three-phase active neutral point clamped inverter, utilizing 650 V gallium nitride enhancement-mode high-electron-mobility transistors. The proposed scaled-down prototype design has been developed for electric traction systems. The modular approach of the presented power converter design is discussed in detail, and the different parts composing the power conversion unit are presented. Moreover, critical issues related to the design of power converters employing gallium nitride technology are highlighted for both power and gate driving sections. In addition, PCB’s thermal analysis and parasitic extraction are discussed, and for the latter the results of the conducted finite element analysis are reported. Experimental characterization of the designed prototype is performed on a 7.5 kW induction motor drive. Experimental results are reported, and the power conversion unit’s performances are evaluated in terms of efficiency at different operating conditions of the electric drive.

Phase change memory (PCM) offers high density and low current operation for energy efficient in-memory computing (IMC). However, conductance drift is still a major challenge affecting the accuracy of IMC-based edge artificial intelligence (AI) accelerators. This work presents a new weight mapping technique to linearly compensate drift, thus ensuring the linearity of matrix-vector multiplication (MVM). We show a new integrated circuit for MVM capable of current subtraction in the analog domain. High inference accuracy is demonstrated for mapped deep neural networks (DNNs) even after extensive annealing, thanks to differential mapping and optimization of read voltage V<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">read</inf> to mitigate drift-induced variations. These results support drift-compensated PCM as a robust technology for IMC-based edge AI inference.

This work presents a comprehensive study of the source of nonlinearities in a novel z-axis MicroElectroMechanical Systems (MEMS) accelerometer fabricated using a two silicon layers fabrication process. The device features a unique mechanical architecture that converts the out-of-plane motion of the proof mass into linear in-plane displacement of the sensing frames, enabling efficient capacitive readout. Initial experimental characterization revealed an unexpected non-linearity, exceeding predictions of the ideal mechanical model. To investigate the origin of this behavior, a detailed 3D finite element method (FEM) analysis was performed, incorporating fabrication-induced effects such as substrate deformation and residual stresses. Simulations demonstrated that substrate deformation has negligible impact within the operational range, while residual pre-stresses on the structural silicon layer strongly influences the device response, producing non-linearity levels consistent with experimental measurements. The close agreement between FEM predictions and experimental data validates the model and identifies residual pre-stresses on the structural silicon layer as the dominant factor affecting the device linearity. These insights provide a clear pathway for future design optimization, suggesting that careful control of residual stress and potential structural modifications can significantly improve the performance, linearity, and reliability of subsequent generations of z-axis MEMS accelerometers.

The adoption of IoT-based wireless sensor networks in aquaculture is hindered by the high cost, maintenance, and continuous power needs of conventional systems, emphasizing the need for practical and low-cost solutions to monitor key water quality parameters. This paper presents an energy-autonomous wireless sensing system that indirectly measures light energy through water samples with turbidity between 1 and 10 NTU. The system operates without a battery, harvests and senses ambient light, providing time-domain data correlating with turbidity levels. Key advantages include improved energy efficiency, fewer components, greater miniaturization, and lower implementation costs. Built using off-the-shelf components on a 2 × 2 cm PCB with a 0.45 cm thickness, the device detects turbidity changes with an elapsed time, tadv sensitivity of 10.67 ms/NTU. This cost-effective, battery-free sensing solution holds promise for more accessible, scalable monitoring of aquaculture environments, contributing to improved fish health and optimized operational efficiency.

This work introduces a CWT-based signal preprocessing method for ultrasonic imaging with PMUTs, aimed at improving signal quality before reconstruction. Validated experimentally using an 18-channel PZT PMUT array and standard algorithms (DAS, DMAS), the approach significantly boosts SNR without adding computational overhead.

In this work, we investigate how the selection of PDMS encapsulant, Sylgard 160 and Sylgard 184, affects the transmit and receive performance of immersion coupled PVD PZT PMUTs under the influence of DC bias. First, the benefit of having PDMS encapsulant has been demonstrated through acoustic characterization. Experiments were performed under immersion to determine the TRX frequency responses of the DUTs to show a 998% to 1085% improvement in TRX signal levels with DC bias as compared to the performance without bias for both types of PDMS, respectively. Second, a performance comparison of the encapsulation polymers was performed. Sylgard 160 encapsulated devices displayed 1.61 times more signal strength compared to Sylgard 184 encapsulated devices at their respective peak frequencies.