We report the rapid growth of pit-free (011) β-Ga2O3 homoepitaxial layers with very low residual Si and Cl concentrations by halide vapor phase epitaxy (HVPE), in which GaCl was generated by the reaction of Ga with HCl. In this process, GaCl and O2 were supplied as the growth precursors, while additional HCl was introduced directly into the growth zone to suppress parasitic gas-phase reactions. The dependences of the residual Si and Cl concentrations on the supplied O2 and additional HCl partial pressures were systematically investigated. The residual Si concentration decreased sharply with increasing O2 partial pressure and then became nearly constant, whereas it increased monotonically with increasing additional HCl partial pressure and fell below the secondary ion mass spectrometry detection limit (1 × 1015 cm−3) in the absence of additional HCl. Thermodynamic analysis of SiO2 etching by H2 and HCl qualitatively reproduced these experimental trends and suggested that residual Si originates predominantly from etching of quartz components in the source zone upstream in the reactor by additional HCl. Even without additional HCl, the growth rate increased with increasing GaCl partial pressure, reaching 13 μm/h while maintaining a residual Si concentration below the detection limit and a residual Cl concentration less than 2 × 1015 cm−3 without particle-induced surface degradation. Capacitance–voltage measurements confirmed that the homoepitaxial layer grown at 13 μm/h was semi-insulating. Furthermore, no pit formation was observed in the absence of additional HCl, and the pits were found to originate from substrate pre-etching by additional HCl before growth. These findings establish HCl-HVPE as a practical route for the rapid growth of pit-free (011) β-Ga2O3 homoepitaxial layers with very low residual donor concentrations, providing a promising platform for cost-effective vertical β-Ga2O3 power devices.
In the present study, we have investigated the role of surface roughness influencing bacterial adhesion as a key to inhibit catheter associated infections. Polydimethylsiloxane (PDMS) substrate was fabricated through the incorporation of polyhydroxy fullerenes (PHF) to induce nanoscale roughness. Surface characteristics of developed PHF-PDMS surfaces were carried out using bio-Atomic Force Microscopy (Bio-AFM), along with fluorescent staining–based adhesion analysis for Staphylococcus (cocci) and Klebsiella (rod-shaped) bacteria. PHF-PDMS exhibited significant nanoscale heterogeneity, reflecting through its altered roughness, and asymmetric height profiles influencing bacterial morphology. Bacteria grown on the PHF-PDMS substrate showed characteristic protruding and membrane morphological changes. Fluorescence imaging corroborated with other findings, showing reduction in bacterial growth, and spatially fragmented adhesion over PHF-PDMS compared to the control surface. Collectively, this study highlights that surface nanoarchitectonics offers antibiotic-free strategies for preventing biofilm formation on biomedical devices.
Short-term memory is essential for temporal information processing, yet its hardware implementation commonly requires recurrent connections or additional state storage. Here, we demonstrate a device-level short-term memory mechanism enabled by stochastic switching in magnetic tunnel junction neurons. Probabilistic RESET failures retain the neuronal state across time steps, generating an exponentially decaying memory trace whose duration is directly controlled by the RESET probability. This relationship was established theoretically and verified through device-level state-retention measurements, delayed-matching, and n-back tasks. We further incorporated stochastic-RESET neurons into an SEW-ResNet18 architecture and evaluated N-CARS classification, Time-Reversed DVS Gesture recognition, and DDD20 steering prediction. Appropriate RESET probabilities consistently outperformed the memory-free condition, while different tasks favored different memory durations. These results demonstrate tunable, compact, device-native short-term memory for efficient spatiotemporal computing.
Solidification cracking is a critical defect in welding, and its occurrence has conventionally been described using temperature-dependent critical-strain curves. However, whether the critical strain is uniquely determined by temperature remains unclear, particularly at the dendrite scale. In this study, solidification-cracking behavior in an Fe–15Mn–10Cr–8Ni–4Si alloy was investigated using in-situ synchrotron X-ray transmission imaging combined with local strain and temperature analyses. Crack-initiation events were identified from the time-resolved X-ray images, and the corresponding critical equivalent strains and critical temperatures were extracted at individual positions along the crack path. Solidification cracking occurred over approximately 1320–1000°C, corresponding to the effective solid–liquid coexistence region predicted by a Scheil–Gulliver calculation. The median critical strain was higher at lower temperatures and generally decreased toward higher temperatures. A broad critical-strain distribution remained within each 30°C temperature range. After subtraction of the temperature-bin median, the remaining variation was spatially nonuniform along the crack path. Comparison with a non-cracking case showed that most equivalent-strain values remained below the critical-strain distribution obtained in the cracking case, although the upper tail partially overlapped its lower part. These results provide experimental evidence that the critical strain at solidification-crack initiation is better represented as a temperature-dependent distribution than as a single-valued curve. The physical factors governing the distribution width and spatial variation remain to be identified.
