Xuejiang Chen

Silicon carbide (SiC), as a wide-bandgap semiconductor material, plays a crucial role in high-power semiconductor devices. The species transport processes within the growth chamber of PVT induction furnace for SiC crystal growth, and the effects of growth conditions such as temperature, pressure, and height of the growth chamber are studied. In the study, the chemical reactions, component transport, and decomposition-deposition phenomena are all considered. And the effects of convection and diffusion are also analyzed in the growth progress. The results indicate that by optimizing growth temperature and pressure, while adjusting the height of the growth chamber, an appropriate convection-diffusion ratio can be determined, thus, the growth rate can be adjusted and the crystal quality can be improved.

Altitude Test Facility (ATF) compressor systems are widely used in aero-engine tests. These systems achieve the control of gas pressure and transport through complex operation processes. With advancements in the aviation industry, there is a growing demand for higher performance, greater safety, and more energy efficiency in digital ATF systems. Hybrid modeling is a technology that combines many methods and can meet these requirements. The Modular Dynamic System Greitzer (MDSG) compressor model, including mechanistic and data-driven modeling approaches, is combined with a neural network to obtain a BPNN-MDSG hybrid modeling method for the digital turbine system. The digital simulation is linked with the physical sensors of the ATF system to realize real-time simulation and monitoring. The steady and dynamic conditions of the actual system are simulated in virtual space. Compared with the actual results, the average error of steady mass flow is less than 3%, and the error of pressure is less than 1%. The average error of dynamic mass flow is less than 5%, and the error of pressure is less than 3%. The simulation and characteristic predictions are carried out in BPNN-MDSG virtual space. The anti-surge characteristics of the ATF system under start-up conditions are obtained. The full-condition anti-surge operation map of the system is obtained, which provides guidance for the actual operation of the ATF system.

Transition metal dichalcogenides (TMDs) are widely used in semiconductor materials, lubricants, catalysts, and various other applications due to their exceptional physical and chemical properties. During the chemical vapor deposition growth process of TMDs, triangular or hexagonal domains can be generated under appropriate conditions, and TMDs may also exhibit fractal growth and manifest diverse morphologies, such as irregular fractals, dendrites, and semicompact structures. In order to thoroughly investigate the phenomenon of fractal growth, a lattice kinetic Monte Carlo model has been developed to quantitatively characterize the morphology of tungsten diselenide (WSe2) on a graphene substrate and simulate its growth process. First, the deposition rate and diffusion rate of atoms are calculated by using the energy parameter, and then, the atomic kinetic behaviors of atoms are simulated. Subsequently, the coordinates of all adsorbed atoms during the growth process are obtained, and, thus, the morphology of the domain is obtained. Then, their morphological characteristics are further analyzed including the nucleation conversion rate, radius of gyration, and fractal dimension. The results indicate that the growth temperature has a significant impact on the fractal type of domain, and the deposition flux has a significant impact on the branching part of the domain. The increase in the temperature leads to a more compact atomic distribution within the domain, resulting in the changes in the overall morphology. Similarly, an increase in the deposition flux leads to an increase in the number of atoms in the branches of the domain.

Due to the lack of research on the microscopic evolution process and the formation of step growth patterns for SiC crystals, it is of great importance to deepen the understanding of the epitaxial growth of a SiC vicinal surface from a microscopic point of view. In this study, a three-dimensional lattice kinetic Monte Carlo algorithm was used to study the step flow growth characteristics of SiC crystals. The microscopic evolution of the step flow growth patterns for SiC vicinal surfaces was shown. C and Si were treated as the basic particles, and the net deposition and diffusion of atoms were considered in this model. The periodic boundary conditions were applied along the step edge and the helical boundary conditions were applied in the direction perpendicular to the step. The surface morphology evolution of SiC crystals grown on step substrates was simulated at different growth temperatures, partial deposition fluxes, and terrace widths. The results indicated that the growth patterns of the SiC vicinal surface could be transformed from a step flow growth to a nucleation growth pattern by decreasing the growth temperature, increasing the deposition flux, and increasing the terrace width.

An on-lattice kinetic Monte Carlo model is constructed to investigate microscopic nucleation behavior during the submonolayer epitaxial growth of GaN islands, where the Ga and N atoms are treated as the basic particles. The input kinetic parameters of Ga and N, including their surface diffusion energy barriers, were obtained from previous ab initio calculations. Furthermore, a simple and effective bond counting rule is applied in our kinetic Monte Carlo model, and the statistics of the GaN islands on the surface are realized via the application of the Hoshen–Kopelman algorithm. The growth temperature range covers the typical growth temperatures used in the molecular beam epitaxy of GaN. The results obtained show that triangular GaN flakes are observed and that the shapes of the GaN islands remain triangular when the growth temperature is changed. Additionally, the power law for the maximum density of islands versus the ratio of the effective diffusion to the deposition rate is obtained; the exponent of this law is −0.506 ± 0.006, indicating that these triplets represent the seeds required for further nucleation. Finally, the coexistence of the Ga-edge and N-edge types of triangular GaN islands is observed. The island formation mechanism is attributed to a local monomer density misbalance, and it is also shown that a slight variation in the Ga/N ratio in the deposition flux changes the proportion of the Ga-edge and N-edge type triangles; this represents a further indication that controllable GaN morphologies can be obtained by tuning the chemical potentials of the constituent elements.

In order to better understand the microscopic nucleation behavior during the epitaxial growth of a SiC crystal, a lattice kinetic Monte Carlo model is developed, in which Si and C particles are set as the basic elements. The events list is built up to implement group search and binary search, which can improve the efficiency of the algorithm. In addition, the Hoshen–Kopelman algorithm is also used to calculate the statistics of the properties of clusters. Then, the cluster size distribution function and the nucleation preference of trimers are analyzed. The results show that the cluster size distribution function obeys the Gauss distribution, and the proportion of crystal nuclei close to the average size gradually increases. Moreover, the growth temperature mainly affects the trapping rate of adatoms by dimers, and the C/Si ratio affects the formation of different types of dimers and the concentration of adatoms.