Computational Materials Science最新文献

Extracting significant quantitative results from SEM images requires feature segmentation with image processing software. The efficiency of segmentation algorithms depends on the image quality, determined by the parameters set on the microscope during acquisitions. By integrating AI within SEM acquisition workflows, it is possible to suggest microscope parameters that will produce images where the features to quantify will be easily segmented. Specifically, a model is trained to automatically suggest the beam energy and probe current to set on the microscope during acquisitions. This paper is the first of two parts, describing workflows for generating a complete training set. The training set is carefully designed, consisting of both simulated data and real data acquired on the SEM by varying the energy and current. Separate workflows are required for generating simulated and acquired training examples. Simulated data generation is accomplished with the MC X-ray simulator in Dragonfly, where multiple virtual samples are created to intentionally diversify the training set. Acquiring data on the SEM for training is a time-consuming process when compared to generating simulations and would ideally be avoided but is included here to determine the degree to which it is required. Using only simulated data for adequate training, we show that our data generation workflow can be fully automated and produces a considerable amount of high quality data rapidly and with minimal effort.

Corrosion is a significant issue for materials, leading to economic losses and potential safety accidents. Corrosion degree detection allows the assessment of its impact on materials, providing crucial safety and performance information essential for maintaining and managing asset integrity. This study proposes an intelligent detection technology based on the pixel-level location of surface corrosion area and corrosion degree recognition of carbon steel samples. First, a corrosion acceleration test was employed to corrode the samples to various degrees. A generative adversarial network (GAN), StyleGAN3-t expands the corrosion image, reducing the experimental workload and sample requirements. A semi-automatic labeling approach using the Segment Anything Model (SAM) was introduced for rapid and high-resolution identification of corroded regions with complex shapes. Lastly, this paper presents the MN-DeepLabv3, which replaces the DeepLabv3 backbone network Xception with MobileNetV2, for training real corroded and generated virtual images, respectively. Experiments show that MN-DeepLabv3 outperforms other algorithms in segmenting the corrosion area and recognizing the corrosion degree. This approach presents a promising technical strategy for intelligent detection of carbon steel surface corrosion.

The rapid development of electric vehicles has promoted researchers to explore the field of high-capacity batteries. Two-dimensional (2D) materials have been proven to have ultra-high storage capacity due to their unique structural advantages. A first-principles approach was applied here to verify the feasibility of the novel AlB₂ monolayer as an anode material with ultra-high storage capacity for Li/Na-ion batteries. The Li capacity of AlB₂ monolayer is up to 3308.6 mAh/g. Meanwhile, the Na capacity up to 1654.3 mAh/g. It is worth noting that the diffusion barrier of the monolayer is extremely low (Li: 0.50 eV; Na 0.26 eV). The results show that AlB₂ monolayer is a kind of anode material for high energy storage, which provides a new choice for the development of long-life battery.

The study presents the use of a novel on-lattice sampling approach to generate titanium oxynitride (TiN_xO_y) structures with potential applications in photovoltaic and water splitting. This approach presents a simple route to overcome challenges with structure-generating tools like Cluster Approach to Statistical Mechanics (CASM), and Ab initio Random Structure Search (AIRSS), CASM faces difficulty in generating ternary structures with large unit cells. With AIRSS, there is an increase in probability of sampling amorphous sample spaces with increased number of atoms in the unit cell. Here an on-lattice sampling approach was used to model the electronic structure of TiN_xO_y as a function of composition. We present results for Ti₂N₂O, Ti₅N₄O₄ and Ti₇N₄O₈, with 33 %, 50 % and 67 % N replaced by O via substitution relative to titanium nitride (TiN), respectively. Koopmans theorem was used correct the Kohn-Sham Density Functional Theory (KS-DFT) bandgaps with corresponding values of 2.68 eV, 3.03 eV, and 3.65 eV for 33, 50 and 67 % O doping respectively. The projected density of states (PDOS) plot for TiN shows that the Fermi level is dominated by the 3d atomic orbitals of Ti, confirming pure TiN’s metallicity. The valence bands of TiN_xO_y structures were dominated by 2p orbitals of O at lower energy levels, but they were dominated by 2p orbitals of N at energies close to the valence band maximum (VBM). The conduction bands were dominated by the 3d atomic orbitals of Ti, with the bandgap increasing with O composition leading to creation of shallow trap states near the VBM, which negatively impacts carrier mobility. In conclusion, the on-lattice sampling approach is an effective tool to generate highly crystalline structures of large unit cells, also keeping O substitution for N below 33 % as seen in Ti₂N₂O is crucial for avoiding shallow traps in TiN_xO_y structures.

This work explores a possibility of improving the mechanical behavior and thermodynamic stability of AlB₂ -type YB₂ through alloying with isostructural VB₂ using first-principles calculations. The analysis derived from the cluster-expansion model suggests Y_0.5V_0.5B₂, whose atomic configuration is represented by periodically alternating YB₂/VB₂ layers in the ¡0001¿ direction, is the only solution in the pseudo-binary YB₂–VB₂ system predicted to be stable from the thermodynamic viewpoint. By evaluating the influence of lattice dynamics on the Gibbs free energies of superlattice-structured Y_0.5V_0.5B₂ and its constituent compounds within the quasiharmonic approximation, the thermodynamic stability, elastic properties, and hardness at a given pressure and temperature of the three diborides can be accessed. The results reveal, at a given temperature, isotropic compression of the diborides to high pressures enhances the stability of Y_0.5V_0.5B₂ measured relative to YB₂ and VB₂, while raising the temperature at a given applied pressure can increasingly result in a driving force toward separation of Y_0.5V_0.5B₂ into YB₂ and VB₂. The thermodynamic stabilization of superlattice-structured Y_0.5V_0.5B₂, despite large distinctions in atomic radius and electronegativity between Y and V, can be explained in terms of band filling induced by introduction of V atoms in YB₂. Within the range of temperatures (0–1200 K) and pressures (0–15 GPa) studied, the band-filling effect is found to result in significant positive deviations in the values of shear strength, stiffness, and hardness of superlattice-structured Y_0.5V_0.5B₂ from those evaluated from its constituent compounds using the Vegard’s law, respectively, by $\sim$8%, $\sim$5%, and $\sim$25%, and the hardness of superlattice-structured Y_0.5V_0.5B₂ is $\sim$40 GPa potentially indicating its superhard nature. These consequences strongly underline the essential impact of band filling on improvement in mechanical behavior and stability of YB₂ through alloying with VB₂, and also they can be served as guidance for further advancement of hard-coating technology based especially on transition-metal diborides.