A multiscale modeling framework integrating density functional theory, machine learning interatomic potentials, molecular dynamics and finite element simulations was developed to overcome the intrinsic limitations of single scale approaches in describing thermal and mechanical processes in GaN heterostructures. Atomic level calculations capture phonon dispersion and interfacial scattering behavior which cannot be resolved at the continuum level, while device scale simulations reveal temperature and stress distributions that are inaccessible to atomistic methods alone. The framework links interfacial phonon scattering with heat accumulation in the active region and stress development under high power operation. The results show that substrate thermal conductivity strongly affects heat spreading behavior, with diamond enabling significantly improved heat dissipation compared with conventional substrates. With increasing power density, the temperature in the near-junction GaN and AlGaN buffer layers approaches critical values, while interfacial stress gradually rises toward reported adhesion limits, indicating an increased risk of thermally induced failure. These observations highlight the key role of coupled thermal and mechanical effects across multiple length scales. The results further suggest that improving heat extraction at the package level, together with reducing interfacial thermal resistance, is essential for enhancing the reliability of GaN based power devices.

The 20CrMnTiH alloy is widely used in the manufacture of gears, gear rings, and gear shafts due to its balanced mechanical properties. However, its limited hardness and wear resistance lead to premature surface failure during long-term service. In this study, high-speed laser directed energy deposition (HS-LDED) was employed to repair the surface of the 20CrMnTiH alloy using Fe-Cr-Ni alloy powder, achieving a sound metallurgical bond between the repair layer and the substrate. The HS-LDED process successfully produced a dense and defect-free repair layer with excellent interfacial bonding. Increasing the scanning speed improved the cooling rate and solidification rate, thereby accelerating the BCC to FCC phase transformation, refining the grains, promoting equiaxed grain nucleation, and increasing the geometrically necessary dislocation density in the Fe-Cr-Ni repair layer. These microstructural evolutions collectively contributed to a hardness increase from 636.3 HV0.5 at 70 mm/s to 669.1 HV0.5 at 85 mm/s, primarily through grain refinement and dislocation strengthening mechanisms. Consequently, compared with the matrix, the wear resistance of the repair layer wass markedly improved, with reductions in friction coefficient, wear width, and wear depth by 35.48%, 12.0%, and 74.85%, respectively. This study demonstrates that HS-LDED enables process-microstructure-property tailoring in Fe-Cr-Ni repaired 20CrMnTiH components, offering a rapid and reliable solution for extending the service life of gear materials in demanding applications.

Understanding the structural evolution during laser stealth dicing of difficult-to-process semiconductors such as AlN is crucial for wafer processing. Here, we establish a molecular dynamics model to investigate the atomic-scale evolution of AlN during laser stealth dicing and tensile fracture, and the effects of laser power, scanning duration, and scanning path on dicing performance are systematically analyzed. The results reveal that higher energy input lowers fracture stress but increases sidewall roughness. Employing sinusoidal or dual-path scanning strategies enables a reduction in fracture stress and improved defect atom distribution under the same or even reduced energy input, achieving energy-efficient, low-defect dicing. This work provides theoretical guidance for optimizing laser stealth dicing of hard materials like AlN.

During the heteroepitaxial growth of GaN, the substantial lattice mismatch and thermal expansion mismatch between the substrate and the epitaxial layer often led to a high density of structural defects, thereby severely degrading the crystalline quality and operational reliability of GaN-based devices fabricated on heterogeneous substrates. In this work, a molecular dynamics framework was established to describe the atomic-scale evolution of GaN epitaxy on patterned AlN substrates with cylindrical and prismatic geometries under different pattern densities. The effects of substrate patterning on surface morphology, crystalline quality, dislocation propagation behavior, dislocation density, and residual stress in GaN epitaxial films were systematically investigated. The results indicate that as the pattern density increases, the spacing between patterns decreased, leading to enhanced coalescence of deposited atoms under identical growth conditions. Furthermore, the prismatic patterned structure exhibited a relatively high pattern coalescence ability and demonstrated a certain degree of preferential growth in the [1-100] crystalline direction.

This study investigates Cu-Cu bonding enhancement through various intermediate metal layers (X = Al, Ni, Au, Ti, and Zr) through molecular dynamics simulations. The introduction of intermediate metal layer significantly enhances interfacial atomic migration. Through atom migration and mean square displacement analysis, comprising Cu(100)-X-Cu(100) heterostructures, reveal that Zr demonstrates the most pronounced facilitation of Cu atom diffusion at 400 K, 600 K, and 800 K, exceeding that of Ti at 800 K. This performance correlates with Zr's higher melting point (2125 K) and larger atomic radius (1.60 & Aring;), which facilitate the diffusion of Cu. Ti metal also exhibits strong interfacial reactivity at 400 K, while Au and Ni require temperatures above 600 K for activation. Al intermediate metal layer shows minimal Cu migration due to mismatched atomic radii and higher interfacial energy. These findings establish Zr as the most effective intermediate layer for low-temperature Cu-Cu bonding in 3D integrated circuits.

Gallium oxide (Ga2O3) is an ultra-wide bandgap semiconductor with excellent potential for high-power device applications. Some defects will inevitably occur during the growth of Ga2O3. Dislocations, as one of the main defects, have a significant impact on the mechanical and electrical properties of materials. In this work, dislocations in beta-Ga2O3 are systematically studied via first principles calculations, with a primary focus on the screw dislocation occurring on the ( 2 & strns;01) plane with b= < 010 >. The concept of Ga-tetra-Ga-octa-dislocation is proposed. The three dislocation core structures, Ga-octa-O-I-Ga-tetra-near-dislocation, Ga-octa-O-I-Ga-tetra-far-dislocation, and Ga-octa-O-III-Ga-octa-far-dislocation have been verified. The study find that the introduction of dislocations leads to a narrowing of the bandgap. This suggests that the presence of dislocations may reduce the breakdown voltage, radiation resistance, and other related properties of beta-Ga2O3. This study provides a new perspective for the theoretical investigation of dislocations in beta-Ga2O3, with significant guiding implications. It also reveals the impact of dislocations on the electrical properties of beta-Ga2O3, laying a foundation for further research.

Laser stealth dicing (LSD) is a critical technology for high-precision, low-damage singulation of brittle semiconductor wafers. In this study, a molecular dynamics (MD) model is established to comprehensively analyze the atomic-scale structural evolution under varying laser powers and scanning times. Both simulation and experiment demonstrate that the extent of the modified layer and the degree of amorphization are critically governed by the total laser energy input, increasing with higher power and longer scanning time. Subsequent tensile simulations quantify how this laser-induced damage degrades the mechanical strength of material, lowering the fracture stress. Furthermore, a non-monotonic relationship between processing parameters and post-fracture surface roughness is identified, with optimal smoothness achieved at intermediate laser energy input. Experimental results on modified layer morphology and surface roughness, obtained by scanning electron microscope (SEM) and 3D optical profiler show excellent qualitative agreement with the MD predictions, thereby validating the simulation model. This work not only elucidates the fundamental atomic-scale mechanisms of phase transformation and fracture in LSD but also establishes MD as a reliable predictive tool for optimizing laser processing parameters to achieve high-quality, controlled fracture in advanced semiconductor manufacturing.

Diamond is regarded as a critical material for the thermal management of high-power electronics. However, the impact of shear strain, which is a prevalent deformation mode in complex stress fields, remains unclear. In this work, the lattice thermal conductivity of diamond under shear strain has been systematically investigated using high-accuracy neuroevolution potential and homogeneous non-equilibrium molecular dynamics. Distinct from isotropic or uniaxial strain cases, shear strain induces pronounced threshold effects and strong anisotropy. Beyond a critical shear strain of 0.01, the thermal conductivity exhibits a sharp decay, with the reduction along the [001] direction (perpendicular to the (001) shear plane) being markedly greater than that along the [100] and [010] directions (parallel to the (001) shear plane). This anisotropy is attributed to a distinct phonon frequency shift compensation mechanism. Along the [100] and [010] directions (parallel to the shear plane), a phonon redshift drives high-frequency modes into the 10-30 THz range critical for thermal transport, thereby partially compensating for scattering-induced losses. Consequently, it can be concluded that the microscopic origins of strain-dependent thermal transport provide further theoretical insights into the thermal management of electronic devices employing diamond heat spreaders under complex stress fields.

The rising power density of advanced electronics demands improved thermal management, while traditional single-scale methods are unable to fully reveal the complex heat transfer mechanisms in heterostructures. This work establishes a multiscale simulation framework by constructing a machine learning potential, enabling accurate cross-scale parameter transfer from atomic to mesoscopic and then to macroscopic levels. Results show that the thermal boundary resistance (TBR) at the beta-Ga2O3/diamond interface is higher than that at the beta-Ga2O3/Si and beta-Ga2O3/SiC interfaces, and that the TBR decreases with increasing temperature, which contradicts conventional understanding. Vibrational density of states and interface conductance modal analysis elucidate the underlying mechanisms. These mesoscale insights are incorporated into macroscopic simulations, showing the beta-Ga2O3/diamond heterostructure's peak power capability reaches 226% of that of beta-Ga2O3/Si. Further analysis reveals that although the thermal conductivity of the heat-spreading substrate remains the dominant factor in overall thermal performance, the thermal bottleneck gradually shifts toward the interface as both substrate conductivity and operating temperature rise. Moreover, crystal orientation significantly influences thermal performance and thermal stress distribution, necessitating careful trade-offs. This study not only provides effective strategies for optimizing beta-Ga2O3-based devices but also establishes a generalizable paradigm for cross-scale thermal management research in heterogeneous material systems.

As a critical safety component and primary load-bearing structure, aircraft landing gear is pivotal to ensuring safe operation during ground movements. However, under continuous dynamic loading, microstructural damage and mechanical property degradation caused by landing gear corrosion may lead to fracture. This work focuses on the 50CrVA landing gear of a lightweight aircraft. The corrosion-resistant GH3536 alloy is employed to repair the 50CrVA alloy via laser remanufacturing technology. The research systematically investigates the microstructural evolution, interfacial characteristics, and high-temperature (700 degrees C) tensile properties. The results show that the laser remanufactured aircraft landing gear demonstrates a defect-free heterogeneous interface with no porosity and cracking. The repaired microstructure is composed of coarse columnar grains in the repair layer, ultra-fine lath grains in the heat affected zone (HAZ), and medium equiaxed grains in the matrix. Besides, the HAZ of 50CrVA undergoes phase transformation: The original ferrite microstructure transforms into austenite microstructure and finally decomposes into martensite microstructure with a small amount of austenite, while the grains are significantly refined. Grain size and dislocation density show a nonlinear gradient distribution. The dislocation densities of the matrix, HAZ, and repair layer are 3.2 & times; 1014 m- 2, 7.6 & times; 1014 m- 2, and 0.4 & times; 1014 m- 2, respectively, which helps alleviate deformation incompatibility under load and improve the plastic deformation capability of the alloy. The repaired sample exhibited increases of 55.9 % in yield strength and 85.4 % in ultimate tensile strength compared with the matrix. The improvement in the high-temperature tensile strength of the repaired sample mainly comes from the interface strengthening induced by phase transformation and hetero-deformation induced (HDI) strengthening. These findings confirm the feasibility of improving the mechanical properties of the repaired 50CrVA alloy by laser remanufacturing, offering a promising approach for extending service life and ensuring operational safety in aviation applications.

This work systematically investigates the interfacial structure evolution, atomic diffusion behavior, strain distribution, and defect formation mechanisms of Au–Si bonding systems with different crystallographic orientation combinations. The results demonstrate that the interfacial properties, atomic mobility, and eutectic bonding performance of the bonding systems strongly depend on the crystal orientation of the Au–Si interfaces. Among the Au–Si (100) bonding systems, the Si (100)–Au (110) interface exhibits the strongest interfacial diffusion capability and atomic mixing behavior due to the relatively open atomic arrangement of the Au (110) surface, which effectively reduces the eutectic nucleation barrier and expands the effective eutectic reaction region. In the Au–Si (110) bonding systems, although the Si (110)–Au (100) and Si (110)–Au (110) interfaces exhibit relatively high mean squared displacement values, the actual number of diffused atoms and the diffusion depth along the z-direction remain limited because of the high vertical diffusion barrier of the Si (110) crystal plane. In addition, obvious stacking faults, localized hexagonal close packed phase transformation, and dislocation structures are observed in these systems, indicating that the interfacial strain could be released through defect-mediated structural evolution. Overall, the simulation results confirm that the Si (100)–Au (110) interface is the most favorable structure for Au–Si eutectic bonding.

Nitrogen-vacancy (NV) center is regarded as one of the most significant point defects in diamond crystals and has emerged as a highly promising candidate for quantum technologies. In this study, the stable diamond NV centers of good coherent properties have been successfully fabricated without relying on a pure 12C source or posttreatment processes, such as electron irradiation or high-temperature annealing. Instead, the nitrogen/ hydrogen (N2/H2) gas ratio has been optimized during the microwave plasma chemical vapor deposition (MPCVD) process. Additionally, the relationship between the nitrogen concentration in the process gas and the coherence times of the NV center, as well as the synergistic effect of boron-nitrogen co-doping on NV center formation have been systematically examined. As a result, it can be indicated that during the fabrication of diamond NV centers via the MPCVD process, the nitrogen doping flow rate has a significant non-monotonic effect on the coherence times (T2 and T2*). The maximum coherence times for T2 (6.83 mu s) and T2* (288.1 ns) are achieved at the nitrogen flow rate of 0.05 sccm. Both coherence times then fluctuate in a similar trend in the range of 2.25-4.79 mu s for T2 and 57.2-193.6 ns for T2* as the flow rate increases up to 5.00 sccm. Furthermore, both experimental and computational results demonstrate that B-N co-doping probably suppresses the formation of NV centers, as boron may act as an inhibitor in the process. Thus, this study could provide critical experimental insights into the controlled synthesis of NV centers and advances the understanding of defect engineering in diamond-based quantum materials.

Chemical mechanical polishing (CMP) is the sole process capable of achieving the required flatness and surface roughness for photolithography without any obvious distortion in the multilevel metal interconnects. As semiconductor manufacturing advances to the next process node, the introduction of new materials and structures has proposed higher performance standards for polishing slurries, resulting in the slurry composition becoming increasingly critical to the overall polishing process. In this work, ReaxFF-based molecular dynamics (MD) is employed to investigate the copper (Cu) CMP process in various slurries, aiming to uncover the chemical interactions of different components and the atomistic mechanisms involved in Cu atom removal. The results demonstrate that the presence of H2O2 cannot only directly oxidize the Cu atoms on the substrate surface, but also inhibit the adsorption of H2O on the Cu surface and promote the dissociation of the adsorbed H2O to indirectly oxidize the Cu atoms. The Cu complexes Cu–C2H5O2N and Cu–H2C2O4 are generated during the reaction due to the addition of glycine and oxalic acid, respectively. The oxidation of H2O2 and the complexation of glycine and oxalic acid significantly enhance the Cu removal. Furthermore, Cu atoms tend to be removed in the form of clusters, and the removal rate is the highest in the mixed solution of H2O2 and glycine. The surface roughness after polishing is 0.082 nm, which closely aligns with the atomic force microscopy (AFM) experimental data of 0.104 nm. This work sheds light on the role of different components in the polishing slurry, which is of great significance to the design of the CMP slurry components for more advanced process nodes.

InGaP/GaAs heterojunction layers are commonly used as semiconductor materials in GaAs solar cells. Nevertheless, challenges endure in the form of poor quality InGaP/GaAs heterojunctions. This is attributed to the diffusion of P atoms caused by the memory effect in the GaAs absorption layer, as well as the As/P exchange and H2 etching effect in the InGaP window layer. In this work, the residual group-V source evacuation (RSE) and stabilizing method have been utilized to InGaP/GaAs heterojunction interface quality, and these effects have been investigated. The variation of process gas concentrations in metal-organic chemical vapor deposition (MOCVD) reactor is numerically studied as an indicator of H2 etching and As/P exchange. Optimization of stabilizing and RSE period times is found to be necessary in order to achieve high-quality heterojunction, balancing the memory effect, etching effect, As/P exchange effect, and GaAs crystal quality. When the RSE and stabilizing period times are both set to 0.5 s, it is observed that the photo luminescence (PL) performance of InGaP/GaAs reached its optimal level. These studies have great significance to the fabrication of InGaP/GaAs heterojunction-based GaAs solar cell, which promotes the further development of multi-junction and high photovoltaic conversion efficiency (PCE) solar applications.

The angle-resolved scatterometer (ARS) is a model-based technique that has demonstrated significant potential for critical dimension (CD) measurements in integrated circuit (IC) manufacturing due to its high speed, accuracy, and non-contact operation. ARS is essentially an inverse problem-solving process. Traditional methods, such as library search, rely on extensive spectral databases with fine grid intervals, which impose substantial storage and computational demands. Nonlinear regression techniques, like the Levenberg-Marquardt (LM) algorithm, require significant computational resources for real-time Jacobian matrix evaluations, limiting their applicability for online measurements. To overcome these challenges, we propose a hybrid method combining library search and particle swarm optimization (LSPSO). This approach employs a sparse spectral library with large grid intervals (LGI), dramatically reducing storage requirements. The LGI library also guides the initial search direction for the PSO, eliminating the need for early model evaluations and improving computational efficiency by reducing the number of model computations. Simulations on both single-parameter (film) and multi-parameter (grating) models demonstrate the accuracy of LSPSO, with noise analysis further confirming its stability. Experimental validation on grating structures reveals that LSPSO improves computational efficiency by 52% compared to standard PSO and operates 1.7 times faster than the LM method. Moreover, the spectral data required by LSPSO is reduced to just 0.1% of the data needed by traditional library search methods. These results position LSPSO as a highly efficient, accurate, and resource-efficient solution for CD measurements in ARS applications.

To explore potential n-type diamond with shallow donor levels, hydrogen–nitrogen co-doped (H-N co-doped) diamonds at various concentrations have been investigated using density functional theory (DFT). The formation energies for Hs-N3 (a substitutional hydrogen atom bonded to three nitrogen atoms) and Hs-N4 (a substitutional hydrogen atom bonded to four nitrogen atoms) co-doped diamonds are relatively low, at − 1.771 eV and 0.022 eV, respectively. Notably, the Hs-N4 defect introduces a shallow n-type donor level of 0.147 eV. Analysis of the density of states (DOS) reveals that the conduction band is mainly composed of C-2 s/2p and N-2 s/2p orbitals, with minor contributions from H-1 s. The formation energy of Hs-N increases under strain (0–10

In the back-end-of-line (BEOL) stage of wafer fabrication, the Damascene process is widely used for interconnect metallization, with plasma-enhanced chemical vapor deposition (PECVD), photolithographic patterning, and copper electroplating as the primary steps. The mismatch in the coefficients of thermal expansion (CTE) between the dielectric materials and the silicon substrate results in inevitable wafer warpage during the high-temperature PECVD process. As the number of Damascene interconnect layers increases, warpage accumulates and intensifies, ultimately compromising process stability and wafer yield. This study investigates wafer warpage induced by high-temperature PECVD in the Damascene process using a finite element simulation approach. To capture the cross-scale structural features and real process conditions of wafers more accurately, a representative volume element (RVE) model is introduced, along with the element birth-and-death technique, to enhance prediction accuracy. A predictive model for wafer warpage was developed and validated experimentally, with a simulation error of less than 11.1%. The accurate warpage prediction model provides valuable insights for process optimization, reducing experimental iterations and improving manufacturing efficiency. Additionally, a warpage mitigation strategy is introduced by pre-patterning trenches on the wafer before the Damascene process, facilitating stress redistribution. Simulation results show that this approach reduces wafer warpage by 12.7%, resulting in a significant improvement in wafer planarity.

GaN-on-diamond power devices exhibit superior thermal performance due to the exceptional thermal conductivity of diamond. Enhancing GaN film quality on diamond involves first depositing an AlN seed layer on (111) single-crystal diamond (SCD) substrates via magnetron sputtering, followed by AlxGa1-xN buffer layer growth. However, the influence of polycrystalline AlN (Poly-AlN) layers on the atomic-scale crystalline quality of AlxGa1xN remains underexplored. This study employs molecular dynamics (MD) simulations to investigate the deposition of AlxGa1-xN on Poly-AlN seed layers with varying grain sizes and AlN mole fraction. Results indicate that larger grain sizes reduce grain boundaries, enhancing the crystalline quality of the buffer layer. Additionally, a lower AlN mole fraction increases the lattice constant of AlxGa1-xN, thereby increasing the lattice mismatch with AlN. Post-annealing simulations reveal a reduction in mean biaxial stress by over 50 % and a decrease in average normal stress in the Z direction by more than 35 % in the final equilibrium state.

High-entropy alloys (HEAs) are renowned for their exceptional mechanical properties, making them ideal candidates for use in demanding engineering applications such as aerospace, automotive, and defense industries. However, optimizing the balance between strength and ductility in HEAs remains a formidable challenge. In this study, we introduce a novel approach to enhance the mechanical properties of FeNiCrCo-based HEAs through in-situ alloying with nano-carbide-framework. By integrating nano-carbon into HEAs during laser powder bed fusion (LPBF), we were able to induce the formation of carbon-enriched nanostructures during the rapid melting and solidification processes of LPBF. Our methodology focuses on the strategic modulation of unique submicronscale dislocation networks and the development of a nano-carbide-frameworks within the HEA matrix. The presence of CrxCy compounds primarily at the grain boundaries and within dislocation networks significantly contributes to the mechanical strength of the alloy. By varying the nano-carbon content, we achieved control over the alloy's microstructures, enabling a tailored balance between ultimate strength and ductility. This in-situ HEA alloying approach leads to the formation of a highly coherent nano-carbide-framework with the matrix, which not only enhances the ultimate strength of the HEAs (achieving values close to 1.4 GPa) but also maintains improved ductility. The nano-carbide-frameworks enabled microstructural design of HEAs provides a potent method for enhancing both the thermal stability and mechanical performance of the alloys. Our study paves the way for future research on the applicability of HEAs for applications in extreme conditions, and offers novel insights and methodologies for developing next-generation HEA with optimized performance.