High-pressure studies of semiconductors reveal unique material behaviors that are critical for applications in extreme environments, such as aerospace, geophysics, and advanced electronics. Computational methods, particularly density functional theory (DFT) and molecular dynamics (MD), have become indispensable tools for predicting semiconductor properties under high-pressure conditions. These methods provide insights into structural stability, electronic band structure modifications, and mechanical responses without requiring costly experimental setups.
DFT is widely used to investigate electronic structure changes in semiconductors under high pressure. The method solves the quantum mechanical many-body problem by approximating electron-electron interactions through exchange-correlation functionals. Under compression, semiconductors often undergo phase transitions, which DFT can predict by comparing the total energies of different crystal structures. For example, silicon transforms from a diamond cubic structure to a beta-tin phase at approximately 11 GPa, a transition accurately captured by DFT calculations. The electronic bandgap, a critical parameter for semiconductor functionality, also evolves under pressure. DFT simulations show that the bandgap of diamond decreases under compression, eventually leading to metallization at around 1400 GPa. Similar predictions have been made for III-V semiconductors like GaAs, where pressure-induced band crossing alters carrier transport properties.
Molecular dynamics complements DFT by simulating the dynamic response of semiconductor lattices to high-pressure conditions. MD employs empirical or machine learning potentials to model atomic trajectories over time, capturing defect formation, amorphization, and plastic deformation. High-pressure MD simulations of silicon reveal that dislocation nucleation occurs at specific critical shear stresses, which depend on crystallographic orientation. The von Mises yield criterion, often applied in continuum mechanics, can be adapted in MD to predict the onset of plastic deformation in semiconductors. For instance, zinc blende structures like GaN exhibit anisotropic yielding under non-hydrostatic pressure, with slip systems activating at different stress thresholds.
The combination of DFT and MD enables a comprehensive understanding of semiconductor behavior under extreme conditions. DFT provides accurate electronic structure predictions, while MD captures kinetic processes such as fracture and phase nucleation. A notable example is the study of wurtzite-to-rock-salt transitions in wide-bandgap semiconductors like ZnO and GaN. DFT identifies the thermodynamic stability of phases, while MD reveals the kinetic barriers governing transformation pathways. High-pressure simulations also predict metastable states, such as the persistence of tetrahedral coordination in silicon at pressures beyond the equilibrium transition point.
Yield criteria under high pressure are essential for assessing mechanical failure in semiconductors. The Tresca and von Mises criteria, originally developed for metals, have been adapted for covalent and ionic crystals. DFT-calculated elastic constants enable the determination of yield surfaces, which define the stress states leading to plastic flow. For example, silicon carbide (SiC) exhibits high shear strength under hydrostatic pressure, but deviatoric stresses induce slip along preferential planes. MD simulations further quantify the pressure dependence of yield strength, showing that dislocation mobility decreases under compression due to enhanced Peierls barriers.
Electronic structure modulations under high pressure are another critical area of investigation. DFT simulations demonstrate that pressure can induce insulator-to-metal transitions, as seen in iodine-doped silicon. The applied pressure alters orbital overlap, leading to band narrowing and eventual closure of the bandgap. In layered semiconductors like MoS2, high pressure reduces interlayer spacing, increasing interlayer coupling and modifying the band structure from indirect to direct. Such transitions are relevant for designing pressure-tunable optoelectronic devices.
Thermal effects under high pressure are also addressed through computational methods. Ab initio MD, which combines DFT with classical MD, predicts thermal conductivity changes in compressed semiconductors. For instance, the thermal conductivity of diamond initially increases with pressure due to enhanced phonon lifetimes but decreases at ultra-high pressures as anharmonic effects dominate. Similar trends are observed in silicon, where pressure-induced phonon softening reduces heat transport efficiency.
Despite their strengths, computational methods have limitations in high-pressure semiconductor studies. DFT approximations, such as the choice of exchange-correlation functional, can affect phase transition pressure predictions. MD simulations rely on interatomic potentials that may not capture all bonding changes under extreme conditions. Nevertheless, ongoing advancements in machine learning potentials and hybrid quantum-classical methods are improving accuracy.
In summary, DFT and MD are powerful tools for predicting semiconductor behavior under high pressure. They provide detailed insights into structural transformations, electronic property modifications, and mechanical responses, guiding the development of materials for extreme-condition applications. Future work will likely integrate these methods with high-throughput screening and experimental validation to further enhance predictive capabilities.