In this work, atomic scale investigation of pressure-induced transformations in the solid state have been carried out. A series of compounds including GaN, ZnO, CaF2, and AgI, in addition to elemental phosphorus have been studied. The corresponding transition mechanisms have been elucidated with a clear description of atomic displacements and intermediate structures involved therein. In the first group of compounds, the long standing debate on the transition path of the wurtzite(WZ)-to-rocksalt(RS) transition in semiconductors, GaN and ZnO was resolved using geometrical modeling combined with molecular dynamics (MD) simulations conducted in the frame of transition path sampling (TPS) method. In GaN, a two-step mechanism through a metastable intermediate phase with a tetragonal structure iT has been revealed from simulations. In ZnO, the tetragonal intermediate structure was kinetically less stable, although still part of the real transition mechanism. It appeared at the interface between WZ and RS as consequence of a layers shearing. The transition regime in ZnO was characterized by a competition between iT structure and another hexagonal intermediate with hexagonal symmetry iH. Although possible, the latter is not functional for the transition. In both cases, GaN and ZnO, two points of agreement with experiments have been revealed. The tilting of structures after transition, and the phonon mode softening associated with atomic displacements leading to the tetragonal structure iT In the second group of compounds, the investigation of transitions in superionic conductors, CaF2 and AgI, demonstrated a different and particular behavior of atomic motion under pressure. The solid-solid reconstruction of CaF2 structure was shown to be initiated and precedented by high disorder of the anionic sublattice. The percolation of fluoride ions through voids in the fluorite structure created a thin interface of liquid like state. The sparce regions caused by the departure of anions facilitates the cation sublattice reconstruction. In AgI, ion diffusion during the wurtzite/zincnlende(ZB)$rocksalt transition was more pronounced due to the extended stacking disorder WZ/ZB. The Ag+ ions profited not only from the structure of the interface but used the combination of interstitial voids offered by both phases, WZ and ZB, to achieve long diffusion paths and cause the cation sublattice to melt. Clearly, a proper account for such phenomena cannot be provided by geometry-designed mechanisms based on symmetry arguments. In phosphorus, the question of how the stereochemically active lone pairs are reorganized during the orthorhombic (PI) to trigonal (PV) structural transition was answered by means of simulations. Computation was performed at different levels theory. First, the mechanism of the transition was obtained from TPS MD simulations. MD runs were performed within density functional tight binding method (DFTB). The analysis of atomic displacements along the real transformation path indicated a fast bond switching mechanism. In a second step, the nature of the interplay between orbitals of phosphorus during the bond switching was investigated. A simultaneous deformation of lone pair and P−P bond showed a mutual switching of roles during the transformation. This interplay caused a low dimensional polymerization of phosphorus under pressure. The corresponding structure formed as zigzag linear chain of fourfold coordinated phosphorus atoms (· · ·(P(P2))n · · ·) at the interface between PI and PV phases. A further result of this work was the development of a simulation strategy to incorporate defects and chemical doping to structural transformations. On top of the transition path sampling iterations, a Monte Carlo like procedure is added to stepwise substitute atoms in the transforming system. Introducing a chemically different dopant to a pure system represents a perturbation to the energy landscape where the walk between different phases is performed. Therefore, any change in the transition regime reflects the kinetic preference of a given structural motif at times of phase formation. This method was applied to the elucidation of WZ-RS transition mechanism in the series of semiconducting compounds AlN, GaN, and InN. Simulations showed that In atoms adopt the same transformation mechanism as in GaN and favor it, while Al atoms demonstrated a significant reluctance to the path going through tetragonal intermediate iT. The difference between transition regime in mixed systems InxGa1−xN and AlxGa1−xN is in agreement with experiments on high pressure behavior of AlN, GaN, and InN. While transitions in GaN and InN are reversible down to ambient conditions, AlN is stable. The work presented in this thesis constitutes the seed of new perspectives in the understanding of pressure-induced phase transformations in the solid state, where the physics and the chemistry are brought together by means of computer simulations.

In this work, atomic scale investigation of pressure-induced transformations in the solid state have been carried out. A series of compounds including GaN, ZnO, CaF2, and AgI, in addition to elemental phosphorus have been studied. The corresponding transition mechanisms have been elucidated with a clear description of atomic displacements and intermediate structures involved therein. In the first group of compounds, the long standing debate on the transition path of the wurtzite(WZ)-to-rocksalt(RS) transition in semiconductors, GaN and ZnO was resolved using geometrical modeling combined with molecular dynamics (MD) simulations conducted in the frame of transition path sampling (TPS) method. In GaN, a two-step mechanism through a metastable intermediate phase with a tetragonal structure iT has been revealed from simulations. In ZnO, the tetragonal intermediate structure was kinetically less stable, although still part of the real transition mechanism. It appeared at the interface between WZ and RS as consequence of a layers shearing. The transition regime in ZnO was characterized by a competition between iT structure and another hexagonal intermediate with hexagonal symmetry iH. Although possible, the latter is not functional for the transition. In both cases, GaN and ZnO, two points of agreement with experiments have been revealed. The tilting of structures after transition, and the phonon mode softening associated with atomic displacements leading to the tetragonal structure iT In the second group of compounds, the investigation of transitions in superionic conductors, CaF2 and AgI, demonstrated a different and particular behavior of atomic motion under pressure. The solid-solid reconstruction of CaF2 structure was shown to be initiated and precedented by high disorder of the anionic sublattice. The percolation of fluoride ions through voids in the fluorite structure created a thin interface of liquid like state. The sparce regions caused by the departure of anions facilitates the cation sublattice reconstruction. In AgI, ion diffusion during the wurtzite/zincnlende(ZB)$rocksalt transition was more pronounced due to the extended stacking disorder WZ/ZB. The Ag+ ions profited not only from the structure of the interface but used the combination of interstitial voids offered by both phases, WZ and ZB, to achieve long diffusion paths and cause the cation sublattice to melt. Clearly, a proper account for such phenomena cannot be provided by geometry-designed mechanisms based on symmetry arguments. In phosphorus, the question of how the stereochemically active lone pairs are reorganized during the orthorhombic (PI) to trigonal (PV) structural transition was answered by means of simulations. Computation was performed at different levels theory. First, the mechanism of the transition was obtained from TPS MD simulations. MD runs were performed within density functional tight binding method (DFTB). The analysis of atomic displacements along the real transformation path indicated a fast bond switching mechanism. In a second step, the nature of the interplay between orbitals of phosphorus during the bond switching was investigated. A simultaneous deformation of lone pair and P−P bond showed a mutual switching of roles during the transformation. This interplay caused a low dimensional polymerization of phosphorus under pressure. The corresponding structure formed as zigzag linear chain of fourfold coordinated phosphorus atoms (· · ·(P(P2))n · · ·) at the interface between PI and PV phases. A further result of this work was the development of a simulation strategy to incorporate defects and chemical doping to structural transformations. On top of the transition path sampling iterations, a Monte Carlo like procedure is added to stepwise substitute atoms in the transforming system. Introducing a chemically different dopant to a pure system represents a perturbation to the energy landscape where the walk between different phases is performed. Therefore, any change in the transition regime reflects the kinetic preference of a given structural motif at times of phase formation. This method was applied to the elucidation of WZ-RS transition mechanism in the series of semiconducting compounds AlN, GaN, and InN. Simulations showed that In atoms adopt the same transformation mechanism as in GaN and favor it, while Al atoms demonstrated a significant reluctance to the path going through tetragonal intermediate iT. The difference between transition regime in mixed systems InxGa1−xN and AlxGa1−xN is in agreement with experiments on high pressure behavior of AlN, GaN, and InN. While transitions in GaN and InN are reversible down to ambient conditions, AlN is stable. The work presented in this thesis constitutes the seed of new perspectives in the understanding of pressure-induced phase transformations in the solid state, where the physics and the chemistry are brought together by means of computer simulations.