Theoretical Study Of The High-pressure Structural Phase Transitions In Zn-VIA Compounds

Zn-VIA compounds (ZnX, X=O, S, Se and Te) are wide band-gap senmiconductors with great industrial applications. Their properties have been a focused topic in the area of condensed-matter physics. Such compounds adopt the simple but typical structures having subtle connection with the bonding nature inside the crystal. As representatives in the ANB8-N semiconductor family, Zn-IVA compounds exhibit rich transition behaviors under high pressure. It has been shown that, with increasing pressures, the covalent, ionic and metallic bonding prevails subsequently which corresponds to the common transition sequence in such compounds: zinc blende or wurtzite (4 coordination)→NaCl (4 coordination)→Cmcm (5+3 coordination)→CsCl (8 coordination). The Cmcm phase occurs with the metallization in the compounds and its formation can be attributed to a distortion of the NaCl structure driven by soft phonon mode. This structure has been confirmed to widely exist in the IIIA-VA and IIIB-IVA compounds. Due to the limitation in experimental techniques, the Cmcm phase was long been mistaken as ?-Sn phase until the advent of angle-dispersive X-ray powder diffraction technique and image-plate area detector. The CsCl structure is empirically thought to be the strongest candidate in these compounds at large compressions because this structure has very large coordination number which is fit for close-packed arrangement in the binary compounds and also has very low Madelung energy which to some extent resists the energetic instability caused by small volume at high pressures.With the rapid development in first-principles calculation, theoretical investigation of the structures in ANB8-N compounds has surged in the literature. Nevertheless, most of the work was content with static energy calculations, which may miss some unexpected yet more stable structures. On the other hand, phonons are believed to load subtle information on the structural stability and known to be a powerful tool to probe the transition mechanisms. Thus this thesis is intended to extensively explore the structural polymorphs in ZnX and to study the stability of the assumed structures from the point of view of lattice dynamics. Moreover, evolutionary algorithm is used in search for the potential structures hidden from our instinct. Specifically, the work can be divided into the following parts:(I) As the one with the strongest ionicity, ZnO adopts wurtzite structure at ambient conditions and transforms to NaCl phase at ~10 GPa. In analogy to the alkali halides and alkaline-earth oxides, there is an empirical yet widely accepted hypothesis on the NaCl→CsCl transition at even higher pressures. However, there was no evident proof both by experiment and theory. In my work, by means of lattice-dynamics calculation, the direction NaCl→CsCl transition is ruled out while an intermediate phase is proposed in between. By analysis of the soft mode in the CsCl phase, the intermediate phase is predicted to be the ambient PbO phase, which can be viewed as alternatively stacked layers of cations and anions with the anion layer composed of zigzag arranged anionic rows. Enthalpy and phonon calculations have confirmed the stability of this phase. The finding modifies the long thought of the direct NaCl→CsCl transition in ZnO.(II) On the basis of predicted NaCl→PbO transition in ZnO, this work further investigated its transition mechanism. It is revealed that the softened transverse phonon mode at the X point of the Brillouin zone boundary is responsible for the transition. Interestingly, this soft mode is two-fold degenerated and should render the transition to diverse structures. In this work, different combinations of the degenerated mode are analyzed and the one contributes to the NaCl→PbO transition is at last confirmed. Using this method, an orthorhombic transition path with Pmmn symmetry is accordingly revealed. Notably, the reconstructive NaCl→PbO transition can not be fully interpreted by soft mode which always corresponds to second-order displacive transitions. A proper explanation is that strong stress-release is involved in the transition.(III) ZnS, ZnSe and ZnTe have weaker ioniciy but stronger covalence than ZnO does. They also have larger ionic radii and can be metallized more easily under high pressure. To this end, little is known about their behaviors beyond the high-pressure Cmcm phase. In this thesis, their structures were intensively explored using the evolutionary simulations up to the high pressure of 250 GPa. A novel tetragonal phase (space group: P4/nmm; Z = 2) was predicted to occur in ZnS and ZnSe. Further lattice-dynamics calculation confirmed the transition is driven by the soft mode at the Y point in the Brillouin zone of the Cmcm phase. Internal stress release is also thought to be responsible for the transition. The transition can be demonstrated by freezing phonon method along a Pmmn transition path. The revealing of the P4/nmm phase enables us to fairly exclude the occurrence of the previously assumed CsCl phase. (IV) Pressure-induced elemental dissociation is discovered in ZnS, ZnSe and ZnTe. The evolutionary simulations reveals sandwich-arranged structures made up of Zn and X (X=S, Se and Te) atomic blocks at very high pressures, which suggests the elemental dissociation tendency. Further enthalpy calculation confirms the energetic priority of the decomposition phase (Zn+X). This discovery enriches our common knowledge of high pressure, i.e. it is not only a powerful tool to bind different constituents into a compound, but also a novel tool to decompose a compound. In the thesis, the dissociation phenomenon was interpreted by the high metallization of the compounds at large compressions. This argument was supported by the investigation on ZnO that remains insulating up to 800 GPa and shows no sign of elemental dissociation.(V) In ZnTe, the dissociation pressure is supposed to be at 38 GPa by our theoretical calculation while there is no sign of dissociation up to at least 83 GPa in experiment. This apparent contradiction lies in the fact that the theoretical description has no account of the kinetic energy barrier for the transition to occur. It is understandable that the transition to decomposition phase Zn+X should involve massive atomic movements to dissociate the two different elements from their regular alternative arrangement in the solid compound, especially when externally subjected to very high pressure.(VI) Different structures with CsCl-like arrangement were predicted as the candidate post-Cmcm phases in ZnTe under the constraint of different cell sizes. Among the predicted phases, the one with P4/nmm symmetry (2 f.u. cell-l) agrees best with the available experimental data. With increasing the unit cell size, enthalpy of the predicted CsCl-like structures decreases monotonously, which reflects the energetic priority of the decomposition phase Zn+Te and on the other hand suggests that the dissociation could be accomplished via a gradual Zn-Te atomic exchange on the basis of a body-centered cubic lattice.