Investigations On The Chemical Bonds, Crystal Structures And Mechanical Properties Of Hard Crystals

This dissertation concentrates on two different topics: the first part is construction of the relations between macroscopic mechanical properties (including hardness and strength) and microscopic structural parameters of hard crystals in the view of chemical bonding, and the second part is prediction of several novel metastable materials, especially superhard materials, by first principles calculations.Connecting the macroscopic mechanical properties with the microscopic structural parameters not only can develop the theoretical methods of quantitatively predicting the mechanical properties of a crystal, but also help us to deeply understand their physical natures. To combine hardness semiempirical equations and population ionicity scale of a chemical bond makes the hardness predictions possible, in which all parameters can be calculated by first principles. Based on the previous works of our group, we further study the determination of population P_c in boron rich compounds and wurtzite-structures and the intrinsic factors influencing the chemical bond ionicity, predict the theoretical hardness of B₆O, B₁₃C₂, BC₂N, wurtzite semiconductors, B₆N, B₆P and B₆As complex crystals, where the theoretical hardness agrees well with the experimental Vicker's hardness for B₆O, B₁₃C₂, BC₂N and wurtzite semiconductors.For transition metal carbides and nitrides, it was found that the semiempirical model is unsuitable for evaluating their hardness. By exploring the difference of chemical bond feature between these crystals and the covalent crystals of diamond and c-BN, we find another two factors determining the hardness of these crystals. One is a negative factor that a small metallic component of chemical bonds decreases obviously hardness, and second is a positive factor that d valence electrons increase hardness. After introducing the two factors, a universal hardness expression is established for covalence-dominant crystals.The abnormal enhancement of superlattice hardness attracts great attention of physical and material scientists. How to expand the hardness model of a crystal to treat the issue of superlattice hardness is a challenge work. By introducing the quantum confinement effect into the hardness calculation of superlattice, we derive an empirical model describing the superlattice hardness. In this model, the existence of the quantum confinement effect enhances significantly the hardness of superlattices with small stipulation period.Strength is another important mechanical property of crystals. Both strength and hardness is dependent on bond strength. In this dissertation, we define the bond strength as the maximum tensile force that a chemical bond can suffer when tensile deformation is along the axis of the bond. By introducing a new concept, effective bonding valence electrons, we construct the quantitative relationship between the bond strength and the bond length and the effective bonding valence electrons. This model can describe the bond strength of pure covalent, polar covalent and ionic bonds, and can predict the tensile strength of some crystals along specific directions. Compared to the cohesive energy of covalent crystals and the lattice energy of ionic crystals, the bond strength calculated with this model has the same tendency to them.Based on the five reported C₃N₄ structures, we construct five phases of B₄C₃. The equilibrium lattice constants, cohesive energy and electron properties are obtained by performing first-principles calculations. Theoretical hardness of four semiconducting B₄C₃ phases is estimated by using the semiempirical hardness model, where the cubic and cubic spinal phase has the hardness comparable to that of cubic boron nitride. In order to determine the crystal structure of previously-synthesized OsC, we construct seven possible OsC structures. Based on total energy calculation, mechanical stability judgment, X ray diffraction simulation, and hardness calculation, we conclude that the experimentally-synthesized OsC should be in NiAs structure not in WC structure.