Establishment And Application Of Testering Potential Model

With the development of computer hardware, scientists can design new materials with special physical and chemical properties based on the atomic and molecular configuration, which opens up an effective way for materials design. However, the existing theoretical models cannot yet answer the possibility that the designed materials with specific atomic and molecular conformation can be grown up into crystal. For example, Cohen predicted theoretically that C3N4crystal is as hard as diamond via the calculations of the atomic and molecular structure, but did not tell us how difficult to prepare its single crystal. Up to date, no C3N4crystal with a size as large as ten microns is obtained though great deals of experimental efforts have been made to prepare the material. It is still unclear if C3N4has greater ability to form single crystals than diamonds. In practice, different materials show considerable difference in the ability to form single crystals. For example, Si is easy to form a single crystal with a few meters, while the diamond grains of several milimetresare difficult to be grown up. Obviously, in the process of material design, we should not only focus on the atomic structure of the new materials, but also predict the ability to form single crystal. How to predict the ability and evaluate their properties such as melting temperature and hardness challenges current theories of material design.To address these issues, we made the following studies:(1) Based on the fact that when a single atom stays on the crystal surface it "feels" the transverse and longitudinal potential field of the crystal surface, a theoretical model to predict the ability of materials to form single crystal (Condensing Potential Model) is established. By molecular dynamics (MD) simulations of crystal growth for mono-component materials, comparing with the experimental phenomena, the model shows its reliability.(2) Condensing potential (CP) model was extended to predict the ability of binary material to form single crystal and its reliability was verified by MD simulations of crystal growth of Ni-Al and Cu-Au system, which successfully explained the experimental phenomena that A1doping no more than6%in the nickel-base single crystal superalloys.(3) Since the quantum effects of nanostructures depend on their sizes and geometric structures, predicting and preparing desired nanostructures on solid surfaces play important roles in many advanced material technologies. CP model was applied to predict the geometric shape of the two-dimensional atomic island on the crystal surface, and successfully predicted the homoepitaxial compact atomic islands on Pt (111) are triangular shape at400K, in accord with experimental observations, while widely used Wulff construction predicts them as hexagons.(4) The catalytic efficiency of the nanoparticles depends on their surface structure, for example, the catalytic efficiency of the Pt (111) surface is higher than that of Pt (100). So, a theoretical method is required to predict the surface structure of the nanoparticles to find possible ways to control the surface growth. CP model was used to predict the surface structure of Pt and Ni nanoparticles, and the results that the surface of Pt and Ni nanoparticles in any size should be mainly composed of (111) faces are in coincidence with the observations of transmission electron microscopy.(5) The melting point is an important property of materials and melting is born to be a key subject in condensed matter physics. It has been shown that commonly used Lindemann and Born criterion are difficult to predict the melting point accurately. The application of CP model to predict the melting point of solid and the introduction of the new concept "surface melting potential" for the prediction of surface melt temperatures prove CP model as a powerful tool to address the issue.(6) Based on CP model, a concept of condensing force (CF) was introduced to evaluate the hardness of materials. It is shown via ab initio calculations for fcc (Ni, Cu, Al, Ir, Rh, Au, Ag, Pd) and hcp Re crystals that materials with larger CF can have greater hardness.Due to the flexibility of the CP model in predicting the various properties of the material and its simplicity of the calculations, which can be easy to transplant into the semi-empirical and ab initio MD calculations, CP model is expected to be widely used in the design of new materials.