Surface Modification Of Si Nanostructures And Systemtical Structure Search For Stable Au-Si Compounds

Silicon is the fundamental material for the electrical semiconductor industry.Si based nanostructures and metal-silicon compounds were widely implemented in electrical devices,such as integrated circuit,chip technology,optical devices,metal gate transistor and metal-semiconductor contact devices,etc.Thus,the nano-scale material design and structure prediction have attracted many interest from researchers over past decades.In this work,we designed many various shaped Si nanostructures for surface modification,and performed a global structure search for Au-Si compounds using genetic algorithm.Both combined with density functional calculations for guiding the selection of the candidate structures from trial ones and further property analysis.The genetic algorithm package we used was re-deployed and the workflow of the package and functions are partly introduced in details.Furthermore,a benchmark work of a variation method based Gutzwiller wave function calculation is performed,the method was aimed to improve the accuracy of electron correlation calculation.To sum up,the major working includes:1)Based on the common experimental preparation technique and theoretical researches,two kinds of Si nanostructures with different scales(Si₇₈ and Si₁₇₂)were designed,and modified onto two kinds of Si（111）surfaces（surface growth and surface adsorption）,including grid,island,pit and sphere.The characteristics of different type of structures are evaluated by using the density functional based tight binding method and density functional calculations.The optical properties are analyzed and compared.The results show that the binding energy of the silicon nanostructures that suspected on the surface are generally high,but the optical absorption performance within visible region show better results than those grown on the surface.2)Genetic algorithm based structure searches were performed for the Au-Si binary system,and the structural optimization is carried out with DFT calculation to obtain the energy of each structure.The lower-energy structures selected from each composition are calculated by DFT with refined settings,followed with analyzations of dynamic stability phonon calculations,Bader charges and electronic properties.Finally,it is determined that four new structures can exist stably under normal temperature and pressure,and the convex hull of the binding energy obtained from the theoretical calculation of Au-Si system are given,which may provide a theoretical reference for the relevant material design and research.3)The structure searching package is re-deployed in this work.The major features of the deployments include the compatibility support of new third-party tool software,the adjustment of overall execution procedure,the development of batch analysis package tool,the control and adjustment of lattice parameters,the minimum bond length determine method that used in the process of structure generation,population similarity check and structure matching,and the control of the fixed type of atoms in the core process of structure matching.Furthermore,auxiliary analysis tools under certain application were included,such as direct/Cartesian transforming of atomic coordinates,lattice cell expansion,etc.At present,the package is still under updating.4)Recently,a conjugate gradient minimization method based on Gutzwiller variable wave function is proposed.Under the guidance of this theory,we have carried out the benchmark test.The potential energy curves of nine dimers are calculated.The results are compared with the other methods and available experimental data.The results show that this method can improve the calculation accuracy of dynamic electronic correlation to a certain extent.