Research Objective
Interfaces in materials, such as surfaces, grain boundaries, stacking faults and twin boundaries, control many properties. These include strength, growth dynamics, ductility, and electronic states. We are currently addressing the accurate quantum mechanical modeling of these features on the scale of thousands of atoms. Only with the advent of massively parallel computers and order-N electronic structure methods have such first principles and tight-binding calculations become feasible.
Computational Approach
Our recent advances in algorithms, combined with the power of the new Cray T3E computers at NERSC, allow us to perform first-principles plane wave pseudopotential calculations using up to 1000 Si atoms. We are currently implementing our density-matrix, order-N tight-binding molecular dynamics code on the T3E, and are also examining parallel matrix diagonalization techniques for practical, direct solutions for the tight-binding electronic structure problem.
Accomplishments
The first-principles code is now in production mode on the T3E. We have completed the first-principles calculation of the structure and energy of eight different (510) grain boundary structures, for both Si and Ge. Each of these calculations involves approximately 400 atoms. A new tight-binding potential is being tested against these calculations, and reproduces the energetics quite well. We have begun calculating the electronic structure of the Si 7x7 reconstructed (111) surface, using a system of approximately 500 atoms, to understand surface electronic states. We are doing finite temperature simulations on this system to understand the phase transitions taking place on Si(111).
Significance
The ability to evaluate the consequences of extended material defects, requiring large numbers of atoms to simulate, is an important challenge for materials science. We are currently using the tools and techniques applied here to such problems, using fundamental science to solve key technical issues involving mechanical and electronic properties.
Publications
J. R. Morris, K. M. Ho and C. L. Fu. 1996.
Tight-binding study of tilt grain boundaries in diamond
Phys. Rev. B 54:132.
J. R. Morris, J. Scharff, K. M. Ho, D. E. Turner, Y. Y. Ye and M. H.
Yoo. 1997. Prediction of a {11-22} hcp stacking fault using a
modified generalized stacking-fault calculation.
Phil. Mag. A 76:1065.
J. R. Morris, Z.-Y. Lu, D. M. Ring, J.-B. Xiang, K.-M. Ho. C.-Z. Wang,
and C.-L. Fu. N.d.
First-principles determination of the {510} symmetric tilt boundary
structure in Si and Ge. In preparation.
Atomic structure of the lowest energy (510) symmetric tilt boundary in Si.