CALYPSO Surface Reconstruction Prediction Method And Its Applications On Diamond Surfaces

The atomic structure of a surface is its most fundamental physical property since itdefines the actual system under study,in the first place. Most other physical quantities sensi-tively depend on the surface structure. Generally speaking, surfaces of semiconductors arenotoriously complex. Experimental studies are often impeded by the complicated surfacereconstructions that semiconductors undergo. For most surfaces, besides various experi-mental studies, theoretical methods have to be carried out to arrive at convincing surfacestructure models. It is crucially important to develop new method to automatically explorethe surface configurations. However, several challenges exist:(1) The number of possiblesurface structures is very huge even for a relatively small surface. Thus it is impossible tocheck every candidate surface model;(2) Except for few well-studied surface, our knowl-edge about the surface structures is very limited, thus methods rely on educated guesses forstructures initialized by chemical intuition are usually inefficient; and (3) Surface structuresare sensitive to the environments. Surface processes, such as oxidation, further complicatethe determination of surface reconstructions.Here we present an automatic surface structure-searching method that employs struc-ture swarm intelligence, a technique recently developed for finding the global minimumin a complex configuration space(such as crystal, cluster,2D layers, and material design).Search efficiency is improved by implementing symmetry groups and by applying chemi-cal constraints on the surface structures. First of all, we use electron-counting rule to ruleout structures that are physical or chemical unstable. Secondly, a well-designed algorithmare used to generate ordered initial surface structures. The surface structure search mod-ule has incorporated into CALYPSO software packages (free for academic use, please visitwww.calypso.cn).The application of our method to analyse the structure reconstructions of the diamond(100) surface unexpectedly reveals a spontaneous formation of arrays of self-assembledCNTs. The formation energy of such a surface reconstruction is competitive with that ofthe known dimer surface structure. Under certain external conditions (a small compres-sive strain or a high temperature), the selfassembled CNTs surface becomes the more stablereconstruction structure. We unveiled an unique feature of the chemical bondings on this surface, that is the distance between two adjacent tubes are1.788, which is longest C-Cbond reported. Its unique chemical bondings lead to a novel electronic structure. The lowestconduction band has a large dispersion in both axial to the tube (X-S) and radial to the tube(X-) directions, whereas the highest valence band has a large dispersion only in the X-Sdirection and is quite flat in the X-direction. The band structure indicates that the electronscan transport equally well axial and radial to the tubes. In contrast, hole transport is mucheasier along the tubes than across them, showing anisotropic or one dimensional transportbehaviour.Ourfindingshighlightthatthesurfaceplaysvitalrolesinthefabricationofnanodevicesbybeingafunctionalpartofthem. Thisstudymighthelptopursuitthelong-termgoal,whichis semiconductor based on diamond.