Abstract
We simulate the amorphous–crystal interface in silicon using a combination of interatomic potential molecular-dynamics and tight- binding conjugate-gradient relaxation. The samples we create have high quality crystalline and amorphous portions. We develop some localized measures of order to characterize the interface, including a missing neighbor vector and the bond angle deviation. We find that the measures of order interpolate smoothly from a bulk crystal value to a bulk amorphous value across a 7 A thick interface region. The interface structures exhibit a number of interesting features. The crystal planes near the interface are nearly perfect, with a few dimer defects similar to the Si(100) 2×1 reconstruction. Interfaces produced with one constant temperature simulation method are rough, with several layers of atoms forming chains and (111) facets. A different simulation method produces more planar interfaces with only a few chains.
Highlights
Silicon, often considered the prototypical covalent semiconductor, has been extensively studied both for its inherent scienti c interest and because it is the basis for a wide range of technologies
The process by which the amorphous phase becomes crystalline by propagation of the interface, which is referred to as solid phase epitaxial growth SPEG, has been the subject of many experiments
We rapidly quench the molten portion with the modi cation of the bond angle forces suggested by Luedtke and Landman 5 to ensure that the quenched part of the system enters the amorphous phase, and anneal the amorphous portion at 1000 K
Summary
Often considered the prototypical covalent semiconductor, has been extensively studied both for its inherent scienti c interest and because it is the basis for a wide range of technologies. The process by which the amorphous phase becomes crystalline by propagation of the interface, which is referred to as solid phase epitaxial growth SPEG, has been the subject of many experiments. It is a thermally activated process with an activation energy of 2.7 eV over ten orders of magnitude in interface speed, strongly indicating a unique growth mechanism 1, 2. The activation strain and e ects of interface orientation and doping have been studied experimentally. Despite all this evidence the microscopic atomic mechanism responsible for SPEG is still unclear. The rst step towards such an understanding of SPEG is the creation and characterization of realistic atomistic models of the amorphous-crystalline interface
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
