PNNL

Abstract

Semiconductor nanostructures are of great interest from both fundamental and applied perspectives. They are intriguing scientifically because their bonding and properties deviate strongly from those of their bulk solids and relevant industrially since nanodevices now approach the scale of large clusters. Understanding the behavior of any molecular system starts from ascertaining its structure. So a colossal effort was expended over the last two decades on characterizing semiconductor cluster geometries. As silicon is the most critical semiconductor, that effort largely focused on Si clusters. Structures of small Si clusters were elucidated early on by ab initio calculations, vibrationally resolved spectroscopy, and optical spectroscopies of matrix-isolated species. Progress for larger systems was enabled in late 1990s by an integrated suite of new tools. It includes ion mobility spectrometry, photoelectron spectroscopy, collisional dissociation, and threshold ionization on the side of experiment, and novel molecular optimization algorithms based on evolution paradigm and fast, but accurate semiempirical protocols for energy evaluation on the theory side. Coherent application of these methods has characterized Si clusters up to the region of radical transition from prolate to spherical growth at ~ 25 atoms.