Binding sites for SiH 2 /Si(0 0 1): A combined ab initio, tight-binding, and classical investigation S. Cereda a , F. Montalenti a, * , M. Cogoni a , D. Branduardi b , M.W. Radny c , P.V. Smith c , Leo Miglio a a L-NESS and Dipartimento di Scienza dei Materiali della Universita ` degli Studi di Milano-Bicocca, Via Cozzi 53, I-20125 Milano, Italy b Department of Chemistry and Applied Biosciences, ETH Zu ¨ rich, USI Campus, Via Giuseppe Buffi 13, Lugano CH-6900, Switzerland c School of Mathematical and Physical Sciences, The University of Newcastle, Callaghan, NSW 2308, Australia Received 30 March 2006; accepted for publication 10 July 2006 Available online 28 July 2006 Abstract Binding sites for SiH 2 on Si(0 0 1) are investigated theoretically by using several different methods. Possible local minima are first sam- pled by classical molecular dynamics simulations of the SiH 2 /Si(0 0 1) impact, allowing for a preliminary, fast selection. A further refine- ment is carried out by geometry optimizations using semiempirical tight-binding and density functional theory calculations, based on both the local density and generalized gradient approximations. In most cases only minor morphological changes are obtained when comparing the ab initio sites with the classical potentials and tight-binding ones. The purely classical treatments here tested, however, overestimate the number of minima and fail in accurately reproducing the relative energy of some of the adsorption sites. Closer agree- ment is obtained with tight-binding, with the noticeable exception of the lowest ab initio minimum (on-dimer site). Ó 2006 Elsevier B.V. All rights reserved. Keywords: Silicon; Atom–surface interaction; Molecular dynamics; SiH 2 binding sites; Density functional theory; Semiempirical tight-binding; Empirical calculations 1. Introduction Silicon thin-film growth on Si(0 0 1) is one of the most widely investigated processes in physical chemistry, due to its fundamental importance in microelectronics, opto- electronics, and solar-cell development. Depending on the application, there is interest in growing strained and unstrained, hydrogenated and hydrogen-free, crystalline, microcrystalline, and amorphous silicon films. Unsurpris- ingly then, there exists a fast-growing community of researchers devoted to the simulation of the growth process under several different conditions, length and time scales, with the final aim of suggesting ideal growth parameters for obtaining a film with the desired structural and chemical properties. While molecular beam epitaxy (MBE) techniques are of- ten employed to grow Si films in basic-science experiments, the industry-driven call for high deposition rates produced great interest in chemical vapour deposition (CVD) growth techniques and, in particular, in plasma-enhanced chemical vapour deposition (PECVD) where Si-film growth is origi- nated by neutral and ionized species such as SiH, SiH 2 and SiH 3 [1,2]. From the theoretical point of view, several dif- ferent approaches can be used. The development of fast computers has made it possible to undertake growth simu- lation from an atomistic perspective, i.e. by explicitly con- sidering the behavior of every individual atom of the system. The description of the interatomic forces makes the big difference among alternative atomistic approaches. Ab initio calculations allow one to consider up to a few 0039-6028/$ - see front matter Ó 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.susc.2006.07.009 * Corresponding author. Fax: +39 026 448 5403. E-mail address: francesco.montalenti@unimib.it (F. Montalenti). www.elsevier.com/locate/susc Surface Science 600 (2006) 4445–4453