Self-organized growth of Ag islands on Si(1 1 1)-(7 × 7)-optimization of an STM experiment by means of KMC simulations

Abstract

A growth model and parameters obtained in our previous experimental (STM) and theoretical (Kinetic Monte Carlo simulations) studies of Ag/ Si(111)-(7×7) heteroepitaxy were used to optimise growth conditions (temperature and deposition rate) for the most ordered self-organized growth of Ag island arrays on the (7×7) reconstructed surface. The conditions were estimated by means of KMC simulations using the preference in occupation of half unit cells (HUCs) of F-type as a criterion of island ordering. Morphology of experimentally prepared island structures was studied by STM. High degree of experimentally obtained island ordering is compared with the simulated data and results are discussed with respect to the model and parameters used at the KMC simulations. IntroductionA lot of effort has been paid to prepare ordered arrays of metal nanoclusters because of interesting possible applications in the field of microelectronics. Reconstructed Si(111)-(7× 7) surface was used as a template for spontaneous ordering. Recently ordered arrays of In, Al, Ga identical clusters have been prepared [1,2]. Highly ordered arrays of identical bimetal clusters of In/Ag and In/Mn were also reported in [2]. Growth of the identical clusters is mediated by the existence of particular magic sizes of atom clusters that are more stable than the others. In case of Tl [3], Sn [4] and Ag [5,6] more or less ordered island arrays were reported but magic clustering was not observed. Preparation of ordered structures requires a delicate control of deposition conditions. Kinetics of self-ordering of clusters on the reconstructed surface depends on many parameters and theoretical description of a role of deposition conditions is difficult and still not available. Kinetic Monte Carlo simulation (KMC) of the growth represents an effective approach for theoretical study of important processes involved. We used the KMC simulations together with growth experiments for studying nucleation and heteroepitaxial growth of Ag on the Si(111)-(7×7) surface [5,6,7,8].The (7×7) reconstructed surface consists of triangular half-unit cells (HUCs), which represent potential wells for adsorbed atoms hopping on the surface [9,10]. Each HUC contains six Si adatoms of a top layer. The HUCs are of two types, one containing structural fault (FHUC), the other is unfaulted (UHUC) [11]. It results in different reactivities and consequently in preferential nucleation of Ag in FHUCs. The Ag adatom hopping between various adsorption positions within a HUC at room and higher temperatures is easy but the mobility of the Ag adatoms on the surface -important for growth processes -is determined by the hopping rate between HUCs (depends on barrier height). In a coarse grained model, we used for the simulations, the surface is represented by adsorption sites of two types (FHUCs, UHUCs) with different barrier heights. The preference in FHUC occupancy, P F , (ratio of number of FHUCs occupied by Ag clusters to number of all occupied HUCs) is determined by the barrier difference...

Thermal switching of the electrical conductivity of Si(111)()Ag due to a surface phase transition

“…After deposition for 180 s (or at a nominal coverage of three monolayers), this model indicates that the surface was covered by a network of Ag islands with a preferential height of five monolayers, which just exceeded the percolation threshold. The result from our model is consistent with the available STM, optical and photoelectron diffraction (PED) measurements [27][28][29][30].…”

Scanning tunneling microscopy at multiple voltage biases of stable “ring-like” Ag clusters on Si(111)–(7×7)

“…To optimize the growth of arrays of well-defined clusters we have varied the deposition rate, the substrate temperature during evaporation, the total Ag coverage and the annealing temperature. Based on the work of Kocán et al [21] and preliminary measurements, we have established that the deposition rate should be minimized to increase preference for FHUC and overall order. The deposition rate was 4 • 10 −5 1ML • s −1 (ML here corresponds to 6.7 atoms/nm 2 = 6.7 • 10 14 atoms/cm 2 ).…”

Growth of honeycomb-symmetrical Mn nanodots arrays on Si(111)-7 × 7 surfaces

“…Our simulation consists of growth in 450K and a subsequent annealing process at room temperature for an hour. By fixing [14] and changing E F , E U , E a , randomly, we found a least square between simulation and experimental data points. The simulated P F as a function of deposition rate is shown in Fig.…”