Buffer-layer-assisted nanostructure growth via two-dimensional cluster–cluster aggregation

• Published in: Surface science Vol. 518; no. 3; pp. 243 - 250
• Main Authors: Haley, Christina, Weaver, J.H
• Format: Journal Article
• Language: English
• Published: Lausanne Elsevier B.V 20-10-2002; Amsterdam Elsevier Science; New York, NY
• Subjects: Adsorption and desorption kinetics; evaporation and condensation; Clusters; Condensed matter: structure, mechanical and thermal properties; Electron microscopy; Evaporation and sublimation; Exact sciences and technology; Gold; Growth; Physics; Solid-fluid interfaces; Surfaces and interfaces; thin films and whiskers (structure and nonelectronic properties); Clusters; Gold; Electron microscopy; Growth; Evaporation and sublimation; Carbonaceous materials; Reactive evaporation; Theoretical study; Metal clusters; Experimental study; Monte Carlo methods; Transmission electron microscopy; Fractal dimension; Transition elements; Gas solid interface; Physical vapor deposition; Atomic clusters
• ISSN: 0039-6028; 1879-2758
• DOI: 10.1016/S0039-6028(02)02197-0

Physical vapor deposition of Au onto Xe multilayers on amorphous carbon at 20 K produces three-dimensional nanoclusters. Warming to room temperature desorbs the Xe and imparts limited mobility to the clusters. Coalescence occurs when two clusters come in contact, and the extent of coalescence depends on the buffer layer thickness. Using transmission electron microscopy images, we determined the spatial distribution of clusters as a function of Xe thickness and applied the scaling concepts of cluster–cluster aggregation to better understand processes associated with growth. Analysis shows a coverage dependent fractal dimension that extends from
D∼1.42 to 1.72 for initial fractional surface coverages of
ρ
0∼0.04–0.21, consistent with Monte Carlo simulations of two-dimensional diffusion-limited cluster aggregation (DLCA). Both the total number of clusters per unit area and the weighted average cluster size show a power law dependence on the Xe layer thickness, where the buffer layer thickness plays the role of time in DLCA modeling. These relationships will facilitate the design of nanostructure arrays generated by desorption-assisted coalescence.