Semester

Summer

Date of Graduation

2007

Document Type

Dissertation

Degree Type

PhD

College

Eberly College of Arts and Sciences

Department

Physics and Astronomy

Committee Chair

David Lederman.

Abstract

Cobalt thin films 4 nm thick were deposited via metal electron-beam epitaxy on sapphire (110). The films were annealed above 500°C and scanned in-situ with atomic force microscopy upon annealing. The films underwent a roughening transition above 500°C which is highly time dependent. The film's roughness increased with time with a power law of rho ∼ (t-- t0)alpha, with alpha ranging from 0.16 to 0.21, and de-wetted the sapphire substrate above 590°C. The dynamics of single clusters showed an exponential growth of their height with increasing time at early stages of the annealing process which is consistent with surface instabilities. These instabilities are due to the interplay between surface tension forces (which favor the smoothing out of the surface) and long range Casimir interactions (which favor the thickening and thus roughening of the film). The nature of the Casimir interaction acting within the metallic film was analyzed by calculating the average intercluster separation at early annealing times. The images' autocorrelation function gave an intercluster separation between 150-300 nm, which agrees with expectations resulting from a fermionic Casimir interaction. The fermionic Casimir interaction should be more important for the de-wetting of metallic thin films like cobalt rather than the bosonic Casimir interaction (Van der Waals) that plays an important role in the de-wetting of insulators like polymers. The clusters' evolution for late annealing times showed mainly a growth perpendicular to the surface plane with their areas remaining almost unchanged. The clusters' height increased with a power law h ∼ (t--t 0)beta with beta ∼ 0.25, which in principle is consistent with diffusion in 2 dimensions for a 3 dimensional system (3D/2D), however coalescence and ripening are not observed in the evolution of cobalt thin films. Hence the standard Ostwald ripening theory does not seem to fit well in this system.

Share

COinS