Thin metal films on weakly-interacting substrates : Nanoscale growth dynamics, stress generation, and morphology manipulation

Sammanfattning: Vapor-based growth of thin metal films with controlled morphology on weakly-interacting substrates (WIS), including oxides and van der Waals materials, is essential for the fabrication of multifunctional metal contacts in a wide array of optoelectronic devices. Achieving this entails a great challenge, since weak film/substrate interactions yield a pronounced and uncontrolled 3D morphology. Moreover, the far-from-equilibrium nature of vapor-based film growth often leads to generation of mechanical stress, which may further compromise device reliability and functionality. The objectives of this thesis are related to metal film growth on WIS and seek to: (i) contribute to the understanding of atomic-scale processes that control film morphological evolution; (ii) elucidate the dynamic competition between nanoscale processes that govern film stress generation and evolution; and (iii) develop methodologies for manipulating and controlling nanoscale film morphology between 2D and 3D. Investigations focus on magnetron sputter-deposited Ag and Cu films on SiO2 and amorphous carbon (a-C) substrates. Research is conducted by strategically combining of in situ and real-time film growth monitoring, ex situ chemical and (micro)-structural analysis, optical modelling, and deterministic growth simulations.In the first part, the scaling behavior of characteristic morphological transition thicknesses (i.e., percolation and continuous film formation thickness) during growth of Ag and Cu films on a-C are established as function of deposition rate and temperature. These data are interpreted using a theoretical framework based on the droplet growth theory and the kinetic freezing model for island coalescence, from which the diffusion rates of film forming species during Ag and Cu growth are estimated. By combining experimental data with ab initio molecular dynamics simulations, diffusion of multiatomic clusters, rather than monomers, is identified as the rate-limiting structure-forming process.In the second part, the effect of minority metallic or gaseous species (Cu, N2, O2) on Ag film morphological evolution on SiO2 is studied. By employing in situ spectroscopic ellipsometry, it is found that addition of minority species at the film growth front promotes 2D morphology, but also yields an increased continuous-layer resistivity. Ex situ analyses show that 2D morphology is favored because minority species hinder the rate of coalescence completion. Hence, a novel growth manipulation strategy is compiled in which minority species are deployed with high temporal precision to selectively target specific film growth stages and achieve 2D morphology, while retaining opto-electronic properties of pure Ag films.In the third part, the evolution of stress during Ag and Cu film growth on a-C and its dependence on growth kinetics (as determined by deposition rate, substrate temperature) is systematically investigated. A general trend toward smaller compressive stress magnitudes with increasing temperature/deposition rate is found, related to increasing grain size/decreasing adatom diffusion length. Exception to this trend is found for Cu films, in which oxygen incorporation from the residual growth atmosphere at low deposition rates inhibits adatom diffusivity and decreases the magnitude of compressive stress. The effect of N2 on stress type and magnitude in Ag films is also studied. While Ag grown in N2-free atmosphere exhibits a typical compressive-tensile-compressive stress evolution as function of thickness, addition of a few percent of N2 yields to a stress turnaround from compressive to tensile stress after film continuity which is attributed to giant grain growth and film roughening.The overall results of the thesis provide the foundation to: (i) determine diffusion rates over a wide range of WIS film/substrates systems; (ii) design non-invasive strategies for multifunctional contacts in optoelectronic devices; (iii) complete important missing pieces in the fundamental understanding of stress, which can be used to expand theoretical descriptions for predicting and tuning stress magnitude.