Metallic thin films play an important role in devices such as microelectromechanical systems (MEMS). Metallic Cu films, for example, transport the electric current between the semiconductor structures and if they are thick enough can also act as heat sinks for the thermal energy generated by Joule heating. During the lifetime of miniaturized Cu structures these are exposed to several billion thermal cycles, which can cause cracks and subsequently premature failure if the strength of the metal structure is not sufficient. Consequently in this thesis the mechanical properties of Cu thin film structures with different grain sizes are studied to gain insight into the deformation mechanisms, allowing lifetime predictions and optimization of the material behavior. Testing of the Cu films was done by shaping tensile samples and bending beams by a lithographic process and applying miniaturized mechanical experiments mainly conducted inside a scanning electron microscope (SEM).
For analysis of the static mechanical properties, such as yield strength, ultimate tensile strength and elongation to fracture, tensile tests between 143 and 873 K were performed. These experiments gave insight into the activation energies as well as the activation volumes, and subsequently to the underlying deformation mechanisms. The experiments above room temperature revealed an embrittlement in the material with 2.7 µm grain size. Chemical analysis and calculations concerning segregation, work of separation and diffusion indicate a segregation of non-metallic impurities to the grain boundaries.
As the dynamic properties are of great interest for practical applications, cyclic creep experiments (ratcheting) between 293 and 673 K were performed, pointing out the same deformation mechanisms as found in the static experiments. The lifetime, which strongly depends on the creep strain per cycle, was described by a modified Basquin equation. However, as the ratio between the maximum and minimum stress is negative in real life Cu microstructures, also bending experiments were conducted to mimic the alternating stress conditions. In contrast to the ratcheting experiments, these show the extensive formation of intrusions and extrusions on the surface and cell structures inside the freestanding Cu beams.
In situ like electron backscatter diffraction (EBSD) studies during the static and cyclic experiments provided insight into the evolution of the geometrically necessary dislocation (GND) density and the microstructure evolution of the surface grains, which dominate the strength of small scale samples. The deformation behavior recorded at the surface of the samples in the in situ SEM experiments, measurements of the GND density, deduced activation volumes and energies point all towards dislocation plasticity as the dominating deformation mechanism. However, with decreasing grain size grain diffusion comes increasingly into play at temperatures exceeding room temperature.
The results of this thesis are enclosed as four publications. Publication A reports on the grain boundary embrittlement observed in static experiments between 293 and 673 K. Publication B demonstrates the deformation behavior in static and dynamic tension-tension experiments at room temperature, while Publication C extends this study into before unreached temperature regimes for micron-sized samples with experiments being performed between 143 and 873 K. Finally, Publication D presents dynamic bending experiments with in situ like EBSD at room temperature.The present thesis demonstrates that the microstructural design must be performed carefully as the effects of the microstructure are changing with the temperature. Static and dynamic experiments accompanied by in situ SEM/EBSD with different stress-ratios gave insight into deformation mechanisms and pointed out a change from bulk-like to small scale behavior in the analyzed grain size regime.