AbstractThe purpose of this Ph.D. thesis is the development of a new generation of predictive tools dedicated to thin film deposition by vacuum-based techniques. Even though these processes are one of the main pillars of our technological society, the understanding of these techniques has not yet come to an end. There are still grey areas that today’s scientific knowledge cannot describe or explain. However, the advent of computer science opens up the door to a brand-new kind of approaches, allowing scientists to simulate complex processes in order to unravel their understanding. With the evolution of High-Performance Computing resources, the complexity of the numerical models is continuously growing. The former basic 1D or 2D models can now give way to realistic 3D models including a very detailed physics and providing scientist new ways to observe and see the processes governing the universe. Rather to aim to explain the whole universe, this PhD thesis focuses on the reactive magnetron sputtering technique which is a widely used method for the deposition of various compound layers both in laboratories and in industries. Therefore, its understanding is of great importance to master the quality and the properties of the products. The involved physical processes have complex and non-linear nature requiring the use of a wide range of simulation techniques and models. This PhD work established a 3D multi-scale simulation chain of plasma deposition process (a.k.a. Virtual CoaterTM), based on a combination of Particle-In-Cell Monte Carlo (plasma phase) algorithms and a kinetic Monte Carlo (film growth) code.
The first step of this work demonstrates the connections between the several codes in order to provide a global picture of the deposition process. In a second time, attention is focused on the role of neutral particles during the deposition process. Experimental characterization of both the plasma phase and the film properties are performed and compared together with simulation results. The experimental results are in agreement with the simulated ones. For a given coater, the plasma phase hysteresis behaviour, film composition and film morphology are predicted and explained.
In a third time, the challenges arisen when modelling charged particles are highlighted and addressed in a dedicated section. The various fluxes of particles flowing towards the substrate during a plasma deposition process occurring at realistic power density are proved to be scalable from 3D PICMC simulations operating at lower power density. The simulations also feature propagating plasma instabilities, so-called spokes. The validity domain of the scaling strategy is discussed in the light of the model constraints.
The last step successfully applies the established 3D multi-scale simulation chain to the growth of TiO2 thin films by means of reactive magnetron sputtering. The model efficiently predicts the densities and fluxes of both charged and neutral particles towards the substrate. It also enables to explain the changes of properties of the deposited films throughout the transition from metallic deposition to stoichiometric TiO2. Moreover, the high energy negative atomic oxygen ions originating from the targets are identified as origin of the abnormally low inclination of the columnar structure experimentally observed for the oxide mode coatings. Also, the change of the crystallographic arrangement of the coatings during the plasma transition from amorphous to anatase is observed by x-ray diffraction. This modification is explained by an increase of the normalized energy flux (NEF) at the substrate location.
Altogether, the development of Virtual Coater through this work enable the improvement of our knowledge of the interaction between plasma properties and thin films quality. This proposed new simulation procedure paved the way to new methods of studying the plasma state as a deposition process.
|Date of Award||18 Jan 2021|
|Sponsors||Université de Namur|
|Supervisor||Stephane Lucas (Supervisor), Andreas Pflug (Co-Supervisor), Luc Henrard (President), Pavel Moskovkin (Jury), Wilmert De Bosscher (Jury), Stephanos Konstantinidis (Jury) & Achim von Keudell (Jury)|
- Monte Carlo
- Magnetron sputtering
Attachment to an Research Institute in UNAMUR