Optimization of Laser Induced Forward Transfer by Finite Element Modeling

Sammanfattning: This thesis explains a comprehensive study on the thermal modeling aspects of Laser Induced Forward Transfer (LIFT), a laser direct write technique. The LIFT process utilizes a focused laser beam to transfer a donor material coated on a laser wavelength transparent substrate to a receiver substrate kept in close proximity. COMSOL Multiphysics is used to implement a two-dimensional time dependent surface and volumetric heat generation model. The thermal model covers the laser induced heating in a pure copper donor material for nanosecond and picosecond pulsed Nd:YAG (Neodymium doped Yttrium Aluminium Garnett) lasers operating at 355 nm wavelength. The model is used to understand the molten regime of the donor material during LIFT process with a stationary laser beam of Gaussian profile in temporal and spatial domain. The input parameters used in the model include both temperature-dependent as well as temperature-independent thermophysical material properties such as heat capacity and thermal conductivity. In addition, theoretical investigations are done to study the optical properties of the material such as absorption coefficient and reflectivity. Simulations are done by changing pulse length, energy per pulse, donor layer thickness and wavelength. Investigation of the heat loss in the donor substrate is also carried out. The influence of wavelengths for 355 nm, 532 nm and 1064 nm at a fixed laser fluence and thickness is also studied. The simulation result shows strengths and weaknesses of both nanosecond and picosecond systems. A picosecond pulse is much dependent on thickness of the material (a few 100 nanometers) whereas a nanosecond pulse is capable of melting a thicker layer (a few micrometers). Choosing a particular laser pulse depends entirely on the type of applications and requirements. It is observed that the peak surface temperature increases linearly with increasing fluence and falls exponentially with increasing donor layer thickness. Also, it is seen that the longer wavelengths require more energy to reach melting temperature at same fluence and thickness due to increased value of reflectivity and less energetic photons. The simulated results give a good approximation to the experimental results of copper LIFT. The model can be used for other materials also by using the relevant material properties.