Abstract
Ultrashort-pulse laser processing provides a versatile route for nanoscale modification of metals through highly localized energy deposition, rapid heating and cooling, transient stress generation, melting, material ejection, and resolidification. These strongly nonequilibrium processes enable both controlled material removal and the formation of new nanostructures, including nanoparticles generated from ablated material and complex surface features produced on irradiated targets. Despite extensive research on ultrashort-pulse laser processing, a comprehensive mechanistic understanding of the transitions between ablation regimes, nanoparticle generation, and surface morphology evolution is still lacking. The response of metals to femtosecond laser irradiation is governed by a complex interplay of material properties, irradiation conditions, and evolving surface morphology, complicating the interpretation of experimental observations and the establishment of mechanistic links between laser parameters and the resulting nanostructures. This dissertation combines atomistic simulations, physics-guided machine learning, and electromagnetic calculations to investigate laser-induced ablation, nanoparticle formation, and surface morphology evolution in metals.
The first part of the dissertation focuses on nanoparticle generation in femtosecond laser ablation of thin Ag films. Thin metal films are widely used in laser-based nanomanufacturing and provide a unique platform for studying ablation because their thickness can be comparable to the characteristic depths of laser energy deposition, heat transfer, and stress relaxation. Large-scale atomistic simulations performed over a broad range of film thicknesses and laser fluences establish a regime map of thin-film decomposition and ejection. The simulations reveal distinct ablation regimes associated with different material-removal pathways, ranging from stress-driven spallation and rupture of molten layers to partial phase explosion and complete explosive decomposition. These regimes provide a mechanistic explanation for experimentally observed transitions between bimodal and unimodal nanoparticle size distributions. Electromagnetic finite-difference time-domain calculations performed for atomistic configurations generated in the simulations further predict transient reflectance signatures that can be directly compared with pump-probe optical imaging measurements.
The second part of the dissertation investigates the factors controlling surface morphology formation in metals irradiated by ultrashort laser pulses. Surface nanostructures emerge through the coupled effects of energy deposition, stress generation, melt flow, capillary instabilities, and rapid resolidification. To identify the material properties and irradiation conditions governing these processes, a physics-guided machine learning framework is developed using a compact set of dimensionless descriptors derived from dimensional analysis. Machine learning models trained on experimentally measured morphology data relate the density of surface nanoprotrusions to descriptors representing key stages of the laser-material response. Interpretation of the trained models identifies the most influential parameters and highlights the roles of absorbed fluence, viscous and capillary forces, and electron-phonon coupling. Large-scale two-temperature molecular dynamics (TTM-MD) simulations of Mo validate the predicted sensitivity of surface morphology to electron-phonon coupling near the spallation threshold and reveal the underlying atomistic mechanisms.
The third part of the dissertation examines how pre-existing surface structures modify the response of a Mo target to a subsequent femtosecond laser pulse. In multipulse processing, structures generated by earlier pulses alter local electromagnetic absorption, heat transfer, stress generation, and molten material dynamics during later irradiation. Atomistic simulations are used to investigate second-pulse interactions with surface morphologies produced by an initial pulse in both sub-spallation and spallation regimes. The simulations provide evidence that pre-existing surface features may influence subsequent energy deposition, melting, stress relaxation, material ejection, and the continued growth, reshaping, or removal of nanostructures. Electromagnetic wave calculations integrated into the TTM-MD framework reveal how evolving surface topography redistributes absorbed laser energy and generates localized absorption hot spots that drive subsequent morphology evolution.
Together, these studies demonstrate that nanoparticle generation and surface nanostructuring are interconnected outcomes of the same nonequilibrium processes of energy deposition, stress relaxation, melting, material redistribution, and resolidification. By combining atomistic simulations, physics-guided machine learning, and electromagnetic calculations, this dissertation identifies the mechanisms governing transitions between thin-film ablation regimes, nanoparticle formation, and surface morphology evolution. The results establish mechanistic links between irradiation conditions, material properties, transient distributions of absorbed energy, and the resulting nanostructures, contributing to the broader goal of establishing quantitative relationships between laser-processing conditions and resulting nanostructures.
