Abstract:
Modeling complex nanophotonics systems : from large scale simulations to machine learning
Numerical modelling is an important tool for understanding light matter interaction in complex nanoscale systems. Simulations bring insight to experimental results by allowing one tease out the most relevant underlying phenomena in very complex scenarios. In this talk, I will focus on one example, relating to the creation of beautiful colors and artwork on metal surfaces and coins via a laser writing process that creates very complex and random nanostructured surfaces. Our computations were able to identify that plasmonic effects were key to colour creation, and able to explain experimental trends. Furthermore, using the experimental and simulation data generated from this work, we used machine learning to predict new colours, though an iterative inverse design algorithm that we developed.