itle:  Inverse Design of Metamaterials for Tailored Linear and Nonlinear Optical Responses Using Deep Learning

Committee:

Dr. Wenshan Cai, ECE, Chair, Advisor

Dr. Azad Naeemi, ECE

Dr. Zhuomin Zhang, ME

Dr. Ali Adibi, ECE

Dr. Andrew Peterson, ECE

Abstract: The conventional process for developing an optimal design for nonlinear optical responses is based on a trial-and-error approach that is largely inefficient and does not necessarily lead to an ideal result. Deep learning can automate this process and widen the realm of nonlinear geometries and devices. This research illustrates a deep learning framework used to create an optimal plasmonic design for metamaterials with specific desired optical responses, both linear and nonlinear. The algorithm can produce plasmonic patterns that can maximize second-harmonic nonlinear effects of a nonlinear metamaterial. A nanolaminate metamaterial is used as a nonlinear material, and a plasmonic patterns are fabricated on the prepared nanolaminate to demonstrate the validity and efficacy of the deep learning algorithm for second-harmonic generation. Photonic upconversion from the infrared regime to the visible spectrum can occur through sum-frequency generation. The deep learning algorithm was improved to optimize a nonlinear plasmonic metamaterial for sum-frequency generation. The framework was then further expanded using transfer learning to lessen computation resources required to optimize metamaterials for new design parameters. The deep learning architecture applied in this research can be expanded to other optical responses and drive the innovation of novel optical applications.