Quasi-Phase-Matching Efficiency Optimization for Coupled Second-Order Nonlinear Processes

It is well known how to tailor the conversion efficiency grid of a single quasi-phase-matching (QPM) grating when the involved processes are uncoupled. However, it becomes much more sophisticated in the presence of coupling between multiple processes. In this case, different processes compete for the same QPM “resources” throughout the grating, and one process can outweigh the others over a certain range of interaction. Here we propose the generalized iterative domino (GID) algorithm to meet these challenges for the first time (to our best knowledge). Instead of tailoring the strength of each “global” Fourier coefficient, GID algorithm can properly adjust the spatially varying “local” Fourier coefficients in favor of the final yield. Three methods, including cascaded single-period (C1P) structure, quasi-periodic optical superlattice (QPOS), and hyperfine aperiodic optical superlattice (HAOS) optimized by GID, are numerically and experimentally investigated under the platform of third-harmonic generation (THG). It shows that the THG efficiency of HAOS + GID can exceed the record achieved by C1P structure by 33%. This method is applicable to general wavelength converters involving with multiple coupled nonlinear processes.