In 1960, Maiman demonstration of the first laser opened the door to the discovery of nonlinear optical processes. Only one year later, Franken et al. used a ruby laser focused into a quartz crystal to convert light at half the original wavelength, making the first demonstration of frequency conversion. Since then, research on nonlinear optics has expanded rapidly, driven both by fundamental discoveries and technological applications. A crucial concept in this field is phase-matching, which ensures that the generated nonlinear signal remains in phase with the incident light, allowing the emission to build up coherently through the crystal and enabling high conversion efficiencies. 

Recently, metasurfaces have emerged as a promising platform to control light-matter interactions, including nonlinear frequency conversion. Metasurfaces are artificially structured, ultrathin layers composed of periodic arrays of nanostructures. Unlike bulk crystals, nonlinear metasurfaces do not rely on the phase-matching condition to increase their conversion efficiency; instead, they achieve enhanced efficiency by engineering resonant fields within carefully designed nanostructures. However, traditional designs strategies based on semi-analytical models and forward design approaches face limitations, particularly when multi objectives such as efficiency, directionality, and polarisation are simultaneously required.

This project aims to overcome these challenges by employing inverse design and machine learning methods to optimise nonlinear metasurfaces. These advanced approaches can generate highly non-intuitive free-form nanostructures, that outperform conventional designs. The student will actively contribute to the fabrication and experimental characterisation of such free-form metasurfaces, to ultimately develop and demonstrate multi-objective nonlinear metasurfaces with tailored frequency conversion emission.

Interest in optics, electromagnetism and nanotechnology.