Plasmonic nanostructures hold promise for the realization of ultra-thin sub-wavelength devices, reducing power operating thresholds and enabling nonlinear optical functionality in metasurfaces. However, this promise is substantially undercut by absorption introduced by resistive losses, causing the metasurface community to turn away from plasmonics in favour of alternative material platforms (e.g., dielectrics) that provide weaker field enhancement, but more tolerable losses. Here, we report a plasmonic metasurface with a quality-factor (Q-factor) of 2340 in the telecommunication C band by exploiting surface lattice resonances (SLRs), exceeding the record by an order of magnitude. Additionally, we show that SLRs retain many of the same benefits as localized plasmonic resonances, such as field enhancement and strong confinement of light along the metal surface. Our results demonstrate that SLRs provide an exciting and unexplored method to tailor incident light fields, and could pave the way to flexible wavelength-scale devices for any optical resonating application.

Plasmonic metasurfaces are promising as enablers of nanoscale nonlinear optics and flat nonlinear optical components. Nonlinear optical responses of such metasurfaces are determined by the nonlinear optical properties of individual plasmonic meta-atoms. Unfortunately, no simple methods exist to determine the nonlinear optical properties (hyperpolarizabilities) of the meta-atoms hindering the design of nonlinear metasurfaces. Here, we develop the equivalent RLC circuit (resistor, inductor, capacitor) model of such metaatoms to estimate their second-order nonlinear optical properties, that is, the first-order hyperpolarizability in the optical spectral range. In parallel, we extract from second-harmonic generation experiments the first-order hyperpolarizabilities of individual meta-atoms consisting of asymmetrically shaped (elongated) plasmonic nanoprisms, verified with detailed calculations using both nonlinear hydrodynamic-FDTD and nonlinear scattering theory. All three approaches, analytical, experimental, and computational, yield results that agree very well. Our empirical RLC model can thus be used as a simple tool to enable an efficient design of nonlinear plasmonic metasurfaces.

No abstract