In the present Letter, we demonstrate how the design of metallic nanoparticle arrays with large electric field enhancement can be performed using the basic paradigm of engineering, namely the optimization of a well-defined objective function. Such optimization is carried out by coupling a genetic algorithm with the analytical multiparticle Mie theory. General design criteria for best enhancement of electric fields are obtained, unveiling the fundamental interplay between the near-field plasmonic and radiative photonic coupling. Our optimization approach is experimentally validated by surface-enhanced Raman scattering measurements, which demonstrate how genetically optimized arrays, fabricated using electron beam lithography, lead to order of ten improvement of Raman enhancement over nanoparticle dimer antennas, and order of one hundred improvement over optimal nanoparticle gratings. A rigorous design of nanoparticle arrays with optimal field enhancement is essential to the engineering of numerous nanoscale optical devices such as plasmon-enhanced biosensors, photodetectors, light sources and more efficient nonlinear optical elements for on chip integration.
© 2012 American Chemical Society