Morphology and plasmonic properties of self-organized arrays of ...

Chapter 5Modelling and analysis of theoptical propertiesIn this chapter we will focus on the development of a theoretical modelling framework,allowing us to account for the optical response of the 2D arrays of gold nanoparticlesdescribed in the previous chapters. A theoretical support to the experimental data is offundamental importance in order to achieve a comprehensive understanding of the originsof the optical properties of these systems, and thus to engineer the optical response byselecting a priori the proper morphological characteristics.In the next sections, we will first characterize the optical properties of the LiF templates,and then we will cover in detail the arrays of gold NPs, overlooking instead on thenanowires.5.1 Lithium fluoride nanostructuresWe start by extracting the optical constant of the nanopatterned LiF substrates. Adetailed characterization of the substrates characteristics is in fact extremely importantfor an proper modelling of the Au NPs arrays. In §4.1 we reported the ellipsometricspectra measured on the LiF(110) substrates prior to and after the further depositionof LiF, so here we proceed, as customary in ellipsometry, by constructing a model todescribe the samples and then fit the optical constants against the experimental data.All the calculation have been performed using the WVASE software supplied with theellipsometer.5.1.1 LiF substratesWe first consider the LiF(110) substrates in the as-received state. In the previous chapterwe saw that the samples exhibit negligible in-plane birefringence, despite the anisotropicatomic environments along the [001] and [1¯10] crystal directions. We can therefore applyan isotropic dielectric model for the bare substrate. Lithium fluoride is an insulator witha very wide band gap, and in the energy range between ∼1 eV and ∼5 eV it behaves likea transparent, non absorbing material; then, we can approximate its optical constants,in the visible and near-IR range, using a Cauchy parametrization of the refractive index(see equation (1.20)). We can further improve the model by accounting for the surface75