Bulk and surface plasmons in highly nanoporous gold films

Abstract

The far-field plasmonic behaviour of nanoporous gold films with void densities, f, ranging from 60% to 90% has been investigated and modelled. These layers have good dc conductivity and quite different nanostructure to traditional porous layers in which the metal percolates. Our gold films with f above 70% have high thermal emittance for a conductor at their thicknesses, and their flat spectral response at visible and near infrared wavelengths is not metal-like. We derive effective optical constants which become plasmonic at wavelengths between 1.8 m and 4 m for f from 72% to 87%. This onset is at much longer wavelengths than that in bulk gold. For f below 70% the onset of plasmonic behaviour is much closer to the dense material. A simple test is implemented to test for surface plasmon polaritons (SPP) under illumination. The more porous films show no evidence of SPP, while the less porous films display weak evidence. Thus by tailoring f in these nanostructures we can tailor the onset of effective plasmonic response across a wide range and emittance from around 0.9 down to low values. An effective uniform metal response is thus found in the presence of surface nanostructure without the interface absorption found in dense gold layers with structured surfaces.

Percolation in nanoporous gold and the principle of universality for two-dimensional to hyperdimensional networks

“…It follows primarily because nanoporous thin films when percolating behave as if they are homogeneous with an effective plasma frequency  P * which is much less than the  P value in dense gold. The quantitative link between  P * and  P has been derived previously [8,9] both in the general Bergman-Milton (B-M) formalism and in terms of the quasistatic effective medium approach due to Bruggeman [33]. The result in the Bruggeman approximation is that…”

“…The fabrication process of the three dimensional nanoporous gold films is similar to that described previously [8][9][10]. Briefly, alloy films of AuAl 2…”

“…Ideally to explore critical percolation parameters, we would like to make a set of samples with the same f c value but different f . However every change in f with the same vacuum deposition conditions leads in practice to a different morphology, with our etching process involving much gold atom movement [8][9][10]. A universal curve that follows may mean that in random self-supporting systems in 3-dimensions, self-assembled porous structures at low f always do this, but further work is needed to rule out such growth along the same critical curve.…”

“…Direct measurement of  P * gives an independent and useful measure of f using the B-M formalism [9]. The fall is due to a combination of the drop in average free electron density, and a rise in carrier effective mass m* in the complex network as f falls towards f c [8,9].…”

“…The properties of this material near percolation are assuming technical importance as a result of the growing interest in composite nanomaterials. Emerging applications of nanoporous gold include surface-enhanced Raman scattering spectroscopy [6,7], solar energy, thermal radiation and photonics [8,9,10], capacitive biosensors [11], supercapacitors [12] and chemical Percolation theory has an extensive history [15], which we will draw on for analysis of our data along with basic principles in phase transition theory [16]. The application of percolation theory to experimental materials has been predominantly for dc and ac electrical data on composites made from powders [17][18][19], while natural structures such as porous rocks [20] and sea ice [21] have also been studied.…”