Strong coupling between plasmon resonance and optical cavity resonance traps more energy in ultrathin solar cell materials
|Resonance||When oscillations are amplified at a certain frequency/wavelength for a system.|
|Plasmon resonance||Resonance for oscillations of the electron gas in a metal and more generally of free charge carriers.|
|Localized surface plasmon resonance||A type of plasmon resonance found in nanostructured metals and which interacts with light.|
|Optical impedance matching||Matching of the combined (effective) optical properties of a system, such that light pass without reflection.|
|Fabry-Perot resonance||A type of cavity resonance, characterized by a standing wave pattern between two planar, reflective surfaces and occurring when a roundtrip of light amounts to a whole number of wavelengths.|
|Resonance modes||Specific patterns of oscillation, corresponding to resonances at different frequencies (and thereby different wavelengths of light).|
|Hybridized modes||Resonance modes resulting from the mixing of more fundamental modes when these are coupled.|
|Strong coupling||Coupling between different type of resonances, where the coupling strength exceeds the damping (friction) of the system.|
|Charge carrier||Negatively charged electrons or their positively charged counterparts, holes. Free charge carriers conduct current.|
|Semiconductor||A material with low conductivity due to few free charge carriers, but where free charge carriers can be created through a small energy supply, for instance by light.|
|Bandgap||The smallest energy that must be supplied to an electron in order to excite it from the valence to the conduction band of a semiconductor, thereby generating free charge carriers (one electron and one hole) and increasing the conductivity.|
In plasmonic solar cells, localized surface plasmon resonances are used to trap light and create free charge carriers in a layer of semiconductor material. This is the first step in the conversion of sunlight to electricity. To accomplish this, one approach is to use metallic nanoparticles whose localized surface plasmon resonances give rise to a strong electromagnetic field in their vicinity. Within reach of this strong near-field, the plasmon energy may be efficiently transferred to electron-pairs in an ultrathin layer of semiconductor material. In parallel with research in this area, the use of optical cavity resonances has been investigated, where an ultrathin absorption layer (around 10 nm) is deposited on a transparent spacer layer separating the absorber from a highly reflective layer on the back. Through variation of the layer thicknesses, this system can be matched to the optical impedance of the incident medium, so that reflections from the front are eliminated and light is absorbed with maximum efficiency. Optical impedance matching is readily accomplished near a Fabry-Perot (FP) resonance of the semiopen nanocavity formed by the absorber/spacer/reflector system.
By combining plasmon active nanostructures with nanocavities, a high optical absorption over a broad spectral range can be produced by contributions from the two type of resonances and their different modes. However, it is challenging to simultaneously impedance match different type of resonances in one and the same geometry. What we have now shown is that strong coupling between basic resonance modes gives rise to new, hybridized resonance modes which make it possible to construct systems where optical impedance matching and high absorption is accomplished at two or more wavelengths simultaneously.
In the experiments we used gold nanoparticles that were regularly ordered in an array on a surface. The gold nanoparticles capture light effectively due to their plasmon resonances and were further coated by an ultrathin layer of the semiconductor tin monosulfide (SnS). The combined layer of SnS and gold nanoparticles worked together as an ultrathin, nanostructured absorption layer. This absorber layer was further located on top of a transparent spacer of silicon dioxide which in turn was supported by a reflective layer of aluminum. Since the absorber layer also caused internal reflections, a semi-open nanocavity was formed so that FP resonances could be supported in this structure.
SnS samples were produced by atomic layer deposition (ALD) on silicon and glass, respectively, for separate characterization of the material. The optical properties of the SnS turned out to vary with thickness and was also sensitive to various parameters that were not fully controlled in this study. A high capability for optical absorption was nevertheless clear across the solar spectrum. Combined with the gold nanostructure the SnS was seen to increase the absorption up to an optimal thickness of 8.6 nm. Compared to other forms of tin sulfide (SnSx, x≈2) previously studied, the SnS layer resulted in a much broader absorption, which is an important advantage for solar cell applications. Interestingly, this absorption was distributed over two separate peaks for some thicknesses of SnS, which could not be explained by different modes of plasmon resonances being involved since measurements by spectroscopic ellipsometry showed the absorption capability of the layer to be concentrated to a single plasmon peak in the gold/SnS absorber layer.
It turned out that the origin of these two peaks is due to strong coupling between a plasmon resonance of the absorber layer and an FP resonance of the gold-SnS /spacer/reflector structure. The coupling is unusually strong because of the high internal reflection of the gold-SnS layer, which produced a large overlap of the field distributions of the different resonances. Strong coupling causes new, mixed (hybridized) modes to form, which inherit properties from the original modes and behave as a mixture of these. By varying the spacer thickness, the degree of mixing could be controlled and the different resonance peaks were more or less dominated by the properties of one or the other basic resonance. However, for a certain spacer thickness, both hybridized modes are equally influenced by the basic modes, so that their properties are similar. By adapting one of the hybrid modes to a desirable condition, such as optical impedance matching, the other will thereby automatically also fulfill a similar condition. This explains how the two peaks with near 100 % absorption is possible in the studied system. This concept of resonance mode equalization through hybridization could moreover be generalized to include additional resonances, so that properties of the system could be matched to a desirable condition at more than two wavelengths.
Using the effects of strong coupling between plasmon- och FP-resonances may thus enable high optical absorption around two or more wavelengths. Due to the broad resonance peaks observed here, a substantial part of the solar spectrum may be covered and contribute to the efficient conversion to electricity in a solar cell. In this study, tin monosulfide has been identified as a particularly promising material for ultrathin solar cells based on this concept, thanks to its high capability for light absorption and also its high refractive index (which is what causes the high internal reflectivity of the absorber layer). More than 60 % of the solar photons with energies exceeding the SnS bandgap were shown to be available for the generation of electricity, which translates to a maximum theoretical efficiency of around 19 % for the barely 10 nm thick semiconductor coating. These results may therefore be important steps towards cheap and efficient solar cells based on nanotechnology.
The original article is freely available for download here: Strong coupling of plasmon and nanocavity modes for dual-band, near-perfect absorbers and ultrathin photovoltaics