Metal nanoparticles smaller than the wavelength of light can support optically driven oscillations of the conduction electrons in the metal - localised surface plasmons - that strongly interact with incident light. The resonances of these oscillations are highly tuneable by changing the geometry or environment of the nanoparticle, and can be strongly scattering. If the particle is placed in the vicinity of a high refractive index material - for example on top of the absorbing layer of an optoelectronic device - the amount of light scattered and the angular distribution will change. This last point is crucial in plasmonic light-trapping: if the geometry of the nanoparticle and the distance from the surface is optimised, the nanoparticles will scatter a large fraction of the incident light into an underlying semiconductor film, reducing both reflection and transmission losses by randomising the incidence angle.
In this work, the authors use spherical nanoparticles on thin spacer layers: a geometry which has been shown by several groups to provide good light trapping when the nanoparticles are located on the front surface of the device. They simulate the structures using a 3D finite differences time domain code to identify an optimum density for the nanoparticle array: too sparse and the interaction with the incident light will be too weak; too dense and the particles interact with each other resulting in absorption losses. The challenge is then to find a way to fabricate these dense arrays. For plasmonic light trapping to be feasible for optoelectronic devices, the fabrication methods used must be cheap, large area and should not damage the electrical properties of the device. This puts stringent restrictions on fabrication. Ono and colleagues have developed a new bonding technique to attach commercially available Au colloidal nanoparticles to the surface of the photodetector with very high density. Colloidal metal nanoparticles have been used before for light-trapping applications, but it has been difficult to achieve sufficiently high nanoparticle density without aggregation. To overcome this they use a seeding layer that can be synthesised in solution consisting of small Au nanoparticles (4 nm diameter) in a polyimide resin. Exposure to O2 plasma exposes these seed nanoparticles and a thiol ligand is then used to anchor the large, scattering Au colloidal particles onto these Au seeds at high density and without aggregation.
This technique opens up the possibility of using dense films of colloidal particles for light-trapping for a variety of optoelectronic devices. As there is now a veritable zoo of different colloidal metal nanoparticles that can be fabricated, there is also the possibility that it could be useful for other plasmonic applications like SERS and other sensing techniques.
You must log in to add comments.