Introductory thoughts

Nanotechnology has taken an ever broader space in advancing technologies. Many applications and devices are getting smaller, tough maintaining the same functionality or having this functionality in a cleaner, more precise way. From lasers to computer chips, from fabrication of lasers to lasers fabricating other devices, the usual demand is more precision and more power. Both can usually be improved by reducing spatial dimensions. This is also true in the field of photonics, where light is crafted with selected properties in small volumes and down to single photons, depending on the demands. One such application is designing and creating photonic devices able to manipulate small amounts of light in nano wires and similar structures cut into gold. Gold is a very stable material with not too big absorptions in the near infra-red which is an acceptable spectral position. In our group the EP 5 Uni Würzburg we chemically grow gold platelets which are then taken to a focused ion beam milling machine. With a precision down to a few nanometers the desired structures are inscribed into the gold material and can be used to guide or manipulate incident photons due to interaction with free electrons, the electro-magnetic field and boundaries of the material. Such structures confine the light to a region smaller than the wavelength of photons, where we need to describe the not only the electromagnetic part of the signal, but also have to take into account for the free electrons in the metallic material. This interaction between the outer electromagnetic wave and the movement of free electrons in the substrate are not longer independent of each other and can be seen as a single quasi-particle, namely the plasmon.

To be able to create those structures a special chemical synthesis method of gold was developed empirically. This generates, contrary to its normal crystalline structure, very flat but wide single crystal gold platelets. We refer to them as gold-flakes or just flakes. The single crystalline structure is achieved by chemical growth instead of other methods (e.g. molecular beam epitaxy) is crucial for the preservation of the photons and plasmons guided and manipulated in those structures. The less loss introduced by the material, the better is the quality and the more predictable the outcome of our experiments. Thus a polycrystalline material introducing lots of scattering is not suitable for our needs. The transport of these plasmons happens on the surface of a waveguide. Since usually a similar lateral and vertical dimension is desired, flakes should be in the range of to . Their quality in this regard is either defined by their aspect ratio or lateral size. The higher this aspect ratio is, the better. We want to be able to generate extremely flat but wide flakes to have a lot of space and still be able to create wires that are in the region of a few nanometers in the radial dimension. The process which creates those slabs have been refined by trial and error. They achieve a size of hundreds of micrometers while being only tens of nanometers thick. This translates into an aspect ratio of up to . Although this is already a enormous improvement compared to a bulk crystal, we want to push this aspect ratio even further and understand the underlying mechanisms of the growth. The recipe is documented and some parts do have explanations. One big aspect are probably the twin planes which are defects in the usual fcc-structure of the crystal. It is believed that those defects are responsible for changing the binding at the edge of the flakes and maintaining a high adsorption probability at the edges allowing the flakes to grow wide but not high, even contrary to the principle of surface minimization.