Superconductor Nanowires as Powerful Generators of THz Radiation

The so-called 'Terahertz gap' is rapidly closing. Originally perceived as a gap in the generation of electromagnetic radiation (it falls between two common methods of generating electromagnetic waves - infrared and microwave), in the last few years, research efforts have been geared towards exploiting this useful band between 0.5 THz and 2 THz. Terahertz devices are just starting to find use in spectroscopy, imaging and sensing particularly in the biological and physical sciences.

Superconductors are exceedingly sensitive detectors (known as bolometers) of THz radiation, owing to the cryogenic temperatures providing a low-noise environment. Cost however, is a severely limiting factor as low-temperature superconductors require expensive liquid helium cooling in order to attain the superconducting state. High-temperature superconductors (HTSC) are showing promise as bolometers as they require only cheap liquid nitrogen cooling and also exhibit fast quasi-particle relaxation times owing to their unique microstructure, yet it is this very microstructure that makes controlled crystal growth difficult.

The reason superconductors work so well in THz technology is that they can be used to create Josephson junctions (JJs) - two superconducting regions separated by an insulating layer. These junctions naturally convert dc voltages into electromagnetic radiation in the THz region. This means that a suitable array of JJs can be used both to detect impinging THz radiation, and also act as coherent emitters of THz radiation. It is this phenomenon that we propose to utilize here in this proposal, in order to create the most powerful and compact THz source to-date.

The current state of the art in HTSC bolometers attempt to exploit the fact that superconducting anisotropy in the crystal structure of these materials create intrinsic JJs. In the case of the superconductor 'bismuth-strontium-calcium-copper-oxide' (BSCCO), the unit cell can be thought of as consisting of layers of superconducting CuO2 planes separated by non-superconducting 'blocking' layers. This means that if the supercurrent could be made to flow along the c-axis of the unit cell, there would be JJs roughly every 15 Å, however herein lies the problem. Crystals of BSCCO can be grown naturally with any crystal axis expressed, but the supercurrent will always flow along the CuO2 planes in the ab axes, thereby avoiding crossing any intrinsic JJs. As HTSCs would be relatively cheap and powerful THz emitters if current could be made to flow across the JJs, efforts have been made to force the supercurrent along the c-axis by focussed-ion-beam (FIB) milling, but these devices containing BSCCO crystals have a very limited output, typically in the 20 - 50 micro-Watt region.

This proposal aims to improve significantly on this value, by using BSCCO nanowires grown via a biotemplated-inspired approach. Rather than have to use expensive and difficult techniques in order to produce nanostructured BSCCO, nanowires grown using this method can be produced spontaneously with the axis of elongation as the c-axis. This means that we can use them as the core of a compact, high-performance THz emitter. With lengths of up to 6 microns, these nanowires will therefore contain around 6,000 intrinsic JJs. As the radiative power increases as the square of the number of JJs, our nanowires should produce THz radiation in the thousands of mW range.