Inspired by the way in which sunlight reflects off the surfaces of skyscrapers on New York’s Manhattan Island, a group of researchers at the Georgia Tech Research Institute is looking to improve on conventional silicon solar cells and imaging chips by expanding the light-collecting surface from two to three dimensions.
Arrays of photovoltaic cells for solar power generation, and the charge-coupled devices at the heart of digital cameras, are planar (2-D) structures that generate an electric current when light impacts on a cell or pixel in the structure.
But there is a fundamental limitation to the 2-D approach. Light sources often move (such as the Sun in the sky) or are non-localised (e.g., wide-angle camera fields-of-view). 2-D cells reflect a significant proportion of the light, reducing the number of photons that can be absorbed and turned into useable electric current.
With solar panels built into the surfaces of buildings, this is a problem that cannot easily be solved by moving the position of the panels. And on orbiting satellites, there is a limit to how much solar panels can be steered while maintaining the spacecraft attitude, or orientation relative to the Sun and the Earth.
Jud Ready and his colleagues are attempting to overcome these problems by building 3-D cells that resemble blocks of high-rise buildings in a city centre. The tower structures, which are around 100 microns tall, and 40×40 microns square, are built from arrays of vertically-aligned, multiwalled carbon nanotubes.
Photoactive materials such as cadmium telluride and cadmium sulphide are deposited on the towers, and the whole structure coated with a transparent film of indium tin oxide, which serves as the top electrode. The vertical surfaces of the towers trap and multiply-reflect light rays incident from a wide range of angles, allowing the structure to absorb more energy.
Problems common to all photo-cells are voltage limitation due to photo-generated current carriers recombining with the holes they leave in the atomic matrix of the coating, and large leakage currents. If these issues are relevant to 2-D cells, they will be even more so with the greatly increased surface area of 3-D structures. ‘Excess dark current is something that we will need to balance against various trade-offs as we move forward,’ says Ready. ‘We have just done a proof-of-concept demonstration at this point. Now we can move towards optimising and improving.’
The chemical vapour deposition process used to produce the cells is not particularly amenable to industrial scale-up, but Ready and his colleagues are working on a chemical bath deposition that should reduce production costs considerably.
Ready stresses that low cost is not the selling point of his 3-D cells. ‘Our selling point is smaller volume and lighter weight for the same power produced. This allows you to use those weight savings for either more payload (for space-based things like satellites), or alternatively you could launch on a smaller/cheaper rocket. For the terrestrial market, your house would not need to look like a science-fair project with gigantic arrays sticking out everywhere.’
Real-world industrial economics may not apply to space applications, but watt/cost is crucial on the ground. Especially when compared with the latest developments in low-cost, organic thin-film solar cells.
Ready’s goal is to have a commercial product ready in 5–10 years, and he sees the most likely initial applications being aerospace and military special operations.
Interesting to note is that the research is part-funded by the US Air Force Office of Scientific Research. But the military interest is primarily in the potential of 3-D cells for hyperspectral imaging and other photonics applications, not solar power generation. Hyperspectral imaging is by no means limited to military applications. It can be used to identify plant matter and minerals, and is of general use in remote sensing, both from space and on the ground.
Further reading: Carbon Nanotube Arrays for Photovoltaic Applications, Camacho et al., JOM – Minerals, Metals and Materials Society, March 2007, pp 39–42: broken web archive, so no working link (and not yet indexed by Google Scholar).
Article first published in Nanomaterials News.