In principle there’s more than enough sunlight falling on the earth to meet all our energy needs in a sustainable way, but the prospects for large scale solar energy are dimmed by a dilemma. We have very efficient solar cells made from conventional semiconductors, but they are too expensive and difficult to manufacture in very large areas to make a big dent in our energy needs. On the other hand, there are prospects for unconventional solar cells – Graetzel cells or polymer photovoltaics – which can perhaps be made cheaply in large areas, but whose efficiencies and lifetimes are too low. In an article in this month’s Nature Materials (abstract, subscription required for full article, see also this press release), Imperial College’s Keith Barnham suggests a way out of the dilemma.
The efficiencies of the best solar cells available today exceed 30%, and there is every reason to suppose that this figure can be substantially increased with more research. These solar cells are based, not on crystalline silicon, like standard solar cell modules, but on carefully nanostructured compound semiconductors like gallium arsenide (III-V semiconductors, in the jargon). By building up complex layered structures it is possible efficiently to harvest the energy of light of all wavelengths. The problem is that these solar cells are expensive to make, relying on sophisticated techniques for building up different semiconductor layers, like molecular beam epitaxy, and currently are generally only used for applications where cost doesn’t matter, such as on satellites. Barnham argues that the cost disadvantage can be overcome by combining these efficient solar cells with low-cost systems for concentrating sunlight – in his words “our answer to this particular problem is ‘Smart Windows’, which use small, transparent plastic lenses that track the sun and act as effective blinds for the direct sunlight, when combined with innovative light collectors and small 3rd-generation cells,” and he adds “Even in London a system like this would enable a typical office behind a south-facing wall to be electrically self-sufficient.”
Even with conventional technologies, Barnham calculates that if all roofs and south-facing walls were covered in solar cells this would represent three times the total generating capacity of the UK’s current nuclear program – that is, 36 GW. This represents a really substantial dent in the energy needs of the UK, and if we believe Barnham’s calculation that his system would deliver about three times as much energy as conventional solar cells, this represents pretty much a complete solution to our energy problems. What is absent from the article, though, is an estimate of the total production capacity that’s likely to be achievable, merely observing that the UK semiconductor industry has substantial spare capacity after the telecoms downturn. This is the missing calculation that needs to be done before we can accept Barnham’s optimism.