Cheaper solar cell materials = more solar cells = decrease in fossil fuel dependence. Good. A recent paper from the new ACS environmental journal caught my attention on this very topic: “Materials Availability Expands the Opportunity for Large-Scale Photovoltaics Deployment”. It even got a write up in Chemistry World: “Analysis hints at solar energy alternatives”. Their findings are nice, if a little overly optimistic, but first some context...
Silicon, being an indirect band gap semiconductor, is at an obvious disadvantage for solar cells: you need thick films to have enough light absorption to generate significant photocurrents. The use of direct band gap semiconductors enables thin-film solar cells, costing less energy (and material) to produce. Since the 1970’s CdTe and Cu(In,Ga)Se2 have been two thin-film prototype absorbers, but they both suffer from their fair share of idiosyncrasies. For an ideal case, the (open-circuit) voltage you can get from a single-junction absorption material will be close to the value of the band gap. To approach this, you will need to have some very nice epitaxial thin films and a well behaved system. Generally the performance is much less than you’d expect, e.g. Cu(In,Ga)Se2 can be tuned to have an almost ideal band gap of around 1.3 eV, while the voltage you get out is on the order of 600 meV. This could be through a combination of (deep level) defects which can pin the Fermi energy where you don’t want it, or as in the case of most Cu based materials, a number of competing phases and defect complexes.
Back to the paper in question: “Twelve composite materials systems were found to have the capacity to meet or exceed the annual worldwide electricity consumption”. CuO, FeS2 and Zn3P2 are the ‘unconventional’ candidates highlighted in the abstract. Unfortunately, you have no hope of extracting voltages on the order of the band gaps, and being modest, you could easily decrease their expected efficiencies by a factor of two. Looking a bit deeper:
• CuO, with the highly correlated Cu 2+ d9 cation, is an indirect band gap (1.4 eV) Mott–Hubbard insulator (i.e. it’s highly resistive). Probably not the best bet for next generation solar cells.
• Zn3P2 appears more promising (if a little old). While it also has an indirect band gap (1.4 eV), it is intrinsically p-type and has reasonable carrier mobility.
• FeS2 combines a low band gap (1 eV) with some decent looking transport properties. However, looking below the surface, it has been tried and tested. 2.8 % efficiency won’t replace Si anytime soon. The study does raise one strong point: we really need to consider component cost and availability when designing new materials, especially for any energy related applications.