Abstract
Photovoltaics has great potential, not only because of the resource it uses, but also because the performance of current systems, despite the impressive progress made in recent decades, is still a long way from that predicted by theory. A first example is the gap between the yields of commercial devices (around 25%), those of the best laboratory devices (47%) and those predicted by theory (93.3%).
We can first try to understand what limits the yields of the various devices by revisiting the basics of photovoltaic conversion theory: starting with the thermodynamics of electromagnetic radiation, then considering more specific models such as those inspired by the theory of Shockley and Queisser. We can then study the various options that have been proposed to go beyond current devices: multijunction, but also alternative approaches such as hot carrier, intermediate band, photon conversion cells... all of which have the potential to approach the ultimate limits of efficiency.
Another area for performance improvement is the amount of active material required to achieve a given solar energy conversion capacity under standard operating conditions. Thanks to advances in photonics, it has been possible to considerably push back the theoretical and experimental limits on this point too.
Experimental results and proofs of concept are available on these two areas of improvement, which will be presented.
