Third generation photovoltaic cells are solar cells that are potentially able to overcome the Shockley–Queisser limit of 31-41% power efficiency for singlebandgap solar cells. This includes a range of alternatives to the so-called “first generation solar cells” (which are solar cells made of semiconducting p-n junctions) and “second generation solar cells” (based on reducing the cost of first generation cells by employing thin film technologies). Common third-generation systems include multi-layer (“tandem”) cells made of amorphous silicon or gallium arsenide, while more theoretical developments include frequency conversion, hot-carrier effects and other multiple-carrier ejection.
Solar cells can be thought of as visible light counterparts to radio receivers. A receiver consists of three basic parts; an antenna that converts the radio waves (light) into wave-like motions of electrons in the antenna material, an electronic valve that traps the electrons as they pop off the end of the antenna, and a tuner that amplifies electrons of a selected frequency. It is possible to build a solar cell identical to a radio, a system known as an optical rectenna, but to date these have not been practical.
Instead, the vast majority of the solar electric market is made up of silicon-based devices. In silicon cells, the silicon acts as both the antenna (or electron donor, technically) as well as the electronic valve. Silicon is almost ideal as a solar cell material; it is widely available, relatively inexpensive, and has a bandgap that is ideal for solar collection. On the downside it is energetically expensive to produce silicon in bulk, and great efforts have been made to reduce or eliminate the silicon in a cell. Moreover it is mechanically fragile, which typically requires a sheet of strong glass to be used as mechanical support and protection from the elements. The glass alone is a significant portion of the cost of a typical solar module.
According to the Shockley–Queisser limit, the majority of a cell’s theoretical efficiency is due to the difference in energy between the bandgap and solar photon. Any photon with more energy than the bandgap can cause photoexcitation, but in this case any energy above and beyond the bandgap energy is lost. Consider the solar spectrum; only a small portion of the light reaching the ground is blue, but those photons have three times the energy of red light. Silicon’s bandgap is 1.1 eV, about that of red light, so in this case the extra energy contained in blue light is lost in a silicon cell. If the bandgap is tuned higher, say to blue, that energy is now captured, but only at the cost of rejecting all the lower energy photons.
It is possible to greatly improve on a single-junction cell by stacking extremely thin cells with different bandgaps on top of each other – the “tandem cell” or “multi-junction” approach. Traditional silicon preparation methods do not lend themselves to this approach. There has been some progress using thin-films of amorphous silicon, notably Uni-Solar’s products, but other issues have prevented these from matching the performance of traditional cells. Most tandem-cell structures are based on higher performance semiconductors, notably gallium arsenide (GaAs). Three-layer GaAs cells hold the production record of 41.6% for experimental examples
Numerical analysis shows that the “perfect” single-layer solar cell should have a bandgap of 1.13 eV, almost exactly that of silicon. Such a cell can have a maximum theoretical power conversion efficiency of 33.7% – the solar power below red (in the infrared) is lost, and the extra energy of the higher colors is also lost. For a two layer cell, one layer should be tuned to 1.64 eV and the other at 0.94 eV, with a theoretical performance of 44%. A three-layer cell should be tuned to 1.83, 1.16 and 0.71 eV, with an efficiency of 48%. A theoretical “infinity-layer” cell would have a theoretical efficiency of 64%