Simple models for complex devices

The search for higher efficiencies in photovoltaics has led us to develop increasingly complex solar cell architectures that rely on increasingly complex physical processes. The desire to overcome the Shockley-Queisser efficiency limit has caused us to consider stacks of junctions with cascading bandgaps, quantum dots and impurities for the creation of intermediate bands, amongst others. The need to improve in-coupling and absorption in thinner layers requires the development of surface textures and nanoparticles working either in the geometric- or wave-optical regimes, or a combination of both. Quantum wells have been employed to tune bandgaps or to induce angle-selective luminescent emission. More recently, photonic methods have been used to control the black-body emission of solar cells and thus influence their operating temperature. Understanding these processes, requires a detailed knowledge of the paths incident photons take in the device, complex generation and recombination mechanisms (sometimes involving multiple sequential transitions), carrier drift and diffusion, and the paths of emitted photons (either luminescent or incandescent). This sounds like a job for a supercomputer. However, with a few carefully made assumptions, accurate, predictive and insightful models can be developed that allow calculations to be performed on a laptop computer. We will show how these models can help to develop nano-photonic space solar cells that are more radiation hard; to better understand quantum-dot intermediate-band solar cells and improve them using light trapping; and to understand and control black-body emission in silicon solar cells.