A Device Model for Intermediate Band Semiconductors

Abstract

Semiconductors with an additional intermediate band (IB) have the potential to greatly improve solar cell efficiency. Their theoretical efficiency limit is over 50% higher than that of standard semiconductor solar cells at full concentration. In practice however, their efficiencies are low compared to this detailed balance limit. Part of the reason is that it has not been possible to optimize IB device geometry because no device model has existed that could capture all the effects present in IB materials (e.g., charge transport inside the IB and self-consistent optics). In this thesis I introduce my new device model for intermediate band semiconductors called Simudo. The software uses the finite element method to solve the coupled Poisson/drift-diffusion (PDD) system of equations that describe the carrier dynamics inside semiconductor (IB or not) devices, along with optical propagation. I benchmark its accuracy on standard semiconductor problems against Synopsys Sentaurus, and I find that not only does it give valid results but in fact converges to the solution with a smaller number of mesh points by having quartic rather than merely quadratic solution convergence with respect to the number of mesh points. I also demonstrate Simudo's immediate usefulness by answering the question of whether IB mobility can compensate for mismatched optical absorption processes in different regions of the device. The device model work is preceded by three introductory chapters bringing the reader up to speed on semiconductor device physics and providing them with a primer on the finite element method. The coupled PDD equations are numerically challenging to solve, and the road to development of Simudo tried a number of formulations of the problem that were not successful. In the final chapter I discuss some of these formulations and why they did not succeed.