| Abstract: | Typical thin‐film photovoltaic (PV) cells incorporate a textured transparent conductive oxide to enhance light trapping and efficiently harvest solar energy. Rigorous coherent optical simulations of these devices and a complete characterization of these textured films are a challenging problem because of the several orders of magnitude difference between the wavelengths of interest and the spatial dimension of the sample that needs to be evaluated. In this paper, a practical approach for rigorous and predictive modeling of optical properties of thin‐film PV cells incorporating a vast variety of light‐trapping structures including semi‐coherent textured films and patterned coherent structures is presented. In contrast to the existing semi‐empirical device models, it is demonstrated that the presented methodology can accurately predict the scattering properties of textured fluorine‐doped tin oxide and aluminum‐doped zinc oxide conductive transparent films. It is further shown that the optical response of single‐junction and tandem‐junction PV devices incorporating such films can also be predicted with good accuracy as compared with the measured results. Next, a methodology to identify the sufficient statistical fingerprints of semi‐coherent textured films that are needed to unambiguously predict the light propagation in thin‐film cells is presented. This comprehensive approach then lends itself to identifying the optimal surface morphology needed for strong light trapping. This rigorous approach automatically includes the effects of important loss mechanisms such as the surface plasmon‐enhanced absorption in textured metal surfaces that are otherwise very difficult to account for semi‐coherent approaches based on scalar scattering theory. Copyright © 2011 John Wiley & Sons, Ltd. |
