Optimization of non-periodic plasmonic light-trapping layers for thin-film solar cells

• Published in: Nature communications Vol. 4; no. 1; p. 2095
• Main Authors: Pala, Ragip A., Liu, John S. Q., Barnard, Edward S., Askarov, Daulet, Garnett, Erik C., Fan, Shanhui, Brongersma, Mark L.
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 01.07.2013; Nature Publishing Group
• Subjects: 639/301/299/946; 639/766/400/1021; Arrays; Design; Humanities and Social Sciences; Light; multidisciplinary; Optimization; Photovoltaic cells; Science; Science (multidisciplinary); Simulation; United States > US; California
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/ncomms3095

Non-periodic arrangements of nanoscale light scatterers allow for the realization of extremely effective broadband light-trapping layers for solar cells. However, their optimization is challenging given the massive number of degrees of freedom. Brute-force, full-field electromagnetic simulations are computationally too time intensive to identify high-performance solutions in a vast design space. Here we illustrate how a semi-analytical model can be used to quickly identify promising non-periodic spatial arrangements of nanoscale scatterers. This model only requires basic knowledge of the scattering behaviour of a chosen nanostructure and the waveguiding properties of the semiconductor layer in a cell. Due to its simplicity, it provides new intuition into the ideal amount of disorder in high-performance light-trapping layers. Using simulations and experiments, we demonstrate that arrays of nanometallic stripes featuring a limited amount of disorder, for example, following a quasi-periodic or Fibonacci sequence, can substantially enhance solar absorption over perfectly periodic and random arrays. Non-periodic plasmonic nanostructures can provide efficient light trapping for solar cells, but their optimization can be computationally hard. Pala et al. present a semi-analytical model that identifies possible arrangements quickly, providing insight to the optimal level of disorder needed.