J Comput Electron (2011) 10:424–436 DOI 10.1007/s10825-011-0377-4 Simulation of dye solar cells: through and beyond one dimension Alessio Gagliardi · Matthias Auf der Maur · Desiree Gentilini · Aldo Di Carlo Published online: 8 November 2011 © Springer Science+Business Media LLC 2011 Abstract In this work we present a Computer Aided De- sign (CAD) software, called TiberCAD, to simulate Dye Sensitized Solar Cells (DSC). DSCs are particularly inter- esting devices due to their high efficiency (more than 11% on small area and 8% on large area) and long stability. Since their first development, much progress has been made in terms of efficiency, stability, lifespan and engineering of the device. However, the field of DSCs still lacks a com- plete model able to simulate the entire device over a general domain including all its components. In our model a drift- diffusion set of equations for the different charge carriers coupled to Poisson equation has been implemented within finite element method. The model takes into account also trap assisted transport for electrons in the mesoporous tita- nium dioxide with a phenomenological model derived from multi-trapping model. Three different applications of the code in 1, 2 and 3D are presented. The first 1D simulation is a study of corre- lation between physical parameters of the cell and energy conversion efficiency. A second application, 2D, discusses the effect on density and current distributions for different contacting of the cell and loss induced by the shadowing of metallic fingers. Finally, the third case, 3D, presents two different and innovative topologies for a DSC. A cell where contacts and illumination surface are completely decoupled and a DSC wrapped around an optical fiber. Keywords Dye sensitized solar cells · Finite element methods · Drift diffusion · Solar cells · Electrochemistry A. Gagliardi ( ) · M. Auf der Maur · D. Gentilini · A. Di Carlo CHOSE, Dep. of Electronic Engineering, University of Rome ‘Tor Vergata’, Via del Politecnico 1, 00133 Rome, Italy e-mail: Gagliardi@ing.uniroma2.it 1 Introduction The importance of renewable energy sources for mankind can be hardly underestimated. Despite the fraction of total energy produced that comes from renewable sources is still small compared to oil or nuclear energy supply, the amount is rapidly growing. Among renewable energy sources, the sun is one of the simplest and most obvious choices. The biggest problem concerning solar power is finding a way to convert the energy of light in electricity in an ef- ficient and cheap way. Differently from fossil fuels, where the main work of entrapping energy in a suitable chemical form has been already done by nature, in the case of photo- voltaics that process must be made artificially. The discovery of the photovoltaic effect in P-N semiconductor junctions immediately gained the attention of industry and scientific community as an easy way to exploit the sun. However, it turned out that things were not so simple as they seemed originally. Improving the efficiency of a semiconductor so- lar cell is a very challenging task. On the one hand, from the Schockley-Queisser limit [1] it is known that the optimal en- ergy bandgap for photovoltaic applications is 1.1 eV, close to the energy bandgap of silicon. On the other hand, sili- con is a very poor light absorbing material due to its indirect bandgap. The consequence is that we need thick (200 μm) silicon junctions to assure an efficient light harvesting. This results in the need of very pure and expensive crystalline material to avoid recombination due to traps and defects. Nevertheless, silicon solar cells have reached today a high standard efficiency close to 25% [2], which is not too far from the limit of 35% calculated by Shockley [1]. In the meantime people have been exploring other pos- sible solutions beyond conventional silicon, or more gener- ally, crystalline solar cells. The landscape of photovoltaics is nowadays extremely variegate and traditionally divided