par Godefroid, Blaise ;Kozyreff, Gregory
Référence International Conference on Hybrid and Organic Photovoltaics (HOPV18) (2018 May 28th - 31st: Benidorm, Spain)
Publication Non publié, 2018-05-30
Communication à un colloque
Résumé : We address the issue of short exciton diffusion length by focussing on radiative, or fluorescence, losses. These losses sensitively depend on the cell design [1] because of the strong confinement that is characteristic of organic solar cells. Therefore, we discuss the cell performance as a function of the cell geometrical parameters and of the internal luminescence quantum efficiency (ILQE), which is the fraction of the exciton decay that is intrinsically due to fluorescence.In order to determine the optimal design, we have established a generalisation of the Shockley-Queisser theory for organic solar cells [2,3] taking into account the diffusive exciton transport and microcavity effects both for exciton radiation and sunlight injection. Hence, given the ILQE, we are able to determine the buffer layers thicknesses in order to increase the exciton lifetime and, hence, the exciton diffusion length.Managing fluorescence losses may significantly improve the cell performance. This is what we demonstrate using realistic material parameters inspired from literature data for which we obtain an increase of power conversion efficiency from 11.3% to 12.7% as the ILQE goes from zero to one. Conversely, to ignore the dependence of the exciton radiative decay on the environment may lead to suboptimal design and to unnecessary low cell efficiency. We illustrated this latter situation with experimental material data. Our results invite us to rediscuss the statement "good solar cells should also be good emitters".Finally, we observe that there is a qualitative change as soon as the bulk radiative decay rate is no zero (ILQE>0). This is due to the quenching effect experienced by radiative excitons in the vicinity of a dissipative medium. This shows that not to consider exciton radiative decay (ILQE=0) is an inaccurate modelling assumption.References:[1] G. Kozyreff et al. Opt. Express, 21:A336-A354, 2013.[2] B. Godefroid and G. Kozyreff. Phys. Rev. Applied, 8:034024, 2017,[3] Uwe Rau et al. Phys. Rev. Applied, 7:044016, 2017.