Kimmerle, AchimAchimKimmerleGreulich, Johannes M.Johannes M.GreulichHaug, HalvardHalvardHaugWolf, AndreasAndreasWolf2022-03-052022-03-052016https://publica.fraunhofer.de/handle/publica/24306810.1063/1.4939888In this work, the recently proposed band-gap narrowing model by Yan and Cuevas [J. Appl. Phys. 114, 044508 (2013)] is evaluated by simulations of the recombination pre-factor J(0) of highly phosphorus doped, passivated crystalline silicon surfaces, which are particularly relevant for solar cell applications. The results were fitted to experimental J(0) data measured on a large range of samples exhibiting different dopant profiles and passivation coatings, both for planar and textured surfaces. For each sample, the surface recombination velocity parameter S-p was extracted by fitting the simulation results to the experimental data. We show that the Yan and Cuevas' model developed for Fermi-Dirac statistics leads to a smooth and monotonically increasing curve for S-p as a function of the surface dopant concentration N-surf, for both investigated passivation layers. We provide a parameterization for this relation and compare the findings with those obtained with the widely used model by Schenk [J. Appl. Phys. 84, 3684 (1998)]. On the other hand, we show that the apparent band gap narrowing of Yan and Cuevas developed for use with Boltzmann statistics cannot be used to describe the experimental data, requiring unphysical negative S-p values for high N-surf.en530Application and comparison of band gap narrowing models for passivated phosphorus doped silicon surfacesjournal article