Multi-scale modeling of TCO coating deposited by reactive magnetron sputtering: application to structured c-Si thin film solar cells

Transparent conducting oxides (TCO) present a large range of applications such as optoelectronic devices, especially transparent front-side contact for photovoltaic cells. In this last case, aluminium doped zinc oxide (ZnO:Al or AZO) can be a good alternative to indium doped tin oxide (ITO). However, the electrical and optical properties of such coatings highly depend on the structuration of the substrate. The present study focuses on nano-scale characterizations of AZO thin film deposited on micro-scale patterns.

The first step consists of nano-scale modelling of AZO deposition by reactive magnetron sputtering [1] on mono-crystalline silicone substrates thanks to a kinetic Monte Carlo model [2]. For these simulations, metallic (Zn, Al), reactive (O) and neutral (Ar) fluxes can be defined individually, with their own angular and energy distributions. Moreover, in order to mimic large samples, the periodic-supercell method [3] is used. Then, electrical and optical characterizations of the coating can be performed. Electrical properties (effective electrical conductivity) are computed by the mean of a finite-element code solving the Maxwell-Faraday equation (hypothesis: near absence of varying magnetic field). Optical properties (effective optical index) are evaluated by using effective medium models (Maxwell-Garnett and/or Bruggeman). During all the process, a special attention is given to the substrate shape.

The second step is based on a micro-scale modelling of a full multi-layered structured c-Si thin film solar cell. The optical characterization (optical efficiency and short circuit current density) is done by RCWA (rigorous coupled wave approximation) [4] allowing to predict complex optical phenomena like scattering or light trapping. The major novelty of such study is the introduction in the RCWA simulation of the effective refractive indices of AZO depending on the position on the substrate (flat or tilted section of the pattern). This optical model is introduced in a full optimization process done by genetic algorithm [5] in order to find the solar cell structural parameters providing the best short circuit current density.

References:
1. K. Ellmer et al., Surf. Coat. Technol., 93 (1) (1997), pp. 21-26
2. R. Tonneau et al., J. Phys. D. Appl. Phys. 51 (2018) 195202
3. A. Zunger et al., Phys. Rev. Lett. 65, 353 (1990)
4. J. Müller and al., Opt Express 2015 Jun 1;23(11):A657-70
5. A. Mayer et al., Proceedings of SPIE, vol. 10671, p.1067127 (2018)