Atomistic simulation of ionic líquids as novel fuel cell electrolytes

Abstract

The study behind this thesis has been carried on by a cooperation between the Electrochemical Process Engineering (IEK-14) and the Computational Material Modeling (IEK-13) departments at the Forschungszentrum Jülich. In this thesis, the computational analysis of the diethylmethylammonium trifluoromethanesulfonate [dema][TfO] proton conductive ionic liquid (PIL) is presented. This dissertation takes into account similar previous studies, both computational and experimental, such as the work of Palumbo et al. [Palumbo et al., 2020] and the work of Mori et al. [Mori et al., 2010]. The importance of this PIL relies on its application on proton exchange membrane fuel cells (PEMFCs) which are promisingly the new way of converting green energy for automotive and small applications. Different techniques have been used to study the properties of the PIL. The structure of this thesis aims to give a general knowledge on the reason why these studies are carried on, and on the techniques used for the analysis. In particular, the focus will be on the ATR spectroscopy which will be the technique used to obtain the experimental data, and on the computational approaches used. Those computational approaches, Molecular Dynamics (MD) and ab-initio simulations by means of Density Functional Theory (DFT), will be described in detail, focusing on the main parameters needed to obtain an optimal set-up of the simulations and the computational models. Finally, the results are represented and discussed, showing excellent agreement between simulation and experiment. The low wave-numbers mid-IR region of the [dema][TfO] spectra is reproduced accurately with a good match with the reality. The high wave-numbers mid-IR region of the spectra has some discrepancies, but many considerations have been done, together with some interesting results, which will be considered as the basis for further studies. The analysis presented in this thesis, based on combining all of these techniques, aims to a better understanding of the nature which could be obtained only computationally