A multiscale study of atomic interactions in the electrochemical double layer applied to electrocatalysis

Abstract

This work is an integrated study of chemical and electrostatic interactions in the electrochemical double layer, and their significance for accurate prediction of reaction kinetics in electrocatalysis. First, a kinetic model of the oxygen reduction reaction (ORR) on platinum, in connexion with first-principles techniques, is developed to illustrate that a self-consistent description of kinetics and reactant coverages on the surface can help to propose new mechanisms when energy prediction and experimental uncertainties still prevail. ORR kinetic limitation is often rationalized in terms of surface poisoning by parallel reactions, namely water oxidation, and/or as a result of the demanding requirements of Sabatier's principle. The sensitivity analysis presented here suggests that additional mechanisms may have to be considered, in particular self-poisoning by transient 02 dissociation in certain regimes. A common assumption of kinetic studies is that the only effect of electrode bias is to modify the electron chemical potential. To refine our understanding of bias effects in the double layer, a correction code applied to plane-wave DFT techniques is used to realistically simulate an electrochemical setup under potential control as an electrode with variable explicit charge screened by ions in solution. The scheme is first used to shed light on the nature of the stretching frequency shift of CO on Pt(1 11) as a function of electrode potential. It is concluded that the Stark effect interpretation is correct, and more generally, that electrochemistry on metal surfaces may often be correctly described in terms of perturbation theory. Then, hydrogen under-potential deposition on platinum is computed as a function of pH. It is shown that modification of the surface dipole by hydrogen electrosorption couples with the surface charge to make the adsorbant chemical potential pH-dependent. This observation is related to the concept of electrosorption valency. The octahedric-to-cubic nanoparticle shape transition resulting from hydrogen adsorption upon cathodic sweep is then predicted to be more pronounced in alkaline media. Inclusion of surface dipole effects is therefore relevant for surface stability and shape-dependent electroactivity. Third, the correction scheme is applied to develop a model of water dielectric saturation in the strong fields of the double layer. The water molecule dipole is computed in real space and Monte-Carlo simulation techniques are performed for the statistics of proton arrangement. DFT is seen to overestimate the permittivity of ice, confirming the difficulty of water simulation at the first-principles level. However, saturation effects are believed to be qualitatively captured and their influence on reaction kinetics in the double layer from the Frumkin effect is assessed. The impact is rather moderate with at most a factor of 3 in exchange current predictions. Finally, DFT occasional errors in chemisorption energies remain an important drawback for heterogeneous catalysis studies. Here, the vdW-DF functional for inclusion of long-range, dispersion interactions is tested on the prediction of CO adsorption on transition metals. Observed improvement on binding energies and adsorption site ordering comes at the expense of the correct description of metal energetics, suggesting the need for alternative schemes in this case. In conclusion, the purpose of this work is to help the design of electrocatalysts by providing a framework to assess chemical and electrostatic contributions to the kinetics, informed by the complexity and uncertainties attached to the surface and double layer structure.