(699f) Quantifying Confidence in DFT Predicted Surface Pourbaix Diagrams at Solid-Liquid Interfaces on Transition Metal Surfaces | AIChE

(699f) Quantifying Confidence in DFT Predicted Surface Pourbaix Diagrams at Solid-Liquid Interfaces on Transition Metal Surfaces


Vinogradova, O. - Presenter, Carnegie Mellon University
Viswanathan, V., Carnegie Mellon University
Krishnamurthy, D., Carnegie Mellon University
Pande, V., Carnegie Mellon University
Emerging electrochemical energy conversion and storage will play an increasingly crucial part in sustainable energy systems. Low temperature Proton-Exchange Membrane (PEM) fuel cells are attracting considerable attention as a means to convert chemical energy conversion. However, it is widely recognized that a major bottleneck for PEM fuel cells is related to sluggish kinetics of the Oxygen Reduction Reaction (ORR) at the cathode. Understanding how to improve ORR and the activity of catalyst materials requires a detailed understanding of atomistic surface dynamics.

Density Functional Theory (DFT) calculations have been widely used to identify active catalysts by constructing free energy diagrams incorporating the electrochemically stable surface structure [1]. We construct surface Pourbaix diagrams to capture the most stable state of the surface under reaction conditions of electrode potential and pH. A unique aspect of this work is the incorporation of error estimation techniques from within the Bayesian Error Estimation Functional with Van der Waals corrections (BEEF-vdW) exchange correlation function [2]. We developed a systematic approach to propagate the uncertainty associated with the energetics for the construction of surface Pourbaix diagrams [3]. Within this methodology, we assign a prediction confidence value (c-value) [4] to adsorption energies of intermediates governing the ORR. As a result, surface phase transitions are no longer sharp and now have a finite width originating from DFT uncertainty.

We illustrate the surface phase stability of varying quantities of adsorbed O* and OH* intermediates relevant to ORR. In particular, we discuss the phase transition boundary from OH* to O* on commonly studied transition metals Pt, Pd, Ir, Rh, and Ru while noting the adsorption characteristics for favorable ORR activity in terms of our prediction confidence. We compare against linear sweep cyclic voltammetry experiments and DFT-based results and observe the most predominant species are in good agreement with our predictions. The OH* descriptor has been shown to be the optimal descriptor for the four electron ORR because it has the lowest uncertainty associated with prediction [5]. Therefore, we report our comparison of transition metal surfaces in terms of their binding strength of OH* as Ru>Rh>Ir>Pd>Pt, where Ru binds OH* the strongest which is in agreement with prior DFT studies [6]. We believe this method presents the opportunity to robustly incorporate multiple surface phases in determining electrochemical reaction mechanisms and thereby accurately determining catalytic activity.


[1] Hansen, H. A., Rossmeisl, J., Nørskov, J. K. Phys. Chem. Chem. Phys. 10, 3722–3730 (2008).

[2] Wellendorff, J., Lundgaard, K. T., Møgelhøj, A., Petzold, V., Landis, D. D., Nørskov, J. K., Bligaard, T., Jacobsen, K. W. Phys. Rev. B 85, 235149 (2012).

[3] Vinogradova, O., Krishnamurthy, D., Pande, V., Viswanathan, V. arXiv:1710.08407 [cond-mat.mtrl-sci]

[4] Houchins, G., Viswanathan, V. Phys. Rev. 96, 134426 (2017).

[5] Krishnamurthy, D., Sumaria, V., Viswanathan, V. J. Phys. Chem. Lett. 9, 3, 588-595 (2018).

[6] Karlberg, G. S. Phys. Rev. B 74, 153414 (2006).