Non-equilibrium Electrochemical Phase Diagrams with Automatic Differentiation

Reaction rate models are a vital tool to understand behavior of electrochemical systems. The Butler-Volmer model is a standard workhorse in the field, but especially at more extreme conditions (large currents, high overpotentials), it fails to capture realistic behavior. We have previously argued for the need to move towards more physical models such as Marcus-Hush-Chidsey, and also introduced an MHC variant that explicitly incorporates the electronic density of states of the solid side of a solid-electrolyte interface. In this talk, I will showcase a software package I developed that implements a variety of standard electrochemical rate models and allows the construction of multidimensional phase diagrams along axes of composition, current, temperature, or any other model parameter. This functionality relies critically on the automatic differentiation capabilities of the Julia Language to be able to smoothly and efficiently handle mathematically nontrivial model formulations in terms of integrals, inverse functions, etc. We will showcase the capabilities of this package to easily compare model types, fit to experimental data, and construct these phase diagrams.