The modeling gap: What we are missing between molecular dynamics of electrode reactions and simulation of battery packs

Abstract

At the very beginning (and still thereafter) electrodes (not exactly the proper designation following W. Nernst) for electrolytic and galvanic processes (think of chloralkaline, aluminum production, copper refining) were flat and smooth ones, only some coarse surface structuring supporting gas bubble transport was applied sometimes. On the contrary electrodes in electrochemical conversion and storage devices were non-flat (with the notable exception of lithium, zinc and copper in primary batteries). Even today this contrast persists, only recently packed bed electrodes, i.e. porous bodies, have been suggested for some electroorganic processes [1]. The reasons are wellknown: Many of the electrode reactions in the latter devices proceed at fairly low rates causing possibly large charge transfer overpotentials. And because overpotentials (in this case more precisely charge transfer or activation overpotentials) are related by the Butler-Volmer equation to the charge transfer current density increasing the operating surface area is the most obvious way to smaller overpotentials. These porous electrodes provide further benefits beyond the large surface area: They enable the establishment of stable three-phase boundaries in gasdiffusion electrodes. A proper distinction between these basically two classes of electrodes has never been clearly established. At first glance the flat ones can be called 2D, the non flat 3D. A rough electrode with a low roughness factor (the ratio of the true area vs. the apparent or geometric surface area; numerous methods to determine electrochemically active surface areas are known [2]) may still appear flat and may thus be assigned to the first class but where