Mass transport limitations in concentrated aqueous electrolyte solutions: Theoretical and experimental study of the hydrogen–bromine flow battery electrolyte

Abstract

Modelling and simulation is a powerful tool to support the development of novel flow cells such as electrolysers and flow batteries. Electrolytes employed in such cells often consist of aqueous solutions of highly concentrated solutes at elevated temperatures. Such conditions pose numerous challenges in conventional model parametrisation because of non-ideal behaviour of the electrolytes. The aim of this work is to study mass transport of electroactive species in highly-concentrated media.We selected the hydrogen–bromine flow battery posolyte, HBr (aq) and Br2, as an exemplary flow battery electrolyte and we leveraged chronoamperometric techniques involving ultramicroelectrodes to study diffusion and migration of bromide and bromine at high concentration and temperature. We successfully simulated the current densities of HBr/Br2 redox reactions in solutions up to 8 mol L–1 using advanced mass transport theory which agreed well with the results obtained with ultramicroelectrodes.While uncharged species transport (Br2) can be credibly modelled using conventional theories such as Fick’s law, charged species (Br–) require special treatment as the diffusion coefficient vary with concentration up to 50 % with respect to the limiting value at infinite dilution. The transport of charged species without added supporting electrolyte occurs via both migration and diffusion and the contribution of migration current may be up to 50 % of the total current. At HBr concentration > 0.6 mol L–1 migration appears to be suppressed due to the “self-screening” effect of the electrolyte.Proper experimental electrolyte characterisation under operating conditions similar to the actual flow cell applications is indispensable to establish predictive models and digital twins of electrochemical devices. Straightforward transfer of concepts known in electro-analytical chemistry to flow cells modelling may lead to erroneous simulations or model overfitting.