Membrane Selection and Resistance Characterisation in Hydrogen-Iron Redox Flow Batteries

Abstract

Hydrogen–iron (HyFe) redox flow batteries are a promising option for long-duration energy storage, but their performance depends strongly on the ion-exchange membrane, which must combine high ionic conductivity with resistance to iron-associated degradation. This thesis develops an integrated experimental and modelling framework to screen membrane candidates, compare their behaviour under HyFe-relevant conditions, quantify retained iron, and evaluate whether an ex-situ-derived resistance-growth coefficient can be transferred to in-situ operation. A literature-informed screening procedure, based on a conductivity-to-permeability figure of merit combined with practical feasibility criteria, selected Nafion NR212, Selemion CMVN, and a confidential hydrocarbon cation-exchange membrane (Membrane X) for experimental evaluation. In their respective documented stack configurations, Membrane X and NR212 sustained approximately 477 and 965 hours of cycling, with apparent beginning-of-charge cell resistance increases of approximately 8% and 29%; the tested CMVN configuration did not sustain repeatable multi-cycle operation. Because the stack configurations and operating histories were not fully matched, these results are interpreted as candidate-level evidence. Controlled immersion experiments showed that water-wash-resistant, sodium-chloride-exchangeable retained iron was membrane- and exposure-specific, with no universal preference for either the Fe(II)-rich or Fe(III)-rich condition on a charge-equivalent basis. Post-cycling SEM-EDS provided qualitative evidence of iron at selected surface locations on NR212 and Membrane X. A symmetric H-cell experiment yielded an effective concentration-normalised resistance-growth coefficient of 1.162e-9 cubic metres per mole per second, which, when embedded in a data-assisted hybrid 0D-1D Nernst-Planck transport and Langmuir fouling model, captured the principal shape of the normalised Membrane X stack-resistance trajectory (R2 = 0.931, NRMSE = 7.13%). Membrane X is identified as the preferred candidate for further matched and replicated HyFe evaluation.