Redox flow batteries present numerous important advantages such as independent scaling of energy and power; high efficiency of energy transformation; cheap (carbon) electrodes without special catalysts, etc. However, energy density of reactants in reservoirs for most of actually employed systems of this type is relatively low. A new approach [1] is based on the use of aqueous solutions of multi-electron oxidants, first of all halate anions, XO₃^-, which would ensure a large redox capacitance owing to their reduction to halides, X^-, e.g. 790 Ah/kg or 1400 Ah/dm³ for LiBrO₃ and 1580 Ah/kg or 3100 Ah/dm³ for LiClO₃ at 25^oC. In combination with a suitable anodic process, e.g. H₂ oxidation, these systems possess high energy densities: 750 Wh/kg for LiBrO₃ and 1130 Wh/kg or 1150 Wh/dm³ for LiClO₃ [2].

Principal problem of these systems originates from the absence of electrochemical activity of halate ions within a suitable potential range. It may be overcome via use of a proper redox mediator cycle. This approach is especially efficient for autocatalytic cycles, e.g. based on combination of electrochemical and chemical steps:

X₂ + 2e- = 2 X^-; XO₃^- + 5 X^- + 6 H⁺ → 3 X₂ + 3 H₂O (*)

where each cycle leads to progressive accumulation of the catalytic species, X₂ and X^- [3]. Under appropriate conditions these transformations may generate very intensive steady-state currents (on the level of about 1 A/cm²) comparable to the diffusion-limited one for XO₃^- species, even for extremely low bulk-solution concentrations of the catalytic component, X₂. These predictions have been confirmed experimentally for rotating disk (Fig. 1) and 3D flow-through porous electrodes [4,5].

For their practical applications in power-generating discharge devices it is of importance that such intensive currents were reached rapidly, even starting from the zero-current state and a very small bulk-solution X₂ concentration (about 1 mM or lower).

Evolution of the halate process, Eq (*), between these states has been analyzed [6], via numerical solution of the non-stationary transport equations for all components of the system, XO₃^-, X^- and X₂. Three different stages of the process have been revealed.

Within the initial stage the chemical step, Eq (*), is inefficient so that the presence of the principal oxidant does not affect the process which consists in electroreduction of X₂ species which diffuse from bulk solution to electrode surface. Thus, passing current is very weak due to a low X₂ concentration, being close to its diffusion-limited current.

After a characteristic time, t ≈ T, determined by the product of the rate constant of the chemical step and the bulk-solution concentration of XO₃^- species, autocatalytic cycle (*) starts to generate extra amounts of X₂ and X^- species in the vicinity of the electrode leading to exponential increase of the current (Fig. 1).

Then, the surface concentration of XO₃^- species decreases strongly, and the current drops after passing a maximum (Fig. 1).

For rapid chemical steps, e.g. for concentrated BrO₃^- solutions, the whole duration of the evolution is within the range of a few seconds.

Fig. 1. Dependence of the maximal current density j^max on the RDE frequency f: comparison of theoretical predictions and experimental data for two bromate-acid solutions

Acknowledgements: Supported financially by Russian Science Foundation (grant RSF20-63-46041).

References:

1. Y. V. Tolmachev, A. Pyatkivskiy, V. V. Ryzhov et al, J. Solid State Electrochem., 2015, 19, 2711
2. M. A. Vorotyntsev, А. Е. Antipov, D. V. Konev, Pure Applied Chemistry, 2017, 89, 1429
3. M. A. Vorotyntsev, A. E. Antipov, Electrochim. Acta, 2019, 323, 134799
4. A. D. Modestov, D. V. Konev, A. E. Antipov et al, Electrochim. Acta, 2018, 259, 655
5. A. D. Modestov, D. V. Konev, A. E. Antipov, M. A. Vorotyntsev, J. Solid State Electrochem., 2019, 23, 3075
6. M. A. Vorotyntsev, D. V. Konev, Electrochim. Acta, 2021, 391, 138914