A series of commercially available phenothiazine dyes was explored for their application in aqueous organic redox flow batteries. Of the dyes explored, Azure-A was found to have the most promising cycling performance and highest solubility, which was improved further with the use of nicotinamide as a ‘hydrotrope’ additive, which improved both solubility and cycle life. However, it was found that only ~50 % of theoretical capacity expected for Azure-A redox could be reached, regardless of cycling conditions. Through in situ and ex situ NMR, UV/vis and EPR spectroscopy as well as battery cycling using low-concentrations of the dye, this was ascribed to dimerisation of the redox-active species, which takes place at concentrations greater than 10 mM. Because of this, Azure-A is, in effect, only capable of a net 1e− redox process under practical conditions. By combining novel electrochemical impedance spectroscopy processing methods (including distribution of relaxation times and general phase element analysis) with symmetric cell cycling, a degradation mechanism involving polymerisation and passivation of the electrode is proposed as one source of the decrease in capacity with cycling.
Grey et al. (Sat,) studied this question.