Comprehensively understanding the behavior of redox-active compounds in organic flow cells is essential to developing low-cost and long service life flow batteries. We develop and validate a zero-dimensional model of the electrochemical performance of an organic flow cell. The model simulates voltage losses from Faradaic charge transfer, Ohmic resistance, and mass transfer, along with the influence of spatial variations in the electrolyte’s state-of-charge between the cell and electrolyte reservoir, on the cell’s cycling behavior. The model’s predictions agree with constant current and constant voltage cycling data for a symmetric ferro-/ferricyanide cell across a wide range of current densities and electrolyte flow rates. We determine the model’s voltage loss parameters from electrochemical impedance spectroscopy and voltammetry measurements acquired prior to cycling, rather than fitted a posteriori. In operando measurements of the electrolyte’s state-of-charge demonstrate that the finite time for electrolyte flow between its reservoir and the electrochemical cell may significantly affect voltage-current behavior. By modelling active reactant decay, we demonstrate how capacity fade measured in a cell depends on the cycling protocol and reactant decay mechanism. This work shows that zero-dimensional electrochemical modeling helps in elucidating capacity fade mechanisms and optimizing the performance of chemistries under consideration for practical organic flow batteries.

No abstract

Aqueous redox flow batteries (RFBs) are promising candidates for low-cost, grid-scale energy storage. However, the polymer-based membranes that are used in most prototypical systems fail to prevent crossover of small-molecule reactants, which results in high rates of capacity fade. In this work, we explore the feasibility of a von Alpen sodium superionic conductor Na3.1Zr1.55Si2.3P0.7O11 (NaSICON) as an RFB membrane by examining its resistance, permeability, and interfacial morphology as a function of electrolyte composition and temperature. The resistance of NaSICON is stable for several weeks while immersed in neutral to strongly alkaline ([OH–] = 3 M) aqueous electrolytes, and its permeability to polysulfide-based and permanganate-based small-molecule RFB reactants is negligible compared to that of Nafion. The glassy phase of the NaSICON microstructure at the membrane–electrolyte interface is susceptible to some etching while in contact with aqueous electrolytes containing sodium ions. This etching becomes more extensive when potassium ions are present in the electrolyte, leading in certain instances to complete disintegration of the membrane. A ∼0.7 mm-thin NaSICON membrane can nevertheless support over three weeks of cycling of a ferrocyanide|permanganate flow cell in a strongly alkaline electrolyte ([OH–] = 3 M), with apparently negligible reactant crossover and very low capacity fade (<0.04%/day). NaSICON’s area-specific resistance also decreases dramatically with increasing temperature and decreasing membrane thickness; there is a 5.6× reduction from a 1.19 mm-thick membrane at 18 °C (101 Ωcm2) to a 0.61 mm-thick one at 70 °C (18 Ωcm2). Lowering the thickness of the membrane to 0.1 mm or lower will result in power densities at above ambient temperatures that are comparable to power densities of polymer membrane-containing flow cells. This work highlights the promise of ceramic membranes as effective separators in RFBs operating under neutral pH to strongly alkaline pH conditions.

We develop and experimentally validate a zero-dimensional model of the electrochemical performance of a flow cell. The model takes into account voltage losses due to Faradaic charge transfer, ohmic and mass transport resistance, as well as the effect of spatial variations in state-of-charge between the cell and electrolyte reservoir. We validate the model by comparing voltage-current data during cycling to equivalent electrochemical measurements from a symmetric cell with organic reactants. With the exception of the mass transfer coefficient, all relevant model parameters are determined independently prior to the experiment using electrochemical impedance spectroscopy and cyclic voltammetry measurements. We find excellent agreement between the model and experiment for constant-current and constant-voltage cycling, across a wide range of current densities (40 – 200 mA/cm2) and volumetric flow rates (~ 10 – 140 mL/min). The relationship between the fitted mass transport coefficient and flow rate conforms to expectations from the literature [1-3], and reasonably approximates the analogous relationship from single-reservoir cell measurements. This work supports the notion that zero-dimensional electrochemical models can yield simple but predictive frameworks for optimizing organic flow battery performance and understanding how reactant decomposition and crossover contribute to capacity fade. References [1] M. Pugach, M. Kondratenko, S. Briola, and A. Bischi, "Zero dimensional dynamic model of vanadium redox flow battery cell incorporating all modes of vanadium ions crossover," Applied Energy, vol. 226, pp. 560-569, 2018/09/15/ 2018, doi: https://doi.org/10.1016/j.apenergy.2018.05.124. [2] J. L. Barton, J. D. Milshtein, J. J. Hinricher, and F. R. Brushett, "Quantifying the impact of viscosity on mass-transfer coefficients in redox flow batteries," Journal of Power Sources, vol. 399, pp. 133-143, 2018/09/30/ 2018, doi: https://doi.org/10.1016/j.jpowsour.2018.07.046. [3] J. D. Milshtein, K. M. Tenny, J. L. Barton, J. Drake, R. M. Darling, and F. R. Brushett, "Quantifying Mass Transfer Rates in Redox Flow Batteries," Journal of The Electrochemical Society, vol. 164, no. 11, pp. E3265-E3275, 2017, doi: 10.1149/2.0201711jes.

Most polymer-based membranes that are used in prototypical aqueous redox-flow batteries (RFBs) do not adequately prevent crossover of small-molecule reactants, causing high rates of capacity fade. Ceramic superionic conductor membranes are an attractive alternative due to their superior abilities to mitigate crossover;1 they can thus enable the deployment of electrolytes containing earth-abundant, small-molecule reactants.2 3 We test the performance and stability of a von Alpen sodium superionic conductor Na3.1Zr1.55Si2.3P0.7O11 (NaSICON) as an RFB membrane by examining its resistance, permeability and interfacial morphology as a function of electrolyte composition and temperature. The resistance of NaSICON is stable for several weeks while immersed in neutral to strongly alkaline ([OH-] = 3 M) aqueous electrolytes, and its permeability to polysulfide-based and permanganate-based small-molecule RFB reactants is negligible compared to that of Nafion. The glassy phase of the NaSICON microstructure at the membrane-electrolyte interface undergoes small amounts of etching while in contact with aqueous electrolytes containing sodium ions, which becomes more extensive when potassium ions are present in the electrolyte, leading in certain instances to complete disintegration of the membrane. We report a ferrocyanide-permanganate flow cell at a pH of 14.5 with a ~ 700 μm-thin NaSICON membrane supporting weeks of cycling with apparently negligible reactant crossover and very low capacity fade (< 0.04 %/day). Area-specific resistance of NaSICON falls dramatically with increasing temperature and decreasing membrane thickness, and a membrane that is 100 µm thick or thinner can enable power densities at above-ambient temperatures that are comparable to power densities of polymer membrane-containing flow cells. (1) Yu, X.; Gross, M. M.; Wang, S.; Manthiram, A. Aqueous Electrochemical Energy Storage with a Mediator-Ion Solid Electrolyte. Advanced Energy Materials 2017, 7 (11), 1602454, https://doi.org/10.1002/aenm.201602454. DOI: https://doi.org/10.1002/aenm.201602454 (acccessed 2021/03/12). (2) Wei, X.; Xia, G.-G.; Kirby, B.; Thomsen, E.; Li, B.; Nie, Z.; Graff, G. G.; Liu, J.; Sprenkle, V.; Wang, W. An aqueous redox flow battery based on neutral alkali metal ferri/ferrocyanide and polysulfide electrolytes. J. Electrochem. Soc. 2016, 163 (1), A5150-A5153. DOI: 10.1149/2.0221601jes]. (3) Colli, A. N.; Peljo, P.; Girault, H. H. High energy density MnO4-/MnO42- redox couple for alkaline redox flow batteries. Chem Commun (Camb) 2016. DOI: 10.1039/c6cc08070g.