High energy density electrolytes for H2/Br2 redox flow batteries, their polybromide composition and influence on battery cycling limits

Hydrogen–bromine redox flow batteries (H2/Br2-RFB) are a promising stationary energy storage solution, offering energy storage densities up to 200 W h L−1. In this study, high energy density electrolytes of concentrated hydrobromic acid of up to 7.7 M are investigated. Particular polybromide ion (Br2n+1−; n = 1–3) concentrations in the electrolyte at different states of charge, their effect on the electrolytic conductivity and cell operation limits are investigated for the first time. The concentrations of individual polybromides in the electrolytes are determined by Raman spectroscopy. Tribromide (Br3−) and pentabromide (Br5−) are predominantly present in equal concentrations over the entire concentration range. Besides Br3− and Br5−, heptabromide (Br7−) exists in the electrolyte solution at higher bromine concentrations. It is shown that polybromide equilibria and their constants of Br3− and Br5− from literature are not applicable for highly concentrated solutions. The conductivity of the electrolytes depends primarily on the high proton concentration. The presence of higher polybromides leads to lower conductivities. The solubility of bromine increases disproportionately with increasing bromide concentration, since higher polybromides such as Br7− or Br5− are preferably formed with increasing bromide concentration. Cycling experiments on electrolyte in a single cell are performed and combined with limitations due to electrolyte conductivity and bromine solubility. Based on these results concentrations of the electrolyte are defined for potential operation in a H2/Br2-RFB in the range 1.0 M < c(HBr) < 7.7 M and c(Br2) < 3.35 M, leading to a theoretical energy density of 196 W h L−1.

In the ESI the reader will find detailed information on the measurement methods used, parameter settings for measurements, chemicals and purities as well as evaluation steps. In addition, Raman results on aqueous HBr/Br2/H2O electrolytes are shown in figures and tables.

Electrolyte preparation
For this electrolyte investigation an aqueous 7.7 M HBr solution is the initial electrolyte solution. A lower concentration compared to commercially available hydrobromic acid (HBr 48 wt%) with a difference of around 1.1 M is used, to allow the replication of other dilutions. Charging and discharging are theoretically possible between 7.7 M HBr and 0 M HBr. This range is evaluated with 24 samples, with a concentration change of Δc(HBr) = -0.335 M and Δc(Br2) = 0.167 M per sample. The correlation between Δc(HBr) and Δc(Br2) is stoichiometric and represented by eq. 1. The volume (V) of each sample is V = 10 mL. Concentrations of HBr and Br2 in solution are shown in table 2 in the main manuscript. Each sample is used for the investigation of (i) electrolytic conductivity of the solution, (ii) detection of a possible miscibility gap of the solutions and (iii) Raman investigations for polybromide detection as a basis for the calculation of polybromide concentrations and equilibrium constants of the polybromide equilibria according to equations 4 and 5 (main manuscript). In addition to HBr/Br2/H2O samples described above, samples without Br2 based on HBr/H2O are mixed in order to investigate reference conductivities of pure HBr/H2O electrolytes and compare them with solutions including Br2. 24 samples are prepared with a volume of V = 10 mL and a concentration change of Δc(HBr) = -0.335 M in aqueous solution. For all samples, the amount of water is constant starting from a solution with c(HBr) = 7.7 M corresponding to the theoretical conditions in a positive bromine half-cell of a H2/Br2-RFB. For a cell test, V = 60 mL of discharged electrolyte with c(HBr) = 7.7 M HBr in H2O was prepared.

Cycling cell test
A cell test is performed with a test cell developed at Fraunhofer ICT. As membrane electrolyte assembly (MEA) a Nafion 117 membrane with single sided Pt/C catalyst loading (3 mg Pt cm -2 ) from Baltic Fuel Cells (Germany) is used as hydrogen half-cell catalyst. The membrane has a geometrically active surface area of 40 cm². The gas diffusion electrode of the negative hydrogen half-cell consists of a gas diffusion layer (GDL) BC29 from SGL Carbon (Germany) and a current collector FU 4369 from Schunk Kohlenstofftechnik (Germany) with a milled meander structure to supply the gas anode with hydrogen and to remove formed hydrogen and water. The anode side is continuously fed with hydrogen from a gas cylinder with 100 mL min -1 at 1.013 bar or higher, leading to stoichiometric factor λ ≥ 1.31. There is 31 % more hydrogen available compared to the amount needed during discharge. The hydrogen produced during the charging process is not stored. No external humidification of the hydrogen is applied during the measurements. The investigation concerns the bromine half-cell only. There, the electrode consists of a GFA 5 graphite felt from SGL Carbon (Germany), embedded in a 3 mm deep frame, which is closed by a glassy carbon (GC) current collector Sigradur G (HTW, Germany). The cell is charged and discharged while a flow rate of 20 mL min -1 at different current densities of +/-50 mA cm -2 , +/-100 mA cm -2 , +/-150 mA cm -2 , +/-200 mA cm -2 and +/-250 mA cm -2 is applied. Voltage thresholds of the experiments are at Emax = 1.5 V and Emin = 0.25 V. The electrolyte is located in a glass tank in which a GC rod Sigradur G (HTW, Germany) and a mercury sulphate reference electrode (RE) Hg/Hg2SO4/K2SO4(sat.) (SI Analytics, Germany) are installed and in contact with the electrolyte (figure 2 in main Electronic Supplementary Material (ESI) for RSC Advances. This journal is © The Royal Society of Chemistry 2021 ESI to Article "High energy density electrolytes for H2/Br2 redox flow batteries, their polybromide composition and influence on battery cycling limits" by Küttinger et al.
ESI to Article "High energy density electrolytes for H2/Br2 redox flow batteries, their polybromide composition and influence on battery cycling limits" by Küttinger et al. 2/9 manuscript). On one hand, the redox potential of the catholyte at the GC rod versus Hg/Hg2SO4/K2SO4(sat.) is measured. On the other hand the half-cell potential φ(Br2/Br -)i ≠ 0 during charge and discharge operation of the bromine half-cell connected via the electrolyte in the tube against Hg/Hg2SO4/K2SO4(sat.) is measured, according to figure 2 in the main manuscript. The negative hydrogen half-cell contains a dynamic hydrogen electrode (DHE) according to 1 to determine the potential of the hydrogen anode. The cell voltage ECell i ≠ 0 under load, the redox potential φ(Br2/Br -)redox versus a Hg/Hg2SO4/K2SO4(sat.) and half-cell potentials of the positive bromine half-cell φ(Br2/Br -)i ≠ 0 vs. the normal hydrogen electrode (NHE) and the negative hydrogen half-cell φ(H + /H2)i ≠ 0 vs. NHE were determined in parallel during cycling experiments. A so-called "residual potential" is calculated from these values according to eq. 1: It essentially reproduces the overvoltages between the anode and cathode, and represents the sum of electrolyte resistance and membrane resistance. Kinetic inhibitions and mass transport limitation at the electrodes are represented by the half-cell potentials. Qualitative noticeable occurrences of cell voltage behaviour are determined from the galvanostatic cycle test of a test cell.
The measurement setup is shown in figure 2 in the manuscript including all voltage and potential measurements.

Conductivity of electrolyte solutions
Electrolytic conductivities are measured for all 48 samples between the concentration limits at ϑ = 23°C using an LF1101 conductivity cell (SI Analytics/Germany). Electrolytes with HBr/Br2/H2O, and electrolytes without Br2 in solution are examined. Ohmic resistances of the electrolytes RElectrolyte are measured by potentiostatic impedance spectroscopy with an excitation of û = 10 mV (amplitude) in the frequency range between 1 MHz and 100 Hz at a potential offset of 0 mV and converted into electrolytic conductivities κ according to eq. 3. The cell constant KConductometer of the conductivity cell is determined using 1 M KCl solution and a known conductivity of 303.9 mS cm -1 at ϑ = 23 °C (determined by linear regression for ϑ = 23 °C from 2 (eq. 2): [K] = cm eq. 2 Ohmic resistances from the conductivity measurement RElectrolyte were corrected by the ohmic resistance of the cables and connections. Based on the miscibility gap and the measured electrolyte conductivities, an applicable operating range is defined in terms of states of charge (SOC) from 0 % to 100 % in the results chapter "Definition of the operating range for HBr/Br2 electrolytes".

Solubility of Br2 in HBr solutions
The The concentrations for c(Br3 -)eq, c(Br5 -)eq, c(Br2)eq, and c(HBr)eq in equilibrium are calculated using eq. 4 and 5 of the main article (K3 and K5) with K3 = 16 L mol -1 and K5 = 37 L 2 mol -2 for individual operating SOC points in each case iteratively including eq. 7.
c(HBr) eq = c(HBr) T − c(Br 3 − ) eq − c(Br 5 − ) eq eq. 7 The operating points are defined by the state of charge (SOC) of the electrolyte. The SOC range will be defined in chapter "Definition of the operating range for HBr/Br2 electrolytes". Concentrations are plotted as a function of SOC in chapters "Theoretical concentrations and polybromide equilibria" and "Polybromide distribution and polybromide concentration verification of polybromide equilibria" both in the main manuscript.
ESI to Article "High energy density electrolytes for H2/Br2 redox flow batteries, their polybromide composition and influence on battery cycling limits" by Küttinger et al.
ESI to Article "High energy density electrolytes for H2/Br2 redox flow batteries, their polybromide composition and influence on battery cycling limits" by Küttinger et al.

Results of investigation on HBr/Br2/H2O electrolytes by helps of Raman spectroscopy
Raman spectra of HBr/Br2/H2O electrolytes  Table S 1. Raman spectra were corrected for Rayleigh scattering. ESI to Article "High energy density electrolytes for H2/Br2 redox flow batteries, their polybromide composition and influence on battery cycling limits" by Küttinger et al.

Raman spectra of bromine Br2
ESI to Article "High energy density electrolytes for H2/Br2 redox flow batteries, their polybromide composition and influence on battery cycling limits" by Küttinger et al.