Galvanostatic charging and discharging (GCD) curves are displayed in Figs. S11a–c (Supporting information), at the applied current density of 3~5 mA/cm
2. Two discernable discharge plateaus are obtained at 1.5~1.1 V and 0.3~0 V, respectively. The presence of ferric dimer [Fe
2O(phen)
4Cl
2]
2+ with low redox potential accounts for the presence of a second discharge plateau (Scheme S1 in Supporting information) [
21,
22]. It derives from the dimerization of Fe(phen)
3Cl
2 with releasing of protonated ligand (H(phen)
+). More dimers are generated at lower current density, owing to extended discharging reaction time. Nevertheless, this trivially affects the volume capacity, since the reverse reaction could happen between [Fe
2O(phen)
4Cl
2]
2+ and H(phen)
+, and affords [Fe(phen)
3]
2+ [
21,
22]. A trade-off effect is obtained between Coulombic efficiency (CE) and voltage efficiency (VE) at different discharge voltage cut-offs (Fig. S12 in Supporting information). Under the current density of 5 mA/cm
2, CE reaches 63% at relatively low VE (88%) at 1.0 V discharge voltage cut-off, whilst 98% CE and 46% VE are obtained at 0.001 V discharge voltage cut-off. The incomplete release of storage energy (utilization of capacity) at high voltage cut-off accounts for relatively poor CE. Under large current density, the first discharge plateau extended at the expense of the second discharge plateau (Figs. S11d–i in Supporting information). This means large current density is beneficial to battery performance. To avoid dimmer formation during the discharge process, the discharge voltage cut-off is set to be 1.0 V, and battery performance is further evaluated at different current densities (7~20 mA/cm
2) for the first discharge plateau (
Fig. 1f). When increasing the current density, the charge plateau ascends, accompanied by a descent of the discharge plateau. The declined capacity utilization efficiency is attributed to increased ohmic resistance at large current density, as well as overpotential related to mass transport (polarization). Under optimized current density (7 mA/cm
2), the charge and discharge volume capacities are 131.73 and 127.70 mAh/L, which correspond to 98.31% and 95.36% of theoretic capacity. At the same time, the battery exhibits 96.45% CE, 92.75% VE, and 89.46% EE (
Fig. 1g). The impact of electrolytes on the electrochemical performance and recyclability of Fe(phen)
3Cl
2 is investigated as well, and glycine-HCl buffer with a pH of 2.5 turns out to be the optimal electrolyte (Fig. S13 in Supporting information). Under optimized conditions (current density 7 mA/cm
2, glycine-HCl buffer, pH 2.5), the battery is charged and discharged between 1.8 V and 0.001 V (
Fig. 1h). During 20 cycles, the average CE, VE, and EE reach 93.25%, 92.61%, and 86.35% (
Fig. 1i and Fig. S14 in Supporting information). The volume capacity at the first cycle is 117.34 mAh/L (87.31% of theoretic value) and declines to 80.48 mAh/L (60.10% of theoretic value). The capacity fading rate is 1.57% per cycle, better than one without glycine-HCl butter (Figs. S8 and S9 in Supporting information). Self-discharge of [Fe
2O(phen)
4Cl
2]
2+ and its precipitation may lead to capacity decay (Scheme S1), together with the crossover of H(phen)
+ [
21,
22]. Another possibility is the decomposition of Fe(phen)
32+ and Fe(phen)
33+ in aqueous electrolyte (Scheme S2 in Supporting information), which, however, can be efficiently retarded by introducing glycine-HCl buffer.