Research Highlights

This work investigated the use of potassium-exchange sulfonated poly(ether ether ketone) (SPEEK-K) membranes in PFRFB as a substitute for the expensive Nafion membranes. SPEEK-K with optimized sulfonation degree enables PFRFB to achieve a high average coulombic efficiency of 99.80% and an outstanding energy efficiency of 90.42% at a current density of 20 mA cm⁻². Meanwhile, to overcome the kinetic limitations of sulfur redox reactions, CuS-modified carbon felt electrodes exhibit superior catalytic performance, enabling PFRFB to have higher and more stable energy efficiency during cycling. In the battery, the combination of cost-effective membranes with catalytic electrodes maintains 99.54% capacity retention after 1180 cycles and exhibits excellent power densities (up to 223 mW cm⁻²). The significantly improved electrochemical performance under cost reduction makes PFRFB more promising for large-scale energy storage systems.

The author first investigated the morphology of the membranes. After K+ exchange, the cross-sectional morphology of SPEEK-K (DS57-K) became rougher relative to SPEEK (Figure 2a), and the surface morphology of the membrane was also rough (Figures 2b–2d). From DS47-K and DS57-K to DS67-K, there were more K+ ions present, which was proportional to the amount of H+ in the sulfonic groups. Additionally, EDS results validated the uniform distribution of K+ ions within the matrix (Figures 2b-2d). AFM images showed that compared to N212 and SPEEK membranes (Figure 2e), N212-K and SPEEK-K membranes (Figures 2f–2h) displayed more pronounced dark areas. As the DS increased, the dark areas on the graph grew larger, indicating that K+ ions altered the water channels in the membrane, thereby increasing hydrophilicity.

Subsequently, the authors compared the physicochemical properties of the membranes. Table 1 measured the thickness, IEC, SR, AR, and K+ conductivity values of the membranes. Due to the involvement of hydrophilic sulfonic acid groups, SPEEK-K membranes exhibited higher IEC and SR values compared to N212-K. As DS increased, the IEC and SR values of SPEEK-K membranes increased from 1.27 to 1.84 and from 3.81% to 4.85%, respectively, confirming the role of sulfonic acid groups in promoting the absorption and retention of water molecules within the membrane. Generally, high SR values are beneficial for ion conductivity, resulting in a decrease in AR values from 1.59 Ω·cm2 for N212-K to 0.80, 0.74, and 0.71 Ω·cm2 for DS47-K, DS57-K, and DS67-K, respectively. Correspondingly, the K+ conductivity of the membranes increased from 16.84 mS cm−1 for N212-K to 17.16, 19.08, and 22.89 mS cm−1 for DS47-K, DS57-K, and DS67-K, respectively.

图3a显示了空白储层中[Fe(CN)₆]^3-浓度的变化以比较膜的渗透性，SPEEK-K膜的渗透速率随DS的增加而增大，因为DS越高，聚合物基质中-SO₃H基团越多，有效提高K⁺电导率，并加速活性物质通过膜。与N212-K相比，所有SPEEK-K膜都表现出较低的活性物质渗透性。根据图3b，膜的渗透速率从2.23×10⁻⁷cm²min⁻¹（N212-K）和2.37×10⁻⁷cm²min⁻¹（DS67-K），进一步降低至1.26×10⁻⁷cm²min⁻¹（DS57-K）以及1.15 × 10⁻⁷cm²min⁻¹（DS47-K）。DS57-K和DS47-K的离子选择性（电导率与渗透率之比）分别为0.1%和0.2%。所有SPEEK-K膜的离子选择性值均明显高于N212-K，其中，DS57-K的离子选择性最高，这归因于铁氰化物渗透的抑制以及K⁺电导率的增强。

Subsequently, the authors investigated the impact of membranes on battery performance. Figure 4a demonstrates that batteries utilizing DS57-K exhibit higher discharge capacity (60.8 mAh) compared to those based on N212-K (53.8 mAh), which aligns with the lower ion permeability and higher ion selectivity of the DS57-K membrane. Furthermore, the batteries display a high CE of approximately 99.80% at mid-rates. However, EE for batteries based on DS57-K is significantly higher and more stable than those based on N212, achieving a steady 90.42% EE after 1051 cycles, whereas using N212-K membranes results in an EE of only 77.82% after 879 cycles (Figure 4b). Batteries based on DS57-K also show higher CR (Figure 4c). Additionally, as shown in Figure 4d, employing DS57-K membranes significantly prolongs the self-discharge duration of PFRFB, with batteries based on DS57-K maintaining for about 717 hours, which is six times longer than those based on N212-K (98 hours). All these findings confirm that due to the SPEEK-K framework and optimized DS, the DS57-K membrane effectively retards the crossover of redox species.

The authors also investigated CuS-CF electrodes for PFRFB. SEM images reveal distinct morphological differences between pristine CF (Fig. 5a and 5b) and CuS-CF (Fig. 5c and 5d), with the former exhibiting a clean and smooth surface, while the latter displays a flower-like protruding structure composed of innumerable interconnected nanosheets, indicating successful growth of CuS on the CF surface. This flower-like structure can provide more active sites for redox reactions.

The author subsequently investigated the impact of electrode characteristics on battery performance. As shown in Figure 6a, multiple pairs of redox peaks were observed on the CuS-CF, whereas only one pair of major redox peaks could be identified on the pristine CF electrode, indicating that the CuS-CF electrode possesses superior catalytic properties. As depicted in Figure 6b, EIS testing revealed a charge transfer resistance of 2.14 Ω for CuS-CF, lower than that of pristine CF (3.34Ω), suggesting that the CuS-CF electrode exhibits faster electron transfer rates between the electrode interface and electrolyte. Consequently, the surface coverage of flower-like CuS provides more active sites for the redox reactions of polysulfides, thereby enhancing electrochemical activity. The author assembled PFRFB batteries using pristine CF and CuS-CF electrodes separately. From the charge/discharge curves (Figure 6c), it can be seen that the CuS-CF based cell exhibits smaller polarization in the first cycle compared to the pristine CF cell, which further expands during cycling, while the pristine CF based cell shows a significant IR drop. Correspondingly, the CuS-CF electrode enables the PFRFB cell to maintain a more stable EE during cycling, remaining at 75.80% after 879 cycles, whereas when using CF electrodes, the EE was only 66.54% after 751 cycles (Figure 6d). These significant improvements in battery performance originating from the electrodes can be attributed to the increased number of catalytic and active sites in CuS-CF.

The author also investigated the comprehensive performance of PFRFB. As shown in Figure 7a, initially, the discharge capacity (DC) can reach 99.54% of the theoretical value, and CE remains very stable during long-term cycling tests (still at 99.95% after 1180 cycles), even at a high current density of 50 mA cm-2, the average EE of PFRFB can reach 75.41%. In this study, the capacity of the battery is determined by the catholyte, and when the concentration of redox active materials (K3[Fe(CN)6]) in the catholyte increases to 0.5 M and 0.8 M, the DC values show significant enhancement (Figure 7b and 7c). When 0.5 M K3[Fe(CN)6]+2.0 M KCl is used as the catholyte, the battery performs excellently at a current density of 20 mA cm-2, with CE reaching 99.28%, VE reaching 80.46%, and EE reaching 79.27%, and these key parameters remain stable after more than 230 cycles (Figure 7b). When the catholyte is 0.8 M K3[Fe(CN)6]+2.0 M KCl (Figure 7c), the battery exhibits minimal capacity decay after 60 cycles (180 test hours) at a current density of 20 mA cm-2, with a capacity decay rate of 0.0709% per hour, and the average CE, EE, and VE of this more concentrated PFRFB battery reach 97.21%, 83.41%, and 85.23%, respectively.

The rate performance of the PFRFB battery was evaluated from 20 to 140 mA cm⁻² with an interval of 20 mA cm⁻². At all applied current densities, CE remained close to 100%, indicating that both the DS57-K membrane and CuS-CF electrode exhibited high stability. As the current density increased from 100 to 140 mA cm⁻², VE and EE significantly decreased, which may be related to the increase in electrochemical polarization and ohmic polarization within the battery. When the current density returned to 20 mA cm⁻², both VE and EE recovered to their initial high values. For comparison, a PFRFB battery (reference battery) assembled and tested with pristine CF as electrodes and N212-K as the ion exchange membrane showed, from the polarization curve, that the peak power density of the battery with optimized electrodes and membranes reached 223 mW cm⁻², higher than that of the reference battery at 199 mW cm⁻² (Figure 7e). Therefore, optimizing membrane and electrode materials can reduce the cost of the PFRFB system and improve the kinetics of polysulfide redox reactions.

In summary, the authors reported a neutral PFRFB system constructed using a low-cost SPEEK-K membrane and a highly catalytically active CuS-CF electrode. Through optimization of DS, the DS57-K membrane enables the PFRFB cell to achieve a high average CE of 99.80% and an excellent EE of 90.42% after 1051 cycles at a current density of 20 mA cm-2, surpassing the commercial N212 membrane in both performance and price. Meanwhile, the CuS-CF electrode exhibits superior catalytic effects on the polysulfide redox reaction, reducing polarization during long-term cycling. The combination of the DS57-K membrane with the CuS-CF electrode allows the PFRFB cell to maintain high efficiency even at a high current density of 50 mA cm-2. Furthermore, adopting this strategy with a higher concentration of 0.8 M K3[Fe(CN)6] as the positive electrolyte in the proposed PFRFB cell achieves an outstanding power density of 223 mW cm-2. Therefore, the progress made in this study promotes the development of neutral PFRFB for use in high-performance, low-cost energy storage systems.

Product Series：