Despite the discovery that polysulfides suffer from slow reaction kinetics, they can still serve as a low-cost, highly soluble, and efficient anodic redox material for redox flow batteries (RFBs). In this study, the authors report a viable continuous ion layer adsorption and reaction method that enables scalable, controllable deposition of highly active electrocatalysts, grafting copper sulfide electrocatalysts onto graphite felt electrodes, which significantly promotes the redox reactions of polysulfides. With [Fe(CN)6]4−/3− as the catholyte and polysulfides as the anolyte, a flow battery was prepared with a power density of 116 mW cm−2 (220 mA cm−2), an energy efficiency of 77.7% (50 mA cm−2), and a capacity retention rate that is significantly superior to that of reported flow batteries. The performance of this polysulfide-[Fe(CN)6]4−/3− flow battery (Figure 1) has been significantly enhanced and has been proven to be an economically efficient, reliable, and durable large-scale grid energy storage solution. It offers the advantages of being environmentally friendly and low-cost, capable of storing large amounts of electrical energy and buffering the impact of the intermittency of wind and solar power generation on the resilience and quality of electrochemical energy storage (EES) systems for the power grid.

In this study, the authors employed the Successive Ionic Layer Adsorption and Reaction (SILAR) method to grow CuS on thermally treated graphite felt (GF), serving as an efficient catalyst for the redox reactions of polysulfide electrolytes. Compared to other methods, SILAR offers several advantages: it can be carried out at room temperature, does not require expensive and complex equipment, and allows for easy control of deposition even on large-area substrates, as illustrated in Figure 2.

Design and Assembly of the Redox Flow Battery：The design and assembly of the redox flow battery are as follows: The active electrode area is 13.2 cm², with a flow-through configuration without fluid channels. The thermally treated GF and CuS-coated GF (CuS-GF) are used as the cathode and anode, respectively. A graphite plate serves as the current collector and is in close contact with the electrodes. The battery is sealed with a silicone gasket, and a Nafion 212 membrane is used as the separator. A peristaltic pump and silicone tubing (outer diameter of 4 mm) are used to circulate the electrolyte between the cell stack and the storage tanks. The catholyte is 30 mL of 0.3 M K3Fe(CN)6 dissolved in a 1 M KCl solution, and the anolyte is 15 mL of 2 M K2Sx dissolved in a 1 M KCl solution, sealed in the electrolyte tanks during testing. The upper and lower voltage limits for battery testing are set at 1.8 V and 0.2 V, respectively.

In the article, the authors used Scanning Electron Microscopy (SEM) methods, as shown in Figures 4a-b, to display the morphology of the CuS catalyst coated on the GF. Overall, a uniform distribution of submicron and nanoparticle CuS can be observed on the graphite fibers. Additionally, Energy-Dispersive X-ray Spectroscopy (EDS) elemental mapping, as illustrated in Figures 4c-f, was utilized to analyze the distribution of elements. It was found that both copper and sulfur elements are clearly present on the surface of the graphite fibers, with identical distributions.

Further, the authors investigated the electrocatalytic activity of CuS. As shown in the CV in Figure 5a, the original GF electrode exhibits poor electrochemical reactivity towards polysulfides, while the thermally treated GF electrode shows a significant increase in current within the same potential range, which may be attributed to the presence of functional groups and enhanced wettability and activation capacity after treatment. Concurrently, the CuS-grafted GF electrode demonstrates a notably enhanced response within the potential range of −0.9 to 0.2 V vs. Hg/HgO. The broad current peaks in both cathodic and anodic processes indicate that polysulfides undergo multi-electron transfer, involving long-chain Sx2− (1<x<8) species. In fact, the redox species in the prepared anolyte are a mixture of various polysulfides with an average oxidation state of approximately S22−, rather than a single polysulfide species.

The GITT in Figure 5b reveals the reaction kinetics of polysulfides at the CuS-grafted electrode during the charge/discharge of the polysulfide-[Fe(CN)6]4−/3− full cell. For the original GF electrode, the overpotential observed during the charging process is significantly higher than that during the discharging process, implying that the reduction of polysulfide species (involving bond breaking) is kinetically slower than the oxidation process. Meanwhile, the CuS-grafted electrode substantially reduces the cell's overpotential during both charging and discharging processes.

Consistent results were also obtained from the electrochemical impedance spectroscopy (EIS) measurements, as shown in Figure 5c, which found significant differences in the Nyquist plots of the EIS spectra of the aforementioned full cells at different states of charge (SOC). The presence of CuS significantly reduces the total impedance of the cell by seven times, indicating a faster electron transfer process. Furthermore, the impedance of both cells at 0% SOC is higher than at 100% SOC, suggesting that the reduction reaction of long-chain polysulfides is relatively stagnant compared to the oxidation reaction of short-chain polysulfides.

In order to further understand the source of the overpotential, the authors used a four-electrode configuration to reliably and separately measure the local voltage drops between different cell components, as shown in Figure 5d. It was found that at a current density of 10 mA cm−2, the transmembrane voltage drop and the cathode potential of the two cells are similar, with the main difference being the anode potential related to the reaction with polysulfides. For the cell using the GF electrode, almost the entire charging overpotential comes from the anode (>120 mV), while the overpotentials of the [Fe(CN)6]4−/3− cathode (~6 mV) and the membrane (~28 mV) are quite small. Moreover, the anode overpotential is particularly large at the initial and final stages of oxidation, implying that the conversion of soluble polysulfides from S62- to S42- and from S22- to S2- requires higher energy barriers, while the conversion of S42- to S22- appears to be relatively faster. The CuS catalyst significantly reduces the overpotential of polysulfides throughout the charging process to about 15 mV, confirming the catalytic effect of CuS, but the catalytic reaction mechanism of polysulfides remains unclear.

In conclusion, the authors assembled and tested a polysulfide-[Fe(CN)6]4−/3− redox flow full cell using a CuS-GF anode and a GF cathode, and compared it with a cell using two GF electrodes. Figure 6a shows the voltage curves of the cell at a current density of 10 mA cm−2, revealing that the CuS-CF electrode plays a significant role in reducing overpotential, especially during the charging process. Figure 6b illustrates the constant current cycling performance of the flow battery at 50 mA cm−2 and 100% SOC, where an energy efficiency (EE) higher than 72.3% (with a maximum of 77.7%) and a Coulombic efficiency (CE) above 99% were observed over 160 full cycles (approximately 100 hours). The study also indicates that during the first 20 cycles, EE remains stable or even slightly increases due to the improved wettability of the electrodes and membranes, as well as the activation process of the catalyst, and then gradually decreases. Considering that CE remains almost constant, the decay of EE can be attributed to the decline in voltage efficiency (VE), which may be due to the precipitation of long-chain polysulfide species on the electrodes at high discharge states, thereby reducing the conductivity of the GF electrode. Additionally, a capacity retention rate of 98.2% indicates good cycling stability.

Figure 6c presents the efficiency of the cell at constant current densities ranging from 20 to 120 mA cm−2, showing that high CE (>99%) was achieved at all current densities, implying negligible self-discharge (which is usually caused by the crossover of redox species). Unprecedented EE of over 90%, 80%, 70%, and 50% were obtained at current densities of 20, 40, 60, and 120 mA cm−2, respectively. The polarization and power density curves of the two types of cells shown in Figure 6d demonstrate that the CuS-GF cell provides a maximum power density of 116 mW cm−2 at approximately 220 mA cm−2, which is more than four times that of the GF cell and the highest reported, confirming the excellent catalytic performance of the SILAR-deposited CuS electrode. Although flow batteries can operate at high currents, their power density is affected by the inherently lower cell voltage compared to other RFBs, such as VRBs.