Tang Ao, Researcher at the Institute of Metal Research, Chinese Academy of Sciences CEJ: Surface Engineered carbon felt for highly reversible iron anode of all-iron flow battery

 

Low-cost all-iron flow batteries are expected to be used for large-scale energy storage. However, the poor reversibility of iron deposition/dissolution at the anode has largely hindered the further development of all-iron flow batteries.

 

Recently, the team of Tang Ao, a researcher at the Institute of Metal Research, Chinese Academy of Sciences, reported a surface-engineered carbon felt with abundant carbon defects, which achieved highly reversible Fe deposition/dissolution in all-iron flow batteries. Nanoscale pores were introduced on the surface of carbon fibers through a multi-step electrode treatment method. The modified carbon felt has abundant carbon defects, which effectively promotes Fe/Fe2+ kinetics, improves the nucleation and growth morphology of Fe, and is conducive to the uniform deposition and complete dissolution of Fe, thereby greatly improving the reversibility of the Fe anode.

 

Theoretical calculations attribute the improvement of Fe/Fe2+ reversibility to the strong adsorption and hybridization enhancement between Fe2+ and carbon defects, and the all-iron flow battery test using modified carbon felt shows a high power density of 80 mW cm-2 and a stable charge-discharge cycle of more than 250 cycles, and the Coulombic efficiency reaches 99%. These results highlight the importance of electrode design in the development of high-performance all-iron flow batteries with stable iron anodes.

 

 

Figure 1 Schematic diagram of electrode modification process

 

“Surface Engineered Carbon Felt Toward Highly Reversible Fe anode for all-iron Flow Batteries”Chemical Engineering Journal

 

Research Background

Among all the new redox electrochemical systems, all-iron flow batteries using ferric chloride in both half-cells have attracted much attention due to their low price and abundant raw material supply. In addition, the use of highly soluble FeCl2 as the active material in both half-cells can technically prevent cross-contamination and provide a moderate cell voltage of 1.21 V (vs. SHE).

 

However, the irreversibility of iron deposition/dissolution at the anode easily leads to low Coulombic efficiency and short cycle life, especially at high operating current density. In addition, the iron clusters formed under uneven Fe deposition may be stripped from the carbon felt electrode, thereby reducing the anode capacity of the all-iron RFB, which brings difficulties to long-term energy storage and release.

 

Fundamentally, the iron deposition and stripping process is closely related to the solvation structure of Fe2+ and the electrode interface activity. At present, the commonly used electrode material for all-iron RFB is a porous carbon felt electrode produced by carbonization of oxide fibers, which has high anti-corrosion stability in aqueous electrodes. Although carbon felt has a large redox reaction surface area, its electrochemical activity as an iron anode is still far from satisfactory.

 

A team led by Tang Ao, a researcher at the Institute of Metal Research, Chinese Academy of Sciences, reported a surface-engineered carbon felt that can provide a highly reversible Fe anode for high-performance all-iron flow batteries. Based on a multi-step surface engineering approach, nanoscale pores were successfully etched on the fiber surface of the carbon felt for the first time, and it exhibited a larger electrochemical specific surface area and better hydrophilicity.

 

Subsequent characterization and electrochemical analysis further confirmed the presence of abundant carbon defects on the nanoscale pores of the modified carbon felt, which significantly promoted the reversibility and electrode kinetics of the Fe/Fe2+ redox reaction. In addition, the iron deposition morphology further demonstrated the uniform and dendrite-free Fe growth on the modified carbon felt, while the simulation and theoretical calculation results attributed the improvement of Fe deposition to the preferred initial fine-grained Fe nucleation on the nanopores and the enhanced interfacial charge transfer between iron and carbon defects.

 

With the modified carbon felt, the symmetric cell achieved good cycling stability over 2000 cycles, while the all-iron full cell achieved a power density of 80 mW cm-2 and a Coulombic efficiency of 99% over 250 cycles, indicating a significant improvement in cycling performance and stability.

 

Core Content

 

b）Cu（OH）2-0.5@CF、Cu（OH）2-1@CF和Cu（OH）2-2@CF; c）CuO-0.5@CF、CuO-1@CF和CuO-2@CF; d）Cu-0.5@CF、Cu-1@CF和Cu-2@CF; e）CF-0.5、CF-1和CF-2的SEM

 

To investigate the effect of Cu(NO3)2 concentration on electrode morphology, three CFs were prepared and immersed in different concentrations of Cu(NO3)2 (i.e., 0.5 M, 1.0 M, and 2.0 M). SEM analysis showed that Cu(OH)2 nanosheets were uniformly and densely formed on the carbon fiber surface by the introduction of Na2CO3, and the density of Cu(OH)2 nanosheets increased with the increase of Cu(NO3)2 concentration.

After further decomposition of Cu(OH)2 by thermal treatment, CuO nanoparticles were uniformly dispersed on the carbon fiber surface (Figure 1c), which were further reduced to Cu nanospheres on the fiber surface by carbothermal reaction (Figure 1d). With the removal of Cu nanospheres, nanoscale pores were finally formed on the carbon fiber surface of all three samples (Figure 1e).

 

 

Figure 2 a) BET surface area and DLC of pcf, tcf, cf-0.5, cf-1, and cf-2; b) corresponding current versus scan rate for CV recorded at 0.04 V (vs. Ag/AgCl); c) electrolyte contact angle; d) average weight change; e) schematic diagram of the experimental setup for contact resistance measurements.

 

To further characterize the physical properties of the carbon felt and determine the ideal Cu(NO3)2 concentration for electrode treatment, the team first performed BET testing. The results in Figure 2a show that the surface area of ​​the Cu(NO3)2-treated CF increased significantly compared to PCF and heat-treated CF (TCF).

 

This was further confirmed by double-layer capacitance testing, where the electrochemical specific surface area (ESSA) value of the Cu(NO3)2-treated CF was smaller (Figure 2a). Compared with the physical adsorption method, the electrochemical test method takes into account the effect of wettability and can provide a more accurate electrochemical active area.

 

From the above two methods, it can be seen that the Cu(NO3)2 treatment not only increases the specific surface area of ​​the carbon felt, but also improves its wettability, which is further confirmed by the contact angle test (Figure 2c).

 

In addition, the weight loss of the treated CF is also given in Figure 2d. After Cu(NO3)2 treatment, the mass loss of CF is obvious. In particular, the weight loss of the treated CFs showed an increasing trend with increasing Cu(NO3)2 concentration, which originates from the oxidation of carbon during treatment and potentially leads to a decrease in mechanical strength.

 

Therefore, it can be seen that the measured contact resistance increases with increasing Cu(NO3)2 concentration (Figure 2 e), with CF-2 showing the maximum value. Considering all the above factors, CF-1 exhibits the best balanced performance.

 

 

Figure 3 a) Raman spectra, b) XPS spectra of C 1s and c) O 1s, and d) XRD spectra collected for PCF, TCF, and CF-1; e) Cyclic voltammogram of Fe/Fe2+ redox reaction and f) electrochemical impedance spectra of PCF, TCF, CF-0.5, CF-1, and CF-2; Constant current nucleation overpotentials of PCF, TCF, CF-0.5, CF-1, and CF-2 at g) 10 mA cm-2 and h) 20 mA cm-2; i) Chronoamperometry (CA) curves of PCF, TCF, CF-0.5, CF-1, and CF-2 at a constant potential of −0.8 V

 

To explore the effect of surface engineering on electrode performance, Raman tests were first performed on PCF, TCF, and CF-1. By comparison, it is observed that CF-1 has a higher ID/IG value of 1.53, which is greater than the ID/IG values ​​of PCF (1.38) and TCF (1.34), indicating that the amount of defects on CF-1 has increased significantly.

 

To further confirm the formation of defects, we also analyzed the XPS spectra of C 1s and O 1s of the samples. The C1s high-resolution XPS spectrum in Figure 3b shows the presence of signals of graphitized carbon (284.7 eV), defective carbon (285.1 eV), C-O (286.6 eV), C = O (287.8 eV), and O-C = O (289 eV). The defective carbon contents of PCF, TCF, and CF-1 were calculated to be 6%, 13%, and 20.9%, respectively, indicating that a large number of edge sites were generated during the carbothermal reaction.

 

Meanwhile, the O 1s spectrum consists of three peaks attributed to C = O (531.3 eV), C-OH (532.8 eV), and O = C-OH (534.2 eV) (Fig. 3f), indicating that the content of oxygen functional groups on CF-1 is higher than that on PCF and TCF.

 

In addition, the surface structural features of PCF, TCF, and CF-1 were compared by XRD (Fig. 3d). When the diffraction peak intensity decreases, the half-peak width narrows in the order of CF-1>TCF>PCF, which also indicates that more disordered structures are formed on CF-1. This also highlights the high degree of defects consistent with the Raman and XPS results.

 

To further explore the effect of Cu(NO3)2-treated CF on the electrochemical performance of iron anode, cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) tests were first performed on PCF TCF, CF-0.5, CF-1, and CF-2. Compared with other samples, CF-1 shows a small potential separation peak in CV as well as a clear reduction peak at −1.23 V (Figure 3e), which means that the Fe/Fe2+ reversibility is greatly enhanced.

 

In addition, the galvanostatic electrochemical deposition test shows that CF-1 has the lowest galvanostatic nucleation overpotential at 10 and 20 mA cm−2 (Figure 3g-h), which is very beneficial for fast and uniform iron deposition. Chronoamperometry (CA) tests further show that both PCF and TCF exhibit much higher response currents, indicating that the uncontrollable 3D growth of Fe can accumulate and eventually form Fe dendrites. In contrast, CF-1 shows a much lower response current, indicating that the 3D growth of Fe deposition is alleviated (Figure 3i).

 

 

Figure 4 Morphology of Fe deposition on PCF and CF-1 charged at 20 mA cm-2 for a-b) 20 s, c-d) 40 s, and e-f) 900 s; g) Simulation of Fe deposition and current density distribution on PCF and h) CF-1; i) SEM morphology of PCF and j) CF-1 charged at increasing current density for 15 min

 

The simulation results are fully consistent with the SEM morphology of Fe plating. When the current density of Fe deposition is further increased from 20 mA cm-2 to 40 mA cm-2, large Fe dendrite clusters can be clearly seen on PCF fibers (Figure 4 i). Such dendrite clusters are easily detached with electrolyte cycling, which may limit the Fe stripping process and lead to low Coulombic efficiency. In contrast, CF-1 can still deposit Fe without dendrites at high current density (Figure 3 j). In addition to the excellent Fe deposition performance, CF-1 also shows a neat fiber surface without residual Fe after complete Fe stripping, which together with the above results highlights the superiority of CF-1 in achieving a highly reversible Fe plating/stripping process.

 

 

Figure 5 a) Comparison of Fe adsorption energies on C 001, S 001, and D 001; b) Diffusion energy barriers of Fe 2+ on C 001, S 001, and D 001; c) Calculated total DOS of Fe on C 001, d) S 001, and e) D 001; f) COHP and charge density difference of C 001, g) S 001, and h) D 001; i) Schematic diagram of Fe nucleation and growth on PCF and CF-1

 

In order to obtain a fundamental understanding of the enhanced Fe-plating treatment of CF, the team performed first-principles calculations based on density functional theory. Figure 5a first shows that the adsorption energy of Fe on single vacancy defects (S 001) and divacancy defects (D 001) is much lower than that on a defect-free carbon surface (C 001), indicating that Fe nucleation is preferentially initiated at carbon defects.

 

In addition, the diffusion barrier of iron on the defect is also relatively large (7.62 eV for S 001 and 6.3 eV for D 001), which is significantly higher than the value on the defect-free site (0.76 eV), indicating the strong anchoring effect of iron on the defect. In addition, the density of states calculation further indicates the hybridization between Fe-3d orbitals and C-2p orbitals (Figure 5c-e). More specifically, D 001 shows a larger overlap than C 001, corresponding to the enhanced d-p hybridization between Fe-3d and C-2p orbitals.

 

This defect-induced modulation of the electrode surface electronic state is very beneficial for the charge transfer of the anodic Fe2+/Fe reaction. This is also confirmed again by ICOPH and charge density difference analysis, where strong Fe-C bond strength and enhanced interfacial charge interaction are clearly observed between iron and carbon defects (Figure 5 f-h).

 

Based on the theoretical analysis, the basic mechanism of iron plating is finally shown in Figure 5i. As mentioned above, Fe tends to form clusters on the carbon surface of the original CF, resulting in localized Fe deposition at the initial nucleation site. In contrast, Fe preferentially and continuously initiates nucleation at the carbon defect sites of the treated CF, which prevents the formation of iron clusters and induces the uniform growth of Fe.

 

 

Figure 6 a) Long-term cycling stability test of symmetric cells; b) Schematic diagram and cycling stability of full cells; c) Rate performance, d) Power density and e) Nyquist plot of all-iron flow batteries using PCF and CF-1; f) Long-term cycling stability of full cells using CF-1 with Fe mesh; g) Comparison of the proposed all-iron RFB with other reported all-iron RFBs

 

After confirming the superiority of the treated CF, flow battery tests were finally performed. Due to the highly reversible Fe anode, the symmetric cell assembled with CF-1 first exhibited lower voltage hysteresis (0.3 V) and repeated cycling stability of 2000 cycles at 20 mA cm-2 (Figure 6a), which was significantly better than that of the cell with pristine CF. In addition, the all-iron flow battery test showed that the capacity of the pristine CF decayed rapidly after only 20 cycles at 20 mA cm−2 (Figure 6b).

 

In addition, the rate performance of the full cell was further compared in Figure 6c. As the current density increases from 10 mA cm-2 to 40 mA cm-2, it is noted that CF-1 always delivers high Coulombic efficiency, which again verifies the significantly enhanced Fe plating/stripping reversibility at the anode.

 

Finally, the cycling stability of the all-iron flow battery is evaluated by further introducing an additional Fe mesh at the anode. Notably, the CF-1-based flow battery achieves an excellent combination of high cycling efficiency (> 99% CE), long cycling stability (260 h), and high capacity retention (> 99%), which is superior to the all-iron flow batteries reported in the open literature.

 

Conclusions and Outlook

 

In this paper, a multi-step surface engineering approach is developed to modify the surface of carbon felt fibers with abundant carbon defects, which facilitates the uniform deposition and complete dissolution of Fe on the carbon felt electrode, thereby obtaining a highly reversible Fe anode for all-iron flow batteries.

 

Based on Cu(NO3)2 solution, nanoscale pores are introduced on the carbon fiber surface for the first time through a multi-step process including initial precipitation reaction, subsequent heat treatment, and acid washing. By evaluating the physical properties of the modified CF, the optimal Cu(NO3)2 concentration for electrode treatment was determined to be 1 M. The modified CF at the optimal Cu(NO3)2 concentration was subsequently characterized as having abundant carbon defects, indicating that the electrochemical reversibility and kinetics of the anodic Fe/Fe2+ redox reaction were greatly enhanced. Further iron deposition morphology characterization also confirmed the uniform deposition and complete stripping of iron on the modified carbon felt. While finite element analysis showed that the nanopores can effectively regulate the iron deposition current density during the Fe2+ reduction process, thereby preventing the formation of iron clusters.

 

In addition, density functional theory also demonstrated the preferential iron nucleation and strong iron anchoring at carbon defects, as well as the strong hybridization of carbon defects on the Fe-3d and C-2p orbitals on the carbon fiber surface. All of the above results reveal the positive role of carbon defects in achieving rapid and uniform iron deposition.

 

Finally, the symmetric cells assembled with modified CF exhibited high cycling stability with low voltage hysteresis over 2000 cycles, while the full cell further showed a high power density of 80 mW cm-2 and an average energy efficiency of 70% over 250 charge-discharge cycles, significantly outperforming the pristine CF-based all-iron flow battery.

 

This study not only demonstrates the effectiveness of carbon defects in improving the reversibility of the iron anode, but also highlights the importance of electrode design as an effective tool for developing high-performance all-iron flow batteries in addition to electrolyte regulation.