Fe and Co containing N-doped graphene sulfur cathode for high performance Li–S batteries and in situ Raman investigation

Lithium–sulfur (Li–S) batteries are promising candidates for next-generation energy storage devices due to their high theoretical specific capacity, low cost, and environmental friendliness.¹ However, the practical application of lithium–sulfur batteries faces several challenges, including poor cycle life, low sulfur utilization, and the shuttle effect of lithium polysulfide (LiPSs).² In order to solve these problems, a large amount of research work has focused on the development of advanced cathode materials.

In recent years, transition metal decorative carbon materials have received much attention for their potential to improve the electrochemical performance of Li–S batteries. These materials combine the advantages of transition metals, such as high catalytic activity and strong affinity for sulfur substances, with the unique properties of carbon materials,^3,4 including good electrical conductivity, large specific surface area, and chemical stability. The integration of transition metals onto carbon substrates provides a promising method for optimizing the electrochemical performance of Li–S batteries. In addition, iron⁵ and cobalt⁶ based materials have become the focus of attention due to their ability to efficiently adsorb and catalyze LiPSs. Compared with mono-metal components,⁴ bimetallic components⁷ exhibit excellent electrochemical properties due to their diversity of composition and excellent catalytic activity. Compared with mono-metal oxides,⁸ bimetallic oxides⁹ can provide sufficient reaction sites. In addition, the synergistic effect of the two metals can produce additional active sites, improving stability. It is exciting that graphene hybrid materials¹⁰ have received much attention in the field of Li–S batteries. Graphene's¹¹ (G) unique two-dimensional structure, combined with its excellent electrical conductivity and mechanical properties, make it an ideal candidate for improving the electrochemical performance of Li–S batteries. The incorporation of G into carbon materials can enhance its electrical conductivity, promote the adsorption of LiPSs,¹² and provide a conductive pathway for electron transfer, thereby improving the utilization rate and cycle stability of sulfur. In addition, in situ Raman spectroscopy has become a powerful tool to understand the mechanism of sulfur reactions in Li–S batteries.¹³ By monitoring the vibration mode of sulfide during charge and discharge, in situ Raman spectroscopy can deeply understand the formation and evolution of intermediate sulfide, reveal the reaction kinetics and mechanism of Li–S batteries, and provide an opportunity to optimize the design of cathode materials. At present, the reaction mechanism of Li–S batteries capacity attenuation is rarely reported, and the application of in situ Raman technology is particularly important.

In this work, we designed a structure of CoFe₂O₄/Co₃Fe₇ nanoparticles coated with an N-carbon layer and a G layer by pyrolytic bimetallic MOFs (G/FeCo@N-C). N-carbon layers and G layers provide powerful conditions for sulfur storage and adaptation to volume expansion, providing polysulfide attachment sites and fast charge transfer channels. The synergistic effect of CoFe₂O₄ bimetallic oxide and Co₃Fe₇ bimetallic alloy offers good catalytic ability for LiPSs. In addition, the abundant catalytic sites of CoFe₂O₄/Co₃Fe₇ can accelerate the transformation of LiPSs. Importantly, by in situ Raman characterization, we revealed that G/FeCo@N-C exhibited superior electrocatalytic activity to Li–S chemistry, including redox processes that occurred during battery charging and discharging. Thanks to these advantages, the G/FeCo@N-C-S cathode at 1.0 C current density can maintain a high specific capacity of 440.3 mAh g⁻¹ after 1000 cycles, showing a very low capacity decay of 0.042% per cycle. This work provides an efficient strategy for the synthesis of bimetal-functionalized carbon materials for high performance lithium sulfur battery.

Figure 1(a) shows the synthesis of G/FeCo-MOF and G/FeCo@N-C. G/FeCo@N-C was synthesized by simple solvothermal and high temperature annealing methods. First, the graphene coated G/FeCo-MOF nanopolyhedron structure was synthesized by combining Co²⁺, Fe²⁺, G, and C₈H₇NO₄ in DMF. Then, after high temperature annealing, organic molecules, Co²⁺ and Fe²⁺ in G/FeCo-MOF polyhedron, are transformed into the N-carbon layer and CoFe₂O₄/Co₃Fe₇ nanoparticles to form G/FeCo@N-C composite. Meanwhile, CoFe₂O₄/Co₃Fe₇ nanoparticles were in situ coated between the G layer and the N-carbon layer, which not only improved the conductivity of the material but also increased the adsorption site of the material.

Figure 1(b) shows the XRD patterns of G/FeCo@N-C. Two types of diffraction peaks appear in G/FeCo@N-C, in which the characteristic peaks at 18.3°, 30.1°, 35.4°, and 43.0° correspond to (111), (220), (311), and (400) crystal planes of CoFe₂O₄ (PDF#22-1086). The characteristic peaks at 44.7° and 65.1° are attributed to the (110) and (200) crystal planes of Co₃Fe₇ (PDF#48-1817). This result further shows that G/FeCo-MOF is transformed into CoFe₂O₄/Co₃Fe₇ nanoparticles after annealing and also confirms the results of HRTEM characterization. The morphology and structure of the prepared samples were characterized by a microscopic technique. In Fig. S1(a), a scanning electron microscope shows the G-coated polyhedron structure with G/FeCo-MOF. The G layer cannot be clearly seen in the scanning electron microscope, and the uniform average particle size of the polyhedron structure is about 300 nm. In the transmission electron microscope (TEM) in Fig. 1(c), the thin G layer coated on the surface of the polyhedron structure can be clearly observed. Figure S1(b) is an SEM image of G/FeCo@N-C. In the figure, it can be seen that the G-coated polyhedron structure of G/FeCo-MOF is transformed into nanoparticles and N-carbon layers after high-temperature annealing and pyrolysis, and the nanoparticles are embedded in the N-carbon layer and tightly wrapped by the G layer. The transmission electron microscope image of G/FeCo@N-C in Fig. 1(d) further illustrates that the nanoparticles are evenly coated by N-carbon layer and G layer. Meanwhile, the presence of N-carbon layer and G layer can not only improve the conductivity of the materials and increase the adsorption sites on the surface of the materials but also provide enough space for the loading of elemental sulfur. In the high-resolution transmission electron microscopy image of G/FeCo@N-C in Fig. 1(e), crystal lattice streaks of nanoparticles can be clearly seen and the nanoparticles are confirmed to be CoFe₂O₄/Co₃Fe₇. Lattice fringes with lattice spacing of 0.25 and 0.20 nm correspond to the (311) crystal planes of CoFe₂O₄ and (110) crystal planes of Co₃Fe₇, respectively. In addition, a clear heterogeneous interface between CoFe₂O₄ and Co₃Fe₇ can be found in the figure. These heterogeneous interfaces can be considered as active sites to enhance interfacial charge transfer and catalyze rapid LiPSs conversion. In addition, lattice fringes with lattice spacing of 0.34 nm correspond to the (002) crystal faces of G. Figure S2(a) is the STEM-HAADF image of G/FeCo@N-C. In the figure, it can be seen that the nanoparticles and the N-carbon layer and the G layer and the signal features correspond well with the nanoparticles that are evenly coated within the N-carbon layer and G layer. At the same time, the distribution of Fe, Co, and O elements in the element mapping (EDX) in Figs. S2(b)–S2(f) further indicates that the nanoparticles are CoFe₂O₄ and Co₃Fe₇, and the uniform distribution of C and N elements further confirms that CoFe₂O₄/Co₃Fe₇ nanoparticles are uniformly coated with N-carbon layer and G layer.

In order to analyze the transformation process of LiPSs in lithium–sulfur battery during charge and discharge, in situ Raman spectroscopy was carried out. Figure 3(a) shows the schematic diagram of in situ Raman test of Li–S batteries. Figure 3(b) shows the voltage–time curve of the G/FeCo@N-C-S composite electrode, and it is clearly observed that there are two voltage platforms in the discharge process, which corresponds to the CV curve results in Fig. 4(a). As shown in Fig. 3(c), the characteristic peaks of S₈ at 147.8, 215.3, and 472.2 cm⁻¹ can be clearly observed at the beginning of discharge (2.48 V).^15,16 With the discharge process, the characteristic peak of S₈ gradually disappears, and the characteristic peak of Li₂S₆ (395.6 cm⁻¹) begins to appear when the discharge reaches 2.23 V. When the discharge reaches 2.06 V, the characteristic peak of Li₂S₆ begins to weaken and is accompanied by the characteristic peak of Li₂S₄ (227.8 cm⁻¹).¹⁷ Raman characteristic peak of Li₂S₆ and Li₂S₄ can be detected during this period of discharge. This result indicates that Li₂S₆ and Li₂S₄ exist simultaneously during the second stage of the discharge process. Surprisingly, when the electrode was fully discharged to 1.70 V, the characteristic peak of Li₂S₄ disappeared completely, and the characteristic peak of Li₂S (451.9 cm⁻¹)¹⁸ appeared. Meanwhile, the characteristic peak of Li₂S₆ was relatively weakened. This result further indicated that there was a transformation process of Li₂S₆ and Li₂S₄ to Li₂S during the discharge. During the charging process, Li₂S gradually oxidizes to Li₂S₄ and Li₂S₆ completely disappears at 2.26 V, and Raman characteristic peaks of Li₂S₄ appear at the same time. With the progress of charging at 2.33 V, the Raman characteristic peak of Li₂S₄ gradually disappears, while the characteristic peak of S₈ begins to appear. At the end of charging (2.68 V), all Li₂S₄ was oxidized to S₈, and the characteristic peaks of S₈ became clear gradually. A small number of characteristic peaks of Li₂S₆ were also observed. Interestingly, the characteristic peak of Li₂S₆ can be observed in the in situ Raman peak during both charging and discharging, which means that the part of the active substance is not effectively used, which is also the reason for the gradual attenuation of battery capacity. To further determine that the attenuation of battery capacity is related to the fact that the part of the active substance (Li₂S₆) is not being used efficiently, we continued to test the in situ Raman battery for the 100th cycle and obtained the same results as the first cycle (Fig. S7). In addition, the peak of the local Raman spectrum at 280 cm⁻¹ band during the whole charging and discharging process is caused by the Raman characteristic peak of LiTFSI in the electrolyte.^16,18 The kinetics of polysulfide conversion was studied by using a S-free electrode and Li₂S₆ electrolyte to assemble a symmetrical cell. The CV curve of G/FeCo@N-C electrode shows an obvious and sharp symmetric redox peak, while the redox peak of G/FeCo-MOF electrode is less obvious and relatively wide [Fig. 3(d)]. In addition, the CV curve of the G/FeCo@N-C electrode showed a higher current response than that of the G/FeCo-MOF electrode, indicating that the LiPSs conversion on the surface of the G/FeCo@N-C electrode was faster and fuller. These results further indicate that the catalytic action of CoFe₂O₄/Co₃Fe₇ nanoparticles in G/FeCo@N-C can improve the redox kinetic performance of batteries. In order to observe the adsorption effect of G/FeCo-MOF and G/FeCo@N-C samples on polysulfide, adsorption experiments were carried out in Li₂S₆ solution (1.5 mmol L⁻¹). As shown in Fig. 3(e), the G/FeCo-MOF and G/FeCo@N-C samples have just been immersed in the Li₂S₆ solution, and the color in all glass bottles is similar to the blank Li₂S₆ solution. G/FeCo-MOF and G/FeCo@N-C samples were soaked for 6 h. G/FeCo@N-C samples were nearly colorless and G/FeCo-MOF samples were slightly yellow. This result directly proved that the chemical affinity of G/FeCo@N-C for Li₂S₆ was stronger than G/FeCo-MOF. Based on the above-mentioned results, the improved chemical properties of G/FeCo@N-C can be attributed to the following advantages [Fig. 3(f)]. The presence of N-carbon layers and G layers not only provide electron channels for rapid charge transfer and abundant internal space but also provide a large number of electrochemically active sites for the anchoring of polysulfide. The synergistic effect of CoFe₂O₄ and Co₃Fe₇ nanoparticles with good catalytic activity in G/FeCo@N-C can adsorb and promote the conversion of LiPSs at an accelerated rate.

In order to study the electrochemical properties of the materials, different properties of the materials were tested and analyzed in the range of 1.7–2.8 V. Figure 4(a) shows CV curves of G/FeCo@N-C-S and G/FeCo-MOF-S at a scanning rate of 0.1 mV s⁻¹. As can be seen from the figure, there are two reduction peaks around 2.3 and 2.05 V, which are associated with the reduction in S₈ to the soluble polysulfide Li₂S_n (4 ≤ n ≤ 8) and further reduction to the insoluble Li₂S₂/Li₂S. The oxidation peak near 2.35 V is associated with the oxidation of Li₂S/Li₂S₂ to long-chain Li₂S_n and eventually to S₈. G/FeCo@N-C-S (ΔE = 0.26 V) has a smaller redox potential gap than G/FeCo-MOF-S (ΔE = 0.34 V), which indicates that G/FeCo@N-C-S can reduce the polarization potential of redox reaction and accelerate the kinetics of redox reaction. Figure 4(b) shows the constant current charge–discharge curves of G/FeCo@N-C-S and G/FeCo-MOF-S at 0.2 C. Both of these electrode materials exhibit two discharge platforms during the discharge process. The first discharge platform (∼2.3 V) is associated with the transformation of sulfur (S₈) into soluble polysulfide (Li₂S_n, 4 ≤ n ≤ 8).¹⁹ The second discharge platform (∼2.1 V) corresponds to the transformation of polysulfide into insoluble Li₂S₂ and Li₂S by the electrochemical reduction process, which corresponds to the CV curve.²⁰ Moreover, the initial specific capacity of the G/FeCo@N-C-S electrode is 970 mAh g⁻¹, which is higher than that of the G/FeCo-MOF-S electrode at 660 mAh g⁻¹. G/FeCo@N-C-S has the smallest polarization (ΔE = 0.21 V) compared with G/FeCo-MOF-S (ΔE = 0.27 V), which further proves the improvement of redox reaction. In addition, the capacity ratios of G/FeCo@N-C-S and G/FeCo-MOF-S discharge platforms (ΔQ₂/ΔQ₁) are 2.06 and 1.99, respectively. Figure 4(c) shows the cycling performance of the G/FeCo@N-C-S and G/FeCo-MOF-S electrodes at 0.2 C. The specific discharge capacity of the G/FeCo@N-C-S electrode after 200 cycles is 650 mAh g⁻¹, the capacity retention rate is 68.4%, and the coulomb efficiency is 98.7%. The capacity of G/FeCo-MOF-S electrode was reduced from 650 to 340 mAh g⁻¹, and the capacity retention rate was only 52.3% after 200 cycles, and the coulomb efficiency was 97.5%. Compared with G/FeCo@N-C-S, the capacity of G/FeCo-MOF-S decays rapidly, which further indicates that G/FeCo@N-C-S has a higher sulfur utilization rate. In addition, the G/FeCo@N-C-S electrode also shows an excellent rate performance, as shown in Fig. 4(d). Compared with G/FeCo-MOF-S, the specific discharge capacity of G/FeCo@N-C-S electrode at different current rates is significantly increased. When the current density of G/FeCo@N-C-S electrode increases from 0.1 to 0.2, 0.5, 1.0, and 2.0 C successively, the corresponding specific discharge capacities are 1017, 900, 810, 705, and 550 mAh g⁻¹, respectively. Impressively, the high specific discharge capacity of 900 mAh g⁻¹ can still be recovered when the current density is switched back to 0.1 C, indicating that G/FeCo@N-C-S has excellent electrochemical reversibility. Figure 4(f) shows the cyclic stability of G/FeCo@N-C-S and G/FeCo-MOF-S at 0.5 C. After 500 cycles, G/FeCo@N-C-S maintains a high specific discharge capacity of 465 mAh g⁻¹, which is much higher than that of G/FeCo-MOF-S (200 mAh g⁻¹). Figure 4(e) shows the Nyquist impedance diagram of the G/FeCo@N-C-S and G/FeCo-MOF-S (the corresponding equivalent circuit diagram is shown in the illustration). The high frequency region of the Nyquist impedance diagram presents a semicircle and the low frequency region is a slanted line representing the charge transfer resistance (R_ct) and the lithium ion diffusion impedance, respectively. The results show that G/FeCo@N-C-S (R_ct = 47.6 Ω) has lower charge transfer resistance than G/FeCo-MOF-S (R_ct = 120.2 Ω), indicating that the presence of N-carbon layer and G layer can accelerate the charge transfer and improve the conductivity of the material. The results show that G/FeCo@N-C-S has better electrochemical kinetic properties. In addition, the long-term cycling performance of G/FeCo@N-C-S was tested at 1.0 C [Fig. 4(g)]. The G/FeCo@N-C-S electrode after 1000 cycles can still maintain 440.3 mAh g⁻¹ discharge specific capacity, and each cycle capacity decay rate is only 0.042%, showing excellent cycle stability. The excellent electrochemical performance above the G/FeCo@N-C-S electrode can be attributed to the good electronic conductivity of the N-carbon layer and the G layer. Second, the abundant space and adsorption sites provide conditions for the storage and conversion of polysulfide. At the same time, the presence of CoFe₂O₄ and Co₃Fe₇ nanoparticles increased the catalytic activity of LiPSs and accelerated the transformation of LiPSs. In addition, we summarized and compared the electrochemical performance of Li–S batteries using different cathode materials (Table S1). Compared with previous studies, G/FeCo@N-C-S has better catalytic ability, and the electrode shows relatively good cyclic performance and stability.

In summary, the structure of double G/N-carbon layer-coated CoFe₂O₄/Co₃Fe₇ nanoparticles (G/FeCo@N-C) was synthesized by a convenient method of high temperature pyrolytic G coating MOFs. G/FeCo@N-C has high specific surface area, abundant active sites provided by nitrogen doping, and unique two-dimensional conductive network structure of G layer is conducive to ion conduction. In addition, the catalytic synergistic effect of CoFe₂O₄ and Co₃Fe₇ nanoparticles can effectively adsorb and catalyze LiPSs, thus accelerating the transformation of LiPSs and improving the redox kinetics of the sulfur cathode. The results show that the G/FeCo@N-C-S cathode has excellent electrochemical performance. The specific discharge capacity remained constant at 440.3 mAh g⁻¹ after 1000 cycles with a current density of 1.0 C, and the decay rate was only 0.042% per cycle. Additionally, even when using a current density of 2.0 C, the specific discharge capacity can still be maintained at 550 mAh g⁻¹. This work provides a convenient method for the preparation of bimetallic structures or bimetallic electrocatalysts and promotes the future research of cathode materials for Li–S batteries.