Ultrathin Two-Dimensional Fe–Co Bimetallic Oxide Nanosheets for Separator Modification of Lithium–Sulfur Batteries

The shuttle effect is understood to be the most significant issue that needs to be solved to improve the performance of lithium–sulfur batteries. In this study, ultrathin two-dimensional Fe–Co bimetallic oxide nanosheets were prepared using graphene as a template, which could rapidly catalyze the conversion of polysulfides and inhibit the shuttle effect. Additionally, such ultrathin nanostructures based on graphene provided sufficient active sites and fast diffusion pathways for lithium ions. Taking into account the aforementioned benefits, the ultrathin two-dimensional Fe–Co bimetallic oxide nanosheets modified separator assembled lithium–sulfur batteries delivered an incredible capacity of 1044.2 mAh g−1 at 1 C and retained an excellent reversible capacity of 859.4 mAh g−1 after 100 cycles. Even under high loading, it still achieved high area capacity and good cycle stability (92.6% capacity retention).


Introduction
Cost-effective lithium-ion batteries were once the linchpin of rechargeable batteries for the electronics, electric vehicles and other energy industries. However, conventional batteries based on graphite and lithium cobalt dioxide are no longer able to meet the growing energy demand [1]. The lithium-sulfur batteries with ultrahigh theoretical capacity have been created for the limitation of poor energy density of these conventional batteries [2]. Theoretically, lithium-sulfur batteries based on Li + S → Li 2 S redox reaction can obtain a high capacity of 1675 mAh g −1 and an amazing energy density of 2600 Wh kg −1 at stable discharge voltage [3]. Unfortunately, the shuttle effect caused by soluble polysulfides (Li 2 S n , 4 ≤ n ≤ 8) diffusion between the anode and cathode is a major factor in why highperformance lithium-sulfur batteries have not completely supplanted lithium-ion batteries as the market leader. This shuttle effect is a result of the particular charging and discharging mechanism of lithium-sulfur batteries. Chain polysulfides accumulate on the surface of the separator or electrode and mutate with fresh polysulfides, finally forming some insoluble substances (mainly Li 2 S/Li 2 S 2 ). This not only consumes the active material and reduces the capacity of the cell, but also blocks the active site and ion channel, resulting in slow reaction kinetics [4,5]. Therefore, how to suppress the shuttle effect is very important.
One of the most widely utilized strategies to address the aforementioned shuttle effect is to modify the design of functionalized separators [6,7]. For modified materials, carbon, polymers and inorganic substances are the most extensive choices [8][9][10][11]. Among them, carbon-based materials, such as activated carbon, carbon nanotubes and graphene, can speed up the charge transfer due to strong electrical conductivity. However, because of the weak interaction between carbon substrates and polysulfides, they perform slightly less Figure 1a shows the fabrication process of the ultrathin CoFe 2 O 4 nanosheets. Using graphene as a template, the Fe-Co precursors reacted with the oxygen-containing functional groups to form a sheet layer on the graphene surface. As-obtained dried precursors were calcined at 400 • C for 2 h in an atmosphere of air. The metal oxide was then directed to grow laterally in the two-dimensional plane to form a nanosheet shape during the subsequent calcination process [30][31][32]. The X-ray powder diffractometer (XRD) data of the substance created by the aforementioned procedure is shown in Figure 1b, and all diffraction peaks could exactly match the distinctive peaks of CoFe 2 O 4 (JCPDS No. 01-1121). The crystal planes (111), (220), (311), (400), (511) and (440) were represented by the diffraction peaks that were positioned at 18.1 • , 30.3 • , 35.7 • , 43.5 • , 57.2 • and 62.7 • , respectively. The nearabsence of graphene features meant that most of the graphene had been removed during high-temperature heat treatment. The very small amount of leftover graphene would somehow boost the conductivity of CoFe 2 O 4 . The substance was further examined using N 2 adsorption-desorption resolution at 77 K, and the result in Figure 1c demonstrated that the two-dimensional CoFe 2 O 4 nanosheets presented a significant Brunauer-Emmett -Teller (BET) surface area of 89.75 m 2 g −1 . This suggested that it possessed the high surface area of two-dimensional materials, which was advantageous for producing a lot of active sites for interacting with polysulfides. In addition, the inset reveals that the pore size distribution was around 5 nm, which would provide lithium ions quick transit pathways.

Results and Discussion
Molecules 2022, 27, x FOR PEER REVIEW 3 of active sites for interacting with polysulfides. In addition, the inset reveals that the size distribution was around 5 nm, which would provide lithium ions quick transit ways. The morphology of the object under transmission electron microscopy (TEM shown in Figure 1e. Obviously, a two-dimensional CoFe2O4 nanosheet was made u many ultra-fine nanoparticles with transverse sizes of about 7 nm. According to prev Reports, the oxygen-containing functional groups on the graphene surface initially pled with Fe 3+ and Co 2+ and then directed the metal oxides to self-assemble in the dimensional plane using graphene as a template, generating this ultrathin two-di sional structure [30]. Figure 1f shows the high-resolution transmission electron mi copy (HRTEM) image. The CoFe2O4 crystal presented a lattice spacing of 0.30 nm fo (220) crystal plane and 0.25 nm for the (311) crystal plane. From the energy dispe spectroscopy (EDS) mapping and scanning electron microscopy (SEM) results in Fig  1g and S1, it can be seen that the material revealed thin morphology and uniform ele distribution. The thicknesses of the nanosheets further measured by atomic force mi copy (AFM) were only around 1.35 nm (Figure 1d), which was consistent with TEM To comprehend the adsorption capability of CoFe2O4 on polysulfides more vi visual adsorption tests are crucial. As seen in Figure 2a, the original Li2S6 solution yellowish brown without adding any material. The same volume and concentratio Li2S6 solution were then added with equal amounts of carbon nanotubes and CoF respectively, and the results were monitored. Obviously, the solution containing ca nanotubes still remained yellowish brown after 12 h, as can be seen with the naked demonstrating the limited adsorption capacity of carbon nanotubes on polysulfides cording to previous studies, pure carbon has only weak adsorption between van Waals forces for polysulfides [33,34]. After adding CoFe2O4, the hue of the solution br ened after 6 h and became transparent after 12 h, demonstrating its potent ability to an soluble polysulfides [35,36]. A lithium polysulphide permeation experiment was fu conducted (Figure 2b), with pure electrolyte ether on one side of an H-shaped bottle The morphology of the object under transmission electron microscopy (TEM) is shown in Figure 1e. Obviously, a two-dimensional CoFe 2 O 4 nanosheet was made up of many ultrafine nanoparticles with transverse sizes of about 7 nm. According to previous Reports, the oxygen-containing functional groups on the graphene surface initially coupled with Fe 3+ and Co 2+ and then directed the metal oxides to self-assemble in the two-dimensional plane using graphene as a template, generating this ultrathin two-dimensional structure [30]. Figure 1f shows the high-resolution transmission electron microscopy (HRTEM) image. The CoFe 2 O 4 crystal presented a lattice spacing of 0.30 nm for the (220) crystal plane and 0.25 nm for the (311) crystal plane. From the energy dispersive spectroscopy (EDS) mapping and scanning electron microscopy (SEM) results in Figure 1g and Figure S1, it can be seen that the material revealed thin morphology and uniform element distribution. The thicknesses of the nanosheets further measured by atomic force microscopy (AFM) were only around 1.35 nm (Figure 1d), which was consistent with TEM.
To comprehend the adsorption capability of CoFe 2 O 4 on polysulfides more visibly, visual adsorption tests are crucial. As seen in Figure 2a, the original Li 2 S 6 solution was yellowish brown without adding any material. The same volume and concentration of Li 2 S 6 solution were then added with equal amounts of carbon nanotubes and CoFe 2 O 4 , respectively, and the results were monitored. Obviously, the solution containing carbon nanotubes still remained yellowish brown after 12 h, as can be seen with the naked eye, demonstrating the limited adsorption capacity of carbon nanotubes on polysulfides. According to previous studies, pure carbon has only weak adsorption between van der Waals forces for polysulfides [33,34]. After adding CoFe 2 O 4 , the hue of the solution brightened after 6 h and became transparent after 12 h, demonstrating its potent ability to anchor soluble polysulfides [35,36]. A lithium polysulphide permeation experiment was further conducted (Figure 2b), with pure electrolyte ether on one side of an H-shaped bottle and a specific concentration of Li 2 S 6 on the other. A significant amount of the yellow Li 2 S 6 infiltrated the other side with the middle barrier layer of pure Celgard (PP separator), indicating that the initial separator had no positive effect on the diffusion of polysulfides. In contrast, the color of the opposite solution was nearly unaltered in the system with the CoFe 2 O 4 -modified separator as the intermediate layer. Therefore, the latter was suitable for inhibition of the shuttle effect.
Molecules 2022, 27, x FOR PEER REVIEW 4 of 11 a specific concentration of Li2S6 on the other. A significant amount of the yellow Li2S6 infiltrated the other side with the middle barrier layer of pure Celgard (PP separator), indicating that the initial separator had no positive effect on the diffusion of polysulfides. In contrast, the color of the opposite solution was nearly unaltered in the system with the CoFe2O4-modified separator as the intermediate layer. Therefore, the latter was suitable for inhibition of the shuttle effect. The outcomes of X-ray photoelectron spectroscopy (XPS) experiments on CoFe2O4 both before and after adsorbing Li2S6 are depicted in Figure 2c,d. The XPS plot of Co 2p demonstrated that the overall curve swung toward the low binding energy after polysulfide adsorption, which might be caused by the creation of CoSX after the interaction of Co atoms with Li2S6. Due to Co acquiring electrons, the valence of Co element in CoSX was lower than that in the +3-valence state of CoFe2O4 [37][38][39][40]. For the negative shift of the XPS curve for Fe 2p orbital, the same cause was investigated. After adsorption of Li2S6, the intensity of the two peaks at ~722.7 eV at 2p1/2 and ~709.2 eV at 2p1/2 increased slightly, similar to previous reports, possibly due to the formation of a small amount of Fe-S [41,42].
Optical images of the separator before and after CoFe2O4 modification are displayed in Figure 3a, where the material is uniformly covered on the surface of the commercial Celgard separator. After folding and re-expanding the modified separator multiple times, the coating on the surface remained intact, demonstrating favorable mechanical stability. The thinnest coating thickness could be achieved without increasing the cell volume to fully increase the diffusion rate of lithium ions. As shown in Figure 3b, the cross-sectional SEM image reveals that the CoFe2O4 coating on the separator was roughly 10 μm thick, lower than many previous reports (Table S1), fully meeting the coating-thickness criteria of the battery [43][44][45][46][47][48][49][50]. Furthermore, the electrolyte and coating should have a high degree of affinity for electrolyte penetration and ion diffusion. The coating materials with low levels of affinity are ineffective. Therefore, the contact angle test with the lithium-sulfur electrolyte was performed ( Figure 3c). The CoFe2O4-modified separator outperformed the commercial separator in terms of contact angle. The contact angle size of the former was around 26°, whilst the latter was only about 15°. This indicated that the prepared coating The outcomes of X-ray photoelectron spectroscopy (XPS) experiments on CoFe 2 O 4 both before and after adsorbing Li 2 S 6 are depicted in Figure 2c,d. The XPS plot of Co 2p demonstrated that the overall curve swung toward the low binding energy after polysulfide adsorption, which might be caused by the creation of CoS X after the interaction of Co atoms with Li 2 S 6 . Due to Co acquiring electrons, the valence of Co element in CoS X was lower than that in the +3-valence state of CoFe 2 O 4 [37][38][39][40]. For the negative shift of the XPS curve for Fe 2p orbital, the same cause was investigated. After adsorption of Li 2 S 6 , the intensity of the two peaks at~722.7 eV at 2p 1/2 and~709.2 eV at 2p 1/2 increased slightly, similar to previous reports, possibly due to the formation of a small amount of Fe-S [41,42].
Optical images of the separator before and after CoFe 2 O 4 modification are displayed in Figure 3a, where the material is uniformly covered on the surface of the commercial Celgard separator. After folding and re-expanding the modified separator multiple times, the coating on the surface remained intact, demonstrating favorable mechanical stability. The thinnest coating thickness could be achieved without increasing the cell volume to fully increase the diffusion rate of lithium ions. As shown in Figure 3b, the cross-sectional SEM image reveals that the CoFe 2 O 4 coating on the separator was roughly 10 µm thick, lower than many previous reports (Table S1), fully meeting the coating-thickness criteria of the battery [43][44][45][46][47][48][49][50]. Furthermore, the electrolyte and coating should have a high degree of affinity for electrolyte penetration and ion diffusion. The coating materials with low levels of affinity are ineffective. Therefore, the contact angle test with the lithium-sulfur electrolyte was performed (Figure 3c). The CoFe 2 O 4 -modified separator outperformed the commercial separator in terms of contact angle. The contact angle size of the former was around 26 • , whilst the latter was only about 15 • . This indicated that the prepared coating not only did not hinder, but rather accelerated the diffusion of the electrolyte, which would be beneficial to the cell dynamics. not only did not hinder, but rather accelerated the diffusion of the electrolyte, which would be beneficial to the cell dynamics. In addition to the adsorption function, the coating is also important for the catalytic acceleration of the polysulfide conversion process. In order to further explore the effect of the CoFe2O4-modified layer on the reaction kinetics of lithium-sulfur batteries, a series of electrochemical behaviors were carried out. The cyclic voltammetry (CV) curves (0.5 mV s −1 ) of the pure separator and the CoFe2O4-modified separator were measured in symmetric cells. As shown in Figure 4a, the CV curve of the original separator almost did not have any peak, while the obvious redox peaks and higher current response appeared in the CoFe2O4-modified system, indicating that the latter provided a strong catalytic activity for polysulfides [10]. To more clearly demonstrate the superior catalytic and lithium-ion diffusion capabilities of the coating, a pure separator and a CoFe2O4-modified separator were assembled into lithium-sulfur batteries (with ~68% sulfur content in CNT@S cathode, Figure S2). The electrochemical impedance spectra (EIS) of the two separators and the equivalent circuit used to fit the obtained EIS spectra are displayed in Figures 4b and S3. The diffusion resistance decreases with the increasing slope of the diagonal line in the lowfrequency area, and Rct (charge transfer resistance) increases with increasing semicircle size in the medium-frequency zone [51]. As can be observed in the figure, the impedance profile of the CoFe2O4-modified separator was significantly lower than the other one in terms of diffusion resistance and Rct. In addition to the adsorption function, the coating is also important for the catalytic acceleration of the polysulfide conversion process. In order to further explore the effect of the CoFe 2 O 4 -modified layer on the reaction kinetics of lithium-sulfur batteries, a series of electrochemical behaviors were carried out. The cyclic voltammetry (CV) curves (0.5 mV s −1 ) of the pure separator and the CoFe 2 O 4 -modified separator were measured in symmetric cells. As shown in Figure 4a, the CV curve of the original separator almost did not have any peak, while the obvious redox peaks and higher current response appeared in the CoFe 2 O 4 -modified system, indicating that the latter provided a strong catalytic activity for polysulfides [10]. To more clearly demonstrate the superior catalytic and lithium-ion diffusion capabilities of the coating, a pure separator and a CoFe 2 O 4 -modified separator were assembled into lithium-sulfur batteries (with~68% sulfur content in CNT@S cathode, Figure S2). The electrochemical impedance spectra (EIS) of the two separators and the equivalent circuit used to fit the obtained EIS spectra are displayed in Figure 4b and Figure  S3. The diffusion resistance decreases with the increasing slope of the diagonal line in the low-frequency area, and R ct (charge transfer resistance) increases with increasing semicircle size in the medium-frequency zone [51]. As can be observed in the figure, the impedance profile of the CoFe 2 O 4 -modified separator was significantly lower than the other one in terms of diffusion resistance and R ct .  At a sweep rate of 0.1 mV s −1 , the CV tests were then conducted (Figure 4c). Both c showed two distinct reduction peaks, representing the conversion of S8 to soluble p sulfides (R2) and the subsequent formation of solid Li2S2/Li2S (R1), respectively. The br oxidation peak (O) corresponded to the opposite process [52]. It was obvious that changed separator had a lower peak spacing (ΔE) and a higher current density than unmodified one. The cell based on the CoFe2O4-modified separator exhibited an O-p and R1-peak spacing of 0.358 V, whereas the cell assembled with the standard PP separ had a peak spacing of 0.425 V. The obvious polarization of the latter indicated that reaction kinetics were slow. Additionally, to derive the matching Tafel slopes, the cur densities of the anode and cathode peaks in Figure 4c were independently fitted to overpotential. As shown in Figure 4d-f, the corresponding results of the modified se rator, respectively, were 65.1 (R1), 66.2 (R2), and 76.8 (O) mV dec −1 , lower than those of unmodified one (from 97.4 to 116.3 mV dec −1 ), which again indicated that CoFe2O4 excellent polysulfide catalytic conversion activity [53]. To evaluate the advantage of modified separator-assembled lithium-sulfur cell in terms of lithium-ion diffusion r CV curves at various sweep rates were performed in Figures 4g and S4. Linear fit achieved using the square root of the sweep rate and the redox peak current (Figure 4 The lithium-ion diffusion coefficient (D) is positively correlated with line slope (Ip/ according to the Randles-Sevcik equation: Ip = (2.69 × 10 5 ) n 1.5 SD 0.5 − Cν 0.5 , where Ip is peak current, n is the number of electrons transferred, S is the active electrode area, the lithium-ion concentration, and ν is the sweep rate [54]. Clearly, the line slopes of cell assembled by the modified separator were significantly higher than those of the o nary cell with pure Celgard separator, indicating that the CoFe2O4 coating promoted diffusion of lithium ions. From the above electrochemical experimental results,  Figure S4, and the triangle and pentagon correspond to the data points of the first reduction peak (R1) and the second reduction peak (R2), respectively.
At a sweep rate of 0.1 mV s −1 , the CV tests were then conducted (Figure 4c). Both cells showed two distinct reduction peaks, representing the conversion of S 8 to soluble polysulfides (R 2 ) and the subsequent formation of solid Li 2 S 2 /Li 2 S (R 1 ), respectively. The broad oxidation peak (O) corresponded to the opposite process [52]. It was obvious that the changed separator had a lower peak spacing (∆E) and a higher current density than the unmodified one. The cell based on the CoFe 2 O 4 -modified separator exhibited an Opeak and R 1 -peak spacing of 0.358 V, whereas the cell assembled with the standard PP separator had a peak spacing of 0.425 V. The obvious polarization of the latter indicated that the reaction kinetics were slow. Additionally, to derive the matching Tafel slopes, the current densities of the anode and cathode peaks in Figure 4c were independently fitted to the overpotential. As shown in Figure 4d-f, the corresponding results of the modified separator, respectively, were 65.1 (R 1 ), 66.2 (R 2 ), and 76.8 (O) mV dec −1 , lower than those of the unmodified one (from 97.4 to 116.3 mV dec −1 ), which again indicated that CoFe 2 O 4 has excellent polysulfide catalytic conversion activity [53]. To evaluate the advantage of the modified separator-assembled lithium-sulfur cell in terms of lithium-ion diffusion rate, CV curves at various sweep rates were performed in Figure 4g and Figure S4. Linear fit was achieved using the square root of the sweep rate and the redox peak current (Figure 4h,i). The lithium-ion diffusion coefficient (D) is positively correlated with line slope (I p /ν 0.5 ) according to the Randles-Sevcik equation: I p = (2.69 × 10 5 ) n 1.5 SD 0.5 − Cν 0.5 , where I p is the peak current, n is the number of electrons transferred, S is the active electrode area, C is the lithium-ion concentration, and ν is the sweep rate [54]. Clearly, the line slopes of the cell assembled by the modified separator were significantly higher than those of the ordinary cell with pure Celgard separator, indicating that the CoFe 2 O 4 coating promoted the diffusion of lithium ions. From the above electrochemical experimental results, the CoFe 2 O 4 -modified separator may have greatly enhanced the reaction kinetics of lithium-sulfur batteries from the standpoint of accelerating polysulfide conversion and ion diffusion rate. Figure 5a reveals the discharge-charge profiles at various rates (0.1-2 C). The cell with the CoFe 2 O 4 -modified separator exhibited steady charge-discharge plateaus, which highlighted the robust redox reaction kinetics. Consistent with the results of earlier CV curves (Figure 4c), the polarization voltage of the CoFe 2 O 4 -based electrode (0.19 V) was lower than that of the pure Celgard separator (Figure 5b). The rate performance of the two cells was captured and displayed in Figure 5c. It was clear that the battery with the Fe-Co bimetallic oxide nanosheets modified separator obtained the better reversible capacities of 1259.9, 1064.8, 895.7 and 720.7 mAh g −1 at 0.2, 0.5, 1 and 2 C, much higher than those of the ordinary one. Especially at the current density of 2 C, with the assistance of CoFe 2 O 4 , the capacity maintained the same degree of reduction, while the cell based on the pure Celgard separator rapidly decreased to only 158.2 mAh g −1 . During the fast charge-discharge process, the CoFe 2 O 4 based lithium-sulfur battery showed more effective lithium-ion diffusion and polysulfide conversion, which could reduce the influence of the polarization phenomenon and exert the capacity of the battery to some extent [27,36]. After returning to 0.2 C, the modified separator still achieved a high capacity of 1095.7 mAh g −1 , while the common separator was only 879.8 mAh g −1 , indicating that the former was more stable.
In order to evaluate the stability of the lithium-sulfur batteries, constant-current charge-discharge cycling tests were carried out at different rates before and after modification (activated for two cycles at 0.2 C). As shown in Figure 5d, the capacity of pure Celgard fluctuated greatly before 60 cycles at 0.5 C, which might be because the shuttle effect makes polysulfide unable to complete the charge-discharge behavior well. For the CoFe 2 O 4 -modified separator, its capacity was stable during the cycle, with an initial capacity of 1072.5 mAh g −1 , which remained at 835.2 mAh g −1 after 100 cycles. Figure 5e shows the cycle comparison at 1 C. The battery based on CoFe 2 O 4 nanosheets maintained excellent catalytic and stability performance, with an impressive initial capacity of 1044.2 mAh g −1 and a capacity retention rate of 82.3% after 100 cycles, while the contrast sample dropped rapidly from 832.5 mAh g −1 to 461 mAh g −1 . In order to further verify the service life of the modified lithium-sulfur battery, 500 charge-discharge cycles were carried out at 2 C. From Figure 5g, the improved cell maintained high capacity and low capacity decay during the first 200 cycles of the long cycle. Even at the 500th cycle, it still obtained a high capacity of 343.3 mAh g −1 , which was a significant advantage over the pure PP separator cell.
The loading of a cathode active substance in the above assembled lithium-sulfur batteries was approximately 1 mg cm −2 . In fact, in order to obtain a practical capacity, the loading often needs to be higher. Figure 5f shows the cycling performance of a CoFe 2 O 4based cell with a high sulfur loading of 3.2 mg cm −2 . A considerable initial capacity of 924.9 mAh g −1 (2.96 mAh cm −2 ) could be assigned to the device at 0.2 C. After 50 cyclings, the outstanding stability resulted in a consistent capacity of 856.7 mAh g −1 . Taken together, with the help of the ultrathin Fe-Co bimetallic oxide nanosheets modified layer, the lithiumsulfur batteries exhibited considerable rate, stability and high loading properties. charge process, the CoFe2O4 based lithium-sulfur battery showed more effective lithiumion diffusion and polysulfide conversion, which could reduce the influence of the polarization phenomenon and exert the capacity of the battery to some extent [27,36]. After returning to 0.2 C, the modified separator still achieved a high capacity of 1095.7 mAh g −1 , while the common separator was only 879.8 mAh g −1 , indicating that the former was more stable. In order to evaluate the stability of the lithium-sulfur batteries, constant-current charge-discharge cycling tests were carried out at different rates before and after modification (activated for two cycles at 0.2 C). As shown in Figure 5d, the capacity of pure Celgard fluctuated greatly before 60 cycles at 0.5 C, which might be because the shuttle effect makes polysulfide unable to complete the charge-discharge behavior well. For the in the lower right corner means that both the CoFe 2 O 4 -modified separator and the PP separator assembled battery are tested for cycle stability at 2C. The two beating data points in the upper right corner are normal phenomena that occur during testing. The lithiumsulfur battery assembled without a modified separator has undergone hundreds of charge and discharge processes inside the battery after a long cycle. The consequences of the shuttle effect in this process will lead to unstable charging and discharging, so the data points corresponding to the Coulombic effect fluctuate.

Preparation of CoFe 2 O 4 and Modified Separator
In the synthesis of ultrathin two-dimensional CoFe 2 O 4 nanosheets, 120 mg of graphene oxide was first uniformly dispersed into 75 mL of ethylene glycol and sonicated to obtain solution A. Solution B was obtained by dispersing 1 mmol of Co (NO 3 ) 2 I6H 2 O and 2 mmol of Fe (NO 3 ) 2 I9H 2 O into 25 mL of ethylene glycol. The above two solutions were mixed together and stirred at room temperature for 1.5 h. Then, the resulting suspension was condensed and refluxed at 170 • C while stirring for 2 h. After natural cooling, the solids were separated by centrifugation, washed with deionized water and absolute ethanol, and vacuum-dried overnight at 80 • C. Finally, the dried powder was calcined in air at 400 • C for 2 h using a heating rate of 0.5 • C min −1 to obtain the final product.
For the CoFe 2 O 4 -modified separator. A slurry consisting of 70% CoFe 2 O 4 , 20% carbon nanotubes and 10% polyvinylidene fluoride (PVDF) binder was coated on one side of a commercial Celgard-2400 separator and dried under vacuum at 60 • C.

Materials Characterization
The morphology and microstructure were observed by SEM (Hitachi Regulus 8100, operated at 5 kV) and TEM (FEI Tecnai G 2 20). The elemental distribution was collected by EDS mapping through SEM at an accelerating voltage of 15 kV. The crystal structure and phases were determined by XRD (Rigaku Smart Lab) at a wavelength of 1.5418 Å with copper Kα radiation. The specific surface area and pore size distribution were analyzed at a particulate ASAP2460 analyzer with BET calculation. The wettability of the electrolyte was measured by a contact angle (Theta) test. Elemental analysis was performed using XPS (Thermo Scientific K-Alpha). Thickness of material was measured by AFM (Bruker Dimension Icon).

Visualization Lithium Polysulfide Adsorption
Briefly, Li 2 S and sulfur power (molar ratio, 1:5) were added to a mixed solution of 1,3-dioxolane (DOL) and 1,2-dimethoxymethane (DME) in the volume ratio of 1:1 at 60 • C with stirring to obtain a Li 2 S 6 solution of 0.027 M. The same mass of CoFe 2 O 4 nanosheets and carbon nanotubes were added to a certain amount of the above solution, respectively.

Battery Assembly and Electrochemical Measurements
The compound sulfur cathode was obtained by melting sublimated sulfur and carbon nanotubes with a mass ratio of 7:3 at 155 • C ( Figure S2). A certain mass of as-obtained sulfur composite material (80 wt%), acetylene black (10 wt%), PVDF (10 wt%) and Nmethyl-pyrrolidone (NMP) was taken and mixed thoroughly and coated on Al foil. The cathode with a loading of about 1-3.2 mg cm −2 was obtained by vacuum drying (1 mg cm −2 for normal tests and 3.2 mg cm −2 for high loading tests). Metallic lithium was used as the anode. The DOL/DME (v/v, 1:1) mixture of dissolved 0.5 M LiTFSI and LiNO 3 was the electrolyte. A commercial Celgard-2400 or CoFe 2 O 4 modified Celgard-2400 was the separator, and the above parts were assembled as a coin cell (CR-2032).
For symmetric cell, carbon nanotubes or CoFe 2 O 4 mixed with acetylene black, PVDF binder, and NMP were coated on Al foil and dried as electrode; 0.5 M Li 2 S 6 (DOL/DME mixed LiTFSI solution) as the electrolyte, commercial Celgard-2400 as the separator and the above parts were assembled into a coin cell.
Constant-current charge-discharge tests were performed on a LAND system (CT-2001) with a voltage range of 1.7-2.8 V. The CV and EIS (10 mHz-100 kHz) were tested on an electrochemical workstation (CHI 660E).

Conclusions
In summary, ultrathin two-dimensional CoFe 2 O 4 nanosheets were synthesized using graphene as template by a simple two-step method. The superb ability of this Fe-Co bimetallic oxide adsorption to capture polysulfides and the ultrathin two-dimensional lamellar structure provided abundant active sites for the redox reaction and fast diffusion channels for lithium ions. These variables promoted the polysulfide catalytic conversion and lithium-ion diffusion, giving lithium-sulfur batteries a high initial capacity and a relatively slower capacity degradation rate. As a result, the lithium-sulfur battery with the CoFe 2 O 4 -modified separator showed an amazing capacity of 1259.9 mAh g -1 at 0.2 C and a high 82.3% capacity retention rate after 100 cycles at 1 C. The bimetallic oxide nanosheets created and manufactured in this study provided a new approach to successfully circumvent the low utilization of metal oxides and increased the potential applications of two-dimensional materials in the field of lithium-sulfur batteries.