High-Density Oxygen Doping of Conductive Metal Sulfides for Better Polysulfide Trapping and Li2 S-S8 Redox Kinetics in High Areal Capacity Lithium-Sulfur Batteries.

Exploring new materials and methods to achieve high utilization of sulfur with lean electrolyte is still a common concern in lithium-sulfur batteries. Here, high-density oxygen doping chemistry is introduced for making highly conducting, chemically stable sulfides with a much higher affinity to lithium polysulfides. It is found that doping large amounts of oxygen into NiCo2 S4 is feasible and can make it outperform the pristine oxides and natively oxidized sulfides. Taking the advantages of high conductivity, chemical stability, the introduced large Li-O interactions, and activated Co (Ni) facets for catalyzing Sn 2- , the NiCo2 (O-S)4 is able to accelerate the Li2 S-S8 redox kinetics. Specifically, lithium-sulfur batteries using free-standing NiCo2 (O-S)4 paper and interlayer exhibit the highest capacity of 8.68 mAh cm-2 at 1.0 mA cm-2 even with a sulfur loading of 8.75 mg cm-2 and lean electrolyte of 3.8 µL g-1 . The high-density oxygen doping chemistry can be also applied to other metal compounds, suggesting a potential way for developing more powerful catalysts towards high performance of Li-S batteries.

Introduction

The rechargeable lithium‐sulfur battery has attracted wide attention because of its high theoretical energy density (2600 Wh kg−1), low cost, and natural abundance of sulfur.[ ] However, achieving high energy density and long cycle life in realistic lithium‐sulfur batteries is still challenging. One of the major causes of the short cycling problem can be attributed to the well‐known “shuttle effect” of the soluble lithium polysulfide (LiPS) intermediates, which results in unexpected reactions with the lithium anode, leading to continuous sulfur loss, electrolyte consumption, and lithium anode degradation. Additionally, the low energy density issue arises because of the sluggish redox kinetics of the insulating S8 and its discharge products (Li2S2/Li2S), which result in low sulfur utilization especially in practical conditions with high areal capacity (>3 mAh cm–2) and lean electrolyte (electrolyte/sulfur (E/S) ratio <5 µL mg–1).[ ] So far, researchers have achieved tremendous success in tackling the “shuttle effect” and prolonging the cycling life of Li–S cells. Among these works, more and more researchers have confirmed that the “shuttle effect” can be greatly alleviated if the lithium anode is well protected by some artificial solid electrolyte interphase (SEI), interlayer, or sufficient coating.[ ] In this sense, if long‐term cycling of Li–S cells can be greatly improved through lithium anode protection, then more efforts should be focused on enhancing the S8‐Li2S redox kinetics to achieve high utilization of sulfur and therefore high energy density and high areal capacity Li–S batteries.[ ]

Using highly conductive and polar transition metal compounds to adsorb LiPS and accelerate its conversion to S8 and Li2S has been considered as one of the most popular and effective ways to improve sulfur utilization in high areal capacity sulfur cathodes.[ ] Plenty of metal oxides,[ ] sulfides,[ ] phosphides, and nitrides have shown unique ability for catalyzing S8‐Li2S redox kinetics based on their high electronic conductivity, as well as both polar‐polar Li‐X (X = O, S, N, P) interactions and Lewis acid‐base bonding (Metal–S) with LiPS.[ ] Superior performance in low areal capacity Li–S cells has been achieved. However, as the areal capacity increases higher and higher, the sulfur utilization drops, and in most works, it is still challenging to push it to above 60% sulfur utilization (1000 mAh/g) with lean electrolyte (E/S < 6 µL mg−1) and areal capacity above 3 mAh cm–2.[ ] To achieve 3 mAh cm–2 and 300 Wh kg−1 Li–S batteries, it is critical to attain more than 60% sulfur utilization with lean electrolyte.[ ] With this concern, the current materials used to catalyze the S8–Li2S redox are still insufficiently effective; thus, exploring new materials and methods to achieve much higher utilization of sulfur with lean electrolyte is still needed.

The previous work has identified that most transition metal oxides are less conducting but more chemically stable and can interact with LiPS via much stronger Li–O binding compared to the metal sulfide (or nitrides, phosphide) analogs, which tend to be more conducting but chemically unstable and display weaker interactions (the Li–S(N,P) binding is less strong than Li–O) with LiPS.[ ] Thus, there could a possibility to combine the advantages of the oxide and sulfide (or phosphide) compounds through materials with mixed anions to enhance the catalyzing effect on S8‐Li2S redox kinetics. To fully realize the potential and benefit of the Li–S couple, it is necessary to study the doping chemistry of special compounds such as like metal oxide‐sulfides, oxide‐phosphides, oxide‐nitrides, or even phosphide‐sulfides, some of which are known but many that are yet undiscovered, which could serve as a large system of electrochemical catalysts for high‐performance Li–S batteries.[ ] Among these compounds containing different polyanions, oxygen‐doped (O‐doped) materials can be especially promising for improving the affinity of LiPS owing to strong Li–O binding. Recent works also show a few potential examples of O‐doped compounds (O‐doped Sb2S3,[ ] O‐doped VN[ ] and natively oxidized CoP,[ ] etc.) for improving the affinity of LiPS in Li–S batteries. However, yet discovered O‐doped compounds can be less conducting, and many methods that yield low‐density of O‐doping. Some compounds are natively oxidizing in the air that cannot be precisely controlled to achieve a higher oxygen content.[ ] It is also necessary to determine if high‐density O‐doped compounds are better at catalyzing the Li–S couples than the natively oxidized compounds. These concerns motivate us to explore a new oxygen doping method to prepare compounds with a controllable, high‐density of O‐doping, which will lead to more possibilities for developing powerful catalyst candidates for lithium‐sulfur batteries. Moreover, it is also important to obtain more understanding of the effect of O‐doping chemistry on catalyzing S8–Li2S redox reactions.

In this work, we for the first time demonstrate the preparation of high‐density O‐doped, highly conductive bimetal sulfide and study its catalyzing performance on accelerating S8–Li2S redox kinetics. We find that by controlling the degree of sulfidation, it is easy to convert spinel NiCo2O4 to cubic NiCo2S4 but also possible to make a high‐density oxygen‐doped NiCo2S4 which we call NiCo2(O–S)4. The NiCo2(O–S)4 morphology is flower‐like, similar to that of NiCo2O4 and NiCo2S4, but the crystalline structure is similar to that of cubic NiCo2S4. When NiCo2S4 is natively oxidized in air, it shows a lower density of O‐doping. However, the O and S atom ratio in NiCo2(O–S)4 can be above 1.5:1 and it can still be as electronically conducting as NiCo2S4. Taking the advantages of its high conductivity, chemical stability, strong interactions with LiPS via the polar‐polar interaction of Li–O–Co/Ni species and Co/Ni‐S bonding, the NiCo2(O–S)4 catalyst outperforms NiCo2O4 and natively oxidized NiCo2S4 in sulfur cathodes for high areal capacity (>3 mAh cm–2) Li–S batteries even upon long‐term cycling. Moreover, by utilizing the high‐density oxygen‐doped NiCo2(O–S)4 as an interlayer (i.e., separator coating), we achieve a high areal capacity of 8.68 mAh cm–2 at 1.0 mA cm–2 even with a sulfur loading of 8.75 mg cm–2 and lean electrolyte of 3.8 µL g–1. We also apply the high‐density O‐doping chemistry to other high conductivity and high LiPS affinity transition metal sulfides and phosphides to demonstrate the generalizability of the approach. Our work provides a potential way for preparing stable and improved electrochemical catalysts toward high sulfur utilization and high areal density sulfur cathodes.

Result and Discussion

Schematic and Characterization of Doping Chemistry Among NiCo2O4, NiCo2(O–S)4 and NiCo2S4

To study the high‐density O‐doping chemistry, NiCo2O4 was first synthesized, which can be easily converted to cubic NiCo2S4 using a simple hydrothermal sulfidation method as frequently reported before.[ ] By controlling the time and the quantity of the sulfidation reactant, we investigated the feasibility of preparing high‐density O‐doped NiCo2S4. Interestingly, when the amount of sulfidation reactant was insufficient (0.02 M Na2S, mole ratio of 1:2 for NiCo2O4: Na2S), the original flower‐like morphology of the NiCo2O4 appeared to have collapsed and followed be reforming of the morphology as the sulfidation time increased, which was also accompanied by a phase transformation as shown in Figure S1a–c and Figure S2a, Supporting Information. When the sulfidation time was sufficient (8 h) and the reactant quantity was increased (0.015 M, 0.02 M,0.04M, 0.2 M; mole ratio of NiCo2O4: Na2S ≥ 1:1.5), the NiCo2O4 also underwent a phase transformation (Figure S2b, Supporting Information). Then, when the mole concentration of Na2S was controlled to be 0.02 M (mole ratio of NiCo2O4: Na2S = 1:2, corresponding to targeted atom ratio of O:S in O‐doped NiCo2S4 can be 1:1) and 0.04 M (corresponding to targeted atom ratio of O:S in O‐doped NiCo2S4 can be 1:2), we were able to obtain high‐density O‐doped NiCo2S4 samples with flower‐like morphology assembled by 1D nanofibers, similar to that in NiCo2O4 and NiCo2S4, but with the same crystal structure as cubic NiCo2S4 as illustrated in Figure S1d,e,f, and Figure S2b, Supporting Information. Figure shows the schematic of the sulfidation process from NiCo2O4 to NiCo2S4. As the sulfidation amount and concentration are carefully controlled, the high‐density O‐doped NiCo2S4 samples, marked as NiCo2(O–S)4 (with 0.02 M Na2S, Sample 1) and S2‐NiCo2(O–S)4 (with 0.04 M Na2S, Sample 2), can be prepared as mentioned above and characterized as following. The scanning electron microscopy (SEM) images in Figure 1b‐d, Figure S1a,d–i, Supporting Information, and TEM image in Figure 1f also indicate that these materials share a similar morphology with a 3D flower‐like structure. It should be emphasized that similar morphologies can enable a better investigation of the effect of composition on the electrochemical performance. Figure 1e and Figure S2b, Supporting Information show the X‐ray diffraction (XRD) patterns of these materials, which indicates that the structure of the as‐synthesized NiCo2O4 and NiCo2S4 particles are in good agreement with spinel NiCo2O4 and cubic NiCo2S4, respectively, while NiCo2(O‐S)4 clearly contains the main NiCo2S4 phase but with lower intensity reflections than those in S2‐NiCo2(O–S)4 and NiCo2S4, which somehow is owing to its transition‐state towards pure NiCo2S4.[ ] It is also verified by Figure S3, Supporting Information, that the synthesized NiCo2(O‐S)4 and S2‐NiCo2(O–S)4 are O‐doped compounds but not simple mixtures of separated phases of NiCo2O4 and NiCo2S4. High‐resolution transmission electron microscopy (HRTEM) images (Figure 1g and Figure S4, Supporting Information) of the NiCo2(O–S)4 reveal distorted lattice fringes with spacings corresponding to (311) planes slightly larger than the theoretical spacings for NiCo2S4, while the (400), (511) and (440) spacings are slightly lower (compared in Table S1, Supporting Information). These differences in lattice spacings in NiCo2O4, NiCo2(O–S)4 and NiCo2S4 could be due to the introduction of oxygen and sulfur vacancies, which may cause defects and rearrangement of the atoms in the lattices.[ ] Furthermore, energy‐dispersive spectroscopy (EDS) and electron energy loss spectroscopy (EELS) were applied for verifying the high‐density O‐doped states of NiCo2(O–S)4 and S2‐NiCo2(O–S)4. Figure 1h shows that O, S, Ni, and Co are uniformly distributed with a particular atomic percentage of 14.72% Ni, 26.10% Co, 21.90% S, and 37.28% O, which is roughly the same as the theoretical ratio of 1(Ni):2(Co):4(O+S). Additional EELS mapping and spectra (Figure S5 and Figure S6, Supporting Information) also confirm the uniform and high‐density O‐doped state of the NiCo2(O–S)4 surface, indicating the O and S atom ratio can be above 1.5:1. Specifically, Figure S7, Supporting Information shows NiCo2S4 is natively oxidized in the air with a low concentration of O‐doping, which is similar to natively oxidized sulfides and phosphides reported in Wang's work.[ ] The atomic percentage of oxygen in natively oxidized NiCo2S4 is quite low (7.60%, O:S = 1:7). However, by simply controlling the concentration of Na2S and sulfidation time, the O‐doping content can be increased to 15.6% for S2‐NiCo2(O‐S)4 (O:S = 1:3) and finally to 37.28% O (O:S = 1.7:1) for NiCo2(O‐S)4 as shown in Figure S8 and Figure S9, Supporting Information. In order to investigate the air stability of these samples with different amounts of O‐doping, XRD was carried out after exposing NiCo2O4, NiCo2(O–S)4, and natively oxidized NiCo2S4 powders to ambient air for one month. Surprisingly, the XRD patterns after the air exposure (Figure S10a, Supporting Information) show that NiCo2S4 completely degraded and no phase changes were observed for NiCo2(O‐S)4 and NiCo2O4. Further exposing S2‐NiCo2(O‐S)4 and NiCo2(O‐S)4 (Figure S10b, Supporting Information) for more days of air exposure, the XRD patterns and optical images of these samples show that S2‐NiCo2(O‐S)4 starts to be partially degraded after 40 days and totally degraded after 70 days, while NiCo2(O‐S)4 shows a lower degradation rate than S2‐NiCo2(O‐S)4, suggesting that high‐density O‐doping greatly increases the stability (mainly against oxidation) of the sulfides and can be a potential strategy for other sulfides or phosphides. Noting that NiCo2(O‐S)4 shows the highest amount of O‐doping (37.28%) and is more stable in the air during extended storage compared to S2‐NiCo2(O‐S)4 (15.60%), we focused on NiCo2(O‐S)4 for subsequent characterization and comparison. Based on the greatly increased oxygen‐content in high‐density O‐doped NiCo2(O‐S)4 (37.28%) compared to natively oxidized NiCo2S4 (less than 10.0%), we can elucidate the role of oxygen content on the properties of these compounds for catalyzing Li‐S redox reactions compared to the sulfur‐free NiCo2O4.

Figure 1

a) Schematic of sulfidation process from NiCo2O4 to NiCo2(O‐S)4 and NiCo2S4. b) Scanning electron microscopy (SEM) images of NiCo2O4 c), NiCo2S4 d) and NiCo2(O–S)4. e) X‐ray diffraction (XRD) patterns with reference patterns. f, g) Different views of transmission electron microscopy (TEM) and High‐resolution transmission electron microscopy (HRTEM) images of NiCo2(O‐S)4. h) SEM image of NiCo2(O–S)4 and corresponding EDS maps of oxygen, sulfur, cobalt, and nickel elements. XPS patterns of Co 2p (a), Ni 2p (b), S 2p (c), and O 1s (d) for NiCo2O4, NiCo2(O–S)4, and NiCo2S4, respectively.

X‐ray photoelectron spectroscopy (XPS) was also conducted to analyze the changes in chemical valence states of the NiCo2O4, NiCo2(O–S)4, and NiCo2S4 powders. The Co 2p and Ni 2p peaks in NiCo2(O–S)4 are similar to those in NiCo2S4, confirming that they share similar chemical states (Figure 1j,k). The S 2p spectra of NiCo2(O–S)4 as shown in Figure 1l also exhibit a lower intensity than NiCo2S4, indicating a lower sulfur content on NiCo2(O‐S)4 surface. In Figure 1m, the O1 (M–O bond, where M = metal) peak is slightly shifted to higher binding energies in NiCo2(O–S)4 compared to NiCo2O4 while the intensity is lower, which could be due to the weakened M–O bond after the substitution of oxygen atoms with sulfur. All the above data confirm that the partial sulfidation process of NiCo2O4 results in the co‐existence of sulfur and oxygen in the cubic NiCo2S4 structure and the high‐density O‐doped NiCo2S4 can be obtained as we call NiCo2(O–S)4. To further understand the effect of oxygen doping on the structure and property changes of pure NiCo2S4, we used density functional theory (DFT) to calculate the (311) surface structure as shown in Figure S11a,b, Supporting Information. The calculated total density of states (DOS) showed that both NiCo2S4 and one O‐atom doped NiCo2S4 (O‐NiCo2S4) exhibit metallic properties, which should be beneficial for electronic transport (Figure S11c,d, Supporting Information). Although there is no marked difference in the DOS of pristine NiCo2S4 and O‐NiCo2S4 (311), the d‐band center of O–NiCo2S4 (311) is −1.3955, −1.2858, and −1.3407 for the spin‐up, spin‐down and average value, respectively, slightly lower than that of pristine NiCo2S4(311) (−1.392, −1.278 and −1.335 for spin‐up, spin‐down and average value, respectively). The location of the d‐band center further under the Fermi level implies the potential ability for trapping small molecules on the O–NiCo2S4 (311) surface.

Conductivity and Chemical Interaction with LiPS of NiCo2O4, NiCo2(O–S)4 and NiCo2S4

We hypothesized that high‐density oxygen doping could make NiCo2(O–S)4 (similar to S2‐NiCo2(O–S)4) less conducting than natively oxidized NiCo2S4 but increase its affinity to LiPS owing to the introduction of more Li‐O interactions for trapping S 2–. To demonstrate this, the electronic conductivity measurements and LiPS adsorption tests were carried out with four‐point probe measurements and UV‐vis spectroscopy, respectively. As shown in Figure , the conductivities of the natively oxidized NiCo2S4 and NiCo2(O–S)4 sheets can reach 51.2 and 30.1 S cm–1, respectively, while NiCo2O4 only showed a conductivity of 0.3S cm–1. Composite papers containing multiwall carbon nanotubes (MWCNTs) and cellulose nanofiber (CNF) mixed with NiCo2O4, NiCo2(O‐S)4, and NiCo2S4 were also prepared for further investigation as LiPS cathodes as described in the experimental section. The conductivity order of these composite papers was NiCo2S4 (2.0 S cm–1), NiCo2(O–S)4 (1.8 S cm–1), and NiCo2O4 (1.3 S cm–1), which was just consistent with reasonable estimation.

Figure 2

a) Conductivities of NiCo2O4, NiCo2(O–S)4, NiCo2S4 sheets and supported MWCNT papers. b) UV‐vis spectra of 0.1mM Li2S6 solution before and after 1 h absorption. c) Optical images of NiCo2O4, NiCo2(O–S)4 and NiCo2S4 soaked in the 10 mM Li2S6 solution. d) High‐resolution XPS spectra of Co 2p for NiCo2(O–S)4 with and without Li2S6 catholyte adsorption. e) The ratio of M3+/M2+ 2p3/2 peak intensity (I p) and the rate of I p change.

To test the difference in LiPS interactions, adsorption tests were carried out by adding 15 mg of each of the materials into 2 mL vials containing 0.1 mM Li2S6‐DME/DOL solution. As shown in Figure 2b, the absorbance of the S4 6– peak drops to much lower intensities after the addition of the NiCo2(O–S)4 sample, indicating its adsorbed more LiPS (had stronger interactions with LiPS) than NiCo2O4 and NiCo2S4, which displayed similar S4 6– absorbance. Further time‐dependence of the adsorption tests were investigated; as shown by the photographs in Figure 2c and Figure S12, Supporting Information, NiCo2(O–S)4 and S2‐NiCo2(O–S)4 show the improved adsorption of LiPS compared to pristine NiCo2S4 and NiCo2O4. The chemical interactions between the LiPS and these adsorbents were further investigated via XPS. The spectra of NiCo2O4, NiCo2(O–S)4, and NiCo2S4 with and without Li2S6 retrieved from the adsorption test solutions were analyzed. Since all these samples exhibit Lewis acid‐base bonding (M–S) with LiPS, the peaks of Ni 2p and Co 2p all show a shift to lower binding energies after interacting with Li2S6 as shown in Figure 2d, Figure S13–S15, Supporting Information.[ , ] It is hard to identify and compare the shifts in the binding energies among these samples because they are quite similar and it seems that each one has a unique preferred interaction site with Li2S6. By comparing these fitted Co 2p and Ni 2p peaks, it is interesting to see obvious intensity changes for the fitted metal (M2+ and M3+, M = Co or Ni) peaks with and without Li2S6 interaction. For example, the intensities of the Co2+ peak in NiCo2(O–S)4 and NiCo2S4 decrease more than the Co3+ peak after interacting with Li2S6 (Figure 2d, Figure S14, Supporting Information), which suggests that the Co2+ site dominates the LiPS adsorption in the Co3+‐Co2+ couple for NiCo2(O‐S)4 and NiCo2S4. Meanwhile, the spectra show that the intensities of the M3+ peaks for NiCo2O4, NiCo2(O‐S)4 and NiCo2S4 decrease more obviously than M2+ after interacting with Li2S6, suggesting that M3+ sites may also dominate the LiPS adsorption depending on the sample and M2+‐ M3+ couple. Thus, we use the ratio of the M3+ to M2+ 2p3/2 peak intensity, which is marked as I p (M3+/ M2+) in Figure 2e, to clarify the preferred LiPS interaction sites. The columns in Figure 2e show that, for NiCo2(O‐S)4 and NiCo2S4, Ip (Co3+/ Co2+) increases while Ip (Ni3+/ Ni2+) decrease after interacting with Li2S6, suggesting that NiCo2(O–S)4 and NiCo2S4 share similar preferred interacting sites (Co2+ and Ni3+) with LiPS. On the other hand, the Co3+ and Ni3+ are the sites that interact with LiPS in NiCo2O4. The I p0 and I p1 represent the intensity of the fitted Co2p3/2 peaks without and with Li2S6 absorption, respectively, while I’p0 and I’p1 are associated with Ni2p3/2. Since NiCo2(O–S)4 and NiCo2S4 share similar preferred interacting sites (Co2+ and Ni3+) with LiPS, the rates of I p change (R(Co3+/Co2+) = (I p1 − I p0)/I p1; R(Ni3+/Ni2+) = (I’p0 − I’p1)/I’p0 were also calculated to roughly compare the dominated LiPS interaction sites of Co2+ and Ni3+. Specifically, NiCo2O4 is Co3+ and Ni3+ dominated, R(Co3+/Co2+) for NiCo2O4 here is calculated by (I p0 − I p1)/I p0 and R(Ni3+/Ni2+) is similarly calculated for NiCo2(O–S)4 and NiCo2S4. As shown in Figure 2e, the R(Co3+/Co2+) of NiCo2O4 and NiCo2(O–S)4 are much higher than NiCo2S4 while R(Ni3+/Ni2+) increases from 4.5, 6.2 to 8.9% as NiCo2O4 gradually changed to be NiCo2(O–S)4 and NiCo2S4. This indicates that the Co dominates the Li2S6 interaction site of the oxide and oxide‐sulfides, but Ni dominates the Li2S6 interaction site of the sulfide. NiCo2(O–S)4 shares the similar Co2+ preferred interaction site with NiCo2S4, but as high‐density O is introduced, the increased R(Co3+/Co2+) suggests that the Co2+ site of NiCo2(O–S)4 is greatly activated and enhanced without too much decrease in Ni2+ site, which can be an explanation for its better absorption properties compared to NiCo2S4 and NiCo2O4. Figure S16, Supporting Information, also shows the fitted O 1s spectra of NiCo2O4 and NiCo2(O–S)4 with and without adsorption of Li2S6. The fitted O1 peak centered near 529.5 eV represents the M–O bond of the samples. After interacting with Li2S6, the O1 peaks decrease in intensity in both samples, which can be attributed to the formation of Li–O–M species, suggesting that there is electron transfer from the LiPS to the O atoms. Using the same method, we can also obtain the R(O2/O1) data, in which NiCo2O4 shows R of 78.1% while NiCo2(O–S)4 shows 60.0%, indicating that our proposed Li–O interaction is greatly introduced to NiCo2(O‐S)4. It should be emphasized that the interactions between LiPS and the samples are complicated and we cannot rely completely on the R(M3+/M2+) or R(O2/O1) to quantify the differences in LiPS absorption ability difference, especially between NiCo2O4 and NiCo2(O–S)4 because they show different Li2S6 interaction sites (Co3+‐Ni3+ for NiCo2O4, Co2+‐Ni3+for NiCo2(O–S)4). However, we suspect that the Co2+ site can be more effective than Co3+ for absorbing LiPS, which can explain why NiCo2(O–S)4 shows enhanced LiPS absorption ability compared to NiCo2O4 during the adsorption and UV‐vis spectroscopy tests. Moreover, to simply understand the O‐doping effect on LiPS absorption for O‐NiCo2S4 and pure NiCo2S4, DFT is also adopted to simulate the adsorption energies (E ads) with Li2S (n = 1, 4, 6, 8) and charge density surrounding the adsorption sites. As shown and discussed in Figure S17, Supporting Information, the adsorption energies on the O‐NiCo2S4 (311) surface are more negative than those for the pure NiCo2S4 (311) surface and the charges are especially aggregated around the oxygen atoms, indicating enhanced interaction of lithium sulfides on the oxygen sites of O‐NiCo2S4 (311) surface. Bader charge analysis can quantitatively describe charge transfer between adsorbates and substrates. It shows in Table S2, Supporting Information that electrons are transferred from the Li in the LiPS to the S/O atoms, with more electrons transferred to O‐NiCo2S4 compared to NiCo2S4, which is consistent with the higher Pauli electronegativity of O compared to S. Interestingly, compared with the S atoms at the pure NiCo2S4 (311) surface, the substituting O atom gains more electrons (over 3 times). The excess electrons going to the O atoms not only come from Li2S , but also are donated by neighboring atoms. Hence, the charge density surrounding the adsorption sites is redistributed to modulate the adsorption behavior of Li‐S species, which may be beneficial for lithium–sulfur redox.

Redox and Deposition Kinetics of S8–Li2S Assisted with NiCo2O4, NiCo2(O–S)4 and NiCo2S4 Papers

The low conductivities of S8 and Li2S2/Li2S, and the solubility of LiPS in the electrolyte greatly limit the conversion kinetics and utilization of S8; thus, materials with both high electronic conductivity and LiPS affinity are more attractive for fabricating sulfur cathodes.[ ] High‐density O doped NiCo2(O‐S)4 shows more LiPS affinity than NiCo2O4 and NiCo2S4 and it is also as conducting as NiCo2S4. Thus, it can be a great candidate for catalyzing the S8‐Li2S redox reaction. To investigate this, free‐standing NiCo2O4, NiCo2(O–S)4 and NiCo2S4 papers were prepared and implemented as current collectors for the S8‐LiPS‐Li2S conversion reaction as described in the experimental section. The characteristic peaks of the MWCNTs, NiCo2O4, NiCo2(O‐S)4 and NiCo2S4 were clearly observed in the XRD diffraction patterns (Figure S18, Supporting Information), suggesting no significant phase change in our preparation process. To first investigate the catalytic ability of these samples, symmetric cells were assembled with two identical papers as electrodes in 40 µL 0.5 mol L–1 Li2S6 as described in the experimental section. Cyclic voltammetry (CV) tests were conducted in the voltage range −1.5 ‐ 1.5 V at 1 mV s–1 (Figure ), 5 mV s–1 (Figure 3b), and 10 mV s–1(Figure 3c). The capacitive current response and the small peak separation between the reduction and oxidation peaks both demonstrate the fast and reversible conversion of LiPSs.[ ] When comparing the capacitive current response from 1 to 10 mV s–1, it is obvious that the NiCo2(O–S)4 paper exhibits much higher current response than that of symmetrical cells with NiCo2O4 and NiCo2S4 papers. Figure 3a also shows the three distinct pairs of redox peaks for all samples, where the reduction peaks labelled a, b and c are attributed to the reduction of S8 to Li2S6, Li2S6 to Li2S4 and Li2S4 to Li2S2/Li2S, respectively. The oxidation peaks a’, b’ and c’ are associated with oxidation of Li2S2/Li2S to Li2S4, Li2S4 to Li2S6 and Li2S6 to S8. Note that the cell with the NiCo2(O‐S)4 paper exhibits a much smaller peak separation than NiCo2O4 and NiCo2S4 papers at all scan rates, confirming that NiCo2(O‐S)4 significantly enhances the kinetics of the lithiation/delithiation reactions for polysulfides conversion. Meanwhile, the smaller current response and bigger peak separation of NiCo2O4 paper should be ascribed to its high charge transfer barrier because of its low conductivity. The weaker interactions between NiCo2S4 with the LiPS contribute to the more sluggish kinetics for LiPS conversion compared to the NiCo2(O‐S)4 paper.

Figure 3

a) CV curves of the symmetric cells at scan rates of 1 mV s–1, b) 5 mV s–1 and c) 10 mV s–1 with NiCo2O4, NiCo2(O–S)4 and NiCo2S4 papers as the electrodes. The electrolyte was 0.5 M Li2S6 in DOL/DME (1:1 in volume). d) CV curves of the Li–S full cells at a scan rate of 0.05 mV s–1. e) Tafel plots were calculated from the CV curves for the low plateau reduction peak at around 2.0 V and f) the first oxidation peak at around 2.35 V. g) CV curves of NiCo2(O–S)4 paper supported Li–S full cells from scan rates of 0.05 to 0.25 mV s–1. Corresponding peak potential difference of EpI‐IV h) and EpII‐III i).

The different catalyzing effects of the materials can also be confirmed by the electrochemical impedance spectra (EIS) in Figure S19, which show that the symmetric cell with NiCo2(O–S)4 paper electrodes has a much smaller semicircle (associated with the charge transfer resistance, R ct) than those with NiCo2O4 or NiCo2S4 electrodes, indicating that NiCo2(O–S)4 is more effective for enhancing the charge transfer for LiPSs conversion. CV tests were used to further verify the ability of NiCo2(O–S)4 for enhancing the kinetics in Li–S cells. As shown in Figure 3d, the reduction peaks I and II represent the reduction of S8 to LiPS and LiPS to Li2S2/Li2S while the oxidation peaks of III and IV are associated with oxidation of Li2S2/Li2S to LiPS and LiPS to S8, respectively. The much higher potential of the reduction peaks and lower potential of oxidation peaks, which are clearly split, suggests that the NiCo2(O–S)4 paper promoted the conversion between S8, soluble LiPS, and Li2S2/Li2S.[ ] The catalytic effects of these three materials were also compared by calculating the Tafel plots of the reduction (I and II) and oxidation (III) peaks in Figure 3d. As shown in Figure 3e,f, and Figure S20, Supporting Information, for both the reduction and oxidation processes, the fitted slopes clearly follow the order of NiCo2(O–S)4 < NiCo2S4 < NiCo2O4, which suggests that the high‐density O‐doped NiCo2(O‐S)4 has a better catalytic activity for the fast reduction and oxidation of S8, LiPSs, and Li2S2/Li2S while the catalytic activity of NiCo2O4 is relatively sluggish because of its much lower conductivity. Further, CV curves of the Li–S cells with NiCo2(O–S)4, NiCo2O4, and NiCo2S4 papers were also recorded at various scan rates from 0.05‐0.25 mV s−1 as shown in Figure 3g and Figure S21–S22, Supporting Information. The peak potential differences of I–IV (EpI‐IV) and II–III (EpII‐III) redox couples were also used for tracking the changes upon going to higher scan rates. As shown in Figure 3h,i, we can clearly verify that the cell with the NiCo2(O–S)4 paper has lowest EpI–IV and EpII–III at all scan rates, and its slopes between two adjacent scan rates seems to gradually overlap with NiCo2S4, suggesting the more conducting NiCo2(O–S)4 can reduce the charge transfer barrier similarly as NiCo2S4. In contrast, NiCo2O4 suffers from its low conductivity and the cell shows the highest EpI–IV, EpII–III and much higher slopes between two scan rates. These results all demonstrate the enhanced polysulfide conversion kinetics in the entire charge/discharge processes for the NiCo2(O–S)4 assisted Li–S cell.

Apart from enhancing polysulfide redox kinetics as mentioned above, we supposed that NiCo2(O–S)4 would also show better catalyzing effects for Li2S nucleation because of its sufficient conductivity and high affinity to LiPS. To prove this, Li2S precipitation experiments were carried out by using the potentiostatic discharge method.[ ] According to our previous research and other reports,[ , ] when LiPS is discharging to form solid Li2S2/Li2S, driven by the so‐called “local concentration effect”, Li2S2/Li2S prefers to precipitate and gradually grow to form thick coatings on these “dual high” (high conductivity and high LiPS affinity) materials. Thus, the easier the Li2S2/Li2S precipitation and thicker the coating, the faster the deposition kinetics and higher sulfur utilization that can be achieved by the LiPS conversion catalysts such as the NiCo2(O–S)4, NiCo2O4 and NiCo2S4. Comparing the Li2S nucleation curves shown in Figure 4a,c, we can clearly see that the capacity of precipitated Li2S on NiCo2(O–S)4 paper is about 688 mAh gs –1, which is much higher than those on NiCo2S4 paper (490 mAh gs –1) and NiCo2O4 paper (372 mAh gs –1). Besides, the Li–S cell with NiCo2(O–S)4 paper exhibited the highest peak current while NiCo2S4 and NiCo2O4 share a similar peak current, suggesting faster LiPSs trapping and nucleation of Li2S on NiCo2(O–S)4 paper. The pristine electrodes and deposited morphology of Li2S in precipitation tests were also characterized by SEM as shown in Figure S23, Supporting Information, and Figure 4d–f, respectively. The pristine electrodes contain flower‐like particles which are physically mixed with MWCNT (Figure S23, Supporting Information). After constant potential depositing Li2S, Figure 4d shows a little bit of Li2S solid deposits on the 1D nanofibers of the NiCo2O4 particle; while for NiCo2S4 and NiCo2(O–S)4 (Figure 4e,f), thick Li2S coating is clearly deposited on the 1D nanofibers and the surface of NiCo2(O–S)4 (Figure 4f) particle is especially rougher than NiCo2S4, in which Li2S deposits almost wrapped the 1D nanofibers. These morphologies verify that with sufficient electronic conductivity (as high as NiCo2S4) and superior interaction with LiPS, NiCo2(O–S)4 is more powerful for accelerate Li2S nucleating and growing to be thicker, thus the high capacity of sulfur can be achieved. On the other hand, the oxidation morphology of deposited Li2S was further investigated by using a normal galvanostatic charging process as marked in Figure 4d–f. When the assembled cells are charged to 2.3V, we can see that the deposited Li2S are gradually dissolved and NiCo2O4, NiCo2S4 and NiCo2(O‐S)4 share its similar morphology. However, when they are charged to 2.8V, some solid participants can be detected on the surface of NiCo2S4 and NiCo2(O–S)4 while the surface of NiCo2O4 particles remains smooth and clean. Besides, the 1D nanofibers of NiCo2(O–S)4 particle becomes much rougher than those of NiCo2S4, indicating the easier dissolution of Li2S and deposition of S8 on NiCo2(O–S)4 surface than that on NiCo2O4 and NiCo2S4. Noting that these SEM images were taken just for 1 cycled electrode, much details for long cycled electrodes below would show more difference for the different materials. Anyway, these results clearly confirm that with the help of high‐density O doping, NiCo2(O‐S)4 facilitates the effective deposition of both Li2S and S8, suggesting to be a desired catalyst for improving sulfur utilization in high areal capacity (>3mAh cm–2) Li–S cells.

Figure 4

a) Li2S nucleation curves of NiCo2O4 paper, b) NiCo2S4 paper and c) NiCo2(O–S)4 paper. Corresponding morphologies electrodes after Li2S nucleation, dissolution to LiPS, and deposition of S8 on surfaces of d) NiCo2O4, e) NiCo2S4, and f) NiCo2(O–S)4.

Electrochemical Performance of Corresponding Li–S Cells

To compare the effect of the differences in catalyzing performance of these NiCo compounds in the cycling behavior of the Li–S cells, coin cells were assembled using lithium metal as anode and NiCo2O4, NiCo2S4 and NiCo2(O–S)4 papers loaded with Li2S8 solution as cathode. Different amounts of Li2S8 were loaded corresponding to sulfur loading from 2.5, 2.9, 3.3 to 4.4 and 8.75 mg cm–2 for >3 mAh cm–2 Li–S cells. Figure first shows the rate performance of the cells with a low sulfur loading of 2.5 mg cm–2 (40.0 wt.% in cathode), which confirms that the NiCo2(O–S)4 paper outperforms the other two at all rates. The corresponding discharge‐charge curves in Figure S24, Supporting Information further show that the NiCo2(O–S)4 supported cell especially has a higher initial capacity at the higher voltage plateau (from S8–Li2S4 redox), suggesting its better catalyzing effect for converting S8 and high‐chain LiPS compared to NiCo2S4 and NiCo2O4. Specifically, the cell with the NiCo2(O–S)4 paper can deliver a high capacity of about 1268 mAh g–1 at 0.5C, which is equal to a sulfur utilization of 75.8% and areal capacity of 3.2 mAh cm–2. The cycling performance of the NiCo2(O–S)4 paper supported cell at 1C is also shown in Figure S25, showing high sulfur utilization of 50.8% and a high areal capacity of 2.1 mAh cm–2 after 230 cycles. The improved LiPS–Li2S conversion by NiCo2(O–S)4 is also verified from the activating discharge and charge curves for even high sulfur loading (3.3 mg cm–2 and 4.4 mg cm–2) cells as shown in Figure 5b,c. As we applied a quite low current density (0.16 mA cm–2) for achieving full conversion of S8 or high‐chain LiPS to Li2S4 at the higher voltage plateau, the capacity of the lower plateau would demonstrate the catalyzing ability of different samples for Li2S4–Li2S conversion. Figure 5b exactly shows that though the three cells share a similar initial activated capacity of LiPS–Li2S4 (0.83 mAh cm–2), the cells with more conducting NiCo2(O–S)4 and NiCo2S4 clearly have higher capacities associated with the low voltage plateau than the NiCo2O4 paper. When cells with even higher sulfur loading were further activated at the 2nd cycle as shown in Figure 5c, the result confirms that NiCo2(O–S)4 still outperforms NiCo2S4 and NiCo2O4 for both S8‐Li2S4 and Li2S4–Li2S conversion. It should be noted that the papers comprising the Ni–Co compounds, as well as the small quantity of Li2S6 additives in the electrolyte can contribute extra capacities mainly during the activating cycles (Figure S26–S28, Supporting Information), thus the areal capacities of the first cycles in Figure 5b,h exceed the theoretical value. Long‐term cycling tests of these cells with a sulfur loading of 2.9 mg cm–2 (43.2 wt.% in cathode) were also carried out to check the stability of these electrodes. As shown in Figure 5d–f, the NiCo2(O–S)4 paper supported cell can realize a capacity of 962 mAh g–1 (2.8 mAh cm–2) at 0.2C after 200 cycles and 922 mAh g–1 (2.7 mAh cm–2) at 0.5C after 150 cycles, while NiCo2O4 paper shows lowest capacity and NiCo2S4 paper suffers rapid capacity fading in the longer cycles. The coulombic efficiency of the NiCo2S4 paper supported cells also decreases faster during long‐term cycling, suggesting its worse stability than NiCo2(O‐S)4 paper. EIS spectra after cycling at 0.2C (Figure 5e) further show the cell with NiCo2(O–S)4 paper has the smallest charge transfer resistance, which is attributed to the stable exposed catalytic surfaces during cycling.

Figure 5

a) Rate performance of different paper supported Li–S cells from 0.1C to 2C with sulfur loading of 2.5 mg cm–2. b) The first discharge curves of NiCo2O4, NiCo2S4 and NiCo2(O–S)4 paper supported Li2S8 catholyte with sulfur loading of 3.3 mg cm–2, c) and the corresponding 3rd discharge‐charge curves at 0.16 mA cm–2. d) Cycling performance of NiCo2O4, NiCo2S4 and NiCo2(O–S)4 paper supported Li–S cells with sulfur loading of 2.9 mg cm–2 at 0.2C, e) and corresponding EIS spectra after cycling. f) Cycling performance of NiCo2O4, NiCo2S4 and NiCo2(O–S)4 paper supported Li–S cells with sulfur loading of 2.9 mg cm–2 at 0.5C. g) Schematic of the long‐term cycling of the different catalyzing surfaces for LiPS conversion. h) Cycling performance of NiCo2(O–S)4 paper supported Li–S cells with sulfur loading of 3.3 and 4.4 mg cm–2 at 0.8 mA cm–2. Noting that the paper electrodes comprising the Ni‐Co compounds, as well as the small quantity of Li2S6 additives in the electrolyte can contribute extra capacities mainly during the activating cycles.

Optical images after cycling (Figure S29, Supporting Information) show less damage to the surfaces of lithium anodes taken from cells with NiCo2(O–S)4 paper compared to the other two, suggesting more suppression of the LiPS shuttle because of strong LiPS trapping and conversion on the NiCo2(O–S)4 surface. The morphologies of the cycled electrodes at 0.2C were also investigated by SEM to characterize the surface and different parts of the electrode cross‐section (Figure S30, Supporting Information). Previous research has suggested that the continuous precipitation and dissolution of S8–LiPS–Li2S during cycling results in sulfur redistribution towards the electrode surface; thus, high concentrations of LiPS gradually aggregate near the electrode surface, leading to sluggish kinetics for LiPS conversion, deposition of inactive sulfide in every cycle, and formation of an insulating passivated layer that will block the electrode active sites.[ ] A good catalyst would greatly alleviate the passivation process and retain a stable catalyzing surface for long‐term cycling.[ ] The particle morphologies on the surface of the electrode in Figure S30a, Supporting Information, clearly confirms that NiCo2(O–S)4 retains a stable particle shape with obvious 1D nanofibers covered by some precipitates, which suggests there is a higher active exposed surface that is able to catalyze the LiPS conversion. In contrast, the NiCo2O4 and NiCo2S4 particles seem to suffer from structural changes that cause the 1D nanofibers to be decomposed and become covered with thick precipitates, which suggests the surfaces have been inactivated. The top cross‐section (closer to the electrode surface) in Figure S30b, Supporting Information, demonstrates that NiCo2(O–S)4 and NiCo2S4 electrodes are still sufficiently porous for LiPS diffusion, but it might be blocked in the NiCo2O4 electrodes. Going deeper to the middle cross‐section, we can see in Figure S30c, Supporting Information (closer to top cross‐section) that NiCo2(O–S)4 particles still show stable flower‐like morphology with clearly‐visible 1D nanofibers that were uniformly covered with sulfide precipitation as usual. However, the flower‐like NiCo2S4 particle is decomposed to form aggregated, bulk precipitates that are less conductive for SEM imaging; while the NiCo2O4 particle is wrapped by thick precipitates that completely block the active catalyzing surface. Figure S30d,e, Supporting Information, further show the deeper part of the electrode (close to positive coin cell cases), which verifies that NiCo2(O‐S)4 is chemically stable during long‐term cycling and can keep an active surface, but NiCo2S4 is chemically unstable and thus suffering from degradation. Owing to its low conductivity, it is much easier for NiCo2O4 to have inactive sulfide precipitation completely block its active surface. These SEM results clearly demonstrate the failure mechanism of NiCo2O4 and NiCo2S4 particles for LiPS conversion upon long‐term cycling. As illustrated in Figure 5g, our results suggest that NiCo2(O–S)4 not only has high conductivity and high LiPS affinity, but also more chemical stability upon long‐term LiPS conversion. This is all attributed to the large amount of O doping that greatly prevents the passivation of the catalyzing surface upon Li2S2/Li2S deposition and dissolution.

The NiCo2(O‐S)4 papers were also used for fabricating higher areal capacity Li–S cells with high sulfur utilization. The cycling performance for the 3.3 mg cm–2 sulfur loading (46.5 wt.% in cathode) cell (Figure 5h) can achieve an areal capacity of 4.205 mAh cm–2 at the beginning and 3.565 mAh cm–2 (sulfur utilization of 73.0%) after 100 cycles. By increasing the sulfur loading to 4.4 mg cm–2 and also using a 7.1 mg cm–2 NiCo2(O–S)4 paper (40.0 wt.% in cathode), an areal capacity of 5.77 mAh cm–2 at the beginning and 4.28 mAh cm–2 (sulfur utilization of 58.5%) after 100 cycles can be realized. The inset in Figure 5h shows the dissembled lithium anodes, which are damaged due to the high capacity of Li deposition and stripping. Though the long‐term cycling performance of these cells cannot be achieved because of the lithium anode, these results clearly demonstrate that the high‐density O‐doped NiCo2(O–S)4 greatly outperforms the oxide and sulfides for improving the redox kinetics of Li–S cells. There is also potential for applying this high‐density O‐doping chemistry for other metal compounds to develop more active and stable catalysts for improving the performance of Li–S cells.

The good effects of high‐density O‐doped NiCo2(O–S)4 for improving the redox kinetics of Li–S cells are also demonstrated in high sulfur loading cathode with lean electrolyte. In order to further enhance the trapping of LiPS, a NiCo2(O–S)4 interlayer was also applied as described in supporting information. Figure shows the improved areal capacity for NiCo2(O–S)4 interlayer assisted cells with a sulfur loading of 8.75 mg cm–2 and E/S of 3.8 µL g–1. An initial areal capacity of 14.05 mAh cm–2 was achieved at 0.17 mA cm–2 in cells with the interlayer. The rate capacity curve is plotted by merging the initial cycles at 0.17 ‐ 1.33 mA cm–2 in Figure 6b, indicating these areal capacities can be all above 5 mAh cm–2 and the highest capacity at 1 mA cm–2 can be 8.68 mAh cm–2 . The capacity after 16 cycles at 1.0 mA cm–2 is as high as 7.43 mAh cm–2. The discharge‐charge curves at different rates and long‐term cycling performance at 1.33 mA cm–2 are further shown in Figure 6c,d. The discharge profile in Figure 6c still exhibits an intact high voltage plateau at 2.3 V and a low voltage plateau at 2.08 V. In addition, the total capacity is still more than 3 times higher than the high voltage plateau capacity (the theoretical ratio is 4) at 0.67 and 1 mA cm–2, owing to the enhanced electrochemical kinetics within the cell. Figure 6d shows the cycling performance at 1.33 mA cm–2. After 50 cycles, a capacity of 4.54 mAh cm–2 was achieved. Taking into account the high areal capacity, lean electrolyte, and the free‐standing character of the paper electrodes, our work can be competitive with these recently reported works (Table S4, Supporting Information).[ , , ] Since our freestanding papers are free of Al foil and use high sulfur loading, they can be used in pouch cells instead of the double‐sided S coated Al foil. In that case, the self‐standing paper with high sulfur loading of 8.75 mg cm–2 would even perform better and lead to higher energy density and utilization of sulfur.

Figure 6

a) The first activating curves of NiCo2(O–S)4 paper supported Li2S8 catholyte with and without interlayer. b) Rate performance of NiCo2(O–S)4 paper and interlayer assisted Li–S cell with lean electrolyte and high sulfur loading. c) Rate discharge‐charge curves at the highest of the capacity cycle. Long cycling performance at 1.33 mA cm–2.

Furthermore, we also applied this high‐density O‐doping approach to other compounds with high conductivity and find that this simple method is able to obtain higher oxygen content in Mo–OS, Ni–OS, Ni–OP, and Co–OS than the natively oxidized MoS2, NiS, Ni2P, and Co3S4. As shown in Figure S31–S39, Supporting Information, these materials had EDS spectra that showed higher O:S or O:P atom ratios and better LiPS adsorption than natively oxidized sulfides or phosphides. XRD patterns of Ni–OS, Ni–OP and Co–OS show mixed oxides or sulfide phases while Mo–OS shows O‐doped states of MoS2 which is similar to NiCo2(O–S)4. Though in some cases (e.g., for Ni–OS and Ni–OP) the oxide (i.e., NiO) displays the best LiPS adsorption, oxides are usually not as conductive as sulfides and may not result in high utilization of sulfur or resistance to passivation upon cycling, similar to NiCo2O4. Noting that high‐density O‐doping is able to maintain the conductivity, improve the LiPS absorption ability and chemical stability of the original sulfides, this approach can be a promising method for preparing more high‐performance sulfur hosts. Moreover, it should be also emphasized that NiCo2(O‐S)4 and Mo‐OS are special cases because they follow the structure of their sulfides and are not a simple mixed composites of oxides and sulfides like the other Ni–OS, Ni–OP, and Co–OS materials, as indicated by the XRD patterns. Overall, our work demonstrates a simple way for doping large amounts of oxygen into sulfide compounds, the special catalyst NiCo2(O–S)4 or Mo–OS, and a series of potential materials for high performance of sulfur cathodes. The unique NiCo2(O–S)4 we prepared can also have potential uses in other research areas.

Conclusions

To the best of our knowledge, this is the first study that uses a high amount of oxygen doped bimetal sulfide for improving the redox kinetics of Li‐S batteries. By simply controlling the reaction time and quantity of the sulfidation reactant, we find the possibility of preparing a high amount oxygen‐doped NiCo2S4 which we called NiCo2(O–S)4. We demonstrated NiCo2(O–S)4 greatly outperformed the NiCo oxides and natively oxidized sulfides for trapping LiPS and catalyzing LiPS conversion. Taking the advantages of chemical stability, high conductivity, enhanced polar‐polar interaction of Li–O–Co (Ni) species, and activated Co (Ni) facets for absorbing S 2–, the as‐made cell with free‐standing NiCo2(O–S)4 paper greatly showed improved potential for fabricating high areal capacity Li–S batteries. By further applying self‐standing NiCo2(O–S)4 paper and its assisted interlayer for lean electrolyte Li‐S cells, we achieve the highest areal capacity of 8.68 mAh cm–2 at 1mA cm–2 even with 8.75 mg cm–2 sulfur loading and low E/S of 3.8 ul g–1. There is also potential for applying this high‐density O‐doping chemistry for preparing other promising metal compounds or composites. We expect that this study will give more impetus for exploring the highly conductive, chemically stable, and high LiPS affinity metal compounds with high‐density O‐doping for high‐performance Li–S batteries.

Conflict of Interest

The authors declare no conflict of interest.

Author Contributions

H.W. conceptualized the work, and assisted in data curation, formal analysis, funding acquisition, methodology, visualization, project administration, resources, supervision, visualization, writing draft‐reviewing and editing. Y.L. did the conceptualization, data curation, formal analysis, methodology, investigation, validation, software, visualization, writing draft‐reviewing and editing. D.W. conducted investigation, data curation, software, visualization (DFT calculation). H.W. and Y.G. did the validation and review. H.C. performed data curation, formal analysis, methodology, software, visualization and writing draft (DFT calculation). Z.L. assisted in funding acquisition and supervision. W.L., C.X., Q.M., and H.L. provided resources and reviewed the work. C.C. reviewed the work.

Supporting Information

Click here for additional data file.