A Lewis Acid–Lewis Base Hybridized Electrocatalyst for Roundtrip Sulfur Conversion in Lithium–Sulfur Batteries

Electrocatalysts can optimize the sulfur/sulfide reaction kinetics in Li–S batteries to compete with the loss of lithium polysulfides (LiPSs) caused by the shuttling effect. However, the design rationale of electrocatalysts to drive roundtrip sulfur/sulfide conversion is lacking. Here, pairing Lewis acidic and Lewis basic active sites to reach collective adsorption of LiPSs and simultaneous activation of electrophiles and nucleophiles in LiPSs is proposed. This concept is validated by doping polyaniline with protonated metatungstate anions, which enables reduced activation energies for both sulfur reduction reaction and sulfide oxidation reaction and results in significantly improved kinetics. Such electrocatalysts enable a Li–S battery to reach a low capacity‐decay rate of 0.029% per cycle for 1000 cycles. This work would offer insights into battery technologies where sulfur electrocatalysis will play pivotal roles.


Introduction
Sulfur (S 8 ), which has a specific capacity of 1675 mAh g −1 , has emerged as a promising alternative to metal-based cathodes (with a specific capacity below 300 mAh g −1 ) to render high energy densities of lithium batteries. [1]Sulfur electrochemistry in lithium batteries involves a complicated transition between S 8 ring molecules and Li 2 S with 16 electrons exchanged upon complete (SRR) and sulfide oxidation reaction (SOR).Heteroatom-doped graphene and the p-block metal sulfides have been adopted to electrocatalytically reduce apparent activation energy to improve SRR kinetics, [8,14] while various metal sulfides were found to be active catalysts to facilitate oxidation of Li 2 S in SOR. [15]However, most of these catalysts only show activity in a single direction (SRR or SOR) by activating electrophilic or nucleophilic reagents, while the improvement of redox kinetics in reverse conversion is often disadvantaged.Therefore, developing roundtrip catalysis is of great importance to mitigate the loss of the active material in both SRR and SOR.For example, Wang et al. investigated the roundtrip catalytic effect of TiO 2 -Ni 3 S 2 heterostructures.The LiPSs can be adsorbed by TiO 2 and catalytically converted by Ni 3 S 2 in SRR, while both TiO 2 and Ni 3 S 2 promote the activation of Li 2 S for renewing the catalyst surface in SOR. [16,17]Through the engineering of the crystal facet, Yu et al. synthesized a (110) facet-dominated VO 2 , which was used to probe the facet effect in roundtrip catalysis. [18]Moreover, other strategies are also reported for the preparation of roundtrip catalysts, including alloy optimizing, defect manipulating, and single-atom tailoring. [19]owever, the catalyst design rationale for the roundtrip electrochemical reaction kinetics improvement of sulfur/sulfide is lacking.
Recognition of the combination of electrophilic Li cations and nucleophilic polysulfide anions in LiPSs has inspired us to propose the concept of combinational Lewis acid-base catalyst toward more generalized development of roundtrip SRR/SOR electrocatalysis, [20] which we consider as a valuable extension to the recent successful applications of bifunctional catalysts in Li-S batteries. [16,21,22]Our postulation is that the Lewis acidic sites can draw electrons from polysulfide to accelerate SOR, and the Lewis basic site can donate electrons through lithium cations to facilitate SRR.Moreover, for a promising roundtrip electrocatalyst, several key criteria should be simultaneously fulfilled: 1) electron donor and acceptor sites to ensure roundtrip catalytic activity, 2) ultrafine size and uniform dispersion of active centers for high catalytic efficiency, and 3) high electronic conductivity to facilitate prompt electron transfer between catalysts and reactants.
Herein, we report a dual-site Lewis acid-Lewis base electrocatalyst that contains oxometallate-doped polyaniline, as a proofof-concept, to expedite roundtrip SRR and SOR kinetics.The Lewis basic site is constituted of protonated metatungstate anions ([H 2 W 12 O 40 ]6 − ) that are embedded within Lewis acidic polyaniline (PANi).The oxometallate can achieve a molecularlevel dispersion in PANi matrix, so as to ensure high catalytic efficiency.In addition, the roundtrip electrocatalyst can inherit electronic conductivity from the conducting polymer PANi.Hence, the Lewis acid-base combination (WO X -PANi) satisfies the criteria for roundtrip electrocatalyst as outlined above.We found that the WO X -PANi can enhance the collective adsorption of LiPSs through the site-specific interactions with the Li + cations and W. Lv Shenzhen Geim Graphene Center Tsinghua Shenzhen International Graduate School Tsinghua University Shenzhen 518055, China E-mail: lv.wei@sz.tsinghua.edu.cnpolysulfide anions, which are expected to provide more efficient charge transfer pathways for roundtrip sulfur and sulfide conversion.Based on apparent activation energy (E a ), exchange current density and Tafel slope, we show that the activity of co-assembled WO X -PANi outperforms the standalone building blocks for both SOR and SRR.The Li-S battery with this dual-site electrocatalyst incorporated in the sulfur cathode shows both enhanced SOR and SRR kinetics with greatly alleviated shuttling effects, rendering good cycling stability with a low capacity-decay rate of 0.029% per cycle for 1000 cycles.

Lewis Acid-Lewis Base Hybridized Electrocatalyst
A conceptual picture of the WO X -PANi catalyzed sulfur conversion is described in Figure 1a.The use of WO X -PANi as an exemplar option for dual-site Lewis acid-Lewis base electrocatalyst is because of its following unique structural features: 1) homogeneous distribution of the [H 2 W 12 O 40 ]6 − in the PANi matrix, instead of forming heterogeneous composites of crystalline oxometalate in PANi; 2) intrinsic electronic conduction inherited from the oxometalate-doped PANi, which facilitates intramolecular electron transport to the Lewis acid-Lewis base sites; and 3) co-adsorption of Li + cations and polysulfide anions in LiPSs molecules on terminal oxygen atoms and polaronic amine/imine groups, respectively, with reinforced intramolecular forces that contribute to the bond activation.
The WO X -PANi was synthesized in acid solution with aniline, ammonium persulfate (APS), and ammonium metatungstate (AMT).Operando Raman spectroscopy was used to determine the temporal changes of molecular structures during the selfassembly of WO X -PANi in the aqueous solution (Figure 1b).The polymerization of aniline oligomers accelerates after 25 min.The characteristic bands of (C═N) in quinonoid rings turn intense at 1486 cm −1 , coinciding with chain-propagation phase of PANi. [23]he appearance of polaronic structures is proved by the (C∼N + ) and (C∼N +• ) bands (1345-1364 cm −1 ), indicating the PANi formation in the state of conductive emeraldine and the positively charged amines and imines are Lewis acidic.Meanwhile, the bands at ≈978 cm −2 suggest the monodispersed [H 2 W 12 O 40 ]6 − where the negatively charged terminal oxygen atoms as the Lewis basic sites are prone to donate electrons. [24]he selective adsorption of LiPSs on the Lewis acidic and Lewis basic sites of WO X -PANi are determined using X-ray photoelectron spectroscopy (XPS).The two peaks located at 36 and 38.2 eV in the W 4f spectra indicate the existence of [H 2 W 12 O 40 ]6 − in WO X− PANi (Figure S1, Supporting Information). [25]From the O 1s spectra (Figure 1c), the W-O (531.6 eV) shifts to higher binding energy after the Li 2 S 6 adsorption, [26] which exhibits the electron donor ability of [H 2 W 12 O 40 ]6 − toward the LiPSs.In addition, the peak situated at 530.7 eV should be related to the Li-O. [27]The negatively charged terminal oxygen atoms on the external surface of [H 2 W 12 O 40 ]6 − would serve as a Lewis basic site to interact with the Lewis acidic Li + in LiPSs. [20]The shoulder peak at ≈401.7 eV can be assigned to polaronic amine (─NH 2 + ─) and imine groups (═NH + ─) in the N 1s spectra of WO X -PANi/Li 2 S 6 in Figure 1d. [28]hese binding energies of amine and imine groups shift to lower values after Li 2 S 6 adsorption.These phenomena suggest the electronegative S moieties of LiPSs are coordinated with the polaronic sites on PANi chains, which act as electron acceptors (i.e., Lewis acid).Therefore, it is considered that the Lewis acid-base combination in WO X -PANi can enhance the collective adsorption of LiPSs through the site-specific interactions with the Li + cations and polysulfide anions.These paired Lewis acid-Lewis base active sites are expected to provide more efficient charge transfer pathways for roundtrip sulfur/sulfide conversion, where LiPSs are the important intermediates.

Entrapment of Sulfur Species
[31][32][33][34] The peak of stacking feature in WO X -PANi shifts to a higher angle with entrapped Li 2 S 6 solution, indicating its interlayer shrinkage, as given in Figure 2a.This phenomenon can be ascribed to the attraction between WO X -PANi and Li 2 S 6 through the Lewis acid-Lewis base interactions as discussed above.When S 8 was melt-infiltrated or when Li 2 S was electrochemically deposited into WO X -PANi, its interlayer expanded due to the steric effects (Figure 2a).As illustrated by the adsorption isotherms in Figure 2b and the color contrast of the adsorbed Li 2 S 6 solutions in the inset, WO X -PANi quickly captured the yellowish Li 2 S 6 , which turned the solution clear and colorless.The UV-vis spectra of adsorbed Li 2 S 6 solution on different intervals (Figure S3, Supporting Information) were deployed to derive the adsorption rate constant (k) as compared in Figure 2c.The Li 2 S 6 concentration was obtained by the standard spectrophotometric method (Figure S4, Supporting Information), and the adsorbents with the equal total surface area were used despite the different specific surface area (SSA) (Figure S5, Supporting Information).The value of k for WO X -PANi (0.056 m 2 mg −1 min −1 ) is larger than those of PANi and WO X .The adsorption kinetics of LiPSs is essential to tackling LiPSs shuttling, as faster adsorption offsets the shuttling more effectively.The Lewis acid-Lewis base WO X -PANi pairs induced strong catalyst-solution interaction for fast adsorption, which can be vital to the SRR activity and battery performance under lean-electrolyte conditions. [35]he ssNMR 7 Li profiles for pure Li 2 S and WO X -PANi/Li 2 S are displayed in Figure 2d.Pure Li 2 S shows a sharp peak centered at ≈2.74 ppm coherent with the crystalline structure. [36]he 7 Li peak for WO X -PANi/Li 2 S is broadened and downshifted to ≈0.21 ppm, which is attributable to the adhesion of Li 2 S on WO X -PANi as a result of the interaction with the active sites.The operando XRD pattern of WO X -PANi/S cathode in Figure 2e reveals the disappearance of sulfur at the end of the first plateau, indicating the formation of soluble LiPSs.Since Li 2 S was not detected by XRD, it is suggestive that the Li 2 S deposit is amorphous, possibly a consequence of the strong interaction at the dual active sites.

Activity and Efficiency for SRR
The use of dissolved S 8 solution provides a simple approach to studying the kinetics of SRR by recording linear sweep voltammetry (LSV) on rotating disk electrodes (RDE). [14]Figure 3a com-pares the S 8 reduction LSV curves for WO X -PANi with PANi and WO X .WO X -PANi demonstrates the highest onset potential (E onset ) of 2.39 V versus Li/Li + , in comparison with those of PANi (2.26 V) and WO X (2.31 V).Likewise, the half-wave potential (E half-wave ) for WO X -PANi is 2.19 V (versus Li/Li + ), higher than those for PANi (2.07 V) and WO X (2.12 V).The E onset suggests the initiation of the S 8 reduction, and the E half-wave indicates the overpotential for the reaction.Meanwhile, WO X -PANi also attains the largest diffusion-limiting current density (j d ) of 5.78 mA cm −2 , which is more than 7 times of WO X (0.82 mA cm −2 ) and almost 10 times of PANi (0.56 mA cm −2 ).The catalytic activity for SRR can be further elaborated by the exchange current density (j 0 ) as shown in Figure 3b,d.The higher j 0 is an important indicator of faster reaction kinetics. [14]The value of j 0 for WO X -PANi is 0.134 mA cm −2 , which is higher than that for PANi (0.117 mA cm −2 ) and WO X (0.127 mA cm −2 ).Tafel slope provides direct information about the kinetics of electron transfer and is considered a key descriptor for catalyst activity. [37]The Tafel slope (Figure 3b,d) for WO X (319 mV dec −1 ) is much smaller than that for PANi (502 mV dec −1 ), implicating that Lewis basic electronrich sites are more favored in SRR.Notably, the Tafel slope value of WO X -PANi (102 mV dec −1 ) suggests that is much more active than the rest control samples for SRR, implicating the synergistic Lewis acid-Lewis base electrocatalysis in SRR.
The conversion efficiency of SRR can be quantified from the electron transfer numbers by deploying the Koutecký-Levich (K-L) equation (Equations S1 and S2, Supporting Information) to analyze the LSV curves at various rotating rates (Figure S6, Supporting Information).As shown in Figure 3c,d, the slope of the K-L plot gives the electron transfer number (n) which is an indicator of the efficiency and mechanism of SRR.A large n is vital to the high sulfur utilization, as a more complete SRR to solid mitigates the shuttle loss of LiPS intermediates.The WO X -PANi shows the highest n of ≈5.8 compared with PANi (2.8) and WO X (3.1).The highest n of WO X -PANi suggests an overall six-electron reduction process, and therefore S 8 can be theoretically converted into S 8 6 − (equivalent to S 4 2 − +2S 2 2 − ), which is a mixture of loworder soluble LiPSs and solid precipitates.Note that solid-solid conversion from Li 2 S 2 to Li 2 S is unlikely to be reflected in the n, [14] as the RDE method can scarcely influence the solid-state ion diffusion and the output current.
To track the phase evolution of sulfur during SRR, the operando Raman spectra of sulfur species were recorded and compared for WO X -PANi.The strong peaks of S 8 at 155, 220 and 472 cm −1 were detected, as shown in Figure 3e,f.The S 8 fully converted to LiPSs as electrode potential gradually decreased to 2.15 V at the end of the first plateau.The intensity of the signature peaks for LiPSs decreased as the discharge proceeded.More importantly, the characteristic peak for Li 2 S 2 (578 cm −1 ) was observed at 1.70 V, [38] suggesting that WO X -PANi accelerated sulfur reduction to insoluble sulfides.However, it was difficult to detect the signal of Li 2 S by operando Raman spectroscopy. [39]

Roundtrip Sulfur Utilization
Temperature-dependent cyclic voltammetry (CV) of SRR and SOR in Li-S cells was conducted to acquire kinetics information at various temperatures and electrode potentials for WO X -PANi, WO X and PANi (Figure 4a; Figure S7, Supporting Information).
The CV curves obtained at 283 K show the smallest onset overpotential of SOR on WO X -PANi (Figure 4b).The Tafel slopes for Li 2 S oxidation on WO X , PANi, and WO X -PANi are 253, 246, and 123 mV dec −1 , respectively (Figure 4c).The lowest Tafel slope for WO X -PANi compared to those of controls suggests the best activity of Lewis acid-base combination for oxidizing Li 2 S during SOR.For SRR, the trend of Tafel slopes is consistent with LSV results as discussed above (Figure S8, Supporting Information).The smallest polarization in CV peaks and Tafel slopes for WO X -PANi reveal its more efficient roundtrip catalysis by the synergy of Lewis acid-base pairing.
The extended CV enables the determination of the activation energy for every single step of sulfur reactions according to Arrhenius' Equations ( 1) and ( 2): Where j is the apparent current density (mA cm −2 ), and the surface area of the electrode is 1.13 cm −2 , k is the apparent rate constant, c is the concentration of LiPSs near the surface of electrodes, E a is the apparent activation energy (J mol −1 ), R is the molar gas constant (R = 8.314 J mol −1 K −1 ), T is the measurement temperature (K).WO X -PANi shows the lowest E a for all four redox peaks, as compared in Figure 4d-f and Figure S9 (Supporting Information).The E a for S 8 reduction to soluble LiPSs at R1 peak is 22 meV for WO X -PANi, in contrast to the value of 117, and 105 meV for PANi and WO X .The reduction of LiPSs to insoluble Li 2 S 2 /Li 2 S requires overcoming a larger barrier, [14] and hence the E a increases to 196 meV at R2 peak for WO X -PANi and to 348, 280 meV for PANi and WO X .For the SOR process, the values of E a for WO X -PANi at O2 and O1 peaks are 138 and 34 meV, respectively, where the former peak relates to Li 2 S conversion to LiPSs and the latter relates to the oxidation of LiPSs to S 8 .Despite the fact that the Li 2 S deposition and dissolution reactions are less favorable, [15,40] WO X -PANi in general presents the best activity in reducing the E a for all reaction steps, which suggests that WO X -PANi is a roundtrip electrocatalyst for sulfur reduction and Li 2 S oxidation.
Interestingly, WO X shows lower E a in SRR while the PANi has lower E a in SOR, indicating their selective catalytic activity that is related to their specific electronic properties.We understand that the electron-rich [H 2 W 12 O 40 ]6 − tends to donate electrons to the reduction reactants (S 8 and LiPSs) and the PANi chains are apt to capture electrons from the oxidation reactants (Li 2 S and LiPSs).As a consequence, the WO X -PANi with the combination of the Lewis acid-Lewis base active sites demonstrates superior activity for the roundtrip reactions.

WO X -PANi Sensitized Reduced Graphene Oxide (rGO) for Sulfur Host
We used WO X -PANi for the surface sensitization of sulfur host, which is rGO with turbostratic stacked structure and large SSA in this study (Figure S10, Supporting Information).This combination can maximize the electrocatalytic activity of WO X -PANi for SRR/SOR by leveraging the large surface area and good conductivity of rGOs, where the small surface area of WO X -PANi can possibly negate its activity (Figure S5, Supporting Information).The distribution of WO X -PANi particles on rGO is clearly observed by the X-ray spectroscopy (EDX) elemental maps (Figure S11, Supporting Information).By immobilizing WO X -PANi on rGO sheets (WO X -PANi/rGO), we show the decreased size of WO X -PANi particles (Figure S12, Supporting Information), which means more active surfaces are now exposed and the active sites are more accessible to LiPSs.While the mass content of WO X -PANi (52.9 wt%, Figure S13, Supporting Information) is adjustable, this study is primarily intended to understand the "sensitization effect" on rGO induced by WO X -PANi.The WO X -PANi/rGO hybrid maintained the stacking feature of WO X -PANi (Figure S14, Supporting Information) and demonstrated a high SSA of 169 m 2 g −1 (Figure S15, Supporting Information).The WO X -PANi sensitization obviously improved the Li 2 S 6 adsorption on nonpolar rGO (Figure S16, Supporting Information), and Raman peaks at ≈419 and 506 cm −1 indicate the appearance of LiPSs on WO X -PANi/rGO (Figure S17, Supporting Information). [12,41]Fourier transform infrared (FTIR) spectrum of WO X -PANi/rGO after Li 2 S 6 adsorption demonstrates the characteristics peak of S-S vibrational modes, which can be ascribed to the adsorbed LiPSs (Figure S18, Supporting Information).
CV curves of symmetric cells show the redox kinetics of SOR at anodic peaks and SRR at cathodic peaks on the working electrodes (Figure 5a).Two anodic peaks at ≈0.07 and 0.37 V should be attributed to the oxidation from Li 2 S to Li 2 S 6 and from Li 2 S 6 to sulfur on the working electrode, respectively.The cathodic peaks at ≈−0.07 and −0.37 V indicate the conversion of sulfur to Li 2 S 6 with the subsequent reduction of Li 2 S 6 to Li 2 S. The WO X -PANi/rGO shows a higher redox current density, suggesting that the WO X -PANi can enhance the roundtrip kinetics and utilization of LiPSs. [16]WO X -PANi/rGO demonstrated a faster process than that of rGO in the Li 2 S deposition tests, implicating the faster LiPSs conversion kinetics toward Li 2 S (Figure 5b,c).As shown in the insets, the Li 2 S deposits on WO X -PANi/rGO have 3D granular morphology, while those on rGO show 2D sheetlike morphology.Moreover, the WO X -PANi/rGO tends to follow the 3D progressive nucleation model of Li 2 S nucleation (Figure S19, Supporting Information), suggesting that the WO X -PANi may also act as nucleation sites to guide the 3D growth of Li 2 S on rGO sheets.The 3D morphology avoids the catalyst surface passivation, which contributes to better catalytical activity during the redox reactions. [42]e used the CV curves of the Li-S batteries (Figure 5d) and the Tafel plots (Figure S20, Supporting Information) to further verify the effect of WO X -PANi/rGO on the roundtrip redox kinetics of LiPSs in cells.The electrochemical impedance spectroscopy (EIS) results of the Li-S batteries with WO X -PANi/rGO and rGO are shown in Figure 5e, and the WO X -PANi/rGO exhibits smaller charge transfer resistance and lower diffusion impedance.Density functional theory (DFT) simulation shows that, compared with rGO, WO X -PANi shows better adsorption toward LiPSs through the Li─O─W bond in WO X -PANi.The potential energy surfaces of the reaction pathways were built by considering the adsorption of sulfur species (S 8 , Li 2 S 8 , Li 2 S 6 , Li 2 S 4 , Li 2 S 2 , and Li 2 S) on the Lewis acid-base pair and rGO, as shown in Figure 5f.The results show that the conversions from S 8 to Li 2 S 8 are spontaneous exothermic, and the following steps are endothermic or nearly thermoneutral.As expected, the energy difference from Li 2 S 6 to Li 2 S 4 is reduced by WO X -PANi.In addition, the ratedetermining step from Li 2 S 2 to Li 2 S on WO X -PANi also shows a lower energy difference than rGO.Overall, WO X -PANi shows a lower energy pathway for sulfur conversion to Li 2 S, indicating SRR is kinetically more favorable on WO X -PANi than rGO.
Li-S coin cells using WO X -PANi/rGO as the sulfur host with a sulfur content of ≈70 wt% (Figure S21, Supporting Information) were assembled to evaluate the performance of the Lewis acid-base electrocatalysts in sulfur cathodes.The galvanostatic charge-discharge profiles (Figure 6a) for the WO X -PANi/rGO cell show a higher initial Coulombic efficiency of 95.0% at 0.1C compared to 92.3% of rGO, which suggests a suppressed shuttle loss of LiPSs in the first cycle of the WO X -PANi/rGO cell.In addition, the sulfur cathode with WO X -PANi/rGO exhibits a smaller polarization (163 mV) than that of the rGO cathode (230 mV), indicating enhanced redox kinetics.The potential of Li 2 S activation is 2.25 V for the WO X -PANi/rGO cell, compared to 2.30 V for the pristine cell.This should be attributed to the catalytic effect of Lewis acidic sites in decreasing the energy barrier of Li 2 S activation.As shown in Figure 6b,c, the WO X -PANi/rGO cell delivers a higher discharge capacity of 736.1 mAh g −1 at 2C. CV tests at various scan rates demonstrate smaller cathodic peak shifts for WO X -PANi/rGO (Figure 6d; Figure S22a,b, Supporting Information).The higher Li + diffusion coefficient (Figure S22c-f, Supporting Information) of WO X -PANi/rGO indicates the favorable reaction kinetics for lithiation and delithiation of sulfur and sulfide, respectively. [42]Moreover, the sulfur cathode with WO X -PANi/rGO retains 683.1 mAh g −1 after 300 cycles at 0.3C (Figure 6e).The WO X -PANi/rGO based sulfur cathode shows a low capacity-decay rate of 0.029% after 1000 cycles at 1C (Figure 6f), while the rGO based cathode shows a capacity loss of 0.055% per cycle.The cycling stability of the WO X -PANi/rGO based sulfur cathode outperforms many reported heterogeneous catalysts, as shown in Figure S23 (Supporting Information).For further comparison, the rate and cycling performance of sulfur cathodes with pristine WO X -PANi, PANi/rGO, and WO X /rGO are inferior to that of WO X -PANi/rGO (Figure S24, Supporting Information), verifying the effectiveness of WO X -PANi sensitization of the rGO sulfur host and the Lewis acid-base catalysts design with dual active sites.Even under the high sulfur loading (3.7 and 5.9 mg cm −2 ) and low electrolyte/sulfur (E/S) ratio (5.7 μL mg −1 ), stable cycling was achieved at 0.1C for 60 cycles by the WO X -PANi/rGO catalysts (Figure 6g).

Conclusion
In summary, this work presents our consideration of the design rationale for electrocatalysts endowed with catalytic activity for roundtrip sulfur/sulfide conversion.The concept of pairing Lewis acidic and Lewis basic active sites in an electrocatalyst is thereby proposed and validated through the self-assembly of Lewis acidic conductive polymer and Lewis basic metatungstate.The key to this finding is the highly dispersed [H 2 W 12 O 40 ]6 − clusters and the positively charged amines/imines of PANi on WO X -PANi, which work in concert to improve the kinetics, efficiency, and activity in both SRR and SOR.The improved kinetics of LiPSs redox conversions contribute to greatly alleviated LiPSs shuttle effect.As a result, the Li-S battery with WO X -PANi demonstrates remarkable enhancement in reaction efficiency and cycling stability, even under harsh conditions with the high sulfur loading and lean electrolyte.Overall, the design rationale is fundamental to addressing key challenges of Li-S batteries, which also provides new insights into optimizing redox reaction kinetics in batteries through electrocatalysis.

Experimental Section
Material Preparation: Aniline and ammonium persulfate (APS) were purchased from Sigma-Aldrich.Ammonium metatungstate (AMT) was purchased from Shanghai Macklin.Tungstic acid (WO X ) was purchased from Shanghai Aladdin.WO X -PANi was prepared at 5 °C by mixing aniline in H 2 SO 4 solution with AMT and APS under continuous stirring.The yielded powder was collected, washed and dried in vacuum at 70 °C.PANi was synthesized using the same procedure without AMT.Graphite oxide (GO) was prepared using a modified Hummers method. [43]Exfoliation of GO was conducted by thermal treatment at 300 °C under Ar for 1 h with a heating rate of 10 °C min −1 .Then, the exfoliated GO was thermally treated at 900 °C for 4 h under Ar with a heating rate of 10 °C min −1 to obtain rGO.The WO X -PANi/rGO, WO X /rGO and PANi/rGO were synthesized using incipient wetness impregnation.
Materials Characterizations: SEM images were obtained using Hitachi S-4800.Transmission electron microscopy (TEM) images were obtained using JEOL JEM-F200.Thermogravimetric analysis (TGA) data was collected on TGA2 (METTLER TOLEDO).To determine the mass ratio of WO X -PANi in WO X -PANi/rGO, the sample was heated to 1000 °C under air at a rate of 5 °C min −1 .The sulfur content in various samples was measured by heating the samples to 600 °C under N 2 with a rate of 5 °C min −1 .XRD measurements were performed on Rigaku SmartLab.For the XRD patterns, WO X -PANi/S was prepared by melt-infiltrating sulfur into WO X -PANi.WO X -PANi was soaked in the 2 mm Li 2 S 6 of 1,2-dimethoxyethane (DME) and 1,3-dioxolane (DOL) solution (1:1 by volume) and washed and dried to prepare WO X -PANi/Li 2 S 6 .WO X -PANi/Li 2 S was obtained by disassembling the Li-S cell to collect the WO X -PANi based sulfur cathode (sulfur loading: 9.7 mg cm −2 ) after discharging to 1.7 V at 0.1C.1C is equal to 1675 mA g −1 .Operando XRD was performed using WO X -PANi based sulfur electrodes with a mass ratio of WO X -PANi:S:carbon nanotube (CNT) as 4:4:2 at 0.1C.The adsorption-desorption isotherms were collected with BELSORP Max at 77 K and the SSA was calculated based on the Brunauer-Emmett-Teller equation.Raman spectra were collected on LabRAM HR800 (HORIBA Scientific) using the 532 nm Laser.The WO X -PANi/rGO/Li 2 S 6 was washed by DME before the Raman test.For the operando Raman test, the WO X -PANi based sulfur cathode was prepared by mixing WO X -PANi, Sulfur, CNT, and polytetrafluoroethylene with a mass ratio of 4:4:1.5:0.5 in ethanol and forming a free-standing film, which was subsequently dried at 60 °C in vacuum for 12 h.The sulfur loading was ≈2.3-2.5 mg cm −2 and the discharge rate was set at 0.3C.The spectra were recorded with an acquisition time of 5 s for 3 times and a frequency of 5%.FTIR spectra were recorded using the Nicolet iS50 FTIR spectrometer (Thermo Fisher Scientific).XPS was recorded on ESCALAB 250Xi (Thermo Fisher Scientific) with a monochromatic Al K source.The UV-vis spectroscopy results were obtained on Cary 5000 (Agilent).
More details of computational methods and measurements including Li 2 S 6 adsorption kinetics, NMR, LSV with RDEs, activation energy and electrochemical measurements are provided in the Supporting Information.

Figure 1 .
Figure 1.A model dual-site Lewis acid-Lewis base electrocatalyst and its functions.a) Illustration of collective adsorption for sulfur electrocatalysis on WO X -PANi.b) Operando Raman spectra of WO X− PANi growth with various time intervals, which indicate the self-assembly of the protonated metatungstate clusters on conductive polyaniline matrix.The WO X -PANi is taken as a model material system to demonstrate the idea of a dual-site Lewis acid-Lewis base electrocatalyst.It is reasonable to extend this procedure to the wide libraries of oxometalates and conductive polymers.The high-resolution XPS c) O 1s and d) N 1s spectra of WO X -PANi before and after the Li 2 S 6 adsorption.

Figure 2 .
Figure 2. Interaction between WO X -PANi and sulfur species.a) XRD patterns of WO X -PANi, WO X -PANi/S, WO X -PANi/Li 2 S 6 , and WO X -PANi/Li 2 S. b) Adsorption isotherms of Li 2 S 6 by WO X -PANi, WO X , and PANi.The inset image shows the color change after adsorption.c) Adsorption rate constant of Li 2 S 6 for WO X -PANi, WO X , and PANi.d)7 Li ssNMR spectra of Li 2 S and WO X -PANi/Li 2 S. e) Operando XRD pattern of WO X -PANi/S cathode during discharge/charge at 0.1C.

Figure 3 .
Figure 3. Activity and conversion efficiency for SRR.a) LSV curves at 1600 rpm, b) Tafel plots, c) Koutecký-Levich plots (j −1 versus  −1/2 ) at 1.9 V, and d) the values of Tafel slope, n, and j 0 of WO X -PANi, PANi and WO X for SRR.Phase evolution of sulfur species during SRR studied by operando Raman spectra of the sulfur cathode with e) WO X -PANi, and f) the corresponding discharge curve.

Figure 4 .
Figure 4. Activation energy analysis for roundtrip sulfur utilization.a) Temperature-dependent CV curves of sulfur on WO X -PANi.b) CV curves of sulfur on WO X -PANi, WO X , and PANi at 283 K, and c) Tafel plots for the Li 2 S oxidation peaks.d,e) Arrhenius plots at (d) R2 peak and (e) O2 peak for WO X -PANi, WO X , and PANi.f) E a for the SRR and SOR of WO X -PANi, WO X , and PANi.

Figure 5 .
Figure 5. Electrode kinetics of WO X -PANi-sensitized rGO for sulfur reaction.a) CV curves of symmetric cells of rGO and WO X -PANi/rGO based electrodes using the Li 2 S 6 as the active material.Potentiostatic discharge curves at 2.05 V for Li 2 S deposition and Scanning electron microscopy (SEM) images (inset) on Li 2 S nucleation of b) rGO, and c) WO X -PANi/rGO.d) CV curves of the WO X -PANi/rGO and the rGO based sulfur cathodes.e) Nyquist plots of rGO and WO X -PANi/rGO based sulfur cathodes before operation.f) Computational simulation of potential energy surfaces for the reaction pathway from S 8 , Li 2 S 8 , Li 2 S 6 , Li 2 S 4 , Li 2 S 2 to Li 2 S adsorbed on the rGO and local tungstate/PANi pair of WO X -PANi substrates.

Figure 6 .
Figure 6.Electrochemical performance of WO X -PANi-sensitized rGO for sulfur cathode.a) Initial discharge-charge profiles at 0.1C of Li-S batteries with WO X -PANi/rGO based sulfur cathode and rGO based sulfur cathode.b) Discharge-charge profiles of Li-S batteries with WO X -PANi/rGO based sulfur cathode.c) Rate performance of Li-S batteries with WO X -PANi/rGO based sulfur cathode and rGO based sulfur cathode.d) Peak shift in CV curves of Li-S batteries at various scan rates.Long cyclic operation at e) 0.3C and f) 1C with WO X -PANi/rGO and rGO.g) Cyclic performance with WO X -PANi/rGO under the high sulfur loading and the low E/S ratio of 5.7 μL mg −1 .