Constructing sulfur and oxygen super-coordinated main-group electrocatalysts for selective and cumulative H2O2 production

Direct electrosynthesis of hydrogen peroxide (H2O2) via the two-electron oxygen reduction reaction presents a burgeoning alternative to the conventional energy-intensive anthraquinone process for on-site applications. Nevertheless, its adoption is currently hindered by inferior H2O2 selectivity and diminished H2O2 yield induced by consecutive H2O2 reduction or Fenton reactions. Herein, guided by theoretical calculations, we endeavor to overcome this challenge by activating a main-group Pb single-atom catalyst via a local micro-environment engineering strategy employing a sulfur and oxygen super-coordinated structure. The main-group catalyst, synthesized using a carbon dot-assisted pyrolysis technique, displays an industrial current density reaching 400 mA cm−2 and elevated accumulated H2O2 concentrations (1358 mM) with remarkable Faradaic efficiencies. Both experimental results and theoretical simulations elucidate that S and O super-coordination directs a fraction of electrons from the main-group Pb sites to the coordinated oxygen atoms, consequently optimizing the *OOH binding energy and augmenting the 2e− oxygen reduction activity. This work unveils novel avenues for mitigating the production-depletion challenge in H2O2 electrosynthesis through the rational design of main-group catalysts.

This catalyst is interesting and has not been reported for 2-e ORR so far.However, a few technical issues must be fully addressed to meet the high standard of Nature Communications.Detailed comments are listed below: 1. Some recent advancements in theoretical understanding and material innovation have led to the development of a series of efficient main group metal single-atom catalysts for the electrosynthesis of H2O2 in alkaline media (Angew. Chem. Int. Ed.2022,134,e202117347).In comparison with the activity, the Pb SA/OSC catalyst showed excellent stability in the flow cell.More theoretical discussion is needed to explain this advantage.
2. In Figure 1c, the DFT calculations on the Gibbs free energy should be reconducted to consider the related species in alkaline media (OH-and HO2-) instead of those in acid (H2O and H2O2).The authors could refer to some references, such as Phys.Chem.Chem. Phys., 2013, 15, 148-153.3. Operando FTIR Spectroscopy of Pb SA/OC and OSC should be offered to prove the adsorption of O2 and OOH on Pb SAs rather than S and O. Also, isotope experiment is required to confirm the origin of OOH.
4. Thiocyanide (SCN-) poisoning experiment is needed to probe the active site of Pb SA/OSC. 5. To demonstrate the high stability of the Pb SA/OSC catalyst, some post characterizations are required.
6.The structural information was gained from the EXAFS fitting analysis, and the fitting errors must be provided.
1.For the RRDE test, the ORR performance of Pb SA/OSC in neutral medium (0.10 M PBS) was determined.Why did the authors not add tests utilizing gas diffusion electrodes under the same conditions?
Reply #1.1: Gas diffusion electrodes were utilized to evaluate the 2e -ORR performance of Pb SA/OSC in a neutral medium.To ensure an adequate electrolyte concentration for high current densities, 1 M Na2SO4 solution was used in the gas diffusion electrode tests.As shown in Figure R1, the Faradaic efficiencies (FEs) of Pb SA/OSC for H2O2 generation in the neutral medium remained above 91.9%across the applied current densities ranging from 50 to 300 mA/cm 2 .It can be concluded that Pb SA/OSC also demonstrated exceptional efficacy in producing H2O2 via gas diffusion electrodes within a neutral medium.
The data depicted in Figure R1 has been incorporated into the revised Supplementary information (Figure S43). 2. For the flow cell test, the authors used carbon paper as a base to form a gas diffusion electrode.However, many literatures have reported carbon-based materials showed good performance towards 2e-ORR.The contribution of the carbon paper to the ORR performance needs to be clarified.Reply #1.2:We agree the reviewer's perspective that carbon-based materials have demonstrated good performance towards 2e -ORR, as corroborated by numerous literatures.Typically, a GDE comprises of two distinct layers: a carbon fiber layer and a carbon black layer infused with polytetrafluoroethylene.The carbon black layer, situated between the carbon fiber layer and the catalyst layer, has the potential to exhibit 2e -ORR activity within a flow-cell electrolyzer.
To clarify the contribution of the carbon paper to the ORR performance, a blank GDE (YLS30, Suzhou Sinero Technology Co., China) devoid of the catalyst layer was employed for H2O2 production at a current density of 50 mA/cm 2 (Figure R2).In the initial 60 seconds of electrolysis, the blank GDE necessitated a notably elevated overpotential to activate the carbon black nanoparticles, thereby reaching the desired current density.In comparison with the Pb SA/OSC catalyst, the blank GDE exhibited a substantially higher overpotential (exceeding 150 mV) to sustain a consistent current density throughout the electrolysis, despite both the blank GDE and Pb SA/OSC showcasing notable H2O2 Faradaic efficiencies.Subsequently, we carried out bulk electrolysis for H2O2 production using the blank GDE at 200 mA/cm 2 .As shown in Figure R3a, the H2O2 Faradaic efficiencies of the blank GDE were lower than those of the Pb SA/OSC sample and exhibited a decline over the course of electrolysis.It has been documented that the oxygen content at the GDE surface undergoes a substantial increase under cathodic conditions, rendering the GDE hydrophilic (Science, 2018, 360, 783-787;Energy Environ. Sci., 2021,14, 1959-2008).It is plausible that the electrolysis for H2O2 production led to the grafting of numerous oxygen groups onto the carbon black layer of the blank GDE.Consequently, the GDE lost its hydrophobicity and became penetrated by the aqueous electrolyte, resulting in a decline in H2O2 Faradaic efficiency and structural damage to the GDE.Subsequent observations revealed the destruction and separation of the GDE into distinct carbon fiber and carbon black layers after 2 h of electrolysis at 200 mA/cm 2 (Figure R3b).
The aforementioned analysis suggests that the carbon paper without the catalyst layer necessitated a substantial overpotential to enable the carbon black layer to exhibit 2e -ORR activity, likely resulting in structural impairment to the GDE.Therefore, the contribution of the carbon paper to the ORR performance appeared to be negligible.3. The authors mentioned "The superior performance of the Pb SA/OSC catalyst indicated that main-group sites with a regulated local coordination environment effectively prevented the self-decomposition or further electrochemical reduction of generated H2O2".However, the experimental and theoretical data do not support this conclusion.The authors need discuss more in-depth.
Reply #1.3: Main-group sites are charactered by fully occupied d-orbitals, which lack of a combination of both empty and occupied host orbitals.Generally, these main-group sites exhibit inert activity towards electron transfer reactions, which, in turn, has the potential to mitigate the risks associated with the self-decomposition or electrochemical reduction of H2O2.
Secondly, in our previous manuscript, we conducted quasi-in situ EPR measurements and H2O2 reduction reaction experiments to elucidate the Fenton and H2O2RR activity of Pb SACs.As shown in Supplementary Fig. 35, no characteristic signals of hydroxyl radical resulting from Fenton reactions were discernible in the EPR spectra of Pb SA/OSC, Pb SA/OC, and OSC during the 2e − ORR process.This finding led to the conclusion that the Pb SACs exhibited diminished activity towards Fenton reactions.Furthermore, the results of the H2O2RR experiments indicated that the rate of H2O2 electroreduction on Pb SA/OSC exhibited only marginal increase with a higher overpotential and H2O2 concentration (Fig. 4f).The current density for H2O2RR on Pb SA/OSC remained below -0.07 mA cm -2 when the potential exceeded 0.40 V vs. RHE, indicating the limited activity of the main-group sites of Pb SA/OSC towards H2O2RR.Hence, it is evident that Pb SA/OSC played a pivotal role in avoiding the decomposition or electrochemical reduction of H2O2.Furthermore, minimal Fenton and H2O2RR activity are beneficial for achieving a high net rate of H2O2 production (i.e., the production rate minus the electroreduction rate of H2O2), thereby facilitating H2O2 accumulation.As show in Fig. 5c, the concentration of accumulated H2O2 increased nearly linearly, reaching 1358 mM (4 wt.%) after 2 h of electrolysis at 200 mA/cm 2 .Notably, the H2O2 FEs exhibited only a slight decrease trend (97% to 91%) throughout the course of electrolysis.These outcomes further underscore the efficacy of the Pb SA/OSC catalyst in preventing the depletion of generated H2O2 during electrolysis.
Finally, accepting the reviewer's suggestion, we have performed supplemented theoretical investigations to assess the stability of the synthesized H2O2 molecules.As demonstrated in the experimental results, the coordination structures of Pb sites exhibited various abilities in accumulating the synthesized H2O2.To give a comprehensive comparative analysis of the impact of coordination environments, we took PbO4 and PbS4 as examples in the control group, while PbS4O2 as the target group.To probe the complex configurations of H2O2 within a solvation environment at the solid-liquid interface, we used the combined explicit-implicit water model to conduct a more in-depth study.Different initial geometries of H2O2 at the catalyst-water interface were constructed and relaxed.For the PbO4 and PbS4 catalysts, we observed the cleavage of the O-O bond in H2O2 concurrent with the formation of adsorbed *OH of *O species on the Pb sites.In contrast, the PbS4O2 catalysts exhibited a capability to stabilize the molecular H2O2 through hydrogen bonding with water molecules.Therefore, the local coordination environment of the Pb SA/OSC catalyst, especially the PbS4O2 catalyst, played a crucial role in preserving H2O2 by inhibiting the O-O bond cleavage (For more details, please refer to our response in Reply #3.1).
To address the reviewer's concern, we have incorporated the related discussion in the revised manuscript (Page 21, Lines 392-397).
4. For stability testing, the authors did not give sufficient characterizations to prove the stability of catalyst after 100 h of electrolysis.

Reply #1.4:
To verify the stability of the catalyst after 100 h of electrolysis, we examined the catalysts with transmission electron microscopy (TEM) X-ray diffraction (XRD) and Raman characterizations.Figure R5 shows the TEM images of the Pb SA/OSC at different magnifications after prolonged electrolysis, indicating the absence of Pb clusters or small PbS species.The XRD patterns (Figure R6) obtained from the Pb SA/OSC on GDE exhibited no distinctive peaks associated with crystalline PbS species after electrolysis, in alignment with the observation from the TEM measurements.Additionally, the Raman spectra (Figure R7) of the Pb SA/OSC revealed the characteristic D and G bands of conductive carbon materials, with calculated ID/IG values comparable before and after electrolysis.These findings demonstrate that there were no discernible structural changes in the Pb SA/OSC after the stability testing.Therefore, the Pb SA/OSC catalyst exhibited outstanding stability after 100 h of electrolysis.
To address the reviewer's concern, we have added the related description into the revised manuscript (Page 22, Lines 407-409), and incorporated the data presented in Figures R5-R7 into the revised Supplementary information (Figures S44-S46).

Response to Reviewer 2's comments
This study shows how Pb single sites coordinated with sulfur can increase the H2O2 selectivity.The study is unique in their approach with stitching together fundamental first principles studies, lab scale tests, a battery of characterisation studies, industrial scale tests, and an application of using the H2O2 for an oxidation reaction.The amalgamation of these factors create a unique study that is potentially of interest to Nature Communications.But, before that, a few technical points need to be revised, the manuscript in its current form is not suited for Nature Communications.Revising these points is critical to justify overall conclusions.The primary concerns are with explaining the rationale for selecting the DFT structures, not considering oxidized sulfur sites when evidence suggests otherwise, and not reporting the extent of sulfur leaching.
Reply: We sincerely appreciate the reviewer's insightful comments as well as the recognition and support of our work.In response to the reviewer's concerns, we have offered a comprehensive explanation regarding the selection of the DFT structures.We have also explored the impact of oxidized sulfur sites on the 2e − ORR catalytic performances and reported the extent of sulfur leaching, all of which have significantly enhanced the depth and rigor of this work.
1. Figure 1 shows the structures used for DFT simulations.Why were oxidized sites not considered, when such oxidation of sulfur is expected and even seen experimentally (XPS studies, C-SOx peaks)?
Given the larger spatial distance between C-SOx group and Pb site, their interactions become weaker.To investigate the potential impact of C-SOx group on catalytic activity, only one benzene ring was used to separate S and Pb.As shown in Figure R8a, the electron-withdrawing C-SOx substitutes are located on the side C atom of the carbon support.Based on a set of intermediates (*OOH, *O and *OH, Figure R8b), the calculated overpotential for 2e − ORR pathway (pH = 0, U = 0 V) was 0.35 V, suggesting that the C-SOx groups are unfavorable for the formation of H2O2 (Figure R8c).Here, the oxidized sulfur groups are not in the first coordination shell of Pb site, but in close proximity.They play a role similar to that of the enzyme pocket residues found in natural peroxidase during the catalytic process, including H2O2 binding and the cleavage of the O-O bond (J.Am.Chem. Soc., 2023, 145, 30, 16835-16842).
As the primary goal of this work is to screen the first-shell coordination structure of the Pb site, a trade-off between computational cost and screening efficiency becomes essential.By focusing on the distinctions among various Pb sites, we assume that the effect of the C-SOx group on the candidate Pb sites can be treated as a background factor.However, this simplified approach may introduce uncertainty, including the spatial distance and the number of the C-SOx group around the Pb site.Since the XPS detection of sulfur types is based on a statistical analysis of the entire sample, it is challenging to determine the specific spatial distribution of C-SOx groups.Accepting the reviewer's valuable suggestion, we have enhanced understanding of the effect of C-SOx group on the oxygen reduction activity of Pb site by incorporating the analysis of the oxidized sulfur.
To address the reviewer's concern, we have incorporated the data presented in Figure R8 into the revised Supplementary information (Figure S5).

What was the rationale for selecting these structures?
Reply #2.2:We select the DFT structures because of the following reasons: Firstly, main-group SACs hold the potential to prevent the depletion of generated H2O2 during electrolysis by suppressing Fenton and H2O2 reduction reactions.For this potential, main-group sites are expected to surpass transition metals in terms of inert activity towards the electron transfer reactions, as they lack a combination of empty and occupied host orbitals.Given the demonstrated high activity of Pb materials in various electrocatalytic reactions, Pb-based catalysts were chosen as the main-group model catalysts.It is crucial to note that the performances in catalytic processes are highly dependent on the coordination environments due to electronic structure alterations.To tailor the 2e − ORR activity of Pb single-atom sites, a coordination environment engineering strategy was applied.Previous reports show that the strong chemical interaction between metal and sulfur atoms facilitates the formation of strong and thermally stable metal-sulfur bonding (Nat. Commun., 2021, 12, 3135;Sci. Adv., 2019, 5, eaax6322), significantly mitigating the aggregation of metal sites.Consequently, structure models of Pb SACs were constructed by supporting Pb atoms on graphene with simultaneous co-coordination of S and O atoms.Subsequently, SACs with the typical M-N4 structure have been extensively demonstrated to exhibit a high activity for various electrochemical reactions.Several potential Pb-X4 models (where X = O, S) were constructed through diverse combinations of coordinated S and O atoms (PbO4, PbS1O3, PbS2O2, PbS3O1, PbS4).Moreover, PbS3O1-f, PbS4-f models were also constructed to investigate the influence of S type within the graphene substrate.Accordingly, the adsorption energies of *OOH, *O and *OH on the Pb SACs structures were calculated.The results indicate that Pb SACs co-coordinated with both S and O and a higher S ratio, result in an optimal adsorption strength of the *OOH intermediate (Supplementary Fig. 2).Based on these predicted theoretical findings, we attempted to synthesize the Pb SACs coordinated with different S/O ratios.
Furthermore, it has been reported in the literature that the most common coordination numbers for main-group Pb sites are 4 and 6 (J.Am.Chem. Soc., 2005, 127, 9495-9505;Inorg. Chem., 1998, 37, 1853-1867).In this regard, a Pb SACs model was constructed in which Pb sites were co-coordinated by 6 S and/or O atoms with a higher S ratio (PbS4O2).Notably, PbS4O2 is positioned closest to the apex of the limiting potential volcano, akin to the PtHg4 catalyst.As a result, PbS4O2 exhibits a high selectivity to for H2O2 formation with a low overpotential of 0.006 V.The EXAFS fitting of the optimized Pb SA/OSC catalyst clearly revealed that Pb sites are cocoordinated by both S and O atoms.The best-fit result of the EXAFS data (Supplementary Table 6) indicate that the coordination numbers of S and O for Pb sites closely resemble the PbS4O2 model.Additionally, the averaged interatomic distances derived from the fitting results also strongly concur with the distances of Pb-O and Pb-S in the PbS4O2 structure.This predicted structure is strongly supported by the experimental results.
To address the reviewer's concern, we have incorporated the related discussion into the revised manuscript (Page 6, Lines 102-104).

Can the free energy diagram be replotted at the equilibrium potential?
Reply #2.3:At pH = 0 and zero cell potential (U = 0), all elementary steps of oxygen reduction are exothermic for PbS4O2.However, when we shift the chemical potential of the electrons by an applied potential of 0.7 V (corresponding to the equilibrium potential for 2e − oxygen reduction), PbS4O2 exhibits a flat free energy diagram for H2O2 production, suggesting a high catalytic activity with zero overpotential (Figure R9a).At the equilibrium potential of 1.23 V for 4e − oxygen reduction, the PbS4O2 catalyst incorporates two uphill steps in the free energy diagram (Figure R9b), suggesting that the formation of H2O2 is more favored on PbS4O2 under these conditions.It should be noted that this thermodynamic analysis may not perfectly capture the complexity of the experimental results, which can be influenced by various kinetic and surface factors.Additionally, to provide a comprehensive perspective on the catalysts' performance, we included PbO4 and PbS4 as examples in the control group.The effect of various coordination environments on the catalytic performance at the equilibrium potential was examined (Figure R9c-f).At the corresponding equilibrium potentials of 2e -and 4e -ORR pathways, no flat free energy diagram was observed for the control group.These comparisons underscore the unique attributes of PbS4O2 as a catalyst in 2e -ORR electrochemical scenarios.Reply #2.4:To find out whether sulfur leached out during the electrolysis, we conducted ICP-OES analysis after 6 h of continuous electrolysis.According to the element analysis of sulfur (Supplementary Fig. 12), the sulfur content in the Pb SA/OSC was determined to be approximately 2.2 wt%.The ICP-OES analysis revealed that the concentration of leached sulfur after H2O2 production reaction was approximately 0.011 mg L -1 , which accounted for 0.2 % of the total sulfur content only, thereby verifying the exceptional stability of the Pb SA/OSC catalyst.
To address the reviewer's concern, we have incorporated the related descriptions into the revised manuscript (Page 22, Lines 410-411).
5. Figure 4b: The three sites, Pb SA/OSC, Pb SA/OC and OSC all show faradaic efficiencies in the same range.What are the error bars on the measurements?Are DFT calculations qualitatively accurate, especially using a simple limiting potential analysis, to differentiate the relative performance of these sites when their farafaic efficiencies vary by less than 10% at 0.6-0.7 V? Reply #2.5:To accurately assess the faradaic efficiencies of the Pb SA/OSC, Pb SA/OC and OSC catalysts, we conducted multiple parallel tests (Figure R10).Specifically, we selected faradaic efficiencies at 0.3, 0.35, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7 V to calculate the error bars associated with the measurements.The small errors across multiple tests indicate the high reliability and repeatability of the faradaic efficiency measurements.In addition, we agree with the reviewer about the limitations of DFT calculations in providing precise qualitative assessments of faradaic efficiency.DFT calculations are a commonly used tool for qualitatively simulating the properties of molecules and materials.In many cases, DFT calculations are useful for predicting trends, relative stabilities, and electronic structure properties.This enables experimentalists to try and make the material according to the information about the energetics of the chemical reactions (Nat. Comput. Sci., 2022, 2, 539-541).However, it is important to note that the accuracy can vary depending on several factors, including the choice of exchangecorrelation functional, basis set, and the system being studied.The limitation in accuracy becomes more pronounced when dealing with systems with strong electronelectron correlation or long-range interactions.Considering the strengths and limitations of DFT, we have made informed decisions when using it in this work.Specifically, we utilized DFT calculations to determine the relative thermodynamic selectivity of the candidate materials, which in turn informed the experimental adjustments to the S/O ratio for synthesizing Pb-based catalysts.For the coordination structure of Pb site (PbO4 and PbS4), which exclusively consists of either O or S atoms, DFT calculations indicated poor 2e -ORR performance, a finding confirmed by the corresponding control experiments.In contrast, when Pb site's coordination structure incorporates a mixture of O and S atoms (PbS1O3, PbS2O2, PbS3O1, and PbS4O2), the predicted performances depend on the various S/O ratios.Notably, a S/O ratio of 2:1 exhibits optimal performance (PbS4O2), as validated by experiments.Therefore, instead of relying solely on absolute overpotential values and expecting exact alignment with experimentally measured values, we utilized the catalytic trends from DFT calculations and combined with the subsequent material synthesis to achieve high 2e -ORR performance.
6.In the SI, the authors state that the adsorption configurations of O*/OH*/OOH* were determined after searching across all possible configurations.Please report the configurations scanned and the adsorption energies determined in the SI, in addition to reporting the most stable structures.
Reply #2.6:For the Pb site of PbS4O2, we incremented the adsorption angles of ORR intermediates in the horizontal plane from 0° to 360° with a step size of 60° (Figure R11b).Such an exploration of adsorption orientations allowed us to assess the potential adsorption behavior.Started from a range of initial structures, the adsorption geometries for *OOH were relaxed to five orientations (Figure R11c), while the geometries for *OH were optimized to two orientations (Figure R11d).The free energies for the adsorption of 2e -ORR intermediates were calculated and are plotted in Figure R11a, resulting an overpotential ranging from 0.11 to 0.17 V, which is larger than the reported value.We acknowledge the discrepancy in the observed overpotential range.However, while screening the coordination structures of Pb sites, we have concurrently observed that the fluctuation in 2e -ORR overpotential on the PbS4O2 catalyst remained notably smaller than that of other catalysts (0.95 V for PbS4).This result substantiates that the 2e -ORR performance of the PbS4O2 catalyst surpassed those of other screened catalysts.Reply #2.7: Accepting the reviewer's suggestion, we have expanded on the captions describing key aspects of the figures in the revised Supplementary information, which significantly improves overall readability.
8. What are the vibrational frequencies used to determine the ZPE? Were solvation effects included?Please report these aspects for greater reproducibility.

Reply #2.8:
The vibrational frequencies used to determine the Zero-Point Energy (ZPE) were supplemented in Table R1.ZPE is calculated as: ZPE = 0.5hcƩυNA, where h is the Planck's Constant, c is the speed of light in m/s, NA is the Avogadro's Number, Ʃυ is the sum of vibrational frequencies of adsorbates.
The reviewer is correct as the influence of solvent environment on the electrochemical catalysis is important, especially when considering the adsorption configurations at the solid-liquid interface in comparison to a vacuum model.There are two main methods to add solvation within DFT: the explicit approach and the implicit approach (ACS Catal., 2019, 9, 2, 920-931).The explicit approach involves adding water molecules into the system, offering superior descriptions of solvation and field effects, but it comes with significant computational costs for exploring configuration space (J.Chem.Theory Comput., 2019, 15, 12, 6895-6906).In contrast, the implicit approach utilizes the response of a continuum dielectric to model the solvation effect, offering the flexibility to conduct constant potential calculations by adjusting the charge at the interface, but these benefits typically come at the cost of a precise representation of solvation effects and interfacial capacitance (Chem. Rev., 2022, 122, 12, 10777-10820).However, studying surface chemical reactions with solvation effects remains challenging due to the complexity of models and high computational costs (Proc.Natl.Acad.Sci.U. S. A. 2017, 114, 1795-1800).Nevertheless, we have combined the explicit and implicit water models to investigate the effect of solvent molecules on the stability of the synthesized H2O2 at the electrochemical interface .In order to probe the solvent environment across various structures of the solids-liquid interface, we took PbO4 and PbS4 as examples in the control group, with PbS4O2 in the target group.More details can be found in our response in Reply #3.1.
To address the reviewer's concern, we have incorporated Table R1 into the revised Supplementary information as Supplementary Table 5 and the data presented in Figures R12-R14 into the revised Supplementary information (Figures S36-S38).9. Line 188, Main text etc: Please round all adsorption energies calculated using DFT to two decimal places keeping in mind the general numerical accuracies of these simulations.
Reply #2.9: Accepting the reviewer's suggestion, we have updated the related descriptions in the revised manuscript and the Supplementary information (Supplementary Tables 3 and 4).

Response to Reviewer 3's comments
This work demonstrates the synthesis of a Pb SA/OSC catalyst, which exhibited relatively promising H2O2 electrosynthesis activity and selectivity with remarkable stability (100h) in an alkaline electrolyte.The Pb SA/OSC catalyst was applied to a more practical gas diffusion electrode-based reactor and a high current density of 400 mA cm−2 was achieved.This catalyst is interesting and has not been reported for 2-e ORR so far.However, a few technical issues must be fully addressed to meet the high standard of Nature Communications.Detailed comments are listed below: Reply: We highly value the reviewer's insightful comments and valuable suggestions, which have notably improved the quality of our manuscript.In this revised version of the manuscript, we have diligently addressed all the questions and concerns raised by the reviewer.
1. Some recent advancements in theoretical understanding and material innovation have led to the development of a series of efficient main group metal single-atom catalysts for the electrosynthesis of H2O2 in alkaline media (Angew.Chem.Int. Ed.2022, 134, e202117347).In comparison with the activity, the Pb SA/OSC catalyst showed excellent stability in the flow cell.More theoretical discussion is needed to explain this advantage.

Reply #3.1:
The remarkable work (Angew.Chem.Int. Ed., 2022, 134, e202117347) set an important research paradigm for the development of main-group SACs through a combination of theoretical understanding and material innovation.This excellent work has been thoroughly read and cited in our previous manuscript.
In the work, we utilized theoretical simulations to investigate the impact of coordination environment of main-group sites on the electrocatalytic activity.Since SACs with the well-defined M-N4 structure have been extensively demonstrated to exhibit a high activity for various electrochemical reactions, we constructed Pb-X4 models (where X = O, S) using various combinations of coordinated S and O atoms (PbO4, PbS1O3, PbS2O2, PbS3O1, PbS4, PbS3O1-f, PbS4-f).Additionally, it has been reported in the literature that the most common coordination numbers for main-group Pb sites are 4 and 6 (J.Am.Chem. Soc., 2005, 127, 9495-9505;Inorg. Chem., 1998, 37, 1853-1867).Therefore, we also constructed a Pb SACs model with 6 coordination atoms (PbS4O2).Based on the theoretical results, we found that Pb SA/OSC (PbS4O2), co-coordinated with both S and O and with a higher S ratio, is positioned closest to the apex of the limiting potential volcano, exhibiting a high selectivity for H2O2 formation.Inspired by these theoretical results, we endeavored to synthesize the Pb SACs coordinated with different S/O ratios experimentally.Notably, the EXAFS fitting of the optimized Pb SA/OSC catalyst revealed that the coordination numbers of S and O for Pb sites, as well as the distances of Pb-O and Pb-S, closely resemble the PbS4O2 model.Therefore, guided by the DFT calculations, we successfully developed main-group Pb SACs with a high intrinsic H2O2-producing selectivity.
The stable performance of the flow-cell system was demonstrated via the continuous and stable production of H2O2 over 100 h (Fig. 5f).Under realistic conditions, the effective long-term accumulation of H2O2 resulted from a balance between the generation and decomposition of H2O2 molecules at the electrode-liquid interface.Accepting the reviewer's suggestion, we have supplemented theoretical investigations.Here, molecular H2O2 not only interacts with the main-group site of Pb SA/OSC catalyst, but also communicates with the solvent molecules.To probe the complex configurations of H2O2 within a solvation environment, we used the combined explicit-implicit water model to conduct a more in-depth study.Different initial geometries of H2O2 at the solid-liquid interface were constructed and relaxed.To compare the various coordination environments of the catalysts, we took PbO4 and PbS4 as examples in the control group, while PbS4O2 as the target group.
At the surface of PbO4 catalyst, the molecular state of H2O2 was co-adsorbed with two H2O on the Pb site, forming hydrogen bonds with the surrounding water molecules (Figure R12a), resulting in a decrease in the corresponding Pb-O distance from the initial 3.28 Å to 2.56 Å.While the decomposition of H2O2 was more thermally favored (-2.07 eV) with the breaking of the O-O bond, resulting in one adsorbed *OH species and one free •OH species (*H2O2 → *OH + •OH, Figure R12b).The *OH species can be further reduced to generate H2O (Phys.Chem.Chem. Phys., 2013, 15, 148-153).Similarly, at the surface of PbS4 catalyst, H2O2 was also co-adsorbed with two H2O molecules (Figure R12d), resulting in a decrease in the Pb-O distance from the initial 3.07 Å to 2.57 Å.While the decomposition of H2O2 was more favored (-2.13 eV) with the cleavage of the O-O bond, generating one adsorbed *O species and one H2O molecule (*H2O2 → *O + H2O, Figure R12e).Therefore, the activation process of H2O2 on the control samples was favored (Figure R12c and f).As a result, *OH-*H2O species were adsorbed at the PbS4 catalyst and the *O-*2H2O adsorbed at PbO4, respectively.As shown in Figure R13, the charge density enriched at the region between the Pb atom and the dissociated O atom from H2O2, forming a chemical bonding of Pb-O.Integral crystal orbital Hamilton populations (ICOHP) results show that the intensity of the new Pb-O bond when H2O2 was adsorbed on PbO4 (−ICOHP = 0.97) was close to that on PbS4 (−ICOHP = 1.01).In contrast, at the surface of PbS4O2 catalyst, the molecular H2O2 can be stabilized through hydrogen bonding with the water molecules (Figure R14).The O-O distance (1.47 Å) in adsorbed H2O2 was close to the value in molecular H2O2.While the Pb-O distance increased to 4.08 Å from the initial value of 3.28 Å.The energy difference of 0.7 eV between the two geometries of adsorbed H2O2 can be attributed to the variation in their initial solvation structures.Hence, the coordination environments of the Pb SA-based catalysts played a critical role in inhibiting the decomposition of H2O2.
Additionally, the Pb SA/OSC catalyst with the unique super-coordinated structure (PbS4O2) exhibited robust tolerance to S-containing contaminant poisoning, which is advantageous for long-term stability during practical applications (More details can be found in our response in Reply #3.4).Specifically, the Pb SA/OSC catalyst demonstrated a notable capability to stabilize and accumulate H2O2 at the solid-liquid  2. In Figure 1c, the DFT calculations on the Gibbs free energy should be reconducted to consider the related species in alkaline media (OH-and HO2-) instead of those in acid (H2O and H2O2).The authors could refer to some references, such as Phys.Chem. Chem. Phys., 2013, 15, 148-153.Reply #3.2: Accepting the reviewer's suggestion, we have calculated the free-energy diagrams for catalysts in alkaline media.In the recommended reference (Phys.Chem.Chem. Phys., 2013, 15, 148-153) the mechanism of oxygen reduction reaction (ORR) on Co-Nx (x = 2, 4) electrocatalysts in both alkaline and acidic media was investigated.It has proven to be highly beneficial for our work.The complete four-electron reaction in alkaline medium is composed of the following elementary steps: Therefore, the calculated computed free energy diagrams (Figure R15) for the catalysts in alkaline medium were supplemented.At the surface of PbS4O2 catalyst, formation of *OOH remains uphill for U > 0 V.All elementary reaction steps for *OOH to H2O2 (2e -pathway) are downhill by applying low potential (U = 0.2 V), while the generation of H2O (4e -pathway) is downhill at the corresponding equilibrium potential (U = 0.4 V).In contrast, at the surface of PbS4 or PbO4, reduction of *OOH to H2O2 or H2O is uphill processes at high potential.
To address the reviewer's concern, we have updated Figure 1c in the revised manuscript, supplemented Figure R15 into the revised Supplementary information (Supplementary Fig. 4).Besides, we have cited the excellent work (Phys.Chem.Chem. Phys., 2013, 15, 148-153) in the revised manuscript.3. Operando FTIR Spectroscopy of Pb SA/OC and OSC should be offered to prove the adsorption of O2 and OOH on Pb SAs rather than S and O. Also, isotope experiment is required to confirm the origin of OOH.

Reply #3.3:
To investigate the adsorption sites of the prepared catalysts for O2 and OOH, we have applied operando FTIR spectroscopy to examine Pb SA/OC and OSC.As show in Figure R16, potential-dependent absorption bands were observed in the FTIR spectra of both Pb SA/OC and OSC, indicating that the *OOH mediated 2e − ORR pathway on these catalysts.Notably, the absorption bands of Pb SA/OC (1238 cm -1 ) and OSC (1243 cm -1 ), associated with *OOH, exhibited distinct shifts compared to that of Pb SA/OSC (1254 cm -1 ).This observation implies that the adsorption sites for *OOH in Pb SA/OSC were Pb single atoms.On the other hand, the adsorption sites of Pb SA/OC and OSC were different and likely to involve non-metallic S, O or C sites.Indeed, an exemplary study (J.Am.Chem. Soc., 2021, 143, 7819−7827)   To further analyze the origin of the *OOH band, we have conducted an in-situ FTIR experiment using D2O instead of H2O as the solvent.The results of the isotopiclabeling study reveal that the vibration band of *OOH underwent a downshift to 1234 cm -1 in the deuterated medium (Figure R17).Similar shift results have been reported in previous studies regarding operando FTIR spectroscopy of the *OOH intermediate (ACS Catal., 2022, 12, 5345−5355;J. Phys. Chem. B, 2005, 109, 16563−16566).These findings strongly suggest the involvement of hydrogen atoms in the vibration mode at 1254 cm -1 , thus confirming the origin of *OOH.
To address the reviewer's concern, we have added the above discussion into the revised manuscript (Page 18, Lines 328-332), and incorporated Figures R16 and R17 into the revised Supplementary information as Figures S31 and S32.Firstly, we conducted a poisoning experiment by collecting the RDE curve of Pb SA/OSC in the presence of thiocyanate ions.Remarkably, the 2e -ORR activity of Pb SA/OSC remained almost unchanged upon the addition of 0.1 M KSCN (Figure R18), indicating that Pb SA/OSC exhibited a robust tolerance to SCN − poisoning.The slight increase in current observed in the presence of KSCN was attributed to the elevated electrolyte concentration.We hypothesis that the resilience to SCN − was likely associated with the unique super-coordinated structure (PbS4O2) of Pb SA/OSC, where the first coordination was modulated by S and O.Given that the main-group Pb sites were already coordinated by four S atoms, it is not surprising that S-containing SCN − showed negligible effects on the 2e -ORR activity of Pb SA/OSC.Secondly, the DFT calculations reveal that the formation of S and O supercoordinated Pb moieties was the origin of the exceptional 2e -ORR selectivity of Pb SA/OSC.Moreover, based on the RRDE measurements, the Pb SA/OSC exhibited a superior 2e -ORR activity compared to OSC without Pb sites.Therefore, the maingroup Pb atom served as the active site of Pb SA/OSC.
To address the reviewer's concern, we have incorporated the information presented in Figure R18 into the revised Supplementary information (Figure S29). 5. To demonstrate the high stability of the Pb SA/OSC catalyst, some post characterizations are required.

Reply #3.5:
To verify the stability of the catalyst after 100 h of electrolysis, we examined the Pb SA/OSC catalyst with the transmission electron microscopy (TEM) X-ray diffraction (XRD) and Raman.Figure R8 shows the TEM images of the Pb SA/OSC at different magnifications after prolonged electrolysis, indicating the absence of Pb clusters or small PbS species.The XRD patterns (Figure R9) obtained from Pb SA/OSC on GDE exhibited no distinctive peaks associated with crystalline PbS species after electrolysis, in alignment with the observation from the TEM measurements.Additionally, the Raman spectra (Figure R10) of the Pb SA/OSC revealed the characteristic D and G bands of conductive carbon materials, with calculated ID/IG values comparable before and after electrolysis.These findings demonstrate that there were no discernible structural changes in Pb SA/OSC after the stability testing.Therefore, the Pb SA/OSC catalyst exhibited an outstanding stability after 100 h of electrolysis.
To address the reviewer's concern, we have added the related descriptions in the revised manuscript (Page 22, Lines 407-409), and incorporated the data in Figures R5-R7into the revised Supplementary information (Figures S44-S46).6.The structural information was gained from the EXAFS fitting analysis, and the fitting errors must be provided.

Figure R1 :
Figure R1: H2O2 Faradaic efficiencies and JH2O2 of Pb SA/OSC using gas diffusion electrodes in a neutral medium.

Figure R2 :
Figure R2: (a, b) Chronopotentionmetry tests and H2O2 Faradaic efficiencies of the blank GDE and Pb SA/OSC at a current density of 50 mA/cm 2 in a flow-cell electrolyzer.

Figure R3 :
Figure R3: (a) H2O2 Faradaic efficiencies of the blank GDE and Pb SA/OSC at 200 mA/cm 2 in a 1 M KOH solution.(b) Digital image of the blank GDE after 2 h of electrolysis.

Figure R4 :
Figure R4: (a) Quasi-in situ EPR spectra of Pb SA/OSC, Pb SA/OC, and OSC.(b) LSV of Pb SA/OSC in N2-saturated 0.10 M KOH electrolyte containing 1 mM, 5 mM, 10 mM, or 20 mM H2O2.(c) Schematics of the electrochemical cumulative H2O2 production.(The figures displayed here are intended to make our response easier to read, but the figures are already included in the previous manuscript)

Figure R5 :
Figure R5: (a-d) TEM images at different magnifications of Pb SA/OSC obtained from

Figure R6 :
Figure R6: XRD patterns of Pb SA/OSC-GDE before and after the stability test.

Figure R7 :
Figure R7: Raman spectra of Pb SA/OSC-GDE before and after the stability test.The intensity ratios (ID/IG) of the D-band to G-band were calculated.

Figure R8 .
Figure R8.Effect of the C-SOx group on the oxygen reduction reaction.(a) DFToptimized geometry of the C-SOx group on the carbon support, denoted as PbS4O2-SO.(b) Intermediates geometries for the 2e − ORR pathway of PbS4O2-SO.(c) Volcano plots for the 2e -and 4e -ORR on various Pb SA/OSC.

Figure R9 .
Figure R9.Free-energy diagrams of 2e -and 4e -pathways for oxygen reduction on (a, b) PbS4O2, (c, d) PbS4 and (e, f) PbO4 catalysts in acidic medium at zero cell potential (U = 0), at the other potentials (U = 0.2, 0.4, 0.7 and 1.23 V). 4. Does the sulfur leach out?Pb leaching has been reported but what about other elements?Can pre/post characterisations be performed to confirm this point?

Figure R10 :
Figure R10: Multiple faradaic efficiency measurements of Pb SA/OSC, Pb SA/OC and OSC catalysts and faradaic efficiency measurements with error bars.

Figure R11 .
Figure R11.Calculated free energies of adsorption of ORR intermediates on PbS4O2.(b) Rotation states of the initial adsorbates (*OH and *OOH).(c, d) The corresponding DFT-optimized structures of *OH and *OOH, respectively.7. The captions in the SI need more detail so that they can be stand-alone with the figure.Please expand on the captions in the SI describing key aspects of the figure.This change will improve readability.

Figure R12 .
Figure R12.DFT-optimized geometries of H2O2 molecule at the solid-liquid interface.(a-c) Solvated H2O2 and the production of *OH intermediates on the PbO4 catalyst, (df) H2O2 and the *O intermediates at the PbS4 catalyst.

Figure R13 .
Figure R13.Charge transfer behavior and molecular bond strength relationship.The charge density differences for H2O2 adsorbed on (a) PbO4 and (c) PbS4.The yellow color in the representation signifies a high electron density, while the blue color indicates scarce electron density.Electron density difference maps of 0.008 e − /bohr 3 .The crystal orbital Hamilton populations (COHP) of (b) *OH adsorbed on PbO4 and (d)*O adsorbed on PbS4 after H2O2 activation.
interface, as confirmed by the long-term flow-cell experiments.The data presented in Figures R12-R14 have been incorporated into the revised Supplementary information (Figures S36-S38).

Figure R12 .
Figure R12.DFT-optimized geometries of H2O2 molecule at the solid-liquid interface.(a-c) Solvated H2O2 and the production of *OH intermediates on the PbO4 catalyst, (df) H2O2 and the *O intermediates at the PbS4 catalyst.(The Figure secondly shown here is intended to make the Reply easier to read, but the Figure is still numbered as Figure R12 as above)
demonstrated that the C atom adjacent to the coordinated O atom served as the adsorption site (in the case of CoO4 SACs) for *OOH.These analyses indicate that the adsorption sites of Pb SA/OSC for O2 and OOH were the Pb single atoms (Pb SAs).

Figure R17 :
Figure R17: In-situ ATR-SEIRAS spectra collected on the Pb SA/OSC catalyst in an O2-saturated 0.10 M KOH solution and a deuterated medium.4. Thiocyanide (SCN-) poisoning experiment is needed to probe the active site of Pb SA/OSC.Reply #3.4: Thiocyanate ions (SCN − ) exhibit a strong binding affinity to metal sites, facilitating the selective blockage of intermediate adsorption on these metal sites.Therefore, SCN − poisoning experiments are commonly used to identify the active sites of single-atom catalysts.

Figure R5 :Figure R6 :Figure R7 :
Figure R5: (a-d) TEM images at different magnifications of the Pb SA/OSC obtained from the GDE electrode after the stability test.(The figure shown here is intended to make the response easier to read, but the figure is still numbered as Figure R5 as above)

Table R1 .
Vibrational frequencies used to determine the zero-point energy correction (ZPE) for adsorbates (*OOH, *O and *OH) at the Pb sites of various catalysts at T = 298K.