Organic hydrogen peroxide-driven low charge potentials for high-performance lithium-oxygen batteries with carbon cathodes

Reducing the high charge potential is a crucial concern in advancing the performance of lithium-oxygen batteries. Here, for water-containing lithium-oxygen batteries with lithium hydroxide products, we find that a hydrogen peroxide aqueous solution added in the electrolyte can effectively promote the decomposition of lithium hydroxide compounds at the ultralow charge potential on a catalyst-free Ketjen Black-based cathode. Furthermore, for non-aqueous lithium-oxygen batteries with lithium peroxide products, we introduce a urea hydrogen peroxide, chelating hydrogen peroxide without any water in the organic, as an electrolyte additive in lithium-oxygen batteries with a lithium metal anode and succeed in the realization of the low charge potential of ∼3.26 V, which is among the best levels reported. In addition, the undesired water generally accompanying hydrogen peroxide solutions is circumvented to protect the lithium metal anode and ensure good battery cycling stability. Our results should provide illuminating insights into approaches to enhancing lithium-oxygen batteries.

This is an interesting as well as important work in the field of Li-O2 battery and the paper is well written. The authors demonstrate that H2O2 plays a significant role in the large reduction of charge potential for decomposing LiOH in Li-O2 battery without using any other metal catalysts in carbonbased cathode. This is the first report and it does provide insights into the understanding of mechanism on H2O influence for Li-O2 battery. More interestingly, they find an organic chemical (urea hydrogen peroxide) as electrolyte additive in Li-O2 battery to reduce the charge potential by virtue of H2O2 effect. In this case, H2O generally existing in H2O2 solution is avoided not to contaminate Li metal anode. Ultralow charge potential of 3.26 V as well as good cycling performance is obtained. This work shows enough novelty and very enlightening significance. The following points are suggested to improve the article. 1. In Figure 2, the authors presented reduced charge potential for decomposing prefilled LiOH in cathode and dissolved LiOH in electrolyte in the presence of H2O2. The decomposition of the prefilled LiOH in cathode was evidenced by XRD measurement. However, they did not provide data to confirm whether the dissolved LiOH is totally decomposed after charge. 2. SEM analysis should be conducted to confirm the decomposition of prefilled Li2O2 at low charge potential ( Figure S4). And in the reviewer's opinion, the improvement of UH2O2 additive for decomposing the prefilled commercial Li2O2 is important and the results should be moved to the main text, not in the Supplementary Information. 3. In Figure 5c, what do the grey sphere corresponding to? 4. What about the ion conductivity of the electrolyte with UH2O2 additive? Is the ion conductivity changed after UH2O2 addition? 5. In Figure 1a and Figure S4, what is the current density based on the weight of KB in cathode? This value should be clearly given.
Reviewer #2 (Remarks to the Author): The authors report on the deleterious effects of the formation of LiOH in water containing electrolytes in Li-O2 batteries and suggest H2O2-like oxidants as a way to remove LiOH. They show that urea-H2O2 as a way to improve the charge features. The results are impressive; however lacks a key missing piece to complete the story -this is quantitative measurements of pressure rise/decay, DEMS, titration of Li2O2 as a way to quantify the actual discharge and charge electrochemistry. Without this information, the manuscript cannot be considered for publication. 1) The abstract is highly misleading. The authors actually are looking at the formation of parasitic LiOH-type compounds at the anode/cathodes in electrolytes containing water. It gives the reader a false impression that they are potentially considering LiOH as a discharge product. This should be modified in the abstract. Also, in the first line, LiOH should be removed as this once again gives the impression that LiOH is a potential discharge product. 2) Noble metals and oxides (Pt, Au, Pd, Ru and RuO2 etc.),22-27 transition metals and compounds (MnO2,TiC,Ti4O7,54 Cu2O,FeOOH,NiOOH,Ni2CoO4 etc.) were introduced to construct the carbonbased composite cathodes, respectively. -It is important to remember that catalysts should be such that they do not break O-O bond. See ACS Energy Lett. 1, 162. Suitable weak bonding characteristics are required to selectively activate 2e-Li2O2 vs 4e-Li2O. The smileys in Figure 3 and 4 are too gimmicky. The authors should remove this and use an X and tick instead to denote the effects. 3) "Recently, Li et al. in our group demonstrated a novel route to achieve the ultralow charge potential of ~3.2 V by introducing hundred ppm of H2O in the dimethyl sulfoxide (DMSO)-based electrolyte and constructing a composite cathode (MnO2 and Ru particles supported Super P carbon)." It is worth highlighting that several groups (ref. 7, Gasteiger group) demonstrated the role of water in enhancing capacity, however, they suffered from larger overpotentials and this was addressed by Li et al. 4) How confident are the authors of the reaction mechanism of the LiOH mediated oxidation? In light of all the issues relating to the LiOH chemistry, it is worthwhile for the authors to remind the readers that the chemistry is still 2e-reduction of oxidation in the work of Li et al. Perhaps, even downplay the LiOH. 5) Page 4 -challenged the reversibility of Li-O2 battery chemistry -add "with LiOH as the discharge product" 6) Figure 1 -corresponding pressure rise/decay measurements need to be done to extract e-/O2. Without quantitative measurements, it is nearly impossible to conclude anything about the parasitic electrochemistry and chemistry. 7) The results of Figure 5 are impressive, but authors must quantify the pressure rise/decay to quantify e-/O2. See the work from McCloskey et al. This is a crucial aspect that needs to be addressed before these results could be used to conclude enhanced rechargeability.

Reviewer #3 (Remarks to the Author):
This manuscript claims a lowering of the charge overpotential of a lithium-oxygen battery by adding H2O2 to the electrolyte. Particularly, the authors state that H2O2 can reversibly decompose LiOH in a lithium-oxygen battery (either LiOH formed from letting the discharge product Li2O2 react with ambient water, or LiOH prepacked in the cathode), and that an anhydrous urea-H2O2 enables reversible Li2O2 formation/decomposition at this low charge overpotential without the deleterious effects of H2O. The use of H2O2 as an electrolyte additive is novel, to my knowledge. The supporting information and methods descriptions are sufficient to reproduce the work, other than electrolyte volume, as mentioned below. The authors clearly show that electrolytes containing H2O2 and urea-H2O2 see lower overpotentials on charge, however, they do not provide sufficient evidence to ascribe this lower charge overpotential to the same reversible lithium-oxygen electrochemistry that occurs without their additives. Their evidence that a reversible Li-O2 electrochemistry is occurring is spectroscopic signatures (or the lack thereof) of the discharge products before and after charging. A lack of a signal for LiOH on a cathode after charging, for example, could also be ascribed to a parasitic reaction having formed a soluble species. In determining reversibility of a lithium-oxygen battery, it is important to quantitatively characterize the consumption and evolution of oxygen and the formation and decomposition of the discharge product, Li2O2. Without quantitative gas analysis and titrations for the discharge products proving that a 2e-/O2 process is maintained on both discharge and charge, a lowering of the charge overpotential and a disappearance of spectroscopic signals could similarly be ascribed to new irreversible side reactions, rather than the desired reversible Li-O2 electrochemistry. Of note, showing multiple cycles (50 cycles) with similar performance could simply point toward steady decomposition in a large electrolyte volume (electrolyte volume was not given), rather than steady high performance. Thus, I believe this article needs additional evidence to support the claim that the observed low charge overpotential indeed corresponds to a reversible lithium-oxygen electrochemistry. Additionally, the article does not give a proposed mechanism for the performance-enhancing effect of H2O2 in a lithium-oxygen battery. Given that H2O2 is already present when Li2O2 is reacted with ambient water to form LiOH, I am skeptical as to the proposed benefit of additional H2O2, as well as its stability, and thus would appreciate a proposed method of action of the H2O2. I also believe a proposed mechanism, whether it is a discussion of a catalytic pathway or solubility enhancements, etc., is critical for this article to influence a wider field, as it could point to broader design principles for electrolytes in electrochemical systems. For these reasons, I do not believe this manuscript should be published in its current form.

Reviewers' comments:
Reviewer #1 (Remarks to the Author): Authors have addressed all the comments. I would support for the publication of this work, which brings innovation thinking to the current Li O2 battery research.
Reviewer #2 (Remarks to the Author): The authors have done an excellent job of addressing the reviewer concerns including the new experiments that were carried out to address the concerns regarding yield and rechargeabilty. In light of this, I think the manuscript is now suitable for publication.
Reviewer #3 (Remarks to the Author): The authors have made significant helpful revisions to their paper, but I believe there are still two points from the original reviews that have not been adequately addressed: 1) Confusion as to when LiOH is the discharge product being oxidized on charge and when Li2O2 is the discharge product being oxidized on charge.
My understanding is that the primary goal of this paper, according to the abstract and the concluding paragraph, is to report if adding hydrogen peroxide (and in particular an anhydrous urea-hydrogen peroxide) to a lithium-oxygen battery can enable the reversible oxidation of lithium hydroxide. This work is seen in Figures 1-4. According to the introduction, it seems this goal is an extension of the authors' group's recent work on the promotion of LiOH as the primary discharge product by using MnO2 and Ru in their cathode and including H2O in their electrolyte. In addition, the authors present whether an anhydrous urea-hydrogen peroxide can serve as a helpful additive in oxidizing Li2O2, as seen in Figures 5-7. I think the authors still need to be clearer in both the abstract and the text (particularly introduction and conclusion) that they are discussing at times a LiOH-based lithiumoxygen electrochemistry and at times a Li2O2-based electrochemistry. I think both ideas are interesting and important to present; I just believe the authors should more clearly separate the two, especially as they only have evidence of a reversible, oxygen-forming oxidation reaction when Li2O2 is the discharge product being oxidized on charge.
2) The critical DEMS results to support reversible LiOH oxidation are still missing, and trying to approximate DEMS results with titration results is incorrect.
In our first reviews, another reviewer and I asked that differential electrochemical mass spectrometry (DEMS) be employed to verify that on charge H2O2 is indeed reversibly oxidizing LiOH to oxygen. Here, the authors provide titration results of extracted cathodes to show that LiOH, present before charge, is absent after charge. While this additional data is useful, the conclusion that this data can serve as a proxy for DEMS is incorrect.
Titration results and DEMS results provide two different pieces of information -the formation and decomposition of Li2O2 and the formation and decomposition of oxygen, respectively. For example, if more Li2O2 has been consumed after charge (as measured via titration) than O2 has been evolved (as measured via DEMS), this implies that some Li2O2 has been oxidized to parasitic products. This idea was recently presented in studying the oxidation of LiOH in Burke et al., ACS Energy Letters, 2016, 1, 747-756. In that article, the authors looked at the reversibility of I-and H2O-induced LiOH formation in lithium-oxygen batteries. Figure 3(c-d) show that a battery with a certain electrolyte exhibits the decomposition of nearly all LiOH on charge. While this seems to imply reversibility, Figure 2(g-i) clearly show that no oxygen is evolved when a battery with that certain electrolyte is charged. Thus, Burke et al. shows that the absence of LiOH after charge is not sufficient evidence to claim reversible oxidation of LiOH.
In the manuscript at hand, I still feel the authors have failed to corroborate the statement in the abstract that they, "find that the hydrogen peroxide (H2O2) aqueous solution added in the electrolyte can effectively promote the decomposition of LiOH compounds at the ultralow charge potential on the catalyst-free Ketjen Black-based cathode," as the authors have indeed proven that LiOH decomposes, but have not yet proven that LiOH decomposes reversibly to oxygen, as I believe is implied.
In the rebuttal letter, the authors expressed the difficulty in measuring oxygen evolution from a cell containing hydrated H2O2, as there is a rapid reaction between water and the lithium metal anode that produces hydrogen. I understand this difficulty, and recommend that if the authors cannot perform accurate DEMS to check for O2 evolution from LiOH during charge, that the authors note that while LiOH has been decomposed on charge, it is unclear if it has been reversibly oxidized to oxygen. I believe this article is still an important addition to the field with this acknowledgment.
The DEMS results presented in Figure 6 do indeed show reasonable reversibility of Li2O2 in a cell containing urea-H2O2, and I appreciate this addition. This is an interesting as well as important work in the field of Li-O 2 battery and the paper is well written. The authors demonstrate that H 2 O 2 plays a significant role in the large reduction of charge potential for decomposing LiOH in Li-O 2 battery without using any other metal catalysts in carbon-based cathode. This is the first report and it does provide insights into the understanding of mechanism on H 2 O influence for Li-O 2 battery. More interestingly, they find an organic chemical (urea hydrogen peroxide) as electrolyte additive in Li-O 2 battery to reduce the charge potential by virtue of H 2 O 2 effect. In this case, H 2 O generally existing in H 2 O 2 solution is avoided not to contaminate Li metal anode. Ultralow charge potential of 3.26 V as well as good cycling performance is obtained. This work shows enough novelty and very enlightening significance.

Response to Reviewers:
The following points are suggested to improve the article.
1. In Figure 2, the authors presented reduced charge potential for decomposing prefilled LiOH in cathode and dissolved LiOH in electrolyte in the presence of H2O2.
The decomposition of the prefilled LiOH in cathode was evidenced by XRD measurement. However, they did not provide data to confirm whether the dissolved LiOH is totally decomposed after charge.
As the reviewer suggested, we conducted Fourier Transform Infrared Spectroscopy (FTIR) to confirm whether the dissolved LiOH could be totally decomposed after charge. The related data is presented in Supplementary Fig. S3 and corresponding revision has been made in manuscript (Page 8).
After charge at high potential in G4-H 2 O-LiOH (l) electrolyte (Fig. 2b), the peak of LiOH at ~3680 cm -1 in FTIR spectra ( Supplementary Fig. S3, iii) still remains, indicating the uncomplete decomposition of liquid LiOH in electrolyte. In contrast, the LiOH can be fully decomposed after charge at low potential in H 2 O 2 -containing electrolyte (Fig. 2b), evidenced from the disappeared peak ( Supplementary Fig. S3, iv). 2. SEM analysis should be conducted to confirm the decomposition of prefilled Li 2 O 2 at low charge potential ( Figure S4). And in the reviewer's opinion, the improvement of  4. What about the ion conductivity of the electrolyte with UH 2 O 2 additive? Is the ion conductivity changed after UH 2 O 2 addition?
In order to evaluate the ion conductivity of the electrolyte with UH 2 O 2 additive, we carried out the electrochemical impedance spectra. The related data is shown in Corresponding revision has been made in Page 13 in manuscript. Figure 1a and Figure S4, what is the current density based on the weight of KB in cathode? This value should be clearly given.

In
As the reviewer indicated, we have given the current density based on the weight of KB in cathode in Figure 1a and Figure S4 ( Figure S8 at present). In Figure 1a and Figure S4, the current densities are 100 mA g -1 based on the weight of KB. In other figures ( Figure 2 and Figure 3), we also added the corresponding current densities.
Reviewer #2: The authors report on the deleterious effects of the formation of LiOH in water containing electrolytes in Li-O 2 batteries and suggest H 2 O 2 -like oxidants as a way to remove LiOH. They show that urea-H 2 O 2 as a way to improve the charge features.
The results are impressive; however, lacks a key missing piece to complete the story -this is quantitative measurements of pressure rise/decay, DEMS, titration of Li 2 O 2 as a way to quantify the actual discharge and charge electrochemistry. Without this information, the manuscript cannot be considered for publication.
Thank you for the helpful suggestions. It is important for further improvement of our work. As the reviewer's suggestions, we have conducted the quantitative measurements of in situ DEMS and titration of Li 2 O 2 . Also, in the manuscript, corresponding revisions have been made.
1) The abstract is highly misleading. The authors actually are looking at the formation of parasitic LiOH-type compounds at the anode/cathodes in electrolytes containing water. It gives the reader a false impression that they are potentially considering LiOH as a discharge product. This should be modified in the abstract. Also, in the first line, LiOH should be removed as this once again gives the impression that LiOH is a potential discharge product.
LiOH in the first line has been deleted and the abstract has been revised to avoid the misleading information. In some reported work related to 3) The smileys in Figure 3 and 4 are too gimmicky. The authors should remove this and use an X and tick instead to denote the effects.
The smileys in Figure 3 and 4 have been substituted with an X and tick. should be performed to figure out the detailed mechanism.
As the reviewer suggested, we added the description about the 2edischarge process to remind the readers (Page 4).  The KB-based cathodes in situ loading solid LiOH compounds were obtained by firstly discharging Li-O 2 pouch cells in the dry electrolyte to 1.50 mAh (Fig. 1a i) with Li 2 O 2 as the products evidenced from the X-ray diffraction (XRD) pattern in Fig. 1c, extracting the discharged cathodes in an Ar glove box and leaving the cathodes in an Ar atmosphere with a relative humidity of 75% for 7 days. The XRD pattern in Fig. 1d confirms all the Li 2 O 2 on the discharged cathode is converted to the mixture of LiOH As shown in Fig. 1b, at each point during charge process, the calculated amount of LiOH compounds consumed is almost equal to the theoretical value and the relationship of consumed amount and the charge capacity is linearly dependent.
These quantitative results indicate the oxidation of LiOH compounds dominates in the charge process and there is no obvious parasitic electrochemistry and chemistry.
Corresponding revisions have been made in Page 6-7 in manuscript. to the different state (i, ii, iii and iv) in Figure 1a. The Li-O 2 pouch cell is firstly discharged to 1.5 mAh (corresponding to ~4000 mAh g -1 KB and 3.5 mAh cm -2 ) in the dry electrolyte to produce Li 2 O 2 in cathode (c). The Li 2 O 2 is converted to the mixture of LiOH and LiOH·H 2 O (d) by keeping the cathode (i) in the Ar atmosphere with a relative humidity of 75% for 7 days.  The authors clearly show that electrolytes containing H 2 O 2 and urea-H 2 O 2 see lower overpotentials on charge, however, they do not provide sufficient evidence to ascribe this lower charge overpotential to the same reversible lithium-oxygen electrochemistry that occurs without their additives. Their evidence that a reversible Li-O 2 electrochemistry is occurring is spectroscopic signatures (or the lack thereof) of the discharge products before and after charging. A lack of a signal for LiOH on a cathode after charging, for example, could also be ascribed to a parasitic reaction having formed a soluble species.
In order to confirm whether the lower charge overpotential in the presence of H 2 O 2 in electrolyte is attributed to the same reversible lithium-oxygen electrochemistry that occurs without the additives, additional evidences were carried out including quantitative titration experiment to examine the charge process and FTIR analysis of electrolyte after charge to examine whether there is soluble species from parasitic reaction.
Quantitative titration experiments (refer to the work of McCloskey et al. in J. Phys. Chem. Lett. 2013, 4, 2989 corresponding to the LiOH oxidation in the H 2 O 2containing electrolyte in Fig. 1, were conducted to determine the consumed amount of LiOH in varied charge states. Fig. 1b showed the titration results.  Figure 1a. The Li-O 2 pouch cell is firstly discharged to 1.5 mAh (corresponding to ~4000 mAh g -1 KB and 3.5 mAh cm -2 ) in the dry electrolyte to The KB-based cathodes in situ loading solid LiOH compounds were obtained by firstly discharging Li-O 2 pouch cells in the dry electrolyte to 1.50 mAh (Fig. 1a i) with Li 2 O 2 as the products evidenced from the X-ray diffraction (XRD) pattern in Fig. 1c, extracting the discharged cathodes in an Ar glove box and leaving the cathodes in an Ar atmosphere with a relative humidity of 75% for 7 days. The XRD pattern in Fig. 1d confirms all the Li 2 O 2 on the discharged cathode is converted to the mixture of LiOH Based on these results, it can be concluded that the low charge potential in electrolyte with H 2 O 2 additive corresponds to the oxidation decomposition of LiOH compounds.
Corresponding revisions have been made in Page 6-7 in manuscript.

Supplementary Fig. 1 丨 FTIR spectra of (i) pristine G4-H 2 O-H 2 O 2 electrolyte and (ii) G4-H 2 O-H 2 O 2 electrolyte after charge, respectively.
In determining reversibility of a lithium-oxygen battery, it is important to quantitatively characterize the consumption and evolution of oxygen and the formation and decomposition of the discharge product, Li 2 O 2 . Without quantitative gas analysis and titrations for the discharge products proving that a 2e -/O 2 process is maintained on both discharge and charge, a lowering of the charge overpotential and a disappearance of spectroscopic signals could similarly be ascribed to new irreversible side reactions, rather than the desired reversible Li-O 2 electrochemistry.
As the reviewer suggested, according to the work from McCloskey et al. (J. Phys. Chem. Lett. 2013, 4, 2989, we carried out the in situ DEMS measurement to observe the pressure rise/decay and titration experiment to quantify einvolved in the reaction. Because the preliminary custom-built cell for DEMS measurement in the lab has too large space volume for gas storage (~10 mL, nearly 80 times than that of McCloskey in J. Phys. Chem. Lett. 2011Lett. , 2, 1161) and the defective design makes that the evolved gas during charge is unable to be completely transferred to MS for quantifying gas and calculating e-/O 2 , instead, we used titration to evaluate the e -/Li 2 O 2 and combined DEMS for qualitative gas analysis. The value of e -/O 2 was confirmed to be in linear proportion to the value of e -/Li 2 O 2 when the overpotentials held at low values in the article of McCloskey. Accordingly, the value of e -/O 2 can be estimated. The corresponding data was provided in Fig. 6 in manuscript. After discharging the cell to 0.1, 0.2, 0.3, 0.4 and 0.5 mAh at 0.05 mA, respectively, the quantify of Li 2 O 2 formation was determined by titration method (Fig. 6c). The yields of the formed Li 2 O 2 (the amount of Li 2 O 2 titrated divided by the amount of Li 2 O 2 expected given the Coulometry) hold at ~ 91%. This value is similar to the yield of McCloskey and indicates a little possible side reactions corresponding to electrolyte or cathode decomposition. The value of e -/Li 2 O 2 can be estimated to be 2.15, close to the theoretical value of 2. Upon charge, the gas evolution rate of O 2 and CO 2 are presented in Fig. 6b. It can be seen that O 2 evolution was detected from the beginning and continued along the charge potential of ~3.3 V during the whole charge process. No evolution of CO 2 was detected. The amounts of Li 2 O 2 consumed during charge were analyzed by titration measurement (Fig. 6d). Li 2 O 2 oxidation follows a ~2.14e -/Li 2 O 2 process during the whole charge process. This value approaches to the ideal value of 2 and clearly demonstrates that the significant reduction of charge overpotential due to the UH 2 O 2 additive in electrolyte can prohibit the high charge potential-induced side reaction and the introduction of UH 2 O 2 can improve the charge performance. The DEMS and titration results suggest that the charge reactions are dominated by evolution of O 2 and consumption of Li 2 O 2 . Therefore, the high rechargeability of the Li-O 2 cell with G4-UH 2 O 2 electrolyte can be concluded.
Corresponding revisions have been made in Page 14-15 in manuscript. Of note, showing multiple cycles (50 cycles) with similar performance could simply point toward steady decomposition in a large electrolyte volume (electrolyte volume was not given), rather than steady high performance. Thus, I believe this article needs additional evidence to support the claim that the observed low charge overpotential indeed corresponds to a reversible lithium-oxygen electrochemistry.
The electrolyte volume added in the 2030-coin cell is 50 µL. This is a general value in most of the reported work and is much less than that (0.2 -1 mL) in Science, 2015, 350, 530. The electrolyte volume has been added in the part of Cell assembly in manuscript.
With regard to the reversibility of this Li-O 2 cell with G4-UH 2 O 2 electrolyte, the in situ DEMS and titration results (Fig. 6) confirmed the discharge and charge reactions corresponded to the consumption and evolution of O 2 and the formation and oxidation of Li 2 O 2 . Accordingly, the observed low charge overpotential corresponds to a reversible Li-O 2 electrochemistry.
The electrolytes before and after discharge/charge process were analyzed by FTIR analysis to confirm whether there was severe electrolyte decomposition. The results were presented in Supplementary Fig. S8. Compared with the pristine G4-UH 2 O 2 electrolyte ( Supplementary Fig. S8 iii), there is nearly no change in FTIR spectra for G4-UH 2 O 2 electrolytes after discharge (iv) and charge (v), indicating no obvious decomposition of electrolyte occurred. Therefore, it is rational to ascribe the steady high performance of Li-O 2 cell to the introduction of UH 2 O 2 additive in electrolyte.
Corresponding descriptions have been made in Page 17 in manuscript. Additionally, the article does not give a proposed mechanism for the performanceenhancing effect of H 2 O 2 in a lithium-oxygen battery. Given that H 2 O 2 is already present when Li 2 O 2 is reacted with ambient water to form LiOH, I am skeptical as to the proposed benefit of additional H 2 O 2 , as well as its stability, and thus would appreciate a proposed method of action of the H 2 O 2 . I also believe a proposed mechanism, whether it is a discussion of a catalytic pathway or solubility enhancements, etc., is critical for this article to influence a wider field, as it could point to broader design principles for electrolytes in electrochemical systems.
For these reasons, I do not believe this manuscript should be published in its current form. LiOH becomes the main discharge products. Upon charge, the potential increases to above ~3.8 V ( Figure Respond 1). Thus, the important role of H 2 O 2 is emphasized.
Concerning the detailed mechanism, we are making many efforts to explore a clear understanding and it is expected to be figured out in our next work. The stability of UH 2 O 2 additive is evaluated through a storage experiment. One cell was firstly discharged and then rested for 7 days. Another cell was firstly performed for one discharge-charge cycle and then rested for 7 days. The open circuit voltages (OCVs) were monitored ( Supplementary Fig. 7). It can be seen that during these 7 days, the OCVs shows good stability and after rest, the Li metal anodes still keep uncontaminated by H 2 O, the possible decomposition product of UH 2 O 2 . In addition, the Li-O 2 cell performs stable discharge-charge ability for 50 cycles (Fig. 7) and there is no obvious change of Li metal anode after cycles (Fig. 4). Therefore, the stability of the UH 2 O 2 additive can be confirmed. Supplementary Fig. 7. OCVs trend of Li-O 2 cells with UH 2 O 2 additive in electrolyte after a discharge process and one discharge-charge cycle, respectively. Insets are the photos of Li metal anodes after rest.
Corresponding revisions have been made in Page 13 in manuscript.

Response to Reviewers:
Reviewer #1: Authors have addressed all the comments. I would support for the publication of this work, which brings innovation thinking to the current Li-O 2 battery research.
Thank you for your appreciation.
Reviewer #2: The authors have done an excellent job of addressing the reviewer concerns including the new experiments that were carried out to address the concerns regarding yield and rechargeabilty. In light of this, I think the manuscript is now suitable for publication.
Thank you for affirming our work.
Reviewer #3: The authors have made significant helpful revisions to their paper, but I believe there are still two points from the original reviews that have not been adequately addressed: 1) Confusion as to when LiOH is the discharge product being oxidized on charge and when Li 2 O 2 is the discharge product being oxidized on charge.
My understanding is that the primary goal of this paper, according to the abstract and the concluding paragraph, is to report if adding hydrogen peroxide (and in particular an anhydrous urea-hydrogen peroxide) to a lithium-oxygen battery can enable the reversible oxidation of lithium hydroxide. This work is seen in Figures 1-4. According to the introduction, it seems this goal is an extension of the authors' group's recent work on the promotion of LiOH as the primary discharge product by using MnO 2 and Ru in their cathode and including H 2 O in their electrolyte. In addition, the authors present whether an anhydrous urea-hydrogen peroxide can serve as a helpful additive in oxidizing Li 2 O 2 , as seen in Figures 5-7. I think the authors still need to be clearer in both the abstract and the text (particularly introduction and conclusion) that they are discussing at times a LiOH-based lithium-oxygen electrochemistry and at times a Li 2 O 2 -based electrochemistry. I think both ideas are interesting and important to present; I just believe the authors should more clearly separate the two, especially as they only have evidence of a reversible, oxygen-forming oxidation reaction when Li 2 O 2 is the discharge product being oxidized on charge.
Thank you for the guidance about the expression in abstract, introduction and conclusion.
As the reviewer's understanding, we initially aimed to achieve low charge potential and simultaneously avoid the utilization of noble metal Ru and transition metal compound MnO 2 in the H 2 O-containing Li-O 2 battery system with LiOH as product (recently reported by our group). Thus, H 2 O 2 was firstly found to efficiently facilitate the decomposition of LiOH at low charge potential. largely reduced performance and safety issues (Fig. 3 in manuscript). Then, an organic urea hydrogen peroxide without H 2 O contamination was developed and confirmed to play the expected role. Ultralow charge potential of ~3.26 V and good cycling stability were finally obtained.
We agree with the opinion the reviewer indicated, that the results of either the promoted decomposition of LiOH at low charge potential by the action of H 2 O 2 or the reduced charge potential as well as the improved performance of Li-O 2 battery with Li 2 O 2 products in the presence of organic urea hydrogen peroxide are interesting and important.
Based on our newly-added DEMS evidence of O 2 generation together with the reversible LiOH oxidation (Figure 1f-g, corresponding discussion is shown in next reply) and according to the reviewer's suggestion, we revised the related parts including abstract, introduction and conclusion in order to clearly manifest our results.
Corresponding revisions are shown below.

Abstract
Reducing the high charge potential is the crucial concern to advance the 2) The critical DEMS results to support reversible LiOH oxidation are still missing, and trying to approximate DEMS results with titration results is incorrect.
A) In our first reviews, another reviewer and I asked that differential electrochemical mass spectrometry (DEMS) be employed to verify that on charge H 2 O 2 is indeed reversibly oxidizing LiOH to oxygen. Here, the authors provide titration results of extracted cathodes to show that LiOH, present before charge, is absent after charge. While this additional data is useful, the conclusion that this data can serve as a proxy for DEMS is incorrect.
Titration results and DEMS results provide two different pieces of information -the formation and decomposition of Li 2 O 2 and the formation and decomposition of oxygen, respectively. For example, if more Li 2 O 2 has been consumed after charge (as measured via titration) than O 2 has been evolved (as measured via DEMS), this implies that some Li 2 O 2 has been oxidized to parasitic products. This idea was recently presented in studying the oxidation of LiOH in Burke et al., ACS Energy Letters, 2016, 1, 747-756. In that article, the authors looked at the reversibility of I-and H 2 O-induced LiOH formation in lithium-oxygen batteries. Figure 3(c-d) show that a battery with a certain electrolyte exhibits the decomposition of nearly all LiOH on charge. While this seems to imply reversibility, Figure 2(g-i) clearly show that no oxygen is evolved when a battery with that certain electrolyte is charged. Thus, Burke et al. shows that the absence of LiOH after charge is not sufficient evidence to claim reversible oxidation of LiOH.
Thank you for reminding us of the difference and relationship between titration and DEMS. We agree with the viewpoint that the Li 2 O 2 or LiOH consumption measured via titration does not mean the corresponding evolution of O 2 from Li 2 O 2 or LiOH decomposition. As the reviewer referred, although Figure 3  We added the DEMS measurement to further improve our work ( Fig. 1f-g, see the next reply). B) In the manuscript at hand, I still feel the authors have failed to corroborate the statement in the abstract that they, "find that the hydrogen peroxide (H 2 O 2 ) aqueous solution added in the electrolyte can effectively promote the decomposition of LiOH compounds at the ultralow charge potential on the catalyst-free Ketjen Black-based cathode," as the authors have indeed proven that LiOH decomposes, but have not yet proven that LiOH decomposes reversibly to oxygen, as I believe is implied.
In the rebuttal letter, the authors expressed the difficulty in measuring oxygen evolution from a cell containing hydrated H 2 O 2 , as there is a rapid reaction between water and the lithium metal anode that produces hydrogen. I understand this difficulty, and recommend that if the authors cannot perform accurate DEMS to check for O 2 evolution from LiOH during charge, that the authors note that while LiOH has been decomposed on charge, it is unclear if it has been reversibly oxidized to oxygen. I believe this article is still an important addition to the field with this acknowledgment.
Thank you for appreciating the importance of our work, even in the previous case that there was no DEMS evidence of O 2 evolution from LiOH decomposition at low charge potential in the presence of H 2 O 2 .
At this time, we finally fabricated a pouch cell as small as possible and successfully put it into the custom-built device for DEMS analysis for investigating whether O 2 is released during charge in the Li-O 2 pouch cell with in situ formed LiOH/KB cathode and H 2 O 2 -containing electrolyte. As the pouch cell was introduced in the main text (Page 5), "on the anode side, the Li metal is protected by utilizing the Li ion conducting glass-ceramic film (LiSICON, Li 2 O-Al 2 O 3 -SiO 2 -P 2 O 5 -TiO 2 -GeO 2 , Ohara Corporation), only allowing the transport of Li + and preventing the penetration of others. On the cathode side, a KB-based cathode and a glass fiber infiltrating electrolyte are constructed".
In this case, the rapid reaction between water and the lithium metal anode during DEMS measurement can be avoided. Figure 1f-g shows the DEMS results.
It can be seen that obvious O 2 evolution was detected at ~3.50 V during charge process and there was no evolution of CO 2 . Combined with the previous titration results, we confirmed LiOH could decompose reversibly to O 2 . Then, our statement in the abstract "find that the hydrogen peroxide (H 2 O 2 ) 8 / 8 aqueous solution added in the electrolyte can effectively promote the decomposition of LiOH compounds at the ultralow charge potential on the catalyst-free Ketjen Black-based cathode," should be rational. 3) The DEMS results presented in Figure 6 do indeed show reasonable reversibility of Li 2 O 2 in a cell containing urea-H 2 O 2 , and I appreciate this addition.
Thank you for appreciating our effort.