Reconstruction of zinc-metal battery solvation structures operating from −50 ~ +100 °C

Serious solvation effect of zinc ions has been considered as the cause of the severe side reactions (hydrogen evolution, passivation, dendrites, and etc.) of aqueous zinc metal batteries. Even though the regulation of cationic solvation structure has been widely studied, effects of the anionic solvation structures on the zinc metal were rarely examined. Herein, co-reconstruction of anionic and cationic solvation structures was realized through constructing a new multi-component electrolyte (Zn(BF4)2-glycerol-boric acid-chitosan-polyacrylamide, simplified as ZGBCP), which incorporates double crosslinking network via the esterification, protonation and polymerization reactions, thereby combining multiple advantages of ‘liquid-like’ high conductivity, ‘gel-like’ robust interface, and ‘solid-like’ high Zn2+ transfer number. Based on the ZGBCP electrolyte, the Zn anodes achieve record-low polarization and stable cycling. Furthermore, the ZGBCP electrolyte renders the AZMBs ultrawide working temperature (−50 °C ~ +100 °C) and ultralong cycle life (30000 cycles), which further validates the feasibility of the dual solvation structure strategy and provides a innovative perspective for the development of high-performance AZMBs.

The paper titled "Co-reconstruction of anionic and cationic solvation structures for ultra-stable operating of zinc-metal batteries under -50~+100oC" by authors L. Yao, J. Liu, F. Zhang, B. Wen, X. Chi and Y. Liu deals with the regulation of cationic/anionic solvation structures in aqueous zinc metal batteries by using multicomponent electrolyte, consisted of Zn(BF4)2, glycerol, boric acid, chitosan and polyacrylamide (ZGBCP).The authors presented a detailed experimental and computational investigation of solvation structure and a comprehensive understanding of the mechanism in proposed battery devices.Also, the performances of proposed water-based zinc batteries are promising.All the experiments are well-thought-out and well-designed resulting in a coherent story even though the investigated system is complicated.Overall, the article is interesting, actual, and novel.
Recommendation: Major revision Below are some comments, hope will help to modify this paper: Major remarks: 1)DFT calculations.The use of DFT calculations to explain the coordination of Zn2+ and BF4-in terms of binding energies and bond order is a great idea.Is there any particular reason why chitosan is excluded from the discussion and calculation of binding energies?From Scheme 1g, chitosan is proposed to be one of the coordination sites for Zn-ion, along with acrylate, which is a reasonable assumption due to the existence of -OH and -NH groups in the chitosan structure.However, there is no data for Zn2+-chitosan binding energies from DFT calculation which is contradictive.Also, does BF4-have any interactions with chitosane or PAM based on DFT calculations?Please, revise the DFT models and change the proposed solvation model accordingly.
2) MD simulations.It is difficult to analyze presented RDFs data from MD simulations and to evaluate the validity of the used force field.Authors should provide the number of molecules (in Supporting information) and the size of the simulation box used in MD simulations during the production run.Did the authors perform any validation of the used force field?I will strongly recommend showing the obtained data in the form of coordination number since presented RDFs (Figure 2e) are impossible to analyze and therefore the corresponding discussion is confusing.Also, please note that chitosan is again missing like in the DFT calculations, so the proposed discussion is ambiguous and needs to be revised accordingly.
3) Please, provide the procedure for treatment of the electrode before post-mortem analysis.SEI formation could be easily influenced by post-treatment of the electrode, and it is necessary to include a detailed handling procedure of the electrode.

Response to Reviewers' Comments
Reviewer #1 (Remarks to the Author): This work utilizes a multi-component electrolyte with Zn(BF4)2-glycerol-boric acidchitosan-polyacrylamide and achieves good performance over a wide temperature range of -50 to 100 °C.A new solvation view is proposed, highlighting reconstructing the anionic and cationic solvation structures, which contribute to high ionic conductivity and cation transfer numbers.The proposed structure is interesting, but comparisons with similar works in gel or liquid electrolytes are lacking.The complexity of the electrolyte composition (with over 3 components except salt) may lead to results or discussions that do not fully explain the real case.Here are some concerns and suggestions for improvement: Our response: We appreciate the reviewer for recognizing the importance and novelty of our work and constructive revision advice.We have addressed the reviewer's comments point by point.

The claim in sentence 80 about the robust interphase of the ZGBCP electrolyte and the dual crosslinking network involving esterification, protonation, and polymerization
reactions needs further explanation to clarify how these reactions contribute to the function.
Our response: Thank you very much for your insightful comments and profound questions.First, esterification between glycerol and boric acid can enhance polymer mobility effectively 1,2 and then promote the infiltration of ZGBCP into the electrode interface considerably; Second, protonation of chitosan can render the ZGBCP superb adhesion capability, thereby generating extremely intimate interfacial contact; Third, polymerization of acrylamide can provide a robust framework for both the esterification and the protonation and increase the crosslinking density, ultimately leading to a multicomponent and multi-function crosslinking network, which contributes the stable interface between the ZGBCP electrolyte and the electrodes.More detailed information is offered to further elaborate the three parts as follows: The adhesion properties of the PAM/CS dual crosslinking network mainly stem from the interactions between functional groups such as -CONH2, -OH, and -NH2 and the material surface, such as hydrogen bonding and coordination, as previously reported 3 .However, when serving under extreme temperature conditions, further improvement of interfacial contact is required to ensure sufficient infiltration and high reversibility of electrode reactions.We believe that the challenge of further enhancing its adhesion strength lies in the limited dynamic diffusion of these functional groups due to slow chain segment motion, making it difficult to form a dynamically and consistently robust adhesion 4 .
To address this, we introduced the dissociation of protons by the reversible reaction process between boric acid and glycerol to dissolve CS, thus constructing a PAM/CS dual crosslinking network.Rheological test results indicate that the energy storage modulus G' of the double crosslinked network GBCP prepared by plasticization through reversible esterification reaction between boric acid and glycerol is reduced to 4.4 kPa, whereas that of the double crosslinked network CP prepared using traditional acetic acid strategy is 13.7 kPa.This reflects that the GBCP hydrogel is more prone to dissipating energy through deformation flow, thus facilitating the formation of lasting and robust adhesion with the electrode interface.As shown in Fig. 1 and Supplementary Fig. 4-6, this esterification reaction can enhance the adhesion performance of the PAM/CS dual crosslinking network, thereby improving the infiltration of the hydrogel into the electrode.In Revision: The sentence 150 was changed to "Fundamentally, this can be attributed to the multiple electrostatic and weak intermolecular interactions induced by the esterification and protonation process in the ZGBCP electrolyte (Fig. 1g) [5][6][7][8] .Besides, the faster motion of chain segment tuned by the esterification also helps the adhesive groups (-OH, -NH2, and -CONH2) diffuse dynamically to form a consistently robust adhesion.In addition, the polymerization of AM can enhance the crosslinking density, ensuring the reinforced stability between the electrolyte and the electrode.In summary, all these three reactions are all essential for the robust interface of the ZGBCP electrolyte with electrodes." to clarify the contributions of three kinds of reactions to the robust interphase of the ZGBCP electrolyte and the dual crosslinking network.
Besides, Fig. 1b was replaced by the initial Fig. 1c, and Fig. 1c was replaced by Fig. R1 to further clarify the effects induced by esterification and protonation.

The reconstruction of cationic and anionic solvation structures is emphasized but not
clearly summarized in the paper.In Revision: The sentences of 'In conclusion, based on the dual cross-linking framework with robust adhesion and faster motion of chain segments as designed, the reconstruction of the cationic solvation structure enables the ZGBCP electrolyte to suppress the vicious HER and accelerate the stripping/plating kinetics of Zn anodes.Furthermore, the anionic solvation structures are reconstructed to prevent the corrosion induced by the BF4 -decomposition and accelerate the charge transfer process towards cathodes, enabling the novel ZGBCP electrolytes for ultra-stable and ultrawidetemperature-range AZMBs (Fig. 5g).' was provided in the discussion part to provide more detailed explanation of reconstruction of solvation structures.

The statement about the strong hydrogen bond between central hydroxyl-H and one of the edge hydroxyl-O within the glycerol molecule contributing to rapid desolvation lacks evidence to support it.
Our response: Thank you very much for your profound question.First, in previous studies about polyols, M. Mostafavi et al. spectroscopically confirmed the rapid solvation/desolvation kinetics contributed by the glycerol 9 .Second, to provide further evidence to support our hypothesis, we systematically investigated the roles of similar compounds such as 1,2-propanediol and 1,3-propanediol with different intramolecular hydrogen bonding in regulating the desolvation.As seen from Fig. R2a and R2b, Raman spectroscopy reveals that the intensity of the symmetric and asymmetric stretching vibrations of O−H bonds in glycerol is much higher than those of 1,2-propanediol and 1,3-propanediol, indicating the stronger coupling of O−H vibrational modes caused by more intense intramolecular hydrogen bonding of glycerol 10 .Furthermore, to identify the influence of the intramolecular hydrogen bonding on the desolvation process, CA test of the Zn electrode with the electrolytes containing glycerol, 1,2-propanediol and 1,3-propanediol was conducted.It was observed from Fig. R2c that the Zn 2+ transfer number of ZGP electrolyte with glycerol is 0.61, whereas for electrolytes (Z12PP and Z13PP) using 1,2-propanediol and 1,3-propanediol, the Zn 2+ transfer number are only 0.25 and 0.33, respectively, proving that glycerol did accelerate the Zn desolvation as well as Zn plating reaction.Due to the accelerated desolvation process, Zn|Zn symmetric cell using ZGP electrolyte exhibits the lowest initial plating polarization voltage at a current density of 1 mA cm −2 , shown in Fig. R2d.In Revision: As advised by the reviewer, to prove the contribution of strong hydrogen bonding in the glycerol to the rapid desolvation reaction, Fig. R2b, R2c, and R2d were provided as Supplementary Fig. 19c-e in the supporting file and the explanation 'To further support the hypothesis, the roles of similar compounds such as 1,2-propanediol and 1,3-propanediol with different intramolecular hydrogen bonding in regulating the desolvation were systematically investigated.As seen from Supplementary Fig. 19c, Raman spectroscopy reveals that the intensity of the symmetric and asymmetric stretching vibrations of O−H bonds in glycerol is much higher than those of 1,2-propanediol and 1,3-propanediol, indicating the stronger coupling of O−H vibrational modes caused by more intense intramolecular hydrogen bonding of glycerol 10 .Furthermore, to identify the influence of the intramolecular hydrogen bonding on the desolvation process, CA test of the Zn|Zn cells with the
It was observed from Supplementary Fig. 19d that the Zn 2+ transfer number of ZGP electrolyte with glycerol is 0.61, whereas for electrolytes (Z12PP and Z13PP) using  In Revision: As advised by the reviewer, DFT calculations of the desolvation energy were provided as the direct evidence for the reliability of transfer number result.
Fig. R3b was added as Supplementary Fig. 19f to the supporting file.The sentence 'Besides, the stronger anionic affinity of boric acid and glycerol is also conducive to the transfer process of cations, which interprets the high transfer number (0.82) and much lower desolvation energy barrier (3.46 eV) of Zn 2+ in the ZGBCP electrolyte as shown in above Fig.2g and Supplementary Fig. 19f.' was added for further explanation. '

5.
The differences in solvation structures of cations and anions compared to other reported electrolytes are not clearly demonstrated in the current simulation results.The reconstruction of anion solvation structures should be more clearly presented.
Our response: Thank you for your suggestion.We have extensively reviewed relevant literature, and a comparison of different Zn 2+ solvation structures was provided and shown in Table R1.It is clear that the previously developed strategy mainly focused on altering the cation solvation structure by adding high concentration salts or organic solvents, while there were no discussions on the solvation structures of the BF4 − anions.
The anionic solvation structure is mainly governed by weak interactions of hydrogen bonding.First, the function groups of −OH in glycerol, −OH in boric acid, and −CONH2 in polyacrylamide acting as hydrogen bond donors replace the water molecules around free BF4 − .Second, these hydrogen bond donors can also confine the free water molecules and inhibit the water-induced vicious hydrolysis reaction of BF4 − .Thereby, the reconstruction of anions can be realized.To further answer your question, we analysis all the potential anionic solvation structures as shown in Fig. R4 and R5.It reveals that all the reconstructed structures exhibit lower solvation energy and narrower energy gap, thus the charge transfer process during cycling can be accelerated on the basis of protecting free BF4 -from the attacking of H2O.In Revision: As advised by the reviewer, in the revised manuscript, we further discuss the reconstruction of the anionic solvation structure in the main text.The Table R1, Fig.
R4 and R5 were added to the supporting file and the following explanation 'The anionic solvation structure is primarily governed by weak interactions of hydrogen bonding.
First, the function groups of −OH in glycerol, −OH in boric acid, and −CONH2 in polyacrylamide acting as hydrogen bond donors replace the water molecules around free BF4 − .Second, these hydrogen bond donors can also confine the free water molecules and inhibit the water-induced vicious hydrolysis reaction of BF4 − .Thereby, the reconstruction of anions can be realized.'was added into the main text.

Reviewer #2 (Remarks to the Author):
The paper titled "Co-reconstruction of anionic and cationic solvation structures for ultra-stable operating of zinc-metal batteries under -50~+100oC" by authors L. Yao, J.
Liu, F. Zhang, B. Wen, X. Chi and Y. Liu deals with the regulation of cationic/anionic solvation structures in aqueous zinc metal batteries by using multicomponent electrolyte, consisted of Zn(BF4)2, glycerol, boric acid, chitosan and polyacrylamide (ZGBCP).The authors presented a detailed experimental and computational investigation of solvation structure and a comprehensive understanding of the mechanism in proposed battery devices.Also, the performances of proposed waterbased zinc batteries are promising.All the experiments are well-thought-out and welldesigned resulting in a coherent story even though the investigated system is complicated.Overall, the article is interesting, actual, and novel.
Our response: We appreciate the reviewer for recognizing the novelty and importance of our work and constructive advice.We have carefully addressed the reviewer's comments point by point.

Recommendation: Major revision
Below are some comments, hope will help to modify this paper: Major remarks:

1.DFT calculations. The use of DFT calculations to explain the coordination of Zn 2+
and BF4 -in terms of binding energies and bond order is a great idea.
Is there any particular reason why chitosan is excluded from the discussion and calculation of binding energies?From Scheme 1g, chitosan is proposed to be one of the coordination sites for Zn-ion, along with acrylate, which is a reasonable assumption due to the existence of -OH and -NH groups in the chitosan structure.However, there is no data for Zn 2+ -chitosan binding energies from DFT calculation which is contradictive.Also, does BF4 -have any interactions with chitosan or PAM based on DFT calculations?Please, revise the DFT models and change the proposed solvation model accordingly.
Our response: Firstly, thank you very much for your insightful question.We agreed with the reviewer that chitosan (CS) has interactions with Zn 2+ .So, the CS was actually taken into account into the calculation system.However, CS constitutes only 4.1 wt% of the entire electrolyte framework as mentioned in the experimental section, thus showing little influence on the average solvation structure from the calculation results.Compared to PAM, chitosan exhibits slightly stronger affinity for Zn 2+ (−1.57eV vs. −1.24eV).Functional groups -OH and -NH in CS have higher degrees of freedom than amide group in PAM, and the N atom in the -NH2 group possesses lone pair electrons, enabling it to act as a Lewis base and form coordination bonds with metal ions more effectively 16 .As to the anion interactions, the calculation results in Fig. R7 show that CS exhibits a stronger affinity to BF4 -(−0.28 eV for PAM vs. −0.41eV for CS).This interaction is dominated mainly by hydrogen bonding of BF4 -with hydrogen bonding donors such as -NH2, and -OH in the CS.In Revision: As advised by the reviewer, in the revised supporting file, we supplemented the Fig. R6 and Fig. R7; in the revised manuscript, we provided more discussions of the interactions between the CS and cations/anions.therefore the corresponding discussion is confusing.Also, please note that chitosan is again missing like in the DFT calculations, so the proposed discussion is ambiguous and needs to be revised accordingly.

2.MD simulations. It is difficult to analyze presented
Our response: Thank you for your profound comments and questions.For molecular dynamics (MD) simulations, the size of the simulation box and numbers of molecules are presented in the following Table R2.As to the force field, GAFF force field was applied in this work since it has been widely used for molecular dynamics simulations of electrolytes 11,17,18 .In the ZGBCP system, the optimal solvation structure of Zn 2+ obtained from MD simulations is Zn(BF4 − )1.4(H2O)3.6(C3H8O3)0.6PAM0.4,while in the AE system, it is Zn(BF4 − )0.8(H2O)5.show that the dissociation energy barrier of H in coordinating water molecules increase over 2 eV, corresponding to a more reduced tendency for HER (Fig. R11 and R12).In Revision: As advised by the reviewer, in the revised supporting file, we supplemented the Table .R2 and Fig. R8-R12. in the revised manuscript, we have incorporated more discussions on the coordination number and clarified the interactions within the cationic solvation structures and roles of the chitosan in the electrolyte.
3.Please, provide the procedure for treatment of the electrode before post-mortem analysis.SEI formation could be easily influenced by post-treatment of the electrode, and it is necessary to include a detailed handling procedure of the electrode.
Our response: We agree with the reviewer's comment on the treatment of the electrode.
The treatment procedure is as followings: first, the gel residues adhering to the surface of the cycled zinc foil was completely peeled off with great care.Given the good solubility of Zn(BF4)2 in ethanol and the relatively low boiling point of ethanol, the zinc In Revision: The treatment procedure 'First, the gel residues adhering to the surface of the cycled zinc foil was completely peeled off with great care.Given the good solubility of Zn(BF4)2 in ethanol and the relatively low boiling point of ethanol, the zinc foil cycled with ZGBCP and AE electrolytes were then rinsed with anhydrous ethanol to prevent any potential influence of secondary hydrolysis of BF4 − on the determination of F species in the SEI.Subsequently, the samples were vacuum dried, transferred under inert atmosphere protection, and promptly subjected to relevant tests to avoid secondary oxidation and radiation damage.' was added to the experimental part.
Minor remarks:

XPS data. It would be interesting to include a wide range of data for XPS
measurements in Supporting information.The shift in XPS data is obvious (Figure 3e and Fig. S14).Did the authors perform any internal calibration of XPS data?
Our response: We sincerely appreciate your good comments.As advised, we have provided detailed XPS full spectra and C1s peak data corrected for charging effects at different depths (Fig. R13).Besides, the shift in XPS data was also calibrated and the revised figures were shown in Fig. R14 and R15.Your feedback is crucial for ensuring the accuracy of our data, and we are grateful once again for your meticulous inquiries.In Revision: The Fig. R13 was added to the revised supporting file.The original Fig. 2f was replaced with the Fig. R14 and Fig. R15 was provided as Fig. 22 in the revised supporting file, respectively.

5.
English should be improved in the entire Manuscript.
Our response: We sincerely appreciate the valuable feedback from the reviewer.We have meticulously reviewed the entire article multiple times and corrected the language errors and inaccuracies to ensure clarity and precision in its language expression.The revised parts have been highlighted in both the main text and the supporting file.We are grateful for the reviewer's suggestions and are committed to putting in the effort to ensure that every aspect of the manuscript reaches the highest standards.

6.
Chapter before "Methods" should be Conclusion, not Discussion.comprehensive Discussion part.Combined with the reviewer's advice, the Chapter before 'Method' was changed to 'Conclusion and Discussion'.We appreciate the rationale behind it and believe that this adjustment will indeed better organize the structure of the article.We are truly grateful for the insightful guidance you have provided, and we are confident that this adjustment will enhance the coherence and readability of the article.Once again, thank you for your invaluable input.
glycerol, −OH in boric acid, and −CONH2 in the polyacrylamide backbone act as hydrogen bond donors, thus replacing water molecules around free BF4 − anions.These hydrogen bond donors can also confine the free water molecules and inhibit the waterinduced vicious hydrolysis reaction of BF4 − and stabilize the BF4 − by the ZGBCP electrolyte.The reconstructed solvation structures of both the cations and anions together contributes faster electrochemical charge transfer reaction, more excellent rate performance, wider temperature range and longer cycle life compared with the traditional aqueous electrolyte and other reported electrolyte systems.
In Revision: The above explanations were added to the Discussion part in the revised manuscript.

2.All experimental electrolyte systems (ZGP, ZCP and ZGBCP) were tested with a single composition ratio, respectively. It seems the authors have drawn mechanistic conclusions based on this single composition, without showing trends for varying amounts of the different compounds in these multi-component electrolytes.
Our response: Thank you for your insightful question.To save space in the article, the description of the gradient orthogonal experiment was omitted.Firstly, the dosage of PAM and CS was referenced from the typical formulations of double cross-linking network electrolyte 19 .Besides, the feeding amount of boric acid (BA, 0.6 g) nearly reached its solubility limit, while an appropriate amount of glycerol ensured the occurrence of esterification reaction, achieving a balance between adhesiveness, conductivity, and mechanical properties.In Revision: The Table R3 was provided in the revised supporting file and the corresponding explanation for the optimization of the electrolyte composition was added in the revised manuscript.

3.How do the different hydrogel electrolytes compare in water content expressed.
Our response: Thank you for your good question.As advised, thermal gravimetric analysis (TGA) was conducted to determine the water content of different hydrogels.
The results were compared in the Fig. R16.It can be seen that the water contents of the ZCP, ZGP and ZGBCP hydrogel electrolytes were 34.2 wt%, 51.6 wt%, and 28.1 wt%, respectively.In Revision: The Fig. R16 was supplemented in the revised supporting file.

4.L82: by "protonation", do the authors refer to hydrogen-bonding as a non-covalent crosslinking mechanism?
Our response: We are very sorry for the unclear explanation.The chitosan (CS) generally requires acid, e.g. the commonly used acetic acid, to dissolve and form a gel in the water.The dissolution and gelation process of CS is a proton-involved process.
Herein, we did not add the acid during the electrolyte synthesis.The protons come from the esterification reaction between boric acid and glycerol according to the literature 20- 22 , and the motion of chain segments is accelerated through the reaction compared with traditional methods.As shown in Fig. R1 plasticization through reversible esterification reaction between boric acid and glycerol is reduced to 4.4 kPa, whereas that of the double crosslinked network CP prepared using traditional acetic acid strategy is 13.7 kPa.This reflects that the GBCP hydrogel exhibits more viscoelastic properties, which is more prone to dissipating energy through deformation flow, and the faster motion of chain segments ensures the faster ions transport and helps the adhesive groups diffuse dynamically to form a consistently robust adhesion.Therefore, the protonation refers to the proton-stimulated dissolution and gelation of CS.In Revision: The Fig. R1 was provided to clarify the effects induced by esterification and protonation process.Besides, the sentence 150 was changed to "Fundamentally, this can be attributed to the multiple electrostatic and weak intermolecular interactions induced by the esterification and protonation process in the ZGBCP electrolyte (Fig. 1g) [5][6][7][8] .Besides, the faster motion of chain segment tuned by the esterification also helps the adhesive groups (-OH, -NH2, and -CONH2) diffuse dynamically to form a consistently robust adhesion.In addition, the polymerization of AM can enhance the crosslinking density, ensuring the reinforced stability between the electrolyte and the electrode.In summary, all these three reactions are all essential for the robust interface of the ZGBCP electrolyte with electrodes." to clarify the contributions of three kinds of reactions to the robust interphase of the ZGBCP electrolyte and the dual crosslinking network.In Revision: In the revised supporting file, the Table R2, Fig. R8, R17 and Fig. R18 and the corresponding explanation were provided.The following explanation 'To investigate the diffusivity of the ZGBCP electrolyte, MD simulations of the ZGBCP system at different temperatures were further conducted.Based on the MSD calculations (Supplementary Fig. 18), the theoretical diffusion coefficients σ at different temperatures can be obtained.As seen from Supplementary Fig. 18b, the diffusion coefficients and conductivities obtained from the simulations exhibit similar trends and orders of magnitudes as those obtained from the experimentally measured conductivity versus temperatures as shown in Supplementary Fig. 18c This confirms the accuracy and reliability of the simulation results.' was added in the theoretical calculation part of the revised supporting file.

7.Even though MD simulations are conducted that allow to collect ensembles of different complete solvation complexes, DFT energies are only computed for strongly simplified, partial solvation shells, even comparing absolute energies of mono-dentate
and bidentate complexes directly.
Our response: Thank you very much for your insightful perspective and profound questions.MD simulations have been widely recognized to be a reliable tool to gain the insights and a comprehensive view of solvation effects, although some models are simplified.To get much deeper understandings, more extensive computational

(c)
resources are needed.Also, the ZGBCP electrolyte developed in this work is a new and relatively complicated system.Due to the limited resources and the uniqueness of the electrolyte, we have tried our best to provide as many calculations as possible in this work.Furthermore, we believe some simplifications used in this work are based on an understanding of the crucial interactions and energy contributions within the new electrolyte system.
To provide more theoretical calculation models and gain a deeper understanding of the new solvation structures built by the CRACSS strategy, first, based on the reviewer's suggestion, we calculated the binding energy of Zn 2+ and BF4 -with two H2O molecules as shown in Fig. R19.Besides, we calculated more cationic solvation structures that are possible in the ZGBCP electrolyte system, which included the CIPtype structures without polymer involvement: [Zn 2+ (BF4 − )(H2O)4(C3H8O3)] + and SSIPtype structures: [Zn 2+ (H2O)5(C3H8O3)] 2+ and [Zn 2+ (H2O)6] 2+ based on the B97-3c functional.The solvation structure [Zn 2+ (BF4 − )(H2O)4(C3H8O3)] + -PAM under primary discussion exhibited the lowest energy, thereby confirming the validity of the MD simulation as illustrated in Fig. R8.What's more, the additional two solvation structures in the ZGBCP electrolyte also exhibit the inhibited HER tendency and more active electron states (Fig. R20 and R21), which can further suppress the side reactions in the Zn anodes and accelerate charge transfer kinetics as well as the solvation structure under primary discussion.In Revision: In the revised supporting file, the Fig. R4, R5, R8, and R19-21 and the corresponding explanation were provided.

8.The DFT functional used is outdated. Given the size of the complexes if sampled from
the MD simulation, B97-3c is a newer, much more accurate and cheaper alternative for optimization.
Our response: Thank you for your sincere suggestions.We have actively adopted your suggestions to improve the efficiency and accuracy of our DFT calculations.By using the B97-3c functional instead of B3LYP, we have recalculated all solvation structures mentioned in the manuscript.We greatly appreciate the valuable feedback from the reviewer, which is crucial for the improvement of our future work.

9.
Polyacrylamide was simplified to a trimer, however, its structure doesn't match the polymer backbone.There should be a terminal CH2, which can be saturated to CH3 to cap the polymer.Further, no explanation is given as to why a trimer was also used in the classical MD simulations.
Our response: Thank you very much for pointing out the modeling issues.To address this issue, we have re-modeled the PAM to more accurately reflect the polymer's structure.We greatly appreciate the valuable feedback from the reviewer, which is crucial for the improvement of our future work.The new calculation result is shown in Fig. R22.In Revision: As advised by the reviewer, in the revised supporting file, we supplemented the Fig. R4-R7, R10, R12, R25 and the corresponding explanation 'Although it's a small amount for CS, a new AGG solvation structure [Zn 2+ (BF4 -)2(H2O)2(C3H8O3)]-CS was found near the CS chain segment, which is conducive to the desolvation process with highest H + dissociation energy barrier (Supplementary Fig. 13).'.
11.In summary, for the analysis of the mechanistic effects in the hydrogel electrolytes, the authors conducted computational studies, which methodology appears to be misleading and/or incomplete.Therefore, the work does not support the conclusion and claims made by the authors.

Our response:
Thank you for your valuable feedback.We are greatly grateful for your thoughtful consideration of our work.To comprehensively investigate and analyze the completely new electrolyte system, we have tried our best to apply as many characterization tools as possible experimentally and conducted as many modeling methods as possible theoretically.experimental data and theoretical calculation results show good match and the calculation helps understand the underlying mechanism of the unique properties and performance of the electrolyte.Furthermore, from the theoretical calculation, for the first time, we found a co-reconstruction of both anionic and cationic solvation structures; from the experimental test, for the first time, we demonstrated both an electrolyte with the stat-of-the-art working temperature window and a full cell with the best cycling stability.There might be some deficiency in the previous version of the manuscript, however, we believe the revised version has well addressed the reviewer's concerns and meets the requirements of the journal.Once again, we appreciate your feedback and constructive criticism, which will undoubtedly contribute to enhancing the quality of our manuscript.

Response to Reviewers' Comments
Reviewer #1 (Remarks to the Author): The reviewer is satisfied with the data provided in the response letter.However, some data are not thoroughly discussed.One point that may require detailed discussion is the uncoordinated anion.The author demonstrated that the reconstructed structures of the anion exhibit lower solvation energy and a narrower energy gap compared to BF4 - Our response: Thanks for your constructive question.Firstly, for the solvation behavior of free anions tuned by the CRACSS strategies as shown in Fig. R4, the 10101 exhibited the highest solvation energy due to the more abundant hydrogen-bond donors (-NH2, -OH) of CS than PAM with only -CONH2 donors.In the presence of H2O and PAM, the single coordination sites make their interaction with BF4 -relatively weak.
Meanwhile, due to the presence of PAM, the coordination of H2O will be limited, resulting in a less stable solvation structure than the strong coordination from glycerol and CS.Notably, all these structures are important for the interfacial stabilization of Zn anodes and the desolvation process for the PANI cathodes.What's more, for the 21000 mentioned in Fig. R5, the electron states in the HOMO level are mainly occupied by the H2O molecule.In contrast, the LUMO level is contributed from H3BO3 since the typical electron deficient characteristics in H3BO3 because of the sp2 hybridization.On the other hand, the LUMO level of C3H8O3 is mainly contributed by anti-bonding orbitals with relatively low energy levels of -OH groups, whereas the HOMO level of C3H8O3 is relatively higher due to the lone-pairs electrons in -OH groups.Thus, the solvation structures such as 10110 and 01110 consisting of C3H8O3 exhibit lower energy gaps.As for the PAM-involved structures, the relatively higher HOMO can be attributed to the non-bonded electronic state of N and O in -CONH2 groups from PAM.Similarly, the HOMO of CS is relatively higher due to the sufficient -NH2 and -OH groups compared with PAM, which ensures the solvation structures involved with CS exhibit lower energy gaps.
In conclusion, the mechanism of reduction of energy gaps of all the solvation structures tuned by the CRACSS strategy can be clarified from the prospective of electronic structures and energy levels.The CRACSS strategy paves a new way in electrolyte design for AZMBs with fast kinetics and stable interface for wide temperature range applications.Thank you so much again for your constructive and  In revision: The sentence 'For the solvation behavior of free anions tuned by the CRACSS strategies, the 10101 exhibited the highest solvation energy due to the more abundant hydrogen-bond donors (-NH2, -OH) of CS than PAM with only -CONH2 donors.In the presence of H2O and PAM, the single coordination sites make their interaction with BF4 -relatively weak.Meanwhile, due to the presence of PAM, the coordination of H2O will be limited, resulting in a less stable solvation structure than the strong coordination from glycerol and CS.Notably, the two structures are important for the interfacial stabilization of Zn anodes and desolvation process for the PANI cathodes.' was added to the Supplementary Fig. 16 to further clarify the differences in solvation energies from different complexes.
The sentence 'For the 21000 mentioned in Supplementary Fig. 39, the electron states in the HOMO level are mainly occupied by the H2O molecule.In contrast, the LUMO level is contributed from H3BO3 since the typical electron deficient characteristics in H3BO3 because of the sp2 hybridization.On the other hand, the LUMO level of C3H8O3 is mainly contributed by anti-bonding orbitals with relatively low energy levels of -OH groups, whereas the HOMO level of C3H8O3 is relatively higher due to the lone-pairs electrons in -OH groups.Thus, the solvation structures such as 10110 and 01110 consisting of C3H8O3 exhibit lower energy gaps.As for the PAMinvolved structures, the relatively higher HOMO can be attributed to the non-bonded electronic state of N and O in -CONH2 groups from PAM.Similarly, the HOMO of CS is relatively higher due to the more sufficient -NH2 and -OH groups compared with PAM, which ensures the solvation structures involved with CS exhibit lower energy gaps.In conclusion, the mechanism of reduction of energy gaps of all the solvation structures tuned by the CRACSS strategy can be clarified from the prospective of electronic structures and energy levels.' was added to the Supplementary Fig. 39 to further clarify the differences in energy levels from different complexes.

Reviewer #3 (Remarks to the Author):
Thank you for the comprehensive response and the effort to address all reviewer's comments.After carefully reading the revised manuscript, only a few minor remarks come to mind: -Please add the computational level to the figure captions you show DFT results.
-As a suggestion (totally optional), consider moving Supplementary Fig. 10 to the main text.
Our response: We appreciate the reviewer for recognizing the innovations of our work and proposing very constructive revision advice.We have carefully and thoroughly addressed the reviewer's comments below.

1.
Please add the computational level to the figure captions you show DFT results.
Our response: Thanks for your sincere advice.The mark '(B97-3c)' of all the calculations about solvation structures using B97-3c is added to the captions of figures.

2.
As a suggestion (totally optional), consider moving Supplementary Fig. 10 to the main text.
Our response: Thanks for your sincere advice.It's meaningful to move Supplementary Fig. 10 to the main text, but it's difficult to make a suitable replacement due to the length limit of the article page.Thus, we decided to keep Supplementary Fig. 10 where it used to be.

Fig. R1 |
Fig. R1 | The variation of storage modulus G' and loss modulus G'' with strain oscillation of PAM/CS crosslinking network with/without the boric acid and glycerol.More importantly, boric acid and glycerol synergistically modulate the solvation structures, suppressing the HER during zinc plating, avoiding the excessive protons introduced by traditional acetic acid methods, and effectively alleviating the vicious hydrolytic corrosion of free BF4 -anions, thus avoiding the hydrogen evolution problem of traditional gel electrolytes.Detailed descriptions of this anion/cation solvation structure reconstruction are provided in Fig. 2-3 and Supplementary Fig. 10-14.

Our response :
Thank you for your good advice.First, for the reconstruction of cationic solvation structure, −OH in glycerol and −CONH2 in the PAM backbone show stronger affinity to Zn 2+ than the water molecules, leading to the partial replacement of water molecules in the cation solvation structure and then forming a new cation solvation composition of Zn(BF4 − )1.4(H2O)3.6(C3H8O3)0.6PAM0.4 in the ZGBCP electrolyte instead of the Zn(BF4 − )0.8(H2O)5.2 in traditional AE, which effectively suppresses HER at the Zn electrode and accelerates the stripping/plating kinetics of the Zn electrode.Second, for the reconstruction of anionic solvation structure, similar to that of the above cationic solvation structure, −OH in glycerol, −OH in boric acid, and −CONH2 in the PAM backbone act as hydrogen bond donors, thus replacing water molecules around free BF4 − anions and forming new solvation structures, such as [BF4 − (H2O)(H3BO3)(C3H8O3)] -.Unlike the vicious hydrolysis reaction of BF4 − induced by the surrounding water molecules in the AE, the BF4 − can be stabilized by the ZGBCP electrolyte.The reconstructed solvation structures of both the cations and anions together contributes faster electrochemical charge transfer reaction, more excellent rate performance, wider temperature range and longer cycle life compared with the traditional aqueous electrolyte and other reported electrolyte systems.

Fig.S12- 13
Fig.S12-13 elaborate on the changes in anion solvation structures and the distribution of the surrounded species in the ZGBCP electrolyte and traditional aqueous electrolyte.
Therefore, the discussions of the DFT calculation sections mainly focused on the interactions of PAM with Zn 2+ .According to the reviewer's advice, we have modified the calculation model and enhanced the calculations of the interactions between PAM, CS, Zn 2+ , and BF4 − .The new calculation results are shown in Fig R6 and Fig R7.
RDFs data from MD simulations and to evaluate the validity of the used force field.Authors should provide the number of molecules (in Supporting information) and the size of the simulation box used in MD simulations during the production run.Did the authors perform any validation of the used force field?I will strongly recommend showing the obtained data in the form of coordination number since presented RDFs (Figure 2e) are impossible to analyze and

2 .
To validate its effectiveness, DFT calculations were systematically performed for various potential solvation structures as shown in Fig R8.The solvation structure under primary discussion exhibited the lowest energy, thereby confirming the validity of the MD simulation.Additionally, MD simulations were conducted at different temperatures, and the simulated conductivities based on Mean Square Displacement (MSD) show consistent trends and magnitudes with experimental results, further confirming the validity of the simulation as seen in Fig. R9.

Fig. R8 |
Fig. R8 | Structures of potential solvation structures of Zn 2+ and corresponding solvation energy Esol in the (a) AE, and (b) ZGBCP electrolyte.

[
Zn 2+ (BF -foil cycled with ZGBCP and AE electrolytes were then rinsed with anhydrous ethanol to prevent any potential influence of secondary hydrolysis of BF4 − on the determination of F species in the SEI.Subsequently, the samples were vacuum dried, transferred under inert atmosphere protection, and promptly subjected to relevant tests to avoid secondary oxidation and radiation damage. Fig. R13 | The XPS survey and corresponding C 1s spectra of Zn anodes at different depths using (a) ZGBCP, and (b) AE electrolyte.

Our response:
Thank you for your suggestion.The Conclusion was actually included in the Discussion part since the Nature Communications journal advises to provide a
, rheological test results indicate that the energy storage modulus G' of the double crosslinked network GBCP prepared by

Fig. R1 |
Fig. R1 | The variation of storage modulus G' and loss modulus G'' with strain oscillation of PAM/CS crosslinking network with/without the boric acid and glycerol.

6 .
, Multiwfn, DFT functionals/basis sets and solvation models are missing.Our response: Thank you for the valuable suggestion.We have included the important references (Ref.65, 69-75) related to the computational tools and theoretical frameworks we used in the revised manuscript.In Revision: In the revised manuscript, citations (Ref.65, 69-75) of Gaussian 09 software package, ORCA software package, as well as the Multiwfn program are cited.Additionally, the citations for the detailed information on the selection of the density functional theory (DFT), basis sets, and implicit solvent models used in our DFT calculations were also supplemented.Thank you for your thorough review and sincere suggestions.Details about the MD simulations, such as structures of all species and their numbers are missing.The validity of the MD simulations is also not addressed.How do modelled transport properties, such as diffusivity, compare to experiments?Our response: Thank you for your constructive advice.Table R2 and Fig. R17 provide detailed information on the box size, species structure, and quantity obtained from the MD simulations.To validate its effectiveness, DFT calculations were systematically performed for various potential solvation structures as shown in Fig R8.The solvation structure under primary discussion exhibited the lowest energy, thereby confirming the validity of the MD simulation.Additionally, we conducted further MD simulations of the ZGBCP system at different temperatures.Based on MSD calculations, diffusion coefficients σ at different temperatures were obtained.The diffusion coefficients obtained from the simulations exhibit similar trends and orders of magnitudes as those obtained from the experimentally measured conductivity versus temperatures as shown in Fig. R18.This confirms the accuracy and reliability of the simulation results.

Fig. R18 |
Fig. R18 | Ionic transport property calculated by the MD simulation under different temperatures.a, MSD of Zn 2+ obtained from MD simulations under different temperatures.b, diffusion coefficients and c, conductivities of ZGBCP electrolytes based on simulation and experiments under different temperatures.

Fig. R8 |
Fig. R8 | Structures of potential solvation structures of Zn 2+ and corresponding solvation energy Esol in the (a) AE, and (b) ZGBCP electrolyte.

Fig. R21 |
Fig. R21 | (a) ESP distribution, (b) energy level, and (c) electrostatic interaction energy (-H) of [Zn 2+ (H2O)5(C3H8O3)] 2+ in the ZGBCP electrolyte and [Zn 2+ (H2O)6] + in the AE electrolyte.In addition, calculations were also performed for various possible solvation structures of BF4 − anions named as (H2O)a(C3H8O3)b(H3BO3)c(PAM)d(CS)e.It can be observed from Fig. R4 and R5 that all possible solvation structures of free anions in ZGBCP exhibit lower solvation energy especially the BF4 -anions near the CS chain, which can effectively protect the free BF4 -from the attacking of H2O.What's more all the potential anionic solvation structures exhibit the lower energy gap than the [BF4 - (H2O)3] -in the AE electrolyte, accelerating the charge transfer kinetics with PANI.

Fig. R22 |
Fig. R22 | (a) Corrected trimer of PAM.(b) g(r) and n(r) of Zn 2+ obtained from MD simulation based on the corrected trimer of PAM.(c) ESP of the cationic solvation structures in the ZGBCP and AE electrolytes.In addition, trimer and pentamer have been widely used in classical molecular dynamics simulations of hydrogel electrolytes15,[23][24][25] .Therefore, we have also conducted molecular dynamics simulations based on pentamer PAM and CS.The simulation results were shown in Fig.R23, Fig.R24and TableR4, which are similar to those of the trimer MD simulation.

Fig. R23 |
Fig. R23 | Molecular models of PAM and CS pentamers used in the MD simulations.

Fig. R24 | 10 .
Fig. R24 | (a) g(r) and n(r) of Zn 2+ .(b) g(r) of F in the ZGBCP_5mer system.(c) n(r) of F-H (H2O) in the ZGBCP_3mer and ZGBCP_5mer systems.In Revision: In the revised supporting file, the Fig.R23, R24 and Table.R4 as well as the corresponding explanation of 'As illustrated in Supplementary Fig.14, the results of MD simulations based on the PAM and CS pentamer models are consistent with those based on the trimer models, verifying the validity of MD simulations.' were provided.10.It is unclear how chitosan was depicted in the simulations.The possible complexation of ions by chitosan wasn't investigated at all.Our response: We apologize for the missing discussions on the simulations and complexation of chitosan.We agreed with the reviewer that chitosan (CS) has interactions with Zn 2+ .So, the CS was actually taken into account into the calculation system.However, CS constitutes only 4.1wt% of the entire electrolyte framework as mentioned in the experimental section, thus showing little influence on the average solvation structure from the calculation results.Therefore, the discussions of the DFT calculation sections mainly focused on the interactions of PAM with Zn 2+ .According to the reviewer's advice, we have modified the calculation model and enhanced the calculations of the interactions between PAM, CS, Zn 2+ , and especially BF4 − .The new

( 1 .
H2O)a(C3H8O3)b(H3BO3)c(PAM)d(CS)e (abcde=30000), indicating an acceleratedcharge transfer process.However, the differences between other complexes, apart from 30000, are not explained.For instance, 21000 shows the second-highest energy gap of 7.17 eV in Fig.R4, 10101 represents the lowest, and 20010 presents the highest solvation energy in Fig.R4.The upper and lower limits of the displayed energy levels in Fig.R5for different complexes should be clearly specified.Our response:We appreciate the reviewer for approving the importance and novelty of our work and constructive revision advice.We have addressed the reviewer's comments point by point.The author demonstrated that the reconstructed structures of the anion exhibit lower solvation energy and a narrower energy gap compared to BF4 -(H2O)a(C3H8O3)b(H3BO3)c(PAM)d(CS)e (abcde=30000), indicating an acceleratedcharge transfer process.However, the differences between other complexes, apart from 30000, are not explained.For instance, 21000 shows the second-highest energy gap of 7.17 eV in Fig.R5, 10101 represents the lowest, and 20010 presents the highest solvation energy in Fig.R4.

Table R2 .
| Box length and molecule numbers of ZGBCP and AE electrolyte systems.

Table .
R2 | Box length and molecule numbers of ZGBCP and AE electrolyte systems.
Fig. R17 | Molecular models used in the MD simulations.Fig.R8 | Structures of potential solvation structures of Zn 2+ and correspondingH

Table . R4
Box length and molecule numbers of ZGBCP_5mer system.