Locking the lattice oxygen in RuO2 to stabilize highly active Ru sites in acidic water oxidation

Ruthenium dioxide is presently the most active catalyst for the oxygen evolution reaction (OER) in acidic media but suffers from severe Ru dissolution resulting from the high covalency of Ru-O bonds triggering lattice oxygen oxidation. Here, we report an interstitial silicon-doping strategy to stabilize the highly active Ru sites of RuO2 while suppressing lattice oxygen oxidation. The representative Si-RuO2−0.1 catalyst exhibits high activity and stability in acid with a negligible degradation rate of ~52 μV h−1 in an 800 h test and an overpotential of 226 mV at 10 mA cm−2. Differential electrochemical mass spectrometry (DEMS) results demonstrate that the lattice oxygen oxidation pathway of the Si-RuO2−0.1 was suppressed by ∼95% compared to that of commercial RuO2, which is highly responsible for the extraordinary stability. This work supplied a unique mentality to guide future developments on Ru-based oxide catalysts’ stability in an acidic environment.

3. Page 4, line 119-122, the authors need to describe the reason why the binding energy of la ce oxygen constantly shi ed toward a higher binding energy with increasing Si content, while that of Ru shi ed toward a lower binding energy (x≤0.1) and then remained almost unchanged (0.1≤x≤0.3). 4. In Figure 4a, the stability follows an order of Si-RuO2-0.1>Si-RuO2-0>Com-RuO2.However, in Figure 4c, the Ru dissolu on rate of Com-RuO2 is significantly lower than that of Si-RuO2-0; meanwhile, this phenomenon is also observed in Figure S11.Is there a contradic on? 5. Page 7, line 197-199, the authors men oned the following: "Considering that Si-RuO2-0.1 has a small par cle size and porous structure compared to Com-RuO2, we deduced that the dissolu on of Ru in Si-RuO2-0.1 could be further inhibited by increasing the par cle size of Si-RuO2-0.1."I don't understand the basis of this deduc on, please give your explana on.
Reviewer #3 (Remarks to the Author): This manuscript reported a Si-doping RuO2 with extend OER stability in acid media, the durability looks become be er, however, there are too conclusion are not solid, at first all, all figures in this manuscript are obscure (low resolu on), I don't think this work is suitable for the publica on in the nature communica on, I suggest transfer to other journal a er the revisions are done as followed: 1.The authors emphasized that "Si tends to inters ally insert into the RuO2 la ce rather than replace the Ru atoms" through XRD simula ons and DFT calcula ons, but I think persuasive evidences should be provided such as spherical aberra on corrected electron microscopy, in fact, experimental data about the existence of Si in RuO2 is too less.
2. There is also inconsistence statement about Si inters ally insert into the RuO2 la ce, in which XRD show shi , there are no obvious different from la ce spacing, why? 3. I think even air calcined 450C, the carbon cannot be depleted completely, which should be related to conduc vity and stability.The authors can test EDS mapping by choosing carbon elements.
4. The similar work (Adv. Sci. 2023, 2207429) should be cited and discussed.5.In line 77 of page 3, the authors stated, "…and then almost unchanged as the Si content further increased from 0.1 to 0.3." that means when the Si is beyond 0.1, it cannot be doped into RuO2, the extra Si will form SiO2, if like this situa on, the author should characterize how many SiO2 formed on RuO2, this SiO2-RuO2 conduc vity is be er than commercial RuO2?Its ac vity s ll beyond the commercial RuO2? 6.In Fig. 2b, there is a much large offset of Si-RuO2-x rela ve to SiO2, why? 7. A er the stability test, whether the morphology, metal content and metal valence of Si-RuO2-0.1 catalyst changed.Please supplement the series of characteriza on a er the stability test.
8. Si prevents Ov forma on should be proved by EPR and so on.9. Si content in the catalyst needs to be further determined by ICP. 10.In Figure 2b and 2c, Com-RuO2 should add as a comparison.

General notes:
We thank all the reviewers for their thorough and valuable comments, which has helped us improve our manuscript.We provide point-to-point responses in the manuscript for all reviewers' comments.The responses are noted in blue, and the revised parts in the manuscript are highlighted in yellow.To better respond to the reviewer's comments, some figures published or in the original manuscript are cited in this reply and relabeled as Fig. R.

Response to Reviewers' Comments
Reviewer #1 (Remarks to the Author): In the manuscript, the authors reported their experimental and computational results on the electrocatalytic activity and stability of Si-doped RuO2 catalysts for oxygen evolution reaction (OER) in acid media.Specifically, the authors had synthesized the catalysts using cation-exchange resin pyrolysis approach, characterized the structure of the catalysts using XRD, TEM, and XPS, and measured the catalytic activity and stability of the catalysts for OER in acid media.It is noted that the authors also performed density functional theory (DFT) calculations to predict the structure of the Si-doped RuO2 and OER reaction pathway on the catalysts.The authors concluded that the interstitial doped Si would enhance both the activity and stability of RuO2 for OER in acid.The presented results only add some incremental knowledge/data to the current understanding of electrocatalysis.Some conclusions are questionable.This reviewer does not believe the current manuscript contain enough innovative, significant contents to be considered for publication in Nature Communications.

Response:
Thanks for your comments and valuable suggestions.

Response:
Improving the stability of Ru-based oxide is a long-term process for the PEM water electrolyzer because of their low cost and high activity.This work aims to provide a new strategy for approaching the highly stable Ru-based oxide toward acidic water oxidation.
(1) Some recent work reported much better OER activity and durability of RuO2 through doping than the presented catalysts.For example, "Non-iridium-based electrocatalyst for durable acidic oxygen evolution reaction in proton exchange membrane water electrolysis", Nature Materials, 2023.(4) Not much new knowledge is generated from this work.

Response:
Many thanks for recommending the paper published in Nature Materials.We have carefully read and analyzed it.In the reported work, the authors presented a Ni-RuO2 catalyst with high activity and durability in acidic OER by Ni-substituted Ru site in RuO2.Notably, there are some noticeable differences between our work and it, such as ① the doping element was metalloid (Si); ② the doping mode was interstitial-site doping.These two characteristics break the traditional thinking by metal atoms (Co, Ni, Mn, Cu, Cr, and so on) replacing the Ru atoms, and result in excellent performance.In addition, our work focuses on how to improve the stability of RuO2 catalysts for acidic water oxidation, which encourages to researchers to explore the stability, because in real applications, OER stability may be an even more critical factor to consider (Joule 5, 1-28, July 21, ( 2021)).Meanwhile, we propose a new concept of locking lattice oxygen to promote the stability of RuO2 catalysts, which is essential for improving the stability of OER catalysts, especially in acidic environments.Therefore, we believe this innovative work will arouse the considerable interest of readers.
(2) The employed experimental and computational techniques are widely used in the catalyst study.

Response:
Advanced characterization techniques are indeed helpful for discovering new phenomena and proposing new viewpoints.However, the experimental and computational methods employed in our work were able to demonstrate our perspectives and phenomena adequately.Moreover, these wellestablished experimental and computational techniques are more easily understood and accepted by the reader.Therefore, we believe that conventional experimental and computational approaches can also promote the development of science and technology.
(3) The authors could not solidly prove that all the Si dopants are in the assumed interstitial positions.

Response:
Thanks for your comment.The reviewer is correct.Although we observed that some Si with low imaging contrast exists in the RuO2 interstice through spherical difference electron microscopy (Fig. R1) and confirmed that Si easily enters the interstitial sites of RuO2 through XRD and other characterization techniques, it is challenging to verify that all the Si dopants are in the RuO2 interstice.In our manuscript, the ideal doping level of Si in RuO2 interstices was approximately 10%, which is only inferred from the nominal ratios in the precursor mixtures and a series of characterizations.1b and c. 1 are not sufficient enough to directly confirm that the Si dopants lie in the interstitial locations of RuO2 crystal.

Response:
Thank you for your valuable and thoughtful comments.According to your suggestion, we performed spherical aberration-corrected electron microscopy analysis.As we predicted, some Si with low imaging contrast were observed in the RuO2 interstice in the HADDF-STEM image (Fig. R2), suggesting the Si was successfully inserted into the RuO2 interstice.
The corresponding discussion has been updated in the revised manuscript as follows: To visually prove that Si was inserted into the RuO2 interstice, spherical aberration-corrected HAADF-STEM measurements were performed.As shown in Fig. 1d, the lattice fringes with interplanar spacings of 0.318 nm and 0.254 nm were assigned to the ( 110) and ( 101) planes of rutile RuO2, respectively.Furthermore, some isolated Si atoms with low imaging contrast, which is characteristic of light elements with lower atomic numbers, were also observed in the lattice interstices of RuO2 (Fig. 1e-f).This assertion was confirmed by atomic line profiles analysis (Fig. 1g-j).2b and c.
In addition, in our manuscript, we provide multidimensional evidence to demonstrate that Si is in the RuO2 interstice.
② Coordination number.To our knowledge, the Si invariably combines with 4-coordinated oxygen in nature (RSC Adv., 5, 74790 (2015)); however, Ru is coordinated with six oxygen atoms.If Si replaces Ru, Si is coordinated with six oxygens to form SiO6 octahedra.It is well-known that SiO6 octahedra are unstable under ambient conditions (Nature 328, 416-417 (1987)).

From an experimental standpoint:
① XRD patterns.In the case of substitution, the XRD peak shifts to higher 2θ value, indicating the incorporation of smaller Si 4+ into the lattice sites of Ru 4+ .However, in our work, the shift of the XRD peaks to lower angle with increasing Si content to 10% indicates an expansion of the crystalline lattice.This phenomenon suggests that Si 4+ ions were inserted into the interstices of the RuO2 lattice, resulting in the expansion of the crystalline lattice (ACS Energy Lett., 3, 970−978 (2018); Nat Commun 13, 3784 (2022)).
② Simulated XRD patterns.In the simulated XRD pattern, we observed some additional peaks at 10°<2θ<20° in the case of substitutional doping.The reason for this phenomenon is that when Si replaces the Ru site, the crystal structure of RuO2 suffers severe distortion because the ionic radius of Si 4+ is much smaller than that of Ru 4+ .However, for the interstitial-doped case, no additional peaks appeared except for the corresponding peak of RuO2, suggesting that interstitialdoping of Si cannot change the crystalline structure of RuO2, which is consistent with the experimental XRD patterns and common in interstitial doping examples (J Mater Sci: Mater Electron., 29, 9137-9141 (2018).;ACS Appl.Energy Mater., 4, 13636−13645 (2021);Chem. Mater., 33, 4135−4145 (2021)).
③ DFT calculations.In our work, DFT calculations were also performed to investigate the location of Si 4+ in RuO2 from an energy point of view.Based on the calculation results, we conclude that Si ions tend to insert into RuO2 interstice rather than replace the Ru atoms in RuO2, which also verifies that SiO6 octahedra are unstable.
Based on these considerations, we believe that the characterization results we present provide enough evidence that Si is doped into the interstitial locations of RuO2.
3. The computational results in Fig. 1(c) are questionable.First, no detailed information of the modelled structures is given.It is unclear if the crystal structure (including volume and shape) has been fully optimized.Secondly, the nearly linear decrease of the so-called "energy cost" with number of Si atoms would lead to an incorrect conclusion that Si would be preferred to be inserted into RuO2 in a very large amount.The experimental data presented in this manuscript do not support this prediction.

Response:
Thank you for your helpful suggestion.After repeated verification, we can confidently answer the reviewer that our computational results are believable.However, we apologize that we did not display detailed calculation information in the original manuscript, causing you to misunderstand.We have added the detailed calculation information in the revised manuscript.
All the calculations were performed after the crystal structure had been fully optimized.During the optimization process, we first constructed several representative structural models based on different doping amounts and doping modes (Fig. R3a-q), in which we empirically excluded unstable and repetitive models.Subsequently, the energies of different structures were calculated.Finally, the most stable structure was selected to perform subsequent operations according to the principle of lowest energy (Fig. R3r).Per your suggestion, we provided detailed information on the modeled structures after optimization (Table .R1) and supplemented them in the revised manuscript (Supplementary Figure 1 and Table 1).
Second, this so-called "energy cost" was derived from the reported literature (Nat.Commun., 13, 3784 ( 2022)).As determined from our calculation results, Si insertion is exothermic and spontaneous, implying that a large amount of Si can be inserted into the RuO2 interstice.As stated by the reviewer, our experimental data presented in this manuscript do not support this prediction.This phenomenon is because the theoretical calculation only considers their thermodynamic trends (initial state and final state) under ideal conditions.However, in practice, the experimental conditions (for example, precursor, annealing temperature and time, pressure.)will also directly affect the solubility of Si in RuO2.Even so, we have demonstrated experimentally and theoretically that a small number of Si atoms (about 10%) are more likely to insert into the RuO2 interstice than replace the Ru site.

Response:
Our experimental results are fully compatible with the theoretical calculations, so we do not understand the reviewer's claim that we did not provide a convincing explanation for the observed enhancement in both the activity and durability of the Si-doped catalysts.In our manuscript, we calculated the free energy change of reaction intermediates (*OH, *O, *OOH) on the Ru site to illustrate the effect of Si on RuO2 activity.We found that the introduction of Si slightly lowers the energy barrier of the rate-determining step (RDS, *O → *OOH), which is consistent with our experimental results that the activity of Si-RuO2-0.1 (an overpotential of 226 mV at 10 mA cm -2 ) is slightly better than that of Si-RuO2-0 (an overpotential of 248 mV at 10 mA cm -2 ).Furthermore, the PDOS calculation results show that the introduced Si weakens the covalency of the Ru-O bond, which is consistent with the XPS and XAFS results, thereby enhancing stability.
Second, in our manuscript, although the difference between the RDS of Si-RuO2 (1.823 eV) and RuO2 (1.914 eV) is only 0.091 eV, the activity difference between Si-RuO2-0.1 (226 mV) and Si-RuO2-0 (248 mV) is also small, indicating that the theoretical result is consistent with the experimental results.In addition, we summarize the relationship between the energy barrier of the RDS and the activity (overpotential at 10 mA cm -2 ) reported in the literature (Table R2).The Δη10/ΔERDS values in our work is within the range reported in the literature; therefore, we believe that the present free energy change is reasonable in our calculation.However, the following points should be addressed before considering the publication of the manuscript.

Response:
We thank the reviewer for the positive comments and helpful suggestions.
Major comments: 1.On page 3, lines 104-105, the EDS mappings of Si, Ru and O and XRD patterns of Si-RuO2-0.05,Si-RuO2-0.1 and Si-RuO2-0.2are similar, why the authors declared that the ideal doping level of Si in RuO2 interstices was around 10%.

Response:
Thanks for your comment.We list several reasons below to explain that the ideal doping level of Si in RuO2 interstices is around 10%. ① In the XRD pattern, the 2θ value of the (110) plane no longer undergoes a significant change as the introduction of Si exceeded 10%, suggesting that the lattice of RuO2 no longer expands and implying that the excess Si was unable to insert into the RuO2 lattice to change the lattice parameters of RuO2.② As seen from the EDS spectra, Si-RuO2-0.05,Si-RuO2-0.1,and Si-RuO2-0.2samples show that Si, Ru, and O were uniformly distributed throughout the entire catalyst.In the first two samples, the segregation of Si was not observed due to the insertion of Si into the RuO2 interstice; for Si-RuO2-0.2,no aggregation of Si and O was observed, which may be because a small amount of SiO2 was formed and dispersed in the sample, and EDS technology could not accurately capture the accumulation of Si and O. ③ In subsequent Si 2p XPS tests, only one characteristic peak was present in the Si-RuO2-0.05and Si-RuO2-0.1 samples; however, in the Si-RuO2-0.2samples, an additional peak consistent with the SiO2 phase appeared, suggesting that the SiO2 phase was present in the Si-RuO2-0.2samples.Based on the above results and analysis, it is reasonable to declare that the ideal doping level of Si in the RuO2 interstice was around 10%.
2. The authors mentioned that the SiO2 phase is generated when the Si content exceeds 0.1.However, no diffraction peaks corresponding to SiO2 were observed in the XRD patterns of Si-RuO2-0.2and Si-RuO2-0.3.Please explain?

Response:
Thanks for the comment.We believe that there are two main reasons why no diffraction peaks corresponding to SiO2 were observed in the XRD patterns of Si-RuO2-0.2and Si-RuO2-0.3:① Under the experimental conditions, the generated SiO2 was amorphous and could not present an obvious diffraction peak in the XRD patterns (Fig. R5); ② Due to the small amount of Si precursor added and some Si entered the lattice of RuO2, the formed SiO2 was present in a trace amount.Therefore, observing diffraction peaks corresponding to SiO2 in XRD patterns is difficult.

Response:
Many thanks for the helpful suggestion.With increasing Si content, the amount of Si-O bonds formed in the samples continuously increased; moreover, the binding energy of the Si-O bonds was much higher than that of the Ru-O bonds, which led to the binding energy of lattice oxygen constantly shifting toward a higher binding energy (Fig. R6).In terms of Ru, when the content of Si was lower than 0.1 (x≤0.1),Si was inserted into the interstitial site of RuO2, which caused the binding energy of Ru to shift toward a lower binding energy.However, when the content of Si was greater than 0.1 (0.1≤x≤0.3), extra Si existed in the form of SiO2, which could not change the electronic structure of Ru, so the binding energy of Ru was almost unchanged.4a, the stability follows an order of Si-RuO2-0.1>Si-RuO2-0>Com-RuO2.However, in Figure 4c, the Ru dissolution rate of Com-RuO2 is significantly lower than that of Si-RuO2-0; meanwhile, this phenomenon is also observed in Figure S11.Is there a contradiction?

Response:
Thank you for your careful consideration.It is well known that the higher the crystallinity, the lower the dissolution rate of ions, that is, the higher the stability (Joule  2018)).Compared with Si-RuO2-0, Com-RuO2 had a higher crystallinity (Fig. R7), suggesting that the Ru dissolution rate of Com-RuO2 was lower than that of Si-RuO2-0.However, the larger Com-RuO2 particles were more prone to peel from support materials than small Si-RuO2-0 particles, leading to a shorter operation time during the stability test, thus the stability follows an order of Si-RuO2-0.1>Si-RuO2-0>Com-RuO2 in Fig. 4a.The S-number was calculated according to the amount of produced oxygen and dissolved Ru ion (S-number = / ).Under the same current density and operation time, the amount of evolved oxygen was constant, such that the S-number was only negatively related to the amount of Ru ions dissolved.Therefore, the phenomena we observed were not inconsistent with the conclusion. 5. Page 7, line 197-199, the authors mentioned the following: "Considering that Si-RuO2-0.1 has a small particle size and porous structure compared to Com-RuO2, we deduced that the dissolution of Ru in Si-RuO2-0.1 could be further inhibited by increasing the particle size of Si-RuO2-0.1."I don't understand the basis of this deduction, please give your explanation.

Response:
Thanks for your kind reminder.After double-checked the deduction, we believe that this deduction has certain limitations.Therefore, we have removed this deduction in the revised manuscript.

Response:
Thanks for your suggestion.We have corrected this issue in the revised manuscript.

Response:
Thanks for your valuable suggestion.This information has been updated in the revised manuscript.

Response:
Thank you for your careful reading of our manuscript.The errors have been corrected in the revised manuscript.
4. Page10, Line 292, regarding the solution resistance value of 30 Ω, please provide more direct evidence.

Response:
Fig. R8b converted from the changed axis of Fig. 8a (from Fig. 3d in the original manuscript).In Fig. R8b to see the solution resistance value.When the vertical axis is zero (Z″ = 0), the intersection points between the EIS pattern of all samples and the horizontal axis are about 30 Ω, suggesting that the solution resistance value is 30 Ω. Reviewer #3 (Remarks to the Author): This manuscript reported a Si-doping RuO2 with extend OER stability in acid media, the durability looks become better, however, there are too conclusion are not solid, at first all, all figures in this manuscript are obscure (low resolution), I don't think this work is suitable for the publication in the nature communication, I suggest transfer to other journal after the revisions are done as followed: Response: Thanks for your comments and valuable suggestions.I apologize for the obscure figures in this manuscript.The obscure figures may have occurred because the PDF version you saw was converted from the original Word version, and the clear figures were compressed and became blurred during this process.In fact, all figures are 600 dpi resolution in the uploaded original Word version and meet the requirements of Nature Communications (300 dpi or higher resolution).
1.The authors emphasized that "Si tends to interstitially insert into the RuO2 lattice rather than replace the Ru atoms" through XRD simulations and DFT calculations, but I think persuasive evidences should be provided such as spherical aberration corrected electron microscopy, in fact, experimental data about the existence of Si in RuO2 is too less.

Response:
Thank you for your valuable and thoughtful comments.According to your suggestion, we performed spherical aberration-corrected electron microscopy analysis.As expected, some Si with low imaging contrast was found in the RuO2 interstice in the HADDF-STEM image (Fig. R9), suggesting the Si was successfully inserted into RuO2 interstice.
The corresponding discussion have been updated in the revised manuscript as follow: To visually prove that Si was inserted into the RuO2 interstice, spherical aberration-corrected HAADF-STEM measurements were performed.As shown in Fig. 1d, the lattice fringes with interplanar spacings of 0.318 nm and 0.254 nm were assigned to the ( 110) and ( 101) planes of rutile RuO2, respectively.Furthermore, some isolated Si atoms with low imaging contrast, which is characteristic of light elements with lower atomic numbers, were also observed in the lattice interstices of RuO2 (Fig. 1e-f).This assertion was confirmed by atomic line profiles analysis (Fig. 1g-j).2022)).② EDS spectra.When the Si content was less than 0.2, Si was uniformly distributed throughout the whole sample, but when the Si content increased to 0.3, SiO2 was formed, indicating that Si was successfully introduced into RuO2.③ XPS.The XPS survey spectra (Fig. R10) and Si 2p spectra directly demonstrate the successful introduction of Si into RuO2.

Fig. R10. XPS survey of all Si-RuO2-x samples
2. There is also inconsistence statement about Si interstitially insert into the RuO2 lattice, in which XRD show shift, there are no obvious different from lattice spacing, why?

Response:
We greatly admire your acute insight.There are two reasons for this phenomenon: ① For XRD, when Cu is used as the radiation source, the wavelength of the X-ray is 1.542 Å (λ = 1.542Å).According to the Bragg equation (2d×sinθ = nλ), we can obtain the relationship between Δd and Δθ, which is . Given that the 2θ of the RuO2 (110) plane is about 28° (θ =14°, radian is 0.244, n=1), we obtain that the Δd Δθ value is equal to 0.043 Å, indicating that the resolution is 0.043 Å at the (110) plane of RuO2.However, the resolution of the TEM instrument (FEI Talos F200S) was 2.5 Å, suggesting that XRD had a much higher resolution than TEM.In fact, X-rays, as electromagnetic waves with short wavelengths, have less inelastic scattering with the sample; however, electrons with a resting mass have more inelastic scattering with the sample, resulting in X-rays having higher resolution for the crystal structure than electrons3.② the Random errors are inevitable when measuring crystal plane spacing in TEM images.Combining the two factors of resolution and measurement error, we believe that it is tough to observe such a slight difference in crystal spacing in TEM images.

Response:
Thank you for your careful evalution.According to your suggestion, we performed EDS mapping of carbon elements by SEM-EDS (Cautions: Considering that the carbon film was used as the sample support in the TEM-EDS test, this will seriously interfere with the measurement of trace carbon elements in the sample).As the reviewer said, carbon is uniformly distributed throughout the sample (Fig. R11c).However, this test method is not reasonable because a large amount of carbon-containing species in the air will adsorb to the sample's interior with a porous structure, consistent with the conclusion that the C1s peak is always present in the XPS survey.
We performed a TGA test to determine whether there was residual carbon in the sample.As shown in Fig. 11g, the TGA results revealed that the mass fraction of carbon in the sample did not exceed 0.9 %.Furthermore, the C content was determined by a carbon sulfur analyzer (LECO CS230).The test result suggests that the carbon mass fraction in the Si-RuO2-0.1 sample is 0.0024%.The two results implied that almost no carbon present in the sample.
We made considerable effort to explore the preparation conditions for entirely removing carbon in the early exploration.Finally, we determined the optimal conditions for sample preparation.Specifically, 1.5 g of CER powder with TEOS and RuCl3 (volume of approximately 2.4 mL) was spread in a porcelain boat (10.5 cm×4.5 cm), making its thickness approximately 0.5 mm (Fig. 11hi).Subsequently, during the annealing process, the air was also continually injected into the muffle furnace by an air pump (1.2 L/min).These strategies are conducive to the full reaction of the CER with air.In addition, it has been reported that an annealing temperature of 450 ℃ can completely remove carbon (Nat. Mater. 22, 100-108 (2023), Adv. Mater. 30, 1801351 (2018); Nat Commun 10, 4. The similar work (Adv. Sci. 2023, 2207429) should be cited and discussed.

Response:
Thanks for your kind reminder.Similar work has been cited in the revised manuscript.
After carefully reading and comparing both, there is a fundamental difference.In this similar work, the researchers claimed the importance of the Ru-Si bond in enhancing both the activity and stability of the Si-RuOx@C catalyst; however, we pay more attention to improving the stability of RuO2 by introducing Si element in our work and believe that the excellent stability can be attributed to the following three factors: ① higher bond dissociation energy of the Si-O bond; ② low Ru-O bond covalency; and ③ the acid resistance of Si.Therefore, our idea is completely different from this similar work.
In addition, we still have some confusion about some of the viewpoints in this article, for example: 1) The authors mistakenly cited the EXAFS results of a RuSi alloy rather than a Ru-based oxide to confirm the formation of a Ru-Si bond (located at 2.0 Å) (Fig. R12a-b).According to the previously reported works (Fig. R12c-f), the peaks at 2.0 Å should be assigned to the Ru-Cl bond, not the Ru-Si bond.In the RuSi alloy, Ru with a negative charge bonded with Si with a positive charge to form the Ru-Si bond (Fig. R12g-h).However, both Ru and Si in the Si-RuOx@C catalyst exhibited a positive charge, as evidenced by XANES spectra (Fig. R12i).Therefore, the formation of the Ru-Si bond is impossible in the Si-RuOx@C catalyst.2) The FT-EXAFS result of the Si-RuOx@C catalyst is inconsistent with its XRD pattern.The Si-RuOx@C catalyst showed a different FT-EXAFS curve than the RuO2 catalyst (Fig. R13a), and the typical Ru-O-Ru bond at 3.2 Å in RuO2 disappeared in the Si-RuOx@C catalyst.Interestingly, the XRD pattern confirmed that the Si-RuOx@C catalyst mainly existed in the form of RuO2 (Fig. R13b); however, the Ru-O-Ru local structure with di-μ-oxo and μ-oxo configurations (Fig. R13c) was not present in the Si-RuOx@C catalyst (Fig. R13a).It is thus not believable.

5.
In line 77 of page 3, the authors stated, "…and then almost unchanged as the Si content further increased from 0.1 to 0.3." that means when the Si is beyond 0.1, it cannot be doped into RuO2, the extra Si will form SiO2, if like this situation, the author should characterize how many SiO2 formed on RuO2, this SiO2-RuO2 conductivity is better than commercial RuO2?Its activity still beyond the commercial RuO2?

Response:
Thank you for your thoughtful evaluation.
It is impossible to determine the precise amount of SiO2 is generated on RuO2 based on current characterization techniques.Moreover, it is not significant to our work to determine the amount of SiO2 because SiO2 as an insulator is harmful to enhancing activity.
The conductivity of SiO2-RuO2 is worse than that of commercial Com-RuO2 because SiO2 is an insulator.
There are two main reasons why the activity of Si-RuO2-0.2and Si-RuO2-0.3samples containing SiO2 is superior to that of Com-RuO2: ① The Si-RuO2-0.2 and Si-RuO2-0.3catalysts have smaller particles and abundant pore structures, which can provide more active sites for the OER reaction; ② Although both samples contained inert and insulative SiO2, some Si was still inserted into RuO2 interstices and improved the intrinsic activity of RuO2.Hence, their activity still outperforms that of Com-RuO2.
6.In Fig. 2b, there is a much large offset of Si-RuO2-x relative to SiO2, why?

Response:
In Si-RuO2-x samples, there are abundant Ru-O bonds and a few Si-O bonds, so the Ru-O bonds mainly determine the binding energy of O.However, in SiO2, there are only Si-O bonds.Hence, the Si-O bond entirely governs the binding energy of O. Compared with the binding energy of the Si-O bond, the binding energy of the Ru-O bond is lower, so the O binding energy of the Si-RuO2-x sample is lower than that of SiO2 (Fig. R14).Therefore, there is a much larger offset of Si-RuO2-x relative to SiO2. 7.After the stability test, whether the morphology, metal content and metal valence of Si-RuO2-0.1 catalyst changed.Please supplement the series of characterization after the stability test.

Response:
Thanks for your comments on how to improve the paper.According to your suggestion, we supplemented the series of characterizations (e.g., XRD, TEM, XPS) after the stability test, and the corresponding descriptions have been provided in the revised manuscript and Supporting Information as follows: Furthermore, a series of characterizations, including XRD, TEM and XPS, were performed for the spent Si-RuO2-0.1 catalysts to investigate the structural evolution of Si-RuO2-0.1.Obviously, the crystalline structure and morphology of Si-RuO2-0.1 still maintained its integrity after the 800h stability test (Fig. 5a-b and Supplementary Fig. 13).Meanwhile, Si, Ru and O were also uniformly distributed in Si-RuO2-0.1,further illustrating the excellent stability of Si-RuO2-0.1 toward acidic OER (Fig. 5c).To prove that the introduction of Si highly improved dissolution and oxidation resistance of RuO2 toward the acidic OER, the chemical state changes for Ru and O in Si-RuO2-0.1 before and after the 24 h stability test were further investigated and compared with those of Si-RuO2-0 and Com-RuO2 (Caution: The spent Com-RuO2 sample was only tested for 18 h).For the Ru 3p spectra of Si-RuO2-0.1, the Ru >+4 / Ru +4 value increased from 0.34 to 0.39 after the stability test, indicating the inevitable oxidation of catalysts under a high anode potential.Despite this, for the Si-RuO2-0 and Com-RuO2 samples, the change in the Ru >+4 / Ru +4 value is more significant, increasing from 0.43 to 0.58 and from 0.38 to 0.49, respectively (Fig. 5d and Supplementary Fig. 14).This result revealed that the introduced Si kept the Ru from overoxidation during the OER.Likewise, the evolution of oxygen species is also revealed by combining O 1s spectra before and after the stability test.As displayed in Fig. 5e and Supplementary Fig. 15, the remarkable increase in the OV/OL value for Si-RuO2-0 and Com-RuO2 suggests that the lattice O is involved in O2 generation to a large extent, which will accelerate the dissolution of active Ru species.In contrast, the OV/OL value was only slightly increased from 1.37 to 1.41 for Si-RuO2-0.1,indicating that the AEM pathway dominated the OER process rather than the LOM pathway.

Response:
Thanks for your valuable suggestion.Per your suggestion, we performed EPR tests on the Si-RuO2-0.1 sample before and after the stability test.As shown in Fig. R16, the EPR signal intensity at 3513 G (g=2.001) attributed to oxygen vacancies hardly changes, proving that Si prevents Ov formation during the OER process.
The corresponding discussion has been added to the revised manuscript as follows: This assertion was confirmed by electron paramagnetic resonance (EPR) (Fig. 5f), in which the signal intensity of OV at around 3513 G (g = 2.001) showed no obvious change.9. Si content in the catalyst needs to be further determined by ICP.

Response:
you for your concern.The Si content in representative Si-RuO2-0.1 catalysts was determined by ICP (Table R3).The experimental results revealed that the atomic ratio of Si and Ru was close to the expected values.structures and calculated adsorption energies in SI.This will help the readers to examine the reliability of the predictions.The authors did not give all the information in this version of the manuscript.In addition, this reviewer pointed out that "the presented free energy change due to Si doping is within the uncertainty of the DFT method.".The authors failed to address this issue in the revised manuscript.

Response:
According to your suggestions, all possible Ru 16 Si 2 O 32 model have been reconstructed and optimized based on the predicted Si content in the main text (the molar ratio of Si and Ru is about 10%).In detail, a Si atom first occupies A-site and serves as the reference position (Fig. R2).Next, another Si atom can be placed at one of the A, B, C and D-site.There is only one case for A-site, whereas there are two cases for B, C, and D-site (top or bottom).Then, the energies of different structures were calculated.Finally, based on the principle of lowest energy, the most stable structure (A-C-b model) was selected to perform subsequent operations (Fig. R2f and Fig. R3).The detailed information on the modeled structures after optimization (Fig. R2 and Table R1) have been provided in the revised manuscript (Supplementary Figure 8, Supplementary Figure 9 and Supplementary Table 2).According to the optimized bulk structure, the partial density of states (PDOS) calculations were first carried out on O atoms and the unsaturated Ru on Ru 16 O 32 and Ru 16 Si 2 O 32 (110) plane (Notes: Based on previous reports, the unsaturated Ru sites have been considered to be catalytic active centers, and (110) plane has been identified as the most stable surface in RuO 2 ).The calculation results shown that, after introducing Si into the interstitial sites, the surface Ru 4d band center (ɛ d ) upshifted from -1.759 eV to -1.626 eV, while the surface O 2p band center (ɛ p ) downshifted from -3.365 eV to -3.850 eV, suggesting that the gap between ɛ d and ɛ p is obviously enlarged (Fig. R3a).Further, we performed the PDOS calculation on their bulk (Fig. R4).Although the band center of Ru 4d on the bulk and the (110) plane present different trends, the gap between Ru 4d band center and O 2p band center is still enlarged due to the downshift of the O 2p band center (Fig. R4).This is still consistent with the conclusion mentioned in the original manuscript that the introduction of Si weakens the covalency of Ru-O bonds.The calculated results have been revised and the detailed information have been provided in the revised manuscript (Figure .2f and Supplementary Figure 10).Mater. 30, 1801351 (2018)).However, Si-RuO 2 exhibited a lower free energy barrier (1.868 eV).
The calculated results (Fig. R5, R6 and Table R2) have been revised and the detailed information have been provided in the revised manuscript (Figure .3f, Supplementary Figure 11 and Supplementary Table 3).

Fig
Fig. R1.a, HAADF-STEM image of Si-RuO2-0.1.b-c, High-resolution HAADF-STEM image obtained from the area highlighted with purple and green in Fig. 1a.d-g, Line-scanning intensity profile obtained from the area highlighted with red lines in Fig. 1b and c.

Fig
Fig. R2.a, HAADF-STEM image of Si-RuO2-0.1.b-c, High-resolution HAADF-STEM image obtained from the area highlighted with purple and green in Fig. 2a.d-g, Line-scanning intensity profile obtained from the area highlighted with red lines in Fig. 2b and c.

Fig
Fig. R3.a-q, Structural models with different Si doping amounts and doping mode in RuO2 (Cautions: SUB1-2 represents the second model in which a Si atom replaces the Ru site, INT2-1 represents the first model in which

Fig. R4 .
Fig.R4.Theoretical calculations of acidic OER activity on the established model of RuO2 and Si-RuO2.

Fig. R5 .
Fig. R5.XRD pattern of SiO2 prepared under the same conditions

Fig. R6 .
Fig. R6.The O1s binding energy of several typical oxygen-containing species (from Handbook of X-ray Photoelectron Spectroscopy).

Fig
Fig. R8.a, EIS plots of different catalysts from Fig. 3d in original manuscript; b, magnification of Fig. 3d in the original manuscript.

Fig
Fig. R9.a, HAADF-STEM image of Si-RuO2-0.1.b-c, High-resolution HAADF-STEM image obtained from the area highlighted with purple and green in Fig. 8a.d-g, Line-scanning intensity profile obtained from the area highlighted with red lines in Fig. 8b and c.

Fig
Fig. R11.a-f, SEM image of Si-RuO2-0.1 and the corresponding EDX elemental maps; g, TGA and DTA curve of Si-RuO2-0.1 sample; h, Optical image of 1.5 g CER in sample tube; i, Optical image of 1.5 g CRE spread in porcelain boat.

Fig. R14 .
Fig. R14.The O1s binding energy of several typical oxygen-containing species (from Handbook of X-ray Photoelectron Spectroscopy).

Fig. R2 .
Fig. R2.The modeled structures of a, Ru16O32 and b-h, different Ru16Si2O32 after optimization (denoted A-M-X, where A represents the first Si atom occupies the A-site; M represents the position occupied by the second Si atom, M = A, B, C or D; X represents the spatial position located by second Si atom, X = t or b, t is top, b is bottom).

Fig. R3
Fig. R3 Gibbs free energy of different structural models.

Fig. R4
Fig. R4 PDOS plots of the Ru 4d and O 2p orbitals of a, Si-RuO2 slab and RuO2 slab and b, Si-RuO2 bulk and RuO2 bulk.

Fig. R6
Fig. R6 Theoretical calculations of the acidic OER activity of the established models of on a, Si-RuO2 and b, RuO2.

Table .
R1. Lattice parameters and unit-cell volume of the modeled structures after optimization.
4. The authors did not provide convincing explanation to the observed enhancement in both activity and durability of the Si-doped catalysts.First, no detailed information of the modelled structures is given.It is unclear if the normally-assumed O-saturated surface structures were employed in these computations.All the structures and adsorption energies should be given in SI.Secondly, the presented free energy change due to Si doping is within the uncertainty of the DFT method.

Table .
R2.The relationship between the energy barrier of RDS and activity.This manuscript reports a strategy to improve the stability of RuO2 by introducing Si element to the lattice interstices of RuO2.The authors attribute the excellent stability to the following three aspects: 1) higher bond dissociation energy of the Si-O bond; 2) low Ru-O bond covalency; and 3) the acid resistance of Si.It is a very interesting idea.Meanwhile, the authors provide enough evidence to support this story.To be specific, in the structural characterization part, the author first verified that Si was inserted into the RuO2 interstices by XRD, the simulated XRD and DFT calculations.Furthermore, combined TEM-EDS and XRD results, they deduced that the ideal doping level of Si in RuO2 interstices was about 10%.To study the Ru-O bond covalency, the authors present XPS and XAFS measurements and DFT calculations, indicating that Si elements in interstitial sites play a key role in weakening the covalency of Ru-O bonds.In the electrochemical test part, the authors demonstrated that the addition of Si not only has a positive effect on OER activity, but also inhibited the dissolution of Ru ions, resulting in excellent stability.Finally, DEMS measurement directly demonstrated that the lattice oxygen oxidation pathway was markedly suppressed due to the introduction of Si.Overall consideration, I recommend it publication in Nature Communications.

Table .
R3. ICP MS analysis of Si and Ru in Si-RuO2-0.1 catalysts

Table R1 .
Lattice parameters and unit-cell volume of the modeled structures after optimization.

Table R2
The detailed information for the free energy change of reaction intermediates (*OH, *O, *OOH) on the unsaturated Ru site.