Atomic-Scale Insights into Surface Reconstruction and Dissolution of Hematite: The Formation of Water Cages and Protonation Effects
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis work disscuss atomic dissolution on the basis of metadynamics, the typical method applied for this type of problem. There many works that provide a very similar analysis such as those below and should be cited:
J. Phys. Chem. C 2018, 122, 28, 16086–16091
DOI: https://doi.org/10.1021/acs.jpcc.8b03743
J. Phys. Chem. Lett. 2018, 9, 7, 1809–1814
DOI: https://doi.org/10.1021/acs.jpclett.8b00484
Besides, authors show Mayer bond order which show a very pronound separation between two cluster and no detailed explanation was provided. Is ther any physical intuition behind the two cluster of data points observed on Figure 1(d)?
Another genral comment is about quality of figures. An example is the figure legends in figure 1(b) which does not show formatted inder and keep underscore which seems could be improved.
The last final comment is that the particular free energy barrier found by the authors is very particular to the site type and this should be emphasized before making it a general feature of the material as a whole. This is explained in the references mentioned above.
Author Response
Comments 1:
This work disscuss atomic dissolution on the basis of metadynamics, the typical method applied for this type of problem. There many works that provide a very similar analysis such as those below and should be cited:
- Phys. Chem. C 2018, 122, 28, 16086–16091
DOI: https://doi.org/10.1021/acs.jpcc.8b03743
Phys. Chem. Lett. 2018, 9, 7, 1809–1814
DOI: https://doi.org/10.1021/acs.jpclett.8b00484
Materials 2025 Jan 24;18(3):538. DOI: 10.3390/ma18030538
Respond 1: We sincerely thank the reviewer for pointing out these important studies. Following your suggestion, we have incorporated the recommended references (J. Phys. Chem. C 2018, J. Phys. Chem. Lett. 2018, and Materials 2025) into the Introduction and Methods sections.
Comments 2: authors show Mayer bond order which show a very pronound separation between two cluster and no detailed explanation was provided. Is ther any physical intuition behind the two cluster of data points observed on Figure 1(d)?
Respond 2: I'm sorry for my mistake that caused you a bad reading experience. The description at line 121 should be the Mayer bond level description shown in Figure 1c, which has now been corrected.
Comments 3: Another genral comment is about quality of figures. An example is the figure legends in figure 1(b) which does not show formatted inder and keep underscore which seems could be improved.
Respond 3: I agree with your comment and we have corrected figure 1(b), the figure legends have been improved into underscore.
Comments 4:The last final comment is that the particular free energy barrier found by the authors is very particular to the site type and this should be emphasized before making it a general feature of the material as a whole. This is explained in the references mentioned above.
Respond 4:
We completely agree with the reviewer’s perspective that the dissolution mechanisms and energy barriers are highly site-specific. The local coordination environments at different sites, such as terraces, steps, and kinks, vary significantly in their connectivity to the bulk and the solution phase, leading to distinct thermodynamic and kinetic behaviors. As suggested, we have added the recommended citations to the Introduction to acknowledge these site-specific differences.
Regarding the choice of the specific site in this work, our motivation was partly inspired by the experimental observations of Eggleston et al., who noted that even atomically flat hematite surfaces away from steps undergo significant morphological changes—developing "dimples and bumps" (surface reconstruction)—after exposure to high-temperature acidic environments. To elucidate the underlying mechanism of this phenomenon, we specifically chose to model the dissolution/reconstruction of an atom on a flat terrace site. By focusing on this site, we were able to observe how individual atoms initiate movement from the matrix and why they tend to reconstruct rather than fully dissolve under certain conditions.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsPlease find the review report in the attached file.
Comments for author File: Comments.pdf
Author Response
Comments 1: In the Results, the authors state that the simulated water density corresponds to ~55.5 mol/L (i.e., ~1 g/mL). However, in the Methods they report building the box with 69 water molecules at “1 g/L,” which is three orders of magnitude lower. This inconsistency should be corrected or clearly explained.
Respond1: Thank you for pointing this out. I have fixed the density into 1 mg/L.
Comments 2:The authors report identifying a minimum free energy pathway using a “pathway search algorithm,” but they do not specify which algorithm was used (e.g., method, discretization, grid resolution, or related settings).
Respond2: We apologize for the lack of technical detail regarding the pathway search. The minimum energy pathway (MEP) was identified using the MEPplot algorithm based on the string method(available at [https://github.com/XinChenQC/MEPplot]), and the detailed have been added into the part3.3 metadynamics simulation.
Comments 3:The manuscript reports a correlation between Fe–O Mayer bond orders and bond distances based on “single-point” calculations performed every 0.25 ps. However, the Methods describe “static calculations... diagonalization... molden files,” which suggests post-processing on selected snapshots. The authors should clarify whether the Mayer bond orders were obtained from postprocessed static calculations and, if so, provide the exact protocol.
Respond 3:We appreciate the reviewer’s keen observation regarding the calculation of Mayer bond orders. We apologize for the inadvertent omission of the specific software and protocol details in the original submission. In fact, the Mayer bond orders were systematically calculated by importing the Molden files—generated from static SCF calculations—into the Multiwfn package. We then extracted the specific bond order values between the relevant atomic indices for each sampled configuration to analyze the bonding evolution. We have now incorporated the reference for Multiwfn into part 3.4 static calculation and clarified this procedure in the revised manuscript.
Comment 4:
In lines 48–51, the authors state that hematite “is directly exposed to high temperature and strong acid environments” and therefore that understanding its dissolution mechanism is essential. However, the simulation setup does not specify the pH conditions being represented, nor does it clarify whether water dissociation was allowed/observed and to what extent (or whether it was effectively not considered). This point should be explicitly addressed so the proposed mechanism can be interpreted within a well-defined acid–base context. Because hematite is widely recognized to be only sparingly soluble under circumneutral to alkaline conditions, the discussion adds limited novelty unless the pH dependence is explicitly considered or, at minimum, the acid–base regime represented by the model is clearly defined. As the authors acknowledge around line 51 (“Thus understanding the mechanism of the dissolution process is also essential for hematite”), pH is a key variable when studying dissolution mechanisms, because the solubility of iron oxides changes markedly with pH. In fact, the authors state that solubility depends on several factors, yet they omit pH as a variable that can significantly alter solubility. For Fe oxides/hydroxides, both solubility and dissolution pathways are pH-dependent because pH controls surface protonation
2 As pH decreases, dissolution typically increases via a proton-promoted mechanism: protons facilitate the breaking of surface Fe–O/Fe–OH bonds and enhance the release of Fe into solution. At high pH, the apparent solubility of Fe(III) can increase again because hydroxo complexes become dominant, particularly Fe(OH)₄⁻, whose concentration rises strongly with pH; this effectively raises the solubility limit under alkaline conditions. In light of the above, the authors should consider simulations representative of both acidic and alkaline conditions to support a robust mechanistic claim, especially given their own motivation that hematite can be exposed to “high temperature and strong acid environments” and that understanding its dissolution mechanism is essential. As noted earlier, expanding the study to include both low- and high-pH regimes would substantially strengthen the manuscript’s novelty and impact.
Respond 4 : In response to this series of comments regarding pH values, I have the following points that need to be clarified.
1.We would like to clarify that in atomistic simulations, it is technically impossible to define a specific, continuous pH value as in experimental conditions. Instead, the acidity or alkalinity of the system is effectively represented by the discrete addition or removal of mobile hydronium or hydroxide ions within the simulation cell. As explicitly stated in the model construction section of the manuscript, our solvent layer was composed of water molecules, which establishes the foundational chemical environment for observing the interfacial reactions and proton-related behaviors described in our study.
2. A primary advantage of Ab Initio Molecular Dynamics (AIMD) is its ability to describe chemical reactions, including bond breaking and formation, without any pre-defined constraints. Therefore, the dissociation of water molecules and the subsequent protonation of the hematite surface are spontaneous processes captured within our simulation, rather than being "omitted" or "not considered."
3. the pivotal role of protons in the dissolution mechanism is a central theme of our analysis, Our Mayer bond order analysis (Fig. i (c)) explicitly quantifies the weakening of Fe–O bonds upon surface protonation, providing direct evidence of how protons activate the surface.We have also observed that dissociated protons (H atoms) can penetrate into Fe vacancies. This prevents the "re-nesting" or recombination of detached Fe atoms, thereby irreversibly driving the dissolution process. The commentator even mentioned whether water separation was not considered at all, which I think is a very irresponsible comment
4. We would like to emphasize that the primary objective of this study is to provide an atomic-level description of the fundamental events governing the dissolution or reconstruction of an individual iron atom from the hematite matrix. Our analysis is focused on the microscopic reaction coordinate rather than macroscopic thermodynamic properties such as solubility.
5.Regarding the dissolution behavior under neutral conditions, we employed well-tempered metadynamics to determine the free energy profile of the process. The calculated high energy barrier is in excellent agreement with the experimentally observed low solubility of hematite. However, low solubility does not preclude mechanistic significance; rather, it underscores a unique kinetic bottleneck that warrants atomic-level investigation.While the reviewer notes basic insights such as proton-promoted dissolution and the dominance of hydroxo complexes (e.g., Fe(OH)4-) at extreme pH, our work provides a much more granular description of these events. Specifically, as shown in Fig. 3, we explicitly captured the formation of an Fe(OH)4- cage structure during the final stage of dissolution in a neutral environment. Crucially, our simulations reveal that this cage is stabilized and "locked" by a dense interfacial hydrogen-bond network, which sterically and energetically hinders the approach of additional protons to the active site.To achieve full dissolution from this state, the system must simultaneously disrupt the three hydrogen bonds between the OH- groups and surface oxygen atoms, while also cleaving the final Fe–O bond in the absence of proton promotion. This discovery reveals the fundamental reason why hematite exhibits a high propensity for surface reconstruction despite its overall low solubility—a distinct kinetic feature that differentiates it from other minerals. Building on this foundation, it becomes clear that strong acidic or alkaline environments promote dissolution by significantly weakening the integrity of this hydrogen-bond network, thereby lowering the activation barrier for the final release of the iron species.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsAccept in present form
