Structural and Mechanical Properties of Dickite and Nacrite Minerals: A Computational Study
, Minh Phi Nguyen 2,3, Hai Hoang 2,3, Nguyen Van Phuoc 4,*, Suresh Alapati 5,* and Kwang-Il Jeong 6
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors1. I believe the authors could provide a more detailed discussion on the significance of these types of minerals, their applications, and the importance of the evaluated parameters. A more specific and focused introduction would strengthen the manuscript, particularly by incorporating insights from the following references:
10.1016/0169-1317(91)90014-Z
10.1016/j.eti.2020.100692
10.1021/acs.jpcc.4c04280
10.1016/j.oregeorev.2023.105446
2. I recommend that the authors maintain consistent accuracy for the reported values and provide possible error margins for their results. This addition would enhance the reliability and transparency of the study.
3. Significant discrepancies remain between the findings of this study and those reported in the literature. Could the authors investigate the possible reasons for these variations? Furthermore, it would be beneficial to specify which results, and under what conditions, best represent the actual structures and mechanical properties of the minerals studied.
Author Response
Comment 1. I believe the authors could provide a more detailed discussion on the significance of these types of minerals, their applications, and the importance of the evaluated parameters. A more specific and focused introduction would strengthen the manuscript, particularly by incorporating insights from the following references:
10.1016/0169-1317(91)90014-Z
10.1016/j.eti.2020.100692
10.1021/acs.jpcc.4c04280
10.1016/j.oregeorev.2023.105446
Response 1: - We would like to thank the reviewer for her/his valuable suggestion. In response, we have revised the Introduction to provide a more detailed discussion on the industrial and scientific significance of kaolin group minerals based on the suggested references, as follows:
The kaolin group minerals are among the most abundant clays in the Earth’s crust [1], extensively mined for numerous scientific and industrial applications, including geophysics, geomechanics, geochemistry, environmental engineering, ceramics, fillers in polymer nanocomposites, and various other applications [1-8]. These diverse applications stem from their distinctive physical and chemical properties, such as softness, low abrasiveness, ease of dispersion in water, chemical stability across a wide pH range, high adsorption capacity for heavy metals, and suitability as functional fillers in various formulations [1-3, 6-8]. Kaolinite, dickite, and nacrite are the primary polytypes of this kaolin group [1, 9]. Therefore, determining properties, particularly the structural and mechanical ones, of these minerals is of critical importance, as these characteristics underpin their functionality in many applications [1-8].
- Three of the suggested references have been added to strengthen the discussion:
10.1016/0169-1317(91)90014-Z
10.1016/j.eti.2020.100692
10.1016/j.oregeorev.2023.105446
- After careful review, we found the remaining reference (10.1021/acs.jpcc.4c04280) to be less relevant to the current study and have decided not to include it.
Comment-2. I recommend that the authors maintain consistent accuracy for the reported values and provide possible error margins for their results. This addition would enhance the reliability and transparency of the study.
Response-2: - We sincerely thank the reviewer for this valuable suggestion. We have re-presented the values with consistent accuracy. In particular, the bulk and shear moduli for Dickite and Nacrite are now consistently reported with the same level of accuracy across the text and the tables, as indicated in the revised manuscript.
- Regarding the suggestion to calculate error margins for the reported results, we agree that this would enhance the transparency and reliability of the study. However, the methodology employed in this research relies on minimization processes using numerical approaches, where a set of initial parameters, i.e. a set of initial interaction potential parameters, produces unique set of outputs. As a result, error margins could not be calculated, consistent with previous studies employing similar methodologies, which also did not report the error margins [e.g. Karmous (2011, 2012); Benazzouz and Zaoui (2012)]. This point has been explicitly stated in the revised manuscript. Furthermore, the results obtained from this methodology show good agreement with available numerical and experimental data, confirming the reliability and accuracy of the approach.
- To calculate error margins, one would need to perform multiple simulations with various interaction potential parameters. This would require significantly much computational resources and time, which were beyond the scope of the current study. Nonetheless, we acknowledge the importance of this approach and will consider incorporating it in future work where computational resources and time are available. This discussion has been added to our new version (lines 182~185)
Comment-3. Significant discrepancies remain between the findings of this study and those reported in the literature. Could the authors investigate the possible reasons for these variations? Furthermore, it would be beneficial to specify which results, and under what conditions, best represent the actual structures and mechanical properties of the minerals studied.
Response-3: - This study employs molecular simulations to determine various structural and mechanical properties of Dickite and Nacrite. As discussed in the Introduction, the results of this methodology depend on the interaction potential parameters used. In this work, we applied a relaxed fitting approach to adjust the interaction potential parameters initially proposed by Schröder et al. for Zeolite to yield accurate results for structural properties. Consequently, discrepancies arise between our results and those obtained from molecular simulations using the interaction potential parameters of Schröder et al., as applied by Karmous [Karmous, 2012].
- Regarding the discrepancies between our results and those from DFT simulations, these can be attributed to the fundamental differences in the two approaches. DFT simulations are ab-initio methods that rely on approximations for exchange-correlation energy and are typically limited to systems with a small number of atoms. In contrast, molecular simulations require experimental data to determine interaction potential parameters but can be applied for larger atomic systems.
- The results presented in this manuscript correspond to zero-pressure conditions, as explicitly stated in the revised manuscript. As mentioned above, the interaction potential parameters were adjusted to yield accurate results for structural properties, making these properties the most reliably predicted. However, the reported mechanical properties also show good agreement with experimental and DFT simulation data, confirming the reliability of the methodology employed in this study. Furthermore, we plan to extend this work by using the interaction potential parameters obtained from this work to investigate the effects of pressure on structural and mechanical properties, as outlined in the Conclusion as a perspective for future research.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsManuscript ID: Minerals-3354632
Structural and Mechanical Properties of Dickite and Nacrite Minerals:
A Computational Study, by Brahim K. Benazzouz, Minh Phi Nguyen, Hai
Hoang, Nguyen Van Phuoc, Suresh Alapati, and Kwang-Il Jeong
This article presents a study of the structure of Dickite and Nacrite, which are clays of the kaolin group, by using molecular simulations and proposing different parameterizations for the interaction potentials. I consider that this work has valuable results, considering that predicts properties that are difficult to determine experimentally in these clays. In addition, the manuscript is well written. Nevertheless, in spite of all this, I found some very minor issues that should be addressed. Therefore, I recommend a minor revision of the manuscript according to the follow:
Section 1. Introduction
Line 37. Check dotted points at the end of the first sentence.
Line 55. ‘... or generalized...’
Line 58. Here references 13 and 14 are included as numbers at the
end of a sentence, and in the next one authors are mentioned, but
the connection between both is not straightforward. Please check
that considering the journal instructions (references format). This
also repeats at lines 69-70. Please check the rest of the text.
Section 2. Models and Computational Methods
Line 120. It is said that ‘... in this work the short-range
interactions between the hydroxyl hydrogen and ions was assumed
to be zero’. A reason for this and/or some references should be
included.
Line 142. ‘.. the hydrogen assigned a charge of +0.426e and the oxygen a charge of −1.426e’. If these particular charge values are consequence of fixing OH, dipole moment, it should better stated.
Line 172. Table 1 is mentioned as a source of the initial structures,
which are experimental (refs 22 and 23), but this table also shows
lattice parameters obtained in this and other works (i.e., results).
So it is quite confusing the role of this table in Section 2. I suggest moving this Table to the Results section, where the majority of its content is analyzed and compared (subsection 3.2). Note that Table 1 caption also indicates that its content is for comparison of the results with reference data, so in the Results section should be better placed. Then, in section 2 (’computational methods’) the starting (experimental) lattice parameters could mentioned as it is in the text, and, if necessary, their values can be included in the supplementary material together with the atomic positions that are in there.
Line 179. What are the units for the tolerance parameters?
Section 3. Results
Table 2. ‘shell’ instead of ‘shel’
Line 198. Table 2 was already introduced. So, I suggest ‘are also presented’ instead of ‘are presented’.
Line 250. When compare with Kaolinite, please include the reference (I think is ref. 14).
Author Response
Comment 1: Line 37. Check dotted points at the end of the first sentence.
Response 1: - "…" has been replaced by "and various other applications"
Comment 2: Line 55. ‘... or generalized...’
Response 2: - "or" has been added.
Comment 3: Line 58. Here references 13 and 14 are included as numbers at the end of a sentence, and in the next one authors are mentioned, but the connection between both is not straightforward. Please check that considering the journal instructions (references format). This also repeats at lines 69-70. Please check the rest of the text.
Response 3: - We sincerely thank the reviewer for pointing out this issue. In the revised manuscript, we have consistently clarified the connection between numerical references and the associated authors, as follows:
"Molecular simulations provided bulk modulus values between 44 and 75 GPa and a shear modulus around 25 GPa, which is influenced by the parameterization of interaction potentials [13-14]. It is worth noting that Karmous [13] used the interaction parameters proposed by Schröder et al. [17], whereas Benazzouz and Zaoui [14] employed the relax fitting method to adjust these parameters, yielding results that better align with experimental data."
"For molecular simulations, Karmous [13] used the interaction potential parameters proposed by Schröder et al. [17], which are the same as those for Kaolinite, to calculate a few of the properties of Nacrite with the bulk modulus of 56.41GPa [18]."
- In addition, we have carefully reviewed the entire manuscript to ensure consistency in referencing format and clarity throughout the text.
Section 2. Models and Computational Methods
Comment 4: Line 120. It is said that ‘... in this work the short-range interactions between the hydroxyl hydrogen and ions was assumed to be zero’. A reason for this and/or some references should be included.
Response 4: - In this work, the interaction potentials used are based on those of Schröder et al., where short-range interactions between the hydroxyl hydrogen and ions are neglected, except for O2-, which has relatively small contributions compared to other ions. In addition, our previous study using the same methodology have shown that neglecting these short-range interactions still yields results in good agreement with experimental data. Therefore, for computational efficiency, we have assumed these short-range interactions to be zero in this work.
- Two references (Schröder et al., 1992; Benazzouz and Zaoui, 2012) supporting this assumption have been added at the end of the sentence in the revised manuscript.
Comment 5: Line 142. ‘.. the hydrogen assigned a charge of +0.426e and the oxygen a charge of −1.426e’. If these particular charge values are consequence of fixing OH, dipole moment, it should better stated.
Response 5: - Yes, the reviewer is correct that these particular charge values are consequence of fixing OH, dipole moment. This is explicitly stated in the preceding sentence as:
"For the hydroxyl (OH) groups, partial charges are assigned to the oxygen and hydro-gen atoms to accurately represent the dipole moment of the OH group, while main-taining a formal charge of -1e for the hydroxyl ion as a whole [14, 17]. This negative charge is distributed between the oxygen and hydrogen within the hydroxyl group, with the hydrogen assigned a charge of +0.426e and the oxygen a charge of −1.426e."
Comment 6: Line 172. Table 1 is mentioned as a source of the initial structures, which are experimental (refs 22 and 23), but this table also shows lattice parameters obtained in this and other works (i.e., results). So it is quite confusing the role of this table in Section 2. I suggest moving this Table to the Results section, where the majority of its content is analyzed and compared (subsection 3.2). Note that Table 1 caption also indicates that its content is for comparison of the results with reference data, so in the Results section should be better placed. Then, in section 2 (’computational methods’) the starting (experimental) lattice parameters could mentioned as it is in the text, and, if necessary, their values can be included in the supplementary material together with the atomic positions that are in there.
Response 6: We agree with the reviewer that Table 1, which contains both the initial experimental lattice parameters (refs 22 and 23) and the calculated lattice parameters from this and other works, is better placed in the Results section, as its primary role is to compare results. Accordingly, we have moved Table 1 to subsection 3.2 (Results), and it is now presented as Table 2.
Comment 7: Line 179. What are the units for the tolerance parameters?
Response 7: The units of the tolerance parameters typically depend on the specific optimization function and variables being considered. However, in the GULP software, the tolerance parameters may have no physical units. For instance, the optimization equation (e.g., Equation 5 in this study) simultaneously applies to quantities with different units, such as lattice parameters (in Å), angles (in degree) and fractional coordinates (dimensionless). In such case, the tolerance parameter for the function only represent a magnitude for convergence and do not carry specific units. Thus, for simplicity and clarity, the units for the tolerance parameters have been omitted for all cases.
Section 3. Results
Comment 8: Table 2. ‘shell’ instead of ‘shel’
Response 8: - It has been corrected.
Comment 9: Line 198. Table 2 was already introduced. So, I suggest ‘are also presented’ instead of ‘are presented’.
Response 9: - It has been changed.
Comment 10: Line 250. When compare with Kaolinite, please include the reference (I think is ref. 14).
Response 10: - The reference has been included.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe authors made the force fields of dickite and nacrite by GULP and its Python utilities. They evaluated their structural and mechanical properties, and there was a good agreement between their theoretical values and those obtained by the experiments. The manuscript is well written. I recommend that this manuscript be published in Minerals.
I recommend that the structural data shown in Supplementary Information be published in CIF format.
Author Response
Comment 1: I recommend that the structural data shown in Supplementary Information be published in CIF format.
Response 1: - The structure data obtained from this work have been also presented in the CIF format.
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsComments
The manuscript, entitled “Structural and Mechanical Properties of Dickite and Nacrite Minerals : A Computational Study”, presents a detailed work on the atomic modeling and simulation about Dickite and Nacrite. The interaction potential parameters proposed in this work may be useful for investigating these minerals via MD simulation method. However, the manuscript needs carefully revised before publishing. Special comments are as follows:
1. Why there is no Morse potential parameters for Dickite O core - H core in Table 2?
2. Line 269, the equation number (11a) should be separated from the equation to avoid misunderstanding.
3. In Table 5, the B0 of Dickite found in this work is significantly greater than those found with DFT, how do you explain this difference?
4. The simulation results of Dickite and Nacrite were compared with those of Kaolinite, and it will be better to present also the corresponding properties of Kaolinite in the tables, which may be more clear for readers to see the difference between these minerals.
5. CLAYFF was a force field specially for clay minerals, which is widely applied in MD simulations. Have you try CLAYFF for Dickite and Nacrite? The force field model proposed in this work is really complex comparing with CLAYFF. Since the cost of MD calculation with big system is really important, a simple forcefield may save a lot of calculation time. How do you balance the accuracy with the efficiency of the simulation?
Comments for author File: Comments.pdf
Author Response
Comment 1: Why there is no Morse potential parameters for Dickite Ocore - Hcore in Table 2?
Response 1: - Table 2 presents only the adjusted potential parameters for Dickite and Nacrite. For the Ocore–Hcore interaction in Dickite, the Morse potential parameters were not adjusted and are directly adopted from Schröder et al. This was clearly stated in the manuscript.
Comment 2: Line 269, the equation number (11a) should be separated from the equation to avoid misunderstanding.
Response 2: - It has been corrected.
Comment 3:. In Table 5, the B0 of Dickite found in this work is significantly greater than those found with DFT, how do you explain this difference?
Response 3: - We sincerely thank the reviewer for pointing out this interesting observation. DFT simulations are ab-initio methods that rely on approximations for exchange-correlation energy, whereas molecular simulations depend on the parameterization of interaction potentials between atoms. Hence, directly comparing results from these two approaches is not straightforward due to their fundamentally different methodologies. However, the B0 value of DFT simulations smaller than that of this work can be attributed to the approximations used in the DFT pseudopotentials that could result in weaker interaction potentials compared to the parameters derived in this study.
Comment 4:. The simulation results of Dickite and Nacrite were compared with those of Kaolinite, and it will be better to present also the corresponding properties of Kaolinite in the tables, which may be more clear for readers to see the difference between these minerals.
Response 4: - The results for Kaolinite have been mentioned to support the discussion on the results for Dickite and Nacrite. Including these results directly in the main manuscript could potentially cause confusion regarding the primary focus of the study. Furthermore, adding these results to the tables in the main text would compromise their conciseness and clarity. Therefore, we have consolidated the corresponding properties of Kaolinite in supplementary materials to ensure the manuscript remains focused and well-organized.
Comment 5: CLAYFF was a force field specially for clay minerals, which is widely applied in MD simulations. Have you try CLAYFF for Dickite and Nacrite? The force field model proposed in this work is really complex comparing with CLAYFF. Since the cost of MD calculation with big system is really important, a simple forcefield may save a lot of calculation time. How do you balance the accuracy with the efficiency of the simulation?
Response 5: - We do agree with the reviewer that CLAYFF is a force field specially for clay minerals, and has been widely applied in MD simulations. However, as emphasized by Cosenza et al., 2023 [Ref. 12] this force field significantly overestimates the modules in the plane of clay sheet. The mathematical formulation of the optimal force field for clay minerals is still the subject of ongoing important research. Hence, we have not tested the CLAYFF force field for calculating the structural and mechanical properties of Dickite and Nacrite. However, we sincerely thank the reviewer for pointing out this point, which is a perspective for our future research.
- We do also agree with the reviewer that the force field used in this work is more complex than the CLAYFF one, which increases the computational time. However, calculating the structural and mechanical properties of these minerals does not need to set up big systems, the computational cost is manageable for the force field employed in this study. In fact, its primary weakness is polytypic-specific dependence, which is explicitly stated in the revised manuscript.
Author Response File: Author Response.pdf
