Self-Diffusion in Sr-Containing Iron-Polyphosphate Glasses by Molecular Dynamics Simulations

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study uses molecular dynamic simulations to estimate diffusion coefficients and molar volumes of strontium-iron-phosphate glasses. It provides interesting results and would benefit nuclear waste management scientific community. Before I can recommend the manuscript for a publication, the authors should consider the comments below:

1.       Line 86: You mention that a fraction of Fe³⁺ was replaced by Fe²⁺. How did you compensate the charge? Did you remove oxygen atoms or replaced some of them with oxygen molecules (O₂)?

2.       Figure 4: Do you have an explanation for the decrease of diffusion coefficients of above 40 mol% at 1200 K?

3.       Figures 7 and 8: The size of the cluster in atoms is based on all atoms and not just number of atoms under investigation, right? For example, P-cluster of size 500 may contain 100 or less P atoms, right?

4.       How do you interpret the oxygen clusters? I am having trouble visualizing them. Do they represent regions with higher proportion of glass network and less modifiers or something else?

5.       I understand that you are studying a model system, but I think a role or potential impacts of other wasteform components should be discussed or at least mentioned. For example, there are studies that not all halides are volatilized during glass melting and a fraction is retained in the final glass.

 

6.       The chemical durability of waste glasses is typically studied by dissolution tests. Is there a way to compare your results to experimentally determined dissolution rates or EDS maps close to glass surface? Or more advancement in either modeling or experimental techniques is still needed? 

Comments on the Quality of English Language

The paper is clearly written but would benefit from English editing.

Author Response

Comments 1:

Line 86: You mention that a fraction of Fe3+ was replaced by Fe2+. How did you compensate the charge? Did you remove oxygen atoms or replaced some of them with oxygen molecules (O2)?

Response 1:

Thank you for pointing this out. Redundant oxygen atoms were removed.

The following sentence has been added to the manuscript in Materials and Methods (p.2, l.91-92):

“Some of the oxygen atoms were removed to ensure the charge neutrality of the system.”

 

Comments 2:

Figure 4: Do you have an explanation for the decrease of diffusion coefficients of above 40 mol% at 1200 K?

Response 2:

Thank you for the comment.

The following explanation can be found in the Discussion (p.12, l.249-265):

In the second region from c.a. 1000 K to c.a. 1900 K the glass is in liquid form. However, its viscosity is still relatively high. In this case, the process of self-diffusion is mostly related, in contrast to the previous region, to the migration of network species P and FeIII, as well. This may be related to the decomposition of the iron-phosphate network in the temperature range. The rapid decrease of the network species coefficients for x = 40 and 50 (Figure 4) is probably related to changes in the glass network structure discussed above and changes in the iron-phosphate-dominating phase to the Sr-rich phase. The second one starts to reflect the crystalline compounds that probably melt at higher temperatures. As such, the migration of P and FeIII may be limited to the small space of the residual iron-phosphate network separated by the Sr-rich. This process prevents the migration of the species outside the residual regions and limits its self-diffusion.

 

Comment 3:

Figures 7 and 8: The size of the cluster in atoms is based on all atoms and not just number of atoms under investigation, right? For example, P-cluster of size 500 may contain 100 or less P atoms, right?

Response 3:

Thank you for the comment.

No, the number of atoms means the number of atoms under investigation. So, the max size is the biggest cluster of the e.g. O atoms that are in the oxygen-enriched regions. The number of oxygen atoms that are involved in the region formation. There can be e.g. one very big cluster and many small ones in the system. The mean size is the average size of all the clusters.

 

The following sentence has been added to the manuscript (3.3. Cluster statistics) (p.9, l.187-190):

So, the max size is the biggest cluster of the e.g. O atoms that are in the oxygen-enriched regions. The number of oxygen atoms that are involved in the region formation. There can be e.g. one very big cluster and many small ones in the system.

 

Comment 4:

How do you interpret the oxygen clusters? I am having trouble visualizing them. Do they represent regions with higher proportion of glass network and less modifiers or something else?

Response 4:

Thank you for the comment.

In all cases, we define the clusters as the regions where the distance between atoms e.g. O-O is smaller than the specific cut-off radius set as the radius of the corresponding coordination sphere e.g. O-O. Thus, the clusters are regions richer in the atoms than stoichiometry indicates.

 

The following sentence has been added to the manuscript in Materials and Methods (p. 3, l. 112-113):

This approach treats the areas they are enriched with the specific atoms that stoichiometry indicates as clusters.

 

Comment 5:

I understand that you are studying a model system, but I think a role or potential impacts of other wasteform components should be discussed or at least mentioned. For example, there are studies that not all halides are volatilized during glass melting and a fraction is retained in the final glass.

Response 5:

 Thank you for the comment.

In the manuscript, we are focused on the influence of SrO on iron-phosphate glass. Introducing halides may be an interesting topic for future studies. The number of such simulations is very limited. Therefore it would be valuable to conduct such simulations that we consider in the near future.

 

The following sentences have been added to the Introduction (p.1, l.39-43):

The IPG can be used to vitrify so-called “difficult waste” which cannot be immobilized in conventional borosilicate glasses. Such waste has a high concentration of species and solubility in silicate glasses is strongly limited. The species may be sulfur and salty waste containing a high concentration of sulfates that cannot be utilized in conventional waste glasses. The solubility of the salts is much higher in phosphate glasses.

 

Comment 6:

The chemical durability of waste glasses is typically studied by dissolution tests. Is there a way to compare your results to experimentally determined dissolution rates or EDS maps close to glass surface? Or more advancement in either modeling or experimental techniques is still needed?

Response 6:

Thank you for the interesting comment.

The lower self-diffusion may indicate higher chemical stability and durability but there may not be a direct impact on the results. The dissolution is a result of element diffusion and influence of water on a glass network. The process of glass hydration and braking of P-O-P bonds seems to be a more important factor. On the other hand, the self-diffusion may be reflected in electrical conductivity measurements.

The glass inhomogeneity (cluster formation) takes place at the atomic level. The size of the inhomogeneities is several Å (10^-10m) whereas the EDS spot is of 10^-6 m order.  Thus, the map is not able to reflect changes.

 

The following sentences have been added to the Conclusions (p.14, l.316-321):

It should be pointed out that the conducted studies are modeled results. The values of the self-diffusion coefficients may influence the chemical durability of the material. Nevertheless, the influence is not direct, and the more important factor is the hydration of P-O bonds and glass network decomposition due to it. On the other hand, the changes in the coefficients may be seen in electrical conductivity measurements. That may be an interesting point for further studies.

Reviewer 2 Report

Comments and Suggestions for Authors

The temperature dependence of the liquidus dynamic behaviors of (100-x)(0.7P2O5-0.3Fe2O3)-xSrO (mol%) ] iron-phosphate glasses with different SrO content has been investigated thoroughly by MD simulations. The role of Sr cation was well explained, which is placed in the open voids. The Sr cation as the next work modifier will lead to a separated Sr-rich regions with high density along with increasing temperature. This work is interesting and meaningful for the field of phosphate glasses, however, there are in general several issues should be addressed before the final publication.

As shown in this work, with increasing SrO content, the molar volume decreases obviously. However, it is confused that why the ratio of the V_M to the molar volume at 300 K increases with increasing SrO content?

The descriptions about figures and Tables are not very clear, for instance, in Figure. 7, it is better to mark (a), (b), (c)…(o). In Table 1, it is better to include x=0-50 in a separated row (composition list).  

Author Response

Comment 1:

As shown in this work, with increasing SrO content, the molar volume decreases obviously. However, it is confused that why the ratio of the VM to the molar volume at 300 K increases with increasing SrO content?

Response 1:

Thank you for the comment.

These are two different dependencies. The Vm decrease is related to changes in SrO content. Whereas, the ratio describes the speed of change in the function of temperature.

 

Comment 2:

The descriptions about figures and Tables are not very clear, for instance, in Figure. 7, it is better to mark (a), (b), (c)…(o). In Table 1, it is better to include x=0-50 in a separated row (composition list).  

Response:

Thank you for the comment.

The Table 1 and Figures 7 and 8  have been changed according to the suggestion.