Review Reports - Characterizing the Suitability of Granular Fe<sup>0</sup> for the Water Treatment Industry

Round 1

Reviewer 1 Report

The authors present an interesting study on the dissolution kinetics of Fe(0) in EDTA and Phen. However, their conclusions seem to be too general and at least should be explained better. In general the text should be revised away from "publicity for the Phen test" towards more neutral descriptions of the experimental results and underlying mechanisms. Only at the very end the authors should draw a conclusion which should be better explained.

In detail the following points should be improved.

I suggest a more precise title, i.e. "Dissolution kinetics of granular Fe0 in EDTA and 1,10-Phenanthroline"

Abstract:

Shift the sentence from line 23 "A total of eight..." directly after "water treatment" in line 20.

line 24: replace "confirmed" by "showed" or "indicated" as no hypothesis has been mentioned before which could be confirmed.

line 25: add "in" before "phen"

line 26/27: Differentiate clearly between solubility (i.e. thermodynamic equilibrium) and dissolution kinetics. The term "reactivity" is too vague. Either "The reaction kinetics were faster" or "The solubility was higher"

line 28: What do you mean by "time-dependent variation of the kinetics"? Kinetics are per definition time-dependent, e.g. for a first order reaction If you talk about variation of the kinetics, does this mean that different reaction orders apply for different reaction periods?

line 29: The dissolution of a solid in an aqueous solutions depends on the composition of the the solution, i.e. you cannot say that the dissolution in a phen solution is "more true" than the dissolution in a EDTA solution.

line 36: Replace "metallic iron materials" by "zero-valent iron" which is also a keyword and has been mentioned as sample code in the abstract.

line 39. It should be mentioned that ZVI may not only produce FeCPs like lepidocrocite, but may also trigger the precipitation of other phases (https://doi.org/10.1016/j.jclepro.2018.10.180).

I suggest to restructure the beginning of the introduction section:Shift the explanation, i.e. the production of FeCps (line 38) down to line 45 ff where you discuss reducing agent vs in-situ nascent contaminant scavengers (a term which needs some explanation, are these species like OH* which are used in so-called "advanced oxidation processes, AOP, like mentioned by https://doi.org/10.1016/j.chemosphere.2018.11.043?).

line 65: Another reason is that metabolites may form.

line 69: As the authors correctly highlight the semantic issues leading to confusion in literature, I suggest to define "intrinsic reactivity" here: "In this study we mean by intrinsic reactivity..."What is extrinsic reactivity then?

As mentioned above the corrosion of Fe(0) depends on the composition of aqueous media corroding it, i.e. the mentioned tests measure different things. It should be explained, why the "rationale Fe(0) corrosion by water" shall be determined not by water, but by aqueous solutions containing different solutes.

line 77: As the authors have not defined what "intrinsic reactivity" is, it is not clear why the reaction with oxidizing agent does not measure this. I guess that in contact with an oxidizing agent Fe(0) will oxidize (i.e. react) and the oxidizing agent gets reduced. This is for sure "reactivity", but maybe "extrinsic reactivity"? Please explain.

line 84: Finally, the meaning and the idea of "EDTA test" is explained. Good! However, this should be discussed in the context of the definitions which I mentioned above. As far as I understand this test aims for the determination of a dissolution rate (and not of the solubility!) of Fe(0) by removing the reaction products from the system which avoids precipitation of HFO phases which would limit the dissolution by thermodynamic equilbirium. Right? Then the information on dissolution rate - solubility - reactivity should be put together with this test approach somewhere.

line 88: "real world situations" vary a lot and in fact mimicking them would be a different rationale of the test than I thought it to be above. In many "real world situations" I would guess that HFOprecipitation would take place, a process which is surpressed in this test by EDTA addition. So either the test shall determine the dissolution rate of Fe(0) without HFO precipitation, or the behaviour off Fe(0) under natural oxidizing conditions, but it cannot do both.

I suggest to introduce these tests like this: "Reactions between Fe(0) and aqueous solutions lead to the formation of hydrous ferric oxides (HFO), whose low solubility limits the dissolution of Fe(0). Therefore, for the determination of the dissolution rate of Fe(0) the precipitation of HFO must be prevented. Two approaches have been developed for this purpose..."

line 110: Having read the introduction I would prefer to talk about "dissolution rate" instead of "intrinsic reactivity".

Table 2: explain SSA (specific surface area?) in the table caption.

line 131 ff: This is an exact repetition of the lines 123 ff! Delete!

line 145ff: Finally, you8 explain the idea behind these tests. Shift this justification to the introduction and merge it with my remarks.

line 155: This means that HFO phases are not dissolved by Phen, so that b is lower, right? Maybe add this information here.

line 156: add "L/S" before factor or write "liquid/solid ratio".

line 157: Add the formula used for the calculation of k values.

line 158: Finally, you clearify that the "intrinsic reactivity" and the "dissolution rate" are the same. Please use "dissolution rate" throughout the text.

line 160: The title sounds too strong and judgmental. I prefer "Dissolution rates of Fe(0) in EDTA and Phen aqueous solutions".

line 161: more precisely: "dissolved iron concentration".

line 164 and later on: replace "value" by "concentration"

line 168: delete "(a value)"

line 169: The research question should be mentioned in the introduction, not in the results. More important: How do you define better? The dissolution rate in different solvents is different. Do you want to approximate the dissolution rate in pure water by using an aqueous solution with EDTA or Phen? Then explain this: We use the dissolution rate in aqueous solutions of EDTA and Phen as a proxy for the dissolution rate in pure water, as the latter cannot be determined due to the rapid precipitation of HFO phases. ...and later on: We think that the dissolution rate of Fe(0) in a Phen solution is closer to the rate in pure water, because..

line 177: As the molar concentration of EDTA and Phen were the same, why does the Phen solution contain only 1/3 of the required amounts for Fe complexation, whereas there was sufficient EDTA for the same purpose? A discussion about the complexation reactions between the two organic compounds and Fe(III) ions is required (dissociation constants of the formed complexes, structural formulae etc.).

line 183: This motivation for the experimental setup, i.e. the "tests", should be mentioned either in the introduction or in the methods section, not in the results section.

line 189: This is not clear to me. If Fe(III) complexes are more stable than Fe(II) complexes, then this should effect the ratio between complexed and free Fe(III) and between complexed and free Fe(II), but how does it effect the Fe(III)/Fe(II) ratio?

line 191: I understand the explanation, but suggest to include chemical reactions here which show that the reaction product, i.e. dissolved Fe(III) is removed from the system by complexation.

line 194: What do you mean by "atmospheric FeCps"? Iron hydroxides which formed at interface between iron and the atmosphere? The term sounds as if these phases occur as aerosols, so I suggest to change it.

line 197: It has to be explained where the dissolved oxygen comes from and why it disturbs Fe(0) dissolution. I don´t understand the conclusion. What I understand is that in presence of Phen Fe(0) is only oxidized to Fe(II) (because the Phen-Fe(II) complex is too stable to be oxidized) whereas it is oxidized to Fe(III) in presence of EDTA. The step towards dissolved oxygen and then to the general statement on "intrinsic reactivity" is not clear to me. Please explain this better.

Figure 3, Table 3 and Figure 4 should be shifted to the position in the text where they are referred to.

line 221: Please restructure the Results section: First describe the EDTA test, then the Phen test and then you can maybe(!) think of "superiority". But you cannot compare the two before you have discussed them individually.

line 237: add "a" before "monotonous"

line 238: Explain in the methods section how the Phen test is related to dissolved oxygen and atmospheric corrosion

General remark on 3.2 and 3.3: Please discuss the differences between individual materials in the context of other differences (e.g. grain size, shape etc.)

line 252: The discussion of other methods to determine the "intrinsic reactivity" of Fe(0) should be shifted to the introduction, as this does not refer to the results of this study.

line 271ff: Fig. 4 & 5 show that there is no correlation between specific surface area/ particle size on the one hand and the a value, i.e. the dissolution rate. Consequently, other properties of the different materials must differ and explain the differences. Discuss!

line 279/280: I don´t understand this conclusion from the previous sentence

line 282: I suggest to shift this to a separate chapter "Discussion and Outlook"

line 292: I suggest to shift those information which was known before this study to the introduction and to shift the conclusions from this study to the discussion.

Author Response

We are very thaksful to reviewer 1 for his extensive review. We have used them as an opportunity to better present our concept. We have considered all the comments.

Reviewer 1

The authors present an interesting study on the dissolution kinetics of Fe(0) in EDTA and Phen.

Many thanks for this evaluation. We would like to add “in an effort to characterize the intrinsic reactivity of Fe(0) materials”. It is not only about iron dissolution.

However, their conclusions seem to be too general and at least should be explained better.

We will try to better explain. However, we have compared all existing approaches to characterize the intrinsic reactivity of Fe(0) materials and insisted on that the Phen method is the best. This makes the conclusions general!

In general the text should be revised away from "publicity for the Phen test" towards more neutral descriptions of the experimental results and underlying mechanisms. Only at the very end the authors should draw a conclusion which should be better explained.

The text has been carefully re-read to understand this important remark. In the introduction (with table 1) all methods were presented including the Phen test. The information that the Phen test is not introduced by the present work is also present. It results from a master thesis of one of the authors, and is yet to be (peer-reviewed) published. The submission is the first application that had insisted on the advantages of the Phen test over the EDTA test, this was clear in the abstract of the original submission (not changed).

In detail the following points should be improved.

I suggest a more precise title, i.e. "Dissolution kinetics of granular Fe⁰ in EDTA and 1,10-Phenanthroline"

This title would suggest that we are investigating the dissoltion of Fe(0) in two complexing agents. No, we are characterizing Fe(0) material for the water treatment industry and we have specified that both tests are superior to two recent one which one was presented as “Standardization of the reducing power of zerovalent iron using iodine” and the other as “A facile method for determining the Fe(0) content and reactivity of zero valent iron”. In both cases, it is about comparing the reactivity of Fe(0) materials.

Abstract:

Shift the sentence from line 23 "A total of eight..." directly after "water treatment" in line 20.

line 24: replace "confirmed" by "showed" or "indicated" (is preferred) as no hypothesis has been mentioned before which could be confirmed.

Done, thanks!

line 25: add "in" before "phen"

Done, thanks!

One thing is clear, it is not about solubility. The other thing is that the dissolution kinetics or better the comparative dissolution kinetics is used to assess the intrinsic reactivity. The intrinsic reactivity is not something you can characterize with a number (a value). It is just a trend. That is why the research community in the quest for a standard protocol to characterize the reactivity (see the excellent review by Li et al. 2019 at Water Research). Our submission is a contribution.

No action, because it is a fundamental point and most comments (including the one on “too general” conclusions) of this reviewers is based on this key issue!

It means that iron does not corrode with the same kinetics over the time and the law of variation is not linear! This is well-known and cannot be addressed in an abstract. This is also the reason why we have recommended long-term “Phen tests”. No action!

line 29: The dissolution of a solid in an aqueous solutions depends on the composition of the solution, i.e. you cannot say that the dissolution in a phen solution is "more true" than the dissolution in a EDTA solution.

We have not said this, again it is the abstract. What we have demonstrated through the text is that in Phen only Fe⁰ forward dissolution is addressed while in EDTA Fe^III species are also dissolved. No action!

line 36: Replace "metallic iron materials" by "zero-valent iron" which is also a keyword and has been mentioned as sample code in the abstract.

Metallic iron refers to metal iron or iron in the oxidation state zero. It is not clear to the authors, why “valence” is used. We disagree with this acronym but because with want our papers to be read, we systematically give "zero-valent iron" as a keyword. No action!

line 39. It should be mentioned that ZVI may not only produce FeCPs like lepidocrocite, but may also trigger the precipitation of other phases (https://doi.org/10.1016/j.jclepro.2018.10.180).

This idea is contained is that statement as the focus is not on contaminant removal. However, the excellent work by Vollprecht and colleagues is added as ref. 12. A recent tutorial review is also added as ref. 13.

line 45. I completely agree and suggest to define the terms for this particular article. I think that "efficiency" should refer to the relative decrease in dissolved concentration of the treated water constituent, whereas "reactivity" refers to the amount of ZVI that reacts (i.e. corrodes) in a certain time span, right? (fully agree! - this is why “adsorption capacity” is very wrong in the context of metallic iron. The amount of adsorbent changes with time)

Thanks, for this remark. Miyajima (2012) was to the best of our knowledge the first scientist to reveal this confusion. She said that “efficiency is the expression of reactivity under operational conditions” This entails the evidence that reactivity does not change. Changing the operational conditions (pH, salts…) modify the efficiency of the same materials. No action because it is not the focus of the submission.

I suggest to restructure the beginning of the introduction section: Shift the explanation, i.e. the production of FeCps (line 38) down to line 45 ff where you discuss reducing agent vs in-situ nascent contaminant scavengers (a term which needs some explanation, are these species like OH* which are used in so-called "advanced oxidation processes, AOP, like mentioned by https://doi.org/10.1016/j.chemosphere.2018.11.043?).

Thanks for these remarks. We would like to maintain the current structure. The focus is not on contaminant removal nor transformation. OH* and other radicals are (H₂O₂)* also (secondary or ternary) corrosion products. “Nascent” is borrowed to the traditional corrosion literature. No action!

line 65: Another reason is that metabolites may form.

Sure, this has rationalized the introduction of nano-ZVI and bimetallic, just because the “density of electron” of the “electron efficiency” is ill-defined. “(ii) metabolites may form in-situ [13]” is added, Thanks!

“ In this study, intrinsic reactivity is characterized by the initial dissolution kinetics of tested Fe⁰ in EDTA and Phen.” is added, Thanks!

In essence, EDTA and Phen as used herein are two diluted aqueous solutions at pH > 4.0! Iron corrodes with two different mechanisms at pH < 4.0 (H₂ evolution) and pH > 4.0 (O₂ adsorption). Because the reactivity is a trend (not a value), the reactivity order determined in EDTA or Phen should be the same. Remember that this is just material screening and long-term experiments are recommended. No action!

The reviewer describes the impact of oxidizing agents on Fe⁰. This is an excellent example for two reasons:

1) The redox potential of the Fe^II/Fe⁰ redox couple is -0.44 V. This value is valid for ALL Fe⁰ materials, but each material has it own intrinsic reactivity.

2) Cu²⁺ would induces differential extent of Fe0 dissolution at two different temperatures.

The presentation herein is limited on material screening, for further experimentation. Answering this question within the text is over the scope of the submission.

line 84: Finally, the meaning and the idea of "EDTA test" is explained. Good! However, this should be discussed in the context of the definitions which I mentioned above. As far as I understand this test aims for the determination of a dissolution rate (and not of the solubility!) of Fe(0) - Exact until here! by removing the reaction products from the system which avoids precipitation of HFO phases which would limit the dissolution by thermodynamic equilbirium. Right? (The second part is not correct, and is complex to understand. Atmospheric corrosion products have complicated reactivity tests using EDTA, Phen is a complexing agent (mostly) not addressing atmospheric corrosion products) Then the information on dissolution rate - solubility - reactivity should be put together with this test approach somewhere.

To the best of our knowledge this is clear from the text. No action!

line 88: "real world situations" vary a lot and in fact mimicking them would be a different rationale of the test than I thought it to be above. In many "real world situations" I would guess that HFO precipitation would take place, a process which is surpressed in this test by EDTA addition. So either the test shall determine the dissolution rate of Fe(0) without HFO precipitation, or the behaviour off Fe(0) under natural oxidizing conditions, but it cannot do both.

The view of the Reviewer is not clear to us. It is definitively clear that taking all these complications in material screening is impossible but that screening materials before bringing them to the real world for more profound investigations is a good point!

No, the idea in the submission is that HFO are dissolved in EDTA and ‘falsify’ the results, as dissolved iron is not (only) from Fe^II. It is later discussed that the Phen test is free from this inherent difficulty.

line 110: Having read the introduction I would prefer to talk about "dissolution rate" instead of "intrinsic reactivity".

Again, there is discrepancy of view! We have used the initial dissolution rate to access the intrinsic reactivity and this is nothing new in Environmental Geochemistry. We would just like refer on the whole literature on ‘mineral dissolution (e.g., MnO₂ or pyrite dissolution)

Table 2: explain SSA (specific surface area?) in the table caption.

Done, thanks!

line 131 ff: This is an exact repetition of the lines 123 ff! Delete!

Done, many thanks!

line 145ff: Finally, you explain the idea behind these tests. Shift this justification to the introduction and merge it with my remarks.

At last, many thanks! We have added the following to the introduction:

“The rationale for the EDTA test is that the initial iron dissolution in a 2 mM EDTA solution is a linear function of the time. The slope of the corresponding line is characteristic (intrinsic reactivity) for the used Fe⁰ (Section 6.5).”

line 155: This means that HFO phases are not dissolved by Phen, so that b is lower, right? (correct, thanks!) Maybe add this information here.

Done, many thanks! (Accordingly, lower b values are expected in Phen tests. )

line 156: add "L/S" before factor or write "liquid/solid ratio".

No, 1/20 corresponds to 50/1000 to transform the mg/L in mg (from 1000 mL to 50 mL)

line 157: Add the formula used for the calculation of k values.

It is stated that the regression parameters are obtained using Orogin (one can use Excel).

line 158: Finally, you clearify that the "intrinsic reactivity" and the "dissolution rate" are the same. Please use "dissolution rate" throughout the text.

No, even with the addition in the introduction, "intrinsic reactivity" is correct and title is okey. We are measuring the trend in parallel experiments. It is not a number like the dissolution rate and the real dissolution rate needs a standard protocol AND a reference material! Both are missing as discussed here and elsewhere.

line 160: The title sounds too strong and judgmental. I prefer "Dissolution rates of Fe(0) in EDTA and Phen aqueous solutions".

Again both tests exist already, even though Phen test has not yet been peer-reviewed published (Master thesis – manuscript in preparation)! The Phen test addresses only the forward dissolution of Fe⁰ and is not disturbed by corrosion products (HFO).

line 161: more precisely: "dissolved iron concentration".

Done, many thanks!

line 164 and later on: replace "value" by "concentration"

[Fe] stands for Fe concentration. [Fe] value is therefore correct!

line 168: delete "(a value)"

Done, thanks!

line 169: The research question should be mentioned in the introduction, not in the results (We are discussing the results). More important: How do you define better? (in a scientific perspective) The dissolution rate in different solvents is different. (The results are recalling this) Do you want to approximate the dissolution rate in pure water by using an aqueous solution with EDTA or Phen? (This is not possible, it is just material screening as discussed) Then explain this: We use the dissolution rate in aqueous solutions of EDTA and Phen as a proxy for the dissolution rate in pure water, as the latter cannot be determined due to the rapid precipitation of HFO phases. ...and later on: We think that the dissolution rate of Fe(0) in a Phen solution is closer to the rate in pure water, because..

No action as already discussed. A treand is assessed and the Phen test is freed from inherent difficulties of the EDTA test (dissolution of HFO)

Not really necessary! The solutions are not saturated and preliminary works and the Master thesis have shown that the Phen test evidence for the first time the time-dependent variability of the kinetics iron dissolution in short-term experiments. It is also seen here, but discussing it is over the scope of this communication!

line 183: This motivation for the experimental setup, i.e. the "tests", should be mentioned either in the introduction or in the methods section, not in the results section.

Again this is the discussion of the results achieved and presented herein!

This is an interesting point: Fe(III)-EDTA is more stable than Fe(II)-EDTA. Fe(II) is quickly transformed to Fe(III) for the formation of Fe-EDTA or Fe(II)-EDTA is quickly oxidized to Fe(III)-EDTA. The result is the same and the real mechanism is not important for the discussion herein. No action!

line 191: I understand the explanation, but suggest to include chemical reactions here which show that the reaction product, i.e. dissolved Fe(III) is removed from the system by complexation.

The reviewer is right. However we have not supported this by Ks values because literature values are very discrepant.

“Atmospheric FeCPs are FeCPs formed at Fe⁰/atmosphere interface before immersion.” is added!

The experiments are performed under laboratory conditions (open systems), O₂ saturation can be assumed. The following reaction is added:

4 Fe²⁺ + O₂ + 4 H⁺  4 Fe³⁺ + 2 H₂O (3)

Figure 3, Table 3 and Figure 4 should be shifted to the position in the text where they are referred to.

Done, thanks!

Again, the tests are known. We have changed “3.1. Evidence for the superiority of the Phen test” to “3.1. Suitability of the experimental protocol”.

We agree with the Reviewer that this corresponds better to our approach of validating the approach before support it with numbers (a and b values).

line 237: add "a" before "monotonous"

Done, thanks!

line 238: Explain in the methods section how the Phen test is related to dissolved oxygen and atmospheric corrosion

We are convinced that adding Eq. 3 has clarified this issue.

General remark on 3.2 and 3.3: Please discuss the differences between individual materials in the context of other differences (e.g. grain size, shape etc.)

This would make the presentation too long without gain of information. We have referenced previous works which have established that none of these factors in preponderant. Even not the grain size (for materials from various manufacturers)!

line 252: The discussion of other methods to determine the "intrinsic reactivity" of Fe(0) should be shifted to the introduction, as this does not refer to the results of this study.

Again, this is the discussion and is related to the introduction were it was already done and summarized in Tab. 1. The fact that the Reviewer is insisting on this is new to us. We have focused the presentation of our main findings, this implies that what is introduced (by showing the gaps in knowledge) is discussed after the results have been presented. In Results and Discussion, even should experimental details (Materials and Methods) are accepted when it is not a key aspect.

No complete discussion is possible. This manufacture process (mostly patented) is needed as well and it is the objective of this communication to show that despite the absence of such parameters, reliable test is possible.

line 279/280: I don´t understand this conclusion from the previous sentence

The discussion based on particle size is the same as the one base on surface area.

line 282: I suggest to shift this to a separate chapter "Discussion and Outlook"

This would be too short and is not really an outlook. It is our last argument in the discussion: Suggesting an experimental protocol for the new text.

line 292: I suggest to shift those information which was known before this study to the introduction and to shift the conclusions from this study to the discussion.

We understand the point of the reviewer. However, given the novelty of the study we have opted to summarize the main findings in this form. There are journal for which “Significance of the results” are the Concluding remarks. We have benefited to the freedom at Processes to shape our take home message in the presented form, followed by a short conclusion.

Reviewer 2 Report

The manuscript is well designed and the argumentation are convincing. It pointed out the necessity of a reliable characterization of the intrinsic reactivity of granular ZVI materials and the method proposed is reliable and reproducible method for establishing common standards.
I think that the paper can be published because it would be useful for research group working on the topic.

Author Response

Many thanks for this evaluation!

I think that the paper can be published because it would be useful for research group working on the topic.

Many thanks for this recommendation!

Round 2

Reviewer 1 Report

The authors carefully considered all my remarks and signficantly improved the article. There are some remaining different views, but this is totally okay in academia (and anywhere else). Consequently, the paper can be accepted in the current form.