Antimony (V) Adsorption at the Hematite–Water Interface: A Macroscopic and In Situ ATR-FTIR Study

Round 1

Reviewer 1 Report

Review on Mierzwa e al.: Antimony (V) Adsorption at the Hematite-Water Interface: a Macroscopic and In Situ ATR-FTIR study

 

Mierzwa et al. report on batch type adsorption experiments of Sb(V) on Hematite, and a corresponding in-situ ATR-FTIR investigation.

Experiments seem to be well performed. The paper is well organized and results are well supported by Figures/ Graphs. Thus, the experimental adsorption data, the spectroscopic data and their interpretation seem sound to me.

The model development, however, requires some corrections. Not all relevant aqueous species are included in the model, which is a fundamental prerequisite to enable correct surface complexation modelling. I recommend to the authors to look into the Thermochimie database¹ for more complete and up to date information. This must be corrected before the model can be published.

I assume the surface site concentration was fitted during model parameter adjustment. This must be mentioned in the model description. If the authors have any chance to measure the specific surface area (e.g. by BET) this would be very much preferable over using a literature value.

I wonder, why the adsorption constants were treated as conditional constants and adjusted separately for every ionic strength. In my opinion it would be preferable to use these constants as global parameters (i.e. one value for all ionic strength) and if necessary vary e.g. the capacitance as function of ionic strength. This would make much more sense to me, especially as data do not vary very much as function of ionic strength.

Uncertainties must be added to the Figures showing adsorption data. And additional graphs demonstrating how much which sorption species contributes to the overall adsorption would be helpful.

Further minor comments are added to the pdf-manuscript file.

Overall the manuscript reports on a relevant and interesting topic, and is on a good way, but requires some further improvement.

 

Reference

1. Giffaut, E.; Grivé, M.; Blanc, P.; Vieillard, P.; Colàs, E.; Gailhanou, H.; Gaboreau, S.; Marty, N.; Made, B.; Duro, L., Andra thermodynamic database for performance assessment: ThermoChimie. Appl Geochem 2014, 49, 225-236.

Comments for author File: Comments.pdf

Author Response

Reviewer 1:

Mierzwa et al. report on batch type adsorption experiments of Sb(V) on Hematite, and a corresponding in-situ ATR-FTIR investigation.

Experiments seem to be well performed. The paper is well organized and results are well supported by Figures/ Graphs. Thus, the experimental adsorption data, the spectroscopic data and their interpretation seem sound to me.

 

Ans. The corresponding author thanks the reviewer for encouraging comments about the manuscript.

 

The model development, however, requires some corrections. Not all relevant aqueous species are included in the model, which is a fundamental prerequisite to enable correct surface complexation modelling. I recommend to the authors to look into the Thermochimie database¹ for more complete and up to date information. This must be corrected before the model can be published.

 

Ans. The relevant aqueous species for antimonate are included in the model. The only potentially significant aqueous species that may occur is KSb(OH)₆⁰; however, the association constant for this species is not available, and it’s formation is assumed to be insignificant. The only aqueous species offered by the Thermochimie database that are not included in the modeling are polymeric Sb(V) species. The formation of these are not significant for the total Sb(V) solution concentrations used her (<10 μM, see Baes and Mesmer, 1986), thus there inclusion is not warranted.

 

I assume the surface site concentration was fitted during model parameter adjustment. This must be mentioned in the model description. If the authors have any chance to measure the specific surface area (e.g. by BET) this would be very much preferable over using a literature value.

 

Ans. The surface area of the hematite could not be determined because there was an insufficient amount remaining to perform BET analysis. Instead, the measured surface area data presented by Elzinga and Kretzschmar (2013) was used, as they employed the same method of hematite preparation (Sugimoto et al., 1993). As indicated in the text, the total concentration of surface sites is an optimized parameter. This parameter is directly dependent on surface area and site density. Even if the surface area were measured, the total site concentration would still be an optimized parameter because the site density is not known.

 

I wonder, why the adsorption constants were treated as conditional constants and adjusted separately for every ionic strength. In my opinion it would be preferable to use these constants as global parameters (i.e. one value for all ionic strength) and if necessary vary e.g. the capacitance as function of ionic strength. This would make much more sense to me, especially as data do not vary very much as function of ionic strength.

 

Ans. The intrinsic constants optimized in FITEQL are conditional constants. Thus, they will vary with ionic strength (a fact that is consistent with their reference state). To use them as a global parameter would violate the conditions of the SCM. While FITEQL will allow for the optimization of adsorption data obtained at several ionic strength values, generating a single intrinsic adsorption constant, this ability should not be used.

 

Uncertainties must be added to the Figures showing adsorption data. And additional graphs demonstrating how much which sorption species contributes to the overall adsorption would be helpful.

Ans. Three Figures (S1, S2, and S3) have been added in the supplementary section for showing contribution of individual species to the overall adsorption.

 

Further minor comments are added to the pdf-manuscript file.

Overall the manuscript reports on a relevant and interesting topic, and is on a good way, but requires some further improvement.

 

Reference

1. Giffaut, E.; Grivé, M.; Blanc, P.; Vieillard, P.; Colàs, E.; Gailhanou, H.; Gaboreau, S.; Marty, N.; Made, B.; Duro, L., Andra thermodynamic database for performance assessment: ThermoChimie. Appl Geochem 2014, 49, 225-236.

 

Ans. The corresponding author thanks the reviewer 1 for useful comments

Reviewer 2 Report

In this paper, the authors to evaluate the surface complexation mechanism of Sb (V) on hematite under a range of solution properties using macroscopic, in situ attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopic, and surface complexation modeling exercise. The adsorption of Sb was monitored using a simple batch adsorption experiment in different of pH. However, this paper is poorly written and does not have a good analysis in the results and discussions. conclusion is missing. Major problems as follows:

 

Major issue:

1. Introduction, line 49-59, author summarized many previous studies on the effect of pH on the adsorption of Sb, but did not put forward the relevant content of this study. The same problem occurred at line 71-79.
2. The “results” should be “results and discussions”. In this section, the author focuses on the adsorption of Sb(V) by goethite under different pH. However, many of discussions is missing. For example, we can obviously observe that there are obvious changes in bands between 1300-1100 and 1000-900 cm^-1 et al, and these changes should be discussed.

 

3. The conclusions should be presented at the end of this study.

 

4. Line 207-208, Which of functional group leads to the adsorption of Sb(V) on goethite surface.

 

5. Abbreviations in figures should be understood without looking at main text.

 

6. Fig4A, B, C should be merged.

 

7. Line 43-44, Sb(III) in the environment has better mobility and solubility than Sb(V), please check it.

 

8. Line 67-68，references is missing.

 

9. Line 104, overnight equilibrated, how many hours.

Author Response

Reviewer 2:

Comments and Suggestions for Authors

In this paper, the authors to evaluate the surface complexation mechanism of Sb (V) on hematite under a range of solution properties using macroscopic, in situ attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopic, and surface complexation modeling exercise. The adsorption of Sb was monitored using a simple batch adsorption experiment in different of pH. However, this paper is poorly written and does not have a good analysis in the results and discussions. conclusion is missing. Major problems as follows:

 

Major issue:

1. Introduction, line 49-59, author summarized many previous studies on the effect of pH on the adsorption of Sb, but did not put forward the relevant content of this study. The same problem occurred at line 71-79.

 

Ans. We discussed the adsorption behavior of Sb on other Fe-oxides. In addition, we have discussed what other researchers reported on the surface complexation of Sb (V) using in situ ATR-FTIR and what specific IR bands they used to justify their surface complexation mechanism We think this is very relevant discussion to build the context of our study. 

 

1. The “results” should be “results and discussions”. In this section, the author focuses on the adsorption of Sb(V) by goethite under different pH. However, many of discussions is missing. For example, we can obviously observe that there are obvious changes in bands between 1300-1100 and 1000-900 cm^-1 et al, and these changes should be discussed.

 

Ans. The section has been named ‘Results and Discussions’ as the reviewer suggested. We did not study the Sb (V) adsorption on goethite, we studied on hematite, which is a different Fe (III) mineral. It is well described in the discussions and in the introduction that what are the important IR bands for Sb (V) adsorption on iron oxide minerals and how the changes in these IR bands can be related to inner- or outer sphere adsorption mechanism. Our discussions are consistent with published studies as well (as cited in the texts: McComb et al., 2007; Muller et al., 2015).  The major IR bands (~1100 cm^-1 and 1015 cm^-1) are observed to be changed upon adsorption (See Figure 1) and are in the spectral range of 1300-1100 and 1000-900 cm^-1 as reviewer is suggesting. These changes are also specified in Figure 2, in which Aqueous Sb (V) species was compared with adsorbed species to further verify the IR band changes due to adsorption. In addition, Figure 3 have been presented to confirm the distribution of IR bands due to weak (~1015-1067 cm^-1) and strong (1111 cm^-1) adsorption at varying pH values. All of these observations are consistent with the published literature and our wetchemical and surface complexation modeling data.  We omitted other minor IR bands to focus the discussions on the topic and to avoid speculations (in literature, the reference for these minor changes is not available). 

 

 

1. The conclusions should be presented at the end of this study.

 Ans. Conclusion is optional, so we did not include.

 

 

1. Line 207-208, Which of functional group leads to the adsorption of Sb(V) on goethite surface.

 

Ans. There’s only one reactive functional group…the singly-coordinated group. The Sb-O-H vibration changes upon bonding. It is clearly demonstrated in past studies (McComb et al., 2007; Muller et al., 2015). We found the similar evidences in case of hematite (not goethite).

 

1. Abbreviations in figures should be understood without looking at main text.

 

Ans. I am not sure what abbreviation does the reviewer meant. Only abbreviation is ATR-FTIR, which is the main topic of this manuscript and well-spelled out in many places in the texts.

 

1. Fig4A, B, C should be merged.

Ans. Figures 4A,B, and C are merged now.

 

1. Line 43-44, Sb(III) in the environment has better mobility and solubility than Sb(V), please check it.

 Ans: Sb (III) is more toxic, but less soluble and mobile than Sb (V) [check the texts: “Among the two major oxidation states (III and V) of Sb in the geochemical environments, Sb (V) is reported as the most dominant species prevalent in wide redox range (360 to -140 mV) while Sb (III) is considered to be more toxic [1, 10-14].  In addition, the mobility and solubility of Sb (V) are greater than Sb (III) [12, 13]”. The reference cited here clearly demonstrated that Sb (V) is the prevalent species in a wide redox range and more soluble. This is because of metastability and slow rate of reduction of Sb (V) to Sb (III) in reducing condition. In addition, in the shooting range soil, the most identified species is Sb (V) as cited in the references.

 

1. Line 67-68，references is missing.

 Ans. The reference [27] is added.

1. Line 104, overnight equilibrated, how many hours.

 

Ans. The number of hour is added (24h).

Round 2

Reviewer 1 Report

2^nd Review on Mierzwa et al.: Antimony (V) Adsorption at the Hematite-Water Interface: a Macroscopic and In Situ ATR-FTIR study

 

The Manuscript by Mierzwa et al. improved during the review process. The authors applied most of the suggested minor corrections. However, the essential lapses concerning the model were not corrected.

The auhors seem to insist that:

“The charge density in each of the inner and outer Helmholtz layers is proportional to the electrical potential in each layer, where the inner and outer layer capacitance values (C1 and C2) are the proportionality factors.”

This is simply not correct. I attach a book chapter by Ohshima et al. Equations 18 and 19 in this document show how the relations are.

The authors may be right that there is no automated routine in Fiteql to obtain intrinsic constants (valid for infinite dilution). However, there are certainly ways around this problem, e.g. by applying a manual activity corrections via dummy species that mimic activity coefficients within a constant ionic strength approach (ionic strength does not vary as function of pH). This is also not entirely correct, but certainly better than the approach chosen in the present paper.

The best solution would be to use PhreeqC and to couple it with a suitable optimization routine. Then, ionic strength and activity corrections would be handled most accurately. Additionally, charge and potential in the diffuse layer would be treated more accurately.

In principle, the correct use of conditional (ionic strength dependent) log Ks is also possible, however, in that case constants must follow a logic trend, i.e. vary with ionic strength in a fashion that they can be extrapolated towards an intrinsic value at infinite dilution.

This is not the case for the presented models. Therefore, I think it is not appropriate to publish the model in the present state.

 

A further minor comment: There are actually surface complexation models for hematite that successfully reproduce potentiometric titration and zetapotential data and rely on crystallographic site densities, i.e. site densities are not treated as adjustable parameters.  

Comments for author File: Comments.pdf

Author Response

File is attached.

Author Response File: Author Response.docx

Reviewer 2 Report

The reference style needs to be consistent. Check first letter of title is uppercase or lowercase.

Author Response

Thanks for the comments.

The reference has been checked and the first letter of the title is always uppercase and is consistent with the Journal style.