Round 1

Reviewer 1 Report

Manuscript Water  

Title: Impact of advanced oxidation products on nanofiltration efficiency

Authors: R. Zylla et al.

The present work studies the influence of oxidation products on NF membrane performances. For this, they investigated the rejection of salicylic acid and its oxidation products and their influences on the rejection rate of unreacted salicylic acid.

For this purpose, the authors investigated the experimental membranes performances and they discussed the obtained results regarding membrane characterization (MEB, contact angle measurements).

The theme of this manuscript is not novel, but the experimental investigations are systematical and the results are interesting. The methodology and the results could have some importance for the  scientific community.

The subject matter of the paper and the experimental investigation are interesting but need improvements. Accordingly, I think that the paper could be published in Water after modifications.

My comments are as below.

Specific comments:

First, It's difficult to understand results from figures 3, 4 and 6. A summary table numbering each test, giving references (concentrations of oxidation products, degree of concentration, ...) should be a useful addition and referred in figure captions.

pH, electric conductivity should be an interesting additional data for each solution tested (concentrate and permeate). Zetametry investigation can also improve the discussion.

part 3.2 line 224 to 230.

native pH (solution) = 3 and HL membrane pie ~3. even if it is unlikely to be exactly at the membrane pie, this doesn't mean that there is no charged groups on the active layer, but just an equality between positive and negative charges.

In this condition, electric exclusion is possible for mixed ionic solution because there is a competition between different charged species.

Adsorption isn't an exclusion phenomenon. It's a limiting factor (unsteady), which decreases membrane performances as filtration tests are performed.

In the literature, the scientific community agrees (see works of Fievet et al., Déon et al., Bandini et al., Bowen et al., van der Bruggen et al., Dutournié et al., Yaroshchuk et al., etc) with 3 exclusion phenomena to explain membrane selectivity (steric, electric, dielectric).

I would be interesting to investigate steric hindrance via filtration test of a neutral solute and perform zeta experiments.

In the same vein, it would be interesting to investigate the polarization concentration layer. Indeed, it can have a significant effect, especially for tests performed at different degrees of concentration (results from fig 3,4,and 6).

Figure 5 (SEM photographs): The magnification did not support the conclusion

micro-porosity and roughness > nano ?

lines 244 to 247 explain why and references ?

Lines 260-262 --> for water ? please specify

Contact angles

polar and apolar components of surface tension (including electron - acceptor and donor contributions) are given from D.Y. Kwok et al., Langmuir 1998, 14, 2548-2553 or R.J. Good et al. in M. Schrader and G; Loeb (eds) Plenum press, NY, 1992, 22.

These works used the Lifshitz-van der Walls approach developed by van Oss to estimate polar and apolar components of surface tension of the material and the contributions gamma+ and gamma-. It would be interesting to perform calculation with this approach.

First calculations seem to indicate a significant apolar contribution and the polar component is due to gamma+.

Author Response

Reviewer 2

1)

First, it's difficult to understand results from figures 3, 4 and 6. A summary table numbering each test, giving references (concentrations of oxidation products, degree of concentration, ...) should be a useful addition and referred in figure captions.

We acknowledge this comment. The figures have been changed according to the Reviewer’s suggestion. The improvement allows the reader to understand and read the results easier.

2)

pH, electric conductivity should be an interesting additional data for each solution tested (concentrate and permeate). Zetametry investigation can also improve the discussion.

Broader research including measurement of zeta potential and electrical conductivity will be continued and published in the next paper. In this work the pH measurement of the salicylic acid solution was made. The pH of filtrates and concentrates during filtration was not studied. Such measurements will be made in the next work. In the "Introduction" section, a figure showing the dependence of zeta potential on pH of the solution determined by different researchers has been introduced.

3)

part 3.2 line 224 to 230

native pH (solution) = 3 and HL membrane pie~3. even if it is unlikely to be exactly at the membrane pie, this doesn't mean that there is no charged groups on the active layer, but just an equality between positive and negative charges.

In this condition, electric exclusion is possible for mixed ionic solution because there is a competition between different charged species.

Adsorption isn't an exclusion phenomenon. It's a limiting factor (unsteady), which decreases membrane performances as filtration tests are performed.

In the literature, the scientific community agrees (see works of Fievet et al., Déon et al., Bandini et al., Bowen et al., van der Bruggen et al., Dutournié et al., Yaroshchuk et al., etc) with 3 exclusion phenomena to explain membrane selectivity (steric, electric, dielectric).

The authors took into account the valuable comments of the Reviewer in the manuscript.

4)

It would be interesting to investigate steric hindrance via filtration test of a neutral solute and perform zeta experiments. In the same vein, it would be interesting to investigate the polarization concentration layer. Indeed, it can have a significant effect, especially for tests performed at different degrees of concentration (results from fig 3,4 and 6).

Broader research including zeta potential measurement will be continued and published in the next work.

5)

Figure 5 (SEM photographs): The magnification did not support  the conclusion micro-porosity and roughness > nano?

The microscopic images show that the membrane surface of TS80 (Figure 5B – currently after review 7B) is more folded and uneven compared to the HL membrane (Figure 5A – currently after review 7A). The authors based their analysis partly on literature data. Based on the work of other researchers, it was assumed that the RMS roughness parameter for the TS80 membrane is approx. 89 nm, for the HL membrane – approx. 10 nm.

6)

lines 244 to 247 explain why and references?

The authors decided to delete the sentence from the manuscript. After analyzing the constructive Reviewer’s comments it was considered that further research is necessary to clearly identify the cause of the phenomena.

7)

Lines 260-262 --> for water? please specify

The analysis of contact angle for selected membranes refers to water. The authors specified this detail in the manuscript.

8)

Contact angles polar and apolar components of surface tension (including electron - acceptor and donor contributions) are given from D.Y. Kwok et al., Langmuir 1998, 14, 2548-2553 or R.J. Good et al. in M. Schrader and G; Loeb (eds) Plenum press, NY, 1992, 22. These works used the Lifshitz-van der Walls approach developed by van Oss to estimate polar and apolar components of surface tension of the material and the contributions gamma+ and gamma-. It would be interesting to perform calculation with this approach.

First calculations seem to indicate a significant apolar contribution and the polar component is due to gamma+.

Analysis of the results using the Lifshitz-van der Walls model will be taken and included in the next work devoted to this subject. The authors are grateful for the valuable and informative suggestions of the Reviewer.

Reviewer 2 Report

Impact of advanced oxidation products on nanofiltration efficiency

Renata Żyłła, Rafał Milala, Irena Kamińska, Marcin Kudzin, Marta Gmurek , Stanisław Ledakowicz

In the paper the authors investigate the effect (positive or negative) degradation products on the efficiency of salicylic acid nanofiltration. The paper is well done and recommended for publication after minor changes.

In the experiments the authors used rather high substrate concentrations in the 10^-3 – 10^-4 mol dm^-3 range. What was the reason to choose this concentration range? Simple analytical possibilities (without solid phase extraction) would have allowed to use two orders of magnitude smaller concentration.

Line 44: However, its limitation is that it only enables separation of compounds in unchanged chemical forms.     This statement is certainly not true!

Line 46. The methods of advanced oxidation are used to chemically modify oxidized molecules          used to oxidize molecules

Line 47. The integration of both methods makes it possible to take advantage of their properties and simultaneously reduce disadvantages.     Explanation is needed!

Line 203-204. I did not understand the table. What are the conversions for 2,3-DHBA and Cathecol?

Line 225. out at natural pH of the solution, i.e. pH = 3.0.     HMMMM

Line 242. characterized by a more developed surface     What do you mean by a more developed surface?

Line 268. practically no different   not different

Comments for author File: Comments.pdf

Author Response

1)

In the experiments the authors used rather high substrate concentrations in the 10^^-3 – 10^^-4 mol dm^^-3 range. What was the reason to choose this concentration range? Simple analytical possibilities (without solid phase extraction) would have allowed to use two orders of magnitude smaller concentration.

The subject of research were the oxidation products of salicylic acid whose concentration was about 5-9×10^-5 mol/L. In order to observe the effect of these products on the filtration of salicylic acid with the use of available analytical methods, it was necessary to use higher substrate concentrations. The authors wanted to observe all the phenomena occurring during the filtration process, including changes in the contact angle and FSE.

2)

Line 44:

However, its limitation is that it only enables separation of compounds in unchanged chemical forms. This statement is certainly not true!

Line 46. The methods of advanced oxidation are used to chemically modify oxidized molecules used to oxidize molecules

Line 47. The integration of both methods makes it possible to take advantage of their properties and simultaneously reduce disadvantages. Explanation is needed!

Line 44-47. The authors propose to change the sentence in lines 44-47 as follows:

The aim of the process is to separate hazardous contaminants on the membrane, i.e. to obtain a clear filtrate and concentrate containing the retained substances. The methods of advanced oxidation are used to chemically modify oxidized molecules. The integration of both methods allows us to use their positive features while reducing disadvantages.

Advanced oxidation processes provide chemical modification of hazardous substances to non-toxic forms, while membrane filtration ensures removal of substrates and their oxidation products from water. As a result of combining both methods, a pure filtrate and concentrate containing oxidized products of hazardous materials are obtained.

3)

Line 203-204. I did not understand the table. What are the conversions for 2,3-DHBA and Cathecol?

It was assumed that from 1 mole of substrate 1 mole of the product can be obtained as a maximum. The degree of conversion for products was calculated as follows:

Substrate concentration (SA)          = 5.00 x 10^-4 mol/dm³ – 100%

Product concentration (2,3-DHBA) = 0.96 x 10^-4 mol/dm³ –   x%

Product concentration (Catechol)   = 0.46 x 10^-4 mol/dm³ –   y%

2,3 DHBA conc.    = 19.2%

Catechol conc.      = 9.2%

4)

Line 225. out at natural pH of the solution, i.e. pH = 3.0. HMMMM

The sentence was replaced by: “out at native pH of the solution, i.e. pH=3.0”.

5)

Line 242. characterized by a more developed surface What do you mean by a more developed surface?

The authors meant that the microscopic image of the TS80 membrane (Figure 5B) shows a more undulated and uneven surface than the HL membrane. The authors did not determine the RMS roughness parameter, but based on the literature data, it was assumed that for the TS80 RMS membrane it was approx. 89 nm, for the HL membrane – approx. 10 nm.

6)

Line 268. practically no different not different

It was replaced in the manuscript.

Reviewer 3 Report

My comments can be find in the attached file.

Comments for author File: Comments.pdf

Author Response

Reviewer 3

1 (pg. 3) –The concentration of salicylic acid (69.06 mg/L) used in the UV/H₂O₂ oxidation experiment requires a justification. What range of salicylic acid concentrations can occur in drinking water?

The subject of the research were the oxidation products of salicylic acid whose concentration was about 5-9×10^-5 mol/L. In order to observe the effect of these products on the filtration of salicylic acid using available analytical methods, it was necessary to use higher substrate concentrations. The authors wanted to observe all the phenomena occurring during the filtration process, including changes in the contact angle and FSE.

Salicylic acid in the natural environment is quite common. Its concentration in the tested samples taken from different sources varies from a few to even 977 ng/L (Čielić et al., 2019, Mezzelani et al., 2018).

2 (pg. 4; line 138) – The authors say that “The process of nanofiltration and reverse osmosis was carried out using the… when the system being 2 L/min at 30 oC”. I did not see results in the manuscript obtained from the application of the reverse osmosis.

The authors apologize for a mistake in the text. The results concerning reverse osmosis are not presented in the paper. The phrase "reverse osmosis" was deleted from the manuscript.

3 (pg. 4) – In Table 2, information on the potential zeta (mV) as a function of pH for both membranes should be given.

In the "Introduction" section, a figure showing the dependence of zeta potential for HL membrane on pH of the solution determined by different researchers has been introduced. Unfortunately, not enough data have been collected to prepare a diagram for the TS80 membrane. In the manuscript the Authors refer to the zeta potential value for the TS80 membrane at pH = 3.

4 (pg. 6) – The temperature should be included in the operating conditions indicated in the legend of Table 3.

The authors took into account the valuable comments of the Reviewer in the manuscript.

5 (pg 6) – The separation of salicylic acid via adsorption through the HL membrane is questionable. Batch experiments should be done in order to confirm the occurrence of the adsorption phenomena. Despite the low molecular weight of salicylic acid to be lower than the MWCO, it is reasonable to admit the mechanism size exclusion and the effect of the interference of oxidation products on the permeation of this species. In the case of TS80 membrane, at pH 3 the membrane surface is charged negatively and consequently the effect of electrostatic interaction should be taken into account.

The authors took into account the valuable comments of the Reviewer in the manuscript.

6 (pg 7) – In order to better understand the performance of the tested membranes, it is important to know the decline of the permeation flux along the filtration and the fouling tendency of HL and TS80 membranes.

As suggested by the Reviewer, a figure presenting the dependence of filtrate flux on the filtration time has been added in the manuscript. It was observed that the flux for the oxidizing mixture was slightly lower than for pure salicylic acid solutions. However, this range of research will be further explored and published in another work.

7 (pg 9) – In Figs. 6C e 6D, the authors present retention coefficients of the salicylic acid in the presence of the oxidation products at pH 8. What results can be expected operating UV/H₂O₂ oxidation under alkaline conditions?

In the work the oxidation process was performed at pH = 3. It is expected that the reaction mixture obtained at pH 8.0 will have a different composition than that obtained at pH = 3. The oxidation process with OH radicals will take place faster at the alkaline pH. In the work, the pH was changed only in the nanofiltration process to affect the phenomenon of electrostatic repulsion. Nevertheless, the authors wanted to compare the filtration parameters of the same reaction mixture at a different pH.

8 – The authors may indicate a strategy to avoid the negative effect of the oxidation products on the overall performance of the integrated treatment system coupling oxidation and nanofiltration processes.

The influence of the oxidation process on the nanofiltration process is a complex phenomenon. In order to indicate a strategy to avoid the negative impact of oxidation products on the overall efficiency of the integrated oxidation and nanofitration process, it requires further in-depth studies, for example measurements of zeta potential. Further research will be continued to answer the question.

Round 2

Reviewer 1 Report

The paper was consequently improved, but some relevant informations lack for having a scientific discussion about experimental results.

The authors say that these results will be added to a next work; 

These measurements are needed  to have a serious discussion of the results. If it's not possible to measure pH and conductivity of filtrated solutions, it is possible, at least; to re-prepare solutions and to performe measurements (pH, zeta, elec cond, etc...),

these measurements (permeate and concentrate) are required for understanding phenomena that act on separation.

Contact angles: It was interesting to perform calculation with van Oss equation. The experimental data are suitable for the calculation.

This calculation increases the accuracy of the polar component, by indentifying the nature of the surface and so, the nature of the interaction with the solution.

Author Response

Dear Editor,

Please find enclosed a revised version of the manuscript ID: water-427839 titled "Impact of advanced oxidation products on nanofiltration efficiency” by Renata Żyłła, Rafał Milala, Irena Kamińska, Marcin Kudzin, Marta Gmurek and Stanisław Ledakowicz.
I would like to thank the reviewers for their work. We acknowledge the reviewers for all their suggestions. All comments were helpful and have been taken under consideration to improve the quality of our manuscript. We are grateful for very kind revision of Reviewer #2 and Reviewer #3. Moreover, we have tried to improve manuscript according to Reviewer #1 suggestion and we truly believe that it meets with approval.
We hope that you will find the revised manuscript suitable for publication.

Yours sincerely,
Renata Żyłła

Round 3

Reviewer 1 Report

This paper can be acceted in the present form