Extraction of Antioxidants from Grape and Apple Pomace: Solvent Selection and Process Kinetics

Round 1

Reviewer 1 Report

OVERALL

In the manuscript, the authors deal with the solid-liquid extraction of polyphenols and antioxidants from natural sources, which are actually waste, a byproduct of industrial fruit processing. Specifically, these are apple (AP) and grape (GP) pomace. The authors describe the efficiency and kinetics of extraction using solvents of industrial importance. They use several types of organic solvents, including water mixtures in different proportions with organic solvents and for several temperatures. They determined the extraction capacities of the solvent mixture as well as the extraction kinetics of some antioxidants and ingredients in the extracts. Based on the presented results, the authors conclude that the water-to-ethanol mixture (30:70) at medium and high temperatures is a simple, fast enough, and efficient way for extraction. However, they also point out some other solvents (ethylene glycol) that have good extraction capacities, especially from grape pomace (GP). However, ethylene glycol is practically inapplicable in the food industry due to its toxicity. They conclude that grape pomace (GP) is more suitable for the extraction of antioxidants (gives a higher content) than apple pomace (AP), but that AP contains more liquids whose composition is not so complex for extraction.

COMMENT 1

In subheading 2.8, listing the adsorption model you used to process the data (Langmuir adsorption isotherm), you give two empirical equations. You say that instead of the Langmuir isotherm equation (Eq.1), you use an empirically established second-order model/relation (Eqs. 2 and 3). Here you specify the concentration parameters q_l, q_e and the kinetic constant k_d. However, the constant k_d does not appear anywhere in Eqs. 2 and 3. On the other hand, there is a proportionality constant k_E in Eqs. 2 and 3 that is not defined in the text. Finally, since the concentration of the extracted component is time-dependant, namely C_{A, e} ~ t ,(Eq. 1), the Eqs. 2 and 3 are actually Langmuir isotherms. Please reconcile your statements.

Also, take care of the subscripts in Eq. 3.

COMMENT 2

In the Abstract and Introduction, you say a solid-liquid extraction. However, in the first sentence of subheading 3.1 you state a liquid-liquid extraction.

COMMENT 3

Page 5, the last sentence is marked red. Seems to be an unfinished paragraph?

COMMENT 4

Here the extraction kinetics is a fast process. Graphs 1 to 5 mostly show that the concentration of the extracted component is quite high even at the close-to-zero time. Therefore, I suggest a generalized Eq. 2., for fitting. Namely, apply the form:

 q_l=q₀+q_e·b·t/(1+bt)

where q₀=q_l  at t=0. This way pushes upward the fitting curve and I believe the parameters will change significantly and become more accurate and reliable. Now you give the same statistical significance to zero point as for all measured points (you forcing the fitting through a zero point) which is inappropriate in most cases and visible on Figure 2, Figure 3, Figure 4, Figure 5. Exclude the zero point from the fitting.

COMMENT 5:

You say, Figure 3., displays the results for TPC extraction….. It is really interesting to observe the dramatic increase in final yields. However, according to the graphs, there is no increase in yield, especially not a dramatic increase. For the kinetic parameters, too. I think there is no difference within 95% confidence. Try t-test. Apply the COMMENT 4., quantify the yields and speeds then compare.

COMMENT 6:

Figure 4. and Figure 5. Apply Comment 4., compare and discuss. Especially valid for Figure 4., since the corresponding kinetic parameters presented in Table 1. are used in Arrhenius calculations of the activation energies. See Comment 8.

 

COMMENT 7:

Figure 6. There is a significant difference in the number of measuring points on Graphs 6. (GP) compared to the graphs related to AP. Also, the variance of individual measurements is given, which is not the case with graphs related to AP. AP and GP groups of measurements seem to be methodologically different. Explain.

COMMENT 8:

Table 1. comprises kinetic parameters of total phenolic content (TPC) extraction for the case of ethanol-water mixtures and AP. You apply Arrhenius calculation to derive activation energies but have measurements for four temperatures only. For the first three temperatures, the kinetic parameters differ within their errors, i.e. there is no significant change. Linear regression on the Arrhenius graph can be implemented through three or two points. (25-70^oC and 70-90 ^oC). According to that, the activation energies specified on page 10, last paragraph, are incorrect. Apply comment 4 and recalculate.

 

COMMENT 9:

Table 3. and Table 4., related to the GP. For none of the graphs given in Figure 6., case TPC, is possible to plot the adsorption curve by using the corresponding parameters given in Table 3. For example, the capacity of about 70 mg TPC is represented for 70% EtOH at 90 oC (Figure 6). From Table 3. q1+q2 =78. However, the plotted curve does not give the same value for the capacity. The reason is the value of the k₂ parameter is too small to give a significant contribution to the final adsorption capacity curve.

Or the case of Acetone 80 % (v/v) 25 °C! q2=10.93 ± 8271.87,   k2=0.002 ± 0.317. Absolutely nonsense.

Reconsider all the calculus.

In this case, the second addition in Eq 3 can be neglected and reduced to Eq.2.  

Author Response

We are very grateful to the reviewer for his/her enlightening comments. We have our best to follow them in this new version of the manuscript, which contains several changes as described in the following paragraphs that are written in red font for its easier identification.

Comment 1

In subheading 2.8, listing the adsorption model you used to process the data (Langmuir adsorption isotherm), you give two empirical equations. You say that instead of the Langmuir isotherm equation (Eq.1), you use an empirically established second-order model/relation (Eqs. 2 and 3). Here you specify the concentration parameters ql, qe and the kinetic constant kd. However, the constant kd does not appear anywhere in Eqs. 2 and 3. On the other hand, there is a proportionality constant kE in Eqs. 2 and 3 that is not defined in the text. Finally, since the concentration of the extracted component is time-dependant, namely CA, e ~ t ,(Eq. 1), the Eqs. 2 and 3 are actually Langmuir isotherms. Please reconcile your statements.

Also, take care of the subscripts in Eq. 3.

Thank you for the observation. In fact, the isotherm is of a Langmuir type, while the kinetics follow second-order kinetic models (first order kinetic models were also fitted to experimental data but they were unable to explain the mild increasing trend in several cases, so we do not consider them further to avoid a too lengthy manuscript). We have reduced the equations to only one,  founded on the observation made by  the reviewer in the comment 4, and define the parameters carefully here, using them in the discussion and tables afterwards.

Comment 2

In the Abstract and Introduction, you say a solid-liquid extraction. However, in the first sentence of subheading 3.1 you state a liquid-liquid extraction.

It is a solid-liquid extraction or lixiviation. Excuse the error, we have corrected it.

Comment 3

Page 5, the last sentence is marked red. Seems to be an unfinished paragraph?

It is an error and we have corrected it. Thank you for the observation.

Comment 4

Here the extraction kinetics is a fast process. Graphs 1 to 5 mostly show that the concentration of the extracted component is quite high even at the close-to-zero time. Therefore, I suggest a generalized Eq. 2., for fitting. Namely, apply the form:

ql=q0+qe·b·t/(1+bt)

where q0=ql  at t=0. This way pushes upward the fitting curve and I believe the parameters will change significantly and become more accurate and reliable. Now you give the same statistical significance to zero point as for all measured points (you forcing the fitting through a zero point) which is inappropriate in most cases and visible on Figure 2, Figure 3, Figure 4, Figure 5. Exclude the zero point from the fitting.

We have done it as indicated by the reviewer and are most grateful for the indication. In fact, this is the only empirical kinetic model that appears in this revised version of the manuscript, as it is adequate for all cases. All the discussion has been modified accordingly, as parameter values and trends are notably different from those observed with the models used in the previous version of the manuscript.

 

Comment 5

You say, Figure 3., displays the results for TPC extraction….. It is really interesting to observe the dramatic increase in final yields. However, according to the graphs, there is no increase in yield, especially not a dramatic increase. For the kinetic parameters, too. I think there is no difference within 95% confidence. Try t-test. Apply the COMMENT 4., quantify the yields and speeds then compare.

Thank you for the comment. Maybe we have not explained it reasonably; we mean that solutions based on glycols lead to higher TPC values in the case of apple pomace in comparison to other solvents. Therefore, we have written the paragraph again to make our meaning clearer.

Comment 6

Figure 4. and Figure 5. Apply Comment 4., compare and discuss. Especially valid for Figure 4., since the corresponding kinetic parameters presented in Table 1. are used in Arrhenius calculations of the activation energies. See Comment 8.

Thank you for the suggestion. We have performed again the fitting, as the kinetic model has changed to the model suggested by the reviewer in Comment 4. Discussion on the new parameters and parameters values can be found in section 3.2 (Kinetic modeling of extraction process).

Comment 7

Figure 6. There is a significant difference in the number of measuring points on Graphs 6. (GP) compared to the graphs related to AP. Also, the variance of individual measurements is given, which is not the case with graphs related to AP. AP and GP groups of measurements seem to be methodologically different. Explain.

Thank you for the comment. We have applied the same methodology for both wastes, performing the tests and analyses in triplicate (some runs has been performed again during this revision, in fact) but we agree that there are differences in sampling time. In any case, we feel that these differences only affects slightly initial and final TPC or AC values, as well as the kinetic parameters, and these effects are within the experimental error observed in most cases (± 10%), so evidences and the subsequent discussion and conclusions are not affected by the experimental procedure followed in each case.

Comment 8

Table 1. comprises kinetic parameters of total phenolic content (TPC) extraction for the case of ethanol-water mixtures and AP. You apply Arrhenius calculation to derive activation energies but have measurements for four temperatures only. For the first three temperatures, the kinetic parameters differ within their errors, i.e. there is no significant change. Linear regression on the Arrhenius graph can be implemented through three or two points. (25-70oC and 70-90 oC). According to that, the activation energies specified on page 10, last paragraph, are incorrect. Apply comment 4 and recalculate.

We are very grateful to the reviewer for this comment. As we have entirely modified the statistical study by changing the kinetic model, please, refer to the new discussion in section 3.2.

Comment 9

Table 3. and Table 4., related to the GP. For none of the graphs given in Figure 6., case TPC, is possible to plot the adsorption curve by using the corresponding parameters given in Table 3. For example, the capacity of about 70 mg TPC is represented for 70% EtOH at 90 oC (Figure 6). From Table 3. q1+q2 =78. However, the plotted curve does not give the same value for the capacity. The reason is the value of the k2 parameter is too small to give a significant contribution to the final adsorption capacity curve. Or the case of Acetone 80 % (v/v) 25 °C! q2=10.93 ± 8271.87,   k2=0.002 ± 0.317. Absolutely nonsense. Reconsider all the calculus.

We are very grateful to the reviewer for this comment. As we have entirely modified the statistical study by changing the kinetic model, please, refer to the new discussion in section 3.2.

Reviewer 2 Report

1. The kinetic models’ analysis and adsorption isotherm in this article is very confusingly written. The author should search for some articles describing adsorption on kinetic models’ analysis and adsorption isotherm and refer to the writing of these documents.
2. The author mentions Langmuir adsorption isotherm but can't see any data analysis and graphics.
3. In addition to Langmuir adsorption isotherm, there are many kinds of adsorption isotherms, such as Freundlich isotherm, Langmuir- Freundlich adsorption isotherm, etc. The author should make a comparative analysis to find the most suitable adsorption isotherm.
4. In all kinetic analyses and graphs, the author directly writes the most suitable kinetic analyses. The author should list the data calculated by the two kinetic equations and graphs at the same time and list the analytical data, such as the R-squared comparison, to show the kinetic equation used by the authors is the optimal kinetic equation.
5. The author did not set uniform extraction conditions. For example, Figure 1 is completely ethanol 70 % (v/v) at comparison ofthe kinetic diagram of different temperatures but the picture 2 becomes acetone 25 °C, acetone 80 % 25 °C, water 25 °C and ethanol 96 % (v/v) 70 °C comparison, in Figure 3 it becomes the comparison of ethylene glycol 90 % 70 °C and propylene glycol 90 % 70 °C, it is suggested that the author use Figure 1 as the standard, and a solution ratio or solution is compared at 4 temperatures, in order to achieve a unified standardization of the analysis.

In my opinion, this article can be published in this journal if the authors explain all the unclear parts and offer enough new datas. Also, the authors should recheck and correct the errors of typos and grammar before sending out this manuscript. By all the aspects enumerated above, the author should make a rigorous revision of the paper before its publication.

Comments for author File: Comments.pdf

Author Response

Comment 1

The kinetic models’ analysis and adsorption isotherm in this article is very confusingly written. The author should search for some articles describing adsorption on kinetic models’ analysis and adsorption isotherm and refer to the writing of these documents.

The author mentions Langmuir adsorption isotherm but can't see any data analysis and graphics.

In addition to Langmuir adsorption isotherm, there are many kinds of adsorption isotherms, such as Freundlich isotherm, Langmuir- Freundlich adsorption isotherm, etc. The author should make a comparative analysis to find the most suitable adsorption isotherm.

We are very grateful to the reviewer for his/her comments. As we have entirely modified the statistical study by changing the kinetic model, please, refer to the new discussion in sections 3.1. and 3.2.

In this case, we have studied the kinetics of extraction processes so, in reality, the focus of the paper is on the dynamics of the process, not on the L-S equilibrium. In fact, no adsorption is studied so it is very complex to determine the equilibrium state at a given temperature as if an adsorbate enters the porous matrix and interacts with the pore surfaces. Moreover, the complex chemical nature of the extracts make very difficult to study their adsorption onto the depleted matrix even if this is our aim, which is not the case. Therefore, we have tried our best to explain the desorption – extraction results, discussing them with those found by other authors, but always keeping in mind that this is an extraction research focused on the process dynamics.

Comment 2

In all kinetic analyses and graphs, the author directly writes the most suitable kinetic analyses. The author should list the data calculated by the two kinetic equations and graphs at the same time and list the analytical data, such as the R-squared comparison, to show the kinetic equation used by the authors is the optimal kinetic equation.

Thank you very much for your comment. We agree with the reviewer that there are several empirical kinetic models that could be of interest; we have tried two of them in the first version of this manuscript and a third one suggested by the first reviewer in this version. In fact, this latter was even better than the previous ones when considering goodness-of-fit and kinetic constants-related statistical parameters, as squared R or Student’s t. As the number of runs is very notable, to keep the size of the manuscript to an adequate extension, we have selected the best kinetic model in terms of goodness-of-fit and parameter reliability and discuss on the values obtained. The complexity here is due to the notable number of proposed kinetic models in the literature (please, see the last reference of the list, by Natolino and Da Porto, to appreciate the number of models proposed by now).

Comment 3

The author did not set uniform extraction conditions. For example, Figure 1 is completely ethanol 70 % (v/v) at comparison of the kinetic diagram of different temperatures but the picture 2 becomes acetone 25 °C, acetone 80 % 25 °C, water 25 °C and ethanol 96 % (v/v) 70 °C comparison, in Figure 3 it becomes the comparison of ethylene glycol 90 % 70 °C and propylene glycol 90 % 70 °C, it is suggested that the author use Figure 1 as the standard, and a solution ratio or solution is compared at 4 temperatures, in order to achieve a unified standardization of the analysis.

Thank you for the suggestion. To facilitate the comparison of the results, we have done again all figures, adding some more in the case of grape pomace extraction. We hope this way it is would be easier to compare the results for diverse solvents at the same temperature. Some of the experiments have been done again to obtain more reliable data, though trends have remain the same. In any case, the aim of this paper is to compare solvents within temperature values where they are known to be effective and to select the best ones for further optimization of the extraction processes (and kinetic analysis linked to solid structure dynamics), results that we intent to publish in the next future.

In my opinion, this article can be published in this journal if the authors explain all the unclear parts and offer enough new datas. Also, the authors should recheck and correct the errors of typos and grammar before sending out this manuscript. By all the aspects enumerated above, the author should make a rigorous revision of the paper before its publication.

We appreciate very much the comments and suggestions of the reviewer, as we feel he/she has helped us to improve our original manuscript

Reviewer 3 Report

Manuscript presented for review: “Extraction of antioxidants from grape and apple pomace: solvent selection and process kinetics” is a very interesting and well written.

The experiment was planned very carefully. The Introduction section includes a new elements of literature review. Aim of work is present at the end of Introduction section.

The collected experimental material and used methods do not raise any objections. The obtained results were very well describe and present. The discussion section presents a very good comparison of the obtained results with other results available in the data basis.

General opinion: I think, that presented manuscript is a very valuable and should be published in Applied Sciences journal in presented form (without any changes)

Author Response

Manuscript presented for review: “Extraction of antioxidants from grape and apple pomace: solvent selection and process kinetics” is a very interesting and well written.

The experiment was planned very carefully. The Introduction section includes a new elements of literature review. Aim of work is present at the end of Introduction section.

The collected experimental material and used methods do not raise any objections. The obtained results were very well describe and present. The discussion section presents a very good comparison of the obtained results with other results available in the data basis.

General opinion: I think, that presented manuscript is a very valuable and should be published in Applied Sciences journal in presented form (without any changes)

We are very grateful to the reviewer for his/her comments; they are really encouraging.

Reviewer 4 Report

The basic question in case of this paper is what is the novelty of all of the presented results. It's really hard to find justification in the introduction or deeper discussion in the Results and discussion part. What kind of solutions where proposed by other authors? What kind of mixtures they applied for extraction, why authors decided to use such mixtures in their studies?  And if you are first who applied this it is worth underlining.
Below you can find more detail comments and doubts concerning your paper
General rule in a good quality papers is that you always explain abbreviations, even if you think that it is obvious in your environment. What I missed in this paper explanation what TEAC, GAE and ONU abbreviations are.
The size of the font in Figure 1 should be the same
Introduction: Check out your calculations about this 63 tons of apple wastes, something's wrong with them.
Make up your mind about the type of letter - capital letters or not -  which you use in the name Organization of vine and wine, as well as two lines below.
I believe that in order to Franz Ritter von Soxhlet  was a very famous chemist we are obligated to use capital letters using his name.
Figure 1 and 5 tell us not too much, maybe about some kinetics, but you use different ranges in oy axis so it is not possible to compare the obtained results. It would be much better to calculate the significance of differences for the results obtained for the 60th minute and then we could say something about differences among these different extraction processes which you applied. Kinetic analysis is very interesting and it's worth analyzing however, I believe that from practical and applied sciences perspective the most important thing is the final result - how much of the substance was extracted in the final minute of the process. Luckily, you refer to this in the abstract but a reader cannot be sure about the statistical significance which of the applied mixture is the best, we need a statistical proof.
Figure 2, 3, 4 and 6  are much better, but still the question is about the significance of the differences of the results obtained at least for the 60th minute.
Consider also additional statistical analysis in figures 7-10
Page 7 viscosity units: If you want to use poise unit you need to remember that the official symbol in this case is P, so it should be cP not cp

Author Response

The authors appreciate very much the comments and suggestions of the reviewer. They have allowed us to improve the paper greatly, in our view.

Comment 1

The basic question in case of this paper is what is the novelty of all of the presented results. It's really hard to find justification in the introduction or deeper discussion in the Results and discussion part. What kind of solutions where proposed by other authors? What kind of mixtures they applied for extraction, why authors decided to use such mixtures in their studies?  And if you are first who applied this it is worth underlining.

We are very grateful to the reviewer for his/her comment. Indeed, it is always critical to highlight what is novel in a manuscript. We have made an additional effort to clarify it in the last sentence of the Abstract and in the last paragraph of the Introduction, also modifying the Results and Discussion section to make our idea clearer. In essence:

“The aim of this work is to develop fast and efficient extraction processes for both wastes (AP and GP), leading to rich antioxidants liquors, by screening different solvents and conditions and to understand process efficiency by a comparative kinetic analysis of total phenolic content (TPC) and antioxidant capacity evolution with extraction time. Furthermore, a compositional study will be performed with the liquors obtained in the best conditions.”

There are indeed some studies in the extraction process of phenolics from agrowastes, as we have referenced in the manuscript. Our manuscript is founded in the need of further systematic studies of   comparison in identical conditions of extraction processes. For that we have taken two very important agrowastes: grape and apple pomaces. A further, and more important, our main innovative and novel approach resides in the kinetic comparison of TPC and AC results in both cases. While TPC kinetic analysis is a common issue, as proved by several references in this work, AC kinetic analysis is not common and, as far as we know, it has not been performed with the residues here studied. We are convinced, as we tried to show in the discussion and conclusions part, that our study will be a basis to inspire and allow systematic kinetic comparisons taken TPC and AC as process targets.

Comment 2

Below you can find more detail comments and doubts concerning your paper
General rule in a good quality papers is that you always explain abbreviations, even if you think that it is obvious in your environment. What I missed in this paper explanation what TEAC, GAE and ONU abbreviations are.

Thank you for the suggestion. We have included the definition of all abbreviations when first presented in the manuscript,

in the Abstract, Introduction and Materials and Methods sections.

Comment 3

The size of the font in Figure 1 should be the same

Thank you for the suggestion. We have corrected the figure accordingly and modified the x- and y-axis scales to facilitate comparison of results between graphs in Figures 1 and 2.

Comment 4

Introduction: Check out your calculations about this 63 tons of apple wastes, something's wrong with them.

Sorry for the mistake. Indeed, this number is too low: in reality, apple wastes in Spain amounts to almost 156,000 tons in 2020.

Comment 5

Make up your mind about the type of letter - capital letters or not - which you use in the name Organization of vine and wine, as well as two lines below.

Thank you for the remark. We have used capital letters, which is the most correct thing to do.

Comment 6

I believe that in order to Franz Ritter von Soxhlet was a very famous chemist we are obligated to use capital letters using his name.

We are grateful to the reviewer for the suggestion. Indeed, it is most proper to use “Soxhlet”, starting in capital letters when referring to the apparatus named after Franz Ritter von Soxhlet. We have ensured that is done correctly along the manuscript.

Comment 7

Figure 1 and 5 tell us not too much, maybe about some kinetics, but you use different ranges in oy axis so it is not possible to compare the obtained results. It would be much better to calculate the significance of differences for the results obtained for the 60th minute and then we could say something about differences among these different extraction processes which you applied.

 

Thanks a lot for the suggestion. We have modified the axis of the figures to allow a better comparison of the results. But we are obliged to respectfully disagree with the reviewer regarding the importance and significance of the full-time courses. In our position a systematic analysis of progress curve analysis is not useful to tell “maybe about some kinetics” but for a comprehensive understanding of the process. We agree that the data at 60 min is probably the most important for one-single point determination of yield parameter and probably very significant for first assessment of the process suitability, but only by an observation, analysis, and modelling of the full time of the extraction process we can gain the knowledge that we need and we aim in the manuscript. Please note that it is here where the main foundation of the innovative approach of our manuscript resides (see Comment 1).

We do not aim at extending unnecessarily the document of responses, but the comment of reviewer is indeed very interesting since it shows a current debate or dilemma in the analysis of performance of extractive processes in the context of economy or resources in research: final point determination versus time course significance. From chemical engineering background, we are convinced of the importance of the latest.

In consequence, we have modified the definition of the kinetic model to define a TPC (total phenolic content) or AC (antioxidant capacity) at equilibrium (q_e) by summing up the values of these parameters in the washing, or fast, initial stage (q₀) and the slower extraction stage (q₁). Although we agree with the reviewer that yield at 60 minutes, with its error, can help in the comparison of the extraction processes to find the most suitable, this can also be done by comparing q_e, profiting from the kinetic analysis, though it is empirical, to have an statistically significant information of the comparison, as the values of this parameter comes with their standard errors in Tables 1 to 4. In any case, all values in the main text are now with their error interval to facilitate a direct comparison.

Comment 8

Kinetic analysis is very interesting and it's worth analyzing however, I believe that from practical and applied sciences perspective the most important thing is the final result - how much of the substance was extracted in the final minute of the process. Luckily, you refer to this in the abstract but a reader cannot be sure about the statistical significance which of the applied mixture is the best, we need a statistical proof.

We are very grateful to the reviewer for the suggestion and we agree that, in the end, what is important is the final yield in TPC and AC terms (possibly, more important in terms of antioxidant capacity). However, the kinetic analysis allows also for the understanding of when most phenolics or antioxidant activity have been achieved, optimizing the use of processing time. At the same time, and due to its statistical nature, this type of analysis permits data smoothing, thus reducing experimental errors between data obtained at different processing times. This is why we suggest to compare q_e values, with their errors, to have a more reasonable view of the differences. Thus, we provide data on final yields and comments on why these yields are obtained in each case together with an in-depth, though empirical, analysis of all time-courses, focusing not only on the final yield but also on what happens till this yield is attained.

Comment 9

Figure 2, 3, 4 and 6 are much better, but still the question is about the significance of the differences of the results obtained at least for the 60th minute.

Thank you for the indication. The statistical analysis can be done in terms of error for data at 60 minutes, but it is more appropriate to profit from the kinetic analysis; in particular from the q_e value and error, to understand differences for diverse wastes, solvents and conditions.

Comment 10

Consider also additional statistical analysis in figures 7-10

We are grateful to the reviewer for the indications. We have added more information regarding average statistical errors and used adequate ranges for y and x-axis in all figures to facilitate comparisons to the readers, apart from the statistical and kinetic data contained in Tables 1 to 4.

Comment 11

Page 7 viscosity units: If you want to use poise unit you need to remember that the official symbol in this case is P, so it should be cP not cp

Thank you for the remark. We have modified all viscosity units to cP.

Round 2

Reviewer 1 Report

After the first review of the manuscript, the authors made corrections by accepting my remarks and significantly improving the visibility and readability of the results. Explanations have been added and graphs of kinetic curves are shown separately. Although there was no need for that since this could be shown in several representative graphics. Especially since the parameters of all kinetic curves are shown in Tables 1, 2, 3, and 4. The tables are a key part of this manuscript however the comments and especially the conclusions derived do not provide substantial pieces of information and are still confusing. This is especially true for the text labeled in red.

 

Ad. 1. The error in determining the kinetic parameter, k_d, is in the range of 10-50%. This applies to both AP and GP and TPC and DPPH. Therefore, the precision expressed with two significant digits is overestimated. In this regard, the conclusion is that k_d the parameter does not depend significantly on temperature and EtOH concentration, Table 1. Even extraction is more efficient with pure water compared to EtOH, (q₀ is 19 vs. about 10, respectively) in the given temperature range. However, in the case of water, the estimated error of the kinetic parameter is 50%.

 

Ad. 2. The key point is Table 3 and Table 4. Tables say that in many cases the parameter q₀ is greater than parameter q_e which is senseless. For example, Acetone 80 % (v/v) 25 °C,  46.83 ± 1.15;  16.76 ± 4.76;  3.4·10 -3 ± 2.1·10 -3   0.815, Table 3. In some cases q₀≈q_e but kinetic parameter k_d>0. For example, Water 25 °C 8.77 ± 0.37;  3.17 ± 0.35;  0.23 ± 0.023;  0.976, Table 4.

I don't know if it's an oversight and columns replacement or something else. In any case, the discussion and conclusion derived from these tables are impossible to follow.

Author Response

The authors are very grateful to the reviewer for the comments and suggestions. They have helped us to improve our manuscript notably, in our view.

Comment 1

The error in determining the kinetic parameter, k_d, is in the range of 10-50%. This applies to both AP and GP and TPC and DPPH. Therefore, the precision expressed with two significant digits is overestimated. In this regard, the conclusion is that k_d the parameter does not depend significantly on temperature and EtOH concentration, Table 1. Even extraction is more efficient with pure water compared to EtOH, (q₀ is 19 vs. about 10, respectively) in the given temperature range. However, in the case of water, the estimated error of the kinetic parameter is 50%.

Thank you for the comment. Given the heterogeneous nature of the solids that we have dealt with, this error is complex to avoid: we have performed the runs in duplicate, while measurements were carried out in triplicate for each sample, notably reducing experimental error. However, even errors as low as 5-10% results in errors in kinetic constants such as those indicated by the reviewer. Thus, we have reduced the significant digits in all tables to consider the experimental error.

Water was more effective to extract compounds with phenolic moieties in the case of apple pomace, but these mixtures showed a notably lower antioxidant power than mixtures obtained with more hydrophobic solvents (most of them including, however, water in their composition). This was interesting and probably related to the nature of the extracted compounds. However, as our intention was to discern the best solvent and best conditions to achieve extracts with a high antioxidant power, only these extracts were analyzed by GC-MS-MS. Water extracts at room temperature showed a relatively low antioxidant power. Reis et al. (2012) indicated that water at that low temperature was good “to extract hydroxycinnamic acids, flavonols, flavanols, dihydrochalcones and flavones present in the AP. However, water was not the ideal solvent to extract the quercetin glycosides”. These quercetin glycosides are just the compounds that are most abundant in the hydroalcoholic extracts. This evidence suggests that these later compounds possess a high antioxidant activity in the DPPH test.

Reis, S. F., Rai, D. K., Abu-Ghannam, N. (2012). Water at room temperature as a solvent for the extraction of apple pomace phenolic compounds. Food Chemistry, 135(3), 1991-1998.

Comment 2

The key point is Table 3 and Table 4. Tables say that in many cases the parameter q₀ is greater than parameter q_e which is senseless. For example, Acetone 80 % (v/v) 25 °C,  46.83 ± 1.15;  16.76 ± 4.76;  3.4·10 -3 ± 2.1·10 -3   0.815, Table 3. In some cases q₀≈q_e but kinetic parameter k_d>0. For example, Water 25 °C 8.77 ± 0.37;  3.17 ± 0.35;  0.23 ± 0.023;  0.976, Table 4.

We appreciate very much this correction by the reviewer. Indeed, we have modified the definition of the model to avoid misunderstandings: q_e have been defined as the equilibrium value of either TPC or DPPH-measured antioxidant activity, discerning a q₀ value for the rapid washing phase and a q₁ value for the slow extraction phase. The sum of these two parameters is q_e. In this way, the comparison can be notably enhanced.

Comment 3

I don't know if it's an oversight and columns replacement or something else. In any case, the discussion and conclusion derived from these tables are impossible to follow.

Thank you for the comment. We have refurbished all tables and corrected the discussion on the results there compiled point by point, to make it clearer.

Reviewer 2 Report

The authors already revised their paper point-by-point according to reviewers’ comments and tried their best to improve our manuscript. In my opinion, this article can be published in this journal. 

Author Response

The authors already revised their paper point-by-point according to reviewers’ comments and tried their best to improve our manuscript. In my opinion, this article can be published in this journal.

We are very grateful to the reviewer for his/her previous comments, as they have helped us to improve notably the first version of the manuscript, and really appreciate his/her acceptance of our answers and modifications in this second version.