Three-Dimensional Particle-Discrete Datasets: Enabling Multidimensional Particle System Characterization Using X-Ray Tomography
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsDear Authors, the paper is interesting but I believe that it would benefit from a greater focus on the application aspects of the research and reduction/removal of information on topics where your descriptions duplicate information published in detail elsewhere.
Section 1.2.2 contains information on stereological bias which has been widely published elsewhere and can be removed and replaced by reference to previous publications.
Section 2.1 contains information on X-ray tomography which has been published in books and other papers and can be replaced by reference to previous publications.
Sections 2.2.3 and 2.2.4 contain widely known information on Artefacts and Image Processing. It would be more useful to the reader for the paper to focus on how these affect the specific examples presented in the paper and to remove/reduce information which is published elsewhere.
Line 120: Is the referenced paper by Ueda et al. one of the original publications which describes stereological bias?
Line 130: Is the paper by Schulz et al. the definitive paper on MLA?
Lines 134-135: "To address stereological bias, one technique involves re–cutting and rotating of already embedded particulate samples to minimize particle size–dependent segregation effects [11]". This method is described in the MLA Operating Manual Module 1 Sample Preparation first published 2003, pre-dating the referenced work, and this original work should be acknowledged.
Lines 154-156: "structure does not refer exclusively to the geometric dimensions, but also to the internal structure with regard to the individual phases [19] or defects [20], such as cracks, or the association within hetero–agglomerates like host–guest particle structures [21]". Can you identify alternative references which define structure in these terms?
Line 222: This refers the reader to the primary author's PhD thesis for details of the measurement system. Please include details of the measurement system in the paper itself and where tomographic data or images are presented include details of the measurement conditions used in each case.
Line 497: "Previously implemented for 2D mineralogical analysis [11,22]". The use of carbon particles has been a standard feature of sample preparation for SEM-based automated mineralogy systems (e.g. QEMSCAN, MLA) since the 1980s and 1990s respectively. This pre-dates the two referenced publications by TU Bergakademie Freiberg researchers and it would be appropriate to acknowledge the earlier work.
Comments on the Quality of English LanguageIn general the standard of English is high.
However it would be better to use the plain English word "example" rather than the word "exemplary" which is used in the title of several Figures.
Author Response
Dear Authors, the paper is interesting but I believe that it would benefit from a greater focus on the application aspects of the research and reduction/removal of information on topics where your descriptions duplicate information published in detail elsewhere.
We totally agree on that. The thing is that within our community of particle characterization referring to process analysis and development we are not using (or too seldomly) these mentioned tools. The last section should serve as a summary but with no quantitative statement maybe not needed. Instead, we shortened the passage and add it as a final paragraph without extra section.
“Although the theoretical principles of the 3D characterization of particles have been discussed in detail in the past and the advantages of this method over 2D imaging and the description with equivalent sizes are obvious, it is still too rarely used in the particle characterization community. This needs to change. Relationships of particle properties are rarely one--dimensional, and when they are, they often lack precision. Particle--discrete multidimensional analysis offers a powerful tool to gain new insights in on processes and parameter correlations.”
As an additional clarification, that we are aware of that fact, but, nonetheless, want to go that way, we changed the introduction and hope this will clarify the importance of the topic and the relation to the Topical Collection of the journal Powders.
“This publication of the Topical Collection \textit{Multidimensional Particle Properties: Characterization, Description, Separation} is part of the DFG--funded \textit{Priority Programme SPP~2045} and summarizes the work on three--dimensional particle characterization within the central characterization subproject. The focus is on the particle--discrete description of the 3D particle data, which is not fundamentally new within the computed tomography and image analysis community, but in our opinion is underrepresented in the process--related particle characterization community and requires further motivation. In the following, we therefore focus on the motivation behind the topic and show how this data can serve as the basis for multidimensional property correlations. In the individual sections, we focus on how the quality of the resulting image data is affected and explore ways to maximize it in order to create reliable datasets suitable for multidimensional particle analysis. Application examples are briefly outlined and further considered, but are described in detail in the project-related publications.”
Section 1.2.2 contains information on stereological bias which has been widely published elsewhere and can be removed and replaced by reference to previous publications.
As a general motivation and to emphasize the importance we think that we need the passage here. But we changed the reference (next point, you are totally right, this is not really a proper basic reference. Thank you for this advice)
Line 120: Is the referenced paper by Ueda et al. one of the original publications which describes stereological bias? No, you are right… we changed that:
@Article{Laetti2001,
author = {Lätti, D. and Adair, B.J.I.},
journal = {Minerals Engineering},
title = {An assessment of stereological adjustment procedures},
year = {2001},
issn = {0892-6875},
month = dec,
number = {12},
pages = {1579--1587},
volume = {14},
doi = {10.1016/s0892-6875(01)00176-5},
publisher = {Elsevier BV},
}
Section 2.1 contains information on X-ray tomography which has been published in books and other papers and can be replaced by reference to previous publications.
Our experience is that the reader wants to have at least a rough idea how the method behind works without reading primary literature. You are right, the passage was a bit too detailed here. We shortened it by approx. 40%.
Sections 2.2.3 and 2.2.4 contain widely known information on Artefacts and Image Processing. It would be more useful to the reader for the paper to focus on how these affect the specific examples presented in the paper and to remove/reduce information which is published elsewhere.
We understand the point but, in our opinion, we think that the passage needs to stay. Our experience from the community shows that these things are not widely known.
Line 130: Is the paper by Schulz et al. the definitive paper on MLA?
We agree on that, we only focus on the applicational site of the MLA. We added Sutherland and Gottlieb as primary source here. Also, with respect to the sample preparation method with carbon black spacer particles.
@Article{Sutherland1991,
author = {Sutherland, D. N. and Gottlieb, P.},
journal = {Minerals Engineering},
title = {Application of automated quantitative mineralogy in mineral processing},
year = {1991},
month = {jan},
number = {7-11},
pages = {753--762},
volume = {4},
doi = {10.1016/0892-6875(91)90063-2},
}
Lines 134-135: "To address stereological bias, one technique involves re–cutting and rotating of already embedded particulate samples to minimize particle size–dependent segregation effects [11]". This method is described in the MLA Operating Manual Module 1 Sample Preparation first published 2003, pre-dating the referenced work, and this original work should be acknowledged.
To the best of our knowledge this reference is the first we could find with not only describing the method itself but also containing an evaluation. After an intense search for the MLA Operations Manual we are only able to find updated versions from different vendors but with no reference to the sample preparation. If you have a citable document, we would be very happy if you could share this with us that we can update on that.
Lines 154-156: "structure does not refer exclusively to the geometric dimensions, but also to the internal structure with regard to the individual phases [19] or defects [20], such as cracks, or the association within hetero–agglomerates like host–guest particle structures [21]". Can you identify alternative references which define structure in these terms?
In the process of shortening the manuscript text, we excluded this passage because it is too specific here. Additionally, so, we also reduced the amount of self-citation and stay more focused.
Line 222: This refers the reader to the primary author's PhD thesis for details of the measurement system. Please include details of the measurement system in the paper itself and where tomographic data or images are presented include details of the measurement conditions used in each case.
We try to find a balance between the amount of information we give in this kind of publication and we think that especially such an info is best kept within the referenced much more detailed papers. Additionally, the parameters are also stored within the OpARA database together with the raw images (and sometimes segmented ones).
Line 497: "Previously implemented for 2D mineralogical analysis [11,22]". The use of carbon particles has been a standard feature of sample preparation for SEM-based automated mineralogy systems (e.g. QEMSCAN, MLA) since the 1980s and 1990s respectively. This pre-dates the two referenced publications by TU Bergakademie Freiberg researchers and it would be appropriate to acknowledge the earlier work.
We also add Sutherland and Gottlieb here. Because Sutherland and Gottlieb are not mentioning that there could be segregation effects, which our colleagues from Helmholtz in Freiberg reported on, we still think that keeping a validating and an application source is still good.
Reviewer 2 Report
Comments and Suggestions for AuthorsThe manuscript relates to the topic of 3D particle-discrete analysis. The topic is important and I can confirm (to the best of my knowledge) that in many industries there are well established workflows based on 2D analysis and not everywhere they are adopted to 3D capabilities provided by modern imaging and image processing techniques. So, I support this direction of research.
At the same time, having experience in this area I didn't find any novelty in presented material. You can find lots of papers about different imaging techniques and associated challenges with regards to 3D particle analysis. You can find lots of papers about different image processing techniques and associated challenges with regards to 3D particle analysis as well (including different morphometrical parameters and their meaning). There are also papers about applications of 3D particle analysis for real industrial needs.
In contrast to abovementioned kind of papers, this manuscript is rather a review (survey) of well-known issues/challenges associated with 3D particle analysis. But as a review, it presents only some specific cases, sometimes going into deeper details, which are described very superficially. For example, when you discuss the problem of under/over segmentation in case of touching particles - you do not mention at all the watershed approach, which is widely used by the community. Instead you are saying that it is better to have the sample with particles physically separated from each other. Everyone will agree, but unfortunately quite often it is not possible.
The manuscript contains lots of qualitative statements. Just as an example the end of paper says: "The ability to describe and analyze a particle collective in multiple dimensions provides a fresh lens on processes and parameter correlations. We should embrace this tool." I agree that 3D analysis can provide you more information about the particles, but why is it important and how can it be beneficial for real applications? There is no answer and no practical examples to these question in the paper. Why someone should embrace this tool? Or let me reformulate - do you know anyone who is against 3D analysis?
Except low (in my opinion) scientific significance, the formal quality (structure of the paper, quality of figures, amount of references) of this work is good. I identified only one error/typo:
1) Fig 1a and b - small/large particles - discrepancy between text and figure caption
My suggestion is to think carefully about the purpose of this manuscript. If you developed something new (measurement or processing technique) and/or very useful (e.g., open source online database) for the community, you should probably focus more on the description of your development. If your work is a review of the challenges associated with 3D particle analysis then my advise would be to make clearer and more quantitative statements instead of current vague ones. Also, in the second case, examples with some real applications with quantitative assessment of positive impact (on some process/study) of modern 3D particle analysis versus 2D analysis will add a lot of sense in the paper and support your final conclusion.
Author Response
The manuscript relates to the topic of 3D particle-discrete analysis. The topic is important and I can confirm (to the best of my knowledge) that in many industries there are well established workflows based on 2D analysis and not everywhere they are adopted to 3D capabilities provided by modern imaging and image processing techniques. So, I support this direction of research.
At the same time, having experience in this area I didn't find any novelty in presented material. You can find lots of papers about different imaging techniques and associated challenges with regards to 3D particle analysis. You can find lots of papers about different image processing techniques and associated challenges with regards to 3D particle analysis as well (including different morphometrical parameters and their meaning). There are also papers about applications of 3D particle analysis for real industrial needs.
In contrast to abovementioned kind of papers, this manuscript is rather a review (survey) of well-known issues/challenges associated with 3D particle analysis. But as a review, it presents only some specific cases, sometimes going into deeper details, which are described very superficially. For example, when you discuss the problem of under/over segmentation in case of touching particles - you do not mention at all the watershed approach, which is widely used by the community. Instead you are saying that it is better to have the sample with particles physically separated from each other. Everyone will agree, but unfortunately quite often it is not possible.
The manuscript contains lots of qualitative statements. Just as an example the end of paper says: "The ability to describe and analyze a particle collective in multiple dimensions provides a fresh lens on processes and parameter correlations. We should embrace this tool." I agree that 3D analysis can provide you more information about the particles, but why is it important and how can it be beneficial for real applications? There is no answer and no practical examples to these question in the paper. Why someone should embrace this tool? Or let me reformulate - do you know anyone who is against 3D analysis?
We totally agree on that. The thing is that within our community of particle characterization referring to process analysis and development we are not using (or too seldomly) these mentioned tools. The last section should serve as a summary but with no quantitative statement maybe not needed. Instead, we shortened the passage and add it as a final paragraph without extra section.
“Although the theoretical principles of the 3D characterization of particles have been discussed in detail in the past and the advantages of this method over 2D imaging and the description with equivalent sizes are obvious, it is still too rarely used in the particle characterization community. This needs to change. Relationships of particle properties are rarely one--dimensional, and when they are, they often lack precision. Particle--discrete multidimensional analysis offers a powerful tool to gain new insights in on processes and parameter correlations.”
As an additional clarification, that we are aware of that fact, but, nonetheless, want to go that way, we changed the introduction and hope this will clarify the importance of the topic and the relation to the Topical Collection of the journal Powders.
“This publication of the Topical Collection \textit{Multidimensional Particle Properties: Characterization, Description, Separation} is part of the DFG--funded \textit{Priority Programme SPP~2045} and summarizes the work on three--dimensional particle characterization within the central characterization subproject. The focus is on the particle--discrete description of the 3D particle data, which is not fundamentally new within the computed tomography and image analysis community, but in our opinion is underrepresented in the process--related particle characterization community and requires further motivation. In the following, we therefore focus on the motivation behind the topic and show how this data can serve as the basis for multidimensional property correlations. In the individual sections, we focus on how the quality of the resulting image data is affected and explore ways to maximize it in order to create reliable datasets suitable for multidimensional particle analysis. Application examples are briefly outlined and further considered, but are described in detail in the project-related publications.”
Except low (in my opinion) scientific significance, the formal quality (structure of the paper, quality of figures, amount of references) of this work is good. I identified only one error/typo:
1) Fig 1a and b - small/large particles - discrepancy between text and figure caption
Thank you for carefully reading! You are right, we changed the text passage
“First of all, large particles result in a fine structure of intensity peaks, shown in Fig. a, smaller particles create a much more pronounced pattern with larger distances between the individual intensity peaks, shown in Fig.-b”
My suggestion is to think carefully about the purpose of this manuscript. If you developed something new (measurement or processing technique) and/or very useful (e.g., open source online database) for the community, you should probably focus more on the description of your development. If your work is a review of the challenges associated with 3D particle analysis then my advise would be to make clearer and more quantitative statements instead of current vague ones. Also, in the second case, examples with some real applications with quantitative assessment of positive impact (on some process/study) of modern 3D particle analysis versus 2D analysis will add a lot of sense in the paper and support your final conclusion.
Thank you for this statement. In addition to the previously mentioned changes we extended the passage describing the PARROT database. Balancing between too less and too much (known) information is (indeed) quite challenging within this special format of the Topical Collection based on the interconnected work of the last years. So, we totally agree with you on that, that the presented work is a mixture of known methods, the adaption of these methods and some use cases showing the results but with the goal to motivate a lot of other studies in this SPP 2045 framework, where the central project (as part of the SPP 20245) layed the foundation regarding characterization and the understanding of particle-discrete “working”, which is logical and seems sometimes trivial but is unfortunately not a common basis (knowledge) within our community although, e.g., the groups of Withers, Lin, Miller etc. did a great work on that.
For the PARROT passage we agree, there was much too less of information. We changed and added info as follows:
“In accordance with the FAIR principle~\cite{Wilkinson2016}, which is designed to ensure that scientific data is structured and stored in a manner that facilitates its findability, accessibility, interoperability, and reusability, we initiated a feasibility study to generate a particle database called \textit{PARROT}~\cite{Ditscherlein2022a}. The database employs a straightforward web interface, allowing users to filter particle-discrete data based on a range of parameters before downloading datasets in different levels of aggregation. The highest level contains the reconstructed tomographic data as a stack of tagged image file format (TIFF) images, accompanied by a segmented image stack. Both datasets are securely stored within an existing repository,\textit{OpARA}~\cite{OpARA}, ensuring long--term availability via DOI references.
A key feature of the PARROT database is its comprehensive metadata framework, which plays a crucial role in describing the context of each dataset. Metadata includes details about the sample preparation method, measurement conditions, and image processing parameters, ensuring that users can accurately interpret the data and integrate it into their research. Automated metadata extraction is implemented to improve dataset consistency and minimize manual errors.
Beyond data storage, the database architecture follows a relational database model, which efficiently organizes particle--discrete data and metadata while reducing redundancy. This enables fast queries for dataset retrieval, making the system scalable for future expansions. The integration with OpARA further enhances data accessibility, allowing researchers to cross-reference datasets using DOIs in scientific publications.
To demonstrate the value of PARROT, three practical use cases are presented in~\cite{Ditscherlein2022a}, showcasing its applications in 3D particle analysis, statistical and multivariate parametric modeling, and numerical process simulations. The latter leverages lattice Boltzmann methods to simulate fluid--particle interactions, providing critical insights into filtration and transport processes.”
We also moved the visualization of 6 particles from the particle database to this section to show some results and to make it clearer where these particles come from.
Reviewer 3 Report
Comments and Suggestions for AuthorsThe manuscript provides a technical summary of studies focusing on the application of X-ray tomography in advancing the multidimensional characterization of particulate systems. It offers valuable insights into enhancing the quality of 3D particle-discrete data and addresses the limitations of traditional particle characterization methods. The introduction of innovative sample preparation methods and the development of a particle database aligned with FAIR data principles are commendable.
However, as an article intended for publication, it would benefit from a clearer statement regarding the novelty and specific contributions of the work. While the manuscript outlines several advancements, it lacks explicit articulation of how these developments push the boundaries of existing research or provide new insights. Additionally, the results should be better supported with evidence, and a more detailed discussion of their implications would strengthen the manuscript’s impact.
Regarding the particle database, the manuscript asserts its alignment with the FAIR principles, but fails to convincingly demonstrate its compliance. There is limited information about key aspects such as the metadata structure, access protocols, and how interoperability with existing databases is achieved. A more thorough explanation and validation of these points would help address this issue.
Author Response
The manuscript provides a technical summary of studies focusing on the application of X-ray tomography in advancing the multidimensional characterization of particulate systems. It offers valuable insights into enhancing the quality of 3D particle-discrete data and addresses the limitations of traditional particle characterization methods. The introduction of innovative sample preparation methods and the development of a particle database aligned with FAIR data principles are commendable.
However, as an article intended for publication, it would benefit from a clearer statement regarding the novelty and specific contributions of the work. While the manuscript outlines several advancements, it lacks explicit articulation of how these developments push the boundaries of existing research or provide new insights. Additionally, the results should be better supported with evidence, and a more detailed discussion of their implications would strengthen the manuscript’s impact.
We totally agree on that. The thing is that within our community of particle characterization referring to process analysis and development we are not using (or too seldomly) these mentioned tools. The last section should serve as a summary but with no quantitative statement maybe not needed. Instead, we shortened the passage and add it as a final paragraph without extra section.
“Although the theoretical principles of the 3D characterization of particles have been discussed in detail in the past and the advantages of this method over 2D imaging and the description with equivalent sizes are obvious, it is still too rarely used in the particle characterization community. This needs to change. Relationships of particle properties are rarely one--dimensional, and when they are, they often lack precision. Particle--discrete multidimensional analysis offers a powerful tool to gain new insights in on processes and parameter correlations.”
As an additional clarification, that we are aware of that fact, but, nonetheless, want to go that way, we changed the introduction and hope this will clarify the importance of the topic and the relation to the Topical Collection of the journal Powders.
“This publication of the Topical Collection \textit{Multidimensional Particle Properties: Characterization, Description, Separation} is part of the DFG--funded \textit{Priority Programme SPP~2045} and summarizes the work on three--dimensional particle characterization within the central characterization subproject. The focus is on the particle--discrete description of the 3D particle data, which is not fundamentally new within the computed tomography and image analysis community, but in our opinion is underrepresented in the process--related particle characterization community and requires further motivation. In the following, we therefore focus on the motivation behind the topic and show how this data can serve as the basis for multidimensional property correlations. In the individual sections, we focus on how the quality of the resulting image data is affected and explore ways to maximize it in order to create reliable datasets suitable for multidimensional particle analysis. Application examples are briefly outlined and further considered, but are described in detail in the project-related publications.”
Regarding the particle database, the manuscript asserts its alignment with the FAIR principles, but fails to convincingly demonstrate its compliance. There is limited information about key aspects such as the metadata structure, access protocols, and how interoperability with existing databases is achieved. A more thorough explanation and validation of these points would help address this issue.
Thank you for this comment! For the PARROT passage we agree, there was much too less of information. We changed and added info as follows:
“In accordance with the FAIR principle~\cite{Wilkinson2016}, which is designed to ensure that scientific data is structured and stored in a manner that facilitates its findability, accessibility, interoperability, and reusability, we initiated a feasibility study to generate a particle database called \textit{PARROT}~\cite{Ditscherlein2022a}. The database employs a straightforward web interface, allowing users to filter particle-discrete data based on a range of parameters before downloading datasets in different levels of aggregation. The highest level contains the reconstructed tomographic data as a stack of tagged image file format (TIFF) images, accompanied by a segmented image stack. Both datasets are securely stored within an existing repository,\textit{OpARA}~\cite{OpARA}, ensuring long--term availability via DOI references.
A key feature of the PARROT database is its comprehensive metadata framework, which plays a crucial role in describing the context of each dataset. Metadata includes details about the sample preparation method, measurement conditions, and image processing parameters, ensuring that users can accurately interpret the data and integrate it into their research. Automated metadata extraction is implemented to improve dataset consistency and minimize manual errors.
Beyond data storage, the database architecture follows a relational database model, which efficiently organizes particle--discrete data and metadata while reducing redundancy. This enables fast queries for dataset retrieval, making the system scalable for future expansions. The integration with OpARA further enhances data accessibility, allowing researchers to cross-reference datasets using DOIs in scientific publications.
To demonstrate the value of PARROT, three practical use cases are presented in~\cite{Ditscherlein2022a}, showcasing its applications in 3D particle analysis, statistical and multivariate parametric modeling, and numerical process simulations. The latter leverages lattice Boltzmann methods to simulate fluid--particle interactions, providing critical insights into filtration and transport processes.”
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsDear Authors
Thank you for making revisions to the paper and providing your comments. Please see my additional comments below.
Lines 20-23 Introduction
"The focus is on the particle–discrete description of the 3D particle data, which is not fundamentally new within the computed tomography and image analysis community, but in our opinion is underrepresented in the process–related particle characterization community and requires further motivation."
Rather than use the phrase “in our opinion” in a technical paper it would be more informative for the reader to see the phrase “the data show…” and then to see data in the paper that demonstrates the under-representation that you are observing.
Line 138
The reference to “MLA" should be replaced by the more general “automated SEM-based systems”.
Also, please include a reference to optical microscopy here as an alternative method to analyse polished sections.
Lines 142-143: "To address stereological bias, one technique involves re–cutting and rotating of already embedded particulate samples to minimize particle size–dependent segregation effects [11]".
Please provide more detail in the text for the reader on how this sample preparation method addresses stereological bias.
The reference to the MLA Operating manual is as follows:
MLA System User Operating Manual Module 1, Version 2.0b published September 2003, Julius Kruttschnitt Mineral Research Centre, University of Queensland, Brisbane, Australia.pp 22.
The method described in the Manual which involves mounting, vertical sectioning and re-mounting particles is often referred to by practitioners as “transverse mounting”. A review of the literature shows the following publications that refer to this technique which predate your existing reference. Please replace the existing reference with one or more of these earlier references. The objective of using this method is to minimise the bias introduced by non-random presentation of the particles due to segregation by size, shape or density.
Shaffer, M. Sample preparation methods for image analysis. In Proceedings of the Geometallurgy and Applied Mineralogy, Conference of Mineralogists, Sudbury, Canada, 22 August 2009
Kwitko-Ribeiro, R. New sample preparation developments to minimize mineral segregation in process mineralogy. In Proceedings of the 10th International Congress for Applied Mineralogy (ICAM), Trondheim, Norway, 1–5 August 2011; Broekmans, M., Ed., 2012
Blaskovich, R.J. Characterizing Waste Rock Using Automated Quantitative Electron Microscopy. MAppSci Thesis, University of British Columbia, Vancouver, BC, Canada, 2013
Grant, D.C.; Goudie, D.J.; Shaffer, M.; Sylvester, P. A single-step trans-vertical epoxy preparation method for maximising throughput of iron-ore samples via SEM-MLA analysis. Appl. Earth Sci. Trans. IMM B 2016, 125, 57–62 (submitted to journal 9 June 2015).
Section 2.2 Main Factors Influencing Quantitative Particle Analysis
There is a feature of X-ray tomographic analyses which needs to be highlighted in this section - the fact that materials which have similar X-ray attenuation cannot be distinguished in the XCT images. This is a disadvantage when applying XCT to analyse particles in mineral processes, where many minerals of interest can have similar attenuations. I recommend including this information.
Line 216-217
"we will focus on a micro–CT system equipped with an additional microscope optic to achieve high voxel resolution down to 0.3 µm"
Please provide details of the system which is being used to generate the data referred to here. The journal website states that “Full experimental details must be provided so that results can be reproduced”. As a reader the minimum essential information I would expect to be given in a paper of this type is the manufacturer, model name/number and a brief description with the key details of the measurement system being used. If including the information in the body of the paper would be difficult please supply the information as a table in an Appendix and refer to this table within the text of the paper.
Line 257
“the requirements of 3D tomographic characterization [43,44].”
In the spirit of minimizing the amount of self-citation and recognising the contributions of others to this field please can you provide alternative references from other researchers in place of 43 and 44 in this context?
Line 497:
"Previously implemented for 2D mineralogical analysis [9,11,17]".
Earlier comments:
The use of carbon particles has been a standard feature of sample preparation for SEM-based automated mineralogy systems (e.g. QEMSCAN, MLA) since the 1980s and 1990s respectively. This pre-dates the two referenced publications by TU Bergakademie Freiberg researchers and it would be appropriate to acknowledge the earlier work.
We also add Sutherland and Gottlieb here. Because Sutherland and Gottlieb are not mentioning that there could be segregation effects, which our colleagues from Helmholtz in Freiberg reported on, we still think that keeping a validating and an application source is still good.
Please see the four references provided above which do refer to segregation – these predate your reference 11 and can be used to replace it. There is an alternative reference to the Sutherland/Gottlieb paper (accessible online), written by the developers of QEM*SEM which does reference segregation. This states: "It is desirable to avoid having the particles touch in polished section, and it is necessary to avoid segregation on the basis of mass, density or size. For these purposes, particles were mechanically diluted and thoroughly mixed with 10:1 by volume of crushed graphite of the same size range, and just sufficient epoxy resin was added to make the mixture fluid. During setting of the epoxy the ore particles were mechanically supported by the graphite particles, and maintained in a random three-dimensional distribution."
Reid, A.F., Gottlieb, P., MacDonald,K.J. and Miller, P.R., 1985, QEM*SEM image analysis of ore minerals : volume fraction, liberation, and observational variances, Proceedings of ICAM '84, 22-25 February 1984, Los Angeles, USA.
Line 586-587
"it is still too rarely used in the particle characterization community."
That is an interesting comment which seems to be a key message in the paper. Having read about the benefits of 3D characterization in the paper, this comment in the final paragraph prompts the question "why is it rarely used?" in the mind of the reader and it would be very useful to the reader to provide a discussion on the possible reasons why 3D characterization of particles is “too rarely used”.
Author Response
Dear Reviewer
thank you for your time! Please go to the attached word-file having the details of our response.
Many greetings
the authors
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsImplemented corrections improve overall quality of the paper. In my opinion, the paper in as-is state can be interesting and useful for those who are starting their studies in the field of image-based analysis of particles. The described PARROT database could be an important element for such studies.
Author Response
Dear Reviewer,
thanks again for your time to review our manuscript text in that detail.
many greetings
the authors
Reviewer 3 Report
Comments and Suggestions for AuthorsPrevious comments have been adequately addressed.
Author Response
Dear Reviewer,
thanks again for your time to review our manuscript in that detail!
many greetings
the authors
