Vacuum-Free and Highly Dense Nanoparticle Based Low-Band-Gap CuInSe2 Thin-Films Manufactured by Face-to-Face Annealing with Application of Uniaxial Mechanical Pressure
Round 1
Reviewer 1 Report
Manuscript by Schuster et al. describes vacuum-free deposition methods to deposit CuInSe2 film, which involves doctor blade coating of the precursor ink followed by Face-to-face annealing. The introduction is weak, and the results are not conclusive. The experiments are not planned systematically, which makes me think that these results are just preliminary, and a significant amount of work needs to be done to make the manuscript strong. Detailed comments are as follows:
1. The introduction needs to be improved by being more specific and quantitative. For example – the authors mention that CIGS is mainly deposited by processes using vacuum technology and toxic substances. Please mention what vacuum technologies and which toxic substances. Additionally, doctor balding of CIGS has been reported before. So, please mention the novelty of this work clearly.
2. CIGS is abbreviated for copper indium gallium diselenide in the abstract (page1, line 11), whereas it is abbreviated for copper indium gallium sulfo-selenide in the introduction (page 1, line 32). Please check and be consistent. Also, be consistent in capitalizing the first letter of the words in the name.
3. Furthermore, the introduction talks about copper indium gallium-CIGS film (first paragraph of the introduction). However, the manuscript talks about the copper indium selenide film. What is the connection between the two? Please add that in the manuscript and modify the introduction.
4. Page 2, Line 60, please correct “Cu-In-NP” to “Cu-In NP” otherwise it looks like NP is some element.
5. Page 2, line 80, sentence “Both of these species…”, which species?
6. Page 3, line 96, “height of 200 μm..” what is that height of?
7. In route I(b), the seeding layer is of Cu-In or Cu-In-Se. Under the “processing of absorber layer section on page 3, Cu-In is mentioned as seeding layer (line 97-98), whereas under results and discussion section, Cu-In-Se is mentioned as seeding layer (page 4, line 150).
8. In the coating methods (both route I and II), are the two inks mixed together (Cu-In and Se in the route I, and CuSe and InSe in route II) and then coated or they are coated separately?
9. Regarding the post-deposition processing of the film, was the pressure applied after annealing at high temperature or during annealing?
10. Please label the XRD peaks in Figure 2 and 4 with the material name and crystal plane. The peak list given is not helpful, it is not clearly visible, and the x-axis is not provided. Moreover, the x-axis scale in Figure 2 and 4 are different but seems like axis in peak list figures (below the XRD figure) are same in Figure 2 and 4, which mislead the peak identification.
11. Looking at the cross-section SEM images of the film (Figure 3, 5, and 7), the films look highly non-uniform and the thickness varies a lot. How is the thickness measured?
12. Authors mention that CIS layers are dense. I would recommend adding porosity measurements of the film (BET surface area and porosity measurement). The films look like they are composed of nanoparticles and may have definitive porosity. This will also provide a quantitative assessment on which deposition route (I(a), I(b), or II) is better.
13. Why is the thickness of the film deposited by the route I(b) is different than that of the film deposited by the route I(a)?
14. Authors mention the films to be one element rich. Elemental analysis (SEM-EDS) is necessary to validate that. Please add SEM-EDS and report elemental analysis results in the manuscript or appendix.
15. Please add top-down SEM of the film deposited by the route I(b) in Figure 5.
16. UV-vis absorption spectrum does not provide an accurate measurement of the bandgap. If possible, I would recommend doing steady state photoluminescence measurement to determine the band-gap.
17. Considering the XRD pattern, FWHM of the CIS peak and the band gap calculated by the UV-vis absorption spectrum, the results of the films deposited by the three methods are similar. Additionally, all three methods provide a rough film. How do you conclude which method is the best? Also, these rough films are not good to be used in solar cells.
Author Response
Response to Reviewer 1 Comments
Manuscript by Schuster et al. describes vacuum-free deposition methods to deposit CuInSe2 film, which involves doctor blade coating of the precursor ink followed by Face-to-face annealing. The introduction is weak, and the results are not conclusive. The experiments are not planned systematically, which makes me think that these results are just preliminary, and a significant amount of work needs to be done to make the manuscript strong. Detailed comments are as follows:
Thank you very much for all the detailed comments and suggestions to improve our manuscript. We are always glad to get the chance to improve our research. Please find our answers to every comment below.
1. The introduction needs to be improved by being more specific and quantitative. For example – the authors mention that CIGS is mainly deposited by processes using vacuum technology and toxic substances. Please mention what vacuum technologies and which toxic substances. Additionally, doctor balding of CIGS has been reported before. So, please mention the novelty of this work clearly.
We made changes to the introduction to specify the vacuum-methods and toxic substances. Moreover we pointed out more clearly the aim of this study, i.e. to use vacuum-free techniques and environmentally non-harmful substances as well as use combination of liquid-phase- and pressure-sintering by application of mechanical pressure during the annealing.
2. CIGS is abbreviated for copper indium gallium diselenide in the abstract (page1, line 11), whereas it is abbreviated for copper indium gallium sulfo-selenide in the introduction (page 1, line 32). Please check and be consistent. Also, be consistent in capitalizing the first letter of the words in the name.
We made the changes to be consistent in wording and used only lower case letters for the elements.
3. Furthermore, the introduction talks about copper indium gallium-CIGS film (first paragraph of the introduction). However, the manuscript talks about the copper indium selenide film. What is the connection between the two? Please add that in the manuscript and modify the introduction.
We modified the introduction to explain why we only investigated CISe, i.e. to keep it simple for better control of composition.
4. Page 2, Line 60, please correct “Cu-In-NP” to “Cu-In NP” otherwise it looks like NP is some element.
We removed the dash to avoid confusing NP (nanoparticles) with nitrogen and phosphor. We changed this in all instances in the manuscript.
5. Page 2, line 80, sentence “Both of these species…”, which species?
We changed the wording to better explain that Na2Se as well as Na2Sex form Cu2-xSe in chemical reactions.
6. Page 3, line 96, “height of 200 μm..” what is that height of?
We changed the wording to “blade to substrate distance”.
7. In route I(b), the seeding layer is of Cu-In or Cu-In-Se. Under the “processing of absorber layer section on page 3, Cu-In is mentioned as seeding layer (line 97-98), whereas under results and discussion section, Cu-In-Se is mentioned as seeding layer (page 4, line 150).
We changed the wording to clarify in the text that sputtered Cu and In is the seeding-layer, that first forms a dense CISe film, then grows from nanoparticle supply.
8. In the coating methods (both route I and II), are the two inks mixed together (Cu-In and Se in the route I, and CuSe and InSe in route II) and then coated or they are coated separately?
In most cases the nanoparticle inks were pre mixed. We added this information to the “coating methods”.
9. Regarding the post-deposition processing of the film, was the pressure applied after annealing at high temperature or during annealing?
The pressure was applied during the annealing procedure. We changed some wording in the text to make it more clear.
10. Please label the XRD peaks in Figure 2 and 4 with the material name and crystal plane. The peak list given is not helpful, it is not clearly visible, and the x-axis is not provided. Moreover, the x-axis scale in Figure 2 and 4 are different but seems like axis in peak list figures (below the XRD figure) are same in Figure 2 and 4, which mislead the peak identification.
We labelled the XRD reflections in figures 2, 4 and 6 to make the figures more clear. We also removed the reference patterns to avoid confusion, although the x-axes were the same scale as the respective images.
11. Looking at the cross-section SEM images of the film (Figure 3, 5, and 7), the films look highly non-uniform and the thickness varies a lot. How is the thickness measured?
The film thickness was measured from the cross-section images via software on a large scale of the sample. We added this information to the characterization methods.
12. Authors mention that CIS layers are dense. I would recommend adding porosity measurements of the film (BET surface area and porosity measurement). The films look like they are composed of nanoparticles and may have definitive porosity. This will also provide a quantitative assessment on which deposition route (I(a), I(b), or II) is better.
Thank you for your suggestion of method here. It was not possible to carry out these measurements for this revision processes, but we will keep it in mind for our future research.
13. Why is the thickness of the film deposited by the route I(b) is different than that of the film deposited by the route I(a)?
The thickness of both films should be (and is) comparable with ca. 3 µm in Route Ia and ca. 2.5 µm in Route Ib. We noticed a typo where we accidentally wrote “1 µm” instead of 2.5 µm and changed the text accordingly. In the figures itself the scale was correct.
14. Authors mention the films to be one element rich. Elemental analysis (SEM-EDS) is necessary to validate that. Please add SEM-EDS and report elemental analysis results in the manuscript or appendix.
We added the atomic percentage of elemental composition that was performed with EDX to the text in the respective paragraphs.
15. Please add top-down SEM of the film deposited by the route I(b) in Figure 5.
We added the top-down image in figure 5a and changed the cross-section images to 5b and 5c.
16. UV-vis absorption spectrum does not provide an accurate measurement of the bandgap. If possible, I would recommend doing steady state photoluminescence measurement to determine the band-gap.
Thank you for your suggestion of method here. Indeed we had some issues with our PL set-up and the measurements during the data acquisition for this manuscript. We had to use the methods that were working, i.e. UV-VIS to estimate the band-gap. When the PL set-up is running again, we will perform PL measurements, however it was not possible for this revision process.
17. Considering the XRD pattern, FWHM of the CIS peak and the band gap calculated by the UV-vis absorption spectrum, the results of the films deposited by the three methods are similar. Additionally, all three methods provide a rough film. How do you conclude which method is the best? Also, these rough films are not good to be used in solar cells.
Indeed all three methods’ results are similar, but have different (dis-) advantages. Route Ia has holes, that Ib does not have, and Route II had the worst surface coverage. We would conclude that Route Ib is best, however it was not completely vacuum-free. We added a sentence that the roughness/smoothness of the films has to be improved for actual application in solar cells.
Reviewer 2 Report
Wellmann et al., report vacuum free nanoparticle low-band-gap CuInSe2 thin films manufacturing processes. The manuscript is well prepared, and it was a pleasure to read it. I’m of course pleased to accept this nice paper for publication in Coatings. This document shows clarity in every single aspect from the preparation to its characterization. I have noted some small clarifications/suggestions within the attached file.
Comments for author File: 
Comments.pdf
Author Response
Response to Reviewer 2 Comments
Point 1:
Wellmann et al., report vacuum free nanoparticle low-band-gap CuInSe2 thin films manufacturing processes. The manuscript is well prepared, and it was a pleasure to read it. I’m of course pleased to accept this nice paper for publication in Coatings. This document shows clarity in every single aspect from the preparation to its characterization. I have noted some small clarifications/suggestions within the attached file.
Thank you for your kind words and for your comments in the attached file. We gladly made all the small changes you suggested to improve the manuscript.
Reviewer 3 Report
Manuscript Number: coatings-515479
The manuscript entitled “Vacuum-Free and Highly Dense Nanoparticle Based Low-Band-Gap CuInSe2 Thin-Films Manufactured by Face-to-Face Annealing with Application of Uniaxial Mechanical Pressure” by Schuster et al. describes the wet chemical deposition of CISe thin films using facile deposition method. Overall, the manuscript is well written and interesting to the broad readership. However, I do have the following concern prior to accept this manuscript for publication. Therefore, Major revision is mandatory.
1. Authors should confirm the formation of CuInSe compound using other methods like XPS, which is a key tool.
2. Fig.3a and 3b SEM images are not so clear. Why? Avoid charging issue while scanning the image
3. Author should map the SEM images with EDS mapping and the corresponding quantitative as well as qualitative description should be provided
4. It would be great if the authors should show the real-time performance of the solarcell using this thinfilm. Otherwise the application of this work is questionable.
Author Response
Response to Reviewer 3 Comments
The manuscript entitled “Vacuum-Free and Highly Dense Nanoparticle Based Low-Band-Gap CuInSe2 Thin-Films Manufactured by Face-to-Face Annealing with Application of Uniaxial Mechanical Pressure” by Schuster et al. describes the wet chemical deposition of CISe thin films using facile deposition method. Overall, the manuscript is well written and interesting to the broad readership. However, I do have the following concern prior to accept this manuscript for publication. Therefore, Major revision is mandatory.
1. Authors should confirm the formation of CuInSe compound using other methods like XPS, which is a key tool.
We unfortunately do not have access to XPS for characterization of our films. For the future we may find a cooperation partner with this method, but for this revision we could not do these measurements. Nevertheless we are confident that our SEM/EDS and XRD characterization, which are also key tools in CIGS, arequalified to confirm the CISe formation.
2. Fig.3a and 3b SEM images are not so clear. Why? Avoid charging issue while scanning the image
We do not see unclear images in Fig.3 a,b. Maybe it was an issue in uploading? We will also upload a pdf version to check if all images can be clearly seen.
3. Author should map the SEM images with EDS mapping and the corresponding quantitative as well as qualitative description should be provided
We measured several EDS mappings with quantitative element distribution and added them to the manuscript.
4. It would be great if the authors should show the real-time performance of the solarcell using this thinfilm. Otherwise the application of this work is questionable.
Unfortunately fabrication of complete solar cells is not possible at the moment, but is definitely planned for the future. However in the “coatings” journal the focus of the manuscript can be on the thin film and the methods.
Reviewer 4 Report
Presented work is written in good language. All results are clearly presented and correctly interpreted. Conclusions are correctly formulated and supported by the relevant discussion. In general, the work is presented at a very good level. However, there are some minor limitations that generally do not prevent the publication of the work.
1. The authors offer some ways to solve problems voiced in the work. In this connection, there is some understatement and incompleteness of work. The work would be more complete if it were done.
2. In general, the not very high quality of the obtained films reduces the practical significance of the work. However, the fact that the fundamental possibility to obtainine such films by the new method is shown is a rather important result.
Author Response
Response to Reviewer 4 Comments
Presented work is written in good language. All results are clearly presented and correctly interpreted. Conclusions are correctly formulated and supported by the relevant discussion. In general, the work is presented at a very good level. However, there are some minor limitations that generally do not prevent the publication of the work.
1. The authors offer some ways to solve problems voiced in the work. In this connection, there is some understatement and incompleteness of work. The work would be more complete if it were done.
2. In general, the not very high quality of the obtained films reduces the practical significance of the work. However, the fact that the fundamental possibility to obtainine such films by the new method is shown is a rather important result.
Thank you for your kind words above. We hope to continue our research and complete it with all the ideas that we still have to solve the issues.
Reviewer 5 Report
In the manuscript "Vacuum-Free and Highly Dense Nanoparticle Based Low-Band-Gap CuInSe2 Thin-Films Manufactured by Face-to-Face Annealing with Application of Uniaxial Mechanical Pressure" the authors propose an original method of deposition of CuInSe2 material as an absorber layer for solar cells. The particular interest of this method is that it avoids using vacuum and toxic selenization processes, which often decrease the potential of even low-toxicity materials such as CuInSe2. By preparing the nanoparticles of copper and indium selenides by chemical methods as well as the bimetallic Cu-In ones the authors manage to optimise the morphology of the absorber layer while the pressure annealing limits the loss of Se and improves the film properties further. The research work is performed on a good experimental level. The results of this paper could be useful for the scientific community as they propose an alternative low toxicity fabrication method of CuInSe2 films.
I would expect all the layers fabricated to be tested in the solar cells to validate the approach and compare different deposition techniques in the real devices. At the same time, for the scope of the Coatings journal it is possible that the final devices are not necessary. If this is indeed the case, I recommend the paper for the publication provided that some points are addressed.
1. p. 5, l. 159: One cannot call the obtained composition as "copper-poor" as the percentages of Cu and In are almost identical. What is the experimental error of the EDX measurements? Have several points been used to obtain an average material atomic composition? Are those measurements reproducible? The same remarks are valid for other EDX measurements.
2.line 189: It is incorrect to call the film "with lower copper deficit" - this one has clearly an excess of copper.
3.L. 268: Why ethanol has better wetting behaviour compared to the IPA? Some explanation should be provided.
4. The last sentence (275) cannot be left as is: it is relatively easy to verify the hypothesis of the better film quality when a combination of the processing techniques is used. Why the authors haven't checked if these conditions result in better films?
5. Coming back to the previous point - the paper would gain a lot in scientific interest if the authors could provide the tests of the films in real solar cells.
Author Response
Response to Reviewer 5 Comments
In the manuscript "Vacuum-Free and Highly Dense Nanoparticle Based Low-Band-Gap CuInSe2 Thin-Films Manufactured by Face-to-Face Annealing with Application of Uniaxial Mechanical Pressure" the authors propose an original method of deposition of CuInSe2 material as an absorber layer for solar cells. The particular interest of this method is that it avoids using vacuum and toxic selenization processes, which often decrease the potential of even low-toxicity materials such as CuInSe2. By preparing the nanoparticles of copper and indium selenides by chemical methods as well as the bimetallic Cu-In ones the authors manage to optimise the morphology of the absorber layer while the pressure annealing limits the loss of Se and improves the film properties further. The research work is performed on a good experimental level. The results of this paper could be useful for the scientific community as they propose an alternative low toxicity fabrication method of CuInSe2 films.
I would expect all the layers fabricated to be tested in the solar cells to validate the approach and compare different deposition techniques in the real devices. At the same time, for the scope of the Coatings journal it is possible that the final devices are not necessary. If this is indeed the case, I recommend the paper for the publication provided that some points are addressed.
1. p. 5, l. 159: One cannot call the obtained composition as "copper-poor" as the percentages of Cu and In are almost identical. What is the experimental error of the EDX measurements? Have several points been used to obtain an average material atomic composition? Are those measurements reproducible? The same remarks are valid for other EDX measurements.
The percentages of Cu and In are indeed almost identical, but this is expected with Cu/In ratios of 0.88 -0.92. CISe films with a ratio smaller than 1 are considered copper poor by a lot of groups from the community, see for example the paper from Siebentritt group (Reference 11: Prog. Photovoltaics Res. Appl. 26 (2018) 437–442. doi:10.1002/pip.3026). The Cu and In signal can be distinguished well enough in EDX to say the ratio is true, even if the absolute numbers might be off due to amount of oxygen etc.
For our EDX in general we always measure areas of 5x5 µm² to 25x25 µm², so the numbers are always averaged. The measurements are reproducible, e.g. when we measure several spots “far away” from each other but on the same sample.
2.line 189: It is incorrect to call the film "with lower copper deficit" - this one has clearly an excess of copper.
We rephrased that sentence. We assumed the film to be copper poor, because of the measured band gap. The sample was measured again with EDX still showing excess of copper, but a lower amount. The film however has Cu2Se left in it, so this can explain some of the excess copper.
3.L. 268: Why ethanol has better wetting behaviour compared to the IPA? Some explanation should be provided.
Ethanol has a significantly lower viscosity, a slightly lower boiling temperature and a slightly lower surface tension compared to IPA. This all contributes to the better wetting behaviour. We added this explanation to the results and discussion sections.
4. The last sentence (275) cannot be left as is: it is relatively easy to verify the hypothesis of the better film quality when a combination of the processing techniques is used. Why the authors haven't checked if these conditions result in better films?
We removed this sentence, as in the short time we can not process the sample series with combined techniques without significantly delaying the publication of this manuscript. We would prefer to follow up this publication with the newest results, when everything is refined.
5. Coming back to the previous point - the paper would gain a lot in scientific interest if the authors could provide the tests of the films in real solar cells.
Unfortunately fabrication of complete solar cells is not possible at the moment, but is definitely planned for the future. However in the “coatings” journal the focus of the manuscript can be on the thin film and the methods.
Round 2
Reviewer 1 Report
1. The introduction talks about CIGS films and problems associated with its fabrication method. However, the work is on CIS film. The introduction needs to be focused on CIS film then. I understand that authors fabricated CIS film to keep the work simple but then they need to change the introduction on CIS.
2. The film resulting from all the three deposition routes have excess Se. Authors should describe a method to remove excess Se from the film.
3. Authors mentioned in the manuscript that the films are very uniform. However, looking at SEM images, all the films deposited by the three methods are highly non-uniform, and therefore, won’t be a good layer for solar cells.
4. Moreover, all the films are porous and have the same bandgap. So, the study is not conclusive which route is better to adopt. Authors said that the route ‘Ib’ produces the best film. However, route ‘Ib’ uses vacuum technology and selenization of the film, which is opposite to the aim of this work according to the introduction in the manuscript.
Author Response
Response to Reviewer 1 round 2 Comments
1. The introduction talks about CIGS films and problems associated with its fabrication method. However, the work is on CIS film. The introduction needs to be focused on CIS film then. I understand that authors fabricated CIS film to keep the work simple but then they need to change the introduction on CIS.
We modified the introduction a little bit more and added mode information and an application on CISe. However CIGS is by far more relevant and is the long term goal, so we did not remove all the CIGS parts from the introduction.
2. The film resulting from all the three deposition routes have excess Se. Authors should describe a method to remove excess Se from the film.
Excess Se and selenides can be removed by KCN or ammonium sulfide. We added this information to the discussion part.
3. Authors mentioned in the manuscript that the films are very uniform. However, looking at SEM images, all the films deposited by the three methods are highly non-uniform, and therefore, won’t be a good layer for solar cells.
We added to the discussion and the conclusions, that the film quality needs to be improved for solar cells.
4. Moreover, all the films are porous and have the same bandgap. So, the study is not conclusive which route is better to adopt. Authors said that the route ‘Ib’ produces the best film. However, route ‘Ib’ uses vacuum technology and selenization of the film, which is opposite to the aim of this work according to the introduction in the manuscript.
Even in the first revision we wrote, that the sputtered seeding layer was to test how a seeding layer in general works. However the film was not “selenized” as of the commonly known selenization process in Se atmosphere. We decided to keep it in the manuscript even though it is not completely vacuum-free, because we wanted to share how good it worked even without the usual selenziation, and wrote about application of a vacuum-free seeding layer as one of the next steps. These processes however can not be tested in this short time and without significantly delaying this publication.
Reviewer 3 Report
Authors responded to my comments. Manuscript in its present form may consider for publication
Round 3
Reviewer 1 Report
The manuscript looks improved now and it can be accepted. Also, authors should submit the manuscript in correct form without any comments and track changes left in the submitted draft, and only the changes should be highlighted.
