Study of Sputtered ITO Films on Flexible Invar Metal Foils for Curved Perovskite Solar Cells
Round 1
Reviewer 1 Report
The manuscript titled "Study of sputtered ITO films on flexible invar metal foils for curved perovskite solar cells" has tried to present the optical and mechanical properties of ITO deposited on invar/SiO2 substrate. The study looks good but some parts of the work should be revised before accepted by the excellent journal. The proposed amendments are listed as follows:
1. What does the reflection curve of bare invar (without SiO2/ITO layers deposited) looks like?
2. The title of the manuscript is "Study of sputtered ITO films on flexible invar metal
foils for curved PEROVSKITE SOLAR CELLS", however, authors have not shown any actual device data for solar cells. Authors should provide the performance of perovskite solar cells fabricated on bare invar and with different thickness of ITO deposited over Invar/SiO2 substrate. Without these data, I think the title of the paper itself is misleading and should not be published without changing it.
Author Response
Reviewer 1
1. What does the reflection curve of bare invar (without SiO2/ITO layers deposited) looks like?
Answer) As reviewer suggested, we measured reflectance of bare invar metal foil substrate like below. Due to absence of SiO2 layer, there is no modulation of reflectance unlike ITO/SiO2/invar samples in Figure 3a.
Figure 1. Reflectance of the bare invar substrate without SiO2 and ITO layer.
2. The title of the manuscript is "Study of sputtered ITO films on flexible invar metal foils for curved PEROVSKITE SOLAR CELLS", however, authors have not shown any actual device data for solar cells. Authors should provide the performance of perovskite solar cells fabricated on bare invar and with different thickness of ITO deposited over Invar/SiO2 substrate. Without these data, I think the title of the paper itself is misleading and should not be published without changing it.
Answer) As reviewer suggested, we added the power conversion efficiency of curved perovskite solar cell fabricated on ITO/SiO2/invar substrate. Although the PCE of curved perovskite solar cell on ITO/SiO2/invar substrate is not optimized at this moment, we’d like to show the feasibility of invar substrate for curved perovskite solar cells.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Authors have performed a good job in the characterization and presentation of the results. However, in my opinion, the use of this subsrates for perovskite solar cell is not trivial and quite a lot optimization work has to be done in order to obtain competitive results. I will suggest no to mention " perovskite solar cells" in the main title as no experiments have done in this regards.
After reading the work, I have some questions:
- Do you think that annealing temperature could affect in the properties of the subtrates and therefore reduce the number of bending cycles.
- How can affect the stability of the perovskite in curve devices.
- Which metal contact do the authors propose in order to fabricate the solar devices.
Author Response
Reviewer 2
a) The topic of the manuscript is very obviously the preparation and analysis of the ITO films which could be used in perovskite-based or other thin film solar cells. However, the introduction almost exclusively deals with perovskite solar cells and metal foils as substrate materials. Not a single citation of other ITO-related work is found which would put the presented work into a context of research on this material. The authors should include some references to state-of-the-art publications on ITO-films.
Answer) As reviewer suggested, we added some references to state-of-the art publications on ITO films for perovskite solar cells.
b) In the section "materials and methods" it is not described how the thickness of the sputtered films has been determined, which is of interest when analyzing the results later on.
Answer) The thickness of the ITO films was measured using a typical profilometer (alpha step).
c) The results presented in Fig 2a and b clearly show the good reproducibility the authors have achieved both with the ITO-deposition but also with the important SiO2 barrier layer. However, in spite of the good agreement between the measured sheet resistance on the barrier layer and on the glass substrates the measured values would need more discussion analysis. If anything one would expect the 50 nm thin layer to be at least twice as resistive as the 100 nm layer and 4 times as resistive as the 200 nm layer. Also comparing with values from the literature the 50 nm thin layer appears to be too conductive in comparison with the other samples within this study and films prepared elsewhere. Therefore the question: how was the thickness of the layers measured, how large is the uncertainty?
Answer) Because our group has investigated ITO-based transparent electrode materials and coating process, we have know-how to fabricate high-quality ITO films at room temperature for flexible displays and photovoltaics. The electrical properties of ITO films on invar or glass substrates are reproducible and easily fabricated by room temperature sputtering process. The thickness of ITO film and sputtering rate was confirmed by TEM and profilometer.
d) A figure of merit is a powerful tool in optimizing a system for a certain application. Often the FoM is directly derived from the implicit impact of the individual properties onto the performance of a particular system - which in this case would be the perovskite solar cell. How do the authors come up with a FoM weighing the average transparency to the 10th power, divided by the sheet resistance? The current produced by a solar cell depends linearly on the incident light intensity (or better photon flux), the detrimental voltage loss depends linearly on the sheet resistance. Perhaps I am missing something here, but I suggest the authors explain the derivation of this FoM in the paper.
Answer) As reviewer suggested, we further explained the derivation of FoM value.
e) What is actually shown in the inset in Fig 2c? Especially in the lower halves? And in the upper halves - is the fringing shown there an indication of extreme variations of film thickness over the sample area?
Answer) The lower halves photograph in Figure 2c is the ITO films on glass substrates as a function of thickness. Due to opacity of invar substrate, the optical transmittance of the ITO films was obtained from the ITO film sputtered on glass substrate at identical sputtering conditions and chamber with invar substrate. The upper halves photograph is the ITO films deposited on SiO2/invar substrate. The fringing was formed by multilayer on invar substrate such as oxide layer on invar and CVD processed SiO2 passivation layer.
f) How is the reflection at the ITO/SiO2/metal foil rear side of the final solar cell affected by the difference in refractive index between air (n=1, this study, Fig 3a) and MAPbI3 (n=2.5)?
Answer) As shown inset of Figure 6a, the reflectance at the ITO/SiO2/invar are mainly affected by NiO buffer layer. However, in this work we showed the reflectance of ITO/SiO2/invar because the reflectance from substrate could affect on the power conversion efficiency.
g) I have to apologize, but at the given contrast, resolution and quality of the FESEM images in Fig 4. I cannot follow the authors' description about the different crystallinity - but I tend to believe that this difference between the layers would be visible on the original FESEM images. The bending tests are very important for supposedly flexible films, and the manuscript shows a well planned and conducted study in this respect. The corresponding results deserve a better presentation.
Answer) As reviewer suggested, we enlarged the FESEM images in manuscript.
g) The upper parts in Fig 5, showing the test setup, are way too small to show the reader how the test was conducted.
Answer) As reviewer suggested, we changed with an enlarged photograph.
h) The curves in Fig 5b actually show nothing. For all these two curves you should choose a different y-axis scale so that at least some variation on the measurement data can be seen - there must be some variation, if nothing else then variation from stochastic errors in the measurement points.
Answer) As reviewer suggested, we have plotted it on the y-axis scale to show the difference between the two curves.
Finally a remark - while I can understand the thriving for solar cells on light-weight and flexible substrates I do not understand why one should want solar cells with a curvature, since only part of these solar cells would ever be oriented optimally towards the sun, while the unavoidable internal shadowing would be worse than for flat solar cells.
Answer) Curved solar cells are expected to be applicable to vehicles with curved wall such as automobiles, airplanes, and ships.
There are some errors in the English language which should be corrected, e.g. "photon reached at active layer" in the abstract, "stability due to low water vapor transmission rate" (it is a too high rate which is detrimental), "Through Hall measurements", "results obtain", and some other minor errors.
Answer) As reviewer suggested, we have corrected some minor errors.
Reviewer 3 Report
a) The topic of the manuscript is very obviously the preparation and analysis of the ITO films which could be used in perovskite-based or other thinfilm solar cells. However, the introduction almost exclusively deals with perovskite solar cells and metal foils as substrate materials. Not a single citation of other ITO-related work is found which would put the presented work into a context of research on this material. The authors should include some references to state-of-the-art publications on ITO-films.
b) In the section "materials and methods" it is not described how the thickness of the sputtered films has been determined, which is of interest when analyzing the results later on.
c) The results presented in Fig 2a and b clearly show the good reproducibility the authors have achieved both with the ITO-deposition but also with the important SiO2 barrier layer. However, in spite of the good agreement between the measured sheet resistance on the barrier layer and on the glass substrates the measured values would need more discussion analysis. If anything one would expect the 50 nm thin layer to be at least twice as resistive as the 100 nm layer and 4 times as resistive as the 200 nm layer. ALso comparing with values from the literature the 50 nm thin layer appears to be too conductive in comparision with the other samples within this study and films prepared elsewhere. Therefore the question: how was the thickness of the layers measured, how large is the uncertainty?
d) A figure of merit is a powerful tool in optimizing a system for a certain application. Often the FoM is directly derived from the implicit impact of the individual properties onto the performance of a particular system - which in this case would be the perovskite solar cell. How do the authors come up with a FoM weighing the average transparency to the 10th power, divided by the sheet resistance? The current produced by a solar cell depends linearly on the incident light intensity (or better photon flux), the detrimental voltage loss depends linearly on the sheet resistance. Perhaps I am missing something here, but I suggest the authors explain the derivation of this FoM in the paper.
e) What is actually shown in the inset in Fig 2c? Especially in the lower halves? And in the upper halves - is the fringing shown there an indication of extreme variations of film thickness over the sample area?
f) How is the reflection at the ITO/SiO2/metal foil rear side of the final solar cell affected by the difference in refractive index between air (n=1, this study, Fig 3a) and MAPbI3 (n=2.5)?
g) I have to apologize, but at the given contrast, resolution and quality of the FESEM images in Fig 4 I cannot follow the authors' description about the different crystallinity - but I tend to believe that this difference between the layers would be visible on the original FESEM images.
The bending tests are very important for supposedly flexible films, and the manuscript shows a well planned and conducted study in this respect. The corresponding results deserve a better presentation.
g) The upper parts in Fig 5, showing the test setup, are way too small to show the reader how the test was conducted.
h) The curves in Fig 5b actually show nothing. For all these two curves you should choose a different y-axis scale so that at least some variation on the measurement data can be seen - there must be some variation, if nothing else then variation from stochastic errors in the measurement points.
Finally a remark - while I can understand the thriving for solar cells on light-weight and flexible substrates I do not understand why one should want solar cells with a curvature, since only part of these solar cells would ever be oriented optimally towards the sun, while the unavoidable internal shadowing would be worse than for flat solar cells.
There are some errors in the English language which should be corrected, e.g. "photon reached at active layer" in the abstract, "stability due to low water vapor transmission rate" (it is a too high rate which is detrimental), "Through Hall measurements", "results obtain", and some other minor errors.
Author Response
Reviewer 2
a) The topic of the manuscript is very obviously the preparation and analysis of the ITO films which could be used in perovskite-based or other thin film solar cells. However, the introduction almost exclusively deals with perovskite solar cells and metal foils as substrate materials. Not a single citation of other ITO-related work is found which would put the presented work into a context of research on this material. The authors should include some references to state-of-the-art publications on ITO-films.
Answer) As reviewer suggested, we added some references to state-of-the art publications on ITO films for perovskite solar cells.
b) In the section "materials and methods" it is not described how the thickness of the sputtered films has been determined, which is of interest when analyzing the results later on.
Answer) The thickness of the ITO films was measured using a typical profilometer (alpha step).
c) The results presented in Fig 2a and b clearly show the good reproducibility the authors have achieved both with the ITO-deposition but also with the important SiO2 barrier layer. However, in spite of the good agreement between the measured sheet resistance on the barrier layer and on the glass substrates the measured values would need more discussion analysis. If anything one would expect the 50 nm thin layer to be at least twice as resistive as the 100 nm layer and 4 times as resistive as the 200 nm layer. Also comparing with values from the literature the 50 nm thin layer appears to be too conductive in comparison with the other samples within this study and films prepared elsewhere. Therefore the question: how was the thickness of the layers measured, how large is the uncertainty?
Answer) Because our group has investigated ITO-based transparent electrode materials and coating process, we have know-how to fabricate high-quality ITO films at room temperature for flexible displays and photovoltaics. The electrical properties of ITO films on invar or glass substrates are reproducible and easily fabricated by room temperature sputtering process. The thickness of ITO film and sputtering rate was confirmed by TEM and profilometer.
d) A figure of merit is a powerful tool in optimizing a system for a certain application. Often the FoM is directly derived from the implicit impact of the individual properties onto the performance of a particular system - which in this case would be the perovskite solar cell. How do the authors come up with a FoM weighing the average transparency to the 10th power, divided by the sheet resistance? The current produced by a solar cell depends linearly on the incident light intensity (or better photon flux), the detrimental voltage loss depends linearly on the sheet resistance. Perhaps I am missing something here, but I suggest the authors explain the derivation of this FoM in the paper.
Answer) As reviewer suggested, we further explained the derivation of FoM value.
e) What is actually shown in the inset in Fig 2c? Especially in the lower halves? And in the upper halves - is the fringing shown there an indication of extreme variations of film thickness over the sample area?
Answer) The lower halves photograph in Figure 2c is the ITO films on glass substrates as a function of thickness. Due to opacity of invar substrate, the optical transmittance of the ITO films was obtained from the ITO film sputtered on glass substrate at identical sputtering conditions and chamber with invar substrate. The upper halves photograph is the ITO films deposited on SiO2/invar substrate. The fringing was formed by multilayer on invar substrate such as oxide layer on invar and CVD processed SiO2 passivation layer.
f) How is the reflection at the ITO/SiO2/metal foil rear side of the final solar cell affected by the difference in refractive index between air (n=1, this study, Fig 3a) and MAPbI3 (n=2.5)?
Answer) As shown inset of Figure 6a, the reflectance at the ITO/SiO2/invar are mainly affected by NiO buffer layer. However, in this work we showed the reflectance of ITO/SiO2/invar because the reflectance from substrate could affect on the power conversion efficiency.
g) I have to apologize, but at the given contrast, resolution and quality of the FESEM images in Fig 4. I cannot follow the authors' description about the different crystallinity - but I tend to believe that this difference between the layers would be visible on the original FESEM images. The bending tests are very important for supposedly flexible films, and the manuscript shows a well planned and conducted study in this respect. The corresponding results deserve a better presentation.
Answer) As reviewer suggested, we enlarged the FESEM images in manuscript.
g) The upper parts in Fig 5, showing the test setup, are way too small to show the reader how the test was conducted.
Answer) As reviewer suggested, we changed with an enlarged photograph.
h) The curves in Fig 5b actually show nothing. For all these two curves you should choose a different y-axis scale so that at least some variation on the measurement data can be seen - there must be some variation, if nothing else then variation from stochastic errors in the measurement points.
Answer) As reviewer suggested, we have plotted it on the y-axis scale to show the difference between the two curves.
Finally a remark - while I can understand the thriving for solar cells on light-weight and flexible substrates I do not understand why one should want solar cells with a curvature, since only part of these solar cells would ever be oriented optimally towards the sun, while the unavoidable internal shadowing would be worse than for flat solar cells.
Answer) Curved solar cells are expected to be applicable to vehicles with curved wall such as automobiles, airplanes, and ships.
There are some errors in the English language which should be corrected, e.g. "photon reached at active layer" in the abstract, "stability due to low water vapor transmission rate" (it is a too high rate which is detrimental), "Through Hall measurements", "results obtain", and some other minor errors.
Answer) As reviewer suggested, we have corrected some minor errors.
Round 2
Reviewer 1 Report
Dear Authors,
Thank you for adding the efficiency curve of perovskite solar cells. Although the result is pretty interesting, I think the study is not complete at all. I would suggest you take some time and do a complete comprehensive study of the photovoltaic performance of solar cells made with all of your samples. I believe, by doing so chances of your paper getting cited world increase. I don't think there is a meaning of publishing an incomplete study at all. Please don't forget to add the detailed fabrication procedure which has been used to make the solar cells.
Author Response
Answer)As reviewer suggested, we added an improved power conversion efficiency of perovskite solar cell fabricated on ITO/SiO2/invar substrate and explained detailed fabrication procedure of perovskite solar cell. Although the performance of perovskite solar cells is not better than that of previously reported glass-based perovskite solar cells, we’d like to show the potential of flexible metal substrate for curved perovskite solar cells in this manuscript. As reviewer recommended, comprehensive study to improve the performance will be made with all of our samples. And the detailed results regarding performance of perovskite solar cells fabricated on invar substrate will be submitted in this journal.
Reviewer 2 Report
After reading the work, I have some questions:
- where is the stadistics of the devices?
- Do you think that annealing temperature could affect in the properties of the subtrates and therefore reduce the number of bending cycles.
- How can affect the stability of the perovskite in curve devices.
- Which metal contact do the authors propose in order to fabricate the solar devices.
Author Response
After reading the work, I have some questions:
- where is the statistics of the devices?
Answer)At this moment, we just demonstrated the feasibility of invar metal substrate for curved perovskite solar cells. Therefore, we didn't get statistics for the devices. However, as reviewer suggested, comprehensive study to improve the performance will be made with all of our samples. And the detailed results regarding performance of perovskite solar cells fabricated on invar substrate will be submitted in this journal.
- Do you think that annealing temperature could affect in the properties of the substrates and therefore reduce the number of bending cycles.
Answer)The PET substrate, which has bee employed as a typical flexible substrate, was degraded at a temperature of 150 ËšC. However the invar substrate is stable even at a temperature of 600ËšC due to its high thermal stability. Therefore, the annealing temperature doesn't affect the properties of the invar substrate and the number of bending cycles.
- How can affect the stability of the perovskite in curve devices.
Answer)In curved perovskite solar cell, the stability of devices could be affected by tensile and compressive stress applied to the perovskite active materials.
- Which metal contact do the authors propose in order to fabricate the solar devices.
Answer)We usually used the Ag metal for the devices.
