Ultrathin CNTs Film Based on Marangoni Effect for Strain Sensing Application
Round 1
Reviewer 1 Report
Ultrathin CNTs Film Based on Marangoni Effect for Strain Sensing Application
This paper is the good study on studying the influence of Marangoni effect on the preparation of ultrathin films. The work is novel and found to be acceptable for publication. However, the application of CNT film for strain sensing application is not clear in this manuscript. It needs to be discussed elaborately. The sensitivity measurement of strain sensors after some constant cycles were observed and reported. Is there any technical or numerical validation available for the same? This makes sense and provides additional support for the experimental data. The images and tables presented in this paper is found to be acceptable. The conclusion should be elaborated with important findings and experimental data.
Ultrathin CNTs Film Based on Marangoni Effect for Strain Sensing Application
This paper is the good study on studying the influence of Marangoni effect on the preparation of ultrathin films. The work is novel and found to be acceptable for publication. However, the application of CNT film for strain sensing application is not clear in this manuscript. It needs to be discussed elaborately. The sensitivity measurement of strain sensors after some constant cycles were observed and reported. Is there any technical or numerical validation available for the same? This makes sense and provides additional support for the experimental data. The images and tables presented in this paper is found to be acceptable. The conclusion should be elaborated with important findings and experimental data.
Author Response
We greatly appreciate the reviewer's comments. We have revisied the manuscript carefully. In response to the reviewer's suggestions, we will make the following explanations.
1. We have further discussed the application of CNTs film for strain sensing in our manuscript. We added sensing principles of strain sensors based on CNTs films in lines 68 to 71. We also explained the sensing principle of our sensor in lines 220 to 222.
2. There are some technical and numerical validation for the same constant cycles tests on strain sensors, and we have cited the related references in our manuscript. The updated content can be found in lines 280 to 293.
3. We have rewritten the conclusion, incorporating some key data and results (in lines 317-328).
Reviewer 2 Report
This paper presents possible application of ultrathin carbon nanotubes for strain measurements. This is interesting topics but I found some week points, mentioned below, and they have to be explained or removed before possible publishing of this manuscript in "Coatings".
1. The method of measuring the mechanical stresses was not described - the authors only provided the devices they used for the measurement. Please correct this part of the manuscript.
2. What were the planar dimensions of the tested strain sensors?
3. As Figs. 3a), 3d) and 3g) show, the relative changes in resistance as a function of stress are non-linear. Thus, GF should have a functional relationship and not a constant value as shown in Figs. 3c) and 3f). Please respond to this matter..
4. On what basis the Authors conclude the stable long-term behavior of the sensor shown in Fig. 4c) - this figure shows a significant decrease in the value of relative resistance changes and probably also significant changes in the initial resistance of this element after a series of mechanical cycles. Please comment more detaily this figure (with proper values).
Based on above remarks I think that this manuscript needs major revision and further re-referring.
Author Response
Many thanks for the valuable feedback. We will respond to the reviewer's questions one by one as follows.
Q1. The method of measuring the mechanical stresses was not described - the authors only provided the devices they used for the measurement. Please correct this part of the manuscript.
A1: For strain sensors, the stress is generally not measured. We focus on the relationship between the strain and the electrical signals (current and sheet resistance) of the sensors. In addition, we have made necessary revisions to the experimental section.
Q2. What were the planar dimensions of the tested strain sensors?
A2: The width of the conductive layer is 0.8 cm and the distance between the two electrodes is about 2 cm as described in scetion 2.2 (in lines 140-142). All sensors were kept the same planar dimensions.
Q3. As Figs. 3a), 3d) and 3g) show, the relative changes in resistance as a function of stress are non-linear. Thus, GF should have a functional relationship and not a constant value as shown in Figs. 3c) and 3f). Please respond to this matter.
A3: ΔR/R0 of the strain sensor changes linearly with strain within a certain strain range. The GF values were measured within the linear regions. In Figures 3(d) and 3(g), we identified the linear ranges by using a black fitted lines.
Q4. On what basis the Authors conclude the stable long-term behavior of the sensor shown in Fig. 4c) - this figure shows a significant decrease in the value of relative resistance changes and probably also significant changes in the initial resistance of this element after a series of mechanical cycles. Please comment more detaily this figure (with proper values).
A4: We have revised the relevant paragraph and provided a more detailed explanation of Figure 4c, please refer to lines 288 to 294. Overall, the ΔR/R0 change does indeed show a downward trend, but in long-term cyclic experiments, the trend is relatively gentle and stable.
Reviewer 3 Report
The manuscript is a research article entitled “Ultrathin CNTs Film Based on Marangoni Effect for Strain Sensing Application”. The following issue should be addressed as part of a major revision of the current form.
Based on its scope, there is a lack of novelty in the current work. Previous authors have taken a similar approach (DOI: 10.1016/j.fmre.2022.05.010, DOI: 10.1002/adfm.201504717, and DOI: 10.1039/c8tc00711j). What is unique about the author's current work compared to the published one? There is a need for the authors to include relevant references on the previous work that has used ultrathin CNTs film based on the Marangoni effect for strain sensing.
I am waiting for a response from the author before making a decision.
Author Response
Thanks a lot for recommending recent-published relative papers to us and we have already read carefully and cited them in our manuscript. In fact, our work has a lot of similarities with the aforementioned papers because, as in all work, the Marangoni effect was used to form ultrathin carbon films,which in turn were used to fabricate the flexible strain sensors. Moreover, our work was inspired by Li et al. (DOI: 10.1002/adfm.201504717). Even so, our research has the characteristics that differ from the literature, such as simpler and more efficient operation, and better sensing performance of the prepared sensors.
Compared with the work of Chen et al. (DOI: 10.1016/j.fmre.2022.05.010), our work shows differences in three aspects. (1) The fabrication process of our sensor is simpler and more time-saving, which can be finished in a few hours. As described in the paper, Chen et al. used pre-stretched and plasma-treated PDMS (spin-coated on a piece of aluminum foil at 600 min-1) as substrate to transfer CNTs film and then removed the aluminum foil by soaking in hydrochloric acid solution for hours. In our work, the ultrathin CNTs film was directly and completely transferred to a plastic substrate (lid of petri dish), and then liquid PDMS was cast on its surface to be cured. This process is fast and can ensure the integrity of the CNTS film. (2) The CNTs films can be transferred to PDMS twice without apparent defects to enhance the conductivity of composite materials by our method. Nevertheless, obvious defects such as cracks and bubbles occurred frequently when we transferred the second layer of CNTs film to PDMS using the method described in the paper of Chen et al. (3) Our senor owns higher sensitivity (GF=3.4) and excellent cyclic stability (ε=30%, 8000 cycles). Indeed, we didn't find the accurate data for describing the sensitivity and cyclic stability in the paper of Chen et al. We can calculate from Figure 3(d) in their paper that (R-R0)/R0 of a 150%-pre-stretched strain sensor was about 0.13 at 15% strain and come to a conclusion that GF was about 0.9. It means we fabricated a more sensitive strain sensor using a simpler strategy. Our manuscript is helpful to supply accurate data for readers to evaluate the sensitivity and stability of high-performance strain sensors.
Compared with the work of Li et al. (DOI: 10.1002/adfm.201504717), which has been cited 295 times, there are several improvements in our work as follows. (1) We used CNTs instead of GN (graphene) to form conductive films based on Marangoni effect to fabricate sensor with high sensitivity and wide working range. GN nanosheets are easy to slip even under subtle strains because the 2D nanosheets didn't connect tightly, resulting in super-high GF (1037, ε = 2%) but much limited working range (3.3%) for the strain sensor. However, the CNTs network is much tighter with 1D curly CNTs intertwining with each other, permitting high conductivity in a wide strain range for the sensor. (2) We optimized the transferring method of ultrathin film on water surface effectively. Our method can avoid defects including bubbles and cracks in manual operation, and the full conductive film can be transferred totally to PDMS repeatedly. (3) Our sensor showed excellent cyclic stability (ε=30%, 8000 cycles).
Compared with the work of Liang et al. (DOI: 10.1039/c8tc00711j), our work is different at following aspects. (1) Our sensor shows better sensing properties, including higher sensitivity (GF=3.4, ε=0-50%), wider working range (87%) and more excellent cyclic stability (ε=30%, 8000 cycles). The sensor of Liang et al. exhibited GF of 0.97 (ε=0.5-10%), 1.23 (ε=10-30%), 1.97 (ε=30-50%), working range of 50% and cyclic stability of 1000 cycles at ε=1%. Our sensor exhibits high comprehensive performance mainly because of the special transferring method for ultrathin films. (2) With our transferring method, the films on water surface can be transferred twice onto PDMS to enhance the conductivity of composite material. However, it looks difficult to repeat again using the method of Liang et al. because the CNTs films have been cured with liquid TPE (thermoplastic elastomer).
Of course, our work is not entirely unique. As mentioned above, we have also been inspired by the work of Li et al. (DOI: 10.1002/adfm.201504717). It can be said that our work has only made slight progress based on the work of others. It should be said that our work, along with the listed literature work, has its own characteristics and jointly promotes the progress of carbon-based strain sensor technology.
Reviewer 4 Report
coatings-2414857
The manuscript reports the fabrication of CNT film for stretchable strain sensors by Marangoni-driven self-assembled. This manuscript can be published in a Coating journal, but the following comments should be considered before publication.
1- Line 32: it is better to categorize strain sensors based on their transition mechanism of them. In this case, you can add magnetism or inductance types.
2- Line 34: Mention more positive advantages of resistive-type sensors, such as stable signal output or simple signal reception.
3- What's the reason of composition conductive films and flexible substrates in resistive-type sensors?
4- What are the sensing mechanisms of resistance-type flexible strain sensors? You can refer to the related published paper (DOI: 10.1016/j.isci.2022.105162).
5- Line 53: Instead of using the term "if necessary", mention the reasons for the combination with other conductive materials exactly.
6- what’s the effect of humidity on resistive-type sensors-based CNT? You can refer to the related published paper (DOI: 10.1109/JSEN.2020.3038647).
7- deposited electrodes by silver paste in your sample in Figure 1 aren't clear.
8- Specify the conditions of microscopy in the SEM figures.
9- Add the relationship between Id and Ig and the quality of CNT. In this case, you can confirm a dense network of high-quality CNTs in the 2 layers-combined films.
10- "CNTs are one-dimensional materials in a curly state." This sentence needs a related reference.
11- What is the base value of sensor resistance?
The article needs to be revised more carefully by the authors. For example, in lines 53 and 55, the reference number is given after the dot!
Author Response
The reviewer has carefully reviewed our manuscript and proposed very helpful suggestions. We deeply appreciate his/her work. Here we reply the questions/suggestions one by one as follows.
Q1. Line 32: it is better to categorize strain sensors based on their transition mechanism of them. In this case, you can add magnetism or inductance types.
A1: We have categorized strain sensors based on their transition mechanism and added magnetism and inductance types with recent-published references.
Q2. Line 34: Mention more positive advantages of resistive-type sensors, such as stable signal output or simple signal reception.
A2: We have mentioned more positive advantages of resistive-type sensors in lines 34 to 36, and the relevant references have been cited.
Q3. What's the reason of composition conductive films and flexible substrates in resistive-type sensors?
A3: We have added the relative explanation and references in lines 38 to 44 in the revised manuscript.
Q4. What are the sensing mechanisms of resistance-type flexible strain sensors? You can refer to the related published paper (DOI: 10.1016/j.isci.2022.105162).
A4: After reading the paper, we explained the sensing mechanism of resistance-type flexible strain sensors as "The sensing mechanism of resistive-type flexible strain sensors is that the electron transport paths change due to the destruction and reconstruction of conductive network under strains, or the contact resistance and tunneling resistance change due to the dis-placement of the conductive nanomaterials." (lines 45-48) The paper has been cited in our manuscript.
Q5. Line 53: Instead of using the term "if necessary", mention the reasons for the combination with other conductive materials exactly.
A5: We have revised the text as "CNTs usually collaborate with other conductive materials, such as carbon black (CB), graphene, silver nanowires (AgNWs), for that the conductive network composed of the nano-materials of different dimensions can enable the sensor to exhibit high sensitivity and wide working range simultaneously." (lines 63-67)
Q6. What's the effect of humidity on resistive-type sensors-based CNT? You can refer to the related published paper (DOI: 10.1109/JSEN.2020.3038647).
A6: Shooshtari et al. fabricated a gas sensor based on aligned CNTs using plasma-enhanced chemical vapor deposition (PECVD) method. The conductivity of the sensor with CNTs exposed in the chambers with different relative humidity reduced about 4% when the relative humidity reduced from 10% to 80%. As for our strain sensor based on CNTs film, we think the relative humidity may have little effect on electrical conductivity of the packaged sensor because of the protection of PDMS. Thanks a lot for your advice, which enlighten us to a great extent in the further work.
Q7. Deposited electrodes by silver paste in your sample in Figure 1 aren't clear.
A1: We have updated the Figure 1 and added comments of conductive silver paste and copper wire.
Q8. Specify the conditions of microscopy in the SEM figures.
A8: The conditions of microscopy in the SEM images have been specified. In addition, in section 2.3, the acceleration voltage and platinum-sputter-coated treatment for SEM characterization are also specified. (lines149-151)
Q9. Add the relationship between Id and Ig and the quality of CNT. In this case, you can confirm a dense network of high-quality CNTs in the 2 layers-combined films.
A9: We have added the relationship between ID and IG and the quality of CNTs in lines 208-212 as "The value of ID/IG reflects the defect degree of the sample. The higher the value of ID/IG, the more defects the sample contains and the more irregular the structure. The ID/IG values of CNTs powder and 2 layers-combined film are 1.29 and 1.33 respectively, and the similar values mean a dense network of high-quality CNTs have been informed in the 2 layers-combined films."
Q10. "CNTs are one-dimensional materials in a curly state." This sentence needs a related reference.
A10: We have added a related reference [19]. (lines 224-225)
Q11. What is the base value of sensor resistance?
A11: Generally, the base value of the strain sensor resistance is about several hundreds of kiloohms. And for our sensor, the value is between 100-200 kiloohms.
Q12. The article needs to be revised more carefully by the authors. For example, in lines 53 and 55, the reference number is given after the dot!
A12: We checked and confirmed the reference numbers and their formats in the manuscript, and carefully polished the full text of the manuscript.
Round 2
Reviewer 2 Report
Many thanks for improving the manuscript. However I am not satisfied for some of your answers, especially on Q3
Q3. As Figs. 3a), 3d) and 3g) show, the relative changes in resistance as a function of stress are non-linear. Thus, GF should have a functional relationship and not a constant value as shown in Figs. 3c) and 3f). Please respond to this matter.
A3: ΔR/R0 of the strain sensor changes linearly with strain within a certain strain range. The GF values were measured within the linear regions. In Figures 3(d) and 3(g), we identified the linear ranges by using a black fitted lines.
In the improved manuscript I found the following sentence (lines 326-329);
When the PCP sensor was stretched by 5%, 10%, 20% and 30% strain, the electrical signals showed obvious difference, with ΔR/R0 of 0.04, 0.12, 0.24, 0.48, respectively. The ΔR/R0 was relatively small under 5% strain and sharply ncreased under the strain of 30%.
This means that for ε = 5 % we have ΔR/R0 = 4% and GF = 0.8, for ε = 10 % we have ΔR/R0 = 12% and GF = 1.2, for ε = 20 % we have ΔR/R0 = 24% and GF = 1.2, and for ε = 30 % we have ΔR/R0 = 48% and GF = 1.6.
This contradicts the authors' earlier statement that ΔR/R0 of the strain sensor changes linearly with strain within a certain strain range. The GF values were measured within the linear regions. Moreover mentioned by Authors linear ranges identifieb by Authors in Figs. 3d) and 3g) are practically ivisible to the readers.
Therefore, this fact has to be clarified before the possible printing of the peer-reviewed manuscript.
Author Response
Thanks for the valuable feedback provided by the reviewer. These comments have also prompted us to think and analyze the experimental results more deeply. We have carefully checked our experimental process and compared the experimental results with other types of strain sensors reported in the literatures. Finally, we have come to the following conclusion.
We have reconsidered the results shown in Figure 3(a), 3(d) and 3(g). We ever though that the ΔR/R0 change increased linearly with strain within a certain strain range. Now we believe that it is incorrect. In fact, ΔR/R0 change of strain sensors increased nonlinearly with the increase of strain, however, the trend of the ΔR/R0 change was not significant yet at low strain range, so it seemed linear. Only when the strain reached a higher level would the ΔR/R0 change sharply increase. We retract the answer for Q3 in last response so that the contradiction between "linear change" and the experimental results no longer exists, at the same time, the fitted lines in Figure 3(d) and 3(g) in the revised manuscript were also removed.
The phenomenon that ΔR/R0 change increased nonlinearly with the strain has also been reported in literatures. For example, from the Figure 4(d) in Jia's paper [1], we can caculate that the ΔR/R0 changes were about 1.39, 7.05, and 21.08 for strain of 10%, 40%, and 50%, respectively, indicating that the GFs were about 139, 176, and 421 for strain of 10%, 40%, and 50%, respectively. As pointed out by the reviewer, from Figure 4 (b) in our paper, it can be calculated that the GF values at 5%, 10%, 20%, and 30% strain were 0.8, 1.2, 1.2, and 1.6, respectively. As discussed before, the GF of the PCP sensor at 50% strain was 3.4. It can be seen that the overall trend of GF change becomes increasingly significant. However, the regularity of this set of experimental results is not optimal, but it is indeed our original experimental result. We will further verify the relationship between ΔR/R0 change and strain in subsequent experiments.
Ref:
[1] Jia, Y.; Chen W.; Ye, C.; et al. Controllable formation of periodic wrinkles in Marangoni-driven self-assembled graphene film for
sensitive strain detection. Science China Materials. 2020, 63(10), 1983-1992. 10.1007/s40843-020-1314-1
Reviewer 3 Report
The authors answered all of my questions and comments. The manuscript has been revised to incorporate all suggestions. The manuscript is suitable for publication in Coatings journal as it is consistent with its content.
Author Response
Thanks for the feedback from the reviewer.
Round 3
Reviewer 2 Report
I am not fully satisfied with the response to my last review. I hope that in the future the Authors will pay more attention to the analysis of the results, therefore I propose to publish the paper in the current version.
