Pushing detectability and sensitivity for subtle force to new limits with shrinkable nanochannel structured aerogel

There is an urgent need for developing electromechanical sensor with both ultralow detection limits and ultrahigh sensitivity to promote the progress of intelligent technology. Here we propose a strategy for fabricating a soft polysiloxane crosslinked MXene aerogel with multilevel nanochannels inside its cellular walls for ultrasensitive pressure detection. The easily shrinkable nanochannels and optimized material synergism endow the piezoresistive aerogel with an ultralow Young’s modulus (140 Pa), numerous variable conductive pathways, and mechanical robustness. This aerogel can detect extremely subtle pressure signals of 0.0063 Pa, deliver a high pressure sensitivity over 1900 kPa−1, and exhibit extraordinarily sensing robustness. These sensing properties make the MXene aerogel feasible for monitoring ultra-weak force signals arising from a human’s deep-lying internal jugular venous pulses in a non-invasive manner, detecting the dynamic impacts associated with the landing and take-off of a mosquito, and performing static pressure mapping of a hair.

1. What are basic design requirements? Are there any reports or references?
2. There are many low-class mistakes, such as the sentence in lines 92-94 "Previous work has reported that ……. in cellular walls13, 14." should be corrected into "Previous works have reported that ……. in cellular walls13, 14." Please carefully double check the main text. 3. Please cited the reported papers in your sentences such as "Overall, although the recent intense research effort has effectively advanced the development of piezoresistive aerogels and achieved high performance for certain aspects of sensing properties." In lines 102-104. Do not forejudge anything in manuscript without evidence or references! 4. What is the intrinsic sense of "3C" or just the abbreviation of words (Composite materials, nanochannel structure, cross-linking interface)? Besides, there are any logical relationships between these three words. The authors obviously oversell the term "3C". 8. The authors claim that "PGPDMS can act as a spacer materials for intercalation into the MXene interlayer and prevent the restacking of the MXene nanosheets via molecular assembly." So how to verify this conclusion? 9. The authors have make numerus assumptions such as "Such multilevel nano-channel structure allows the detection of tiny external force by shrinking (or expanding) the space between the nanochannels (meeting requirement (i)). Moreover, numerous new conductive pathways can be created between the neighboring MXene nanosheets during nano-channel shrinking (meeting requirement (ii)), leading to considerable resistance changes inside the cellular walls……" in lines 178-183.
10. The authors have claimed that "ICP-MX-AG are investigated using molecular dynamics (MD) simulations ( Figure S4) and Supporting note 2)". However, there is no any direct evidence for verifying the structure of ICP-MX-AG, so what is the sense of simulation? What is the base for using MD simulation? 11. The density is very important parameter! So the calculation procedure details should be provided.
12. The authors claim that "……expansion of nano-channels in the cellular walls of BBP-MX-AG" in line 256, however, there is no any direct evidence to verify this conclusion. The in-situ characterization only just shows the curving of MXene bundle sheets under loading.
Reviewer #3 (Remarks to the Author): Here the composite aerogel by Shi et al. could respond to subtle stimuli down to 0.0063 Pa with a high sensitivity over 1400/kPa. The result, of course, is impressive. However, I do have concerns on the novelty of this work. Please address the following points for further submission.
1. The authors stressed high sensitivity of their sensor. However, high sensitivity at low pressures is not a challenging work. There have been many published papers reporting a maximum sensitivity far higher than 1900/kPa. For example, ACS Nano 2021, 15, 1795-1804 (sensitivity= 380000/kPa); and ACS Appl. Mater. Interfaces 2018, 10, 40880-40889. The authors should do a more comprehensive research on the state of the art of this field.
2. The demonstration of the change in the d-spacings of the MXene layers via in-situ TEM inspection seems amazing ( Figure 3). However, I question about the applied stress level in this observation (by using indentation). Since the specimen is significantly deformed, the stress is estimated to be on GPa level, while the sensing test is conducted on 0.001-1 Pa level. The difference is at about 10 orders of magnitude. Obviously, it is not convincing to clarify the sensing mechanism by using the in-situ indentation and TEM observation.
3. The authors claimed that: To realize accurate, continuous, and ultrasensitive monitoring of ultraweak pressure stimuli, a desired piezoresistive material should have all of the following structural and material properties: (i) ultralow Young's modulus to significantly reduce the critical stress value that triggers the material deformation; (ii) multilevel or hierarchical structure with ultrahigh densities of variable conductive pathways to allow very large changes in the electrical conductivity of the sensing materials during structural deformation; (iii) excellent mechanical elasticity and robustness to prevent the collapse and disintegration of the material during repeated structure deformations. I agree that a low Young's modulus may help. However, there should be more ways to achieve ultrasensitivity other than what are claimed here. Also, the second and the third points are not convincing. 4. Appling a small pressure lower than 0.1 Pa is difficult. I like the method offered by the authors. They may provide more details on the use of acoustic pressure.
5. There are too much typo and grammatic errors. They should ask a native English speaker to further improve the language. 5. More references had been added to support our statements and conclusion (on Page 3, 4, 5, 8 11, 14, 17, 21, and 22) according to the comments from reviewer#2.
6. The abbreviations bad been improved according to the comments from reviewer#2 (on Page 7, 8, and 9). 10. The density calculation details had been added in the Method part (on Page 25) according to the comments from reviewer#2. 11. Figure 3 had been modified and more in-situ HRTEM images (Supplementary Figure S18) had been added to confirm the shrinkable nature of nanochannel structure in the cellular walls of BBP-MX-AG according to the comments from reviewer#2.
12. More reference had been added in Supplementary Table 1 according to the comments from  reviewer#3. 13. XRD patterns (Supplementary Figure S19) had been added to further confirm the shrinkable nature of nanochannel structure in the cellular walls of BBP-MX-AG under subtle pressure according to the comments from reviewer#3.
14. More details on the use of acoustic pressure had been added in the Method part (on Page 26) according to the comments from reviewer#3.
15. The language has been fully polished by a native English speaker according to the comments from reviewer#3.
Reply to the reviewer#1:  Figure 6e is too large to construct a high sensing density. A sensing array with smaller pixels and higher sensing resolution should be provided.

REPLIES:
Thanks for the suggestion! We have made a 5 × 5 sensing array with an active pixel area of 2 mm in diameter and the total device area of 1.5 × 1.5 cm 2 as demonstrated in Figure   S30 in Comments: Why the maximum compressive strain in Figure S10 for the MX-AG was only 30%?

REPLIES:
Thanks for the question! Pure MXene aerogel (MX-AG) without the addition of polysiloxane exhibited poor mechanical properties, and the aerogel would collapse when the compressive strain was larger than 30%.
Comments: Whether the BBP-MX-AG with a density of ~7 mg/cm2 and Young's modulus of 58 Pa (Figure 2e) can exhibit an even smaller pressure detectable limit?

REPLIES:
Thanks for the question! Piezoresistive BBP-MX-AG (density = ~7 mg/cm 2 ) exhibited nearly the same pressure detection limit (0.0063 Pa) as the one with a density of ~10 mg/cm 2 (Supplementary Figure S28). This is because the pressure detection limit is mainly determined by the shrinkable nanochannels in the cellular walls of BBP-MX-AG. Both BBP-

MX-AG samples have similar multilevel nanochannels in their cellular walls because their initial
MXene-to-silane weight ratio and fabrication conditions were the same; however, the sensitivity of BBP-MX-AG with a density of ~7 mg/cm 2 was lower than the one with a density of ~10 mg/cm 2 (Supplementary Figure S28). This is because, at a lower mass density, fewer conductive pathways can be formed during compression, which results in a smaller resistance change under the same applied pressure. We have added the corresponding data and discussion in the revised  The sensing mechanism of pressure sensors includes piezoelectricity, capacitance, and piezoresistivity, and the sensitivity of a pressure sensor is defined as the slope of relative electrical signal change versus the applied pressure.
By contrast, strain sensors (such as the sensing device reported in Science Bulletin, 2020, 65(11): 899) transduce mechanical deformation into electrical signal when stretching the sensing devices.
The sensitivity (gauge factor) of strain sensors refers to the slope of the curve of relative electrical signal change versus applied strain. Cracks generating and propagating in conductive thin film/network during stretching and thus greatly limiting the electrical conduction through the thin film/network is the main mechanism exploited in strain sensors. The strain sensors, which was in the form of densely packed film reported in Science Bulletin, 2020, 65(11): 899, is based on this crack-propagation mechanism.
Reply to the reviewer#2: Comments: What are basic design requirements? Are there any reports or references?

REPLIES:
Thanks for the question! We have deleted the corresponding statements of "basic design requirements", reorganized the ideas, and rewritten the introduction part in the revised manuscript. In addition, more comprehensive references were cited to support our statement and analysis (Highlight in abstract and summary part, and on Page 3, 4, and 5 of the manuscript).  Comments: The cartoon images in Fig 1 looks beautiful Figure S18). Although the sensitivity of our BBP-MX-AG sensor was lower than that of the devices reported in ACS Nano 2021, 15, 1795 and ACS Appl. Mater. Interfaces 2018, 10, 40880, the detectable pressure limit of our sensing device (0.0063 Pa) was still better than that of the sensing devices published in ACS Nano 2021, 15, 1795 (0.025 Pa) and ACS Appl. Mater. Interfaces 2018, 10, 40880 (5 Pa).

Comments: The demonstration of the change in the d-spacings of the MXene layers via in-situ
TEM inspection seems amazing (Figure 3). However, I question about the applied stress level in this observation (by using indentation). Since the specimen is significantly deformed, the stress is estimated to be on GPa level, while the sensing test is conducted on 0.001-1 Pa level. The difference is at about 10 orders of magnitude. Obviously, it is not convincing to clarify the sensing mechanism by using the in-situ indentation and TEM observation.

REPLIES:
Thanks for the question! Although the specific value of the external force supplied by the nano-indenter during the in-situ HRTEM characterization in Figure 3 was unclear, it should not be at the GPa level since the critical stress value that triggers the buckling or bending of the cellular walls in the aerogels was lower than 1 Pa, as calculated in the Supporting note 1 in the original SI. To verify the shrinkable nanochannel structure in the cellular walls of BBP-MX-AG, we further characterize the change of interlayer distance of the cellular walls in BBP-MX-AG under tiny force using XRD, and the corresponding discussion and data were added in the revised manuscript and SI (Highlight on Page 13 and 14 in the manuscript and Supplementary Figure S19). As can be observed from the XRD patterns in Figure S19, Comments: There are too much typo and grammatic errors. They should ask a native English speaker to further improve the language.

REPLIES:
Thanks for pointing out this problem! A native English speaker had helped us to fully polished the language of the revised manuscript.
(HRTEM), and X-ray diffraction (XRD)." But there're no SEM and HRTEM data for ICP-MX-AG in main text and supporting information. Please make sure all the data you want to present are added correctly.
4. There are two terms "MXene-AG" and "MX-AG" appear in the manuscript at the same time. I believe these two nouns are actually refer to the same thing. Please choose only one term and keep consistent in the article.
5. The main difference between BBP-MX-AG and ICP-MX-AG is compressive modulus. In this kind of material system with hierarchical structures, lighter compressive modulus should be directly related to the detection limit that the material can reach. So in this point of view, the authors may consider exploring the relationship between different R1 (substituent of polysiloxane) and compressive modulus.
6. I noticed that in Fig  7. It's quite interesting to learn that authors tried to use a homemade acoustic pressure measurement method to applying pressure lower than 0.1 Pa. More details about calibration of these measurement should be provided. This is not a standard or at least not that common test method when it comes to piezoresistive pressure sensors. Is there anyone else using this method? Please add the references if possible.

REPLY to REVIEWERS for MANUSCRIPT NCOMMS-21-05897A
Dear Editor: We appreciate your consideration of our manuscript! We also appreciate the valuable comments and advises from the reviewer#4, which can help to further improve the quality of our  Figure S24 in the revised SI (highlighted in Page 16 and Figure   S24).

REPLIES:
Thanks for pointing out this problem. We have modified the Figure 6 in the revised manuscript (highlighted in Figure 6).

Comments: The main difference between BBP-MX-AG and ICP-MX-AG is compressive modulus.
In this kind of material system with hierarchical structures, lighter compressive modulus should be directly related to the detection limit that the material can reach. So in this point of view, the authors may consider exploring the relationship between different R1 (substituent of polysiloxane) and compressive modulus.

REPLIES:
Thanks for the suggestion! In fact, the relationship between different R I (substituent of polysiloxane) and compressive modulus is complicated, and more work will be carried out in our future investigations. In the present work, to simply explore the relationship between detection limit and compressive modulus of our sensing aerogels, we have measured the pressure sensing performance of BBP-MX-AG with two different mass densities: ~7 mg/cm 2 and ~10 mg/cm 2 . BBP-MX-AG with density of ~7 mg/cm 2 and compressive modulus of 58 Pa exhibited nearly the same pressure detection limit (0.0063 Pa) as the one with a density of ~10 mg/cm 2 and compressive modulus of 140 Pa (Supplementary Figure S28). These results indicated that the compressive modulus may not be the critical factor to determine the pressure sensing limit of our sensing aerogels. This is because the pressure detection limit is mainly determined by the shrinkable nanochannels in the cellular walls of BBP-MX-AG. Both BBP-MX-AG samples have similar multilevel nanochannels in their cellular walls because their initial MXene-to-silane weight ratio and fabrication conditions were the same; however, the sensitivity of BBP-MX-AG with a density of ~7 mg/cm 2 was lower than the one with a density of ~10 mg/cm 2 (Supplementary Figure S28). This is because, at a lower mass density, fewer conductive pathways can be formed during compression, which results in a smaller resistance change under the same applied pressure. Moreover, the detection limit of the pressure sensors assembled from ICP-MX-AG (with density of ~10 mg/cm 2 ) was only 0.1 Pa (Supplementary Figure S27), despite the presence of shrinkable nanochannels in its cellular walls. MD simulations revealed that the crosslinked PGPTMS network exhibited a much higher compressive modulus than that of bottlebrush-like PGPDMS (Supplementary Figure S9). Thus, a larger force stimulus was required to trigger the deformation of the crosslinked network of PGPTMS in the nanochannels.
All these results confirm that the intercalation of a soft polymer with small modulus as the molecular spacer into the nanochannels of multilevel cellular walls plays a critical role in the fabrication of highly sensitive piezoresistive aerogels. More thorough studies will be carried out in our coming work to investigate this question.
Comments: I noticed that in Fig S11. EDS Figure S11 in the revised SI (highlighted in Figure S11).

REPLIES:
Thanks for the question! In fact, using a homemade acoustic pressure measurement method to applying pressure has been commonly reported in some previously published works (ACS Nano 2021, 15, 1795; Nat. Commun., 2015, 6, 6269). We have cited these two references in the revised manuscript (highlighted in Page 16). As we have descripted in the Method part of our manuscript, the sound pressure is provided by a homemade loudspeaker box which can produce sound or acoustic waves with various frequencies (similar to Figure S31 in ACS Nano 2021, 15, 1795 and Figure S13 in Nat. Commun., 2015, 6, 6269). This sound source was positioned 5 cm over the sensing device and was faced directly at the sensor surface. The output sound wave is edited by "GoldWave" software, and the sound wave template is in "freq" mode under "wave". The volume of the output acoustic signal was controlled by computer, and the precise volume or sound level reaching the sensor surface was calibrated and determined by a decibel meter. The resistance and current changes were measured using a Keithley 2000 digital multimeter. The corresponding sound pressure (P) can be calculated by , where P 0 and L dB denote the reference sound pressure in air (20 µPa) and the measured sound pressure level, respectively (Aquacult. Eng. 2007, 37, 125). Thus, the sound pressure that reaching to the surface of our sensing samples can be precisely controlled by the homemade loudspeaker box. It should be pointed out that to prevent environmental noise interference, the whole measurement setup should be installed inside a noise isolation chamber. We have provided more details about this part in the method section of the revised manuscript (highlighted in Page 26 and 27).