Additive Manufacturing of Stretchable Multi-Walled Carbon Nanotubes/Thermoplastic Polyurethanes Conducting Polymers for Strain Sensing

Abstract

With the development of science and technology, flexible sensors play an indispensable role in body monitoring. Rapid prototyping of high-performance flexible sensors has become an important method to develop flexible sensors. The purpose of this study was to develop a flexible resin with multi-walled carbon nanotubes (MWCNTs) for the rapid fabrication of flexible sensors using digital light processing additive manufacturing. In this study, MWCNTs were mixed in thermoplastic polyurethane (TPU) photosensitive resin to prepare polymer–matrix composites, and a flexible strain sensor was prepared using self-developed additive equipment. The results showed that the 1.2 wt% MWCNTs/TPU composite flexible sensor had high gauge factor of 9.988 with a linearity up to 45% strain and high mechanical durability (1000 cycles). Furthermore, the sensor could be used for gesture recognition and monitoring and has good performance. This method is expected to provide a new idea for the rapid personalized forming of flexible sensors.

Introduction

With the development of science and technology, wearable devices containing numerous sensing elements have brought great changes to our lives. Wearable flexible sensors provide important information about human health status in a timely and economical manner by measuring the human behavior parameters and sensing the surrounding environmental indicators, which have a positive impact on human health.¹ Polymer–matrix composites with excellent flexibility are not only widely used as carriers for flexible electronic devices, but also conductive polymer composites constructed based on them are used to prepare flexible pressure, strain, and other sensors.^2–8

Flexible sensors are suitable for different sensing principles. Commonly used preparation methods include liquid phase mixing method, chemical vapor deposition method, coating method, spin coating method, compression molding, and electrostatic assembly.^9–11 Gao and colleagues used carbonized melamine sponges to fabricate wearable flexible sensors.¹² Liang et al. used a covalent assembly-annealing strategy to prepare nitrogen-containing covalently cross-linked graphene foams with excellent compressibility and electrical conductivity.¹³ These traditional manufacturing methods have certain disadvantages such as long processing cycles and difficult fabrication of composite polymer microstructures. Additive manufacturing of the polymer-based flexible sensors has become a current research hot spot and has broad development prospects.^14–16

The commonly used additive manufacturing materials for flexible sensors include carbon nanocomposites, metals and composites, liquid metals, and PVD polymers.¹⁷ Among them, carbon materials are inexpensive and easy to fabricate due to their excellent electrical conductivity, mechanical properties, chemical stability, and thermal stability. In recent years, with the development of advanced carbon materials with unique dimensions such as one-dimensional carbon nanotubes and two-dimensional graphene, carbon materials have become a research hot spot for flexible strain sensors.^18–22 Furthermore, carbon nanotubes are used for the development of multifunctional composite materials with their excellent mechanical and electrical properties.^23,24 Furthermore, high aspect ratio multi-walled carbon nanotubes (MWCNTs) with low percolation threshold enable low-cost and high-sensitivity flexible sensor fabrication.²⁰

Presently, additive manufacturing is widely used in the manufacturing of metal parts, cell culture chips, scaffolds for tissue engineering,²⁵ and medical aids.^26–28 Among them, digital light processing (DLP) printing technology has small curing shrinkage, high printing accuracy, excellent surface quality, fast processing speed, and no emission of harmful gases during the curing. It can also process samples with complex structures and has a wide range of applications.²⁹ Wu and colleagues fabricated self-healing three-dimensional (3D) smart structures using DLP additive manufacturing technology, providing a unique strategy for 3D smart composites.³⁰ Thermoplastic polyurethanes (TPU) have high elongation and good affinity with the carbon fillers,³¹ and the fabrication of TPU-like conductive polymer-based strain sensors is widely used.³² However, there are only a few studies on photocuring forming of carbon nanotube/TPU composite flexible sensors.

In this work, high aspect ratio (tube length/diameter 3000) MWCNTs were uniformly dispersed in TPU photosensitive resin material using stirring and ultrasonic vibration, and the rapid prototyping of flexible sensors was realized using photocuring molding. The distribution of the carbon nanotubes in conductive polymers was further studied using field emission scanning electron microscopy (SEM) and Raman spectroscopy. The strain sensing behavior of the flexible sensor was analyzed using a digital bridge combined with a universal testing machine. From the results, it was found that the highest sensitivity of the fabricated sensor was 9.988, and the stability of the sensor was evaluated using cyclic stretching. The fabricated sensor is also found to satisfy 1000 stretching cycles and is used to monitor human movements.

Experimental Section

Materials

TPU photosensitive resin was purchased from Dongguan Shenshuo Technology Co., Ltd.; MWCNTs (diameter 10–20 nm, tube length <50 μm, purity >99.5%) were purchased from Shenzhen Yuechuang Evolution Technology Co., Ltd.; dispersant DISPERBYK-163 was purchased from Shanghai Kane Chemical Co., Ltd.; and ultraviolet (UV) thinner TPGDA was purchased from Shanghai Kaiyin Chemical Co., Ltd.

Preparation of composite material slurry

A certain amount of MWCNTs is added to the TPU photosensitive resin, and the weight percentage of MWCNTs is controlled at 0.1%, 0.3%, 0.5%, 0.7%, 0.9%, 1.1%, and 1.2% (sample names are TPU, TPU-0.1, TPU-0.3, TPU-0.5, TPU-0.7, TPU-0.9, TPU-1.1, TPU-1.2). When adding, the dispersant leveling agent is evenly added to the resin. The mixture was stirred in a YM-100S3 ultrasonic cleaner (Shenzhen Fangao Microelectronics Co., Ltd.) using a LC-ES-60 digital display mixer (Shanghai Lichen Instrument Technology Co., Ltd.) at 1200 rpm to obtain a composite material slurry. The preparation process was carried out in a dark environment.

Manufacture of strain sensors

The composite material slurry was placed in a self-developed photocuring forming equipment, the UV light intensity was 50 mW/cm², the printing layer thickness was 0.1 mm, and the exposure time was 40 s. After forming, the samples were cleaned in isopropyl alcohol cleaning solution for 2 min and then placed in a UV box for secondary curing for 5 min. After curing, the copper sheets are glued to both ends of the sample, and the copper wires are connected to complete the fabrication of the sensor.

Figure 1 shows the preparation process of the material and the principle of photocuring molding. The MWCNTs are uniformly dispersed in the TPU resin by stirring and ultrasonic vibration. The surface morphology of the formed samples was characterized by SEM, and the prepared samples were electrically measured by a digital bridge. All measurements were performed in air and at room temperature.

FIG. 1.

Schematic diagram of forming.

Performance characterization

The morphologies of MWCNTs, TPU, and MWCNTs/TPU composites were characterized by field emission SEM (Hitachi Regulus 8220). The chemical composition of MWCNTs, TPU, and MWCNTs/TPU was investigated using Raman spectrometer (Horiba Scientific-LabRAM HR Evolution, ion laser at 514 nm). These composite samples were measured with a digital bridge (RK2830; Shenzhen Merrick Electronic Technology Co., Ltd.) and a multimeter [UT60BT; Uni-Trend Technology (China) Co., Ltd.] resistance of the piece. The stress–strain curve and electromechanical properties of the strain sensor were measured by a universal material testing machine (China-Sansi Zongheng-UTM2203) and a digital bridge (RK2830; Shenzhen Meiruike Electronic Technology Co., Ltd.).

Conductivity is calculated according to Equation (1)³³: $ρ = \frac{R A}{L},$ (1)

where L is the distance between the two probes, R is the ohmic resistance measured by the digital bridge, and A is the cross-sectional area of the sample.

The sensitivity coefficient (gauge factor [GF]) of the sensor is calculated by Equation (2): $G F = \frac{(R - R_{0}) ∕ R_{0}}{ε} = \frac{Δ R ∕ R_{0}}{ε} .$ (2)

Here, ɛ is the strain of the sensor, and $R_{0}$ , $R$ , $Δ R$ are the initial resistance, the resistance at deformation ɛ, and the resistance change of the sensor, respectively.

Results and Discussion

The results of liquid conductivity of MWCNTs/TPU composites with different contents, characterization of photocurable forming conductive composites, sensing properties of composite materials, and the potential applications of the flexible sensors based on MWCNTs/TPU are briefly described below.

Liquid conductivity of MWCNTs/TPU composites with different contents

In this work, a multimeter was used to test the change of the resistance of the liquid composite material with the ultrasonic vibration time. Figure 2a shows the schematic diagram of the measurement. The measurement results are shown in Figure 2c, and from the figure, it is observed that with the increase of ultrasonic vibration stirring time, the liquid resistance gradually decreases and finally stabilizes. The optimum ultrasonic vibration stirring time is found to be 20 min. After the curing, the resistance of the sample was measured using Figure 2b, and the resistivity was calculated as shown in Equation (1). The results are shown in Figure 2d.

FIG. 2.

Electrical measurement method and characterization of MWCNTs/TPU composites. (a) Schematic diagram of liquid resistance test of MWCNTs/TPU composites. (b) Resistance test device of MWCNTs/TPU composite samples. (c) Liquid composite dispersion time-resistance. (d) Electrical conductivity of MWCNTs/TPU composites with various MWCNT mass contents. (e) SEM image of TPU. (f) SEM image of TPU-0.9. (g) SEM image of TPU-1.2. (h) SEM image of TPU cross section. (i) SEM image of TPU-0.9 cross section. (j) SEM image of TPU-1.2 cross section. MWCNTs, multi-walled carbon nanotubes; SEM, scanning electron microscopy; TPU, thermoplastic polyurethanes.

Figure 2d shows that the resistivity of the composite decreases with the increase of MWCNT content; however, it is not a simple linear relationship. The red line area is the seepage area, which is consistent with the literature.^34,35 Furthermore, according to the Ponte Castaneda–Willis model, the percolation threshold of MWCNTs is found to be about 0.3 wt%, which is close to 0.4 wt% in the experiment, indicating that MWCNTs are well dispersed in TPU. Figure 2e–j shows the SEM images of the surface of the MWCNTs/TPU samples at different concentrations and the SEM images of the pure TPU. The bright filamentous and punctate parts in the figure are the distribution state of MWCNTs in the resin, indicating that the MWCNTs are uniformly dispersed in the TPU.

Characterization of photocurable forming conductive composites

Figure 3a shows that the TPU characteristic peaks at 1294, 1458, 2870, and 2930 appear in the Raman spectrum of TPU. The peak at 1294 is the stretching vibration of the C = C double bond, and the peak at 1458 is attributed to the deformation of C-H. Furthermore, the observed bands are found to originate from C-H stretching vibrations, where 1348 is the D-band, 1594 is the G-band, and 2697 is the 2D-band for MWCNTs. Raman bands assigned to TPU and MWCNTs are observed for TPU/1.2MWCNTs, and it is considered that MWCNTs penetrate well into TPU.

FIG. 3.

Mechanical and electrical properties testing and results of MWCNTs/TPU composites. (a) Raman spectra of TPU, MWCNTs, and TPU/1.2MWCNTs. (b) Tensile properties of TPU and TPU/MWCNTs composites. (c) Stress–strain curves of TPU and TPU/MWCNTs composites. (d) Relative resistance change (ΔR/R₀) strain relationship of MWCNT/TPU flexible strain sensor under different MWCNT mass ratios. (e) Sensor response time and recovery time. (f) Comparison of sensor performance.

The mechanical tensile test was carried out using the formed tensile samples.³⁶ Figure 3b plots the comparison of the tensile properties of composites with different contents of MWCNTs, and Figure 3c shows the stress–strain curve. From the figures, it is found that the trend is consistent with the literature³⁷ Furthermore, the test results showed that with the increase of MWCNTs content, the elastic modulus initially increased and then decreased, and the elongation at break showed a decreasing trend as a whole. It is also observed that Young's modulus of the composite increases from 1.053 to 2.331 MPa when the MWCNT content is 0.5 wt%, an increase of 121%.

Moreover, the MWCNTs are uniformly dispersed in the TPU, and the hydrogen bonding between the MWCNTs and the TPU matrix is found to result in a strong interfacial bonding force between them.³⁸ The increase in Young's modulus may be due to the high stiffness of MWCNTs and the stronger interfacial bonding between MWCNTs and the matrix. Furthermore, with the increase of MWCNT content, the agglomeration will also increase. The internal defects of the MWCNT-based composites are found to increase with the increase of the content of MWCNTs, resulting in a decrease in elongation at break. Moreover, with the increase of MWCNT content, the degree of freedom of the TPU macromolecular chain is limited to a certain extent.³⁵

Figure 3d shows the sensitivity of the sensor with different contents of MWCNTs/TPU. The results showed that with the increase of MWCNT content, the sensitivity of the device showed an upward trend, and the highest sensitivity GF was obtained as 9.988 at 1.2 wt%. Moreover, the maximum detectable strain was found to be 45%, and it had good linearity within 45% deformation. Furthermore, at low levels, it is observed that there are fewer conductive channels, and when strain is applied, the channels are less on and off, exhibiting less sensitivity. Figure 3e shows the response time and recovery time of the sample tested at a high tensile rate of 1000 mm/min under a small strain of 5%. The results show that the response time of the composite is 400 ms and the recovery time is 1200 ms. The longer recovery time may be due to the hysteresis of the material itself. Figure 3f compares the performance of the sensor prepared in this study with the current study. The relevant data are shown in Table 1.

Table 1.

Comparison of Sensor Performance

Material	Elongation at break (%)	Young's modulus (MPa)	Conductivity (S/cm)	Sensitivity	Refs.
FCMS	80	—	—	18.7	¹²
RGO/ER	199.66	0.118	1.03E-5	6.7	³⁹
PANI/PDMS	23	—	—	10	⁴⁰
EGaIn/SEBS microfiber	1000	0.92	2.6E3	1.6	⁴¹
CNTs/UV-curable resin	0.05	2500	1.0E-4	2	⁴²
CNT/EA	201	0.122	4.73E-5	8.9	³⁷
CNTs/TMTCTW	7	350	1.0E-3	38	⁴³
MWCNTs/TPU-resin	48	0.41	2.79E-4	9.988	This work

CNT, carbon nanotube; EA, elastomer; EGaIn, Gallium–indium eutectic; ER, elastomer resin; FCMS, fragmented carbonized melamine sponges; MWCNTs, multi-walled carbon nanotubes; PANI, polyaniline; PDMS, polydimethylsiloxane; RGO, graphene oxide; SEBS, Styrene-ethylenebutylene-styrene; TMTCW, Name of a product of Zhilei Electronics And Technology (Wuhan) Co., Ltd; TPU, thermoplastic polyurethanes; UV, ultraviolet.

Sensing properties of composite materials

Figure 4a and bshows the physical drawings and drawings of the tensile sample, respectively. Figure 4c and d demonstrates the good flexibility of TPU-1.2 composites, respectively. Figure 4e shows the cyclic loading stress–strain curves of the three MWCNTs with a tensile specimen type variation of 45%. From the figures, it is observed that TPU-0.3 and TPU-0.9 have large plastic deformation and poor elasticity in the loading and unloading stage, and TPU-1.2 shows good elasticity. It is also found that after a certain number of cyclic stretching, the internal structure of the TPU-1.2 composite tends to be stable, and the mechanical response behavior has good repeatability and recoverability.^44–46

FIG. 4.

Cyclic tensile test results of MWCNTs/TPU composites. (a) The drawing of the tensile sample. (b) The actual drawing of the tensile sample. (c) The straight state of the sample. (d) The bent state of the sample. (e) TPU-0.3, TPU-0.9, cyclic tensile stress–strain curves of TPU-1.2. (f) Resistance recording curves of different strains of TPU-1.2. (g) Enlarged view of part of area of (f).

Mechanical durability is an important characteristic to ensure the long-term usability of strain sensors,²² and Figure 4f illustrates the resistance response curve of TPU-1.2 composites cycled 1000 times at the same tensile speed. Figure 4f illustrates the cycling stability of the DLP 3D-printed sensor based on the MWCNT content of 1.2 wt% under 15%, 30%, and 45% strain conditions. The results indicate that the developed strain sensor has high mechanical durability. Furthermore, the sensor resistance showed a trend of increasing first and then decreasing. This may be due to the fact that during the stretching process, the MWCNTs in the MWCNTs/TPU composite, the MWCNTs in the matrix are rearranged along the stretching direction, and the conductive channels parallel to the stretching direction will increase due to the larger aspect ratio and winding structure. Moreover, due to the after-effect caused by the viscoelastic behavior of the composite, during the stretching–shrinking cycle, an optimized conductive network was gradually formed, leading to final stabilization.

Figure 4g shows an enlarged view of a portion of Figure 4f, showing a clear right shoulder. This may be because there is competition for the destruction and remodeling of the conducting channel. Furthermore, the change mechanism is shown in Figure 5. In the initial state, Figure 5a corresponds to the conductive network shown in Figure 5d. It is observed that during the initial state, the uniformly dispersed MWCNTs overlapped each other to form stable and efficient conductive paths. When the material is in a stretched state, as shown in Figure 5b corresponding to the conductive network shown in Figure 5e. The MWCNTs stretched with the movement of the TPU molecular chains, leading to the destruction of some conductive channels, and the distance between the conductive fillers is found to increase leading to the disconnection of the network (indicated by the red area).

FIG. 5.

Variation mechanism of tensile strain and conductive network of MWCNTs/TPU composites under various strains. (a) Initial state. (b) State during stretching. (c) State when returning to the initial state position. (d) Conductive network in the initial state. (e) Conductive network in the stretched state. (f) Conductive network after restoring the initial state.

According to the tunnel effect model of resistance, increasing the distance between conductive fillers reduces the probability of electron transitions, resulting in an increase in the resistance value of the material.⁴⁷ At the same time, the MWCNTs will re-lap together to form a new conductive network and generate new conductive channels (green area). The destruction of the conductive channels during the stretching process is dominant, so the resistance value of the composites shows an increasing trend during the stretching process. After returning to the initial position as shown in Figure 5c, the corresponding conductive network is shown in Figure 5f.

Part of the conductive network was restored to the initial state due to the tensile failure of the conductive channel, and the resistance showed a downward trend. However, due to the hysteresis effect of the polymer material and the permanent destruction of some channels during the stretching process, residual resistance will appear during the cycle recovery process. Thus, there is a competition between disruption and reconfiguration of the conductive network, causing fluctuations in the strain response of the material, exhibiting a shoulder effect. Finally, after cyclic stretching, the conductive network is stabilized, and the material exhibits good repeatability and recoverability.

Cyclic resistance–strain response mechanism

To describe the mechanical response mechanism and to quantify the resistance–strain relationship of MWCNTs/TPU composites under cyclic action, the following mathematical model was established.

The main reason for the various resistance of the composite material under different strains is the change in the conductive network. Based on the Krauss model, the number of interconnections between particles per unit volume is defined as $N_{1} (ε)$ ⁴⁸:

In the equation: $N_{0}$ is the initial interconnection number per unit volume; $ε$ is the corresponding strain; $_{ε c}$ is a constant, usually replaced by yield strain; $m$ is a parameter related to the fractal structure of the conductive network.

In the large deformation state, assuming that the conductive network is in the viscous deformation stage, the number of particle connections per unit volume $N_{2} (ε)$ increases linearly with time⁴⁸: $N_{2} (ε) = k_{2} t = k_{2} ε ∕ \dot{ε},$ (4)

where $k_{2}$ is a constant related to the conductive network of matrix properties; $t$ is the loading time; $\dot{ε}$ is the strain rate.

During the change in the conductive network due to the deformation, a part of the network is pulled off, resulting in irreversible damage. At the same time, due to the viscoelasticity of the material, the destruction and reconstruction of the conductive network accompany the entire cycle process, which is the same as the change mechanism of the conductive network. According to the Smoluchowski formula, the following equation is used to describe the complex quantitative change $N_{3} (ε)$ relationship between the destruction and reconstruction of the conductive network during the deformation process.⁴⁹ $N_{3} (ε) = N_{0} (η_{1} - η_{2} e^{- k \frac{ε}{\dot{ε}}}),$ (5)

where $η_{1}$ , $η_{2}$ , and $k$ are constants related to the destruction and reconstruction of the conductive network. Since in the unloading stage, the conductive network has a similar loading network reconstruction in the TPU, the total number of conductive networks $N (ε)$ in the loading and unloading stages can be composed of $N_{1} (ε)$ , $N_{2} (ε)$ , $N_{3} (ε)$ . $N (ε) = N_{1} (ε) + N_{2} (ε) + N_{3} (ε) .$ (6)

Referring to the theory of percolation, the relationship between resistivity and the number of conductive networks is as follows: $ρ \propto {(N (ε))}^{- n_{ε}} = {(N_{1} (ε) + N_{2} (ε) + N_{3} (ε))}^{- n_{ε}} .$ (7)

Combining Equations (3) to (7) are:

where $ρ$ is resistivity; $n_{ε}$ is scaling index; $ξ = k_{2} N_{0}$ .

The relationship between strain and the spatial spacing of conductive particles is adjusted by the parameter $E$ , and the expression for the variation of the spatial spacing between conductive particles is as follows: $L = L_{0} (1 + E ε),$ (9)

where $L$ and $L_{0}$ are the spatial distance and initial distance between conductive particles.

According to the volume invariance of composites and the relationship between resistance and resistivity: $\frac{R}{R_{0}} = \frac{ρ}{ρ_{0}} {(E ε + 1)}^{2} .$ (10)

Combining the above Equations (8) and (10), we can get:

In the equation: $R$ and $R_{0}$ are the resistance value and the initial resistance value corresponding to the strain $ε$ .

The sample is loaded with constant strain (the strain curve is shown in Fig. 6a), and the measurement results are well fitted with the theoretical model [Eq. (11)]. The fitted results are shown in Figure 6b. The corresponding fitting parameters are listed in Table 2. Due to the viscoelastic properties of the material, a shoulder peak appears to a greater extent during the unloading phase. From the curve in the figure, it is observed that the experimental value and the theoretical model have a high degree of fit in the loading and unloading stage. Therefore, the model can be used to describe the strain–resistance response of this test.

FIG. 6.

Strain resistance fitting curve. (a) Strain curve. (b) Strain resistance test/fitting curve.

Table 2.

Parameters Obtained by Fitting R/R₀-Strain Curves for Multi-Walled Carbon Nanotubes/Thermoplastic Polyurethanes Composites

Graphene content	$E$	ɛ_c	$m$	$ξ$	$η_{1}$	$η_{2}$	$k$	$n_{ε}$
Compression	0.2691	0.4228	0.5479	0.9427	0.4177	0.9831	0.3015	0.7011
Release	0.1887	0.2875	0.0911	0.5762	0.6834	0.5466	0.4257	0.6444

Potential applications of flexible sensors based on MWCNTs/TPU

Since additive manufacturing can produce complex structures, this experiment attempted to form complex-structured composite materials. The designs are shown in Figure 7a.I and a.IV, and the formed parts are shown in Figure 7a.II and a.V. This method can realize the preparation of flexible parts with complex two-dimensional and 3D structures. In this article, wearable flexible sensors were fabricated from TPU/MWCNTs composites. Human joint movements are detected by attaching sensors to different parts of the human body with the help of medical tape. Figure 7b shows the recognition of human fingers using strain sensors, which can detect the finger straightening and bending movements. Furthermore, the resistance is found to change significantly, and the resistance is recovered after the recovery action, which proves the repeatability of the sensor. Therefore, the developed sensor can be used to detect finger gestures and has significant potential in the control of gesture robots and other virtual display applications.^50,51

FIG. 7.

Potential applications. (a) Design drawing of complex structure: (a.I, a.II) 2.5D design drawings and formed parts, (a.III) is the expanded view of Fermat spiral, and (a.IV, a.V) the design drawings and physical drawings of the three-dimensional structure. (b) Finger gesture recognition (bending and straightening).

Conclusions

In this work, MWCNTs were uniformly dispersed in TPU photosensitive resin using ultrasonic vibration and stirring, and low-cost flexible sensors were fabricated using self-developed DLP photocuring equipment. From the results, it was found that the prepared flexible sensor had high sensitivity and large tensile strain, with a sensitivity factor of 9.988 and a working strain range of 45%. The composites prepared in this article will fluctuate in the strain–resistance response of the material due to the competition between the destruction and reconstruction of the conductive network during stretching, thereby exhibiting a shoulder effect, and can form a stable after cyclic stretching. conductive network.

Furthermore, the fabricated strain sensor exhibited excellent cyclic stability under loading and unloading cycles, which could provide stable and repeatable electrical signals. The sensor could also reach thousands of cycles or more, demonstrating the application of strain sensors to detect the motion of human knuckles. This preparation method showed the prospect of DLP 3D printing technology in flexible sensors and provided an idea for personalized custom wearable electronic devices. Therefore, it is concluded that DLP 3D printing technology is a promising approach for wearable flexible devices, which could further promote the development of wearable electronics.

Footnotes

Acknowledgments

The authors extend their science thanks to those who contributed in instructions and experiments work.

Authors' Contributions

F.Li.: Conceptualization, methodology, investigation, formal analysis, and writing—original draft. D.B.: Software and validation. D.X.: Resources and supervision. Supervision. L.S.: Funding acquisition, resources, and supervision. Z.T.: Project administration. J.Z.: Resources and supervision.

Author Disclosure Statement

There are no conflicts to declare.

Funding Information

This work was supported by Jiangsu Provincial Key Research and Development Program (No. BE2019002). This research was partly financed by the JITRI Institute of Precision Manufacturing.

References

Amjadi

, Kyung

, Park

, et al. Stretchable, skin-mountable, and wearable strain sensors and their potential applications: a review. Adv Funct Mater, 2016; 26:1678–1698.

Gunasekaran

, Ponnan

, Zheng

, et al. Facile fabrication of highly sensitive thermoplastic polyurethane sensors with surface- and interface-impregnated 3D conductive networks. ACS Appl Mater Interfaces, 2022; 14:22615–22625.

Wang

, Shirzaei Sani

, Gao

. Wearable bioelectronics for chronic wound management. Adv Funct Mater, 2021; 32:2111022.

Ray

, Choi

, Bandodkar

, et al. Bio-integrated wearable systems: a comprehensive review. Chem Rev, 2019; 119:5461–5533.

Sun

, Wang

, Ren

, et al. Smart band-aid: multifunctional and wearable electronic device for self-powered motion monitoring and human-machine interaction. Nano Energy, 2022; 92:106840.

Jiang

, Zhang

, Wang

, et al. Topological supramolecular network enabled high-conductivity, stretchable organic bioelectronics. Science, 2022; 375:1411–1417.

Zhang

, Wang

, Jiang

, et al. High-brightness all-polymer stretchable LED with charge-trapping dilution. Nature, 2022; 603:624–630.

Gao

, Wang

, et al. Highly sensitive pseudocapacitive iontronic pressure sensor with broad sensing range. Nanomicro Lett, 2021; 13:140.

, Song

. Printable smart materials and devices: strategies and applications. Chem Rev, 2022; 122:5144–5164.

10.

Chang

, Guo

, Sun

, et al. Direct writing and repairable paper flexible electronics using nickel-liquid metal ink. Adv Mater Interfaces, 2018; 5:1800571.

11.

, Li

, Zhang

, et al. Piezoresistive sensors based on rGO 3D microarchitecture: coupled properties tuning in local/integral deformation. Adv Electron Mater, 2019; 5:1800461.

12.

Fang

, Tan

, Gao

, et al. High-performance wearable strain sensors based on fragmented carbonized melamine sponges for human motion detection. Nanoscale, 2017; 9:17948–17956.

13.

Liang

, Zhao

, Tang

, et al. Covalent bonds-integrated graphene foam with superb electromechanical properties as elastic conductor and compressive sensor. Carbon, 2019; 147:206–213.

14.

Liu

, Zhang

, Han

, et al. 3D printed flexible strain sensors: from printing to devices and signals. Adv Mater, 2021; 33:2004782.

15.

Deng

, Hu

, Lei

, et al. Stimuli-responsive conductive nanocomposite hydrogels with high stretchability, self-healing, adhesiveness, and 3D printability for human motion sensing. ACS Appl Mater Interfaces, 2019; 11:6796–6808.

16.

Chen

, Liu

, Tian

, et al. 3D-printed anisotropic polymer materials for functional applications. Adv Mater, 2022; 34:e2102877.

17.

Wang

, Cen

, Zeng

. Highly sensitive flexible tactile sensor mimicking the microstructure perception behavior of human skin. ACS Appl Mater Interfaces, 2021; 13:28538–28545.

18.

Kang

, Pikhitsa

, Choi

, et al. Ultrasensitive mechanical crack-based sensor inspired by the spider sensory system. Nature, 2014; 516:222–226.

19.

Patel

, Park

, Bonato

, et al. A review of wearable sensors and systems with application in rehabilitation. J Neuroeng Rehabil, 2012; 9:21.

20.

Mannsfeld

SCB

, Tee

, Stoltenberg

, et al. Highly sensitive flexible pressure sensors with microstructured rubber dielectric layers. Nat Mater, 2010; 9:859–864.

21.

, Chen

, Jiang

, et al. Recent progress in active mechanical metamaterials and construction principles. Adv Sci, 2022; 9:e2102662.

22.

Sui

, Downing

, Hersam

, et al. Additive manufacturing and applications of nanomaterial-based sensors. Mater Today, 2021; 48:135–154.

23.

Lin

, Zheng

, Yang

, et al. A high-performance, sensitive, wearable multifunctional sensor based on rubber/CNT for human motion and skin temperature detection. Adv Mater, 2022; 34:e2107309.

24.

Wang

, Zhou

, Li

, et al. High-performance foam-shaped strain sensor based on carbon nanotubes and Ti3 C2 Tx MXene for the monitoring of human activities. ACS Nano, 2021; 15:9690–9700.

25.

Qiu

, Zhao

, Haghiashtiani

, et al. 3D printed organ models with physical properties of tissue and integrated sensors. Adv Mater Technol, 2018; 3:1700235.

26.

Alizadehgiashi

, Gevorkian

, Tebbe

, et al. 3D-printed microfluidic devices for materials science. Adv Mater Technol, 2018; 3:1800068.

27.

Bhattacharjee

, Urrios

, Kang

, et al. The upcoming 3D-printing revolution in microfluidics. Lab Chip, 2016; 16:1720–1742.

28.

CMB

, Ng

, Li

KHH

, et al. 3D printed microfluidics for biological applications. Lab Chip, 2015; 15:3627–3637.

29.

Dong

, Vuckovac

, Cui

, et al. 3D printing of superhydrophobic objects with bulk nanostructure. Adv Mater, 2021; 33:2106068.

30.

Miao

, Ge

, Wu

, et al. 3D printing of sacrificial thermosetting mold for building near-infrared irradiation induced self-healable 3D smart structures. Chem Eng J, 2022; 427:131580.

31.

, Zhou

, Byun

, et al. Highly stretchable multi-walled carbon nanotube/thermoplastic polyurethane composite fibers for ultrasensitive, wearable strain sensors. Nanoscale, 2019; 11:5884–5890.

32.

Zhang

, Zhang

, Huang

, et al. A new class of electronic devices based on flexible porous substrates. Adv Sci (Weinh), 2022; 9:e2105084.

33.

Peng

, Guo

, Thirunavukkarasu

, et al. Tailoring of photocurable ionogel toward high resilience and low hysteresis 3D printed versatile porous flexible sensor. Chem Eng J, 2022; 439:135593.

34.

Christ

, Aliheidari

, Ameli

, et al. 3D printed highly elastic strain sensors of multiwalled carbon nanotube/thermoplastic polyurethane nanocomposites. Mater Des, 2017; 131:394–401.

35.

Wang

, Liu

, Schubert

. Highly sensitive ultrathin flexible thermoplastic polyurethane/carbon black fibrous film strain sensor with adjustable scaffold networks. Nanomicro Lett, 2021; 13:64.

36.

Zhou

, Gao

, Fu

, et al. Multimaterial 3D printing of highly stretchable silicone elastomers. ACS Appl Mater Interfaces, 2019; 11:23573–23583.

37.

Xiao

, Qian

, Yin

, et al. 3D printing of flexible strain sensor array based on UV-curable multiwalled carbon nanotube/elastomer composite. Adv Mater Technol, 2020; 6:2000745.

38.

, Wang

, Zhou

, et al. An electrically conductive polymer composite with a co-continuous segregated structure for enhanced mechanical performance. J Mater Chem C, 2020; 8:11546–11554.

39.

Qian

, Xiao

, Chen

, et al. 3D printed reduced graphene oxide/elastomer resin composite with structural modulated sensitivity for flexible strain sensor. Adv Eng Mater, 2022; 24:2101068.

40.

Zheng

, Wu

, Huang

, et al. Highly sensitive and multifunctional piezoresistive sensor based on polyaniline foam for wearable human-activity monitoring. Compos Part A Appl Sci Manuf, 2019; 121:510–516.

41.

, Zhang

, et al. Highly permeable and ultrastretchable liquid metal micromesh for skin-attachable electronics. ACS Mater Lett, 2022; 4:634–641.

42.

Cortés

, Sánchez-Romate

, Jiménez-Suárez

, et al. Mechanical and strain-sensing capabilities of carbon nanotube reinforced composites by digital light processing 3D printing technology. Polymers, 2020; 12:975.

43.

, Wang

, Dunn

, et al. Digital light processing 3D printing of conductive complex structures. Addit Manuf, 2017; 18:74–83.

44.

Coleman

, Khan

, Gun'Ko

. Mechanical reinforcement of polymers using carbon nanotubes. Adv Mater, 2006; 18:689–706.

45.

Andrews

, Jacques

, Qian

, et al. Multiwall carbon nanotubes: synthesis and application. Acc Chem Res, 2002; 35:1008–1017.

46.

Shofner

, Rodrı´guez-Macı´as

, Vaidyanathan

, et al. Single wall nanotube and vapor grown carbon fiber reinforced polymers processed by extrusion freeform fabrication. Compos Part A Appl Sci Manuf, 2003; 34:1207–1217.

47.

Gong

, Zhu

. On the mechanism of piezoresistivity of carbon nanotube polymer composites. Polymers, 2014; 55:4136–4149.

48.

Boland

, Khan

, Ryan

, et al. Sensitive electromechanical sensors using viscoelastic graphene-polymer nanocomposites. Science, 2016; 354:1257–1260.

49.

Yang

, Yao

, Zheng

, et al. Highly sensitive and stretchable graphene-silicone rubber composites for strain sensing. Compos Sci Technol, 2018; 167:371–378.

50.

Kwon

, Kim

, Kwon

, et al. All-printed nanomembrane wireless bioelectronics using a biocompatible solderable graphene for multimodal human-machine interfaces. Nat Commun, 2020; 11:3450.

51.

Zhou

, Chen

, Li

, et al. Sign-to-speech translation using machine-learning-assisted stretchable sensor arrays. Nat Electron, 2020; 3:571–578.