Highly Stretchable and Wearable Strain Sensor ... - ACS Publications

Subscriber access provided by MT ROYAL COLLEGE

Article

Highly Stretchable and Wearable Strain Sensor Based on Printable Carbon Nanotube Layers/ Polydimethylsiloxane Composites with Adjustable Sensitivity Xin Wang, Jinfeng Li, Haonan Song, Helen Huang, and Jan Gou ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b17766 • Publication Date (Web): 12 Feb 2018 Downloaded from http://pubs.acs.org on February 16, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Highly Stretchable and Wearable Strain Sensor Based on Printable Carbon Nanotube Layers/Polydimethylsiloxane Composites with Adjustable Sensitivity Xin Wang1, Jinfeng Li1, Haonan Song1, Helen Huang1, Jihua Gou1,* 1

Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando,

FL 32816, USA *Corresponding author, Email: [email protected] ABSTRACT Strain sensors that are capable of monitoring complex human motions with high accuracy are highly desirable for developing wearable electronics. This paper reports the fabrication of highly stretchable and sensitive multi-directional strain sensors with tunable strain gauge factors by employing a digitally controlled printer to incorporate carbon nanotube (CNT) layers into polydimethylsiloxane (PDMS) substrate. The fabricated sensors exhibit a high stretchability (up to 45%) and sensitivity with a gauge factor of 35.75. The gauge factors could be easily modulated by tuning the number of CNT printing cycles to accommodate diverse requirements. The cyclic loading-unloading test results revealed that the composite strain sensors exhibited excellent long-term durability. Particularly, in this work, for the first time, human motioninduced strain was measured by a motion capture system and compared with the strain data obtained from the fabricated strain sensors. The deviation of strains measured by composite sensors is less than 20%, indicating the great accuracy of CNT/PDMS sensors to quantify the amount of motion-induced strain. Of significant importance is that due to the flexibility of the printing technique used, rosette-typed sensors were fabricated to simultaneously measure strains along multiple axes. These superior sensing capabilities of the fabricated CNT/PDMS strain sensors give them great application potential in motion detecting systems. Keywords: Carbon nanotube, PDMS, composite strain sensor, digital manufacturing, human motion monitoring

1. Introduction Monitoring human body motion has drawn great interests for developing artificial skin and healthcare-related wearable electronics1-3. Strain sensors that allow the mechanical deformations to be quantified through corresponding electrical signals have been a promising engineering tool for human body motion detection. Commercially available metal based strain sensors have 1

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

limited stretchability (typically ~5%) and sensitivity(~2 of gauge factor) to detect large-scale human motions like bending of arms or hands4. As such, there is a pressing need to develop stretchable and wearable strain sensors with high sensitivity for monitoring human motions. Composites have attracted considerable attention due to their ability to combine the matrix and reinforcement to achieve excellent functional properties5-7, and composites consist of flexible polymer matrix and piezo-resistive filler have been a promising candidate for flexible and wearable strain sensors8-9. To fabricate strain sensors with desired properties, various piezo-resistive filler have been used to date, including metal nanomaterials10, graphene11-14, graphene oxide15 and carbon nanotubes(CNT)16-17. In particular, CNTs have been used extensively for flexible strain sensors due to their unique electrical and mechanical properties18-19. The high aspect ratio of carbon nanotubes makes them entangled to form internal conductive network, and the network could breakdown when stretched and reconstruct when the strain released, showing a measurable electrical resistance change. Several studies have been conducted on the flexible strain sensors based on carbon nanotube and elastomers. For instance, CNTs were spray-deposited on a PDMS substrate to prepare a strain sensor, which could accommodate strain up to 150% for detection of large-scale human motions but shows low sensitivity (~6 of gauge factor)16. Besides this, a CNT based strain sensor produced by the compression molding method exhibited a high sensitivity with a gauge factor of 152.93 at 30% strain20. Very recently, a CNT paper based strain sensor has been shown to have a very high sensitivity, with the highest gauge factor approaching 107,21, but its long response time limited its application for human motion monitoring. Most of previously-reported sensors were fabricated through expensive and complex processes without controllable sensitivity or good response linearity for measuring diverse human motions. Additionally, these techniques such as film transfer and solution casting are only capable of producing strain sensors with simple rectangle geometry to measure the strain in a single direction, but for monitoring human motion with a complicated strain condition, strain sensors that can measure strain from multi-direction are favorable. Furthermore, in most of previous reports, the ability of strain sensors to detect human motion was just demoed and the accuracy of the strain data obtained from the sensors has rarely been reported. Recent works on the fabrication of nanocomposites with printing technology offer a promising way to overcome the above-mentioned problems22-24. Digital printing could enable us to fabricate arbitrary sensor configuration with well-controlled sensing properties. In this work, 2


Page 2 of 22

Page 3 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


highly stretchable and wearable strain sensors with adjustable sensitivity was fabricated by a digitally-controlled spray deposition modeling (SDM) printer to incorporate CNT layers into PDMS substrate. The additive feature and exquisite print path control of SDM enable the continuous fabrication of arbitrary sensor configuration with low cost. The CNT layers were printed layer-by-layer on the PDMS substrate with controllable thickness and desired patterns. The resultant strain sensor exhibited a gauge factor of 35.75 and a detection limit of 45%, which is qualified for multi-scale human motion. To our best knowledge, the accuracy of CNT/PDMS strain sensors to monitor skin stretch during wrist flexion movements is first demonstrated in our study through comparison with data recorded by a motion capture system. We further demonstrate the capabilities of rosette-typed sensors to elaborate complex strain condition. The high sensitivity of our sensors, which is comparable to most of recently reported CNT/elastomer strain sensors fabricated by printing technique, coupled with its remarkable advantages, such as high accuracy, cost-effective and scalable, gives them great application potential in wearable electronics, human-machine interaction and other related areas. 2. Experimental Details 2.1 Preparation of CNT inks Multi-walled carbon nanotubes (MWCNTs), diameter of 30-50 nm, length of ~20 µm, provided by Chengdu Organic Chemical, were used as received. Triton X-100(Fisher Bioreagents) was used as surfactant to disperse MWCNTs in deionized water. The CNT aqueous inks were obtained by adding 0.1 g CNTs and 0.25 mL Triton X-100 to 150 mL deionized water, followed by sonication (Q1375 Sonicator from Qsonica) for 30 mins at room temperature and centrifugation (Thinky, USA) at 2200 rpm for 5 mins. The final received CNT ink has a good stability for at least two weeks, as shown in Figure S1. 2.2 Preparation of CNT/PDMS composites First, PDMS and its curing agent (Sylgard 184 Silicone Elastomer Kit, Dow Corning) were mixed in 10:1 weight ratio and cured in an oven at 100°C for 1.5 hrs. Acid treatment (H2O2:H2SO4 = 1:3, H2O2 30%; H2SO4 98%; purchased from Sigma-Aldrich) was performed to increase the hydrophilicity of PDMS substrate for aqueous ink printing. PDMS substrate was treated in acid for 5min and then rinsed with deionized water, and dried in the oven. After the surface modification process, PDMS were obtained as a substrate with a thickness of 0.2 mm and used for the further printing process. Secondly, the prepared CNT ink was printed on the PDMS substrate using a home-made spray deposition modeling printer. The pattern design of sensors 3



was done in a CAD software. Printer set-up and printing parameters could be found in our previous study25-26. Specimens were printed at the speed of 15 mm/s and during the printing process, PDMS substrates were place on a heating platen with temperature of 90°C to induce the water evaporation. After the printing process, copper wires were attached to the end of CNT patterns and finally another layer of PDMS was cured on top of printed patterns to obtain the composite strain sensors. 2.3 Characterization Scanning electron microscopy (SEM, Zeiss-Ultra 55) was used to examine the morphology of the composite strain sensors under an accelerating voltage of 10 kV. The strain sensing tests were carried out on a MTS hydraulic test system. Specimens were stretched/released at a constant frequency of 0.2 Hz. Real-time electrical resistance was measured during the tests using a Fluke 45 digital multi-meter connected to a PC that records data through LabView software. Silver paste electrodes were coated to the two ends of sensors to reduce the contact resistance. Human wrist motion was monitored by a motion capture system (the motion capture set up is shown in Figure S6). The passive reflective markers, 3 mm diameter, were attached to the corners of the strain sensor (4 on unidirectional-type sensor and 12 on rosette-type sensor). A 9camera motion capture system (OptiTrack Prime 13W, NaturalPoint Inc., USA) sampling at a frame rate of 240 frames per second was used to record the coordinates of the markers. The maximum residual value was set to 2.7 mm, and the system calibration was completed with a mean 3D error of 0.411 mm before the single session assessment.

4


Page 4 of 22

Page 5 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. Schematic of the digital fabrication of composite strain sensors through the SDM technique

3. Results and Discussions Figure 1 schematically depicts the digital fabrication process of composite strain sensors carried out through the deposition of CNT on PDMS substrate using the SDM printing process. In this study, polydimethylsilicone (PDMS) was utilized as flexible substrate for composite strain sensors owing to their high elasticity and fast response speed to strain8. However, the surface of PDMS substrate is highly hydrophobic with very low surface energy27. Therefore, appropriate surface modification of PDMS is required to increase the wetting of aqueous CNT ink. In this work, an inexpensive approach involving chemical treatment in acid solution was used to render the hydrophilicity of PDMS surface. Static contact angles were measured with water by using a calibrated syringe and a digital camera. The contact angle value measured is shown in Figure S3. It is observed that with 5 mins treatment, the contact angle reduced from 97° to 55°, which is beneficial for printing of aqueous ink. The hydrophilicity improvement of PDMS surface could be attributed to an oxidized surface where Si-CH3 bonds were attacked by the strong acid oxidizer to form a hydrophilic Si-OH group28. The detailed mechanism is described in Figure S4. To control the thickness of the CNT layer in composites, different numbers of printing cycles were performed. Figure S2 shows the sheet resistance of CNT layers on PDMS substrate. After 50 printing cycles, the decreasing of resistance is less obvious, thus in this study, specimens fabricated with different numbers of printing cycles from 10 to 50 with an interval of 10 layers were investigated. Figure 2 shows printed CNT layers on PDMS substrates. It shows that the CNT layer became darker as printing cycle increases, indicating the increasing of the thickness of CNT layers. It should be noticed that when the printing number is small, the deposited CNT layer is not perfectly uniform because of the “coffee ring” effect with CNT dispersion29. As the printing number increases, printing on deposited CNT layers is easier than the printing on the smooth PDMS substrate that has pretty low surface energy, so more homogenous CNT layers formed. The fabricated strain sensor is shown in Figure 2(b), which indicating the excellent stretchability.

5



Figure 2. (a) Optical images of CNT layers printed with 10, 20, 30, 40 cycles; (b) photograph of the strain sensor subjected to 40% strain

To identify the morphology of the printed pristine CNT layer on PDMS substrate, cross-section SEM images of CNT layers obtained by 10 and 50 printing cycles were shown in Figure 3(a) and (b), respectively. Figure 3(c) clearly shows that CNTs were uniformly deposited on PDMS substrate to form a dense and continuous conductive network without the agglomeration of CNTs, owing to the homogenous dispersion of CNTs in water as well as the optimized and effective SDM printing process. The thickness of CNT layers printed by different numbers of cycles could be acquired from the SEM images (Figure S5), as plotted in Figure 3(d). As the number of printing cycles increase, the CNT layer becomes thicker correspondingly, indicating the tunable CNT layer thickness by controlling the number of printing cycles. The thickness of CNT layers was 1.6, 3.2, 5.4, 7.2, 8.4 µm, which corresponds to 10, 20, 30, 40, 50 printing cycles. The corresponding CNT concentration in composites is also shown in Figure 3(d). Figure 3(e) and (f) shows the morphology of the fractured surface of CNT/PDMS composite, where the PDMS resin fully impregnated throughout CNT networks indicating a good interfacial bonding between matrix and fillers. The “sandwich” layered structure with one CNT embedded composite layer was also confirmed from Figure 3(e). 6


Page 6 of 22

Page 7 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Cross-section SEM images of pristine CNT layers on PDMS substrate with (a) 10; (b-c) 50 printing cycles; (d) CNT layer thickness and concentration in composites as a function of printing cycle number; (e-f) Cross-section SEM images of CNT/PDMS composite with 50 printing cycles of CNT layer

The thickness effect of CNT layers on the strain sensing property of the composite sensors was investigated by comparing the sensors made from different CNT printing cycles. The strain sensors with an area of 15x20 mm2 and thickness of 0.4 mm were prepared for sensing testing, and were named as CPx, where x denotes the printing cycles of CNT layer. These composite strain sensors were characterized with regard to the applied strain under tension, and the strain sensors were stretched up to 45% strain until the damage of substrates was found. Figure 4(a) shows the representative resistance change-strain data. All the sensors exhibited similar piezoresistive responses to the strain, and with the increasing of strain, the electrical resistance of all the sensors increased. The strain-dependent responses of the sensors were found to have two regions. A simple model was proposed to describe this sensing mechanism of printed strain sensors.

7



18 16 14

50

(b) 45

CP10 CP20 CP30 CP40 CP50

40 30 20

40 35 Gauge Factor

Gauge Factor

(a) 20

R/R0

10

12

12 17 22 27 32 37 42 47 Strain(%)

10 8 6

30 25 20 15

4

10

2

5

0

0

0

5

10

15

20 25 30 Strain(%)

35

40

45

50

10

20 30 40 Printing Cyle Number

50

(c) 1000 Ref. 37 100 Gauge Factor

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 22

Our Work Ref. 34 Ref. 30 10

Ref. 33 Ref. 35

Ref. 36 Ref. 31

Ref. 32

1 Ref. 17 0.1 0

100

200

300

400

Maximum Sensing Strain

Figure 4. (a) Resistance change–strain relationship of composite sensors with different CNT layer thickness (b) Corresponding gauge factors. (Inset in panel (a) shows the gauge factor-strain relationship after initiation stage) (c) Comparison of the gauge factor and maximum sensing strain of CNT/PDMS strain sensor with those of the counterparts reported printed strain sensors in the literature ref 17, 30-37

At low strain region (less than 5%), the stretching of CNT network was just initiated in the loading direction, and there are still many CNT junctions due to the compression of transverse direction, so the resistance change was slow at the beginning of the stretching. There is no obvious crack or gap appeared in the network. According to the previous report, CNTs could be regarded as elastic materials38, thus at low strain, most of the CNTs are still connected, the change of tunneling resistance is a dominant phenomenon. The transition from the non-sensitive to sensitive region, i.e. the strain at which the contact areas of CNT network start reducing significantly and the crack starts to form, is defined as critical strain. At high strain, the cracks generate and propagate in CNT networks under stretching, which significantly limit the electrical 8


Page 9 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


conduction through CNT networks. With increasing of strain, the number of cracks increases, but as shown in SEM image in Figure 5, some CNT networks still bridge the cracks, preventing the entire rupture of conduction network. Figure 5 schematically depicts the resistance change mechanism under stretching in CNT networks. According to this model, the overall resistance change as strain increases could be described by the following empirical equation (Supporting information, equation 2-11):

∆R (1 + ε)exp(aε) - 1 ≈  Ro bε + c(ε − ε c )n Where ∆ R , R0 and

(0 ≤ ε < εc ) (ε ≥ εc )

ε are the resistance change, initial resistance and the applied strain,

respectively. a is the constant related to the distance between CNT ends at un-stretched state, b is the constant related to the piezoeletric coefficient of the composites, c and n are the constants related to the mechanical properties of CNT films and the cracking behavior of the composites. For strain sensors fabricated by certain printing cycles, these constants could be reasonably obtained by fitting the experimental curve. For CS50, a, b, c, n was obtained as 3.95, 7.89, 31.05 and 0.95 respectively, and a good correlation is achieved with R2 of 0.996 between the experimental data and empirical equation, as shown in Figure 6. The value of exponent constant n in the equation is close to 1, in a range of 0.93 to 1.03 for composites fabricated by different printing cycles, indicating a moderately strong linear relationship between resistance change and strain change. The good correlation between the empirical equation and experimental data allows us to deduce the relative resistance change according to the strain. It also validates that the tunneling mechanism is the dominant sensing mechanism at low strain, and crack initiation and propagation is the dominant one at high strain.

9



Figure 5. (a) Schematic illustrations of the proposed structural change of the CNT network under applied strain. SEM images of the surface of CNT network under (b) 0% (c) 5% (d) 40% strain (a) 0.3

(b) 20 18 Experimental

0.25

Experimental

16

Theoretical

Theoretical

14

0.2 R/R0

 R/R0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 22

0.15

12 10

0.1

8 6 4

0.05

2 0

0 0

1

2 3 Strain(%)

4

5

5

15

25 Strain(%)

35

45

Figure 6. Comparison between theoretical prediction and experimental test results of resistance change– strain relationship for CS50 sensors under (a) less than 5% strain (b)5 to 45% strain

In particular, the sensitivity of the fabricated sensors could be modulated by simply tailoring the CNT printing cycles to accommodate diverse sensing requirements. As shown in Figure 4(a), the resistance of composite strain sensors with more printed CNT layers increased significantly, indicating that the sensors with more CNT layers were more sensitive to applied strain. It should be noted that our result is somewhat different from the previous reports, in which the sensitivity increases as the piezo-resistive filler concentration decreases and the change in percolation network state is stated as sensing mechanism8,

39

. In our study, the fabricated sensor has a

sandwich structure, and the CNT concentration in the middle layer functions as the sensing part is much higher than the percolation value. Under high strain, the cracks initiate and propagate, and some CNT networks still bridge the cracks, so there should still be a significant number of CNT joints to provide electrical conductivity. Thus, a crack related film breakage mechanism is proposed to explain the sensing behavior of our fabricated sensors. At un-stretched state, strain sensors with thicker CNT layer have a higher density CNT network and more CNT joints, thus the contact resistance between CNTs dominantly determines the overall resistance. When the strain was applied, the cracking occurred in the CNT network resulting in the disconnection of most CNT joints so that the connection mode transfers to the tunneling state where a large increase of electrical resistance exists. However, for sensors with thinner CNT layer, the CNT

10


Page 11 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


joint is few at original state and the conducting mechanism transition is less visible so that they are insensitive to external stimulus. Generally, the gauge factor [Κ = (∆R/R0 ) / ε]was calculated to determine the sensitivity of composite strain sensors

40

, where R0 is the resistance of carbon nanotube strain sensor without

strain, ∆R is the resistance change under strain, εis the strain and K is the gauge factor. For all the sensors, the electrical resistance increased almost linearly with the strain (R2>0.98), and the gauge factors are almost constant after the initiation stage (Figure 4 inset). This high linearity is of vital importance in precise sensing of strains, and the gauge factor of all sensors are plotted in Figure 4(b), suggesting that the desired controllability of gauge factors, which is beneficial to diverse applications. The gauge factor obtained in this study is 35.75 of the CP50 sensors, which is comparable to the value of most recently reported composite strain sensors (Table S1). Moreover, the CNT/PDMS strain sensors were capable of measuring tensile strain up to 45%, which is more desirable for monitoring human motions compared to commercially available metallic strain gauges. Generally, the joint movement of human body results in less than 50% strain9, thus the fabricated composite strain sensors hold a great promise for monitoring human motions. To examine the response repeatability and durability of the strain sensors, repeated stretching and releasing cycles were applied to the strain sensors. Figure 7(a) shows the resistance change of CP50 sensor at the first stretching/releasing response. It demonstrated that the sensors exhibited a fully recoverable electrical resistance with negligible hysteresis upon the releasing of strain. This could be attributed to the strong interfacial bonding between the CNTs and PDMS in composite sensors, where the polymer matrix was tightly trapped in the dense CNT network. As shown in Figure 7(b), sensors under three different strain all present excellent resistance recoverability and reproducibility. Figure 7(c) shows the long-time piezo-resistive behavior of CP50 composite strain sensor during 1000 stretching-releasing cycles. The electrical resistance change at 45% strain maintained constantly at the beginning couples of cycles, but over 650 cycles, the electrical resistance change starts to show a slight increase gradually. This could be interpreted by the permanent damage of the CNT conductive networks. After a large number of loading-unloading cycles, the cracks in the network propagate resulting in the unrecoverable resistance change. Figure 7(d) represents the resistance change at 45% strain plotted as a function of the loading cycle numbers and the drift of resistance change is only less than 20% 11



after 1000 cycles, indicating a good durability for practical application. In terms of linearity, sensitivity, hysteresis, response time and durability, the printed sensors exhibited excellent comprehensive performance compared to the previously reported CNT/PDMS sensors (Table S1).

Figure 7. Piezoresistive response of CP50 (a) at first stretching/releasing cycle (b) under different strains (c) during 1000 loading-unloading cycles; (d) at 45% strain plotted as a function of the loading cycle numbers

In practical application of the strain sensor, environmental factors such as temperature should be taken into consideration. Hence, the dependence of piezoresistive behavior of composite strain sensors on temperature was studied. Figure 8 shows that the resistance increases sharply as temperature increases at low temperature (less than 50˚C), indicating a positive temperature coefficient trend. This could be explained by the difference between thermal expansion coefficient of PDMS and CNTs (for PDMS: ~3.2x10-4 K-1, for CNT: ~1 to 2x10-5 K-1)41-42. When temperature increases, PDMS expands resulting in larger distance between conductive CNTs so that the resistance increases. Interestingly, at higher temperature, the electrical resistance remains 12


Page 12 of 22

Page 13 of 22

unchanged, and even slightly decreases as temperature increases. This is possibly because CNTs themselves usually exhibit a negative thermal coefficient behavior (the resistivity decreases as temperature increases)43. As temperature increases, the electrons at inner metallic shells of CNTs could be thermally activated to participate electric conductance. Thus, at higher temperature, there is a competition between the resistance reduction results from expansion of CNT networks and the resistance increase results from the intrinsic resistivity decrease of CNTs, which leads to a slightly changed electrical resistance. As shown in Figure 8(inset), the gauge factor decreases when temperature increases. The most possible reason proposed for this is the weaker binding resulting from the different thermal expansion ratio between PDMS and CNTs. When the binding is weaker, load transfer between PDMS and CNT network is inefficient so that the sensors turn to be less sensitive to the applied strain38. In addition, the un-stretched resistance of strain sensors increases a lot when temperature increases, thus compared with this original value, the resistance change is smaller and a lower gauge factor is achieved. The change of gauge factor is as low as 0.17%/˚C. Therefore, the temperature will not have significant effects on sensitivity when measuring strains using the fabricated strain sensors. 1.40 1.35 38

1.30

37 Gauge Factor

Resistance(kΩ)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1.25

36 35 34 33

1.20

32 10 20 30 40 50 60 70 80 90 100 110 Temperature(°C)

1.15 10

20

30

40 50 60 70 80 Temperature(°C)

90 100 110

Figure 8. Dependence of piezoresistive behavior of composite strain sensors on temperature

The fabricated strain sensors can be used for wearable electronics applications, because they are flexible to be easily attached to human body parts. To examine their capability of monitoring human motion with mechanical robustness and electrical stability, the CP50 strain sensors were attached to a human wrist to monitor physiological range of motions. To best of our knowledge, 13



so far, all sensors only demonstrate the capability of showing resistance change when human body moves, but the accuracy of quantifying skin stretch or strain was rarely reported. In this study, the strain data obtained from the correlated piezoresistive signal was compared with strain data calculated from measuring the 3D position of reflective markers attached to the strain sensor during wrist flexion movements using a motion capture. As shown in Figure 9(a), the straight neutral wrist was considered as the initial state. Three motions at different flexion angles were carried out to evaluate the sensing performance of the composite strain sensor. Figure 9(b) shows the location differences between markers during wrist flexion. Strains during wrist movements were calculated from the changes of distance between markers from the initial state to the bent state in the flexed position. The calculated strains for different ranges of wrist flexion (25°, 40°, and 50°) are shown in Figure 9(c). During the wrist movements, the strain sensors recorded specific resistance signal changes (Figure 9(d)) for each motion. The uniform signal changes were observed when the similar ranges of motion were repeated, which represents a good reliability of fabricated sensors for monitoring motions. By using the gauge factors obtained from previous tensile tests, maximum trains for three motions were calculated as 5.49%, 13.39% and 23.65%, respectively. The strains obtained from strain sensors were slightly larger than the strain obtained from the marker positions, as shown in Figure 9(e). This could be because the motion capture based strain was calculated from the linear distance between two markers, but the strain sensors were curved during the wrist movements. The actual curved strain should be larger than the strain calculated from the change in linear distance. Thus, there is a slightly difference between the motion capture recorded strain and strain sensor-recorded strain. Figure 9(f) shows the maximum strains recorded by motion capture and strain sensors under diverse ranges of flexion, and the difference is less than 20%, indicating that strain sensors could be used to accurately quantify the amount of strain during human movement.

14


Page 14 of 22

Page 15 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 9. (a) Photographs of the unidirectional strain sensor attached to the tester’s wrist with three ranges of flexion; (b) location differences between markers when tester’s wrist moves (c) Strain profile obtain from motion capture; (d) electrical resistance signal changes of strain sensor; (e) Comparison of strain obtained by motion capture with strain obtained by strain sensor in first flexion and extension cycle (f) maximum strains recorded by motion capture and strain sensors in first flexion and extension cycle under a range of flexion angles 15



Human body deformation is generally multidirectional, which cannot be accurately monitored by single-directional strain sensors. Due to flexibility of the SDM fabrication, multi-axial strain sensors were fabricated to monitor complicated strain. A rosette-shaped strain sensor was fabricated with an intersecting angle of 120°. By using this type of strain sensor, the magnitude and direction of the principal strain could be measured. As shown in Figure 10(a), the fabricated rosette-shaped strain sensor was mounted on the tester’s wrist. Figure 10(b) shows the strain recorded by the motion capture of each individual strain sensors during the cyclic flexion and extension motions of the wrist. Figure 10(c) shows the corresponding electrical resistance changes of each individual sensors. Maximum strains calculated from the signal changes of three individual sensors were 6.29%, 2.45% and 0.82%, respectively, which is only slightly different from the strain recorded by motion capture, indicating the accuracy of the strain sensors to monitor human movement. In particular, the strain sensor panel that was placed on top of and in the same direction as the wrist extensors is larger than the angled sensor panels, which is reasonable considering that our testing motion was wrist flexion and extension movements. Additionally, using this multi-direction strain sensors, the strain along other direction could be identified. According to the defined coordinated system, the principal strain and their angle of orientation was calculated to be 6.43% and 8.4° by solving the strain transformation equations (Supporting information, equation 12-17) using strains obtained from three sensors in the rosette configuration. It indicates that during wrist flexion, the direction of the skin stretch is not completely identical with wrist flexion due to the complexity of human motion. When this rosette-type strain sensor is subjected to any strain state, the principal strain and its orientation could be calculated through the strain data obtained from the three independent strain, enabling to monitor diverse human motions.

16


Page 16 of 22

Page 17 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 10. (a) Photographs of the rosette-type strain sensor attached to tester’s wrist (b) strain recorded

by camera of each individual strain sensors (c) electrical resistance changes of each individual sensors 4. Conclusions In summary, CNT/PDMS strain sensors were fabricated by printing a CNT dispersion on PDMS substrate using SDM device, which is a digitally-controlled, cost-effective, additive and continuous process. The fabricated strain sensors show a high stretchablity (~45%) and high sensitivity (~35 gauge factor) as well as good linear relationship between the applied strain and electrical resistance change. Furthermore, the CNT/PDMS strain sensor showed small hysteresis and excellent durability under cyclic loading-unloading. The printing cycles of CNT layer could be controlled to tune strain gauge factor according to the requirement. It was also demonstrated that using the digital printing technique, unidirectional or multidirectional strain sensors could be manufactured to monitor complicated deformation of human motion. Besides, the strains obtained from CNT/PDMS sensors were compared with strain calculated from motion capture data, and only less than 20% deviation was found, demonstrating the great accuracy of fabricated sensors to quantify the amount of strain. The rosette type strain sensors are also highly applicable to monitor diverse human motion due to their sensing capability of principal strain and its 17



direction. The facile fabrication of this high-performance layered CNT/PDMS strain sensors, combining their high sensitivity, stretchability and accuracy, paves a new way for the fabrication of wearable strain sensors for monitoring human motions.

Associated Content Supporting Information Photograph of CNT suspension; Sheet resistance of CNT layers printed with different numbers of cycles; Contact angle of water on PDMS substrates before and after surface treatment; Mechanism of hydrophilicity improvement for PDMS through acid treatment; Cross-section SEM images of pristine CNT layers on PDMS substrate; Detailed explanation for equations of proposed sensing models. Table of the performance of strain sensors reported in the previous literature; Motion capture system set up; The equations for calculating the principal strain and its orientation in the rosette-type strain sensor.

Acknowledgements No funding to declare

References (1) Son, D.; Lee, J.; Qiao, S.; Ghaffari, R.; Kim, J.; Lee, J. E.; Song, C.; Kim, S. J.; Lee, D. J.; Jun, S. W. Multifunctional Wearable Devices for Diagnosis and Therapy of Movement Disorders. Nat. Nanotechnol. 2014, 9 (5), 397-404. (2) Park, S.; Vosguerichian, M.; Bao, Z. A Review of Fabrication and Applications of Carbon Nanotube Film-Based Flexible Electronics. Nanoscale 2013, 5 (5), 1727-1752. (3) Jung, S.; Kim, J. H.; Kim, J.; Choi, S.; Lee, J.; Park, I.; Hyeon, T.; Kim, D. H. Reverse‐ Micelle‐Induced Porous Pressure‐Sensitive Rubber for Wearable Human–Machine Interfaces. Adv. Mater. 2014, 26 (28), 4825-4830. (4) Yamada, T.; Hayamizu, Y.; Yamamoto, Y.; Yomogida, Y.; Izadi-Najafabadi, A.; Futaba, D. N.; Hata, K. A Stretchable Carbon Nanotube Strain Sensor for Human-Motion Detection. Nat. Nanotechnol. 2011, 6 (5), 296-301. (5) Shabahang, S.; Tao, G.; Kaufman, J. J.; Qiao, Y.; Wei, L.; Bouchenot, T.; Gordon, A. P.; Fink, Y.; Bai, Y.; Hoy, R. S. Controlled Fragmentation of Multimaterial Fibres and Films Via Polymer Cold-Drawing. Nature 2016, 534, 529-533. (6) Lu, H.; Wang, X.; Yao, Y.; Gou, J.; Hui, D.; Xu, B.; Fu, Y. Synergistic Effect of Siloxane Modified Aluminum Nanopowders and Carbon Fiber on Electrothermal Efficiency of Polymeric Shape Memory Nanocomposite. Composites, Part B 2015, 80, 1-6. (7) Qiao, Y.; Liu, J.; Jia, Y.; Xu, C.; An, L.; Bai, Y. Study on Coexistence of Brittle and Ductile Fractures in Nano Reinforcement Composites under Different Loading Conditions. Int. J. Fract. 2016, 204, 1-20. 18


Page 18 of 22

Page 19 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(8) Zhao, H.; Bai, J. Highly Sensitive Piezo-Resistive Graphite Nanoplatelet–Carbon Nanotube Hybrids/Polydimethylsilicone Composites with Improved Conductive Network Construction. ACS Appl. Mater. Interfaces 2015, 7 (18), 9652-9659. (9) Tang, Y.; Zhao, Z.; Hu, H.; Liu, Y.; Wang, X.; Zhou, S.; Qiu, J. Highly Stretchable and Ultrasensitive Strain Sensor Based on Reduced Graphene Oxide Microtubes–Elastomer Composite. ACS Appl. Mater. Interfaces 2015, 7 (49), 27432-27439. (10) Amjadi, M.; Pichitpajongkit, A.; Lee, S.; Ryu, S.; Park, I. Highly Stretchable and Sensitive Strain Sensor Based on Silver Nanowire–Elastomer Nanocomposite. ACS nano 2014, 8 (5), 5154-5163. (11) Tian, H.; Shu, Y.; Cui, Y.-L.; Mi, W.-T.; Yang, Y.; Xie, D.; Ren, T.-L. Scalable Fabrication of High-Performance and Flexible Graphene Strain Sensors. Nanoscale 2014, 6 (2), 699-705. (12) Wang, Y.; Wang, L.; Yang, T.; Li, X.; Zang, X.; Zhu, M.; Wang, K.; Wu, D.; Zhu, H. Wearable and Highly Sensitive Graphene Strain Sensors for Human Motion Monitoring. Adv. Funct. Mater. 2014, 24 (29), 4666-4670. (13) Park, J. J.; Hyun, W. J.; Mun, S. C.; Park, Y. T.; Park, O. O. Highly Stretchable and Wearable Graphene Strain Sensors with Controllable Sensitivity for Human Motion Monitoring. ACS Appl. Mater. Interfaces 2015, 7 (11), 6317-6324. (14) Chen, S.; Wei, Y.; Wei, S.; Lin, Y.; Liu, L. Ultrasensitive Cracking-Assisted Strain Sensors Based on Silver Nanowires/Graphene Hybrid Particles. ACS Appl. Mater. Interfaces 2016, 8 (38), 25563-25570. (15) Yin, B.; Wen, Y.; Hong, T.; Xie, Z.; Yuan, G.; Ji, Q.; Jia, H. Highly Stretchable, Ultrasensitive, and Wearable Strain Sensors Based on Facilely Prepared Reduced Graphene Oxide Woven Fabrics in an Ethanol Flame. ACS Appl. Mater. Interfaces 2017, 9 (37), 3205432064. (16) Lipomi, D. J.; Vosgueritchian, M.; Tee, B. C.; Hellstrom, S. L.; Lee, J. A.; Fox, C. H.; Bao, Z. Skin-Like Pressure and Strain Sensors Based on Transparent Elastic Films of Carbon Nanotubes. Nat. Nanotechnol. 2011, 6 (12), 788-792. (17) Michelis, F.; Bodelot, L.; Bonnassieux, Y.; Lebental, B. Highly Reproducible, Hysteresis-Free, Flexible Strain Sensors by Inkjet Printing of Carbon Nanotubes. Carbon 2015, 95, 1020-1026. (18) De Volder, M. F.; Tawfick, S. H.; Baughman, R. H.; Hart, A. J. Carbon Nanotubes: Present and Future Commercial Applications. Science 2013, 339 (6119), 535-539. (19) Wood, J. R.; Zhao, Q.; Frogley, M. D.; Meurs, E. R.; Prins, A. D.; Peijs, T.; Dunstan, D. J.; Wagner, H. D. Carbon Nanotubes: From Molecular to Macroscopic Sensors. Phys. Rev. B 2000, 62 (11), 7571. (20) Liu, H.; Gao, J.; Huang, W.; Dai, K.; Zheng, G.; Liu, C.; Shen, C.; Guo, J.; Yan, X.; Guo, Z. Electrically Conductive Strain Sensing Polyurethane Nanocomposites with Synergistic Carbon Nanotubes and Graphene Bifillers. Nanoscale 2016, 8 (26), 12977-12989. (21) Zhou, J.; Yu, H.; Xu, X.; Han, F.; Lubineau, G. Ultrasensitive, Stretchable Strain Sensors Based on Fragmented Carbon Nanotube Papers. ACS Appl. Mater. Interfaces 2017, 9 (5), 48354842. (22) Chizari, K.; Daoud, M. A.; Ravindran, A. R.; Therriault, D. 3d Printing of Highly Conductive Nanocomposites for the Functional Optimization of Liquid Sensors. Small 2016, 12 (44), 6076-6082.

19



(23) Kim, K.; Park, J.; Suh, J.-h.; Kim, M.; Jeong, Y.; Park, I. 3d Printing of Multiaxial Force Sensors Using Carbon Nanotube (Cnt)/Thermoplastic Polyurethane (Tpu) Filaments. Sens. Actuators, A 2017, 263, 493-500. (24) Wang, X.; Jiang, M.; Zhou, Z.; Gou, J.; Hui, D. 3d Printing of Polymer Matrix Composites: A Review and Prospective. Composites, Part B 2017, 110, 442-458. (25) Wang, X.; Xu, X.; Zhou, Z.; Gou, J. Shape Memory Polymer Composite Coatings with Enhanced Mechanical and Antimicrobial Properties. Pigm. Resin Technol. 2018, 47 (1), 63-71. (26) Wang, X.; Sparkman, J.; Gou, J. Electrical Actuation and Shape Memory Behavior of Polyurethane Composites Incorporated with Printed Carbon Nanotube Layers. Compos. Sci. Technol. 2017, 141, 8-15. (27) Maji, D.; Lahiri, S.; Das, S. Study of Hydrophilicity and Stability of Chemically Modified Pdms Surface Using Piranha and Koh Solution. Surf. Interface Anal. 2012, 44 (1), 6269. (28) Koh, K.-S.; Chin, J.; Chia, J.; Chiang, C.-L. Quantitative Studies on Pdms-Pdms Interface Bonding with Piranha Solution and Its Swelling Effect. Micromachines 2012, 3 (2), 427-441. (29) Shimoni, A.; Azoubel, S.; Magdassi, S. Inkjet Printing of Flexible High-Performance Carbon Nanotube Transparent Conductive Films by “Coffee Ring Effect”. Nanoscale 2014, 6 (19), 11084-11089. (30) Kim, T.; Song, H.; Ha, J.; Kim, S.; Kim, D.; Chung, S.; Lee, J.; Hong, Y. Inkjet-Printed Stretchable Single-Walled Carbon Nanotube Electrodes with Excellent Mechanical Properties. Appl. Phys. Lett. 2014, 104 (11), 113103. (31) Lee, H.; Seong, B.; Moon, H.; Byun, D. Directly Printed Stretchable Strain Sensor Based on Ring and Diamond Shaped Silver Nanowire Electrodes. RSC Adv. 2015, 5 (36), 28379-28384. (32) Muth, J. T.; Vogt, D. M.; Truby, R. L.; Mengüç, Y.; Kolesky, D. B.; Wood, R. J.; Lewis, J. A. Embedded 3d Printing of Strain Sensors within Highly Stretchable Elastomers. Adv. Mater. 2014, 26 (36), 6307-6312. (33) Gonçalves, B.; Oliveira, J.; Costa, P.; Correia, V.; Martins, P.; Botelho, G.; LancerosMendez, S. Development of Water-Based Printable Piezoresistive Sensors for Large Strain Applications. Composites, Part B 2017, 112, 344-352. (34) Agarwala, S.; Goh, G. L.; Yap, Y. L.; Goh, G. D.; Yu, H.; Yeong, W. Y.; Tran, T. Development of Bendable Strain Sensor with Embedded Microchannels Using 3d Printing. Sens. Actuators, A 2017, 263, 593-599. (35) Giffney, T.; Bejanin, E.; Kurian, A. S.; Travas-Sejdic, J.; Aw, K. Highly Stretchable Printed Strain Sensors Using Multi-Walled Carbon Nanotube/Silicone Rubber Composites. Sens. Actuators, A 2017, 259, 44-49. (36) Borghetti, M.; Serpelloni, M.; Sardini, E.; Pandini, S. Mechanical Behavior of Strain Sensors Based on Pedot: Pss and Silver Nanoparticles Inks Deposited on Polymer Substrate by Inkjet Printing. Sens. Actuators, A 2016, 243, 71-80. (37) Christ, J. F.; Aliheidari, N.; Ameli, A.; Pötschke, P. 3d Printed Highly Elastic Strain Sensors of Multiwalled Carbon Nanotube/Thermoplastic Polyurethane Nanocomposites. Mater. Des. 2017, 131, 394-401. (38) Amjadi, M.; Kyung, K. U.; Park, I.; Sitti, M. Stretchable, Skin‐Mountable, and Wearable Strain Sensors and Their Potential Applications: A Review. Adv. Funct. Mater. 2016, 26 (11), 1678-1698. (39) Roh, E.; Hwang, B.-U.; Kim, D.; Kim, B.-Y.; Lee, N.-E. Stretchable, Transparent, Ultrasensitive, and Patchable Strain Sensor for Human–Machine Interfaces Comprising a Nanohybrid of Carbon Nanotubes and Conductive Elastomers. ACS nano 2015, 9 (6), 6252-6261. 20


Page 20 of 22

Page 21 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(40) Lu, N.; Lu, C.; Yang, S.; Rogers, J. Highly Sensitive Skin‐Mountable Strain Gauges Based Entirely on Elastomers. Adv. Funct. Mater. 2012, 22 (19), 4044-4050. (41) Schubert, B. E.; Floreano, D. Variable Stiffness Material Based on Rigid Low-MeltingPoint-Alloy Microstructures Embedded in Soft Poly (Dimethylsiloxane)(Pdms). RSC Adv. 2013, 3 (46), 24671-24679. (42) Wu, F.; Cheng, H. Structure and Thermal Expansion of Multi-Walled Carbon Nanotubes before and after High Temperature Treatment. J. Phys. D: Appl. Phys. 2005, 38 (24), 4302. (43) Barberio, M.; Camarca, M.; Barone, P.; Bonanno, A.; Oliva, A.; Xu, F. Electric Resistivity of Multi-Walled Carbon Nanotubes at High Temperatures. Surf. Sci. 2007, 601 (13), 2814-2818.

21



For Table of Contents Only

22


Page 22 of 22

Highly Stretchable and Wearable Strain Sensor ... - ACS Publications

Recommend Documents