See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/318675231

Printable low-cost and flexible carbon nanotube buckypaper motion

sensors

Article  in  Materials & design · July 2017

DOI: 10.1016/j.matdes.2017.07.048

CITATION READS

1 124

6 authors, including:

Joshua Degraff Richard Liang

Florida State University Florida State University
3 PUBLICATIONS   7 CITATIONS    118 PUBLICATIONS   1,559 CITATIONS   

SEE PROFILE SEE PROFILE

Jean-Fabien Capsal
Institut National des Sciences Appliquées de Lyon
47 PUBLICATIONS   331 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Carbon nanotube assembly conductivity mechanism by chemical doping View project

carbon Nanotube Buckypaper Sensor View project

All content following this page was uploaded by Joshua Degraff on 28 December 2017.

The user has requested enhancement of the downloaded file.

 Materials and Design 133 (2017) 47–53

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier.com/locate/matdes

Printable low-cost and flexible carbon nanotube buckypaper

motion sensors
Joshua DeGraff a, Richard Liang a,⁎, Minh Quyen Le b, Jean-Fabien Capsal b,
Florent Ganet b, Pierre-Jean Cottinet b,⁎
a
High Performance Materials Institute (HPMI), Florida State University, Tallahassee, FL, United States
b
Laboratoire de Génie Electrique et Ferroélectricité (LGEF), INSA Lyon, Villeurbanne, France

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Low-profile tactile sensors can be

manufactured using low-cost printing
technology and ultra-thin carbon nano-
tube buckypaper.
• The printed buckypaper sensor can
sense tensile and compressive strains
as low as 0.005% and as high as 1%.
• The printed buckypaper sensors are
flexible and can be designed to conform
to surfaces on the body.
• Finger movements were easily detected
using a glove that was integrated with
sensors.

a r t i c l e i n f o a b s t r a c t

Article history: Wearable technology, which features affordable and flexible sensors integrated into fabrics and garments to de-
Received 4 April 2017 tect both deliberate and subtle body movements, will reshape the way we approach self-rehabilitation, physical
Received in revised form 23 July 2017 training, and many high-dexterity tasks by harvesting data about the wearer's activity. Metallic and semi-
Accepted 24 July 2017
conductor sensors are currently the most commercially viable sensors. Metallic sensors designs are low profile
Available online 25 July 2017
and flexible; however, they are limited by low sensitivity and complex manufacturing. Semi-conductor sensor
Keywords:
designs are highly sensitive but limited by their rigidity and brittle nature. Wearable sensors that are low profile,
Carbon nanotubes flexible, and sensitive to micro-strains are highly desired. We have developed a printable and low profile strain
Buckypaper sensor using multi-wall carbon nanotube thin films called buckypaper (MWCNT-BP). Our tests indicate that
Flexible electronics the buckypaper sensors are 77% more sensitive than similar sensor designs. This paper explains the low-cost
Printed electronics printing technology and displays the sensors' performance after integration into a fabric glove.
Wearable technology © 2017 Elsevier Ltd. All rights reserved.

1. Introduction capabilities could provide concurrent feedback for self-rehabilitation,

intelligent prosthetics, and physical training [4–6]. New human-
Wearable technology will reform the methods of collecting biome- machine interfaces have the potential to significantly improve the qual-
chanical data [1–6]. Smart clothing and garments with tactile sensing ity of life. Tactile sensors must be affordable, low profile and flexible
enough to conform to the arbitrary curves and crevices of the human
⁎ Corresponding authors.
body. They must also provide quick and stable responses to micro-
E-mail addresses: liang@eng.fsu.edu (R. Liang), pierre-jean.cottinet@insa-lyon.fr strains with high sensitivity and low power requirements. Given these
(P.-J. Cottinet). design specifications, micro-strain (b100 μm/m) has remained a

http://dx.doi.org/10.1016/j.matdes.2017.07.048
0264-1275/© 2017 Elsevier Ltd. All rights reserved.
48 J. DeGraff et al. / Materials and Design 133 (2017) 47–53

challenge to the scientific community. Tactile sensors that are based on improvement slows as CNT concentration increases beyond this thresh-
piezoresistivity mechanisms, in which electrical resistance changes old value [21,25–27]. Darabi et al. recently showcased the potential of
under applied strain, are the most common due to their simplicity and carbon nanotubes by dispersing a small amount of CNTs (b10 wt%)
affordability. Metal-foil sensors have been a standard due to their flexi- into chewing gum to make flexible tactile sensors that can detect
bility and low profile; however, they are limited by low sensitivity with sneezes and breathing patterns [1].
a gauge factor (GF) ≈ 2. The resistance variation derives from changes The use of individual carbon nanotubes is showing promise, but the
in the effective length of the sensing element. As a result, they behave sensor designs are limited to the laboratory scale. In addition, sensitivity
as fixed directional sensors that are limited to sensing strain in a specific is considerably low, and manufacturing methods are difficult to repeat
direction. Alternatively, highly sensitive (GF ≈ 200) semi-conducting [10]. Carbon nanotube buckypaper (CNT-BP) films have helped elimi-
sensors made of crystalline silicone (Si) are limited by their rigidity nate process complexity [13,14,23,28,29]. Paper materials, such as
and a restricted working range. Achieving wearable tactile sensing re- CNT-BP, have recently attracted increased research and commercial in-
quires novel approaches in design and structural engineering. terest for next-generation wearable technology due to their lightweight,
Recent research has focused on manufacturing carbon nanotube scalable manufacturing processes, and conformability [30].
(CNT)-based sensors. CNTs demonstrate high tensile strength Buckypaper is a dense, carbon nanotube network that exhibits im-
(~ 100 GPa), Young's modulus (~ 1 TPa), and electrical conductivity proved sensitivity to local distributions of stress compared to individual
(106–107 S/m), and feature a large aspect ratio (N106). The unique com- CNTs. Multi-walled CNT-BP (MWCNT-BP) has a well-distributed, yet
bination of high elastic modulus and impressive electrical properties random structure that provides several degrees of freedom, which is a
promotes excellent sensor characteristics compared to traditional sen- key advantage over metallic foil strain gauges [3]. MWCNT-BP film sen-
sors [1–3,5,7–16]. The structure of carbon nanotubes has a significant sors have displayed a greater working range than single-walled carbon
influence on their electrical properties. Thus, mechanical deformation nanotube buckypaper [10].
alters the structure, which then alters the conductivity of the CNT. Recently, Suzuki et al. designed highly sensitive (GF N 10) buckypaper
This is the essence of the carbon nanotube piezo-resistivity [17]. strain sensors by spin-coating aligned MWCNT buckypaper films with
Researchers have developed CNT-based sensors following a low-modulus polyurethane (PU) resin and an elasticity-assist layer
manufacturing process that involves dispersing CNTs into a viable poly- made of polytetrahydrofuran (PTHF) [14]. The sensors were successfully
mer matrix. Hu et al. performed a detailed investigation on various pro- integrated into fabric gloves to sense a pianist's finger motions. Compat-
cess parameters including different polymers, CNT types, curing ibility with a printing process is highly desirable for next-generation tac-
processes, mixing processes, and types of additives. Hu discovered tile sensors. With the advances of printing technology, printing is simple,
that the process complexity affects how well the CNTs form the conduc- cost-efficient, and capable of large-area and high-throughput production.
tive pathways within the matrix to achieve high sensing performance In this report, we present a highly sensitive, low-cost strain sensor that
[18]. CNT/polymer films follow a percolation-like power law, in which takes advantage of the advances in printed electronics in conjunction
a CNT concentration threshold (wt%) is required to begin cultivating with the low-profile (7 μm), lightweight (5 g/m2) and multi-
the percolation network in the host matrix [19–24]. As concentration functionality of MWCNT buckypaper. This type of affordable, flexible
surpasses the percolation threshold, electrical conductivity in the ma- strain sensor can be deployed in various applications ranging from tactile
trix significantly increases. However, once the network forms, the sensing in wearable technology to structural health monitoring.

Fig. 1. Multi-wall carbon nanotube buckypaper; a. continuous roll of MWCNT buckypaper, b. scanning electron microscopy images at 100× and 100,000× magnification.
 J. DeGraff et al. / Materials and Design 133 (2017) 47–53 49

2. Materials and methods between the sensing element and polymer is critical. Laminating the
structure ensures interfacial binding between the materials to facilitate
2.1. Multi-walled carbon nanotube buckypaper fabrication fast responses and cyclic performance with limited hysteresis [6,10,33].
Fig. 2 presents the manufacturing process, which is low-cost and
The fabrication process for MWCNT-BP first involves creating sus- scalable given the simplicity and commercial availability on ink-jet
pension of uniformly dispersed CNTs through sonication. The fabrica- printing technology. Silver ink electrodes were printed on a thin poly-
tion procedure has been documented in detail by Smalley [31] and ethylene terephthalate (PET) substrate using a one-pass printer. Com-
Wang [32]; however, Triton-X surfactant has been added to the suspen- mercially available, water-based silver ink from NovaCentrix was used
sion to further improve the CNT dispersions by reducing the surface en- to print the electrodes. The composition included 25 wt% silver particles
ergies between CNT bundles. Highly dispersed CNTs promote high (d ≈ 60–80 nm), 1–15 wt% ethylene glycol, and 60–75 wt% water. The
quality films in terms of both mechanical and electrical properties. ink was emission-free and dries quickly. Once the silver electrodes
After dispersing the nanotubes, the suspension is then filtrated onto a were printed, a buckypaper strip was positioned between the sub-
substrate through a mesh filter and dried. The resultant free-standing strate's printed contacts and laminated.
films are repeatedly washed with distilled water and heated at 850 °C The buckypaper is considerably thin and avoids influencing wavi-
under argon gas for 4 h to remove residual surfactant and impurities. ness in the final laminated structure. The printed electrodes provide
Fig. 1 presents a 2 m-long continuous roll of buckypaper along with each end of the buckypaper strips with a fixed electrical contact. The
scanning electron microscopy (SEM) images. Buckypaper consists of closely paired ends of the printed silver electrodes were crimped with
tightly packed CNT networks compared to solvent-cast CNT/polymer male connectors using a Nicomatic CrimpFlex tool for electrical con-
sensors. The dense and conductive network of buckypaper provides a tacts. Three fundamental principles define this manufacturing strategy.
more sensitive response to strain changes. The MWCNT-BPs used in The first exploits an observation in basic mechanics. Polymers that are
this research were 6 μm thick, which exhibited an electrical conductivity sufficiently thin will also be flexible. Fig. 3 displays the sensor design's
of 200 S/cm and a possessed an elastic modulus of almost 3GPa. low profile and flexibility. The second idea is based on the
piezoresistivity response to stress in buckypaper [7,11,19]. A dense net-
2.2. Ink-jet printed sensor design work of conductive CNTs promotes a seamless flow of current; however,
under tensile stress, the network stretches and becomes less dense. This
The sensor structure includes a strip of buckypaper (20 mm × 3 mm reduces the number of percolative pathways and ultimately reduces
× 0.006 mm) and a PET substrate with printed circuitry on its surface. conductivity. Fig. 4 illustrates the reversible changes in the CNT network
The structure was laminated with a film to protect the sensor compo- due to both tensile and compression strains. As the network stretches,
nents and hold the buckypaper strips in position. The interaction the number of conductive paths reduces. This leads to limited tunneling

Fig. 2. Manufacturing process flow for buckypaper strain sensors. It includes printing, laminating, and cutting.
50 J. DeGraff et al. / Materials and Design 133 (2017) 47–53

3. Results and discussion

Buckypaper has an electrical resistance that encompasses three

major components: 1) the intrinsic resistance (RIndividual) of the individ-
ual CNTs, 2) the contact resistance (RContact) between the CNTs, and
3) the tunneling resistance (RTunneling) between the neighboring CNTs
[7,10,28]. The increasing number of gaps in the network causes an in-
crease in electrical resistance [7,28]. As percolation pathways are
strained and disconnected, less current can flow through the printed
sensor altering the conductivity of the structure [7,11,27,28].

RBP ¼ RIndividual þ RContact þ RTunneling ð1Þ

Individual CNTs have an elastic modulus that approaches 1 TPa. This

high stiffness indicates that their contributions (RIndividual) to the global
piezoresistivity are negligible in motion sensing. In addition, the
intertube resistance is much greater compared to the individual resis-
tance. The weak interactions at the nanotubes' joints influence interfa-
cial sliding and dictate piezoresistivity. As the nanotubes separate
from each other, the network becomes less conductive. Therefore,
RContact and RTunneling dictate global piezoresistivity [28]. To determine
the gauge factor, the buckypaper sensors were connected to a Wheat-
stone Quarter-Bridge circuit, as shown in Fig. 5. A small voltage (Vin
= 3 V) was applied to the circuit, and the voltage (Vout) across the
Fig. 3. BP strain sensors offer a low-profile and flexible design. buckypaper sensor was monitored. As presented in Eq. (2), Vout can be
expressed in terms of the resistors in the circuit (Rx) and Vin [35].

effects and increased resistance [19,23,28]. As the network compresses, RBP R4 −R2 R3
Vout ¼ VA −VB ¼ V ð2Þ
the contrary occurs leading to increased conductivity [34]. The third ðRBP þ R2 ÞðR3 þ R4 Þ in
manufacturing principle exploits the recent advances in printed elec-
tronics, more specifically inkjet printing (IJP), which has the advantages R2–4 are known values that were chosen with the purpose of
of short production time, low cost, minimal waste, high spatial resolu- balancing the bridge. The bridge is balanced when Vout is null when
tion, and good reproducibility. the sensor is at rest. Eq. (3) presents the general rule for balancing a
Wheatstone bridge [35].

2.3. Strain gauge characterization RBP R3

¼ ð3Þ
R2 R4
Strain gauge measurements were taken using a DEWE-SIRUS-STG-
Multi card. The printed buckypaper sensors were fastened to a flexible
polyvinyl chloride (PVC) substrate (150 mm × 25 mm × 3 mm) using
an epoxy-based adhesive. The PVC substrate was fixed to a linear actu-
ator (NITEK GDI 350ES500S), and the actuator applied periodic (1 Hz)
tensile strains of 0.4% with highly accurate positional control. To better
access sensor performance, metallic, commercial strain gauges with a
gauge factor of 2.5 were fastened to the PVC substrate. The strain gauges
were characterized using a Wheatstone bridge configuration, and the
measurements were recorded on a DEWESoft data acquisition card.

Fig. 4. Schematic of buckypaper before and after being stretched. The induced gaps in the Fig. 5. Schematic of the Wheatstone bridge characterization setup, where Vin = 3 V and
conductive network increases electrical resistance. RBP ≈ R2 = R3 = R4 to balance the bridge.
 J. DeGraff et al. / Materials and Design 133 (2017) 47–53 51

Fig. 6. BP strain sensor's dynamic response: (a) Measured sensitivity based on a Wheatstone bridge configuration. (b) Gauge factor versus strain comparison.

Eq. (4) expresses the change in the output voltage with respect to can be simplified to determine sensor sensitivity, as expressed in
the induced change in resistance (ΔRBP) due to an applied strain [35]. Eq. (5).

½ðRBP þ ΔRBP ÞR4 −R2 R3  ΔVout ΔRBP

ΔVout ¼ V −0 ð4Þ ¼ ð5Þ
½ðRBP þ ΔRBP Þ þ R2 ÞðR3 þ R4 Þ in Vin 4R

By using equivalent resistors for R2–4, the three resistors can be rep- By solving for ΔRBP, the gauge factor (GF) can be determined using
resented as a single resistance (RBP ≈ R2 = R3 = R4 = R). Now, Eq. (4) Eq. (6). The gauge factor expresses the sensitivity of the piezoresistive

Fig. 7. Comparison of commercial and other CNT-based sensors' gauge factors. The printed BP sensor's gauge factor remains consistent for N10,000 cycles.
52 J. DeGraff et al. / Materials and Design 133 (2017) 47–53

response to strain (Δl

LÞ
Printed buckypaper sensor offers two main advantages. The first is
that the manufacturing methods avoid traditional, low-throughput
ΔRBP methods that have limited carbon nanotube-based devices to the labo-
R ratory scale. After establishing the printing process, various low profile
GF ¼ BP ð6Þ
Δl substrates with high thermal and chemical resistance like polyimide
L (PI) polydimethylsiloxane (PDMS) can be used to improve perfor-
mance, provide transparency and impede device degradation [6]. The
Fig. 6a presents the percent change in resistance as a function of second advantage is that paper materials are lightweight and
strain, and shows good linearity for calibration. A strong, linear relation- deformability [30]. They have become excellent candidates in tactile
ship with respect to both extension and compression modes resulted in sensing. Sensing performance can be enhanced by improving the
a large strain window [− 0.4%, 0.4%]. The strain range was also rather Young's modulus of MWCNT buckypaper [10].
large and symmetric, as the normalized resistance change reached
6.23 × 10−3 under 380 μm/m deformation in extension mode and 4. Conclusion
reached 6.22 × 10−3 under 380 μm/m in compression mode. Fig. 6b dis-
plays steady behavior at various levels of strain, as the gauge factor re- The integration of a carbon nanotube buckypaper with affordable
mains constant up to 0.1% strain. printing technology has enabled the manufacturing of flexible tactile
In contrast, the metallic semiconductor gauge strain range reached sensors. The results demonstrated significant improvement in response
±0.05%. Most notably, the gauge factor of the printed buckypaper sen- sensitivity in comparison to commercial metallic strain gauges. The sen-
sor was eight times higher than the metallic gauge GF = (2.5). Note that sors can detect micro-strains and the responses display good linearity in
gauge factor of the printed buckypaper sensor was low compared to both extension and compression modes, which makes for easy calibra-
classical semiconductor strain gauge (50 b GF b 200). However, semi- tion. The sensors were fastened to human fingers to display their ability
conductor strain gauges are fragile, brittle, expensive, temperature- to collect biomechanical data with high repeatability. The advances in
dependent, and possess a limited strain range. The buckypaper sensor the materials and processes described here provide several promising
offers much higher sensitivity than metallic sensors while promoting engineering options for printing low profile electronic components
the benefits of printing technology and thin polymer substrates for and integrated systems. Successful outcomes from these efforts have
conformability. Fig. 7 provides a comprehensive comparison among the potential to fundamentally change our conception of electronics,
the presented CNT-BP sensor and recent CNT-based strain sensors [3, from hard, rigid, planar chips to soft, curvilinear sheets that can decode
14,28]. The printed buckypaper sensor exhibits good stability as it main- and store large amounts of biological information. Printable tactile sen-
tains the same gauge factor for N10,000 cycles (at 1 Hz) of strain. The sors like the buckypaper design will continue to strengthen progress in
sensors can detect strains as low as 0.005% within a 10 ms period. The human/machine interfacing, soft-robotics, and biomedicine.
sensors were fastened to a latex glove using an adhesive to demonstrate
their ability to detect subtle body movements. Fig. 8 displays the sensor Acknowledgements
output given obvious and subtle finger bending movements. The
buckypaper sensor rapidly and consistently detected both forms of The authors acknowledge support from the Florida Education Fund's
bending. McKnight Doctoral Fellowship program. We also acknowledge Région
Monitoring, analyzing, and quantifying subtle, dynamic motions are Auvergne Rhône-Alpes for their “CMIRA Acceuil Doc” fellowship Pro-
vital for the advancement of human informatics. The motions detected gram. This work was also partially supported by NSF Scalable
by the printed buckypaper sensor were impressively small showing Nanomanufacturing Program project (SNM 1344672). We also thank
promise in measuring critical health data like heartbeat, breathing pat- the High-Performance Materials Institute and Laboratoire de Génie
terns and recognizing the intended movements of amputees [1,36]. Électrique et Ferroélectricité for the collaborating facilities.

Fig. 8. Experimental results from monitoring fingers using BP strain gauge that was fastened to a glove.
 J. DeGraff et al. / Materials and Design 133 (2017) 47–53 53

References [19] S. Gong, Z.H. Zhu, On the mechanism of piezoresistivity of carbon nanotube polymer
composites, Polymer 55 (16) (2014) 4136–4149.
[1] M.A. Darabi, et al., Gum sensor: a stretchable, wearable, and foldable sensor based [20] E. Kymakis, G.A.J. Amaratunga, Electrical properties of single-wall carbon nanotube-
on carbon nanotube/chewing gum membrane, ACS Appl. Mater. Interfaces 7 (47) polymer composite films, J. Appl. Phys. 99 (8) (2006), 084302.
(2015) 26195–26205. [21] C.A. Martin, et al., Formation of percolating networks in multi-wall carbon-
[2] J.D. Shi, D. X, M. Li, Graphene reinforced carbon nanotube networks for wearable nanotube–epoxy composites, Compos. Sci. Technol. 64 (15) (2004) 2309–2316.
strain sensors, Adv. Funct. Mater. 26 (2016) 2078–2084. [22] W. Bauhofer, J.Z. Kovacs, A review and analysis of electrical percolation in carbon
[3] M.D. Rein, O. Breuer, H.D. Wagner, Sensors and sensitivity: carbon nanotube nanotube polymer composites, Compos. Sci. Technol. 69 (10) (2009) 1486–1498.
buckypaper films as strain sensing devices, Compos. Sci. Technol. 71 (3) (2011) [23] B.M. Lee, K.J. Loh, A 2D percolation-based model for characterizing the
373–381. piezoresistivity of carbon nanotube-based films, J. Mater. Sci. 50 (7) (2015)
[4] P. Polygerinos, et al., Soft robotic glove for combined assistance and at-home reha- 2973–2983.
bilitation, Robot. Auton. Syst. 73 (2015) 135–143. [24] J.N. Coleman, et al., Percolation-dominated conductivity in a conjugated-polymer-
[5] W. Long, J.L. Kenneth, Wearable carbon nanotube-based fabric sensors for monitor- carbon-nanotube composite, Phys. Rev. B 58 (12) (1998) R7492.
ing human physiological performance, Smart Mater. Struct. 26 (5) (2017), 055018. [25] G.J. Hu, et al., Low percolation thresholds of electrical conductivity and rheology in
[6] T. Yang, et al., Recent advances in wearable tactile sensors: materials, sensing mech- poly(ethylene terephthalate) through the networks of multi-walled carbon nano-
anisms, and device performance, Mater. Sci. Eng. R. Rep. 115 (2017) 1–37. tubes, Polymer 47 (1) (2006) 480–488.
[7] Alamusi, et al., Piezoresistive strain sensors made from carbon nanotubes based [26] A. Nogales, et al., Low percolation threshold in nanocomposites based on oxidized
polymer nanocomposites, Sensors 11 (11) (2011) 10691. Single Wall carbon nanotubes and poly(butylene terephthalate), Macromolecules
[8] J.R. Bautista-Quijano, et al., Strain sensing capabilities of a piezoresistive MWCNT- 37 (20) (2004) 7669–7672.
polysulfone film, Sensors Actuators A Phys. 159 (2) (2010) 135–140. [27] T.C. Theodosiou, D.A. Saravanos, Numerical investigation of mechanisms affecting
[9] L. Cai, et al., Super-stretchable, transparent carbon nanotube-based capacitive strain the piezoresistive properties of CNT-doped polymers using multi-scale models,
sensors for human motion detection, Sci Rep 3 (2013) 3048. Compos. Sci. Technol. 70 (9) (2010) 1312–1320.
[10] W. Obitayo, T. Liu, A review: carbon nanotube-based piezoresistive strain sensors, J. [28] X. Wang, et al., Tensile strain sensing of buckypaper and buckypaper composites,
Sens. 2012 (2012) 15. Mater. Des. 88 (2015) 414–419.
[11] S. Li, et al., Working mechanisms of strain sensors utilizing aligned carbon nanotube [29] L. Ang, E.B. Alexander, D.B. Philip, Aligned carbon nanotube sheet piezoresistive
network and aerosol jet printed electrodes, Carbon 73 (2014) 303–309. strain sensors, Smart Mater. Struct. 24 (9) (2015), 095004.
[12] K.J. Loh, et al., Carbon nanotube sensing skins for spatial strain and impact damage [30] H. Liu, et al., Paper: a promising material for human-friendly functional wearable
identification, J. Nondestruct. Eval. 28 (1) (2009) 9–25. electronics, Mater. Sci. Eng. R. Rep. 112 (2017) 1–22.
[13] C.-W. Ma, et al., A flexible tactile and shear sensing array fabricated using a novel [31] B.W. Smith, et al., Structural anisotropy of magnetically aligned single wall carbon
buckypaper patterning technique, Sensors Actuators A Phys. 231 (2015) 21–27. nanotube films, Appl. Phys. Lett. 77 (5) (2000) 663–665.
[14] K. Suzuki, et al., Rapid-response, widely stretchable sensor of aligned MWCNT/elas- [32] P. Gonnet, et al., Thermal conductivity of magnetically aligned carbon nanotube
tomer composites for human motion detection, ACS Sens. 1 (6) (2016) 817–825. buckypapers and nanocomposites, Curr. Appl. Phys. 6 (1) (2006) 119–122.
[15] P. Tran Hoang, et al., Engineering crack formation in carbon nanotube-silver nano- [33] H. Deng, et al., Towards tunable resistivity-strain behavior through construction of
particle composite films for sensitive and durable piezoresistive sensors, Nanoscale oriented and selectively distributed conductive networks in conductive polymer
Res. Lett. 11 (2016) 422. composites, J. Mater. Chem. A 2 (26) (2014) 10048–10058.
[16] J. Zhou, et al., Deformable and wearable carbon nanotube microwire-based sensors [34] V. Kumar, et al., Leveraging compressive stresses to attenuate the electrical resistiv-
for ultrasensitive monitoring of strain, pressure and torsion, Nano 9 (2) (2017) ity of buckypaper, Carbon 110 (2016) 62–68.
604–612. [35] C. De Silva, Sensors and Actuators: Control Systems Instrumentation, CRC Press, FL,
[17] M.A. Cullinan, M.L. Culpepper, Carbon nanotubes as piezoresistive 2007.
microelectromechanical sensors: theory and experiment, Phys. Rev. B 82 (11) [36] C. Nissler, N. Mouriki, C. Castellini, Optical myography: detecting finger movements
(2010) 115428. by looking at the forearm, Front. Neurorobotic 10 (3) (2016).
[18] N. Hu, et al., Effect of fabrication process on electrical properties of polymer/multi-
wall carbon nanotube nanocomposites, Compos. A: Appl. Sci. Manuf. 39 (5)
(2008) 893–903.

View publication stats