We fabricated thermally driven metal-free soft-actuators consisting of poly(ethylene terephthalate) (PET) threads as the actuator and carbon nanotube (CNT) yarns as the heating source. The mechanical force, displacement, and response behavior of various structures of the coil-shaped soft-actuators were characterized. The actuation performance of the soft-actuators containing a homogeneous arrangement of PET threads and CNT yarns in their cross-sectional profile was the highest. The results of the calculations based on the heat diffusion equations indicated that inhomogeneous heat generation in the soft-actuator causes parts of the actuator to remain unheated and this interferes with the mechanical motions. Homogeneous thermal distribution in the soft-actuators, namely, the use of a multifilament structure, yields the highest performance in terms of the mechanical force and displacement.

As an alternative type of motor, the motions generated by soft-actuators have attracted considerable interests because they are not only lightweight and inexpensive, but they also realize human-mimetic mechanical motions that are in strong demand in the field of robotics.1,2 Soft-actuators produced from organic materials such as polymer gels,3–6 conductive polymers,7–9 elastomers,10,11 and carbon nanotubes (CNTs),12–14 are devices that convert thermal, electric, or optical energy into complex mechanical motions. Haines et al. reported that coil-shaped soft-actuators fabricated from nylon-6, nylon-6,6, and polyethylene exhibit extremely high mechanical work per unit weight (2 kJ/kg), which is approximately 1,000 times higher than that of the human muscle.15 The motion of a soft-actuator induces reversible deformation of polymers, and soft-actuators with a coil-shaped structure efficiently convert thermally induced entropy into mechanical work. The thermo-mechanical behaviors of nylon-based soft-actuators have been thoroughly investigated in terms of its hysteresis, repeatability, and predictability.16 The availability of nylon-based soft-actuators in a wide range atmospheric temperature has also been demonstrated.17 Hiraoka et al. measured the mechanical force and response behavior of thermally driven actuators fabricated from nylon-6,6, linear low-density polyethylene (LLDPE), and high-density polyethylene (HDPE), and found that actuators fabricated using LLDPE delivered the highest performance.18 

One of the important challenges presented by coil-shaped soft-actuators is in their thermal absorption and extraction design. In a previous study, a coil-shaped polymer thread was reeled with copper wire or CNT sheets to transfer heat to the polymer threads.15 In another study, polymer thread wound with copper wire was covered with flexible silver paste to prevent the transfer of thermal energy from the actuators.18 However, the use of copper wire and silver paste increases the weight of soft-actuators and suppresses their flexibility. In addition, the difference in the thermal expansion coefficients between the polymer-thread and accompanying metal frame can degrade the performance of the soft-actuators. In this study, we fabricated metal-free soft-actuators using poly(ethylene terephthalate) (PET) threads and heat-introducing CNT yarns in various designs. We analyzed the effects of thermal absorbance and extraction on their performance in terms of their mechanical force and response and simulated their respective thermal distributions using heat diffusion equations.

High-density and vertically aligned CNTs (VACNT) were synthesized using the chemical vapor deposition (CVD) method (“Black Magic II,” Aixtron Ltd.).19–25 The catalyst, which was a thin layer of iron, was deposited on the substrate (Al2O3/SiO2/Si) by electron beam deposition (“VTR-350M/ERH,” ULVAC KIKO Inc.). In the CVD equipment, the substrate was annealed at 350°C under an atmosphere of hydrogen gas to remove moisture and reduce the iron film. The iron film was deformed into nanoparticles at temperatures above 500°C and acetylene gas was then introduced as carbon source at the synthesis temperature (700°C). During this process, iron nanoparticles continuously absorbed carbon atoms and recomposed them into tube-shaped carbon molecules. The resulting substrate in which the CNTs are oriented in the vertical direction is known as a CNT array. The individual CNTs are multi-wall CNTs with a wall-number of approximately 3 and an outer tube diameter of approximately 4 nm. The CNT yarn was fabricated from this CNT array by bringing the side of the array into contact with a carbon tape and then withdrawing and rotating it using a servomotor on a slider system.26–31 The rotation speed of the motor was 80 mm/min, while the drawing speed of the slider was 1000 mm/min. The diameter of the CNT yarn was measured as ∼20 μm using scanning electron microscopy (SEM) (“JSM-6060LA,” JEOL). The electrical conductivity of the CNT yarns was measured as 5×104 S/m.

Three types of coil-shaped polymer soft-actuators were fabricated using PET threads as actuators and CNT yarns as heating wires. PET is a suitable polymeric material for this soft-actuator because it has a glass transition temperature above room temperature (∼80°C) and a high melting point (∼260°C). The diameter of the PET fiber was approximately 20 μm and the lengths of the PET threads and CNT yarns were 15 cm. The structures of the soft-actuator and their optical micrographs (“VHX-Z100UR,” KEYENCE) are shown in Fig. 1. The “multifilament structure-A” (Fig. 1(a)) has a homogeneous composition of PET threads and CNT yarns. These fibers were contained randomly in their cross-sectional profile and were rotated until they formed a coil-shaped structure. The “multifilament structure-B” (Fig. 1(b)) is composed of some fine coil-shaped threads. First, two CNT yarns and four PET threads were bundled and rotated; then, a certain number of fine coil-shaped PET/CNT fibers were bundled. The “monofilament structure” (Fig. 1(c)) was fabricated from a bundle of PET threads and a bundle of CNT yarns. This structure is similar to the reported soft-actuator fabricated from one polymer fiber and one heating-wire.15,18

FIG. 1.

Micrographs and schematics of the three types of coil-shaped soft-actuators fabricated from PET threads and CNT yarns. (a) Multifilament structure-A, (b) multifilament structure-B, and (c) monofilament structure. The upper left and lower micrograph in each subfigure show the structure before and after coiling with scale bars. In the schematics (upper right panels), the white and black cylinders indicate the PET threads and CNT yarns, respectively.

FIG. 1.

Micrographs and schematics of the three types of coil-shaped soft-actuators fabricated from PET threads and CNT yarns. (a) Multifilament structure-A, (b) multifilament structure-B, and (c) monofilament structure. The upper left and lower micrograph in each subfigure show the structure before and after coiling with scale bars. In the schematics (upper right panels), the white and black cylinders indicate the PET threads and CNT yarns, respectively.

Close modal

These three types of actuators were connected to a tensile strength measurement machine (“AGS-5NX,” SHIMADZU) and a displacement meter (“LD701-1/2,” OMEGA Engineering Inc.). Electric power (5 mW per one CNT yarn) was applied to measure the actuating force of the samples. The electric power cycle consists of 10 s ON and 10 s OFF and was repeated for 100 cycles to measure the peak mechanical force, activation time constant, and cooling time constant. In the displacement measurements, the applied electric power cycle of 10 s ON and 10 s OFF was repeated for 10 cycles under a load of 70 mN.

Figure 2(a) shows the mechanical force of the soft-actuator of the “multifilament structure-A” as a function of the number ratio of PET threads and CNT yarns obtained by varying the number of PET threads and fixing the number of CNT yarns. The applied electric power was maintained constant at 50 mW (5 mW per CNT yarn) because the number of CNT yarns was fixed as 10. The results in the figure show that the mechanical force is saturated when the number of PET threads exceeds 20. This result suggests that on average, the thermal energy generated in a CNT yarn spreads to only one or two neighboring PET threads. Thus, for the following discussions, we fixed the ratio of PET threads and CNT yarns at 2:1. Figure 2(b) shows the mechanical force of the three types of soft-actuators as a function of the number of PET threads, where the number of CNT yarns was half of the PET threads and the applied electric power was 5 mW per CNT yarn. As shown in Fig. 2(b), the mechanical force linearly increases as the number of PET threads is increased in all types of soft-actuators. The mechanical force generated in multifilament structure-A and -B is 1.6 times higher than that generated in the monofilament structure. Figures 2(c) and (d) show the responsivities (heating and cooling processes) of the soft-actuators as a function of the number of PET threads. The response behavior of the soft-actuator strongly depends on the time constant of the deformation of PET threads; therefore, the response behaviors of the three types of soft-actuators are identical.

FIG. 2.

Mechanical force and response of the soft-actuators. (a) Mechanical force of the soft-actuator (multifilament structure-A) as a function of the number of PET threads with 10 CNT yarns. The red dashed line is intended as a visual guide. Mechanical force (b) and response (c) and (d) of the three soft-actuators as a function of the number of PET threads. The number of CNT yarns is half the number of PET threads. Mechanical force (e) and response (f) of the soft-actuator (multifilament structure-A) as a function of the input electric power. The soft-actuator deforms elastically below electric power input of 220 mW, whereas permanent damage is induced when the electric power reaches 220 mW. The blue dashed line is intended as a visual guide.

FIG. 2.

Mechanical force and response of the soft-actuators. (a) Mechanical force of the soft-actuator (multifilament structure-A) as a function of the number of PET threads with 10 CNT yarns. The red dashed line is intended as a visual guide. Mechanical force (b) and response (c) and (d) of the three soft-actuators as a function of the number of PET threads. The number of CNT yarns is half the number of PET threads. Mechanical force (e) and response (f) of the soft-actuator (multifilament structure-A) as a function of the input electric power. The soft-actuator deforms elastically below electric power input of 220 mW, whereas permanent damage is induced when the electric power reaches 220 mW. The blue dashed line is intended as a visual guide.

Close modal

The dependence of the electric power on the mechanical force and response of the soft-actuator in “multifilament structure-A” (30 PET threads and 15 CNT yarns) are represented in Figs. 2(e) and (f). Initially, the temperature of the soft-actuator rises linearly with the electric power, such that the mechanical force increases as the electric power increases (Fig. 2(e)). Above the glass transition temperature of PET, the mechanical force drops (when the electric power reaches 220 mW), and irreversible deformation is induced. The response of the actuator remains constant in the temperature range at which the reversible deformation occurs (Fig. 2(f)).

Figure 3(a) shows the actuating behavior of three types of soft-actuators at the input electric power of 50 mW. The number of PET threads and CNT yarns in the soft-actuators are fixed to be 20 and 10, respectively. As shown in Fig. 3(a), the soft-actuators in all the structures (multifilament structure-A, -B, and monofilament) yield well-controlled actuation, where the tensile actuation is fitted by the combination of a single exponential growth at 0 s and a single exponential decay at 10 s. The soft-actuators in the multifilament structures show increased torsional and tensile actuation compared to the monofilament structure. Figure 3(b) shows the displacement of the soft actuators as a function of the electric power. The displacement of multifilament structure-A and -B is 1.4–1.8 times higher than that of the monofilament structure at input electric power of 50 mW. This indicates good agreement with the result obtained by measuring the mechanical force.

FIG. 3.

Tensile actuation of the soft-actuators. (a) Actuating behavior of the soft-actuators at applied electric power (50 mW) cycle consisting of 10 s ON and 10 s OFF with 20 PET threads and 10 CNT yarns. The fitting curves are a combination of a single exponential growth and a single exponential decay. (b) Peak tensile actuation as a function of the input electric power.

FIG. 3.

Tensile actuation of the soft-actuators. (a) Actuating behavior of the soft-actuators at applied electric power (50 mW) cycle consisting of 10 s ON and 10 s OFF with 20 PET threads and 10 CNT yarns. The fitting curves are a combination of a single exponential growth and a single exponential decay. (b) Peak tensile actuation as a function of the input electric power.

Close modal

A more in-depth analysis of the thermal effects on the soft-actuator was carried out by constructing a model including both multifilaments and monofilaments, as shown in Figs. 4(a) and (b), respectively. The diffusion equations were then used to calculate (COMSOL Multiphysics®)32 the thermalization process in these multifilament and monofilament structures. These filaments, including 80 PET threads and 41 CNT yarns, were considered to be exposed to atmospheric air at a temperature of 20°C. The thermal energy induced in the filaments was fixed at 3.2 mW/mm. The transient thermal distributions in the multifilament and monofilament structures are shown in Figs. 4(c) and (d) (Multimedia view), respectively, 2 s after inducing thermal energy. Videos of the thermalization processes in the monofilament (Multimedia view) and multifilament (Multimedia view) structures are provided as supplementary material. The multifilament structure is homogeneously thermalized with the input energy (Fig. 4(c) (Multimedia view)); however, the monofilament structure shows inhomogeneous thermal distribution. Figure 4(e) shows the time evolution of the temperature in multifilament and monofilament structures. The temperature of the part of the PET thread farthest from the CNT yarns in the monofilament structure is 10°C lower than that of the closest part (Figs. 4(d) and (e)). This suggests that the mechanical force of the soft-actuator is determined by the low-temperature part of the system. The motions generated in the high-temperature part are subjected to interference by the inertial force in the low-temperature part. This is consistent with the experimental mechanical forces and displacements in the multifilament and monofilament structures. Homogeneous thermal distribution yields the best performance in terms of the mechanical force and displacement. The performance of the multifilament structure-A and -B is expected to be the same. It is also worth mentioning that the required time to attain the state of thermal equilibrium in the multifilament and monofilament structures is identical as shown in Fig. 4(e), which also agrees with the experimental responsivities in all the proposed structures.

FIG. 4.

Simulated thermal distributions in the soft-actuators. Multifilament (a) and monofilament (b) structures used in the model. The white and black cylinders indicate PET threads and CNT yarns, respectively. Simulated thermal distributions in the model structures ((c) (Multimedia view) and (d) (Multimedia view)) 2 s after applying electric power input. The temperature gradient is according to the colored thermo-scales that appear as inset in the figures. (e) Time-evolution of the temperature in the model structures. Multimedia views: https://doi.org/10.1063/1.5033487.1 and https://doi.org/10.1063/1.5033487.2

FIG. 4.

Simulated thermal distributions in the soft-actuators. Multifilament (a) and monofilament (b) structures used in the model. The white and black cylinders indicate PET threads and CNT yarns, respectively. Simulated thermal distributions in the model structures ((c) (Multimedia view) and (d) (Multimedia view)) 2 s after applying electric power input. The temperature gradient is according to the colored thermo-scales that appear as inset in the figures. (e) Time-evolution of the temperature in the model structures. Multimedia views: https://doi.org/10.1063/1.5033487.1 and https://doi.org/10.1063/1.5033487.2

Close modal

In summary, we fabricated different types of coil-shaped soft-actuators using PET threads as the actuator and CNT yarns as the heating source. The mechanical forces, displacements, and responsivities of the soft-actuators were experimentally characterized and the thermodynamics of the soft-actuator models were simulated using diffusion equations. The homogeneous thermal distribution in the soft-actuators, namely, the use of a multifilament structure, yielded the highest performance in terms of the mechanical force and displacement. The response behavior of the soft-actuators depends on the particular organic material used as the actuator.

We would like to thank Prof. Shuji Tsuruoka at Shinshu University for valuable discussions on the study. This work was supported by the Saitama Industrial Promotion Public Corporation and partially supported by the JSPS KAKENHI Grant Numbers 17K20065 and 18H01708.

1.
A.
Katchalsky
and
M.
Zwick
,
J. Polym. Sci.
16
(
82
),
221
234
(
1955
).
2.
Y.
Osada
,
Adv. Polym. Sci.
82
,
1
46
(
1987
).
3.
K.
Oguro
,
Y.
Kawami
, and
H.
Takenaka
,
J. Micromachine. Soc.
5
,
27
30
(
1992
).
4.
T.
Fukushima
,
K.
Asaka
,
A.
Kosaka
, and
T.
Aida
,
Angew. Chem. Int. Ed.
44
(
16
),
2410
2413
(
2005
).
5.
T.
Hirai
,
H.
Nemoto
,
M.
Hirai
, and
S.
Hayashi
,
J. Appl. Polym. Sci.
53
(
1
),
79
84
(
1994
).
6.
T.
Hirai
,
T.
Ogiwara
,
K.
Fujii
,
T.
Ueki
,
K.
Kinoshita
, and
M.
Takasaki
,
Adv. Mater.
21
(
28
),
2886
2888
(
2009
).
7.
K.
Kaneto
,
M.
Kaneko
,
Y.
Min
, and
A. G.
MacDiarmid
,
Synth. Met.
71
,
2211
2212
(
1995
).
8.
W.
Takashima
,
M.
Kaneko
,
K.
Kaneto
, and
A. G.
MacDiarmid
,
Synth. Met.
71
,
2265
2266
(
1995
).
9.
H.
Okuzaki
,
H.
Suzuki
, and
T.
Ito
,
J. Phys. Chem. B
113
(
33
),
11378
11383
(
2009
).
10.
R. E.
Pelrine
,
R. D.
Kornbluh
, and
J. P.
Joseph
,
Sens. Actuators A
64
(
1
),
77
85
(
1998
).
11.
Y.
Yu
,
M.
Nakano
, and
T.
Ikeda
,
Nature
425
,
145
(
2003
).
12.
R. H.
Baughman
,
C.
Cui
,
A. A.
Zakhidov
,
Z.
Iqbal
,
J. N.
Barisci
,
G. M.
Spinks
,
G. G.
Wallace
,
A.
Mazzoldi
,
D.
DeRossi
,
A. G.
Rinzler
,
O.
Jaschinski
,
S.
Roth
, and
M.
Kertesz
,
Science
284
(
5418
),
1340
1344
(
1999
).
13.
A. E.
Aliev
,
J.
Oh
,
E.
Kozlov
,
A. A.
Kunznetsov
,
S.
Fang
,
A. F.
Fonseca
,
R.
Ovalle
,
M. D.
Lima
,
H.
Haque
,
Y. N.
Gartstein
,
M.
Zhang
,
A. A.
Zakhidov
, and
R. H.
Baughman
,
Science
323
(
5921
),
1575
1578
(
2009
).
14.
M. D.
Lima
,
N.
Li
,
M. J.
Andrade
,
S.
Fang
,
J.
Oh
,
G. M.
Spinks
,
M. E.
Kozlov
,
C. S.
Haines
,
D.
Suh
,
J.
Foroughi
,
J.
Kim
,
Y.
Chen
,
T.
Ware
,
M. K.
Shin
,
L. D.
Machado
,
A. F.
Fonseca
,
J. D. W.
Madden
,
W. E.
Voit
,
D. S.
Galvao
, and
R. H.
Baughman
,
Science
338
(
6109
),
928
932
(
2012
).
15.
C. S.
Haines
,
M. D.
Lima
,
N.
Li
,
G. M.
Spinks
,
J.
Foroughi
,
J. D. W.
Madden
,
S. H.
Kim
,
S.
Fang
,
M.
Jung de Andrade
,
F.
Göktepe
,
Ö.
Göktepe
,
S. M.
Mirvakili
,
S.
Naficy
,
X.
Lepró
,
J.
Oh
,
M. E.
Kozlov
,
S. J.
Kim
,
X.
Xu
,
B. J.
Swedlove
,
G. G.
Wallace
, and
R. H.
Baughman
,
Science
343
(
6173
),
868
872
(
2014
).
16.
A.
Cherubini
,
G.
Moretti
,
R.
Vertechy
, and
M.
Fontana
,
AIP Adv.
5
,
067158
(
2015
).
17.
S.
Kianzad
,
M.
Pandit
,
A.
Bahi
,
A.
Rafiee
,
F.
Ko
,
G. M.
Spinks
, and
J. D. W.
Madden
,
Proc. SPIE
9430
,
94301X
(
2015
).
18.
M.
Hiraoka
,
K.
Nakamura
,
H.
Arase
,
K.
Asai
,
Y.
Kaneko
,
S. W.
John
,
K.
Tagashira
, and
A.
Omote
,
Sci. Rep.
6
,
36358
(
2016
).
19.
W. Z.
Li
,
S. S.
Xie
,
L. X.
Qian
,
B. H.
Chang
,
B. S.
Zou
,
W. Y.
Zhou
,
R. A.
Zhao
, and
G.
Wang
,
Science
274
(
5293
),
1701
1703
(
1996
).
20.
I. T.
Han
,
H. J.
Kim
,
Y. J.
Park
,
N.
Lee
,
J. E.
Jang
,
J. W.
Kim
,
J. E.
Jung
, and
J. M.
Kim
,
Appl. Phys. Lett.
81
,
2070
2072
(
2002
).
21.
Z. P.
Yang
,
L.
Ci
,
J. A.
Bur
,
S. Y.
Lin
, and
P. M.
Ajayan
,
Nano Lett.
8
(
2
),
446
451
(
2008
).
22.
K.
Mizuno
,
J.
Ishii
,
H.
Kishida
,
Y.
Hayamizu
,
S.
Yasuda
,
D. N.
Futaba
,
M.
Yumura
, and
K.
Hata
,
Proc. Nat. Am. Sci.
106
(
15
),
6044
6047
(
2009
).
23.
R.
Das
,
M. E.
Ali
,
S. B. A.
Hamid
,
S.
Ramakrishna
, and
Z. Z.
Chowdhury
,
Desalination
336
,
97
109
(
2014
).
24.
H.
Matsumoto
,
S.
Tsuruoka
,
Y.
Hayashi
,
K.
Abe
,
K.
Hata
,
S.
Zhang
,
Y.
Saito
,
M.
Aiba
,
T.
Tokunaga
,
T.
Iijima
,
T.
Hayashi
,
H.
Inoue
, and
G. A. J.
Amaratung
,
Carbon
120
,
358
365
(
2017
).
25.
T.
Iijima
,
H.
Oshima
,
Y.
Hayashi
,
U. B.
Suryavanshi
,
A.
Hayashi
, and
M.
Tanemura
,
Phys. Status Solidi A
208
(
10
),
2332
2334
(
2011
).
26.
K.
Jiang
,
Q.
Li
, and
S.
Fan
,
Nature
419
,
801
(
2002
).
27.
M.
Zhang
,
K. R.
Atkinson
, and
R. H.
Baughman
,
Science
306
(
5700
),
1358
1361
(
2004
).
28.
Y.
Inoue
,
K.
Kakihata
,
Y.
Hirono
,
T.
Horie
,
A.
Ishida
, and
H.
Mimura
,
Appl. Phys. Lett.
92
,
213113
(
2008
).
29.
T.
Iijima
,
H.
Oshima
,
Y.
Hayashi
,
U. B.
Suryavanshi
,
A.
Hayashi
, and
M.
Tanemura
,
Diam. Relat. Mater.
24
,
158
160
(
2012
).
30.
T.
Iijima
,
Y.
Inagaki
,
H.
Oshima
,
T.
Iwata
,
R.
Sato
,
G.
Kalita
,
T.
Kuzumaki
,
Y.
Hayashi
, and
M.
Tanemura
,
Mater. Express
2
(
4
),
357
362
(
2012
).
31.
M.
Scholz
,
Y.
Hayashi
,
V.
Khavrus
,
D.
Chujo
,
H.
Inoue
,
M.
Hada
,
A.
Leonhardt
,
B.
Büchner
, and
S.
Hampel
,
Carbon
133
,
232
238
(
2018
).
32.
COMSOL AB
,
Comsol Multiphysics: Structural Mechanics Module. COMSOL AB
(
Stockholm
,
Sweden
,
2005
).