Embed Size (px)
Citation preview
Thermal, mechanical, and shape-memory properties of nanorubber-toughened, epoxy-based shape-memory nanocomposites
Hong-Qiu Wei,1 Ye Chen,2 Tao Zhang,3 Liwu Liu,4 Jinliang Qiao,5 Yanju Liu ,4 Jinsong Leng1
1Center for Composite Materials and Structures, Harbin Institute of Technology, Harbin 150080, People’s Republic of China2Department of Materials Science and Engineering, Harbin Institute of Technology at Weihai, Weihai 264209, People’s Republic ofChina3School of Mechanical and Automotive Engineering, Kingston University, London SW15 3DW, United Kingdom4Department of Astronautical Science and Mechanics, Harbin Institute of Technology, Harbin 150080, People’s Republic of China5Beijing Research Institute of Chemical Industry, SINOPEC, Beijing 100013, People’s Republic of ChinaCorrespondence to: Y. Liu (E - mail: [email protected])
ABSTRACT: Epoxy-based shape-memory polymers (ESMPs) are a type of the most promising engineering smart polymers. However,
their inherent brittleness limits their applications. Existing modification approaches are either based on complicated chemical reac-
tions or done at the cost of the thermal properties of the ESMPs. In this study, a simple approach was used to fabricate ESMPs with
the aim of improving their overall properties by introducing crosslinked carboxylic nitrile–butadiene nanorubber (CNBNR) into the
ESMP network. The results show that the toughness of the CNBNR–ESMP nanocomposites greatly improved at both room tempera-
ture and the glass-transition temperature (Tg) over that of the pure ESMP. Meanwhile, the increase in the toughness did not nega-
tively affect other macroscopic properties. The CNBNR–ESMP nanocomposites presented improved thermal properties with a Tg in a
stable range around 100 8C, enhanced thermal stabilities, and superior shape-memory performance in terms of the shape-fixing ratio,
shape-recovery ratio, shape-recovery time, and repeatability of shape-memory cycles. The combined property improvements and the
simplicity of the manufacturing process demonstrated that the CNBNR–ESMP nanocomposites are desirable candidates for large-
scale applications in the engineering field as smart structural materials. VC 2017 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2018, 135, 45780.
KEYWORDS: mechanical properties; stimuli-sensitive polymers; thermal properties
Received 30 March 2017; accepted 11 September 2017DOI: 10.1002/app.45780
INTRODUCTION
Shape-memory polymers (SMPs) as a class of smart materials
are capable of returning to their initial shape from a temporary
shape when they are actuated by heat, light, electrical or mag-
netic fields, moisture, pH, and so on.1–6 These smart character-
istics enable SMPs to potentially be applied in areas ranging
from deployable aerospace structures to biomedical devices.7–12
SMPs are generally divided into either thermoplastic or thermo-
set classes according to their chemical structures. The networks
of thermoset SMPs are based on covalent crosslinking points,
whereas those of thermoplastic SMPs usually consist of physical
crosslinking points.1,10,13,14 Compared with thermoplastic SMPs,
thermoset SMPs always present a high reliability and superior
shape-memory behavior because of their stable covalent cross-
linking networks.15,16 Because of such unique features, thermo-
set SMPs have been investigated extensively in recent years.
Epoxy-based shape-memory polymers (ESMPs) are one of the
most important groups of thermoset SMPs. In addition to hav-
ing the general capabilities of thermoset SMPs, ESMPs also pro-
vide their own attractive properties, including an adjustable and
high glass-transition temperature (Tg; >80 8C), excellent thermal
performance, high modulus and fracture strength, rapid respon-
sive speed, chemical durability, and easy processing.17–22 These
characteristics make ESMPs extremely attractive for applications
as engineering smart and structural materials. However, pristine
ESMPs suffer from inherent brittleness, which is caused by their
relatively high crosslinking structures.23–26 Such drawbacks give
ESMPs a sensitivity to fatigue and cracks; this severely hinders
their potential for practical applications.25 To overcome this
challenge, extensive studies to improve the toughness of ESMPs
have been carried out worldwide.25,27–30 For example, Arnebold
and Hartwig27 prepared ESMPs with partial crystal structures
by cationic polymerization with poly(x-pentadecalactone)
Additional Supporting Information may be found in the online version of this article.
VC 2017 Wiley Periodicals, Inc.
J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.4578045780 (1 of 8)
(PPDL). Such epoxy–PPDL composite showed good shape-
memory cycle (SMC) stability and enhanced strength and
toughness. Huang et al.28 fabricated a novel ESMP through the
introduction of oxyethylene groups to diglycidyl ether of 9,9-
bis(4-hydroxyphenyl) fluorine. The cured 9,9-bis[4-(2-hydroxye-
thoxy) phenyl] fluorine/4,40-diaminodiphenyl sulfone polymeric
network exhibited a superior low water uptake, mechanical
properties, and thermal stability. Xie et al.25 recently reported a
kind of ESMP with a strain at break of 111% above the Tg and
212% at Tg, respectively. They greatly improved the toughness
of the ESMPs at high temperature.
The aforementioned studies have significantly paved the way for
improving the toughness of ESMPs. However, some of the mod-
ification methods have relied on complicated chemical reac-
tions; this restricts the large-scale applications of these ESMPs.
Other researchers have either sacrificed the thermal capabilities
or mechanical properties of ESMPs at room temperature. It is
still urgently needs to enhance the overall performance further
and simplify the fabrication methods of ESMPs to enable their
large-scale applications as smart structural materials in engi-
neering and other necessary areas. To achieve this, simple and
effective composite approaches are necessary for the develop-
ment of ESMPs. Previous composite strategies have generally
focused on carbon nanotubes, carbon black, carbon fibers, iron
oxide, silicon carbide, and so on.18,31–36 These fillers greatly
improved the thermal properties and mechanical strength of
ESMPs. However, the toughness of such stiff filler-reinforced
ESMP composites seriously decreases. Moreover, it is quite diffi-
cult to use them as composite matrixes for smart structures.
Crosslinked carboxylic nitrile–butadiene nanorubber (CNBNR)
powder is well known for toughening epoxies without sacrific-
ing their heat resistance.37,38 Meanwhile, CNBNR is easy to uni-
formly disperse in epoxy resins and preserve the characteristics
of epoxies for composite matrixes.39,40 Therefore, such elasto-
meric nanoparticles may be one of the best choices of fillers for
ESMP-based composites to achieve excellent comprehensive per-
formances. On the basis of these advantages, in this study,
CNBNR was introduced into the ESMP network to improve its
toughness and maintain other macroscopic properties of such
smart materials to broaden their engineering applications. Sys-
tematic research was constructed on the CNBNR–ESMP nano-
composites. Their macroscopic capabilities, including their
thermal properties, thermal stabilities, and mechanical proper-
ties, were investigated in detail. The reinforcing mechanism of
CNBNR for improving the overall properties of the CNBNR–
ESMP nanocomposites is discussed. We achieved an optimized
content of CNBNR for excellent overall properties, and the
shape-memory performance of nanorubber-toughed materials
was systematically investigated.
EXPERIMENTAL
Materials
ESMPs were prepared according to our previous studies; they
were thermoset crosslinking networks containing three ingre-
dients: epoxy resin, hardener, and accelerator.30 Crosslinked
CNBNR powder (Narpow VP-501, particle size distribution-
5 50–100 nm, Tg 5 229.1 8C, acrylonitrile content 5 26 wt %)
was supplied by Beijing Research Institute of Chemical Industry
(China). To remove the moisture, CNBNR was dried in an oven
at 50 8C for 24 h in vacuo before mixing.
Methods
Sample Preparation. First, a masterbatch was prepared.
CNBNR (20 phr) was added to a beaker containing 100 phr
epoxy resin and premixed by manual stirring for 10 min. The
obtained mixture was transferred to an ARE 500 planetary cen-
trifugal mixer for speed mixing at 300 rpm for 3 min, 500 rpm
for 2 min, and 1000 rpm for 1 min. To achieve a homogeneous
masterbatch, this blend was further dispersed six times by a
three-roll mill with a gradually reduced gap between rollers
(from 30 to 5mm). The ready masterbatch was then divided
into four parts with equal weights. Different amounts of epoxy
resin (75, 25, 8.3, and 0 g) and hardener (80, 40, 26.7, and 20 g)
were added to each part, respectively, to prepare the 5, 10, 15,
and 20 phr CNBNR–ESMP nanocomposites. The mixing pro-
cess was proceeded by continuous mechanical stirring with
speeds gradually increasing from 300 to 1000 rpm at 80 8C. After
45 min, the accelerator was added at a weight ratio of accelera-
tor to epoxy resin plus hardener of 1:50, and the mixture was
stirred for another 10 min at 1000 rpm. This blend was degassed
in a vacuum oven at 80 8C for 15 min to remove bubbles. Then,
it was put into preheated glass molds with different thicknesses
and cured at 80 8C for 2 h, 100 8C for 3 h, and 150 8C for
another 5 h. The CNBNR–ESMP nanocomposites were achieved
by cooling to room temperature. For comparison, a pure ESMP
was fabricated by direct mixturing epoxy resin, hardener, and
accelerator using mechanical stirring at 300 rpm. The curing
conditions of the pure ESMP were set to be the same as those
of the CNBNR–ESMP nanocomposites. The nomenclature and
composition of the prepared materials are given in Table I.
Characterization. Thermal analysis. Differential scanning calo-
rimetry (DSC; DSC/700/1410, Mettler-Toledo) was used to
study the phase-transition behaviors. The temperature was
increased from 25 to 200 8C at a rate of 5 8C/min under a nitro-
gen atmosphere. The thermal stabilities were tested by thermog-
ravimetric analysis (TGA; TGA/DSC1 SF1942, Mettler-Toledo)
under a nitrogen atmosphere with increasing temperature from
25 to 600 8C at 10 8C/min. The chemical structures were evalu-
ated with a Fourier transform infrared (FTIR) spectrometer
(Spectrum One, PerkinElmer) through dispersal of a powder
sample in a KBr pellet. The test was undertaken in the spectral
range 4000–370 cm21 at a resolution of 4 cm21.
Table I. Compositions of the CNBNR–ESMP Nanocomposites
Ingredient
SampleEpoxyresin (g)
Hardener(g)
Accelerator(g)
CNBNR(g)
ESMP 25 20 0.9 0
5 phr 100 80 3.6 5
10 phr 50 40 1.8 5
15 phr 33.3 26.7 1.2 5
20 phr 25 20 0.9 5
ARTICLE WILEYONLINELIBRARY.COM/APP
J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.4578045780 (2 of 8)
Static mechanical analysis. The static mechanical properties at
room temperature (25 6 2 8C) and 100 8C were characterized by
a universal material testing machine (Z050, Zwick/Roell) with a
thermal chamber under tensional mode. Dog-bone-shaped
specimens (ASTM D 638, Type IV) were cut off from the pre-
pared plates with a laser cutting machine. A crosshead speed of
5 mm/min was used for tests at both room temperature and
100 8C. The fractured surfaces were observed by scanning elec-
tron microscopy (SEM; VEGA3SB, Tescan) after sputter coating
with gold.
Shape-memory behavior. Dynamic mechanical analysis (DMA;
Q800, TA Instruments) under single-cantilever mode was car-
ried out to evaluate the thermomechanical properties. The tem-
perature ranged from 25 to 150 8C at a rate of 5 8C/min. A
frequency of 1 Hz was used during the whole test. The sample
dimension was 35 3 13 3 3 mm3. A bending-recovery test was
carried out to qualitatively investigate the shape-memory behav-
ior. A rectangular specimen with dimensions of 30 3 5 3 1 mm3
was used to study the shape memory behaviors. First, this was
heated to Tg and bent into a U-like shape after it became flexi-
ble. Then, a temporary shape was achieved by cooling the
deformed shape to room temperature under external force. The
shape-recovery phenomenon was monitored by an imaging pro-
cess when the deformed sample was reheated to Tg. The shape-
recovery time and shape-recovery angle as a function of time
were obtained from the recorded images. SMC testing on a
DMA Q800 (TA Instruments) was used to quantitatively
characterize the shape-memory behavior. The test was per-
formed under tensional film fixture, and a preload of 0.01 N
was imposed (30 3 4 3 1 mm3) to protect the specimen from
slide. The procedures for the SMC were as follows (Figure S1,
Supporting Information):
1. Heat the sample to Tg for deformation.
2. Isothermally stretch the soft sample to a certain strain with
a constant stress.
3. Cooling down the deformed sample to room temperature
under the same load to maintain the temporary shape.
4. Elevate the temperature to Tg to induce active shape
recovery.
The shape fixed ratio (Rf) and shape-recovery ratio (Rr) were
obtained by the following equations19,41–43:
Rf 5ef
eload
3100% (1)
Rfr5ef 2erec
ef
3100% (2)
where eload is the strain under external stress, ef is the strain
after unloading, and erec is the residual strain after recovery.
RESULTS AND DISCUSSION
Thermal Properties
Figure 1 shows the thermal analysis results for the pure ESMP,
CNBNR, and CNBNR–ESMP nanocomposites. The typical DSC
Figure 1. Thermal properties of the pure ESMP and CNBNR–ESMP nanocomposites: (a) DSC thermograms, (b) magnification of the selected part in
panel a, (c) thermogravimetry curves, and (d) DTG curves. [Color figure can be viewed at wileyonlinelibrary.com]
ARTICLE WILEYONLINELIBRARY.COM/APP
J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.4578045780 (3 of 8)
curves shown in Figure 1(a) reflect the thermal transition
behavior of the prepared materials. To eliminate the influence
of the heating history, the secondary heating scan was used in
this analysis. The step in Figure 1(b) indicates the occurrence of
the glass transition during the heating process. The DSC results
qualitatively showed that the introduction of CNBNR had no
negative effects on Tg. Instead, most Tgs of the CNBNR–ESMP
nanocomposites were slightly higher than that of the pure
ESMP. These results agreed with previous studies done for
CNBNR-toughened commercial epoxy composites.38,39
The thermal stabilities of the CNBNR–ESMP nanocomposites
are shown in Figure 1(c,d). For each sample, only one degrada-
tive stage was found [Figure 1(c)]; this was caused by the
decomposition of the polymeric backbone. Important thermal
elements, including the temperature related to the 5% weight
loss (Td), char yield, and temperature of the peak value in the
derivative thermogravimetry (DTG) curves (Tmax) were
extracted and are listed in Table II. Obviously, the decomposi-
tion of all of the materials (related to Td) started at above
345 8C (the Td of ESMP), and most of them were higher than
347 8C. The major decomposition, judging by Tmax, occurred at
about 380 8C. These results demonstrate that the fabricated
smart nanocomposites could be manipulated safely around their
Tg range. Further observation showed that the CNBNR–ESMP
nanocomposites had higher Td, Tmax and char yield values than
the pure ESMP (except for the 20 phr sample). This indicated
that the thermal stabilities of the CNBNR–ESMP nanocompo-
sites were slightly improved by the introduction of CNBNR into
the polymeric network.
From the consequences, it is easy to conclude that the CNBNR–
ESMP nanocomposites with addition of CNBNR at less than 15
phr possess better thermal capabilities than the pure ESMP net-
work. To study the possible reasons, FTIR tests were carried
out, and the results are shown in Figure 2. As shown in Figure
2(a), the improvement in the thermal properties was not caused
by chemical reactions between CNBNR and ESMP because there
was no difference in their FTIR characteristic peaks. A close
observation from the selected part in Figure 2(a) shows that the
absorption peaks belonging to the hydroxyl groups gradually
moved to a lower frequency when the content of CNBNR
increased [Figure 2(b)]. This indicated the presence of enhanced
hydrogen bonding between ESMP and CNBNR, which
explained the increased thermal performance.37,38 However, the
20 phr sample was discovered to have decreased thermal prop-
erties and thermal stabilities compared with the pure ESMP.
This was because when large amount of low-Tg material was
added to the system, the part not joining the network not only
decreased the interactions between the nanofillers and polymer
matrix but also resulted in a lower degree of crosslinking in the
material system. Therefore, a lower Tg and Td were achieved.
Static Mechanical Properties
To investigate the effectiveness of our proposed modification
method on the static mechanical properties, tensile tests were
performed on the CNBNR–ESMP nanocomposites at room
temperature. Corresponding data are plotted in Figure 3(a). The
typical stress–strain curves gradually varied from linearity to
nonlinearity; this indicated that the mechanical behaviors of the
prepared materials changed from brittle to ductile when the
content of CNBNR increased. Such a phenomenon was also ver-
ified by the SEM images shown in Figure 4. The fractured sur-
face of the pure ESMP was extremely smooth; this revealed a
highly brittle breakage [Figure 4(a)]. However, the fractured cross
Table II. Thermal Properties of the Pure ESMP and CNBNR and the
CNBNR–ESMP Nanocomposites
TGA
Sample Td (8C) Tmax (8C)Charyield (%) DMA: Tg (8C)
ESMP 344.9 378.6 0 102.9
5 phr 349.6 380.3 0.5 106.5
10 phr 347.8 378.9 0.7 103.5
15 phr 348.9 378.7 1.1 101.2
20 phr 329.7 375.0 2.3 99.5
CNBNR 357.3 457.1 9.2 —
Figure 2. Chemical structures of the pure ESMP and CNBNR–ESMP nanocomposites: (a) FTIR spectra and (b) selected area in panel A at a high magni-
fication [(1) pure ESMP, (2) 5 phr, (3) 10 phr, (4) 15 phr, and (5) 20 phr]. [Color figure can be viewed at wileyonlinelibrary.com]
ARTICLE WILEYONLINELIBRARY.COM/APP
J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.4578045780 (4 of 8)
Figure 3. Mechanical properties of the pure ESMP and CNBNR–ESMP nanocomposites: (a) typical stress–strain curves at room temperature, (b) results
of a tensional test at room temperature, (c) typical stress–strain curves at 100 8C, and (d) toughness at room temperature and 100 8C. [Color figure can
be viewed at wileyonlinelibrary.com]
Figure 4. SEM images of the tensional fractured cross sections of the CNBNR–ESMP nanocomposite series: (a) pure ESMP, (b) 5 phr, (c) 10 phr, (d) 15
phr, (e) 20 phr, and (f) magnification of the selected part in panel e. [Color figure can be viewed at wileyonlinelibrary.com]
ARTICLE WILEYONLINELIBRARY.COM/APP
J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.4578045780 (5 of 8)
section shown in Figure 4(b–e) became increasingly rough. Varied
crack propagations, plastic zones, and shear flow become more
and more visible, and their amounts increased markedly as the
content of CNBNR was changed from 5 to 20 phr. This clearly
indicated that the fractured mode of the materials turned into
ductile breakage.18,44 In Figure 3(b), an increase in the elongation
at break (eb) in the range 49.5–736.4% was observed with chang-
ing content of CNBNR from 5 to 20 phr. However, only a slight
drop in the Young’s modulus (E; 5.6–16.4%) was found when the
CNBNR content was varied from 5 to 20 phr; this indicated that
the remarkable improvement in eb had no dramatically negative
effect on E. Meanwhile, the maximum strength (rm) improved by
29.2% when 10 phr CNBNR was used, although it presented a
small decline after the first increase. These results suggest that the
introduction of CNBNR greatly improved the ductility of the
CNBNR–ESMP nanocomposites and could maintain or slightly
increase the strength at room temperature.
For ESMPs, it was equally important to investigate the static
mechanical properties at temperatures above Tg because eb in
the rubbery state is an important parameter for their deforming
performance. In Figure 3(c), the typical stress–strain curves
demonstrate that the ductility at rubbery state was boosted by
the addition of CNBNR. The eb values were improved from
22.1% for the pure ESMP to 44.7% for the 20 phr CNBNR
sample. This showed the improvement in the deformation capa-
bilities of the CNBNR–ESMP nanocomposites. A large deforma-
tion was achieved in the rubbery state without breakage of the
specimen, and this will allow us to increase the safety in the
applications of ESMPs.
The area of the strain–stress curve is always used to evaluate the
toughness of materials. As shown in Figure 3(d), we observed
that the toughness of the CNBNR–ESMP nanocomposites at both
room temperature (KIcrt) and 100 8C (KIc100 8C) greatly improved
in comparison with those of the pure ESMP. Such results indicate
the effectiveness of CNBNR for toughening ESMPs.
The enhancement of the mechanical properties at both room
temperature and high temperature were due to the introduction
of flexible nanofillers into the polymeric network. The ductile fea-
ture of CNBNR contributed to the high toughness of the
materials, whereas the large surface energy caused by their nano-
scale sizes and enhanced hydrogen bonding between CNBNR and
ESMP were key driving elements for the high strength and modu-
lus.38,39 In general, the CNBNR–ESMP nanocomposites showed
significantly enhanced mechanical properties.
Shape-Memory Properties
The thermomechanical properties were first studied because
they have close relationships with the shape-memory behavior.
Figure 5(a) shows that all of the tested samples presented a
glassy platform at low temperatures; this was followed by a sud-
den drop to the rubbery platform at high temperature. Such a
transformation revealed the occurrence of a glass–rubber transi-
tion during the heating process. Importantly, the changes in the
storage modulus between the glassy state and rubbery state
reached two orders of magnitude. This is the necessary precon-
dition for shape-memory functions.30 The peak value of the loss
factor is an important reference for Tg. As shown in Figure 5(b)
and Table II, the Tg values varied in a narrow range around
100 8C. Moreover, the variation trend of the Tg values obtained
from DMA was similar to that obtained from DSC [Figure
2(a)]. The increased Tg of the CNBNR–ESMP nanocomposites
in comparison with that of pure ESMP was attributed to the
enhanced hydrogen bonding as we illustrated before. However,
still, no higher Tg was observed for the 20 phr sample. The
underlying explanation was found from the SEM images, shown
in Figure 4. In contrast to others, some holes appeared on the
fractured surface of the 20 phr sample [Figure 4(e)]; these can
be clearly seen in Figure 4(f). The presence of these holes indi-
cated that the nanofillers aggregated into microscale particles;
this led to less interfacial adhesion between CNBNR and ESMP.
In that case, CNBNR acted as a normal rubber.39 Although it
was still effective for improving the toughness, no enhancement
to the thermal properties was found.38
The DMA results reveal that all of the CNBNR–ESMP samples
displayed glass–rubber transitions, dramatic changes in the stor-
age modulus during the heating process, and final flat rubbery
platforms. They indicate that all of the CNBNR–ESMP speci-
mens should have possessed shape-memory functions.25 Herein,
the 15 phr sample is chosen for deep investigation of the shape-
Figure 5. DMA results for the pure ESMP and CNBNR–ESMP nanocomposites: (a) storage modulus and (b) tan d as a function of temperature. [Color
figure can be viewed at wileyonlinelibrary.com]
ARTICLE WILEYONLINELIBRARY.COM/APP
J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.4578045780 (6 of 8)
memory behavior on the basis of its relatively excellent thermal
and mechanical properties. A bending-recovery experiment was
conducted first to qualitatively characterize the shape-memory
behavior. The corresponding results are shown in Figure 6(a,b).
The temporary U-like shape unfolded to the initial dashlike
shape with a fairly fast responsive speed. The shape-recovery
angle increased rapidly with time, and the entire shape-recovery
process occurred in merely 25 s, with nearly 100% shape recov-
ery. Such consequences visually demonstrated the excellent
shape-memory performance of the CNBNR–ESMP nanocompo-
site. The SMC test was carried out on DMA to quantitatively
analyze the shape-memory function of the 15 phr sample. To
eliminate the effect of the stress history, a secondary SMC is
presented in Figure 6(c) for analysis. According to eqs. (1) and
(2), the calculated Rf and Rr were up to 96.3 and 97.6%, respec-
tively; this implied the high shape-fixing and recovery perfor-
mance of the CNBNR–ESMP nanocomposite.
We concluded from both the quantitative and qualitative char-
acterizations that the CNBNR–ESMP nanocomposites presented
excellent shape-memory performance. Such results were due to
the introduction of CNBNR into the ESMP network. As
reported, the shape-memory behavior of a polymer is based on
its two-phase structures. One is a soft domain that functions in
shape deformation, and the other is a hard domain that is
responsible for memorizing the original shape and shape fixa-
tion.2,4,13 The introduction of the elastic CNBNR nanofillers
provided an increase in the soft domains in the ESMP network.
More movable molecular segments contributed to the easy
deformation and fast recovery of the CNBNR–ESMP nanocom-
posite. In the meanwhile, the enhanced hydrogen bonds
between CNBNR and ESMP constructed more hard domains in
the network; this enabled high Rf and Rr values for the
CNBNR–ESMP nanocomposite.
Ten SMCs were further carried out to verify the repeatability of
the shape-memory behavior. As shown in Figure 6(d), we
observed that these cycles presented reproducibility and consis-
tency with high Rf and Rr values. Importantly, no damage
appeared, even after 10 testing cycles. This suggested that the
shape-memory behavior of the CNBNR–ESMP nanocomposite
was highly repeatable with superior fatigue resistance due to the
high toughness at both room temperature and 100 8C. The near
perfect shape-memory behavior make CNBNR–ESMP desirable
for smart engineering applications.
CONCLUSIONS
A series of thermosetting shape-memory nanocomposites of
CNBNR–ESMP were successfully constructed, and their overall
properties were systematically investigated. The thermal proper-
ties of the prepared nanocomposites were improved slightly in
comparison to those of pure ESMP. The CNBNR–ESMP nano-
composites showed Tg values that varied in a narrow range
around 100 8C (according to the DMA test) and high Td and
Tmax values up to 350 and 380 8C, respectively. We also found
Figure 6. Characterization of the shape-memory behavior of the 15 phr CNBNR–ESMP nanocomposite: (a) shape recovery as a function of time, (b)
visual demonstration of the shape-memory behavior at 100 8C, (c) secondary SMC proceeding on DMA, and (d) 10 SMCs performed on DMA. [Color
figure can be viewed at wileyonlinelibrary.com]
ARTICLE WILEYONLINELIBRARY.COM/APP
J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.4578045780 (7 of 8)
that the CNBNR–ESMP nanocomposites provided greatly
enhanced ductility and strength at both room temperature and
above Tg without an excessive decrease in the modulus. FTIR
and SEM analysis indicated that the enhanced hydrogen bond-
ing and dispersion status of the nanofillers were two controlling
elements for the improvement of the macroscopic properties of
the CNBNR–ESMP nanocomposites. Moreover, the CNBNR–
ESMP nanocomposite possessed near perfect and highly repeat-
able shape-memory behavior. This research provides a feasible
method for the fabrication of ESMP-based nanocomposites with
improved toughness. Meanwhile, without the sacrifice of other
properties, especially the thermal properties and E. Such a com-
bined improvement of properties will pave a way for their
large-scale applications in engineering fields.
ACKNOWLEDGMENTS
This work was supported by the National Natural Science Founda-
tion of China (contract grant number 11225211), to which the
authors are very grateful.
REFERENCES
1. Zhao, Q.; Qi, H. J.; Xie, T. Prog. Polym. Sci. 2015, 49, 79.
2. Leng, J. S.; Lan, X.; Liu, Y. J.; Du, S. Y. Prog. Mater. Sci.
2011, 56, 1077.
3. Huang, W. M.; Yang, B.; Zhao, Y.; Ding, Z. J. Mater. Chem.
2010, 20, 3367.
4. Liu, C.; Qin, H.; Mather, P. T. J. Mater. Chem. 2007, 17,
1543.
5. Lendlein, A.; Jiang, H. Y.; Junger, O.; Langer, R. Nature
2005, 434, 879.
6. Glock, S.; Canal, L. P.; Grize, C. M.; Michaud, V. Compos.
Sci. Technol. 2015, 114, 110.
7. Lendlein, A.; Langer, R. Science 2002, 296, 1673.
8. Liu, Y. J.; Du, H. Y.; Liu, L. W.; Leng, J. S. Smart Mater.
Struct. 2014, 23, 023001.
9. Zarek, M.; Layani, M.; Cooperstein, I.; Sachyani, E.; Cohn,
D.; Magdassi, S. Adv. Mater. 2016, 28, 4449.
10. Hu, J. L.; Zhu, Y.; Huang, H. H.; Lu, J. Prog. Polym. Sci.
2012, 37, 1720.
11. Wang, Y. K.; Tian, W. C.; Xie, J. Q.; Liu, Y. Micromachines
2016, 7, 145.
12. Hu, J. L.; Meng, H. P.; Li, G. Q.; Ibekwe, S. I. Smart Mater.
Struct. 2012, 21, 053001.
13. Meng, H.; Li, G. Q. Polymer 2013, 54, 2199.
14. Ratna, D.; Karger-Kocsis, J. J. Mater. Sci. 2008, 43, 254.
15. Dong, Y. B.; Ni, Q. Q.; Fu, Y. Q. Compos. A 2015, 72, 1.
16. Yu, K.; McClung, A. J. W.; Tandon, G. P.; Baur, J. W.; Qi, H.
J. Mech. Time-Depend. Mater. 2014, 18, 453.
17. Kumar, K. S. S.; Biju, R.; Nair, C. P. R. React. Funct. Polym.
2013, 73, 421.
18. Liu, Y. Y.; Zhao, J.; Zhao, L. Y.; Li, W. W.; Zhang, H.; Yu,
X.; Zhang, Z. ACS Appl. Mater. Interfaces 2016, 8, 311.
19. Rousseau, I. A.; Xie, T. J. Mater. Chem. 2010, 20, 3431.
20. Wei, K.; Ma, B.; Liu, Y.; Wang, H. N.; Li, N. J. Mater. Res.
2015, 30, 2179.
21. Zou, Q.; Ba, L. H.; Tan, X. C.; Tu, M. J.; Cheng, J.; Zhang,
J. Y. J. Mater. Sci. 2016, 51, 10596.
22. Wei, H. Q.; Yao, Y. T.; Liu, Y. J.; Leng, J. S. Composites B
2015, 83, 7.
23. Fan, M. J.; Li, X. Y.; Zhang, J. Y.; Cheng, J. J. Therm. Anal.
Calorim. 2015, 119, 537.
24. Zhang, C. H.; Wei, H. G.; Liu, Y. Y.; Tan, H. F.; Guo, Z. H.
High Perform. Polym. 2012, 24, 702.
25. Zheng, N.; Fang, G. Q.; Cao, Z. L.; Zhao, Q.; Xie, T. Polym.
Chem. 2015, 6, 3046.
26. Liu, Y. Y.; Han, C. M.; Tan, H. F.; Du, X. W. Mater. Sci.
Eng. A 2010, 527, 2510.
27. Arnebold, A.; Hartwig, A. Polymer 2016, 83, 40.
28. Wu, X.; Yang, X.; Zhang, Y.; Huang, W. J. Mater. Sci. 2016,
51, 3231.
29. Leonardi, A. B.; Fasce, L. A.; Zucchi, I. A.; Hoppe, C. E.;
Soule, E. R.; Perez, C. J.; Williams, R. J. J. Eur. Polym. J.
2011, 47, 362.
30. Leng, J. S.; Wu, X. L.; Liu, Y. J. Smart Mater. Struct. 2009,
18, 095031.
31. Chen, L.; Li, W. B.; Liu, Y. J.; Leng, J. S. Composites B 2016, 91, 75.
32. Dong, Y. B.; Ni, Q. Q. Polym. Compos. 2015, 36, 1712.
33. Zhang, Z. Y.; Zhang, H.; Shou, J. Q.; Sun, Y. Y.; Liu, Y. Q.
New Carbon Mater. 2015, 30, 404.
34. Dong, Y. B.; Ding, J.; Wang, J.; Fu, X.; Hu, H. M.; Li, S.;
Yang, H. B.; Xu, C. S.; Du, M. L.; Fu, Y. Q. Compos. Sci.
Technol. 2013, 76, 8.
35. Ding, J.; Zhu, Y. F.; Fu, Y. Q. Polym. Compos. 2014, 35, 412.
36. Wong, T. T.; Lau, K. T.; Tam, W. Y.; Leng, J. S.; Wang, W.
X.; Li, W. B.; Wei, H. Q. Compos. Struct. 2015, 132, 1056.
37. Qi, G. C.; Zhang, X. H.; Li, B. H.; Song, Z. H.; Qiao, J. L.
Polym. Chem. 2011, 2, 1271.
38. Huang, F.; Liu, Y. Q.; Zhang, X. H.; Wei, G. S.; Gao, J. M.;
Song, Z. H.; Zhang, M. L.; Qiao, J. L. Macromol. Rapid
Commun. 2002, 23, 786.
39. Ozdemir, N. G.; Zhang, T.; Aspin, I.; Scarpa, F.; Hadavinia,
H.; Song, Y. eXPRESS Polym. Lett. 2016, 10, 394.
40. Ozdemir, N. G.; Zhang, T.; Hadavinia, H.; Aspin, I.; Scarpa,
F. J. Compos. Mater. 2016, 50, 2633.
41. Xie, T. Nature 2010, 464, 267.
42. Dolog, R.; Weiss, R. A. Macromolecules 2013, 46, 7845.
43. Li, X. X.; Zhu, Y. F.; Dong, Y. B.; Liu, M.; Ni, Q. Q.; Fu, Y.
Q. J. Nanomater. 2015, 2015, 475316.
44. Li, B. H.; Zhang, X. H.; Gao, J. M.; Song, Z. H.; Qi, G. C.;
Liu, Y. Q.; Qiao, J. L. J. Nanosci. Nanotechnol. 2010, 10,
5864.
ARTICLE WILEYONLINELIBRARY.COM/APP
J. APPL. POLYM. SCI. 2018, DOI: 10.1002/APP.4578045780 (8 of 8)
