Side chain dendritic polyurethanes with shape-memory effect

Article abstract of DOI:10.1039/b910614f

In this study, thermally reversible polyurethanes (PUs) with shape memory effect were developed by using hydrogen bonding to enhance physical interactions. Two different types of PUs were synthesized: (1) a linear PU whose hard segment consists of methyle

Full text of DOI:10.1039/b910614f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Side chain dendritic polyurethanes with shape-memory effect
Cheng-Che Tsai,a Chia-Cheng Chang,a Ching-Shiang Yu,a Shenghong A. Dai,a Tzong-Ming Wu,a
a
Wen-Chiung Su,b Chang-Nan Chen,c Franklin M. C. Chenad and Ru-Jong Jeng
*
Received 29th May 2009, Accepted 11th September 2009
First published as an Advance Article on the web 2nd October 2009
DOI: 10.1039/b910614f
In this study, thermally reversible polyurethanes (PUs) with shape memory effect were developed by
using hydrogen bonding to enhance physical interactions. Two different types of PUs were synthesized:
(1) a linear PU whose hard segment consists of methylene di-para-phenyl isocyanate (MDI) and
di(ethylene glycol) (DEG); its soft segment is made of Tone 0260 polyol, a polyester polyol; (2) a side-
chain dendritic PU which replaces DEG with different generations of dendritic chain extenders. By
incorporation of the dendritic structure with peripheral long alkyl chains (strong van der Waals forces),
the miscibility between hard and soft segments can be significantly improved. Consequently, the
hydrogen bonding reinforced hard segments (malonamide linkages) of side chain dendritic PUs result
in greatly enhanced mechanical properties and shape memory effect. During cyclic shape memory tests,
one of the series can effortlessly achieve complete recovery in less than 10 second without any
deficiency.
response properties. Traditional linear PUs exhibiting note-
1
Introduction
worthy phase separation and proper molecular weight of the soft
segment might exhibit certain shape memory effects.13,16 Several
studies were set forth to overcome the poor recovery ratio and
mechanical properties by developing either chemical or physical
modifications such as adjusting the hard-to-soft segment ratio,13
incorporating ionic forces and hydrogen bonding,15,16 creating an
irreversible chemical cross-linking mechanism,17–19 or blending
with inorganic materials.20–22 Based on the above studies, it is
evident that complex processes and compositions are required to
achieve materials with excellent mechanical and shape memory
properties. Therefore, finding the essential factors of a simple
procedure, better shape recovery characteristics and environ-
mental practicality is still in great demand for future applica-
bility.
Recently, shape-memory materials which can sense and respond
to external stimulus have attracted great interest due to their
potential applications in various fields such as medical devices1–3
and sensor development.4,5 Organic materials (i.e. polymers) can
be triggered by pH value, electrical voltage or entropy elasticity,
which are more prevalent in applications because of low cost,
ease of processing, and biocompatibility.6
Thermo-responsive polymers such as poly(cyclooctene)
(PCO),4 oligomers,7 poly(ethylene-vinyl acetate),8 polyethylene
(PE),9 poly(vinyl alcohol) (PVA)10 and polyurethane (PU)11–16
were developed as part of the above-mentioned endeavor.
Specifically, the PU system has the advantages of low cost and
excellent processability. Facile modification of the PU system
can be accomplished so that they can be applied as biomedical
materials or further processed in the plastics industry.
A new type of polyurea/malonamide dendron based on an
industrial raw material, methylene di-para-phenyl isocyanate
(MDI), was synthesized previously in our laboratory.23 By
incorporating various functional groups in the exteriors, these
dendrons with strong inter/intramolecular interactions were
utilized in various areas such as modification of thermally
reversible polyurethane,24,25 surface modification of montmoril-
lonite,26,27 and self-organization of nonlinear optical materials.28
In addition, the pendant architecture of the side chain type
polymer with different chemical functionalities might restrict the
backbone conformation and enhance the assembly ability to
approach microphase segregation of incompatible constitu-
ents.29,30 Herein, a series of PUs containing dendritic side-chains
with peripheral long alkyl chains were synthesized. The dendritic
chain extenders encouraged phase-mixing of soft and hard
segments, and provided more opportunities for hydrogen
bonding. The resulting materials are compared to linear poly-
urethanes (as control materials). Moreover, the hydrogen
bonding effect and microphase morphology relating to thermal
and mechanical properties were also studied.
In linear PUs, each PU chain contains hard segments
(domains) of highly polar urethane groups from the reaction
between isocyanate and chain-extender derivatives, and rela-
tively non-polar soft segments (domains). During the deforma-
tion process, the hard domains must provide sufficient strength
to stabilize the entire architecture, whereas the soft domains act
as a flexible phase to respond to the external stress. Additionally,
the morphology of polymer matrices and the molecular weight of
the soft segment dominate the mechanical and shape memory
aDepartments of Chemical Engineering, and Materials Science and
Engineering, National Chung Hsing University, Taichung, 402, Taiwan.
E-mail: rjjeng@dragon.nchu.edu.tw
bChung-Shan Institute of Science and Technology, Lungtan, Taoyuan, 325,
Taiwan
cDepartment of Applied Chemistry, Chaoyang University of Technology,
Taichung, 413, Taiwan
dDepartment of Natural and Applied Sciences (Chemistry), University of
Wisconsin-Green Bay, Wisconsin, 54311, USA
8484 | J. Mater. Chem., 2009, 19, 8484–8494
This journal is ª The Royal Society of Chemistry 2009

View Article Online
Equipment
2
Experimental
1H NMR spectra were taken on a Varian Gemini-200 FT-NMR
spectrometer with chloroform-d and dimethyl sulfoxide-d6. IR
measurements were performed on a PerkinElmer Spectrum One
Fourier transform infrared (FTIR) spectrometer. Elemental
CHN analysis was performed on a Heraeus CHN-OS Rapid
Analyzer. Thermal analysis was performed in N2 on a TA
Instruments DSC2010 at a heating rate of 10 ꢀC/min. Ther-
mogravimetric analysis (TGA) was performed under nitrogen
using a Seiko SSC-5200 Thermogravimetric Analyzer at a heat-
ing rate of 10 ꢀC/min. Fast atom bombardment mass spec-
trometry (FABMS) analysis was performed on a JEOL JMS SX/
SX 102A mass spectrometer equipped with the standard FAB
source, whose upper limit for measuring molecular weight is
Materials
MDI, DEG, isobutyryl chloride, triethylamine, diethyltriamine
(DETA), octadecyl alcohol and N-(3-aminopropyl)diethanol-
amine (APDEA) were purchased from Sigma-Aldrich (St. Louis,
MO, USA). Tone 0260 polyol (Mn ¼ 3000), which is a capro-
lactone based difunctional polyester polyol, was purchased from
Dow Chemical Company (Midland, MI, USA). Tetrahydrofuran
(THF) was distilled under nitrogen from sodium/benzophenone
before use. The building block, 4-isocyanate-4-(3,3-dimethyl-2,4-
dioxo-azetidine)diphenyl methane (IDD) with selective reactivity
nature was synthesized according to procedures described in the
literature.31
Scheme 1 Synthesis of G0.5 diol, G1.5 diol and G2.5 diol from IDD and 1-octadecyl alcohol.
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 8484–8494 | 8485

View Article Online
2000. Matrix-assisted laser desorption ionization with time of
flight (MALDI-TOF) mass spectra were recorded on a Voyager
DE-PRO (Applied Biosystems, Houston, TX) equipped with
a nitrogen laser (337 nm) operating in linear detection mode to
generate positive ion spectra with dithranol as a matrix, dimethyl
sulfoxide (DMSO) as a solvent and sodium trifluoroacetate as an
additive agent. Gel permeation chromatography (GPC) was
performed in THF as an eluent with a Waters Apparatus
equipped with Waters Stygel columns with a refractive index (RI)
detector and polystyrene calibration. Tensile measurements were
carried out on specimens having dimensions of 5 mm wide and 20
mm length using a Shimadzu testing machine with a cross-head
speed of 50 mm/min at room temperature. The specimens were
prepared and measured according to ASTM D638. Four samples
were evaluated for each measurement. Standard deviations were
typically 4% for modulus, 3% for tensile strength and 5% for
elongation. Molecular arrangement of PUs was analyzed by an
X-ray powder diffractometer (XRD, Shimadzu SD-D1 using
a Cu target at 35 kV, 30 mA).
a nitrogen atmosphere for 2 h. The solvent was evaporated and
the product was purified by recrystallization from acetone to give
G1 (4.6 g, 85%) as a white powder. Found: C, 72.71; H, 9.52; N,
7.41. C78H121N7O8 requires C, 72.91; H, 9.49; N, 7.63%;
n
max/cmꢁ1 3356 (NH), 1706 [(NH)C]O(O)] and 1652
[C]O(NH)]; dH (200 MHz; DMSO) 0.83 (6H, t, CH3), 1.23
(28H, s, CH2), 1.38 (12H, s, CH3), 1.57 (4H, t, CH2), 2.55 (4H, t,
CH2(NH)), 3.30 (4H, m, CH2(N)), 3.86 (4H, s, Ar–CH2–Ar),
4.02 (4H, t, CH2), 7.06–7.62 (16H, m, Ar–H); m/z (FAB)
1285 (M+).
Synthesis of dendron generation 1.5 (G1.5). To a solution of G1
(4.0 g, 3.11 mmol) in dry THF (25 cm3), IDD (1.24 g, 3.89 mmol)
was added. The solution was stirred at 60 ꢀC under nitrogen for
2 h. The solvent was evaporated and the product was purified by
precipitation from methanol to give G1.5 (4.76 g, 91%) as a white
powder. Found: C, 72.04; H, 7.99; N, 7.80. C97H137N9O11
requires C, 72.58; H, 8.60; N, 7.85%; nmax/cmꢁ1 3336 (NH), 1854
(C]O), 1738 (C]O), 1706 [(NH)C]O(O)] and 1652
[C]O(NH)]; dH (200 MHz; DMSO) 0.83 (6H, t, CH3), 1.23
(60H, s, CH2), 1.38 (12H, s, CH3), 1.44 (6H, s, CH3), 1.57 (4H, t,
CH2), 3.20 (8H, m, CH2(N)), 3.77 (4H, s, Ar–CH2–Ar), 3.81 (2H,
s, Ar–CH2–Ar), 4.01 (4H, t, CH2), 6.90–7.70 (24H, m, Ar–H);
m/z (FAB) 1605 (M+).
Synthesis of dendrons (Scheme 1)
Synthesis of dendron generation 0.5 (G0.5). 1-Octadecyl
alcohol (5.0 g, 18.49 mmol) was added to a solution of IDD
(7.402 g, 23.12mmol) in dry THF (40 cm3). The solution was
ꢀ
stirred at 60 C under nitrogen for 2 h. The solvent was evapo-
Synthesis of dendron generation 2 (G2). To a solution of G1.5
(4.78 g, 2.98 mmol) in dry THF (25 cm3), DETA (0.15 g, 1.45
mmol) was added. The solution was stirred at 60 ꢀC under
a nitrogen atmosphere for 2 h. The solvent was evaporated and
the product was purified by silica gel chromatography with ethyl
acetate to give G2 (3.3g, 67%) as a yellow powder. Found: C,
71.55; H, 8.36; N, 8.82. C198H287N21O22 requires C, 71.77; H,
8.73; N, 8.88%; nmax/cmꢁ1 3334 (NH), 1704 [(NH)C]O(O)] and
1654 [C]O(NH)]; dH (200 MHz; DMSO) 0.82 (12H, t, CH3),
1.23 (120H, s, CH2), 1.34 (36H, s, CH3), 1.57 (8H, t, CH2), 2.55
(4H, t, CH2(NH)), 3.30 (20H, m, CH2(N)), 3.86 (12H, s, Ar–
CH2–Ar), 4.02 (8H, t, CH2), 7.06–7.62 (48H, m, Ar–H); m/z
(MALDI-TOF) 3312 (M + Na+).
rated and the product was purified by precipitation from meth-
anol to give G0.5 (10.6 g, 85%) as a white powder. Found: C,
75.29; H, 8.91; N, 4.72. C37H54N2O4 requires C, 75.21; H, 9.21;
N, 4.74%; nmax/cmꢁ1 3300 (NH), 1854 (C]O), 1736 (C]O) and
1700 [(NH)C]O(O)]; dH(200 MHz; CDCl3) 0.83 (3H, t, CH3),
1.23 (30H, s, CH2), 1.39 (6H, s, CH3), 1.57 (2H, t, CH2), 3.85
(2H, s, Ar–CH2–Ar), 4.07 (2H, t, CH2), 6.98–7.68 (8H, m, Ar–
H); m/z (FAB) 591 (M+).
Synthesis of dendron generation 1 (G1). To a solution of G0.5
(5.0 g, 8.46 mmol) in dry THF (25 cm3), DETA (0.426 g,
ꢀ
4.13 mmol) was added. The solution was stirred at 60 C under
Scheme 2 Preparation of linear and side chain dendritic PUs.
8486 | J. Mater. Chem., 2009, 19, 8484–8494
This journal is ª The Royal Society of Chemistry 2009

View Article Online
Table 1 Formulations of all series of PUs
Synthesis of dendron generation 2.5 (G2.5). To a solution of G2
(4.50 g, 1.37 mmol) in dry THF (25 cm3), IDD (0.48 g, 1.50
mmol) was added. The solution was stirred at 60 ꢀC under
a nitrogen atmosphere for 2 h. The solvent was evaporated and
the product was purified by silica gel chromatography with ethyl
acetate to give G2.5 (4.17 g, 68%) as a yellow powder. Found: C,
70.77; H, 8.00; N, 8.70. C217H303N23O25 requires C, 71.72; H,
8.40; N, 8.87%; nmax/cmꢁ1 3328 (NH), 1854 (C]O), 1742(C]O)
and 1676 [(NH)C]O(O)]; dH (200 MHz; DMSO) 0.83 (12H, t,
CH3), 1.23 (56H, s, CH2), 1.38 (36H, s, CH3), 1.44 (6H, s, CH3),
1.57 (8H, t, CH2), 3.30 (24H, m, CH2(N)), 3.86 (12H, s, Ar–CH2–
Ar), 3.94 (2H, s, Ar–CH2–Ar), 4.02 (8H, t, CH2), 7.06–7.62 (48H,
m, Ar–H); m/z (MALDI-TOF) 3655 (M + Na+).
Molar ratio
Sample
Chain extender
Tone 0260 polyol
MDI
Hc (wt%)a
L-45
L-50
L-55
1
1
1
1
1
1
1
1
1
1
1
1
0.616
0.130
0.104
0.455
0.365
0.293
0.915
0.730
0.590
1.830
1.470
1.180
1.616
1.130
1.104
1.455
1.365
1.293
1.915
1.730
1.590
2.830
2.470
2.180
45
50
55
45
50
55
45
50
55
45
50
55
D0-45
D0-50
D0-55
D1-45
D1-50
D1-55
D2-45
D2-50
D2-55
Synthesis of dendritic diols
a
Hard segment content (Hc) (wt%) ¼ [MDI (g) + Chain extender (g)]/
[MDI (g) + Chain extender (g) + Tone 0260 polyol (g)].
Synthesis of G0.5 diol. To a solution of G0.5 (3.0 g, 5.07 mmol)
in dry THF (25 cm3), N-(3-aminopropyl)diethanolamine
(APDEA) (0.82 g, 5.05 mmol) was added. The solution was
stirred at 60 ꢀC under a nitrogen atmosphere for 4 h. The solvent
was evaporated and the product was purified by silica gel chro-
matography with acetone to give G0.5 diol (4.05 g, 80%). Found:
C, 70.30; H, 9.07; N, 7.27. C44H72N4O6 requires C, 70.81; H,
9.64; N, 7.44%; nmax/cmꢁ1 3450 (OH), 1706 [(NH)C]O(O)] and
1652 [C]O(NH)]; dH (200 MHz; DMSO) 0.83 (3H, t, CH3), 1.23
(30H, s, CH2), 1.39 (6H, s, CH3), 1.57 (4H, m, CH2), 3.0–3.1 (8H,
m, CH2), 3.37 (4H, t, CH2), 3.79 (2H, s, Ar–CH2–Ar), 4.07 (2H, t,
CH2), 6.98–7.68 (8H, m, Ar–H); m/z (FAB) 753 (M+).
toluene. These two reactants were heated at 60 ꢀC for 0.5 h
forming isocyanate terminated prepolymer, and then the chain
extender (DEG, G0.5 diol, G1.5 diol or G2.5 diol) was added
dropwise into the stirring reaction at 80 ꢀC for 3 h to give linear or
side chain dendritic type PUs. All of these PUs are identified with
labels in the form ‘‘WX-YZ’’, where the first letter (W) denotes the
type of PU (L: linear or D: dendritic), and the following numbers
denote the generation of dendritic diol (X) and hard segment
content (YZ %). Hard segment weight fraction is then obtained
from the total weight of the PUs divided by the hard segment
weight. For example, D1-50 possesses 50% weight fraction of hard
segment, using the G1.5 diol, whereas D2-50 possesses 50%weight
fraction of hard segment, using the G2.5 diol (Scheme 2). The
formulations of these PUs are shown in Table 1.
Synthesis of G1.5 diol. To a solution of G1.5 (3.0 g, 1.87 mmol)
in dry THF (25 cm3), APDEA (0.3 g, 1.85 mmol) was added. The
solution was stirred at 60 ꢀC under a nitrogen atmosphere for 4 h.
The solvent was evaporated and the product was purified by silica
gel chromatography with acetone to give G1.5 diol (2.31 g, 70%).
Found: C, 68.7; H, 8.22; N, 8.41. C104H155N11O13 requires C,
PU films were prepared as follows. The as-polymerized solu-
tions (10 wt %) were poured into Teflon circular disks at room
temperature for 24 h and then heated to 70 ꢀC for an additional
24 h. Films were on the order of 5 mm thick.
70.67; H, 8.84; N, 8.72%;
n
max/cmꢁ1 3330 (OH), 1704
[(NH)C]O(O)] and 1654 [C]O(NH)]; dH (200 MHz; DMSO)
0.83 (6H, t, CH3), 1.23 (60H, s, CH2), 1.38 (18H, s, CH3), 1.57 (6H,
m, CH2), 3.0–3.4 (20H, m, CH2), 3.78 (6H, s, Ar–CH2–Ar), 4.01
(4H, t, CH2), 6.90–7.70 (24H, m, Ar–H); m/z (FAB) 1767 (M+).
Shape memory test (Scheme 3)
A digital camera (Electrophysics Micro-Viewer 7290A) equipped
with an image capture card (NI IMAQ PCI 1407) was used to
evaluate the shape memory effect. The recovery ratio is based on
the ratio of the angle between two planes to right angle.32–34
A
Synthesis of G2.5 diol. To a solution of G2.5 (3.0 g, 0.94 mmol)
in dry THF (25 cm3), APDEAꢀ(0.15 g, 0.93 mmol) was added.
The solution was stirred at 60 C under a nitrogen atmosphere
for 4 h. The solvent was evaporated and the product was purified
by silica gel chromatography with acetone to give G2.5 diol (1.89
g, 60%). Found: C, 69.57; H, 8.22; N, 8.95. C224H321N25O27
requires C, 70.87; H, 8.52; N, 9.22%; nmax/cmꢁ1 3330 (OH), 1700
[(NH)C]O(O)] and 1652 [C]O(NH)]; dH (200 MHz; DMSO)
0.83 (12H, t, CH3), 1.23 (56H, s, CH2), 1.38 (42H, s, CH3), 1.57
(10H, m, CH2), 3.0–3.4 (36H, m, CH2), 3.79 (14H, s, Ar–CH2–
Ar), 4.02 (8H, t, CH2), 7.06–7.62 (48H, m, Ar–H); m/z (MALDI-
TOF) 3795.2 (M+).
tensile tester equipped with a thermochamber was also utilized to
study the shape memory effect of the dendritic PUs.35 The
specimen was loaded to the strain (3m: 50%) at a constant
crosshead speed of 10 mm/min at 60 ꢀC (Process 1). Subse-
quently, it was cooled to 20 ꢀC under the same strain (Process 2).
ꢀ
After 10 min at 20 C, the load on the specimen was taken off
(holding for 5 min), the recovery of strain (3u) was recorded
ꢀ
ꢀ
(Process 3), then the specimen was heated from 20 C to 60 C,
and the residual strain (3p) was recorded (Process 4). The shape
fixity and shape recovery could be calculated based on the
following equations:
shape fixity (%) ¼ 3u/3m ꢂ 100
Synthesis of linear and side chain dendritic PUs (Scheme 2)
Different compositions of PUs were synthesized via a two-step
condensation reaction. The reactions were carried out by adding
MDI to a solution of Tone 2060 polyol (Mn ¼ 3000) in dry
shape recovery (%) ¼ (3m ꢁ 3p)/3m ꢂ 100
J. Mater. Chem., 2009, 19, 8484–8494 | 8487
This journal is ª The Royal Society of Chemistry 2009

View Article Online
Scheme 3 Schematic diagram of shape memory test.
additional ‘‘pseudo-cross-linked’’ interactions would greatly
improve the mechanical properties. The urethane (1702 cmꢁ1
and urea or amide (1654 cmꢁ1) carbonyl absorption peaks orig-
3
Results and discussion
)
Characterization
inated from hydrogen bonding between intra- or inter-polymer
Recently, we presented the facile synthesis of polyurea/malona-
mide dendrons with phenyl or decyl alkyl groups located on the
exterior.23,26 In this work, a series of dendrons (G0.5 to G2.5)
with peripheral octadecyl alkyl chains have been synthesized in
the same manner. Subsequently, the dendritic diols (G0.5 diol,
G1.5 diol, and G2.5 diol) were synthesized by ring-opening
addition reaction of azetidine-2,4-dione groups toward the
primary amine of APDEA (Scheme 1). These dendritic diols
serve as chain extenders for side chain dendritic PUs. The
chemical structures of the resultant compounds were fully
1
confirmed by FT-IR, H NMR, elemental analysis (CHN), and
MALDI-TOF spectroscopy. Fig. 1 shows the FT-IR and mass
spectrum of G2.5 diol. The absorption peaks of azetidine-2,4-
dione, 1854 cmꢁ1 (C]O asymmetric stretching) and 1742 cmꢁ1
(C]O symmetric stretching), disappeared completely, whereas
the peaks at 3300 cmꢁ1 (OH stretching of hydroxyl) and 1652
cmꢁ1 (C]O stretching) emerged. In addition, G2.5 diol was
confirmed by mass spectrometry (MALDI-TOF ¼ 3795.23 (M+)
m/z) which is consistent with the theoretical molecular weight.
Hydrogen bonding effect
The FT-IR spectroscopy was utilized to investigate the extent of
the hydrogen bonding effect among hard/soft segments.25,36,37 As
shown in Fig. 2, the NH stretching peak, 3330 cmꢁ1, CH asym-
metric (2941 cmꢁ1) and symmetric stretching peak (2865 cmꢁ1
)
were observed. Furthermore, the carbonyl stretching vibration
peaks from 1600 to 1800 cmꢁ1 were assigned to the urethane, urea
and malonamide groups. According to previous studies,24,25,37
analogous polymers possessing malonamide and urea groups
were utilized to prepare supramolecules via hydrogen bonding.
With the incorporation of abundant hydrogen bonding sites (i.e.,
the presence of urea, urethane, and amide linkages), the
Fig. 1 (a) FT-IR spectrum and (b) MALDI-TOF MS spectrum for
G2.5 diol.
8488 | J. Mater. Chem., 2009, 19, 8484–8494
This journal is ª The Royal Society of Chemistry 2009

View Article Online
malonamide and urea groups was found for D2-50. Nevertheless,
it is difficult to identify which functional group is responsible for
a certain absorption peak. Those physical interactions between
inter- or intra-polymer chains were observed at ambient
temperature, but apparently the interactions weakened as the
temperature was increased above 110 ꢀC. Hence, it is postulated
that the ‘‘pseudo-cross-linked’’ interactions via hydrogen
bonding could effectively sustain optimal mechanical stress and
provide a process window for shape memory test performed
between 25 and 110 ꢀC.
Morphology
The microstructure of the PUs samples was analyzed using DSC
and XRD measurements. Thermal properties of all samples are
shown in Table 2. According to TGA analysis, the materials
maintain great thermal stability at 220 ꢀC without any weight
loss. DSC analysis (Fig. 4a) was utilized to study the micro-
structure of PUs. The deviation of glass transition temperature
from that of soft segments (TgS) was used to determine the degree
of phase separation of the polyurethane.38–40 A larger deviation
of TgS suggests a larger degree of phase mixing between hard and
soft segments. On the other hand, the order degree within the
hard segments depends upon the rigidity and hydrogen bonding
within the hard segments. Indeed, only a higher order structure
would reveal the melting temperature of hard segments (TmH). In
the L-series PUs, the soft segments revealed a TgS of ꢁ50 ꢀC,
a crystallization temperature (TcS) of ꢁ9 ꢀC, and a sharp melting
Fig. 2 FT-IR spectra for L-45 and D0-45 samples.
chains. Therefore, considerable quantities of additional non-
covalent interactions or polarity discrepancy between hard and
soft segments were expected to induce the reinforcing effect of
hard domains. In addition, the variable-temperature FT-IR
spectra of L-50 (Fig. 3(a) and (b)) and D2-50 (Fig. 3(c) and (d))
PUs were utilized to analyze the dissociation of hydrogen
bonding. During the heating process, the NH and carbonyl peaks
were found to continuously shift to higher wavenumbers.
Further contribution to hydrogen bonding interactions via
Fig. 3 Variable-temperature FT-IR spectra for L-50 (a and b) and D2-50 (c and d).
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 8484–8494 | 8489

View Article Online
Table 2 Mechanical and thermal properties of PUs
a
b
c
d
e
g
3b (%)
Sample
Td (ꢀC)
TgS (ꢀC)
TcS (ꢀC)
TmS (ꢀC)
TmH (ꢀC)
sbf (MPa)
L-45
L-50
L-55
287
289
287
262
266
256
273
271
264
263
273
261
ꢁ52.5
ꢁ46.6
ꢁ52.7
ꢁ39.5
ꢁ36.8
ꢁ26.7
ꢁ44.6
ꢁ36.8
ꢁ32.6
ꢁ57.4
ꢁ55.2
ꢁ47.3
ꢁ7.2
ꢁ9.2
ꢁ9.2
34.8
39.7
40.1
N/A
N/A
N/A
N/A
N/A
N/A
39.6
41.6
39.7
75.9
74.9
77.2
31.8
27.9
31.1
33.0
32.2
30.4
155.3
153.6
161.4
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
8.5
11.7
12.8
8.0
240
245
242
350
373
364
680
695
671
525
509
511
D0-45
D0-50
D0-55
D1-45
D1-50
D1-55
D2-45
D2-50
D2-55
9.6
10.4
18.2
18.7
17.4
10.6
10.4
11.3
a
5% weight loss. b Glass-transition temperature of soft segment. c Crystallization temperature of soft segment. d Melting temperature of soft segment.
Melting temperature of hard segment. f Tensile strength at break. g Elongation at break.
e
temperature (TmS) of 40 ꢀC, whereas the hard segments displayed
For D2 series PUs, the abundance of hydrogen bonding
interactions originated from the malonamide dendrons which
could lead to self-associated segments. Hence, the peripheral
alkyl chains have no discernable influence on miscibility between
the hard and soft segments in this case. As the size of the side
a TmH of 155 ꢀC. This individual thermal phenomenon shows
that phase segregation of the L-series PUs is due to the incom-
patibility between the well-ordered soft segment crystallites and
the hard segments. Furthermore, its thermal behavior is inde-
pendent of the hard segment content.
For the D0 series PUs, the soft segments displayed the TgS
from ꢁ39.5 ꢀC to ꢁ26.7 ꢀC which increased with increasing hard
segment content. Furthermore, the TmH of hard segments vanish
in all cases, while a distinct transition temperature appeared
around 10.0 ꢀC between the pristine TgS and TmS of the soft
segments. In order to confirm this newly emerged transition
temperature, a control polymer comprising 100% hard segments
(MDI + G0.5 diol) was synthesized and a Tg was detected at
30.0 ꢀC (Fig. 4b). Compared to the control polymer and L-55, the
decreased TgH of the hard segments (from 30.0 ꢀC to 10.0 ꢀC) and
an enhanced TgS of the soft segment (from ꢁ52.7 ꢀC to ꢁ26.7 ꢀC)
were observed for D0-55. This phase-mixing phenomenon was
ascribed to the dissolution between hard and soft segments
because of the interpenetration of the long alkyl chains of the
dendritic structures. Hence, the emerged transition temperature
could be related to the glass transition temperature of hard
segments (TgH). Additionally, a regular structure could be
derived from the pendant octadecyl chains while integrating with
the polymer backbone chains. Consequently, the TcS and TmS of
soft segments were greatly increased to around 40 ꢀC and 75 ꢀC,
respectively. A similar phenomenon of long alkyl chains and
hydrogen bonds influencing the original arrangement of soft
segments was also observed in the literature.41,42
The phase-mixing behavior was also investigated for the D1
series PUs that exhibited an increase of the TgS. Compared to D0
series PUs, the pronounced branched steric effect from the G1.5
diol architecture not only brings about enlarged distances
between polymer chains, but also prohibits metastable crystalli-
zation (Fig. 4b). As a result, the TmS of the D1 series soft
segments was detected at about 40.0 ꢀC without delaying the
melting progress. In addition, a control polymer comprising only
MDI and G1.5 diol (100% hard segments) exhibited a Tg of
55.0 ꢀC. Owing to the hard segment TgH relaxation acting in
conjunction with the melting of soft segment, only miniature
crystallization of soft segments was observed on DSC traces.
Fig. 4 DSC thermograms of (a) L-50, D0-50, D1-50, D2-50, and (b) D0-
55, MDI + G0.5 diol, D1-55, MDI + G1.5 diol, D2-55 and MDI + G2.5
diol samples.
8490 | J. Mater. Chem., 2009, 19, 8484–8494
This journal is ª The Royal Society of Chemistry 2009

View Article Online
with polydispersities between 2.1 and 2.8. The hydrodynamic
volumes and polar functional groups of the branched polymer
would influence the retention times and, therefore, the absolute
molecular weight might differ from the Mn measured by
GPC.46,47 Nevertheless, the sufficient film formability allowed us
to evaluate the mechanical properties of these PUs. Mechanical
properties as presented by stress-strain analysis are summarized
in Table 2. The miscibility between hard/soft segments, and
strong inter/intramolecular interactions would greatly influence
the mechanical properties of PUs.48 For each series, the tensile
strength increased with increasing hard segment content, which is
ascribed to the accumulated interactions of a great many
hydrogen bonding sites. However, the characteristics of elonga-
tion at break were only related to different types of chain
extenders, but independent of hard segment content. Fig. 6
shows various types of PUs with the same hard segment content
of 50%. Based on the DSC results, both hard segment and soft
segment of L series would exhibit crystalline peaks, indicating
a clear phase separation. Consequently, the L series samples
revealed hard plastic behavior with relatively poor elongations,
about 240%. After respective incorporation of G0.5 and G1.5
diol, the appearance of phase-mixed morphology and the
increase of hydrogen bonding sites could improve the tensile
properties. Thus, the polymer matrices progressively switched to
elastic-plastic behavior and exhibited excellent elongations
higher than 350%. In particular, the D1 series samples even
exhibited elongations higher than 650%, as well as tensile
strengths of approximately 18 MPa. With similar hard segment
contents, the samples with higher generation dendritic diols were
reported to show relatively lower density of dendritic structure.23
As for the D2 series, the slight phase mixing and dilution of
dendritic structure per repeat unit led to relatively inferior
mechanical properties.
Fig. 5 XRD patterns for L-55, D0-55, D1-55 and D2-55 samples.
chain dendritic structure increased, the microstructure
morphology relapsed into slight phase separation. The tangible
evidence is that TgS of the soft segment decreased to approxi-
ꢀ
mately ꢁ55.0 C, which is lower than those of D0 series or D1
series. In addition, a 100% hard segment polymer comprising
only MDI and G2.5 diol exhibited a Tg of 65.0 ꢀC. As antici-
pated, the concurrent hard segment glass-transition relaxation
and soft segment melting along with the increased steric
hindrance would certainly lead to reduced crystallinity of soft
segments.
XRD measurements were carried out to support the above-
mentioned morphology study (Fig. 5). In general, the linear
flexible polymer chains could effortlessly stack well and form
regular structures. Thus, a small peak and two sharp peaks were
observed at 9.8ꢀ, 21.4ꢀ and 23.8ꢀ for the L-55 PU sample.
According to the literature,43–45 the two sharp peaks were
attributed to the (110) and (200) reflections of orthorhombic
crystals from soft segments. However, the crystallized hard
segments showed no explicit diffraction peak. This is ascribed to
the overlapping of the strong soft segment signals. When G0.5
diol was incorporated as the chain extender, the peripheral alkyl
groups of the side chains would slightly merge into the soft
segment forming a metastable structure. Because of this, two
insignificant peaks (19.3ꢀ and 23.8ꢀ) along with an amorphous
halo appeared and differed from those of L-55 PU. As the
generation of dendritic diol increased, the pendant bulky
dendritic structure would prevent the polymer main chains from
forming ordered structures like the linear PUs. Consequently,
X-ray diffraction patterns revealed only an amorphous halo
because of the above-mentioned interruption of regular structure
by the bulky dendritic structure. As for the D2 series, the large
size of hard segments and abundant hydrogen bonding sites
provided by dendritic structures would eventually result in slight
phase mixing and prevent soft segments from crystallizing.
Indeed, X-ray diffraction patterns revealed only an amorphous
halo, which is also consistent with the DSC analysis.
Shape memory effect
In this work, the ꢀtemperature for stress deformation studies was
selected to be 60 C which corresponded to the original melting
temperature of Tone 2060 polyol. The PU film usually collapses
and becomes viscous at temperatures higher than 60 ꢀC in those
Mechanical properties
The number-average molecular weights (Mn) of these PUs were
determined by GPC, ranging from 3.1 ꢂ 104 to 5.8 ꢂ 104 (g/mol)
Fig. 6 Stress-strain behavior of PUs prepared from different dendritic
diols and DEG with similar hard segment contents.
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 8484–8494 | 8491

View Article Online
Table 3 Recovery time and ratio for shape memory test
1st Test
2nd Test
Sample
Rest Angle (ꢀ)
Recovery Time (sec)
Recovery Ratio (%)
Rest Angle (ꢀ)
Recovery Time (sec)
Recovery Ratio (%)
L-45
L-50
L-55
76.5
77.0
75.0
81.5
82.0
83.0
88.0
90.0
90.0
90.0
90.0
90.0
12
10
11
7
8
7
6
4
3
2
85.0
85.6
83.3
90.6
91.1
92.2
97.8
100.0
100.0
100.0
100.0
100.0
76.5
76.0
74.5
83.0
82.0
82.5
87.5
89.0
87.5
90.0
90.0
90.0
13
13
12
8
8
7
5
3
3
2
85.0
84.4
82.8
92.2
91.1
91.7
97.2
98.9
97.2
100.0
100.0
100.0
D0-45
D0-50
D0-55
D1-45
D1-50
D1-55
D2-45
D2-50
D2-55
2
2
2
2
films when the hard segment was less than 40%. The hard
segments of PU play an important role of providing sufficient
physical interactions to sustain permanent shape. On the other
hand, the soft segments act as the switching components to fulfill
external influence and memorize temporary shape after the
deforming progress. The samples were also examined by ther-
momechanical tensile tests (Table 4). The L-series samples were
examined at 0 and 40 ꢀC because the tensile tests were not
shape. Hence, increased recovery ratios and reduced recovery
times could be achieved for the D1 series and D2 series PUs,
particularly the D2 series samples with complete and rapid
recovery. For the thermomechanical tests, the D1 series samples
exhibited good shape fixity of 93.0–95.7% and reasonable shape
recovery of 73.9–81.7%, whereas the D2 series samples exhibited
the best shape fixity and shape recovery of 91.3–99.1% and 91.3–
99.1%, respectively. These results show that the shape memory
polymers produced via incorporating the dendritic structures can
effectively enhance the shape memory properties of the materials.
For cyclic bending tests, the dendron-containing D series PUs
still maintained excellent recovery properties from sufficient and
reversible physical interactions as compared to the L series PUs.
This phenomenon could also be observed for the thermo
mechanical tensile tests. This suggests that the physically
‘‘pseudo-cross-linked’’ interactions via hydrogen bonding can
not only sufficiently stabilize the permanent shape, but also
enhance the mechanical properties of an elastomer.49 In addition,
a series of continuous photographs demonstrates the stepwise
recovery progress of D2-55 PU (Fig. 7). In the beginning, the D2-
55 PU was deformed and wound around a rod at 60 ꢀC and then
cooled to room temperature. This was followed by immersing the
sample in 60 ꢀC water and the sample was photographed every 3
ꢀ
applicable at 60 C for these samples.
As shown in Table 3, the L series samples achieved only an
average recovery ratio of 85%, and a recovery time greater than
10 seconds. Apparently, as shown in Table 4, the L series samples
show good shape fixity, but rather poor shape recovery. The
good shape fixity of the L series samples is because the soft
segments exhibited a TgS around ꢁ50 ꢀC. This indicates that the
ꢀ
amorphous phase remained in a rubbery state at 0 C and thus
underwent immediate shrinkage after the release of the tensile
load, while the crystalline phase helped retain the stretched
shape. The poor shape recovery is due to the presence of phase-
segregated morphology and the shortage of hydrogen bonding
interactions. With the introduction of dendrons to the PUs,
however, the shape recovery increases due to the peripheral alkyl
chains of dendron which promote the miscibility between hard/
soft segments and subsequently induce phase-mixing
morphology. In addition, the hydrogen bonding rich malona-
mide dendrons were also capable of reinforcing the physical
cross-linking effect of hard segments. By the shape memory
bending test, the recovery ratio of the D0 series was increased to
90%, whereas the recovery time was decreased to less than 10
seconds. Similar shape memory behavior was also observed in
the thermomechanical tensile tests, The shape recovery of the D0
series was increased to 96.5%–99.1%. However, the shape fixity
of the D0 series was decreased to 34.8%–40.9%. This is because
the operating temperatures (20 ꢀC and 60 ꢀC) were within the
range of TgH and TmS of the D0 series samples. Therefore, the
samples were under relatively large stress during the loading and
subsequent cooling process. This would cause a large shrinkage
upon removing the grip from the samples and subsequently
induce a large recovery during the reheating process.
Table 4 Shape fixity and shape recovery of the dendritic PUs determined
by thermomechanical cycle tests
Shape Fixity (%)
Shape Recovery (%)
1st
2nd
3rd
1st
2nd
3rd
L-45a
L-50
95.7
96.0
96.8
98.3
94.4
96.8
—
38.3
36.5
94.8
91.3
94.8
96.5
92.2
87.0
97.4
94.4
96.0
—
37.4
37.4
94.8
91.3
93.9
94.8
93.0
85.2
63.5
43.2
36.8
—
96.5
99.1
73.9
79.1
81.7
91.3
99.1
93.0
63.5
44.0
34.4
—
97.4
97.4
73.9
78.3
80.0
88.7
92.2
91.3
62.6
45.6
36.0
—
97.4
97.4
73.9
77.4
79.1
87.8
93.0
92.2
L-55
b
D0-45
D0-50
D0-55
D1-45
D1-50
D1-55
D2-45
D2-50
D2-55
—
40.9
34.8
93.0
93.0
95.7
99.1
97.4
91.3
As the size of dendritic structure and the number of hydrogen
bonding sites increased, the hard segments would further self-
associate to provide sufficient strength in maintaining permanent
a
Examined at 40 ꢀC/10 ꢀC. b Not applicable.
8492 | J. Mater. Chem., 2009, 19, 8484–8494
This journal is ª The Royal Society of Chemistry 2009

View Article Online
Fig. 7 A series of continuous photographs for the stepwise recovery progress of D2-55 PU.
seconds. Full recovery of D2-55 PU required less than 10 seconds
without any creases or defects. It is important to note that the
dendron-containing D2 series PUs displayed excellent shape
fixity and shape recovery comparable to those of the most
advanced chemical cross-linking type shape memory polymers.49–51
These thermally reversible properties and long term reliability of
the dendron-containing PUs are highly effective in enhancing
shape memory effect, providing a new route to prepare shape
memory polymers via the incorporation of dendritic structures.
and Technology, Taiwan. This work is also supported in part by
the Ministry of Education, Taiwan under ATU plan.
References
1 C. M. Yakacki, R. Shandas, C. Lanning, B. Rech, A. Eckstein and
K. Gall, Biomaterials, 2007, 28, 2255.
2 A. Lendlein and R. Langer, Science, 2002, 296, 1673.
3 D. J. Maitland, A. P. Lee, D. L. Schumann and L. D. Silva, US Pat., 6
102 917, 2000.
4 J. Kunzelman, T. Chung, P. T. Mather and C. Weder, J. Mater.
Chem., 2008, 18, 1082.
€
5 A. Lendlein, H. Jiang, O. Junger and R. Langer, Nature, 2005, 434,
879.
4
Conclusions
A series of side chain dendritic PUs with excellent mechanical
and shape memory properties has been synthesized. The
peripheral alkyl chains of malonamide dendron would improve
miscibility between the hard and soft segments. Furthermore,
strong hydrogen bonding interactions could further reinforce
hard segment association. The variations of hydrogen bonding-
induced microphase morphology were investigated by FT-IR,
DSC, and XRD measurements. During the cyclic shape memory
tests, the reversible physical cross-linking segments of side chain
dendritic PUs significantly improved the recovery ratio and time.
In particular, the D2 series sample effortlessly achieved complete
and rapid recovery. By incorporating the dendritic chain
extenders derived from MDI, this study is able to provide a novel
route to modify PUs with excellent shape memory effect. These
results clearly show that mechanical and shape memory prop-
erties could be simultaneously enhanced in these side-chain
dendritic PUs. It is important to point out that few modified
systems are able to achieve the improvements simulta-
neously.13,15–22
6 C. Liu, H. Qin and P. T. Mather, J. Mater. Chem., 2007, 17, 1543.
7 S. Kelch, S. Steuer, A. M. Schmidt and A. Lendlein,
Biomacromolecules, 2007, 8, 1018.
8 F. Li, W. Zhu, X. Zhang, C. Zhao and M. Xu, J. Appl. Polym. Sci.,
1999, 71, 1063.
9 S. Rezanejad and M. Kokabi, Eur. Polym. J., 2007, 43, 2856.
ꢀ
10 P. Miaudet, A. Derre, M. Maugey, C. Zakri, P. M. Piccione,
R. Inoubli and P. Poulin, Science, 2007, 318, 1294.
11 Y. Shirai and S. Hayashi, Mitsubishi Tech. Bull., 1988, 184, 213.
12 K. Kobayashi and S. Hayashi, US Pat., 5 128 197, 1992.
13 P. Ping, W. Wang, X. Chen and X. Jing, Biomacromolecules, 2005, 6,
587.
14 K. J. Kim and M. Shahinpoor, Polymer, 2002, 43, 797.
15 B. K. Kim, S. Y. Lee, J. S. Lee, S. H. Baek, Y. J. Choi, J. O. Lee and
M. Xu, Polymer, 1998, 39, 2803.
16 B. C. Chun, T. K. Cho and Y.-C. Chung, Eur. Polym. J., 2006, 42,
3367.
17 E. Wornyo, K. Gall, F. Yang and W. Kingc, Polymer, 2007, 48, 3213.
18 T. Chung, A. Romo-Uribe and P. T. Mather, Macromolecules, 2008,
41, 184.
19 M. Nagata and Y. Yamamoto, J. Polym. Sci., Part A: Polym. Chem.,
2009, 47, 2422.
20 P. T. Mather, H. G. Jeon, A. Romo-Uribe, T. S. Haddad and
J. D. Lichtenhan, Macromolecules, 1999, 32, 1194.
21 Q. Q. Ni, C. S. Zhang, Y. Fu, G. Dai and T. Kimura, Compos. Struct.,
2007, 81, 176.
22 M. K. Jang, A. Hartwig and B. K. Kim, J. Mater. Chem., 2009, 19,
1166.
23 C. P. Chen, S. A. Dai, H. L. Chang, W. C. Su and R. J. Jeng, J. Polym.
Sci., Part A: Polym. Chem., 2005, 43, 682.
Acknowledgements
The authors are very grateful for the financial support from
National Science Council and Chung Shan Institute of Science
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 8484–8494 | 8493

View Article Online
24 C. P. Chen, S. A. Dai, H. L. Chang, W. C. Su, T. M. Wu and
R. J. Jeng, Polymer, 2005, 46, 11849.
38 C. M. Brunette, S. L. Hsu and W. J. MacKnight, Macromolecules,
1982, 15, 71.
25 S. A. Dai, C. P. Chen, C. C. Lin, C. C. Chang, T. M. Wu, W. C. Su,
H. L. Chang and R. J. Jeng, Macromol. Mater. Eng., 2006, 291, 395.
26 C. C. Tsai, T. Y. Juang, S. A. Dai, T. M. Wu, W. C. Su, Y. L. Liu and
R. J. Jeng, J. Mater. Chem., 2006, 16, 2056.
39 L. T. J. Korley, B. D. Pate, E. L. Thomas and P. T. Hammond,
Polymer, 2006, 47, 3073.
40 H. M. Jeong, S. Y. Lee and B. K. Kim, J. Mater. Sci., 2000, 35,
1579.
27 T. Y. Juang, C. C. Tsai, T. M. Wu, S. A. Dai, C. P. Chen, J. J. Lin,
Y. L. Liu and R. J. Jeng, Nanotechnology, 2007, 18, 205606.
28 Y. C. Chen, T. Y. Juang, S. A. Dai, T. M. Wu, J. J. Lin and R. J. Jeng,
Macromol. Rapid Commun., 2008, 29, 535.
29 H. Frauenrath, Prog. Polym. Sci., 2005, 30, 325.
30 C. X. Sun, M. A. J. van der Mee, J. G. P. Goossens and M. van Duin,
Macromolecules, 2006, 39, 3441.
31 S. A. Dai, T. Y. Juang, C. P. Chen, H. Y. Chang, W. J. Kuo, W. C. Su
and R. J. Jeng, J. Appl. Polym. Sci., 2007, 103, 3591.
32 H. C. Lin and S. K. Wu, Scr. Metall. Mater., 1992, 26, 59.
33 Y. P. Cao, Y. Guan, J. Du, J. Luo, Y. X. Peng, C. W. Yip and Albert
S. C. Chan, J. Mater. Chem., 2002, 12, 2957.
41 S. A. Greenberg and T. Alfrey, J. Am. Chem. Soc., 1954, 76, 6280.
42 J. L. Lee, E. M. Pearce and T. K. Kwei, Macromolecules, 1997, 30,
6877.
43 C. S. P. Sung, C. B. Hu and C. S. Wu, Macromolecules, 1980, 13, 111.
44 Y. Zhang, V. r. Leblanc-Boily, Y. Zhao and R. E. Prud’homme,
Polymer, 2005, 46, 8141.
45 C. Prisacariu, R. H. Olley, A. A. Caraculacu, D. C. Bassett and
C. Martin, Polymer, 2003, 44, 5407.
46 K. E. Uhrich, C. J. Hawker and J. M. J. Frechet, Macromolecules,
1992, 25, 4583.
47 D. C. Lee, T. A. Speckhard, A. D. Sorensen and S. L. Cooper,
Macromolecules, 1986, 19, 2383.
34 G. Q. Liu, X. B. Ding, Y. P. Cao, Z. H. Zheng and Y. X. Peng,
Macromolecules, 2004, 37, 2228.
35 B. K. Kim, S. Y. Lee and M. Xu, Polymer, 1996, 37, 5781.
36 C. S. P. Sung and N. S. Schneiderl, Macromolecules, 1977, 10, 452.
37 M. C. Kuo, R. J. Jeng, W. C. Su and S. A. Dai, Macromolecules, 2008,
41, 682.
48 C. S. P. Sung, C. B. Hu and C. S. Wu, Macromolecules, 1980, 13, 117.
49 J. Li, J. A. Viveros, M. H. Wrue and M. Anthamatten, Adv. Mater.,
2007, 19, 2851.
50 N. Choi and A. Lendlein, Soft Matter, 2007, 3, 901.
51 M. K. Jang, A. Hartwig and B. K. Kim, J. Mater. Chem., 2009, 19,
1166.
8494 | J. Mater. Chem., 2009, 19, 8484–8494
This journal is ª The Royal Society of Chemistry 2009