Polymer International

Polym Int 49:17141721 (2000)

Temperature sensitive water vapour

permeability and shape memory effect of
polyurethane with crystalline reversible phase
and hydrophilic segments
Han Mo Jeong,1* Byoung Kun Ahn2 and Byung Kyu Kim2
1

Research Centre for Machine Parts and Materials Processing, Department of Chemistry, University of Ulsan, Ulsan 680-749, Republic of
Korea
2
Department of Polymer Science and Engineering, Pusan National University, Pusan 609-749, Republic of Korea

Abstract: Polyurethanes based on poly(caprolactone) (PCL) diol, hexamethylene diisocyanate, 4,4'diphenylmethane diisocyanate and hexamethylene diamine were modied by hydrophilic segments,
diol-terminated poly(ethylene oxide) or dimethylol propionic acid (DMPA). Differential scanning
calorimetry, dynamic mechanical tests, tensile tests, and measurement of water vapour permeability
were carried out to characterize these polyurethanes. Temperature sensitive water vapour permeability, that is, the abrupt increase of water vapour permeability at the melting temperature of the PCL
phase, was enhanced by modication with hydrophilic segments. Fatigue in shape memory effects was
minimized by introducing some amount of DMPA units into the polyurethane chain.
# 2000 Society of Chemical Industry

Keywords: polyurethane; water vapour permeability; shape memory; thermomechanical properties

INTRODUCTION

Shape memory polymers basically consist of two

phases, a frozen phase and a reversible phase.13
Segmented polyurethanes can present a microphase
separated structure due to incompatibility of the
constituent segments.4 In our previous papers, we
synthesized segmented thermoplastic polyurethanes
(TPUs) from diol-terminated poly(caprolactone)
(PCL diol) as a soft segment, 4,4'-diphenylmethane
diisocyanate (MDI) as a diisocyanate, and 1,4-butane
diol or 4,4'-bis-(o-hydroxy alkoxy) biphenyl as chain
extenders.57 In these TPUs, rubber elasticity was
observed in the temperature range between the
melting temperature of the soft segments (Tms) and
the melting temperature of the hard segments (Tmh)
because the micro-Brownian motion of the soft
segments and the restricted molecular motion due to
the crystalline hard segment phase. When they were
deformed in the temperature range between Tms and
Tmh and subsequently cooled below Tms under
constant strain, the deformed shape was xed because
the micro-Brownian movement was frozen. When they
were reheated within the temperature range TmsTmh,
a shape memory effect was observed, ie. the original
shape was recovered through the elastic force generated during deformation.
In the application of waterproof polyurethanes, the

temperature dependency of water vapour permeability

is an important factor to be considered for the effective
utilization of the materials. For example, for improved user comfort of textiles coated with waterproof
polyurethane, the coated polyurethane should have
high water vapour permeability at higher temperature
and low water vapour permeability at lower temperature.8,9 A hydrophilic segment such as poly(ethylene
oxide) (PEO) has high water vapour permeability,10
which increase signicantly when the micro-Brownian
motion of the polymer chains is activated in the
temperature range above its glass transition temperature (Tg) or its melting temperature (Tm).11
In this study, shape memory polyurethanes based on
PCL diol (number average molecular weight
4000 g mol1; PCL 4000), hexamethylene diisocyanate (HDI), MDI, and hexamethylene diamine
(HDA), were modied to obtain enhanced and
temperature sensitive moisture permeability by introducing a hydrophibilic segment, diol-terminated PEO
(number average molecular weight 200 g mol1, PEG
200) or dimethylol propionic acid (DMPA). The
temperature dependency of water vapour permeability
and shape memory effect were studied. During
synthesis, to maximize the effect of micro-Brownian
motion of the PCL segments due to phase transition
(melting) on the chain mobility of the hydrophilic

* Correspondence to: Han Mo Jeong, Research Centre for Machine Parts and Materials Processing, Department of Chemistry, University of
Ulsan, Ulsan 680-749, Republic of Korea
Contract/grant sponsor: Korea Research Foundation; contract/grant number: 1998-017-E00196
(Received 9 May 2000; revised version received 29 June 2000; accepted 31 July 2000)

# 2000 Society of Chemical Industry. Polym Int 09598103/2000/$30.00

1714

Water vapour permeability and shape memory of PU

segments, the monomeric materials were fed into the

reactor in the following sequence to produce closely
linked PCL and hydrophilic segments: PCL 4000
HDI PEG200 or DMPA MDI HDA.

DMPA was used as a hydrophilic segment instead of

PEG 200. The IPA series was obtained by neutralizing
the carboxylic acid group of DMPA unit in the NPA
series with the same number of moles of triethylamine.
Neutralization was carried out for 2 h after the
polymerization reaction. The nal polymer concentration in DMF was about 25 wt%. The feed mole ratio of
HDI MDI/PCL 4000 PEG 200 (or DMPA)
HDA was kept at 1.0/1.0 to obtain a linear polymer.
In the sample designation code for PEG 1 and PEG 2
series, the number following PEG means the weight
percentage of PCL 4000 HDI in the monomer feed.
For example, the weight percentage of PCL 4000 and
HDI in the monomer feed for the synthesis of PEG 60
is 60.0 wt%. The feed mole ratio of PCL 4000/HDI
was kept at 2.0/1.0 for all the polyurethanes in Table 1
to produce the dimer of PCL 4000 at the initial stage
of polymerization.

EXPERIMENTAL
Materials

PCL 4000 and PEG 200 were dried under vacuum at

50 C for 12 h. DMPA was dried under vacuum at
100 C for 3 h. MDI was melted at 45 C and used after
excluding the white MDI dimer precipitated from the
melt. HDI, HDA, triethylamine and dibutyltin dilaurate were used as received. N,N-Dimethylformamide
molecular
(DMF) was used after dehydration with 4 A
sieves for 2 days.
Synthesis of polyurethanes

A 500 ml round-bottomed, four-necked separable

ask equipped with a mechanical stirrer, nitrogen
inlet, thermometer and condenser with drying tube
was used as the reactor. Polyurethanes were synthesized by solution polymerization in DMF under dry
N2. For the synthesis of PEG 1 and PEG 2 series in
Table 1, PCL diol was reacted with HDI at 60 C for
30 min and for another 2 h at 80 C in the presence of
dibutylin dilaurate (0.03 % wt wise) as a catalyst.
DMF was added to the reactor occasionally when the
viscosity of the reaction mixture became too high.
Then, PEG 200 and MDI were fed into the reactor
sequentially and reacted for another 1 h at 80 C.
Finally, HDA was fed dropwise into the reactor at
60 C and reacted for 1 h. The NPA series in Table 1
was synthesized by the same procedure except that

Characterization

The viscosity of polyurethane solution in DMF was

measured with an Ostwald viscometer at 25 C.
Differential scanning calorimetry (DSC) was carried
out with a TA Instruments DSC-2100 at a heating and
cooling rate of 10 C min1. After melting for 3 min at
200 C, a cooling scan down to 100 C and a
subsequent heating scan up to 200 C were carried
out. All runs were carried out with samples of about
5 mg.
Films for the measurement of mechanical properties
were prepared by casting the polyurethane solution in
DMF on a glass plate. After evaporation of DMF at
80 C for 24 h, the lms were further dried at 100 C
under vacuum for the next 24 h.

Table 1. Characteristics of polyurethanes prepared

Sample
designation
PEG 1 series
PEG50
PEG60
PEG70
PEG80
PEG 2 series
PEG70
1PEG70
2PEG70
NPA series
NPA1
NPA2
NPA3
IPA series
IPA1
IPA2
IPA3
a

Content in feed (wt%)

Content in feed (mol)

PCL 4000

HDI

PEG 200

DMPA

MDI

HDA

PCL 4000

HDI

PEG 200

DMPA

MDI

HDA

Intrinsic
viscosity
(dlg1)

48.9
58.7
68.5
78.3

1.1
1.3
1.5
1.7

18.8
14.8
10.7
6.7

29.7
24.0
18.4
12.7

1.5
1.2
0.9
0.6

2.0
2.0
2.0
2.0

1.0
1.0
1.0
1.0

15.4
10.1
6.3
3.3

18.5
12.5
8.2
4.8

2.1
1.4
0.9
0.5

0.769
0.769
0.769
0.385

68.5
68.5
68.5

1.5
1.5
1.5

10.7
9.5
6.9

18.4
18.7
19.5

0.9
1.8
3.6

2.0
2.0
2.0

1.0
1.0
1.0

6.3
5.5
4.1

8.2
8.3
8.7

0.9
1.8
3.6

0.769
0.769
0.769

68.5
68.5
68.5

1.5
1.5
1.5

1.5
3.0
4.5

21.4
21.2
21.1

7.2
5.8
4.5

2.0
2.0
2.0

1.0
1.0
1.0

1.3
2.6
3.9

9.5
9.4
9.3

7.2
5.8
4.4

0.385
0.385
0.385

68.5
68.5
68.5

1.5
1.5
1.5

1.5a
3.0a
4.5a

21.4
21.2
21.1

7.2
5.8
4.5

2.0
2.0
2.0

1.0
1.0
1.0

1.3a
2.6a
3.9a

9.5
9.4
9.3

7.2
5.8
4.4

1.154
1.154
1.154

Neutralized by triethylamine.

Polym Int 49:17141721 (2000)

1715

HM Jeong, BK Ahn, BK Kim

Figure 1. DSC thermograms of PEG 1

series obtained on cooling (upper
thermograms with exothermic peaks)
and subsequent heating (lower
thermograms with endothermic peaks)
scans: (a) PEG 50; (b) PEG 60; (c)
PEG 70; (d) PEG 80.

Dynamic mechanical properties were determined

with a dynamic mechanical thermal analyser (Rheometry Scientic DMTA MK III), using a bending
mode at a heating rate of 4 C min1 and 10 Hz.
Tensile tests were carried out using a tensile tester
(Tinius Olsen Series 1000), attached to a constant
temperature heating chamber. Microtensile specimens
for test had the dimensions 25 mm length, 5 mm width
and 1 mm thickness.
Water vapour permeability (WVP) was measured
according to ASTM D1653-93. That is, an open
cup containing water was sealed with the cast lm of
polyurethane, and the assembly was placed in a
test chamber with a controlled atmosphere. The
steady water vapour ow (per unit time through
unit area of the lm induced by unit vapour
pressure difference between the two surfaces) was
measured by plotting the weight change of the
cup containing water against elapsed time. The
thickness of the cast lm was 0.03 mm and the
1716

test chamber was maintained at a relative humidity

of 50  2 %.

RESULTS AND DISCUSSION

Thermal properties

The thermograms of all the polyurethanes used in this

study showed the exothermic crystallization peak and
endothermic melting peak of the PCL segment on
cooling and heating scans, respectively, as shown in
Fig 1. However, no distinct melting or crystallization
peak of the hard segment was observed by DSC. In Fig
1 and Table 2, we can see that the crystallization
temperature (Tc) and the heat of crystallization (DHc)
measured on the cooling scan, and the melting
temperature (Tms) and the heat of fusion (DHms)
measured on the subsequent heating scan, generally
decrease as the content of PEG 200 increases in the
PEG 1 and PEG 2 series. This shows that the
crystallization of PCL segments is hindered by some
Polym Int 49:17141721 (2000)

Water vapour permeability and shape memory of PU

Table 2. Thermal properties of the

polyurethanes

Sample

Tc ( C)

DHc (J g1PCL segment)

Tms ( C)

DHmg (J g1 PCL segment)

PEG 50
PEG 60
PEG 70
PEG 80

5.0
5.2
8.4
14.4

19.7
28.5
33.7
47.2

48.5
48.7
48.9
50.1

25.6
36.8
39.7
49.6

1PEG 70
2PEG 70

10.9
18.4

39.0
42.4

50.1
51.9

41.4
42.1

NPA 1
NPA 2
NPA 3

20.3
20.9
22.3

41.4
41.6
41.9

49.1
50.6
50.8

41.4
41.6
42.1

IPA 1
IPA 2
IPA 3

18.8
19.2
20.4

39.2
41.2
43.0

49.1
49.5
49.8

39.2
41.2
43.1

amount of PEG segment mixed in the PCL phase,12,13

and the degree of this phase mixing is increased as the
content of PEG segments in polyurethane is increased.
However, when DMPA instead of PEG 200 is used as
the hydrophilic segment, we can see in Table 2 that all
the Tc, DHc, Tms and DHms values generally increase as
the content of DMPA is increased, in both the NPA
and IPA series. This shows that the miscibility
between hard and PCL segments is reduced when
the content of DMPA is increased.
All the polyurethanes showed broad heat capacity
change at the glass transition of the PCL phase at
around 50 C; however, this change was not distinct
enough to measure and discuss quantitatively.

Dynamic mechanical properties

The changes in tensile storage modulus E ' of polyurethanes on heating are shown in Fig 2. A gradual
decrease of E ' in the temperature range between the
glass transition temperature (Tgs, about 50 C) and
Tms of PCL segment, a sudden drop at Tms, and a
rubbery plateau up to 150200 C can be observed.
Figure 2(a) shows that, as the content of PCL segment
is increased with an accompanying decrease of the
other segments contents, both the crystalline state
modulus below Tms and the crystalrubber transition
temperature (Tms) increase, and the rubbery plateau
modulus above Tms decreases. The increased amount
of crystalline PCL phase, the increase of Tms (see

Figure 2. Tensile storage modulus of: (a) PEG 1 series; (b) PEG 2 series; (c) NPA series; (d) IPA series.

Polym Int 49:17141721 (2000)

1717

HM Jeong, BK Ahn, BK Kim

Because the intermolecular hydrogen bond of the urea

linkage is stronger than the urethane linkage,14 the
reduced intermolecular interaction when HDA is
replaced by DMPA seems to be the cause of the
rubbery plateau modulus decrease. The decrease of
the rubbery plateau modulus caused by DMPA units is
reduced in Fig 2(d), when the DMPA unit is
neutralized by triethylamine. Increased cohesion of
hard segments due to the strong coulombic force of the
ionic sites seems to be the reason for this.15 The
increased crystalline modulus and crystalrubber
transition temperature, when HDA is replaced by
DMPA (Fig 2c, d), seem to be the result of the
increase in crystallinity (DHms) and Tms of the PCL
phase (see Table 2), caused by the reduced miscibility
between hard and PCL segments.
Tensile properties

Figure 3. Stressstrain curves of: (a) PEG 70; (b) 1PEG 70; and (c) 2PEG
70.

Table 2) by reduced phase mixing, and the decreased

amount of rigid hard segment that can sustain the
rubbery state seem to be the respective causes of the
above results. In the PEG 2 series (Fig 2b), we can see
that both the crystalline modulus and rubbery plateau
modulus decrease as the content of PEG 200 is
increased. The reduction in DHms caused by the
miscibility of PEG 200 and PCL segments seems to
be the cause of the decrease in the crystalline state
modulus. The reduced amount of hard segment seems
to cause the decrease of the rubbery plateau modulus.
Figure 2(c) shows that the rubbery plateau modulus
decreases as some HDA is replaced by DMPA.

Table 3. Mechanical properties of the

polyurethanes

1718

Stressstrain curves for the PEG 2 series are shown in

Fig 3 and the tensile properties measured at 25 C are
summarized in Table 3. The 3 % secant modulus and
yield strength generally decrease in both the PEG 1
and PEG 2 series when the content of PEG 200 in
polyurethane is increased. In the NPA and IPA series,
it can be seen in Table 3 that both the 3 % secant
modulus and the yield strength generally increase
when HDA is replaced by DMPA, probably due to the
increased crystallinity of the PCL domain (see Table
2). However, the relatively small increase of DHms,
when HAD is replaced by DMPA, compared with the
evident variation of various properties of the NPA and
IPA series, ie crystalline modulus in Fig 2(c, d), 3 %
secant modulus and yield strength in Table 2, water
vapour permeability in the following Fig 3(c, d), leaves
room for doubt that the increased crystallinity of the
PCL phase might not be the only factor causing these
variations of physical properties.
Water vapour permeability

The variations of water vapour permeability of polyurethanes on heating are shown in Fig 4. It can be

Sample

3% secant
modulus (MPa)

Yield strength
(MPa)

Tensile strength
(MPa)

Elongation at
break (%)

PEG 50
PEG 60
PEG 70
PEG 80

3.62
3.80
3.86
4.17

5.17
5.47
5.61
5.72

45.1
20.5
22.6
17.8

800
780
860
800

1PEG 70
2PEG 70

6.01
6.61

7.91
9.49

31.3
28.6

780
630

NPA 1
NPA 2
NPA 3

6.33
7.61
8.03

7.96
10.82
11.90

48.9
63.7
66.0

760
700
680

IPA 1
IPA 2
IPA 3

7.60
7.95
8.95

10.45
11.43
12.09

50.5
65.1
73.3

790
700
820

Polym Int 49:17141721 (2000)

Water vapour permeability and shape memory of PU

Figure 4. Water vapour permeability of polyurethanes: (a) PEG 1 series; (b) PEG 2 series; (c) NPA series; (d) IPA series.

generally observed that the water vapour permeability

increases abruptly at Tms, and this phenomenon is
more evident when the content of hydrophilic PEG
200 or DMPA units in the polyurethane is increased.
This shows that when the mobility of the PCL
segments is enhanced by melting, the mobility of
closely linked hydrophilic PEG 200 or DMPA
segments is also increased making a path along which
water vapour molecules can easily pass.11 Thus, Fig 4
shows that the melting of the hydrophobic PCL phase
can trigger water vapour transfer through the closely
linked hydrophilic segments by enhancing the chain
mobility. The results shown in Fig 4(c, d) show that
DMPA units play their role as a water vapour barrier at
temperatures below Tms by increasing the crystallinity
of PCL segments and their role as a path for water
vapour transfer at temperatures above Tms thanks to
their hydrophilicity.
Shape memory effect

The results of cyclic tensile tests to characterize the

shape memory effect of polyurethanes are shown in
Fig 5.57 The sample was elongated at 65 C (in the
rubbery state) to 100% strain (em) at a constant
elongation rate of 10 mm min1. While maintaining
the strain at em, the sample was cooled to 25 C (to its
crystalline state) and unloaded. Upon removing the
constraint at 25 C, a small recovery of strain to eu
occured. The sample was subsequently heated to
65 C in 5 min, and kept at that temperature for the
next 10 min, allowing recovery of strain. This completes one thermomechanical cycle (N = 1) leaving a
residual strain, ep, where the next cycle (N = 2) starts.
In Fig 5, the PEG 2 series has ep values of around 20 %
and shows the fatigue that occurs on repeated deformaPolym Int 49:17141721 (2000)

tions when these values increase. These results are

similar to those of other shape memory polyurethanes
that we have studied previously.57 However, we can
observe that this fatigue behaviour is minimized in NPA
and IPA series when some of the HDA is replaced by
DMPA. NPA 3 shows almost identical stressstrain
curves for repeated thermomechanical cycles. Because
fatigue by repeated deformations in polyurethanes
stems from softhard phase mixing and orientation,16,17
these results show that changes in phase mixing or
segmental orientation upon repeated deformations are
minimized when DMPA units are introduced into the
polyurethane chain.

CONCLUSIONS

When polyurethanes based on PCL 4000, HDI, MDI

and HDA were modied by hydrophilic segments
(PEG 200 or DMPA) which are closely linked to
PCL 4000 segments the following changes were
observed:
(1) PEG 200 segments reduced the crystallization of
PCL segments through phase mixing, whereas
DMPA segments enhanced the crystallization of
PCL segments by reducing the miscibility between hard and PCL segments.
(2) Temperature sensitive water vapour permeability
was amplied. That is, water vapour permeability
of the polyurethane lm was triggered at temperatures around 50 C, when the chain mobility of the
PCL segment and closely linked hydrophilic
segment were enhanced by phase transition, ie
melting of the PCL segment occurred.
(3) Fatigue in the shape memory effect was reduced
1719

HM Jeong, BK Ahn, BK Kim

Figure 5. Cyclic tensile behavior of: (a) PEG 2 series; (b) NPA series; (c) IPA series.

by introducing some DMPA units into the polyurethane chain.

ACKNOWLEDGEMENTS

This work was nancially supported by a research

grant (1998-017-E00196) from the Korea Research
Foundation.
1720

REFERENCES
1 Li H-L, Ito K, Ujihira Y, Nanasawa A and Iwamoto T, Kobunshi
Ronbunshu 55:448 (1998).
2 Li F, Chen Y, Zhu W, Zhang X and Xu M, Polymer 39:6929
(1998).
3 Hayashi S, Int Prog Urethanes 6:90 (1993).
4 Gogolewski S, Colloid Polym Sci 267:757 (1989).
5 Kim BK, Lee SY and Xu M, Polymer 37:5781 (1996).

Polym Int 49:17141721 (2000)

Water vapour permeability and shape memory of PU

6 Kim BK, Lee SY, Lee JS, Baek SH, Choi YJ, Lee JO and Xu M,
Polymer 39:2803 (1998).
7 Jeong HM, Lee JB, Lee SY and Kim BK, J Mater Sci 35:279
(2000).
8 Kleinerman GJ, Cell Polym 8:95 (1989).
9 Morioka A, Sen'i Gakkaishi 41:P-410 (1985).
10 Schultze D, US Patent 35840812 (1998).
11 Hayashi S, Ishikawa N and Giordano C, in Polyurethanes World
Congress Proceedings, Technomic, Lancaster, p 400 (1993).
12 Nishi T and Wang TT, Macromolecules 8:909 (1975).

Polym Int 49:17141721 (2000)

13 Ahn TO, Oh M-H, Yoo KS and Jeong HM, Polym Int 36:239
(1995).
14 Wang CB and Cooper SL, Macromolecules 16:775 (1983).
15 Kim SJ, Kim BK and Jeong HM, J Appl Polym Sci 51:2187
(1994).
16 Shibayama M, Kawauchi T, Kotani T, Nomura S and Matsuda
T, Polym J 18:719 (1986).
17 Tang W, MacKnight WJ and Hsu SL, Macromolecules 28:4284
(1995).

1721