Impact of hydrogen bonds dynamics on mechanical behavior of supramolecular elastomer

Polymer 105 (2016) 221e226

Contents lists available at ScienceDirect

Polymer journal homepage: www.elsevier.com/locate/polymer

Impact of hydrogen bonds dynamics on mechanical behavior of supramolecular elastomer Ming-Chao Luo, Jian Zeng, Zheng-Tian Xie, Lai-Yun Wei, Guangsu Huang*, Jinrong Wu** State Key Laboratory of Polymer Material Engineering, College of Polymer Science and Engineering, Sichuan University, Chengdu 610065, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 July 2016 Received in revised form 15 September 2016 Accepted 15 October 2016 Available online 19 October 2016

Here our special consideration is devoted to the relationship between dynamics and mechanical behaviours of supramolecular elastomer (SE) based on 2-ureido-4[1H]-pyrimidinone (UPy) groups. We find that SE exhibits a new relaxation mode (a0 relaxation) which differs from segmental relaxation mode (a relaxation) and normal relaxation mode (NM). Calculated by Arrhenius model, supramolecular interactions are much lower than the bonds energy of covalent bonds; this enables high energy dissipation as the elastomer is subjected to deformation. Moreover, unlike covalent bonds, the hydrogen bonds of UPy groups are dynamic and longer waiting time leads to better re-association efficiency, as evidenced by recovery of hysteresis loop during cyclic tensile tests. This work on the relationship between dynamics and mechanical properties will not only improve the understanding of reversible bonds relaxation, but also provide an idea on preparing mechanically robust SE for us. © 2016 Published by Elsevier Ltd.

Keywords: Dynamics of hydrogen bonds Mechanical behaviors Supramolecular elastomer

1. Introduction Supramolecular polymers are assembled from monomeric building blocks through noncovalent interactions, such as hydrogen bonds and metal-ligand interactions [1]. Among such noncovalent bonds, hydrogen bond is a promising candidate because of its intrinsic directionality and versatility. The structural reversibility, caused by the transient nature of secondary interactions, often endows polymers with a wide range of tunable characteristics including self-healing properties, long stacking structure, and shape memory properties [2e12], but hydrogen bonds as weaker physical interactions lead to supramolecular materials with low mechanical properties, limiting their applications [13e15]. To exploit mechanically robust supramolecular elastomers (SE), it is very essential for us to systematically explore the relationship between hydrogen bonds dynamics and mechanical behaviors. Recently, constructing supramolecular polymers through selfassembly of building blocks has appealed to many researchers' attention. Numerous experimental and theoretical studies have shown that the morphology and domain spacing of separated nanophase can be controlled by varying the degree of

* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (G. Huang), wujinrong@scu. edu.cn (J. Wu). http://dx.doi.org/10.1016/j.polymer.2016.10.042 0032-3861/© 2016 Published by Elsevier Ltd.

polymerization, the volume fraction, and the interaction parameter between blocks [15,16]. However, less attention has been paid to relate the dynamics with mechanical properties of supramolecular polymers. Previous article has analyzed the dynamics of reversible bonds in detail from theoretical viewpoint, which can be described well by hindered reputation model [17]. For example, enhanced relaxation phenomenon, a new relaxation process combined with an intermolecular association and dissociation process, is observed in reversible bonds [18e23]. This model provides very useful method for us to study the dynamics of supramolecular interactions and calculate the activation energy between them. In addition, inspired by the research methods of double networks toughening mechanism [5,24e27], we can probe the weaker bonds rupture by the energy dissipation. We envision that the dynamics of reversible bonds from microscopic perspective can be considered as the cause for the change of mechanical behaviors. Thus, the bridge is developed between supramolecular interactions and mechanical properties. Here telechelic oligoisoprene is reacted with 2-ureido-4[1H]pyrimidinone (UPy) groups, as shown in Scheme 1. The SE is based on the self-complementary quadruple hydrogen bonds of UPy groups. We find that SE exhibits a new relaxation mode (a0 relaxation) which differs from segmental relaxation mode (a relaxation) and normal relaxation mode (NM). Calculated by Arrhenius model, supramolecular interactions is much lower than the bonds energy of covalent bonds. That means supramolecular interactions can preferentially break before that rupture of covalent bonds. Thus we

222

M.-C. Luo et al. / Polymer 105 (2016) 221e226

CH3

H O N C R N N N H H H O N N

O

R

Cl

H N C

O H

O

H 5 IO 6

O

O OH

m

O

m-x

x

n

O

NaBH 4 CH 3 N O

N

OCN

C 6H 12

CH 3

NCO

NH 2

N N

O

H N H

O C

N H

C 6H 12

NCO

HO

n

OH

N CH3 CH 3

Scheme 1. UPy groups.

N O

N

H N H

O C

N H

C 6H 12

N H

O C

CH 3 O

can use energy dissipation to increase mechanical properties such as toughness upon deformation. This work on the relationship between dynamics and mechanical properties will not only improve the understanding of reversible bonds relaxation behavior, but also provide an idea on preparing mechanically robust SE for us.

n

O

O C

N H

C 6H 12 N H

O C

H N H

N N

O

DCP

2. Experimental section 2.1. Materials Polyisoprene (PI, GPC data: Mn ¼ 241468 g/mol, DPI ¼ 3.893) was purchased from Shanghai Sanlian Co., Ltd. 3Chloroperoxybenzoic acid (MCPBA, 85%), periodic acid (99%), sodium borohydride (NaBH4, 98%), DCP (98%), 2-amino-4-hydroxy-6methylpyrimidine (98%), hexyldiisocyanate (98%), and dibutyltindilaurate (95%) were purchased from Adamas Reagent Co., Ltd. Tetrahydrofuran (THF) was purchased from Shanghai Titan Scientific Co., Ltd and used without further purification. 2.2. Preparation of HTPI The synthesis of HTPI was performed according to the previous literature [28,29]. First, we used MCPBA to prepare epoxidized PI in THF. Through oxidative chain cleavage reaction, we used periodic acid to prepare carbonyl telechelic oligoisoprene (CTPI). Through the reduction of CTPI, HTPI was obtained. GPC data of HTPI: Mn ¼ 23000 g/mol. 2.3. Preparation of SE According to the method invented by Sijbesma and Meijer [30], the UPy unit linked to a reactive isocyanate group (UPy-NCO) was prepared. To a solution of HTPI in 700 mL chloroform, UPy-NCO was added. After addition of 2 drops dibutyltindilaurate, the resulting solution was stirred at 60  C for 16 h. After the end of reaction, chloroform solution was cooled to room temperature. Then 1 phr DCP crosslinker (based on 100 phr SE) was added to chloroform solution. After chloroform was removed, the resulting material was cured in a hydraulic press at 160  C for 25 min. The synthesis process is shown in Scheme 2.

Covalent crosslinking UPy dimers Scheme 2. Synthesis process of SE.

FTIR spectra. The wavenumber range was from 4000 cm1 to 650 cm1. 2.4.2. Dynamic mechanical analysis (DMA) Dynamic mechanical properties were measured in a tensile mode on a Q800 DMA (TA Instruments) under a nitrogen atmosphere. The dimensions of the specimens tested were 12 mm  6.5 mm  1 mm. The DMA spectra were performed at heating rate of 3  C/min and a frequency of 1 Hz. The temperature range was 100  C to 100  C and a preload force of 0.01 N was applied.

2.4. Characterization 2.4.1. Fourier transform infrared (FTIR) The FTIR spectra were measured at room temperature using Thermo Scientific Nicolet iS50 FTIR with a resolution of 4 cm1. Attenuated total reflection with SeZn crystal was used to obtain the

2.4.3. Broadband dielectric spectroscopy (BDS) Dielectric measurements were performed over the frequency range of 101e107 Hz on a Novocontrol Concept 50 system with Alpha impedance analyzer and Quatro Cryosystem temperature control. The disk-shaped film of about 1 mm thickness was placed

M.-C. Luo et al. / Polymer 105 (2016) 221e226

223

between two parallel electrodes with 20 mm diameter. The temperature range was from 70  C to 50  C with 10  C intervals. The analyses of the dielectric spectra are made by using empirical equation of Havriliak and Negami (HN) [31]. In this model, the frequency dependence of the dielectric complex (ε*) can be described by

ε* ðuÞ ¼ ε∞ þ


Dε b 1 þ ðiutHN Þa

(1)

where Dε ¼ εs  ε∞ is the dielectric strength, εs and ε∞ are the relaxed and unrelaxed values of dielectric constant, the parameters a and b (0< a, ab  1) define the symmetrical and asymmetrical broadening of the loss peak, and tHN is the characteristic relaxation time. The relation between tHN and tmax is given by Refs. [32,33]



tmax ¼ tHN sin

1a 

pab

2ð1 þ bÞ

sin

pa

1

2ð1 þ bÞ

a

;

fmax ¼

1 2ptmax (2)

where fmax is the frequency at which ε00 passes through the maximum value. This characteristic relaxation time (tmax) obtained from the HN equation fit can be correlated with the temperature through the Vogel-Fulcher-Tamman (VFT) equation [34,35]:



tmax ¼ t0 exp

B T  T0

 (3)

where t0 and B are empirical parameters and T0 is the so-called Vogel temperature. Fig. 2. Relaxation behavior of SE. (a) Tand and E0 measured by DMA; (b) ε00 as a function of frequency for a relaxation; (c) ε00 as a function of frequency for a0 relaxation.

2.4.4. Mechanical analysis Mechanical properties were measured on a universal testing machine (Instron 5966) at room temperature with a cross-head speed of 100 mm/min. The specimen was a dumbbell shaped thin strip with dimension of 25 mm  4 mm  1 mm. For each data, three measurements were carried out and the average value was taken. Incremental loading and unloading cycles were performed with the same experimental setup as tensile tests at 100 mm/min. Cycles were applied between nearly l ¼ 1 to li (l is the local elongation in the tensile direction). To quantify the energy dissipation, we calculate the integrated area of the hysteresis loop. For every cycle, the dissipated energy (U) was calculated from Equation (4).

Z U¼

sdl 

loading

Z

sdl

(4)

unloading

where s is stress. 3. Results and discussion 3.1. Molecular structure analysis To investigate the effect of hydrogen bonds dynamics on mechanical behaviors of SE, we prepare SE with hydroxyl telechelic cis-1,

Fig. 1. Molecular structure analysis. (a) FTIR spectra confirm the forming of SE by reacting HTPI with UPy-NCO; (b) Schematic diagram of fabricating SE by reacting HTPI with UPyNCO.

224

M.-C. Luo et al. / Polymer 105 (2016) 221e226

Fig. 4. Mechanical behaviors of the first wise-step cyclic tensile. (a) Stress-strain curves of SE subjected to a cyclic uniaxial tension with a maximum strain of 250%; (b) Cyclic stress-strain curves during the first wise-step cyclic tensile; (c) Hysteresis loss during first for the first wise-step cyclic tensile.

of eNHe groups, as presented in Fig. 1a. These facts confirm that HTPI is successfully reacted with two UPy termini, as illustrated by Fig. 1b. Fig. 3. The comparison of dielectric analysis between a relaxation and a0 relaxation. (a) The temperature dependence of Dε for a relaxation; (b) The temperature dependence of Dε for a0 relaxation; (c) Temperature dependence of tmax for a relaxation and a0 relaxation. Solid lines represent fitting curves; (d) Arrhenius model for a0 relaxation. Solid line represents fitting curve.

4-oligoisoprene (HTPI) and UPy groups. The HTPI is prepared by oxidative chain cleavage reaction of epoxided PI (EPI) and reduction of carbonyl telechelic oligoisoprene (CTPI), as shown in Fig. S1a. The existence of hydroxyl groups is confirmed by a broad characteristic peak at 3500-3200 cm1 on the FTIR spectrum, as shown in Fig. 1a. This HTPI is reacted with 2(6-isocyanatohexylaminocarbonylamino)6-methyl- 4[1H]pyrimidinone (UPy-NCO, see Fig. S1b) and dicumyl peroxide (DCP) to form SE. FTIR measurements show that after reacting with UPy groups, the characteristic peaks of hydroxyl and isocyanate groups both disappear; meanwhile SE exhibits a characteristic band at 3324 cm1, corresponding to the stretching vibrations

3.2. Changes in relaxation behavior By strong dimerization of UPy groups, oligoisoprenes are bonded end to end to give virtual polymer chains with high molecular weight. To investigate the relaxation behaviors of SE, we perform DMA. We find that glassy region is obviously observed, which is the characteristic for conventional polymers, as shown in Fig. 2a. At higher temperature than glassy regions, the transition from the glassy state to the rubbery state is present. In rubber region, the significant decrease of the storage modulus (E0 ) appears with increasing temperature. Moreover, tand curve exhibits the second weaker transition in the higher temperature, which could be attributed to the disassociation of UPy groups. The changes in molecular dynamics can be further investigated by BDS analysis. The dielectric loss (ε00 ) as a function of frequency

Scheme 3. The equilibrium between disassociated UPy groups and associated UPy groups.

M.-C. Luo et al. / Polymer 105 (2016) 221e226

225

Notably, Dε of a0 relaxation exhibits the opposite trend with the increase of temperature, as presented in Fig. 3b. PI with dipole moment parallel to the backbone shows NM [37,38]. The increase of temperature also leads to the decreased Dε of NM [39]. We compare the temperature dependence of Dε for a0 relaxation and NM. Obviously, a0 relaxation does not belong to NM. UPy dimers, as shown in Scheme 1, are centrosymmetric and should not have any dipole moment. This centrosymmetric structure is not observed by BDS. However, as temperature rises, the associated UPy groups gradually shift to the disassociated state, as illustrated by Scheme 3. This equilibrium shift breaks centrosymmetric structure of UPy dimers, which increases the dipole moment of system. Such process leads to the increase of Dε. Thus, it is reasonable to conclude that a0 relaxation originates from the dissociation of UPy dimers. To further describe a relaxation and a0 relaxation dynamics, we extract the characteristic relaxation time (tmax) from the HN equation. The temperature dependence of tmax can be well described with Vogel-Fulcher-Tamman (VFT) equation, as shown in Fig. 3c. Compared with a relaxation, a0 relaxation time is slower by several orders of magnitude. While a relaxation of polymers conforms to the VFT equation, the polymers containing UPy groups also conform to the Arrhenius relationship. Thus, we use Arrhenius equation to fit relaxation time. The apparent activation energy of about 56 KJ/mol derived from the slope of the plot is similar to the previous works [6,40].

Fig. 5. Understanding of Hydrogen Bonds Disassociation and Re-association from Energy Dissipation Viewpoint. (a) Stress-strain curves for SE after different waiting time; (b) Hysteresis loss with different waiting time; (c) Re-association efficiency for SE.

exhibits loss peak at temperature ranging from 223 K to 273 K, as presented in Fig. 2b. The loss peak is associated with a relaxation and this peak shifts toward higher frequency with increasing temperature. Furthermore, a comparison of selected ε00 spectra for SE from 263 K to 293 K is displayed in Fig. 2c. Interestingly, a new shoulder peak (a0 relaxation) is clearly observed in the frequency range of 100e102 Hz, which apparently differs from a relaxation. 3.3. a0 relaxation To make sure the origin of a0 relaxation, the comparison of dielectric analysis between a relaxation and a0 relaxation is shown in Fig. 3. Fitting the dielectric spectra with two HN equations, we can extract the dielectric strength (Dε) of a relaxation and a0 relaxation. The typical fitting curves for a0 relaxation is shown in Fig. S2. For conventional glass forming materials, a slight decrease of Dε with temperature is characteristic, reflecting the decreasing cooperativity of a relaxation process at higher temperature [36]. Such phenomenon is also observed in SE, as shown in Fig. 3a.

3.4. Energy dissipation analysis The stress on reloading is less than that on the initial loading for the same strain during cyclic loading and unloading process. This stress softening phenomenon is referred to as the well-known Mullins effect [41,42]. Such phenomenon is also observed in Fig. 4a. The bond energies between UPy dimers are about 56 KJ/mol, as confirmed by Fig. 3, which is much lower than 360 KJ/mol of CeC covalent bond. Obviously, these hydrogen bonds should preferentially break before that rupture of covalent bonds upon deformation. We believe that the reversible break and reformation of physical bonds leads to energy dissipation [43e45]. To investigate the relationship between energy dissipation and UPy interactions, we perform wise-step cyclic tensile for SE, as shown in Fig. 4b and c. We calculate the integrated area of the hysteresis loop to quantify the energy dissipation, as shown in Fig. 4c. It can been seen that SE exhibits higher hysteresis loss with higher extension, which means that much more UPy groups are disassociated upon deformation. The reason for this phenomenon is that the hydrogen bonds energy of UPy groups is much lower that of covalent bonds and it does not take much energy to break these hydrogen bonds before the rupture of covalent bonds. Therefore, these suggest that the hydrogen bonds between UPy groups indeed play a role of sacrificial bonds, which effectively dissipate energy upon deformation.

Scheme 4. The mechanism of hydrogen bonds disassociation and re-association.

226

M.-C. Luo et al. / Polymer 105 (2016) 221e226

3.5. Understanding of hydrogen bonds disassociation and re-association from energy dissipation viewpoint Unlike covalent bonds that irreversibly break upon deformation, hydrogen bonds can reform after releasing of stress. To investigate the disassociation and re-association of hydrogen bonds, we perform repeated cyclic tensile tests after different waiting time, as shown in Fig. 5a. We find that the first cyclic test of a fresh sample which has not been previously deformed shows remarkable hysteresis loop, but the second cyclic test to the same maximum strain immediately after the first one has much less hysteresis loss, as shown in Fig. 5a. Such phenomenon can be attributed to the fact that most of the disassociated hydrogen bonds cannot reform during this time scale. However, as the waiting time increases, the hysteresis loop is gradually recovered due to the re-association of UPy groups, as illustrated by Scheme 4. Meanwhile, the hysteresis loss curve gradually approaches to the original one (the first stepcycle hysteresis loss curve) with waiting time increasing. It suggests that the longer waiting time leads to higher the efficiency of hydrogen bonds re-association, as show in Fig. 5b. To quantify the efficiency of hydrogen bonds re-association, we define it as the ratio of the hysteresis loss of the repeated cyclic tensile tests to that of the first cyclic tensile test. This re-association efficiency is plotted as a function of time, as shown in Fig. 5c. Due to the slow dynamics and strong direction of UPy groups, short waiting time leads to low re-association effect. As waiting time increases, the re-association efficiency of hydrogen bonds increases. 4. Conclusion In this article, we investigate the effect of hydrogen bonds dynamics on mechanical properties of SE. The HTPI is reacted with UPy groups to prepare SE which is covalently crosslinked by DCP. Compared with a relaxation and NM, SE exhibits another new relaxation behavior (a0 relaxation), which is attributed to the association and disassociation of UPy groups. Calculated by Arrhenius model, supramolecular interactions are much lower than the bonds energy of covalent bonds; this enables higher energy dissipation as the elastomer is subjected to deformation. Finally, due to the dynamic of hydrogen bonds, longer waiting time leads to better hydrogen bonds re-association efficiency. Acknowledgements This work was financially supported by National Natural Science Foundation of China (grant No. 51333003) and Special Fund for Agro-scientific Research in the Public Interest (grant No. 201403066). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.polymer.2016.10.042. References [1] L. Brunsveld, B.J.B. Folmer, E.W. Meijer, R.P. Sijbesma, Chem. Rev. 101 (2001)

4071e4098. [2] Y. Chen, A.M. Kushner, G.A. Williams, Z. Guan, Nat. Chem. 4 (2012) 467e472. [3] K.E. Feldman, M.J. Kade, E.W. Meijer, C.J. Hawker, E.J. Kramer, Macromolecules 42 (2009) 9072e9081. [4] M.G. McKee, C.L. Elkins, T. Park, T.E. Long, Macromolecules 38 (2005) 6015e6023. €velmann, C. Weiss, A. Radulescu, J. Allgaier, W. Pyckhout[5] B.J. Gold, C.H. Ho Hintzen, A. Wischnewski, D. Richter, Polymer 87 (2016) 123e128. [6] C.L. Lewis, K. Stewart, M. Anthamatten, Macromolecules 47 (2014) 729e740. [7] A. Shabbir, H. Goldansaz, O. Hassager, E. van Ruymbeke, N.J. Alvarez, Macromolecules 48 (2015) 5988e5996. chambre, C. Fonteneau, S. Pensec, J.M. Chenal, L. Chazeau, [8] X. Callies, C. Ve L. Bouteiller, G. Ducouret, C. Creton, Macromolecules 48 (2015) 7320e7326. [9] J.R. Kumpfer, S.J. Rowan, J. Am. Chem. Soc. 133 (2011) 12866e12874. [10] J. Li, J.A. Viveros, M.H. Wrue, M. Anthamatten, Adv. Mater. 19 (2007) 2851e2855. [11] T.L. Chantawansri, I.-C. Yeh, A.J. Hsieh, Polymer 81 (2015) 50e61. [12] I.-C. Yeh, B.C. Rinderspacher, J.W. Andzelm, L.T. Cureton, J. La Scala, Polymer 55 (2014) 166e174. [13] P. Cordier, F. Tournilhac, C. Soulie-Ziakovic, L. Leibler, Nature 451 (2008) 977e980. -Ziakovic, F. Tournilhac, L. Leibler, J. Polym. [14] D. Montarnal, P. Cordier, C. Soulie Sci. Part A Polym. Chem. 46 (2008) 7925e7936. [15] J. Scavuzzo, S. Tomita, S. Cheng, H. Liu, M. Gao, J.P. Kennedy, S. Sakurai, S.Z.D. Cheng, L. Jia, Macromolecules 48 (2015) 1077e1086. [16] R.H. Zha, B.F. de Waal, M. Lutz, A.J. Teunissen, E.W. Meijer, J. Am. Chem. Soc. 138 (2016) 5693e5698. [17] L. Leibler, M. Rubinstein, R.H. Colby, Macromolecules 24 (1991) 4701e4707. [18] M. Müller, E.W. Fischer, F. Kremer, U. Seidel, R. Stadler, Colloid Polym. Sci. 273 (1995) 38e46. [19] M. Müller, U. Seidel, R. Stadler, Polymer 36 (1995) 3143e3150. [20] A. Dimopoulos, J.-L. Wietor, M. Wübbenhorst, S. Napolitano, R.A.T.M. van Benthem, G. de With, R.P. Sijbesma, Macromolecules 43 (2010) 8664e8669. [21] M.-C. Luo, X.-K. Zhang, J. Zeng, X.-X. Gao, G.-S. Huang, Polymer 91 (2016) 81e88. [22] K. Xing, S. Chatterjee, T. Saito, C. Gainaru, A.P. Sokolov, Macromolecules 49 (2016) 3138e3147. [23] M. Hayashi, S. Matsushima, A. Noro, Y. Matsushita, Macromolecules 48 (2015) 421e431. [24] J.P. Gong, Y. Katsuyama, T. Kurokawa, Y. Osada, Adv. Mater. 15 (2003) 1155e1158. [25] H.R. Brown, Macromolecules 40 (2007) 3815e3818. [26] R.E. Webber, C. Creton, H.R. Brown, J.P. Gong, Macromolecules 40 (2007) 2919e2927. [27] E. Ducrot, Y. Chen, M. Bulters, R.P. Sijbesma, C. Creton, Science 344 (2014) 186e189. bir, I. Campistron, A. Laguerre, J.-F. Pilard, C. Bunel, J.-P. Couvercelle, [28] N. Ke C. Gondard, Polymer 46 (2005) 6869e6877. bir, G. Morandi, I. Campistron, A. Laguerre, J.-F. Pilard, Polymer 46 (2005) [29] N. Ke 6844e6854. [30] B.J.B. Folmer, R.P. Sijbesma, R.M. Versteegen, J.A.J. van der Rijt, E.W. Meijer, Adv. Mater. 12 (2000) 874e878. [31] S. Havriliak, S. Negami, Polymer 8 (1967) 161e210. [32] R. Díaz-Calleja, Macromolecules 33 (2000), 8924e8924. [33] J. Liu, S. Wu, Z. Tang, T. Lin, B. Guo, G. Huang, Soft Matter 11 (2015) 2290e2299. [34] P. Ortiz-Serna, R. Díaz-Calleja, M.J. Sanchis, G. Floudas, R.C. Nunes, A.F. Martins, L.L. Visconte, Macromolecules 43 (2010) 5094e5102. [35] Z. Tang, L. Zhang, W. Feng, B. Guo, F. Liu, D. Jia, Macromolecules 47 (2014) 8663e8673. €nhals, F. Kremer, E. Schlosser, Phys. Rev. Lett. 67 (1991) 999e1002. [36] A. Scho [37] K. Adachi, T. Kotaka, Macromolecules 17 (1984) 120e122. [38] K. Adachi, T. Kotaka, Prog. Polym. Sci. 18 (1993) 585e622. [39] D. Boese, F. Kremer, Macromolecules 23 (1990) 829e835. € ntjens, R.P. Sijbesma, M.H.P. van Genderen, E.W. Meijer, J. Am. Chem. [40] S.H.M. So Soc. 122 (2000) 7487e7493. [41] L. Mullins, Rubber Chem. Technol. 42 (1969) 339e362. [42] R.W. Ogden, D.G. Roxburgh, Proc. R. Soc. A Math. Phys. Eng. Sci. 455 (1999) 2861e2877. [43] J. Liu, Z. Tang, J. Huang, B. Guo, G. Huang, Polymer 97 (2016) 580e588. [44] J. Huang, Z. Tang, Z. Yang, B. Guo, Macromol. Rapid Commun. 37 (2016) 1040e1045. [45] Z. Tang, J. Huang, B. Guo, L. Zhang, F. Liu, Macromolecules 49 (2016) 1781e1789.