Improved shape memory performance of star-shaped POSS-polylactide based polyurethanes (POSS-PLAUs)

Abstract

Star-shaped polyhedral oligomeric silsesquioxane multi-arm polylactides (POSS-PLA) with various arm lengths were synthesized by ring opening polymerization of D,L-lactide. Then, the star-shaped POSS-PLA based polyurethanes (POSS-PLAUs) were synthesized by cross-linking POSS-PLAs with various PLA arm lengths and polytetramethylene ether (PTMEG) with hexamethylene diisocyanate (HDI). Fourier transform infrared spectroscopy (FTIR) and nuclear magnetic resonance (NMR) were utilized to characterize the structures of the materials. Differential scanning calorimetry (DSC) experiments were carried out to investigate the glass transition temperature (T_g) of POSS-PLAUs and crystallinity of PTMEG. Dynamic mechanical analysis (DMA) and stress relaxation experiments were used to study the dynamic mechanical properties and stress relaxation behaviors of POSS-PLAUs. Cyclic thermal mechanical and physical shape recovery tests were used to study the shape memory properties. The effects of POSS and PLA arm length on the mechanical, thermal properties and the shape memory behaviors were investigated. In our system, PLA arm length has no significant effect on glass transition temperature (T_g). T_g increases with the POSS content in the networks due to the obstructed movement of the polymer chains caused by caged POSS. E′ increases from POSS-PLAU110 to POSS-PLAU530 due to the decreasing of PTMEG content. Dynamic mechanical analysis reveals a similar relationship of glass transition temperature to POSS content. The stress relaxation curves show an increase in initial stress in POSS-PLAUs with longer arm length due to lower PTMEG content. The relaxation ratio is higher for the polymer with shorter PLA arm length. The quick relaxation above the triggering temperature of POSS-PLAU with shorter PLA arm length is favorable for quick shape recovery. All the POSS-PLAUs have excellent shape memory properties with high shape fixity ratios above 99%, shape recovery ratios around 84% for the first cycle and above 89% for the second cycle. POSS-PLAUs with the shorter arm length show faster recovery speed due to the higher content of POSS cores.

---

1. Introduction

A shape memory polymer (SMP) is a stimuli responsive material which can rapidly change its shape from a temporary shape to its original shape under an external stimulus such as heat,^1–3 electricity^4–6 alternating magnetic field,^7–9 microwave radiation,^10–12 solvent^13,14 etc. Since the mid-1980s, SMPs have been developing rapidly. So far, various SMPs, such as polyurethanes,^1–4,6 shape memory epoxy resins,¹⁵ styrene-based shape memory polymers,¹⁶ shape memory polynorbornenes,⁹ shape memory cyanates¹⁷ etc. have been reported. However, SMPs have not been widely used due to their scientific and technical barriers, such as low mechanical properties or poor shape memory properties. Incorporation of inorganic fillers or networks, especially nanoparticles with high modulus to SMPs is a common method to improve mechanical and shape memory properties.^1,18

In the past decade, organic–inorganic hybrid materials have received significant attention for their greatly improved physical and mechanical properties or new functional properties at relatively low inorganic filler loading levels comparing with conventional microcomposites.^19,20 Polyhedral oligomeric silsesquioxane (POSS), as a hybrid material, consists of silicon and oxygen atoms arranged in an inner eight-cornered cage with the Si atoms positioned at the corners. The inorganic core of POSS has a diameter of approximately 0.45 nm.²¹ Each POSS molecule contains organic functionalities on its eight reactive sites which makes the POSS nanostructure compatible with various polymer matrices,^21,22 such as polynorbornene,²³ polyurethane,^3,4 epoxy resin,²⁴ poly(methyl methacrylate),²⁵ poly(4-methylstyrene)²⁶ and siloxane polymers.²⁷ Generally, the mechanical properties, thermal stability, oxidative resistance, anti-flammability and surface hardness of the polymers can be improved due to the incorporation of POSS.

Polyurethanes (PUs) have various applications in daily life such as hard plastics, flexible films, coatings, adhesives, or elastomers.²⁸ Polyurethane block copolymers are a typical shape memory polymers received increasing attention, owing to their adjustable transition temperatures and good biocompatibility.²⁹ Basically, shape memory polyurethanes (SMPUs) consist of two segments, hard segments and soft segments. The soft segments form the reversible phase through molecular motion in a rubbery state, while the hard segments form chemical crosslinks or physical crosslinks due to polar interaction, hydrogen bonding or crystallization in the hard domains.³⁰ Ionic networks may be suitable for forming the permanent networks and also the temporary networks due to their long relaxation time and thermal liability.³¹ The effect of molecular weight of soft segments and ionomers on the mechanical and shape memory behaviors of PUs were discussed by Kim et al. It was reported that ionomers gave higher hardness, modulus and strength, higher recovery strain and lower residual strain due to coulombic forces.³²

In order to enhance the shape memory performance, the cross-linking SMPUs consisting of organic segments or inorganic units have been reported.^33–37 Tensile modulus and strength of SMPUs were improved due to the nanoscale reinforcement of incorporated POSS cores. Triple-shape memory properties of star-shaped polyhedral oligomeric silsesquioxane-poly(3-caprolactone) polyurethanes (SPOSS-PUs) were reported by Bothe et al.³⁵ It was revealed that the triple-shape properties of SPOSS-PUs networks considerably depended on the chain length of PCL networks. Improved shape fixities, recovery abilities and stress storage capabilities were revealed at high POSS content for star-shaped POSS-polycaprolactone polyurethanes.³⁶ Excellent mechanical properties and stable temporary shape fixing at room and body temperature, fast and complete shape recovery and tunable bioactivities of POSS-PLA networks were reported by Xu et al.³⁷ POSS cores led to rapid shape recovery compared with a simple organic crosslinker.³⁷

PLA is one of the most intensively studied biodegradable polymers because of its good mechanical properties. Shape memory PLA and its copolymers or blends have been widely investigated. Poly(L-lactic acid) (PLLA) homopolymer and poly(L-lactide-co-ε-caprolactone) exhibited good shape memory behaviors.³⁸ To improve the mechanical and shape memory properties of PLA, hydroxyapatite was solution blended and grafted.^39,40 Shape memory behaviors had been improved obviously due to the incorporation of stereocomplex crystallite network of PLLA and poly(D-lactic acid) (PDLA) as reported by Li et al.⁴¹ Magnetic-field-induced shape memory PLLA was reported by Zhang et al. The elastic modulus, tensile strength, elongation at break and the shape memory properties were greatly improved due to the incorporation of Fe₃O₄ nanoparticles.⁴² The effect of arm length on the thermal, mechanical properties and shape memory behaviors of shape memory properties of branched polylactide with inositol as core were investigated. The six-arm star-shaped polylactide showed short recovery time, high recovery ratio and fixity ratio.⁴³

In this work, to obtain excellent polylactide-based shape memory polymers and broaden the properties or the applications of polylactide-based shape memory polymers, a series of star-shaped POSS-polylactide based polyurethanes (POSS-PLAUs) were synthesized. Flexible polytetramethylene ether (PTMEG) chains and POSS cores were introduced into the networks. The effects of POSS content and PLA arm length on the thermal, mechanical and shape memory behaviors were investigated. The star-shaped POSS-PLAUs were expected to have better shape recovery performances compared with normal PLA-based polyurethanes.

2. Experimental

2.1. Materials

(3-Chloropropyl)trimethoxysilane (98%), silver nitrate (AgNO₃, AR), sodium hydroxide (NaOH, AR), alcohol (AR), tetrahydrofuran (AR), methanol (AR), hydrochloric acid (HCl, AR), methylene dichloride (AR) and toluene (AR) were purchased from Sinopharm Chemical Reagent Co., Ltd. D,L-lactide was purchased from Jinan Daigang biological engineering Co., Ltd. Stannous octoate (Sn(Oct)₂, 95%) and hexamethylene diisocyanate (HDI, 99%) were purchased from Aldrich. Polytetramethylene ether (PTMEG, 2000 g mol⁻¹) was purchased from Mitsubishi Chemical Corp.

2.2. Synthesis of octa-(3-hydroxypropyl) polyhedral oligomeric silsesquioxane (POSS-(OH)₈)

Octa-(3-chloropropyl) polyhedral oligomeric silsesquioxane (POSS-Cl₈) was synthesized by the method described by Liu et al.⁴⁴ 79.5 g (0.4 mol) 3-chloropropyltrimethoxysilane, 1800 ml methanol and 90 ml concentrated hydrochloric acid were mixed and the mixture was stirred at room temperature for at least 5 weeks. 13.4 g (yield: 25.9%) colorless crystals were obtained after being washed by deionized water and dried in vacuo at 65 °C for 48 h. The synthesis route is presented in Scheme 1. ¹H NMR (DMSO-d₆, ppm): 3.58–3.62 (m, SiCH₂CH₂CH₂Cl); 1.73–1.80 (m, SiCH₂CH₂CH₂Cl); 0.75–0.79 (m, SiCH₂CH₂CH₂OH) (see ESI† for ¹H NMR spectrum).

Scheme 1 Synthesis route of octa-(3-chloropropyl) polyhedral oligomeric silsesquioxane (POSS-Cl₈).

Octa-(3-hydroxypropyl) polyhedral oligomeric silsesquioxane (POSS-(OH)₈) was synthesized by the hydrolysis of POSS-Cl₈ in the presence of fresh silver oxide (Ag₂O).⁴⁴ Ag₂O was prepared by dissolving AgNO₃ first (2.355 g, 13.95 mmol) in 10 ml deionized water and aqueous NaOH (0.56 g, 14.08 mmol) dropwise with vigorous stirring. The silver oxide precipitates were centrifuged and washed with deionized water for at least three times. POSS-Cl₈ (1.50 g, 1.44 mmol), ethanol (38 ml) and tetrahydrofuran (38 ml) were mixed in a round-bottom flask to form a transparent solution. Then, the newly prepared Ag₂O together with 1 ml deionized water were added to the solution. The reactive system was refluxed with vigorous stirring in the dark for 48 h. The solution was evaporated by rotary evaporation to obtain white solids (1.12 g) with a yield of 87% after filtering. The synthesis route is presented in Scheme 2. ¹H NMR (DMSO-d₆, ppm): 3.58–3.62 (m, SiCH₂CH₂CH₂OH); 3.39 (s, SiCH₂CH₂CH₂OH), 1.73–1.80 (m, SiCH₂CH₂CH₂OH); 0.75–0.79 (m, SiCH₂CH₂CH₂OH) (see ESI† for ¹H NMR spectrum).

Scheme 2 Synthesis route of octa-(3-hydroxypropyl) polyhedral oligomeric silsesquioxane (POSS-(OH)₈).

2.3. Synthesis of star-shaped POSS-PLAs

A series of star-shaped POSS-PLAs were prepared through ring-opening polymerization (ROP) of D,L-lactide catalyzed by stannous(II) octoate [Sn(Oct)₂]. D,L-Lactide (7.3 g, 0.05 mol), POSS-(OH)₈ (0.087 g, 0.104 mmol), Sn(Oct)₂ (0.10 wt%) were added into a round-bottom flask with a magnetic stir bar. The air in the flask was exchanged with argon. The reactive mixture was heated at 130 °C for 24 h. The crude products were dissolved with a small amount of chloroform and precipitated in an excessive amount of cold hexane. The precipitation procedure was repeated for 3 times and the polymers were dried in vacuo at 40 °C for 48 h. The number of repeating units of PLA arm was calculated according to integral area ratio from ¹H NMR spectra. A series of POSS-PLAs with different composition ratios (POSS-PLAn, n = 110, 220, 420, 530, the number of lactide repeating units) were prepared in the same way. The synthesis route is presented in Scheme 3. ¹H NMR (CDCl₃, ppm): 3.56–3.52 (SiCH₂CH₂CH₂–PDLLA), 1.72–1.71 (SiCH₂CH₂CH₂–PDLLA), 0.82–0.78 (SiCH₂CH₂CH₂–PDLLA), 5.23–5.14 (O–CO–CH(CH₃)–O), 1.59–1.55 (O–CO–CH(CH₃)–O). FTIR (ATR, cm⁻¹): 2958, 2860 (ν-CH); 1748 (ν-C=O); 1183 (ν-O–CO) (see ESI† for ¹H NMR and FTIR spectra).

Scheme 3 Synthesis route of star-shaped POSS-PLA.

2.4. Synthesis of star-shaped POSS-PLAUs

Pre-prepared POSS-PLA, PTMEG (the molar ratio of POSS to PTMEG is 1 : 8, i.e., the molar ratio of PLA arm to PTMEG is 1 : 1) and an appropriate amount of HDI (the ratio of HDI to OH group is 1.02) were dissolved in dried chloroform (0.17 g ml⁻¹) in a round-bottom flask with a magnetic stir bar. A catalytic amount of Sn(Oct)₂ was added into this system when the reactants were completely dissolved. The solution was stirred at 25 °C for 6 h before being poured into Teflon molds. The solvent was evaporated at room temperature for 24 h under N₂ and the material was further cross-linked at 80 °C for 24 h. The residual uncross-linked reactants were removed by washing with chloroform. Then the material was dried in vacuo for 12 h to obtain the final POSS-PLAU films. The synthesis route is presented in Scheme 4. FTIR (ATR, cm⁻¹): 3400 (ν-NH) 2958, 2860 (ν-CH); 1748 (ν-C=O); 1183 (ν-O–CO) (see ESI† for ¹H NMR and FTIR spectra).

Scheme 4 Synthesis route of star-shaped POSS-PLAU.

2.5. Methods

Fourier transform infrared spectroscopy (FTIR) spectra were obtained using a BRUKER AVATAR 360 ESP FT-IR by ART mode at room temperature (25 °C). The wavelength range is 500–4000 cm⁻¹. ¹H NMR spectra were carried out using a Bruker DMX 500 NMR spectrometer with CDCl₃ as the solvent at room temperature (25 °C). Tetramethylsilane (TMS) was used as an internal standard for the analysis of chemical shifts. The number of repeating units of PLA arm was calculated according to integral area ratio of methane proton peak (δ = 5.23–5.14 ppm, CH) to the proton peak next to the terminated hydroxyl groups (δ = 3.56–3.52 ppm, CH) from ¹H NMR spectra. The detailed results of the number of repeating units and M_n of PLA arms are summarized in Table 1. Differential scanning calorimetry (DSC) experiments were carried out using TA Q100. The specimens were heated to 180 °C at a rate of 30 °C min⁻¹ and kept 180 °C for 2 min. Then the specimens were cooled down to −60 °C at a rate of 10 °C min⁻¹. After being kept at −60 °C for 2 min, the specimens were heated to 180 °C at a heating rate of 10 °C min⁻¹. The transition temperatures were obtained from the second heating scan. Dynamic mechanical properties such as storage modules (E′) and tangent value of loss angle (tan δ) were determined using DMA-800 in tensile mode at a frequency of 1 Hz, a strain amplitude of 15.00 μm and a pre-load force of 0.01 N. Rectangular samples with a dimension of 10 mm × 4 mm × 0.4 mm were used. The samples were quenched to 0 °C and equilibrated for 5 min, then ramped to 100 °C at a heating rate of 3 °C min⁻¹. Stress relaxation experiments were also carried out in tensile mode. The rectangular samples were stretched to a strain of 50% at a strain rate of 3% per second at 80 °C. Then the samples were held at 50% strain while the stress relaxation occurred. The values of stress were recorded with time. The relaxation ratio was defined as (σ₀ − σ_t)/σ₀, where σ₀ is the initial stress and σ_t is the stress at time t.

Table 1 Components and properties of POSS-PLAUs

Sample name	The number of repeating unit of PLA arm	Mn (g mol^−1)	POSS (wt%)	PLA (wt%)	PTMEG (wt%)	Tg (°C)
POSS-PLAU110	110	7900	1	79	20	55.7
POSS-PLAU220	220	15 800	0.58	88.3	11.12	54.3
POSS-PLAU420	420	30 200	0.32	93.5	6.18	53.8
POSS-PLAU530	530	38 100	0.26	94.8	4.94	52.8

---

As reported by Xu et al.,³⁷ the cyclic thermal mechanical recovery properties were carried out using DMA-800 in a force control mode. The rectangular samples were isothermal at 20 °C for 2 min. A preload force of 1 mN was applied to the samples. Then the temperature was ramped to 80 °C at a heating rate of 3 °C min⁻¹ and kept isothermal for 5 min. The samples were stretched at 80 °C with a stress ramping rate of 0.2 MPa min⁻¹ from its original shape at the beginning of the N^th testing cycle, ε_p(N − 1), to the elongated shape under a final tensile stress. The temperature was rapidly cooled to 0 °C with the stress loaded, and the sample length was recorded as ε_l(N) (the strained sample length at the lower temperature at the N^th cycle). Then the load force was released to 1 mN after being held at 0 °C for 5 min, and the sample length was recorded as ε_u(N) (the sample length at the lower temperature after unloading the tensile stress at the N^th cycle). At last, the temperature was ramped to 80 °C at a heating rate of 3 °C min⁻¹ and kept isothermal for 30 min, and the final sample length was recorded as ε_p(N) (the recovered length at the N^th cycle or permanent length at the beginning of the (N + 1)^th cycle). Shape memory performance can be evaluated by the strain fixing ratio (R_f), and the strain recovery ratio (R_r). R_f and R_r in a given cycle N were determined by the following formulas:

The shape recovery procedures were also observed in a hot water bath. The materials were cut into rectangular samples with a dimension of 40 mm × 4 mm × 0.4 mm. After being bended to ‘‘U’’ shape at 55 °C, the samples were quenched in ice-water to obtain the temporary shape with the deformation force loaded. The samples were then put in the hot water bath to recover to their original shapes and their shape recovery procedures were recorded.

3. Results and discussion

3.1. Thermal properties

POSS-PLAU110, POSS-PLAU220, POSS-PLAU420, POSS-PLAU530 were synthesized based on the POSS-PLAn (n = 110, 220, 420, 530, the number of repeating units of PLA arms) with different PLA arm length. M_n of each PLA arm and the weight content of POSS cores, PLA arms and PTMEG chains were calculated and presented in Table 1 (the molar ratio of PLA arm to PTMEG is 1 : 1). Fig. 1 presents the second heating trace of POSS-PLAUs. Broad melting peaks in the heating scan for POSS-PLAU110 and POSS-PLAU220 (shown by arrows in Fig. 1) are ascribed to the melting of PTMEG segments. No melting peaks are observed in the heating scan for the POSS-PLAU420 and POSS-PLAU530, which might be due to the more significant influence of longer PLA arms on PTMEG's crystallization. As shown in Fig. 1, glass transition temperature (T_g) of PLA chain segments decreases from 55.7 °C to 52.8 °C from POSS-PLAU110 to POSS-PLAU530 as the increasing of PLA arm length. The result is inconsistent with the result of six-arm shaped polylactide by Xie et al.⁴³ T_g of six-arm shaped polylactide increased with increasing molecular weight of PLA chains.⁴³ The inconsistency might be due to the different range of molecular weight of PLA arms. Molecular weight of the reported six-arm PLAs located in the range from 9000 to 29 000 g mol⁻¹, and 1500 to 4833 g mol⁻¹ for each PLA arm. T_g increases with molecular weight in the lower molecular weight areas. No dependence of T_g from molecular weight appeared for PLA with higher molecular, ranging from 20 000 to 66 000 g mol⁻¹ reported by Perego et al.⁴⁵ In our work, the molecular weight of PLA arms (from 7900 to 38 100 g mol⁻¹) is much higher than those reported by Xie et al., no dependence of T_g from molecular weight should be expected. The higher glass transition temperature of POSS-PLAUs with shorter PLA arm length might be due to the higher content of POSS in the system as shown in Table 1. An increase in the content of the caged POSS contributes large volume obstructed movement of the polymer chains as reported by Ni et al.⁴⁶ The shape memory behaviors of SMPs are mainly determined by the fixed and reversible phases. In our system, the fixed phases are ascribed to hydrogen bonded urethane linkages, entanglement of macromolecules and the cross-linking of POSS cores and crystalline PTMEG for POSS-PLAU110 and POSS-PLAU220. The amorphous PLA phase and PTMEG phase play the role of the reversible phases. Thus, the glass transition of the amorphous PLA chains and chain segments acts as the switch for the control of the shape fixity and recoverability.

Fig. 1 DSC curves of the POSS-PLAUs.

3.2. Dynamic mechanical properties

The storage modulus (E′) and tan δ as a function of temperature are presented in Fig. 2. The storage modulus descends steeply when the temperature increases above T_g. At low temperatures (T < T_g), E′ increases as the increasing of PLA arm length which is inconsistent with the result obtained by Bothe et al.³⁵ This might be due to the higher weight content of PTMEG in POSS-PLAU with shorter PLA arm length as shown in Table 1. Tan δ represents the ratio of the loss modulus E′′ (not shown here) to storage modulus (E′). Generally, tan δ peak corresponds to the temperature of glass transition. The narrow glass transitions as shown in Fig. 2b indicate that homogenous networks formed in all POSS-PLAUs.³⁷ Tan δ peaks occur at the temperature range from 60 to 80 °C, which seems to be inconsistent with the glass transition temperature obtained by DSC in Fig. 1. However, the temperature where the storage modulus begins to decrease is in the range from 45 to 60 °C, which is consistent with the glass transition temperature obtained by DSC in Fig. 1. Dynamic mechanical analysis reveals a similar relationship of glass transition temperature to POSS content. The temperature at the tan δ peak decreases as the increasing of PLA arm due to lower POSS cross-linking density.

Fig. 2 DMA curves of the POSS-PLAUs. (a) Storage modulus (E′), (b) tan δ.

3.3. Stress relaxation behaviors

The stress relaxation curves of the POSS-PLAUs at 50% strain at 80 °C (above switching temperature) are presented in Fig. 3a. The stress relaxation curves consist of two zones which can be named the “fast relaxation zone” and the plateau zone afterwards. Stress decreases fast in the first 10 minutes (fast relaxation zone) and the rate of decreasing becomes slighter after 20 minutes (the plateau zone). Among the factors which influence the shape recovery, the “fast relaxation zone” plays the key role.⁴⁷ The initial stress increases with the increasing of PLA arm length from 0.46 MPa to 0.69 MPa for POSS-PLAU110 to POSS-PLAU530. The higher initial stress of POSS-PLAUs with longer PLA arm length is due to less PTMEG weight content in POSS-PLAUs. For all POSS-PLAUs, the stress does not decay to zero and reaches a plateau value due to the cross-linking of POSS cores.

Fig. 3 Stress relaxation (a) and relaxation ratios (b) of POSS-PLAUs at 80 °C.

Relaxation ratios are calculated and shown in Fig. 3b. For all the POSS-PLAUs, the stress decays fast in the first 10 minutes. The stress decays about 48% for POSS-PLAU110 and POSS-PLAU220, 44% for POSS-PLAU420 and POSS-PLAU530. After 20 minutes, the relaxation becomes very slow. As shown in Fig. 3b, the relaxation ratio is higher for the polymers with shorter PLA arm length due to more PTMEG weight content. The quick relaxation above trigger temperature is favorable for quick shape recovery.

3.4. Shape memory properties

The stress–strain–temperature diagrams of POSS-PLAUs are presented in Fig. 4. Two consecutive cycles were performed to examine repeatability. The shape fixity ratio (R_f) and shape recovery ratio (R_r) for the first and second cycles are calculated and shown in Fig. 5. R_f and R_r are normally used to describe the shape memory properties of SMPs. R_r qualifies the ability of the material for memorizing its original shape. The excellent micro-phase separation contributes to good shape recovery performance. The degree of cross-linking is also one of the factors which influence the shape recovery performance for chemical cross-linking SMPs. R_f describes the mechanical deformation fixing ability of the switching segments which is related to two factors.¹ The first one is the amount of ‘unlocked’ oriented chains and chain segments. When the deformation is occurred, all chains can be regarded as experiencing the same level of deformation in the stretching step. In our system, the orientation of these stretched chains is preserved by the hydrogen bonded urethane linkages, the entanglement of macromolecules and the cross-linking of POSS cores and crystalline PTMEG for POSS-PLAU110 and POSS-PLAU220. However, not all the chains and chain segments are locked. A certain fraction of the chains and chain segments preserve mobility. These ‘unlocked’ chains generate an instantaneously retractive force when the tensile load is removed. The second factor is the modulus of the material at low temperature. When the low temperature modulus is high, once the external force is removed, instantaneous recovery strain is formed by instantaneously retractive force. Combination of the two factors makes it hard to achieve 100% shape fixity ratio. Fig. 5a presents the shape fixity of POSS-PLAUs. All the samples show high shape fixity above 99% (nearly 100%), higher than the system without cross-linking POSS cores (95%).^1,38 High shape fixity is also reported for POSS-PLA networks by Xu et al. due to the cross-linking networks formed by POSS cores.³⁷

Fig. 4 Typical cyclic thermomechanical tensile tests for (a) POSS-PLAU110, (b) POSS-PLAU220, (c) POSS-PLAU420 and (d) POSS-PLAU530, plotted as stress–strain–temperature diagrams.

Fig. 5 Cyclic shape fixity ratios (a) and recovery ratios (b) of the POSS-PLAUs.

The driving force for shape recovery of SMPs is of entropic origin as ramified by elasticity of polymer chains. SMPs possess structures similar to elastomeric networks. The structures are usually of multiphase materials that contain fixed (or hard) phases and reversible (or soft) phases. In our case, the hydrogen bonded urethane linkages, entanglement of macromolecules and the cross-linking of POSS cores and crystalline PTMEG for POSS-PLAU110 and POSS-PLAU220 serve as the fixed phases (hard segments) and the amorphous PLA and PTMEG chains and chain segments serve as the reversible phases. The hard segments play an important role of providing sufficient physical interaction to sustain permanent shape. Moreover, the soft segments act as the switching components to fulfil external response and memorize temporary shape after the deforming progress. T_g of the soft segments can be thought as the switch temperature. The conformational rearrangement of the reversible phase induced by stretching deformation is severely restricted at temperatures below T_g, hence chains cannot recover. The recovery temperature can be set a little above T_g of PLA chains and chain segments. As shown in Fig. 4, POSS-PLAUs started to recover its shape around 60 °C. When the temperature reached 60 °C or above, the frozen oriented chains retracted and the conformational entropy of the system reduced. Large scale movement of the chain backbone experiences results in the shape recovery.⁴⁸ The POSS-PLAUs with different PLA arm length have good shape recovery performance. R_r of all the POSS-PLAUs is around 84% for the first cycle. There is a significant increase in R_r in cycle 2 comparing with cycle 1. The materials have higher R_r for cycle 2, which is above 90%. Comparing with the polylactide-based thermoplastic shape memory polymers, POSS-PLAUs have excellent shape recovery properties.^1,38 All POSS-PLAUs exhibit high shape fixity ratio above 99% for the first and second tensile cycles (Fig. 5b). The difference of arm length in POSS-PLAUs has no significant effect on the shape fixity of SMPs. R_f decreases in cycle 2 which might be due to the reasons that certain inner structure was remarkably destroyed or the entanglement of polymer chains was unwound during the stretching processes.³⁶ The above results showed that all the POSS-PLAUs showed good shape memory properties, such as fast recovery speed, good recovery and excellent fixity.

The physical shape fixing and recovery procedures were also observed. The quenched ‘‘U’’ shape could be fixed completely at room temperature (25 °C) for all POSS-PLAUs. After being immersed in hot water, images of the shape recovery procedures were recorded every 2 seconds and presented in Fig. 6. All samples nearly recovered to its original shapes in about 8 s, showing a good shape recovery performance at 55 °C. However, sample with the shorter arm length (POSS-PLAU110) showed faster recovery speed which is consistent with the higher relaxation ratios as shown in Fig. 3b. The quick relaxation is favorable for quick shape recovery.

Fig. 6 The unconstrained bend recovery of the POSS-PLAUs at 55 °C.

4. Conclusions

A series of star-shaped POSS-polylactide-based polyurethanes (POSS-PLAUs) with shape memory properties were prepared. Octa-(3-hydroxypropyl) polyhedral oligomeric silsesquioxane was synthesized first. Then, the star-shaped POSS-polylactides (POSS-PLAs) with various PLA arm lengths were synthesized through the ring-opening polymerization. At last, the star-shaped POSS-PLAUs were synthesized by cross-linking POSS-PLA and PTMEG with HDI. The influences of POSS content and PLA arm length on the shape memory behaviors were investigated. In our system, PLA arm length has no significant effect on glass transition temperature (T_g). T_g of the POSS-PLAUs decreased from 55.7 °C to 52.8 °C from POSS-PLAU110 to POSS-PLAU530 due to the obstructed movement of the polymer chains caused by caged POSS. As detected by DMA, E′ increases from POSS-PLAU110 to POSS-PLAU530 due to the decreasing of PTMEG content. Dynamic mechanical analysis reveals a similar relationship of glass transition temperature to POSS content. The stress relaxation curves show an increasing of initial stress in POSS-PLAUs with longer arm length due to lower PTMEG content. Relaxation ratio is higher for the polymer with shorter PLA arm length. The quick relaxation above trigger temperature of POSS-PLAU with shorter PLA arm length is favorable for quick shape recovery. R_r of all the POSS-PLAUs are around 84% for the first cycle. There is a significant increase in R_r in cycle 2 compared with cycle 1. All POSS-PLAUs exhibit high R_f, above 99% for the first and second tensile cycles. However, there are no significant changes in R_r and R_f for all the POSS-PLAUs. All the POSS-PLAUs have excellent shape memory properties. POSS-PLAUs with shorter arm length show faster recovery speed.

Acknowledgements

The work was supported by the national key technology R&D program (Grant No. 2012BAI17B05).

Notes and references

1. B. Yan, S. Gu and Y. Zhang, Eur. Polym. J., 2013, 49, 366.
2. S. Gu, B. Yan, L. Liu and J. Ren, Eur. Polym. J., 2013, 49, 3867.
3. T. M. Filion, J. Xu, M. L. Prasad and J. Song, Biomaterials, 2011, 32, 985.
4. M. Xie, L. Wang, J. Ge, B. Guo and P. X. Ma, ACS Appl. Mater. Interfaces, 2015, 7, 6772.
5. H. F. Lee and H. H. Yu, Soft Matter, 2011, 7, 3801.
6. G. Fei, C. Tuinea-Bobe, D. Li, G. Li, B. Whiteside, P. Coates and H. Xia, RSC Adv., 2013, 3, 24132.
7. Y. Cai, J. S. Jiang, B. Zheng and M. R. Xie, J. Appl. Polym. Sci., 2013, 127, 49.
8. M. Y. Razzaq, M. Behl, K. Kratz and A. Lendlein, Adv. Mater., 2013, 25, 5730.
9. S. Bai, H. Zou, H. Dietsch, Y. C. Simon and C. Weder, Macromol. Chem. Phys., 2014, 215, 398.
10. H. Li, H. Li and H. Tzou, J. Vib. Acoust., 2015, 137.
11. K. Yu, Y. Liu and J. Leng, RSC Adv., 2014, 4, 2961.
12. Y. Zheng, J. Li, E. Lee and S. Yang, RSC Adv., 2015, 5, 30495.
13. H. Du and J. Zhang, Soft Matter, 2010, 6, 3370.
14. X. Gu and P. T. Mather, RSC Adv., 2013, 3, 15783.
15. Y. Dong, Q. Q. Ni and Y. Fu, Composites, Part A, 2015, 72, 1.
16. K. Yu, Y. Liu, Y. Liu, H. X. Peng and J. Leng, J. Intell. Mater. Syst. Struct., 2014, 25, 1264.
17. F. Xie, L. Huang, Y. Liu and J. Leng, Polymer, 2014, 55, 5873.
18. C. H. Wu, S. M. Shau, S. C. Liu, S. A. Dai, S. C. Chen, R. H. Lee, C. F. Hsieh and R. J. Jeng, RSC Adv., 2015, 5, 16897.
19. D. R. Paul and L. M. Robeson, Polymer, 2008, 49, 3187.
20. H. Kim, A. A. Abdala and C. W. Macosko, Macromolecules, 2010, 43, 6515.
21. A. J. Waddon and E. B. Coughlin, Chem. Mater., 2003, 15, 4555.
22. G. Li, L. Wang, H. Ni and C. U. Pittman, J. Inorg. Organomet. Polym., 2001, 11, 123.
23. D. Yang, D. Gao, C. Zeng, J. Jiang and M. Xie, React. Funct. Polym., 2011, 71, 1096.
24. Y. Ni, S. Zheng and K. Nie, Polymer, 2004, 45, 5557.
25. E. T. Kopesky, G. H. Mckinley and R. E. Cohen, Polymer, 2006, 47, 299.
26. A. Romo-Uribe, P. T. Mather, T. S. Haddad and J. D. Lichtenhan, J. Polym. Sci., Part B: Polym. Phys., 1998, 36, 1857.
27. R. A. Mantz, P. F. Jones, K. P. Chaffee, J. D. Lichtenhan, J. W. Gilman, I. M. K. Ismail and M. J. Burmeister, Chem. Mater., 1996, 8, 1250.
28. D. Randall, The Polyurethanes Book, John Wiley & Sons Inc, 2003.
29. B. S. Lee, B. C. Chun, Y. C. Chung, K. I. Sul and J. W. Cho, Macromolecules, 2001, 34, 6431.
30. M. Barikani, K. M. Zia, I. A. Bhatti, M. Zuber and H. N. Bhatti, Carbohydr. Polym., 2008, 74, 621.
31. L. Zhang, N. R. Brostowitz, K. A. Cavicchi and R. A. Weiss, Macromol. React. Eng., 2014, 8, 81.
32. 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.
33. Y. C. Chung, J. W. Choi, S. J. Shin and B. C. Chun, Fibers Polym., 2012, 13, 815.
34. M. K. Jang, A. Hartwig and B. K. Kim, J. Mater. Chem., 2009, 19, 1166.
35. M. Bothe, K. Y. Mya, E. M. J. Lin, C. C. Yeo, X. Lu, C. He and T. Pretsch, Soft Matter, 2012, 8, 965.
36. K. Y. Mya, H. B. Gose, T. Pretsch, M. Bothe and C. He, J. Mater. Chem., 2011, 13, 4827.
37. J. Xu and J. Song, PNAS, 2010, 107, 7652.
38. X. L. Lu, W. Cai, Z. Gao and W. J. Tang, Polym. Bull., 2007, 58, 381.
39. A. B. Kutikov, K. A. Reyer and J. Song, Macromol. Chem. Phys., 2014, 215, 2482.
40. K. Du and Z. Gan, J. Mater. Chem. B, 2014, 2, 3340.
41. Y. Li, S. Xin, Y. Bian, Q. Dong, C. Han, K. Xu and L. Dong, RSC Adv., 2015, 5, 24352.
42. X. Zhang, X. Lu, Z. Wang, J. Wang and Z. Sun, J. Biomater. Sci., Polym. Ed., 2013, 24, 1057.
43. M. Xie, L. Wang, J. Ge, B. Guo and P. X. Ma, ACS Appl. Mater. Interfaces, 2015, 7, 6772.
44. Y. Liu, X. Yang, W. Zhang and S. Zheng, Polymer, 2006, 47, 6814.
45. G. Perego, G. D. Cella and C. Bastioli, J. Appl. Polym. Sci., 1996, 59, 37.
46. C. Ni, G. Ni, L. Zhang, J. Mi, B. Yao and C. Zhu, J. Colloid Interface Sci., 2011, 362, 94.
47. B. Rajendran, S. K. S. Kumar, R. S. Rajeev and C. P. R. Nair, Polym. Adv. Technol., 2013, 24, 623.
48. I. S. Gunes and S. C. Jana, J. Nanosci. Nanotechnol., 2008, 8, 1616.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra14742e

---