Methylene diphenyl diisocyanate (MDI) and toluene diisocyanate (TDI) based polyurethanes: thermal, shape-memory and mechanical behavior

Abstract

Shape memory polymers (SMPs) have attracted extensive attention from basic and fundamental research to industrial and practical applications. Among them, shape memory polyurethanes (SMPUs) have different applications such as in textile finishings, adhesives, coatings, automotive parts, furniture, construction materials, thermal insulation materials and footwear industries because they can be synthesized with different types of molecular architectures by manipulating their composition and properly choosing the chemical structure of their individual components. In this work, the synthesis and characterization of SMPUs, based on two-step polymerization, are reported. The hard segment of SMPU was composed of diisocyanate (toluene 2,4-diisocyanate (TDI) or 4,4′-methylene diphenyl diisocyanate (MDI)) and a chain extender, 1,4-butanediol (BD). On the other hand, the soft segment was prepared by a polyol, poly(oxytetramethylene) glycol (PTMG). By selectively choosing the hard-to-soft segment content, the glass transition temperature of SMPUs could be varied from −52.1 °C to 8.6 °C, while the proper combination of both segments imparts combined ductility and strength to our materials. Furthermore, the shape memory effect was found to depend on hydrogen bonding molecular interactions, making TDI-based SMPUs more appropriate for their commercial use.

---

1. Introduction

Shape memory polymers (SMPs) present the ability of modifying their shape in a predefined manner in response to externally imposed stimuli. A deformation-induced temporary shape is transformed to an initial equilibrium configuration that is defined by the chemical or physical crosslinking within the polymer.¹ SMPs are cheap, easily processed, light (ρ ∼ 1–1.3 g cm⁻³), could be deformed to high strains and high recovery ratios can be achieved in contrast to shape memory alloys.¹ To date, light,² thermal,^3,4 electrical,⁵ magnetic⁶ and solvent^7,8 stimuli have been mainly used for triggering the shape-memory effect in polymeric materials. Due to the inherent advantages over other kind of stimuli, temperature-responsive SMPs are the most explored materials where either the glass transition temperature (T_g) or melting temperature (T_m) are considered as the transition temperature of the shape memory effect (T_trans).⁹

Usually, shape memory processes consist of two different phases known as “programming” and “recovery”.¹⁰ During programming, the material is deformed above the transition temperature where the conformational entropy of the material is decreased.¹¹ Subsequently, the material is cooled to temperatures below the segmental transition under constrained conditions to reach the “temporary shape” owing to the reduced molecular mobility at temperatures below T_trans.¹² Finally, the material recovers its initial permanent “fixed shape” upon the application of the external stimulus, which triggers the shape memory effect. Depending on their microstructure and chemical nature, several classes of SMPs could be found. Among all the available types of materials showing a shape-memory effect, shape memory polyurethanes (SMPUs) have shown suitable physical–mechanical properties to be used in applications as stents,¹³ micro-actuators¹⁴ and wrinkle free fabrics,¹⁵ among others.

Since the shape-memory behavior is related to the switching ability of different domains formed physically or chemically crosslinked structures, this effect could be tailored by selectively modifying the microphase morphology of the material.¹⁶ In polyurethanes, the shape-memory properties depend on the phase-morphology resulting from the molecular structure of alternating hard and soft segments. Typically, those soft segments present T_g values below the service temperature; while hard segments, mainly obtained by the reaction of diisocyanates with a diol or a diamine chain extenders, have T_g values above ambient temperature due to the presence of strongly hydrogen bonded moieties.¹⁷ Whereas soft segments could be considered as elastomeric reversible phases, hard segments behave as rigid fixed units. It has been shown that the rigid units are able to memorize the original shape while the soft phase forms the temporary shape. In this framework, through proper design, it would be possible to synthesize polyurethanes with desired mechanical flexibility/strength and transition temperatures.

Two main routes exist for the development of segmented polyurethanes. The first one is known as a “one-shot method” and involves mixing together the required amounts of diisocyanate, soft segment oligomer and the chain extender.¹⁸ On the contrary, in the “prepolymer method”, an excess of diisocyanate reacts with the soft segment to yield a prepolymer, which is further reacted with the chain extender to form the high molecular weight PU. This second method is preferred because it allows one to obtain well-defined hard segments with probable distributions.¹⁹ It has been reported that the electron withdrawing ability of phenylene rings makes aromatic diisocyanates more reactive than aliphatic diisocyanates, enabling the synthesis of polyurethanes without catalysts.¹⁷ Typically, solvents such as dimethylacetamide (DMAC) or dimethylformamide (DMF) are used for the fabrication of polyurethanes, leading to side-reactions when 4,4-diphenylmethane diisocyanate (MDI) is used as the diisocyanate. In addition, their use represents a risk to human health and the environment. In view to these facts, in this work, SMPUs have been obtained via the “prepolymer method” in a solvent-free process, where aromatic toluene 2,4-diisocyanate (TDI) and 4,4-diphenylmethane diisocyanate (MDI) have been selected as diisocyanates, avoiding the environmental issues associated with the release of volatile organic compounds (VOCs).

Though extensive work has been devoted in developing SMPUs, little attention has been paid in understanding how the chain microstructure affects the shape-memory behavior and mechanical properties of the resulting materials. In this scenario, this work deals with the synthesis and physico-mechanical characterization of MDI and TDI-based SMPUs. Transition temperatures have been determined by differential scanning calorimetry (DSC) and dynamic mechanical analysis (DMA). Thermomechanical programming experiments were carried out to examine the shape-memory effect of the developed materials. Finally, the mechanical properties (elastic modulus, elongation at break and fracture toughness) have been further determined by tensile testing. Overall, the results reveal a marked influence of the soft–hard segments over the transition temperature, shape-memory effect and mechanical behavior of MDI and TDI-based polyurethanes.

2. Experimental

2.1 Materials

All the raw materials employed in the polyurethane synthesis were purchased from Sigma-Aldrich: poly(oxytetramethylene) glycol (PTMG, M_n = 1000 g mol⁻¹), 4,4′-methylene diphenyl diisocyanate (MDI) and toluene 2,4-diisocyanate (TDI) were used as received; whereas 1,4-butanediol (BD), used as a chain extender, was dried under vacuum for 3 h at 65 °C before use.

2.2 Synthesis of shape memory polyurethanes (SMPUs)

All of the SMPUs were synthesized by a two-step method (prepolymer method) varying the relationship between the hard-segment/soft-segment content between 30 and 70%. Briefly, SMPUs were prepared by the reaction of stoichiometric amounts of polyol/diisocyanate/chain extender with block ratios of 1 : n + 1 : n (where n, mole ratio of reactants, is between 0.5 and 5.5). Therefore, following the synthetic route outlined in Scheme 1, the hard segments of the SMPUs would be composed of 4,4′-methylenediphenyl diisocyanate (MDI) or toluene 2,4-diisocyanate (TDI) and the chain extender, 1,4-butanediol (BD); whilst the soft segment would be composed of poly(oxytetramethylene) glycol (PTMG).

Scheme 1 Synthetic route for 4,4′-methylene diphenyl diisocyanate (MDI) and toluene 2,4-diisocyanate (TDI) based polyurethanes.

The synthesis was carried out at 70 °C in a 150 mL 5-neck round-bottom flask equipped with a mechanical stirrer and a nitrogen inlet. In the first step, the polyol (PTMG) was added into the dried reactor, and after 30 min under nitrogen atmosphere, MDI or TDI was added dropwise. The reaction continued at 70 °C for 2 h to obtain a –NCO terminated prepolymer. A vigorous nitrogen flow has been used to prevent the reaction of the isocyanate groups with air moisture. In the second step, the BD chain extender was added dropwise into the reaction system. The reaction mixture was continuously stirred for 2 minutes until a significant increase in viscosity was detected. Then, the viscous mixture was poured into a preheated stainless steel mold at 100 °C, and put into a hydraulic press overnight. Thus, SMPUs sheets were manufactured by compression molding under a pressure of 150 bar. Two Teflon sheets were placed on both sides of the mold to reduce the surface roughness of the SMPUs obtained. After curing, the obtained specimens were cooled to room temperature in the mold under constant pressure. FTIR was used to confirm that the isocyanate groups were completely reacted.^20,21

Table 1 summarizes all the SMPUs synthesized both with MDI and with TDI. It should be noted that neither the MDI-based polyurethanes with a molar ratio above 4 nor the TDI-based polyurethanes with a molar ratio below 2.5 could be characterized properly due to its mechanical properties. This impediment is because of the procedure of synthesis. In the case of MDI-based polyurethanes, they were too rigid and solid. In the case of TDI-based polyurethanes, they were too liquid to be put into a press under pressure.

Table 1 Summary of the MDI and TDI-based SMPUs synthesized (n represents the molar ratio)

Polymer code	Composition of polyurethane PTMG/diisocyanate/BD (1 : n + 1 : n)
	PTMG	MDI	TDI	BD
MDI-0.5	1	1.5	0	0.5
MDI-1	1	2	0	1
MDI-2	1	3	0	2
MDI-3	1	4	0	3
MDI-4	1	5	0	4
TDI-3	1	0	4	3
TDI-3.5	1	0	4.5	3.5
TDI-4.5	1	0	5.5	4.5
TDI-5	1	0	6	5

---

2.3 Attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR)

ATR-FTIR was used to assess the extent of reaction between the isocyanate and hydroxyl groups. Infrared spectra were collected on a Nicolet Nexus FTIR spectrophotometer (Thermo Electron Corporation). The spectra were obtained using an ATR tool in the range 500 to 4000 cm⁻¹ at a resolution of 4 cm⁻¹ and 64 scans per spectrum.

2.4 Differential scanning calorimetry (DSC)

Thermal properties of all samples were measured by differential scanning calorimetry (DSC 822e from Mettler Toledo) to identify the thermal transitions of the obtained materials. The transition temperature (T_trans) was defined from the glass transition temperature measured in the second heating cycle (T_g in Table 2). Samples containing 10–15 mg were sealed in aluminum pans and were characterized under constant nitrogen flow (20 mL min⁻¹). First, samples were equilibrated at −100 °C, and then heated at a rate of 10 °C min⁻¹ from −100 to 250 °C. In this first cycle, the thermal history of the sample was erased. It was then cooled down to −100 °C at a cooling rate of 10 °C min⁻¹. Subsequently, a second heating scan to 250 °C was conducted at the same heating rate.

Table 2 Thermal properties for MDI and TDI-based polyurethanes measured by DSC and DMA

Sample	Tg,DSC (°C)	ΔHm (J g^−1)	Tg,DMA (°C)
MDI-0.5	−52.1	—	−30.9
MDI-1	−51.3	—	−30.4
MDI-2	−49.7	106.9	−18.4
MDI-3	*	117.1	−12.4
MDI-4	*	347.7	−3.9
TDI-3	−19.4	—	22.1
TDI-3.5	−15.7	—	25.3
TDI-4.5	−7.8	—	37.4
TDI-5	1.5	—	41.9

---

2.5 Dynamic mechanical analysis (DMA)

DMA was performed on a Mettler-Toledo DMA1 analyzer in the tensile mode. 1.5 mm thick, 6 mm wide and 10 mm long specimens were used. Curves displaying storage modulus (E′) and the loss factor (tan δ) were recorded in the range of −100 to 150 °C at a heating rate of 3 °C min⁻¹. Deformation frequencies of 1, 3 and 10 Hz (only data for 10 Hz frequency are shown) and displacement of 20 μm were used, which were found within the Linear Viscoelastic Region (LVR) of the synthesized SMPUs.

Thermo-mechanical analysis (TMA) was also conducted on the Mettler Toledo DMA1 at a heating rate of 4 °C min⁻¹ in the temperature range of −100 to 30 °C or −20 to 80 °C for MDI and TDI-based SMPUs, respectively. Rectangular samples of about 10 mm × 6 mm × 1.5 mm were used in shape memory performances. First, the sample was heated to a desired programming temperature T_prog at least 20 °C above the transition temperature (T_trans) defined from the glass transition temperature measured by DSC (30 °C or 80 °C), and deformed by applying a force. This force (5 N for MDI-based polyurethanes and 2 N for TDI-based polyurethanes) was applied in 0.1 minute and held for 5 min at the deformation temperature. This procedure was identical for all polyurethane samples thereby avoiding the effect of the loading rate on the recovery performance reported by Xiao et al.²² Once the sample has been stretched, ε_m, the next stage is to cool it below the transition temperature T_low (−100 °C or −20 °C) in order to fix the temporary shape. After unloading (F = 0), the deformation of the sample is ε_u. The shape-memory effect is triggered by heating the sample above its transition temperature. The heating rate during shape recovery was 4 °C min⁻¹. The amount of non-recoverable deformation at the end of programming is ε_p. The fixing (R_f) and recovery (R_r) ratios were calculated for each sample using eqn (1) and (2).²³

2.6 Tensile test

25 mm long, 1 mm thick and 4 mm wide specimens were tested on a Shimadzu Autograph using a 500 N load cell at a stretching speed of 2.5 mm min⁻¹. Specimens were prepared according to ISO 527-1:2012 and they were tested at 23 °C. Young's modulus (E′), stress and strain at yield (σ_y, ε_y) and stress and strain at break (σ_b, ε_b) were determined. Reported values represent mean average values and standard deviations over 5 specimens.

2.7 Morphological characterization

SMPUs have been analyzed in a Hitachi S-4800 field emission scanning electron microscope (FE-SEM) at an acceleration voltage of 15 kV. Analyzed surfaces were copper-coated in a Quorum Q150T ES turbo-pumped sputter coater (5 nm thick coating).

3. Results and discussion

3.1 ATR-FTIR analysis

Polyurethanes based on MDI and TDI were examined by ATR-FTIR spectroscopy (Fig. 1). ATR-FTIR was used to follow and control the degree of the reaction.^24–26 The stretching band corresponding to the isocyanate group was not found at 2270 cm⁻¹, indicating that the initial isocyanate groups completely reacted during the synthesis.²⁷ Moreover, in Fig. 1, it can be observed, for the isocyanate segments, the symmetric and asymmetric stretching vibrations of N–H, corresponding to the broad absorption bands near 3320 cm⁻¹; the medium-strong peak at 1590 cm⁻¹, which confirms the in-plane bending vibration of N–H; the sharp absorption peaks around 1730–1700 cm⁻¹, which are typical for the stretching vibration of esters C=O; the 1220 cm⁻¹ absorption peak caused by the vibration of C=C in benzene ring; and several weak peaks near 900–700 cm⁻¹ that belong to out-of-plane bending vibration of C–H in multisubstituted benzene ring. On the other hand, Fig. 1 shows, for the soft segment polyether polyols, the peak groups near 2940 to 2850 cm⁻¹ that are caused by the stretching vibration of C–H₂, and the broad and strong peak at 1100 cm⁻¹ from the ether bond C–O–C stretch. Furthermore, the peaks observed at around 1530 and 1450 cm⁻¹, corresponding to the C–N and CH₂ bonds, respectively, support the formation of the urethane group.^24,28

Fig. 1 ATR-FTIR spectra for the MDI and TDI-based polyurethanes synthesized.

3.2 Differential scanning calorimetry (DSC)

It is well known that SMPUs have an amorphous reversible phase, due to their microstructure, composed by hard and soft segments. In this study, the glass transition temperatures (T_g,DSC) of the SMPUs were obtained with DSC. Fig. 2 shows the second heating DSC scans of the SMPUs synthesized whereas Table 2 summarizes the measured glass transition temperatures.

Fig. 2 Second heating DSC thermograms for MDI (a) and TDI-based SMPUs (b).

DSC results show that the glass transition temperature of polyurethanes increases with the hard segment content (higher n). The glass transition temperatures of MDI-based polyurethanes are between −52.1 °C and −49.6 °C, whilst for TDI-based polyurethanes are between −19.4 and 1.5 °C. This suggests that the high hard segments can achieve a more well-oriented position within the polymeric structure.²⁷ Additionally, it was found that, for a given n the T_g values, TDI-based polyurethanes were lower than those of the MDI-based polyurethanes. Moreover, in the second heating scan, a melting peak can be observed for some MDI-based SMPUs (n = 2, 3 and 4), which is attributed to the crystallization of the soft domains. As a result, MDI-2, MDI-3 and MDI-4 polyurethanes present melting temperatures (T_m) between 160 and 210 °C. In addition, the melting enthalpy (ΔH_m) measured by DSC became higher, from 106.9 to 347.7 J g⁻¹, as the hard segment content increases (n = 2, 3 and 4).^29–31

3.3 Mechanical and morphological features

The viscoelastic behavior of the synthesized polyurethanes was investigated using DMA. Fig. 3 depicts the temperature dependence of tan δ as a function of n, both for MDI-based and TDI-based SMPUs. Results reveal a well-defined single α-relaxation associated with the glass transition (T_g,DMA values determined as the maximum in tan δ are shown in Table 2) for all the obtained polyurethanes, which gradually increases with n. This increase in T_g, from −30.9 to −3.9 °C for MDI-based polyurethanes (n from 0.5 to 4) and from 22.1 to 41.9 °C for TDI-based polyurethanes (n from 3 to 5), may be ascribed to the reduction in chain-segment mobility induced by the increased presence of hard segments. Similar T_g increase have been found in semi-crystalline thermoplastics via the development of more rigid crystalline domains.³²

Fig. 3 Evolution of the loss factor (tan δ) with temperature for MDI (a) and TDI-based SMPUs (b).

From the experimental point of view, several differences exist between DSC and DMA. In standard DSC, the T_g is defined in relation to the measured heat flow, preferably in cooling experiments. Thus, DSC is sensitive to the C_p changes associated with the glass transition, while DMA is sensitive to mechanical relaxation, also associated with the glass transition, but the expression of which depends on the mechanical frequency imposed by the test.³³ The glass transition temperature could be determined by both techniques, but it has been demonstrated in macromolecular systems that methods such as DSC are less sensitive to the glass transition phenomenon than the DMA employed in this study³⁴ because DMA is a method with great sensitivity in detecting changes in internal molecular mobility. Furthermore, the predominant heat transfer mechanism in DSC is conduction, while in DMA, it is convection.³⁴ Therefore, it should be noted that dynamic experiments (DMA) exhibit higher glass transition temperatures than static experiments (DSC), i.e. T_g,DMA > T_g,DSC.

Therefore, to compare the values of glass transition temperature, the values measured by DMA (T_g,DMA) have been taken into account. From DMA results, it could be observed that MDI-based SMPUs present lower T_g values (−12.4 °C for MDI-3 vs. 22.1 °C for TDI-3). Despite this, MDI-based polyurethanes present higher flow temperatures, especially high for SMPUs synthesized with n up to 2 (MDI-2, MDI-3, MDI-4), which is attributed to the stronger hydrogen bonding interactions and partially crystallized hard segments.²¹ In fact, while the flow zone is located at temperatures above 60–70 °C for all MDI-based polyurethanes (90 °C above their T_g), glass transition is closely followed by the flow region in TDI-based SMPUs (almost overlapping for n = 4.5).

As denoted by the smaller loss factor peak intensity and area, the energy dissipation of synthesized polyurethanes through the studied temperature range decreases upon the addition of hard segments. This decrease in damping is due to the restriction of the hard diisocyanates in the viscous component of the soft poly(tetrahydrofuran) segment.³⁵ In fact, it has been reported that the constraints introduced by the hydrogen bonding between the hard segments provide an increased rigidity to the whole system.^36–38 It should be pointed out that the range of tan δ values obtained here through the introduction of different amounts of MDI and TDI is notably larger than those previously reported by Tan et al.³¹ for TPU/TPS blends using a polyolefin elastomer as compatibilizer. This would be useful for developing polyurethanes with rather different damping behavior simply by tuning the soft/hard segment ratio.

Stress–strain curves of the synthesized SMPUs are shown in Fig. 4 while the statistic values of elastic modulus (E), secant modulus at 2% (E*), elongation at break (ε_b) and fracture toughness (U_T) are summarized in Table 3. While all the polyurethanes present ductile or elastomeric behavior, the results reveal a similar effect of n on the mechanical response of the synthesized materials. The ductility of the specimens decreases as the hard phase increases, yielding a 49 and 29-fold increase in the elastic modulus for TDI and MDI-based PUs respectively. More precisely, as shown in Fig. 4, the elongation at break is reduced from 291% to 182% and from 945% to 408% for MDI and TDI-based polyurethanes, respectively, upon the increase of n. This behavior is usually found in polyurethanes and it is ascribed to the formation of a densely crosslinked structure when the presence of hard segment increases. The increased ε_b of TDI-based SMPUs in regard to the MDI-based one for a given n is associated with a reduced molecular mobility of chains in the last one.^35,39,40 This is in concordance with DMA results where MDI-based polyurethanes exhibit fairly lower glass transition temperatures when compared with TDI-based polyurethanes.

Fig. 4 Stress–strain curves for MDI (a) and TDI-based SMPUs (b). Note that stress–strain curves are partially enlarged in the inset.

Table 3 Main mechanical parameters for the MDI and TDI-based SMPUs synthesized

n	Elastic modulus E (MPa)	Secant modulus E* (MPa)	Elongation at break εb (%)	Fracture toughness UT (MJ m^−3)
MDI
0.5	8.2 ± 0.1	8.1	291	4.0
1	22.8 ± 0.1	22.6	404	16.5
2	37.5 ± 0.7	37.2	463	91.2
3	124.4 ± 0.8	121.8	212	36.3
4	239.0 ± 1.9	232.3	182	42.3
TDI
3	1.6 ± 0.1	1.6	945	2.3
3.5	1.9 ± 0.1	2.0	1167	11.1
4.5	2.5 ± 0.1	2.5	622	20.9
5	78.6 ± 0.7	75.3	408	77.9

---

It is interesting to note that intermediate n values result in a concurrent increase of strength and elongation at break, yielding fracture toughness values up to 91.2 (n = 2) and 77.9 (n = 5) MJ m⁻³ for MDI-based and TDI-based SMPUs, respectively. This 23-fold and 34-fold increase in fracture toughness arises from the synergetic effect obtained when both rigid and soft segments are found at an appropriate ratio. Indeed, the hard phase is responsible for holding together the rubbery soft phase by intermolecular hydrogen bonding and crosslinked networks, whereas the soft segments are able to extend with no rupture upon stretching.⁴¹ The proper combination of both segments imparts combined ductility and strength. When the soft phase is very large (low n) the material cannot withstand applied stress and is easily deformed. On the contrary, at high n values, hard segments prevent macromolecules moving too far out of position, yielding lower ε_b values but increased elastic modulus.

As shown in Fig. 5, the morphology of MDI-based and TDI-based polyurethanes has been studied by FE-SEM. No phase separation is observed in all of the synthesized SMPUs indicating that homogeneous materials were obtained during the synthesis. Although all the samples were cryogenically fractured, according to their morphological features, they show different fracture modes. It could be seen that the surface of specimens containing large fractions of hard segment present a rather smooth surface. More interestingly, at medium n values, where higher fracture toughness has been achieved (Table 3), surfaces present more shear zones and fibrils, indicating that more energy has been dissipated by the material during the plastic deformation. Korley et al.⁴² mentioned that as the hard segment composition increases, mechanical data are consistent with a shift in the continuous domain morphology, producing materials with inter-connected hard domains, exhibit limited extensibility, but increased initial modulus, as shown in Table 3.

Fig. 5 Representative FE-SEM images showing the morphology of cryogenically fractured SMPU surfaces: MDI (a.1 at 500 μm scale and a.2 at 50 μm scale) and TDI-based SMPUs (b.1 at 500 μm scale and b.2 at 50 μm scale).

3.4 Shape memory behavior

Thermally-induced shape memory behavior of SMPU samples was quantitatively evaluated by TMA monitoring the shape recovery process. Rectangular samples, with a cross-section area of 6 mm × 1.5 mm and initial clamps distance of 10 mm, were deformed above its transition temperature of shape memory effect (30 °C for MDI-based SMPUs or 80 °C for TDI-based SMPUs).

Regarding thermally activated shape-memory properties, soft segments will be responsible for shape fixity, acting as the switching segments; while hard segments will be responsible for shape recovery, determining the permanent shape.⁴³ First, the sample is heated without force at a temperature above the T_trans in order to allow relaxation of the polymer chain. Then, force is applied and the sample is deformed. The stress is then maintained for 5 min. It can be seen that, applying the same stress to all samples, deformation depends on the hard segment content. Indeed, at low hard-segment content, the deformation is notably increases, which agrees well with the reported tensile data. The temporary shape was fixed cooling down its transition temperature of shape memory effect (−100 °C for MDI-based SMPUs or −20 °C for TDI-based SMPUs) and, finally, the samples were heated-up above the transition temperature, so the thermally-induced recovery process was observed.^44,45

The thermally activated shape memory behavior of MDI-based SMPUs and TDI-based SMPUs is shown in Fig. 6. Furthermore, Table 4 summarizes the values obtained of fixity ratio (R_f) and shape recovery ratio (R_r) which were calculated employing eqn (1) and (2). It can be seen that the R_f values obtained for the TDI-based SMPUs are larger than those previously showed by Gu et al.⁴⁶ for polyethylene glycol (PEG)-based SMPUs, which obtained a maximum of 75%. Moreover, the hard segment domain formed by MDI and BD provides stable physical net-points to achieve good shape recovery, approximately near 100%. For MDI-based SMPUs, shape fixity and shape recovery ratios decreased with the increase in hard segment content. The shape fixity values for MDI-based SMPUs with molar ratio more than 2, are less than 0%. These negative values are directly related to the TMA test procedure developed in order to measure the shape memory behavior. It should be noted that the maximum displacement that the Mettler Toledo DMA1 can register is 1 mm (1000 μm). Thus, on the one hand, when forces higher than 5 N were applied this limit can be overcome and, on the other hand, the applied force (5 N) is not enough to fix the temporary shape due to a thermal process contraction, which decreases values used in eqn (2) to calculate R_f.

Fig. 6 Thermomechanical response of MDI (a) and TDI-based SMPUs (b).

Table 4 Shape memory parameters for the MDI- and TDI-based SMPUs calculated from eqn (1) and (2)

n	Rf (%)	Rr (%)
MDI
0.5	94.8	89.9
1	60.6	73.1
2	59.4	78.6
3	*	76.5
4	*	41.1
TDI
3	89.8	100.1
3.5	93.4	99.8
4.5	89.6	99.9
5	84.9	99.8

---

It can be concluded that the shape memory properties for the TDI-based SMPUs (R_r (%)) are superior to MDI-based SMPUs. One possible reason is that their physical net-points formed by molecular interactions are weak in the MDI-based SMPUs, whereas TDI-based SMPUs show higher hydrogen bonded molecular interactions as mentioned above.

4. Conclusions

In this study, we have synthesized shape-memory polyurethanes via the “prepolymer method” in a solvent-free process using 4,4-diphenylmethane diisocyanate (MDI) and toluene 2,4-diisocyanate (TDI) as hard segments, poly(tetrahydrofuran) as the soft segment and 1,4-butanediol as the chain extender. Polyurethanes with tunable thermal and mechanical properties have been obtained by varying the hard-to-soft segment content between 30% and 70% by weight. ATR-FTIR and DSC measurements show that reaction procedure was appropriate and the glass transition temperature of polyurethanes increases with hard segment content.

Regarding the mechanical properties, both types of polyurethanes show huge versatility, ranging from highly elastic soft materials with an elastic modulus of 1.6 MPa and elongation at break up to 1167% to tough–hard materials. Furthermore, depending on their microstructure their shape-memory effect varies between MDI-based polyurethanes and TDI-based polyurethanes. TDI-shape memory properties were found to be superior to MDI-SMPUs.

Overall, the obtained experimental findings through this work highlight the potential of both MDI and TDI-based SMPUs for applications in which vibration isolation is needed over a wide temperature range. These may include the manufacturing of soles for footwear, isolators for large industrial equipment, and isolation systems for vibration-sensitive instruments such as scanning electron microscopes, among others.

Acknowledgements

Authors thank Government of La Rioja and the Footwear Technology Center of La Rioja for the financial support.

References

1. A. Lendlein and S. Kelch, Angew. Chem., Int. Ed., 2002, 41, 2034–2057.
2. A. Lendlein, H. Jiang, O. Jünger and R. Langer, Nature, 2005, 434, 879–882.
3. W. Zhang, L. Chen and Y. Zhang, Polymer, 2009, 50, 1311–1315.
4. C. C. Wang, W. M. Huang, Z. Ding, Y. Zhao and H. Purnawali, Compos. Sci. Technol., 2012, 72, 1178–1182.
5. Q. Meng and J. Hu, Composites, Part A, 2009, 40, 1661–1672.
6. X. Zheng, S. Zhou, Y. Xiao, X. Yu, X. Li and P. Wu, Colloids Surf., B, 2009, 71, 67–72.
7. R. Xiao, RSC Adv., 2016, 6, 6378–6383.
8. R. Xiao, J. Guo, D. L. Safranski and T. D. Nguyen, Soft Matter, 2015, 11, 3977–3985.
9. I. Bellin, S. Kelch and A. Lendlein, J. Mater. Chem., 2007, 17, 2885.
10. T. Xie, Nature, 2010, 464, 267–270.
11. L. R. Treloar, The Physics of Rubber Elasticity, 1975.
12. C. Azra, Y. Ding, C. J. G. Plummer and J.-A. E. Månson, Eur. Polym. J., 2013, 49, 184–193.
13. M. C. Serrano and G. A. Ameer, Macromol. Biosci., 2012, 12, 1156–1171.
14. T. S. Wilson, J. P. Bearinger, J. L. Herberg, J. E. Marion, W. J. Wright, C. L. Evans and D. J. Maitland, J. Appl. Polym. Sci., 2007, 106, 540–551.
15. J. Hu, H. Meng, G. Li, S. I. Ibekwe, H. Jinlian, M. Harper, L. Guoqiang and I. I. Samuel, Smart Mater. Struct., 2012, 21, 53001.
16. W. M. Huang, Y. Zhao, C. C. Wang, Z. Ding, H. Purnawali, C. Tang and J. L. Zhang, J. Polym. Res., 2012, 19, 9952.
17. I. Yilgör, E. Yilgör and G. L. Wilkes, Polymer, 2015, 58, A1–A36.
18. J. Bernardini, P. Cinelli, I. Anguillesi, M.-B. Coltelli and A. Lazzeri, Eur. Polym. J., 2015, 64, 147–156.
19. R. Hernandez, J. Weksler, A. Padsalgikar, T. Choi, E. Angelo, J. S. Lin, L.-C. Xu, C. A. Siedlecki and J. Runt, Macromolecules, 2008, 41, 9767–9776.
20. M. Charlon, B. Heinrich, Y. Matter, E. Couzigné, B. Donnio and L. Avérous, Eur. Polym. J., 2014, 61, 197–205.
21. Z. Chang, M. Zhang, A. G. Hudson, E. B. Orler, R. B. Moore, G. L. Wilkes and S. R. Turner, Polymer, 2013, 54, 6910–6917.
22. R. Xiao, C. M. Yakacki, J. Guo, C. P. Frick and T. D. Nguyen, J. Polym. Sci., Part B: Polym. Phys., 2016, 54, 1405–1414.
23. E. Axpe, N. García-Huete, J. M. Cuevas, C. Ribeiro, D. Mérida, J. M. Laza, J. Á. García, J. L. Vilas, S. Lanceros-Méndez, F. Plazaola and L. M. León, J. Polym. Sci., Part B: Polym. Phys., 2015, 53, 1080–1088.
24. M.-C. Kuo, S.-M. Shau, J.-M. Su, R.-J. Jeng, T.-Y. Juang and S. A. Dai, Macromolecules, 2012, 45, 5358–5370.
25. J. T. Kim, B. K. Kim, E. Y. Kim, S. H. Kwon and H. M. Jeong, Eur. Polym. J., 2013, 49, 3889–3896.
26. L. Jiao, H. Xiao, Q. Wang and J. Sun, Polym. Degrad. Stab., 2013, 98, 2687–2696.
27. 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, 16897–16910.
28. Z. Wang, L. Yu, M. Ding, H. Tan, J. Li and Q. Fu, Polym. Chem., 2011, 2, 601–607.
29. W. Panwiriyarat, V. Tanrattanakul, J.-F. Pilard, F. Burel and N. Kébir, J. Appl. Polym. Sci., 2016, 133, 42943–42950.
30. H. Lu, C. Lu, W. Huang and J. Leng, Smart Mater. Struct., 2015, 24, 035018.
31. L. Tan, Q. Su, S. Zhang and H. Huang, RSC Adv., 2015, 5, 80884–80892.
32. E. Lizundia, S. Petisco and J.-R. Sarasua, J. Mech. Behav. Biomed. Mater., 2013, 17, 242–251.
33. C. A. Gracia-Fernández, S. Gómez-Barreiro, J. López-Beceiro, J. Tarrío Saavedra, S. Naya and R. Artiaga, Polym. Test., 2010, 29, 1002–1006.
34. S. Rimdusit and H. Ishida, J. Polym. Sci., Part B: Polym. Phys., 2000, 38, 1687–1698.
35. L. Shao, J. Dai, Z. Zhang, J. Yang, N. Zhang, T. Huang and Y. Wang, RSC Adv., 2015, 5, 101455–101465.
36. P. Krol, Prog. Mater. Sci., 2007, 52, 915–1015.
37. Z. S. Petrović and J. Ferguson, Prog. Polym. Sci., 1991, 16, 695–836.
38. D. K. Chattopadhyay and D. C. Webster, Prog. Polym. Sci., 2009, 34, 1068–1133.
39. Y.-S. He, J.-B. Zeng, G.-C. Liu, Q.-T. Li and Y.-Z. Wang, RSC Adv., 2014, 4, 12857.
40. Y. Zhou, H. Xiu, J. Dai, H. Bai, Q. Zhang and Q. Fu, RSC Adv., 2015, 5, 30912–30919.
41. D. Ionita, C. Gaina, M. Cristea and D. Banabic, RSC Adv., 2015, 5, 76852–76861.
42. L. T. J. Korley, B. D. Pate, E. L. Thomas and P. T. Hammond, Polymer., 2006, 47, 3073–3082.
43. A. Saralegi, M. L. Gonzalez, A. Valea, A. Eceiza and M. A. Corcuera, Compos. Sci. Technol., 2014, 92, 27–33.
44. H. M. Jeong, B. K. Ahn and B. K. Kim, Eur. Polym. J., 2001, 37, 2245–2252.
45. Y. Kim, H. Park and B. Kim, High Perform. Polym., 2015, 27, 886–897.
46. X. Gu and P. T. Mather, Polymer, 2012, 53, 5924–5934.

---