Introduction

A polymer melt-blending process is in general an environmentally friendly process due to the limited usage of solvent. The process was often employed to prepare novel materials in terms of the aforementioned advantage and cost competitive merit. In addition, bio-based polymers have received much attention due to an increasing environmental concern. Particularly, polylactic acid (PLA), originated from the synthesis of lactide via the fermentation of the sugar feedstock, such as corn, etc., has been considered as a commercial interest because of its biocompatibility, non-toxicity, and biodegradability [1]. In light of these fascinating characteristics, PLA could serve in some biomedical and environmental applications, such as bone screw, plastic articles, etc. An attempt to expand the application of PLA by preparing the derived blends is still increasing, and hopefully this effort could reduce the environmental impact.

Recently, among those expanding fields of application, shape memory polymer (SMP), one type of self-healing polymer, shows a fast response in changing their temporary shapes to original shapes when subjected to certain stimulus, such as temperature, light, electric field, pH, etc. Thermal stimulus SMP types, by far, are the most commonly investigated systems. In the early development of SMP, the sample, exhibiting block copolymer type morphology with a hard domain representing the reversible phase and a soft block representing the switch phase, was able to fix at a temporary shape at room temperature following an immediate cooling from the predeformed state above the switching temperature (Ts). And the further heating to above the switching temperature, SMP could regain their original shape. The switching temperature could be either the glass transition temperature or crystalline melting temperature. Several recent reviews discussed the recent progress and potential applications of SMPs [2–5] in several systems [6–9]. However, in considering the versatility and cost of SMP, a polymer blending technology being a facile and exciting approach to the novel development of SMP has just received much attention recently [10–15]. Mather group [10–12] investigated a series of shape memory behavior of amorphous (switching phase)/crystalline polymer (reversible phase) blends. The amorphous polymer included ply(vinyl acetate), poly(methyl acrylate), poly(ethyl acrylate), poly(alkyl methacrylate) and the crystalline polymer phase included poly(vinylidene fluoride), polylactide, poly(hydroxybutyrate), poly(ethylene–glycol), polyethylene, poly(ethylene-co-vinyl acetate), copolymers of poly(vinylidene chloride) and poly(vinyl chloride).

To further endow SMP with different characteristic, a bio-based SMP polymer exhibiting environmental benefit deserved much focus in terms of sustainable development. Only limited PLA-based SMP blends were prepared recently [16, 17]. Zhang et al. [16] employed a novel bio-based nylon elastomer, to tough PLA, and accidently found the shape memory effect for this toughening blend, but no detail investigation on the shape memory behavior was reported. This work aims to unveil this important observation by preparing PLA/thermoplastic polyurethane (TPU) blends and focuses on the shape memory behaviors under three different predeformation temperatures (25, 80, 120 °C). The predeformation temperatures were chosen below or above the glass transition temperature of PLA near 80 °C, so we can observe the effect of predeformation temperatures to the shape recovery processes at various recovery temperatures. In addition, some material like conventional rubber, it showed very good elastic recovery, but failed to keep a fixed deformation. Thus, it was considered as an elastic material, instead of shape memory polymer. In order to consider both fixing and recovery ratios together, an additional denotation of shape memory index, Rf*Rr, to couple both ratios by the multiplication of fixing ratio and recovery ratio was considered. To further have an in-depth understanding on the shape recovery mechanism, the recovery stress was recorded against temperature while the predeformed samples were heated under full-strain constraint, instead of conventional free-strain recovery to evaluate the SMP performance of bio-based blends.

Experimental

Materials

The materials used in this study were poly(lactic acid) (PLA) and thermoplastic polyurethane (TPU). PLA (4032D) was supplied from Nature Works LLC. The density is 1.24 g/cm3. TPU (Desmopan, KU2-8785) was supplied from Bayer Taiwan Co., Ltd.

Sample preparations

All pristine resins were predried at 80 °C for 4 h in a vacuum oven prior to mixing. The mixing of PLA and TPU at a composition ratio of 70/30 or 50/50 was conducted in the Brabender internal mixer (815605, Plastograph) under 50 rpm at 170 °C for 10 min. The prepared PLA/TPU batch was hot-pressed to form a thin sheet at 180 °C for about 10 min.

Measurements

Thermal characterizations

The thermal behaviors of samples were characterized by a differential scanning calorimeter (DSC, TA Instrument, Q10). First, all test samples were heated to 250 °C at a heating rate of 20 °C/min to eliminate the thermal history, and then were cooled to 0 °C at a cooling rate of 10 °C/min, followed by the second heating to 250 °C with the same heating rate of 20 °C/min to record the glass transition temperature and melting temperature. The crystallinity of PLA was calculated by taking the heat of fusion divided by the actual PLA content and the enthalpy required for 100 % crystallinity, equal to 93.7 J/g [18].

Morphological characterizations

The transmission electron microscope (TEM, Hitachi, H-7100) was employed to observe the TPU dispersion within the PLA matrix at acceleration voltage of 75 kV The samples were cryo-microtomed to ultrathin films for better observation.

Mechanical properties

Shape memory test was conducted using a Universal Tensile Tester (Instron 4469) according to the following steps, (1) Recording the length before extending, (2) Heating the sample to temperatures (80 °C, 120 °C) and holding for 5 min.(no heating for 25 °C), (3) Extending the sample to the strain of 100 % (εm), (4) Cooling the sample to room temperature (25 °C) and holding for 5 min, (5) Unloading the stress to zero and recording the strain (εu). (6) Heating the sample at various recovery temperatures from 25 °C to 160 °C again and holding for 5 min, (7) Recording the sample strain (εp). The shape fixing ratio (Rf) and shape recovery ratio (Rr) are determined as follows [3, 8],

$$ {R_f}=\frac{{{\varepsilon_u}}}{{{\varepsilon_m}}} $$
(1)
$$ {R_r}=\frac{{{\varepsilon_m}-{\varepsilon_p}}}{{{\varepsilon_m}}} $$
(2)

In order to determine the recovery stress, the predeformed samples at 100 % strain at various predeformation temperatures of 25, 80, 120 °C were heated with full-strain constraint, i.e. maintaining 100 % strain. The recovery stress was recorded at various recovery temperatures from room temperature to 160 °C.

Results and discussion

Morphological observation

TEM experiments were carried out to delineate the dispersion status of TPU in the PLA matrix. Figure 1 illustrates the TEM micrographs of PLA/TPU blends for 70/30 and 50/50 composition. Basically, the images indicated a good compatibility between PLA and TPU as mentioned in the literature [19–21]. A good dispersion of TPU particles was found for the PLA/TPU (70/30) case. A co-continuous phase of TPU and PLA was found for the blend composition of 50/50. Surprisingly, the investigated blends showed the improved shape memory effect of PLA with the addition of TPU. The illustration of shape recovery process of PLA/TPU blends (50/50) was shown in Fig. 2. In a very short time of a few seconds, the predeformed sample was recovered to the original shape. This work attempts to investigate this shape memory effect in great detail as discussed in the following sections.

Fig. 1
figure 1

TEM images of PLA/TPU blends (× 5000, scale bar: 2 μm) (a) 70/30, (b) 50/50

Full size image
Fig. 2
figure 2

Illustration of shape recovery process of PLA/TPU blends (50/50) at 90 °C (a) original sample (b) 0 s (c) 1 s (d )2 s

Full size image

Shape memory effect

The good shape memory effect depends on both good shape fixing ratio and shape recovery ratio. Figures 3, 4 and 5 show the effect of recovery temperatures for PLA/TPU blends at various compositions under different predeformation temperatures. Three predeformation temperatures, 25 °C, 80 °C, and 120 °C, were conducted to predeform the test samples. At the predeformation temperature of 25 °C, PLA was hard and brittle at room temperature, so it couldn’t be deformed up to 100 % without failure during the test and no shape memory effect was discussed. As for TPU, it showed highest recovery ratio due to its segmented structure serving as physical crosslinks, and possessed strong elastic behaviors. Surprisingly, the addition of TPU significantly increased the shape recovery ratio of PLA up to 93.5 ± 0.4 % with increasing recovery temperature, reaching 160 °C for the PLA/TPU (50/50) blends. This recovery ratio decreased slightly at lower TPU composition. The first work to observe this shape recovery phenomenon at room temperature was surprisingly reported recently while a polyamide elastomer was incorporated into PLA in order to increase its impact strength [16]. Yet, no detail study in terms of shape recovery and fixing ratios was disclosed in their work. Our work attempts to investigate this recovery mechanism in detail. It was suggested that TPU in our work acted as a stress concentrator to dissipate deformed energy, and the good compatibility of PLA and TPU contributed to the observed predeformation without premature failure through restructured PLA at room temperature. This high internal strain stored energy and low configurational state entropy at low predeformation temperature was easily relieved upon heating to recover to its original shape [21]. Thus, an instant, but not significant, shape recovery even at a lower recovery temperature, such as 40 °C, was observed, suggesting a higher recovery onset speed for samples predeformed at this low predeformation temperature.

Fig. 3
figure 3

Shape recovery ratio of PLA/TPU blends predeformed at 25 °C with different recovery temperatures

Full size image
Fig. 4
figure 4

Shape recovery ratio of PLA/TPU blends predeformed at 80 °C with different recovery temperatures

Full size image
Fig. 5
figure 5

Shape recovery ratio of PLA/TPU blends predeformed at 120 °C with different recovery temperatures

Full size image

For the predeformation at 80 °C, PLA began to show shape memory effect at various recovery temperatures. The crystalline region of PLA acted as the physical crosslink points was partially dislocated in a same way like the deformed amorphous region through molecular orientation at this predeformation temperature above its glass transition temperature of 62.4 °C. On releasing the tensile load, the sample shape was kept in a certain degree while cooled to the room temperature. The stored energy was in general smaller than that at low predeformation temperature due to a typical viscoelastic nature of general polymer. On the other hand, the relatively higher configuration state entropy was expected owing to sufficient predeformation. Thus, the PLA sample still could regain its shape on further heating above its glass transition temperature. Apparently, the addition of TPU again significantly improved the recovery ratio of PLA as seen in the samples predeformed at low temperature regardless of TPU compositions. The highest value reached up to 91.4 ± 1.4 % for PLA/TPU (50/50) blends. Interestingly, the sharp increases in the recovery ratio for blends predeformed near the glass transition temperature of PLA were observed in comparison with the gradual, but incipient, increases in the shape recovery curves for those predeformed at room temperature. This would lead to different practical applications depending on the requirement of predeformed condition.

For the high predeformation temperature of 120 °C, it is interesting to see if the shape memory effect improved above the glass transition temperature. However, PLA was too weak to deform at high elongation, leading to no shape memory results available in this case. Yet, the addition of TPU surprisingly did assist the deformation of PLA and increased the recovery ratio of PLA with increasing recovery temperature as seen in the low predeformation temperature. This internal strain stored energy and configurational state entropy at this high predeformation temperature above the glass transition temperature were among the lowest values due to the high molecular chain mobility at high temperature. So, the least incipient and gradual increases in the shape recovery curves with increasing recovery temperature for the blends was observed, and the recovery ratio gradually increased till 160 °C for PLA/TPU (50/50) blends with increasing recovery temperature unlike an abrupt change near the glass transition temperature as seen in the medium predeformation temperature of 80 °C. For better comparison, Fig. 6 shows the effect of the predeformation temperatures on the recovery ratio for PLA/TPU blends at a typical recovery temperature of 80 °C. As discussed earlier, the high internal strain stored energy and low configurational state entropy at low predeformation temperature gave highest recovery ratio leading to the instantly stress relieved upon heating. In addition, since neat TPU showed highest recovery ratio in all cases due to its segmented structure serving as physical crosslinks to possess strong elastic behaviors, the recovery ratio increased at higher TPU content through a higher degree of physical crosslinks. This was also attributed to TPU acting as a stress concentrator to avoid the premature failure of PLA and the good compatibility of PLA and TPU.

Fig. 6
figure 6

Shape recovery ratio of PLA/TPU blends recovery at 80 °C with different predeformation temperatures

Full size image

To better understand the shape fixing ratio for the typical PLA/TPU (50/50) blends, Fig. 7 shows the results at different predeformation temperatures for a specific recovery temperature of 80 °C. With the increase of predeformation temperature, the shape fixing ability enhanced, but the ability of shape recovery decreased as seen earlier in Fig. 6. As discussed in the earlier section, this was attributed to the high strain energy and entropy loss for the predeformation at room temperature. When the stress was released, the polymer chains would tend to easily revert to its original shape in a certain degree, resulting in poor shape fixing ability, but relatively better shape recovery ability. On the other hand, for the predeformation at high temperature, especially above the glass transition temperature, the polymer chains with dislocated crystalline domains tended to move more easily with less degree of molecular orientation. When samples were cooled to room temperature with relieved stress, high degree of permanent set with high shape fixing ratio was observed due to the high degree of dislocation of crystalline domains or disentangled molecular chains. A similar situation was found for PLA/TPU (70/30) blends.

Fig. 7
figure 7

Shape fixing ratio of PLA/TPU blends with different predeformation temperatures

Full size image

To further briefly consider the effect of the thermal behaviors of PLA after modification on the shape memory effect, the glass transition temperature, melting temperatures, and the crystallinity were also reported. The DSC traces were omitted here for brevity. It was found that the glass transition temperature of neat PLA (about 62.4 °C) didn’t vary much for the PLA/TPU (70/30) and (50/50) blends at 61.2 °C and 62.4 °C, respectively. Therefore, the effect of the variation of glass temperatures was not considered. Similarly, the melting temperatures of PLA about 164.4 °C remained roughly the same as 162.7 °C and 163.2 °C for PLA/TPU (70/30) and (50/50) blends, except for additional minor shoulder peaks at 167.7 °C and 167.5 °C for respective cases. This could be attributed to the melting behaviors of the reorganized perfect crystals from imperfect crystals at higher temperature [22]. In addition, the crystallinity of neat PLA at 47.0 % decreased slightly to 42.1 % and 41.4 % for PLA/TPU (70/30) and (50/50) blends, respectively. Although, TPU tended to suppress the crystallization of PLA in our cases, the shape memory effects were still improved. A further study is needed to investigate those thermal behavior effects in detail.

Note that the storage time at room temperature would affect the shape memory effect in a certain degree through the decrement of stored energy and configurational state entropy due to molecular chain relaxation. Yet, the glass transition temperature of PLA is still higher than room temperature. Therefore, we didn’t expect a large degree in deterioration. It is quite interesting to evaluate this storage effect in terms of practical applications of shape memory polymers further.

Shape memory index

Some material like conventional rubber, it showed very good elastic recovery, but failed to keep a fixed deformation. Thus, it was considered as an elastic material, instead of shape memory polymer. In order to consider both fixing and recovery ratios together, an additional denotation of shape memory index, Rf*Rr, to couple both ratios by the multiplication of fixing ratio and recovery ratio was considered. Even though there is no much physical meaning for this coupling denotation, but it is quite informative to have this distinct value as an index. When the value of shape memory index was closed to 1, the materials conferred the highest shape memory effect. Figures 8, 9 and 10 show the results of the shape memory index at different predeformation temperatures. At room temperature, although the TPU exhibited good shape recovery ratio, but the shape fixing ratio was poor, so the values of shape memory index were only between 14 and 19 %. The shape memory index for PLA/TPU (50/50) blends reached equilibrium value of about 62.1 ± 2.6 % above the recovery temperature of 60 °C. For the PLA/TPU(70/30) blends, the shape memory index reached about 66.9 ± 1.7 % above the recovery temperature of 80 °C. Therefore, PLA/TPU (70/30) appeared to show the slightly higher shape memory index at room temperature.

Fig. 8
figure 8

Shape memory index of PLA/TPU blends predeformed at 25 °C with different recovery temperatures

Full size image
Fig. 9
figure 9

Shape memory index of PLA/TPU blends predeformed at 80 °C with different recovery temperatures

Full size image
Fig. 10
figure 10

Shape memory index of PLA/TPU blends predeformed at 120 °C with different recovery temperatures

Full size image

At the predeformation temperature of 80 °C, the shape memory index for PLA reached about 78.0 ± 1.4 % at the recovery temperature of 160 °C. The shape memory index for PLA/TPU(50/50) reached equilibrium about 85.5 ± 1.9 %, slightly higher than that of PLA/TPU(70/30) at 81.6 ± 1.7 % . At the predeformation temperature of 120 °C, in Fig. 10, the shape memory index slowly increased with the increase of recovery temperature. When the recovery temperature reached 160 °C, shape memory index for the PLA/TPU(50/50) blend was 87.0 ± 1.2 %, which was the highest value among other blend compositions and neat resins.

The shape memory index for a typical PLA/TPU (70/30) blend at different predeformation temperatures with different recovery temperatures was also compared in Figs. 8, 9 and 10. Below the recovery temperature of 60 °C (close to the glass transition temperature), the shape memory index at the predeformation temperature of room temperature was the highest. However, when the recovery temperature was above the glass transition temperature, the shape memory index at the predeformation temperature of 80 °C was the highest value. Interestingly, a continuous increase of this index was found for the predeformation temperature of 120 °C. This difference in the onset of recovery curves with various recovery temperatures was mentioned earlier in the discussion of shape recovery ratio. A similar observation was found for PLA/TPU (50/50) blends as shown in Figs. 8, 9 and 10.

To further have an in-depth understanding on the shape recovery mechanism, the recovery stress was recorded against temperature while the predeformed samples were heated under full-strain constraint, instead of conventional free-strain recovery. Figure 11 shows the recovery stress responses for PLA/TPU blends. For the blend composition of PLA/TPU (70/30), the recovery stress showed a maximum peak with increasing temperature for all samples at three different predeformation temperatures. The recovery stress in general stemmed from both the internal strain stored energy and configurational state entropy. At low predeformation temperature, the high internal strain stored energy and low configurational state entropy was easily relieved upon heating to recover to its original undeformed shape [21]. Thus, its maximum recovery stress occurred at the low temperature range. In particular, for samples predeformed at 80 °C, the recovery stress reached the maximum value near 80 °C, close to the glass transition temperature of PLA at 62.4 °C. The recovery stress then decreased further with increasing temperature, as most molecular chains were disorientated above this temperature range. This maximum peak indicated the highest recovery stress among three different predeformation temperatures. For the predeformation temperature at 80 °C, the crystalline region of PLA acted as the physical crosslink points was dislocated in a same way like the deformed amorphous region through molecular orientation within the PLA/TPU phases. Yet, TPU exhibited physical crosslink points were capable of stored elastic energy, leading to the highest recovery response near the glass transition temperature of PLA. For the predeformation temperature at 120 °C, this internal strain stored energy and configurational state entropy at this high predeformation temperature above the glass transition temperature were among the lowest values due to the high molecular chain mobility at this temperature. Interestingly, the recovery stress peak temperatures largely corresponded to the predeformation temperatures, indicating a unique “memory” effect in terms of recovery stress, which was first observed for this type of bio-based blends in the literature, to our best knowledge. Likewise, a similar observation was observed for the blend composition of PLA/TPU (50/50) as shown in Fig. 12. The recovery stress peak temperatures largely corresponded to the predeformation temperatures, and increased with increasing the predeformation temperatures. Miaudet et al. [23] observed a similar behavior for the PVA/nanotube nanocomposites, indicating that the nanotubes induced a broadening of the glass transition and a “memory” temperature with a recovery stress peak at their predeformation temperature. In addition, Amirian et al. [24] reported an increase in the stress recovery stress by adding nanotube into the poly(L-lactide-co-Ε-caprolactone) matrix. Although, in our cases, there were no nanotube fibers filled in the matrix, yet the TEM images portrayed earlier in Fig. 1 indicated a similar role of TPU phase in our work as their nanotube to serve as a reversible phase. Thus, the prepared blends still showed the correlation between recovery stress peak temperatures and predeformation temperatures. The current results presented the first work to layout the recovery stress mechanism of the bio-based blends under different predeformation temperatures.

Fig. 11
figure 11

Recovery stress and temperature of PLA/TPU(70/30) blends with different predeformation temperatures

Full size image
Fig. 12
figure 12

Recovery stress and temperature of PLA/TPU(50/50) blends with different predeformation temperatures

Full size image

Conclusions

To improve the shape memory effect of PLA, TPU was melt-blended with PLA to form shape memory bio-based blends. PLA/TPU blends (70/30 and 50/50) were predeformed under three different predeformation temperatures of 25, 80, and 120 °C, respectively. The shape memory behaviors of predeformed samples were investigated at various recovery temperatures. At the predeformation temperature of 25 °C, the addition of TPU significantly increased the shape recovery ratio of PLA up to 93.5 ± 0.4 % at 160 °C for the PLA/TPU (50/50) blends. With the increase of predeformation temperature, the shape fixing ability enhanced, but the ability of shape recovery decreased. In order to consider both fixing and recovery ratios together, an additional denotation of shape memory index to couple both ratios by the multiplication of fixing ratio (Rf) and recovery ratio (Rr) was considered. The sharp increases in the shape memory index curves against recovery temperatures for blends predeformed near the glass transition temperature of PLA were observed in comparison with the gradual, and earliest or latest, increases in the shape recovery curves for those blends predeformed at room temperature or high temperature. This difference in the onset of shape memory index curves with various recovery temperatures stemmed from the different degrees of internal strain stored energy and configurational state entropy at three predeformation temperatures. To further have an in-depth understanding on the shape recovery mechanism, the recovery stress was recorded against temperature while the predeformed samples were heated under full-strain constraint, instead of conventional free-strain recovery. The recovery stress showed a maximum peak with increasing temperature for all samples at three different predeformation temperatures. The recovery stress peak temperatures largely corresponded to the predeformation temperatures and increased with increasing the predeformation temperatures, indicating a unique “memory” effect in terms of recovery stress, which was first observed in the literature for this type of bio-based blends.