Compatibilisation of various PLA/thermoplastic elastomer blends with diisocyanate coupling agent

Abstract

Poly(lactic acid) (PLA) is characterised by its inherent brittleness, a detrimental feature for the production of durable bioplastics. PLA has been toughened by a low amount (12 wt-%) of various thermoplastic elastomers (TPE) including poly(ether-b-ester) (PEEs), poly(ether-b-amide) (PEBA) and poly(ether-b-urethane) (PEU). PLA–TPE blends were prepared by using a twin screw extruder. Ductility and impact resistance can be slightly improved with the incorporation of TPEs but but PEBA appears the most efficient. Reactive compatibilisation has been performed with the addition in the melt of a low amount (2 wt-%) of 4,4-methylene diphenyl diisocyanate. All compatibilised blends exhibit high toughness with similar ductility. These blends preserve good stiffness and high tensile strength. Compatibilised PEBA blends can be considered as super tough poly(lactic acid) materials. This work confirms that the flexibility of the elastomer together with the quality of the interfacial adhesion between the rigid (PLA) and the soft (TPE) phases are the primary factors influencing the toughness.

Keywords

PLA thermoplastic elastomer (TPE)blends MDI compatibilisation toughening mechanical properties thermal properties

Introduction

Poly(lactic acid) (PLA) is a bio-based, biodegradable and biocompatible aliphatic polyester (see Figure 1 for the chemical structure) with promising properties. First used in the biomedical sector, PLA is currently investigated for a large number of commodity applications and represents a viable alternative to substitute oil-based plastics. However, amorphous PLA is characterised not only by its inherent brittleness but also by its poor heat resistance which is detrimental for many applications notably for the elaboration of durable bioplastics.

High yield strength and high ductility are the key characteristics in order to toughen the brittle polymers. Among the many strategies tested to reinforce PLA [1], the incorporation of immiscible thermoplastic elastomer (TPE) remains one of the most effective and cost efficient [2]. TPEs are linear segmented copolymers alternating rigid (polyamide, polyester, polyurethane, polystyrene) and flexible (polyester, polyether or polysiloxane, polyethylene-co-butylene) segments. Various commercial (A–B) _n type TPEs have been incorporated into PLA including poly(ether-b-ester) (PEEs) [3, 4], poly(ether-b-amide) (PEBA) [5-7], poly(ether-b-urethane) (PEU) [8, 9] and poly(ester-b-urethane) [10-12]. The chemical structures of PEBA, PEEs and PEU are given in Figure 2. PLA/TPE blends can be obtained by simply mixing the two components in the melt state. As PLA and TPE are immiscible and incompatible, the addition of a compatibiliser is mandatory to increase the cohesion between the two phases and maximise the physico-chemical properties. Multiple strategies have been applied to compatibilise PLA–TPE blends via reactive processing [10, 13-16]. For instance, the addition of small molecular weight chemicals or polymers with reactive groups has been used to form in situ bloc copolymers that act as compatibilisers. Polymers containing reactive glycidyl [17] or maleic groups [3] and small molecules with carbodiimide [18], epoxide [19], caprolactam [20], oxazoline [21] and isocyanate moieties [22-25] have been used for such purpose.

Recently, the authors studied the thermal, morphological and mechanical properties of both PLA/PEEs and PLA/PEBA blends prepared by melt blending [3, 6]. It was shown that these blends need to be compatibilised in order to promote interfacial adhesion and enhance mechanical properties. The performances of various coupling agents to compatibilise these blends were compared. 4,4-Methylene diphenyl diisocyanate (MDI) was found to be as one of the more efficient coupling agent for toughening PLA/PEBA blends compared to polymeric poly(ethylene-co-methylacrylate-co-glycidyl methacrylate) (PEAGM), poly(maleic anhydride-alt-1-octadecene) (PMAOD) and small difunctional reactive coupling agents such as 1,1′-carbonyl-bis-caprolactam (CBC). MDI has a positive impact in terms of toughening effect [26]. Notched Izod and Gardner impact strengths were increased compared to their homologous blends without MDI. The toughening effect was ascribed to the cross-linking of the rigid segments of PEBA (see Figure 2(a) for the chemical structure of PEBA) together with the in situ reactive compatibilisation of bloc copolymers making the interfacial adhesion optimal [6]. MDI is known to react primarily with hydroxyl groups to form urethane linkage. Some possible coupling reactions prevailing between MDI and hydroxyl end groups are exemplified in Figure 3(a–c) in the case of PLA–PEBA blends. Carboxylic groups can also react with isocyanate to form amide linkage (reaction not shown) while when used in excess MDI can also react with urethane and amide groups to form branches or even cross-linked structures via the formation of allophanate or urea bonds. The possible reactions are detailed in several recent articles [6, 10, 11].

Comparative studies of PLA blends containing different types of TPE, prepared and characterised in the same experimental conditions, are rather scarce [27]. In addition, such a comparison has not yet been reported for homologous blends compatibilised with MDI. In the present study, we report and compare the effect of a low amount of coupling agent, MDI (2 wt-%), on morphological, thermal and mechanical properties of three different types of blends, namely PLA–PEEs, PLA–PEBA and PLA–PEU. All the elastomers are characterised by flexible poly(tetraoxymethylene) (PTMO) segments but differ from their rigid segments. Polyamide 12 (PA₁₂), polybutylene terephthalate (PBT) and aromatic polyurethane (PU) are the rigid segments of PEA, PEEs and PEU elastomers, respectively. The physico-chemical characteristics and the chemical structures of the elastomers are given in Table 1 and Figure 2(a–c), respectively. The selected TPEs correspond to the most flexible grades commercially available that can be processed by the conventional extrusion technique. There are still significant differences in the level of flexibility between the different elastomers based on the flexural modulus reported (see Table 1). Among the three selected TPEs, PEBA is the most flexible with the highest fluidity. PEU is the most rigid and has the highest glass transition temperature for its soft phase. PEEs has an intermediate flexibility but disposes of a high viscosity compared to the others. All the TPEs are biocompatible, sterilisable and can be used in various medical applications such as for catheters, connectors, flexible tubings, films, seals and gaskets for medical devices. The series of blends studied in this work are characterised by a low elastomer content (12 wt-%). This composition has proved to confer relatively little improvement in the mechanical performances to the blends compared to neat PLA and was chosen for further compatibilisation studies with MDI.

Table 1

Physical characteristics of the various TPE elastomers.

Samples	MFR^a	T _g ^s _(DMTA) ^b	T _{d (5%)} ^c	E _(Flexural) ^d
(g per 10 min)	(^oC)	(^oC)	(MPa)
B (PEBA)	42.9 (1.6)	−52	381	15.9
T (PEU)	20.7 (1.0)	−15	323	26.9
H (PEEs)	10.8 (0.3)	−53	377	21.0

^aMeasured according ASTM 1238 (see the Experimental part for more details).

^bT_g^s is the glass transition temperature of the soft phase (see detection conditions described in the Experimental part).

^cSee the Experimental part for more details.

^dObtained from technical data sheets (measured according ASTM D 790).

Experimental part

Materials

Poly(lactic acid) (PLA 2003D, 4.1% D-unit) was supplied by NatureWorks LLC. Hytrel® 3078 (shore D: 30, ASTM D 2240), obtained from DuPont, is a plasticiser free segmented copolymer alternating PBT (34 wt-%) and PTMO segments (PTMO; 64 wt-%) and contains colour-stable antioxidants. Pebax® 2533 SA 01 (shore D: 27, ASTM D 2240) was bought from Arkema. It is a segmented copolymer (PEBA) alternating PTMO (85 wt-%) and PA₁₂ (12 wt-%) segments and also contains 3 wt-% of adipic acid used for carboxylic acid termination of PA₁₂ [28]. TEXIN® 985, an aromatic PEU, characterised by a low nominal hardness (shore D: 33, ASTM D 2240), was bought from Bayer Material Science. MDI was purchased from Sigma Aldrich. Notations ‘P’, ‘B’, ‘H’, ‘T’ and ‘M’ are reserved for PLA, PEBAX®, HYTREL®, TEXIN® and MDI, respectively. For instance, PT₁₂M₂ is a blend containing 86 wt-% of PLA, 12 wt-% of PEU (TEXIN®) and 2 wt-% of MDI.

Compounding procedures

PLA and TPEs were dried at 70°C during 4 hours in a Conair dryer before extrusion. MDI was used as received. All blends were prepared by melt mixing using a co-rotating Coperion ZSK25WLE twin screw extruder. The screw has a length of 1040 mm, a diameter of 25 mm and L/D ratio of 41.6. The screw configuration and the mode of incorporation of the various ingredients were detailed elsewhere [6]. The rotation speed was set at 80 rev min⁻¹ and the temperatures profile was 180/190/200/210/210/200°C. A vacuum pump placed at the exit of the extrusion barrel was used to eliminate gas and humidity. All blends were extruded under the form of threads (4 mm diameter), dipped into a cold water bath, dried by compressed air and granulated by a strand pelletiser granulator (Scheer Bay BT25).

Specimens preparation (injection moulding)

Compounded pellets were dried under the conditions described above. Specimens were moulded using an ARBURG Allrounder 370 A 66 tons injection press having a 25-mm screw and an L/D ratio of 17.5. Mould temperature was set at 25°C and melt temperature at 190°C. Temperature profile was adjusted at 185/204/200/196/196°C while the other technical parameters were described elsewhere [6].

Melt flow rate (MFR) measurements

MFR measurements were carried out with a Dynisco polymer test system (LMI 40002) according to ASTM 1238 ‘procedure A’ (2.16 kg, 210°C). MFR was measured from pellets stored at 23°C and 40% humidity. At least five tests with ∼6 g of material were performed to calculate the mean and standard deviation.

Solubility tests

Solvent tests were performed in order to reveal any cross-linking reaction taking place during melt processing due to the addition of a coupling agent. Neat polymers, compatibilised and incompatibilised blends were tentatively dissolved at room temperature in various solvents [chloroform (CHCl₃), hot (100°C), dimethylformamide (DMF), tetrahydrofuran (THF)] at a concentration of ∼1% (w/v). A visual examination was carried out in order to detect the presence of any gels in the preparation.

Differential scanning calorimetry (DSC)

DSC measurements were performed on a Mettler Toledo 822^e calibrated with indium (In), flushed with a nitrogen flow (50 mL min⁻¹) and operated in sub-ambient temperature mode with an Intracooler Haake model EK90/MT. The experiments were run on 40 µL aluminium crucibles with perforated aluminium lids. Granules of blends obtained from extrusion were scanned between −30°C and 210°C at a heating rate of 10°C min⁻¹. After heating, samples were cooled to −30°C at the rate of 10°C min⁻¹. The glass transition temperatures (T_g) were taken from midpoint of the heat capacity jump. The scans were reproducible after the initial heating scan. The crystallisation temperatures (T_C) were obtained from the maximum of the recrystallisation peak on the heating scan for PLA. The melting temperatures (T_m) were taken from the maximum of the peak. T_g, T_c and T_m were obtained from the second heating scan. The degrees of crystallinity, X, were obtained from the second scan. The area of the melting peak was used to calculate the crystallinity content for PLA in the blends by using ΔH_m, ΔH_c and the heat of fusion of the 100% crystalline polymers (93.7 J g⁻¹ for PLA [29]). The degree of crystallinity calculated using Equation (1) was corrected by taking into account the PLA fraction contained in the blend. (1)

Thermogravimetric analysis (TGA)

Thermogravimetric analysis was performed on a Mettler Toledo TGA/SDT/A851e/LF/1600 between 25°C and 600°C at a heating rate of 10°C min⁻¹ in a nitrogen flow (50 mL min⁻¹). The experiments were run on samples of ∼10 mg disposed in a 100-µL aluminium crucible without lids. Thermal degradation temperatures, T_d and T_p, were measured at 5% weight loss relative to the weight at 25°C of the undried samples and at the maximum of the peak of the derivative curve, respectively.

Dynamical mechanical thermal analysis (DMTA)

The dynamical mechanical thermal analysis was performed by using DMA Q800 V20.24 Build 43 (TA Instruments) calibrated with stainless steel and polycarbonate specimens in dual cantilever mode. The dimensions of the specimens were 35 × 10 × 2 (±0.02) mm. Injection-moulded specimens were conditioned at 23°C and 40% moisture for at least 1 week. Samples were scanned from −120°C to 140°C at a rate of 3°C min⁻¹ using liquid nitrogen. Frequency and drive amplitude were set to 1 Hz and 15 μm, respectively. The damping function (tan δ) was recorded as a function of temperature. For blends, glass transition temperatures of the rigid and soft phases were determined at the maximum of the peak seen in the tan δ curve.

Heat distortion temperature (HDT)

The heat distortion temperature was obtained with a HDT-Vicat Ceast 22837 according to ASTM D648 ‘procedure B’ (width between the support points is 100 ± 0.05 mm). Injection-moulded specimens were conditioned at 23°C and 40% moisture for at least 1 week. The specimen was loaded with 0.455 MPa with three independent measuring stations. Temperature is increased at 2°C min⁻¹ from 25°C until the specimen deflects of 0.25 mm. Three temperatures were measured to calculate the mean and the standard deviation.

Scanning electron microscopy (SEM)

The fracture surfaces of neat polymers and blends were observed on a Hitachi model SU1510 scanning electron microscope. Samples were submerged in liquid nitrogen for 10 min before being fractured using the notched Izod test. Surfaces were coated with a thin layer of gold/palladium and observed at an acceleration voltage of 7 kV and a magnification of 8000×.

Traction tests

Traction tests were performed with a Zwick/Roell Z030 machine equipped with a 30-kN load cell and a video extensometer. Injection-moulded dog-bone shaped specimens (type 1 according to ASTM D638-10) were conditioned at 23°C and 40% moisture for at least 1 week. Young's modulus (E), maximum constraint (σ) and deformation at rupture (ε) were measured at a traction speed of 5 mm min⁻¹. The gap between grips was 115 mm and the gap between targets was 50 mm. Inconsistent data were eliminated and a minimum of eight samples were used to calculate the mean and standard deviation.

Izod impact (notched)

Tests were performed on notched samples with an Instron Ceast 9050 Impact Pendulum System according to ASTM D256-10. Notches were made by a Notching Cutter model TMI 22-05 from Testing Machines Inc. and measured with a Mitutoyo 547-312S Absolute Digimatic Thickness Gage. After notching, samples were conditioned at 23°C and 40% moisture for at least 2 days. Hammers with 1–2.75 J of potential energy were used depending on the velocity variation. At least 10 specimens served to calculate the mean and standard deviation.

Results and discussion

Thermal properties of the blends

DSC traces of neat PLA, PEBA, PEU, PEEs and blends are shown in Figure 4 and data are given in Table 2. Trace of neat PLA was characterised by a glass transition (T_g) at 60°C evidenced by a clear jump in the heat flow. The endothermic transition observed at 117°C is associated with the PLA cold crystallisation and points out the difficulty to crystallise. Melting transition was observed at 158°C.

Table 2

Data obtained from DSC and DMTA measurements for PLA and PLA/TPE blends.

Sample	T _g _(DSC)	T _g,s _(DMTA)	T _g,h _(DMTA)	T _{c,o (DSC)}	T _{c,m (DSC)}	T _{m (DSC)}	ΔH_{m (DSC)}	X _(DSC)
Sample	(°C)	(°C)	(°C)	(°C)	(°C)	(°C)	(J g⁻¹)	(%)
PLA	60	–	71	102	117	151	26	28
PH₁₂^a	60	−62	72	111	123	152	20	21
PH₁₂M₂	61	−60	64	112	∼124	154	16	17
PB₁₂^b	60	∼−65, −47	72	112	∼121	152	27	29
PB₁₂M₂^b	62	−65	63	119	∼133	154	5	5
PT₁₂	60	−30	68	109	∼120	152	25	26
PT₁₂M₂	62	∼−24	63	111	∼122	153	24	26

^aSee Ref. [3] for other thermal and thermomechanical data.

^bSee Ref. [6] for a more detailed discussion about the thermal properties of this blend.

Thermal behaviour of PEEs and PEBA elastomers was described in details elsewhere [3, 6]. For the three elastomers, no glass transition can be observed in the second DSC scan. For PEBA, the two endotherms seen at low (11°C) and high (∼165°C) temperature correspond to the melting transition of PTMO and PA₁₂ segments, respectively [6]. Similar observations were reported elsewhere for PEEs [3] with two transitions observed at 4 and 170°C and associated with the melting of PTMO and PBT segments, respectively. In contrast with latter's, PEU does not present any specific feature in the second and subsequent DSC scans since no crystallisation peak corresponding for PTMO and urethane segments of the PEU can be seen.

For incompatibilised blends (PT₁₂, PB₁₂, PH₁₂), the position of the glass and melting transitions was observed at 60°C and 158°C, respectively (see Table 1). For PH and PB series varying the elastomer content, similar results were reported elsewhere [3, 6]. These works based on DSC, DMA and SEM also lead to the conclusion that PLA/PEEs and PLA/PEBA blends were immiscible. Based only on DSC results, the incorporation of PEU into PLA seems not influence the glass transition of the rigid phase reflecting an apparent purity of the latter. Other works have suggested that some degree of miscibility exists between PLA and thermoplastic PU [30].

PEEs, PEU and PEBA elastomers shift the position of the exothermic crystallisation peak of PLA at higher temperature compared to neat PLA. The onset of crystallisation, T_c,o, is increased by 7–10°C while the maximum of the peak, T_c, increased between 4°C and 7°C compared to neat PLA. This indicates that non-isothermal crystallisation is hindered by the incorporation of TPEs into PLA. However, while the interference of PEBA and PEEs with the crystallisation process is pretty similar, the impact of PEU is less marked. It can also be noticed that for PH₁₂ and PB₁₂ blends, PTMO melting peak is barely observed on the second DSC scan.

Based on DSC measurements, the glass transition of the PLA rigid phase is slightly increased (1–2°C) by the incorporation of MDI in the blend (Figure 4). MDI has a different impact on the crystallisation of PLA depending on the type of elastomer. In combination with PEU, the PLA crystallisation rate is only slightly affected but crystallinity degree is unchanged when compared to the incompatibilised blend. With PEEs, MDI has a moderate effect on the rate of PLA crystallisation while the extent of the PLA crystalline phase is markedly reduced. In contrast, when used with PEBA, MDI almost suppressed the crystallinity of the PLA rigid phase. For PH and PB blends, the impact of MDI on the crystallisation process of the TPE rigid segments cannot be obtained by DSC due to the weak extent of crystallinity of the rigid phase and to the superposition of PLA and TPE crystallisation process occurring in the same temperature range. However, although barely visible in Figure 4, the crystallisation rate of the soft PTMO segments is markedly reduced by the MDI incorporation since the melting peak is shifted to higher temperatures.

Thermomechanical properties of the blends

Dynamic mechanical experiments were employed to investigate the effect of MDI coupling agent on the compatibility of the PLA–elastomer blends. The damping function (tan δ) of the blends is shown in Figure 2 and the data obtained from the latter's given in Table 2. In the high temperature range (50–100°C), the damping traces of the blends show a clear α-relaxation peak associated with the glass transition of the rigid PLA phase. For neat PLA, the T_g value is observed at 71°C. Compared to PLA, PB₁₂ and PH₁₂ blends display a slight increase (1–2°C) – which is within the experimental error related to the DMA measurements – while for PT₁₂ a moderate decrease of T_g was observed at 68°C. This could indicate a slight miscibility of PEU in the PLA rigid phase as reported elsewhere [8, 11, 30, 31]. Surprisingly, in contrast with what was observed by DSC measurements, all the blends compatibilised with MDI exhibit a significant decrease in T_g (5–9°C) suggesting a plasticisation effect. It could be conjectured that the flexible PTMO segments of the in situ bloc copolymers formed during melt mixing act as plasticisers in the PLA rigid phase. Indeed similar polyethers such as polyethylene glycol (PEG) [32-35], polypropylene glycol (PPG) [32, 36], polytrimethylene glycol (PTG) [37] and block copolymer of PEG and PPG (Pluronic) [38] are known to be efficient plasticisers. However, the discrepancies between DSC and DMTA data mentioned above still remain unclear and would deserve additional investigations. To close on this point, it is worth mentioning the difference existing in the experimental procedures between the two techniques. DSC data were obtained after melting the sample (see the Experimental Part for more details), i.e. after the thermal and mechanical history have been cleared. This is not the case for samples analysed by DMTA since specimens were simply stored for at least 1 week at 23°C and 50% RH before being tested (see the Experimental Part) (Figure 5).

HDT is an index of heat resistance towards applied static load. HDT values of PLA and PLA/TPE blends are reported in Table 3. Neat PLA is characterised by an HDT value of 54°C which is in agreement with the value of 55°C given in the literature for the amorphous PLA [2]. As expected, the incorporation of a small amount (12 wt-%) of flexible PEU or PEEs elastomers into PLA decreases the HDT value by 3°C. For PEBA blend, no change is observed but the HDT value (55°C) remains close to, but still below, the PLA T_g (61°C). In addition, a very low percentage of PLA and PEBA crystallinity cannot be completely excluded (see ref [6]) and can slightly impede the chain mobility and limit the loss of heat resistance.

Table 3

Data obtained from MFR, TGA and HDT measurements for PLA and PLA/TPE blends.

Sample	MFR (s.d.)	T _d	HDT (s.d.)
Sample	(g per 10 min)	(°C)	(°C)
PLA	8.5 (0.3)	340	54 (0.2)
PB₁₂	15.6 (0.5)	337	55 (0.2)
PB₁₂M₂	3.5 (0.0)	324	57 (0.2)
PT₁₂	9.0 (0.0)	322	51 (0.0)
PT₁₂M₂	3.0 (0.1)	316	57 (0.2)
PH₁₂	7.1 (0.2)	326	51 (0.0)
PH₁₂M₂	2.8 (0.1)	321	57 (0.1)

The incorporation of MDI into the binary blends allows overcoming the loss of heat resistance due the presence of PEU and PEEs. MDI increases the HDT value by 6°C for PEEs and PEU but less (2°C) for PEBA. The HDT values of the blends even exceed that of the neat PLA. This is not insignificant if we consider that an amount of at least 10 wt-% of talc into an amorphous PLA is required to increase the HDT value of PLA by only 3°C [39]. This trend observed in particular for PEU and PHE blends points out the impact of the better quality of the interface between PLA and elastomeric phases on the thermomechanical improvement.

Melt viscosity and blend's behaviour in various solvents

The melt flow rate (index) is related to the melt viscosity and therefore to the molecular weight of the polymer. For immiscible blends for which a relative limited number of specific interactions between the two components of the blend are expected, MFI increases with the incorporation of the polymer of higher MFI. MFI results are given in Table 3. MFI of PEBA is significantly higher than that of pure extruded PLA (8.5 g per 10 min) and for this reason PB₁₂ has a greater fluidity than PLA. PT₁₂ has a similar melt flow to that of PLA while PH₁₂ exhibits a lower melt viscosity than PLA even though MFR of PEEs (H) is significantly higher (10.8 g per 10 min). During the blending process of PLA/PEEs blends (PH), exchange reactions (intermolecular alcoholysis and acidolysis and transesterification) were suspected [3]. However, it should be noticed that for all the blends, solutions prepared in either CHCl₃ or THF were found free of gels, indicating the absence of any cross-linked structures.

For all the elastomer-based blends, the addition of MDI coupling agent leads to an unequivocal decrease in MFI (see Table 3). This indicates that some specific interactions or reactions occur in the rigid phase (PLA), in the soft phase (TPE) and/or at the interface between the rigid and soft phases. The incorporation of MDI leads to the insolubility of the blends whatever the type of TPE used. The insolubility of PB₁₂M₂ in CHCl₃ was described in details elsewhere [6]. It has been invoked that the increase in the melt viscosity and the gelation of the medium of PB₁₂M₂ was related to an increase in the PLA molecular weights together with the formation of a chemical network/cross-linking of the amide segments, respectively.

The same observations have been done for PH₁₂ and PH₁₂M₂ which are also soluble and insoluble in CHCl₃, respectively. PT₁₂ is not soluble in CHCl₃ but it is in cold DMF. With MDI, PT₁₂ becomes insoluble, even in hot (100°C) DMF suggesting the formation of a cross-linked structure. It is well known that MDI reacts with both hydroxyl and less reactive carboxyl terminated end groups of PLA and TPE for forming extended homo PLA, homo TPE or linear segmented PLA–TPE copolymers (see Figure 2). It must be stressed that MDI can also react with urethane groups and forms allophanate grafted copolymers [10]. However, the fact that the reaction of isocyanates with hydroxyl groups is much faster than the reaction with urethane promotes the formation of PLA–PEU block copolymers which helps the slight miscibility of PEU into the PLA matrix.

Blends morphology

The morphology of both incompatibilised and MDI-compatibilised PLA/TPE blends was investigated by SEM. SEM images are presented in Figure 6. Incompatibilised blends are characterised by TPE globular domains, typical of the sea-island morphology. For all incompatibilised blends, domains are homogenously dispersed in the PLA matrix and characterised by submicronic dimension varying roughly between 230 and 870 nm (see Table 4). As shown in Figure 6 and in Table 4, sizes of PEU domains are smaller than those of PEEs and PEA. In addition, the cohesion between PEU and PLA phases seems better because the presence of clear gaps between the two phases is less visible. This is possibly due to the better compatibility of PLA and PEU as supported above by DMTA measurements.

Table 4

Size of the elastomeric domains measured from SEM images for the PLA/TPE blends.

Sample	Size
Sample	(µm)
PB₁₂	0.27–0.87
PB₁₂M₂	0.44–1.50
PT₁₂	0.23–0.51
PT₁₂M₂	0.46–1.20
PH₁₂	0.23–0.69
PH₁₂M₂	0.37–0.97

With the incorporation of MDI, the size of the elastomeric domains increased whatever the type of elastomer used (Figure 6(b, d, f)). The smallest domain sizes were found to measure around 400 nm while the biggest domains were found roughly between 1 and 1.5 μm. The increase in the size of the domains and the broadening of the size distribution are particularly evident for PB₁₂M₂ and PT₁₂M₂ (see Figure 7(b, d) and Table 4). For PH₁₂M₂, the size of the elastomeric domains is still in the submicronic range (Figure 7(f) and Table 4). It is also clear that these domains have been more or less deformed due to the cryogenic fracture. Still, this observation is more visible with PEBA than for PEEs. For PEEs based blends such deformation has been reported with the incorporation of different coupling agents such as PMAOD and PEAGM but in higher amount (12 wt-%). For PB₁₂M₂, clear stretched connections are present (Figure 6(b)) between the elastomeric domains and the matrix indicating the ductile tearing character of the blend.

Mechanical properties

Tensile tests were performed to compare the effects of the various types of elastomers on the mechanical properties of amorphous injection-moulded specimen. Young's modulus, tensile strength and elongation at break of the various blends are presented in Figure 7.

PLA samples show a tensile strength of 64.5 MPa, an elastic modulus of 3654 MPa and an elongation at break of 5%. Compared to processed PLA, incompatibilised blends (PT₁₂, PB₁₂ and PH₁₂) dispose of a lower Young's modulus (E) and tensile strength (σ) (Figure 7(a, b)). Such reduction in strength and rigidity is common with the incorporation of a softer second polymer. However, mechanical properties can be easily modulated according to the TPE used in the blends. Given its intrinsic flexibility (see Table 1), PEBA reduces more the Young's modulus (29%) and the tensile strength (42%) than do PEU and PEEs but is more efficient in increasing the elongation at break (533%) (Figure 7(c)). It can be seen that the elastic modulus of PT₁₂ is much lowered (16%) than PH₁₂ (8%) despite the fact that PEU is more rigid than PEEs (see Table 1). This result is possibly related to the slight miscibility of PEU in the rigid PLA phase resulting from hydrogen bonding prevailing between PLA and PEU [11].

Tensile tests were also performed to investigate the effects of MDI on the mechanical properties of the amorphous injection-moulded specimen. Surprisingly, the Young's modulus (stiffness) of the compatibilised blends is pretty the same (∼2600 ± 100 MPa) despite the fact that TPEs are different in terms of resistance to flexibility and fluidity. For the blends containing PEEs or PEU, the addition of MDI results in a decrease in the Young's modulus and tensile strength by 11–19% and 15–18% (with respect to the blends without MDI), respectively. This contrasts, with PB₁₂M₂, for which Young's modulus is maintained at a similar value and the tensile strength is increased by 5%. The elongation at break of all the PLA/TPE blends is greatly improved by the addition of MDI (326–1273%). Given the uncertainties, the ductility of the MDI blends is quite similar, with an average value of the elongation at break of 180 ± 20%.

Impact resistance is an important property to characterise the toughness of a polymeric material. It can be measured thanks to the notched Izod test which measures the energy needed to propagate the craze. Figure 8 compares the Izod impact strength of compatibilised and incompatibilised PLA/TPE blends and neat PLA. PLA presents an impact strength of 31 J m⁻¹, which corresponds to the typical energy of a brittle fracture during the impact. The incorporation of a low amount of a TPE elastomer (≤12 wt-%) into PLA provides a limited improvement (26–61%) of the impact strength. PEBA is more efficient than PEU and PEEs to increase the impact strength. The value of 50 J m⁻¹ measured for PB₁₂M₂ is probably due to its greater flexibility (see Table 1). For incompatibilised TPE/blends substantial improvement of the notched Izod impact strength can be observed at higher elastomer amount with a minimum of 16–20 wt-% depending on the TPE used [3, 6, 40].

When MDI is used as a coupling agent to compatibilise, impact toughening of PLA/elastomer blends is markedly improved (170–341%). A notched Izod impact strength of 341 J m⁻¹ is observed for PB₁₂M₂. This characteristic qualifies this blend as a super toughened PLA blend ((PLA blends can be considered as « super tough » when the Izod (notched) impact strength reached ∼350 J/m) 2. When compared to neat PLA, the impact resistance is increased 11, 8.5 and 5.5 times for PB₁₂M₂, PH₁₂M₂ and PT₁₂M₂, respectively. This trend is coherent with the flexibility of the TPEs used in this work (see Table 1). Recently the authors prepared other super toughened PLA blends with PEAGM (as compatibilising agent instead of MDI) [3, 6]. The minimum amount of PEAGM required to achieve a value beyond ∼350 J m⁻¹ for PB₁₂ and PH₁₂ was 8 and 12 wt-%, respectively. However, for the latter's, severe reductions in mechanical properties [stiffness (∼15%), yield strength (∼20%) and ductility (∼60%)] were observed.

Conclusion

Reinforced PLA blends were prepared with a twin screw extruder by using three different TPEs, namely PEEs, PEBA and PEU. All these elastomers combine similar polyether-type (PTMO) soft segments but are different from each other based on the chemical nature – ester for PEEs, amide for PEBA and urethane for PEU – of the rigid segments. PEEs and PEBA elastomers form immiscible blends with PLA. In contrast, PEU seems slightly miscible with PLA based on DMTA measurements. All PLA–TPE blends are characterised by a sea-island morphology where elastomers’ domains are of a submicronic size. Mechanical properties were investigated on the amorphous blends prepared by injection moulding. Owing to its high flexibility, PEBA markedly decreases the Young's modulus and the tensile strength but is the most efficient in increasing the ductility and the impact resistance. PLA–TPE blends were further compatibilised by diisocyanate small molecules (MDI) via reactive extrusion. The addition of the MDI coupling agent decreases severely the fluidity of the blends and leads to their insolubility in classical solvents indicating the presence of a cross-linked network. It is clear from the MEB observations that the compatibilised blends are still phase separated systems. However, elastomeric (TPE) domains are larger in size than those observed in incompatibilised blends. Mechanical, thermomechanical (static and dynamic) experiments and impact strength measurements show the positive effects of MDI in compatibilising the soft elastomeric phase with the PLA rigid phase. The heat resistance values of the compatibilised blends (HDT) exceed that of the neat PLA. PLA–TPE blends are also characterised by a good stiffness (E > 2500 MPa), high tensile strength (σ > 40 MPa) and appropriate ductility (ε > 160%). This work confirms that the elastomer flexibility together with the quality of the interfacial adhesion between the PLA rigid phase and the TPE phases are the primary factors that influence the toughness of these blends.

Footnotes

Acknowledgements

The authors would like to thank Mr. S. Carrier, Mr. S. Lacasse, Mrs. J. Alain, Mrs. M. Poulin, Mrs. K. Bourgeault and Mr. W. Gagné Monfette for their technical contributions. The authors are also grateful to Mr. E. Leclair for its invaluable help in the preparation of the blends and to Oleotek for its collaboration.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Pascal Yves Vuillaume

References

Krishnan

, Pandey

, Mohanty

, Toughening of polylactic acid: an overview of research progress. Polym Plast Tech Eng. 2016;55(15):1623–1652. doi: 10.1080/03602559.2015.1098698

Nagarajan

, Mohanty

, Misra

Perspective on polylactic acid (PLA) based sustainable materials for durable applications: focus on toughness and heat resistance. ACS Sustain Chem Eng. 2016;4(6):2899–2916. doi: 10.1021/acssuschemeng.6b00321

Vachon

, Pépin

, Béland

, Thermal, mechanical and morphological properties of binary and ternary PLA blends containing a poly(ether ester) elastomer. J Biobased Mater Bio. 2015;9(2):205–217. doi: 10.1166/jbmb.2015.1507

Nagarajan

, Mohanty

, Misra

Blends of polylactic acid with thermoplastic copolyester elastomer: effect of functionalized terpolymer type on reactive toughening. Polym Eng Sci. 2017. doi 10.1002/pen.24566.

Zhang

, Nagarajan

, Misra

, Supertoughened renewable PLA reactive multiphase blends system: phase morphology and performance. ACS App Mater Interfaces. 2014;6(15):12436–12448. doi: 10.1021/am502337u

Vachon

, Pépin

, Balampanis

, Compatibilization of PLA/PEBA blends via reactive extrusion: a comparison of different coupling agents. J Polym Environ. 2017;25(3):812–827. doi: 10.1007/s10924-016-0860-x

Jauzein

, Huneault

, Heuzey

MC.

Crystallinity and mechanical properties of polylactide/ether–amide copolymer blends. J App Polym Sci. 2017;134(14):338. doi:10.1002/app.44677.

, Shimizu

Toughening of polylactide by melt blending with a biodegradable poly(ether)urethane elastomer. Macromol Biosci. 2007;7(7):921–928. doi: 10.1002/mabi.200700027

Zeng

, Li

, He

, Improving flexibility of poly(l-lactide) by blending with poly(l-lactic acid) based poly(ester–urethane): morphology, mechanical properties, and crystallization behaviors. Ind Eng Chem Res. 2011;50(10):6124–6131. doi: 10.1021/ie102422q

10.

Dogan

, Reyes

, Rastogi

, Reactive compatibilization of PLA/TPU blends with a diisocyanate. J App Polym Sci. 2014;131(10):40251–40260. doi: 10.1002/app.40251

11.

Oliaei

, Kaffashi

, Davoodi

Investigation of structure and mechanical properties of toughened poly(l-lactide)/thermoplastic poly(ester urethane) blends. J App Polym Sci. 2016;133(15):43104–43116. doi: 10.1002/app.43104

12.

Yuan

, Ruckenstein

Polyurethane toughened polylactide. Polym Bull. 1998;40(4–5):485–490. doi: 10.1007/s002890050280

13.

Liu

, He

, Zeng

, In situ formed crosslinked polyurethane toughened polylactide. Polym Chem. 2014;5(7):2530–2539. doi: 10.1039/c3py01649h

14.

Zhao

, Huang

, Zhang

SD.

Largely toughening biodegradable poly(lactic acid)/thermoplastic polyurethane blends by adding MDI. J App Polym Sci. 2015;132(48):42511–42519.

15.

Imre

, Bedo

, Domján

, Structure, properties and interfacial interactions in poly(lactic acid)/polyurethane blends prepared by reactive processing. Eur Polym J. 2013;49(10):3104–3113. doi: 10.1016/j.eurpolymj.2013.07.007

16.

, Wei

, Huang

, Supertoughened poly(lactic acid)/polyurethane blend material by in situ reactive interfacial compatibilization via dynamic vulcanization. Ind Eng Chem Res. 2014;53(44):17386–17393. doi: 10.1021/ie503092w

17.

, Loi

, Delgado

, Reactive compatibilization of polylactide/polypropylene blends. Ind Eng Chem Res. 2015;54(23):6108–6114. doi: 10.1021/acs.iecr.5b00882

18.

Janorkar

, Fritz

Jr , Burg

KJL

, Grafting amine-terminated branched architectures from poly(l-lactide) film surfaces for improved cell attachment. J Biomed Materials Res B. 2007;81B(1):142–152. doi: 10.1002/jbm.b.30647

19.

Al-Itry

, Lamnawar

, Maazouz

Reactive extrusion of PLA, PBAT with a multi-functional epoxide: physico-chemical and rheological properties. Eur Polym J. 2014;58:90–102. doi: 10.1016/j.eurpolymj.2014.06.013

20.

Yang

, Hirata

, Hsu

, Improved thermal and mechanical properties of poly(butylene succinate) by polymer blending with a thermotropic liquid crystalline polyester. J App Polym Sci. 2014;131(4):39952–39959.

21.

Dong

, Zou

, Yan

, Effect of chain-extenders on the properties and hydrolytic degradation behavior of the poly(lactide)/ poly(butylene adipate-co-terephthalate) blends. Int J Mol Sci. 2013;14(10):20189–20203. doi: 10.3390/ijms141020189

22.

, Sun

XS.

Thermal and mechanical properties of poly(lactic acid)/starch/methylenediphenyl diisocyanate blending with triethyl citrate. J App Polym Sci. 2003;88(13):2947–2955. doi: 10.1002/app.12112

23.

Kim

, Hong

, Lee

, Chain extension effects of para-phenylene diisocyanate on crystallization behavior and biodegradability of poly(lactic acid)/poly(butylene terephthalate) blends. Adv Compos Mater. 2010;19(4):331–348. doi: 10.1163/092430409X12605406698471

24.

Shen

, Lu

, Liang

Modification of the properties of polylactide/polycaprolactone blends by incorporation of blocked polyisocyanates. J Macromol Sci Pure App Chem. 2013;50(5):547–554. doi: 10.1080/10601325.2013.781463

25.

Shen

, Lu

, Zhou

, Effect of reactive blocked polyisocyanate on the properties of solvent cast blends from poly(lactic acid) and poly(ethylene glycol). J App Polym Sci. 2012;125(3):2071–2077. doi: 10.1002/app.36276

26.

Vuillaume

, Zang

, Leclair

, Creating super-tough, bio-based polylactic acid materials. Plast Eng. 2017;73(2):24–27.

27.

Yang

, Nie

, Zhu

A comparative study on different rubbery modifiers: effect on morphologies, mechanical, and thermal properties of PLA blends. J App Polym Sci. 2016;133(17). 10.1002/app.43340.

28.

Sheth

, Xu

, Wilkes

GL.

Solid state structure-property behavior of semicrystalline poly(ether-block-amide) PEBAX® thermoplastic elastomers. Polymer. 2003;44(3):743–756. doi: 10.1016/S0032-3861(02)00798-X

29.

Fisher

, Sterzel

, Wegner

Kolloid Z Z Polym. 1973;251:981–990.

30.

Hong

, Wei

, Yuan

, A novel composite coupled hardness with flexibleness-polylactic acid toughen with thermoplastic polyurethane. J App Polym Sci. 2011;121(2):855–861. doi: 10.1002/app.33675

31.

Feng

, Ye

Morphologies and mechanical properties of polylactide/thermoplastic polyurethane elastomer blends. J App Polym Sci. 2011;119(5):2778–2783. doi: 10.1002/app.32863

32.

Martin

, Avérous

Poly(lactic acid): plasticization and properties of biodegradable multiphase systems. Polymer (Guildf). 2001;42(14):6209–6219. doi: 10.1016/S0032-3861(01)00086-6

33.

Jacobsen

, Fritz

HG.

Plasticizing polylactide – the effect of different plasticizers on the mechanical properties. Polym Eng Sci. 1999;39(7):1303–1310. doi: 10.1002/pen.11517

34.

Chieng

, Ibrahim

, Yunus

WMZW

, Plasticized poly(lactic acid) with low molecular weight poly(ethylene glycol): mechanical, thermal, and morphology properties. J App Polym Sci. 2013;130(6):4576–4580.

35.

Baiardo

, Frisoni

, Scandola

, Thermal and mechanical properties of plasticized poly(l-lactic acid). J App Polym Sci. 2003;90(7):1731–1738. doi: 10.1002/app.12549

36.

Piorkowska

, Kulinski

, Galeski

, Plasticization of semicrystalline poly(l-lactide) with poly(propylene glycol). Polymer (Guildf). 2006;47(20):7178–7188. doi: 10.1016/j.polymer.2006.03.115

37.

Andrusyszyn

, Sunkara

HB.

Methods for improving physical properties of polyesters. United States Patent US 0105665 A1. 2011.

38.

Jun

CL.

Reactive blending of biodegradable polymers: PLA and starch. J Polym Env. 2000;8(1):33–37. doi: 10.1023/A:1010172112118

39.

, Liu

, Zhao

, Effects of talc on the mechanical and thermal properties of polylactide. J App Polym Sci. 2012;125(suppl. 2):E99–E109. doi: 10.1002/app.36260

40.

Han

, Han

, Dong

Morphology and properties of the biosourced poly(lactic acid)/poly(ethylene oxide-b-amide-12) blends. Polym Compos. 2013;34(1):122–130. doi: 10.1002/pc.22383