Novel super-toughened bio-based blend from polycarbonate and poly(lactic acid) for durable applications

Abstract

High-performance bio-based polycarbonate (PC) and poly(lactic acid) (PLA) blends are created using poly(ethylene-n-butylene-acrylate-co-glycidyl methacrylate) (EBA-GMA) as a reactive compatibilizer and impact modifier, with the aid of an epoxy-based chain extender (CE). It is found that the use of a specific acrylic impact modifier and a high blending temperature are the major factors in reaching high toughness and superior heat resistance. Toughened blends containing 32 wt% of biosourced PLA showed impact strength and heat resistance comparable to that of neat PC, while exceeding the mechanical properties of PC. Atomic force microscopy (AFM) studies have shown that an EBA-GMA phase was concentrated exclusively within a brittle PLA phase, ensuring high-efficiency impact-strength modification with a relatively small amount of modifier. As little as 6 wt% of EBA-GMA in a PC/PLA blend was enough to increase the toughness by more than 10 times. It is found that PLA and PC phases form a co-continuous morphology in the blend, allowing it to reach a high heat resistance that is almost unaffected by an impact modifier content of up to 10 wt%. It is shown that targeted toughening of the brittle PLA phase allows the achievement of excellent mechanical properties due to a lower required acrylic rubber content. A significant bio-based content in the PC/PLA blend allows increased sustainability of manufacturing and opens possibilities for use in a wide range of industrial applications.

---

1. Introduction

A growing demand for polymers manufactured from renewable sources induced significant interest in biopolymers having good mechanical properties for demanding applications, such as automotive and electronics industry use. One of the most promising polymers for these kinds of applications is poly(lactic acid) (PLA). PLA is a biodegradable plant-sourced polymer with good mechanical properties, which makes it suitable for a wide range of consumer and industrial applications.^1–3 Currently, PLA application is limited by its low toughness and mediocre heat resistance, due to its low glass transition temperature in the range 55–60 °C.^4–6 Polymer blending is an efficient and versatile technique to create new materials with a necessary combination of properties.⁷ Blending polycarbonate (PC) with PLA is appealing because of the high heat resistance and toughness of PC.⁸ The Fujitsu and Toray companies have developed a PC/PLA blend with 50 wt% PLA, which contains ignition-resistant agents intended for use in notebook computer cases.⁹ The preparation of PC/PLA blends having more than 50 wt% PC requires an elevated processing temperature, typically exceeding 260–270 °C. In this range of temperatures, the thermal decomposition rate of PLA is significant, making the preparation of these types of blends challenging.

Non-compatibilized PC/PLA blends are intrinsically brittle and have low elasticity and a low heat deflection temperature in many cases. Thus, PC/PLA blending aims to toughen the blends using various compatibilizers and impact modifiers. Typically, a blend's elasticity, as well as its heat resistance, mostly depends on the PLA content. A number of compatibilizers were tried in PC/PLA blends. Lee et al.¹⁰ used three types of compatibilizers: poly(styrene-g-acrylonitrile)-maleic anhydride (SAN-g-MAH), poly(ethylene-co-octene) rubber-maleic anhydride (EOR-MAH), and poly(ethylene-co-glycidyl methacrylate) (EGMA). It was found that SAN-g-MAH shows significantly better efficiency in most aspects of its mechanical properties. A 70 wt% PC/30 wt% PLA blend showed an impact strength of ∼360 J m⁻¹ with the addition of 5 phr of SAN-g-MAH and ∼220 J m⁻¹ with no additives. Researchers related this improvement to a smaller droplet size of PLA as compared to the effect of other additives. It was shown that the addition of 5 phr SAN-g-MAH resulted in a droplet-size reduction from 1.42 μm to 0.19 μm.

Another approach to PC/PLA blending is introducing another polymeric component, i.e. via ternary blends. This approach was used by Kanzawa and Tokumitsu,¹¹ who introduced poly(butylene adipate-co-terephthalate) (PBAT), along with a dicumyl peroxide (DCP) radical initiator, to a PC/PLA blend. Despite a certain improvement in the blend miscibility, the mechanical properties of the blend remained poor, which was attributed to the lack of a toughening effect from PBAT. Wang et al.¹² investigated poly(butylene succinate-co-lactate) (PBSL) and epoxy (EP) as compatibilizers for a PLA/PC binary blend. The researchers did not find a noticeable improvement in the mechanical and thermal properties for up to 20% compatibilizer content. Hashima et al.¹³ toughed PLA using hydrogenated styrene–butadiene–styrene block copolymer (SEBS), which was premixed with poly(ethylene-co-glycidyl methacrylate) (EGMA). In this system, EGMA was a reactive compatibilizer. The heat resistance and resistance against thermal aging were further improved by incorporating polycarbonate (PC). The resulting good elongational properties of the blend were attributed to the effect of the negative pressure of the dilating SEBS phase in the PLA/PC matrix, which, in turn, enhanced the local segment motions.

Wang et al.¹⁴ noted that the molecular weight of PC might affect the compatibility of PC/PLA/PBSL ternary blends. Blends containing low-molecular-weight PC blends showed lower impact strength, though the observed effect was small, and these blends showed a significantly higher heat deflection temperature (HDT) as compared to a blend composed from high-molecular-weight PC. The improvement in the HDT was significant: from 61 °C to 114 °C for the 50% PLA/50% PC blend. This observation was related to the lower viscosity of low-M_w PC at the processing temperatures, as compared to high-M_w PC, which allowed the blend to form a more intricate structure, having a finer phase dispersion.

Significant efforts were made to enhance the interface between PLA and PC in blends. One of the approaches was transesterification of PC and PLA. Liu et al.¹⁵ studied the effect of a catalyst on transesterification reactions between poly(lactic acid) (PLA) and polycarbonate (PC) under intense shear. Three catalysts (zinc borate, titanium pigment, and tetrabutyl titanate) were studied in PC/PLA blends. It was found that, while the transesterification reaction occurs in the absence of catalyst, all three catalysts significantly promote the transesterification reaction between poly(lactic acid) and polycarbonate. Phuong et al.¹⁶ added tetrabutylammonium tetraphenylborate (TBATPB) and triacetin during extrusion to melt blends of polylactic acid (PLA) and polycarbonate (PC) through a reactive compatibilization approach, in order to improve the materials' mechanical properties and heat resistance. The compatibility between PC and PLA in this study was improved through the formation of a PLA–PC copolymer. It was found that catalytic compatibilization noticeably improves the elastic modulus of the blend, while its elongational properties remain mostly unchanged. The latter was related to the fact that, despite the interaction between PLA and PC being improved, the sizes of the dispersed domains increased. Wang et al.¹⁷ attempted to enhance the properties of PC/PLA blends using functionalized multi-walled carbon nanotubes (FMWCNTs). Researchers found that carbon nanotubes concentrated exclusively in the PC phase. Even 0.5% FMWCNT noticeably improved the mechanical properties, including the impact strength. The impact strength more than doubled (from 5.0 kJ m⁻² to 11.0 kJ m⁻²), and the elongation at break increased from ∼20% to ∼90%. Further increasing the FMWCNT content had a reverse effect on these properties.

Srithep et al.¹⁸ used an epoxy-based chain extender to control the molecular weight of PLA during reactive blending of PLA and PC. They concluded that the chain extender acted both as a chain extender and as a compatibilizer. Moreover, the chain extender also promotes crystallization behaviour of PLA. The blend's HDT was also improved; for example, the HDT of a 50 wt% PLA/50 wt% PC blend improved from 62 °C to 106 °C after mixing with 1.1 wt% of chain extender. This behavior was related to PLA crystallization promoted by the chain extender, PLA. The positive effect of the epoxy-based chain extender on the crystallization behavior of PLA was also shown by Meng et al.¹⁹ and Najafi et al.²⁰

PLA toughening using various impact modifiers and additives was intensively studied in recent years. Researchers found that PLA toughness can be modified most efficiently by ethylene-acrylic-based rubbers. Numerous studies of binary and ternary systems of PLA and modifiers, such as ethylene methyl acrylate-glycidyl methacrylate (EMA-GMA) and poly(ether-b-amide) elastomeric copolymer (PEBA),²¹ poly(ethylene-glycidyl methacrylate) (EGMA),²² and PEG derivatives,²³ were attempted. Nevertheless, the PLA/EBA-GMA binary blend has not been thoroughly studied. Direct blending of EBA-GMA and PLA resulted in relatively modest improvement in PLA toughness.^24,25 Liu et al.²⁵ attempted PLA/EBA-GMA system blending at two different temperatures and found no significant improvement in toughness. It was found that the overall improvement in impact strength was negligibly small when the processing temperature was changed from 185 °C to 240 °C. The observed improvement in toughness was moderate, which was related to PLA degradation at elevated temperatures. It was concluded that reactive compatibilization at the PLA/EBA-GMA interface might be successful when a catalyst and/or an additive that lowers the activation energy of the reaction is present.

It is possible, nevertheless, to achieve a high rate of reactive blending without employing a catalyst, with the help of an elevated temperature alone. According to Arrhenius' law, the PLA thermal degradation rate obeys an exponential law; therefore, a means to mitigate PLA degradation must be considered. One of the ways to achieve this is the use of a so-called chain extender, having a specific reactivity regarding PLA to maintain its molecular weight, which helps it to withstand thermal degradation. It was shown in several studies that a Joncryl series of PLA chain extenders is highly efficient over a wide range of temperatures.^20,26,27 These studies show that, not only can Joncryl prevent the thermal degradation of PLA, but it can also increase the molecular weight of PLA more than comparable chain extenders. Also Joncryl contains active epoxy groups necessary for interphase compatibilization (see Fig. 1). Viscosity is another factor that favorably affects toughness in a PLA/EBA-GMA binary blend processed at elevated temperatures, due to smaller EBA-GMA droplets being achieved during the blending of a low-viscosity melt.

Fig. 1 The chemical structures of chain extender Joncryl ADR-4368C and poly(ethylene n-butylene acrylate glycidyl methacrylate) acrylic impact modifier Elvaloy PTW (EBA-GMA).

It can be deducted that, for significant improvement in the performance of PC/PLA blends, several conditions must be met: firstly, good compatibility between phases, and secondly, a suitable impact modifier specific to the brittle phase, namely, PLA. The most efficient toughness modification of PLA was achieved using acrylic core/shell modifiers. This approach was used for the successful creation of a high-performance PC/PLA blend. The novelty of the proposed study is a targeted reinforcement, where an impact modifier was distributed exclusively in the brittle PLA phase, along with high-temperature processing. The proposed approach allowed a significantly decreased amount of the required modifier, while maintaining excellent heat resistance and mechanical properties.

2. Experimental section

Materials

Low-viscosity general purpose polycarbonate was used in this study. PC Hylex P1025L1HB (MFI = 25 g/10 min at 300 °C/1.2 kg according to manufacturer) was produced and supplied by Entec Polymers. An injection-grade PLA Ingeo 3251D, having a weight-average molecular weight (M_w) of 55 000 and a polydispersity index (PI) of 1.62 (ref. 22) was purchased from Nature Works LLC (USA). The mechanical properties and heat resistance data of neat materials used in this study are given in Table 1.

Table 1 Mechanical properties of neat materials (processed in DSM microcompounder)

Material	Tensile strength, MPa	Tensile modulus, MPa	Flexural strength, MPa	Flexural modulus, MPa	Elongation at break, %	Izod notched impact strength, J m^−1	HDT, °C
PLA Ingeo 3251D (processed at 180 °C)	63.7 ± 2.5	3650 ± 96	117.8 ± 1.7	3786 ± 45	2.8 ± 0.2	18.3 ± 2.5	56.4 ± 1.1
PC Hylex P1025L1HB (low-viscosity general purpose) (processed at 270 °C)	59.6 ± 0.96	2310 ± 43	91.8 ± 1.3	2458 ± 29	63.4 ± 6.8	748 ± 31	139.4 ± 1.8

---

Poly(ethylene n-butylene acrylate glycidyl methacrylate) (EBA-GMA) ethylene terpolymer was supplied by DuPont company under the trade name Elvaloy PTW. Elvaloy PTW is a terpolymer. Its chain consists of 66.75 wt% ethylene, 28 wt% butyl acrylate, and 5.25 wt% glycidyl methacrylate.²⁸ The melt flow index of Elvaloy PTW of 12 g/10 min (190 °C/2.16 kg) was measured according to ASTM method D1238; the melting point of 72 °C and glass transition temperature (T_g) of −55 °C were determined using DSC.

The commercially available chain extender, Joncryl ADR-4368C, was supplied by BASF. Joncryl ADR-4368C is in the form of glassy brittle transparent pellets, and consists of a multi-functional reactive polymer having an epoxy equivalent weight of 285 g mol⁻¹ and a T_g of 54 °C, according to the manufacturer's data. Joncryl ADR-4368C is typically used as an additive for thermal degradation mitigation in a number of polymers, including PLA and PC. The chemical structure of the EBA-GMA terpolymer, Joncryl ADR-4368C, has significant similarities, but these terpolymers have a different role in polymer blending. Joncryl ADR-4368C exhibits partial miscibility with a wide range of polyesters, including PLA, and contains highly reactive epoxy groups capable of reacting with the carboxylic end groups of PLA chains, resulting in molecular weight increase. At the same time, EBA-GMA is immiscible in PLA and forms a separate phase with relatively poor adhesion under normal blending conditions. The structures of both additives are shown in Fig. 1.

Blend and sample preparation

PLA and PC pellets were dried for 12 hours at 80 °C to remove the moisture and then dry mixed with EBA-GMA pellets and powdered chain extender. Blend compounding and subsequent injection molding of the blends was done with a DSM Xplore 15 cm³ microcompounder (Netherlands), using a pneumatic feeder and a 10 cm³ injection molding machine. DSM Xplore 15 is a lab-scale co-rotating twin-screw extruder. It has separate temperature control in three barrel zones and, in all experiments, the temperature in all zones was kept the same. The screw speed ranged from 100 rpm to 150 rpm, and the mixing time was 2 minutes. The experiments where a 150 rpm screw speed was used are specified in the figure captions. The mold temperature was 40 °C, the injection pressure was 5 bars, and the holding pressure was 8 bars, with a holding time of 22 seconds. In all blends in this study, 0.3 phr of chain extender was used. During the preliminary experiments, it was found that low chain-extender loadings do not prevent degradation, and amounts exceeding 0.5 phr result in lower blend toughness, presumably due to the development of a higher degree of crosslinking in the PLA phase. The blending temperature was kept at 270 °C, except for experiments where the effect of temperature on toughness was investigated.

Some of the test samples were prepared using a Leistritz (Germany) pilot-scale extruder with a screw speed of 100 rpm, followed by extrudate pelletizing. The extruder was equipped with co-rotating twin screws with 4 kneading zones, with a screw diameter of 27 mm and an L/D ratio of 48. Prior to the injection molding, the extrudate was pelletized and dried in an oven at 80 °C for 24 h. The dried extruded pellets were injection molded in a pilot-scale Arburg Allrounder 370C (Germany) injection molding machine to obtain test specimens. The Arburg Allrounder 370C injection molding machine had a maximum injection pressure of 1500 bar and a screw diameter of 35 mm. The samples were conditioned prior to testing for 48 hours at room temperature and 50% relative humidity.

Testing and characterization

Mechanical properties testing. An Instron Universal Testing Machine (Model 3382) was used to perform tensile (ASTM standard D 638) and flexural tests (ASTM standard D 790). Crosshead speeds of 14 mm min⁻¹ for the flexural test and 50 mm min⁻¹ for the tensile test were used, as recommended by the respective standards. For each test, at least 5 samples were used.

A Testing Machine Inc. (TMI) Instrument equipped with a 5 lbs per ft Izod impact hammer was used for notched Izod impact strength measurements, according to ASTM standard D256. Each test set consisted of at least 6 samples.

Differential scanning calorimetry (DSC). DSC analysis was performed in a TA Instruments Q200 setup, and heating and cooling rates of 10 °C min⁻¹ were used in all tests. The samples were heated under a nitrogen flow of 50 ml min⁻¹. The melting enthalpy was calculated by measuring the area under the corresponding DSC curves, using TA Universal Analysis software.

Heat deflection measurement. The heat deflection temperature was determined using dynamic mechanical analysis (DMA) equipment Q800 from TA Instruments, using a 3-point bending test with a 0.455 MPa load and a heating rate of 2 °C min⁻¹, according to ASTM D648.

Thermogravimetric analysis (TGA). TGA was performed under nitrogen, using a thermogravimetric analyzer (TGA Q500, TA Instruments, Inc.). The samples were heated from room temperature at a heating rate of 10 °C min⁻¹. The derivatives of the TGA curves (DTG) were also obtained using TA analysis software.

Atomic force microscopy (AFM). An AFM instrument (Multimode 8) from Bruker Nano Inc., CA, U.S.A., which was equipped with a Nanoscope V controller and Nanoscope Software, version 8.15, was used for the AFM measurements. Image processing and subsequent data analysis were performed using Nanoscope Analysis software. Imaging in PeakForce (PF) Mode was done using RTESPA and TAP525A silicon cantilevers (Bruker AFM probes), with spring constants of 40 N m⁻¹ (RTESPA) and 200 N m⁻¹ (TAP525A) in air. PeakForce Quantitative Nanomechanical Mapping (QNM) is a relatively new atomic force microscopy technique for measuring the Young's modulus of materials with nanoscale resolutions. In PeakForce QNM, the Young's modulus is calculated using a DMT (Derjaguin, Muller, Toropov) model from a known tip indentation depth, the tip's geometry, and the force applied to the tip. For precise DMT modulus measurements, cantilevers were calibrated on a polystyrene standard sample (∼2.7 GPa DMT modulus). PeakForce Quantitative Nanomechanical Property Mapping (PF-QNM) AFM was done at constant oscillation of the sample at 2 kHz using an amplitude in the range 100–250 nm. The AFM analysis requires a smooth area for scanning, with outstanding features not exceeding 1–2 μm. Therefore, the specimens for AFM imaging were prepared by microtoming at room temperature with a sapphire knife, in order to create a smooth surface, using a Leica UC7 Microtome (Germany).

3. Results and discussion

The mechanical properties of PLA/PC composites, such as tensile and flexural strength, are mostly determined by the composition of the blend and the compatibilizer content. In this regard, PC/PLA blends generally follow the mixture rule, and the blend's stiffness increases with increasing PLA content. Therefore, most efforts in the development of PC/PLA blends are aimed at improving the impact strength, elongational properties, and heat resistance. All these critical properties are related to the PLA phase, and this is the reason why PLA phase toughening attracts the most attention.

PC/PLA blend toughening

Effect of processing temperature on PC/PLA blend toughness. The toughness of the PC/PLA blends demonstrated significant dependence on the blending temperature. It should be noted that the effect of the temperature factor on compatibilization in PC/PLA blends has not been studied at all to date, and many aspects of this phenomenon remain unclear. While the overall toughness increased with increasing temperature and decreased after reaching a maximum, the highest toughness was achieved when the chain extender was used at 270 °C (Fig. 2). This temperature is extremely high for conventional PLA processing (180–220 °C), and little is known regarding PLA behaviour at such elevated temperatures. Clearly, the processing was possible only in the presence of a chain extender that prevented PLA from rapid degradation during processing.^27,28 The impact strength of the blends was determined to a significant degree by interface adhesion between the matrix and the impact modifier. From Fig. 2, it can be concluded that the compatibilization reaction rate continuously increases up to 270 °C, after which the degradation rate of the chain extender becomes very high, limiting further improvement. From the TGA analysis of the chain extender (Fig. 3), it can be seen that, even at 250 °C, the chain extender sample experienced a weight loss of 6.9%. Considering that weight loss due to vaporization to low-molecular-weight compounds is preceded by high-molecular-weight substance decomposition, the real degree of chain extender degradation might be even higher. At 270 °C, where the top performance for the PC/PLA blend was achieved, the chain extender weight loss was 14.8% (Fig. 3). It may be assumed that the reactivity of the epoxy groups in the chain extender increases at a higher rate than its decomposition due to high temperature. This assumption is in agreement with observations by Ojijo and Ray,²⁹ who noticed that a time significantly exceeding 5 minutes at typical processing temperatures of 180–200 °C is required for Joncryl to start the reaction with PLA. Their experiments also showed the existence of a pronounced induction time in the PLA/chain extender system, during which no reaction is happening. Considering this, it might be advantageous to increase the processing temperature significantly in order to accelerate the reaction, despite concerns regarding chain extender and PLA degradation, in order to decrease the induction time and increase the compatibilization reaction rate. It should be noted, nevertheless, that there are few studies on the effects of the temperature of blending on the PC/PLA blend performance. In this regard, this research provides a valuable insight into PC/PLA blending and toughness modification.

Fig. 2 Notched Izod impact strength of PC/PLA blends containing 32 wt% of PLA and 6% EBA-GMA, blended at different temperatures. DSM injection molding at 150 rpm screw speed.

Fig. 3 TGA curve of Joncryl ADR-4368C chain extender at 10 °C min⁻¹ scan rate.

Effect of EBA-GMA content on PC/PLA blends mechanical properties. In considering the toughening of PC/PLA blends, most attention is paid to the brittle PLA phase. Finding an efficient impact modifier for PLA is one of the most important factors in the overall blend performance. Impact modification by rubbery elastomers is typically characterized by a sharp increase in impact strength, called the brittle-ductile (BD) transition, when a certain content of impact modifier is added.³⁰ The PLA BD transition can be observed at as low as 15 wt% of impact modifier, or as high as 40 wt%.³¹ Eventually, the impact modifier content needed for successful toughening will be determined not only by interface bonding, but also the particle size, its distribution, and the average interparticular distance.³² Considering this, EBA-GMA is one of the most efficient high temperature impact modifiers for PLA, allowing a high impact strength to be reached at a relatively low content (Fig. 4).

Fig. 4 Notched Izod impact strength of PC/PLA blends containing 32 wt% of PLA at different impact modifier contents and PLA toughened with the same impact modifier; x represents the amount of EBA-GMA in the blend. All blends were blended at 270 °C.

It was found that, unlike the PLA/EBA-GMA blend, the ternary PC/PLA/EBA-GMA blend requires much smaller amounts of EBA-GMA to reach its optimum strength (Fig. 4). Overall, the blend's behaviour was typical for brittle polymers toughened by rubber modifiers, but it required three times less impact modifier loading. It can be seen that a typical brittle-ductile (BD) transition occurs at 4–6 wt% impact modifier content in ternary PC/PLA/EBA-GMA blends, unlike the typical 15 wt% required for PLA-based binary blends. This BD transition onset was characterized by a wide range of observed sample toughness values (wide error bars) (Fig. 4). Since, for efficient impact modification, certain amounts of impact modifier are necessary, it can be assumed that in the case of the ternary blend, the impact modifier was predominantly distributed in the brittle PLA phase; thus, a significantly lower content was required to reach a high toughness. For the blend under investigation, the 4–6% BD transition range corresponds to 11.1–15.7 wt% relative to the PLA phase, which is even lower than for the binary PLA/EBA-GMA blend. The higher efficiency of EBA-GMA in PC/PLA blends can be explained by PLA phase containment within the PC phase, which increases the effective concentration of EBA-GMA in the co-continuous PLA phase. The PC/PLA interface, having good interface bonding, can be an efficient crack propagation limiter, increasing the efficiency of the impact modifier. This observation will be complemented by a morphology study later. It should be noted that the chain extender itself has little to no effect on the impact strength of the blends. The impact strength of non-impact-modified blends remained low: 40.5 J m⁻¹ without a chain extender and 43.7 J m⁻¹ in the presence of 0.3 phr of chain extender.

In order to maintain comparability of the experiments, the pilot-scale processing temperature and the average residence times were the same as in lab-scale experiments. The impact strength of the blends increased with increasing shear rate in lab-scale experiments and further increased in pilot-scale extrusion, where an even higher shear rate was applied to the melt. The investigated ternary blend showed noticeable dependence of impact strength on shearing conditions during processing. In the lab-scale DSM setup, the Izod notched impact strength increased from 540 J m⁻¹ at 100 RPM blending to 683 J m⁻¹ at 150 RPM blending at 270 °C for the 32 wt% PLA/62 wt% PC/6 wt% EBA-GMA blend. The impact strength of the same blend further increased to 715 J m⁻¹ when it was processed in a pilot-scale Leistritz twin-screw extruder setup. Eventually, this was explained by the finer dispersion of the acrylic terpolymer in PLA due to a higher shear.

Overall, the mechanical properties of toughened PC/PLA blends follow a linear decreasing trend with an increasing impact modifier loading (Fig. 5). The tensile and flexural strengths of the blends were more affected by an increasing content of EBA-GMA; they decreased by about 20% as the impact modifier content increased from 0% to 10%. The tensile and flexural moduli decreased by 15% with the same increase in the EBA-GMA content. The blends maintained high elasticity with an elongation at break of over 70% over the whole range of impact modifier content. Interestingly, the highest elongation at break was achieved at 6% of the EBA-GMA content, which was the onset of the BD transition.

Fig. 5 Mechanical properties of toughened 68% PC/32 wt% PLA blends having various EBA-GMA contents replacing PC: (a) tensile properties; (b) flexural properties; (c) elongation at break. All samples were prepared at 270 °C.

The crystallization and melting behaviour of toughened PC/PLA blends. PC is a very slowly crystallizing polymer, having very low spherulitic growth rates, even when facilitated by the presence of a solvent.³³ Therefore, PC can be considered as an amorphous polymer, in terms of processing and thermal testing times. During the relatively fast cooling in the mold during blend processing, PLA does not crystallize and remains amorphous (Fig. 6(b)). In our studies, PC did not crystallize in this blend during the cooling run under testing conditions (Fig. 6(b)). PLA, as a slow crystallizing polymer, does not show crystallization during cooling from the melt. Also, there was little to no effect of the impact modifier on PLA crystallization in the subsequent heating run (Fig. 6(c)). Since processing of these blends was done at a very high temperature, it is interesting to assess its effect on the PLA thermal degradation. Despite this, the melting point depression of PLA, and thus its degradation, is relatively small after processing (Fig. 6(a)). It increases noticeably during the DSC scan and, during the third run, not only the melting point of PLA was lower, but also a second melting peak appeared, which means that the prevalent degradation mechanism in this case was PLA intramolecular transesterification, leading to cyclic oligomers of lactic acid and lactide.³⁴ A similar splitting of the PLA melting peak due to thermal degradation was observed by Le Marec et al.³⁵ Generally, this means that, while the chain extender is very efficient, in this blend, the residence time at high temperatures is limited and property deterioration may be expected with prolonged exposure to high temperatures. As mentioned before (Fig. 3), this observation can be explained by the low thermal stability of the chain extender.

Fig. 6 DSC curves for PLA, non-modified, and impact-modified PC/32% PLA blends: first heating run (a); first cooling run (b); second heating run (c). Curves are shifted for viewing convenience.

It can also be noted that the maximum relative crystallinity achieved by PLA in both modified and non-modified blends is lower than that in neat PLA. Considering that the enthalpy of crystallization of 100% crystalline PLA is equal to 93.7 J g⁻¹,³⁶ the crystallinity of PLA in PC/PLA blends was ∼39%, while for unblended PLA, it was ∼43%. This can be explained by steric difficulties for PLA crystallization in blends. In other words, the presence of domains of foreign phases in the blend changes the pattern of crystallization from unlimited spherulitic growth to limited growth confined by the PC phase.³⁷ This effect is relatively small but still noticeable. Further depression of the crystallization rate might be expected if the PLA phase had a smaller characteristic size.

A comparison of differential TGA weight loss curves for neat materials and PC/PLA blends is given in Fig. 7. The addition of EBA-GMA significantly slows the rate of decomposition of the blend as compared to the non-modified blend. This might be explained by the ablation effect of EBA-GMA, which protects the PLA phase and slows down the PLA decomposition during TGA analysis.

Fig. 7 Differential TGA curves for PLA, PC, non-modified, and impact-modified PC/32% PLA blends.

The heat resistance of PC/PLA/EBA-GMA blends. At 32 wt% PLA content, the heat deflection temperature of the blend demonstrated little dependence on modifier content and remained over 130 °C (0.455 MPa stress load) over the whole range of impact modifier loading from 0% to 10%. Only for EBA-GMA contents over 7 wt%, the blend demonstrated a weak tendency for a decrease in the HDT (Fig. 8). The reason for this remarkable behaviour is containment of EBA-GMA within the PLA phase, which does not affect the PC phase continuity with increasing EBA-GMA content. This observation suggests that the PC content is a major factor with respect to the heat resistance of the PC/PLA blends. Eventually, a continuous PC phase forms a three-dimensional structure, which defines its heat resistance. While the PLA phase loses its heat resistance at low temperatures below 60 °C, the unaffected PC phase remains intact and continues to withstand increasing temperatures during testing. Clearly, the amount of PC in the blend is critical from this point of view, and as the PC content decreases in this structure, an increasing number of weak elements start to appear, leading to an accelerating decrease in heat resistance (Fig. 9). For the investigated ternary PC/PLA/EBA-GMA blend, 32 wt% of PLA appears to be the critical limit, after which heat resistance starts to decrease in an accelerating manner. As a result of the PC three-dimensional structure losing its integrity, the heat deflection temperature drops from 125.3 °C at 36 wt% PLA content to 86.6 °C at just 40 wt% PLA content. It is interesting to note that an increasing amount of EBA-GMA in tested blends has less effect on the heat deflection temperature, unlike an increasing amount of PLA, although the increase in EBA-GMA was accompanied by a decreasing PC content (Fig. 8 and 9). For example, the 32 wt% PLA blend represented in Fig. 8 at 10 wt% of EBA-GMA content has 58 wt% of PC, which is identical to the PC content in the blend of 35 wt% PLA with 7 wt% of EBA-GMA, represented in Fig. 9; nevertheless, the first blend has a noticeably higher HDT of 128.6 °C vs. 125.3 °C. This can also be explained by the assumption that a noticeable amount of EBA-GMA is located in the PC/PLA interphase and is therefore less disruptive to the structure of the PC phase. This also supported by TGA analysis, where noticeable ablation was observed.

Fig. 8 The effect of impact modifier loading on heat deflection temperature of PC/32 wt% PLA blends. All samples were prepared at 270 °C.

Fig. 9 The effect of PLA content on the heat deflection temperature of PC/PLA/EBA-GMA blends. The amount of EBA-GMA was 20% of the amount of PLA in all blends. All samples were prepared at 270 °C with 0.3 phr of chain extender.

Morphology of PC/PLA blends

While a number of studies attempted to achieve tough PC/PLA blends, the toughness modification of PC/PLA blends usually involved a significant impact-modifier loading, typically exceeding 15% by weight. In this study, a dramatic improvement was achieved at a significantly lower impact modifier content in the range 5–10 wt%. The key to this was the use of selective phase modification. Indeed, since PC has a high impact strength, it does not need impact modification in this case, while the intrinsically brittle PLA phase does require modification. Therefore, successful toughening of the PC/PLA blend includes selective modification of the PLA phase, which was achieved in this study. For better understanding of the investigated PC/PLA blends, an AFM study was attempted (Fig. 10).

Fig. 10 Atomic force microscopy images of (a) impact-modified 32 wt% PLA/62 wt% PC/6 wt% EBA-GMA blend (b) non-modified 32 wt% PLA/68 wt% PC blend. All samples were prepared at 270 °C; 30 μm DMT modulus images are shown.

PeakForce AFM is a novel technique for surface imaging, which includes actual nano-indentation of the investigated surface and continuous measuring of indentation forces and surface deformation. From the known geometry of the indenter (tip), it is possible to evaluate the local properties of the material, such as DMT modulus, adhesion, and dissipation energy. Images of impact-modified and non-modified blends are shown in Fig. 10. In DMT modulus images, the modulus of the material is represented via a color-coded scale. Areas of higher modulus are shown in a lighter color, while a darker color represents areas with a lower modulus. The image of an impact-modified blend shows that the EBA-GMA phase is distributed exclusively in a continuous PLA phase, which explains its high efficiency at relatively small loads (Fig. 10(a)). The distribution is not uniform and the EBA-GMA droplet distribution is wide, which suggests a significant potential for further improvement of the toughness of the PC/PLA blend. The dark spots observed in Fig. 10(b) are areas where the material was chipped out during microtoming, which resulted in dark spots during the AFM imaging. The sensitivity of the blend's toughness to shear was observed during a comparison of the toughness of blends prepared in different experimental setups. Lab-scale production, where a lower shear was applied, resulted in the blend having a notched Izod impact strength of 540 J m⁻¹, while a pilot-scale setup, which applies a higher shear during blending, yielded a blend having an impact strength of 715 J m⁻¹. It should be noted that the observations of brittle-ductile transition behaviour for typical brittle polymers toughened by an acrylic rubber phase remain valid in this case. The actual impact modifier load ratio to the brittle phase is more than 3 times higher in the PC/PLA blend and ranges from 13.5 to 23.8 wt%, when the EBA-GMA load varies from 5 to 10 wt%.

It can be also noted that, even when PLA forms a droplet in the continuous PC phase, EBA-GMA tends to stay inside that droplet. This suggests high surface tension in the PC/EBA-GMA interphase, making distribution of EBA-GMA in the PLA phase more favourable. One of the reasons for this observation is the relative viscosity of the phases. There is a significant gradient in the viscosity of PLA, EBA-GMA, and PC under high-temperature processing conditions, which also contributes to this particular phase distribution. The continuous morphology of the PC phase also explains the unusually high heat resistance of the blend. It can be expected that when this continuity is broken, for example, when the PLA content in the blend increases, the heat deflection temperature will experience a sharp drop.

The PC/PLA binary blend shows a noticeably different morphology (Fig. 10(b)). While maintaining its co-continuous morphology, there is significant distribution of small PLA droplets in the PC phase. These inclusions of a brittle phase, along with the poor interphase strength, act as stress concentrators and dramatically decrease its toughness. The observed DMT moduli of the PLA and PC phases were ∼3.7 GPa and ∼2.5 GPa, respectively (red cross-section line). Delamination can be seen between phases in Fig. 10(b) due to stress imposed by the microtome cutting knife. This delamination is also reflected in the DMT modulus scan when an indenter sinks into the crevice between the delaminated phases (see blue cross-section line). It is not clear why the blend containing EBA-GMA does not show phase delamination during microtoming, but it can be hypothesized that the EBA-GMA phase in the modified blend acts as a dumper and force redistributor during the cutting, resulting in a surface without delamination and an actual PC/PLA interphase strength that is not significantly different in either case. It is also a possibility that EBA-GMA participates in interface compatibilization between PC and PLA.

4. Conclusions

In this study, the toughness of the PC/PLA blends was successfully modified using poly(ethylene-n-butylene-acrylate-co-glycidyl methacrylate) (EBA-GMA). A more than 10 times increase in the impact strength was observed when as little as 6 wt% of EBA-GMA was added to the blend. Two main factors were identified for efficient PC/PLA blend toughening: a specific acrylic core/shell rubber as impact modifier and a non-typical high processing temperature, which allowed an increase in the rate of reactive compatibilization. While most PC/PLA blend studies focused mostly on PC and PLA compatibilization, it was shown that temperature plays a very important role in toughening, and EBA-GMA is most efficient as an impact modifier in the temperature range 260–270 °C. It was demonstrated that PLA thermal degradation at very high processing temperatures of 270 °C and higher can be mitigated by a chain extender. It was thought that both of these factors play an important role in achieving a high-performance blend.

Studies showed that the EBA-GMA phase was distributed exclusively in the brittle PLA phase, allowing targeted toughening of the PC/PLA blend and achieving a dramatic improvement in toughness with low impact modifier contents in the blend. This observation was supported by a novel AFM PeakForce surface imaging technique, which showed the morphology and distribution of EBA-GMA predominantly in the PLA phase. The investigated PC/PLA blend maintained co-continuous morphology even at EBA-GMA loads of up to at least 10 wt%, which resulted in excellent heat resistance (HDT exceeding 130 °C) even at high impact modifier loads.

It was shown that PLA in PC/PLA blends can be used as a bio-based renewable component, and also that PC/PLA blends can have high toughness and heat resistance, making them suitable for high-performance applications.

Conflict of interest

The authors declare no competing financial interest.

Acknowledgements

This research is financially supported by (1) the Ontario Ministry of Agriculture, Food, and Rural Affairs (OMAFRA)-University of Guelph Bioeconomy Industrial uses Research Program (Project # 200358, 030054 and 030055); and (2) the Natural Sciences and Engineering Research Council (NSERC), Canada, Discovery Grants (Project # 401111); and Ontario Research Fund, Research Excellence Program; Round-7 (ORF-RE07) from the Ontario Ministry of Research and Innovation (MRI), currently known as the Ontario Ministry of Research, Innovation and Science (MRIS) (Project # 052644 and # 052665).

Notes and references

1. H. Tsuji and S. Miyauchi, Poly(L-lactide): 7. Enzymatic hydrolysis of free and restricted amorphous regions in poly(L-lactide) films with different crystallinities and a fixed crystalline thickness, Polymer, 2001, 42, 4463–4467.
2. Y. Doi and K. Fukuda, in Biodegradable plastics and polymers, Elsevier, Amsterdam, 1994.
3. S. S. Ray, K. Yamada, M. Okamoto and K. Ueda, New polylactide-layered silicate nanocomposites. 2. Concurrent improvements of material properties, biodegradability and melt rheology, Polymer, 2003, 44, 857–866.
4. J. Ren, Z. Liu and T. Ren, Mechanical and thermal properties of poly(lactic acid)/starch/montmorillonite biodegradable blends, Polym. Polym. Compos., 2007, 15, 633–638.
5. T. Semba, K. Kitagawa, U. S. Ishiaku and H. Hamada, The effect of crosslinking on the mechanical properties of polylactic acid/polycaprolactone blends, J. Appl. Polym. Sci., 2006, 101, 1816–1825.
6. V. Vilay, M. Mariatti, Z. Ahmad, K. Pasomsouk and M. J. Todo, Characterization of the Mechanical and Thermal Properties and Morphological Behavior of Biodegradable Poly(L-lactide)/Poly(epsilon-caprolactone) and Poly(L-lactide)/Poly(butylene succinate-co-L-lactate) Polymeric Blends, J. Appl. Polym. Sci., 2009, 114, 1784–1792.
7. P. Nugroho, H. Mitomo, F. Yoshii, T. Kume and K. Nishimura, Improvement of Processability of PCL and PBS blend by Irradiation and its Biodegradability, Macromol. Mater. Eng., 2001, 286, 316–323.
8. NatureWorks, Technology Focus Report: Blends of PLA with Other Thermoplastics, Ver.2/7/2007.
9. Fujitsu and Toray, Develop PLA Notebook Housing, Modern Plastics Worldwide, March, 11, 2005.
10. J. B. Lee, Y. K. Lee, G. D. Choi, S. W. Na, T. S. Park and W. N. Kim, Compatibilizing effects for improving mechanical properties of biodegradable poly(lactic acid) and polycarbonate blends, Polym. Degrad. Stab., 2011, 96, 553–560.
11. T. Kanzawa and T. Katsuhisa, Mechanical Properties and Morphological Changes of Poly(lactic acid)/Polycarbonate/Poly(butylene adipate-co-terephthalate) Blend Through Reactive Processing, J. Appl. Polym. Sci., 2011, 121, 2908–2918.
12. Y. Wang, S. M. Chiao, T.-F. Hung and S.-Y. Yang, Improvement in toughness and heat resistance of poly(lactic acid)/polycarbonate blend through twin-screw blending: influence of compatibilizer type, J. Appl. Polym. Sci., 2012, 125, E402–E412.
13. K. Hashima, S. Nishitsuji and T. Inoue, Structure-properties of super-tough PLA alloy with excellent heat resistance, Polymer, 2010, 51, 3934–3939.
14. Y. Wang, M.-T. Chiao, S.-Y. Lai and S.-Y. Yang, The role of polycarbonate molecular weight in the poly(L-lactide) blends compatibilized with poly(butylene succinate-co-L-lactate), Polym. Eng. Sci., 2013, 53, 1171–1180.
15. C. Liu, S. Lin, C. Zhou and W. Yu, Influence of catalyst on transesterification between poly(lactic acid) and polycarbonate under flow field, Polymer, 2013, 54, 310–319.
16. V. T. Phuong, M.-B. Coltelli, P. Cinelli, M. Cifelli, S. Verstichel and A. Lazzeri, Compatibilization and property enhancement of poly(lactic acid)/polycarbonate blends through triacetin-mediated interchange reactions in the melt, Polymer, 2014, 55, 4498–4513.
17. Y. Wang, Y.-Y. Shi, J. Dai, J. Yang, T. Huang, N. Zhang, Y. Peng and Y. Wang, Morphology and property changes of immiscible polycarbonate/poly(L-lactide) blends induced by carbon nanotubes, Polym. Int., 2013, 62, 957–965.
18. Y. Srithep, W. Rungseesantivanon, B. Hararak and K. Suchiva, Processing and characterization of poly(lactic acid) blended with polycarbonate and chain extender, J. Polym. Eng., 2014, 34, 665–672.
19. Q. Meng, M.-C. Heuzey and P. J. Carreau, Control of thermal degradation of polylactide/clay nanocomposites during melt processing by chain extension reaction, Polym. Degrad. Stab., 2012, 97, 2010–2020.
20. N. Najafi, M.-C. Heuzey, P. J. Carreau and P. Wood-Adams, Control of thermal degradation of polylactide (PLA)–clay nanocomposites using chain extenders, Polym. Degrad. Stab., 2012, 97, 554–565.
21. K. Zhang, V. Nagarajan, M. Misra and A. K. Mohanty, Supertoughened Renewable PLA Reactive Multiphase Blends System: Phase Morphology and Performance, ACS Appl. Mater. Interfaces, 2014, 6, 12436–12448.
22. H. T. Oyama, Super-tough poly(lactic acid) materials: reactive blending with ethylene copolymer, Polymer, 2009, 50, 747–751.
23. G. Kfoury, F. Hassouna, J.-M. Raquez, V. Toniazzo, D. Ruch and P. Dubois, Tunable and Durable Toughening of Polylactide Materials Via Reactive Extrusion, Macromol. Mater. Eng., 2014, 299, 583–595.
24. H. Liu, F. Chen, B. Liu, G. Estep and J. Zhang, Super Toughened Poly(lactic acid) Ternary Blends by Simultaneous Dynamic Vulcanization and Interfacial Compatibilization, Macromolecules, 2010, 43, 6058–6066.
25. H. Liu, L. Guo, X. Guo and J. Zhang, Effects of reactive blending temperature on impact toughness of poly(lactic acid) ternary blends, Polymer, 2012, 53, 272–276.
26. Q. Meng, M.-C. Heuzey and P. J. Carreau, Control of thermal degradation of polylactide/clay nanocomposites during melt processing by chain extension reaction, Polym. Degrad. Stab., 2012, 97, 2010–2020.
27. N. Najafi, M.-C. Heuzey and P. J. Carreau, Polylactide (PLA)–clay nanocomposites prepared by melt compounding in the presence of a chain extender, Compos. Sci. Technol., 2012, 72, 608–615.
28. M. Kaci, S. Cimmino, C. Silvestre, D. Duraccio, A. Benhamida and L. Zaidi, Ethylene butyl acrylate glycidyl methacrylate terpolymer as an interfacial agent for isotactic poly(propylene)/wood flour composites, Macromol. Mater. Eng., 2006, 291, 869–876.
29. V. Ojijo and S. S. Ray, Super toughened biodegradable polylactide blends with non-linear copolymer interfacial architecture obtained via facile in situ reactive compatibilization, Polymer, 2015, 80, 1–17.
30. Plastics additives: An A-Z reference, ed. G. Pritchard, Chapmann Hill, 1998.
31. Y. Wang, K. Chen, C. Xu and Y. Chen, Supertoughened biobased poly(lactic acid)-epoxidized natural rubber thermoplastic vulcanizates: fabrication, co-continuous phase structure, interfacial in situ compatibilization, and toughening mechanism, J. Phys. Chem., 2015, 119, 12138–12146.
32. H. Ge, F. Yang, Y. Hao, G. Wu, H. Zhang and L. Dong, Thermal, mechanical, and rheological properties of plasticized poly(L-lactic acid), J. Appl. Polym. Sci., 2013, 127, 2832–2839.
33. R. Legras and J.-P. Mercier, Crystallization of bisphenol-a polycarbonate. 3. Spherulitic growth rate of the plasticized polymer, J. Polym. Sci., Part B: Polym. Phys., 1979, 17, 1171–1181.
34. O. Wachsen, K. H. Reichert, R. P. Krüger, H. Much and G. Schulz, Thermal decomposition of biodegradable polyesters – III. Studies on the mechanisms of thermal degradation of oligo-L-lactide using SEC, LACCC and MALDI-TOF-MS, Polym. Degrad. Stab., 1997, 55, 225–231.
35. P. E. LeMarec, L. Ferry, J.-C. Quantin, J.-C. Benezet, F. Bonfils, S. Guilbert and A. Bergeret, Influence of melt processing conditions on poly(lactic acid) degradation: molar mass distribution and crystallization, Polym. Degrad. Stab., 2014, 110, 353–363.
36. D. Garlotta, A literature review of poly(lactic acid), J. Polym. Environ., 2001, 9, 63–84.
37. Y. Yuryev and P. Wood-Adams, Effect of surface nucleation on isothermal crystallization kinetics: theory, simulation and experiment, Polymer, 2011, 52, 708–717.

---