Biobased Cyclic Enoate Monomers: Enabling Intrinsically Crystalline, Chemically Recyclable, Ultratough Polymers

ABSTRACT

Chemically recyclable polymers that depolymerize back to their monomers offer a promising alternative to non‑recyclable petroleum‑based plastics. However, three seemingly intractable trade‑offs have long hindered the rational design of circular polymers that combine high chemical recyclability with high performance: polymerizability versus depolymerizability, depolymerizability versus material performance, and crystallinity versus ductility. Here, we introduce a monomer design strategy based on renewable cyclic enoates that enables highly regioselective (exclusive head‑to‑tail) and stereoselective (E‑selective) ring‑opening metathesis polymerization (ROMP), yielding polyolefins/polyesters with high crystallinity and full chemical recyclability. Through modulation of the ring–chain equilibrium, both the forward polymerization and the reverse depolymerization proceeded to near‐quantitative conversion. These polymers defy the aforementioned trade‑offs by exhibiting an unusual combination of desirable properties, including intrinsic crystallinity, chemical recyclability, and excellent performance metrics such as high thermal stability, high mechanical strength, ductility, and toughness.

A monomer design strategy based on renewable cyclic enoates enables highly regioselective and stereoselective ring‑opening metathesis polymerization. The resulting polyesters exhibit an unusual combination of intrinsic crystallinity, quantitative depolymerizability, high thermal stability, high mechanical strength, ductility, and toughness.

1. Introduction

Crystalline polymers dominate the field of commodity plastics (75%) [1]. Although these materials have become ubiquitous and indispensable in modern society, their linear life cycles and limited end‐of‐life management options contribute to severe environmental challenges, including plastic waste accumulation and the depletion of finite resources [2, 3, 4, 5]. Closed‐loop chemical recycling offers a promising alternative, enabling polymers to be efficiently depolymerized into their constituent monomers or oligomers, which can subsequently be repolymerized into pristine materials with preserved properties [6, 7, 8, 9, 10, 11, 12, 13, 14]. Among chemically recyclable systems, crystalline polymers that undergo depolymerization are often synthesized via ring‐opening polymerization (ROP), with polyesters being representative examples [15, 16, 17, 18, 19, 20, 21]. As a variant of ROP, ring‐opening metathesis polymerization (ROMP) also provides access to recyclable polymers [22, 23, 24, 25], particularly those with polyolefin backbones [26, 27, 28, 29, 30]. ROMP‐based chemically recyclable polymers are typically derived from low‐strain cyclic monomers (Figure 1). However, the resulting polyolefin chains often lack structural regularity (regio‐ and stereoregularity), leading to amorphous plastics that are predominantly non‐crystalline [31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44]. As a result, their mechanical performance, such as ductility and toughness, is dwarfed by that of crystalline commodity polyolefins (e.g., HDPE, LDPE), which significantly restricts their practical applicability.

FIGURE 1.

Properties of representative chemically recyclable polymers prepared by ROMP. [Correction added on May 27, 2026, after first online publication: The “green checkmark symbol” in Figure 1 has been updated.]

While incorporating polyethylene‐like segments into the main [45, 46, 47, 48, 49, 50, 51, 52, 53, 54] or side chains [55] can render polyolefins crystallizable, the intrinsic backbone structure still suffers from poor stereochemical and regiochemical control during the ROMP of cycloolefins. The ring‐opening and recombination steps lack regio‐ and stereoselectivity, leading to limited chain regularity—a key bottleneck in the development of crystalline [56, 57, 58], recyclable polymers via ROMP. While certain tungsten and molybdenum catalysts have demonstrated the ability to control cycloolefin polymerization with enhanced selectivity, air‐ and water‐stable, commercially available catalysts such as Grubbs‐type complexes generally fail to achieve high regiocontrol in the polymerization of conventional cycloolefins [59].

The design of novel low‐strain cycloolefin monomers that enable highly regio‐ and stereoselective ROMP under Grubbs catalysis represents a promising strategy for constructing well‐defined polyolefins with intrinsic crystallinity and chemical recyclability. It has been established that when an olefin with high reactivity is reacted with a sterically hindered and electron‐deficient olefin partner, selective cross‐metathesis and stereoselective outcomes can be achieved even under equimolar conditions [60, 61]. Drawing inspiration from this small‐molecule organic chemistry methodology, we aim to develop a new class of cycloolefin platforms tailored for macromolecular engineering, wherein ROMP proceeds with high regio‐ and stereocontrol, thereby enabling precise regulation of chain regularity. In this context, the ester group emerges as an attractive functionality—it is sterically demanding (beneficial for regiocontrol), electron‐withdrawing (favorable for stereocontrol), and serves as a versatile linkage for coupling functional building blocks. Placing an ester group adjacent to the reacting olefin yields an α, β‐unsaturated lactone [62, 63, 64, 65], also referred to as a cyclic enoate [66].

In this work, by shifting the ester group position within the cycloolefin skelenton, we constructed a series of biobased lactone isomers via ring‐closing metathesis and systematically investigated their polymerization behavior with respect to regio‐ and stereoselectivity. This approach screened cyclic enoate monomers, enabled the synthesis of highly ordered crystalline polymers. We further characterized their thermal stability and mechanical properties to elucidate the influence of regio‐ and stereochemistry on polymer performance. Finally, we explored the chemical recyclability of these crystalline polymers back to oligomers/monomers under Grubbs catalyst conditions.

2. Results and Discussion

2.1. Monomer Design and Polymerization Studies

As shown in Scheme 1, we employed undecenol, a castor oil‐derived bio‐based feedstock, as the starting material. Esterification with acrylic acid followed by ring‐closing metathesis smoothly afforded the 13‐membered cyclic monomer M1. Using M1 as the reference monomer, we reduced the enoate ring size to 10‐membered (M3), 6‐membered (M6), and 5‐membered (M5) rings for comparison. To investigate the influence of the enoate structure on polymerization regio‐ and stereoselectivity, we designed and synthesized isomeric monomers M2 and M4 via an analogous synthetic route, in which the ester group is not conjugated with the olefin but is instead remote from or separated from the olefin unit. The structural difference between the two types of monomers—those with the ester group adjacent to the olefin unit versus those nonadjacent to it (e.g., M1 and M2)—was unambiguously confirmed by ¹H NMR spectroscopy (Figure 2a).

SCHEME 1.

Reaction scheme and selectivity for the ROMP of cyclic olefinic monomers with different ring sizes and C═C bond positions.

FIGURE 2.

Characterization of monomers and polymers. (a) ¹H NMR spectra (400 MHz, CDCl₃) of M1, P1 _95k, M2, and P2. (b) FTIR spectra of M1, P1 _95k and P2. (c) ¹³C NMR spectra (101 MHz, CDCl₃) of the olefinic region for P1 _95k and P2. (d) ¹H–¹H COSY spectrum (olefinic region) of P1 _95k. (e) Proposed ROMP mechanism for the formation of P1 from M1, featuring exceptional regio‐ and stereoregularity.

Considering that M5 and M6 are commercially available, ROMP of these monomers was tested initially. Despite extensive optimization of the solvent type, monomer concentration (up to bulk polymerization), monomer‐to‐catalyst feed ratio, reaction time, and polymerization temperature (Table S1), no polymeric products with molecular weights above 1000 Da were obtained, indicating that the monomers are nonpolymerizable, largely due to their low ring strain energy (RSE, Table S4), similar to that of cyclohexene. For the remaining monomers (M1–M4), which have larger ring sizes and therefore a greater entropy gain, polymerization using Grubbs second generation (G2) catalyst proceeded very smoothly (Table S2), with high conversions achieved in all cases. For the 13‐membered‐ring‐based monomers (M1 and M2), both yielded polymers with comparable molecular weights (P1 _95k and P2 _94k, Table 1). We then systematically investigated the regio‐ and stereoselectivity of the polymerization leading to P1 _95k. FTIR spectroscopy (Figure 2b) revealed similar overall profiles for M1 and P1 _95k, consistent with their identical chemical compositions. A notable difference was observed at approximately 1600 cm⁻¹: P1 _95k exhibited a distinct absorption band at this frequency, which was absent in P2. This band is attributed to the C═C stretching vibration. Typically, symmetric C═C bonds in P2 are Raman‐active but IR‐inactive. However, conjugation with the adjacent ester group in the enoate moiety breaks the symmetry, rendering the C═C bond in P1 _95k IR‐active. However, distinguishing between cis and trans olefin configurations via the C–H bending vibrations, often in the 1000–800 cm⁻¹ region [67, 68], proved challenging due to signal congestion.

TABLE 1.

Characterization of ROMP‐prepared Polymers.

Polymer	M n a (kDa)	Đ a	Alkene trans ratio b	T m c (°C)	ΔH m c (J/g)	χ d	T d,5% e (°C)	E f (MPa)	σ b f (MPa)	ε b f (%)
P1 95k	94.5	2.26	95%	59.0	47.0	40%	386.9	93.3 ± 2.2	24.0 ± 3.5	2315 ± 374
P2	94.1	2.21	80%	34.2	49.1	39%	385.6	36.1 ± 3.4	10.1 ± 0.4	931 ± 77
P3	193.8	1.99	96%	82.1	38.2	32%	385.3	94.0 ± 0.1	20.2 ± 1.5	2221 ± 233
P4	182.0	1.92	81%	27.5	38.8	29%	381.5	/	/	/
P1 179k	178.7	1.75	97%	61.7	45.5	42%	382.1	101.3 ± 0.9	28.5 ± 2.2	2120 ± 442
P1 289k	288.9	1.74	97%	61.9	45.6	43%	393.9	101.6 ± 4.5	33.7 ± 2.6	2118 ± 261
P1 412k	412.1	1.76	97%	60.6	38.8	41%	395.8	77.5 ± 10.2	26.4 ± 1.9	1512 ± 308

^a Determined by gel permeation chromatography (GPC) in THF against polystyrene standards.

^b Obtained from integration of characteristic peaks in the ¹H NMR spectra.

^c Obtained from the second heating scan of DSC thermograms.

^d Calculated by peak deconvolution of XRD patterns.

^e Recorded by thermogravimetric analysis (TGA); decomposition temperature corresponds to 5% weight loss.

^f Determined from tensile testing at a crosshead speed of 10 mm/min.

We therefore turned to ¹³C NMR spectroscopy to elucidate the configuration of the main‐chain double bonds. The spectrum of P1 _95k exhibited two sharp signals in the olefinic region (Figure 2c), indicating precise control over both regio‐ and stereochemistry during chain growth. This pattern is diagnostic of either an exclusively head‐to‐tail enchainment or a perfectly alternating head‐to‐head/tail‐to‐tail sequence [57, 69]. Complementary ¹H NMR measurements revealed a pair of vinylic protons with a coupling constant of 15.6 Hz (Figure 2d), characteristic of a trans double bond. The preferential formation of trans olefins under G2 catalysis is likely governed by steric interactions within the metallacyclobutane intermediate, where the adjacent ester and alkyl chains disfavor the cis pathway [60]. A corresponding mechanism is proposed in Figure 2e to validate this process. Definitive assignment of the regiochemistry required two‐dimensional NMR methods, as one‐dimensional NMR could not distinguish between the possible regioisomers—both would produce identical multiplet patterns for H_a (doublet) and H_b (doublet of triplets). The ¹H–¹H COSY spectrum displayed clear cross‐peaks correlating H_a and H_b (Figure 2d), providing direct evidence for vicinal coupling and thus supporting a head‐to‐tail connectivity. The alternative head‐to‐head/tail‐to‐tail arrangement would lack this through‐bond correlation. Collectively, these spectroscopic data confirm that P1 _95k possesses a highly ordered microstructure with exceptional regio‐ and stereoregularity. This level of ROMP selectivity control over M1 can be extended to other commercially available Grubbs catalysts, such as Grubbs third generation catalyst (G3, Figure S1), indicating that the selectivity is not catalyst‐dependent but rather dictated by the enoate structure itself.

To assess whether this level of control extends to other members of the cyclic enoate family, we examined a ten‐membered ring analog (Figure S2). The resulting polymer P3, despite having a different repeat unit length, exhibited comparable regio‐ and stereochemical fidelity, underscoring the generality of the catalyst's selectivity. By contrast, the polymerization of M2 or M4 under identical conditions produced a material with poorly defined microstructure. For example, the olefinic region of the ¹³C NMR spectrum for P2 (Figure 2c) showed a complex envelope of resonances, reflecting a statistical distribution of enchainment patterns—head‐to‐tail, head‐to‐head, and tail‐to‐tail—further compounded by mixed cis/trans olefin geometries. Quantitative analysis of the aliphatic carbon signals (Figure S3) provided insight into the extent of disorder: regioselectivity was essentially absent (∼50% head‐to‐tail junctions), while stereoselectivity remained modest (∼80% trans content, Table 1).

To evaluate the control over molecular weight, a series of polymerizations of M1 was conducted at various monomer‐to‐catalyst ratios ([M1]/[G2] = 200/1, 300/1, 500/1, and 1000/1). Gel permeation chromatography (GPC) analysis (Figure 3a) revealed a clear dependence of the number‐average molecular weight (M _n) on the feed ratio (Figure 3b), demonstrating good molecular weight control despite the relatively high dispersity. The resulting high‐molecular‐weight polymers are denoted as P1 _95k, P1 _179k, P1 _289k, and P1 _412k, respectively (Table 1), enabling subsequent investigations into the effect of molecular weight on the properties of this structurally regular polymer. Furthermore, the matched molecular weights between P1 _95k and P2, as well as between P1 _179k, P3 and P4, provide a well‐defined platform for comparative studies, allowing for systematic elucidation of structure–property relationships.

FIGURE 3.

Elucidating the driving force and control in ROMP of M1. (a) GPC traces of P1 with controlled molecular weights. (b) Linear relationship between M _n and the monomer/catalyst ratio. (c) Time‐conversion plot for polymerization kinetics. (d) van't Hoff analysis for the polymerization of M1 at 0.1 M.

We next investigated the polymerization thermodynamics of M1, aiming to precisely determine its ceiling temperature (T _c) or floor temperature (T _f). However, this seemingly straightforward process proved nontrivial. Initial experiments were conducted under standard conditions (1 M monomer concentration, [M1]: [G2] = 100:1). To evaluate polymerization thermodynamics, it is necessary to determine the equilibrium monomer concentration at various temperatures. After 3 h at 55°C, the polymerization mixture had solidified; the reaction was quenched and conversion was determined by ¹H NMR to be >95% (Table S3). However, accurate integration proved difficult, precluding reliable quantification of conversion. Raising the temperature to 65°C for 3 h still resulted in >95% conversion. Given that reactions were stopped after 3 h, equilibrium may not have been reached; thus, the true equilibrium conversion at these temperatures is expected to be even higher—certainly exceeding 95%. Under such conditions, reliable conversion values cannot be obtained for fitting the van't Hoff equation.

Generally, decreasing monomer concentration reduces the entropic driving force for polymerization, shifting ΔG toward positive values. We therefore lowered the monomer concentration to 0.5 M in order to establish an equilibrium between polymerization and depolymerization. At 0.5 M monomer concentration, kinetic studies at 35°C revealed that equilibrium was reached within approximately 4 h, yet conversion remained above 95% (Figure S5). Further dilution to 0.2 M at 35°C afforded an equilibrium monomer conversion of 87.7%. Ultimately, at 0.1 M concentration, the equilibrium conversion at 35°C was 79.1%, falling within a range amenable to precise conversion analysis. We systematically investigated the polymerization kinetics at this concentration over a temperature range of 35–50°C (Figure 3c). As expected, higher temperatures accelerated the rate of polymerization and shortened the time required to reach equilibrium. Notably, equilibrium conversion decreased with increasing temperature, confirming a negative entropy change for the polymerization at this concentration. Fitting the experimental data obtained at a monomer concentration of 0.1 M to the van't Hoff equation (Figure 3d) yielded the following thermodynamic parameters for M1 polymerization: an enthalpy change (ΔH) of −2.99 kcal mol⁻¹, an entropy change (ΔS) of −1.98 cal mol⁻¹ K⁻¹, and a Gibbs free energy change (ΔG) of −2.44 kcal mol⁻¹ at 25°C. These figures suggest a thermodynamically more favorable polymerization than is typical for many previously reported ROMP‐derived depolymerizable polyolefins [22]. Importantly, while the measured ΔH aligns with the expectation of low ring strain for this monomer, it is marginally lower than the computationally estimated ring‐strain energy (RSE) of 4.17 kcal mol⁻¹ (Table S4). This minor divergence, a common observation in ROMP systems, can be rationalized by considering side reactions such as the formation of cyclic oligomers [31] and the energy associated with adopting specific polymer chain conformations [70]. Because concentration only affects the ΔS of polymerization [41], the enthalpy change at 1.0 M will be the same as that at 0.1 M [71]. The ΔS at 1.0 M can is derived as +2.59 cal mol⁻¹ K⁻¹, reflecting an entropy‑favorable ROMP process for M1 at high concentrations, consistent with its large‑ring structure. The T _f value is then calculated to be −876°C, which would place depolymerization below this theoretical temperature, making it experimentally inaccessible at this concentration. Considering the relative low ΔH value of M1, the ΔS value can be substantially reduced by further dilution to increase entropy gain during polymerization, enabling depolymerization to occur under mild T _c conditions [53, 72]. Consequently, the system is anticipated to achieve high conversion efficiencies in both polymerization and depolymerization directions by shifting the ring‐chain equilibrium.

2.2. Thermal Properties and Crystalline Behavior of Polymers

The structural regularity imparts semicrystalline characteristics to both P1 and P3, as evidenced by differential scanning calorimetry (DSC) thermograms (Figure 4a). P2 and P4, despite possessing flexible chain segments, also exhibits semicrystalline behavior; however, their melting temperature (T _m = 34°C for P2, T _m = 27.5°C for P4) is substantially lower than that of their isomeric counterpart P1 _95k (T _m = 59°C). For P1, variations in molecular weight lead to modest changes in both T _m and melting enthalpy (ΔH _m) (Table 1). Given that approximately 20% of the double bonds in P2 adopt a cis configuration, which reduces chain regularity and thereby limits lamellar thickening [73, 74], the observed depression in T _m is unsurprising. However, such a low T _m (close to room temperature) will inevitably compromise key material properties, including heat resistance, modulus, and mechanical strength, thereby limiting its practical applications.

FIGURE 4.

Thermal properties of polymers. (a) DSC heating scan of P1 _95k, P2, P3, and P4. (b) Hydrogenation of P1 and P2 and microstructure of HP1 and HP2. (c) DSC heating scan of HP1 and HP2. (d) XRD patterns of P1 _95k, P2, P3, P4 and HDPE. (e) TGA curves of P1 _95k, P2, P3, and P4.

To decouple the contributions of regio‐ and stereochemistry to crystallization behavior, we subjected both P1 and P2 to catalytic hydrogenation using p‐toluenesulfonyl hydrazide, thereby eliminating olefinic unsaturation along the polymer backbone (Figure 4b). It should be note that the cleavage of backbone ester bonds during such catalytic hydrogenation process was unavoidable [46, 54], resulting in lower molecular weights for HP1 and HP2 compared to the original polymers. Nevertheless, by carefully controlling the hydrogenation conditions, we maintained their molecular weights at approximately 40 kDa, ensuring that their crystallization behavior could be compared without interference from molecular weight effects. The resulting hydrogenated polymers (denoted HP1 and HP2) differ solely in their regiochemical microstructure: HP1 features exclusively head‐to‐tail enchainment, whereas HP2 comprises a statistical mixture of head‐to‐tail, head‐to‐head, and tail‐to‐tail linkages. As shown in Figure 4c, HP1 exhibits a higher T _m and ΔH _m than HP2; unequivocally demonstrating that regioregularity alone promotes more favorable chain packing. This behavior mirrors that before hydrogenation: P1 had a higher T _m than P2, and after hydrogenation, HP1 retained both a higher T _m and greater crystallinity than HP2, highlighting the advantage of the regio and stereoselective polymerization method. Collectively, these findings establish that the crystal of P1, with its high regio‐ and stereochemical purity, exhibits higher thermostability than that of P2. Notably, P3—despite its lower melting enthalpy—displays an even higher melting point (T _m = 82.1°C, Figure 4a) than P1, further underscoring the impact of structural regularity on thermal transitions.

X‐ray diffraction (XRD) measurements were performed to interrogate crystallinity and lattice parameters (Figure 4d). The diffraction pattern of P1 _95k closely resembles that of high‐density polyethylene (HDPE); however, the two characteristic reflections appear at 19.5° and 23.6°—shifted to lower angles relative to the (110) and (200) reflections of HDPE (21.5° and 23.9°, respectively). This shift, consistent with Bragg's law, indicates an expansion of the d‐spacing and suggests inclusion of the enoate functionality within the crystal lattice, implying looser lattice packing of P1 compared to HDPE. P2 displays a nearly identical diffraction profile, implying formation of a similar lattice structure. The degree of crystallinity (χ) for each polymer was quantified by deconvolution of the XRD patterns (Figures S20–S26). P1 and P2 exhibit comparable crystallinities (approximately 40%), indicating that chain regularity primarily affects their crystal thermostability, but does not influence total crystallinity or lattice packing. P3 with a shorter repeating unit length shows a markedly lower crystallinity value (32%), manifesting as two broad reflections at 19.8° and 21.6°. This reduced crystallinity correlates well with the lower melting enthalpy observed for P3 by DSC (ΔH _m = 38.2 J/g, Figure 4a). P4 exhibited the lowest crystallinity, likely due to its T _m (27.5°C) being near room temperature, leading to partial melting.

Thermostability was assessed by thermogravimetric analysis (TGA) under a nitrogen atmosphere. Whereas chemically recyclable polyesters typically exhibit 5% weight‐loss temperatures (T _d,5%) around 300°C [17], with values exceeding 350°C considered exceptional [75]. P1 _95k displays a remarkable T _d,5% of 387°C (Figure 4e). For the highest‐molecular‐weight sample, P1 _412k, this value reaches 395.8°C, approaching 400°C. These data substantially surpass those reported for most known ROP‐derived recyclable polyesters (see Table S6 for detailed data) [16, 17], representing a significant advancement in the upper bounds of polyester thermal stability. This exceptional performance likely stems from a combination of high molecular weight and the relatively low density of ester linkages along the polymer backbone [76, 77]. Notably, all polymers exhibit comparable thermal stabilities, indicating that adjacency of the double bond to the ester group does not compromise thermostability.

2.3. Mechanical Properties of Polymers

The combination of high thermal transition temperatures and excellent thermal stability suggests a broad processing window for these crystalline polymers, rendering them amenable to conventional melt processing techniques such as heat molding for the fabrication of plastics. Because P4 is partially melted at room temperature, uniaxial tensile testing was conducted only on specimens of P1, P2, and P3; the resulting stress–strain curves are presented in Figure 5a, and representative mechanical parameters are summarized in Table 1. These data collectively reveal a pronounced dependence of mechanical behavior on crystalline microstructure and chain structure.

FIGURE 5.

Mechanical properties of polymers. (a) Stress–strain curves of P1 _95k, P2 and P3. (b) Stress–strain curves of P1 with different molecular weights. (c) T _d,5% and toughness of representative chemically recyclable polyesters.

P1 _95k and P3 exhibit characteristic ductile plastic deformation, evidenced by distinct necking during stretching—hallmarks of tough plastic materials. Despite this shared qualitative behavior, quantitative mechanical properties vary substantially across the series. P2, although possessing a crystallinity comparable to that of P1 _95k, displays a melting point near ambient temperature (T _m = 34°C), resulting in a soft material with a low Young's modulus (36.1 MPa) and a tensile strength at break of approximately 10 MPa. P3, despite its lower crystallinity, benefits from a high melting point (82.1°C) and accordingly exhibits intermediate modulus, strength, and extensibility. P1 _95k combines the favorable attributes of high crystallinity and elevated T _m (59°C), yielding a Young's modulus roughly three times that of P2—a value (∼100 MPa) comparable to that of low‐density polyethylene (LDPE). Among the three materials, P1 demonstrates the most balanced mechanical profile, excelling in both tensile strength and elongation at break.

To further elucidate the influence of molecular weight on mechanical performance, we examined a series of P1 samples spanning molecular weights from approximately 100 kDa–400 kDa (Figure 5b). Given the flexible backbone of P1, even the lowest‐molecular‐weight sample (94.5 kDa) is expected to lie well above its critical entanglement molecular weight. Nevertheless, molecular weight exerts a discernible effect on mechanical properties. Optimal performance is achieved at an intermediate molecular weight of 288.9 kDa, which affords a tensile strength of 33.7 MPa, an elongation at break exceeding 2000%, and a toughness of 534 MJ/m³—values that surpass those of both HDPE and LDPE. Although the polymers constructed in this system are obtained via olefin metathesis polymerization, their structure resembles polyesters more than polyolefins. Conventional polyesters are typically brittle, yet the polyesters reported here exhibit excellent ductility. Benchmarking against representative chemically recyclable polyesters reported in the literature reveals that P1 exhibits a particularly favorable combination of thermostability and toughness (Figure 5c, see Table S6 for detailed data) [16, 17]. Rather than relying on anionic ROP, our strategy provides an alternative route to high‑toughness polyesters via olefin metathesis polymerization.

2.4. Chemical Recycling of Polymers

The regular distribution of depolymerizable C═C bonds along the backbone of P1 renders this structurally uniform polymer a promising candidate for chemical recycling via ring‐closing metathesis depolymerization, ideally affording the pristine monomer in a closed‐loop fashion. To evaluate this potential, polymer samples were dissolved in toluene at a concentration of 20 mM (with respect to olefin units) and treated with 1 mol% G2 at 50°C for 4 h—conditions typical for the depolymerization of low‐strain cycloolefins [31, 38, 40, 53, 78]. Under these conditions, ring‐closing metathesis proceeded smoothly, achieving >98% conversion and yielding a viscous liquid mixture comprising macrocyclic oligomers of varying ring sizes along with the monomer as the predominant species (∼43% yield). This product distribution reflects an equilibrium among cyclic species of different ring dimensions [79, 80], as corroborated by GPC elution profiles (Figure 6a). Kinetic analysis of the depolymerization process revealed complete consumption of the parent polymer within 2 h, followed by gradual conversion of oligomeric macrocycles into monomer until equilibrium was attained at approximately 4 h (Figure 6b). Quantification of component fractions by integration of GPC refractive index traces showed excellent agreement with ¹H NMR spectroscopic analysis (Figure S6). Following removal of residual catalyst using Quadrapure TU particles, the remaining macrocyclic oligomers could then be directly repolymerized via ROMP to afford virgin polymer with high molecular weights (M _n = 60 kDa) and high conversion (>80%, Figure 6a).

FIGURE 6.

Chemical recycling of P1. (a) GPC elution curves of and M1, P1 _95k, depolymerized system, and regenerated P1. (b) Depolymerization kinetics of P1 _95k at 50°C (olefin units = 20 mM). (c) Depolymerization temperature v.s. M1 yield (olefin units = 20 mM). (d) ¹H NMR spectra (400 M, CDCl₃) of P1 _95k, recovered M1 and virgin M1.

To optimize the conditions for closed‐loop recycling, we systematically varied depolymerization concentraion, catalyst loading, and temperature while monitoring monomer recovery ratio. Consistent with an entropy‐driven depolymerization process, monomer yield increased with decreasing polymer concentration (Figure S7) or rising temperature (Figure 6c). Optimal results were obtained at 85°C with a repeat‐unit concentration of 20 mM and a reaction time of 1 h using 1 mol% G2, affording M1 in 91% yield. The recovered monomer was purified by recrystallization from methanol to remove residual ruthenium catalyst, achieving high purity (GC >97%, Figure S9) and structural identity with the original monomer as confirmed by ¹H NMR spectroscopy (Figure 6d).

3. Conclusion

Based on bio‐based feedstock, we have developed a new class of lactone monomers containing reactive carbon–carbon double bonds. By varying the alkyl chain length, the ring strain can be systematically tuned. These monomers undergo efficient ROMP with excellent forward polymerization efficiency and reverse depolymerization efficiency under appropriate conditions that balance ring–chain equilibria. The polymerization proceeds with exceptional regio‐ and stereoselectivity, ensuring high chain regularity to the resulting polymers. Through systematic elucidation of structure–property relationships, we reveal that the ester groups along the polyolefin backbone exerts surprisingly large “positional effects” [81] on the crystallization behavior, thermomechanical performance, and recyclability of the resulting polymer variants. The enoate‐based isomer enables chemically recyclable crystalline bioplastics that exhibit remarkable heat resistance, high thermal stability, and ultra‐tough mechanical properties. This work overcomes the long‐standing challenge in conventional ROMP—namely, the difficulty of accessing crystalline plastics due to poor polymerization selectivity—thereby substantially broadening the application landscape of chemically recyclable polymers. Our design embodies the principles of renewable sourcing, end‐of‐life recyclability, and high‐performance during service. Looking forward, the regularly distributed ester bonds along the polymer backbone also render these materials degradable to heterotelechelic monomers or macromonomers, enabling recycling via polycondensation routes. Furthermore, the cyclic enoate motif serves as a versatile template: incorporation of functional building blocks into this platform does not compromise the regio‐ and stereoselectivity of the polymerization, establishing a new platform for the development of chemically recyclable polymers with tailored functionalities. Finally, this strategy provides a new route to circular and high‐performance polyesters via olefin metathesis polymerization.

Author Contributions

Ying Wang: investigation, writing – original draft. Jianhua Tang: methodology. Yiyang Liang: data curation. Zhenyang Luo: project administration. Yucheng He: validation.Puyou Jia: funding acquisition. Enhua Xu: supervision. Ye Sha: supervision, resources, writing – review and editing, conceptualization.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

The authors have cited additional references within the Supporting Information [82–85].

Supporting File: anie72841‐sup‐0001‐SuppMat.pdf.

Data Availability Statement

The data that supports the findings of this study are available in the supplementary material of this article