Intramolecularly Synergic Catalysis Enables Efficient Closed‐Loop Recycling of Polyesters and Polycarbonates

ABSTRACT

Efficient and selective chemical recycling of commercial plastics back to monomers is a pivotal strategy for achieving a circular plastics economy. However, rare catalysts can deliver versatility in both polymerization and depolymerization processes. Herein, we report an efficient and tolerant dinuclear zinc catalyst for the bulk depolymerization of various polyesters and polycarbonates to the corresponding cyclic monomers with >99% selectivity, and for repolymerization to virgin materials, through an intramolecular bimetallic synergistic dual‐activation mechanism. A record‐breaking activity of up to 43 kg of polymer/g of catalyst per hour was observed in the catalytic bulk poly(_L‐lactic acid) depolymerization with microwave assistance. The catalyst could be recycled several times without obvious loss in activity and selectivity. This work provides an effective guideline for catalyst design that closes the loop for commercial plastics, offering a direct route to transform plastic waste into renewable feedstocks and advancing a truly sustainable materials economy.

This article reports a versatile and tolerant dinuclear zinc catalyst for highly efficient and selective closed‐loop recycling of polyesters and polycarbonates to pristine monomers and virgin polymers. The results reveal that rational catalyst design enables potential extension to certain commercial polymers, where both polymerization and depolymerization can be achieved using the same catalyst, thus realizing a fully closed‐loop material lifecycle.

1. Introduction

Polyesters and polycarbonates are ubiquitous in modern society because of their light weight, versatility, and highly tunable properties. As of 2021, the annual global market size of commercial polyesters and polycarbonates stood at 75 million metric tons [1]. However, the durability and resistance of these plastic materials, combined with inadequate disposal strategies, pose significant threats to both the environment and human health. Although mechanical processing provides a widely practiced method for recycling a fraction of plastic waste, it invariably leads to the deterioration of material properties [2, 3]. Consequently, developing efficient and sustainable plastic recycling technologies is critical to establishing a circular plastics economy. Chemical recycling to monomers (CRM) represents a pivotal strategy for achieving a sustainable plastics economy [4]. By converting plastic waste into its constituent monomers, CRM not only diminishes the consumption of virgin feedstocks but also effectively addresses the end‐of‐life issue of plastic materials. To this end, CRM strategy has been actively explored for various polyesters and polycarbonates [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]. Despite these advancements, effective CRM technologies for polyesters and polycarbonates still face challenges regarding energy and catalyst efficiency.

Aliphatic polyesters and polycarbonates exhibit outstanding mechanical performance, versatility, and processability, making them suitable for diverse applications, including construction materials, electronic components, food packaging, and tissue engineering scaffolds [18, 19, 20]. The ring‐opening polymerization (ROP) of cyclic monomers provides precise control over molecular architecture, molar mass, and stereochemistry [21]. However, the transition to a circular plastics economy for ROP‐derived polyesters and polycarbonates is impeded by the limited and often fossil‐based sources of their cyclic monomers. Therefore, achieving true sustainability requires efficient CRM technologies that close the loop by regenerating the essential molecular building blocks for repolymerization [21, 22]. In addition to thermodynamics, the success of CRM also depends on favorable kinetics, because depolymerization must overcome a significant activation barrier [23, 24]. Therefore, developing a catalyst to lower the barrier is crucial for achieving high efficiency and selectivity. A series of metal catalysts, including Al(III)‐, Mg(II)‐, Sn(II)‐, Co(II)‐, and Zn(II) complexes, have been developed for the CRM of polyesters and polycarbonates [6, 7, 25, 26, 27]. For example, Zhang et al. reported that Mg‐based Lewis acid catalysts facilitate the solvent‐free chemical recycling of polyesters and polycarbonates [6]. Prior to this work, Williams et al. achieved the chemical recycling of commercial poly(_L‐lactic acid) (PLLA) to _L‐lactide (_L‐LA) using Sn(II)/alcohol catalyst [7]. Subsequently, they developed a heterometallic Mg(II)Co(II) complex for mediating the copolymerization of epoxides with CO₂ at 80°C as well as the corresponding depolymerization of the resulting polycarbonates at elevated temperatures [27]. Among these metal‐based catalysts, Zn(II) exhibits high depolymerization efficiency due to its intermediate Lewis acidity, which optimally balances carbonyl activation and alkoxide labilization [23]. More recently, mononuclear zinc catalysts bearing various ancillary ligands have been developed (see Figure S1) [26, 28, 29, 30, 31, 32]. In 2020, Jones et al. reported the first dinuclear zinc catalyst for PLA depolymerization [33]. Since then, numerous dinuclear Zn catalysts have been reported, exhibiting excellent efficiency toward the depolymerization of polyesters and polycarbonates [10, 34, 35, 36, 37]. Despite these advancements, no catalytic systems capable of mediating both polymerization and depolymerization of diverse polyesters and polycarbonates at low catalyst loadings have been reported to date.

Generally, the rate‐determining step for both depolymerization of polyesters and polycarbonates and repolymerization of the corresponding monomers is the nucleophilic addition of an alkoxide to an acyl group [22, 38]. While a Lewis acid facilitates this process by activating the acyl group, it simultaneously complicates the process by coordinating with the alkoxide, which introduces a dissociation energy penalty. Balancing this contradiction is thus the key to achieving the efficient depolymerization of polyesters and polycarbonates and the reverse process. Inspired by the proven efficacy of bimetallic synergistic catalysis in the ROP of cyclic esters [39, 40], we hypothesized that a catalyst integrating acyl activation and alkoxide binding would facilitate nucleophilic addition, enabling efficient depolymerization and polymerization. In this study, we developed a highly efficient dinuclear zinc catalyst for the CRM of various polyesters and polycarbonates, spanning polylactides, poly(thio)lactones, and polycarbonates, alongside the corresponding reverse reactions (Figure 1). This work establishes an efficient design principle for catalysts capable of closing the materials loop of commercial plastics, affording a straightforward strategy to convert plastic waste into renewable feedstocks and thereby advancing the development of a truly sustainable materials economy.

FIGURE 1.

Dinuclear zinc catalyst for the closed‐loop recycling of polyesters and polycarbonates through an intramolecular synergistic dual‐activation pathway.

2. Results and Discussion

PLLA, which has a molecular weight (M _n) of 23.0 kg/mol with a dispersity (Đ) of 1.20, as well as a decomposition temperature (T _d) of 336°C, was selected as a model polymer for optimizing the depolymerization conditions (Scheme S1 and Figures S2–S10), using the well‐established bimetallic zinc complex 1 as the catalyst (Figure 2a) [40]. With a feed ratio of [PLLA]₀/[1]₀ = 600:1 at 180°C, the conversion of PLLA depolymerization reached 99% after 0.5 h, corresponding to a turnover frequency (TOF) of 1200 h⁻¹ (Figure 2b). Notably, no meso‐LA was detected (>99% _L‐LA selectivity), indicating the absence of epimerization [41]. Under identical conditions, the monometallic analog 2 exhibited decreased efficiency with a TOF of 300 h⁻¹. The superiority of 1 is more evident at low catalyst loadings. At a [PLLA]₀/[catalyst]₀ ratio of 1000, 1 is 10 times more active than 2 (TOF = 667 vs. 63 h⁻¹). Moreover, at a feed ratio of 4000, 1 retains high activity (TOF = 363 h⁻¹), whereas 2 loses activity. Therefore, 1 exhibits exceptional activity and _L‐LA selectivity (>99%), surpassing not only 2 but also reported recycling catalytic systems (Table S1).

FIGURE 2.

Catalytic activity of dinuclear (1) and mononuclear (2) zinc catalysts for catalyzing the CRM of PLLA. (a) Structures and (b) catalytic activity of catalysts. Logarithmic plots of the initial rate versus catalyst concentration for (c) 1 (The bulk depolymerization reactions were conducted with a feed ratio of [PLLA]₀/[1]₀ = 6000:1, 7000:1, 8000:1, and 9000:1, respectively, at 180°C) and (d) 2 (The bulk depolymerization reactions were conducted with a feed ratio of [PLLA]₀/[2]₀ = 150:1, 175:1, 200:1, and 225:1, respectively, at 180°C).

We further investigated the kinetic features of the CRM of PLLA to rationalize the superior performance of catalyst 1 by monitoring the reaction at various catalyst concentrations using ¹H nuclear magnetic resonance (NMR) spectroscopy (Figure 2c). A reaction order of 1.09 was obtained for the 1‐catalyzed depolymerization of PLLA, indicating a first‐order dependence on the catalyst concentration. In contrast, 2‐mediated PLLA depolymerization proceeded via an intermolecular bimetallic synergistic catalysis mechanism with a second‐order dependence on the catalyst concentration, as confirmed with a reaction order of 2.03 (Figure 2d). These findings confirm that 1‐catalyzed PLLA depolymerization proceeds through an intramolecular bimetallic synergistic catalysis mechanism, and thus, the intramolecular bimetallic synergistic catalysis strategy was verified as a viable candidate to circumvent the dilemma caused by the catalyst concentration effect of the monometallic complex.

We next focus on the mechanism of dinuclear catalyst 1‐mediated PLLA depolymerization. During the initiation stage, two possible pathways can lead to the formation of the active alkoxide chain end: deprotonation of the hydroxyl chain end (Figure 3a) or transesterification via nucleophilic attack of the catalyst‐bound ethoxide on the carbonyl group (Figure 3b). Density functional theory (DFT) calculations show the deprotonation path proceeds with a lower energy barrier (3.3 kcal/mol) than that of transesterification (9.9 kcal/mol) (Figures 3 and S11). This computational result is further supported by two experimental observations: (1) the ethanol, the byproduct of deprotonation, was detected upon heating the mixture of 1 and PLLA at 180°C for 5 min (Scheme S2 and Figure S12) and (2) the acetate‐terminated polymer, in which the hydroxyl chain end is protected as an acetate, exhibited significantly lower activity (TOF = 142 vs. 363 h⁻¹) and a reduced isolated yield of _L‐LA (39% vs. 99%) (Table S2, Entries 1 and 2; Figure S13). Therefore, deprotonation of the hydroxyl chain end is the dominant pathway for generating the active chain‐end alkoxide during the initiation stage, which subsequently undergoes cyclization, leading to further degradation. The further degradation at 180°C with a feed ratio of [PLLA]₀/[1]₀ = 1000:1 was monitored using ¹H NMR spectroscopy and gel permeation chromatography (GPC) (Figure S14). A linear decrease in the M _n of the residual polymer with conversion, alongside consistently narrow dispersities, corroborates an unzipping depolymerization mechanism. This process is proposed to proceed via nucleophilic attack of a zinc‐bound alkoxide on a carbonyl group activated by the adjacent metal center.

FIGURE 3.

Depolymerization of PLLA with different terminated groups: (a) hydroxyl‐terminated PLLA; (b) acetate‐terminated PLLA.

DFT calculations were also performed to gain further insight into the catalytic cycle of 1‐catalyzed PLLA depolymerization (Figure 4a). To simplify the calculations, an isopropoxide was used to model the polymer chain end (I‐1). Once the active alkoxide species enters the catalytic depolymerization cycle, it undergoes nucleophilic attack on the adjacent carbonyl, yielding the tetrahedral intermediate (I‐2) with an energy barrier of ΔG ^‡ = 5.9 kcal/mol (TS‐1). Subsequently, I‐2 further undergoes the elimination step with the energy barrier of 5.6 kcal/mol (ΔG _TS‐2  = 4.4 kcal/mol), generating I‐3. Moreover, _L‐LA release and concomitant coordination of the zinc center to the carbonyl group of another polymer chain result in the formation of I‐4 through an exergonic process (ΔG = −8.1 kcal/mol). Furthermore, the lower energy barrier of ΔG ^‡ = 0.7 kcal/mol (TS‐3) is surpassed through a proton transfer process in a slightly exergonic step of ΔG = −0.4 kcal/mol, resulting in the formation of I‐5. Finally, isopropanol is released in an endergonic step with ΔG = 4.9 kcal/mol. This step regenerated the alkoxide species (I‐1), completing the catalytic depolymerization cycle. Overall, the rate‐determining step of depolymerization is the active alkoxide nucleophilic attacking the adjacent carbonyl (TS‐1), and the calculated energy barrier is 5.9 kcal/mol for this catalytic cycle.

FIGURE 4.

(a) Calculated catalytic cycle mechanism for the dinuclear zinc catalyst 1‐mediated PLLA depolymerization via an intramolecular bimetallic synergistic catalysis mechanism. (b) The rate‐determining step of mononuclear catalyst 2‐mediated PLLA depolymerization via an intermolecular bimetallic synergistic catalysis mechanism. (c) Charge analyses of the two bimetallic pathways to account for the observed differences in activity. The Gibbs free energies are given in kcal/mol.

In addition, we also calculated the possible catalytic cycles of mononuclear catalyst 2. The energy barriers for the monometallic and intermolecular bimetallic degradations are 18.0 and 10.3 kcal/mol, respectively (Figures S17 and S18). The lower barrier for the intermolecular bimetallic pathway is consistent with the experimentally observed second‐order dependence on catalyst 2 concentration during the degradation (Figure 2d). Moreover, this barrier also remains significantly higher than that of the intramolecular pathway mediated by catalyst 1 (ΔG ^‡ = 5.9 kcal/mol, Figure 4a,b). These findings align with the experimentally observed activity differences between the catalyst systems, with bimetallic complex 1 exhibiting higher activity (TOF = 363 h⁻¹ at 0.0125 mol%) than monometallic complex 2 (TOF < 1 h⁻¹ at 0.025 mol%, Figure 2b).

We further analyzed the key transition states of bimetallic catalyst 1 and monometallic catalyst 2 (TS‐1 in Figures 4a and M‐TS‐1 in Figure 4b) for further understanding the origin of the superior catalyst 1. In TS‐1, the ethoxide ligand bridges the two zinc centers simultaneously, which delocalizes charge and mitigates the energetic penalty associated with alkoxide dissociation, thereby stabilizing the overall system [42]. By contrast, in M‐TS‐1, the intermolecular bimetallic pathway, it lacks such stabilizing interactions; the unassisted dissociation of the alkoxide leads to a pronounced increase in system energy during the nucleophilic attack. This hypothesis is supported by charge analyses (Table S3). Moreover, this intramolecular dinuclear architecture facilitates synergistic dual activation essential for catalytic enhancement (Table S4 and Figure 4c): (1) zinc‐bound alkoxide exhibits substantially increased electron density (q Hirshfeld: −0.242 e vs. −0.228 e; q NPA: −0.903 e vs. −0.849 e); and (2) the carbonyl substrates displayed a pronounced electrophilic character enhancement (q Hirshfeld: +0.222 e vs. +0.207 e; q NPA: +0.895 e vs. +0.875 e). This cooperative electronic restructuring results in a dramatic 4.4 kcal/mol reduction in the energy barrier of the rate‐determining step (from ΔG ^‡ = 10.3 to 5.9 kcal/mol, Figure S19). Consequently, this smoother energy pathway rationalizes the exceptional performance of the dinuclear catalyst observed experimentally, even at extremely low concentrations.

With the efficient depolymerization catalyst 1 in hand, we further explored its performance in larger‐scale PLLA recycling. The depolymerization was performed using 10 g of PLLA with a feed ratio of [PLLA]₀/[1]₀ = 8000:1 for 22 h. Remarkably, this process afforded _L‐LA in 92% yield and >99% selectivity, demonstrating its potential for larger‐scale PLLA recycling. However, a comparison with small‐scale reactions revealed that efficacy was compromised upon scale‐up; the isolated yield of _L‐LA decreased from 99% at the 0.4 g scale to 92% in the 10 g reaction (Table S1, Entry 6). This result indicates the critical need for further enhanced depolymerization rates to establish a practical, large‐scale PLLA recycling process. To ascertain the underlying cause of the slow depolymerization rate observed during larger‐scale PLLA recycling, conventional heating was initially employed to raise the polymer temperature to 180°C, while monitoring the temperature at the central region of the polymer (Figures S20–S22). These results showed that thermal transfer was notably inefficient, and it took 45 min to heat 3 g of PLLA to the target temperature of 180°C. This process generated significant thermal gradients between the surface and central regions, leading to uneven depolymerization conditions and significantly reducing overall efficiency. Microwave assistance has been established as a powerful tool for enhancing the chemical recycling of some polymers [43, 44, 45]. Microwave heating enabled the target reaction temperature to be reached within 2 min. Under these optimized conditions, _L‐LA was obtained in 97% yield and >99% purity after 4 h, demonstrating a significant acceleration of the depolymerization.

Next, our investigation focused on understanding the effect of the microwave‐assisted reaction conditions on the catalytic activity of 1 for PLLA depolymerization (Table 1). A control experiment revealed that microwave heating did not degrade PLLA in the absence of a catalyst (Table 1, Entry 1). With a feed ratio of [PLLA]₀/[1]₀ = 8000:1, nearly quantitative conversion of PLLA to _L‐LA with >99% selectivity was achieved at 180°C with microwave heating within 4 h. At the same time, a significantly lower activity and a reduced yield were observed under the conventional heating method (Table 1, Entries 2 and 3). Further decreasing the catalyst loading, no obvious change in activity and _L‐LA selectivity was discovered (Entries 4−6). For example, even at such a low catalyst loading (0.0033 mol%), the depolymerization proceeded without any detectable epimerization. This represents the lowest catalyst loading reported for the CRM of PLLA (Table S5).

TABLE 1.

Bulk depolymerization of PLLA catalyzed by dinuclear zinc catalyst 1 with microwave assistance. ^a

Entry	T (°C)	t (h)	Conv. b (%)	Activity c (g PLLA h−1 g Cat. −1)	Selec. d (%)
1	180	4.0	—	—	—
2	180	4.0	97	173	>99
3 e	180	4.0	29	52	>99
4 f	180	5.5	99	162	>99
5 g	180	10.0	97	173	>99
6 h	180	16.0	95	162	>99
7	190	1.5	97	464	>99
8	200	0.5	96	1382	>99
9	220	0.1	99	7199	>99
10	240	1 min	99	43 204	>99
11	250	0.5 min	97	83 816	96
12 i	240	3 min	98	14 117	>99
13 j	240	0.1	92	6621	>99

^a Reaction conditions: 1.0 g PLLA, [PLLA]₀/[1]₀ = 8000:1, 0.8 mbar, in a 20 mL quartz tube unless otherwise stated.

^b The conversion was determined using the gravimetric method, with the average conversion calculated from three experiments.

^c Activity = mass of polymer/mass of catalyst per hour.

^d The selectivity of _L‐LA was determined by ¹H NMR spectroscopy and GC‐MS.

^e Depolymerization reaction under the conventional heating method.

^f [PLLA]₀/[1]₀ = 10000:1.

^g [PLLA]₀/[1]₀ = 20000:1.

^h [PLLA]₀ /[1]₀ = 30000:1.

ⁱ PLLA sample with a M _n of 166.0 kg/mol.

^j PLLA sample with a M _n of 707.0 kg/mol.

The reaction temperature has a strong influence on the depolymerization rate (Table 1, Entries 7−11). Fixed the feed ratio of [PLLA]₀/[1]₀ = 8000:1, increasing the temperature to 200°C resulted in a substantial activity enhancement, approximately 8 times faster than that performed at 180°C. Further increase of reaction temperature to 220 or 240°C resulted in rapid quantitative conversion, with activities of 7199 and 4.3 × 10⁴ g _PLLA g _Cat. ⁻¹ h⁻¹, respectively, while retaining >99% _L‐LA selectivity (Entries 9 and 10). However, an enhanced reaction temperature of 250°C caused the formation of meso‐LA via epimerization to a certain extent (Entry 11). Moreover, the effect of the molecular weight on PLLA depolymerization was also investigated. Remarkably, with the use of higher‐molecular‐weight PLLA (M _n = 166.0 kg/mol) and a feed ratio of [PLLA]₀/[1]₀ = 8000:1, 98% conversion was achieved in only 3 min, while maintaining excellent selectivity (>99%) toward _L‐LA (Entry 12). Notably, even for ultra‐high‐molecular‐weight PLLA (M _n = 707.0 kg/mol), 1 afforded 92% conversion within 6 min without epimerization byproducts under the identical conditions (Entry 13).Overall, the dinuclear catalyst addresses this issue by being remarkably tolerant, maintaining >99% selectivity for _L‐LA across a broad temperature window (170°C–240°C). This effective epimerization suppression enabled the production of high‐purity _L‐LA and facilitated the subsequent synthesis of high‐quality stereoregular PLLA via repolymerization.

Furthermore, catalyst recyclability and tolerance were examined in detail. Catalyst recycling experiments were performed for PLLA depolymerization at a fixed feed ratio of [PLLA]₀/[1]₀ = 5000:1 (3.0 g of PLLA was added per cycle). Both activity and selectivity for _L‐LA have no obvious change after five cycles (Figure 5a). Thermo‐gravimetric analysis confirmed the high thermal stability of complex 1 with a T _d of 291°C (Figures S23 and S24), significantly higher than the reaction temperature, thereby ensuring its durability. The excellent activity and tolerance of the dinuclear zinc catalyst 1 enable cost‐efficient and sustainable continuous flow processes for industrial‐scale applications.

FIGURE 5.

(a) The reusability performance of dinuclear zinc catalyst 1 was investigated at 180°C with a feed ratio of [PLLA]₀/[1]₀ = 5000:1 over five different CRM cycles. (b) CRM of PLLA from a mixed commodity plastics feedstock initially yields dyed _L‐LA monomer at 180°C. Recrystallization was used to decolorize the dyed monomer, yielding pure _L‐LA monomer with 96% yield and >99% selectivity. PET, poly(ethylene terephthalate); PP, polypropylene; PE, polyethylene; BPA‐PC, bisphenol A polycarbonate; PBAT, poly(butylene adipate‐co‐terephthalate).

Mixed plastic waste typically comprises a complex feedstock containing additives, such as colorants, plasticizers, and stabilizers, along with other impurities. A key advantage of CRM is its ability to recover pure monomers directly from such heterogeneous mixtures, thereby avoiding energy‐ and cost‐intensive separation steps. To evaluate the efficacy of selective PLLA depolymerization and _L‐LA recovery from realistic plastic mixtures, we conducted CRM experiments using PLLA blended with common commercial plastics, including poly(ethylene terephthalate) (PET), polypropylene, polyethylene, bisphenol A polycarbonate (BPA‐PC), and poly(butylene adipate‐co‐terephthalate) (Figure 5b). The bulk depolymerization reaction was carried out with 1 (0.2 wt%) using a heterogeneous mixture of PLLA (2.0 g) and mixed plastics (3.0 g) at 180°C and 0.8 mbar for 6.0 h. Catalyst 1 demonstrated excellent tolerance and selectively converted PLLA to _L‐LA from the mixed plastic matrix. The crude monomer product, which initially contained colorants, was decolorized by recrystallization from a DCM/n‐hexane (1:99, v/v) mixture. Subsequent isolation and vacuum drying afforded pure _L‐LA as a white crystalline solid in 96% yield (1.92 g) with a selectivity of >99% (Figure S25). These results confirm that PLLA can be efficiently and selectively recycled from complex plastic waste streams via CRM, circumventing the need for energy‐intensive upstream separation.

In addition, the bimetallic complex was demonstrated to be also very efficient in the catalytic depolymerization of a broad range of polyesters and polycarbonates, including poly(glycolic acid) (PGA), poly(ε‐caprolactone) (PCL), poly(δ‐valerolactone) (PVL), polythioester (PtCL), poly(trimethylene carbonate) (PTMC), poly(cyclopentene carbonate) (PCPC) and poly(cyclohexene carbonate) (PCHC) at different temperatures ranging from 160°C to 230°C, back to the corresponding cyclic monomers in nearly complete conversion and >99% selectivity (Figures 6 and S26−S33). It is worth emphasizing that being contrary to those of Williams et al., who reported a heterometallic Mg(II)Co(II) complex for PCHC selectively depolymerizing back to cyclohexene oxide (CHO) [46], the present dinuclear zinc catalyst 1 enabled the quantitative depolymerization of PCHC to trans‐cyclohexene carbonate (trans‐CHC) with >99% selectivity. To elucidate the mechanism of the 1‐mediated PCHC depolymerization, interval time sampling was performed in the reaction system at 180°C with a feed ratio of [PCHC]₀/[1]₀ = 500:1 to monitor by ¹H NMR spectroscopy and GPC (Figure S34). Throughout the process, the molecular weight of the residual polymer decreased linearly with increasing PCHC conversion, with consistently narrow dispersities, supporting an unzipping depolymerization mechanism. DFT calculations were performed to give Gibbs free energy profiles (including intermediate and transition states) for the 1‐mediated PCHC depolymerization to produce CHO or trans‐CHC (Figure S35). The energy profiles revealed that the formation of trans‐CHC was kinetically favored over that of CHO. Notably, when (R, R)‐PCHC (95.5% ee) was treated with 1 (2 mol %) at 200°C, it underwent quantitative conversion to (R, R)‐trans‐CHC with 95.0% ee within 3 h (Scheme S3). This near‐perfect maintenance of stereochemical integrity confirms that the catalyst exhibits both chemoselectivity and stereochemical preservation during ring‐closing depolymerization, thereby establishing a promising route for the closed‐loop recycling of stereoregular polymers. Furthermore, the catalytic system also quantitatively depolymerized poly(cyclopentene carbonate) (PCPC) to cis‐CPC with >99% selectivity (Figure S36).

FIGURE 6.

Depolymerization of diverse polyesters and polycarbonates catalyzed by dinuclear zinc catalyst 1 (see the Supporting Information for details).

We further expanded the substrate scope to commercial aromatic polyesters, including PET, poly(trimethylene terephthalate) (PTT), poly(butylene terephthalate) (PBT), and poly(ethylene naphthalate) (PEN), which were synthesized via the step‐growth polymerization of aromatic diacids and diols. Although the catalytic bulk pyrolysis of these polyesters proved to be impractical due to high energy input, complex product distributions, and poor monomer recovery [47, 48, 49, 50], their methanolysis mediated by catalyst 1 was very efficient. With a very low catalyst loading of 0.01 mol%, the complete methanolysis of PET, PTT, and PBT was achieved, yielding dimethyl terephthalate (DMT) with >99% selectivity (Figures S37–S39). The methanolysis of PEN afforded dimethyl 2,6‐naphthalenedicarboxylate (NDC) in 95% yield (Figure S40). In addition, the catalyst achieved the complete depolymerization of BPA‐PC under similarly low loading conditions (0.01 mol%) with >99% monomer selectivity (Figure S41). This study established a tolerant and general catalytic strategy for polymer recycling, demonstrating exceptional efficiency across the broadest substrate scope reported to date, encompassing diverse polyesters and polycarbonates, and representing a significant advancement toward circular plastic economies.

Having demonstrated the efficiency of bimetallic complex 1 in the depolymerization of various polyesters and polycarbonates, we further explored its versatility by employing it as a catalyst for the repolymerization of the recovered monomers, thereby establishing a closed‐loop material lifecycle (Table 2). With a feed ratio of [_L‐LA]₀/[1]₀ = 1000:1 in toluene ([Monomer]₀ = 3.3 M), complex 1 efficiently mediates the ROP of isolated _L‐LA at 25°C, achieving >99% conversion within 0.5 h (Table 2, Entry 1). The resulting PLLA has an M _n of 68.1 kg/mol, corresponding to a TOF of 2000 h⁻¹. Increasing the feed ratio to 5000:1 leads to an _L‐LA conversion of 69% after 2.0 h at 80°C, yielding PLLA with an M _n of 255.6 kg/mol and a TOF of 1725 h⁻¹. The thermal properties of the regenerated polymer matched those of the original PLLA, demonstrating the closed‐loop circularity of PLLA with industrial‐grade properties (Table 2, Entry 2; Figure S26). Remarkably, the 1‐catalyzed ROP of _L‐LA proceeds in a controlled manner. Intermittent sampling experiments revealed that the M _n of the resultant PLLA increases linearly with _L‐LA conversion. Experiments with varying initiator loadings also showed that the M _n increases linearly with the feed ratio [_L‐LA]₀/[1]₀. Moreover, the M _ns at different _L‐LA conversions and initiator loadings are close to the calculated MW (M _n ^theo.), together with low Ðs (Figure S27). Notably, beyond _L‐LA, the monomers recovered from the depolymerization of other polyesters and polycarbonates, including GL, ε‐CL, δ‐VL, TMC, trans‐CHC, and tnCL, could also be efficiently repolymerized into their respective polymers, with the use of the same catalyst. Under feed ratios of [Monomer]₀/[1]₀ ranging from 500:1 to 8000:1, the ROP proceeded smoothly with TOFs of up to 10 000 h⁻¹, producing the corresponding polymers with M _ns ranging from 3.8 to 214.7 kg/mol (Table 2, Entries 3–9; Figures S28–S33). Furthermore, DMT obtained from PET methanolysis was also successfully used in a 1‐catalyzed step‐growth polymerization with ethylene glycol to regenerate PET (Figure S37). The exceptional efficiency of dinuclear zinc catalyst 1 for both depolymerization and repolymerization of the recovered monomers across a broad range of polyesters and polycarbonates represents a notable advancement toward a sustainable circular plastics economy.

TABLE 2.

Re‐polymerization of various recycled monomers catalyzed by the bimetallic complex 1. ^a

Entry	Mon.	Mon./1 (molar ratio)	T (°C)	t (h)	Conv. b (%)	TOF c (h−1)	M n d (kg/mol)	Ð d
1	L‐LA	1000/1	25	0.5	>99	2000	68.1	1.22
2	L‐LA	5000/1	80	2.0	69	1725	255.6	1.82
3 e	GL	500/1	120	2.0	93	233	3.8	1.23
4	ε‐CL	3000/1	80	0.3	>99	10000	154.6	1.28
5	ε‐CL	8000/1	80	2.0	54	2160	214.7	1.41
6	δ‐VL	2000/1	25	0.5	86	3440	79.3	1.26
7	TMC	2000/1	60	0.5	91	3640	102.2	1.31
8 f	trans‐CHC	500/1	80	60.0	91	8	22.7	1.29
9	tnCL	2000/1	100	2.0	73	730	83.1	1.69

^a Conditions: Toluene was used as the solvent, with an initial monomer concentration of 3.3 M unless otherwise stated.

^b Determined by ¹H NMR analysis of the crude reaction mixture.

^c Turnover frequency (TOF) = mol of monomer/mol of catalyst per hour.

^d Determined by GPC in DMF at 40°C versus polystyrene standards.

^e Determined by GPC in 1,1,1,3,3,3‐hexafluoroisopropanol containing trifluoroacetic acid sodium (10 × 10⁻³ mol/L) at 40°C versus polymethyl methacrylate standards.

^f [Monomer]₀ = 0.8 M.

3. Conclusion

In summary, we have developed an efficient, tolerant, and versatile catalyst for bulk depolymerization of various polyesters and polycarbonates at 180°C−240°C, including polylactides, poly(thio)lactones, and polycarbonates, back to the corresponding cyclic monomers with >99% selectivity via an intramolecularly synergistic dual‐activation and unzipping mechanism, concerning a first‐order dependence on the bimetallic catalyst concentration. Exemplified by the depolymerization of PLLA, the catalyst exhibits remarkable tolerance, maintaining high performance across a broad temperature range (180°C−240°C), and could be recycled several times without obvious loss in activity and selectivity. Microwave heating greatly promoted the depolymerization rate over conventional heating, particularly at very low catalyst loadings and large‐scale production, but it did not affect the catalytic pyrolysis mechanism and product selectivity. A record‐breaking activity of up to 43 kg PLLA/g of catalyst per hour was achieved, with >99% selectivity for _L‐LA. Notably, the catalytic thermolysis of PLLA can afford pure _L‐LA monomer selectively in the presence of commercial PET, BPA‐PC, PP, PE, and PBAT waste, avoiding the need for time‐ and cost‐intensive separation processes. In addition, the present dinuclear zinc catalyst was highly active for the methanolysis of PET, PTT, PBT, PEN, and BPA‐PC, producing a monomer with >99% selectivity. Of importance, the present catalyst was also very efficient in the repolymerization of the recovered monomers at 25°C−120°C, affording polymers with high molecular weights of up to 255.6 kg/mol. This catalyst is unique in that it possesses high activity and >99% selectivity for both polymerization and depolymerization processes, thereby establishing a real closed‐loop material lifecycle. The results indicated that, with judicious design of the catalyst, it is possible to create a perfect closed‐loop lifecycle of some commercial polymers that their synthesis and chemical recycling processes can be performed using the same catalyst under industrially relevant conditions, allowing the establishment of an economically viable material recycling model.

Author Contributions

Gan‐Tao Ma: writing – original draft, formal analysis, data curation, methodology, validation, writing – review and editing, investigation. Bai‐Hao Ren: investigation, writing – original draft, writing – review and editing, methodology, data curation. Tian‐Jun Yue: writing – original draft, methodology, writing – review and editing. Wen‐Xue Peng: investigation, validation, data curation. Wei‐Min Ren: conceptualization, funding acquisition, writing – review and editing, methodology, supervision, resources. Ye Liu: writing – original draft, formal analysis. Xiao‐Bing Lu: conceptualization, funding acquisition, writing – original draft, writing – review and editing, methodology, project administration, formal analysis, supervision, resources, data curation.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: anie73498‐sup‐0001‐SuppMat.pdf.

Data Availability Statement

The data that support the findings of this study are available in the Supporting Information of this article.