A Tert‐Alkyl Terephthalate Derivative as a Universal Degradable Unit for Condensation Polymers

ABSTRACT

Incorporating cleavable triggers into the backbones of parent polymers enables programmed degradation, offering a pathway to chemically recyclable plastics. Although various triggers have been reported to date, simple esters bearing a C─O bond cleavable by photocatalytic reactions have not been successfully employed as triggers. This is because the C─O bond of simple esters is inert. In this study, we demonstrated that efficient ester C─O bond cleavage can be achieved by generating a stable tertiary alkyl radical upon bond scission. Based on this principle, we designed a dicarboxylic acid core featuring a tertiary alkyl ester incorporated into a terephthalic acid scaffold. The designed core was incorporated into polyarylate, polyurethane, and polyester, and all three polymers underwent selective C─O bond cleavage of the tertiary alkyl ester under identical photocatalytic conditions, achieving degradation efficiencies of up to 99% within 2–4 h. Unlike conventional hydrolysis, our method selectively cleaves only the targeted ester bond, leaving others intact. This yields fewer, simpler products and greatly simplifies monomer recovery. Notably, our degradable polymers retained physical properties comparable to parent plastics despite trigger incorporation.

A sterically bulky dicarboxylic acid core bearing a tertiary alkyl ester enables photocatalytic C─O bond cleavage via generation of a stabilized tertiary alkyl radical. Incorporated into polyarylate, polyurethane, and polyester, this universal trigger facilitates efficient degradation under identical conditions (up to 99% within 2–4 h). Remarkably, the resulting polymers retain material properties comparable to the parent polymers.

1. Introduction

Plastic pollution presents a global challenge that impedes the transition to a sustainable society [1, 2]. To overcome this impasse, an approach incorporating degradation triggers directly into polymer chains has been advocated [3, 4, 5, 6]. This strategy—embedding degradability into polymer chains during synthesis—dramatically facilitates the chemical decomposition of these robust polymers [7, 8, 9, 10, 11, 12, 13, 14, 15]. Recent advances have enabled sophisticated designs that program photodegradable or photocatalytically cleavable triggers into polymer chains, whose backbones are composed solely of C─C bonds. For example, triggers such as carbonyls (CO) [16, 17, 18], acylsilanes [19], hydroxyl groups [20], acetals [21], and activated esters [22, 23] have been reported, bringing controlled polymer degradation closer to reality (Figure 1, part 1). However, when polymer structures are greatly modified to facilitate degradation, the material properties are often significantly compromised compared to those of the parent polymer.

FIGURE 1.

Previous issues.

In contrast, for condensation polymers containing carbon–heteroatom bonds such as esters, urethanes, or amides in their backbone (e.g., polyesters [24, 25, 26, 27], polyamides [28, 29, 30, 31, 32, 33], polycarbonates [34, 35, 36], and other engineering plastics [37, 38], hydrolysis is in principle feasible. The degradation of these plastics often requires harsh conditions—high energy input and/or large amounts of chemical reagents—with only a few exceptions [39]. Furthermore, conventional methods suffer from the critical drawback of non‑selective, indiscriminate cleavage of ester, urethane, and amide bonds. As a result, unintended sites are also degraded, forcing the separation of the target monomers from complex mixtures of hydrolyzed by‑products. This renders the overall chemical recycling process highly inefficient (Figure 1, part 2). Therefore, there is a strong need to develop reactions that enable selective degradation of condensation polymers under mild conditions.

Photocatalytic reductive cleavage using simple esters as degradation triggers is attractive due to their ready availability and the fact that, unlike hydrolysis [40, 41, 42, 43, 44], they do not require large amounts of base or heating. Classically, cleavage of the alcohol‐derived C─O bond in small molecules has required unstable activating groups [45, 46]. Transformations of a simple ester require strong reductants. Consequently, it has been frequently targeted in the development of molecular transformations for small molecules, and extensive research has been conducted on its cleavage [47]. However, although numerous examples of ester C─O bond cleavage have been reported for small molecules, such reactions have rarely been applied as photocatalytic degradation triggers for polymers.

We therefore hypothesized that if the radical generated upon degradation were a tertiary alkyl radical, a degradation trigger (breakable unit) could be designed without the need for an active ester. Based on this hypothesis, we developed a degradable trigger core applicable to photocatalytic C─O cleavage reactions (via single‐electron transfer (SET)) driven by the generation of stable tertiary alkyl radicals. Specifically, we designed a dicarboxylic acid core featuring a tertiary alkyl ester incorporated into a terephthalic acid scaffold (Figure 2). This core possesses two key characteristics: (1) Smooth and selective cleavage: Even in the presence of ester or urethane bonds along the polymer chain, only the tertiary alkyl ester moiety is selectively cleaved. Under mild conditions, this C─O bond undergoes reductive cleavage, efficiently generating a tertiary alkyl radical that serves as the initiation point for degradation (Figure 2, part 1). (2) Broad versatility: The core bears carboxylic acid termini, allowing the introduction of desired monomer structures and incorporation into diverse polymers, with all resulting materials being degradable under identical conditions (Figure 2, part 2).

FIGURE 2.

Degradation concept.

In this paper, we describe the unified photocatalytic degradation of diverse plastics (with efficiencies up to 99%) and the material properties of the resulting polymers. The carboxylic acid termini of the core can be converted to enable incorporation of the same degradation functionality into commodity, performance polymers and engineering plastics, such as polyarylate (PAR), polyurethane (PU), and polyethylene terephthalate (PET). Although the introduction of degradation initiation sites often compromises material properties, the degradable polymers developed in this study exhibit material properties comparable to those of parent plastics. This is attributed to the common and robust structural motif of the degradation core.

2. Results and Discussion

2.1. Design and Degradability of Model Polymer

We previously demonstrated that tertiary alkyl bromides undergo efficient C─Br bond cleavage to generate tertiary alkyl radicals using copper catalysis [48, 49, 50, 51, 52]. However, when the corresponding tertiary alkyl esters were employed as substrates, the same copper catalyst (reduction potential: ‐0.8 V vs Fc [53]) failed to induce C─O bond cleavage. Reductive cleavage of ester C─O bonds is generally challenging [54], representing a significant barrier in developing degradation reactions.

During the course of our study, we discovered that excited‐state organic photocatalysts with high reducing power promote single‐electron transfer (SET) to the ester, leading to C─O bond cleavage and the generation of tertiary alkyl radicals via a Markó‐Lam‐type reaction [55]. Based on this insight, we designed and synthesized monomer 1—a tertiary alkyl ester bearing polymerizable allyl groups at both termini—as a degradable trigger (core block) for incorporation into polymer chains. Using this monomer, the metathesis polymerization was carried out at 600 Pa to remove the evolved ethylene gas. When ethylene gas was not removed, the degree of polymerization was extremely low. As the result, monomer 1 readily polymerized via metathesis to afford Polymer 1 (M _n = 20000) (Figure 3a). Although the target polymer is designed with degradability in mind, both the monomer used in this reaction and the resulting polymer are chemically stable, as exemplified by their electrochemical stability and resistance to hydrolysis. For example, cyclic voltammetry revealed reduction potentials of ‐2.08 V and ‐2.12 V (vs Fc) for monomer 1 and Polymer 1, respectively. We also evaluated the hydrolytic stability of this bond under basic hydrolysis (See Supporting Information). As a result, the steric hindrance of the bulky tertiary alkyl structure imparts high hydrolytic stability under use conditions (approximately 12‑fold greater stability compared to conventional primary esters under basic conditions.). These results indicate that the designed degradable core is not a fragment that compromises the stability of the polymer.

FIGURE 3.

Synthesis and degradation of model polymer A. ^aA mixture of Polymer A, catalyst (5.0 mol%), DPE 2 (4 equiv.) additive (10 mol%), and THF/H₂O (95/5, 0.25 M) was stirred under irradiation (365 nm) at room temperature for 24 h. Yields were determined by ¹H NMR analysis. Isolated yields were shown in parentheses. The apparent molecular weights of all polymers were determined by size exclusion chromatography (SEC) in tetrahydrofuran (THF) using polystyrene (PS) standards. E _deg was calculated from the reported method [56]. ^bThe reaction time was 48 h. ^c450 nm LED was used.

To identify conditions for C─O bond cleavage in Polymer 1, we screened various photocatalysts. Reactions were conducted under 365 nm irradiation in the presence of photocatalyst, 1,1‐diphenylethylene (DPE, 2), and additives (Figure 3b). In this study, DPE (2) was chosen as a reliable radical trapping reagent, and the resultant lactone 3 was used as an indicator for degradation efficiency. With photocatalyst PC1, the degree of degradation (E _deg [56]) reached 46%, yielding lactone 3 (3%) from reaction of the generated tertiary radical with 2. However, the other degradation product 4 was not obtained (Figure 3b, Run 1). Efficient SET is critical for C─O bond cleavage in photochemical reactions. For carbonyl compounds, Lewis or Brønsted acids are known to facilitate SET by lowering the HOMO energy [57]. To accelerate the reaction, we tested Lewis acids (ZnBr₂, Mg(ClO₄)₂) and Brønsted acids (CSA: 10‐Camphorsulfonic Acid, PTSA: p‐toluenesulfonic acid)) (Runs 2–6). PTSA gave the highest degree of degradation and lactone 3 yield, though product 4 remained undetected even with high‐molecular‐weight Polymer 1 (Run 7). This was presumably due to incomplete degradation of the core, leaving some residues as oligomers. To achieve complete degradation of the remaining core, extended reaction times and additional photocatalysts were then examined (Runs 8–13). PC2 and PC3 showed high degrees of degradation, while PC4‐6 were sluggish. Ultimately, extending the reaction time to 48 h with PC1 afforded both 3 and 4 in high yields, confirming the effectiveness of the degradable core in Polymer A.

From the results of control experiments (see, section S3 for details: radical inhibitor tests, the effect of light and PC1, Stern–Volmer plots, CV results with or without additives), a plausible reaction mechanism is shown in Figure 3c. Initially, BNPB is excited to BNPB* upon light irradiation. Subsequently, Polymer A (–2.12 V vs Fc) is reduced by a SET from the photo‐excited BNPB* (E*_ox(BNPB^•+/BNPB *) = –2.89 V) to afford the anion radical A. The formed anion radical (A) undergoes smooth fragmentation into tert‐alkyl radical intermediate (B) in the presence of proton. Yeung and Wickens suggested that LA promotes fragmentation rather than single electron reduction [58], whereas Zeitler suggested the proton lowers the LUMO of ester [57]. We also measured redox potential in the presence/absence of proton. But the potential was not changed significantly. These results may suggest that LA or protons do not facilitate the SET process (Polymer A to A) but fragmentation [58, 59, 60]. After the formation of B, B reacts with diphenylethylene (2) to form the benzylic radical C which is oxidized by BNPB^•+ to form D. Finally, after E is formed in reaction with water, intramolecular esterification proceeds to yield lactone 3.

3. Applications

3.1. Degradable Polyarylate

To demonstrate the universality of our core, we examined three engineering plastics: polyarylate (PAR, recalcitrant with no precedent for chemical recycling), polyethylene terephthalate (PET, a benchmark in chemical recycling studies), and polyurethane (a representative functional polymer) (Figure 4). We first synthesized degradable polyarylates by replacing the ester units of commercial polyarylate with our core (Figure 4a). Monomers 5 bearing various carbon substituents were polymerized with 6 to afford the corresponding degradable polyarylates (Polymer B) (Figure 4b). Thermal properties were evaluated by TGA and DSC. All polymers exhibited decomposition temperatures around 320°C, showing no correlation with the structure of monomer 5. Since no significant change in decomposition temperature was observed upon variation of the core structure, this temperature is presumed to reflect the upper limit of the thermal stability of the tertiary alkyl ester bond in the core. Glass transition temperatures (T _g) ranged from 124–152°C, though these values are molecular‐weight‐dependent and thus only indicative. Applying our optimized degradation conditions, the polymer derived from 5‐Me,Me achieved near‐complete degradation (E _deg = 99%), affording the corresponding degradation products (3–6) in high yields. Similarly, the polymer from 5‐Me,Et degraded efficiently, but the yield of 6 was slightly low. The polymer derived from 5‑(CH₂)₅‑ and 5‑Ph,Ph did not give good yield of lactone 3 and bisphenol A (6). This result indicates that although the tertiary alkyl ester bond in the degradable core was cleaved, the ester bonds derived from bisphenol A remained intact. The reason for this is not clear, but it is presumed that either the cyclic alkyl radical or the diphenylmethyl radical generated upon cleavage of the tertiary alkyl ester bond did not react with diphenylethylene (Runs 3 and 4), or that the subsequent cyclization after addition did not proceed (Figure 3c, step E to 3). This represents a limitation of the current reaction system. These results indicate that the Me,Me‑substituted core performs optimally.

FIGURE 4.

Synthesis, degradation and properties of PAR. ^aDetermined by thermogravimetric analysis (TGA). ^bDetermined from the second cycle of differential scanning calorimetry (DSC) with a heating rate of 10°C min⁻¹. ^cThe yields were determined by ¹H NMR analysis. ^dDetermined by thermogravimetric analysis (TGA). ^eDetermined from the second cycle of differential scanning calorimetry (DSC) with a heating rate of 10°C min⁻¹. Conc. of Core = the number of core fragment (equiv.)/ core + other fragments (equiv.). ^fcore (1) / core (1) + BPA (1) = 0.50 (50%), ^g core (1) / core (1) + BPA (1) + TPA (2) = 0.25 (25%), ^hcore (1) / core (1) + BPA (5) + TPA (4) = 0.10 (10%). ⁱ Polymer B (R¹,R² = Me, Me), M _n = 37000 for tensile test, M _n = 44400 for other tests, Young: Young's modulus, CLTE: Coefficient of linear thermal expansion, TC: Thermal conductivity, DC: Dielectric constant, δ: Dielectric loss tangent, T: Transparency at 450 nm. The apparent molecular weights of all polymers were determined by SEC in THF using PS standards. E _deg was calculated from the reported method [56]. See details (Section S5).

The polymer from 5‐Me,Me was readily processable into films or plates by hot pressing. Soluble in THF (like commercial polyarylate), it enabled degradation studies in homogeneous solutions. Post‐reaction, monomers were recovered by simple filtration (Figure 4c). The recovered degradation product (5‐Me,Me) could be repolymerized, establishing an efficient chemical recycling process.

3.2. Polyarylates with Varying Degradable Core Concentrations

We next synthesized Polymers B with varying degradable core concentrations (Figure 4d). Theoretically, the lower the incorporation ratio of the core, the closer the properties of the resulting polymer approach those of commercial PAR. Commercial polyarylate (M _n = 30000) exhibits T _d = 482°C and T _g = 193°C (Run 1). To approximate this molecular weight, we prepared Polymer B (5‐Me,Me) with 50% core loading (Run 2: M _n = 37000, Đ = 1.72; Run 3: M _n = 42000, Đ = 1.82). Polymer B’ (concentration of core (CC) = 25%) with 25% core loading was obtained by polymerizing core monomer (5‐Me,Me) with polyarylate/terephthalate monomer 7 in a 1:1 ratio (71% yield, M _n = 26000, Đ = 1.87). Polymer B″ (CC = 10%) with 10% core loading was synthesized from bisphenol A (6), terephthaloyl chloride (8), and core monomer 9 (6:8:9 = 3:4:1, M _n = 46000, Đ = 1.85). Monomers 6 and 9 exhibited comparable reactivity, with ¹H NMR analysis confirming their uniform incorporation into the polymer chain. Thermal properties of Polymers B were evaluated. While decomposition temperatures (T _d) remained largely independent of core concentration, glass transition temperatures (T _g) approached that of commercial polyarylate as core loading decreased. This result suggests that as the core concentration decreases, the polymer structure approaches that of the original PAR containing no core. Polymers B’ and B″ (CC = 25% and 10%) underwent efficient degradation similar to Polymer B (Figure 4e). Degradation of Polymer B’ (CC = 25%) yielded 3 (86%), 4 (80%) and monomer 7 (83%), whereas Polymer B″ (CC = 10%) produced 3 (77%) and the corresponding oligomer 10 (M _n = 3200). Lowering the core concentration does not affect the degradation efficiency; however, the resulting degradation products become oligomers rather than monomers. Therefore, from the perspective of chemical recycling, setting the core concentration at approximately 25% is considered to be a well‐balanced option. Interestingly, because no bonds other than the ester bond of the core were cleaved, monomer 7 could be used directly for the synthesis of Polymer B’ (CC = 25%). This result indicates that the reaction achieved selective cleavage at the targeted ester bond.

3.3. Properties of Degradable PAR

We next compared the properties of degradable PAR with 50% core loading (Polymer B, R¹,R² = Me) against those of the parent commercial PAR (Figure 4f, section S5). Specifically, we evaluated tensile strength, Young's modulus, CLTE, TC, DC, tangent, and T (Tables S10–S12; Figure S54). The degradable PAR retained most of the properties required for super‐engineering plastic applications. These results demonstrate that even a newly synthesized degradable PAR exhibits properties that are largely predictable from the parent PAR. However, the clear scientific reason why some physical property values were reproduced between the original PAR and the core‐containing PAR is not yet fully understood. The benzoate‐based core synthesized in this study is structurally similar to the parent plastics containing a benzene ring in the main chain, and this structural similarity is presumed to have resulted in the good reproducibility of physical properties. In any case, it is clear that the higher the content of benzene rings in the parent plastic structure, the more effectively the core‐incorporated materials can inherit the majority of the original physical properties.

3.4. Degradable Polyurethane & PET

Having established that even 50% core loading maintains sufficient material properties, we extended our technology to polyurethane and PET (Figure 5). We incorporated the degradable core into PDDI (p‐phenylene diisocyanate)‐based polyurethane bearing 1,6‐hexanediol units and evaluated its properties (Figure 5a). Degradable polyurethane (Polymer C) with 50% core loading was synthesized efficiently and exhibited tensile strength comparable to the parent polyurethane. For example, tensile strength of Polymer C was 48 MPa (average of 3 runs), compared to 42 MPa for the parent material. T _g and T _d were 60 and 315°C (Figure S35 and S36), respectively. Furthermore, subjecting Polymer C to optimized degradation conditions afforded the corresponding products 3, 4, and 13 in high yields (Figure 5b). The recovered monomer 13 could be repolymerized with 5‐Me,Me to regenerate Polymer C. Next, degradable PET (Polymer D) was synthesized efficiently and exhibited properties comparable to the parent material. For example, tensile strength of Polymer D was 37 MPa (average of 3 runs), compared to 48 MPa for the parent PET (Figure 5c). T _g and T _d were 69 and 318°C (Figure S39 and S40), respectively. Polymers C and D both exhibited similar T _d values, which are presumably attributed to the tert‐alkyl ester linkage. Furthermore, subjecting Polymer D to optimized degradation conditions afforded the corresponding products 3, 4, and 14 in high yields (Figure 5d). The recovered monomer 14 can be repolymerized with 5‐Me,Me to regenerate Polymer D. One of the highlights of these results is that all ester and urethane bonds other than the tertiary alkyl ester contained in the core remained completely intact. As a result, the number of degradation products was minimized, making their separation very easy. This point differs significantly from the outcomes of conventional hydrolysis.

FIGURE 5.

Synthesis and degradation of degradable polyurethane & PET.

3.5. Efficient Degradation

The complete degradation of polymers in this reaction requires 48 h. To shorten this reaction time, we conducted various optimization studies (Table 1). Although the reaction exhibits high degradation efficiency even at shorter times, incomplete cleavage of polymer chains yields only oligomers (dimers and trimers) rather than the corresponding monomers. While oligomers may suffice for certain applications, we pursued further optimization to achieve complete degradation.

TABLE 1.

Reaction acceleration. ^a

^a A mixture of polymer, BNPB (5.0 mol%), DPE 2 (4 equiv.), PTSA·H₂O (10 mol%), and THF/H₂O (95/5, 0.25 M) was stirred under irradiation (365 nm) at room temperature for 2–12 h. Yields were isolated. The apparent molecular weights of all polymers were determined by SEC in THF using PS standards. E _deg was calculated from the reported method [56].

We found that light irradiation efficiency is critical for reducing reaction time. In our initial setup, side irradiation of the reaction vessel limited light exposure to the solution. In this case, the reaction time was 48 h. Photocatalytic reactions are generally initiated by photoexcitation of the catalyst. Therefore, to ensure efficient reaction progress, a certain proportion of the catalyst must be maintained in the excited state. We reasoned that this requires continuous irradiation of the reaction system with light of at least a certain intensity and thus investigated the method of light irradiation onto the reaction vessel. As a result, simply repositioning the vessel horizontally and irradiating from above maximized light exposure to the solution. This straightforward modification increased the irradiated surface area and substantially reduced the reaction time. For example, Polymer B (degradable polyarylate) afforded results equivalent to the 48‐h reaction in just 12 h (Table 1, Run 1). Further optimization yielded excellent results comparable to the 12‐h reaction within only 2 h (Run 2). Polymers C and D also underwent efficient degradation within these shortened reaction times (Runs 3 and 4).

4. Conclusion

We have established a new strategy for selective photocatalytic C─O bond cleavage in polymer backbones by designing ester‐based triggers that generate stable tertiary alkyl radicals upon single‐electron transfer. Unlike conventional approaches requiring activated esters, our simple tertiary alkyl ester core overcomes the high reduction potential barrier of ester C─O bonds through radical stabilization. The dicarboxylic acid trigger exhibits two key features: selective reductive cleavage under mild photocatalytic conditions, and broad synthetic versatility. We incorporated this core into polyarylate, polyurethane, and PET. All three polymers underwent efficient photocatalytic degradation under identical conditions (up to 99% within 2–4 h) using an organic photocatalyst and p‐toluenesulfonic acid as additive. In this C─O bond cleavage reaction, DPE was used to trap the tertiary alkyl radical generated after bond scission. However, the generated radical can react with various substrates, and we are currently developing new reactions to achieve upcycling. Although the core contains an ester linkage and can therefore be degraded under harsh hydrolytic conditions, such treatment would cleave all ester, and urethane, and amide bonds present in the condensation polymer, leading to a complex mixture of products. In contrast, the dicarboxylic acid core we have developed enables selective cleavage of only the targeted tertiary alkyl ester C─O bond via a photocatalytic reaction. As a result, monomer separation, recovery, and reuse are substantially easier compared to conventional hydrolysis methods. We term this new polymer design and degradation method utilizing a common degradation trigger “Core Block Technology.” Mechanistically, photoexcited PC1 reduces the ester to an anion radical; protonation facilitates fragmentation to a tertiary alkyl radical, which propagates depolymerization. Notably, despite the general tendency that the introduction of degradation triggers compromises polymer properties, the degradable polymers we designed retained mechanical and thermal properties comparable to those of the parent plastics—tensile strength, Young's modulus, and thermal stability were all maintained. The materialisosterism of the designed core represents another highlight of this study. This enables the properties of plastics incorporating the degradable core to be predicted from those of the parent polymers. The “Core Block Technology” provides a unified platform for chemical recycling across diverse plastics using identical photocatalytic conditions. Recovered monomers were repolymerized, establishing formal closed‐loop recycling. Simple optimization of light irradiation geometry reduced reaction times from 48 to 2 h. Thus, this work offers a generalizable approach to chemically recyclable plastics without sacrificing material performance.

Author Contributions

Yasuyuki Nitta: methodology, data curation, formal analysis, validation, investigation. Ryosuke Nishikuma: investigation. Souya Kouda: investigation. Masatoshi Ohyama: investigation, methodology, data curation, formal analysis, validation. Meiki Shinoda: investigation. Tsuyoshi Taniguchi: formal analysis, supervision, data curation, validation, writing – original draft, writing – review and editing. Takashi Koike: validation, formal analysis, data curation, supervision, writing – review and editing, writing – original draft. Takashi Nishikata: conceptualization, funding acquisition, project administration, writing – review and editing, writing – original draft, data curation, supervision, formal analysis, validation.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Experimental details and characteristics of compounds are indicated in Supporting Information.Supporting File: anie73948‐sup‐0001‐SuppMat.pdf.

Data Availability Statement

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