CO₂‐Derived Degradable Polythioester Adhesives From Proton‐Trap‐Assisted S/O Isomerization‐Driven Cationic Ring‐Opening Polymerization

ABSTRACT

The development of degradable pressure‐sensitive adhesives (PSAs) holds great promise for enabling the recycling of adhesive‐containing materials. However, existing degradable PSAs are often constrained by insufficient or moderate adhesion strength, reliance on fossil‐based feedstocks, and/or incomplete degradation. Addressing these challenges requires the development of new polymerization strategies and material designs. Herein, we report a proton‐trap‐assisted S/O isomerization‐driven cationic ring‐opening polymerization of the CO₂‐derived thionolactone 3,6‐diethyltetrahydro‐2H‐pyran‐2‐thione, which enables the well‐controlled synthesis of high‐molecular‐weight CO₂‐based (co)polythioesters. The resulting materials exhibit a unique combination of high and tunable peel strength (up to 16.43 N/cm), excellent optical clarity (over 96% transmittance, low haze and low yellowness index), and the ability to undergo either complete degradation or quantitative depolymerization under mild conditions, delivering the first practical example of polythioester‐based PSAs. This study establishes a sustainable platform for PSA design that integrates robust performance with full life‐cycle management, thereby advancing the utilization of CO₂‐derived materials and circular polymer design.

This work presents a proton‐trap‐assisted S/O isomerization‐driven cationic ring‐opening polymerization of the CO₂‐derived thionolactone DET, which enables the well‐controlled synthesis of high‐molecular‐weight CO₂‐based (co)polythioesters. The resulting materials exhibit a unique combination of high and tunable peel strength, excellent optical clarity, and the ability to undergo degradation and depolymerization, delivering the first practical example of polythioester‐based PSAs.

1. Introduction

Sustainability has become a central focus of contemporary polymer chemistry, where modern polymeric material design must go beyond functionality to encompass the entire lifecycle for achieving genuine sustainability [1, 2]. This paradigm shift requires an integrated approach that spans every stage, from the selection of renewable or low‐impact feedstocks [3, 4, 5] and energy‐efficient synthesis [6] to end‐of‐life management strategies such as reuse, upcycling, and controlled degradation [7, 8]. Despite the remarkable progress made in developing sustainable plastics [9, 10, 11, 12, 13, 14, 15] and elastomers [16, 17, 18, 19], sustainable adhesives [20, 21, 22, 23, 24], another crucial class of polymeric materials, remain largely underexplored.

Pressure‐sensitive adhesives (PSAs) are among the most widely used adhesive materials, with a global market value at $9.8 billion in 2020 and projected to reach $17.6 billion in 2030 [23]. Depending on specific application requirements, their peel strength can be tuned over a broad range [25], typically below 10 N/cm for labels, general adhesive tapes, consumer products, and display lamination [26, 27], and exceeding 10 N/cm for high‐performance uses such as electronic manufacturing, automotive, and structural bonding [28]. However, the predominantly carbon–carbon backbones of conventional acrylic‐based PSAs, coupled with the cross‐linked networks introduced to improve cohesion, lead to poor degradability and persistent adhesive residues [29, 30]. These residues contaminate recycling equipment and hinder the reuse of otherwise recyclable substrates such as plastics, paper, cardboard, and multifunctional display modules. This long‐standing “stickies problem” has become a major barrier to recycling of adhesive‐containing materials [31, 32]. A promising strategy to address this issue is the development of degradable PSAs capable of eliminating adhesive residues [33] (Figure 1A). Bates [29] and Roth [34] respectively incorporated disulfide and thioester groups into acrylic backbones via radical copolymerization, enabling the resulting acrylic‐based PSAs to degrade upon external triggering. But these acrylic PSAs show only moderate adhesive performance (peel strength of 4–6 N/cm), degrade merely into oligomeric fragments, and rely on fossil‐based backbones. Kwon and coworkers [35] further developed degradable acrylic PSAs through the radical ring‐opening polymerization (ROP) of vinylcyclopropane. But the preparation of crosslinked PSA samples typically requires relatively complex processing steps. Beyond acrylic‐based PSAs, polycarbonate‐ and polyester‐based systems have also been explored, featuring fully degradable carbonate or ester linkages. For instance, Grinstaff [36] reported polycarbonate‐based PSAs with moderate peel strengths comparable to commercial tapes, while Williams and coworkers [37] developed bioderived triblock copolyester PSAs through epoxide/anhydride/lactone copolymerizations, achieving tunable and high peel strengths of 4–13 N/cm. However, these systems typically require long degradation times or environmentally hazardous reagents, and the use of metal catalysts in polymerization may hinder their applicability in optical devices. Hillmyer [38] and others prepared polyester PSAs by the ROP of lactones, but high tackifier loading was demanded to achieve sufficient adhesion. Moreover, depolymerizable or chemically recyclable PSAs, highly desirable for closed‐loop recycling with substrate plastics, remain exceedingly rare.

FIGURE 1.

(A) Schematic overview of the reported degradable PSAs in previous work. (B) The high‐performance and degradable CO₂‐derived polythioester PSAs obtained by proton‐trap‐assisted S/O isomerization‐driven cationic ROP.

3‐ethylidene‐6‐vinyltetrahydro‐2H‐pyran‐2‐one (EVP), which can be synthesized in a single step with high yield and selectivity from CO₂ and butadiene [39, 40], has emerged as a very promising CO₂‐based monomer owing to its high CO₂ content [41, 42]. Considerable efforts have been devoted to transforming EVP and its derivatives into sustainable polymeric materials through well‐controlled polymerization strategies [43, 44]. Nevertheless, EVP shows very low ROP activity due to the low ring strain of its stable six‐membered disubstituted structure, rendering the process both thermodynamically and kinetically unfavorable [45, 46, 47, 48]. Although elimination of the conjugated double bonds improves the ROP ability, polymerizations still suffer from low monomer conversion and afford only low‐molecular‐weight polyesters even after prolonged reaction times [49, 50]. We recently developed a dual isomerization‐driven cationic ROP [51], in which EVP was converted into a vinyl‐functional thionolactone EtVT, demonstrating that combined S/O and vinyl isomerization markedly enhances ROP activity and enables the rapid synthesis of high‐molecular‐weight CO₂‐based polythioesters. However, the presence of two distinct structural units within the backbone hampers depolymerization and chemical recycling. Moreover, although sulfur incorporation [52] is expected to enhance the properties of CO₂‐based polymers, its effect remains largely unexplored. Notably, sulfur‐containing backbones can substantially improve interfacial interactions [53, 54], while the inherent susceptibility of thioesters provides a built‐in platform for mild degradation and on‐demand depolymerization [34, 55, 56, 57, 58, 59].

Herein, we report a controlled cationic ring‐opening polymerization (CROP) of a CO₂‐derived thionolactone, 3,6‐diethyltetrahydro‐2H‐pyran‐2‐thione (DET), which is obtained from EVP through consecutive hydrogenation and thiocarbonylation [51]. A previous study revealed that the CROP of DET was limited by poor control and low molecular weights. In this work, we demonstrate that the addition of 2,6‐di‐tert‐butylpyridine (DTBP) as a proton trap effectively suppresses side reactions, thereby enhancing control and enabling the synthesis of high‐molecular‐weight CO₂‐derived polythioesters (Figure 2B). This DTBP‐assisted CROP of DET offers the following advantages. First, DET features a high theoretical CO₂ utilization, and this CROP proceeds efficiently using simple organic catalysts. The high controllability also enables copolymerization with rigid thionolactone for property tuning. Second, the resulting (co)polythioesters serve as the first polythioester‐based PSAs, offering tunable and good adhesive performance (peel strength up to 16.43 N/cm), high thermal stability, and excellent transparency without requiring a tackifier. Third, the embedded thioester groups enable complete PSA degradation under mild conditions. Finally, the near‐quantitative depolymerization can also be conducted on‐demand to recover these PSAs to thiolactone feedstock. This study presents a robust strategy for transforming CO₂ into useful polymers and offers a practical platform for designing sustainable, degradable/recyclable, and high‐performance PSAs, addressing long‐standing challenges.

FIGURE 2.

(A) ¹H NMR spectrum of produced PDET. (B) ¹³C NMR spectrum of produced PDET. (C) Overlay of SEC profiles at different [DET]₀/[[Et₃O]⁺[B(C₆F₅)₄]^–]₀/[DTBP]₀ ratios. (D) Plots of M _n and Đ as a function of [DET]₀/[[Et₃O]⁺[B(C₆F₅)₄]^–]₀/[DTBP]₀ ratio. (E) Pyridine‐based proton traps investigated in this study.

2. Results and Discussion

Starting from CO₂ and 1,3‐butadiene, EVP was first synthesized, followed by hydrogenation and subsequent thiocarbonylation with phosphorus pentasulfide (P₂S₅) to afford the DET thionolactone. The CROP of DET was first conducted with the commonly used Lewis acid BF₃•Et₂O as the catalyst at room temperature in dichloromethane (Entry 1 in Table 1, [DET]/[BF₃•Et₂O] = 100/1, [DET] = 2 M). As the reaction proceeded, a gradual increase in viscosity was observed, indicating polymer formation. After quenching the polymerization, the structure of the polymer was fully characterized by ¹H NMR and ¹³C NMR (Figure 2A,B). The characteristic resonance of the carbonyl groups at 203 ppm clearly indicated the complete S/O isomerization during the ring opening process and the formation of polythioester. However, the ¹H NMR spectrum of the quenched reaction solution showed that the monomer conversion was only 47% within 8 h, revealing the relatively poor polymerization activity of DET. Moreover, the molecular weight of PDET significantly deviated from the theoretical value, suggesting poor initiation efficiency of BF₃•Et₂O (Figures S1 and S2). Given these unsatisfactory results, we speculated that more active onium salt catalysts bearing non‐coordinating/non‐nucleophilic counterions could enhance propagation rates and monomer conversion. Specifically, Hong's group [60, 61] previously conducted the cationic ROP of γ‐thionovalerolactones and obtained high‐performance polythioesters by using the robust [Ph₃C]⁺[B(C₆F₅)₄]^– and [Et₃O]⁺[B(C₆F₅)₄]^– cationic catalysts, which can easily form free/loose ion pair active species during the polymerization process due to the bulky, weakly coordinating of [B(C₆F₅)₄]^–.

TABLE 1.

Results of cationic ROP of DET ^{a .}

Entry	Catalyst	[M]0/[Cat]0/[DTBP]0	Time (h)	Conv. (%) b	M n,theo (kDa) b	M n,SEC (kDa) c	Ð c
1	BF3·Et2O	100/1/0	8	47	8.1	26.2	1.19
2	Ph3C+B(C6F5)4 –	100/1/0	1	95	16.6	15.0	1.05
3	Et3O+B(C6F5)4 –	100/1/0	1	87	15.0	15.2	1.08
4	Et3O+B(C6F5)4 –	200/1/0	4	90	31.0	25.1	1.21
5	Et3O+B(C6F5)4 –	300/1/0	6	89	46.0	32.5	1.23
6	Et3O+B(C6F5)4 –	400/1/0	8	89	61.3	36.5	1.30
7	Et3O+B(C6F5)4 –	500/1/0	12	92	79.2	45.9	1.36
8	Et3O+B(C6F5)4 –	200/1/0.3	4	87	30.0	34.1	1.25
9	Et3O+B(C6F5)4 –	300/1/0.3	6	85	43.9	39.7	1.24
10	Et3O+B(C6F5)4 –	400/1/0.3	10	84	57.9	51.1	1.30
11	Et3O+B(C6F5)4 –	500/1/0.3	12	85	73.2	70.2	1.36

^a Polymerizations were conducted at room temperature in DCM, [M]₀ = 2 M.

^b Conversions and M _n,theo were calculated by ¹H NMR analysis of the quenched solution in CDCl₃.

^c Determined by SEC at 40°C in THF with a PSt standard.

Consistent with our hypothesis, both [Ph₃C]⁺[B(C₆F₅)₄]^– and [Et₃O]⁺[B(C₆F₅)₄]^– onium catalysts displayed significantly enhanced catalytic activity compared to BF₃•Et₂O. Under identical conditions ([DET]/[catalyst] = 100/1), DET conversions reached 95% and 87% within 1 h, respectively (Entries 2 and 3 in Table 1, Figures S3 and S4). Notably, the resulting polymers exhibited molecular weights close to the theoretical values and narrow dispersities (Đ < 1.10). The chain‐end fidelity of PDET was further confirmed by matrix‐assisted laser desorption/ionization‐time of flight mass spectroscopy (MALDI‐TOF MS) analysis of a low‐molecular‐weight sample initiated by [Et₃O]⁺[B(C₆F₅)₄]^– and quenched with MeOH. As shown in Figure S5, the interval between the adjacent molecular ion peaks is consistent with the monomer mass, which confirms the definite structure of PDET with an α‐end group corresponding to the mass of CH₃CH₂‐ and an ω‐end group corresponding to ‐OCH₃, respectively. Actually, non‐coordinating/non‐nucleophilic counterions generally reduce the controllability of polymerization while enhancing the polymerization activity. The good controllability and low distribution in the polymerization with [DET]/[Catalyst] = 100/1 may be attributed to the S_N2 ring‐opening mechanism, where the C = S of monomer has a strong interaction with the cationic active center to suppress side reactions [60]. However, when the initial DET‐to‐catalyst ratio was increased to 200–500/1, the size exclusion chromatography (SEC)‐determined molecular weights were substantially lower than the theoretical values despite high monomer conversions (Entries 4–7 in Table 1, Entries 1–4 in Table S1, Figures S6–S13). We supposed that the deviation from the theoretical molecular weight was likely caused by trace acidic species in the system, which may originate from adventitious water and normally led to unfavorable initiation and chain transfer side reactions at high monomer‐to‐initiator ratios (low initiator and propagation species concentration) [62, 63]. This phenomenon is common in cationic addition polymerization and ring‐opening polymerization, and the addition of proton traps usually helps to suppress side reactions. However, whether such additives can also promote the controlled CROP of thionolactones has not yet been reported. Pyridine and its derivatives are often employed as proton traps to achieve controlled cationic polymerization, among which 2,6‐di‐tert‐butylpyridine (DTBP) exhibits exceptional proton scavenging selectivity and minimizes unfavorable interactions with Lewis acids and carbocationic active centers owing to its distinctive steric hindrance [64]. We therefore conducted the polymerizations with the addition of DTBP as a proton trap. We tried different ratios of DTBP to catalyst to evaluate its regulatory effect on polymerizations and molecular weights. At low DTBP loadings, the polymer molecular weights increased markedly; however, excessive DTBP inhibited polymerization (Entries 5–8 in Table S1), likely because this strong base competes with monomers for coordination to the cationic active species, thereby lowering monomer conversion [64]. An optimal DTBP loading of 0.3 equiv relative to the catalyst was thus identified. Using this optimized condition, the ROP of DET proceeded smoothly, and the molecular weights of PDET increased linearly with the ratio of monomer to catalyst, indicating well‐controlled polymerization (Figure 2C,D; Entries 8–11 in Table 1). Although the M _n values were determined by SEC using polystyrene standards and may therefore deviate somewhat from the true molecular weights of these sulfur‐containing polymers, the linear trend clearly supports the excellent controllability of the polymerization. This stands in stark contrast to the uncontrolled behavior observed without DTBP (Figure S14). Remarkably, at an initial ratio of [DET]₀/[[Et₃O]⁺[B(C₆F₅)₄]^–]₀/[DTBP]₀ = 500/1/0.3, a high‐molecular‐weight PDET (M _n = 70.2 kDa) was obtained. To further elucidate the role of DTBP in the polymerization, other pyridine‐based compounds were also evaluated (Figure 2E and Table S2). Pyridine, 2,6‐lutidine, and 2,4,6‐collidine, which possess lower steric hindrance, were likewise able to increase the molecular weight of PDET under the optimized conditions, but afforded polymers with slightly broader dispersities than those obtained using DTBP. In contrast, a more sterically hindered base, 2,6‐bis(1,1‐diphenylethyl)pyridine (2,6‐DPEP), which was very recently reported in the literature during the review process of this work [65], was also examined. However, it showed a less satisfactory effect, likely because excessive steric hindrance is unfavorable for efficient proton scavenging during DET polymerization. These comparisons indicate that DTBP, with its moderate steric hindrance, is the most effective pyridine‐based additive for DET polymerization through a proton‐trapping mechanism. Taken together, these results establish a new DTBP‐assisted CROP of CO₂‐derived thionolatone, enabling access to high‐molecular‐weight polythioesters with excellent control.

The thermogravimetric analysis (TGA) of PDET shows a single‐step mass loss with an onset decomposition temperature (T _d, 5% weight loss) of 197.2°C, indicative of moderate thermal stability (Figure S15). Differential scanning calorimetry (DSC) further reveals that PDET is amorphous, exhibiting a low glass‐transition temperature (T _g) of –23.8°C (Figure S16). The low T _g and good thermal stability of PDET make it highly promising for use as a PSA. Therefore, the 180° peel strength test was carried out to measure the peel strength of PDET under GB/T 2792‐2014 standard (Figures 3A and S17). A stainless steel plate was utilized as the rigid substrate, and polyethylene terephthalate (PET) film was used as the face substrate. Remarkably, PDET exhibited a high peel strength of 6.62 N/cm, significantly exceeding that of commercial 3 M Scotch tape (∼3.48 N/cm), demonstrating its potential as a novel CO₂‐based PSA (Entry 1 in Table 2, Figure 3B). However, cohesive failure occurred during peeling, leaving sticky residues on the substrate (Figure 3C). This behavior is likely due to the insufficient cohesive strength of PDET during the debonding process [66]. An ideal PSA should be soft and fluid‐like during bonding to ensure complete surface wetting, yet stiffen markedly under rapid deformation while maintaining high viscosity and stretchability. To achieve this balance, high‐T _g functional monomers are often incorporated as comonomer units to enhance the cohesive strength of the polymer.

FIGURE 3.

(A) The 180° peel strength test model. (B) The curves of 180° peel strength of PDET and 3 M Scotch tape. (C) Photo of the substrate surface after the debonding process of PDET.

TABLE 2.

The adhesive and rheological properties of the resulting PSA samples.

Entry	DOT content (mol%) a	M n,SEC (kDa) b	G′ at 25°C (MPa) c	G″ at 25°C (MPa) c	T g (°C) d	Peel strength (N/cm) e
1	0	49.2	0.126	0.058	−23.7	6.62 (±0.16)
2	9.9	35.0	0.094	0.037	−13.4	16.43 (±0.32)
3	17.4	39.4	0.089	0.050	−7.2	11.47 (±0.33)
4	20.6	38.6	0.102	0.068	−5.6	8.06 (±0.18)

^a DOT contents were calculated by ¹H NMR analysis of the copolythioesters in CDCl₃.

^b Determined by SEC at 40°C in THF with a PSt standard.

^c Storage modulus and loss modulus at 25°C and 1 Hz (∼6.3 rad/s), measured using the frequency sweep method.

^d Determined by DSC.

^e Peel strength values were obtained from 3–5 measurements.

Based on the successfully achieved DTBP‐assisted well‐controlled CROP of DET, we next explored its copolymerization with suitable comonomers to further tailor PSA properties. In our recent work [51], we reported the first example of CROP between distinct thionolactones and demonstrated a tunable transition in material properties from plastic to elastomer, underscoring the effectiveness of this strategy for property modulation. Guided by this insight, we investigated the cationic copolymerization of DET with the biaryl‐fused thionolactone dibenzo[c,e]‐oxepine‐5(7H)‐thione (DOT). DOT is a well‐studied thionolactone known for preparing degradable vinyl polymers through radical polymerizations with vinyl monomers [67, 68, 69]. The homopolymer of DOT is rigid with a high T _g of 98°C [70, 71], making it very suitable for copolymerization with DET. We anticipated that incorporating this rigid DOT unit into the polymer backbone would impart enhanced and well‐balanced adhesive properties. Using [Ph₃C]⁺[B(C₆F₅)₄]^– as the catalyst, we performed copolymerizations at various initial [DET]₀/[DOT]₀ ratios (Figure 4A). Encouragingly, the reactions proceeded smoothly and achieved nearly quantitative conversion of both monomers. The ¹H NMR spectrum confirmed that the molar incorporation ratio of DET and DOT closely matched the initial feed composition (Figure 4B). Moreover, the unimodal SEC trace and the uniform diffusion coefficient observed in the DOSY NMR spectrum provided strong evidence for the successful formation of true DET–DOT copolymers (Figure 4C,D). We next carried out a kinetic study on the copolymerization of DET and DOT. Using an initial DET:DOT feed ratio of 1:1, the reactivity ratios were determined by fitting the conversion data with a nonterminal model (Figures S18 and S19). The resulting values, r _DET = 2.632 and r _DOT = 0.438, indicate that DET is incorporated more rapidly than DOT, leading to a relatively gradient copolymer. The slightly low ring strain of DOT [72] and its greater steric hindrance likely retard its incorporation relative to DET. Thanks to the well‐controlled DTBP‐assisted CROP and the near‐quantitative monomer conversions, the copolymer composition can be readily tuned by simply adjusting the initial feed ratios, enabling the facile preparation of P(DET‐co‐DOT) copolythioesters with DET molar contents ranging from 100 mol% to 50 mol% (Figures S20–S31). The incorporation of DOT resulted in a slight improvement in thermal stability (Figure S32). Notably, DSC analysis showed that all P(DET‐co‐DOT) copolythioesters were amorphous, exhibiting T _gs ranging from –13.4 to 17.4°C (Figures S33–S39), with T _g increasing as the DOT content increased. Importantly, copolymers with DET molar contents above 79.4% displayed T _g values below –5°C, falling within the desirable range for general PSA applications.

FIGURE 4.

(A) Well controlled cationic CROP of DET and DOT. (B) The ¹H NMR spectrum of produced copolythioester of DET with DOT ([DET]/[DOT] = 1/0.1 in feed). (C) SEC curve of produced copolythioester of DET with DOT ([DET]/[DOT] = 1/0.1 in feed). (D) ¹H DOSY NMR spectrum of produced copolythioester of DET with DOT.

Building on the above copolymerization results, the adhesive performance of the P(DET‐co‐DOT) series was systematically evaluated. As shown in Figure 5A, introducing DOT into the PDET backbone significantly enhanced adhesion. The peel strength increased markedly with the introduction of DOT and reached a maximum value of 16.43 N/cm at 9.9 mol% DOT, surpassing the performance of most previously reported degradable PSAs and those PSAs made from renewable raw materials, which typically exhibit peel strengths below 10 N/cm. However, at this composition, the samples still displayed cohesive failure, indicating insufficient internal strength during debonding. Strikingly, further increasing the DOT content not only modulated cohesion but also enabled residue‐free debonding. At 17.4 mol% DOT, the PSA transitioned from cohesive to complete adhesive failure while still maintaining an impressively high peel strength of 11.47 N/cm (Figure S40), again outperforming the majority of degradable PSA systems. Increasing the DOT content to 20.6 mol% preserved clean debonding, with the peel strength remaining substantial at 8.06 N/cm. The slight decrease in adhesion at this composition likely arises from reduced interfacial interactions as the material becomes stiffer. In addition to adhesive strength, the creep performance of PSAs is also of great significance for their applications. To evaluate creep resistance, creep–recovery experiments were performed on the PSA samples by applying a constant stress of 20 kPa for 3 min, followed by a 9 min recovery period. As shown in Figure S41, PDET exhibits a very large creep strain and only limited recovery, with a residual strain exceeding 1600%, indicative of substantial irreversible deformation. With increasing incorporation of the rigid DOT comonomer, the creep strain decreases markedly, while the recoverable strain increases significantly. In particular, the PSAs containing 17.4 and 20.6 mol% DOT exhibit residual strains of only 69% and 31%, respectively, demonstrating substantially resistance to creep‐induced deformation. The improvement in cohesive performance was further confirmed by static shear‐holding tests. A 25 mm × 25 mm area of PSA film was applied to a stainless‐steel substrate and subjected to a 1 kg load. Neat PDET failed within approximately 1 h, reflecting its low cohesive strength and susceptibility to flow under sustained stress. In contrast, when the DOT content was increased to 17.4 mol% and 20.6 mol%, the shear failure time was significantly prolonged to 138 h and more than 240 h respectively, demonstrating that DOT effectively enhances both the cohesive strength and shear resistance of the PSA (Figure S42). In conventional PSAs, such creep and shear resistance are typically achieved through covalent crosslinking. Remarkably, linear P(DET‐co‐DOT) system in this work exhibits comparable resistance without crosslinking. This improvement likely arises not only from the increased chain rigidity imparted by the DOT units, but also from possible π–π stacking interactions between the DOT‐derived biphenyl moieties, which may strengthen interchain associations and suppress chain slippage under load. To further elucidate the viscoelastic behavior of the copolythioester‐based PSAs, rheological measurements were carried out using a rheometer. The storage (G′) and loss (G″) moduli were collected from frequency sweep experiments (Figures S43–S46), and the resulting viscoelastic windows are presented in Figure 5B. All samples exhibited viscoelastic windows that fell entirely within Chang's window, and their bonding moduli (i.e., the bases of the windows) were well below the Dahlquist criterion (G′ ≤ 3 × 10⁵ Pa) [73], confirming that these polythioesters behave as typical PSAs capable of good interfacial adhesion. Neat PDET displayed a sufficiently low G′ at bonding frequencies, enabling excellent substrate wetting. However, its overly soft chains provided limited shear resistance, resulting in only moderate peel strength and cohesive failure with residual adhesive remaining after detachment. These deficiencies were effectively mitigated by incorporating an appropriate amount of rigid DOT units. As the DOT content increased, the viscoelastic windows progressively shifted toward the upper‐right region (i.e., the high‐shear PSA domain), consistent with the corresponding increases in T _g. Taken together, the adhesion measurements and rheological analysis reveal that incorporating a small amount of DOT provides an effective lever to simultaneously enhance interfacial adhesion and regulate cohesion, ultimately enabling CO₂‐derived polythioester PSAs that deliver both high peel strength, reasonable creep resistance, and clean, residue‐free debonding, an uncommon yet highly desirable performance profile among degradable PSA systems.

FIGURE 5.

(A) 180° peel strength results for various PSA samples and literature degradable PSAs. (B) The viscoelastic window of CO₂‐based PSAs with different DOT contents. (C) UV−vis transmission spectra of the PSAs with different DOT contents. (D) Haze and yellowness index of the PSAs with different DOT contents.

To the best of our knowledge, this work represents the first example of a polythioester‐based PSA. The enhanced adhesion performance arises from the stronger interfacial interactions between thioester groups and substrates compared with the ester moieties in conventional acrylic PSAs and other ester‐containing systems [29, 53], as well as from the tunable incorporation of rigid units. Traditionally, polythioesters are prepared via the anionic ROP of thiolactones, a process that typically requires specialized catalysts, offers limited controllability, and renders copolymerization between structurally distinct thiolactones particularly challenging. In contrast, this DTBP‐assisted CROP developed in this study provides an efficient and well‐controlled copolymerization pathway, owing to the unique electrophile‐reactive behavior of the C = S group of thionolactone. This enables the synthesis of copolythioesters with precisely tunable compositions. This capability is key to achieving the high‐performance PSA here. Moreover, besides general‐purpose applications, functional PSAs are also used in high‐value fields such as optically clear adhesives, which, in addition to the general requirements of PSAs, also need high transparency, low haze, and low yellowness. For example, Kwon and coworkers developed a cleanly removable and degradable bio‐based adhesive for flexible displays [30]. In this study, the high copolymerization activity affords CO₂‐based polythioester PSAs that are inherently colorless, exhibiting over 96% transmittance at wavelengths above 380 nm (Figure 5C), low haze (< 1.00%) and low yellowness index (< 0.5) (Figure 5D), indicating their good optical properties and promising potential for display applications.

Another advantage of these (co)polythioester PSAs lie in the intrinsically higher degradability of thioester linkages compared with conventional ester and carbonate bonds, enabling efficient breakdown under much milder conditions. This feature offers a gentle yet practical solution to the long‐standing challenge of poor degradability in conventional PSAs. To demonstrate this capability, we investigated the methanolysis of a representative P(DET‐co‐DOT) PSA in a dilute potassium hydroxide solution (0.5 M) at 80°C with stirring for 2 h (Figure 6A). Under these conditions, the PSA sample, which is otherwise insoluble in methanol, was completely dissolved, indicating extensive main‐chain cleavage into sufficiently small fragments. SEC analysis further confirmed the absence of any detectable high‐molecular‐weight species, evidencing complete degradation. After simple purification, the two corresponding thiol‐containing methyl ester derivatives were readily isolated in a combined yield of 82% (Figures S47 and S48). These products may serve as valuable new building blocks for the synthesis of sustainable materials.

FIGURE 6.

(A) Illustration of the methanolysis of the polythioester‐based PSA. (B) ¹H NMR spectra of the PDET (top) and the depolymerization product (bottom). (C) ¹H NMR spectra of the original copolymer (top) and the crude depolymerization products with the molar ratio of DET/DOT = 1/1 (bottom).

Beyond excellent degradability, we further hypothesized that PDET and P(DET‐co‐DOT) PSAs could undergo on‐demand depolymerization to enable chemical recycling. To evaluate this possibility, we first examined the depolymerization behavior of PDET. Depolymerization was carried out in solution at 160°C using zinc chloride (ZnCl₂) as the catalyst. Under these conditions, complete depolymerization was readily achieved within 12 h, and the resulting ¹H and ¹³C NMR spectra confirmed the formation of the corresponding thiolactone (Figure 6B). To further reduce the energy required for depolymerization, we next employed the organic superbase 1,5,7‐triazabicyclo[4.4.0]dec‐5‐ene (TBD), a highly efficient catalyst widely used for polyester depolymerization. Gratifyingly, TBD significantly enhanced the depolymerization efficiency, enabling complete conversion to the same thiolactone product within 12 h at a much milder temperature of 75°C. Depolymerization of the copolythioester P(DET‐co‐DOT) was also examined under identical conditions. After cooling to room temperature, direct ¹H NMR analysis of the reaction mixture confirmed full depolymerization into the two thiolactones corresponding to the monomeric composition of the parent copolymer (Figure 6C), and these could be readily separated (Figure S49). Although the recycled CO₂‐based six‐membered disubstituted thiolactone cannot undergo ROP again due to its low ring strain, thiolactones are well known to undergo efficient nucleophilic ring‐opening with primary amines to generate thiols [74]. Importantly, the depolymerized CO₂‐derived thiolactone underwent quantitative ring‐opening with model amine compound 1‐(3‐aminopropyl)imidazole, demonstrating that the recovered monomers can be cleanly and fully transformed into value‐added thiol compounds, thereby establishing a viable closed‐loop recycling pathway for this polythioester system (Figures S50 and S51). Since chemically recyclable and closed‐loop plastics are increasingly demonstrating superior performance and sustainability [75, 76, 77], these depolymerizable poly(co)thioesters have the potential to serve as PSAs that can be recycled simultaneously with recyclable substrate plastics, thereby simplifying the overall recycling process and reducing associated costs.

3. Conclusion

In summary, we have developed a proton‐trap‐assisted S/O isomerization‐driven CROP of the CO₂‐derived thionolactone DET. Enabled by simple onium salt catalysis, S/O isomerization and the assist of DTBP effectively suppress the side reactions and contribute to the well‐controlled synthesis of high‐molecular‐weight CO₂‐based polythioester with M _n up to 70.2 kDa. The high controllability further enables copolymerization with the rigid thionolactone DOT, providing an effective handle for fine‐tuning material properties. As a result, the resulting (co)polythioesters exhibit a unique combination of high and tunable peel strength (up to 16.43 N/cm), excellent optical clarity (over 96% transmittance, low haze and low yellowness index), and the ability to undergo either complete degradation or quantitative depolymerization under mild conditions, delivering the first practical example of polythioester‐based PSAs. Overall, this study presents a robust strategy for transforming CO₂ into useful polymers and offers a practical platform for designing sustainable, degradable/recyclable, and high‐performance PSAs, thereby addressing long‐standing challenges in adhesive recycling and circular polymer design. Our future research is focused on creating CO₂‐based PSAs with broader tunability of viscoelastic properties and T _g, particularly toward softer adhesive regimes.

Author Contributions

Zong‐Bin Lu: investigation, methodology. Yu Xiong: investigation. Guang Chen: investigation. Li‐Min Wu: writing – original draft, investigation. Xuan Nie: investigation. Lei Xia: investigation. Chun‐Yan Hong: investigation, funding acquisition, writing – original draft. Ze Zhang: conceptualization, investigation, methodology, writing – original draft, writing – review and editing, supervision, funding acquisition. Ye‐Zi You: writing – original draft, writing – review and editing, funding acquisition, supervision, conceptualization, investigation, methodology.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: anie73197‐sup‐0001‐SuppMat.pdf.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.