To achieve performance and sustainability goals in material design, there are two scientific requirements: (1) the optimal ratio of cleavage group incorporation to maintain desired properties, and (2) ensuring homogeneous distribution of cleavage groups along the polymer backbone to prevent the formation of large microplastic fragments during degradation. Therefore, the reactivity of corresponding monomers in radical polymerization is crucial for controlling polymerization kinetics and tailoring molecular architecture. The photochemical [2+2] cycloaddition of succinimides with acetylene or 1,2-dichloroethylene has been reported to has been reported to efficiently produce cyclobutene-fused succinimides (Table 1).33,34 However, only activated cyclobutene-based monomers such as alkyl cyclobutenecarboxylates were reported to produce radical copolymers with n-butyl acrylate.35,36 Unfortunately, little was known about the reactivity of nonactivated cyclobutene-based monomers in radical polymerization reactions. Hence, we embarked on exploring the radical copolymerization of methyl acrylate (MA) with a cyclobutene-fused succinimide (CBS) comonomer. The CBS comonomer consists of a strained cyclobutene and a fused succinimide; the former may serve as the polymerization site moiety, while the latter acts as the preinstalled cleavable group. To establish a controllable polymerization system, the photo-iniferter RAFT polymerization of MA was initially conducted with a [MA]/[CTA] ratio of 200/1 in DMSO under blue-light irradiation. The kinetic plot exhibited a first-order relationship between ln([M]0/[M]t) over reaction time, strongly indicating of a constant radical concentration (Supplementary Fig. 6a). Despite variations in monomer conversions, the number average molecular weight (Mn,SEC) of the resulting polymethylacrylate (PMA) agreed well with theoretical values (Mn,NMR), yielding low dispersities (Ð = Mw/Mn < 1.08, Supplementary Figs. 6b and 6c).
Afterwards, photo-iniferter RAFT copolymerization of MA and CBS were conducted with different feed ratios (Table 1, entries 1-9). After precipitated from methanol, the composition of the purified product was determined by 1H NMR spectrum through comparison of the resonances of PMA segments and CBS units, respectively (Fig. 2a). To unambiguously verify the insertion of CBS, the diffusion-ordered spectroscopy (DOSY) NMR of the resultant products exhibited that all the characteristic signals shared single one diffusion coefficient with a sharp peak (Fig. 2b). Meanwhile, the size-exclusion chromatography (SEC) curve gave a monomodal distribution (Supplementary Figs. 8, 12, 14, 16, 21, and 24), indicating the formation of PMA bearing in-chain CBS units. With different comonomer feed ratio, a range of PMA copolymers incorporating CBS units (0.9-9.7 mol%) were synthesized to evaluate the impact of the comonomer content on the thermal properties. Differential scanning calorimetry (DSC) analysis of PMA exhibited a singular glass transition temperature (Tg) at 7.8 °C. The incorporation of CBS units augmented the thermal properties of PMA. With an increase of CBS incorporation ratio from 0.9 to 9.7 mol%, the glass transition temperatures rose from 19.5 to 38.8 °C (Fig. 2c).37 Thermogravimetric analysis (TGA) displayed a thermal decomposition profile closely mirroring that of commercial PMA materials, with a 5 wt% loss observed at 360-365 °C, comparable to their Td value of 355 °C (Fig. 2d). Consequently, the inclusion of CBS units does not compromise thermal stability but rather offers a means to modulate the thermomechanical properties of PMA. Furthermore, the synthesized copolymer in Table 1, entry 6 was subjected to treatment with trifluoroacetic acid water (TFA/H2O) at 40 oC for 24 hours. The SEC (Fig. 2e) and 1H NMR (Supplementary Fig. 44) analyses revealed no alternations, indicating that polyacrylates bearing in-chain CBS units displayed desired chemical stability with no unintended chain scission under acidic conditions, owing to the relative stability of cyclic imide.
To gain deeper insights into the impact of monomer composition on copolymerization kinetics and controllability, we closely monitored the evolution of monomer conversion and molecular weight over time for two distinct monomer feed ratios (MA : CBS = 9 : 1 and 7 : 3). When the feed ratio of CBS was higher, the conversions of MA and CBS both increased slower throughout the reaction (Fig. 3a, Supplementary Tables 2 and 3). As a typical controlled character of the copolymerization, a proximate first-order kinetic plots suggested constant radical concentrations during copolymerization, discernible from the slope (Fig. 3b, kobs, rate = kobs × [monomer], kobs= kp × [radical]). However, the polymerization rate increased with decreasing CBS fraction in the feed (kobs, rate = 0.61 and 0.33 h-1 for MA : CBS = 9 : 1 and 7 : 3, respectively). Furthermore, the linear relationships observed between the molecular weights and total monomer conversions of propagating polymers, along with monomodal SEC traces, indicated controlled polymerizations (Fig. 3c). Taking the advantage of photo-iniferter RAFT copolymerization, “ON/OFF” experiments were designed under a periodic process. At regular intervals, samples were withdrawn from the reaction mixture for 1H NMR and SEC analyses (Supplementary Figs. 29 and 30). Fig. 3d demonstrates that light exposure enables precise "ON/OFF" control over the conversions of MA with CBS simultaneously. Mn rises gradually with reaction time indicating the continuous chain propagation, which can be entirely terminated under dark conditions (Fig. 3e). The narrow dispersities of SEC traces demonstrated that our systems behaved perfect temporal control.
To resolve the monomer distribution and composition of the resultant copolymers, we determined the reactivity of MA and CBS by conducting a series of copolymerizations with varied monomer feed ratios (0.3 ≤ fMA ≤ 0.9) and low monomer conversion to prevent significant composition drift (<10%, Supplementary Table 5). After purification, the Mayo–Lewis plot was generated from the MA content in the copolymer (FMA) against the MA molar fraction in feed (fMA, Fig. 3f). It is evident that the MA content in copolymer surpasses the monomer feed ratio, indicating a preference for MA insertion over CBS. The value of rMA was calculated as 19.5, which compares with the value 18.1 and 18.8 Fineman-Ross (FR) and Kelen-Tüdös (KT) methods, respectively (Supplementary Figs. 32 and 33). Nevertheless, the value, rCBS = 0, suggests that the consecutive radical insertions of CBS are assumed to be impossible, which is consistent with the poor radical homopolymerization reactivity of CBS (Table 1, entry 9). Additionally, copolymerizations with high monomer conversions (feed ratio: MA : CBS = 7 : 3) demonstrated consistent incorporation ratios of CBS units (Table 1, entries 3-6), implying that the CBS units are evenly distributed along the PMA chain. This even distribution is crucial for maximizing the influence of cleavage groups from the degradable moieties.
Inspired by these results, we expanded the applicability of this methodology beyond MA by exploring copolymerization with other vinyl monomers. Acrylate derivatives such as n-butyl acrylate (nBA), benzyl acrylate (BnA), and N,N-dimethylacrylamide (DMA) were all able to copolymerize with CBS in high reactivity, resulting in the formation of corresponding polymers containing in-chain CBS units (Table 1, entries 10-12). Although similar CBS incorporation ratios were observed, the copolymerization with BnA gave the highest molecular weight (Table 1, entry 11, Mn = 103.5 kg/mol). Unfortunately, low polymerization reactivity and no CBS incorporation were observed in the case of styrene (Sty), possibly due to the radical chain termination after radical addition to a CBS monomer (Table 1, entry 13).
Cyclobutane-fused units on polymer backbone have the potential to undergo force-induced cycloreversion, leading to the formation of unsaturated linear polymers.38-42 Although the force-induced cycloreversion of CBS units have not been reported so far, efforts were taken to study the mechanochemical cycloreversion reactivity of the obtained polyacrylate copolymers. Ring-opening of cyclobutane were explored through pulsed ultrasonication of the THF solutions at 0 oC. Aliquots were removed at regular time intervals and analyzed by both 1H NMR and SEC spectra (Supplementary Figs. 46, 47, 49 and 50). As sonication time increased, 1H NMR analysis revealed a continuous cycloreversion of cyclobutane, yielding linear acyclic imide segments (Fig. 4a). The rate of ring-opening exhibited a dependency on molecular weight, which is contributed to the nature of mechanochemical activation.43 Meanwhile, decrease in molecular weight was observed due to elongational forces causing chain scission (Fig. 4b).
As sonication time increased, the activation of units and the occurrence of unintended chain scission both escalated. To assess the balance between mechanical activation of CBS and chain scission, we introduced a function of the scission cycle (SC, eq. 1).43-47
Where, Mn,0 is the initial molecular weight, Mn,t is the current molecular weight after specific sonication time, the slope of the obtained plot ϕi serves as a quantitative parameter of the competition between mechanophore activation and polymer chain scission. A higher slope (larger ϕi value) indicates a greater degree of mechanophore reactivity under sonication conditions, implying more ring-opening events per scission cycle. Notably, copolymer with two distinct molecular weights showed similar ϕi values (Fig. 4c, Mn = 47.4 kg/mol, ϕi = 0.29; Mn = 63.2 kg/mol, ϕi = 0.28), which is higher than the previous observations by Bowser et al. with copolymer bearing in-chain alkyl cyclobutanecarboxylate units (Mn = 114 kg/mol, ϕi = 0.24).36 This result indicates that cyclobutane-fused succinimide mechanophores are more responsive to the reported alkyl cyclobutanecarboxylate mechanophores, which is beneficial for efficient backbone editing via mechanochemical cycloreversion.
The activation of CBS units leads to the unveiled acyclic imide groups into the polymer backbone (Fig. 4d, red line). It is known that acyclic imides exhibit substantially greater activity in hydrolysis compared to cyclic imides.48 The sonication activated products were then subjected to a solution of TFA/H2O (10/1 v/v) for 24 h at 40 oC. The 1H NMR spectrum of crude product revealed the absence of acyclic imide groups, which indicates the successful hydrolysis of the polymer backbone. After precipitation, the 1H NMR of polymeric material only showed the unactivated CBS units, similarly with the initial polymers. The SEC trace consistently shifts to longer elution time and the relative intensity of oligomer fraction increased obviously (ca. 4.7 kg/mol, Fig. 4e). The oligomers were collected and analyzed by high-resolution electrospray ionization mass spectrometry (HR-ESI-MS), and three types of bifunctional molecules were detected: diacid, diamide, and w-amide-a-acid telechelic oligomers (Supplementary Tables 6 and 7), which constitute important intermediates in recycling materials, such as preparation of polyamides and polyesters.
Significantly, it is also possible to achieve bulk mechanochemical activation of the obtained polyacrylate copolymers using ball-milling grinding. After 1.0 h of ball-milling, 1H NMR analysis revealed a 13% cycloreversion of the CBS units for the copolymer in Table 1, entry 6 (Fig. 4d, orange line). Meanwhile, a greater reduction in molecular weight (Fig. 4f, from 63.2 kg/mol to 19.6 kg/mol after 1.0 h) due to mechanically induced chain scission was observed comparing to that using ultrasonication activation (Fig. 4b, from 63.2 kg/mol to 30.5 kg/mol after 20 h). Moreover, hydrolysis of the unveiled acyclic imide groups under acidic conditions (10/1 v/v TFA/H2O) or basic conditions (0.5 M NaOH aq.) both led to evident degradation to low molecular weight oligomers (Fig. 4f, green and purple lines). This degradation ability holds significant importance, particularly when these materials are released into seawater, where they undergo prolonged exposure to mechanical forces from tides and waves.49