7-Oxa-2,3-Diazanorbornene: A One-Step Accessible Monomer for Living Ring-Opening Metathesis Polymerization to Produce Backbone-Biodegradable Polymers

Introduction

The development of living polymerization enables the synthesis of precision polymers with well-controlled structures and properties.^1,2 Among the diverse polymer properties, backbone biodegradability holds significant importance in the fields of environmental sustainability, biomedicine, and nanoetching materials.^3–5 However, using living polymerization tools to precisely control the distribution of the biodegradable functional groups poses considerable challenges, as traditional living polymerization reactions give completely biodegradable backbones (e.g., polyesters, polyamides, and polyacetals) or nondegradable backbones (e.g., vinylic addition polymers). Although living polymerization reactions can generate block copolymers (BCPs) with different structures and functions, the covalent-tethering of degradable and nondegradable backbones usually requires complex chain end control and the use of macroinitiators to switch polymerization techniques.^6,7

Olefin metathesis reactions, especially ring-opening metathesis polymerization (ROMP) reactions using commercially available Ru-based precatalysts, are highly versatile approaches for synthesizing functionalized macromolecular materials with diverse applications ranging from drug delivery to large-scale engineering plastics.^8–11 The ROMP reactions that feature living polymerization characteristics enable the precise control of polymer architectures, such as linear, block, bottlebrush, and star shapes.^12,13 ROMP polymers are intrinsically degradable due to the presence of alkenes that are cleavable under metathesis and strong oxidation conditions.¹⁴ To make the polymers capable of undergoing degradation into smaller fragments under physiological conditions, diverse heteroatoms and cleavable linkages need to be introduced,^15,16 such as silicon,^17–21 phosphorous,^22–24 acetal/ketal,^25–32 vinyl ether,^33–42 ester,^43–45 carbonate,⁴⁶ and hemiaminal ether.^47,48 However, mostly due to low to moderate ring strain (Figure 1a),⁴⁹ when making BCPs containing biodegradable components, these monomers cannot form “pure homo-blocks” by living ROMP of themselves. They can only be used as sacrificial “end blocks,”^24–29 or be copolymerized with highly reactive alkenes to form nearly “alternating blocks.”^{17,34,35,37–40}

• Download figure
• Download PowerPoint

The insufficient ring strain in previously reported biodegradable ROMP monomers can be regarded as a compromise of structural design to the synthetic feasibility,¹⁶ as formation of strained rings usually requires more energy input to overcome the enthalpy penalty in a highly strained system. To overcome this challenge, we are inspired by norbornenes, which are the most common living ROMP monomers bearing [2.2.1]-bicyclic motifs and can be easily accessed through Diels–Alder (DA) reactions. It is hypothesized that if multiple heteroatoms are used to substitute norbornenes, the living ROMP feature and biodegradability can be simultaneously achieved. More specifically, as shown in Figure 1b, 7-oxa-2,3-diazanorbornene (ODAN) consisting of one oxygen atom and two nitrogen atoms within the [2.2.1]-scaffold is utilized for living ROMP to form structurally well-defined biodegradable polymers. Like norbornenes, highly strained ODAN derivatives can be prepared through a one-step solvent-free hetero-DA cycloaddition reaction using commercial furan and azodicarboxylate as starting materials.⁵⁰ More importantly, the resulting ROMP polymers are demonstrated to have hydrolyzable hemiaminal ether groups, and the hydrolysis rates are tunable by varying the substituents on ODAN derivatives. This approach allows for the synthesis of diverse biodegradable ROMP multiblock and statistical copolymers that possess well-controlled biodegradable sequence, controlled nanoetching morphology, controlled fragments after degradation, and controlled degradation rates.

Experimental Methods

Materials and characterization

Materials and characterization by nuclear magnetic resonance (NMR) ( Supporting Information Figures S57–S74), gel permeation chromatography (GPC), thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), scanning electron microscope (SEM), and small-angle X-ray scattering (SAXS) involved in this work are all represented in the Supporting Information.

Monomers synthesis

Diethyl azodicarboxylate (DEAD; 1.74 g, 10 mmol) and distilled furan (2.2 mL, 30 mmol) were placed in a 10 mL flask equipped with a magnetic stir bar, and the mixtures were stirred at 25 °C for 2.0 h until the color changed from orange red to colorless. The ODAN1 was obtained in quantitative yield as colorless oil without further purification after the excess furan was removed through evaporation. It is noteworthy that the monomer is unstable at room temperature, but it can be kept in the freezer (−20 °C) for six months without deterioration. The detail for the synthesis of ODAN2, ODAN3, and ODAN4 and the scalable synthesis of ODAN1 are all represented in the Supporting Information.

Homopolymers synthesis

In a N₂-filled glovebox, the desired amounts of monomers were added to 5 mL vials equipped with a magnetic stir bar, then anhydrous tetrahydrofuran (THF) was added to the vials. The required amount of Grubbs III catalyst solution in THF was then added to the vial and the concentration of monomers was ensured to be 0.5 M. The solution was stirred at 0 °C. When the polymerization was completed, the reactions were quenched with excess ethyl vinyl ether and the SiliaMetS DMT metal scavenger was added to remove the catalyst residue. After filtration, the polymers were obtained without further purification after the THF was removed through evaporation.

Penta-BCP synthesis

In a N₂-filled glovebox, NB1 (51 mg, 0.2 mmol, 50 equiv) were dissolved in 0.5 mL of THF. 0.5 mL of freshly prepared 1.5 mg Grubbs III catalyst in 0.5 mL THF was added. The mixture was stirred at room temperature for 0.5 h. Next, a solution of ODAN1 (48 mg, 0.2 mmol, 50 equiv) in 0.5 mL THF was added, and after the solution was stirred at 0 °C for 1.0 h, a solution of NB2 (41 mg, 0.2 mmol, 50 equiv) in 0.5 mL THF was added, where the resultant mixture was stirred at room temperature for 0.5 h. Then, a solution of ODAN4 (60 mg, 0.2 mmol, 50 equiv) in 0.5 mL THF was added and stirred at 0 °C for 1.0 h. Finally, a solution of NB3 (41 mg, 0.2 mmol, 50 equiv) in 0.5 mL THF was added, and the solution was stirred at room temperature for 0.5 h, quenched with excess ethyl vinyl ether, concentrated under vacuum, and the formation of penta-BCPs with five different components was finished. Details for the synthesis of copolymers and bottlebrush polymers are available in the Supporting Information.

Acidic hydrolysis of the polymers

Twenty milligrams of polymer were dissolved in 2 mL of THF, and 20 μL 1 M HCl was added to the solution. The mixture was stirred for 4 h. Excess sodium bicarbonate and sodium sulfate were added and the mixture was allowed to sit for 10 min. Finally, the mixture was extracted with dichloromethane (DCM), concentrated, and subjected to GPC analysis.

Casting and etching of the diblock copolymer films

Diblock copolymers P-NB1₁₀₀- b -ODAN1 _m films were prepared by dissolving the polymer (100 mg) in dimethylformamide (10 mL) and casting on a clean 6 cm × 6 cm glass plate with a glass dish to slow down the rate of solvent evaporation at 40 °C for 72 h. The final thicknesses of the films were between 90 and 120 μm. The diblock copolymer films were soaked in 1 M HCl methanol solution for 5 days. After that, the films were soaked and washed in methanol five times for 20 min each time. Finally, the residual methanol was removed through evaporation.

Degradation of bottlebrush polymers under buffered conditions

Ten milligrams of bottlebrush polymers prepared as described above were placed in a vial containing 2 mL of the requisite pH ≈ 7.4 phosphate-buffered saline (PBS) buffer. The solution was stirred at 25 °C for the indicated times. The water was removed through evaporation and the mixture was suspended in DCM, then excess sodium sulfate was added and the reaction mixture was allowed to sit for 5 min. Finally, the DCM solution was filtered with a 0.2 μm Nylon filter, concentrated under vacuum, dissolved in THF, and subjected to GPC analysis.

Results and Discussion

Controlled sequence

The hetero-DA cycloaddition reactions between furan and DEAD were first reported in the 1950s, and the cycloadduct ODAN1 can be produced under ambient temperatures and neat conditions.^50,51 The reaction conditions are much milder than traditional DA reactions to form norbornenes, because the diene component furan is electron-rich and the dienophile azodicarboxylate is electron-deficient.⁴² In this study, we revised the synthesis of ODAN1 to a scale exceeding 100 g ( Supporting Information Figure S1), characterized its rotational isomers by two-dimensional nuclear magnetic resonance (2D-NMR) analysis ( Supporting Information Figures S3–S11),⁵² and confirmed its stable storage at −20 °C for more than 6 months ( Supporting Information Figure S12). Other ODAN derivatives, ODAN2– ODAN4, were synthesized from other commercial azodicarboxylate compounds (Table 1) and stored in the same way as ODAN1. Notably, density functional theory (DFT) was utilized to estimate the ring-strain energy of ODAN1, which was found to be ∼22.8 kcal/mol ( Supporting Information Figure S2). This value was even higher than that of norbornene (15.8 kcal/mol when calculated using the same method),^17,53 showing apparent differences to previously reported biodegradable ROMP monomers.

When the ODAN monomers were polymerized using the Grubbs III catalyst in THF, complete monomer conversion and obvious living polymerization features were observed (Figure 2a,b, Table 1, and Supporting Information Table S1). The number average molecular weight (M_n) increased linearly with the monomer to initiator ratio ([M]/[I]), while the polydispersity (Đ) remained narrow (Figure 2b). At a [M]/[I] ratio of 100, the reactions performed at 0 and 25 °C gave similar M_n and Đ results (Table 1, entries 2 and 7). However, at a [M]/[I] ratio of 500:1, the ROMP of ODAN1 was better controlled at 0 than at 25 °C, possibly due to higher monomer stability at colder temperatures (Table 1, entries 6 and 8). The ODAN monomers are prone to slowly undergo retro-DA reactions in solution at room temperatures, which would regenerate furan and azodicarboxylate.⁵⁰ It was confirmed that the presence of furan had minimal influence on the living ROMP reaction (Table 1, entry 9). However, the presence of DEAD led to incomplete monomer conversion and significantly broadened polymer distribution (Table 1, entry 10). Nevertheless, the M_n of the biodegradable homopolymers reached 71.2 kDa, while the polydispersity remained narrow (Đ = 1.16, Table 1, entry 6), likely owing to the high ring strain of the ODAN derivatives. The homopolymers of other ODAN derivatives can also be prepared through living ROMP (Table 1, entries 11–13) with diverse glass transition and thermal decomposition temperatures ( Supporting Information Figures S82–S88 and Tables S2–S3).

• Download figure
• Download PowerPoint

The living ROMP of ODAN derivatives was then applied to prepare multiblock copolymers with well-defined sequences. Through sequential addition of classical living ROMP monomers ( NB1– NB3) and ODAN monomers ( ODAN1 and ODAN4), a penta-BCP with five different components was formed with narrow dispersity, further confirming the well-controlled living ROMP process (Figure 2c). This is in stark contrast to previously reported biodegradable ROMP polymers. For example, the ABABA-type penta-block and ABABABA-type hepta-block hydrolysable ROMP polymers prepared by Kilbinger and coworkers²⁸ used cyclic acetals as the sacrificial “A” blocks. However, due to insufficient ring strain of the degradable dioxepine monomers, the chain-extension from the “A” blocks to the nondegradable B blocks displayed low efficiency, resulting in obvious polymodal GPC peaks.²⁸ Therefore, partially biodegradable ROMP BCPs more frequently use the sacrificial components as “end-blocks” without further block growth due to the loss of controllability.^24–29 More recently, the use of 1,2-dihydrofuran (DHF) in biodegradable ROMP polymers has received much attention.^{34,35,37–40} Although its entropy-driven homo-polymerization require neat conditions,³³ DHF can undergo nearly alternating copolymerization with reactive alkenes.³⁷ By regarding the alternating copolymers as a whole biodegradable block, well-controlled biodegradable ROMP copolymers were prepared.^{34,35,37–40} However, the DHF-based alternating “mixed-block” can only be extended by another DHF-based block, and no controlled growing of nondegradable blocks was reported. This is mostly because in these alternating copolymerization systems, DHF has to be used in excess to ensure degradability, while DHF cannot be fully converted. Therefore, the ABCDE-type penta-block results reported herein (Figure 2c and Supporting Information Figures S75–S81) clearly indicate that the ODAN monomers can not only lead to efficient chain-extension from the Ru-carbene active centers, but can also maintain the viability for other ROMP monomers to propagate, showing obvious advancement over other biodegradable ROMP monomers with moderate to low ring-strain energies.

Controlled nanoetching morphology

After the ROMP reactions, the ODAN monomers are transformed into 1,3,4-oxadiazolidine-3,4-dicarboxylate scaffolds within the polymer backbones, featuring directly connected allyl hemiaminal ether and carbamate units. As a consequence, it is hypothesized that the polymer backbones are hydrolyzable under biorelevant conditions.^47,54 Specifically, although the ODAN-derived polymers can be stably stored at room temperature as solids for over 3 months without significant changes in M_n and Đ ( Supporting Information Figure S13), rapid degradation of P-ODAN1₁₀₀ homopolymers was observed when subjected to a dilute HCl/THF solution (0.02 M) for 4 h, as evidenced by comparison of GPC traces in Supporting Information Figures S14–S18. The degradation processes were accelerated to completion within 2 and 1 h when the temperatures were raised to 40 and 60 °C, respectively. By analyzing degradation products of the small molecular model compound ROM1, the degradation mechanism was proposed to be hydrolysis of the 1,3,4-oxadiazolidine-3,4-dicarboxylate core into α,β-unsaturated aldehydes and 1,2-hydrazinedicarboxylate (Scheme 1 and Supporting Information Figure S20).^47,54 The former acrolein derivatives are volatile and unstable, thus giving rise to complicated further decomposition products,⁵⁵ while the latter one has the potential to be oxidized to DEAD,⁵⁶ indicating partial recyclability to ODAN1.

• Download figure
• Download PowerPoint

Self-assembly is one of the most charming features of BCPs,⁵⁷ and the attachment of a sacrificial degradable block enables the preparation of nanoetching materials.⁵ The use of ROMP reactions has the potential to generate these etchable BCPs through a one-step process, avoiding the complicated controlling of chain-ends and switching of polymerization methods via traditional methods.^6,58,59 However, the self-assembly properties were not evaluated in earlier biodegradable ROMP BCPs, and there are two plausible reasons: On the one hand, the controllability of biodegradable ROMP polymers from low to moderately strained monomers does not meet the requirements of BCP self-assembly study;^24–29 On the other hand, the BCPs involving the multicomponent alternating DHF blocks may have too complicated microphase separation and self-assembly behaviors to be thoroughly evaluated.^{17,34,35,37–40} While in the ODAN-based biodegradable polymers reported herein, both problems are easily circumvented. The diblock copolymer P-NB1₁₀₀- b -ODAN1 _m, featuring one nondegradable block of NB1 and a second degradable block ODAN1, was subjected to heterogeneous acidic hydrolysis and maintained a polydispersity index of 1.02 afterwards (Figure 3a,b and Supporting Information Figures S21–S26). By varying the composition of ODAN1 blocks, we obtained various microscopically ordered morphologies of well-defined P-NB1₁₀₀- b -ODAN1 _m copolymers through a one-step process. By selectively etching the degradable ODAN1 block, we prepared ordered nanoporous P-NB1₁₀₀ blocks with different morphologies (Figure 3c–e).

• Download figure
• Download PowerPoint

The solvent-casting films exhibited specific morphologies as the molar fraction of degradable ODAN1 block increased from 0.20 to 0.43 (see cross-sectional SEM images in Supporting Information Figures S27–S29), although their one-dimensional small-angle X-ray scattering (1D SAXS) signals were not strong enough to identify the corresponding structures due to the similar electron cloud density differences between the two blocks (Figure 3c–e, black symbols). After the selective etching of ODAN1 blocks under heterogeneous acidic conditions, the nondegradable NB1₁₀₀ blocks showed spherical, cylindrical, and lamellar morphologies (Figure 3c–e, blue symbols and Supporting Information Figures S30–S32). The microstructures and interdomain distance (d) of the etched P-NB1₁₀₀- b -ODAN1_x films were determined in detail based on the positions of the principal and multiple high-order reflections in the 1D SAXS profiles. A body-centered cubic structure with peak ratios of 1:$\sqrt{2}$ was detected for degraded P-NB1₁₀₀- b -ODAN1₂₅ (Figure 3c). The d was determined to be approximately 17.0 nm by the magnitude of scattering vector at the first intensity peak (q_max), which was consistent with that measured from the SEM images. The etched P-NB1₁₀₀- b -ODAN1₅₀ showed a well-evident hexagonal-packing cylinder (HEX) structure, as revealed by both SEM images and SAXS profiles (Figure 3d). The detected peak ratios were 1:$\sqrt{3}$:$\sqrt{7}$:3:4.4, and some local misalignments of the HEX packing may occur in the microphase domains by the position of the fifth-order scattering peak. The interdomain (or intercylinder) distance of the HEX structure was determined to be 27.3 nm, while the average cylinder diameter measured from the SEM images was 13.1 ± 2.7 nm. With further increase in the ODAN1 block content to approximately 43 mol % ( P-NB1₁₀₀- b -ODAN1₇₅), noticeable lamellar stacks (LAM) were observed (Figure 3e). The interdomain distance (or the distance between stacks) decreased from 26.0 to 16.6 nm, interpreted as a collapse of the LAM due to the lack of support from ODAN1 blocks. Therefore, by simply modulating the block length (or the degree of microphase separation) during the living ROMP of ODAN monomers, we can precisely control the shapes and sizes of the nanopores in a classical ROMP polymer like P-NB1₁₀₀ ( Supporting Information Figures S30–S33). This straightforward approach provides a convenient and promising platform for the preparation and development of well-defined etchable materials with ordered nanostructures.

Controlled fragments

Statistical copolymerization is another approach to efficiently incorporate degradable functional groups onto polymer backbones.^17,18 Compared to previously reported biodegradable ROMP monomers, ODAN derivatives have significantly higher ring strain, thus leading to a broader scope of comonomers, ranging from strained norbornene derivatives to moderately/low strained cis-cyclooctene (COE) and cyclopentene (CPE) (Table 2). The statistical copolymerization with norbornene derivatives remained in the living manner, producing polymers with narrow polydispersity (Đ < 1.2). By increasing the [M]/[I] ratio, statistical copolymers with high molecular weights (M_n > 200 kDa) can be easily achieved (Table 2, entries 2 and 5). When COE was subjected to the statistical copolymerization with ODAN1, both monomers were completely converted, although the polydispersity was relatively broad due to the secondary metathesis of polycyclooctene units (entry 9). Even in the presence of CPE, ODAN1 achieved complete conversion, with approximately 25% unreacted CPE remaining (entry 10).

For backbone-degradable statistical copolymers, the distribution of the degradable functional groups directly determines the degradability and the degradable fragments, while the monomer distribution is contributed by both reactivity ratios and monomer feeding ratios.⁶⁰ The monomer distribution can be easily tuned by the feeding ratios only when the reactivity ratios of the comonomers are comparable. Likely because of the synergistic effect of the thermodynamic driving force of high ring-strain and the appropriate chelation kinetic factors of the urethane groups during ROMP of ODAN monomers ( Supporting Information Figure S34),⁶¹ the reactivity ratios of ODAN1 with classical ROMP monomers varied in a broad range (Mayo-Lewis plots in Supporting Information Figures S35–S38). For example, the mixture of ODAN1 and the exo-derivative NB2, which is one of the most common living ROMP monomers, gave nearly ideal copolymerization, as both reactivity ratios were close to 1 (r _ODAN1 = 1.0, r _NB2 = 1.1).⁶⁰ When copolymerizing with the endo-derivative NB3, ODAN1 converted significantly faster (r _ODAN1 = 4.2, r _NB3 = 0.8). In contrast, when copolymerizing with the nonsubstituted NB4, ODAN1 converted more slowly (r _ODAN1 = 0.3, r _NB4 = 1.9), possibly due to functional group chelation of ODAN1 during ROMP ( Supporting Information Figure S34).⁶¹ It is noteworthy that copolymers of COE and ODAN1 exhibited both r values smaller than 1 (r _ODAN1 = 0.6, r _COE = 0.4), suggesting the propensity to form alternating copolymers, which is likely due to the lower hindrance and smaller strain in COE than ODAN1 ( Supporting Information Figure S39).^31,35,37,62 In contrast, according to earlier reports, the reactivity ratios of 7-membered cyclic acetal monomers with NB3-like endo-monomers were 0.19 and 3.48, respectively,²⁹ and those of DHF and NB2-like exo-monomers were 0.088 and 0.41, respectively.³⁷ Therefore, utilizing ODAN as monomers yields well-controlled ROMP polymers with respect to molecular weights, polydispersity, and functional group distribution, thus leading to a new dimension to tune degradability.

The statistical copolymers were then subjected to acidic hydrolysis and the degradation behaviors were consistent with expectation. The statistical copolymers of ODAN1 and NB4 with equal molar contents ( P-ODAN1 _m - s -NB4 _n, m = n) readily degraded into oligomers with similar molecular weights after identical acidic treatment ( Supporting Information Figures S40–S46). In contrast, when the proportions of NB4 in the copolymers were varied ( P-ODAN1 _m - s -NB4 _n, m = 150, n = 150–750), the molecular weights of the degraded products linearly correlated with the NB4 feeding equivalents (Figure 4a,b and Supporting Information Figures S47 and S48). These results confirm that the ODAN monomers can be efficiently incorporated into polynorbornene backbones, even for the highly reactive nonsubstituted norbornene NB4, and this result is consistent with the reactivity ratio analysis ( Supporting Information Figures S49 and S50). By utilizing this method, it is possible to control both the distribution of degradable units and the size of fragments after degradation, resulting in an unprecedented level of tunability in biodegradable ROMP systems.¹⁶

• Download figure
• Download PowerPoint

Controlled degradation rates

Another advantage of the ODAN monomers is the simple modification of substituents, as four different derivatives can be easily prepared from four inexpensive commercial azodicarboxylates. Their degradation rates were compared as both homopolymers ( Supporting Information Figure S19) and water-soluble bottlebrush copolymers ( P-NB_PEG20- s -ODAN1₄₀, Figure 5a). The latter one was prepared by copolymerizing monomer ODAN1 with the poly(ethylene glycol) (PEG)-grafted norbornene monomer ( NB_PEG, M_n ≈ 5 kDa). Additionally, one unit of N-hydroxysuccinimide (NHS)-functionalized monomer was incorporated for the introduction of functional groups. A nondegradable version of the PEG-grafted polynorbornene ( P-NB_PEG20) was also prepared for comparison. The results showed that P-NB_PEG20- s -ODAN1₄₀ underwent gradual degradation when dissolved in PBS (pH ≈ 7.4) at 25 °C. After 7 days, GPC analysis indicated that the copolymer had predominantly converted to the ring-opened macromonomers (Figure 5b and Supporting Information Figure S51). In contrast, the benzyl-substituted copolymer P-NB_PEG20- s -ODAN3₄₀ showed complete hydrolysis in only 48 h ( Supporting Information Figure S52), while the tert-butyl-substituted copolymer P-NB_PEG20- s -ODAN4₄₀ required nearly two weeks for complete hydrolysis (Figure 5b and Supporting Information Figure S53). It is obvious that the major degradation products have similar molecular weights to the macromonomer from all three copolymers. This confirmed the effective incorporation of ODAN and the controllable degradability of the resultant copolymers ( Supporting Information Figure S54).³⁷ In contrast, the M_n of the nondegradable P-NB_PEG20 remained almost unchanged after the same treatment ( Supporting Information Figure S55).

• Download figure
• Download PowerPoint

To evaluate the biodegradability of the ODAN-based ROMP copolymers, we synthesized Cyanine 5 (Cy5)-labeled degradable PEG-grafted polymers through NHS coupling to enable in vivo fluorescence imaging (the fluorescent degradable polymer is abbreviated as DPPEG). Cy5-labeled nondegradable PEG-grafted polymers (NPPEG), free Cy5, and PBS were used as controls. DPPEG, NPPEG, free Cy5, or PBS was intravenously injected through the tail vein into BALB/C nude mice at equal dose of Cy5. We monitored the clearance of the Cy5 signals from mice by an in vivo imaging system for up to 168 h. As shown in Figure 5c, the fluorescent signals of free Cy5 rapidly decreased from 0.5 to 48 h and reached to the background level at 120 h post-injection, suggesting rapid clearance of Cy5 due to its small molecular weight. On the other hand, Cy5 signals from NPPEG-treated mice slowly declined and maintained much longer than free Cy5, suggesting prolonged blood circulation time and slower clearance rate, which are often observed for PEG brush polymers with high molecular weights.⁶³ The retention time of Cy5-DPPEG in mice was between that of free Cy5 and NPPEG, suggesting that Cy5 was attached to DPPEG upon injection and was then slowly cleared due to the degradation of polymer backbones. Ex vivo imaging of major organs showed that NPPEG mainly accumulated in the liver, which is typical for PEG brush polymers (Figure 5d).⁶⁴ Free Cy5 primarily accumulated in the kidney and was cleared from the body through glomerular filtration.⁶⁵ Quantification of Cy5 through either imaging data or supernatants of homogenized organs both showed that DPPEG and free Cy5-treated mice have similar fluorescent signals which are significantly lower than NPPEG-treated mice in the liver (Figure 5e and Supporting Information Figure S56). These results further demonstrate that DPPEG was degraded in 7 days post-injection, which is consistent with the GPC analysis (Figure 5b and Supporting Information Figure S51). Collectively, these data suggest that DPPEG can serve as a biodegradable material to tune the blood circulation time and biodistribution of small molecules, thus serving as a promising candidate as drug delivery vehicles. The fast degradation kinetics of DPPEG could alleviate toxicity or side effects associated with the prolonged accumulation of traditional ROMP-polymer-based drug carriers and enable repeated administration of drugs in short periods.

Conclusion

In summary, ROMP monomers namely ODAN are expediently synthesized through a one-step, solvent-free hetero-DA cycloaddition process from furan and azodicarboxylates. ODAN derivatives have high ring strain, allowing for living polymerization to afford well-defined homo- and block-copolymers with narrow polydispersity, as well as statistical copolymers with a broad scope of cyclic olefin comonomers and diverse reactivity ratios. After ROMP, ODAN groups convert into 1,3,4-oxadiazolidine-3,4-dicarboxylate scaffolds that are hydrolyzable with tunable rates. The BCPs can be efficiently degraded, and the heterogeneous degradation successfully produced nanoetched microporous polymers with well-defined microstructures. Meanwhile, the statistical copolymers can also be degraded into fragments with controllable molecular weights, and the water-soluble brush copolymers are successfully degraded under biological conditions. Overall, this straightforward method enables ROMP to produce polymers that simultaneously have both backbone-biodegradability and precise structural tunability, which will find potential applications ranging from nanolithography to drug delivery.

Supporting Information

Supporting Information is available and includes experimental procedures, characterization, NMR, GPC, DSC, DFT calculation, and relevant discussions.

Disclosure

The authors declare no conflict of interest. All animal experiments reported here were performed according to a protocol approved by the Peking University Institutional Animal Care and Use Committee (Approval No. LA2021362).

Data Availability Statement

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

Preprint Acknowledgment

Research presented in this article was posted on a preprint server prior to publication in CCS Chemistry. The corresponding preprint article can be found here: 10.26434/chemrxiv-2023-ltf08-v3.

Funding Information

This work was financially supported by the National Natural Science Foundation of China (grant nos. 22001254 and 22175188).