Modular Approach to Degradable Acetal Polymers Using Cascade Enyne Metathesis Polymerization

Abstract

A modular synthetic approach to degradable metathesis polymers is presented using acetal-containing enyne monomers. The monomers are prepared in a short and divergent synthetic sequence that features two points of modification to tune polymerization behavior and introduce molecular cargo. Steric and stereochemical elements are critical in the monomer design in order to provide rapid and living polymerizations capable of generating block polymers. The developed polyacetal materials readily undergo pH-dependent degradation in aqueous mixtures, and the rate of hydrolysis can be tuned through post-polymerization modification with triazolinedione click chemistry. This presents a new scaffold for responsive metathesis polymers that may find use in applications that requires controllable breakdown and release of small molecules.

Graphical Abstract

Well-defined polyacetals are designed that readily degrade in aqueous environments through the use of enyne metathesis polymerization. This approach permits modular design of the monomer scaffolds, and post-polymerization functionalization with triazolinediones modulates the rate of hydrolytic degradation.

Ruthenium-based olefin metathesis polymerization has greatly expanded the types of monomers that can be used in living polymerization due to the wide-range of functional groups and reaction media tolerated.^[1],[2] This has led to the direct polymerization of unprotected sugars, peptides, and nucleosides for use in biomedical applications and targeted drug-delivery systems.^[3] Historically promoted by cycloolefins with high ring strain, nearly all polymers produced through ring-opening metathesis polymerization (ROMP) possess all-carbon backbones that are not easily broken down in the environment or in the body (Figure 1A). Given the significant potential of ROMP in the biomedical arena, methods to create well-defined polymers that readily degrade under mild conditions are greatly needed.

Figure 1.

(A-C) Previous monomer designs for metathesis polymerization. (D) New monomers for synthesis of degradable polyacetals via enyne metathesis.

A number of strategies have been reported in the literature to create metathesis-derived materials with the ability to degrade through hydrolytic, redox, or photolytic events. While acyclic diene metathesis (ADMET) has produced many responsive polymers, the step-growth mechanism limits the ability to control molecular weight and generate block polymers.^[4] Ring-opening metathesis polymerization (ROMP) has the potential to introduce these features, but it has proven difficult to maintain livingness of the polymerizations along with the introduction of labile backbone functionality. Entropy-driven ROMP has granted access to degradable polyesters with generally high dispersities, though recent developments by Meyer have shown that polymerization control can be obtained with the use of cycic Z-olefins.^[5],[6],[7] The installation of a disulfide, phosphate, or acetal unit into 7 and 8-membered cycloolefin derivatives results in degradable materials, but the monomers cannot homopolymerize to high molecular weights due to reduced ring strain.^[8],[9],[10] A cellulose-derived levoglucosenol monomer recently described by Schlaad and coworkers produced high molecular weight polyacetal thermoplastics, though without control of molecular weight and dispersity.^[11] The most successful strained monomer to date is Kiessling’s bicyclic oxazinone system, which features an easily modified synthetic handle for introducing fluorophores and bioactive epitopes (Figure 1B).^{[12],[13],[14]} While the dispersities reported were modest, targetable molecular weights were obtained in good agreement with theory. Further, the examples by Schlaad and Kiessling required significantly acidic (pH < 4) or basic conditions (pH > 9) to degrade, leading to materials that would persist in a biological environment for long periods of time.

An alternative to the use of high ring strain monomers^[15] for ROMP was reported by Choi and coworkers in 2012 (Figure 1C).^[16] Enyne monomers with an unstrained cyclohexene ring polymerized rapidly with the 3^rd generation Grubbs initiator (G3) due to the facile addition across the terminal alkyne followed by rapid intramolecular ring-opening. While this polymer featured an all carbon backbone,^[17] Hawker recognized the potential to expand the ring size to macrocycles for preparation of sequence-defined polyesters.^[18] This strategy was effective but the monomer synthesis required lengthy reaction sequences and suffered chain transfer at high conversions. In this communication, a modular synthetic strategy is reported to produce well-defined polyacetals using cascade enyne metathesis polymerization that readily degrade under mild conditions (Figure 1D).

The general strategy for the preparation of the enyne acetal monomers M1-M6 used in this study is shown in Scheme 1 starting from tert-butyl carbonate 1, which was prepared on multigram scale through the Achmatowicz rearrangement of furfuryl alcohol.^[19] Tsuji-Trost substitution of the carbonate is the first point of diversification in the monomer design and cleanly delivered a number of racemic alkyl acetal derivatives (2) for exploration using a slight excess of alcohol (72-95% yield).^[20],[21] The silyl acetal derivative was alternatively prepared through direct reaction of the carbonate precursor with tert-butyldimethylsilyl chloride (Scheme S4).^[22] While the alkyl acetal derivatives could also be synthesized from this intermediate using acid catalysis, large excesses of alcohol were required and resulted in overall lower yields. Luche reduction of the acetals in methanol resulted in highly diastereoselective reduction to the racemic cis-allylic alcohols 3.^[23] A final Mitsunobu inversion of the alcohol with tolyl (Tol) or mesityl (Mes) propargyl sulfonamide occurred in 83-94% yield to deliver the trans-enyne monomers M1-M6.

Scheme 1.

Modular synthesis of enyne acetal monomers.

Tolyl benzyl monomer M1_Tol was the first structure evaluated for potential in enyne metathesis polymerization. Adapting Choi’s protocol, G3 was added to a THF solution of M1_Tol (0.08 M) at room temperature with a monomer to initiator (M:I) ratio of 50:1. The polymerization reached full conversion in 15 minutes to give polymer P1_Tol with a number average molecular weight (M_n) of 26.1 kDa and a dispersity (Ð) of 1.18 according to size-exclusion chromatography (SEC) using polystyrene standards (Table 1, entry 1). Unfortunately, decreasing the amount of G3 to target higher degrees of polymerization (DP) gave increased molecular weight dispersities, with loss of control at M/I = 200 (entries 2 and 3). This was rationalized to arise from secondary metathesis of the backbone olefins, and mesityl derivative M1_Mes was prepared to provide greater steric shielding.^{[16d, 16e, 17]} This modification improved polymerization control and gave targetable molecular weights with low dispersities in targets up to M/I = 200, though an increase in dispersity was seen at M/I = 500 (entries 4-8 and Figure 2). Consistent with previous observations by Choi, a 6:4 mixture of E:Z olefin isomers were produced in the polymer backbone (Figure S9).^[16b]

Table 1.

Polymerization data for M1-M6.

entry	monomer	DP (M/I)	Time (min)	Conv (%)[a]	Mn,SECapp (kDa)[a]	Ð[a]
1	M1Tol	50	15	>95%	26.1	1.18
2	M1Tol	100	30	>95%	47.1	1.27
3	M1Tol	200	60	91%	79.1	1.77
4	M1Mes	25	6	>95%	10.6	1.13
5	M1Mes	50	6	>95%	23.5	1.12
6	M1Mes	100	10	>95%	43.9	1.18
7	M1Mes	200	30	>95%	83.8	1.23
8	M1Mes	500	45	92%	164.7	1.88
9	M2Mes	100	10	>95%	47.2	1.13
10	M3Mes	100	12	>95%	34.0	1.17
11	M4Mes	100	10	>95%	37.4	1.24
12	M5Mes	200	30	>95%	78.7	1.22
13	M5Mes	500	45	>95%	171.1	1.42
14	M6Mes	100	25	10%	7.0	1.50
15	cis-M4Mes	100	20	29%	18.3	1.27

^[a]Determined by CHCl₃ size-exclusion chromatography (SEC) calibrated using polystyrene standards.

Figure 2.

(A) SEC chromatogram of P1_Mes at different targeted degrees of polymerization (DP) and (B) linear correlation of M_n with DP.

With the successful polymerization of this scaffold established, other mesityl enyne monomers were evaluated with different substitutions at the acetal position. Methyl and N-hydroxysuccinimide (NHS) benzyl ester derivatives M2_Mes and M3_Mes also polymerized to high conversions with low dispersities at M/I = 100 (entries 9 and 10). Notably, NHS ester M3_Mes permits introduction of molecular cargo into the polymer chain at a later stage, which would then be released upon degradation. Given the improvement of the mesityl substituent, steric and stereochemical effects at the acetal position were explored. Reduction of the substituent size in methyl acetal M4_Mes resulted in slightly higher dispersities relative to M1_Mes at M/I = 100 (Ð = 1.24, entry 11), consistent with the steric argument proposed. The bulkier isopropyl M5_Mes further improved polymerization characteristics and provided reduced dispersities at M/I = 500 (Ð = 1.44, entry 13). There appears to be a limit on the benefits of steric crowding, as the tert-butyl dimethylsilyl (TBS) acetal M6_Mes polymerized at a much slower rate than the other monomers examined and halted at low conversions (entry 14). Through slight modification of the synthetic strategy (Scheme S3), the cis-methyl acetal cis-M4_Mes was prepared and polymerized under the standard reaction conditions. A lower polymerization rate was observed relative to the trans-M4_Mes diastereomer, with only 29% conversion after 20 minutes (entry 15). This is possibly due to the conformational equilibrium of the half-chair monomer structure, if cyclization is rate-limiting (Figure S11). In order to cyclize onto the six-membered ring after alkyne addition, the sulfonamide substituent needs to be in a pseudo-axial position. In the trans series, this requires the acetal to also be pseudo-axial, which is assisted by the anomeric effect. In the ground state of cis-M4_Mes, the equilibrium favors the large sulfonamide to be pseudo-equatorial and must ring-flip into a disfavored conformation for cyclization, thereby reducing the overall rate of the polymerization

In order to further verify the livingness of the polymerization, kinetic studies were performed with M1_Mes. By sampling the polymerization at different time points, a linear increase in molecular weight with conversion was observed, along with a first order kinetic profile (Figure S12). Since the polymerization displays living characteristics, fully degradable block polymers were accessible using these monomers, which has not been realized in previous metathesis methodologies (Figure 3). This was shown by the sequential polymerization of benzyl M1_Mes (50 equiv) for 6 minutes, followed by addition of isopropyl M5_Mes (100 equiv) for 10 minutes to give a P1_Mes-b-P5_Mes diblock polymer with M_n = 58.4 kDa and Ð = 1.20. A significant shift in the SEC chromatography to an earlier elution time from the initial M1_Mes block supports a high degree of polymer chain reinitiation and successful diblock formation.

Figure 3.

(A) Synthesis of diblock P1_Mes-b-P5_Mes through sequential monomer addition and (B) SEC chromatogram of the first block and chain-extension.

These polyacetals fully degraded into small molecular enal products in the presence of water with concomitant release of the acetal side chain (Figure 4). To study the rates of degradation, experiments with P1_Mes (M_n = 36.2 kDa; Ð = 1.15) were performed under neutral, basic and acidic conditions using tetrahydrofuran/water (20/1 vol/vol) mixtures. Due to the hydrophobic nature of these polymers, a significant volume fraction of organic solvent was needed to obtain homogeneous solution, which was confirmed by dynamic light scattering (Figure S16). Five equivalents of triethylamine (TEA), acetic acid (AcOH), or trifluoroacetic acid (TFA) relative to each monomer unit were added to the solutions to alter pH for the experiments, and the degradations were measured by SEC to monitor the decrease in molecular weight over time. The polymer showed gradual hydrolysis under neutral conditions, with a 33% reduction in molecular weight observed at 24 hours and 63% at 48 hours. When triethylamine (TEA) was added, the polymer fully maintained its integrity over the course of the experiment and no change in molecular weight could be detected. Addition of acetic acid to the solutions produced faster degradation relative to the neutral conditions with an 80% reduction in molecular weight measured at 48 hours. Rapid degradation occurred with trifluoroacetic acid (TFA) additive with only small oligomers remaining after 2 hours and full conversion to the small molecule product at 24 hours. Using the same series of experiments, P5_Mes showed similar overall degradation profiles (Figure S18). These pH effects are consistent with established acetal reactivity, as acid is needed to form an oxocarbenium intermediate for hydrolysis and acetals are stable to basic conditions. While readily degraded in the presence of water, the polyacetal materials were found to be stable for over four weeks in the solid state without any change in molecular weight or dispersity (Figure S14A). Further characterization of P1_Mes supports general thermal stability with 5% mass loss at 266 °C using thermogravimetric analysis, and differential scanning calorimetry shows a glass transition temperature at 82 °C (Figures S20 and S21).

Figure 4.

Degradation of P1_Mes in neutral, basic, and acidic aqueous media.

In order to further tune the rates of degradation and increase stability in aqueous media, post-polymerization functionalization was explored. The lability of these polymers is likely due to the neighboring olefin that stabilizes the intermediate oxocarbenium species generated during hydrolysis. Therefore, removal of this double bond was considered as a means to increase hydrolytic stability. While attempts to hydrogenate the backbone olefins were ineffective, success was found through triazolinedione click chemistry with the highly reactive dienophile 4-phenyl-1,2,4-triazole-3,5-dione (PTAD) ^{[16a, 24]}. The reaction of P1_Mes with PTAD proceed rapidly and quantitatively at room temperature to give the Diels–Alder adduct, P1_Mes-PTAD, as a monomodal peak with low dispersity by SEC analysis (M_n = 31.7 kDa and Ð = 1.22) (Figure 5A). The observed reduction in molecular weight is presumably due to the reduced hydrodynamic volume of the rigid Diels–Alder product that requires reaction of the diene in a cis configuration, as no aldehyde peaks indicative of hydrolysis were observed in the ¹H NMR spectrum. Hydrolysis studies confirmed the more robust nature of P1_Mes-PTAD. Significantly slower rates of degradation of the polymer compared to P1_Mes were observed when exposed to solutions containing TFA and HCl, with only 9% and 24% reductions in molecular weight at 24 hours, respectively (Figure 5B). In addition to serving as a means to modulate degradation rates, the emergence of functional triazolinediones in the literature presents an additional method to bring functionality into the degradable enyne polymers.^{[16b, 25]}

Figure 5.

(A) Triazolinedione click functionalization of P1_Mes with PTAD and (B) increased hydrolytic stability of P1_Mes-PTAD under acidic conditions.

In conclusion, a new platform for the generation of degradable polymers from modular acetal monomers is reported using cascade enyne metathesis polymerization. By tuning the size and stereochemistry of the monomer substituents, living polymerizations are achieved that occur rapidly at room temperature and can be used to prepare fully degradable metathesis diblock polymers for the first time. The introduction of different small molecule substituents at the acetal position provides opportunities to release cargo upon degradation, and the rates of hydrolytic breakdown can be modulated using post-polymerization functionalization. It is anticipated that this strategy will be broadened through the introduction of heteroatom acetals to further modify the degradation windows and explore the use of these polymers in targeted drug delivery applications.