3,6-Anhydrogalactal as A Single-Addition Monomer for Chain-End Functionalization of Metathesis Polymers

Abstract

Controlling the structure and reactivity of the chain-end group is a central objective in modern polymer chemistry. Here, we introduce 3,6-anhydrogalactal as a single-addition monomer that enables efficient and versatile chain-end functionalization of metathesis polymers. Readily synthesized from biomass-derived galactal, 3,6-anhydrogalactal exhibits excellent single-addition reactivity, allowing precise chain-end modifications even when introduced simultaneously with the propagating monomer. Theoretical calculations provide mechanistic insights into the unique reactivities governing its single-addition behavior. Its broad functional group compatibility facilitates diverse applications, including block copolymer synthesis, polymer-polymer coupling, and bioconjugation, demonstrating significant potential for advancing polymer materials and bioconjugation strategies.

Graphical Abstract

Placing customized functional groups at the chain ends of polymers not only enhances the utility of polymers by expanding their functionality beyond the properties of the polymer backbone, but also enables the construction of more sophisticated molecular architectures.¹ Ring-opening metathesis polymerization (ROMP) is a living polymerization technique that has demonstrated utility in diverse applications. However, ROMP mediated by the Grubbs catalysts typically produces an unstable Ru-carbene intermediate at the ω-chain end that requires chain-end capping/quenching. Reagents for the functionalization of the ω-chain end of metathesis polymers typically belong to two classes. The first class of reagents, such as internal cis-olefin², vinylene carbonate³, 3H-furanone³, 2-methoxy-3,4-dihydro-2H-pyran⁴, linear enol ether⁵, enynes⁶, diazoacetates⁷ and allyl halides⁸, convert the R-ucarbene into a less-reactive olefin and cleavage the Ru species from polymer backbone (Figure 1A). However, with the exception of enynes, reagents of this class often suffer from low chain-end functionalization efficiency and can only install a limited scope of functional groups. The enyne-type reagents developed by Gutekunst et al.⁶ represent an elegant approach that uses cascade metathesis to achieve efficient chain-end capping, but require lengthy synthesis and result in dead chain ends that cannot be further extended. The second class of reagents are single-addition monomers that react with the Ru-carbene propagating species and form a less reactive Ru-carbene species unable to homo-propagate. Examples of single-addition monomers in ROMP are limited, and so far only include a 1,1-cyclopropene reported by Xia et al.⁹, an oxanorbornadiene derivative reported by Kilbinger et al.¹⁰ and an oxanorbornene imide reported by Foster et al.¹¹ (Figure 1B). However, the cyclopropane derivatives require lengthy syntheses (> 6 steps) and precious reagents/catalysts; functional group compatibility of oxanorbornadiene/oxanorbornene derivatives remains limited.

Figure 1.

Chain-end functionalization of metathesis polymers via (A) chain termination/transfer or (B) single addition. (C) This work: 3,6-anhydrogalactal as a single-addition monomer. FG = functional group.

To address these challenges, herein we report 3,6-anhydrogalactal, a carbohydrate bicyclic enol ether, as a single-addition monomer for the chain-end functionalization of metathesis polymers (Figure 1C). The first cyclic enol ether monomer for ROMP was 2,3-dihydrofuran (DHF) reported in 2019 by Feist and Xia¹². Subsequently, DHF and 3,4-dihydropyran (DHP) was utilized in the copolymerization with high ring strain monomers such as norbornene derivatives^13,14, tricyclodecadienes¹⁵, enynes¹⁶, diynes¹⁷, and terminal alkynes¹⁸. Notably, bicyclic enol ethers have not been used in ROMP to date. Compared to DHF and DHP, 3,6-anhydroglycals have a higher ring strain as the thermodynamic driving force for ring-opening, and form a Fischer-type Ru-carbene whose reactivity is further reduced by steric hindrance and oxygen coordination, leading to single monomer addition. We demonstrated that 3,6-anhydrogalactal can be facilely synthesized from biomass-derived D-galactals through a straightforward two-step synthetic route and can install a broad scope of chain-end functionalities including acyl, alkyl, halides, azides, alkynes, N-hydroxysuccimide (NHS) esters for polymer chain-end coupling, block copolymer synthesis, and bioconjugation.

First, we prepared a 3,6-anhydro-d-glucal (2) following the report by Tucker and coworker¹⁹ from 3,4-di-O-acetyl-6-O-toluene-p-sulphonyl-d-glucal (1) in the presence of basic OH resin. Intermediate 1 could be readily synthesized starting from inexpensive D-glucal through tandem deacetylation and cyclization (Figure 2A). Following the same route, we also prepared 3,6-anhydro-d-galactal (4), a geometric isomer of 2, in high yield (83%). The metathesis reactivity of 2 and 4 was found to differ significantly. Nearly no initiation was observed when 2 was treated with Grubbs first-generation catalyst (G1) at room temperature (R.T.), or Hoveyda-Grubbs second-generation catalyst (HG2), at 50 °C over 24 hours (Figure S1). A slow initiation was observed when Grubbs third-generation catalyst (G3) was employed (43% conversion, 24 hours, R.T.) (Figure. 2B). The ¹H NMR signal of the product at 13.51 ppm was assigned to the alkylidene proton of Fischer-type Ru-carbene (Figure S2). The low reactivity of 2 was attributed to steric hindrance and the Ru coordination by the -OH group on C4 endo to the enol moiety, impeding monomer coordination to the Ru center. Consistent with this rationale, 4 with a -OH group on C4 exo to the enol moiety demonstrated a significantly faster initiation with G3, reaching 100% conversion into the Fischer-type Ru-carbene in an hour (Figure. 2B and S3). Only one equivalent (relative to the catalyst) of 4 was consumed after 24 hours, suggesting that 4 did not undergo homo-propagation and was a single-addition monomer in ROMP.

Figure 2.

Monomer synthesis and initial reactivity studies. (A) Synthetic route of 3,6-anhydro-d-glucal (2) and 3,6-anhydro-d-galactal (4). (B) Reaction of G3 with 2 or 4 monitored by the conversion of the Ru-alkylidene signal in 1H NMR. (C) ROMP of M1 and chain-end functionalization using 5. (D) 1H NMR spectra of the Ru-alkylidene peak before and after reaction with 5. (E) MALDI-TOF mass spectrum of P3 indicated high chain-end fidelity. (F) ¹H NMR spectrum of P3, showing matching integrals of the protons at the ω-chain end and those at the α-chain end. (G) Simultaneously added M1 and 5 successful chain-end functionalization. (H) SEC traces of the polymers (DP = 8) obtained from simultaneous addition of M1 and 5 and sequential addition of these monomers, respectively.

We then studied ROMP of a model system consisting of 4-methyl-3,6-anhydro-d-galactal 5 as the single-addition monomer and norbornene exo-dicarboximide M1 as the propagating monomer (Figure 2C). The polymerization of M1 is rapid, consuming all monomers in 5 min and generating P1 with an Ru-alkylidene peak at 18.54 ppm in ¹H NMR (Figure 2D). After the addition of 2 equiv. of 5, the Ru-alkylidene signal (18.54 ppm) was converted into a Fischer-type Ru-carbene signal at 13.76 ppm over four hours at 22 °C (denoted as P2) (Figure 2D). Complete conversion could also be accomplished using 1.5 equiv. of 5 under 45 °C in 2 hours. A polymer with a stable ω-chain end, P3, was formed by quenching the Fischer-type Ru-carbene in P2 with ethyl vinyl ether (EVE). MALDI-TOF mass spectrum of P3 showed only one series of peaks corresponding to the expected mass (Figure 2E). In 1H NMR, the ratio of the integrals of the ω-chain end to the α-chain end was consistent with the stoichiometry expected from the molecular formula (Figure 2F and S4). Taking together, all evidence suggested that quantitative ω-chain end functionalization of the polymer.

Given the reactivity difference between 5 and M1, we wondered whether they can be added into the reaction simultaneously without affecting the reaction outcome. Indeed, at degree of polymerization (DP) of 8, when 5 and M1 were added simultaneously, a polymer was generated with identical ¹H NMR spectrum and molecular weight compared to the one generated by the sequential addition of these two monomers (Figure 2G–H and S5). Chain-end incorporation of 5 that was added simultaneously with M1 remains highly efficient at a higher DP of 90 (Figure S5). Similarly, successful ω-chain end functionalization was also achieved for the simultaneous addition of 5 and a PEGylated norbornene exo-dicarboximide M2 (Figure S6), or cis-cyclooctene (COE) (Figure S7). To the best of our knowledge, 3,6-anhydrogalactal represents the first single-addition monomer in ROMP that can be added simultaneously with the propagating monomer without compromising the latter’s reactivity.

Density functional theory (DFT) calculations provided mechanistic insights into the reactivity of 5 in the presence of propagating monomers in ROMP (Figure 3 and S8). Compared to M1 (Precat 1 to ts_{G8_G9} ΔG^‡ = 19.9 kcal/mol), the initiation of 5 (Precat 1 to ts_{G3_G4} ΔG^‡ = 22.7 kcal/mol) required to overcome a higher energy barrier, with a ΔΔG^‡ of 2.8 kcal/mol, suggesting that the initiation of M1 was faster than that of 5. Moreover, double addition of 5 was also met with a high energy barrier (G5 to ts_{G12_G13} ΔG^‡ = 16.5 kcal/mol), and was significantly endergonic (ΔG = 10.8 kcal/mol) from the single-addition product G5, making the double addition of 5 energetically disfavored. In addition, our calculations also suggest that the bicyclic enol ether structure of 5 raised the ring strain energy (RSE = 5.2 kcal/mol) compared to that of a simple galactal (RSE = 0.5 kcal/mol) (Figure S8). The higher RSE was necessary for 5 to react efficiently with the propagating Ru-carbene. When tri-O-acetyl-d-galactal was used to react with the Fischer-type Ru-carbene at the chain end of P2, more than one Ru-carbene peaks (19.10, 18.54, 13.22 and 13.17 ppm) were observed after the treatment of 45 °C for 2 hours (Figure S9), indicating that monocyclic galactals could not achieve efficient chain-end functionalization.

Figure 3.

Transition states and key intermediates in the initiation and propagation of 5 or M1 in ROMP. Free energies (kcal/mol) were obtained at the M06(D3)/def2-TVZPP/PCM(CHCl₃)//BP86(D3)/def2-SVP level. NHC = 1,3-dimesitylimidazol-2-ylidene; Ph = phenyl; Me = methyl.

Despite the reduced reactivity of the Fischer-type Ru-carbenes²⁰, reports have shown that cyclic enol ether such as DHF could ready react with this species in ROMP.^11–17 Indeed, chain extension of P2 by a mixture of DHF and norbornene N-isopropyl-exo-dicarboximide (M3) successfully generated a block copolymer P4, with a number-average molecular weight (M_n) of 7.6 kg/mol and dispersity (Đ) of 1.08 (Figure 4A and 4B). Diffusion-ordered spectroscopy (DOSY) NMR analysis showed only one diffusion signal corresponding to the ¹H NMR signals from all repeating units, confirming the block copolymer structure (Figure S10). Consistent with previous reports on the ROMP of DHF¹², we also found that the vinyl ether motif from the incorporated DHF conferred degradability for this block copolymer under acidic conditions (Figure S11).

Figure 4.

Chain extension and chain-end functionalization enabled by 3,6-anhydrogalactals. (A)–(B) Chain extension of P2 by ROMP of DHF and M3, confirmed by SEC. (C)–(D) Monomers 6, 7, and 8 were used to generate P5, P6 and P7 with defined chain ends, confirmed by MALDI-TOF.

Leveraging their single-addition reactivity, we prepared a series of derivatives of 3,6-anhydro-d-galactal for the chain-end functionalization. We found that 3,6-anhydro-d-galactals 6, 7, and 8, consisting of an NHS ester, a bromide, and a 2-bromoisobutyryl group, respectively, all achieved successful chain end functionalization when they were added with M1 simultaneously. The ¹H NMR and MALDI-TOF analyses confirmed the successful chain-end functionalization of the resulting polymers, P5, P6, and P7 (Figures 4D, S12–S14). The NHS ester group and bromide could readily react with various nucleophiles, enabling diverse functional groups to be introduced at the ω-chain ends of metathesis polymers. For example, the NHS group of P5 could react with propargyl amine to generate P8 with an alkyne chain end, while the bromide group of P6 can react with sodium azide to generate P9 with an azido chain end (Figures S15–S16). Facile coupling between P8 and P9 could be achieved via copper-catalyzed azide-alkyne cycloaddition (CuAAC) reaction to give P10 with a molecular weight consistent with the sum of those of P8 and P9 (Figures 5A–B, and S17). Next, P7 with a α-bromoisobutyryl chain-end group was prepared using a 3,6-anhydro-d-galactal 8 bearing this group. Subsequent atom-transfer radical polymerization (ATRP) of styrene using P7 as the macroinitiator generated block copolymer P11 (M_n = 10.4 kg/mol, Đ = 1.07) (Figures 5C and 5D). DOSY NMR analysis confirmed a single polymer species consisting of both repeating units (Figure S18–S19).

Figure 5.

Utilities of 3,6-anhydrogalactal in block copolymer synthesis and bioconjugation. (A) Polymer-polymer coupling of P8 and P9 using click reaction. (B) SEC traces of P8, P9 and P10. (C)–(D)Chain extension of P7 by ATRP of styrene produced P11, confirmed by SEC. (E) Bioconjugation of P13 with sfGFP-151-AzK generated sfGFP-151-P13. (F) SDS-PAGE gel of sfGFP-151-AzK, sfGFP-151-DBCO and sfGFP-151-P13.

Finally, we investigated the utility of our chain-end functionalization strategy for polymer bioconjugation. A water-soluble polymer P13 was prepared using M2 and 7 followed by chain-end azidation (Figures 5E, S20–S22). A green fluorescent protein with an azido lysine incorporated at the position 151 (sfGFP-151-AzK) by non-canonical amino acid (ncAA) mutagenesis was functionalized with a dicyclooctyne DBCO-PEG8-DBCO (see Supporting Information for structure), before coupling with P13. Both SDS-PAGE and MALDI-TOF analyses suggest that P13 was successfully conjugated to sfGFP-151-AzK in 87% yield (Figure 5F and S23).

In conclusion, we demonstrated 3,6-anhydrogalactals as a class of single-addition monomers for efficient and versatile chain-end functionalization of metathesis polymers. Readily synthesized from biomass-derived galactal, these 3,6-anhydrogalactals exhibit excellent single-addition reactivity even when introduced simultaneously with the propagating monomer. Compatible with diverse functional groups, 3,6-anhydrogalactals facilitated chain-end functionalization of metathesis polymers for a variety of applications, such as block copolymer synthesis, polymer-polymer coupling, and bioconjugation. DFT calculations elucidated the unique reactivities of 3,6-anhydrogalactals underlying their single-addition behavior in the presence of a propagating monomer. These findings underscore the potential of 3,6-anhydrogalactals as cost-effective and sustainable reagents for advanced polymers in material science and bioconjugation.

Supplementary Material

ASSOCIATED CONTENT

Supplementary figures and tables, characterization dataand ¹H/¹³C NMR spectra, and detailed experimental protocols.