Ultrastrong underwater adhesion on diverse substrates using non-canonical phenolic groups

Achieving strong adhesion in wet environments is vital for medical and marine applications, yet hydration layers on surfaces hinder adhesive–substrate interactions. Conventional dry-environment adhesives (cyanoacrylates, epoxies, polyurethanes) underperform underwater. Bioinspired phenolic polymers derived from mussel adhesion, which relies on catechol-containing amino acids (DOPA), have been widely explored with variations in polymer backbone, topology, comonomers, molecular weight, and coacervation. While these advances improve underwater adhesion, catechol-based polymers still lag behind dry adhesives in strength. Gallol groups (three hydroxyls) from plants have shown improved underwater adhesion relative to catechols. This motivates exploring phenolic units bearing four or five hydroxyls per ring, which could further enhance adhesion; however, such adhesive polymers do not occur in nature and had not been synthesized prior to this study. The purpose of this work is to synthesize non-canonical phenolic polymers bearing four and five hydroxyl groups on a styrenic unit, quantify their underwater adhesion across diverse substrates, and elucidate structure–property relationships involving number and position of hydroxyl groups, polymer composition, and molecular weight.

Prior efforts in underwater adhesives draw inspiration from mussel byssal proteins where catechol groups penetrate hydration layers and bind diverse substrates. Synthetic catechol-functionalized polymers have been optimized via backbone architecture, topology, co-monomer selection, molecular weight control, and coacervation strategies, achieving stronger bonds but still below dry adhesive benchmarks. Gallol-based systems (with one extra hydroxyl vs catechol) have demonstrated stronger underwater adhesion than catechol analogues and have been pursued in underwater and biomedical contexts. Nevertheless, literature lacks polymers incorporating phenolic motifs with four or five hydroxyl groups per aromatic ring. Prior reports also show that increasing phenolic content in catechol or gallol copolymers can reduce performance by creating hydrophilic microenvironments, upsetting interfacial interactions, and increasing bound water, highlighting the need for designs that concentrate phenolics while preserving hydrophobicity.

Synthesis: Non-canonical styrenic monomers 2,3,4,5-tetramethoxystyrene (TMS) and 2,3,4,5,6-pentamethoxystyrene (PMS) were synthesized. TMS was prepared by radical bromination of 2,3,4,5-tetramethoxytoluene followed by a one-pot Wittig reaction. PMS required additional steps to install five methoxy groups on the styrenic ring (details in SI). Free-radical copolymerization of these monomers with styrene (S), followed by deprotection of methoxy groups, yielded poly(2,3,4,5-tetrahydroxystyrene)-co-styrene [P(4HS-co-S)] and poly(2,3,4,5,6-pentahydroxystyrene)-co-styrene [P(5HS-co-S)]. Reference polymers bearing 1–3 hydroxyls (P1HS–P3HS) were similarly synthesized. Isomeric monomers with different hydroxyl positions for 2HS, 3HS, and 4HS were also prepared to probe positional effects.

Backbone-modified monomers N-(2,3,4,5-tetramethoxybenzyl) methacrylamide (TMA) and N-(2,3,4,5,6-pentamethoxybenzyl) methacrylamide (PMA) were synthesized, copolymerized with styrene to M_n ~80 kDa, and deprotected to give P(4HMA8%-co-S92%) and P(5HMA9%-co-S91%). Reactivity ratios were estimated via Fineman–Ross (r_4HMA = 0.19, r_styrene = 3.0).

Polymer characterization: 1H/13C NMR, HRMS, UV–Vis, FT-IR, viscometry, and GPC were used to confirm structures, monitor oxidation, assess molecular weights (M_n) and polydispersity. Viscosity measurements determined the critical entanglement concentration c* (~30 mg/mL even for low M_n 4HS/5HS polymers).

Underwater adhesion tests (tensile on rods): Copolymer solutions were prepared at 300 mg/mL in CHCl3/MeOH (9:1 v/v). A polished aluminum (A1050) or steel (SUS304) cylindrical rod (JIS-K6849, D = 1.27 cm, h = 3.80 cm) was fully submerged. Polymer solution (typically 40 µL) was deposited and painted uniformly on one rod underwater. A second rod was overlapped. A preload (25–300 g) was applied and joints incubated underwater for set times (10 s to ≥1 week). Samples were removed and tested on a SHIMAZU AGS-X 10 kN load cell at crosshead speeds 1–100 mm/min. Adhesion strength was reported; cohesive vs adhesive failure modes were noted.

Lap-shear tests (plates): For glass, wood, PE, and PTFE, flat plates (5 cm × 1 cm × 0.1 cm) were used. Polymer solution (typically 20 µL) was applied underwater, plates overlapped (area 1 cm2 for wood/PE/PTFE; 0.5 cm2 for glass), a 10 g weight was applied, and samples incubated before lap-shear testing.

Parameter studies: Molecular weight dependence was assessed (M_n ~20–80 kDa). Composition studies compared P(2HS26%-co-S74%) versus P(5HS11%-co-S89%) (similar total hydroxyl content), and P(2HS11%-co-S89%) to decouple styrene content effects. Hydroxyl positional isomers were synthesized for 2HS, 3HS, 4HS to probe structure–adhesion relationships. Preload and adhesive quantity effects, and crosshead speed dependence were evaluated.

Surface/interfacial analyses: Quartz crystal microbalance (QCM) measured adsorption of homopolymers (P1HS–P5HS) and copolymers on Au, Si-based, carbon-based, and SUS316 chips. Water uptake of coated films in DI water was quantified by QCM. Static contact angles (SCA) in air and air-bubble contact angles underwater were measured on polymer-coated aluminum. Scanning electron microscopy (SEM) examined fracture morphologies. X-ray photoelectron spectroscopy (XPS, C 1s) quantified oxidation states post-fracture (take-off angles 90° and 30°). Molecular dynamics simulations estimated adsorption energies of isolated phenolic units (1HS–5HS) on Fe.

Oxidation monitoring: 1H NMR tracked oxidation of P(4HS-co-S) and P(5HS-co-S) in air over days to weeks. UV–Vis and FT-IR (C=O band ~1650 cm−1) monitored oxidation of homopolymers (P3HS–P5HS) in water over weeks. GPC assessed changes in M_n/PDI after 2–4 weeks to evaluate crosslinking.

Recycling tests: Fractured adhesive joints were re-used by briefly redissolving the adhesive surface (3 s in 20 µL CHCl3/MeOH 9:1), re-submerging underwater, rejoining, and re-testing over multiple cycles. Solvent-based removal (rinsing or sonication) in common organics (acetone, THF, DCM, ethyl acetate) was demonstrated.

• Non-canonical phenolic copolymers bearing 4 or 5 hydroxyl groups per ring [P(4HS-co-S), P(5HS-co-S)] exhibit strong underwater adhesion across diverse substrates. Adhesion strength increased with M_n from ~20 to ~40 kDa (from ~3.4 MPa to ~6.0 MPa) and plateaued above ~40 kDa; M_n ~70–80 kDa used thereafter.
• Rapid bonding: P(5HS-co-S) reached ~1.4 MPa after only 10 s underwater (5–100× stronger than commercial/literature benchmarks for that timeframe), lifting a 10 kg weight after 2 min. After 72 h, both P(4HS-co-S) and P(5HS-co-S) reached a plateau of ~7.0 MPa, stable for at least 1 month.
• Hydroxyl number and positional effects: Adhesion strength increased approximately linearly with hydroxyl count: 1HS < 2HS < 3HS < 4HS ≈ 5HS, with ~1–2 MPa increments per additional OH up to 4. QCM adsorption showed the same trend (P1HS < P2HS < P3HS < P4HS ≤ P5HS). MD simulations indicated higher adsorption energies for 4HS and 5HS on Fe. Among 2HS isomers, 3,5-2HS (1.9 MPa) > 3,4-2HS catechol (1.8 MPa) > 2,3-2HS (1.5 MPa). Positions 2 and 6 contributed less to adhesion; certain 4HS isomers with adjacent OH near the backbone (2,3,5,6-4HS) showed reduced strength (4.3 MPa).
• Spatial confinement and composition: With similar total hydroxyl content, P(5HS11%-co-S89%) outperformed P(2HS26%-co-S74%) after 72 h (6.9 MPa vs 3.1 MPa). QCM showed lower water uptake for P(5HS11%-co-S89%) (1.4% water by polymer weight) vs P(2HS26%-co-S74%) (6.9%), and higher adsorption (~14 vs ~10 µg/cm2). Higher styrene content increased hydrophobicity and water repellence while local clustering of OH within 5HS created a favorable microenvironment at the interface.
• Interfacial and morphological observations: SEM of fractured surfaces showed more homogeneous, globular morphology for P(5HS11%-co-S89%) versus sponge-like morphology for P(2HS26%-co-S74%), indicating reduced water intrusion for 5HS-rich copolymers.
• Oxidation behavior: In air, ~30% of OHs in 4HS units and ~20% in 5HS units oxidized after 7 days; after 1 month, ~60% (4HS) vs ~20% (5HS). In water during adhesion, XPS after 7 days underwater showed only 10.0% (4HS) and 4.1% (5HS) oxidized (C 1s), with higher apparent oxidation at very surface likely post-fracture/measurement. GPC showed no changes in M_n/PDI after 2–4 weeks, indicating negligible crosslinking.
• Backbone spacing improves cohesion: Methacrylamide spacing (P(4HMA8%-co-S92%), P(5HMA9%-co-S91%)) increased cohesion and set unprecedented underwater tensile adhesion: 10.3 MPa (4HMA) and 9.9 MPa (5HMA) after ~1 week; work of adhesion ~11.8 kJ/m2. Adhesion strength increased with higher preload (e.g., P(4HS-co-S): ~7 → ~9 MPa; P(4HMA-co-S): ~10 → ~11 MPa from 25 g to 300 g). Crosshead speed (1–100 mm/min) had negligible effect, consistent with solidification during incubation.
• Broad substrate performance: P(4HMA8%-co-S92%) outperformed literature values on metals and polymeric substrates (Al, SUS, wood, glass, PTFE, PE), with particularly high strengths on metals. It also outperformed several commercial underwater adhesives on steel under matched mass conditions.
• Recyclability: Adhesives could be removed with common organic solvents and reused via brief redissolution. P(4HMA8%-co-S92%) maintained ~2 MPa after the third reuse cycle; even after four cycles, joints could lift ~8 kg underwater.

The study demonstrates that moving beyond bioinspired catechol/gallol motifs to non-canonical phenolic units with four or five hydroxyl groups dramatically enhances underwater adhesion. Increased hydroxyl count boosts interfacial interactions and adsorption, but equal or superior performance of 5HS at lower overall phenolic loadings arises from spatial confinement of hydroxyls on a single aromatic ring within a largely hydrophobic copolymer. This architecture establishes a hydrophobic microenvironment at the substrate–adhesive interface that repels water, maintains effective interfacial interactions (chelation, hydrogen bonding, electrostatics, hydrophobic interactions), and reduces phenolic oxidation during underwater setting. Positional analysis shows that hydroxyls at 2 and 6 are less effective due to steric and backbone proximity, explaining why performance plateaus from 4HS to 5HS and why certain isomers underperform. Minimal oxidation and unchanged GPC profiles indicate that adhesion is primarily non-covalent, differing from oxidation-mediated crosslinking often invoked in catechol/gallol systems. Further increasing the distance between the phenolic moiety and backbone via methacrylamide linkers enhances cohesive interactions and yields record-high underwater tensile adhesion (>10 MPa), approaching dry adhesive performance. The adhesives’ non-covalent nature also enables recyclability, providing practical advantages over crosslinking commercial glues.

This work introduces a new class of non-canonical phenolic polymers bearing four and five hydroxyl groups per ring that deliver ultrastrong, versatile underwater adhesion. Systematic tuning of hydroxyl number and position, molecular weight, and copolymer composition reveals that clustered hydroxyls combined with high hydrophobic content create favorable interfacial microenvironments that exclude water, suppress oxidation, and maximize interfacial interactions. Backbone modification with methacrylamide linkers further increases cohesion, achieving >10 MPa tensile adhesion underwater—surpassing reported literature and commercial underwater adhesives and approaching dry adhesive benchmarks. The non-covalent, minimally oxidizing mechanism allows solvent-based removal and reuse. Future work should focus on implementing living polymerization to better control molecular weight and dispersity of these non-canonical polymers and developing organic solvent-free underwater processing (e.g., coacervation or emulsions) to broaden applicability, including biomedical uses.

• The underwater adhesion process in this study requires organic solvents (e.g., CHCl3/MeOH), which may limit biomedical applicability.
• Living polymerization methods were not used; more precise control over molecular weight and dispersity remains a challenge for these new monomers.
• Although reusable, bond strength diminishes over repeated reuse cycles, potentially due to limited chain mobility during brief redissolution and oxidation upon air exposure.
• Certain positional isomers with high phenolic content can suffer from solubility issues (e.g., some 2,4,6- and 2,3,4-isomer copolymers at 10% phenolic content were insoluble, requiring 2.5%).