Revisiting the adhesion mechanism of mussel-inspired chemistry

Mussel-inspired chemistry has become an ideal platform to engineer a myriad of functional materials, but fully understanding the underlying adhesion mechanism is still missing. Particularly, one of the most pivotal questions is whether catechol still plays a dominant role in molecular-scale adhesion like that in mussel adhesive proteins. Herein, for the first time, we reveal an unexplored adhesion mechanism of mussel-inspired chemistry that is strongly dictated by 5,6-dihydroxyindole (DHI) moieties, amending the conventional viewpoint of catechol-dominated adhesion. We demonstrate that polydopamine (PDA) delivers an unprecedented adhesion of 71.62 mN m−1, which surpasses that of many mussel-inspired derivatives and is even 121-fold higher than that of polycatechol. Such a robust adhesion mainly stems from a high yield of DHI moieties through a delicate synergy of leading oxidation and subsidiary cyclization within self-polymerization, allowing for governing mussel-inspired adhesion by the substituent chemistry and self-polymerization manner. The adhesion mechanisms revealed in this work offer a useful paradigm for the exploitation of functional mussel-inspired materials.

Adhesion measurement in a symmetric configuration using SFA. The normalized force (F/R)distance (D) profile of mussel-inspired derivatives was in situ measured by SFA according to the previous reports. 1 -3 In a typical symmetric force measurement, back-silvered thin mica sheets (1-5 μm) were glued onto cylindrical silica disks (R = 2 cm) by an epoxy glue and then mounted into SFA chamber in a crosscylinder configuration. After that, a droplet of 70 L freshly prepared mussel-inspired derivatives Tris buffer solution was injected between two mica surfaces and the system was allowed to equilibrate for 30 min. The normal forces F between in situ polymerized coatings were detected as a function of surface separation D with the distance accuracy down to 0.1 nm. The measured normal force or "pull-off" force F ad was correlated to the adhesion energy per unit area between two flat surfaces W ad by F ad /R = 1.5πW ad , dictated by the Johnson-Kendall-Roberts (JKR) model. 4 Note that, the reference distance (D = 0) was determined at the contact between two bare mica surfaces in air before the introduction of the musselinspired precursor. To study the synergy of catechol and amine on the impact of PDA adhesion, we measured the adhesion of catechol and PEI 800 with various molar ratios from 1:2 to 1:1 and 2:1 as controls. For obtaining repeatable results, the force measurements were conducted on at least two pairs of surfaces independently prepared with three different positions on each pair of the surfaces under the same experimental conditions. Adhesion measurement in an asymmetric configuration using SFA. To measure the adhesion between PDA and different surfaces, we first fabricated Phe-, OH-, and N(CH 3 ) 3 + -terminated surfaces by functionalizing the gold-coated micas. Briefly, freshly gold-coated micas were immersed in 10 mM methanol solutions of phenylethyl mercaptan, 2-Mercaptoethanol, 2-(dimethylamino)ethanethiolhydrochloride for 24 h incubation. For the N(CH 3 ) 3 + -terminated surface, the as-formed 2-(dimethylamino)ethanethiolhydrochloride surface was required to carry out the quaternization treatment of the amine groups by the immersion into dichloromethane of CH 3 I. Finally, the functionalized micas were washed by ethanol and dried with nitrogen gas before use.
In a typical asymmetric force measurement, PDA-coated and functionalized mica sheets (1-5 μm) were glued onto cylindrical silica disks (R = 2 cm) by an epoxy glue and then mounted into SFA chamber in a cross-cylinder configuration. After that, a droplet of 70 L freshly Tris buffer solution was injected between two mica surfaces and the system was allowed to equilibrate for 30 min. The normal forces F between PDA and Phe-, OH-, and N(CH 3 ) 3 + -terminated surfaces were detected as a function of surface separation D with the distance accuracy down to 0.1 nm.
Simulation calculation of the standard electrode potentials of catechol-to-quinone translation and the energy barrier of Michael addition. Geometry optimizations of all molecules were performed by density functional theory (DFT) at the M06-2X level of theory 5 with 6-311+G(d) basis set 6 . Single point energies were calculated at the M06-2X/6-31+G(d) level of theory including Grimme's D3 (zerodamping) dispersion corrections. 7 In particular, when calculating the oxidation-reduction potential, the gaseous ground state single point energies of all molecules were calculated by the thermodynamic combination method CBS-QB3. 8 Energy minima and transition states were verified through vibrational analysis. 9 All minima were found to have no imaginary frequency, while all transition states had a single imaginary frequency. The associated eigenvectors were confirmed to correspond to the motion along the reaction coordinate using the intrinsic reaction coordinate (IRC) method. 10 All calculations in aqueous solutions were used the SMD model of the self-consistent reaction field (SCRF) to describe the influence of the solvent. 11 All calculations were conducted with the Gaussian 16 software package. 12 Optimized structures were illustrated using CYLview. 13 The redox potential of dopamine and its derivatives can be calculated from Nerst equation by the Gibbs free energy change in the aqueous solution (ΔG (sol) ) in the following thermodynamic cycle.

G (sol)
Where the Gibbs free energy changes at each step (ΔG (sol,i) ) were obtained by DFT calculations. The oxidation-reduction potential relative to the standard hydrogen electrode (SHE) can be calculated by the Nernst equation: Where n is the number of transferred electrons and F is the Faraday constant (F= 23.061 kcal mol -1 ).

Molecular-scale simulations of the electron density and adsorption energies of cation- interactions.
To reveal the electron density of the cyclized mussel-inspired derivatives, the density functional theory

Calculation of pair interaction energy between DHI/catechol and different groups
Molclus 14 was used to randomly generate 30 molecule pairs with different relative positions, and xtb 15 was invoked to roughly screen the lowest-energy configurations at GFN1-xTB level of theory 16 Figure S7. Adhesion between various coatings (polycatechol, poly(catechol-amine), and PDA) and mica using an asymmetric SFA measurement configuration. The results indicate that the adhesion between PDA and mica (12.7 mNm -1 ) is 3-fold and 15-fold higher than that of poly(catechol-amine) (4.2 mNm -1 ) and polycatechol (0.83 mNm -1 ), respectively. Figure S8. (a) Electrostatic potential (ESP) surfaces of DHI moieties and catechol. The ESP distribution at van der Waals surface was calculated at a M062X-D3/def2-TZVP level of theory considering the aqueous environment via SMD implicit solvent model. [19][20][21] (b) Ratio of electrostatic potential below -20 kcalmol -1 for catechol and DHI. The result shows that DHI has higher electron densities than catechol, owing to its inherent -conjugated structure. Figure S9. Adsorption energy of catechol-based moieties (catechol and DHI) and various groups (Benzene, CH 3 , NH 4 + , OH). Here, the interactions between these groups and DHI/catechol represent the - interaction, hydrophobic interaction, cation- interaction, and hydrogen bond, respectively. The simulation results strongly support that DHI moieties have better capacities to form these four interactions than that of catechol group.