Weak Hydrogen Bonds Lead to Self-healable and Bio-adhesive

4 days ago - Furthermore, the hydrogel can serve as a robust template for bio-mineralization to produce hydrogel composite that facilitate cell attach...
0 downloads 8 Views 1MB Size
Subscriber access provided by READING UNIV

Article

Weak Hydrogen Bonds Lead to Self-healable and Bio-adhesive Hybrid Polymeric Hydrogels with Mineralization-Active Functions Lijing Teng, Yunhua Chen, Min Jin, Yongguang Jia, Yingjun Wang, and Li Ren Biomacromolecules, Just Accepted Manuscript • DOI: 10.1021/acs.biomac.7b01688 • Publication Date (Web): 23 Jan 2018 Downloaded from http://pubs.acs.org on January 24, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Biomacromolecules is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Weak Hydrogen Bonds Lead to Self-healable and Bio-adhesive Hybrid Polymeric Hydrogels with Mineralization-Active Functions Lijing Teng,§,‡ Yunhua Chen,*,†,‡ Min Jin,† Yongguang Jia, ‡ Yingjun Wang†,‡ and Li Ren*,†,‡ §



School of medicine, South China University of Technology, Guangzhou 510006, P. R. China

School of Materials Science and Engineering, South China University of Technology,

Guangzhou 510640, P. R. China ‡

National Engineering Research Center for Tissue Restoration and Reconstruction, South China

University of Technology, Guangzhou 510006, P. R. China

Keywords: Weak hydrogen bonds, Self-healing hydrogels, Bio-adhesion, Mineralization

ACS Paragon Plus Environment

1

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 36

ABSTRACT: Hydrogels with self-healing features that can spontaneously repair themselves upon mechanical damage are increasingly attractive for biomedical applications. Many attempts have been made to develop unique hydrogels possessing this property, as well as stimuliresponsiveness and biocompatibility, however, the hydrogel fabrication strategies often involve specific design of functional monomers that are able to optimally provide reversible physical or chemical interactions. Here, we report that weak hydrogen bonds, provided by oligo (ethylene glycol) methacrylate (OEGMA) and methacrylic acid (MAA), a monomer combination that commonly used to prepare chemically crosslinking hydrogels, can generate self-healable hydrogels with mechanically resilient and adhesive properties through a facile one-step free radical copolymerization. The hydrogen bonds break and reform, providing an effective energy dissipation mechanism and synergic mechanical reinforcement. The physical properties can be simply tuned by OEGMA/MAA ratio control and reversible pH adjustment. Furthermore, the hydrogel can serve as a robust template for bio-mineralization to produce hydrogel composite that facilitate cell attachment and proliferations. This work is synthetically simple and dramatically increases the choice of amendable and adhesive hydrogels for industrial and biomedical applications.

ACS Paragon Plus Environment

2

Page 3 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

INTRODUCTION The remarkable ability of living organisms to spontaneously repair their damage and functions, which has been found in many biological tissues including bones,1 woods2 and skins,3 has attracted intensive attentions and inspired researchers to explore synthetic materials that can self-heal,4 in particular self-healing hydrogels, which are being increasingly utilized for biomedical applications, such as regenerative medicine, drug delivery, tissue engineering and implantable bioelectronics.5 The self-healing feature also offers great benefit for the artificial structural biomaterials as it could promote longer working life of the resulting products.6 However, for most synthetic hydrogels, achieving self-healing character has remained elusive because of the presence of irreversible cross-links, which originates from their very low resistance to crack propagation due to the lack of efficient energy dissipation mechanism in the hydrogel network.7 Nevertheless, reversibility in hydrogels can be accomplished by either employing dynamic covalent bonds8 or relying on reversible physical interactions including hydrogen bonding,9-11 ionic bonding,12 host-guest interactions,13-15 metal-ligand coordination16 and even hydrophobic associations.17 The introduction of reversible cross-linking will increase the overall viscoelastic dissipation of the hydrogel sample through imparting dissipative mechanisms at molecular levels. Among the physical interactions, hydrogen bonding, particularly cooperative hydrogen bonding, plays an important role in the construction of secondary and tertiary structures of biosystems.18,19 Up to date, hydrogen bonding has been widely employed to fabricate synthetic tough hydrogels and other functional materials. For example, a self-complementary quadruple hydrogen-bond array named UPy (ureido-pyrimidinone) with a dramatically high intermolecular

ACS Paragon Plus Environment

3

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 36

bonding association constant (6×107 M-1 in chloroform) can be imparted into polymers to prepare modular and injectable thermoplastic elastomers,20,21 moreover, with the support of hydrophobic micro-environment, extremely stretchable and fast self-healing hydrogels can be generated.22 We also developed UPy moieties functionalized macroporous hydrogels for promising cell adhesion.11,23 Recently, weak complementary hydrogen bonding associated monomers, for example, N, N-Dimethylacrylamide (DMAA) and methacrylic acid (MAA), have been utilized to produce tough hydrogels with shape memory behaviors. Due to the hydrophobic methyl substituent, the MAA and DMAA groups undergo strong association, which leads to the formation of hydrogen bonding clusters and generates tough hydrogels.9,24 However, the selfrepairing ability was not explored, and for most current hydrogen bond crosslinked self-healable hydrogels, the fabrication approaches often involve employing specifically designed monomers, such as acryloyl derivatized 6-aminocaproic acid25 and glycinamide.26 The development of selfhealable hydrogels with highly tunable multi-functions in a facile route is still strongly desirable, but also remains a challenge. In this work, we report that weak hydrogen bonds can generate self-healable hydrogels with mechanically resilient and adhesive properties through a simple one-step free radical copolymerization of oligo (ethylene glycol) methacrylate (OEGMA) and methacrylic acid (MAA), without using crosslinking agents. The formed hydrogel network contains tunable composition of covalently-bonded backbones and hydrogen bonded nanoclusters (as sacrificial and recoverable crosslinks). OEGMA and MAA monomer combination has been chemically crosslinked to prepare mucosal-adhesive hydrogels for oral delivery of proteins.27 The molecular design of the oral delivery system was based on the capability of PEG tethered chains to promote mucosal and protein protection and the calcium chelating capability of the carboxylic pendent

ACS Paragon Plus Environment

4

Page 5 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

groups to promote epithelial cell junction opening. Our choice of this monomer couple here was based on the following three reasons. First, the electronegative O atom of OEG is known as a strong hydrogen-bond acceptor, while MAA is a potent hydrogen-bond donor. The hydrogen bonding interactions of OEG/MAA will endow the hydrogels with energy dissipating mechanism and self-healing ability.28-30 Second, the physical hydrogen bonding linkages can also be tailored through monomer ratio variation to generate tunable physical properties. Moreover, the reactive carboxyl groups can provide hydrogels further functionalizing potentials. The physical properties of resulting hydrogel are highly tunable and reversible via OEG/MAA ratio variation and pH adjustment, and the hydrogel can serve as a robust template for bio-mineralization to produce hydrogel composite that facilitates cell attachment and proliferations. MATERIALS AND METHODS Materials Methacrylic acid (MAA, 99%), acrylic acid (AA, 99.5%), oligo(ethylene glycol) methyl ether methacrylate (OEGMA, average molecular weight Mn 500), oligo(ethylene glycol) methacrylate (-OH terminated OEGMA, average molecular weight Mn 500), potassium persulfate (KPS, 99%), N,N,N’N’-tetramethylethylenediamine (TEMED, 99%), poly(ethylene glycol) diacrylate (PEGDA, average molecular weight Mn 600), poly(acrylic acid) (PAA, average Mv 450,000), poly(ethylene oxide) (PEO, Mv 10,000) were used as purchased (Sigma-Aldrich) without further purification. Calcium chloride anhydrous (CaCl2, AR) and dipotassium hydrogen phosphate trihydrate (K2HPO4·3H2O, AR) were purchased from Sinopharm and used as received. Simulated body fluid (SBF, pH=7.4) was used as purchased (MACGENE Biotech). Deionized water was used in all experiments.

ACS Paragon Plus Environment

5

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 36

Preparation of hydrogels crosslinked by weak hydrogen bonding All hydrogels were synthesized by one-step free radical copolymerization of MAA and OEGMA. Briefly, different proportions of MAA to OEGMA were dissolved in deionized water with designed total monomer concentration. The initiator KPS (0.2 wt%, with respect to the water weight) was added with nitrogen bubbling for 10 min to remove the dissolved oxygen which acts as free radical scavengers, then the accelerator TEMED (0.1 wt%, with respect to the water weight) was added to the solution. Afterwards, the resulting solutions were sealed with Teflon tape and then placed in 25 °C water bath for 24 h to form hydrogels. In this work, the mass ratios of OEGMA/MAA and total monomer concentrations were set as the variations. Hereafter, samples will be denoted as OEG/MAA-x-y-z, where x and y are the mass ratios of OEGMA relative to the MAA, z is the total monomer fraction. Mineralization of hydrogels through alternate soaking process Mineralized hydrogels were prepared through alternate soaking process as described in the literatures with slight modification.31-33 In brief, the hydrogel was first immersed in SBF (pH=7.4) overnight to reach the equilibrium. This process could adjust the pH within hydrogel and simultaneously the hydrogel could be partially crosslinked by calcium ions (2.5 mM) in SBF. The hydrogel was then immersed in 0.5 M CaCl2 (pH=8) for 1 h and washed with deionized water to remove the superfluous calcium ions. Subsequently, the hydrogel was soaked into 0.3 M K2HPO4 (pH=9) solution for 1 h and washed with deionized water to remove the superfluous HPO42-. This alternative immersion process was repeated three cycles to mineralize calcium phosphate. After the alternative immersion, the matrix was then equilibrated in SBF for 6 h, and briefly rinsed with deionized water.

ACS Paragon Plus Environment

6

Page 7 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Characterization The hydrodynamic size of nanoclusters generated by the polymerization of monomers was measured by dynamic light scattering (DLS) using Zetasizer Nano ZS (Malvern, UK). The Samples (1 uL) were taken out and diluted 1000 times for DLS analysis. Atomic force microscope (AFM) images were acquired using an Asylum MFP-3D (Asylum Research, USA) under a tapping mode. Fourier-transform infrared (FTIR) spectra of PEO, PAA and PEO/PAA composites were measured on a Bruker VERTEX70 FTIR spectrometer. DSC analysis was carried out using a DSC Q2000 (TA Instruments) under nitrogen flow. Samples were weighted in an aluminum pan sealed with an aluminum lid. The samples were analyzed using a heat/cool/heat mode to eliminate their thermal history. To that purpose, samples were heated from 25 °C to 100 °C at 10 °C/min, held for 3 min, then rapidly cooled down to 50 °C at 10 °C /min to reach the amorphous state, and then heated to 100 °C at 5 °C /min. The glass transition temperatures were read from the second heating traces. Lyophilized hydrogels were sputter coated with platinum and imaged using field-emission scanning electron microscope (FEI Quanta NANO 350) to study the micro-structures, and the composition of hydrogel-mineral composites was studied using an EDX energy-dispersive spectrometer. The hydrogels were frozen with liquid nitrogen and immediately lyophilized for 48 h before observation. The actual Ca/P content was quantified by thermogravimetric analysis (STA449C/4/G, Germany). Prior to the analysis, all hydrogel samples were completely dried in vacuum oven and

ACS Paragon Plus Environment

7

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 36

heated to 800 °C at the rate of 10 °C /min-1 under an N2 atmosphere. Uniaxial tensile tests were performed on the as-prepared cylindrical hydrogel samples with dimension of 8 mm in diameter and 50 mm in length using a Shimadzu AG-X plus tensile tester with crosshead speed of 100 mm/min at room temperature. Three parallel samples per measurement were performed, and the obtained values were averaged. The strain was defined as the length change relative to the original length l0, the elongation ratio λ = l/l0 was the deformed length l to the original one l0. The elastic modulus was determined from the slope of the initial linear region of the stress-strain curve. The compressive stress-strain measurements were performed by using a dynamic mechanical analyzer (DMA Q800, USA) under an unconfined compression mode. The maximum compressive force of the instrument is 18 N. The cylindrical hydrogel samples, 12 mm in diameter and 5 mm in thickness, were put on the lower plate and compressed by the upper plate, the linear ramp force 3 N/min up to 18 N was designed to conduct the tests. The strain from 10% to 15% was used to determine the values of the compressive moduli.

34

Then, the shape recovery properties of the hydrogels were characterized

by 6 cycles of ramp force loading and unloading. At least three duplicates were tested for each hydrogel, and the mean value was calculated. The dissipated energy was calculated from the area between the loading-unloading profiles. Swelling tests were used to determine the swelling ratio (SR) and stability of hydrogels. The hydrogels were immersed into 20 mL of acidic bath (pH 4.0), and PBS (pH 7.4), respectively. Then the samples were put into sealed vials at 37 °C. When reaching the preset time interval, the hydrogels were taken out from the solution and the superficial water was removed by filter papers, and then the swelling hydrogels were weighed. SR was calculated using the following equation:

ACS Paragon Plus Environment

8

Page 9 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

ܴܵ =

ܹ௧ − ܹ௜ ܹ௜

where Wi and Wt represents the initial weight of the wet hydrogels and the weight after swelling equilibrium, respectively. Each test was repeated three times. Self-healing test of hydrogels The self-healing efficiency was characterized by comparing the tensile fracture behaviors of the virgin sample and the healed sample. Cylindrical hydrogel samples (8 mm diameter × 50 mm length) were cut into two pieces, and the cut parts were brought into contact at 25 °C immediately. To ensure good contact between the cut surfaces, a slight handing pressure was applied to the two cut pieces. During the self-healing process, no stress was applied. After different self-healing time, the self-healed samples were then subjected to stress-strain tests (Shimadzu AG-X plus tensile tester) at room temperature at a pulling rate of 100 mm/min. The elongation ratio of the pristine (lb,0) and healed (lb) hydrogels were used to determine self-healing efficiency (SE): 6

SE =

l l

b

×100

b ,0

The pH responsive self-healing of hydrogels was also studied. Disk-shaped hydrogels were put together in an acidic solution (pH=4.0) for 5 min, and then immersed in sodium hydroxide solution (pH=12) for 5 min. The separated hydrogels were then briefly rinsed in deionized water to remove excess sodium hydroxide solution and reintroduced into an acidic solution (pH=4.0) and kept the surfaces in contact for less than 5 min. Adhesive strength test of hydrogels

ACS Paragon Plus Environment

9

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 36

The adhesive strength test of the hydrogel was evaluated as references described.35,36 Porcine skin tissues were used and cut into 20 mm × 50 mm rectangular shapes. Then, hydrogel slides were placed onto the surface of the fresh skin tissue and another skin was placed on the top of the polymer mixture. The contact area of two skin tissues is 20 mm × 20 mm. After that, the upper fixture was then withdrawn at a rate of 1 mm/min and the load-displacement data was recorded by a Shimadzu AG-X plus tensile tester software until the polymer became completely detached from the skin. All measurements were triplicate. In vitro cell assay Cytocompatibility was first investigated, mouse bone mesenchymal stem cells (BMSCs, passage 8, Cyagen Biosciences) were seeded in 24-well plates at a density of 2500 cells/cm2 and cultured in Dulbecco’s modified Eagle medium (DMEM, Gibco, Life Technologies, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, Life Technologies, USA) and 1% penicillin/streptomycin overnight to spread. Afterward, the extracts of the hydrogels (0.1g/mL) were added to replace the culture medium (with 1 mL per well) of the cells. The sample without hydrogel extract was served as the control group. The cell viability was determined at days 1, 3 and 5 by a CCK-8 assay according to the manufacturer’s instructions, and the absorbance of the solution at 450 nm was tested using an Elx 800 instrument (Biotek, USA) after incubation for 1 h. The relative cell viability (%) was calculated as follows:

Cell viability ( % ) =

[ A]test × 100% [ A]control

where [A]test is the absorbance of the test sample, and [A]control is the absorbance of the control sample. To study the cell proliferation, BMSCs were cultured in the same medium mentioned

ACS Paragon Plus Environment

10

Page 11 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

above. Hydrogel samples sterilized with electron beam irradiation at 16 kGy for 4 h. After pretreatment with culture medium for 12 h, the BMSCs were seeded on the surface of the hydrogel at a density of 1.5×104 per well. The culture medium was refreshed every two days. The cell proliferation onto the pristine hydrogel and mineralized hydrogel were determined quantitatively using theCCK-8assay according to the manufacturer’s instruction. After 1, 3 and 5 days of incubation, cell medium was removed and washed with PBS. Then 200 µL of CCK-8 working solution was added to each sample and incubated at 37 °C for 1 h. Subsequently, the supernatant medium was extracted and the absorbance at 450 nm was measured by Thermo 3001 microplate reader (Thermo Scientific, USA). The cell viability and morphology on the hydrogels were also assessed by Live-Dead staining kit (Dojindo Laboratories, Japan) according to manufacturer’s guidelines. Calcein AM and ethidium homodimer were used as staining dyes. Green-labeled cells represent viable cells with no membrane disruption (live cells), while red-labeled nuclei indicate cell necrosis (dead cells).37 After culturing for 1, 3 and 5 days, respectively, the hydrogel samples were washed with PBS and Live/Dead staining solution was added and incubated for 40 min. Then the cells were washed with PBS again and observed by fluorescence microscope (Eclipse Ti-U, Nikon, Japan). Statistical analysis All experiments were performed at least three times, and the data were presented as mean ± standard deviation. The statistical difference of samples was analyzed by using the paired Student's t-test. A statistically significant difference was reported if p < 0.05. RESULTS AND DISCUSSION Polymeric hydrogel formation

ACS Paragon Plus Environment

11

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 36

The hydrogels were prepared by one-batch radical copolymerization of OEGMA and MAA. Theoretically, for the constant total monomer concentrations the equivalent molar number of CH2CH2O- and -COOH groups (molar ratio close to 1:1) could generate the largest amount of hyrogen bonds that are important for hydrogelation and mechanical performance of hydrogels, however, with OEGMA/MAA mass ratio lower than 2:1 (molar ratio 3.4:1), we found the reaction mixture is turbid and phase separation occurs at room temperature (Figure S1 and Table 1), while the mixture becomes a clear homogeneous solution with the mass ratio of 2:1 and above, We conjecture that the formed OEGMA and MAA monomer complex is no longer miscible with water due to the abundant hydrogen bonds generated when the amount of carboxyl and oxethyl groups is close. OEGMA/MAA mass ratio of 2:1 was then chosen for the hydrogel synthesis, and the hydrogel with mass ratio of 3:1 serves a control sample in some cases.

Figure 1. Polymeric hydrogels generated by weak complementary hydrogen bonds. (a) Schematic

ACS Paragon Plus Environment

12

Page 13 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

illustration of hydrogen bonding nanoclusters as connecting joints to build physical crosslinking networks after the copolymerization of OEGMA and MAA. (b) The chemical structures of monomers studied in this work. (c) Photograph of the resulting products after 24 h of polymerization with monomer OEG/MAA mass ratio of 2:1. (d) Self-healing behavior of the OEG/MAA-2-1-25% hydrogel. (e) Hydrogel blocks bent and connected to form a ring shape, a strong lump formed after twisting and could withstand stretching.

There is a critical monomer concentration to induce the physcial gelation, as shown in Figure 1c, low weight fraction of the total monomers (10 wt%) led to a transparent polymer solution which can flow freely. An increase of monomer fraction to 15 wt% results in physcial gelation that is confirmed by the inverted vial test. However, the hydrogel remains very weak until reaching a polymer composition of 25 wt%, where a more rigid network can be formed. We expect the hydrogelation is attributed to the formation of nanoclusters or nanoparticles which are crosslinked by the inter- and intra-molecular hydrogen-bonds of -COOH and -CH2CH2O- groups and possible hydrophobic interactions of polymer backbones. The formation of nanoclusters was confirmed by DLS measurements and AFM imaging (Figure S2). After 24 h polymerization of OEG/MAA-2-1-10%, DLS measurement displays a unimodal distribution of particles sizes, with an average diamater about 31 nm. AFM imaging also reveals the broadly dispersed nanoparticles. The disk shape morphology of the nanoparticles could be due to the deformation of particles upon water evaporation. Based on these results, we conjecture the hydrogelation processas follows. At the early stage of the copolymerization of OEGMA and MAA, the initiator decomposes and reacts with monomers to generate monomer radicals. With continuous radical propagation, oligomeric chain radicals are formed. These oligomers will grow, then precipitate out and aggregate to form

ACS Paragon Plus Environment

13

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 36

primary particles due to the hydrogen bonding interations between -COOH and -CH2CH2Ogroups, which is analogous to the hydrophobicity induced particle nucleation in precipitation polymerization.

However,

unlike

the

low

monomer

concentration

in

precipitation

polymerization, here the concentration of OEGMA and MAA is relatively high, as the formed nanoparticles continuously grow, the distance between particles dramatically decreases, therefore, the susequent polymerization will facilitate the connection of these particles through radical termination and possible chain transfer reactions, finally resulting a 3D crosslinking networks with numerous nanoparticles homogeneously distributed (Figure 1a). Self-healing property of hydrogels Given that the breakage and recombination of reversible bonds (sacrificial bonds) could induce the damaged area to autonomously self-heal. We expect that the reversible hydrogen bonding in the soft matrix can afford self-healing properties. As shown in Figure 1d, we found that such weak hydrogen bonding cross-linked hydrogels exhibit spontaneous self-healing behavior at room temperature without imparting any healing agents or external stimuli (also see the supplementary video). Two disc-shaped hydrogels (one dyed with rhodamine B for visualization of the interface) were cut into two pieces using a razor and the two halves taken from each hydrogel were put together rapidly to have their freshly created fracture surfaces brought into contact (Figure 1d-ii and iii). A whole piece of integrated hydrogel disk was formed and could keep intact subjecting to its own weight (Figure 1d-iv). After 30 min, the self-healed hydrogel could withstand mechanical stretching without failure at the interface (Figure 1d-v). The self-healing ability of the hydrogels was also demonstrated in other processing shapes, for example rod-shape, as shown in Figure 1e, the hydrogel rod blocks could re-joint firmly at interfaces and form a ring, a hydrogel lump was generated after twisting the ring with tweezers

ACS Paragon Plus Environment

14

Page 15 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

due to the strong tackiness on the surfaces, which can also withstand large mechanical stretching. Additionally, we brought 6.0-cm-long hydrogel blocks into contact, after 24 h they merged into an integral hydrogel string and could be stretched to a great extent without breaking, as shown in Figure S3, displaying its remarkable self-healing ability. To quantitatively evaluate the self-healing ability of the hydrogels, uniaxial elongation measurements have been performed on cylindrical OEG/MAA-2-1-25% hydrogels that have been cut and the pieces pressed together for a specified healing time (Figure 2a). Similar to the observation of reported hydrogen-bonded self-healing rubber system,38,39 longer healing time leads to better healing efficiency, with optimal healing of up to 90% recovery of extensibility relative to a pristine sample (Figure 2a and 2b), which can be attributed to the dangling hydrogen bonds exposed on the freshly cutting surface being more likely to find co-facial interactive partners upon increasing contact time. The physical cross-linking networks formed between the hydrogen atoms of the carboxylic groups from PMAA chains and the oxygens in oxethyl groups of the OEG chains brought viscoelastic properties to the resulting hydrogels. At large deformation covalent bonds may be broken but these physical bonds could rebuild and to some extent self-heal the overall network structures (Figure 2c). Moreover, the stress-strain curves of cut hydrogels being healed for 24 h at 37 oC were also investigated. The healing efficiency at 37 oC can reach up to 85%, which is comparable to that value (80%) at room temperature (Figure S4), indicating the trivial effect of temperature on the healing efficiency. The designed OEG/MAA-2-1-25% hydrogel demonstrates that the hydrogen bond associations are reversible responding to pH conditions. As shown in Figure 2d, when two disc-shaped hydrogels contact each other, the hydrogen bonds are established for healing at pH=4, and the healed samples are able to sustain large deformations and recover their size and shape when the

ACS Paragon Plus Environment

15

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 36

stress is released, which is attributed to the synergistic effect of complementary hydrogen bonding between ethoxy groups and protonated carboxylic groups occurs at the interfaces. On the contrary, the healed hydrogel separated into two pieces again when exposed to basic solution (pH=12) as a result of electrostatic repulsions respecting to the deprotonation of terminal carboxylic acid groups. Remarkably, the separated hydrogels are able to re-heal upon reintroduction into a low-pH environment, indicating the total reversibility of self-healing behavior. Reversible healing phenomenon was also observed between pH 7 and pH 4, as shown in Figure S5.

Figure 2. The self-healing tests of OEG/MAA-2-1-25% hydrogel. (a) The stress-strain curves of uncut

ACS Paragon Plus Environment

16

Page 17 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

hydrogel and cut hydrogels being healed at varied time under identical conditions. (b) Self-healing efficiency of the hydrogel as a function of healing time. (c) Hydrogel recovery due to the reversible formation of the hydrogen bonded nanoclusters. (d) Hydrogels heal in low pH solution and withstand strong stretching. Synergistic effect of complementary hydrogen bonding between ethoxy groups and pronated carboxylic groups occurs at the interfaces. While the hydrogel pieces separate in a high pH solution, displaying no healing ability due to the lack of hydrogen bonding formation attributed to the deprotonation of carboxylic groups. The hydrogels can re-heal upon exposure back into the low pH solution.

We also directly utilized polyethylene oxide (PEO) and polyacrylic acid (PAA) polymers to investigate the physical interactions between the ethoxy and carboxyl groups. After the aqueous PAA solution (concentration of 100 mg/mL) mixing with PEO solutions (varied concentrations), instant precipitation occurred (Figure S6a). We conjecture this phenomenon could attribute to the hydrogen bonds between the ethoxy and carboxyl groups, which results in the formation of insoluble PEO/PAA complex. The generated PEO/PAA complex behaves like soft elastomer and can be remolded into desired shape. The hydrogen bonds are also confirmed by FTIR spectroscopy characterization (Figure S6b). Pure PAA polymer by itself can form intramolecular hydrogen bonds through COOH dimerization, the absorption peaks located at 1710 cm-1 correspond to the COOH groups. When PEO is added, a new peak shift at 1728 cm-1 appears (Figure S6c), which could be attributed to the intermolecular hydrogen bonds between PAA and PEO.40 Moreover, the DSC curves of the PEO/PAA composites clearly show the strong effect of PAA proportion on the glass transition temperature of the composite, also verifying the hydrogen-bonding interactions between the two polymers, as shown in Figure S7.41 The complementary hydrogen bonding interactions are essential to gel the polymer chains

ACS Paragon Plus Environment

17

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 36

but might not be only key contributor for the self-healing performance of our hydrogels. The copolymerization of three other monomer combinations with CH2CH2O- and -COOH groups were first investigated, including -OH terminated OEG/MAA, -OH terminated OEG/AA and CH3 terminated OEG/AA couple. It is interesting to note that when MAA was added into the aqueous -OH terminated OEG solution, the mixture became turbid and phase separation occurred subsequently at room temperature, which indicates the difficulty to produce homogeneous hydrogel with this monomer combination. Unlike -CH3 terminated OEG/MAA hydrogels, both OH terminated OEG/AA-2-1-25% hydrogel and -CH3 terminated OEG-AA-2-1-25% hydrogel with the same mass composition and water fraction display no self-healing properties (Figure S8). We conjecture that the poor self-healing ability could attribute to the absence of -CH3 groups of PAA chains. -CH3 groups on the polymer backbones can increase the hydrophobicity of polymers, and a more hydrophobic microenvironment could significantly promote hydrogen bonding association thus benefit for the physical network rebuilt, therefore, without -CH3 groups, the self-healing performance is limited. While in the case of -OH terminated OEG/AA, the resulting hydrogel is quite rigid and no healing behavior was found. This is probably due to the self-crosslinking of polymer chains resulting from the extra radical chain transfer reactions of OH terminated OEGMA,42-44 which eventually limits the polymer chain movements. Overall, the excellent self-healing performance of the -CH3 terminated OEG/MAA-2-1-25% hydrogels could be mainly attributed to the formed hydrogen-bonded clusters mediated by hydrophobic interactions. As chemical crosslinking could limit polymer chain movements, the crosslinking density might have dramatic effect on the self-healing performance of hydrogels. A common cross-linker polyethylene glycol diacrylate (PEGDA) with different concentrations was introduced into the

ACS Paragon Plus Environment

18

Page 19 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

OEG-MAA-2-1-25% hydrogels on purpose to investigate this effect. As shown in Figure S9, it is clear that the hydrogel displays very weak self-healing ability when only 0.2 wt% PEGDA was employed, and no self-healing behavior can be found for the hydrogels with 1 wt%, 2 wt% of PEGDA. Since the movement of polymer chains and hydrogen bonded nanocluster re-unions are restricted by the crosslinking points, the chemically-crosslinked hydrogels are difficult to acquire mechanical healing ability. Mechanical properties of hydrogels The broadly dispersed hydrogen bonding nanoclusters as sacrificial cross-linkers could break and recombine during the stretching process, providing good mechanical performance to the hydrogels. Moreover, by varying the relative concentrations of the OEG/MAA mass ratio and monomer concentrations, we can control both the density of hydrogen bonding and permanent crosslinking that directly affect the mechanical properties of resultant hydrogels. Figure 3a and b shows the tensile stress-strain curves of hydrogels synthesized with different monomer concentrations with a fixed OEG/MAA mass ratio. Figure 3c and Figure 3d depict two properties: the breaking elongation and Young’s modulus as a function of the monomer concentrations. It is clear that the breaking elongation and Young’s modulus of hydrogel with OEG/MAA mass ratio 2:1 are considerably higher than those values of hydrogel with OEG/MAA mass ratio 3:1, which is attributed to the more amounts of hydrogen bonds generated in the former. Similarity, the higher total monomer concentrations could allow for a higher number of hydrogen bonds in hydrogels, which results in high crosslinking density (higher modulus) and lower breaking elongation, for example, the OEG/MAA-2-1-35% displays a breaking elongation of 158±7% but with a highest modulus of 13.7±0.9 kPa (Table 2).

ACS Paragon Plus Environment

19

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 36

Figure 3. Tensile mechanical behaviors of hydrogels. (a) Tensile stress-strain curves of hydrogels with OEG/MAA ratio of 2:1 under varied total monomer concentrations. (b) Tensile stress-strain curves of hydrogels with OEG/MAA ratio of 3:1 under varied total monomer concentrations. (c) The breaking elongation and (d) the Young’s modulus of the hydrogels.

Hydrogel samples were also tested for their mechanical behaviors under compression. Figure 4a and b show the compressive stress-strain curves of the hydrogels. From the resulting compressive stress-strain plots, the compressive modulus of hydrogels was determined. The OEG/MAA mass ratio and monomer concentrations have considerable effect on the stress-strain behavior and compressive modulus. It can be seen that a higher monomer concentration led to a higher compressive modulus, for example, OEG/MAA-2-1-35% had the highest value of compressive modulus at about 55 kPa. Meanwhile, the compressive modulus decreased as the mass ratio of OEG to MAA increased with a fixed monomer concentration (Figure S10).

ACS Paragon Plus Environment

20

Page 21 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

Figure 4. Compressive and recovery properties of hydrogels. (a) Compressive stress-strain curves of hydrogels with a OEG/MAA ratio (2:1) in different monomer concentrations. (b) Compressive stressstrain curves of hydrogels with a OEG/MAA ratio (2:1) in different monomer concentrations. (c) Cyclic compressive stress-strain curves. (d) Stress-time curves at 50% strain. (e) The area of the hysteresis loop of cyclic compressive stress-strain curves. (f) Hydrogel exhibiting an excellent ability to withstand compression and (g) high resistance to fracture when compressed with a sharp blade.

To study the self-recovery behavior, a OEG/MAA-2-1-25% hydrogel sample was subjected

ACS Paragon Plus Environment

21

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 36

to six consecutive loading-unloading processes at 50% strain without any lapse time between the cycles. As shown in Figure 4c, a hysteresis loop occurred in each cycle between the loading curve and the unloading curve. This demonstrates the energy dissipation after each cycle. Figure 4d further plots the variation of the compressive stress over time upon cyclic loading and unloading tests, which also indicates the good self-recoverability of the hydrogel. The quantified results are shown in Figure 4e. It was clearly shown that the as-prepared hydrogel could dissipate energy effectively as much as 73 KJ/m3 at the first cycle. When the second test was conducted immediately, the dissipated energy recovered to 80% of its original value, clearly demonstrating the good shape recovery of our hydrogel,34,45 attributing to the contribution of hydrogen bonds which served as reversible sacrificial bonds and cracked to effectively dissipate energies. Moreover, our experiments suggest that all hydrogel did not fracture even under a maximum force and recovered to the initial shape immediately upon unloading, and the hydrogen bonding cross-linked hydrogels shows high resistance to fracture when compressed with a sharp blade (Figure 4f and g). Adhesive property of hydrogels The adherence of hydrogels is critically important in medical materials such as bio-binding agents utilized in surgical operations. The softness of hydrogels is an advantage in adhesion to tissues, as most tissues have relatively rough surfaces and undergo frequent movement. The adhesive ability of hydrogels is primarily attributed to the incorporated functional groups, which can interact and bind with surrounding tissues through chemical or physical linkages including imine, amide, urea and hydrogen bonding, etc.46 Figure 5a clearly shows the strong adhesion of the OEG/MAA-2-1-25% hydrogel to varied surfaces from glass slide, polypropylene (PP) dish, chromium and ceramic surface to human finger. More remarkably, the hydrogel can adhere and

ACS Paragon Plus Environment

22

Page 23 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

adapt to the shapes of substrates, for example, the hydrogel cylindrical string can firmly stick to the rectangular PP substrate to adapt the shapes even when the substrate is curvedly bent (Figure 5b).

Figure 5. Adhesion of OEG/MAA-2-1-25% hydrogel. (a) Hydrogel adheres to varied substrate surfaces, (i) glass slide, (ii) polypropylene, (iii) chromium, (iv) ceramics, and (v) human skin. (b) The hydrogel can adhere and adapt to the substrate (polypropylene) shapes with extreme bending circumstances. (c) The hydrogels adhere to varied objects and build multilayered complex. (d) Adhesion testing geometry. (e) Adhesive force-displacement curves of hydrogel adhering to porcine dermis, contact area 4 cm2.

In addition, the hydrogel could strongly bridge varied materials and form multilayered complex (Figure 5c). We characterized quantitatively the adhesive properties of hydrogels by a lap shear test (Figure 5d). Adhesive force-displacement curves of hydrogel/porcine skin multilayered complex are shown in Figure 5e. It is clear that the OEG/MAA-2-1-25% hydrogel exhibits a higher adhesion force (0.2 N/cm2) than that of OEG/MAA-3-1-25% hydrogel (0.1

ACS Paragon Plus Environment

23

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 36

N/cm2), which could be due to the larger amount of hydrogen-bonding clusters induced when more MAA is employed in the hydrogel.29 Overall, we conjecture the strong adhesion of our hydrogel to substrates could attribute to the synergic effect of multiple molecular interactions including van der Waals forces, hydrogen bonding, flexibility of chains and possible hydrophobic interactions contributed from the polymer backbones.47 Swelling behavior of hydrogels Dynamic swelling behavior of the OEG/MAA hydrogels was investigated in response to the changes of monomer concentration and pH. Figure S11 depicts the swelling plots of OEG/MAA-2-1 hydrogels with different monomer concentrations in PBS (pH 7.4). Owing to the correlation between the crosslinking density and monomer concentration, the swelling ratio and rate of OEG/MAA-2-1 hydrogels gradually decrease when increasing the total monomer concentrations, with OEG/MAA-2-1-35% hydrogel demonstrating the lowest swelling degree of approximately 8 in 36 h. Figure S12 shows the effect of medium pH on the swelling behavior of OEG/MAA-2-1-25% and OEG/MAA-3-1-25% hydrogels. At low pH value 4.0, the carboxyl groups of PMAA are protonated and hydrogen bonding network are well preserved within the hydrogels, which leads to lower swelling ratio and rate of both hydrogels comparing to the situation of high pH value 7.4, where carboxyl groups are deprotonated and hydrogen bonds are disassociated.48 Overall, the swelling degrees of these hydrogels can be modulated by adjusting the monomer concentration, OEG/MAA mass ratio or medium pH. Bioinspired mineralization of hydrogels and cell attachment It has been shown that synthetic hydrogels, for example hydrolyzed poly(2-hydroxyethyl methacrylate) hydrogel and poly(acrylic acid) hydrogel, could provide a versatile template for

ACS Paragon Plus Environment

24

Page 25 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biomacromolecules

the surface mineralization of calcium phosphate due to the strong interaction between carboxyl group and the calcium ions.49,50 Inspired by these findings, it is believed that the carboxyl groups of our OEG/MAA hydrogel can also chelate calcium ions, providing active nanoparticle nucleating sites and inducing biomimetic Ca/P mineralization. To carry out the Ca/P mineralization, we alternatively dipped the hydrogel (OEG/MAA-2-125%) into aqueous solutions of calcium chloride (CaCl2) and dipotassium hydrogen phosphate (K2HPO4). After repeated dipping, the transparent hydrogel gradually became opaque white with a slight swelling in volume. Figure 6a clearly shows the microstructures produced by Ca/P mineralization. The as-prepared hydrogel has highly smooth surface (Figure 6a-i), however, after mineralization numerous spherical calcium phosphate particles were formed, leading to rough surface morphology of mineralized hydrogel (Figure 6a-ii and iii). A view from the rough fracture surface of mineralized hydrogel also shows the wide distribution of calcium phosphate microspheres (Figure 6a-iv). To estimate the chemical compositions of the minerals, energy dispersive X-ray spectroscopy (EDS) analysis was performed. EDS results clearly reveal the presence of calcium and phosphorus elements in the mineralized hydrogels (Figure 6b). Additionally, TGA was performed to determine the total amount of generated calcium phosphate. The two stages of weight loss include the evaporation of occluded water at low temperature from 50-220 °C and the decomposition of organic components from 220-450 °C. Accordingly, the amount of inorganic composition in the mineralized hydrogels was determined to be about 24.2%. Overall, these results indicate that our hydrogel network could provide active nucleating sites for the formation of apatite due to the strong interaction between the carboxyl groups and the calcium ions. The compressive mechanical property variation of the hydrogel after mineralization was

ACS Paragon Plus Environment

25

Biomacromolecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 36

also evaluated. The stress-strain curves of the pristine and mineralized hydrogel are almost overlapped in the low strain range (