Boronic ester-based self-healing hydrogels formed by using intermolecular B-N coordination

Article abstract of DOI:10.1016/j.polymer.2020.122624

The copolymer poly(acrylamide-co-2-acrylamido glucose) (PA) and poly(acrylamide-co-N-acryloyl-3-aminophenylboronic acid-co-N-(3-dimethylaminopropyl) acrylamide) (PB) were separately prepared by radical polymerization, and the hydrogel was readily construc

Full text of DOI:10.1016/j.polymer.2020.122624

Journal Pre-proof
Boronic Ester-Based Self-Healing Hydrogels Formed by Using Intermolecular B-N
Coordination
Yusheng Liu, Yigang Liu, Qiuxia Wang, Yugui Han, Yebang Tan
PII:
S0032-3861(20)30455-9
DOI:
https://doi.org/10.1016/j.polymer.2020.122624
JPOL 122624
Reference:
To appear in: Polymer
Received Date: 17 December 2019
Revised Date: 16 April 2020
Accepted Date: 17 May 2020
Please cite this article as: Liu Y, Liu Y, Wang Q, Han Y, Tan Y, Boronic Ester-Based Self-Healing
Hydrogels Formed by Using Intermolecular B-N Coordination, Polymer (2020), doi: https://
doi.org/10.1016/j.polymer.2020.122624.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Ltd.

CRediT author statement
Yusheng Liu: Conceptualization, Methodology, Data Curation, Formal analysis,
Writing - Original Draft
Yigang Liu: Funding acquisition, Resources, Project administration
Qiuxia Wang: Visualization, Funding acquisition
Yugui Han: Validation Data, Curation
Yebang Tan: Conceptualization, Supervision, Project administration Writing -
Review & Editing

Graphical abstract
for
Boronic Ester-Based Self-Healing Hydrogels Formed by using
Intermolecular B-N Coordination
Yusheng Liu a,b, Yigang Liu b, Qiuxia Wang b, Yugui Han b, Yebang Tan a,b*

Boronic Ester-Based Self-Healing Hydrogels Formed by Using
Intermolecular B-N Coordination
Yusheng Liu a,b, Yigang Liu b, Qiuxia Wang b, Yugui Han b, Yebang Tan a,b
*
a School of Chemistry and Chemical Engineering, Shandong University, Jinan 250100,
People’s Republic of China；The Key Laboratory of Special Functional Aggregated
Materials, Ministry of Education, Shandong University, Jinan 250100, People’s
Republic of China
b CNOOC, Ltd, Tianjin Branch, Bohai Oilfield Research Institute, Tanggu Tianjin
300452, China
Corresponding author. Tel.: +86-531-88363502; Fax: +86-531-88564464
E-mail address: ybtan@sdu.edu.cn

Abstract
The copolymer
poly(acrylamide-co-N-acryloyl-3-aminophenylboronic
poly(acrylamide-co-2-acrylamido
glucose)
(PA)
and
acid-co-N-(3-dimethylaminopropyl) acrylamide) (PB) were separately prepared by
radical polymerization, and the hydrogel was readily constructed by simple mixing
the
two
kinds
of
copolymers.
The
amine-containing
monomer,
N-(3-dimethylaminopropyl) acrylamide was incorporated into the polymer chain to
enhance the stability of the complex between boronic acid and glucose units under
physiological condition. The mechanical properties of the hydrogel can be readily
engineered by adjusting the pH, solid content, and [phenylboronic acid]/[glucose]
ratios. Moreover, the hydrogel possesses excellent intrinsic self-healing ability.
Benefitting from the reversible property of the boronic ester, the hydrogel can be
dissociated by adding competitive sugar molecules, which makes it easy to remove
from medical gauze or skin. Additionally, cell cytotoxicity studies revealed that the
hydrogel has good biocompatibility. With its superior properties, this dynamic
phenylboronic ester-based hydrogel has great potential to be a component of new
generation of wound dressings.
Keywords: hydrogel, water soluble polymer; self-healing, rheological behaviour,
biocompatibility

1. Introduction
Recently, hydrogels have been considered as ideal materials for wound dressing
owing to their biocompatibility and high water content.[1-4] The high water content
provides a moist environment, which has been shown to be beneficial for wound
healing.[5] However, traditional hydrogel dressings adhere to or entangle with the
wound to some degree, which can cause secondary injuries to newly formed tissues
and thus additional pain to the patient during the removal of the dressing material.[6]
It is therefore necessary to develop easily removable dressing materials, which can be
degraded in a rapid and controllable manner to prevent secondary injuries to
patients.[7] Use of dynamic covalent bonds (DCBs), which can selectively undergo
reversible breakage and reformation without irreversible side reactions,[8, 9]
represents a new strategy for the design and preparation of wound dressings. The
dynamic nature of DCBs endows the resulting hydrogels with self-healing,
controllable degradation, and stimuli-response properties.
As a representative DCB, the boronic ester bond, which is formed through the
condensation of boronic acid and 1,2- or 1,3-diols, has been successfully used to
prepare hydrogels with self-healing and stimuli-response properties.[10-16] However,
the equilibrium between the breakage and reformation of the boronic ester bond is
controlled by the pH of the environment.[17-19] Typically, a dynamic bond forms

only at a pH near or above the pKa of the boronic acid derivatives, which is typically
8–10.[20] In aqueous media, phenylboronic acid (PBA) exists in equilibrium between
a hydrophilic triangular form and hydrophilic tetrahedral form, depending on the
pH.[20,21] This pH-dependent dynamic bond formation severely inhibits the practical
application of these materials.
Significant effort has been devoted to reducing the apparent pKa of PBA-based
materials in order to facilitate the formation of dynamic boronic esters under
physiological conditions.[22,23] A typical method is to synthesize boronic
acid derivatives with a lower pKa by modifying the structure of PBA with
electron-withdrawing groups.[10,24,25] However, the modification of the molecular
structure or synthesis of these special PBA derivatives often involves complicated
synthetic chemistry, thus limiting its applicability. Recent reports have shown that the
pH dependence and viscoelasticity of boronic ester-based hydrogels can be
engineered by designing the chemical structure of the polymer chain.[11,26]
Accordingly, a suitable amino group was introduced into the polymer backbone to
decrease the apparent pKa of the PBA moiety.[27,28] The amino groups present in
PBA-containing polymers enhanced the stability of the boronic ester owing to the
coordination interaction between PBA and the amide groups.[27,29-31] Okano et al.
used this approach to prepare microgels to control the release of insulin under
physiological conditions.[31] However, to the best of our knowledge, there has been
no report on boronic ester-based self-healing bulk hydrogels with excellent
physiological usability formed via B-N intermolecular coordination.

Herein, we report a boronic ester-based hydrogel formed under physiological
conditions as an ideal wound dressing material. The hydrogel was readily constructed
by simple mixing the boronic acid-containing poly(acrylamide-co-2-acrylamido
glucose) and diol-containing poly(acrylamide-co-N-acryloyl-3-aminophenylboronic
acid-co-N-(3-dimethylaminopropyl) acrylamide). To fabricate the hydrogel under
physiological pH, amine-containing N-(3-dimethylaminopropyl) acrylamide
(DMAPAA), was introduced into the polymer chain. The composition-property
relationship of the hydrogels in terms of the pH, [PBA]/[glucose] ratios, solid content
was investigated by the rheological study. The self-healing property and degradability
of the hydrogel were investigated comprehensively. Moreover, the effective
biocompatibility of the hydrogel renders it an ideal platform for developing easily
removable wound dressing materials.
2. EXPERIMENTAL SECTION
2.1 Materials
D-glucosamine hydrochloride, aminophenylboronic acid, acryloyl chloride,
2,2′-azobis(isobutylronitrile) (AIBN), and N-(3-dimethylaminopropyl) acrylamide
(DMAPAA) were obtained from J&K Scientific and used as received. Acrylamide
(AM), sodium carbonate, sodium hydrogen carbonate, hydrochloric acid, dimethyl
sulfoxide (DMSO), and acetone were obtained from Sinopharm Chemical Reagent Co.
and used as received.
2.2 Characterisation
1
The H NMR spectra were recorded on Bruker Avance 400 or 300 NMR

spectrometer. The feed ratios of the copolymers were determined by elemental
analyse using an Elementar Vario E1 III elemental analyser (Germany) at 1150 °C.
Static light scattering (SLS) was performed on a Dawn Heleos light scattering
instrument (Wyatt Technology) equipped 18-angle light scattering detectors.
Morphological observations were performed using a Hitachi S-4800 scanning electron
microscope (SEM). Samples of the hydrogel were freeze-dried and sputter-coated
with platinum. The rheological behaviour of the samples was characterised using a
Thermo Scientific Haake MARS III rheometer equipped with a 20 mm parallel plate
at 25 °C. All hydrogels were aged for 24 h before testing.
2.3 Synthesis of a 2-acrylamido glucose (AAG)
AAG was synthesized according the literature via the reaction of D-glucosamine
hydrochloride and acryloyl chloride.[32] D-glucosamine hydrochloride (8.6 g) and
NaNO2 (0.14 g) were dissolved in 40 mL of 2 M K2CO3 aqueous and cooled in a
low-temperature reactor at -5 °C with magnetic stirring. Acryloyl chloride (4.0 g) was
added dropwise to the solution. The mixture was reacted at -5 °C for 120 min, and
then warmed to room temperature for 24 h. Subsequently, absolute ethyl alcohols (200
mL) were poured into the reactant and then the white solid was removed by filtration.
The solution was concentrated by rotary evaporation and the crude product was
obtained. The crude product was purified on silica gel column chromatography and
recrystallization with a methanol/ethyl acetate mixture. The product was obtained and
1
the yield of the product is 41.5%. H NMR (400 MHz, D2O) δ 6.34 – 6.08 (m, 4H),
5.72 (dd, J = 10.0, 1.7 Hz, 2H), 5.15 (d, J = 3.6 Hz, 1H), 4.68 (d, J = 8.4 Hz, 1H),

3.96 – 3.33 (m, 10H).
2.4 Synthesis of N-acryloyl-3-aminophenylboronic acid (AAPBA)
3-aminophenylboronic acid (5.478 g) was dissolved in 80 mL of 2 M NaOH and
cooled in a low-temperature reactor at -5 °C with magnetic stirring for 10 min.
Subsequently, acryloyl chloride (5.20 mL) was added dropwise. The reaction mixture
was reacted at -5 °C for 30 min and then was carried out at room temperature for 120
min, and the pH of the reaction mixture was adjusted to 2 with a 2 M HCl solution.
The precipitate was filtered off and washed 3 times with 25 mL of cold deionized
water. The precipitate was dissolved in hot water (70 °C) and filtered off the insoluble
impurities. The hot filtrate was left to a refrigerator overnight and the product was
1
obtained by filtered off. The yield of AAPBA was 69.8%. H NMR (300 MHz,
DMSO-d6) δ 10.08 (d, J = 16.0 Hz, 1H), 8.01 (s, 2H), 7.94 – 7.77 (m, 2H), 7.50 (dt, J
= 7.3, 1.2 Hz, 1H), 7.30 (q, J = 8.5, 7.7 Hz, 1H), 6.45 (dd, J = 17.0, 10.0 Hz, 1H),
6.24 (dd, J = 17.0, 2.2 Hz, 1H), 5.73 (dd, J = 10.0, 2.2 Hz, 1H).
2.5 Preparation of copolymers
The copolymer, poly(AM-co-AAG) with 5.0 mol% AAG, referred to as PA-5,
was synthesized by radical polymerization. First, AM (6.752 g) and AAG (1.168 g,)
were dissolved in 52.8 mL of DMSO. The mixture was degassed under a nitrogen
atmosphere with magnetic stirring for 1 h, and then an initiator solution (1.0 mL)
containing 0.025 g of AIBN was injected into the reaction mixture. The mixture was
allowed to react at 70 °C for 24 h. After polymerization, the solution was transferred
to a semi-permeable membrane (MWCO: 3500) and purified against deionized water

for 7 d to remove impurities. The product was obtained after lyophilisation. Another
polymer referred to as PA-10 containing 10 mol% AAG was prepared in the same
way.
The copolymer, poly(AM-co-AAPBA-co-DMAPAA) with 5.0 mol% AAPBA,
referred to as PB-5, was also synthesized by radical copolymerization. First, AM (6.04
g), AAPBA (0.96 g), and DMAPAA (1.56 g) were dissolved in 57.1 mL of DMSO.
The mixture was degassed under a nitrogen atmosphere with magnetic stirring for 1 h,
and then an initiator solution (1.0 mL) consisting of 0.025 g of AIBN in DMSO was
injected into the reaction mixture. The mixture was allowed to react at 70 °C for 24 h.
After polymerisation, the solution was transferred to a semi-permeable membrane
(MWCO: 3500) and purified against deionized water for 7 d to remove impurities.
The product was obtained after lyophilisation. Another polymer referred to as PB-10
containing 10 mol% AAPBA was prepared in the same way.
2.6 Fabrication of hydrogels
Hydrogels were prepared by mixing PB and PA solutions with mild stirring at
room temperature. To investigate the effect of pH, the polymers were dissolved in
phosphate-buffered saline (PBS) solutions with different pH values (ranging from 5 to
11). The pH value in experiment was repeatedly adjusted by adding 1 M HCl aqueous
solution or 1 M NaOH aqueous solution to the PBS solutions. The solid contents were
kept constant at 5.0 wt%. The [PBA]/[glucose] ratio was adjusted by varying the
volumes of the PA and PB polymer solutions used for hydrogel fabrication.
2.7 Cell cytotoxicity

The cytotoxicity of the dynamic hydrogels was evaluated by Cell Counting Kit-8
(CCK-8) assay using both human embryonic kidney (HEK293) cells and HeLa cells.
Typically, hydrogels with different compositions and amounts were incubated in 4 mL
of the Dulbecco’s modified eagle medium (DMEM) (Corning, USA) for 24 h. Cells
were seeded in 96-well plates at a density of 7 × 103 cells/well and incubated at 37 °C
in a 5% CO2 atmosphere for 24 h. Then, the cells were incubated with either fresh
DMEM medium or DMEM medium containing gel extracts for another 24 h. Each
well was washed twice with PBS after removing the medium. Then, a mixture of
CCK-8 (Shanghai Yeasen Biological Engineering Technology, China) and DMEM
medium (1:10; 100 µL/well) was added to the wells and the cells were cultured for 2
h. The cell numbers were counted using a microplate reader (Model 680 Bio-RAD) at
450 nm; DMEM medium without cells but with added CCK-8 was used to record the
background signal.
2.8 Decomposition of hydrogels
The hydrogel formed from PB-5 and PA-5 in PBS (polymer content of 10 wt%,
[PBA]/[glucose] = 1) was used to investigate their responsiveness to sugar. To test the
sugar responsiveness, a hydrogel plate coloured by carmine was cut into two equal
pieces, and the halves were incubated in a PBS (pH 7.3) solution with or without 2 M
fructose. The dissociation of the hydrogels was monitored visually at different time
intervals. Moreover, a piece of hydrogel was attached to a medical gauze, and a
fructose-soaked gauze was applied to half the hydrogel to test its controllable
degradability.

3. Results and discussion
3.1 Characterization of copolymers and fabrication of hydrogels at various pH
The boronic acid-containing copolymer PB and diol-containing copolymer PA
were separately prepared by conventional radical polymerisation. The structures of
1
the polymers were characterized by H NMR spectra as shown in Figure 1, and the
molecular weight of the copolymers was determined by SLS (Table 1). The actual
molar ratios of monomers in copolymers were calculated from the results of elemental
analyse (Table S1). Rheological properties of the PB solutions with varying
concentrations were also investigated, as shown in Figure S3. For the PB-5 and PB-10
solutions, the loss modulus is always prevailing on the storage modulus over entire
investigated range of the oscillatory strain. Moreover, loss modulus dominates the
whole frequency range with increasing tendency of both storage and loss modulus.
The PB itself cannot form gels using B-N as cross-linking at the polymer
concentration use in experiments.

Figure 1. 400 M 1H NMR spectra of a) PA-5 and b) PB-5.The spectra were recorded
at 298 K in D2O.
Hydrogels were prepared by mixing boronic acid-containing PB copolymer and
diol-containing PA copolymer solutions (Scheme 1a). The amino groups present in the
boronic-containing polymer enhanced the stability of the boronic ester, which is likely
related to the electrostatic attraction between electron-deficient boron in the PBA
group and electron-rich nitrogen in the amino group at an intermediate pH. Scheme 1b
and Figure S4 qualitatively illustrates the intermolecular B-N interactions in the
polymer mixture. [33,34] It is clear that the middle system (pH = 7) has stronger
attractive interactions between boron in the PBA moiety and nitrogen in the amino
moiety. It is worth noting that other weak non-covalent interactions occur in parts

other than those involving B-N interactions, when the pH is greater than or less than
7.
Table 1 Characterization of copolymers.
b
b
Determined ratioa
(%)
Mw, app
A2 (10-3
Yield
(%)
sample Feed ratio (%)
(104 g/mol) mol mL/g2)
PB-5 85.0: 10.0: 5.0c 80.53: 14.11: 5.36c 7.29 ± 0.63 0.112
PB-10 80.0: 10.0: 10.0c 73.23: 17.56: 9.21c 8.10 ± 1.42 -0.098
92.4
93.8
87.5
86.4
PA-5 95.0: 5.0d
PA-10 90.0: 10.0d
92.12: 7.88d
11.2 ± 1.58 0.692
9.63 ± 1.88 0.743
83.37: 16.63d
b
c
a Calculated from elemental analyse；Obtained by SLS measurements； Molar ratios
of AM: DMAPAA: AAPBA in PB；d Molar ratios of AM: AAG in PA.
Hydrogel formation at varying pH values was qualitatively confirmed through a
vial inversion test, as shown in Figure 2a. The mixture was converted from a viscous
liquid to an elastic hydrogel with a change in pH from acidic to alkaline one. At pH
6.5, the system exhibited a semisolid state and flowed under gravity on a long-time
scale, indicating the formation of a hydrogel. Moreover, gelation occurred
immediately upon mixing and gently stirring the two polymer solutions. To obtain
further insight into the gel-formation kinetics, a dynamic time sweep experiment was
conducted to determine the hydrogelation process, as shown in Figure S5. The
hydrogel could form within 60 s, as indicated by the crossover point of Gʹ and Gʺ
curves. These results indicated that the hydrogel can formed with the pH > 6.5. To

demonstrate the influence of the intermolecular B-N coordination on this system, the
control experiments were carried out. The copolymer, poly(AM-co-AAPBA) with 5.0
mol% AAPBA and without the DMAPAA, referred to as PN-5, was synthesized.
Rheological properties of hydrogels (formed from PN-5 and PA-5, 5.0 wt%,
[PBA]/[glucose] = 1) at different pH were also investigated as shown in Figure S6 and
Figure S7. It was clearly that the hydrogel can formed when the pH > 8.2. The
existence of the DMAPAA in the polymer backbone can provide a weak alkaline
microenvironment due to the coordination interaction between PBA and the amide
groups, which can decrease the apparent pKa of the AAPBA.
Scheme 1. (a) Schematic illustration of the formation of the dynamic hydrogels. (b)

Non-covalent interactions of the polymer mixture system. Regions of red, green, and
blue indicate strongly repulsive, weakly attractive, and strongly attractive interactions,
respectively.
To quantitatively investigate the effects of pH on the mechanical properties,
hydrogels prepared at various pH were subjected to oscillatory rheological
measurements as a function of angular frequency. As shown in Figure 2b,
frequency-dependent modulus curves were observed for all hydrogels. At high
frequencies, the hydrogels behaved like an elastomer with G′ > G″, whereas at low
frequencies, a liquid-like state was observed with G′ < G″. Typically, unlike gels
cross-linked by permanent covalent bonds, gels cross-linked by dynamic covalent
bonds
usually
display
frequency-dependent
moduli.
This
reversible
association-dissociation mechanism often contributes to a characteristic relaxation
time τc, which characterises the dynamics of the system.
The τc value was obtained from the inverse of the crossover frequency fc, where G′
equals G″.[35,36] At low angular frequencies, there is sufficient time for the bonds to
restructure when perturbed because the time scale probed in the test is longer than the
lifetime of the reversible cross-linking, whereas the system does not have enough time
to dissociate at high angular frequencies. The hydrogel formed at pH 6.5 has the
highest fc of 0.16 Hz corresponding to a relaxation time τc of 6.67 s. With increasing
pH, τc increased as fc shifted to the low-frequency region. For example, τc increased to
7.38 s and 20.11 s for the hydrogels formed at pH 7.0 (fc = 0.135 Hz) and 10.0 (fc =
0.043 Hz), respectively. Moreover, the storage modulus in the frequency-independent

region increased with increasing pH. The plateau G′ values are summarised in Figure
2c. The improved mechanical properties and lifetime of the network can be
reasonably attributed to the increased binding affinity between the PBA group and
glucose moiety at an alkaline pH. The relative binding constants of PBA and sugar are
pH-dependent, and a higher pH favours the formation of the boronic ester bond under
specific conditions.[37,38]
Figure 2. (a) Photographs, (b) dynamic oscillatory angular frequency sweep, (c)
storage modulus of the hydrogels formed at different pH. (d) SEM images of the
dynamic polymer dry gels formed at different pH. The hydrogels were formed from
PB-5 and PA-5 (5.0 wt%, [PBA]/[glucose] = 1).
The structural features of the hydrogels were also characterised according to the
network parameters (Table S3) calculated from Equations 1–3.[39,40]

ꢀꢁ = ꢂꢃꢄ
1
2
3
ꢅꢆ
ꢇ
ꢈꢁ = ꢅꢆ ꢉ
ꢇ
ꢊ
ꢃꢄ
ꢋ = ꢌꢎ ꢈ ꢏꢐꢑ
ꢍ
ꢁ
where c (g/L) is the polymer concentration, R is the universal gas constant, NA is
the Avogadro constant and T is the temperature, the G′p was determined from the
frequency independent regime of the storage modulus.
The cross-linking density υe increased with increasing pH, and therefore, the
cross-linking distance ξ between effective cross-links decreased. The calculated
molecular weight of the polymer segment between effective cross-links Me was used
to evaluate the number of effective cross-links per polymer chain. To investigate the
interior morphology of the hydrogels formed at different pH values, SEM images
were obtained, and are shown in Figure 2d. All the hydrogels possessed an
interconnected porous network structure. The pore size decreased at an alkaline pH,
indicating an increase in the cross-linking density.
Table 2. Network parameters for polymer gels formed with varying pH.
pH
6.5
7.0
7.5
8.0
9.0
10.0
τc (s)
5.03
G′p (Pa)
21.5
Me (kg/mol)
5756.4
2381.4
981.6
νe (1021 m-3)
5.2
ξ (nm)
57.6
57.6
42.9
31.9
27.0
25.3
7.38
52.0
5.2
10.84
15.91
18.06
20.11
126.2
209.1
253.2
284.3
12.6
592.7
30.7
489.6
50.8
436.2
61.5

The rheological properties of the hydrogel formed from PB-10 and PA-10 were
also investigated (Figure S9). The moduli of the polymer gels formed by complexing
the PB-10 and PA-10 copolymers were much greater than those of the hydrogels
consisting of PB-5 and PA-5 at the same pH. Moreover, the crossover frequency
shifted to lower angular frequencies as compared with those of the hydrogels formed
from PB-5 and PA-5, indicating a slowing of the hydrogel dynamics.[36] Increasing
the number of cross-linking moieties in the polymer chain increased the number of
effective cross-links per polymer chain, which is correlated with τc.
3.2 Effect of [PBA]/[glucose] on hydrogels

Figure 3. (a) Dynamic oscillatory angular frequency sweep of the hydrogels formed
from PB-5 and PA-5 (pH = 7.0, 5.0 wt%) with different [PBA]/[glucose] ratios. The
data have been vertically shifted by 10a to avoid overlapping. (b) summarization of
plateau storage modulus with various [PBA]/[glucose] ratios.
The mechanical properties of the polymer gels could also be manipulated by
altering the ratio of [PBA]/[glucose]. Figure 3a and Figure S10 show the G′ and G′′ as
a function of f for the gels obtained from PB-5 and PA-5 with different

[PBA]/[glucose] ratios at
a
constant 5.0 wt% solid content. The
frequency-independent plateau G′ are summarised in Figure 3b. As expected, the
largest plateau G′ occurred at the stoichiometric ratio (1: 1) of [PBA]/[glucose] due to
the increased cross linkages. The network parameters were calculated, and are
presented in Table S2. The hydrogel with a [PBA]/[glucose] ratio of 5:5 has the
highest cross-linking density υe and thus the lowest cross-link distance ξ between two
points.
3.3 Effect of solid content on hydrogels
The mechanical properties of the dynamic boronic ester-based hydrogels can be
further engineered by adjusting the solid content. Polymer gels formed from PB-5 and
PA-5 (pH = 7.0, [PBA]/[glucose] = 1) with different polymer concentrations were
chosen for testing. The G′ and G′′ as functions of strain obtained from oscillatory
amplitude sweep are shown in Figure S11. As expected, both storage and loss moduli
increased with increasing polymer content due to the increase in the cross-linking
density. The hydrogels with various concentrations demonstrated similar linear
viscoelastic regions up to 100% strain. In this region, G′ remained above G″,
indicating that the hydrogel networks were rigid and that the networks remained intact
under relatively large deformations. With increasing strain amplitude, gel-liquid
transition points, where the G″ curve crosses over the G′ curve occurred at 500%
strain, indicating the beginning of the destruction of the hydrogel network. However,
both G′ and G″ show little strain-hardening behaviours under oscillatory strain before
the destruction of the network structure, implying that both the creation and

destruction of dynamic bonds occurred during this process and that their creation
occurred more quickly than their destruction.[41,42]
Figure 4. (a) Dynamic oscillatory angular frequency sweep of the hydrogels formed
from PB-5 and PA-5 (pH = 7.0, [PBA]/[glucose] = 1) with different polymer
concentrations. (b) storage modulus as functions of the polymer concentrations.
Figure 4a depicts the storage and loss moduli versus angular frequency at a
constant strain (10%) for hydrogels with various polymer concentrations. The
crossover frequency of the G′ and G″ curves was independent of the concentration at

approximately 0.083 Hz in the concentration range of 30–150 g L-1. There have been
reports that both the G′ and G″ curves of dynamic hydrogels with different solid
contents can be superimposed by only vertical shifting.[36] This indicated that the
polymer concentration did not alter the lifetime of the cross-links arising from the
reversible nature of the DCBs. The plateau storage modulus versus polymer
concentration plots are presented in Figure 4b, showing a curve approximating a
straight line on a double logarithmic coordinate system. The correlation between the
mechanical strength and polymer concentration offers an effective approach to design
hydrogels with tailored mechanical properties.
The rheological properties of the hydrogel formed from PB-10 and PA-10 were
also investigated (Figure S12). The modulus of the polymer gels formed by
complexing the PB-10 and PA-10 copolymers were much higher than those of the
PB-5 and PA-5 gels. The plateau storage modulus of the hydrogel with a polymer
concentration of 15 wt% was 8.6 KPa, indicating that the mechanical strength could
be tuned over a large range.
3.4 Hydrogel self-healing behaviour
Hydrogel wound dressings with self-healing abilities can improve the reliability
of dressings and provide better wound protection.[43-45] To qualitatively evaluate the
self-healing behaviour of the dynamic polymer gels at intermediate pH, three parts of
a cylindrical hydrogel (formed from PB-5 and PA-5, 10 wt%, [PBA]/[glucose] = 1)
prepared under physiological conditions were placed together and contacted without
external treatment. As expected, the gels merged autonomously into a whole

cylindrical hydrogel after contact for 40 min. Furthermore, the cracks in the
attachment region disappeared completely (Figure 5a). The hydrogel after healing was
strong enough to be stretched to a length several times its initial length without
breakage. This self-healing behaviour can be explained by the dynamic equilibrium of
the formation and breakage of the phenylboronic ester bond.[46] In aqueous media,
free boronic acid and diol groups still exist in the hydrogels due to the reversibility of
the boronic ester bonds. Thus, the boronic ester formation and rearrangement across
the adjacent interfaces contributed to the self-healing of the hydrogels.
Figure 5. a) Optical images showing the self-healing behaviour of dynamic hydrogels,
the red hydrogels were dyed by carmine, and self-healing of hydrogels after fracture,
the plot b) shows the change in modulus during the strain ramp described by the
bottom plot c).
The self-healing properties of the dynamic polymer gels were also investigated by

dynamic rheological experiments. The hydrogel was subjected to increasing strain to
induce a large-amplitude deformation (1300% oscillatory strain) far beyond the linear
viscoelastic region, followed by a recovery period under small-amplitude (10%
oscillatory strain) deformation conditions. As shown in Figure 5b, the gel recovered
its original mechanical properties in 100 s at 10% strain after being destroyed,
indicating its excellent self-healing performance. Furthermore, the modulus value of
the hydrogel did not decrease significantly with each repeated cycle of recovery. This
repeatable and efficient self-healing performance of the hydrogel is beneficial for
bioengineering functions. To study the injectable properties, hydrogel was loaded into
a syringe, then injected on glass substrate. The self-healing hydrogel could pass
through a 26-gauge needle without clogging (Figure S13), indicating its
excellent injectability. This injectable self-healing hydrogel dressing could be easily
applied to irregular wound sites and adhered to wounds or medical gauze in situ. To
illustrate the adhesion properties, the hydrogel was adhered to human skin, and the
image was shown in Figure S14.
3.5 Cell cytotoxicity
Biocompatibility is an essential requirement for well-designed hydrogels for
wound dressing.[4,5] Therefore, a relative cell viability assay based on the standard
CCK-8 was performed for both HEK293 cells and HeLa cells, and the results are
shown in Figure 6. Considering the appropriate mechanical strength and good
self-healing ability,[5,7,47] the hydrogel formed from PB-5 and PA-5 (10 wt%,
[PBA]/[glucose] = 1) was chosen for evaluation. Mechanical tensile test showed that

the hydrogel possessed good stretchability (21 kPa, 210% of the elongation-at-break),
which could assure their integrity during the application (Figure S15). The hydrogel
exhibited low cytotoxicity, with over 95% cell viability remaining for both HEK293
and HeLa cells at a concentration of 400 mg gel in 4 mL of DMEM. When the
hydrogel content was increased to 800 mg, slightly lower yet still good viabilities of
88.7% and 87.5%, respectively, were obtained for HEK293 and HeLa cells. Typically,
boronic derivatives are considered to have low cytotoxicity for most cell lines.[48]
The good biocompatibility of the hydrogel can be attributed to the good
biocompatibility of the glycopolymer PA and the low content of PBA groups in PB-5.
Therefore, this boronic ester-based hydrogel formed under physiological conditions
has great potential as a wound dressing material.
Figure 6. Cell viability of HeLa cancer cells and HEK293 cells after 24 incubation
with gel extracts, the hydrogels were formed from PB-5 and PA-5 (10 wt%,
[PBA]/[glucose] = 1).
3.6 Decomposition of hydrogels

Wound dressing materials are usually mechanically removed from the wound,
which may cause secondary injury and additional pain to the patient. The dynamic
boronic acid ester bond can be dissociated in the presence of another competitive
molecule due to a shift in the equilibrium,[49,50] which can be effectively used to
prevent any new injuries arising from the removal of the material. Fructose was
selected for experiments owing to its high sensitivity to complexing with PBA.[16,51]
The hydrogel could completely degrade into a viscous liquid in 3 min after the
addition of excess fructose (Figure 7a), indicating the efficient dissociation of
phenylboronic-glucosamine interactions. The dissociation of carmine-dyed hydrogels
in 2 M fructose is depicted in Figure 7b. The hydrogel incubated in the fructose
solution completely disappeared within 30 min because of the cleavage of the
cross-links due to competition from fructose for binding with the PBA groups in the
system. Moreover, there was no obvious volume change of the hydrogel in the PBS
solution without fructose, indicating that the hydrogel was stable under physiological
conditions. Figure 7c shows the dissolution process on medical gauze. One half of the
carmine-dyed hydrogel coated on fructose-soaked gauze disappeared completely
within 5 min; that is, only half of the hydrogel remained. These results confirmed that
the demonstrated dynamic hydrogel with its controllable degradability can effectively
improve the performance of wound dressings, as it can be easily removed to avoid
damage to newly formed tissues.

Figure 7. (a) Degradation of the hydrogel by adding excess of the fructose, (b)
Hydrogel stability in PBS with (right vial) and without (left vial) 2.0 M of fructose, (c)
photographs of gel degradation process. (c, 1) original hydrogel, (c, 2)
fructose-soaked gauze was applied to half of the hydrogel, (c, 3) after 5.0 min and (c,
4) gauze was removed and only half of the hydrogel remained. Hydrogels were dyed
by carmine for better observation and the gauze was immersed in a 10 M fructose
solution for 10 min before used.
4. CONCLUSIONS
In summary, we fabricated bulk boronic ester-based hydrogels with excellent
physiological usability using the B-N coordination interaction for wound dressings.
The hydrogels were readily prepared by mixing the copolymer poly(AM-co-AAG)
and poly(AM-co-APBA-co-DMAPAA) in aqueous solution. Amine-containing
DMAPAA was incorporated into the polymer backbone to facilitate the formation and
increase the stability of the cross-links. The cross-linking density and inner structure
of the polymer hydrogels could be readily engineered by controlling the pH,

[PBA]/[glucose] ratio, solid content of the polymer, thus offering great versatility for
preparing functional hydrogels with tuneable mechanical properties. Moreover, the
hydrogels exhibited excellent biocompatibility for both HEK293 cells and HeLa cells.
Benefitting from the dynamic nature of the boronic ester, the hydrogels can be
dissociated by adding competitive sugar molecules, making them easy to remove from
medical gauze or skin. This approach of designing a polymer architecture to regulate
the pH dependence and mechanical properties of a gel is a promising strategy for
preparing functional hydrogels. The amine-containing dynamic hydrogel is a practical
platform for designing multifunctional, smart soft materials, and has great potential
for biomedical applications, especially for wound dressings.
Acknowledgements
This work was supported by the Major Research of Science and Technology, China
[grant number 2016ZX05058-003-012].
Declaration of interest: There are no conflicts to declare.
REFERENCES
[1] H. Chen, J. Cheng, L. Ran, K. Yu, B. Lu, G. Lan, F. Dai, F. Lu, An injectable self-healing hydrogel
with adhesive and antibacterial properties effectively promotes wound healing, Carbohydrate
Polymers, 201 (2018) 522-531.
[2] J.C. Dumville, N. Stubbs, S.J. Keogh, R.M. Walker, Z. Liu, Hydrogel dressings for treating
pressure ulcers, Cochrane Database of Systematic Reviews, 2 (2015) 1-76.
[3] J. Qu, X. Zhao, Y. Liang, T. Zhang, P.X. Ma, B. Guo, Antibacterial adhesive injectable hydrogels
with rapid self-healing, extensibility and compressibility as wound dressing for joints skin wound
healing, Biomaterials, 183 (2018) 185-199.
[4] X. Zhao, H. Wu, B. Guo, R. Dong, Y. Qiu, P. Ma, Antibacterial anti-oxidant electroactive injectable
hydrogel as self-healing wound dressing with hemostasis and adhesiveness for cutaneous wound
healing, Biomaterials, 122 (2017) 34- 47.
[5] J.S. Boateng, K.H. Matthews, H.N.E. Stevens, G.M. Eccleston, Wound Healing Dressings and Drug
Delivery Systems: A Review, Journal of Pharmaceutical Sciences, 97 (2008) 2892-2923.
[6] F.J. Vermolen, E.C.M.M. Arkesteijn, A. Gefen, Modelling the immune system response to