Influence of material properties on the damage-reporting and self-healing performance of a mechanically active dynamic network polymer in coating applications

Article abstract of DOI:10.3390/molecules26092468

We conducted a detailed investigation of the influence of the material properties of dynamic polymer network coatings on their self-healing and damage-reporting performance. A series of reversible polyacrylate urethane networks containing the damage-reporting diarylbibenzofuranone unit were synthesized, and their material properties (e.g., indentation modulus, hardness modulus, and glass-transition temperature) were measured conducting nanoindentation and differential scanning calorimetry experiments. The damage-reporting and self-healing performances of the dynamic polymer network coatings exhibited opposite tendencies with respect to the material properties of the polymer network coatings. Soft polymer network coatings with low glass-transition temperature (~10^?C) and indentation hardness (20 MPa) exhibited better self-healing performance (almost 100%) but two times worse damage-reporting properties than hard polymer network coatings with high glass-transition temperature (35~50^?C) and indentation hardness (150~200 MPa). These features of the dynamic polymer network coatings are unique; they are not observed in elastomers, films, and hydrogels, whereby the polymer networks are bound to the substrate surface. Evidence indicates that controlling the polymer’s physical properties is a key factor in designing high-performance self-healing and damage-reporting polymer coatings based on mechanophores.

Full text of DOI:10.3390/molecules26092468

molecules
Article
Influence of Material Properties on the Damage-Reporting and
Self-Healing Performance of a Mechanically Active Dynamic
Network Polymer in Coating Applications
Da Hae Son 1,†, Gi Young Kim 1,†, Ji-Eun Jeong 1, Sang-Ho Lee 1 , Young Il Park 1 , Hoyoul Kong 1,
In Woo Cheong 2,* and Jin Chul Kim 1,
*
1
Center for Advanced Specialty Chemicals, Division of Specialty and Bio-based Chemicals Technology,
Korea Research Institute of Chemical Technology (KRICT), Ulsan 44412, Korea; sdf1114@krict.re.kr (D.H.S.);
giyoung@krict.re.kr (G.Y.K.); jieunj@krict.re.kr (J.-E.J.); slee@krict.re.kr (S.-H.L.); ypark@krict.re.kr (Y.I.P.);
hkong2@krict.re.kr (H.K.)
School of Applied Chemistry, Kyungpook National University, Daegu 41566, Korea
Correspondence: inwoo@knu.ac.kr (I.W.C.); jckim81@krict.re.kr (J.C.K.)
These authors contributed equally to this work.
2
*
†
Abstract: We conducted a detailed investigation of the influence of the material properties of dynamic
polymer network coatings on their self-healing and damage-reporting performance. A series of
reversible polyacrylate urethane networks containing the damage-reporting diarylbibenzofuranone
unit were synthesized, and their material properties (e.g., indentation modulus, hardness modulus,
and glass-transition temperature) were measured conducting nanoindentation and differential scan-
ning calorimetry experiments. The damage-reporting and self-healing performances of the dynamic
polymer network coatings exhibited opposite tendencies with respect to the material properties of
the polymer network coatings. Soft polymer network coatings with low glass-transition temperature
ꢀꢁꢂꢀꢃꢄꢅꢆꢇ
ꢀꢁꢂꢃꢄꢅꢆ
Citation: Son, D.H.; Kim, G.Y.; Jeong,
J.-E.; Lee, S.-H.; Park, Y.I.; Kong, H.;
Cheong, I.W.; Kim, J.C. Influence of
Material Properties on the
◦
(~10 C) and indentation hardness (20 MPa) exhibited better self-healing performance (almost 100%)
Damage-Reporting and Self-Healing
Performance of a Mechanically Active
Dynamic Network Polymer in
Coating Applications. Molecules 2021,
26, 2468. https://doi.org/10.3390/
molecules26092468
but two times worse damage-reporting properties than hard polymer network coatings with high
◦
glass-transition temperature (35~50 C) and indentation hardness (150~200 MPa). These features of
the dynamic polymer network coatings are unique; they are not observed in elastomers, films, and
hydrogels, whereby the polymer networks are bound to the substrate surface. Evidence indicates
that controlling the polymer’s physical properties is a key factor in designing high-performance
self-healing and damage-reporting polymer coatings based on mechanophores.
Academic Editors: Ricardo Calhelha
and Hom Nath Dhakal
Keywords: polymer coating; dynamic polymer network; mechanophore; self-healing coating;
self-reporting polymer
Received: 18 February 2021
Accepted: 20 April 2021
Published: 23 April 2021
1. Introduction
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Mechanophores are molecules that can undergo chemical changes in response to
an externally applied mechanical force [
chromophores, meaning that they can both undergo chemical changes and alter their
light-emitting characteristics in response to a mechanical stimulus [ ]. Chemical motifs
1–6]. Notably, some mechanophores are also
7,8
that are able to behave as both mechanophores and chromophores have been embedded
into polymers to fabricate systems that could function as damage-reporting materials,
such as stress sensors and visual stress–strain detectors. Recently, extensive efforts have
been devoted to using these light-emitting polymers as self-healing materials that are
Copyright:
© 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
able to signal their damaged or healed status [9–11]. Thus far, however, most of these
investigations have focused exclusively on films and elastomers, although the combination
of these technologies (self-healing and damage-reporting technologies) is also important in
the development of novel self-healing polymer coatings.
Self-healing coatings can be classified into two categories: extrinsic and intrinsic.
Extrinsic self-healing coatings are obtained by including in the said coatings microcapsules
Molecules 2021, 26, 2468. https://doi.org/10.3390/molecules26092468
https://www.mdpi.com/journal/molecules

Molecules 2021, 26, 2468
2 of 12
with the ability to heal the damaged area [12]. In the case of these coatings, the damage-
reporting function can be easily achieved by embedding damage indicators, such as
pH-sensitive dyes and fluorescent molecules, into the microcapsules comprising healing
agents [13–15]. In this system, the material properties and forms of matrix polymers rarely
influence self-healing and damage-reporting performances, since self-healing and damage-
reporting are elicited by the diffusion of the solutions of the healing agents and/or the
damage indicators. However, the extrinsic self-healing mechanism limits repeated use of
self-healing and damage-reporting functions because the said mechanism depends on the
collapse of the microcapsules.
In the case of intrinsic self-healing coatings based on reversible processes, the self-
healing process differs completely from those occurring with other forms of polymer
network materials, such as elastomers, adhesives, and gels, because the polymer networks
are bound to the substrate surface (Figure 1). Since the diffusion of polymer chains in the
coatings is much more restricted than the diffusion of other polymer forms, elastic coatings
with a low elastic modulus (G0) value are preferred over hard coatings to achieve high
self-healing efficiency [16–20]. However, the effect of the material properties on the damage-
reporting performance of a mechanically active dynamic network self-healing polymer
coating has not yet been systematically reported. Recently, Geitner et al. reported a detailed
investigation conducted on an intrinsic self-healing polymer coating using molecular
coherent anti-Stokes Raman scattering analysis; notably, this investigation focused on
the morphological and chemical aspects, and it did not deal with the effect that material
properties have on damage-reporting and self-healing coatings [21].
Figure 1. Differences in the self-healing mechanisms of intrinsic self-healing elastomers, films, and coatings.
In the present paper, we report a detailed investigation of the influence that the ma-
terial properties of dynamic polymer network coatings have on their self-healing and
damage-reporting performance. As a mechanophore, diarylbibenzofuranone (DABBF),
first introduced by Hideyuki Otsuka’s group, was employed because this compound’s
disintegration (triggered by mechanical activation) and recombination (proceeding via aryl-
benzofuranone radical coupling) can be visualized by a color change [22–24]. In addition,
since the radical species generated as a result of the cleavage of a DABBF unit is known to
tolerate the presence of oxygen and to be stable for prolonged periods of time under ambi-
ent conditions, the self-healing and damage-reporting performance of the dynamic polymer
network can be reliably analyzed. A series of DABBF-containing poly(methyl methacrylate)-
co-[poly(hydroxyethyl methacrylate)-graft-poly(oligo-caprolactone)] urethane networks
exhibiting different material properties and different oligo-caprolactone bristle lengths of
the polymer networks were prepared. The material properties (e.g., thermal transition
temperature, thermal stability, and indentation hardness) were characterized by differen-
tial scanning calorimetry (DSC), thermogravimetric analysis (TGA), and nanoindentation
testing. The surface damage-reporting and self-healing performance of the polymer net-
works were characterized using a micro-scratch tester equipped with an optical microscope.

Molecules 2021, 26, 2468
3 of 12
Finally, with these measurement data in hand, the relationship between the material prop-
erties and the damage-reporting and self-healing performance of the dynamic network
polymers was systematically investigated.
2. Experiment
2.1. Materials
2,4-Di-tert-butylphenol (99%), acetic acid (
tin(II) 2-ethylhexanoate ( 92.5%), -caprolactone (97%), toluene (
drous, 99.8%), 2-butanone (99.7%), 2-hydroxyethyl methacrylate (HEMA, 97%), methyl
≥
99%), methanesulfonic acid (
≥
99%),
≥
ε
≥
99.5%), benzene (anhy-
≥
methacrylate (MMA, 99%), and dibutyltin dilaurate (DBTDL, 95%) were purchased from
Merck (Darmstadt, Germany). 3-chloro-1-propanol (>98%), di-tert-butyl peroxide (>98%),
and 4-hydroxymandelic acid monohydrate were obtained from TCI (Tokyo, Japan). Azobi-
sisobutyronitrile (AIBN, 98%) and sodium hydroxide (97%) were purchased from Junsei
(Tokyo, Japan). For ring-opening polymerization, ε-caprolactone and toluene were distilled
before use. MMA was passed through a basic alumina column to remove inhibitors before
conducting the polymerization reaction. AIBN was recrystallized twice from methanol
prior to use.
2.2. Synthesis
Synthesis of DABBF [22]. In a 100 mL two-neck round-bottom flask equipped with
a reflux condenser, 2,4-di-tert-butylphenol (15 g, 73 mmol) and 4-hydroxymandelic acid
monohydrate (10 g, 54 mmol) were dissolved in acetic acid (29 mL), and the resultant
solution was stirred at 95 ◦C. Methanesulfonic acid (0.1 mL, 0.8 mmol) was then added
to the acetic acid solution, and the reaction mixture thus obtained was stirred at 95 ◦C for
3 h. The mixture was then allowed to cool to room temperature, and it was made to stand
overnight; the product that precipitated as a result was filtered and washed several times
with water and hexane. After recrystallization from chloroform and hexane, precursor
1
compound
1
was obtained as a white solid (yield = 56%). H-NMR (300 MHz, CDCl3) peak
δ
values (ppm): 7.31 (s, 1H), 7.09 (d, 2H), 7.04 (s, 1H), 6.82 (d, 2H), 4.94 (s, 1H), 4.77 (s, 1H),
1.43 (s, 9H), 1.29 (s, 9H).
Compound (8 g, 24 mmol) was added to a 250 mL three-neck round-bottom flask
1
equipped with a dropping funnel and reflux condenser; an aqueous sodium hydroxide
solution (3 g NaOH in 60 mL of water) was then added to the flask. After the reaction
mixture thus obtained was stirred at 80 ◦C under an Ar atmosphere, 3-chloro-1-propanol
(4.2 mL, 50 mmol) was slowly droppe◦d into the flask using the dropping funnel. The
reaction mixture was maintained at 80 C for 3 h. After the mixture was allowed to cool to
room temperature, a solution of concentrated hydrochloric acid (7.6 mL in 60 mL water) was
added to it, and the resulting mixture was kept at 80 ◦C for another 3 h. After the reaction
mixture was allowed to cool to room temperature, the resultant precipitate was dissolved
in ethyl ether, washed several times with brine, and then concentrated under reduced
pressure. The crude product thus isolated was purified by column chromatography (using
ethyl acetate:hexane = 3:7, v:v, as eluent). Precursor compound
2
was thus obtained as a
1
white solid (yield = 64%). H-NMR (300 MHz, CDCl3) peak
δ values (ppm): 7.31 (s, 1H),
7.16 (d, 2H), 7.04 (s, 1H), 6.92 (d, 2H), 4.78 (s, 1H), 4.12 (t, 2H), 3.86 (t, 2H), 2.05 (m, 2H),
1.43 (s, 9H), 1.29 (s, 9H).
Precursor
2 (5 g, 13 mmol), di-tert-butyl peroxide (19 g, 130 mmol), and benzene
(
25 mL) were added to a 100 mL round-bottom flask. The solution thus obtained was irra-
diated with UV light (365 nm) for 2 h at 30 ◦C using a photoreactor (LZC 4V, LUZCHEM,
Canada). After completion of the reaction, the crude product was purified by flash col-
umn chromatography (using ethyl acetate:hexane = 3:1, v:v, as eluent) and recrystallized
from a mixed-solvent system of chloroform and hexane to give DABBF as a white solid
1
(
yield = 40%). H-NMR (300 MHz, CDCl3) peak
δ
values (ppm): 7.29 (br, 8H), 6.82 (d, 4H),
4.14 (m, 4H), 3.88 (m, 4H), 2.06 (m, 4H), 1.31 (br, 18H), 1.17 (br, 18H).

Molecules 2021, 26, 2468
4 of 12
Synthesis of oligo(
ε
-caprolactone) methacrylate monomers (MCLn). In a glovebox,
ε
-caprolactone, hydroxyethyl methacrylate, and tin(II) 2-ethylhexanoate were dissolved in
toluene, and the solution thus obtained was stirred at 120 ◦C. After 2 h, excess methanol
was added to the mixture to quench the reaction. The reaction mixture was precipitated in
cold hexane, washed three times with cold hexane, and dried under vacuum at 30 ◦C for
12 h.
MCL1. Yield: 66%; 1H-NMR (300 MHz, CDCl3) peak
δ
values (ppm): 6.15 (s, 1H), 5.61
(s, 1H), 4.38 (s, 4H), 3.68 (t, 2H), 2.35 (m, 2H), 1.99 (s, 3H), 1.67 (m, 4H), 1.42 (m, 2H).
MCL5. Yield: 65%; 1H-NMR (300 MHz, CDCl3) peak
values (ppm): 6.15 (s, 1H), 5.61
δ
(s, 1H), 4.38 (s, 4H), 4.09 (t, 8H), 3.68 (t, 2H), 2.35 (m, 10H), 1.99 (s, 3H), 1.67 (m, 20H), 1.42
(m, 10H).
MCL10. Yield: 72%; 1H-NMR (300 MHz, CDCl3) peak
δ values (ppm): 6.15 (s, 1H),
5.61 (s, 1H), 4.38 (s, 4H), 4.09 (t, 18H), 3.68 (t, 2H), 2.35 (m, 20H), 1.99 (s, 3H), 1.67 (br, 40H),
1.42 (br, 20H).
Polymerization of hydroxyl-terminated graft copolymers (GCLs). GCL1, GCL5, and
GCL10 were synthesized starting from the corresponding MCLs (20 mol%) and methyl
methacrylate (80 mol%) in 2-butanone via a conventional free-radical polymerization
carried out in the presence of AIBN (2 mol%) at 75 ◦C for 3 h under Ar atmosphere. After
being precipitated from diethyl ether, the polymer was washed and then dried under
vacuum for 24 h.
GCL1. Yield: 70%; 1H-NMR (300 MHz, CDCl3) peak
3.61 (br, 14H), 2.37 (br, 2H), 1.68 (br, 4H), 1.42 (br, 2H).
GCL5. Yield: 58%; 1H-NMR (300 MHz, CDCl3) peak
δ
values (ppm): 4.30 (br, 4H),
δ
values (ppm): 4.30 (br, 4H),
4.19 (br, 8H), 3.61 (br, 14H), 2.37 (br, 10H), 1.68 (br, 20H), 1.42 (br, 10H).
GCL10. Yield: 45%; 1H-NMR (300 MHz, CDCl3) peak
δ
values (ppm): 4.30 (br, 4H),
4.19 (br, 18H), 3.61 (br, 14H), 2.37 (br, 20H), 1.68 (br, 40H), 1.42 (br, 20H).
2.3. Preparation of the Polymer Network Coatings (C-GCLs)
The GCL polymers, DABBF, Desmodur N3300, and a catalytic amount of DBTDL
were dissolved in chloroform. The solution thus obtained was coated onto a glass slide
implementing the conventional drawdown bar coating method. The coated glass slide
was then cured in a convection oven at 70 ◦C for 12 h. The thickness of the coatings was
adjusted to 40 µm. The weight contents of the DABBF units in the C-GCL coatings were
adjusted to 22 wt%.
2.4. Chemical Structure Confirmation
1H-NMR spectra of the organic compounds and polymers were recorded using a
300 MHz NMR spectrometer (Bruker, Ultrashield, Fällanden, Switzerland) under ambient
conditions. The singlet resonance peak due to residual CHCl3 in CDCl3 appearing at 7.26
ppm was selected as the reference standard.
2.5. Molecular-Weight Determination
The number-average molecular weight (Mn) and dispersity (ÐM, = Mw/Mn, where
Mw is the weight-average molecular weight) of the synthesized polymers were deter-
mined by size-exclusion chromatography using an apparatus (Agilent Tech 1260, Agilent
Technologies, Santa Clara, CA, USA) equipped with a set of gel columns (Agilent PLgel
5 µm Mixed-D column, Agilent Technologies, Santa Clara, CA, USA). The system was
equilibrated at 40 ◦C in anhydrous tetrahydrofuran (THF), which was used as both the
polymer solvent and the eluent (flow rate of 1 mL min−1), and calibrated using polystyrene
standards (650 ≤ Mw ≤ 6,375,000).
2.6. Thermal Property Determination
The thermal stability of the polymers was investigated by TGA (TA Instruments TGA
Q500, New Castle, DE, USA) conducted under an N2 atmosphere between 25 and 600 ◦C

Molecules 2021, 26, 2468
5 of 12
◦
applying a heating rate of 10 C min−1. The thermal transitions of the polymers were
determined from the second DSC heating ramp under a N2 atmosphere between
−
50 and
100 ◦C at a heating rate of 10 ◦C min−1 using a TA Instruments DSC Q2000.
2.7. Nanoindentation Test
Loading, holding, and unloading indentation measurements were performed using
a nanoindentation tester (Anton Paar, NHT3, Graz, Austria) equipped with a Berkovich-
type indenter. During the loading step, the force imparted by the indenter was gradually
increased from 0 to 20 mN at a rate of 20 mN min−1. During the holding step, the applied
force was maintained at a value of 20 mN for 30 s. Finally, the indenter was unloaded at
the same absolute value of the rate used for the loading. Through these three steps, the
values for the maximum displacement (hmax), permanent depth of penetration (final depth,
hf), elastic unloading stiffness (S = dP/dh), indentation hardness (HIT), and indentation
modulus (EIT) were obtained [17].
2.8. Micro-Scratch Test
Micro-scratch tests were performed using a spheroconical indenter (tip radius: 5 µm)
mounted on a scratch test machine (Anton Paar, nano-scratch tester). The force exerted on
the coating was progressively increased from 2 to 100 mN to initiate fracture formation
on the coating surface. The indenter speed and the scratch length were 2 mm min−1 and
2.0 mm, respectively. The mechanisms for the fracture of the polymer coatings and for
the dissociation of the DABBF units present in the polymer networks (dissociation of the
polymer networks) were analyzed using an optical microscope attached to the scratch test
machine [17].
2.9. Scratch-Healing Efficiency Determination
Scratch healing was induced by heating the material at 60 ◦C for 6 h. The healing of
the scratches and the recombination of the DABBF at the damaged surfaces were observed
using an optical microscope attached to the scratch test machine.
2.10. Measurement of DABBF Dissociation and Recombination Efficiencies
The dissociation efficiency of DABBF caused by the generation of the surface scratch
and the recombination efficiency of DABBF influenced by the self-healing process of the
C-GCL coatings were quantitatively determined in terms of the changes in the value of Db*.
In the CIELAB color space (L*a*b*), the b* axis represents the blue-yellow components,
with number toward blue and positive toward yellow. The Db* transition measurement
of the scratched and healed surface of the C-GCL coatings was conducted by analyzing
the color coordination of pixels in the optical microscopy images. The CIE 1931 XYZ color
space values of the pixels in the image were converted to the L*a*b* values using the Adobe
Photoshop®program (version 12.0 x64). Notably, the Db* value at a certain point was the
average value of 100 pixels.
3. Results and discussion
3.1. Material Design and Preparation
The detailed procedures for the syntheses of DABBF, MCLs, and GCLs are reported in
Scheme 1.

Molecules 2021, 26, 2468
6 of 12
Scheme 1. Procedures implemented to synthesize (
a) diarylbibenzofuranone and (
b
) oligo(
ε-
caprolactone) methacrylate monomers and hydroxyl-terminated graft copolymers; (
structure of polyisocyanate (Desmodur N3300).
c) chemical
DABBF was prepared from 4-hydroxymandelic acid and 2,4-di-tert-butylphenol in
three steps following a previously reported protocol. MCL1, MCL5, and MCL10 were
synthesized via ring-opening polymerization of the ε-caprolactone monomer using the
hydroxyethyl methacrylate monomer as initiator, in the presence of tin(II) 2-ethylhexanoate
acting as a catalyst. GCL1, GCL5, and GCL10 were synthesized from MMA and the
corresponding MCL monomers in the presence of AIBN acting as the initiator by way of
free-radical polymerization. The chemical structures of the synthesized materials were
confirmed by 1H-NMR spectroscopy (Figure 2). Values for the average molecular weights,
degrees of polymerization, and monomer ratios of MCLs and GCL polymers are listed in
Tables 1 and 2.
1
Figure 2. Chemical structure confirmation of the synthesized materials: H-NMR spectra of (
a
) di-
) hydroxyl-
) Fourier-transform infrared spectra of DABBF, hydroxyl-
arylbibenzofuranone (DABBF), (
b) oligo(
ε
-caprolactone) methacrylate 10 monomer, and (
c
terminated graft 10 polymer (GCL10). (
d
terminated graft polymer (GCL5, polyisocyanate (Desmodur N3300)), and C-GCL10.

Molecules 2021, 26, 2468
7 of 12
Table 1. Molecular weights and molecular-weight distributions of the various oligo(
ε
-caprolactone)
methacrylate monomers (MCLs).
a
b
c
Monomer Code
Mn (g/mol) c Mw (g/mol) c
ÐM
DPTV
DPCV
MCL1
MCL5
MCL10
1
5
10
1.8
5.5
10.3
340
1080
1500
450
1310
1710
1.3
1.2
1.1
b
a
DPTV and DPCV indicate the theoretical and calculated degrees of polymerization, respectively. Determined
by 1H-NMR spectroscopy. Determined by size-exclusion chromatography.
c
Table 2. Chemical compositions, molecular weights, and molecular-weight distributions of the
hydroxyl-terminated graft copolymers (GCLs).
MCL a Contents (mol%) b
c
Polymer Code
ÐM
Mn (kg/mol) Mw (kg/mol)
Theoretical
Calculated
GCL1
GCL5
GCL10
20
20
20
20
21
21
b
17.0
25.0
34.0
25.6
32.5
44.2
1.5
1.3
1.3
a
c
MCL: oligo(
ε
-caprolactone) methacrylate monomer. Determined by 1H-NMR spectroscopy. Determined by
size-exclusion chromatography.
The poly(urethane acrylate) network coatings containing the self-healing unit were
prepared from DABBF, the corresponding GCL polymers, and Desmodur N3300 imple-
menting the conventional drawdown bar coating method. The thermal stabilities of the
C-GCL coatings were investigated by TGA (Figure 3a). All C-GCLs were stable up to
180 ◦C. The values for the glass-transition temperature (Tg) of the C-GCLs were measured
by DSC (Figure 3b). The Tg values of the C-GCLs decreased as the bristle length of the
repeating unit increased; notably, increases in the mentioned parameter are associated with
decreases in the crosslinking density of the polymer networks.
Figure 3. Thermal properties of the polymer coatings: (a) thermogravimetric analysis pyrograms
and (b) scanning calorimetry thermograms of the C-GCL1, C-GCL5, and C-GCL10.
3.2. Results of Nanoindentation Tests Conducted on the C-GCL Coatings
In Figure 4 are reported the typical load–displacement curves obtained following
implementation of nanoindentation tests on the C-GCL coatings. All C-GCLs exhibited
plastic deformation in response to applied normal forces. Moreover, as the bristle length
of the polymer increased, the hmax and hf values increased, but the dP/dh values in the
unloading step decreased, which we attributed to the decrease in crosslinking density. The
calculated HIT and EIT values of the C-GCL coatings also decreased as the bristle length of
the polymer was increased. Especially, compared to the C-GCL1 and GCL5 coatings, the
corresponding C-GCL10 coating exhibited much lower HIT and EIT values due to its low
crosslinking density and glass-transition temperature. Both C-GCL1 and C-GCL5 coatings
with high HIT and EIT values tend to resist deformation against the applied stress. In this
case, when scratched, the applied stress will easily be accumulated on the surface and then

Molecules 2021, 26, 2468
8 of 12
dissipated by the thermal energy. On the other hand, the surface of the C-GCL10 coating
will be deformed to dissipate the applied stress.
Figure 4. Nanoindentation tests conducted on the C-GCL coatings: (
) load–displacement curves, and ( ) values for the indentation hardness (HIT) and indentation
modulus (EIT) of the C-GCL coatings.
a) nanoindentation tester setup,
(b
c
3.3. Results of the Scratch and Damage-Reporting Tests Conducted on the C-GCL Coatings
The influence of the scratch load on the breakage response of the C-GCL coatings was
determined by implementing the scratch test method, whereby a deforming load with a
force that increased progressively from 2 to 100 mN was applied (Figure 5).
Figure 5. Optical microscopy images of the scratched and healed surfaces of the C-GCL coatings.
All C-GCL coatings exhibited plastic deformation at the initial stage of the test (against
a small load). As the deforming load increased, however, the mechanophore unit (DABBF)
in the C-GCL coatings began to dissociate, dissipating the loading energy.

Molecules 2021, 26, 2468
9 of 12
The color transitions of the scratched C-GCLs, which reflect the degree of dissoci-
ation of the DABBF unit, are quantitatively presented in Figure 6 in terms of the vari-
ous C-GCLs’ Db* values. As the applied force increased, the Db* value drastically de-
creased, as indicated by the appearance of the scratched position of the C-GCL surfaces,
which changed from transparent (colorless) to blue. The Db* values of hard C-GCL coat-
ings decreased more rapidly at smaller deforming loads than the soft C-GCL coatings
(C-GCL1 > C-GCL5 > C-GCL10).
Figure 6. The color transitions (expressed in Db* values) of the damaged and healed surfaces of the
C-GCL coatings as a function of the applied loads: (a) C-GCL1, (b) C-GCL5, and (c) C-GCL10.
In general, hard coatings with high crosslinking density tend to exhibit greater scratch
resistance than soft coatings. However, in the present work, the hard C-GCL1 and C-GCL5,
which are characterized by high EIT and Tg values (Tg >> 25 ◦C), proved to be more vul-
nerable to surface rupture than the soft C-GCL10 in the rubbery state (Tg < 25 ◦C). This
interesting observation is attributable to a massive dissociation of the DABBF units, which
resulted in the loss of numerous crosslinking points in the hard coating. Indeed, after a
massive loss of crosslinking points, the coating is no longer a tough material characterized
by high crosslinking density. In addition, C-GCL1 exhibited greater plastic deformation
than C-GCL5 and C-GCL10 under the experimental conditions (room temperature). In
contrast to this result, evidence from our previous investigations indicated that a ther-
mally reversible dynamic poly(urethane arylate) network with similar chemical structure
and material properties to the C-GCL coatings had much higher surface toughness than
C-GCL coatings.
Previously, Ostuka et al. investigated the activation and recombination process of
DABBF units in the hard and soft domains of a polymer film in response to the application
of stretching and grinding stress modes. In their experiments, they observed that the
DABBF units in the hard polystyrene domains were not cleaved as a result of the stretching
process, but they were activated when the film was ground. In our system, the failure
of the DABBF units to elicit a response to the external stimulus in the C-GCL coatings is
similar to the result of the grinding stress mode, since two-dimensional stresses (vertical
and horizontal stresses) are applied to C-GCL polymers bound to their substrates [25].

Molecules 2021, 26, 2468
10 of 12
3.4. Scratch-Healing Performance of the C-GCL Coatings
The scratch-healing performance of the C-GCL coatings was characterized by heating
each material at 60 ◦C for 6 h.
The width-based scratch-healing efficiency (%WSHE) was calculated employing
Equation (1):
ꢀ
ꢁ
Rw,s − Rw,h
%WSHE =
× 100
(1)
Rw,s
where Rw,s and Rw are the residual width after scratch formation and the residual width
,h
after the healing process, respectively.
The calculated %WSHE values for the C-GCL coatings at a thickness of 2000 µm are
presented in Figure 7a. As expected, the self-healing performance of the C-GCL coatings
increased as the crosslinking density increased (C-GCL10 >> C-GCL5 > C-GCL1) because
of the enhanced polymer-chain mobility. The ruptured surface area of the C-GCL10 coating
was also efficiently diminished by the scratch-zipping process (polymer-chain reflow). By
contrast, the zipping process of the ruptured C-GCL5 was much more limited than that
of C-GCL10, producing the remaining ruptured surface. In the case of C-GCL1, even the
remaining scratch formed by plastic deformation as well as the ruptured region was not
healed due to the low polymer-chain mobility that characterizes this coating.
Figure 7. Self-healing and damage-reporting performance of the C-GCL coatings: (a) scratch widths
and width-based scratch-healing efficiencies of the C-GCL coatings (loading force = 100 mN); (
b)
damage-reporting and diarylbibenzofuranone recombination efficiencies of the C-GCL coatings.
We investigated the reconstruction of destroyed crosslinking points in the C-GCL
coatings by observing the disappearance of the color due to the DABBF units. As can be
evinced from Figures 6 and 7b, the color of the damaged regions of C-GCL10 and C-GCL5
completely disappeared after the healing process, indicating that the recombination of
DABBF occurred in high yield. By contrast, the blue color of C-GCL1 did not disappear
completely after the scratch-healing process, and it remained in the central position of the
scratch path and the massively ruptured area. This result was attributed mainly to C-GCL1
being characterized by higher crosslinking density than C-GCL5 and C-GCL10. In the case
of C-GCL5, the blue color of the scratch path and the massively ruptured area after the
scratch-healing process disappeared, but the damaged surface did not fully recover its
original shape.
The contact of the DABBF units present in the polymer networks with high crosslink-
ing density was highly restricted by the limited polymer-chain movements. Notably, the
disappearance of the blue color of the damaged area indicates recombination of the DABBF
units. In the case of C-GCL5, under the described healing conditions, the polymer-chain
movement is sufficient for the recombination of nearby DABBF units in both deformed
and ruptured area to take place, but the said movement is not enough for the polymer
to flow over a wide range. Meanwhile, the movements of the polymer chains in C-GCL1
are more highly restricted than those of the polymer chains in C-GCL5 due to C-GCL10s
high crosslinking density. Under the described healing conditions, the DABBF units in the
ruptured area (not in the scratch path) of C-GCL1 cannot recombine.

Molecules 2021, 26, 2468
11 of 12
Notably, Otsuka et al. observed similar trends to those reported in the present study.
In their system, the recombination of the DABBF units in the rigid silica network was much
slower than that observed in the soft poly(butyl acrylate) domain [26].
Based on the results of the aforementioned experiments, we can conclude that, in
order to achieve a satisfactory damage-reporting performance, damage-reporting coatings
based on dynamic covalent bonds need to be characterized by high G0 and Tg values (Tg
>> room temperature). However, if the values of the said parameters are too high, the
polymer-chain mobility is restricted, resulting in a low recombination yield of the dynamic
covalent bond in the polymer. Indeed, high values for G0 and Tg limit the repeated use of
the damage-reporting capability of the polymer coatings. However, if the G0 value of the
coating is too low, the stress required to activate the cleavage of the dynamic covalent bond
increases, which means that the sensitivity of the damage-reporting system decreases.
In terms of self-healing, the polymer coatings with low G0 values exhibit high self-
healing performance. This trend is mainly attributed to an increase in polymer-chain
mobility. Unlike the conventional self-healing coating system based on dynamic cova-
lent bonding, the surfaces of hard polymers exhibited inferior toughness to those of soft
polymers because the crosslinking point of hard coatings is more fragile than that of
soft coatings. In addition, ruptures in the surfaces of hard coatings are permanent be-
cause recombination of the DABBF units (crosslinking point) only occurs with adjacent
DABBF units, which means that polymer-chain interdiffusion in the polymer coating is no
longer available.
4. Conclusions
In this paper is described a detailed investigation of the influence of the material prop-
erties of stress-sensitive dynamic polymer network coatings on their damage-reporting and
damage-healing performances. Our observations of the damage-reporting and self-healing
process of the dynamic polymer network coatings revealed that material properties such
as those identified by the parameters EIT, HIT, and Tg, as well as the toughness of the
said coatings, greatly influence the self-healing performance of the reversible polymer
network coatings containing the DABBF units. In terms of the damage-reporting perfor-
mance, hard polymer coatings exhibited superior performance to soft polymer coatings
because, in the case of hard polymer coatings, the applied stress can mainly be used to
break the DABBF units, thus minimizing other energy dissipation modes, such as elastic
and plastic deformations. By contrast, hard polymer coatings showed inferior self-healing
performance to their soft counterparts because the self-healing process of the intrinsic
self-healing polymers was strongly influenced by the polymer-chain interdiffusion effect.
These features of the dynamic polymer network coatings are unique; they are not observed
in systems like elastomers, films, and hydrogels, where the polymer networks are not
bound to the substrate surface. Therefore, we conclude that controlling the polymer’s
physical properties is a key factor in designing scratch-healing polymer network coatings
based on damage-reporting mechanophores. Notably, in the present system, C-GCL5
exhibited excellent damage-reporting and self-healing performance.
Author Contributions: Conceptualization, I.W.C. and J.C.K.; methodology, J.-E.J., S.-H.L., Y.I.P. and
H.K.; validation, J.-E.J.; investigation, D.H.S. and G.Y.K.; data curation, J.-E.J.; writing—original draft
preparation, J.C.K. and I.W.C.; writing—review and editing, J.C.K. and I.W.C.; supervision, J.C.K.;
project administration, J.C.K.; funding acquisition, J.C.K. All authors have read and agreed to the
published version of the manuscript.
Funding: This work was supported by a National Research Foundation of Korea (NRF) grant funded
by the Korean government (MSIT) (No. 2020R1C1C1005569 and No. 2020R1I1A3074491).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.

Molecules 2021, 26, 2468
12 of 12
Conflicts of Interest: The authors declare no conflict of interest.
Sample Availability: Samples of the compounds are not available.
References
1.
2.
3.
May, P.A.; Moore, J.S. Polymer Mechanochemistry: Techniques to Generate Molecular Force via Elongational Flows. Chem. Soc.
Rev. 2013, 42, 7497–7506. [CrossRef]
Caruso, M.M.; Davis, D.A.; Shen, Q.; Odom, S.A.; Sottos, N.R.; White, S.R.; Moore, J.S. Mechanically-Induced Chemical Changes
in Polymeric Materials. Chem. Rev. 2009, 109, 5755–5798. [CrossRef] [PubMed]
Bowser, B.H.; Craig, S.L. Empowering Mechanochemistry with Multi-Mechanophore Polymer Architectures. Polym. Chem. 2018
9, 3583–3593. [CrossRef]
,
4.
5.
Do, J.-L.; Frišcˇic´, T. Mechanochemistry: A Force of Synthesis. ACS Cent. Sci. 2017, 3, 13–19. [CrossRef]
Stolar, T.; Užarevic´, K. Mechanochemistry: An Efficient and Versatile Toolbox for Synthesis, Transformation, and Functionalization
of Porous Metal–Organic Frameworks. Cryst. Eng. Comm. 2020, 22, 4511–4525. [CrossRef]
Gomes, C.; Vinagreiro, C.S.; Damas, L.; Aquino, G.; Quaresma, J.; Chaves, C.; Pimenta, J.; Campos, J.; Pereira, M.; Pineiro, M.
Advanced Mechanochemistry Device for Sustainable Synthetic Processes. ACS Omega 2020, 5, 10868–10877. [CrossRef] [PubMed]
Willis-Fox, N.; Rognin, E.; Aljohani, T.A.; Daly, R. Polymer Mechanochemistry: Manufacturing Is Now a Force to Be Reckoned
With. Chem 2018, 4, 2499–2537. [CrossRef]
Akbulatov, S.; Boulatov, R. Experimental Polymer Mechanochemistry and Its Interpretational Frameworks. Chem. Phys. Chem.
2017, 18, 1422–1450. [CrossRef]
Li, J.; Nagamani, C.; Moore, J.S. Polymer Mechanochemistry: From Destructive to Productive. Acc. Chem. Res. 2015, 48, 2181–2190.
[CrossRef] [PubMed]
6.
7.
8.
9.
10. Diesendruck, C.E.; Moore, J.S. Mechanophores for Self-Healing Applications. In Self-Healing Polymers; John Wiley & Sons, Ltd.:
Hoboken, NJ, USA, 2013; pp. 193–214, ISBN 978-3-527-67018-5.
11. Hu, H.; Ma, Z.; Jia, X. Reaction Cascades in Polymer Mechanochemistry. Mater. Chem. Front. 2020, 4, 3115–3129. [CrossRef]
12. White, S.R.; Sottos, N.R.; Geubelle, P.H.; Moore, J.S.; Kessler, M.R.; Sriram, S.R.; Brown, E.N.; Viswanathan, S. Autonomic Healing
of Polymer Composites. Nature 2001, 409, 794–797. [CrossRef]
13. Robb, M.J.; Li, W.; Gergely, R.C.R.; Matthews, C.C.; White, S.R.; Sottos, N.R.; Moore, J.S. A Robust Damage-Reporting Strategy for
Polymeric Materials Enabled by Aggregation-Induced Emission. Acs Cent. Sci. 2016, 2, 598–603. [CrossRef]
14. Li, W.; Matthews, C.C.; Yang, K.; Odarczenko, M.T.; White, S.R.; Sottos, N.R. Autonomous Indication of Mechanical Damage in
Polymeric Coatings. Adv. Mater. 2016, 28, 2189–2194. [CrossRef] [PubMed]
15. Lee, T.H.; Song, Y.K.; Park, S.H.; Park, Y.I.; Noh, S.M.; Kim, J.C. Dual Stimuli Responsive Self-Reporting Material for Chemical
Reservoir Coating. Appl. Surf. Sci. 2018, 434, 1327–1335. [CrossRef]
16. Lee, H.M.; Perumal, S.; Kim, G.Y.; Kim, J.C.; Kim, Y.-R.; Kim, M.P.; Ko, H.; Rho, Y.; Cheong, I.W. Enhanced Thermomechanical
Property of a Self-Healing Polymer via Self-Assembly of a Reversibly Cross-Linkable Block Copolymer. Polym. Chem. 2020, 11,
3701–3708. [CrossRef]
17. Kim, S.Y.; Lee, T.H.; Park, Y.I.; Nam, J.H.; Noh, S.M.; Cheong, I.W.; Kim, J.C. Influence of Material Properties on Scratch-Healing
Performance of Polyacrylate-Graft-Polyurethane Network That Undergo Thermally Reversible Crosslinking. Polymer 2017, 128,
135–146. [CrossRef]
18. Kim, G.Y.; Sung, S.; Kim, M.P.; Kim, S.C.; Lee, S.-H.; Park, Y.I.; Noh, S.M.; Cheong, I.W.; Kim, J.C. Reversible Polymer Networks
Based on the Dynamic Hindered Urea Bond for Scratch Healing in Automotive Clearcoats. Appl. Surf. Sci. 2020, 505, 144546.
[CrossRef]
19. Park, J.I.; Choe, A.; Kim, M.P.; Ko, H.; Lee, T.H.; Noh, S.M.; Kim, J.C.; Cheong, I.W. Water-Adaptive and Repeatable Self-Healing
Polymers Bearing Bulky Urea Bonds. Polym. Chem. 2018, 9, 11–19. [CrossRef]
20. Sung, S.; Kim, S.Y.; Lee, T.H.; Favaro, G.; Park, Y.I.; Lee, S.-H.; Ahn, J.B.; Noh, S.M.; Kim, J.C. Thermally Reversible Polymer
Networks for Scratch Resistance and Scratch Healing in Automotive Clear Coats. Prog. Org. Coat. 2019, 127, 37–44. [CrossRef]
21. Geitner, R.; Legesse, F.-B.; Kuhl, N.; Bocklitz, T.W.; Zechel, S.; Vitz, J.; Hager, M.; Schubert, U.S.; Dietzek, B.; Schmitt, M.; et al. Do
You Get What You See? Understanding Molecular Self-Healing. Chem. Eur. J. 2018, 24, 2493–2502. [CrossRef]
22. Imato, K.; Kanehara, T.; Nojima, S.; Ohishi, T.; Higaki, Y.; Takahara, A.; Otsuka, J. Repeatable Mechanochemical Activation of
Dynamic Covalent Bonds in Thermoplastic Elastomers. Chem. Commun. 2016, 52, 10482–10485. [CrossRef]
23. Imato, K.; Natterodt, J.C.; Sapkota, J.; Goseki, R.; Weder, C.; Takahara, A.; Otsuka, H. Dynamic Covalent Diarylbibenzofuranone-
Modified Nanocellulose: Mechanochromic Behaviour and Application in Self-Healing Polymer Composites. Polym. Chem. 2017
8, 2115–2122. [CrossRef]
,
24. Kosuge, T.; Aoki, D.; Otsuka, H. Network reorganization in cross-linked polymer/silica composites based on exchangeable
dynamic covalent carbon–carbon bonds. Polymer 2019, 177, 10–18. [CrossRef]
25. Ishizuki, K.; Aoki, D.; Goseki, R.; Otsuka, H. Multicolor Mechanochromic Polymer Blends That Can Discriminate between
Stretching and Grinding. ACS Macro Lett. 2018, 7, 556–560. [CrossRef]
26. Kosuge, T.; Imato, K.; Goseki, R.; Otsuka, H. Polymer–Inorganic Composites with Dynamic Covalent Mechanochromophore.
Macromolecules 2016, 49, 5903–5911. [CrossRef]