Progress in Organic Coatings 127 (2019) 37–44

Contents lists available at ScienceDirect

Progress in Organic Coatings

journal homepage: www.elsevier.com/locate/porgcoat

Thermally reversible polymer networks for scratch resistance and scratch T

healing in automotive clear coats
Sujin Sunga,1, So Young Kima,1, Tae Hee Leea, Gregory Favarob, Young Il Parka, Sang-Ho Leea,
⁎ ⁎
Jae Beom Ahnc, Seung Man Noha, , Jin Chul Kima,
a
Research Center for Green Fine Chemicals, Korea Research Institute of Chemical Technology, Ulsan, 44412, Republic of Korea
b
TriTec SA, Anton Paar, CH-2034, Peseux, Switzerland
c
NOROO BEE Chemical R&D Center, Cheonan, 31094, Republic of Korea

A R T I C LE I N FO A B S T R A C T

Keywords: This study investigated the application of a thermally reversible polymer network fabricated using Diels-Alder
Self-healing polymer (DA)/retro-DA (rDA) reactions for use as a scratch-healing automotive clear coat. For this purpose, a new
Diels-Alder reaction scratch-healing poly(urethane acrylate) network containing a DA adduct unit (DA-CL) was prepared, and its
Thermally reversible crosslinking material properties and scratch-healing performance were compared to the properties of a commercial clear coat
Automotive clear coat
system (C-CL). The thermally reversible crosslinking and de-crosslinking reactions among the DA-CL, via DA and
In-situ oscillatory rheology
rDA reactions, were systematically evaluated using in-situ oscillatory rheology coupled with FT-IR spectroscopy.
The material properties of the DA-CL and C-CL materials, including the thermal stability, thermal transitions,
hardness, and scratch resistance, were measured using thermogravimetric analysis (TGA), differential scanning
calorimetry (DSC), nanoindentation, and nanoscratch test methods. The scratch-healing performance of the DA-
CL was quantitatively characterized and compared to the performance of the C-CL using a nanoscratch tester
equipped with an optical microscope (OM) and an atomic force microscope (AFM). The DA-CL polymer network
exhibited superior scratch healing and scratch resistance compared to the C-CL polymer network. These data
indicated that the DA self-healing polymer network is potentially useful as a scratch-healing clear coat for the
automotive industry.

1. Introduction scratch healing without an external stimulus. Microcapsules cannot,

however, be used in transparent coatings because the differences in the
Polymer coatings have been used to protect substrate surfaces from refractive indices of the microcapsules and coating material render the
the external environment and/or to enhance aesthetic appearance. coating opaque. Additionally, the microcapsules within a damaged area
Recently, however, polymer coats have evolved to perform various collapse during scratch healing and are not available to heal subsequent
‘smart’ functions such as self-healing, self-cleaning, and stimulus-re- scratches within the damaged area. In intrinsic scratch-healing systems,
sponse functions [1,2]. Self-healing coatings are among the most im- on the other hand, diverse interactions and mechanisms are present; i)
portant technologies for the use of high-value products that require physical interactions such as hydrogen bonds and charge transfer, ii)
long-term mechanical stability and chemical resistance, as required in metal-ligand coordination (zinc acetate, CuCl2, and dibutyltin dilaurate
automotive, architectural, medical, and electronic devices [3–6]. Au- (DBTDL)), and iii) dynamic crosslinking through Diels-Alder (DA) re-
tomotive coatings, in particular, require scratch-healing systems be- actions or other reactions that form hindered urea bonds, alkox-
cause they are exposed to harsh environments for prolonged periods of yamines, boronic esters, or di-thiolate bonds. The healing process is
time [7,8]. triggered by an external stimulus, such as thermal energy, ultraviolet
Scratch-healing techniques can use extrinsic or intrinsic healing light, a change in pH, or the introduction of moisture [12–16]. Intrinsic
mechanisms. Extrinsic scratch-healing systems rely on the presence of scratch-healing techniques are, therefore, superior because they are
microcapsules within the coating that are capable of healing scratches applicable to a wide range of organic coatings and can be used with
[9–11]. This approach is advantageous as it can provide autonomous transparent coatings that may require repeated scratch-healing.

⁎
Corresponding authors.
E-mail addresses: smnoh@krict.re.kr (S.M. Noh), jckim81@krict.re.kr (J.C. Kim).
1
S. Sung and S.Y. Kim contributed equally to this work.

https://doi.org/10.1016/j.porgcoat.2018.10.023
Received 10 March 2018; Received in revised form 11 October 2018; Accepted 22 October 2018
0300-9440/ © 2018 Elsevier B.V. All rights reserved.
S. Sung et al. Progress in Organic Coatings 127 (2019) 37–44

Fig. 1. Schematic diagram illustrating the self-healing mechanisms underlying (a) the commercial (C-CL) and (b) the DA adduct-based isocyanate (DA-CL) polymer
networks.

Scheme 1. (a) Synthetic scheme and chemical structure of the furan-maleimide adduct diisocyanate (FMADI) crosslinking reagent; chemical structures of (b) the
polyisocyanate (Desmodur N3300) and (c) the acryl polyol.

Unfortunately, the scratch-healing performance of intrinsic healing demonstrated using in-situ oscillatory rheology combined with FT-IR
systems tends to decrease as their mechanical properties such as the spectroscopy. The physical properties of the DA-CL coating, such as the
elastic modulus and hardness are enhanced, i.e., as they increase, be- hardness (HIT), modulus (E), glass transition temperature (Tg), and
cause scratch healing relies on polymer chain mobility. thermal stability (Td), were characterized using nanoindentation
Intrinsic scratch-healing mechanisms are suitable for use in auto- testing, dynamic mechanical analysis (DMA), differential scanning ca-
motive scratch-healing coatings because they provide transparent and lorimetry (DSC), and thermal gravimetric analysis (TGA). The proper-
repeated scratch healing. Thermally reversible polymer networks based ties of the DA-CL coating were compared with those of a commercial
on DA/retro-DA reactions are remarkable for their transparent, color- automotive clear coat (C-CL) network system. Finally, the scratch re-
less, durable, and catalyst-free scratch-healing properties [17,18]. In sistance and healing performances of the DA-CL and C-CL polymer
this paper, we demonstrated the effectiveness of a DA/retro-DA reac- networks were quantitatively analyzed using a nanoscratch tester
tion-based scratch-healing system as an automotive clear coat. For this equipped with an optical microscope (OM) and atomic force micro-
purpose, a furan-maleimide adduct diisocyanate crosslinking agent scope (AFM). These materials were evaluated for their potential utility
(FMADI) was synthesized and reacted with a commercial clear coat as self-healing polymer systems.
system (DA-CL). The reversible formation of crosslinked and de-cross-
linked structures at the DA and retro-DA reaction temperatures was

38
S. Sung et al. Progress in Organic Coatings 127 (2019) 37–44

Fig. 2. (a) 1H NMR spectra of the synthesized DA adduct and FMADI, and (b) FT-IR spectra of FMADI for various urethane linkage formation reaction times.

2. Experimental method with a wire-wound rod. The solution was bar-coated onto the
basecoat layer and cured in an oven at 70 °C for 12 h. The completion of
2.1. Materials the crosslinking reactions in the polymer coatings was confirmed by
monitoring the characteristic ratio of the intensity of the peak due to
2.1.1. Materials and synthesis of FMADI the methylene unit (eCH2e) FT-IR band at 2950–2850 cm−1 to that of
Maleic anhydride, furan, ethanol amine, furfuryl alcohol, iso- the peak due to the isocyanate unit (eNCO) band at 2250 cm−1. As
phorone diisocyanate (IPDI), dibutyltin dilaurate (DBTDL), ethyl shown in Fig. S1, the isocyanate bands in the spectra of the polymer
acetate, ethanol, toluene, and 2-butanone were purchased from Sigma- coatings disappear after their crosslinking reactions, which indicates
Aldrich, and all materials were used without further purification. that the isocyanate units have been completely converted to urethane
Acrylic polyol as a representative clear coat resin and polyisocyanate units.
(Desmodur N3300) as a crosslinking reagent, obtained from NOROO
BEE Chemical Co., Ltd., were used as received. 2.2. Characterization
The furan-maleimide adduct was synthesized as described pre-
viously [19]. The furan-maleimide adduct and IPDI were added to 2- 2.2.1. Chemical structure confirmation
butanone in a 1:2 M ratio in the presence of DBTDL (1 wt% DA adduct), 1
H NMR spectra of the synthesized organic compounds were col-
and the mixture was heated at 35 °C while stirring for 6 h. Upon reac- lected using a 300 MHz NMR spectrometer (Bruker, Ultrashield) with
tion completion, the solvent mixture was evaporated under reduced DMSO-d6 (δ = 2.50 ppm) as a deuterated solvent. FT-IR spectra of the
pressure, and the product was stored under an inert atmosphere. FMADI (IPDI end-capped furan-maleimide adduct) were recorded over
the range 4000–800 cm−1 using an FT-IR spectrometer (Thermo Fisher
2.1.2. Preparation of the clear coats Scientific Inc., Nicolet 6700/Nicolet Continuum).
The acrylic polyol resin (vOH = 55, Mn = 3.5 kg/mol, 10 g,
2.9 mmol) and FMADI (vNCO = 123, 4.5 g, 6.6 mmol) or Desmodur 2.2.2. Thermal and viscoelastic properties
N3300 (vNCO = 208, 2.6 g, 5.2 mmol) were mixed in mole equivalent The thermal stabilities of the crosslinked DA-CL and C-CL polymer
ratios, which had been dissolved in 2-butanone. Each solution was network films were determined using TGA. The TGA (TA Instrument,
poured into a custom Teflon mold under ambient conditions, and was TGA Q500) measurements were performed between 30 and 800 °C at a
heated in an oven at 70 °C for 12 h. For the scratch test, a 30 μm wa- heating rate of 10 °C min−1 under nitrogen and air. DSC (TA
terborne primer layer and a 15 μm basecoat layer were applied in series Instrument, DSC Q2000) measurements were conducted by heating and
to a steel substrate using the conventional draw-down bar-coating cooling from −30 to 170 °C at a rate of 10 °C min−1 under a N2

39
S. Sung et al. Progress in Organic Coatings 127 (2019) 37–44

Variations in the chemical structure of the DA-CL polymer at the DA

and retro-DA temperature were characterized by tracking the CeH out-
of-plane bending vibrations attributed to the maleimide moiety at
827 cm–1. The variations in the intensities of the characteristic peaks of
the furan group (eCeH at 3100 cm-1, eC]Ce at 1510 cm-1 and
eCeOeCe at 1180 cm-1) in the DA-CL polymer film could not be
monitored during the thermocycle at the DA and retro-DA temperatures
due to the overlap between these peaks and those due to other chemical
groups in the clear coat (Fig. S2).
The evolution of heat during the thermally reversible crosslinking of
the DA-CL coating was monitored with DSC measurements over the
range −30 °C to 170 °C at a heating rate of 10 °C min−1. The enthalpy
changes and degree of conversion of the rDA reaction in the DA-CL
coating were calculated by using the following equations:
ΔHtotal = Wf × ΔHrDA (1)

ΔHexp
Degree of conversion (DC , %) = × 100
ΔHtotal (2)
where ΔHexp, ΔHtotal, ΔHrDA, and Wf are the enthalpy change of the rDA
reaction in the DA-CL coating obtained with DSC, the calculated total
enthalpy change of the rDA reaction in the DA-CL coating, the calcu-
lated total enthalpy change of the rDA reaction in FMADI, and the
weight fraction of FMADI in the DA-CL coating, respectively. ΔHrDA is
known to be approximately 18 kcal/mol (317 J/g) [20]. The re-
combination ratio of maleimides and furan groups after the thermal
rDA reaction was calculated with the following equation:
ΔHexp, n + 1
Recombination ratio =
ΔHexp, n (3)
where ΔHexp, n and ΔHexp, n + 1 are the ΔHexp values in the n-th cycle and
the (n+1)th cycle respectively.

2.2.4. Nanoindentation (NI)

Loading, holding, and unloading indentation measurements were
collected using a nanoindentation tester (Anton Paar, NHT3) with a
Berkovich-type indenter. Specimens with a thickness of 40 μm were
prepared with the method described in Section 2.1.2. During the
loading step, the force was gradually increased from 0 to 50 mN at a
rate of 20 mN min–1. The force was then maintained at 50 mN for 10 s
during the holding step, after which the indenter was unloaded in a
process that reversed the loading process. Through these three steps,
the maximum displacement (hmax), permanent depth of penetration
(final depth, hf), elastic unloading stiffness (S =dP/dh), indentation
Fig. 3. Thermal analyses: (a) TGA curves, (b) DSC thermograms, and (c) DMA
curves showing the viscoelastic properties of the crosslinked DA-CL and C-CL hardness (HIT), and indentation modulus (EIT) were obtained.
polymers.
2.2.5. Nanoscratch test (NST)
The NST was performed using a sphero-conical indenter (tip radius
environment. The glass transition temperatures (Tg values) of the
5 μm) mounted on a scratch test machine (Anton Paar, NST3). The force
crosslinked films were determined from the thermal transition mea-
on the coating was increased gradually from 0.40 to 20.4, 40.4, and
sured during the second heating run. The viscoelastic properties of the
60.4 mN to allow the formation of a scratch on the coating surface. The
crosslinked polymer films were evaluated by collecting DMA (TA
indenter speed and the scratch length were 1.6 mm min–1 and 0.8 mm,
Instruments, DMA Q800) measurements between −50 and 130 °C,
respectively. The scratch recovery results were achieved by heating the
using a 5 °C min−1 heating rate, an applied strain of 0.01%, and a
scratched films at 50, 70, or 150 °C for 1 h. After the scratching and
constant frequency of 1 Hz.
healing processes reached completion, the scratch-recovered sample
was analyzed using optical microscopy (OM) and atomic force micro-
2.2.3. Reversible formation and disintegration of the polymer network scopy (AFM).
structure
The reversible formation and disintegration of the DA-CL polymer 3. Results and discussions
network structure at the DA and retro-DA temperatures were monitored
by measuring the viscosity changes and chemical structures using an 3.1. Material design and synthesis
oscillatory rheometer coupled with FT-IR spectrometry (Thermo
Scientific Inc., HAAKE MARS III coupled with a Rheonaut). The oscil- In this study, we designed a polyacrylate urethane network con-
latory rheometer was operated between 70 and 130 °C at a constant taining a DA adduct (DA-CL) to create a scratch-healing clear coat that
frequency of 1 Hz, with a strain of 0.1%, the parallel plate at 8 mm, and exhibited thermally-reversible crosslinking. As a control, a commercial
an axial force of 5 N. The rate of heating and cooling was 2 °C min−1. clear coat (C-CL) was prepared, and its material properties, thermal

40
S. Sung et al. Progress in Organic Coatings 127 (2019) 37–44

Fig. 4. Reversible crosslinking and de-crosslinking of the DA-CL coating at the DA and rDA temperatures. (a) The rheological properties of the DA-CL and C-CL
polymer network structures, as revealed by measurements of the complex viscosity; (b) FT-IR spectra of the DA-CL polymer network structure during the ther-
mocycles, obtained with a Rheonaut (2 °C min–1); (c) FT-IR spectra collected at 827 cm–1; (d) DSC integrals for the DA-CL sample: first, second, and third cycles.

Table 1
Degrees of conversion (DC) and furan and maleimide recombination ratios after
repeated rDA reactions.
Cycle ΔHtotal (J/g) ΔHexp (J/g) DC (%) Recombination ratio

1 6.3 2.40 38 –
2 6.3 2.19 35 0.91
3 6.3 2.18 35 0.99

behavior, and scratch-healing performance were compared with the

corresponding properties of the DA-CL (Fig. 1). The FMADI crosslinker
was prepared through a multi-step synthesis approach (Scheme 1). The
chemical structure of the di-hydroxyl-terminated DA adduct and FMADI
were confirmed using 1H NMR and IR spectroscopy (Fig. 2). In the 1H
NMR spectra, the furan-maleimide adduct units of both the di-hydroxyl-
terminated DA adduct and FMADA produce clear chemical shifts at 6.5,
5.07, and 3.02–2.89 ppm corresponding to the eCH]CHe, eCHeOe,
and eCHeCHe groups respectively. For the di-hydroxyl terminated DA
adduct, the characteristic peaks of the two eOH groups appear at 4.92
and 4.75 ppm (Fig. 2a). However, these peaks completely disappear
after the reaction of urethane with IPDI, and a characteristic peak due
to the -NH group in the urethane linkage appears at 5.5 ppm. The
completion of the urethane reaction was also confirmed by monitoring
the ratio of the intensity of the peak due to the methylene unit at
2950–2850 cm−1 to that of the band due to the isocyanate unit (eNCO)
at 2250 cm−1. As shown in Fig. 2b, this characteristic peak ratio gra-
dually decreases as the reaction time increases but is constant after 6 h, Fig. 5. (a) Nanoindentation experiments determined the load–displacement
which indicates that the reaction is complete. curves, and (b) nanoscratch tests determined the scratch resistance profiles and
provided the images of the scratches (inset) on the polymer coatings.
3.2. Thermal and viscoelastic properties of the polymer networks

The relative thermal stabilities of the DA-CL and C-CL were in-
vestigated using TGA. All polymers were stable up to at least 150 °C

41
S. Sung et al. Progress in Organic Coatings 127 (2019) 37–44

two polymers were similar, although Tg for C-CL (10 °C) was slightly
lower than the corresponding value of DA-CL (20 °C).

3.3. Thermally reversible crosslinking in the polymer networks

Quantitative analysis was performed using in-situ oscillatory

rheology and FT-IR spectroscopy measurements to demonstrate the
thermally reversible formation and disintegration of the crosslinked DA
adduct (DA-CL) polymer network (Fig. 4). Fig. 4a presents the complex
viscosity variations in the crosslinked films measured using an oscilla-
tory rheometer. The thermal reversibility and repeatability were ob-
served over the temperature range 70–130 °C. The film must display
thermal stability at 130 °C, the temperature at which the retro-DA re-
action occurred. The thermal stabilities of the polymer networks were
characterized using TGA, and all polymers were found to be stable up to
150 °C. The polymer resin complex viscosity decreased gradually due to
the destruction of the polymer network as the temperature increased to
130 °C. As the temperature decreased to 70 °C, the complex viscosity
Fig. 6. Optical microscopy images obtained after the nanoscratch tests (NST), steadily returned to its original values. The increase and decrease in the
after scratching and healing at 50 °C and 150 °C under a stepwise force applied viscosity of the DA-CL polymers during the heating and cooling cycles
from 0.4 mN to 60.4 mN to the DA-CL and C-CL polymer coatings. were similar to the properties observed in the commercial crosslinker
(C-CL) polymer network. Interestingly, the gap between the highest and
(Fig. 3a). The dependence of the polymer viscoelastic properties on the lowest viscosity values was much higher in DA-CL than in C-CL [25,26].
temperature was determined using DMA (Fig. 3c). As shown in the plots This result likely arose from the formation of highly mobile phases due
of G’ vs. temperature, the G’ values of the two polymeric systems in the to the disintegration of the DA-CL polymer network structure, in
glassy state were stable, but the moduli dropped steeply over ap- agreement with the results obtained from the DMA and DSC experi-
proximately four orders of magnitude as the temperature increased, ments. An in-situ investigation of the chemical structure variations in
indicating that the polymer networks assumed a rubbery state. The G’ the DA-CL polymer network at both the DA and retro-DA reaction
value of the DA-CL continued to decrease rapidly as the temperature temperatures was conducted by using an FT-IR spectrometer connected
increased further up to 130 °C, signaling that the network structure of to a rheometer. As shown in Fig. 4b, the formation of a maleimide band
the polymer disintegrated due to the retro-DA reaction, whereas the G’ at 827 cm–1 (CeH out-of-plane bending vibration) during the retro-DA
value of the C-CL remained stable [21,22]. The DSC results revealed a process was detected, indicating that the polymer network structure
broad endothermic transition over the temperature range 100–170 °C had disintegrated [27]. Furthermore, the maleimide band disappeared
for the DA-CL polymer networks. This transition corresponded to the when the polymer was heated to the DA reaction temperature, in-
breaking of the DA bonds, as shown in Fig. 3b (inset) [23,24]. The Tg dicating that the polymer network structure was recovered. The pre-
values of the polymers were determined using DSC. The Tg values of the sence of residual maleimide bands after the DA process indicated that
recombination between the furan and maleimide moieties in the DA-CL

Fig. 7. Penetration depth before (black) and after (red) heat treatment (50 °C and 150 °C) of the DA-CL and C-CL polymer coatings. (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article).

42
S. Sung et al. Progress in Organic Coatings 127 (2019) 37–44

Fig. 8. Representative AFM images of the scratched and healed surfaces heated to 150 °C upon application of constant forces (0.4 mN, 20.4 mN, 40.4 mN, or 60.4 mN)
to the DA-CL and C-CL polymer coatings.

Table 2 adduct moieties and the cyclic chemical structure of IPDI in the DA-CL
Scratch-depth-based healing efficiency (%DSHE) values for the polymer coat- polymer increase its HIT value [28]. In fact, the DA-CL polymer network
ings. exhibited a higher Tg (20 °C) than the C-CL polymer network (10 °C)
Polymer Code Load (mN) %DSHE due to the steric effects of the bulky FMADI moiety and the cyclic
structure (Fig. 3b). The scratch hardness was assessed by measuring the
50 °C 150 °C scratch resistance of the polymer coatings in response to a step-wise
increase in the deformation load using the scratch test machine to exert
DA-CL 20.4 29.3 100
40.4 30.9 100 four sequentially increasing loads applied under 0.40–60.4 mN, as
60.4 29.7 100 shown in Fig. 5b. The optical microscopy images show that the scratch
C-CL 20.4 74.2 93.9 path width of the DA-CL polymer film was narrower than that of the C-
40.4 58.6 87.0 CL polymer film (Fig. 5b, inset). The surface of the DA-CL polymer
60.4 57.5 85.2
displayed smooth scratch paths, whereas the surface of the C-CL
polymer exhibited tears and large ruptures in the scratched groove. It is
polymer was incomplete due to the short reaction time. The re- possible that a longer chain length between the crosslinked networks
combination ratio between the furan and maleimide moieties in the DA- increased the material flexibility. We concluded that the hard DA-CL
CL polymer was approximately 0.50 per cycle. The DA and retro-DA polymer film provided better scratch resistance than the soft C-CL
reactions proceeded repeatedly in the DA-CL, as indicated by the IR polymer film. This conclusion is consistent with the results obtained
peak at 827 cm–1, which depended on the measurement time and was from the nanoindentation studies.
not observed in C-CL (Fig. 4c).
The enthalpy changes in the DA and rDA reactions were determined
3.5. Scratch healing in the clear coat systems
over multiple heating and cooling cycles by using DSC (Fig. 4d). The
ΔHexp values of DA-CL for the three heating and cooling cycles are listed
The scratch recovery results obtained from the DA-CL and C-CL
in Table 1. The degree of conversion of the thermal rDA reaction is
polymers were compared by analyzing images obtained using OM or
approximately 40% after the first heating cycle. The ΔHexp values are
AFM. The scratched polymer films were subjected to the target tem-
slightly lower for the later cycles, which indicates that the yield of the
perature for 1 h using the Peltier heating module of the scratch in-
DA reaction is not 100% under these conditions. However, a network
dentation machine (Fig. 6). After healing at 50 °C, the scratched areas in
structure forms in the DA-CL coating with a high recombination ratio of
both polymer-coated films were shown to decrease, but the soft C-CL
at least 0.90 per cycle.
coating showed superior scratch healing compared to the DA-CL
polymer film due to its lower Tg and HIT values [29,30]. Elastic de-
3.4. Mechanical properties of the clear coat systems formation is a necessary component of the self-recovery process, espe-
cially at temperatures above Tg, at which recovery is dominated by
Fig. 5a shows typical load–displacement curves obtained from the elasticity. These results were expected at a temperature of 50 °C, con-
nanoindentation tests carried out on the DA-CL and C-CL polymer sidering that Tg was calculated from the DMA results to be approxi-
networks. The mechanical properties of the films were evaluated to mately 45 °C for C-CL and 65 °C for DA-CL. Heating at 70 °C sig-
determine whether the clear coat provided an appropriate hardness nificantly improved the scratch-healing performances of both the C-CL
(107–108 Pa). All coatings exhibited plastic deformation in response to and DA-CL compared to the results obtained from heating at 50 °C (data
the applied normal forces; however, the DA-CL polymer showed lower not shown). The healing process occurred at 150 °C, the maximum
hmax and hf values than the C-CL polymer. Similarly, the calculated HIT temperature at which the retro-DA reactions occurred, without indu-
value of the DA-CL polymer (HIT (DA-CL) = 85 MPa) was higher than cing polymer decomposition. The scratched groove in the DA-CL
the corresponding value obtained from the C-CL polymer (HIT (C- polymer heated to 70 °C completely disappeared when treated at 150 °C
CL) = 69 MPa). We conclude that the steric bulky furan-maleimide for 1 h, whereas the scratched surface of the C-CL polymer film was

43
S. Sung et al. Progress in Organic Coatings 127 (2019) 37–44

imperfectly healed. These results were attributed to the greater polymer Self-healing gels based on constitutional dynamic chemistry and their potential
chain mobility at 150 °C than at 70 °C due to polymer network disin- applications, Chem. Soc. Rev. 43 (2014) 8114–8131, https://doi.org/10.1039/
C4CS00219A.
tegration in the DA-CL [31]. The depth profiles of the scratched and [6] L. Zedler, M.D. Hager, U.S. Schubert, M.J. Harrington, M. Schmitt, J. Popp,
healed surfaces of the polymer films were acquired using an AFM at- B. Dietzek, Monitoring the chemistry of self-healing by vibrational spectroscopy –
current state and perspectives, Mater. Today. 17 (2014) 57–69, https://doi.org/10.
tached to the scratch indentation machine (Figs. 7 and 8). The depth 1016/j.mattod.2014.01.020.
profile data were used to calculate the scratch depth healing efficiency [7] S.M. Noh, J.H. Nam, J.K. Oh, H.W. Jung, Scratch and recovery characteristics of
(%DSHE) values according to the equation automotive clearcoats containing blocked polyisocyanate crosslinkers, J. Coat.
Technol. Res. 12 (2015) 85–95, https://doi.org/10.1007/s11998-014-9617-4.
D − Dr ⎤ [8] S.M. Noh, J.W. Lee, J.H. Nam, J.M. Park, H.W. Jung, Analysis of scratch char-
%DSHE = ⎡ i × 100 acteristics of automotive clearcoats containing silane modified blocked isocyanates
⎢
⎣ Di ⎥
⎦ via carwash and nano-scratch tests, Prog. Org. Coat. 74 (2012) 192–203, https://
doi.org/10.1016/j.porgcoat.2011.12.009.
where Di and Dr indicate the scratch depth and residual depth, re- [9] S.H. Cho, S.R. White, P.V. Braun, Self-healing polymer coatings, Adv. Mater. 21
spectively. The calculated %DSHE is presented in Table 2. The scratch- (2009) 645–649, https://doi.org/10.1002/adma.200802008.
[10] Y.K. Song, B. Kim, T.H. Lee, J.C. Kim, J.H. Nam, S.M. Noh, Y.I. Park, Fluorescence
healing performances of the DA-CL and C-CL polymer films increased detection of microcapsule-type self-healing, based on aggregation-induced emis-
with increasing temperature; however, the DA-CL polymer exhibited a sion, Macromol. Rapid Commun. 38 (2017), https://doi.org/10.1002/marc.
201600657 n/a-n/a.
superior scratch-healing performance against deep scratches generated [11] Y.K. Song, B. Kim, T.H. Lee, S.Y. Kim, J.C. Kim, S.M. Noh, Y.I. Park, Monitoring
from 40.4 and 60.4 mN compared to the C-CL polymer. As mentioned in fluorescence colors to separately identify cracks and healed cracks in microcapsule-
the context of the NST experiments, the C-CL polymer formed tears and containing self-healing coating, Sens. Actuators B Chem. 257 (2018) 1001–1008,
https://doi.org/10.1016/j.snb.2017.11.019.
large ruptures in the scratched groove. These features may have re- [12] N. Roy, B. Bruchmann, J.-M. Lehn, DYNAMERS: dynamic polymers as self-healing
sulted from a loss of mass resulting from the material softness and materials, Chem. Soc. Rev. 44 (2015) 3786–3807, https://doi.org/10.1039/
C5CS00194C.
ductile film coating. Although chemical recombination is important for
[13] A.J.R. Amaral, G. Pasparakis, Stimuli responsive self-healing polymers: gels, elas-
scratch healing, scratch resistance is of comparable importance to tomers and membranes, Polym. Chem. 8 (2017) 6464–6484, https://doi.org/10.
performance. 1039/C7PY01386H.
[14] S. Young An, D. Arunbabu, S. Man Noh, Y. Kyu Song, J. Kwon Oh, Recent strategies
to develop self-healable crosslinked polymeric networks, Chem. Commun. 51
4. Conclusions (2015) 13058–13070, https://doi.org/10.1039/C5CC04531B.
[15] H. Ying, Y. Zhang, J. Cheng, Dynamic urea bond for the design of reversible and
self-healing polymers, Nat. Commun. 5 (2014) 3218, https://doi.org/10.1038/
This report describes the application of a thermally reversible DA/ ncomms4218.
retro-DA reaction-based scratch-healing polymer network system for [16] J.I. Park, A. Choe, M.P. Kim, H. Ko, T.H. Lee, S.M. Noh, J.C. Kim, I.W. Cheong,
Water-adaptive and repeatable self-healing polymers bearing bulky urea bonds,
use in automotive clear coats. In-situ oscillatory rheology measurements Polym. Chem. 9 (2018) 11–19, https://doi.org/10.1039/C7PY01655G.
coupled with IR spectroscopy revealed that thermally reversible cross- [17] Y.-L. Liu, T.-W. Chuo, Self-healing polymers based on thermally reversible
linking and de-crosslinking in a DA-CL polymer network containing an Diels–Alder chemistry, Polym. Chem. 4 (2013) 2194–2205, https://doi.org/10.
1039/C2PY20957H.
FMADI moiety successfully resisted and administered scratches. The [18] M.A. Tasdelen, Diels–Alder “click” reactions: recent applications in polymer and
scratch-healing performance and material properties of the DA-CL material science, Polym. Chem. 2 (2011) 2133–2145, https://doi.org/10.1039/
polymer network, including the Tg, Td, HIT and complex viscosity va- C1PY00041A.
[19] S.Y. Kim, T.H. Lee, Y.I. Park, J.H. Nam, S.M. Noh, I.W. Cheong, J.C. Kim, Influence
lues, were characterized and found to be superior to those of the C-CL of material properties on scratch-healing performance of polyacrylate-graft-poly-
polymer network. Our results indicate that thermo-reversible polymer urethane network that undergo thermally reversible crosslinking, Polymer 128
(2017) 135–146, https://doi.org/10.1016/j.polymer.2017.09.021.
networks based on a DA-functionalized polymer are potentially useful [20] R.C. Boutelle, B.H. Northrop, Substituent effects on the reversibility of
as scratch-healing systems for automotive clear coats. Furan–Maleimide cycloadditions, J. Org. Chem. 76 (2011) 7994–8002, https://doi.
org/10.1021/jo201606z.
[21] A. Wang, H. Niu, Z. He, Y. Li, Thermoreversible cross-linking of ethylene/propylene
Acknowledgements copolymer rubbers, Polym. Chem. 8 (2017) 4494–4502, https://doi.org/10.1039/
C7PY00896A.
[22] L.M. Polgar, M. van Duin, A.A. Broekhuis, F. Picchioni, Use of Diels–Alder chemistry
The authors are very grateful for financial support from the Ministry
for thermoreversible cross-linking of rubbers: the next step toward recycling of
of Trade, Industry & Energy of Korea (Industrial Technology Innovation rubber products? Macromolecules. 48 (2015) 7096–7105, https://doi.org/10.
Program No. 10067082). Technical support (scratch tester) from Anton 1021/acs.macromol.5b01422.
[23] G. Postiglione, S. Turri, M. Levi, Effect of the plasticizer on the self-healing prop-
Paar is gratefully acknowledged. Additionally, the authors thank Mr. erties of a polymer coating based on the thermoreversible Diels–Alder reaction,
Gregory Favaro, Ms. Mihaela Dubuisson, Wonkeun Park and Prog. Org. Coat. 78 (2015) 526–531, https://doi.org/10.1016/j.porgcoat.2014.05.
Jungwoong Ma for their efforts and valuable comments related to the 022.
[24] X. Kuang, G. Liu, X. Dong, D. Wang, Triple-shape memory epoxy based on
scratch-healing experiments. Diels–Alder adduct molecular switch, Polymer 84 (2016) 1–9, https://doi.org/10.
1016/j.polymer.2015.12.033.
[25] D.H. Turkenburg, H.R. Fischer, Diels-Alder based, thermo-reversible cross-linked
Appendix A. Supplementary data epoxies for use in self-healing composites, Polymer 79 (2015) 187–194, https://doi.
org/10.1016/j.polymer.2015.10.031.
Supplementary material related to this article can be found, in the [26] R.K. Bose, J. Kötteritzsch, S.J. Garcia, M.D. Hager, U.S. Schubert, S. van der Zwaag,
A rheological and spectroscopic study on the kinetics of self-healing in a single-
online version, at doi:https://doi.org/10.1016/j.porgcoat.2018.10.023. component diels–alder copolymer and its underlying chemical reaction, J. Polym.
Sci. Part Polym. Chem. 52 (2014) 1669–1675, https://doi.org/10.1002/pola.
27164.
References
[27] L.-T.T. Nguyen, T. Thu Truong, H. Tran Nguyen, L. Le, V. Quoc Nguyen, T.V. Le,
A. Tuan Luu, Healable shape memory (thio)urethane thermosets, Polym. Chem. 6
[1] S.B. Ulaeto, R. Rajan, J.K. Pancrecious, T.P.D. Rajan, B.C. Pai, Developments in (2015) 3143–3154, https://doi.org/10.1039/C5PY00126A.
smart anticorrosive coatings with multifunctional characteristics, Prog. Org. Coat. [28] D.N. Amato, G.A. Strange, J.P. Swanson, A.D. Chavez, S.E. Roy, K.L. Varney,
111 (2017) 294–314, https://doi.org/10.1016/j.porgcoat.2017.06.013. C.A. Machado, D.V. Amato, P.J. Costanzo, Synthesis and evaluation of thermally-
[2] V. Anand Ganesh, H. Kumar Raut, A. Sreekumaran Nair, S. Ramakrishna, A review responsive coatings based upon Diels–Alder chemistry and renewable materials,
on self-cleaning coatings, J. Mater. Chem. 21 (2011) 16304–16322, https://doi. Polym. Chem. 5 (2014) 69–76, https://doi.org/10.1039/C3PY01024D.
org/10.1039/C1JM12523K. [29] S.J. Garcia, Effect of polymer architecture on the intrinsic self-healing character of
[3] K. Urdl, A. Kandelbauer, W. Kern, U. Müller, M. Thebault, E. Zikulnig-Rusch, Self- polymers, Eur. Polym. J. 53 (2014) 118–125, https://doi.org/10.1016/j.eurpolymj.
healing of densely crosslinked thermoset polymers—a critical review, Prog. Org. 2014.01.026.
Coat. 104 (2017) 232–249, https://doi.org/10.1016/j.porgcoat.2016.11.010. [30] Y. Yang, M.W. Urban, Self-healing polymeric materials, Chem. Soc. Rev. 42 (2013)
[4] D.G. Bekas, K. Tsirka, D. Baltzis, A.S. Paipetis, Self-healing materials: A review of 7446–7467, https://doi.org/10.1039/C3CS60109A.
advances in materials, evaluation, characterization and monitoring techniques, [31] S. Jung, S. Young Kim, J. Chul Kim, S. Man Noh, J. Kwon Oh, Ambient temperature
Compos. Part B Eng. 87 (2016) 92–119, https://doi.org/10.1016/j.compositesb. induced Diels–Alder crosslinked networks based on controlled methacrylate copo-
2015.09.057. lymers for enhanced thermoreversibility and self-healability, RSC Adv. 7 (2017)
[5] Z. Wei, J. Hai Yang, J. Zhou, F. Xu, M. Zrínyi, P.H. Dussault, Y. Osada, Y. Mei Chen, 26496–26506, https://doi.org/10.1039/C7RA04222A.

44