Influence of material properties on scratch-healing performance of polyacrylate-graft-polyurethane network that undergo thermally reversible crosslinking

Polymer 128 (2017) 135e146

Contents lists available at ScienceDirect

Polymer journal homepage: www.elsevier.com/locate/polymer

Influence of material properties on scratch-healing performance of polyacrylate-graft-polyurethane network that undergo thermally reversible crosslinking So Young Kim a, b, 1, Tae Hee Lee a, c, 1, Young Il Park a, Joon Hyun Nam a, Seung Man Noh a, **, In Woo Cheong b, ***, Jin Chul Kim a, * a b c

Research Center for Green Fine Chemicals, Korea Research Institute of Chemical Technology, Ulsan 44412, Republic of Korea School of Applied Chemistry, Kyungpook National University (KNU), Daegu 41566, Republic of Korea Department of Chemical Engineering, Ulsan National Institute of Science and Technology, Ulsan 44919, Republic of Korea

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 June 2017 Received in revised form 4 September 2017 Accepted 10 September 2017 Available online 13 September 2017

Scratch-healing poly (methyl methacrylate)-co-[poly (methyl metharyleate)-graft-(oligo-caprolactone)] urethane networks containing a Diels Alder (DA) adduct unit (GCPNp-DAs) were successfully synthesized and shown to be capable of undergoing thermally reversible crosslinking. The synthesized polymers were coated on steel substrates to investigate the influence of their material properties on their scratchhealing performance. The reversible formation of crosslinked and de-crosslinked structures of the GCPNp-DAs at DA and retro-DA (rDA) reaction temperatures was demonstrated using FT-IR spectroscopy, differential scanning calorimetry (DSC), oscillatory rheology, and nanoindentation. The scratchresistance and healing performances of the GCPNp-DA coatings were evaluated quantitatively using a scratch test machine equipped with an optical microscope (OM) and an atomic force microscope (AFM). These results were found to be greatly influenced by the material properties of the coatings such as the elastic modulus, indentation hardness (HIT), crosslinking density (vc), and thermal transition temperature as well as by whether the deforming load that produced the scratches was increased in a progressive (gradual) or step-wise manner. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Intrinsic self-healing Graft copolymer Diels alder reaction Reversible crosslinking Scratch resistance Scratch-healing

1. Introduction Polymeric coatings are employed in industrial applications such as electronics, transportation, construction and medical devices to protect substrate surfaces from physical damage, corrosion and microbial contamination [1e8]. In such systems, however, scratching of the polymeric coating quite frequently leads to corrosion or contamination of the substrate material, degrading the function or appearance of the substrate and reducing the service life of the product. Considerable research effort has therefore been devoted to eliminating or reducing the influence of scratches. The main strategies in this regard fall into two categories:

* Corresponding author. ** Corresponding author. *** Corresponding author. E-mail address: [email protected] (J.C. Kim). 1 S. Y. Kim and T. H. Lee contributed equally to this work. http://dx.doi.org/10.1016/j.polymer.2017.09.021 0032-3861/© 2017 Elsevier Ltd. All rights reserved.

preventing scratches from forming in the first place, or healing them if they do form [9e11]. The basic principle behind preventing scratch formation is to improve the hardness and toughness of the coating by increasing the matrix crosslinking density (nc) and/or adding inorganic particles. While this approach is attractive in that it blocks scratch formation, it cannot be used for applications that require high elasticity, processability, compatibility, dispersion stability and/or optical transparency [12]. The techniques for healing scratches, on the other hand, use extrinsic or intrinsic healing mechanisms. In general, extrinsic scratch-healing of a polymeric coating is achieved by including microcapsules into the coating that can heal the scratches. This approach has the advantage that it provides autonomous (self) scratch healing without the need for an external stimulus [13e16]. However, it cannot be used for transparent coatings because the different reflective indices of micro-sized capsules and coating materials often cause the coating to become cloudy [17,18]. In addition, when a scratch heals in these systems, the microcapsules

136

S.Y. Kim et al. / Polymer 128 (2017) 135e146

collapse. Thus, scratch healing is effective the first time a region is scratched, but is increasingly ineffective if repeated scratching occurs [14]. Moreover, the inclusion of microcapsules into the coating lowers the mechanical strength and thermal stability of the coating [19,20]. Intrinsic scratch-healing of polymer coatings can occur via diverse mechanisms, including the formation of physical interactions including hydrogen bonds and charge-transfer complexes; the use of catalytic reactions; and the inclusion of dynamic crosslinking systems [21,22] such as Diels-Alder (DA) reactions [23e27] and reactions that form hindered urea bonds [28e30], alkoxyamines [31,32], boronic esters and di-thiolate bonds [33e37], with the healing process triggered by an external stimulus such as thermal energy, ultraviolet light, a change in pH, or the introduction of moisture. The advantages of this approach over the extrinsic scratch-healing technique are that it can be applied to a wide range of organic coating materials [11], it can be used with transparent coatings [38,39], it allows the repeatedly healing of scratches [40,41]. Overall, however, the durability of the coating is inversely related to its scratch-healing performance because the former requires high nc, while the latter relies on enhanced polymer chain mobility so that polymer chain rearrangements and hence healing can occur more easily. Therefore, striking the right balance between the material properties and the self-healing performance is a key factor in designing high-performance scratch-healing coating materials. From this point of view, the graft polymer would be a feasible candidate because its material properties can be easily tailored by varying the type, length, contents, and chemical structure of its bristles [8,42,43]. Intrinsic self-healing polymers based on the thermally reversible DA and retro-DA (rDA) reaction systems are versatile in that they provide transparent, colorless, durable, and catalyst-free coatings. When designing the system, it is necessary to quantitatively analyze two important factors: the reversible formation of the crosslinked and de-crosslinked structures, and the dependence of the scratch-healing performance on material properties such as hardness, scratch resistance, and thermal and viscoelastic properties To date, the former has been well investigated but only a few studies were found to have analyzed the latter, therefore, more systematic research is needed [21,44e47]. In the current study, we conducted a detailed investigation of the relationship between the material properties and scratchhealing performance of poly (methyl methacrylate)-co-[poly (methyl metharyleate)-graft-(oligo-caprolactone)] urethane networks containing a DA adduct unit (GCPNp-DAs). In the first part of the paper we demonstrate the reversible formation of the crosslinked and de-crosslinked structures of the GCPNp-DAs at DA and rDA reaction temperatures. In addition, the difference between the viscoelastic properties of GCPNp-DAs and those of conventional crosslinked poly (urethane acrylate)s (GCPNp-Hs), which otherwise have similar chemical structures and vc levels, was also investigated when both polymers were subjected to DA and rDA temperatures. The second part of the paper describes the quantitative analysis of the scratch-resistance and healing performances of the coatings as a function of material properties such as elastic modulus, indentation hardness (HIT), vc , glass transition temperature (Tg) and thermal stability. 2. Experimental section 2.1. Materials Maleic anhydride, furan, ethanol amine, furfuryl alcohol, isophorone diisocyanate (IPDI), dibutyltin dilaurate (DBTDL), 2,20 azobis (2-methylpropionitrile) (AIBN), ethyl acetate, ethanol,

toluene and 2-butanone were purchased from Sigma-Aldrich and were used as received. Methyl methacrylate (MMA) obtained from Sigma-Aldrich and hydroxyl-terminated caprolactone methacrylate monomers provided from Mirae Chemicals were passed through a basic alumina column to remove inhibitors before polymerization. 2.2. Synthesis 2.2.1. 4,10-Dioxatricyclo[5.2.1.02,6]dec-8-ene-3,5-dione (1) Maleic anhydride (10 g, 102 mmol) and furan (9.24 g, 136 mmol) were mixed and stirred in ethyl acetate (20 mL) at room temperature for 24 h. After completion of the reaction, precipitated white powder was filtered, washed with toluene, and dried in vacuo. No further purification step was needed (yield: 85.6%). 1H NMR (300 Hz, DMSO-d6) peak d values: 6.59 (s, 2H), 5.36 (s, 2H), and 3.32 (s, 2H) ppm. 2.2.2. 4-(2-Hydroxy-ethyl)-10-oxa-4-aza-tricyclo[5.2.1.02,6]dec-8ene-3,5-dione (2) To a 250 mL two-neck round-bottom flask in an ice bath, the product 1 (30 g, 120 mmol) and ethanol (40 mL) were added and stirred for 15 min. Afterwards, a solution of ethanolamine (7.58 g, 124 mmol) in ethanol (10 mL) was slowly dropped over the course of 30 min into the flask in the ice bath under stirring and the reaction mixture was then kept at room temperature for an additional 30 min until 1 completely dissolved. Then, the solution was refluxed for 24 h. After the reaction completed, the solution was cooled down. The obtained yellowish solid powder was filtered and dried in vacuo (yield 46.6%). 1H NMR (300 Hz, DMSO-d6) peak d values: 6.53 (s, 2H), 5.10 (s, 2H), 4.74 (br, 1H), 3.40 (br, 4H), and 2.90 (s, 2H) ppm. 2.2.3. N-(2-Hydroxyethyl)-maleimide (3) To a 250 mL two-neck round-bottom flask equipped with a reflux condenser, the product 2 (10 g, 47.8 mmol) was added and refluxed in toluene (100 mL) for over 24 h. The degree of conversion of 2 to HEM was monitored by tracking the decrease in the intensities of the 1H NMR peaks at 6.520 ppm and 5.104 ppm, which are characteristic peaks of the furan moiety. After the completion of the reaction, the solution was cooled down. The white powders crystallized from the solution were filtered and dried in vacuo (yield 83.02%). 1H NMR (300 Hz, DMSO-d6) peak d values: 6.97 (s, 2H), 4.80 (br, 1H), and 3.44 (br, 4H) ppm. 2.2.4. DA adducts (DAA) To a one-neck round-bottom flask, HEM and furfuryl alcohol in a 1:1 M ratio were added into toluene and then the resulting mixture was heated at 75  C under stirring for 12 h. Light-yellowish precipitate formed and this product was filtered, washed twice with ether, and dried in vacuo. 1H NMR (300 MHz, CDCl3) peak d values (ppm): 6.52 (br, 2H), 5.07 (s, 1H), 4.92 (br, 1H), 4.75 (br, 1H), 4.01 (d, 1H), 3.71 (d, 1H), 3.41 (s, 4H), 3.03 (d, 1H), and 2.89 (d, 1H). 2.2.5. DA diisocyanate crosslinker (DADI) For DAA and IPDI, each monomer was added into 2-butanone with a 1:2 M ratio and then the mixtures were heated at 35  C in the presence of DBTDL under stirring for 6 h. After the reaction completed, the reaction solvent was evaporated under reduced pressure and the product was stored under a nitrogen atmosphere. FT-IR bands (cm1): 3354 (-NH), 2950e2850 (-C-H2), 2270 (-NCO), 1720 (C¼O), and 1540 (C-N). 2.2.6. Diisocyanate crosslinker (DI) For 1,6-hexanediol and IPDI, each monomer was added into 2-

S.Y. Kim et al. / Polymer 128 (2017) 135e146

butanone with a 1:2 M ratio and then the mixtures were heated at 70  C in presence of DBTDL under stirring for 3 h. After the reaction completed, the reaction solvent was evaporated under reduced pressure and the product was stored under a nitrogen atmosphere. FT-IR bands (cm1): 3354 (-NH), 2950e2850 (-C-H2), 2270 (-NCO), and 1540 (C-N). 2.2.7. Hydroxyl-terminated graft copolymers (GCPp-OHs) GCP1-OH, GCP2-OH and GCP4-OH were synthesized in 2butanone from MMA (80 mol%) and the corresponding hydroxyl terminated oligocaprolactone acrylates (20 mol%) by carrying out a conventional free radical polymerization in the presence of AIBN at 70  C for 3 h. After being precipitated and hence purified from diethyl ether, the polymers were filtered and dried in vacuo for 12 h. GCP1-OH: yield: 72%, 1H NMR (300 MHz, CDCl3) peak d values: 4.3 (br, 2H), 4.1 (br, 2H), 3.5 (br, 13H), 2.4 (br, 3H), 1.9 (br, 6H), and 0.8e1.6 (br, 27H) ppm; ¼ 3.7 kg/mol, ÐM ¼ 2.20. GCP2-OH: yield: 66%, 1H NMR (300 MHz, CDCl3) peak d values: 4.3 (br, 2H), 4.1 (br, 2H), 3.5 (br, 16H), 2.4 (br, 4H), 1.9 (br, 10H), and 0.8e1.6 (br, 26H) ppm; ¼ 3.8 kg/mol, ÐM ¼ 2.53. GCP4-OH: yield: 50%, 1H NMR (300 MHz, CDCl3) peak d values: 4.3 (br, 2H), 4.1 (br, 2H), 3.5 (br, 19H), 2.4 (br, 4H), 1.9 (br, 26H), and 0.8e1.6 (br, 25H) ppm; ¼ 4.1 kg/mol, ÐM ¼ 2.15. 2.2.8. Preparation of GCPNp-DA and GCPNp-H coatings As a test substrate, phosphated steel panels were coated with a 20-mm-thick layer of epoxy amine using electrical deposition.
basecoat layers were Waterborne primer and waterborne blase deposited in series with thicknesses of 30 mm and 15 mm, respectively, using the conventional draw-down bar coating method. Finally, the prepared GCPNp series were coated on the base coat with a 40e50 mm thickness and subsequently cured at 70  C for 1 h. 2.3. Characterizations 2.3.1. Chemical structure confirmation 1 H NMR spectra of the organic compounds and GCPp-OH polymers were recorded using a 300 MHz NMR spectrometer (Bruker, Ultrashield) at ambient condition. The DMSO-d6 multiplet at 2.50 ppm was selected as the reference standards. FT-IR spectra of the DA and DADI were recorded in the region 4000-400 cm1 on a FT-IR spectrometer (Nicolet 6700/Nicolet Continuum FT-IR spectrometer, Thermo Fisher Scientific Inc.) using transmittance mode at ambient condition. 2.3.2. Molecular weight determination The number average molecular weight (Mn ) and molecular weight distribution (ÐM) of the synthesized GCPp-OHs were determined using a size-exclusion chromatography (SEC) apparatus (Agilent Tech., 1260 Infinity) equipped with a set of gel columns (Agilent PLgel 5 mm MIXED-D column). The system was equilibrated at 25  C in anhydrous tetrahydrofuran (THF), which was used as the polymer solvent and eluent with a flow rate of 1 mL min1, and calibrated with polystyrene standards (650e6,375,000 Mw ; Mw is the weight-average molecular weight). 2.3.3. Characterization of the thermally reversible crosslinking reaction The change in the chemical structure of the DA adduct unit in the GCPNp-DAs at DA (70  C) and rDA (130  C) temperatures were investigated using an attenuated total reflectance (ATR) mode of the FT-IR spectrometer (Thermo Fisher Scientific Inc., Nicolet 6700/ Nicolet Continuum). The evolution of heat during the thermally reversible crosslinking reaction of the GCPNp-DAs was monitored

137

by taking differential scanning calorimetry (DSC) measurements (TA Instruments, DSC Q2000) from 25  C to 170  C at a rate of 10  C min1. Enthalpy changes and degree of conversions of rDA reaction in the GCPNp-DAs were calculated from the following equations:

DHtotal ¼ Wf X DHrDA

(1)

Degree of conversion ðDC; %Þ ¼

DHexp X 100 DHtotal

(2)

where, DHexp, DHtotal, DHrDA, and Wf indicate the enthalpy changes of rDA reaction in GCPNp-DAs measured by DSC, the calculated total enthalpy changes of rDA reaction in GCPNp-DAs, the calculated total enthalpy change of rDA reaction in DA adduct, and the weight fraction of DA adduct in GCPNp-DAs, respectively. The DHrDA value is known to be approximately 18 kcal/mol (317 J/g) [48]. The recombination ratio of maleimides and furan groups after the thermal rDA reaction was calculated from the following equation:

Recombination ratio ¼

DHexp; nþ1 DHexp; n

(3)

where, DHexp; n and DHexp; nþ1 indicate that the DHexp value in n-th cycle and in (nþ1)th cycle respectively. The changes in the complex viscosity of the GCPNp series at DA and rDA temperatures were characterized with an oscillatory rheometer (Thermo Scientific Inc., MARS III) operated at a constant frequency of 1 Hz, with a strain of 0.1%, 8 mm parallel plate, and axial force of 5 N. The heating and cooling rate was 10  C min1. For multiple cycles of FT-IR, DSC, and rheology measurements, the de-crosslinked samples were kept in isothermal conditions at the DA temperature for 1 h to regenerate the network structure of the polymer films. The initial cycle of each measurement was ignored in order to remove possible influence of residual solvent evaporation and chain reorientation effects of the polymer samples.

2.3.4. Thermal and viscoelastic properties The thermal stability levels of the GCPp-OH and GCPNp series were determined by carrying out thermogravimetric analysis (TGA) (TA Instruments, TGAQ500) from 25  C to 600  C at a rate of 10  C min1 under a nitrogen atmosphere. The thermal transitions of the GCPp-OHs and GCPNp polymers were determined from the heating ramp between 0 and 170  C at a rate of 10  C min1 using a DSC (TA Instruments, DSC Q2000) under a nitrogen atmosphere. The viscoelastic properties of the GCPNp polymers were evaluated by taking a dynamic mechanical analysis (DMA) measurements (TA Instruments, DMA Q800) with a strain of 0.01% and constant frequency of 1 Hz. The data were recorded by heating the samples at 5  C min1 from 40 to 150  C. The dimensions of the specimens were 60 mm  12 mm x 3 mm. The vc values and the loss factors (tan d ¼ G”/G0 ) of the GCPNp series were calculated from the tensile storage modulus (G0 ) and loss modulus (G00 ) data. The Tg at the surface of the GCPp-OHs and GCPNp series was measured using a rigid-body pendulum type physical properties testing instrument (RPT) (A&D Co., RPT-3000w) with a cylindrical edge type pendulum (Fig. S1). In the RPT experiment, the test piece coating plate was fixed on the cooling/heating block and was in contact with a cylinder via a knife. A free vibration was applied to the pendulum as the temperature was increased, and the rate and period (damping) of the motion were recorded. The Tg at the surface of the polymeric coating was defined as the point at which the speed of the

138

S.Y. Kim et al. / Polymer 128 (2017) 135e146

pendulum increased since the segmental motion of the polymer chain at this point would start to interfere with the motion of the pendulum. RPT measurements were taken by gradually increasing the temperature from 40 to 160  C at a rate of 10  C min1 [49,50]. 2.3.5. Nanoindentation Loading, holding, and unloading indentation measurements were taken using a nanoindentation tester (Anton-Paar, NHT3) by employing a Berkovich-type indenter. During the loading step, the force imparted by the indenter was gradually increased from 0 mN to 10 mN at a rate of 20 mN min1. Then, in the holding step, the force was maintained at 10 mN for 10 s. And finally, the indenter was unloaded at the same absolute value of the rate as that of the loading. Through these three steps, the maximum displacement (hmax), permanent depth of penetration (final depth, hf), elastic unloading stiffness (S ¼ dP/dh), HIT, and indentation modulus (EIT) were obtained (Figs. S2 and S3) [51]. After obtaining the indentation properties, the coatings were subjected to the rDA temperature for 1 h using the Feltier module and then annealed at the DA temperature for 12 h. The indentation properties were measured again after completion of the crosslinking process. 2.3.6. Characterization of the scratch-resistance and scratchhealing performances Micro-scratch tests were performed using a Rockwell C indenter (tip radius 10 mm) mounted on a scratch test machine (Anton Paar, micro-scratch tester). In a progressive load mode, the test was conducted using three stages; pre-scan, scratching and post-scan. First, in the pre-scan stage, the indenter was applied onto the surface with a low load of 0.4 mN to record the surface profile. Then, during scratching, a progressively increasing normal force from 0.4 mN up to 1000 mN at a constant rate of about 7980 mN min1 was applied to the GCPNp-DA coating surface, and scratches with lengths of 2 mm each were made at a rate of 4 mm min1. As an alternative, the force was also increased stepwise to 250, 500, 750, and 1000 mN on the coating to initiate formation of a fracture on the coating surface. The indenter speed and the scratch length here were the same as those used when the load was increased progressively. After recording the scratch position and topography data, the scratch-healing performances of the coatings were evaluated by heating the scratched surfaces using a Feltier module installed in the scratch test machine. Scratch healing was achieved by heating the material at 80  C for 1 h or 130  C for 10 min. After the healing process, the healed scratches (the locations of the original scratches) were analyzed using an optical microscope (OM) and atomic force microscope (AFM) attached to the scratch test machine. Nano-scratch tests were performed using a sphero-conical indenter (tip radius 5 mm) mounted on a scratch test machine (Anton Paar, nano-scratch tester). The force was increased stepwise to 0.40, 20.4, 40.4, and 60.4 mN on the coating to initiate formation of a fracture on the coating surface. The indenter speed and the scratch length here were 1.6 mm min1 and 0.8 mm respectively. Scratch healing was achieved by heating the material at 50  C or 70  C or 150  C for 1 h. After the healing process, the healed scratches (the locations of the original scratches) were analyzed using an OM and AFM attached to the scratch test machine. 3. Results and discussions 3.1. Material design and synthesis In this study, we designed poly (methyl methacrylate)-co[poly (methyl metharyleate)-graft-(oligo-caprolactone)] urethane

networks containing a DA adduct unit (GCPNp-DAs) in an effort to make scratch-healing coating materials that display thermally reversible crosslinking. Three types of polyacrylate-graft-polyurethane network, each having a different bristle length and denoted as GCPN1-DA, GCPN2-DA and GCPN4-DA, were prepared to control the thermal and viscoelastic properties of the coating materials. As a control sample, conventional crosslinked poly (urethane acrylate)s denoted as GCPN1-H, GCPN2-H and GCPN4-H were also synthesized and their thermal and viscoelastic properties were compared with the corresponding GCPNp-DA at DA and rDA temperatures. The detailed synthesis procedure is shown in Scheme 1. The DAA was synthesized in four steps: a DA reaction of furan with maleic anhydride (1), imidization of 1, rDA reaction of 2, and DA reaction of furfuryl alcohol with 3. The chemical structures of the materials synthesized at each step were confirmed using 1H NMR (Figs. S4(a)-(d)). After the synthesis of DAA, the DADI was prepared by reacting DAA with two equivalents of IPDI in the presence of a DBTDL catalyst. The completion of the reaction was confirmed by the observation of the characteristic ratio of the intensity of the methylene unit (-CH2-) FT-IR band at 2950e2850 cm1 to that of the isocyanate unit (eNCO) band at 2250 cm1 (Fig. S5). In a similar manner, the DI was prepared by reacting 1,6-hexanediol with two equivalents of IPDI. The GCPp-OHs were synthesized via conventional radical polymerization. The obtained GCP-OHs were characterized using 1 H NMR spectroscopy (Fig. S6). The relative amount of hydroxylterminated bristles in GCPp-OHs was calculated from the characteristic ratio of the intensity of the peak corresponding to the eOCH3 group in MMA to that of the eCH2- group in the oligocaprolactone bristle. The M n and ÐM of the GCPp-OHs were determined using a SEC to be in the range of 3.7e4.1 kg/mol and 2.15e2.53, respectively (Fig. S7). Detailed information about the polymers is provided in Table S1. 3.2. Thermally reversible crosslinking of polymer networks The reversibility of the de-crosslinking and crosslinking of the GCPNp-DAs was confirmed using FT-IR spectroscopy, DSC and oscillatory rheology at the DA and rDA reaction temperatures (Fig. 1). The occurrence of repeated DA and rDA reactions was confirmed by tracking the ratio of the intensity of the maleimide double bond absorbance peak at 654 cm1 to that of the carbonyl group stretch absorbance peak at 1720 cm1. As shown in Fig. 1(a), the maleimide double bond peak appeared after the de-crosslinking at the rDA reaction temperature (130  C) and completely disappeared after the crosslinking process at the DA reaction temperature (70  C), confirming the occurrence of the reversible DA/rDA reactions [52]. Enthalpy changes in the DA and rDA reactions during multiple heating and cooling cycles were determined using DSC (Fig. 1(b)). In the repeated heating runs, a broad endothermic transition (DHexp) was detected in the range 100e170  C, which corresponded to the breaking of the DA bonds. The starting point, end point and maximum of the DHexp peak were observed at 100  C, 170  C, and 150  C, respectively. The DHexp values of GCPNp-DAs for three heating and cooling cycles are listed in Table 1. The degree of conversion of thermal rDA reaction was 54e59% on average after first heating cycle. These DHexp values were slightly lower for the later cycles, which indicated that the reaction yield of the DA reaction was not 100% under this condition. However, all GCPNp-DAs were found to have been converted to network structures with high recombination ratio of at least 0.90 per cycle (Table S3) [53]. Fig. 1(c) shows the complex viscosity changes of the GCPNp-DAs resulting from cycling between the DA and rDA reaction

S.Y. Kim et al. / Polymer 128 (2017) 135e146

139

Scheme 1. Procedures used to synthesize the hydroxyl terminated graft copolymers (GCPp-OHs), diisocyanate crosslinkers (DADI and DI), and graft copolymer urethane networks (GCPNp-DAs and GCPNp-Hs).

temperatures five times. As the temperature was increased to 150  C, the complex viscosity values of the polymers decreased rapidly due to destruction of the polymer network structure, while upon cooling to 70  C, the complex viscosity values of the polymers returned to their respective original values, indicating that the polymer network in each case was completely recovered. In addition, the complex viscosity levels of GCPNp-DAs decreased as the length of the repeating unit of the bristle was increased. This result was attributed to the different crosslinking densities of the different GCPNp-DAs. As shown in Fig. 1(d), the complex viscosity levels of the GCPNp-Hs during the heating and cooling cycles showed patterns similar to those shown by the GCPNp-DAs, but the differences between the highest and lowest values of the GCPNpHs were smaller than those of the corresponding GCPNp-DAs. These smaller differences were attributed mainly to the absence of a disintegration of the polymer network of the GCPNp-Hs at the DA reaction temperature [54,55]. 3.3. Thermal and viscoelastic properties of polymers The thermal stabilities of the GCPp-OH and GCPNp series were investigated using TGA. All polymers were found to be stable up to 180  C (Fig. S8). The dependence of the viscoelastic properties of

the polymers on temperature was determined with DMA (Fig. 2). As shown in the G0 vs temperature plots (Fig. 2(a) and (b)), the G0 values of the GCPp-DA series in the glassy state decreased proportionally with the increase in the bristle length of the repeating unit (G0 of GCPN1-DA > G0 of GCPN2-DA > G0 of GCPN4-DA). The G0 values of the GCPNp-Hs in the glassy state were stable, but as the temperature was increased, a steep drop in each modulus of approximately four orders of magnitude was observed, which indicated that the GCPNp-H samples were in a rubbery state. However, similar to the results of the rheology experiment, the G0 values of the GCPp-DA series continued to decrease as the temperature was further increased, which indicated that the network structure of the polymer disintegrated due to the rDA reaction. Plots of the loss factor vs temperature for the GCPNp-DAs and GCPNp-Hs are shown in Fig. 2(a) and (b), respectively. The temperature at which the maximum tan d occurs is the temperature of a relaxation of the network and is closely related to Tg of the GCPNp polymers. The Tg values of the GCPNp series were thus indicated to have decreased with increasing bristle length of the repeating unit (Tg of GCPN1 > Tg of GCPN2 > Tg of GCPN4). The thermal properties of the GCP polymers are summarized in Table S2. The Tg values of the GCPNp series determined using RPT were slightly lower on average than those of the GCPNp series determined using DSC and

140

S.Y. Kim et al. / Polymer 128 (2017) 135e146

Fig. 1. Reversible crosslinking and de-crosslinking of the GCPNp series at DA and rDA temperatures. (a) FT-IR spectra of the GCPN1-DA sample. (b) DSC integrals for the first, second, and third cycles for the GCPN1-DA sample. Complex viscosity values of the (c) GCPNp-DA and (d) GCPNp-Hs in response to repeated heating and cooling cycles.

Table 1 DSC integrals for the first, second, and third cycles for the GCPNp-DA polymers. Polymer code

GCPN1-DA GCPN2-DA GCPN4-DA

DHtotal (J/g)

42 36 26

DHexp (J/g) Cycle 1

Cycle 2

Cycle 3

23 20 15

21 18 14

20 16 13

DMA (Figs. S9 and S10). Note that the Tg of a polymer at its interface with air is generally lower than that of the polymer chain in bulk because the rotation of the polymer chain at the interface is less restricted. Tg measured using RPT reflects the Tg at the coating surface rather than in bulk and is hence more suitable for understanding scratch-resistance and healing performances of the coating. The vc of a cured polymer is defined as the number of moles of network chains per unit volume. The vc value of a crosslinked thermoset can be determined by taking modulus measurements in

Fig. 2. Viscoelastic properties of films of the GCP series: Variations in G0 of GCPNp-DAs (a) and GCPNp-Hs (b) with temperature (solid lines), and plots of the calculated loss factors of GCPNp-DAs (a) and GCPNp-Hs (b) versus temperature (dotted lines) obtained with DMA measurements.

S.Y. Kim et al. / Polymer 128 (2017) 135e146

the rubbery plateau and using the equation of state for rubber elasticity (eq. (4)) [56].

vc ¼ G

0

. 3RT

(4)

where G0 is tensile storage modulus in the rubbery plateau, T is the temperature in K corresponding to the storage modulus value, and R is the gas constant. The vc values of the GCPN1-H, GCPN2-H, and GCPN4-H were calculated to be 200.3, 103.3, and 77.28 mmol L1, respectively, and hence showed a decrease with increasing bristle length of the polymers. The vc values of the GCPNp-DAs should be similar to those of the GCPNp-Hs since the chemical structures and functionalities of the polymer films are similar. 3.4. Nanoindentation test of GCPNp coatings Fig. 3 shows typical loadedisplacement curves for nanoindentation tests carried out on the GCPNp coatings. All GCPNp coatings exhibited plastic deformation in response to applied normal forces. As the bristle length of the polymer was increased, the hmax and hf values increased, while dP/dh in the unloading step decreased, with this decrease attributed to the lower vc values. Similarly, the calculated HIT and EIT values of the GCPNp coatings also increased as the bristle length of the polymer was decreased (Table S4). Compared to the GCPNp-H coatings, the corresponding GCPNp-DA coatings exhibited 2e3 times higher HIT and EIT values. The cyclic structure of the DA adduct unit in the GCPNp-DAs apparently restricted the chain movement more efficiently than

141

did the linear hexamethylene moiety in the GCPNp-Hs. The tendency was quite consistent with that derived from the rheology and DMA measurements. The recovery of the mechanical properties of the de-crosslinked GCPNp-DA coatings after the DA reaction process was also investigated by carrying out sequential de-crosslinking and crosslinking. As shown in the curves of Fig. 3(b)-(d), the HIT and EIT values of the re-crosslinked coatings were similar to those of the corresponding original samples, indicating that the coatings successfully recovered their network structure after the DA process. 3.5. Scratch-resistance and healing performances of GCPNp-DA coatings in response to a progressively increasing deforming load The influence of the scratch load on the breakage response of the GCPNp-DA coatings was determined using the scratch test method, in which a progressively increasing deforming load from 0.4 to 1000 mN was exerted to initiate the fracture and plastic deformation of the GCPNp-DA surface. The scratch characteristics were quantified in terms of the load at the fracture point (Lc1), and the width, penetration depth (Pd), recovery depth (Rd) and pile up of the scratch. In general, the load at Lc1 is closely related to the intrinsic resistance of coatings to being scratched. OM images of the scratched and healed surfaces of the GCPNp-DAs coatings (Fig. 4) revealed increasing loads at Lc1 of these coating surfaces with decreasing bristle length of the polymer, which we attributed to the increased hardness and elastic modulus. The scratch depth and the percentage of recovery can together

Fig. 3. Load-displacement curves of GCPNp coatings: (a) GCPNp-H coatings, (b) GCPN1-DA, (c) GCPN2-DA, and (d) GCPN4-DA coatings. Dotted and solid lines indicate the curves before and after the rDA/DA process, respectively.

142

S.Y. Kim et al. / Polymer 128 (2017) 135e146

Fig. 5. Scratch-resistance profiles of the GCPNp-DA coatings: Penetration depths and recovered depths were measured as a function of lateral position of a line of the coating subjected to (a) progressively increasing loads from 0.4 mN to 1000 mN, and (b) a step-wise increase in deformation-producing loads from 250 mN to 1000 mN.

Fig. 4. OM images of scratched and healed surfaces of the GCPNp-DAs of coatings. The healing process was performed at 80  C for 1 h and 130  C for 10 min. White dashed line indicates the Lc1 point.

be a good indicator for assessing the resistance of organic coatings to being scratched. For instance, a deep penetration and high residual depth would indicate the coating to be in a fracture deformation mode rather than plastic deformation mode. As clearly seen in the load vs penetration plot for the GCPNp-DA coatings (Fig. 5(a)), the penetration depth of the GCPNp-DA coating increased with increasing bristle length of the polymer (Pd of GCPN1-DA < Pd of GCPN2-DA < Pd of GCPN4-DA), which we attributed to their lower HIT values. In contrast, the percentage of recovery (%rec) of the GCPNp-DA coating was inversely proportional to the bristle length of the graft polymers (%rec of GCPN1DA < %rec of GCPN2-DA < %rec of GCPN4-DA). The flexible long oligo-caprolactone bristles apparently facilitated the movement of the polymer chains at the scratched area more efficiently than did the short bristles. The thermally reversible scratch-healing performances of the GCPNp-DA coatings were quantitatively measured based on the healing of the main trace of the scratch (Fig. 4). The scratched GCPNp-DA coatings were subjected to the target temperature for 1 h using a Peltier heating module of the scratch indentation machine. After carrying out the healing process at 80  C, the area of the main scratch for each of the GCPNp-DA coatings was obviously reduced. The scratch-healing performance was also found to improve with increasing bristle length of the polymer (GCPN1-

DA < GCPN2-DA < GCPN4-DA). This result was mainly attributed to the different Tg values of the GCPNp-DA coatings. The polymer chain mobility levels of the GCPN2-DA and GCPN4-DA coatings were much improved, compared to that of the GCPN1-DA coating, since their Tg values were near the process temperature while the Tg of the GCPN1-DA polymer was above the process temperature. When the scratched GCPN1-DA coatings were subjected to a temperature of 130  C, their scratch-healing performances were greatly improved compared to those resulting from the treatment at 80  C, and the remaining scratched groove seen at 80  C completely disappeared (Fig. 4). We attributed this observation to a much greater polymer chain mobility at 130  C than at 80  C resulting from disintegration of the network structure as well as the Tg. 3.6. Scratch-resistance and healing performances of GCPNp-DA coatings in response to a step-wise-increasing deforming load The scratch-resistance and healing performances of the GCPNpDA coatings in response to a step-wise increase in the deforming load was investigated using the scratch test machine to exert four sequentially increasing loads from 250 to 1000 mN and initiate formation of the fracture on the coating surface (Fig. 6). Interestingly, all of the GCPNp-DA coatings showed much inferior resistance to scratches produced in the step-wise manner than to those produced by the progressive increase in load. Both the GCPN1-DA and GCPN2-DA coatings showed smooth scratch paths below 500 mN, but tears and large ruptures in the scratched groove above 500 mN. The surface of the GCPN4-DA coating exhibited scratch resistance much inferior to those of the GCPN1-DA and GCPN2-DA coatings.

S.Y. Kim et al. / Polymer 128 (2017) 135e146

143

between 0.4 and 0.8 mm, and the applied force to create the scratches on the coatings was 500 mN. Scratch-depth-based healing efficiency (%DSHE) values of the GCPNp-DA coatings were calculated from the equations

Ef ¼

ðPd  Rd Þ R ¼ 1  d ¼ 1  Pf Pd Pd

  %DSHE ¼ 1  Pf  100

(5)

(6)

where Rd, Pd, Ef and Pf indicate the residual depth, penetration depth, elastic factor, and plastic factor, respectively. Similarly, scratch-width-based healing efficiency (%WSHE) values of the GCPNp-DA coatings were derived using the equation

 %WSHE ¼

Fig. 6. OM images of scratched and healed surface of the GCPNp-DAs of coatings. Here, the healing process was performed at 130  C for 10 min.

That the scratches of the GCPNp-DA coatings showed different features for the different ways the load was increased can be explained by different energy-dissipation modes of the viscoelastic materials in these two cases. When a load is increased gradually, as it was in our progressive mode, the polymeric coatings can have enough time to dissipate the applied stress. In this case, a smooth scratch path would be generated by plastic deformation if the magnitude of the applied force exceeds the elasticity of the polymeric materials. However, when the load is increased in discrete steps, a large increase in force is applied on the polymeric coating surface at once. In this case, the energy dissipation would mainly result in the destruction of the surface rather than formation of a scratch path. Fig. 5(b) shows load vs penetration plots for the GCPNp-DA coatings when the load was increased as function of lateral distance in a step-wise manner. The GCPN1-DA and GCPN2-DA coatings mainly showed plastic deformations as expected but the GCPN4-DA coating showed a non-uniform fracture according to the depth profile pattern. The scratch-healing process operated at 80  C resulted in efficient reduction of the areas of the main scratches of both the GCPN1-DA and GCPN2-DA coatings but the GCPN2-DA coating showed a better healing performance than did the GCPN1-DA coating (Fig. S11). GCPN4-DA showed a poor scratchhealing ability despite it having the lowest Tg, mainly because of the very large rupture in the scratched groove. The scratch-healing performances of the GCPNp-DA coatings were greatly improved at 130  C (Fig. 6), which was mainly attributed to the rupture of the network structure of the GCPNp-DA polymers. The GCPN1-DA coating in particular showed a dramatic improvement of the scratch-healing performance compared to the improvements displayed by the other GCPNp-DA coatings due to its increased polymer chain mobility at the temperature above its Tg. However, the very large ruptured region of each GCPNp-DA coating was not healed even at the rDA temperature. Based on the results of these experiments, the GCPNp-DA coatings (except GCPN4-DA) were able to eliminate a scratch resulting from an applied load of as high as 500 mN using the system displaying thermally reversible crosslinking. The depth profiles of scratched and healed surfaces of the GCPNp-DA coatings were acquired using an AFM attached to a scratch indentation machine (Fig. 7). The scanned positions were

1

Wr Wi

  100

(7)

where Wi and Wr indicate, respectively, the initial scratch width and residual scratch width after the healing process. The calculated values are listed in Table S5 and Table 2. The %DSHE and %WSHE values of both the GCPN1-DA and GCPN2-DA coatings increased as the scratch-healing temperature was increased. However, the %SHEs of GCPN4-DA coating could not be determined because the main scratch path was essentially destroyed. The ratio of Ef to Pf at a specified temperature can be used as a measure of the scratch-healing performance of polymeric coating materials since the value is closely related to polymer chain mobility. We observed the main scratch trace of the GCPNp-DA coatings to be completely eliminated when the ratio exceeded approximately 9. In the weak scratch generation and healing experiment using a nano-scratch tester, all GCPNp-DA coatings showed excellent scratch-healing performances, especially at the rDA temperature (Fig. S12eS15). 4. Conclusions Scratch-healing GCPNp-DAs were successfully synthesized. They were coated on steel substrates to investigate the influence of their material properties such as elastic modulus, HIT, and thermal transition temperature on their scratch-healing performance. The maleimide FT-IR peak at 654 cm1 appeared after decrosslinking at the rDA reaction temperature and completely disappeared again after crosslinking at the DA reaction temperature. In the multiple cycles of DSC measurements, a repeated evolution of heat was detected during the heating runs, indicating a disintegration of the crosslinked structure of GCPNp-DA polymers. The complex viscosity levels of these polymers were observed in the rheology experiments to decrease rapidly as the temperature was increased to 150  C, and then to return to their original values after cooling them to 70  C in each of the five heating and cooling cycles. The GCPNp-H polymers showed patterns of complex viscosity values similar to those of the GCPNp-DA polymers, but with smaller differences between the highest and lowest values in response to the changes in temperature, due to the absence of a rupture of the polymer network at the DA reaction temperature. In the NST experiment, mechanical properties of GCPNp-DA coatings were maintained after de-crosslinking and crosslinking. All of above results provided good support for the occurrence of the reversible formation of crosslinked and de-crosslinked GCPNp-DA structures. In response to a progressively increasing deforming load, the graft copolymer with the short bristles exhibited higher resistance to being scratched than did the graft copolymer with the long

144

S.Y. Kim et al. / Polymer 128 (2017) 135e146

Fig. 7. AFM images of scratched and healed surfaces of the GCPNp-DA coatings. The scanned positions and applied load were 0.4e0.8 mm and 500 mN, respectively.

Table 2 Ef, Pf, and %DSHE values of the GCPNp-DA coatings. Polymer code

80  C

130  C

Ef

Pf

%DSHE

Ef

Pf

%DSHE

GCPN1-DA GCPN2-DA GCPN4-DA

0.57 0.90 e

0.43 0.10 e

56.8 89.8 e

0.90 0.93 e

0.10 0.07 e

90.0 92.8 e

surfaces did not fully heal even when treated at a temperature above their Tg and de-crosslinking temperatures. Taken together, our results revealed that both the scratchresistance and healing performances of the GCPNp-DA coatings were greatly influenced by their material properties as well as by whether the scratch was made with a progressively increasing or step-wise-increasing deforming load. Acknowledgements

bristles because of their higher vc, elastic modulus, HIT and Tg values. However, scratches that did form on the coatings containing the graft polymers with shorter bristles did not heal when treated at 80  C unlike did those with the longer bristles. These results were mainly attributed to the different Tg values and hence different chain mobility levels of the graft polymers with different bristle lengths. Note, however, that all graft polymers exhibited excellent scratch-healing performance at 130  C. The graft polymers subjected to a step-wise increase in the deforming load showed a much worse scratch resistance than did those subjected to the progressively increasing load. The hard GCPN1-DA and GCPN2-DA coatings showed smooth scratch paths when subjected to loads of up to 500 mN, while the soft GCPN4-DA coating showed a very large rupture in the scratch groove. This result was attributed mainly to the different energy dissipation modes related to the magnitude of the applied stress. The energy added to the polymeric coatings was mainly dissipated by plastic deformation when the deforming load was progressively increased, but by the formation of a surface rupture when the load was increased step-wise. The very large ruptures in the scratched grooves on the GCPNp-DA coating surfaces were found to be disadvantageous relative to the smooth scratch paths: the ruptured

The authors are very grateful for financial support from the Ministry of Trade, Industry & Energy of Korea (Industrial Technology Innovation Program No.10068082). Technical support (scratch tester) from Anton Paar is gratefully acknowledged. Additionally, the authors thank Mr. Gregory FAVARO, Ms. Mihaela Dubuisson, Wonkeun Park and Jungwoong Ma for their efforts and valuable comments related to the scratch-healing experiments. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.polymer.2017.09.021. References [1] M. Gerard, A. Chaubey, B.D. Malhotra, Application of conducting polymers to biosensors, Biosens. Bioelectron. 17 (2002) 345e359, http://dx.doi.org/ 10.1016/S0956-5663(01)00312-8. [2] Z.-M. Huang, Y.-Z. Zhang, M. Kotaki, S. Ramakrishna, A review on polymer nanofibers by electrospinning and their applications in nanocomposites, Compos. Sci. Technol. 63 (2003) 2223e2253, http://dx.doi.org/10.1016/S02663538(03)00178-7.
mann, F. De Jaeghere, E. Doelker, R. Gurny, Biodegradable [3] J.-C. Leroux, E. Alle

S.Y. Kim et al. / Polymer 128 (2017) 135e146

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17] [18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

nanoparticles d from sustained release formulations to improved site specific drug delivery, J. Control. Release 39 (1996) 339e350, http://dx.doi.org/ 10.1016/0168-3659(95)00164-6. H. Jiang, R. Browning, J.D. Whitcomb, M. Ito, M. Shimouse, T.A. Chang, H.J. Sue, Mechanical modeling of scratch behavior of polymeric coatings on hard and soft substrates, Tribol. Lett. 37 (2010) 159e167, http://dx.doi.org/10.1007/ s11249-009-9505-8. €pferich, Polymer coating of quantum dots e a A.F.E. Hezinger, J. Teßmar, A. Go powerful tool toward diagnostics and sensorics, Eur. J. Pharm. Biopharm. 68 (2008) 138e152, http://dx.doi.org/10.1016/j.ejpb.2007.05.013. J. Jung, J.C. Kim, Y. Rho, M. Kim, W. Kwon, H. Kim, M. Ree, Molecular layer-bylayer self-assembly and mercury sensing characteristics of novel brush polymers bearing thymine moieties, ACS Appl. Mater. Interfaces 3 (2011) 2655e2664, http://dx.doi.org/10.1021/am200454x. J.C. Kim, J. Jung, Y. Rho, M. Kim, W. Kwon, H. Kim, I.J. Kim, J.R. Kim, M. Ree, Well-defined DNA-mimic brush polymers bearing adenine moieties: synthesis, layer-by-layer self-assembly, and biocompatibility, Biomacromolecules 12 (2011) 2822e2833, http://dx.doi.org/10.1021/bm200572t. J.C. Kim, Y. Rho, G. Kim, M. Kim, H. Kim, I.J. Kim, J.R. Kim, M. Ree, New selfassembled brush glycopolymers: synthesis, structure and properties, Polym. Chem. 4 (2013) 2260e2271, http://dx.doi.org/10.1039/C3PY21147A. S.K. Ghosh, Self-healing materials: fundamentals, design strategies, and applications, in: S.K. Ghosh (Ed.), Self-Heal. Mater, Wiley-VCH Verlag GmbH & Co. KGaA, 2008, pp. 1e28, http://dx.doi.org/10.1002/9783527625376.ch1. B.J. Blaiszik, S.L.B. Kramer, S.C. Olugebefola, J.S. Moore, N.R. Sottos, S.R. White, Self-healing polymers and composites, Annu. Rev. Mater. Res. 40 (2010) 179e211, http://dx.doi.org/10.1146/annurev-matsci-070909-104532. S.Y. An, D. Arunbabu, S.M. Noh, Y.K. Song, J.K. Oh, Recent strategies to develop self-healable crosslinked polymeric networks, Chem. Commun. 51 (2015) 13058e13070, http://dx.doi.org/10.1039/C5CC04531B. H. Yari, M. Mohseni, M. Messori, A scratch resistant yet healable automotive clearcoat containing hyperbranched polymer and POSS nanostructures, RSC Adv. 6 (2016) 76028e76041, http://dx.doi.org/10.1039/C6RA07824A. S.R. White, N.R. Sottos, P.H. Geubelle, J.S. Moore, M.R. Kessler, S.R. Sriram, E.N. Brown, S. Viswanathan, Autonomic healing of polymer composites, Nature 409 (2001) 794e797, http://dx.doi.org/10.1038/35057232. K. Wazarkar, D. Patil, A. Rane, D. Balgude, M. Kathalewar, A. Sabnis, Microencapsulation: an emerging technique in the modern coating industry, RSC Adv. 6 (2016) 106964e106979, http://dx.doi.org/10.1039/C6RA13237E. H. Wei, Y. Wang, J. Guo, N.Z. Shen, D. Jiang, X. Zhang, X. Yan, J. Zhu, Q. Wang, L. Shao, H. Lin, S. Wei, Z. Guo, Advanced micro/nanocapsules for self-healing smart anticorrosion coatings, J. Mater. Chem. A 3 (2015) 469e480, http:// dx.doi.org/10.1039/C4TA04791E. D.Y. Zhu, M.Z. Rong, M.Q. Zhang, Self-healing polymeric materials based on microencapsulated healing agents: from design to preparation, Prog. Polym. Sci. 49 (2015) 175e220, http://dx.doi.org/10.1016/j.progpolymsci.2015.07. 002. S.H. Cho, S.R. White, P.V. Braun, Self-healing polymer coatings, Adv. Mater 21 (2009) 645e649, http://dx.doi.org/10.1002/adma.200802008. Y.-K. Song, Y.-H. Jo, Y.-J. Lim, S.-Y. Cho, H.-C. Yu, B.-C. Ryu, S.-I. Lee, C.M. Chung, Sunlight-induced self-healing of a microcapsule-type protective coating, ACS Appl. Mater. Interfaces 5 (2013) 1378e1384, http://dx.doi.org/ 10.1021/am302728m. S.M. Mirabedini, I. Dutil, L. Gauquelin, N. Yan, R.R. Farnood, Preparation of selfhealing acrylic latex coatings using novel oil-filled ethyl cellulose microcapsules, Prog. Org. Coat. 85 (2015) 168e177, http://dx.doi.org/10.1016/ j.porgcoat.2015.03.024. H. Es-Haghi, S.M. Mirabedini, M. Imani, R.R. Farnood, Mechanical and selfhealing properties of a water-based acrylic latex containing linseed oil filled microcapsules: effect of pre-silanization of microcapsules' shell compound, Compos. Part B Eng. 85 (2016) 305e314, http://dx.doi.org/10.1016/ j.compositesb.2015.09.015. S.J. Garcia, Effect of polymer architecture on the intrinsic self-healing character of polymers, Eur. Polym. J. 53 (2014) 118e125, http://dx.doi.org/ 10.1016/j.eurpolymj.2014.01.026. M.Q. Zhang, M.Z. Rong, Intrinsic self-healing of covalent polymers through bond reconnection towards strength restoration, Polym. Chem. 4 (2013) 4878e4884, http://dx.doi.org/10.1039/C3PY00005B. J. Bai, H. Li, Z. Shi, J. Yin, An eco-friendly Scheme for the cross-linked polybutadiene elastomer via thioleene and dielsealder click chemistry, Macromolecules 48 (2015) 3539e3546, http://dx.doi.org/10.1021/acs.macromol. 5b00389.
, A. Gandini, P. Hodge, Application of the DielsAlder reaction to C. Gousse polymers bearing furan moieties. 2. DielsAlder and retro-diels-alder reactions involving furan rings in some styrene copolymers, Macromolecules 31 (1998) 314e321, http://dx.doi.org/10.1021/ma9710141. Y. Chujo, K. Sada, T. Saegusa, Reversible gelation of polyoxazoline by means of Diels-Alder reaction, Macromolecules 23 (1990) 2636e2641, http:// dx.doi.org/10.1021/ma00212a007. Y.-L. Liu, T.-W. Chuo, Self-healing polymers based on thermally reversible DielseAlder chemistry, Polym. Chem. 4 (2013) 2194e2205, http://dx.doi.org/ 10.1039/C2PY20957H. S.A. Canary, M.P. Stevens, Thermally reversible crosslinking of polystyrene via the furanemaleimide DielseAlder reaction, J. Polym. Sci. Part Polym. Chem. 30 (1992) 1755e1760, http://dx.doi.org/10.1002/pola.1992.080300829.

145

[28] N. Roy, B. Bruchmann, J.-M. Lehn, DYNAMERS: dynamic polymers as selfhealing materials, Chem. Soc. Rev. 44 (2015) 3786e3807, http://dx.doi.org/ 10.1039/C5CS00194C. [29] H. Ying, J. Cheng, Hydrolyzable polyureas bearing hindered urea bonds, J. Am. Chem. Soc. 136 (2014) 16974e16977, http://dx.doi.org/10.1021/ja5093437. [30] Y. Zhang, H. Ying, K.R. Hart, Y. Wu, A.J. Hsu, A.M. Coppola, T.A. Kim, K. Yang, N.R. Sottos, S.R. White, J. Cheng, Malleable and recyclable poly(urea-urethane) thermosets bearing hindered urea bonds, Adv. Mater 28 (2016) 7646e7651, http://dx.doi.org/10.1002/adma.201601242. [31] Z.P. Zhang, M.Z. Rong, M.Q. Zhang, C.’e Yuan, Alkoxyamine with reduced homolysis temperature and its application in repeated autonomous selfhealing of stiff polymers, Polym. Chem. 4 (2013) 4648e4654, http:// dx.doi.org/10.1039/C3PY00679D. [32] C. Yuan, M.Z. Rong, M.Q. Zhang, Z.P. Zhang, Y.C. Yuan, Self-healing of polymers via synchronous covalent bond fission/radical recombination, chem. Mater 23 (2011) 5076e5081, http://dx.doi.org/10.1021/cm202635w. [33] Z. Wei, J. Hai Yang, J. Zhou, F. Xu, M. Zrínyi, P.H. Dussault, Y. Osada, Y. Mei Chen, Self-healing gels based on constitutional dynamic chemistry and their potential applications, Chem. Soc. Rev. 43 (2014) 8114e8131, http:// dx.doi.org/10.1039/C4CS00219A. [34] S.Y. An, S.M. Noh, J.H. Nam, J.K. Oh, Dual sulfideedisulfide crosslinked networks with rapid and room temperature self-healability, Macromol. Rapid Commun. 36 (2015) 1255e1260, http://dx.doi.org/10.1002/marc.201500123. [35] Z.Q. Lei, H.P. Xiang, Y.J. Yuan, M.Z. Rong, M.Q. Zhang, Room-temperature selfhealable and remoldable cross-linked Polymer based on the dynamic exchange of disulfide bonds, Chem. Mater 26 (2014) 2038e2046, http:// dx.doi.org/10.1021/cm4040616. [36] J.A. Yoon, J. Kamada, K. Koynov, J. Mohin, R. Nicolaÿ, Y. Zhang, A.C. Balazs, T. Kowalewski, K. Matyjaszewski, Self-healing polymer films based on thioledisulfide exchange reactions and self-healing kinetics measured using atomic force microscopy, Macromolecules 45 (2012) 142e149, http:// dx.doi.org/10.1021/ma2015134. [37] M. Pepels, I. Filot, B. Klumperman, H. Goossens, Self-healing systems based on disulfide e thiol exchange reactions, Polym. Chem. 4 (2013) 4955e4965, http://dx.doi.org/10.1039/C3PY00087G. [38] L. He, D. Szopinski, Y. Wu, G.A. Luinstra, P. Theato, Toward self-healing hydrogels using one-pot thioleene click and borax-diol chemistry, ACS Macro Lett. 4 (2015) 673e678, http://dx.doi.org/10.1021/acsmacrolett. 5b00336. [39] F. Zeng, Y. Han, Z.-C. Yan, C.-Y. Liu, C.-F. Chen, Supramolecular polymer gel with multi stimuli responsive, self-healing and erasable properties generated by hosteguest interactions, Polymer 54 (2013) 6929e6935, http://dx.doi.org/ 10.1016/j.polymer.2013.10.048. [40] X. Kuang, G. Liu, X. Dong, X. Liu, J. Xu, D. Wang, Facile fabrication of fast recyclable and multiple self-healing epoxy materials through diels-alder adduct cross-linker, J. Polym. Sci. Part Polym. Chem. 53 (2015) 2094e2103, http://dx.doi.org/10.1002/pola.27655. [41] S. Thakur, N. Karak, Tuning of sunlight-induced self-cleaning and self-healing attributes of an elastomeric nanocomposite by judicious compositional variation of the TiO 2ereduced graphene oxide nanohybrid, J. Mater. Chem. A 3 (2015) 12334e12342, http://dx.doi.org/10.1039/C5TA02162F. [42] C.M. Bates, A.B. Chang, M.W. Schulze, N. Mom cilovi
c, S.C. Jones, R.H. Grubbs, Brush polymer ion gels, J. Polym. Sci. Part B Polym. Phys. 54 (2016) 292e300, http://dx.doi.org/10.1002/polb.23927. [43] K. Kim, Y.Y. Kim, S. Park, Y.-G. Ko, Y. Rho, W. Kwon, T.J. Shin, J. Kim, M. Ree, Nanostructure- and orientation-controlled digital memory behaviors of linear-brush diblock copolymers in nanoscale thin films, Macromolecules 47 (2014) 4397e4407, http://dx.doi.org/10.1021/ma500884q. [44] H.-Y. Lee, S.-H. Cha, Enhancement of self-healing property by introducing ethylene glycol group into thermally reversible Diels-Alder reaction based self-healable materials, Macromol. Res. 25 (2017) 640e647, http://dx.doi.org/ 10.1007/s13233-017-5120-y. [45] G. Scheltjens, M.M. Diaz, J. Brancart, G. Van Assche, B. Van Mele, A self-healing polymer network based on reversible covalent bonding, React. Funct. Polym 73 (2013) 413e420, http://dx.doi.org/10.1016/j.reactfunctpolym.2012.06.017. [46] R.J. Sheridan, C.N. Bowman, Understanding the process of healing of thermoreversible covalent adaptable networks, Polym. Chem. 4 (2013) 4974e4979, http://dx.doi.org/10.1039/C2PY20960H. [47] P.A. Pratama, A.M. Peterson, G.R. Palmese, The role of maleimide structure in the healing of furan-functionalized epoxyeamine thermosets, Polym. Chem. 4 (2013) 5000e5006, http://dx.doi.org/10.1039/C3PY00084B. [48] R.C. Boutelle, B.H. Northrop, Substituent effects on the reversibility of furanemaleimide cycloadditions, J. Org. Chem. 76 (2011) 7994e8002, http:// dx.doi.org/10.1021/jo201606z. [49] R. Lu, Y.-Y. Wan, T. Honda, T. Ishimura, Y. Kamiya, T. Miyakoshi, Design and characterization of modified urethane lacquer coating, Prog. Org. Coat. 57 (2006) 215e222, http://dx.doi.org/10.1016/j.porgcoat.2006.08.014. [50] J.-I. Moon, Y.-H. Lee, H.-J. Kim, Synthesis and characterization of elastomeric polyester coatings for automotive pre-coated metal, Prog. Org. Coat. 74 (2012) 125e133, http://dx.doi.org/10.1016/j.porgcoat.2011.11.025. [51] W.C. Oliver, G.M. Pharr, Measurement of hardness and elastic modulus by instrumented indentation: advances in understanding and refinements to methodology, J. Mater. Res. 19 (2004) 3e20, http://dx.doi.org/10.1557/ jmr.2004.19.1.3. €tteritzsch, S.J. Garcia, M.D. Hager, U.S. Schubert, S. van der [52] R.K. Bose, J. Ko

146

S.Y. Kim et al. / Polymer 128 (2017) 135e146

Zwaag, A rheological and spectroscopic study on the kinetics of self-healing in a single-component dielsealder copolymer and its underlying chemical reaction, J. Polym. Sci. Part Polym. Chem. 52 (2014) 1669e1675, http:// dx.doi.org/10.1002/pola.27164. [53] C. Toncelli, D.C. De Reus, F. Picchioni, A.A. Broekhuis, Properties of reversible dielsealder furan/maleimide polymer networks as function of crosslink density, Macromol. Chem. Phys. 213 (2012) 157e165, http://dx.doi.org/ 10.1002/macp.201100405. [54] K.K. Oehlenschlaeger, J.O. Mueller, J. Brandt, S. Hilf, A. Lederer, M. Wilhelm, R. Graf, M.L. Coote, F.G. Schmidt, C. Barner-Kowollik, Adaptable hetero

dielsealder networks for fast self-healing under mild conditions, Adv. Mater 26 (2014) 3561e3566, http://dx.doi.org/10.1002/adma.201306258. [55] D.H. Turkenburg, H.R. Fischer, Diels-Alder based, thermo-reversible crosslinked epoxies for use in self-healing composites, Polymer 79 (2015) 187e194, http://dx.doi.org/10.1016/j.polymer.2015.10.031. [56] T.H. Lee, Y.I. Park, S.M. Noh, J.C. Kim, In-situ visualization of the kinetics of low temperature thiol-epoxy crosslinking reactions by using a pH-responsive epoxy resin, Prog. Org. Coat. 104 (2017) 20e27, http://dx.doi.org/10.1016/ j.porgcoat.2016.11.007.