Study on Microstructure and Properties of the UV Curing Acrylic Epoxy/SiO2 Nanocomposite Coating

This study is aimed at exploring the effects of SiO2 nanoparticles on the crosslinking and mechanical and thermal properties of UV curing acrylic epoxy coating. The curing polymerization process and thermal and mechanical properties of UV-curable acrylate epoxy system have been evaluated with or without the presence of SiO2 nanoparticles. To fabricate the UV curing acrylic epoxy/SiO2 nanocomposite coating, nano-SiO2 particles (0.5–5wt.% by weight of resin) were added in the photocurable system using sonication for 3 h. Various techniques for characterization have been used, such as FESEM (field emission scanning electron microscope), FTIR (Fourier-transform infrared spectroscopy), TGA (thermogravimetry analysis), gel fraction, and swelling degree analyses. FESEM data indicated that at the content of 2.5 wt.%, nanosilica was homogeneously dispersed in the coating procedure. However, once added 5wt.%, large aggregation portions were found inside the coating matrices. Surprisingly, nano-SiO2 could play dual roles, as both UV absorbers and nanoreinforcers, in this nanocomposite coating. Besides, data from FTIR, gel fraction, and swelling degree analyses confirmed the role of SiO2 nanoparticles as UV absorbers that reduced the conversion performance of acrylate double bonds, thus increased slightly the swelling degree of coating. In addition, incorporation of SiO2 nanoparticles (as nanofillers, at content of 2.5 wt.%) in the polymer matrix enhanced significantly the abrasion resistance and thermal stability of the coating, by 60% (from 98.3 to 158.4 lite/mil) and 9C (from 348C to 357C), respectively.


Introduction
As can be processed at room temperature and organic solventfree, acrylate resin-based UV curing paints exhibit many advantages, such as transparency, moisture resistance, chemical resistance, and environmental friendly. As reported, they have been widely applied to various finish surfaces, such as steel or wooden floors [1][2][3][4][5]. However, the main disadvantage of this resin system is the high curing shrinkage (up to 15%) [6,7]. Its shrinkage could be enhanced by three pathways such as the following: (i) by selecting components (with suitable structure), (ii) by increasing the curing density, and (iii) by using fillers [7][8][9][10][11][12].
However, the addition of nanoparticles (as nanofillers) can affect the curing reaction of thermoset resin systems depending on the nature and content of the nanoparticles. In the case of the epoxy matrix, Fe 3 O 4 nanoparticles can act as a bridging link molecule, thereby reducing the total free mass and increasing the crosslinking density [43,44]. In this direction, to study the healing dynamics of the epoxy/amine system, Fe 3 O 4 nanoparticles were used as nanocontainers for loading the acid functional groups [45], amino functional groups [46,47], or hydroxyl functional groups [46,48]. The previous reported publications mentioned that the reaction between acidic groups and amine groups of curing agent can deactivate the hardener [45]. And the existence of hydroxyl and amine groups might boost the epoxy groups. It resulted to an increase in the content of thermal curing [46]. For example, Yari et al. investigated that nanosilica caused to an incomplete curing procedure of the acrylic melamine clearcoat, suggesting to a lower crosslinked density, as compared with the pure clearcoat and improved weathering capacity. It might be assumed that nano-SiO 2 absorbed the harmful incident UV rays, protecting the clearcoat against weathering degradation [49]. In our reported reviews [50], when conducting the curing reaction of acrylic polyols with isocyanate, we found that SiO 2 nanoparticles involved in reaction with isocyanate group to form a tight inorganic-organic hybrid structure. For UV curing systems, the degree of influence of additives depends not only on the content but also on their ability to absorb UV rays and photocatalytic activity. For Fe 3 O 4 -Ag hybrid nanoparticles, despite a strong UV absorption, their low content (0.1 wt.%) might not significantly affect the kinetics of curing reaction [51]. The organic UV absorber T384 reduced the efficiency of double bonds while nanoparticles absorb UV anatas-TiO 2 and ZnO increased the conversion of the acrylate groups of the system due to strong photocatalytic activity [52].
Although polymers/SiO 2 nanocomposites have been reported intensively in the literature, effect of nanoparticles on the kinetics of photo-curing process for UV curing acrylic epoxy resin system has not been well clarified. In this work, the photo-crosslinking polymerization process of the system of acrylate epoxy resin and 1.6 hexanediol diacrylate diluent in the presence of SiO 2 nanoparticles has been evaluated by using various techniques, such as the quantitative infrared analysis, relative hardness, gel fraction, and swelling degree measurements. In addition, the effect of SiO 2 nanoparticles on the mechanical and thermal properties of nanocomposite coating is evaluated.

Preparation of Nanocomposite
Coating. SiO 2 nanoparticles, at the contents from 0.5 to 5 wt.%, were firstly dispersed in HDDA by using a TPC-25 supersonic bath (Switzerland) for 3 h; then, these mixtures were mixed with E284 and I.184 by stirring in the Ika RW16 Basic Mixer (England) for 30 min. The ratio of E284 : HDDA : I.184 was 55 : 45 : 3.
Coatings with the thickness of~25 μm were prepared on KBr pellets for IR spectral analysis, on Teflon sheets with the size of 100 × 100 × 10 mm for UV-Vis spectral, gel fraction, swelling degree, and FESEM analyses, on glass plates with the size of 100 × 100 × 2 mm for relative hardness and abrasion resistance tests by using a Quadruple Film Applicator Model 360. The coating process was applied for an UV radiation of a medium-pressure mercury lamp (250 mW/cm 2 ) at 25°C in an UV device, model F300S, USA. The product was conducted various times under the light with a web rate ranged of 5-40 m/s.

Methods for Characterization of Coatings
2.3.1. IR Spectral Analysis. Changes of IR absorption band at 983 cm -1 (=CH stretching) assigning to acrylate double bonds during the crosslinking reaction were quantitatively studied by using a FTIR spectroscopy NEXUS 670 from Nicolet. Experiments are carried out at the same position of each sample and after various time of UV exposure [51]. Benzene ring at 1510 cm -1 was selected as a band to eliminate the changes in the coating thickness.
The optical density (D) of groups was estimated by the following formula: The relationship between I 0 and I with H and U is shown in Figure 2 where H and U were calculated by using the software of FTIR spectroscopy.
Remaining acrylate group was calculated as follows: where D 0 and D t are the optical densities of groups before and after curing.

Gel Fraction and Swelling Degree Analyses.
Gel fraction and swelling degree analyses of coating product are conducted in Soxhlet tool in accordance with the ASTM (American Society for Testing and Materials) D 2765 [51]. Briefly, the dried coating was immersed in acetone solution for 24 h; then, the insoluble portion was dried at 50°C. The gel fraction is a weight ratio of insoluble portion and the initial coating.
Similarity, the swelling degree is a weight ratio of swollen film and the dried film.

Determination of Mechanical Resistance and Surface
Morphology. The value of coating abrasion resistance was calculated by using abrasive falling methods, according to the ASTM D968 [28].
The surface chemistry was observed in a scanning electron microscope (SEM-S-4800, Hitachi, Japan). To increase the electrical conductivity of products, their surface was covered by a carbon layer.
2.3.4. Thermal Analysis. Thermogravimetry experiments were performed using a well-equipped thermogravimetry analyzer (TGA-50, Shimadzu, Japan). The samples were heated in ranged of 25-600°C, at a rate of 10°C min −1 (under an argon atmosphere with a flow rate of 50 cm 3 min −1 ).

Results and Discussions
3.1. IR Spectra Studies. IR spectra measurement has been used widely to evaluate the crosslinking process of UV-curable acrylate epoxy coating, thought the chemical conversion of acrylate double bonds [51,52]. Figure 3 shows the IR spectra of the neat UV curing acrylate epoxy coating (UVAE) and the nanocomposite coating with 2.5 wt.% nano-SiO 2 (UVAE/SiO 2 ) before and after 9.6 s UV light exposure.
As shown in Figure 3, a decrease in intensity of the characteristic peaks (1636, 1409, 983, and 812 cm -1 ) for =C-H stretching vibrations in acrylate groups was clearly observed after 9.6 s of UV exposure. Besides, the absorbance intensity of 1510 cm -1 band (=C-H stretching of aromatic ring) did not change after UV light exposure [51]. The peaks at 983 and 1510 cm -1 changed clearly but not overlapped with the neighboring bands. Therefore, the peak at 983 cm -1 was chosen to quantitatively evaluate the variation of the acrylate group during UV light irradiation exposure, whereas peaks at 1510 cm -1 can be used as the reference. Conversion of acrylate double bonds in the neat coating (UVAE) and the nanocomposite containing 2.5 wt.% nano-SiO 2 (UVAE/-SiO 2 ) during the UV light exposure process is presented in Figure 4. For the comparative study, we also added the conversion of acrylate double bonds in the acrylate urethane coating (UVAU) [50] in Figure 4.
As can be observed in Figure 4, the acrylate double bonds are conversed rapidly in the first 0.3 s of UV light exposure, and their conversion rate is reduced till 9.6 s of UV light exposure. Conversion of acrylate double bonds in the neat coating was higher than that in the nanocomposite. After 4.8 s of UV exposure, content of acrylate group conversed 87.3 and 81.6% in the neat coating and the nanocomposite, respectively. Further exposure to UV light irradiation could only improve slightly the conversion of acrylate double bonds. After 9.6 s of UV exposure, the content of acrylate groups conversed 89.4 and 85.0% in the neat coating and the nanocomposite, respectively.
It was reported in the literature that the photoinitiator was very sensitive to UV radiation [2,51,52]. Under UV light from a medium-pressure mercury lamp, photoinitiator I.184 was decomposed into the free radicals. Then, these free radicals reacted with acrylate double bonds, thus initiated photo-crosslinking polymerization reaction. Briefly, in the first 0.15 s of UV exposure, the content of photoinitiators was high (around 3%), and the curing system was relatively flexible (low viscosity). Therefore, the reaction of the acrylate double bonds was insignificantly affected by UV absorbers; however, the concentration of the photoinitiators and acrylate double bonds was reduced rapidly; thus, the conversion of the acrylate double bonds decreased.
This finding can be explained by the fact that nano-SiO 2 absorbed the UV light [20,49,50], leading to protect the photoinitiators from photon absorption. On the other hand, the conjugated double bonds of the benzene ring in BGDM that strongly absorb in the 270 nm wavelength region also prevent the UV energy absorption of the photoinitiators, which is a reason why the conversion performance of acrylate groups in this coating system (UV  3 Journal of Nanomaterials curing acrylic epoxy resin) was lower than that in the UV curing acrylate urethane resin [52].

Gel Fraction and Swelling Degree Studies.
Gel fraction and swelling degree were important parameters to evaluate the crosslinking of the thermoset resin. Figure 5 presents the variations of gel fraction and swelling degree of the neat coating and the nanocomposite with 2.5 wt.% nano-SiO 2 during the UV light exposure process. As shown, after 0.3 s of the reaction, the gel fraction of the coatings without and with 2.5 wt.% nano-SiO 2 appeared. The gel fraction increased, and the swelling degree reduced rapidly in the first 2.4 s, and after then, it slowed down. In the presence of 2.5 wt.% nano-SiO 2 , the gel fraction decreased insignificantly, while its swelling degree increased slightly. After 4.8 s of the reaction, the gel fraction and swelling degree reached the maximum values of 95.5 and 365.2 (for the neat coating) and 95.0 and 380.4% (for the nanocomposite), respectively.
Thus, nano-SiO 2 did not affect to the gel fraction but affected slightly on the swelling of the coating. This obtained data could be explained by the fact that UV-absorbing nano-SiO 2 reduced the conversion efficiency of acrylate groups and did not significantly affect on the curing participation of BGDM resin and HDDA diluent but reduced slightly the crosslinking density of the coating so the gel fraction of the coating changed insignificantly, while its swelling increased slightly.

Coating Hardness Study.
For thermoset resin systems, monitoring the hardness of system is also a simple and effective method to study the kinetics of curing reaction because during curing the resin system gradually changes from liquid to solid state and becomes stiffer over time. The changes in relative hardness of the neat coating and the nanocomposite containing 2.5 wt.% nano-SiO 2 during exposure to UV light radiation are presented in Figure 6. As can be seen in Figure 6, the hardness of the coatings increased rapidly in the first 1.2 UV exposure and then slowed down.  Journal of Nanomaterials concentrations of initiator and the acrylate group, as well as the higher coating flexibility. However, after 1.2 s of UV exposure, the initiator concentration was not much and the most of the acrylate double bonds has converted as well as the coatings were less flexible due to the high crosslinking density as shown in Figures 4 and 5 so the reaction rate slowed down resulting to that the hardness of the coatings increased also slowly. Addition of 2.5 wt.% nano-SiO 2 into the coating matrix can make their structure became harder. However, it could also reduce the coating hardness due to the higher roughness of coating surface. These two contradictory impacts might compete with each other, leading to the slight variation in the coating hardness [51].

Coating Abrasion Resistance Study. SiO 2 nanoparticles
have been used widely as nanofillers to enhance the abrasive resistance of polymer materials, due to their high hardness and abrasion resistance [49]. Figure 7 presents the effect of nano-SiO 2 contents on the abrasion resistance of the nanocomposite coating. As can be observed in Figure 7, abrasion resistance of the nanocoating increased with increasing the content of nano-SiO 2 (from 0 to 2.5 wt.%). In the introduction of 2.5 wt.% nanoparticles into the polymer matrix, the abrasion resistance of the coating increased 60% (from 98.3 to 158.4 lite/mil). However, when 5 wt.% nanoparticles were used, the abrasion resistance of the nanocoating was only 133.7 lite/mile (36% increase).  (Figure 8(c)). In this case, these small nanoparticles could not only serve as nanofillers for the polymer matrix, but also actively participate in the polymer-curing process. We expect that these small nanoparticles can penetrate into small holes/indentation/capillaries in the polymer matrices and also can act as nuclei for the growth of crosslinking polymer networks. In case that the content of nano-SiO 2 is 2.5 wt.% is used, large aggregations (~0.5 μm in size) were found in the nanocomposite coating (Figure 8(d)). These aggregations might lead to reduce the mechanical property of the nanocomposite coating.  Figure 9 is TGA curves of the neat coating and the nanocomposite coating with 2.5 wt.% nano-SiO 2 . Table 1 shows the values of temperature at the weight loss of 5%, 50%, and 75% for the coating samples. The data from Figure 9 and Table 1 showed that the initial loss stage of the coatings with weight loss of 5% was observed around 345-360°C. The possible reason could be attributed to the release of both adsorbed water and low molecular organic substances. The T 5% (T initial ) of the neat coating and the nanocomposite with 2.5 wt.% nano-SiO 2 were 348 and 357°C. The T 5% increase of the coating (9°C) due to the presence of the nanoparticles could be interpreted in the terms of their role as thermal stabilizers. The stage of major weight loss of 50% occurred at   Journal of Nanomaterials 420°C for the neat coating, and at 425°C for the nanocomposite was attributed to scission of backbone chain [25,28] in acrylate urethane polymer matrix. The coatings lost 75% of its weight at the temperature of 440°C for the neat coating and 445°C for the nanocomposite. As compared to the styrene acrylic polyurethane and acrylic emulsion coatings [25,28], thermal stability of this nanocomposite coatings is higher. This finding can be explained by the higher crosslinking density and higher thermal resistance chemical bonds in this nanocomposite.

Conclusions
The effect of SiO 2 nanoparticles on the photo-crosslinking polymerization process of UV-curable acrylate epoxy system has been evaluated. Microstructure and properties of the UV curing acrylic epoxy/SiO 2 nanocomposite coating are also characterized.
The main findings of this study were as follows: (i) FESEM images indicated that SiO 2 nanoparticles were homogeneously dispersed into the coatings, when used at the content of 2.5 wt.%. However, large aggregations were found in the nanocomposite coating with 5 wt.% nano-SiO 2 (ii) In the photo-crosslinking polymerization process, FTIR data indicated that nanoparticle reduced the conversion performance of acrylate double bonds, thus slightly increased the swelling degree of nanocomposite coating (iii) Incorporation of SiO 2 nanoparticles into the coating matrix (as nanofillers, at content of 2.5 wt.%) enhanced significantly its abrasion resistance and thermal stability by 60% (from 98.3 to 158.4 lite/mil) and 9°C (from 348°C to 357°C), respectively (iv) In the UV curing acrylic epoxy/SiO 2 nanocomposite coating, SiO 2 nanoparticles can play dual roles, as both UV absorbers and nanoreinforcers

Data Availability
All the data and supporting materials are included within the article.

Conflicts of Interest
All authors declare that there are no conflicts of interest regarding the publication of this paper.