Enhancement in Nanomechanical, Thermal, and Abrasion Properties of SiO 2 Nanoparticle-Modiﬁed Epoxy Coatings

: Epoxy formulations containing 1%, 3%, and 5% SiO 2 nanoparticles (SNPs) were produced and applied to mild steel substrates in order to improve their thermal, nanomechanical, and abrasion resistance. Field emission scanning electron microscopy (FE-SEM) was used to analyze the dispersion of nanoparticles in the ﬁnal coating samples, and Energy-dispersive X-ray spectroscopy (EDX) was used to conﬁrm the presence of nanoparticles. Thermogravimetric analysis (TGA) was employed to measure the thermal resistance of the prepared coatings. Conventional techniques were used to measure the impact and scratch resistance. For nanomechanical testing, nanoindentation was performed using a Berkovich-type indenter. Using a taber abraser, the abrasion properties of the coatings were measured. The FE-SEM images indicated good dispersion of the nanoparticles at all three di ﬀ erent loading levels. The scratch, impact, and hardness of coatings improved with the addition of the SNPs. Nanomechanical properties, such as hardness and elastic modulus, improved when compared to the unmodiﬁed coatings. The thermal and abrasion resistances of the coatings improved with the increase in the SNPs content of the coatings. The highest mechanical, thermal, and abrasion properties were obtained for the coatings with 5% SNP content.


Introduction
The application of organic and inorganic coatings has been well studied with respect to the protection of metals against corrosion and their effectiveness for various structures. However, despite the significant development in coating technologies, challenges remain in the long-term protection of metals from damage and aggressive environments. One of the main reasons for the limited number of high-performance coating systems is the complexity of coating substrate systems and the number of factors that affect the performance and service life of such protective coatings. In addition to the composition of the coating-which consists of resin, solvents, fillers, and additives, such as UV stabilizers and wetting agents-the performance and durability of the coatings depends on several other parameters, such as the type of substrate, pretreatment of the substrate, curing thickness of the coating, and adhesion between the substrate and the coating. For example, a high-performance protective and durable coating must have adequate flexibility and toughness while withstanding impact and cracking as well as maintaining its qualities when subjected to stress, mechanical damage, and well dispersed into the epoxy matrix. Mixing was carried out using a mechanical mixer and ultrasonication. The mechanical, thermal, and abrasion properties of these SNP modified coatings are reported herein, whereas the electrochemical properties have already been studied and reported earlier [24].

Materials and Methods
Diglycidyl ether bisphenol-A (DGEBA; Hexion, Iserlohn, Germany) was the main formulating ingredient as a resin matrix, along with polyamidoamine adduct (Aradur D-450 BD; Huntsman Advance Materials, Germany), which was used as a hardener/crosslinker. Xylene and MIBK were used as solvents to prepare formulations to facilitate mixing and viscosity adjustments. The SNPs used in this study were purchased from Sigma-Aldrich (catalog number: 637238, St. Louis, MO, USA) having a particle size of 10-20 nm.
Three formulations with different percentages of SNPs (1%, 3%, and 5%) were prepared while using stoichiometric balanced amounts of epoxy resin and hardener. Initially, the resin was placed in a beaker and diluted using xylene as a solvent to facilitate mixing by means of a mechanical stirrer (Sheen S2 disperse master, Sheen instruments, Surrey, UK). SNPs, in contrast, were dispersed in acetone in the presence of silane while using the sonication technique to promote mixing in the resin. The nanoparticle mixture was subjected to sonication for 30 min. The dispersed nanoparticle solution was then added slowly to the diluted epoxy resin with continuous stirring using a mechanical stirrer at 500 RPM. After the addition of the dispersed nanoparticles, the formulation was subjected to mechanical mixing at high RPM (5000 RPM) for 45 min. to facilitate the dispersion and removal of the excess solvent. After complete mixing, the stoichiometric amount of hardener was added to complete the formulation. Table 1, below, shows the complete formulation ingredients and the quantities used to prepare the formulations. The FE-SEM (Field emission scanning electron microscope, JEOL, Tokyo, Japan) model JSM-7400F was used to check the morphology of the coating samples along with Energy-dispersive X-ray spectroscopy (EDX) to check the distribution of the added nanoparticles in the final coatings. The samples were mounted to stubs using carbon tape and were then platinum-coated via sputtering.
Thermogravimetric analysis (TGA) was performed on the prepared coatings using a Q600 by TA instruments (New Castle, DE, USA) to observe the thermal stability and decomposition temperatures. A standard aluminum pan was used for this purpose and the samples were heated from room temperature (25 • C) to 600 • C under a N 2 environment with a ramping rate of 10 • C/min.
The surface hardness of the coatings was measured using a Koenig pendulum tester (model 707/K) by Sheen Instruments, Surrey, UK, following the ASTM D-4366, where a higher number of oscillations defines a higher surface hardness. A scratch tester (model 705, Sheen Instruments, Surrey, UK) following the ASTM D-7027 was used in order to measure the scratch resistance of the prepared coatings. The maximum load (in kg) that is sustained by the coatings determines the coatings' scratch resistance. To check the impact resistance of the coatings, a BYK-Gardener impact tester was used (model IG-1120, BYK, Columbia, USA) following the ASTM D-2794 while using a standard weight of 8 lb. The results for pendulum hardness, scratch, and impact resistance are provided as average of the results of the three test specimens.
Using the Nanotest platform from micromaterials, UK, nanomechanical properties, such as the elastic modulus and the hardness of the prepared coatings were determined using a Berkovich-type indenter (Micromaterials, Wrexham, UK). A load control program was used to determine the coating properties by subjecting them to a maximum load of 100 mN. The samples were indented at 1 mN·s −1 until a predetermined load of 100 mN was achieved, followed by a holding period of 60 s. Finally unloading was started at the same strain rate of 1 mN·s −1 until the complete removal of the load. The results obtained through indentation were analyzed using software that was provided with the machine. To obtain result uniformity, at least 10 indentations were performed, although the final obtained results were averaged.
A taber abrader with a dual rotary platform (model 5155) was used using CS-17 wheels (Sheen Instruments, Surrey, UK) to measure the abrasion resistance of the coated samples. Initially, the samples were subjected to abrasion for 500 cycles at a speed of 60 cycles per minute. After 500 cycles, the abrading wheels were refaced for 50 cycles; subsequently, another 500 cycles of abrasion were performed, amounting to a total of 1000 cycles on each sample. Finally, the weight loss was calculated after 1000 cycles by subtracting the final weight from the initial weight. Three samples were tested and the average value was reported for each formulation. Figure 1A-D show the FE-SEM images for unmodified and SNP-modified coatings; all of the images were taken at the same magnification. Figure 1A shows a smooth and plain surface of the epoxy material without the presence of any particles. With the addition of nanoparticles in the epoxy, starting from 1%, followed by 3%, and 5%, changes can be seen from Figure 1B-D. The images show that the nanoparticles are well dispersed and distributed throughout the matrix, with quite a few aggregate formations. The formation of aggregates with the use of nanoparticles is common, happening because of the high attraction forces between nanoparticles due to their high surface area. Another reason is the mixing of particles in a high viscous matrix [25]. EDX was simultaneously performed with SEM; Table 2 shows the overall composition of elements obtained in the EDX.

Field Emission Scanning Electron Microscopy (FE-SEM)
Coatings 2020, 10, x FOR PEER REVIEW 4 of 11 until a predetermined load of 100 mN was achieved, followed by a holding period of 60 s. Finally unloading was started at the same strain rate of 1 mN·s −1 until the complete removal of the load. The results obtained through indentation were analyzed using software that was provided with the machine. To obtain result uniformity, at least 10 indentations were performed, although the final obtained results were averaged. A taber abrader with a dual rotary platform (model 5155) was used using CS-17 wheels (Sheen Instruments, Surrey, UK) to measure the abrasion resistance of the coated samples. Initially, the samples were subjected to abrasion for 500 cycles at a speed of 60 cycles per minute. After 500 cycles, the abrading wheels were refaced for 50 cycles; subsequently, another 500 cycles of abrasion were performed, amounting to a total of 1000 cycles on each sample. Finally, the weight loss was calculated after 1000 cycles by subtracting the final weight from the initial weight. Three samples were tested and the average value was reported for each formulation. Figure 1A-D show the FE-SEM images for unmodified and SNP-modified coatings; all of the images were taken at the same magnification. Figure 1A shows a smooth and plain surface of the epoxy material without the presence of any particles. With the addition of nanoparticles in the epoxy, starting from 1%, followed by 3%, and 5%, changes can be seen from Figure 1B to Figure 1D. The images show that the nanoparticles are well dispersed and distributed throughout the matrix, with quite a few aggregate formations. The formation of aggregates with the use of nanoparticles is common, happening because of the high attraction forces between nanoparticles due to their high surface area. Another reason is the mixing of particles in a high viscous matrix [25]. EDX was simultaneously performed with SEM; Table 2 shows the overall composition of elements obtained in the EDX.    Table 2. Energy-dispersive X-ray spectroscopy (EDX) element detection and weight percentages. 3.2. Thermogravimetric Analysis (TGA) Figure 2 shows the thermal behavior of the epoxy and SNP-modified epoxy coatings and Table 3 summarizes their degradation temperatures (Td). All of the curves depicted a similar weight loss pattern. A two-step degradation can be observed from the graphs. First, a step involves weight loss starting at approximately 100 • C, proceeding until approximately 300 • C, which involves the removal of solvent traces and other volatile substances, the degradation of residual reactants, or the decomposition of the resin's low molecular weight fractions [26][27][28]. Approximately 15% of the weight loss occurred in all of the formulations at this first stage of degradation.

Sample Nanoparticle Elements Present Weight Percentage
Coatings 2020, 10, x FOR PEER REVIEW 5 of 11  Figure 2 shows the thermal behavior of the epoxy and SNP-modified epoxy coatings and Table  3 summarizes their degradation temperatures (Td). All of the curves depicted a similar weight loss pattern. A two-step degradation can be observed from the graphs. First, a step involves weight loss starting at approximately 100 °C, proceeding until approximately 300 °C, which involves the removal of solvent traces and other volatile substances, the degradation of residual reactants, or the decomposition of the resin's low molecular weight fractions [26][27][28]. Approximately 15% of the weight loss occurred in all of the formulations at this first stage of degradation.  The second and major weight loss pattern started from 300 °C and continued to 500 °C. In this degradation profile, the decomposition percentages were recorded, as follows: epoxy (91%), SN-1 (83%), SN-3 (89%), and SN-5 (83%), respectively. This major weight loss at the second stage of degradation above 300 °C is the result of the degradation of the epoxy resin [27,28]. With the addition of SNPs, an increase in thermal resistance is observed, as summarized in Table 3, where temperature at 15%, 25%, 50%, and 75% degradation can be seen in increasing order. It can be clearly observed that the epoxy without nanoparticles at 15%, 25%, 50%, and 75% degradation withstands 335, 367, 410, and 431 °C, while the coating with 5%  The second and major weight loss pattern started from 300 • C and continued to 500 • C. In this degradation profile, the decomposition percentages were recorded, as follows: epoxy (91%), SN-1 (83%), SN-3 (89%), and SN-5 (83%), respectively. This major weight loss at the second stage of degradation above 300 • C is the result of the degradation of the epoxy resin [27,28]. With the addition of SNPs, an increase in thermal resistance is observed, as summarized in Table 3, where temperature at 15%, 25%, 50%, and 75% degradation can be seen in increasing order. It can be clearly observed that the epoxy without nanoparticles at 15%, 25%, 50%, and 75% degradation withstands 335, 367, 410, and 431 • C, while the coating with 5% nanoparticles withstands 344, 377, 420, and 477 • C, which is significantly higher than the unmodified epoxy.

Thermogravimetric Analysis (TGA)
It is apparent from the results ( Table 3) that the addition of nanoparticles increased the thermal stability of coatings, irrespective of the percentages added. The highest thermal stability was observed for the coatings with 5% SNPs. With an increasing silica content, the thermal stability of the epoxy SNPs increased. At higher temperatures, the SNPs move toward the surface because of silica's low surface potential energy. This silica acts as a thermal insulator and protects the polymer [29][30][31].

Mechanical Properties and Nanoindentation Test
The mechanical properties of coatings, such as impact strength and scratch resistance, were analyzed after complete curing (Table 4). An increase in both the impact and scratch resistance was observed in all of formulations, as compared with the epoxy without nanoparticles. With the addition of 5% SNPs, the impact strength increased from 64 to 128 lb/in 2 when compared to the epoxy without nanoparticles, while, in the case of scratch resistance, there is an increase from a 6 kg failure load for the epoxy to a 9 kg failure load with the addition of 5% nanoparticles. The epoxy imparts brittle behavior [32,33], and the increase in its mechanical properties is due to the contribution of the SNPs that increase the modulus [34,35]. This induced a plasticizing effect on the coatings, which, in turn, increased the scratch and impact properties.
Nanoindentation was performed on all of the coating formulations, attaining a maximum load of 100 mN, and typical load vs. displacement curves for all of the coating samples are shown in Figure 3. From the curves, it can be seen that no discontinuities are present (i.e. no significant jump in depth values), which suggests that no cracking occurred during the indenting phase. All of the formulations were subjected to similar loading conditions. From Figure 3, it can be seen that the incorporation of nanoparticles shifted the depth of all curves to lower values, which suggests that the incorporation increased the resistance of the coating against the indentation force; this, in turn, represents an increase in the load-bearing capacity of all the coatings. At the maximum load, the depth values for the epoxy formulation was 10,405 nm, with the addition of 1%, 3%, and 5% the maximum depth values were significantly reduced and restricted to 6360, 6249, and 6184 nm, respectively. This reduction of depth values suggests an increase in the hardness properties.
These load/depth curves were analyzed using software provided by Micro Materials to obtain the hardness and elastic modulus values. This analysis software used the Oliver and Pharr method [36] to derive the desired properties. The equations below show the parameters that were used to derive the hardness and modulus values: Coatings 2020, 10, 310 where H is the hardness, F max is the maximum load, A is the projected contact area at maximum load, E is the modulus of sample, υ is the Poisson's ratio (0.35 for polymer samples), E r is the as reduced modulus, E i (1141 GPa) is the modulus of diamond indenter, and υ i (0.07) is the Poisson's ratio of diamond indenter [37][38][39].
to the contribution of the SNPs that increase the modulus [34,35]. This induced a plasticizing effect on the coatings, which, in turn, increased the scratch and impact properties. Nanoindentation was performed on all of the coating formulations, attaining a maximum load of 100 mN, and typical load vs. displacement curves for all of the coating samples are shown in Figure  3. From the curves, it can be seen that no discontinuities are present (i.e. no significant jump in depth values), which suggests that no cracking occurred during the indenting phase. All of the formulations were subjected to similar loading conditions. From Figure 3, it can be seen that the incorporation of nanoparticles shifted the depth of all curves to lower values, which suggests that the incorporation increased the resistance of the coating against the indentation force; this, in turn, represents an increase in the load-bearing capacity of all the coatings. At the maximum load, the depth values for the epoxy formulation was 10,405 nm, with the addition of 1%, 3%, and 5% the maximum depth values were significantly reduced and restricted to 6360, 6249, and 6184 nm, respectively. This reduction of depth values suggests an increase in the hardness properties.   Figure 4 shows an increase in hardness with the addition of nanoparticles from 0.08 GPa for the unmodified epoxy coating to 0.13 GPa for 1% addition, 0.14 GPa for 3% addition, and 0.15 GPa for 5% nanoparticle addition. Although an increase is observed in all of the added nanoparticle percentages, the highest hardness was obtained for the 5% formulation. Similarly, we can see that the modulus profile of all the coatings increased in all formulations from 3.04 GPa for the unmodified epoxy to 3.55, 3.67, and 3.73 GPa for 1%, 3%, and 5% formulations, respectively. These load/depth curves were analyzed using software provided by Micro Materials to obtain the hardness and elastic modulus values. This analysis software used the Oliver and Pharr method [36] to derive the desired properties. The equations below show the parameters that were used to derive the hardness and modulus values: where H is the hardness, Fmax is the maximum load, A is the projected contact area at maximum load, E is the modulus of sample, υ is the Poisson's ratio (0.35 for polymer samples), Er is the as reduced modulus, Ei (1141 GPa) is the modulus of diamond indenter, and υi (0.07) is the Poisson's ratio of diamond indenter [37][38][39]. Figures 4 and 5 show the graphical representation of hardness and elastic modulus for all of the coating samples. Figure 4 shows an increase in hardness with the addition of nanoparticles from 0.08 GPa for the unmodified epoxy coating to 0.13 GPa for 1% addition, 0.14 GPa for 3% addition, and 0.15 GPa for 5% nanoparticle addition. Although an increase is observed in all of the added nanoparticle percentages, the highest hardness was obtained for the 5% formulation. Similarly, we can see that the modulus profile of all the coatings increased in all formulations from 3.04 GPa for the unmodified epoxy to 3.55, 3.67, and 3.73 GPa for 1%, 3%, and 5% formulations, respectively.  These load/depth curves were analyzed using software provided by Micro Materials to obtain the hardness and elastic modulus values. This analysis software used the Oliver and Pharr method [36] to derive the desired properties. The equations below show the parameters that were used to derive the hardness and modulus values: where H is the hardness, Fmax is the maximum load, A is the projected contact area at maximum load, E is the modulus of sample, υ is the Poisson's ratio (0.35 for polymer samples), Er is the as reduced modulus, Ei (1141 GPa) is the modulus of diamond indenter, and υi (0.07) is the Poisson's ratio of diamond indenter [37][38][39]. Figures 4 and 5 show the graphical representation of hardness and elastic modulus for all of the coating samples. Figure 4 shows an increase in hardness with the addition of nanoparticles from 0.08 GPa for the unmodified epoxy coating to 0.13 GPa for 1% addition, 0.14 GPa for 3% addition, and 0.15 GPa for 5% nanoparticle addition. Although an increase is observed in all of the added nanoparticle percentages, the highest hardness was obtained for the 5% formulation. Similarly, we can see that the modulus profile of all the coatings increased in all formulations from 3.04 GPa for the unmodified epoxy to 3.55, 3.67, and 3.73 GPa for 1%, 3%, and 5% formulations, respectively. The high modulus values of the modified coatings were due to the addition of the SNPs [40][41][42], which in turn, increased the modulus of the coatings [43]. The increase in hardness was mainly due to the free volume reduction in the cross-linked epoxy matrix. This reduction in the free volume happens due to the addition of nanoparticles [44,45].

Abrasion Resistance
The abrasion resistance (ASTM D4060) of silica incorporated coatings was determined using a taber abraser running 1000 cycles at each sample with the refacing of the wheels after 500 cycles. Refacing was performed for 50 cycles. The results that are shown in Figures 6 and 7 were obtained from the following equations: Mass where Minitial represents the initial mass of intact coating (mg) and Mdamaged represents the mass of the coating after abrasion (mg).
The high modulus values of the modified coatings were due to the addition of the SNPs [40][41][42], which in turn, increased the modulus of the coatings [43]. The increase in hardness was mainly due to the free volume reduction in the cross-linked epoxy matrix. This reduction in the free volume happens due to the addition of nanoparticles [44,45].

Abrasion Resistance
The abrasion resistance (ASTM D4060) of silica incorporated coatings was determined using a taber abraser running 1000 cycles at each sample with the refacing of the wheels after 500 cycles. Refacing was performed for 50 cycles. The results that are shown in Figures 6 and 7 were obtained from the following equations: where Minitial represents the initial mass of intact coating (mg) and Mdamaged represents the mass of the coating after abrasion (mg).   Figure 6 shows that the highest recorded weight loss was for the unmodified epoxy coating: 60 mg at 500 cycles and 110 mg at 1000 cycles. The addition of nanoparticles from 1% to 5% decreased the abrasion of the coatings, and the lowest abrasion was recorded with the 5% nanoparticles formulation. At 1000 cycles, a significant increase in the abrasion resistance was witnessed in comparison with the unmodified epoxy coating. The high modulus values of the modified coatings were due to the addition of the SNPs [40][41][42], which in turn, increased the modulus of the coatings [43]. The increase in hardness was mainly due to the free volume reduction in the cross-linked epoxy matrix. This reduction in the free volume happens due to the addition of nanoparticles [44,45].

Abrasion Resistance
The abrasion resistance (ASTM D4060) of silica incorporated coatings was determined using a taber abraser running 1000 cycles at each sample with the refacing of the wheels after 500 cycles. Refacing was performed for 50 cycles. The results that are shown in Figures 6 and 7 were obtained from the following equations: where Minitial represents the initial mass of intact coating (mg) and Mdamaged represents the mass of the coating after abrasion (mg).  Figure 6 shows that the highest recorded weight loss was for the unmodified epoxy coating: 60 mg at 500 cycles and 110 mg at 1000 cycles. The addition of nanoparticles from 1% to 5% decreased the abrasion of the coatings, and the lowest abrasion was recorded with the 5% nanoparticles formulation. At 1000 cycles, a significant increase in the abrasion resistance was witnessed in comparison with the unmodified epoxy coating.  Figure 6 shows that the highest recorded weight loss was for the unmodified epoxy coating: 60 mg at 500 cycles and 110 mg at 1000 cycles. The addition of nanoparticles from 1% to 5% decreased the abrasion of the coatings, and the lowest abrasion was recorded with the 5% nanoparticles formulation. At 1000 cycles, a significant increase in the abrasion resistance was witnessed in comparison with the unmodified epoxy coating.
The results indicated an increase in the abrasion resistance with the addition of nanoparticles [7,46]. Figure 7 shows that the weight loss started to decrease with the addition of nanoparticles. The total weight loss calculated was 60 mg with the addition of 5% SNPs, which was a 45% increase in the abrasion resistance when compared to the 110 mg weight loss of the epoxy without nanoparticles.

Conclusions
The effect of incorporation of SNPs on the mechanical, thermal, and abrasion properties of coatings using epoxy as a base matrix was evaluated. Morphological studies showed a good dispersion of the nanoparticles and some well distributed aggregates of the nanoparticles. The addition of the SNPs increased the overall mechanical properties; highest properties were obtained for the coatings with 5% SNPs. The coatings with 5% SNPs content showed approximately 50% improvements in the scratch resistance and 45% higher abrasion resistance when compared to the unmodified coatings. Nanoindentation results, such as hardness and elastic modulus were found to increase by 87% and 23%, respectively. Irrespective of mechanical properties, coatings with 5% SNPs resulted in better thermal properties and the decomposition temperature increased with an increasing content of the nanoparticles. These coatings provide excellent properties for application on various metal substrates and machinery parts where excellent mechanical properties are desirable.