Revealing the Impact of Micro-SiO2 Filer Content on the Anti-Corrosion Performance of Water-Borne Epoxy Resin

Due to green development in recent years, water-borne epoxy resins (WBE) have become increasingly popular since they generate the lowest level of volatile organic compounds (VOC) during curing. However, because of the large surface tension of water, it is easy to produce voids and cracks during the curing process of the coating. An electrochemical strategy was used in this study to assess the impact of different SiO2 content on the corrosion performance of a WBE coating, in which micron spherical SiO2 particles were synthesized in a liquid phase reduction. The results showed that the synthesized micron spherical SiO2 particles were about 800 ± 50 nm in diameter and in an amorphous state. By hydrophilizing the surfaces of these SiO2 particles, uniform dispersion in an aqueous solvent and a WBE can be achieved. It is important to note that adding a small or excessive amount of SiO2 to a coating will not improve corrosion resistance and may even reduce corrosion resistance. With the appropriate modification of SiO2, corrosion resistance of composite coatings is greatly enhanced, as is the adhesion between the coatings and the metallic substrates. Because the appropriately modified SiO2 can effectively fill the pores that are formed during the curing process, a corrosive medium is less likely to react with the matrix when the medium comes into contact with the matrix. Based on their incorporation content of 3 wt.%, their corrosion resistance is the best after 16 cycles of AC-DC-AC accelerated corrosion tests.


Introduction
Over 2.2 trillion dollars are lost annually as a result of metal corrosion, approximately 3% of gross domestic product. Approximately half of this cost is attributed to corrosion prevention and control globally [1][2][3][4][5][6]. According to a study conducted in the US, metallic corrosion has a direct economic impact of approximately USD 276 billion per year, and corrosion control practices will reduce this cost by 15-35% [7,8]. Epoxy networks have been widely used in coatings among different thermosetting networks, and approximately 90% of metallic materials are protected by organic coatings to provide physical barriers against corrosion [9,10]. At present, polyurethane, phenolic resin, acrylic acid, and epoxy resin are commonly used anticorrosive coatings [11][12][13][14][15]. It is important to note that anticorrosive pigments are composed of various volatile organic compounds (VOCs) which may ingress into the environment during the production, application, or curing of the solvent-based coating formulations [16]. In contrast to other anti-corrosion coatings, water-borne epoxy resin (WBE) is a stable dispersion system in which an epoxy resin is dispersed as small particles or droplets in a continuous phase of water to reduce the use of VOCs as solvents and thinners. Due to the reduction of VOCs, it produces the least number of VOCs during curing [16][17][18]. In recent years, WBE has become increasingly popular due to its environmentally friendly characteristics [19][20][21]. In the film formation process, water's high SiO 2 to improve its dispersion in organic solution. Despite the fact that nano-SiO 2 is a costly material, Anitha et al. developed a hydrophobic coating without fluorine that is stabilized at the air-water interface and that was exhibited to possess exceptional corrosion resistance when exposed to salt spray for more than 1000 h [44]. These studies were mainly conducted with the modification of nano-SiO 2 and micro-SiO 2 for better dispersal with oil epoxy resin, but as more attention has been paid to environmental protection, water-based epoxy resin now tends to be used instead of oil-based epoxy. In spite of this, there are limited studies examining the effects of adding different concentrations of SiO 2 to water-based epoxy resin in order to increase their corrosion resistance.
In this study, different amounts of micro-SiO 2 were used to determine how much micro-SiO 2 was required to maximize anticorrosion properties of WBE coatings. SiO 2 particles with a dimension of 800 ± 50 nm were synthesized and then subjected to hydrophilic modification to achieve uniform dispersion in the epoxy coating independently. A scanning electron microscope (SEM) and Fourier transform infrared spectroscopy (FT-IR) analysis were used to characterize the morphology and the changes of function groups on the SiO 2 surface before and after modification. Instead of using traditional EIS tests to determine the corrosion resistance of composite coatings, AC-DC-AC tests have been recognized as a method of determining corrosion resistance of composite coatings. In order to determine the best anticorrosive properties of modified SiO 2 , both macroscopic and microscopic morphologies of the cross-section of the samples were evaluated after AC-DC-AC testing. As a final step, we conducted scratch tests on composite coatings with varying SiO 2 content on metallic substrates before and after the electrochemical test.

Synthesis of the Modified Spherical SiO 2
The solution-gel method is used to synthesize spherical SiO 2 . TEOS is used as a silicon source, and the CTAB and DDA are used as a surfactant. A homogeneous solution was prepared by dissolving 0.3 g of CTAB and 3.0 g of DDA in a mixture with 200 mL of ethanol, 80 mL of deionized water, and 60 mL of IPA until homogeneously dissolved. Then, the pH was adjusted with a 25% ammonia solution. A continuous stirring process was performed in the reactor in preparation for the next step. In addition, 10 mL of TEOS was added dropwise to the above compound, which was stirred continuously for 4 h at room temperature. Following centrifugation at 6000 rpm/min, the mixture was washed four times with ethanol and then dried under a vacuum for twelve hours at 70 • C.
The KH550 is used as a surface modifier to increase the dispersion of SiO 2 in WBE. Self-synthesized SiO 2 was added to the mixed solution of 100 mL ethanol and 4.3 mL KH550. A centrifuge was used to disperse the compound at a speed of 6000 rpm/min after stirring for four hours. The precipitate was poured into an appropriate amount of ethanol by using ultrasonication for 20 min and then repeated three times to remove the excess KH550. The sediment was dried in an oven.

Preparation of Composite Coating with Different Content of SiO 2
Next, 1 wt.%, 3 wt.%, 5 wt.%, and 10 wt.% modified SiO 2 was completely dispersed in 3 mL of deionized water with ultrasonic vibration, followed by 3 g of aqueous epoxy resin being thoroughly mixed with the SiO 2 dispersion. To the above epoxy resin component, 3 g of the curing agent was added and mixed thoroughly. To ensure uniform coating thickness, the compound was applied directly to pretreated Q235 carbon steel and allowed to cure for 48 h at room temperature. In the course of curing, the mixture gradually changed from white to colorless. Furthermore, a neat epoxy coating was fabricated as a comparison. Upon curing, the coating had a thickness of approximately 120 µm ± 12 µm. This coating was prepared under the same conditions as the control group but without adding SiO 2 . According to the amount of modified SiO 2 in the epoxy resin, different samples were referred to as EP-0, EP-1, EP-3, EP-5, and EP-10. Figure 1 illustrates the schematic diagram of the specific process.

Synthesis of the Modified Spherical SiO2
The solution-gel method is used to synthesize spherical SiO2. TEOS is used as con source, and the CTAB and DDA are used as a surfactant. A homogeneous so was prepared by dissolving 0.3 g of CTAB and 3.0 g of DDA in a mixture with 200 ethanol, 80 mL of deionized water, and 60 mL of IPA until homogeneously diss Then, the pH was adjusted with a 25% ammonia solution. A continuous stirring p was performed in the reactor in preparation for the next step. In addition, 10 mL of was added dropwise to the above compound, which was stirred continuously for room temperature. Following centrifugation at 6000 rpm/min, the mixture was w four times with ethanol and then dried under a vacuum for twelve hours at 70 °C.
The KH550 is used as a surface modifier to increase the dispersion of SiO2 in Self-synthesized SiO2 was added to the mixed solution of 100 mL ethanol and 4 KH550. A centrifuge was used to disperse the compound at a speed of 6000 rpm/min stirring for four hours. The precipitate was poured into an appropriate amount of et by using ultrasonication for 20 min and then repeated three times to remove the e KH550. The sediment was dried in an oven.

Preparation of Composite Coating with Different Content of SiO2
Next, 1 wt.%, 3 wt.%, 5 wt.%, and 10 wt.% modified SiO2 was completely disp in 3 mL of deionized water with ultrasonic vibration, followed by 3 g of aqueous resin being thoroughly mixed with the SiO2 dispersion. To the above epoxy resin co nent, 3 g of the curing agent was added and mixed thoroughly. To ensure uniform co thickness, the compound was applied directly to pretreated Q235 carbon steel a lowed to cure for 48 h at room temperature. In the course of curing, the mixture grad changed from white to colorless. Furthermore, a neat epoxy coating was fabricate comparison. Upon curing, the coating had a thickness of approximately 120 µm ± 1 This coating was prepared under the same conditions as the control group but withou ing SiO2. According to the amount of modified SiO2 in the epoxy resin, different sa were referred to as EP-0, EP-1, EP-3, EP-5, and EP-10. Figure 1 illustrates the schemat gram of the specific process.

Characterization of Modified SiO 2
Field emission scanning electron microscope (FESEM, Zeiss, ULTRATM 55, Jena, Oberkochen, Germany) images of SiO 2 particles and microstructures were taken. X-ray diffraction (XRD, Rigaku Ultima IV instrument, Tokyo, Japan) patterns were obtained by using monochromatic CuK α radiation at a speed of 5 • /min in the range of 5-80 • at 40 kV and 20 mA. Fourier transform infrared spectroscopy (FT-IR, Nicolet iS50 + iN10, PerkinElmer, UK) was used to confirm the successful modification of SiO 2 .

Electrochemical Test
Gamry Framework electrochemical workstations equipped with three electrodes were used to collect the electrochemical data of a variety of samples immersed in NaCl solution of 3.5 percent. A saturated calomel electrode (SCE) was used as the reference electrode, a platinum plate with a surface area of 10 cm 2 was used as the counter electrode, and the different samples were used as the working electrodes. During the measurement, a Faraday cage was placed inside the electrochemical cell to minimize external disturbances to the system. Impedance spectra were recorded after an open circuit potential (OCP) test by using the Gamry Framework. During the AC-DC-AC test, all EIS tests were performed at frequencies ranging from 10 −1 Hz to 10 5 Hz.
As a test method for observing the effect of adding SiO 2 on the anticorrosion performance of the coating more quickly, the composite coating of SiO 2 and epoxy resin was immersed in the NaCl solution of 3.5 percent. As soon as the OCP stabilized in 2 h, the AC-DC-AC test was performed as follows. In the first instance, the EIS was measured under OCP at this time, followed by cathodic polarization for 30 min at a DC voltage of −4 V and anodic polarization for 10 min at a DC voltage of +4 V, and then the system was relaxed again at the OCP until it returned to a steady state. Last but not least, we measured the EIS under an OCP that returned to steady state after a short period of time. A total of 15 cycles of AC-DC-AC tests were performed. The schematic diagram of an AC-DC-AC test is shown in Figure 2. For each sample, three parallel samples were prepared with different SiO 2 additions. Using Zview software, the EIS data were fitted to electrical equivalent circuit models. diffraction (XRD, Rigaku Ultima IV instrument, Tokyo, Japan) patterns were obtained by using monochromatic CuKα radiation at a speed of 5°/min in the range of 5-80° at 40 kV and 20 mA. Fourier transform infrared spectroscopy (FT-IR, Nicolet iS50 + iN10, Perki-nElmer, UK) was used to confirm the successful modification of SiO2.

Electrochemical Test
Gamry Framework electrochemical workstations equipped with three electrodes were used to collect the electrochemical data of a variety of samples immersed in NaCl solution of 3.5 percent. A saturated calomel electrode (SCE) was used as the reference electrode, a platinum plate with a surface area of 10 cm 2 was used as the counter electrode, and the different samples were used as the working electrodes. During the measurement, a Faraday cage was placed inside the electrochemical cell to minimize external disturbances to the system. Impedance spectra were recorded after an open circuit potential (OCP) test by using the Gamry Framework. During the AC-DC-AC test, all EIS tests were performed at frequencies ranging from 10 −1 Hz to 10 5 Hz.
As a test method for observing the effect of adding SiO2 on the anticorrosion performance of the coating more quickly, the composite coating of SiO2 and epoxy resin was immersed in the NaCl solution of 3.5 percent. As soon as the OCP stabilized in 2 h, the AC-DC-AC test was performed as follows. In the first instance, the EIS was measured under OCP at this time, followed by cathodic polarization for 30 min at a DC voltage of −4 V and anodic polarization for 10 min at a DC voltage of +4 V, and then the system was relaxed again at the OCP until it returned to a steady state. Last but not least, we measured the EIS under an OCP that returned to steady state after a short period of time. A total of 15 cycles of AC-DC-AC tests were performed. The schematic diagram of an AC-DC-AC test is shown in Figure 2. For each sample, three parallel samples were prepared with different SiO2 additions. Using Zview software, the EIS data were fitted to electrical equivalent circuit models.

Surface Characterization
By using digital optical microscopy, the surface topography of the coated samples was recorded before and after the cross-orthogonal cycle. Scanning electron microscopes

Surface Characterization
By using digital optical microscopy, the surface topography of the coated samples was recorded before and after the cross-orthogonal cycle. Scanning electron microscopes were used to examine the microscopic morphology of the coating in its cross-section under an electron accelerating potential of 1.5 kV.
The adhesion of composite coatings with different SiO 2 content before and after the AC-DC-AC test was studied by using an automatic scratch testing machine (WS-2005, Lanzhou Zhongke Kaihua Science and Technology Development Co., Ltd., Lanzhou, China). The diamond indenter tip radius was 200 µm, and the cone angle was 120 • . A linear load increase of 50 N per minute was set from 0 to 100 N. A scratch length of 4 mm was set.

Characterization of SiO 2 and Composite Coating
The morphology of the synthesized SiO 2 particles was observed by using SEM. As shown in Figure 3a, the cluster of SiO 2 particles reveals a regular spherical shape and quite a narrow distribution in particle size. To obtain the details of the SiO 2 particles, ultrasonic dispersion was employed to get individual SiO 2 particles for TEM observation. As shown in Figure 3b, the exemplified SiO 2 particle displays an opaque morphology, which suggests that it could hardly be penetrated by the electron beam under the specific energy. In addition, the average diameter of the SiO 2 particles was determined to be 800 ± 50 nm. An XRD analysis was performed on the synthesized SiO 2 particles in order to determine their phase composition. As shown in Figure 3c, a typical broad peak is found at the vicinity of 21 • , which suggests the amorphous state of the SiO 2 particles synthesized via the modified Stober method [45][46][47][48].    Figure 4a and Table 3, it can be seen that the final residual content of the composite coating increased with the theoretical content increasing, and because the melting point of SiO2 is 1700 °C, its weight will not change at 600 °C. This proves that the theoretical addition value of SiO2 is close to the actual addition. It should also be noted that the maximum decomposition temperature (T d peak) is an important indicator to evaluate the thermal stability of a composite coating [56,57]. With the addition of SiO2, the maximum decomposition temperature increases, and it may be evident that the thermal stability of the coating gradually improves when the SiO2 content increases as shown in Figure 4b [58]. When the waterborne epoxy resin is modified with -NH2 on the surface of SiO2 and dehydration condensation of -COOH in the waterborne epoxy resin, the compactness and glass transition temperature of the composite coating can be improved as a result of the Si-O-Si bonds in SiO2 and the high degree of cross-linking between KH550 and epoxy resin. Therefore, the addition of the modified SiO2 can enhance the coating's stability.  Figure 3d exhibits the FT-IR spectra of SiO 2 particles before and after surface modification. As can be seen, the absorption peak at 1104 cm −1 and 802 cm −1 represents vibrations of Si-O-Si bonds, and the absorption peak at 3449 cm −1 and 1629 cm −1 recirculates with vibrations of -OH bonds [49][50][51]. As a result of the hydrogen bond between water molecules and SiO 2 for the peaks observed at around 1646 cm −1 , it may be possible to demonstrate that the process of synthesis of the SiO 2 is responsible for the observed peaks. SiO 2 modified by KH550 displayed characteristic absorption peaks at 3449 cm −1 (-OH), 1629 cm −1 (-OH), and 1074 cm −1 (C-O-C) compared with self-synthesized SiO 2 . It is likely that the enhancement of this peak at 3449 cm −1 is a consequence of the -NH 2 in KH550, as the introduction of amino groups can form hydrogen bonds with water molecules, thus increasing the hydrophilicity of SiO 2 and promoting their good dispersion in aqueous epoxy resin. A new absorption peak was observed at 2924 cm −1 as a result of the stretching vibration peaks on the -CH 3 in KH550 [52][53][54]. KH550 was grafted successfully onto SiO 2 based on such results. Figure S1 shows the dispersibility of SiO 2 in water before and after modification and various concentrations. According to Figure S1, the different contents of SiO 2 immersed in water after 24 h have different dispersibility and stability. The unmodified SiO 2 shows an obvious sediment in the water. In contrast, KH550 modified SiO 2 (1 wt.% and 3 wt.%) can be uniformly suspended in water and stood for 24 h. This is primarily due to the formation of hydrogen bonds that were formed between the -NH 2 on the surface of modified SiO 2 and the -OH in water. Then, the water with a content 5 wt.% and 10 wt.% SiO 2 also displays different degrees of settlement. Therefore, as the SiO 2 content increases, the dispersibility in aqueous solution will decrease. Due to its internal structure of siloxane, and its surface consisting of dihydroxyl groups, isolated hydroxyl groups, and adjacent hydroxyl groups, SiO 2 itself is easy to reunite due to its large number of hydroxyl groups [55]. In contrast, hydrogen bonds are formed when amino groups grafted onto SiO 2 modified by KH550 are bonded with water molecules, thereby improving the dispersibility of SiO 2 . However, when too much SiO 2 is added, the strong electrostatic attraction of the hydroxyl groups on silicon's surface will cause the SiO 2 to agglomerate into large particles which will deposit as a result of their own weight. Figure 4 illustrates both the thermal analysis curve for a coating without fillers and the composite coatings containing various concentrations of SiO 2 . Based on Figure 4a and Table 3, it can be seen that the final residual content of the composite coating increased with the theoretical content increasing, and because the melting point of SiO 2 is 1700 • C, its weight will not change at 600 • C. This proves that the theoretical addition value of SiO 2 is close to the actual addition. It should also be noted that the maximum decomposition temperature (T d peak ) is an important indicator to evaluate the thermal stability of a composite coating [56,57]. With the addition of SiO 2 , the maximum decomposition temperature increases, and it may be evident that the thermal stability of the coating gradually improves when the SiO 2 content increases as shown in Figure 4b [58]. When the waterborne epoxy resin is modified with -NH 2 on the surface of SiO 2 and dehydration condensation of -COOH in the waterborne epoxy resin, the compactness and glass transition temperature of the composite coating can be improved as a result of the Si-O-Si bonds in SiO 2 and the high degree of cross-linking between KH550 and epoxy resin. Therefore, the addition of the modified SiO 2 can enhance the coating's stability.     A comparison of epoxy resin and composite coating with different levels of modified SiO 2 is shown in Figure 5. According to Figure 5(a1), the surface of the pure epoxy coating is relatively smooth, whereas Figure 5(b1-e1) shows that the amount of modified SiO 2 increases with increasing SiO 2 content. It is possible to observe that SiO 2 is evenly distributed over the surface of all composite coating. It can be shown in Figure 5(b1,c1) that SiO 2 is evenly dispersed in the coating and also presents different dispersion states on the coating surface. However, there is an obvious agglomeration of fillers on the surface of the coating when SiO 2 is added at an amount of 5 wt.% as shown in Figure 5(d2). However, when the addition of SiO 2 was 10 wt.%, a large amount of SiO 2 piled up on the coating surface.  A comparison of epoxy resin and composite coating with different levels of modified SiO2 is shown in Figure 5. According to Figure 5(a1), the surface of the pure epoxy coating is relatively smooth, whereas Figure 5(b1-e1) shows that the amount of modified SiO2 increases with increasing SiO2 content. It is possible to observe that SiO2 is evenly distributed over the surface of all composite coating. It can be shown in Figure 5(b1,c1) that SiO2 is evenly dispersed in the coating and also presents different dispersion states on the coating surface. However, there is an obvious agglomeration of fillers on the surface of the coating when SiO2 is added at an amount of 5 wt.% as shown in Figure 5(d2). However, when the addition of SiO2 was 10 wt.%, a large amount of SiO2 piled up on the coating surface.

Electrochemical Test
The AC-DC-AC test is used in this study to assess the coating's protective properties more rapidly. The difference between AC-DC-AC testing methods and natural immersion in composite coating is shown in Figure 6. Through the defects in the coating, the corrosion medium is brought into contact with the metallic substrate directly, and the reactions take place through cathodic and anodic reactions, resulting in corrosion as shown in Figure 6a. When the water penetrates the metallic substrate, oxygen reduction and alkalization occur through cathodic polarization, which pushes the coating away from the metallic substrate to promote the damage of the coating under the action of external electric field during AC-DC-AC tests. It has been observed that large quantities of oxides and hydroxides form on the metallic substrate during anodic polarization, further promoting the delamination between the coating and the metallic substrate. take place through cathodic and anodic reactions, resulting in corrosion as shown in Fig-ure 6a. When the water penetrates the metallic substrate, oxygen reduction and alkalization occur through cathodic polarization, which pushes the coating away from the metallic substrate to promote the damage of the coating under the action of external electric field during AC-DC-AC tests. It has been observed that large quantities of oxides and hydroxides form on the metallic substrate during anodic polarization, further promoting the delamination between the coating and the metallic substrate. In order to regulate the degradation of the WBE coatings without and with the addition of SiO2 fillers, preliminary AC-DC-AC experiment was conducted by employing different DC excitation voltages upon the blank WBE coatings. As such, the Bode spectra obtained under ±2 V, ±3 V, and ±4 V cyclic DC polarization are shown in Figure 7. The terminated impedance values obtained at 10 −1 Hz were used to evaluate the variation of the anti-corrosion performance of different systems. As shown in Figure 7a, an expired impedance of approximately 10 6 Ω·cm 2 is obtained after two cycles of ±2 V polarization, after which this impedance resistance value reduces to about one decade lower after the 4th cycle, and it remains with relatively limited changes until the end (10 cycles) of testing. The difference in impedance obtained between the 2nd and the following polarization cycles could be caused by the diffusion of corrosive species, which would move into the substrate through the inherent tiny pores, and the stable EIS test results could demonstrate that the coating without fillers is not damaged under this polarization condition. Increasing the cyclic excitation voltage to ±3 V, the impedance value drops to 10 5 Ω·cm 2 after two cycles polarization as shown in Figure 7b. In spite of this, the DC voltage of ±3 V is still not enough to quickly evaluate the corrosion resistance of the coating. A clear but not aggressive degradation of the coating without fillers is achieved by applying the excitation voltage of ±4 V as shown in Figure 7c. The terminated impedance values are 2.2 × 10 5 , 1.7 × 10 4 , 2.6 × 10 4 , and 4.7 × 10 4 Ω·cm 2 after 2nd, 4th, 6th, and 8th polarization, respectively, the results are also coupled with the gradual shift of the phase towards lower degrees. However, further increasing the polarization repetition to 10 cycles, the impedance In order to regulate the degradation of the WBE coatings without and with the addition of SiO 2 fillers, preliminary AC-DC-AC experiment was conducted by employing different DC excitation voltages upon the blank WBE coatings. As such, the Bode spectra obtained under ±2 V, ±3 V, and ±4 V cyclic DC polarization are shown in Figure 7. The terminated impedance values obtained at 10 −1 Hz were used to evaluate the variation of the anti-corrosion performance of different systems. As shown in Figure 7a, an expired impedance of approximately 10 6 Ω·cm 2 is obtained after two cycles of ±2 V polarization, after which this impedance resistance value reduces to about one decade lower after the 4th cycle, and it remains with relatively limited changes until the end (10 cycles) of testing. The difference in impedance obtained between the 2nd and the following polarization cycles could be caused by the diffusion of corrosive species, which would move into the substrate through the inherent tiny pores, and the stable EIS test results could demonstrate that the coating without fillers is not damaged under this polarization condition. Increasing the cyclic excitation voltage to ±3 V, the impedance value drops to 10 5 Ω·cm 2 after two cycles polarization as shown in Figure 7b. In spite of this, the DC voltage of ±3 V is still not enough to quickly evaluate the corrosion resistance of the coating. A clear but not aggressive degradation of the coating without fillers is achieved by applying the excitation voltage of ±4 V as shown in Figure 7c. The terminated impedance values are 2.2 × 10 5 , 1.7 × 10 4 , 2.6 × 10 4 , and 4.7 × 10 4 Ω·cm 2 after 2nd, 4th, 6th, and 8th polarization, respectively, the results are also coupled with the gradual shift of the phase towards lower degrees. However, further increasing the polarization repetition to 10 cycles, the impedance increases to 8.3 × 10 4 Ω·cm 2 and a new time constant appears at the middle frequency region. A significant change in the Bode spectrum can be attributed to the corrosion of the metallic substrate, whose corrosion products conceal the defects caused by the polarization in the resin coatings [59]. With the consideration of the difference in EIS results obtained under different DC polarization values, the anti-corrosion performance of WBE resin coatings and composite coatings with different additions of SiO 2 has been evaluated under ±4 V. region. A significant change in the Bode spectrum can be attributed to the corrosion of the metallic substrate, whose corrosion products conceal the defects caused by the polarization in the resin coatings [59]. With the consideration of the difference in EIS results obtained under different DC polarization values, the anti-corrosion performance of WBE resin coatings and composite coatings with different additions of SiO2 has been evaluated under ±4 V.  Figure 8 depicts the Bode plots for the WBE resin coating and composite coatings containing different amounts of modified SiO2. For better comparison, the Bode impedance and phase plots of cycle 3, cycle 6, cycle 9, cycle 12, and cycle 15 DC bipolarization are illustrated. EP-0 obtained the highest impedance value of 2.2 × 10 5 Ω·cm 2 in the low frequency and also the highest phase angle −4.1° after the 3rd polarization of the AC-DC-AC test. But the impedance value dropped to the lowest value after the 9th polarization of the AC-DC-AC test, which is 1.6 × 10 5 Ω·cm 2 . After the 12th and 15th polarizations, corrosion products filled in the damaged area of the coating, resulting in a slight increase in low frequency impedance. The phase angle displays a new time constant at the middle frequency region, which is due to interfacial corrosion caused by the corrosive medium penetrating the metallic substrate [59]. In addition to revealing significant fluctuations in EP-1 impedance at low frequencies, the AC-DC-AC tests also reveal significant phase angle fluctuations. Perhaps this is due to the fact that the addition of a small quantity of SiO2 cannot fully plug the pores that were produced during the coating cures. EP-3 maintained the best corrosion resistance, as shown in Figure 8c, and the impedance value in low frequency and the phase angle in high frequency were nearly constant after the 9th AC-DC-AC cycles, which are 9.4 × 10 5 Ω·cm 2 and −0.9°, respectively. Consequently, it is possible to demonstrate that the modified SiO2 in an appropriate amount could improve the corrosion resistance of the coating. Additionally, due to the high amounts of amino groups in KH550, cross-linking is enhanced during coating curing, which results in a higher  For better comparison, the Bode impedance and phase plots of cycle 3, cycle 6, cycle 9, cycle 12, and cycle 15 DC bipolarization are illustrated. EP-0 obtained the highest impedance value of 2.2 × 10 5 Ω·cm 2 in the low frequency and also the highest phase angle −4.1 • after the 3rd polarization of the AC-DC-AC test. But the impedance value dropped to the lowest value after the 9th polarization of the AC-DC-AC test, which is 1.6 × 10 5 Ω·cm 2 . After the 12th and 15th polarizations, corrosion products filled in the damaged area of the coating, resulting in a slight increase in low frequency impedance. The phase angle displays a new time constant at the middle frequency region, which is due to interfacial corrosion caused by the corrosive medium penetrating the metallic substrate [59]. In addition to revealing significant fluctuations in EP-1 impedance at low frequencies, the AC-DC-AC tests also reveal significant phase angle fluctuations. Perhaps this is due to the fact that the addition of a small quantity of SiO 2 cannot fully plug the pores that were produced during the coating cures. EP-3 maintained the best corrosion resistance, as shown in Figure 8c, and the impedance value in low frequency and the phase angle in high frequency were nearly constant after the 9th AC-DC-AC cycles, which are 9.4 × 10 5 Ω·cm 2 and −0.9 • , respectively. Consequently, it is possible to demonstrate that the modified SiO 2 in an appropriate amount could improve the corrosion resistance of the coating. Additionally, due to the high amounts of amino groups in KH550, cross-linking is enhanced during coating curing, which results in a higher coating density. It lacks a significant change in phase angle at high frequency, but the impedance value in low frequency of the samples decreased significantly within the same number of cycles when the SiO 2 content increased to 5 wt.% and 10 wt.%. In comparison with other samples, EP-10 has the lowest impedance value and small flow during all AC-DC-AC tests. The impedance value of EP-10 in low frequency is 9.1 × 10 4 Ω·cm 2 after the 15th polarization, indicating that the sample maintains the lowest corrosion resistance. This may be due to an excessive amount of SiO 2 causing the agglomeration of fillers, which leads to a reduction in the uniformity of the coating and an appearance of larger defects during curing.
DC-AC tests. The impedance value of EP-10 in low frequency is 9.1 × 10 4 Ω·cm 2 after the 15th polarization, indicating that the sample maintains the lowest corrosion resistance. This may be due to an excessive amount of SiO2 causing the agglomeration of fillers, which leads to a reduction in the uniformity of the coating and an appearance of larger defects during curing. Based on the AC-DC-AC testing, Figure 9 shows the evolution of the coating's characteristic parameters over the cycle times. In Figure 8f, an equivalent circuit model is presented, which can be used to interpret the experimental EIS data for pure and composite Based on the AC-DC-AC testing, Figure 9 shows the evolution of the coating's characteristic parameters over the cycle times. In Figure 8f, an equivalent circuit model is presented, which can be used to interpret the experimental EIS data for pure and composite coatings on metallic substrates. R s refers to the resistance of the electrolyte, which is the resistance between the reference electrode and working electrode. An electrolyte's porous resistance (R p ) is determined by the diameter of pores, the porosity, and the characteristics of capillary channels by which the ions penetrate an interface. In Figure 9a, the R p changes of different samples are relatively different during the first six cycles. Relative to other samples, the R p values of EP-0 and EP-1 are relatively high, but there is a large drop in the first four AC-DC cycles, suggesting that an aggressive species might have penetrated into the metal substrate surface of EP-0 and EP-1 due to there being more micropores on their surfaces [18]. The R p values of EP-5 and EP-10 are low and show a gradual upward trend in the first six cycles. It was found that the R p of EP-3 was changed by a small amount in comparison with other composite coatings during the 1st to 6th cycles. During the 6th to 15th cycles, the R p values of EP-0, EP-1, EP-5, and EP-10 all increased slightly. The corrosion resistance of the coating is increased because the corrosion products produced during anodic polarization temporarily resist defects created during coating degradation [60]. However, during the 6th to 15th cycles, the R p of EP-3 remains relatively stable, indicating that it has the best corrosion resistance. A greater number of electrolytes will be absorbed by the coating as moisture permeates into it, thus increasing the coating capacitance C C [61]. In the electrical equivalent of Figure 8f, C CPE, T and C CPE, P are parallel to R p , making it possible to calculate its effective coating capacitance by using Equation (1): where R p is the resistance of the resistance parallel to the C CPE,T and C CPE,P. All capacitance values presented were determined according to Equation (1) [62].
the metal substrate surface of EP-0 and EP-1 due to there being more micropores on their surfaces [18]. The Rp values of EP-5 and EP-10 are low and show a gradual upward trend in the first six cycles. It was found that the Rp of EP-3 was changed by a small amount in comparison with other composite coatings during the 1st to 6th cycles. During the 6th to 15th cycles, the Rp values of EP-0, EP-1, EP-5, and EP-10 all increased slightly. The corrosion resistance of the coating is increased because the corrosion products produced during anodic polarization temporarily resist defects created during coating degradation [60]. However, during the 6th to 15th cycles, the Rp of EP-3 remains relatively stable, indicating that it has the best corrosion resistance. A greater number of electrolytes will be absorbed by the coating as moisture permeates into it, thus increasing the coating capacitance CC [61]. In the electrical equivalent of Figure 8f, CCPE, T and CCPE, P are parallel to Rp, making it possible to calculate its effective coating capacitance by using Equation (1): where Rp is the resistance of the resistance parallel to the CCPE,T and CCPE,P. All capacitance values presented were determined according to Equation (1) [62]. The changes of CC of different samples in the AC-DC-AC test are shown in Figure 9b. The CC of EP-0 gradually increases, indicating that the capacitive effect of the pure coating is increased. The CC of EP-3 increased slightly during the 1st to 6th cycles but remained relatively stable overall, which indicates that no further reaction occurred at the interface between coating and metallic substrate after the coating reached water saturation. From the 6th to the 15th cycles, there is no significant change in the corrosion coefficient of all samples, indicating that the coating has reached saturation in terms of water uptake [63].

Corrosion Evaluation of Composite Coating
The surface morphology of the coating should be assessed upon completion of the AC-DC-AC test in order to determine the most appropriate amount of SiO2 to add to the coating. Based on the AC-DC-AC test results, it can be seen that pure coatings and composite coatings with different SiO2 content exhibit different surface morphologies as The changes of C C of different samples in the AC-DC-AC test are shown in Figure 9b. The C C of EP-0 gradually increases, indicating that the capacitive effect of the pure coating is increased. The C C of EP-3 increased slightly during the 1st to 6th cycles but remained relatively stable overall, which indicates that no further reaction occurred at the interface between coating and metallic substrate after the coating reached water saturation. From the 6th to the 15th cycles, there is no significant change in the corrosion coefficient of all samples, indicating that the coating has reached saturation in terms of water uptake [63].

Corrosion Evaluation of Composite Coating
The surface morphology of the coating should be assessed upon completion of the AC-DC-AC test in order to determine the most appropriate amount of SiO 2 to add to the coating. Based on the AC-DC-AC test results, it can be seen that pure coatings and composite coatings with different SiO 2 content exhibit different surface morphologies as shown in Figure 10. According to Figure 10(a1,b1,d1), there are some bulges on the surface of the EP-0, EP-1, EP-5 coatings. Further observation of these areas of bubble formation through an optical microscope shows that there are corrosion products precipitated around these areas as shown in Figure 10(a2,b2,d2). Due to cathodic polarization, oxygen reduction and alkalinization are generated, which stretches out the coating surface, reduces the coating density, and facilitates the corrosive medium permeating around this large bulge, which in turn makes corrosion products more likely to accumulate around these areas during anodic polarization. Therefore, corrosion may first occur in these areas. As can be seen in Figure 10(e1), the coating is almost completely stripped from the metallic substrate, and there is a great deal of rusting products around the exfoliated coating as shown in Figure 10(e2). Through the cracks and voids of the composite coating, the corrosion medium had penetrated the metallic substrate. Cathodic polarization also produced oxygen reduction and alkalinization, which would exacerbate the degradation and spalling process of the coating. As illustrated in Figure 10(c1), corrosion damage on the coating surface is relatively slight, and there is almost no trace of corrosion damage as shown in Figure 10(c2). The EP-3 coating has the best barrier properties to aggressive media. The observed results of the coating surface morphology are consistent with the AC-DC-AC test results as shown in Figure 8a-e.
through an optical microscope shows that there are corrosion products precipitated around these areas as shown in Figure 10(a2,b2,d2). Due to cathodic polarization, oxygen reduction and alkalinization are generated, which stretches out the coating surface, reduces the coating density, and facilitates the corrosive medium permeating around this large bulge, which in turn makes corrosion products more likely to accumulate around these areas during anodic polarization. Therefore, corrosion may first occur in these areas. As can be seen in Figure 10(e1), the coating is almost completely stripped from the metallic substrate, and there is a great deal of rusting products around the exfoliated coating as shown in Figure 10(e2). Through the cracks and voids of the composite coating, the corrosion medium had penetrated the metallic substrate. Cathodic polarization also produced oxygen reduction and alkalinization, which would exacerbate the degradation and spalling process of the coating. As illustrated in Figure 10(c1), corrosion damage on the coating surface is relatively slight, and there is almost no trace of corrosion damage as shown in Figure 10(c2). The EP-3 coating has the best barrier properties to aggressive media. The observed results of the coating surface morphology are consistent with the AC-DC-AC test results as shown in Figure 8a-e. According to the results of the AC-DC-AC test, the corrosion damage of the coating can be determined by detecting the blistering and peeling that takes place at the interface, which would be broken between the coating and the metallic substrate due to metallic substrate corrosion. Figure 11 shows the cross-sectional morphology of EP-0, EP-3, and EP-10. According to Figure 11(a1), the coating has been stripped off from the metallic substrate and some corrosion products have been stripped in the delamination. In Figure  11(a2), there is an obvious delamination between the corrosion product and the substrate, leading to a significant decrease in the adhesion of the coating. Figure 11(b1) indicates that there is no obvious corrosion product between the coating and the metallic substrate. Furthermore, there is a little delamination between the coating and the metallic substrate as shown in Figure 11(b2). In Figure 11(c1), there is an obvious bulge in the composite coating, and it is severely separated from the metallic substrate. It can be shown than there are many loose corrosions products deposited on the top of the metallic substrate as shown in Figure 11(c2), indicating that the interface was subject to serious corrosion. In comparison with other samples, from the cross-section morphology between the EP-3 coating and the metallic substrate, there was a slight peeling of the coating away from the metallic substrate but no obvious damage to the coating after the AC-DC-AC test, indicating a low degree of corrosion on the metallic substrate. By adding a sufficient and appropriate amount of SiO2 to the coating, the pores generated during the curing process of the coating can be effectively blocked, thereby increasing the difficulty for water molecules or other corrosive media to penetrate the coating and enter the metallic substrate. According to the results of the AC-DC-AC test, the corrosion damage of the coating can be determined by detecting the blistering and peeling that takes place at the interface, which would be broken between the coating and the metallic substrate due to metallic substrate corrosion. Figure 11 shows the cross-sectional morphology of EP-0, EP-3, and EP-10. According to Figure 11(a1), the coating has been stripped off from the metallic substrate and some corrosion products have been stripped in the delamination. In Figure 11(a2), there is an obvious delamination between the corrosion product and the substrate, leading to a significant decrease in the adhesion of the coating. Figure 11(b1) indicates that there is no obvious corrosion product between the coating and the metallic substrate. Furthermore, there is a little delamination between the coating and the metallic substrate as shown in Figure 11(b2). In Figure 11(c1), there is an obvious bulge in the composite coating, and it is severely separated from the metallic substrate. It can be shown than there are many loose corrosions products deposited on the top of the metallic substrate as shown in Figure 11(c2), indicating that the interface was subject to serious corrosion. In comparison with other samples, from the cross-section morphology between the EP-3 coating and the metallic substrate, there was a slight peeling of the coating away from the metallic substrate but no obvious damage to the coating after the AC-DC-AC test, indicating a low degree of corrosion on the metallic substrate. By adding a sufficient and appropriate amount of SiO 2 to the coating, the pores generated during the curing process of the coating can be effectively blocked, thereby increasing the difficulty for water molecules or other corrosive media to penetrate the coating and enter the metallic substrate. The adhesion test results of composite coatings containing different amounts of SiO2 are shown in Figure 12. It is evident that adhesion strength increased with increased SiO2 content as shown in Figure 12. There is an average adhesion of 81 N for the EP-10 coating. This is due to the fact that SiO2 has a certain reinforcing effect as a hard phase, and it disperses widely on the coating's surfaces. As a result of the interaction between particles and the active groups of the epoxy resin or curing agent, the coating is more tightly bonded with the metallic substrate, improving the coating's adhesion. Compared with other samples, the adhesion between the EP-3 coating and the metallic substrate decreased the least after the AC-DC-AC test, which was only 37 N. However, the adhesion between the EP-10 coating and the metallic substrate decreased to 18 N. In this case, excessive SiO2 was added to the epoxy resin, resulting in an agglomeration at the interface that reduces the uniformity of the coating and causes larger defects. This, in turn, will result in more corrosion media entering the metallic substrate, resulting in more oxygen reduction and alkalinization being produced during the AC-DC-AC test. Cathodic polarization results in oxygen reduction and alkalinization, which accelerates coating spalling, resulting in a reduction of adhesion between the coating and the metallic substrate [62]. The adhesion test results of composite coatings containing different amounts of SiO 2 are shown in Figure 12. It is evident that adhesion strength increased with increased SiO 2 content as shown in Figure 12. There is an average adhesion of 81 N for the EP-10 coating. This is due to the fact that SiO 2 has a certain reinforcing effect as a hard phase, and it disperses widely on the coating's surfaces. As a result of the interaction between particles and the active groups of the epoxy resin or curing agent, the coating is more tightly bonded with the metallic substrate, improving the coating's adhesion. Compared with other samples, the adhesion between the EP-3 coating and the metallic substrate decreased the least after the AC-DC-AC test, which was only 37 N. However, the adhesion between the EP-10 coating and the metallic substrate decreased to 18 N. In this case, excessive SiO 2 was added to the epoxy resin, resulting in an agglomeration at the interface that reduces the uniformity of the coating and causes larger defects. This, in turn, will result in more corrosion media entering the metallic substrate, resulting in more oxygen reduction and alkalinization being produced during the AC-DC-AC test. Cathodic polarization results in oxygen reduction and alkalinization, which accelerates coating spalling, resulting in a reduction of adhesion between the coating and the metallic substrate [62]. The corrosion protection mechanism of composite coatings is shown in Figure 13. According to Figure 13a, a single water-borne epoxy coating is easily porous during curing, because of the large surface tension of water. This does not provide adequate barrier properties, which results in corrosion of the metallic substrate through the easy entry of corrosive media through micropores. Figure 13b shows that a small amount of SiO2 is not capable of effectively filling the pores in epoxy resin, which does not improve its anticorrosion performance. Figure 13c illustrates that when the coating is uniformly dispersed with an appropriate amount of SiO2, pores can be effectively plugged and corrosive media are less likely to penetrate. Consequently, the coating performs better as a barrier and is more corrosion resistant. In Figure 13d, the agglomeration of SiO2 will lead to the decrease of the specific surface area of fillers, resulting in greater defects in the interface between fillers and resin during volume shrinkage during resin curing, and the agglomerated fillers will also reduce the uniformity of the coating. Therefore, excessive water absorption of the composite coating, increased internal stress, and bubbling occur. The composite coating may lose its protective effect in the application process. The coating's protective properties can, therefore, be increased by adding a suitable amount of SiO2 to the resin. It is possible to block micropores during curing by adding sufficient quantities of SiO2 to the coating. SiO2 has the characteristics of a large specific surface area, high specific surface energy, and strong activity, which enable it to facilitate the cross-linking reaction between epoxy resins and improve the coating's compactness and mechanical properties [64][65][66]. It is as a result of this improvement in density that electrolyte permeation is slowed, thus increasing the coating's protection and allowing it to last for a long period of time. The corrosion protection mechanism of composite coatings is shown in Figure 13. According to Figure 13a, a single water-borne epoxy coating is easily porous during curing, because of the large surface tension of water. This does not provide adequate barrier properties, which results in corrosion of the metallic substrate through the easy entry of corrosive media through micropores. Figure 13b shows that a small amount of SiO 2 is not capable of effectively filling the pores in epoxy resin, which does not improve its anticorrosion performance. Figure 13c illustrates that when the coating is uniformly dispersed with an appropriate amount of SiO 2 , pores can be effectively plugged and corrosive media are less likely to penetrate. Consequently, the coating performs better as a barrier and is more corrosion resistant. In Figure 13d, the agglomeration of SiO 2 will lead to the decrease of the specific surface area of fillers, resulting in greater defects in the interface between fillers and resin during volume shrinkage during resin curing, and the agglomerated fillers will also reduce the uniformity of the coating. Therefore, excessive water absorption of the composite coating, increased internal stress, and bubbling occur. The composite coating may lose its protective effect in the application process. The coating's protective properties can, therefore, be increased by adding a suitable amount of SiO 2 to the resin. It is possible to block micropores during curing by adding sufficient quantities of SiO 2 to the coating. SiO 2 has the characteristics of a large specific surface area, high specific surface energy, and strong activity, which enable it to facilitate the cross-linking reaction between epoxy resins and improve the coating's compactness and mechanical properties [64][65][66]. It is as a result of this improvement in density that electrolyte permeation is slowed, thus increasing the coating's protection and allowing it to last for a long period of time.

Conclusions
The AC-DC-AC test was performed in this study to accelerate corrosion damage in composite coatings containing micro-SiO2. According to the result, the following conclusions can be drawn: (1) Spherical SiO2 particles with a particle size of 800 ± 50 nm were synthesized by a solution-gel method. (2) By modifying SiO2 with the silane coupling agent KH550, appropriate amounts of SiO2 can be dispersed in water-borne epoxy resins more effectively. (3) The corrosion resistance increases first with an increase in SiO2 content in the composite coating, and then decreases as SiO2 fills the pores formed in the WBE resin coating during curing. It is observed that the composite coating has the best corrosion resistance when the addition of SiO2 amounts to 3 wt.%; it has the lowest impedance value of 9.4 × 10 5 Ω·cm 2 after the AC-DC-AC test; and its adhesion decreases the least before and after the AC-DC-AC test.
Supplementary Materials: The following supporting information can be downloaded at: www.mdpi.com/xxx/s1, Figure S1: Dispersion of SiO2 with different contents in water. Figure

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare that they have no known competing financial interest or personal relationship that could have appeared to influence the work reported in this paper.

Conclusions
The AC-DC-AC test was performed in this study to accelerate corrosion damage in composite coatings containing micro-SiO 2 . According to the result, the following conclusions can be drawn: (1) Spherical SiO 2 particles with a particle size of 800 ± 50 nm were synthesized by a solution-gel method. (2) By modifying SiO 2 with the silane coupling agent KH550, appropriate amounts of SiO 2 can be dispersed in water-borne epoxy resins more effectively. (3) The corrosion resistance increases first with an increase in SiO 2 content in the composite coating, and then decreases as SiO 2 fills the pores formed in the WBE resin coating during curing. It is observed that the composite coating has the best corrosion resistance when the addition of SiO 2 amounts to 3 wt.%; it has the lowest impedance value of 9.4 × 10 5 Ω·cm 2 after the AC-DC-AC test; and its adhesion decreases the least before and after the AC-DC-AC test.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare that they have no known competing financial interest or personal relationship that could have appeared to influence the work reported in this paper.