Four clean beakers were respectively filled with 9.7 g of the mixture of ternary epoxy and quaternary epoxy at a ratio of 1:1, and then four photoinitiators were added with 0.3 g each to obtain the basic test. The results of the photo-DSC testing showed that the four cationic photoinitiators had high photoinitiation rates. As shown in Figure 4, four different cationic photoinitiators (structures shown in Figure 5) (Changzhou Qiangli Company, technology grade) were selected to test their photocuring rate through photo-DSC, and their heat release per unit mass was calculated by integrating the four curves. Photoinitiator-1 is triaryl sulfonium hexafluoroantimonate. Photoinitiator-2 is triaryl sulfonium hexafluorophosphate (single and double salt mixture). Photoinitiator-3 is triaryl sulfonium hexafluorophosphate (single and double salt mixture), and Photoinitiator-4 is diaryliodinium hexafluoroantimonate. It can be seen from Figure 4 that Photoinitiator-4 has the fastest photoinitiation rate and the lowest heat release per unit. Photoinitiator-4 has high photoinitiation efficiency and the curing degree of its system is low. The photocuring rate of Photoinitiator-3 is slightly lower than that of Photoinitiator-4, and the highest heat release per unit mass indicates that the curing degree of the system is the highest. Because the cost of iodonium salt is much higher than that of sulfonium salt, it is difficult to be used in the field of ordinary photocuring, so Photoinitiator-3 is chosen as the photoinitiator. As displayed in Figure 6 and Figure 7, the ultraviolet absorption spectra of four cationic photoinitiators and one anthracene sensitizer were tested. Figure 6 shows that the four cationic photoinitiators all had strong absorption at about 260 nm and 300 nm, on which Photoinitiator-3 had the strongest absorption at about 300 nm. Because the light source UV-LED’s spectral range arrive at 385 nm, and the four photoinitiators have no absorption at 385 nm, an anthracene sensitizer with strong absorption at 385 nm is selected to sensitize the cationic photoinitiator, so as to indirectly increase the maximum absorption wavelength.

As shown in Figure 8 and Figure 9, the effects of different dosage of photoinitiator on curing rate and curing degree were tested. Photoinitiator-3 triaryliodonium hexafluorophosphate (single salt) was used as cationic photoinitiator with the addition ratio of 1% - 4% and active thinner of 96% - 99%, in which the ratio of ternary ring and quaternary ring was in 1:1. As from Figure 8, when the dosage of photoinitiator is 2.5%, the curing rate is the fastest. And when the dosage of photoinitiator is higher than 2.5%, the curing rate decreases, the degree of curing increases slightly either. Therefore, 2.5% dosage was selected as the optimal addition amount.

As shown in Figure 10 and Figure 11, the effects of the dosage of sensitizer on curing rate and degree were tested. 2.5% Photoinitiator-3 triaryliodonium hexafluorophosphate (single salt) was used as cationic photoinitiator, an anthracene compound was used as sensitizer, and the dosage of active thinner was 96.5% - 97.5%. The ratio of ternary and quaternary rings is in 1:1. In Figure 10, it has the fastest reaction rate when the sensitizer is 0.6%, when the amount of sensitizer is greater than 0.6%, the curing rate decreases and the heat release per unit mass increases slightly as well. Considering comprehensively, choose 0.6% as the optimal dosage of sensitizer.

As shown in Figure 12 and Figure 13, the effects of different proportions of oxacyclobutane monomer on curing rate and curing degree were tested. 2.5% Photoinitiator-3 triaryliodonium hexafluorophosphate (single salt) was used as cationic photoinitiator, 0.5% anthracene compound was used as sensitizer, and active diluent was 97%. The ratios of ternary epoxy and quaternary epoxy are 10:0, 8:2, 6:4, 4:6, 2:8, and 0:10, respectively. From Figure 12 and Figure 13, with four-membered epoxy monomer content increasing, the light curing rate is also increasing, when four-membered epoxy monomers and three-membered to speak when the proportion of the curing rate of epoxy monomers fastest, continue to improve the four-membered monomer ratio of epoxy curing rate decreased, comprehensive consider choosing four-membered epoxy: three-membered epoxy = 8:2 as the best added.

Studies have showed that photopolymerization is initiated after a certain amount of heat is given, which can also effectively reduce the induction period with the help of external energy input. Taking 2.5% Photoinitiator-3 triaryliodonium hexafluorophosphate (single salt) as cationic photoinitiator, 0.5% anthracene compounds as sensitizer, 97% as active diluent, and the ratio of quaternary epoxy to ternary epoxy is 8:2. Photo-DSC tests were conducted at temperatures of 20˚C, 40˚C, 60˚C, 80˚C and 100˚C, respectively. The testing results listed in Figure 14 and Figure 15, show that the photocuring rate increases with the increase of curing temperature, and the fastest curing rate is achieved at 80˚C, while the photocuring rate decreases with the temperature increasing. Possibly due to the Photoinitiator-3 cationic photoinitiator will decompose and lose its photoactivity at about 100˚C. Wherefore, when the curing temperature is 80˚C, it has the fastest light curing rate and the highest heat release per unit mass, so 80˚C is selected as the best curing temperature.

Nano-SiO₂ has a strong UV absorption capacity, and it will compete with photoinitiators to absorb UV light. As shown in Figure 16 and Figure 17, photoinitiators cannot generate enough free radicals or protic acid, resulting in low curing degree of the film and affecting the overall performance of the film, so the addition amount of nano-SiO₂ needs to be optimized. Figure 16 and Figure 17 have shown that the different nanometer SiO₂ content light curing rate and the influence of the curing degree. It can be seen from Figure 16 and Figure 17 that with the increasing amount of nano-SiO₂ added, the curing rate and curing degree of the system are constantly decreasing. When the amount of addition is less than 0.4%, the effect on the curing rate and degree of curing is small. Hence, the optimal addition amount of nano-SiO₂ should be less than or equal to 0.4%.

As shown in Table 2 and Figure 18, when the addition amount of nano-SiO₂ is none, the corrosion resistance of the coating is poor, and the corrosion area of the 72-h neutral salt spray experiment is greater than 10%. With the increase of the addition amounts of nano-SiO₂ (0.2% - 1.0%), the white rust area of the coating for 72 h neutral salt spray corrosion is stable in the range of 3% - 4%. Considering the influence of nano-SiO₂ on the curing rate and degree of the coating, 0.2% - 0.4% is selected as the optimal addition amount of nano-SiO₂.

Take a piece of hot aluminum-zinc steel plate with the size of 75 mm × 150 mm × 1 mm. First, smooth the burres on the edge of the steel plate with sandpaper. Then, soak the steel plate in water solution with 2% degreasing agent at room temperature for 10 min before taking it out. Finally, the surface of the steel plate is wiped with acetone, process as shown in Figure 19.

Take a clean small beaker and add 2 - 3 g of epoxy oligomer into it, then add 5 - 6 g of active diluent mixed by 8:2 of quaternary epoxy and ternary epoxy, and then add 0.2 g of nano-SiO₂ dispersion (Shanghai Chemical Company, China). After that, 0.25 g of Photoinitiator-3 cationic photoinitiator and 0.05 g of anthracene sensitizer were added. Finally, 0.5 - 1.5 g of additive was added and stirred for 30 min to obtain the cationic photocurable passivation solution. Passivation solution is evenly coated on the surface of the steel plate with a 1.5 μm extruding coating rod. Thus, a cationic photocurable passivation film is obtained by putting it into a UV-LED exposure machine with an energy of 150 mJ/cm².

In order to meet the application requirements, there is a general requirement for the performance of passivation film in the industry, as shown in Table 3.

The neutral salt spray testing was carried out according to GB 10125-2012 (Wuxi, China). The corrosion solution was 5% NaCl solution, the pH is 6.5 - 7.2, the experimental temperature was 35˚C ± 2˚C, the settlement was 1 - 2 mL per hour, the sample plate size was 75 mm × 150 mm, the surface was flat without oil stains, no damage, and the edge was without burns. The edges of the samples were sealed and protected with transparent tape. After continuous spraying for 72 h, the samples were evaluated according to the corrosion area of the samples. At present, the basic requirement of corrosion resistance of passivation film of commercial hot aluminum-zinc steel plate is that the neutral salt spray corrosion area of 72 h is less than 5%.

The electrochemical performance of the coating is mainly measured by AC impedance spectroscopy (EIS) and Tafel polarization curve, which is an effective method to evaluate the corrosion behavior of the material. The specific device and circuit connection are shown in Figure 20.

The electrochemistry workstation with three electrodes system was used for the testing (Shanghai, China). The photocurable hot aluminum-zinc steel plate was the working electrode, the saturated calomel electrode as the reference electrode, and the platinum plate electrode was the auxiliary electrode. NaCl solution with mass fraction of 3.5% was used as the corrosive medium, and the working area was 1 cm². The testing was carried out at room temperature. AC impedance spectrum testing frequency range was set as 0.01 - 100,000 Hz. Open circuit voltage (OCP) testing was carried out after the circuit connection was completed, and AC impedance spectrum (EIS) testing was carried out after the open circuit voltage (OCP) was stabilized. The experimental data was fitted with Z-View software.

The scanning range of Tafel polarization curve testing was from −1.7 V to −0.5 V, and the scanning rate was 20 mV/s. The testing was carried out at room temperature. Open circuit voltage (OCP) testing was carried out after the circuit connection was completed, and Tafel polarization curve test was carried out after the open circuit voltage (OCP) was stabilized. The experimental data were processed by the software attached to the electrochemical workstation.

Fourier transform infrared spectroscopy (FT-IR) is an important characterization method in the field of organic chemistry. Different functional groups have different absorption peaks in different wavelength ranges. The degree of curing of coatings was characterized by the comparison of the changes of absorption peaks of characteristic functional groups before and after curing. The infrared spectrometer used in this experiment is the IR Tracer-100 infrared spectrometer of Shimadzu Company, Kyoto, Japan, and the scanning range is 4000 cm⁻¹ - 400 cm⁻¹.

Scanning electron microscopy (SEM) can directly image the surface of materials as an effective method to observe the microstructure of materials. Hitachi FlexSEM 1000II (Japan, Tokyo) type scanning electron microscope was used to observe the microstructure of the coating surface, the main magnification was 1000 and 3000 times, respectively, the testing acceleration voltage was 20 kV.

Electrochemical tests were carried out on different surface treatments (fingerprint resistance, chromate passivation, cationic photocurable coating) and untreated hot coated aluminum-zinc steel plates to compare with the corrosion resistance of four different steel plates. The Tafel polarization curve and AC impedance spectrum (EIS) were measured at room temperature using 3.5% neutral NaCl solution as corrosive medium.

1) Tafel polarization curve testing

Figure 22 shows that the comparison of hot-dip aluminum-zinc steel plate, chromate passivation film, fingerprint resistant film and cationic UV-curable coatings from the corrosion potential was increased, and the anode polarization curve and the polarization current of cathodic polarization curve is relatively lower than that of hot-dip aluminum-zinc steel plate, an indicative of the process of cationic UV-curable coatings for corrosion inhibition effect of cathode and anode reactions. The data onto Table 6, electrochemical workstation with the software were obtained from the calculation on the basis of by the table to know, since the cationic UV-curable coatings corrosion current with hot-dip aluminum-zinc steel plate is two orders of magnitude, slightly below the chromate passivation film and commercial fingerprint resistant passivation film, it shows that cationic UV-curable coating corrosion resistance is relatively lower than the chromate passivation film and fingerprint resistant coating but the overall difference.

2) AC impedance spectroscopy (EIS) testing

Figure 23 shows the alternating current impedance spectra (Nyquist) of four samples. It can be seen that the impedance of cationic photocurable coating, chromate passivation film and fingerprint resistant film is greatly improved compared with that of hot aluminum-zinc steel plate without surface treatment, and the impedance value of cationic photocurable coating is about 35 times that of hot aluminum-zinc steel plate, but lower than that of chromate passivation film and fingerprint resistant film on the market, an indicative of the diffusion of the corrosive medium in the cationic photocurable coating being more difficult and the coating has better corrosion resistance. At the same time, the corrosion resistance of the coating is slightly lower than that of chromate passivation film and fingerprint resistant film, which is consistent with the results in Figure 22 and Table 6.

Using 3.5% neutral NaCl solution as a corrosive medium, the cationic photocurable coating plate was soaked in brine. Before soaking, the steel plate was edge sealed with 3M tape. The soaking temperature was room temperature, and the soaking time was 0 h, 6 h, 12 h, 24 h, 48 h and 72 h, respectively. The obtained results were fitted with Z-View software, and the fitted data were plotted by Origin software.

The experimental results show that the cationic photocurable coating exhibits the excellent barrier properties at the initial soaking stage (0 h, 6 h), and its equivalent circuit is shown in Figure 24(a). The corrosive medium slowly penetrates into the surface of the metal substrate, thus causing corrosion on the surface of the metal substrate with an extension of immersion time (12 h, 24 h, 48 h). At this time, the corrosion products will enter the pores of the coating, thus inhibiting the diffusion of the corrosive medium in the coating. The equivalent circuit of this process is shown in Figure 24(b). In the late soaking period (72 h), the coating has basically lost its shielding function, and the corrosive medium can easily diffuse from the coating surface into the surface of the metal matrix. Its equivalent circuit is shown in Figure 24(c).

Generally, the resistance of the coating to corrosive media can be indicated by the magnitude of the alternating impedance modulus of the coating in the low frequency region (0.01 Hz), that is, the corrosion resistance of the coating is better with the larger alternating impedance modulus of the coating at low frequency. It can be seen from Figure 25 and Figure 26 that the barrier of the cationic photocurable coating decreases gradually with the extension of immersion time in 3.5% NaCl solution, which indicates that the corrosive medium penetrates from the coating surface to the surface of the metal substrate and causing corrosion of the metal substrate with the extension of immersion time. According to the alternating current impedance spectra (Nyquist), it can be found that the low-frequency alternating impedance modulus of the coating decreased slightly at the beginning of soaking (0 h, 6 h), but the change was small, and its frequency/phase diagram basically did not change. This indicates that the coating has a good barrier to the corrosive medium in this time. In the middle soaking period (12 h, 24 h, 48 h), the low frequency AC impedance modulus of the coating decreased significantly compared with the initial soaking period. However, the low frequency AC impedance modulus of the coating decreased to a small extent during 12 - 48 h. At this stage, the corrosive medium has diffused from the coating surface to the surface of the metal matrix. At this time, the corrosion

products will fill the pores of the coating, thus inhibiting the continuous corrosion. It can also be seen from Nyquist diagram that the arc radius of the coating does not change much after soaking for 12 h, 24 h and 48 h, indicating that the coating still has good protective performance after soaking for 48 h. At the latest soaking stage (72 h), the low-frequency AC impedance modulus of the coating decreased sharply, indicating that the structure of the coating had been seriously damaged at this stage, but the low-frequency AC impedance modulus of the coating was still higher than that of the hot aluminum-zinc steel plate.