Study on the Preparation of High-Temperature Resistant and Electrically Insulating h-BN Coating in Ethanol Solution by Electrophoretic Deposition

A hexagonal boron nitride (h-BN) coating of micron thickness is deposited directly on 316L stainless steel (SS316L) cathode through efficient, adjustable electrophoretic deposition (EPD) in a suspension system containing surfactant and ethanol. It is based on the mixing of h-BN with polyethyleneimine (PEI) resulting in positively charged ceramic powder making cathodic electrophoretic deposition possible. The thickness of the resulting h-BN coatings deposited on SS316L could be controlled by varying the time and the voltage of electrophoretic deposition. The deposition kinetics and mechanism have been discussed. After soaking in Al(H2PO4)3 solution and high-temperature annealing, the h-BN coatings exhibited good adhesive strength. Furthermore, a novel method has been used for the evaluation of the adhesive strength to explore the appropriate experimental conditions. X-ray diffraction spectroscopy (XRD), Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) were employed to characterize the h-BN coatings. The h-BN coatings are applied for the DC breakdown performance test and exhibit remarkable breakdown voltage and breakdown strength.


Introduction
In response to the ever-increasing climate change, the use and transmission of clean energy have become a primary focus for the industry and the scientific community. As a form of energy that is widely used and environmentally friendly, electricity gives great hope for a better future for mankind. There has been increasing interest in long-distance power transmission equipment. Electrical steel with high magnetic induction and mechanical properties is an important component of transformers and electronic components [1][2][3]. To reduce eddy current losses, the coatings have been applied to the surface of electrical steel sheets to prolong the service life of the device [4,5]. Such coatings should exhibit good adhesion and insulation properties to meet the requirements of equipment for components. The heat resistance and heat dissipation of coating material, furthermore, is considered to pose a great deal of help to the use of electrical steel due to overheating issues at high voltage [6,7].
In the previous studies, several different methods have been developed, such as the solgel method [8], spray method [9], dipping method [10], vapor deposition method [11] and so on. While the traditional dipping method for industrial production is dominating the market, they suffer from many drawbacks originating from non-uniformity of the coating and high energy consumption. Besides, large-scale applications of the vapor deposition method are not feasible due to their complexity and high cost. These methods, such as spraying and the sol-gel method, cannot enable efficient control over the structure of the coatings and rate of qualified products at a low processing cost, and cannot be employed extensively in the coating industries. In recent years, DC electrophoretic deposition (EPD) has attracted tremendous research interest as an efficient preparation method. Compared with the several methods introduced before, it has the advantages of simpler equipment, more convenient operation and the ability to control the thickness of the coating. EPD can not only be applied to the suspension of ceramic particles but can also be used to deposit polymers [12], graphene [13], and biological materials [14]. Prior to EPD, various powders are suspended in an electrophoretic solvent containing surfactants to form a colloidal solution. The EPD process drives particulate materials in a solvent towards substrate with an applied electric field [15,16]. Although the preparation of the coating can be controlled by the deposition voltage and time, the EPD process still needs to be carried out in a solvent with a higher decomposition voltage. In this system, surfactants that provide effective surface charge to the particles have also been considered in detail [17]. These considerations become particularly important.
Hexagonal boron nitride (h-BN), an artificially prepared material with a two-dimensional (2D) nanostructure, has attracted much attention in ceramic coating research [18,19]. In a two-dimensional plane, boron atoms and nitrogen atoms are covalently bonded; van der Waals forces exist in these two plane layers [20]. It has been widely recognized as a promising ceramic material for coatings as it carries many superior properties, including exceptional thermal stability, optimal anti-oxidation, and excellent electrical insulation properties [21]. The reason why h-BN is used as the main component of the coating is that: (1) the material itself has good insulation and oxidation resistance, which is beneficial to be used under more severe conditions; (2) the excellent thermal conductivity of h-BN can ensure heat dissipation and avoid damage to electrical steel; and (3) the excellent mechanical properties of h-BN are conducive to various complex processing [22].
The h-BN is being widely explored because of its layered structure and special surface properties. However, the surface of boron nitride is too stable to have a binding site to bond with a functional group [23,24]. To overcome this problem, surfactants have been used to modify the surface of boron nitride to enhance the wettability, dispersibility, and charge on the particle surface in the solvent. According to related literature [25], we found that common cationic surfactants with only a single charge can hardly meet the above requirements. Therefore, polyethyleneimine (PEI) has become a suitable choice, which owes its advantages to a large number of amino groups and good solubility in ethanol.
To reduce the eddy current loss caused by hard magnetic materials in use, herein, we report the facile electrophoretic deposition of hexagonal boron nitride on 316L stainless steel as an insulation coating. We found that combining two low-cost materials, hexagonal boron nitride and phosphate, it is possible to obtain an insulation coating with controllable thickness and high-temperature resistance. We produced coatings with good adhesion and breakdown properties. The direct deposition of hexagonal boron nitride 316L stainless steel would provide suitable structure and porosity, thus facilitating the diffusion of phosphate ions, improving the tightness of the coating, the adhesion between the particles and the insulation performance.

Preparation of Solution
A total of 0.1 g of h-BN was dispersed in 50 mL of ethanol with different pH and dispersed through ultrasonication for 0.5 h. These suspensions of h-BN were magnetically stirred for 12 h. Acetic acid (99.9%, Sinopharm Group, Shanghai, China) and sodium hydroxide (99.6%, Sinopharm Group, Shanghai, China) were used as the pH regulators of the suspension. The digital display pH meter was used as an indicator for pH adjustment. A series of ethanol suspension containing 0.1 g of h-BN powder and polyethyleneimine (PEI, 99.9%, M.W.10000) (Aladdin, Shanghai, China) was prepared. The concentrations of PEI varied from 0.1 to 0.8 g/L. The pH of these suspensions was adjusted to 4 with acetic acid. These suspensions were dispersed through ultrasonication for 0.5 h and magnetically stirred for 12 h. The supernatant of all of the suspensions was taken out for the measurement of zeta potential after standing for 12 h.
A suspension containing 2 g/L of h-BN, 0.4 mol/L of PEI and 500 mL of ethanol was adjusted to pH = 4 with acetic acid. The suspension was used for electrophoretic deposition of the insulating coating after dispersion for 0.5 h and magnetic stirring for 12 h.
The pH of the suspension was measured with a pH meter (Leici PHS-3E, Shanghai, China) at 25 • C. Due to the lack of standard buffer solutions for ethanol three aqueous standards of pH = 4.00, 6.86 and 9.18 were used for standardization. Thus, the so-called "operational pH values" of a non-aqueous suspension can be directly read by a pH meter [26].
where pH is the pH meter reading, and pa H is the negative logarithm of the proton activity in a non-aqueous suspension. The ∆E j is the residual liquid-junction potential. For the ethanol suspension, (∆E j /0.05916) = 1.23. [26]. In this paper, the uncorrected pH read by the pH meter was used for convenience.

Preparation of h-BN-PEI Samples
A total of 20 mL of h-BN suspension for deposition was taken out and centrifuged for 10 min at a speed of 8000 r/min. After centrifugation, the precipitate was washed with ethanol and centrifuged several times. The processed precipitate was dried at 50 • C for 12 h, and the dried sample was used to test.

Surface Treatment of the Substrate
Electrodes consisted of 40 × 10 × 0.01 mm SS316L sheet were polished with sandpaper with a grain size of 1000 and 2000, and were electrochemically degreased in an alkaline solution at 70 • C for 10 min. These electrodes were thoroughly washed with deionized water, dried and weighed on a microbalance. Polyimide tapes were used to encapsulate these electrodes, and the same area of 20 mm × 10 mm was exposed.

Electrophoretic Deposition and Post-Processing of Sample
The EPD process parameters were given in Table 1. SS316L was used as the working electrode and counter electrode. The prepared h-BN suspension was ultrasonicated for 30 min before use. The two electrodes were placed parallel to each other in a two-electrode cell and were separated by a distance of 10 mm. After depositing the coatings at 25 • C, the samples were placed in a drying oven followed by drying at 50 • C for 12 h. The dried samples with the packaging tape removed were weighed on an analytical balance. Subsequently, the coating samples were placed in the prepared aluminum dehydration phosphate (Al(H 2 PO 4 ) 3 , 99.9%, Aladdin, Los Angeles, CA, USA) aqueous solution and soaked at 25 • C for 12 h. The sample was placed in a tube furnace and heat-treated in a nitrogen atmosphere after the immersion. The samples were annealed at 100, 200 and 400 • C, respectively, for 2 h, using a heating rate of 5 • C/min.

Characterization
All prepared samples of h-BN suspension were used to determine the zeta potential of the particles in ethanol at 25 • C. Each sample was tested three times with a zeta potential meter (Nano ZS, Malvern, UK) and the average was taken.
XRD analysis was performed on the powder and samples to determine the substance and showed the crystal phase state of the coating surface. XRD patterns were obtained with an X-ray diffractometer (D8-Focus, Bruker, Germany) with CuKα radiation (λ = 1.5419 Å), in the 2θ range of 20 to 60 • with a scan speed of 4 • /min.
Coatings and suspension were analyzed by Fourier infrared transform spectroscopy (Nicolet 6700, Nicoli, MA, USA), with potassium bromide as a reference, in the wavenumber range of 500 to 4000.
The morphology of all coatings was analyzed by using a scanning electron microscope (SEM), and the elements were analyzed by energy dispersive spectroscopy (EDS). The coatings were observed by using a projection electron microscope.
An ultrasonic oscillation (KQ-300B, Jiangsu, China) was employed to evaluate the adhesive strength (Ad %) between the coating and the steel substrate. The sintered samples ultrasonically oscillated for 30 min in deionized water at 300 W, 40 kHz, followed by drying and weighing. The expulsion rate (Er %) and adhesive strength of the coating were then calculated by the following equations [27].
where M 1 (g) and M 2 (g) are the quality of samples before and after ultrasonic treatment, respectively. In addition, the cross-test was used to test the adhesion strength of the coating prepared under different conditions according to the ASTM D3359-09-B. Several coating samples were deposited at 30 V and 10 min, and then immersed (0.2 mol/L Al(H 2 PO 4 ) 3 ) and annealed. Next, the samples were annealed in an environment filled with nitrogen and at different temperatures, with a heating rate of 5 • C/min.
A digital micrometer (AWT-CHY01, Henan, China) was used to test the thickness of the sample. Five points were measured for each sample, and the average value was taken. With a boosted rate of 0.1 kV/s, a breakdown voltage tester (DDJT-50kV, Beijing, China) was used to test the breakdown voltage of the processed samples.

Zeta Potential
In this work, the surface electrical characteristic of the h-BN particles in ethanol is also investigated by zeta potential measurements, which changes as a function of pH values of the dispersion medium. As shown in Figure 1, the isoelectric point is achieved in the dispersion medium in the pH interval of 5 to 6. When the pH is lower than the isoelectric point, then the positive zeta potential of the h-BN particles increases with the decrease of pH, which is mainly caused by the adsorption of protons by h-BN. However, when the pH is higher than the isoelectric point, the negative zeta potential of the h-BN particles decreases with the increase of pH. It may be ascribed to the introduction of acetate or hydroxide group. Generally, if the absolute value of the zeta potential of the particles is greater than 30 mV, the suspension is considered to be stable for a long time. As mentioned in Figure 1, the suspension cannot remain stable by adjusting the pH, requiring the addition of an appropriate amount of PEI. To further identify the impact of PEI concentration at pH 4, the surface charge conditions of the h-BN/PEI particles were monitored by the zeta potential. The result of zetapotential charge analysis of the h-BN/PEI particles is shown in Figure 2. The results indicate that the zeta potential increases with the concentration of PEI from 0.1 to 0.4 mol/L, and then decreases with the concentration higher than 0.4 mol/L. The highest potential of the h-BN/PEI particles reaches 38.1 mV. The h-BN/PEI particles show an extensive positive zeta potential over a suitable PEI concentration range, which is predominantly due to the amine groups carried by PEI. Through the acid-base interaction, large amounts of PEI molecules are better accommodated on the surface of the h-BN particle, which contains electron-deficient boron atoms [28]. As mentioned in Figure 3, the h-BN/PEI particles contain large quantities of residual amine or imine groups liable to be protonated under acidic conditions, thus causing the particles to carry a highly positive charge density [29]. However, an excessively high PEI concentration can cause bridging effects between the boron nitride particles, resulting in a decrease in zeta potential.  To further identify the impact of PEI concentration at pH 4, the surface charge conditions of the h-BN/PEI particles were monitored by the zeta potential. The result of zeta-potential charge analysis of the h-BN/PEI particles is shown in Figure 2. The results indicate that the zeta potential increases with the concentration of PEI from 0.1 to 0.4 mol/L, and then decreases with the concentration higher than 0.4 mol/L. The highest potential of the h-BN/PEI particles reaches 38.1 mV. The h-BN/PEI particles show an extensive positive zeta potential over a suitable PEI concentration range, which is predominantly due to the amine groups carried by PEI. Through the acid-base interaction, large amounts of PEI molecules are better accommodated on the surface of the h-BN particle, which contains electron-deficient boron atoms [28]. As mentioned in Figure 3, the h-BN/PEI particles contain large quantities of residual amine or imine groups liable to be protonated under acidic conditions, thus causing the particles to carry a highly positive charge density [29]. However, an excessively high PEI concentration can cause bridging effects between the boron nitride particles, resulting in a decrease in zeta potential. To further identify the impact of PEI concentration at pH 4, the surface charge conditions of the h-BN/PEI particles were monitored by the zeta potential. The result of zetapotential charge analysis of the h-BN/PEI particles is shown in Figure 2. The results indicate that the zeta potential increases with the concentration of PEI from 0.1 to 0.4 mol/L, and then decreases with the concentration higher than 0.4 mol/L. The highest potential of the h-BN/PEI particles reaches 38.1 mV. The h-BN/PEI particles show an extensive positive zeta potential over a suitable PEI concentration range, which is predominantly due to the amine groups carried by PEI. Through the acid-base interaction, large amounts of PEI molecules are better accommodated on the surface of the h-BN particle, which contains electron-deficient boron atoms [28]. As mentioned in Figure 3, the h-BN/PEI particles contain large quantities of residual amine or imine groups liable to be protonated under acidic conditions, thus causing the particles to carry a highly positive charge density [29]. However, an excessively high PEI concentration can cause bridging effects between the boron nitride particles, resulting in a decrease in zeta potential.

XRD Patterns
The structure of the h-BN particles, SS316L and coating sample was also characterized by XRD patterns, as shown in Figure 4. The five characteristic peaks displayed by the h-BN particles correspond to the diffraction peaks in the standard spectrum of h-BN. The coating sample displays a diffraction peak at the left end of the spectrum, which corresponds to (002) planes of the hexagonal crystal of h-BN [30]. It suggests that the structure of the h-BN particles is not destroyed after electrophoretic deposition. The remaining two diffraction peaks of the coating sample correspond to the two diffraction peaks of SS316L, indicating that the stainless steel is not affected by the deposited coatings.

FTIR Spectra
To ensure the successful adsorption of PEI into the h-BN particles, the FTIR spectra of PEI, the h-BN particles and the h-BN-PEI sample were collected, as shown in Figure 5. The spectrum of the h-BN particles with no adsorption shows predominant peaks at 1388 cm-1 and 804 cm-1. These are two typical peaks that are caused by the in-plane B-N transverse optical modes of the sp2 bonded BN and the B-N-B out-of-plane bending vibration [31]. Compared to the h-BN particles, new peaks of the h-BN-PEI sample at 3392 cm −1 ,

XRD Patterns
The structure of the h-BN particles, SS316L and coating sample was also characterized by XRD patterns, as shown in Figure 4. The five characteristic peaks displayed by the h-BN particles correspond to the diffraction peaks in the standard spectrum of h-BN. The coating sample displays a diffraction peak at the left end of the spectrum, which corresponds to (002) planes of the hexagonal crystal of h-BN [30]. It suggests that the structure of the h-BN particles is not destroyed after electrophoretic deposition. The remaining two diffraction peaks of the coating sample correspond to the two diffraction peaks of SS316L, indicating that the stainless steel is not affected by the deposited coatings.

XRD Patterns
The structure of the h-BN particles, SS316L and coating sample was also characterized by XRD patterns, as shown in Figure 4. The five characteristic peaks displayed by the h-BN particles correspond to the diffraction peaks in the standard spectrum of h-BN. The coating sample displays a diffraction peak at the left end of the spectrum, which corresponds to (002) planes of the hexagonal crystal of h-BN [30]. It suggests that the structure of the h-BN particles is not destroyed after electrophoretic deposition. The remaining two diffraction peaks of the coating sample correspond to the two diffraction peaks of SS316L, indicating that the stainless steel is not affected by the deposited coatings.

FTIR Spectra
To ensure the successful adsorption of PEI into the h-BN particles, the FTIR spectra of PEI, the h-BN particles and the h-BN-PEI sample were collected, as shown in Figure 5. The spectrum of the h-BN particles with no adsorption shows predominant peaks at 1388 cm-1 and 804 cm-1. These are two typical peaks that are caused by the in-plane B-N transverse optical modes of the sp2 bonded BN and the B-N-B out-of-plane bending vibration [31]. Compared to the h-BN particles, new peaks of the h-BN-PEI sample at 3392 cm −1 ,

FTIR Spectra
To ensure the successful adsorption of PEI into the h-BN particles, the FTIR spectra of PEI, the h-BN particles and the h-BN-PEI sample were collected, as shown in Figure 5. The spectrum of the h-BN particles with no adsorption shows predominant peaks at 1388 cm −1 and 804 cm −1 . These are two typical peaks that are caused by the in-plane B-N transverse optical modes of the sp2 bonded BN and the B-N-B out-of-plane bending vibration [31]. Compared to the h-BN particles, new peaks of the h-BN-PEI sample at 3392 cm −1 , 3320 cm −1 , 2919 cm −1 and 2848 cm −1 , representing -NH 2 [32] and -CH 2 group [33], indicate that PEI was successfully adsorbed on the h-BN particles.
Processes 2021, 9,871 7 of 17 3320 cm −1 , 2919 cm −1 and 2848 cm −1 , representing -NH2 [32] and -CH2 group [33], indicate that PEI was successfully adsorbed on the h-BN particles.  Figure 6 shows the mass-per-unit area of the PEI modified h-BN coatings deposited onto stainless steel after EPD at different voltages. The deposition process for the EPD can be seen in Figure 7. While the mass per unit area is continuously increasing at fixed voltage as the deposition time increases, the slope of the deposition curve is gradually decreasing. It is attributed to the shielding effect of the deposited coating on the electric field and the h-BN particles with slowed migration. Furthermore, the mass per unit area increases with the increase of the deposition voltage in the same time frame due to the faster migration and deposition of particles at the higher voltage.   Figure 6 shows the mass-per-unit area of the PEI modified h-BN coatings deposited onto stainless steel after EPD at different voltages. The deposition process for the EPD can be seen in Figure 7. While the mass per unit area is continuously increasing at fixed voltage as the deposition time increases, the slope of the deposition curve is gradually decreasing. It is attributed to the shielding effect of the deposited coating on the electric field and the h-BN particles with slowed migration. Furthermore, the mass per unit area increases with the increase of the deposition voltage in the same time frame due to the faster migration and deposition of particles at the higher voltage.

Weight Deposition Curves
Processes 2021, 9,871 7 of 17 3320 cm −1 , 2919 cm −1 and 2848 cm −1 , representing -NH2 [32] and -CH2 group [33], indicate that PEI was successfully adsorbed on the h-BN particles.  Figure 6 shows the mass-per-unit area of the PEI modified h-BN coatings deposited onto stainless steel after EPD at different voltages. The deposition process for the EPD can be seen in Figure 7. While the mass per unit area is continuously increasing at fixed voltage as the deposition time increases, the slope of the deposition curve is gradually decreasing. It is attributed to the shielding effect of the deposited coating on the electric field and the h-BN particles with slowed migration. Furthermore, the mass per unit area increases with the increase of the deposition voltage in the same time frame due to the faster migration and deposition of particles at the higher voltage.   where W is the mass deposition, μ is the particle electrophoretic mobility, E is the electric field, t is the deposition time, A is the surface area of the electrode, and Cs is the concentration of colloidal particles in suspension. The changes in the particle concentration and conductivity caused by long-term deposition can limit the conditions of use of this equation. causing the mass curve at the same voltage not to follow the Hamarck equation.

SEM and EDS
SEM was used to investigate the surface morphology of the h-BN coatings without the added binder. Figure 8 shows the SEM images of the h-BN coatings prepared under different deposition conditions via EPD. With the increase of voltage within the same EPD time, the coating becomes more uneven and cracks are also visible, as shown in Figure 8a-c. On the microscale, the coating deposited at higher voltage is more prone to contain more pores. This is attributed to the fact that the migrated flake particles are deposited on the SS316L cathode with a non-compact spatial state at the higher voltage. The increase of porosity in the coating leads to the appearance of continuous cracks. The SEM images of the coatings with different magnifications displayed in Figure 8d-f show that the coatings deposited during different times at the same voltage. Obviously, it can be observed from the images that the unevenness of the coating surface becomes more obvious with increasing deposition times. Besides, the membrane surface exhibits a non-dense structure with some aggregates of the h-BN particles, which is mainly attributed to the shielding effect of deposited coating and shedding of particles with non-compact deposition for a long time. Figure 9 shows more related mechanisms in detail. As seen from Figure 10, the coating thickness decreases as the voltage and time increase. Repeated measurements of deposition thickness were carried out in all samples and the same results were observed. The tightness of the coating is thought to be related to EPD voltage and time too. Therefore, we can consider that the relevant explanation proposed has been proved. Increasing the voltage from 15 to 50 V resulted in a higher deposition yield, which is in rough agreement with the prediction of Hamaker's equation [34]: where W is the mass deposition, µ is the particle electrophoretic mobility, E is the electric field, t is the deposition time, A is the surface area of the electrode, and C s is the concentration of colloidal particles in suspension. The changes in the particle concentration and conductivity caused by long-term deposition can limit the conditions of use of this equation. causing the mass curve at the same voltage not to follow the Hamarck equation.

SEM and EDS
SEM was used to investigate the surface morphology of the h-BN coatings without the added binder. Figure 8 shows the SEM images of the h-BN coatings prepared under different deposition conditions via EPD. With the increase of voltage within the same EPD time, the coating becomes more uneven and cracks are also visible, as shown in Figure 8a-c. On the microscale, the coating deposited at higher voltage is more prone to contain more pores. This is attributed to the fact that the migrated flake particles are deposited on the SS316L cathode with a non-compact spatial state at the higher voltage. The increase of porosity in the coating leads to the appearance of continuous cracks. The SEM images of the coatings with different magnifications displayed in Figure 8d-f show that the coatings deposited during different times at the same voltage. Obviously, it can be observed from the images that the unevenness of the coating surface becomes more obvious with increasing deposition times. Besides, the membrane surface exhibits a non-dense structure with some aggregates of the h-BN particles, which is mainly attributed to the shielding effect of deposited coating and shedding of particles with non-compact deposition for a long time. Figure 9 shows more related mechanisms in detail. As seen from Figure 10, the coating thickness decreases as the voltage and time increase. Repeated measurements of deposition thickness were carried out in all samples and the same results were observed. The tightness of the coating is thought to be related to EPD voltage and time too. Therefore, we can consider that the relevant explanation proposed has been proved.     Figure 11 features the surface SEM images of annealed h-BN coatings soaked with Al(H2PO4)3 via the different deposition conditions. It can be observed from the image that the morphology of the treated coatings still follows the mechanism previously discussed. Comparing to the pristine coatings, Figure 11 shows the apparent distinction in the surface modified with post-treatment. The coating surface morphology with fewer and smaller pores confirms the diffusion of Al(H2PO4)3 into the coating. When the h-BN coatings are soaked, ions dissociated can easily diffuse to the coatings. The metaphosphates formed under annealing conditions seals the pores and increases the compactness of the coating [35]. The element uniformity in the post-treated h-BN coating is shown by EDS maps presented in Figure 12a, with the bright spots indicating the presence of some related, respectively. As shown in Figure 12b, it is apparent that the two elements of phosphorus and  Figure 11 features the surface SEM images of annealed h-BN coatings soaked with Al(H 2 PO 4 ) 3 via the different deposition conditions. It can be observed from the image that the morphology of the treated coatings still follows the mechanism previously discussed. Comparing to the pristine coatings, Figure 11 shows the apparent distinction in the surface modified with post-treatment. The coating surface morphology with fewer and smaller pores confirms the diffusion of Al(H 2 PO 4 ) 3 into the coating. When the h-BN coatings are soaked, ions dissociated can easily diffuse to the coatings. The metaphosphates formed under annealing conditions seals the pores and increases the compactness of the coating [35].  Figure 11 features the surface SEM images of annealed h-BN coatings soaked with Al(H2PO4)3 via the different deposition conditions. It can be observed from the image that the morphology of the treated coatings still follows the mechanism previously discussed. Comparing to the pristine coatings, Figure 11 shows the apparent distinction in the surface modified with post-treatment. The coating surface morphology with fewer and smaller pores confirms the diffusion of Al(H2PO4)3 into the coating. When the h-BN coatings are soaked, ions dissociated can easily diffuse to the coatings. The metaphosphates formed under annealing conditions seals the pores and increases the compactness of the coating [35]. The element uniformity in the post-treated h-BN coating is shown by EDS maps presented in Figure 12a, with the bright spots indicating the presence of some related, respectively. As shown in Figure 12b, it is apparent that the two elements of phosphorus and The element uniformity in the post-treated h-BN coating is shown by EDS maps presented in Figure 12a, with the bright spots indicating the presence of some related, respectively. As shown in Figure 12b, it is apparent that the two elements of phosphorus and aluminum are dispersed uniformly. The molar ratio of aluminum and phosphorus is 0.51/2.64, although it does not completely match the ratio of the two elements in Al(H 2 PO 4 ) 3 . These element distribution maps prove that Al(H 2 PO 4 ) 3 diffuses successfully into the coating.
aluminum are dispersed uniformly. The molar ratio of aluminum and phosphorus is 0.51/2.64, although it does not completely match the ratio of the two elements in Al(H2PO4)3. These element distribution maps prove that Al(H2PO4)3 diffuses successfully into the coating.  Figure 13 shows the effect of the Al(H2PO4)3 concentration on the adhesive strength of the post-treated coatings under the same deposition, annealing and test conditions. The Al(H2PO4)3 solution entering the coating can form metaphosphates as a binder at high temperature to strengthen the adhesion between the stainless steel and the coating [36,37]. As observed in Figure 13a, the adhesive strength increases gradually with the increasing Al(H2PO4)3 concentration from 0.05 to 0.2 mol/L and then decreases with the concentration from 0.2 to 0.4 mol/L. It could be seen that the adhesive strength of the coating soaked in a 0.2 mol/L solution is the highest, presenting the least damage on the coating surface in Figure 12b,c. Compared to the coating with the best adhesion, it is clear that the coatings soaked in 0.05 and 0.1 mol/L solutions show obvious cracking and peeling since a lower concentration solution could diffuse ions into the coating with a lower rate. Moreover, the coatings soaked in a solution higher than 0.2 mol/L still showed slight peeling off after the adhesion test. This peeling off could be caused by a solution with higher viscosity that slows down the diffusion of ions and decreases the content of metaphosphate.  Figure 13 shows the effect of the Al(H 2 PO 4 ) 3 concentration on the adhesive strength of the post-treated coatings under the same deposition, annealing and test conditions. The Al(H 2 PO 4 ) 3 solution entering the coating can form metaphosphates as a binder at high temperature to strengthen the adhesion between the stainless steel and the coating [36,37]. As observed in Figure 13a, the adhesive strength increases gradually with the increasing Al(H 2 PO 4 ) 3 concentration from 0.05 to 0.2 mol/L and then decreases with the concentration from 0.2 to 0.4 mol/L. It could be seen that the adhesive strength of the coating soaked in a 0.2 mol/L solution is the highest, presenting the least damage on the coating surface in Figure 12b,c. Compared to the coating with the best adhesion, it is clear that the coatings soaked in 0.05 and 0.1 mol/L solutions show obvious cracking and peeling since a lower concentration solution could diffuse ions into the coating with a lower rate. Moreover, the coatings soaked in a solution higher than 0.2 mol/L still showed slight peeling off after the adhesion test. This peeling off could be caused by a solution with higher viscosity that slows down the diffusion of ions and decreases the content of metaphosphate.   Figure 14a, the adhesive strength of the coatings decreases distinctly as the deposition voltage increases, mainly owing to uneven and loose structures of the post-treated coating caused by fast migration of the flake h-BN particle at a higher voltage. The process of soaking and annealing cannot repair the loose coating structure, leading to the cracking and peeling-off of the edges of the coating after the test, as shown in Figure 14b,c. Figure 15a shows the relationship between the adhesive strength of the coating and the EPD time. The corresponding microscope images of the coatings after the adhesion test are shown in Figure 15b. The post-treated coatings deposited at 30 V for different times were soaked in a 0.2 mol/L Al(H2PO4)3 solution under the same annealing conditions. It could be seen from Figure 15a that the adhesive strength of the coatings deposited for 8 min is closer to the other one deposited for 10 min. As the deposition time increases, the adhesive strength of the coating decreases. The adhesive strength reaches the highest value at 10 min, which is 83.9%. Since the thick coating could inhibit the migration of h-BN particles in an electric field, some nanoparticles were deposited with an irregular state and formed a loose structure on the surface of the coating. The decreased adhesive strength caused by a large number of loose structures results in the peeling-off of the coating after the test, as observed in Figure 15b,c.   Figure 14a, the adhesive strength of the coatings decreases distinctly as the deposition voltage increases, mainly owing to uneven and loose structures of the post-treated coating caused by fast migration of the flake h-BN particle at a higher voltage. The process of soaking and annealing cannot repair the loose coating structure, leading to the cracking and peeling-off of the edges of the coating after the test, as shown in Figure 14b,c. Figure 15a shows the relationship between the adhesive strength of the coating and the EPD time. The corresponding microscope images of the coatings after the adhesion test are shown in Figure 15b. The post-treated coatings deposited at 30 V for different times were soaked in a 0.2 mol/L Al(H 2 PO 4 ) 3 solution under the same annealing conditions. It could be seen from Figure 15a that the adhesive strength of the coatings deposited for 8 min is closer to the other one deposited for 10 min. As the deposition time increases, the adhesive strength of the coating decreases. The adhesive strength reaches the highest value at 10 min, which is 83.9%. Since the thick coating could inhibit the migration of h-BN particles in an electric field, some nanoparticles were deposited with an irregular state and formed a loose structure on the surface of the coating. The decreased adhesive strength caused by a large number of loose structures results in the peeling-off of the coating after the test, as observed in Figure 15b,c.

Heat Resistance Test of Coating
The uniform and crack-free h-BN coatings surfaces were obtained by EPD. To determine that the coating can withstand high temperatures, the soaked sample were annealed

Heat Resistance Test of Coating
The uniform and crack-free h-BN coatings surfaces were obtained by EPD. To determine that the coating can withstand high temperatures, the soaked sample were annealed

Heat Resistance Test of Coating
The uniform and crack-free h-BN coatings surfaces were obtained by EPD. To determine that the coating can withstand high temperatures, the soaked sample were annealed with an appropriate heating rate. Figure 16 shows the surface of the boron nitride coating at different temperatures. There are no obvious cracks and color changes on the surface of the annealed h-BN coatings. Al(H 2 PO 4 ) 3 forms different crystalline metaphosphates and phosphates with strong adhesion at higher temperatures. The properties of the two materials determine that the coating can withstand high temperatures.
Processes 2021, 9,871 14 of 17 with an appropriate heating rate. Figure 16 shows the surface of the boron nitride coating at different temperatures. There are no obvious cracks and color changes on the surface of the annealed h-BN coatings. Al(H2PO4)3 forms different crystalline metaphosphates and phosphates with strong adhesion at higher temperatures. The properties of the two materials determine that the coating can withstand high temperatures.

Breakdown Test of Coating
When the applied voltage is as high as a certain value, the dielectric loss or breakdown occurs due to the inability of the electrically insulating material to withstand extreme conditions. This voltage is called the breakdown voltage. The breakdown strength of an insulating coating is defined as the ratio of the breakdown voltage to the thickness of the coating. Figure 17 shows the effect of EPD time on the thickness, breakdown voltage and breakdown strength of the coating. The post-treated coatings deposited at 30 V for different times were soaked in a 0.2 mol/L Al(H2PO4)3 solution under the same annealing conditions. With an increase in EPD time, the thickness of the post-treated coating increases. The breakdown voltage increases with the EPD time from 8 to 10 min, and then decreases with the EPD time from 10 to 16 min. The breakdown voltage of the coating under the EPD time of 10 min is the highest, which is 1.22 kV. Under a long period of deposition, the loose structure is formed inside the coating due to the irregular deposition of h-BN particles. This is unfavorable for the formation of a dense insulating medium with high voltage resistance. Based on the above two sets of data, we can obtain the relationship between the EPD time and the breakdown strength of the coating. The results indicate that the breakdown strength decreases with an increase in EPD time. The coating prepared under the EPD time of 10 min can achieve an excellent balance between suitable thickness and compact structure, showing the highest breakdown voltage and great breakdown strength.

Breakdown Test of Coating
When the applied voltage is as high as a certain value, the dielectric loss or breakdown occurs due to the inability of the electrically insulating material to withstand extreme conditions. This voltage is called the breakdown voltage. The breakdown strength of an insulating coating is defined as the ratio of the breakdown voltage to the thickness of the coating. Figure 17 shows the effect of EPD time on the thickness, breakdown voltage and breakdown strength of the coating. The post-treated coatings deposited at 30 V for different times were soaked in a 0.2 mol/L Al(H 2 PO 4 ) 3 solution under the same annealing conditions. With an increase in EPD time, the thickness of the post-treated coating increases. The breakdown voltage increases with the EPD time from 8 to 10 min, and then decreases with the EPD time from 10 to 16 min. The breakdown voltage of the coating under the EPD time of 10 min is the highest, which is 1.22 kV. Under a long period of deposition, the loose structure is formed inside the coating due to the irregular deposition of h-BN particles. This is unfavorable for the formation of a dense insulating medium with high voltage resistance. Based on the above two sets of data, we can obtain the relationship between the EPD time and the breakdown strength of the coating. The results indicate that the breakdown strength decreases with an increase in EPD time. The coating prepared under the EPD time of 10 min can achieve an excellent balance between suitable thickness and compact structure, showing the highest breakdown voltage and great breakdown strength.

Conclusions
We have combined the EPD, inorganic binder and sintering process to prepare h-BN coating on SS316L stainless steel substrate to make up for the problems of traditional insulating coating. After the inorganic binder immersion and annealing treatment, the adhesion between the insulating coating and the stainless steel substrate increases. The h-BN coating deposited at 30 V for 10 min has a uniform surface, a suitable thickness and a relatively compact structure. Through the corresponding experiment, the appropriate concentration of the inorganic binder was selected, which was 0.2 mol/L. Through the adhesion test, the coatings after soaking and sintering showed good adhesion, and the adhesive strength was 83.9%. In addition to having good heat resistance, the coating prepared under this condition also has good insulation, with the highest breakdown voltage of 1.22 kV and excellent breakdown strength of 0.254 kV/mm. In summary, this method can shorten the preparation cycle of the h-BN coatings, thereby obtaining a thicker and denser ceramic coating, with ideal and controllable performance, and showing the potential for insulation applications. This method can be used to prepare high-performance inorganic ceramic materials.

Conclusions
We have combined the EPD, inorganic binder and sintering process to prepare h-BN coating on SS316L stainless steel substrate to make up for the problems of traditional insulating coating. After the inorganic binder immersion and annealing treatment, the adhesion between the insulating coating and the stainless steel substrate increases. The h-BN coating deposited at 30 V for 10 min has a uniform surface, a suitable thickness and a relatively compact structure. Through the corresponding experiment, the appropriate concentration of the inorganic binder was selected, which was 0.2 mol/L. Through the adhesion test, the coatings after soaking and sintering showed good adhesion, and the adhesive strength was 83.9%. In addition to having good heat resistance, the coating prepared under this condition also has good insulation, with the highest breakdown voltage of 1.22 kV and excellent breakdown strength of 0.254 kV/mm. In summary, this method can shorten the preparation cycle of the h-BN coatings, thereby obtaining a thicker and denser ceramic coating, with ideal and controllable performance, and showing the potential for insulation applications. This method can be used to prepare high-performance inorganic ceramic materials.  Informed Consent Statement: Not applicable.

Data Availability Statement:
All data used to support the findings of this study are included within the article.

Conflicts of Interest:
The authors declare no conflict of interest.