Study on Improving Interface Performance of HVDC Composite Insulators by Plasma Etching

: High-voltage direct-current composite insulators are faced with various challenges during operation, such as creeping discharge, umbrella skirt damage, abnormal heating and insulator breakage. Among them, the aging of the interface between the core rod and the sheath is one of the important causes of composite insulator failure. In order to improve the electrical resistance of the composite insulator interface, this study uses plasma etching to modify the surface of the glass-ﬁber-reinforced epoxy resin plastic to prepare the high-voltage direct-current composite insulator core rod–sheath samples. By analyzing the surface morphology of the epoxy resin, static contact angle and surface charge transfer characteristics, the control mechanism of the plasma etching treatment on the interface bonding performance and leakage current of composite insulator core rod–sheath samples were studied. The results show that proper etching time treatment can improve the trap energy level distribution and microstructure of epoxy resin and increase the discharge voltage along the surface; chemical bonding plasma etching can improve the interfacial bonding performance of core rod–sheath samples sheaths, reduce the leakage current of composite insulator core rod–sheath samples sheath specimens and improve their interfacial performance. validation, Y.W.; formal analysis, Y.X. and J.L.; investigation, H.L.; resources, Y.X.; data curation, Y.W.; writing—original draft preparation, Y.W.; writing—review and editing, Y.X.; visualization, Y.W.; supervision, Y.X.; project administration, Y.X.; funding acquisition, Y.X. All authors have read and agreed to the published version of the manuscript.


Introduction
High-voltage DC composite insulators are mainly composed of internal glass fiber reinforced epoxy resin core bar and external high-temperature vulcanized silicone rubber (SIR) skirt sheath [1,2], the integrated performance of the core bar and sheath interface is directly related to the operational reliability of composite insulators [3][4][5]. Poor interface combination is prone to produce small voids, especially at high temperature, humidity, dirt and other harsh environments, the core rod-sheath samples surface may produce chemical aging due to nitric acid corrosion, oxidative decomposition and hydrolysis, the formation of low-density gas areas, causing local electric field distortion, inducing local discharge, accelerating insulation aging, ultimately leading to the failure of composite insulator insulation [6]. Therefore, in order to ensure the reliable operation of composite insulators under harsh climatic conditions, there is an urgent need to conduct research on the interface performance improvement of high-voltage DC composite insulators.
High-voltage DC composite insulators typically use a blunt polishing treatment and a silane coupling agent to improve interfacial bonding performance. The blunt polishing process causes a certain degree of inhomogeneity at the interface of the composite insulator, which affects the microstructural properties and trap parameters of the surface of the material, reduces the electrical resistance of the surface, and cannot guarantee the adequate adhesion of the silane coupling agent. Surface modification technology can reduce the inhomogeneity to a certain extent and become one of the alternatives to the blunt polishing process. Direct fluorination, dielectric blocking discharge, plasma etching, magnetron sputtering, surface coating are the common methods of surface modification of materials [7][8][9][10]. In order to regulate the interfacial charge migration properties and improve the electrical resistance along the surface, researchers at home and abroad have adopted a variety of different methods to modify the surface of polymer materials and come to fruition [4,11,12]. Du Boxue et al. used direct fluorination to treat the polymer surface by replacing the H atoms with F atoms to form C-F bonds, shielding the internal deep traps and increasing the surface charge dissipation rate and flashing voltage along the surface [13,14]. Cheng Yonghong et al. used the magnetron sputtering technique to sputter TiO 2 , Al 2 O 3 , and Cr 2 O 3 films on the surface of epoxy resin samples, resulting in faster surface charge dissipation [15]. Zhang Cheng et al. used atmospheric pressure dielectric blocking discharge and plasma-enhanced chemical vapor deposition methods to modify the epoxy resin surface, which promoted the material surface entrapment charge depression and reduced charge accumulation [16,17].
These studies mainly aim at improving the surface insulation of a single polymer material, but few studies have been conducted on the interface properties of different polymer materials. In this study, the effects of plasma etching time on the surface characteristics and the insulating properties along the surface of the epoxy resin samples were investigated. The effects of different etching times and surface roughness of the epoxy resin on the binding properties of the composite insulator interface were investigated.

Sample Preparation
Bisphenol A epoxy resin (Yanhai Chemical, HY-611, Tianjin, China) and low molecular weight polyamide resin (Yanhai Chemical, HY-651, Tianjin, China) were used to prepare square epoxy resin sample with thickness of 0.5 mm and side length of 9 cm. Additionally, 150 mesh sandpaper was used to grind some epoxy resin samples. A 12 mm diameter glass-fiber-reinforced epoxy resin (FRP, glass-fiber-reinforced epoxy resin plastic) core rod-sheath sample was selected and part of the FRP core rod-sheath samples was sanded with 80 mesh sandpaper. Using ME-3A multi-functional magnetically enhanced reactiveion etching machine (Institute of Microelectronics of the Chinese Academy, Beijing, China), plasma etching of epoxy resin samples and core rod-sheath samples were carried out for 5 min and 30 min, respectively, using CF 4 as working gas. Then silane coupling agent (KH550) is applied uniformly to the surface of the core rod-sheath samples, injection and high-temperature vulcanization of silicone rubber at 165 • C and 23 MPa, and finally cut into 25 mm lengths to complete the composite insulator short core rod-sheath samples.

Test Methods
(1) Scanning electron microscopy: Scanning electron microscope uses secondary electron signal imaging to observe the surface morphology of the sample. Scan the surface of the sample with a very thin electron beam to excite the surface of the sample to release secondary electrons. The secondary electrons generated are collected with a special detector to form an electrical signal and transport it to the picture tube to display the object on the fluorescent screen. It is mainly used to observe the surface morphology of the sample, the structure of the split surface, the structure of the inner surface of the lumen, etc.
(2) X-ray photoelectron spectroscopy: The main application of XPS is to measure the binding energy of electrons to achieve the qualitative analysis of surface elements, including valence states. This article is mainly used for the quantitative analysis of elements, reflecting the content or relative concentration of atoms based on the intensity of the photoelectron spectrum (the area of the photoelectron peak) in the energy spectrum.
(3) Surface potential measurement: experiment methods refer to previous papers [18], before each measurement with anhydrous ethanol cleaning epoxy resin sample surface, high voltage DC power supply through the current limiting resistor and needle electrode connected to the sample for corona charging, surface potentiometer through Kelvin probe to collect potential information, needle electrode 5 mm from the sample surface and Kelvin probe 3 mm from the sample surface in 25 • C experimental temperature and 30% humidity.
(4) Flashing voltage measurement along the surface: According to experiment method reference [19], the ungritted FRP core rod-sheath samples are placed between the high-voltage electrode and the ground electrode. The electrode spacing is 15 mm and the DC voltage is applied at a boost rate of 2 kV/min until the flashing, each sample conducts ten flashing tests and records the flashing voltage.
(5) Interfacial bonding performance test: the composite insulator short core rod-sheath samples are dissected. According to the size of the material bonding area of the core rod-sheath samples, sheath interfacial bonding performance is divided into five levels: 1 level-silicone rubber can be almost completely peeled off with the core rod-sheath samples; 2 level-sample non-stick area accounted for more than 50% of the bonding area; 3 level-sample non-stick area accounted for about 25% of the bonding area; 4 level-sample non-stick area accounted for less than 5% of the bonding area; 5 level-sample does not appear in the silicone rubber and core rod-sheath samples separation phenomenon.
(6) Leakage current measurement: According to experiment methods reference [20], using composite insulator short core rod-sheath samples, electrode in accordance with GB/T 22079-2008 [21] and IEC 62217 [22]. The leakage current is measured by a micro-analyzer connected to a computer. The DC voltage rises at 2 kV/min and increases to 30 kV, recording the peak leakage current during a minute period of the applied voltage. Figure 1 shows the surface morphology scanning electron microscopy (SEM) images of epoxy resin samples with plasma etching times of 0, 5, and 30 min. As the surface reaction of the epoxy resin at the initial etching stage is mainly the breaking of the polymer's own chemical bonds and the formation of new bonds, the surface of the epoxy resin reacts with the fluorine atoms in the CF 4 plasma as the etching time increasing, produces volatile gases and results in a significant increase in the sample surface roughness.

Effect of Plasma Etching on the Surface Characteristics of Epoxy Resin Samples
Coatings 2020, 10, x FOR PEER REVIEW 3 of 12 concentration of atoms based on the intensity of the photoelectron spectrum (the area of the photoelectron peak) in the energy spectrum.
(3) Surface potential measurement: experiment methods refer to previous papers [18], before each measurement with anhydrous ethanol cleaning epoxy resin sample surface, high voltage DC power supply through the current limiting resistor and needle electrode connected to the sample for corona charging, surface potentiometer through Kelvin probe to collect potential information, needle electrode 5 mm from the sample surface and Kelvin probe 3 mm from the sample surface in 25 °C experimental temperature and 30% humidity.
(4) Flashing voltage measurement along the surface: According to experiment method reference [19], the ungritted FRP core rod-sheath samples are placed between the high-voltage electrode and the ground electrode. The electrode spacing is 15 mm and the DC voltage is applied at a boost rate of 2 kV/min until the flashing, each sample conducts ten flashing tests and records the flashing voltage.
(5) Interfacial bonding performance test: the composite insulator short core rod-sheath samples are dissected. According to the size of the material bonding area of the core rod-sheath samples, sheath interfacial bonding performance is divided into five levels: 1 level-silicone rubber can be almost completely peeled off with the core rod-sheath samples; 2 level-sample non-stick area accounted for more than 50% of the bonding area; 3 level-sample non-stick area accounted for about 25% of the bonding area; 4 level-sample non-stick area accounted for less than 5% of the bonding area; 5 level-sample does not appear in the silicone rubber and core rod-sheath samples separation phenomenon.
(6) Leakage current measurement: According to experiment methods reference [20], using composite insulator short core rod-sheath samples, electrode in accordance with GB/T 22079-2008 [21] and IEC 62217 [22]. The leakage current is measured by a micro-analyzer connected to a computer. The DC voltage rises at 2 kV/min and increases to 30 kV, recording the peak leakage current during a minute period of the applied voltage. Figure 1 shows the surface morphology scanning electron microscopy (SEM) images of epoxy resin samples with plasma etching times of 0, 5, and 30 min. As the surface reaction of the epoxy resin at the initial etching stage is mainly the breaking of the polymer's own chemical bonds and the formation of new bonds, the surface of the epoxy resin reacts with the fluorine atoms in the CF4 plasma as the etching time increasing, produces volatile gases and results in a significant increase in the sample surface roughness. The X-ray photoelectron spectroscopy (XPS) of the epoxy samples at different plasma etching times is shown in Figure 2 and the surface atomic percentages of the samples at different etching times are shown in Table 1. After 30 min of etching, F1s peaks appear in the sample and the number of photonics of O1s peaks decreases, which indicates that after plasma etching, a large number of F  Table 1. After 30 min of etching, F1s peaks appear in the sample and the number of photonics of O1s peaks decreases, which indicates that after plasma etching, a large number of F elements appear on the surface of the sample. The C-O bond on the surface of the sample is replaced by C-F bond and the percentage of F atoms increases while the percentage of O atoms decreases.  To investigate the effect of plasma etching on the surface hydrophobicity of epoxy resin, the static contact angles of epoxy resin samples at different etching times were measured as shown in Figure 3. As the etching time increases, the CF4 plasma bombards the surface with free radicals and the fluorine atoms combine with the polymer group to form a stable C-F bond that forms a water-repellent layer on the sample surface, then results in a larger contact angle.  To investigate the effect of plasma etching on the surface hydrophobicity of epoxy resin, the static contact angles of epoxy resin samples at different etching times were measured as shown in Figure 3. As the etching time increases, the CF 4 plasma bombards the surface with free radicals and the fluorine Coatings 2020, 10, 1036 5 of 12 atoms combine with the polymer group to form a stable C-F bond that forms a water-repellent layer on the sample surface, then results in a larger contact angle.  Figure 4 shows the surface potential decay curves of epoxy samples with different plasma etching times after applying 6 kV DC voltage for 10 min. Due to exposure to the atmosphere, many factors can cause the attenuation of the surface potential. This article discusses the changes in surface potential caused by the sinking and de-sinking behavior of surface charges on the sample surface. In the initial stage, the drop in surface potential is mainly caused by the charge trapped in shallow traps, and the surface potential decays relatively faster. With increasing time, the charge in the shallow traps becomes less and less charged, while the charge in the deep traps is more difficult to detach, resulting in a gradual decrease in the rate of surface potential decay. To further investigate the effect of plasma etching on the surface potential decay rate, this paper defines the decay rate to describe the surface potential decay rate in a certain minute, which is expressed as:

Effect of Plasma Etching on the Surface Insulation Properties of Epoxy Resin Samples
In the equation, VS(0) is the initial surface potential and VS(t) is the t-time surface potential.   Figure 4 shows the surface potential decay curves of epoxy samples with different plasma etching times after applying 6 kV DC voltage for 10 min. Due to exposure to the atmosphere, many factors can cause the attenuation of the surface potential. This article discusses the changes in surface potential caused by the sinking and de-sinking behavior of surface charges on the sample surface. In the initial stage, the drop in surface potential is mainly caused by the charge trapped in shallow traps, and the surface potential decays relatively faster. With increasing time, the charge in the shallow traps becomes less and less charged, while the charge in the deep traps is more difficult to detach, resulting in a gradual decrease in the rate of surface potential decay. To further investigate the effect of plasma etching on the surface potential decay rate, this paper defines the decay rate to describe the surface potential decay rate in a certain minute, which is expressed as:

Effect of Plasma Etching on the Surface Insulation Properties of Epoxy Resin Samples
Coatings 2020, 10, x FOR PEER REVIEW 5 of 12  Figure 4 shows the surface potential decay curves of epoxy samples with different plasma etching times after applying 6 kV DC voltage for 10 min. Due to exposure to the atmosphere, many factors can cause the attenuation of the surface potential. This article discusses the changes in surface potential caused by the sinking and de-sinking behavior of surface charges on the sample surface. In the initial stage, the drop in surface potential is mainly caused by the charge trapped in shallow traps, and the surface potential decays relatively faster. With increasing time, the charge in the shallow traps becomes less and less charged, while the charge in the deep traps is more difficult to detach, resulting in a gradual decrease in the rate of surface potential decay. To further investigate the effect of plasma etching on the surface potential decay rate, this paper defines the decay rate to describe the surface potential decay rate in a certain minute, which is expressed as:

Effect of Plasma Etching on the Surface Insulation Properties of Epoxy Resin Samples
In the equation, VS(0) is the initial surface potential and VS(t) is the t-time surface potential.  In the equation, V S (0) is the initial surface potential and V S (t) is the t-time surface potential. Figures 1-5 gives the surface potential decay rates for epoxy samples with different plasma etching times at 60 s dissipation time, corresponding to 10.45%, 7.37%, and 15.25% for samples with 0, 5, and 30 min etching times. As the etching time increases, the surface potential decay of the sample first slows down and then accelerates.
Coatings 2020, 10, x FOR PEER REVIEW 6 of 12 0, 5, and 30 min etching times. As the etching time increases, the surface potential decay of the sample first slows down and then accelerates. The surface potential attenuation is closely related to the charge entrapment and deprogramming process [23]. Figure 6 shows the trap energy level distribution of epoxy resin samples with different plasma etching times calculated according to reference [24]. As the etching time increases from 0 min to 5 min, the curve peak trap energy level increases from 0.83 to 0.85 eV. As the etching time increases to 30 min, the curve peak trap energy level decreases to 0.81 eV. As the CF4 plasma bombards the epoxy resin surface at the beginning of etching, the original chemical bond is broken and new chemical bond forms. Bond breaking and re-crosslinking process as the main part of the reaction, making the sample trap energy level deepen. As the etching time increases, a large number of fluorine atoms bind to the active groups on the surface of the sample and due to the strong electronegativity of the fluorine element, the sample surface can bind a large number of electrons, forming a shield on the surface of the material, blocking the injection of charge into the interior of the material, reducing the depth of the trap [25]. The surface potential attenuation is closely related to the charge entrapment and deprogramming process [23]. Figure 6 shows the trap energy level distribution of epoxy resin samples with different plasma etching times calculated according to reference [24]. As the etching time increases from 0 min to 5 min, the curve peak trap energy level increases from 0.83 to 0.85 eV. As the etching time increases to 30 min, the curve peak trap energy level decreases to 0.81 eV. As the CF 4 plasma bombards the epoxy resin surface at the beginning of etching, the original chemical bond is broken and new chemical bond forms. Bond breaking and re-crosslinking process as the main part of the reaction, making the sample trap energy level deepen. As the etching time increases, a large number of fluorine atoms bind to the active groups on the surface of the sample and due to the strong electronegativity of the fluorine element, the sample surface can bind a large number of electrons, forming a shield on the surface of the material, blocking the injection of charge into the interior of the material, reducing the depth of the trap [25].
Coatings 2020, 10, x FOR PEER REVIEW 6 of 12 0, 5, and 30 min etching times. As the etching time increases, the surface potential decay of the sample first slows down and then accelerates. The surface potential attenuation is closely related to the charge entrapment and deprogramming process [23]. Figure 6 shows the trap energy level distribution of epoxy resin samples with different plasma etching times calculated according to reference [24]. As the etching time increases from 0 min to 5 min, the curve peak trap energy level increases from 0.83 to 0.85 eV. As the etching time increases to 30 min, the curve peak trap energy level decreases to 0.81 eV. As the CF4 plasma bombards the epoxy resin surface at the beginning of etching, the original chemical bond is broken and new chemical bond forms. Bond breaking and re-crosslinking process as the main part of the reaction, making the sample trap energy level deepen. As the etching time increases, a large number of fluorine atoms bind to the active groups on the surface of the sample and due to the strong electronegativity of the fluorine element, the sample surface can bind a large number of electrons, forming a shield on the surface of the material, blocking the injection of charge into the interior of the material, reducing the depth of the trap [25]. Weibull distribution results, it can be concluded that the average flashing voltages of samples with 0, 5, and 30 min etching times are 25.00, 31.99, and 22.56 kV. Proper plasma etching time increases the surface trap depth of the sample, enhances the ability of the surface to trap the isopolar charge, limits the initial emission of electrons and weakens the electric field along the surface. In addition, plasma etching can increase the surface roughness of the sample and increase the creepage distance. The surface bumps formed by the etching can also block electron migration and impede the development process of secondary electron emission. However, increasing the surface roughness can cause local electric field distortion, resulting in an uneven electric field distribution and lower the flash voltage. When the etching time is short, the surface roughness of the sample is low and the main factor affecting the flash voltage along the surface is a deep trap, the flash voltage is increased. As the etching time increases, the surface roughness of the sample is further improved, the surface shape of some parts fluctuates violently. Electric field distortion becomes the main factor affecting the flashing along the surface, making the flashing voltage lower.
Coatings 2020, 10, x FOR PEER REVIEW 7 of 12 time increases the surface trap depth of the sample, enhances the ability of the surface to trap the isopolar charge, limits the initial emission of electrons and weakens the electric field along the surface. In addition, plasma etching can increase the surface roughness of the sample and increase the creepage distance. The surface bumps formed by the etching can also block electron migration and impede the development process of secondary electron emission. However, increasing the surface roughness can cause local electric field distortion, resulting in an uneven electric field distribution and lower the flash voltage. When the etching time is short, the surface roughness of the sample is low and the main factor affecting the flash voltage along the surface is a deep trap, the flash voltage is increased. As the etching time increases, the surface roughness of the sample is further improved, the surface shape of some parts fluctuates violently. Electric field distortion becomes the main factor affecting the flashing along the surface, making the flashing voltage lower.

Effect of Plasma Etching on Interface Bonding Performance of Core Rod-Sheath Samples
Samples with different mold and plasma etching times were numbered, as shown in Table 2. Interfacial binding performance tests were performed on the samples according to the methods in Section 2 and the results are shown in Table 3. The surface bonding properties of a smooth sample improve with increasing etching time. The surface bonding properties of the rough samples does not change significantly with a short etching time but improves with further increase in etching time.  The chemical bond theory is typically used to explain the micromechanism of the silane coupling agent in coupling two materials. Its role is usually divided into the following processes [20,27]: First of all, the silane coupling agent RSi(OR')3 coated on the core rod is hydrolyzed in a

Effect of Plasma Etching on Interface Bonding Performance of Core Rod-Sheath Samples
Samples with different mold and plasma etching times were numbered, as shown in Table 2. Interfacial binding performance tests were performed on the samples according to the methods in Section 2 and the results are shown in Table 3. The surface bonding properties of a smooth sample improve with increasing etching time. The surface bonding properties of the rough samples does not change significantly with a short etching time but improves with further increase in etching time.  The chemical bond theory is typically used to explain the micromechanism of the silane coupling agent in coupling two materials. Its role is usually divided into the following processes [20,27]: First of all, the silane coupling agent RSi(OR') 3 coated on the core rod is hydrolyzed in a humid environment to yield reactive polyhydric silanol RSi(OH) 3 and liberating alcohol R'OH simultaneously. During the hydrolyzing process, the concomitant condensation between silicon hydroxyls in silanols also takes place to liberate water. The activated silanol monomers or oligomers are physically adsorbed to hydroxyl groups on the core rod surface by hydrogen bonds. The silicon hydroxyls and the hydroxyl groups can be converted into the covalent Si-O-C bonds through a dehydration condensation reaction. At higher temperatures, the residual silicon hydroxyls in silanols on the rod surface will further condense with each other, thereby form rigid polysiloxane structures linked with a stable Si-O-Si bond. On the surface of the high-temperature vulcanized silicone rubber, the organic group R in the silane coupling agent forms a Si-O-Si chemical bond with the HTV SIR sheath by crosslinking, as shown in Figure 8. As described above, the covalent bond Si-O-C between the core rod-sheath samples and the sheath is hydrolyzed during immersion of the short rod-sheath samples in boiling deionized water and aqueous solution due to the reversibility of this bond. The process shown in Figure 9 is affected by the application time of the silane coupling agent. Correspondingly, the inverse of the process shown in Figure 9, the hydrolysis reaction of the covalent Si-O-C bonds, is also affected by the coating times.
Coatings 2020, 10, x FOR PEER REVIEW 8 of 12 humid environment to yield reactive polyhydric silanol RSi(OH)3 and liberating alcohol R'OH simultaneously. During the hydrolyzing process, the concomitant condensation between silicon hydroxyls in silanols also takes place to liberate water. The activated silanol monomers or oligomers are physically adsorbed to hydroxyl groups on the core rod surface by hydrogen bonds. The silicon hydroxyls and the hydroxyl groups can be converted into the covalent Si-O-C bonds through a dehydration condensation reaction. At higher temperatures, the residual silicon hydroxyls in silanols on the rod surface will further condense with each other, thereby form rigid polysiloxane structures linked with a stable Si-O-Si bond. On the surface of the high-temperature vulcanized silicone rubber, the organic group R in the silane coupling agent forms a Si-O-Si chemical bond with the HTV SIR sheath by crosslinking, as shown in Figure 8. As described above, the covalent bond Si-O-C between the core rod-sheath samples and the sheath is hydrolyzed during immersion of the short rod-sheath samples in boiling deionized water and aqueous solution due to the reversibility of this bond. The process shown in Figure 9 is affected by the application time of the silane coupling agent. Correspondingly, the inverse of the process shown in Figure 9, the hydrolysis reaction of the covalent Si-O-C bonds, is also affected by the coating times.  humid environment to yield reactive polyhydric silanol RSi(OH)3 and liberating alcohol R'OH simultaneously. During the hydrolyzing process, the concomitant condensation between silicon hydroxyls in silanols also takes place to liberate water. The activated silanol monomers or oligomers are physically adsorbed to hydroxyl groups on the core rod surface by hydrogen bonds. The silicon hydroxyls and the hydroxyl groups can be converted into the covalent Si-O-C bonds through a dehydration condensation reaction. At higher temperatures, the residual silicon hydroxyls in silanols on the rod surface will further condense with each other, thereby form rigid polysiloxane structures linked with a stable Si-O-Si bond. On the surface of the high-temperature vulcanized silicone rubber, the organic group R in the silane coupling agent forms a Si-O-Si chemical bond with the HTV SIR sheath by crosslinking, as shown in Figure 8. As described above, the covalent bond Si-O-C between the core rod-sheath samples and the sheath is hydrolyzed during immersion of the short rod-sheath samples in boiling deionized water and aqueous solution due to the reversibility of this bond. The process shown in Figure 9 is affected by the application time of the silane coupling agent. Correspondingly, the inverse of the process shown in Figure 9, the hydrolysis reaction of the covalent Si-O-C bonds, is also affected by the coating times.

Effect of Plasma Etching on Leakage Current of Composite Insulator Samples
According to the national standard GB/T 22079-2008 and IEC 62217:2005, composite insulator samples were immersed in deionized water and 0.1% NaCl solution. The samples were boiled for 100 h to simulate the humidity and salt spray environment. Figure 10a,b show the DC leakage currents of smooth samples and rough samples after deionized water treatment. For the smooth samples etched for 0, 5, and 30 min, with the increase of voltage, the leakage current of the etched 30 min sample is the largest. The leakage current of the etched 5 min and unetched smooth samples are smaller, the difference between the two is not large. The shorter etching time has less effect on the leakage current performance of smooth samples treated with deionized water, while the leakage current increases significantly with longer etching time. For coarse samples, the leakage current is greatest for unetched samples with increasing voltage and differs greatly from that of etched samples. For rough samples etched for 0, 5, and 30 min, leakage currents were minimal for the 5 min etched samples and slightly greater for the 30 min etched samples than for the 5 min etched samples. Therefore, the proper etching time can significantly reduce the leakage current of the rough sample after deionized water treatment, but if a certain etching time is exceeded, the leakage current will increase, making the modification effect worse.

Effect of Plasma Etching on Leakage Current of Composite Insulator Samples
According to the national standard GB/T 22079-2008 and IEC 62217:2005, composite insulator samples were immersed in deionized water and 0.1% NaCl solution. The samples were boiled for 100 h to simulate the humidity and salt spray environment. Figure 10a,b show the DC leakage currents of smooth samples and rough samples after deionized water treatment. For the smooth samples etched for 0, 5, and 30 min, with the increase of voltage, the leakage current of the etched 30 min sample is the largest. The leakage current of the etched 5 min and unetched smooth samples are smaller, the difference between the two is not large. The shorter etching time has less effect on the leakage current performance of smooth samples treated with deionized water, while the leakage current increases significantly with longer etching time. For coarse samples, the leakage current is greatest for unetched samples with increasing voltage and differs greatly from that of etched samples. For rough samples etched for 0, 5, and 30 min, leakage currents were minimal for the 5 min etched samples and slightly greater for the 30 min etched samples than for the 5 min etched samples. Therefore, the proper etching time can significantly reduce the leakage current of the rough sample after deionized water treatment, but if a certain etching time is exceeded, the leakage current will increase, making the modification effect worse.

Conclusions
In order to improve the interface performance of the composite insulator core rod-sheath samples sheath, this paper adopts plasma etching to modify the surface of the epoxy resin core rodsheath samples, observes the epoxy resin surface shape, static contact angle and insulating

Conclusions
In order to improve the interface performance of the composite insulator core rod-sheath samples sheath, this paper adopts plasma etching to modify the surface of the epoxy resin core rod-sheath samples, observes the epoxy resin surface shape, static contact angle and insulating properties along the surface at different etching times. The bonding performance of the core rod sheath interface and the leakage current of the composite insulator sample under different conditions are also measured. The main conclusions are as follows: (1) When the plasma etching time is 5 min, the surface roughness of the epoxy resin sample changes less and when the etching time is 30 min, the surface roughness increases significantly. After plasma etching, the C-O bond on the surface of the sample is replaced with a C-F bond and the F-atom percentage increases while the O-atom percentage decreases.