**2. Experiment**

#### *2.1. Test Samples and Electrode Arrangement*

This paper selects three types of polycyclic compounds. The polycyclic compound 2-hydroxy-2- phenylacetophenone is denoted by A, which is produced by J&K Scientific Ltd. The polycyclic compound 4-phenylbenzophenone is denoted by B, which is produced by Shanghai Macklin Biochemical Co., Ltd. The polycyclic compound 4,4-difluorobenzophenone is denoted by C, which is produced by J&K Scientific Ltd. The molecular structures of these polycyclic compounds are shown in Figure 1. The neat XLPE was supplied by Borealis Company. At 110 ◦C, we put the appropriate amount of XLPE into the mixer and mixed for 3 min. Then we weighed an amount of 0.5 wt% polycyclic compound and added it to the internal mixer. We then mixed for 10 min to make it evenly dispersed in XLPE. Then we used the flat vulcanizing machine and the special mold to make the needle-plate electrode samples [30]. The distance between the tip and the ground electrode was 2 ± 0.1 mm. The needle electrode diameter and curvature radius were 300 μm and 3 μm, respectively.

**Figure 1.** Molecular structures of polycyclic compounds.

#### *2.2. Experimental Apparatus and Procedure*

Figure 2 shows the schematic diagram of the experimental setup. The electrical tree experiment was carried out in the high temperature environment experimental platform. The sample was placed in a heat-resistant glass cylinder with dimethyl silicone oil, which could e ffectively prevent surface flashover of the sample during the experiment. In addition, silicone oil could also fill the surface of the sample to increase the transparency of the sample, improving the clarity of the electrical tree observation. The high temperature environment was provided by the resistance wires on both sides of the heat-resistant glass cylinder, and the maximum operating temperature was 200 ◦C. The experimental temperature was set to 30, 60, and 90 ◦C to make a temperature gradient. After the temperature reached the set temperature during the experiment, the sample was allowed to stand in the incubator for 10 min to ensure that the temperature of the material was consistent with the environment. The DC-impulse voltage was generated by the DC-impulse power source, which consisted of a DC power source and impulse power source [31]. The DC voltage was applied at a rate of 1 kV/s. After 1 min, the impulse voltage was applied at a rate of 1 kV/s. In order to reduce the error, each set of experiments was repeated 20 times. Because the electrical treeing initiation and breakdown phenomenon had a grea<sup>t</sup> relationship with the relative polarity of the impulse voltage with DC-impulse voltage [25], the electrical treeing characteristics with −25 kV DC and ±35 kV impulse voltage were measured to analyze the suppression e ffect of polycyclic compounds with DC-impulse voltage. The impulse voltage frequency was 400 Hz. The equivalent circuit of the experimental configuration is shown in Figure A1 in our Appendix A. The electrical tree imaging system included a computer, microscope unit, and a cold light source. The microscope unit consisted of an objective lens, an eyepiece, and a charge-coupled device (CCD). The CCD was a high-resolution imaging device. The camera multiplier was 1× and the highest resolution was 1024 × 768 pixels. The three objective magnifications were 4×, 10×, 40×, and the eyepiece magnification was 10×. The accumulated damage was used to analyze the electrical treeing characteristics. The accumulated damage refers to the number of pixels in the area covered by the electrical tree, which can be used to characterize the damage area of electrical tree to insulating materials, and to describe the development trend of electrical tree in space [29,32]. The accumulated damage was calculated using Matlab language. The specific calculation method was divided into three steps. The first step was to take a photo of the entire electrical branch with a pixel value of 500 × 500 pixels. In the second step, the image was subjected to filtering binarization to obtain a black and white image. The third step was to count the total number of pixels in the black area of the image. The value obtained is the accumulated damage value of the electrical tree.

The surface potential decay (SPD) technique is an e ffective method for measuring trap distribution behaviors [33]. The surface charge test system consisted of an HVDC power supply, a pin-gate-plate electrode system, a TREK type surface potentiometer (including a Kelvin type vibrating probe), and a constant temperature and humidity chamber. The vertical distance from the tip of the needle electrode to the gate electrode, and the vertical distance from the gate electrode to the surface of the sample were both 5 mm. The Kelvin vibrating probe was fixed with an epoxy holder with a vertical distance of 3 mm from the surface of the specimen. Charge was injected onto the surface of the sample for a charge time of 10 min. After the corona was over, we moved the center of the sample quickly below the surface potential measurement probe. The experimental temperature was set to 30, 60, and 90 ◦C with a relative humidity of ~20%. In this paper, the distribution of trap energy level (*Et*) and the trap density (*Nt*) are calculated to analyze the electrical property behaviors [34]. The carrier mobility was also calculated [35], which can obtain the characteristics of charge transportation.

**Figure 2.** Schematic diagram of experimental setup.
