Preparation Methods of Polypropylene/Nano-Silica/Styrene-Ethylene-Butylene-Styrene Composite and Its Effect on Electrical Properties

Compared with traditional insulation materials, such as cross-linked polyethylene (XLPE), polypropylene (PP) is famous for its better recyclable and thermal properties, as well as its good electrical performance. However, the problem of poor impact strength has restricted the application of pure PP in high-voltage, direct current (HVDC) cables. In this paper, styrene-ethylene-butylene-styrene block copolymer (SEBS) was used as a toughening filler, and nano-SiO2 was expected to improve the electric properties of the nano-composite. By controlling the masterbatch system, the dispersion characteristics of nano-SiO2 in the ternary composite system were changed. When PP/SiO2 was used as the masterbatch and then blended with SEBS, nano-SiO2 tended to disperse in the PP phase, and the number of nano-particles in the SEBS phase was lower. When PP/SEBS was used as the masterbatch, nano-SiO2 was distributed in both the PP phase and the SEBS phase. When SEBS/SiO2 was used as the masterbatch, nano-SiO2 tended to be dispersed in the SEBS phase. The different dispersion characteristics of nano-SiO2 changed the crystallization and mechanical properties of the ternary composite system and produced different electrical performance improvement effects. The results of our experiment revealed that the space charge suppression capability was positively correlated with the direct current (DC) breakdown strength improvement effect. Compared with the DC performance of 500 kV commercial XLPE materials, the self-made PP-based ternary composite system has better space charge suppression effects and higher DC breakdown strength. When nano-SiO2 was more dispersed in the PP phase, the space charge improvement effect was best. When the nano-SiO2 particles were more dispersed in the SEBS phase, the expected electrical property improvement was not obtained. Scanning electron microscopy showed that the nano-SiO2 particles in the SEBS phase were more dispersed at the interface than in the SEBS matrix, indicating that the nano-particles were poorly dispersed, which may be a reason why the electrical properties of the composite system were not significantly improved.


Introduction
In recent years, with the introduction of concepts such as the super grid and the global energy interconnection, the demand for high-voltage power cables has increased. Intercontinental power interconnection over long distances across the sea needs to be achieved with high-voltage DC (direct current) cables [1]. The application of cross-linked polyethylene (XLPE) to high-voltage DC cables, Polymers 2019, 11, 797 3 of 14 was 75:25:1. All the samples melted at 190 • C and 60 rpm by using an internal mixer, and PP, SEBS, and nano-SiO 2 were melted respectively for 3, 3, and 15 min. The preparation process was as follows. First, PP and SEBS were separately melt-blended into the internal mixer to prepare masterbatch 1, which we called PP/SEBS, masterbatch 2 was prepared by blending PP with nano-SiO 2 , and masterbatch 3 was blended by using SEBS and nano-SiO 2 . Second, the nano-composites of the PP, SEBS, and nano-SiO 2 ternary systems were prepared by using three masterbatches, respectively. Nano-SiO 2 was blended into masterbatch 1, SEBS was added into masterbatch 2, and PP was melted into masterbatch 3, according to different blending orders. We referred to these three nano-composites as PP/SEBS/SiO 2 , PP/SiO 2 /SEBS, SEBS/SiO 2 /PP, respectively, and the blending order and abbreviations of the composites are shown in Table 1. PP samples were pressed to the required thickness with compression molding at 190 • C, and XLPE samples were placed in a mold and heated, first to 110 • C at 2 × 10 6 Pa pressure to melt and then to 175 • C at 15 × 10 6 Pa to crosslink the polymer.

Sample Preparation for Scanning Electron Microscope (SEM) Observations
All the samples were pressed to 1 mm thickness and fractured by liquid nitrogen at low temperature. The fractured surfaces of the samples were then immersed for 7 h in the solution recommended by M. Aboulfaraj for the observation of spherulitic structures in PP [14]. The formula is 1.3 wt % potassium permanganate, 32.9 wt % concentrated H 3 PO 4 , and 65.8 wt % concentrated H 2 SO 4 . This permanganic acid preferentially etches the amorphous part of the polymer in the spherulites, in such a way that the lamellae then appear clearly. This method proved particularly useful in the case of PP. Subsequently, the specimens were carefully washed with detergent, which was a mixture of concentrated sulfuric acid water and hydrogen peroxide with a volume ratio of 2:7:1 [15]. Then, all the fractured surfaces of the samples were sputtered with a very thin layer of gold in order to eliminate any undesirable charge effects during the SEM observations. Field emission scanning electron microscopy (Hitachi SU8020, Tokyo, Japan) was used to observe the microscopic morphology of the samples.

Thermal Properties
Crystallization and melting curves were measured by differential scanning calorimetry (DSC, DSC822e, Mettler-Toledo International, Inc., Switzerland). A specimen weighing approximately 5 mg was heated to 200 • C at a rate of 20 • C/min and kept at that temperature for 2 min to eliminate thermal history. Then, it was cooled to 30 • C at the rate of 10 • C/min and kept at that temperature for 2 min to obtain its crystallization curve. Finally, the specimens were heated to 200 • C again at a rate of 10 • C/min to record their melting curves. The crystallinity (X c /%) was calculated from the DSC result by Equation (1).
where ∆H m (J/g) was the melting enthalpy; H 0 was the theoretical melting enthalpy of the completely crystallized form, and H 0 was 209 J/g for the isotactic PP [16]. x was the mass fraction of the inorganic.

The Dynamic Mechanical Properties
The dynamic temperature relaxation spectrum was tested by the dynamic mechanical analysis (DMA) method using DMAQ800 manufactured by TA instruments (New Castle, DE, USA). The selection Polymers 2019, 11, 797 4 of 14 mode was the tensile mode, in which the target amplitude was 15 µm, the frequency was 1 Hz, the static force was 0.375 N, and the dynamic force was 0.3 N. The sample was cuboid with a length, width, and thickness of 15 mm, 6 mm, and 1 mm, respectively. The sample was first cooled to −80 • C for 5 min, and then, the elastic modulus E', the loss modulus E", and the loss factor tan δ from −80 • C to 160 • C were measured at a linear heating rate of 3 • C/min.

Thermally Stimulated Depolarization Current
The charge trap characteristic was tested using the thermally stimulated depolarization current (TSDC) method. The prepared samples were electrically polarized by applying 40 kV/mm DC high voltage for 1 h in a vacuum environment at 60 • C. The liquid nitrogen was then used for rapid cooling the samples to −80 • C so that all kinds of charge carriers had been "frozen", after which the DC high voltage was removed and the samples were short-circuited for about 10 min. After that, the sample was heated linearly under a constant heating rate of 3 • C/min, and the short-circuit current was measured using a 6517B electrometer (Keithley Instrument Inc., Cleveland, OH, USA).

DC Breakdown Test
A DC breakdown test was undertaken on the film samples. The PP samples were tested at room temperature, 90 and 120 • C, and the XLPE samples were tested at room temperature, 50, 70, and 90 • C. Samples with thickness of 100 µm were placed between two electrodes. The samples were immersed in transformer oil to prevent surface flashover, and the voltage ramp was 2 kV/s.

Space Charge Measurement
The space charge distribution within the samples under DC electric field was measured using the pulsed electro-acoustic (PEA) system at room temperature. The space charge measurement was performed with a pulsed electro-acoustic (PEA) system produced by Shang Hai Xiangtie electromechanical device Co., Ltd., Shanghai, China. All the samples were kept under a 40 kV/mm DC electrical field for 1 h and then short-circuited for 1 h. Space charge profiles were recorded at various times for analysis.

Morphology
The SEM images of all the samples are shown in Figure 1. It can be seen from the images that the structure of the PP crystal was regular, and the spherulite diameter was about 100-120 µm. Many scholars have also confirmed the morphological structure of PP crystals [17]. The high-field electrical properties of PP, particularly the breakdown strength, are associated with the features of spherulites [18].
Because SEBS material contains polystyrene block copolymer and the material is amorphous in the aggregation state, PP and SEBS cannot be homogeneous blends [19]. In Figure 1c, the "voids" in the PP/SEBS composites are the etched SEBS phase. The "sea-island" structure distribution with SEBS as the "island" phase and PP as the "sea" phase were presented in the PP/SEBS composite [20]. The white bright spots in Figure 1f,h,j are the nano-SiO 2 particles. Due to the different preparation methods, in Figure 1f,h,j, it was found that nano-SiO 2 in the PP/SiO 2 /SEBS sample was mostly dispersed in the PP phase and a few in the interface between PP and SEBS. In PP/SEBS/SiO 2 nano-composites, both the PP phase and the SEBS phase had nano-SiO 2 particles, while in the SEBS/SiO2/PP sample, nano-SiO 2 was concentrated in the PP and SEBS interface, partly in the SEBS phase and a few in the PP phase. The dispersion of nano-SiO 2 in the PP phase was relatively uniform, but the distribution of nano-SiO 2 in the SEBS phase was not uniform.  In the large field of view in Figure 1c,e,g,i, it was observed that there was a spherulite morphology in the PP/SEBS composite, and the crystal structure was smaller and looser than in PP. It was also found that SEBS was distributed along the direction of the lamella growth. The dispersion state of SEBS in PP/SiO 2 /SEBS and PP/SEBS/SiO 2 was similar, while SEBS in SEBS/SiO 2 /PP was significantly smaller. This indicates that nano-SiO 2 in the SEBS phase caused SEBS to form small islands and facilitated the SEBS distribution in the PP matrix. One possible mechanism was that the surface of the nano-SiO 2 particles modified by hexamethyldisilazane was coated by dense alkane molecules, which reduced the polarity of nano-SiO 2 . It can be seen from the SEM that the dispersion of nano-SiO 2 in the non-polar PP phase was better than that in the weak polar SEBS phase, and a large number of nano-SiO 2 particles were distributed in the interface between SEBS and PP, which improved the compatibility between SEBS and PP, showing that the dispersion of SEBS in PP had been improved.

Thermal Properties
The DSC curves of all the samples are shown in Figure 2, and the crystallization temperature (T c ), melting temperature (T m ), and crystallinity (X c ) are summarized in Table 2. In the large field of view in Figure 1c,e,g,i, it was observed that there was a spherulite morphology in the PP/SEBS composite, and the crystal structure was smaller and looser than in PP. It was also found that SEBS was distributed along the direction of the lamella growth. The dispersion state of SEBS in PP/SiO2/SEBS and PP/SEBS/SiO2 was similar, while SEBS in SEBS/SiO2/PP was significantly smaller. This indicates that nano-SiO2 in the SEBS phase caused SEBS to form small islands and facilitated the SEBS distribution in the PP matrix. One possible mechanism was that the surface of the nano-SiO2 particles modified by hexamethyldisilazane was coated by dense alkane molecules, which reduced the polarity of nano-SiO2. It can be seen from the SEM that the dispersion of nano-SiO2 in the non-polar PP phase was better than that in the weak polar SEBS phase, and a large number of nano-SiO2 particles were distributed in the interface between SEBS and PP, which improved the compatibility between SEBS and PP, showing that the dispersion of SEBS in PP had been improved.

Thermal Properties
The DSC curves of all the samples are shown in Figure 2, and the crystallization temperature (Tc), melting temperature (Tm), and crystallinity (Xc) are summarized in Table 2.  When SEBS was added into PP, the Tc and Tm of PP/SEBS were reduced compared with that of PP. A similar result was reported by Jyotishkumar; a low concentration of SEBS resulted in little change in the Tc and a higher Tm of the polymer, but when the amount of SEBS increased to 20%, the Tc and Tm of composite were decreased [21]. When SEBS was less filled, SEBS acted more as a nucleating agent to promote the crystallization, but when the concentration of SEBS increased to a larger extent, more SEBS entangled with the molecular chain of PP, suppressing the crystallization process of PP and causing the decrease of the Tc. As shown in Figure 1, SEBS made PP spherulites smaller, and the structure became looser. This phenomenon allowed the PP crystal to melt at a lower temperature, so the Tm peak of PP/SEBS was lowered.
As shown in Table 2, the differences in the Tc and Tm between the nano-composites and PP/SEBS were very small. Many studies indicated that nano-particles cause heterogeneous nucleation and promote the formation of crystals [22,23]. However, there was no obvious change in the Tc and Tm of  Table 2. Thermal parameters for the crystallization and melting processes. When SEBS was added into PP, the T c and T m of PP/SEBS were reduced compared with that of PP. A similar result was reported by Jyotishkumar; a low concentration of SEBS resulted in little change in the T c and a higher T m of the polymer, but when the amount of SEBS increased to 20%, the T c and T m of composite were decreased [21]. When SEBS was less filled, SEBS acted more as a nucleating agent to promote the crystallization, but when the concentration of SEBS increased to a larger extent, more SEBS entangled with the molecular chain of PP, suppressing the crystallization process of PP and causing the decrease of the T c . As shown in Figure 1, SEBS made PP spherulites smaller, and the structure became looser. This phenomenon allowed the PP crystal to melt at a lower temperature, so the T m peak of PP/SEBS was lowered.
As shown in Table 2, the differences in the T c and T m between the nano-composites and PP/SEBS were very small. Many studies indicated that nano-particles cause heterogeneous nucleation and promote the formation of crystals [22,23]. However, there was no obvious change in the T c and T m of nano-composites in this paper. One possible reason was that the content of nano-SiO 2 was only 1 phr (parts per hundreds of resin); thus, nano-SiO 2 did not produce a significant change in the T c and T m of the nano-composite.

The Dynamic Mechanical Properties
The dynamic mechanical spectra of PP, PP/SEBS, and the three nano-composites are illustrated in Figure 3. From the change in the storage modulus (E') of the different materials in Figure 3a, it can be seen that the E' of the composites decreased after SEBS was added in the whole temperature spectrum, and E' of the PP/SiO 2 /SEBS nano-composite had a significantly lower E' in the temperature range of −80 to −20 • C. With the increasing of the temperature, the E' of all the samples gradually decreased. It was reported that the glass transition temperature of PP was about 10 • C [24]; in the temperature spectrum of the loss factor (tan δ) in Figure 3b, the loss peak appeared at this temperature. Therefore, in the temperature range of −80 to −20 • C, the molecular chains of PP in the crystal region and the amorphous region were in a frozen state, and the E' of the composites were mainly affected by the SEBS phase. In the PP/SiO 2 /SEBS nano-composite, nano-SiO 2 was more dispersed in the PP phase, which induced the formation of smaller-scale spherulites, so much SEBS dispersed in the interface of PP spherulite and the E' of PP/SiO 2 /SEBS nano-composites were more significantly reduced. As the temperature gradually increased and the lamella gradually melted, the increasing SEBS surrounded by spherulites began to decrease E', and then, the E' of all the samples were not much different at high temperatures. nano-composites in this paper. One possible reason was that the content of nano-SiO2 was only 1 phr (parts per hundreds of resin); thus, nano-SiO2 did not produce a significant change in the Tc and Tm of the nano-composite.

The Dynamic Mechanical Properties
The dynamic mechanical spectra of PP, PP/SEBS, and the three nano-composites are illustrated in Figure 3. From the change in the storage modulus (E') of the different materials in Figure 3a, it can be seen that the E' of the composites decreased after SEBS was added in the whole temperature spectrum, and E' of the PP/SiO2/SEBS nano-composite had a significantly lower E' in the temperature range of −80 to −20 °C. With the increasing of the temperature, the E' of all the samples gradually decreased. It was reported that the glass transition temperature of PP was about 10 °C [24]; in the temperature spectrum of the loss factor (tan δ) in Figure 3b, the loss peak appeared at this temperature. Therefore, in the temperature range of −80 to −20 °C, the molecular chains of PP in the crystal region and the amorphous region were in a frozen state, and the E' of the composites were mainly affected by the SEBS phase. In the PP/SiO2/SEBS nano-composite, nano-SiO2 was more dispersed in the PP phase, which induced the formation of smaller-scale spherulites, so much SEBS dispersed in the interface of PP spherulite and the E' of PP/SiO2/SEBS nano-composites were more significantly reduced. As the temperature gradually increased and the lamella gradually melted, the increasing SEBS surrounded by spherulites began to decrease E', and then, the E' of all the samples were not much different at high temperatures. From the change in the loss factor shown in Figure 3b, it was determined that the loss factor of the composite system increased after filling with SEBS compared with PP. After the addition of nano-SiO2, the loss factor of the nano-composite further increased. This was due to the entanglement of SEBS and nano-SiO2 with PP molecules, and the increase in friction during mechanical relaxation led to an increase in internal friction. At the same time, it was found that the mechanical relaxation peak of the PP/SiO2/SEBS nano-composite near 120 °C, which corresponded to the structural relaxation generated when the PP crystal region melts, was higher than those of other material systems, because the nano-SiO2 was more dispersed in the PP phase. Therefore, the biggest loss was produced in PP/SiO2/SEBS. The loss factor temperature spectra of the three nano-composites were complicated and involved various mechanisms, such as the crystal morphology change and interface zone action; thus, the specific influence needs further study.

Space Charge Distribution
The space charge distributions of all the samples under a polarization electric field are illustrated in Figure 4. It can be seen from Figure 4a that a significant space charge distribution appeared in the From the change in the loss factor shown in Figure 3b, it was determined that the loss factor of the composite system increased after filling with SEBS compared with PP. After the addition of nano-SiO 2 , the loss factor of the nano-composite further increased. This was due to the entanglement of SEBS and nano-SiO 2 with PP molecules, and the increase in friction during mechanical relaxation led to an increase in internal friction. At the same time, it was found that the mechanical relaxation peak of the PP/SiO 2 /SEBS nano-composite near 120 • C, which corresponded to the structural relaxation generated when the PP crystal region melts, was higher than those of other material systems, because the nano-SiO 2 was more dispersed in the PP phase. Therefore, the biggest loss was produced in PP/SiO 2 /SEBS. The loss factor temperature spectra of the three nano-composites were complicated and involved various mechanisms, such as the crystal morphology change and interface zone action; thus, the specific influence needs further study.

Space Charge Distribution
The space charge distributions of all the samples under a polarization electric field are illustrated in Figure 4. It can be seen from Figure 4a that a significant space charge distribution appeared in the XLPE after the electric field was applied. As the polarization time increased, the space charge migrated inside the sample. Different from XLPE, as shown in Figure 4b, when the polarization time was 5 s, there was no obvious space charge distribution in PP. When the polarization time reached 3600 s, a large amount of heterocharge accumulation appeared in the PP samples. With the addition of SEBS into PP, a large amount of heterogeneous space charge was found near the two electrodes. The amount of charge increased the longer the increasing voltage was applied, and more charges moved to the interior in PP/SEBS. With the addition of nano-SiO 2 , the space charge was decreased, and a small amount of space charge was accumulated in the three nano-composites under 40 kV/mm for 1 h. This indicated that the nano-SiO 2 particles could suppress the transfer of space charge in the material.  The space charge distribution of the samples under short circuit at 5 s and 3600 s are presented in the Figure 5. In order to further study the influence of SEBS and nano-SiO2 on the PP space charge, we calculated the space charge density of the samples by integrating the short circuit data of 5 s. The linear average space charge density Q(t) of the samples was obtained by dividing the middle space charge amount of each sample by the thickness of the specimen [25].
where x1 and x2 are the positive and negative electrode positions, respectively. ρ(x,t) is the space charge profile obtained at the short circuit data of 5 s.
The average space charge densities under short circuit at 5 s and 3600 s are shown in Table 3. It can be seen from the table that the average space charge density of XLPE was the highest under the short circuit at 5 s compared with those of the other materials. The space charge density decreased obviously after adding nano-SiO2 into PP/SEBS. After 3600 s, the space charge obviously decayed in The space charge distribution of the samples under short circuit at 5 s and 3600 s are presented in the Figure 5. In order to further study the influence of SEBS and nano-SiO 2 on the PP space charge, we calculated the space charge density of the samples by integrating the short circuit data of 5 s. The linear average space charge density Q(t) of the samples was obtained by dividing the middle space charge amount of each sample by the thickness of the specimen [25].
where x 1 and x 2 are the positive and negative electrode positions, respectively. ρ(x,t) is the space charge profile obtained at the short circuit data of 5 s. XLPE, and the average space charge density in XLPE changed from the highest to the lowest, whereas the space charge in PP and its composites showed no obvious change. This shows that the trap depth of XLPE was shallower than that of PP.

Thermally Stimulated Depolarization Current
The TSDC curves of all the samples are shown in Figure 6. The TSDC spectrum of XLPE has only one current release peak, and the peak position was at about 55 °C. Ieda's research showed that the carrier trap corresponding to this temperature was mainly due to some structural defects in the polyethylene crystal region [26]. The depth of this trap was relatively shallow, and it resulted in the space charge easily migrating into the sample and decaying at a faster rate, as shown in Figure 4a and Figure 5a.
As the TSDC curves of PP and its composites show in Figure 6, the two peaks were corresponded to the α relaxation process in PP [27]. The relaxation peak at low temperature, which was named peak 1, was mainly derived from the movement of amorphous linked molecular chains and loose coils between lamellas, and part of the trap charge bound by this area was released. The molecular chain  Table 3. It can be seen from the table that the average space charge density of XLPE was the highest under the short circuit at 5 s compared with those of the other materials. The space charge density decreased obviously after adding nano-SiO 2 into PP/SEBS. After 3600 s, the space charge obviously decayed in XLPE, and the average space charge density in XLPE changed from the highest to the lowest, whereas the space charge in PP and its composites showed no obvious change. This shows that the trap depth of XLPE was shallower than that of PP.

Thermally Stimulated Depolarization Current
The TSDC curves of all the samples are shown in Figure 6. The TSDC spectrum of XLPE has only one current release peak, and the peak position was at about 55 • C. Ieda's research showed that the carrier trap corresponding to this temperature was mainly due to some structural defects in the polyethylene crystal region [26]. The depth of this trap was relatively shallow, and it resulted in the space charge easily migrating into the sample and decaying at a faster rate, as shown in Figures 4a  and 5a. movement of the molecules in the crystalline region. Ions and electrons usually migrate between molecular chains [13].
As shown in Figure 6, it can be seen from the curves that the current release peaks of PP were 26 and 49 Pa at 80 and 140 °C, respectively. After the addition of SEBS, peak 1 of PP/SEBS appeared at 78 °C, and the peak value of peak 1 increased. The peak position of peak 2 in the PP/SEBS composite decreased to 120 °C, and the peak value increased to 54 Pa. In the TSDC measurement, the current released at high temperature corresponded to a deeper energy level [29]. When SEBS was blended into PP, as shown in Figure 1, SEBS destroyed the PP crystal structure and introduced a large number of interface phases between SEBS and PP. As SEM and DSC test results showed, compared with PP, the Tm of PP/SEBS decreased, and the regularity of crystallization worsened; therefore, the current release peaks were likely to occur at low temperature. When nano-SiO2 was added into the polymer, the TSDC value of the three nano-composites was obviously reduced, which implied that there was a small amount of charges accumulated in the three nano-composites samples compared with PP. This phenomenon was also illustrated by the space charge test results as shown in Figure 5. Many studies have proven that nano-particles can influence the trap density, and the deep trap mechanism may suppress carrier migration [30,31]. When nano-SiO2 was introduced, a lot of deep traps around the interface between nano-SiO2 and PP were formed, and the positive and negative charges that would otherwise move freely in PP were bound in a nearby nano-SiO2 [31]. When the electric field was applied, the deep traps near nano-SiO2 could prevent the transfer of charges and suppress the formation of charges accumulated in the samples [31][32][33].
Comparing the TSDC curves of the three nano-composites, the peak value of the TSDC current in the SEBS/SiO2/PP nano-composite near 80 °C was significantly higher than those of the other two nano-composites, indicating that there was more space charge accumulation between lamellas in the SEBS/SiO2/PP nano-composite. The space charge test results also show that there was more space accumulated in the SEBS/SiO2/PP nano-composite. From the results of the SEM images, it was determined that nano-SiO2 in SEBS/SiO2/PP was more distributed in the SEBS phase, whereas nano-SiO2 in the PP phase was less distributed, which may be the reason for the decrease in the space charge suppression effect. It also can be seen from the TSDC graph that the current values of the three nano-composites still had an upward trend at 160 °C, and it was likely that a current peak would appear in higher temperature. Because the temperature was already higher than the melting temperature of PP, the current after this temperature was more likely to come from the release process of the trapped charge of nano-SiO2, suggesting that nano-SiO2 can form deeper trap levels. As the TSDC curves of PP and its composites show in Figure 6, the two peaks were corresponded to the α relaxation process in PP [27]. The relaxation peak at low temperature, which was named peak 1, was mainly derived from the movement of amorphous linked molecular chains and loose coils between lamellas, and part of the trap charge bound by this area was released. The molecular chain movement in the amorphous region was due to the internal strain of the amorphous phase and the space hindrance caused by the presence of the crystalline phase [27,28]. The high temperature peak, which was named peak 2, was the crystallization pre-melting peak, which originated from the movement of the molecules in the crystalline region. Ions and electrons usually migrate between molecular chains [13].

DC Breakdown Strength
As shown in Figure 6, it can be seen from the curves that the current release peaks of PP were 26 and 49 Pa at 80 and 140 • C, respectively. After the addition of SEBS, peak 1 of PP/SEBS appeared at 78 • C, and the peak value of peak 1 increased. The peak position of peak 2 in the PP/SEBS composite decreased to 120 • C, and the peak value increased to 54 Pa. In the TSDC measurement, the current released at high temperature corresponded to a deeper energy level [29]. When SEBS was blended into PP, as shown in Figure 1, SEBS destroyed the PP crystal structure and introduced a large number of interface phases between SEBS and PP. As SEM and DSC test results showed, compared with PP, the T m of PP/SEBS decreased, and the regularity of crystallization worsened; therefore, the current release peaks were likely to occur at low temperature.
When nano-SiO 2 was added into the polymer, the TSDC value of the three nano-composites was obviously reduced, which implied that there was a small amount of charges accumulated in the three nano-composites samples compared with PP. This phenomenon was also illustrated by the space charge test results as shown in Figure 5. Many studies have proven that nano-particles can influence the trap density, and the deep trap mechanism may suppress carrier migration [30,31]. When nano-SiO 2 was introduced, a lot of deep traps around the interface between nano-SiO 2 and PP were formed, and the positive and negative charges that would otherwise move freely in PP were bound in a nearby nano-SiO 2 [31]. When the electric field was applied, the deep traps near nano-SiO 2 could prevent the transfer of charges and suppress the formation of charges accumulated in the samples [31][32][33].
Comparing the TSDC curves of the three nano-composites, the peak value of the TSDC current in the SEBS/SiO 2 /PP nano-composite near 80 • C was significantly higher than those of the other two nano-composites, indicating that there was more space charge accumulation between lamellas in the SEBS/SiO 2 /PP nano-composite. The space charge test results also show that there was more space accumulated in the SEBS/SiO 2 /PP nano-composite. From the results of the SEM images, it was determined that nano-SiO 2 in SEBS/SiO 2 /PP was more distributed in the SEBS phase, whereas nano-SiO 2 in the PP phase was less distributed, which may be the reason for the decrease in the space charge suppression effect. It also can be seen from the TSDC graph that the current values of the three nano-composites still had an upward trend at 160 • C, and it was likely that a current peak would appear in higher temperature. Because the temperature was already higher than the melting temperature of PP, the current after this temperature was more likely to come from the release process of the trapped charge of nano-SiO 2 , suggesting that nano-SiO 2 can form deeper trap levels.

DC Breakdown Strength
The electrical breakdown strength of XLPE, PP, and the PP composites was analyzed by Weibull statistics, and the data were shown in Figure 7. The highest long-term operating temperature of the HVDC XLPE insulation was only 70 • C; therefore, the highest test temperature for XLPE was 90 • C in this study. The electrical breakdown strength of XLPE, PP, and the PP composites was analyzed by Weibull statistics, and the data were shown in Figure 7. The highest long-term operating temperature of the HVDC XLPE insulation was only 70 °C; therefore, the highest test temperature for XLPE was 90 °C in this study. It can be seen from Figure 7 that XLPE had the highest breakdown field strength at room temperature, while the breakdown field strength decreased rapidly as the temperature increased. The breakdown strength of XLPE at 70 °C was about 290 kV/mm, and when the temperature increased to 90 °C, the breakdown strength of XPLE decreased to 230 kV/mm. For PP and its composites, it can be seen from Figure 7d that the breakdown strength of PP/SEBS was significantly decreased compared with PP at room temperature. This is mainly because SEBS was a thermoplastic elastomer, and its breakdown strength was lower than that of PP. In addition, it can be seen from the DSC and SEM results that SEBS also destroyed the crystal regularity of PP and also reduced the breakdown strength. As the temperature increased gradually, the breakdown strength of PP decreased significantly, while the breakdown strength of PP/SEBS decreased less. The breakdown strength of PP and PP/SEBS were quite close at 90 °C. The crystallization of the PP phase was gradually melted, resulting in a significant decrease in the breakdown strength of the PP phase. In addition, an aromatic hydrocarbon structure existed at the end of the SEBS macromolecular chain, which is likely to have a function as a voltage stabilizer and to improve insulation properties. The breakdown strength of the three nanocomposites were obviously increased. Many scholars believe that when nano-particles are added into It can be seen from Figure 7 that XLPE had the highest breakdown field strength at room temperature, while the breakdown field strength decreased rapidly as the temperature increased. The breakdown strength of XLPE at 70 • C was about 290 kV/mm, and when the temperature increased to 90 • C, the breakdown strength of XPLE decreased to 230 kV/mm. For PP and its composites, it can be seen from Figure 7d that the breakdown strength of PP/SEBS was significantly decreased compared with PP at room temperature. This is mainly because SEBS was a thermoplastic elastomer, and its breakdown strength was lower than that of PP. In addition, it can be seen from the DSC and SEM results that SEBS also destroyed the crystal regularity of PP and also reduced the breakdown strength. As the temperature increased gradually, the breakdown strength of PP decreased significantly, while the breakdown strength of PP/SEBS decreased less. The breakdown strength of PP and PP/SEBS were quite close at 90 • C. The crystallization of the PP phase was gradually melted, resulting in a significant decrease in the breakdown strength of the PP phase. In addition, an aromatic hydrocarbon structure existed at the end of the SEBS macromolecular chain, which is likely to have a function as a voltage stabilizer and to improve insulation properties. The breakdown strength of the three nano-composites were obviously increased. Many scholars believe that when nano-particles are added into a polymer, a certain amount of deep traps appear in the interfaces, which can capture the space charge and reduce the carrier mobility in the interior material, and this results the improvement of DC breakdown strength [7]. It has been confirmed in the TSDC test above nano-SiO 2 made the deep trap mechanism. Nano-SiO 2 increased the depth and density of the traps inside PP, which enhanced the chance that the trap captured carriers. The breakdown field strength of PP/SiO 2 /SEBS was larger than that of the SEBS/SiO 2 /PP nano-composite. A possible reason for this was that the dispersion of nano-SiO 2 in the SEBS phase was not uniform, and more nano-particles accumulated in the interface of SEBS and PP as shown in Figure 1. Nano-particles need to be uniformly and sufficiently dispersed in the polymer matrix material to produce significant electrical property improvement effects. When most of the nano-SiO 2 particles were dispersed at the interface between the SEBS and the PP phases, the amount of nano-SiO 2 particles in the PP phase and the SEBS phase was insufficient, and the electrical properties of the PP and SEBS phases could not be effectively improved. On the other hand, the carriers could obtain greater mobility and free paths in the SEBS phase, causing a decrease in electrical properties.
By comparing the breakdown strength of 500 kV XLPE materials and PP/SiO 2 /SEBS nano-composites, it was found that the breakdown strength of the self-made nano-composites at 90 • C was higher than that of 500 kV XLPE materials at 70 • C. Thus, 70 • C is the limitative operating temperature of 500 kV XLPE, which implies that the self-made nano-composite has a higher operating temperature.

Conclusions
The dispersion characteristics of nano-particles in ternary composites composed by nano-particles, SEBS, and PP can be effectively controlled by changing the masterbatch system. The effect of nano-particle dispersion characteristics on the DC performance of ternary composites was more obvious than that on the mechanical properties. For PP/SiO 2 /SEBS, more nano-SiO 2 was dispersed in the PP phase, and the optimal space charge suppression effect and the highest DC breakdown strength were obtained. The TSDC results confirmed that the traps in the composites were closely related to the structural defects in the PP phase, and the deep traps introduced by these nano-particles can effectively suppress the formation of space charge so that the electrical properties of PP/SiO 2 /SEBS were better than those of the other two composites. It was found that when the nano-SiO 2 particles were more dispersed in the SEBS phase, the expected electrical property improvement was not obtained. Scanning electron microscopy showed that the nano-SiO 2 particles in the SEBS phase were more dispersed at the interface than in the SEBS matrix, indicating that nano-SiO 2 was poorly dispersed, which may be a reason why the electrical properties of the composite system were not significantly improved. The horizontal comparison with the DC performance of 500 kV HVDC XLPE materials showed that PP/SiO 2 /SEBS nano-composites have better space charge suppression and higher DC breakdown strength at 90 • C, which means that such nano-composites have good potential for use as recyclable HVDC insulation materials with engineering application value.