Anthrone/XLPE: an adaptive charge capture intelligent insulation material for advanced electric power transmission

Introduction

Electric energy is the most widely used form of energy globally, yet it cannot be stored in large quantities. Ensuring the reliability of the power transmission channels between power plants and end users is crucial. Power cables are the primary components of the transmission and distribution network. High voltage (HV) power cables play a crucial role in hydropower, wind power, solar energy, and other forms of distributed renewable energy, as well as cross-sea power supply, urban power supply, and permafrost regions. The unique advantages of HV power cables are irreplaceable, and their usage continues to increase annually. The insulation deterioration of alternating current (AC) cables and direct current (DC) cables results from the concentration of electric field caused by local defects.

In the early 1960s, HV cables started using extruded polyethylene (PE). Many 13 kV and 23 kV underground residential distribution cables in the United States transitioned from oil-soaked paper insulation and lead shielding to PE. This shift resulted in significant cost savings in cables, connectors, terminals, and installation processes¹. In 1963, cross-linked polyethylene (XLPE) was introduced as an insulating material for medium-voltage cables, surpassing PE in both electrical insulation and mechanical strength². However, within a few years, many reports of electrical tree damage in XLPE-insulated cables emerged. In 1968, tree-like aging was identified as the main factor compromising the reliability of extruded cable systems. This discovery prompted further research into the resistance of XLPE insulation against electrical tree aging³. In 1973, the formation of electrical trees stemmed from electromechanical disturbances due to Maxwell stresses was proposed⁴. Later, in 1980, the resistance of PE against treeing was improved by introducing three organic compounds, resulting in a two-to-four-fold increase in its treeing onset voltage⁵. In that same year, researchers discovered that organic semiconductors contributed to improved conductivity and an increased the electrical tree initiation voltage (TIV)⁶. The enol form of acetylacetone played a crucial role. Its unique structure, with conjugated configurations and specific ketone groups, enhanced the tree onset voltage by enabling energy dispersion.

Conducting density functional theory (DFT) calculations were performed to determine the electron affinity, ionization potential, highest occupied molecular orbital (HOMO), and lowest unoccupied molecular orbital (LUMO) energy gap of fifty-four organic semiconductor materials⁷. A linear relationship was subsequently observed by correlating the computed data with the electrical TIV of organic semiconductor/PE. Later, similar conclusions were reached through comparable DFT calculations and electrical tree branch experiments⁸. Many researchers have focused on studying the effects of organic semiconductors on electrical tree properties and have made progress^{9,10,11,12,13,14,15,16}. Additives of organic semiconductors can effectively increase the electrical TIV of XLPE^17,18,19,20. Some studies indicate that organic semiconductors can suppress charge injection and charge accumulation^14,21,22. Organic semiconductors that can improve the performance of insulating materials share two key characteristics: first, they have a conjugated structure, and second, they exhibit higher electron affinity than the insulating material.

Previous studies have utilized organic semiconductors to introduce traps, aiming to increase AC treeing initiation voltage and suppress DC charge accumulation. Organic semiconductors are added at relatively low concentrations (~1%) and are uniformly distributed within the XLPE. However, the low probability of overlap between defect locations and the distribution of organic semiconductors, combined with the low concentration, means that organic semiconductors may not precisely address local defect issues. In this study, the electrophoretic migration characteristics of organic semiconductors were utilized to precisely reinforce weak points associated with local defects in XLPE. Anthracene (EN) and anthrone (ET) possess conjugated structures, and DFT calculations reveal that both EN and ET exhibit high electron affinity. Therefore, organic semiconductors, EN, and ET were selected. Samples were prepared by high temperature crosslinking method, named EN/XLPE and ET/XLPE, respectively. The electrical tree, AC breakdown strength (BDS), and space charge characteristics were then tested. Tools such as DFT calculations, dielectric spectrum, ultraviolet photoelectron spectroscopy (UPS), ultraviolet-visible spectroscopy (UV-vis), differential scanning calorimetry (DSC), X-ray diffraction (XRD), and scanning electron microscope (SEM) were employed to investigate the insulation mechanisms through which these additives affected the insulating properties of XLPE. The results of the study demonstrated that organic semiconductors improved insulation in ET/XLPE, increasing TIV and reducing internal electric fields. These findings highlight the potential as promising additives for advanced insulating materials.

Results

Space charge

The accumulation of space charges can induce localized electric field distortion, accelerating insulation deterioration and ultimately lesding to breakdown. Figure 1 illustrates the time-dependent distribution of space charges in XLPE, EN/XLPE, and ET/XLPE at 30 °C under a −50 kV/mm uniform electric field. In XLPE and EN/XLPE, charges transport from the anode to the cathode under a high electric field and take approximately ten minutes to reach equilibrium, with homopolar charge accumulation near the electrode (Fig. 1A1–A3 and B1–B3). The electric field intensity is higher in the middle along the thickness direction and lower towards both sides. Interestingly, a substantial injection of charges occurs immediately upon applying voltage. Most of the injected charges are confined near the interface between the electrode and ET/XLPE, while a small portion of the charges migrates towards the cathode (Fig. 1C1–C3). These charges stabilized within approximately five minutes, resulting in a significant accumulation of homopolar charges near the anode electrode. A small quantity of heteropolar charges accumulates at a distance of 15 μm from the anode and near the cathode. In the thickness direction, the electric field intensity is lower in the middle and higher on both sides. The peak electric field measures 59.6 kV/mm at 165 μm for XLPE, and 56.2 kV/mm at 150 μm for EN/XLPE. The minimum electric field measures 41 kV/mm at 84 μm, while the peak electric field measures 63 kV/mm at 196 μm for ET/XLPE. The charge transport and electric field distribution characteristics of ET/XLPE differ from those of XLPE and EN/XLPE.

Electrical insulation strength

The electrical TIV is a crucial indicator of the insulation strength of materials. Figure 2A and Supplementary Table. 1 present the Weibull eigenvalues of TIV for XLPE, EN/XLPE, and ET/XLPE at 30 °C. The TIV of XLPE in the control group is 7.8 kV. In contrast, the TIV of EN/XLPE increases to 12 kV, reflecting a 54% increase, while the TIV of ET/XLPE is 17.4 kV, indicating a remarkable 123% enhancement, far outperforming existing dielectric polymers and polymer composites (Fig. 2B and Supplementary Table. 2)^{17,18,19,20,23,24,25,26,27,28}. As shown in Fig. 2C and Supplementary Fig. 1, under the influence of the electric field, the electrical tree in XLPE continues to grow, forming a bush-like morphology²⁹, with a length of 2182 μm. EN/XLPE and ET/XLPE grow slowly and enter a stagnation phase after approximately 600 seconds. EN/XLPE exhibits a branch-like morphology with a length of 434 μm, while ET/XLPE displays a clustered morphology with a length of 378 μm. The organic semiconductor effectively suppresses electrical tree development. The PD results show a similar trend to TIV (Supplementary Fig. 2).

AC BDS reflects the capability of insulation materials to withstand high electric fields. Figure 2D and Supplementary Table 3 present the Weibull eigenvalues of AC BDS of XLPE, EN/XLPE, and ET/XLPE at 30 °C. The AC BDS of XLPE is 102 kV/mm. In comparison, the AC BDS for EN/XLPE is slightly lower at 97.4 kV/mm, representing a 4.5% decrease, while the shape parameter increases by 2.7. On the other hand, ET/XLPE exhibits a higher AC BDS of 111.5 kV/mm, indicating a 9.3% increase, while its shape parameter decreases by 2.3.

Physicochemical properties

The trap depth was calculated based on surface potential measurements (Fig. 2E). XLPE exhibits two pronounced trap distribution peaks: a high shallow trap density concentrated around 0.93 eV, and a low deep trap density concentrated around 1.04 eV. EN/XLPE shows two distinct trap distribution peaks: a low shallow trap density concentrated around 0.84 eV, and a higher deep trap density concentrated around 0.92 eV. The trap depth of ET/XLPE exhibits only one trap distribution peak, a high-density deep trap concentrated ~1.01 eV. In general, EN/XLPE introduces many shallow traps, while ET/XLPE introduces a large number of deep traps compared to XLPE.

The energy band diagrams for EN and ET are shown in Fig. 2F. Φ_EA refers to electron affinity, and E_vac denotes the vacuum level. The ionization potential (Φ_IP) was obtained from the UPS measurements (Supplementary Fig. 3). The acquired Φ_IP enables the determination of the HOMO energy level, while the band gap (Φ_g) is obtained from the UV-vis measurements (Supplementary Fig. 4). Thus, the position of LUMO level, calculated as the difference between the HOMO energy level and the band gap, is 2.42 eV for EN and 3.33 eV for ET, which are also their respective Φ_EA values. The energy values are reported as absolute values relative to a vacuum. The energy band calculation results of EN and ET are shown in Supplementary Table 4, and the trends of experimental results are similar.

The results of DSC and XRD are shown in Supplementary Fig. 5 and Supplementary Table 5. The crystallinity, calculated by DSC, is the ratio of the enthalpy of fusion and the enthalpy of complete crystallization. The heat of crystalline fusion is represented by the area under the heating curve. The complete crystalline melting heat for XLPE is 290 J/g³⁰. The crystallinities of XLPE, EN/XLPE, and ET/XLPE are 40%, 46%, and 46%, respectively. Peak 1 at 19.8° is an amorphous diffraction peak, while peaks 2 at 21.3° and 3 at 23.6° are crystalline diffraction peaks for the 110 and 200 crystal planes. The proportions of crystallinity calculated from the sum of the integral areas of the XRD are the ratios of the integral areas of the crystalline peaks to the integral areas of diffraction peaks³¹. The crystallinities for XLPE, EN/XLPE, and ET/XLPE are 40%, 45%, and 44%, respectively.

Discussion

The phenomenon of heteropolar charge accumulation in insulating material

XLPE and EN/XLPE are typical examples of homo-polarity charge injection and accumulation. The shallow traps introduced by EN promote charge transport and reduce the accumulation of homo-polarity charges within XLPE. Notably, the charge transport characteristics of ET/XLPE differ from those of XLPE and EN/XLPE, resulting in the formation of uncommon hetero-polarity charge accumulation near the anode (Fig. 1C-2). After applying voltage, charges are injected into the XLPE from the anode. Homo-polarity charges are captured by the deep traps formed by ET and XLPE, while others enter the interior of XLPE and migrate toward the cathode. As the net charge injection decreases, equilibrium is reached ~5 minutes, and hetero-polarity charge accumulation is observed near both the anode and cathode. Ultimately, a distinct positive-negative-positive charge region is formed in the thickness direction of XLPE (Fig. 3A).

**Fig. 3: Space charge transport mechanism of anthrone (ET)/XLPE.**

Mechanism of charge regulation

Around 1–5 minutes after applying voltage, the concentration of ET near the anode is low due to its uniform distribution of ET in XLPE (Fig. 3B and Supplementary Fig. 6A). The deep traps formed by ET and XLPE can capture only a portion of the charges injected from the anode, while the remainder enters the interior of XLPE. The electric field formed by the captured positive charges weakens the interface electric field (between the positive charge trapping region and the anode) and strengthens the internal electric field of XLPE (between the charge trapping region and the cathode). This inhibits the net injection of charges from the anode, promoting the electrophoretic migration of ET, which accumulates near the anode (Fig. 3C and Supplementary Fig. 6B). As the number of deep traps introduced by ET increases, the accumulation of interface charges grows, and the electric field formed by the trapped charges further inhibits the injection of charges from the anode. The net charge injection in the negative charge region is less than the migration amount towards the cathode, resulting in the formation of a negative charge region. The negative charge region establishes an inherent electric field that opposes the applied electric field direction, reducing the electric field intensity experienced by the insulation material internally. This serves as a new protective mechanism for XLPE. The improvement in the DC BDS of ET/XLPE further verifies this conclusion (Supplementary Fig. 7). The discovery of this composite material is significant for the development of ultra-HV cables.

Mechanism of insulation enhancement for charge regulation

EN and ET enhance insulation strength by changing trap distribution and crystallinity, thereby reducing the energy or accumulation characteristics of electrons. EN/XLPE and ET/XLPE have higher TIV than XLPE. The analysis focuses on two aspects: firstly, high-energy electron absorption and energy dissipation. The LUMO and HOMO correspond to the conduction and valence bands in polymers. EN and ET, as small molecular organic semiconductors, introduce certain electronic orbits between the LUMO and HOMO of XLPE. The introduced electron orbitals are located between the LUMO of XLPE and the LUMO of organic semiconductors, or the HOMO of XLPE and the HOMO of the organic semiconductors (Fig. 2F). The discontinuity of the introduced electronic orbitals in XLPE creates traps in the polymer. Polymers inherently lack the capacity to capture electrons, while traps are more likely to seize them. The addition of EN introduces shallow traps, while ET introduces a greater number of deep traps (Fig. 2E). EN and ET preferentially trap high-energy charges and absorb and release energy through the conversion of ground and excited states (Fig. 4a)^32,33, dissipating the energy of high-energy electrons into weaker forms such as fluorescence and heat. Hence, EN and ET equalize the electric field by trapping high-energy electrons and reducing charge accumulation. Secondly, increasing crystallinity increases the total energy of the grain boundary barrier, impeding charge movement. The electron conductance in polymers is mainly composed of intramolecular conduction band transport and intermolecular jump conduction. As shown in the Supplementary Table 5, the crystallinity of EN/XLPE and ET/XLPE has improved. The increase in crystallinity will increase the grain boundary barrier that the jump conductance needs to cross, obstructing charge movement. Reduce the energy obtained by electrons under the influence of an external electric field.

**Fig. 4: Organic semiconductor/XLPE electric tree enhancement mechanism.**

Polar organic semiconductors enhance charge regulation

On the one hand, existing research indicates a positive correlation between the electron affinity of organic semiconductors and TIV of electrical treeing, while an inverse correlation is between the energy gap and the TIV³⁴. Compared to EN, the molecular structure of ET contains a polar carbonyl group (Supplementary Table 4). Hence, ET has higher electron affinity and a narrower energy gap, providing it with a greater capability to capture high-energy electrons (Fig. 2F). On the other hand, the adaptive charge capture (ACC) of organic semiconductors containing polar groups. Due to the low molecular weight and short chain length of organic small molecules, dielectrophoretic effects occur under a gradient electric field³⁵. Before polarization, ET is evenly distributed (Fig. 4e). After polarization, ET migrates by dielectrophoresis and gathers near the needle tip, where the electric field strength is highest, and the insulation strength is weakest (Fig. 4f). Organic semiconductors adaptively reposition themselves as the electric field distribution alters; they relocate to the insulating weak point with high field strength, inhibiting further destruction of the insulating material by high-energy electrons through ACC. However, the ACC phenomenon is not observed in EN/XLPE (Supplementary Fig. 8). Therefore, the addition of ET precisely limits charge transfer at the insulating weak point, increases the initiation threshold of discharge for electrical trees, and reduces the maximum discharge during the electrical tree growth process (Supplementary Fig. 2). ET/XLPE exhibits a higher TIV than EN/XLPE.

The weak point breakdown of insulating material

Under a strong electric field, field-emitted or thermally emitted electrons are generated, leading to a certain number of electrons in the dielectric’s conduction band. The molecular chains of XLPE absorb energy and vibrate when they are hit by high-energy electrons, causing the electrons to lose energy. When the energies are balanced, no irreversible damage occurs to the molecular chain. If the energy of the electron exceeds what the molecular chain can absorb, the molecular chain breaks, leading to a breakdown. EN and ET are small molecule materials in organic semiconductors. Both have conjugated π bonds in their molecular structures, providing vacant orbitals for electrons, serving the function of absorbing high-energy electrons and dissipating electron energy (Fig. 4a). As the voltage rises, if electrons’ energy obtained through acceleration in the field surpasses the energy from the organic semiconductor’s trap energy, the organic semiconductor can no longer trap electrons. Simultaneously, its conjugated structure provides an empty orbit for electrons to accelerate in the electric field. This indicates that the organic semiconductor becomes the conductor at a certain voltage, providing beneficial conditions for electron acceleration or energy absorption.

Charge transport during breakdown

ET contains an additional carbonyl group with stronger polarity compared to EN. As a result, ET possesses greater electron affinity and a smaller band gap than EN (Fig. 2F and Supplementary Table 4). This result is confirmed by the surface potential measurement experiments (Fig. 2E). EN/XLPE introduces a significant number of shallow traps compared to XLPE, while ET/XLPE introduces a substantial number of deep traps. Therefore, electrons in ET/XLPE require more for energy level transitions than in XLPE, while EN/XLPE requires less. Compared with XLPE, the electron transport in EN/XLPE is less hindered by shallow trap, while the electron transport of ET/XLPE is more hindered by deep trap. The electron transport path becomes longer under the application of an applied electric field (Fig. 5). This is verified in the DC conductance test, where the order of DC conductivity from highest to lowest is EN/XLPE > XLPE > ET/XLPE (Supplementary Fig. 9). Thus, EN/XLPE has a lower BDS than XLPE, while ET/XLPE has a higher BDS than both XLPE and EN/XLPE.

**Fig. 5: Schematic diagram of the breakdown path in organic semiconductor/XLPE.**

Conclusions

In this work, organic semiconductors were used to directionally strengthen insulation weak points through dielectrophoresis, significantly increasing the composite TIV. The TIV of ET/XLPE increased by 123%. By constructing a charge barrier that forms a reverse electric field, the level of charge injection and accumulation in XLPE was reduced, decreasing the internal electric field strength of XLPE by up to 18%. The results of this study affirm the directional reinforcement effect of organic semiconductors. Repurposing organic semiconductors as additives to polyolefins instead of using them as active materials in optoelectronic devices, may greatly improve the application range of this promising class of materials. This is crucial for the successful commercialization and practical application of organic semiconductors in the field of insulating media.

Methods

Sample preparation

All reagents were ready for use without the need for further preparation. Low-density polyethylene (LDPE) produced by Yangzi BASF Corporation, was used as the base material. The LDPE had a melt flow rate of 3.3–3.6 g/10 min, a density of 0.922 g/cm², and belongs to grade 2220H. Its melting point was 110 °C. EN (CAS: 120-12-7) and ET (CAS: 90-44-8), were purchased from Macklin. The crosslinking agent, dicumyl peroxide (DCP, CAS: 80-43-3) was procured from Aladdin. The purity of the additives used exceeded 98%. The sample preparation method is provided in Supplementary Note 1.

Electrical performance test

Electrical tree-partial discharge (PD) joint measuring device was used (Supplementary Fig. 10). It represents a schematic of the electrical tree and PD measurement system. The setup comprises a real-time microscopic imaging system and a PD measurement device. The apparatus conformed to the IEC 60270 standard³⁶. The electrical tree experiment used a pin-plate electrode model (Supplementary Fig. 11). The equipment parameter and detailed testing processes are provided in Supplementary Note 2.1.

The detailed process of the AC BDS experiment is provided in Supplementary Note 2.2. The BDS was statistically evaluated using a bi-parametric Weibull distribution function³⁷, with the failure probability of 63.2% serving as the characteristic BDS. The transport of space charge within XLPE was investigated using the pulsed electroacoustic method³⁸. Equipment and experimental details are provided in Supplementary Note 2.3. Equipment and experimental details for surface potential measurements are provided in Supplementary Note 2.4. Finally, the trap distribution of the insulating material was calculated through an inversion process^39,40. The DFT calculation process is provided in Supplementary Note 2.5.

Physicochemical properties test

Test instruments and conditions for DSC, XRD, UPS, UV-vis, SEM, and conductivity are provided in Supplementary Notes 2.6–2.11, respectively.