Mikael Unge

Communication: Band bending at the interface in polyethylene-MgO nanocomposite dielectric

Polymer nanocomposite dielectrics are promising materials for electrical insulation in high voltage applications. However, the physics behind their performance is not yet fully understood. We use density functional theory to investigate the electronic properties of the interfacial area in magnesium oxide-polyethylene nanocomposite. Our results demonstrate polyethylene conduction band matching with conduction bands of different surfaces of magnesium oxide. Such band bending results in long range potential wells of up to 2.6 eV deep. Furthermore, the fundamental influence of silicon treatment on magnesium oxide surface properties is assessed. We report a reduction of the surface-induced states at the silicon-treated interface. The simulations provide information used to propose a new model for charge trapping in nanocomposite dielectrics.

First-principle simulations of electronic structure in semicrystalline polyethylene

In order to increase our fundamental knowledge about high-voltage cable insulation materials, realistic polyethylene (PE) structures, generated with a novel molecular modeling strategy, have been analyzed using first principle electronic structure simulations. The PE structures were constructed by first generating atomistic PE configurations with an off-lattice Monte Carlo method and then equilibrating the structures at the desired temperature and pressure using molecular dynamics simulations. Semicrystalline, fully crystalline and fully amorphous PE, in some cases including crosslinks and short-chain branches, were analyzed. The modeled PE had a structure in agreement with established experimental data. Linear-scaling density functional theory (LS-DFT) was used to examine the electronic structure (e.g., spatial distribution of molecular orbitals, bandgaps and mobility edges) on all the materials, whereas conventional DFT was used to validate the LS-DFT results on small systems. When hybrid functionals were used, the simulated bandgaps were close to the experimental values. The localization of valence and conduction band states was demonstrated. The localized states in the conduction band were primarily found in the free volume (result of gauche conformations) present in the amorphous regions. For branched and crosslinked structures, the localized electronic states closest to the valence band edge were positioned at branches and crosslinks, respectively. At 0 K, the activation energy for transport was lower for holes than for electrons. However, at room temperature, the effective activation energy was very low (∼0.1 eV) for both holes and electrons, which indicates that the mobility will be relatively high even below the mobility edges and suggests that charge carriers can be hot carriers above the mobility edges in the presence of a high electrical field.

Conductivity simulations of field-grading composites

The electrical conductivity and the percolation threshold of field grading polymer composites intended for high voltage applications were examined with representative elementary volume simulation methods based on percolation threshold modeling (PTM) and electrical network modeling (ENM). Comparisons were made with experimental conductivity data for SiC-EPDM composites with spherical and angular particles, using different filler fractions and electrical field strengths. With a known conductivity of the filler particles (powder), the simulations could predict the percolation threshold and the composite conductivity as functions of the electrical field for a wide range of SiC-filler fractions. The effects of morphology, dispersion and filler shape were examined and the simulations were able to explain the experimental difficulty of reaching sufficient reproducibility when designing composites with filler fractions close to a percolation threshold. PTM of composites containing hard-core/soft-shell spheres revealed a y = (a + bx)(−1/c) relationship (R 2 = 0.9997) between filler fraction and relative soft-shell thickness.

Ab-initio Modeling of Interfacial Region in Nanocomposite Dielectrics

The interfacial region between a base matrix and nanoparticles in nanocomposite dielectrics is often referred to as the main cause of good performance of nanocomposites as insulating materials. In the present work we compare electronic structure of the interfacial region in the polyethylene magnesium oxide nanocomposite with the electronic structures of its bulk constituents. The calculations were performed with density functional theory, the LDA and AM05 functionals were used. Hydroxylated, silanol-terminated (-SiOH) MgO surfaces and an interface (a surface with grafted through Si alkane chains) were studied. Investigation has shown the presence of surface states in untreated (hydroxylated) MgO (111) surface, while for both silanized surfaces these states are removed. It results in 1.7 eV higher band gap energy compared to the untreated case. Untreated regions present in treated nanoparticle are proposed to behave as traps for electrons.

Electronic properties of magnesium oxide - polyethylene interface

We use density functional theory to calculate band offsets for different configurations of magnesium oxide - polyethylene interface. Our study shows that band bending occurs at the contact between the polymer and the MgO particle. As a result, the surface of the MgO nanoparticle introduces deep traps up to 2 eV in the material. Also, the calculations indicate the possibility of a double layer formation within the nanoparticle.

ZnO-polyethylene interface: Band alignment

Zinc oxide-polyethylene nanocomposites have a potential to be used for high voltage insulation. Here, we investigate electronic properties of zinc oxide-polyethylene interface using density functional theory. Interface states up to 1 eV below conduction band edge are found in polyethylene. Our results suggest that the energy bands of ZnO and polyethylene align to comply with vacuum level equality. We use our findings to establish mechanisms of band alignment at the interfaces between polyethylene and crystal materials.

High field ion mobility in dielectric polymers: A molecular dynamics study of water in poly(dimethylsiloxane)

Detailed understanding of charge transport phenomena in dielectrics is important in order to understand the influence of chemical composition on the dielectric properties. Material simulation methods such as density functional theory (DFT) and molecular dynamics (MD) can be used to calculate basic properties of the material. Here we focus on mobility of ions at high fields by applying MD simulations. In a recent study of ionic mobility in polyethylene, it was shown that ion mobility starts to deviate from the Einstein relation at high fields, resulting in a linear increase of the mobility [1]. We argue that this is due to conformational change of the surrounding polymers through elastic scattering which facilitates and thus accelerates the charged molecule. Here we investigate this effect of water in poly(dimethylsiloxane) (PDMS).

Electron mobility edge in amorphous polyethylene

The conduction mechanism in a material is to a large extent determined by the nature of the electronic states. Localized states give hopping conduction and delocalized states band transport. In amorphous materials there may be a transition from localized states at the band edges to delocalized states higher up in the band. Here we use linear scaling density functional theory and a percolation method to determine electron mobility in amorphous polyethylene. The electron mobility edge is determined to ~ 0.2 eV.

Development of simulation methods for dielectrics

Small organic molecules, e.g. antioxidants, byproducts or additives, are often present in polymeric materials that are used as insulation in power products. The impact on the performance of the insulation due to these molecules can be significant. In the current study the impact of a polyol (mesoerythritol) on a polypropylene system was studied with the focus on the dielectric permittivity and losses. Material modeling was applied in order to compute these properties for various polyol concentrations. Experiments results were in order to verify the findings of the simulations and the results show that the agreement is very good, especially for the simulations performed in high-resolution. With a so-called multi-scale scheme almost equivalent results was obtained but at significantly lower computational cost. The long term goal is to develop an effective and fast simulation tool.