A novel method to evaluate insulation performance of SF6 alternative gases based on modified GIPF-Local properties model

The development of structure-activity relationships emerges as a crucial strategy to establish the correlations between molecular micro-parameters and the macroscopic dielectric strength, thus markedly improving the gases screening efficiency. However, conventional prediction approaches have failed to characterize the molecule electronegativity accurately, resulting in inadequate prediction precision within extensive gas databases. To address this limitation, we have devised an insulation prediction method that incorporates corrections based on molecular local properties, including local ionization energy and local electron affinity. They enable us to determine both the electron acceptance and donation capacity of the whole molecular orbital electron cloud: a more accuracy assessment of electronegativity. Compared with previous method, our enhanced approach combined local properties with structural and general interaction properties function (GIPF) parameters. It has yielded substantial improvements based on an extensive database comprising 91 gases, as evidenced by an increase in the correlation coefficient from 0.848 to 0.969. Utilizing the modified GIPF-Local properties model, over ten potential gases were found and categorized according to their application. This innovative screening strategy provides crucial insights for exploration and design of new eco-friendly gases as substitute for SF6 used in high-voltage power equipment.