First principles based computational scheme for designing new SF6 replacements

Searching for new SF6replacement gases from a large number of candidate molecular gases usually requires laborious experimental works with high costs. Here we report an efficient and economic computation-based scheme which can semi-quantitatively predict the thermodynamics and electric properties of new SF6substitute insulation gases with a statistical model combined with quantum chemical calculations. Tests for many representative small molecules with comparison to available experiment data shows that high correlation coefficients can be achieved for our multiple regression analysis. Further investigations for new SF6substitute molecule gases with this model were also given and discussed. Such a framework is expected to be used for the economic and efficient pre-selection for further time-consuming and high-cost experiments.