Prediction of Crystal Density of CHONF High Energetic Explosophoric Materials Based on Molecular Volume Method.

Most of the existing studies on crystal density prediction focus on CHON energetic materials. Due to the different application ranges of the formulas, large prediction errors appear in predicting the crystal density of CHONF highly energetic explosophoric material. In order to improve the performance evaluation accuracy, at the B3PW91/6-31g(d, p) evel, the performances of the Rice and Politzer formulas in predicting the crystal density of 56 such materials screened from the Cambridge Crystal Data Center were first systematically studied. In addition, on the basis of the typical parameters related to the molecular surface electrostatic potential, the density prediction formula suitable for this system was constructed by using symbolic regression method and four operators of "addition, subtraction, multiplication, and division". The results showed that the root-mean-square error of the symbolic regression formula was reduced from 0. 106g/cm³ of the Rice formula to 0. 045g/ cm³, and the predicted error of over 50% of materials were less than 0.030g/cm³. The symbolic regression method greatly improved the crystal density prediction accuracy of fluorine-containing highly energetic explosophoric materials. [ABSTRACT FROM AUTHOR]