Abstract:Machine learning (ML) can map the complex relationship between structures and properties, thus exhibiting great promise in various fields. The mechanical property and the sensitivity are crucial for performance of energetic compounds. However, relationships have not been elucidated between the molecular structure and the mechanical property and between the two important properties. Thus, we introduce here ANN and SISSO methods coupled with selected descriptors involving molecular structures to explore their relationship models for 240 nitroaromatic compounds. In total, six ML models are constructed. ANN exhibits significantly better performances than SISSO for the mechanical property and the sensitivity models. The feature evaluation reveals that the average molecular weight, the number of unsaturated and oxygen-containing groups, hydrophilicity and 3D structure play major roles in influencing the bulk modulus while main contributions to the impact sensitivity come from the structure features involving oxygen-containing group, distribution of atomic properties and hydrophilicity. However, SISSO exhibits better prediction accuracy with R2 of 0.91 than ANN with R2 of 0.89 for the relationship model between the impact sensitivity and the bulk modulus. Furthermore, the SISSO model explicitly gives an inverse-like function relationship for the two important properties. It indicates that the impact sensitivity sharply decreases with increasing bulk modulus when the bulk modulus is small, while an opposite trend occurs in the case of relatively large bulk modulus. These findings could provide useful structural information for improving comprehensive properties of energetic materials.

Impact sensitivity analysis on aliphatic nitro energetic compounds by QSPR

QSPR Study on Impact Sensitivity of Nitroaromatic Compounds

Prediction of Impact Sensitivity of Nitramine with QSPR Method

Prediction of Impact Sensitivity of Nonheterocyclic Nitroenergetic Compounds Using Genetic Algorithm and Artificial Neural Network

QSPR Study on Electric Spark Sensitivity of Nitro Arenes

Prediction of Impact Sensitivity of Nitro Energetic Compounds by Neural Network Based on Electrotopological-State Indices.

Research on the Prediction of Impact Sensitivity of Nitro Energetic Materials

Prediction of impact sensitivity of nitro energetic compounds by using structural parameters

Prediction on impact sensitivity of nitro energetic materials by means of electrotopological state indices

QSPR Study of Correlation Between Impact Sensitivity of M-Nitroaromatics and Electrotopological State Indices

Quantitative Structure Activity Relationship Models For The Prediction Of Rat Lethal Dose 50% Of Aldehydes Based On Heuristic Method And Support Vector Machine

Prediction of thermal stability of some reactive chemicals using the QSPR approach

QSPR Study of N‐octanol/water Partition Coefficient of Some Aromatic Compounds Using Support Vector Regression

QSPR Prediction of GC Retention Indices for Nitrogen-Containing Polycyclic Aromatic Compounds from Heuristically Computed Molecular Descriptors

Prediction of the Aquatic Toxicity of Aromatic Compounds to Tetrahymena Pyriformis Through Support Vector Regression

Probing impact of molecular structure on bulk modulus and impact sensitivity of energetic materials by machine learning methods

Prediction of the Upper Flammability Limits of Organic Compounds from Molecular Structures

Linear and Nonlinear Qspr Models for Predicting Thermal Stabilities of Nitroaromatic Compounds

Prediction of impact sensitivity and electrostatic spark sensitivity for energetic compounds by machine learning and density functional theory

Probing impact of molecular structure on mechanical property and sensitivity of energetic materials by machine learning methods