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

Which molecular properties determine the impact sensitivity of an explosive? A machine learning quantitative investigation of nitroaromatic explosives

“…75 Most of these investigations employed QSPR packages to generate and select sometimes over 1000 molecular descriptors 76 that are afterward reduced by statistical techniques. 77 compositional and topological descriptors gave better results compared to those composed of electronic descriptors, which may be due to their use of the not very accurate semiempirical AM1 method. 79 Keshavarz and Jaafari also employed neural networks, with 275 molecules comprising the training set and 14 explosive molecules the test set.…”

Recent Progress on Synthesis, Characterization, and Performance of Energetic Cocrystals: A Review

“…The synthesis of new EMs is cumbersome due to complex reaction steps and escalated processing time. On the other hand, crystal modification via coating and doping may reduce the sensitivity of EMs but at the cost of higher energy losses and limited applicability [ 3 , 8 ]. As an alternative strategy to produce materials possessing high energy and lower sensitivity, or in other words, to allow for the energy–safety contradiction, cocrystallization is introduced as a promising technique to produce enhanced EMs, generally referred to as energetic cocrystals (ECCs) [ 9 ].…”

Can a shock-induced phonon up-pumping model relate to impact sensitivity of molecular crystals, polymorphs and cocrystals?

“…In an attempt to go beyond mere molecular properties, Bondarchuk 25 has considered solid-state properties of 24 energetic crystals, calculated the pressure at which the electronic band gap is lower than 1 meV, and ended up with a shock sensitivity model with a correlation coefficient R 2 of 0.83. The method is a bit expensive, and considering the bulk modulus instead may be more efficient, as done by Deng et al 26 for 240 nitroaromatics. They achieve a correlation coefficient R 2 of 0.91.…”