Micro-scale crystallization thermodynamics study of typical energetic compounds integrating optofluidics and machine learning

With the aim of investigating the changing law of crystallization driving force of typical energetic compounds under micro-scale crystallization conditions, a thermodynamic parameter determination method based on optofluidics was proposed. Aimed at nitro, nitramine and nitrate explosives in energetic compounds, hexanitrostilbene (HNS), cyclotetramethylene tetranitramine (HMX) and pentaerythritol tetranitrate (PETN) were selected as representatives, the solubility of the three kinds of energetic compounds in their respective commonly used solvents (HNS: in DMF, DMSO, NMP; HMX: in DMF, DMSO, CYC; PETN: in DMF, DMSO, EAc) at different temperatures were determined. Furthermore, microfluidics and machine learning were combined, the solubility data of the explosives were processed using BP neural network. Moreover, the metastable zone widths of HNS, HMX and PETN in each solvent were determined using on-line Raman technique. Additionally, crystalline thermodynamic parameters such as solid–liquid interfacial tension, crystal surface entropy factor, enthalpy of dissolution and etc. were calculated for each system.