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

• Published in: RSC advances 2022-10, Vol.12 (48), p.31282-31292
• Main Authors: Bidault, X, Chaudhuri, S
• Format: Article
• Language: English
• Subjects: Chemistry; Coupling (molecular); Criteria; Crystals; Hammers; HMX; Impact tests; INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY; Phonons; Prediction models; Pumping; Sensitivity
• ISSN: 2046-2069; 2046-2069
• DOI: 10.1039/d2ra05062e

Impact sensitivity engineering of high-energy molecular crystals requires accurate predictive models. For this purpose, the promising multi-phonon based approach is selected, assessing a bit more its strengths and weaknesses. Presently used with high-quality phonon calculations of 22 molecular crystals, using a physics-based criterion to determine the phonon bath extent, the resulting intrinsic shock sensitivity index (SSI) is compared to the most common marker of impact sensitivity, h 50 , as determined from drop-weight impact tests. Selecting a data subset from experiments performed under very similar conditions (2.5 kg hammer with grit and 30-40 mg samples), the model can predict h 50 values for mono-molecular crystals with very good accuracy, including the ability to discriminate the polymorphs of HMX and CL20. This very good agreement validates an initial indirect up-pumping mechanism occurring under these conditions, where the doorway modes also interact with the phonon bath. However, the phonon bath criterion for mono-molecular crystals does not transfer well to cocrystals. Owing to the vibrational coupling of the co-molecules, it seems a broader phonon bath should be considered. Additionally recalling experimental uncertainty and various experimental factors affecting h 50 values for a given compounds, we recommend that the density of the sample, granularity and morphology be systematically considered and reported along with measurements, which will in turn allow for more systematic data and predictive capabilities for sensitivity models. The model predicts impact sensitivity of mono-molecular crystals with 90% accuracy on a log-scale, with the ability to discriminate polymorphs, but it does not transfer well to cocrystals.