Towards predicting and tailoring properties of energetic materials

Abstract

The field of energetic materials (EMs) involves the study of materials (explosives, propellants, and pyrotechnics) that can release a significant amount of energy when initiated. This property renders EMs particularly useful to a wide array of industries including space travel (rocket propellants), mining (demolition charges), and defence applications. The propensity to release a significant amount of energy upon initiation means these materials are inherently dangerous, as such they are subjected to stringent safety requirements, and must be rigorously characterised prior to use. The safety of an EM is often quantified through the evaluation of the sensitivity (propensity to initiate) with respect to different stimuli such as impact, shock, friction, and electric spark. The focus of this work is the impact sensitivity, a solid-state property which can be influenced through changes in the orientation of molecules in 3D space (polymorphism or co-crystallisation) as well as through changing the structure or bonding environment of the molecules comprising the material. Prediction of this metric has been shown in previous work within the group to be computationally achievable for molecular EMs if the crystal structure of the material is known. This is completed through use of the vibrational up-pumping methodology.

Vibrational up-pumping refers to the process by which mechanical impact energy excites delocalised low energy motions in a material and is subsequently channelled upwards into localised molecular vibrations. The vibrational states excited through up-pumping are termed the two-phonon density of states, which represents a measure of how efficiently the initial energy can become trapped on the molecular vibrations. Projection of the twophonon density of states onto the underlying vibrational character yields the up-pumped density which shows a correlation with experimental impact sensitivity. To this date, this method has been applied exclusively to molecular EMs, successfully reproducing experimental sensitivities. While important, focusing on solely molecular materials overlooks those of growing importance such as co-crystals, salts and coordination polymers. Application of the vibrational up-pumping methodology to materials from these areas of growing interest forms the backbone for the work presented in this thesis.

Chapter 2 addresses a number of areas within the vibrational up-pumping methodology that could be improved upon, namely, the generation of consistent phonon density of states (g(w)) spectra as well as partial g(w) spectra, the determination of the location of uppermost phonon frequency (Wmax) and the interrogation of vibrational modes within the solid-state vibrations to track the local modes of vibration (bond stretches and angle bends). Three Python scripts have been developed to address these problems and improve the efficiency and applicability of the process by which the impact sensitivity of an EM is predicted via the vibrational up-pumping methodology.

Chapter 3 focuses on two unexpected findings that had recently come to light in the EMs group at Edinburgh: a co-crystal of FOX-7 with the non-energetic p-phenylenediamine (PPD) that appeared to be more hazardous to mechanical impact than the pure EM, and a new high-pressure polymorph of 3,4,5-trinitro-1H-pyrazole (TNP) that was markedly more sensitive to initiation than the ambient pressure polymorph. For the former study, strong hydrogen bonding interactions significantly altered the molecular conformation of FOX- 7. For the latter, the molecular conformation remained unchanged in the ambient and high-pressure polymorphs, meaning that crystal packing or pressure-induced vibrational mode hardening must account for the increase in mechanical sensitivity. Taken together both studies present challenges for the up-pumping model, which if successful would allow important structure/property connections to be made.

Chapter 4 focuses on salt coordination polymers, all of which present as exceptionally sensitive EMs. The study began with lead azide (LA), which is often used in small quantities as a detonator for a much larger mass of a less sensitive EM. It is well documented that lead has drastic adverse effects to both people and the environment and as such REACH (Registration, Evaluation, Authorisation and Restriction of Chemicals) has issued a ban on the use of LA. This has necessitated the development of a number of ‘green’ copper-containing replacements (DBX-1, DBX-2, DBX-3 and Cu(ADNP)) with comparable impact sensitivity and detonation characteristics such that they could potentially be used as drop-in replacements. This type of EM has not been studied before using the vibrational up-pumping procedure; they present a number of unique challenges, exemplified primarily by the need to separate the lattice modes from the molecular modes, which is a key requirement of the vibrational up-pumping model. In this chapter a full discussion on a range of mechanochemical models are investigated, from simple phonon heating, through to up-pumping and consideration of target (i.e. trigger mode) activation. Culminating in the development of a workflow for the treatment of such materials in the future within the vibrational up-pumping methodology.

In Chapter 5 the emphasis switches towards applying the up-pumping model in a wider capacity to explore the effects of molecular structure on the impact sensitivity of molecular energetics. Here, the investigation centred on a series of chemically related EMs from three common families, namely pyrazoles, tetrazoles and nitrate esters. A number of these materials only differ by the location or substitution of a single functional group, and yet taken together cover a wide range of impact sensitivity response. Successful predictions of their respective impact sensitivities by the up-pumping model would therefore present a unique opportunity to fully explore structure/property relationships, with molecular flexibility, functional group identity and proximity being key structural features to explore. The data set also allowed further exploration of the trigger mode activation introduced in Chapter 4, where only the weakest bonds in the molecules are vibrationally excited by up-pumping. This approach improves the physical basis for impact sensitivity prediction.

Collectively, this thesis explores the application of the vibrational up-pumping methodology to various EMs that present with greater structural complexity than the single-component molecular materials that it was initially designed to model. This work has been aided by the development of supplementary Python scripts which attempt to improve both the efficiency and applicability of the vibrational up-pumping methodology. If successful this work will act to considerably validate vibrational up-pumping, as well as to provide the opportunity to explore in-depth structure/property relationships, to understand the physical basis of impact sensitivity. Such understanding may lead to the development of tailored EMs with desired physical properties in the future.