Reactive neural network potentials: optimization of dataset construction and application to mechanochemical reactions
The spontaneous decomposition of organic molecules during synthesis, handling, or storage causes significant safety issues in the field of energetic materials. Besides thermal activation, recent studies suggest that intramolecular deformations, such as those induced by shock waves, significantly influence chemical reactivity and may alter decomposition mechanisms.
Molecular-level studies of these phenomena present significant challenges because they require both quantum-level accuracy for bond breaking and formation and the inclusion of condensed phase effect.
To bridge this gap, we propose the development and application of machine learning-based interatomic potentials (MLIPs),
In particular, we aim to significantly advance methodologies for building reactive structural datasets, specifically tailored to complex thermal and mechanochemical reactions with multiple decomposition pathways. Leveraging these improved datasets, we will develop MLIPs to study molecular decomposition under varying temperature and pressure conditions. Besides the safety concerns inherent to energetic molecules, the tools and knowledge developed during the project are expected to be of great value to the mechanochemistry community who currently lacks a molecular-level understanding of transformations in mechanochemical systems.
Microscopic description of fission fragment properties at scission
Fission is one of the most difficult nuclear reactions to describe, reflecting the diversity of dynamic aspects of the N-body problem. During this process, the nucleus explores extreme deformation states leading to the formation of two fragments. While the number of degrees of freedom (DOF) involved is extremely large, the mean-field approximation is a good starting point that drastically reduces the DOF, with elongation and asymmetry being unavoidable. This reduction introduces discontinuities in the successive generation of states through which the nucleus transits, since continuity in energy does not ensure the continuity of states resulting from a variational principle. Recently, a new method based on constraints associated with wave function overlaps has been implemented to ensure this continuity up to and beyond the scission (Coulomb valley). This continuity is crucial for describing the dynamics of the process.
The objective of the proposed thesis is to carry out for the first time a two-dimensional implementation of this new approach in order to take into account the whole collectivity generated by elongation and asymmetry DOF. The theoretical and numerical developments will be done within the framework of the time-dependent generator coordinate method. This type of approach contains a first static step, which consists of generating potential energy surfaces (PES) obtained by constrained Hartree-Fock-Bogoliubov calculations, and a second dynamic step, which describes the dynamic propagation of a wave packet on these surfaces by solving the time-dependent Schrödinger equation. It is from this second step that the observables are generally extracted.
As part of this thesis, the PhD student will:
- as a first step, construct continuous two-dimensional PESs for the adiabatic and excited states. This will involve the three algorithms Link, drop and Deflation
- secondly, extract observables that are accessible using this type of approach: yields, the energy balance at scission, fragment deformation and the average number of emitted neutrons. In particular, we want to study the impact of intrinsic excitations on the fission observables, which are essentially manifested in the descent from the saddle point to the scission.
Finally, these results will be compared with experimental data, in actinides and pre-actinides of interest. In particular, the recent very precise measurements obtained by the SOFIA experiments for moderate to very exotic nuclei should help to test the precision and predictivity of our approaches, and guide future developments of N-body approaches and nuclear interaction in fission.
Characterization methods for LMJ’s layered cryogenic targets
Inertial Fusion on the Laser Megajoule facilty requires to form a spherical DT layer at cryogenic temperature. A major topic of interest for fusion experiments is the characterization of the layer quality and thickness. The characterization will be done using two technics : optical shadowgraphy and X-ray phase contrast analysis. A cryostat developed by CEA is already available to work on future target designs and layer characterization.
The objectives of the PhD are to understand and model (theorically and numerically) the physics of the layer observation and to develop the characterization test bench in the cryostat’s environment and the image analysis for the 3D description of the layer.
The student will have to learn to use the cryostat, its command system and its simple actual characterization system. After a bibliography research, he will have to study the physics governing the characterization (multiple reflections, refractions, phase contrast, …) and develop the acquisition and image analysis system allowing the 3D description of the layer using images obtained during experiments with the cryostat. Lastly, the coupling between the command of the cryostat and the characterization will be developed. For all these developments, the student will have access to extensive bibliographical data and the expertise of the host team