Study of energetic materials by first-principles methods. Theoretical development and applications.

A postdoctoral position is available in France, at the "Commissariat à l'Énergie Atomique" (CEA, Bruyères le Châtel, south of Paris, France).

We are interested here in the properties of energetic materials based on molecular solids. The goal is to determine their structural and thermodynamic properties (equation of state) using first-principles methods, based on electronic properties, such as Density-Functional Theory (DFT).

In this type of materials, long range van der Waals interactions play a preponderant role. As they are not correctly modeled by the Standard Density-Functional Theory, it is necessary to go beyond the usual approximations [1].

Moreover, the materials of interest are molecular solids based on organic molecules. Therefore they contain a large number of hydrogen atoms whose nuclei are known to have a quantum behavior (zero point energy).

In this project, we propose to use a non-local van der Waals functional in the following framework:

• The ABINIT software [2] will be used developed as part of an high visibility international project, and especially in our group. ABINIT is computational software in the DFT formalism, using a plane wave basis. It has recently been adapted to many-core computer architectures (e.g. Intel MIC).

• The non-local van der Waals functional will be implemented in the complete "Projector Augmented-Wave" (PAW) formalism, which allows obtaining results with a high precision. To date, there is no code implementing a complete "PAW + van der Waals" formalism.

• The calculation of van der Waals forces will be coupled with the "Path-Integral Molecular Dynamics" method, in order to take into account both long-range interactions and the quantum effect of hydrogen nuclei. To our knowledge, this has not yet been achieved on the materials studied here.

The applicant will first have to develop the theoretical framework (complete non-local functional in the PAW formalism); then implement it in the ABINIT code, with particular attention paid to parallelism on many-core architecture. Finally the code will be used for the targeted applications (properties of the molecular solids).

The CEA is a leading national center for numerical simulations and has also access to multiple computational resources in several HPC centers.

Requisites:

The candidate should have obtained a PhD in computational/theoretical physics for less than two years; they should have an interest in the use of electronic structure methods to solve challenging condensed matter problems and a solid background in solid state physics.

Experience in development and good knowledge of computing language(s) (Fortran, C) is a prerequisite. A parallel programming experience is welcome.

The postdoctoral position is available for one year, renewable upon mutual agreement.

Because of restrictions due to European supercomputers access we are looking for candidate nationals from Europe or north-America.

If you are interested and motivated by this project, please send a detailed CV and a letter of application to Jordan Bieder (Jordan DOT Bieder AT cea.fr) and Marc Torrent (marc DOT torrent AT cea.fr), CEA, Bruyères-le-Chatel, 91297 Arpajon - France

[1] Dion et al., Phys. Rev. Lett. 92, 246401 (2004)

[2] http://www.abinit.org

[3] Torrent et al., Comp. Mat. Sc. 42, 337 (2008)