Dietrich FOERSTER est membre de l’équipe Théorie de la Matière Condensée, thématique Physique Statistique.

 

Techniques de recherche

Techniques de recherche

Informations à venir…

Thèmes

Thèmes

Informations à venir…

ANR Orgavolt

ANR Orgavolt

A brief summary of the ANR project “ORGAVOLT” 2013-2016

A previous internet site of this project disappeared when the internet site of LOMA, Universite de Bordeaux got revamped. This site is rebuilt here in a preliminary way by the coordinator of the project. Later links towards the authors and their home pages should appear and the coordinator’s text should be replaced by text provided by the various members of the team.

The origin of the project “ORGAVOLT”.

During 2008-2010, at LOMA at the university of Bordeaux and with funds provided by “Agence Nationale de Recherche” (ANR), Peter Koval and Dietrich Foerster developed computational techniques of “lower complexity” for electronic structure. We found algorithms of N3 complexity where N is the number of atoms, both for TDDFT linear response and for Hedin’s GW approximation for molecules and clusters. In both cases, the reduction of complexity used a local orbital representation (LCAO) and a local basis for response that one of us had developped earlier.

Having obtained satisfactory results from GW for lumo and homo levels of large molecules at a modest computational cost, we wished to deal next with organic semiconductors in the organic solar cells pioneered by Heeger and coworkers. Photo emission experiments on these systems reveal standard bands, but with quasi particle peaks softening at higher temperatures.

Experimental bands and band gaps of organic semi conductors agree well with results obtained using Hedin’s GW approximation, but these computations are laborious. We wanted to extend our reduced complexity approach to GW from molecules to crystals in order to help nd improved organic semi conductors for solar cells. In parallel, Remi Avriller of LOMA would construct a simpli ed model of a donor acceptor interface and apply his model to donor acceptor interfaces studied by Mark E. Casida at Grenoble and his group.

A joint project between the teams at San Sebastian, Grenoble and Bordeaux was indeed accepted by ANR “modeles numeriques” with a kick o meeting end of October 2012.

Some results obtained so far

Results obtained at Grenoble

Peter Koval at San Sebatian has worked on a number of related questions and broadened the basis of our project. Amongst his results are a dominant product version of electron energy loss spectroscopy and atom centered basis functions. He fi nds a reduction of the dimension of the basis that should translate into a 100 fold acceleration of our code. He also implemented the Hartree-Fock approximation using a real space representation.

Mathias Ljungberg at San Sebastian/Marburg has implemented an N3 algorithm of Shirley for the Bethe Salpeter equation of clusters, going beyond the Tamm Danko approximation. Using a real time GW starting point, Mathias Ljungberg found many satellites accompanying the quasi particle peaks and disentangling these has been one of the concerns of Mathias Ljungberg.

Rémi Avriller has formulated on a minimal model of the donor acceptor interface in the presence of external electrodes. Excitons and Coulomb interactions are important in organic systems and setting up and solving the equations for an organic hetero interface interface is challenging. So far Remi Avriller focuses on the stability of the interface without external photons.

Saber Gueddida joined the Bordeaux team in the middle of October 2014. Since then he has helped to implement about 30-50% of a GW algorithm for crystals where operations scale as O(NBrillouin) rather than as O(N2B rillouin) with NBrillouin the number of sampling points in the Brillouin zone. This code will now be interfaced with the siesta code of our partner Daniel Sanchez-Portal at San Sebastian.

Some publications of this project

  • P. Koval, D. Foerster, and D. Sanchez-Portal, Fully self-consistent GW and quasi-particle GW for molecules, Phys. Rev. B89, 155417 (2014);
  • P. Koval, M.P. Ljungberg, D. Foerster and D. Sanchez-Portal, Computation of electron energy loss spectra by an iterative method, Nuclear Instruments and Methods in Physics Research B (2014), in press.

Presentations

Collaborations

Collaborations

Informations à venir…

Publications

Publications

Informations à venir…

Curriculum vitae

Curriculum vitae

Informations à venir…

Pas de photo

 

Dietrich FOERSTER

Laboratoire Ondes et Matière d’aquitaine (LOMA)
351 cours de la libération
33405 Talence Cedex

Phone : + 33 (0)5 40 00 25 07
Fax : + 33 (0)5 40 00 69 70
E-maildietrich.foerster@u-bordeaux.fr