Ab Initio Quantum Many Body Calculations on Real Materials
This activity has really taken off dramatically over the past year. The possibility of performing ab initio many body calculations has recently been transformed from a dream to a realistic goal. This most exciting development is due to two factors. First, the steady development of the GW and Quantum Monte Carlo methods have made it possible to calculate accurate total energies for realistic systems. Second, the combination of Dynamical Mean Field Theory, a new many body theory, with ab initio electronic structure calculations promises to deliver ab initio quantum many body calculations. For the first time it will be possible to perform materials specific calculations for strongly interacting or correlated electron systems as well as low-dimensional and nanostructured materials of technological interest.
Correlated electron systems pose fundamental challenges in the description of their electronic, magnetic and superconducting properties due to the strong coupling of the charge, spin, structural and orbital degrees of freedom. These challenges do not only go to the heart of the electron correlation problem, but they also occur in, and even more importantly determine, technological materials such as the high temperature superconducting copper oxides and the manganite colossal magneto-resistance materials. The past decade has seen major efforts and progress in the many body theory of correlated electron systems. The complexity of the materials leads to many unknown parameters, which ab initio techniques should be able to deliver. The hugely challenging aim is to incorporate many body techniques into ab initio techniques to create highly accurate first principles modelling. The role of correlation effects becomes more center stage, also driven by nanotechnology's search for low dimensional materials.
For weakly correlated systems, the Density Functional Theory (DFT) in the local density approximation (LDA) has been remarkably successful most notably through its molecular dynamics implementation (Car-Parrinello). However there are some major failures such as in the determination of reaction barriers. For strongly correlated materials (most notably those involving partially filled d- and f- shells), DFT-LDA methods (as well as their extensions like GGA) are not successful, and sometimes suffer major failures even in the predictions of ground-state properties. Furthermore, accurate predictions for excited states certainly requires extensions of these methods. A profound and wide ranging revolution could be expected if correlated electron systems, strong and weak alike, could be modelled with higher accuracy.
The goal of ab initio many body calculations on real materials will be accomplished by three working groups, namely GW Method (spokesperson: Angel Rubio), Quantum Monte Carlo (spokesperson: Claudia Filippi), DMFT (spokesperson: Antoine Georges), with the overall coordination of their activities by Walter Temmerman. The overall coordination will also ensure that other relevant ab initio many body approaches not incorporated in the working groups are represented. This would include Time Dependent DFT (TD-DFT), DFT for superconductors (Kohn-Sham Bogoliubov-de-Gennes), better GGA's, etc..
WG1: GW Method (Spokesperson Angel Rubio, U Pais Vasco, Spain)
WG2: Quantum Monte Carlo Method (Spokesperson Claudia Filippi, U Leiden, the Netherlands)
WG3: Dynamical Mean Field Theory (DMFT) (Spokesperson Antoine Georges, ENS Paris, France)