Organisers:

Angel Rubio, Dept. of Theoretical Physics, University of Valladolid, Spain (arubio@mileto.fam.cie.uva.es)

Lucia Reining, LSI-SESI Ecole Polytechnique Palaiseau France (Lucia.Reining@polytechnique.fr)

Giovanni Onida, Physics Department, U. of Roma Tor Vergata, I- 00133 Roma (onida@roma2.infn.it)

(Full programme now available)

The workshop is concerned with ab initio calculations of excited electronic states in condensed matter. One aim is to establish useful connections between methods based on Density Functional Theory (TDDFT) and those using Green's-function techniques. We plan an informal, relatively small workshop with sufficient time to maximise intense discussion.

U.Valladolid - Workshop page (with travel and accommodation information)

List of Participants, Updated August 1998

Links to abstracts included as soon as they become available

(*= to be confirmed)

---

List of abstracts:

available here.

Sessions:

Monday 7 September

9:30 Start
13.20-15.30 lunch break
18:40 End

Tuesday 8 September

09:00 Start
12.50-15.00 Lunch break
19:00 End
20.00 Workshop Dinner

Wednesday 9 September

09:30 Start (including the Special French-Swedish session)
13.05-15.00 Lunch break
17.35 End

Full Programme:

now available here, including links to abstracts of talks.

Inscription fee:

Due to the limited funding available there will be a small inscription fee, in order to maintain the possibility of some (about four) bursary subsidies for young participants (PhD students). The fee has been fixed to 11.700 pesetas (70 euros,about 85 dollars), and is covered by the organization for all the non-excam speakers. Bursary applications must be sent as soon as possible to one of us.

Scientific background

The workshop is concerned with ab initio calculations of excited electronic states, and related spectroscopic properties, in condensed matter applications. In solid state physics, electronic excitations are normally studied using a Green's-function formalism, based on self-energy operators. In the 1960s the theory reached a relatively advanced stage for the homogeneous electron gas, but, due to computational limitations, one had to wait the mid-1980s for the theory starting to be sistematically applied to realistic, inhomogeneous systems. Since then there has been considerable activity in exploring applications of the self-energy theory, notably at the level of the GW approximation, to solids and structures and technical advances in computation at the GW level. Great success has been obtained in calculating the quasiparticle band structure of semiconductors, insulators and simple metals. Now, however, interest is expanding into three main directions: (i) applications should include more complex systems, like surfaces, interfaces, nanostructures, clusters and defects; (ii) more information than just the improvement of bandstructure should be inferred from GW calculations, like updated charge densities or total energies; (iii) besides the one-electron addition and removal energies, which are often well described on the GW level, a realistic description of two-particle excitations is now starting to be feasible, but calculations are still cumbersome.

Since work on all points (i)-(iii) demands for a considerable conceptual and numerical effort, it is important that a continuous contact to other scientific communities is established. In fact, similar approaches are used in different contexts, and similar problems may be solved with different methods. Therefore, in 1997 a CECAM workshop has been organized together with theoretical chemists (''Excited electrons in Molecules, Solids and Atoms"), with obvious benefit for both communities.

In 1998, it is planned to put an accent on links to Time Dependent Density Functional approaches, which work with effective potentials rather than with explicit self-energies (or their derivatives). This link can be regarded as the natural counterpart to the relation with theoretical chemistry: the latter are, roughly speaking, designed for very high precision with corresponding numerical effort, whereas DFT like approaches promise improved efficiency with, at least at present, lower precision. On the other hand, the present knowledge on self-energy calculations and new variational expressions for the many-body total energy and response functions are very relevant for the development of new non-local exchange-correlation functionals in time and space that are of key importance for the practical implementation of TDDFT approaches.

Motivation and objectives

We want to bring together researchers of two different backgrounds: density functional and many-body approaches in order to assess the present status of the computation of excitation spectra. We believe that from the discussion will emerge new ideas, helpful to design the orientation of the future developement of a field of research which still suffers from exploding difficulties, both when going to materials of realistic applications, and when asking for the description of new types of phenomena.

The increased interest of many, often experimental, groups for a reliable description of electronic spectroscopy is a strong motivation for our effort in this direction.

Back to the top (U. Rome Tor Vergata)