Psi-k - Ab initio (from electronic structure) calculation of complex processes in materials


Continuum Quantum Monte Carlo Methods

1. Title and purpose of the workshop:

Title: Continuum Quantum Monte Carlo Methods

Purpose: A hands-on tutorial on QMC methods for continuum systems:
- Variational,
- Projection, in the diffusion/reptation/auxiliary_fields flavors,
- Path_integral.

Organizers:
David Ceperley
NCSA and Dept. of Phys, UIUC, Urbana, Illinois

(see item 9)

Saverio Moroni
INFM, Dept. of Phys, Univ. "La Sapienza", Roma, Italy


Stefano Baroni
SISSA and INFM DEMOCRITOS, Trieste, Italy

2. The scientific content and why a workshop would be useful at this time.

At this time, predictive electronic structure calculations more and more often require a treatment of electronic correlation beyond the existing DFT approximations. QMC has a role to play in this game, but relatively few research groups have QMC expertise. It is important to attract now new researchers into this subject, in order to exploit the potential of available QMC techniques and foster further development.

The topics to be covered in the School include:

Introduction to Monte Carlo, statistics errors, Metropolis method.

Variational Monte Carlo: Correlated trial functions, optimization; molecules, forces;

Diffusion Monte Carlo: Fixed node, Fixed phase;
Electron gas, Twist-average boundary conditions;

Pseudopotentials in QMC;
VMC and DMC electronic structure calculations for solids;
Auxiliary-field Monte Carlo for electronic structure calcs.;

Reptation QMC; response functions, excited states; Maximum Entropy;

Path Integral Monte Carlo; density matrix, pair action, sampling methods;
Bose statistics, superfluidity, condensation.

See also item 5 below.

3. A tentative list of speakers whom it is hoped to have.

A corresponding number of young scientists, with expertise in the field, will help with the hands-on computer labs.

4. The number of participants:

We expect a number of about 40 graduate students and young post-doctoral scientists to participate in the school.

5. Plans for a tutorial element and for attracting new researchers into the subject of the workshop.

The proposed activity is definitely intended as a tutorial. For each of the QMC methods presented, we plan on having elementary lectures to introduce the basic formalism, with hands-on computer sessions using simple codes for test case applications in order to familiarize with concepts and algorithms; at later stage, more advanced topics will be covered, with applications of full-fledged codes to physically relevant problems.

6. Budget, expressed in Euro, and how much is being applied for from the ESF Ψk Programme. In the past, this has been limited to about 9k euro.

The budget to cover the expenses of lecturers and computer lab assistants is estimated in the range 12k-14k euro. Further funding will be devoted to offer lump sum contributions to selected participants, with a total budget of 20k-25k euro. Encouraging applications through partial financial support will arguably make it possible to select very motivated students. Co-funding from Ψk would also raise the international profile of the proposed activity. We apply for a Ψk contribution of 6k euro.

7. A statement about other organisations which will be applied to for
co-sponsorship and additional funding, e.g. any EU RTN Network, CECAM,
CCP-9 in the UK, etc..

The School is presently funded by ICTP (www.ictp.trieste.it) and DEMOCRITOS (http://democritos.infm.it). With the Ψk contribution, it would be possible to offer more support to selected participants (e.g. from Europe, or according to Ψk criterea). DEMOCRITOS is willing to double the contribution given by Ψk to european participants, i.e. for every euro we get from Ψk, we get another euro from DEMOCRITOS. See also item 9 below, concerning US participants.

NOTE: Since we had a strict deadline from the ICTP for the organization of this activity, we had no time to apply for Ψk support before the School was announced (see http://agenda.ictp.trieste.it/age.php?ida=a0332). Therefore it is not possible to mention Ψk in the printed copy of the ICTP poster which is being distributed. Should we get Ψk support, we will make every effort to acknowledge the Ψk contribution in the web announcement as well as in any further advertisement.

8. Where it is hoped the workshop would be held.

The school will be held at ICTP, Trieste, Italy, on Jan 12-23 2004.

9. Plans about inviting and financing any participants from America. See separate announcement directly after this one, about having an appropriate
American co-organiser of the workshop.

David Ceperley is a co-organizer of this activity. Shiwei Zhang is among the invited lecturers. It is anticipated that funds (up to $3000) from the Materials Computational Center at the University of Illinois (funded by the US NSF) will be available to support student and postdoc travel to this school, assuming that Ψk sponsors the school. The funds and procedures are just being put into place in August 2003, so such funding is not certain.

More information at: http://www.democritos.it/events/montecarlo.php