Aucun titre de diapositive

1
Virtual Laboratory for Nanotechnology Applications
1) Massimo Celino, 2) Giulio Gianese, 1) Simone
Giusepponi, 1) Michele Gusso, 1,2) Vittorio
Rosato

1. ENEA, Ente per le Nuove Tecnologie, lEnergia e
   lAmbiente, Casaccia Research Centre, Via
   Anguillarese 301, 00123 S. Maria di Galeria
   (Roma)
2. Ylichron srl , Casaccia Research Centre, Via
   Anguillarese 301, 00123 S. Maria di Galeria (Roma)

Abstract
The availability of high performance computing
platforms has opened the way to their use in
Nanotechnology applications, also of interest for
many industrial sectors. A virtual laboratory is
a common environment where scientists and
researchers from universities and industries can
work together by sharing competences, software
and specialized services. An example is the
ENEA-GRID environment for Materials Science
applications. Examples and main results are
reported.
Introduction
Nanotechnology applications
Only recently numerical modelling has become a
critical issue for industrial RD. Thanks to the
advent of high performance computing platforms
(the so-called supercomputers), mathematical
and physical models can be deployed for the study
and the design of objects which have a strong
interest even for industrial applications. An
example, among others, is the solution of
Schröedinger equations in quantum Physics that
allows the accurate description of materials at
the atomic scale. For many years, this equation
has been numerically intractable. This has
inhibited its deployment for practical purposes.
Due to significant improvements on both software
and hardware, the study and the visualization of
complex physical, chemical and engineering
processes has become possible with significant
advantages for science and technology. Several
technical fields have found, in the numerical
approach, an effective methodology for the
improvements of knowledge and the reduction of
RD costs. Current generation of supercomputers,
due to their intrinsic complexity and fast
evolution, needs large investments and can be
managed only by specialized professionals. In
this field ENEA has a long lasting experience
coming from the field of nuclear research done in
the sixties. Moreover recently ENEA has received
a significant financial support (CRESCO Project)
for the acquisition of a last generation
supercomputer, now installed at the Portici ENEA
Centre, nearby Naples (Italy). The CRESCO
platform is not a standalone platform, but has
been inserted into the ENEA-GRID environment
which is integrated and opened to the most used
european GRID infrastructures. The CRESCO
platform is currently the 125th most powerful
computational platform in the world, with respect
to both peak and sustained performances.
Materials science has received significant
benefits by the availability of powerful
supercomputers. These infrastructures allow the
simulation of nanotechnology materials at the
atomic scale with sufficient accuracy to impact
on real applications. The main numerical
technique used in this field is Molecular
Dynamics (MD hereafter). In this approach, the
equations of motion of an interacting N-particles
system (ions, electrons) are iteratively solved.
The behaviour of these particles can be simulated
in different aggregation conditions (liquid,
solid, amorphous etc.), at different
thermodynamic conditions and under the effect of
different external conditions and constraints.
The evaluation of the dynamic behaviour of each
particle of the system allows to evaluate its
macroscopic properties and its global response to
external conditions or perturbations. This is a
key point because this opens the way to the
evaluation of a large set of macroscopic
quantities, often of interest for the industrial
applications. Simulations can be seen as thought
experiments which can be performed on a given
system, under very well controlled conditions.
Organic-inorganic materials
Green spheres are carbons, blue nitrogens and red
oxygens. The numerical experiment is performed in
water (water molecules are not displayed).
Spheres represent atoms and sticks are the bonds
among the atoms.
Hydrogen storage for automotive applications
The increasing request for downsizing in
microelectronics, implies the development of new
materials able to mimic processes which
spontaneously occur in organic or biological
domains. New frontiers of research Atomic level
characterization of the adhesion of a polypeptide
molecule on a carbon nanotube.
Multi-scale MD simulations enable an accurate
description of the adhesion of a peptide on a
carbon nanotube. From an experimental point of
view, there is not a clear evidence on the
peptide region which mainly affects its binding
property with a given substrate and which are the
energies involved. Simulations, as a powerful
probe, can selectively test the adhesion of each
part of the peptide and compute the electronic
modifications.

• On the left scanning electron microscope image of
  the hydrogen desorption from magnesium hydride
  (MgH2), thanks to dr. A.Montone, ENEA. On the
  right, the numerical model for the atomic
  simulation of the chemical and physical processes
  at the interface between MgH2 and Mg during
  hydrogen desorption.

Highly ionic conductive materials
Constant temperature and constant volume
ab-initio molecular dynamics simulations (using
the CPMD code) are performed from room
temperature to 900 K for system. Atoms on the
free surfaces in the system are kept fixed to
mimic bulk behavior and to minimize interactions
among the interface and the free surfaces. The
final configurations reveal an increased mobility
of hydrogen atoms near the interface. This
mobility is much higher than for bulk hydrogen
atoms. This difference increases at higher
temperatures. No hydrogen atoms are observed to
diffuse in the Mg bulk.
The concept of short range order is largely
developed to describe disordered systems, since
it is based on the idea that structural
correlations do not extend beyond the first shell
of neighbours. However, a more extended level of
structural order involving atomic nanostructures
can be found in disordered network forming
systems, as those belonging to the AX2 family (A
Si, Ge  XO, Se, S). Experimentally this long
range order manifests itself through the
appearance of a first sharp diffraction peak in
the total structure factor. Disordered network
forming systems have a high technological impact,
since they can play the role of host glassy
matrices whose structure can be changed by
addition of fast ionic species, the so-called
network modifiers. MD simulations is the method
of choice to link the microscopic origins of the
long correlations to macroscopic properties.
T 500 K
T 300 K
T 700 K
T 900 K