Diapositive 1 - Industrial Technologies 2012

Report
Modelling of Materials and Nanotechnologies:
From Ab Initio to Macro-Simulations
Dr. Jérôme Cornil
[email protected]
Laboratory for Chemistry of Novel Materials
University of Mons, Belgium
Industrial Technologies, June 2012, Aarhus
Multiscale Modelling
ms to ks
104-106
100-104 / fs to ns
10-100
http://www.cse.sc.edu/~heyden/Multi-ScaleModelling.html
Quantum-Chemistry
Key quantity : Wavefunction
Time-independent Schrödinger equation
H=E
Approximations
F i (r) = i i (r)
Averaged electronic interactions 
Correlation error : E0 – EHF
Hartree-Fock equation
Quantum-Chemistry
Post-Hartree Fock methods
Perturbative
Variational
Configuration Interaction (CI)
Möller Plesset (MP2, MP3,..)
Very time consuming 
CI = C0 SCF
+
C 
J
J
J
ab
+
C 
K
K
K
ab
cd
Truncations needed 
Access to excited states 
Coupled Cluster (CC)
CASSCF
+ ...
Density Functional Theory (DFT)
Key quantity : Electronic density
Hohenberg-Kohn functional
E0 [0] = FHK [0] +  0(r ) Vext dr
F [] = TS [] + J [] + EXC []
Exchange-correlation energy
fKS i = i i
Need to find expressions for EXC 
Self-interaction error 
Access to excited states (TD-DFT) 
Kohn-Sham equation
DFT and HF
- Explicit account of electrons 
- Limitation in the size of the system (a few hundred atoms) 
Semi-empirical Hartree-Fock techniques
- Poor description of van der Waals interactions 
- Use of Periodic Boundary Conditions (PBC) 
CH3S on gold
Slab approach
Gaussian, Turbomole, VASP, SIESTA, DMOL, CRYSTAL, MOPAC, ABINIT…
Phenomenological Models
- Charge transport
Sites Vibrations Coupling
Tight-binding models (Tight-binding DFT)
Force Field Calculations
E=
Stretching
Torsions
Bending
Out-of-plane
H-bonds
Non bonded interactions
(van der Waals, Coulombic)
Force Fields
Parameterized from quantum-chemical calculations and/or experimental data
Stretching
Bending
Vs = ks (r – r0)2
Vb = kb ( – 0)2
Harmonic approximation
van der Waals
Coulomb
Vc = 
qi qj
4  0 r r
Vw = (Aij / r12 – Bij / r6)
Lennard-Jones potential
Force Fields
- No explicit account of electrons 
- Parameterization from QC calculations or experimental data 
- Large size of the systems (many thousands) 
- Explicit term for vdW interactions 
Specific force fields 
Organic
MM2, MM3
UFF (Universal Force Field)
COMPASS (Condensed-phase Optimized Molecular Potentials for
Atomistic Simulation Studies)
Biologic (proteins, nucleic acids)
AMBER (Assisted Model Building with Energy Refinement)
CHARMM (Chemistry at HARvard using Molecular Mechanics)
Reliable description of H-bonds
Materials Studio, GROMACS, NAMD, TINKER, LAMMPS
Molecular Simulations
Molecular Mechanics
Molecular Dynamics
E(x)
E(x)
x
Search for
equilibrium structures
x
Exploration of
the conformational space
+ Simulated Annealing + Monte Carlo
PBC : Available
Quantum dynamics : Car Parrinello (CPMD)
QM/MM
- Embedding of a quantum part in a classical medium
- Interest in using polarizable force field
T. Van Voorhis and co, Acc. Chem. Res. 7, 43, 995
Coarse-Graining
Atomistic description  Effective interaction potentials
Key role of the parameterization 
Claudio Zannoni, University of Bologna
Coarse-Graining
C. Zannoni, J. Mater. Chem., 11, 2637, 2001
Kinetic Monte Carlo (KMC)
- Random choice of the destination : ij
i
j
- Probability of transfer
 ij
Pij 
 ik
k i
- Random number between 0 and 1 : 
If   Pij, the transfer is accepted with a time ij-1
If   Pij, the transfer is not accepted
vx
dx
x  
F t F
Continuum Models – Device Modelling
- Tools to solve classical differential equations fed with input parameters coming
from experiment or molecular simulations (Finite Element Methods – FEM)
- Specific to the applications (fluid, charges…)
Comsol, Abaqus
- Charges  Drift-diffusion model
Charge density profile !!
- Fluid  Navier-Stokes equation
Fluid density profile !!
- Structural mechanics
- Implicit account of chemistry concepts
AMCOS 233502 – Materials as CO2 removers
Zeolite framework for selective sorption of CO2
Force Field (Dreiding)
Molecular Dynamics, Monte Carlo
DFT (partial charges)
Amount of gas sorbed (isotherms)
CO2 diffusion
J. Mater. Chem., 2010, 20, 7676-7681
HYMEC 263073
Non volatile hybrid memories
- Gold nanoparticles in organic semiconductors
S
D
- Switch via charging of the nanoparticles
G
Quantum-chemistry (Electronic structure of the NPs and molecules)
Kinetic Monte Carlo or Master Equation (mobility)
Drift – Diffusion Model (I/V characteristics)
MAHEATT 227541
Search for new cathode materials in
Li-ion batteries for increased storage
DFT – Structure and stability of the materials
Intercalation and deintercalation of Li
DFT - MD
Li-ion mobility
MODIFY 228320
Stress induced mechanical properties of soft-based adhesives (acrylic polymers)
Force field (interaction polymer/substrate – formation of H bonds)
Coarse graining (interaction between particles)
Finite element (bulk rheological properties – debonding mechanisms)
MORDRED 261868
Nanoelectronics
Random defects + trapped charges
DFT
MULTIHY 263335
DFT (energy barriers / binding energies)
Tight binding  Grain boundaries
Kinetic Monte Carlo (hydrogen diffusion as a function of defect, T, stress)
Finite Element Method (macroscopic model for hydrogen diffusion)
HYPOMAP 233482
New materials for hydrogen storage / proton exchange membranes
DFT Chemisorption / physisorption of H2 in metal hydrides and MOFs
Proton diffusion rates
DFTB up to 1000 – 10000 atoms
POCO 213939
Interactions of functionalized carbon nanotubes with polymer matrices
Force field calculations
Interaction energies
Pull-out energy
FEM
Mechanical properties
ADGLASS 229205
Adhesion and cohesion at interfaces in high performance glassy systems
Anti-adherent properties for proteins
Adhesion between glassy SiO2 and crystalline
TiO2 for solar collectors
Hybrid quantum-mechanical atomistic modelling
Quantum chemistry involved for the formation of chemical bonds, proton transfer
Model development !
Dynamag 233552 and Magnonics 228673
DFT (Exchange energy)
Finite Element Micromagnetic package
Spin wave spectra and dispersion
Magnonic devices
New models for new physics !!
MINOTOR 228424
Self-Assembled Monolayers
HS-C16H33
120 nm metal
220 nm metal + HS-C16H33
320 nm metal + HS-C2H4C8F17
Work function [eV]
6
5.5
5
HS-C2H4C8F17
4.5
4
3.5
1
2
Au
3
4
5
6
Ag
7
metal (+SAM)
Tuning of injection barriers
Bert de Boer et al., Adv. Mater., 2005, 17, 621
Theoretical Methodology
CH3S
 (bridge) = -1.19 eV
 (fcc) = -1.52 eV
SIESTA – DFT – GGA (PBE)
Perspectives
- Linear scaling
- Dispersion forces at the quantum level
- Hybrid approaches QM/MM
- Polarizable force fields, reactive force fields
- More links between the atomistic and macroscopic world
- Automatized procedure for multiscale modelling
[email protected]: e-infrastructure
Workflows
and Services
• diverse & modular
• open & extendable
• maintainable & accessible
Application
Integration
• efficient (HPC)
• adaptable (UNICORE/
DEISA)
• secure
• OLED (MINOTOR/BASF)
• Li-Ion Batteries (CEA)
• Carbon Devices (NOKIA)
• Molecular Electronics
(SONY)
• Developers (Academics)
• Resources (HPC Providers)
• Users
(Industry/SME/Academics)
Key
Applications
Community
Building
Proposal , pp. 5−8
Partners
Participant
Karlsruhe Institute of
Technology
Commissariat à l'énergie
atomique
Acronym Country
Nokia
KIT
Germany
CSC
CEA
France
CINECA Bologna
CIN
Italy
CSC - IT Center for Science
CSC
Finland
Korea Institute of Science
and Technology
KIST
Korea
UMons
Sony
SONY
Germany
CEA
Science and Technology
Facilities Council
STFC
UK
University of Mons
Umons
Belgium
University of Patras
UPA
Greece
STFC
KI
T Sony
CIN
UPA
www.multiscale-modelling.eu
Application Example
• Film deposition (or MD)
– Generate disordered film
morphologies
• QM calculations of
hopping sites
– Calculate HOMO, LUMO,
LUMO+1 etc energies.
– Electronic couplings
reorganization energies
– Calculate charge hopping
rates
• Kinetic Monte Carlo
(KMC)
– Calculate charge
(electron-hole) mobility
– Calculate current density
J. J. Kwiatkowski, J. Nelson, H. Li,
J. L. Bredas, W. Wenzel, and C.
Lennartz, Phys. Chem. Chem.
Phys., 2008, 10, 1852–1858.
07/07/2015
For the science: see talk by J.
Cornil
32
Graphical User Interface
for individual codes and
entire workflows
Gridbean
s
Parameters can be adjusted
UNICORE Rich Client
Workflow
07/07/2015
Embedded visualisation with Jm
33
Control flow:
Example
07/07/2015
Stefan Bozic – ISGC, Academia
Sinica Taipei, February 29, 2012
34
Coverage of different
scales
macroscopic scale
~ 10-6
m
molecular scale
~ 10-8
m
electronic scale
~ 10-10
m
continuum
model (FEA)
coarse-grained
model
(CG)
Atomistic
model
(MM)
QM model
(QM)
Elmer
ToFeT (KMC)
DEPOSIT
MOPAC
FEAP
End-bridging MC
LAMMPS
TURBOMOLE
Transporter
DL_POLY
BigDFT
07/07/2015
Many gridbeans for high performance
35

similar documents