Dear Imad,
would you please send me a .pdf of your article ([EMAIL
PROTECTED])? Regarding Van del Waals. It has been demonstrated that LDA can
reasonable reproduce pi stacking, which is dominated by dispersion
interactions. You may check Zhao and Truhlar´s articles. LDA is among the best
functionals for stacking and some Van der waals interactions.
Y. Zhao, D.G. Truhlar J. Chem. and Theory. Comput. 3 (2006) 289.
I have recently submited an article about stacking. If you compare the benzene
dimer results at the levels: MP2/CBS resutls and LDA against CCSD(T)/CBS, you
will find that LDA is far better than MP2. MP2 seriously overestimates
stacking. Just an example (in kcal/mol):
BBPD
LDA 2.6
CCSD(T)/CBS 2.73
MP2/CBS 5.03
So I would not try to reproduce MP2 results, but CCSD(T)/CBS.
Regards,
pablo
----- Original Message -----
From: Belabbas Imad
To: [email protected]
Sent: Friday, July 27, 2007 6:20 PM
Subject: [SIESTA-L] RE : Re: [SIESTA-L] stacking
Dear Denis,
Let me coment a bit your last message. I am not really aware about the
specific system that you are treating. However, as far as I have understood
from the massage, you are treating a system involving aromatic molecules. Note
that the source of you problem is related to the shortcoming of XC-functionals
used in DFT to describe Van Der Waals interactions !!. And I guess, even if you
move to a PW code you will run with similar problems !!?????. What you need is
a correction of DFT to improve its description of these Van Der Waals
interactions.
A very interesting approach proposed by a group at EPFL (PRB. 75, 205131
(2007)), consists of making corrections at the pseudopotential level to
reproduce MP2 or CC calculations !!! these pseudopots are designed for CPMD
!!!, One can try to convert their format to that supported by SIESTA and try
them !!!!!!!
Best wishes,
Imad
"Pablo A. Denis" <[EMAIL PROTECTED]> a écrit :
Dear Oleksandr,
Thank you very much for your answer. I have been playing a little with the
basis sets. The problem was basis set. BSSE optimization is important, but
only for the default DZP. The interaction energies without BSSE are
atractive by 10-18 kcal/mol, very huge as usual for siesta.
BBS BBPD
BBT BB2Hover
deault DZP -3.2 (0.3 bsseopt) -1.34
2.7 3.25
DZPshort -1.4 0.54
3.35
DZPlong 0.5 2.35
3.2 3.6
target values 1 2.6
3.1 3.3
negative = repulsion
BBS and BBPD are parallel aromatics wheras BB2Hover and BBT not. So the
problem is to find a basis set that describe the stacking and not overbind
the other complexes that have hydrogens over the pi clouds. Even with the
DZP Long the interaction between the parallel aromatics is not well
described. Tournus employed DZP+3s on carbon and DZ on H. The reported good
values for BBS and BBPD but no result is available for BBT. I will keep
working or maybe I will change to plane wave, the problem is that pw takes
more resources...
Thank you very much for you help.
Regards,
pablo
P.S. below there is a full .fdf for BSSE maybe the cause of everything is a
mistake...
# FDF file for bb
# General System descriptors
SystemName bb # Descriptive name of the system
SystemLabel bb # Short name for naming files
NumberOfAtoms 24 # Number of atoms
NumberOfSpecies 4 # Number of species
%block Chemical_Species_Label
1 6 C
2 1 H
3 -6 C_G
4 -1 H_G
%endblock Chemical_Species_Label
%block PAO.Basis
C 3 0.01
n=2 0 2 E 69.86 4.68
6.004 4.195
1.000 1.000
n=2 1 2 E 18.95 3.78
4.995 3.100
1.000 1.000
n=3 2 1 E 16.84 0.05
4.187
1.000
H 2 0.22
n=1 0 2 E 2.07 0.00
4.971 1.771
1.000 1.000
n=2 1 1 E 0.89 0.01
4.988
1.000
C_G 3 0.01
n=2 0 2 E 69.86 4.68
6.004 4.195
1.000 1.000
n=2 1 2 E 18.95 3.78
4.995 3.100
1.000 1.000
n=3 2 1 E 16.84 0.05
4.187
1.000
H_G 2 0.22
n=1 0 2 E 2.07 0.00
4.971 1.771
1.000 1.000
n=2 1 1 E 0.89 0.01
4.988
1.000
%endblock PAO.Basis
# Lattice, coordinates, k-sampling
LatticeConstant 15.0000 Ang
%block LatticeVectors
1.000000 0.000000 0.000000
0.000000 1.000000 0.000000
0.000000 0.000000 1.000000
%endblock LatticeVectors
%block kgrid_Monkhorst_Pack
1 0 0 0.0
0 1 0 0.0
0 0 1 0.0
%endblock kgrid_Monkhorst_Pack
AtomicCoordinatesFormat NotScaledCartesianAng # Format for coordinates
AtomicCoorFormatOut Ang
%block AtomicCoordinatesAndAtomicSpecies
-0.00001305 1.39137696 -0.58199054 1
0.00001559 -1.39131893 -0.58209854 1
1.20541798 0.69582469 -0.58201361 1
1.20540902 -0.69575812 -0.58202552 1
-1.20535060 0.69582449 -0.58216644 1
-1.20539960 -0.69574182 -0.58217128 1
0.00001963 2.48284460 -0.57741741 2
0.00000245 -2.48281112 -0.57757235 2
2.15127460 1.24139401 -0.57738336 2
2.15129256 -1.24137904 -0.57747607 2
-2.15125940 1.24140624 -0.57767286 2
-2.15127288 -1.24137847 -0.57773904 2
0.00003196 1.39125978 2.77519589 3
0.00000545 -1.39125731 2.77530704 3
1.20542647 0.69575598 2.77474408 3
1.20541822 -0.69577865 2.77485974 3
-1.20544732 0.69581618 2.77474221 3
-1.20545658 -0.69579639 2.77485844 3
0.00001154 2.48268398 2.77220809 4
-0.00002279 -2.48268283 2.77248520 4
2.15118306 1.24170370 2.77070140 4
2.15115349 -1.24173433 2.77084865 4
-2.15116586 1.24172703 2.77071756 4
-2.15117033 -1.24170091 2.77086670 4
%endblock AtomicCoordinatesAndAtomicSpecies
# DFT, Grid, SCF
XC.functional LDA # Exchange-correlation functional type
XC.authors CA # Particular parametrization of xc func
SpinPolarized .false. # Spin unpolarized calculation
MeshCutoff 200. Ry # Equivalent planewave cutoff for the
grid
MaxSCFIterations 450 # Maximum number of SCF iterations per
step
DM.MixingWeight 0.3 # New DM amount for next SCF cycle
DM.Tolerance 1.d-6 # Tolerance in maximum difference
# between input and output DM
DM.NumberPulay 3 # Number of SCF steps between pulay
mixing
# Eigenvalue problem: order-N or diagonalization
SolutionMethod diagon # OrderN or Diagon
ElectronicTemperature 0 K # Temp. for Fermi smearing
# Molecular dynamics and relaxations
#MD.TypeOfRun CG # Type of dynamics:
#MD.NumCGsteps 100 # Number of CG steps for
# coordinate optimization
#MD.MaxCGDispl 0.05 Ang # Maximum atomic displacement
# in one CG step (Bohr)
#MD.MaxForceTol 0.01 eV/Ang # Tolerance in the maximum
# atomic force (Ry/Bohr)
# Output options
WriteCoorInitial .true.
WriteCoorStep .true.
WriteForces .true.
WriteKpoints .true.
WriteEigenvalues .true.
WriteKbands .true.
WriteBands .true.
WriteMullikenPop 1 # Write Mulliken Population Analysis
WriteCoorXmol .true.
WriteMDCoorXmol .true.
WriteMDhistory .true.
WriteCoorXmol .true.
# Options for saving/reading information
DM.UseSaveDM # Use DM Continuation files
MD.UseSaveXV .true. # Use stored positions and velocities
MD.UseSaveCG .true. # Use stored positions and velocities
SaveRho # Write valence pseudocharge at the
mesh
SaveDeltaRho # Write RHOscf-RHOatm at the mesh
SaveElectrostaticPotential .false. # Write the total elect. pot. at the
mesh
SaveTotalPotential .false. # Write the total pot. at the mesh
WriteSiestaDim .false. # Write minimum dim to siesta.h and
stop
WriteDenchar # Write information for DENCHAR
DirectPhi .true.
SaveMemory .true.
------------------------------------------------------------------------------
Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! Mail
