[Octopus-users] Multipole Kicks

[Rudolf Repges]

Hello,

I use single multipole kicks of the order l=1 and l=2 to calculate
dynamic polarizabilities.

After the ground state calculation the first steps of the time
propagation following l=1 and m=1,0,-1 kicks result in strong (relative
to the others) induced multipole moments in a direction opposite to the
kicks but following l=2 and m=-2,-1,0,1,2 kicks in strong induced
multipole moments in the direction of the kicks.

But I think that for the first steps of the time propagation all single
kicks must be followed by relatively strong induced multipole moments in
the direction of the kicks or all by multipole moments in the opposite
direction. If not I get some inconsistency regarding the symmetry of
some polarizability matrix.

It would be very nice if somebody could explain me the different sign of
the directions of the induced multipole moments related to the kicks
when going from l=1 to l=2.

Rudolf


PS: I attach the inp "file" I used with one of the multipole kicks.
After the "td" run I multiply the induced dipole moment given in the
"multipoles" file stored in the "td.general" folder in x- and
y-direction by -0.488602512 and the induced dipole moment in z-direction
by 0.488602512 to get the corresponding l=1,m=1 and l=1,m=-1 and l=1,m=0
spherical harmonic moment.
 
-------------- next part --------------
#Generalities.
###############################################################################
verbose = 30
SystemName = "MeCONH2"
CalculationMode = td 
fromScratch = yes
#Units = eV_angstrom
b = 0.529177
eV = 27.21161
#b = 1
###############################################################################
# Species.
###############################################################################
%Coordinates
"O"  | -0.293377/b  |   1.354350/b  |   0.000000 |  no
"C"  |  0.000000/b  |   0.163045/b  |   0.000000 |  no
"N"  |  1.304042/b  |  -0.265454/b  |   0.000000 |  no
"C"  | -1.057712/b  |  -0.931753/b  |   0.000000 |  no
"H"  |  1.558460/b  |  -1.242280/b  |   0.000000 |  no
"H"  |  2.038175/b  |   0.431461/b  |   0.000000 |  no
"H"  | -1.692477/b  |  -0.804177/b  |   0.881055/b |  no
"H"  | -1.692477/b  |  -0.804177/b  |  -0.881055/b |  no
"H"  |  0.646689/b  |  -1.945200/b  |   0.000000 |  no
%

###############################################################################
# Mesh description.
###############################################################################
BoxShape = minimum
Radius  = 3.5/b
Spacing = 0.25/b

###############################################################################
# Exchange and Correlation.
###############################################################################
#XFunctional = lda_x
#CFunctional = lda_c_pw
###############################################################################
#States
###############################################################################
#SpinComponents = spin_polarized
#ExtraStates = 5
##################SCF########################################################
MaximumIter = 200
EigenSolverMaxIter = 100
EigenSolverFinalToleranceIteration = 15
EigenSolverInitTolerance = 1e-2
EigenSolverFinalTolerance = 1e-5
LCAOStart = no
###############################################################################
# 11. Time Dependent.
###############################################################################
TDMaximumIter = 3
TDTimeStep = 0.012*eV

TDExponentialMethod = lanczos
TDExpOrder = 15

#TDDeltaStrength = 0.0004886
TDDeltaStrength = 0.001
TDDeltaStrengthMode = 0

%TDKickFunction
1|-1|1
%

OutputEvery = 200
TDOutput = multipoles + angular
TDDipoleLMax = 2