[CPMD-list] Surface Setup -- inconsistent behaviour?

Urs Zimmerli zimmerli at inf.ethz.ch
Thu Aug 15 11:50:43 CEST 2002


Dear all,

I'm currently experiencing peculiar behaviour of
simulations using the SURFACE setup. I do not know
whether this is a problem of my input statements
or of CPMD. To illustrate how I use the SURFACE
setup, please refer to the inserted input fragment.


----- BEGIN INPUT FRAGMENT ----------------

&CPMD
   OPTIMIZE GEOMETRY
&END
&DFT
   FUNCTIONAL HCTH
&END
&SYSTEM
  SYMMETRY
  0
  CELL
  27.9025   0.865908160639909   1.0    0.0  0.0  0.0
  CUTOFF
  70.
  SURFACE
&END

&ATOMS
*O_MT_HCTH.psp KLEINMANN-BYLANDER
   LMAX=P
    1
    11.811 10.000 10.000   O

*H_MT_HCTH.psp  KLEINMAN-BYLANDER
   LMAX=S
    2
    10.000 10.000 10.000  H
    12.265 10.111  8.250  H
&END

----- END INPUT FRAGMENT ----------------


I'm confronting this simulation to different simulations
in an orthorhombic setup of the cell, thus the only
difference in the input file being in the SYSTEM section
(Note that the lateral dimensions of the cell remain the
same):


----- BEGIN DIFFERENCE ------------------

&SYSTEM
  SYMMETRY
  8
  CELL
  27.9025   0.865908160639909   1.0    0.0  0.0  0.0
  CUTOFF
  70.
&END

----- BEGIN DIFFERENCE ------------------


Not only does this lead to completely different total
Energies:

SURFACE        -17.3248
ORTHORHOMBIC   -17.2737

but also the resulting geometry of the water molecule
is different:

SETUP           Bond length [a.u.]        Bond Angle
SURFACE            1.851                    111.5
ORTHORHOMBIC       1.825                    104.7

Furthermore does the total energy in the SURFACE setup
strongly depend on the orientation of the water molecule.
The final geometry is orientated along the longest
diagonal of the box.
In the case of the 'normal' orthorhombic setup the total
Energy is almost independent on the rotation of the water
molecule (variation approximately 1e-4 a.u.)

The differences do not get smaller upon changing the
dimensions of the simulation cells.

Did anybody experience similar behaviour at some point?
What might be the reason for this behaviour?
Am I using the wrong input options?

With best regards and thank you for any help
urs






More information about the CPMD-list mailing list