Hello everyone,

I intend to calculate the electrostatic potential of periodic structures (more precisely, zeolites) using DMol3. I successfully performed such calculations some time ago, running Materials Studio 4.X under Windows XP. Now, I am using Materials Studio 5.5 under Windows Vista. Unfortunately, the calculation delivers very strange results, with a very "edgy" ESP with many kinks and local oscillations (see attached picture). The older calculations gave much smoother results which looked more reasonable.

I have attached the input below. The structure is the SOD zeolite from the provided structures, replacing all Al atoms with Si for charge neutrality.

Does anybody have any hints where this problem comes from? Is it related to the operating system?

Best regards,

Michael

DMol3 input:

# Task parameters

Calculate                     energy

Symmetry                      on

Max_memory                    2048

# Electronic parameters

Spin_polarization             restricted

Charge                        0

Basis                         dnp

Pseudopotential               none

Functional                    pwc

Harris                        off

Aux_density                   octupole

Integration_grid              fine

Occupation                    fermi

Cutoff_Global                 4.6000 angstrom

Scf_density_convergence       1.0000e-006

Scf_charge_mixing             0.2000

Scf_iterations                50

Scf_diis                      6 pulay

# Kpoint definition file (intervals/offset):

Kpoints                       file     2 2 2 0.0000 0.0000 0.0000

SOD.kpoints

# Calculated properties

Plot                          potential

Grid                          box  3 50 50 50 0