[QE-users] Unoccupied molecular orbitals near the vacuum level.

[Lorenzo Paulatto]

Dear Steven,
Any pseudopotential (and PAW is no different) will eventually fail for
sufficiently high energy. In the specific case of H2, it is worst because
the two protons are closer than the typical pseudization radius. I would
bet that generating a hard pseudopotential with a couple of high energy
projectors could solve the problem, but I have no proof.

Also the "all-electron" pseudopotential available here should work (beware,
it requires a cutoff of 500+ Ry)
http://quantum-espresso.org/upf_files/H.coulomb-ae.UPF

-- 
Lorenzo Paulatto

On Tue, Dec 21, 2021, 09:06 Steven Best <steven.best at hdr.qut.edu.au> wrote:

> I have used PWSCF v6.3 and optimized the geometry of an isolated H2
> molecule in a 24 Angstrom box.
> I have then used pp.x to plot the occupied and un-occupied molecular
> orbitals.
>
> With nbnd=2 chosen for the scf calculation I obtain energy
> bands of:
>    -10.3337   0.1108  eV
>      occupation numbers
>      1.0000   0.0000
>
> With nbnd=4 chosen I obtain energy bands of:
> -10.3343  -0.0797   0.1110  0.1375  eV
>      occupation numbers
>      1.0000   0.0000  0.0000   0.0000
>
> With nbnd=2 the plots of the occupied and unoccupied molecular orbitals
> look like the expected bonding and anti-bonding
> H2 molecular orbitals. However, if nbnd=4 is chosen, an unoccupied orbital
> (band 2) with a shape consisting of a number of shells is
> located below the anti-bonding unoccupied orbital (band 3). And another
> orbital with a similar shell structure (band 4) is located
> above the anti-bonding orbital which appears to have the opposite phase to
> band 2.
>
> I have calculated the vacuum level energy far away from the molecule to be
> approximately 0.003 eV by plotting the average potential.
> I think that the shape of the orbitals for nbnd=4 is possibly a result of
> a limitation in the plane wave method that occurs for unoccupied orbitals
> with an energy greater than or close to the vacuum level? Can anybody
> provide some insight into this behaviour?
>
> Kind regards,
>
> Steven Best
>
> PhD Student,
>
> School of Chemistry and Physics,
>
> Queensland University of Technology, Brisbane, Australia
>
> Below are my script files.
>
>
> #######################################################################################
>
> # H2 molecule with nbnd = 2.
>
> &CONTROL
>     calculation = 'scf',
>     verbosity = 'high',
>     restart_mode = 'from_scratch',
>     wf_collect = .T.,
>     nstep = 200,
>     tstress = .true.,
>     tprnfor = .true.,
>     outdir = './',
>     prefix = 'SCF_H2_molecule_nbnd_2_t007',
>     etot_conv_thr = 1.0d-6,
>     forc_conv_thr = 1.0d-5,
>     pseudo_dir = './pseudo'
> /
> &SYSTEM
>     ibrav = 1, A = 24.0,
>     nat = 2, ntyp = 1,
>     nbnd = 2,
>     ecutwfc = 60.0, ecutrho = 480.0
> /
> &ELECTRONS
>     electron_maxstep = 100,
>     conv_thr =  1.0d-8,
>     mixing_mode = 'plain',
>     mixing_beta = 0.4, mixing_ndim = 8,
>     diagonalization = 'david'
> /
> ATOMIC_SPECIES
>    H   1.008   H.pbe-kjpaw_psl.1.0.0.UPF
> ATOMIC_POSITIONS (angstrom)
>    H   12.00         12.00   12.00   0   0   0
>    H   12.750482666  12.00   12.00
> K_POINTS (gamma)
>
>
> #######################################################################################
>
> # Post processing for H2 molecular orbitals.
> # Number of bands = 2.
>
>  &INPUTPP
>     prefix = 'SCF_H2_molecule_nbnd_2_t007',
>     outdir = './',
>     filplot = 'H2_molecule_nbnd_2_wfc_t007',
>     plot_num = 7,
>     kpoint(1) = 1,
>     kband(1) = 1,
>     kband(2) = 2,
>     lsign=.true.
>  /
>  &PLOT
>     nfile = 1,
>     filepp(1) = 'H2_molecule_nbnd_2_wfc_t007', weight(1)=1.0,
>     iflag = 3,
>     output_format = 6,
>     fileout = '.cube'
>  /
>
>
> #######################################################################################
>
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20211221/46792469/attachment.html>