Page 1 of 1

How to get (Effective Core Potential) ECP into CASINO format

PostPosted: Thu Nov 22, 2018 12:32 pm
by Philip_Hoggan
Dear All,

We've prepared some molecular wave-functions with Molpro using an ECP (we prefer not to use pseudo-potentials). The wave-function was converted using molden2qmc into the gwfn.data form and CASINO reads it. However, it does not cater for the ECP and we do not know how to do this. From the manual and elsewhere, it seems to be possible. Would anybody be able to tell me how to do it?

Thanks in advance!

Philip

Re: How to get (Effective Core Potential) ECP into CASINO fo

PostPosted: Mon Nov 26, 2018 5:40 am
by Vladimir_Konjkov
Hello Philip

For calculations with pseudopotential (ECP) is usually required to have two formats for the pseudopotential (ECP) — one for conventional quantum chemical programs, the other for CASINO. It is also necessary to have a basis corresponding to this pseudopotential (ECP).

On the page https://vallico.net/casinoqmc/pplib/ you can find relevant information.

For example C atom Dirac-Fock AREP Trail & Needs pseudopotential's data consists of:
http://www.tcm.phy.cam.ac.uk/~mdt26/pse ... /1/pp.data - ECP in CASINO format
http://www.tcm.phy.cam.ac.uk/~mdt26/pse ... p_gaussian - ECP in Gaussian format
http://www.tcm.phy.cam.ac.uk/~mdt26/pse ... /pp_gamess - ECP in GAMMES format
http://www.tcm.phy.cam.ac.uk/~mdt26/pse ... ussian.txt - corresponding bases (aug-cc-pVDZ-CDF, aug-cc-pVTZ-CDF, aug-cc-pVQZ-CDF, aug-cc-pV5Z-CDF) in Gaussian format.

You should convert Gaussian format to Molpro first.
It is necessary to make sure that the atomic numbers for atoms with a pseudopotential in the gwfn.data file are +200.
molden2qmc has an optional argument --pseudoatoms as it cannot detect if a pseudopotential was used, set it to:
none = pseudopotential was not used for any atoms in this calculation (default).
all = pseudopotential was used for all atoms in this calculation.
white space separated numbers = list of pseudoatom's number (started from 1).
According to the CASINO manual "7.5 Pseudopotential file" you should rename ECP for C atom in CASINO format to c_pp.data file and place it in the directory where Casino is run, adjust up- and down- electron numbers in the input file, then run CASINO as usual, switch on USE_TMOVE in DMC which is on by default since 2018.

Best Vladimir.

Re: How to get (Effective Core Potential) ECP into CASINO fo

PostPosted: Sun Dec 02, 2018 4:31 pm
by Philip_Hoggan
Thanks Vladimir!

My wavefunction is fine and expects the ECP I used in Molpro but in the CASINO format. I have not managed to do that part yet. I have:

1 The ECP (a very concise file for use with a gaussian code)

2 The basis set

It is not yet clear to me how to put this into x_pp.data format.

If I just put in another x_pp.data for the right element but not made like the ECP then the casino code does run but the energies are way off, as you might expect.

I will look at the documents in detail to see if I can sort it out but if you have any tipe, they would be welcome.

Philip

Re: How to get (Effective Core Potential) ECP into CASINO fo

PostPosted: Mon Dec 03, 2018 3:26 pm
by Vladimir_Konjkov
Hello Philip

Formats for ECP used by different programs are similar. On the https://bse.pnl.gov/bse/portal you can get the desired potential in various formats but not in Casino.
And I do not know how to convert to, but maybe some clues can be found in the directory CASINO/utils/pseudo_converters of the Casino tarball

Best Vladimir.

Re: How to get (Effective Core Potential) ECP into CASINO fo

PostPosted: Tue Dec 04, 2018 9:57 am
by Mike Towler
Hi Philip,

Use the ptm utility.

See the various README files in CASINO/utils/ptm

Cheers,
Mike