Messages

1

General Questions and Answers / DFT+U + SOC

« on: July 12, 2019, 15:13 »

Dear QATK developers

I have a question concerning the use of QATK to perform DFT+U calculations including spin-orbit coupling since I am interested in the magnetic anisotropy (using the MAE study object) of a magnetic system that necessitate a +U treatment.
It seems from the vnl interface that it is not allowed. However I still tried by explictely defining the exchange correlation  in my python script (I alrady did that to perform DFT+U with VdW which is also in principle not allowed):

#----------------------------------------
# Exchange-Correlation
#----------------------------------------
exchange_correlation = ExchangeCorrelation(
  exchange=PerdewBurkeErnzerhofExchange,
  correlation=PerdewBurkeErnzerhofCorrelation,
  number_of_spins=4,
  hubbard_term=Dual,
  spin_orbit=True,
  )
exchange_correlation._setShortName("None")

However I get the following error message when the code start the calculation that includes SOC:
** Back Engine Exception : The SpinMatrix is NONCOLLINEAR, hence it can only be accessed with ALL_SPIN
** Location of Exception : spinmatrix.h:520

Is there a way to solve this issue?

best
Cyrille

ps: I am using 2019 version.

2

General Questions and Answers / Re: density mesh cutoff

« on: December 12, 2017, 10:42 »

OK thanks for your reply. In terms of computational cost what is the "best" solution?
increase density mesh cutoff or reciprocal_energy_cutoff
Cyrille

3

General Questions and Answers / density mesh cutoff

« on: December 11, 2017, 16:15 »

Dear atk experts

I have recently faced the following problem when working on MOF (metal-organic framework).
My system contains 3 metallic ions (Cobalt) which are strictly equivalent (when in "ideal" position: see image).  However during the scf loop the 3 atoms sometimes acquire different charges and magnetization which can in some cases have a rather drastic influence on the electronic structure.
I manage to "solve" this problem by increasing the density mesh cutoff (up to 150 Hartree).

Are there other possible options to solve this issue?

Cyrille

4

General Questions and Answers / DFT+U Anisotropic or SphericalSymmetric

« on: October 13, 2017, 15:53 »

Dear all

I have noticed a strange behaviour of the DFT+U with the "Anisotropic" option.
I am performing DFT+U calculation on a MOF (metal-organic framework) and when using the Anisotropic implementation the band structure looks totally weird while everything looks "nice" with the SphericalSymmetric implementation.
I have the feeling that there is something going wrong with Anisotropic implementation of DFT+U.

I have attached my python input.

Thanks in advance

Cyrille

5

General Questions and Answers / Re: "unrealistic" relaxation of H2O molecules

« on: October 4, 2017, 11:30 »

Thanks a lot Daniele

I thought that the system should find its  minimum quite "naturally" .
But it seems that it is more tricky than I thought.

I will do some extra checks/tests...

Cyrille

6

General Questions and Answers / Re: "unrealistic" relaxation of H2O molecules

« on: October 4, 2017, 11:05 »

Dear Petr

I refer to some "chemist know-how". Not being a chemist myself I have to follow the instinct of people with a longer knowledge in the domain:-)

I have added a U because it seems to be necessary for these systems. However the relaxed configuration is the same without U!

Cyrille

7

General Questions and Answers / "unrealistic" relaxation of H2O molecules

« on: October 4, 2017, 09:26 »

Dear all

I am trying to relax a MOF (metal-organic framework) in presence of water molecules.
According to my colleagues chemist H2O should stay ontop (above and below) of the metallic atoms of the MOF.
However during the relaxation process with  ATK H2Omolecules  lays down on the MOF and starts forming H bonds with the Oxygens of the MOF.

Is there a reason for this behavior? Is there a way to cure this "problem"?


Thanks in advance

Cyrille

PS: here is a link towards my py and nc file
https://www.dropbox.com/sh/hjonz6hda1ocw5i/AADjfebjuHydbic8iovJIyBba?dl=0

8

General Questions and Answers / Re: NEB calculation

« on: February 24, 2017, 16:04 »

I am still struggling with the NEB calculation. I have converged the initial and final states with a more stringent criterion on the force (below 0.01eV/A) . However I still get the following message:

 WARNING: Product configuration is not optimized.                             |
|          Max. atom force is     3.528e-02 eV/Ang.                            |
| This means the product energy is most likely not a local minimum. 

Any idea what is going wrong?

I attach my python script.

thanks in advance

Cyrille

9

General Questions and Answers / Re: NEB calculation

« on: February 23, 2017, 10:06 »

OK I will try to run the scf calculation of end points with a more stringent convergence threshold on forces.
But if I understand well the NEB code is also running a preliminary convergence of end points to check if they are OK?

Cyrille

10

General Questions and Answers / NEB calculation

« on: February 23, 2017, 09:16 »

Dear all

I have some difficulties regarding a NEB calculation.
I am investigating the configuration of a rather complex molecule.
I have first performed DFT+U+VdW calculations and found two possible conformations of the molecule that I call C1 and C2.
The energy of C2 is 0.3 eV lower  than C1 and I want to calculate the energy barrier between these two conformations.
I have therefore generated a NEB path and run the calculation.

First of all I get several warning messages:
1)  WARNING: The Grimme DFT-D2 correction is only intended for GGA.PBE and GGA.BLYP exchange-correlation functionals.
2)  WARNING: Reactant configuration is not optimized.
3)  WARNING: Product configuration is not optimized

For the first warning I know it is related to VdW and DFT+U

But what is the meaning of warning 2 and 3.

In addition at the end of the calculation I find that the energy of C2 is now higher  than the one of C1..
I do not understand how it is possible since these two configurations have been fully relaxed in my preliminary DFT calculation.

Thanks a lot in advance

Cyrille

11

General Questions and Answers / Re: advice on basis set/pseudo for complex molecule

« on: February 6, 2017, 14:42 »

Hi Daniele

OK thanks I will try DZP  (with DFT+U and VdW DFT-D2) first for all atoms and see if it cures my Jahn-Teller problem.
These spin crossover molecules are really complicated! Everything is inter-connected...

best
Cyrille

12

General Questions and Answers / Re: advice on basis set/pseudo for complex molecule

« on: February 3, 2017, 16:04 »

Hi Daniele

I knew you would answer this question for chemist:-)
I am now trying SZP for H, C and B. DZP for N (directly connected to Fe) and SG15 (medium) for Fe.

The comparison with QE is not straightforward because this is a very delicate molecule. A so called spin crossover molecule with two spin state: Low spin (S=0) and High Spin (S=2). DFT+U is absolutely mandatory (other wise LS is drastically too low in energy) and VdW is also rather important.
Therefore there are many factors (and parameters) that make the comparison difficult.

However I realized that with QW in SZP I found that the equilibrium configuration was slightly distorted and asymmetric (a kind of Jahn-Teller distorsion) which is not the case with QE . In fact both configurations exist with QE (asymmetric and symmetric) but the symmetric one is the lowest in energy contrary to QW..

By the way, which density_mesh_cutoff do you suggest. I took 150.0*Hartree but maybe it is not necessary.?

Cheers

Cyrille

13

General Questions and Answers / advice on basis set/pseudo for complex molecule

« on: February 3, 2017, 08:43 »

Dear all

I am having difficulties to model some complex molecules and I need your advice.
I have started using a small basis set (FHI SZP) to avoid too lengthy calculation.
However I realized by comparing with Quantum espresso that it was probably not sufficient.
I have switched to a much more complete basis set SG15 (medium) for all the atoms (C,H,B,N,Fe).
The calculation are becoming much heavier and I wonder if I could use SG15 only for the transition metal and keep smaller basis set for C,H,B and N.

I also need to perform DFT+U with VdW (as discussed yesterday).

Let me also add that the python file attached to my message correspond to an isolated molecule but I want to perform calculation for organized layers with a super-cell containing 3 molecules and test many configurations! Therefore I should find the optimal solution in terms of precision/speed.

Thanks a lot for your help.

Cyrille

14

General Questions and Answers / Re: DFT+U with VdW

« on: February 2, 2017, 14:55 »

OK thanks a lot. I will try it.
If I understand well: VNL does not allow DFT+U with VdW but it is possible to perform such calculation within atk.

Cyrille

15

General Questions and Answers / DFT+U with VdW

« on: February 2, 2017, 11:35 »

Dear all

I have noticed that VdW is disabled in DFT+U calculations. Is there a particular reason for this "limitation"?
I am presently doing calculations on molecular layers where DFT+U must be taken into account to describe properly the magnetic properties but the inter- and intra-molecular interaction probably needs to also include VdW.
What do you propose?

thanks in advance

Cyrille