setting orbital occupancy matrix
Posted: Thu May 03, 2012 4:21 pm
I'm investigating the UO2 system with VASP L(S)DA+U, which has been known to significantly improve the results compared to standard L(S)DA whilst creating many meta-stable states, and making it difficult to find the ground state. This is well documented in these papers :
http://link.aps.org/doi/10.1103/PhysRevB.79.235125
http://dx.doi.org.libproxy.ucl.ac.uk/10 ... 011.03.012
The approach that has been developed by these authors to avoid the meta-stable states is to explicitly set the occupation matrix of the f-orbitals (that can be viewed with LDAUPRINT>0) to a near GS configuration prior to the simulation. Does anyone have suggestions as to how this can be achieved in VASP?
It must be possible as the authors of these papers claim to have used VASP to perform their calculations.
I am aware that FERWE may allow for the diagonal of the occupation matrix to be set, but a further requirement to break the symmetry requires the off-diagonal elements be set as well.
I hope one of the VASP wizards is able to help....
Regards.
http://link.aps.org/doi/10.1103/PhysRevB.79.235125
http://dx.doi.org.libproxy.ucl.ac.uk/10 ... 011.03.012
The approach that has been developed by these authors to avoid the meta-stable states is to explicitly set the occupation matrix of the f-orbitals (that can be viewed with LDAUPRINT>0) to a near GS configuration prior to the simulation. Does anyone have suggestions as to how this can be achieved in VASP?
It must be possible as the authors of these papers claim to have used VASP to perform their calculations.
I am aware that FERWE may allow for the diagonal of the occupation matrix to be set, but a further requirement to break the symmetry requires the off-diagonal elements be set as well.
I hope one of the VASP wizards is able to help....
Regards.