[Wien] LDA+U Manual Initialisation of Density Matrix

[David Tompsett]

Dear Peter and All on the list,
                                            thank you for the help on 
the manual initialisation of the density matrix. I have a related 
question to follow up which is related to Peter's second comment.

The question relates to the local rotation matrices for my crystal and 
how to specify the correct orbital state in the density matrix. The 
rotation matrix of the f-ion that I wish to consider interchanges the z 
and x axes. However, I want to be able to set the orbital state | Lz=3, 
Sz = -1/2 >  with respect to the initial global cartesian z-axis.

Is the best approach to change the rotation matrix to the indentity and 
then add in all of the relevant LM for the resulting symmetry to case.in2?
If so, how should I determine which point group of Table 7.42 of the UG 
to get the LM from? Initially  (with the rotation matrix found by 
sgroup) the point group is mm2, but if I set the rotation matrix back to 
the identity the required LM basis will change.

Thank you for any help.

Best,
David.

-- 
David A. Tompsett
Quantum Matter Group
Cavendish Laboratory
J. J. Thomson Avenue
Cambridge CB3 0HE
U.K.
Tel: +44 7907 566351 (mobile)
Fax: +44 1223 768140
http://www-qm.phy.cam.ac.uk/



Peter Blaha wrote:
> Most likely, you don't need to run LDA+SO before, but you should probably
> use x orb after you created the dmat files and then use runsp -orbc -so for some time.
>
> Also note, the description:    Lz=3, Sz = -1/2 >   is not "unique" in a
> solid, but your solution will depend a lot on how you place this ion into the solid,
> i.e. where your local z axis points.
>
> David Tompsett schrieb:
>   
>> Dear Users,
>> I have a question about the correct procedure for the initialisation of
>> an LDA+U+SO calculation when I want to manually start f-orbitals in a
>> particular spin state. eg. a hole in |Lz=3, Sz = -1/2 >.
>>
>> To initialise this case I need to create case.dmatup/dn by hand. My
>> question is that in general the UG advises us to converge an LDA+SO
>> calculation first, then add +U. But this would cause the spin
>> configuration of the f-orbitals to be quite different to the one that I
>> want as shown by running lapwdm. Therefore, should I run an initial
>> LDA+SO calculation when I want to start my +U calculation with a
>> particular density matrix?
>>
>> Thank you,
>> David.
>>
>>     
>
>