[Pw_forum] how to build cell for diamagnetic material?

[Gabriele Sclauzero]

duchl06 wrote:
> hi,Gabriele Sclauzero
> thanks for your advice and hoping your further guide.
> i am reading a paper "Charge-Transfer-Based Mechanism for 
> Half-Metallicity and Ferromagnetism in one-Dimensional 
> OrganometallicSandwich Molecular Wires"
> http://pubs.acs.org/doi/abs/10.1021/ja804053a
> in this paper they consider a sandwich molecular wires (such as 1D 
> (VBz)∞ wire, (MnCp)∞ wire, Bz is benzene, Cp** is cyclopentadienyl), and 
> they said "Furthermore, we investigate the effect of doubling the unit 
> cell size along the SMW direction to evaluate the ground state of 
> ferromagnetic (FM), antiferromagnetic (AFM), and diamagnetic (DM) 
> configurations in our spin-polarized calculations with structural 
> optimization,

In this paper, I think they refer to a non-magnetic ground-state as "DM configuration".
A priori you do not know if the DFT ground state (which does not necessarily always 
correspond to the true physical GS) will be FM, AFM or non-magnetic. You have to allow the 
system to "explore" an enlarged phase space with non-zero magnetization and see if the 
energy of a magnetic configuration (either FM or AFM) is lower than the non-magnetic 
solution.
Actually, if the GS is non-magnetic, you should end to the non-magnetic GS even if you 
start from a magnetic configuration (unless you get trapped in some local minimum).
I think that's why \Delta E in table 1 of the paper you pointed out is always negative (or 
not given): if the GS is magnetic, it will have a lower energy than the lowest energy 
non-magnetic solution, otherwise the GS is not magnetic and a spin-polarized calculation 
will give the same GS of a non-spin polarized one.

HTH

GS



> and assess possibility of the Peierls transition."
>  
> 
>   (and i also found in some other papers when they do not know the groud 
> state is FM or DM, they would always say "To examine the stability of 
> the local magnetic state, we calculate the energy difference between the 
> diamagnetic and ferromagnetic states".
> 
> 在2009-10-02，"Gabriele Sclauzero" <sclauzer at sissa.it> 写道：
>>
>>
>>duchl06 wrote:
>>> thanks every one's reply!
>>> hi, Hui Wang
>>> the example you supplied is a calculation for a antiferromagnetic 
>>> material, i mean how to set the start magnetic moment for the 
>>> diamagnetic material,
>>
>>Diamagnetic material in absence of an external magnetic field should have no net magnetic 
>>moment per atom, if I'm not wrong. So it is not clear to me what do you want to study and 
>>if all these suggestions are pointing you to the correct solution. If you want to study a 
>>diamagnetic material in absence of magnetic field there should be no need for a 
>>spin-polarized calculation.
>>So, just to agree on the starting point of your question: what you mean with diamagnetic 
>>material is really what people usually refer to? Can you give us more details on the 
>>system you want to address? What properties do you want to study? ...
>>
>>Regards,
>>
>>
>>GS
>>

-- 


o ------------------------------------------------ o
| Gabriele Sclauzero, PhD Student                  |
| c/o:   SISSA & CNR-INFM Democritos,              |
|        via Beirut 2-4, 34014 Trieste (Italy)     |
| email: sclauzer at sissa.it                         |
| phone: +39 040 3787 511                          |
| skype: gurlonotturno                             |
o ------------------------------------------------ o