[CP2K-user] [CP2K:13141] spin of unpaired electrons

[pavan kumar behara]

Hi Lucas,

Yeah, I tried to build a minimal example and I realized that this system 
doesn't convey what I really wanted to ask, that's why I deleted the 
question after I posted it. I apologize for the inconvenience, thank you 
though for taking time to look into it.

I wanted to look at the c-direction electron transfer in spin bi-layer bulk 
Hematite. I could always generate one state where Fe(III) has all five 
d-electrons spin down and the excess electron pairs up with one d-electron 
making it Fe(II). Let's say this is my initial state. And all the Fe(III) 
in one layer above or below this layer in c-dir have 5 d-electrons spin-up. 
When I try to localize an excess electron on one of these iron atoms in the 
second layer to generate a final state for electron transfer pathway I 
always end up with the electron going to the other spin-layer. I tried to 
relax multiplicity and started with a good initial guess using broken 
symmetry section. But, I see that more MO_beta orbitals are not accepted in 
cp2k by design, and it seems to be the case in all chemistry codes I came 
across. And also I can't specify the multiplicity to be zero considering I 
have an excess beta electron [2*(-0.5)+1 = 0], zero is the default value 
and multiplicity is evaluated to 2 with an excess electron, again placing 
it in alpha channel. It seems I am missing something basic here since 
negative spin multiplicities are not accepted. As you explained I think 
both states give the same energy and there is no need to place the excess 
charge in beta channel. There might be some other way to do these 
calculations if one wants to study anisotropy in c-dir of Hematite.

Thanks again.

Best regards,
Pavan.

On Friday, April 17, 2020 at 9:23:27 PM UTC-4, Lucas Lodeiro wrote:
>
> Hi Pavan, 
>
> I am not an expert on CP2K code, but in a theoretical point of view what 
> you want to calculate is the same state, just in a different micro-state. 
> The case you show, or what ever case of duplet spin multiplicity has two 
> micro-states (in the spin freedom degree), and both have the same energy 
> (degenerated) and are described by the same spectroscopic term, but with 
> different z spin projection. In both cases the quantum state for all 
> properties is the same (except for Ŝ_z), then most of program, for 
> simplicity, uses the alpha channel to project the spin.
> For higher spin multiplicities, you have more micro-states for the same 
> spectroscopic term, but is complex in some cases. The triplet example, you 
> have 3 micro-states, S_z=+1,0,-1. The former and the latter are a single 
> determinant and the only difference is the channels of the unpaired 
> electrons (two), is a simple case as duplet multiplicity, but, the S_z=0 
> case is complex, due to this state is a linear combination of two 
> determinants, and needs multi reference description. Then the full spin 
> projection is a simpler case.
>
> Only if you use a magnetic field, the degeneration is broken, and both 
> cases you mention are differents.
>
> Regards - Lucas Lodeiro
>
>
> El jue., 16 abr. 2020 a las 17:38, pavan kumar behara (<pa... at gmail.com 
> <javascript:>>) escribió:
>
>> Hello CP2K developers,
>>
>> I notice that excess electrons are always assigned to alpha channel, is 
>> it possible to generate a configuration where I can have more beta 
>> electrons? Let's consider a minimal example of hydrogen molecule with 
>> excess charge, H-H(-), making it a 3-electron system. I can always 
>> calculate [2 spin-up, 1 spin-down] configuration
>>
>> ↑
>>
>> ↑*↓*
>>
>>
>> but never [1 spin-up, 2 spin-down]
>>
>> ↓
>>
>> ↓*↑*
>>
>>
>> I started with a spin-down hydrogen as initial atomic guess but the 
>> density matrix is re-scaled before starting the SCF iterations making it 2 
>> alpha and 1 beta. 
>>
>> Is there a way to get the second configuration or there is a reason not 
>> to do so? I attached the inputs and output for your reference.
>>
>> Any help is appreciated, thank you very much for your time.
>>
>>
>> Best regards,
>> Pavan.
>>
>> -- 
>> You received this message because you are subscribed to the Google Groups 
>> "cp2k" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to c... at googlegroups.com <javascript:>.
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/cp2k/631dd3cc-55b5-4d01-854b-c2255130c39b%40googlegroups.com 
>> <https://groups.google.com/d/msgid/cp2k/631dd3cc-55b5-4d01-854b-c2255130c39b%40googlegroups.com?utm_medium=email&utm_source=footer>
>> .
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200418/8269854f/attachment.htm>