[CP2K-user] [CP2K:13797] SMEAR vs. LSD+RELAX_MULTIPLICITY/MULTIPLICITY

Ivan Gladich igla... at gmail.com
Wed Aug 19 20:29:25 UTC 2020


Dear Marcella,

This answers my question.
Thank you very much
Grazie
Best regards

Ivan


On Wednesday, August 19, 2020 at 9:15:09 AM UTC+3, Marcella Iannuzzi wrote:
>
> Dear Ivan,
>
> The projected density of states, calculated individually for the alpha and 
> beta channels, give the information on how the electrons are distributed. 
> Also the population analysis can provide information, or at least trends, 
> on the fluctuations on the atomic charges and spin.
>
> Kind regards
> Marcella
>
>
> On Tuesday, August 18, 2020 at 5:52:58 PM UTC+2 ig... at gmail.com wrote:
>
>> Dear Matthias and Marcella
>>
>> thank you very much for your reply. 
>> Now is much more clear.
>>
>> However, I still have one doubt.
>>
>> While in LSD we know the spin state of the system (i.e., S**2 and 
>> integrated spin mentioned above) at any step,
>> it seems to me that in SMEAR we lost such information: as Matthias said, 
>> .S**2 and integrated spin cannot be calculated for fractional occupation in 
>> SMEAR.
>>
>> Using SMEAR, is it possible to define some average quantities defining 
>> the spin state of the system?
>> I do not know, for example,  the average occupation of the alpha and beta 
>> channels?
>> Or something else?
>>
>> Thank you very very much for your replies and help
>>
>> Best
>> Ivan 
>>
>>
>>
>> On Tuesday, August 18, 2020 at 5:54:48 PM UTC+3, Marcella Iannuzzi wrote:
>>>
>>> Dear Ivan, 
>>>
>>> in addition to what Matthias already wrote, by default the multiplicity 
>>> is not fixed with SMEAR + LSD. 
>>>
>>> Working with fractional orbital occupations requires the use of 
>>> diagonalisation schemes, which are in general  more demanding in terms of 
>>> computational costs if compared with the orbital transformation (OT) 
>>> method. 
>>>
>>> Kind regards
>>> Marcella
>>>
>>>
>>> On Tuesday, August 18, 2020 at 4:42:07 PM UTC+2 Matthias Krack wrote:
>>>
>>>> Dear Ivan
>>>>
>>>>  
>>>>
>>>> a)       SMEAR allows for fractional orbital occupations whereas LSD 
>>>> with RELAX_MULTIPLICITY is meant for systems with integer occupations, e.g. 
>>>> to allow for the spin flip needed to switch automatically from a singlet to 
>>>> a possibly lower lying triplet state. Thus, these schemes are not 
>>>> equivalent.
>>>>
>>>> b)      The expectation value of S**2 is only defined for calculations 
>>>> with integer orbital occupations Therefore this properties is not 
>>>> calculated/printed when SMEAR is active.
>>>>
>>>>  
>>>>
>>>> HTH
>>>>
>>>>  
>>>>
>>>> Matthias
>>>>
>>>>  
>>>>
>>>> *From:* c... at googlegroups.com <c... at googlegroups.com> *On Behalf Of *Ivan 
>>>> Gladich
>>>> *Sent:* Dienstag, 18. August 2020 15:56
>>>> *To:* cp2k <c... at googlegroups.com>
>>>> *Subject:* [CP2K:13797] SMEAR vs. LSD+RELAX_MULTIPLICITY/MULTIPLICITY
>>>>
>>>>  
>>>>
>>>> Dear CP2K users,
>>>>
>>>>  
>>>>
>>>> I would like some clarifications about SMEAR vs. LSD +RELAX_MULTIPLICTY 
>>>> in cp2k. I am trying to simulate a metallic nanoparticle (few tens of 
>>>> atoms) in water
>>>>
>>>>  
>>>>
>>>> I have two (maybe simple) questions:
>>>>
>>>>  
>>>>
>>>> a) From this post 
>>>>
>>>> https://groups.google.com/forum/#!topic/cp2k/GEKLwRQat_M
>>>>
>>>>  
>>>>
>>>> my understanding is that the SMEAR option populates the alpha and beta 
>>>> channel according to a Fermi-Dirac 
>>>>
>>>>  
>>>>
>>>> In LSD + MULTIPLICITY (or LSD + RELAX MULTIPLICITY)
>>>>
>>>>
>>>> https://groups.google.com/forum/#!searchin/cp2k/smearing$20spin%7Csort:date/cp2k/ugkJLz4PSSI/C-ProBR6BwAJ
>>>>
>>>>  
>>>>
>>>> my understating is that after energy levels are calculated, electrons 
>>>> are placed according to Aufbau principle, from the lowest to the highest in 
>>>> an unrestricted scheme. If MULTIPLICITY is set, the multiplicity is fixed, 
>>>> with RELAX_MULTIPLICITY no
>>>>
>>>>  
>>>>
>>>> Please correct me if my understanding is mistaken.
>>>>
>>>>  
>>>>
>>>> Question: are the two approaches (SMEAR vs. LSD +RELAX_MULTIPLICTY) 
>>>> equivalent? One approach is more appropriate than the other for particular 
>>>> systems? Is the computational cost the same?
>>>>
>>>>  
>>>>
>>>> b) If I use LSD +MULTIPLICTY in the output I get the
>>>>
>>>>  
>>>>
>>>> “Integrated absolute spin density” and  “Ideal and single determinant 
>>>> S**2”
>>>>
>>>>  
>>>>
>>>> That are equations (2.271) and (2.272) in “Modern Quantum Chemistry: 
>>>> Introduction to Advanced Electronic Structure Theory,Attila Bangha 
>>>> Szabo, Neil S. Ostlund “
>>>>
>>>>  
>>>>
>>>> https://groups.google.com/forum/#!topic/cp2k/flEEaseAd7c
>>>>
>>>>  
>>>>
>>>>  
>>>>
>>>> If I see the output of my calculation with SMEAR I do not see such or 
>>>> any other information about the spin state of my system. 
>>>>
>>>>  
>>>>
>>>> Question: Is this expected? How I can easily get information about the 
>>>> spin state of my system when using SMEAR?
>>>>
>>>>  
>>>>
>>>> Thank you very much for any possible help
>>>>
>>>>  
>>>>
>>>> Best regards
>>>>
>>>> Ivan
>>>>
>>>> -- 
>>>> 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 cp... at googlegroups.com.
>>>> To view this discussion on the web visit 
>>>> https://groups.google.com/d/msgid/cp2k/80e157d7-5559-4066-bbf5-bca66ce52addo%40googlegroups.com 
>>>> <https://groups.google.com/d/msgid/cp2k/80e157d7-5559-4066-bbf5-bca66ce52addo%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>> .
>>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200819/9791f478/attachment.htm>


More information about the CP2K-user mailing list