[CP2K-user] Annihilating spin contamination in CP2K

[Stephen Vicchio]

Hello All, 

I hope all is well and everyone is staying safe during these interesting 
times! 

I have a question related to spin contamination in CP2K. Specifically, how 
can account for that spin contamination and annihilate it (as can be down 
in a software package such as Gaussian)? 

The system I am studying is a Ni(II) cation complex supported in a 
metal-organic framework (MOF). When looking at the system in the high spin 
state configuration (every Ni atom contains two unpaired electrons | 
MULTIPLICITY = 9 ), I'm seeing the following result: 

  Ideal and single determinant S**2 :                   20.000000      
20.012256

However, when I switch to the low spin state configuration (no unpaired 
electrons | MULTIPLICITY = 1), I'm seeing the following result: 

  Ideal and single determinant S**2 :                    0.000000      
 3.078170

Comparing the ideal and single determinant values of S**2 are a good 
indicator of spin contamination in the system. I see as I lower the 
MULTIPLICITY value, I am seeing more spin contamination. Does anyone have 
any tips for annihilating the spin contamination of my system?

For reference, I have attached sample input and output files for one of the 
calculations showing my problem. At present, I am not using any smearing in 
my calculations. All comments, questions, and concerns are welcome! 

With much appreciation, 

Stephen   

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200407/9e03c07b/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: singlet-ni4-bare.inp
Type: chemical/x-gamess-input
Size: 56538 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200407/9e03c07b/attachment.inp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: singlet-ni4-bare.zip
Type: application/zip
Size: 2937906 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200407/9e03c07b/attachment.zip>