Hirshfeld charge of MoS2 in cp2k

[Gabriele]

Hi Yun,

It might be because the Mo basis set is effectively larger than the S one. 
For instance, if you were to use a SZV-MOLOPT-SR for Mo and DZVP-MOLOPT-SR 
for S and look at the mulliken charges you would see that the signs are 
also correct, while they are wrong if you use DZVP for both elements. I am 
not suggesting to use a SZV for production though. The point on the 
mulliken charges is just to illustrate the effect of a different basis set 
size for the two elements.

Best
Gabriele

On Tuesday, November 21, 2017 at 3:22:02 PM UTC+1, any... at gmail.com wrote:
>
> Dear Gabriele,
>
> Thank you very much for the kind reply! I tried rerun the calculations 
> with your suggestions(I used the self-consistent hirshfeld and shape 
> function density, also the new TZVP basis for Mo ), now it gives the Net 
> charge for Mo 0.319 and -0.160 for S. Is this means the basis set on Mo 
> is stronger than S? I'll rerun several tests. Thank you again.
>
> Best,
> Yun
> 在 2017年11月20日星期一 UTC+1下午3:55:53，Gabriele写道：
>
>> Dear Yun,
>>
>> You can try to compute the charges using self-consistent hirshfeld 
>> partitioning (use SELF_CONSISTENT T inside the section &HIRSHFELD..&END) 
>> and using SHAPE_FUNCTION DENSITY. It has been highlighted before that there 
>> is in general a dependence of the charges on the basis set on MoS2 (see 
>> https://groups.google.com/forum/#!topic/cp2k/gZgiyqZURs4)
>>
>> For the DZVP-MOLOPT-SR basis, setting self consistent partitioning will 
>> give the right sign of the charges. We have also noticed that using this 
>> scheme there is not a significant dependence of the charges wrt to the size 
>> of the basis set. However, it is worth checking the dependence on the basis 
>> set size of the geometry and of other properties that may be relevant for 
>> your specific case.
>>
>> For that, you may want to check the file BASIS_MOLOPT_UCL, which is an 
>> extended molopt basis set library.
>>
>>
>> https://sourceforge.net/p/cp2k/code/HEAD/tree/trunk/cp2k/data/BASIS_MOLOPT_UCL
>>
>> Best
>> Gabriele
>>
>>
>> On Monday, November 20, 2017 at 2:26:15 PM UTC+1, any... at gmail.com 
>> wrote:
>>>
>>> Dear all,
>>>
>>> I run the cell_opt  for 4x4x1 MoS2 system and I want to analyze 
>>> the Hirshfeld charges, the result shows the Net charge of Mo and S are 
>>> -1.243 and 0.622, respectively.  But I thought the charge of Mo should be 
>>> positive while S should be negative.  
>>>
>>> Can anybody help me to see if there are any problems or I misunderstood 
>>> it?  Thank you very much in advance. 
>>>
>>> Kind regards,
>>> Yun
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20171121/60ca9156/attachment.htm>