[CP2K:8686] negative homo - lumo gap

Ian Hamilton ianpha... at gmail.com
Fri Feb 24 20:07:19 UTC 2017


Hi Marcella,

Thank you for your reply. Let me tell you a little bit about the system and 
why, although a negative HOMO-LUMO gap doesn't make sense, I'm no longer
worried about the situation.

The student is doing calculations on graphene quantum dots which (unlike 
graphene) are well-known to gave a gap. He is looking at the effect on the 
gap 
of replacing hydrogens with either electron donating or electron 
withdrawing groups. He is also looking at the effect on the gap of 
replacing hydrogens with
bulkier groups at either armchair or zigzag positions since it has been 
suggested that a gap can form in graphene if it is strained and that the 
effect is different
for strain in the armchair versus zigzag directions. 

The calculation for which he got the negative HOMO-LUMO gap, was one in 
which, due to too much strain or congestion, the "optimized" geometry was 
completely 
nonphysical. 

Best,
Ian

On Wednesday, February 22, 2017 at 4:12:39 AM UTC-5, Marcella Iannuzzi 
wrote:
>
> Hi Ian,
>
> without more information about the calculation you are performing it is 
> hard to guess what is going on.
> By default, the  unoccupied states are calculated only a posteriori. Here 
> the problem seems to be that the ground state is not properly optimised.
> Either the system has no gap, i.e., it is metallic, or there is some 
> problem with the DFT settings. 
> If the system is metallic, one should use a different scheme for the 
> optimisation that is able to apply the partial occupation of the KS states, 
> by assigning the occupation number as a function of the KS energy and of 
> the Fermi energy. 
> This is not possible using OT. From your message it is not clear what did 
> Claudio actually try to correct the problem.
> Best
> Marcella
>
>
> On 13 Feb 2017, at 17:58, Ian Hamilton <ianp... at gmail.com <javascript:>> 
> wrote:
>
> Hi Marcella,
>
> One of my students has encountered this problem. He tried changing the OT 
> and increasing the number of MOs (we don't know what you mean by "smearing 
> the occupation numbers around the Fermi energy") but it didn't correct the 
> problem. I’m posting this on his behalf because he was unable to- he got a 
> message that he had “limited access to public groups in his domain.”
>
> Do you (or anyone else) have another suggestion? In his pdos files all the 
> levels look fine except for the HOMO and LUMO.
>
> From his output:
>  Fermi Energy [eV] :   -5.966796
>   Lowest Eigenvalues of the unoccupied subspace spin            1
>  -----------------------------------------------------
>   Reached convergence in          149  iterations 
>       -0.23489989     -0.21167052     -0.18538586     -0.17943768
>       -0.13796253     -0.13007839
>  HOMO - LUMO gap [eV] :   -0.425155
>  ENERGY| Total FORCE_EVAL ( QS ) energy (a.u.):           
>  -1076.972424168399129
>
> Best regards,
> Ian Haminton
>
> On Sunday, November 22, 2009 at 6:13:06 AM UTC-5, Claudio wrote:
>>
>> Dear all, 
>>
>> sometimes after I converge a calculation to an accuracy of say 1.0E-5 
>> I get things like 
>>
>> HOMO - LUMO gap [eV] :   -0.980594 
>>
>> This is because the Homo Energy is 0.22417642 and the Lumo Energy is 
>> 0.18814027 in AU. 
>>
>> I have requested the same accuracy for homo and lumo  and have added 
>> enough iterations to the lumo calculation loop to make sure the 
>> program claims it is converged. I am using the OT method. How do I 
>> prevent this from happening? I mean how do I ensure that the 
>> calculation yields Lumo energies that are higher than the homo energy? 
>>
>>
>> Sorry to bother you and thanks for your help! 
>>
>> Claudio 
>>
>>
> -- 
> 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 cp2k+... at googlegroups.com <javascript:>.
> To post to this group, send email to cp... at googlegroups.com <javascript:>.
> Visit this group at https://groups.google.com/group/cp2k.
> For more options, visit https://groups.google.com/d/optout.
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20170224/72a455b3/attachment.htm>


More information about the CP2K-user mailing list