[CP2K:8462] Need advice for setting CP2K input for all electron CAM-B3LYP calculations

[Matt W]

This paper cites a couple of sources saying the C60 HOMO-LUMO gap as 2.77 
at B3LYP/6-31G**

https://journals.aps.org/prb/abstract/10.1103/PhysRevB.83.115103

On Sunday, June 17, 2018 at 10:58:00 PM UTC+1, Samir ABDELOUAHED wrote:
>
>
> Dear Juerg,
> I am facing a serious problem getting the right HOMO-LUMO gap for an 
> isolated C60. I've a pending draft on C60 adsorption and I don't have any 
> more the ADF license. So I thought cp2k might be the  best alternative. So 
> I started calculating the C60 gap. The problem is that I am still far from 
> the C60 experimental gap of around 5 eV.
> the PBE, B3LYP , and PBE0 (using either GTH or MOLOPT DZVP basis set) 
> calculations gave 1.62 eV, 2.64, and 2.90 eV, respectively.
> As the ADF B3LYP All electron-STO basis set calculations gave around 4.7 
> eV I thought that a cp2k B3LYP all electron-GTO basis set would give a 
> similar value...But apparently it is not the case. Indeed, I've got similar 
> values to the pseudopotential calculations: 1.65, 2.69, and 2.95 eV.
> So I am wondering whether this is due to the fact that the ADF 
> calculations were done using STO, which are more efficient than GTO. If it 
> is the case which other GTO basis should I take to get the right gap. Or is 
> there some problems in my settings (inputs).
> Please find attached the inputs for these calculations (6 inputs).
> I would like to tell you that the calculations are converged without any 
> problem even the hybrid ones as I started these calculations from a PBE 
> pre-converged calculation.  
> Many many thanks for your help,
> Samir
> Le mardi 6 décembre 2016 14:05:45 UTC+1, jgh a écrit :
>>
>> Hi 
>>
>> your definition of the HF section is not correct: 
>>
>>       &HF 
>>        &SCREENING 
>>          EPS_SCHWARZ 1.0E-12 
>>        &END 
>>        &MEMORY 
>>          MAX_MEMORY 100 
>>        &END 
>>        &INTERACTION_POTENTIAL 
>>          POTENTIAL_TYPE MIX_CL 
>>          OMEGA 0.33 
>>          SCALE_LONGRANGE 0.94979 
>>          SCALE_COULOMB 0.18352 
>>        &END 
>>       &END 
>>
>> If you want to reproduce the G09 results, I would also take the 
>> WAVELET solver (not MT), remove the XC_GRID section (no smoothing) 
>> and maybe increase the cutoff. 
>>
>> You can find some examples in tests/QS/regtest-hybrid-3 and 4. 
>>
>> If you want to optimize speed I would use GPW and pseudopotentials 
>> together with the ADMM method. 
>>
>> regards 
>>
>> Juerg 
>> -------------------------------------------------------------- 
>> Juerg Hutter                         Phone : ++41 44 635 4491 
>> Institut für Chemie C                FAX   : ++41 44 635 6838 
>> Universität Zürich                   E-mail: hut... at chem.uzh.ch 
>> Winterthurerstrasse 190 
>> CH-8057 Zürich, Switzerland 
>> --------------------------------------------------------------- 
>>
>> -----cp... at googlegroups.com wrote: -----To: cp2k <cp... at googlegroups.com> 
>>
>> From: Kun-Han Lin 
>> Sent by: cp... at googlegroups.com 
>> Date: 12/06/2016 01:10PM 
>> Subject: [CP2K:8462] Need advice for setting CP2K input for all electron 
>> CAM-B3LYP calculations 
>>
>> Dear CP2K users, 
>>
>> Recently, I want to use CP2K to do umbrella 
>> sampling with the QM/MM method. The system is composed of  a Cl anion 
>> and neutral molecule (closed shell system).  
>>
>>
>>
>>
>> I 
>>  first compared the static energy-distance curve calculated by CP2K and 
>> Gaussian under the same QM condition (CAM-B3LYP, 6-311G**, 
>> I also tried adding diffuse function in Gaussian, the results are 
>> similar to that without diffuse function). However, the results from 
>> CP2K give incorrect long-range behavior and the SCF also does not 
>> converge (especially in when the monomer are largely separated, 
>> I put in CP2K result.pdf file). Therefore, I'd like to know if there is 
>> some error in my input file. I attach it in this file as well 
>> (CAM-B3LYP.inp). Note that we do need to use a range-separated 
>> functional for describing this system properly. 
>>
>>
>>
>>
>>
>>
>>
>> Finally, 
>>  I'd like to know if there is some setting to make this long-range 
>> exchange computations accurate and efficient. Since 
>> we want to do sampling, efficiency is a critical issue. If you have 
>> advice or any suggestion, please let me know. Many thanks in advance. 
>>
>> Best, 
>> Kun-Han Lin 
>>
>>
>>
>>
>>
>>
>> PhD student LCMD 
>>
>> Ecole Polytechnique Fédérale de Lausanne (EPFL) 
>>
>>         EPFL SB ISIC LCMD 
>>
>>         BCH 5312 (Bât. BCH) 
>>
>>         CH-1015 Lausanne 
>>
>>         Switzerland 
>>
>>
>> E-mail: kun-... at epfl.ch 
>>
>>
>>
>> -- 
>>
>> 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. 
>>
>> To post to this group, send email to cp... at googlegroups.com. 
>>
>> Visit this group at https://groups.google.com/group/cp2k. 
>>
>> For more options, visit https://groups.google.com/d/optout. 
>>
>>
>>
>> [attachment "CP2K result.pdf" removed by Jürg Hutter/at/UZH] 
>> [attachment "CAM-B3LYP.inp" removed by Jürg Hutter/at/UZH] 
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20180617/4ea62fb8/attachment.htm>