[CP2K:3175] Re: error in bond dissociation energy of gas phase molecule

[Fawzi Mohamed]

On 6-apr-11, at 15:27, Jun wrote:

> I just recalled that we also did the calculations using GAPW. The bond
> dissociation energy of H2O2 using BLYP increases to about 2.8 eV. This
> result places some doubt on the pseudopotential of O. I should admit
> that this GAPW number is subject to some uncertainty because we don't
> have much experience on GAPW. We just modified an input sample file
> from /cp2k/tests. Could any expert on GAPW  quickly do a testing
> calculation on this, please. The optimised geometries can be found in
> the input files attached above.
> Many thanks,
> Jun

That is exactly what I wanted to comment on:
1) GTH is a pseudopetential, and also has scalar relativistic  
correction (should be small in this case, but you never know)
2) basiset for GTH pseudopotential are different from those on all  
electrons, they should be more or less comparable, but there is no  
guarantee, what can be compared is the basset limit in both cases
3) a cutoff of 280 for OH is a bit on the low side, especially if you  
don't do any smoothing of the xc part, I would increase that

Fawzi
>
> On Apr 6, 12:40 pm, Jun <chen... at gmail.com> wrote:
>> Thank both of you very much !
>> Just quickly recalculated the H2O2 bond dissociation energy using MT.
>> The number hardly changes. I will try WAVELET, but don't think it  
>> will
>> make massive difference either.
>> Any other comments?
>> Jun
>>
>> On Apr 6, 12:12 pm, Teodoro Laino <teodor... at gmail.com> wrote:
>>
>>> ops.. sorry my mistake now..
>>> reported relative error is of ~ 1.0E-3 Hartree.. nonetheless..
>>
>>> On Apr 6, 2011, at 1:09 PM, Teodoro Laino wrote:
>>
>>>>> Sure -
>>>>> but for grid steps (fine density grid on which both decoupler  
>>>>> are applied) of 0.2-0.3 Angstrom we are comparing errors of  
>>>>> 1.0E-5 (MT) with 1.0E-7(WAVELET).
>>>>> Given the intrinsic errors reported by Jun (we are talking about  
>>>>> errors of 1.0E-2 Hartree here!) for me the above numbers are  
>>>>> comparable (I wish to see all the papers out there with an  
>>>>> overall accuracy of 1.0E-5 Hartree :) ).
>>
>>>>> ciao,
>>>>> Teo
>>
>>>>> On Apr 6, 2011, at 1:02 PM, hut... at pci.uzh.ch wrote:
>>
>>>>>> My mistake, I thought I had read OH-.
>>>>>> Regarding the relative accuracy of the decoupling methods, have a
>>>>>> look in the paper
>>>>>> DOI=10.1063/1.2335442
>>>>>> where the two methods are compared.
>>
>>>>>> regards
>>
>>>>>> Juerg
>>
>>>>>> --------------------------------------------------------------
>>>>>> Juerg Hutter                         Phone : ++41 44 635 4491
>>>>>> Physical Chemistry Institute   FAX   : ++41 44 635 6838
>>>>>> University of Zurich               E-mail:  hut... at pci.uzh.ch
>>>>>> Winterthurerstrasse 190
>>>>>> CH-8057 Zurich, Switzerland
>>>>>> ---------------------------------------------------------------
>>
>>>>>> -----cp... at googlegroups.com wrote: -----
>>>>>> To: cp... at googlegroups.com
>>>>>> From: Teodoro Laino
>>>>>> Sent by: cp... at googlegroups.com
>>>>>> Date: 04/06/2011 12:58PM
>>>>>> Subject: Re: [CP2K:3169] error in bond dissociation energy of  
>>>>>> gas phase molecule
>>
>>>>>> Hi Juerg,
>>
>>>>>> the species are not charged - these are all homolytic  
>>>>>> dissociations.
>>>>>> Nonetheless the H2O2 may have quite a strong dipole and the  
>>>>>> decoupling may affect the energetics.
>>
>>>>>> Just a comment: WAVELET is not superior to MT. The main  
>>>>>> difference is the cell size. For such small molecules MT may  
>>>>>> even be better because you can use a smaller cell or at least  
>>>>>> comparable to the one of WAVELET :).
>>
>>>>>> Teo
>>
>>>>>> On Apr 6, 2011, at 12:53 PM, hut... at pci.uzh.ch wrote:
>>
>>>>>>> Hi
>>
>>>>>>> as you are generating charged species you should use a non- 
>>>>>>> periodic
>>>>>>> steup.
>>
>>>>>>> Try PERIODIC NONE in the &CELL section together with
>>
>>>>>>>  &POISSON
>>>>>>>    PERIODIC NONE
>>>>>>>    PSOLVER WAVELET    <- MT and ANALYTIC could also work but  
>>>>>>> this is best
>>>>>>>  &END
>>
>>>>>>> in the DFT section.
>>
>>>>>>> regards
>>
>>>>>>> Juerg
>>
>>>>>>> --------------------------------------------------------------
>>>>>>> Juerg Hutter                         Phone : ++41 44 635 4491
>>>>>>> Physical Chemistry Institute   FAX   : ++41 44 635 6838
>>>>>>> University of Zurich               E-mail:  hut... at pci.uzh.ch
>>>>>>> Winterthurerstrasse 190
>>>>>>> CH-8057 Zurich, Switzerland
>>>>>>> ---------------------------------------------------------------
>>
>>>>>>> -----cp... at googlegroups.com wrote: -----
>>>>>>> To: cp2k <cp... at googlegroups.com>
>>>>>>> From: Jun
>>>>>>> Sent by: cp... at googlegroups.com
>>>>>>> Date: 04/06/2011 12:43PM
>>>>>>> Subject: [CP2K:3165] error in bond dissociation energy of gas  
>>>>>>> phase molecule
>>
>>>>>>> Hello everyone,
>>
>>>>>>> We did some testing calculations on bond dissociation energies  
>>>>>>> of HO-
>>>>>>> OH in vacuum using BLYP and B3LYP. We found a large difference  
>>>>>>> from
>>>>>>> reported results using the same functionals in literature; our  
>>>>>>> numbers
>>>>>>> are about 0.5 eV too small for both functionals. We also cross- 
>>>>>>> checked
>>>>>>> the numbers using gaussian program, indeed cp2k gives smaller  
>>>>>>> bond
>>>>>>> dissociation energy. It is not likely to be BSSE, as we use  
>>>>>>> quite
>>>>>>> large basis set TZV2P and also checked QZV2P. It is obvious  
>>>>>>> that we
>>>>>>> got either the total energy of H2O2 too high or the OH energy  
>>>>>>> too low.
>>>>>>> We checked the energy of H2O --> OH + H, which is very close  
>>>>>>> to what
>>>>>>> we found in literature. So, the problem appears to be the too  
>>>>>>> high
>>>>>>> energy of H2O2.
>>>>>>> Also, we checked CH3-CH3, and found the same: CH3-CH3 bond
>>>>>>> dissociation energy is 0.3 eV too small for both functionals  
>>>>>>> compared
>>>>>>> to literature numbers, while the energy of CH4-->CH3+H agrees  
>>>>>>> well.
>>>>>>> The bond dissociation energies (total energies without ZPE) are
>>>>>>> summarised below, and inputs for H2O2-->2OH (H2O2.inp and  
>>>>>>> OH.inp) are
>>>>>>> uploaded.
>>>>>>>                               BLYP             B3LYP
>>>>>>> H2O2-->2OH       2.04 eV (2.67)      1.80 eV  (2.27)
>>>>>>> H2O-->OH+H        5.10 eV              5.16 eV (5.2)
>>>>>>> C2H6-->2CH3     3.64 eV (3.92)       3.69 eV (3.97)
>>>>>>> CH4-->CH3+H    4.69 eV (4.77)       4.75 eV (4.67)
>>>>>>> Numbers in parentheses are from literature. The geometry is  
>>>>>>> optimised
>>>>>>> one taken from literature, which is almost the same as  
>>>>>>> optimised by
>>>>>>> cp2k.
>>>>>>> Can anyone comment on this? Many thanks in advance.
>>
>>>>>>> Cheers,
>>>>>>> Jun
>>
>>>>>>> --
>>>>>>> You received this message because you are subscribed to the  
>>>>>>> Google Groups "cp2k" group.
>>>>>>> To post to this group, send email to cp... at googlegroups.com.
>>>>>>> To unsubscribe from this group, send email to cp2k+uns... at googlegroups.com 
>>>>>>> .
>>>>>>> For more options, visit this group athttp://groups.google.com/group/cp2k?hl=en 
>>>>>>> .
>>
>>>>>>> --
>>>>>>> You received this message because you are subscribed to the  
>>>>>>> Google Groups "cp2k" group.
>>>>>>> To post to this group, send email to cp... at googlegroups.com.
>>>>>>> To unsubscribe from this group, send email to cp2k+uns... at googlegroups.com 
>>>>>>> .
>>>>>>> For more options, visit this group athttp://groups.google.com/group/cp2k?hl=en 
>>>>>>> .
>>
>>>>>> --
>>>>>> You received this message because you are subscribed to the  
>>>>>> Google Groups "cp2k" group.
>>>>>> To post to this group, send email to cp... at googlegroups.com.
>>>>>> To unsubscribe from this group, send email to cp2k+uns... at googlegroups.com 
>>>>>> .
>>>>>> For more options, visit this group athttp://groups.google.com/group/cp2k?hl=en 
>>>>>> .
>>
>>>>>> --
>>>>>> You received this message because you are subscribed to the  
>>>>>> Google Groups "cp2k" group.
>>>>>> To post to this group, send email to cp... at googlegroups.com.
>>>>>> To unsubscribe from this group, send email to cp2k+uns... at googlegroups.com 
>>>>>> .
>>>>>> For more options, visit this group athttp://groups.google.com/group/cp2k?hl=en 
>>>>>> .
>
> -- 
> You received this message because you are subscribed to the Google  
> Groups "cp2k" group.
> To post to this group, send email to cp... at googlegroups.com.
> To unsubscribe from this group, send email to cp2k+uns... at googlegroups.com 
> .
> For more options, visit this group at http://groups.google.com/group/cp2k?hl=en 
> .
>